diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,115533 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.2159252590139156, + "eval_steps": 500, + "global_step": 16500, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 3.084646557341652e-05, + "grad_norm": 81.36478424072266, + "learning_rate": 1.5422578655151142e-08, + "loss": 10.3893, + "step": 1 + }, + { + "epoch": 6.169293114683304e-05, + "grad_norm": 82.38279724121094, + "learning_rate": 3.0845157310302284e-08, + "loss": 10.403, + "step": 2 + }, + { + "epoch": 9.253939672024955e-05, + "grad_norm": 82.81998443603516, + "learning_rate": 4.626773596545343e-08, + "loss": 10.4267, + "step": 3 + }, + { + "epoch": 0.00012338586229366608, + "grad_norm": 83.8257064819336, + "learning_rate": 6.169031462060457e-08, + "loss": 10.4424, + "step": 4 + }, + { + "epoch": 0.00015423232786708257, + "grad_norm": 83.32088470458984, + "learning_rate": 7.711289327575571e-08, + "loss": 10.3924, + "step": 5 + }, + { + "epoch": 0.0001850787934404991, + "grad_norm": 82.46728515625, + "learning_rate": 9.253547193090685e-08, + "loss": 10.3717, + "step": 6 + }, + { + "epoch": 0.00021592525901391563, + "grad_norm": 82.39380645751953, + "learning_rate": 1.07958050586058e-07, + "loss": 10.3754, + "step": 7 + }, + { + "epoch": 0.00024677172458733215, + "grad_norm": 81.09526062011719, + "learning_rate": 1.2338062924120914e-07, + "loss": 10.3638, + "step": 8 + }, + { + "epoch": 0.0002776181901607486, + "grad_norm": 81.36984252929688, + "learning_rate": 1.388032078963603e-07, + "loss": 10.3897, + "step": 9 + }, + { + "epoch": 0.00030846465573416515, + "grad_norm": 81.59973907470703, + "learning_rate": 1.5422578655151142e-07, + "loss": 10.3627, + "step": 10 + }, + { + "epoch": 0.0003393111213075817, + "grad_norm": 81.62650299072266, + "learning_rate": 1.6964836520666255e-07, + "loss": 10.3901, + "step": 11 + }, + { + "epoch": 0.0003701575868809982, + "grad_norm": 81.47443389892578, + "learning_rate": 1.850709438618137e-07, + "loss": 10.3711, + "step": 12 + }, + { + "epoch": 0.0004010040524544147, + "grad_norm": 80.487060546875, + "learning_rate": 2.0049352251696484e-07, + "loss": 10.3361, + "step": 13 + }, + { + "epoch": 0.00043185051802783125, + "grad_norm": 81.62635803222656, + "learning_rate": 2.15916101172116e-07, + "loss": 10.3585, + "step": 14 + }, + { + "epoch": 0.0004626969836012477, + "grad_norm": 79.68704223632812, + "learning_rate": 2.3133867982726715e-07, + "loss": 10.3338, + "step": 15 + }, + { + "epoch": 0.0004935434491746643, + "grad_norm": 82.13473510742188, + "learning_rate": 2.467612584824183e-07, + "loss": 10.3685, + "step": 16 + }, + { + "epoch": 0.0005243899147480808, + "grad_norm": 79.47209167480469, + "learning_rate": 2.621838371375694e-07, + "loss": 10.3407, + "step": 17 + }, + { + "epoch": 0.0005552363803214972, + "grad_norm": 78.69096374511719, + "learning_rate": 2.776064157927206e-07, + "loss": 10.2991, + "step": 18 + }, + { + "epoch": 0.0005860828458949138, + "grad_norm": 78.3796157836914, + "learning_rate": 2.930289944478717e-07, + "loss": 10.2931, + "step": 19 + }, + { + "epoch": 0.0006169293114683303, + "grad_norm": 76.12564086914062, + "learning_rate": 3.0845157310302284e-07, + "loss": 10.2695, + "step": 20 + }, + { + "epoch": 0.0006477757770417469, + "grad_norm": 77.07991027832031, + "learning_rate": 3.23874151758174e-07, + "loss": 10.238, + "step": 21 + }, + { + "epoch": 0.0006786222426151633, + "grad_norm": 75.79402160644531, + "learning_rate": 3.392967304133251e-07, + "loss": 10.2583, + "step": 22 + }, + { + "epoch": 0.0007094687081885798, + "grad_norm": 75.79705047607422, + "learning_rate": 3.5471930906847623e-07, + "loss": 10.2446, + "step": 23 + }, + { + "epoch": 0.0007403151737619964, + "grad_norm": 68.02273559570312, + "learning_rate": 3.701418877236274e-07, + "loss": 10.0908, + "step": 24 + }, + { + "epoch": 0.0007711616393354129, + "grad_norm": 64.18199157714844, + "learning_rate": 3.8556446637877854e-07, + "loss": 10.0332, + "step": 25 + }, + { + "epoch": 0.0008020081049088294, + "grad_norm": 63.33767318725586, + "learning_rate": 4.0098704503392967e-07, + "loss": 10.0079, + "step": 26 + }, + { + "epoch": 0.0008328545704822459, + "grad_norm": 63.122013092041016, + "learning_rate": 4.1640962368908085e-07, + "loss": 10.0156, + "step": 27 + }, + { + "epoch": 0.0008637010360556625, + "grad_norm": 62.6221923828125, + "learning_rate": 4.31832202344232e-07, + "loss": 10.0014, + "step": 28 + }, + { + "epoch": 0.000894547501629079, + "grad_norm": 59.73037338256836, + "learning_rate": 4.472547809993831e-07, + "loss": 9.951, + "step": 29 + }, + { + "epoch": 0.0009253939672024954, + "grad_norm": 59.53206253051758, + "learning_rate": 4.626773596545343e-07, + "loss": 9.9388, + "step": 30 + }, + { + "epoch": 0.000956240432775912, + "grad_norm": 59.05270767211914, + "learning_rate": 4.780999383096854e-07, + "loss": 9.9583, + "step": 31 + }, + { + "epoch": 0.0009870868983493286, + "grad_norm": 58.37528610229492, + "learning_rate": 4.935225169648366e-07, + "loss": 9.9287, + "step": 32 + }, + { + "epoch": 0.001017933363922745, + "grad_norm": 41.95176315307617, + "learning_rate": 5.089450956199877e-07, + "loss": 9.6617, + "step": 33 + }, + { + "epoch": 0.0010487798294961615, + "grad_norm": 28.621204376220703, + "learning_rate": 5.243676742751388e-07, + "loss": 9.527, + "step": 34 + }, + { + "epoch": 0.0010796262950695781, + "grad_norm": 27.27179718017578, + "learning_rate": 5.3979025293029e-07, + "loss": 9.5315, + "step": 35 + }, + { + "epoch": 0.0011104727606429945, + "grad_norm": 26.737306594848633, + "learning_rate": 5.552128315854412e-07, + "loss": 9.5164, + "step": 36 + }, + { + "epoch": 0.001141319226216411, + "grad_norm": 25.745166778564453, + "learning_rate": 5.706354102405923e-07, + "loss": 9.4997, + "step": 37 + }, + { + "epoch": 0.0011721656917898276, + "grad_norm": 24.734355926513672, + "learning_rate": 5.860579888957434e-07, + "loss": 9.4851, + "step": 38 + }, + { + "epoch": 0.0012030121573632442, + "grad_norm": 23.643321990966797, + "learning_rate": 6.014805675508946e-07, + "loss": 9.4658, + "step": 39 + }, + { + "epoch": 0.0012338586229366606, + "grad_norm": 22.227819442749023, + "learning_rate": 6.169031462060457e-07, + "loss": 9.4352, + "step": 40 + }, + { + "epoch": 0.0012647050885100772, + "grad_norm": 21.384784698486328, + "learning_rate": 6.323257248611968e-07, + "loss": 9.4199, + "step": 41 + }, + { + "epoch": 0.0012955515540834938, + "grad_norm": 19.936220169067383, + "learning_rate": 6.47748303516348e-07, + "loss": 9.3921, + "step": 42 + }, + { + "epoch": 0.0013263980196569101, + "grad_norm": 16.649097442626953, + "learning_rate": 6.631708821714991e-07, + "loss": 9.2965, + "step": 43 + }, + { + "epoch": 0.0013572444852303267, + "grad_norm": 13.570879936218262, + "learning_rate": 6.785934608266502e-07, + "loss": 9.214, + "step": 44 + }, + { + "epoch": 0.0013880909508037433, + "grad_norm": 13.162702560424805, + "learning_rate": 6.940160394818013e-07, + "loss": 9.1518, + "step": 45 + }, + { + "epoch": 0.0014189374163771596, + "grad_norm": 13.2677640914917, + "learning_rate": 7.094386181369525e-07, + "loss": 9.1449, + "step": 46 + }, + { + "epoch": 0.0014497838819505762, + "grad_norm": 12.907882690429688, + "learning_rate": 7.248611967921037e-07, + "loss": 9.1151, + "step": 47 + }, + { + "epoch": 0.0014806303475239928, + "grad_norm": 12.483068466186523, + "learning_rate": 7.402837754472548e-07, + "loss": 9.07, + "step": 48 + }, + { + "epoch": 0.0015114768130974094, + "grad_norm": 11.745848655700684, + "learning_rate": 7.55706354102406e-07, + "loss": 9.0454, + "step": 49 + }, + { + "epoch": 0.0015423232786708257, + "grad_norm": 11.16760540008545, + "learning_rate": 7.711289327575571e-07, + "loss": 9.0275, + "step": 50 + }, + { + "epoch": 0.0015731697442442423, + "grad_norm": 10.60039234161377, + "learning_rate": 7.865515114127083e-07, + "loss": 9.0316, + "step": 51 + }, + { + "epoch": 0.001604016209817659, + "grad_norm": 10.370899200439453, + "learning_rate": 8.019740900678593e-07, + "loss": 8.9997, + "step": 52 + }, + { + "epoch": 0.0016348626753910753, + "grad_norm": 10.163037300109863, + "learning_rate": 8.173966687230106e-07, + "loss": 8.9961, + "step": 53 + }, + { + "epoch": 0.0016657091409644918, + "grad_norm": 10.336933135986328, + "learning_rate": 8.328192473781617e-07, + "loss": 8.945, + "step": 54 + }, + { + "epoch": 0.0016965556065379084, + "grad_norm": 10.267864227294922, + "learning_rate": 8.482418260333127e-07, + "loss": 8.9297, + "step": 55 + }, + { + "epoch": 0.001727402072111325, + "grad_norm": 10.498688697814941, + "learning_rate": 8.63664404688464e-07, + "loss": 8.9201, + "step": 56 + }, + { + "epoch": 0.0017582485376847414, + "grad_norm": 10.419842720031738, + "learning_rate": 8.79086983343615e-07, + "loss": 8.8583, + "step": 57 + }, + { + "epoch": 0.001789095003258158, + "grad_norm": 10.13261890411377, + "learning_rate": 8.945095619987662e-07, + "loss": 8.7875, + "step": 58 + }, + { + "epoch": 0.0018199414688315745, + "grad_norm": 10.167094230651855, + "learning_rate": 9.099321406539174e-07, + "loss": 8.7495, + "step": 59 + }, + { + "epoch": 0.0018507879344049909, + "grad_norm": 10.329927444458008, + "learning_rate": 9.253547193090686e-07, + "loss": 8.705, + "step": 60 + }, + { + "epoch": 0.0018816343999784075, + "grad_norm": 10.758321762084961, + "learning_rate": 9.407772979642196e-07, + "loss": 8.6599, + "step": 61 + }, + { + "epoch": 0.001912480865551824, + "grad_norm": 10.68518352508545, + "learning_rate": 9.561998766193708e-07, + "loss": 8.6209, + "step": 62 + }, + { + "epoch": 0.0019433273311252404, + "grad_norm": 10.591952323913574, + "learning_rate": 9.716224552745219e-07, + "loss": 8.5942, + "step": 63 + }, + { + "epoch": 0.001974173796698657, + "grad_norm": 10.513204574584961, + "learning_rate": 9.87045033929673e-07, + "loss": 8.5421, + "step": 64 + }, + { + "epoch": 0.0020050202622720736, + "grad_norm": 10.280865669250488, + "learning_rate": 1.0024676125848241e-06, + "loss": 8.4896, + "step": 65 + }, + { + "epoch": 0.00203586672784549, + "grad_norm": 10.49332332611084, + "learning_rate": 1.0178901912399754e-06, + "loss": 8.4551, + "step": 66 + }, + { + "epoch": 0.0020667131934189067, + "grad_norm": 10.475257873535156, + "learning_rate": 1.0333127698951264e-06, + "loss": 8.4182, + "step": 67 + }, + { + "epoch": 0.002097559658992323, + "grad_norm": 10.291110038757324, + "learning_rate": 1.0487353485502776e-06, + "loss": 8.3717, + "step": 68 + }, + { + "epoch": 0.0021284061245657395, + "grad_norm": 10.081697463989258, + "learning_rate": 1.0641579272054289e-06, + "loss": 8.323, + "step": 69 + }, + { + "epoch": 0.0021592525901391563, + "grad_norm": 9.863335609436035, + "learning_rate": 1.07958050586058e-06, + "loss": 8.2487, + "step": 70 + }, + { + "epoch": 0.0021900990557125726, + "grad_norm": 9.875226020812988, + "learning_rate": 1.0950030845157311e-06, + "loss": 8.2423, + "step": 71 + }, + { + "epoch": 0.002220945521285989, + "grad_norm": 9.859506607055664, + "learning_rate": 1.1104256631708823e-06, + "loss": 8.1533, + "step": 72 + }, + { + "epoch": 0.0022517919868594058, + "grad_norm": 10.05392074584961, + "learning_rate": 1.1258482418260334e-06, + "loss": 8.113, + "step": 73 + }, + { + "epoch": 0.002282638452432822, + "grad_norm": 10.032729148864746, + "learning_rate": 1.1412708204811846e-06, + "loss": 8.0704, + "step": 74 + }, + { + "epoch": 0.0023134849180062385, + "grad_norm": 10.190704345703125, + "learning_rate": 1.1566933991363356e-06, + "loss": 7.9496, + "step": 75 + }, + { + "epoch": 0.0023443313835796553, + "grad_norm": 9.874703407287598, + "learning_rate": 1.1721159777914869e-06, + "loss": 7.9291, + "step": 76 + }, + { + "epoch": 0.0023751778491530717, + "grad_norm": 9.836445808410645, + "learning_rate": 1.1875385564466379e-06, + "loss": 7.844, + "step": 77 + }, + { + "epoch": 0.0024060243147264885, + "grad_norm": 9.5654878616333, + "learning_rate": 1.2029611351017891e-06, + "loss": 7.7807, + "step": 78 + }, + { + "epoch": 0.002436870780299905, + "grad_norm": 9.756566047668457, + "learning_rate": 1.2183837137569401e-06, + "loss": 7.6939, + "step": 79 + }, + { + "epoch": 0.002467717245873321, + "grad_norm": 10.02169132232666, + "learning_rate": 1.2338062924120914e-06, + "loss": 7.6251, + "step": 80 + }, + { + "epoch": 0.002498563711446738, + "grad_norm": 9.633440971374512, + "learning_rate": 1.2492288710672424e-06, + "loss": 7.5551, + "step": 81 + }, + { + "epoch": 0.0025294101770201543, + "grad_norm": 9.357256889343262, + "learning_rate": 1.2646514497223936e-06, + "loss": 7.5057, + "step": 82 + }, + { + "epoch": 0.0025602566425935707, + "grad_norm": 8.978069305419922, + "learning_rate": 1.2800740283775449e-06, + "loss": 7.408, + "step": 83 + }, + { + "epoch": 0.0025911031081669875, + "grad_norm": 8.58599853515625, + "learning_rate": 1.295496607032696e-06, + "loss": 7.3447, + "step": 84 + }, + { + "epoch": 0.002621949573740404, + "grad_norm": 8.267987251281738, + "learning_rate": 1.3109191856878471e-06, + "loss": 7.2681, + "step": 85 + }, + { + "epoch": 0.0026527960393138202, + "grad_norm": 7.737511157989502, + "learning_rate": 1.3263417643429982e-06, + "loss": 7.2126, + "step": 86 + }, + { + "epoch": 0.002683642504887237, + "grad_norm": 7.343693256378174, + "learning_rate": 1.3417643429981494e-06, + "loss": 7.1278, + "step": 87 + }, + { + "epoch": 0.0027144889704606534, + "grad_norm": 6.952764511108398, + "learning_rate": 1.3571869216533004e-06, + "loss": 7.0721, + "step": 88 + }, + { + "epoch": 0.0027453354360340698, + "grad_norm": 7.1587324142456055, + "learning_rate": 1.3726095003084516e-06, + "loss": 6.9841, + "step": 89 + }, + { + "epoch": 0.0027761819016074865, + "grad_norm": 6.741302490234375, + "learning_rate": 1.3880320789636027e-06, + "loss": 6.9357, + "step": 90 + }, + { + "epoch": 0.002807028367180903, + "grad_norm": 5.784395217895508, + "learning_rate": 1.403454657618754e-06, + "loss": 6.8746, + "step": 91 + }, + { + "epoch": 0.0028378748327543193, + "grad_norm": 5.607021808624268, + "learning_rate": 1.418877236273905e-06, + "loss": 6.8294, + "step": 92 + }, + { + "epoch": 0.002868721298327736, + "grad_norm": 5.084137916564941, + "learning_rate": 1.4342998149290562e-06, + "loss": 6.794, + "step": 93 + }, + { + "epoch": 0.0028995677639011524, + "grad_norm": 4.953973770141602, + "learning_rate": 1.4497223935842074e-06, + "loss": 6.7081, + "step": 94 + }, + { + "epoch": 0.0029304142294745692, + "grad_norm": 4.368112087249756, + "learning_rate": 1.4651449722393586e-06, + "loss": 6.6548, + "step": 95 + }, + { + "epoch": 0.0029612606950479856, + "grad_norm": 4.1827006340026855, + "learning_rate": 1.4805675508945097e-06, + "loss": 6.5679, + "step": 96 + }, + { + "epoch": 0.002992107160621402, + "grad_norm": 4.07691764831543, + "learning_rate": 1.4959901295496609e-06, + "loss": 6.5314, + "step": 97 + }, + { + "epoch": 0.0030229536261948188, + "grad_norm": 3.782886505126953, + "learning_rate": 1.511412708204812e-06, + "loss": 6.5007, + "step": 98 + }, + { + "epoch": 0.003053800091768235, + "grad_norm": 3.6685757637023926, + "learning_rate": 1.5268352868599631e-06, + "loss": 6.4223, + "step": 99 + }, + { + "epoch": 0.0030846465573416515, + "grad_norm": 3.4558212757110596, + "learning_rate": 1.5422578655151142e-06, + "loss": 6.3922, + "step": 100 + }, + { + "epoch": 0.0031154930229150683, + "grad_norm": 3.084496021270752, + "learning_rate": 1.5576804441702654e-06, + "loss": 6.3968, + "step": 101 + }, + { + "epoch": 0.0031463394884884846, + "grad_norm": 3.0880043506622314, + "learning_rate": 1.5731030228254166e-06, + "loss": 6.3173, + "step": 102 + }, + { + "epoch": 0.003177185954061901, + "grad_norm": 3.040515661239624, + "learning_rate": 1.5885256014805675e-06, + "loss": 6.2513, + "step": 103 + }, + { + "epoch": 0.003208032419635318, + "grad_norm": 3.030207872390747, + "learning_rate": 1.6039481801357187e-06, + "loss": 6.264, + "step": 104 + }, + { + "epoch": 0.003238878885208734, + "grad_norm": 2.9090206623077393, + "learning_rate": 1.61937075879087e-06, + "loss": 6.1888, + "step": 105 + }, + { + "epoch": 0.0032697253507821505, + "grad_norm": 2.632591962814331, + "learning_rate": 1.6347933374460212e-06, + "loss": 6.1465, + "step": 106 + }, + { + "epoch": 0.0033005718163555673, + "grad_norm": 2.591071128845215, + "learning_rate": 1.6502159161011722e-06, + "loss": 6.1074, + "step": 107 + }, + { + "epoch": 0.0033314182819289837, + "grad_norm": 2.3332560062408447, + "learning_rate": 1.6656384947563234e-06, + "loss": 6.0744, + "step": 108 + }, + { + "epoch": 0.0033622647475024, + "grad_norm": 2.4703032970428467, + "learning_rate": 1.6810610734114747e-06, + "loss": 6.0198, + "step": 109 + }, + { + "epoch": 0.003393111213075817, + "grad_norm": 2.214493989944458, + "learning_rate": 1.6964836520666255e-06, + "loss": 6.0518, + "step": 110 + }, + { + "epoch": 0.003423957678649233, + "grad_norm": 2.396679162979126, + "learning_rate": 1.7119062307217767e-06, + "loss": 6.0044, + "step": 111 + }, + { + "epoch": 0.00345480414422265, + "grad_norm": 2.2926342487335205, + "learning_rate": 1.727328809376928e-06, + "loss": 5.973, + "step": 112 + }, + { + "epoch": 0.0034856506097960664, + "grad_norm": 1.9853686094284058, + "learning_rate": 1.7427513880320792e-06, + "loss": 5.9312, + "step": 113 + }, + { + "epoch": 0.0035164970753694827, + "grad_norm": 1.9319242238998413, + "learning_rate": 1.75817396668723e-06, + "loss": 5.8605, + "step": 114 + }, + { + "epoch": 0.0035473435409428995, + "grad_norm": 1.9075840711593628, + "learning_rate": 1.7735965453423812e-06, + "loss": 5.8535, + "step": 115 + }, + { + "epoch": 0.003578190006516316, + "grad_norm": 2.0626113414764404, + "learning_rate": 1.7890191239975324e-06, + "loss": 5.8171, + "step": 116 + }, + { + "epoch": 0.0036090364720897323, + "grad_norm": 1.8688559532165527, + "learning_rate": 1.8044417026526837e-06, + "loss": 5.7626, + "step": 117 + }, + { + "epoch": 0.003639882937663149, + "grad_norm": 1.6804627180099487, + "learning_rate": 1.8198642813078347e-06, + "loss": 5.7491, + "step": 118 + }, + { + "epoch": 0.0036707294032365654, + "grad_norm": 1.792115569114685, + "learning_rate": 1.835286859962986e-06, + "loss": 5.7243, + "step": 119 + }, + { + "epoch": 0.0037015758688099818, + "grad_norm": 1.6904889345169067, + "learning_rate": 1.8507094386181372e-06, + "loss": 5.6928, + "step": 120 + }, + { + "epoch": 0.0037324223343833986, + "grad_norm": 1.6751915216445923, + "learning_rate": 1.8661320172732884e-06, + "loss": 5.6764, + "step": 121 + }, + { + "epoch": 0.003763268799956815, + "grad_norm": 1.521457552909851, + "learning_rate": 1.8815545959284392e-06, + "loss": 5.6797, + "step": 122 + }, + { + "epoch": 0.0037941152655302313, + "grad_norm": 1.6277779340744019, + "learning_rate": 1.8969771745835905e-06, + "loss": 5.6559, + "step": 123 + }, + { + "epoch": 0.003824961731103648, + "grad_norm": 1.4592853784561157, + "learning_rate": 1.9123997532387417e-06, + "loss": 5.664, + "step": 124 + }, + { + "epoch": 0.0038558081966770645, + "grad_norm": 1.413446068763733, + "learning_rate": 1.9278223318938927e-06, + "loss": 5.5794, + "step": 125 + }, + { + "epoch": 0.003886654662250481, + "grad_norm": 1.38154137134552, + "learning_rate": 1.9432449105490437e-06, + "loss": 5.5492, + "step": 126 + }, + { + "epoch": 0.003917501127823898, + "grad_norm": 1.629254698753357, + "learning_rate": 1.958667489204195e-06, + "loss": 5.5247, + "step": 127 + }, + { + "epoch": 0.003948347593397314, + "grad_norm": 1.6704044342041016, + "learning_rate": 1.974090067859346e-06, + "loss": 5.5003, + "step": 128 + }, + { + "epoch": 0.00397919405897073, + "grad_norm": 1.2934266328811646, + "learning_rate": 1.9895126465144977e-06, + "loss": 5.5024, + "step": 129 + }, + { + "epoch": 0.004010040524544147, + "grad_norm": 1.3899742364883423, + "learning_rate": 2.0049352251696483e-06, + "loss": 5.5377, + "step": 130 + }, + { + "epoch": 0.004040886990117564, + "grad_norm": 1.3179103136062622, + "learning_rate": 2.0203578038247997e-06, + "loss": 5.4767, + "step": 131 + }, + { + "epoch": 0.00407173345569098, + "grad_norm": 1.1811275482177734, + "learning_rate": 2.0357803824799507e-06, + "loss": 5.4289, + "step": 132 + }, + { + "epoch": 0.004102579921264397, + "grad_norm": 1.3001583814620972, + "learning_rate": 2.051202961135102e-06, + "loss": 5.4465, + "step": 133 + }, + { + "epoch": 0.0041334263868378135, + "grad_norm": 1.2545536756515503, + "learning_rate": 2.0666255397902528e-06, + "loss": 5.4184, + "step": 134 + }, + { + "epoch": 0.004164272852411229, + "grad_norm": 1.1940233707427979, + "learning_rate": 2.0820481184454042e-06, + "loss": 5.378, + "step": 135 + }, + { + "epoch": 0.004195119317984646, + "grad_norm": 1.3556828498840332, + "learning_rate": 2.0974706971005552e-06, + "loss": 5.3728, + "step": 136 + }, + { + "epoch": 0.004225965783558063, + "grad_norm": 1.3728647232055664, + "learning_rate": 2.1128932757557067e-06, + "loss": 5.2926, + "step": 137 + }, + { + "epoch": 0.004256812249131479, + "grad_norm": 1.1112613677978516, + "learning_rate": 2.1283158544108577e-06, + "loss": 5.3196, + "step": 138 + }, + { + "epoch": 0.004287658714704896, + "grad_norm": 0.9983944892883301, + "learning_rate": 2.1437384330660087e-06, + "loss": 5.3063, + "step": 139 + }, + { + "epoch": 0.0043185051802783125, + "grad_norm": 0.9900311231613159, + "learning_rate": 2.15916101172116e-06, + "loss": 5.244, + "step": 140 + }, + { + "epoch": 0.004349351645851728, + "grad_norm": 1.1906732320785522, + "learning_rate": 2.1745835903763108e-06, + "loss": 5.2281, + "step": 141 + }, + { + "epoch": 0.004380198111425145, + "grad_norm": 1.2363054752349854, + "learning_rate": 2.1900061690314622e-06, + "loss": 5.2855, + "step": 142 + }, + { + "epoch": 0.004411044576998562, + "grad_norm": 1.1300644874572754, + "learning_rate": 2.2054287476866132e-06, + "loss": 5.2867, + "step": 143 + }, + { + "epoch": 0.004441891042571978, + "grad_norm": 0.9496215581893921, + "learning_rate": 2.2208513263417647e-06, + "loss": 5.1697, + "step": 144 + }, + { + "epoch": 0.004472737508145395, + "grad_norm": 1.7752348184585571, + "learning_rate": 2.2362739049969153e-06, + "loss": 5.1822, + "step": 145 + }, + { + "epoch": 0.0045035839737188115, + "grad_norm": 1.28167724609375, + "learning_rate": 2.2516964836520667e-06, + "loss": 5.2159, + "step": 146 + }, + { + "epoch": 0.0045344304392922275, + "grad_norm": 0.9614664912223816, + "learning_rate": 2.2671190623072178e-06, + "loss": 5.1627, + "step": 147 + }, + { + "epoch": 0.004565276904865644, + "grad_norm": 0.9931091070175171, + "learning_rate": 2.282541640962369e-06, + "loss": 5.1464, + "step": 148 + }, + { + "epoch": 0.004596123370439061, + "grad_norm": 1.0743926763534546, + "learning_rate": 2.2979642196175202e-06, + "loss": 5.1513, + "step": 149 + }, + { + "epoch": 0.004626969836012477, + "grad_norm": 1.0678631067276, + "learning_rate": 2.3133867982726713e-06, + "loss": 5.1062, + "step": 150 + }, + { + "epoch": 0.004657816301585894, + "grad_norm": 1.236189365386963, + "learning_rate": 2.3288093769278223e-06, + "loss": 5.1572, + "step": 151 + }, + { + "epoch": 0.004688662767159311, + "grad_norm": 0.9001237750053406, + "learning_rate": 2.3442319555829737e-06, + "loss": 5.1229, + "step": 152 + }, + { + "epoch": 0.0047195092327327265, + "grad_norm": 0.8681719303131104, + "learning_rate": 2.3596545342381248e-06, + "loss": 5.0788, + "step": 153 + }, + { + "epoch": 0.004750355698306143, + "grad_norm": 0.898551344871521, + "learning_rate": 2.3750771128932758e-06, + "loss": 5.0359, + "step": 154 + }, + { + "epoch": 0.00478120216387956, + "grad_norm": 1.0069688558578491, + "learning_rate": 2.3904996915484272e-06, + "loss": 5.0729, + "step": 155 + }, + { + "epoch": 0.004812048629452977, + "grad_norm": 1.0892609357833862, + "learning_rate": 2.4059222702035782e-06, + "loss": 5.071, + "step": 156 + }, + { + "epoch": 0.004842895095026393, + "grad_norm": 0.8369899392127991, + "learning_rate": 2.4213448488587293e-06, + "loss": 5.0815, + "step": 157 + }, + { + "epoch": 0.00487374156059981, + "grad_norm": 0.9352590441703796, + "learning_rate": 2.4367674275138803e-06, + "loss": 5.0481, + "step": 158 + }, + { + "epoch": 0.0049045880261732264, + "grad_norm": 1.1740933656692505, + "learning_rate": 2.4521900061690317e-06, + "loss": 5.0617, + "step": 159 + }, + { + "epoch": 0.004935434491746642, + "grad_norm": 0.860466480255127, + "learning_rate": 2.4676125848241828e-06, + "loss": 5.0145, + "step": 160 + }, + { + "epoch": 0.004966280957320059, + "grad_norm": 0.8278803825378418, + "learning_rate": 2.4830351634793338e-06, + "loss": 5.0424, + "step": 161 + }, + { + "epoch": 0.004997127422893476, + "grad_norm": 1.0816013813018799, + "learning_rate": 2.498457742134485e-06, + "loss": 5.031, + "step": 162 + }, + { + "epoch": 0.005027973888466892, + "grad_norm": 0.9863463640213013, + "learning_rate": 2.5138803207896363e-06, + "loss": 4.9475, + "step": 163 + }, + { + "epoch": 0.005058820354040309, + "grad_norm": 0.7756854295730591, + "learning_rate": 2.5293028994447873e-06, + "loss": 4.9759, + "step": 164 + }, + { + "epoch": 0.0050896668196137255, + "grad_norm": 0.8434688448905945, + "learning_rate": 2.5447254780999383e-06, + "loss": 4.9679, + "step": 165 + }, + { + "epoch": 0.005120513285187141, + "grad_norm": 0.8642579913139343, + "learning_rate": 2.5601480567550897e-06, + "loss": 4.9397, + "step": 166 + }, + { + "epoch": 0.005151359750760558, + "grad_norm": 0.7430039048194885, + "learning_rate": 2.5755706354102408e-06, + "loss": 4.9488, + "step": 167 + }, + { + "epoch": 0.005182206216333975, + "grad_norm": 0.8508272171020508, + "learning_rate": 2.590993214065392e-06, + "loss": 4.9249, + "step": 168 + }, + { + "epoch": 0.005213052681907391, + "grad_norm": 0.7853840589523315, + "learning_rate": 2.606415792720543e-06, + "loss": 4.9059, + "step": 169 + }, + { + "epoch": 0.005243899147480808, + "grad_norm": 0.724083423614502, + "learning_rate": 2.6218383713756943e-06, + "loss": 4.8948, + "step": 170 + }, + { + "epoch": 0.0052747456130542245, + "grad_norm": 0.7432600855827332, + "learning_rate": 2.6372609500308453e-06, + "loss": 4.9173, + "step": 171 + }, + { + "epoch": 0.0053055920786276405, + "grad_norm": 0.674286425113678, + "learning_rate": 2.6526835286859963e-06, + "loss": 4.8329, + "step": 172 + }, + { + "epoch": 0.005336438544201057, + "grad_norm": 0.7284206748008728, + "learning_rate": 2.6681061073411473e-06, + "loss": 4.9136, + "step": 173 + }, + { + "epoch": 0.005367285009774474, + "grad_norm": 0.6132212281227112, + "learning_rate": 2.6835286859962988e-06, + "loss": 4.8715, + "step": 174 + }, + { + "epoch": 0.00539813147534789, + "grad_norm": 0.7052930593490601, + "learning_rate": 2.69895126465145e-06, + "loss": 4.856, + "step": 175 + }, + { + "epoch": 0.005428977940921307, + "grad_norm": 0.6743758916854858, + "learning_rate": 2.714373843306601e-06, + "loss": 4.8642, + "step": 176 + }, + { + "epoch": 0.005459824406494724, + "grad_norm": 0.68760746717453, + "learning_rate": 2.7297964219617523e-06, + "loss": 4.8625, + "step": 177 + }, + { + "epoch": 0.0054906708720681395, + "grad_norm": 0.7693434953689575, + "learning_rate": 2.7452190006169033e-06, + "loss": 4.8377, + "step": 178 + }, + { + "epoch": 0.005521517337641556, + "grad_norm": 0.7133588790893555, + "learning_rate": 2.7606415792720547e-06, + "loss": 4.8462, + "step": 179 + }, + { + "epoch": 0.005552363803214973, + "grad_norm": 0.8171933889389038, + "learning_rate": 2.7760641579272053e-06, + "loss": 4.8486, + "step": 180 + }, + { + "epoch": 0.005583210268788389, + "grad_norm": 0.8579136729240417, + "learning_rate": 2.7914867365823568e-06, + "loss": 4.8297, + "step": 181 + }, + { + "epoch": 0.005614056734361806, + "grad_norm": 0.7044315338134766, + "learning_rate": 2.806909315237508e-06, + "loss": 4.8475, + "step": 182 + }, + { + "epoch": 0.005644903199935223, + "grad_norm": 0.7016251683235168, + "learning_rate": 2.8223318938926593e-06, + "loss": 4.8035, + "step": 183 + }, + { + "epoch": 0.0056757496655086385, + "grad_norm": 0.7976976633071899, + "learning_rate": 2.83775447254781e-06, + "loss": 4.7734, + "step": 184 + }, + { + "epoch": 0.005706596131082055, + "grad_norm": 0.7739298939704895, + "learning_rate": 2.8531770512029613e-06, + "loss": 4.7937, + "step": 185 + }, + { + "epoch": 0.005737442596655472, + "grad_norm": 0.6487588286399841, + "learning_rate": 2.8685996298581123e-06, + "loss": 4.7722, + "step": 186 + }, + { + "epoch": 0.005768289062228888, + "grad_norm": 0.7361035346984863, + "learning_rate": 2.8840222085132638e-06, + "loss": 4.8198, + "step": 187 + }, + { + "epoch": 0.005799135527802305, + "grad_norm": 0.6988633871078491, + "learning_rate": 2.899444787168415e-06, + "loss": 4.7622, + "step": 188 + }, + { + "epoch": 0.005829981993375722, + "grad_norm": 0.6415325403213501, + "learning_rate": 2.914867365823566e-06, + "loss": 4.7322, + "step": 189 + }, + { + "epoch": 0.0058608284589491385, + "grad_norm": 0.6974925994873047, + "learning_rate": 2.9302899444787173e-06, + "loss": 4.7219, + "step": 190 + }, + { + "epoch": 0.005891674924522554, + "grad_norm": 0.8437878489494324, + "learning_rate": 2.9457125231338683e-06, + "loss": 4.7332, + "step": 191 + }, + { + "epoch": 0.005922521390095971, + "grad_norm": 0.6838964223861694, + "learning_rate": 2.9611351017890193e-06, + "loss": 4.7063, + "step": 192 + }, + { + "epoch": 0.005953367855669388, + "grad_norm": 0.5635040402412415, + "learning_rate": 2.9765576804441703e-06, + "loss": 4.7088, + "step": 193 + }, + { + "epoch": 0.005984214321242804, + "grad_norm": 0.5678243041038513, + "learning_rate": 2.9919802590993218e-06, + "loss": 4.7236, + "step": 194 + }, + { + "epoch": 0.006015060786816221, + "grad_norm": 0.5828536748886108, + "learning_rate": 3.0074028377544724e-06, + "loss": 4.7278, + "step": 195 + }, + { + "epoch": 0.0060459072523896375, + "grad_norm": 0.6387608647346497, + "learning_rate": 3.022825416409624e-06, + "loss": 4.7229, + "step": 196 + }, + { + "epoch": 0.0060767537179630534, + "grad_norm": 0.5523983240127563, + "learning_rate": 3.038247995064775e-06, + "loss": 4.6855, + "step": 197 + }, + { + "epoch": 0.00610760018353647, + "grad_norm": 0.6500614881515503, + "learning_rate": 3.0536705737199263e-06, + "loss": 4.7098, + "step": 198 + }, + { + "epoch": 0.006138446649109887, + "grad_norm": 0.6111279726028442, + "learning_rate": 3.069093152375077e-06, + "loss": 4.6977, + "step": 199 + }, + { + "epoch": 0.006169293114683303, + "grad_norm": 0.5749845504760742, + "learning_rate": 3.0845157310302283e-06, + "loss": 4.6636, + "step": 200 + }, + { + "epoch": 0.00620013958025672, + "grad_norm": 0.6371256113052368, + "learning_rate": 3.0999383096853794e-06, + "loss": 4.6415, + "step": 201 + }, + { + "epoch": 0.0062309860458301366, + "grad_norm": 0.6452645659446716, + "learning_rate": 3.115360888340531e-06, + "loss": 4.6894, + "step": 202 + }, + { + "epoch": 0.0062618325114035525, + "grad_norm": 0.5660443902015686, + "learning_rate": 3.130783466995682e-06, + "loss": 4.6811, + "step": 203 + }, + { + "epoch": 0.006292678976976969, + "grad_norm": 0.6819775104522705, + "learning_rate": 3.1462060456508333e-06, + "loss": 4.6205, + "step": 204 + }, + { + "epoch": 0.006323525442550386, + "grad_norm": 0.7046071290969849, + "learning_rate": 3.161628624305984e-06, + "loss": 4.6737, + "step": 205 + }, + { + "epoch": 0.006354371908123802, + "grad_norm": 0.6260894536972046, + "learning_rate": 3.177051202961135e-06, + "loss": 4.627, + "step": 206 + }, + { + "epoch": 0.006385218373697219, + "grad_norm": 0.5110775232315063, + "learning_rate": 3.1924737816162864e-06, + "loss": 4.652, + "step": 207 + }, + { + "epoch": 0.006416064839270636, + "grad_norm": 0.5455238223075867, + "learning_rate": 3.2078963602714374e-06, + "loss": 4.6651, + "step": 208 + }, + { + "epoch": 0.0064469113048440515, + "grad_norm": 0.523145854473114, + "learning_rate": 3.223318938926589e-06, + "loss": 4.6382, + "step": 209 + }, + { + "epoch": 0.006477757770417468, + "grad_norm": 0.602391242980957, + "learning_rate": 3.23874151758174e-06, + "loss": 4.6366, + "step": 210 + }, + { + "epoch": 0.006508604235990885, + "grad_norm": 0.6408519148826599, + "learning_rate": 3.2541640962368913e-06, + "loss": 4.5729, + "step": 211 + }, + { + "epoch": 0.006539450701564301, + "grad_norm": 0.6153237223625183, + "learning_rate": 3.2695866748920423e-06, + "loss": 4.5909, + "step": 212 + }, + { + "epoch": 0.006570297167137718, + "grad_norm": 0.5435067415237427, + "learning_rate": 3.285009253547193e-06, + "loss": 4.6055, + "step": 213 + }, + { + "epoch": 0.006601143632711135, + "grad_norm": 0.5963379740715027, + "learning_rate": 3.3004318322023444e-06, + "loss": 4.6907, + "step": 214 + }, + { + "epoch": 0.006631990098284551, + "grad_norm": 0.528070867061615, + "learning_rate": 3.3158544108574954e-06, + "loss": 4.6003, + "step": 215 + }, + { + "epoch": 0.006662836563857967, + "grad_norm": 0.5825647115707397, + "learning_rate": 3.331276989512647e-06, + "loss": 4.6308, + "step": 216 + }, + { + "epoch": 0.006693683029431384, + "grad_norm": 0.56778883934021, + "learning_rate": 3.346699568167798e-06, + "loss": 4.557, + "step": 217 + }, + { + "epoch": 0.0067245294950048, + "grad_norm": 0.533927321434021, + "learning_rate": 3.3621221468229493e-06, + "loss": 4.5746, + "step": 218 + }, + { + "epoch": 0.006755375960578217, + "grad_norm": 0.6435202360153198, + "learning_rate": 3.3775447254781003e-06, + "loss": 4.6072, + "step": 219 + }, + { + "epoch": 0.006786222426151634, + "grad_norm": 0.5651525259017944, + "learning_rate": 3.392967304133251e-06, + "loss": 4.5994, + "step": 220 + }, + { + "epoch": 0.00681706889172505, + "grad_norm": 0.5186142325401306, + "learning_rate": 3.408389882788402e-06, + "loss": 4.5814, + "step": 221 + }, + { + "epoch": 0.006847915357298466, + "grad_norm": 0.6316450238227844, + "learning_rate": 3.4238124614435534e-06, + "loss": 4.5875, + "step": 222 + }, + { + "epoch": 0.006878761822871883, + "grad_norm": 0.5337743759155273, + "learning_rate": 3.4392350400987044e-06, + "loss": 4.5799, + "step": 223 + }, + { + "epoch": 0.0069096082884453, + "grad_norm": 0.5976253747940063, + "learning_rate": 3.454657618753856e-06, + "loss": 4.5745, + "step": 224 + }, + { + "epoch": 0.006940454754018716, + "grad_norm": 0.5289996266365051, + "learning_rate": 3.470080197409007e-06, + "loss": 4.5571, + "step": 225 + }, + { + "epoch": 0.006971301219592133, + "grad_norm": 0.5527687668800354, + "learning_rate": 3.4855027760641583e-06, + "loss": 4.5608, + "step": 226 + }, + { + "epoch": 0.0070021476851655495, + "grad_norm": 0.5198066234588623, + "learning_rate": 3.5009253547193094e-06, + "loss": 4.5388, + "step": 227 + }, + { + "epoch": 0.0070329941507389655, + "grad_norm": 0.5727770924568176, + "learning_rate": 3.51634793337446e-06, + "loss": 4.5667, + "step": 228 + }, + { + "epoch": 0.007063840616312382, + "grad_norm": 0.5684541463851929, + "learning_rate": 3.5317705120296114e-06, + "loss": 4.5446, + "step": 229 + }, + { + "epoch": 0.007094687081885799, + "grad_norm": 0.4761212170124054, + "learning_rate": 3.5471930906847624e-06, + "loss": 4.5625, + "step": 230 + }, + { + "epoch": 0.007125533547459215, + "grad_norm": 0.6094976663589478, + "learning_rate": 3.562615669339914e-06, + "loss": 4.5027, + "step": 231 + }, + { + "epoch": 0.007156380013032632, + "grad_norm": 0.5641266107559204, + "learning_rate": 3.578038247995065e-06, + "loss": 4.5569, + "step": 232 + }, + { + "epoch": 0.007187226478606049, + "grad_norm": 0.5055134892463684, + "learning_rate": 3.5934608266502163e-06, + "loss": 4.504, + "step": 233 + }, + { + "epoch": 0.0072180729441794645, + "grad_norm": 0.5287951827049255, + "learning_rate": 3.6088834053053674e-06, + "loss": 4.5005, + "step": 234 + }, + { + "epoch": 0.007248919409752881, + "grad_norm": 0.48975786566734314, + "learning_rate": 3.624305983960519e-06, + "loss": 4.5481, + "step": 235 + }, + { + "epoch": 0.007279765875326298, + "grad_norm": 0.5442043542861938, + "learning_rate": 3.6397285626156694e-06, + "loss": 4.5392, + "step": 236 + }, + { + "epoch": 0.007310612340899714, + "grad_norm": 0.5715160965919495, + "learning_rate": 3.6551511412708204e-06, + "loss": 4.5103, + "step": 237 + }, + { + "epoch": 0.007341458806473131, + "grad_norm": 0.49980753660202026, + "learning_rate": 3.670573719925972e-06, + "loss": 4.4945, + "step": 238 + }, + { + "epoch": 0.007372305272046548, + "grad_norm": 0.5992623567581177, + "learning_rate": 3.685996298581123e-06, + "loss": 4.5146, + "step": 239 + }, + { + "epoch": 0.0074031517376199635, + "grad_norm": 0.5263873934745789, + "learning_rate": 3.7014188772362744e-06, + "loss": 4.5327, + "step": 240 + }, + { + "epoch": 0.00743399820319338, + "grad_norm": 0.5474061965942383, + "learning_rate": 3.7168414558914254e-06, + "loss": 4.5363, + "step": 241 + }, + { + "epoch": 0.007464844668766797, + "grad_norm": 0.518379271030426, + "learning_rate": 3.732264034546577e-06, + "loss": 4.4871, + "step": 242 + }, + { + "epoch": 0.007495691134340213, + "grad_norm": 0.5149906873703003, + "learning_rate": 3.747686613201728e-06, + "loss": 4.4816, + "step": 243 + }, + { + "epoch": 0.00752653759991363, + "grad_norm": 0.4773837625980377, + "learning_rate": 3.7631091918568784e-06, + "loss": 4.4808, + "step": 244 + }, + { + "epoch": 0.007557384065487047, + "grad_norm": 0.7583971619606018, + "learning_rate": 3.7785317705120295e-06, + "loss": 4.4675, + "step": 245 + }, + { + "epoch": 0.007588230531060463, + "grad_norm": 0.45772066712379456, + "learning_rate": 3.793954349167181e-06, + "loss": 4.4349, + "step": 246 + }, + { + "epoch": 0.007619076996633879, + "grad_norm": 0.6742913126945496, + "learning_rate": 3.809376927822332e-06, + "loss": 4.5062, + "step": 247 + }, + { + "epoch": 0.007649923462207296, + "grad_norm": 0.5423294305801392, + "learning_rate": 3.824799506477483e-06, + "loss": 4.5019, + "step": 248 + }, + { + "epoch": 0.007680769927780712, + "grad_norm": 0.5376593470573425, + "learning_rate": 3.840222085132635e-06, + "loss": 4.4581, + "step": 249 + }, + { + "epoch": 0.007711616393354129, + "grad_norm": 0.4981565475463867, + "learning_rate": 3.8556446637877854e-06, + "loss": 4.4683, + "step": 250 + }, + { + "epoch": 0.007742462858927546, + "grad_norm": 0.532468855381012, + "learning_rate": 3.871067242442936e-06, + "loss": 4.4757, + "step": 251 + }, + { + "epoch": 0.007773309324500962, + "grad_norm": 0.599917471408844, + "learning_rate": 3.8864898210980875e-06, + "loss": 4.4348, + "step": 252 + }, + { + "epoch": 0.0078041557900743784, + "grad_norm": 0.45797646045684814, + "learning_rate": 3.901912399753239e-06, + "loss": 4.4501, + "step": 253 + }, + { + "epoch": 0.007835002255647795, + "grad_norm": 0.5251150131225586, + "learning_rate": 3.91733497840839e-06, + "loss": 4.4498, + "step": 254 + }, + { + "epoch": 0.007865848721221211, + "grad_norm": 0.5094879269599915, + "learning_rate": 3.932757557063541e-06, + "loss": 4.4645, + "step": 255 + }, + { + "epoch": 0.007896695186794629, + "grad_norm": 0.5041837096214294, + "learning_rate": 3.948180135718692e-06, + "loss": 4.4419, + "step": 256 + }, + { + "epoch": 0.007927541652368045, + "grad_norm": 0.5862743258476257, + "learning_rate": 3.963602714373844e-06, + "loss": 4.4767, + "step": 257 + }, + { + "epoch": 0.00795838811794146, + "grad_norm": 0.5594961047172546, + "learning_rate": 3.979025293028995e-06, + "loss": 4.4598, + "step": 258 + }, + { + "epoch": 0.007989234583514878, + "grad_norm": 0.4845632016658783, + "learning_rate": 3.994447871684145e-06, + "loss": 4.3948, + "step": 259 + }, + { + "epoch": 0.008020081049088294, + "grad_norm": 0.5313999056816101, + "learning_rate": 4.0098704503392965e-06, + "loss": 4.4159, + "step": 260 + }, + { + "epoch": 0.00805092751466171, + "grad_norm": 0.4561810791492462, + "learning_rate": 4.025293028994448e-06, + "loss": 4.4154, + "step": 261 + }, + { + "epoch": 0.008081773980235128, + "grad_norm": 0.5163814425468445, + "learning_rate": 4.040715607649599e-06, + "loss": 4.4678, + "step": 262 + }, + { + "epoch": 0.008112620445808544, + "grad_norm": 0.5073924660682678, + "learning_rate": 4.05613818630475e-06, + "loss": 4.4045, + "step": 263 + }, + { + "epoch": 0.00814346691138196, + "grad_norm": 0.4644468426704407, + "learning_rate": 4.0715607649599014e-06, + "loss": 4.4321, + "step": 264 + }, + { + "epoch": 0.008174313376955377, + "grad_norm": 0.47268030047416687, + "learning_rate": 4.086983343615053e-06, + "loss": 4.4327, + "step": 265 + }, + { + "epoch": 0.008205159842528793, + "grad_norm": 0.5479356646537781, + "learning_rate": 4.102405922270204e-06, + "loss": 4.4353, + "step": 266 + }, + { + "epoch": 0.00823600630810221, + "grad_norm": 0.4277481436729431, + "learning_rate": 4.117828500925355e-06, + "loss": 4.4215, + "step": 267 + }, + { + "epoch": 0.008266852773675627, + "grad_norm": 0.5674469470977783, + "learning_rate": 4.1332510795805055e-06, + "loss": 4.3995, + "step": 268 + }, + { + "epoch": 0.008297699239249043, + "grad_norm": 0.4517870247364044, + "learning_rate": 4.148673658235657e-06, + "loss": 4.4281, + "step": 269 + }, + { + "epoch": 0.008328545704822459, + "grad_norm": 0.48754599690437317, + "learning_rate": 4.1640962368908084e-06, + "loss": 4.3923, + "step": 270 + }, + { + "epoch": 0.008359392170395876, + "grad_norm": 0.43422752618789673, + "learning_rate": 4.17951881554596e-06, + "loss": 4.4144, + "step": 271 + }, + { + "epoch": 0.008390238635969292, + "grad_norm": 0.4442002475261688, + "learning_rate": 4.1949413942011105e-06, + "loss": 4.3933, + "step": 272 + }, + { + "epoch": 0.008421085101542708, + "grad_norm": 0.42256155610084534, + "learning_rate": 4.210363972856262e-06, + "loss": 4.3781, + "step": 273 + }, + { + "epoch": 0.008451931567116126, + "grad_norm": 0.4278530478477478, + "learning_rate": 4.225786551511413e-06, + "loss": 4.4194, + "step": 274 + }, + { + "epoch": 0.008482778032689542, + "grad_norm": 0.491337388753891, + "learning_rate": 4.241209130166564e-06, + "loss": 4.4132, + "step": 275 + }, + { + "epoch": 0.008513624498262958, + "grad_norm": 0.44742700457572937, + "learning_rate": 4.256631708821715e-06, + "loss": 4.3971, + "step": 276 + }, + { + "epoch": 0.008544470963836375, + "grad_norm": 0.5613349676132202, + "learning_rate": 4.272054287476866e-06, + "loss": 4.4148, + "step": 277 + }, + { + "epoch": 0.008575317429409791, + "grad_norm": 0.45742297172546387, + "learning_rate": 4.2874768661320175e-06, + "loss": 4.3786, + "step": 278 + }, + { + "epoch": 0.008606163894983207, + "grad_norm": 0.46094775199890137, + "learning_rate": 4.302899444787169e-06, + "loss": 4.3577, + "step": 279 + }, + { + "epoch": 0.008637010360556625, + "grad_norm": 0.4680476486682892, + "learning_rate": 4.31832202344232e-06, + "loss": 4.3904, + "step": 280 + }, + { + "epoch": 0.008667856826130041, + "grad_norm": 0.5356520414352417, + "learning_rate": 4.333744602097471e-06, + "loss": 4.3874, + "step": 281 + }, + { + "epoch": 0.008698703291703457, + "grad_norm": 0.4329937696456909, + "learning_rate": 4.3491671807526216e-06, + "loss": 4.4098, + "step": 282 + }, + { + "epoch": 0.008729549757276875, + "grad_norm": 0.4725155830383301, + "learning_rate": 4.364589759407773e-06, + "loss": 4.3782, + "step": 283 + }, + { + "epoch": 0.00876039622285029, + "grad_norm": 0.5237762928009033, + "learning_rate": 4.3800123380629245e-06, + "loss": 4.3781, + "step": 284 + }, + { + "epoch": 0.008791242688423706, + "grad_norm": 0.4305984377861023, + "learning_rate": 4.395434916718075e-06, + "loss": 4.4087, + "step": 285 + }, + { + "epoch": 0.008822089153997124, + "grad_norm": 0.4980679154396057, + "learning_rate": 4.4108574953732265e-06, + "loss": 4.3396, + "step": 286 + }, + { + "epoch": 0.00885293561957054, + "grad_norm": 0.5026816129684448, + "learning_rate": 4.426280074028378e-06, + "loss": 4.4125, + "step": 287 + }, + { + "epoch": 0.008883782085143956, + "grad_norm": 0.4740833640098572, + "learning_rate": 4.441702652683529e-06, + "loss": 4.3658, + "step": 288 + }, + { + "epoch": 0.008914628550717374, + "grad_norm": 0.5047520399093628, + "learning_rate": 4.45712523133868e-06, + "loss": 4.3355, + "step": 289 + }, + { + "epoch": 0.00894547501629079, + "grad_norm": 0.5570448040962219, + "learning_rate": 4.472547809993831e-06, + "loss": 4.3989, + "step": 290 + }, + { + "epoch": 0.008976321481864205, + "grad_norm": 0.4474649131298065, + "learning_rate": 4.487970388648982e-06, + "loss": 4.3168, + "step": 291 + }, + { + "epoch": 0.009007167947437623, + "grad_norm": 0.5314493775367737, + "learning_rate": 4.5033929673041335e-06, + "loss": 4.3089, + "step": 292 + }, + { + "epoch": 0.009038014413011039, + "grad_norm": 0.42806127667427063, + "learning_rate": 4.518815545959285e-06, + "loss": 4.3573, + "step": 293 + }, + { + "epoch": 0.009068860878584455, + "grad_norm": 0.4819366931915283, + "learning_rate": 4.5342381246144355e-06, + "loss": 4.3466, + "step": 294 + }, + { + "epoch": 0.009099707344157873, + "grad_norm": 0.4401512145996094, + "learning_rate": 4.549660703269587e-06, + "loss": 4.3441, + "step": 295 + }, + { + "epoch": 0.009130553809731289, + "grad_norm": 0.4702405631542206, + "learning_rate": 4.565083281924738e-06, + "loss": 4.3156, + "step": 296 + }, + { + "epoch": 0.009161400275304704, + "grad_norm": 0.4395723044872284, + "learning_rate": 4.58050586057989e-06, + "loss": 4.271, + "step": 297 + }, + { + "epoch": 0.009192246740878122, + "grad_norm": 0.41895321011543274, + "learning_rate": 4.5959284392350405e-06, + "loss": 4.3357, + "step": 298 + }, + { + "epoch": 0.009223093206451538, + "grad_norm": 0.46962985396385193, + "learning_rate": 4.611351017890191e-06, + "loss": 4.3675, + "step": 299 + }, + { + "epoch": 0.009253939672024954, + "grad_norm": 0.4045839309692383, + "learning_rate": 4.6267735965453425e-06, + "loss": 4.3324, + "step": 300 + }, + { + "epoch": 0.009284786137598372, + "grad_norm": 0.46376198530197144, + "learning_rate": 4.642196175200494e-06, + "loss": 4.3325, + "step": 301 + }, + { + "epoch": 0.009315632603171788, + "grad_norm": 0.48544812202453613, + "learning_rate": 4.6576187538556446e-06, + "loss": 4.3197, + "step": 302 + }, + { + "epoch": 0.009346479068745204, + "grad_norm": 0.38740357756614685, + "learning_rate": 4.673041332510796e-06, + "loss": 4.332, + "step": 303 + }, + { + "epoch": 0.009377325534318621, + "grad_norm": 0.4687975347042084, + "learning_rate": 4.6884639111659475e-06, + "loss": 4.2901, + "step": 304 + }, + { + "epoch": 0.009408171999892037, + "grad_norm": 0.4003926217556, + "learning_rate": 4.703886489821098e-06, + "loss": 4.3017, + "step": 305 + }, + { + "epoch": 0.009439018465465453, + "grad_norm": 0.5348908305168152, + "learning_rate": 4.7193090684762495e-06, + "loss": 4.3359, + "step": 306 + }, + { + "epoch": 0.00946986493103887, + "grad_norm": 0.4703734219074249, + "learning_rate": 4.7347316471314e-06, + "loss": 4.2996, + "step": 307 + }, + { + "epoch": 0.009500711396612287, + "grad_norm": 0.4117778241634369, + "learning_rate": 4.7501542257865515e-06, + "loss": 4.3093, + "step": 308 + }, + { + "epoch": 0.009531557862185703, + "grad_norm": 0.4532212018966675, + "learning_rate": 4.765576804441703e-06, + "loss": 4.2798, + "step": 309 + }, + { + "epoch": 0.00956240432775912, + "grad_norm": 0.47473230957984924, + "learning_rate": 4.7809993830968544e-06, + "loss": 4.2749, + "step": 310 + }, + { + "epoch": 0.009593250793332536, + "grad_norm": 0.43723064661026, + "learning_rate": 4.796421961752005e-06, + "loss": 4.3216, + "step": 311 + }, + { + "epoch": 0.009624097258905954, + "grad_norm": 0.39337158203125, + "learning_rate": 4.8118445404071565e-06, + "loss": 4.3094, + "step": 312 + }, + { + "epoch": 0.00965494372447937, + "grad_norm": 0.4164552688598633, + "learning_rate": 4.827267119062307e-06, + "loss": 4.2723, + "step": 313 + }, + { + "epoch": 0.009685790190052786, + "grad_norm": 0.4631640911102295, + "learning_rate": 4.8426896977174585e-06, + "loss": 4.2768, + "step": 314 + }, + { + "epoch": 0.009716636655626203, + "grad_norm": 0.5098804235458374, + "learning_rate": 4.85811227637261e-06, + "loss": 4.2946, + "step": 315 + }, + { + "epoch": 0.00974748312119962, + "grad_norm": 0.44128119945526123, + "learning_rate": 4.873534855027761e-06, + "loss": 4.2574, + "step": 316 + }, + { + "epoch": 0.009778329586773035, + "grad_norm": 0.4407106935977936, + "learning_rate": 4.888957433682912e-06, + "loss": 4.2569, + "step": 317 + }, + { + "epoch": 0.009809176052346453, + "grad_norm": 0.43998202681541443, + "learning_rate": 4.9043800123380635e-06, + "loss": 4.2705, + "step": 318 + }, + { + "epoch": 0.009840022517919869, + "grad_norm": 0.4628184735774994, + "learning_rate": 4.919802590993215e-06, + "loss": 4.291, + "step": 319 + }, + { + "epoch": 0.009870868983493285, + "grad_norm": 0.46835052967071533, + "learning_rate": 4.9352251696483655e-06, + "loss": 4.2519, + "step": 320 + }, + { + "epoch": 0.009901715449066702, + "grad_norm": 0.41583818197250366, + "learning_rate": 4.950647748303516e-06, + "loss": 4.3101, + "step": 321 + }, + { + "epoch": 0.009932561914640118, + "grad_norm": 0.5304214358329773, + "learning_rate": 4.9660703269586676e-06, + "loss": 4.2609, + "step": 322 + }, + { + "epoch": 0.009963408380213534, + "grad_norm": 0.4371168613433838, + "learning_rate": 4.981492905613819e-06, + "loss": 4.3039, + "step": 323 + }, + { + "epoch": 0.009994254845786952, + "grad_norm": 0.47491657733917236, + "learning_rate": 4.99691548426897e-06, + "loss": 4.3083, + "step": 324 + }, + { + "epoch": 0.010025101311360368, + "grad_norm": 0.467857301235199, + "learning_rate": 5.012338062924121e-06, + "loss": 4.3043, + "step": 325 + }, + { + "epoch": 0.010055947776933784, + "grad_norm": 0.4635056257247925, + "learning_rate": 5.0277606415792725e-06, + "loss": 4.2442, + "step": 326 + }, + { + "epoch": 0.010086794242507201, + "grad_norm": 0.5144153833389282, + "learning_rate": 5.043183220234424e-06, + "loss": 4.2596, + "step": 327 + }, + { + "epoch": 0.010117640708080617, + "grad_norm": 0.5868098139762878, + "learning_rate": 5.0586057988895746e-06, + "loss": 4.2636, + "step": 328 + }, + { + "epoch": 0.010148487173654033, + "grad_norm": 0.4412139654159546, + "learning_rate": 5.074028377544725e-06, + "loss": 4.2661, + "step": 329 + }, + { + "epoch": 0.010179333639227451, + "grad_norm": 0.5700720548629761, + "learning_rate": 5.089450956199877e-06, + "loss": 4.2581, + "step": 330 + }, + { + "epoch": 0.010210180104800867, + "grad_norm": 0.46195289492607117, + "learning_rate": 5.104873534855028e-06, + "loss": 4.2283, + "step": 331 + }, + { + "epoch": 0.010241026570374283, + "grad_norm": 0.4880393147468567, + "learning_rate": 5.1202961135101795e-06, + "loss": 4.308, + "step": 332 + }, + { + "epoch": 0.0102718730359477, + "grad_norm": 0.5247568488121033, + "learning_rate": 5.13571869216533e-06, + "loss": 4.3101, + "step": 333 + }, + { + "epoch": 0.010302719501521116, + "grad_norm": 0.46867644786834717, + "learning_rate": 5.1511412708204815e-06, + "loss": 4.2733, + "step": 334 + }, + { + "epoch": 0.010333565967094532, + "grad_norm": 0.5221483707427979, + "learning_rate": 5.166563849475633e-06, + "loss": 4.2343, + "step": 335 + }, + { + "epoch": 0.01036441243266795, + "grad_norm": 0.4837551414966583, + "learning_rate": 5.181986428130784e-06, + "loss": 4.2178, + "step": 336 + }, + { + "epoch": 0.010395258898241366, + "grad_norm": 0.5579172372817993, + "learning_rate": 5.197409006785935e-06, + "loss": 4.2333, + "step": 337 + }, + { + "epoch": 0.010426105363814782, + "grad_norm": 0.4567427635192871, + "learning_rate": 5.212831585441086e-06, + "loss": 4.239, + "step": 338 + }, + { + "epoch": 0.0104569518293882, + "grad_norm": 0.5036153793334961, + "learning_rate": 5.228254164096237e-06, + "loss": 4.243, + "step": 339 + }, + { + "epoch": 0.010487798294961615, + "grad_norm": 0.4032018184661865, + "learning_rate": 5.2436767427513885e-06, + "loss": 4.2381, + "step": 340 + }, + { + "epoch": 0.010518644760535031, + "grad_norm": 0.4462509751319885, + "learning_rate": 5.25909932140654e-06, + "loss": 4.2055, + "step": 341 + }, + { + "epoch": 0.010549491226108449, + "grad_norm": 0.39857831597328186, + "learning_rate": 5.2745219000616906e-06, + "loss": 4.2427, + "step": 342 + }, + { + "epoch": 0.010580337691681865, + "grad_norm": 0.5531516671180725, + "learning_rate": 5.289944478716842e-06, + "loss": 4.2291, + "step": 343 + }, + { + "epoch": 0.010611184157255281, + "grad_norm": 0.40396034717559814, + "learning_rate": 5.305367057371993e-06, + "loss": 4.2352, + "step": 344 + }, + { + "epoch": 0.010642030622828699, + "grad_norm": 0.45114099979400635, + "learning_rate": 5.320789636027144e-06, + "loss": 4.1742, + "step": 345 + }, + { + "epoch": 0.010672877088402115, + "grad_norm": 0.40345871448516846, + "learning_rate": 5.336212214682295e-06, + "loss": 4.2107, + "step": 346 + }, + { + "epoch": 0.01070372355397553, + "grad_norm": 0.47255831956863403, + "learning_rate": 5.351634793337446e-06, + "loss": 4.1835, + "step": 347 + }, + { + "epoch": 0.010734570019548948, + "grad_norm": 0.39346346259117126, + "learning_rate": 5.3670573719925976e-06, + "loss": 4.2179, + "step": 348 + }, + { + "epoch": 0.010765416485122364, + "grad_norm": 0.4031015634536743, + "learning_rate": 5.382479950647749e-06, + "loss": 4.1866, + "step": 349 + }, + { + "epoch": 0.01079626295069578, + "grad_norm": 0.43677273392677307, + "learning_rate": 5.3979025293029e-06, + "loss": 4.1982, + "step": 350 + }, + { + "epoch": 0.010827109416269198, + "grad_norm": 0.35807621479034424, + "learning_rate": 5.413325107958051e-06, + "loss": 4.2293, + "step": 351 + }, + { + "epoch": 0.010857955881842614, + "grad_norm": 0.3888252079486847, + "learning_rate": 5.428747686613202e-06, + "loss": 4.2068, + "step": 352 + }, + { + "epoch": 0.01088880234741603, + "grad_norm": 0.3722546696662903, + "learning_rate": 5.444170265268353e-06, + "loss": 4.1939, + "step": 353 + }, + { + "epoch": 0.010919648812989447, + "grad_norm": 0.43199339509010315, + "learning_rate": 5.4595928439235045e-06, + "loss": 4.2083, + "step": 354 + }, + { + "epoch": 0.010950495278562863, + "grad_norm": 0.40493977069854736, + "learning_rate": 5.475015422578655e-06, + "loss": 4.1939, + "step": 355 + }, + { + "epoch": 0.010981341744136279, + "grad_norm": 0.44482532143592834, + "learning_rate": 5.490438001233807e-06, + "loss": 4.2286, + "step": 356 + }, + { + "epoch": 0.011012188209709697, + "grad_norm": 0.420952171087265, + "learning_rate": 5.505860579888958e-06, + "loss": 4.1903, + "step": 357 + }, + { + "epoch": 0.011043034675283113, + "grad_norm": 0.42233625054359436, + "learning_rate": 5.5212831585441095e-06, + "loss": 4.1995, + "step": 358 + }, + { + "epoch": 0.011073881140856529, + "grad_norm": 0.3894343972206116, + "learning_rate": 5.536705737199259e-06, + "loss": 4.2041, + "step": 359 + }, + { + "epoch": 0.011104727606429946, + "grad_norm": 0.3671377897262573, + "learning_rate": 5.552128315854411e-06, + "loss": 4.1937, + "step": 360 + }, + { + "epoch": 0.011135574072003362, + "grad_norm": 0.43794602155685425, + "learning_rate": 5.567550894509562e-06, + "loss": 4.2181, + "step": 361 + }, + { + "epoch": 0.011166420537576778, + "grad_norm": 0.359497994184494, + "learning_rate": 5.5829734731647136e-06, + "loss": 4.1524, + "step": 362 + }, + { + "epoch": 0.011197267003150196, + "grad_norm": 0.3978222608566284, + "learning_rate": 5.598396051819864e-06, + "loss": 4.1848, + "step": 363 + }, + { + "epoch": 0.011228113468723612, + "grad_norm": 0.40741071105003357, + "learning_rate": 5.613818630475016e-06, + "loss": 4.2073, + "step": 364 + }, + { + "epoch": 0.011258959934297028, + "grad_norm": 0.3350363075733185, + "learning_rate": 5.629241209130167e-06, + "loss": 4.1895, + "step": 365 + }, + { + "epoch": 0.011289806399870445, + "grad_norm": 0.395192950963974, + "learning_rate": 5.6446637877853185e-06, + "loss": 4.1645, + "step": 366 + }, + { + "epoch": 0.011320652865443861, + "grad_norm": 0.3470457196235657, + "learning_rate": 5.660086366440469e-06, + "loss": 4.1806, + "step": 367 + }, + { + "epoch": 0.011351499331017277, + "grad_norm": 0.4110138416290283, + "learning_rate": 5.67550894509562e-06, + "loss": 4.1713, + "step": 368 + }, + { + "epoch": 0.011382345796590695, + "grad_norm": 0.40001338720321655, + "learning_rate": 5.690931523750771e-06, + "loss": 4.1576, + "step": 369 + }, + { + "epoch": 0.01141319226216411, + "grad_norm": 0.3540880084037781, + "learning_rate": 5.706354102405923e-06, + "loss": 4.1829, + "step": 370 + }, + { + "epoch": 0.011444038727737527, + "grad_norm": 0.38489672541618347, + "learning_rate": 5.721776681061074e-06, + "loss": 4.1868, + "step": 371 + }, + { + "epoch": 0.011474885193310944, + "grad_norm": 0.3404908776283264, + "learning_rate": 5.737199259716225e-06, + "loss": 4.1566, + "step": 372 + }, + { + "epoch": 0.01150573165888436, + "grad_norm": 0.38503584265708923, + "learning_rate": 5.752621838371376e-06, + "loss": 4.1382, + "step": 373 + }, + { + "epoch": 0.011536578124457776, + "grad_norm": 0.3913092017173767, + "learning_rate": 5.7680444170265275e-06, + "loss": 4.169, + "step": 374 + }, + { + "epoch": 0.011567424590031194, + "grad_norm": 0.41242536902427673, + "learning_rate": 5.783466995681678e-06, + "loss": 4.1759, + "step": 375 + }, + { + "epoch": 0.01159827105560461, + "grad_norm": 0.41931062936782837, + "learning_rate": 5.79888957433683e-06, + "loss": 4.1612, + "step": 376 + }, + { + "epoch": 0.011629117521178026, + "grad_norm": 0.3932725787162781, + "learning_rate": 5.81431215299198e-06, + "loss": 4.1423, + "step": 377 + }, + { + "epoch": 0.011659963986751443, + "grad_norm": 0.40188419818878174, + "learning_rate": 5.829734731647132e-06, + "loss": 4.1382, + "step": 378 + }, + { + "epoch": 0.01169081045232486, + "grad_norm": 0.43479123711586, + "learning_rate": 5.845157310302283e-06, + "loss": 4.2059, + "step": 379 + }, + { + "epoch": 0.011721656917898277, + "grad_norm": 0.4158079922199249, + "learning_rate": 5.8605798889574345e-06, + "loss": 4.1977, + "step": 380 + }, + { + "epoch": 0.011752503383471693, + "grad_norm": 0.3654467463493347, + "learning_rate": 5.876002467612585e-06, + "loss": 4.1619, + "step": 381 + }, + { + "epoch": 0.011783349849045109, + "grad_norm": 0.48977863788604736, + "learning_rate": 5.8914250462677366e-06, + "loss": 4.1582, + "step": 382 + }, + { + "epoch": 0.011814196314618526, + "grad_norm": 0.3477798104286194, + "learning_rate": 5.906847624922887e-06, + "loss": 4.1281, + "step": 383 + }, + { + "epoch": 0.011845042780191942, + "grad_norm": 0.3933609127998352, + "learning_rate": 5.922270203578039e-06, + "loss": 4.1433, + "step": 384 + }, + { + "epoch": 0.011875889245765358, + "grad_norm": 0.41575169563293457, + "learning_rate": 5.937692782233189e-06, + "loss": 4.1331, + "step": 385 + }, + { + "epoch": 0.011906735711338776, + "grad_norm": 0.37122276425361633, + "learning_rate": 5.953115360888341e-06, + "loss": 4.1138, + "step": 386 + }, + { + "epoch": 0.011937582176912192, + "grad_norm": 0.36738064885139465, + "learning_rate": 5.968537939543492e-06, + "loss": 4.1464, + "step": 387 + }, + { + "epoch": 0.011968428642485608, + "grad_norm": 0.36541691422462463, + "learning_rate": 5.9839605181986436e-06, + "loss": 4.1188, + "step": 388 + }, + { + "epoch": 0.011999275108059025, + "grad_norm": 0.38406893610954285, + "learning_rate": 5.999383096853794e-06, + "loss": 4.1371, + "step": 389 + }, + { + "epoch": 0.012030121573632441, + "grad_norm": 0.35382723808288574, + "learning_rate": 6.014805675508945e-06, + "loss": 4.1478, + "step": 390 + }, + { + "epoch": 0.012060968039205857, + "grad_norm": 0.8129196166992188, + "learning_rate": 6.030228254164096e-06, + "loss": 4.1425, + "step": 391 + }, + { + "epoch": 0.012091814504779275, + "grad_norm": 0.34851983189582825, + "learning_rate": 6.045650832819248e-06, + "loss": 4.1256, + "step": 392 + }, + { + "epoch": 0.012122660970352691, + "grad_norm": 0.3599962294101715, + "learning_rate": 6.061073411474399e-06, + "loss": 4.157, + "step": 393 + }, + { + "epoch": 0.012153507435926107, + "grad_norm": 0.3392893970012665, + "learning_rate": 6.07649599012955e-06, + "loss": 4.131, + "step": 394 + }, + { + "epoch": 0.012184353901499525, + "grad_norm": 0.3513313829898834, + "learning_rate": 6.091918568784701e-06, + "loss": 4.1572, + "step": 395 + }, + { + "epoch": 0.01221520036707294, + "grad_norm": 0.3460153043270111, + "learning_rate": 6.107341147439853e-06, + "loss": 4.1571, + "step": 396 + }, + { + "epoch": 0.012246046832646356, + "grad_norm": 0.3889797329902649, + "learning_rate": 6.122763726095004e-06, + "loss": 4.1073, + "step": 397 + }, + { + "epoch": 0.012276893298219774, + "grad_norm": 0.35544922947883606, + "learning_rate": 6.138186304750154e-06, + "loss": 4.1416, + "step": 398 + }, + { + "epoch": 0.01230773976379319, + "grad_norm": 0.3499937355518341, + "learning_rate": 6.153608883405305e-06, + "loss": 4.1378, + "step": 399 + }, + { + "epoch": 0.012338586229366606, + "grad_norm": 0.3560405373573303, + "learning_rate": 6.169031462060457e-06, + "loss": 4.1063, + "step": 400 + }, + { + "epoch": 0.012369432694940024, + "grad_norm": 0.3555552065372467, + "learning_rate": 6.184454040715608e-06, + "loss": 4.0931, + "step": 401 + }, + { + "epoch": 0.01240027916051344, + "grad_norm": 0.8508663177490234, + "learning_rate": 6.199876619370759e-06, + "loss": 4.1245, + "step": 402 + }, + { + "epoch": 0.012431125626086855, + "grad_norm": 0.3405144214630127, + "learning_rate": 6.21529919802591e-06, + "loss": 4.1404, + "step": 403 + }, + { + "epoch": 0.012461972091660273, + "grad_norm": 0.3584420382976532, + "learning_rate": 6.230721776681062e-06, + "loss": 4.1238, + "step": 404 + }, + { + "epoch": 0.012492818557233689, + "grad_norm": 0.34069547057151794, + "learning_rate": 6.246144355336213e-06, + "loss": 4.1214, + "step": 405 + }, + { + "epoch": 0.012523665022807105, + "grad_norm": 0.3547484576702118, + "learning_rate": 6.261566933991364e-06, + "loss": 4.1331, + "step": 406 + }, + { + "epoch": 0.012554511488380523, + "grad_norm": 0.34348946809768677, + "learning_rate": 6.276989512646515e-06, + "loss": 4.0785, + "step": 407 + }, + { + "epoch": 0.012585357953953939, + "grad_norm": 0.3440205752849579, + "learning_rate": 6.2924120913016666e-06, + "loss": 4.1196, + "step": 408 + }, + { + "epoch": 0.012616204419527354, + "grad_norm": 0.34636321663856506, + "learning_rate": 6.307834669956818e-06, + "loss": 4.1319, + "step": 409 + }, + { + "epoch": 0.012647050885100772, + "grad_norm": 0.36342504620552063, + "learning_rate": 6.323257248611968e-06, + "loss": 4.1306, + "step": 410 + }, + { + "epoch": 0.012677897350674188, + "grad_norm": 0.37424859404563904, + "learning_rate": 6.338679827267119e-06, + "loss": 4.0613, + "step": 411 + }, + { + "epoch": 0.012708743816247604, + "grad_norm": 0.35635513067245483, + "learning_rate": 6.35410240592227e-06, + "loss": 4.09, + "step": 412 + }, + { + "epoch": 0.012739590281821022, + "grad_norm": 0.3671311140060425, + "learning_rate": 6.369524984577421e-06, + "loss": 4.1319, + "step": 413 + }, + { + "epoch": 0.012770436747394438, + "grad_norm": 0.3544754683971405, + "learning_rate": 6.384947563232573e-06, + "loss": 4.1545, + "step": 414 + }, + { + "epoch": 0.012801283212967854, + "grad_norm": 0.43713635206222534, + "learning_rate": 6.400370141887724e-06, + "loss": 4.0801, + "step": 415 + }, + { + "epoch": 0.012832129678541271, + "grad_norm": 0.3339393138885498, + "learning_rate": 6.415792720542875e-06, + "loss": 4.0721, + "step": 416 + }, + { + "epoch": 0.012862976144114687, + "grad_norm": 0.3465556204319, + "learning_rate": 6.431215299198026e-06, + "loss": 4.1107, + "step": 417 + }, + { + "epoch": 0.012893822609688103, + "grad_norm": 0.32445189356803894, + "learning_rate": 6.446637877853178e-06, + "loss": 4.0729, + "step": 418 + }, + { + "epoch": 0.01292466907526152, + "grad_norm": 0.3739950954914093, + "learning_rate": 6.462060456508329e-06, + "loss": 4.0985, + "step": 419 + }, + { + "epoch": 0.012955515540834937, + "grad_norm": 0.35849428176879883, + "learning_rate": 6.47748303516348e-06, + "loss": 4.1012, + "step": 420 + }, + { + "epoch": 0.012986362006408353, + "grad_norm": 0.33864596486091614, + "learning_rate": 6.492905613818631e-06, + "loss": 4.064, + "step": 421 + }, + { + "epoch": 0.01301720847198177, + "grad_norm": 0.34660792350769043, + "learning_rate": 6.508328192473783e-06, + "loss": 4.1147, + "step": 422 + }, + { + "epoch": 0.013048054937555186, + "grad_norm": 0.31942033767700195, + "learning_rate": 6.523750771128934e-06, + "loss": 4.1377, + "step": 423 + }, + { + "epoch": 0.013078901403128602, + "grad_norm": 0.34794819355010986, + "learning_rate": 6.539173349784085e-06, + "loss": 4.0791, + "step": 424 + }, + { + "epoch": 0.01310974786870202, + "grad_norm": 0.345765084028244, + "learning_rate": 6.554595928439234e-06, + "loss": 4.0743, + "step": 425 + }, + { + "epoch": 0.013140594334275436, + "grad_norm": 0.34186556935310364, + "learning_rate": 6.570018507094386e-06, + "loss": 4.0727, + "step": 426 + }, + { + "epoch": 0.013171440799848852, + "grad_norm": 0.29015296697616577, + "learning_rate": 6.585441085749537e-06, + "loss": 4.0795, + "step": 427 + }, + { + "epoch": 0.01320228726542227, + "grad_norm": 0.841758668422699, + "learning_rate": 6.600863664404689e-06, + "loss": 4.08, + "step": 428 + }, + { + "epoch": 0.013233133730995685, + "grad_norm": 0.32446324825286865, + "learning_rate": 6.616286243059839e-06, + "loss": 4.0773, + "step": 429 + }, + { + "epoch": 0.013263980196569101, + "grad_norm": 0.31437793374061584, + "learning_rate": 6.631708821714991e-06, + "loss": 4.0991, + "step": 430 + }, + { + "epoch": 0.013294826662142519, + "grad_norm": 0.8302384614944458, + "learning_rate": 6.647131400370142e-06, + "loss": 4.0766, + "step": 431 + }, + { + "epoch": 0.013325673127715935, + "grad_norm": 0.35288137197494507, + "learning_rate": 6.662553979025294e-06, + "loss": 4.0918, + "step": 432 + }, + { + "epoch": 0.01335651959328935, + "grad_norm": 0.36966627836227417, + "learning_rate": 6.677976557680444e-06, + "loss": 4.067, + "step": 433 + }, + { + "epoch": 0.013387366058862768, + "grad_norm": 0.3782534897327423, + "learning_rate": 6.693399136335596e-06, + "loss": 4.1021, + "step": 434 + }, + { + "epoch": 0.013418212524436184, + "grad_norm": 0.3488578498363495, + "learning_rate": 6.708821714990747e-06, + "loss": 4.083, + "step": 435 + }, + { + "epoch": 0.0134490589900096, + "grad_norm": 0.3412225544452667, + "learning_rate": 6.724244293645899e-06, + "loss": 4.0543, + "step": 436 + }, + { + "epoch": 0.013479905455583018, + "grad_norm": 0.35565343499183655, + "learning_rate": 6.739666872301049e-06, + "loss": 4.0351, + "step": 437 + }, + { + "epoch": 0.013510751921156434, + "grad_norm": 0.3810979127883911, + "learning_rate": 6.755089450956201e-06, + "loss": 4.0696, + "step": 438 + }, + { + "epoch": 0.01354159838672985, + "grad_norm": 0.3236299157142639, + "learning_rate": 6.770512029611352e-06, + "loss": 4.0607, + "step": 439 + }, + { + "epoch": 0.013572444852303267, + "grad_norm": 0.38692525029182434, + "learning_rate": 6.785934608266502e-06, + "loss": 4.0716, + "step": 440 + }, + { + "epoch": 0.013603291317876683, + "grad_norm": 0.3023357689380646, + "learning_rate": 6.801357186921653e-06, + "loss": 4.0773, + "step": 441 + }, + { + "epoch": 0.0136341377834501, + "grad_norm": 0.39024314284324646, + "learning_rate": 6.816779765576804e-06, + "loss": 4.0971, + "step": 442 + }, + { + "epoch": 0.013664984249023517, + "grad_norm": 0.33914390206336975, + "learning_rate": 6.832202344231955e-06, + "loss": 4.0873, + "step": 443 + }, + { + "epoch": 0.013695830714596933, + "grad_norm": 0.5535599589347839, + "learning_rate": 6.847624922887107e-06, + "loss": 4.0982, + "step": 444 + }, + { + "epoch": 0.013726677180170349, + "grad_norm": 0.3078445792198181, + "learning_rate": 6.863047501542258e-06, + "loss": 4.0287, + "step": 445 + }, + { + "epoch": 0.013757523645743766, + "grad_norm": 0.34544292092323303, + "learning_rate": 6.878470080197409e-06, + "loss": 4.0832, + "step": 446 + }, + { + "epoch": 0.013788370111317182, + "grad_norm": 0.29909199476242065, + "learning_rate": 6.89389265885256e-06, + "loss": 4.0461, + "step": 447 + }, + { + "epoch": 0.0138192165768906, + "grad_norm": 0.3227236866950989, + "learning_rate": 6.909315237507712e-06, + "loss": 4.0703, + "step": 448 + }, + { + "epoch": 0.013850063042464016, + "grad_norm": 0.33257341384887695, + "learning_rate": 6.924737816162863e-06, + "loss": 4.0686, + "step": 449 + }, + { + "epoch": 0.013880909508037432, + "grad_norm": 0.30151569843292236, + "learning_rate": 6.940160394818014e-06, + "loss": 4.0568, + "step": 450 + }, + { + "epoch": 0.01391175597361085, + "grad_norm": 0.32288119196891785, + "learning_rate": 6.955582973473165e-06, + "loss": 4.0762, + "step": 451 + }, + { + "epoch": 0.013942602439184265, + "grad_norm": 0.3460620045661926, + "learning_rate": 6.971005552128317e-06, + "loss": 4.0301, + "step": 452 + }, + { + "epoch": 0.013973448904757681, + "grad_norm": 0.34772971272468567, + "learning_rate": 6.986428130783468e-06, + "loss": 4.0689, + "step": 453 + }, + { + "epoch": 0.014004295370331099, + "grad_norm": 0.37752974033355713, + "learning_rate": 7.001850709438619e-06, + "loss": 4.059, + "step": 454 + }, + { + "epoch": 0.014035141835904515, + "grad_norm": 0.3308340013027191, + "learning_rate": 7.01727328809377e-06, + "loss": 4.0559, + "step": 455 + }, + { + "epoch": 0.014065988301477931, + "grad_norm": 0.3896198570728302, + "learning_rate": 7.03269586674892e-06, + "loss": 4.0632, + "step": 456 + }, + { + "epoch": 0.014096834767051349, + "grad_norm": 0.37239792943000793, + "learning_rate": 7.048118445404071e-06, + "loss": 4.0225, + "step": 457 + }, + { + "epoch": 0.014127681232624765, + "grad_norm": 0.3392679989337921, + "learning_rate": 7.063541024059223e-06, + "loss": 4.0473, + "step": 458 + }, + { + "epoch": 0.01415852769819818, + "grad_norm": 0.542512059211731, + "learning_rate": 7.078963602714373e-06, + "loss": 4.0829, + "step": 459 + }, + { + "epoch": 0.014189374163771598, + "grad_norm": 0.3591475188732147, + "learning_rate": 7.094386181369525e-06, + "loss": 4.0489, + "step": 460 + }, + { + "epoch": 0.014220220629345014, + "grad_norm": 0.32008180022239685, + "learning_rate": 7.109808760024676e-06, + "loss": 4.0366, + "step": 461 + }, + { + "epoch": 0.01425106709491843, + "grad_norm": 0.35058602690696716, + "learning_rate": 7.125231338679828e-06, + "loss": 3.9878, + "step": 462 + }, + { + "epoch": 0.014281913560491848, + "grad_norm": 0.3103954792022705, + "learning_rate": 7.140653917334978e-06, + "loss": 4.0255, + "step": 463 + }, + { + "epoch": 0.014312760026065264, + "grad_norm": 0.34683337807655334, + "learning_rate": 7.15607649599013e-06, + "loss": 4.0328, + "step": 464 + }, + { + "epoch": 0.01434360649163868, + "grad_norm": 0.3370136320590973, + "learning_rate": 7.171499074645281e-06, + "loss": 4.022, + "step": 465 + }, + { + "epoch": 0.014374452957212097, + "grad_norm": 0.3209238648414612, + "learning_rate": 7.186921653300433e-06, + "loss": 4.0476, + "step": 466 + }, + { + "epoch": 0.014405299422785513, + "grad_norm": 0.36398598551750183, + "learning_rate": 7.202344231955583e-06, + "loss": 4.0605, + "step": 467 + }, + { + "epoch": 0.014436145888358929, + "grad_norm": 0.2936258018016815, + "learning_rate": 7.217766810610735e-06, + "loss": 4.0201, + "step": 468 + }, + { + "epoch": 0.014466992353932347, + "grad_norm": 0.3543979227542877, + "learning_rate": 7.233189389265886e-06, + "loss": 4.0646, + "step": 469 + }, + { + "epoch": 0.014497838819505763, + "grad_norm": 0.3126312494277954, + "learning_rate": 7.248611967921038e-06, + "loss": 3.9785, + "step": 470 + }, + { + "epoch": 0.014528685285079179, + "grad_norm": 0.3114073872566223, + "learning_rate": 7.264034546576187e-06, + "loss": 4.0503, + "step": 471 + }, + { + "epoch": 0.014559531750652596, + "grad_norm": 0.3436780571937561, + "learning_rate": 7.279457125231339e-06, + "loss": 4.027, + "step": 472 + }, + { + "epoch": 0.014590378216226012, + "grad_norm": 0.2992337942123413, + "learning_rate": 7.294879703886489e-06, + "loss": 4.0518, + "step": 473 + }, + { + "epoch": 0.014621224681799428, + "grad_norm": 0.3278312683105469, + "learning_rate": 7.310302282541641e-06, + "loss": 4.0048, + "step": 474 + }, + { + "epoch": 0.014652071147372846, + "grad_norm": 0.2938264012336731, + "learning_rate": 7.325724861196792e-06, + "loss": 4.0572, + "step": 475 + }, + { + "epoch": 0.014682917612946262, + "grad_norm": 0.3216952085494995, + "learning_rate": 7.341147439851944e-06, + "loss": 4.0103, + "step": 476 + }, + { + "epoch": 0.014713764078519678, + "grad_norm": 0.6791350841522217, + "learning_rate": 7.356570018507094e-06, + "loss": 3.9776, + "step": 477 + }, + { + "epoch": 0.014744610544093095, + "grad_norm": 0.28557831048965454, + "learning_rate": 7.371992597162246e-06, + "loss": 4.0484, + "step": 478 + }, + { + "epoch": 0.014775457009666511, + "grad_norm": 0.3088971674442291, + "learning_rate": 7.387415175817397e-06, + "loss": 4.035, + "step": 479 + }, + { + "epoch": 0.014806303475239927, + "grad_norm": 0.31441935896873474, + "learning_rate": 7.402837754472549e-06, + "loss": 4.0746, + "step": 480 + }, + { + "epoch": 0.014837149940813345, + "grad_norm": 0.2995738685131073, + "learning_rate": 7.418260333127699e-06, + "loss": 4.0112, + "step": 481 + }, + { + "epoch": 0.01486799640638676, + "grad_norm": 0.323941707611084, + "learning_rate": 7.433682911782851e-06, + "loss": 3.9873, + "step": 482 + }, + { + "epoch": 0.014898842871960177, + "grad_norm": 0.31433212757110596, + "learning_rate": 7.449105490438002e-06, + "loss": 4.0311, + "step": 483 + }, + { + "epoch": 0.014929689337533594, + "grad_norm": 0.31832069158554077, + "learning_rate": 7.464528069093154e-06, + "loss": 4.0199, + "step": 484 + }, + { + "epoch": 0.01496053580310701, + "grad_norm": 0.31982940435409546, + "learning_rate": 7.479950647748304e-06, + "loss": 3.9888, + "step": 485 + }, + { + "epoch": 0.014991382268680426, + "grad_norm": 0.33108749985694885, + "learning_rate": 7.495373226403456e-06, + "loss": 4.0261, + "step": 486 + }, + { + "epoch": 0.015022228734253844, + "grad_norm": 0.339418888092041, + "learning_rate": 7.5107958050586054e-06, + "loss": 3.9838, + "step": 487 + }, + { + "epoch": 0.01505307519982726, + "grad_norm": 0.32247382402420044, + "learning_rate": 7.526218383713757e-06, + "loss": 4.0081, + "step": 488 + }, + { + "epoch": 0.015083921665400676, + "grad_norm": 0.36072561144828796, + "learning_rate": 7.541640962368908e-06, + "loss": 4.0195, + "step": 489 + }, + { + "epoch": 0.015114768130974093, + "grad_norm": 0.39254429936408997, + "learning_rate": 7.557063541024059e-06, + "loss": 3.9788, + "step": 490 + }, + { + "epoch": 0.01514561459654751, + "grad_norm": 0.4011254906654358, + "learning_rate": 7.57248611967921e-06, + "loss": 3.9898, + "step": 491 + }, + { + "epoch": 0.015176461062120925, + "grad_norm": 0.3294040858745575, + "learning_rate": 7.587908698334362e-06, + "loss": 3.9981, + "step": 492 + }, + { + "epoch": 0.015207307527694343, + "grad_norm": 0.35888224840164185, + "learning_rate": 7.603331276989513e-06, + "loss": 3.9913, + "step": 493 + }, + { + "epoch": 0.015238153993267759, + "grad_norm": 0.32174989581108093, + "learning_rate": 7.618753855644664e-06, + "loss": 3.9653, + "step": 494 + }, + { + "epoch": 0.015269000458841175, + "grad_norm": 0.3233802616596222, + "learning_rate": 7.634176434299816e-06, + "loss": 4.0182, + "step": 495 + }, + { + "epoch": 0.015299846924414592, + "grad_norm": 0.3296842575073242, + "learning_rate": 7.649599012954967e-06, + "loss": 3.9836, + "step": 496 + }, + { + "epoch": 0.015330693389988008, + "grad_norm": 0.33686113357543945, + "learning_rate": 7.665021591610117e-06, + "loss": 3.9973, + "step": 497 + }, + { + "epoch": 0.015361539855561424, + "grad_norm": 0.33309200406074524, + "learning_rate": 7.68044417026527e-06, + "loss": 4.0077, + "step": 498 + }, + { + "epoch": 0.015392386321134842, + "grad_norm": 0.3097420930862427, + "learning_rate": 7.69586674892042e-06, + "loss": 4.0056, + "step": 499 + }, + { + "epoch": 0.015423232786708258, + "grad_norm": 0.6101787090301514, + "learning_rate": 7.711289327575571e-06, + "loss": 3.9681, + "step": 500 + }, + { + "epoch": 0.015454079252281674, + "grad_norm": 0.3292408585548401, + "learning_rate": 7.726711906230723e-06, + "loss": 3.9763, + "step": 501 + }, + { + "epoch": 0.015484925717855091, + "grad_norm": 0.3890770971775055, + "learning_rate": 7.742134484885872e-06, + "loss": 4.0064, + "step": 502 + }, + { + "epoch": 0.015515772183428507, + "grad_norm": 0.3332178592681885, + "learning_rate": 7.757557063541024e-06, + "loss": 4.0235, + "step": 503 + }, + { + "epoch": 0.015546618649001923, + "grad_norm": 0.37621328234672546, + "learning_rate": 7.772979642196175e-06, + "loss": 4.0219, + "step": 504 + }, + { + "epoch": 0.015577465114575341, + "grad_norm": 0.35572606325149536, + "learning_rate": 7.788402220851326e-06, + "loss": 3.9833, + "step": 505 + }, + { + "epoch": 0.015608311580148757, + "grad_norm": 0.30865949392318726, + "learning_rate": 7.803824799506478e-06, + "loss": 4.0145, + "step": 506 + }, + { + "epoch": 0.015639158045722173, + "grad_norm": 0.38788342475891113, + "learning_rate": 7.819247378161628e-06, + "loss": 4.0001, + "step": 507 + }, + { + "epoch": 0.01567000451129559, + "grad_norm": 0.29719364643096924, + "learning_rate": 7.83466995681678e-06, + "loss": 4.0195, + "step": 508 + }, + { + "epoch": 0.015700850976869008, + "grad_norm": 0.3571527302265167, + "learning_rate": 7.850092535471931e-06, + "loss": 4.028, + "step": 509 + }, + { + "epoch": 0.015731697442442422, + "grad_norm": 0.36732667684555054, + "learning_rate": 7.865515114127082e-06, + "loss": 3.996, + "step": 510 + }, + { + "epoch": 0.01576254390801584, + "grad_norm": 0.31795406341552734, + "learning_rate": 7.880937692782234e-06, + "loss": 4.027, + "step": 511 + }, + { + "epoch": 0.015793390373589258, + "grad_norm": 0.39692309498786926, + "learning_rate": 7.896360271437385e-06, + "loss": 3.9967, + "step": 512 + }, + { + "epoch": 0.015824236839162672, + "grad_norm": 0.3340602517127991, + "learning_rate": 7.911782850092535e-06, + "loss": 3.962, + "step": 513 + }, + { + "epoch": 0.01585508330473609, + "grad_norm": 0.36588260531425476, + "learning_rate": 7.927205428747688e-06, + "loss": 3.9514, + "step": 514 + }, + { + "epoch": 0.015885929770309507, + "grad_norm": 0.33053624629974365, + "learning_rate": 7.942628007402838e-06, + "loss": 3.9997, + "step": 515 + }, + { + "epoch": 0.01591677623588292, + "grad_norm": 0.3271799683570862, + "learning_rate": 7.95805058605799e-06, + "loss": 3.9924, + "step": 516 + }, + { + "epoch": 0.01594762270145634, + "grad_norm": 0.3428313434123993, + "learning_rate": 7.973473164713141e-06, + "loss": 3.972, + "step": 517 + }, + { + "epoch": 0.015978469167029757, + "grad_norm": 0.35157400369644165, + "learning_rate": 7.98889574336829e-06, + "loss": 3.9693, + "step": 518 + }, + { + "epoch": 0.01600931563260317, + "grad_norm": 0.33497878909111023, + "learning_rate": 8.004318322023442e-06, + "loss": 3.972, + "step": 519 + }, + { + "epoch": 0.01604016209817659, + "grad_norm": 0.31749898195266724, + "learning_rate": 8.019740900678593e-06, + "loss": 3.9658, + "step": 520 + }, + { + "epoch": 0.016071008563750006, + "grad_norm": 0.36618101596832275, + "learning_rate": 8.035163479333745e-06, + "loss": 3.9839, + "step": 521 + }, + { + "epoch": 0.01610185502932342, + "grad_norm": 0.3277404308319092, + "learning_rate": 8.050586057988896e-06, + "loss": 3.9642, + "step": 522 + }, + { + "epoch": 0.016132701494896838, + "grad_norm": 0.35166555643081665, + "learning_rate": 8.066008636644047e-06, + "loss": 4.0083, + "step": 523 + }, + { + "epoch": 0.016163547960470256, + "grad_norm": 0.32125118374824524, + "learning_rate": 8.081431215299199e-06, + "loss": 4.0143, + "step": 524 + }, + { + "epoch": 0.01619439442604367, + "grad_norm": 0.3105856776237488, + "learning_rate": 8.09685379395435e-06, + "loss": 3.9787, + "step": 525 + }, + { + "epoch": 0.016225240891617088, + "grad_norm": 0.32057681679725647, + "learning_rate": 8.1122763726095e-06, + "loss": 3.9371, + "step": 526 + }, + { + "epoch": 0.016256087357190505, + "grad_norm": 0.2897120714187622, + "learning_rate": 8.127698951264652e-06, + "loss": 3.991, + "step": 527 + }, + { + "epoch": 0.01628693382276392, + "grad_norm": 0.4307211637496948, + "learning_rate": 8.143121529919803e-06, + "loss": 4.0334, + "step": 528 + }, + { + "epoch": 0.016317780288337337, + "grad_norm": 0.30313020944595337, + "learning_rate": 8.158544108574955e-06, + "loss": 4.0073, + "step": 529 + }, + { + "epoch": 0.016348626753910755, + "grad_norm": 0.3366444706916809, + "learning_rate": 8.173966687230106e-06, + "loss": 3.964, + "step": 530 + }, + { + "epoch": 0.01637947321948417, + "grad_norm": 0.3471347391605377, + "learning_rate": 8.189389265885256e-06, + "loss": 3.9614, + "step": 531 + }, + { + "epoch": 0.016410319685057587, + "grad_norm": 0.34622722864151, + "learning_rate": 8.204811844540409e-06, + "loss": 3.9622, + "step": 532 + }, + { + "epoch": 0.016441166150631004, + "grad_norm": 0.39090725779533386, + "learning_rate": 8.220234423195558e-06, + "loss": 3.9922, + "step": 533 + }, + { + "epoch": 0.01647201261620442, + "grad_norm": 0.3334308862686157, + "learning_rate": 8.23565700185071e-06, + "loss": 3.981, + "step": 534 + }, + { + "epoch": 0.016502859081777836, + "grad_norm": 0.3182835280895233, + "learning_rate": 8.25107958050586e-06, + "loss": 3.9878, + "step": 535 + }, + { + "epoch": 0.016533705547351254, + "grad_norm": 0.341772198677063, + "learning_rate": 8.266502159161011e-06, + "loss": 3.9452, + "step": 536 + }, + { + "epoch": 0.016564552012924668, + "grad_norm": 0.32429179549217224, + "learning_rate": 8.281924737816163e-06, + "loss": 3.9664, + "step": 537 + }, + { + "epoch": 0.016595398478498086, + "grad_norm": 0.4072919189929962, + "learning_rate": 8.297347316471314e-06, + "loss": 3.9748, + "step": 538 + }, + { + "epoch": 0.016626244944071503, + "grad_norm": 0.35796207189559937, + "learning_rate": 8.312769895126466e-06, + "loss": 3.9905, + "step": 539 + }, + { + "epoch": 0.016657091409644918, + "grad_norm": 0.287631630897522, + "learning_rate": 8.328192473781617e-06, + "loss": 3.9545, + "step": 540 + }, + { + "epoch": 0.016687937875218335, + "grad_norm": 0.3627811670303345, + "learning_rate": 8.343615052436767e-06, + "loss": 3.9628, + "step": 541 + }, + { + "epoch": 0.016718784340791753, + "grad_norm": 0.33657750487327576, + "learning_rate": 8.35903763109192e-06, + "loss": 3.9744, + "step": 542 + }, + { + "epoch": 0.016749630806365167, + "grad_norm": 0.30754533410072327, + "learning_rate": 8.37446020974707e-06, + "loss": 3.9528, + "step": 543 + }, + { + "epoch": 0.016780477271938585, + "grad_norm": 0.3261829614639282, + "learning_rate": 8.389882788402221e-06, + "loss": 3.9674, + "step": 544 + }, + { + "epoch": 0.016811323737512002, + "grad_norm": 0.3323749005794525, + "learning_rate": 8.405305367057373e-06, + "loss": 3.9341, + "step": 545 + }, + { + "epoch": 0.016842170203085417, + "grad_norm": 0.3549530506134033, + "learning_rate": 8.420727945712524e-06, + "loss": 3.954, + "step": 546 + }, + { + "epoch": 0.016873016668658834, + "grad_norm": 0.3383442163467407, + "learning_rate": 8.436150524367674e-06, + "loss": 3.977, + "step": 547 + }, + { + "epoch": 0.016903863134232252, + "grad_norm": 0.7313940525054932, + "learning_rate": 8.451573103022827e-06, + "loss": 3.9477, + "step": 548 + }, + { + "epoch": 0.016934709599805666, + "grad_norm": 0.3055112957954407, + "learning_rate": 8.466995681677976e-06, + "loss": 3.9562, + "step": 549 + }, + { + "epoch": 0.016965556065379084, + "grad_norm": 0.39761391282081604, + "learning_rate": 8.482418260333128e-06, + "loss": 3.9381, + "step": 550 + }, + { + "epoch": 0.0169964025309525, + "grad_norm": 0.347799152135849, + "learning_rate": 8.497840838988279e-06, + "loss": 3.9632, + "step": 551 + }, + { + "epoch": 0.017027248996525916, + "grad_norm": 0.31029877066612244, + "learning_rate": 8.51326341764343e-06, + "loss": 3.94, + "step": 552 + }, + { + "epoch": 0.017058095462099333, + "grad_norm": 0.31401655077934265, + "learning_rate": 8.528685996298581e-06, + "loss": 3.9808, + "step": 553 + }, + { + "epoch": 0.01708894192767275, + "grad_norm": 0.28758901357650757, + "learning_rate": 8.544108574953732e-06, + "loss": 3.943, + "step": 554 + }, + { + "epoch": 0.017119788393246165, + "grad_norm": 0.31496456265449524, + "learning_rate": 8.559531153608884e-06, + "loss": 3.9921, + "step": 555 + }, + { + "epoch": 0.017150634858819583, + "grad_norm": 0.32990941405296326, + "learning_rate": 8.574953732264035e-06, + "loss": 3.9486, + "step": 556 + }, + { + "epoch": 0.017181481324393, + "grad_norm": 0.3185062110424042, + "learning_rate": 8.590376310919186e-06, + "loss": 3.9434, + "step": 557 + }, + { + "epoch": 0.017212327789966415, + "grad_norm": 0.31472209095954895, + "learning_rate": 8.605798889574338e-06, + "loss": 3.9202, + "step": 558 + }, + { + "epoch": 0.017243174255539832, + "grad_norm": 0.31313541531562805, + "learning_rate": 8.621221468229488e-06, + "loss": 3.8714, + "step": 559 + }, + { + "epoch": 0.01727402072111325, + "grad_norm": 0.34916338324546814, + "learning_rate": 8.63664404688464e-06, + "loss": 3.9243, + "step": 560 + }, + { + "epoch": 0.017304867186686664, + "grad_norm": 0.4029117524623871, + "learning_rate": 8.652066625539791e-06, + "loss": 3.9196, + "step": 561 + }, + { + "epoch": 0.017335713652260082, + "grad_norm": 0.299302339553833, + "learning_rate": 8.667489204194942e-06, + "loss": 3.9699, + "step": 562 + }, + { + "epoch": 0.0173665601178335, + "grad_norm": 0.3029027283191681, + "learning_rate": 8.682911782850094e-06, + "loss": 3.911, + "step": 563 + }, + { + "epoch": 0.017397406583406914, + "grad_norm": 0.33063191175460815, + "learning_rate": 8.698334361505243e-06, + "loss": 3.9544, + "step": 564 + }, + { + "epoch": 0.01742825304898033, + "grad_norm": 0.2955205738544464, + "learning_rate": 8.713756940160395e-06, + "loss": 3.9628, + "step": 565 + }, + { + "epoch": 0.01745909951455375, + "grad_norm": 0.3073122203350067, + "learning_rate": 8.729179518815546e-06, + "loss": 3.9507, + "step": 566 + }, + { + "epoch": 0.017489945980127163, + "grad_norm": 0.34223443269729614, + "learning_rate": 8.744602097470697e-06, + "loss": 3.9048, + "step": 567 + }, + { + "epoch": 0.01752079244570058, + "grad_norm": 0.3347774147987366, + "learning_rate": 8.760024676125849e-06, + "loss": 3.9179, + "step": 568 + }, + { + "epoch": 0.017551638911274, + "grad_norm": 0.6813507080078125, + "learning_rate": 8.775447254781e-06, + "loss": 3.9043, + "step": 569 + }, + { + "epoch": 0.017582485376847413, + "grad_norm": 0.3213602900505066, + "learning_rate": 8.79086983343615e-06, + "loss": 3.9097, + "step": 570 + }, + { + "epoch": 0.01761333184242083, + "grad_norm": 0.34326374530792236, + "learning_rate": 8.806292412091302e-06, + "loss": 3.9623, + "step": 571 + }, + { + "epoch": 0.017644178307994248, + "grad_norm": 0.3662053942680359, + "learning_rate": 8.821714990746453e-06, + "loss": 3.887, + "step": 572 + }, + { + "epoch": 0.017675024773567662, + "grad_norm": 0.33156415820121765, + "learning_rate": 8.837137569401605e-06, + "loss": 3.9543, + "step": 573 + }, + { + "epoch": 0.01770587123914108, + "grad_norm": 0.3644343614578247, + "learning_rate": 8.852560148056756e-06, + "loss": 3.9557, + "step": 574 + }, + { + "epoch": 0.017736717704714498, + "grad_norm": 0.3337916135787964, + "learning_rate": 8.867982726711906e-06, + "loss": 3.915, + "step": 575 + }, + { + "epoch": 0.017767564170287912, + "grad_norm": 0.37422192096710205, + "learning_rate": 8.883405305367059e-06, + "loss": 3.9023, + "step": 576 + }, + { + "epoch": 0.01779841063586133, + "grad_norm": 0.31736183166503906, + "learning_rate": 8.89882788402221e-06, + "loss": 3.9239, + "step": 577 + }, + { + "epoch": 0.017829257101434747, + "grad_norm": 0.34419509768486023, + "learning_rate": 8.91425046267736e-06, + "loss": 3.901, + "step": 578 + }, + { + "epoch": 0.01786010356700816, + "grad_norm": 0.3142429292201996, + "learning_rate": 8.92967304133251e-06, + "loss": 3.9351, + "step": 579 + }, + { + "epoch": 0.01789095003258158, + "grad_norm": 0.3003634810447693, + "learning_rate": 8.945095619987661e-06, + "loss": 3.9499, + "step": 580 + }, + { + "epoch": 0.017921796498154997, + "grad_norm": 0.3625619411468506, + "learning_rate": 8.960518198642813e-06, + "loss": 3.8841, + "step": 581 + }, + { + "epoch": 0.01795264296372841, + "grad_norm": 0.2891596257686615, + "learning_rate": 8.975940777297964e-06, + "loss": 3.9305, + "step": 582 + }, + { + "epoch": 0.01798348942930183, + "grad_norm": 0.34846970438957214, + "learning_rate": 8.991363355953115e-06, + "loss": 3.9013, + "step": 583 + }, + { + "epoch": 0.018014335894875246, + "grad_norm": 0.3184296786785126, + "learning_rate": 9.006785934608267e-06, + "loss": 3.8971, + "step": 584 + }, + { + "epoch": 0.01804518236044866, + "grad_norm": 0.34674254059791565, + "learning_rate": 9.022208513263418e-06, + "loss": 3.9698, + "step": 585 + }, + { + "epoch": 0.018076028826022078, + "grad_norm": 0.3140704333782196, + "learning_rate": 9.03763109191857e-06, + "loss": 3.9379, + "step": 586 + }, + { + "epoch": 0.018106875291595496, + "grad_norm": 0.31047022342681885, + "learning_rate": 9.05305367057372e-06, + "loss": 3.91, + "step": 587 + }, + { + "epoch": 0.01813772175716891, + "grad_norm": 0.3194608986377716, + "learning_rate": 9.068476249228871e-06, + "loss": 3.9092, + "step": 588 + }, + { + "epoch": 0.018168568222742328, + "grad_norm": 0.32686567306518555, + "learning_rate": 9.083898827884023e-06, + "loss": 3.8592, + "step": 589 + }, + { + "epoch": 0.018199414688315745, + "grad_norm": 0.3055470287799835, + "learning_rate": 9.099321406539174e-06, + "loss": 3.9232, + "step": 590 + }, + { + "epoch": 0.01823026115388916, + "grad_norm": 0.3288993835449219, + "learning_rate": 9.114743985194325e-06, + "loss": 3.9142, + "step": 591 + }, + { + "epoch": 0.018261107619462577, + "grad_norm": 0.29851123690605164, + "learning_rate": 9.130166563849477e-06, + "loss": 3.8743, + "step": 592 + }, + { + "epoch": 0.018291954085035995, + "grad_norm": 0.3596423864364624, + "learning_rate": 9.145589142504627e-06, + "loss": 3.9155, + "step": 593 + }, + { + "epoch": 0.01832280055060941, + "grad_norm": 0.3067793548107147, + "learning_rate": 9.16101172115978e-06, + "loss": 3.8864, + "step": 594 + }, + { + "epoch": 0.018353647016182827, + "grad_norm": 0.33642008900642395, + "learning_rate": 9.176434299814929e-06, + "loss": 3.9094, + "step": 595 + }, + { + "epoch": 0.018384493481756244, + "grad_norm": 0.32907581329345703, + "learning_rate": 9.191856878470081e-06, + "loss": 3.9049, + "step": 596 + }, + { + "epoch": 0.01841533994732966, + "grad_norm": 0.3088388741016388, + "learning_rate": 9.207279457125232e-06, + "loss": 3.871, + "step": 597 + }, + { + "epoch": 0.018446186412903076, + "grad_norm": 0.30994969606399536, + "learning_rate": 9.222702035780382e-06, + "loss": 3.9013, + "step": 598 + }, + { + "epoch": 0.018477032878476494, + "grad_norm": 0.2986358404159546, + "learning_rate": 9.238124614435534e-06, + "loss": 3.9002, + "step": 599 + }, + { + "epoch": 0.018507879344049908, + "grad_norm": 0.3357411026954651, + "learning_rate": 9.253547193090685e-06, + "loss": 3.9433, + "step": 600 + }, + { + "epoch": 0.018538725809623326, + "grad_norm": 0.326695054769516, + "learning_rate": 9.268969771745836e-06, + "loss": 3.8807, + "step": 601 + }, + { + "epoch": 0.018569572275196743, + "grad_norm": 0.3374023139476776, + "learning_rate": 9.284392350400988e-06, + "loss": 3.9321, + "step": 602 + }, + { + "epoch": 0.018600418740770158, + "grad_norm": 0.36673521995544434, + "learning_rate": 9.299814929056139e-06, + "loss": 3.9033, + "step": 603 + }, + { + "epoch": 0.018631265206343575, + "grad_norm": 0.3046558201313019, + "learning_rate": 9.315237507711289e-06, + "loss": 3.853, + "step": 604 + }, + { + "epoch": 0.018662111671916993, + "grad_norm": 0.8939979076385498, + "learning_rate": 9.330660086366441e-06, + "loss": 3.9437, + "step": 605 + }, + { + "epoch": 0.018692958137490407, + "grad_norm": 0.31382232904434204, + "learning_rate": 9.346082665021592e-06, + "loss": 3.929, + "step": 606 + }, + { + "epoch": 0.018723804603063825, + "grad_norm": 0.30473509430885315, + "learning_rate": 9.361505243676744e-06, + "loss": 3.9136, + "step": 607 + }, + { + "epoch": 0.018754651068637242, + "grad_norm": 0.313920795917511, + "learning_rate": 9.376927822331895e-06, + "loss": 3.8776, + "step": 608 + }, + { + "epoch": 0.018785497534210657, + "grad_norm": 0.30601373314857483, + "learning_rate": 9.392350400987046e-06, + "loss": 3.9318, + "step": 609 + }, + { + "epoch": 0.018816343999784074, + "grad_norm": 0.3050955832004547, + "learning_rate": 9.407772979642196e-06, + "loss": 3.8919, + "step": 610 + }, + { + "epoch": 0.018847190465357492, + "grad_norm": 0.29702016711235046, + "learning_rate": 9.423195558297347e-06, + "loss": 3.9003, + "step": 611 + }, + { + "epoch": 0.018878036930930906, + "grad_norm": 0.3080318868160248, + "learning_rate": 9.438618136952499e-06, + "loss": 3.879, + "step": 612 + }, + { + "epoch": 0.018908883396504324, + "grad_norm": 0.3264477849006653, + "learning_rate": 9.45404071560765e-06, + "loss": 3.8449, + "step": 613 + }, + { + "epoch": 0.01893972986207774, + "grad_norm": 0.2932049036026001, + "learning_rate": 9.4694632942628e-06, + "loss": 3.8891, + "step": 614 + }, + { + "epoch": 0.018970576327651156, + "grad_norm": 0.31368303298950195, + "learning_rate": 9.484885872917952e-06, + "loss": 3.903, + "step": 615 + }, + { + "epoch": 0.019001422793224573, + "grad_norm": 0.34329792857170105, + "learning_rate": 9.500308451573103e-06, + "loss": 3.8949, + "step": 616 + }, + { + "epoch": 0.01903226925879799, + "grad_norm": 0.3153548538684845, + "learning_rate": 9.515731030228255e-06, + "loss": 3.9038, + "step": 617 + }, + { + "epoch": 0.019063115724371405, + "grad_norm": 0.331134170293808, + "learning_rate": 9.531153608883406e-06, + "loss": 3.9104, + "step": 618 + }, + { + "epoch": 0.019093962189944823, + "grad_norm": 0.3146257698535919, + "learning_rate": 9.546576187538557e-06, + "loss": 3.8827, + "step": 619 + }, + { + "epoch": 0.01912480865551824, + "grad_norm": 0.30860722064971924, + "learning_rate": 9.561998766193709e-06, + "loss": 3.8943, + "step": 620 + }, + { + "epoch": 0.019155655121091658, + "grad_norm": 0.2976592779159546, + "learning_rate": 9.57742134484886e-06, + "loss": 3.8782, + "step": 621 + }, + { + "epoch": 0.019186501586665072, + "grad_norm": 0.34351709485054016, + "learning_rate": 9.59284392350401e-06, + "loss": 3.8516, + "step": 622 + }, + { + "epoch": 0.01921734805223849, + "grad_norm": 0.28913614153862, + "learning_rate": 9.608266502159162e-06, + "loss": 3.8669, + "step": 623 + }, + { + "epoch": 0.019248194517811908, + "grad_norm": 0.33480632305145264, + "learning_rate": 9.623689080814313e-06, + "loss": 3.8301, + "step": 624 + }, + { + "epoch": 0.019279040983385322, + "grad_norm": 0.2897937297821045, + "learning_rate": 9.639111659469465e-06, + "loss": 3.8961, + "step": 625 + }, + { + "epoch": 0.01930988744895874, + "grad_norm": 0.33788126707077026, + "learning_rate": 9.654534238124614e-06, + "loss": 3.8754, + "step": 626 + }, + { + "epoch": 0.019340733914532157, + "grad_norm": 0.49476876854896545, + "learning_rate": 9.669956816779765e-06, + "loss": 3.8883, + "step": 627 + }, + { + "epoch": 0.01937158038010557, + "grad_norm": 0.3149275481700897, + "learning_rate": 9.685379395434917e-06, + "loss": 3.8531, + "step": 628 + }, + { + "epoch": 0.01940242684567899, + "grad_norm": 0.30240175127983093, + "learning_rate": 9.700801974090068e-06, + "loss": 3.8375, + "step": 629 + }, + { + "epoch": 0.019433273311252407, + "grad_norm": 0.3183809518814087, + "learning_rate": 9.71622455274522e-06, + "loss": 3.8823, + "step": 630 + }, + { + "epoch": 0.01946411977682582, + "grad_norm": 0.3466890752315521, + "learning_rate": 9.73164713140037e-06, + "loss": 3.8798, + "step": 631 + }, + { + "epoch": 0.01949496624239924, + "grad_norm": 0.30030015110969543, + "learning_rate": 9.747069710055521e-06, + "loss": 3.8627, + "step": 632 + }, + { + "epoch": 0.019525812707972656, + "grad_norm": 0.3037448227405548, + "learning_rate": 9.762492288710673e-06, + "loss": 3.8871, + "step": 633 + }, + { + "epoch": 0.01955665917354607, + "grad_norm": 0.3259693682193756, + "learning_rate": 9.777914867365824e-06, + "loss": 3.835, + "step": 634 + }, + { + "epoch": 0.019587505639119488, + "grad_norm": 0.2735373377799988, + "learning_rate": 9.793337446020975e-06, + "loss": 3.872, + "step": 635 + }, + { + "epoch": 0.019618352104692906, + "grad_norm": 0.3367472290992737, + "learning_rate": 9.808760024676127e-06, + "loss": 3.8798, + "step": 636 + }, + { + "epoch": 0.01964919857026632, + "grad_norm": 0.29907289147377014, + "learning_rate": 9.824182603331278e-06, + "loss": 3.8411, + "step": 637 + }, + { + "epoch": 0.019680045035839738, + "grad_norm": 0.33275094628334045, + "learning_rate": 9.83960518198643e-06, + "loss": 3.8503, + "step": 638 + }, + { + "epoch": 0.019710891501413155, + "grad_norm": 0.2822853624820709, + "learning_rate": 9.85502776064158e-06, + "loss": 3.8699, + "step": 639 + }, + { + "epoch": 0.01974173796698657, + "grad_norm": 0.3147752583026886, + "learning_rate": 9.870450339296731e-06, + "loss": 3.8575, + "step": 640 + }, + { + "epoch": 0.019772584432559987, + "grad_norm": 0.30845269560813904, + "learning_rate": 9.885872917951882e-06, + "loss": 3.8598, + "step": 641 + }, + { + "epoch": 0.019803430898133405, + "grad_norm": 0.2955753207206726, + "learning_rate": 9.901295496607032e-06, + "loss": 3.9069, + "step": 642 + }, + { + "epoch": 0.01983427736370682, + "grad_norm": 0.30815261602401733, + "learning_rate": 9.916718075262185e-06, + "loss": 3.8988, + "step": 643 + }, + { + "epoch": 0.019865123829280237, + "grad_norm": 0.31427499651908875, + "learning_rate": 9.932140653917335e-06, + "loss": 3.87, + "step": 644 + }, + { + "epoch": 0.019895970294853654, + "grad_norm": 0.29957282543182373, + "learning_rate": 9.947563232572486e-06, + "loss": 3.8821, + "step": 645 + }, + { + "epoch": 0.01992681676042707, + "grad_norm": 0.2958481013774872, + "learning_rate": 9.962985811227638e-06, + "loss": 3.9227, + "step": 646 + }, + { + "epoch": 0.019957663226000486, + "grad_norm": 0.3128184378147125, + "learning_rate": 9.978408389882789e-06, + "loss": 3.8495, + "step": 647 + }, + { + "epoch": 0.019988509691573904, + "grad_norm": 0.27730700373649597, + "learning_rate": 9.99383096853794e-06, + "loss": 3.854, + "step": 648 + }, + { + "epoch": 0.020019356157147318, + "grad_norm": 0.29813817143440247, + "learning_rate": 1.0009253547193092e-05, + "loss": 3.8494, + "step": 649 + }, + { + "epoch": 0.020050202622720736, + "grad_norm": 0.29771730303764343, + "learning_rate": 1.0024676125848242e-05, + "loss": 3.8704, + "step": 650 + }, + { + "epoch": 0.020081049088294153, + "grad_norm": 0.2928963899612427, + "learning_rate": 1.0040098704503394e-05, + "loss": 3.8405, + "step": 651 + }, + { + "epoch": 0.020111895553867568, + "grad_norm": 0.3140575885772705, + "learning_rate": 1.0055521283158545e-05, + "loss": 3.8728, + "step": 652 + }, + { + "epoch": 0.020142742019440985, + "grad_norm": 0.28014761209487915, + "learning_rate": 1.0070943861813696e-05, + "loss": 3.8656, + "step": 653 + }, + { + "epoch": 0.020173588485014403, + "grad_norm": 0.34011006355285645, + "learning_rate": 1.0086366440468848e-05, + "loss": 3.8377, + "step": 654 + }, + { + "epoch": 0.020204434950587817, + "grad_norm": 0.302679181098938, + "learning_rate": 1.0101789019123999e-05, + "loss": 3.8739, + "step": 655 + }, + { + "epoch": 0.020235281416161235, + "grad_norm": 0.3164663314819336, + "learning_rate": 1.0117211597779149e-05, + "loss": 3.8412, + "step": 656 + }, + { + "epoch": 0.020266127881734652, + "grad_norm": 0.32334667444229126, + "learning_rate": 1.01326341764343e-05, + "loss": 3.8283, + "step": 657 + }, + { + "epoch": 0.020296974347308067, + "grad_norm": 0.3251613676548004, + "learning_rate": 1.014805675508945e-05, + "loss": 3.8733, + "step": 658 + }, + { + "epoch": 0.020327820812881484, + "grad_norm": 0.31220871210098267, + "learning_rate": 1.0163479333744603e-05, + "loss": 3.871, + "step": 659 + }, + { + "epoch": 0.020358667278454902, + "grad_norm": 0.3261125683784485, + "learning_rate": 1.0178901912399753e-05, + "loss": 3.8507, + "step": 660 + }, + { + "epoch": 0.020389513744028316, + "grad_norm": 0.30035650730133057, + "learning_rate": 1.0194324491054904e-05, + "loss": 3.8466, + "step": 661 + }, + { + "epoch": 0.020420360209601734, + "grad_norm": 0.30099645256996155, + "learning_rate": 1.0209747069710056e-05, + "loss": 3.8813, + "step": 662 + }, + { + "epoch": 0.02045120667517515, + "grad_norm": 0.32628387212753296, + "learning_rate": 1.0225169648365207e-05, + "loss": 3.8663, + "step": 663 + }, + { + "epoch": 0.020482053140748566, + "grad_norm": 0.28582054376602173, + "learning_rate": 1.0240592227020359e-05, + "loss": 3.8164, + "step": 664 + }, + { + "epoch": 0.020512899606321983, + "grad_norm": 0.30528751015663147, + "learning_rate": 1.025601480567551e-05, + "loss": 3.8138, + "step": 665 + }, + { + "epoch": 0.0205437460718954, + "grad_norm": 0.28473252058029175, + "learning_rate": 1.027143738433066e-05, + "loss": 3.8562, + "step": 666 + }, + { + "epoch": 0.020574592537468815, + "grad_norm": 1.3712400197982788, + "learning_rate": 1.0286859962985812e-05, + "loss": 3.8533, + "step": 667 + }, + { + "epoch": 0.020605439003042233, + "grad_norm": 0.29366111755371094, + "learning_rate": 1.0302282541640963e-05, + "loss": 3.85, + "step": 668 + }, + { + "epoch": 0.02063628546861565, + "grad_norm": 0.3264636695384979, + "learning_rate": 1.0317705120296114e-05, + "loss": 3.8194, + "step": 669 + }, + { + "epoch": 0.020667131934189065, + "grad_norm": 0.2953530251979828, + "learning_rate": 1.0333127698951266e-05, + "loss": 3.8762, + "step": 670 + }, + { + "epoch": 0.020697978399762482, + "grad_norm": 0.31250807642936707, + "learning_rate": 1.0348550277606417e-05, + "loss": 3.8612, + "step": 671 + }, + { + "epoch": 0.0207288248653359, + "grad_norm": 0.30696356296539307, + "learning_rate": 1.0363972856261567e-05, + "loss": 3.8528, + "step": 672 + }, + { + "epoch": 0.020759671330909314, + "grad_norm": 0.34691116213798523, + "learning_rate": 1.0379395434916718e-05, + "loss": 3.8713, + "step": 673 + }, + { + "epoch": 0.020790517796482732, + "grad_norm": 0.3042965233325958, + "learning_rate": 1.039481801357187e-05, + "loss": 3.8245, + "step": 674 + }, + { + "epoch": 0.02082136426205615, + "grad_norm": 0.35854077339172363, + "learning_rate": 1.041024059222702e-05, + "loss": 3.8394, + "step": 675 + }, + { + "epoch": 0.020852210727629564, + "grad_norm": 0.2972158193588257, + "learning_rate": 1.0425663170882171e-05, + "loss": 3.8524, + "step": 676 + }, + { + "epoch": 0.02088305719320298, + "grad_norm": 0.33702990412712097, + "learning_rate": 1.0441085749537324e-05, + "loss": 3.8833, + "step": 677 + }, + { + "epoch": 0.0209139036587764, + "grad_norm": 0.30110645294189453, + "learning_rate": 1.0456508328192474e-05, + "loss": 3.8367, + "step": 678 + }, + { + "epoch": 0.020944750124349813, + "grad_norm": 0.285474568605423, + "learning_rate": 1.0471930906847625e-05, + "loss": 3.8562, + "step": 679 + }, + { + "epoch": 0.02097559658992323, + "grad_norm": 0.33502164483070374, + "learning_rate": 1.0487353485502777e-05, + "loss": 3.8321, + "step": 680 + }, + { + "epoch": 0.02100644305549665, + "grad_norm": 0.31119444966316223, + "learning_rate": 1.0502776064157928e-05, + "loss": 3.8522, + "step": 681 + }, + { + "epoch": 0.021037289521070063, + "grad_norm": 0.3040304481983185, + "learning_rate": 1.051819864281308e-05, + "loss": 3.7952, + "step": 682 + }, + { + "epoch": 0.02106813598664348, + "grad_norm": 0.3004648685455322, + "learning_rate": 1.053362122146823e-05, + "loss": 3.8618, + "step": 683 + }, + { + "epoch": 0.021098982452216898, + "grad_norm": 0.2952640652656555, + "learning_rate": 1.0549043800123381e-05, + "loss": 3.8376, + "step": 684 + }, + { + "epoch": 0.021129828917790312, + "grad_norm": 0.27945801615715027, + "learning_rate": 1.0564466378778533e-05, + "loss": 3.8268, + "step": 685 + }, + { + "epoch": 0.02116067538336373, + "grad_norm": 0.298737108707428, + "learning_rate": 1.0579888957433684e-05, + "loss": 3.8283, + "step": 686 + }, + { + "epoch": 0.021191521848937148, + "grad_norm": 0.30837610363960266, + "learning_rate": 1.0595311536088835e-05, + "loss": 3.875, + "step": 687 + }, + { + "epoch": 0.021222368314510562, + "grad_norm": 0.28406745195388794, + "learning_rate": 1.0610734114743985e-05, + "loss": 3.8829, + "step": 688 + }, + { + "epoch": 0.02125321478008398, + "grad_norm": 0.29412877559661865, + "learning_rate": 1.0626156693399136e-05, + "loss": 3.853, + "step": 689 + }, + { + "epoch": 0.021284061245657397, + "grad_norm": 0.29812148213386536, + "learning_rate": 1.0641579272054288e-05, + "loss": 3.8391, + "step": 690 + }, + { + "epoch": 0.02131490771123081, + "grad_norm": 0.27533063292503357, + "learning_rate": 1.0657001850709439e-05, + "loss": 3.8353, + "step": 691 + }, + { + "epoch": 0.02134575417680423, + "grad_norm": 0.2835785150527954, + "learning_rate": 1.067242442936459e-05, + "loss": 3.8132, + "step": 692 + }, + { + "epoch": 0.021376600642377647, + "grad_norm": 0.30375924706459045, + "learning_rate": 1.0687847008019742e-05, + "loss": 3.8701, + "step": 693 + }, + { + "epoch": 0.02140744710795106, + "grad_norm": 0.2857644855976105, + "learning_rate": 1.0703269586674892e-05, + "loss": 3.8421, + "step": 694 + }, + { + "epoch": 0.02143829357352448, + "grad_norm": 0.3047102987766266, + "learning_rate": 1.0718692165330045e-05, + "loss": 3.8386, + "step": 695 + }, + { + "epoch": 0.021469140039097896, + "grad_norm": 0.2885618507862091, + "learning_rate": 1.0734114743985195e-05, + "loss": 3.8238, + "step": 696 + }, + { + "epoch": 0.02149998650467131, + "grad_norm": 0.2908279001712799, + "learning_rate": 1.0749537322640346e-05, + "loss": 3.8549, + "step": 697 + }, + { + "epoch": 0.021530832970244728, + "grad_norm": 0.3101877272129059, + "learning_rate": 1.0764959901295498e-05, + "loss": 3.8375, + "step": 698 + }, + { + "epoch": 0.021561679435818146, + "grad_norm": 0.28522205352783203, + "learning_rate": 1.0780382479950649e-05, + "loss": 3.8095, + "step": 699 + }, + { + "epoch": 0.02159252590139156, + "grad_norm": 0.27365583181381226, + "learning_rate": 1.07958050586058e-05, + "loss": 3.7773, + "step": 700 + }, + { + "epoch": 0.021623372366964978, + "grad_norm": 0.2819003760814667, + "learning_rate": 1.0811227637260951e-05, + "loss": 3.8065, + "step": 701 + }, + { + "epoch": 0.021654218832538395, + "grad_norm": 0.299411803483963, + "learning_rate": 1.0826650215916102e-05, + "loss": 3.8255, + "step": 702 + }, + { + "epoch": 0.02168506529811181, + "grad_norm": 0.3127776086330414, + "learning_rate": 1.0842072794571253e-05, + "loss": 3.8005, + "step": 703 + }, + { + "epoch": 0.021715911763685227, + "grad_norm": 0.2765767276287079, + "learning_rate": 1.0857495373226403e-05, + "loss": 3.8508, + "step": 704 + }, + { + "epoch": 0.021746758229258645, + "grad_norm": 0.32821616530418396, + "learning_rate": 1.0872917951881554e-05, + "loss": 3.8894, + "step": 705 + }, + { + "epoch": 0.02177760469483206, + "grad_norm": 0.2624841332435608, + "learning_rate": 1.0888340530536706e-05, + "loss": 3.827, + "step": 706 + }, + { + "epoch": 0.021808451160405477, + "grad_norm": 0.3216800391674042, + "learning_rate": 1.0903763109191857e-05, + "loss": 3.8557, + "step": 707 + }, + { + "epoch": 0.021839297625978894, + "grad_norm": 0.2962445616722107, + "learning_rate": 1.0919185687847009e-05, + "loss": 3.8097, + "step": 708 + }, + { + "epoch": 0.02187014409155231, + "grad_norm": 0.25820592045783997, + "learning_rate": 1.093460826650216e-05, + "loss": 3.8191, + "step": 709 + }, + { + "epoch": 0.021900990557125726, + "grad_norm": 0.29291245341300964, + "learning_rate": 1.095003084515731e-05, + "loss": 3.848, + "step": 710 + }, + { + "epoch": 0.021931837022699144, + "grad_norm": 0.2778274118900299, + "learning_rate": 1.0965453423812463e-05, + "loss": 3.8051, + "step": 711 + }, + { + "epoch": 0.021962683488272558, + "grad_norm": 0.3009317219257355, + "learning_rate": 1.0980876002467613e-05, + "loss": 3.7935, + "step": 712 + }, + { + "epoch": 0.021993529953845976, + "grad_norm": 0.2975328266620636, + "learning_rate": 1.0996298581122764e-05, + "loss": 3.8636, + "step": 713 + }, + { + "epoch": 0.022024376419419393, + "grad_norm": 0.2989310324192047, + "learning_rate": 1.1011721159777916e-05, + "loss": 3.7862, + "step": 714 + }, + { + "epoch": 0.022055222884992808, + "grad_norm": 0.31424394249916077, + "learning_rate": 1.1027143738433067e-05, + "loss": 3.8059, + "step": 715 + }, + { + "epoch": 0.022086069350566225, + "grad_norm": 0.26154422760009766, + "learning_rate": 1.1042566317088219e-05, + "loss": 3.8113, + "step": 716 + }, + { + "epoch": 0.022116915816139643, + "grad_norm": 0.3319043815135956, + "learning_rate": 1.105798889574337e-05, + "loss": 3.7965, + "step": 717 + }, + { + "epoch": 0.022147762281713057, + "grad_norm": 0.27671509981155396, + "learning_rate": 1.1073411474398518e-05, + "loss": 3.7943, + "step": 718 + }, + { + "epoch": 0.022178608747286475, + "grad_norm": 0.34975022077560425, + "learning_rate": 1.108883405305367e-05, + "loss": 3.8079, + "step": 719 + }, + { + "epoch": 0.022209455212859892, + "grad_norm": 0.2801985442638397, + "learning_rate": 1.1104256631708821e-05, + "loss": 3.8438, + "step": 720 + }, + { + "epoch": 0.022240301678433307, + "grad_norm": 0.3319854438304901, + "learning_rate": 1.1119679210363974e-05, + "loss": 3.7926, + "step": 721 + }, + { + "epoch": 0.022271148144006724, + "grad_norm": 0.3099897503852844, + "learning_rate": 1.1135101789019124e-05, + "loss": 3.8134, + "step": 722 + }, + { + "epoch": 0.022301994609580142, + "grad_norm": 0.3087681233882904, + "learning_rate": 1.1150524367674275e-05, + "loss": 3.7694, + "step": 723 + }, + { + "epoch": 0.022332841075153556, + "grad_norm": 0.30060893297195435, + "learning_rate": 1.1165946946329427e-05, + "loss": 3.8387, + "step": 724 + }, + { + "epoch": 0.022363687540726974, + "grad_norm": 0.31602710485458374, + "learning_rate": 1.1181369524984578e-05, + "loss": 3.8141, + "step": 725 + }, + { + "epoch": 0.02239453400630039, + "grad_norm": 0.2778055965900421, + "learning_rate": 1.1196792103639728e-05, + "loss": 3.8131, + "step": 726 + }, + { + "epoch": 0.022425380471873806, + "grad_norm": 0.2846052050590515, + "learning_rate": 1.121221468229488e-05, + "loss": 3.8168, + "step": 727 + }, + { + "epoch": 0.022456226937447223, + "grad_norm": 0.2775716781616211, + "learning_rate": 1.1227637260950031e-05, + "loss": 3.8198, + "step": 728 + }, + { + "epoch": 0.02248707340302064, + "grad_norm": 0.2708854079246521, + "learning_rate": 1.1243059839605184e-05, + "loss": 3.7913, + "step": 729 + }, + { + "epoch": 0.022517919868594055, + "grad_norm": 0.30115175247192383, + "learning_rate": 1.1258482418260334e-05, + "loss": 3.8331, + "step": 730 + }, + { + "epoch": 0.022548766334167473, + "grad_norm": 0.29232510924339294, + "learning_rate": 1.1273904996915485e-05, + "loss": 3.8413, + "step": 731 + }, + { + "epoch": 0.02257961279974089, + "grad_norm": 0.31017541885375977, + "learning_rate": 1.1289327575570637e-05, + "loss": 3.7986, + "step": 732 + }, + { + "epoch": 0.022610459265314305, + "grad_norm": 0.2907472252845764, + "learning_rate": 1.1304750154225788e-05, + "loss": 3.7994, + "step": 733 + }, + { + "epoch": 0.022641305730887722, + "grad_norm": 0.2705024480819702, + "learning_rate": 1.1320172732880938e-05, + "loss": 3.8147, + "step": 734 + }, + { + "epoch": 0.02267215219646114, + "grad_norm": 0.3327040374279022, + "learning_rate": 1.1335595311536089e-05, + "loss": 3.7601, + "step": 735 + }, + { + "epoch": 0.022702998662034554, + "grad_norm": 0.2983090877532959, + "learning_rate": 1.135101789019124e-05, + "loss": 3.8042, + "step": 736 + }, + { + "epoch": 0.022733845127607972, + "grad_norm": 0.3492867052555084, + "learning_rate": 1.1366440468846392e-05, + "loss": 3.8145, + "step": 737 + }, + { + "epoch": 0.02276469159318139, + "grad_norm": 0.3645930588245392, + "learning_rate": 1.1381863047501542e-05, + "loss": 3.8085, + "step": 738 + }, + { + "epoch": 0.022795538058754804, + "grad_norm": 0.3540519177913666, + "learning_rate": 1.1397285626156695e-05, + "loss": 3.8218, + "step": 739 + }, + { + "epoch": 0.02282638452432822, + "grad_norm": 0.27309057116508484, + "learning_rate": 1.1412708204811845e-05, + "loss": 3.8243, + "step": 740 + }, + { + "epoch": 0.02285723098990164, + "grad_norm": 0.34920039772987366, + "learning_rate": 1.1428130783466996e-05, + "loss": 3.8145, + "step": 741 + }, + { + "epoch": 0.022888077455475053, + "grad_norm": 0.2896840274333954, + "learning_rate": 1.1443553362122148e-05, + "loss": 3.764, + "step": 742 + }, + { + "epoch": 0.02291892392104847, + "grad_norm": 0.334908664226532, + "learning_rate": 1.1458975940777299e-05, + "loss": 3.8159, + "step": 743 + }, + { + "epoch": 0.02294977038662189, + "grad_norm": 0.2779608368873596, + "learning_rate": 1.147439851943245e-05, + "loss": 3.7784, + "step": 744 + }, + { + "epoch": 0.022980616852195303, + "grad_norm": 0.33188578486442566, + "learning_rate": 1.1489821098087602e-05, + "loss": 3.7778, + "step": 745 + }, + { + "epoch": 0.02301146331776872, + "grad_norm": 0.2916290760040283, + "learning_rate": 1.1505243676742752e-05, + "loss": 3.7673, + "step": 746 + }, + { + "epoch": 0.023042309783342138, + "grad_norm": 0.32541099190711975, + "learning_rate": 1.1520666255397903e-05, + "loss": 3.8002, + "step": 747 + }, + { + "epoch": 0.023073156248915552, + "grad_norm": 0.28803524374961853, + "learning_rate": 1.1536088834053055e-05, + "loss": 3.8266, + "step": 748 + }, + { + "epoch": 0.02310400271448897, + "grad_norm": 0.4567795693874359, + "learning_rate": 1.1551511412708204e-05, + "loss": 3.7911, + "step": 749 + }, + { + "epoch": 0.023134849180062388, + "grad_norm": 0.2888467311859131, + "learning_rate": 1.1566933991363356e-05, + "loss": 3.7916, + "step": 750 + }, + { + "epoch": 0.023165695645635802, + "grad_norm": 0.30702659487724304, + "learning_rate": 1.1582356570018507e-05, + "loss": 3.81, + "step": 751 + }, + { + "epoch": 0.02319654211120922, + "grad_norm": 0.287707656621933, + "learning_rate": 1.159777914867366e-05, + "loss": 3.812, + "step": 752 + }, + { + "epoch": 0.023227388576782637, + "grad_norm": 0.2877368927001953, + "learning_rate": 1.161320172732881e-05, + "loss": 3.7867, + "step": 753 + }, + { + "epoch": 0.02325823504235605, + "grad_norm": 0.3211654722690582, + "learning_rate": 1.162862430598396e-05, + "loss": 3.7714, + "step": 754 + }, + { + "epoch": 0.02328908150792947, + "grad_norm": 0.3304784595966339, + "learning_rate": 1.1644046884639113e-05, + "loss": 3.7825, + "step": 755 + }, + { + "epoch": 0.023319927973502887, + "grad_norm": 0.29197365045547485, + "learning_rate": 1.1659469463294263e-05, + "loss": 3.7811, + "step": 756 + }, + { + "epoch": 0.023350774439076304, + "grad_norm": 0.3132404088973999, + "learning_rate": 1.1674892041949414e-05, + "loss": 3.8209, + "step": 757 + }, + { + "epoch": 0.02338162090464972, + "grad_norm": 0.3158780634403229, + "learning_rate": 1.1690314620604566e-05, + "loss": 3.7811, + "step": 758 + }, + { + "epoch": 0.023412467370223136, + "grad_norm": 0.2968407869338989, + "learning_rate": 1.1705737199259717e-05, + "loss": 3.78, + "step": 759 + }, + { + "epoch": 0.023443313835796554, + "grad_norm": 0.3355177044868469, + "learning_rate": 1.1721159777914869e-05, + "loss": 3.8193, + "step": 760 + }, + { + "epoch": 0.023474160301369968, + "grad_norm": 0.26430603861808777, + "learning_rate": 1.173658235657002e-05, + "loss": 3.8021, + "step": 761 + }, + { + "epoch": 0.023505006766943386, + "grad_norm": 0.3128180503845215, + "learning_rate": 1.175200493522517e-05, + "loss": 3.7395, + "step": 762 + }, + { + "epoch": 0.023535853232516803, + "grad_norm": 0.2812933325767517, + "learning_rate": 1.1767427513880323e-05, + "loss": 3.7464, + "step": 763 + }, + { + "epoch": 0.023566699698090218, + "grad_norm": 0.29529500007629395, + "learning_rate": 1.1782850092535473e-05, + "loss": 3.7629, + "step": 764 + }, + { + "epoch": 0.023597546163663635, + "grad_norm": 0.3013310134410858, + "learning_rate": 1.1798272671190624e-05, + "loss": 3.791, + "step": 765 + }, + { + "epoch": 0.023628392629237053, + "grad_norm": 0.28242063522338867, + "learning_rate": 1.1813695249845774e-05, + "loss": 3.7707, + "step": 766 + }, + { + "epoch": 0.023659239094810467, + "grad_norm": 0.32069265842437744, + "learning_rate": 1.1829117828500925e-05, + "loss": 3.7704, + "step": 767 + }, + { + "epoch": 0.023690085560383885, + "grad_norm": 0.3139113187789917, + "learning_rate": 1.1844540407156077e-05, + "loss": 3.7741, + "step": 768 + }, + { + "epoch": 0.023720932025957302, + "grad_norm": 0.3525616526603699, + "learning_rate": 1.1859962985811228e-05, + "loss": 3.7976, + "step": 769 + }, + { + "epoch": 0.023751778491530717, + "grad_norm": 0.28328701853752136, + "learning_rate": 1.1875385564466378e-05, + "loss": 3.7475, + "step": 770 + }, + { + "epoch": 0.023782624957104134, + "grad_norm": 0.37036004662513733, + "learning_rate": 1.189080814312153e-05, + "loss": 3.7679, + "step": 771 + }, + { + "epoch": 0.023813471422677552, + "grad_norm": 0.2939198315143585, + "learning_rate": 1.1906230721776681e-05, + "loss": 3.7894, + "step": 772 + }, + { + "epoch": 0.023844317888250966, + "grad_norm": 0.34323012828826904, + "learning_rate": 1.1921653300431834e-05, + "loss": 3.759, + "step": 773 + }, + { + "epoch": 0.023875164353824384, + "grad_norm": 0.29486748576164246, + "learning_rate": 1.1937075879086984e-05, + "loss": 3.7615, + "step": 774 + }, + { + "epoch": 0.0239060108193978, + "grad_norm": 0.3175199329853058, + "learning_rate": 1.1952498457742135e-05, + "loss": 3.7879, + "step": 775 + }, + { + "epoch": 0.023936857284971216, + "grad_norm": 0.3154183030128479, + "learning_rate": 1.1967921036397287e-05, + "loss": 3.7572, + "step": 776 + }, + { + "epoch": 0.023967703750544633, + "grad_norm": 0.27494016289711, + "learning_rate": 1.1983343615052438e-05, + "loss": 3.7437, + "step": 777 + }, + { + "epoch": 0.02399855021611805, + "grad_norm": 0.3085651993751526, + "learning_rate": 1.1998766193707588e-05, + "loss": 3.7666, + "step": 778 + }, + { + "epoch": 0.024029396681691465, + "grad_norm": 0.30204513669013977, + "learning_rate": 1.201418877236274e-05, + "loss": 3.8042, + "step": 779 + }, + { + "epoch": 0.024060243147264883, + "grad_norm": 0.26955679059028625, + "learning_rate": 1.202961135101789e-05, + "loss": 3.8009, + "step": 780 + }, + { + "epoch": 0.0240910896128383, + "grad_norm": 0.2756097912788391, + "learning_rate": 1.2045033929673042e-05, + "loss": 3.7439, + "step": 781 + }, + { + "epoch": 0.024121936078411715, + "grad_norm": 0.292546808719635, + "learning_rate": 1.2060456508328192e-05, + "loss": 3.7931, + "step": 782 + }, + { + "epoch": 0.024152782543985132, + "grad_norm": 0.2733784019947052, + "learning_rate": 1.2075879086983343e-05, + "loss": 3.7578, + "step": 783 + }, + { + "epoch": 0.02418362900955855, + "grad_norm": 0.3072745203971863, + "learning_rate": 1.2091301665638495e-05, + "loss": 3.7537, + "step": 784 + }, + { + "epoch": 0.024214475475131964, + "grad_norm": 0.292239785194397, + "learning_rate": 1.2106724244293646e-05, + "loss": 3.7869, + "step": 785 + }, + { + "epoch": 0.024245321940705382, + "grad_norm": 0.2938590943813324, + "learning_rate": 1.2122146822948798e-05, + "loss": 3.7789, + "step": 786 + }, + { + "epoch": 0.0242761684062788, + "grad_norm": 0.28691089153289795, + "learning_rate": 1.2137569401603949e-05, + "loss": 3.8233, + "step": 787 + }, + { + "epoch": 0.024307014871852214, + "grad_norm": 0.3152560591697693, + "learning_rate": 1.21529919802591e-05, + "loss": 3.755, + "step": 788 + }, + { + "epoch": 0.02433786133742563, + "grad_norm": 0.27569302916526794, + "learning_rate": 1.2168414558914252e-05, + "loss": 3.7619, + "step": 789 + }, + { + "epoch": 0.02436870780299905, + "grad_norm": 0.3155033588409424, + "learning_rate": 1.2183837137569402e-05, + "loss": 3.8078, + "step": 790 + }, + { + "epoch": 0.024399554268572463, + "grad_norm": 0.29878127574920654, + "learning_rate": 1.2199259716224553e-05, + "loss": 3.7783, + "step": 791 + }, + { + "epoch": 0.02443040073414588, + "grad_norm": 0.3166087865829468, + "learning_rate": 1.2214682294879705e-05, + "loss": 3.6987, + "step": 792 + }, + { + "epoch": 0.0244612471997193, + "grad_norm": 0.29612547159194946, + "learning_rate": 1.2230104873534856e-05, + "loss": 3.7659, + "step": 793 + }, + { + "epoch": 0.024492093665292713, + "grad_norm": 0.3289998173713684, + "learning_rate": 1.2245527452190008e-05, + "loss": 3.805, + "step": 794 + }, + { + "epoch": 0.02452294013086613, + "grad_norm": 0.30642297863960266, + "learning_rate": 1.2260950030845159e-05, + "loss": 3.7551, + "step": 795 + }, + { + "epoch": 0.024553786596439548, + "grad_norm": 0.29994380474090576, + "learning_rate": 1.2276372609500308e-05, + "loss": 3.7502, + "step": 796 + }, + { + "epoch": 0.024584633062012962, + "grad_norm": 0.3158094584941864, + "learning_rate": 1.229179518815546e-05, + "loss": 3.7483, + "step": 797 + }, + { + "epoch": 0.02461547952758638, + "grad_norm": 0.303558886051178, + "learning_rate": 1.230721776681061e-05, + "loss": 3.7635, + "step": 798 + }, + { + "epoch": 0.024646325993159798, + "grad_norm": 0.32628655433654785, + "learning_rate": 1.2322640345465763e-05, + "loss": 3.7901, + "step": 799 + }, + { + "epoch": 0.024677172458733212, + "grad_norm": 0.26948291063308716, + "learning_rate": 1.2338062924120913e-05, + "loss": 3.7374, + "step": 800 + }, + { + "epoch": 0.02470801892430663, + "grad_norm": 0.32616546750068665, + "learning_rate": 1.2353485502776064e-05, + "loss": 3.7533, + "step": 801 + }, + { + "epoch": 0.024738865389880047, + "grad_norm": 0.3068939447402954, + "learning_rate": 1.2368908081431216e-05, + "loss": 3.8072, + "step": 802 + }, + { + "epoch": 0.02476971185545346, + "grad_norm": 0.29877421259880066, + "learning_rate": 1.2384330660086367e-05, + "loss": 3.749, + "step": 803 + }, + { + "epoch": 0.02480055832102688, + "grad_norm": 0.3255801498889923, + "learning_rate": 1.2399753238741517e-05, + "loss": 3.7438, + "step": 804 + }, + { + "epoch": 0.024831404786600297, + "grad_norm": 0.29108139872550964, + "learning_rate": 1.241517581739667e-05, + "loss": 3.7736, + "step": 805 + }, + { + "epoch": 0.02486225125217371, + "grad_norm": 0.2848063111305237, + "learning_rate": 1.243059839605182e-05, + "loss": 3.7383, + "step": 806 + }, + { + "epoch": 0.02489309771774713, + "grad_norm": 0.27210620045661926, + "learning_rate": 1.2446020974706973e-05, + "loss": 3.7391, + "step": 807 + }, + { + "epoch": 0.024923944183320546, + "grad_norm": 0.2803117334842682, + "learning_rate": 1.2461443553362123e-05, + "loss": 3.749, + "step": 808 + }, + { + "epoch": 0.02495479064889396, + "grad_norm": 0.27191978693008423, + "learning_rate": 1.2476866132017274e-05, + "loss": 3.7538, + "step": 809 + }, + { + "epoch": 0.024985637114467378, + "grad_norm": 0.2790079414844513, + "learning_rate": 1.2492288710672426e-05, + "loss": 3.7227, + "step": 810 + }, + { + "epoch": 0.025016483580040796, + "grad_norm": 0.2575942277908325, + "learning_rate": 1.2507711289327575e-05, + "loss": 3.7653, + "step": 811 + }, + { + "epoch": 0.02504733004561421, + "grad_norm": 0.26819485425949097, + "learning_rate": 1.2523133867982727e-05, + "loss": 3.7548, + "step": 812 + }, + { + "epoch": 0.025078176511187628, + "grad_norm": 0.2777901589870453, + "learning_rate": 1.2538556446637878e-05, + "loss": 3.7137, + "step": 813 + }, + { + "epoch": 0.025109022976761045, + "grad_norm": 0.2727816700935364, + "learning_rate": 1.255397902529303e-05, + "loss": 3.7635, + "step": 814 + }, + { + "epoch": 0.02513986944233446, + "grad_norm": 0.24618196487426758, + "learning_rate": 1.256940160394818e-05, + "loss": 3.7412, + "step": 815 + }, + { + "epoch": 0.025170715907907877, + "grad_norm": 0.2714095115661621, + "learning_rate": 1.2584824182603333e-05, + "loss": 3.7628, + "step": 816 + }, + { + "epoch": 0.025201562373481295, + "grad_norm": 0.2777657210826874, + "learning_rate": 1.2600246761258484e-05, + "loss": 3.7379, + "step": 817 + }, + { + "epoch": 0.02523240883905471, + "grad_norm": 0.26680323481559753, + "learning_rate": 1.2615669339913636e-05, + "loss": 3.7495, + "step": 818 + }, + { + "epoch": 0.025263255304628127, + "grad_norm": 0.30911222100257874, + "learning_rate": 1.2631091918568785e-05, + "loss": 3.7269, + "step": 819 + }, + { + "epoch": 0.025294101770201544, + "grad_norm": 0.25889989733695984, + "learning_rate": 1.2646514497223936e-05, + "loss": 3.7286, + "step": 820 + }, + { + "epoch": 0.02532494823577496, + "grad_norm": 0.3499143719673157, + "learning_rate": 1.2661937075879088e-05, + "loss": 3.762, + "step": 821 + }, + { + "epoch": 0.025355794701348376, + "grad_norm": 0.25442570447921753, + "learning_rate": 1.2677359654534238e-05, + "loss": 3.783, + "step": 822 + }, + { + "epoch": 0.025386641166921794, + "grad_norm": 0.3143722116947174, + "learning_rate": 1.269278223318939e-05, + "loss": 3.7467, + "step": 823 + }, + { + "epoch": 0.025417487632495208, + "grad_norm": 0.26888367533683777, + "learning_rate": 1.270820481184454e-05, + "loss": 3.7511, + "step": 824 + }, + { + "epoch": 0.025448334098068626, + "grad_norm": 0.2947244942188263, + "learning_rate": 1.2723627390499694e-05, + "loss": 3.8046, + "step": 825 + }, + { + "epoch": 0.025479180563642043, + "grad_norm": 0.3088867962360382, + "learning_rate": 1.2739049969154843e-05, + "loss": 3.7272, + "step": 826 + }, + { + "epoch": 0.025510027029215458, + "grad_norm": 0.27410751581192017, + "learning_rate": 1.2754472547809995e-05, + "loss": 3.7814, + "step": 827 + }, + { + "epoch": 0.025540873494788875, + "grad_norm": 0.3128851652145386, + "learning_rate": 1.2769895126465145e-05, + "loss": 3.7251, + "step": 828 + }, + { + "epoch": 0.025571719960362293, + "grad_norm": 0.275599867105484, + "learning_rate": 1.2785317705120298e-05, + "loss": 3.7394, + "step": 829 + }, + { + "epoch": 0.025602566425935707, + "grad_norm": 0.31116756796836853, + "learning_rate": 1.2800740283775448e-05, + "loss": 3.7416, + "step": 830 + }, + { + "epoch": 0.025633412891509125, + "grad_norm": 0.26396146416664124, + "learning_rate": 1.28161628624306e-05, + "loss": 3.7396, + "step": 831 + }, + { + "epoch": 0.025664259357082542, + "grad_norm": 0.2749943137168884, + "learning_rate": 1.283158544108575e-05, + "loss": 3.722, + "step": 832 + }, + { + "epoch": 0.025695105822655957, + "grad_norm": 0.2681169807910919, + "learning_rate": 1.2847008019740902e-05, + "loss": 3.7462, + "step": 833 + }, + { + "epoch": 0.025725952288229374, + "grad_norm": 0.29859334230422974, + "learning_rate": 1.2862430598396052e-05, + "loss": 3.7112, + "step": 834 + }, + { + "epoch": 0.025756798753802792, + "grad_norm": 0.29883649945259094, + "learning_rate": 1.2877853177051203e-05, + "loss": 3.7722, + "step": 835 + }, + { + "epoch": 0.025787645219376206, + "grad_norm": 0.4629368782043457, + "learning_rate": 1.2893275755706355e-05, + "loss": 3.7348, + "step": 836 + }, + { + "epoch": 0.025818491684949624, + "grad_norm": 0.28303563594818115, + "learning_rate": 1.2908698334361504e-05, + "loss": 3.7385, + "step": 837 + }, + { + "epoch": 0.02584933815052304, + "grad_norm": 0.2830965518951416, + "learning_rate": 1.2924120913016658e-05, + "loss": 3.7055, + "step": 838 + }, + { + "epoch": 0.025880184616096456, + "grad_norm": 0.2728595733642578, + "learning_rate": 1.2939543491671807e-05, + "loss": 3.7193, + "step": 839 + }, + { + "epoch": 0.025911031081669873, + "grad_norm": 0.2926602363586426, + "learning_rate": 1.295496607032696e-05, + "loss": 3.7474, + "step": 840 + }, + { + "epoch": 0.02594187754724329, + "grad_norm": 0.2836267948150635, + "learning_rate": 1.297038864898211e-05, + "loss": 3.7392, + "step": 841 + }, + { + "epoch": 0.025972724012816705, + "grad_norm": 0.30341917276382446, + "learning_rate": 1.2985811227637262e-05, + "loss": 3.7451, + "step": 842 + }, + { + "epoch": 0.026003570478390123, + "grad_norm": 0.26635509729385376, + "learning_rate": 1.3001233806292413e-05, + "loss": 3.7439, + "step": 843 + }, + { + "epoch": 0.02603441694396354, + "grad_norm": 0.2951584458351135, + "learning_rate": 1.3016656384947565e-05, + "loss": 3.7656, + "step": 844 + }, + { + "epoch": 0.026065263409536955, + "grad_norm": 0.37562039494514465, + "learning_rate": 1.3032078963602714e-05, + "loss": 3.7811, + "step": 845 + }, + { + "epoch": 0.026096109875110372, + "grad_norm": 0.2551366090774536, + "learning_rate": 1.3047501542257868e-05, + "loss": 3.7284, + "step": 846 + }, + { + "epoch": 0.02612695634068379, + "grad_norm": 0.26856693625450134, + "learning_rate": 1.3062924120913017e-05, + "loss": 3.7496, + "step": 847 + }, + { + "epoch": 0.026157802806257204, + "grad_norm": 0.2613849341869354, + "learning_rate": 1.307834669956817e-05, + "loss": 3.7563, + "step": 848 + }, + { + "epoch": 0.026188649271830622, + "grad_norm": 0.2666480243206024, + "learning_rate": 1.309376927822332e-05, + "loss": 3.727, + "step": 849 + }, + { + "epoch": 0.02621949573740404, + "grad_norm": 0.27196207642555237, + "learning_rate": 1.3109191856878469e-05, + "loss": 3.7269, + "step": 850 + }, + { + "epoch": 0.026250342202977454, + "grad_norm": 0.24855084717273712, + "learning_rate": 1.3124614435533623e-05, + "loss": 3.7064, + "step": 851 + }, + { + "epoch": 0.02628118866855087, + "grad_norm": 0.2611757516860962, + "learning_rate": 1.3140037014188772e-05, + "loss": 3.7222, + "step": 852 + }, + { + "epoch": 0.02631203513412429, + "grad_norm": 0.2625925838947296, + "learning_rate": 1.3155459592843924e-05, + "loss": 3.7472, + "step": 853 + }, + { + "epoch": 0.026342881599697703, + "grad_norm": 0.2653435170650482, + "learning_rate": 1.3170882171499075e-05, + "loss": 3.7534, + "step": 854 + }, + { + "epoch": 0.02637372806527112, + "grad_norm": 0.30051863193511963, + "learning_rate": 1.3186304750154227e-05, + "loss": 3.7091, + "step": 855 + }, + { + "epoch": 0.02640457453084454, + "grad_norm": 0.273875892162323, + "learning_rate": 1.3201727328809377e-05, + "loss": 3.7102, + "step": 856 + }, + { + "epoch": 0.026435420996417953, + "grad_norm": 0.252520352602005, + "learning_rate": 1.321714990746453e-05, + "loss": 3.7557, + "step": 857 + }, + { + "epoch": 0.02646626746199137, + "grad_norm": 0.2921120822429657, + "learning_rate": 1.3232572486119679e-05, + "loss": 3.766, + "step": 858 + }, + { + "epoch": 0.026497113927564788, + "grad_norm": 0.2669920325279236, + "learning_rate": 1.3247995064774833e-05, + "loss": 3.7171, + "step": 859 + }, + { + "epoch": 0.026527960393138202, + "grad_norm": 0.2685324549674988, + "learning_rate": 1.3263417643429982e-05, + "loss": 3.7172, + "step": 860 + }, + { + "epoch": 0.02655880685871162, + "grad_norm": 0.32729142904281616, + "learning_rate": 1.3278840222085134e-05, + "loss": 3.7592, + "step": 861 + }, + { + "epoch": 0.026589653324285038, + "grad_norm": 0.2939786911010742, + "learning_rate": 1.3294262800740284e-05, + "loss": 3.7217, + "step": 862 + }, + { + "epoch": 0.026620499789858452, + "grad_norm": 0.30078810453414917, + "learning_rate": 1.3309685379395437e-05, + "loss": 3.7285, + "step": 863 + }, + { + "epoch": 0.02665134625543187, + "grad_norm": 0.41784414649009705, + "learning_rate": 1.3325107958050587e-05, + "loss": 3.7297, + "step": 864 + }, + { + "epoch": 0.026682192721005287, + "grad_norm": 0.4627390503883362, + "learning_rate": 1.3340530536705736e-05, + "loss": 3.7424, + "step": 865 + }, + { + "epoch": 0.0267130391865787, + "grad_norm": 0.8131214380264282, + "learning_rate": 1.3355953115360889e-05, + "loss": 3.7158, + "step": 866 + }, + { + "epoch": 0.02674388565215212, + "grad_norm": 1.1677910089492798, + "learning_rate": 1.3371375694016039e-05, + "loss": 3.7438, + "step": 867 + }, + { + "epoch": 0.026774732117725537, + "grad_norm": 0.2978474497795105, + "learning_rate": 1.3386798272671191e-05, + "loss": 3.6942, + "step": 868 + }, + { + "epoch": 0.02680557858329895, + "grad_norm": 0.7441917657852173, + "learning_rate": 1.3402220851326342e-05, + "loss": 3.6889, + "step": 869 + }, + { + "epoch": 0.02683642504887237, + "grad_norm": 0.39667338132858276, + "learning_rate": 1.3417643429981494e-05, + "loss": 3.7403, + "step": 870 + }, + { + "epoch": 0.026867271514445786, + "grad_norm": 0.9737300276756287, + "learning_rate": 1.3433066008636643e-05, + "loss": 3.7231, + "step": 871 + }, + { + "epoch": 0.0268981179800192, + "grad_norm": 0.3513216972351074, + "learning_rate": 1.3448488587291797e-05, + "loss": 3.7253, + "step": 872 + }, + { + "epoch": 0.026928964445592618, + "grad_norm": 0.3086363673210144, + "learning_rate": 1.3463911165946946e-05, + "loss": 3.7404, + "step": 873 + }, + { + "epoch": 0.026959810911166036, + "grad_norm": 0.3279983401298523, + "learning_rate": 1.3479333744602098e-05, + "loss": 3.7269, + "step": 874 + }, + { + "epoch": 0.02699065737673945, + "grad_norm": 0.32062339782714844, + "learning_rate": 1.3494756323257249e-05, + "loss": 3.69, + "step": 875 + }, + { + "epoch": 0.027021503842312868, + "grad_norm": 0.2709603011608124, + "learning_rate": 1.3510178901912401e-05, + "loss": 3.7297, + "step": 876 + }, + { + "epoch": 0.027052350307886285, + "grad_norm": 0.29741281270980835, + "learning_rate": 1.3525601480567552e-05, + "loss": 3.7257, + "step": 877 + }, + { + "epoch": 0.0270831967734597, + "grad_norm": 0.30818837881088257, + "learning_rate": 1.3541024059222704e-05, + "loss": 3.6997, + "step": 878 + }, + { + "epoch": 0.027114043239033117, + "grad_norm": 0.285917192697525, + "learning_rate": 1.3556446637877853e-05, + "loss": 3.7159, + "step": 879 + }, + { + "epoch": 0.027144889704606535, + "grad_norm": 0.30020007491111755, + "learning_rate": 1.3571869216533004e-05, + "loss": 3.7393, + "step": 880 + }, + { + "epoch": 0.02717573617017995, + "grad_norm": 0.2583557963371277, + "learning_rate": 1.3587291795188156e-05, + "loss": 3.7401, + "step": 881 + }, + { + "epoch": 0.027206582635753367, + "grad_norm": 0.3311845362186432, + "learning_rate": 1.3602714373843307e-05, + "loss": 3.7363, + "step": 882 + }, + { + "epoch": 0.027237429101326784, + "grad_norm": 0.2711980640888214, + "learning_rate": 1.3618136952498459e-05, + "loss": 3.6677, + "step": 883 + }, + { + "epoch": 0.0272682755669002, + "grad_norm": 0.30283892154693604, + "learning_rate": 1.3633559531153608e-05, + "loss": 3.7343, + "step": 884 + }, + { + "epoch": 0.027299122032473616, + "grad_norm": 0.28681305050849915, + "learning_rate": 1.3648982109808762e-05, + "loss": 3.6807, + "step": 885 + }, + { + "epoch": 0.027329968498047034, + "grad_norm": 0.3009156882762909, + "learning_rate": 1.366440468846391e-05, + "loss": 3.7195, + "step": 886 + }, + { + "epoch": 0.027360814963620448, + "grad_norm": 0.267649382352829, + "learning_rate": 1.3679827267119063e-05, + "loss": 3.7001, + "step": 887 + }, + { + "epoch": 0.027391661429193866, + "grad_norm": 0.26894611120224, + "learning_rate": 1.3695249845774214e-05, + "loss": 3.742, + "step": 888 + }, + { + "epoch": 0.027422507894767283, + "grad_norm": 0.2482168972492218, + "learning_rate": 1.3710672424429366e-05, + "loss": 3.6788, + "step": 889 + }, + { + "epoch": 0.027453354360340698, + "grad_norm": 0.28737691044807434, + "learning_rate": 1.3726095003084516e-05, + "loss": 3.7326, + "step": 890 + }, + { + "epoch": 0.027484200825914115, + "grad_norm": 0.25951385498046875, + "learning_rate": 1.3741517581739669e-05, + "loss": 3.7206, + "step": 891 + }, + { + "epoch": 0.027515047291487533, + "grad_norm": 0.2581038177013397, + "learning_rate": 1.3756940160394818e-05, + "loss": 3.7476, + "step": 892 + }, + { + "epoch": 0.027545893757060947, + "grad_norm": 0.2570895552635193, + "learning_rate": 1.3772362739049972e-05, + "loss": 3.7114, + "step": 893 + }, + { + "epoch": 0.027576740222634365, + "grad_norm": 0.26339101791381836, + "learning_rate": 1.378778531770512e-05, + "loss": 3.7102, + "step": 894 + }, + { + "epoch": 0.027607586688207782, + "grad_norm": 0.2418684959411621, + "learning_rate": 1.3803207896360273e-05, + "loss": 3.6926, + "step": 895 + }, + { + "epoch": 0.0276384331537812, + "grad_norm": 0.8760995864868164, + "learning_rate": 1.3818630475015423e-05, + "loss": 3.7272, + "step": 896 + }, + { + "epoch": 0.027669279619354614, + "grad_norm": 0.25834810733795166, + "learning_rate": 1.3834053053670572e-05, + "loss": 3.7069, + "step": 897 + }, + { + "epoch": 0.027700126084928032, + "grad_norm": 0.25935572385787964, + "learning_rate": 1.3849475632325726e-05, + "loss": 3.6812, + "step": 898 + }, + { + "epoch": 0.02773097255050145, + "grad_norm": 0.2621622383594513, + "learning_rate": 1.3864898210980875e-05, + "loss": 3.699, + "step": 899 + }, + { + "epoch": 0.027761819016074864, + "grad_norm": 0.42784878611564636, + "learning_rate": 1.3880320789636028e-05, + "loss": 3.6795, + "step": 900 + }, + { + "epoch": 0.02779266548164828, + "grad_norm": 0.2929039001464844, + "learning_rate": 1.3895743368291178e-05, + "loss": 3.7161, + "step": 901 + }, + { + "epoch": 0.0278235119472217, + "grad_norm": 0.42287907004356384, + "learning_rate": 1.391116594694633e-05, + "loss": 3.6992, + "step": 902 + }, + { + "epoch": 0.027854358412795113, + "grad_norm": 0.27959880232810974, + "learning_rate": 1.3926588525601481e-05, + "loss": 3.671, + "step": 903 + }, + { + "epoch": 0.02788520487836853, + "grad_norm": 0.2500801980495453, + "learning_rate": 1.3942011104256633e-05, + "loss": 3.7074, + "step": 904 + }, + { + "epoch": 0.02791605134394195, + "grad_norm": 0.26861193776130676, + "learning_rate": 1.3957433682911782e-05, + "loss": 3.7554, + "step": 905 + }, + { + "epoch": 0.027946897809515363, + "grad_norm": 0.25102856755256653, + "learning_rate": 1.3972856261566936e-05, + "loss": 3.6787, + "step": 906 + }, + { + "epoch": 0.02797774427508878, + "grad_norm": 0.25203025341033936, + "learning_rate": 1.3988278840222085e-05, + "loss": 3.6956, + "step": 907 + }, + { + "epoch": 0.028008590740662198, + "grad_norm": 0.2556506097316742, + "learning_rate": 1.4003701418877237e-05, + "loss": 3.7116, + "step": 908 + }, + { + "epoch": 0.028039437206235612, + "grad_norm": 0.2512381970882416, + "learning_rate": 1.4019123997532388e-05, + "loss": 3.672, + "step": 909 + }, + { + "epoch": 0.02807028367180903, + "grad_norm": 0.24656182527542114, + "learning_rate": 1.403454657618754e-05, + "loss": 3.7093, + "step": 910 + }, + { + "epoch": 0.028101130137382448, + "grad_norm": 0.27085721492767334, + "learning_rate": 1.4049969154842691e-05, + "loss": 3.707, + "step": 911 + }, + { + "epoch": 0.028131976602955862, + "grad_norm": 0.269015908241272, + "learning_rate": 1.406539173349784e-05, + "loss": 3.7175, + "step": 912 + }, + { + "epoch": 0.02816282306852928, + "grad_norm": 0.25532352924346924, + "learning_rate": 1.4080814312152992e-05, + "loss": 3.6776, + "step": 913 + }, + { + "epoch": 0.028193669534102697, + "grad_norm": 0.33153173327445984, + "learning_rate": 1.4096236890808143e-05, + "loss": 3.7039, + "step": 914 + }, + { + "epoch": 0.02822451599967611, + "grad_norm": 0.2629305124282837, + "learning_rate": 1.4111659469463295e-05, + "loss": 3.6919, + "step": 915 + }, + { + "epoch": 0.02825536246524953, + "grad_norm": 0.27931731939315796, + "learning_rate": 1.4127082048118446e-05, + "loss": 3.7014, + "step": 916 + }, + { + "epoch": 0.028286208930822947, + "grad_norm": 0.24561835825443268, + "learning_rate": 1.4142504626773598e-05, + "loss": 3.7136, + "step": 917 + }, + { + "epoch": 0.02831705539639636, + "grad_norm": 0.2790784537792206, + "learning_rate": 1.4157927205428747e-05, + "loss": 3.7119, + "step": 918 + }, + { + "epoch": 0.02834790186196978, + "grad_norm": 0.26437896490097046, + "learning_rate": 1.41733497840839e-05, + "loss": 3.694, + "step": 919 + }, + { + "epoch": 0.028378748327543196, + "grad_norm": 0.295870304107666, + "learning_rate": 1.418877236273905e-05, + "loss": 3.7099, + "step": 920 + }, + { + "epoch": 0.02840959479311661, + "grad_norm": 0.25946521759033203, + "learning_rate": 1.4204194941394202e-05, + "loss": 3.6978, + "step": 921 + }, + { + "epoch": 0.028440441258690028, + "grad_norm": 0.2986140847206116, + "learning_rate": 1.4219617520049353e-05, + "loss": 3.6867, + "step": 922 + }, + { + "epoch": 0.028471287724263446, + "grad_norm": 0.27081575989723206, + "learning_rate": 1.4235040098704505e-05, + "loss": 3.7493, + "step": 923 + }, + { + "epoch": 0.02850213418983686, + "grad_norm": 0.28031253814697266, + "learning_rate": 1.4250462677359655e-05, + "loss": 3.6822, + "step": 924 + }, + { + "epoch": 0.028532980655410278, + "grad_norm": 0.26246970891952515, + "learning_rate": 1.4265885256014808e-05, + "loss": 3.6957, + "step": 925 + }, + { + "epoch": 0.028563827120983695, + "grad_norm": 0.2817685604095459, + "learning_rate": 1.4281307834669957e-05, + "loss": 3.6731, + "step": 926 + }, + { + "epoch": 0.02859467358655711, + "grad_norm": 0.6078115105628967, + "learning_rate": 1.4296730413325107e-05, + "loss": 3.6757, + "step": 927 + }, + { + "epoch": 0.028625520052130527, + "grad_norm": 0.2579318881034851, + "learning_rate": 1.431215299198026e-05, + "loss": 3.7101, + "step": 928 + }, + { + "epoch": 0.028656366517703945, + "grad_norm": 0.30226069688796997, + "learning_rate": 1.432757557063541e-05, + "loss": 3.7097, + "step": 929 + }, + { + "epoch": 0.02868721298327736, + "grad_norm": 0.2638665735721588, + "learning_rate": 1.4342998149290562e-05, + "loss": 3.7063, + "step": 930 + }, + { + "epoch": 0.028718059448850777, + "grad_norm": 0.2707088589668274, + "learning_rate": 1.4358420727945713e-05, + "loss": 3.7343, + "step": 931 + }, + { + "epoch": 0.028748905914424194, + "grad_norm": 0.2567636966705322, + "learning_rate": 1.4373843306600865e-05, + "loss": 3.7262, + "step": 932 + }, + { + "epoch": 0.02877975237999761, + "grad_norm": 0.3092184066772461, + "learning_rate": 1.4389265885256014e-05, + "loss": 3.6844, + "step": 933 + }, + { + "epoch": 0.028810598845571026, + "grad_norm": 0.2806922495365143, + "learning_rate": 1.4404688463911167e-05, + "loss": 3.7048, + "step": 934 + }, + { + "epoch": 0.028841445311144444, + "grad_norm": 0.2572122812271118, + "learning_rate": 1.4420111042566317e-05, + "loss": 3.6283, + "step": 935 + }, + { + "epoch": 0.028872291776717858, + "grad_norm": 0.29012930393218994, + "learning_rate": 1.443553362122147e-05, + "loss": 3.6596, + "step": 936 + }, + { + "epoch": 0.028903138242291276, + "grad_norm": 0.2595019042491913, + "learning_rate": 1.445095619987662e-05, + "loss": 3.6794, + "step": 937 + }, + { + "epoch": 0.028933984707864693, + "grad_norm": 0.2839815616607666, + "learning_rate": 1.4466378778531772e-05, + "loss": 3.7148, + "step": 938 + }, + { + "epoch": 0.028964831173438108, + "grad_norm": 0.24855473637580872, + "learning_rate": 1.4481801357186921e-05, + "loss": 3.6817, + "step": 939 + }, + { + "epoch": 0.028995677639011525, + "grad_norm": 0.28319910168647766, + "learning_rate": 1.4497223935842075e-05, + "loss": 3.7053, + "step": 940 + }, + { + "epoch": 0.029026524104584943, + "grad_norm": 0.24601833522319794, + "learning_rate": 1.4512646514497224e-05, + "loss": 3.6729, + "step": 941 + }, + { + "epoch": 0.029057370570158357, + "grad_norm": 0.2708101272583008, + "learning_rate": 1.4528069093152375e-05, + "loss": 3.6935, + "step": 942 + }, + { + "epoch": 0.029088217035731775, + "grad_norm": 0.24636927247047424, + "learning_rate": 1.4543491671807527e-05, + "loss": 3.6613, + "step": 943 + }, + { + "epoch": 0.029119063501305192, + "grad_norm": 0.26215115189552307, + "learning_rate": 1.4558914250462678e-05, + "loss": 3.6977, + "step": 944 + }, + { + "epoch": 0.029149909966878607, + "grad_norm": 0.24626219272613525, + "learning_rate": 1.457433682911783e-05, + "loss": 3.6926, + "step": 945 + }, + { + "epoch": 0.029180756432452024, + "grad_norm": 0.2676321268081665, + "learning_rate": 1.4589759407772979e-05, + "loss": 3.7192, + "step": 946 + }, + { + "epoch": 0.029211602898025442, + "grad_norm": 0.2934280037879944, + "learning_rate": 1.4605181986428131e-05, + "loss": 3.7198, + "step": 947 + }, + { + "epoch": 0.029242449363598856, + "grad_norm": 0.24768953025341034, + "learning_rate": 1.4620604565083282e-05, + "loss": 3.663, + "step": 948 + }, + { + "epoch": 0.029273295829172274, + "grad_norm": 0.2797356843948364, + "learning_rate": 1.4636027143738434e-05, + "loss": 3.707, + "step": 949 + }, + { + "epoch": 0.02930414229474569, + "grad_norm": 0.29276788234710693, + "learning_rate": 1.4651449722393585e-05, + "loss": 3.6984, + "step": 950 + }, + { + "epoch": 0.029334988760319106, + "grad_norm": 0.2631774842739105, + "learning_rate": 1.4666872301048737e-05, + "loss": 3.6876, + "step": 951 + }, + { + "epoch": 0.029365835225892523, + "grad_norm": 0.2560730278491974, + "learning_rate": 1.4682294879703888e-05, + "loss": 3.6803, + "step": 952 + }, + { + "epoch": 0.02939668169146594, + "grad_norm": 0.27775639295578003, + "learning_rate": 1.469771745835904e-05, + "loss": 3.6859, + "step": 953 + }, + { + "epoch": 0.029427528157039355, + "grad_norm": 0.2647665739059448, + "learning_rate": 1.4713140037014189e-05, + "loss": 3.6843, + "step": 954 + }, + { + "epoch": 0.029458374622612773, + "grad_norm": 0.2986468970775604, + "learning_rate": 1.4728562615669341e-05, + "loss": 3.6612, + "step": 955 + }, + { + "epoch": 0.02948922108818619, + "grad_norm": 0.24196229875087738, + "learning_rate": 1.4743985194324492e-05, + "loss": 3.682, + "step": 956 + }, + { + "epoch": 0.029520067553759605, + "grad_norm": 0.2793794870376587, + "learning_rate": 1.4759407772979644e-05, + "loss": 3.718, + "step": 957 + }, + { + "epoch": 0.029550914019333022, + "grad_norm": 0.25466838479042053, + "learning_rate": 1.4774830351634795e-05, + "loss": 3.679, + "step": 958 + }, + { + "epoch": 0.02958176048490644, + "grad_norm": 0.24869388341903687, + "learning_rate": 1.4790252930289943e-05, + "loss": 3.708, + "step": 959 + }, + { + "epoch": 0.029612606950479854, + "grad_norm": 0.27537330985069275, + "learning_rate": 1.4805675508945097e-05, + "loss": 3.6793, + "step": 960 + }, + { + "epoch": 0.029643453416053272, + "grad_norm": 0.2654569149017334, + "learning_rate": 1.4821098087600246e-05, + "loss": 3.676, + "step": 961 + }, + { + "epoch": 0.02967429988162669, + "grad_norm": 0.25466564297676086, + "learning_rate": 1.4836520666255399e-05, + "loss": 3.6548, + "step": 962 + }, + { + "epoch": 0.029705146347200104, + "grad_norm": 0.25466087460517883, + "learning_rate": 1.485194324491055e-05, + "loss": 3.6767, + "step": 963 + }, + { + "epoch": 0.02973599281277352, + "grad_norm": 0.27027562260627747, + "learning_rate": 1.4867365823565701e-05, + "loss": 3.6925, + "step": 964 + }, + { + "epoch": 0.02976683927834694, + "grad_norm": 0.25940579175949097, + "learning_rate": 1.4882788402220852e-05, + "loss": 3.6703, + "step": 965 + }, + { + "epoch": 0.029797685743920353, + "grad_norm": 0.24430632591247559, + "learning_rate": 1.4898210980876004e-05, + "loss": 3.6478, + "step": 966 + }, + { + "epoch": 0.02982853220949377, + "grad_norm": 0.270147442817688, + "learning_rate": 1.4913633559531153e-05, + "loss": 3.6912, + "step": 967 + }, + { + "epoch": 0.02985937867506719, + "grad_norm": 0.2524145543575287, + "learning_rate": 1.4929056138186307e-05, + "loss": 3.6578, + "step": 968 + }, + { + "epoch": 0.029890225140640603, + "grad_norm": 0.2540969252586365, + "learning_rate": 1.4944478716841456e-05, + "loss": 3.6843, + "step": 969 + }, + { + "epoch": 0.02992107160621402, + "grad_norm": 0.26526594161987305, + "learning_rate": 1.4959901295496608e-05, + "loss": 3.6856, + "step": 970 + }, + { + "epoch": 0.029951918071787438, + "grad_norm": 0.27158161997795105, + "learning_rate": 1.4975323874151759e-05, + "loss": 3.6697, + "step": 971 + }, + { + "epoch": 0.029982764537360852, + "grad_norm": 0.25596413016319275, + "learning_rate": 1.4990746452806911e-05, + "loss": 3.7073, + "step": 972 + }, + { + "epoch": 0.03001361100293427, + "grad_norm": 0.26933181285858154, + "learning_rate": 1.5006169031462062e-05, + "loss": 3.6801, + "step": 973 + }, + { + "epoch": 0.030044457468507688, + "grad_norm": 0.254189670085907, + "learning_rate": 1.5021591610117211e-05, + "loss": 3.6876, + "step": 974 + }, + { + "epoch": 0.030075303934081102, + "grad_norm": 0.28820884227752686, + "learning_rate": 1.5037014188772363e-05, + "loss": 3.692, + "step": 975 + }, + { + "epoch": 0.03010615039965452, + "grad_norm": 0.2732674479484558, + "learning_rate": 1.5052436767427514e-05, + "loss": 3.677, + "step": 976 + }, + { + "epoch": 0.030136996865227937, + "grad_norm": 0.24574854969978333, + "learning_rate": 1.5067859346082666e-05, + "loss": 3.7034, + "step": 977 + }, + { + "epoch": 0.03016784333080135, + "grad_norm": 0.25243595242500305, + "learning_rate": 1.5083281924737817e-05, + "loss": 3.6652, + "step": 978 + }, + { + "epoch": 0.03019868979637477, + "grad_norm": 0.2572830021381378, + "learning_rate": 1.5098704503392969e-05, + "loss": 3.6252, + "step": 979 + }, + { + "epoch": 0.030229536261948187, + "grad_norm": 0.25325626134872437, + "learning_rate": 1.5114127082048118e-05, + "loss": 3.6678, + "step": 980 + }, + { + "epoch": 0.0302603827275216, + "grad_norm": 0.2620638906955719, + "learning_rate": 1.5129549660703272e-05, + "loss": 3.6547, + "step": 981 + }, + { + "epoch": 0.03029122919309502, + "grad_norm": 0.25061291456222534, + "learning_rate": 1.514497223935842e-05, + "loss": 3.656, + "step": 982 + }, + { + "epoch": 0.030322075658668436, + "grad_norm": 0.40104347467422485, + "learning_rate": 1.5160394818013573e-05, + "loss": 3.7118, + "step": 983 + }, + { + "epoch": 0.03035292212424185, + "grad_norm": 0.2893005609512329, + "learning_rate": 1.5175817396668724e-05, + "loss": 3.6635, + "step": 984 + }, + { + "epoch": 0.030383768589815268, + "grad_norm": 0.24886508285999298, + "learning_rate": 1.5191239975323876e-05, + "loss": 3.6525, + "step": 985 + }, + { + "epoch": 0.030414615055388686, + "grad_norm": 0.350188285112381, + "learning_rate": 1.5206662553979027e-05, + "loss": 3.6628, + "step": 986 + }, + { + "epoch": 0.0304454615209621, + "grad_norm": 0.2784472405910492, + "learning_rate": 1.5222085132634179e-05, + "loss": 3.6729, + "step": 987 + }, + { + "epoch": 0.030476307986535518, + "grad_norm": 0.26195716857910156, + "learning_rate": 1.5237507711289328e-05, + "loss": 3.6491, + "step": 988 + }, + { + "epoch": 0.030507154452108935, + "grad_norm": 0.2760872542858124, + "learning_rate": 1.5252930289944478e-05, + "loss": 3.664, + "step": 989 + }, + { + "epoch": 0.03053800091768235, + "grad_norm": 0.25018903613090515, + "learning_rate": 1.5268352868599632e-05, + "loss": 3.6483, + "step": 990 + }, + { + "epoch": 0.030568847383255767, + "grad_norm": 0.3006135821342468, + "learning_rate": 1.528377544725478e-05, + "loss": 3.659, + "step": 991 + }, + { + "epoch": 0.030599693848829185, + "grad_norm": 0.27858635783195496, + "learning_rate": 1.5299198025909934e-05, + "loss": 3.6459, + "step": 992 + }, + { + "epoch": 0.0306305403144026, + "grad_norm": 0.30452170968055725, + "learning_rate": 1.5314620604565082e-05, + "loss": 3.6259, + "step": 993 + }, + { + "epoch": 0.030661386779976017, + "grad_norm": 0.2777387201786041, + "learning_rate": 1.5330043183220235e-05, + "loss": 3.6914, + "step": 994 + }, + { + "epoch": 0.030692233245549434, + "grad_norm": 0.30385351181030273, + "learning_rate": 1.5345465761875387e-05, + "loss": 3.6797, + "step": 995 + }, + { + "epoch": 0.03072307971112285, + "grad_norm": 0.26444870233535767, + "learning_rate": 1.536088834053054e-05, + "loss": 3.662, + "step": 996 + }, + { + "epoch": 0.030753926176696266, + "grad_norm": 0.30009475350379944, + "learning_rate": 1.5376310919185688e-05, + "loss": 3.6327, + "step": 997 + }, + { + "epoch": 0.030784772642269684, + "grad_norm": 0.30921223759651184, + "learning_rate": 1.539173349784084e-05, + "loss": 3.7028, + "step": 998 + }, + { + "epoch": 0.030815619107843098, + "grad_norm": 0.24391913414001465, + "learning_rate": 1.540715607649599e-05, + "loss": 3.6592, + "step": 999 + }, + { + "epoch": 0.030846465573416516, + "grad_norm": 0.30715638399124146, + "learning_rate": 1.5422578655151142e-05, + "loss": 3.6826, + "step": 1000 + }, + { + "epoch": 0.030877312038989933, + "grad_norm": 0.2625090777873993, + "learning_rate": 1.5438001233806294e-05, + "loss": 3.7091, + "step": 1001 + }, + { + "epoch": 0.030908158504563348, + "grad_norm": 0.3016837537288666, + "learning_rate": 1.5453423812461446e-05, + "loss": 3.6842, + "step": 1002 + }, + { + "epoch": 0.030939004970136765, + "grad_norm": 0.27689453959465027, + "learning_rate": 1.5468846391116595e-05, + "loss": 3.67, + "step": 1003 + }, + { + "epoch": 0.030969851435710183, + "grad_norm": 0.2897111475467682, + "learning_rate": 1.5484268969771744e-05, + "loss": 3.7152, + "step": 1004 + }, + { + "epoch": 0.031000697901283597, + "grad_norm": 0.25096049904823303, + "learning_rate": 1.5499691548426896e-05, + "loss": 3.6601, + "step": 1005 + }, + { + "epoch": 0.031031544366857015, + "grad_norm": 0.29074880480766296, + "learning_rate": 1.551511412708205e-05, + "loss": 3.6387, + "step": 1006 + }, + { + "epoch": 0.031062390832430432, + "grad_norm": 0.2553057074546814, + "learning_rate": 1.55305367057372e-05, + "loss": 3.688, + "step": 1007 + }, + { + "epoch": 0.031093237298003847, + "grad_norm": 0.2785859704017639, + "learning_rate": 1.554595928439235e-05, + "loss": 3.6576, + "step": 1008 + }, + { + "epoch": 0.031124083763577264, + "grad_norm": 0.2730870246887207, + "learning_rate": 1.5561381863047502e-05, + "loss": 3.6793, + "step": 1009 + }, + { + "epoch": 0.031154930229150682, + "grad_norm": 0.26524606347084045, + "learning_rate": 1.557680444170265e-05, + "loss": 3.5895, + "step": 1010 + }, + { + "epoch": 0.031185776694724096, + "grad_norm": 0.24456842243671417, + "learning_rate": 1.5592227020357807e-05, + "loss": 3.6802, + "step": 1011 + }, + { + "epoch": 0.031216623160297514, + "grad_norm": 0.2679893374443054, + "learning_rate": 1.5607649599012956e-05, + "loss": 3.6839, + "step": 1012 + }, + { + "epoch": 0.03124746962587093, + "grad_norm": 0.25774675607681274, + "learning_rate": 1.5623072177668108e-05, + "loss": 3.6516, + "step": 1013 + }, + { + "epoch": 0.031278316091444346, + "grad_norm": 0.26357972621917725, + "learning_rate": 1.5638494756323257e-05, + "loss": 3.7191, + "step": 1014 + }, + { + "epoch": 0.03130916255701777, + "grad_norm": 0.29672062397003174, + "learning_rate": 1.565391733497841e-05, + "loss": 3.6237, + "step": 1015 + }, + { + "epoch": 0.03134000902259118, + "grad_norm": 0.2386469691991806, + "learning_rate": 1.566933991363356e-05, + "loss": 3.6662, + "step": 1016 + }, + { + "epoch": 0.031370855488164595, + "grad_norm": 0.28304237127304077, + "learning_rate": 1.5684762492288714e-05, + "loss": 3.6457, + "step": 1017 + }, + { + "epoch": 0.031401701953738016, + "grad_norm": 0.2777099609375, + "learning_rate": 1.5700185070943863e-05, + "loss": 3.6921, + "step": 1018 + }, + { + "epoch": 0.03143254841931143, + "grad_norm": 0.25186648964881897, + "learning_rate": 1.571560764959901e-05, + "loss": 3.7028, + "step": 1019 + }, + { + "epoch": 0.031463394884884845, + "grad_norm": 0.25551241636276245, + "learning_rate": 1.5731030228254164e-05, + "loss": 3.6683, + "step": 1020 + }, + { + "epoch": 0.031494241350458266, + "grad_norm": 0.26814645528793335, + "learning_rate": 1.5746452806909316e-05, + "loss": 3.6219, + "step": 1021 + }, + { + "epoch": 0.03152508781603168, + "grad_norm": 0.23856498301029205, + "learning_rate": 1.576187538556447e-05, + "loss": 3.6472, + "step": 1022 + }, + { + "epoch": 0.031555934281605094, + "grad_norm": 0.2672717869281769, + "learning_rate": 1.5777297964219617e-05, + "loss": 3.6072, + "step": 1023 + }, + { + "epoch": 0.031586780747178515, + "grad_norm": 0.24357722699642181, + "learning_rate": 1.579272054287477e-05, + "loss": 3.6386, + "step": 1024 + }, + { + "epoch": 0.03161762721275193, + "grad_norm": 0.25208452343940735, + "learning_rate": 1.580814312152992e-05, + "loss": 3.6146, + "step": 1025 + }, + { + "epoch": 0.031648473678325344, + "grad_norm": 0.2432018369436264, + "learning_rate": 1.582356570018507e-05, + "loss": 3.6251, + "step": 1026 + }, + { + "epoch": 0.031679320143898765, + "grad_norm": 0.2613210082054138, + "learning_rate": 1.5838988278840223e-05, + "loss": 3.6245, + "step": 1027 + }, + { + "epoch": 0.03171016660947218, + "grad_norm": 0.24120832979679108, + "learning_rate": 1.5854410857495375e-05, + "loss": 3.6213, + "step": 1028 + }, + { + "epoch": 0.03174101307504559, + "grad_norm": 0.2843015491962433, + "learning_rate": 1.5869833436150524e-05, + "loss": 3.6316, + "step": 1029 + }, + { + "epoch": 0.031771859540619014, + "grad_norm": 0.24481900036334991, + "learning_rate": 1.5885256014805677e-05, + "loss": 3.6186, + "step": 1030 + }, + { + "epoch": 0.03180270600619243, + "grad_norm": 0.27210554480552673, + "learning_rate": 1.5900678593460826e-05, + "loss": 3.6299, + "step": 1031 + }, + { + "epoch": 0.03183355247176584, + "grad_norm": 0.26963165402412415, + "learning_rate": 1.591610117211598e-05, + "loss": 3.6574, + "step": 1032 + }, + { + "epoch": 0.031864398937339264, + "grad_norm": 0.2450762391090393, + "learning_rate": 1.593152375077113e-05, + "loss": 3.6569, + "step": 1033 + }, + { + "epoch": 0.03189524540291268, + "grad_norm": 0.2657122313976288, + "learning_rate": 1.5946946329426282e-05, + "loss": 3.6499, + "step": 1034 + }, + { + "epoch": 0.03192609186848609, + "grad_norm": 0.23751893639564514, + "learning_rate": 1.596236890808143e-05, + "loss": 3.6315, + "step": 1035 + }, + { + "epoch": 0.03195693833405951, + "grad_norm": 0.2657988965511322, + "learning_rate": 1.597779148673658e-05, + "loss": 3.6401, + "step": 1036 + }, + { + "epoch": 0.03198778479963293, + "grad_norm": 0.23756220936775208, + "learning_rate": 1.5993214065391736e-05, + "loss": 3.6491, + "step": 1037 + }, + { + "epoch": 0.03201863126520634, + "grad_norm": 0.4764891564846039, + "learning_rate": 1.6008636644046885e-05, + "loss": 3.6593, + "step": 1038 + }, + { + "epoch": 0.03204947773077976, + "grad_norm": 0.25385531783103943, + "learning_rate": 1.6024059222702037e-05, + "loss": 3.63, + "step": 1039 + }, + { + "epoch": 0.03208032419635318, + "grad_norm": 0.3246819078922272, + "learning_rate": 1.6039481801357186e-05, + "loss": 3.6364, + "step": 1040 + }, + { + "epoch": 0.03211117066192659, + "grad_norm": 0.26391276717185974, + "learning_rate": 1.605490438001234e-05, + "loss": 3.6599, + "step": 1041 + }, + { + "epoch": 0.03214201712750001, + "grad_norm": 0.6525307297706604, + "learning_rate": 1.607032695866749e-05, + "loss": 3.636, + "step": 1042 + }, + { + "epoch": 0.03217286359307343, + "grad_norm": 0.2657068073749542, + "learning_rate": 1.6085749537322643e-05, + "loss": 3.668, + "step": 1043 + }, + { + "epoch": 0.03220371005864684, + "grad_norm": 0.260219931602478, + "learning_rate": 1.6101172115977792e-05, + "loss": 3.6506, + "step": 1044 + }, + { + "epoch": 0.03223455652422026, + "grad_norm": 0.25239136815071106, + "learning_rate": 1.6116594694632944e-05, + "loss": 3.6283, + "step": 1045 + }, + { + "epoch": 0.032265402989793676, + "grad_norm": 0.24527081847190857, + "learning_rate": 1.6132017273288093e-05, + "loss": 3.6385, + "step": 1046 + }, + { + "epoch": 0.03229624945536709, + "grad_norm": 0.24446243047714233, + "learning_rate": 1.6147439851943245e-05, + "loss": 3.6475, + "step": 1047 + }, + { + "epoch": 0.03232709592094051, + "grad_norm": 0.24436137080192566, + "learning_rate": 1.6162862430598398e-05, + "loss": 3.6424, + "step": 1048 + }, + { + "epoch": 0.032357942386513926, + "grad_norm": 0.26519346237182617, + "learning_rate": 1.617828500925355e-05, + "loss": 3.6437, + "step": 1049 + }, + { + "epoch": 0.03238878885208734, + "grad_norm": 0.25255683064460754, + "learning_rate": 1.61937075879087e-05, + "loss": 3.6307, + "step": 1050 + }, + { + "epoch": 0.03241963531766076, + "grad_norm": 0.2429829239845276, + "learning_rate": 1.6209130166563848e-05, + "loss": 3.6625, + "step": 1051 + }, + { + "epoch": 0.032450481783234175, + "grad_norm": 0.26336541771888733, + "learning_rate": 1.6224552745219e-05, + "loss": 3.6669, + "step": 1052 + }, + { + "epoch": 0.03248132824880759, + "grad_norm": 0.2534208595752716, + "learning_rate": 1.6239975323874152e-05, + "loss": 3.5895, + "step": 1053 + }, + { + "epoch": 0.03251217471438101, + "grad_norm": 0.23858462274074554, + "learning_rate": 1.6255397902529305e-05, + "loss": 3.6433, + "step": 1054 + }, + { + "epoch": 0.032543021179954425, + "grad_norm": 0.25510114431381226, + "learning_rate": 1.6270820481184453e-05, + "loss": 3.6019, + "step": 1055 + }, + { + "epoch": 0.03257386764552784, + "grad_norm": 0.24468813836574554, + "learning_rate": 1.6286243059839606e-05, + "loss": 3.6275, + "step": 1056 + }, + { + "epoch": 0.03260471411110126, + "grad_norm": 0.24675990641117096, + "learning_rate": 1.6301665638494755e-05, + "loss": 3.6118, + "step": 1057 + }, + { + "epoch": 0.032635560576674674, + "grad_norm": 0.24188446998596191, + "learning_rate": 1.631708821714991e-05, + "loss": 3.6115, + "step": 1058 + }, + { + "epoch": 0.03266640704224809, + "grad_norm": 0.24869203567504883, + "learning_rate": 1.633251079580506e-05, + "loss": 3.612, + "step": 1059 + }, + { + "epoch": 0.03269725350782151, + "grad_norm": 0.2719830274581909, + "learning_rate": 1.634793337446021e-05, + "loss": 3.612, + "step": 1060 + }, + { + "epoch": 0.032728099973394924, + "grad_norm": 0.24006858468055725, + "learning_rate": 1.636335595311536e-05, + "loss": 3.6244, + "step": 1061 + }, + { + "epoch": 0.03275894643896834, + "grad_norm": 0.2772569954395294, + "learning_rate": 1.6378778531770513e-05, + "loss": 3.6332, + "step": 1062 + }, + { + "epoch": 0.03278979290454176, + "grad_norm": 0.30463480949401855, + "learning_rate": 1.6394201110425665e-05, + "loss": 3.6917, + "step": 1063 + }, + { + "epoch": 0.03282063937011517, + "grad_norm": 0.24636127054691315, + "learning_rate": 1.6409623689080817e-05, + "loss": 3.6532, + "step": 1064 + }, + { + "epoch": 0.03285148583568859, + "grad_norm": 0.25556159019470215, + "learning_rate": 1.6425046267735966e-05, + "loss": 3.6279, + "step": 1065 + }, + { + "epoch": 0.03288233230126201, + "grad_norm": 0.26886314153671265, + "learning_rate": 1.6440468846391115e-05, + "loss": 3.6558, + "step": 1066 + }, + { + "epoch": 0.03291317876683542, + "grad_norm": 0.24051351845264435, + "learning_rate": 1.6455891425046267e-05, + "loss": 3.6092, + "step": 1067 + }, + { + "epoch": 0.03294402523240884, + "grad_norm": 0.2625601887702942, + "learning_rate": 1.647131400370142e-05, + "loss": 3.6527, + "step": 1068 + }, + { + "epoch": 0.03297487169798226, + "grad_norm": 0.2678801417350769, + "learning_rate": 1.6486736582356572e-05, + "loss": 3.6226, + "step": 1069 + }, + { + "epoch": 0.03300571816355567, + "grad_norm": 0.25998008251190186, + "learning_rate": 1.650215916101172e-05, + "loss": 3.665, + "step": 1070 + }, + { + "epoch": 0.03303656462912909, + "grad_norm": 0.27184319496154785, + "learning_rate": 1.6517581739666873e-05, + "loss": 3.6095, + "step": 1071 + }, + { + "epoch": 0.03306741109470251, + "grad_norm": 0.24924013018608093, + "learning_rate": 1.6533004318322022e-05, + "loss": 3.6229, + "step": 1072 + }, + { + "epoch": 0.03309825756027592, + "grad_norm": 0.26190465688705444, + "learning_rate": 1.6548426896977174e-05, + "loss": 3.6239, + "step": 1073 + }, + { + "epoch": 0.033129104025849336, + "grad_norm": 0.2987116575241089, + "learning_rate": 1.6563849475632327e-05, + "loss": 3.6209, + "step": 1074 + }, + { + "epoch": 0.03315995049142276, + "grad_norm": 0.23914960026741028, + "learning_rate": 1.657927205428748e-05, + "loss": 3.6415, + "step": 1075 + }, + { + "epoch": 0.03319079695699617, + "grad_norm": 0.2802031338214874, + "learning_rate": 1.6594694632942628e-05, + "loss": 3.6485, + "step": 1076 + }, + { + "epoch": 0.033221643422569586, + "grad_norm": 0.25585564970970154, + "learning_rate": 1.661011721159778e-05, + "loss": 3.6394, + "step": 1077 + }, + { + "epoch": 0.03325248988814301, + "grad_norm": 0.2521415054798126, + "learning_rate": 1.6625539790252933e-05, + "loss": 3.6395, + "step": 1078 + }, + { + "epoch": 0.03328333635371642, + "grad_norm": 0.25594228506088257, + "learning_rate": 1.6640962368908085e-05, + "loss": 3.6519, + "step": 1079 + }, + { + "epoch": 0.033314182819289835, + "grad_norm": 0.252485454082489, + "learning_rate": 1.6656384947563234e-05, + "loss": 3.625, + "step": 1080 + }, + { + "epoch": 0.033345029284863256, + "grad_norm": 0.23983298242092133, + "learning_rate": 1.6671807526218383e-05, + "loss": 3.6465, + "step": 1081 + }, + { + "epoch": 0.03337587575043667, + "grad_norm": 0.2741076946258545, + "learning_rate": 1.6687230104873535e-05, + "loss": 3.6624, + "step": 1082 + }, + { + "epoch": 0.033406722216010085, + "grad_norm": 0.24161919951438904, + "learning_rate": 1.6702652683528687e-05, + "loss": 3.6143, + "step": 1083 + }, + { + "epoch": 0.033437568681583506, + "grad_norm": 0.24817213416099548, + "learning_rate": 1.671807526218384e-05, + "loss": 3.5755, + "step": 1084 + }, + { + "epoch": 0.03346841514715692, + "grad_norm": 0.2342797815799713, + "learning_rate": 1.673349784083899e-05, + "loss": 3.6184, + "step": 1085 + }, + { + "epoch": 0.033499261612730334, + "grad_norm": 0.2644317150115967, + "learning_rate": 1.674892041949414e-05, + "loss": 3.6678, + "step": 1086 + }, + { + "epoch": 0.033530108078303755, + "grad_norm": 0.2590143084526062, + "learning_rate": 1.676434299814929e-05, + "loss": 3.6289, + "step": 1087 + }, + { + "epoch": 0.03356095454387717, + "grad_norm": 0.23853063583374023, + "learning_rate": 1.6779765576804442e-05, + "loss": 3.659, + "step": 1088 + }, + { + "epoch": 0.033591801009450584, + "grad_norm": 0.2801627218723297, + "learning_rate": 1.6795188155459594e-05, + "loss": 3.607, + "step": 1089 + }, + { + "epoch": 0.033622647475024005, + "grad_norm": 0.2631864845752716, + "learning_rate": 1.6810610734114747e-05, + "loss": 3.5919, + "step": 1090 + }, + { + "epoch": 0.03365349394059742, + "grad_norm": 0.24828413128852844, + "learning_rate": 1.6826033312769895e-05, + "loss": 3.637, + "step": 1091 + }, + { + "epoch": 0.03368434040617083, + "grad_norm": 0.28948974609375, + "learning_rate": 1.6841455891425048e-05, + "loss": 3.6888, + "step": 1092 + }, + { + "epoch": 0.033715186871744254, + "grad_norm": 0.26564595103263855, + "learning_rate": 1.6856878470080197e-05, + "loss": 3.6571, + "step": 1093 + }, + { + "epoch": 0.03374603333731767, + "grad_norm": 0.29195690155029297, + "learning_rate": 1.687230104873535e-05, + "loss": 3.6231, + "step": 1094 + }, + { + "epoch": 0.03377687980289108, + "grad_norm": 0.26382818818092346, + "learning_rate": 1.68877236273905e-05, + "loss": 3.6488, + "step": 1095 + }, + { + "epoch": 0.033807726268464504, + "grad_norm": 0.2809063494205475, + "learning_rate": 1.6903146206045653e-05, + "loss": 3.6077, + "step": 1096 + }, + { + "epoch": 0.03383857273403792, + "grad_norm": 0.2656170427799225, + "learning_rate": 1.6918568784700802e-05, + "loss": 3.6453, + "step": 1097 + }, + { + "epoch": 0.03386941919961133, + "grad_norm": 0.28957322239875793, + "learning_rate": 1.693399136335595e-05, + "loss": 3.6268, + "step": 1098 + }, + { + "epoch": 0.03390026566518475, + "grad_norm": 0.24462205171585083, + "learning_rate": 1.6949413942011107e-05, + "loss": 3.6345, + "step": 1099 + }, + { + "epoch": 0.03393111213075817, + "grad_norm": 0.271310955286026, + "learning_rate": 1.6964836520666256e-05, + "loss": 3.6336, + "step": 1100 + }, + { + "epoch": 0.03396195859633158, + "grad_norm": 0.2429295927286148, + "learning_rate": 1.6980259099321408e-05, + "loss": 3.6318, + "step": 1101 + }, + { + "epoch": 0.033992805061905, + "grad_norm": 0.24515590071678162, + "learning_rate": 1.6995681677976557e-05, + "loss": 3.5992, + "step": 1102 + }, + { + "epoch": 0.03402365152747842, + "grad_norm": 0.25581103563308716, + "learning_rate": 1.701110425663171e-05, + "loss": 3.6145, + "step": 1103 + }, + { + "epoch": 0.03405449799305183, + "grad_norm": 0.24518103897571564, + "learning_rate": 1.702652683528686e-05, + "loss": 3.6212, + "step": 1104 + }, + { + "epoch": 0.03408534445862525, + "grad_norm": 0.22654704749584198, + "learning_rate": 1.7041949413942014e-05, + "loss": 3.5945, + "step": 1105 + }, + { + "epoch": 0.03411619092419867, + "grad_norm": 0.26193296909332275, + "learning_rate": 1.7057371992597163e-05, + "loss": 3.6226, + "step": 1106 + }, + { + "epoch": 0.03414703738977208, + "grad_norm": 0.2511600852012634, + "learning_rate": 1.7072794571252315e-05, + "loss": 3.6396, + "step": 1107 + }, + { + "epoch": 0.0341778838553455, + "grad_norm": 0.259863018989563, + "learning_rate": 1.7088217149907464e-05, + "loss": 3.647, + "step": 1108 + }, + { + "epoch": 0.034208730320918916, + "grad_norm": 0.2640058696269989, + "learning_rate": 1.7103639728562616e-05, + "loss": 3.6652, + "step": 1109 + }, + { + "epoch": 0.03423957678649233, + "grad_norm": 0.23210087418556213, + "learning_rate": 1.711906230721777e-05, + "loss": 3.6268, + "step": 1110 + }, + { + "epoch": 0.03427042325206575, + "grad_norm": 0.2683172821998596, + "learning_rate": 1.713448488587292e-05, + "loss": 3.6358, + "step": 1111 + }, + { + "epoch": 0.034301269717639166, + "grad_norm": 0.222483828663826, + "learning_rate": 1.714990746452807e-05, + "loss": 3.6178, + "step": 1112 + }, + { + "epoch": 0.03433211618321258, + "grad_norm": 0.22700117528438568, + "learning_rate": 1.716533004318322e-05, + "loss": 3.605, + "step": 1113 + }, + { + "epoch": 0.034362962648786, + "grad_norm": 0.23491081595420837, + "learning_rate": 1.718075262183837e-05, + "loss": 3.6086, + "step": 1114 + }, + { + "epoch": 0.034393809114359415, + "grad_norm": 0.2399303913116455, + "learning_rate": 1.7196175200493523e-05, + "loss": 3.6029, + "step": 1115 + }, + { + "epoch": 0.03442465557993283, + "grad_norm": 0.22389774024486542, + "learning_rate": 1.7211597779148676e-05, + "loss": 3.6062, + "step": 1116 + }, + { + "epoch": 0.03445550204550625, + "grad_norm": 0.24863201379776, + "learning_rate": 1.7227020357803825e-05, + "loss": 3.5859, + "step": 1117 + }, + { + "epoch": 0.034486348511079665, + "grad_norm": 0.2404118776321411, + "learning_rate": 1.7242442936458977e-05, + "loss": 3.6357, + "step": 1118 + }, + { + "epoch": 0.03451719497665308, + "grad_norm": 0.23977839946746826, + "learning_rate": 1.7257865515114126e-05, + "loss": 3.5903, + "step": 1119 + }, + { + "epoch": 0.0345480414422265, + "grad_norm": 0.23838968575000763, + "learning_rate": 1.727328809376928e-05, + "loss": 3.6324, + "step": 1120 + }, + { + "epoch": 0.034578887907799914, + "grad_norm": 0.2386476993560791, + "learning_rate": 1.728871067242443e-05, + "loss": 3.5816, + "step": 1121 + }, + { + "epoch": 0.03460973437337333, + "grad_norm": 0.2461887001991272, + "learning_rate": 1.7304133251079583e-05, + "loss": 3.6425, + "step": 1122 + }, + { + "epoch": 0.03464058083894675, + "grad_norm": 0.24141962826251984, + "learning_rate": 1.731955582973473e-05, + "loss": 3.5723, + "step": 1123 + }, + { + "epoch": 0.034671427304520164, + "grad_norm": 0.227415531873703, + "learning_rate": 1.7334978408389884e-05, + "loss": 3.5825, + "step": 1124 + }, + { + "epoch": 0.03470227377009358, + "grad_norm": 0.24212077260017395, + "learning_rate": 1.7350400987045036e-05, + "loss": 3.6168, + "step": 1125 + }, + { + "epoch": 0.034733120235667, + "grad_norm": 0.24072545766830444, + "learning_rate": 1.736582356570019e-05, + "loss": 3.6483, + "step": 1126 + }, + { + "epoch": 0.03476396670124041, + "grad_norm": 0.2578463554382324, + "learning_rate": 1.7381246144355337e-05, + "loss": 3.6445, + "step": 1127 + }, + { + "epoch": 0.03479481316681383, + "grad_norm": 0.2504970133304596, + "learning_rate": 1.7396668723010486e-05, + "loss": 3.615, + "step": 1128 + }, + { + "epoch": 0.03482565963238725, + "grad_norm": 0.24855287373065948, + "learning_rate": 1.741209130166564e-05, + "loss": 3.6076, + "step": 1129 + }, + { + "epoch": 0.03485650609796066, + "grad_norm": 0.24227218329906464, + "learning_rate": 1.742751388032079e-05, + "loss": 3.6473, + "step": 1130 + }, + { + "epoch": 0.03488735256353408, + "grad_norm": 0.25308817625045776, + "learning_rate": 1.7442936458975943e-05, + "loss": 3.6057, + "step": 1131 + }, + { + "epoch": 0.0349181990291075, + "grad_norm": 0.2329581379890442, + "learning_rate": 1.7458359037631092e-05, + "loss": 3.6105, + "step": 1132 + }, + { + "epoch": 0.03494904549468091, + "grad_norm": 0.2715435028076172, + "learning_rate": 1.7473781616286244e-05, + "loss": 3.5814, + "step": 1133 + }, + { + "epoch": 0.034979891960254327, + "grad_norm": 0.24830839037895203, + "learning_rate": 1.7489204194941393e-05, + "loss": 3.6254, + "step": 1134 + }, + { + "epoch": 0.03501073842582775, + "grad_norm": 0.25527727603912354, + "learning_rate": 1.7504626773596546e-05, + "loss": 3.5893, + "step": 1135 + }, + { + "epoch": 0.03504158489140116, + "grad_norm": 0.24462352693080902, + "learning_rate": 1.7520049352251698e-05, + "loss": 3.6232, + "step": 1136 + }, + { + "epoch": 0.035072431356974576, + "grad_norm": 0.25070035457611084, + "learning_rate": 1.753547193090685e-05, + "loss": 3.5991, + "step": 1137 + }, + { + "epoch": 0.035103277822548, + "grad_norm": 0.2614418864250183, + "learning_rate": 1.7550894509562e-05, + "loss": 3.6069, + "step": 1138 + }, + { + "epoch": 0.03513412428812141, + "grad_norm": 0.237568661570549, + "learning_rate": 1.756631708821715e-05, + "loss": 3.603, + "step": 1139 + }, + { + "epoch": 0.035164970753694826, + "grad_norm": 0.23892706632614136, + "learning_rate": 1.75817396668723e-05, + "loss": 3.584, + "step": 1140 + }, + { + "epoch": 0.03519581721926825, + "grad_norm": 0.22511692345142365, + "learning_rate": 1.7597162245527456e-05, + "loss": 3.6101, + "step": 1141 + }, + { + "epoch": 0.03522666368484166, + "grad_norm": 0.24253378808498383, + "learning_rate": 1.7612584824182605e-05, + "loss": 3.5975, + "step": 1142 + }, + { + "epoch": 0.035257510150415075, + "grad_norm": 0.23971866071224213, + "learning_rate": 1.7628007402837754e-05, + "loss": 3.5952, + "step": 1143 + }, + { + "epoch": 0.035288356615988496, + "grad_norm": 0.2371843159198761, + "learning_rate": 1.7643429981492906e-05, + "loss": 3.5835, + "step": 1144 + }, + { + "epoch": 0.03531920308156191, + "grad_norm": 0.2605656683444977, + "learning_rate": 1.7658852560148055e-05, + "loss": 3.6252, + "step": 1145 + }, + { + "epoch": 0.035350049547135325, + "grad_norm": 0.23523151874542236, + "learning_rate": 1.767427513880321e-05, + "loss": 3.6083, + "step": 1146 + }, + { + "epoch": 0.035380896012708746, + "grad_norm": 0.24945168197155, + "learning_rate": 1.768969771745836e-05, + "loss": 3.5905, + "step": 1147 + }, + { + "epoch": 0.03541174247828216, + "grad_norm": 0.25454258918762207, + "learning_rate": 1.7705120296113512e-05, + "loss": 3.6624, + "step": 1148 + }, + { + "epoch": 0.035442588943855574, + "grad_norm": 0.2623962163925171, + "learning_rate": 1.772054287476866e-05, + "loss": 3.5896, + "step": 1149 + }, + { + "epoch": 0.035473435409428995, + "grad_norm": 0.23829463124275208, + "learning_rate": 1.7735965453423813e-05, + "loss": 3.6092, + "step": 1150 + }, + { + "epoch": 0.03550428187500241, + "grad_norm": 0.23063145577907562, + "learning_rate": 1.7751388032078965e-05, + "loss": 3.5833, + "step": 1151 + }, + { + "epoch": 0.035535128340575824, + "grad_norm": 0.28522586822509766, + "learning_rate": 1.7766810610734118e-05, + "loss": 3.6264, + "step": 1152 + }, + { + "epoch": 0.035565974806149245, + "grad_norm": 0.25926533341407776, + "learning_rate": 1.7782233189389266e-05, + "loss": 3.6339, + "step": 1153 + }, + { + "epoch": 0.03559682127172266, + "grad_norm": 0.25707268714904785, + "learning_rate": 1.779765576804442e-05, + "loss": 3.6073, + "step": 1154 + }, + { + "epoch": 0.03562766773729607, + "grad_norm": 0.2491319179534912, + "learning_rate": 1.7813078346699568e-05, + "loss": 3.5707, + "step": 1155 + }, + { + "epoch": 0.035658514202869494, + "grad_norm": 0.2459113746881485, + "learning_rate": 1.782850092535472e-05, + "loss": 3.6334, + "step": 1156 + }, + { + "epoch": 0.03568936066844291, + "grad_norm": 0.2363123595714569, + "learning_rate": 1.7843923504009872e-05, + "loss": 3.5988, + "step": 1157 + }, + { + "epoch": 0.03572020713401632, + "grad_norm": 0.24272987246513367, + "learning_rate": 1.785934608266502e-05, + "loss": 3.6317, + "step": 1158 + }, + { + "epoch": 0.035751053599589744, + "grad_norm": 0.2398492842912674, + "learning_rate": 1.7874768661320173e-05, + "loss": 3.6068, + "step": 1159 + }, + { + "epoch": 0.03578190006516316, + "grad_norm": 0.2455689162015915, + "learning_rate": 1.7890191239975322e-05, + "loss": 3.5994, + "step": 1160 + }, + { + "epoch": 0.03581274653073657, + "grad_norm": 0.2311619222164154, + "learning_rate": 1.7905613818630475e-05, + "loss": 3.592, + "step": 1161 + }, + { + "epoch": 0.03584359299630999, + "grad_norm": 0.24495993554592133, + "learning_rate": 1.7921036397285627e-05, + "loss": 3.5917, + "step": 1162 + }, + { + "epoch": 0.03587443946188341, + "grad_norm": 0.23029939830303192, + "learning_rate": 1.793645897594078e-05, + "loss": 3.5882, + "step": 1163 + }, + { + "epoch": 0.03590528592745682, + "grad_norm": 0.25308966636657715, + "learning_rate": 1.7951881554595928e-05, + "loss": 3.5878, + "step": 1164 + }, + { + "epoch": 0.03593613239303024, + "grad_norm": 0.24489742517471313, + "learning_rate": 1.796730413325108e-05, + "loss": 3.5802, + "step": 1165 + }, + { + "epoch": 0.03596697885860366, + "grad_norm": 0.24986471235752106, + "learning_rate": 1.798272671190623e-05, + "loss": 3.5962, + "step": 1166 + }, + { + "epoch": 0.03599782532417707, + "grad_norm": 0.2545029819011688, + "learning_rate": 1.7998149290561385e-05, + "loss": 3.6047, + "step": 1167 + }, + { + "epoch": 0.03602867178975049, + "grad_norm": 0.26220276951789856, + "learning_rate": 1.8013571869216534e-05, + "loss": 3.6008, + "step": 1168 + }, + { + "epoch": 0.03605951825532391, + "grad_norm": 0.23879912495613098, + "learning_rate": 1.8028994447871686e-05, + "loss": 3.6196, + "step": 1169 + }, + { + "epoch": 0.03609036472089732, + "grad_norm": 0.251747190952301, + "learning_rate": 1.8044417026526835e-05, + "loss": 3.5858, + "step": 1170 + }, + { + "epoch": 0.03612121118647074, + "grad_norm": 0.22924289107322693, + "learning_rate": 1.8059839605181987e-05, + "loss": 3.5824, + "step": 1171 + }, + { + "epoch": 0.036152057652044156, + "grad_norm": 0.23638056218624115, + "learning_rate": 1.807526218383714e-05, + "loss": 3.5806, + "step": 1172 + }, + { + "epoch": 0.03618290411761757, + "grad_norm": 0.4851248562335968, + "learning_rate": 1.8090684762492292e-05, + "loss": 3.5852, + "step": 1173 + }, + { + "epoch": 0.03621375058319099, + "grad_norm": 0.22384408116340637, + "learning_rate": 1.810610734114744e-05, + "loss": 3.5515, + "step": 1174 + }, + { + "epoch": 0.036244597048764406, + "grad_norm": 0.25209060311317444, + "learning_rate": 1.812152991980259e-05, + "loss": 3.612, + "step": 1175 + }, + { + "epoch": 0.03627544351433782, + "grad_norm": 0.22604218125343323, + "learning_rate": 1.8136952498457742e-05, + "loss": 3.5933, + "step": 1176 + }, + { + "epoch": 0.03630628997991124, + "grad_norm": 0.2491193264722824, + "learning_rate": 1.8152375077112894e-05, + "loss": 3.6018, + "step": 1177 + }, + { + "epoch": 0.036337136445484655, + "grad_norm": 0.23011355102062225, + "learning_rate": 1.8167797655768047e-05, + "loss": 3.5491, + "step": 1178 + }, + { + "epoch": 0.03636798291105807, + "grad_norm": 0.2525694668292999, + "learning_rate": 1.8183220234423196e-05, + "loss": 3.6518, + "step": 1179 + }, + { + "epoch": 0.03639882937663149, + "grad_norm": 0.22281931340694427, + "learning_rate": 1.8198642813078348e-05, + "loss": 3.6385, + "step": 1180 + }, + { + "epoch": 0.036429675842204905, + "grad_norm": 0.23714599013328552, + "learning_rate": 1.8214065391733497e-05, + "loss": 3.5416, + "step": 1181 + }, + { + "epoch": 0.03646052230777832, + "grad_norm": 0.24523532390594482, + "learning_rate": 1.822948797038865e-05, + "loss": 3.6062, + "step": 1182 + }, + { + "epoch": 0.03649136877335174, + "grad_norm": 0.22915101051330566, + "learning_rate": 1.82449105490438e-05, + "loss": 3.5923, + "step": 1183 + }, + { + "epoch": 0.036522215238925154, + "grad_norm": 0.23222216963768005, + "learning_rate": 1.8260333127698954e-05, + "loss": 3.5543, + "step": 1184 + }, + { + "epoch": 0.03655306170449857, + "grad_norm": 0.24661925435066223, + "learning_rate": 1.8275755706354103e-05, + "loss": 3.605, + "step": 1185 + }, + { + "epoch": 0.03658390817007199, + "grad_norm": 0.25159862637519836, + "learning_rate": 1.8291178285009255e-05, + "loss": 3.6003, + "step": 1186 + }, + { + "epoch": 0.036614754635645404, + "grad_norm": 0.2549530863761902, + "learning_rate": 1.8306600863664404e-05, + "loss": 3.5822, + "step": 1187 + }, + { + "epoch": 0.03664560110121882, + "grad_norm": 0.23194114863872528, + "learning_rate": 1.832202344231956e-05, + "loss": 3.5749, + "step": 1188 + }, + { + "epoch": 0.03667644756679224, + "grad_norm": 0.25352391600608826, + "learning_rate": 1.833744602097471e-05, + "loss": 3.5539, + "step": 1189 + }, + { + "epoch": 0.03670729403236565, + "grad_norm": 0.24706782400608063, + "learning_rate": 1.8352868599629857e-05, + "loss": 3.6102, + "step": 1190 + }, + { + "epoch": 0.03673814049793907, + "grad_norm": 0.26822182536125183, + "learning_rate": 1.836829117828501e-05, + "loss": 3.5818, + "step": 1191 + }, + { + "epoch": 0.03676898696351249, + "grad_norm": 0.5403900146484375, + "learning_rate": 1.8383713756940162e-05, + "loss": 3.5641, + "step": 1192 + }, + { + "epoch": 0.0367998334290859, + "grad_norm": 0.24710623919963837, + "learning_rate": 1.8399136335595314e-05, + "loss": 3.5849, + "step": 1193 + }, + { + "epoch": 0.03683067989465932, + "grad_norm": 0.25822627544403076, + "learning_rate": 1.8414558914250463e-05, + "loss": 3.5728, + "step": 1194 + }, + { + "epoch": 0.03686152636023274, + "grad_norm": 0.2418903112411499, + "learning_rate": 1.8429981492905615e-05, + "loss": 3.5888, + "step": 1195 + }, + { + "epoch": 0.03689237282580615, + "grad_norm": 0.253540575504303, + "learning_rate": 1.8445404071560764e-05, + "loss": 3.5789, + "step": 1196 + }, + { + "epoch": 0.036923219291379566, + "grad_norm": 0.2441854476928711, + "learning_rate": 1.8460826650215917e-05, + "loss": 3.5702, + "step": 1197 + }, + { + "epoch": 0.03695406575695299, + "grad_norm": 0.230422705411911, + "learning_rate": 1.847624922887107e-05, + "loss": 3.5023, + "step": 1198 + }, + { + "epoch": 0.0369849122225264, + "grad_norm": 0.2559676766395569, + "learning_rate": 1.849167180752622e-05, + "loss": 3.6242, + "step": 1199 + }, + { + "epoch": 0.037015758688099816, + "grad_norm": 0.2357252836227417, + "learning_rate": 1.850709438618137e-05, + "loss": 3.5861, + "step": 1200 + }, + { + "epoch": 0.03704660515367324, + "grad_norm": 0.23260195553302765, + "learning_rate": 1.8522516964836522e-05, + "loss": 3.5757, + "step": 1201 + }, + { + "epoch": 0.03707745161924665, + "grad_norm": 0.24571110308170319, + "learning_rate": 1.853793954349167e-05, + "loss": 3.5719, + "step": 1202 + }, + { + "epoch": 0.037108298084820066, + "grad_norm": 0.2535451352596283, + "learning_rate": 1.8553362122146824e-05, + "loss": 3.5972, + "step": 1203 + }, + { + "epoch": 0.03713914455039349, + "grad_norm": 0.24934551119804382, + "learning_rate": 1.8568784700801976e-05, + "loss": 3.5492, + "step": 1204 + }, + { + "epoch": 0.0371699910159669, + "grad_norm": 0.24289588630199432, + "learning_rate": 1.8584207279457125e-05, + "loss": 3.5772, + "step": 1205 + }, + { + "epoch": 0.037200837481540315, + "grad_norm": 0.25230348110198975, + "learning_rate": 1.8599629858112277e-05, + "loss": 3.5899, + "step": 1206 + }, + { + "epoch": 0.037231683947113736, + "grad_norm": 0.24599787592887878, + "learning_rate": 1.8615052436767426e-05, + "loss": 3.5733, + "step": 1207 + }, + { + "epoch": 0.03726253041268715, + "grad_norm": 0.26710233092308044, + "learning_rate": 1.8630475015422578e-05, + "loss": 3.5407, + "step": 1208 + }, + { + "epoch": 0.037293376878260565, + "grad_norm": 0.23738428950309753, + "learning_rate": 1.864589759407773e-05, + "loss": 3.5773, + "step": 1209 + }, + { + "epoch": 0.037324223343833986, + "grad_norm": 0.26103922724723816, + "learning_rate": 1.8661320172732883e-05, + "loss": 3.58, + "step": 1210 + }, + { + "epoch": 0.0373550698094074, + "grad_norm": 0.27288007736206055, + "learning_rate": 1.8676742751388032e-05, + "loss": 3.5852, + "step": 1211 + }, + { + "epoch": 0.037385916274980814, + "grad_norm": 0.2429853081703186, + "learning_rate": 1.8692165330043184e-05, + "loss": 3.5833, + "step": 1212 + }, + { + "epoch": 0.037416762740554235, + "grad_norm": 0.27991747856140137, + "learning_rate": 1.8707587908698336e-05, + "loss": 3.592, + "step": 1213 + }, + { + "epoch": 0.03744760920612765, + "grad_norm": 0.236264169216156, + "learning_rate": 1.872301048735349e-05, + "loss": 3.5766, + "step": 1214 + }, + { + "epoch": 0.037478455671701064, + "grad_norm": 0.30811959505081177, + "learning_rate": 1.8738433066008638e-05, + "loss": 3.5672, + "step": 1215 + }, + { + "epoch": 0.037509302137274485, + "grad_norm": 0.2466907650232315, + "learning_rate": 1.875385564466379e-05, + "loss": 3.5725, + "step": 1216 + }, + { + "epoch": 0.0375401486028479, + "grad_norm": 0.2624056041240692, + "learning_rate": 1.876927822331894e-05, + "loss": 3.604, + "step": 1217 + }, + { + "epoch": 0.03757099506842131, + "grad_norm": 0.24161642789840698, + "learning_rate": 1.878470080197409e-05, + "loss": 3.5713, + "step": 1218 + }, + { + "epoch": 0.037601841533994734, + "grad_norm": 0.24529233574867249, + "learning_rate": 1.8800123380629243e-05, + "loss": 3.5822, + "step": 1219 + }, + { + "epoch": 0.03763268799956815, + "grad_norm": 0.23915189504623413, + "learning_rate": 1.8815545959284392e-05, + "loss": 3.5325, + "step": 1220 + }, + { + "epoch": 0.03766353446514156, + "grad_norm": 0.2308206558227539, + "learning_rate": 1.8830968537939545e-05, + "loss": 3.6194, + "step": 1221 + }, + { + "epoch": 0.037694380930714984, + "grad_norm": 0.23960153758525848, + "learning_rate": 1.8846391116594693e-05, + "loss": 3.576, + "step": 1222 + }, + { + "epoch": 0.0377252273962884, + "grad_norm": 0.24290457367897034, + "learning_rate": 1.8861813695249846e-05, + "loss": 3.5667, + "step": 1223 + }, + { + "epoch": 0.03775607386186181, + "grad_norm": 0.24723567068576813, + "learning_rate": 1.8877236273904998e-05, + "loss": 3.5821, + "step": 1224 + }, + { + "epoch": 0.03778692032743523, + "grad_norm": 0.24689695239067078, + "learning_rate": 1.889265885256015e-05, + "loss": 3.5496, + "step": 1225 + }, + { + "epoch": 0.03781776679300865, + "grad_norm": 0.24324090778827667, + "learning_rate": 1.89080814312153e-05, + "loss": 3.592, + "step": 1226 + }, + { + "epoch": 0.03784861325858206, + "grad_norm": 0.22780157625675201, + "learning_rate": 1.892350400987045e-05, + "loss": 3.5446, + "step": 1227 + }, + { + "epoch": 0.03787945972415548, + "grad_norm": 0.2404559701681137, + "learning_rate": 1.89389265885256e-05, + "loss": 3.5481, + "step": 1228 + }, + { + "epoch": 0.0379103061897289, + "grad_norm": 0.24415996670722961, + "learning_rate": 1.8954349167180753e-05, + "loss": 3.5235, + "step": 1229 + }, + { + "epoch": 0.03794115265530231, + "grad_norm": 0.23945604264736176, + "learning_rate": 1.8969771745835905e-05, + "loss": 3.5774, + "step": 1230 + }, + { + "epoch": 0.03797199912087573, + "grad_norm": 0.25586655735969543, + "learning_rate": 1.8985194324491057e-05, + "loss": 3.5954, + "step": 1231 + }, + { + "epoch": 0.03800284558644915, + "grad_norm": 0.23772770166397095, + "learning_rate": 1.9000616903146206e-05, + "loss": 3.573, + "step": 1232 + }, + { + "epoch": 0.03803369205202256, + "grad_norm": 0.25049713253974915, + "learning_rate": 1.901603948180136e-05, + "loss": 3.5985, + "step": 1233 + }, + { + "epoch": 0.03806453851759598, + "grad_norm": 0.25597381591796875, + "learning_rate": 1.903146206045651e-05, + "loss": 3.5868, + "step": 1234 + }, + { + "epoch": 0.038095384983169396, + "grad_norm": 0.2428811639547348, + "learning_rate": 1.9046884639111663e-05, + "loss": 3.6062, + "step": 1235 + }, + { + "epoch": 0.03812623144874281, + "grad_norm": 0.24497875571250916, + "learning_rate": 1.9062307217766812e-05, + "loss": 3.5772, + "step": 1236 + }, + { + "epoch": 0.03815707791431623, + "grad_norm": 0.23955394327640533, + "learning_rate": 1.907772979642196e-05, + "loss": 3.5356, + "step": 1237 + }, + { + "epoch": 0.038187924379889646, + "grad_norm": 0.25154635310173035, + "learning_rate": 1.9093152375077113e-05, + "loss": 3.5902, + "step": 1238 + }, + { + "epoch": 0.03821877084546307, + "grad_norm": 0.24646180868148804, + "learning_rate": 1.9108574953732265e-05, + "loss": 3.5386, + "step": 1239 + }, + { + "epoch": 0.03824961731103648, + "grad_norm": 0.2428087443113327, + "learning_rate": 1.9123997532387418e-05, + "loss": 3.5828, + "step": 1240 + }, + { + "epoch": 0.038280463776609895, + "grad_norm": 0.2695709466934204, + "learning_rate": 1.9139420111042567e-05, + "loss": 3.586, + "step": 1241 + }, + { + "epoch": 0.038311310242183316, + "grad_norm": 0.25732165575027466, + "learning_rate": 1.915484268969772e-05, + "loss": 3.5661, + "step": 1242 + }, + { + "epoch": 0.03834215670775673, + "grad_norm": 0.24684856832027435, + "learning_rate": 1.9170265268352868e-05, + "loss": 3.5683, + "step": 1243 + }, + { + "epoch": 0.038373003173330145, + "grad_norm": 0.30270200967788696, + "learning_rate": 1.918568784700802e-05, + "loss": 3.5423, + "step": 1244 + }, + { + "epoch": 0.038403849638903566, + "grad_norm": 0.26587527990341187, + "learning_rate": 1.9201110425663172e-05, + "loss": 3.5711, + "step": 1245 + }, + { + "epoch": 0.03843469610447698, + "grad_norm": 0.2415309101343155, + "learning_rate": 1.9216533004318325e-05, + "loss": 3.5577, + "step": 1246 + }, + { + "epoch": 0.038465542570050394, + "grad_norm": 0.3087371289730072, + "learning_rate": 1.9231955582973474e-05, + "loss": 3.5577, + "step": 1247 + }, + { + "epoch": 0.038496389035623815, + "grad_norm": 0.24598470330238342, + "learning_rate": 1.9247378161628626e-05, + "loss": 3.5754, + "step": 1248 + }, + { + "epoch": 0.03852723550119723, + "grad_norm": 0.25953561067581177, + "learning_rate": 1.9262800740283775e-05, + "loss": 3.5615, + "step": 1249 + }, + { + "epoch": 0.038558081966770644, + "grad_norm": 0.2345045804977417, + "learning_rate": 1.927822331893893e-05, + "loss": 3.5496, + "step": 1250 + }, + { + "epoch": 0.038588928432344065, + "grad_norm": 0.6001016497612, + "learning_rate": 1.929364589759408e-05, + "loss": 3.5985, + "step": 1251 + }, + { + "epoch": 0.03861977489791748, + "grad_norm": 0.258253276348114, + "learning_rate": 1.930906847624923e-05, + "loss": 3.5633, + "step": 1252 + }, + { + "epoch": 0.03865062136349089, + "grad_norm": 0.2593851685523987, + "learning_rate": 1.932449105490438e-05, + "loss": 3.5633, + "step": 1253 + }, + { + "epoch": 0.038681467829064314, + "grad_norm": 0.27315664291381836, + "learning_rate": 1.933991363355953e-05, + "loss": 3.5722, + "step": 1254 + }, + { + "epoch": 0.03871231429463773, + "grad_norm": 0.26458290219306946, + "learning_rate": 1.9355336212214685e-05, + "loss": 3.577, + "step": 1255 + }, + { + "epoch": 0.03874316076021114, + "grad_norm": 0.2754521667957306, + "learning_rate": 1.9370758790869834e-05, + "loss": 3.6069, + "step": 1256 + }, + { + "epoch": 0.038774007225784564, + "grad_norm": 0.2590654194355011, + "learning_rate": 1.9386181369524986e-05, + "loss": 3.5261, + "step": 1257 + }, + { + "epoch": 0.03880485369135798, + "grad_norm": 0.2615162134170532, + "learning_rate": 1.9401603948180135e-05, + "loss": 3.5964, + "step": 1258 + }, + { + "epoch": 0.03883570015693139, + "grad_norm": 0.27616482973098755, + "learning_rate": 1.9417026526835288e-05, + "loss": 3.5807, + "step": 1259 + }, + { + "epoch": 0.03886654662250481, + "grad_norm": 0.2569907307624817, + "learning_rate": 1.943244910549044e-05, + "loss": 3.5785, + "step": 1260 + }, + { + "epoch": 0.03889739308807823, + "grad_norm": 0.22871632874011993, + "learning_rate": 1.9447871684145592e-05, + "loss": 3.5124, + "step": 1261 + }, + { + "epoch": 0.03892823955365164, + "grad_norm": 0.24584931135177612, + "learning_rate": 1.946329426280074e-05, + "loss": 3.5533, + "step": 1262 + }, + { + "epoch": 0.03895908601922506, + "grad_norm": 0.24364745616912842, + "learning_rate": 1.9478716841455893e-05, + "loss": 3.5558, + "step": 1263 + }, + { + "epoch": 0.03898993248479848, + "grad_norm": 0.25596144795417786, + "learning_rate": 1.9494139420111042e-05, + "loss": 3.5382, + "step": 1264 + }, + { + "epoch": 0.03902077895037189, + "grad_norm": 0.23552606999874115, + "learning_rate": 1.9509561998766195e-05, + "loss": 3.5602, + "step": 1265 + }, + { + "epoch": 0.03905162541594531, + "grad_norm": 0.22799737751483917, + "learning_rate": 1.9524984577421347e-05, + "loss": 3.5401, + "step": 1266 + }, + { + "epoch": 0.03908247188151873, + "grad_norm": 0.22352305054664612, + "learning_rate": 1.9540407156076496e-05, + "loss": 3.5546, + "step": 1267 + }, + { + "epoch": 0.03911331834709214, + "grad_norm": 0.23427830636501312, + "learning_rate": 1.9555829734731648e-05, + "loss": 3.5664, + "step": 1268 + }, + { + "epoch": 0.03914416481266556, + "grad_norm": 0.23677901923656464, + "learning_rate": 1.9571252313386797e-05, + "loss": 3.5775, + "step": 1269 + }, + { + "epoch": 0.039175011278238976, + "grad_norm": 0.2310684770345688, + "learning_rate": 1.958667489204195e-05, + "loss": 3.5833, + "step": 1270 + }, + { + "epoch": 0.03920585774381239, + "grad_norm": 0.23794253170490265, + "learning_rate": 1.96020974706971e-05, + "loss": 3.5702, + "step": 1271 + }, + { + "epoch": 0.03923670420938581, + "grad_norm": 0.25184792280197144, + "learning_rate": 1.9617520049352254e-05, + "loss": 3.5704, + "step": 1272 + }, + { + "epoch": 0.039267550674959226, + "grad_norm": 0.23849056661128998, + "learning_rate": 1.9632942628007403e-05, + "loss": 3.5782, + "step": 1273 + }, + { + "epoch": 0.03929839714053264, + "grad_norm": 0.23943911492824554, + "learning_rate": 1.9648365206662555e-05, + "loss": 3.5815, + "step": 1274 + }, + { + "epoch": 0.03932924360610606, + "grad_norm": 0.2384975701570511, + "learning_rate": 1.9663787785317704e-05, + "loss": 3.5426, + "step": 1275 + }, + { + "epoch": 0.039360090071679475, + "grad_norm": 0.2306309938430786, + "learning_rate": 1.967921036397286e-05, + "loss": 3.5963, + "step": 1276 + }, + { + "epoch": 0.03939093653725289, + "grad_norm": 0.24108190834522247, + "learning_rate": 1.969463294262801e-05, + "loss": 3.5772, + "step": 1277 + }, + { + "epoch": 0.03942178300282631, + "grad_norm": 0.2322525531053543, + "learning_rate": 1.971005552128316e-05, + "loss": 3.5917, + "step": 1278 + }, + { + "epoch": 0.039452629468399725, + "grad_norm": 0.2337523251771927, + "learning_rate": 1.972547809993831e-05, + "loss": 3.5616, + "step": 1279 + }, + { + "epoch": 0.03948347593397314, + "grad_norm": 0.24989357590675354, + "learning_rate": 1.9740900678593462e-05, + "loss": 3.5747, + "step": 1280 + }, + { + "epoch": 0.03951432239954656, + "grad_norm": 0.21804283559322357, + "learning_rate": 1.9756323257248614e-05, + "loss": 3.5979, + "step": 1281 + }, + { + "epoch": 0.039545168865119974, + "grad_norm": 0.2411736249923706, + "learning_rate": 1.9771745835903763e-05, + "loss": 3.5435, + "step": 1282 + }, + { + "epoch": 0.03957601533069339, + "grad_norm": 0.22679747641086578, + "learning_rate": 1.9787168414558916e-05, + "loss": 3.6054, + "step": 1283 + }, + { + "epoch": 0.03960686179626681, + "grad_norm": 0.2510559558868408, + "learning_rate": 1.9802590993214064e-05, + "loss": 3.567, + "step": 1284 + }, + { + "epoch": 0.039637708261840224, + "grad_norm": 0.2638252377510071, + "learning_rate": 1.9818013571869217e-05, + "loss": 3.5692, + "step": 1285 + }, + { + "epoch": 0.03966855472741364, + "grad_norm": 0.23445942997932434, + "learning_rate": 1.983343615052437e-05, + "loss": 3.5611, + "step": 1286 + }, + { + "epoch": 0.03969940119298706, + "grad_norm": 0.23105166852474213, + "learning_rate": 1.984885872917952e-05, + "loss": 3.5323, + "step": 1287 + }, + { + "epoch": 0.03973024765856047, + "grad_norm": 0.25006103515625, + "learning_rate": 1.986428130783467e-05, + "loss": 3.5413, + "step": 1288 + }, + { + "epoch": 0.03976109412413389, + "grad_norm": 0.3394722640514374, + "learning_rate": 1.9879703886489823e-05, + "loss": 3.593, + "step": 1289 + }, + { + "epoch": 0.03979194058970731, + "grad_norm": 0.24632370471954346, + "learning_rate": 1.989512646514497e-05, + "loss": 3.5607, + "step": 1290 + }, + { + "epoch": 0.03982278705528072, + "grad_norm": 0.23306435346603394, + "learning_rate": 1.9910549043800124e-05, + "loss": 3.567, + "step": 1291 + }, + { + "epoch": 0.03985363352085414, + "grad_norm": 0.24986478686332703, + "learning_rate": 1.9925971622455276e-05, + "loss": 3.5708, + "step": 1292 + }, + { + "epoch": 0.03988447998642756, + "grad_norm": 0.33817416429519653, + "learning_rate": 1.994139420111043e-05, + "loss": 3.555, + "step": 1293 + }, + { + "epoch": 0.03991532645200097, + "grad_norm": 0.2295769453048706, + "learning_rate": 1.9956816779765577e-05, + "loss": 3.565, + "step": 1294 + }, + { + "epoch": 0.03994617291757439, + "grad_norm": 0.2355661541223526, + "learning_rate": 1.997223935842073e-05, + "loss": 3.5577, + "step": 1295 + }, + { + "epoch": 0.03997701938314781, + "grad_norm": 0.22775959968566895, + "learning_rate": 1.998766193707588e-05, + "loss": 3.5596, + "step": 1296 + }, + { + "epoch": 0.04000786584872122, + "grad_norm": 0.2537963390350342, + "learning_rate": 2.000308451573103e-05, + "loss": 3.5582, + "step": 1297 + }, + { + "epoch": 0.040038712314294636, + "grad_norm": 0.222590371966362, + "learning_rate": 2.0018507094386183e-05, + "loss": 3.5629, + "step": 1298 + }, + { + "epoch": 0.04006955877986806, + "grad_norm": 0.23731957376003265, + "learning_rate": 2.0033929673041332e-05, + "loss": 3.5514, + "step": 1299 + }, + { + "epoch": 0.04010040524544147, + "grad_norm": 0.22455862164497375, + "learning_rate": 2.0049352251696484e-05, + "loss": 3.5009, + "step": 1300 + }, + { + "epoch": 0.040131251711014886, + "grad_norm": 0.23213304579257965, + "learning_rate": 2.0064774830351633e-05, + "loss": 3.5578, + "step": 1301 + }, + { + "epoch": 0.04016209817658831, + "grad_norm": 0.22166290879249573, + "learning_rate": 2.008019740900679e-05, + "loss": 3.5786, + "step": 1302 + }, + { + "epoch": 0.04019294464216172, + "grad_norm": 0.22895874083042145, + "learning_rate": 2.0095619987661938e-05, + "loss": 3.5203, + "step": 1303 + }, + { + "epoch": 0.040223791107735135, + "grad_norm": 0.2219117283821106, + "learning_rate": 2.011104256631709e-05, + "loss": 3.5716, + "step": 1304 + }, + { + "epoch": 0.040254637573308556, + "grad_norm": 0.22446542978286743, + "learning_rate": 2.012646514497224e-05, + "loss": 3.5564, + "step": 1305 + }, + { + "epoch": 0.04028548403888197, + "grad_norm": 0.21967339515686035, + "learning_rate": 2.014188772362739e-05, + "loss": 3.5631, + "step": 1306 + }, + { + "epoch": 0.040316330504455385, + "grad_norm": 0.2846149504184723, + "learning_rate": 2.0157310302282544e-05, + "loss": 3.5297, + "step": 1307 + }, + { + "epoch": 0.040347176970028806, + "grad_norm": 0.2349005937576294, + "learning_rate": 2.0172732880937696e-05, + "loss": 3.5723, + "step": 1308 + }, + { + "epoch": 0.04037802343560222, + "grad_norm": 0.22943007946014404, + "learning_rate": 2.0188155459592845e-05, + "loss": 3.5561, + "step": 1309 + }, + { + "epoch": 0.040408869901175634, + "grad_norm": 0.23881907761096954, + "learning_rate": 2.0203578038247997e-05, + "loss": 3.5423, + "step": 1310 + }, + { + "epoch": 0.040439716366749055, + "grad_norm": 0.21609054505825043, + "learning_rate": 2.0219000616903146e-05, + "loss": 3.552, + "step": 1311 + }, + { + "epoch": 0.04047056283232247, + "grad_norm": 0.2340463548898697, + "learning_rate": 2.0234423195558298e-05, + "loss": 3.5516, + "step": 1312 + }, + { + "epoch": 0.040501409297895884, + "grad_norm": 0.2358899563550949, + "learning_rate": 2.024984577421345e-05, + "loss": 3.5606, + "step": 1313 + }, + { + "epoch": 0.040532255763469305, + "grad_norm": 0.2364603579044342, + "learning_rate": 2.02652683528686e-05, + "loss": 3.5193, + "step": 1314 + }, + { + "epoch": 0.04056310222904272, + "grad_norm": 0.23557406663894653, + "learning_rate": 2.0280690931523752e-05, + "loss": 3.5672, + "step": 1315 + }, + { + "epoch": 0.04059394869461613, + "grad_norm": 0.23582200706005096, + "learning_rate": 2.02961135101789e-05, + "loss": 3.5374, + "step": 1316 + }, + { + "epoch": 0.040624795160189554, + "grad_norm": 0.21977734565734863, + "learning_rate": 2.0311536088834053e-05, + "loss": 3.5685, + "step": 1317 + }, + { + "epoch": 0.04065564162576297, + "grad_norm": 0.2410944402217865, + "learning_rate": 2.0326958667489205e-05, + "loss": 3.5484, + "step": 1318 + }, + { + "epoch": 0.04068648809133638, + "grad_norm": 0.2355090230703354, + "learning_rate": 2.0342381246144357e-05, + "loss": 3.5154, + "step": 1319 + }, + { + "epoch": 0.040717334556909804, + "grad_norm": 0.2414407879114151, + "learning_rate": 2.0357803824799506e-05, + "loss": 3.5659, + "step": 1320 + }, + { + "epoch": 0.04074818102248322, + "grad_norm": 0.2462792992591858, + "learning_rate": 2.037322640345466e-05, + "loss": 3.5518, + "step": 1321 + }, + { + "epoch": 0.04077902748805663, + "grad_norm": 0.22879518568515778, + "learning_rate": 2.0388648982109808e-05, + "loss": 3.5291, + "step": 1322 + }, + { + "epoch": 0.04080987395363005, + "grad_norm": 0.23841850459575653, + "learning_rate": 2.0404071560764963e-05, + "loss": 3.5357, + "step": 1323 + }, + { + "epoch": 0.04084072041920347, + "grad_norm": 0.23835696280002594, + "learning_rate": 2.0419494139420112e-05, + "loss": 3.5458, + "step": 1324 + }, + { + "epoch": 0.04087156688477688, + "grad_norm": 0.2242366075515747, + "learning_rate": 2.0434916718075264e-05, + "loss": 3.5422, + "step": 1325 + }, + { + "epoch": 0.0409024133503503, + "grad_norm": 0.23499555885791779, + "learning_rate": 2.0450339296730413e-05, + "loss": 3.5639, + "step": 1326 + }, + { + "epoch": 0.04093325981592372, + "grad_norm": 0.2334563285112381, + "learning_rate": 2.0465761875385566e-05, + "loss": 3.5587, + "step": 1327 + }, + { + "epoch": 0.04096410628149713, + "grad_norm": 0.21829728782176971, + "learning_rate": 2.0481184454040718e-05, + "loss": 3.5167, + "step": 1328 + }, + { + "epoch": 0.04099495274707055, + "grad_norm": 0.24331443011760712, + "learning_rate": 2.0496607032695867e-05, + "loss": 3.5697, + "step": 1329 + }, + { + "epoch": 0.04102579921264397, + "grad_norm": 0.21950975060462952, + "learning_rate": 2.051202961135102e-05, + "loss": 3.5417, + "step": 1330 + }, + { + "epoch": 0.04105664567821738, + "grad_norm": 0.24037432670593262, + "learning_rate": 2.0527452190006168e-05, + "loss": 3.5241, + "step": 1331 + }, + { + "epoch": 0.0410874921437908, + "grad_norm": 0.23271732032299042, + "learning_rate": 2.054287476866132e-05, + "loss": 3.5279, + "step": 1332 + }, + { + "epoch": 0.041118338609364216, + "grad_norm": 0.2422005981206894, + "learning_rate": 2.0558297347316473e-05, + "loss": 3.5183, + "step": 1333 + }, + { + "epoch": 0.04114918507493763, + "grad_norm": 0.22964638471603394, + "learning_rate": 2.0573719925971625e-05, + "loss": 3.5576, + "step": 1334 + }, + { + "epoch": 0.04118003154051105, + "grad_norm": 0.2575171887874603, + "learning_rate": 2.0589142504626774e-05, + "loss": 3.5735, + "step": 1335 + }, + { + "epoch": 0.041210878006084466, + "grad_norm": 0.34039193391799927, + "learning_rate": 2.0604565083281926e-05, + "loss": 3.5575, + "step": 1336 + }, + { + "epoch": 0.04124172447165788, + "grad_norm": 0.24074313044548035, + "learning_rate": 2.0619987661937075e-05, + "loss": 3.5291, + "step": 1337 + }, + { + "epoch": 0.0412725709372313, + "grad_norm": 0.23729640245437622, + "learning_rate": 2.0635410240592227e-05, + "loss": 3.5201, + "step": 1338 + }, + { + "epoch": 0.041303417402804715, + "grad_norm": 0.25678882002830505, + "learning_rate": 2.065083281924738e-05, + "loss": 3.5066, + "step": 1339 + }, + { + "epoch": 0.04133426386837813, + "grad_norm": 0.2372741401195526, + "learning_rate": 2.0666255397902532e-05, + "loss": 3.6048, + "step": 1340 + }, + { + "epoch": 0.04136511033395155, + "grad_norm": 0.2526729106903076, + "learning_rate": 2.068167797655768e-05, + "loss": 3.5493, + "step": 1341 + }, + { + "epoch": 0.041395956799524965, + "grad_norm": 0.23766325414180756, + "learning_rate": 2.0697100555212833e-05, + "loss": 3.5512, + "step": 1342 + }, + { + "epoch": 0.04142680326509838, + "grad_norm": 0.24464574456214905, + "learning_rate": 2.0712523133867982e-05, + "loss": 3.5405, + "step": 1343 + }, + { + "epoch": 0.0414576497306718, + "grad_norm": 0.2344249188899994, + "learning_rate": 2.0727945712523134e-05, + "loss": 3.5847, + "step": 1344 + }, + { + "epoch": 0.041488496196245214, + "grad_norm": 0.23646190762519836, + "learning_rate": 2.0743368291178287e-05, + "loss": 3.504, + "step": 1345 + }, + { + "epoch": 0.04151934266181863, + "grad_norm": 0.25140804052352905, + "learning_rate": 2.0758790869833436e-05, + "loss": 3.57, + "step": 1346 + }, + { + "epoch": 0.04155018912739205, + "grad_norm": 0.243734210729599, + "learning_rate": 2.0774213448488588e-05, + "loss": 3.6014, + "step": 1347 + }, + { + "epoch": 0.041581035592965464, + "grad_norm": 0.2374454140663147, + "learning_rate": 2.078963602714374e-05, + "loss": 3.511, + "step": 1348 + }, + { + "epoch": 0.04161188205853888, + "grad_norm": 0.2782422602176666, + "learning_rate": 2.0805058605798892e-05, + "loss": 3.5134, + "step": 1349 + }, + { + "epoch": 0.0416427285241123, + "grad_norm": 0.23073843121528625, + "learning_rate": 2.082048118445404e-05, + "loss": 3.5501, + "step": 1350 + }, + { + "epoch": 0.04167357498968571, + "grad_norm": 0.25780609250068665, + "learning_rate": 2.0835903763109194e-05, + "loss": 3.5566, + "step": 1351 + }, + { + "epoch": 0.04170442145525913, + "grad_norm": 0.23252767324447632, + "learning_rate": 2.0851326341764343e-05, + "loss": 3.5363, + "step": 1352 + }, + { + "epoch": 0.04173526792083255, + "grad_norm": 0.24425554275512695, + "learning_rate": 2.0866748920419495e-05, + "loss": 3.5562, + "step": 1353 + }, + { + "epoch": 0.04176611438640596, + "grad_norm": 0.24849188327789307, + "learning_rate": 2.0882171499074647e-05, + "loss": 3.5104, + "step": 1354 + }, + { + "epoch": 0.04179696085197938, + "grad_norm": 0.2308211475610733, + "learning_rate": 2.08975940777298e-05, + "loss": 3.5625, + "step": 1355 + }, + { + "epoch": 0.0418278073175528, + "grad_norm": 0.23234239220619202, + "learning_rate": 2.0913016656384948e-05, + "loss": 3.5399, + "step": 1356 + }, + { + "epoch": 0.04185865378312621, + "grad_norm": 0.23724502325057983, + "learning_rate": 2.09284392350401e-05, + "loss": 3.5198, + "step": 1357 + }, + { + "epoch": 0.041889500248699627, + "grad_norm": 0.2291116863489151, + "learning_rate": 2.094386181369525e-05, + "loss": 3.5203, + "step": 1358 + }, + { + "epoch": 0.04192034671427305, + "grad_norm": 0.23582908511161804, + "learning_rate": 2.0959284392350402e-05, + "loss": 3.5749, + "step": 1359 + }, + { + "epoch": 0.04195119317984646, + "grad_norm": 0.24304983019828796, + "learning_rate": 2.0974706971005554e-05, + "loss": 3.5576, + "step": 1360 + }, + { + "epoch": 0.041982039645419876, + "grad_norm": 0.23693828284740448, + "learning_rate": 2.0990129549660703e-05, + "loss": 3.515, + "step": 1361 + }, + { + "epoch": 0.0420128861109933, + "grad_norm": 0.2157559096813202, + "learning_rate": 2.1005552128315855e-05, + "loss": 3.5071, + "step": 1362 + }, + { + "epoch": 0.04204373257656671, + "grad_norm": 0.22167052328586578, + "learning_rate": 2.1020974706971004e-05, + "loss": 3.5404, + "step": 1363 + }, + { + "epoch": 0.042074579042140126, + "grad_norm": 0.22148047387599945, + "learning_rate": 2.103639728562616e-05, + "loss": 3.5298, + "step": 1364 + }, + { + "epoch": 0.04210542550771355, + "grad_norm": 0.23352032899856567, + "learning_rate": 2.105181986428131e-05, + "loss": 3.5763, + "step": 1365 + }, + { + "epoch": 0.04213627197328696, + "grad_norm": 0.23195956647396088, + "learning_rate": 2.106724244293646e-05, + "loss": 3.5195, + "step": 1366 + }, + { + "epoch": 0.042167118438860375, + "grad_norm": 0.22991780936717987, + "learning_rate": 2.108266502159161e-05, + "loss": 3.5478, + "step": 1367 + }, + { + "epoch": 0.042197964904433796, + "grad_norm": 0.23234671354293823, + "learning_rate": 2.1098087600246762e-05, + "loss": 3.526, + "step": 1368 + }, + { + "epoch": 0.04222881137000721, + "grad_norm": 0.26615241169929504, + "learning_rate": 2.1113510178901915e-05, + "loss": 3.5811, + "step": 1369 + }, + { + "epoch": 0.042259657835580625, + "grad_norm": 0.2556490898132324, + "learning_rate": 2.1128932757557067e-05, + "loss": 3.5438, + "step": 1370 + }, + { + "epoch": 0.042290504301154046, + "grad_norm": 0.2422417402267456, + "learning_rate": 2.1144355336212216e-05, + "loss": 3.5292, + "step": 1371 + }, + { + "epoch": 0.04232135076672746, + "grad_norm": 0.256850004196167, + "learning_rate": 2.1159777914867368e-05, + "loss": 3.5366, + "step": 1372 + }, + { + "epoch": 0.042352197232300874, + "grad_norm": 0.23579657077789307, + "learning_rate": 2.1175200493522517e-05, + "loss": 3.5832, + "step": 1373 + }, + { + "epoch": 0.042383043697874295, + "grad_norm": 0.24447029829025269, + "learning_rate": 2.119062307217767e-05, + "loss": 3.5083, + "step": 1374 + }, + { + "epoch": 0.04241389016344771, + "grad_norm": 0.22327739000320435, + "learning_rate": 2.120604565083282e-05, + "loss": 3.5127, + "step": 1375 + }, + { + "epoch": 0.042444736629021124, + "grad_norm": 0.24778009951114655, + "learning_rate": 2.122146822948797e-05, + "loss": 3.522, + "step": 1376 + }, + { + "epoch": 0.042475583094594545, + "grad_norm": 0.21934135258197784, + "learning_rate": 2.1236890808143123e-05, + "loss": 3.5423, + "step": 1377 + }, + { + "epoch": 0.04250642956016796, + "grad_norm": 0.24259278178215027, + "learning_rate": 2.125231338679827e-05, + "loss": 3.4991, + "step": 1378 + }, + { + "epoch": 0.04253727602574137, + "grad_norm": 0.21583212912082672, + "learning_rate": 2.1267735965453424e-05, + "loss": 3.5192, + "step": 1379 + }, + { + "epoch": 0.042568122491314794, + "grad_norm": 0.26673391461372375, + "learning_rate": 2.1283158544108576e-05, + "loss": 3.5542, + "step": 1380 + }, + { + "epoch": 0.04259896895688821, + "grad_norm": 0.21618486940860748, + "learning_rate": 2.129858112276373e-05, + "loss": 3.5241, + "step": 1381 + }, + { + "epoch": 0.04262981542246162, + "grad_norm": 0.25244414806365967, + "learning_rate": 2.1314003701418877e-05, + "loss": 3.5385, + "step": 1382 + }, + { + "epoch": 0.042660661888035044, + "grad_norm": 0.23253880441188812, + "learning_rate": 2.132942628007403e-05, + "loss": 3.5527, + "step": 1383 + }, + { + "epoch": 0.04269150835360846, + "grad_norm": 0.24498514831066132, + "learning_rate": 2.134484885872918e-05, + "loss": 3.5571, + "step": 1384 + }, + { + "epoch": 0.04272235481918187, + "grad_norm": 0.23731690645217896, + "learning_rate": 2.1360271437384334e-05, + "loss": 3.5521, + "step": 1385 + }, + { + "epoch": 0.04275320128475529, + "grad_norm": 0.23879051208496094, + "learning_rate": 2.1375694016039483e-05, + "loss": 3.5512, + "step": 1386 + }, + { + "epoch": 0.04278404775032871, + "grad_norm": 0.2181149572134018, + "learning_rate": 2.1391116594694636e-05, + "loss": 3.5374, + "step": 1387 + }, + { + "epoch": 0.04281489421590212, + "grad_norm": 0.23935148119926453, + "learning_rate": 2.1406539173349784e-05, + "loss": 3.5187, + "step": 1388 + }, + { + "epoch": 0.04284574068147554, + "grad_norm": 0.21894851326942444, + "learning_rate": 2.1421961752004937e-05, + "loss": 3.5352, + "step": 1389 + }, + { + "epoch": 0.04287658714704896, + "grad_norm": 0.22181573510169983, + "learning_rate": 2.143738433066009e-05, + "loss": 3.5181, + "step": 1390 + }, + { + "epoch": 0.04290743361262237, + "grad_norm": 0.22065898776054382, + "learning_rate": 2.1452806909315238e-05, + "loss": 3.5401, + "step": 1391 + }, + { + "epoch": 0.04293828007819579, + "grad_norm": 0.2280014157295227, + "learning_rate": 2.146822948797039e-05, + "loss": 3.5633, + "step": 1392 + }, + { + "epoch": 0.04296912654376921, + "grad_norm": 0.21253809332847595, + "learning_rate": 2.148365206662554e-05, + "loss": 3.5347, + "step": 1393 + }, + { + "epoch": 0.04299997300934262, + "grad_norm": 0.23606958985328674, + "learning_rate": 2.149907464528069e-05, + "loss": 3.5232, + "step": 1394 + }, + { + "epoch": 0.04303081947491604, + "grad_norm": 0.22454626858234406, + "learning_rate": 2.1514497223935844e-05, + "loss": 3.555, + "step": 1395 + }, + { + "epoch": 0.043061665940489456, + "grad_norm": 0.23860299587249756, + "learning_rate": 2.1529919802590996e-05, + "loss": 3.5659, + "step": 1396 + }, + { + "epoch": 0.04309251240606287, + "grad_norm": 0.23637306690216064, + "learning_rate": 2.1545342381246145e-05, + "loss": 3.51, + "step": 1397 + }, + { + "epoch": 0.04312335887163629, + "grad_norm": 0.23484346270561218, + "learning_rate": 2.1560764959901297e-05, + "loss": 3.5484, + "step": 1398 + }, + { + "epoch": 0.043154205337209706, + "grad_norm": 0.23739397525787354, + "learning_rate": 2.1576187538556446e-05, + "loss": 3.5125, + "step": 1399 + }, + { + "epoch": 0.04318505180278312, + "grad_norm": 0.21771050989627838, + "learning_rate": 2.15916101172116e-05, + "loss": 3.4939, + "step": 1400 + }, + { + "epoch": 0.04321589826835654, + "grad_norm": 0.23829257488250732, + "learning_rate": 2.160703269586675e-05, + "loss": 3.5104, + "step": 1401 + }, + { + "epoch": 0.043246744733929955, + "grad_norm": 0.22412773966789246, + "learning_rate": 2.1622455274521903e-05, + "loss": 3.5402, + "step": 1402 + }, + { + "epoch": 0.04327759119950337, + "grad_norm": 0.24093563854694366, + "learning_rate": 2.1637877853177052e-05, + "loss": 3.5014, + "step": 1403 + }, + { + "epoch": 0.04330843766507679, + "grad_norm": 0.21504248678684235, + "learning_rate": 2.1653300431832204e-05, + "loss": 3.5098, + "step": 1404 + }, + { + "epoch": 0.043339284130650205, + "grad_norm": 0.2347370833158493, + "learning_rate": 2.1668723010487353e-05, + "loss": 3.5334, + "step": 1405 + }, + { + "epoch": 0.04337013059622362, + "grad_norm": 0.21235711872577667, + "learning_rate": 2.1684145589142505e-05, + "loss": 3.5279, + "step": 1406 + }, + { + "epoch": 0.04340097706179704, + "grad_norm": 0.24033792316913605, + "learning_rate": 2.1699568167797658e-05, + "loss": 3.5171, + "step": 1407 + }, + { + "epoch": 0.043431823527370454, + "grad_norm": 0.22948524355888367, + "learning_rate": 2.1714990746452807e-05, + "loss": 3.5285, + "step": 1408 + }, + { + "epoch": 0.04346266999294387, + "grad_norm": 0.2268013209104538, + "learning_rate": 2.173041332510796e-05, + "loss": 3.5046, + "step": 1409 + }, + { + "epoch": 0.04349351645851729, + "grad_norm": 0.2327979952096939, + "learning_rate": 2.1745835903763108e-05, + "loss": 3.5284, + "step": 1410 + }, + { + "epoch": 0.043524362924090704, + "grad_norm": 0.231210395693779, + "learning_rate": 2.1761258482418263e-05, + "loss": 3.5125, + "step": 1411 + }, + { + "epoch": 0.04355520938966412, + "grad_norm": 0.2307559698820114, + "learning_rate": 2.1776681061073412e-05, + "loss": 3.515, + "step": 1412 + }, + { + "epoch": 0.04358605585523754, + "grad_norm": 0.24152453243732452, + "learning_rate": 2.1792103639728565e-05, + "loss": 3.552, + "step": 1413 + }, + { + "epoch": 0.04361690232081095, + "grad_norm": 0.2135423719882965, + "learning_rate": 2.1807526218383714e-05, + "loss": 3.5418, + "step": 1414 + }, + { + "epoch": 0.04364774878638437, + "grad_norm": 0.22993570566177368, + "learning_rate": 2.1822948797038866e-05, + "loss": 3.5572, + "step": 1415 + }, + { + "epoch": 0.04367859525195779, + "grad_norm": 0.21019817888736725, + "learning_rate": 2.1838371375694018e-05, + "loss": 3.5447, + "step": 1416 + }, + { + "epoch": 0.0437094417175312, + "grad_norm": 0.22750435769557953, + "learning_rate": 2.185379395434917e-05, + "loss": 3.5347, + "step": 1417 + }, + { + "epoch": 0.04374028818310462, + "grad_norm": 0.21124528348445892, + "learning_rate": 2.186921653300432e-05, + "loss": 3.5181, + "step": 1418 + }, + { + "epoch": 0.04377113464867804, + "grad_norm": 0.24319998919963837, + "learning_rate": 2.188463911165947e-05, + "loss": 3.5157, + "step": 1419 + }, + { + "epoch": 0.04380198111425145, + "grad_norm": 0.22778835892677307, + "learning_rate": 2.190006169031462e-05, + "loss": 3.4926, + "step": 1420 + }, + { + "epoch": 0.043832827579824866, + "grad_norm": 0.22901290655136108, + "learning_rate": 2.1915484268969773e-05, + "loss": 3.5543, + "step": 1421 + }, + { + "epoch": 0.04386367404539829, + "grad_norm": 0.2374235987663269, + "learning_rate": 2.1930906847624925e-05, + "loss": 3.5624, + "step": 1422 + }, + { + "epoch": 0.0438945205109717, + "grad_norm": 0.20798209309577942, + "learning_rate": 2.1946329426280074e-05, + "loss": 3.4904, + "step": 1423 + }, + { + "epoch": 0.043925366976545116, + "grad_norm": 0.25810760259628296, + "learning_rate": 2.1961752004935226e-05, + "loss": 3.5523, + "step": 1424 + }, + { + "epoch": 0.04395621344211854, + "grad_norm": 0.24110932648181915, + "learning_rate": 2.1977174583590375e-05, + "loss": 3.5111, + "step": 1425 + }, + { + "epoch": 0.04398705990769195, + "grad_norm": 0.23917022347450256, + "learning_rate": 2.1992597162245528e-05, + "loss": 3.5505, + "step": 1426 + }, + { + "epoch": 0.044017906373265366, + "grad_norm": 0.2326761782169342, + "learning_rate": 2.200801974090068e-05, + "loss": 3.5368, + "step": 1427 + }, + { + "epoch": 0.04404875283883879, + "grad_norm": 0.25121384859085083, + "learning_rate": 2.2023442319555832e-05, + "loss": 3.5293, + "step": 1428 + }, + { + "epoch": 0.0440795993044122, + "grad_norm": 0.21831677854061127, + "learning_rate": 2.203886489821098e-05, + "loss": 3.5172, + "step": 1429 + }, + { + "epoch": 0.044110445769985615, + "grad_norm": 0.224837988615036, + "learning_rate": 2.2054287476866133e-05, + "loss": 3.5209, + "step": 1430 + }, + { + "epoch": 0.044141292235559036, + "grad_norm": 0.21752701699733734, + "learning_rate": 2.2069710055521282e-05, + "loss": 3.5329, + "step": 1431 + }, + { + "epoch": 0.04417213870113245, + "grad_norm": 0.23671011626720428, + "learning_rate": 2.2085132634176438e-05, + "loss": 3.5107, + "step": 1432 + }, + { + "epoch": 0.044202985166705865, + "grad_norm": 0.21372224390506744, + "learning_rate": 2.2100555212831587e-05, + "loss": 3.524, + "step": 1433 + }, + { + "epoch": 0.044233831632279286, + "grad_norm": 0.24517932534217834, + "learning_rate": 2.211597779148674e-05, + "loss": 3.5003, + "step": 1434 + }, + { + "epoch": 0.0442646780978527, + "grad_norm": 0.21590539813041687, + "learning_rate": 2.2131400370141888e-05, + "loss": 3.5487, + "step": 1435 + }, + { + "epoch": 0.044295524563426114, + "grad_norm": 0.2485089898109436, + "learning_rate": 2.2146822948797037e-05, + "loss": 3.5598, + "step": 1436 + }, + { + "epoch": 0.044326371028999535, + "grad_norm": 0.22446881234645844, + "learning_rate": 2.2162245527452193e-05, + "loss": 3.5338, + "step": 1437 + }, + { + "epoch": 0.04435721749457295, + "grad_norm": 0.2304956018924713, + "learning_rate": 2.217766810610734e-05, + "loss": 3.4816, + "step": 1438 + }, + { + "epoch": 0.044388063960146364, + "grad_norm": 0.231170192360878, + "learning_rate": 2.2193090684762494e-05, + "loss": 3.5578, + "step": 1439 + }, + { + "epoch": 0.044418910425719785, + "grad_norm": 0.23642979562282562, + "learning_rate": 2.2208513263417643e-05, + "loss": 3.5073, + "step": 1440 + }, + { + "epoch": 0.0444497568912932, + "grad_norm": 0.22129635512828827, + "learning_rate": 2.2223935842072795e-05, + "loss": 3.4923, + "step": 1441 + }, + { + "epoch": 0.04448060335686661, + "grad_norm": 0.26146626472473145, + "learning_rate": 2.2239358420727947e-05, + "loss": 3.5167, + "step": 1442 + }, + { + "epoch": 0.044511449822440034, + "grad_norm": 0.22627177834510803, + "learning_rate": 2.22547809993831e-05, + "loss": 3.5305, + "step": 1443 + }, + { + "epoch": 0.04454229628801345, + "grad_norm": 0.243395134806633, + "learning_rate": 2.227020357803825e-05, + "loss": 3.5285, + "step": 1444 + }, + { + "epoch": 0.04457314275358686, + "grad_norm": 0.22408559918403625, + "learning_rate": 2.22856261566934e-05, + "loss": 3.511, + "step": 1445 + }, + { + "epoch": 0.044603989219160284, + "grad_norm": 0.2630980908870697, + "learning_rate": 2.230104873534855e-05, + "loss": 3.532, + "step": 1446 + }, + { + "epoch": 0.0446348356847337, + "grad_norm": 0.22398975491523743, + "learning_rate": 2.2316471314003702e-05, + "loss": 3.5072, + "step": 1447 + }, + { + "epoch": 0.04466568215030711, + "grad_norm": 0.2416234016418457, + "learning_rate": 2.2331893892658854e-05, + "loss": 3.5017, + "step": 1448 + }, + { + "epoch": 0.04469652861588053, + "grad_norm": 0.21780169010162354, + "learning_rate": 2.2347316471314007e-05, + "loss": 3.5222, + "step": 1449 + }, + { + "epoch": 0.04472737508145395, + "grad_norm": 0.22839099168777466, + "learning_rate": 2.2362739049969155e-05, + "loss": 3.5478, + "step": 1450 + }, + { + "epoch": 0.04475822154702736, + "grad_norm": 0.24433141946792603, + "learning_rate": 2.2378161628624308e-05, + "loss": 3.5323, + "step": 1451 + }, + { + "epoch": 0.04478906801260078, + "grad_norm": 0.23032960295677185, + "learning_rate": 2.2393584207279457e-05, + "loss": 3.5459, + "step": 1452 + }, + { + "epoch": 0.0448199144781742, + "grad_norm": 0.2241591066122055, + "learning_rate": 2.240900678593461e-05, + "loss": 3.504, + "step": 1453 + }, + { + "epoch": 0.04485076094374761, + "grad_norm": 0.21492105722427368, + "learning_rate": 2.242442936458976e-05, + "loss": 3.4766, + "step": 1454 + }, + { + "epoch": 0.04488160740932103, + "grad_norm": 0.21381311118602753, + "learning_rate": 2.243985194324491e-05, + "loss": 3.492, + "step": 1455 + }, + { + "epoch": 0.04491245387489445, + "grad_norm": 0.22054703533649445, + "learning_rate": 2.2455274521900062e-05, + "loss": 3.4974, + "step": 1456 + }, + { + "epoch": 0.04494330034046786, + "grad_norm": 0.21623928844928741, + "learning_rate": 2.247069710055521e-05, + "loss": 3.5126, + "step": 1457 + }, + { + "epoch": 0.04497414680604128, + "grad_norm": 0.21905499696731567, + "learning_rate": 2.2486119679210367e-05, + "loss": 3.4884, + "step": 1458 + }, + { + "epoch": 0.045004993271614696, + "grad_norm": 0.2146964967250824, + "learning_rate": 2.2501542257865516e-05, + "loss": 3.5295, + "step": 1459 + }, + { + "epoch": 0.04503583973718811, + "grad_norm": 0.23316438496112823, + "learning_rate": 2.2516964836520668e-05, + "loss": 3.5171, + "step": 1460 + }, + { + "epoch": 0.04506668620276153, + "grad_norm": 0.22618906199932098, + "learning_rate": 2.2532387415175817e-05, + "loss": 3.5264, + "step": 1461 + }, + { + "epoch": 0.045097532668334946, + "grad_norm": 0.24775470793247223, + "learning_rate": 2.254780999383097e-05, + "loss": 3.4908, + "step": 1462 + }, + { + "epoch": 0.04512837913390836, + "grad_norm": 0.22847643494606018, + "learning_rate": 2.2563232572486122e-05, + "loss": 3.5023, + "step": 1463 + }, + { + "epoch": 0.04515922559948178, + "grad_norm": 0.21266405284404755, + "learning_rate": 2.2578655151141274e-05, + "loss": 3.5216, + "step": 1464 + }, + { + "epoch": 0.045190072065055195, + "grad_norm": 0.21662791073322296, + "learning_rate": 2.2594077729796423e-05, + "loss": 3.4994, + "step": 1465 + }, + { + "epoch": 0.04522091853062861, + "grad_norm": 0.238960862159729, + "learning_rate": 2.2609500308451575e-05, + "loss": 3.5526, + "step": 1466 + }, + { + "epoch": 0.04525176499620203, + "grad_norm": 0.22335056960582733, + "learning_rate": 2.2624922887106724e-05, + "loss": 3.5208, + "step": 1467 + }, + { + "epoch": 0.045282611461775445, + "grad_norm": 0.22191718220710754, + "learning_rate": 2.2640345465761876e-05, + "loss": 3.5446, + "step": 1468 + }, + { + "epoch": 0.04531345792734886, + "grad_norm": 0.21444527804851532, + "learning_rate": 2.265576804441703e-05, + "loss": 3.5029, + "step": 1469 + }, + { + "epoch": 0.04534430439292228, + "grad_norm": 0.2250947207212448, + "learning_rate": 2.2671190623072178e-05, + "loss": 3.4909, + "step": 1470 + }, + { + "epoch": 0.045375150858495694, + "grad_norm": 0.21456120908260345, + "learning_rate": 2.268661320172733e-05, + "loss": 3.5058, + "step": 1471 + }, + { + "epoch": 0.04540599732406911, + "grad_norm": 0.22407831251621246, + "learning_rate": 2.270203578038248e-05, + "loss": 3.4816, + "step": 1472 + }, + { + "epoch": 0.04543684378964253, + "grad_norm": 0.21760649979114532, + "learning_rate": 2.271745835903763e-05, + "loss": 3.5167, + "step": 1473 + }, + { + "epoch": 0.045467690255215944, + "grad_norm": 0.20336735248565674, + "learning_rate": 2.2732880937692783e-05, + "loss": 3.5064, + "step": 1474 + }, + { + "epoch": 0.04549853672078936, + "grad_norm": 0.22252929210662842, + "learning_rate": 2.2748303516347936e-05, + "loss": 3.5367, + "step": 1475 + }, + { + "epoch": 0.04552938318636278, + "grad_norm": 0.21197614073753357, + "learning_rate": 2.2763726095003085e-05, + "loss": 3.5362, + "step": 1476 + }, + { + "epoch": 0.04556022965193619, + "grad_norm": 0.22530899941921234, + "learning_rate": 2.2779148673658237e-05, + "loss": 3.4871, + "step": 1477 + }, + { + "epoch": 0.04559107611750961, + "grad_norm": 0.2419249415397644, + "learning_rate": 2.279457125231339e-05, + "loss": 3.5135, + "step": 1478 + }, + { + "epoch": 0.04562192258308303, + "grad_norm": 0.21957167983055115, + "learning_rate": 2.280999383096854e-05, + "loss": 3.4857, + "step": 1479 + }, + { + "epoch": 0.04565276904865644, + "grad_norm": 0.22422190010547638, + "learning_rate": 2.282541640962369e-05, + "loss": 3.5158, + "step": 1480 + }, + { + "epoch": 0.04568361551422986, + "grad_norm": 0.22538581490516663, + "learning_rate": 2.2840838988278843e-05, + "loss": 3.4778, + "step": 1481 + }, + { + "epoch": 0.04571446197980328, + "grad_norm": 0.2307002991437912, + "learning_rate": 2.285626156693399e-05, + "loss": 3.5419, + "step": 1482 + }, + { + "epoch": 0.04574530844537669, + "grad_norm": 0.21909424662590027, + "learning_rate": 2.2871684145589144e-05, + "loss": 3.5142, + "step": 1483 + }, + { + "epoch": 0.045776154910950106, + "grad_norm": 0.21141724288463593, + "learning_rate": 2.2887106724244296e-05, + "loss": 3.507, + "step": 1484 + }, + { + "epoch": 0.04580700137652353, + "grad_norm": 0.22339200973510742, + "learning_rate": 2.2902529302899445e-05, + "loss": 3.5273, + "step": 1485 + }, + { + "epoch": 0.04583784784209694, + "grad_norm": 0.2271803617477417, + "learning_rate": 2.2917951881554597e-05, + "loss": 3.5206, + "step": 1486 + }, + { + "epoch": 0.045868694307670356, + "grad_norm": 0.21025048196315765, + "learning_rate": 2.2933374460209746e-05, + "loss": 3.5174, + "step": 1487 + }, + { + "epoch": 0.04589954077324378, + "grad_norm": 0.22294940054416656, + "learning_rate": 2.29487970388649e-05, + "loss": 3.5245, + "step": 1488 + }, + { + "epoch": 0.04593038723881719, + "grad_norm": 0.21334421634674072, + "learning_rate": 2.296421961752005e-05, + "loss": 3.4948, + "step": 1489 + }, + { + "epoch": 0.045961233704390606, + "grad_norm": 0.22682546079158783, + "learning_rate": 2.2979642196175203e-05, + "loss": 3.5184, + "step": 1490 + }, + { + "epoch": 0.04599208016996403, + "grad_norm": 0.22177337110042572, + "learning_rate": 2.2995064774830352e-05, + "loss": 3.5023, + "step": 1491 + }, + { + "epoch": 0.04602292663553744, + "grad_norm": 0.20898549258708954, + "learning_rate": 2.3010487353485504e-05, + "loss": 3.4709, + "step": 1492 + }, + { + "epoch": 0.046053773101110855, + "grad_norm": 0.239373579621315, + "learning_rate": 2.3025909932140653e-05, + "loss": 3.5389, + "step": 1493 + }, + { + "epoch": 0.046084619566684276, + "grad_norm": 0.2138511836528778, + "learning_rate": 2.3041332510795806e-05, + "loss": 3.5331, + "step": 1494 + }, + { + "epoch": 0.04611546603225769, + "grad_norm": 0.22621172666549683, + "learning_rate": 2.3056755089450958e-05, + "loss": 3.4887, + "step": 1495 + }, + { + "epoch": 0.046146312497831105, + "grad_norm": 0.21689970791339874, + "learning_rate": 2.307217766810611e-05, + "loss": 3.5057, + "step": 1496 + }, + { + "epoch": 0.046177158963404526, + "grad_norm": 0.2137114405632019, + "learning_rate": 2.308760024676126e-05, + "loss": 3.536, + "step": 1497 + }, + { + "epoch": 0.04620800542897794, + "grad_norm": 0.21465514600276947, + "learning_rate": 2.3103022825416408e-05, + "loss": 3.4978, + "step": 1498 + }, + { + "epoch": 0.046238851894551354, + "grad_norm": 0.2036920189857483, + "learning_rate": 2.3118445404071564e-05, + "loss": 3.5221, + "step": 1499 + }, + { + "epoch": 0.046269698360124775, + "grad_norm": 0.22266119718551636, + "learning_rate": 2.3133867982726713e-05, + "loss": 3.5202, + "step": 1500 + }, + { + "epoch": 0.04630054482569819, + "grad_norm": 0.22999392449855804, + "learning_rate": 2.3149290561381865e-05, + "loss": 3.4582, + "step": 1501 + }, + { + "epoch": 0.046331391291271604, + "grad_norm": 0.21618644893169403, + "learning_rate": 2.3164713140037014e-05, + "loss": 3.5035, + "step": 1502 + }, + { + "epoch": 0.046362237756845025, + "grad_norm": 0.21795512735843658, + "learning_rate": 2.3180135718692166e-05, + "loss": 3.4697, + "step": 1503 + }, + { + "epoch": 0.04639308422241844, + "grad_norm": 0.21234484016895294, + "learning_rate": 2.319555829734732e-05, + "loss": 3.4726, + "step": 1504 + }, + { + "epoch": 0.04642393068799185, + "grad_norm": 0.22512637078762054, + "learning_rate": 2.321098087600247e-05, + "loss": 3.4825, + "step": 1505 + }, + { + "epoch": 0.046454777153565274, + "grad_norm": 0.2178320586681366, + "learning_rate": 2.322640345465762e-05, + "loss": 3.4759, + "step": 1506 + }, + { + "epoch": 0.04648562361913869, + "grad_norm": 0.23933248221874237, + "learning_rate": 2.3241826033312772e-05, + "loss": 3.5163, + "step": 1507 + }, + { + "epoch": 0.0465164700847121, + "grad_norm": 0.2156447172164917, + "learning_rate": 2.325724861196792e-05, + "loss": 3.4762, + "step": 1508 + }, + { + "epoch": 0.046547316550285524, + "grad_norm": 0.25216352939605713, + "learning_rate": 2.3272671190623073e-05, + "loss": 3.4965, + "step": 1509 + }, + { + "epoch": 0.04657816301585894, + "grad_norm": 0.21366186439990997, + "learning_rate": 2.3288093769278225e-05, + "loss": 3.4966, + "step": 1510 + }, + { + "epoch": 0.04660900948143235, + "grad_norm": 0.24079008400440216, + "learning_rate": 2.3303516347933378e-05, + "loss": 3.5042, + "step": 1511 + }, + { + "epoch": 0.04663985594700577, + "grad_norm": 0.21490442752838135, + "learning_rate": 2.3318938926588527e-05, + "loss": 3.4707, + "step": 1512 + }, + { + "epoch": 0.04667070241257919, + "grad_norm": 0.2328464686870575, + "learning_rate": 2.3334361505243675e-05, + "loss": 3.521, + "step": 1513 + }, + { + "epoch": 0.04670154887815261, + "grad_norm": 0.23060698807239532, + "learning_rate": 2.3349784083898828e-05, + "loss": 3.4986, + "step": 1514 + }, + { + "epoch": 0.04673239534372602, + "grad_norm": 0.24272359907627106, + "learning_rate": 2.336520666255398e-05, + "loss": 3.4813, + "step": 1515 + }, + { + "epoch": 0.04676324180929944, + "grad_norm": 0.21782554686069489, + "learning_rate": 2.3380629241209132e-05, + "loss": 3.5021, + "step": 1516 + }, + { + "epoch": 0.04679408827487286, + "grad_norm": 0.3457781970500946, + "learning_rate": 2.339605181986428e-05, + "loss": 3.4979, + "step": 1517 + }, + { + "epoch": 0.04682493474044627, + "grad_norm": 0.21481551229953766, + "learning_rate": 2.3411474398519434e-05, + "loss": 3.4976, + "step": 1518 + }, + { + "epoch": 0.04685578120601969, + "grad_norm": 0.2521607577800751, + "learning_rate": 2.3426896977174582e-05, + "loss": 3.5056, + "step": 1519 + }, + { + "epoch": 0.04688662767159311, + "grad_norm": 0.2154860943555832, + "learning_rate": 2.3442319555829738e-05, + "loss": 3.502, + "step": 1520 + }, + { + "epoch": 0.04691747413716652, + "grad_norm": 0.2475544512271881, + "learning_rate": 2.3457742134484887e-05, + "loss": 3.5173, + "step": 1521 + }, + { + "epoch": 0.046948320602739936, + "grad_norm": 0.22401884198188782, + "learning_rate": 2.347316471314004e-05, + "loss": 3.5241, + "step": 1522 + }, + { + "epoch": 0.04697916706831336, + "grad_norm": 0.22211973369121552, + "learning_rate": 2.3488587291795188e-05, + "loss": 3.5156, + "step": 1523 + }, + { + "epoch": 0.04701001353388677, + "grad_norm": 0.22315163910388947, + "learning_rate": 2.350400987045034e-05, + "loss": 3.5071, + "step": 1524 + }, + { + "epoch": 0.047040859999460186, + "grad_norm": 0.22305959463119507, + "learning_rate": 2.3519432449105493e-05, + "loss": 3.5181, + "step": 1525 + }, + { + "epoch": 0.04707170646503361, + "grad_norm": 0.21782657504081726, + "learning_rate": 2.3534855027760645e-05, + "loss": 3.5083, + "step": 1526 + }, + { + "epoch": 0.04710255293060702, + "grad_norm": 0.22208473086357117, + "learning_rate": 2.3550277606415794e-05, + "loss": 3.5126, + "step": 1527 + }, + { + "epoch": 0.047133399396180435, + "grad_norm": 0.22705870866775513, + "learning_rate": 2.3565700185070946e-05, + "loss": 3.4664, + "step": 1528 + }, + { + "epoch": 0.047164245861753856, + "grad_norm": 0.21963652968406677, + "learning_rate": 2.3581122763726095e-05, + "loss": 3.5062, + "step": 1529 + }, + { + "epoch": 0.04719509232732727, + "grad_norm": 0.22861681878566742, + "learning_rate": 2.3596545342381248e-05, + "loss": 3.4938, + "step": 1530 + }, + { + "epoch": 0.047225938792900685, + "grad_norm": 0.21688520908355713, + "learning_rate": 2.36119679210364e-05, + "loss": 3.4949, + "step": 1531 + }, + { + "epoch": 0.047256785258474106, + "grad_norm": 0.2363302856683731, + "learning_rate": 2.362739049969155e-05, + "loss": 3.4745, + "step": 1532 + }, + { + "epoch": 0.04728763172404752, + "grad_norm": 0.23529206216335297, + "learning_rate": 2.36428130783467e-05, + "loss": 3.465, + "step": 1533 + }, + { + "epoch": 0.047318478189620934, + "grad_norm": 0.22213689982891083, + "learning_rate": 2.365823565700185e-05, + "loss": 3.4974, + "step": 1534 + }, + { + "epoch": 0.047349324655194355, + "grad_norm": 0.23797348141670227, + "learning_rate": 2.3673658235657002e-05, + "loss": 3.5383, + "step": 1535 + }, + { + "epoch": 0.04738017112076777, + "grad_norm": 0.22555601596832275, + "learning_rate": 2.3689080814312154e-05, + "loss": 3.5338, + "step": 1536 + }, + { + "epoch": 0.047411017586341184, + "grad_norm": 0.23291316628456116, + "learning_rate": 2.3704503392967307e-05, + "loss": 3.4869, + "step": 1537 + }, + { + "epoch": 0.047441864051914605, + "grad_norm": 0.22997817397117615, + "learning_rate": 2.3719925971622456e-05, + "loss": 3.4741, + "step": 1538 + }, + { + "epoch": 0.04747271051748802, + "grad_norm": 0.23283976316452026, + "learning_rate": 2.3735348550277608e-05, + "loss": 3.4593, + "step": 1539 + }, + { + "epoch": 0.04750355698306143, + "grad_norm": 0.2462237924337387, + "learning_rate": 2.3750771128932757e-05, + "loss": 3.5056, + "step": 1540 + }, + { + "epoch": 0.047534403448634854, + "grad_norm": 0.2257695198059082, + "learning_rate": 2.3766193707587913e-05, + "loss": 3.5086, + "step": 1541 + }, + { + "epoch": 0.04756524991420827, + "grad_norm": 0.22382386028766632, + "learning_rate": 2.378161628624306e-05, + "loss": 3.5333, + "step": 1542 + }, + { + "epoch": 0.04759609637978168, + "grad_norm": 0.29678046703338623, + "learning_rate": 2.3797038864898214e-05, + "loss": 3.4909, + "step": 1543 + }, + { + "epoch": 0.047626942845355104, + "grad_norm": 0.2368413209915161, + "learning_rate": 2.3812461443553363e-05, + "loss": 3.4855, + "step": 1544 + }, + { + "epoch": 0.04765778931092852, + "grad_norm": 0.22945816814899445, + "learning_rate": 2.382788402220851e-05, + "loss": 3.4811, + "step": 1545 + }, + { + "epoch": 0.04768863577650193, + "grad_norm": 0.23738142848014832, + "learning_rate": 2.3843306600863667e-05, + "loss": 3.5229, + "step": 1546 + }, + { + "epoch": 0.04771948224207535, + "grad_norm": 0.23493649065494537, + "learning_rate": 2.3858729179518816e-05, + "loss": 3.5277, + "step": 1547 + }, + { + "epoch": 0.04775032870764877, + "grad_norm": 0.22167985141277313, + "learning_rate": 2.387415175817397e-05, + "loss": 3.4658, + "step": 1548 + }, + { + "epoch": 0.04778117517322218, + "grad_norm": 0.2860850691795349, + "learning_rate": 2.3889574336829117e-05, + "loss": 3.5062, + "step": 1549 + }, + { + "epoch": 0.0478120216387956, + "grad_norm": 0.2233666628599167, + "learning_rate": 2.390499691548427e-05, + "loss": 3.4864, + "step": 1550 + }, + { + "epoch": 0.04784286810436902, + "grad_norm": 0.24985814094543457, + "learning_rate": 2.3920419494139422e-05, + "loss": 3.4549, + "step": 1551 + }, + { + "epoch": 0.04787371456994243, + "grad_norm": 0.2171219140291214, + "learning_rate": 2.3935842072794574e-05, + "loss": 3.4959, + "step": 1552 + }, + { + "epoch": 0.04790456103551585, + "grad_norm": 0.2480589747428894, + "learning_rate": 2.3951264651449723e-05, + "loss": 3.5244, + "step": 1553 + }, + { + "epoch": 0.04793540750108927, + "grad_norm": 0.21455474197864532, + "learning_rate": 2.3966687230104875e-05, + "loss": 3.5187, + "step": 1554 + }, + { + "epoch": 0.04796625396666268, + "grad_norm": 0.23398911952972412, + "learning_rate": 2.3982109808760024e-05, + "loss": 3.4655, + "step": 1555 + }, + { + "epoch": 0.0479971004322361, + "grad_norm": 0.20742109417915344, + "learning_rate": 2.3997532387415177e-05, + "loss": 3.4871, + "step": 1556 + }, + { + "epoch": 0.048027946897809516, + "grad_norm": 0.24831591546535492, + "learning_rate": 2.401295496607033e-05, + "loss": 3.4788, + "step": 1557 + }, + { + "epoch": 0.04805879336338293, + "grad_norm": 0.21937377750873566, + "learning_rate": 2.402837754472548e-05, + "loss": 3.5284, + "step": 1558 + }, + { + "epoch": 0.04808963982895635, + "grad_norm": 0.2338317632675171, + "learning_rate": 2.404380012338063e-05, + "loss": 3.5156, + "step": 1559 + }, + { + "epoch": 0.048120486294529766, + "grad_norm": 0.22903822362422943, + "learning_rate": 2.405922270203578e-05, + "loss": 3.4881, + "step": 1560 + }, + { + "epoch": 0.04815133276010318, + "grad_norm": 0.22326608002185822, + "learning_rate": 2.407464528069093e-05, + "loss": 3.4725, + "step": 1561 + }, + { + "epoch": 0.0481821792256766, + "grad_norm": 0.22868721187114716, + "learning_rate": 2.4090067859346084e-05, + "loss": 3.5242, + "step": 1562 + }, + { + "epoch": 0.048213025691250015, + "grad_norm": 0.2345913052558899, + "learning_rate": 2.4105490438001236e-05, + "loss": 3.5077, + "step": 1563 + }, + { + "epoch": 0.04824387215682343, + "grad_norm": 0.2205367535352707, + "learning_rate": 2.4120913016656385e-05, + "loss": 3.5227, + "step": 1564 + }, + { + "epoch": 0.04827471862239685, + "grad_norm": 0.24369174242019653, + "learning_rate": 2.4136335595311537e-05, + "loss": 3.5218, + "step": 1565 + }, + { + "epoch": 0.048305565087970265, + "grad_norm": 0.20132991671562195, + "learning_rate": 2.4151758173966686e-05, + "loss": 3.5051, + "step": 1566 + }, + { + "epoch": 0.04833641155354368, + "grad_norm": 0.3743053674697876, + "learning_rate": 2.4167180752621842e-05, + "loss": 3.4889, + "step": 1567 + }, + { + "epoch": 0.0483672580191171, + "grad_norm": 0.21329358220100403, + "learning_rate": 2.418260333127699e-05, + "loss": 3.5009, + "step": 1568 + }, + { + "epoch": 0.048398104484690514, + "grad_norm": 0.24265195429325104, + "learning_rate": 2.4198025909932143e-05, + "loss": 3.486, + "step": 1569 + }, + { + "epoch": 0.04842895095026393, + "grad_norm": 0.22309410572052002, + "learning_rate": 2.4213448488587292e-05, + "loss": 3.4955, + "step": 1570 + }, + { + "epoch": 0.04845979741583735, + "grad_norm": 0.2558354139328003, + "learning_rate": 2.4228871067242444e-05, + "loss": 3.5098, + "step": 1571 + }, + { + "epoch": 0.048490643881410764, + "grad_norm": 0.21615538001060486, + "learning_rate": 2.4244293645897596e-05, + "loss": 3.4965, + "step": 1572 + }, + { + "epoch": 0.04852149034698418, + "grad_norm": 0.2284162938594818, + "learning_rate": 2.425971622455275e-05, + "loss": 3.4687, + "step": 1573 + }, + { + "epoch": 0.0485523368125576, + "grad_norm": 0.2327912449836731, + "learning_rate": 2.4275138803207898e-05, + "loss": 3.4886, + "step": 1574 + }, + { + "epoch": 0.04858318327813101, + "grad_norm": 0.22298315167427063, + "learning_rate": 2.4290561381863047e-05, + "loss": 3.4806, + "step": 1575 + }, + { + "epoch": 0.04861402974370443, + "grad_norm": 0.22061601281166077, + "learning_rate": 2.43059839605182e-05, + "loss": 3.4663, + "step": 1576 + }, + { + "epoch": 0.04864487620927785, + "grad_norm": 0.2382347732782364, + "learning_rate": 2.432140653917335e-05, + "loss": 3.5147, + "step": 1577 + }, + { + "epoch": 0.04867572267485126, + "grad_norm": 0.21713759005069733, + "learning_rate": 2.4336829117828503e-05, + "loss": 3.4629, + "step": 1578 + }, + { + "epoch": 0.04870656914042468, + "grad_norm": 0.23343555629253387, + "learning_rate": 2.4352251696483652e-05, + "loss": 3.4769, + "step": 1579 + }, + { + "epoch": 0.0487374156059981, + "grad_norm": 0.2330176681280136, + "learning_rate": 2.4367674275138805e-05, + "loss": 3.5035, + "step": 1580 + }, + { + "epoch": 0.04876826207157151, + "grad_norm": 0.2343248724937439, + "learning_rate": 2.4383096853793953e-05, + "loss": 3.4824, + "step": 1581 + }, + { + "epoch": 0.048799108537144927, + "grad_norm": 0.21932755410671234, + "learning_rate": 2.4398519432449106e-05, + "loss": 3.5076, + "step": 1582 + }, + { + "epoch": 0.04882995500271835, + "grad_norm": 0.2257014662027359, + "learning_rate": 2.4413942011104258e-05, + "loss": 3.481, + "step": 1583 + }, + { + "epoch": 0.04886080146829176, + "grad_norm": 0.22502362728118896, + "learning_rate": 2.442936458975941e-05, + "loss": 3.5101, + "step": 1584 + }, + { + "epoch": 0.048891647933865176, + "grad_norm": 0.2313494086265564, + "learning_rate": 2.444478716841456e-05, + "loss": 3.4957, + "step": 1585 + }, + { + "epoch": 0.0489224943994386, + "grad_norm": 0.3219587802886963, + "learning_rate": 2.446020974706971e-05, + "loss": 3.5344, + "step": 1586 + }, + { + "epoch": 0.04895334086501201, + "grad_norm": 0.23284834623336792, + "learning_rate": 2.447563232572486e-05, + "loss": 3.4692, + "step": 1587 + }, + { + "epoch": 0.048984187330585426, + "grad_norm": 0.23583771288394928, + "learning_rate": 2.4491054904380016e-05, + "loss": 3.4755, + "step": 1588 + }, + { + "epoch": 0.04901503379615885, + "grad_norm": 0.22149714827537537, + "learning_rate": 2.4506477483035165e-05, + "loss": 3.4547, + "step": 1589 + }, + { + "epoch": 0.04904588026173226, + "grad_norm": 0.2283712476491928, + "learning_rate": 2.4521900061690317e-05, + "loss": 3.487, + "step": 1590 + }, + { + "epoch": 0.049076726727305675, + "grad_norm": 0.23345622420310974, + "learning_rate": 2.4537322640345466e-05, + "loss": 3.5224, + "step": 1591 + }, + { + "epoch": 0.049107573192879096, + "grad_norm": 0.22915932536125183, + "learning_rate": 2.4552745219000615e-05, + "loss": 3.4959, + "step": 1592 + }, + { + "epoch": 0.04913841965845251, + "grad_norm": 0.2595207691192627, + "learning_rate": 2.456816779765577e-05, + "loss": 3.5047, + "step": 1593 + }, + { + "epoch": 0.049169266124025925, + "grad_norm": 0.23516450822353363, + "learning_rate": 2.458359037631092e-05, + "loss": 3.4881, + "step": 1594 + }, + { + "epoch": 0.049200112589599346, + "grad_norm": 0.239092618227005, + "learning_rate": 2.4599012954966072e-05, + "loss": 3.5181, + "step": 1595 + }, + { + "epoch": 0.04923095905517276, + "grad_norm": 0.22223152220249176, + "learning_rate": 2.461443553362122e-05, + "loss": 3.5051, + "step": 1596 + }, + { + "epoch": 0.049261805520746174, + "grad_norm": 0.2143213003873825, + "learning_rate": 2.4629858112276373e-05, + "loss": 3.451, + "step": 1597 + }, + { + "epoch": 0.049292651986319595, + "grad_norm": 0.22331278026103973, + "learning_rate": 2.4645280690931526e-05, + "loss": 3.5244, + "step": 1598 + }, + { + "epoch": 0.04932349845189301, + "grad_norm": 0.21433137357234955, + "learning_rate": 2.4660703269586678e-05, + "loss": 3.4684, + "step": 1599 + }, + { + "epoch": 0.049354344917466424, + "grad_norm": 0.22210028767585754, + "learning_rate": 2.4676125848241827e-05, + "loss": 3.4937, + "step": 1600 + }, + { + "epoch": 0.049385191383039845, + "grad_norm": 0.2254543900489807, + "learning_rate": 2.469154842689698e-05, + "loss": 3.4883, + "step": 1601 + }, + { + "epoch": 0.04941603784861326, + "grad_norm": 0.21790625154972076, + "learning_rate": 2.4706971005552128e-05, + "loss": 3.4804, + "step": 1602 + }, + { + "epoch": 0.04944688431418667, + "grad_norm": 0.23129907250404358, + "learning_rate": 2.472239358420728e-05, + "loss": 3.4894, + "step": 1603 + }, + { + "epoch": 0.049477730779760094, + "grad_norm": 0.21605758368968964, + "learning_rate": 2.4737816162862433e-05, + "loss": 3.4684, + "step": 1604 + }, + { + "epoch": 0.04950857724533351, + "grad_norm": 0.21472887694835663, + "learning_rate": 2.4753238741517585e-05, + "loss": 3.5139, + "step": 1605 + }, + { + "epoch": 0.04953942371090692, + "grad_norm": 0.21763825416564941, + "learning_rate": 2.4768661320172734e-05, + "loss": 3.4707, + "step": 1606 + }, + { + "epoch": 0.049570270176480344, + "grad_norm": 0.22559358179569244, + "learning_rate": 2.4784083898827883e-05, + "loss": 3.4947, + "step": 1607 + }, + { + "epoch": 0.04960111664205376, + "grad_norm": 0.24255552887916565, + "learning_rate": 2.4799506477483035e-05, + "loss": 3.4851, + "step": 1608 + }, + { + "epoch": 0.04963196310762717, + "grad_norm": 0.2175436019897461, + "learning_rate": 2.4814929056138187e-05, + "loss": 3.4922, + "step": 1609 + }, + { + "epoch": 0.04966280957320059, + "grad_norm": 0.23830674588680267, + "learning_rate": 2.483035163479334e-05, + "loss": 3.4786, + "step": 1610 + }, + { + "epoch": 0.04969365603877401, + "grad_norm": 0.2207736223936081, + "learning_rate": 2.484577421344849e-05, + "loss": 3.4609, + "step": 1611 + }, + { + "epoch": 0.04972450250434742, + "grad_norm": 0.22492298483848572, + "learning_rate": 2.486119679210364e-05, + "loss": 3.5192, + "step": 1612 + }, + { + "epoch": 0.04975534896992084, + "grad_norm": 0.22953468561172485, + "learning_rate": 2.4876619370758793e-05, + "loss": 3.5361, + "step": 1613 + }, + { + "epoch": 0.04978619543549426, + "grad_norm": 0.22060205042362213, + "learning_rate": 2.4892041949413945e-05, + "loss": 3.4555, + "step": 1614 + }, + { + "epoch": 0.04981704190106767, + "grad_norm": 0.24701526761054993, + "learning_rate": 2.4907464528069094e-05, + "loss": 3.5104, + "step": 1615 + }, + { + "epoch": 0.04984788836664109, + "grad_norm": 0.2296900451183319, + "learning_rate": 2.4922887106724247e-05, + "loss": 3.4592, + "step": 1616 + }, + { + "epoch": 0.04987873483221451, + "grad_norm": 0.2218155413866043, + "learning_rate": 2.4938309685379395e-05, + "loss": 3.4711, + "step": 1617 + }, + { + "epoch": 0.04990958129778792, + "grad_norm": 0.24320748448371887, + "learning_rate": 2.4953732264034548e-05, + "loss": 3.4811, + "step": 1618 + }, + { + "epoch": 0.04994042776336134, + "grad_norm": 0.22385169565677643, + "learning_rate": 2.49691548426897e-05, + "loss": 3.4762, + "step": 1619 + }, + { + "epoch": 0.049971274228934756, + "grad_norm": 0.22502224147319794, + "learning_rate": 2.4984577421344852e-05, + "loss": 3.4971, + "step": 1620 + }, + { + "epoch": 0.05000212069450817, + "grad_norm": 0.2289135605096817, + "learning_rate": 2.5e-05, + "loss": 3.4699, + "step": 1621 + }, + { + "epoch": 0.05003296716008159, + "grad_norm": 0.21431808173656464, + "learning_rate": 2.501542257865515e-05, + "loss": 3.4526, + "step": 1622 + }, + { + "epoch": 0.050063813625655006, + "grad_norm": 0.24819952249526978, + "learning_rate": 2.5030845157310302e-05, + "loss": 3.4621, + "step": 1623 + }, + { + "epoch": 0.05009466009122842, + "grad_norm": 0.21426181495189667, + "learning_rate": 2.5046267735965455e-05, + "loss": 3.4805, + "step": 1624 + }, + { + "epoch": 0.05012550655680184, + "grad_norm": 0.2392580509185791, + "learning_rate": 2.5061690314620607e-05, + "loss": 3.4958, + "step": 1625 + }, + { + "epoch": 0.050156353022375255, + "grad_norm": 0.2326546162366867, + "learning_rate": 2.5077112893275756e-05, + "loss": 3.4716, + "step": 1626 + }, + { + "epoch": 0.05018719948794867, + "grad_norm": 0.24908195436000824, + "learning_rate": 2.5092535471930905e-05, + "loss": 3.518, + "step": 1627 + }, + { + "epoch": 0.05021804595352209, + "grad_norm": 0.21967516839504242, + "learning_rate": 2.510795805058606e-05, + "loss": 3.4729, + "step": 1628 + }, + { + "epoch": 0.050248892419095505, + "grad_norm": 0.21294231712818146, + "learning_rate": 2.512338062924121e-05, + "loss": 3.4963, + "step": 1629 + }, + { + "epoch": 0.05027973888466892, + "grad_norm": 0.21470512449741364, + "learning_rate": 2.513880320789636e-05, + "loss": 3.4661, + "step": 1630 + }, + { + "epoch": 0.05031058535024234, + "grad_norm": 0.23325732350349426, + "learning_rate": 2.515422578655151e-05, + "loss": 3.4766, + "step": 1631 + }, + { + "epoch": 0.050341431815815754, + "grad_norm": 0.21641597151756287, + "learning_rate": 2.5169648365206666e-05, + "loss": 3.4885, + "step": 1632 + }, + { + "epoch": 0.05037227828138917, + "grad_norm": 0.2179093360900879, + "learning_rate": 2.5185070943861815e-05, + "loss": 3.5195, + "step": 1633 + }, + { + "epoch": 0.05040312474696259, + "grad_norm": 0.2136697620153427, + "learning_rate": 2.5200493522516967e-05, + "loss": 3.4539, + "step": 1634 + }, + { + "epoch": 0.050433971212536004, + "grad_norm": 0.2099001258611679, + "learning_rate": 2.5215916101172116e-05, + "loss": 3.5046, + "step": 1635 + }, + { + "epoch": 0.05046481767810942, + "grad_norm": 0.202602818608284, + "learning_rate": 2.5231338679827272e-05, + "loss": 3.4769, + "step": 1636 + }, + { + "epoch": 0.05049566414368284, + "grad_norm": 0.2144894301891327, + "learning_rate": 2.524676125848242e-05, + "loss": 3.467, + "step": 1637 + }, + { + "epoch": 0.05052651060925625, + "grad_norm": 0.20605291426181793, + "learning_rate": 2.526218383713757e-05, + "loss": 3.4814, + "step": 1638 + }, + { + "epoch": 0.05055735707482967, + "grad_norm": 0.21372255682945251, + "learning_rate": 2.5277606415792722e-05, + "loss": 3.4647, + "step": 1639 + }, + { + "epoch": 0.05058820354040309, + "grad_norm": 0.20304235816001892, + "learning_rate": 2.529302899444787e-05, + "loss": 3.4613, + "step": 1640 + }, + { + "epoch": 0.0506190500059765, + "grad_norm": 0.21924416720867157, + "learning_rate": 2.5308451573103027e-05, + "loss": 3.4984, + "step": 1641 + }, + { + "epoch": 0.05064989647154992, + "grad_norm": 0.20707330107688904, + "learning_rate": 2.5323874151758176e-05, + "loss": 3.4908, + "step": 1642 + }, + { + "epoch": 0.05068074293712334, + "grad_norm": 0.21555553376674652, + "learning_rate": 2.5339296730413325e-05, + "loss": 3.4176, + "step": 1643 + }, + { + "epoch": 0.05071158940269675, + "grad_norm": 0.21145038306713104, + "learning_rate": 2.5354719309068477e-05, + "loss": 3.4783, + "step": 1644 + }, + { + "epoch": 0.050742435868270166, + "grad_norm": 0.23238115012645721, + "learning_rate": 2.537014188772363e-05, + "loss": 3.4971, + "step": 1645 + }, + { + "epoch": 0.05077328233384359, + "grad_norm": 0.20536011457443237, + "learning_rate": 2.538556446637878e-05, + "loss": 3.4705, + "step": 1646 + }, + { + "epoch": 0.050804128799417, + "grad_norm": 0.25116732716560364, + "learning_rate": 2.540098704503393e-05, + "loss": 3.4825, + "step": 1647 + }, + { + "epoch": 0.050834975264990416, + "grad_norm": 0.20755918323993683, + "learning_rate": 2.541640962368908e-05, + "loss": 3.4543, + "step": 1648 + }, + { + "epoch": 0.05086582173056384, + "grad_norm": 0.22777311503887177, + "learning_rate": 2.5431832202344235e-05, + "loss": 3.4492, + "step": 1649 + }, + { + "epoch": 0.05089666819613725, + "grad_norm": 0.2121964544057846, + "learning_rate": 2.5447254780999387e-05, + "loss": 3.4893, + "step": 1650 + }, + { + "epoch": 0.050927514661710666, + "grad_norm": 0.22980041801929474, + "learning_rate": 2.5462677359654536e-05, + "loss": 3.4978, + "step": 1651 + }, + { + "epoch": 0.05095836112728409, + "grad_norm": 0.22009587287902832, + "learning_rate": 2.5478099938309685e-05, + "loss": 3.4563, + "step": 1652 + }, + { + "epoch": 0.0509892075928575, + "grad_norm": 0.2229454070329666, + "learning_rate": 2.5493522516964834e-05, + "loss": 3.4724, + "step": 1653 + }, + { + "epoch": 0.051020054058430915, + "grad_norm": 0.21955549716949463, + "learning_rate": 2.550894509561999e-05, + "loss": 3.4797, + "step": 1654 + }, + { + "epoch": 0.051050900524004336, + "grad_norm": 0.23641623556613922, + "learning_rate": 2.5524367674275142e-05, + "loss": 3.5358, + "step": 1655 + }, + { + "epoch": 0.05108174698957775, + "grad_norm": 0.21525338292121887, + "learning_rate": 2.553979025293029e-05, + "loss": 3.4705, + "step": 1656 + }, + { + "epoch": 0.051112593455151165, + "grad_norm": 0.23000042140483856, + "learning_rate": 2.555521283158544e-05, + "loss": 3.4876, + "step": 1657 + }, + { + "epoch": 0.051143439920724586, + "grad_norm": 0.21894347667694092, + "learning_rate": 2.5570635410240595e-05, + "loss": 3.5025, + "step": 1658 + }, + { + "epoch": 0.051174286386298, + "grad_norm": 0.2685544490814209, + "learning_rate": 2.5586057988895744e-05, + "loss": 3.4612, + "step": 1659 + }, + { + "epoch": 0.051205132851871414, + "grad_norm": 0.2158500999212265, + "learning_rate": 2.5601480567550897e-05, + "loss": 3.4845, + "step": 1660 + }, + { + "epoch": 0.051235979317444835, + "grad_norm": 0.2513350248336792, + "learning_rate": 2.5616903146206046e-05, + "loss": 3.4773, + "step": 1661 + }, + { + "epoch": 0.05126682578301825, + "grad_norm": 0.21366776525974274, + "learning_rate": 2.56323257248612e-05, + "loss": 3.4615, + "step": 1662 + }, + { + "epoch": 0.051297672248591664, + "grad_norm": 0.2432953268289566, + "learning_rate": 2.564774830351635e-05, + "loss": 3.4774, + "step": 1663 + }, + { + "epoch": 0.051328518714165085, + "grad_norm": 0.21470889449119568, + "learning_rate": 2.56631708821715e-05, + "loss": 3.4656, + "step": 1664 + }, + { + "epoch": 0.0513593651797385, + "grad_norm": 0.23149427771568298, + "learning_rate": 2.567859346082665e-05, + "loss": 3.4939, + "step": 1665 + }, + { + "epoch": 0.05139021164531191, + "grad_norm": 0.21098364889621735, + "learning_rate": 2.5694016039481804e-05, + "loss": 3.5253, + "step": 1666 + }, + { + "epoch": 0.051421058110885334, + "grad_norm": 0.22476863861083984, + "learning_rate": 2.5709438618136956e-05, + "loss": 3.4895, + "step": 1667 + }, + { + "epoch": 0.05145190457645875, + "grad_norm": 0.22441533207893372, + "learning_rate": 2.5724861196792105e-05, + "loss": 3.4754, + "step": 1668 + }, + { + "epoch": 0.05148275104203216, + "grad_norm": 0.21920987963676453, + "learning_rate": 2.5740283775447254e-05, + "loss": 3.4777, + "step": 1669 + }, + { + "epoch": 0.051513597507605584, + "grad_norm": 0.23310036957263947, + "learning_rate": 2.5755706354102406e-05, + "loss": 3.484, + "step": 1670 + }, + { + "epoch": 0.051544443973179, + "grad_norm": 0.21342644095420837, + "learning_rate": 2.577112893275756e-05, + "loss": 3.4747, + "step": 1671 + }, + { + "epoch": 0.05157529043875241, + "grad_norm": 0.22101429104804993, + "learning_rate": 2.578655151141271e-05, + "loss": 3.4544, + "step": 1672 + }, + { + "epoch": 0.05160613690432583, + "grad_norm": 0.20460771024227142, + "learning_rate": 2.580197409006786e-05, + "loss": 3.4292, + "step": 1673 + }, + { + "epoch": 0.05163698336989925, + "grad_norm": 0.21634137630462646, + "learning_rate": 2.581739666872301e-05, + "loss": 3.4475, + "step": 1674 + }, + { + "epoch": 0.05166782983547266, + "grad_norm": 0.21042339503765106, + "learning_rate": 2.5832819247378164e-05, + "loss": 3.4893, + "step": 1675 + }, + { + "epoch": 0.05169867630104608, + "grad_norm": 0.22934842109680176, + "learning_rate": 2.5848241826033316e-05, + "loss": 3.4494, + "step": 1676 + }, + { + "epoch": 0.0517295227666195, + "grad_norm": 0.2168094664812088, + "learning_rate": 2.5863664404688465e-05, + "loss": 3.4326, + "step": 1677 + }, + { + "epoch": 0.05176036923219291, + "grad_norm": 0.21527744829654694, + "learning_rate": 2.5879086983343614e-05, + "loss": 3.4852, + "step": 1678 + }, + { + "epoch": 0.05179121569776633, + "grad_norm": 0.21454358100891113, + "learning_rate": 2.589450956199877e-05, + "loss": 3.4535, + "step": 1679 + }, + { + "epoch": 0.05182206216333975, + "grad_norm": 0.2184884399175644, + "learning_rate": 2.590993214065392e-05, + "loss": 3.4867, + "step": 1680 + }, + { + "epoch": 0.05185290862891316, + "grad_norm": 0.21877513825893402, + "learning_rate": 2.592535471930907e-05, + "loss": 3.4513, + "step": 1681 + }, + { + "epoch": 0.05188375509448658, + "grad_norm": 0.22647695243358612, + "learning_rate": 2.594077729796422e-05, + "loss": 3.4801, + "step": 1682 + }, + { + "epoch": 0.051914601560059996, + "grad_norm": 0.2559882700443268, + "learning_rate": 2.595619987661937e-05, + "loss": 3.506, + "step": 1683 + }, + { + "epoch": 0.05194544802563341, + "grad_norm": 0.2312449812889099, + "learning_rate": 2.5971622455274525e-05, + "loss": 3.4822, + "step": 1684 + }, + { + "epoch": 0.05197629449120683, + "grad_norm": 0.2170533686876297, + "learning_rate": 2.5987045033929673e-05, + "loss": 3.4573, + "step": 1685 + }, + { + "epoch": 0.052007140956780246, + "grad_norm": 0.2396116852760315, + "learning_rate": 2.6002467612584826e-05, + "loss": 3.4556, + "step": 1686 + }, + { + "epoch": 0.05203798742235366, + "grad_norm": 0.2444802075624466, + "learning_rate": 2.6017890191239975e-05, + "loss": 3.4621, + "step": 1687 + }, + { + "epoch": 0.05206883388792708, + "grad_norm": 0.22087012231349945, + "learning_rate": 2.603331276989513e-05, + "loss": 3.4778, + "step": 1688 + }, + { + "epoch": 0.052099680353500495, + "grad_norm": 0.23571957647800446, + "learning_rate": 2.604873534855028e-05, + "loss": 3.4243, + "step": 1689 + }, + { + "epoch": 0.05213052681907391, + "grad_norm": 0.21988171339035034, + "learning_rate": 2.6064157927205428e-05, + "loss": 3.4617, + "step": 1690 + }, + { + "epoch": 0.05216137328464733, + "grad_norm": 0.23251059651374817, + "learning_rate": 2.607958050586058e-05, + "loss": 3.4423, + "step": 1691 + }, + { + "epoch": 0.052192219750220745, + "grad_norm": 0.237881600856781, + "learning_rate": 2.6095003084515736e-05, + "loss": 3.4853, + "step": 1692 + }, + { + "epoch": 0.05222306621579416, + "grad_norm": 0.22688201069831848, + "learning_rate": 2.6110425663170885e-05, + "loss": 3.4763, + "step": 1693 + }, + { + "epoch": 0.05225391268136758, + "grad_norm": 0.24191492795944214, + "learning_rate": 2.6125848241826034e-05, + "loss": 3.4879, + "step": 1694 + }, + { + "epoch": 0.052284759146940994, + "grad_norm": 0.2161644548177719, + "learning_rate": 2.6141270820481183e-05, + "loss": 3.4705, + "step": 1695 + }, + { + "epoch": 0.05231560561251441, + "grad_norm": 0.22941170632839203, + "learning_rate": 2.615669339913634e-05, + "loss": 3.4798, + "step": 1696 + }, + { + "epoch": 0.05234645207808783, + "grad_norm": 0.20582911372184753, + "learning_rate": 2.617211597779149e-05, + "loss": 3.462, + "step": 1697 + }, + { + "epoch": 0.052377298543661244, + "grad_norm": 0.2113560289144516, + "learning_rate": 2.618753855644664e-05, + "loss": 3.4684, + "step": 1698 + }, + { + "epoch": 0.05240814500923466, + "grad_norm": 0.20596130192279816, + "learning_rate": 2.620296113510179e-05, + "loss": 3.4188, + "step": 1699 + }, + { + "epoch": 0.05243899147480808, + "grad_norm": 0.21067427098751068, + "learning_rate": 2.6218383713756938e-05, + "loss": 3.471, + "step": 1700 + }, + { + "epoch": 0.05246983794038149, + "grad_norm": 0.21156297624111176, + "learning_rate": 2.6233806292412093e-05, + "loss": 3.4391, + "step": 1701 + }, + { + "epoch": 0.05250068440595491, + "grad_norm": 0.21239334344863892, + "learning_rate": 2.6249228871067246e-05, + "loss": 3.4404, + "step": 1702 + }, + { + "epoch": 0.05253153087152833, + "grad_norm": 0.21857401728630066, + "learning_rate": 2.6264651449722394e-05, + "loss": 3.5017, + "step": 1703 + }, + { + "epoch": 0.05256237733710174, + "grad_norm": 0.22375430166721344, + "learning_rate": 2.6280074028377543e-05, + "loss": 3.4778, + "step": 1704 + }, + { + "epoch": 0.05259322380267516, + "grad_norm": 0.20808011293411255, + "learning_rate": 2.62954966070327e-05, + "loss": 3.4719, + "step": 1705 + }, + { + "epoch": 0.05262407026824858, + "grad_norm": 0.2232886701822281, + "learning_rate": 2.6310919185687848e-05, + "loss": 3.454, + "step": 1706 + }, + { + "epoch": 0.05265491673382199, + "grad_norm": 0.21551771461963654, + "learning_rate": 2.6326341764343e-05, + "loss": 3.4543, + "step": 1707 + }, + { + "epoch": 0.052685763199395406, + "grad_norm": 0.22981752455234528, + "learning_rate": 2.634176434299815e-05, + "loss": 3.4146, + "step": 1708 + }, + { + "epoch": 0.05271660966496883, + "grad_norm": 0.20702801644802094, + "learning_rate": 2.6357186921653305e-05, + "loss": 3.4672, + "step": 1709 + }, + { + "epoch": 0.05274745613054224, + "grad_norm": 0.22586989402770996, + "learning_rate": 2.6372609500308454e-05, + "loss": 3.4564, + "step": 1710 + }, + { + "epoch": 0.052778302596115656, + "grad_norm": 0.22369450330734253, + "learning_rate": 2.6388032078963603e-05, + "loss": 3.4215, + "step": 1711 + }, + { + "epoch": 0.05280914906168908, + "grad_norm": 0.2211315631866455, + "learning_rate": 2.6403454657618755e-05, + "loss": 3.4932, + "step": 1712 + }, + { + "epoch": 0.05283999552726249, + "grad_norm": 0.2329055666923523, + "learning_rate": 2.641887723627391e-05, + "loss": 3.4399, + "step": 1713 + }, + { + "epoch": 0.052870841992835906, + "grad_norm": 0.20834533870220184, + "learning_rate": 2.643429981492906e-05, + "loss": 3.4113, + "step": 1714 + }, + { + "epoch": 0.05290168845840933, + "grad_norm": 0.23837414383888245, + "learning_rate": 2.644972239358421e-05, + "loss": 3.4741, + "step": 1715 + }, + { + "epoch": 0.05293253492398274, + "grad_norm": 0.2133001834154129, + "learning_rate": 2.6465144972239357e-05, + "loss": 3.4496, + "step": 1716 + }, + { + "epoch": 0.052963381389556155, + "grad_norm": 0.22828367352485657, + "learning_rate": 2.648056755089451e-05, + "loss": 3.4541, + "step": 1717 + }, + { + "epoch": 0.052994227855129576, + "grad_norm": 0.238260880112648, + "learning_rate": 2.6495990129549665e-05, + "loss": 3.4245, + "step": 1718 + }, + { + "epoch": 0.05302507432070299, + "grad_norm": 0.23143097758293152, + "learning_rate": 2.6511412708204814e-05, + "loss": 3.4651, + "step": 1719 + }, + { + "epoch": 0.053055920786276405, + "grad_norm": 0.25080132484436035, + "learning_rate": 2.6526835286859963e-05, + "loss": 3.4563, + "step": 1720 + }, + { + "epoch": 0.053086767251849826, + "grad_norm": 0.22002437710762024, + "learning_rate": 2.6542257865515112e-05, + "loss": 3.4707, + "step": 1721 + }, + { + "epoch": 0.05311761371742324, + "grad_norm": 0.23904365301132202, + "learning_rate": 2.6557680444170268e-05, + "loss": 3.4433, + "step": 1722 + }, + { + "epoch": 0.053148460182996654, + "grad_norm": 0.2141844630241394, + "learning_rate": 2.657310302282542e-05, + "loss": 3.4697, + "step": 1723 + }, + { + "epoch": 0.053179306648570075, + "grad_norm": 0.2282436341047287, + "learning_rate": 2.658852560148057e-05, + "loss": 3.4291, + "step": 1724 + }, + { + "epoch": 0.05321015311414349, + "grad_norm": 0.22688381373882294, + "learning_rate": 2.6603948180135718e-05, + "loss": 3.4505, + "step": 1725 + }, + { + "epoch": 0.053240999579716904, + "grad_norm": 0.23070664703845978, + "learning_rate": 2.6619370758790873e-05, + "loss": 3.427, + "step": 1726 + }, + { + "epoch": 0.053271846045290325, + "grad_norm": 0.24702197313308716, + "learning_rate": 2.6634793337446022e-05, + "loss": 3.462, + "step": 1727 + }, + { + "epoch": 0.05330269251086374, + "grad_norm": 0.2243805080652237, + "learning_rate": 2.6650215916101175e-05, + "loss": 3.4566, + "step": 1728 + }, + { + "epoch": 0.05333353897643715, + "grad_norm": 0.2188766598701477, + "learning_rate": 2.6665638494756324e-05, + "loss": 3.4563, + "step": 1729 + }, + { + "epoch": 0.053364385442010574, + "grad_norm": 0.2104344666004181, + "learning_rate": 2.6681061073411472e-05, + "loss": 3.4853, + "step": 1730 + }, + { + "epoch": 0.05339523190758399, + "grad_norm": 0.22990034520626068, + "learning_rate": 2.6696483652066628e-05, + "loss": 3.4533, + "step": 1731 + }, + { + "epoch": 0.0534260783731574, + "grad_norm": 0.21389706432819366, + "learning_rate": 2.6711906230721777e-05, + "loss": 3.4608, + "step": 1732 + }, + { + "epoch": 0.053456924838730824, + "grad_norm": 0.2813508212566376, + "learning_rate": 2.672732880937693e-05, + "loss": 3.5114, + "step": 1733 + }, + { + "epoch": 0.05348777130430424, + "grad_norm": 0.2410842776298523, + "learning_rate": 2.6742751388032078e-05, + "loss": 3.4655, + "step": 1734 + }, + { + "epoch": 0.05351861776987765, + "grad_norm": 0.21866106986999512, + "learning_rate": 2.6758173966687234e-05, + "loss": 3.4648, + "step": 1735 + }, + { + "epoch": 0.05354946423545107, + "grad_norm": 0.3115043044090271, + "learning_rate": 2.6773596545342383e-05, + "loss": 3.4909, + "step": 1736 + }, + { + "epoch": 0.05358031070102449, + "grad_norm": 0.2226424366235733, + "learning_rate": 2.6789019123997532e-05, + "loss": 3.4429, + "step": 1737 + }, + { + "epoch": 0.0536111571665979, + "grad_norm": 0.22843101620674133, + "learning_rate": 2.6804441702652684e-05, + "loss": 3.4886, + "step": 1738 + }, + { + "epoch": 0.05364200363217132, + "grad_norm": 0.23172159492969513, + "learning_rate": 2.681986428130784e-05, + "loss": 3.4791, + "step": 1739 + }, + { + "epoch": 0.05367285009774474, + "grad_norm": 0.22075214982032776, + "learning_rate": 2.683528685996299e-05, + "loss": 3.4551, + "step": 1740 + }, + { + "epoch": 0.05370369656331815, + "grad_norm": 0.22216755151748657, + "learning_rate": 2.6850709438618138e-05, + "loss": 3.4588, + "step": 1741 + }, + { + "epoch": 0.05373454302889157, + "grad_norm": 0.22201314568519592, + "learning_rate": 2.6866132017273286e-05, + "loss": 3.4729, + "step": 1742 + }, + { + "epoch": 0.05376538949446499, + "grad_norm": 0.21917986869812012, + "learning_rate": 2.6881554595928442e-05, + "loss": 3.4929, + "step": 1743 + }, + { + "epoch": 0.0537962359600384, + "grad_norm": 0.2057534009218216, + "learning_rate": 2.6896977174583594e-05, + "loss": 3.4729, + "step": 1744 + }, + { + "epoch": 0.05382708242561182, + "grad_norm": 0.20642000436782837, + "learning_rate": 2.6912399753238743e-05, + "loss": 3.4615, + "step": 1745 + }, + { + "epoch": 0.053857928891185236, + "grad_norm": 0.21536225080490112, + "learning_rate": 2.6927822331893892e-05, + "loss": 3.4539, + "step": 1746 + }, + { + "epoch": 0.05388877535675865, + "grad_norm": 0.21887952089309692, + "learning_rate": 2.694324491054904e-05, + "loss": 3.4456, + "step": 1747 + }, + { + "epoch": 0.05391962182233207, + "grad_norm": 0.22926804423332214, + "learning_rate": 2.6958667489204197e-05, + "loss": 3.4619, + "step": 1748 + }, + { + "epoch": 0.053950468287905486, + "grad_norm": 0.44517430663108826, + "learning_rate": 2.697409006785935e-05, + "loss": 3.4462, + "step": 1749 + }, + { + "epoch": 0.0539813147534789, + "grad_norm": 0.2265234738588333, + "learning_rate": 2.6989512646514498e-05, + "loss": 3.4911, + "step": 1750 + }, + { + "epoch": 0.05401216121905232, + "grad_norm": 0.21711771190166473, + "learning_rate": 2.7004935225169647e-05, + "loss": 3.4404, + "step": 1751 + }, + { + "epoch": 0.054043007684625735, + "grad_norm": 0.22019663453102112, + "learning_rate": 2.7020357803824803e-05, + "loss": 3.5026, + "step": 1752 + }, + { + "epoch": 0.05407385415019915, + "grad_norm": 0.2202875167131424, + "learning_rate": 2.703578038247995e-05, + "loss": 3.4416, + "step": 1753 + }, + { + "epoch": 0.05410470061577257, + "grad_norm": 0.23327292501926422, + "learning_rate": 2.7051202961135104e-05, + "loss": 3.4753, + "step": 1754 + }, + { + "epoch": 0.054135547081345985, + "grad_norm": 0.23495887219905853, + "learning_rate": 2.7066625539790253e-05, + "loss": 3.4567, + "step": 1755 + }, + { + "epoch": 0.0541663935469194, + "grad_norm": 0.24440249800682068, + "learning_rate": 2.708204811844541e-05, + "loss": 3.4692, + "step": 1756 + }, + { + "epoch": 0.05419724001249282, + "grad_norm": 0.24275097250938416, + "learning_rate": 2.7097470697100557e-05, + "loss": 3.4287, + "step": 1757 + }, + { + "epoch": 0.054228086478066234, + "grad_norm": 0.21187008917331696, + "learning_rate": 2.7112893275755706e-05, + "loss": 3.4633, + "step": 1758 + }, + { + "epoch": 0.05425893294363965, + "grad_norm": 0.2267332524061203, + "learning_rate": 2.712831585441086e-05, + "loss": 3.4526, + "step": 1759 + }, + { + "epoch": 0.05428977940921307, + "grad_norm": 0.2141246795654297, + "learning_rate": 2.7143738433066007e-05, + "loss": 3.4838, + "step": 1760 + }, + { + "epoch": 0.054320625874786484, + "grad_norm": 0.2341742068529129, + "learning_rate": 2.7159161011721163e-05, + "loss": 3.4465, + "step": 1761 + }, + { + "epoch": 0.0543514723403599, + "grad_norm": 0.258120596408844, + "learning_rate": 2.7174583590376312e-05, + "loss": 3.4457, + "step": 1762 + }, + { + "epoch": 0.05438231880593332, + "grad_norm": 0.2300146222114563, + "learning_rate": 2.719000616903146e-05, + "loss": 3.4457, + "step": 1763 + }, + { + "epoch": 0.05441316527150673, + "grad_norm": 0.22238121926784515, + "learning_rate": 2.7205428747686613e-05, + "loss": 3.4477, + "step": 1764 + }, + { + "epoch": 0.05444401173708015, + "grad_norm": 0.22210079431533813, + "learning_rate": 2.722085132634177e-05, + "loss": 3.492, + "step": 1765 + }, + { + "epoch": 0.05447485820265357, + "grad_norm": 0.20831458270549774, + "learning_rate": 2.7236273904996918e-05, + "loss": 3.4513, + "step": 1766 + }, + { + "epoch": 0.05450570466822698, + "grad_norm": 0.21987122297286987, + "learning_rate": 2.7251696483652067e-05, + "loss": 3.4116, + "step": 1767 + }, + { + "epoch": 0.0545365511338004, + "grad_norm": 0.2244614064693451, + "learning_rate": 2.7267119062307216e-05, + "loss": 3.4394, + "step": 1768 + }, + { + "epoch": 0.05456739759937382, + "grad_norm": 0.21605640649795532, + "learning_rate": 2.728254164096237e-05, + "loss": 3.4964, + "step": 1769 + }, + { + "epoch": 0.05459824406494723, + "grad_norm": 0.22724957764148712, + "learning_rate": 2.7297964219617524e-05, + "loss": 3.4722, + "step": 1770 + }, + { + "epoch": 0.054629090530520646, + "grad_norm": 0.20767779648303986, + "learning_rate": 2.7313386798272672e-05, + "loss": 3.4372, + "step": 1771 + }, + { + "epoch": 0.05465993699609407, + "grad_norm": 0.3012153208255768, + "learning_rate": 2.732880937692782e-05, + "loss": 3.4342, + "step": 1772 + }, + { + "epoch": 0.05469078346166748, + "grad_norm": 0.2299102395772934, + "learning_rate": 2.7344231955582977e-05, + "loss": 3.4343, + "step": 1773 + }, + { + "epoch": 0.054721629927240896, + "grad_norm": 0.21756939589977264, + "learning_rate": 2.7359654534238126e-05, + "loss": 3.414, + "step": 1774 + }, + { + "epoch": 0.05475247639281432, + "grad_norm": 0.23542167246341705, + "learning_rate": 2.7375077112893278e-05, + "loss": 3.4476, + "step": 1775 + }, + { + "epoch": 0.05478332285838773, + "grad_norm": 0.21443048119544983, + "learning_rate": 2.7390499691548427e-05, + "loss": 3.4648, + "step": 1776 + }, + { + "epoch": 0.054814169323961145, + "grad_norm": 0.2455175668001175, + "learning_rate": 2.7405922270203576e-05, + "loss": 3.4733, + "step": 1777 + }, + { + "epoch": 0.05484501578953457, + "grad_norm": 0.21511417627334595, + "learning_rate": 2.7421344848858732e-05, + "loss": 3.4461, + "step": 1778 + }, + { + "epoch": 0.05487586225510798, + "grad_norm": 0.21865415573120117, + "learning_rate": 2.743676742751388e-05, + "loss": 3.4416, + "step": 1779 + }, + { + "epoch": 0.054906708720681395, + "grad_norm": 0.4178365468978882, + "learning_rate": 2.7452190006169033e-05, + "loss": 3.4263, + "step": 1780 + }, + { + "epoch": 0.054937555186254816, + "grad_norm": 0.2794886529445648, + "learning_rate": 2.7467612584824182e-05, + "loss": 3.4718, + "step": 1781 + }, + { + "epoch": 0.05496840165182823, + "grad_norm": 0.21826724708080292, + "learning_rate": 2.7483035163479338e-05, + "loss": 3.4397, + "step": 1782 + }, + { + "epoch": 0.054999248117401645, + "grad_norm": 0.22300389409065247, + "learning_rate": 2.7498457742134486e-05, + "loss": 3.453, + "step": 1783 + }, + { + "epoch": 0.055030094582975066, + "grad_norm": 0.209990993142128, + "learning_rate": 2.7513880320789635e-05, + "loss": 3.4929, + "step": 1784 + }, + { + "epoch": 0.05506094104854848, + "grad_norm": 0.2539306879043579, + "learning_rate": 2.7529302899444788e-05, + "loss": 3.4318, + "step": 1785 + }, + { + "epoch": 0.055091787514121894, + "grad_norm": 0.30704203248023987, + "learning_rate": 2.7544725478099943e-05, + "loss": 3.4441, + "step": 1786 + }, + { + "epoch": 0.055122633979695315, + "grad_norm": 0.23566915094852448, + "learning_rate": 2.7560148056755092e-05, + "loss": 3.4676, + "step": 1787 + }, + { + "epoch": 0.05515348044526873, + "grad_norm": 0.21709467470645905, + "learning_rate": 2.757557063541024e-05, + "loss": 3.4135, + "step": 1788 + }, + { + "epoch": 0.05518432691084215, + "grad_norm": 0.24418558180332184, + "learning_rate": 2.759099321406539e-05, + "loss": 3.4833, + "step": 1789 + }, + { + "epoch": 0.055215173376415565, + "grad_norm": 0.22187206149101257, + "learning_rate": 2.7606415792720546e-05, + "loss": 3.4888, + "step": 1790 + }, + { + "epoch": 0.05524601984198898, + "grad_norm": 0.23263989388942719, + "learning_rate": 2.7621838371375698e-05, + "loss": 3.4917, + "step": 1791 + }, + { + "epoch": 0.0552768663075624, + "grad_norm": 0.2174757421016693, + "learning_rate": 2.7637260950030847e-05, + "loss": 3.4495, + "step": 1792 + }, + { + "epoch": 0.055307712773135814, + "grad_norm": 0.2329171895980835, + "learning_rate": 2.7652683528685996e-05, + "loss": 3.4755, + "step": 1793 + }, + { + "epoch": 0.05533855923870923, + "grad_norm": 0.2216310054063797, + "learning_rate": 2.7668106107341145e-05, + "loss": 3.4341, + "step": 1794 + }, + { + "epoch": 0.05536940570428265, + "grad_norm": 0.21517391502857208, + "learning_rate": 2.76835286859963e-05, + "loss": 3.4557, + "step": 1795 + }, + { + "epoch": 0.055400252169856064, + "grad_norm": 0.23165732622146606, + "learning_rate": 2.7698951264651453e-05, + "loss": 3.4213, + "step": 1796 + }, + { + "epoch": 0.05543109863542948, + "grad_norm": 0.22708043456077576, + "learning_rate": 2.77143738433066e-05, + "loss": 3.4308, + "step": 1797 + }, + { + "epoch": 0.0554619451010029, + "grad_norm": 0.2199406921863556, + "learning_rate": 2.772979642196175e-05, + "loss": 3.488, + "step": 1798 + }, + { + "epoch": 0.05549279156657631, + "grad_norm": 0.21627213060855865, + "learning_rate": 2.7745219000616906e-05, + "loss": 3.4296, + "step": 1799 + }, + { + "epoch": 0.05552363803214973, + "grad_norm": 0.21556402742862701, + "learning_rate": 2.7760641579272055e-05, + "loss": 3.4069, + "step": 1800 + }, + { + "epoch": 0.05555448449772315, + "grad_norm": 0.2228003740310669, + "learning_rate": 2.7776064157927207e-05, + "loss": 3.4625, + "step": 1801 + }, + { + "epoch": 0.05558533096329656, + "grad_norm": 0.22190430760383606, + "learning_rate": 2.7791486736582356e-05, + "loss": 3.4447, + "step": 1802 + }, + { + "epoch": 0.05561617742886998, + "grad_norm": 0.2160140424966812, + "learning_rate": 2.7806909315237512e-05, + "loss": 3.4668, + "step": 1803 + }, + { + "epoch": 0.0556470238944434, + "grad_norm": 0.22586770355701447, + "learning_rate": 2.782233189389266e-05, + "loss": 3.4476, + "step": 1804 + }, + { + "epoch": 0.05567787036001681, + "grad_norm": 0.2572517693042755, + "learning_rate": 2.783775447254781e-05, + "loss": 3.4454, + "step": 1805 + }, + { + "epoch": 0.055708716825590227, + "grad_norm": 0.22243262827396393, + "learning_rate": 2.7853177051202962e-05, + "loss": 3.4402, + "step": 1806 + }, + { + "epoch": 0.05573956329116365, + "grad_norm": 0.23458904027938843, + "learning_rate": 2.786859962985811e-05, + "loss": 3.4443, + "step": 1807 + }, + { + "epoch": 0.05577040975673706, + "grad_norm": 0.22993530333042145, + "learning_rate": 2.7884022208513267e-05, + "loss": 3.5052, + "step": 1808 + }, + { + "epoch": 0.055801256222310476, + "grad_norm": 0.21548877656459808, + "learning_rate": 2.7899444787168416e-05, + "loss": 3.4207, + "step": 1809 + }, + { + "epoch": 0.0558321026878839, + "grad_norm": 0.2317223846912384, + "learning_rate": 2.7914867365823564e-05, + "loss": 3.4812, + "step": 1810 + }, + { + "epoch": 0.05586294915345731, + "grad_norm": 0.22535400092601776, + "learning_rate": 2.7930289944478717e-05, + "loss": 3.4417, + "step": 1811 + }, + { + "epoch": 0.055893795619030726, + "grad_norm": 0.21792486310005188, + "learning_rate": 2.7945712523133872e-05, + "loss": 3.4435, + "step": 1812 + }, + { + "epoch": 0.05592464208460415, + "grad_norm": 0.2178640514612198, + "learning_rate": 2.796113510178902e-05, + "loss": 3.442, + "step": 1813 + }, + { + "epoch": 0.05595548855017756, + "grad_norm": 0.2069755494594574, + "learning_rate": 2.797655768044417e-05, + "loss": 3.4403, + "step": 1814 + }, + { + "epoch": 0.055986335015750975, + "grad_norm": 0.21587003767490387, + "learning_rate": 2.799198025909932e-05, + "loss": 3.4808, + "step": 1815 + }, + { + "epoch": 0.056017181481324396, + "grad_norm": 0.26982352137565613, + "learning_rate": 2.8007402837754475e-05, + "loss": 3.4547, + "step": 1816 + }, + { + "epoch": 0.05604802794689781, + "grad_norm": 0.21154671907424927, + "learning_rate": 2.8022825416409627e-05, + "loss": 3.4774, + "step": 1817 + }, + { + "epoch": 0.056078874412471225, + "grad_norm": 0.22037993371486664, + "learning_rate": 2.8038247995064776e-05, + "loss": 3.4635, + "step": 1818 + }, + { + "epoch": 0.056109720878044646, + "grad_norm": 0.2136654108762741, + "learning_rate": 2.8053670573719925e-05, + "loss": 3.4816, + "step": 1819 + }, + { + "epoch": 0.05614056734361806, + "grad_norm": 0.21136468648910522, + "learning_rate": 2.806909315237508e-05, + "loss": 3.4318, + "step": 1820 + }, + { + "epoch": 0.056171413809191474, + "grad_norm": 0.21630056202411652, + "learning_rate": 2.808451573103023e-05, + "loss": 3.4401, + "step": 1821 + }, + { + "epoch": 0.056202260274764895, + "grad_norm": 0.22875455021858215, + "learning_rate": 2.8099938309685382e-05, + "loss": 3.4566, + "step": 1822 + }, + { + "epoch": 0.05623310674033831, + "grad_norm": 0.21445931494235992, + "learning_rate": 2.811536088834053e-05, + "loss": 3.4462, + "step": 1823 + }, + { + "epoch": 0.056263953205911724, + "grad_norm": 0.21266081929206848, + "learning_rate": 2.813078346699568e-05, + "loss": 3.4592, + "step": 1824 + }, + { + "epoch": 0.056294799671485145, + "grad_norm": 0.21856887638568878, + "learning_rate": 2.8146206045650835e-05, + "loss": 3.4869, + "step": 1825 + }, + { + "epoch": 0.05632564613705856, + "grad_norm": 0.1961311250925064, + "learning_rate": 2.8161628624305984e-05, + "loss": 3.4283, + "step": 1826 + }, + { + "epoch": 0.05635649260263197, + "grad_norm": 0.22128155827522278, + "learning_rate": 2.8177051202961137e-05, + "loss": 3.4679, + "step": 1827 + }, + { + "epoch": 0.056387339068205394, + "grad_norm": 0.21163570880889893, + "learning_rate": 2.8192473781616285e-05, + "loss": 3.4855, + "step": 1828 + }, + { + "epoch": 0.05641818553377881, + "grad_norm": 0.21586070954799652, + "learning_rate": 2.820789636027144e-05, + "loss": 3.4355, + "step": 1829 + }, + { + "epoch": 0.05644903199935222, + "grad_norm": 0.21511267125606537, + "learning_rate": 2.822331893892659e-05, + "loss": 3.3891, + "step": 1830 + }, + { + "epoch": 0.056479878464925644, + "grad_norm": 0.2117455005645752, + "learning_rate": 2.823874151758174e-05, + "loss": 3.4525, + "step": 1831 + }, + { + "epoch": 0.05651072493049906, + "grad_norm": 0.22034800052642822, + "learning_rate": 2.825416409623689e-05, + "loss": 3.4123, + "step": 1832 + }, + { + "epoch": 0.05654157139607247, + "grad_norm": 0.20253965258598328, + "learning_rate": 2.8269586674892047e-05, + "loss": 3.4361, + "step": 1833 + }, + { + "epoch": 0.05657241786164589, + "grad_norm": 0.20383915305137634, + "learning_rate": 2.8285009253547196e-05, + "loss": 3.4521, + "step": 1834 + }, + { + "epoch": 0.05660326432721931, + "grad_norm": 0.2251727432012558, + "learning_rate": 2.8300431832202345e-05, + "loss": 3.3977, + "step": 1835 + }, + { + "epoch": 0.05663411079279272, + "grad_norm": 0.20456455647945404, + "learning_rate": 2.8315854410857494e-05, + "loss": 3.4429, + "step": 1836 + }, + { + "epoch": 0.05666495725836614, + "grad_norm": 0.21201221644878387, + "learning_rate": 2.833127698951265e-05, + "loss": 3.4293, + "step": 1837 + }, + { + "epoch": 0.05669580372393956, + "grad_norm": 0.21813391149044037, + "learning_rate": 2.83466995681678e-05, + "loss": 3.4673, + "step": 1838 + }, + { + "epoch": 0.05672665018951297, + "grad_norm": 0.2039433866739273, + "learning_rate": 2.836212214682295e-05, + "loss": 3.4487, + "step": 1839 + }, + { + "epoch": 0.05675749665508639, + "grad_norm": 0.21206477284431458, + "learning_rate": 2.83775447254781e-05, + "loss": 3.4147, + "step": 1840 + }, + { + "epoch": 0.05678834312065981, + "grad_norm": 0.2126823365688324, + "learning_rate": 2.8392967304133252e-05, + "loss": 3.419, + "step": 1841 + }, + { + "epoch": 0.05681918958623322, + "grad_norm": 0.2116871029138565, + "learning_rate": 2.8408389882788404e-05, + "loss": 3.4197, + "step": 1842 + }, + { + "epoch": 0.05685003605180664, + "grad_norm": 0.20872777700424194, + "learning_rate": 2.8423812461443556e-05, + "loss": 3.4378, + "step": 1843 + }, + { + "epoch": 0.056880882517380056, + "grad_norm": 0.21398591995239258, + "learning_rate": 2.8439235040098705e-05, + "loss": 3.4356, + "step": 1844 + }, + { + "epoch": 0.05691172898295347, + "grad_norm": 0.20052435994148254, + "learning_rate": 2.8454657618753854e-05, + "loss": 3.4632, + "step": 1845 + }, + { + "epoch": 0.05694257544852689, + "grad_norm": 0.227495476603508, + "learning_rate": 2.847008019740901e-05, + "loss": 3.4246, + "step": 1846 + }, + { + "epoch": 0.056973421914100306, + "grad_norm": 0.20209872722625732, + "learning_rate": 2.848550277606416e-05, + "loss": 3.4446, + "step": 1847 + }, + { + "epoch": 0.05700426837967372, + "grad_norm": 0.20936089754104614, + "learning_rate": 2.850092535471931e-05, + "loss": 3.4426, + "step": 1848 + }, + { + "epoch": 0.05703511484524714, + "grad_norm": 0.20340096950531006, + "learning_rate": 2.851634793337446e-05, + "loss": 3.4172, + "step": 1849 + }, + { + "epoch": 0.057065961310820555, + "grad_norm": 0.20751719176769257, + "learning_rate": 2.8531770512029616e-05, + "loss": 3.4406, + "step": 1850 + }, + { + "epoch": 0.05709680777639397, + "grad_norm": 0.20529621839523315, + "learning_rate": 2.8547193090684764e-05, + "loss": 3.464, + "step": 1851 + }, + { + "epoch": 0.05712765424196739, + "grad_norm": 0.5479379892349243, + "learning_rate": 2.8562615669339913e-05, + "loss": 3.4171, + "step": 1852 + }, + { + "epoch": 0.057158500707540805, + "grad_norm": 0.22995363175868988, + "learning_rate": 2.8578038247995066e-05, + "loss": 3.4415, + "step": 1853 + }, + { + "epoch": 0.05718934717311422, + "grad_norm": 0.22254493832588196, + "learning_rate": 2.8593460826650215e-05, + "loss": 3.429, + "step": 1854 + }, + { + "epoch": 0.05722019363868764, + "grad_norm": 0.23682139813899994, + "learning_rate": 2.860888340530537e-05, + "loss": 3.4321, + "step": 1855 + }, + { + "epoch": 0.057251040104261054, + "grad_norm": 0.22272108495235443, + "learning_rate": 2.862430598396052e-05, + "loss": 3.4491, + "step": 1856 + }, + { + "epoch": 0.05728188656983447, + "grad_norm": 0.2289854884147644, + "learning_rate": 2.8639728562615668e-05, + "loss": 3.4392, + "step": 1857 + }, + { + "epoch": 0.05731273303540789, + "grad_norm": 0.22187405824661255, + "learning_rate": 2.865515114127082e-05, + "loss": 3.459, + "step": 1858 + }, + { + "epoch": 0.057343579500981304, + "grad_norm": 0.23092707991600037, + "learning_rate": 2.8670573719925976e-05, + "loss": 3.4329, + "step": 1859 + }, + { + "epoch": 0.05737442596655472, + "grad_norm": 0.22693145275115967, + "learning_rate": 2.8685996298581125e-05, + "loss": 3.4502, + "step": 1860 + }, + { + "epoch": 0.05740527243212814, + "grad_norm": 0.23615223169326782, + "learning_rate": 2.8701418877236274e-05, + "loss": 3.407, + "step": 1861 + }, + { + "epoch": 0.05743611889770155, + "grad_norm": 0.23411785066127777, + "learning_rate": 2.8716841455891426e-05, + "loss": 3.4754, + "step": 1862 + }, + { + "epoch": 0.05746696536327497, + "grad_norm": 0.22939209640026093, + "learning_rate": 2.873226403454658e-05, + "loss": 3.4015, + "step": 1863 + }, + { + "epoch": 0.05749781182884839, + "grad_norm": 0.21883904933929443, + "learning_rate": 2.874768661320173e-05, + "loss": 3.4455, + "step": 1864 + }, + { + "epoch": 0.0575286582944218, + "grad_norm": 0.22478966414928436, + "learning_rate": 2.876310919185688e-05, + "loss": 3.4431, + "step": 1865 + }, + { + "epoch": 0.05755950475999522, + "grad_norm": 0.20155398547649384, + "learning_rate": 2.877853177051203e-05, + "loss": 3.4298, + "step": 1866 + }, + { + "epoch": 0.05759035122556864, + "grad_norm": 0.22553269565105438, + "learning_rate": 2.8793954349167184e-05, + "loss": 3.4488, + "step": 1867 + }, + { + "epoch": 0.05762119769114205, + "grad_norm": 0.20956219732761383, + "learning_rate": 2.8809376927822333e-05, + "loss": 3.4249, + "step": 1868 + }, + { + "epoch": 0.057652044156715467, + "grad_norm": 0.21265731751918793, + "learning_rate": 2.8824799506477485e-05, + "loss": 3.4265, + "step": 1869 + }, + { + "epoch": 0.05768289062228889, + "grad_norm": 0.20762094855308533, + "learning_rate": 2.8840222085132634e-05, + "loss": 3.4644, + "step": 1870 + }, + { + "epoch": 0.0577137370878623, + "grad_norm": 0.21406421065330505, + "learning_rate": 2.8855644663787783e-05, + "loss": 3.476, + "step": 1871 + }, + { + "epoch": 0.057744583553435716, + "grad_norm": 0.21880683302879333, + "learning_rate": 2.887106724244294e-05, + "loss": 3.4468, + "step": 1872 + }, + { + "epoch": 0.05777543001900914, + "grad_norm": 0.21398574113845825, + "learning_rate": 2.8886489821098088e-05, + "loss": 3.4186, + "step": 1873 + }, + { + "epoch": 0.05780627648458255, + "grad_norm": 0.21096795797348022, + "learning_rate": 2.890191239975324e-05, + "loss": 3.414, + "step": 1874 + }, + { + "epoch": 0.057837122950155966, + "grad_norm": 0.22035695612430573, + "learning_rate": 2.891733497840839e-05, + "loss": 3.4428, + "step": 1875 + }, + { + "epoch": 0.05786796941572939, + "grad_norm": 0.2058515101671219, + "learning_rate": 2.8932757557063545e-05, + "loss": 3.426, + "step": 1876 + }, + { + "epoch": 0.0578988158813028, + "grad_norm": 0.20540335774421692, + "learning_rate": 2.8948180135718694e-05, + "loss": 3.4617, + "step": 1877 + }, + { + "epoch": 0.057929662346876215, + "grad_norm": 0.20237484574317932, + "learning_rate": 2.8963602714373843e-05, + "loss": 3.4432, + "step": 1878 + }, + { + "epoch": 0.057960508812449636, + "grad_norm": 0.2084232121706009, + "learning_rate": 2.8979025293028995e-05, + "loss": 3.414, + "step": 1879 + }, + { + "epoch": 0.05799135527802305, + "grad_norm": 0.2332303673028946, + "learning_rate": 2.899444787168415e-05, + "loss": 3.4208, + "step": 1880 + }, + { + "epoch": 0.058022201743596465, + "grad_norm": 0.21020448207855225, + "learning_rate": 2.90098704503393e-05, + "loss": 3.3975, + "step": 1881 + }, + { + "epoch": 0.058053048209169886, + "grad_norm": 0.2196374088525772, + "learning_rate": 2.9025293028994448e-05, + "loss": 3.4668, + "step": 1882 + }, + { + "epoch": 0.0580838946747433, + "grad_norm": 0.20378413796424866, + "learning_rate": 2.90407156076496e-05, + "loss": 3.4122, + "step": 1883 + }, + { + "epoch": 0.058114741140316714, + "grad_norm": 0.22171954810619354, + "learning_rate": 2.905613818630475e-05, + "loss": 3.4286, + "step": 1884 + }, + { + "epoch": 0.058145587605890135, + "grad_norm": 0.20878738164901733, + "learning_rate": 2.9071560764959905e-05, + "loss": 3.4066, + "step": 1885 + }, + { + "epoch": 0.05817643407146355, + "grad_norm": 0.2062613070011139, + "learning_rate": 2.9086983343615054e-05, + "loss": 3.4447, + "step": 1886 + }, + { + "epoch": 0.058207280537036964, + "grad_norm": 0.2066197246313095, + "learning_rate": 2.9102405922270203e-05, + "loss": 3.4601, + "step": 1887 + }, + { + "epoch": 0.058238127002610385, + "grad_norm": 0.22173580527305603, + "learning_rate": 2.9117828500925355e-05, + "loss": 3.4328, + "step": 1888 + }, + { + "epoch": 0.0582689734681838, + "grad_norm": 0.21132344007492065, + "learning_rate": 2.9133251079580508e-05, + "loss": 3.4299, + "step": 1889 + }, + { + "epoch": 0.05829981993375721, + "grad_norm": 0.2304607629776001, + "learning_rate": 2.914867365823566e-05, + "loss": 3.473, + "step": 1890 + }, + { + "epoch": 0.058330666399330634, + "grad_norm": 0.23437480628490448, + "learning_rate": 2.916409623689081e-05, + "loss": 3.4282, + "step": 1891 + }, + { + "epoch": 0.05836151286490405, + "grad_norm": 0.2233826369047165, + "learning_rate": 2.9179518815545958e-05, + "loss": 3.4104, + "step": 1892 + }, + { + "epoch": 0.05839235933047746, + "grad_norm": 0.220653235912323, + "learning_rate": 2.9194941394201113e-05, + "loss": 3.437, + "step": 1893 + }, + { + "epoch": 0.058423205796050884, + "grad_norm": 0.2122262567281723, + "learning_rate": 2.9210363972856262e-05, + "loss": 3.4444, + "step": 1894 + }, + { + "epoch": 0.0584540522616243, + "grad_norm": 0.20449016988277435, + "learning_rate": 2.9225786551511415e-05, + "loss": 3.4656, + "step": 1895 + }, + { + "epoch": 0.05848489872719771, + "grad_norm": 0.22559310495853424, + "learning_rate": 2.9241209130166563e-05, + "loss": 3.4128, + "step": 1896 + }, + { + "epoch": 0.05851574519277113, + "grad_norm": 0.1998181939125061, + "learning_rate": 2.925663170882172e-05, + "loss": 3.4127, + "step": 1897 + }, + { + "epoch": 0.05854659165834455, + "grad_norm": 0.22212405502796173, + "learning_rate": 2.9272054287476868e-05, + "loss": 3.4129, + "step": 1898 + }, + { + "epoch": 0.05857743812391796, + "grad_norm": 0.21330450475215912, + "learning_rate": 2.928747686613202e-05, + "loss": 3.439, + "step": 1899 + }, + { + "epoch": 0.05860828458949138, + "grad_norm": 0.21083107590675354, + "learning_rate": 2.930289944478717e-05, + "loss": 3.4114, + "step": 1900 + }, + { + "epoch": 0.0586391310550648, + "grad_norm": 0.2244800180196762, + "learning_rate": 2.9318322023442318e-05, + "loss": 3.4261, + "step": 1901 + }, + { + "epoch": 0.05866997752063821, + "grad_norm": 0.20370268821716309, + "learning_rate": 2.9333744602097474e-05, + "loss": 3.4117, + "step": 1902 + }, + { + "epoch": 0.05870082398621163, + "grad_norm": 0.2192079722881317, + "learning_rate": 2.9349167180752623e-05, + "loss": 3.433, + "step": 1903 + }, + { + "epoch": 0.05873167045178505, + "grad_norm": 0.21812570095062256, + "learning_rate": 2.9364589759407775e-05, + "loss": 3.441, + "step": 1904 + }, + { + "epoch": 0.05876251691735846, + "grad_norm": 0.20656733214855194, + "learning_rate": 2.9380012338062924e-05, + "loss": 3.4207, + "step": 1905 + }, + { + "epoch": 0.05879336338293188, + "grad_norm": 0.2946585416793823, + "learning_rate": 2.939543491671808e-05, + "loss": 3.3967, + "step": 1906 + }, + { + "epoch": 0.058824209848505296, + "grad_norm": 0.21053946018218994, + "learning_rate": 2.941085749537323e-05, + "loss": 3.4444, + "step": 1907 + }, + { + "epoch": 0.05885505631407871, + "grad_norm": 0.223206028342247, + "learning_rate": 2.9426280074028377e-05, + "loss": 3.4462, + "step": 1908 + }, + { + "epoch": 0.05888590277965213, + "grad_norm": 0.24549809098243713, + "learning_rate": 2.944170265268353e-05, + "loss": 3.4261, + "step": 1909 + }, + { + "epoch": 0.058916749245225546, + "grad_norm": 0.2082710862159729, + "learning_rate": 2.9457125231338682e-05, + "loss": 3.411, + "step": 1910 + }, + { + "epoch": 0.05894759571079896, + "grad_norm": 0.2191745787858963, + "learning_rate": 2.9472547809993834e-05, + "loss": 3.3978, + "step": 1911 + }, + { + "epoch": 0.05897844217637238, + "grad_norm": 0.3993918299674988, + "learning_rate": 2.9487970388648983e-05, + "loss": 3.4215, + "step": 1912 + }, + { + "epoch": 0.059009288641945795, + "grad_norm": 0.21324734389781952, + "learning_rate": 2.9503392967304132e-05, + "loss": 3.4015, + "step": 1913 + }, + { + "epoch": 0.05904013510751921, + "grad_norm": 0.21376772224903107, + "learning_rate": 2.9518815545959288e-05, + "loss": 3.4243, + "step": 1914 + }, + { + "epoch": 0.05907098157309263, + "grad_norm": 0.21787624061107635, + "learning_rate": 2.9534238124614437e-05, + "loss": 3.4439, + "step": 1915 + }, + { + "epoch": 0.059101828038666045, + "grad_norm": 0.21584294736385345, + "learning_rate": 2.954966070326959e-05, + "loss": 3.4717, + "step": 1916 + }, + { + "epoch": 0.05913267450423946, + "grad_norm": 0.20482271909713745, + "learning_rate": 2.9565083281924738e-05, + "loss": 3.4445, + "step": 1917 + }, + { + "epoch": 0.05916352096981288, + "grad_norm": 0.2685428261756897, + "learning_rate": 2.9580505860579887e-05, + "loss": 3.4424, + "step": 1918 + }, + { + "epoch": 0.059194367435386294, + "grad_norm": 0.21901677548885345, + "learning_rate": 2.9595928439235043e-05, + "loss": 3.4221, + "step": 1919 + }, + { + "epoch": 0.05922521390095971, + "grad_norm": 0.21212553977966309, + "learning_rate": 2.9611351017890195e-05, + "loss": 3.4112, + "step": 1920 + }, + { + "epoch": 0.05925606036653313, + "grad_norm": 0.20657961070537567, + "learning_rate": 2.9626773596545344e-05, + "loss": 3.4191, + "step": 1921 + }, + { + "epoch": 0.059286906832106544, + "grad_norm": 0.2073073387145996, + "learning_rate": 2.9642196175200493e-05, + "loss": 3.4369, + "step": 1922 + }, + { + "epoch": 0.05931775329767996, + "grad_norm": 0.22598090767860413, + "learning_rate": 2.9657618753855648e-05, + "loss": 3.4618, + "step": 1923 + }, + { + "epoch": 0.05934859976325338, + "grad_norm": 0.22057843208312988, + "learning_rate": 2.9673041332510797e-05, + "loss": 3.4461, + "step": 1924 + }, + { + "epoch": 0.05937944622882679, + "grad_norm": 0.21379521489143372, + "learning_rate": 2.968846391116595e-05, + "loss": 3.4517, + "step": 1925 + }, + { + "epoch": 0.05941029269440021, + "grad_norm": 0.21776433289051056, + "learning_rate": 2.97038864898211e-05, + "loss": 3.3986, + "step": 1926 + }, + { + "epoch": 0.05944113915997363, + "grad_norm": 0.2149236500263214, + "learning_rate": 2.9719309068476254e-05, + "loss": 3.4269, + "step": 1927 + }, + { + "epoch": 0.05947198562554704, + "grad_norm": 0.2113804966211319, + "learning_rate": 2.9734731647131403e-05, + "loss": 3.4119, + "step": 1928 + }, + { + "epoch": 0.05950283209112046, + "grad_norm": 0.21773701906204224, + "learning_rate": 2.9750154225786552e-05, + "loss": 3.4466, + "step": 1929 + }, + { + "epoch": 0.05953367855669388, + "grad_norm": 0.21750815212726593, + "learning_rate": 2.9765576804441704e-05, + "loss": 3.4163, + "step": 1930 + }, + { + "epoch": 0.05956452502226729, + "grad_norm": 0.21445490419864655, + "learning_rate": 2.9780999383096853e-05, + "loss": 3.4621, + "step": 1931 + }, + { + "epoch": 0.059595371487840706, + "grad_norm": 0.2431734949350357, + "learning_rate": 2.979642196175201e-05, + "loss": 3.3962, + "step": 1932 + }, + { + "epoch": 0.05962621795341413, + "grad_norm": 0.19495953619480133, + "learning_rate": 2.9811844540407158e-05, + "loss": 3.3944, + "step": 1933 + }, + { + "epoch": 0.05965706441898754, + "grad_norm": 0.22002215683460236, + "learning_rate": 2.9827267119062307e-05, + "loss": 3.4232, + "step": 1934 + }, + { + "epoch": 0.059687910884560956, + "grad_norm": 0.22005951404571533, + "learning_rate": 2.984268969771746e-05, + "loss": 3.4485, + "step": 1935 + }, + { + "epoch": 0.05971875735013438, + "grad_norm": 0.20731881260871887, + "learning_rate": 2.9858112276372615e-05, + "loss": 3.4204, + "step": 1936 + }, + { + "epoch": 0.05974960381570779, + "grad_norm": 0.20859135687351227, + "learning_rate": 2.9873534855027763e-05, + "loss": 3.481, + "step": 1937 + }, + { + "epoch": 0.059780450281281206, + "grad_norm": 0.2111263871192932, + "learning_rate": 2.9888957433682912e-05, + "loss": 3.4359, + "step": 1938 + }, + { + "epoch": 0.05981129674685463, + "grad_norm": 0.20675989985466003, + "learning_rate": 2.990438001233806e-05, + "loss": 3.4528, + "step": 1939 + }, + { + "epoch": 0.05984214321242804, + "grad_norm": 0.20195706188678741, + "learning_rate": 2.9919802590993217e-05, + "loss": 3.4082, + "step": 1940 + }, + { + "epoch": 0.059872989678001455, + "grad_norm": 0.20345716178417206, + "learning_rate": 2.993522516964837e-05, + "loss": 3.432, + "step": 1941 + }, + { + "epoch": 0.059903836143574876, + "grad_norm": 0.21025702357292175, + "learning_rate": 2.9950647748303518e-05, + "loss": 3.4366, + "step": 1942 + }, + { + "epoch": 0.05993468260914829, + "grad_norm": 0.21388451755046844, + "learning_rate": 2.9966070326958667e-05, + "loss": 3.4076, + "step": 1943 + }, + { + "epoch": 0.059965529074721705, + "grad_norm": 0.20194201171398163, + "learning_rate": 2.9981492905613823e-05, + "loss": 3.445, + "step": 1944 + }, + { + "epoch": 0.059996375540295126, + "grad_norm": 0.2132629007101059, + "learning_rate": 2.999691548426897e-05, + "loss": 3.4298, + "step": 1945 + }, + { + "epoch": 0.06002722200586854, + "grad_norm": 0.21632251143455505, + "learning_rate": 3.0012338062924124e-05, + "loss": 3.4363, + "step": 1946 + }, + { + "epoch": 0.060058068471441954, + "grad_norm": 0.2120971977710724, + "learning_rate": 3.0027760641579273e-05, + "loss": 3.4493, + "step": 1947 + }, + { + "epoch": 0.060088914937015375, + "grad_norm": 0.20364229381084442, + "learning_rate": 3.0043183220234422e-05, + "loss": 3.4199, + "step": 1948 + }, + { + "epoch": 0.06011976140258879, + "grad_norm": 0.2095443457365036, + "learning_rate": 3.0058605798889577e-05, + "loss": 3.4408, + "step": 1949 + }, + { + "epoch": 0.060150607868162204, + "grad_norm": 0.20152756571769714, + "learning_rate": 3.0074028377544726e-05, + "loss": 3.4596, + "step": 1950 + }, + { + "epoch": 0.060181454333735625, + "grad_norm": 0.20441056787967682, + "learning_rate": 3.008945095619988e-05, + "loss": 3.4296, + "step": 1951 + }, + { + "epoch": 0.06021230079930904, + "grad_norm": 0.1979372650384903, + "learning_rate": 3.0104873534855028e-05, + "loss": 3.4408, + "step": 1952 + }, + { + "epoch": 0.06024314726488245, + "grad_norm": 0.21745909750461578, + "learning_rate": 3.0120296113510183e-05, + "loss": 3.4194, + "step": 1953 + }, + { + "epoch": 0.060273993730455874, + "grad_norm": 0.2214238941669464, + "learning_rate": 3.0135718692165332e-05, + "loss": 3.4712, + "step": 1954 + }, + { + "epoch": 0.06030484019602929, + "grad_norm": 0.21675528585910797, + "learning_rate": 3.015114127082048e-05, + "loss": 3.4065, + "step": 1955 + }, + { + "epoch": 0.0603356866616027, + "grad_norm": 0.19783656299114227, + "learning_rate": 3.0166563849475633e-05, + "loss": 3.4338, + "step": 1956 + }, + { + "epoch": 0.060366533127176124, + "grad_norm": 0.2173328995704651, + "learning_rate": 3.018198642813079e-05, + "loss": 3.4245, + "step": 1957 + }, + { + "epoch": 0.06039737959274954, + "grad_norm": 0.2074468582868576, + "learning_rate": 3.0197409006785938e-05, + "loss": 3.4472, + "step": 1958 + }, + { + "epoch": 0.06042822605832295, + "grad_norm": 0.19781413674354553, + "learning_rate": 3.0212831585441087e-05, + "loss": 3.4268, + "step": 1959 + }, + { + "epoch": 0.06045907252389637, + "grad_norm": 0.20908915996551514, + "learning_rate": 3.0228254164096236e-05, + "loss": 3.4348, + "step": 1960 + }, + { + "epoch": 0.06048991898946979, + "grad_norm": 0.20833493769168854, + "learning_rate": 3.0243676742751388e-05, + "loss": 3.4397, + "step": 1961 + }, + { + "epoch": 0.0605207654550432, + "grad_norm": 0.19716520607471466, + "learning_rate": 3.0259099321406544e-05, + "loss": 3.391, + "step": 1962 + }, + { + "epoch": 0.06055161192061662, + "grad_norm": 0.21940232813358307, + "learning_rate": 3.0274521900061693e-05, + "loss": 3.3871, + "step": 1963 + }, + { + "epoch": 0.06058245838619004, + "grad_norm": 0.21686503291130066, + "learning_rate": 3.028994447871684e-05, + "loss": 3.4271, + "step": 1964 + }, + { + "epoch": 0.06061330485176345, + "grad_norm": 0.20457163453102112, + "learning_rate": 3.030536705737199e-05, + "loss": 3.4148, + "step": 1965 + }, + { + "epoch": 0.06064415131733687, + "grad_norm": 0.21155069768428802, + "learning_rate": 3.0320789636027146e-05, + "loss": 3.3834, + "step": 1966 + }, + { + "epoch": 0.06067499778291029, + "grad_norm": 0.5501000285148621, + "learning_rate": 3.03362122146823e-05, + "loss": 3.4372, + "step": 1967 + }, + { + "epoch": 0.0607058442484837, + "grad_norm": 0.2190013974905014, + "learning_rate": 3.0351634793337447e-05, + "loss": 3.4559, + "step": 1968 + }, + { + "epoch": 0.06073669071405712, + "grad_norm": 0.20739711821079254, + "learning_rate": 3.0367057371992596e-05, + "loss": 3.4573, + "step": 1969 + }, + { + "epoch": 0.060767537179630536, + "grad_norm": 0.2387748658657074, + "learning_rate": 3.0382479950647752e-05, + "loss": 3.4103, + "step": 1970 + }, + { + "epoch": 0.06079838364520395, + "grad_norm": 0.21647633612155914, + "learning_rate": 3.03979025293029e-05, + "loss": 3.4529, + "step": 1971 + }, + { + "epoch": 0.06082923011077737, + "grad_norm": 0.23701339960098267, + "learning_rate": 3.0413325107958053e-05, + "loss": 3.4374, + "step": 1972 + }, + { + "epoch": 0.060860076576350786, + "grad_norm": 0.2134370505809784, + "learning_rate": 3.0428747686613202e-05, + "loss": 3.3817, + "step": 1973 + }, + { + "epoch": 0.0608909230419242, + "grad_norm": 0.22932444512844086, + "learning_rate": 3.0444170265268358e-05, + "loss": 3.4192, + "step": 1974 + }, + { + "epoch": 0.06092176950749762, + "grad_norm": 0.2218402624130249, + "learning_rate": 3.0459592843923507e-05, + "loss": 3.4101, + "step": 1975 + }, + { + "epoch": 0.060952615973071035, + "grad_norm": 0.20628316700458527, + "learning_rate": 3.0475015422578655e-05, + "loss": 3.4436, + "step": 1976 + }, + { + "epoch": 0.06098346243864445, + "grad_norm": 0.22715243697166443, + "learning_rate": 3.0490438001233808e-05, + "loss": 3.4111, + "step": 1977 + }, + { + "epoch": 0.06101430890421787, + "grad_norm": 0.20170097053050995, + "learning_rate": 3.0505860579888957e-05, + "loss": 3.4174, + "step": 1978 + }, + { + "epoch": 0.061045155369791285, + "grad_norm": 0.2229090929031372, + "learning_rate": 3.052128315854411e-05, + "loss": 3.3318, + "step": 1979 + }, + { + "epoch": 0.0610760018353647, + "grad_norm": 0.2197667360305786, + "learning_rate": 3.0536705737199265e-05, + "loss": 3.4315, + "step": 1980 + }, + { + "epoch": 0.06110684830093812, + "grad_norm": 0.20435510575771332, + "learning_rate": 3.055212831585441e-05, + "loss": 3.4349, + "step": 1981 + }, + { + "epoch": 0.061137694766511534, + "grad_norm": 0.21378149092197418, + "learning_rate": 3.056755089450956e-05, + "loss": 3.442, + "step": 1982 + }, + { + "epoch": 0.06116854123208495, + "grad_norm": 0.20936156809329987, + "learning_rate": 3.0582973473164715e-05, + "loss": 3.4449, + "step": 1983 + }, + { + "epoch": 0.06119938769765837, + "grad_norm": 0.20702391862869263, + "learning_rate": 3.059839605181987e-05, + "loss": 3.4146, + "step": 1984 + }, + { + "epoch": 0.061230234163231784, + "grad_norm": 0.21421600878238678, + "learning_rate": 3.061381863047502e-05, + "loss": 3.4131, + "step": 1985 + }, + { + "epoch": 0.0612610806288052, + "grad_norm": 0.21546876430511475, + "learning_rate": 3.0629241209130165e-05, + "loss": 3.4152, + "step": 1986 + }, + { + "epoch": 0.06129192709437862, + "grad_norm": 0.2126457393169403, + "learning_rate": 3.0644663787785324e-05, + "loss": 3.4376, + "step": 1987 + }, + { + "epoch": 0.06132277355995203, + "grad_norm": 0.21183928847312927, + "learning_rate": 3.066008636644047e-05, + "loss": 3.4032, + "step": 1988 + }, + { + "epoch": 0.06135362002552545, + "grad_norm": 0.21436192095279694, + "learning_rate": 3.067550894509562e-05, + "loss": 3.3957, + "step": 1989 + }, + { + "epoch": 0.06138446649109887, + "grad_norm": 0.21033377945423126, + "learning_rate": 3.0690931523750774e-05, + "loss": 3.3983, + "step": 1990 + }, + { + "epoch": 0.06141531295667228, + "grad_norm": 0.2270636409521103, + "learning_rate": 3.0706354102405926e-05, + "loss": 3.4128, + "step": 1991 + }, + { + "epoch": 0.0614461594222457, + "grad_norm": 0.21009035408496857, + "learning_rate": 3.072177668106108e-05, + "loss": 3.3965, + "step": 1992 + }, + { + "epoch": 0.06147700588781912, + "grad_norm": 0.2125236839056015, + "learning_rate": 3.0737199259716224e-05, + "loss": 3.4148, + "step": 1993 + }, + { + "epoch": 0.06150785235339253, + "grad_norm": 0.20463745296001434, + "learning_rate": 3.0752621838371376e-05, + "loss": 3.4564, + "step": 1994 + }, + { + "epoch": 0.061538698818965946, + "grad_norm": 0.20858001708984375, + "learning_rate": 3.076804441702653e-05, + "loss": 3.392, + "step": 1995 + }, + { + "epoch": 0.06156954528453937, + "grad_norm": 0.21894840896129608, + "learning_rate": 3.078346699568168e-05, + "loss": 3.4005, + "step": 1996 + }, + { + "epoch": 0.06160039175011278, + "grad_norm": 0.20636305212974548, + "learning_rate": 3.079888957433683e-05, + "loss": 3.4044, + "step": 1997 + }, + { + "epoch": 0.061631238215686196, + "grad_norm": 0.19780008494853973, + "learning_rate": 3.081431215299198e-05, + "loss": 3.4719, + "step": 1998 + }, + { + "epoch": 0.06166208468125962, + "grad_norm": 0.20638731122016907, + "learning_rate": 3.082973473164713e-05, + "loss": 3.4227, + "step": 1999 + }, + { + "epoch": 0.06169293114683303, + "grad_norm": 0.2198038101196289, + "learning_rate": 3.0845157310302283e-05, + "loss": 3.392, + "step": 2000 + }, + { + "epoch": 0.061723777612406446, + "grad_norm": 0.20377279818058014, + "learning_rate": 3.0860579888957436e-05, + "loss": 3.4518, + "step": 2001 + }, + { + "epoch": 0.06175462407797987, + "grad_norm": 0.20545874536037445, + "learning_rate": 3.087600246761259e-05, + "loss": 3.4413, + "step": 2002 + }, + { + "epoch": 0.06178547054355328, + "grad_norm": 0.20359331369400024, + "learning_rate": 3.0891425046267734e-05, + "loss": 3.4127, + "step": 2003 + }, + { + "epoch": 0.061816317009126695, + "grad_norm": 0.2104749232530594, + "learning_rate": 3.090684762492289e-05, + "loss": 3.4306, + "step": 2004 + }, + { + "epoch": 0.061847163474700116, + "grad_norm": 0.20123988389968872, + "learning_rate": 3.092227020357804e-05, + "loss": 3.4276, + "step": 2005 + }, + { + "epoch": 0.06187800994027353, + "grad_norm": 0.20600078999996185, + "learning_rate": 3.093769278223319e-05, + "loss": 3.3818, + "step": 2006 + }, + { + "epoch": 0.061908856405846945, + "grad_norm": 0.20709122717380524, + "learning_rate": 3.095311536088834e-05, + "loss": 3.4301, + "step": 2007 + }, + { + "epoch": 0.061939702871420366, + "grad_norm": 0.2167259156703949, + "learning_rate": 3.096853793954349e-05, + "loss": 3.4105, + "step": 2008 + }, + { + "epoch": 0.06197054933699378, + "grad_norm": 0.22504229843616486, + "learning_rate": 3.098396051819865e-05, + "loss": 3.4625, + "step": 2009 + }, + { + "epoch": 0.062001395802567194, + "grad_norm": 0.2237553894519806, + "learning_rate": 3.099938309685379e-05, + "loss": 3.4369, + "step": 2010 + }, + { + "epoch": 0.062032242268140615, + "grad_norm": 0.2153075486421585, + "learning_rate": 3.1014805675508945e-05, + "loss": 3.444, + "step": 2011 + }, + { + "epoch": 0.06206308873371403, + "grad_norm": 0.20248685777187347, + "learning_rate": 3.10302282541641e-05, + "loss": 3.427, + "step": 2012 + }, + { + "epoch": 0.062093935199287444, + "grad_norm": 0.20900161564350128, + "learning_rate": 3.104565083281925e-05, + "loss": 3.4459, + "step": 2013 + }, + { + "epoch": 0.062124781664860865, + "grad_norm": 0.1952146291732788, + "learning_rate": 3.10610734114744e-05, + "loss": 3.4088, + "step": 2014 + }, + { + "epoch": 0.06215562813043428, + "grad_norm": 0.2027866244316101, + "learning_rate": 3.107649599012955e-05, + "loss": 3.4158, + "step": 2015 + }, + { + "epoch": 0.06218647459600769, + "grad_norm": 0.20888455212116241, + "learning_rate": 3.10919185687847e-05, + "loss": 3.4075, + "step": 2016 + }, + { + "epoch": 0.062217321061581114, + "grad_norm": 0.8513734340667725, + "learning_rate": 3.110734114743986e-05, + "loss": 3.455, + "step": 2017 + }, + { + "epoch": 0.06224816752715453, + "grad_norm": 0.21954794228076935, + "learning_rate": 3.1122763726095004e-05, + "loss": 3.4398, + "step": 2018 + }, + { + "epoch": 0.06227901399272794, + "grad_norm": 0.2275015264749527, + "learning_rate": 3.113818630475016e-05, + "loss": 3.4391, + "step": 2019 + }, + { + "epoch": 0.062309860458301364, + "grad_norm": 0.20472222566604614, + "learning_rate": 3.11536088834053e-05, + "loss": 3.4054, + "step": 2020 + }, + { + "epoch": 0.06234070692387478, + "grad_norm": 0.2231948971748352, + "learning_rate": 3.116903146206046e-05, + "loss": 3.3989, + "step": 2021 + }, + { + "epoch": 0.06237155338944819, + "grad_norm": 0.2525116801261902, + "learning_rate": 3.1184454040715614e-05, + "loss": 3.4135, + "step": 2022 + }, + { + "epoch": 0.06240239985502161, + "grad_norm": 0.2393421232700348, + "learning_rate": 3.119987661937076e-05, + "loss": 3.3966, + "step": 2023 + }, + { + "epoch": 0.06243324632059503, + "grad_norm": 0.23293356597423553, + "learning_rate": 3.121529919802591e-05, + "loss": 3.4243, + "step": 2024 + }, + { + "epoch": 0.06246409278616844, + "grad_norm": 0.22581049799919128, + "learning_rate": 3.123072177668106e-05, + "loss": 3.4418, + "step": 2025 + }, + { + "epoch": 0.06249493925174186, + "grad_norm": 0.21449503302574158, + "learning_rate": 3.1246144355336216e-05, + "loss": 3.4708, + "step": 2026 + }, + { + "epoch": 0.06252578571731528, + "grad_norm": 0.24455158412456512, + "learning_rate": 3.126156693399137e-05, + "loss": 3.4221, + "step": 2027 + }, + { + "epoch": 0.06255663218288869, + "grad_norm": 0.20686639845371246, + "learning_rate": 3.1276989512646514e-05, + "loss": 3.4509, + "step": 2028 + }, + { + "epoch": 0.0625874786484621, + "grad_norm": 0.2410569041967392, + "learning_rate": 3.1292412091301666e-05, + "loss": 3.394, + "step": 2029 + }, + { + "epoch": 0.06261832511403553, + "grad_norm": 0.2171521931886673, + "learning_rate": 3.130783466995682e-05, + "loss": 3.3935, + "step": 2030 + }, + { + "epoch": 0.06264917157960895, + "grad_norm": 0.24406246840953827, + "learning_rate": 3.132325724861197e-05, + "loss": 3.4022, + "step": 2031 + }, + { + "epoch": 0.06268001804518236, + "grad_norm": 0.21540981531143188, + "learning_rate": 3.133867982726712e-05, + "loss": 3.4393, + "step": 2032 + }, + { + "epoch": 0.06271086451075578, + "grad_norm": 0.21342718601226807, + "learning_rate": 3.135410240592227e-05, + "loss": 3.4003, + "step": 2033 + }, + { + "epoch": 0.06274171097632919, + "grad_norm": 0.21459811925888062, + "learning_rate": 3.136952498457743e-05, + "loss": 3.4124, + "step": 2034 + }, + { + "epoch": 0.0627725574419026, + "grad_norm": 0.2167055904865265, + "learning_rate": 3.138494756323257e-05, + "loss": 3.3966, + "step": 2035 + }, + { + "epoch": 0.06280340390747603, + "grad_norm": 0.22842122614383698, + "learning_rate": 3.1400370141887725e-05, + "loss": 3.4596, + "step": 2036 + }, + { + "epoch": 0.06283425037304945, + "grad_norm": 0.22313036024570465, + "learning_rate": 3.141579272054288e-05, + "loss": 3.4199, + "step": 2037 + }, + { + "epoch": 0.06286509683862286, + "grad_norm": 0.21979369223117828, + "learning_rate": 3.143121529919802e-05, + "loss": 3.4374, + "step": 2038 + }, + { + "epoch": 0.06289594330419628, + "grad_norm": 0.22079747915267944, + "learning_rate": 3.144663787785318e-05, + "loss": 3.3565, + "step": 2039 + }, + { + "epoch": 0.06292678976976969, + "grad_norm": 0.22101916372776031, + "learning_rate": 3.146206045650833e-05, + "loss": 3.425, + "step": 2040 + }, + { + "epoch": 0.0629576362353431, + "grad_norm": 0.2227390706539154, + "learning_rate": 3.147748303516348e-05, + "loss": 3.4034, + "step": 2041 + }, + { + "epoch": 0.06298848270091653, + "grad_norm": 0.22670558094978333, + "learning_rate": 3.149290561381863e-05, + "loss": 3.4181, + "step": 2042 + }, + { + "epoch": 0.06301932916648995, + "grad_norm": 0.20350992679595947, + "learning_rate": 3.1508328192473785e-05, + "loss": 3.4301, + "step": 2043 + }, + { + "epoch": 0.06305017563206336, + "grad_norm": 0.227951318025589, + "learning_rate": 3.152375077112894e-05, + "loss": 3.372, + "step": 2044 + }, + { + "epoch": 0.06308102209763677, + "grad_norm": 0.21859876811504364, + "learning_rate": 3.153917334978408e-05, + "loss": 3.4141, + "step": 2045 + }, + { + "epoch": 0.06311186856321019, + "grad_norm": 0.22683504223823547, + "learning_rate": 3.1554595928439235e-05, + "loss": 3.4072, + "step": 2046 + }, + { + "epoch": 0.0631427150287836, + "grad_norm": 0.1992632895708084, + "learning_rate": 3.157001850709439e-05, + "loss": 3.424, + "step": 2047 + }, + { + "epoch": 0.06317356149435703, + "grad_norm": 0.2015453428030014, + "learning_rate": 3.158544108574954e-05, + "loss": 3.4295, + "step": 2048 + }, + { + "epoch": 0.06320440795993044, + "grad_norm": 0.19983847439289093, + "learning_rate": 3.160086366440469e-05, + "loss": 3.4306, + "step": 2049 + }, + { + "epoch": 0.06323525442550386, + "grad_norm": 0.19916681945323944, + "learning_rate": 3.161628624305984e-05, + "loss": 3.442, + "step": 2050 + }, + { + "epoch": 0.06326610089107727, + "grad_norm": 0.19918814301490784, + "learning_rate": 3.1631708821714996e-05, + "loss": 3.413, + "step": 2051 + }, + { + "epoch": 0.06329694735665069, + "grad_norm": 0.20947112143039703, + "learning_rate": 3.164713140037014e-05, + "loss": 3.4148, + "step": 2052 + }, + { + "epoch": 0.0633277938222241, + "grad_norm": 0.20504380762577057, + "learning_rate": 3.1662553979025294e-05, + "loss": 3.4075, + "step": 2053 + }, + { + "epoch": 0.06335864028779753, + "grad_norm": 0.19851969182491302, + "learning_rate": 3.1677976557680446e-05, + "loss": 3.4495, + "step": 2054 + }, + { + "epoch": 0.06338948675337094, + "grad_norm": 0.19776245951652527, + "learning_rate": 3.169339913633559e-05, + "loss": 3.4196, + "step": 2055 + }, + { + "epoch": 0.06342033321894436, + "grad_norm": 0.20685745775699615, + "learning_rate": 3.170882171499075e-05, + "loss": 3.4195, + "step": 2056 + }, + { + "epoch": 0.06345117968451777, + "grad_norm": 0.20347362756729126, + "learning_rate": 3.1724244293645896e-05, + "loss": 3.4221, + "step": 2057 + }, + { + "epoch": 0.06348202615009119, + "grad_norm": 0.2071980983018875, + "learning_rate": 3.173966687230105e-05, + "loss": 3.4008, + "step": 2058 + }, + { + "epoch": 0.0635128726156646, + "grad_norm": 0.20847073197364807, + "learning_rate": 3.17550894509562e-05, + "loss": 3.4098, + "step": 2059 + }, + { + "epoch": 0.06354371908123803, + "grad_norm": 0.2059977650642395, + "learning_rate": 3.177051202961135e-05, + "loss": 3.3812, + "step": 2060 + }, + { + "epoch": 0.06357456554681144, + "grad_norm": 0.2109186053276062, + "learning_rate": 3.1785934608266506e-05, + "loss": 3.4002, + "step": 2061 + }, + { + "epoch": 0.06360541201238486, + "grad_norm": 0.20708821713924408, + "learning_rate": 3.180135718692165e-05, + "loss": 3.3997, + "step": 2062 + }, + { + "epoch": 0.06363625847795827, + "grad_norm": 0.2191806137561798, + "learning_rate": 3.1816779765576803e-05, + "loss": 3.3925, + "step": 2063 + }, + { + "epoch": 0.06366710494353169, + "grad_norm": 0.202923983335495, + "learning_rate": 3.183220234423196e-05, + "loss": 3.3923, + "step": 2064 + }, + { + "epoch": 0.0636979514091051, + "grad_norm": 0.21116903424263, + "learning_rate": 3.184762492288711e-05, + "loss": 3.3949, + "step": 2065 + }, + { + "epoch": 0.06372879787467853, + "grad_norm": 0.23237818479537964, + "learning_rate": 3.186304750154226e-05, + "loss": 3.4264, + "step": 2066 + }, + { + "epoch": 0.06375964434025194, + "grad_norm": 0.20173528790473938, + "learning_rate": 3.1878470080197406e-05, + "loss": 3.417, + "step": 2067 + }, + { + "epoch": 0.06379049080582536, + "grad_norm": 0.22888559103012085, + "learning_rate": 3.1893892658852565e-05, + "loss": 3.4399, + "step": 2068 + }, + { + "epoch": 0.06382133727139877, + "grad_norm": 0.2120388150215149, + "learning_rate": 3.190931523750772e-05, + "loss": 3.4063, + "step": 2069 + }, + { + "epoch": 0.06385218373697218, + "grad_norm": 0.22129811346530914, + "learning_rate": 3.192473781616286e-05, + "loss": 3.3995, + "step": 2070 + }, + { + "epoch": 0.0638830302025456, + "grad_norm": 0.2097272127866745, + "learning_rate": 3.1940160394818015e-05, + "loss": 3.4242, + "step": 2071 + }, + { + "epoch": 0.06391387666811903, + "grad_norm": 0.21853765845298767, + "learning_rate": 3.195558297347316e-05, + "loss": 3.4255, + "step": 2072 + }, + { + "epoch": 0.06394472313369244, + "grad_norm": 0.2274811863899231, + "learning_rate": 3.197100555212832e-05, + "loss": 3.4049, + "step": 2073 + }, + { + "epoch": 0.06397556959926586, + "grad_norm": 0.20414668321609497, + "learning_rate": 3.198642813078347e-05, + "loss": 3.3726, + "step": 2074 + }, + { + "epoch": 0.06400641606483927, + "grad_norm": 0.2223559468984604, + "learning_rate": 3.200185070943862e-05, + "loss": 3.4212, + "step": 2075 + }, + { + "epoch": 0.06403726253041268, + "grad_norm": 0.19728447496891022, + "learning_rate": 3.201727328809377e-05, + "loss": 3.4336, + "step": 2076 + }, + { + "epoch": 0.0640681089959861, + "grad_norm": 0.21955351531505585, + "learning_rate": 3.203269586674892e-05, + "loss": 3.4412, + "step": 2077 + }, + { + "epoch": 0.06409895546155953, + "grad_norm": 0.2026480883359909, + "learning_rate": 3.2048118445404074e-05, + "loss": 3.392, + "step": 2078 + }, + { + "epoch": 0.06412980192713294, + "grad_norm": 0.20542500913143158, + "learning_rate": 3.2063541024059227e-05, + "loss": 3.4034, + "step": 2079 + }, + { + "epoch": 0.06416064839270635, + "grad_norm": 0.20528581738471985, + "learning_rate": 3.207896360271437e-05, + "loss": 3.4295, + "step": 2080 + }, + { + "epoch": 0.06419149485827977, + "grad_norm": 0.2028125375509262, + "learning_rate": 3.209438618136953e-05, + "loss": 3.4269, + "step": 2081 + }, + { + "epoch": 0.06422234132385318, + "grad_norm": 0.20300909876823425, + "learning_rate": 3.210980876002468e-05, + "loss": 3.3888, + "step": 2082 + }, + { + "epoch": 0.0642531877894266, + "grad_norm": 0.20136982202529907, + "learning_rate": 3.212523133867983e-05, + "loss": 3.3794, + "step": 2083 + }, + { + "epoch": 0.06428403425500002, + "grad_norm": 0.2015087902545929, + "learning_rate": 3.214065391733498e-05, + "loss": 3.4053, + "step": 2084 + }, + { + "epoch": 0.06431488072057344, + "grad_norm": 0.1986033171415329, + "learning_rate": 3.215607649599013e-05, + "loss": 3.4204, + "step": 2085 + }, + { + "epoch": 0.06434572718614685, + "grad_norm": 0.2077392339706421, + "learning_rate": 3.2171499074645286e-05, + "loss": 3.3677, + "step": 2086 + }, + { + "epoch": 0.06437657365172027, + "grad_norm": 0.20460297167301178, + "learning_rate": 3.218692165330043e-05, + "loss": 3.3988, + "step": 2087 + }, + { + "epoch": 0.06440742011729368, + "grad_norm": 0.1942853033542633, + "learning_rate": 3.2202344231955584e-05, + "loss": 3.4154, + "step": 2088 + }, + { + "epoch": 0.0644382665828671, + "grad_norm": 0.19786542654037476, + "learning_rate": 3.2217766810610736e-05, + "loss": 3.4115, + "step": 2089 + }, + { + "epoch": 0.06446911304844052, + "grad_norm": 0.1927604228258133, + "learning_rate": 3.223318938926589e-05, + "loss": 3.4411, + "step": 2090 + }, + { + "epoch": 0.06449995951401394, + "grad_norm": 0.19770218431949615, + "learning_rate": 3.224861196792104e-05, + "loss": 3.3715, + "step": 2091 + }, + { + "epoch": 0.06453080597958735, + "grad_norm": 0.21352867782115936, + "learning_rate": 3.2264034546576186e-05, + "loss": 3.4088, + "step": 2092 + }, + { + "epoch": 0.06456165244516077, + "grad_norm": 0.20318326354026794, + "learning_rate": 3.227945712523134e-05, + "loss": 3.3952, + "step": 2093 + }, + { + "epoch": 0.06459249891073418, + "grad_norm": 0.2053423970937729, + "learning_rate": 3.229487970388649e-05, + "loss": 3.4181, + "step": 2094 + }, + { + "epoch": 0.0646233453763076, + "grad_norm": 0.20328623056411743, + "learning_rate": 3.231030228254164e-05, + "loss": 3.3834, + "step": 2095 + }, + { + "epoch": 0.06465419184188102, + "grad_norm": 0.20960350334644318, + "learning_rate": 3.2325724861196795e-05, + "loss": 3.4165, + "step": 2096 + }, + { + "epoch": 0.06468503830745444, + "grad_norm": 0.2146020233631134, + "learning_rate": 3.234114743985194e-05, + "loss": 3.4073, + "step": 2097 + }, + { + "epoch": 0.06471588477302785, + "grad_norm": 0.18932585418224335, + "learning_rate": 3.23565700185071e-05, + "loss": 3.393, + "step": 2098 + }, + { + "epoch": 0.06474673123860127, + "grad_norm": 0.21996153891086578, + "learning_rate": 3.2371992597162245e-05, + "loss": 3.4379, + "step": 2099 + }, + { + "epoch": 0.06477757770417468, + "grad_norm": 0.22012515366077423, + "learning_rate": 3.23874151758174e-05, + "loss": 3.4304, + "step": 2100 + }, + { + "epoch": 0.0648084241697481, + "grad_norm": 0.20004871487617493, + "learning_rate": 3.240283775447255e-05, + "loss": 3.4102, + "step": 2101 + }, + { + "epoch": 0.06483927063532152, + "grad_norm": 0.2265087366104126, + "learning_rate": 3.2418260333127695e-05, + "loss": 3.3941, + "step": 2102 + }, + { + "epoch": 0.06487011710089494, + "grad_norm": 0.20094022154808044, + "learning_rate": 3.2433682911782854e-05, + "loss": 3.4072, + "step": 2103 + }, + { + "epoch": 0.06490096356646835, + "grad_norm": 0.22657480835914612, + "learning_rate": 3.2449105490438e-05, + "loss": 3.4099, + "step": 2104 + }, + { + "epoch": 0.06493181003204176, + "grad_norm": 0.2038571536540985, + "learning_rate": 3.246452806909315e-05, + "loss": 3.4293, + "step": 2105 + }, + { + "epoch": 0.06496265649761518, + "grad_norm": 0.23913468420505524, + "learning_rate": 3.2479950647748305e-05, + "loss": 3.3842, + "step": 2106 + }, + { + "epoch": 0.06499350296318859, + "grad_norm": 0.2176472693681717, + "learning_rate": 3.249537322640346e-05, + "loss": 3.3801, + "step": 2107 + }, + { + "epoch": 0.06502434942876202, + "grad_norm": 0.2000254988670349, + "learning_rate": 3.251079580505861e-05, + "loss": 3.3969, + "step": 2108 + }, + { + "epoch": 0.06505519589433544, + "grad_norm": 0.20656849443912506, + "learning_rate": 3.2526218383713755e-05, + "loss": 3.4259, + "step": 2109 + }, + { + "epoch": 0.06508604235990885, + "grad_norm": 0.2046579271554947, + "learning_rate": 3.254164096236891e-05, + "loss": 3.3966, + "step": 2110 + }, + { + "epoch": 0.06511688882548226, + "grad_norm": 0.19389435648918152, + "learning_rate": 3.2557063541024066e-05, + "loss": 3.4079, + "step": 2111 + }, + { + "epoch": 0.06514773529105568, + "grad_norm": 0.23137325048446655, + "learning_rate": 3.257248611967921e-05, + "loss": 3.3688, + "step": 2112 + }, + { + "epoch": 0.06517858175662909, + "grad_norm": 0.1950756311416626, + "learning_rate": 3.2587908698334364e-05, + "loss": 3.3973, + "step": 2113 + }, + { + "epoch": 0.06520942822220252, + "grad_norm": 0.22262491285800934, + "learning_rate": 3.260333127698951e-05, + "loss": 3.4179, + "step": 2114 + }, + { + "epoch": 0.06524027468777593, + "grad_norm": 0.21427585184574127, + "learning_rate": 3.261875385564467e-05, + "loss": 3.374, + "step": 2115 + }, + { + "epoch": 0.06527112115334935, + "grad_norm": 0.21886688470840454, + "learning_rate": 3.263417643429982e-05, + "loss": 3.4007, + "step": 2116 + }, + { + "epoch": 0.06530196761892276, + "grad_norm": 0.2136942744255066, + "learning_rate": 3.2649599012954966e-05, + "loss": 3.4368, + "step": 2117 + }, + { + "epoch": 0.06533281408449618, + "grad_norm": 0.22468233108520508, + "learning_rate": 3.266502159161012e-05, + "loss": 3.4216, + "step": 2118 + }, + { + "epoch": 0.06536366055006959, + "grad_norm": 0.1934579312801361, + "learning_rate": 3.268044417026527e-05, + "loss": 3.3732, + "step": 2119 + }, + { + "epoch": 0.06539450701564302, + "grad_norm": 0.21430541574954987, + "learning_rate": 3.269586674892042e-05, + "loss": 3.3882, + "step": 2120 + }, + { + "epoch": 0.06542535348121643, + "grad_norm": 0.2076197862625122, + "learning_rate": 3.2711289327575575e-05, + "loss": 3.3931, + "step": 2121 + }, + { + "epoch": 0.06545619994678985, + "grad_norm": 0.2125728279352188, + "learning_rate": 3.272671190623072e-05, + "loss": 3.3835, + "step": 2122 + }, + { + "epoch": 0.06548704641236326, + "grad_norm": 0.21765129268169403, + "learning_rate": 3.274213448488587e-05, + "loss": 3.3591, + "step": 2123 + }, + { + "epoch": 0.06551789287793668, + "grad_norm": 0.20928022265434265, + "learning_rate": 3.2757557063541026e-05, + "loss": 3.3757, + "step": 2124 + }, + { + "epoch": 0.06554873934351009, + "grad_norm": 0.21833543479442596, + "learning_rate": 3.277297964219618e-05, + "loss": 3.3826, + "step": 2125 + }, + { + "epoch": 0.06557958580908352, + "grad_norm": 0.20201896131038666, + "learning_rate": 3.278840222085133e-05, + "loss": 3.4156, + "step": 2126 + }, + { + "epoch": 0.06561043227465693, + "grad_norm": 0.2026110589504242, + "learning_rate": 3.2803824799506476e-05, + "loss": 3.4011, + "step": 2127 + }, + { + "epoch": 0.06564127874023035, + "grad_norm": 0.20605751872062683, + "learning_rate": 3.2819247378161635e-05, + "loss": 3.3868, + "step": 2128 + }, + { + "epoch": 0.06567212520580376, + "grad_norm": 0.19882026314735413, + "learning_rate": 3.283466995681678e-05, + "loss": 3.4174, + "step": 2129 + }, + { + "epoch": 0.06570297167137717, + "grad_norm": 0.21622247993946075, + "learning_rate": 3.285009253547193e-05, + "loss": 3.4397, + "step": 2130 + }, + { + "epoch": 0.06573381813695059, + "grad_norm": 0.19859492778778076, + "learning_rate": 3.2865515114127085e-05, + "loss": 3.393, + "step": 2131 + }, + { + "epoch": 0.06576466460252402, + "grad_norm": 0.21501639485359192, + "learning_rate": 3.288093769278223e-05, + "loss": 3.344, + "step": 2132 + }, + { + "epoch": 0.06579551106809743, + "grad_norm": 0.22692202031612396, + "learning_rate": 3.289636027143739e-05, + "loss": 3.4197, + "step": 2133 + }, + { + "epoch": 0.06582635753367085, + "grad_norm": 0.2179371565580368, + "learning_rate": 3.2911782850092535e-05, + "loss": 3.4147, + "step": 2134 + }, + { + "epoch": 0.06585720399924426, + "grad_norm": 0.21961720287799835, + "learning_rate": 3.292720542874769e-05, + "loss": 3.3905, + "step": 2135 + }, + { + "epoch": 0.06588805046481767, + "grad_norm": 0.23526102304458618, + "learning_rate": 3.294262800740284e-05, + "loss": 3.4212, + "step": 2136 + }, + { + "epoch": 0.06591889693039109, + "grad_norm": 0.2077738344669342, + "learning_rate": 3.295805058605799e-05, + "loss": 3.4225, + "step": 2137 + }, + { + "epoch": 0.06594974339596452, + "grad_norm": 0.21232767403125763, + "learning_rate": 3.2973473164713144e-05, + "loss": 3.4009, + "step": 2138 + }, + { + "epoch": 0.06598058986153793, + "grad_norm": 0.20083066821098328, + "learning_rate": 3.298889574336829e-05, + "loss": 3.3827, + "step": 2139 + }, + { + "epoch": 0.06601143632711134, + "grad_norm": 0.2030404508113861, + "learning_rate": 3.300431832202344e-05, + "loss": 3.4137, + "step": 2140 + }, + { + "epoch": 0.06604228279268476, + "grad_norm": 0.20716333389282227, + "learning_rate": 3.3019740900678594e-05, + "loss": 3.4177, + "step": 2141 + }, + { + "epoch": 0.06607312925825817, + "grad_norm": 0.22966015338897705, + "learning_rate": 3.3035163479333747e-05, + "loss": 3.4075, + "step": 2142 + }, + { + "epoch": 0.06610397572383159, + "grad_norm": 0.1993637979030609, + "learning_rate": 3.30505860579889e-05, + "loss": 3.425, + "step": 2143 + }, + { + "epoch": 0.06613482218940502, + "grad_norm": 0.22874094545841217, + "learning_rate": 3.3066008636644044e-05, + "loss": 3.3486, + "step": 2144 + }, + { + "epoch": 0.06616566865497843, + "grad_norm": 0.2073153406381607, + "learning_rate": 3.30814312152992e-05, + "loss": 3.4042, + "step": 2145 + }, + { + "epoch": 0.06619651512055184, + "grad_norm": 0.2060299664735794, + "learning_rate": 3.309685379395435e-05, + "loss": 3.4417, + "step": 2146 + }, + { + "epoch": 0.06622736158612526, + "grad_norm": 0.21374693512916565, + "learning_rate": 3.31122763726095e-05, + "loss": 3.4135, + "step": 2147 + }, + { + "epoch": 0.06625820805169867, + "grad_norm": 0.2174546867609024, + "learning_rate": 3.3127698951264653e-05, + "loss": 3.3983, + "step": 2148 + }, + { + "epoch": 0.06628905451727209, + "grad_norm": 0.24982640147209167, + "learning_rate": 3.31431215299198e-05, + "loss": 3.3849, + "step": 2149 + }, + { + "epoch": 0.06631990098284551, + "grad_norm": 0.2063010334968567, + "learning_rate": 3.315854410857496e-05, + "loss": 3.4112, + "step": 2150 + }, + { + "epoch": 0.06635074744841893, + "grad_norm": 0.22141946852207184, + "learning_rate": 3.3173966687230104e-05, + "loss": 3.4046, + "step": 2151 + }, + { + "epoch": 0.06638159391399234, + "grad_norm": 0.22367358207702637, + "learning_rate": 3.3189389265885256e-05, + "loss": 3.3965, + "step": 2152 + }, + { + "epoch": 0.06641244037956576, + "grad_norm": 0.5073530077934265, + "learning_rate": 3.320481184454041e-05, + "loss": 3.3758, + "step": 2153 + }, + { + "epoch": 0.06644328684513917, + "grad_norm": 0.21497415006160736, + "learning_rate": 3.322023442319556e-05, + "loss": 3.4456, + "step": 2154 + }, + { + "epoch": 0.06647413331071259, + "grad_norm": 0.2261999547481537, + "learning_rate": 3.323565700185071e-05, + "loss": 3.3977, + "step": 2155 + }, + { + "epoch": 0.06650497977628601, + "grad_norm": 0.21819202601909637, + "learning_rate": 3.3251079580505865e-05, + "loss": 3.4143, + "step": 2156 + }, + { + "epoch": 0.06653582624185943, + "grad_norm": 0.21841171383857727, + "learning_rate": 3.326650215916101e-05, + "loss": 3.3776, + "step": 2157 + }, + { + "epoch": 0.06656667270743284, + "grad_norm": 0.20023421943187714, + "learning_rate": 3.328192473781617e-05, + "loss": 3.3899, + "step": 2158 + }, + { + "epoch": 0.06659751917300626, + "grad_norm": 0.20807917416095734, + "learning_rate": 3.3297347316471315e-05, + "loss": 3.4283, + "step": 2159 + }, + { + "epoch": 0.06662836563857967, + "grad_norm": 0.20303897559642792, + "learning_rate": 3.331276989512647e-05, + "loss": 3.424, + "step": 2160 + }, + { + "epoch": 0.06665921210415308, + "grad_norm": 0.20589613914489746, + "learning_rate": 3.332819247378162e-05, + "loss": 3.3951, + "step": 2161 + }, + { + "epoch": 0.06669005856972651, + "grad_norm": 0.20855313539505005, + "learning_rate": 3.3343615052436765e-05, + "loss": 3.4022, + "step": 2162 + }, + { + "epoch": 0.06672090503529993, + "grad_norm": 0.20803438127040863, + "learning_rate": 3.3359037631091924e-05, + "loss": 3.3687, + "step": 2163 + }, + { + "epoch": 0.06675175150087334, + "grad_norm": 0.2043534219264984, + "learning_rate": 3.337446020974707e-05, + "loss": 3.376, + "step": 2164 + }, + { + "epoch": 0.06678259796644676, + "grad_norm": 0.20601396262645721, + "learning_rate": 3.338988278840222e-05, + "loss": 3.4008, + "step": 2165 + }, + { + "epoch": 0.06681344443202017, + "grad_norm": 0.2058686465024948, + "learning_rate": 3.3405305367057374e-05, + "loss": 3.3977, + "step": 2166 + }, + { + "epoch": 0.06684429089759358, + "grad_norm": 0.20091381669044495, + "learning_rate": 3.342072794571253e-05, + "loss": 3.3786, + "step": 2167 + }, + { + "epoch": 0.06687513736316701, + "grad_norm": 0.20174796879291534, + "learning_rate": 3.343615052436768e-05, + "loss": 3.389, + "step": 2168 + }, + { + "epoch": 0.06690598382874043, + "grad_norm": 0.21135219931602478, + "learning_rate": 3.3451573103022825e-05, + "loss": 3.4312, + "step": 2169 + }, + { + "epoch": 0.06693683029431384, + "grad_norm": 0.19780483841896057, + "learning_rate": 3.346699568167798e-05, + "loss": 3.4264, + "step": 2170 + }, + { + "epoch": 0.06696767675988725, + "grad_norm": 0.2003423571586609, + "learning_rate": 3.348241826033313e-05, + "loss": 3.4337, + "step": 2171 + }, + { + "epoch": 0.06699852322546067, + "grad_norm": 0.20774656534194946, + "learning_rate": 3.349784083898828e-05, + "loss": 3.3939, + "step": 2172 + }, + { + "epoch": 0.06702936969103408, + "grad_norm": 0.19919563829898834, + "learning_rate": 3.3513263417643434e-05, + "loss": 3.4212, + "step": 2173 + }, + { + "epoch": 0.06706021615660751, + "grad_norm": 0.2070872038602829, + "learning_rate": 3.352868599629858e-05, + "loss": 3.364, + "step": 2174 + }, + { + "epoch": 0.06709106262218092, + "grad_norm": 0.1939745545387268, + "learning_rate": 3.354410857495374e-05, + "loss": 3.3822, + "step": 2175 + }, + { + "epoch": 0.06712190908775434, + "grad_norm": 0.20394283533096313, + "learning_rate": 3.3559531153608884e-05, + "loss": 3.3334, + "step": 2176 + }, + { + "epoch": 0.06715275555332775, + "grad_norm": 0.2071804404258728, + "learning_rate": 3.3574953732264036e-05, + "loss": 3.374, + "step": 2177 + }, + { + "epoch": 0.06718360201890117, + "grad_norm": 0.20661425590515137, + "learning_rate": 3.359037631091919e-05, + "loss": 3.4188, + "step": 2178 + }, + { + "epoch": 0.06721444848447458, + "grad_norm": 0.21057258546352386, + "learning_rate": 3.3605798889574334e-05, + "loss": 3.4172, + "step": 2179 + }, + { + "epoch": 0.06724529495004801, + "grad_norm": 0.21579520404338837, + "learning_rate": 3.362122146822949e-05, + "loss": 3.411, + "step": 2180 + }, + { + "epoch": 0.06727614141562142, + "grad_norm": 0.2027140110731125, + "learning_rate": 3.363664404688464e-05, + "loss": 3.4025, + "step": 2181 + }, + { + "epoch": 0.06730698788119484, + "grad_norm": 0.2025529444217682, + "learning_rate": 3.365206662553979e-05, + "loss": 3.3998, + "step": 2182 + }, + { + "epoch": 0.06733783434676825, + "grad_norm": 0.19489412009716034, + "learning_rate": 3.366748920419494e-05, + "loss": 3.4062, + "step": 2183 + }, + { + "epoch": 0.06736868081234167, + "grad_norm": 0.20050688087940216, + "learning_rate": 3.3682911782850095e-05, + "loss": 3.4193, + "step": 2184 + }, + { + "epoch": 0.06739952727791508, + "grad_norm": 0.21472159028053284, + "learning_rate": 3.369833436150525e-05, + "loss": 3.3841, + "step": 2185 + }, + { + "epoch": 0.06743037374348851, + "grad_norm": 0.19987699389457703, + "learning_rate": 3.371375694016039e-05, + "loss": 3.3861, + "step": 2186 + }, + { + "epoch": 0.06746122020906192, + "grad_norm": 0.21908175945281982, + "learning_rate": 3.3729179518815546e-05, + "loss": 3.3869, + "step": 2187 + }, + { + "epoch": 0.06749206667463534, + "grad_norm": 0.2117312252521515, + "learning_rate": 3.37446020974707e-05, + "loss": 3.4112, + "step": 2188 + }, + { + "epoch": 0.06752291314020875, + "grad_norm": 0.21072886884212494, + "learning_rate": 3.376002467612585e-05, + "loss": 3.4147, + "step": 2189 + }, + { + "epoch": 0.06755375960578217, + "grad_norm": 0.19951456785202026, + "learning_rate": 3.3775447254781e-05, + "loss": 3.3982, + "step": 2190 + }, + { + "epoch": 0.06758460607135558, + "grad_norm": 0.21620365977287292, + "learning_rate": 3.379086983343615e-05, + "loss": 3.3939, + "step": 2191 + }, + { + "epoch": 0.06761545253692901, + "grad_norm": 0.20994195342063904, + "learning_rate": 3.380629241209131e-05, + "loss": 3.394, + "step": 2192 + }, + { + "epoch": 0.06764629900250242, + "grad_norm": 0.20235881209373474, + "learning_rate": 3.382171499074645e-05, + "loss": 3.4083, + "step": 2193 + }, + { + "epoch": 0.06767714546807584, + "grad_norm": 0.20897099375724792, + "learning_rate": 3.3837137569401605e-05, + "loss": 3.4173, + "step": 2194 + }, + { + "epoch": 0.06770799193364925, + "grad_norm": 0.22658559679985046, + "learning_rate": 3.385256014805676e-05, + "loss": 3.3881, + "step": 2195 + }, + { + "epoch": 0.06773883839922266, + "grad_norm": 0.21376964449882507, + "learning_rate": 3.38679827267119e-05, + "loss": 3.4288, + "step": 2196 + }, + { + "epoch": 0.06776968486479608, + "grad_norm": 0.2069305181503296, + "learning_rate": 3.388340530536706e-05, + "loss": 3.4121, + "step": 2197 + }, + { + "epoch": 0.0678005313303695, + "grad_norm": 0.20327669382095337, + "learning_rate": 3.3898827884022214e-05, + "loss": 3.4046, + "step": 2198 + }, + { + "epoch": 0.06783137779594292, + "grad_norm": 0.2165224850177765, + "learning_rate": 3.391425046267736e-05, + "loss": 3.4127, + "step": 2199 + }, + { + "epoch": 0.06786222426151634, + "grad_norm": 0.19119709730148315, + "learning_rate": 3.392967304133251e-05, + "loss": 3.3521, + "step": 2200 + }, + { + "epoch": 0.06789307072708975, + "grad_norm": 0.2222154140472412, + "learning_rate": 3.3945095619987664e-05, + "loss": 3.386, + "step": 2201 + }, + { + "epoch": 0.06792391719266316, + "grad_norm": 0.19499298930168152, + "learning_rate": 3.3960518198642816e-05, + "loss": 3.4135, + "step": 2202 + }, + { + "epoch": 0.06795476365823659, + "grad_norm": 0.20685115456581116, + "learning_rate": 3.397594077729797e-05, + "loss": 3.3666, + "step": 2203 + }, + { + "epoch": 0.06798561012381, + "grad_norm": 0.2515808939933777, + "learning_rate": 3.3991363355953114e-05, + "loss": 3.4046, + "step": 2204 + }, + { + "epoch": 0.06801645658938342, + "grad_norm": 0.20811563730239868, + "learning_rate": 3.400678593460827e-05, + "loss": 3.4164, + "step": 2205 + }, + { + "epoch": 0.06804730305495683, + "grad_norm": 0.21029651165008545, + "learning_rate": 3.402220851326342e-05, + "loss": 3.3507, + "step": 2206 + }, + { + "epoch": 0.06807814952053025, + "grad_norm": 0.1936984658241272, + "learning_rate": 3.403763109191857e-05, + "loss": 3.3802, + "step": 2207 + }, + { + "epoch": 0.06810899598610366, + "grad_norm": 0.2079598605632782, + "learning_rate": 3.405305367057372e-05, + "loss": 3.4013, + "step": 2208 + }, + { + "epoch": 0.06813984245167709, + "grad_norm": 0.20088674128055573, + "learning_rate": 3.406847624922887e-05, + "loss": 3.3945, + "step": 2209 + }, + { + "epoch": 0.0681706889172505, + "grad_norm": 0.5514625906944275, + "learning_rate": 3.408389882788403e-05, + "loss": 3.4022, + "step": 2210 + }, + { + "epoch": 0.06820153538282392, + "grad_norm": 0.19842158257961273, + "learning_rate": 3.4099321406539173e-05, + "loss": 3.3734, + "step": 2211 + }, + { + "epoch": 0.06823238184839733, + "grad_norm": 0.21692365407943726, + "learning_rate": 3.4114743985194326e-05, + "loss": 3.381, + "step": 2212 + }, + { + "epoch": 0.06826322831397075, + "grad_norm": 0.22138965129852295, + "learning_rate": 3.413016656384948e-05, + "loss": 3.386, + "step": 2213 + }, + { + "epoch": 0.06829407477954416, + "grad_norm": 0.23183368146419525, + "learning_rate": 3.414558914250463e-05, + "loss": 3.3999, + "step": 2214 + }, + { + "epoch": 0.06832492124511759, + "grad_norm": 0.20775827765464783, + "learning_rate": 3.416101172115978e-05, + "loss": 3.3932, + "step": 2215 + }, + { + "epoch": 0.068355767710691, + "grad_norm": 0.2868039608001709, + "learning_rate": 3.417643429981493e-05, + "loss": 3.3954, + "step": 2216 + }, + { + "epoch": 0.06838661417626442, + "grad_norm": 0.21427685022354126, + "learning_rate": 3.419185687847008e-05, + "loss": 3.4108, + "step": 2217 + }, + { + "epoch": 0.06841746064183783, + "grad_norm": 0.2302980124950409, + "learning_rate": 3.420727945712523e-05, + "loss": 3.4067, + "step": 2218 + }, + { + "epoch": 0.06844830710741125, + "grad_norm": 0.2073662281036377, + "learning_rate": 3.4222702035780385e-05, + "loss": 3.4008, + "step": 2219 + }, + { + "epoch": 0.06847915357298466, + "grad_norm": 0.21738730370998383, + "learning_rate": 3.423812461443554e-05, + "loss": 3.3819, + "step": 2220 + }, + { + "epoch": 0.06851000003855809, + "grad_norm": 0.2095523178577423, + "learning_rate": 3.425354719309068e-05, + "loss": 3.3788, + "step": 2221 + }, + { + "epoch": 0.0685408465041315, + "grad_norm": 0.2104516625404358, + "learning_rate": 3.426896977174584e-05, + "loss": 3.3998, + "step": 2222 + }, + { + "epoch": 0.06857169296970492, + "grad_norm": 0.23955868184566498, + "learning_rate": 3.428439235040099e-05, + "loss": 3.3769, + "step": 2223 + }, + { + "epoch": 0.06860253943527833, + "grad_norm": 0.1989419311285019, + "learning_rate": 3.429981492905614e-05, + "loss": 3.4042, + "step": 2224 + }, + { + "epoch": 0.06863338590085175, + "grad_norm": 0.23568353056907654, + "learning_rate": 3.431523750771129e-05, + "loss": 3.3949, + "step": 2225 + }, + { + "epoch": 0.06866423236642516, + "grad_norm": 0.21141743659973145, + "learning_rate": 3.433066008636644e-05, + "loss": 3.3829, + "step": 2226 + }, + { + "epoch": 0.06869507883199859, + "grad_norm": 0.19941067695617676, + "learning_rate": 3.4346082665021597e-05, + "loss": 3.3848, + "step": 2227 + }, + { + "epoch": 0.068725925297572, + "grad_norm": 0.20242057740688324, + "learning_rate": 3.436150524367674e-05, + "loss": 3.3811, + "step": 2228 + }, + { + "epoch": 0.06875677176314542, + "grad_norm": 0.2005254626274109, + "learning_rate": 3.4376927822331894e-05, + "loss": 3.3859, + "step": 2229 + }, + { + "epoch": 0.06878761822871883, + "grad_norm": 0.2063940465450287, + "learning_rate": 3.439235040098705e-05, + "loss": 3.4097, + "step": 2230 + }, + { + "epoch": 0.06881846469429224, + "grad_norm": 0.2154732197523117, + "learning_rate": 3.44077729796422e-05, + "loss": 3.4386, + "step": 2231 + }, + { + "epoch": 0.06884931115986566, + "grad_norm": 0.20537547767162323, + "learning_rate": 3.442319555829735e-05, + "loss": 3.3865, + "step": 2232 + }, + { + "epoch": 0.06888015762543909, + "grad_norm": 0.21631598472595215, + "learning_rate": 3.44386181369525e-05, + "loss": 3.3941, + "step": 2233 + }, + { + "epoch": 0.0689110040910125, + "grad_norm": 0.20317526161670685, + "learning_rate": 3.445404071560765e-05, + "loss": 3.3835, + "step": 2234 + }, + { + "epoch": 0.06894185055658592, + "grad_norm": 0.214723601937294, + "learning_rate": 3.446946329426281e-05, + "loss": 3.4073, + "step": 2235 + }, + { + "epoch": 0.06897269702215933, + "grad_norm": 0.20306679606437683, + "learning_rate": 3.4484885872917954e-05, + "loss": 3.3872, + "step": 2236 + }, + { + "epoch": 0.06900354348773274, + "grad_norm": 0.2050723433494568, + "learning_rate": 3.4500308451573106e-05, + "loss": 3.4119, + "step": 2237 + }, + { + "epoch": 0.06903438995330616, + "grad_norm": 0.20345287024974823, + "learning_rate": 3.451573103022825e-05, + "loss": 3.3844, + "step": 2238 + }, + { + "epoch": 0.06906523641887959, + "grad_norm": 0.21071217954158783, + "learning_rate": 3.4531153608883404e-05, + "loss": 3.3824, + "step": 2239 + }, + { + "epoch": 0.069096082884453, + "grad_norm": 0.20518271625041962, + "learning_rate": 3.454657618753856e-05, + "loss": 3.4209, + "step": 2240 + }, + { + "epoch": 0.06912692935002641, + "grad_norm": 0.19240964949131012, + "learning_rate": 3.456199876619371e-05, + "loss": 3.3727, + "step": 2241 + }, + { + "epoch": 0.06915777581559983, + "grad_norm": 0.19431184232234955, + "learning_rate": 3.457742134484886e-05, + "loss": 3.3801, + "step": 2242 + }, + { + "epoch": 0.06918862228117324, + "grad_norm": 0.20083647966384888, + "learning_rate": 3.4592843923504006e-05, + "loss": 3.3842, + "step": 2243 + }, + { + "epoch": 0.06921946874674666, + "grad_norm": 0.20022986829280853, + "learning_rate": 3.4608266502159165e-05, + "loss": 3.39, + "step": 2244 + }, + { + "epoch": 0.06925031521232008, + "grad_norm": 0.2015189230442047, + "learning_rate": 3.462368908081432e-05, + "loss": 3.3992, + "step": 2245 + }, + { + "epoch": 0.0692811616778935, + "grad_norm": 0.20478710532188416, + "learning_rate": 3.463911165946946e-05, + "loss": 3.4187, + "step": 2246 + }, + { + "epoch": 0.06931200814346691, + "grad_norm": 0.20219206809997559, + "learning_rate": 3.4654534238124615e-05, + "loss": 3.3978, + "step": 2247 + }, + { + "epoch": 0.06934285460904033, + "grad_norm": 0.21414610743522644, + "learning_rate": 3.466995681677977e-05, + "loss": 3.3908, + "step": 2248 + }, + { + "epoch": 0.06937370107461374, + "grad_norm": 0.2131543904542923, + "learning_rate": 3.468537939543492e-05, + "loss": 3.3496, + "step": 2249 + }, + { + "epoch": 0.06940454754018716, + "grad_norm": 0.21704840660095215, + "learning_rate": 3.470080197409007e-05, + "loss": 3.4094, + "step": 2250 + }, + { + "epoch": 0.06943539400576058, + "grad_norm": 0.21469895541667938, + "learning_rate": 3.471622455274522e-05, + "loss": 3.3839, + "step": 2251 + }, + { + "epoch": 0.069466240471334, + "grad_norm": 0.21161885559558868, + "learning_rate": 3.473164713140038e-05, + "loss": 3.3777, + "step": 2252 + }, + { + "epoch": 0.06949708693690741, + "grad_norm": 0.2013760209083557, + "learning_rate": 3.474706971005552e-05, + "loss": 3.3544, + "step": 2253 + }, + { + "epoch": 0.06952793340248083, + "grad_norm": 0.22102198004722595, + "learning_rate": 3.4762492288710675e-05, + "loss": 3.3629, + "step": 2254 + }, + { + "epoch": 0.06955877986805424, + "grad_norm": 0.21624456346035004, + "learning_rate": 3.477791486736583e-05, + "loss": 3.3788, + "step": 2255 + }, + { + "epoch": 0.06958962633362765, + "grad_norm": 0.19586117565631866, + "learning_rate": 3.479333744602097e-05, + "loss": 3.4111, + "step": 2256 + }, + { + "epoch": 0.06962047279920108, + "grad_norm": 0.21072803437709808, + "learning_rate": 3.480876002467613e-05, + "loss": 3.374, + "step": 2257 + }, + { + "epoch": 0.0696513192647745, + "grad_norm": 0.19480249285697937, + "learning_rate": 3.482418260333128e-05, + "loss": 3.3815, + "step": 2258 + }, + { + "epoch": 0.06968216573034791, + "grad_norm": 0.22180776298046112, + "learning_rate": 3.483960518198643e-05, + "loss": 3.3877, + "step": 2259 + }, + { + "epoch": 0.06971301219592133, + "grad_norm": 0.20228804647922516, + "learning_rate": 3.485502776064158e-05, + "loss": 3.4107, + "step": 2260 + }, + { + "epoch": 0.06974385866149474, + "grad_norm": 0.19721619784832, + "learning_rate": 3.4870450339296734e-05, + "loss": 3.401, + "step": 2261 + }, + { + "epoch": 0.06977470512706815, + "grad_norm": 0.1939246952533722, + "learning_rate": 3.4885872917951886e-05, + "loss": 3.3783, + "step": 2262 + }, + { + "epoch": 0.06980555159264158, + "grad_norm": 0.20124535262584686, + "learning_rate": 3.490129549660703e-05, + "loss": 3.3885, + "step": 2263 + }, + { + "epoch": 0.069836398058215, + "grad_norm": 0.19684995710849762, + "learning_rate": 3.4916718075262184e-05, + "loss": 3.3836, + "step": 2264 + }, + { + "epoch": 0.06986724452378841, + "grad_norm": 0.2110070139169693, + "learning_rate": 3.4932140653917336e-05, + "loss": 3.4116, + "step": 2265 + }, + { + "epoch": 0.06989809098936182, + "grad_norm": 0.20053498446941376, + "learning_rate": 3.494756323257249e-05, + "loss": 3.3754, + "step": 2266 + }, + { + "epoch": 0.06992893745493524, + "grad_norm": 0.21133041381835938, + "learning_rate": 3.496298581122764e-05, + "loss": 3.3673, + "step": 2267 + }, + { + "epoch": 0.06995978392050865, + "grad_norm": 0.19401776790618896, + "learning_rate": 3.4978408389882786e-05, + "loss": 3.3472, + "step": 2268 + }, + { + "epoch": 0.06999063038608208, + "grad_norm": 0.2215632051229477, + "learning_rate": 3.4993830968537946e-05, + "loss": 3.3647, + "step": 2269 + }, + { + "epoch": 0.0700214768516555, + "grad_norm": 0.20779970288276672, + "learning_rate": 3.500925354719309e-05, + "loss": 3.3724, + "step": 2270 + }, + { + "epoch": 0.07005232331722891, + "grad_norm": 0.2155521661043167, + "learning_rate": 3.502467612584824e-05, + "loss": 3.3865, + "step": 2271 + }, + { + "epoch": 0.07008316978280232, + "grad_norm": 0.20989775657653809, + "learning_rate": 3.5040098704503396e-05, + "loss": 3.3981, + "step": 2272 + }, + { + "epoch": 0.07011401624837574, + "grad_norm": 0.21262621879577637, + "learning_rate": 3.505552128315854e-05, + "loss": 3.3636, + "step": 2273 + }, + { + "epoch": 0.07014486271394915, + "grad_norm": 0.21381108462810516, + "learning_rate": 3.50709438618137e-05, + "loss": 3.4051, + "step": 2274 + }, + { + "epoch": 0.07017570917952258, + "grad_norm": 0.2104123830795288, + "learning_rate": 3.5086366440468846e-05, + "loss": 3.3728, + "step": 2275 + }, + { + "epoch": 0.070206555645096, + "grad_norm": 0.2073395848274231, + "learning_rate": 3.5101789019124e-05, + "loss": 3.4069, + "step": 2276 + }, + { + "epoch": 0.07023740211066941, + "grad_norm": 0.19871261715888977, + "learning_rate": 3.511721159777915e-05, + "loss": 3.3783, + "step": 2277 + }, + { + "epoch": 0.07026824857624282, + "grad_norm": 0.19530366361141205, + "learning_rate": 3.51326341764343e-05, + "loss": 3.391, + "step": 2278 + }, + { + "epoch": 0.07029909504181624, + "grad_norm": 0.21921902894973755, + "learning_rate": 3.5148056755089455e-05, + "loss": 3.3529, + "step": 2279 + }, + { + "epoch": 0.07032994150738965, + "grad_norm": 0.1929784119129181, + "learning_rate": 3.51634793337446e-05, + "loss": 3.3862, + "step": 2280 + }, + { + "epoch": 0.07036078797296308, + "grad_norm": 0.20282575488090515, + "learning_rate": 3.517890191239975e-05, + "loss": 3.4131, + "step": 2281 + }, + { + "epoch": 0.0703916344385365, + "grad_norm": 0.21615637838840485, + "learning_rate": 3.519432449105491e-05, + "loss": 3.395, + "step": 2282 + }, + { + "epoch": 0.07042248090410991, + "grad_norm": 0.19481438398361206, + "learning_rate": 3.520974706971006e-05, + "loss": 3.3704, + "step": 2283 + }, + { + "epoch": 0.07045332736968332, + "grad_norm": 0.19658657908439636, + "learning_rate": 3.522516964836521e-05, + "loss": 3.3867, + "step": 2284 + }, + { + "epoch": 0.07048417383525674, + "grad_norm": 0.20892782509326935, + "learning_rate": 3.5240592227020355e-05, + "loss": 3.402, + "step": 2285 + }, + { + "epoch": 0.07051502030083015, + "grad_norm": 0.21655449271202087, + "learning_rate": 3.525601480567551e-05, + "loss": 3.3804, + "step": 2286 + }, + { + "epoch": 0.07054586676640358, + "grad_norm": 0.1972951591014862, + "learning_rate": 3.5271437384330666e-05, + "loss": 3.3728, + "step": 2287 + }, + { + "epoch": 0.07057671323197699, + "grad_norm": 0.2005433440208435, + "learning_rate": 3.528685996298581e-05, + "loss": 3.3875, + "step": 2288 + }, + { + "epoch": 0.0706075596975504, + "grad_norm": 0.19993597269058228, + "learning_rate": 3.5302282541640964e-05, + "loss": 3.3855, + "step": 2289 + }, + { + "epoch": 0.07063840616312382, + "grad_norm": 0.2663000226020813, + "learning_rate": 3.531770512029611e-05, + "loss": 3.3858, + "step": 2290 + }, + { + "epoch": 0.07066925262869724, + "grad_norm": 0.21002621948719025, + "learning_rate": 3.533312769895127e-05, + "loss": 3.412, + "step": 2291 + }, + { + "epoch": 0.07070009909427065, + "grad_norm": 0.20145952701568604, + "learning_rate": 3.534855027760642e-05, + "loss": 3.3815, + "step": 2292 + }, + { + "epoch": 0.07073094555984408, + "grad_norm": 0.32326796650886536, + "learning_rate": 3.536397285626157e-05, + "loss": 3.386, + "step": 2293 + }, + { + "epoch": 0.07076179202541749, + "grad_norm": 0.2082718461751938, + "learning_rate": 3.537939543491672e-05, + "loss": 3.4002, + "step": 2294 + }, + { + "epoch": 0.0707926384909909, + "grad_norm": 0.20994830131530762, + "learning_rate": 3.539481801357187e-05, + "loss": 3.4097, + "step": 2295 + }, + { + "epoch": 0.07082348495656432, + "grad_norm": 0.2395908683538437, + "learning_rate": 3.5410240592227024e-05, + "loss": 3.3651, + "step": 2296 + }, + { + "epoch": 0.07085433142213773, + "grad_norm": 0.21231397986412048, + "learning_rate": 3.5425663170882176e-05, + "loss": 3.3684, + "step": 2297 + }, + { + "epoch": 0.07088517788771115, + "grad_norm": 0.2105076014995575, + "learning_rate": 3.544108574953732e-05, + "loss": 3.3968, + "step": 2298 + }, + { + "epoch": 0.07091602435328458, + "grad_norm": 0.21642151474952698, + "learning_rate": 3.545650832819248e-05, + "loss": 3.4275, + "step": 2299 + }, + { + "epoch": 0.07094687081885799, + "grad_norm": 0.2022365778684616, + "learning_rate": 3.5471930906847626e-05, + "loss": 3.4053, + "step": 2300 + }, + { + "epoch": 0.0709777172844314, + "grad_norm": 2.203265905380249, + "learning_rate": 3.548735348550278e-05, + "loss": 3.3871, + "step": 2301 + }, + { + "epoch": 0.07100856375000482, + "grad_norm": 0.5730552077293396, + "learning_rate": 3.550277606415793e-05, + "loss": 3.3919, + "step": 2302 + }, + { + "epoch": 0.07103941021557823, + "grad_norm": 0.21592040359973907, + "learning_rate": 3.5518198642813076e-05, + "loss": 3.3876, + "step": 2303 + }, + { + "epoch": 0.07107025668115165, + "grad_norm": 0.22680355608463287, + "learning_rate": 3.5533621221468235e-05, + "loss": 3.3884, + "step": 2304 + }, + { + "epoch": 0.07110110314672508, + "grad_norm": 0.22031250596046448, + "learning_rate": 3.554904380012338e-05, + "loss": 3.3875, + "step": 2305 + }, + { + "epoch": 0.07113194961229849, + "grad_norm": 0.33466649055480957, + "learning_rate": 3.556446637877853e-05, + "loss": 3.4066, + "step": 2306 + }, + { + "epoch": 0.0711627960778719, + "grad_norm": 0.26727697253227234, + "learning_rate": 3.5579888957433685e-05, + "loss": 3.3934, + "step": 2307 + }, + { + "epoch": 0.07119364254344532, + "grad_norm": 0.23648983240127563, + "learning_rate": 3.559531153608884e-05, + "loss": 3.4022, + "step": 2308 + }, + { + "epoch": 0.07122448900901873, + "grad_norm": 0.23143622279167175, + "learning_rate": 3.561073411474399e-05, + "loss": 3.4178, + "step": 2309 + }, + { + "epoch": 0.07125533547459215, + "grad_norm": 0.22429712116718292, + "learning_rate": 3.5626156693399135e-05, + "loss": 3.3938, + "step": 2310 + }, + { + "epoch": 0.07128618194016557, + "grad_norm": 0.23114806413650513, + "learning_rate": 3.564157927205429e-05, + "loss": 3.4062, + "step": 2311 + }, + { + "epoch": 0.07131702840573899, + "grad_norm": 0.2190990000963211, + "learning_rate": 3.565700185070944e-05, + "loss": 3.3805, + "step": 2312 + }, + { + "epoch": 0.0713478748713124, + "grad_norm": 0.20547468960285187, + "learning_rate": 3.567242442936459e-05, + "loss": 3.3547, + "step": 2313 + }, + { + "epoch": 0.07137872133688582, + "grad_norm": 0.21876582503318787, + "learning_rate": 3.5687847008019745e-05, + "loss": 3.4344, + "step": 2314 + }, + { + "epoch": 0.07140956780245923, + "grad_norm": 0.23046596348285675, + "learning_rate": 3.570326958667489e-05, + "loss": 3.4021, + "step": 2315 + }, + { + "epoch": 0.07144041426803265, + "grad_norm": 0.21365049481391907, + "learning_rate": 3.571869216533004e-05, + "loss": 3.4304, + "step": 2316 + }, + { + "epoch": 0.07147126073360607, + "grad_norm": 0.20101305842399597, + "learning_rate": 3.5734114743985195e-05, + "loss": 3.3698, + "step": 2317 + }, + { + "epoch": 0.07150210719917949, + "grad_norm": 0.21535101532936096, + "learning_rate": 3.574953732264035e-05, + "loss": 3.4093, + "step": 2318 + }, + { + "epoch": 0.0715329536647529, + "grad_norm": 0.2113439440727234, + "learning_rate": 3.57649599012955e-05, + "loss": 3.4359, + "step": 2319 + }, + { + "epoch": 0.07156380013032632, + "grad_norm": 0.20685942471027374, + "learning_rate": 3.5780382479950645e-05, + "loss": 3.3309, + "step": 2320 + }, + { + "epoch": 0.07159464659589973, + "grad_norm": 0.20910879969596863, + "learning_rate": 3.5795805058605804e-05, + "loss": 3.3841, + "step": 2321 + }, + { + "epoch": 0.07162549306147314, + "grad_norm": 0.5808777213096619, + "learning_rate": 3.581122763726095e-05, + "loss": 3.391, + "step": 2322 + }, + { + "epoch": 0.07165633952704657, + "grad_norm": 0.20344363152980804, + "learning_rate": 3.58266502159161e-05, + "loss": 3.3776, + "step": 2323 + }, + { + "epoch": 0.07168718599261999, + "grad_norm": 0.21006573736667633, + "learning_rate": 3.5842072794571254e-05, + "loss": 3.3462, + "step": 2324 + }, + { + "epoch": 0.0717180324581934, + "grad_norm": 0.20828330516815186, + "learning_rate": 3.5857495373226406e-05, + "loss": 3.425, + "step": 2325 + }, + { + "epoch": 0.07174887892376682, + "grad_norm": 0.2024289220571518, + "learning_rate": 3.587291795188156e-05, + "loss": 3.373, + "step": 2326 + }, + { + "epoch": 0.07177972538934023, + "grad_norm": 0.718325674533844, + "learning_rate": 3.5888340530536704e-05, + "loss": 3.4003, + "step": 2327 + }, + { + "epoch": 0.07181057185491364, + "grad_norm": 0.20285579562187195, + "learning_rate": 3.5903763109191856e-05, + "loss": 3.3768, + "step": 2328 + }, + { + "epoch": 0.07184141832048707, + "grad_norm": 0.21225397288799286, + "learning_rate": 3.5919185687847015e-05, + "loss": 3.412, + "step": 2329 + }, + { + "epoch": 0.07187226478606049, + "grad_norm": 0.20483987033367157, + "learning_rate": 3.593460826650216e-05, + "loss": 3.3861, + "step": 2330 + }, + { + "epoch": 0.0719031112516339, + "grad_norm": 0.5745860934257507, + "learning_rate": 3.595003084515731e-05, + "loss": 3.3753, + "step": 2331 + }, + { + "epoch": 0.07193395771720731, + "grad_norm": 0.21706050634384155, + "learning_rate": 3.596545342381246e-05, + "loss": 3.3441, + "step": 2332 + }, + { + "epoch": 0.07196480418278073, + "grad_norm": 0.21063153445720673, + "learning_rate": 3.598087600246761e-05, + "loss": 3.3676, + "step": 2333 + }, + { + "epoch": 0.07199565064835414, + "grad_norm": 0.24414435029029846, + "learning_rate": 3.599629858112277e-05, + "loss": 3.352, + "step": 2334 + }, + { + "epoch": 0.07202649711392757, + "grad_norm": 0.20755471289157867, + "learning_rate": 3.6011721159777916e-05, + "loss": 3.3406, + "step": 2335 + }, + { + "epoch": 0.07205734357950098, + "grad_norm": 0.22744202613830566, + "learning_rate": 3.602714373843307e-05, + "loss": 3.3836, + "step": 2336 + }, + { + "epoch": 0.0720881900450744, + "grad_norm": 0.2657777667045593, + "learning_rate": 3.604256631708821e-05, + "loss": 3.3554, + "step": 2337 + }, + { + "epoch": 0.07211903651064781, + "grad_norm": 0.24022512137889862, + "learning_rate": 3.605798889574337e-05, + "loss": 3.3873, + "step": 2338 + }, + { + "epoch": 0.07214988297622123, + "grad_norm": 0.20154209434986115, + "learning_rate": 3.6073411474398525e-05, + "loss": 3.3688, + "step": 2339 + }, + { + "epoch": 0.07218072944179464, + "grad_norm": 0.21628080308437347, + "learning_rate": 3.608883405305367e-05, + "loss": 3.3842, + "step": 2340 + }, + { + "epoch": 0.07221157590736807, + "grad_norm": 0.20339369773864746, + "learning_rate": 3.610425663170882e-05, + "loss": 3.3959, + "step": 2341 + }, + { + "epoch": 0.07224242237294148, + "grad_norm": 0.21172209084033966, + "learning_rate": 3.6119679210363975e-05, + "loss": 3.3906, + "step": 2342 + }, + { + "epoch": 0.0722732688385149, + "grad_norm": 0.20046332478523254, + "learning_rate": 3.613510178901913e-05, + "loss": 3.3763, + "step": 2343 + }, + { + "epoch": 0.07230411530408831, + "grad_norm": 0.21263329684734344, + "learning_rate": 3.615052436767428e-05, + "loss": 3.3699, + "step": 2344 + }, + { + "epoch": 0.07233496176966173, + "grad_norm": 0.19564978778362274, + "learning_rate": 3.6165946946329425e-05, + "loss": 3.3826, + "step": 2345 + }, + { + "epoch": 0.07236580823523514, + "grad_norm": 0.20621994137763977, + "learning_rate": 3.6181369524984584e-05, + "loss": 3.3881, + "step": 2346 + }, + { + "epoch": 0.07239665470080857, + "grad_norm": 0.20782920718193054, + "learning_rate": 3.619679210363973e-05, + "loss": 3.3666, + "step": 2347 + }, + { + "epoch": 0.07242750116638198, + "grad_norm": 0.21729828417301178, + "learning_rate": 3.621221468229488e-05, + "loss": 3.419, + "step": 2348 + }, + { + "epoch": 0.0724583476319554, + "grad_norm": 0.20702719688415527, + "learning_rate": 3.6227637260950034e-05, + "loss": 3.4009, + "step": 2349 + }, + { + "epoch": 0.07248919409752881, + "grad_norm": 0.20346401631832123, + "learning_rate": 3.624305983960518e-05, + "loss": 3.3939, + "step": 2350 + }, + { + "epoch": 0.07252004056310223, + "grad_norm": 0.20334944128990173, + "learning_rate": 3.625848241826034e-05, + "loss": 3.3652, + "step": 2351 + }, + { + "epoch": 0.07255088702867564, + "grad_norm": 0.20587801933288574, + "learning_rate": 3.6273904996915484e-05, + "loss": 3.3666, + "step": 2352 + }, + { + "epoch": 0.07258173349424907, + "grad_norm": 0.21118751168251038, + "learning_rate": 3.6289327575570637e-05, + "loss": 3.4046, + "step": 2353 + }, + { + "epoch": 0.07261257995982248, + "grad_norm": 0.2052147090435028, + "learning_rate": 3.630475015422579e-05, + "loss": 3.375, + "step": 2354 + }, + { + "epoch": 0.0726434264253959, + "grad_norm": 0.20711584389209747, + "learning_rate": 3.632017273288094e-05, + "loss": 3.3793, + "step": 2355 + }, + { + "epoch": 0.07267427289096931, + "grad_norm": 0.21170659363269806, + "learning_rate": 3.6335595311536093e-05, + "loss": 3.3973, + "step": 2356 + }, + { + "epoch": 0.07270511935654272, + "grad_norm": 0.1940140426158905, + "learning_rate": 3.635101789019124e-05, + "loss": 3.3759, + "step": 2357 + }, + { + "epoch": 0.07273596582211614, + "grad_norm": 0.21804232895374298, + "learning_rate": 3.636644046884639e-05, + "loss": 3.4118, + "step": 2358 + }, + { + "epoch": 0.07276681228768957, + "grad_norm": 0.1959022879600525, + "learning_rate": 3.6381863047501544e-05, + "loss": 3.3975, + "step": 2359 + }, + { + "epoch": 0.07279765875326298, + "grad_norm": 0.19172319769859314, + "learning_rate": 3.6397285626156696e-05, + "loss": 3.3923, + "step": 2360 + }, + { + "epoch": 0.0728285052188364, + "grad_norm": 0.2019728124141693, + "learning_rate": 3.641270820481185e-05, + "loss": 3.3783, + "step": 2361 + }, + { + "epoch": 0.07285935168440981, + "grad_norm": 0.19640541076660156, + "learning_rate": 3.6428130783466994e-05, + "loss": 3.354, + "step": 2362 + }, + { + "epoch": 0.07289019814998322, + "grad_norm": 0.20142635703086853, + "learning_rate": 3.6443553362122146e-05, + "loss": 3.3685, + "step": 2363 + }, + { + "epoch": 0.07292104461555664, + "grad_norm": 0.20425637066364288, + "learning_rate": 3.64589759407773e-05, + "loss": 3.3716, + "step": 2364 + }, + { + "epoch": 0.07295189108113007, + "grad_norm": 0.22039559483528137, + "learning_rate": 3.647439851943245e-05, + "loss": 3.3806, + "step": 2365 + }, + { + "epoch": 0.07298273754670348, + "grad_norm": 0.2139904648065567, + "learning_rate": 3.64898210980876e-05, + "loss": 3.3583, + "step": 2366 + }, + { + "epoch": 0.0730135840122769, + "grad_norm": 0.20606957376003265, + "learning_rate": 3.650524367674275e-05, + "loss": 3.3714, + "step": 2367 + }, + { + "epoch": 0.07304443047785031, + "grad_norm": 0.2141675502061844, + "learning_rate": 3.652066625539791e-05, + "loss": 3.376, + "step": 2368 + }, + { + "epoch": 0.07307527694342372, + "grad_norm": 0.20437400043010712, + "learning_rate": 3.653608883405305e-05, + "loss": 3.382, + "step": 2369 + }, + { + "epoch": 0.07310612340899714, + "grad_norm": 0.2082064002752304, + "learning_rate": 3.6551511412708205e-05, + "loss": 3.3712, + "step": 2370 + }, + { + "epoch": 0.07313696987457056, + "grad_norm": 0.19795627892017365, + "learning_rate": 3.656693399136336e-05, + "loss": 3.3579, + "step": 2371 + }, + { + "epoch": 0.07316781634014398, + "grad_norm": 0.2195448875427246, + "learning_rate": 3.658235657001851e-05, + "loss": 3.3719, + "step": 2372 + }, + { + "epoch": 0.0731986628057174, + "grad_norm": 0.1979299932718277, + "learning_rate": 3.659777914867366e-05, + "loss": 3.3503, + "step": 2373 + }, + { + "epoch": 0.07322950927129081, + "grad_norm": 0.22444194555282593, + "learning_rate": 3.661320172732881e-05, + "loss": 3.373, + "step": 2374 + }, + { + "epoch": 0.07326035573686422, + "grad_norm": 0.2189752757549286, + "learning_rate": 3.662862430598396e-05, + "loss": 3.3718, + "step": 2375 + }, + { + "epoch": 0.07329120220243764, + "grad_norm": 0.23614458739757538, + "learning_rate": 3.664404688463912e-05, + "loss": 3.4108, + "step": 2376 + }, + { + "epoch": 0.07332204866801106, + "grad_norm": 0.6868945956230164, + "learning_rate": 3.6659469463294264e-05, + "loss": 3.3645, + "step": 2377 + }, + { + "epoch": 0.07335289513358448, + "grad_norm": 0.22306658327579498, + "learning_rate": 3.667489204194942e-05, + "loss": 3.343, + "step": 2378 + }, + { + "epoch": 0.07338374159915789, + "grad_norm": 0.22898761928081512, + "learning_rate": 3.669031462060456e-05, + "loss": 3.3888, + "step": 2379 + }, + { + "epoch": 0.0734145880647313, + "grad_norm": 0.216342955827713, + "learning_rate": 3.6705737199259715e-05, + "loss": 3.4017, + "step": 2380 + }, + { + "epoch": 0.07344543453030472, + "grad_norm": 0.21122458577156067, + "learning_rate": 3.6721159777914874e-05, + "loss": 3.3969, + "step": 2381 + }, + { + "epoch": 0.07347628099587813, + "grad_norm": 0.2105821669101715, + "learning_rate": 3.673658235657002e-05, + "loss": 3.3841, + "step": 2382 + }, + { + "epoch": 0.07350712746145156, + "grad_norm": 0.2142922729253769, + "learning_rate": 3.675200493522517e-05, + "loss": 3.3969, + "step": 2383 + }, + { + "epoch": 0.07353797392702498, + "grad_norm": 0.34897446632385254, + "learning_rate": 3.6767427513880324e-05, + "loss": 3.3885, + "step": 2384 + }, + { + "epoch": 0.07356882039259839, + "grad_norm": 0.22882933914661407, + "learning_rate": 3.6782850092535476e-05, + "loss": 3.3931, + "step": 2385 + }, + { + "epoch": 0.0735996668581718, + "grad_norm": 0.21215006709098816, + "learning_rate": 3.679827267119063e-05, + "loss": 3.3593, + "step": 2386 + }, + { + "epoch": 0.07363051332374522, + "grad_norm": 0.2150963544845581, + "learning_rate": 3.6813695249845774e-05, + "loss": 3.3809, + "step": 2387 + }, + { + "epoch": 0.07366135978931863, + "grad_norm": 0.22674615681171417, + "learning_rate": 3.6829117828500926e-05, + "loss": 3.395, + "step": 2388 + }, + { + "epoch": 0.07369220625489206, + "grad_norm": 0.18895819783210754, + "learning_rate": 3.684454040715608e-05, + "loss": 3.3842, + "step": 2389 + }, + { + "epoch": 0.07372305272046548, + "grad_norm": 0.24363841116428375, + "learning_rate": 3.685996298581123e-05, + "loss": 3.3816, + "step": 2390 + }, + { + "epoch": 0.07375389918603889, + "grad_norm": 0.2091073840856552, + "learning_rate": 3.687538556446638e-05, + "loss": 3.3806, + "step": 2391 + }, + { + "epoch": 0.0737847456516123, + "grad_norm": 0.21637371182441711, + "learning_rate": 3.689080814312153e-05, + "loss": 3.394, + "step": 2392 + }, + { + "epoch": 0.07381559211718572, + "grad_norm": 0.20760980248451233, + "learning_rate": 3.690623072177668e-05, + "loss": 3.4074, + "step": 2393 + }, + { + "epoch": 0.07384643858275913, + "grad_norm": 0.22051464021205902, + "learning_rate": 3.692165330043183e-05, + "loss": 3.3387, + "step": 2394 + }, + { + "epoch": 0.07387728504833256, + "grad_norm": 0.2126236855983734, + "learning_rate": 3.6937075879086985e-05, + "loss": 3.355, + "step": 2395 + }, + { + "epoch": 0.07390813151390598, + "grad_norm": 0.2109571248292923, + "learning_rate": 3.695249845774214e-05, + "loss": 3.3684, + "step": 2396 + }, + { + "epoch": 0.07393897797947939, + "grad_norm": 0.21767102181911469, + "learning_rate": 3.696792103639728e-05, + "loss": 3.3696, + "step": 2397 + }, + { + "epoch": 0.0739698244450528, + "grad_norm": 0.2137502282857895, + "learning_rate": 3.698334361505244e-05, + "loss": 3.3528, + "step": 2398 + }, + { + "epoch": 0.07400067091062622, + "grad_norm": 0.23181135952472687, + "learning_rate": 3.699876619370759e-05, + "loss": 3.4011, + "step": 2399 + }, + { + "epoch": 0.07403151737619963, + "grad_norm": 0.21160924434661865, + "learning_rate": 3.701418877236274e-05, + "loss": 3.355, + "step": 2400 + }, + { + "epoch": 0.07406236384177306, + "grad_norm": 0.19536587595939636, + "learning_rate": 3.702961135101789e-05, + "loss": 3.3784, + "step": 2401 + }, + { + "epoch": 0.07409321030734647, + "grad_norm": 0.21203386783599854, + "learning_rate": 3.7045033929673045e-05, + "loss": 3.3771, + "step": 2402 + }, + { + "epoch": 0.07412405677291989, + "grad_norm": 0.8249675035476685, + "learning_rate": 3.70604565083282e-05, + "loss": 3.4136, + "step": 2403 + }, + { + "epoch": 0.0741549032384933, + "grad_norm": 0.23180986940860748, + "learning_rate": 3.707587908698334e-05, + "loss": 3.4218, + "step": 2404 + }, + { + "epoch": 0.07418574970406672, + "grad_norm": 0.22549548745155334, + "learning_rate": 3.7091301665638495e-05, + "loss": 3.4068, + "step": 2405 + }, + { + "epoch": 0.07421659616964013, + "grad_norm": 0.23500971496105194, + "learning_rate": 3.710672424429365e-05, + "loss": 3.4156, + "step": 2406 + }, + { + "epoch": 0.07424744263521356, + "grad_norm": 0.4167129099369049, + "learning_rate": 3.71221468229488e-05, + "loss": 3.3814, + "step": 2407 + }, + { + "epoch": 0.07427828910078697, + "grad_norm": 0.21847569942474365, + "learning_rate": 3.713756940160395e-05, + "loss": 3.3925, + "step": 2408 + }, + { + "epoch": 0.07430913556636039, + "grad_norm": 0.23915210366249084, + "learning_rate": 3.71529919802591e-05, + "loss": 3.3641, + "step": 2409 + }, + { + "epoch": 0.0743399820319338, + "grad_norm": 0.21835319697856903, + "learning_rate": 3.716841455891425e-05, + "loss": 3.357, + "step": 2410 + }, + { + "epoch": 0.07437082849750722, + "grad_norm": 0.2102622389793396, + "learning_rate": 3.71838371375694e-05, + "loss": 3.3803, + "step": 2411 + }, + { + "epoch": 0.07440167496308063, + "grad_norm": 0.2076576203107834, + "learning_rate": 3.7199259716224554e-05, + "loss": 3.3898, + "step": 2412 + }, + { + "epoch": 0.07443252142865406, + "grad_norm": 0.23679427802562714, + "learning_rate": 3.7214682294879706e-05, + "loss": 3.3862, + "step": 2413 + }, + { + "epoch": 0.07446336789422747, + "grad_norm": 0.20853009819984436, + "learning_rate": 3.723010487353485e-05, + "loss": 3.3398, + "step": 2414 + }, + { + "epoch": 0.07449421435980089, + "grad_norm": 0.20370687544345856, + "learning_rate": 3.724552745219001e-05, + "loss": 3.4106, + "step": 2415 + }, + { + "epoch": 0.0745250608253743, + "grad_norm": 0.20892341434955597, + "learning_rate": 3.7260950030845156e-05, + "loss": 3.3881, + "step": 2416 + }, + { + "epoch": 0.07455590729094771, + "grad_norm": 0.32478293776512146, + "learning_rate": 3.727637260950031e-05, + "loss": 3.4004, + "step": 2417 + }, + { + "epoch": 0.07458675375652113, + "grad_norm": 0.2696201205253601, + "learning_rate": 3.729179518815546e-05, + "loss": 3.3889, + "step": 2418 + }, + { + "epoch": 0.07461760022209456, + "grad_norm": 0.2123071551322937, + "learning_rate": 3.730721776681061e-05, + "loss": 3.3507, + "step": 2419 + }, + { + "epoch": 0.07464844668766797, + "grad_norm": 0.2255987524986267, + "learning_rate": 3.7322640345465766e-05, + "loss": 3.3723, + "step": 2420 + }, + { + "epoch": 0.07467929315324139, + "grad_norm": 0.22614224255084991, + "learning_rate": 3.733806292412091e-05, + "loss": 3.3454, + "step": 2421 + }, + { + "epoch": 0.0747101396188148, + "grad_norm": 0.20029862225055695, + "learning_rate": 3.7353485502776063e-05, + "loss": 3.3938, + "step": 2422 + }, + { + "epoch": 0.07474098608438821, + "grad_norm": 0.2079920619726181, + "learning_rate": 3.736890808143122e-05, + "loss": 3.3685, + "step": 2423 + }, + { + "epoch": 0.07477183254996163, + "grad_norm": 0.33636412024497986, + "learning_rate": 3.738433066008637e-05, + "loss": 3.3741, + "step": 2424 + }, + { + "epoch": 0.07480267901553506, + "grad_norm": 0.22085709869861603, + "learning_rate": 3.739975323874152e-05, + "loss": 3.3642, + "step": 2425 + }, + { + "epoch": 0.07483352548110847, + "grad_norm": 0.20084267854690552, + "learning_rate": 3.741517581739667e-05, + "loss": 3.3574, + "step": 2426 + }, + { + "epoch": 0.07486437194668188, + "grad_norm": 0.2182771861553192, + "learning_rate": 3.743059839605182e-05, + "loss": 3.3399, + "step": 2427 + }, + { + "epoch": 0.0748952184122553, + "grad_norm": 0.21303170919418335, + "learning_rate": 3.744602097470698e-05, + "loss": 3.382, + "step": 2428 + }, + { + "epoch": 0.07492606487782871, + "grad_norm": 0.21619661152362823, + "learning_rate": 3.746144355336212e-05, + "loss": 3.372, + "step": 2429 + }, + { + "epoch": 0.07495691134340213, + "grad_norm": 0.20257771015167236, + "learning_rate": 3.7476866132017275e-05, + "loss": 3.3745, + "step": 2430 + }, + { + "epoch": 0.07498775780897556, + "grad_norm": 0.20304733514785767, + "learning_rate": 3.749228871067243e-05, + "loss": 3.3814, + "step": 2431 + }, + { + "epoch": 0.07501860427454897, + "grad_norm": 0.2240181714296341, + "learning_rate": 3.750771128932758e-05, + "loss": 3.3952, + "step": 2432 + }, + { + "epoch": 0.07504945074012238, + "grad_norm": 0.19493453204631805, + "learning_rate": 3.752313386798273e-05, + "loss": 3.3832, + "step": 2433 + }, + { + "epoch": 0.0750802972056958, + "grad_norm": 0.22061482071876526, + "learning_rate": 3.753855644663788e-05, + "loss": 3.3627, + "step": 2434 + }, + { + "epoch": 0.07511114367126921, + "grad_norm": 0.20937775075435638, + "learning_rate": 3.755397902529303e-05, + "loss": 3.3756, + "step": 2435 + }, + { + "epoch": 0.07514199013684263, + "grad_norm": 0.20881856977939606, + "learning_rate": 3.756940160394818e-05, + "loss": 3.3836, + "step": 2436 + }, + { + "epoch": 0.07517283660241605, + "grad_norm": 0.21888893842697144, + "learning_rate": 3.7584824182603334e-05, + "loss": 3.3862, + "step": 2437 + }, + { + "epoch": 0.07520368306798947, + "grad_norm": 0.19525942206382751, + "learning_rate": 3.760024676125849e-05, + "loss": 3.3804, + "step": 2438 + }, + { + "epoch": 0.07523452953356288, + "grad_norm": 0.19792258739471436, + "learning_rate": 3.761566933991363e-05, + "loss": 3.3744, + "step": 2439 + }, + { + "epoch": 0.0752653759991363, + "grad_norm": 0.19679009914398193, + "learning_rate": 3.7631091918568784e-05, + "loss": 3.3234, + "step": 2440 + }, + { + "epoch": 0.07529622246470971, + "grad_norm": 0.19670075178146362, + "learning_rate": 3.764651449722394e-05, + "loss": 3.4055, + "step": 2441 + }, + { + "epoch": 0.07532706893028313, + "grad_norm": 0.20845529437065125, + "learning_rate": 3.766193707587909e-05, + "loss": 3.3728, + "step": 2442 + }, + { + "epoch": 0.07535791539585655, + "grad_norm": 0.17978599667549133, + "learning_rate": 3.767735965453424e-05, + "loss": 3.4038, + "step": 2443 + }, + { + "epoch": 0.07538876186142997, + "grad_norm": 0.19545479118824005, + "learning_rate": 3.769278223318939e-05, + "loss": 3.3879, + "step": 2444 + }, + { + "epoch": 0.07541960832700338, + "grad_norm": 0.1952594816684723, + "learning_rate": 3.7708204811844546e-05, + "loss": 3.4025, + "step": 2445 + }, + { + "epoch": 0.0754504547925768, + "grad_norm": 0.2115422934293747, + "learning_rate": 3.772362739049969e-05, + "loss": 3.3801, + "step": 2446 + }, + { + "epoch": 0.07548130125815021, + "grad_norm": 0.20287831127643585, + "learning_rate": 3.7739049969154844e-05, + "loss": 3.3818, + "step": 2447 + }, + { + "epoch": 0.07551214772372362, + "grad_norm": 0.18777671456336975, + "learning_rate": 3.7754472547809996e-05, + "loss": 3.3958, + "step": 2448 + }, + { + "epoch": 0.07554299418929705, + "grad_norm": 0.1997004896402359, + "learning_rate": 3.776989512646515e-05, + "loss": 3.3476, + "step": 2449 + }, + { + "epoch": 0.07557384065487047, + "grad_norm": 0.22454766929149628, + "learning_rate": 3.77853177051203e-05, + "loss": 3.3737, + "step": 2450 + }, + { + "epoch": 0.07560468712044388, + "grad_norm": 0.18502606451511383, + "learning_rate": 3.7800740283775446e-05, + "loss": 3.3889, + "step": 2451 + }, + { + "epoch": 0.0756355335860173, + "grad_norm": 0.20038342475891113, + "learning_rate": 3.78161628624306e-05, + "loss": 3.3853, + "step": 2452 + }, + { + "epoch": 0.07566638005159071, + "grad_norm": 0.19232326745986938, + "learning_rate": 3.783158544108575e-05, + "loss": 3.386, + "step": 2453 + }, + { + "epoch": 0.07569722651716412, + "grad_norm": 0.21437883377075195, + "learning_rate": 3.78470080197409e-05, + "loss": 3.3464, + "step": 2454 + }, + { + "epoch": 0.07572807298273755, + "grad_norm": 0.2047601342201233, + "learning_rate": 3.7862430598396055e-05, + "loss": 3.4015, + "step": 2455 + }, + { + "epoch": 0.07575891944831097, + "grad_norm": 0.19663874804973602, + "learning_rate": 3.78778531770512e-05, + "loss": 3.405, + "step": 2456 + }, + { + "epoch": 0.07578976591388438, + "grad_norm": 0.21925503015518188, + "learning_rate": 3.789327575570635e-05, + "loss": 3.3824, + "step": 2457 + }, + { + "epoch": 0.0758206123794578, + "grad_norm": 0.18708688020706177, + "learning_rate": 3.7908698334361505e-05, + "loss": 3.3417, + "step": 2458 + }, + { + "epoch": 0.07585145884503121, + "grad_norm": 0.20734362304210663, + "learning_rate": 3.792412091301666e-05, + "loss": 3.3765, + "step": 2459 + }, + { + "epoch": 0.07588230531060462, + "grad_norm": 0.1934652030467987, + "learning_rate": 3.793954349167181e-05, + "loss": 3.3973, + "step": 2460 + }, + { + "epoch": 0.07591315177617805, + "grad_norm": 0.19445182383060455, + "learning_rate": 3.7954966070326955e-05, + "loss": 3.3446, + "step": 2461 + }, + { + "epoch": 0.07594399824175146, + "grad_norm": 0.1949009895324707, + "learning_rate": 3.7970388648982115e-05, + "loss": 3.3824, + "step": 2462 + }, + { + "epoch": 0.07597484470732488, + "grad_norm": 0.2011028230190277, + "learning_rate": 3.798581122763727e-05, + "loss": 3.3695, + "step": 2463 + }, + { + "epoch": 0.0760056911728983, + "grad_norm": 0.2032075971364975, + "learning_rate": 3.800123380629241e-05, + "loss": 3.3463, + "step": 2464 + }, + { + "epoch": 0.07603653763847171, + "grad_norm": 0.20624615252017975, + "learning_rate": 3.8016656384947565e-05, + "loss": 3.3726, + "step": 2465 + }, + { + "epoch": 0.07606738410404512, + "grad_norm": 0.20308269560337067, + "learning_rate": 3.803207896360272e-05, + "loss": 3.3815, + "step": 2466 + }, + { + "epoch": 0.07609823056961855, + "grad_norm": 0.19995644688606262, + "learning_rate": 3.804750154225787e-05, + "loss": 3.3148, + "step": 2467 + }, + { + "epoch": 0.07612907703519196, + "grad_norm": 0.2028922736644745, + "learning_rate": 3.806292412091302e-05, + "loss": 3.3826, + "step": 2468 + }, + { + "epoch": 0.07615992350076538, + "grad_norm": 0.23445071280002594, + "learning_rate": 3.807834669956817e-05, + "loss": 3.3639, + "step": 2469 + }, + { + "epoch": 0.07619076996633879, + "grad_norm": 0.1960982382297516, + "learning_rate": 3.8093769278223326e-05, + "loss": 3.3583, + "step": 2470 + }, + { + "epoch": 0.0762216164319122, + "grad_norm": 0.21753659844398499, + "learning_rate": 3.810919185687847e-05, + "loss": 3.3337, + "step": 2471 + }, + { + "epoch": 0.07625246289748562, + "grad_norm": 0.22121013700962067, + "learning_rate": 3.8124614435533624e-05, + "loss": 3.406, + "step": 2472 + }, + { + "epoch": 0.07628330936305905, + "grad_norm": 0.20100191235542297, + "learning_rate": 3.8140037014188776e-05, + "loss": 3.3493, + "step": 2473 + }, + { + "epoch": 0.07631415582863246, + "grad_norm": 0.20199702680110931, + "learning_rate": 3.815545959284392e-05, + "loss": 3.3697, + "step": 2474 + }, + { + "epoch": 0.07634500229420588, + "grad_norm": 0.19837570190429688, + "learning_rate": 3.817088217149908e-05, + "loss": 3.3237, + "step": 2475 + }, + { + "epoch": 0.07637584875977929, + "grad_norm": 0.19400954246520996, + "learning_rate": 3.8186304750154226e-05, + "loss": 3.3515, + "step": 2476 + }, + { + "epoch": 0.0764066952253527, + "grad_norm": 0.2048172503709793, + "learning_rate": 3.820172732880938e-05, + "loss": 3.3529, + "step": 2477 + }, + { + "epoch": 0.07643754169092613, + "grad_norm": 0.23420263826847076, + "learning_rate": 3.821714990746453e-05, + "loss": 3.3966, + "step": 2478 + }, + { + "epoch": 0.07646838815649955, + "grad_norm": 0.19640710949897766, + "learning_rate": 3.823257248611968e-05, + "loss": 3.3749, + "step": 2479 + }, + { + "epoch": 0.07649923462207296, + "grad_norm": 0.21068759262561798, + "learning_rate": 3.8247995064774836e-05, + "loss": 3.3435, + "step": 2480 + }, + { + "epoch": 0.07653008108764638, + "grad_norm": 0.19237467646598816, + "learning_rate": 3.826341764342998e-05, + "loss": 3.3673, + "step": 2481 + }, + { + "epoch": 0.07656092755321979, + "grad_norm": 0.22359590232372284, + "learning_rate": 3.827884022208513e-05, + "loss": 3.3795, + "step": 2482 + }, + { + "epoch": 0.0765917740187932, + "grad_norm": 0.19620725512504578, + "learning_rate": 3.8294262800740286e-05, + "loss": 3.3688, + "step": 2483 + }, + { + "epoch": 0.07662262048436663, + "grad_norm": 0.20824258029460907, + "learning_rate": 3.830968537939544e-05, + "loss": 3.3846, + "step": 2484 + }, + { + "epoch": 0.07665346694994005, + "grad_norm": 0.19797927141189575, + "learning_rate": 3.832510795805059e-05, + "loss": 3.3513, + "step": 2485 + }, + { + "epoch": 0.07668431341551346, + "grad_norm": 0.201787069439888, + "learning_rate": 3.8340530536705736e-05, + "loss": 3.3764, + "step": 2486 + }, + { + "epoch": 0.07671515988108688, + "grad_norm": 0.2039644420146942, + "learning_rate": 3.835595311536089e-05, + "loss": 3.4099, + "step": 2487 + }, + { + "epoch": 0.07674600634666029, + "grad_norm": 0.21225886046886444, + "learning_rate": 3.837137569401604e-05, + "loss": 3.3321, + "step": 2488 + }, + { + "epoch": 0.0767768528122337, + "grad_norm": 0.20790500938892365, + "learning_rate": 3.838679827267119e-05, + "loss": 3.373, + "step": 2489 + }, + { + "epoch": 0.07680769927780713, + "grad_norm": 0.20241466164588928, + "learning_rate": 3.8402220851326345e-05, + "loss": 3.379, + "step": 2490 + }, + { + "epoch": 0.07683854574338055, + "grad_norm": 0.1945326328277588, + "learning_rate": 3.841764342998149e-05, + "loss": 3.3498, + "step": 2491 + }, + { + "epoch": 0.07686939220895396, + "grad_norm": 0.19986413419246674, + "learning_rate": 3.843306600863665e-05, + "loss": 3.345, + "step": 2492 + }, + { + "epoch": 0.07690023867452737, + "grad_norm": 0.18908998370170593, + "learning_rate": 3.8448488587291795e-05, + "loss": 3.3209, + "step": 2493 + }, + { + "epoch": 0.07693108514010079, + "grad_norm": 0.20536305010318756, + "learning_rate": 3.846391116594695e-05, + "loss": 3.3929, + "step": 2494 + }, + { + "epoch": 0.0769619316056742, + "grad_norm": 0.18323969841003418, + "learning_rate": 3.84793337446021e-05, + "loss": 3.3644, + "step": 2495 + }, + { + "epoch": 0.07699277807124763, + "grad_norm": 0.19898656010627747, + "learning_rate": 3.849475632325725e-05, + "loss": 3.344, + "step": 2496 + }, + { + "epoch": 0.07702362453682104, + "grad_norm": 0.18760496377944946, + "learning_rate": 3.8510178901912404e-05, + "loss": 3.3409, + "step": 2497 + }, + { + "epoch": 0.07705447100239446, + "grad_norm": 0.1903563290834427, + "learning_rate": 3.852560148056755e-05, + "loss": 3.4187, + "step": 2498 + }, + { + "epoch": 0.07708531746796787, + "grad_norm": 0.21152320504188538, + "learning_rate": 3.85410240592227e-05, + "loss": 3.3784, + "step": 2499 + }, + { + "epoch": 0.07711616393354129, + "grad_norm": 0.1859981119632721, + "learning_rate": 3.855644663787786e-05, + "loss": 3.3933, + "step": 2500 + }, + { + "epoch": 0.0771470103991147, + "grad_norm": 0.20654451847076416, + "learning_rate": 3.8571869216533007e-05, + "loss": 3.3409, + "step": 2501 + }, + { + "epoch": 0.07717785686468813, + "grad_norm": 0.20259824395179749, + "learning_rate": 3.858729179518816e-05, + "loss": 3.3424, + "step": 2502 + }, + { + "epoch": 0.07720870333026154, + "grad_norm": 0.22506731748580933, + "learning_rate": 3.8602714373843304e-05, + "loss": 3.3409, + "step": 2503 + }, + { + "epoch": 0.07723954979583496, + "grad_norm": 0.1909119039773941, + "learning_rate": 3.861813695249846e-05, + "loss": 3.36, + "step": 2504 + }, + { + "epoch": 0.07727039626140837, + "grad_norm": 0.19505606591701508, + "learning_rate": 3.8633559531153616e-05, + "loss": 3.385, + "step": 2505 + }, + { + "epoch": 0.07730124272698179, + "grad_norm": 0.1996556669473648, + "learning_rate": 3.864898210980876e-05, + "loss": 3.3716, + "step": 2506 + }, + { + "epoch": 0.0773320891925552, + "grad_norm": 0.18668098747730255, + "learning_rate": 3.8664404688463914e-05, + "loss": 3.3509, + "step": 2507 + }, + { + "epoch": 0.07736293565812863, + "grad_norm": 0.19117596745491028, + "learning_rate": 3.867982726711906e-05, + "loss": 3.344, + "step": 2508 + }, + { + "epoch": 0.07739378212370204, + "grad_norm": 0.1993093490600586, + "learning_rate": 3.869524984577422e-05, + "loss": 3.3747, + "step": 2509 + }, + { + "epoch": 0.07742462858927546, + "grad_norm": 0.1971176117658615, + "learning_rate": 3.871067242442937e-05, + "loss": 3.3441, + "step": 2510 + }, + { + "epoch": 0.07745547505484887, + "grad_norm": 0.21173429489135742, + "learning_rate": 3.8726095003084516e-05, + "loss": 3.3784, + "step": 2511 + }, + { + "epoch": 0.07748632152042229, + "grad_norm": 0.18697303533554077, + "learning_rate": 3.874151758173967e-05, + "loss": 3.3775, + "step": 2512 + }, + { + "epoch": 0.0775171679859957, + "grad_norm": 0.19404345750808716, + "learning_rate": 3.875694016039482e-05, + "loss": 3.3947, + "step": 2513 + }, + { + "epoch": 0.07754801445156913, + "grad_norm": 0.1887557953596115, + "learning_rate": 3.877236273904997e-05, + "loss": 3.3942, + "step": 2514 + }, + { + "epoch": 0.07757886091714254, + "grad_norm": 0.2015741765499115, + "learning_rate": 3.8787785317705125e-05, + "loss": 3.3782, + "step": 2515 + }, + { + "epoch": 0.07760970738271596, + "grad_norm": 0.21329519152641296, + "learning_rate": 3.880320789636027e-05, + "loss": 3.3735, + "step": 2516 + }, + { + "epoch": 0.07764055384828937, + "grad_norm": 0.18830743432044983, + "learning_rate": 3.881863047501542e-05, + "loss": 3.3598, + "step": 2517 + }, + { + "epoch": 0.07767140031386278, + "grad_norm": 0.19113928079605103, + "learning_rate": 3.8834053053670575e-05, + "loss": 3.3831, + "step": 2518 + }, + { + "epoch": 0.0777022467794362, + "grad_norm": 0.19682565331459045, + "learning_rate": 3.884947563232573e-05, + "loss": 3.3657, + "step": 2519 + }, + { + "epoch": 0.07773309324500963, + "grad_norm": 0.19550418853759766, + "learning_rate": 3.886489821098088e-05, + "loss": 3.3679, + "step": 2520 + }, + { + "epoch": 0.07776393971058304, + "grad_norm": 0.19702666997909546, + "learning_rate": 3.8880320789636025e-05, + "loss": 3.4076, + "step": 2521 + }, + { + "epoch": 0.07779478617615646, + "grad_norm": 0.20497384667396545, + "learning_rate": 3.8895743368291184e-05, + "loss": 3.3793, + "step": 2522 + }, + { + "epoch": 0.07782563264172987, + "grad_norm": 0.18429400026798248, + "learning_rate": 3.891116594694633e-05, + "loss": 3.3122, + "step": 2523 + }, + { + "epoch": 0.07785647910730328, + "grad_norm": 0.19060133397579193, + "learning_rate": 3.892658852560148e-05, + "loss": 3.3473, + "step": 2524 + }, + { + "epoch": 0.0778873255728767, + "grad_norm": 0.18940526247024536, + "learning_rate": 3.8942011104256635e-05, + "loss": 3.3623, + "step": 2525 + }, + { + "epoch": 0.07791817203845013, + "grad_norm": 0.19012922048568726, + "learning_rate": 3.895743368291179e-05, + "loss": 3.3757, + "step": 2526 + }, + { + "epoch": 0.07794901850402354, + "grad_norm": 0.19598953425884247, + "learning_rate": 3.897285626156694e-05, + "loss": 3.3687, + "step": 2527 + }, + { + "epoch": 0.07797986496959695, + "grad_norm": 0.19693982601165771, + "learning_rate": 3.8988278840222085e-05, + "loss": 3.3666, + "step": 2528 + }, + { + "epoch": 0.07801071143517037, + "grad_norm": 0.19388945400714874, + "learning_rate": 3.900370141887724e-05, + "loss": 3.382, + "step": 2529 + }, + { + "epoch": 0.07804155790074378, + "grad_norm": 0.20897674560546875, + "learning_rate": 3.901912399753239e-05, + "loss": 3.3717, + "step": 2530 + }, + { + "epoch": 0.0780724043663172, + "grad_norm": 0.20036879181861877, + "learning_rate": 3.903454657618754e-05, + "loss": 3.3973, + "step": 2531 + }, + { + "epoch": 0.07810325083189062, + "grad_norm": 0.18970496952533722, + "learning_rate": 3.9049969154842694e-05, + "loss": 3.3694, + "step": 2532 + }, + { + "epoch": 0.07813409729746404, + "grad_norm": 0.20213910937309265, + "learning_rate": 3.906539173349784e-05, + "loss": 3.3551, + "step": 2533 + }, + { + "epoch": 0.07816494376303745, + "grad_norm": 0.20053638517856598, + "learning_rate": 3.908081431215299e-05, + "loss": 3.3532, + "step": 2534 + }, + { + "epoch": 0.07819579022861087, + "grad_norm": 0.2257729470729828, + "learning_rate": 3.9096236890808144e-05, + "loss": 3.3725, + "step": 2535 + }, + { + "epoch": 0.07822663669418428, + "grad_norm": 0.21659474074840546, + "learning_rate": 3.9111659469463296e-05, + "loss": 3.3255, + "step": 2536 + }, + { + "epoch": 0.0782574831597577, + "grad_norm": 0.21585169434547424, + "learning_rate": 3.912708204811845e-05, + "loss": 3.3886, + "step": 2537 + }, + { + "epoch": 0.07828832962533112, + "grad_norm": 0.21433813869953156, + "learning_rate": 3.9142504626773594e-05, + "loss": 3.3279, + "step": 2538 + }, + { + "epoch": 0.07831917609090454, + "grad_norm": 0.1943010687828064, + "learning_rate": 3.915792720542875e-05, + "loss": 3.4032, + "step": 2539 + }, + { + "epoch": 0.07835002255647795, + "grad_norm": 0.21334430575370789, + "learning_rate": 3.91733497840839e-05, + "loss": 3.423, + "step": 2540 + }, + { + "epoch": 0.07838086902205137, + "grad_norm": 0.1955939084291458, + "learning_rate": 3.918877236273905e-05, + "loss": 3.3362, + "step": 2541 + }, + { + "epoch": 0.07841171548762478, + "grad_norm": 0.19827419519424438, + "learning_rate": 3.92041949413942e-05, + "loss": 3.3553, + "step": 2542 + }, + { + "epoch": 0.0784425619531982, + "grad_norm": 0.18901658058166504, + "learning_rate": 3.9219617520049355e-05, + "loss": 3.3448, + "step": 2543 + }, + { + "epoch": 0.07847340841877162, + "grad_norm": 0.20474925637245178, + "learning_rate": 3.923504009870451e-05, + "loss": 3.393, + "step": 2544 + }, + { + "epoch": 0.07850425488434504, + "grad_norm": 0.2934277355670929, + "learning_rate": 3.925046267735965e-05, + "loss": 3.3784, + "step": 2545 + }, + { + "epoch": 0.07853510134991845, + "grad_norm": 0.19160786271095276, + "learning_rate": 3.9265885256014806e-05, + "loss": 3.3776, + "step": 2546 + }, + { + "epoch": 0.07856594781549187, + "grad_norm": 0.20257166028022766, + "learning_rate": 3.9281307834669965e-05, + "loss": 3.3305, + "step": 2547 + }, + { + "epoch": 0.07859679428106528, + "grad_norm": 0.19790059328079224, + "learning_rate": 3.929673041332511e-05, + "loss": 3.3949, + "step": 2548 + }, + { + "epoch": 0.0786276407466387, + "grad_norm": 0.2091807723045349, + "learning_rate": 3.931215299198026e-05, + "loss": 3.3696, + "step": 2549 + }, + { + "epoch": 0.07865848721221212, + "grad_norm": 0.20120273530483246, + "learning_rate": 3.932757557063541e-05, + "loss": 3.374, + "step": 2550 + }, + { + "epoch": 0.07868933367778554, + "grad_norm": 0.19538873434066772, + "learning_rate": 3.934299814929056e-05, + "loss": 3.3767, + "step": 2551 + }, + { + "epoch": 0.07872018014335895, + "grad_norm": 0.20536457002162933, + "learning_rate": 3.935842072794572e-05, + "loss": 3.3551, + "step": 2552 + }, + { + "epoch": 0.07875102660893236, + "grad_norm": 0.20150189101696014, + "learning_rate": 3.9373843306600865e-05, + "loss": 3.382, + "step": 2553 + }, + { + "epoch": 0.07878187307450578, + "grad_norm": 0.20816241204738617, + "learning_rate": 3.938926588525602e-05, + "loss": 3.3658, + "step": 2554 + }, + { + "epoch": 0.07881271954007919, + "grad_norm": 0.21211804449558258, + "learning_rate": 3.940468846391116e-05, + "loss": 3.3363, + "step": 2555 + }, + { + "epoch": 0.07884356600565262, + "grad_norm": 0.19702959060668945, + "learning_rate": 3.942011104256632e-05, + "loss": 3.3906, + "step": 2556 + }, + { + "epoch": 0.07887441247122604, + "grad_norm": 0.21939271688461304, + "learning_rate": 3.9435533621221474e-05, + "loss": 3.3247, + "step": 2557 + }, + { + "epoch": 0.07890525893679945, + "grad_norm": 0.19041141867637634, + "learning_rate": 3.945095619987662e-05, + "loss": 3.3491, + "step": 2558 + }, + { + "epoch": 0.07893610540237286, + "grad_norm": 0.19639268517494202, + "learning_rate": 3.946637877853177e-05, + "loss": 3.3491, + "step": 2559 + }, + { + "epoch": 0.07896695186794628, + "grad_norm": 0.2021104097366333, + "learning_rate": 3.9481801357186924e-05, + "loss": 3.3376, + "step": 2560 + }, + { + "epoch": 0.07899779833351969, + "grad_norm": 0.19578948616981506, + "learning_rate": 3.9497223935842076e-05, + "loss": 3.3635, + "step": 2561 + }, + { + "epoch": 0.07902864479909312, + "grad_norm": 0.20835258066654205, + "learning_rate": 3.951264651449723e-05, + "loss": 3.3834, + "step": 2562 + }, + { + "epoch": 0.07905949126466653, + "grad_norm": 0.3204231560230255, + "learning_rate": 3.9528069093152374e-05, + "loss": 3.3588, + "step": 2563 + }, + { + "epoch": 0.07909033773023995, + "grad_norm": 0.19895651936531067, + "learning_rate": 3.9543491671807527e-05, + "loss": 3.3883, + "step": 2564 + }, + { + "epoch": 0.07912118419581336, + "grad_norm": 0.2079656720161438, + "learning_rate": 3.955891425046268e-05, + "loss": 3.3488, + "step": 2565 + }, + { + "epoch": 0.07915203066138678, + "grad_norm": 0.1987820565700531, + "learning_rate": 3.957433682911783e-05, + "loss": 3.3824, + "step": 2566 + }, + { + "epoch": 0.07918287712696019, + "grad_norm": 0.20994390547275543, + "learning_rate": 3.9589759407772983e-05, + "loss": 3.3536, + "step": 2567 + }, + { + "epoch": 0.07921372359253362, + "grad_norm": 0.19891560077667236, + "learning_rate": 3.960518198642813e-05, + "loss": 3.3874, + "step": 2568 + }, + { + "epoch": 0.07924457005810703, + "grad_norm": 0.2108825445175171, + "learning_rate": 3.962060456508329e-05, + "loss": 3.3343, + "step": 2569 + }, + { + "epoch": 0.07927541652368045, + "grad_norm": 0.2914073169231415, + "learning_rate": 3.9636027143738434e-05, + "loss": 3.3602, + "step": 2570 + }, + { + "epoch": 0.07930626298925386, + "grad_norm": 0.20113776624202728, + "learning_rate": 3.9651449722393586e-05, + "loss": 3.3781, + "step": 2571 + }, + { + "epoch": 0.07933710945482728, + "grad_norm": 0.18633724749088287, + "learning_rate": 3.966687230104874e-05, + "loss": 3.3414, + "step": 2572 + }, + { + "epoch": 0.07936795592040069, + "grad_norm": 0.20951920747756958, + "learning_rate": 3.968229487970389e-05, + "loss": 3.3809, + "step": 2573 + }, + { + "epoch": 0.07939880238597412, + "grad_norm": 0.2070133239030838, + "learning_rate": 3.969771745835904e-05, + "loss": 3.3378, + "step": 2574 + }, + { + "epoch": 0.07942964885154753, + "grad_norm": 0.20170538127422333, + "learning_rate": 3.971314003701419e-05, + "loss": 3.3697, + "step": 2575 + }, + { + "epoch": 0.07946049531712095, + "grad_norm": 0.2124345600605011, + "learning_rate": 3.972856261566934e-05, + "loss": 3.3808, + "step": 2576 + }, + { + "epoch": 0.07949134178269436, + "grad_norm": 0.19907541573047638, + "learning_rate": 3.974398519432449e-05, + "loss": 3.367, + "step": 2577 + }, + { + "epoch": 0.07952218824826777, + "grad_norm": 0.1992700695991516, + "learning_rate": 3.9759407772979645e-05, + "loss": 3.3607, + "step": 2578 + }, + { + "epoch": 0.07955303471384119, + "grad_norm": 0.19982734322547913, + "learning_rate": 3.97748303516348e-05, + "loss": 3.3831, + "step": 2579 + }, + { + "epoch": 0.07958388117941462, + "grad_norm": 0.24415159225463867, + "learning_rate": 3.979025293028994e-05, + "loss": 3.3823, + "step": 2580 + }, + { + "epoch": 0.07961472764498803, + "grad_norm": 0.25259530544281006, + "learning_rate": 3.9805675508945095e-05, + "loss": 3.3587, + "step": 2581 + }, + { + "epoch": 0.07964557411056145, + "grad_norm": 0.19578760862350464, + "learning_rate": 3.982109808760025e-05, + "loss": 3.3848, + "step": 2582 + }, + { + "epoch": 0.07967642057613486, + "grad_norm": 0.19705119729042053, + "learning_rate": 3.98365206662554e-05, + "loss": 3.3415, + "step": 2583 + }, + { + "epoch": 0.07970726704170827, + "grad_norm": 0.20686139166355133, + "learning_rate": 3.985194324491055e-05, + "loss": 3.3556, + "step": 2584 + }, + { + "epoch": 0.07973811350728169, + "grad_norm": 0.19089742004871368, + "learning_rate": 3.98673658235657e-05, + "loss": 3.4126, + "step": 2585 + }, + { + "epoch": 0.07976895997285512, + "grad_norm": 0.2011856585741043, + "learning_rate": 3.988278840222086e-05, + "loss": 3.365, + "step": 2586 + }, + { + "epoch": 0.07979980643842853, + "grad_norm": 0.3910815417766571, + "learning_rate": 3.9898210980876e-05, + "loss": 3.3563, + "step": 2587 + }, + { + "epoch": 0.07983065290400194, + "grad_norm": 0.20249450206756592, + "learning_rate": 3.9913633559531154e-05, + "loss": 3.321, + "step": 2588 + }, + { + "epoch": 0.07986149936957536, + "grad_norm": 0.19444622099399567, + "learning_rate": 3.992905613818631e-05, + "loss": 3.3768, + "step": 2589 + }, + { + "epoch": 0.07989234583514877, + "grad_norm": 0.2267790585756302, + "learning_rate": 3.994447871684146e-05, + "loss": 3.3545, + "step": 2590 + }, + { + "epoch": 0.07992319230072219, + "grad_norm": 0.18243993818759918, + "learning_rate": 3.995990129549661e-05, + "loss": 3.3283, + "step": 2591 + }, + { + "epoch": 0.07995403876629562, + "grad_norm": 0.21813629567623138, + "learning_rate": 3.997532387415176e-05, + "loss": 3.3618, + "step": 2592 + }, + { + "epoch": 0.07998488523186903, + "grad_norm": 0.1932801902294159, + "learning_rate": 3.999074645280691e-05, + "loss": 3.3864, + "step": 2593 + }, + { + "epoch": 0.08001573169744244, + "grad_norm": 0.19603987038135529, + "learning_rate": 4.000616903146206e-05, + "loss": 3.3312, + "step": 2594 + }, + { + "epoch": 0.08004657816301586, + "grad_norm": 0.19983482360839844, + "learning_rate": 4.0021591610117214e-05, + "loss": 3.3464, + "step": 2595 + }, + { + "epoch": 0.08007742462858927, + "grad_norm": 0.339275598526001, + "learning_rate": 4.0037014188772366e-05, + "loss": 3.3653, + "step": 2596 + }, + { + "epoch": 0.08010827109416269, + "grad_norm": 0.19830270111560822, + "learning_rate": 4.005243676742751e-05, + "loss": 3.3298, + "step": 2597 + }, + { + "epoch": 0.08013911755973611, + "grad_norm": 0.19002267718315125, + "learning_rate": 4.0067859346082664e-05, + "loss": 3.354, + "step": 2598 + }, + { + "epoch": 0.08016996402530953, + "grad_norm": 0.19882524013519287, + "learning_rate": 4.008328192473782e-05, + "loss": 3.3609, + "step": 2599 + }, + { + "epoch": 0.08020081049088294, + "grad_norm": 0.19939708709716797, + "learning_rate": 4.009870450339297e-05, + "loss": 3.3689, + "step": 2600 + }, + { + "epoch": 0.08023165695645636, + "grad_norm": 0.19590599834918976, + "learning_rate": 4.011412708204812e-05, + "loss": 3.3818, + "step": 2601 + }, + { + "epoch": 0.08026250342202977, + "grad_norm": 0.20920996367931366, + "learning_rate": 4.0129549660703266e-05, + "loss": 3.3776, + "step": 2602 + }, + { + "epoch": 0.08029334988760319, + "grad_norm": 0.21368682384490967, + "learning_rate": 4.0144972239358425e-05, + "loss": 3.371, + "step": 2603 + }, + { + "epoch": 0.08032419635317661, + "grad_norm": 0.20281541347503662, + "learning_rate": 4.016039481801358e-05, + "loss": 3.368, + "step": 2604 + }, + { + "epoch": 0.08035504281875003, + "grad_norm": 0.1899651288986206, + "learning_rate": 4.017581739666872e-05, + "loss": 3.3629, + "step": 2605 + }, + { + "epoch": 0.08038588928432344, + "grad_norm": 0.21566277742385864, + "learning_rate": 4.0191239975323875e-05, + "loss": 3.3651, + "step": 2606 + }, + { + "epoch": 0.08041673574989686, + "grad_norm": 0.20472025871276855, + "learning_rate": 4.020666255397903e-05, + "loss": 3.3439, + "step": 2607 + }, + { + "epoch": 0.08044758221547027, + "grad_norm": 0.22107873857021332, + "learning_rate": 4.022208513263418e-05, + "loss": 3.3793, + "step": 2608 + }, + { + "epoch": 0.08047842868104368, + "grad_norm": 0.20095393061637878, + "learning_rate": 4.023750771128933e-05, + "loss": 3.359, + "step": 2609 + }, + { + "epoch": 0.08050927514661711, + "grad_norm": 0.20816835761070251, + "learning_rate": 4.025293028994448e-05, + "loss": 3.3452, + "step": 2610 + }, + { + "epoch": 0.08054012161219053, + "grad_norm": 0.21216146647930145, + "learning_rate": 4.026835286859963e-05, + "loss": 3.3305, + "step": 2611 + }, + { + "epoch": 0.08057096807776394, + "grad_norm": 0.19066286087036133, + "learning_rate": 4.028377544725478e-05, + "loss": 3.3712, + "step": 2612 + }, + { + "epoch": 0.08060181454333736, + "grad_norm": 0.2172059714794159, + "learning_rate": 4.0299198025909935e-05, + "loss": 3.3823, + "step": 2613 + }, + { + "epoch": 0.08063266100891077, + "grad_norm": 0.18973661959171295, + "learning_rate": 4.031462060456509e-05, + "loss": 3.319, + "step": 2614 + }, + { + "epoch": 0.08066350747448418, + "grad_norm": 0.23043496906757355, + "learning_rate": 4.033004318322023e-05, + "loss": 3.357, + "step": 2615 + }, + { + "epoch": 0.08069435394005761, + "grad_norm": 0.19412288069725037, + "learning_rate": 4.034546576187539e-05, + "loss": 3.3743, + "step": 2616 + }, + { + "epoch": 0.08072520040563103, + "grad_norm": 0.20770113170146942, + "learning_rate": 4.036088834053054e-05, + "loss": 3.3682, + "step": 2617 + }, + { + "epoch": 0.08075604687120444, + "grad_norm": 0.18844597041606903, + "learning_rate": 4.037631091918569e-05, + "loss": 3.3338, + "step": 2618 + }, + { + "epoch": 0.08078689333677785, + "grad_norm": 0.2101985216140747, + "learning_rate": 4.039173349784084e-05, + "loss": 3.3526, + "step": 2619 + }, + { + "epoch": 0.08081773980235127, + "grad_norm": 0.1888302117586136, + "learning_rate": 4.0407156076495994e-05, + "loss": 3.3714, + "step": 2620 + }, + { + "epoch": 0.08084858626792468, + "grad_norm": 0.2072719782590866, + "learning_rate": 4.0422578655151146e-05, + "loss": 3.3768, + "step": 2621 + }, + { + "epoch": 0.08087943273349811, + "grad_norm": 0.18707838654518127, + "learning_rate": 4.043800123380629e-05, + "loss": 3.3717, + "step": 2622 + }, + { + "epoch": 0.08091027919907152, + "grad_norm": 0.2121260166168213, + "learning_rate": 4.0453423812461444e-05, + "loss": 3.3777, + "step": 2623 + }, + { + "epoch": 0.08094112566464494, + "grad_norm": 0.1800328493118286, + "learning_rate": 4.0468846391116596e-05, + "loss": 3.3645, + "step": 2624 + }, + { + "epoch": 0.08097197213021835, + "grad_norm": 0.9002565741539001, + "learning_rate": 4.048426896977175e-05, + "loss": 3.3411, + "step": 2625 + }, + { + "epoch": 0.08100281859579177, + "grad_norm": 0.19062620401382446, + "learning_rate": 4.04996915484269e-05, + "loss": 3.358, + "step": 2626 + }, + { + "epoch": 0.08103366506136518, + "grad_norm": 0.22968816757202148, + "learning_rate": 4.0515114127082047e-05, + "loss": 3.3836, + "step": 2627 + }, + { + "epoch": 0.08106451152693861, + "grad_norm": 0.6461688876152039, + "learning_rate": 4.05305367057372e-05, + "loss": 3.3639, + "step": 2628 + }, + { + "epoch": 0.08109535799251202, + "grad_norm": 0.22390641272068024, + "learning_rate": 4.054595928439235e-05, + "loss": 3.3636, + "step": 2629 + }, + { + "epoch": 0.08112620445808544, + "grad_norm": 0.21422497928142548, + "learning_rate": 4.0561381863047503e-05, + "loss": 3.3996, + "step": 2630 + }, + { + "epoch": 0.08115705092365885, + "grad_norm": 0.5630872249603271, + "learning_rate": 4.0576804441702656e-05, + "loss": 3.3666, + "step": 2631 + }, + { + "epoch": 0.08118789738923227, + "grad_norm": 0.2180708348751068, + "learning_rate": 4.05922270203578e-05, + "loss": 3.3624, + "step": 2632 + }, + { + "epoch": 0.08121874385480568, + "grad_norm": 0.2137863039970398, + "learning_rate": 4.060764959901296e-05, + "loss": 3.352, + "step": 2633 + }, + { + "epoch": 0.08124959032037911, + "grad_norm": 0.2460484653711319, + "learning_rate": 4.0623072177668106e-05, + "loss": 3.4012, + "step": 2634 + }, + { + "epoch": 0.08128043678595252, + "grad_norm": 0.25203582644462585, + "learning_rate": 4.063849475632326e-05, + "loss": 3.333, + "step": 2635 + }, + { + "epoch": 0.08131128325152594, + "grad_norm": 0.2272152304649353, + "learning_rate": 4.065391733497841e-05, + "loss": 3.3633, + "step": 2636 + }, + { + "epoch": 0.08134212971709935, + "grad_norm": 0.2053152322769165, + "learning_rate": 4.066933991363356e-05, + "loss": 3.3704, + "step": 2637 + }, + { + "epoch": 0.08137297618267277, + "grad_norm": 0.2101818472146988, + "learning_rate": 4.0684762492288715e-05, + "loss": 3.3485, + "step": 2638 + }, + { + "epoch": 0.08140382264824618, + "grad_norm": 0.2103378027677536, + "learning_rate": 4.070018507094386e-05, + "loss": 3.3463, + "step": 2639 + }, + { + "epoch": 0.08143466911381961, + "grad_norm": 0.21136096119880676, + "learning_rate": 4.071560764959901e-05, + "loss": 3.3514, + "step": 2640 + }, + { + "epoch": 0.08146551557939302, + "grad_norm": 0.20804111659526825, + "learning_rate": 4.0731030228254165e-05, + "loss": 3.3499, + "step": 2641 + }, + { + "epoch": 0.08149636204496644, + "grad_norm": 0.20744380354881287, + "learning_rate": 4.074645280690932e-05, + "loss": 3.3707, + "step": 2642 + }, + { + "epoch": 0.08152720851053985, + "grad_norm": 0.20122183859348297, + "learning_rate": 4.076187538556447e-05, + "loss": 3.3658, + "step": 2643 + }, + { + "epoch": 0.08155805497611326, + "grad_norm": 0.20985284447669983, + "learning_rate": 4.0777297964219615e-05, + "loss": 3.3096, + "step": 2644 + }, + { + "epoch": 0.08158890144168668, + "grad_norm": 0.21078509092330933, + "learning_rate": 4.079272054287477e-05, + "loss": 3.3477, + "step": 2645 + }, + { + "epoch": 0.0816197479072601, + "grad_norm": 0.19095174968242645, + "learning_rate": 4.0808143121529927e-05, + "loss": 3.3784, + "step": 2646 + }, + { + "epoch": 0.08165059437283352, + "grad_norm": 0.21873871982097626, + "learning_rate": 4.082356570018507e-05, + "loss": 3.338, + "step": 2647 + }, + { + "epoch": 0.08168144083840694, + "grad_norm": 0.1984986960887909, + "learning_rate": 4.0838988278840224e-05, + "loss": 3.3456, + "step": 2648 + }, + { + "epoch": 0.08171228730398035, + "grad_norm": 0.20458723604679108, + "learning_rate": 4.085441085749537e-05, + "loss": 3.3593, + "step": 2649 + }, + { + "epoch": 0.08174313376955376, + "grad_norm": 0.19422349333763123, + "learning_rate": 4.086983343615053e-05, + "loss": 3.3396, + "step": 2650 + }, + { + "epoch": 0.08177398023512718, + "grad_norm": 0.2083793431520462, + "learning_rate": 4.088525601480568e-05, + "loss": 3.3682, + "step": 2651 + }, + { + "epoch": 0.0818048267007006, + "grad_norm": 0.18382851779460907, + "learning_rate": 4.090067859346083e-05, + "loss": 3.3471, + "step": 2652 + }, + { + "epoch": 0.08183567316627402, + "grad_norm": 0.21529421210289001, + "learning_rate": 4.091610117211598e-05, + "loss": 3.3314, + "step": 2653 + }, + { + "epoch": 0.08186651963184743, + "grad_norm": 0.20703855156898499, + "learning_rate": 4.093152375077113e-05, + "loss": 3.3563, + "step": 2654 + }, + { + "epoch": 0.08189736609742085, + "grad_norm": 0.20387372374534607, + "learning_rate": 4.0946946329426284e-05, + "loss": 3.3651, + "step": 2655 + }, + { + "epoch": 0.08192821256299426, + "grad_norm": 0.19959570467472076, + "learning_rate": 4.0962368908081436e-05, + "loss": 3.3371, + "step": 2656 + }, + { + "epoch": 0.08195905902856768, + "grad_norm": 0.19471944868564606, + "learning_rate": 4.097779148673658e-05, + "loss": 3.358, + "step": 2657 + }, + { + "epoch": 0.0819899054941411, + "grad_norm": 0.21234434843063354, + "learning_rate": 4.0993214065391734e-05, + "loss": 3.3566, + "step": 2658 + }, + { + "epoch": 0.08202075195971452, + "grad_norm": 0.19934958219528198, + "learning_rate": 4.1008636644046886e-05, + "loss": 3.3771, + "step": 2659 + }, + { + "epoch": 0.08205159842528793, + "grad_norm": 0.2151016741991043, + "learning_rate": 4.102405922270204e-05, + "loss": 3.3601, + "step": 2660 + }, + { + "epoch": 0.08208244489086135, + "grad_norm": 0.2276594489812851, + "learning_rate": 4.103948180135719e-05, + "loss": 3.3511, + "step": 2661 + }, + { + "epoch": 0.08211329135643476, + "grad_norm": 0.198682501912117, + "learning_rate": 4.1054904380012336e-05, + "loss": 3.348, + "step": 2662 + }, + { + "epoch": 0.08214413782200818, + "grad_norm": 0.33139920234680176, + "learning_rate": 4.1070326958667495e-05, + "loss": 3.3914, + "step": 2663 + }, + { + "epoch": 0.0821749842875816, + "grad_norm": 0.19979533553123474, + "learning_rate": 4.108574953732264e-05, + "loss": 3.3412, + "step": 2664 + }, + { + "epoch": 0.08220583075315502, + "grad_norm": 0.23429690301418304, + "learning_rate": 4.110117211597779e-05, + "loss": 3.3998, + "step": 2665 + }, + { + "epoch": 0.08223667721872843, + "grad_norm": 0.20819495618343353, + "learning_rate": 4.1116594694632945e-05, + "loss": 3.372, + "step": 2666 + }, + { + "epoch": 0.08226752368430185, + "grad_norm": 0.24374166131019592, + "learning_rate": 4.11320172732881e-05, + "loss": 3.3397, + "step": 2667 + }, + { + "epoch": 0.08229837014987526, + "grad_norm": 0.22758780419826508, + "learning_rate": 4.114743985194325e-05, + "loss": 3.3935, + "step": 2668 + }, + { + "epoch": 0.08232921661544867, + "grad_norm": 0.20706790685653687, + "learning_rate": 4.1162862430598395e-05, + "loss": 3.3731, + "step": 2669 + }, + { + "epoch": 0.0823600630810221, + "grad_norm": 0.20777761936187744, + "learning_rate": 4.117828500925355e-05, + "loss": 3.3577, + "step": 2670 + }, + { + "epoch": 0.08239090954659552, + "grad_norm": 0.20772960782051086, + "learning_rate": 4.11937075879087e-05, + "loss": 3.4154, + "step": 2671 + }, + { + "epoch": 0.08242175601216893, + "grad_norm": 0.19667896628379822, + "learning_rate": 4.120913016656385e-05, + "loss": 3.3554, + "step": 2672 + }, + { + "epoch": 0.08245260247774235, + "grad_norm": 0.19903163611888885, + "learning_rate": 4.1224552745219005e-05, + "loss": 3.3487, + "step": 2673 + }, + { + "epoch": 0.08248344894331576, + "grad_norm": 0.20571006834506989, + "learning_rate": 4.123997532387415e-05, + "loss": 3.3576, + "step": 2674 + }, + { + "epoch": 0.08251429540888917, + "grad_norm": 0.21406254172325134, + "learning_rate": 4.12553979025293e-05, + "loss": 3.3619, + "step": 2675 + }, + { + "epoch": 0.0825451418744626, + "grad_norm": 0.2022210955619812, + "learning_rate": 4.1270820481184455e-05, + "loss": 3.3619, + "step": 2676 + }, + { + "epoch": 0.08257598834003602, + "grad_norm": 0.788866400718689, + "learning_rate": 4.128624305983961e-05, + "loss": 3.3453, + "step": 2677 + }, + { + "epoch": 0.08260683480560943, + "grad_norm": 0.209885835647583, + "learning_rate": 4.130166563849476e-05, + "loss": 3.3307, + "step": 2678 + }, + { + "epoch": 0.08263768127118284, + "grad_norm": 0.22237123548984528, + "learning_rate": 4.1317088217149905e-05, + "loss": 3.3753, + "step": 2679 + }, + { + "epoch": 0.08266852773675626, + "grad_norm": 0.22307905554771423, + "learning_rate": 4.1332510795805064e-05, + "loss": 3.3739, + "step": 2680 + }, + { + "epoch": 0.08269937420232967, + "grad_norm": 0.2002476304769516, + "learning_rate": 4.134793337446021e-05, + "loss": 3.3597, + "step": 2681 + }, + { + "epoch": 0.0827302206679031, + "grad_norm": 0.22157321870326996, + "learning_rate": 4.136335595311536e-05, + "loss": 3.37, + "step": 2682 + }, + { + "epoch": 0.08276106713347652, + "grad_norm": 0.20179133117198944, + "learning_rate": 4.1378778531770514e-05, + "loss": 3.3438, + "step": 2683 + }, + { + "epoch": 0.08279191359904993, + "grad_norm": 0.19338715076446533, + "learning_rate": 4.1394201110425666e-05, + "loss": 3.3706, + "step": 2684 + }, + { + "epoch": 0.08282276006462334, + "grad_norm": 0.21736972033977509, + "learning_rate": 4.140962368908082e-05, + "loss": 3.3574, + "step": 2685 + }, + { + "epoch": 0.08285360653019676, + "grad_norm": 0.1990717202425003, + "learning_rate": 4.1425046267735964e-05, + "loss": 3.3561, + "step": 2686 + }, + { + "epoch": 0.08288445299577017, + "grad_norm": 0.214432954788208, + "learning_rate": 4.1440468846391116e-05, + "loss": 3.353, + "step": 2687 + }, + { + "epoch": 0.0829152994613436, + "grad_norm": 0.20148485898971558, + "learning_rate": 4.145589142504627e-05, + "loss": 3.3281, + "step": 2688 + }, + { + "epoch": 0.08294614592691701, + "grad_norm": 0.21057872474193573, + "learning_rate": 4.147131400370142e-05, + "loss": 3.3671, + "step": 2689 + }, + { + "epoch": 0.08297699239249043, + "grad_norm": 0.19244736433029175, + "learning_rate": 4.148673658235657e-05, + "loss": 3.3403, + "step": 2690 + }, + { + "epoch": 0.08300783885806384, + "grad_norm": 0.21521297097206116, + "learning_rate": 4.1502159161011726e-05, + "loss": 3.3605, + "step": 2691 + }, + { + "epoch": 0.08303868532363726, + "grad_norm": 0.21871209144592285, + "learning_rate": 4.151758173966687e-05, + "loss": 3.3803, + "step": 2692 + }, + { + "epoch": 0.08306953178921067, + "grad_norm": 0.21768032014369965, + "learning_rate": 4.153300431832203e-05, + "loss": 3.3658, + "step": 2693 + }, + { + "epoch": 0.0831003782547841, + "grad_norm": 0.19349735975265503, + "learning_rate": 4.1548426896977176e-05, + "loss": 3.3321, + "step": 2694 + }, + { + "epoch": 0.08313122472035751, + "grad_norm": 0.2121044248342514, + "learning_rate": 4.156384947563233e-05, + "loss": 3.322, + "step": 2695 + }, + { + "epoch": 0.08316207118593093, + "grad_norm": 0.3339119553565979, + "learning_rate": 4.157927205428748e-05, + "loss": 3.3583, + "step": 2696 + }, + { + "epoch": 0.08319291765150434, + "grad_norm": 0.19966255128383636, + "learning_rate": 4.159469463294263e-05, + "loss": 3.3585, + "step": 2697 + }, + { + "epoch": 0.08322376411707776, + "grad_norm": 0.2002507895231247, + "learning_rate": 4.1610117211597785e-05, + "loss": 3.3222, + "step": 2698 + }, + { + "epoch": 0.08325461058265117, + "grad_norm": 0.19989249110221863, + "learning_rate": 4.162553979025293e-05, + "loss": 3.3478, + "step": 2699 + }, + { + "epoch": 0.0832854570482246, + "grad_norm": 0.19477616250514984, + "learning_rate": 4.164096236890808e-05, + "loss": 3.3625, + "step": 2700 + }, + { + "epoch": 0.08331630351379801, + "grad_norm": 0.20153042674064636, + "learning_rate": 4.1656384947563235e-05, + "loss": 3.3336, + "step": 2701 + }, + { + "epoch": 0.08334714997937143, + "grad_norm": 0.19155806303024292, + "learning_rate": 4.167180752621839e-05, + "loss": 3.3339, + "step": 2702 + }, + { + "epoch": 0.08337799644494484, + "grad_norm": 0.19387607276439667, + "learning_rate": 4.168723010487354e-05, + "loss": 3.3422, + "step": 2703 + }, + { + "epoch": 0.08340884291051825, + "grad_norm": 0.2064722776412964, + "learning_rate": 4.1702652683528685e-05, + "loss": 3.3479, + "step": 2704 + }, + { + "epoch": 0.08343968937609167, + "grad_norm": 0.19646550714969635, + "learning_rate": 4.171807526218384e-05, + "loss": 3.3691, + "step": 2705 + }, + { + "epoch": 0.0834705358416651, + "grad_norm": 0.20612645149230957, + "learning_rate": 4.173349784083899e-05, + "loss": 3.3598, + "step": 2706 + }, + { + "epoch": 0.08350138230723851, + "grad_norm": 0.19935554265975952, + "learning_rate": 4.174892041949414e-05, + "loss": 3.3673, + "step": 2707 + }, + { + "epoch": 0.08353222877281193, + "grad_norm": 0.19303660094738007, + "learning_rate": 4.1764342998149294e-05, + "loss": 3.3312, + "step": 2708 + }, + { + "epoch": 0.08356307523838534, + "grad_norm": 0.20376873016357422, + "learning_rate": 4.177976557680444e-05, + "loss": 3.3205, + "step": 2709 + }, + { + "epoch": 0.08359392170395875, + "grad_norm": 0.19535961747169495, + "learning_rate": 4.17951881554596e-05, + "loss": 3.3568, + "step": 2710 + }, + { + "epoch": 0.08362476816953217, + "grad_norm": 0.19834627211093903, + "learning_rate": 4.1810610734114744e-05, + "loss": 3.3739, + "step": 2711 + }, + { + "epoch": 0.0836556146351056, + "grad_norm": 0.21676325798034668, + "learning_rate": 4.1826033312769897e-05, + "loss": 3.3629, + "step": 2712 + }, + { + "epoch": 0.08368646110067901, + "grad_norm": 0.1881699115037918, + "learning_rate": 4.184145589142505e-05, + "loss": 3.3482, + "step": 2713 + }, + { + "epoch": 0.08371730756625242, + "grad_norm": 0.21231327950954437, + "learning_rate": 4.18568784700802e-05, + "loss": 3.3611, + "step": 2714 + }, + { + "epoch": 0.08374815403182584, + "grad_norm": 0.19231107831001282, + "learning_rate": 4.1872301048735353e-05, + "loss": 3.3428, + "step": 2715 + }, + { + "epoch": 0.08377900049739925, + "grad_norm": 0.2127636820077896, + "learning_rate": 4.18877236273905e-05, + "loss": 3.3676, + "step": 2716 + }, + { + "epoch": 0.08380984696297267, + "grad_norm": 0.1976720094680786, + "learning_rate": 4.190314620604565e-05, + "loss": 3.3823, + "step": 2717 + }, + { + "epoch": 0.0838406934285461, + "grad_norm": 0.20641964673995972, + "learning_rate": 4.1918568784700804e-05, + "loss": 3.388, + "step": 2718 + }, + { + "epoch": 0.08387153989411951, + "grad_norm": 0.19578008353710175, + "learning_rate": 4.1933991363355956e-05, + "loss": 3.3092, + "step": 2719 + }, + { + "epoch": 0.08390238635969292, + "grad_norm": 0.21170960366725922, + "learning_rate": 4.194941394201111e-05, + "loss": 3.3376, + "step": 2720 + }, + { + "epoch": 0.08393323282526634, + "grad_norm": 0.2007468193769455, + "learning_rate": 4.1964836520666254e-05, + "loss": 3.3513, + "step": 2721 + }, + { + "epoch": 0.08396407929083975, + "grad_norm": 0.20648978650569916, + "learning_rate": 4.1980259099321406e-05, + "loss": 3.3849, + "step": 2722 + }, + { + "epoch": 0.08399492575641317, + "grad_norm": 0.18994596600532532, + "learning_rate": 4.199568167797656e-05, + "loss": 3.3463, + "step": 2723 + }, + { + "epoch": 0.0840257722219866, + "grad_norm": 0.20676660537719727, + "learning_rate": 4.201110425663171e-05, + "loss": 3.3732, + "step": 2724 + }, + { + "epoch": 0.08405661868756001, + "grad_norm": 0.19431574642658234, + "learning_rate": 4.202652683528686e-05, + "loss": 3.3328, + "step": 2725 + }, + { + "epoch": 0.08408746515313342, + "grad_norm": 0.21687105298042297, + "learning_rate": 4.204194941394201e-05, + "loss": 3.3714, + "step": 2726 + }, + { + "epoch": 0.08411831161870684, + "grad_norm": 0.19710446894168854, + "learning_rate": 4.205737199259717e-05, + "loss": 3.3435, + "step": 2727 + }, + { + "epoch": 0.08414915808428025, + "grad_norm": 0.2103477269411087, + "learning_rate": 4.207279457125232e-05, + "loss": 3.3268, + "step": 2728 + }, + { + "epoch": 0.08418000454985367, + "grad_norm": 0.6918432712554932, + "learning_rate": 4.2088217149907465e-05, + "loss": 3.3674, + "step": 2729 + }, + { + "epoch": 0.0842108510154271, + "grad_norm": 0.19798190891742706, + "learning_rate": 4.210363972856262e-05, + "loss": 3.3507, + "step": 2730 + }, + { + "epoch": 0.08424169748100051, + "grad_norm": 0.30688410997390747, + "learning_rate": 4.211906230721777e-05, + "loss": 3.3572, + "step": 2731 + }, + { + "epoch": 0.08427254394657392, + "grad_norm": 0.5762542486190796, + "learning_rate": 4.213448488587292e-05, + "loss": 3.3422, + "step": 2732 + }, + { + "epoch": 0.08430339041214734, + "grad_norm": 0.19840769469738007, + "learning_rate": 4.2149907464528074e-05, + "loss": 3.3339, + "step": 2733 + }, + { + "epoch": 0.08433423687772075, + "grad_norm": 0.21585023403167725, + "learning_rate": 4.216533004318322e-05, + "loss": 3.3663, + "step": 2734 + }, + { + "epoch": 0.08436508334329416, + "grad_norm": 0.2066277116537094, + "learning_rate": 4.218075262183837e-05, + "loss": 3.3695, + "step": 2735 + }, + { + "epoch": 0.08439592980886759, + "grad_norm": 0.21537823975086212, + "learning_rate": 4.2196175200493525e-05, + "loss": 3.3529, + "step": 2736 + }, + { + "epoch": 0.084426776274441, + "grad_norm": 0.20247608423233032, + "learning_rate": 4.221159777914868e-05, + "loss": 3.3473, + "step": 2737 + }, + { + "epoch": 0.08445762274001442, + "grad_norm": 0.20479565858840942, + "learning_rate": 4.222702035780383e-05, + "loss": 3.3263, + "step": 2738 + }, + { + "epoch": 0.08448846920558784, + "grad_norm": 0.19302469491958618, + "learning_rate": 4.2242442936458975e-05, + "loss": 3.3788, + "step": 2739 + }, + { + "epoch": 0.08451931567116125, + "grad_norm": 0.2339627742767334, + "learning_rate": 4.2257865515114134e-05, + "loss": 3.3364, + "step": 2740 + }, + { + "epoch": 0.08455016213673466, + "grad_norm": 0.1899828463792801, + "learning_rate": 4.227328809376928e-05, + "loss": 3.3483, + "step": 2741 + }, + { + "epoch": 0.08458100860230809, + "grad_norm": 0.19357599318027496, + "learning_rate": 4.228871067242443e-05, + "loss": 3.3299, + "step": 2742 + }, + { + "epoch": 0.0846118550678815, + "grad_norm": 0.4601619243621826, + "learning_rate": 4.2304133251079584e-05, + "loss": 3.3805, + "step": 2743 + }, + { + "epoch": 0.08464270153345492, + "grad_norm": 0.20335645973682404, + "learning_rate": 4.2319555829734736e-05, + "loss": 3.3598, + "step": 2744 + }, + { + "epoch": 0.08467354799902833, + "grad_norm": 0.1961936205625534, + "learning_rate": 4.233497840838989e-05, + "loss": 3.332, + "step": 2745 + }, + { + "epoch": 0.08470439446460175, + "grad_norm": 0.19972912967205048, + "learning_rate": 4.2350400987045034e-05, + "loss": 3.3532, + "step": 2746 + }, + { + "epoch": 0.08473524093017516, + "grad_norm": 0.19993628561496735, + "learning_rate": 4.2365823565700186e-05, + "loss": 3.3597, + "step": 2747 + }, + { + "epoch": 0.08476608739574859, + "grad_norm": 0.1919040083885193, + "learning_rate": 4.238124614435534e-05, + "loss": 3.3221, + "step": 2748 + }, + { + "epoch": 0.084796933861322, + "grad_norm": 0.19453957676887512, + "learning_rate": 4.239666872301049e-05, + "loss": 3.3448, + "step": 2749 + }, + { + "epoch": 0.08482778032689542, + "grad_norm": 0.19384317100048065, + "learning_rate": 4.241209130166564e-05, + "loss": 3.3391, + "step": 2750 + }, + { + "epoch": 0.08485862679246883, + "grad_norm": 0.19803744554519653, + "learning_rate": 4.242751388032079e-05, + "loss": 3.3436, + "step": 2751 + }, + { + "epoch": 0.08488947325804225, + "grad_norm": 0.1915515661239624, + "learning_rate": 4.244293645897594e-05, + "loss": 3.3417, + "step": 2752 + }, + { + "epoch": 0.08492031972361568, + "grad_norm": 0.20768973231315613, + "learning_rate": 4.245835903763109e-05, + "loss": 3.3729, + "step": 2753 + }, + { + "epoch": 0.08495116618918909, + "grad_norm": 0.20628094673156738, + "learning_rate": 4.2473781616286246e-05, + "loss": 3.3479, + "step": 2754 + }, + { + "epoch": 0.0849820126547625, + "grad_norm": 0.20580290257930756, + "learning_rate": 4.24892041949414e-05, + "loss": 3.3586, + "step": 2755 + }, + { + "epoch": 0.08501285912033592, + "grad_norm": 0.20103509724140167, + "learning_rate": 4.250462677359654e-05, + "loss": 3.3622, + "step": 2756 + }, + { + "epoch": 0.08504370558590933, + "grad_norm": 0.20382331311702728, + "learning_rate": 4.25200493522517e-05, + "loss": 3.3111, + "step": 2757 + }, + { + "epoch": 0.08507455205148275, + "grad_norm": 0.20917150378227234, + "learning_rate": 4.253547193090685e-05, + "loss": 3.3637, + "step": 2758 + }, + { + "epoch": 0.08510539851705617, + "grad_norm": 0.19537918269634247, + "learning_rate": 4.2550894509562e-05, + "loss": 3.3496, + "step": 2759 + }, + { + "epoch": 0.08513624498262959, + "grad_norm": 0.25880640745162964, + "learning_rate": 4.256631708821715e-05, + "loss": 3.3487, + "step": 2760 + }, + { + "epoch": 0.085167091448203, + "grad_norm": 0.19837796688079834, + "learning_rate": 4.2581739666872305e-05, + "loss": 3.3706, + "step": 2761 + }, + { + "epoch": 0.08519793791377642, + "grad_norm": 0.20428059995174408, + "learning_rate": 4.259716224552746e-05, + "loss": 3.3592, + "step": 2762 + }, + { + "epoch": 0.08522878437934983, + "grad_norm": 0.20971404016017914, + "learning_rate": 4.26125848241826e-05, + "loss": 3.3551, + "step": 2763 + }, + { + "epoch": 0.08525963084492325, + "grad_norm": 0.19929170608520508, + "learning_rate": 4.2628007402837755e-05, + "loss": 3.3603, + "step": 2764 + }, + { + "epoch": 0.08529047731049667, + "grad_norm": 0.2054266482591629, + "learning_rate": 4.264342998149291e-05, + "loss": 3.354, + "step": 2765 + }, + { + "epoch": 0.08532132377607009, + "grad_norm": 0.18601316213607788, + "learning_rate": 4.265885256014806e-05, + "loss": 3.3636, + "step": 2766 + }, + { + "epoch": 0.0853521702416435, + "grad_norm": 0.19906768202781677, + "learning_rate": 4.267427513880321e-05, + "loss": 3.3875, + "step": 2767 + }, + { + "epoch": 0.08538301670721692, + "grad_norm": 0.18842346966266632, + "learning_rate": 4.268969771745836e-05, + "loss": 3.3595, + "step": 2768 + }, + { + "epoch": 0.08541386317279033, + "grad_norm": 0.2725621461868286, + "learning_rate": 4.270512029611351e-05, + "loss": 3.3593, + "step": 2769 + }, + { + "epoch": 0.08544470963836374, + "grad_norm": 0.19419445097446442, + "learning_rate": 4.272054287476867e-05, + "loss": 3.363, + "step": 2770 + }, + { + "epoch": 0.08547555610393717, + "grad_norm": 0.20476365089416504, + "learning_rate": 4.2735965453423814e-05, + "loss": 3.3639, + "step": 2771 + }, + { + "epoch": 0.08550640256951059, + "grad_norm": 0.20523402094841003, + "learning_rate": 4.2751388032078966e-05, + "loss": 3.3804, + "step": 2772 + }, + { + "epoch": 0.085537249035084, + "grad_norm": 0.2008276879787445, + "learning_rate": 4.276681061073411e-05, + "loss": 3.3373, + "step": 2773 + }, + { + "epoch": 0.08556809550065742, + "grad_norm": 0.19244664907455444, + "learning_rate": 4.278223318938927e-05, + "loss": 3.3529, + "step": 2774 + }, + { + "epoch": 0.08559894196623083, + "grad_norm": 0.1850292831659317, + "learning_rate": 4.279765576804442e-05, + "loss": 3.3059, + "step": 2775 + }, + { + "epoch": 0.08562978843180424, + "grad_norm": 0.2000182569026947, + "learning_rate": 4.281307834669957e-05, + "loss": 3.3329, + "step": 2776 + }, + { + "epoch": 0.08566063489737767, + "grad_norm": 0.1853158324956894, + "learning_rate": 4.282850092535472e-05, + "loss": 3.3031, + "step": 2777 + }, + { + "epoch": 0.08569148136295109, + "grad_norm": 0.18885692954063416, + "learning_rate": 4.2843923504009873e-05, + "loss": 3.3363, + "step": 2778 + }, + { + "epoch": 0.0857223278285245, + "grad_norm": 0.18492357432842255, + "learning_rate": 4.2859346082665026e-05, + "loss": 3.3492, + "step": 2779 + }, + { + "epoch": 0.08575317429409791, + "grad_norm": 0.18545740842819214, + "learning_rate": 4.287476866132018e-05, + "loss": 3.3528, + "step": 2780 + }, + { + "epoch": 0.08578402075967133, + "grad_norm": 0.19296813011169434, + "learning_rate": 4.2890191239975324e-05, + "loss": 3.2829, + "step": 2781 + }, + { + "epoch": 0.08581486722524474, + "grad_norm": 0.18670670688152313, + "learning_rate": 4.2905613818630476e-05, + "loss": 3.3084, + "step": 2782 + }, + { + "epoch": 0.08584571369081817, + "grad_norm": 0.19296932220458984, + "learning_rate": 4.292103639728563e-05, + "loss": 3.3682, + "step": 2783 + }, + { + "epoch": 0.08587656015639158, + "grad_norm": 0.19214306771755219, + "learning_rate": 4.293645897594078e-05, + "loss": 3.3321, + "step": 2784 + }, + { + "epoch": 0.085907406621965, + "grad_norm": 0.18648181855678558, + "learning_rate": 4.295188155459593e-05, + "loss": 3.3849, + "step": 2785 + }, + { + "epoch": 0.08593825308753841, + "grad_norm": 0.19258899986743927, + "learning_rate": 4.296730413325108e-05, + "loss": 3.3234, + "step": 2786 + }, + { + "epoch": 0.08596909955311183, + "grad_norm": 0.1796867996454239, + "learning_rate": 4.298272671190624e-05, + "loss": 3.3299, + "step": 2787 + }, + { + "epoch": 0.08599994601868524, + "grad_norm": 0.19268573820590973, + "learning_rate": 4.299814929056138e-05, + "loss": 3.3671, + "step": 2788 + }, + { + "epoch": 0.08603079248425867, + "grad_norm": 0.18392758071422577, + "learning_rate": 4.3013571869216535e-05, + "loss": 3.3552, + "step": 2789 + }, + { + "epoch": 0.08606163894983208, + "grad_norm": 0.19293926656246185, + "learning_rate": 4.302899444787169e-05, + "loss": 3.327, + "step": 2790 + }, + { + "epoch": 0.0860924854154055, + "grad_norm": 0.18796870112419128, + "learning_rate": 4.304441702652684e-05, + "loss": 3.3141, + "step": 2791 + }, + { + "epoch": 0.08612333188097891, + "grad_norm": 0.19382908940315247, + "learning_rate": 4.305983960518199e-05, + "loss": 3.3287, + "step": 2792 + }, + { + "epoch": 0.08615417834655233, + "grad_norm": 0.26595741510391235, + "learning_rate": 4.307526218383714e-05, + "loss": 3.3231, + "step": 2793 + }, + { + "epoch": 0.08618502481212574, + "grad_norm": 0.1912512481212616, + "learning_rate": 4.309068476249229e-05, + "loss": 3.3454, + "step": 2794 + }, + { + "epoch": 0.08621587127769917, + "grad_norm": 0.21578483283519745, + "learning_rate": 4.310610734114744e-05, + "loss": 3.3496, + "step": 2795 + }, + { + "epoch": 0.08624671774327258, + "grad_norm": 0.22776292264461517, + "learning_rate": 4.3121529919802594e-05, + "loss": 3.3639, + "step": 2796 + }, + { + "epoch": 0.086277564208846, + "grad_norm": 0.19329920411109924, + "learning_rate": 4.313695249845775e-05, + "loss": 3.3542, + "step": 2797 + }, + { + "epoch": 0.08630841067441941, + "grad_norm": 0.21114347875118256, + "learning_rate": 4.315237507711289e-05, + "loss": 3.3078, + "step": 2798 + }, + { + "epoch": 0.08633925713999283, + "grad_norm": 0.1825011670589447, + "learning_rate": 4.3167797655768045e-05, + "loss": 3.33, + "step": 2799 + }, + { + "epoch": 0.08637010360556624, + "grad_norm": 0.23104767501354218, + "learning_rate": 4.31832202344232e-05, + "loss": 3.3202, + "step": 2800 + }, + { + "epoch": 0.08640095007113967, + "grad_norm": 0.20440798997879028, + "learning_rate": 4.319864281307835e-05, + "loss": 3.3445, + "step": 2801 + }, + { + "epoch": 0.08643179653671308, + "grad_norm": 0.20943528413772583, + "learning_rate": 4.32140653917335e-05, + "loss": 3.3617, + "step": 2802 + }, + { + "epoch": 0.0864626430022865, + "grad_norm": 0.2250903695821762, + "learning_rate": 4.322948797038865e-05, + "loss": 3.3498, + "step": 2803 + }, + { + "epoch": 0.08649348946785991, + "grad_norm": 0.19094592332839966, + "learning_rate": 4.3244910549043806e-05, + "loss": 3.3553, + "step": 2804 + }, + { + "epoch": 0.08652433593343332, + "grad_norm": 0.20880888402462006, + "learning_rate": 4.326033312769895e-05, + "loss": 3.3398, + "step": 2805 + }, + { + "epoch": 0.08655518239900674, + "grad_norm": 0.18832914531230927, + "learning_rate": 4.3275755706354104e-05, + "loss": 3.3337, + "step": 2806 + }, + { + "epoch": 0.08658602886458017, + "grad_norm": 0.20854760706424713, + "learning_rate": 4.3291178285009256e-05, + "loss": 3.3146, + "step": 2807 + }, + { + "epoch": 0.08661687533015358, + "grad_norm": 0.2370426207780838, + "learning_rate": 4.330660086366441e-05, + "loss": 3.364, + "step": 2808 + }, + { + "epoch": 0.086647721795727, + "grad_norm": 0.18580815196037292, + "learning_rate": 4.332202344231956e-05, + "loss": 3.3611, + "step": 2809 + }, + { + "epoch": 0.08667856826130041, + "grad_norm": 0.2364961951971054, + "learning_rate": 4.3337446020974706e-05, + "loss": 3.3442, + "step": 2810 + }, + { + "epoch": 0.08670941472687382, + "grad_norm": 0.192682683467865, + "learning_rate": 4.335286859962986e-05, + "loss": 3.3414, + "step": 2811 + }, + { + "epoch": 0.08674026119244724, + "grad_norm": 0.2041468769311905, + "learning_rate": 4.336829117828501e-05, + "loss": 3.3442, + "step": 2812 + }, + { + "epoch": 0.08677110765802067, + "grad_norm": 0.1970762014389038, + "learning_rate": 4.338371375694016e-05, + "loss": 3.3708, + "step": 2813 + }, + { + "epoch": 0.08680195412359408, + "grad_norm": 0.19378793239593506, + "learning_rate": 4.3399136335595315e-05, + "loss": 3.3425, + "step": 2814 + }, + { + "epoch": 0.0868328005891675, + "grad_norm": 0.21640771627426147, + "learning_rate": 4.341455891425046e-05, + "loss": 3.3561, + "step": 2815 + }, + { + "epoch": 0.08686364705474091, + "grad_norm": 0.18924130499362946, + "learning_rate": 4.342998149290561e-05, + "loss": 3.3407, + "step": 2816 + }, + { + "epoch": 0.08689449352031432, + "grad_norm": 0.20679333806037903, + "learning_rate": 4.344540407156077e-05, + "loss": 3.3366, + "step": 2817 + }, + { + "epoch": 0.08692533998588774, + "grad_norm": 0.19670429825782776, + "learning_rate": 4.346082665021592e-05, + "loss": 3.3065, + "step": 2818 + }, + { + "epoch": 0.08695618645146116, + "grad_norm": 0.18830956518650055, + "learning_rate": 4.347624922887107e-05, + "loss": 3.3506, + "step": 2819 + }, + { + "epoch": 0.08698703291703458, + "grad_norm": 0.21200081706047058, + "learning_rate": 4.3491671807526216e-05, + "loss": 3.3037, + "step": 2820 + }, + { + "epoch": 0.087017879382608, + "grad_norm": 0.20346686244010925, + "learning_rate": 4.3507094386181375e-05, + "loss": 3.3728, + "step": 2821 + }, + { + "epoch": 0.08704872584818141, + "grad_norm": 0.2163609117269516, + "learning_rate": 4.352251696483653e-05, + "loss": 3.3562, + "step": 2822 + }, + { + "epoch": 0.08707957231375482, + "grad_norm": 0.2176007628440857, + "learning_rate": 4.353793954349167e-05, + "loss": 3.3383, + "step": 2823 + }, + { + "epoch": 0.08711041877932824, + "grad_norm": 0.1972505748271942, + "learning_rate": 4.3553362122146825e-05, + "loss": 3.346, + "step": 2824 + }, + { + "epoch": 0.08714126524490166, + "grad_norm": 0.20174719393253326, + "learning_rate": 4.356878470080198e-05, + "loss": 3.3375, + "step": 2825 + }, + { + "epoch": 0.08717211171047508, + "grad_norm": 0.44402673840522766, + "learning_rate": 4.358420727945713e-05, + "loss": 3.3171, + "step": 2826 + }, + { + "epoch": 0.08720295817604849, + "grad_norm": 0.21038414537906647, + "learning_rate": 4.359962985811228e-05, + "loss": 3.3076, + "step": 2827 + }, + { + "epoch": 0.0872338046416219, + "grad_norm": 0.197340726852417, + "learning_rate": 4.361505243676743e-05, + "loss": 3.3434, + "step": 2828 + }, + { + "epoch": 0.08726465110719532, + "grad_norm": 0.1961524486541748, + "learning_rate": 4.363047501542258e-05, + "loss": 3.3138, + "step": 2829 + }, + { + "epoch": 0.08729549757276873, + "grad_norm": 0.2162128984928131, + "learning_rate": 4.364589759407773e-05, + "loss": 3.3521, + "step": 2830 + }, + { + "epoch": 0.08732634403834216, + "grad_norm": 0.2022390067577362, + "learning_rate": 4.3661320172732884e-05, + "loss": 3.3353, + "step": 2831 + }, + { + "epoch": 0.08735719050391558, + "grad_norm": 0.20434404909610748, + "learning_rate": 4.3676742751388036e-05, + "loss": 3.3298, + "step": 2832 + }, + { + "epoch": 0.08738803696948899, + "grad_norm": 0.21595284342765808, + "learning_rate": 4.369216533004318e-05, + "loss": 3.3701, + "step": 2833 + }, + { + "epoch": 0.0874188834350624, + "grad_norm": 0.2110062539577484, + "learning_rate": 4.370758790869834e-05, + "loss": 3.3039, + "step": 2834 + }, + { + "epoch": 0.08744972990063582, + "grad_norm": 0.24639087915420532, + "learning_rate": 4.3723010487353486e-05, + "loss": 3.3776, + "step": 2835 + }, + { + "epoch": 0.08748057636620923, + "grad_norm": 0.21353690326213837, + "learning_rate": 4.373843306600864e-05, + "loss": 3.3372, + "step": 2836 + }, + { + "epoch": 0.08751142283178266, + "grad_norm": 0.21406129002571106, + "learning_rate": 4.375385564466379e-05, + "loss": 3.3522, + "step": 2837 + }, + { + "epoch": 0.08754226929735608, + "grad_norm": 0.22999198734760284, + "learning_rate": 4.376927822331894e-05, + "loss": 3.3344, + "step": 2838 + }, + { + "epoch": 0.08757311576292949, + "grad_norm": 0.19109223783016205, + "learning_rate": 4.3784700801974096e-05, + "loss": 3.3157, + "step": 2839 + }, + { + "epoch": 0.0876039622285029, + "grad_norm": 0.20954139530658722, + "learning_rate": 4.380012338062924e-05, + "loss": 3.3029, + "step": 2840 + }, + { + "epoch": 0.08763480869407632, + "grad_norm": 0.21779035031795502, + "learning_rate": 4.3815545959284393e-05, + "loss": 3.3773, + "step": 2841 + }, + { + "epoch": 0.08766565515964973, + "grad_norm": 0.19796329736709595, + "learning_rate": 4.3830968537939546e-05, + "loss": 3.3575, + "step": 2842 + }, + { + "epoch": 0.08769650162522316, + "grad_norm": 0.21369200944900513, + "learning_rate": 4.38463911165947e-05, + "loss": 3.3356, + "step": 2843 + }, + { + "epoch": 0.08772734809079658, + "grad_norm": 0.1900564432144165, + "learning_rate": 4.386181369524985e-05, + "loss": 3.3424, + "step": 2844 + }, + { + "epoch": 0.08775819455636999, + "grad_norm": 0.2024858444929123, + "learning_rate": 4.3877236273904996e-05, + "loss": 3.3452, + "step": 2845 + }, + { + "epoch": 0.0877890410219434, + "grad_norm": 0.4500417113304138, + "learning_rate": 4.389265885256015e-05, + "loss": 3.3135, + "step": 2846 + }, + { + "epoch": 0.08781988748751682, + "grad_norm": 0.23754726350307465, + "learning_rate": 4.39080814312153e-05, + "loss": 3.3451, + "step": 2847 + }, + { + "epoch": 0.08785073395309023, + "grad_norm": 0.2164919525384903, + "learning_rate": 4.392350400987045e-05, + "loss": 3.367, + "step": 2848 + }, + { + "epoch": 0.08788158041866366, + "grad_norm": 0.20863096415996552, + "learning_rate": 4.3938926588525605e-05, + "loss": 3.33, + "step": 2849 + }, + { + "epoch": 0.08791242688423707, + "grad_norm": 0.22313566505908966, + "learning_rate": 4.395434916718075e-05, + "loss": 3.3312, + "step": 2850 + }, + { + "epoch": 0.08794327334981049, + "grad_norm": 0.2269708514213562, + "learning_rate": 4.396977174583591e-05, + "loss": 3.3332, + "step": 2851 + }, + { + "epoch": 0.0879741198153839, + "grad_norm": 0.20381459593772888, + "learning_rate": 4.3985194324491055e-05, + "loss": 3.3361, + "step": 2852 + }, + { + "epoch": 0.08800496628095732, + "grad_norm": 0.19764374196529388, + "learning_rate": 4.400061690314621e-05, + "loss": 3.3557, + "step": 2853 + }, + { + "epoch": 0.08803581274653073, + "grad_norm": 0.2319285124540329, + "learning_rate": 4.401603948180136e-05, + "loss": 3.3597, + "step": 2854 + }, + { + "epoch": 0.08806665921210416, + "grad_norm": 0.20340394973754883, + "learning_rate": 4.403146206045651e-05, + "loss": 3.3524, + "step": 2855 + }, + { + "epoch": 0.08809750567767757, + "grad_norm": 0.7850560545921326, + "learning_rate": 4.4046884639111664e-05, + "loss": 3.3041, + "step": 2856 + }, + { + "epoch": 0.08812835214325099, + "grad_norm": 0.2080986499786377, + "learning_rate": 4.406230721776681e-05, + "loss": 3.3416, + "step": 2857 + }, + { + "epoch": 0.0881591986088244, + "grad_norm": 0.2954613268375397, + "learning_rate": 4.407772979642196e-05, + "loss": 3.3517, + "step": 2858 + }, + { + "epoch": 0.08819004507439782, + "grad_norm": 0.217209592461586, + "learning_rate": 4.4093152375077114e-05, + "loss": 3.348, + "step": 2859 + }, + { + "epoch": 0.08822089153997123, + "grad_norm": 0.20990023016929626, + "learning_rate": 4.410857495373227e-05, + "loss": 3.3346, + "step": 2860 + }, + { + "epoch": 0.08825173800554466, + "grad_norm": 0.2330152839422226, + "learning_rate": 4.412399753238742e-05, + "loss": 3.32, + "step": 2861 + }, + { + "epoch": 0.08828258447111807, + "grad_norm": 0.4731096923351288, + "learning_rate": 4.4139420111042564e-05, + "loss": 3.3123, + "step": 2862 + }, + { + "epoch": 0.08831343093669149, + "grad_norm": 0.20732606947422028, + "learning_rate": 4.415484268969772e-05, + "loss": 3.3372, + "step": 2863 + }, + { + "epoch": 0.0883442774022649, + "grad_norm": 0.2092141956090927, + "learning_rate": 4.4170265268352876e-05, + "loss": 3.3485, + "step": 2864 + }, + { + "epoch": 0.08837512386783831, + "grad_norm": 0.20797084271907806, + "learning_rate": 4.418568784700802e-05, + "loss": 3.3372, + "step": 2865 + }, + { + "epoch": 0.08840597033341173, + "grad_norm": 0.20797227323055267, + "learning_rate": 4.4201110425663174e-05, + "loss": 3.3673, + "step": 2866 + }, + { + "epoch": 0.08843681679898516, + "grad_norm": 0.32398197054862976, + "learning_rate": 4.421653300431832e-05, + "loss": 3.3527, + "step": 2867 + }, + { + "epoch": 0.08846766326455857, + "grad_norm": 0.1908559650182724, + "learning_rate": 4.423195558297348e-05, + "loss": 3.3476, + "step": 2868 + }, + { + "epoch": 0.08849850973013199, + "grad_norm": 0.32882335782051086, + "learning_rate": 4.424737816162863e-05, + "loss": 3.3071, + "step": 2869 + }, + { + "epoch": 0.0885293561957054, + "grad_norm": 0.2124207317829132, + "learning_rate": 4.4262800740283776e-05, + "loss": 3.3777, + "step": 2870 + }, + { + "epoch": 0.08856020266127881, + "grad_norm": 0.20464357733726501, + "learning_rate": 4.427822331893893e-05, + "loss": 3.3522, + "step": 2871 + }, + { + "epoch": 0.08859104912685223, + "grad_norm": 0.19471187889575958, + "learning_rate": 4.4293645897594074e-05, + "loss": 3.3478, + "step": 2872 + }, + { + "epoch": 0.08862189559242566, + "grad_norm": 0.4823291301727295, + "learning_rate": 4.430906847624923e-05, + "loss": 3.3441, + "step": 2873 + }, + { + "epoch": 0.08865274205799907, + "grad_norm": 0.21979153156280518, + "learning_rate": 4.4324491054904385e-05, + "loss": 3.3553, + "step": 2874 + }, + { + "epoch": 0.08868358852357248, + "grad_norm": 0.2141953706741333, + "learning_rate": 4.433991363355953e-05, + "loss": 3.3418, + "step": 2875 + }, + { + "epoch": 0.0887144349891459, + "grad_norm": 0.21636809408664703, + "learning_rate": 4.435533621221468e-05, + "loss": 3.3371, + "step": 2876 + }, + { + "epoch": 0.08874528145471931, + "grad_norm": 0.19768057763576508, + "learning_rate": 4.4370758790869835e-05, + "loss": 3.3726, + "step": 2877 + }, + { + "epoch": 0.08877612792029273, + "grad_norm": 0.220245361328125, + "learning_rate": 4.438618136952499e-05, + "loss": 3.3596, + "step": 2878 + }, + { + "epoch": 0.08880697438586616, + "grad_norm": 0.20556168258190155, + "learning_rate": 4.440160394818014e-05, + "loss": 3.3294, + "step": 2879 + }, + { + "epoch": 0.08883782085143957, + "grad_norm": 0.19977232813835144, + "learning_rate": 4.4417026526835285e-05, + "loss": 3.3265, + "step": 2880 + }, + { + "epoch": 0.08886866731701298, + "grad_norm": 0.22520457208156586, + "learning_rate": 4.4432449105490445e-05, + "loss": 3.3182, + "step": 2881 + }, + { + "epoch": 0.0888995137825864, + "grad_norm": 0.21174365282058716, + "learning_rate": 4.444787168414559e-05, + "loss": 3.3105, + "step": 2882 + }, + { + "epoch": 0.08893036024815981, + "grad_norm": 0.29617831110954285, + "learning_rate": 4.446329426280074e-05, + "loss": 3.3672, + "step": 2883 + }, + { + "epoch": 0.08896120671373323, + "grad_norm": 0.20254312455654144, + "learning_rate": 4.4478716841455895e-05, + "loss": 3.3755, + "step": 2884 + }, + { + "epoch": 0.08899205317930665, + "grad_norm": 0.2055642306804657, + "learning_rate": 4.449413942011105e-05, + "loss": 3.3693, + "step": 2885 + }, + { + "epoch": 0.08902289964488007, + "grad_norm": 0.2071792483329773, + "learning_rate": 4.45095619987662e-05, + "loss": 3.3847, + "step": 2886 + }, + { + "epoch": 0.08905374611045348, + "grad_norm": 0.19736438989639282, + "learning_rate": 4.4524984577421345e-05, + "loss": 3.3398, + "step": 2887 + }, + { + "epoch": 0.0890845925760269, + "grad_norm": 0.20706476271152496, + "learning_rate": 4.45404071560765e-05, + "loss": 3.3635, + "step": 2888 + }, + { + "epoch": 0.08911543904160031, + "grad_norm": 0.31341680884361267, + "learning_rate": 4.455582973473165e-05, + "loss": 3.3491, + "step": 2889 + }, + { + "epoch": 0.08914628550717373, + "grad_norm": 0.20538076758384705, + "learning_rate": 4.45712523133868e-05, + "loss": 3.3211, + "step": 2890 + }, + { + "epoch": 0.08917713197274715, + "grad_norm": 0.20398926734924316, + "learning_rate": 4.4586674892041954e-05, + "loss": 3.3447, + "step": 2891 + }, + { + "epoch": 0.08920797843832057, + "grad_norm": 0.21009023487567902, + "learning_rate": 4.46020974706971e-05, + "loss": 3.3416, + "step": 2892 + }, + { + "epoch": 0.08923882490389398, + "grad_norm": 0.19358469545841217, + "learning_rate": 4.461752004935225e-05, + "loss": 3.3435, + "step": 2893 + }, + { + "epoch": 0.0892696713694674, + "grad_norm": 0.200425386428833, + "learning_rate": 4.4632942628007404e-05, + "loss": 3.3274, + "step": 2894 + }, + { + "epoch": 0.08930051783504081, + "grad_norm": 0.20847219228744507, + "learning_rate": 4.4648365206662556e-05, + "loss": 3.3039, + "step": 2895 + }, + { + "epoch": 0.08933136430061422, + "grad_norm": 0.19111108779907227, + "learning_rate": 4.466378778531771e-05, + "loss": 3.3075, + "step": 2896 + }, + { + "epoch": 0.08936221076618765, + "grad_norm": 0.18883879482746124, + "learning_rate": 4.4679210363972854e-05, + "loss": 3.3348, + "step": 2897 + }, + { + "epoch": 0.08939305723176107, + "grad_norm": 0.1907830685377121, + "learning_rate": 4.469463294262801e-05, + "loss": 3.3531, + "step": 2898 + }, + { + "epoch": 0.08942390369733448, + "grad_norm": 0.18545696139335632, + "learning_rate": 4.471005552128316e-05, + "loss": 3.3402, + "step": 2899 + }, + { + "epoch": 0.0894547501629079, + "grad_norm": 0.7598090171813965, + "learning_rate": 4.472547809993831e-05, + "loss": 3.3539, + "step": 2900 + }, + { + "epoch": 0.08948559662848131, + "grad_norm": 0.19593116641044617, + "learning_rate": 4.474090067859346e-05, + "loss": 3.3377, + "step": 2901 + }, + { + "epoch": 0.08951644309405472, + "grad_norm": 0.19318519532680511, + "learning_rate": 4.4756323257248616e-05, + "loss": 3.3336, + "step": 2902 + }, + { + "epoch": 0.08954728955962815, + "grad_norm": 0.2076471745967865, + "learning_rate": 4.477174583590377e-05, + "loss": 3.3594, + "step": 2903 + }, + { + "epoch": 0.08957813602520157, + "grad_norm": 0.19760699570178986, + "learning_rate": 4.478716841455891e-05, + "loss": 3.3522, + "step": 2904 + }, + { + "epoch": 0.08960898249077498, + "grad_norm": 0.19716069102287292, + "learning_rate": 4.4802590993214066e-05, + "loss": 3.3447, + "step": 2905 + }, + { + "epoch": 0.0896398289563484, + "grad_norm": 0.19863244891166687, + "learning_rate": 4.481801357186922e-05, + "loss": 3.379, + "step": 2906 + }, + { + "epoch": 0.08967067542192181, + "grad_norm": 0.3060222566127777, + "learning_rate": 4.483343615052437e-05, + "loss": 3.3526, + "step": 2907 + }, + { + "epoch": 0.08970152188749522, + "grad_norm": 0.23228976130485535, + "learning_rate": 4.484885872917952e-05, + "loss": 3.364, + "step": 2908 + }, + { + "epoch": 0.08973236835306865, + "grad_norm": 0.1998637318611145, + "learning_rate": 4.486428130783467e-05, + "loss": 3.3556, + "step": 2909 + }, + { + "epoch": 0.08976321481864206, + "grad_norm": 0.2080475091934204, + "learning_rate": 4.487970388648982e-05, + "loss": 3.3555, + "step": 2910 + }, + { + "epoch": 0.08979406128421548, + "grad_norm": 0.19170713424682617, + "learning_rate": 4.489512646514498e-05, + "loss": 3.31, + "step": 2911 + }, + { + "epoch": 0.0898249077497889, + "grad_norm": 0.1989683359861374, + "learning_rate": 4.4910549043800125e-05, + "loss": 3.3521, + "step": 2912 + }, + { + "epoch": 0.08985575421536231, + "grad_norm": 0.19802476465702057, + "learning_rate": 4.492597162245528e-05, + "loss": 3.3289, + "step": 2913 + }, + { + "epoch": 0.08988660068093572, + "grad_norm": 0.18914775550365448, + "learning_rate": 4.494139420111042e-05, + "loss": 3.355, + "step": 2914 + }, + { + "epoch": 0.08991744714650915, + "grad_norm": 0.20064964890480042, + "learning_rate": 4.495681677976558e-05, + "loss": 3.3348, + "step": 2915 + }, + { + "epoch": 0.08994829361208256, + "grad_norm": 0.19386447966098785, + "learning_rate": 4.4972239358420734e-05, + "loss": 3.3621, + "step": 2916 + }, + { + "epoch": 0.08997914007765598, + "grad_norm": 0.1980847418308258, + "learning_rate": 4.498766193707588e-05, + "loss": 3.3309, + "step": 2917 + }, + { + "epoch": 0.09000998654322939, + "grad_norm": 0.19557712972164154, + "learning_rate": 4.500308451573103e-05, + "loss": 3.3693, + "step": 2918 + }, + { + "epoch": 0.0900408330088028, + "grad_norm": 0.21206030249595642, + "learning_rate": 4.5018507094386184e-05, + "loss": 3.3339, + "step": 2919 + }, + { + "epoch": 0.09007167947437622, + "grad_norm": 0.19607120752334595, + "learning_rate": 4.5033929673041337e-05, + "loss": 3.3449, + "step": 2920 + }, + { + "epoch": 0.09010252593994965, + "grad_norm": 0.20223961770534515, + "learning_rate": 4.504935225169649e-05, + "loss": 3.3689, + "step": 2921 + }, + { + "epoch": 0.09013337240552306, + "grad_norm": 0.18960212171077728, + "learning_rate": 4.5064774830351634e-05, + "loss": 3.3332, + "step": 2922 + }, + { + "epoch": 0.09016421887109648, + "grad_norm": 0.20983242988586426, + "learning_rate": 4.508019740900679e-05, + "loss": 3.3459, + "step": 2923 + }, + { + "epoch": 0.09019506533666989, + "grad_norm": 0.20799261331558228, + "learning_rate": 4.509561998766194e-05, + "loss": 3.3285, + "step": 2924 + }, + { + "epoch": 0.0902259118022433, + "grad_norm": 0.20191559195518494, + "learning_rate": 4.511104256631709e-05, + "loss": 3.3199, + "step": 2925 + }, + { + "epoch": 0.09025675826781672, + "grad_norm": 0.19653718173503876, + "learning_rate": 4.5126465144972244e-05, + "loss": 3.3243, + "step": 2926 + }, + { + "epoch": 0.09028760473339015, + "grad_norm": 0.19786091148853302, + "learning_rate": 4.514188772362739e-05, + "loss": 3.3703, + "step": 2927 + }, + { + "epoch": 0.09031845119896356, + "grad_norm": 0.18781918287277222, + "learning_rate": 4.515731030228255e-05, + "loss": 3.3165, + "step": 2928 + }, + { + "epoch": 0.09034929766453698, + "grad_norm": 0.19750717282295227, + "learning_rate": 4.5172732880937694e-05, + "loss": 3.3624, + "step": 2929 + }, + { + "epoch": 0.09038014413011039, + "grad_norm": 0.19398446381092072, + "learning_rate": 4.5188155459592846e-05, + "loss": 3.3416, + "step": 2930 + }, + { + "epoch": 0.0904109905956838, + "grad_norm": 0.21179702877998352, + "learning_rate": 4.5203578038248e-05, + "loss": 3.3478, + "step": 2931 + }, + { + "epoch": 0.09044183706125722, + "grad_norm": 0.225688636302948, + "learning_rate": 4.521900061690315e-05, + "loss": 3.3691, + "step": 2932 + }, + { + "epoch": 0.09047268352683065, + "grad_norm": 0.1899472326040268, + "learning_rate": 4.52344231955583e-05, + "loss": 3.3256, + "step": 2933 + }, + { + "epoch": 0.09050352999240406, + "grad_norm": 0.21146425604820251, + "learning_rate": 4.524984577421345e-05, + "loss": 3.3435, + "step": 2934 + }, + { + "epoch": 0.09053437645797748, + "grad_norm": 0.1919640749692917, + "learning_rate": 4.52652683528686e-05, + "loss": 3.303, + "step": 2935 + }, + { + "epoch": 0.09056522292355089, + "grad_norm": 0.21584808826446533, + "learning_rate": 4.528069093152375e-05, + "loss": 3.3569, + "step": 2936 + }, + { + "epoch": 0.0905960693891243, + "grad_norm": 0.18293891847133636, + "learning_rate": 4.5296113510178905e-05, + "loss": 3.3156, + "step": 2937 + }, + { + "epoch": 0.09062691585469772, + "grad_norm": 0.2147713452577591, + "learning_rate": 4.531153608883406e-05, + "loss": 3.3631, + "step": 2938 + }, + { + "epoch": 0.09065776232027115, + "grad_norm": 0.19920262694358826, + "learning_rate": 4.53269586674892e-05, + "loss": 3.328, + "step": 2939 + }, + { + "epoch": 0.09068860878584456, + "grad_norm": 0.23202724754810333, + "learning_rate": 4.5342381246144355e-05, + "loss": 3.3686, + "step": 2940 + }, + { + "epoch": 0.09071945525141797, + "grad_norm": 0.23045672476291656, + "learning_rate": 4.535780382479951e-05, + "loss": 3.3259, + "step": 2941 + }, + { + "epoch": 0.09075030171699139, + "grad_norm": 0.1940208226442337, + "learning_rate": 4.537322640345466e-05, + "loss": 3.3352, + "step": 2942 + }, + { + "epoch": 0.0907811481825648, + "grad_norm": 0.23026016354560852, + "learning_rate": 4.538864898210981e-05, + "loss": 3.3275, + "step": 2943 + }, + { + "epoch": 0.09081199464813822, + "grad_norm": 0.19774900376796722, + "learning_rate": 4.540407156076496e-05, + "loss": 3.3261, + "step": 2944 + }, + { + "epoch": 0.09084284111371164, + "grad_norm": 0.2146570384502411, + "learning_rate": 4.541949413942012e-05, + "loss": 3.3473, + "step": 2945 + }, + { + "epoch": 0.09087368757928506, + "grad_norm": 0.20855890214443207, + "learning_rate": 4.543491671807526e-05, + "loss": 3.3568, + "step": 2946 + }, + { + "epoch": 0.09090453404485847, + "grad_norm": 0.19810424745082855, + "learning_rate": 4.5450339296730415e-05, + "loss": 3.3497, + "step": 2947 + }, + { + "epoch": 0.09093538051043189, + "grad_norm": 0.21047735214233398, + "learning_rate": 4.546576187538557e-05, + "loss": 3.3518, + "step": 2948 + }, + { + "epoch": 0.0909662269760053, + "grad_norm": 0.1985594630241394, + "learning_rate": 4.548118445404071e-05, + "loss": 3.3516, + "step": 2949 + }, + { + "epoch": 0.09099707344157872, + "grad_norm": 0.21211275458335876, + "learning_rate": 4.549660703269587e-05, + "loss": 3.3123, + "step": 2950 + }, + { + "epoch": 0.09102791990715214, + "grad_norm": 0.20532649755477905, + "learning_rate": 4.551202961135102e-05, + "loss": 3.3082, + "step": 2951 + }, + { + "epoch": 0.09105876637272556, + "grad_norm": 0.20332859456539154, + "learning_rate": 4.552745219000617e-05, + "loss": 3.3107, + "step": 2952 + }, + { + "epoch": 0.09108961283829897, + "grad_norm": 0.276006817817688, + "learning_rate": 4.554287476866132e-05, + "loss": 3.3313, + "step": 2953 + }, + { + "epoch": 0.09112045930387239, + "grad_norm": 0.19614242017269135, + "learning_rate": 4.5558297347316474e-05, + "loss": 3.3202, + "step": 2954 + }, + { + "epoch": 0.0911513057694458, + "grad_norm": 0.1924978345632553, + "learning_rate": 4.5573719925971626e-05, + "loss": 3.3379, + "step": 2955 + }, + { + "epoch": 0.09118215223501921, + "grad_norm": 0.20211930572986603, + "learning_rate": 4.558914250462678e-05, + "loss": 3.3726, + "step": 2956 + }, + { + "epoch": 0.09121299870059264, + "grad_norm": 0.19706374406814575, + "learning_rate": 4.5604565083281924e-05, + "loss": 3.3337, + "step": 2957 + }, + { + "epoch": 0.09124384516616606, + "grad_norm": 0.2629600763320923, + "learning_rate": 4.561998766193708e-05, + "loss": 3.3255, + "step": 2958 + }, + { + "epoch": 0.09127469163173947, + "grad_norm": 0.18834328651428223, + "learning_rate": 4.563541024059223e-05, + "loss": 3.3345, + "step": 2959 + }, + { + "epoch": 0.09130553809731289, + "grad_norm": 0.18925586342811584, + "learning_rate": 4.565083281924738e-05, + "loss": 3.3268, + "step": 2960 + }, + { + "epoch": 0.0913363845628863, + "grad_norm": 0.19260460138320923, + "learning_rate": 4.566625539790253e-05, + "loss": 3.3665, + "step": 2961 + }, + { + "epoch": 0.09136723102845971, + "grad_norm": 0.185893252491951, + "learning_rate": 4.5681677976557685e-05, + "loss": 3.3143, + "step": 2962 + }, + { + "epoch": 0.09139807749403314, + "grad_norm": 0.19040271639823914, + "learning_rate": 4.569710055521284e-05, + "loss": 3.3195, + "step": 2963 + }, + { + "epoch": 0.09142892395960656, + "grad_norm": 0.20019884407520294, + "learning_rate": 4.571252313386798e-05, + "loss": 3.3326, + "step": 2964 + }, + { + "epoch": 0.09145977042517997, + "grad_norm": 0.1922605335712433, + "learning_rate": 4.5727945712523136e-05, + "loss": 3.3434, + "step": 2965 + }, + { + "epoch": 0.09149061689075338, + "grad_norm": 0.19210003316402435, + "learning_rate": 4.574336829117829e-05, + "loss": 3.3194, + "step": 2966 + }, + { + "epoch": 0.0915214633563268, + "grad_norm": 0.18562525510787964, + "learning_rate": 4.575879086983344e-05, + "loss": 3.3549, + "step": 2967 + }, + { + "epoch": 0.09155230982190021, + "grad_norm": 0.1933554857969284, + "learning_rate": 4.577421344848859e-05, + "loss": 3.328, + "step": 2968 + }, + { + "epoch": 0.09158315628747364, + "grad_norm": 0.17766843736171722, + "learning_rate": 4.578963602714374e-05, + "loss": 3.3338, + "step": 2969 + }, + { + "epoch": 0.09161400275304706, + "grad_norm": 0.18533197045326233, + "learning_rate": 4.580505860579889e-05, + "loss": 3.3381, + "step": 2970 + }, + { + "epoch": 0.09164484921862047, + "grad_norm": 0.1885988861322403, + "learning_rate": 4.582048118445404e-05, + "loss": 3.3316, + "step": 2971 + }, + { + "epoch": 0.09167569568419388, + "grad_norm": 0.18882116675376892, + "learning_rate": 4.5835903763109195e-05, + "loss": 3.3442, + "step": 2972 + }, + { + "epoch": 0.0917065421497673, + "grad_norm": 0.18418645858764648, + "learning_rate": 4.585132634176435e-05, + "loss": 3.3477, + "step": 2973 + }, + { + "epoch": 0.09173738861534071, + "grad_norm": 0.20199550688266754, + "learning_rate": 4.586674892041949e-05, + "loss": 3.3375, + "step": 2974 + }, + { + "epoch": 0.09176823508091414, + "grad_norm": 0.21880711615085602, + "learning_rate": 4.588217149907465e-05, + "loss": 3.3191, + "step": 2975 + }, + { + "epoch": 0.09179908154648755, + "grad_norm": 0.18784774839878082, + "learning_rate": 4.58975940777298e-05, + "loss": 3.3493, + "step": 2976 + }, + { + "epoch": 0.09182992801206097, + "grad_norm": 0.32304131984710693, + "learning_rate": 4.591301665638495e-05, + "loss": 3.3385, + "step": 2977 + }, + { + "epoch": 0.09186077447763438, + "grad_norm": 0.18569597601890564, + "learning_rate": 4.59284392350401e-05, + "loss": 3.3584, + "step": 2978 + }, + { + "epoch": 0.0918916209432078, + "grad_norm": 0.19400227069854736, + "learning_rate": 4.5943861813695254e-05, + "loss": 3.3102, + "step": 2979 + }, + { + "epoch": 0.09192246740878121, + "grad_norm": 0.18191656470298767, + "learning_rate": 4.5959284392350406e-05, + "loss": 3.3346, + "step": 2980 + }, + { + "epoch": 0.09195331387435464, + "grad_norm": 0.20629017055034637, + "learning_rate": 4.597470697100555e-05, + "loss": 3.3305, + "step": 2981 + }, + { + "epoch": 0.09198416033992805, + "grad_norm": 0.18220998346805573, + "learning_rate": 4.5990129549660704e-05, + "loss": 3.3605, + "step": 2982 + }, + { + "epoch": 0.09201500680550147, + "grad_norm": 0.1854933649301529, + "learning_rate": 4.6005552128315856e-05, + "loss": 3.3556, + "step": 2983 + }, + { + "epoch": 0.09204585327107488, + "grad_norm": 0.18353435397148132, + "learning_rate": 4.602097470697101e-05, + "loss": 3.3316, + "step": 2984 + }, + { + "epoch": 0.0920766997366483, + "grad_norm": 0.1942308098077774, + "learning_rate": 4.603639728562616e-05, + "loss": 3.3346, + "step": 2985 + }, + { + "epoch": 0.09210754620222171, + "grad_norm": 0.1804037243127823, + "learning_rate": 4.6051819864281307e-05, + "loss": 3.3553, + "step": 2986 + }, + { + "epoch": 0.09213839266779514, + "grad_norm": 0.18673424422740936, + "learning_rate": 4.606724244293646e-05, + "loss": 3.3733, + "step": 2987 + }, + { + "epoch": 0.09216923913336855, + "grad_norm": 0.20638056099414825, + "learning_rate": 4.608266502159161e-05, + "loss": 3.3382, + "step": 2988 + }, + { + "epoch": 0.09220008559894197, + "grad_norm": 0.2144937962293625, + "learning_rate": 4.6098087600246763e-05, + "loss": 3.3312, + "step": 2989 + }, + { + "epoch": 0.09223093206451538, + "grad_norm": 0.2046356052160263, + "learning_rate": 4.6113510178901916e-05, + "loss": 3.3491, + "step": 2990 + }, + { + "epoch": 0.0922617785300888, + "grad_norm": 0.1894908845424652, + "learning_rate": 4.612893275755706e-05, + "loss": 3.3508, + "step": 2991 + }, + { + "epoch": 0.09229262499566221, + "grad_norm": 0.19151093065738678, + "learning_rate": 4.614435533621222e-05, + "loss": 3.3448, + "step": 2992 + }, + { + "epoch": 0.09232347146123564, + "grad_norm": 0.4084566831588745, + "learning_rate": 4.6159777914867366e-05, + "loss": 3.3284, + "step": 2993 + }, + { + "epoch": 0.09235431792680905, + "grad_norm": 0.18614082038402557, + "learning_rate": 4.617520049352252e-05, + "loss": 3.3427, + "step": 2994 + }, + { + "epoch": 0.09238516439238247, + "grad_norm": 0.1809227466583252, + "learning_rate": 4.619062307217767e-05, + "loss": 3.3321, + "step": 2995 + }, + { + "epoch": 0.09241601085795588, + "grad_norm": 0.18519705533981323, + "learning_rate": 4.6206045650832816e-05, + "loss": 3.2982, + "step": 2996 + }, + { + "epoch": 0.0924468573235293, + "grad_norm": 0.18102823197841644, + "learning_rate": 4.6221468229487975e-05, + "loss": 3.3051, + "step": 2997 + }, + { + "epoch": 0.09247770378910271, + "grad_norm": 0.19139650464057922, + "learning_rate": 4.623689080814313e-05, + "loss": 3.328, + "step": 2998 + }, + { + "epoch": 0.09250855025467614, + "grad_norm": 0.5375152230262756, + "learning_rate": 4.625231338679827e-05, + "loss": 3.3427, + "step": 2999 + }, + { + "epoch": 0.09253939672024955, + "grad_norm": 0.19338785111904144, + "learning_rate": 4.6267735965453425e-05, + "loss": 3.3557, + "step": 3000 + }, + { + "epoch": 0.09257024318582296, + "grad_norm": 0.18206456303596497, + "learning_rate": 4.628315854410858e-05, + "loss": 3.3299, + "step": 3001 + }, + { + "epoch": 0.09260108965139638, + "grad_norm": 0.23149137198925018, + "learning_rate": 4.629858112276373e-05, + "loss": 3.3199, + "step": 3002 + }, + { + "epoch": 0.09263193611696979, + "grad_norm": 0.18586365878582, + "learning_rate": 4.631400370141888e-05, + "loss": 3.3515, + "step": 3003 + }, + { + "epoch": 0.09266278258254321, + "grad_norm": 0.18953582644462585, + "learning_rate": 4.632942628007403e-05, + "loss": 3.3365, + "step": 3004 + }, + { + "epoch": 0.09269362904811664, + "grad_norm": 0.20097455382347107, + "learning_rate": 4.634484885872919e-05, + "loss": 3.3159, + "step": 3005 + }, + { + "epoch": 0.09272447551369005, + "grad_norm": 0.19055789709091187, + "learning_rate": 4.636027143738433e-05, + "loss": 3.3019, + "step": 3006 + }, + { + "epoch": 0.09275532197926346, + "grad_norm": 0.18591275811195374, + "learning_rate": 4.6375694016039484e-05, + "loss": 3.3467, + "step": 3007 + }, + { + "epoch": 0.09278616844483688, + "grad_norm": 0.19050711393356323, + "learning_rate": 4.639111659469464e-05, + "loss": 3.3523, + "step": 3008 + }, + { + "epoch": 0.09281701491041029, + "grad_norm": 0.18715515732765198, + "learning_rate": 4.640653917334979e-05, + "loss": 3.321, + "step": 3009 + }, + { + "epoch": 0.0928478613759837, + "grad_norm": 0.1921118050813675, + "learning_rate": 4.642196175200494e-05, + "loss": 3.3185, + "step": 3010 + }, + { + "epoch": 0.09287870784155713, + "grad_norm": 0.18194974958896637, + "learning_rate": 4.643738433066009e-05, + "loss": 3.3361, + "step": 3011 + }, + { + "epoch": 0.09290955430713055, + "grad_norm": 0.18438059091567993, + "learning_rate": 4.645280690931524e-05, + "loss": 3.3758, + "step": 3012 + }, + { + "epoch": 0.09294040077270396, + "grad_norm": 0.19211071729660034, + "learning_rate": 4.646822948797039e-05, + "loss": 3.3434, + "step": 3013 + }, + { + "epoch": 0.09297124723827738, + "grad_norm": 0.1908780187368393, + "learning_rate": 4.6483652066625544e-05, + "loss": 3.3393, + "step": 3014 + }, + { + "epoch": 0.09300209370385079, + "grad_norm": 0.1820296198129654, + "learning_rate": 4.6499074645280696e-05, + "loss": 3.3268, + "step": 3015 + }, + { + "epoch": 0.0930329401694242, + "grad_norm": 0.2007494419813156, + "learning_rate": 4.651449722393584e-05, + "loss": 3.3335, + "step": 3016 + }, + { + "epoch": 0.09306378663499763, + "grad_norm": 0.18609727919101715, + "learning_rate": 4.6529919802590994e-05, + "loss": 3.3419, + "step": 3017 + }, + { + "epoch": 0.09309463310057105, + "grad_norm": 0.18813715875148773, + "learning_rate": 4.6545342381246146e-05, + "loss": 3.3768, + "step": 3018 + }, + { + "epoch": 0.09312547956614446, + "grad_norm": 0.19365821778774261, + "learning_rate": 4.65607649599013e-05, + "loss": 3.3334, + "step": 3019 + }, + { + "epoch": 0.09315632603171788, + "grad_norm": 0.18375615775585175, + "learning_rate": 4.657618753855645e-05, + "loss": 3.3386, + "step": 3020 + }, + { + "epoch": 0.09318717249729129, + "grad_norm": 0.1908135563135147, + "learning_rate": 4.6591610117211596e-05, + "loss": 3.3465, + "step": 3021 + }, + { + "epoch": 0.0932180189628647, + "grad_norm": 0.19265779852867126, + "learning_rate": 4.6607032695866755e-05, + "loss": 3.3387, + "step": 3022 + }, + { + "epoch": 0.09324886542843813, + "grad_norm": 0.203319251537323, + "learning_rate": 4.66224552745219e-05, + "loss": 3.3475, + "step": 3023 + }, + { + "epoch": 0.09327971189401155, + "grad_norm": 0.21685557067394257, + "learning_rate": 4.663787785317705e-05, + "loss": 3.3184, + "step": 3024 + }, + { + "epoch": 0.09331055835958496, + "grad_norm": 0.19145062565803528, + "learning_rate": 4.6653300431832205e-05, + "loss": 3.3525, + "step": 3025 + }, + { + "epoch": 0.09334140482515838, + "grad_norm": 0.1945759356021881, + "learning_rate": 4.666872301048735e-05, + "loss": 3.358, + "step": 3026 + }, + { + "epoch": 0.09337225129073179, + "grad_norm": 0.18635056912899017, + "learning_rate": 4.668414558914251e-05, + "loss": 3.3128, + "step": 3027 + }, + { + "epoch": 0.09340309775630522, + "grad_norm": 0.19186289608478546, + "learning_rate": 4.6699568167797655e-05, + "loss": 3.3459, + "step": 3028 + }, + { + "epoch": 0.09343394422187863, + "grad_norm": 0.19135165214538574, + "learning_rate": 4.671499074645281e-05, + "loss": 3.3235, + "step": 3029 + }, + { + "epoch": 0.09346479068745205, + "grad_norm": 0.18125827610492706, + "learning_rate": 4.673041332510796e-05, + "loss": 3.3286, + "step": 3030 + }, + { + "epoch": 0.09349563715302546, + "grad_norm": 0.18503139913082123, + "learning_rate": 4.674583590376311e-05, + "loss": 3.3454, + "step": 3031 + }, + { + "epoch": 0.09352648361859887, + "grad_norm": 0.17820127308368683, + "learning_rate": 4.6761258482418265e-05, + "loss": 3.3353, + "step": 3032 + }, + { + "epoch": 0.09355733008417229, + "grad_norm": 0.19364699721336365, + "learning_rate": 4.677668106107341e-05, + "loss": 3.3676, + "step": 3033 + }, + { + "epoch": 0.09358817654974572, + "grad_norm": 0.17954213917255402, + "learning_rate": 4.679210363972856e-05, + "loss": 3.3605, + "step": 3034 + }, + { + "epoch": 0.09361902301531913, + "grad_norm": 0.19674040377140045, + "learning_rate": 4.680752621838372e-05, + "loss": 3.3357, + "step": 3035 + }, + { + "epoch": 0.09364986948089254, + "grad_norm": 0.1842634528875351, + "learning_rate": 4.682294879703887e-05, + "loss": 3.3218, + "step": 3036 + }, + { + "epoch": 0.09368071594646596, + "grad_norm": 0.19484500586986542, + "learning_rate": 4.683837137569402e-05, + "loss": 3.294, + "step": 3037 + }, + { + "epoch": 0.09371156241203937, + "grad_norm": 0.22278261184692383, + "learning_rate": 4.6853793954349165e-05, + "loss": 3.3214, + "step": 3038 + }, + { + "epoch": 0.09374240887761279, + "grad_norm": 0.18324948847293854, + "learning_rate": 4.6869216533004324e-05, + "loss": 3.3207, + "step": 3039 + }, + { + "epoch": 0.09377325534318622, + "grad_norm": 0.17751923203468323, + "learning_rate": 4.6884639111659476e-05, + "loss": 3.3169, + "step": 3040 + }, + { + "epoch": 0.09380410180875963, + "grad_norm": 0.18578030169010162, + "learning_rate": 4.690006169031462e-05, + "loss": 3.3469, + "step": 3041 + }, + { + "epoch": 0.09383494827433304, + "grad_norm": 0.19650425016880035, + "learning_rate": 4.6915484268969774e-05, + "loss": 3.3361, + "step": 3042 + }, + { + "epoch": 0.09386579473990646, + "grad_norm": 0.18860600888729095, + "learning_rate": 4.693090684762492e-05, + "loss": 3.3279, + "step": 3043 + }, + { + "epoch": 0.09389664120547987, + "grad_norm": 0.18144741654396057, + "learning_rate": 4.694632942628008e-05, + "loss": 3.3622, + "step": 3044 + }, + { + "epoch": 0.09392748767105329, + "grad_norm": 0.1914285570383072, + "learning_rate": 4.696175200493523e-05, + "loss": 3.3535, + "step": 3045 + }, + { + "epoch": 0.09395833413662671, + "grad_norm": 0.18238362669944763, + "learning_rate": 4.6977174583590376e-05, + "loss": 3.3307, + "step": 3046 + }, + { + "epoch": 0.09398918060220013, + "grad_norm": 0.18726398050785065, + "learning_rate": 4.699259716224553e-05, + "loss": 3.3286, + "step": 3047 + }, + { + "epoch": 0.09402002706777354, + "grad_norm": 0.19188641011714935, + "learning_rate": 4.700801974090068e-05, + "loss": 3.3527, + "step": 3048 + }, + { + "epoch": 0.09405087353334696, + "grad_norm": 0.1899179220199585, + "learning_rate": 4.702344231955583e-05, + "loss": 3.3396, + "step": 3049 + }, + { + "epoch": 0.09408171999892037, + "grad_norm": 0.20190322399139404, + "learning_rate": 4.7038864898210986e-05, + "loss": 3.3118, + "step": 3050 + }, + { + "epoch": 0.09411256646449379, + "grad_norm": 0.18019166588783264, + "learning_rate": 4.705428747686613e-05, + "loss": 3.3327, + "step": 3051 + }, + { + "epoch": 0.09414341293006721, + "grad_norm": 0.1908770352602005, + "learning_rate": 4.706971005552129e-05, + "loss": 3.3356, + "step": 3052 + }, + { + "epoch": 0.09417425939564063, + "grad_norm": 0.18528999388217926, + "learning_rate": 4.7085132634176436e-05, + "loss": 3.3668, + "step": 3053 + }, + { + "epoch": 0.09420510586121404, + "grad_norm": 0.19300785660743713, + "learning_rate": 4.710055521283159e-05, + "loss": 3.3566, + "step": 3054 + }, + { + "epoch": 0.09423595232678746, + "grad_norm": 0.1969001442193985, + "learning_rate": 4.711597779148674e-05, + "loss": 3.348, + "step": 3055 + }, + { + "epoch": 0.09426679879236087, + "grad_norm": 0.18546324968338013, + "learning_rate": 4.713140037014189e-05, + "loss": 3.3, + "step": 3056 + }, + { + "epoch": 0.09429764525793428, + "grad_norm": 0.2043047845363617, + "learning_rate": 4.7146822948797045e-05, + "loss": 3.3587, + "step": 3057 + }, + { + "epoch": 0.09432849172350771, + "grad_norm": 0.18661901354789734, + "learning_rate": 4.716224552745219e-05, + "loss": 3.3535, + "step": 3058 + }, + { + "epoch": 0.09435933818908113, + "grad_norm": 0.18587970733642578, + "learning_rate": 4.717766810610734e-05, + "loss": 3.3138, + "step": 3059 + }, + { + "epoch": 0.09439018465465454, + "grad_norm": 0.18820445239543915, + "learning_rate": 4.7193090684762495e-05, + "loss": 3.2687, + "step": 3060 + }, + { + "epoch": 0.09442103112022796, + "grad_norm": 0.18724325299263, + "learning_rate": 4.720851326341765e-05, + "loss": 3.3154, + "step": 3061 + }, + { + "epoch": 0.09445187758580137, + "grad_norm": 0.18037928640842438, + "learning_rate": 4.72239358420728e-05, + "loss": 3.2928, + "step": 3062 + }, + { + "epoch": 0.09448272405137478, + "grad_norm": 0.18468180298805237, + "learning_rate": 4.7239358420727945e-05, + "loss": 3.3174, + "step": 3063 + }, + { + "epoch": 0.09451357051694821, + "grad_norm": 0.1913236826658249, + "learning_rate": 4.72547809993831e-05, + "loss": 3.3146, + "step": 3064 + }, + { + "epoch": 0.09454441698252163, + "grad_norm": 0.21230743825435638, + "learning_rate": 4.727020357803825e-05, + "loss": 3.2895, + "step": 3065 + }, + { + "epoch": 0.09457526344809504, + "grad_norm": 0.19073060154914856, + "learning_rate": 4.72856261566934e-05, + "loss": 3.3368, + "step": 3066 + }, + { + "epoch": 0.09460610991366845, + "grad_norm": 0.19638466835021973, + "learning_rate": 4.7301048735348554e-05, + "loss": 3.2986, + "step": 3067 + }, + { + "epoch": 0.09463695637924187, + "grad_norm": 0.1950896680355072, + "learning_rate": 4.73164713140037e-05, + "loss": 3.3067, + "step": 3068 + }, + { + "epoch": 0.09466780284481528, + "grad_norm": 0.20599783957004547, + "learning_rate": 4.733189389265886e-05, + "loss": 3.3095, + "step": 3069 + }, + { + "epoch": 0.09469864931038871, + "grad_norm": 0.18368323147296906, + "learning_rate": 4.7347316471314004e-05, + "loss": 3.3233, + "step": 3070 + }, + { + "epoch": 0.09472949577596212, + "grad_norm": 0.19701489806175232, + "learning_rate": 4.736273904996916e-05, + "loss": 3.328, + "step": 3071 + }, + { + "epoch": 0.09476034224153554, + "grad_norm": 0.19212469458580017, + "learning_rate": 4.737816162862431e-05, + "loss": 3.3478, + "step": 3072 + }, + { + "epoch": 0.09479118870710895, + "grad_norm": 0.19731910526752472, + "learning_rate": 4.7393584207279454e-05, + "loss": 3.3619, + "step": 3073 + }, + { + "epoch": 0.09482203517268237, + "grad_norm": 0.20496459305286407, + "learning_rate": 4.7409006785934614e-05, + "loss": 3.3219, + "step": 3074 + }, + { + "epoch": 0.09485288163825578, + "grad_norm": 0.2006881684064865, + "learning_rate": 4.742442936458976e-05, + "loss": 3.3151, + "step": 3075 + }, + { + "epoch": 0.09488372810382921, + "grad_norm": 0.18833547830581665, + "learning_rate": 4.743985194324491e-05, + "loss": 3.2999, + "step": 3076 + }, + { + "epoch": 0.09491457456940262, + "grad_norm": 0.2073565274477005, + "learning_rate": 4.7455274521900064e-05, + "loss": 3.3299, + "step": 3077 + }, + { + "epoch": 0.09494542103497604, + "grad_norm": 0.17863507568836212, + "learning_rate": 4.7470697100555216e-05, + "loss": 3.3461, + "step": 3078 + }, + { + "epoch": 0.09497626750054945, + "grad_norm": 0.20467273890972137, + "learning_rate": 4.748611967921037e-05, + "loss": 3.3043, + "step": 3079 + }, + { + "epoch": 0.09500711396612287, + "grad_norm": 0.1936817616224289, + "learning_rate": 4.7501542257865514e-05, + "loss": 3.3218, + "step": 3080 + }, + { + "epoch": 0.09503796043169628, + "grad_norm": 0.19123460352420807, + "learning_rate": 4.7516964836520666e-05, + "loss": 3.3198, + "step": 3081 + }, + { + "epoch": 0.09506880689726971, + "grad_norm": 0.19976411759853363, + "learning_rate": 4.7532387415175825e-05, + "loss": 3.3357, + "step": 3082 + }, + { + "epoch": 0.09509965336284312, + "grad_norm": 0.17974047362804413, + "learning_rate": 4.754780999383097e-05, + "loss": 3.3514, + "step": 3083 + }, + { + "epoch": 0.09513049982841654, + "grad_norm": 0.21281373500823975, + "learning_rate": 4.756323257248612e-05, + "loss": 3.3432, + "step": 3084 + }, + { + "epoch": 0.09516134629398995, + "grad_norm": 0.21655049920082092, + "learning_rate": 4.757865515114127e-05, + "loss": 3.3773, + "step": 3085 + }, + { + "epoch": 0.09519219275956337, + "grad_norm": 0.18886660039424896, + "learning_rate": 4.759407772979643e-05, + "loss": 3.3107, + "step": 3086 + }, + { + "epoch": 0.09522303922513678, + "grad_norm": 0.19202342629432678, + "learning_rate": 4.760950030845158e-05, + "loss": 3.3305, + "step": 3087 + }, + { + "epoch": 0.09525388569071021, + "grad_norm": 0.18886899948120117, + "learning_rate": 4.7624922887106725e-05, + "loss": 3.3282, + "step": 3088 + }, + { + "epoch": 0.09528473215628362, + "grad_norm": 0.2077721655368805, + "learning_rate": 4.764034546576188e-05, + "loss": 3.3418, + "step": 3089 + }, + { + "epoch": 0.09531557862185704, + "grad_norm": 0.18542811274528503, + "learning_rate": 4.765576804441702e-05, + "loss": 3.328, + "step": 3090 + }, + { + "epoch": 0.09534642508743045, + "grad_norm": 0.19663546979427338, + "learning_rate": 4.767119062307218e-05, + "loss": 3.3406, + "step": 3091 + }, + { + "epoch": 0.09537727155300386, + "grad_norm": 0.19963832199573517, + "learning_rate": 4.7686613201727335e-05, + "loss": 3.343, + "step": 3092 + }, + { + "epoch": 0.09540811801857728, + "grad_norm": 0.1961738020181656, + "learning_rate": 4.770203578038248e-05, + "loss": 3.33, + "step": 3093 + }, + { + "epoch": 0.0954389644841507, + "grad_norm": 0.2028144747018814, + "learning_rate": 4.771745835903763e-05, + "loss": 3.3227, + "step": 3094 + }, + { + "epoch": 0.09546981094972412, + "grad_norm": 0.18312743306159973, + "learning_rate": 4.7732880937692785e-05, + "loss": 3.3039, + "step": 3095 + }, + { + "epoch": 0.09550065741529754, + "grad_norm": 0.19446969032287598, + "learning_rate": 4.774830351634794e-05, + "loss": 3.305, + "step": 3096 + }, + { + "epoch": 0.09553150388087095, + "grad_norm": 0.18956603109836578, + "learning_rate": 4.776372609500309e-05, + "loss": 3.3403, + "step": 3097 + }, + { + "epoch": 0.09556235034644436, + "grad_norm": 0.19702552258968353, + "learning_rate": 4.7779148673658235e-05, + "loss": 3.3208, + "step": 3098 + }, + { + "epoch": 0.09559319681201778, + "grad_norm": 0.1963428258895874, + "learning_rate": 4.7794571252313394e-05, + "loss": 3.3484, + "step": 3099 + }, + { + "epoch": 0.0956240432775912, + "grad_norm": 0.18758831918239594, + "learning_rate": 4.780999383096854e-05, + "loss": 3.3476, + "step": 3100 + }, + { + "epoch": 0.09565488974316462, + "grad_norm": 0.19379684329032898, + "learning_rate": 4.782541640962369e-05, + "loss": 3.3318, + "step": 3101 + }, + { + "epoch": 0.09568573620873803, + "grad_norm": 0.1890457421541214, + "learning_rate": 4.7840838988278844e-05, + "loss": 3.3079, + "step": 3102 + }, + { + "epoch": 0.09571658267431145, + "grad_norm": 0.191565603017807, + "learning_rate": 4.7856261566933996e-05, + "loss": 3.3086, + "step": 3103 + }, + { + "epoch": 0.09574742913988486, + "grad_norm": 0.18727917969226837, + "learning_rate": 4.787168414558915e-05, + "loss": 3.3188, + "step": 3104 + }, + { + "epoch": 0.09577827560545828, + "grad_norm": 0.1797633320093155, + "learning_rate": 4.7887106724244294e-05, + "loss": 3.3291, + "step": 3105 + }, + { + "epoch": 0.0958091220710317, + "grad_norm": 0.18286117911338806, + "learning_rate": 4.7902529302899446e-05, + "loss": 3.3733, + "step": 3106 + }, + { + "epoch": 0.09583996853660512, + "grad_norm": 0.19720973074436188, + "learning_rate": 4.79179518815546e-05, + "loss": 3.3372, + "step": 3107 + }, + { + "epoch": 0.09587081500217853, + "grad_norm": 0.18651039898395538, + "learning_rate": 4.793337446020975e-05, + "loss": 3.3235, + "step": 3108 + }, + { + "epoch": 0.09590166146775195, + "grad_norm": 0.5777696371078491, + "learning_rate": 4.79487970388649e-05, + "loss": 3.3141, + "step": 3109 + }, + { + "epoch": 0.09593250793332536, + "grad_norm": 0.18405193090438843, + "learning_rate": 4.796421961752005e-05, + "loss": 3.3321, + "step": 3110 + }, + { + "epoch": 0.09596335439889878, + "grad_norm": 0.19993756711483002, + "learning_rate": 4.79796421961752e-05, + "loss": 3.3363, + "step": 3111 + }, + { + "epoch": 0.0959942008644722, + "grad_norm": 0.1982763409614563, + "learning_rate": 4.799506477483035e-05, + "loss": 3.3467, + "step": 3112 + }, + { + "epoch": 0.09602504733004562, + "grad_norm": 0.19820217788219452, + "learning_rate": 4.8010487353485506e-05, + "loss": 3.3646, + "step": 3113 + }, + { + "epoch": 0.09605589379561903, + "grad_norm": 1.0764055252075195, + "learning_rate": 4.802590993214066e-05, + "loss": 3.3522, + "step": 3114 + }, + { + "epoch": 0.09608674026119245, + "grad_norm": 0.24970798194408417, + "learning_rate": 4.8041332510795803e-05, + "loss": 3.3141, + "step": 3115 + }, + { + "epoch": 0.09611758672676586, + "grad_norm": 0.21166352927684784, + "learning_rate": 4.805675508945096e-05, + "loss": 3.2907, + "step": 3116 + }, + { + "epoch": 0.09614843319233927, + "grad_norm": 1.5539162158966064, + "learning_rate": 4.807217766810611e-05, + "loss": 3.3383, + "step": 3117 + }, + { + "epoch": 0.0961792796579127, + "grad_norm": 0.22541797161102295, + "learning_rate": 4.808760024676126e-05, + "loss": 3.2777, + "step": 3118 + }, + { + "epoch": 0.09621012612348612, + "grad_norm": 0.23259076476097107, + "learning_rate": 4.810302282541641e-05, + "loss": 3.3364, + "step": 3119 + }, + { + "epoch": 0.09624097258905953, + "grad_norm": 0.21946358680725098, + "learning_rate": 4.811844540407156e-05, + "loss": 3.3284, + "step": 3120 + }, + { + "epoch": 0.09627181905463295, + "grad_norm": 0.23317955434322357, + "learning_rate": 4.813386798272672e-05, + "loss": 3.3424, + "step": 3121 + }, + { + "epoch": 0.09630266552020636, + "grad_norm": 0.2220587432384491, + "learning_rate": 4.814929056138186e-05, + "loss": 3.3634, + "step": 3122 + }, + { + "epoch": 0.09633351198577977, + "grad_norm": 0.2408958524465561, + "learning_rate": 4.8164713140037015e-05, + "loss": 3.3241, + "step": 3123 + }, + { + "epoch": 0.0963643584513532, + "grad_norm": 0.21283207833766937, + "learning_rate": 4.818013571869217e-05, + "loss": 3.306, + "step": 3124 + }, + { + "epoch": 0.09639520491692662, + "grad_norm": 0.21675877273082733, + "learning_rate": 4.819555829734732e-05, + "loss": 3.3337, + "step": 3125 + }, + { + "epoch": 0.09642605138250003, + "grad_norm": 0.2404310554265976, + "learning_rate": 4.821098087600247e-05, + "loss": 3.3409, + "step": 3126 + }, + { + "epoch": 0.09645689784807344, + "grad_norm": 0.22955885529518127, + "learning_rate": 4.822640345465762e-05, + "loss": 3.3219, + "step": 3127 + }, + { + "epoch": 0.09648774431364686, + "grad_norm": 0.1925782710313797, + "learning_rate": 4.824182603331277e-05, + "loss": 3.3306, + "step": 3128 + }, + { + "epoch": 0.09651859077922027, + "grad_norm": 1.1006418466567993, + "learning_rate": 4.825724861196793e-05, + "loss": 3.3326, + "step": 3129 + }, + { + "epoch": 0.0965494372447937, + "grad_norm": 0.2213466912508011, + "learning_rate": 4.8272671190623074e-05, + "loss": 3.3421, + "step": 3130 + }, + { + "epoch": 0.09658028371036712, + "grad_norm": 0.20027276873588562, + "learning_rate": 4.8288093769278227e-05, + "loss": 3.3495, + "step": 3131 + }, + { + "epoch": 0.09661113017594053, + "grad_norm": 0.23761235177516937, + "learning_rate": 4.830351634793337e-05, + "loss": 3.3512, + "step": 3132 + }, + { + "epoch": 0.09664197664151394, + "grad_norm": 0.33305734395980835, + "learning_rate": 4.831893892658853e-05, + "loss": 3.3461, + "step": 3133 + }, + { + "epoch": 0.09667282310708736, + "grad_norm": 1.3541532754898071, + "learning_rate": 4.8334361505243683e-05, + "loss": 3.3414, + "step": 3134 + }, + { + "epoch": 0.09670366957266077, + "grad_norm": 0.5150039792060852, + "learning_rate": 4.834978408389883e-05, + "loss": 3.3616, + "step": 3135 + }, + { + "epoch": 0.0967345160382342, + "grad_norm": 0.33250412344932556, + "learning_rate": 4.836520666255398e-05, + "loss": 3.3258, + "step": 3136 + }, + { + "epoch": 0.09676536250380761, + "grad_norm": 0.570968747138977, + "learning_rate": 4.838062924120913e-05, + "loss": 3.3566, + "step": 3137 + }, + { + "epoch": 0.09679620896938103, + "grad_norm": 0.28834614157676697, + "learning_rate": 4.8396051819864286e-05, + "loss": 3.3344, + "step": 3138 + }, + { + "epoch": 0.09682705543495444, + "grad_norm": 0.25053268671035767, + "learning_rate": 4.841147439851944e-05, + "loss": 3.3287, + "step": 3139 + }, + { + "epoch": 0.09685790190052786, + "grad_norm": 0.2598170340061188, + "learning_rate": 4.8426896977174584e-05, + "loss": 3.3468, + "step": 3140 + }, + { + "epoch": 0.09688874836610127, + "grad_norm": 0.23727662861347198, + "learning_rate": 4.8442319555829736e-05, + "loss": 3.3408, + "step": 3141 + }, + { + "epoch": 0.0969195948316747, + "grad_norm": 0.35075482726097107, + "learning_rate": 4.845774213448489e-05, + "loss": 3.3273, + "step": 3142 + }, + { + "epoch": 0.09695044129724811, + "grad_norm": 0.25904375314712524, + "learning_rate": 4.847316471314004e-05, + "loss": 3.3099, + "step": 3143 + }, + { + "epoch": 0.09698128776282153, + "grad_norm": 0.23824749886989594, + "learning_rate": 4.848858729179519e-05, + "loss": 3.3291, + "step": 3144 + }, + { + "epoch": 0.09701213422839494, + "grad_norm": 0.26023241877555847, + "learning_rate": 4.850400987045034e-05, + "loss": 3.3281, + "step": 3145 + }, + { + "epoch": 0.09704298069396836, + "grad_norm": 0.23633742332458496, + "learning_rate": 4.85194324491055e-05, + "loss": 3.3189, + "step": 3146 + }, + { + "epoch": 0.09707382715954177, + "grad_norm": 0.23257866501808167, + "learning_rate": 4.853485502776064e-05, + "loss": 3.3061, + "step": 3147 + }, + { + "epoch": 0.0971046736251152, + "grad_norm": 0.2185048907995224, + "learning_rate": 4.8550277606415795e-05, + "loss": 3.3455, + "step": 3148 + }, + { + "epoch": 0.09713552009068861, + "grad_norm": 0.2483467161655426, + "learning_rate": 4.856570018507095e-05, + "loss": 3.3046, + "step": 3149 + }, + { + "epoch": 0.09716636655626203, + "grad_norm": 0.2264622300863266, + "learning_rate": 4.858112276372609e-05, + "loss": 3.315, + "step": 3150 + }, + { + "epoch": 0.09719721302183544, + "grad_norm": 0.2331334948539734, + "learning_rate": 4.859654534238125e-05, + "loss": 3.3391, + "step": 3151 + }, + { + "epoch": 0.09722805948740885, + "grad_norm": 0.22386357188224792, + "learning_rate": 4.86119679210364e-05, + "loss": 3.3209, + "step": 3152 + }, + { + "epoch": 0.09725890595298227, + "grad_norm": 0.19012847542762756, + "learning_rate": 4.862739049969155e-05, + "loss": 3.3252, + "step": 3153 + }, + { + "epoch": 0.0972897524185557, + "grad_norm": 0.22813504934310913, + "learning_rate": 4.86428130783467e-05, + "loss": 3.3317, + "step": 3154 + }, + { + "epoch": 0.09732059888412911, + "grad_norm": 0.23285217583179474, + "learning_rate": 4.8658235657001854e-05, + "loss": 3.3036, + "step": 3155 + }, + { + "epoch": 0.09735144534970253, + "grad_norm": 0.20340359210968018, + "learning_rate": 4.867365823565701e-05, + "loss": 3.3435, + "step": 3156 + }, + { + "epoch": 0.09738229181527594, + "grad_norm": 0.20836354792118073, + "learning_rate": 4.868908081431215e-05, + "loss": 3.314, + "step": 3157 + }, + { + "epoch": 0.09741313828084935, + "grad_norm": 0.22066618502140045, + "learning_rate": 4.8704503392967305e-05, + "loss": 3.3333, + "step": 3158 + }, + { + "epoch": 0.09744398474642277, + "grad_norm": 0.2056196630001068, + "learning_rate": 4.871992597162246e-05, + "loss": 3.3171, + "step": 3159 + }, + { + "epoch": 0.0974748312119962, + "grad_norm": 0.2182077318429947, + "learning_rate": 4.873534855027761e-05, + "loss": 3.3462, + "step": 3160 + }, + { + "epoch": 0.09750567767756961, + "grad_norm": 0.19837097823619843, + "learning_rate": 4.875077112893276e-05, + "loss": 3.3007, + "step": 3161 + }, + { + "epoch": 0.09753652414314302, + "grad_norm": 0.30279016494750977, + "learning_rate": 4.876619370758791e-05, + "loss": 3.3251, + "step": 3162 + }, + { + "epoch": 0.09756737060871644, + "grad_norm": 0.1928970068693161, + "learning_rate": 4.8781616286243066e-05, + "loss": 3.3154, + "step": 3163 + }, + { + "epoch": 0.09759821707428985, + "grad_norm": 0.2223213016986847, + "learning_rate": 4.879703886489821e-05, + "loss": 3.3322, + "step": 3164 + }, + { + "epoch": 0.09762906353986327, + "grad_norm": 0.1905486285686493, + "learning_rate": 4.8812461443553364e-05, + "loss": 3.33, + "step": 3165 + }, + { + "epoch": 0.0976599100054367, + "grad_norm": 0.21444550156593323, + "learning_rate": 4.8827884022208516e-05, + "loss": 3.3488, + "step": 3166 + }, + { + "epoch": 0.09769075647101011, + "grad_norm": 0.19812120497226715, + "learning_rate": 4.884330660086366e-05, + "loss": 3.3574, + "step": 3167 + }, + { + "epoch": 0.09772160293658352, + "grad_norm": 0.21528403460979462, + "learning_rate": 4.885872917951882e-05, + "loss": 3.3188, + "step": 3168 + }, + { + "epoch": 0.09775244940215694, + "grad_norm": 0.20758476853370667, + "learning_rate": 4.8874151758173966e-05, + "loss": 3.3212, + "step": 3169 + }, + { + "epoch": 0.09778329586773035, + "grad_norm": 0.2018652856349945, + "learning_rate": 4.888957433682912e-05, + "loss": 3.357, + "step": 3170 + }, + { + "epoch": 0.09781414233330377, + "grad_norm": 0.19137854874134064, + "learning_rate": 4.890499691548427e-05, + "loss": 3.3125, + "step": 3171 + }, + { + "epoch": 0.0978449887988772, + "grad_norm": 0.184129536151886, + "learning_rate": 4.892041949413942e-05, + "loss": 3.3469, + "step": 3172 + }, + { + "epoch": 0.09787583526445061, + "grad_norm": 0.18742451071739197, + "learning_rate": 4.8935842072794575e-05, + "loss": 3.2946, + "step": 3173 + }, + { + "epoch": 0.09790668173002402, + "grad_norm": 0.18472890555858612, + "learning_rate": 4.895126465144972e-05, + "loss": 3.3161, + "step": 3174 + }, + { + "epoch": 0.09793752819559744, + "grad_norm": 0.1967545449733734, + "learning_rate": 4.896668723010487e-05, + "loss": 3.3288, + "step": 3175 + }, + { + "epoch": 0.09796837466117085, + "grad_norm": 0.19031459093093872, + "learning_rate": 4.898210980876003e-05, + "loss": 3.3302, + "step": 3176 + }, + { + "epoch": 0.09799922112674427, + "grad_norm": 0.18739113211631775, + "learning_rate": 4.899753238741518e-05, + "loss": 3.2958, + "step": 3177 + }, + { + "epoch": 0.0980300675923177, + "grad_norm": 0.17898733913898468, + "learning_rate": 4.901295496607033e-05, + "loss": 3.3293, + "step": 3178 + }, + { + "epoch": 0.09806091405789111, + "grad_norm": 0.19223682582378387, + "learning_rate": 4.9028377544725476e-05, + "loss": 3.3435, + "step": 3179 + }, + { + "epoch": 0.09809176052346452, + "grad_norm": 0.19596529006958008, + "learning_rate": 4.9043800123380635e-05, + "loss": 3.3237, + "step": 3180 + }, + { + "epoch": 0.09812260698903794, + "grad_norm": 0.20192646980285645, + "learning_rate": 4.905922270203579e-05, + "loss": 3.3303, + "step": 3181 + }, + { + "epoch": 0.09815345345461135, + "grad_norm": 0.17634546756744385, + "learning_rate": 4.907464528069093e-05, + "loss": 3.2977, + "step": 3182 + }, + { + "epoch": 0.09818429992018476, + "grad_norm": 0.1840737760066986, + "learning_rate": 4.9090067859346085e-05, + "loss": 3.3018, + "step": 3183 + }, + { + "epoch": 0.09821514638575819, + "grad_norm": 0.18555255234241486, + "learning_rate": 4.910549043800123e-05, + "loss": 3.3047, + "step": 3184 + }, + { + "epoch": 0.0982459928513316, + "grad_norm": 0.18093739449977875, + "learning_rate": 4.912091301665639e-05, + "loss": 3.3475, + "step": 3185 + }, + { + "epoch": 0.09827683931690502, + "grad_norm": 0.18553368747234344, + "learning_rate": 4.913633559531154e-05, + "loss": 3.3176, + "step": 3186 + }, + { + "epoch": 0.09830768578247844, + "grad_norm": 0.18860766291618347, + "learning_rate": 4.915175817396669e-05, + "loss": 3.3243, + "step": 3187 + }, + { + "epoch": 0.09833853224805185, + "grad_norm": 0.1778855323791504, + "learning_rate": 4.916718075262184e-05, + "loss": 3.2914, + "step": 3188 + }, + { + "epoch": 0.09836937871362526, + "grad_norm": 0.17951224744319916, + "learning_rate": 4.918260333127699e-05, + "loss": 3.2654, + "step": 3189 + }, + { + "epoch": 0.09840022517919869, + "grad_norm": 0.18969368934631348, + "learning_rate": 4.9198025909932144e-05, + "loss": 3.3239, + "step": 3190 + }, + { + "epoch": 0.0984310716447721, + "grad_norm": 0.182065948843956, + "learning_rate": 4.9213448488587296e-05, + "loss": 3.3079, + "step": 3191 + }, + { + "epoch": 0.09846191811034552, + "grad_norm": 0.17960132658481598, + "learning_rate": 4.922887106724244e-05, + "loss": 3.3333, + "step": 3192 + }, + { + "epoch": 0.09849276457591893, + "grad_norm": 0.18197940289974213, + "learning_rate": 4.92442936458976e-05, + "loss": 3.3239, + "step": 3193 + }, + { + "epoch": 0.09852361104149235, + "grad_norm": 0.187059685587883, + "learning_rate": 4.9259716224552747e-05, + "loss": 3.2973, + "step": 3194 + }, + { + "epoch": 0.09855445750706576, + "grad_norm": 0.1858416199684143, + "learning_rate": 4.92751388032079e-05, + "loss": 3.3231, + "step": 3195 + }, + { + "epoch": 0.09858530397263919, + "grad_norm": 0.18581296503543854, + "learning_rate": 4.929056138186305e-05, + "loss": 3.2901, + "step": 3196 + }, + { + "epoch": 0.0986161504382126, + "grad_norm": 0.18723860383033752, + "learning_rate": 4.93059839605182e-05, + "loss": 3.3403, + "step": 3197 + }, + { + "epoch": 0.09864699690378602, + "grad_norm": 0.18958723545074463, + "learning_rate": 4.9321406539173356e-05, + "loss": 3.3214, + "step": 3198 + }, + { + "epoch": 0.09867784336935943, + "grad_norm": 0.17753449082374573, + "learning_rate": 4.93368291178285e-05, + "loss": 3.3121, + "step": 3199 + }, + { + "epoch": 0.09870868983493285, + "grad_norm": 0.18684126436710358, + "learning_rate": 4.9352251696483653e-05, + "loss": 3.3483, + "step": 3200 + }, + { + "epoch": 0.09873953630050626, + "grad_norm": 0.1801404058933258, + "learning_rate": 4.9367674275138806e-05, + "loss": 3.3488, + "step": 3201 + }, + { + "epoch": 0.09877038276607969, + "grad_norm": 0.17653816938400269, + "learning_rate": 4.938309685379396e-05, + "loss": 3.3474, + "step": 3202 + }, + { + "epoch": 0.0988012292316531, + "grad_norm": 0.19884654879570007, + "learning_rate": 4.939851943244911e-05, + "loss": 3.3465, + "step": 3203 + }, + { + "epoch": 0.09883207569722652, + "grad_norm": 0.1897125393152237, + "learning_rate": 4.9413942011104256e-05, + "loss": 3.3019, + "step": 3204 + }, + { + "epoch": 0.09886292216279993, + "grad_norm": 0.1842755824327469, + "learning_rate": 4.942936458975941e-05, + "loss": 3.3282, + "step": 3205 + }, + { + "epoch": 0.09889376862837335, + "grad_norm": 0.19552266597747803, + "learning_rate": 4.944478716841456e-05, + "loss": 3.3159, + "step": 3206 + }, + { + "epoch": 0.09892461509394676, + "grad_norm": 0.18758255243301392, + "learning_rate": 4.946020974706971e-05, + "loss": 3.3337, + "step": 3207 + }, + { + "epoch": 0.09895546155952019, + "grad_norm": 0.1898254156112671, + "learning_rate": 4.9475632325724865e-05, + "loss": 3.3421, + "step": 3208 + }, + { + "epoch": 0.0989863080250936, + "grad_norm": 0.1831645965576172, + "learning_rate": 4.949105490438001e-05, + "loss": 3.3164, + "step": 3209 + }, + { + "epoch": 0.09901715449066702, + "grad_norm": 0.18461881577968597, + "learning_rate": 4.950647748303517e-05, + "loss": 3.3105, + "step": 3210 + }, + { + "epoch": 0.09904800095624043, + "grad_norm": 0.17614035308361053, + "learning_rate": 4.9521900061690315e-05, + "loss": 3.3296, + "step": 3211 + }, + { + "epoch": 0.09907884742181385, + "grad_norm": 0.1715174913406372, + "learning_rate": 4.953732264034547e-05, + "loss": 3.3135, + "step": 3212 + }, + { + "epoch": 0.09910969388738726, + "grad_norm": 0.19899846613407135, + "learning_rate": 4.955274521900062e-05, + "loss": 3.3114, + "step": 3213 + }, + { + "epoch": 0.09914054035296069, + "grad_norm": 0.16977179050445557, + "learning_rate": 4.9568167797655765e-05, + "loss": 3.2946, + "step": 3214 + }, + { + "epoch": 0.0991713868185341, + "grad_norm": 0.1807067096233368, + "learning_rate": 4.9583590376310924e-05, + "loss": 3.2956, + "step": 3215 + }, + { + "epoch": 0.09920223328410752, + "grad_norm": 0.1823662966489792, + "learning_rate": 4.959901295496607e-05, + "loss": 3.2808, + "step": 3216 + }, + { + "epoch": 0.09923307974968093, + "grad_norm": 0.17971082031726837, + "learning_rate": 4.961443553362122e-05, + "loss": 3.3248, + "step": 3217 + }, + { + "epoch": 0.09926392621525434, + "grad_norm": 0.309326708316803, + "learning_rate": 4.9629858112276374e-05, + "loss": 3.3216, + "step": 3218 + }, + { + "epoch": 0.09929477268082776, + "grad_norm": 0.18088534474372864, + "learning_rate": 4.964528069093153e-05, + "loss": 3.2963, + "step": 3219 + }, + { + "epoch": 0.09932561914640119, + "grad_norm": 0.18433576822280884, + "learning_rate": 4.966070326958668e-05, + "loss": 3.34, + "step": 3220 + }, + { + "epoch": 0.0993564656119746, + "grad_norm": 0.1914118230342865, + "learning_rate": 4.9676125848241825e-05, + "loss": 3.3615, + "step": 3221 + }, + { + "epoch": 0.09938731207754802, + "grad_norm": 0.17931032180786133, + "learning_rate": 4.969154842689698e-05, + "loss": 3.3475, + "step": 3222 + }, + { + "epoch": 0.09941815854312143, + "grad_norm": 0.179707333445549, + "learning_rate": 4.9706971005552136e-05, + "loss": 3.3146, + "step": 3223 + }, + { + "epoch": 0.09944900500869484, + "grad_norm": 0.19599418342113495, + "learning_rate": 4.972239358420728e-05, + "loss": 3.3104, + "step": 3224 + }, + { + "epoch": 0.09947985147426826, + "grad_norm": 0.18052034080028534, + "learning_rate": 4.9737816162862434e-05, + "loss": 3.3385, + "step": 3225 + }, + { + "epoch": 0.09951069793984169, + "grad_norm": 0.17302420735359192, + "learning_rate": 4.9753238741517586e-05, + "loss": 3.3263, + "step": 3226 + }, + { + "epoch": 0.0995415444054151, + "grad_norm": 0.18204817175865173, + "learning_rate": 4.976866132017273e-05, + "loss": 3.3505, + "step": 3227 + }, + { + "epoch": 0.09957239087098851, + "grad_norm": 0.17606443166732788, + "learning_rate": 4.978408389882789e-05, + "loss": 3.3585, + "step": 3228 + }, + { + "epoch": 0.09960323733656193, + "grad_norm": 0.17607229948043823, + "learning_rate": 4.9799506477483036e-05, + "loss": 3.3094, + "step": 3229 + }, + { + "epoch": 0.09963408380213534, + "grad_norm": 0.1797540783882141, + "learning_rate": 4.981492905613819e-05, + "loss": 3.3255, + "step": 3230 + }, + { + "epoch": 0.09966493026770876, + "grad_norm": 0.17534451186656952, + "learning_rate": 4.983035163479334e-05, + "loss": 3.3349, + "step": 3231 + }, + { + "epoch": 0.09969577673328218, + "grad_norm": 0.17869730293750763, + "learning_rate": 4.984577421344849e-05, + "loss": 3.2981, + "step": 3232 + }, + { + "epoch": 0.0997266231988556, + "grad_norm": 0.1744428426027298, + "learning_rate": 4.9861196792103645e-05, + "loss": 3.3262, + "step": 3233 + }, + { + "epoch": 0.09975746966442901, + "grad_norm": 0.18115711212158203, + "learning_rate": 4.987661937075879e-05, + "loss": 3.3046, + "step": 3234 + }, + { + "epoch": 0.09978831613000243, + "grad_norm": 0.1839175671339035, + "learning_rate": 4.989204194941394e-05, + "loss": 3.3248, + "step": 3235 + }, + { + "epoch": 0.09981916259557584, + "grad_norm": 0.6871650218963623, + "learning_rate": 4.9907464528069095e-05, + "loss": 3.3378, + "step": 3236 + }, + { + "epoch": 0.09985000906114926, + "grad_norm": 0.18206176161766052, + "learning_rate": 4.992288710672425e-05, + "loss": 3.3048, + "step": 3237 + }, + { + "epoch": 0.09988085552672268, + "grad_norm": 0.1818610578775406, + "learning_rate": 4.99383096853794e-05, + "loss": 3.3389, + "step": 3238 + }, + { + "epoch": 0.0999117019922961, + "grad_norm": 0.1860790103673935, + "learning_rate": 4.9953732264034546e-05, + "loss": 3.3224, + "step": 3239 + }, + { + "epoch": 0.09994254845786951, + "grad_norm": 0.18702733516693115, + "learning_rate": 4.9969154842689705e-05, + "loss": 3.3119, + "step": 3240 + }, + { + "epoch": 0.09997339492344293, + "grad_norm": 0.1856156587600708, + "learning_rate": 4.998457742134485e-05, + "loss": 3.3388, + "step": 3241 + }, + { + "epoch": 0.10000424138901634, + "grad_norm": 0.18047502636909485, + "learning_rate": 5e-05, + "loss": 3.3384, + "step": 3242 + }, + { + "epoch": 0.10003508785458975, + "grad_norm": 0.18722227215766907, + "learning_rate": 4.999999985507001e-05, + "loss": 3.3472, + "step": 3243 + }, + { + "epoch": 0.10006593432016318, + "grad_norm": 0.1855047345161438, + "learning_rate": 4.9999999420280055e-05, + "loss": 3.3155, + "step": 3244 + }, + { + "epoch": 0.1000967807857366, + "grad_norm": 0.19277696311473846, + "learning_rate": 4.999999869563012e-05, + "loss": 3.3099, + "step": 3245 + }, + { + "epoch": 0.10012762725131001, + "grad_norm": 0.5375243425369263, + "learning_rate": 4.999999768112023e-05, + "loss": 3.3458, + "step": 3246 + }, + { + "epoch": 0.10015847371688343, + "grad_norm": 0.1934397965669632, + "learning_rate": 4.999999637675038e-05, + "loss": 3.3001, + "step": 3247 + }, + { + "epoch": 0.10018932018245684, + "grad_norm": 0.19725164771080017, + "learning_rate": 4.99999947825206e-05, + "loss": 3.2987, + "step": 3248 + }, + { + "epoch": 0.10022016664803025, + "grad_norm": 0.1836722493171692, + "learning_rate": 4.99999928984309e-05, + "loss": 3.3364, + "step": 3249 + }, + { + "epoch": 0.10025101311360368, + "grad_norm": 0.18390928208827972, + "learning_rate": 4.999999072448131e-05, + "loss": 3.3268, + "step": 3250 + }, + { + "epoch": 0.1002818595791771, + "grad_norm": 0.19068914651870728, + "learning_rate": 4.999998826067185e-05, + "loss": 3.2987, + "step": 3251 + }, + { + "epoch": 0.10031270604475051, + "grad_norm": 0.19883136451244354, + "learning_rate": 4.999998550700255e-05, + "loss": 3.3221, + "step": 3252 + }, + { + "epoch": 0.10034355251032392, + "grad_norm": 0.18167293071746826, + "learning_rate": 4.9999982463473443e-05, + "loss": 3.2726, + "step": 3253 + }, + { + "epoch": 0.10037439897589734, + "grad_norm": 0.19155319035053253, + "learning_rate": 4.999997913008456e-05, + "loss": 3.2835, + "step": 3254 + }, + { + "epoch": 0.10040524544147075, + "grad_norm": 0.22523657977581024, + "learning_rate": 4.999997550683595e-05, + "loss": 3.3477, + "step": 3255 + }, + { + "epoch": 0.10043609190704418, + "grad_norm": 0.1894044280052185, + "learning_rate": 4.999997159372763e-05, + "loss": 3.3099, + "step": 3256 + }, + { + "epoch": 0.1004669383726176, + "grad_norm": 0.1807062327861786, + "learning_rate": 4.9999967390759675e-05, + "loss": 3.3253, + "step": 3257 + }, + { + "epoch": 0.10049778483819101, + "grad_norm": 0.199574813246727, + "learning_rate": 4.999996289793212e-05, + "loss": 3.3429, + "step": 3258 + }, + { + "epoch": 0.10052863130376442, + "grad_norm": 0.18564322590827942, + "learning_rate": 4.999995811524501e-05, + "loss": 3.3302, + "step": 3259 + }, + { + "epoch": 0.10055947776933784, + "grad_norm": 0.19742420315742493, + "learning_rate": 4.999995304269842e-05, + "loss": 3.3238, + "step": 3260 + }, + { + "epoch": 0.10059032423491125, + "grad_norm": 0.1955946534872055, + "learning_rate": 4.9999947680292395e-05, + "loss": 3.3108, + "step": 3261 + }, + { + "epoch": 0.10062117070048468, + "grad_norm": 0.19052699208259583, + "learning_rate": 4.9999942028027e-05, + "loss": 3.3411, + "step": 3262 + }, + { + "epoch": 0.1006520171660581, + "grad_norm": 0.1937655806541443, + "learning_rate": 4.99999360859023e-05, + "loss": 3.3075, + "step": 3263 + }, + { + "epoch": 0.10068286363163151, + "grad_norm": 0.18422894179821014, + "learning_rate": 4.9999929853918364e-05, + "loss": 3.322, + "step": 3264 + }, + { + "epoch": 0.10071371009720492, + "grad_norm": 0.18531158566474915, + "learning_rate": 4.999992333207527e-05, + "loss": 3.3341, + "step": 3265 + }, + { + "epoch": 0.10074455656277834, + "grad_norm": 0.18828345835208893, + "learning_rate": 4.999991652037308e-05, + "loss": 3.2993, + "step": 3266 + }, + { + "epoch": 0.10077540302835175, + "grad_norm": 0.2402905523777008, + "learning_rate": 4.999990941881188e-05, + "loss": 3.3135, + "step": 3267 + }, + { + "epoch": 0.10080624949392518, + "grad_norm": 0.19088023900985718, + "learning_rate": 4.999990202739176e-05, + "loss": 3.2992, + "step": 3268 + }, + { + "epoch": 0.1008370959594986, + "grad_norm": 0.17918163537979126, + "learning_rate": 4.99998943461128e-05, + "loss": 3.2988, + "step": 3269 + }, + { + "epoch": 0.10086794242507201, + "grad_norm": 0.20269142091274261, + "learning_rate": 4.999988637497508e-05, + "loss": 3.3238, + "step": 3270 + }, + { + "epoch": 0.10089878889064542, + "grad_norm": 0.18483948707580566, + "learning_rate": 4.999987811397871e-05, + "loss": 3.3154, + "step": 3271 + }, + { + "epoch": 0.10092963535621884, + "grad_norm": 0.1899690330028534, + "learning_rate": 4.999986956312377e-05, + "loss": 3.2806, + "step": 3272 + }, + { + "epoch": 0.10096048182179225, + "grad_norm": 0.1814526468515396, + "learning_rate": 4.9999860722410365e-05, + "loss": 3.3286, + "step": 3273 + }, + { + "epoch": 0.10099132828736568, + "grad_norm": 0.185224249958992, + "learning_rate": 4.99998515918386e-05, + "loss": 3.3255, + "step": 3274 + }, + { + "epoch": 0.10102217475293909, + "grad_norm": 0.19371874630451202, + "learning_rate": 4.999984217140858e-05, + "loss": 3.3101, + "step": 3275 + }, + { + "epoch": 0.1010530212185125, + "grad_norm": 0.1833687126636505, + "learning_rate": 4.999983246112041e-05, + "loss": 3.3493, + "step": 3276 + }, + { + "epoch": 0.10108386768408592, + "grad_norm": 0.1975378543138504, + "learning_rate": 4.999982246097421e-05, + "loss": 3.3204, + "step": 3277 + }, + { + "epoch": 0.10111471414965933, + "grad_norm": 0.1872277408838272, + "learning_rate": 4.999981217097009e-05, + "loss": 3.3274, + "step": 3278 + }, + { + "epoch": 0.10114556061523275, + "grad_norm": 0.1906963586807251, + "learning_rate": 4.999980159110817e-05, + "loss": 3.3369, + "step": 3279 + }, + { + "epoch": 0.10117640708080618, + "grad_norm": 0.17991626262664795, + "learning_rate": 4.9999790721388574e-05, + "loss": 3.309, + "step": 3280 + }, + { + "epoch": 0.10120725354637959, + "grad_norm": 0.18799155950546265, + "learning_rate": 4.9999779561811426e-05, + "loss": 3.3345, + "step": 3281 + }, + { + "epoch": 0.101238100011953, + "grad_norm": 0.1892252117395401, + "learning_rate": 4.999976811237686e-05, + "loss": 3.3014, + "step": 3282 + }, + { + "epoch": 0.10126894647752642, + "grad_norm": 0.17889635264873505, + "learning_rate": 4.999975637308501e-05, + "loss": 3.3084, + "step": 3283 + }, + { + "epoch": 0.10129979294309983, + "grad_norm": 0.2070685178041458, + "learning_rate": 4.9999744343936e-05, + "loss": 3.3009, + "step": 3284 + }, + { + "epoch": 0.10133063940867325, + "grad_norm": 0.19796882569789886, + "learning_rate": 4.999973202492998e-05, + "loss": 3.296, + "step": 3285 + }, + { + "epoch": 0.10136148587424668, + "grad_norm": 0.1839163601398468, + "learning_rate": 4.999971941606709e-05, + "loss": 3.3058, + "step": 3286 + }, + { + "epoch": 0.10139233233982009, + "grad_norm": 0.19450467824935913, + "learning_rate": 4.9999706517347475e-05, + "loss": 3.3298, + "step": 3287 + }, + { + "epoch": 0.1014231788053935, + "grad_norm": 0.20226646959781647, + "learning_rate": 4.9999693328771294e-05, + "loss": 3.3196, + "step": 3288 + }, + { + "epoch": 0.10145402527096692, + "grad_norm": 0.18587514758110046, + "learning_rate": 4.999967985033869e-05, + "loss": 3.3493, + "step": 3289 + }, + { + "epoch": 0.10148487173654033, + "grad_norm": 0.1835005134344101, + "learning_rate": 4.9999666082049814e-05, + "loss": 3.3139, + "step": 3290 + }, + { + "epoch": 0.10151571820211375, + "grad_norm": 0.19689838588237762, + "learning_rate": 4.999965202390483e-05, + "loss": 3.3377, + "step": 3291 + }, + { + "epoch": 0.10154656466768718, + "grad_norm": 0.20360714197158813, + "learning_rate": 4.9999637675903916e-05, + "loss": 3.3112, + "step": 3292 + }, + { + "epoch": 0.10157741113326059, + "grad_norm": 0.19090254604816437, + "learning_rate": 4.999962303804721e-05, + "loss": 3.3177, + "step": 3293 + }, + { + "epoch": 0.101608257598834, + "grad_norm": 0.20364172756671906, + "learning_rate": 4.9999608110334915e-05, + "loss": 3.2539, + "step": 3294 + }, + { + "epoch": 0.10163910406440742, + "grad_norm": 0.17960935831069946, + "learning_rate": 4.999959289276718e-05, + "loss": 3.3049, + "step": 3295 + }, + { + "epoch": 0.10166995052998083, + "grad_norm": 0.189431831240654, + "learning_rate": 4.999957738534419e-05, + "loss": 3.3436, + "step": 3296 + }, + { + "epoch": 0.10170079699555425, + "grad_norm": 0.18327675759792328, + "learning_rate": 4.9999561588066123e-05, + "loss": 3.2994, + "step": 3297 + }, + { + "epoch": 0.10173164346112767, + "grad_norm": 0.18027912080287933, + "learning_rate": 4.999954550093316e-05, + "loss": 3.3277, + "step": 3298 + }, + { + "epoch": 0.10176248992670109, + "grad_norm": 0.18122026324272156, + "learning_rate": 4.999952912394549e-05, + "loss": 3.3205, + "step": 3299 + }, + { + "epoch": 0.1017933363922745, + "grad_norm": 0.17647527158260345, + "learning_rate": 4.99995124571033e-05, + "loss": 3.3176, + "step": 3300 + }, + { + "epoch": 0.10182418285784792, + "grad_norm": 0.18041309714317322, + "learning_rate": 4.99994955004068e-05, + "loss": 3.3093, + "step": 3301 + }, + { + "epoch": 0.10185502932342133, + "grad_norm": 0.3042145371437073, + "learning_rate": 4.9999478253856155e-05, + "loss": 3.312, + "step": 3302 + }, + { + "epoch": 0.10188587578899476, + "grad_norm": 0.2950230836868286, + "learning_rate": 4.9999460717451596e-05, + "loss": 3.3154, + "step": 3303 + }, + { + "epoch": 0.10191672225456817, + "grad_norm": 0.17624938488006592, + "learning_rate": 4.9999442891193304e-05, + "loss": 3.3071, + "step": 3304 + }, + { + "epoch": 0.10194756872014159, + "grad_norm": 0.1898709386587143, + "learning_rate": 4.9999424775081496e-05, + "loss": 3.3016, + "step": 3305 + }, + { + "epoch": 0.101978415185715, + "grad_norm": 0.17497974634170532, + "learning_rate": 4.999940636911638e-05, + "loss": 3.3036, + "step": 3306 + }, + { + "epoch": 0.10200926165128842, + "grad_norm": 0.18768882751464844, + "learning_rate": 4.999938767329818e-05, + "loss": 3.3373, + "step": 3307 + }, + { + "epoch": 0.10204010811686183, + "grad_norm": 0.1884714514017105, + "learning_rate": 4.999936868762709e-05, + "loss": 3.3057, + "step": 3308 + }, + { + "epoch": 0.10207095458243526, + "grad_norm": 0.1814107745885849, + "learning_rate": 4.999934941210335e-05, + "loss": 3.3048, + "step": 3309 + }, + { + "epoch": 0.10210180104800867, + "grad_norm": 0.1834474354982376, + "learning_rate": 4.9999329846727175e-05, + "loss": 3.3184, + "step": 3310 + }, + { + "epoch": 0.10213264751358209, + "grad_norm": 0.17832155525684357, + "learning_rate": 4.999930999149879e-05, + "loss": 3.3225, + "step": 3311 + }, + { + "epoch": 0.1021634939791555, + "grad_norm": 0.18951041996479034, + "learning_rate": 4.999928984641843e-05, + "loss": 3.3021, + "step": 3312 + }, + { + "epoch": 0.10219434044472891, + "grad_norm": 0.18992483615875244, + "learning_rate": 4.999926941148633e-05, + "loss": 3.3013, + "step": 3313 + }, + { + "epoch": 0.10222518691030233, + "grad_norm": 0.2016969472169876, + "learning_rate": 4.999924868670273e-05, + "loss": 3.3266, + "step": 3314 + }, + { + "epoch": 0.10225603337587576, + "grad_norm": 0.21133044362068176, + "learning_rate": 4.999922767206785e-05, + "loss": 3.3556, + "step": 3315 + }, + { + "epoch": 0.10228687984144917, + "grad_norm": 0.18829838931560516, + "learning_rate": 4.999920636758195e-05, + "loss": 3.3323, + "step": 3316 + }, + { + "epoch": 0.10231772630702259, + "grad_norm": 0.18533146381378174, + "learning_rate": 4.999918477324529e-05, + "loss": 3.297, + "step": 3317 + }, + { + "epoch": 0.102348572772596, + "grad_norm": 0.1869489848613739, + "learning_rate": 4.999916288905809e-05, + "loss": 3.3027, + "step": 3318 + }, + { + "epoch": 0.10237941923816941, + "grad_norm": 0.1905023753643036, + "learning_rate": 4.999914071502062e-05, + "loss": 3.3205, + "step": 3319 + }, + { + "epoch": 0.10241026570374283, + "grad_norm": 0.18320856988430023, + "learning_rate": 4.999911825113315e-05, + "loss": 3.333, + "step": 3320 + }, + { + "epoch": 0.10244111216931626, + "grad_norm": 0.19005292654037476, + "learning_rate": 4.999909549739591e-05, + "loss": 3.314, + "step": 3321 + }, + { + "epoch": 0.10247195863488967, + "grad_norm": 0.180525541305542, + "learning_rate": 4.999907245380919e-05, + "loss": 3.3095, + "step": 3322 + }, + { + "epoch": 0.10250280510046308, + "grad_norm": 0.18411119282245636, + "learning_rate": 4.999904912037324e-05, + "loss": 3.316, + "step": 3323 + }, + { + "epoch": 0.1025336515660365, + "grad_norm": 0.18557080626487732, + "learning_rate": 4.999902549708835e-05, + "loss": 3.2957, + "step": 3324 + }, + { + "epoch": 0.10256449803160991, + "grad_norm": 0.22438733279705048, + "learning_rate": 4.9999001583954776e-05, + "loss": 3.3148, + "step": 3325 + }, + { + "epoch": 0.10259534449718333, + "grad_norm": 0.1872452199459076, + "learning_rate": 4.99989773809728e-05, + "loss": 3.2999, + "step": 3326 + }, + { + "epoch": 0.10262619096275676, + "grad_norm": 0.19069842994213104, + "learning_rate": 4.99989528881427e-05, + "loss": 3.3474, + "step": 3327 + }, + { + "epoch": 0.10265703742833017, + "grad_norm": 0.1779751479625702, + "learning_rate": 4.999892810546477e-05, + "loss": 3.3659, + "step": 3328 + }, + { + "epoch": 0.10268788389390358, + "grad_norm": 0.19572308659553528, + "learning_rate": 4.999890303293928e-05, + "loss": 3.3059, + "step": 3329 + }, + { + "epoch": 0.102718730359477, + "grad_norm": 0.18305616080760956, + "learning_rate": 4.9998877670566545e-05, + "loss": 3.3334, + "step": 3330 + }, + { + "epoch": 0.10274957682505041, + "grad_norm": 0.17881959676742554, + "learning_rate": 4.9998852018346844e-05, + "loss": 3.3122, + "step": 3331 + }, + { + "epoch": 0.10278042329062383, + "grad_norm": 0.1765105128288269, + "learning_rate": 4.999882607628048e-05, + "loss": 3.2964, + "step": 3332 + }, + { + "epoch": 0.10281126975619725, + "grad_norm": 0.19179509580135345, + "learning_rate": 4.999879984436774e-05, + "loss": 3.2903, + "step": 3333 + }, + { + "epoch": 0.10284211622177067, + "grad_norm": 0.17893697321414948, + "learning_rate": 4.9998773322608936e-05, + "loss": 3.3255, + "step": 3334 + }, + { + "epoch": 0.10287296268734408, + "grad_norm": 0.1871093511581421, + "learning_rate": 4.999874651100439e-05, + "loss": 3.349, + "step": 3335 + }, + { + "epoch": 0.1029038091529175, + "grad_norm": 0.1941724270582199, + "learning_rate": 4.9998719409554394e-05, + "loss": 3.303, + "step": 3336 + }, + { + "epoch": 0.10293465561849091, + "grad_norm": 0.19404487311840057, + "learning_rate": 4.999869201825928e-05, + "loss": 3.3161, + "step": 3337 + }, + { + "epoch": 0.10296550208406433, + "grad_norm": 0.17863723635673523, + "learning_rate": 4.999866433711934e-05, + "loss": 3.2956, + "step": 3338 + }, + { + "epoch": 0.10299634854963775, + "grad_norm": 0.18528734147548676, + "learning_rate": 4.999863636613491e-05, + "loss": 3.3294, + "step": 3339 + }, + { + "epoch": 0.10302719501521117, + "grad_norm": 0.19432272017002106, + "learning_rate": 4.999860810530632e-05, + "loss": 3.3243, + "step": 3340 + }, + { + "epoch": 0.10305804148078458, + "grad_norm": 0.18716971576213837, + "learning_rate": 4.999857955463389e-05, + "loss": 3.2987, + "step": 3341 + }, + { + "epoch": 0.103088887946358, + "grad_norm": 0.17419342696666718, + "learning_rate": 4.9998550714117956e-05, + "loss": 3.3561, + "step": 3342 + }, + { + "epoch": 0.10311973441193141, + "grad_norm": 0.19633474946022034, + "learning_rate": 4.999852158375884e-05, + "loss": 3.3143, + "step": 3343 + }, + { + "epoch": 0.10315058087750482, + "grad_norm": 0.18125250935554504, + "learning_rate": 4.9998492163556896e-05, + "loss": 3.2986, + "step": 3344 + }, + { + "epoch": 0.10318142734307825, + "grad_norm": 0.18899033963680267, + "learning_rate": 4.999846245351245e-05, + "loss": 3.3078, + "step": 3345 + }, + { + "epoch": 0.10321227380865167, + "grad_norm": 0.1853909194469452, + "learning_rate": 4.9998432453625865e-05, + "loss": 3.2993, + "step": 3346 + }, + { + "epoch": 0.10324312027422508, + "grad_norm": 0.17944081127643585, + "learning_rate": 4.999840216389747e-05, + "loss": 3.2738, + "step": 3347 + }, + { + "epoch": 0.1032739667397985, + "grad_norm": 0.1972704380750656, + "learning_rate": 4.999837158432763e-05, + "loss": 3.3202, + "step": 3348 + }, + { + "epoch": 0.10330481320537191, + "grad_norm": 0.18860214948654175, + "learning_rate": 4.999834071491669e-05, + "loss": 3.3133, + "step": 3349 + }, + { + "epoch": 0.10333565967094532, + "grad_norm": 0.23290641605854034, + "learning_rate": 4.9998309555665014e-05, + "loss": 3.3097, + "step": 3350 + }, + { + "epoch": 0.10336650613651875, + "grad_norm": 0.17754070460796356, + "learning_rate": 4.999827810657295e-05, + "loss": 3.3339, + "step": 3351 + }, + { + "epoch": 0.10339735260209217, + "grad_norm": 0.18130706250667572, + "learning_rate": 4.9998246367640886e-05, + "loss": 3.2935, + "step": 3352 + }, + { + "epoch": 0.10342819906766558, + "grad_norm": 0.17560678720474243, + "learning_rate": 4.9998214338869175e-05, + "loss": 3.2955, + "step": 3353 + }, + { + "epoch": 0.103459045533239, + "grad_norm": 0.19405971467494965, + "learning_rate": 4.999818202025819e-05, + "loss": 3.3413, + "step": 3354 + }, + { + "epoch": 0.10348989199881241, + "grad_norm": 0.18025660514831543, + "learning_rate": 4.99981494118083e-05, + "loss": 3.3304, + "step": 3355 + }, + { + "epoch": 0.10352073846438582, + "grad_norm": 0.19806750118732452, + "learning_rate": 4.99981165135199e-05, + "loss": 3.3268, + "step": 3356 + }, + { + "epoch": 0.10355158492995925, + "grad_norm": 0.1962517946958542, + "learning_rate": 4.999808332539336e-05, + "loss": 3.3393, + "step": 3357 + }, + { + "epoch": 0.10358243139553266, + "grad_norm": 0.18349343538284302, + "learning_rate": 4.999804984742906e-05, + "loss": 3.2745, + "step": 3358 + }, + { + "epoch": 0.10361327786110608, + "grad_norm": 0.19720867276191711, + "learning_rate": 4.99980160796274e-05, + "loss": 3.3147, + "step": 3359 + }, + { + "epoch": 0.1036441243266795, + "grad_norm": 0.17655695974826813, + "learning_rate": 4.999798202198877e-05, + "loss": 3.3084, + "step": 3360 + }, + { + "epoch": 0.10367497079225291, + "grad_norm": 0.19355450570583344, + "learning_rate": 4.9997947674513545e-05, + "loss": 3.3073, + "step": 3361 + }, + { + "epoch": 0.10370581725782632, + "grad_norm": 0.18428459763526917, + "learning_rate": 4.999791303720215e-05, + "loss": 3.3322, + "step": 3362 + }, + { + "epoch": 0.10373666372339975, + "grad_norm": 0.18302352726459503, + "learning_rate": 4.9997878110054975e-05, + "loss": 3.3564, + "step": 3363 + }, + { + "epoch": 0.10376751018897316, + "grad_norm": 0.17808756232261658, + "learning_rate": 4.999784289307242e-05, + "loss": 3.2903, + "step": 3364 + }, + { + "epoch": 0.10379835665454658, + "grad_norm": 0.18380799889564514, + "learning_rate": 4.99978073862549e-05, + "loss": 3.3526, + "step": 3365 + }, + { + "epoch": 0.10382920312011999, + "grad_norm": 0.16822147369384766, + "learning_rate": 4.999777158960283e-05, + "loss": 3.2921, + "step": 3366 + }, + { + "epoch": 0.1038600495856934, + "grad_norm": 0.17791573703289032, + "learning_rate": 4.999773550311662e-05, + "loss": 3.2945, + "step": 3367 + }, + { + "epoch": 0.10389089605126682, + "grad_norm": 0.1754847913980484, + "learning_rate": 4.999769912679668e-05, + "loss": 3.3334, + "step": 3368 + }, + { + "epoch": 0.10392174251684025, + "grad_norm": 0.18128268420696259, + "learning_rate": 4.999766246064345e-05, + "loss": 3.3131, + "step": 3369 + }, + { + "epoch": 0.10395258898241366, + "grad_norm": 0.16844724118709564, + "learning_rate": 4.999762550465734e-05, + "loss": 3.3241, + "step": 3370 + }, + { + "epoch": 0.10398343544798708, + "grad_norm": 0.190501406788826, + "learning_rate": 4.999758825883879e-05, + "loss": 3.3037, + "step": 3371 + }, + { + "epoch": 0.10401428191356049, + "grad_norm": 0.1707083284854889, + "learning_rate": 4.9997550723188227e-05, + "loss": 3.3265, + "step": 3372 + }, + { + "epoch": 0.1040451283791339, + "grad_norm": 0.18640010058879852, + "learning_rate": 4.999751289770608e-05, + "loss": 3.3434, + "step": 3373 + }, + { + "epoch": 0.10407597484470732, + "grad_norm": 0.17536069452762604, + "learning_rate": 4.9997474782392786e-05, + "loss": 3.3065, + "step": 3374 + }, + { + "epoch": 0.10410682131028075, + "grad_norm": 0.1784447282552719, + "learning_rate": 4.9997436377248805e-05, + "loss": 3.3395, + "step": 3375 + }, + { + "epoch": 0.10413766777585416, + "grad_norm": 0.18110015988349915, + "learning_rate": 4.999739768227456e-05, + "loss": 3.2816, + "step": 3376 + }, + { + "epoch": 0.10416851424142758, + "grad_norm": 0.1779746264219284, + "learning_rate": 4.999735869747052e-05, + "loss": 3.3245, + "step": 3377 + }, + { + "epoch": 0.10419936070700099, + "grad_norm": 0.18244200944900513, + "learning_rate": 4.999731942283713e-05, + "loss": 3.3097, + "step": 3378 + }, + { + "epoch": 0.1042302071725744, + "grad_norm": 0.18767547607421875, + "learning_rate": 4.999727985837483e-05, + "loss": 3.2888, + "step": 3379 + }, + { + "epoch": 0.10426105363814782, + "grad_norm": 0.17650184035301208, + "learning_rate": 4.99972400040841e-05, + "loss": 3.3262, + "step": 3380 + }, + { + "epoch": 0.10429190010372125, + "grad_norm": 0.1800835281610489, + "learning_rate": 4.999719985996539e-05, + "loss": 3.2801, + "step": 3381 + }, + { + "epoch": 0.10432274656929466, + "grad_norm": 0.19015465676784515, + "learning_rate": 4.999715942601918e-05, + "loss": 3.3221, + "step": 3382 + }, + { + "epoch": 0.10435359303486808, + "grad_norm": 0.17965519428253174, + "learning_rate": 4.999711870224592e-05, + "loss": 3.3002, + "step": 3383 + }, + { + "epoch": 0.10438443950044149, + "grad_norm": 0.49841514229774475, + "learning_rate": 4.999707768864609e-05, + "loss": 3.3411, + "step": 3384 + }, + { + "epoch": 0.1044152859660149, + "grad_norm": 0.18489326536655426, + "learning_rate": 4.9997036385220164e-05, + "loss": 3.3341, + "step": 3385 + }, + { + "epoch": 0.10444613243158832, + "grad_norm": 0.18158771097660065, + "learning_rate": 4.999699479196863e-05, + "loss": 3.3053, + "step": 3386 + }, + { + "epoch": 0.10447697889716175, + "grad_norm": 0.17973248660564423, + "learning_rate": 4.999695290889195e-05, + "loss": 3.2987, + "step": 3387 + }, + { + "epoch": 0.10450782536273516, + "grad_norm": 0.17849789559841156, + "learning_rate": 4.999691073599063e-05, + "loss": 3.3156, + "step": 3388 + }, + { + "epoch": 0.10453867182830857, + "grad_norm": 0.18805348873138428, + "learning_rate": 4.999686827326515e-05, + "loss": 3.317, + "step": 3389 + }, + { + "epoch": 0.10456951829388199, + "grad_norm": 0.17653459310531616, + "learning_rate": 4.999682552071601e-05, + "loss": 3.3084, + "step": 3390 + }, + { + "epoch": 0.1046003647594554, + "grad_norm": 0.19058586657047272, + "learning_rate": 4.9996782478343696e-05, + "loss": 3.3316, + "step": 3391 + }, + { + "epoch": 0.10463121122502882, + "grad_norm": 0.19452250003814697, + "learning_rate": 4.9996739146148707e-05, + "loss": 3.2952, + "step": 3392 + }, + { + "epoch": 0.10466205769060224, + "grad_norm": 0.18459412455558777, + "learning_rate": 4.9996695524131555e-05, + "loss": 3.3394, + "step": 3393 + }, + { + "epoch": 0.10469290415617566, + "grad_norm": 0.1846790760755539, + "learning_rate": 4.9996651612292736e-05, + "loss": 3.2865, + "step": 3394 + }, + { + "epoch": 0.10472375062174907, + "grad_norm": 0.17277312278747559, + "learning_rate": 4.999660741063276e-05, + "loss": 3.3192, + "step": 3395 + }, + { + "epoch": 0.10475459708732249, + "grad_norm": 0.17733895778656006, + "learning_rate": 4.999656291915215e-05, + "loss": 3.317, + "step": 3396 + }, + { + "epoch": 0.1047854435528959, + "grad_norm": 0.1796792894601822, + "learning_rate": 4.99965181378514e-05, + "loss": 3.3183, + "step": 3397 + }, + { + "epoch": 0.10481629001846932, + "grad_norm": 0.18023866415023804, + "learning_rate": 4.999647306673105e-05, + "loss": 3.3497, + "step": 3398 + }, + { + "epoch": 0.10484713648404274, + "grad_norm": 0.18083477020263672, + "learning_rate": 4.999642770579163e-05, + "loss": 3.3302, + "step": 3399 + }, + { + "epoch": 0.10487798294961616, + "grad_norm": 0.17874060571193695, + "learning_rate": 4.999638205503363e-05, + "loss": 3.3375, + "step": 3400 + }, + { + "epoch": 0.10490882941518957, + "grad_norm": 0.1830933392047882, + "learning_rate": 4.999633611445762e-05, + "loss": 3.3377, + "step": 3401 + }, + { + "epoch": 0.10493967588076299, + "grad_norm": 0.1655663549900055, + "learning_rate": 4.9996289884064094e-05, + "loss": 3.3006, + "step": 3402 + }, + { + "epoch": 0.1049705223463364, + "grad_norm": 0.17937901616096497, + "learning_rate": 4.9996243363853626e-05, + "loss": 3.3237, + "step": 3403 + }, + { + "epoch": 0.10500136881190981, + "grad_norm": 0.1713716834783554, + "learning_rate": 4.999619655382673e-05, + "loss": 3.3086, + "step": 3404 + }, + { + "epoch": 0.10503221527748324, + "grad_norm": 0.1902548223733902, + "learning_rate": 4.999614945398395e-05, + "loss": 3.2722, + "step": 3405 + }, + { + "epoch": 0.10506306174305666, + "grad_norm": 0.17114098370075226, + "learning_rate": 4.999610206432584e-05, + "loss": 3.2982, + "step": 3406 + }, + { + "epoch": 0.10509390820863007, + "grad_norm": 0.17967788875102997, + "learning_rate": 4.999605438485296e-05, + "loss": 3.2894, + "step": 3407 + }, + { + "epoch": 0.10512475467420349, + "grad_norm": 0.18977786600589752, + "learning_rate": 4.9996006415565846e-05, + "loss": 3.3117, + "step": 3408 + }, + { + "epoch": 0.1051556011397769, + "grad_norm": 0.1926335096359253, + "learning_rate": 4.999595815646505e-05, + "loss": 3.3132, + "step": 3409 + }, + { + "epoch": 0.10518644760535031, + "grad_norm": 0.194233700633049, + "learning_rate": 4.9995909607551144e-05, + "loss": 3.3139, + "step": 3410 + }, + { + "epoch": 0.10521729407092374, + "grad_norm": 0.1748219132423401, + "learning_rate": 4.9995860768824685e-05, + "loss": 3.3277, + "step": 3411 + }, + { + "epoch": 0.10524814053649716, + "grad_norm": 0.18551526963710785, + "learning_rate": 4.999581164028624e-05, + "loss": 3.3465, + "step": 3412 + }, + { + "epoch": 0.10527898700207057, + "grad_norm": 0.1864136904478073, + "learning_rate": 4.999576222193639e-05, + "loss": 3.3047, + "step": 3413 + }, + { + "epoch": 0.10530983346764398, + "grad_norm": 0.18414752185344696, + "learning_rate": 4.999571251377569e-05, + "loss": 3.3477, + "step": 3414 + }, + { + "epoch": 0.1053406799332174, + "grad_norm": 0.19219420850276947, + "learning_rate": 4.999566251580472e-05, + "loss": 3.3459, + "step": 3415 + }, + { + "epoch": 0.10537152639879081, + "grad_norm": 0.19528885185718536, + "learning_rate": 4.999561222802407e-05, + "loss": 3.2927, + "step": 3416 + }, + { + "epoch": 0.10540237286436424, + "grad_norm": 0.18277589976787567, + "learning_rate": 4.9995561650434316e-05, + "loss": 3.3271, + "step": 3417 + }, + { + "epoch": 0.10543321932993766, + "grad_norm": 0.18709416687488556, + "learning_rate": 4.999551078303605e-05, + "loss": 3.3269, + "step": 3418 + }, + { + "epoch": 0.10546406579551107, + "grad_norm": 0.19617217779159546, + "learning_rate": 4.9995459625829846e-05, + "loss": 3.333, + "step": 3419 + }, + { + "epoch": 0.10549491226108448, + "grad_norm": 0.17565526068210602, + "learning_rate": 4.999540817881631e-05, + "loss": 3.2804, + "step": 3420 + }, + { + "epoch": 0.1055257587266579, + "grad_norm": 0.17910531163215637, + "learning_rate": 4.999535644199604e-05, + "loss": 3.2994, + "step": 3421 + }, + { + "epoch": 0.10555660519223131, + "grad_norm": 0.1847343146800995, + "learning_rate": 4.999530441536964e-05, + "loss": 3.3179, + "step": 3422 + }, + { + "epoch": 0.10558745165780474, + "grad_norm": 0.18192370235919952, + "learning_rate": 4.99952520989377e-05, + "loss": 3.2848, + "step": 3423 + }, + { + "epoch": 0.10561829812337815, + "grad_norm": 0.17899948358535767, + "learning_rate": 4.9995199492700826e-05, + "loss": 3.3153, + "step": 3424 + }, + { + "epoch": 0.10564914458895157, + "grad_norm": 0.19504155218601227, + "learning_rate": 4.999514659665964e-05, + "loss": 3.2574, + "step": 3425 + }, + { + "epoch": 0.10567999105452498, + "grad_norm": 0.18261562287807465, + "learning_rate": 4.999509341081475e-05, + "loss": 3.3158, + "step": 3426 + }, + { + "epoch": 0.1057108375200984, + "grad_norm": 0.1993151158094406, + "learning_rate": 4.999503993516678e-05, + "loss": 3.3257, + "step": 3427 + }, + { + "epoch": 0.10574168398567181, + "grad_norm": 0.1814647912979126, + "learning_rate": 4.999498616971633e-05, + "loss": 3.2912, + "step": 3428 + }, + { + "epoch": 0.10577253045124524, + "grad_norm": 0.18754279613494873, + "learning_rate": 4.999493211446403e-05, + "loss": 3.2674, + "step": 3429 + }, + { + "epoch": 0.10580337691681865, + "grad_norm": 0.18945197761058807, + "learning_rate": 4.9994877769410525e-05, + "loss": 3.3106, + "step": 3430 + }, + { + "epoch": 0.10583422338239207, + "grad_norm": 0.19297227263450623, + "learning_rate": 4.999482313455643e-05, + "loss": 3.3157, + "step": 3431 + }, + { + "epoch": 0.10586506984796548, + "grad_norm": 0.22474107146263123, + "learning_rate": 4.999476820990238e-05, + "loss": 3.3333, + "step": 3432 + }, + { + "epoch": 0.1058959163135389, + "grad_norm": 0.1940387636423111, + "learning_rate": 4.9994712995449014e-05, + "loss": 3.3087, + "step": 3433 + }, + { + "epoch": 0.10592676277911231, + "grad_norm": 0.18936224281787872, + "learning_rate": 4.999465749119697e-05, + "loss": 3.3112, + "step": 3434 + }, + { + "epoch": 0.10595760924468574, + "grad_norm": 0.19818831980228424, + "learning_rate": 4.999460169714689e-05, + "loss": 3.2911, + "step": 3435 + }, + { + "epoch": 0.10598845571025915, + "grad_norm": 0.18859925866127014, + "learning_rate": 4.9994545613299423e-05, + "loss": 3.3278, + "step": 3436 + }, + { + "epoch": 0.10601930217583257, + "grad_norm": 0.20131611824035645, + "learning_rate": 4.999448923965523e-05, + "loss": 3.2903, + "step": 3437 + }, + { + "epoch": 0.10605014864140598, + "grad_norm": 0.19577428698539734, + "learning_rate": 4.9994432576214936e-05, + "loss": 3.3227, + "step": 3438 + }, + { + "epoch": 0.1060809951069794, + "grad_norm": 0.19605201482772827, + "learning_rate": 4.999437562297923e-05, + "loss": 3.3258, + "step": 3439 + }, + { + "epoch": 0.10611184157255281, + "grad_norm": 0.1802118867635727, + "learning_rate": 4.999431837994875e-05, + "loss": 3.2856, + "step": 3440 + }, + { + "epoch": 0.10614268803812624, + "grad_norm": 0.19358879327774048, + "learning_rate": 4.999426084712418e-05, + "loss": 3.3099, + "step": 3441 + }, + { + "epoch": 0.10617353450369965, + "grad_norm": 0.19701731204986572, + "learning_rate": 4.9994203024506166e-05, + "loss": 3.3135, + "step": 3442 + }, + { + "epoch": 0.10620438096927307, + "grad_norm": 0.1819622665643692, + "learning_rate": 4.999414491209539e-05, + "loss": 3.3159, + "step": 3443 + }, + { + "epoch": 0.10623522743484648, + "grad_norm": 0.19390825927257538, + "learning_rate": 4.999408650989252e-05, + "loss": 3.2833, + "step": 3444 + }, + { + "epoch": 0.1062660739004199, + "grad_norm": 0.19048139452934265, + "learning_rate": 4.9994027817898236e-05, + "loss": 3.3059, + "step": 3445 + }, + { + "epoch": 0.10629692036599331, + "grad_norm": 0.20054057240486145, + "learning_rate": 4.9993968836113224e-05, + "loss": 3.3372, + "step": 3446 + }, + { + "epoch": 0.10632776683156674, + "grad_norm": 0.17625309526920319, + "learning_rate": 4.999390956453816e-05, + "loss": 3.3102, + "step": 3447 + }, + { + "epoch": 0.10635861329714015, + "grad_norm": 0.18569478392601013, + "learning_rate": 4.9993850003173743e-05, + "loss": 3.3103, + "step": 3448 + }, + { + "epoch": 0.10638945976271356, + "grad_norm": 0.17970232665538788, + "learning_rate": 4.999379015202065e-05, + "loss": 3.2709, + "step": 3449 + }, + { + "epoch": 0.10642030622828698, + "grad_norm": 0.1818823218345642, + "learning_rate": 4.9993730011079575e-05, + "loss": 3.2791, + "step": 3450 + }, + { + "epoch": 0.10645115269386039, + "grad_norm": 0.18568813800811768, + "learning_rate": 4.999366958035122e-05, + "loss": 3.3055, + "step": 3451 + }, + { + "epoch": 0.10648199915943381, + "grad_norm": 0.17829188704490662, + "learning_rate": 4.999360885983629e-05, + "loss": 3.2938, + "step": 3452 + }, + { + "epoch": 0.10651284562500724, + "grad_norm": 0.19560858607292175, + "learning_rate": 4.999354784953548e-05, + "loss": 3.3222, + "step": 3453 + }, + { + "epoch": 0.10654369209058065, + "grad_norm": 0.18836185336112976, + "learning_rate": 4.999348654944951e-05, + "loss": 3.2849, + "step": 3454 + }, + { + "epoch": 0.10657453855615406, + "grad_norm": 0.1893051713705063, + "learning_rate": 4.999342495957907e-05, + "loss": 3.3295, + "step": 3455 + }, + { + "epoch": 0.10660538502172748, + "grad_norm": 0.1741720736026764, + "learning_rate": 4.99933630799249e-05, + "loss": 3.3018, + "step": 3456 + }, + { + "epoch": 0.10663623148730089, + "grad_norm": 0.1800323724746704, + "learning_rate": 4.99933009104877e-05, + "loss": 3.307, + "step": 3457 + }, + { + "epoch": 0.1066670779528743, + "grad_norm": 0.18508324027061462, + "learning_rate": 4.999323845126819e-05, + "loss": 3.3006, + "step": 3458 + }, + { + "epoch": 0.10669792441844773, + "grad_norm": 0.18977946043014526, + "learning_rate": 4.99931757022671e-05, + "loss": 3.3479, + "step": 3459 + }, + { + "epoch": 0.10672877088402115, + "grad_norm": 0.18039223551750183, + "learning_rate": 4.999311266348516e-05, + "loss": 3.3124, + "step": 3460 + }, + { + "epoch": 0.10675961734959456, + "grad_norm": 0.17955940961837769, + "learning_rate": 4.9993049334923103e-05, + "loss": 3.3246, + "step": 3461 + }, + { + "epoch": 0.10679046381516798, + "grad_norm": 0.1785697340965271, + "learning_rate": 4.999298571658165e-05, + "loss": 3.2967, + "step": 3462 + }, + { + "epoch": 0.10682131028074139, + "grad_norm": 0.17975465953350067, + "learning_rate": 4.999292180846156e-05, + "loss": 3.3045, + "step": 3463 + }, + { + "epoch": 0.1068521567463148, + "grad_norm": 0.17501945793628693, + "learning_rate": 4.999285761056355e-05, + "loss": 3.3121, + "step": 3464 + }, + { + "epoch": 0.10688300321188823, + "grad_norm": 0.18812477588653564, + "learning_rate": 4.999279312288837e-05, + "loss": 3.3043, + "step": 3465 + }, + { + "epoch": 0.10691384967746165, + "grad_norm": 0.17862798273563385, + "learning_rate": 4.9992728345436777e-05, + "loss": 3.2968, + "step": 3466 + }, + { + "epoch": 0.10694469614303506, + "grad_norm": 0.18015655875205994, + "learning_rate": 4.9992663278209515e-05, + "loss": 3.2927, + "step": 3467 + }, + { + "epoch": 0.10697554260860848, + "grad_norm": 0.18778593838214874, + "learning_rate": 4.999259792120735e-05, + "loss": 3.2985, + "step": 3468 + }, + { + "epoch": 0.10700638907418189, + "grad_norm": 0.18031704425811768, + "learning_rate": 4.9992532274431025e-05, + "loss": 3.2871, + "step": 3469 + }, + { + "epoch": 0.1070372355397553, + "grad_norm": 0.18704503774642944, + "learning_rate": 4.9992466337881314e-05, + "loss": 3.2712, + "step": 3470 + }, + { + "epoch": 0.10706808200532873, + "grad_norm": 0.17448969185352325, + "learning_rate": 4.999240011155897e-05, + "loss": 3.298, + "step": 3471 + }, + { + "epoch": 0.10709892847090215, + "grad_norm": 0.18642328679561615, + "learning_rate": 4.999233359546476e-05, + "loss": 3.3038, + "step": 3472 + }, + { + "epoch": 0.10712977493647556, + "grad_norm": 0.19519497454166412, + "learning_rate": 4.9992266789599464e-05, + "loss": 3.3611, + "step": 3473 + }, + { + "epoch": 0.10716062140204898, + "grad_norm": 0.1869397610425949, + "learning_rate": 4.999219969396385e-05, + "loss": 3.3223, + "step": 3474 + }, + { + "epoch": 0.10719146786762239, + "grad_norm": 0.17541193962097168, + "learning_rate": 4.9992132308558707e-05, + "loss": 3.3102, + "step": 3475 + }, + { + "epoch": 0.1072223143331958, + "grad_norm": 0.1799476444721222, + "learning_rate": 4.9992064633384804e-05, + "loss": 3.3048, + "step": 3476 + }, + { + "epoch": 0.10725316079876923, + "grad_norm": 0.18567822873592377, + "learning_rate": 4.999199666844293e-05, + "loss": 3.2982, + "step": 3477 + }, + { + "epoch": 0.10728400726434265, + "grad_norm": 0.18453890085220337, + "learning_rate": 4.999192841373387e-05, + "loss": 3.2994, + "step": 3478 + }, + { + "epoch": 0.10731485372991606, + "grad_norm": 0.18541987240314484, + "learning_rate": 4.999185986925843e-05, + "loss": 3.3269, + "step": 3479 + }, + { + "epoch": 0.10734570019548947, + "grad_norm": 0.17370837926864624, + "learning_rate": 4.999179103501738e-05, + "loss": 3.3341, + "step": 3480 + }, + { + "epoch": 0.10737654666106289, + "grad_norm": 0.19038929045200348, + "learning_rate": 4.9991721911011536e-05, + "loss": 3.309, + "step": 3481 + }, + { + "epoch": 0.1074073931266363, + "grad_norm": 0.17455418407917023, + "learning_rate": 4.999165249724169e-05, + "loss": 3.3252, + "step": 3482 + }, + { + "epoch": 0.10743823959220973, + "grad_norm": 0.17780065536499023, + "learning_rate": 4.999158279370866e-05, + "loss": 3.3532, + "step": 3483 + }, + { + "epoch": 0.10746908605778314, + "grad_norm": 0.18349096179008484, + "learning_rate": 4.999151280041324e-05, + "loss": 3.3018, + "step": 3484 + }, + { + "epoch": 0.10749993252335656, + "grad_norm": 0.18458937108516693, + "learning_rate": 4.999144251735625e-05, + "loss": 3.3132, + "step": 3485 + }, + { + "epoch": 0.10753077898892997, + "grad_norm": 0.1786295771598816, + "learning_rate": 4.99913719445385e-05, + "loss": 3.3268, + "step": 3486 + }, + { + "epoch": 0.10756162545450339, + "grad_norm": 0.18439500033855438, + "learning_rate": 4.999130108196082e-05, + "loss": 3.2738, + "step": 3487 + }, + { + "epoch": 0.1075924719200768, + "grad_norm": 0.18632535636425018, + "learning_rate": 4.999122992962401e-05, + "loss": 3.2828, + "step": 3488 + }, + { + "epoch": 0.10762331838565023, + "grad_norm": 0.1704372763633728, + "learning_rate": 4.999115848752891e-05, + "loss": 3.3107, + "step": 3489 + }, + { + "epoch": 0.10765416485122364, + "grad_norm": 0.18230462074279785, + "learning_rate": 4.999108675567635e-05, + "loss": 3.2968, + "step": 3490 + }, + { + "epoch": 0.10768501131679706, + "grad_norm": 0.22375071048736572, + "learning_rate": 4.999101473406715e-05, + "loss": 3.3129, + "step": 3491 + }, + { + "epoch": 0.10771585778237047, + "grad_norm": 0.1732882708311081, + "learning_rate": 4.999094242270216e-05, + "loss": 3.3043, + "step": 3492 + }, + { + "epoch": 0.10774670424794389, + "grad_norm": 0.176077201962471, + "learning_rate": 4.99908698215822e-05, + "loss": 3.312, + "step": 3493 + }, + { + "epoch": 0.1077775507135173, + "grad_norm": 0.21131952106952667, + "learning_rate": 4.9990796930708125e-05, + "loss": 3.3183, + "step": 3494 + }, + { + "epoch": 0.10780839717909073, + "grad_norm": 0.17301832139492035, + "learning_rate": 4.999072375008078e-05, + "loss": 3.2768, + "step": 3495 + }, + { + "epoch": 0.10783924364466414, + "grad_norm": 0.1789105087518692, + "learning_rate": 4.999065027970101e-05, + "loss": 3.2978, + "step": 3496 + }, + { + "epoch": 0.10787009011023756, + "grad_norm": 0.17487169802188873, + "learning_rate": 4.9990576519569674e-05, + "loss": 3.2954, + "step": 3497 + }, + { + "epoch": 0.10790093657581097, + "grad_norm": 0.17780061066150665, + "learning_rate": 4.999050246968762e-05, + "loss": 3.3017, + "step": 3498 + }, + { + "epoch": 0.10793178304138439, + "grad_norm": 0.17495974898338318, + "learning_rate": 4.99904281300557e-05, + "loss": 3.2775, + "step": 3499 + }, + { + "epoch": 0.1079626295069578, + "grad_norm": 0.1815287470817566, + "learning_rate": 4.999035350067478e-05, + "loss": 3.306, + "step": 3500 + }, + { + "epoch": 0.10799347597253123, + "grad_norm": 0.17953164875507355, + "learning_rate": 4.999027858154574e-05, + "loss": 3.2784, + "step": 3501 + }, + { + "epoch": 0.10802432243810464, + "grad_norm": 0.1700781285762787, + "learning_rate": 4.999020337266943e-05, + "loss": 3.2875, + "step": 3502 + }, + { + "epoch": 0.10805516890367806, + "grad_norm": 0.17545944452285767, + "learning_rate": 4.9990127874046734e-05, + "loss": 3.3119, + "step": 3503 + }, + { + "epoch": 0.10808601536925147, + "grad_norm": 0.16844959557056427, + "learning_rate": 4.999005208567852e-05, + "loss": 3.3083, + "step": 3504 + }, + { + "epoch": 0.10811686183482488, + "grad_norm": 0.17031016945838928, + "learning_rate": 4.9989976007565666e-05, + "loss": 3.2698, + "step": 3505 + }, + { + "epoch": 0.1081477083003983, + "grad_norm": 0.18277114629745483, + "learning_rate": 4.998989963970906e-05, + "loss": 3.2729, + "step": 3506 + }, + { + "epoch": 0.10817855476597173, + "grad_norm": 0.16861160099506378, + "learning_rate": 4.998982298210958e-05, + "loss": 3.2943, + "step": 3507 + }, + { + "epoch": 0.10820940123154514, + "grad_norm": 0.17036497592926025, + "learning_rate": 4.9989746034768134e-05, + "loss": 3.2859, + "step": 3508 + }, + { + "epoch": 0.10824024769711856, + "grad_norm": 0.1726498007774353, + "learning_rate": 4.9989668797685586e-05, + "loss": 3.2806, + "step": 3509 + }, + { + "epoch": 0.10827109416269197, + "grad_norm": 0.8556423783302307, + "learning_rate": 4.998959127086285e-05, + "loss": 3.3104, + "step": 3510 + }, + { + "epoch": 0.10830194062826538, + "grad_norm": 0.17346590757369995, + "learning_rate": 4.9989513454300824e-05, + "loss": 3.253, + "step": 3511 + }, + { + "epoch": 0.1083327870938388, + "grad_norm": 0.19476467370986938, + "learning_rate": 4.9989435348000405e-05, + "loss": 3.3333, + "step": 3512 + }, + { + "epoch": 0.10836363355941223, + "grad_norm": 0.17359820008277893, + "learning_rate": 4.9989356951962493e-05, + "loss": 3.3147, + "step": 3513 + }, + { + "epoch": 0.10839448002498564, + "grad_norm": 0.19462791085243225, + "learning_rate": 4.998927826618801e-05, + "loss": 3.3502, + "step": 3514 + }, + { + "epoch": 0.10842532649055905, + "grad_norm": 0.18725159764289856, + "learning_rate": 4.9989199290677865e-05, + "loss": 3.3565, + "step": 3515 + }, + { + "epoch": 0.10845617295613247, + "grad_norm": 0.19450603425502777, + "learning_rate": 4.9989120025432964e-05, + "loss": 3.3376, + "step": 3516 + }, + { + "epoch": 0.10848701942170588, + "grad_norm": 0.18601059913635254, + "learning_rate": 4.998904047045424e-05, + "loss": 3.2961, + "step": 3517 + }, + { + "epoch": 0.1085178658872793, + "grad_norm": 0.43926966190338135, + "learning_rate": 4.99889606257426e-05, + "loss": 3.2832, + "step": 3518 + }, + { + "epoch": 0.10854871235285272, + "grad_norm": 1.03969144821167, + "learning_rate": 4.998888049129898e-05, + "loss": 3.325, + "step": 3519 + }, + { + "epoch": 0.10857955881842614, + "grad_norm": 0.20160779356956482, + "learning_rate": 4.998880006712432e-05, + "loss": 3.2727, + "step": 3520 + }, + { + "epoch": 0.10861040528399955, + "grad_norm": 0.20334096252918243, + "learning_rate": 4.9988719353219525e-05, + "loss": 3.3349, + "step": 3521 + }, + { + "epoch": 0.10864125174957297, + "grad_norm": 0.21944884955883026, + "learning_rate": 4.998863834958555e-05, + "loss": 3.2893, + "step": 3522 + }, + { + "epoch": 0.10867209821514638, + "grad_norm": 0.206107959151268, + "learning_rate": 4.998855705622333e-05, + "loss": 3.2897, + "step": 3523 + }, + { + "epoch": 0.1087029446807198, + "grad_norm": 0.2015969157218933, + "learning_rate": 4.998847547313381e-05, + "loss": 3.2911, + "step": 3524 + }, + { + "epoch": 0.10873379114629322, + "grad_norm": 0.20365990698337555, + "learning_rate": 4.998839360031793e-05, + "loss": 3.2927, + "step": 3525 + }, + { + "epoch": 0.10876463761186664, + "grad_norm": 0.2282821387052536, + "learning_rate": 4.998831143777664e-05, + "loss": 3.3134, + "step": 3526 + }, + { + "epoch": 0.10879548407744005, + "grad_norm": 0.2136220633983612, + "learning_rate": 4.99882289855109e-05, + "loss": 3.3115, + "step": 3527 + }, + { + "epoch": 0.10882633054301347, + "grad_norm": 0.20693953335285187, + "learning_rate": 4.998814624352166e-05, + "loss": 3.3248, + "step": 3528 + }, + { + "epoch": 0.10885717700858688, + "grad_norm": 0.21230480074882507, + "learning_rate": 4.9988063211809874e-05, + "loss": 3.3502, + "step": 3529 + }, + { + "epoch": 0.1088880234741603, + "grad_norm": 0.19183026254177094, + "learning_rate": 4.9987979890376523e-05, + "loss": 3.2799, + "step": 3530 + }, + { + "epoch": 0.10891886993973372, + "grad_norm": 0.20151972770690918, + "learning_rate": 4.998789627922256e-05, + "loss": 3.3313, + "step": 3531 + }, + { + "epoch": 0.10894971640530714, + "grad_norm": 0.22326618432998657, + "learning_rate": 4.998781237834894e-05, + "loss": 3.3022, + "step": 3532 + }, + { + "epoch": 0.10898056287088055, + "grad_norm": 0.20115555822849274, + "learning_rate": 4.998772818775666e-05, + "loss": 3.2802, + "step": 3533 + }, + { + "epoch": 0.10901140933645397, + "grad_norm": 0.19886514544487, + "learning_rate": 4.9987643707446695e-05, + "loss": 3.3336, + "step": 3534 + }, + { + "epoch": 0.10904225580202738, + "grad_norm": 0.18681584298610687, + "learning_rate": 4.998755893742002e-05, + "loss": 3.3123, + "step": 3535 + }, + { + "epoch": 0.1090731022676008, + "grad_norm": 0.35414254665374756, + "learning_rate": 4.9987473877677605e-05, + "loss": 3.3366, + "step": 3536 + }, + { + "epoch": 0.10910394873317422, + "grad_norm": 0.18487648665905, + "learning_rate": 4.998738852822045e-05, + "loss": 3.3215, + "step": 3537 + }, + { + "epoch": 0.10913479519874764, + "grad_norm": 0.18295404314994812, + "learning_rate": 4.998730288904953e-05, + "loss": 3.345, + "step": 3538 + }, + { + "epoch": 0.10916564166432105, + "grad_norm": 0.17347915470600128, + "learning_rate": 4.9987216960165865e-05, + "loss": 3.2931, + "step": 3539 + }, + { + "epoch": 0.10919648812989446, + "grad_norm": 0.17329807579517365, + "learning_rate": 4.998713074157042e-05, + "loss": 3.3352, + "step": 3540 + }, + { + "epoch": 0.10922733459546788, + "grad_norm": 0.19399598240852356, + "learning_rate": 4.998704423326422e-05, + "loss": 3.3544, + "step": 3541 + }, + { + "epoch": 0.10925818106104129, + "grad_norm": 0.19531983137130737, + "learning_rate": 4.998695743524825e-05, + "loss": 3.2783, + "step": 3542 + }, + { + "epoch": 0.10928902752661472, + "grad_norm": 0.18948781490325928, + "learning_rate": 4.998687034752353e-05, + "loss": 3.3217, + "step": 3543 + }, + { + "epoch": 0.10931987399218814, + "grad_norm": 0.19211016595363617, + "learning_rate": 4.9986782970091064e-05, + "loss": 3.273, + "step": 3544 + }, + { + "epoch": 0.10935072045776155, + "grad_norm": 0.2101781815290451, + "learning_rate": 4.998669530295186e-05, + "loss": 3.3117, + "step": 3545 + }, + { + "epoch": 0.10938156692333496, + "grad_norm": 0.19445650279521942, + "learning_rate": 4.998660734610694e-05, + "loss": 3.3241, + "step": 3546 + }, + { + "epoch": 0.10941241338890838, + "grad_norm": 0.19461935758590698, + "learning_rate": 4.998651909955733e-05, + "loss": 3.2941, + "step": 3547 + }, + { + "epoch": 0.10944325985448179, + "grad_norm": 0.1888042539358139, + "learning_rate": 4.9986430563304044e-05, + "loss": 3.2936, + "step": 3548 + }, + { + "epoch": 0.10947410632005522, + "grad_norm": 0.1757163405418396, + "learning_rate": 4.99863417373481e-05, + "loss": 3.3227, + "step": 3549 + }, + { + "epoch": 0.10950495278562863, + "grad_norm": 0.21121983230113983, + "learning_rate": 4.9986252621690546e-05, + "loss": 3.3062, + "step": 3550 + }, + { + "epoch": 0.10953579925120205, + "grad_norm": 0.19525475800037384, + "learning_rate": 4.998616321633241e-05, + "loss": 3.3166, + "step": 3551 + }, + { + "epoch": 0.10956664571677546, + "grad_norm": 0.1951005905866623, + "learning_rate": 4.9986073521274727e-05, + "loss": 3.2826, + "step": 3552 + }, + { + "epoch": 0.10959749218234888, + "grad_norm": 4.015944480895996, + "learning_rate": 4.9985983536518535e-05, + "loss": 3.301, + "step": 3553 + }, + { + "epoch": 0.10962833864792229, + "grad_norm": 0.2913748323917389, + "learning_rate": 4.998589326206488e-05, + "loss": 3.3084, + "step": 3554 + }, + { + "epoch": 0.10965918511349572, + "grad_norm": 0.3239012062549591, + "learning_rate": 4.99858026979148e-05, + "loss": 3.3058, + "step": 3555 + }, + { + "epoch": 0.10969003157906913, + "grad_norm": 0.22042788565158844, + "learning_rate": 4.9985711844069364e-05, + "loss": 3.3191, + "step": 3556 + }, + { + "epoch": 0.10972087804464255, + "grad_norm": 0.2505595088005066, + "learning_rate": 4.99856207005296e-05, + "loss": 3.3032, + "step": 3557 + }, + { + "epoch": 0.10975172451021596, + "grad_norm": 0.23933666944503784, + "learning_rate": 4.9985529267296585e-05, + "loss": 3.3511, + "step": 3558 + }, + { + "epoch": 0.10978257097578938, + "grad_norm": 0.23436015844345093, + "learning_rate": 4.998543754437138e-05, + "loss": 3.2908, + "step": 3559 + }, + { + "epoch": 0.10981341744136279, + "grad_norm": 0.20283307135105133, + "learning_rate": 4.998534553175503e-05, + "loss": 3.3113, + "step": 3560 + }, + { + "epoch": 0.10984426390693622, + "grad_norm": 0.2262822538614273, + "learning_rate": 4.998525322944862e-05, + "loss": 3.3213, + "step": 3561 + }, + { + "epoch": 0.10987511037250963, + "grad_norm": 0.19032466411590576, + "learning_rate": 4.998516063745321e-05, + "loss": 3.3112, + "step": 3562 + }, + { + "epoch": 0.10990595683808305, + "grad_norm": 0.210026815533638, + "learning_rate": 4.998506775576988e-05, + "loss": 3.2672, + "step": 3563 + }, + { + "epoch": 0.10993680330365646, + "grad_norm": 0.1894656866788864, + "learning_rate": 4.99849745843997e-05, + "loss": 3.3198, + "step": 3564 + }, + { + "epoch": 0.10996764976922987, + "grad_norm": 0.2077430933713913, + "learning_rate": 4.9984881123343756e-05, + "loss": 3.3224, + "step": 3565 + }, + { + "epoch": 0.10999849623480329, + "grad_norm": 0.20057053864002228, + "learning_rate": 4.9984787372603126e-05, + "loss": 3.3337, + "step": 3566 + }, + { + "epoch": 0.11002934270037672, + "grad_norm": 0.19933946430683136, + "learning_rate": 4.998469333217891e-05, + "loss": 3.2927, + "step": 3567 + }, + { + "epoch": 0.11006018916595013, + "grad_norm": 0.20539860427379608, + "learning_rate": 4.998459900207218e-05, + "loss": 3.2755, + "step": 3568 + }, + { + "epoch": 0.11009103563152355, + "grad_norm": 0.21225950121879578, + "learning_rate": 4.998450438228404e-05, + "loss": 3.3088, + "step": 3569 + }, + { + "epoch": 0.11012188209709696, + "grad_norm": 0.23594075441360474, + "learning_rate": 4.998440947281559e-05, + "loss": 3.3177, + "step": 3570 + }, + { + "epoch": 0.11015272856267037, + "grad_norm": 0.2017408311367035, + "learning_rate": 4.998431427366792e-05, + "loss": 3.2986, + "step": 3571 + }, + { + "epoch": 0.11018357502824379, + "grad_norm": 0.22241966426372528, + "learning_rate": 4.998421878484214e-05, + "loss": 3.3308, + "step": 3572 + }, + { + "epoch": 0.11021442149381722, + "grad_norm": 0.1825525015592575, + "learning_rate": 4.998412300633936e-05, + "loss": 3.2774, + "step": 3573 + }, + { + "epoch": 0.11024526795939063, + "grad_norm": 0.2261403501033783, + "learning_rate": 4.998402693816069e-05, + "loss": 3.3181, + "step": 3574 + }, + { + "epoch": 0.11027611442496404, + "grad_norm": 0.19871625304222107, + "learning_rate": 4.9983930580307235e-05, + "loss": 3.2881, + "step": 3575 + }, + { + "epoch": 0.11030696089053746, + "grad_norm": 0.20646138489246368, + "learning_rate": 4.998383393278012e-05, + "loss": 3.2646, + "step": 3576 + }, + { + "epoch": 0.11033780735611087, + "grad_norm": 0.20095112919807434, + "learning_rate": 4.998373699558047e-05, + "loss": 3.3077, + "step": 3577 + }, + { + "epoch": 0.1103686538216843, + "grad_norm": 0.18502047657966614, + "learning_rate": 4.99836397687094e-05, + "loss": 3.3155, + "step": 3578 + }, + { + "epoch": 0.11039950028725772, + "grad_norm": 0.19816845655441284, + "learning_rate": 4.998354225216804e-05, + "loss": 3.3039, + "step": 3579 + }, + { + "epoch": 0.11043034675283113, + "grad_norm": 0.24686378240585327, + "learning_rate": 4.9983444445957514e-05, + "loss": 3.3456, + "step": 3580 + }, + { + "epoch": 0.11046119321840454, + "grad_norm": 0.19157792627811432, + "learning_rate": 4.9983346350078973e-05, + "loss": 3.293, + "step": 3581 + }, + { + "epoch": 0.11049203968397796, + "grad_norm": 0.17823117971420288, + "learning_rate": 4.998324796453354e-05, + "loss": 3.2906, + "step": 3582 + }, + { + "epoch": 0.11052288614955137, + "grad_norm": 0.186862051486969, + "learning_rate": 4.9983149289322365e-05, + "loss": 3.3181, + "step": 3583 + }, + { + "epoch": 0.1105537326151248, + "grad_norm": 0.18160760402679443, + "learning_rate": 4.998305032444658e-05, + "loss": 3.3233, + "step": 3584 + }, + { + "epoch": 0.11058457908069821, + "grad_norm": 0.18276464939117432, + "learning_rate": 4.998295106990734e-05, + "loss": 3.3094, + "step": 3585 + }, + { + "epoch": 0.11061542554627163, + "grad_norm": 0.1791532039642334, + "learning_rate": 4.99828515257058e-05, + "loss": 3.3246, + "step": 3586 + }, + { + "epoch": 0.11064627201184504, + "grad_norm": 0.17645162343978882, + "learning_rate": 4.9982751691843106e-05, + "loss": 3.2984, + "step": 3587 + }, + { + "epoch": 0.11067711847741846, + "grad_norm": 0.18302425742149353, + "learning_rate": 4.998265156832041e-05, + "loss": 3.3525, + "step": 3588 + }, + { + "epoch": 0.11070796494299187, + "grad_norm": 0.18899068236351013, + "learning_rate": 4.99825511551389e-05, + "loss": 3.3116, + "step": 3589 + }, + { + "epoch": 0.1107388114085653, + "grad_norm": 0.18391938507556915, + "learning_rate": 4.998245045229971e-05, + "loss": 3.2395, + "step": 3590 + }, + { + "epoch": 0.11076965787413871, + "grad_norm": 0.18890558183193207, + "learning_rate": 4.998234945980403e-05, + "loss": 3.3052, + "step": 3591 + }, + { + "epoch": 0.11080050433971213, + "grad_norm": 0.18602125346660614, + "learning_rate": 4.998224817765301e-05, + "loss": 3.309, + "step": 3592 + }, + { + "epoch": 0.11083135080528554, + "grad_norm": 0.16960081458091736, + "learning_rate": 4.998214660584785e-05, + "loss": 3.2851, + "step": 3593 + }, + { + "epoch": 0.11086219727085896, + "grad_norm": 0.17643417418003082, + "learning_rate": 4.9982044744389695e-05, + "loss": 3.2943, + "step": 3594 + }, + { + "epoch": 0.11089304373643237, + "grad_norm": 0.1677800714969635, + "learning_rate": 4.998194259327975e-05, + "loss": 3.2682, + "step": 3595 + }, + { + "epoch": 0.1109238902020058, + "grad_norm": 0.18686912953853607, + "learning_rate": 4.998184015251919e-05, + "loss": 3.319, + "step": 3596 + }, + { + "epoch": 0.11095473666757921, + "grad_norm": 0.16958196461200714, + "learning_rate": 4.998173742210921e-05, + "loss": 3.2723, + "step": 3597 + }, + { + "epoch": 0.11098558313315263, + "grad_norm": 0.18530364334583282, + "learning_rate": 4.9981634402050994e-05, + "loss": 3.3398, + "step": 3598 + }, + { + "epoch": 0.11101642959872604, + "grad_norm": 0.1779954731464386, + "learning_rate": 4.998153109234575e-05, + "loss": 3.2988, + "step": 3599 + }, + { + "epoch": 0.11104727606429945, + "grad_norm": 0.19276855885982513, + "learning_rate": 4.998142749299465e-05, + "loss": 3.3196, + "step": 3600 + }, + { + "epoch": 0.11107812252987287, + "grad_norm": 0.18582689762115479, + "learning_rate": 4.998132360399892e-05, + "loss": 3.2943, + "step": 3601 + }, + { + "epoch": 0.1111089689954463, + "grad_norm": 0.1763857901096344, + "learning_rate": 4.9981219425359747e-05, + "loss": 3.3262, + "step": 3602 + }, + { + "epoch": 0.11113981546101971, + "grad_norm": 0.19776564836502075, + "learning_rate": 4.9981114957078356e-05, + "loss": 3.3143, + "step": 3603 + }, + { + "epoch": 0.11117066192659313, + "grad_norm": 0.1747395545244217, + "learning_rate": 4.998101019915594e-05, + "loss": 3.2686, + "step": 3604 + }, + { + "epoch": 0.11120150839216654, + "grad_norm": 0.19240416586399078, + "learning_rate": 4.998090515159373e-05, + "loss": 3.306, + "step": 3605 + }, + { + "epoch": 0.11123235485773995, + "grad_norm": 0.17541013658046722, + "learning_rate": 4.998079981439294e-05, + "loss": 3.3024, + "step": 3606 + }, + { + "epoch": 0.11126320132331337, + "grad_norm": 0.19793447852134705, + "learning_rate": 4.9980694187554776e-05, + "loss": 3.2907, + "step": 3607 + }, + { + "epoch": 0.1112940477888868, + "grad_norm": 0.17914198338985443, + "learning_rate": 4.998058827108049e-05, + "loss": 3.2865, + "step": 3608 + }, + { + "epoch": 0.11132489425446021, + "grad_norm": 0.17495201528072357, + "learning_rate": 4.998048206497128e-05, + "loss": 3.2992, + "step": 3609 + }, + { + "epoch": 0.11135574072003362, + "grad_norm": 0.18027393519878387, + "learning_rate": 4.998037556922839e-05, + "loss": 3.3613, + "step": 3610 + }, + { + "epoch": 0.11138658718560704, + "grad_norm": 0.17533859610557556, + "learning_rate": 4.998026878385307e-05, + "loss": 3.3067, + "step": 3611 + }, + { + "epoch": 0.11141743365118045, + "grad_norm": 0.1730884164571762, + "learning_rate": 4.9980161708846535e-05, + "loss": 3.3526, + "step": 3612 + }, + { + "epoch": 0.11144828011675387, + "grad_norm": 0.1870054304599762, + "learning_rate": 4.998005434421004e-05, + "loss": 3.3137, + "step": 3613 + }, + { + "epoch": 0.1114791265823273, + "grad_norm": 0.17241360247135162, + "learning_rate": 4.9979946689944824e-05, + "loss": 3.3019, + "step": 3614 + }, + { + "epoch": 0.11150997304790071, + "grad_norm": 0.18816670775413513, + "learning_rate": 4.997983874605214e-05, + "loss": 3.3077, + "step": 3615 + }, + { + "epoch": 0.11154081951347412, + "grad_norm": 0.17806260287761688, + "learning_rate": 4.997973051253324e-05, + "loss": 3.3158, + "step": 3616 + }, + { + "epoch": 0.11157166597904754, + "grad_norm": 0.1960720419883728, + "learning_rate": 4.9979621989389366e-05, + "loss": 3.2786, + "step": 3617 + }, + { + "epoch": 0.11160251244462095, + "grad_norm": 0.16863718628883362, + "learning_rate": 4.997951317662179e-05, + "loss": 3.3379, + "step": 3618 + }, + { + "epoch": 0.11163335891019437, + "grad_norm": 0.18077003955841064, + "learning_rate": 4.9979404074231774e-05, + "loss": 3.3092, + "step": 3619 + }, + { + "epoch": 0.1116642053757678, + "grad_norm": 0.18791459500789642, + "learning_rate": 4.997929468222058e-05, + "loss": 3.2558, + "step": 3620 + }, + { + "epoch": 0.11169505184134121, + "grad_norm": 0.16868191957473755, + "learning_rate": 4.997918500058947e-05, + "loss": 3.2683, + "step": 3621 + }, + { + "epoch": 0.11172589830691462, + "grad_norm": 0.1845010668039322, + "learning_rate": 4.997907502933972e-05, + "loss": 3.3082, + "step": 3622 + }, + { + "epoch": 0.11175674477248804, + "grad_norm": 0.17824125289916992, + "learning_rate": 4.9978964768472605e-05, + "loss": 3.3327, + "step": 3623 + }, + { + "epoch": 0.11178759123806145, + "grad_norm": 0.17513148486614227, + "learning_rate": 4.9978854217989415e-05, + "loss": 3.2901, + "step": 3624 + }, + { + "epoch": 0.11181843770363487, + "grad_norm": 0.16942158341407776, + "learning_rate": 4.9978743377891404e-05, + "loss": 3.2642, + "step": 3625 + }, + { + "epoch": 0.1118492841692083, + "grad_norm": 0.17750954627990723, + "learning_rate": 4.9978632248179885e-05, + "loss": 3.323, + "step": 3626 + }, + { + "epoch": 0.11188013063478171, + "grad_norm": 0.17834600806236267, + "learning_rate": 4.997852082885613e-05, + "loss": 3.2928, + "step": 3627 + }, + { + "epoch": 0.11191097710035512, + "grad_norm": 0.17390327155590057, + "learning_rate": 4.997840911992144e-05, + "loss": 3.2803, + "step": 3628 + }, + { + "epoch": 0.11194182356592854, + "grad_norm": 0.1729860156774521, + "learning_rate": 4.997829712137711e-05, + "loss": 3.2951, + "step": 3629 + }, + { + "epoch": 0.11197267003150195, + "grad_norm": 0.1753891259431839, + "learning_rate": 4.997818483322443e-05, + "loss": 3.3053, + "step": 3630 + }, + { + "epoch": 0.11200351649707536, + "grad_norm": 0.17241115868091583, + "learning_rate": 4.997807225546471e-05, + "loss": 3.2875, + "step": 3631 + }, + { + "epoch": 0.11203436296264879, + "grad_norm": 0.1839093416929245, + "learning_rate": 4.997795938809925e-05, + "loss": 3.3626, + "step": 3632 + }, + { + "epoch": 0.1120652094282222, + "grad_norm": 0.18194931745529175, + "learning_rate": 4.997784623112936e-05, + "loss": 3.3098, + "step": 3633 + }, + { + "epoch": 0.11209605589379562, + "grad_norm": 0.16710638999938965, + "learning_rate": 4.9977732784556355e-05, + "loss": 3.3155, + "step": 3634 + }, + { + "epoch": 0.11212690235936904, + "grad_norm": 0.19656315445899963, + "learning_rate": 4.9977619048381545e-05, + "loss": 3.3044, + "step": 3635 + }, + { + "epoch": 0.11215774882494245, + "grad_norm": 0.175246924161911, + "learning_rate": 4.997750502260625e-05, + "loss": 3.2668, + "step": 3636 + }, + { + "epoch": 0.11218859529051586, + "grad_norm": 0.18929347395896912, + "learning_rate": 4.9977390707231806e-05, + "loss": 3.3208, + "step": 3637 + }, + { + "epoch": 0.11221944175608929, + "grad_norm": 0.18033218383789062, + "learning_rate": 4.997727610225952e-05, + "loss": 3.3016, + "step": 3638 + }, + { + "epoch": 0.1122502882216627, + "grad_norm": 0.17017494142055511, + "learning_rate": 4.9977161207690723e-05, + "loss": 3.2913, + "step": 3639 + }, + { + "epoch": 0.11228113468723612, + "grad_norm": 0.17036853730678558, + "learning_rate": 4.997704602352675e-05, + "loss": 3.2657, + "step": 3640 + }, + { + "epoch": 0.11231198115280953, + "grad_norm": 0.171138197183609, + "learning_rate": 4.997693054976894e-05, + "loss": 3.2875, + "step": 3641 + }, + { + "epoch": 0.11234282761838295, + "grad_norm": 0.17292875051498413, + "learning_rate": 4.9976814786418637e-05, + "loss": 3.2971, + "step": 3642 + }, + { + "epoch": 0.11237367408395636, + "grad_norm": 0.17295347154140472, + "learning_rate": 4.997669873347717e-05, + "loss": 3.3126, + "step": 3643 + }, + { + "epoch": 0.11240452054952979, + "grad_norm": 0.18535088002681732, + "learning_rate": 4.997658239094589e-05, + "loss": 3.3046, + "step": 3644 + }, + { + "epoch": 0.1124353670151032, + "grad_norm": 0.18005934357643127, + "learning_rate": 4.9976465758826144e-05, + "loss": 3.3087, + "step": 3645 + }, + { + "epoch": 0.11246621348067662, + "grad_norm": 0.18560752272605896, + "learning_rate": 4.997634883711929e-05, + "loss": 3.3081, + "step": 3646 + }, + { + "epoch": 0.11249705994625003, + "grad_norm": 0.17409463226795197, + "learning_rate": 4.9976231625826684e-05, + "loss": 3.2794, + "step": 3647 + }, + { + "epoch": 0.11252790641182345, + "grad_norm": 0.2068367600440979, + "learning_rate": 4.997611412494968e-05, + "loss": 3.2918, + "step": 3648 + }, + { + "epoch": 0.11255875287739686, + "grad_norm": 0.17318220436573029, + "learning_rate": 4.997599633448964e-05, + "loss": 3.3119, + "step": 3649 + }, + { + "epoch": 0.11258959934297029, + "grad_norm": 0.1804666966199875, + "learning_rate": 4.997587825444793e-05, + "loss": 3.2955, + "step": 3650 + }, + { + "epoch": 0.1126204458085437, + "grad_norm": 0.17320570349693298, + "learning_rate": 4.9975759884825915e-05, + "loss": 3.295, + "step": 3651 + }, + { + "epoch": 0.11265129227411712, + "grad_norm": 0.1759348064661026, + "learning_rate": 4.997564122562498e-05, + "loss": 3.3116, + "step": 3652 + }, + { + "epoch": 0.11268213873969053, + "grad_norm": 0.17414630949497223, + "learning_rate": 4.9975522276846494e-05, + "loss": 3.3211, + "step": 3653 + }, + { + "epoch": 0.11271298520526395, + "grad_norm": 0.17947585880756378, + "learning_rate": 4.997540303849183e-05, + "loss": 3.2928, + "step": 3654 + }, + { + "epoch": 0.11274383167083736, + "grad_norm": 0.17253351211547852, + "learning_rate": 4.9975283510562386e-05, + "loss": 3.2944, + "step": 3655 + }, + { + "epoch": 0.11277467813641079, + "grad_norm": 0.16803789138793945, + "learning_rate": 4.997516369305953e-05, + "loss": 3.3156, + "step": 3656 + }, + { + "epoch": 0.1128055246019842, + "grad_norm": 0.18313227593898773, + "learning_rate": 4.997504358598466e-05, + "loss": 3.3391, + "step": 3657 + }, + { + "epoch": 0.11283637106755762, + "grad_norm": 0.17216624319553375, + "learning_rate": 4.9974923189339177e-05, + "loss": 3.3195, + "step": 3658 + }, + { + "epoch": 0.11286721753313103, + "grad_norm": 0.17919155955314636, + "learning_rate": 4.997480250312445e-05, + "loss": 3.2745, + "step": 3659 + }, + { + "epoch": 0.11289806399870445, + "grad_norm": 0.18106186389923096, + "learning_rate": 4.997468152734191e-05, + "loss": 3.3, + "step": 3660 + }, + { + "epoch": 0.11292891046427786, + "grad_norm": 0.19383864104747772, + "learning_rate": 4.997456026199294e-05, + "loss": 3.3168, + "step": 3661 + }, + { + "epoch": 0.11295975692985129, + "grad_norm": 0.17318539321422577, + "learning_rate": 4.997443870707896e-05, + "loss": 3.2836, + "step": 3662 + }, + { + "epoch": 0.1129906033954247, + "grad_norm": 0.1881897896528244, + "learning_rate": 4.997431686260136e-05, + "loss": 3.3282, + "step": 3663 + }, + { + "epoch": 0.11302144986099812, + "grad_norm": 0.19086118042469025, + "learning_rate": 4.9974194728561576e-05, + "loss": 3.3242, + "step": 3664 + }, + { + "epoch": 0.11305229632657153, + "grad_norm": 0.16967318952083588, + "learning_rate": 4.9974072304960995e-05, + "loss": 3.2917, + "step": 3665 + }, + { + "epoch": 0.11308314279214494, + "grad_norm": 0.18842490017414093, + "learning_rate": 4.997394959180107e-05, + "loss": 3.2769, + "step": 3666 + }, + { + "epoch": 0.11311398925771836, + "grad_norm": 0.1741948127746582, + "learning_rate": 4.9973826589083195e-05, + "loss": 3.2989, + "step": 3667 + }, + { + "epoch": 0.11314483572329179, + "grad_norm": 0.1732313632965088, + "learning_rate": 4.9973703296808816e-05, + "loss": 3.2981, + "step": 3668 + }, + { + "epoch": 0.1131756821888652, + "grad_norm": 0.1902502179145813, + "learning_rate": 4.997357971497935e-05, + "loss": 3.3, + "step": 3669 + }, + { + "epoch": 0.11320652865443862, + "grad_norm": 0.19505035877227783, + "learning_rate": 4.997345584359624e-05, + "loss": 3.2929, + "step": 3670 + }, + { + "epoch": 0.11323737512001203, + "grad_norm": 0.1682652086019516, + "learning_rate": 4.997333168266091e-05, + "loss": 3.3188, + "step": 3671 + }, + { + "epoch": 0.11326822158558544, + "grad_norm": 0.20012547075748444, + "learning_rate": 4.997320723217481e-05, + "loss": 3.2874, + "step": 3672 + }, + { + "epoch": 0.11329906805115886, + "grad_norm": 0.17603623867034912, + "learning_rate": 4.997308249213938e-05, + "loss": 3.3158, + "step": 3673 + }, + { + "epoch": 0.11332991451673229, + "grad_norm": 0.17056377232074738, + "learning_rate": 4.9972957462556066e-05, + "loss": 3.2741, + "step": 3674 + }, + { + "epoch": 0.1133607609823057, + "grad_norm": 0.1739414930343628, + "learning_rate": 4.997283214342631e-05, + "loss": 3.2964, + "step": 3675 + }, + { + "epoch": 0.11339160744787911, + "grad_norm": 0.1843598186969757, + "learning_rate": 4.997270653475158e-05, + "loss": 3.2744, + "step": 3676 + }, + { + "epoch": 0.11342245391345253, + "grad_norm": 0.17534342408180237, + "learning_rate": 4.997258063653333e-05, + "loss": 3.2839, + "step": 3677 + }, + { + "epoch": 0.11345330037902594, + "grad_norm": 0.23478800058364868, + "learning_rate": 4.9972454448773e-05, + "loss": 3.3231, + "step": 3678 + }, + { + "epoch": 0.11348414684459936, + "grad_norm": 0.19646640121936798, + "learning_rate": 4.997232797147208e-05, + "loss": 3.341, + "step": 3679 + }, + { + "epoch": 0.11351499331017278, + "grad_norm": 0.182975634932518, + "learning_rate": 4.997220120463201e-05, + "loss": 3.3128, + "step": 3680 + }, + { + "epoch": 0.1135458397757462, + "grad_norm": 0.18843714892864227, + "learning_rate": 4.997207414825429e-05, + "loss": 3.2818, + "step": 3681 + }, + { + "epoch": 0.11357668624131961, + "grad_norm": 0.178920179605484, + "learning_rate": 4.9971946802340366e-05, + "loss": 3.3536, + "step": 3682 + }, + { + "epoch": 0.11360753270689303, + "grad_norm": 0.18375566601753235, + "learning_rate": 4.997181916689173e-05, + "loss": 3.2946, + "step": 3683 + }, + { + "epoch": 0.11363837917246644, + "grad_norm": 0.1940091848373413, + "learning_rate": 4.997169124190985e-05, + "loss": 3.3093, + "step": 3684 + }, + { + "epoch": 0.11366922563803986, + "grad_norm": 0.19213984906673431, + "learning_rate": 4.997156302739622e-05, + "loss": 3.2986, + "step": 3685 + }, + { + "epoch": 0.11370007210361328, + "grad_norm": 0.18389150500297546, + "learning_rate": 4.9971434523352324e-05, + "loss": 3.2965, + "step": 3686 + }, + { + "epoch": 0.1137309185691867, + "grad_norm": 0.1944877803325653, + "learning_rate": 4.997130572977965e-05, + "loss": 3.3087, + "step": 3687 + }, + { + "epoch": 0.11376176503476011, + "grad_norm": 0.17997895181179047, + "learning_rate": 4.99711766466797e-05, + "loss": 3.3238, + "step": 3688 + }, + { + "epoch": 0.11379261150033353, + "grad_norm": 0.2022680640220642, + "learning_rate": 4.997104727405395e-05, + "loss": 3.3065, + "step": 3689 + }, + { + "epoch": 0.11382345796590694, + "grad_norm": 0.17354895174503326, + "learning_rate": 4.997091761190392e-05, + "loss": 3.3008, + "step": 3690 + }, + { + "epoch": 0.11385430443148035, + "grad_norm": 0.20029045641422272, + "learning_rate": 4.99707876602311e-05, + "loss": 3.2458, + "step": 3691 + }, + { + "epoch": 0.11388515089705378, + "grad_norm": 0.18336305022239685, + "learning_rate": 4.9970657419037006e-05, + "loss": 3.286, + "step": 3692 + }, + { + "epoch": 0.1139159973626272, + "grad_norm": 0.19654493033885956, + "learning_rate": 4.997052688832315e-05, + "loss": 3.269, + "step": 3693 + }, + { + "epoch": 0.11394684382820061, + "grad_norm": 0.1779557466506958, + "learning_rate": 4.997039606809103e-05, + "loss": 3.3214, + "step": 3694 + }, + { + "epoch": 0.11397769029377403, + "grad_norm": 0.18274502456188202, + "learning_rate": 4.9970264958342184e-05, + "loss": 3.2603, + "step": 3695 + }, + { + "epoch": 0.11400853675934744, + "grad_norm": 0.18184857070446014, + "learning_rate": 4.9970133559078115e-05, + "loss": 3.2823, + "step": 3696 + }, + { + "epoch": 0.11403938322492085, + "grad_norm": 0.18235072493553162, + "learning_rate": 4.997000187030035e-05, + "loss": 3.3037, + "step": 3697 + }, + { + "epoch": 0.11407022969049428, + "grad_norm": 0.20492002367973328, + "learning_rate": 4.996986989201042e-05, + "loss": 3.2944, + "step": 3698 + }, + { + "epoch": 0.1141010761560677, + "grad_norm": 0.17608332633972168, + "learning_rate": 4.996973762420987e-05, + "loss": 3.3153, + "step": 3699 + }, + { + "epoch": 0.11413192262164111, + "grad_norm": 0.1867474764585495, + "learning_rate": 4.996960506690019e-05, + "loss": 3.2898, + "step": 3700 + }, + { + "epoch": 0.11416276908721452, + "grad_norm": 0.1672419011592865, + "learning_rate": 4.9969472220082965e-05, + "loss": 3.3124, + "step": 3701 + }, + { + "epoch": 0.11419361555278794, + "grad_norm": 0.18086916208267212, + "learning_rate": 4.99693390837597e-05, + "loss": 3.3187, + "step": 3702 + }, + { + "epoch": 0.11422446201836135, + "grad_norm": 0.1721484363079071, + "learning_rate": 4.9969205657931964e-05, + "loss": 3.3142, + "step": 3703 + }, + { + "epoch": 0.11425530848393478, + "grad_norm": 0.18074984848499298, + "learning_rate": 4.996907194260129e-05, + "loss": 3.2874, + "step": 3704 + }, + { + "epoch": 0.1142861549495082, + "grad_norm": 0.1767139732837677, + "learning_rate": 4.9968937937769234e-05, + "loss": 3.2586, + "step": 3705 + }, + { + "epoch": 0.11431700141508161, + "grad_norm": 0.18568934500217438, + "learning_rate": 4.996880364343735e-05, + "loss": 3.2596, + "step": 3706 + }, + { + "epoch": 0.11434784788065502, + "grad_norm": 0.1751554310321808, + "learning_rate": 4.996866905960718e-05, + "loss": 3.2775, + "step": 3707 + }, + { + "epoch": 0.11437869434622844, + "grad_norm": 0.18126653134822845, + "learning_rate": 4.99685341862803e-05, + "loss": 3.2967, + "step": 3708 + }, + { + "epoch": 0.11440954081180185, + "grad_norm": 0.1657930463552475, + "learning_rate": 4.996839902345828e-05, + "loss": 3.2897, + "step": 3709 + }, + { + "epoch": 0.11444038727737528, + "grad_norm": 0.1724555790424347, + "learning_rate": 4.996826357114267e-05, + "loss": 3.2611, + "step": 3710 + }, + { + "epoch": 0.1144712337429487, + "grad_norm": 0.1666932851076126, + "learning_rate": 4.996812782933505e-05, + "loss": 3.3003, + "step": 3711 + }, + { + "epoch": 0.11450208020852211, + "grad_norm": 0.17296102643013, + "learning_rate": 4.996799179803699e-05, + "loss": 3.2727, + "step": 3712 + }, + { + "epoch": 0.11453292667409552, + "grad_norm": 0.16497687995433807, + "learning_rate": 4.996785547725007e-05, + "loss": 3.3123, + "step": 3713 + }, + { + "epoch": 0.11456377313966894, + "grad_norm": 0.17010633647441864, + "learning_rate": 4.996771886697587e-05, + "loss": 3.2547, + "step": 3714 + }, + { + "epoch": 0.11459461960524235, + "grad_norm": 0.18171672523021698, + "learning_rate": 4.996758196721597e-05, + "loss": 3.3244, + "step": 3715 + }, + { + "epoch": 0.11462546607081578, + "grad_norm": 0.17803052067756653, + "learning_rate": 4.996744477797197e-05, + "loss": 3.3178, + "step": 3716 + }, + { + "epoch": 0.1146563125363892, + "grad_norm": 0.1783168911933899, + "learning_rate": 4.996730729924545e-05, + "loss": 3.2893, + "step": 3717 + }, + { + "epoch": 0.11468715900196261, + "grad_norm": 0.18175843358039856, + "learning_rate": 4.9967169531037996e-05, + "loss": 3.3156, + "step": 3718 + }, + { + "epoch": 0.11471800546753602, + "grad_norm": 0.17267893254756927, + "learning_rate": 4.996703147335122e-05, + "loss": 3.3215, + "step": 3719 + }, + { + "epoch": 0.11474885193310944, + "grad_norm": 0.187903493642807, + "learning_rate": 4.996689312618672e-05, + "loss": 3.2416, + "step": 3720 + }, + { + "epoch": 0.11477969839868285, + "grad_norm": 0.17062242329120636, + "learning_rate": 4.99667544895461e-05, + "loss": 3.3043, + "step": 3721 + }, + { + "epoch": 0.11481054486425628, + "grad_norm": 0.20040011405944824, + "learning_rate": 4.996661556343095e-05, + "loss": 3.318, + "step": 3722 + }, + { + "epoch": 0.11484139132982969, + "grad_norm": 0.18205013871192932, + "learning_rate": 4.9966476347842914e-05, + "loss": 3.2895, + "step": 3723 + }, + { + "epoch": 0.1148722377954031, + "grad_norm": 0.19235800206661224, + "learning_rate": 4.996633684278358e-05, + "loss": 3.2944, + "step": 3724 + }, + { + "epoch": 0.11490308426097652, + "grad_norm": 0.16647131741046906, + "learning_rate": 4.9966197048254574e-05, + "loss": 3.2865, + "step": 3725 + }, + { + "epoch": 0.11493393072654993, + "grad_norm": 0.1845467984676361, + "learning_rate": 4.996605696425751e-05, + "loss": 3.28, + "step": 3726 + }, + { + "epoch": 0.11496477719212335, + "grad_norm": 0.20665651559829712, + "learning_rate": 4.996591659079403e-05, + "loss": 3.2873, + "step": 3727 + }, + { + "epoch": 0.11499562365769678, + "grad_norm": 0.18732000887393951, + "learning_rate": 4.996577592786574e-05, + "loss": 3.2862, + "step": 3728 + }, + { + "epoch": 0.11502647012327019, + "grad_norm": 0.18081390857696533, + "learning_rate": 4.996563497547429e-05, + "loss": 3.2936, + "step": 3729 + }, + { + "epoch": 0.1150573165888436, + "grad_norm": 0.20782619714736938, + "learning_rate": 4.9965493733621296e-05, + "loss": 3.291, + "step": 3730 + }, + { + "epoch": 0.11508816305441702, + "grad_norm": 0.19195471704006195, + "learning_rate": 4.99653522023084e-05, + "loss": 3.2997, + "step": 3731 + }, + { + "epoch": 0.11511900951999043, + "grad_norm": 0.1939123570919037, + "learning_rate": 4.9965210381537265e-05, + "loss": 3.2779, + "step": 3732 + }, + { + "epoch": 0.11514985598556385, + "grad_norm": 0.27132448554039, + "learning_rate": 4.9965068271309504e-05, + "loss": 3.3067, + "step": 3733 + }, + { + "epoch": 0.11518070245113728, + "grad_norm": 0.17582838237285614, + "learning_rate": 4.996492587162679e-05, + "loss": 3.2836, + "step": 3734 + }, + { + "epoch": 0.11521154891671069, + "grad_norm": 0.19140923023223877, + "learning_rate": 4.996478318249075e-05, + "loss": 3.3051, + "step": 3735 + }, + { + "epoch": 0.1152423953822841, + "grad_norm": 0.17172355949878693, + "learning_rate": 4.996464020390306e-05, + "loss": 3.298, + "step": 3736 + }, + { + "epoch": 0.11527324184785752, + "grad_norm": 0.18931171298027039, + "learning_rate": 4.9964496935865366e-05, + "loss": 3.2949, + "step": 3737 + }, + { + "epoch": 0.11530408831343093, + "grad_norm": 0.1711784154176712, + "learning_rate": 4.996435337837933e-05, + "loss": 3.307, + "step": 3738 + }, + { + "epoch": 0.11533493477900435, + "grad_norm": 0.1972791850566864, + "learning_rate": 4.996420953144662e-05, + "loss": 3.302, + "step": 3739 + }, + { + "epoch": 0.11536578124457778, + "grad_norm": 0.17172034084796906, + "learning_rate": 4.99640653950689e-05, + "loss": 3.2904, + "step": 3740 + }, + { + "epoch": 0.11539662771015119, + "grad_norm": 0.1782234162092209, + "learning_rate": 4.9963920969247854e-05, + "loss": 3.3226, + "step": 3741 + }, + { + "epoch": 0.1154274741757246, + "grad_norm": 0.1688428372144699, + "learning_rate": 4.996377625398513e-05, + "loss": 3.3025, + "step": 3742 + }, + { + "epoch": 0.11545832064129802, + "grad_norm": 0.1978594958782196, + "learning_rate": 4.9963631249282435e-05, + "loss": 3.2842, + "step": 3743 + }, + { + "epoch": 0.11548916710687143, + "grad_norm": 0.1755095273256302, + "learning_rate": 4.996348595514143e-05, + "loss": 3.2933, + "step": 3744 + }, + { + "epoch": 0.11552001357244485, + "grad_norm": 0.18074259161949158, + "learning_rate": 4.996334037156381e-05, + "loss": 3.2898, + "step": 3745 + }, + { + "epoch": 0.11555086003801827, + "grad_norm": 0.18181811273097992, + "learning_rate": 4.996319449855126e-05, + "loss": 3.2716, + "step": 3746 + }, + { + "epoch": 0.11558170650359169, + "grad_norm": 0.1764526218175888, + "learning_rate": 4.9963048336105466e-05, + "loss": 3.3216, + "step": 3747 + }, + { + "epoch": 0.1156125529691651, + "grad_norm": 0.17518813908100128, + "learning_rate": 4.996290188422813e-05, + "loss": 3.301, + "step": 3748 + }, + { + "epoch": 0.11564339943473852, + "grad_norm": 0.17122553288936615, + "learning_rate": 4.9962755142920946e-05, + "loss": 3.3052, + "step": 3749 + }, + { + "epoch": 0.11567424590031193, + "grad_norm": 0.18407489359378815, + "learning_rate": 4.996260811218563e-05, + "loss": 3.2837, + "step": 3750 + }, + { + "epoch": 0.11570509236588535, + "grad_norm": 0.16705070436000824, + "learning_rate": 4.9962460792023855e-05, + "loss": 3.3007, + "step": 3751 + }, + { + "epoch": 0.11573593883145877, + "grad_norm": 0.1736566722393036, + "learning_rate": 4.996231318243736e-05, + "loss": 3.2911, + "step": 3752 + }, + { + "epoch": 0.11576678529703219, + "grad_norm": 0.18534529209136963, + "learning_rate": 4.996216528342784e-05, + "loss": 3.2893, + "step": 3753 + }, + { + "epoch": 0.1157976317626056, + "grad_norm": 0.1768919825553894, + "learning_rate": 4.9962017094997015e-05, + "loss": 3.2816, + "step": 3754 + }, + { + "epoch": 0.11582847822817902, + "grad_norm": 0.16207824647426605, + "learning_rate": 4.99618686171466e-05, + "loss": 3.2987, + "step": 3755 + }, + { + "epoch": 0.11585932469375243, + "grad_norm": 0.16719326376914978, + "learning_rate": 4.996171984987832e-05, + "loss": 3.284, + "step": 3756 + }, + { + "epoch": 0.11589017115932584, + "grad_norm": 0.16617241501808167, + "learning_rate": 4.9961570793193904e-05, + "loss": 3.2963, + "step": 3757 + }, + { + "epoch": 0.11592101762489927, + "grad_norm": 0.16991601884365082, + "learning_rate": 4.996142144709506e-05, + "loss": 3.2776, + "step": 3758 + }, + { + "epoch": 0.11595186409047269, + "grad_norm": 0.17839877307415009, + "learning_rate": 4.996127181158355e-05, + "loss": 3.3146, + "step": 3759 + }, + { + "epoch": 0.1159827105560461, + "grad_norm": 0.16772601008415222, + "learning_rate": 4.996112188666109e-05, + "loss": 3.2986, + "step": 3760 + }, + { + "epoch": 0.11601355702161951, + "grad_norm": 0.18146122992038727, + "learning_rate": 4.996097167232941e-05, + "loss": 3.2666, + "step": 3761 + }, + { + "epoch": 0.11604440348719293, + "grad_norm": 0.17753370106220245, + "learning_rate": 4.9960821168590275e-05, + "loss": 3.3101, + "step": 3762 + }, + { + "epoch": 0.11607524995276634, + "grad_norm": 0.17836636304855347, + "learning_rate": 4.996067037544542e-05, + "loss": 3.2992, + "step": 3763 + }, + { + "epoch": 0.11610609641833977, + "grad_norm": 0.17417164146900177, + "learning_rate": 4.996051929289658e-05, + "loss": 3.3273, + "step": 3764 + }, + { + "epoch": 0.11613694288391319, + "grad_norm": 0.17538928985595703, + "learning_rate": 4.996036792094553e-05, + "loss": 3.3131, + "step": 3765 + }, + { + "epoch": 0.1161677893494866, + "grad_norm": 0.18053771555423737, + "learning_rate": 4.9960216259594e-05, + "loss": 3.3031, + "step": 3766 + }, + { + "epoch": 0.11619863581506001, + "grad_norm": 0.16851434111595154, + "learning_rate": 4.996006430884378e-05, + "loss": 3.2988, + "step": 3767 + }, + { + "epoch": 0.11622948228063343, + "grad_norm": 0.17809858918190002, + "learning_rate": 4.9959912068696603e-05, + "loss": 3.2699, + "step": 3768 + }, + { + "epoch": 0.11626032874620684, + "grad_norm": 0.17091569304466248, + "learning_rate": 4.995975953915425e-05, + "loss": 3.2978, + "step": 3769 + }, + { + "epoch": 0.11629117521178027, + "grad_norm": 0.18414345383644104, + "learning_rate": 4.995960672021848e-05, + "loss": 3.2918, + "step": 3770 + }, + { + "epoch": 0.11632202167735368, + "grad_norm": 0.17186233401298523, + "learning_rate": 4.995945361189106e-05, + "loss": 3.2516, + "step": 3771 + }, + { + "epoch": 0.1163528681429271, + "grad_norm": 0.28597110509872437, + "learning_rate": 4.9959300214173786e-05, + "loss": 3.2823, + "step": 3772 + }, + { + "epoch": 0.11638371460850051, + "grad_norm": 0.17760534584522247, + "learning_rate": 4.9959146527068427e-05, + "loss": 3.2912, + "step": 3773 + }, + { + "epoch": 0.11641456107407393, + "grad_norm": 0.1793055236339569, + "learning_rate": 4.9958992550576754e-05, + "loss": 3.2702, + "step": 3774 + }, + { + "epoch": 0.11644540753964734, + "grad_norm": 0.19760476052761078, + "learning_rate": 4.9958838284700565e-05, + "loss": 3.3155, + "step": 3775 + }, + { + "epoch": 0.11647625400522077, + "grad_norm": 0.1690884828567505, + "learning_rate": 4.995868372944165e-05, + "loss": 3.2904, + "step": 3776 + }, + { + "epoch": 0.11650710047079418, + "grad_norm": 0.1857103705406189, + "learning_rate": 4.995852888480178e-05, + "loss": 3.3045, + "step": 3777 + }, + { + "epoch": 0.1165379469363676, + "grad_norm": 0.1769459992647171, + "learning_rate": 4.995837375078278e-05, + "loss": 3.2972, + "step": 3778 + }, + { + "epoch": 0.11656879340194101, + "grad_norm": 0.17999009788036346, + "learning_rate": 4.995821832738643e-05, + "loss": 3.2797, + "step": 3779 + }, + { + "epoch": 0.11659963986751443, + "grad_norm": 0.1905067265033722, + "learning_rate": 4.995806261461454e-05, + "loss": 3.2853, + "step": 3780 + }, + { + "epoch": 0.11663048633308784, + "grad_norm": 0.17743104696273804, + "learning_rate": 4.9957906612468905e-05, + "loss": 3.3081, + "step": 3781 + }, + { + "epoch": 0.11666133279866127, + "grad_norm": 0.1862344592809677, + "learning_rate": 4.995775032095135e-05, + "loss": 3.2506, + "step": 3782 + }, + { + "epoch": 0.11669217926423468, + "grad_norm": 0.17346693575382233, + "learning_rate": 4.995759374006367e-05, + "loss": 3.319, + "step": 3783 + }, + { + "epoch": 0.1167230257298081, + "grad_norm": 0.17995642125606537, + "learning_rate": 4.9957436869807694e-05, + "loss": 3.3013, + "step": 3784 + }, + { + "epoch": 0.11675387219538151, + "grad_norm": 0.1909841001033783, + "learning_rate": 4.995727971018523e-05, + "loss": 3.2909, + "step": 3785 + }, + { + "epoch": 0.11678471866095493, + "grad_norm": 0.17080987989902496, + "learning_rate": 4.995712226119811e-05, + "loss": 3.3123, + "step": 3786 + }, + { + "epoch": 0.11681556512652834, + "grad_norm": 0.18901073932647705, + "learning_rate": 4.9956964522848154e-05, + "loss": 3.3076, + "step": 3787 + }, + { + "epoch": 0.11684641159210177, + "grad_norm": 0.18396559357643127, + "learning_rate": 4.995680649513719e-05, + "loss": 3.3008, + "step": 3788 + }, + { + "epoch": 0.11687725805767518, + "grad_norm": 0.1928137242794037, + "learning_rate": 4.9956648178067054e-05, + "loss": 3.2766, + "step": 3789 + }, + { + "epoch": 0.1169081045232486, + "grad_norm": 0.1692982316017151, + "learning_rate": 4.9956489571639577e-05, + "loss": 3.249, + "step": 3790 + }, + { + "epoch": 0.11693895098882201, + "grad_norm": 0.18703845143318176, + "learning_rate": 4.99563306758566e-05, + "loss": 3.3245, + "step": 3791 + }, + { + "epoch": 0.11696979745439542, + "grad_norm": 0.17601241171360016, + "learning_rate": 4.9956171490719975e-05, + "loss": 3.3035, + "step": 3792 + }, + { + "epoch": 0.11700064391996884, + "grad_norm": 0.18174943327903748, + "learning_rate": 4.995601201623152e-05, + "loss": 3.2853, + "step": 3793 + }, + { + "epoch": 0.11703149038554227, + "grad_norm": 0.17289750277996063, + "learning_rate": 4.995585225239312e-05, + "loss": 3.2874, + "step": 3794 + }, + { + "epoch": 0.11706233685111568, + "grad_norm": 0.1806178092956543, + "learning_rate": 4.99556921992066e-05, + "loss": 3.2963, + "step": 3795 + }, + { + "epoch": 0.1170931833166891, + "grad_norm": 0.169141486287117, + "learning_rate": 4.9955531856673833e-05, + "loss": 3.2749, + "step": 3796 + }, + { + "epoch": 0.11712402978226251, + "grad_norm": 0.19368493556976318, + "learning_rate": 4.995537122479666e-05, + "loss": 3.2786, + "step": 3797 + }, + { + "epoch": 0.11715487624783592, + "grad_norm": 0.17801842093467712, + "learning_rate": 4.995521030357696e-05, + "loss": 3.293, + "step": 3798 + }, + { + "epoch": 0.11718572271340934, + "grad_norm": 0.18392382562160492, + "learning_rate": 4.99550490930166e-05, + "loss": 3.3136, + "step": 3799 + }, + { + "epoch": 0.11721656917898277, + "grad_norm": 0.24419118463993073, + "learning_rate": 4.9954887593117436e-05, + "loss": 3.3031, + "step": 3800 + }, + { + "epoch": 0.11724741564455618, + "grad_norm": 0.19947226345539093, + "learning_rate": 4.995472580388134e-05, + "loss": 3.2737, + "step": 3801 + }, + { + "epoch": 0.1172782621101296, + "grad_norm": 0.18659885227680206, + "learning_rate": 4.99545637253102e-05, + "loss": 3.2798, + "step": 3802 + }, + { + "epoch": 0.11730910857570301, + "grad_norm": 0.16860777139663696, + "learning_rate": 4.995440135740589e-05, + "loss": 3.2685, + "step": 3803 + }, + { + "epoch": 0.11733995504127642, + "grad_norm": 0.18382276594638824, + "learning_rate": 4.9954238700170296e-05, + "loss": 3.3066, + "step": 3804 + }, + { + "epoch": 0.11737080150684984, + "grad_norm": 0.17440412938594818, + "learning_rate": 4.9954075753605296e-05, + "loss": 3.297, + "step": 3805 + }, + { + "epoch": 0.11740164797242326, + "grad_norm": 0.16364949941635132, + "learning_rate": 4.995391251771278e-05, + "loss": 3.2845, + "step": 3806 + }, + { + "epoch": 0.11743249443799668, + "grad_norm": 0.17847363650798798, + "learning_rate": 4.995374899249465e-05, + "loss": 3.3159, + "step": 3807 + }, + { + "epoch": 0.1174633409035701, + "grad_norm": 0.1732705533504486, + "learning_rate": 4.995358517795279e-05, + "loss": 3.2676, + "step": 3808 + }, + { + "epoch": 0.11749418736914351, + "grad_norm": 0.17717574536800385, + "learning_rate": 4.995342107408911e-05, + "loss": 3.3016, + "step": 3809 + }, + { + "epoch": 0.11752503383471692, + "grad_norm": 0.17216923832893372, + "learning_rate": 4.99532566809055e-05, + "loss": 3.287, + "step": 3810 + }, + { + "epoch": 0.11755588030029034, + "grad_norm": 0.17400088906288147, + "learning_rate": 4.995309199840389e-05, + "loss": 3.3053, + "step": 3811 + }, + { + "epoch": 0.11758672676586376, + "grad_norm": 0.17065110802650452, + "learning_rate": 4.995292702658616e-05, + "loss": 3.3049, + "step": 3812 + }, + { + "epoch": 0.11761757323143718, + "grad_norm": 0.16946947574615479, + "learning_rate": 4.9952761765454236e-05, + "loss": 3.2998, + "step": 3813 + }, + { + "epoch": 0.11764841969701059, + "grad_norm": 0.16844843327999115, + "learning_rate": 4.995259621501004e-05, + "loss": 3.2742, + "step": 3814 + }, + { + "epoch": 0.117679266162584, + "grad_norm": 0.1719282865524292, + "learning_rate": 4.9952430375255486e-05, + "loss": 3.2838, + "step": 3815 + }, + { + "epoch": 0.11771011262815742, + "grad_norm": 0.17916220426559448, + "learning_rate": 4.995226424619249e-05, + "loss": 3.2656, + "step": 3816 + }, + { + "epoch": 0.11774095909373083, + "grad_norm": 0.17115333676338196, + "learning_rate": 4.995209782782299e-05, + "loss": 3.2825, + "step": 3817 + }, + { + "epoch": 0.11777180555930426, + "grad_norm": 0.16609306633472443, + "learning_rate": 4.995193112014891e-05, + "loss": 3.2747, + "step": 3818 + }, + { + "epoch": 0.11780265202487768, + "grad_norm": 0.1862884759902954, + "learning_rate": 4.9951764123172184e-05, + "loss": 3.2819, + "step": 3819 + }, + { + "epoch": 0.11783349849045109, + "grad_norm": 0.16677211225032806, + "learning_rate": 4.995159683689474e-05, + "loss": 3.3225, + "step": 3820 + }, + { + "epoch": 0.1178643449560245, + "grad_norm": 0.16397804021835327, + "learning_rate": 4.995142926131853e-05, + "loss": 3.2859, + "step": 3821 + }, + { + "epoch": 0.11789519142159792, + "grad_norm": 0.1779787689447403, + "learning_rate": 4.9951261396445494e-05, + "loss": 3.2765, + "step": 3822 + }, + { + "epoch": 0.11792603788717133, + "grad_norm": 0.16412948071956635, + "learning_rate": 4.995109324227758e-05, + "loss": 3.2951, + "step": 3823 + }, + { + "epoch": 0.11795688435274476, + "grad_norm": 0.173862487077713, + "learning_rate": 4.9950924798816724e-05, + "loss": 3.3116, + "step": 3824 + }, + { + "epoch": 0.11798773081831818, + "grad_norm": 0.16601316630840302, + "learning_rate": 4.995075606606489e-05, + "loss": 3.2581, + "step": 3825 + }, + { + "epoch": 0.11801857728389159, + "grad_norm": 0.16762664914131165, + "learning_rate": 4.9950587044024036e-05, + "loss": 3.2914, + "step": 3826 + }, + { + "epoch": 0.118049423749465, + "grad_norm": 0.1767447143793106, + "learning_rate": 4.995041773269612e-05, + "loss": 3.265, + "step": 3827 + }, + { + "epoch": 0.11808027021503842, + "grad_norm": 0.17082534730434418, + "learning_rate": 4.99502481320831e-05, + "loss": 3.2772, + "step": 3828 + }, + { + "epoch": 0.11811111668061183, + "grad_norm": 0.17433147132396698, + "learning_rate": 4.995007824218695e-05, + "loss": 3.2698, + "step": 3829 + }, + { + "epoch": 0.11814196314618526, + "grad_norm": 0.1855974793434143, + "learning_rate": 4.994990806300963e-05, + "loss": 3.3167, + "step": 3830 + }, + { + "epoch": 0.11817280961175868, + "grad_norm": 0.17799992859363556, + "learning_rate": 4.994973759455312e-05, + "loss": 3.2955, + "step": 3831 + }, + { + "epoch": 0.11820365607733209, + "grad_norm": 0.16666212677955627, + "learning_rate": 4.9949566836819404e-05, + "loss": 3.307, + "step": 3832 + }, + { + "epoch": 0.1182345025429055, + "grad_norm": 0.21706143021583557, + "learning_rate": 4.994939578981045e-05, + "loss": 3.2752, + "step": 3833 + }, + { + "epoch": 0.11826534900847892, + "grad_norm": 0.17634814977645874, + "learning_rate": 4.9949224453528245e-05, + "loss": 3.276, + "step": 3834 + }, + { + "epoch": 0.11829619547405233, + "grad_norm": 0.1715715527534485, + "learning_rate": 4.994905282797477e-05, + "loss": 3.2618, + "step": 3835 + }, + { + "epoch": 0.11832704193962576, + "grad_norm": 0.28393256664276123, + "learning_rate": 4.994888091315202e-05, + "loss": 3.2734, + "step": 3836 + }, + { + "epoch": 0.11835788840519917, + "grad_norm": 0.18201537430286407, + "learning_rate": 4.9948708709062e-05, + "loss": 3.2831, + "step": 3837 + }, + { + "epoch": 0.11838873487077259, + "grad_norm": 0.1798393428325653, + "learning_rate": 4.994853621570668e-05, + "loss": 3.3099, + "step": 3838 + }, + { + "epoch": 0.118419581336346, + "grad_norm": 0.18856699764728546, + "learning_rate": 4.9948363433088085e-05, + "loss": 3.2689, + "step": 3839 + }, + { + "epoch": 0.11845042780191942, + "grad_norm": 1.097495198249817, + "learning_rate": 4.9948190361208205e-05, + "loss": 3.2825, + "step": 3840 + }, + { + "epoch": 0.11848127426749283, + "grad_norm": 0.1780746430158615, + "learning_rate": 4.994801700006905e-05, + "loss": 3.284, + "step": 3841 + }, + { + "epoch": 0.11851212073306626, + "grad_norm": 0.1978173404932022, + "learning_rate": 4.994784334967263e-05, + "loss": 3.2574, + "step": 3842 + }, + { + "epoch": 0.11854296719863967, + "grad_norm": 0.6922443509101868, + "learning_rate": 4.994766941002096e-05, + "loss": 3.3123, + "step": 3843 + }, + { + "epoch": 0.11857381366421309, + "grad_norm": 0.22216275334358215, + "learning_rate": 4.994749518111604e-05, + "loss": 3.2864, + "step": 3844 + }, + { + "epoch": 0.1186046601297865, + "grad_norm": 0.20907555520534515, + "learning_rate": 4.994732066295993e-05, + "loss": 3.2903, + "step": 3845 + }, + { + "epoch": 0.11863550659535992, + "grad_norm": 0.30870312452316284, + "learning_rate": 4.994714585555461e-05, + "loss": 3.3015, + "step": 3846 + }, + { + "epoch": 0.11866635306093333, + "grad_norm": 0.1944558173418045, + "learning_rate": 4.9946970758902136e-05, + "loss": 3.2926, + "step": 3847 + }, + { + "epoch": 0.11869719952650676, + "grad_norm": 0.3746813237667084, + "learning_rate": 4.994679537300452e-05, + "loss": 3.2987, + "step": 3848 + }, + { + "epoch": 0.11872804599208017, + "grad_norm": 0.1975131630897522, + "learning_rate": 4.9946619697863804e-05, + "loss": 3.2676, + "step": 3849 + }, + { + "epoch": 0.11875889245765359, + "grad_norm": 0.19565124809741974, + "learning_rate": 4.994644373348203e-05, + "loss": 3.25, + "step": 3850 + }, + { + "epoch": 0.118789738923227, + "grad_norm": 0.31717318296432495, + "learning_rate": 4.9946267479861223e-05, + "loss": 3.3083, + "step": 3851 + }, + { + "epoch": 0.11882058538880041, + "grad_norm": 0.21934707462787628, + "learning_rate": 4.9946090937003444e-05, + "loss": 3.3176, + "step": 3852 + }, + { + "epoch": 0.11885143185437384, + "grad_norm": 0.20590052008628845, + "learning_rate": 4.9945914104910727e-05, + "loss": 3.2754, + "step": 3853 + }, + { + "epoch": 0.11888227831994726, + "grad_norm": 0.18323637545108795, + "learning_rate": 4.994573698358513e-05, + "loss": 3.283, + "step": 3854 + }, + { + "epoch": 0.11891312478552067, + "grad_norm": 0.19993147253990173, + "learning_rate": 4.99455595730287e-05, + "loss": 3.3117, + "step": 3855 + }, + { + "epoch": 0.11894397125109409, + "grad_norm": 0.18442073464393616, + "learning_rate": 4.994538187324349e-05, + "loss": 3.3076, + "step": 3856 + }, + { + "epoch": 0.1189748177166675, + "grad_norm": 0.19373932480812073, + "learning_rate": 4.9945203884231584e-05, + "loss": 3.2926, + "step": 3857 + }, + { + "epoch": 0.11900566418224091, + "grad_norm": 0.19132766127586365, + "learning_rate": 4.994502560599502e-05, + "loss": 3.3056, + "step": 3858 + }, + { + "epoch": 0.11903651064781434, + "grad_norm": 0.2132892608642578, + "learning_rate": 4.994484703853588e-05, + "loss": 3.2696, + "step": 3859 + }, + { + "epoch": 0.11906735711338776, + "grad_norm": 0.19707615673542023, + "learning_rate": 4.994466818185622e-05, + "loss": 3.2912, + "step": 3860 + }, + { + "epoch": 0.11909820357896117, + "grad_norm": 0.17522555589675903, + "learning_rate": 4.994448903595813e-05, + "loss": 3.2954, + "step": 3861 + }, + { + "epoch": 0.11912905004453458, + "grad_norm": 0.1795511096715927, + "learning_rate": 4.9944309600843675e-05, + "loss": 3.2934, + "step": 3862 + }, + { + "epoch": 0.119159896510108, + "grad_norm": 0.17678096890449524, + "learning_rate": 4.994412987651494e-05, + "loss": 3.3449, + "step": 3863 + }, + { + "epoch": 0.11919074297568141, + "grad_norm": 0.18615111708641052, + "learning_rate": 4.9943949862974e-05, + "loss": 3.3301, + "step": 3864 + }, + { + "epoch": 0.11922158944125484, + "grad_norm": 0.1728835105895996, + "learning_rate": 4.9943769560222966e-05, + "loss": 3.2868, + "step": 3865 + }, + { + "epoch": 0.11925243590682826, + "grad_norm": 0.18306703865528107, + "learning_rate": 4.994358896826391e-05, + "loss": 3.2251, + "step": 3866 + }, + { + "epoch": 0.11928328237240167, + "grad_norm": 0.16857045888900757, + "learning_rate": 4.994340808709893e-05, + "loss": 3.3289, + "step": 3867 + }, + { + "epoch": 0.11931412883797508, + "grad_norm": 0.1794869750738144, + "learning_rate": 4.9943226916730116e-05, + "loss": 3.2877, + "step": 3868 + }, + { + "epoch": 0.1193449753035485, + "grad_norm": 0.4495128393173218, + "learning_rate": 4.994304545715958e-05, + "loss": 3.2805, + "step": 3869 + }, + { + "epoch": 0.11937582176912191, + "grad_norm": 0.1651013344526291, + "learning_rate": 4.994286370838942e-05, + "loss": 3.2574, + "step": 3870 + }, + { + "epoch": 0.11940666823469534, + "grad_norm": 0.19752277433872223, + "learning_rate": 4.994268167042174e-05, + "loss": 3.2995, + "step": 3871 + }, + { + "epoch": 0.11943751470026875, + "grad_norm": 0.1734388917684555, + "learning_rate": 4.994249934325866e-05, + "loss": 3.2652, + "step": 3872 + }, + { + "epoch": 0.11946836116584217, + "grad_norm": 0.19003283977508545, + "learning_rate": 4.994231672690229e-05, + "loss": 3.3156, + "step": 3873 + }, + { + "epoch": 0.11949920763141558, + "grad_norm": 0.7811154127120972, + "learning_rate": 4.9942133821354745e-05, + "loss": 3.3339, + "step": 3874 + }, + { + "epoch": 0.119530054096989, + "grad_norm": 0.6561194658279419, + "learning_rate": 4.994195062661814e-05, + "loss": 3.2892, + "step": 3875 + }, + { + "epoch": 0.11956090056256241, + "grad_norm": 0.3895606994628906, + "learning_rate": 4.994176714269462e-05, + "loss": 3.2873, + "step": 3876 + }, + { + "epoch": 0.11959174702813584, + "grad_norm": 0.41113415360450745, + "learning_rate": 4.994158336958629e-05, + "loss": 3.2891, + "step": 3877 + }, + { + "epoch": 0.11962259349370925, + "grad_norm": 0.26596876978874207, + "learning_rate": 4.9941399307295286e-05, + "loss": 3.2878, + "step": 3878 + }, + { + "epoch": 0.11965343995928267, + "grad_norm": 0.23346969485282898, + "learning_rate": 4.9941214955823745e-05, + "loss": 3.2924, + "step": 3879 + }, + { + "epoch": 0.11968428642485608, + "grad_norm": 0.20183660089969635, + "learning_rate": 4.994103031517381e-05, + "loss": 3.3431, + "step": 3880 + }, + { + "epoch": 0.1197151328904295, + "grad_norm": 0.20199674367904663, + "learning_rate": 4.9940845385347615e-05, + "loss": 3.2863, + "step": 3881 + }, + { + "epoch": 0.11974597935600291, + "grad_norm": 0.19917036592960358, + "learning_rate": 4.99406601663473e-05, + "loss": 3.3007, + "step": 3882 + }, + { + "epoch": 0.11977682582157634, + "grad_norm": 0.18916761875152588, + "learning_rate": 4.994047465817502e-05, + "loss": 3.2752, + "step": 3883 + }, + { + "epoch": 0.11980767228714975, + "grad_norm": 0.1833811104297638, + "learning_rate": 4.994028886083293e-05, + "loss": 3.2745, + "step": 3884 + }, + { + "epoch": 0.11983851875272317, + "grad_norm": 0.3927651047706604, + "learning_rate": 4.994010277432317e-05, + "loss": 3.2906, + "step": 3885 + }, + { + "epoch": 0.11986936521829658, + "grad_norm": 0.17558392882347107, + "learning_rate": 4.9939916398647915e-05, + "loss": 3.3003, + "step": 3886 + }, + { + "epoch": 0.11990021168387, + "grad_norm": 0.9533789753913879, + "learning_rate": 4.993972973380931e-05, + "loss": 3.2982, + "step": 3887 + }, + { + "epoch": 0.11993105814944341, + "grad_norm": 0.1870499700307846, + "learning_rate": 4.9939542779809526e-05, + "loss": 3.3031, + "step": 3888 + }, + { + "epoch": 0.11996190461501684, + "grad_norm": 0.1943441480398178, + "learning_rate": 4.9939355536650736e-05, + "loss": 3.3066, + "step": 3889 + }, + { + "epoch": 0.11999275108059025, + "grad_norm": 0.19711712002754211, + "learning_rate": 4.99391680043351e-05, + "loss": 3.2409, + "step": 3890 + }, + { + "epoch": 0.12002359754616367, + "grad_norm": 0.19165292382240295, + "learning_rate": 4.9938980182864794e-05, + "loss": 3.3002, + "step": 3891 + }, + { + "epoch": 0.12005444401173708, + "grad_norm": 0.1886846125125885, + "learning_rate": 4.9938792072242005e-05, + "loss": 3.2971, + "step": 3892 + }, + { + "epoch": 0.1200852904773105, + "grad_norm": 0.20419704914093018, + "learning_rate": 4.993860367246891e-05, + "loss": 3.2795, + "step": 3893 + }, + { + "epoch": 0.12011613694288391, + "grad_norm": 0.19810377061367035, + "learning_rate": 4.993841498354769e-05, + "loss": 3.2944, + "step": 3894 + }, + { + "epoch": 0.12014698340845734, + "grad_norm": 0.1875728815793991, + "learning_rate": 4.993822600548053e-05, + "loss": 3.2866, + "step": 3895 + }, + { + "epoch": 0.12017782987403075, + "grad_norm": 0.20550711452960968, + "learning_rate": 4.993803673826963e-05, + "loss": 3.331, + "step": 3896 + }, + { + "epoch": 0.12020867633960416, + "grad_norm": 0.18017417192459106, + "learning_rate": 4.9937847181917175e-05, + "loss": 3.2673, + "step": 3897 + }, + { + "epoch": 0.12023952280517758, + "grad_norm": 0.18098901212215424, + "learning_rate": 4.993765733642537e-05, + "loss": 3.2432, + "step": 3898 + }, + { + "epoch": 0.12027036927075099, + "grad_norm": 0.17539949715137482, + "learning_rate": 4.9937467201796415e-05, + "loss": 3.2967, + "step": 3899 + }, + { + "epoch": 0.12030121573632441, + "grad_norm": 0.17753224074840546, + "learning_rate": 4.993727677803251e-05, + "loss": 3.3224, + "step": 3900 + }, + { + "epoch": 0.12033206220189784, + "grad_norm": 0.17469505965709686, + "learning_rate": 4.993708606513587e-05, + "loss": 3.3172, + "step": 3901 + }, + { + "epoch": 0.12036290866747125, + "grad_norm": 0.1857665777206421, + "learning_rate": 4.993689506310871e-05, + "loss": 3.3215, + "step": 3902 + }, + { + "epoch": 0.12039375513304466, + "grad_norm": 0.17383983731269836, + "learning_rate": 4.993670377195323e-05, + "loss": 3.2903, + "step": 3903 + }, + { + "epoch": 0.12042460159861808, + "grad_norm": 0.18013307452201843, + "learning_rate": 4.9936512191671655e-05, + "loss": 3.3057, + "step": 3904 + }, + { + "epoch": 0.12045544806419149, + "grad_norm": 0.18734177947044373, + "learning_rate": 4.993632032226621e-05, + "loss": 3.3539, + "step": 3905 + }, + { + "epoch": 0.1204862945297649, + "grad_norm": 0.17778846621513367, + "learning_rate": 4.993612816373911e-05, + "loss": 3.2959, + "step": 3906 + }, + { + "epoch": 0.12051714099533833, + "grad_norm": 0.18231722712516785, + "learning_rate": 4.993593571609259e-05, + "loss": 3.3197, + "step": 3907 + }, + { + "epoch": 0.12054798746091175, + "grad_norm": 0.18399296700954437, + "learning_rate": 4.993574297932888e-05, + "loss": 3.3044, + "step": 3908 + }, + { + "epoch": 0.12057883392648516, + "grad_norm": 0.17699283361434937, + "learning_rate": 4.993554995345022e-05, + "loss": 3.3184, + "step": 3909 + }, + { + "epoch": 0.12060968039205858, + "grad_norm": 0.16961337625980377, + "learning_rate": 4.993535663845884e-05, + "loss": 3.2878, + "step": 3910 + }, + { + "epoch": 0.12064052685763199, + "grad_norm": 0.17185397446155548, + "learning_rate": 4.993516303435698e-05, + "loss": 3.2819, + "step": 3911 + }, + { + "epoch": 0.1206713733232054, + "grad_norm": 0.17200416326522827, + "learning_rate": 4.993496914114689e-05, + "loss": 3.303, + "step": 3912 + }, + { + "epoch": 0.12070221978877883, + "grad_norm": 0.1787317544221878, + "learning_rate": 4.9934774958830824e-05, + "loss": 3.3094, + "step": 3913 + }, + { + "epoch": 0.12073306625435225, + "grad_norm": 0.7050204277038574, + "learning_rate": 4.993458048741102e-05, + "loss": 3.3008, + "step": 3914 + }, + { + "epoch": 0.12076391271992566, + "grad_norm": 0.17545925080776215, + "learning_rate": 4.993438572688974e-05, + "loss": 3.3284, + "step": 3915 + }, + { + "epoch": 0.12079475918549908, + "grad_norm": 0.21847964823246002, + "learning_rate": 4.9934190677269235e-05, + "loss": 3.274, + "step": 3916 + }, + { + "epoch": 0.12082560565107249, + "grad_norm": 0.17803841829299927, + "learning_rate": 4.9933995338551784e-05, + "loss": 3.3015, + "step": 3917 + }, + { + "epoch": 0.1208564521166459, + "grad_norm": 0.17236389219760895, + "learning_rate": 4.993379971073964e-05, + "loss": 3.3041, + "step": 3918 + }, + { + "epoch": 0.12088729858221933, + "grad_norm": 0.17285862565040588, + "learning_rate": 4.9933603793835064e-05, + "loss": 3.2869, + "step": 3919 + }, + { + "epoch": 0.12091814504779275, + "grad_norm": 0.16563667356967926, + "learning_rate": 4.993340758784034e-05, + "loss": 3.3089, + "step": 3920 + }, + { + "epoch": 0.12094899151336616, + "grad_norm": 0.17397160828113556, + "learning_rate": 4.993321109275774e-05, + "loss": 3.2719, + "step": 3921 + }, + { + "epoch": 0.12097983797893958, + "grad_norm": 0.17132100462913513, + "learning_rate": 4.9933014308589535e-05, + "loss": 3.2994, + "step": 3922 + }, + { + "epoch": 0.12101068444451299, + "grad_norm": 0.1788635551929474, + "learning_rate": 4.993281723533801e-05, + "loss": 3.2838, + "step": 3923 + }, + { + "epoch": 0.1210415309100864, + "grad_norm": 0.1697738915681839, + "learning_rate": 4.993261987300546e-05, + "loss": 3.3108, + "step": 3924 + }, + { + "epoch": 0.12107237737565983, + "grad_norm": 0.1717386394739151, + "learning_rate": 4.993242222159416e-05, + "loss": 3.2611, + "step": 3925 + }, + { + "epoch": 0.12110322384123325, + "grad_norm": 0.1746242493391037, + "learning_rate": 4.993222428110642e-05, + "loss": 3.2767, + "step": 3926 + }, + { + "epoch": 0.12113407030680666, + "grad_norm": 0.17377768456935883, + "learning_rate": 4.99320260515445e-05, + "loss": 3.3198, + "step": 3927 + }, + { + "epoch": 0.12116491677238007, + "grad_norm": 0.17524808645248413, + "learning_rate": 4.9931827532910734e-05, + "loss": 3.2768, + "step": 3928 + }, + { + "epoch": 0.12119576323795349, + "grad_norm": 0.1618947684764862, + "learning_rate": 4.9931628725207415e-05, + "loss": 3.2741, + "step": 3929 + }, + { + "epoch": 0.1212266097035269, + "grad_norm": 0.16221055388450623, + "learning_rate": 4.993142962843683e-05, + "loss": 3.2555, + "step": 3930 + }, + { + "epoch": 0.12125745616910033, + "grad_norm": 0.16396845877170563, + "learning_rate": 4.993123024260131e-05, + "loss": 3.2958, + "step": 3931 + }, + { + "epoch": 0.12128830263467374, + "grad_norm": 0.16880400478839874, + "learning_rate": 4.9931030567703154e-05, + "loss": 3.2965, + "step": 3932 + }, + { + "epoch": 0.12131914910024716, + "grad_norm": 0.17384275794029236, + "learning_rate": 4.993083060374468e-05, + "loss": 3.3047, + "step": 3933 + }, + { + "epoch": 0.12134999556582057, + "grad_norm": 0.17721746861934662, + "learning_rate": 4.99306303507282e-05, + "loss": 3.2419, + "step": 3934 + }, + { + "epoch": 0.12138084203139399, + "grad_norm": 0.1731419563293457, + "learning_rate": 4.993042980865605e-05, + "loss": 3.2547, + "step": 3935 + }, + { + "epoch": 0.1214116884969674, + "grad_norm": 0.1671038717031479, + "learning_rate": 4.993022897753056e-05, + "loss": 3.2607, + "step": 3936 + }, + { + "epoch": 0.12144253496254083, + "grad_norm": 0.1738748848438263, + "learning_rate": 4.993002785735403e-05, + "loss": 3.2861, + "step": 3937 + }, + { + "epoch": 0.12147338142811424, + "grad_norm": 0.17665423452854156, + "learning_rate": 4.9929826448128814e-05, + "loss": 3.3193, + "step": 3938 + }, + { + "epoch": 0.12150422789368766, + "grad_norm": 0.1668185293674469, + "learning_rate": 4.992962474985724e-05, + "loss": 3.2742, + "step": 3939 + }, + { + "epoch": 0.12153507435926107, + "grad_norm": 0.16406366229057312, + "learning_rate": 4.9929422762541645e-05, + "loss": 3.2501, + "step": 3940 + }, + { + "epoch": 0.12156592082483449, + "grad_norm": 0.165787011384964, + "learning_rate": 4.992922048618437e-05, + "loss": 3.2789, + "step": 3941 + }, + { + "epoch": 0.1215967672904079, + "grad_norm": 0.16674834489822388, + "learning_rate": 4.9929017920787774e-05, + "loss": 3.2914, + "step": 3942 + }, + { + "epoch": 0.12162761375598133, + "grad_norm": 0.16661494970321655, + "learning_rate": 4.992881506635419e-05, + "loss": 3.2771, + "step": 3943 + }, + { + "epoch": 0.12165846022155474, + "grad_norm": 0.16650833189487457, + "learning_rate": 4.992861192288598e-05, + "loss": 3.3105, + "step": 3944 + }, + { + "epoch": 0.12168930668712816, + "grad_norm": 0.17752507328987122, + "learning_rate": 4.992840849038549e-05, + "loss": 3.3479, + "step": 3945 + }, + { + "epoch": 0.12172015315270157, + "grad_norm": 0.1651119589805603, + "learning_rate": 4.992820476885509e-05, + "loss": 3.312, + "step": 3946 + }, + { + "epoch": 0.12175099961827499, + "grad_norm": 0.16742385923862457, + "learning_rate": 4.992800075829713e-05, + "loss": 3.3267, + "step": 3947 + }, + { + "epoch": 0.1217818460838484, + "grad_norm": 0.16215075552463531, + "learning_rate": 4.992779645871399e-05, + "loss": 3.2986, + "step": 3948 + }, + { + "epoch": 0.12181269254942183, + "grad_norm": 0.16843323409557343, + "learning_rate": 4.9927591870108024e-05, + "loss": 3.3014, + "step": 3949 + }, + { + "epoch": 0.12184353901499524, + "grad_norm": 0.17127962410449982, + "learning_rate": 4.9927386992481604e-05, + "loss": 3.2548, + "step": 3950 + }, + { + "epoch": 0.12187438548056866, + "grad_norm": 0.16568636894226074, + "learning_rate": 4.992718182583712e-05, + "loss": 3.2838, + "step": 3951 + }, + { + "epoch": 0.12190523194614207, + "grad_norm": 0.16260802745819092, + "learning_rate": 4.992697637017694e-05, + "loss": 3.3287, + "step": 3952 + }, + { + "epoch": 0.12193607841171548, + "grad_norm": 0.16885177791118622, + "learning_rate": 4.992677062550345e-05, + "loss": 3.3407, + "step": 3953 + }, + { + "epoch": 0.1219669248772889, + "grad_norm": 0.17274171113967896, + "learning_rate": 4.992656459181903e-05, + "loss": 3.294, + "step": 3954 + }, + { + "epoch": 0.12199777134286233, + "grad_norm": 0.16951894760131836, + "learning_rate": 4.992635826912607e-05, + "loss": 3.2658, + "step": 3955 + }, + { + "epoch": 0.12202861780843574, + "grad_norm": 0.1733904629945755, + "learning_rate": 4.9926151657426976e-05, + "loss": 3.2937, + "step": 3956 + }, + { + "epoch": 0.12205946427400916, + "grad_norm": 0.16862952709197998, + "learning_rate": 4.992594475672413e-05, + "loss": 3.2878, + "step": 3957 + }, + { + "epoch": 0.12209031073958257, + "grad_norm": 0.1654517501592636, + "learning_rate": 4.9925737567019925e-05, + "loss": 3.2923, + "step": 3958 + }, + { + "epoch": 0.12212115720515598, + "grad_norm": 0.16730013489723206, + "learning_rate": 4.9925530088316775e-05, + "loss": 3.3015, + "step": 3959 + }, + { + "epoch": 0.1221520036707294, + "grad_norm": 0.17082266509532928, + "learning_rate": 4.992532232061708e-05, + "loss": 3.2772, + "step": 3960 + }, + { + "epoch": 0.12218285013630283, + "grad_norm": 0.1786566972732544, + "learning_rate": 4.9925114263923255e-05, + "loss": 3.3069, + "step": 3961 + }, + { + "epoch": 0.12221369660187624, + "grad_norm": 0.17098164558410645, + "learning_rate": 4.992490591823771e-05, + "loss": 3.3097, + "step": 3962 + }, + { + "epoch": 0.12224454306744965, + "grad_norm": 0.1689991056919098, + "learning_rate": 4.992469728356286e-05, + "loss": 3.2712, + "step": 3963 + }, + { + "epoch": 0.12227538953302307, + "grad_norm": 0.16885438561439514, + "learning_rate": 4.9924488359901114e-05, + "loss": 3.2586, + "step": 3964 + }, + { + "epoch": 0.12230623599859648, + "grad_norm": 0.16750849783420563, + "learning_rate": 4.9924279147254904e-05, + "loss": 3.2982, + "step": 3965 + }, + { + "epoch": 0.1223370824641699, + "grad_norm": 0.1652289181947708, + "learning_rate": 4.992406964562666e-05, + "loss": 3.2814, + "step": 3966 + }, + { + "epoch": 0.12236792892974332, + "grad_norm": 0.17100952565670013, + "learning_rate": 4.992385985501881e-05, + "loss": 3.265, + "step": 3967 + }, + { + "epoch": 0.12239877539531674, + "grad_norm": 0.17197376489639282, + "learning_rate": 4.992364977543378e-05, + "loss": 3.2654, + "step": 3968 + }, + { + "epoch": 0.12242962186089015, + "grad_norm": 0.17214030027389526, + "learning_rate": 4.992343940687401e-05, + "loss": 3.2651, + "step": 3969 + }, + { + "epoch": 0.12246046832646357, + "grad_norm": 0.17029306292533875, + "learning_rate": 4.992322874934193e-05, + "loss": 3.3094, + "step": 3970 + }, + { + "epoch": 0.12249131479203698, + "grad_norm": 0.18102781474590302, + "learning_rate": 4.992301780283999e-05, + "loss": 3.2836, + "step": 3971 + }, + { + "epoch": 0.1225221612576104, + "grad_norm": 0.16612353920936584, + "learning_rate": 4.992280656737064e-05, + "loss": 3.3175, + "step": 3972 + }, + { + "epoch": 0.12255300772318382, + "grad_norm": 0.16919675469398499, + "learning_rate": 4.992259504293632e-05, + "loss": 3.2941, + "step": 3973 + }, + { + "epoch": 0.12258385418875724, + "grad_norm": 0.17363549768924713, + "learning_rate": 4.992238322953949e-05, + "loss": 3.2683, + "step": 3974 + }, + { + "epoch": 0.12261470065433065, + "grad_norm": 0.1687847226858139, + "learning_rate": 4.99221711271826e-05, + "loss": 3.2986, + "step": 3975 + }, + { + "epoch": 0.12264554711990407, + "grad_norm": 0.18183737993240356, + "learning_rate": 4.9921958735868126e-05, + "loss": 3.2621, + "step": 3976 + }, + { + "epoch": 0.12267639358547748, + "grad_norm": 0.1632104218006134, + "learning_rate": 4.99217460555985e-05, + "loss": 3.2801, + "step": 3977 + }, + { + "epoch": 0.1227072400510509, + "grad_norm": 0.173264741897583, + "learning_rate": 4.9921533086376215e-05, + "loss": 3.3001, + "step": 3978 + }, + { + "epoch": 0.12273808651662432, + "grad_norm": 0.1661836951971054, + "learning_rate": 4.992131982820373e-05, + "loss": 3.3113, + "step": 3979 + }, + { + "epoch": 0.12276893298219774, + "grad_norm": 0.1787884384393692, + "learning_rate": 4.9921106281083515e-05, + "loss": 3.2621, + "step": 3980 + }, + { + "epoch": 0.12279977944777115, + "grad_norm": 0.16649362444877625, + "learning_rate": 4.992089244501804e-05, + "loss": 3.279, + "step": 3981 + }, + { + "epoch": 0.12283062591334457, + "grad_norm": 0.18053847551345825, + "learning_rate": 4.9920678320009804e-05, + "loss": 3.3279, + "step": 3982 + }, + { + "epoch": 0.12286147237891798, + "grad_norm": 0.17607910931110382, + "learning_rate": 4.9920463906061274e-05, + "loss": 3.2997, + "step": 3983 + }, + { + "epoch": 0.1228923188444914, + "grad_norm": 0.17626075446605682, + "learning_rate": 4.9920249203174945e-05, + "loss": 3.2646, + "step": 3984 + }, + { + "epoch": 0.12292316531006482, + "grad_norm": 0.33467715978622437, + "learning_rate": 4.99200342113533e-05, + "loss": 3.293, + "step": 3985 + }, + { + "epoch": 0.12295401177563824, + "grad_norm": 0.16601401567459106, + "learning_rate": 4.9919818930598824e-05, + "loss": 3.2872, + "step": 3986 + }, + { + "epoch": 0.12298485824121165, + "grad_norm": 0.18758869171142578, + "learning_rate": 4.9919603360914034e-05, + "loss": 3.2863, + "step": 3987 + }, + { + "epoch": 0.12301570470678506, + "grad_norm": 0.16710814833641052, + "learning_rate": 4.991938750230142e-05, + "loss": 3.2725, + "step": 3988 + }, + { + "epoch": 0.12304655117235848, + "grad_norm": 0.18392477929592133, + "learning_rate": 4.991917135476347e-05, + "loss": 3.2692, + "step": 3989 + }, + { + "epoch": 0.12307739763793189, + "grad_norm": 0.1696988046169281, + "learning_rate": 4.9918954918302715e-05, + "loss": 3.2866, + "step": 3990 + }, + { + "epoch": 0.12310824410350532, + "grad_norm": 0.19578687846660614, + "learning_rate": 4.9918738192921645e-05, + "loss": 3.3068, + "step": 3991 + }, + { + "epoch": 0.12313909056907874, + "grad_norm": 0.17722634971141815, + "learning_rate": 4.9918521178622786e-05, + "loss": 3.2809, + "step": 3992 + }, + { + "epoch": 0.12316993703465215, + "grad_norm": 0.19373881816864014, + "learning_rate": 4.991830387540865e-05, + "loss": 3.3242, + "step": 3993 + }, + { + "epoch": 0.12320078350022556, + "grad_norm": 0.17650818824768066, + "learning_rate": 4.991808628328174e-05, + "loss": 3.2689, + "step": 3994 + }, + { + "epoch": 0.12323162996579898, + "grad_norm": 0.19353678822517395, + "learning_rate": 4.991786840224461e-05, + "loss": 3.2835, + "step": 3995 + }, + { + "epoch": 0.12326247643137239, + "grad_norm": 0.17987975478172302, + "learning_rate": 4.9917650232299756e-05, + "loss": 3.309, + "step": 3996 + }, + { + "epoch": 0.12329332289694582, + "grad_norm": 0.19691701233386993, + "learning_rate": 4.991743177344973e-05, + "loss": 3.2611, + "step": 3997 + }, + { + "epoch": 0.12332416936251923, + "grad_norm": 0.18467189371585846, + "learning_rate": 4.9917213025697054e-05, + "loss": 3.2651, + "step": 3998 + }, + { + "epoch": 0.12335501582809265, + "grad_norm": 0.1801314502954483, + "learning_rate": 4.991699398904427e-05, + "loss": 3.2812, + "step": 3999 + }, + { + "epoch": 0.12338586229366606, + "grad_norm": 0.1853683739900589, + "learning_rate": 4.99167746634939e-05, + "loss": 3.2851, + "step": 4000 + }, + { + "epoch": 0.12341670875923948, + "grad_norm": 0.17283210158348083, + "learning_rate": 4.991655504904852e-05, + "loss": 3.2951, + "step": 4001 + }, + { + "epoch": 0.12344755522481289, + "grad_norm": 0.18331453204154968, + "learning_rate": 4.9916335145710644e-05, + "loss": 3.2828, + "step": 4002 + }, + { + "epoch": 0.12347840169038632, + "grad_norm": 0.17027145624160767, + "learning_rate": 4.991611495348284e-05, + "loss": 3.2934, + "step": 4003 + }, + { + "epoch": 0.12350924815595973, + "grad_norm": 0.18196551501750946, + "learning_rate": 4.991589447236767e-05, + "loss": 3.2789, + "step": 4004 + }, + { + "epoch": 0.12354009462153315, + "grad_norm": 0.1716972142457962, + "learning_rate": 4.991567370236765e-05, + "loss": 3.3026, + "step": 4005 + }, + { + "epoch": 0.12357094108710656, + "grad_norm": 0.17714859545230865, + "learning_rate": 4.991545264348538e-05, + "loss": 3.2954, + "step": 4006 + }, + { + "epoch": 0.12360178755267998, + "grad_norm": 0.17289084196090698, + "learning_rate": 4.991523129572341e-05, + "loss": 3.2724, + "step": 4007 + }, + { + "epoch": 0.12363263401825339, + "grad_norm": 0.1688702255487442, + "learning_rate": 4.99150096590843e-05, + "loss": 3.2776, + "step": 4008 + }, + { + "epoch": 0.12366348048382682, + "grad_norm": 0.1839548647403717, + "learning_rate": 4.991478773357063e-05, + "loss": 3.3339, + "step": 4009 + }, + { + "epoch": 0.12369432694940023, + "grad_norm": 0.16529589891433716, + "learning_rate": 4.9914565519184965e-05, + "loss": 3.3205, + "step": 4010 + }, + { + "epoch": 0.12372517341497365, + "grad_norm": 0.1867317259311676, + "learning_rate": 4.9914343015929886e-05, + "loss": 3.2701, + "step": 4011 + }, + { + "epoch": 0.12375601988054706, + "grad_norm": 0.16998717188835144, + "learning_rate": 4.991412022380797e-05, + "loss": 3.2981, + "step": 4012 + }, + { + "epoch": 0.12378686634612047, + "grad_norm": 0.18607886135578156, + "learning_rate": 4.9913897142821805e-05, + "loss": 3.3009, + "step": 4013 + }, + { + "epoch": 0.12381771281169389, + "grad_norm": 0.17489704489707947, + "learning_rate": 4.9913673772973966e-05, + "loss": 3.3278, + "step": 4014 + }, + { + "epoch": 0.12384855927726732, + "grad_norm": 0.17095430195331573, + "learning_rate": 4.991345011426706e-05, + "loss": 3.3034, + "step": 4015 + }, + { + "epoch": 0.12387940574284073, + "grad_norm": 0.19255024194717407, + "learning_rate": 4.991322616670367e-05, + "loss": 3.2843, + "step": 4016 + }, + { + "epoch": 0.12391025220841415, + "grad_norm": 0.16227734088897705, + "learning_rate": 4.99130019302864e-05, + "loss": 3.2643, + "step": 4017 + }, + { + "epoch": 0.12394109867398756, + "grad_norm": 0.2024138867855072, + "learning_rate": 4.991277740501783e-05, + "loss": 3.3174, + "step": 4018 + }, + { + "epoch": 0.12397194513956097, + "grad_norm": 0.17042967677116394, + "learning_rate": 4.991255259090059e-05, + "loss": 3.2791, + "step": 4019 + }, + { + "epoch": 0.12400279160513439, + "grad_norm": 0.20447425544261932, + "learning_rate": 4.991232748793726e-05, + "loss": 3.2773, + "step": 4020 + }, + { + "epoch": 0.12403363807070782, + "grad_norm": 0.1697845607995987, + "learning_rate": 4.9912102096130465e-05, + "loss": 3.2972, + "step": 4021 + }, + { + "epoch": 0.12406448453628123, + "grad_norm": 0.17559324204921722, + "learning_rate": 4.991187641548282e-05, + "loss": 3.2679, + "step": 4022 + }, + { + "epoch": 0.12409533100185464, + "grad_norm": 0.17251412570476532, + "learning_rate": 4.991165044599694e-05, + "loss": 3.3024, + "step": 4023 + }, + { + "epoch": 0.12412617746742806, + "grad_norm": 0.16030369699001312, + "learning_rate": 4.991142418767544e-05, + "loss": 3.2703, + "step": 4024 + }, + { + "epoch": 0.12415702393300147, + "grad_norm": 0.1904028058052063, + "learning_rate": 4.991119764052096e-05, + "loss": 3.325, + "step": 4025 + }, + { + "epoch": 0.12418787039857489, + "grad_norm": 0.1719515323638916, + "learning_rate": 4.991097080453609e-05, + "loss": 3.2903, + "step": 4026 + }, + { + "epoch": 0.12421871686414832, + "grad_norm": 0.16558638215065002, + "learning_rate": 4.9910743679723504e-05, + "loss": 3.281, + "step": 4027 + }, + { + "epoch": 0.12424956332972173, + "grad_norm": 0.18427683413028717, + "learning_rate": 4.9910516266085805e-05, + "loss": 3.3319, + "step": 4028 + }, + { + "epoch": 0.12428040979529514, + "grad_norm": 0.17138423025608063, + "learning_rate": 4.991028856362564e-05, + "loss": 3.2416, + "step": 4029 + }, + { + "epoch": 0.12431125626086856, + "grad_norm": 0.17196306586265564, + "learning_rate": 4.991006057234565e-05, + "loss": 3.3106, + "step": 4030 + }, + { + "epoch": 0.12434210272644197, + "grad_norm": 0.17580083012580872, + "learning_rate": 4.990983229224847e-05, + "loss": 3.2775, + "step": 4031 + }, + { + "epoch": 0.12437294919201539, + "grad_norm": 0.18424974381923676, + "learning_rate": 4.9909603723336764e-05, + "loss": 3.2525, + "step": 4032 + }, + { + "epoch": 0.12440379565758881, + "grad_norm": 0.17311161756515503, + "learning_rate": 4.9909374865613166e-05, + "loss": 3.3183, + "step": 4033 + }, + { + "epoch": 0.12443464212316223, + "grad_norm": 0.1855415552854538, + "learning_rate": 4.9909145719080336e-05, + "loss": 3.2725, + "step": 4034 + }, + { + "epoch": 0.12446548858873564, + "grad_norm": 0.19908744096755981, + "learning_rate": 4.990891628374093e-05, + "loss": 3.3101, + "step": 4035 + }, + { + "epoch": 0.12449633505430906, + "grad_norm": 0.16487163305282593, + "learning_rate": 4.990868655959761e-05, + "loss": 3.2975, + "step": 4036 + }, + { + "epoch": 0.12452718151988247, + "grad_norm": 0.19639359414577484, + "learning_rate": 4.990845654665304e-05, + "loss": 3.2828, + "step": 4037 + }, + { + "epoch": 0.12455802798545589, + "grad_norm": 0.17484693229198456, + "learning_rate": 4.9908226244909875e-05, + "loss": 3.2908, + "step": 4038 + }, + { + "epoch": 0.12458887445102931, + "grad_norm": 0.1856653243303299, + "learning_rate": 4.99079956543708e-05, + "loss": 3.2531, + "step": 4039 + }, + { + "epoch": 0.12461972091660273, + "grad_norm": 0.18174414336681366, + "learning_rate": 4.990776477503849e-05, + "loss": 3.2774, + "step": 4040 + }, + { + "epoch": 0.12465056738217614, + "grad_norm": 0.17337164282798767, + "learning_rate": 4.990753360691561e-05, + "loss": 3.2645, + "step": 4041 + }, + { + "epoch": 0.12468141384774956, + "grad_norm": 0.1692575365304947, + "learning_rate": 4.990730215000484e-05, + "loss": 3.2866, + "step": 4042 + }, + { + "epoch": 0.12471226031332297, + "grad_norm": 0.17427895963191986, + "learning_rate": 4.9907070404308877e-05, + "loss": 3.2798, + "step": 4043 + }, + { + "epoch": 0.12474310677889638, + "grad_norm": 0.16771353781223297, + "learning_rate": 4.99068383698304e-05, + "loss": 3.279, + "step": 4044 + }, + { + "epoch": 0.12477395324446981, + "grad_norm": 0.1713591068983078, + "learning_rate": 4.990660604657209e-05, + "loss": 3.2717, + "step": 4045 + }, + { + "epoch": 0.12480479971004323, + "grad_norm": 0.17350855469703674, + "learning_rate": 4.990637343453666e-05, + "loss": 3.2796, + "step": 4046 + }, + { + "epoch": 0.12483564617561664, + "grad_norm": 0.17005522549152374, + "learning_rate": 4.990614053372679e-05, + "loss": 3.3026, + "step": 4047 + }, + { + "epoch": 0.12486649264119005, + "grad_norm": 0.16635553538799286, + "learning_rate": 4.9905907344145196e-05, + "loss": 3.3106, + "step": 4048 + }, + { + "epoch": 0.12489733910676347, + "grad_norm": 0.17471268773078918, + "learning_rate": 4.990567386579457e-05, + "loss": 3.2705, + "step": 4049 + }, + { + "epoch": 0.12492818557233688, + "grad_norm": 0.1652374267578125, + "learning_rate": 4.990544009867761e-05, + "loss": 3.2637, + "step": 4050 + }, + { + "epoch": 0.12495903203791031, + "grad_norm": 0.18619538843631744, + "learning_rate": 4.990520604279705e-05, + "loss": 3.3135, + "step": 4051 + }, + { + "epoch": 0.12498987850348373, + "grad_norm": 0.15650416910648346, + "learning_rate": 4.9904971698155595e-05, + "loss": 3.2885, + "step": 4052 + }, + { + "epoch": 0.12502072496905714, + "grad_norm": 0.18990401923656464, + "learning_rate": 4.990473706475596e-05, + "loss": 3.2887, + "step": 4053 + }, + { + "epoch": 0.12505157143463055, + "grad_norm": 0.16529621183872223, + "learning_rate": 4.990450214260086e-05, + "loss": 3.2942, + "step": 4054 + }, + { + "epoch": 0.12508241790020397, + "grad_norm": 0.1724453866481781, + "learning_rate": 4.990426693169302e-05, + "loss": 3.3132, + "step": 4055 + }, + { + "epoch": 0.12511326436577738, + "grad_norm": 0.17204904556274414, + "learning_rate": 4.990403143203517e-05, + "loss": 3.334, + "step": 4056 + }, + { + "epoch": 0.1251441108313508, + "grad_norm": 0.1654684990644455, + "learning_rate": 4.990379564363005e-05, + "loss": 3.3058, + "step": 4057 + }, + { + "epoch": 0.1251749572969242, + "grad_norm": 0.16814616322517395, + "learning_rate": 4.990355956648038e-05, + "loss": 3.265, + "step": 4058 + }, + { + "epoch": 0.12520580376249762, + "grad_norm": 0.17028898000717163, + "learning_rate": 4.9903323200588905e-05, + "loss": 3.2751, + "step": 4059 + }, + { + "epoch": 0.12523665022807107, + "grad_norm": 0.17380426824092865, + "learning_rate": 4.9903086545958354e-05, + "loss": 3.2459, + "step": 4060 + }, + { + "epoch": 0.12526749669364448, + "grad_norm": 0.17502833902835846, + "learning_rate": 4.9902849602591483e-05, + "loss": 3.2542, + "step": 4061 + }, + { + "epoch": 0.1252983431592179, + "grad_norm": 0.18191954493522644, + "learning_rate": 4.9902612370491044e-05, + "loss": 3.275, + "step": 4062 + }, + { + "epoch": 0.1253291896247913, + "grad_norm": 0.1842563897371292, + "learning_rate": 4.990237484965977e-05, + "loss": 3.2723, + "step": 4063 + }, + { + "epoch": 0.12536003609036472, + "grad_norm": 0.17629922926425934, + "learning_rate": 4.990213704010043e-05, + "loss": 3.3117, + "step": 4064 + }, + { + "epoch": 0.12539088255593814, + "grad_norm": 0.17481808364391327, + "learning_rate": 4.990189894181577e-05, + "loss": 3.2418, + "step": 4065 + }, + { + "epoch": 0.12542172902151155, + "grad_norm": 0.19351117312908173, + "learning_rate": 4.990166055480856e-05, + "loss": 3.2966, + "step": 4066 + }, + { + "epoch": 0.12545257548708497, + "grad_norm": 0.1720219999551773, + "learning_rate": 4.9901421879081564e-05, + "loss": 3.2944, + "step": 4067 + }, + { + "epoch": 0.12548342195265838, + "grad_norm": 0.16556911170482635, + "learning_rate": 4.9901182914637533e-05, + "loss": 3.2963, + "step": 4068 + }, + { + "epoch": 0.1255142684182318, + "grad_norm": 0.16982616484165192, + "learning_rate": 4.990094366147926e-05, + "loss": 3.277, + "step": 4069 + }, + { + "epoch": 0.1255451148838052, + "grad_norm": 0.17676085233688354, + "learning_rate": 4.990070411960951e-05, + "loss": 3.2733, + "step": 4070 + }, + { + "epoch": 0.12557596134937862, + "grad_norm": 0.1692405790090561, + "learning_rate": 4.9900464289031056e-05, + "loss": 3.299, + "step": 4071 + }, + { + "epoch": 0.12560680781495207, + "grad_norm": 0.1724795252084732, + "learning_rate": 4.9900224169746676e-05, + "loss": 3.284, + "step": 4072 + }, + { + "epoch": 0.12563765428052548, + "grad_norm": 0.17163844406604767, + "learning_rate": 4.989998376175917e-05, + "loss": 3.267, + "step": 4073 + }, + { + "epoch": 0.1256685007460989, + "grad_norm": 0.16459213197231293, + "learning_rate": 4.989974306507131e-05, + "loss": 3.2865, + "step": 4074 + }, + { + "epoch": 0.1256993472116723, + "grad_norm": 0.1762913316488266, + "learning_rate": 4.9899502079685896e-05, + "loss": 3.2955, + "step": 4075 + }, + { + "epoch": 0.12573019367724572, + "grad_norm": 0.1644686460494995, + "learning_rate": 4.9899260805605713e-05, + "loss": 3.2607, + "step": 4076 + }, + { + "epoch": 0.12576104014281914, + "grad_norm": 0.19255681335926056, + "learning_rate": 4.989901924283357e-05, + "loss": 3.2428, + "step": 4077 + }, + { + "epoch": 0.12579188660839255, + "grad_norm": 0.16723397374153137, + "learning_rate": 4.989877739137225e-05, + "loss": 3.2907, + "step": 4078 + }, + { + "epoch": 0.12582273307396596, + "grad_norm": 0.19441314041614532, + "learning_rate": 4.9898535251224584e-05, + "loss": 3.2942, + "step": 4079 + }, + { + "epoch": 0.12585357953953938, + "grad_norm": 0.16277137398719788, + "learning_rate": 4.9898292822393356e-05, + "loss": 3.2794, + "step": 4080 + }, + { + "epoch": 0.1258844260051128, + "grad_norm": 0.1939517855644226, + "learning_rate": 4.989805010488138e-05, + "loss": 3.2603, + "step": 4081 + }, + { + "epoch": 0.1259152724706862, + "grad_norm": 0.18139518797397614, + "learning_rate": 4.989780709869148e-05, + "loss": 3.3118, + "step": 4082 + }, + { + "epoch": 0.12594611893625962, + "grad_norm": 0.19544975459575653, + "learning_rate": 4.9897563803826474e-05, + "loss": 3.2937, + "step": 4083 + }, + { + "epoch": 0.12597696540183306, + "grad_norm": 0.17794841527938843, + "learning_rate": 4.989732022028917e-05, + "loss": 3.2734, + "step": 4084 + }, + { + "epoch": 0.12600781186740648, + "grad_norm": 0.19395282864570618, + "learning_rate": 4.9897076348082404e-05, + "loss": 3.2964, + "step": 4085 + }, + { + "epoch": 0.1260386583329799, + "grad_norm": 0.18060192465782166, + "learning_rate": 4.9896832187209e-05, + "loss": 3.298, + "step": 4086 + }, + { + "epoch": 0.1260695047985533, + "grad_norm": 0.17027854919433594, + "learning_rate": 4.989658773767178e-05, + "loss": 3.2584, + "step": 4087 + }, + { + "epoch": 0.12610035126412672, + "grad_norm": 0.17679479718208313, + "learning_rate": 4.989634299947359e-05, + "loss": 3.2319, + "step": 4088 + }, + { + "epoch": 0.12613119772970013, + "grad_norm": 0.16427400708198547, + "learning_rate": 4.989609797261727e-05, + "loss": 3.2772, + "step": 4089 + }, + { + "epoch": 0.12616204419527355, + "grad_norm": 0.17549437284469604, + "learning_rate": 4.989585265710565e-05, + "loss": 3.2692, + "step": 4090 + }, + { + "epoch": 0.12619289066084696, + "grad_norm": 0.1706358939409256, + "learning_rate": 4.9895607052941575e-05, + "loss": 3.2628, + "step": 4091 + }, + { + "epoch": 0.12622373712642038, + "grad_norm": 0.1852363646030426, + "learning_rate": 4.98953611601279e-05, + "loss": 3.2754, + "step": 4092 + }, + { + "epoch": 0.1262545835919938, + "grad_norm": 0.16209323704242706, + "learning_rate": 4.989511497866748e-05, + "loss": 3.2533, + "step": 4093 + }, + { + "epoch": 0.1262854300575672, + "grad_norm": 0.17697864770889282, + "learning_rate": 4.989486850856315e-05, + "loss": 3.3067, + "step": 4094 + }, + { + "epoch": 0.12631627652314062, + "grad_norm": 0.16706211864948273, + "learning_rate": 4.989462174981778e-05, + "loss": 3.2539, + "step": 4095 + }, + { + "epoch": 0.12634712298871406, + "grad_norm": 0.1829005777835846, + "learning_rate": 4.989437470243424e-05, + "loss": 3.3106, + "step": 4096 + }, + { + "epoch": 0.12637796945428748, + "grad_norm": 0.1800529807806015, + "learning_rate": 4.989412736641538e-05, + "loss": 3.2701, + "step": 4097 + }, + { + "epoch": 0.1264088159198609, + "grad_norm": 0.32103338837623596, + "learning_rate": 4.9893879741764075e-05, + "loss": 3.2986, + "step": 4098 + }, + { + "epoch": 0.1264396623854343, + "grad_norm": 0.18233367800712585, + "learning_rate": 4.9893631828483194e-05, + "loss": 3.2755, + "step": 4099 + }, + { + "epoch": 0.12647050885100772, + "grad_norm": 0.16861025989055634, + "learning_rate": 4.9893383626575614e-05, + "loss": 3.3075, + "step": 4100 + }, + { + "epoch": 0.12650135531658113, + "grad_norm": 0.18712987005710602, + "learning_rate": 4.98931351360442e-05, + "loss": 3.2771, + "step": 4101 + }, + { + "epoch": 0.12653220178215455, + "grad_norm": 0.16870272159576416, + "learning_rate": 4.989288635689186e-05, + "loss": 3.2926, + "step": 4102 + }, + { + "epoch": 0.12656304824772796, + "grad_norm": 0.18091551959514618, + "learning_rate": 4.989263728912145e-05, + "loss": 3.2809, + "step": 4103 + }, + { + "epoch": 0.12659389471330137, + "grad_norm": 0.17273981869220734, + "learning_rate": 4.9892387932735864e-05, + "loss": 3.3196, + "step": 4104 + }, + { + "epoch": 0.1266247411788748, + "grad_norm": 0.1798972636461258, + "learning_rate": 4.989213828773801e-05, + "loss": 3.2798, + "step": 4105 + }, + { + "epoch": 0.1266555876444482, + "grad_norm": 0.1793389469385147, + "learning_rate": 4.989188835413076e-05, + "loss": 3.301, + "step": 4106 + }, + { + "epoch": 0.12668643411002162, + "grad_norm": 0.16423439979553223, + "learning_rate": 4.989163813191703e-05, + "loss": 3.2961, + "step": 4107 + }, + { + "epoch": 0.12671728057559506, + "grad_norm": 0.18110865354537964, + "learning_rate": 4.989138762109971e-05, + "loss": 3.2417, + "step": 4108 + }, + { + "epoch": 0.12674812704116847, + "grad_norm": 0.16640253365039825, + "learning_rate": 4.989113682168172e-05, + "loss": 3.2302, + "step": 4109 + }, + { + "epoch": 0.1267789735067419, + "grad_norm": 0.17187567055225372, + "learning_rate": 4.9890885733665935e-05, + "loss": 3.3053, + "step": 4110 + }, + { + "epoch": 0.1268098199723153, + "grad_norm": 0.17063114047050476, + "learning_rate": 4.989063435705531e-05, + "loss": 3.277, + "step": 4111 + }, + { + "epoch": 0.12684066643788872, + "grad_norm": 0.1748674362897873, + "learning_rate": 4.989038269185272e-05, + "loss": 3.2821, + "step": 4112 + }, + { + "epoch": 0.12687151290346213, + "grad_norm": 0.17032447457313538, + "learning_rate": 4.9890130738061116e-05, + "loss": 3.3039, + "step": 4113 + }, + { + "epoch": 0.12690235936903554, + "grad_norm": 0.17678290605545044, + "learning_rate": 4.988987849568339e-05, + "loss": 3.3015, + "step": 4114 + }, + { + "epoch": 0.12693320583460896, + "grad_norm": 0.175283282995224, + "learning_rate": 4.988962596472249e-05, + "loss": 3.28, + "step": 4115 + }, + { + "epoch": 0.12696405230018237, + "grad_norm": 0.18609757721424103, + "learning_rate": 4.988937314518132e-05, + "loss": 3.2719, + "step": 4116 + }, + { + "epoch": 0.1269948987657558, + "grad_norm": 0.1640259474515915, + "learning_rate": 4.988912003706284e-05, + "loss": 3.28, + "step": 4117 + }, + { + "epoch": 0.1270257452313292, + "grad_norm": 0.17468442022800446, + "learning_rate": 4.9888866640369955e-05, + "loss": 3.2924, + "step": 4118 + }, + { + "epoch": 0.12705659169690262, + "grad_norm": 0.17437702417373657, + "learning_rate": 4.988861295510563e-05, + "loss": 3.28, + "step": 4119 + }, + { + "epoch": 0.12708743816247606, + "grad_norm": 0.17334939539432526, + "learning_rate": 4.988835898127279e-05, + "loss": 3.2686, + "step": 4120 + }, + { + "epoch": 0.12711828462804947, + "grad_norm": 0.18309688568115234, + "learning_rate": 4.988810471887439e-05, + "loss": 3.2749, + "step": 4121 + }, + { + "epoch": 0.12714913109362289, + "grad_norm": 0.17171594500541687, + "learning_rate": 4.988785016791336e-05, + "loss": 3.278, + "step": 4122 + }, + { + "epoch": 0.1271799775591963, + "grad_norm": 0.17339842021465302, + "learning_rate": 4.988759532839267e-05, + "loss": 3.2119, + "step": 4123 + }, + { + "epoch": 0.12721082402476971, + "grad_norm": 0.16137386858463287, + "learning_rate": 4.988734020031527e-05, + "loss": 3.2849, + "step": 4124 + }, + { + "epoch": 0.12724167049034313, + "grad_norm": 0.17452572286128998, + "learning_rate": 4.9887084783684114e-05, + "loss": 3.2836, + "step": 4125 + }, + { + "epoch": 0.12727251695591654, + "grad_norm": 0.16111211478710175, + "learning_rate": 4.9886829078502164e-05, + "loss": 3.249, + "step": 4126 + }, + { + "epoch": 0.12730336342148996, + "grad_norm": 0.16924528777599335, + "learning_rate": 4.988657308477238e-05, + "loss": 3.3297, + "step": 4127 + }, + { + "epoch": 0.12733420988706337, + "grad_norm": 0.18263304233551025, + "learning_rate": 4.9886316802497745e-05, + "loss": 3.2454, + "step": 4128 + }, + { + "epoch": 0.12736505635263679, + "grad_norm": 0.1705182045698166, + "learning_rate": 4.988606023168122e-05, + "loss": 3.2612, + "step": 4129 + }, + { + "epoch": 0.1273959028182102, + "grad_norm": 0.17444615066051483, + "learning_rate": 4.988580337232578e-05, + "loss": 3.2713, + "step": 4130 + }, + { + "epoch": 0.12742674928378364, + "grad_norm": 0.1779797077178955, + "learning_rate": 4.9885546224434405e-05, + "loss": 3.3049, + "step": 4131 + }, + { + "epoch": 0.12745759574935706, + "grad_norm": 0.180414617061615, + "learning_rate": 4.9885288788010074e-05, + "loss": 3.2884, + "step": 4132 + }, + { + "epoch": 0.12748844221493047, + "grad_norm": 0.18817654252052307, + "learning_rate": 4.9885031063055774e-05, + "loss": 3.2472, + "step": 4133 + }, + { + "epoch": 0.12751928868050388, + "grad_norm": 0.17532357573509216, + "learning_rate": 4.9884773049574494e-05, + "loss": 3.301, + "step": 4134 + }, + { + "epoch": 0.1275501351460773, + "grad_norm": 0.18152430653572083, + "learning_rate": 4.988451474756923e-05, + "loss": 3.3469, + "step": 4135 + }, + { + "epoch": 0.1275809816116507, + "grad_norm": 0.16888050734996796, + "learning_rate": 4.9884256157042964e-05, + "loss": 3.2578, + "step": 4136 + }, + { + "epoch": 0.12761182807722413, + "grad_norm": 0.17627964913845062, + "learning_rate": 4.9883997277998704e-05, + "loss": 3.2685, + "step": 4137 + }, + { + "epoch": 0.12764267454279754, + "grad_norm": 0.17441615462303162, + "learning_rate": 4.988373811043945e-05, + "loss": 3.286, + "step": 4138 + }, + { + "epoch": 0.12767352100837095, + "grad_norm": 0.17555947601795197, + "learning_rate": 4.988347865436821e-05, + "loss": 3.2697, + "step": 4139 + }, + { + "epoch": 0.12770436747394437, + "grad_norm": 0.16684241592884064, + "learning_rate": 4.988321890978798e-05, + "loss": 3.2399, + "step": 4140 + }, + { + "epoch": 0.12773521393951778, + "grad_norm": 0.17062078416347504, + "learning_rate": 4.988295887670178e-05, + "loss": 3.2767, + "step": 4141 + }, + { + "epoch": 0.1277660604050912, + "grad_norm": 0.1697029024362564, + "learning_rate": 4.988269855511263e-05, + "loss": 3.2924, + "step": 4142 + }, + { + "epoch": 0.12779690687066464, + "grad_norm": 0.17384955286979675, + "learning_rate": 4.988243794502354e-05, + "loss": 3.2705, + "step": 4143 + }, + { + "epoch": 0.12782775333623805, + "grad_norm": 0.18560023605823517, + "learning_rate": 4.9882177046437536e-05, + "loss": 3.2822, + "step": 4144 + }, + { + "epoch": 0.12785859980181147, + "grad_norm": 0.16998769342899323, + "learning_rate": 4.9881915859357645e-05, + "loss": 3.2803, + "step": 4145 + }, + { + "epoch": 0.12788944626738488, + "grad_norm": 0.22233806550502777, + "learning_rate": 4.988165438378688e-05, + "loss": 3.3078, + "step": 4146 + }, + { + "epoch": 0.1279202927329583, + "grad_norm": 0.17226916551589966, + "learning_rate": 4.988139261972829e-05, + "loss": 3.2741, + "step": 4147 + }, + { + "epoch": 0.1279511391985317, + "grad_norm": 0.1786569505929947, + "learning_rate": 4.988113056718492e-05, + "loss": 3.261, + "step": 4148 + }, + { + "epoch": 0.12798198566410512, + "grad_norm": 0.16944526135921478, + "learning_rate": 4.988086822615978e-05, + "loss": 3.2544, + "step": 4149 + }, + { + "epoch": 0.12801283212967854, + "grad_norm": 0.1635713130235672, + "learning_rate": 4.988060559665592e-05, + "loss": 3.2743, + "step": 4150 + }, + { + "epoch": 0.12804367859525195, + "grad_norm": 1.523328185081482, + "learning_rate": 4.988034267867639e-05, + "loss": 3.2906, + "step": 4151 + }, + { + "epoch": 0.12807452506082537, + "grad_norm": 0.637599527835846, + "learning_rate": 4.988007947222425e-05, + "loss": 3.3131, + "step": 4152 + }, + { + "epoch": 0.12810537152639878, + "grad_norm": 0.49915942549705505, + "learning_rate": 4.987981597730253e-05, + "loss": 3.2515, + "step": 4153 + }, + { + "epoch": 0.1281362179919722, + "grad_norm": 0.5478315353393555, + "learning_rate": 4.98795521939143e-05, + "loss": 3.276, + "step": 4154 + }, + { + "epoch": 0.12816706445754564, + "grad_norm": 0.32671916484832764, + "learning_rate": 4.987928812206261e-05, + "loss": 3.3214, + "step": 4155 + }, + { + "epoch": 0.12819791092311905, + "grad_norm": 0.43144550919532776, + "learning_rate": 4.9879023761750523e-05, + "loss": 3.3139, + "step": 4156 + }, + { + "epoch": 0.12822875738869247, + "grad_norm": 0.45757222175598145, + "learning_rate": 4.987875911298111e-05, + "loss": 3.2872, + "step": 4157 + }, + { + "epoch": 0.12825960385426588, + "grad_norm": 0.23079228401184082, + "learning_rate": 4.987849417575743e-05, + "loss": 3.2559, + "step": 4158 + }, + { + "epoch": 0.1282904503198393, + "grad_norm": 0.29676687717437744, + "learning_rate": 4.987822895008256e-05, + "loss": 3.317, + "step": 4159 + }, + { + "epoch": 0.1283212967854127, + "grad_norm": 0.20697267353534698, + "learning_rate": 4.987796343595958e-05, + "loss": 3.2832, + "step": 4160 + }, + { + "epoch": 0.12835214325098612, + "grad_norm": 0.23655235767364502, + "learning_rate": 4.987769763339157e-05, + "loss": 3.2723, + "step": 4161 + }, + { + "epoch": 0.12838298971655954, + "grad_norm": 0.19712191820144653, + "learning_rate": 4.98774315423816e-05, + "loss": 3.2781, + "step": 4162 + }, + { + "epoch": 0.12841383618213295, + "grad_norm": 0.3442574143409729, + "learning_rate": 4.987716516293276e-05, + "loss": 3.2995, + "step": 4163 + }, + { + "epoch": 0.12844468264770637, + "grad_norm": 0.19315965473651886, + "learning_rate": 4.987689849504813e-05, + "loss": 3.2924, + "step": 4164 + }, + { + "epoch": 0.12847552911327978, + "grad_norm": 0.1958157867193222, + "learning_rate": 4.9876631538730824e-05, + "loss": 3.258, + "step": 4165 + }, + { + "epoch": 0.1285063755788532, + "grad_norm": 0.19057568907737732, + "learning_rate": 4.9876364293983923e-05, + "loss": 3.2361, + "step": 4166 + }, + { + "epoch": 0.12853722204442664, + "grad_norm": 0.27216729521751404, + "learning_rate": 4.987609676081052e-05, + "loss": 3.2736, + "step": 4167 + }, + { + "epoch": 0.12856806851000005, + "grad_norm": 0.19809067249298096, + "learning_rate": 4.987582893921373e-05, + "loss": 3.2616, + "step": 4168 + }, + { + "epoch": 0.12859891497557346, + "grad_norm": 0.18784843385219574, + "learning_rate": 4.9875560829196654e-05, + "loss": 3.3027, + "step": 4169 + }, + { + "epoch": 0.12862976144114688, + "grad_norm": 0.20806553959846497, + "learning_rate": 4.987529243076239e-05, + "loss": 3.304, + "step": 4170 + }, + { + "epoch": 0.1286606079067203, + "grad_norm": 0.1772603839635849, + "learning_rate": 4.987502374391407e-05, + "loss": 3.2857, + "step": 4171 + }, + { + "epoch": 0.1286914543722937, + "grad_norm": 0.20258212089538574, + "learning_rate": 4.987475476865479e-05, + "loss": 3.2904, + "step": 4172 + }, + { + "epoch": 0.12872230083786712, + "grad_norm": 0.188033789396286, + "learning_rate": 4.987448550498768e-05, + "loss": 3.3014, + "step": 4173 + }, + { + "epoch": 0.12875314730344053, + "grad_norm": 0.20739367604255676, + "learning_rate": 4.987421595291585e-05, + "loss": 3.2439, + "step": 4174 + }, + { + "epoch": 0.12878399376901395, + "grad_norm": 0.18567243218421936, + "learning_rate": 4.987394611244244e-05, + "loss": 3.2926, + "step": 4175 + }, + { + "epoch": 0.12881484023458736, + "grad_norm": 0.18406109511852264, + "learning_rate": 4.987367598357058e-05, + "loss": 3.2879, + "step": 4176 + }, + { + "epoch": 0.12884568670016078, + "grad_norm": 0.17630530893802643, + "learning_rate": 4.987340556630339e-05, + "loss": 3.2802, + "step": 4177 + }, + { + "epoch": 0.1288765331657342, + "grad_norm": 0.18828067183494568, + "learning_rate": 4.9873134860644e-05, + "loss": 3.2651, + "step": 4178 + }, + { + "epoch": 0.12890737963130763, + "grad_norm": 0.1835431009531021, + "learning_rate": 4.9872863866595566e-05, + "loss": 3.2791, + "step": 4179 + }, + { + "epoch": 0.12893822609688105, + "grad_norm": 0.18762077391147614, + "learning_rate": 4.987259258416122e-05, + "loss": 3.2748, + "step": 4180 + }, + { + "epoch": 0.12896907256245446, + "grad_norm": 0.1843879520893097, + "learning_rate": 4.9872321013344114e-05, + "loss": 3.256, + "step": 4181 + }, + { + "epoch": 0.12899991902802788, + "grad_norm": 0.18683266639709473, + "learning_rate": 4.9872049154147386e-05, + "loss": 3.3049, + "step": 4182 + }, + { + "epoch": 0.1290307654936013, + "grad_norm": 0.16669850051403046, + "learning_rate": 4.98717770065742e-05, + "loss": 3.2695, + "step": 4183 + }, + { + "epoch": 0.1290616119591747, + "grad_norm": 0.17329885065555573, + "learning_rate": 4.98715045706277e-05, + "loss": 3.2749, + "step": 4184 + }, + { + "epoch": 0.12909245842474812, + "grad_norm": 0.19122308492660522, + "learning_rate": 4.9871231846311054e-05, + "loss": 3.281, + "step": 4185 + }, + { + "epoch": 0.12912330489032153, + "grad_norm": 0.16511136293411255, + "learning_rate": 4.987095883362743e-05, + "loss": 3.2805, + "step": 4186 + }, + { + "epoch": 0.12915415135589495, + "grad_norm": 0.18157579004764557, + "learning_rate": 4.9870685532579964e-05, + "loss": 3.2705, + "step": 4187 + }, + { + "epoch": 0.12918499782146836, + "grad_norm": 0.17718128859996796, + "learning_rate": 4.987041194317186e-05, + "loss": 3.2963, + "step": 4188 + }, + { + "epoch": 0.12921584428704178, + "grad_norm": 0.16911713778972626, + "learning_rate": 4.987013806540627e-05, + "loss": 3.2871, + "step": 4189 + }, + { + "epoch": 0.1292466907526152, + "grad_norm": 0.17347672581672668, + "learning_rate": 4.9869863899286376e-05, + "loss": 3.2756, + "step": 4190 + }, + { + "epoch": 0.12927753721818863, + "grad_norm": 0.1664588451385498, + "learning_rate": 4.986958944481535e-05, + "loss": 3.2923, + "step": 4191 + }, + { + "epoch": 0.12930838368376205, + "grad_norm": 0.173716202378273, + "learning_rate": 4.986931470199638e-05, + "loss": 3.2509, + "step": 4192 + }, + { + "epoch": 0.12933923014933546, + "grad_norm": 0.18195740878582, + "learning_rate": 4.986903967083266e-05, + "loss": 3.2735, + "step": 4193 + }, + { + "epoch": 0.12937007661490887, + "grad_norm": 0.1796097308397293, + "learning_rate": 4.986876435132736e-05, + "loss": 3.3056, + "step": 4194 + }, + { + "epoch": 0.1294009230804823, + "grad_norm": 0.17957358062267303, + "learning_rate": 4.986848874348369e-05, + "loss": 3.2873, + "step": 4195 + }, + { + "epoch": 0.1294317695460557, + "grad_norm": 0.17354370653629303, + "learning_rate": 4.986821284730483e-05, + "loss": 3.2794, + "step": 4196 + }, + { + "epoch": 0.12946261601162912, + "grad_norm": 0.1750769466161728, + "learning_rate": 4.9867936662793994e-05, + "loss": 3.2801, + "step": 4197 + }, + { + "epoch": 0.12949346247720253, + "grad_norm": 0.18452559411525726, + "learning_rate": 4.986766018995437e-05, + "loss": 3.2432, + "step": 4198 + }, + { + "epoch": 0.12952430894277595, + "grad_norm": 0.17519842088222504, + "learning_rate": 4.986738342878916e-05, + "loss": 3.2586, + "step": 4199 + }, + { + "epoch": 0.12955515540834936, + "grad_norm": 0.263101190328598, + "learning_rate": 4.9867106379301596e-05, + "loss": 3.267, + "step": 4200 + }, + { + "epoch": 0.12958600187392277, + "grad_norm": 0.17609894275665283, + "learning_rate": 4.986682904149488e-05, + "loss": 3.3045, + "step": 4201 + }, + { + "epoch": 0.1296168483394962, + "grad_norm": 0.17743276059627533, + "learning_rate": 4.986655141537221e-05, + "loss": 3.2693, + "step": 4202 + }, + { + "epoch": 0.12964769480506963, + "grad_norm": 0.18292424082756042, + "learning_rate": 4.986627350093683e-05, + "loss": 3.2828, + "step": 4203 + }, + { + "epoch": 0.12967854127064304, + "grad_norm": 0.1688043624162674, + "learning_rate": 4.986599529819195e-05, + "loss": 3.261, + "step": 4204 + }, + { + "epoch": 0.12970938773621646, + "grad_norm": 0.17756988108158112, + "learning_rate": 4.98657168071408e-05, + "loss": 3.2666, + "step": 4205 + }, + { + "epoch": 0.12974023420178987, + "grad_norm": 0.17362377047538757, + "learning_rate": 4.986543802778661e-05, + "loss": 3.3027, + "step": 4206 + }, + { + "epoch": 0.1297710806673633, + "grad_norm": 0.17600569128990173, + "learning_rate": 4.98651589601326e-05, + "loss": 3.305, + "step": 4207 + }, + { + "epoch": 0.1298019271329367, + "grad_norm": 0.17682971060276031, + "learning_rate": 4.9864879604182015e-05, + "loss": 3.2775, + "step": 4208 + }, + { + "epoch": 0.12983277359851012, + "grad_norm": 0.16834643483161926, + "learning_rate": 4.986459995993808e-05, + "loss": 3.2759, + "step": 4209 + }, + { + "epoch": 0.12986362006408353, + "grad_norm": 0.2668224573135376, + "learning_rate": 4.9864320027404075e-05, + "loss": 3.3019, + "step": 4210 + }, + { + "epoch": 0.12989446652965694, + "grad_norm": 0.17391376197338104, + "learning_rate": 4.986403980658321e-05, + "loss": 3.2851, + "step": 4211 + }, + { + "epoch": 0.12992531299523036, + "grad_norm": 0.17739024758338928, + "learning_rate": 4.986375929747874e-05, + "loss": 3.2844, + "step": 4212 + }, + { + "epoch": 0.12995615946080377, + "grad_norm": 0.1792248785495758, + "learning_rate": 4.986347850009393e-05, + "loss": 3.2844, + "step": 4213 + }, + { + "epoch": 0.12998700592637719, + "grad_norm": 0.16803067922592163, + "learning_rate": 4.986319741443203e-05, + "loss": 3.2645, + "step": 4214 + }, + { + "epoch": 0.13001785239195063, + "grad_norm": 0.18898513913154602, + "learning_rate": 4.986291604049629e-05, + "loss": 3.3061, + "step": 4215 + }, + { + "epoch": 0.13004869885752404, + "grad_norm": 0.1704539656639099, + "learning_rate": 4.986263437828998e-05, + "loss": 3.2913, + "step": 4216 + }, + { + "epoch": 0.13007954532309746, + "grad_norm": 0.18239301443099976, + "learning_rate": 4.9862352427816366e-05, + "loss": 3.2857, + "step": 4217 + }, + { + "epoch": 0.13011039178867087, + "grad_norm": 0.16446149349212646, + "learning_rate": 4.9862070189078716e-05, + "loss": 3.2798, + "step": 4218 + }, + { + "epoch": 0.13014123825424428, + "grad_norm": 0.18155325949192047, + "learning_rate": 4.98617876620803e-05, + "loss": 3.2699, + "step": 4219 + }, + { + "epoch": 0.1301720847198177, + "grad_norm": 0.1649923324584961, + "learning_rate": 4.98615048468244e-05, + "loss": 3.2372, + "step": 4220 + }, + { + "epoch": 0.1302029311853911, + "grad_norm": 0.1817777156829834, + "learning_rate": 4.98612217433143e-05, + "loss": 3.2776, + "step": 4221 + }, + { + "epoch": 0.13023377765096453, + "grad_norm": 0.16668801009655, + "learning_rate": 4.9860938351553265e-05, + "loss": 3.2785, + "step": 4222 + }, + { + "epoch": 0.13026462411653794, + "grad_norm": 0.193259134888649, + "learning_rate": 4.986065467154459e-05, + "loss": 3.2371, + "step": 4223 + }, + { + "epoch": 0.13029547058211136, + "grad_norm": 0.16677862405776978, + "learning_rate": 4.986037070329156e-05, + "loss": 3.2565, + "step": 4224 + }, + { + "epoch": 0.13032631704768477, + "grad_norm": 0.16659867763519287, + "learning_rate": 4.9860086446797475e-05, + "loss": 3.2523, + "step": 4225 + }, + { + "epoch": 0.13035716351325818, + "grad_norm": 0.18651621043682098, + "learning_rate": 4.985980190206563e-05, + "loss": 3.3411, + "step": 4226 + }, + { + "epoch": 0.13038800997883163, + "grad_norm": 0.1779502034187317, + "learning_rate": 4.9859517069099314e-05, + "loss": 3.2946, + "step": 4227 + }, + { + "epoch": 0.13041885644440504, + "grad_norm": 0.18515056371688843, + "learning_rate": 4.9859231947901835e-05, + "loss": 3.2876, + "step": 4228 + }, + { + "epoch": 0.13044970290997845, + "grad_norm": 0.1817583292722702, + "learning_rate": 4.98589465384765e-05, + "loss": 3.3006, + "step": 4229 + }, + { + "epoch": 0.13048054937555187, + "grad_norm": 0.1750667840242386, + "learning_rate": 4.985866084082663e-05, + "loss": 3.312, + "step": 4230 + }, + { + "epoch": 0.13051139584112528, + "grad_norm": 0.17252254486083984, + "learning_rate": 4.9858374854955517e-05, + "loss": 3.2755, + "step": 4231 + }, + { + "epoch": 0.1305422423066987, + "grad_norm": 0.17324164509773254, + "learning_rate": 4.985808858086649e-05, + "loss": 3.2689, + "step": 4232 + }, + { + "epoch": 0.1305730887722721, + "grad_norm": 0.16819295287132263, + "learning_rate": 4.985780201856286e-05, + "loss": 3.2877, + "step": 4233 + }, + { + "epoch": 0.13060393523784553, + "grad_norm": 0.17405398190021515, + "learning_rate": 4.985751516804795e-05, + "loss": 3.269, + "step": 4234 + }, + { + "epoch": 0.13063478170341894, + "grad_norm": 0.17365345358848572, + "learning_rate": 4.98572280293251e-05, + "loss": 3.2881, + "step": 4235 + }, + { + "epoch": 0.13066562816899235, + "grad_norm": 0.18100503087043762, + "learning_rate": 4.985694060239762e-05, + "loss": 3.3324, + "step": 4236 + }, + { + "epoch": 0.13069647463456577, + "grad_norm": 0.16726872324943542, + "learning_rate": 4.985665288726885e-05, + "loss": 3.2883, + "step": 4237 + }, + { + "epoch": 0.13072732110013918, + "grad_norm": 0.18246503174304962, + "learning_rate": 4.985636488394214e-05, + "loss": 3.26, + "step": 4238 + }, + { + "epoch": 0.13075816756571262, + "grad_norm": 0.17548123002052307, + "learning_rate": 4.98560765924208e-05, + "loss": 3.2934, + "step": 4239 + }, + { + "epoch": 0.13078901403128604, + "grad_norm": 0.18885013461112976, + "learning_rate": 4.9855788012708196e-05, + "loss": 3.2656, + "step": 4240 + }, + { + "epoch": 0.13081986049685945, + "grad_norm": 0.18354550004005432, + "learning_rate": 4.985549914480766e-05, + "loss": 3.2782, + "step": 4241 + }, + { + "epoch": 0.13085070696243287, + "grad_norm": 0.18490073084831238, + "learning_rate": 4.985520998872256e-05, + "loss": 3.2994, + "step": 4242 + }, + { + "epoch": 0.13088155342800628, + "grad_norm": 0.1717410385608673, + "learning_rate": 4.985492054445623e-05, + "loss": 3.2688, + "step": 4243 + }, + { + "epoch": 0.1309123998935797, + "grad_norm": 0.18157845735549927, + "learning_rate": 4.985463081201203e-05, + "loss": 3.2589, + "step": 4244 + }, + { + "epoch": 0.1309432463591531, + "grad_norm": 0.17159734666347504, + "learning_rate": 4.985434079139333e-05, + "loss": 3.249, + "step": 4245 + }, + { + "epoch": 0.13097409282472652, + "grad_norm": 0.16669434309005737, + "learning_rate": 4.985405048260348e-05, + "loss": 3.2482, + "step": 4246 + }, + { + "epoch": 0.13100493929029994, + "grad_norm": 0.1822511851787567, + "learning_rate": 4.985375988564585e-05, + "loss": 3.2552, + "step": 4247 + }, + { + "epoch": 0.13103578575587335, + "grad_norm": 0.16265197098255157, + "learning_rate": 4.98534690005238e-05, + "loss": 3.2881, + "step": 4248 + }, + { + "epoch": 0.13106663222144677, + "grad_norm": 0.1762874573469162, + "learning_rate": 4.9853177827240724e-05, + "loss": 3.298, + "step": 4249 + }, + { + "epoch": 0.13109747868702018, + "grad_norm": 0.1659260243177414, + "learning_rate": 4.985288636579998e-05, + "loss": 3.2992, + "step": 4250 + }, + { + "epoch": 0.13112832515259362, + "grad_norm": 0.17966608703136444, + "learning_rate": 4.985259461620495e-05, + "loss": 3.2979, + "step": 4251 + }, + { + "epoch": 0.13115917161816704, + "grad_norm": 0.16822966933250427, + "learning_rate": 4.985230257845903e-05, + "loss": 3.2885, + "step": 4252 + }, + { + "epoch": 0.13119001808374045, + "grad_norm": 0.1743391752243042, + "learning_rate": 4.98520102525656e-05, + "loss": 3.29, + "step": 4253 + }, + { + "epoch": 0.13122086454931386, + "grad_norm": 0.164505735039711, + "learning_rate": 4.985171763852803e-05, + "loss": 3.2801, + "step": 4254 + }, + { + "epoch": 0.13125171101488728, + "grad_norm": 0.1693854033946991, + "learning_rate": 4.9851424736349737e-05, + "loss": 3.2814, + "step": 4255 + }, + { + "epoch": 0.1312825574804607, + "grad_norm": 0.17061007022857666, + "learning_rate": 4.9851131546034103e-05, + "loss": 3.305, + "step": 4256 + }, + { + "epoch": 0.1313134039460341, + "grad_norm": 0.16675443947315216, + "learning_rate": 4.985083806758454e-05, + "loss": 3.2614, + "step": 4257 + }, + { + "epoch": 0.13134425041160752, + "grad_norm": 0.17071682214736938, + "learning_rate": 4.9850544301004434e-05, + "loss": 3.2739, + "step": 4258 + }, + { + "epoch": 0.13137509687718094, + "grad_norm": 0.17598356306552887, + "learning_rate": 4.98502502462972e-05, + "loss": 3.2836, + "step": 4259 + }, + { + "epoch": 0.13140594334275435, + "grad_norm": 0.16707125306129456, + "learning_rate": 4.984995590346626e-05, + "loss": 3.2943, + "step": 4260 + }, + { + "epoch": 0.13143678980832776, + "grad_norm": 0.17140009999275208, + "learning_rate": 4.984966127251501e-05, + "loss": 3.299, + "step": 4261 + }, + { + "epoch": 0.13146763627390118, + "grad_norm": 0.1630203276872635, + "learning_rate": 4.984936635344687e-05, + "loss": 3.273, + "step": 4262 + }, + { + "epoch": 0.13149848273947462, + "grad_norm": 0.17158755660057068, + "learning_rate": 4.9849071146265254e-05, + "loss": 3.2844, + "step": 4263 + }, + { + "epoch": 0.13152932920504803, + "grad_norm": 0.18605051934719086, + "learning_rate": 4.984877565097359e-05, + "loss": 3.2803, + "step": 4264 + }, + { + "epoch": 0.13156017567062145, + "grad_norm": 0.16154755651950836, + "learning_rate": 4.984847986757532e-05, + "loss": 3.2583, + "step": 4265 + }, + { + "epoch": 0.13159102213619486, + "grad_norm": 0.18271642923355103, + "learning_rate": 4.984818379607384e-05, + "loss": 3.254, + "step": 4266 + }, + { + "epoch": 0.13162186860176828, + "grad_norm": 0.17378337681293488, + "learning_rate": 4.984788743647261e-05, + "loss": 3.2799, + "step": 4267 + }, + { + "epoch": 0.1316527150673417, + "grad_norm": 0.17794597148895264, + "learning_rate": 4.984759078877506e-05, + "loss": 3.227, + "step": 4268 + }, + { + "epoch": 0.1316835615329151, + "grad_norm": 0.18702450394630432, + "learning_rate": 4.9847293852984615e-05, + "loss": 3.2884, + "step": 4269 + }, + { + "epoch": 0.13171440799848852, + "grad_norm": 0.18407604098320007, + "learning_rate": 4.9846996629104734e-05, + "loss": 3.2867, + "step": 4270 + }, + { + "epoch": 0.13174525446406193, + "grad_norm": 0.1642162799835205, + "learning_rate": 4.9846699117138865e-05, + "loss": 3.2741, + "step": 4271 + }, + { + "epoch": 0.13177610092963535, + "grad_norm": 0.17656287550926208, + "learning_rate": 4.9846401317090443e-05, + "loss": 3.2642, + "step": 4272 + }, + { + "epoch": 0.13180694739520876, + "grad_norm": 0.1681586056947708, + "learning_rate": 4.984610322896294e-05, + "loss": 3.2906, + "step": 4273 + }, + { + "epoch": 0.13183779386078218, + "grad_norm": 0.1779032051563263, + "learning_rate": 4.9845804852759785e-05, + "loss": 3.2707, + "step": 4274 + }, + { + "epoch": 0.13186864032635562, + "grad_norm": 0.16506646573543549, + "learning_rate": 4.984550618848446e-05, + "loss": 3.3146, + "step": 4275 + }, + { + "epoch": 0.13189948679192903, + "grad_norm": 0.18575721979141235, + "learning_rate": 4.984520723614042e-05, + "loss": 3.2784, + "step": 4276 + }, + { + "epoch": 0.13193033325750245, + "grad_norm": 0.16154596209526062, + "learning_rate": 4.984490799573114e-05, + "loss": 3.2765, + "step": 4277 + }, + { + "epoch": 0.13196117972307586, + "grad_norm": 0.17756785452365875, + "learning_rate": 4.9844608467260076e-05, + "loss": 3.2835, + "step": 4278 + }, + { + "epoch": 0.13199202618864928, + "grad_norm": 0.17285758256912231, + "learning_rate": 4.984430865073071e-05, + "loss": 3.2937, + "step": 4279 + }, + { + "epoch": 0.1320228726542227, + "grad_norm": 0.23359346389770508, + "learning_rate": 4.984400854614651e-05, + "loss": 3.289, + "step": 4280 + }, + { + "epoch": 0.1320537191197961, + "grad_norm": 0.1662876158952713, + "learning_rate": 4.984370815351096e-05, + "loss": 3.2284, + "step": 4281 + }, + { + "epoch": 0.13208456558536952, + "grad_norm": 0.18468192219734192, + "learning_rate": 4.9843407472827544e-05, + "loss": 3.2751, + "step": 4282 + }, + { + "epoch": 0.13211541205094293, + "grad_norm": 0.17644059658050537, + "learning_rate": 4.984310650409975e-05, + "loss": 3.3271, + "step": 4283 + }, + { + "epoch": 0.13214625851651635, + "grad_norm": 0.17096178233623505, + "learning_rate": 4.984280524733107e-05, + "loss": 3.2543, + "step": 4284 + }, + { + "epoch": 0.13217710498208976, + "grad_norm": 0.22940045595169067, + "learning_rate": 4.9842503702524984e-05, + "loss": 3.3129, + "step": 4285 + }, + { + "epoch": 0.13220795144766317, + "grad_norm": 0.18103857338428497, + "learning_rate": 4.9842201869685e-05, + "loss": 3.2468, + "step": 4286 + }, + { + "epoch": 0.13223879791323662, + "grad_norm": 0.1666063666343689, + "learning_rate": 4.984189974881461e-05, + "loss": 3.2935, + "step": 4287 + }, + { + "epoch": 0.13226964437881003, + "grad_norm": 0.17041908204555511, + "learning_rate": 4.984159733991732e-05, + "loss": 3.2555, + "step": 4288 + }, + { + "epoch": 0.13230049084438344, + "grad_norm": 0.16612794995307922, + "learning_rate": 4.984129464299664e-05, + "loss": 3.3007, + "step": 4289 + }, + { + "epoch": 0.13233133730995686, + "grad_norm": 0.1660345494747162, + "learning_rate": 4.984099165805608e-05, + "loss": 3.2807, + "step": 4290 + }, + { + "epoch": 0.13236218377553027, + "grad_norm": 0.16496767103672028, + "learning_rate": 4.9840688385099144e-05, + "loss": 3.2529, + "step": 4291 + }, + { + "epoch": 0.1323930302411037, + "grad_norm": 0.16340798139572144, + "learning_rate": 4.9840384824129354e-05, + "loss": 3.2849, + "step": 4292 + }, + { + "epoch": 0.1324238767066771, + "grad_norm": 0.17519621551036835, + "learning_rate": 4.9840080975150234e-05, + "loss": 3.2583, + "step": 4293 + }, + { + "epoch": 0.13245472317225052, + "grad_norm": 0.16471408307552338, + "learning_rate": 4.9839776838165296e-05, + "loss": 3.3084, + "step": 4294 + }, + { + "epoch": 0.13248556963782393, + "grad_norm": 0.15688876807689667, + "learning_rate": 4.983947241317808e-05, + "loss": 3.2844, + "step": 4295 + }, + { + "epoch": 0.13251641610339734, + "grad_norm": 0.16986924409866333, + "learning_rate": 4.98391677001921e-05, + "loss": 3.2959, + "step": 4296 + }, + { + "epoch": 0.13254726256897076, + "grad_norm": 0.16066347062587738, + "learning_rate": 4.9838862699210896e-05, + "loss": 3.277, + "step": 4297 + }, + { + "epoch": 0.13257810903454417, + "grad_norm": 0.1752752810716629, + "learning_rate": 4.9838557410238016e-05, + "loss": 3.3179, + "step": 4298 + }, + { + "epoch": 0.13260895550011761, + "grad_norm": 0.16812099516391754, + "learning_rate": 4.983825183327698e-05, + "loss": 3.2624, + "step": 4299 + }, + { + "epoch": 0.13263980196569103, + "grad_norm": 0.17000482976436615, + "learning_rate": 4.983794596833135e-05, + "loss": 3.2602, + "step": 4300 + }, + { + "epoch": 0.13267064843126444, + "grad_norm": 0.16576538980007172, + "learning_rate": 4.9837639815404654e-05, + "loss": 3.252, + "step": 4301 + }, + { + "epoch": 0.13270149489683786, + "grad_norm": 0.16904211044311523, + "learning_rate": 4.983733337450045e-05, + "loss": 3.2759, + "step": 4302 + }, + { + "epoch": 0.13273234136241127, + "grad_norm": 0.1622731238603592, + "learning_rate": 4.983702664562229e-05, + "loss": 3.2715, + "step": 4303 + }, + { + "epoch": 0.13276318782798469, + "grad_norm": 0.16308985650539398, + "learning_rate": 4.9836719628773735e-05, + "loss": 3.2775, + "step": 4304 + }, + { + "epoch": 0.1327940342935581, + "grad_norm": 0.16984327137470245, + "learning_rate": 4.983641232395834e-05, + "loss": 3.2569, + "step": 4305 + }, + { + "epoch": 0.13282488075913151, + "grad_norm": 0.16340065002441406, + "learning_rate": 4.983610473117967e-05, + "loss": 3.2853, + "step": 4306 + }, + { + "epoch": 0.13285572722470493, + "grad_norm": 0.16766329109668732, + "learning_rate": 4.983579685044128e-05, + "loss": 3.3019, + "step": 4307 + }, + { + "epoch": 0.13288657369027834, + "grad_norm": 0.16460008919239044, + "learning_rate": 4.983548868174677e-05, + "loss": 3.2709, + "step": 4308 + }, + { + "epoch": 0.13291742015585176, + "grad_norm": 0.16766341030597687, + "learning_rate": 4.983518022509968e-05, + "loss": 3.2898, + "step": 4309 + }, + { + "epoch": 0.13294826662142517, + "grad_norm": 0.16629843413829803, + "learning_rate": 4.9834871480503604e-05, + "loss": 3.2443, + "step": 4310 + }, + { + "epoch": 0.1329791130869986, + "grad_norm": 0.18005692958831787, + "learning_rate": 4.983456244796212e-05, + "loss": 3.2475, + "step": 4311 + }, + { + "epoch": 0.13300995955257203, + "grad_norm": 0.16504539549350739, + "learning_rate": 4.98342531274788e-05, + "loss": 3.2987, + "step": 4312 + }, + { + "epoch": 0.13304080601814544, + "grad_norm": 0.1698792427778244, + "learning_rate": 4.983394351905724e-05, + "loss": 3.2904, + "step": 4313 + }, + { + "epoch": 0.13307165248371886, + "grad_norm": 0.1616796851158142, + "learning_rate": 4.983363362270103e-05, + "loss": 3.2823, + "step": 4314 + }, + { + "epoch": 0.13310249894929227, + "grad_norm": 0.17692942917346954, + "learning_rate": 4.983332343841376e-05, + "loss": 3.2804, + "step": 4315 + }, + { + "epoch": 0.13313334541486568, + "grad_norm": 0.16531874239444733, + "learning_rate": 4.983301296619903e-05, + "loss": 3.2754, + "step": 4316 + }, + { + "epoch": 0.1331641918804391, + "grad_norm": 0.16648058593273163, + "learning_rate": 4.983270220606044e-05, + "loss": 3.2796, + "step": 4317 + }, + { + "epoch": 0.1331950383460125, + "grad_norm": 0.1667051613330841, + "learning_rate": 4.983239115800159e-05, + "loss": 3.2858, + "step": 4318 + }, + { + "epoch": 0.13322588481158593, + "grad_norm": 0.16820621490478516, + "learning_rate": 4.9832079822026084e-05, + "loss": 3.276, + "step": 4319 + }, + { + "epoch": 0.13325673127715934, + "grad_norm": 0.17225679755210876, + "learning_rate": 4.983176819813754e-05, + "loss": 3.2709, + "step": 4320 + }, + { + "epoch": 0.13328757774273275, + "grad_norm": 0.1680578589439392, + "learning_rate": 4.983145628633956e-05, + "loss": 3.274, + "step": 4321 + }, + { + "epoch": 0.13331842420830617, + "grad_norm": 0.163648784160614, + "learning_rate": 4.983114408663577e-05, + "loss": 3.2902, + "step": 4322 + }, + { + "epoch": 0.1333492706738796, + "grad_norm": 0.16607610881328583, + "learning_rate": 4.983083159902978e-05, + "loss": 3.2783, + "step": 4323 + }, + { + "epoch": 0.13338011713945302, + "grad_norm": 0.1628633588552475, + "learning_rate": 4.983051882352523e-05, + "loss": 3.2792, + "step": 4324 + }, + { + "epoch": 0.13341096360502644, + "grad_norm": 0.1624702662229538, + "learning_rate": 4.983020576012572e-05, + "loss": 3.2596, + "step": 4325 + }, + { + "epoch": 0.13344181007059985, + "grad_norm": 0.17309343814849854, + "learning_rate": 4.982989240883491e-05, + "loss": 3.2893, + "step": 4326 + }, + { + "epoch": 0.13347265653617327, + "grad_norm": 0.17247071862220764, + "learning_rate": 4.9829578769656414e-05, + "loss": 3.2871, + "step": 4327 + }, + { + "epoch": 0.13350350300174668, + "grad_norm": 0.17127816379070282, + "learning_rate": 4.9829264842593874e-05, + "loss": 3.2467, + "step": 4328 + }, + { + "epoch": 0.1335343494673201, + "grad_norm": 0.171305313706398, + "learning_rate": 4.982895062765093e-05, + "loss": 3.2743, + "step": 4329 + }, + { + "epoch": 0.1335651959328935, + "grad_norm": 0.1781681925058365, + "learning_rate": 4.982863612483122e-05, + "loss": 3.3133, + "step": 4330 + }, + { + "epoch": 0.13359604239846692, + "grad_norm": 0.16841472685337067, + "learning_rate": 4.9828321334138387e-05, + "loss": 3.2923, + "step": 4331 + }, + { + "epoch": 0.13362688886404034, + "grad_norm": 0.1651521772146225, + "learning_rate": 4.98280062555761e-05, + "loss": 3.2722, + "step": 4332 + }, + { + "epoch": 0.13365773532961375, + "grad_norm": 0.1769537478685379, + "learning_rate": 4.982769088914799e-05, + "loss": 3.2721, + "step": 4333 + }, + { + "epoch": 0.13368858179518717, + "grad_norm": 0.16146880388259888, + "learning_rate": 4.9827375234857735e-05, + "loss": 3.2822, + "step": 4334 + }, + { + "epoch": 0.1337194282607606, + "grad_norm": 0.1690809428691864, + "learning_rate": 4.982705929270898e-05, + "loss": 3.2909, + "step": 4335 + }, + { + "epoch": 0.13375027472633402, + "grad_norm": 0.16275392472743988, + "learning_rate": 4.9826743062705395e-05, + "loss": 3.2039, + "step": 4336 + }, + { + "epoch": 0.13378112119190744, + "grad_norm": 0.16719765961170197, + "learning_rate": 4.9826426544850635e-05, + "loss": 3.2889, + "step": 4337 + }, + { + "epoch": 0.13381196765748085, + "grad_norm": 0.15920498967170715, + "learning_rate": 4.9826109739148385e-05, + "loss": 3.2807, + "step": 4338 + }, + { + "epoch": 0.13384281412305427, + "grad_norm": 0.1666439324617386, + "learning_rate": 4.982579264560231e-05, + "loss": 3.2928, + "step": 4339 + }, + { + "epoch": 0.13387366058862768, + "grad_norm": 0.16663549840450287, + "learning_rate": 4.982547526421609e-05, + "loss": 3.2974, + "step": 4340 + }, + { + "epoch": 0.1339045070542011, + "grad_norm": 0.1606433391571045, + "learning_rate": 4.98251575949934e-05, + "loss": 3.2945, + "step": 4341 + }, + { + "epoch": 0.1339353535197745, + "grad_norm": 0.16060656309127808, + "learning_rate": 4.982483963793793e-05, + "loss": 3.2794, + "step": 4342 + }, + { + "epoch": 0.13396619998534792, + "grad_norm": 0.16248968243598938, + "learning_rate": 4.982452139305335e-05, + "loss": 3.2869, + "step": 4343 + }, + { + "epoch": 0.13399704645092134, + "grad_norm": 0.1607799232006073, + "learning_rate": 4.9824202860343374e-05, + "loss": 3.2606, + "step": 4344 + }, + { + "epoch": 0.13402789291649475, + "grad_norm": 0.15395472943782806, + "learning_rate": 4.982388403981167e-05, + "loss": 3.2913, + "step": 4345 + }, + { + "epoch": 0.13405873938206816, + "grad_norm": 0.16167041659355164, + "learning_rate": 4.982356493146196e-05, + "loss": 3.3088, + "step": 4346 + }, + { + "epoch": 0.1340895858476416, + "grad_norm": 0.15879443287849426, + "learning_rate": 4.982324553529794e-05, + "loss": 3.2663, + "step": 4347 + }, + { + "epoch": 0.13412043231321502, + "grad_norm": 0.16443784534931183, + "learning_rate": 4.9822925851323286e-05, + "loss": 3.3, + "step": 4348 + }, + { + "epoch": 0.13415127877878844, + "grad_norm": 0.18042148649692535, + "learning_rate": 4.982260587954173e-05, + "loss": 3.2312, + "step": 4349 + }, + { + "epoch": 0.13418212524436185, + "grad_norm": 0.1658790111541748, + "learning_rate": 4.9822285619956985e-05, + "loss": 3.2706, + "step": 4350 + }, + { + "epoch": 0.13421297170993526, + "grad_norm": 0.1624440997838974, + "learning_rate": 4.982196507257274e-05, + "loss": 3.2642, + "step": 4351 + }, + { + "epoch": 0.13424381817550868, + "grad_norm": 0.16288474202156067, + "learning_rate": 4.982164423739274e-05, + "loss": 3.2734, + "step": 4352 + }, + { + "epoch": 0.1342746646410821, + "grad_norm": 0.165771022439003, + "learning_rate": 4.982132311442068e-05, + "loss": 3.2904, + "step": 4353 + }, + { + "epoch": 0.1343055111066555, + "grad_norm": 0.16501617431640625, + "learning_rate": 4.9821001703660294e-05, + "loss": 3.2577, + "step": 4354 + }, + { + "epoch": 0.13433635757222892, + "grad_norm": 0.16988520324230194, + "learning_rate": 4.9820680005115314e-05, + "loss": 3.2697, + "step": 4355 + }, + { + "epoch": 0.13436720403780233, + "grad_norm": 0.1675434410572052, + "learning_rate": 4.982035801878947e-05, + "loss": 3.2325, + "step": 4356 + }, + { + "epoch": 0.13439805050337575, + "grad_norm": 0.1761053055524826, + "learning_rate": 4.982003574468649e-05, + "loss": 3.3112, + "step": 4357 + }, + { + "epoch": 0.13442889696894916, + "grad_norm": 0.17468300461769104, + "learning_rate": 4.98197131828101e-05, + "loss": 3.2644, + "step": 4358 + }, + { + "epoch": 0.1344597434345226, + "grad_norm": 0.18501943349838257, + "learning_rate": 4.981939033316406e-05, + "loss": 3.301, + "step": 4359 + }, + { + "epoch": 0.13449058990009602, + "grad_norm": 0.15605033934116364, + "learning_rate": 4.98190671957521e-05, + "loss": 3.2567, + "step": 4360 + }, + { + "epoch": 0.13452143636566943, + "grad_norm": 0.17949777841567993, + "learning_rate": 4.981874377057797e-05, + "loss": 3.277, + "step": 4361 + }, + { + "epoch": 0.13455228283124285, + "grad_norm": 0.1749545782804489, + "learning_rate": 4.981842005764542e-05, + "loss": 3.257, + "step": 4362 + }, + { + "epoch": 0.13458312929681626, + "grad_norm": 0.16551657021045685, + "learning_rate": 4.98180960569582e-05, + "loss": 3.2677, + "step": 4363 + }, + { + "epoch": 0.13461397576238968, + "grad_norm": 0.17575904726982117, + "learning_rate": 4.981777176852008e-05, + "loss": 3.2811, + "step": 4364 + }, + { + "epoch": 0.1346448222279631, + "grad_norm": 0.16760894656181335, + "learning_rate": 4.98174471923348e-05, + "loss": 3.2297, + "step": 4365 + }, + { + "epoch": 0.1346756686935365, + "grad_norm": 0.16807280480861664, + "learning_rate": 4.981712232840614e-05, + "loss": 3.2603, + "step": 4366 + }, + { + "epoch": 0.13470651515910992, + "grad_norm": 0.17074941098690033, + "learning_rate": 4.9816797176737855e-05, + "loss": 3.2643, + "step": 4367 + }, + { + "epoch": 0.13473736162468333, + "grad_norm": 0.1687806248664856, + "learning_rate": 4.9816471737333724e-05, + "loss": 3.3035, + "step": 4368 + }, + { + "epoch": 0.13476820809025675, + "grad_norm": 0.16561339795589447, + "learning_rate": 4.9816146010197516e-05, + "loss": 3.294, + "step": 4369 + }, + { + "epoch": 0.13479905455583016, + "grad_norm": 0.1770816445350647, + "learning_rate": 4.9815819995333004e-05, + "loss": 3.3156, + "step": 4370 + }, + { + "epoch": 0.1348299010214036, + "grad_norm": 0.1739879548549652, + "learning_rate": 4.981549369274398e-05, + "loss": 3.2643, + "step": 4371 + }, + { + "epoch": 0.13486074748697702, + "grad_norm": 0.17250460386276245, + "learning_rate": 4.9815167102434215e-05, + "loss": 3.2877, + "step": 4372 + }, + { + "epoch": 0.13489159395255043, + "grad_norm": 0.17549116909503937, + "learning_rate": 4.981484022440749e-05, + "loss": 3.2571, + "step": 4373 + }, + { + "epoch": 0.13492244041812385, + "grad_norm": 0.1623540222644806, + "learning_rate": 4.9814513058667614e-05, + "loss": 3.2688, + "step": 4374 + }, + { + "epoch": 0.13495328688369726, + "grad_norm": 0.17916224896907806, + "learning_rate": 4.9814185605218363e-05, + "loss": 3.273, + "step": 4375 + }, + { + "epoch": 0.13498413334927067, + "grad_norm": 0.16495463252067566, + "learning_rate": 4.981385786406355e-05, + "loss": 3.2957, + "step": 4376 + }, + { + "epoch": 0.1350149798148441, + "grad_norm": 0.1827060431241989, + "learning_rate": 4.981352983520697e-05, + "loss": 3.2835, + "step": 4377 + }, + { + "epoch": 0.1350458262804175, + "grad_norm": 0.5992274284362793, + "learning_rate": 4.981320151865242e-05, + "loss": 3.2891, + "step": 4378 + }, + { + "epoch": 0.13507667274599092, + "grad_norm": 0.17269723117351532, + "learning_rate": 4.98128729144037e-05, + "loss": 3.2219, + "step": 4379 + }, + { + "epoch": 0.13510751921156433, + "grad_norm": 0.20060989260673523, + "learning_rate": 4.981254402246463e-05, + "loss": 3.2732, + "step": 4380 + }, + { + "epoch": 0.13513836567713775, + "grad_norm": 0.17188921570777893, + "learning_rate": 4.981221484283903e-05, + "loss": 3.3234, + "step": 4381 + }, + { + "epoch": 0.13516921214271116, + "grad_norm": 0.19479280710220337, + "learning_rate": 4.981188537553072e-05, + "loss": 3.2806, + "step": 4382 + }, + { + "epoch": 0.1352000586082846, + "grad_norm": 0.175693079829216, + "learning_rate": 4.981155562054349e-05, + "loss": 3.3021, + "step": 4383 + }, + { + "epoch": 0.13523090507385802, + "grad_norm": 0.18074245750904083, + "learning_rate": 4.981122557788119e-05, + "loss": 3.2818, + "step": 4384 + }, + { + "epoch": 0.13526175153943143, + "grad_norm": 0.18070147931575775, + "learning_rate": 4.981089524754764e-05, + "loss": 3.2918, + "step": 4385 + }, + { + "epoch": 0.13529259800500484, + "grad_norm": 0.17844350636005402, + "learning_rate": 4.981056462954667e-05, + "loss": 3.2211, + "step": 4386 + }, + { + "epoch": 0.13532344447057826, + "grad_norm": 0.17533612251281738, + "learning_rate": 4.981023372388211e-05, + "loss": 3.2965, + "step": 4387 + }, + { + "epoch": 0.13535429093615167, + "grad_norm": 0.19241419434547424, + "learning_rate": 4.98099025305578e-05, + "loss": 3.2913, + "step": 4388 + }, + { + "epoch": 0.1353851374017251, + "grad_norm": 0.16980203986167908, + "learning_rate": 4.9809571049577576e-05, + "loss": 3.2731, + "step": 4389 + }, + { + "epoch": 0.1354159838672985, + "grad_norm": 0.17976103723049164, + "learning_rate": 4.980923928094529e-05, + "loss": 3.2948, + "step": 4390 + }, + { + "epoch": 0.13544683033287191, + "grad_norm": 0.1622900515794754, + "learning_rate": 4.980890722466478e-05, + "loss": 3.2619, + "step": 4391 + }, + { + "epoch": 0.13547767679844533, + "grad_norm": 0.17316129803657532, + "learning_rate": 4.98085748807399e-05, + "loss": 3.2775, + "step": 4392 + }, + { + "epoch": 0.13550852326401874, + "grad_norm": 0.16873133182525635, + "learning_rate": 4.9808242249174505e-05, + "loss": 3.3141, + "step": 4393 + }, + { + "epoch": 0.13553936972959216, + "grad_norm": 0.17536520957946777, + "learning_rate": 4.9807909329972445e-05, + "loss": 3.2448, + "step": 4394 + }, + { + "epoch": 0.1355702161951656, + "grad_norm": 0.16886116564273834, + "learning_rate": 4.980757612313759e-05, + "loss": 3.2589, + "step": 4395 + }, + { + "epoch": 0.135601062660739, + "grad_norm": 0.17351631820201874, + "learning_rate": 4.980724262867379e-05, + "loss": 3.2744, + "step": 4396 + }, + { + "epoch": 0.13563190912631243, + "grad_norm": 0.16863107681274414, + "learning_rate": 4.980690884658492e-05, + "loss": 3.2652, + "step": 4397 + }, + { + "epoch": 0.13566275559188584, + "grad_norm": 0.17854295670986176, + "learning_rate": 4.9806574776874856e-05, + "loss": 3.2903, + "step": 4398 + }, + { + "epoch": 0.13569360205745926, + "grad_norm": 0.17048589885234833, + "learning_rate": 4.9806240419547454e-05, + "loss": 3.2885, + "step": 4399 + }, + { + "epoch": 0.13572444852303267, + "grad_norm": 0.17576617002487183, + "learning_rate": 4.9805905774606606e-05, + "loss": 3.2673, + "step": 4400 + }, + { + "epoch": 0.13575529498860608, + "grad_norm": 0.18355131149291992, + "learning_rate": 4.9805570842056196e-05, + "loss": 3.2774, + "step": 4401 + }, + { + "epoch": 0.1357861414541795, + "grad_norm": 0.17218703031539917, + "learning_rate": 4.9805235621900085e-05, + "loss": 3.2918, + "step": 4402 + }, + { + "epoch": 0.1358169879197529, + "grad_norm": 0.16837652027606964, + "learning_rate": 4.9804900114142184e-05, + "loss": 3.2592, + "step": 4403 + }, + { + "epoch": 0.13584783438532633, + "grad_norm": 0.17215074598789215, + "learning_rate": 4.980456431878636e-05, + "loss": 3.2605, + "step": 4404 + }, + { + "epoch": 0.13587868085089974, + "grad_norm": 0.16618385910987854, + "learning_rate": 4.980422823583653e-05, + "loss": 3.2546, + "step": 4405 + }, + { + "epoch": 0.13590952731647318, + "grad_norm": 0.17143367230892181, + "learning_rate": 4.980389186529657e-05, + "loss": 3.2773, + "step": 4406 + }, + { + "epoch": 0.1359403737820466, + "grad_norm": 0.15916141867637634, + "learning_rate": 4.98035552071704e-05, + "loss": 3.2654, + "step": 4407 + }, + { + "epoch": 0.13597122024762, + "grad_norm": 0.1623084545135498, + "learning_rate": 4.980321826146191e-05, + "loss": 3.2399, + "step": 4408 + }, + { + "epoch": 0.13600206671319343, + "grad_norm": 0.16557151079177856, + "learning_rate": 4.9802881028175e-05, + "loss": 3.2532, + "step": 4409 + }, + { + "epoch": 0.13603291317876684, + "grad_norm": 0.1590295284986496, + "learning_rate": 4.980254350731359e-05, + "loss": 3.2652, + "step": 4410 + }, + { + "epoch": 0.13606375964434025, + "grad_norm": 0.17789655923843384, + "learning_rate": 4.98022056988816e-05, + "loss": 3.2594, + "step": 4411 + }, + { + "epoch": 0.13609460610991367, + "grad_norm": 0.16024303436279297, + "learning_rate": 4.980186760288294e-05, + "loss": 3.2889, + "step": 4412 + }, + { + "epoch": 0.13612545257548708, + "grad_norm": 0.16866715252399445, + "learning_rate": 4.980152921932153e-05, + "loss": 3.2442, + "step": 4413 + }, + { + "epoch": 0.1361562990410605, + "grad_norm": 0.15540017187595367, + "learning_rate": 4.980119054820128e-05, + "loss": 3.2642, + "step": 4414 + }, + { + "epoch": 0.1361871455066339, + "grad_norm": 0.16560377180576324, + "learning_rate": 4.980085158952614e-05, + "loss": 3.312, + "step": 4415 + }, + { + "epoch": 0.13621799197220733, + "grad_norm": 0.16818462312221527, + "learning_rate": 4.980051234330003e-05, + "loss": 3.2848, + "step": 4416 + }, + { + "epoch": 0.13624883843778074, + "grad_norm": 0.16516484320163727, + "learning_rate": 4.980017280952688e-05, + "loss": 3.285, + "step": 4417 + }, + { + "epoch": 0.13627968490335418, + "grad_norm": 0.16259191930294037, + "learning_rate": 4.979983298821063e-05, + "loss": 3.2568, + "step": 4418 + }, + { + "epoch": 0.1363105313689276, + "grad_norm": 0.16681967675685883, + "learning_rate": 4.9799492879355224e-05, + "loss": 3.2433, + "step": 4419 + }, + { + "epoch": 0.136341377834501, + "grad_norm": 0.16149897873401642, + "learning_rate": 4.97991524829646e-05, + "loss": 3.2348, + "step": 4420 + }, + { + "epoch": 0.13637222430007442, + "grad_norm": 0.16826646029949188, + "learning_rate": 4.97988117990427e-05, + "loss": 3.2539, + "step": 4421 + }, + { + "epoch": 0.13640307076564784, + "grad_norm": 0.16792282462120056, + "learning_rate": 4.979847082759348e-05, + "loss": 3.2318, + "step": 4422 + }, + { + "epoch": 0.13643391723122125, + "grad_norm": 0.16046777367591858, + "learning_rate": 4.9798129568620894e-05, + "loss": 3.253, + "step": 4423 + }, + { + "epoch": 0.13646476369679467, + "grad_norm": 0.1759956330060959, + "learning_rate": 4.97977880221289e-05, + "loss": 3.2955, + "step": 4424 + }, + { + "epoch": 0.13649561016236808, + "grad_norm": 0.16305217146873474, + "learning_rate": 4.979744618812145e-05, + "loss": 3.2772, + "step": 4425 + }, + { + "epoch": 0.1365264566279415, + "grad_norm": 0.1663176417350769, + "learning_rate": 4.9797104066602515e-05, + "loss": 3.2805, + "step": 4426 + }, + { + "epoch": 0.1365573030935149, + "grad_norm": 0.18016061186790466, + "learning_rate": 4.979676165757606e-05, + "loss": 3.279, + "step": 4427 + }, + { + "epoch": 0.13658814955908832, + "grad_norm": 0.1640796661376953, + "learning_rate": 4.979641896104606e-05, + "loss": 3.2668, + "step": 4428 + }, + { + "epoch": 0.13661899602466174, + "grad_norm": 0.16875068843364716, + "learning_rate": 4.9796075977016476e-05, + "loss": 3.2775, + "step": 4429 + }, + { + "epoch": 0.13664984249023518, + "grad_norm": 0.15832407772541046, + "learning_rate": 4.979573270549129e-05, + "loss": 3.2633, + "step": 4430 + }, + { + "epoch": 0.1366806889558086, + "grad_norm": 0.16690123081207275, + "learning_rate": 4.979538914647449e-05, + "loss": 3.2523, + "step": 4431 + }, + { + "epoch": 0.136711535421382, + "grad_norm": 0.16531774401664734, + "learning_rate": 4.9795045299970054e-05, + "loss": 3.2549, + "step": 4432 + }, + { + "epoch": 0.13674238188695542, + "grad_norm": 0.15714576840400696, + "learning_rate": 4.979470116598197e-05, + "loss": 3.2721, + "step": 4433 + }, + { + "epoch": 0.13677322835252884, + "grad_norm": 0.1653933972120285, + "learning_rate": 4.979435674451422e-05, + "loss": 3.2844, + "step": 4434 + }, + { + "epoch": 0.13680407481810225, + "grad_norm": 0.16031493246555328, + "learning_rate": 4.97940120355708e-05, + "loss": 3.2734, + "step": 4435 + }, + { + "epoch": 0.13683492128367566, + "grad_norm": 2.497419595718384, + "learning_rate": 4.9793667039155725e-05, + "loss": 3.2924, + "step": 4436 + }, + { + "epoch": 0.13686576774924908, + "grad_norm": 0.17429262399673462, + "learning_rate": 4.979332175527296e-05, + "loss": 3.2828, + "step": 4437 + }, + { + "epoch": 0.1368966142148225, + "grad_norm": 0.1767251044511795, + "learning_rate": 4.979297618392654e-05, + "loss": 3.3019, + "step": 4438 + }, + { + "epoch": 0.1369274606803959, + "grad_norm": 0.17777341604232788, + "learning_rate": 4.979263032512046e-05, + "loss": 3.2593, + "step": 4439 + }, + { + "epoch": 0.13695830714596932, + "grad_norm": 0.17437654733657837, + "learning_rate": 4.979228417885873e-05, + "loss": 3.2788, + "step": 4440 + }, + { + "epoch": 0.13698915361154274, + "grad_norm": 0.18742695450782776, + "learning_rate": 4.9791937745145354e-05, + "loss": 3.2627, + "step": 4441 + }, + { + "epoch": 0.13702000007711618, + "grad_norm": 0.1828894466161728, + "learning_rate": 4.979159102398437e-05, + "loss": 3.2688, + "step": 4442 + }, + { + "epoch": 0.1370508465426896, + "grad_norm": 0.19740375876426697, + "learning_rate": 4.979124401537978e-05, + "loss": 3.3083, + "step": 4443 + }, + { + "epoch": 0.137081693008263, + "grad_norm": 0.18466410040855408, + "learning_rate": 4.979089671933561e-05, + "loss": 3.2889, + "step": 4444 + }, + { + "epoch": 0.13711253947383642, + "grad_norm": 0.1933760941028595, + "learning_rate": 4.9790549135855904e-05, + "loss": 3.2752, + "step": 4445 + }, + { + "epoch": 0.13714338593940983, + "grad_norm": 0.17593181133270264, + "learning_rate": 4.979020126494466e-05, + "loss": 3.2823, + "step": 4446 + }, + { + "epoch": 0.13717423240498325, + "grad_norm": 0.18062882125377655, + "learning_rate": 4.978985310660594e-05, + "loss": 3.2293, + "step": 4447 + }, + { + "epoch": 0.13720507887055666, + "grad_norm": 0.17870059609413147, + "learning_rate": 4.9789504660843775e-05, + "loss": 3.3099, + "step": 4448 + }, + { + "epoch": 0.13723592533613008, + "grad_norm": 0.16935209929943085, + "learning_rate": 4.978915592766219e-05, + "loss": 3.2741, + "step": 4449 + }, + { + "epoch": 0.1372667718017035, + "grad_norm": 0.17899614572525024, + "learning_rate": 4.9788806907065245e-05, + "loss": 3.2939, + "step": 4450 + }, + { + "epoch": 0.1372976182672769, + "grad_norm": 0.1711122989654541, + "learning_rate": 4.978845759905697e-05, + "loss": 3.2725, + "step": 4451 + }, + { + "epoch": 0.13732846473285032, + "grad_norm": 0.17888745665550232, + "learning_rate": 4.978810800364144e-05, + "loss": 3.2776, + "step": 4452 + }, + { + "epoch": 0.13735931119842373, + "grad_norm": 0.17566344141960144, + "learning_rate": 4.978775812082269e-05, + "loss": 3.275, + "step": 4453 + }, + { + "epoch": 0.13739015766399718, + "grad_norm": 0.1728859841823578, + "learning_rate": 4.978740795060477e-05, + "loss": 3.2482, + "step": 4454 + }, + { + "epoch": 0.1374210041295706, + "grad_norm": 0.16505973041057587, + "learning_rate": 4.978705749299176e-05, + "loss": 3.2442, + "step": 4455 + }, + { + "epoch": 0.137451850595144, + "grad_norm": 0.16696438193321228, + "learning_rate": 4.978670674798771e-05, + "loss": 3.2829, + "step": 4456 + }, + { + "epoch": 0.13748269706071742, + "grad_norm": 0.1585579514503479, + "learning_rate": 4.9786355715596695e-05, + "loss": 3.2395, + "step": 4457 + }, + { + "epoch": 0.13751354352629083, + "grad_norm": 0.16876330971717834, + "learning_rate": 4.978600439582277e-05, + "loss": 3.2985, + "step": 4458 + }, + { + "epoch": 0.13754438999186425, + "grad_norm": 0.20089177787303925, + "learning_rate": 4.9785652788670035e-05, + "loss": 3.2722, + "step": 4459 + }, + { + "epoch": 0.13757523645743766, + "grad_norm": 0.1720040738582611, + "learning_rate": 4.978530089414254e-05, + "loss": 3.275, + "step": 4460 + }, + { + "epoch": 0.13760608292301107, + "grad_norm": 0.1661665141582489, + "learning_rate": 4.9784948712244374e-05, + "loss": 3.3047, + "step": 4461 + }, + { + "epoch": 0.1376369293885845, + "grad_norm": 0.17583027482032776, + "learning_rate": 4.978459624297962e-05, + "loss": 3.2593, + "step": 4462 + }, + { + "epoch": 0.1376677758541579, + "grad_norm": 0.170674130320549, + "learning_rate": 4.9784243486352374e-05, + "loss": 3.2895, + "step": 4463 + }, + { + "epoch": 0.13769862231973132, + "grad_norm": 0.16773806512355804, + "learning_rate": 4.978389044236671e-05, + "loss": 3.2911, + "step": 4464 + }, + { + "epoch": 0.13772946878530473, + "grad_norm": 0.16902872920036316, + "learning_rate": 4.978353711102674e-05, + "loss": 3.281, + "step": 4465 + }, + { + "epoch": 0.13776031525087817, + "grad_norm": 0.1711973398923874, + "learning_rate": 4.978318349233654e-05, + "loss": 3.275, + "step": 4466 + }, + { + "epoch": 0.1377911617164516, + "grad_norm": 0.16387507319450378, + "learning_rate": 4.978282958630023e-05, + "loss": 3.2342, + "step": 4467 + }, + { + "epoch": 0.137822008182025, + "grad_norm": 0.17394636571407318, + "learning_rate": 4.978247539292189e-05, + "loss": 3.2899, + "step": 4468 + }, + { + "epoch": 0.13785285464759842, + "grad_norm": 0.16521285474300385, + "learning_rate": 4.9782120912205655e-05, + "loss": 3.2632, + "step": 4469 + }, + { + "epoch": 0.13788370111317183, + "grad_norm": 0.1630503088235855, + "learning_rate": 4.9781766144155614e-05, + "loss": 3.2509, + "step": 4470 + }, + { + "epoch": 0.13791454757874524, + "grad_norm": 0.1824638992547989, + "learning_rate": 4.9781411088775884e-05, + "loss": 3.2335, + "step": 4471 + }, + { + "epoch": 0.13794539404431866, + "grad_norm": 0.1560610979795456, + "learning_rate": 4.978105574607059e-05, + "loss": 3.2776, + "step": 4472 + }, + { + "epoch": 0.13797624050989207, + "grad_norm": 0.17696477472782135, + "learning_rate": 4.978070011604384e-05, + "loss": 3.251, + "step": 4473 + }, + { + "epoch": 0.1380070869754655, + "grad_norm": 0.16657568514347076, + "learning_rate": 4.978034419869977e-05, + "loss": 3.2913, + "step": 4474 + }, + { + "epoch": 0.1380379334410389, + "grad_norm": 0.1708914339542389, + "learning_rate": 4.97799879940425e-05, + "loss": 3.2879, + "step": 4475 + }, + { + "epoch": 0.13806877990661232, + "grad_norm": 0.17629171907901764, + "learning_rate": 4.9779631502076155e-05, + "loss": 3.2853, + "step": 4476 + }, + { + "epoch": 0.13809962637218573, + "grad_norm": 0.16718782484531403, + "learning_rate": 4.977927472280488e-05, + "loss": 3.2672, + "step": 4477 + }, + { + "epoch": 0.13813047283775917, + "grad_norm": 0.18314146995544434, + "learning_rate": 4.97789176562328e-05, + "loss": 3.2754, + "step": 4478 + }, + { + "epoch": 0.13816131930333259, + "grad_norm": 0.16856105625629425, + "learning_rate": 4.9778560302364066e-05, + "loss": 3.2991, + "step": 4479 + }, + { + "epoch": 0.138192165768906, + "grad_norm": 0.15849007666110992, + "learning_rate": 4.9778202661202805e-05, + "loss": 3.2284, + "step": 4480 + }, + { + "epoch": 0.13822301223447941, + "grad_norm": 0.16621273756027222, + "learning_rate": 4.977784473275318e-05, + "loss": 3.2886, + "step": 4481 + }, + { + "epoch": 0.13825385870005283, + "grad_norm": 0.1699475795030594, + "learning_rate": 4.9777486517019334e-05, + "loss": 3.2707, + "step": 4482 + }, + { + "epoch": 0.13828470516562624, + "grad_norm": 0.16660906374454498, + "learning_rate": 4.9777128014005424e-05, + "loss": 3.3074, + "step": 4483 + }, + { + "epoch": 0.13831555163119966, + "grad_norm": 0.16545118391513824, + "learning_rate": 4.97767692237156e-05, + "loss": 3.2905, + "step": 4484 + }, + { + "epoch": 0.13834639809677307, + "grad_norm": 0.17138329148292542, + "learning_rate": 4.977641014615403e-05, + "loss": 3.2675, + "step": 4485 + }, + { + "epoch": 0.13837724456234649, + "grad_norm": 0.17290225625038147, + "learning_rate": 4.977605078132488e-05, + "loss": 3.2693, + "step": 4486 + }, + { + "epoch": 0.1384080910279199, + "grad_norm": 0.1687026470899582, + "learning_rate": 4.977569112923229e-05, + "loss": 3.2659, + "step": 4487 + }, + { + "epoch": 0.1384389374934933, + "grad_norm": 0.17281897366046906, + "learning_rate": 4.977533118988047e-05, + "loss": 3.2606, + "step": 4488 + }, + { + "epoch": 0.13846978395906673, + "grad_norm": 0.15868844091892242, + "learning_rate": 4.977497096327356e-05, + "loss": 3.2573, + "step": 4489 + }, + { + "epoch": 0.13850063042464017, + "grad_norm": 0.1783865988254547, + "learning_rate": 4.9774610449415756e-05, + "loss": 3.269, + "step": 4490 + }, + { + "epoch": 0.13853147689021358, + "grad_norm": 0.16919733583927155, + "learning_rate": 4.9774249648311235e-05, + "loss": 3.2676, + "step": 4491 + }, + { + "epoch": 0.138562323355787, + "grad_norm": 0.22980152070522308, + "learning_rate": 4.9773888559964166e-05, + "loss": 3.2554, + "step": 4492 + }, + { + "epoch": 0.1385931698213604, + "grad_norm": 0.17400731146335602, + "learning_rate": 4.977352718437875e-05, + "loss": 3.2447, + "step": 4493 + }, + { + "epoch": 0.13862401628693383, + "grad_norm": 0.18812967836856842, + "learning_rate": 4.9773165521559186e-05, + "loss": 3.2605, + "step": 4494 + }, + { + "epoch": 0.13865486275250724, + "grad_norm": 0.17188286781311035, + "learning_rate": 4.9772803571509644e-05, + "loss": 3.2822, + "step": 4495 + }, + { + "epoch": 0.13868570921808065, + "grad_norm": 0.17861875891685486, + "learning_rate": 4.977244133423433e-05, + "loss": 3.2773, + "step": 4496 + }, + { + "epoch": 0.13871655568365407, + "grad_norm": 0.16970722377300262, + "learning_rate": 4.977207880973745e-05, + "loss": 3.2645, + "step": 4497 + }, + { + "epoch": 0.13874740214922748, + "grad_norm": 0.16380074620246887, + "learning_rate": 4.9771715998023195e-05, + "loss": 3.2464, + "step": 4498 + }, + { + "epoch": 0.1387782486148009, + "grad_norm": 0.25747865438461304, + "learning_rate": 4.9771352899095785e-05, + "loss": 3.2584, + "step": 4499 + }, + { + "epoch": 0.1388090950803743, + "grad_norm": 0.16673386096954346, + "learning_rate": 4.977098951295942e-05, + "loss": 3.2561, + "step": 4500 + }, + { + "epoch": 0.13883994154594773, + "grad_norm": 0.16514673829078674, + "learning_rate": 4.977062583961832e-05, + "loss": 3.2745, + "step": 4501 + }, + { + "epoch": 0.13887078801152117, + "grad_norm": 0.1799771785736084, + "learning_rate": 4.97702618790767e-05, + "loss": 3.3152, + "step": 4502 + }, + { + "epoch": 0.13890163447709458, + "grad_norm": 0.17618346214294434, + "learning_rate": 4.976989763133878e-05, + "loss": 3.283, + "step": 4503 + }, + { + "epoch": 0.138932480942668, + "grad_norm": 0.1594364047050476, + "learning_rate": 4.9769533096408784e-05, + "loss": 3.2626, + "step": 4504 + }, + { + "epoch": 0.1389633274082414, + "grad_norm": 0.17455527186393738, + "learning_rate": 4.976916827429093e-05, + "loss": 3.245, + "step": 4505 + }, + { + "epoch": 0.13899417387381482, + "grad_norm": 0.17309744656085968, + "learning_rate": 4.976880316498945e-05, + "loss": 3.2576, + "step": 4506 + }, + { + "epoch": 0.13902502033938824, + "grad_norm": 0.1617107093334198, + "learning_rate": 4.976843776850859e-05, + "loss": 3.2922, + "step": 4507 + }, + { + "epoch": 0.13905586680496165, + "grad_norm": 0.16852818429470062, + "learning_rate": 4.976807208485258e-05, + "loss": 3.2407, + "step": 4508 + }, + { + "epoch": 0.13908671327053507, + "grad_norm": 0.1632247269153595, + "learning_rate": 4.9767706114025655e-05, + "loss": 3.2524, + "step": 4509 + }, + { + "epoch": 0.13911755973610848, + "grad_norm": 0.16726276278495789, + "learning_rate": 4.9767339856032056e-05, + "loss": 3.2595, + "step": 4510 + }, + { + "epoch": 0.1391484062016819, + "grad_norm": 0.1767667829990387, + "learning_rate": 4.976697331087604e-05, + "loss": 3.2682, + "step": 4511 + }, + { + "epoch": 0.1391792526672553, + "grad_norm": 0.17374172806739807, + "learning_rate": 4.9766606478561854e-05, + "loss": 3.2843, + "step": 4512 + }, + { + "epoch": 0.13921009913282872, + "grad_norm": 0.18297544121742249, + "learning_rate": 4.976623935909375e-05, + "loss": 3.2556, + "step": 4513 + }, + { + "epoch": 0.13924094559840217, + "grad_norm": 0.16966083645820618, + "learning_rate": 4.976587195247597e-05, + "loss": 3.2891, + "step": 4514 + }, + { + "epoch": 0.13927179206397558, + "grad_norm": 0.4128846228122711, + "learning_rate": 4.9765504258712804e-05, + "loss": 3.2981, + "step": 4515 + }, + { + "epoch": 0.139302638529549, + "grad_norm": 0.16268014907836914, + "learning_rate": 4.976513627780849e-05, + "loss": 3.2877, + "step": 4516 + }, + { + "epoch": 0.1393334849951224, + "grad_norm": 0.17121167480945587, + "learning_rate": 4.976476800976731e-05, + "loss": 3.2819, + "step": 4517 + }, + { + "epoch": 0.13936433146069582, + "grad_norm": 0.16777539253234863, + "learning_rate": 4.9764399454593514e-05, + "loss": 3.2802, + "step": 4518 + }, + { + "epoch": 0.13939517792626924, + "grad_norm": 0.1610599011182785, + "learning_rate": 4.976403061229139e-05, + "loss": 3.2605, + "step": 4519 + }, + { + "epoch": 0.13942602439184265, + "grad_norm": 0.19788581132888794, + "learning_rate": 4.976366148286522e-05, + "loss": 3.2741, + "step": 4520 + }, + { + "epoch": 0.13945687085741607, + "grad_norm": 0.1772819459438324, + "learning_rate": 4.976329206631928e-05, + "loss": 3.277, + "step": 4521 + }, + { + "epoch": 0.13948771732298948, + "grad_norm": 0.1754777431488037, + "learning_rate": 4.976292236265784e-05, + "loss": 3.27, + "step": 4522 + }, + { + "epoch": 0.1395185637885629, + "grad_norm": 0.1729850322008133, + "learning_rate": 4.976255237188519e-05, + "loss": 3.2928, + "step": 4523 + }, + { + "epoch": 0.1395494102541363, + "grad_norm": 0.16343702375888824, + "learning_rate": 4.976218209400564e-05, + "loss": 3.2507, + "step": 4524 + }, + { + "epoch": 0.13958025671970972, + "grad_norm": 0.17813202738761902, + "learning_rate": 4.9761811529023465e-05, + "loss": 3.2698, + "step": 4525 + }, + { + "epoch": 0.13961110318528316, + "grad_norm": 0.17219087481498718, + "learning_rate": 4.976144067694297e-05, + "loss": 3.2787, + "step": 4526 + }, + { + "epoch": 0.13964194965085658, + "grad_norm": 0.17157769203186035, + "learning_rate": 4.976106953776844e-05, + "loss": 3.2625, + "step": 4527 + }, + { + "epoch": 0.13967279611643, + "grad_norm": 0.16731424629688263, + "learning_rate": 4.976069811150419e-05, + "loss": 3.2732, + "step": 4528 + }, + { + "epoch": 0.1397036425820034, + "grad_norm": 0.1907815933227539, + "learning_rate": 4.9760326398154535e-05, + "loss": 3.2857, + "step": 4529 + }, + { + "epoch": 0.13973448904757682, + "grad_norm": 0.18330229818820953, + "learning_rate": 4.9759954397723763e-05, + "loss": 3.29, + "step": 4530 + }, + { + "epoch": 0.13976533551315024, + "grad_norm": 0.1680546998977661, + "learning_rate": 4.9759582110216206e-05, + "loss": 3.28, + "step": 4531 + }, + { + "epoch": 0.13979618197872365, + "grad_norm": 0.17463763058185577, + "learning_rate": 4.9759209535636166e-05, + "loss": 3.2531, + "step": 4532 + }, + { + "epoch": 0.13982702844429706, + "grad_norm": 0.18591313064098358, + "learning_rate": 4.975883667398797e-05, + "loss": 3.2884, + "step": 4533 + }, + { + "epoch": 0.13985787490987048, + "grad_norm": 0.16143400967121124, + "learning_rate": 4.9758463525275956e-05, + "loss": 3.2942, + "step": 4534 + }, + { + "epoch": 0.1398887213754439, + "grad_norm": 0.17338085174560547, + "learning_rate": 4.975809008950442e-05, + "loss": 3.2825, + "step": 4535 + }, + { + "epoch": 0.1399195678410173, + "grad_norm": 0.16267740726470947, + "learning_rate": 4.9757716366677714e-05, + "loss": 3.2664, + "step": 4536 + }, + { + "epoch": 0.13995041430659072, + "grad_norm": 0.16696123778820038, + "learning_rate": 4.975734235680016e-05, + "loss": 3.279, + "step": 4537 + }, + { + "epoch": 0.13998126077216416, + "grad_norm": 0.1633608490228653, + "learning_rate": 4.9756968059876106e-05, + "loss": 3.2922, + "step": 4538 + }, + { + "epoch": 0.14001210723773758, + "grad_norm": 0.16833023726940155, + "learning_rate": 4.975659347590988e-05, + "loss": 3.2543, + "step": 4539 + }, + { + "epoch": 0.140042953703311, + "grad_norm": 0.16473132371902466, + "learning_rate": 4.975621860490583e-05, + "loss": 3.2651, + "step": 4540 + }, + { + "epoch": 0.1400738001688844, + "grad_norm": 0.16977083683013916, + "learning_rate": 4.97558434468683e-05, + "loss": 3.265, + "step": 4541 + }, + { + "epoch": 0.14010464663445782, + "grad_norm": 0.17483434081077576, + "learning_rate": 4.975546800180164e-05, + "loss": 3.2956, + "step": 4542 + }, + { + "epoch": 0.14013549310003123, + "grad_norm": 0.17522822320461273, + "learning_rate": 4.975509226971021e-05, + "loss": 3.2318, + "step": 4543 + }, + { + "epoch": 0.14016633956560465, + "grad_norm": 0.15692546963691711, + "learning_rate": 4.975471625059837e-05, + "loss": 3.2065, + "step": 4544 + }, + { + "epoch": 0.14019718603117806, + "grad_norm": 0.19937792420387268, + "learning_rate": 4.975433994447045e-05, + "loss": 3.2808, + "step": 4545 + }, + { + "epoch": 0.14022803249675148, + "grad_norm": 0.16736826300621033, + "learning_rate": 4.9753963351330846e-05, + "loss": 3.2678, + "step": 4546 + }, + { + "epoch": 0.1402588789623249, + "grad_norm": 0.18287503719329834, + "learning_rate": 4.975358647118391e-05, + "loss": 3.2496, + "step": 4547 + }, + { + "epoch": 0.1402897254278983, + "grad_norm": 0.16414296627044678, + "learning_rate": 4.975320930403402e-05, + "loss": 3.2952, + "step": 4548 + }, + { + "epoch": 0.14032057189347172, + "grad_norm": 0.171206533908844, + "learning_rate": 4.975283184988553e-05, + "loss": 3.2677, + "step": 4549 + }, + { + "epoch": 0.14035141835904516, + "grad_norm": 0.16830268502235413, + "learning_rate": 4.975245410874284e-05, + "loss": 3.2376, + "step": 4550 + }, + { + "epoch": 0.14038226482461857, + "grad_norm": 0.1689106673002243, + "learning_rate": 4.975207608061032e-05, + "loss": 3.2298, + "step": 4551 + }, + { + "epoch": 0.140413111290192, + "grad_norm": 0.16707070171833038, + "learning_rate": 4.975169776549234e-05, + "loss": 3.2297, + "step": 4552 + }, + { + "epoch": 0.1404439577557654, + "grad_norm": 0.1966152787208557, + "learning_rate": 4.975131916339332e-05, + "loss": 3.2977, + "step": 4553 + }, + { + "epoch": 0.14047480422133882, + "grad_norm": 0.16533666849136353, + "learning_rate": 4.975094027431761e-05, + "loss": 3.3127, + "step": 4554 + }, + { + "epoch": 0.14050565068691223, + "grad_norm": 0.1789119988679886, + "learning_rate": 4.975056109826963e-05, + "loss": 3.2634, + "step": 4555 + }, + { + "epoch": 0.14053649715248565, + "grad_norm": 0.1813502162694931, + "learning_rate": 4.9750181635253756e-05, + "loss": 3.259, + "step": 4556 + }, + { + "epoch": 0.14056734361805906, + "grad_norm": 0.1762274205684662, + "learning_rate": 4.974980188527441e-05, + "loss": 3.267, + "step": 4557 + }, + { + "epoch": 0.14059819008363247, + "grad_norm": 0.17657209932804108, + "learning_rate": 4.974942184833598e-05, + "loss": 3.2482, + "step": 4558 + }, + { + "epoch": 0.1406290365492059, + "grad_norm": 0.18124596774578094, + "learning_rate": 4.974904152444288e-05, + "loss": 3.2567, + "step": 4559 + }, + { + "epoch": 0.1406598830147793, + "grad_norm": 0.17977620661258698, + "learning_rate": 4.974866091359951e-05, + "loss": 3.2564, + "step": 4560 + }, + { + "epoch": 0.14069072948035272, + "grad_norm": 0.172136127948761, + "learning_rate": 4.9748280015810295e-05, + "loss": 3.2688, + "step": 4561 + }, + { + "epoch": 0.14072157594592616, + "grad_norm": 0.1849406361579895, + "learning_rate": 4.9747898831079644e-05, + "loss": 3.2717, + "step": 4562 + }, + { + "epoch": 0.14075242241149957, + "grad_norm": 0.1757396161556244, + "learning_rate": 4.9747517359411974e-05, + "loss": 3.3107, + "step": 4563 + }, + { + "epoch": 0.140783268877073, + "grad_norm": 0.1720939576625824, + "learning_rate": 4.974713560081171e-05, + "loss": 3.2804, + "step": 4564 + }, + { + "epoch": 0.1408141153426464, + "grad_norm": 0.16695691645145416, + "learning_rate": 4.9746753555283285e-05, + "loss": 3.2047, + "step": 4565 + }, + { + "epoch": 0.14084496180821982, + "grad_norm": 0.5233209133148193, + "learning_rate": 4.974637122283112e-05, + "loss": 3.2529, + "step": 4566 + }, + { + "epoch": 0.14087580827379323, + "grad_norm": 0.17415495216846466, + "learning_rate": 4.974598860345966e-05, + "loss": 3.2616, + "step": 4567 + }, + { + "epoch": 0.14090665473936664, + "grad_norm": 0.173706516623497, + "learning_rate": 4.9745605697173324e-05, + "loss": 3.2625, + "step": 4568 + }, + { + "epoch": 0.14093750120494006, + "grad_norm": 0.17822778224945068, + "learning_rate": 4.974522250397656e-05, + "loss": 3.3138, + "step": 4569 + }, + { + "epoch": 0.14096834767051347, + "grad_norm": 0.17068231105804443, + "learning_rate": 4.974483902387382e-05, + "loss": 3.3053, + "step": 4570 + }, + { + "epoch": 0.14099919413608689, + "grad_norm": 0.17449095845222473, + "learning_rate": 4.974445525686954e-05, + "loss": 3.2849, + "step": 4571 + }, + { + "epoch": 0.1410300406016603, + "grad_norm": 0.16263394057750702, + "learning_rate": 4.974407120296816e-05, + "loss": 3.2654, + "step": 4572 + }, + { + "epoch": 0.14106088706723371, + "grad_norm": 0.1799393594264984, + "learning_rate": 4.9743686862174154e-05, + "loss": 3.2471, + "step": 4573 + }, + { + "epoch": 0.14109173353280716, + "grad_norm": 0.17009973526000977, + "learning_rate": 4.9743302234491965e-05, + "loss": 3.2911, + "step": 4574 + }, + { + "epoch": 0.14112257999838057, + "grad_norm": 0.16540977358818054, + "learning_rate": 4.974291731992605e-05, + "loss": 3.2735, + "step": 4575 + }, + { + "epoch": 0.14115342646395398, + "grad_norm": 0.18697576224803925, + "learning_rate": 4.974253211848089e-05, + "loss": 3.2601, + "step": 4576 + }, + { + "epoch": 0.1411842729295274, + "grad_norm": 0.16857928037643433, + "learning_rate": 4.974214663016093e-05, + "loss": 3.299, + "step": 4577 + }, + { + "epoch": 0.1412151193951008, + "grad_norm": 0.19210399687290192, + "learning_rate": 4.9741760854970645e-05, + "loss": 3.2532, + "step": 4578 + }, + { + "epoch": 0.14124596586067423, + "grad_norm": 0.16434305906295776, + "learning_rate": 4.974137479291452e-05, + "loss": 3.2745, + "step": 4579 + }, + { + "epoch": 0.14127681232624764, + "grad_norm": 0.17392119765281677, + "learning_rate": 4.9740988443997016e-05, + "loss": 3.2698, + "step": 4580 + }, + { + "epoch": 0.14130765879182106, + "grad_norm": 0.16925543546676636, + "learning_rate": 4.974060180822262e-05, + "loss": 3.2383, + "step": 4581 + }, + { + "epoch": 0.14133850525739447, + "grad_norm": 0.16943661868572235, + "learning_rate": 4.9740214885595816e-05, + "loss": 3.2484, + "step": 4582 + }, + { + "epoch": 0.14136935172296788, + "grad_norm": 0.16816939413547516, + "learning_rate": 4.973982767612108e-05, + "loss": 3.2963, + "step": 4583 + }, + { + "epoch": 0.1414001981885413, + "grad_norm": 0.16942843794822693, + "learning_rate": 4.973944017980291e-05, + "loss": 3.2692, + "step": 4584 + }, + { + "epoch": 0.1414310446541147, + "grad_norm": 0.17516188323497772, + "learning_rate": 4.9739052396645805e-05, + "loss": 3.2423, + "step": 4585 + }, + { + "epoch": 0.14146189111968815, + "grad_norm": 0.1648404449224472, + "learning_rate": 4.973866432665425e-05, + "loss": 3.2618, + "step": 4586 + }, + { + "epoch": 0.14149273758526157, + "grad_norm": 0.18379832804203033, + "learning_rate": 4.973827596983275e-05, + "loss": 3.2856, + "step": 4587 + }, + { + "epoch": 0.14152358405083498, + "grad_norm": 0.1763308048248291, + "learning_rate": 4.97378873261858e-05, + "loss": 3.3081, + "step": 4588 + }, + { + "epoch": 0.1415544305164084, + "grad_norm": 0.18472643196582794, + "learning_rate": 4.973749839571793e-05, + "loss": 3.2774, + "step": 4589 + }, + { + "epoch": 0.1415852769819818, + "grad_norm": 0.16440507769584656, + "learning_rate": 4.9737109178433616e-05, + "loss": 3.2741, + "step": 4590 + }, + { + "epoch": 0.14161612344755523, + "grad_norm": 0.1891566962003708, + "learning_rate": 4.973671967433738e-05, + "loss": 3.2418, + "step": 4591 + }, + { + "epoch": 0.14164696991312864, + "grad_norm": 0.23498940467834473, + "learning_rate": 4.973632988343376e-05, + "loss": 3.2969, + "step": 4592 + }, + { + "epoch": 0.14167781637870205, + "grad_norm": 0.1747361719608307, + "learning_rate": 4.973593980572725e-05, + "loss": 3.2561, + "step": 4593 + }, + { + "epoch": 0.14170866284427547, + "grad_norm": 0.16682656109333038, + "learning_rate": 4.973554944122239e-05, + "loss": 3.2337, + "step": 4594 + }, + { + "epoch": 0.14173950930984888, + "grad_norm": 0.16854506731033325, + "learning_rate": 4.97351587899237e-05, + "loss": 3.259, + "step": 4595 + }, + { + "epoch": 0.1417703557754223, + "grad_norm": 0.17435652017593384, + "learning_rate": 4.97347678518357e-05, + "loss": 3.2942, + "step": 4596 + }, + { + "epoch": 0.1418012022409957, + "grad_norm": 0.17078542709350586, + "learning_rate": 4.973437662696293e-05, + "loss": 3.2844, + "step": 4597 + }, + { + "epoch": 0.14183204870656915, + "grad_norm": 0.16284428536891937, + "learning_rate": 4.973398511530994e-05, + "loss": 3.2551, + "step": 4598 + }, + { + "epoch": 0.14186289517214257, + "grad_norm": 0.16277723014354706, + "learning_rate": 4.9733593316881235e-05, + "loss": 3.2523, + "step": 4599 + }, + { + "epoch": 0.14189374163771598, + "grad_norm": 0.17323996126651764, + "learning_rate": 4.973320123168139e-05, + "loss": 3.2872, + "step": 4600 + }, + { + "epoch": 0.1419245881032894, + "grad_norm": 0.1952444612979889, + "learning_rate": 4.973280885971494e-05, + "loss": 3.2585, + "step": 4601 + }, + { + "epoch": 0.1419554345688628, + "grad_norm": 0.15817788243293762, + "learning_rate": 4.973241620098643e-05, + "loss": 3.2767, + "step": 4602 + }, + { + "epoch": 0.14198628103443622, + "grad_norm": 0.1703493744134903, + "learning_rate": 4.973202325550043e-05, + "loss": 3.2191, + "step": 4603 + }, + { + "epoch": 0.14201712750000964, + "grad_norm": 0.16055740416049957, + "learning_rate": 4.973163002326147e-05, + "loss": 3.2072, + "step": 4604 + }, + { + "epoch": 0.14204797396558305, + "grad_norm": 0.17484186589717865, + "learning_rate": 4.9731236504274114e-05, + "loss": 3.3057, + "step": 4605 + }, + { + "epoch": 0.14207882043115647, + "grad_norm": 0.16503332555294037, + "learning_rate": 4.9730842698542945e-05, + "loss": 3.2377, + "step": 4606 + }, + { + "epoch": 0.14210966689672988, + "grad_norm": 0.17812606692314148, + "learning_rate": 4.9730448606072516e-05, + "loss": 3.3158, + "step": 4607 + }, + { + "epoch": 0.1421405133623033, + "grad_norm": 0.1695965677499771, + "learning_rate": 4.9730054226867395e-05, + "loss": 3.2896, + "step": 4608 + }, + { + "epoch": 0.1421713598278767, + "grad_norm": 0.16289865970611572, + "learning_rate": 4.9729659560932154e-05, + "loss": 3.2392, + "step": 4609 + }, + { + "epoch": 0.14220220629345015, + "grad_norm": 0.17098335921764374, + "learning_rate": 4.9729264608271375e-05, + "loss": 3.257, + "step": 4610 + }, + { + "epoch": 0.14223305275902356, + "grad_norm": 0.19916769862174988, + "learning_rate": 4.972886936888963e-05, + "loss": 3.2796, + "step": 4611 + }, + { + "epoch": 0.14226389922459698, + "grad_norm": 0.17384274303913116, + "learning_rate": 4.972847384279151e-05, + "loss": 3.2663, + "step": 4612 + }, + { + "epoch": 0.1422947456901704, + "grad_norm": 0.16994468867778778, + "learning_rate": 4.972807802998159e-05, + "loss": 3.2841, + "step": 4613 + }, + { + "epoch": 0.1423255921557438, + "grad_norm": 0.18010026216506958, + "learning_rate": 4.972768193046446e-05, + "loss": 3.2851, + "step": 4614 + }, + { + "epoch": 0.14235643862131722, + "grad_norm": 0.3014155924320221, + "learning_rate": 4.9727285544244737e-05, + "loss": 3.2477, + "step": 4615 + }, + { + "epoch": 0.14238728508689064, + "grad_norm": 0.17652180790901184, + "learning_rate": 4.9726888871326984e-05, + "loss": 3.2701, + "step": 4616 + }, + { + "epoch": 0.14241813155246405, + "grad_norm": 0.17518949508666992, + "learning_rate": 4.972649191171581e-05, + "loss": 3.2746, + "step": 4617 + }, + { + "epoch": 0.14244897801803746, + "grad_norm": 0.18382926285266876, + "learning_rate": 4.972609466541583e-05, + "loss": 3.2701, + "step": 4618 + }, + { + "epoch": 0.14247982448361088, + "grad_norm": 0.17557008564472198, + "learning_rate": 4.972569713243164e-05, + "loss": 3.266, + "step": 4619 + }, + { + "epoch": 0.1425106709491843, + "grad_norm": 0.18309666216373444, + "learning_rate": 4.972529931276785e-05, + "loss": 3.2914, + "step": 4620 + }, + { + "epoch": 0.1425415174147577, + "grad_norm": 0.21145054697990417, + "learning_rate": 4.972490120642906e-05, + "loss": 3.2812, + "step": 4621 + }, + { + "epoch": 0.14257236388033115, + "grad_norm": 0.16776135563850403, + "learning_rate": 4.9724502813419915e-05, + "loss": 3.2561, + "step": 4622 + }, + { + "epoch": 0.14260321034590456, + "grad_norm": 0.18322841823101044, + "learning_rate": 4.972410413374501e-05, + "loss": 3.2709, + "step": 4623 + }, + { + "epoch": 0.14263405681147798, + "grad_norm": 0.16100217401981354, + "learning_rate": 4.9723705167408974e-05, + "loss": 3.2819, + "step": 4624 + }, + { + "epoch": 0.1426649032770514, + "grad_norm": 0.1790846437215805, + "learning_rate": 4.9723305914416443e-05, + "loss": 3.2442, + "step": 4625 + }, + { + "epoch": 0.1426957497426248, + "grad_norm": 0.17188335955142975, + "learning_rate": 4.9722906374772026e-05, + "loss": 3.2642, + "step": 4626 + }, + { + "epoch": 0.14272659620819822, + "grad_norm": 0.16654010117053986, + "learning_rate": 4.9722506548480376e-05, + "loss": 3.2584, + "step": 4627 + }, + { + "epoch": 0.14275744267377163, + "grad_norm": 0.1713726967573166, + "learning_rate": 4.972210643554611e-05, + "loss": 3.2581, + "step": 4628 + }, + { + "epoch": 0.14278828913934505, + "grad_norm": 0.17092378437519073, + "learning_rate": 4.972170603597388e-05, + "loss": 3.284, + "step": 4629 + }, + { + "epoch": 0.14281913560491846, + "grad_norm": 0.1754758358001709, + "learning_rate": 4.972130534976832e-05, + "loss": 3.3022, + "step": 4630 + }, + { + "epoch": 0.14284998207049188, + "grad_norm": 0.16047930717468262, + "learning_rate": 4.9720904376934084e-05, + "loss": 3.259, + "step": 4631 + }, + { + "epoch": 0.1428808285360653, + "grad_norm": 0.17678943276405334, + "learning_rate": 4.972050311747582e-05, + "loss": 3.2523, + "step": 4632 + }, + { + "epoch": 0.1429116750016387, + "grad_norm": 0.16436053812503815, + "learning_rate": 4.9720101571398175e-05, + "loss": 3.2444, + "step": 4633 + }, + { + "epoch": 0.14294252146721215, + "grad_norm": 0.16274523735046387, + "learning_rate": 4.9719699738705804e-05, + "loss": 3.2741, + "step": 4634 + }, + { + "epoch": 0.14297336793278556, + "grad_norm": 0.17794060707092285, + "learning_rate": 4.9719297619403375e-05, + "loss": 3.2734, + "step": 4635 + }, + { + "epoch": 0.14300421439835898, + "grad_norm": 0.1614210456609726, + "learning_rate": 4.9718895213495544e-05, + "loss": 3.2823, + "step": 4636 + }, + { + "epoch": 0.1430350608639324, + "grad_norm": 0.16563859581947327, + "learning_rate": 4.971849252098698e-05, + "loss": 3.2746, + "step": 4637 + }, + { + "epoch": 0.1430659073295058, + "grad_norm": 0.16676698625087738, + "learning_rate": 4.971808954188234e-05, + "loss": 3.264, + "step": 4638 + }, + { + "epoch": 0.14309675379507922, + "grad_norm": 0.16957929730415344, + "learning_rate": 4.9717686276186316e-05, + "loss": 3.2325, + "step": 4639 + }, + { + "epoch": 0.14312760026065263, + "grad_norm": 0.15884146094322205, + "learning_rate": 4.9717282723903566e-05, + "loss": 3.2298, + "step": 4640 + }, + { + "epoch": 0.14315844672622605, + "grad_norm": 0.15922687947750092, + "learning_rate": 4.971687888503878e-05, + "loss": 3.2521, + "step": 4641 + }, + { + "epoch": 0.14318929319179946, + "grad_norm": 0.1787968873977661, + "learning_rate": 4.971647475959663e-05, + "loss": 3.2891, + "step": 4642 + }, + { + "epoch": 0.14322013965737287, + "grad_norm": 0.15875202417373657, + "learning_rate": 4.9716070347581814e-05, + "loss": 3.2966, + "step": 4643 + }, + { + "epoch": 0.1432509861229463, + "grad_norm": 0.17252448201179504, + "learning_rate": 4.9715665648999016e-05, + "loss": 3.2857, + "step": 4644 + }, + { + "epoch": 0.1432818325885197, + "grad_norm": 0.18042588233947754, + "learning_rate": 4.971526066385292e-05, + "loss": 3.2624, + "step": 4645 + }, + { + "epoch": 0.14331267905409315, + "grad_norm": 0.1601305603981018, + "learning_rate": 4.971485539214824e-05, + "loss": 3.2573, + "step": 4646 + }, + { + "epoch": 0.14334352551966656, + "grad_norm": 0.16241754591464996, + "learning_rate": 4.971444983388965e-05, + "loss": 3.2859, + "step": 4647 + }, + { + "epoch": 0.14337437198523997, + "grad_norm": 0.6730847954750061, + "learning_rate": 4.971404398908188e-05, + "loss": 3.2813, + "step": 4648 + }, + { + "epoch": 0.1434052184508134, + "grad_norm": 0.18233583867549896, + "learning_rate": 4.971363785772961e-05, + "loss": 3.2633, + "step": 4649 + }, + { + "epoch": 0.1434360649163868, + "grad_norm": 0.17620688676834106, + "learning_rate": 4.9713231439837574e-05, + "loss": 3.2468, + "step": 4650 + }, + { + "epoch": 0.14346691138196022, + "grad_norm": 0.1802932769060135, + "learning_rate": 4.971282473541046e-05, + "loss": 3.2733, + "step": 4651 + }, + { + "epoch": 0.14349775784753363, + "grad_norm": 0.22499626874923706, + "learning_rate": 4.971241774445299e-05, + "loss": 3.244, + "step": 4652 + }, + { + "epoch": 0.14352860431310704, + "grad_norm": 0.17222407460212708, + "learning_rate": 4.9712010466969896e-05, + "loss": 3.2832, + "step": 4653 + }, + { + "epoch": 0.14355945077868046, + "grad_norm": 0.1815350353717804, + "learning_rate": 4.971160290296588e-05, + "loss": 3.2782, + "step": 4654 + }, + { + "epoch": 0.14359029724425387, + "grad_norm": 0.17150402069091797, + "learning_rate": 4.9711195052445684e-05, + "loss": 3.2472, + "step": 4655 + }, + { + "epoch": 0.1436211437098273, + "grad_norm": 0.17507225275039673, + "learning_rate": 4.971078691541403e-05, + "loss": 3.265, + "step": 4656 + }, + { + "epoch": 0.1436519901754007, + "grad_norm": 0.1666886806488037, + "learning_rate": 4.971037849187566e-05, + "loss": 3.2538, + "step": 4657 + }, + { + "epoch": 0.14368283664097414, + "grad_norm": 0.1793103963136673, + "learning_rate": 4.970996978183529e-05, + "loss": 3.264, + "step": 4658 + }, + { + "epoch": 0.14371368310654756, + "grad_norm": 0.18008531630039215, + "learning_rate": 4.970956078529767e-05, + "loss": 3.2964, + "step": 4659 + }, + { + "epoch": 0.14374452957212097, + "grad_norm": 0.1772540807723999, + "learning_rate": 4.970915150226755e-05, + "loss": 3.27, + "step": 4660 + }, + { + "epoch": 0.14377537603769439, + "grad_norm": 0.17883415520191193, + "learning_rate": 4.970874193274966e-05, + "loss": 3.2591, + "step": 4661 + }, + { + "epoch": 0.1438062225032678, + "grad_norm": 0.3270735740661621, + "learning_rate": 4.9708332076748755e-05, + "loss": 3.2267, + "step": 4662 + }, + { + "epoch": 0.14383706896884121, + "grad_norm": 0.2221916913986206, + "learning_rate": 4.970792193426959e-05, + "loss": 3.277, + "step": 4663 + }, + { + "epoch": 0.14386791543441463, + "grad_norm": 0.17583003640174866, + "learning_rate": 4.970751150531692e-05, + "loss": 3.2453, + "step": 4664 + }, + { + "epoch": 0.14389876189998804, + "grad_norm": 0.1893818974494934, + "learning_rate": 4.97071007898955e-05, + "loss": 3.2835, + "step": 4665 + }, + { + "epoch": 0.14392960836556146, + "grad_norm": 0.18609246611595154, + "learning_rate": 4.9706689788010085e-05, + "loss": 3.2512, + "step": 4666 + }, + { + "epoch": 0.14396045483113487, + "grad_norm": 0.16173283755779266, + "learning_rate": 4.970627849966546e-05, + "loss": 3.2377, + "step": 4667 + }, + { + "epoch": 0.14399130129670829, + "grad_norm": 0.16029773652553558, + "learning_rate": 4.970586692486638e-05, + "loss": 3.2876, + "step": 4668 + }, + { + "epoch": 0.1440221477622817, + "grad_norm": 0.19377712905406952, + "learning_rate": 4.970545506361762e-05, + "loss": 3.2868, + "step": 4669 + }, + { + "epoch": 0.14405299422785514, + "grad_norm": 0.16322551667690277, + "learning_rate": 4.970504291592395e-05, + "loss": 3.2595, + "step": 4670 + }, + { + "epoch": 0.14408384069342856, + "grad_norm": 0.16035325825214386, + "learning_rate": 4.970463048179016e-05, + "loss": 3.2359, + "step": 4671 + }, + { + "epoch": 0.14411468715900197, + "grad_norm": 0.16938719153404236, + "learning_rate": 4.970421776122102e-05, + "loss": 3.2785, + "step": 4672 + }, + { + "epoch": 0.14414553362457538, + "grad_norm": 0.1662767231464386, + "learning_rate": 4.9703804754221326e-05, + "loss": 3.2889, + "step": 4673 + }, + { + "epoch": 0.1441763800901488, + "grad_norm": 0.16679804027080536, + "learning_rate": 4.9703391460795855e-05, + "loss": 3.2854, + "step": 4674 + }, + { + "epoch": 0.1442072265557222, + "grad_norm": 0.16875532269477844, + "learning_rate": 4.9702977880949406e-05, + "loss": 3.2649, + "step": 4675 + }, + { + "epoch": 0.14423807302129563, + "grad_norm": 0.17042727768421173, + "learning_rate": 4.9702564014686774e-05, + "loss": 3.243, + "step": 4676 + }, + { + "epoch": 0.14426891948686904, + "grad_norm": 0.16052016615867615, + "learning_rate": 4.970214986201276e-05, + "loss": 3.2462, + "step": 4677 + }, + { + "epoch": 0.14429976595244245, + "grad_norm": 0.16725608706474304, + "learning_rate": 4.970173542293216e-05, + "loss": 3.2408, + "step": 4678 + }, + { + "epoch": 0.14433061241801587, + "grad_norm": 0.16650241613388062, + "learning_rate": 4.970132069744978e-05, + "loss": 3.2604, + "step": 4679 + }, + { + "epoch": 0.14436145888358928, + "grad_norm": 0.15799301862716675, + "learning_rate": 4.970090568557043e-05, + "loss": 3.2519, + "step": 4680 + }, + { + "epoch": 0.14439230534916273, + "grad_norm": 0.16440193355083466, + "learning_rate": 4.9700490387298926e-05, + "loss": 3.2753, + "step": 4681 + }, + { + "epoch": 0.14442315181473614, + "grad_norm": 0.16553303599357605, + "learning_rate": 4.970007480264008e-05, + "loss": 3.2448, + "step": 4682 + }, + { + "epoch": 0.14445399828030955, + "grad_norm": 0.16871048510074615, + "learning_rate": 4.9699658931598704e-05, + "loss": 3.2485, + "step": 4683 + }, + { + "epoch": 0.14448484474588297, + "grad_norm": 0.16033823788166046, + "learning_rate": 4.969924277417963e-05, + "loss": 3.2725, + "step": 4684 + }, + { + "epoch": 0.14451569121145638, + "grad_norm": 0.17537014186382294, + "learning_rate": 4.969882633038768e-05, + "loss": 3.2912, + "step": 4685 + }, + { + "epoch": 0.1445465376770298, + "grad_norm": 0.175490602850914, + "learning_rate": 4.969840960022768e-05, + "loss": 3.2856, + "step": 4686 + }, + { + "epoch": 0.1445773841426032, + "grad_norm": 0.16546660661697388, + "learning_rate": 4.9697992583704454e-05, + "loss": 3.2721, + "step": 4687 + }, + { + "epoch": 0.14460823060817662, + "grad_norm": 0.17601613700389862, + "learning_rate": 4.969757528082286e-05, + "loss": 3.2895, + "step": 4688 + }, + { + "epoch": 0.14463907707375004, + "grad_norm": 0.17294315993785858, + "learning_rate": 4.9697157691587715e-05, + "loss": 3.2882, + "step": 4689 + }, + { + "epoch": 0.14466992353932345, + "grad_norm": 0.17879542708396912, + "learning_rate": 4.969673981600386e-05, + "loss": 3.2898, + "step": 4690 + }, + { + "epoch": 0.14470077000489687, + "grad_norm": 0.16930875182151794, + "learning_rate": 4.969632165407616e-05, + "loss": 3.2284, + "step": 4691 + }, + { + "epoch": 0.14473161647047028, + "grad_norm": 0.17244942486286163, + "learning_rate": 4.9695903205809444e-05, + "loss": 3.2596, + "step": 4692 + }, + { + "epoch": 0.14476246293604372, + "grad_norm": 0.18789468705654144, + "learning_rate": 4.969548447120858e-05, + "loss": 3.2473, + "step": 4693 + }, + { + "epoch": 0.14479330940161714, + "grad_norm": 0.1745796501636505, + "learning_rate": 4.96950654502784e-05, + "loss": 3.2447, + "step": 4694 + }, + { + "epoch": 0.14482415586719055, + "grad_norm": 0.1633003056049347, + "learning_rate": 4.969464614302378e-05, + "loss": 3.278, + "step": 4695 + }, + { + "epoch": 0.14485500233276397, + "grad_norm": 0.18118520081043243, + "learning_rate": 4.9694226549449585e-05, + "loss": 3.253, + "step": 4696 + }, + { + "epoch": 0.14488584879833738, + "grad_norm": 0.16673274338245392, + "learning_rate": 4.969380666956067e-05, + "loss": 3.2658, + "step": 4697 + }, + { + "epoch": 0.1449166952639108, + "grad_norm": 0.17565107345581055, + "learning_rate": 4.969338650336191e-05, + "loss": 3.2734, + "step": 4698 + }, + { + "epoch": 0.1449475417294842, + "grad_norm": 0.17259076237678528, + "learning_rate": 4.969296605085816e-05, + "loss": 3.2557, + "step": 4699 + }, + { + "epoch": 0.14497838819505762, + "grad_norm": 0.17177660763263702, + "learning_rate": 4.969254531205432e-05, + "loss": 3.2835, + "step": 4700 + }, + { + "epoch": 0.14500923466063104, + "grad_norm": 0.16820871829986572, + "learning_rate": 4.9692124286955244e-05, + "loss": 3.2727, + "step": 4701 + }, + { + "epoch": 0.14504008112620445, + "grad_norm": 0.1789037436246872, + "learning_rate": 4.9691702975565836e-05, + "loss": 3.2771, + "step": 4702 + }, + { + "epoch": 0.14507092759177787, + "grad_norm": 0.18064187467098236, + "learning_rate": 4.969128137789096e-05, + "loss": 3.2193, + "step": 4703 + }, + { + "epoch": 0.14510177405735128, + "grad_norm": 0.17389065027236938, + "learning_rate": 4.969085949393552e-05, + "loss": 3.2774, + "step": 4704 + }, + { + "epoch": 0.14513262052292472, + "grad_norm": 0.1644088625907898, + "learning_rate": 4.969043732370441e-05, + "loss": 3.2701, + "step": 4705 + }, + { + "epoch": 0.14516346698849814, + "grad_norm": 0.17321191728115082, + "learning_rate": 4.969001486720251e-05, + "loss": 3.2915, + "step": 4706 + }, + { + "epoch": 0.14519431345407155, + "grad_norm": 0.16478665173053741, + "learning_rate": 4.9689592124434714e-05, + "loss": 3.3159, + "step": 4707 + }, + { + "epoch": 0.14522515991964496, + "grad_norm": 0.18289022147655487, + "learning_rate": 4.968916909540594e-05, + "loss": 3.2785, + "step": 4708 + }, + { + "epoch": 0.14525600638521838, + "grad_norm": 0.32661446928977966, + "learning_rate": 4.968874578012109e-05, + "loss": 3.2586, + "step": 4709 + }, + { + "epoch": 0.1452868528507918, + "grad_norm": 0.17452916502952576, + "learning_rate": 4.968832217858507e-05, + "loss": 3.2682, + "step": 4710 + }, + { + "epoch": 0.1453176993163652, + "grad_norm": 0.1658896654844284, + "learning_rate": 4.968789829080279e-05, + "loss": 3.2746, + "step": 4711 + }, + { + "epoch": 0.14534854578193862, + "grad_norm": 0.16727152466773987, + "learning_rate": 4.968747411677916e-05, + "loss": 3.2822, + "step": 4712 + }, + { + "epoch": 0.14537939224751203, + "grad_norm": 0.1764206737279892, + "learning_rate": 4.968704965651911e-05, + "loss": 3.2697, + "step": 4713 + }, + { + "epoch": 0.14541023871308545, + "grad_norm": 0.1624937653541565, + "learning_rate": 4.9686624910027545e-05, + "loss": 3.2817, + "step": 4714 + }, + { + "epoch": 0.14544108517865886, + "grad_norm": 0.17091849446296692, + "learning_rate": 4.968619987730941e-05, + "loss": 3.2843, + "step": 4715 + }, + { + "epoch": 0.14547193164423228, + "grad_norm": 0.16098274290561676, + "learning_rate": 4.9685774558369614e-05, + "loss": 3.271, + "step": 4716 + }, + { + "epoch": 0.14550277810980572, + "grad_norm": 0.17163509130477905, + "learning_rate": 4.96853489532131e-05, + "loss": 3.2556, + "step": 4717 + }, + { + "epoch": 0.14553362457537913, + "grad_norm": 0.16163617372512817, + "learning_rate": 4.96849230618448e-05, + "loss": 3.2486, + "step": 4718 + }, + { + "epoch": 0.14556447104095255, + "grad_norm": 0.1585303097963333, + "learning_rate": 4.968449688426965e-05, + "loss": 3.2337, + "step": 4719 + }, + { + "epoch": 0.14559531750652596, + "grad_norm": 0.16764803230762482, + "learning_rate": 4.968407042049259e-05, + "loss": 3.2739, + "step": 4720 + }, + { + "epoch": 0.14562616397209938, + "grad_norm": 0.1666874885559082, + "learning_rate": 4.968364367051857e-05, + "loss": 3.2782, + "step": 4721 + }, + { + "epoch": 0.1456570104376728, + "grad_norm": 0.1685873568058014, + "learning_rate": 4.968321663435254e-05, + "loss": 3.2728, + "step": 4722 + }, + { + "epoch": 0.1456878569032462, + "grad_norm": 0.16596581041812897, + "learning_rate": 4.968278931199943e-05, + "loss": 3.2626, + "step": 4723 + }, + { + "epoch": 0.14571870336881962, + "grad_norm": 0.16654403507709503, + "learning_rate": 4.968236170346422e-05, + "loss": 3.2533, + "step": 4724 + }, + { + "epoch": 0.14574954983439303, + "grad_norm": 0.16331537067890167, + "learning_rate": 4.968193380875186e-05, + "loss": 3.2595, + "step": 4725 + }, + { + "epoch": 0.14578039629996645, + "grad_norm": 0.172569140791893, + "learning_rate": 4.968150562786731e-05, + "loss": 3.2573, + "step": 4726 + }, + { + "epoch": 0.14581124276553986, + "grad_norm": 0.16655175387859344, + "learning_rate": 4.968107716081553e-05, + "loss": 3.2532, + "step": 4727 + }, + { + "epoch": 0.14584208923111328, + "grad_norm": 0.1608237475156784, + "learning_rate": 4.9680648407601495e-05, + "loss": 3.2593, + "step": 4728 + }, + { + "epoch": 0.14587293569668672, + "grad_norm": 0.16653253138065338, + "learning_rate": 4.968021936823018e-05, + "loss": 3.29, + "step": 4729 + }, + { + "epoch": 0.14590378216226013, + "grad_norm": 0.16896679997444153, + "learning_rate": 4.967979004270654e-05, + "loss": 3.298, + "step": 4730 + }, + { + "epoch": 0.14593462862783355, + "grad_norm": 0.16114725172519684, + "learning_rate": 4.967936043103557e-05, + "loss": 3.2562, + "step": 4731 + }, + { + "epoch": 0.14596547509340696, + "grad_norm": 0.1658214032649994, + "learning_rate": 4.967893053322225e-05, + "loss": 3.2662, + "step": 4732 + }, + { + "epoch": 0.14599632155898037, + "grad_norm": 0.16426178812980652, + "learning_rate": 4.967850034927155e-05, + "loss": 3.2534, + "step": 4733 + }, + { + "epoch": 0.1460271680245538, + "grad_norm": 1.0007482767105103, + "learning_rate": 4.9678069879188476e-05, + "loss": 3.2745, + "step": 4734 + }, + { + "epoch": 0.1460580144901272, + "grad_norm": 0.17359444499015808, + "learning_rate": 4.967763912297802e-05, + "loss": 3.2662, + "step": 4735 + }, + { + "epoch": 0.14608886095570062, + "grad_norm": 0.1774689108133316, + "learning_rate": 4.9677208080645154e-05, + "loss": 3.2448, + "step": 4736 + }, + { + "epoch": 0.14611970742127403, + "grad_norm": 0.165381520986557, + "learning_rate": 4.967677675219489e-05, + "loss": 3.2645, + "step": 4737 + }, + { + "epoch": 0.14615055388684745, + "grad_norm": 0.18467961251735687, + "learning_rate": 4.9676345137632236e-05, + "loss": 3.2534, + "step": 4738 + }, + { + "epoch": 0.14618140035242086, + "grad_norm": 0.16828742623329163, + "learning_rate": 4.967591323696218e-05, + "loss": 3.2339, + "step": 4739 + }, + { + "epoch": 0.14621224681799427, + "grad_norm": 0.3097487986087799, + "learning_rate": 4.967548105018975e-05, + "loss": 3.2834, + "step": 4740 + }, + { + "epoch": 0.14624309328356772, + "grad_norm": 0.17473027110099792, + "learning_rate": 4.9675048577319936e-05, + "loss": 3.2626, + "step": 4741 + }, + { + "epoch": 0.14627393974914113, + "grad_norm": 0.17475572228431702, + "learning_rate": 4.967461581835777e-05, + "loss": 3.2776, + "step": 4742 + }, + { + "epoch": 0.14630478621471454, + "grad_norm": 0.1830933392047882, + "learning_rate": 4.9674182773308254e-05, + "loss": 3.2754, + "step": 4743 + }, + { + "epoch": 0.14633563268028796, + "grad_norm": 0.19199053943157196, + "learning_rate": 4.967374944217642e-05, + "loss": 3.3011, + "step": 4744 + }, + { + "epoch": 0.14636647914586137, + "grad_norm": 0.17823760211467743, + "learning_rate": 4.967331582496729e-05, + "loss": 3.2316, + "step": 4745 + }, + { + "epoch": 0.1463973256114348, + "grad_norm": 0.17637988924980164, + "learning_rate": 4.967288192168589e-05, + "loss": 3.2918, + "step": 4746 + }, + { + "epoch": 0.1464281720770082, + "grad_norm": 0.18042968213558197, + "learning_rate": 4.967244773233724e-05, + "loss": 3.2429, + "step": 4747 + }, + { + "epoch": 0.14645901854258161, + "grad_norm": 0.17884038388729095, + "learning_rate": 4.9672013256926394e-05, + "loss": 3.2516, + "step": 4748 + }, + { + "epoch": 0.14648986500815503, + "grad_norm": 0.17371395230293274, + "learning_rate": 4.9671578495458384e-05, + "loss": 3.2739, + "step": 4749 + }, + { + "epoch": 0.14652071147372844, + "grad_norm": 0.4506058990955353, + "learning_rate": 4.967114344793824e-05, + "loss": 3.2777, + "step": 4750 + }, + { + "epoch": 0.14655155793930186, + "grad_norm": 0.16754110157489777, + "learning_rate": 4.9670708114371024e-05, + "loss": 3.258, + "step": 4751 + }, + { + "epoch": 0.14658240440487527, + "grad_norm": 0.18666549026966095, + "learning_rate": 4.967027249476176e-05, + "loss": 3.2821, + "step": 4752 + }, + { + "epoch": 0.1466132508704487, + "grad_norm": 1.6723783016204834, + "learning_rate": 4.9669836589115516e-05, + "loss": 3.2726, + "step": 4753 + }, + { + "epoch": 0.14664409733602213, + "grad_norm": 0.23592552542686462, + "learning_rate": 4.966940039743734e-05, + "loss": 3.1955, + "step": 4754 + }, + { + "epoch": 0.14667494380159554, + "grad_norm": 0.3137739300727844, + "learning_rate": 4.9668963919732295e-05, + "loss": 3.2622, + "step": 4755 + }, + { + "epoch": 0.14670579026716896, + "grad_norm": 0.2950877249240875, + "learning_rate": 4.966852715600544e-05, + "loss": 3.2714, + "step": 4756 + }, + { + "epoch": 0.14673663673274237, + "grad_norm": 0.2438563108444214, + "learning_rate": 4.9668090106261834e-05, + "loss": 3.2569, + "step": 4757 + }, + { + "epoch": 0.14676748319831578, + "grad_norm": 0.230081245303154, + "learning_rate": 4.9667652770506544e-05, + "loss": 3.2491, + "step": 4758 + }, + { + "epoch": 0.1467983296638892, + "grad_norm": 0.2358185350894928, + "learning_rate": 4.966721514874464e-05, + "loss": 3.2559, + "step": 4759 + }, + { + "epoch": 0.1468291761294626, + "grad_norm": 0.22605015337467194, + "learning_rate": 4.966677724098121e-05, + "loss": 3.2937, + "step": 4760 + }, + { + "epoch": 0.14686002259503603, + "grad_norm": 0.22722893953323364, + "learning_rate": 4.966633904722132e-05, + "loss": 3.2262, + "step": 4761 + }, + { + "epoch": 0.14689086906060944, + "grad_norm": 0.23074941337108612, + "learning_rate": 4.966590056747005e-05, + "loss": 3.292, + "step": 4762 + }, + { + "epoch": 0.14692171552618286, + "grad_norm": 0.20846964418888092, + "learning_rate": 4.9665461801732474e-05, + "loss": 3.2732, + "step": 4763 + }, + { + "epoch": 0.14695256199175627, + "grad_norm": 0.20465363562107086, + "learning_rate": 4.96650227500137e-05, + "loss": 3.2839, + "step": 4764 + }, + { + "epoch": 0.1469834084573297, + "grad_norm": 0.18582074344158173, + "learning_rate": 4.966458341231881e-05, + "loss": 3.2703, + "step": 4765 + }, + { + "epoch": 0.14701425492290313, + "grad_norm": 0.2007139027118683, + "learning_rate": 4.96641437886529e-05, + "loss": 3.2764, + "step": 4766 + }, + { + "epoch": 0.14704510138847654, + "grad_norm": 0.18205274641513824, + "learning_rate": 4.966370387902106e-05, + "loss": 3.2637, + "step": 4767 + }, + { + "epoch": 0.14707594785404995, + "grad_norm": 0.205934539437294, + "learning_rate": 4.966326368342839e-05, + "loss": 3.2937, + "step": 4768 + }, + { + "epoch": 0.14710679431962337, + "grad_norm": 0.18179312348365784, + "learning_rate": 4.966282320188001e-05, + "loss": 3.2677, + "step": 4769 + }, + { + "epoch": 0.14713764078519678, + "grad_norm": 0.1765410602092743, + "learning_rate": 4.9662382434381004e-05, + "loss": 3.2969, + "step": 4770 + }, + { + "epoch": 0.1471684872507702, + "grad_norm": 0.18051041662693024, + "learning_rate": 4.966194138093649e-05, + "loss": 3.2818, + "step": 4771 + }, + { + "epoch": 0.1471993337163436, + "grad_norm": 0.16892342269420624, + "learning_rate": 4.96615000415516e-05, + "loss": 3.2528, + "step": 4772 + }, + { + "epoch": 0.14723018018191703, + "grad_norm": 0.16674546897411346, + "learning_rate": 4.966105841623143e-05, + "loss": 3.2866, + "step": 4773 + }, + { + "epoch": 0.14726102664749044, + "grad_norm": 0.16530844569206238, + "learning_rate": 4.966061650498111e-05, + "loss": 3.2434, + "step": 4774 + }, + { + "epoch": 0.14729187311306385, + "grad_norm": 0.1690329909324646, + "learning_rate": 4.966017430780576e-05, + "loss": 3.2439, + "step": 4775 + }, + { + "epoch": 0.14732271957863727, + "grad_norm": 0.16553524136543274, + "learning_rate": 4.965973182471051e-05, + "loss": 3.2649, + "step": 4776 + }, + { + "epoch": 0.1473535660442107, + "grad_norm": 0.17220225930213928, + "learning_rate": 4.9659289055700484e-05, + "loss": 3.2699, + "step": 4777 + }, + { + "epoch": 0.14738441250978412, + "grad_norm": 0.5796443223953247, + "learning_rate": 4.965884600078082e-05, + "loss": 3.2667, + "step": 4778 + }, + { + "epoch": 0.14741525897535754, + "grad_norm": 0.17073753476142883, + "learning_rate": 4.965840265995666e-05, + "loss": 3.2876, + "step": 4779 + }, + { + "epoch": 0.14744610544093095, + "grad_norm": 0.1693287491798401, + "learning_rate": 4.9657959033233133e-05, + "loss": 3.2524, + "step": 4780 + }, + { + "epoch": 0.14747695190650437, + "grad_norm": 0.23084601759910583, + "learning_rate": 4.9657515120615396e-05, + "loss": 3.2491, + "step": 4781 + }, + { + "epoch": 0.14750779837207778, + "grad_norm": 0.17548713088035583, + "learning_rate": 4.965707092210858e-05, + "loss": 3.3085, + "step": 4782 + }, + { + "epoch": 0.1475386448376512, + "grad_norm": 0.1717001348733902, + "learning_rate": 4.965662643771785e-05, + "loss": 3.2507, + "step": 4783 + }, + { + "epoch": 0.1475694913032246, + "grad_norm": 0.16281794011592865, + "learning_rate": 4.965618166744835e-05, + "loss": 3.2386, + "step": 4784 + }, + { + "epoch": 0.14760033776879802, + "grad_norm": 0.1651829034090042, + "learning_rate": 4.9655736611305235e-05, + "loss": 3.2736, + "step": 4785 + }, + { + "epoch": 0.14763118423437144, + "grad_norm": 0.16399723291397095, + "learning_rate": 4.965529126929368e-05, + "loss": 3.2745, + "step": 4786 + }, + { + "epoch": 0.14766203069994485, + "grad_norm": 0.1735386699438095, + "learning_rate": 4.965484564141884e-05, + "loss": 3.2618, + "step": 4787 + }, + { + "epoch": 0.14769287716551827, + "grad_norm": 0.16746537387371063, + "learning_rate": 4.965439972768588e-05, + "loss": 3.2867, + "step": 4788 + }, + { + "epoch": 0.1477237236310917, + "grad_norm": 0.16455253958702087, + "learning_rate": 4.9653953528099964e-05, + "loss": 3.2543, + "step": 4789 + }, + { + "epoch": 0.14775457009666512, + "grad_norm": 0.163854718208313, + "learning_rate": 4.965350704266627e-05, + "loss": 3.2679, + "step": 4790 + }, + { + "epoch": 0.14778541656223854, + "grad_norm": 0.16958965361118317, + "learning_rate": 4.9653060271389984e-05, + "loss": 3.2605, + "step": 4791 + }, + { + "epoch": 0.14781626302781195, + "grad_norm": 0.21758243441581726, + "learning_rate": 4.965261321427628e-05, + "loss": 3.2695, + "step": 4792 + }, + { + "epoch": 0.14784710949338536, + "grad_norm": 0.16629241406917572, + "learning_rate": 4.9652165871330334e-05, + "loss": 3.2874, + "step": 4793 + }, + { + "epoch": 0.14787795595895878, + "grad_norm": 0.16789160668849945, + "learning_rate": 4.9651718242557344e-05, + "loss": 3.2229, + "step": 4794 + }, + { + "epoch": 0.1479088024245322, + "grad_norm": 0.16943174600601196, + "learning_rate": 4.965127032796249e-05, + "loss": 3.2442, + "step": 4795 + }, + { + "epoch": 0.1479396488901056, + "grad_norm": 0.16282840073108673, + "learning_rate": 4.965082212755098e-05, + "loss": 3.2114, + "step": 4796 + }, + { + "epoch": 0.14797049535567902, + "grad_norm": 0.16453880071640015, + "learning_rate": 4.965037364132799e-05, + "loss": 3.2531, + "step": 4797 + }, + { + "epoch": 0.14800134182125244, + "grad_norm": 0.16105620563030243, + "learning_rate": 4.964992486929874e-05, + "loss": 3.2683, + "step": 4798 + }, + { + "epoch": 0.14803218828682585, + "grad_norm": 0.16390547156333923, + "learning_rate": 4.964947581146842e-05, + "loss": 3.2491, + "step": 4799 + }, + { + "epoch": 0.14806303475239926, + "grad_norm": 0.17444349825382233, + "learning_rate": 4.964902646784223e-05, + "loss": 3.2659, + "step": 4800 + }, + { + "epoch": 0.1480938812179727, + "grad_norm": 0.16603389382362366, + "learning_rate": 4.96485768384254e-05, + "loss": 3.2634, + "step": 4801 + }, + { + "epoch": 0.14812472768354612, + "grad_norm": 0.16233660280704498, + "learning_rate": 4.9648126923223135e-05, + "loss": 3.2669, + "step": 4802 + }, + { + "epoch": 0.14815557414911953, + "grad_norm": 0.1621933877468109, + "learning_rate": 4.964767672224065e-05, + "loss": 3.3026, + "step": 4803 + }, + { + "epoch": 0.14818642061469295, + "grad_norm": 0.16264520585536957, + "learning_rate": 4.964722623548316e-05, + "loss": 3.2531, + "step": 4804 + }, + { + "epoch": 0.14821726708026636, + "grad_norm": 0.16965407133102417, + "learning_rate": 4.96467754629559e-05, + "loss": 3.2638, + "step": 4805 + }, + { + "epoch": 0.14824811354583978, + "grad_norm": 0.16725698113441467, + "learning_rate": 4.964632440466409e-05, + "loss": 3.2776, + "step": 4806 + }, + { + "epoch": 0.1482789600114132, + "grad_norm": 0.17545932531356812, + "learning_rate": 4.964587306061295e-05, + "loss": 3.2619, + "step": 4807 + }, + { + "epoch": 0.1483098064769866, + "grad_norm": 0.1715695559978485, + "learning_rate": 4.964542143080773e-05, + "loss": 3.2802, + "step": 4808 + }, + { + "epoch": 0.14834065294256002, + "grad_norm": 0.1819733828306198, + "learning_rate": 4.964496951525366e-05, + "loss": 3.2956, + "step": 4809 + }, + { + "epoch": 0.14837149940813343, + "grad_norm": 0.16129861772060394, + "learning_rate": 4.9644517313955976e-05, + "loss": 3.2344, + "step": 4810 + }, + { + "epoch": 0.14840234587370685, + "grad_norm": 0.16703347861766815, + "learning_rate": 4.9644064826919916e-05, + "loss": 3.2669, + "step": 4811 + }, + { + "epoch": 0.14843319233928026, + "grad_norm": 0.16450734436511993, + "learning_rate": 4.964361205415074e-05, + "loss": 3.2065, + "step": 4812 + }, + { + "epoch": 0.1484640388048537, + "grad_norm": 0.1676148921251297, + "learning_rate": 4.964315899565369e-05, + "loss": 3.2613, + "step": 4813 + }, + { + "epoch": 0.14849488527042712, + "grad_norm": 0.15913426876068115, + "learning_rate": 4.9642705651434024e-05, + "loss": 3.2471, + "step": 4814 + }, + { + "epoch": 0.14852573173600053, + "grad_norm": 0.1730932891368866, + "learning_rate": 4.964225202149699e-05, + "loss": 3.2704, + "step": 4815 + }, + { + "epoch": 0.14855657820157395, + "grad_norm": 0.16108107566833496, + "learning_rate": 4.964179810584786e-05, + "loss": 3.2542, + "step": 4816 + }, + { + "epoch": 0.14858742466714736, + "grad_norm": 0.16446267068386078, + "learning_rate": 4.9641343904491874e-05, + "loss": 3.2746, + "step": 4817 + }, + { + "epoch": 0.14861827113272078, + "grad_norm": 0.15606912970542908, + "learning_rate": 4.964088941743433e-05, + "loss": 3.2582, + "step": 4818 + }, + { + "epoch": 0.1486491175982942, + "grad_norm": 0.16628825664520264, + "learning_rate": 4.964043464468047e-05, + "loss": 3.2872, + "step": 4819 + }, + { + "epoch": 0.1486799640638676, + "grad_norm": 0.16660381853580475, + "learning_rate": 4.963997958623558e-05, + "loss": 3.2737, + "step": 4820 + }, + { + "epoch": 0.14871081052944102, + "grad_norm": 0.18474750220775604, + "learning_rate": 4.963952424210494e-05, + "loss": 3.2505, + "step": 4821 + }, + { + "epoch": 0.14874165699501443, + "grad_norm": 0.16833369433879852, + "learning_rate": 4.963906861229381e-05, + "loss": 3.2452, + "step": 4822 + }, + { + "epoch": 0.14877250346058785, + "grad_norm": 0.16233254969120026, + "learning_rate": 4.9638612696807496e-05, + "loss": 3.2736, + "step": 4823 + }, + { + "epoch": 0.14880334992616126, + "grad_norm": 0.16467872262001038, + "learning_rate": 4.9638156495651265e-05, + "loss": 3.263, + "step": 4824 + }, + { + "epoch": 0.1488341963917347, + "grad_norm": 0.16409412026405334, + "learning_rate": 4.963770000883042e-05, + "loss": 3.2489, + "step": 4825 + }, + { + "epoch": 0.14886504285730812, + "grad_norm": 0.1620059460401535, + "learning_rate": 4.963724323635025e-05, + "loss": 3.2698, + "step": 4826 + }, + { + "epoch": 0.14889588932288153, + "grad_norm": 0.16235756874084473, + "learning_rate": 4.963678617821606e-05, + "loss": 3.2834, + "step": 4827 + }, + { + "epoch": 0.14892673578845494, + "grad_norm": 0.19505342841148376, + "learning_rate": 4.963632883443312e-05, + "loss": 3.2649, + "step": 4828 + }, + { + "epoch": 0.14895758225402836, + "grad_norm": 0.20234687626361847, + "learning_rate": 4.9635871205006765e-05, + "loss": 3.2947, + "step": 4829 + }, + { + "epoch": 0.14898842871960177, + "grad_norm": 0.16790929436683655, + "learning_rate": 4.9635413289942276e-05, + "loss": 3.2592, + "step": 4830 + }, + { + "epoch": 0.1490192751851752, + "grad_norm": 0.16684135794639587, + "learning_rate": 4.963495508924499e-05, + "loss": 3.2731, + "step": 4831 + }, + { + "epoch": 0.1490501216507486, + "grad_norm": 0.16655127704143524, + "learning_rate": 4.9634496602920196e-05, + "loss": 3.2545, + "step": 4832 + }, + { + "epoch": 0.14908096811632202, + "grad_norm": 0.16980183124542236, + "learning_rate": 4.963403783097321e-05, + "loss": 3.289, + "step": 4833 + }, + { + "epoch": 0.14911181458189543, + "grad_norm": 0.15877673029899597, + "learning_rate": 4.963357877340937e-05, + "loss": 3.2451, + "step": 4834 + }, + { + "epoch": 0.14914266104746884, + "grad_norm": 0.15135134756565094, + "learning_rate": 4.963311943023398e-05, + "loss": 3.28, + "step": 4835 + }, + { + "epoch": 0.14917350751304226, + "grad_norm": 0.16327962279319763, + "learning_rate": 4.9632659801452374e-05, + "loss": 3.2621, + "step": 4836 + }, + { + "epoch": 0.1492043539786157, + "grad_norm": 0.18689268827438354, + "learning_rate": 4.9632199887069883e-05, + "loss": 3.2638, + "step": 4837 + }, + { + "epoch": 0.14923520044418911, + "grad_norm": 0.16302859783172607, + "learning_rate": 4.963173968709184e-05, + "loss": 3.2852, + "step": 4838 + }, + { + "epoch": 0.14926604690976253, + "grad_norm": 0.15443575382232666, + "learning_rate": 4.9631279201523575e-05, + "loss": 3.2657, + "step": 4839 + }, + { + "epoch": 0.14929689337533594, + "grad_norm": 0.15963247418403625, + "learning_rate": 4.963081843037042e-05, + "loss": 3.2504, + "step": 4840 + }, + { + "epoch": 0.14932773984090936, + "grad_norm": 0.15946954488754272, + "learning_rate": 4.963035737363774e-05, + "loss": 3.2854, + "step": 4841 + }, + { + "epoch": 0.14935858630648277, + "grad_norm": 0.18052928149700165, + "learning_rate": 4.962989603133086e-05, + "loss": 3.2506, + "step": 4842 + }, + { + "epoch": 0.14938943277205619, + "grad_norm": 0.161152645945549, + "learning_rate": 4.962943440345513e-05, + "loss": 3.3035, + "step": 4843 + }, + { + "epoch": 0.1494202792376296, + "grad_norm": 0.15720172226428986, + "learning_rate": 4.9628972490015925e-05, + "loss": 3.3046, + "step": 4844 + }, + { + "epoch": 0.149451125703203, + "grad_norm": 0.16268287599086761, + "learning_rate": 4.9628510291018574e-05, + "loss": 3.2265, + "step": 4845 + }, + { + "epoch": 0.14948197216877643, + "grad_norm": 0.175294890999794, + "learning_rate": 4.9628047806468446e-05, + "loss": 3.2475, + "step": 4846 + }, + { + "epoch": 0.14951281863434984, + "grad_norm": 0.16569076478481293, + "learning_rate": 4.962758503637091e-05, + "loss": 3.2626, + "step": 4847 + }, + { + "epoch": 0.14954366509992326, + "grad_norm": 0.1650940328836441, + "learning_rate": 4.962712198073132e-05, + "loss": 3.2448, + "step": 4848 + }, + { + "epoch": 0.1495745115654967, + "grad_norm": 0.1648792177438736, + "learning_rate": 4.962665863955505e-05, + "loss": 3.2631, + "step": 4849 + }, + { + "epoch": 0.1496053580310701, + "grad_norm": 0.1626644730567932, + "learning_rate": 4.9626195012847466e-05, + "loss": 3.2617, + "step": 4850 + }, + { + "epoch": 0.14963620449664353, + "grad_norm": 0.1664755940437317, + "learning_rate": 4.962573110061396e-05, + "loss": 3.2711, + "step": 4851 + }, + { + "epoch": 0.14966705096221694, + "grad_norm": 0.1566053181886673, + "learning_rate": 4.9625266902859894e-05, + "loss": 3.3067, + "step": 4852 + }, + { + "epoch": 0.14969789742779036, + "grad_norm": 0.1621129810810089, + "learning_rate": 4.962480241959066e-05, + "loss": 3.2477, + "step": 4853 + }, + { + "epoch": 0.14972874389336377, + "grad_norm": 0.15930095314979553, + "learning_rate": 4.962433765081164e-05, + "loss": 3.2652, + "step": 4854 + }, + { + "epoch": 0.14975959035893718, + "grad_norm": 0.17287112772464752, + "learning_rate": 4.9623872596528206e-05, + "loss": 3.316, + "step": 4855 + }, + { + "epoch": 0.1497904368245106, + "grad_norm": 0.17251881957054138, + "learning_rate": 4.962340725674579e-05, + "loss": 3.2547, + "step": 4856 + }, + { + "epoch": 0.149821283290084, + "grad_norm": 0.16060978174209595, + "learning_rate": 4.962294163146976e-05, + "loss": 3.2613, + "step": 4857 + }, + { + "epoch": 0.14985212975565743, + "grad_norm": 0.161659374833107, + "learning_rate": 4.962247572070551e-05, + "loss": 3.2226, + "step": 4858 + }, + { + "epoch": 0.14988297622123084, + "grad_norm": 0.16655972599983215, + "learning_rate": 4.9622009524458445e-05, + "loss": 3.2873, + "step": 4859 + }, + { + "epoch": 0.14991382268680425, + "grad_norm": 0.17146266996860504, + "learning_rate": 4.962154304273399e-05, + "loss": 3.2601, + "step": 4860 + }, + { + "epoch": 0.1499446691523777, + "grad_norm": 0.15883347392082214, + "learning_rate": 4.962107627553753e-05, + "loss": 3.2924, + "step": 4861 + }, + { + "epoch": 0.1499755156179511, + "grad_norm": 0.16876061260700226, + "learning_rate": 4.962060922287449e-05, + "loss": 3.2668, + "step": 4862 + }, + { + "epoch": 0.15000636208352452, + "grad_norm": 0.16723060607910156, + "learning_rate": 4.962014188475028e-05, + "loss": 3.2782, + "step": 4863 + }, + { + "epoch": 0.15003720854909794, + "grad_norm": 0.1674908697605133, + "learning_rate": 4.961967426117032e-05, + "loss": 3.2649, + "step": 4864 + }, + { + "epoch": 0.15006805501467135, + "grad_norm": 0.16104605793952942, + "learning_rate": 4.9619206352140033e-05, + "loss": 3.2466, + "step": 4865 + }, + { + "epoch": 0.15009890148024477, + "grad_norm": 0.19596314430236816, + "learning_rate": 4.961873815766484e-05, + "loss": 3.2507, + "step": 4866 + }, + { + "epoch": 0.15012974794581818, + "grad_norm": 0.16101159155368805, + "learning_rate": 4.9618269677750175e-05, + "loss": 3.2865, + "step": 4867 + }, + { + "epoch": 0.1501605944113916, + "grad_norm": 0.15806853771209717, + "learning_rate": 4.9617800912401466e-05, + "loss": 3.2613, + "step": 4868 + }, + { + "epoch": 0.150191440876965, + "grad_norm": 0.1645890474319458, + "learning_rate": 4.961733186162415e-05, + "loss": 3.2652, + "step": 4869 + }, + { + "epoch": 0.15022228734253842, + "grad_norm": 0.16268663108348846, + "learning_rate": 4.961686252542366e-05, + "loss": 3.2927, + "step": 4870 + }, + { + "epoch": 0.15025313380811184, + "grad_norm": 0.17168381810188293, + "learning_rate": 4.9616392903805445e-05, + "loss": 3.252, + "step": 4871 + }, + { + "epoch": 0.15028398027368525, + "grad_norm": 0.16817320883274078, + "learning_rate": 4.961592299677495e-05, + "loss": 3.2873, + "step": 4872 + }, + { + "epoch": 0.1503148267392587, + "grad_norm": 0.18276207149028778, + "learning_rate": 4.9615452804337616e-05, + "loss": 3.2633, + "step": 4873 + }, + { + "epoch": 0.1503456732048321, + "grad_norm": 0.16038858890533447, + "learning_rate": 4.96149823264989e-05, + "loss": 3.2816, + "step": 4874 + }, + { + "epoch": 0.15037651967040552, + "grad_norm": 0.19128237664699554, + "learning_rate": 4.961451156326425e-05, + "loss": 3.2639, + "step": 4875 + }, + { + "epoch": 0.15040736613597894, + "grad_norm": 0.16235826909542084, + "learning_rate": 4.9614040514639146e-05, + "loss": 3.2352, + "step": 4876 + }, + { + "epoch": 0.15043821260155235, + "grad_norm": 0.1687467098236084, + "learning_rate": 4.961356918062902e-05, + "loss": 3.2683, + "step": 4877 + }, + { + "epoch": 0.15046905906712577, + "grad_norm": 0.16106100380420685, + "learning_rate": 4.9613097561239355e-05, + "loss": 3.259, + "step": 4878 + }, + { + "epoch": 0.15049990553269918, + "grad_norm": 0.1607065349817276, + "learning_rate": 4.9612625656475615e-05, + "loss": 3.2864, + "step": 4879 + }, + { + "epoch": 0.1505307519982726, + "grad_norm": 0.1560932993888855, + "learning_rate": 4.961215346634327e-05, + "loss": 3.2748, + "step": 4880 + }, + { + "epoch": 0.150561598463846, + "grad_norm": 0.16275466978549957, + "learning_rate": 4.9611680990847795e-05, + "loss": 3.2958, + "step": 4881 + }, + { + "epoch": 0.15059244492941942, + "grad_norm": 0.16200841963291168, + "learning_rate": 4.9611208229994665e-05, + "loss": 3.2459, + "step": 4882 + }, + { + "epoch": 0.15062329139499284, + "grad_norm": 0.16362279653549194, + "learning_rate": 4.961073518378937e-05, + "loss": 3.2815, + "step": 4883 + }, + { + "epoch": 0.15065413786056625, + "grad_norm": 0.16478024423122406, + "learning_rate": 4.96102618522374e-05, + "loss": 3.2708, + "step": 4884 + }, + { + "epoch": 0.1506849843261397, + "grad_norm": 0.17296183109283447, + "learning_rate": 4.960978823534421e-05, + "loss": 3.2691, + "step": 4885 + }, + { + "epoch": 0.1507158307917131, + "grad_norm": 0.17651128768920898, + "learning_rate": 4.960931433311533e-05, + "loss": 3.28, + "step": 4886 + }, + { + "epoch": 0.15074667725728652, + "grad_norm": 0.17361515760421753, + "learning_rate": 4.960884014555624e-05, + "loss": 3.249, + "step": 4887 + }, + { + "epoch": 0.15077752372285994, + "grad_norm": 0.16308118402957916, + "learning_rate": 4.960836567267243e-05, + "loss": 3.2693, + "step": 4888 + }, + { + "epoch": 0.15080837018843335, + "grad_norm": 0.16680626571178436, + "learning_rate": 4.960789091446941e-05, + "loss": 3.2775, + "step": 4889 + }, + { + "epoch": 0.15083921665400676, + "grad_norm": 0.1882774829864502, + "learning_rate": 4.960741587095268e-05, + "loss": 3.2348, + "step": 4890 + }, + { + "epoch": 0.15087006311958018, + "grad_norm": 0.1630047708749771, + "learning_rate": 4.9606940542127755e-05, + "loss": 3.2382, + "step": 4891 + }, + { + "epoch": 0.1509009095851536, + "grad_norm": 0.1660354882478714, + "learning_rate": 4.960646492800014e-05, + "loss": 3.25, + "step": 4892 + }, + { + "epoch": 0.150931756050727, + "grad_norm": 0.1599220186471939, + "learning_rate": 4.960598902857535e-05, + "loss": 3.253, + "step": 4893 + }, + { + "epoch": 0.15096260251630042, + "grad_norm": 0.175490602850914, + "learning_rate": 4.96055128438589e-05, + "loss": 3.2455, + "step": 4894 + }, + { + "epoch": 0.15099344898187383, + "grad_norm": 0.16041481494903564, + "learning_rate": 4.9605036373856316e-05, + "loss": 3.2622, + "step": 4895 + }, + { + "epoch": 0.15102429544744725, + "grad_norm": 0.1833489090204239, + "learning_rate": 4.960455961857313e-05, + "loss": 3.2646, + "step": 4896 + }, + { + "epoch": 0.1510551419130207, + "grad_norm": 0.1606464833021164, + "learning_rate": 4.9604082578014855e-05, + "loss": 3.2545, + "step": 4897 + }, + { + "epoch": 0.1510859883785941, + "grad_norm": 0.1604328453540802, + "learning_rate": 4.9603605252187025e-05, + "loss": 3.2569, + "step": 4898 + }, + { + "epoch": 0.15111683484416752, + "grad_norm": 0.16964644193649292, + "learning_rate": 4.960312764109518e-05, + "loss": 3.2613, + "step": 4899 + }, + { + "epoch": 0.15114768130974093, + "grad_norm": 0.16518700122833252, + "learning_rate": 4.9602649744744855e-05, + "loss": 3.2722, + "step": 4900 + }, + { + "epoch": 0.15117852777531435, + "grad_norm": 0.17594009637832642, + "learning_rate": 4.9602171563141586e-05, + "loss": 3.243, + "step": 4901 + }, + { + "epoch": 0.15120937424088776, + "grad_norm": 0.16555596888065338, + "learning_rate": 4.9601693096290926e-05, + "loss": 3.2364, + "step": 4902 + }, + { + "epoch": 0.15124022070646118, + "grad_norm": 0.1745307445526123, + "learning_rate": 4.960121434419842e-05, + "loss": 3.2455, + "step": 4903 + }, + { + "epoch": 0.1512710671720346, + "grad_norm": 0.16032159328460693, + "learning_rate": 4.960073530686962e-05, + "loss": 3.2761, + "step": 4904 + }, + { + "epoch": 0.151301913637608, + "grad_norm": 0.17860792577266693, + "learning_rate": 4.9600255984310076e-05, + "loss": 3.2489, + "step": 4905 + }, + { + "epoch": 0.15133276010318142, + "grad_norm": 0.16142180562019348, + "learning_rate": 4.9599776376525335e-05, + "loss": 3.2689, + "step": 4906 + }, + { + "epoch": 0.15136360656875483, + "grad_norm": 0.1651962250471115, + "learning_rate": 4.959929648352098e-05, + "loss": 3.2773, + "step": 4907 + }, + { + "epoch": 0.15139445303432825, + "grad_norm": 0.16024741530418396, + "learning_rate": 4.9598816305302564e-05, + "loss": 3.2549, + "step": 4908 + }, + { + "epoch": 0.1514252994999017, + "grad_norm": 0.1667727828025818, + "learning_rate": 4.959833584187565e-05, + "loss": 3.2439, + "step": 4909 + }, + { + "epoch": 0.1514561459654751, + "grad_norm": 0.16312876343727112, + "learning_rate": 4.9597855093245824e-05, + "loss": 3.2737, + "step": 4910 + }, + { + "epoch": 0.15148699243104852, + "grad_norm": 0.16414479911327362, + "learning_rate": 4.9597374059418646e-05, + "loss": 3.2614, + "step": 4911 + }, + { + "epoch": 0.15151783889662193, + "grad_norm": 0.16802190244197845, + "learning_rate": 4.959689274039969e-05, + "loss": 3.2844, + "step": 4912 + }, + { + "epoch": 0.15154868536219535, + "grad_norm": 0.1630798578262329, + "learning_rate": 4.959641113619455e-05, + "loss": 3.2336, + "step": 4913 + }, + { + "epoch": 0.15157953182776876, + "grad_norm": 0.159814715385437, + "learning_rate": 4.95959292468088e-05, + "loss": 3.261, + "step": 4914 + }, + { + "epoch": 0.15161037829334217, + "grad_norm": 0.16394798457622528, + "learning_rate": 4.959544707224803e-05, + "loss": 3.2688, + "step": 4915 + }, + { + "epoch": 0.1516412247589156, + "grad_norm": 0.1608845293521881, + "learning_rate": 4.959496461251784e-05, + "loss": 3.2727, + "step": 4916 + }, + { + "epoch": 0.151672071224489, + "grad_norm": 0.17372170090675354, + "learning_rate": 4.959448186762381e-05, + "loss": 3.2321, + "step": 4917 + }, + { + "epoch": 0.15170291769006242, + "grad_norm": 0.16373145580291748, + "learning_rate": 4.959399883757154e-05, + "loss": 3.2624, + "step": 4918 + }, + { + "epoch": 0.15173376415563583, + "grad_norm": 0.1731775552034378, + "learning_rate": 4.959351552236664e-05, + "loss": 3.2772, + "step": 4919 + }, + { + "epoch": 0.15176461062120924, + "grad_norm": 0.9140037894248962, + "learning_rate": 4.95930319220147e-05, + "loss": 3.2731, + "step": 4920 + }, + { + "epoch": 0.1517954570867827, + "grad_norm": 0.18629176914691925, + "learning_rate": 4.959254803652134e-05, + "loss": 3.2761, + "step": 4921 + }, + { + "epoch": 0.1518263035523561, + "grad_norm": 0.1629156917333603, + "learning_rate": 4.959206386589215e-05, + "loss": 3.2622, + "step": 4922 + }, + { + "epoch": 0.15185715001792952, + "grad_norm": 0.17113175988197327, + "learning_rate": 4.9591579410132774e-05, + "loss": 3.26, + "step": 4923 + }, + { + "epoch": 0.15188799648350293, + "grad_norm": 0.1798003762960434, + "learning_rate": 4.95910946692488e-05, + "loss": 3.2684, + "step": 4924 + }, + { + "epoch": 0.15191884294907634, + "grad_norm": 0.17467984557151794, + "learning_rate": 4.9590609643245864e-05, + "loss": 3.2766, + "step": 4925 + }, + { + "epoch": 0.15194968941464976, + "grad_norm": 0.16692452132701874, + "learning_rate": 4.959012433212959e-05, + "loss": 3.2547, + "step": 4926 + }, + { + "epoch": 0.15198053588022317, + "grad_norm": 0.17625364661216736, + "learning_rate": 4.95896387359056e-05, + "loss": 3.3058, + "step": 4927 + }, + { + "epoch": 0.1520113823457966, + "grad_norm": 0.164394810795784, + "learning_rate": 4.958915285457952e-05, + "loss": 3.2209, + "step": 4928 + }, + { + "epoch": 0.15204222881137, + "grad_norm": 0.1889483630657196, + "learning_rate": 4.9588666688157e-05, + "loss": 3.276, + "step": 4929 + }, + { + "epoch": 0.15207307527694341, + "grad_norm": 0.16135607659816742, + "learning_rate": 4.958818023664366e-05, + "loss": 3.2452, + "step": 4930 + }, + { + "epoch": 0.15210392174251683, + "grad_norm": 0.17044365406036377, + "learning_rate": 4.958769350004514e-05, + "loss": 3.2607, + "step": 4931 + }, + { + "epoch": 0.15213476820809024, + "grad_norm": 0.2044897824525833, + "learning_rate": 4.95872064783671e-05, + "loss": 3.2639, + "step": 4932 + }, + { + "epoch": 0.15216561467366368, + "grad_norm": 0.17938600480556488, + "learning_rate": 4.958671917161517e-05, + "loss": 3.2675, + "step": 4933 + }, + { + "epoch": 0.1521964611392371, + "grad_norm": 0.18962593376636505, + "learning_rate": 4.958623157979501e-05, + "loss": 3.2782, + "step": 4934 + }, + { + "epoch": 0.1522273076048105, + "grad_norm": 0.18420933187007904, + "learning_rate": 4.958574370291227e-05, + "loss": 3.2646, + "step": 4935 + }, + { + "epoch": 0.15225815407038393, + "grad_norm": 0.17158657312393188, + "learning_rate": 4.95852555409726e-05, + "loss": 3.2768, + "step": 4936 + }, + { + "epoch": 0.15228900053595734, + "grad_norm": 0.17865897715091705, + "learning_rate": 4.958476709398166e-05, + "loss": 3.2695, + "step": 4937 + }, + { + "epoch": 0.15231984700153076, + "grad_norm": 0.1741965115070343, + "learning_rate": 4.958427836194513e-05, + "loss": 3.2682, + "step": 4938 + }, + { + "epoch": 0.15235069346710417, + "grad_norm": 0.19694389402866364, + "learning_rate": 4.958378934486867e-05, + "loss": 3.2535, + "step": 4939 + }, + { + "epoch": 0.15238153993267758, + "grad_norm": 0.16845452785491943, + "learning_rate": 4.958330004275794e-05, + "loss": 3.2355, + "step": 4940 + }, + { + "epoch": 0.152412386398251, + "grad_norm": 0.2537862956523895, + "learning_rate": 4.958281045561862e-05, + "loss": 3.2973, + "step": 4941 + }, + { + "epoch": 0.1524432328638244, + "grad_norm": 0.16566653549671173, + "learning_rate": 4.958232058345638e-05, + "loss": 3.2802, + "step": 4942 + }, + { + "epoch": 0.15247407932939783, + "grad_norm": 0.17486628890037537, + "learning_rate": 4.958183042627691e-05, + "loss": 3.2829, + "step": 4943 + }, + { + "epoch": 0.15250492579497124, + "grad_norm": 0.18834872543811798, + "learning_rate": 4.9581339984085886e-05, + "loss": 3.2761, + "step": 4944 + }, + { + "epoch": 0.15253577226054468, + "grad_norm": 0.16853976249694824, + "learning_rate": 4.9580849256889e-05, + "loss": 3.2699, + "step": 4945 + }, + { + "epoch": 0.1525666187261181, + "grad_norm": 0.17501822113990784, + "learning_rate": 4.958035824469194e-05, + "loss": 3.2616, + "step": 4946 + }, + { + "epoch": 0.1525974651916915, + "grad_norm": 0.16598618030548096, + "learning_rate": 4.957986694750039e-05, + "loss": 3.2362, + "step": 4947 + }, + { + "epoch": 0.15262831165726493, + "grad_norm": 0.16809597611427307, + "learning_rate": 4.957937536532006e-05, + "loss": 3.2659, + "step": 4948 + }, + { + "epoch": 0.15265915812283834, + "grad_norm": 0.16271275281906128, + "learning_rate": 4.957888349815664e-05, + "loss": 3.2365, + "step": 4949 + }, + { + "epoch": 0.15269000458841175, + "grad_norm": 0.16921526193618774, + "learning_rate": 4.957839134601583e-05, + "loss": 3.2731, + "step": 4950 + }, + { + "epoch": 0.15272085105398517, + "grad_norm": 0.16587011516094208, + "learning_rate": 4.957789890890335e-05, + "loss": 3.2861, + "step": 4951 + }, + { + "epoch": 0.15275169751955858, + "grad_norm": 0.1650896519422531, + "learning_rate": 4.95774061868249e-05, + "loss": 3.2792, + "step": 4952 + }, + { + "epoch": 0.152782543985132, + "grad_norm": 0.17532193660736084, + "learning_rate": 4.957691317978619e-05, + "loss": 3.2248, + "step": 4953 + }, + { + "epoch": 0.1528133904507054, + "grad_norm": 0.1619632989168167, + "learning_rate": 4.9576419887792946e-05, + "loss": 3.2723, + "step": 4954 + }, + { + "epoch": 0.15284423691627883, + "grad_norm": 0.15695635974407196, + "learning_rate": 4.9575926310850885e-05, + "loss": 3.254, + "step": 4955 + }, + { + "epoch": 0.15287508338185227, + "grad_norm": 0.7661869525909424, + "learning_rate": 4.957543244896572e-05, + "loss": 3.2689, + "step": 4956 + }, + { + "epoch": 0.15290592984742568, + "grad_norm": 0.16421161592006683, + "learning_rate": 4.957493830214319e-05, + "loss": 3.2634, + "step": 4957 + }, + { + "epoch": 0.1529367763129991, + "grad_norm": 0.17198564112186432, + "learning_rate": 4.957444387038901e-05, + "loss": 3.248, + "step": 4958 + }, + { + "epoch": 0.1529676227785725, + "grad_norm": 0.22358699142932892, + "learning_rate": 4.957394915370893e-05, + "loss": 3.2828, + "step": 4959 + }, + { + "epoch": 0.15299846924414592, + "grad_norm": 0.17943261563777924, + "learning_rate": 4.957345415210867e-05, + "loss": 3.2683, + "step": 4960 + }, + { + "epoch": 0.15302931570971934, + "grad_norm": 0.16573187708854675, + "learning_rate": 4.957295886559398e-05, + "loss": 3.2588, + "step": 4961 + }, + { + "epoch": 0.15306016217529275, + "grad_norm": 0.17881207168102264, + "learning_rate": 4.95724632941706e-05, + "loss": 3.2464, + "step": 4962 + }, + { + "epoch": 0.15309100864086617, + "grad_norm": 0.17267073690891266, + "learning_rate": 4.957196743784427e-05, + "loss": 3.3143, + "step": 4963 + }, + { + "epoch": 0.15312185510643958, + "grad_norm": 0.17468969523906708, + "learning_rate": 4.957147129662074e-05, + "loss": 3.2886, + "step": 4964 + }, + { + "epoch": 0.153152701572013, + "grad_norm": 2.188491106033325, + "learning_rate": 4.9570974870505774e-05, + "loss": 3.3073, + "step": 4965 + }, + { + "epoch": 0.1531835480375864, + "grad_norm": 0.18727795779705048, + "learning_rate": 4.9570478159505115e-05, + "loss": 3.2608, + "step": 4966 + }, + { + "epoch": 0.15321439450315982, + "grad_norm": 0.3191048204898834, + "learning_rate": 4.9569981163624535e-05, + "loss": 3.2724, + "step": 4967 + }, + { + "epoch": 0.15324524096873327, + "grad_norm": 0.19630883634090424, + "learning_rate": 4.956948388286978e-05, + "loss": 3.2636, + "step": 4968 + }, + { + "epoch": 0.15327608743430668, + "grad_norm": 0.22655586898326874, + "learning_rate": 4.956898631724662e-05, + "loss": 3.2841, + "step": 4969 + }, + { + "epoch": 0.1533069338998801, + "grad_norm": 0.19971731305122375, + "learning_rate": 4.9568488466760845e-05, + "loss": 3.2426, + "step": 4970 + }, + { + "epoch": 0.1533377803654535, + "grad_norm": 0.1702352911233902, + "learning_rate": 4.95679903314182e-05, + "loss": 3.2577, + "step": 4971 + }, + { + "epoch": 0.15336862683102692, + "grad_norm": 0.18703612685203552, + "learning_rate": 4.956749191122446e-05, + "loss": 3.2781, + "step": 4972 + }, + { + "epoch": 0.15339947329660034, + "grad_norm": 0.18263591825962067, + "learning_rate": 4.956699320618543e-05, + "loss": 3.2916, + "step": 4973 + }, + { + "epoch": 0.15343031976217375, + "grad_norm": 0.19911512732505798, + "learning_rate": 4.956649421630687e-05, + "loss": 3.2981, + "step": 4974 + }, + { + "epoch": 0.15346116622774716, + "grad_norm": 0.1836591511964798, + "learning_rate": 4.956599494159457e-05, + "loss": 3.2702, + "step": 4975 + }, + { + "epoch": 0.15349201269332058, + "grad_norm": 0.18024536967277527, + "learning_rate": 4.956549538205433e-05, + "loss": 3.2438, + "step": 4976 + }, + { + "epoch": 0.153522859158894, + "grad_norm": 0.19936147332191467, + "learning_rate": 4.9564995537691924e-05, + "loss": 3.2595, + "step": 4977 + }, + { + "epoch": 0.1535537056244674, + "grad_norm": 0.1905021369457245, + "learning_rate": 4.9564495408513165e-05, + "loss": 3.2433, + "step": 4978 + }, + { + "epoch": 0.15358455209004082, + "grad_norm": 0.18673278391361237, + "learning_rate": 4.9563994994523845e-05, + "loss": 3.2507, + "step": 4979 + }, + { + "epoch": 0.15361539855561426, + "grad_norm": 0.19303405284881592, + "learning_rate": 4.9563494295729754e-05, + "loss": 3.2514, + "step": 4980 + }, + { + "epoch": 0.15364624502118768, + "grad_norm": 0.22284410893917084, + "learning_rate": 4.956299331213671e-05, + "loss": 3.2456, + "step": 4981 + }, + { + "epoch": 0.1536770914867611, + "grad_norm": 0.18351349234580994, + "learning_rate": 4.956249204375052e-05, + "loss": 3.271, + "step": 4982 + }, + { + "epoch": 0.1537079379523345, + "grad_norm": 0.1958288699388504, + "learning_rate": 4.9561990490577e-05, + "loss": 3.2853, + "step": 4983 + }, + { + "epoch": 0.15373878441790792, + "grad_norm": 0.18697895109653473, + "learning_rate": 4.9561488652621956e-05, + "loss": 3.2613, + "step": 4984 + }, + { + "epoch": 0.15376963088348133, + "grad_norm": 0.8752386569976807, + "learning_rate": 4.9560986529891205e-05, + "loss": 3.2591, + "step": 4985 + }, + { + "epoch": 0.15380047734905475, + "grad_norm": 0.3066278100013733, + "learning_rate": 4.956048412239058e-05, + "loss": 3.2469, + "step": 4986 + }, + { + "epoch": 0.15383132381462816, + "grad_norm": 0.2687603831291199, + "learning_rate": 4.955998143012591e-05, + "loss": 3.25, + "step": 4987 + }, + { + "epoch": 0.15386217028020158, + "grad_norm": 0.227837473154068, + "learning_rate": 4.9559478453103e-05, + "loss": 3.2616, + "step": 4988 + }, + { + "epoch": 0.153893016745775, + "grad_norm": 0.23618102073669434, + "learning_rate": 4.95589751913277e-05, + "loss": 3.2556, + "step": 4989 + }, + { + "epoch": 0.1539238632113484, + "grad_norm": 0.28624001145362854, + "learning_rate": 4.955847164480585e-05, + "loss": 3.2897, + "step": 4990 + }, + { + "epoch": 0.15395470967692182, + "grad_norm": 0.26215532422065735, + "learning_rate": 4.955796781354327e-05, + "loss": 3.2632, + "step": 4991 + }, + { + "epoch": 0.15398555614249526, + "grad_norm": 0.19647450745105743, + "learning_rate": 4.9557463697545815e-05, + "loss": 3.2987, + "step": 4992 + }, + { + "epoch": 0.15401640260806868, + "grad_norm": 0.21127091348171234, + "learning_rate": 4.955695929681932e-05, + "loss": 3.2551, + "step": 4993 + }, + { + "epoch": 0.1540472490736421, + "grad_norm": 0.24274970591068268, + "learning_rate": 4.955645461136964e-05, + "loss": 3.2777, + "step": 4994 + }, + { + "epoch": 0.1540780955392155, + "grad_norm": 0.21584510803222656, + "learning_rate": 4.955594964120263e-05, + "loss": 3.2428, + "step": 4995 + }, + { + "epoch": 0.15410894200478892, + "grad_norm": 0.19241192936897278, + "learning_rate": 4.955544438632414e-05, + "loss": 3.2801, + "step": 4996 + }, + { + "epoch": 0.15413978847036233, + "grad_norm": 0.1884203553199768, + "learning_rate": 4.955493884674003e-05, + "loss": 3.2475, + "step": 4997 + }, + { + "epoch": 0.15417063493593575, + "grad_norm": 0.19531120359897614, + "learning_rate": 4.955443302245615e-05, + "loss": 3.2486, + "step": 4998 + }, + { + "epoch": 0.15420148140150916, + "grad_norm": 0.2517591416835785, + "learning_rate": 4.9553926913478384e-05, + "loss": 3.236, + "step": 4999 + }, + { + "epoch": 0.15423232786708257, + "grad_norm": 0.23665404319763184, + "learning_rate": 4.955342051981259e-05, + "loss": 3.2508, + "step": 5000 + }, + { + "epoch": 0.154263174332656, + "grad_norm": 0.22400502860546112, + "learning_rate": 4.9552913841464634e-05, + "loss": 3.2761, + "step": 5001 + }, + { + "epoch": 0.1542940207982294, + "grad_norm": 0.21951726078987122, + "learning_rate": 4.955240687844041e-05, + "loss": 3.2583, + "step": 5002 + }, + { + "epoch": 0.15432486726380282, + "grad_norm": 0.20452192425727844, + "learning_rate": 4.955189963074576e-05, + "loss": 3.2396, + "step": 5003 + }, + { + "epoch": 0.15435571372937626, + "grad_norm": 0.19163084030151367, + "learning_rate": 4.95513920983866e-05, + "loss": 3.2731, + "step": 5004 + }, + { + "epoch": 0.15438656019494967, + "grad_norm": 0.20017342269420624, + "learning_rate": 4.9550884281368803e-05, + "loss": 3.2686, + "step": 5005 + }, + { + "epoch": 0.1544174066605231, + "grad_norm": 0.2891629934310913, + "learning_rate": 4.9550376179698264e-05, + "loss": 3.2734, + "step": 5006 + }, + { + "epoch": 0.1544482531260965, + "grad_norm": 0.21304449439048767, + "learning_rate": 4.954986779338086e-05, + "loss": 3.2564, + "step": 5007 + }, + { + "epoch": 0.15447909959166992, + "grad_norm": 0.2030935138463974, + "learning_rate": 4.954935912242248e-05, + "loss": 3.2798, + "step": 5008 + }, + { + "epoch": 0.15450994605724333, + "grad_norm": 0.17711767554283142, + "learning_rate": 4.954885016682905e-05, + "loss": 3.2738, + "step": 5009 + }, + { + "epoch": 0.15454079252281674, + "grad_norm": 0.1997876763343811, + "learning_rate": 4.9548340926606445e-05, + "loss": 3.2708, + "step": 5010 + }, + { + "epoch": 0.15457163898839016, + "grad_norm": 0.17561303079128265, + "learning_rate": 4.954783140176058e-05, + "loss": 3.2497, + "step": 5011 + }, + { + "epoch": 0.15460248545396357, + "grad_norm": 0.17869065701961517, + "learning_rate": 4.9547321592297366e-05, + "loss": 3.2737, + "step": 5012 + }, + { + "epoch": 0.154633331919537, + "grad_norm": 0.17286501824855804, + "learning_rate": 4.954681149822271e-05, + "loss": 3.276, + "step": 5013 + }, + { + "epoch": 0.1546641783851104, + "grad_norm": 0.16722184419631958, + "learning_rate": 4.954630111954252e-05, + "loss": 3.2878, + "step": 5014 + }, + { + "epoch": 0.15469502485068382, + "grad_norm": 0.17402136325836182, + "learning_rate": 4.954579045626272e-05, + "loss": 3.2603, + "step": 5015 + }, + { + "epoch": 0.15472587131625726, + "grad_norm": 0.16229982674121857, + "learning_rate": 4.954527950838923e-05, + "loss": 3.2908, + "step": 5016 + }, + { + "epoch": 0.15475671778183067, + "grad_norm": 0.17238011956214905, + "learning_rate": 4.9544768275927975e-05, + "loss": 3.2316, + "step": 5017 + }, + { + "epoch": 0.15478756424740409, + "grad_norm": 0.16625909507274628, + "learning_rate": 4.9544256758884884e-05, + "loss": 3.2273, + "step": 5018 + }, + { + "epoch": 0.1548184107129775, + "grad_norm": 0.262405127286911, + "learning_rate": 4.954374495726588e-05, + "loss": 3.272, + "step": 5019 + }, + { + "epoch": 0.15484925717855091, + "grad_norm": 0.16060039401054382, + "learning_rate": 4.9543232871076904e-05, + "loss": 3.2907, + "step": 5020 + }, + { + "epoch": 0.15488010364412433, + "grad_norm": 0.16547344624996185, + "learning_rate": 4.9542720500323894e-05, + "loss": 3.2745, + "step": 5021 + }, + { + "epoch": 0.15491095010969774, + "grad_norm": 0.1801033765077591, + "learning_rate": 4.954220784501279e-05, + "loss": 3.2266, + "step": 5022 + }, + { + "epoch": 0.15494179657527116, + "grad_norm": 0.2001509964466095, + "learning_rate": 4.954169490514953e-05, + "loss": 3.2487, + "step": 5023 + }, + { + "epoch": 0.15497264304084457, + "grad_norm": 0.2213430553674698, + "learning_rate": 4.9541181680740065e-05, + "loss": 3.2422, + "step": 5024 + }, + { + "epoch": 0.15500348950641799, + "grad_norm": 0.1790754497051239, + "learning_rate": 4.954066817179035e-05, + "loss": 3.2439, + "step": 5025 + }, + { + "epoch": 0.1550343359719914, + "grad_norm": 0.1730213463306427, + "learning_rate": 4.954015437830633e-05, + "loss": 3.2504, + "step": 5026 + }, + { + "epoch": 0.1550651824375648, + "grad_norm": 0.1672257035970688, + "learning_rate": 4.953964030029397e-05, + "loss": 3.2855, + "step": 5027 + }, + { + "epoch": 0.15509602890313826, + "grad_norm": 0.17408542335033417, + "learning_rate": 4.9539125937759225e-05, + "loss": 3.2474, + "step": 5028 + }, + { + "epoch": 0.15512687536871167, + "grad_norm": 0.16789494454860687, + "learning_rate": 4.953861129070807e-05, + "loss": 3.2416, + "step": 5029 + }, + { + "epoch": 0.15515772183428508, + "grad_norm": 0.17053617537021637, + "learning_rate": 4.953809635914645e-05, + "loss": 3.2192, + "step": 5030 + }, + { + "epoch": 0.1551885682998585, + "grad_norm": 0.17894123494625092, + "learning_rate": 4.953758114308036e-05, + "loss": 3.2343, + "step": 5031 + }, + { + "epoch": 0.1552194147654319, + "grad_norm": 0.18016429245471954, + "learning_rate": 4.953706564251575e-05, + "loss": 3.2866, + "step": 5032 + }, + { + "epoch": 0.15525026123100533, + "grad_norm": 0.2084878385066986, + "learning_rate": 4.953654985745861e-05, + "loss": 3.2759, + "step": 5033 + }, + { + "epoch": 0.15528110769657874, + "grad_norm": 0.18386776745319366, + "learning_rate": 4.953603378791493e-05, + "loss": 3.2458, + "step": 5034 + }, + { + "epoch": 0.15531195416215215, + "grad_norm": 0.1858201026916504, + "learning_rate": 4.9535517433890676e-05, + "loss": 3.227, + "step": 5035 + }, + { + "epoch": 0.15534280062772557, + "grad_norm": 0.17631526291370392, + "learning_rate": 4.953500079539185e-05, + "loss": 3.2627, + "step": 5036 + }, + { + "epoch": 0.15537364709329898, + "grad_norm": 0.16301394999027252, + "learning_rate": 4.953448387242442e-05, + "loss": 3.241, + "step": 5037 + }, + { + "epoch": 0.1554044935588724, + "grad_norm": 0.1751694679260254, + "learning_rate": 4.9533966664994404e-05, + "loss": 3.2837, + "step": 5038 + }, + { + "epoch": 0.1554353400244458, + "grad_norm": 0.1892436295747757, + "learning_rate": 4.953344917310778e-05, + "loss": 3.2306, + "step": 5039 + }, + { + "epoch": 0.15546618649001925, + "grad_norm": 0.17708417773246765, + "learning_rate": 4.9532931396770554e-05, + "loss": 3.2865, + "step": 5040 + }, + { + "epoch": 0.15549703295559267, + "grad_norm": 0.1847384124994278, + "learning_rate": 4.953241333598874e-05, + "loss": 3.2732, + "step": 5041 + }, + { + "epoch": 0.15552787942116608, + "grad_norm": 0.1631636768579483, + "learning_rate": 4.953189499076833e-05, + "loss": 3.2599, + "step": 5042 + }, + { + "epoch": 0.1555587258867395, + "grad_norm": 0.18079182505607605, + "learning_rate": 4.9531376361115345e-05, + "loss": 3.254, + "step": 5043 + }, + { + "epoch": 0.1555895723523129, + "grad_norm": 0.16709601879119873, + "learning_rate": 4.9530857447035786e-05, + "loss": 3.2805, + "step": 5044 + }, + { + "epoch": 0.15562041881788632, + "grad_norm": 0.1652534008026123, + "learning_rate": 4.953033824853568e-05, + "loss": 3.2528, + "step": 5045 + }, + { + "epoch": 0.15565126528345974, + "grad_norm": 0.16792792081832886, + "learning_rate": 4.952981876562104e-05, + "loss": 3.269, + "step": 5046 + }, + { + "epoch": 0.15568211174903315, + "grad_norm": 0.16769404709339142, + "learning_rate": 4.952929899829789e-05, + "loss": 3.2661, + "step": 5047 + }, + { + "epoch": 0.15571295821460657, + "grad_norm": 0.15537570416927338, + "learning_rate": 4.952877894657226e-05, + "loss": 3.2401, + "step": 5048 + }, + { + "epoch": 0.15574380468017998, + "grad_norm": 0.170246422290802, + "learning_rate": 4.9528258610450184e-05, + "loss": 3.2687, + "step": 5049 + }, + { + "epoch": 0.1557746511457534, + "grad_norm": 0.15983198583126068, + "learning_rate": 4.952773798993768e-05, + "loss": 3.2772, + "step": 5050 + }, + { + "epoch": 0.1558054976113268, + "grad_norm": 0.1641712635755539, + "learning_rate": 4.95272170850408e-05, + "loss": 3.2656, + "step": 5051 + }, + { + "epoch": 0.15583634407690025, + "grad_norm": 0.16032570600509644, + "learning_rate": 4.9526695895765586e-05, + "loss": 3.232, + "step": 5052 + }, + { + "epoch": 0.15586719054247367, + "grad_norm": 0.17078077793121338, + "learning_rate": 4.952617442211806e-05, + "loss": 3.2554, + "step": 5053 + }, + { + "epoch": 0.15589803700804708, + "grad_norm": 0.19367174804210663, + "learning_rate": 4.9525652664104284e-05, + "loss": 3.2335, + "step": 5054 + }, + { + "epoch": 0.1559288834736205, + "grad_norm": 0.16119880974292755, + "learning_rate": 4.95251306217303e-05, + "loss": 3.2464, + "step": 5055 + }, + { + "epoch": 0.1559597299391939, + "grad_norm": 0.16806229948997498, + "learning_rate": 4.952460829500217e-05, + "loss": 3.2836, + "step": 5056 + }, + { + "epoch": 0.15599057640476732, + "grad_norm": 0.17701968550682068, + "learning_rate": 4.952408568392593e-05, + "loss": 3.2373, + "step": 5057 + }, + { + "epoch": 0.15602142287034074, + "grad_norm": 0.16682659089565277, + "learning_rate": 4.952356278850767e-05, + "loss": 3.2524, + "step": 5058 + }, + { + "epoch": 0.15605226933591415, + "grad_norm": 0.17153021693229675, + "learning_rate": 4.952303960875343e-05, + "loss": 3.2699, + "step": 5059 + }, + { + "epoch": 0.15608311580148757, + "grad_norm": 0.16692857444286346, + "learning_rate": 4.9522516144669286e-05, + "loss": 3.2607, + "step": 5060 + }, + { + "epoch": 0.15611396226706098, + "grad_norm": 0.17432662844657898, + "learning_rate": 4.9521992396261294e-05, + "loss": 3.2331, + "step": 5061 + }, + { + "epoch": 0.1561448087326344, + "grad_norm": 0.17472466826438904, + "learning_rate": 4.952146836353555e-05, + "loss": 3.2442, + "step": 5062 + }, + { + "epoch": 0.1561756551982078, + "grad_norm": 0.16457226872444153, + "learning_rate": 4.9520944046498107e-05, + "loss": 3.2651, + "step": 5063 + }, + { + "epoch": 0.15620650166378125, + "grad_norm": 0.16888590157032013, + "learning_rate": 4.9520419445155055e-05, + "loss": 3.2907, + "step": 5064 + }, + { + "epoch": 0.15623734812935466, + "grad_norm": 0.1752483993768692, + "learning_rate": 4.9519894559512476e-05, + "loss": 3.2577, + "step": 5065 + }, + { + "epoch": 0.15626819459492808, + "grad_norm": 0.17124004662036896, + "learning_rate": 4.9519369389576454e-05, + "loss": 3.2385, + "step": 5066 + }, + { + "epoch": 0.1562990410605015, + "grad_norm": 0.16725899279117584, + "learning_rate": 4.9518843935353075e-05, + "loss": 3.2725, + "step": 5067 + }, + { + "epoch": 0.1563298875260749, + "grad_norm": 0.18012472987174988, + "learning_rate": 4.951831819684844e-05, + "loss": 3.2377, + "step": 5068 + }, + { + "epoch": 0.15636073399164832, + "grad_norm": 0.1674269437789917, + "learning_rate": 4.951779217406863e-05, + "loss": 3.2496, + "step": 5069 + }, + { + "epoch": 0.15639158045722173, + "grad_norm": 0.17372557520866394, + "learning_rate": 4.9517265867019766e-05, + "loss": 3.2462, + "step": 5070 + }, + { + "epoch": 0.15642242692279515, + "grad_norm": 0.1825776845216751, + "learning_rate": 4.951673927570793e-05, + "loss": 3.3062, + "step": 5071 + }, + { + "epoch": 0.15645327338836856, + "grad_norm": 0.4309515953063965, + "learning_rate": 4.9516212400139237e-05, + "loss": 3.2438, + "step": 5072 + }, + { + "epoch": 0.15648411985394198, + "grad_norm": 0.1652100831270218, + "learning_rate": 4.9515685240319785e-05, + "loss": 3.2557, + "step": 5073 + }, + { + "epoch": 0.1565149663195154, + "grad_norm": 0.17862007021903992, + "learning_rate": 4.9515157796255706e-05, + "loss": 3.2703, + "step": 5074 + }, + { + "epoch": 0.1565458127850888, + "grad_norm": 0.1701839417219162, + "learning_rate": 4.951463006795311e-05, + "loss": 3.273, + "step": 5075 + }, + { + "epoch": 0.15657665925066225, + "grad_norm": 0.19014954566955566, + "learning_rate": 4.951410205541809e-05, + "loss": 3.2733, + "step": 5076 + }, + { + "epoch": 0.15660750571623566, + "grad_norm": 0.16852445900440216, + "learning_rate": 4.951357375865681e-05, + "loss": 3.2656, + "step": 5077 + }, + { + "epoch": 0.15663835218180908, + "grad_norm": 0.1889410763978958, + "learning_rate": 4.9513045177675354e-05, + "loss": 3.2615, + "step": 5078 + }, + { + "epoch": 0.1566691986473825, + "grad_norm": 0.20511171221733093, + "learning_rate": 4.951251631247988e-05, + "loss": 3.2252, + "step": 5079 + }, + { + "epoch": 0.1567000451129559, + "grad_norm": 0.16845525801181793, + "learning_rate": 4.951198716307651e-05, + "loss": 3.233, + "step": 5080 + }, + { + "epoch": 0.15673089157852932, + "grad_norm": 0.19031542539596558, + "learning_rate": 4.951145772947137e-05, + "loss": 3.2384, + "step": 5081 + }, + { + "epoch": 0.15676173804410273, + "grad_norm": 0.16873422265052795, + "learning_rate": 4.9510928011670614e-05, + "loss": 3.2521, + "step": 5082 + }, + { + "epoch": 0.15679258450967615, + "grad_norm": 0.19277308881282806, + "learning_rate": 4.951039800968038e-05, + "loss": 3.2681, + "step": 5083 + }, + { + "epoch": 0.15682343097524956, + "grad_norm": 0.16925457119941711, + "learning_rate": 4.95098677235068e-05, + "loss": 3.275, + "step": 5084 + }, + { + "epoch": 0.15685427744082298, + "grad_norm": 0.16767023503780365, + "learning_rate": 4.9509337153156044e-05, + "loss": 3.2477, + "step": 5085 + }, + { + "epoch": 0.1568851239063964, + "grad_norm": 0.1733572632074356, + "learning_rate": 4.9508806298634245e-05, + "loss": 3.2807, + "step": 5086 + }, + { + "epoch": 0.1569159703719698, + "grad_norm": 0.17318566143512726, + "learning_rate": 4.950827515994757e-05, + "loss": 3.2614, + "step": 5087 + }, + { + "epoch": 0.15694681683754325, + "grad_norm": 0.17000605165958405, + "learning_rate": 4.950774373710217e-05, + "loss": 3.2713, + "step": 5088 + }, + { + "epoch": 0.15697766330311666, + "grad_norm": 0.17482033371925354, + "learning_rate": 4.9507212030104214e-05, + "loss": 3.2388, + "step": 5089 + }, + { + "epoch": 0.15700850976869007, + "grad_norm": 0.18365848064422607, + "learning_rate": 4.950668003895985e-05, + "loss": 3.2815, + "step": 5090 + }, + { + "epoch": 0.1570393562342635, + "grad_norm": 0.16443480551242828, + "learning_rate": 4.950614776367526e-05, + "loss": 3.2373, + "step": 5091 + }, + { + "epoch": 0.1570702026998369, + "grad_norm": 0.16289249062538147, + "learning_rate": 4.950561520425662e-05, + "loss": 3.2537, + "step": 5092 + }, + { + "epoch": 0.15710104916541032, + "grad_norm": 0.16816362738609314, + "learning_rate": 4.95050823607101e-05, + "loss": 3.2493, + "step": 5093 + }, + { + "epoch": 0.15713189563098373, + "grad_norm": 0.16007623076438904, + "learning_rate": 4.9504549233041865e-05, + "loss": 3.2525, + "step": 5094 + }, + { + "epoch": 0.15716274209655715, + "grad_norm": 0.15872062742710114, + "learning_rate": 4.950401582125811e-05, + "loss": 3.2534, + "step": 5095 + }, + { + "epoch": 0.15719358856213056, + "grad_norm": 0.15487977862358093, + "learning_rate": 4.950348212536502e-05, + "loss": 3.2577, + "step": 5096 + }, + { + "epoch": 0.15722443502770397, + "grad_norm": 0.1591082513332367, + "learning_rate": 4.9502948145368785e-05, + "loss": 3.2581, + "step": 5097 + }, + { + "epoch": 0.1572552814932774, + "grad_norm": 0.15923410654067993, + "learning_rate": 4.950241388127558e-05, + "loss": 3.2641, + "step": 5098 + }, + { + "epoch": 0.1572861279588508, + "grad_norm": 0.1674949824810028, + "learning_rate": 4.950187933309162e-05, + "loss": 3.2631, + "step": 5099 + }, + { + "epoch": 0.15731697442442424, + "grad_norm": 0.23836016654968262, + "learning_rate": 4.950134450082309e-05, + "loss": 3.2656, + "step": 5100 + }, + { + "epoch": 0.15734782088999766, + "grad_norm": 0.15854543447494507, + "learning_rate": 4.9500809384476186e-05, + "loss": 3.2677, + "step": 5101 + }, + { + "epoch": 0.15737866735557107, + "grad_norm": 0.1651347279548645, + "learning_rate": 4.950027398405713e-05, + "loss": 3.282, + "step": 5102 + }, + { + "epoch": 0.1574095138211445, + "grad_norm": 0.16426068544387817, + "learning_rate": 4.949973829957212e-05, + "loss": 3.2708, + "step": 5103 + }, + { + "epoch": 0.1574403602867179, + "grad_norm": 0.16260802745819092, + "learning_rate": 4.949920233102736e-05, + "loss": 3.2676, + "step": 5104 + }, + { + "epoch": 0.15747120675229132, + "grad_norm": 0.15914282202720642, + "learning_rate": 4.949866607842908e-05, + "loss": 3.2963, + "step": 5105 + }, + { + "epoch": 0.15750205321786473, + "grad_norm": 0.15798868238925934, + "learning_rate": 4.9498129541783475e-05, + "loss": 3.2856, + "step": 5106 + }, + { + "epoch": 0.15753289968343814, + "grad_norm": 0.15715812146663666, + "learning_rate": 4.94975927210968e-05, + "loss": 3.2589, + "step": 5107 + }, + { + "epoch": 0.15756374614901156, + "grad_norm": 0.15984931588172913, + "learning_rate": 4.9497055616375245e-05, + "loss": 3.2605, + "step": 5108 + }, + { + "epoch": 0.15759459261458497, + "grad_norm": 0.1595090627670288, + "learning_rate": 4.9496518227625046e-05, + "loss": 3.2399, + "step": 5109 + }, + { + "epoch": 0.15762543908015839, + "grad_norm": 0.1592245250940323, + "learning_rate": 4.949598055485245e-05, + "loss": 3.2353, + "step": 5110 + }, + { + "epoch": 0.1576562855457318, + "grad_norm": 0.15746371448040009, + "learning_rate": 4.949544259806368e-05, + "loss": 3.2645, + "step": 5111 + }, + { + "epoch": 0.15768713201130524, + "grad_norm": 0.1606309860944748, + "learning_rate": 4.949490435726497e-05, + "loss": 3.2567, + "step": 5112 + }, + { + "epoch": 0.15771797847687866, + "grad_norm": 0.1599702388048172, + "learning_rate": 4.949436583246257e-05, + "loss": 3.2218, + "step": 5113 + }, + { + "epoch": 0.15774882494245207, + "grad_norm": 0.16376841068267822, + "learning_rate": 4.9493827023662696e-05, + "loss": 3.2705, + "step": 5114 + }, + { + "epoch": 0.15777967140802548, + "grad_norm": 0.1579059362411499, + "learning_rate": 4.949328793087163e-05, + "loss": 3.2524, + "step": 5115 + }, + { + "epoch": 0.1578105178735989, + "grad_norm": 0.16295067965984344, + "learning_rate": 4.9492748554095614e-05, + "loss": 3.2559, + "step": 5116 + }, + { + "epoch": 0.1578413643391723, + "grad_norm": 0.15238815546035767, + "learning_rate": 4.94922088933409e-05, + "loss": 3.27, + "step": 5117 + }, + { + "epoch": 0.15787221080474573, + "grad_norm": 0.17440186440944672, + "learning_rate": 4.949166894861373e-05, + "loss": 3.2504, + "step": 5118 + }, + { + "epoch": 0.15790305727031914, + "grad_norm": 0.16455793380737305, + "learning_rate": 4.949112871992039e-05, + "loss": 3.2712, + "step": 5119 + }, + { + "epoch": 0.15793390373589256, + "grad_norm": 0.17487560212612152, + "learning_rate": 4.9490588207267116e-05, + "loss": 3.265, + "step": 5120 + }, + { + "epoch": 0.15796475020146597, + "grad_norm": 0.16060605645179749, + "learning_rate": 4.94900474106602e-05, + "loss": 3.2525, + "step": 5121 + }, + { + "epoch": 0.15799559666703938, + "grad_norm": 0.16163669526576996, + "learning_rate": 4.9489506330105894e-05, + "loss": 3.244, + "step": 5122 + }, + { + "epoch": 0.1580264431326128, + "grad_norm": 0.24013714492321014, + "learning_rate": 4.948896496561048e-05, + "loss": 3.2458, + "step": 5123 + }, + { + "epoch": 0.15805728959818624, + "grad_norm": 0.16084490716457367, + "learning_rate": 4.948842331718023e-05, + "loss": 3.2654, + "step": 5124 + }, + { + "epoch": 0.15808813606375965, + "grad_norm": 0.1647738367319107, + "learning_rate": 4.948788138482143e-05, + "loss": 3.242, + "step": 5125 + }, + { + "epoch": 0.15811898252933307, + "grad_norm": 0.17829729616641998, + "learning_rate": 4.948733916854036e-05, + "loss": 3.2359, + "step": 5126 + }, + { + "epoch": 0.15814982899490648, + "grad_norm": 0.16709645092487335, + "learning_rate": 4.9486796668343315e-05, + "loss": 3.2679, + "step": 5127 + }, + { + "epoch": 0.1581806754604799, + "grad_norm": 0.16874200105667114, + "learning_rate": 4.948625388423657e-05, + "loss": 3.2679, + "step": 5128 + }, + { + "epoch": 0.1582115219260533, + "grad_norm": 0.1753886640071869, + "learning_rate": 4.948571081622643e-05, + "loss": 3.2842, + "step": 5129 + }, + { + "epoch": 0.15824236839162673, + "grad_norm": 0.1543329358100891, + "learning_rate": 4.9485167464319184e-05, + "loss": 3.244, + "step": 5130 + }, + { + "epoch": 0.15827321485720014, + "grad_norm": 0.17394110560417175, + "learning_rate": 4.948462382852114e-05, + "loss": 3.2417, + "step": 5131 + }, + { + "epoch": 0.15830406132277355, + "grad_norm": 0.1595219522714615, + "learning_rate": 4.948407990883859e-05, + "loss": 3.2308, + "step": 5132 + }, + { + "epoch": 0.15833490778834697, + "grad_norm": 0.17002761363983154, + "learning_rate": 4.948353570527785e-05, + "loss": 3.2694, + "step": 5133 + }, + { + "epoch": 0.15836575425392038, + "grad_norm": 0.17001476883888245, + "learning_rate": 4.948299121784523e-05, + "loss": 3.258, + "step": 5134 + }, + { + "epoch": 0.1583966007194938, + "grad_norm": 0.17998187243938446, + "learning_rate": 4.9482446446547033e-05, + "loss": 3.291, + "step": 5135 + }, + { + "epoch": 0.15842744718506724, + "grad_norm": 0.16716520488262177, + "learning_rate": 4.9481901391389585e-05, + "loss": 3.2361, + "step": 5136 + }, + { + "epoch": 0.15845829365064065, + "grad_norm": 0.2178860306739807, + "learning_rate": 4.94813560523792e-05, + "loss": 3.2544, + "step": 5137 + }, + { + "epoch": 0.15848914011621407, + "grad_norm": 0.1802007555961609, + "learning_rate": 4.94808104295222e-05, + "loss": 3.2439, + "step": 5138 + }, + { + "epoch": 0.15851998658178748, + "grad_norm": 0.15786397457122803, + "learning_rate": 4.948026452282492e-05, + "loss": 3.2173, + "step": 5139 + }, + { + "epoch": 0.1585508330473609, + "grad_norm": 0.17490628361701965, + "learning_rate": 4.947971833229368e-05, + "loss": 3.2585, + "step": 5140 + }, + { + "epoch": 0.1585816795129343, + "grad_norm": 0.16424062848091125, + "learning_rate": 4.947917185793482e-05, + "loss": 3.2902, + "step": 5141 + }, + { + "epoch": 0.15861252597850772, + "grad_norm": 0.17739297449588776, + "learning_rate": 4.947862509975467e-05, + "loss": 3.2368, + "step": 5142 + }, + { + "epoch": 0.15864337244408114, + "grad_norm": 0.16937091946601868, + "learning_rate": 4.947807805775958e-05, + "loss": 3.2331, + "step": 5143 + }, + { + "epoch": 0.15867421890965455, + "grad_norm": 0.16198143362998962, + "learning_rate": 4.9477530731955875e-05, + "loss": 3.23, + "step": 5144 + }, + { + "epoch": 0.15870506537522797, + "grad_norm": 0.17157551646232605, + "learning_rate": 4.947698312234991e-05, + "loss": 3.2636, + "step": 5145 + }, + { + "epoch": 0.15873591184080138, + "grad_norm": 0.19790445268154144, + "learning_rate": 4.9476435228948034e-05, + "loss": 3.2816, + "step": 5146 + }, + { + "epoch": 0.1587667583063748, + "grad_norm": 0.1713201105594635, + "learning_rate": 4.947588705175661e-05, + "loss": 3.2614, + "step": 5147 + }, + { + "epoch": 0.15879760477194824, + "grad_norm": 0.16260211169719696, + "learning_rate": 4.9475338590781975e-05, + "loss": 3.2289, + "step": 5148 + }, + { + "epoch": 0.15882845123752165, + "grad_norm": 0.17121002078056335, + "learning_rate": 4.94747898460305e-05, + "loss": 3.2603, + "step": 5149 + }, + { + "epoch": 0.15885929770309506, + "grad_norm": 0.17171815037727356, + "learning_rate": 4.947424081750855e-05, + "loss": 3.2223, + "step": 5150 + }, + { + "epoch": 0.15889014416866848, + "grad_norm": 0.16816698014736176, + "learning_rate": 4.9473691505222474e-05, + "loss": 3.2525, + "step": 5151 + }, + { + "epoch": 0.1589209906342419, + "grad_norm": 0.16709205508232117, + "learning_rate": 4.947314190917866e-05, + "loss": 3.2526, + "step": 5152 + }, + { + "epoch": 0.1589518370998153, + "grad_norm": 0.16169977188110352, + "learning_rate": 4.947259202938347e-05, + "loss": 3.2518, + "step": 5153 + }, + { + "epoch": 0.15898268356538872, + "grad_norm": 0.16118082404136658, + "learning_rate": 4.947204186584328e-05, + "loss": 3.2294, + "step": 5154 + }, + { + "epoch": 0.15901353003096214, + "grad_norm": 0.1581030786037445, + "learning_rate": 4.947149141856447e-05, + "loss": 3.2943, + "step": 5155 + }, + { + "epoch": 0.15904437649653555, + "grad_norm": 0.16606515645980835, + "learning_rate": 4.947094068755343e-05, + "loss": 3.2703, + "step": 5156 + }, + { + "epoch": 0.15907522296210896, + "grad_norm": 0.16308431327342987, + "learning_rate": 4.947038967281653e-05, + "loss": 3.2847, + "step": 5157 + }, + { + "epoch": 0.15910606942768238, + "grad_norm": 0.16027168929576874, + "learning_rate": 4.946983837436017e-05, + "loss": 3.2681, + "step": 5158 + }, + { + "epoch": 0.1591369158932558, + "grad_norm": 0.1627826988697052, + "learning_rate": 4.946928679219074e-05, + "loss": 3.2311, + "step": 5159 + }, + { + "epoch": 0.15916776235882923, + "grad_norm": 0.17343318462371826, + "learning_rate": 4.946873492631463e-05, + "loss": 3.2597, + "step": 5160 + }, + { + "epoch": 0.15919860882440265, + "grad_norm": 0.15421143174171448, + "learning_rate": 4.9468182776738245e-05, + "loss": 3.2307, + "step": 5161 + }, + { + "epoch": 0.15922945528997606, + "grad_norm": 0.16953212022781372, + "learning_rate": 4.946763034346799e-05, + "loss": 3.2523, + "step": 5162 + }, + { + "epoch": 0.15926030175554948, + "grad_norm": 0.15537965297698975, + "learning_rate": 4.946707762651025e-05, + "loss": 3.2434, + "step": 5163 + }, + { + "epoch": 0.1592911482211229, + "grad_norm": 0.16437523066997528, + "learning_rate": 4.9466524625871456e-05, + "loss": 3.2693, + "step": 5164 + }, + { + "epoch": 0.1593219946866963, + "grad_norm": 0.15914316475391388, + "learning_rate": 4.946597134155802e-05, + "loss": 3.2434, + "step": 5165 + }, + { + "epoch": 0.15935284115226972, + "grad_norm": 0.2008916288614273, + "learning_rate": 4.946541777357634e-05, + "loss": 3.2083, + "step": 5166 + }, + { + "epoch": 0.15938368761784313, + "grad_norm": 0.1623932421207428, + "learning_rate": 4.946486392193284e-05, + "loss": 3.2332, + "step": 5167 + }, + { + "epoch": 0.15941453408341655, + "grad_norm": 0.16488276422023773, + "learning_rate": 4.946430978663395e-05, + "loss": 3.276, + "step": 5168 + }, + { + "epoch": 0.15944538054898996, + "grad_norm": 0.16334927082061768, + "learning_rate": 4.946375536768608e-05, + "loss": 3.2736, + "step": 5169 + }, + { + "epoch": 0.15947622701456338, + "grad_norm": 0.16385987401008606, + "learning_rate": 4.946320066509568e-05, + "loss": 3.2577, + "step": 5170 + }, + { + "epoch": 0.1595070734801368, + "grad_norm": 0.15962348878383636, + "learning_rate": 4.946264567886917e-05, + "loss": 3.2453, + "step": 5171 + }, + { + "epoch": 0.15953791994571023, + "grad_norm": 0.16855120658874512, + "learning_rate": 4.9462090409012974e-05, + "loss": 3.2205, + "step": 5172 + }, + { + "epoch": 0.15956876641128365, + "grad_norm": 0.15675127506256104, + "learning_rate": 4.946153485553354e-05, + "loss": 3.2644, + "step": 5173 + }, + { + "epoch": 0.15959961287685706, + "grad_norm": 0.15664103627204895, + "learning_rate": 4.9460979018437314e-05, + "loss": 3.245, + "step": 5174 + }, + { + "epoch": 0.15963045934243048, + "grad_norm": 0.15808463096618652, + "learning_rate": 4.9460422897730734e-05, + "loss": 3.2846, + "step": 5175 + }, + { + "epoch": 0.1596613058080039, + "grad_norm": 0.15595227479934692, + "learning_rate": 4.945986649342025e-05, + "loss": 3.2339, + "step": 5176 + }, + { + "epoch": 0.1596921522735773, + "grad_norm": 0.19584767520427704, + "learning_rate": 4.945930980551231e-05, + "loss": 3.244, + "step": 5177 + }, + { + "epoch": 0.15972299873915072, + "grad_norm": 0.15890933573246002, + "learning_rate": 4.945875283401338e-05, + "loss": 3.2499, + "step": 5178 + }, + { + "epoch": 0.15975384520472413, + "grad_norm": 0.157992884516716, + "learning_rate": 4.94581955789299e-05, + "loss": 3.2435, + "step": 5179 + }, + { + "epoch": 0.15978469167029755, + "grad_norm": 0.16019250452518463, + "learning_rate": 4.945763804026834e-05, + "loss": 3.2316, + "step": 5180 + }, + { + "epoch": 0.15981553813587096, + "grad_norm": 0.16065867245197296, + "learning_rate": 4.945708021803517e-05, + "loss": 3.256, + "step": 5181 + }, + { + "epoch": 0.15984638460144437, + "grad_norm": 0.15561993420124054, + "learning_rate": 4.945652211223685e-05, + "loss": 3.2591, + "step": 5182 + }, + { + "epoch": 0.1598772310670178, + "grad_norm": 0.15632303059101105, + "learning_rate": 4.945596372287985e-05, + "loss": 3.2543, + "step": 5183 + }, + { + "epoch": 0.15990807753259123, + "grad_norm": 0.18018019199371338, + "learning_rate": 4.945540504997064e-05, + "loss": 3.2563, + "step": 5184 + }, + { + "epoch": 0.15993892399816464, + "grad_norm": 0.15954814851284027, + "learning_rate": 4.9454846093515716e-05, + "loss": 3.2541, + "step": 5185 + }, + { + "epoch": 0.15996977046373806, + "grad_norm": 0.15694363415241241, + "learning_rate": 4.9454286853521545e-05, + "loss": 3.2563, + "step": 5186 + }, + { + "epoch": 0.16000061692931147, + "grad_norm": 0.15912853181362152, + "learning_rate": 4.9453727329994615e-05, + "loss": 3.2802, + "step": 5187 + }, + { + "epoch": 0.1600314633948849, + "grad_norm": 0.1535356491804123, + "learning_rate": 4.9453167522941403e-05, + "loss": 3.2193, + "step": 5188 + }, + { + "epoch": 0.1600623098604583, + "grad_norm": 0.158924400806427, + "learning_rate": 4.9452607432368416e-05, + "loss": 3.2421, + "step": 5189 + }, + { + "epoch": 0.16009315632603172, + "grad_norm": 0.1614316999912262, + "learning_rate": 4.945204705828214e-05, + "loss": 3.2543, + "step": 5190 + }, + { + "epoch": 0.16012400279160513, + "grad_norm": 0.1578250676393509, + "learning_rate": 4.9451486400689065e-05, + "loss": 3.2562, + "step": 5191 + }, + { + "epoch": 0.16015484925717854, + "grad_norm": 0.15800388157367706, + "learning_rate": 4.945092545959571e-05, + "loss": 3.2677, + "step": 5192 + }, + { + "epoch": 0.16018569572275196, + "grad_norm": 0.16637781262397766, + "learning_rate": 4.9450364235008564e-05, + "loss": 3.2298, + "step": 5193 + }, + { + "epoch": 0.16021654218832537, + "grad_norm": 0.15763193368911743, + "learning_rate": 4.944980272693413e-05, + "loss": 3.2332, + "step": 5194 + }, + { + "epoch": 0.1602473886538988, + "grad_norm": 0.1618143618106842, + "learning_rate": 4.9449240935378935e-05, + "loss": 3.2546, + "step": 5195 + }, + { + "epoch": 0.16027823511947223, + "grad_norm": 0.16066136956214905, + "learning_rate": 4.944867886034947e-05, + "loss": 3.2465, + "step": 5196 + }, + { + "epoch": 0.16030908158504564, + "grad_norm": 0.49417954683303833, + "learning_rate": 4.9448116501852285e-05, + "loss": 3.24, + "step": 5197 + }, + { + "epoch": 0.16033992805061906, + "grad_norm": 0.17687663435935974, + "learning_rate": 4.944755385989387e-05, + "loss": 3.2088, + "step": 5198 + }, + { + "epoch": 0.16037077451619247, + "grad_norm": 0.18198779225349426, + "learning_rate": 4.944699093448076e-05, + "loss": 3.2733, + "step": 5199 + }, + { + "epoch": 0.16040162098176589, + "grad_norm": 0.16597862541675568, + "learning_rate": 4.9446427725619484e-05, + "loss": 3.2379, + "step": 5200 + }, + { + "epoch": 0.1604324674473393, + "grad_norm": 0.17662928998470306, + "learning_rate": 4.944586423331656e-05, + "loss": 3.2795, + "step": 5201 + }, + { + "epoch": 0.16046331391291271, + "grad_norm": 0.17082589864730835, + "learning_rate": 4.9445300457578534e-05, + "loss": 3.2653, + "step": 5202 + }, + { + "epoch": 0.16049416037848613, + "grad_norm": 0.1631575971841812, + "learning_rate": 4.944473639841195e-05, + "loss": 3.2581, + "step": 5203 + }, + { + "epoch": 0.16052500684405954, + "grad_norm": 0.17260149121284485, + "learning_rate": 4.944417205582332e-05, + "loss": 3.2448, + "step": 5204 + }, + { + "epoch": 0.16055585330963296, + "grad_norm": 0.1642199605703354, + "learning_rate": 4.944360742981922e-05, + "loss": 3.2557, + "step": 5205 + }, + { + "epoch": 0.16058669977520637, + "grad_norm": 0.17572905123233795, + "learning_rate": 4.9443042520406176e-05, + "loss": 3.2375, + "step": 5206 + }, + { + "epoch": 0.16061754624077978, + "grad_norm": 0.18026100099086761, + "learning_rate": 4.944247732759074e-05, + "loss": 3.2527, + "step": 5207 + }, + { + "epoch": 0.16064839270635323, + "grad_norm": 0.16720978915691376, + "learning_rate": 4.944191185137947e-05, + "loss": 3.2698, + "step": 5208 + }, + { + "epoch": 0.16067923917192664, + "grad_norm": 0.16229526698589325, + "learning_rate": 4.9441346091778916e-05, + "loss": 3.2538, + "step": 5209 + }, + { + "epoch": 0.16071008563750006, + "grad_norm": 0.16166798770427704, + "learning_rate": 4.9440780048795645e-05, + "loss": 3.2643, + "step": 5210 + }, + { + "epoch": 0.16074093210307347, + "grad_norm": 0.16378483176231384, + "learning_rate": 4.9440213722436214e-05, + "loss": 3.2311, + "step": 5211 + }, + { + "epoch": 0.16077177856864688, + "grad_norm": 0.1639784872531891, + "learning_rate": 4.9439647112707195e-05, + "loss": 3.2638, + "step": 5212 + }, + { + "epoch": 0.1608026250342203, + "grad_norm": 0.1637689769268036, + "learning_rate": 4.943908021961515e-05, + "loss": 3.2516, + "step": 5213 + }, + { + "epoch": 0.1608334714997937, + "grad_norm": 0.16160638630390167, + "learning_rate": 4.943851304316666e-05, + "loss": 3.2871, + "step": 5214 + }, + { + "epoch": 0.16086431796536713, + "grad_norm": 0.15892310440540314, + "learning_rate": 4.9437945583368294e-05, + "loss": 3.2227, + "step": 5215 + }, + { + "epoch": 0.16089516443094054, + "grad_norm": 0.16828978061676025, + "learning_rate": 4.943737784022664e-05, + "loss": 3.2691, + "step": 5216 + }, + { + "epoch": 0.16092601089651395, + "grad_norm": 0.1631450355052948, + "learning_rate": 4.943680981374827e-05, + "loss": 3.2622, + "step": 5217 + }, + { + "epoch": 0.16095685736208737, + "grad_norm": 0.1671726107597351, + "learning_rate": 4.943624150393977e-05, + "loss": 3.2482, + "step": 5218 + }, + { + "epoch": 0.16098770382766078, + "grad_norm": 0.1602620631456375, + "learning_rate": 4.9435672910807754e-05, + "loss": 3.2658, + "step": 5219 + }, + { + "epoch": 0.16101855029323422, + "grad_norm": 0.17207834124565125, + "learning_rate": 4.943510403435877e-05, + "loss": 3.2897, + "step": 5220 + }, + { + "epoch": 0.16104939675880764, + "grad_norm": 0.16594988107681274, + "learning_rate": 4.9434534874599454e-05, + "loss": 3.205, + "step": 5221 + }, + { + "epoch": 0.16108024322438105, + "grad_norm": 0.15841303765773773, + "learning_rate": 4.943396543153639e-05, + "loss": 3.2554, + "step": 5222 + }, + { + "epoch": 0.16111108968995447, + "grad_norm": 0.16177067160606384, + "learning_rate": 4.943339570517617e-05, + "loss": 3.2876, + "step": 5223 + }, + { + "epoch": 0.16114193615552788, + "grad_norm": 0.17131003737449646, + "learning_rate": 4.943282569552541e-05, + "loss": 3.2613, + "step": 5224 + }, + { + "epoch": 0.1611727826211013, + "grad_norm": 0.16342315077781677, + "learning_rate": 4.943225540259073e-05, + "loss": 3.2502, + "step": 5225 + }, + { + "epoch": 0.1612036290866747, + "grad_norm": 0.17191362380981445, + "learning_rate": 4.943168482637872e-05, + "loss": 3.2553, + "step": 5226 + }, + { + "epoch": 0.16123447555224812, + "grad_norm": 0.15714429318904877, + "learning_rate": 4.9431113966896014e-05, + "loss": 3.2397, + "step": 5227 + }, + { + "epoch": 0.16126532201782154, + "grad_norm": 0.1712082177400589, + "learning_rate": 4.9430542824149216e-05, + "loss": 3.2331, + "step": 5228 + }, + { + "epoch": 0.16129616848339495, + "grad_norm": 0.17659780383110046, + "learning_rate": 4.942997139814496e-05, + "loss": 3.2287, + "step": 5229 + }, + { + "epoch": 0.16132701494896837, + "grad_norm": 0.1648412048816681, + "learning_rate": 4.9429399688889865e-05, + "loss": 3.2577, + "step": 5230 + }, + { + "epoch": 0.1613578614145418, + "grad_norm": 0.17965780198574066, + "learning_rate": 4.9428827696390554e-05, + "loss": 3.25, + "step": 5231 + }, + { + "epoch": 0.16138870788011522, + "grad_norm": 0.1625843346118927, + "learning_rate": 4.942825542065368e-05, + "loss": 3.2465, + "step": 5232 + }, + { + "epoch": 0.16141955434568864, + "grad_norm": 0.16788439452648163, + "learning_rate": 4.942768286168585e-05, + "loss": 3.2345, + "step": 5233 + }, + { + "epoch": 0.16145040081126205, + "grad_norm": 0.16154329478740692, + "learning_rate": 4.9427110019493725e-05, + "loss": 3.2772, + "step": 5234 + }, + { + "epoch": 0.16148124727683547, + "grad_norm": 0.175032839179039, + "learning_rate": 4.9426536894083944e-05, + "loss": 3.2678, + "step": 5235 + }, + { + "epoch": 0.16151209374240888, + "grad_norm": 0.15937110781669617, + "learning_rate": 4.9425963485463133e-05, + "loss": 3.2697, + "step": 5236 + }, + { + "epoch": 0.1615429402079823, + "grad_norm": 0.17241647839546204, + "learning_rate": 4.942538979363796e-05, + "loss": 3.2512, + "step": 5237 + }, + { + "epoch": 0.1615737866735557, + "grad_norm": 0.16168391704559326, + "learning_rate": 4.942481581861507e-05, + "loss": 3.2593, + "step": 5238 + }, + { + "epoch": 0.16160463313912912, + "grad_norm": 0.16236217319965363, + "learning_rate": 4.942424156040112e-05, + "loss": 3.2644, + "step": 5239 + }, + { + "epoch": 0.16163547960470254, + "grad_norm": 0.163023442029953, + "learning_rate": 4.942366701900276e-05, + "loss": 3.2749, + "step": 5240 + }, + { + "epoch": 0.16166632607027595, + "grad_norm": 0.15920129418373108, + "learning_rate": 4.9423092194426665e-05, + "loss": 3.2554, + "step": 5241 + }, + { + "epoch": 0.16169717253584937, + "grad_norm": 0.166967511177063, + "learning_rate": 4.942251708667949e-05, + "loss": 3.2507, + "step": 5242 + }, + { + "epoch": 0.1617280190014228, + "grad_norm": 0.165352925658226, + "learning_rate": 4.942194169576791e-05, + "loss": 3.2487, + "step": 5243 + }, + { + "epoch": 0.16175886546699622, + "grad_norm": 0.16675402224063873, + "learning_rate": 4.942136602169858e-05, + "loss": 3.2806, + "step": 5244 + }, + { + "epoch": 0.16178971193256964, + "grad_norm": 0.16289041936397552, + "learning_rate": 4.942079006447819e-05, + "loss": 3.2716, + "step": 5245 + }, + { + "epoch": 0.16182055839814305, + "grad_norm": 0.17533095180988312, + "learning_rate": 4.942021382411343e-05, + "loss": 3.2757, + "step": 5246 + }, + { + "epoch": 0.16185140486371646, + "grad_norm": 0.16252867877483368, + "learning_rate": 4.941963730061094e-05, + "loss": 3.2454, + "step": 5247 + }, + { + "epoch": 0.16188225132928988, + "grad_norm": 0.17630621790885925, + "learning_rate": 4.941906049397745e-05, + "loss": 3.2661, + "step": 5248 + }, + { + "epoch": 0.1619130977948633, + "grad_norm": 0.1673036515712738, + "learning_rate": 4.9418483404219614e-05, + "loss": 3.2556, + "step": 5249 + }, + { + "epoch": 0.1619439442604367, + "grad_norm": 0.2375011146068573, + "learning_rate": 4.941790603134415e-05, + "loss": 3.243, + "step": 5250 + }, + { + "epoch": 0.16197479072601012, + "grad_norm": 0.17507901787757874, + "learning_rate": 4.941732837535773e-05, + "loss": 3.2172, + "step": 5251 + }, + { + "epoch": 0.16200563719158353, + "grad_norm": 0.1637374758720398, + "learning_rate": 4.9416750436267065e-05, + "loss": 3.2336, + "step": 5252 + }, + { + "epoch": 0.16203648365715695, + "grad_norm": 0.17361392080783844, + "learning_rate": 4.941617221407885e-05, + "loss": 3.2287, + "step": 5253 + }, + { + "epoch": 0.16206733012273036, + "grad_norm": 0.1580812633037567, + "learning_rate": 4.941559370879979e-05, + "loss": 3.2564, + "step": 5254 + }, + { + "epoch": 0.1620981765883038, + "grad_norm": 0.17246080935001373, + "learning_rate": 4.94150149204366e-05, + "loss": 3.2361, + "step": 5255 + }, + { + "epoch": 0.16212902305387722, + "grad_norm": 0.166893869638443, + "learning_rate": 4.941443584899598e-05, + "loss": 3.2484, + "step": 5256 + }, + { + "epoch": 0.16215986951945063, + "grad_norm": 0.15724053978919983, + "learning_rate": 4.941385649448464e-05, + "loss": 3.2636, + "step": 5257 + }, + { + "epoch": 0.16219071598502405, + "grad_norm": 0.1613057553768158, + "learning_rate": 4.941327685690931e-05, + "loss": 3.2325, + "step": 5258 + }, + { + "epoch": 0.16222156245059746, + "grad_norm": 0.16395303606987, + "learning_rate": 4.94126969362767e-05, + "loss": 3.2502, + "step": 5259 + }, + { + "epoch": 0.16225240891617088, + "grad_norm": 0.1669962853193283, + "learning_rate": 4.941211673259355e-05, + "loss": 3.2489, + "step": 5260 + }, + { + "epoch": 0.1622832553817443, + "grad_norm": 0.16886164247989655, + "learning_rate": 4.941153624586657e-05, + "loss": 3.2702, + "step": 5261 + }, + { + "epoch": 0.1623141018473177, + "grad_norm": 0.16340328752994537, + "learning_rate": 4.94109554761025e-05, + "loss": 3.2638, + "step": 5262 + }, + { + "epoch": 0.16234494831289112, + "grad_norm": 0.15857265889644623, + "learning_rate": 4.941037442330807e-05, + "loss": 3.2377, + "step": 5263 + }, + { + "epoch": 0.16237579477846453, + "grad_norm": 0.16358056664466858, + "learning_rate": 4.9409793087490016e-05, + "loss": 3.2453, + "step": 5264 + }, + { + "epoch": 0.16240664124403795, + "grad_norm": 0.16176742315292358, + "learning_rate": 4.940921146865508e-05, + "loss": 3.2366, + "step": 5265 + }, + { + "epoch": 0.16243748770961136, + "grad_norm": 0.15656152367591858, + "learning_rate": 4.9408629566810006e-05, + "loss": 3.2474, + "step": 5266 + }, + { + "epoch": 0.1624683341751848, + "grad_norm": 0.17586496472358704, + "learning_rate": 4.940804738196154e-05, + "loss": 3.2916, + "step": 5267 + }, + { + "epoch": 0.16249918064075822, + "grad_norm": 0.15335796773433685, + "learning_rate": 4.940746491411643e-05, + "loss": 3.2376, + "step": 5268 + }, + { + "epoch": 0.16253002710633163, + "grad_norm": 0.1649249941110611, + "learning_rate": 4.940688216328143e-05, + "loss": 3.2646, + "step": 5269 + }, + { + "epoch": 0.16256087357190505, + "grad_norm": 0.1633840948343277, + "learning_rate": 4.94062991294633e-05, + "loss": 3.2277, + "step": 5270 + }, + { + "epoch": 0.16259172003747846, + "grad_norm": 0.17776328325271606, + "learning_rate": 4.94057158126688e-05, + "loss": 3.2764, + "step": 5271 + }, + { + "epoch": 0.16262256650305187, + "grad_norm": 0.1639615297317505, + "learning_rate": 4.940513221290469e-05, + "loss": 3.2623, + "step": 5272 + }, + { + "epoch": 0.1626534129686253, + "grad_norm": 0.17413559556007385, + "learning_rate": 4.940454833017774e-05, + "loss": 3.2714, + "step": 5273 + }, + { + "epoch": 0.1626842594341987, + "grad_norm": 0.16448141634464264, + "learning_rate": 4.940396416449472e-05, + "loss": 3.2246, + "step": 5274 + }, + { + "epoch": 0.16271510589977212, + "grad_norm": 0.17172129452228546, + "learning_rate": 4.940337971586239e-05, + "loss": 3.2917, + "step": 5275 + }, + { + "epoch": 0.16274595236534553, + "grad_norm": 0.170777827501297, + "learning_rate": 4.9402794984287546e-05, + "loss": 3.2294, + "step": 5276 + }, + { + "epoch": 0.16277679883091895, + "grad_norm": 0.1614876538515091, + "learning_rate": 4.9402209969776957e-05, + "loss": 3.2632, + "step": 5277 + }, + { + "epoch": 0.16280764529649236, + "grad_norm": 0.16197209060192108, + "learning_rate": 4.9401624672337407e-05, + "loss": 3.2688, + "step": 5278 + }, + { + "epoch": 0.1628384917620658, + "grad_norm": 0.16356134414672852, + "learning_rate": 4.940103909197568e-05, + "loss": 3.2058, + "step": 5279 + }, + { + "epoch": 0.16286933822763922, + "grad_norm": 0.19306348264217377, + "learning_rate": 4.940045322869857e-05, + "loss": 3.2388, + "step": 5280 + }, + { + "epoch": 0.16290018469321263, + "grad_norm": 0.16499212384223938, + "learning_rate": 4.9399867082512865e-05, + "loss": 3.2206, + "step": 5281 + }, + { + "epoch": 0.16293103115878604, + "grad_norm": 0.1604006588459015, + "learning_rate": 4.939928065342537e-05, + "loss": 3.2405, + "step": 5282 + }, + { + "epoch": 0.16296187762435946, + "grad_norm": 0.17077814042568207, + "learning_rate": 4.939869394144287e-05, + "loss": 3.2483, + "step": 5283 + }, + { + "epoch": 0.16299272408993287, + "grad_norm": 0.15643924474716187, + "learning_rate": 4.939810694657219e-05, + "loss": 3.2289, + "step": 5284 + }, + { + "epoch": 0.1630235705555063, + "grad_norm": 0.17708991467952728, + "learning_rate": 4.93975196688201e-05, + "loss": 3.2377, + "step": 5285 + }, + { + "epoch": 0.1630544170210797, + "grad_norm": 0.16496233642101288, + "learning_rate": 4.939693210819345e-05, + "loss": 3.2432, + "step": 5286 + }, + { + "epoch": 0.16308526348665311, + "grad_norm": 0.17816780507564545, + "learning_rate": 4.939634426469902e-05, + "loss": 3.2539, + "step": 5287 + }, + { + "epoch": 0.16311610995222653, + "grad_norm": 0.16092383861541748, + "learning_rate": 4.939575613834365e-05, + "loss": 3.2825, + "step": 5288 + }, + { + "epoch": 0.16314695641779994, + "grad_norm": 0.183834969997406, + "learning_rate": 4.939516772913414e-05, + "loss": 3.245, + "step": 5289 + }, + { + "epoch": 0.16317780288337336, + "grad_norm": 0.1755627989768982, + "learning_rate": 4.939457903707732e-05, + "loss": 3.2563, + "step": 5290 + }, + { + "epoch": 0.1632086493489468, + "grad_norm": 0.16392485797405243, + "learning_rate": 4.939399006218002e-05, + "loss": 3.2466, + "step": 5291 + }, + { + "epoch": 0.1632394958145202, + "grad_norm": 0.1615988165140152, + "learning_rate": 4.9393400804449064e-05, + "loss": 3.2566, + "step": 5292 + }, + { + "epoch": 0.16327034228009363, + "grad_norm": 0.16843098402023315, + "learning_rate": 4.939281126389128e-05, + "loss": 3.251, + "step": 5293 + }, + { + "epoch": 0.16330118874566704, + "grad_norm": 0.16285288333892822, + "learning_rate": 4.9392221440513506e-05, + "loss": 3.2429, + "step": 5294 + }, + { + "epoch": 0.16333203521124046, + "grad_norm": 0.16477616131305695, + "learning_rate": 4.9391631334322584e-05, + "loss": 3.2558, + "step": 5295 + }, + { + "epoch": 0.16336288167681387, + "grad_norm": 0.15775734186172485, + "learning_rate": 4.939104094532536e-05, + "loss": 3.2326, + "step": 5296 + }, + { + "epoch": 0.16339372814238728, + "grad_norm": 0.1753617376089096, + "learning_rate": 4.9390450273528674e-05, + "loss": 3.2608, + "step": 5297 + }, + { + "epoch": 0.1634245746079607, + "grad_norm": 0.1659909188747406, + "learning_rate": 4.938985931893937e-05, + "loss": 3.2651, + "step": 5298 + }, + { + "epoch": 0.1634554210735341, + "grad_norm": 0.16302818059921265, + "learning_rate": 4.9389268081564296e-05, + "loss": 3.2385, + "step": 5299 + }, + { + "epoch": 0.16348626753910753, + "grad_norm": 0.15513300895690918, + "learning_rate": 4.938867656141033e-05, + "loss": 3.2595, + "step": 5300 + }, + { + "epoch": 0.16351711400468094, + "grad_norm": 0.1671561598777771, + "learning_rate": 4.938808475848431e-05, + "loss": 3.2507, + "step": 5301 + }, + { + "epoch": 0.16354796047025436, + "grad_norm": 0.2202921062707901, + "learning_rate": 4.938749267279309e-05, + "loss": 3.2685, + "step": 5302 + }, + { + "epoch": 0.1635788069358278, + "grad_norm": 0.16536468267440796, + "learning_rate": 4.9386900304343566e-05, + "loss": 3.2412, + "step": 5303 + }, + { + "epoch": 0.1636096534014012, + "grad_norm": 0.15754759311676025, + "learning_rate": 4.9386307653142583e-05, + "loss": 3.2598, + "step": 5304 + }, + { + "epoch": 0.16364049986697463, + "grad_norm": 0.15585927665233612, + "learning_rate": 4.938571471919701e-05, + "loss": 3.2214, + "step": 5305 + }, + { + "epoch": 0.16367134633254804, + "grad_norm": 0.15924951434135437, + "learning_rate": 4.9385121502513734e-05, + "loss": 3.2398, + "step": 5306 + }, + { + "epoch": 0.16370219279812145, + "grad_norm": 0.16699600219726562, + "learning_rate": 4.938452800309963e-05, + "loss": 3.2342, + "step": 5307 + }, + { + "epoch": 0.16373303926369487, + "grad_norm": 0.15951672196388245, + "learning_rate": 4.938393422096157e-05, + "loss": 3.2537, + "step": 5308 + }, + { + "epoch": 0.16376388572926828, + "grad_norm": 0.1623559445142746, + "learning_rate": 4.938334015610645e-05, + "loss": 3.2718, + "step": 5309 + }, + { + "epoch": 0.1637947321948417, + "grad_norm": 0.15788517892360687, + "learning_rate": 4.9382745808541155e-05, + "loss": 3.2515, + "step": 5310 + }, + { + "epoch": 0.1638255786604151, + "grad_norm": 0.16253617405891418, + "learning_rate": 4.9382151178272574e-05, + "loss": 3.268, + "step": 5311 + }, + { + "epoch": 0.16385642512598853, + "grad_norm": 0.16106592118740082, + "learning_rate": 4.938155626530761e-05, + "loss": 3.2598, + "step": 5312 + }, + { + "epoch": 0.16388727159156194, + "grad_norm": 0.1722228229045868, + "learning_rate": 4.938096106965314e-05, + "loss": 3.2237, + "step": 5313 + }, + { + "epoch": 0.16391811805713535, + "grad_norm": 0.16995534300804138, + "learning_rate": 4.938036559131608e-05, + "loss": 3.2439, + "step": 5314 + }, + { + "epoch": 0.1639489645227088, + "grad_norm": 0.16903941333293915, + "learning_rate": 4.9379769830303335e-05, + "loss": 3.2982, + "step": 5315 + }, + { + "epoch": 0.1639798109882822, + "grad_norm": 0.9169728755950928, + "learning_rate": 4.93791737866218e-05, + "loss": 3.2952, + "step": 5316 + }, + { + "epoch": 0.16401065745385562, + "grad_norm": 0.16958270967006683, + "learning_rate": 4.937857746027841e-05, + "loss": 3.2261, + "step": 5317 + }, + { + "epoch": 0.16404150391942904, + "grad_norm": 0.177603617310524, + "learning_rate": 4.937798085128006e-05, + "loss": 3.2539, + "step": 5318 + }, + { + "epoch": 0.16407235038500245, + "grad_norm": 0.17706185579299927, + "learning_rate": 4.937738395963366e-05, + "loss": 3.2487, + "step": 5319 + }, + { + "epoch": 0.16410319685057587, + "grad_norm": 0.18088555335998535, + "learning_rate": 4.937678678534615e-05, + "loss": 3.2639, + "step": 5320 + }, + { + "epoch": 0.16413404331614928, + "grad_norm": 0.1761385053396225, + "learning_rate": 4.937618932842445e-05, + "loss": 3.2204, + "step": 5321 + }, + { + "epoch": 0.1641648897817227, + "grad_norm": 0.17576797306537628, + "learning_rate": 4.937559158887548e-05, + "loss": 3.2285, + "step": 5322 + }, + { + "epoch": 0.1641957362472961, + "grad_norm": 0.17880894243717194, + "learning_rate": 4.9374993566706175e-05, + "loss": 3.2574, + "step": 5323 + }, + { + "epoch": 0.16422658271286952, + "grad_norm": 0.17717207968235016, + "learning_rate": 4.9374395261923464e-05, + "loss": 3.2797, + "step": 5324 + }, + { + "epoch": 0.16425742917844294, + "grad_norm": 0.1692977398633957, + "learning_rate": 4.937379667453429e-05, + "loss": 3.2214, + "step": 5325 + }, + { + "epoch": 0.16428827564401635, + "grad_norm": 0.1696232408285141, + "learning_rate": 4.937319780454559e-05, + "loss": 3.2622, + "step": 5326 + }, + { + "epoch": 0.1643191221095898, + "grad_norm": 0.16904324293136597, + "learning_rate": 4.9372598651964306e-05, + "loss": 3.2151, + "step": 5327 + }, + { + "epoch": 0.1643499685751632, + "grad_norm": 0.1706136167049408, + "learning_rate": 4.93719992167974e-05, + "loss": 3.2877, + "step": 5328 + }, + { + "epoch": 0.16438081504073662, + "grad_norm": 0.17810219526290894, + "learning_rate": 4.937139949905179e-05, + "loss": 3.2635, + "step": 5329 + }, + { + "epoch": 0.16441166150631004, + "grad_norm": 0.16782554984092712, + "learning_rate": 4.937079949873445e-05, + "loss": 3.2179, + "step": 5330 + }, + { + "epoch": 0.16444250797188345, + "grad_norm": 0.16846567392349243, + "learning_rate": 4.937019921585235e-05, + "loss": 3.2782, + "step": 5331 + }, + { + "epoch": 0.16447335443745686, + "grad_norm": 0.17257942259311676, + "learning_rate": 4.9369598650412426e-05, + "loss": 3.2681, + "step": 5332 + }, + { + "epoch": 0.16450420090303028, + "grad_norm": 0.16067089140415192, + "learning_rate": 4.9368997802421646e-05, + "loss": 3.2705, + "step": 5333 + }, + { + "epoch": 0.1645350473686037, + "grad_norm": 0.16547802090644836, + "learning_rate": 4.936839667188699e-05, + "loss": 3.2744, + "step": 5334 + }, + { + "epoch": 0.1645658938341771, + "grad_norm": 0.15938159823417664, + "learning_rate": 4.936779525881541e-05, + "loss": 3.2677, + "step": 5335 + }, + { + "epoch": 0.16459674029975052, + "grad_norm": 0.16324838995933533, + "learning_rate": 4.936719356321389e-05, + "loss": 3.2416, + "step": 5336 + }, + { + "epoch": 0.16462758676532394, + "grad_norm": 0.16904045641422272, + "learning_rate": 4.93665915850894e-05, + "loss": 3.2637, + "step": 5337 + }, + { + "epoch": 0.16465843323089735, + "grad_norm": 0.16649554669857025, + "learning_rate": 4.936598932444893e-05, + "loss": 3.2779, + "step": 5338 + }, + { + "epoch": 0.1646892796964708, + "grad_norm": 0.16923311352729797, + "learning_rate": 4.936538678129945e-05, + "loss": 3.2528, + "step": 5339 + }, + { + "epoch": 0.1647201261620442, + "grad_norm": 0.1663179248571396, + "learning_rate": 4.9364783955647954e-05, + "loss": 3.2356, + "step": 5340 + }, + { + "epoch": 0.16475097262761762, + "grad_norm": 0.1675247997045517, + "learning_rate": 4.936418084750143e-05, + "loss": 3.2445, + "step": 5341 + }, + { + "epoch": 0.16478181909319103, + "grad_norm": 0.16472448408603668, + "learning_rate": 4.936357745686687e-05, + "loss": 3.2279, + "step": 5342 + }, + { + "epoch": 0.16481266555876445, + "grad_norm": 0.16752289235591888, + "learning_rate": 4.936297378375128e-05, + "loss": 3.2095, + "step": 5343 + }, + { + "epoch": 0.16484351202433786, + "grad_norm": 0.16952724754810333, + "learning_rate": 4.936236982816164e-05, + "loss": 3.285, + "step": 5344 + }, + { + "epoch": 0.16487435848991128, + "grad_norm": 0.1610931009054184, + "learning_rate": 4.936176559010496e-05, + "loss": 3.2964, + "step": 5345 + }, + { + "epoch": 0.1649052049554847, + "grad_norm": 0.17924444377422333, + "learning_rate": 4.936116106958825e-05, + "loss": 3.2837, + "step": 5346 + }, + { + "epoch": 0.1649360514210581, + "grad_norm": 0.16708187758922577, + "learning_rate": 4.936055626661851e-05, + "loss": 3.2452, + "step": 5347 + }, + { + "epoch": 0.16496689788663152, + "grad_norm": 0.1638604998588562, + "learning_rate": 4.935995118120277e-05, + "loss": 3.2316, + "step": 5348 + }, + { + "epoch": 0.16499774435220493, + "grad_norm": 0.1689494401216507, + "learning_rate": 4.935934581334803e-05, + "loss": 3.3028, + "step": 5349 + }, + { + "epoch": 0.16502859081777835, + "grad_norm": 0.18167054653167725, + "learning_rate": 4.935874016306131e-05, + "loss": 3.2478, + "step": 5350 + }, + { + "epoch": 0.1650594372833518, + "grad_norm": 0.18528801202774048, + "learning_rate": 4.9358134230349645e-05, + "loss": 3.2352, + "step": 5351 + }, + { + "epoch": 0.1650902837489252, + "grad_norm": 0.16573837399482727, + "learning_rate": 4.935752801522004e-05, + "loss": 3.272, + "step": 5352 + }, + { + "epoch": 0.16512113021449862, + "grad_norm": 0.16953372955322266, + "learning_rate": 4.935692151767955e-05, + "loss": 3.2638, + "step": 5353 + }, + { + "epoch": 0.16515197668007203, + "grad_norm": 0.1639828234910965, + "learning_rate": 4.9356314737735175e-05, + "loss": 3.2535, + "step": 5354 + }, + { + "epoch": 0.16518282314564545, + "grad_norm": 0.16256718337535858, + "learning_rate": 4.9355707675393984e-05, + "loss": 3.2485, + "step": 5355 + }, + { + "epoch": 0.16521366961121886, + "grad_norm": 0.15894383192062378, + "learning_rate": 4.9355100330662985e-05, + "loss": 3.2339, + "step": 5356 + }, + { + "epoch": 0.16524451607679227, + "grad_norm": 0.15688514709472656, + "learning_rate": 4.935449270354924e-05, + "loss": 3.2563, + "step": 5357 + }, + { + "epoch": 0.1652753625423657, + "grad_norm": 0.16007670760154724, + "learning_rate": 4.935388479405978e-05, + "loss": 3.2725, + "step": 5358 + }, + { + "epoch": 0.1653062090079391, + "grad_norm": 0.16614651679992676, + "learning_rate": 4.935327660220167e-05, + "loss": 3.2849, + "step": 5359 + }, + { + "epoch": 0.16533705547351252, + "grad_norm": 0.16494926810264587, + "learning_rate": 4.935266812798195e-05, + "loss": 3.2633, + "step": 5360 + }, + { + "epoch": 0.16536790193908593, + "grad_norm": 0.17661947011947632, + "learning_rate": 4.935205937140767e-05, + "loss": 3.244, + "step": 5361 + }, + { + "epoch": 0.16539874840465935, + "grad_norm": 0.15940260887145996, + "learning_rate": 4.9351450332485904e-05, + "loss": 3.261, + "step": 5362 + }, + { + "epoch": 0.1654295948702328, + "grad_norm": 0.1674206256866455, + "learning_rate": 4.93508410112237e-05, + "loss": 3.2571, + "step": 5363 + }, + { + "epoch": 0.1654604413358062, + "grad_norm": 0.15950293838977814, + "learning_rate": 4.9350231407628126e-05, + "loss": 3.2569, + "step": 5364 + }, + { + "epoch": 0.16549128780137962, + "grad_norm": 0.16608074307441711, + "learning_rate": 4.934962152170626e-05, + "loss": 3.2759, + "step": 5365 + }, + { + "epoch": 0.16552213426695303, + "grad_norm": 0.16845589876174927, + "learning_rate": 4.9349011353465156e-05, + "loss": 3.2003, + "step": 5366 + }, + { + "epoch": 0.16555298073252644, + "grad_norm": 0.15534758567810059, + "learning_rate": 4.9348400902911904e-05, + "loss": 3.2313, + "step": 5367 + }, + { + "epoch": 0.16558382719809986, + "grad_norm": 0.1653418242931366, + "learning_rate": 4.9347790170053575e-05, + "loss": 3.2288, + "step": 5368 + }, + { + "epoch": 0.16561467366367327, + "grad_norm": 0.16719330847263336, + "learning_rate": 4.934717915489725e-05, + "loss": 3.2478, + "step": 5369 + }, + { + "epoch": 0.1656455201292467, + "grad_norm": 0.16158272325992584, + "learning_rate": 4.934656785745001e-05, + "loss": 3.2682, + "step": 5370 + }, + { + "epoch": 0.1656763665948201, + "grad_norm": 0.17626558244228363, + "learning_rate": 4.934595627771895e-05, + "loss": 3.2536, + "step": 5371 + }, + { + "epoch": 0.16570721306039352, + "grad_norm": 0.1632303148508072, + "learning_rate": 4.934534441571115e-05, + "loss": 3.281, + "step": 5372 + }, + { + "epoch": 0.16573805952596693, + "grad_norm": 0.16988343000411987, + "learning_rate": 4.934473227143372e-05, + "loss": 3.2727, + "step": 5373 + }, + { + "epoch": 0.16576890599154034, + "grad_norm": 0.15535968542099, + "learning_rate": 4.934411984489374e-05, + "loss": 3.2093, + "step": 5374 + }, + { + "epoch": 0.16579975245711379, + "grad_norm": 0.15952348709106445, + "learning_rate": 4.9343507136098325e-05, + "loss": 3.2389, + "step": 5375 + }, + { + "epoch": 0.1658305989226872, + "grad_norm": 0.15784990787506104, + "learning_rate": 4.934289414505458e-05, + "loss": 3.2128, + "step": 5376 + }, + { + "epoch": 0.16586144538826061, + "grad_norm": 0.1554117053747177, + "learning_rate": 4.93422808717696e-05, + "loss": 3.2181, + "step": 5377 + }, + { + "epoch": 0.16589229185383403, + "grad_norm": 0.16415826976299286, + "learning_rate": 4.93416673162505e-05, + "loss": 3.2514, + "step": 5378 + }, + { + "epoch": 0.16592313831940744, + "grad_norm": 0.17048437893390656, + "learning_rate": 4.934105347850439e-05, + "loss": 3.2521, + "step": 5379 + }, + { + "epoch": 0.16595398478498086, + "grad_norm": 0.15855832397937775, + "learning_rate": 4.93404393585384e-05, + "loss": 3.2219, + "step": 5380 + }, + { + "epoch": 0.16598483125055427, + "grad_norm": 0.16196973621845245, + "learning_rate": 4.9339824956359636e-05, + "loss": 3.26, + "step": 5381 + }, + { + "epoch": 0.16601567771612769, + "grad_norm": 0.15473559498786926, + "learning_rate": 4.9339210271975236e-05, + "loss": 3.2413, + "step": 5382 + }, + { + "epoch": 0.1660465241817011, + "grad_norm": 0.1738855540752411, + "learning_rate": 4.933859530539232e-05, + "loss": 3.2345, + "step": 5383 + }, + { + "epoch": 0.1660773706472745, + "grad_norm": 0.16031700372695923, + "learning_rate": 4.9337980056618006e-05, + "loss": 3.2718, + "step": 5384 + }, + { + "epoch": 0.16610821711284793, + "grad_norm": 0.15946753323078156, + "learning_rate": 4.9337364525659446e-05, + "loss": 3.2008, + "step": 5385 + }, + { + "epoch": 0.16613906357842134, + "grad_norm": 0.16371312737464905, + "learning_rate": 4.933674871252377e-05, + "loss": 3.2562, + "step": 5386 + }, + { + "epoch": 0.16616991004399478, + "grad_norm": 0.16357718408107758, + "learning_rate": 4.9336132617218114e-05, + "loss": 3.2559, + "step": 5387 + }, + { + "epoch": 0.1662007565095682, + "grad_norm": 0.15950819849967957, + "learning_rate": 4.9335516239749626e-05, + "loss": 3.2372, + "step": 5388 + }, + { + "epoch": 0.1662316029751416, + "grad_norm": 0.1567741334438324, + "learning_rate": 4.933489958012546e-05, + "loss": 3.2293, + "step": 5389 + }, + { + "epoch": 0.16626244944071503, + "grad_norm": 0.1513269990682602, + "learning_rate": 4.9334282638352744e-05, + "loss": 3.2423, + "step": 5390 + }, + { + "epoch": 0.16629329590628844, + "grad_norm": 0.16708910465240479, + "learning_rate": 4.933366541443865e-05, + "loss": 3.2552, + "step": 5391 + }, + { + "epoch": 0.16632414237186186, + "grad_norm": 0.15492254495620728, + "learning_rate": 4.9333047908390325e-05, + "loss": 3.249, + "step": 5392 + }, + { + "epoch": 0.16635498883743527, + "grad_norm": 0.1599360704421997, + "learning_rate": 4.933243012021493e-05, + "loss": 3.2266, + "step": 5393 + }, + { + "epoch": 0.16638583530300868, + "grad_norm": 0.16060534119606018, + "learning_rate": 4.9331812049919636e-05, + "loss": 3.2194, + "step": 5394 + }, + { + "epoch": 0.1664166817685821, + "grad_norm": 0.15926790237426758, + "learning_rate": 4.93311936975116e-05, + "loss": 3.2589, + "step": 5395 + }, + { + "epoch": 0.1664475282341555, + "grad_norm": 0.15353329479694366, + "learning_rate": 4.9330575062998e-05, + "loss": 3.2299, + "step": 5396 + }, + { + "epoch": 0.16647837469972893, + "grad_norm": 0.17746596038341522, + "learning_rate": 4.9329956146386e-05, + "loss": 3.2719, + "step": 5397 + }, + { + "epoch": 0.16650922116530234, + "grad_norm": 0.17048855125904083, + "learning_rate": 4.9329336947682776e-05, + "loss": 3.2382, + "step": 5398 + }, + { + "epoch": 0.16654006763087578, + "grad_norm": 0.1935863345861435, + "learning_rate": 4.932871746689551e-05, + "loss": 3.2741, + "step": 5399 + }, + { + "epoch": 0.1665709140964492, + "grad_norm": 0.23712190985679626, + "learning_rate": 4.932809770403138e-05, + "loss": 3.273, + "step": 5400 + }, + { + "epoch": 0.1666017605620226, + "grad_norm": 0.15800894796848297, + "learning_rate": 4.932747765909759e-05, + "loss": 3.2593, + "step": 5401 + }, + { + "epoch": 0.16663260702759602, + "grad_norm": 0.1643064320087433, + "learning_rate": 4.93268573321013e-05, + "loss": 3.2613, + "step": 5402 + }, + { + "epoch": 0.16666345349316944, + "grad_norm": 0.17029047012329102, + "learning_rate": 4.9326236723049734e-05, + "loss": 3.2079, + "step": 5403 + }, + { + "epoch": 0.16669429995874285, + "grad_norm": 0.16731765866279602, + "learning_rate": 4.9325615831950064e-05, + "loss": 3.276, + "step": 5404 + }, + { + "epoch": 0.16672514642431627, + "grad_norm": 0.2548476457595825, + "learning_rate": 4.93249946588095e-05, + "loss": 3.2708, + "step": 5405 + }, + { + "epoch": 0.16675599288988968, + "grad_norm": 0.16315743327140808, + "learning_rate": 4.9324373203635236e-05, + "loss": 3.2577, + "step": 5406 + }, + { + "epoch": 0.1667868393554631, + "grad_norm": 0.159129798412323, + "learning_rate": 4.9323751466434484e-05, + "loss": 3.2351, + "step": 5407 + }, + { + "epoch": 0.1668176858210365, + "grad_norm": 0.1596859097480774, + "learning_rate": 4.932312944721445e-05, + "loss": 3.26, + "step": 5408 + }, + { + "epoch": 0.16684853228660992, + "grad_norm": 0.15753522515296936, + "learning_rate": 4.932250714598235e-05, + "loss": 3.2237, + "step": 5409 + }, + { + "epoch": 0.16687937875218334, + "grad_norm": 0.16543327271938324, + "learning_rate": 4.932188456274539e-05, + "loss": 3.2806, + "step": 5410 + }, + { + "epoch": 0.16691022521775678, + "grad_norm": 0.16697128117084503, + "learning_rate": 4.932126169751079e-05, + "loss": 3.2329, + "step": 5411 + }, + { + "epoch": 0.1669410716833302, + "grad_norm": 0.1647985279560089, + "learning_rate": 4.932063855028579e-05, + "loss": 3.2716, + "step": 5412 + }, + { + "epoch": 0.1669719181489036, + "grad_norm": 0.15586380660533905, + "learning_rate": 4.932001512107759e-05, + "loss": 3.2383, + "step": 5413 + }, + { + "epoch": 0.16700276461447702, + "grad_norm": 0.1565781533718109, + "learning_rate": 4.9319391409893436e-05, + "loss": 3.2457, + "step": 5414 + }, + { + "epoch": 0.16703361108005044, + "grad_norm": 0.17868772149085999, + "learning_rate": 4.931876741674055e-05, + "loss": 3.255, + "step": 5415 + }, + { + "epoch": 0.16706445754562385, + "grad_norm": 1.3282335996627808, + "learning_rate": 4.9318143141626174e-05, + "loss": 3.2567, + "step": 5416 + }, + { + "epoch": 0.16709530401119727, + "grad_norm": 0.16155600547790527, + "learning_rate": 4.9317518584557535e-05, + "loss": 3.2738, + "step": 5417 + }, + { + "epoch": 0.16712615047677068, + "grad_norm": 0.1688619703054428, + "learning_rate": 4.931689374554188e-05, + "loss": 3.258, + "step": 5418 + }, + { + "epoch": 0.1671569969423441, + "grad_norm": 0.17811337113380432, + "learning_rate": 4.931626862458646e-05, + "loss": 3.2333, + "step": 5419 + }, + { + "epoch": 0.1671878434079175, + "grad_norm": 0.17508146166801453, + "learning_rate": 4.931564322169852e-05, + "loss": 3.2469, + "step": 5420 + }, + { + "epoch": 0.16721868987349092, + "grad_norm": 0.16840232908725739, + "learning_rate": 4.93150175368853e-05, + "loss": 3.2561, + "step": 5421 + }, + { + "epoch": 0.16724953633906434, + "grad_norm": 0.18571530282497406, + "learning_rate": 4.931439157015407e-05, + "loss": 3.2577, + "step": 5422 + }, + { + "epoch": 0.16728038280463778, + "grad_norm": 0.16178865730762482, + "learning_rate": 4.9313765321512077e-05, + "loss": 3.2451, + "step": 5423 + }, + { + "epoch": 0.1673112292702112, + "grad_norm": 0.17309381067752838, + "learning_rate": 4.931313879096659e-05, + "loss": 3.2369, + "step": 5424 + }, + { + "epoch": 0.1673420757357846, + "grad_norm": 0.16540902853012085, + "learning_rate": 4.931251197852487e-05, + "loss": 3.2477, + "step": 5425 + }, + { + "epoch": 0.16737292220135802, + "grad_norm": 0.16782544553279877, + "learning_rate": 4.9311884884194174e-05, + "loss": 3.2359, + "step": 5426 + }, + { + "epoch": 0.16740376866693144, + "grad_norm": 0.17496806383132935, + "learning_rate": 4.931125750798179e-05, + "loss": 3.2625, + "step": 5427 + }, + { + "epoch": 0.16743461513250485, + "grad_norm": 0.16030563414096832, + "learning_rate": 4.9310629849894985e-05, + "loss": 3.222, + "step": 5428 + }, + { + "epoch": 0.16746546159807826, + "grad_norm": 0.16999784111976624, + "learning_rate": 4.931000190994104e-05, + "loss": 3.2318, + "step": 5429 + }, + { + "epoch": 0.16749630806365168, + "grad_norm": 0.16190531849861145, + "learning_rate": 4.930937368812723e-05, + "loss": 3.2368, + "step": 5430 + }, + { + "epoch": 0.1675271545292251, + "grad_norm": 0.16203095018863678, + "learning_rate": 4.930874518446083e-05, + "loss": 3.2705, + "step": 5431 + }, + { + "epoch": 0.1675580009947985, + "grad_norm": 0.1637336164712906, + "learning_rate": 4.9308116398949144e-05, + "loss": 3.2873, + "step": 5432 + }, + { + "epoch": 0.16758884746037192, + "grad_norm": 0.15693162381649017, + "learning_rate": 4.930748733159945e-05, + "loss": 3.273, + "step": 5433 + }, + { + "epoch": 0.16761969392594533, + "grad_norm": 0.16235341131687164, + "learning_rate": 4.9306857982419064e-05, + "loss": 3.2508, + "step": 5434 + }, + { + "epoch": 0.16765054039151878, + "grad_norm": 0.1569259762763977, + "learning_rate": 4.930622835141525e-05, + "loss": 3.258, + "step": 5435 + }, + { + "epoch": 0.1676813868570922, + "grad_norm": 0.15794681012630463, + "learning_rate": 4.930559843859533e-05, + "loss": 3.2501, + "step": 5436 + }, + { + "epoch": 0.1677122333226656, + "grad_norm": 0.16721121966838837, + "learning_rate": 4.9304968243966606e-05, + "loss": 3.2643, + "step": 5437 + }, + { + "epoch": 0.16774307978823902, + "grad_norm": 0.15681643784046173, + "learning_rate": 4.930433776753638e-05, + "loss": 3.2388, + "step": 5438 + }, + { + "epoch": 0.16777392625381243, + "grad_norm": 0.18603910505771637, + "learning_rate": 4.9303707009311954e-05, + "loss": 3.2384, + "step": 5439 + }, + { + "epoch": 0.16780477271938585, + "grad_norm": 0.16213132441043854, + "learning_rate": 4.930307596930066e-05, + "loss": 3.2344, + "step": 5440 + }, + { + "epoch": 0.16783561918495926, + "grad_norm": 0.8784133791923523, + "learning_rate": 4.93024446475098e-05, + "loss": 3.2555, + "step": 5441 + }, + { + "epoch": 0.16786646565053268, + "grad_norm": 0.17438602447509766, + "learning_rate": 4.93018130439467e-05, + "loss": 3.2293, + "step": 5442 + }, + { + "epoch": 0.1678973121161061, + "grad_norm": 0.2001504749059677, + "learning_rate": 4.930118115861868e-05, + "loss": 3.2592, + "step": 5443 + }, + { + "epoch": 0.1679281585816795, + "grad_norm": 0.2138446420431137, + "learning_rate": 4.9300548991533066e-05, + "loss": 3.2779, + "step": 5444 + }, + { + "epoch": 0.16795900504725292, + "grad_norm": 0.1868847757577896, + "learning_rate": 4.92999165426972e-05, + "loss": 3.1939, + "step": 5445 + }, + { + "epoch": 0.16798985151282633, + "grad_norm": 0.1881057769060135, + "learning_rate": 4.929928381211839e-05, + "loss": 3.2581, + "step": 5446 + }, + { + "epoch": 0.16802069797839977, + "grad_norm": 0.3971162736415863, + "learning_rate": 4.929865079980399e-05, + "loss": 3.238, + "step": 5447 + }, + { + "epoch": 0.1680515444439732, + "grad_norm": 0.16883911192417145, + "learning_rate": 4.929801750576134e-05, + "loss": 3.2503, + "step": 5448 + }, + { + "epoch": 0.1680823909095466, + "grad_norm": 0.7525451183319092, + "learning_rate": 4.929738392999778e-05, + "loss": 3.2471, + "step": 5449 + }, + { + "epoch": 0.16811323737512002, + "grad_norm": 0.18650057911872864, + "learning_rate": 4.9296750072520645e-05, + "loss": 3.2111, + "step": 5450 + }, + { + "epoch": 0.16814408384069343, + "grad_norm": 0.2615026831626892, + "learning_rate": 4.929611593333731e-05, + "loss": 3.288, + "step": 5451 + }, + { + "epoch": 0.16817493030626685, + "grad_norm": 0.3474155366420746, + "learning_rate": 4.929548151245509e-05, + "loss": 3.3047, + "step": 5452 + }, + { + "epoch": 0.16820577677184026, + "grad_norm": 0.18574047088623047, + "learning_rate": 4.929484680988138e-05, + "loss": 3.2626, + "step": 5453 + }, + { + "epoch": 0.16823662323741367, + "grad_norm": 0.22568759322166443, + "learning_rate": 4.92942118256235e-05, + "loss": 3.3014, + "step": 5454 + }, + { + "epoch": 0.1682674697029871, + "grad_norm": 0.2144346982240677, + "learning_rate": 4.929357655968885e-05, + "loss": 3.2222, + "step": 5455 + }, + { + "epoch": 0.1682983161685605, + "grad_norm": 0.2251526415348053, + "learning_rate": 4.9292941012084784e-05, + "loss": 3.2874, + "step": 5456 + }, + { + "epoch": 0.16832916263413392, + "grad_norm": 0.19240890443325043, + "learning_rate": 4.9292305182818654e-05, + "loss": 3.2759, + "step": 5457 + }, + { + "epoch": 0.16836000909970733, + "grad_norm": 0.18474330008029938, + "learning_rate": 4.929166907189784e-05, + "loss": 3.2602, + "step": 5458 + }, + { + "epoch": 0.16839085556528077, + "grad_norm": 0.17822012305259705, + "learning_rate": 4.929103267932973e-05, + "loss": 3.2339, + "step": 5459 + }, + { + "epoch": 0.1684217020308542, + "grad_norm": 0.17974050343036652, + "learning_rate": 4.9290396005121684e-05, + "loss": 3.2623, + "step": 5460 + }, + { + "epoch": 0.1684525484964276, + "grad_norm": 0.18127568066120148, + "learning_rate": 4.928975904928111e-05, + "loss": 3.2893, + "step": 5461 + }, + { + "epoch": 0.16848339496200102, + "grad_norm": 0.17738036811351776, + "learning_rate": 4.928912181181535e-05, + "loss": 3.2396, + "step": 5462 + }, + { + "epoch": 0.16851424142757443, + "grad_norm": 0.20136874914169312, + "learning_rate": 4.928848429273184e-05, + "loss": 3.2626, + "step": 5463 + }, + { + "epoch": 0.16854508789314784, + "grad_norm": 0.18578103184700012, + "learning_rate": 4.928784649203795e-05, + "loss": 3.2352, + "step": 5464 + }, + { + "epoch": 0.16857593435872126, + "grad_norm": 0.17471304535865784, + "learning_rate": 4.928720840974106e-05, + "loss": 3.2612, + "step": 5465 + }, + { + "epoch": 0.16860678082429467, + "grad_norm": 0.16574621200561523, + "learning_rate": 4.9286570045848594e-05, + "loss": 3.2222, + "step": 5466 + }, + { + "epoch": 0.1686376272898681, + "grad_norm": 0.16691939532756805, + "learning_rate": 4.9285931400367944e-05, + "loss": 3.2151, + "step": 5467 + }, + { + "epoch": 0.1686684737554415, + "grad_norm": 0.19102473556995392, + "learning_rate": 4.928529247330651e-05, + "loss": 3.2841, + "step": 5468 + }, + { + "epoch": 0.16869932022101491, + "grad_norm": 0.17095233500003815, + "learning_rate": 4.9284653264671705e-05, + "loss": 3.2612, + "step": 5469 + }, + { + "epoch": 0.16873016668658833, + "grad_norm": 0.1691117137670517, + "learning_rate": 4.928401377447094e-05, + "loss": 3.2637, + "step": 5470 + }, + { + "epoch": 0.16876101315216177, + "grad_norm": 0.176065593957901, + "learning_rate": 4.928337400271162e-05, + "loss": 3.258, + "step": 5471 + }, + { + "epoch": 0.16879185961773518, + "grad_norm": 0.16292783617973328, + "learning_rate": 4.928273394940118e-05, + "loss": 3.2712, + "step": 5472 + }, + { + "epoch": 0.1688227060833086, + "grad_norm": 0.20953421294689178, + "learning_rate": 4.928209361454703e-05, + "loss": 3.2327, + "step": 5473 + }, + { + "epoch": 0.168853552548882, + "grad_norm": 0.16154377162456512, + "learning_rate": 4.9281452998156596e-05, + "loss": 3.2682, + "step": 5474 + }, + { + "epoch": 0.16888439901445543, + "grad_norm": 0.16824528574943542, + "learning_rate": 4.928081210023731e-05, + "loss": 3.2253, + "step": 5475 + }, + { + "epoch": 0.16891524548002884, + "grad_norm": 0.1653156280517578, + "learning_rate": 4.928017092079659e-05, + "loss": 3.2406, + "step": 5476 + }, + { + "epoch": 0.16894609194560226, + "grad_norm": 0.17981620132923126, + "learning_rate": 4.927952945984188e-05, + "loss": 3.2549, + "step": 5477 + }, + { + "epoch": 0.16897693841117567, + "grad_norm": 0.48751237988471985, + "learning_rate": 4.927888771738062e-05, + "loss": 3.2565, + "step": 5478 + }, + { + "epoch": 0.16900778487674908, + "grad_norm": 0.3295288681983948, + "learning_rate": 4.927824569342025e-05, + "loss": 3.2279, + "step": 5479 + }, + { + "epoch": 0.1690386313423225, + "grad_norm": 0.1686423122882843, + "learning_rate": 4.927760338796821e-05, + "loss": 3.2844, + "step": 5480 + }, + { + "epoch": 0.1690694778078959, + "grad_norm": 0.18133464455604553, + "learning_rate": 4.927696080103194e-05, + "loss": 3.255, + "step": 5481 + }, + { + "epoch": 0.16910032427346933, + "grad_norm": 0.18118740618228912, + "learning_rate": 4.927631793261891e-05, + "loss": 3.2902, + "step": 5482 + }, + { + "epoch": 0.16913117073904277, + "grad_norm": 0.9923518300056458, + "learning_rate": 4.927567478273655e-05, + "loss": 3.3169, + "step": 5483 + }, + { + "epoch": 0.16916201720461618, + "grad_norm": 0.16924935579299927, + "learning_rate": 4.927503135139233e-05, + "loss": 3.2342, + "step": 5484 + }, + { + "epoch": 0.1691928636701896, + "grad_norm": 0.175795778632164, + "learning_rate": 4.927438763859372e-05, + "loss": 3.2571, + "step": 5485 + }, + { + "epoch": 0.169223710135763, + "grad_norm": 0.17047622799873352, + "learning_rate": 4.927374364434816e-05, + "loss": 3.2403, + "step": 5486 + }, + { + "epoch": 0.16925455660133643, + "grad_norm": 0.18428853154182434, + "learning_rate": 4.927309936866314e-05, + "loss": 3.2359, + "step": 5487 + }, + { + "epoch": 0.16928540306690984, + "grad_norm": 0.15633320808410645, + "learning_rate": 4.927245481154611e-05, + "loss": 3.23, + "step": 5488 + }, + { + "epoch": 0.16931624953248325, + "grad_norm": 0.18356038630008698, + "learning_rate": 4.927180997300456e-05, + "loss": 3.2634, + "step": 5489 + }, + { + "epoch": 0.16934709599805667, + "grad_norm": 0.1743268221616745, + "learning_rate": 4.927116485304596e-05, + "loss": 3.2543, + "step": 5490 + }, + { + "epoch": 0.16937794246363008, + "grad_norm": 0.17614096403121948, + "learning_rate": 4.927051945167778e-05, + "loss": 3.2557, + "step": 5491 + }, + { + "epoch": 0.1694087889292035, + "grad_norm": 0.21309420466423035, + "learning_rate": 4.9269873768907514e-05, + "loss": 3.2492, + "step": 5492 + }, + { + "epoch": 0.1694396353947769, + "grad_norm": 0.7734467387199402, + "learning_rate": 4.926922780474265e-05, + "loss": 3.2574, + "step": 5493 + }, + { + "epoch": 0.16947048186035032, + "grad_norm": 0.17950372397899628, + "learning_rate": 4.926858155919067e-05, + "loss": 3.2433, + "step": 5494 + }, + { + "epoch": 0.16950132832592377, + "grad_norm": 0.18244484066963196, + "learning_rate": 4.926793503225908e-05, + "loss": 3.2501, + "step": 5495 + }, + { + "epoch": 0.16953217479149718, + "grad_norm": 0.18050865828990936, + "learning_rate": 4.926728822395536e-05, + "loss": 3.2583, + "step": 5496 + }, + { + "epoch": 0.1695630212570706, + "grad_norm": 0.16541600227355957, + "learning_rate": 4.926664113428701e-05, + "loss": 3.2222, + "step": 5497 + }, + { + "epoch": 0.169593867722644, + "grad_norm": 0.1635267436504364, + "learning_rate": 4.9265993763261544e-05, + "loss": 3.2845, + "step": 5498 + }, + { + "epoch": 0.16962471418821742, + "grad_norm": 0.22023038566112518, + "learning_rate": 4.926534611088647e-05, + "loss": 3.2214, + "step": 5499 + }, + { + "epoch": 0.16965556065379084, + "grad_norm": 0.17181135714054108, + "learning_rate": 4.926469817716928e-05, + "loss": 3.2428, + "step": 5500 + }, + { + "epoch": 0.16968640711936425, + "grad_norm": 0.16974052786827087, + "learning_rate": 4.92640499621175e-05, + "loss": 3.2477, + "step": 5501 + }, + { + "epoch": 0.16971725358493767, + "grad_norm": 0.23929612338542938, + "learning_rate": 4.9263401465738636e-05, + "loss": 3.253, + "step": 5502 + }, + { + "epoch": 0.16974810005051108, + "grad_norm": 0.16887523233890533, + "learning_rate": 4.9262752688040215e-05, + "loss": 3.2287, + "step": 5503 + }, + { + "epoch": 0.1697789465160845, + "grad_norm": 0.1735258847475052, + "learning_rate": 4.926210362902976e-05, + "loss": 3.2086, + "step": 5504 + }, + { + "epoch": 0.1698097929816579, + "grad_norm": 0.1714419424533844, + "learning_rate": 4.9261454288714795e-05, + "loss": 3.2663, + "step": 5505 + }, + { + "epoch": 0.16984063944723135, + "grad_norm": 0.17644034326076508, + "learning_rate": 4.926080466710284e-05, + "loss": 3.2357, + "step": 5506 + }, + { + "epoch": 0.16987148591280476, + "grad_norm": 0.16153153777122498, + "learning_rate": 4.926015476420144e-05, + "loss": 3.2174, + "step": 5507 + }, + { + "epoch": 0.16990233237837818, + "grad_norm": 0.1701391041278839, + "learning_rate": 4.925950458001812e-05, + "loss": 3.2689, + "step": 5508 + }, + { + "epoch": 0.1699331788439516, + "grad_norm": 0.1622985154390335, + "learning_rate": 4.9258854114560425e-05, + "loss": 3.2441, + "step": 5509 + }, + { + "epoch": 0.169964025309525, + "grad_norm": 0.1587572991847992, + "learning_rate": 4.925820336783589e-05, + "loss": 3.263, + "step": 5510 + }, + { + "epoch": 0.16999487177509842, + "grad_norm": 0.16746772825717926, + "learning_rate": 4.925755233985207e-05, + "loss": 3.2499, + "step": 5511 + }, + { + "epoch": 0.17002571824067184, + "grad_norm": 0.1665186882019043, + "learning_rate": 4.925690103061651e-05, + "loss": 3.2316, + "step": 5512 + }, + { + "epoch": 0.17005656470624525, + "grad_norm": 0.17088228464126587, + "learning_rate": 4.925624944013675e-05, + "loss": 3.2572, + "step": 5513 + }, + { + "epoch": 0.17008741117181866, + "grad_norm": 0.16618993878364563, + "learning_rate": 4.925559756842036e-05, + "loss": 3.2432, + "step": 5514 + }, + { + "epoch": 0.17011825763739208, + "grad_norm": 0.17093425989151, + "learning_rate": 4.925494541547489e-05, + "loss": 3.2748, + "step": 5515 + }, + { + "epoch": 0.1701491041029655, + "grad_norm": 0.16891497373580933, + "learning_rate": 4.92542929813079e-05, + "loss": 3.2021, + "step": 5516 + }, + { + "epoch": 0.1701799505685389, + "grad_norm": 0.1701062172651291, + "learning_rate": 4.925364026592697e-05, + "loss": 3.2622, + "step": 5517 + }, + { + "epoch": 0.17021079703411235, + "grad_norm": 0.5384010672569275, + "learning_rate": 4.925298726933965e-05, + "loss": 3.2311, + "step": 5518 + }, + { + "epoch": 0.17024164349968576, + "grad_norm": 0.18710483610630035, + "learning_rate": 4.925233399155351e-05, + "loss": 3.2436, + "step": 5519 + }, + { + "epoch": 0.17027248996525918, + "grad_norm": 0.16488826274871826, + "learning_rate": 4.925168043257614e-05, + "loss": 3.2567, + "step": 5520 + }, + { + "epoch": 0.1703033364308326, + "grad_norm": 0.17604230344295502, + "learning_rate": 4.9251026592415105e-05, + "loss": 3.2654, + "step": 5521 + }, + { + "epoch": 0.170334182896406, + "grad_norm": 0.17223124206066132, + "learning_rate": 4.9250372471077986e-05, + "loss": 3.2455, + "step": 5522 + }, + { + "epoch": 0.17036502936197942, + "grad_norm": 0.18314115703105927, + "learning_rate": 4.924971806857238e-05, + "loss": 3.2307, + "step": 5523 + }, + { + "epoch": 0.17039587582755283, + "grad_norm": 0.19193525612354279, + "learning_rate": 4.924906338490586e-05, + "loss": 3.2529, + "step": 5524 + }, + { + "epoch": 0.17042672229312625, + "grad_norm": 0.18079963326454163, + "learning_rate": 4.924840842008602e-05, + "loss": 3.3088, + "step": 5525 + }, + { + "epoch": 0.17045756875869966, + "grad_norm": 0.16793200373649597, + "learning_rate": 4.924775317412047e-05, + "loss": 3.2654, + "step": 5526 + }, + { + "epoch": 0.17048841522427308, + "grad_norm": 0.22637256979942322, + "learning_rate": 4.924709764701678e-05, + "loss": 3.2316, + "step": 5527 + }, + { + "epoch": 0.1705192616898465, + "grad_norm": 0.16328346729278564, + "learning_rate": 4.9246441838782564e-05, + "loss": 3.2747, + "step": 5528 + }, + { + "epoch": 0.1705501081554199, + "grad_norm": 0.16767065227031708, + "learning_rate": 4.924578574942542e-05, + "loss": 3.208, + "step": 5529 + }, + { + "epoch": 0.17058095462099335, + "grad_norm": 0.1573832631111145, + "learning_rate": 4.924512937895297e-05, + "loss": 3.2383, + "step": 5530 + }, + { + "epoch": 0.17061180108656676, + "grad_norm": 0.19132335484027863, + "learning_rate": 4.9244472727372816e-05, + "loss": 3.2402, + "step": 5531 + }, + { + "epoch": 0.17064264755214018, + "grad_norm": 0.1618880033493042, + "learning_rate": 4.924381579469256e-05, + "loss": 3.2522, + "step": 5532 + }, + { + "epoch": 0.1706734940177136, + "grad_norm": 0.1839229017496109, + "learning_rate": 4.9243158580919835e-05, + "loss": 3.2183, + "step": 5533 + }, + { + "epoch": 0.170704340483287, + "grad_norm": 0.1744617223739624, + "learning_rate": 4.924250108606226e-05, + "loss": 3.2475, + "step": 5534 + }, + { + "epoch": 0.17073518694886042, + "grad_norm": 0.16469042003154755, + "learning_rate": 4.9241843310127444e-05, + "loss": 3.2262, + "step": 5535 + }, + { + "epoch": 0.17076603341443383, + "grad_norm": 0.16997839510440826, + "learning_rate": 4.924118525312302e-05, + "loss": 3.2318, + "step": 5536 + }, + { + "epoch": 0.17079687988000725, + "grad_norm": 0.16286039352416992, + "learning_rate": 4.9240526915056634e-05, + "loss": 3.2626, + "step": 5537 + }, + { + "epoch": 0.17082772634558066, + "grad_norm": 0.1625300645828247, + "learning_rate": 4.92398682959359e-05, + "loss": 3.2354, + "step": 5538 + }, + { + "epoch": 0.17085857281115407, + "grad_norm": 0.16041867434978485, + "learning_rate": 4.923920939576845e-05, + "loss": 3.2238, + "step": 5539 + }, + { + "epoch": 0.1708894192767275, + "grad_norm": 0.163848876953125, + "learning_rate": 4.923855021456194e-05, + "loss": 3.2586, + "step": 5540 + }, + { + "epoch": 0.1709202657423009, + "grad_norm": 0.17130720615386963, + "learning_rate": 4.923789075232401e-05, + "loss": 3.2219, + "step": 5541 + }, + { + "epoch": 0.17095111220787435, + "grad_norm": 0.16730085015296936, + "learning_rate": 4.92372310090623e-05, + "loss": 3.2764, + "step": 5542 + }, + { + "epoch": 0.17098195867344776, + "grad_norm": 0.17194291949272156, + "learning_rate": 4.923657098478446e-05, + "loss": 3.2432, + "step": 5543 + }, + { + "epoch": 0.17101280513902117, + "grad_norm": 0.16869519650936127, + "learning_rate": 4.923591067949814e-05, + "loss": 3.2529, + "step": 5544 + }, + { + "epoch": 0.1710436516045946, + "grad_norm": 0.1552906632423401, + "learning_rate": 4.9235250093211005e-05, + "loss": 3.2493, + "step": 5545 + }, + { + "epoch": 0.171074498070168, + "grad_norm": 0.16217263042926788, + "learning_rate": 4.9234589225930705e-05, + "loss": 3.2171, + "step": 5546 + }, + { + "epoch": 0.17110534453574142, + "grad_norm": 0.15547417104244232, + "learning_rate": 4.9233928077664905e-05, + "loss": 3.2594, + "step": 5547 + }, + { + "epoch": 0.17113619100131483, + "grad_norm": 0.16590429842472076, + "learning_rate": 4.9233266648421274e-05, + "loss": 3.2572, + "step": 5548 + }, + { + "epoch": 0.17116703746688824, + "grad_norm": 0.16018567979335785, + "learning_rate": 4.9232604938207483e-05, + "loss": 3.2348, + "step": 5549 + }, + { + "epoch": 0.17119788393246166, + "grad_norm": 0.15638624131679535, + "learning_rate": 4.92319429470312e-05, + "loss": 3.2613, + "step": 5550 + }, + { + "epoch": 0.17122873039803507, + "grad_norm": 0.16829760372638702, + "learning_rate": 4.9231280674900094e-05, + "loss": 3.2503, + "step": 5551 + }, + { + "epoch": 0.1712595768636085, + "grad_norm": 0.16028040647506714, + "learning_rate": 4.923061812182185e-05, + "loss": 3.2307, + "step": 5552 + }, + { + "epoch": 0.1712904233291819, + "grad_norm": 0.181573748588562, + "learning_rate": 4.9229955287804155e-05, + "loss": 3.247, + "step": 5553 + }, + { + "epoch": 0.17132126979475534, + "grad_norm": 0.1598082035779953, + "learning_rate": 4.922929217285468e-05, + "loss": 3.2281, + "step": 5554 + }, + { + "epoch": 0.17135211626032876, + "grad_norm": 0.16809159517288208, + "learning_rate": 4.9228628776981135e-05, + "loss": 3.238, + "step": 5555 + }, + { + "epoch": 0.17138296272590217, + "grad_norm": 0.16950859129428864, + "learning_rate": 4.9227965100191185e-05, + "loss": 3.2232, + "step": 5556 + }, + { + "epoch": 0.17141380919147559, + "grad_norm": 0.16116948425769806, + "learning_rate": 4.9227301142492556e-05, + "loss": 3.2127, + "step": 5557 + }, + { + "epoch": 0.171444655657049, + "grad_norm": 0.16271495819091797, + "learning_rate": 4.922663690389291e-05, + "loss": 3.2237, + "step": 5558 + }, + { + "epoch": 0.17147550212262241, + "grad_norm": 0.16164690256118774, + "learning_rate": 4.922597238439998e-05, + "loss": 3.2695, + "step": 5559 + }, + { + "epoch": 0.17150634858819583, + "grad_norm": 0.1625596135854721, + "learning_rate": 4.9225307584021455e-05, + "loss": 3.247, + "step": 5560 + }, + { + "epoch": 0.17153719505376924, + "grad_norm": 0.18244145810604095, + "learning_rate": 4.922464250276505e-05, + "loss": 3.2598, + "step": 5561 + }, + { + "epoch": 0.17156804151934266, + "grad_norm": 0.15659435093402863, + "learning_rate": 4.922397714063847e-05, + "loss": 3.2508, + "step": 5562 + }, + { + "epoch": 0.17159888798491607, + "grad_norm": 0.1569335162639618, + "learning_rate": 4.9223311497649424e-05, + "loss": 3.2243, + "step": 5563 + }, + { + "epoch": 0.17162973445048949, + "grad_norm": 0.1557430624961853, + "learning_rate": 4.9222645573805646e-05, + "loss": 3.2422, + "step": 5564 + }, + { + "epoch": 0.1716605809160629, + "grad_norm": 0.17298273742198944, + "learning_rate": 4.9221979369114854e-05, + "loss": 3.2513, + "step": 5565 + }, + { + "epoch": 0.17169142738163634, + "grad_norm": 0.17457899451255798, + "learning_rate": 4.922131288358476e-05, + "loss": 3.2952, + "step": 5566 + }, + { + "epoch": 0.17172227384720976, + "grad_norm": 0.16601890325546265, + "learning_rate": 4.92206461172231e-05, + "loss": 3.2876, + "step": 5567 + }, + { + "epoch": 0.17175312031278317, + "grad_norm": 0.17166993021965027, + "learning_rate": 4.9219979070037606e-05, + "loss": 3.2123, + "step": 5568 + }, + { + "epoch": 0.17178396677835658, + "grad_norm": 0.16467584669589996, + "learning_rate": 4.9219311742036e-05, + "loss": 3.2386, + "step": 5569 + }, + { + "epoch": 0.17181481324393, + "grad_norm": 0.1672755479812622, + "learning_rate": 4.9218644133226035e-05, + "loss": 3.2128, + "step": 5570 + }, + { + "epoch": 0.1718456597095034, + "grad_norm": 0.1715662032365799, + "learning_rate": 4.921797624361544e-05, + "loss": 3.2731, + "step": 5571 + }, + { + "epoch": 0.17187650617507683, + "grad_norm": 0.1599506139755249, + "learning_rate": 4.9217308073211974e-05, + "loss": 3.2367, + "step": 5572 + }, + { + "epoch": 0.17190735264065024, + "grad_norm": 0.16138838231563568, + "learning_rate": 4.921663962202337e-05, + "loss": 3.2466, + "step": 5573 + }, + { + "epoch": 0.17193819910622365, + "grad_norm": 0.158944770693779, + "learning_rate": 4.921597089005737e-05, + "loss": 3.2473, + "step": 5574 + }, + { + "epoch": 0.17196904557179707, + "grad_norm": 0.15880084037780762, + "learning_rate": 4.9215301877321753e-05, + "loss": 3.2566, + "step": 5575 + }, + { + "epoch": 0.17199989203737048, + "grad_norm": 0.16381363570690155, + "learning_rate": 4.921463258382426e-05, + "loss": 3.2576, + "step": 5576 + }, + { + "epoch": 0.1720307385029439, + "grad_norm": 0.15560594201087952, + "learning_rate": 4.921396300957265e-05, + "loss": 3.2479, + "step": 5577 + }, + { + "epoch": 0.17206158496851734, + "grad_norm": 0.16565269231796265, + "learning_rate": 4.921329315457469e-05, + "loss": 3.2661, + "step": 5578 + }, + { + "epoch": 0.17209243143409075, + "grad_norm": 0.1628248393535614, + "learning_rate": 4.9212623018838155e-05, + "loss": 3.2304, + "step": 5579 + }, + { + "epoch": 0.17212327789966417, + "grad_norm": 0.15923869609832764, + "learning_rate": 4.92119526023708e-05, + "loss": 3.2824, + "step": 5580 + }, + { + "epoch": 0.17215412436523758, + "grad_norm": 0.15636910498142242, + "learning_rate": 4.92112819051804e-05, + "loss": 3.2466, + "step": 5581 + }, + { + "epoch": 0.172184970830811, + "grad_norm": 0.16353723406791687, + "learning_rate": 4.921061092727474e-05, + "loss": 3.2649, + "step": 5582 + }, + { + "epoch": 0.1722158172963844, + "grad_norm": 0.16018183529376984, + "learning_rate": 4.9209939668661596e-05, + "loss": 3.2502, + "step": 5583 + }, + { + "epoch": 0.17224666376195782, + "grad_norm": 0.17488808929920197, + "learning_rate": 4.920926812934875e-05, + "loss": 3.2482, + "step": 5584 + }, + { + "epoch": 0.17227751022753124, + "grad_norm": 0.15651226043701172, + "learning_rate": 4.9208596309343985e-05, + "loss": 3.2743, + "step": 5585 + }, + { + "epoch": 0.17230835669310465, + "grad_norm": 0.16217747330665588, + "learning_rate": 4.92079242086551e-05, + "loss": 3.2361, + "step": 5586 + }, + { + "epoch": 0.17233920315867807, + "grad_norm": 0.15962299704551697, + "learning_rate": 4.920725182728987e-05, + "loss": 3.2408, + "step": 5587 + }, + { + "epoch": 0.17237004962425148, + "grad_norm": 0.17051124572753906, + "learning_rate": 4.9206579165256114e-05, + "loss": 3.2413, + "step": 5588 + }, + { + "epoch": 0.1724008960898249, + "grad_norm": 0.16652680933475494, + "learning_rate": 4.920590622256161e-05, + "loss": 3.2354, + "step": 5589 + }, + { + "epoch": 0.17243174255539834, + "grad_norm": 0.16126345098018646, + "learning_rate": 4.9205232999214175e-05, + "loss": 3.2972, + "step": 5590 + }, + { + "epoch": 0.17246258902097175, + "grad_norm": 0.1655939519405365, + "learning_rate": 4.92045594952216e-05, + "loss": 3.2749, + "step": 5591 + }, + { + "epoch": 0.17249343548654517, + "grad_norm": 0.1643146127462387, + "learning_rate": 4.920388571059171e-05, + "loss": 3.2544, + "step": 5592 + }, + { + "epoch": 0.17252428195211858, + "grad_norm": 0.15452904999256134, + "learning_rate": 4.920321164533231e-05, + "loss": 3.2429, + "step": 5593 + }, + { + "epoch": 0.172555128417692, + "grad_norm": 0.33788999915122986, + "learning_rate": 4.9202537299451215e-05, + "loss": 3.2131, + "step": 5594 + }, + { + "epoch": 0.1725859748832654, + "grad_norm": 0.25616708397865295, + "learning_rate": 4.920186267295624e-05, + "loss": 3.2777, + "step": 5595 + }, + { + "epoch": 0.17261682134883882, + "grad_norm": 0.17731039226055145, + "learning_rate": 4.9201187765855224e-05, + "loss": 3.2601, + "step": 5596 + }, + { + "epoch": 0.17264766781441224, + "grad_norm": 0.1643645018339157, + "learning_rate": 4.920051257815597e-05, + "loss": 3.2515, + "step": 5597 + }, + { + "epoch": 0.17267851427998565, + "grad_norm": 0.1879621148109436, + "learning_rate": 4.919983710986632e-05, + "loss": 3.2357, + "step": 5598 + }, + { + "epoch": 0.17270936074555907, + "grad_norm": 0.1641184240579605, + "learning_rate": 4.919916136099409e-05, + "loss": 3.2344, + "step": 5599 + }, + { + "epoch": 0.17274020721113248, + "grad_norm": 0.18150295317173004, + "learning_rate": 4.919848533154713e-05, + "loss": 3.2366, + "step": 5600 + }, + { + "epoch": 0.1727710536767059, + "grad_norm": 0.2605171799659729, + "learning_rate": 4.919780902153328e-05, + "loss": 3.2469, + "step": 5601 + }, + { + "epoch": 0.17280190014227934, + "grad_norm": 0.7542741298675537, + "learning_rate": 4.919713243096037e-05, + "loss": 3.278, + "step": 5602 + }, + { + "epoch": 0.17283274660785275, + "grad_norm": 0.17916887998580933, + "learning_rate": 4.9196455559836255e-05, + "loss": 3.2078, + "step": 5603 + }, + { + "epoch": 0.17286359307342616, + "grad_norm": 0.17293444275856018, + "learning_rate": 4.919577840816878e-05, + "loss": 3.2395, + "step": 5604 + }, + { + "epoch": 0.17289443953899958, + "grad_norm": 0.1851482093334198, + "learning_rate": 4.919510097596579e-05, + "loss": 3.2629, + "step": 5605 + }, + { + "epoch": 0.172925286004573, + "grad_norm": 0.18134455382823944, + "learning_rate": 4.919442326323515e-05, + "loss": 3.2801, + "step": 5606 + }, + { + "epoch": 0.1729561324701464, + "grad_norm": 0.17397987842559814, + "learning_rate": 4.91937452699847e-05, + "loss": 3.2418, + "step": 5607 + }, + { + "epoch": 0.17298697893571982, + "grad_norm": 0.18992166221141815, + "learning_rate": 4.919306699622231e-05, + "loss": 3.2446, + "step": 5608 + }, + { + "epoch": 0.17301782540129323, + "grad_norm": 0.17248409986495972, + "learning_rate": 4.919238844195586e-05, + "loss": 3.2647, + "step": 5609 + }, + { + "epoch": 0.17304867186686665, + "grad_norm": 0.18140846490859985, + "learning_rate": 4.91917096071932e-05, + "loss": 3.2275, + "step": 5610 + }, + { + "epoch": 0.17307951833244006, + "grad_norm": 0.1666371077299118, + "learning_rate": 4.91910304919422e-05, + "loss": 3.2017, + "step": 5611 + }, + { + "epoch": 0.17311036479801348, + "grad_norm": 0.1866127848625183, + "learning_rate": 4.919035109621074e-05, + "loss": 3.2294, + "step": 5612 + }, + { + "epoch": 0.1731412112635869, + "grad_norm": 0.18153680860996246, + "learning_rate": 4.91896714200067e-05, + "loss": 3.278, + "step": 5613 + }, + { + "epoch": 0.17317205772916033, + "grad_norm": 0.17260168492794037, + "learning_rate": 4.9188991463337954e-05, + "loss": 3.2613, + "step": 5614 + }, + { + "epoch": 0.17320290419473375, + "grad_norm": 0.1810598075389862, + "learning_rate": 4.918831122621239e-05, + "loss": 3.2646, + "step": 5615 + }, + { + "epoch": 0.17323375066030716, + "grad_norm": 0.1905767321586609, + "learning_rate": 4.9187630708637886e-05, + "loss": 3.263, + "step": 5616 + }, + { + "epoch": 0.17326459712588058, + "grad_norm": 0.18559877574443817, + "learning_rate": 4.918694991062234e-05, + "loss": 3.2478, + "step": 5617 + }, + { + "epoch": 0.173295443591454, + "grad_norm": 0.17006038129329681, + "learning_rate": 4.918626883217365e-05, + "loss": 3.2315, + "step": 5618 + }, + { + "epoch": 0.1733262900570274, + "grad_norm": 0.18757474422454834, + "learning_rate": 4.918558747329971e-05, + "loss": 3.2489, + "step": 5619 + }, + { + "epoch": 0.17335713652260082, + "grad_norm": 0.2323518842458725, + "learning_rate": 4.918490583400841e-05, + "loss": 3.2763, + "step": 5620 + }, + { + "epoch": 0.17338798298817423, + "grad_norm": 0.16822831332683563, + "learning_rate": 4.918422391430766e-05, + "loss": 3.2352, + "step": 5621 + }, + { + "epoch": 0.17341882945374765, + "grad_norm": 0.16566374897956848, + "learning_rate": 4.9183541714205374e-05, + "loss": 3.2408, + "step": 5622 + }, + { + "epoch": 0.17344967591932106, + "grad_norm": 0.21193082630634308, + "learning_rate": 4.918285923370944e-05, + "loss": 3.2546, + "step": 5623 + }, + { + "epoch": 0.17348052238489448, + "grad_norm": 0.16889400780200958, + "learning_rate": 4.9182176472827804e-05, + "loss": 3.2585, + "step": 5624 + }, + { + "epoch": 0.1735113688504679, + "grad_norm": 0.15866480767726898, + "learning_rate": 4.9181493431568346e-05, + "loss": 3.2512, + "step": 5625 + }, + { + "epoch": 0.17354221531604133, + "grad_norm": 0.16283276677131653, + "learning_rate": 4.918081010993901e-05, + "loss": 3.2638, + "step": 5626 + }, + { + "epoch": 0.17357306178161475, + "grad_norm": 0.1575610488653183, + "learning_rate": 4.918012650794771e-05, + "loss": 3.2577, + "step": 5627 + }, + { + "epoch": 0.17360390824718816, + "grad_norm": 0.17074809968471527, + "learning_rate": 4.917944262560239e-05, + "loss": 3.2924, + "step": 5628 + }, + { + "epoch": 0.17363475471276157, + "grad_norm": 0.15431946516036987, + "learning_rate": 4.917875846291095e-05, + "loss": 3.213, + "step": 5629 + }, + { + "epoch": 0.173665601178335, + "grad_norm": 0.15913432836532593, + "learning_rate": 4.9178074019881325e-05, + "loss": 3.2445, + "step": 5630 + }, + { + "epoch": 0.1736964476439084, + "grad_norm": 0.15690810978412628, + "learning_rate": 4.917738929652148e-05, + "loss": 3.2361, + "step": 5631 + }, + { + "epoch": 0.17372729410948182, + "grad_norm": 0.15811942517757416, + "learning_rate": 4.917670429283932e-05, + "loss": 3.2454, + "step": 5632 + }, + { + "epoch": 0.17375814057505523, + "grad_norm": 0.1580636203289032, + "learning_rate": 4.917601900884281e-05, + "loss": 3.2712, + "step": 5633 + }, + { + "epoch": 0.17378898704062865, + "grad_norm": 0.1566624492406845, + "learning_rate": 4.9175333444539886e-05, + "loss": 3.2852, + "step": 5634 + }, + { + "epoch": 0.17381983350620206, + "grad_norm": 0.1545908898115158, + "learning_rate": 4.917464759993849e-05, + "loss": 3.2028, + "step": 5635 + }, + { + "epoch": 0.17385067997177547, + "grad_norm": 0.1614721417427063, + "learning_rate": 4.91739614750466e-05, + "loss": 3.2277, + "step": 5636 + }, + { + "epoch": 0.1738815264373489, + "grad_norm": 0.15552574396133423, + "learning_rate": 4.9173275069872145e-05, + "loss": 3.2104, + "step": 5637 + }, + { + "epoch": 0.17391237290292233, + "grad_norm": 0.16885705292224884, + "learning_rate": 4.917258838442309e-05, + "loss": 3.2443, + "step": 5638 + }, + { + "epoch": 0.17394321936849574, + "grad_norm": 0.15964674949645996, + "learning_rate": 4.91719014187074e-05, + "loss": 3.2614, + "step": 5639 + }, + { + "epoch": 0.17397406583406916, + "grad_norm": 0.1596319079399109, + "learning_rate": 4.9171214172733036e-05, + "loss": 3.2404, + "step": 5640 + }, + { + "epoch": 0.17400491229964257, + "grad_norm": 0.1535930037498474, + "learning_rate": 4.917052664650798e-05, + "loss": 3.208, + "step": 5641 + }, + { + "epoch": 0.174035758765216, + "grad_norm": 0.15444503724575043, + "learning_rate": 4.916983884004018e-05, + "loss": 3.2601, + "step": 5642 + }, + { + "epoch": 0.1740666052307894, + "grad_norm": 0.15615436434745789, + "learning_rate": 4.9169150753337625e-05, + "loss": 3.2572, + "step": 5643 + }, + { + "epoch": 0.17409745169636281, + "grad_norm": 0.38937458395957947, + "learning_rate": 4.916846238640829e-05, + "loss": 3.2177, + "step": 5644 + }, + { + "epoch": 0.17412829816193623, + "grad_norm": 0.1606808751821518, + "learning_rate": 4.916777373926016e-05, + "loss": 3.2133, + "step": 5645 + }, + { + "epoch": 0.17415914462750964, + "grad_norm": 0.23600997030735016, + "learning_rate": 4.916708481190122e-05, + "loss": 3.2335, + "step": 5646 + }, + { + "epoch": 0.17418999109308306, + "grad_norm": 0.17866362631320953, + "learning_rate": 4.916639560433945e-05, + "loss": 3.2449, + "step": 5647 + }, + { + "epoch": 0.17422083755865647, + "grad_norm": 0.16141939163208008, + "learning_rate": 4.916570611658285e-05, + "loss": 3.2551, + "step": 5648 + }, + { + "epoch": 0.17425168402422989, + "grad_norm": 0.17272594571113586, + "learning_rate": 4.91650163486394e-05, + "loss": 3.2527, + "step": 5649 + }, + { + "epoch": 0.17428253048980333, + "grad_norm": 0.15898382663726807, + "learning_rate": 4.9164326300517116e-05, + "loss": 3.2576, + "step": 5650 + }, + { + "epoch": 0.17431337695537674, + "grad_norm": 0.1702224314212799, + "learning_rate": 4.9163635972223984e-05, + "loss": 3.2218, + "step": 5651 + }, + { + "epoch": 0.17434422342095016, + "grad_norm": 0.15665078163146973, + "learning_rate": 4.916294536376802e-05, + "loss": 3.2287, + "step": 5652 + }, + { + "epoch": 0.17437506988652357, + "grad_norm": 0.16273769736289978, + "learning_rate": 4.916225447515722e-05, + "loss": 3.2548, + "step": 5653 + }, + { + "epoch": 0.17440591635209698, + "grad_norm": 0.16716790199279785, + "learning_rate": 4.9161563306399606e-05, + "loss": 3.2639, + "step": 5654 + }, + { + "epoch": 0.1744367628176704, + "grad_norm": 0.1630450338125229, + "learning_rate": 4.9160871857503175e-05, + "loss": 3.2691, + "step": 5655 + }, + { + "epoch": 0.1744676092832438, + "grad_norm": 0.15389300882816315, + "learning_rate": 4.916018012847596e-05, + "loss": 3.2926, + "step": 5656 + }, + { + "epoch": 0.17449845574881723, + "grad_norm": 0.16631849110126495, + "learning_rate": 4.9159488119325966e-05, + "loss": 3.245, + "step": 5657 + }, + { + "epoch": 0.17452930221439064, + "grad_norm": 0.16834397614002228, + "learning_rate": 4.915879583006123e-05, + "loss": 3.2018, + "step": 5658 + }, + { + "epoch": 0.17456014867996406, + "grad_norm": 0.1594254970550537, + "learning_rate": 4.915810326068978e-05, + "loss": 3.2797, + "step": 5659 + }, + { + "epoch": 0.17459099514553747, + "grad_norm": 0.15773820877075195, + "learning_rate": 4.915741041121964e-05, + "loss": 3.275, + "step": 5660 + }, + { + "epoch": 0.17462184161111088, + "grad_norm": 0.16824163496494293, + "learning_rate": 4.9156717281658834e-05, + "loss": 3.2765, + "step": 5661 + }, + { + "epoch": 0.17465268807668433, + "grad_norm": 0.17333295941352844, + "learning_rate": 4.9156023872015416e-05, + "loss": 3.2676, + "step": 5662 + }, + { + "epoch": 0.17468353454225774, + "grad_norm": 0.16442686319351196, + "learning_rate": 4.915533018229742e-05, + "loss": 3.2597, + "step": 5663 + }, + { + "epoch": 0.17471438100783115, + "grad_norm": 0.17759975790977478, + "learning_rate": 4.915463621251287e-05, + "loss": 3.2418, + "step": 5664 + }, + { + "epoch": 0.17474522747340457, + "grad_norm": 0.15852299332618713, + "learning_rate": 4.915394196266985e-05, + "loss": 3.2334, + "step": 5665 + }, + { + "epoch": 0.17477607393897798, + "grad_norm": 0.1614086776971817, + "learning_rate": 4.915324743277637e-05, + "loss": 3.2322, + "step": 5666 + }, + { + "epoch": 0.1748069204045514, + "grad_norm": 0.1654549390077591, + "learning_rate": 4.915255262284051e-05, + "loss": 3.262, + "step": 5667 + }, + { + "epoch": 0.1748377668701248, + "grad_norm": 0.17205996811389923, + "learning_rate": 4.915185753287031e-05, + "loss": 3.2518, + "step": 5668 + }, + { + "epoch": 0.17486861333569823, + "grad_norm": 0.1491578370332718, + "learning_rate": 4.9151162162873845e-05, + "loss": 3.2342, + "step": 5669 + }, + { + "epoch": 0.17489945980127164, + "grad_norm": 0.20873326063156128, + "learning_rate": 4.915046651285916e-05, + "loss": 3.2455, + "step": 5670 + }, + { + "epoch": 0.17493030626684505, + "grad_norm": 0.16318689286708832, + "learning_rate": 4.914977058283433e-05, + "loss": 3.2015, + "step": 5671 + }, + { + "epoch": 0.17496115273241847, + "grad_norm": 0.16300688683986664, + "learning_rate": 4.9149074372807426e-05, + "loss": 3.256, + "step": 5672 + }, + { + "epoch": 0.17499199919799188, + "grad_norm": 0.1616784930229187, + "learning_rate": 4.914837788278651e-05, + "loss": 3.2608, + "step": 5673 + }, + { + "epoch": 0.17502284566356532, + "grad_norm": 0.1693800389766693, + "learning_rate": 4.9147681112779664e-05, + "loss": 3.2734, + "step": 5674 + }, + { + "epoch": 0.17505369212913874, + "grad_norm": 0.15577176213264465, + "learning_rate": 4.914698406279497e-05, + "loss": 3.2696, + "step": 5675 + }, + { + "epoch": 0.17508453859471215, + "grad_norm": 0.16944487392902374, + "learning_rate": 4.914628673284051e-05, + "loss": 3.2427, + "step": 5676 + }, + { + "epoch": 0.17511538506028557, + "grad_norm": 0.19537094235420227, + "learning_rate": 4.9145589122924365e-05, + "loss": 3.2463, + "step": 5677 + }, + { + "epoch": 0.17514623152585898, + "grad_norm": 0.16046921908855438, + "learning_rate": 4.914489123305461e-05, + "loss": 3.2429, + "step": 5678 + }, + { + "epoch": 0.1751770779914324, + "grad_norm": 0.1722228080034256, + "learning_rate": 4.914419306323936e-05, + "loss": 3.2922, + "step": 5679 + }, + { + "epoch": 0.1752079244570058, + "grad_norm": 0.2147040069103241, + "learning_rate": 4.9143494613486704e-05, + "loss": 3.2752, + "step": 5680 + }, + { + "epoch": 0.17523877092257922, + "grad_norm": 0.15917521715164185, + "learning_rate": 4.914279588380474e-05, + "loss": 3.2555, + "step": 5681 + }, + { + "epoch": 0.17526961738815264, + "grad_norm": 0.18410907685756683, + "learning_rate": 4.914209687420156e-05, + "loss": 3.2483, + "step": 5682 + }, + { + "epoch": 0.17530046385372605, + "grad_norm": 0.17522354423999786, + "learning_rate": 4.914139758468527e-05, + "loss": 3.2426, + "step": 5683 + }, + { + "epoch": 0.17533131031929947, + "grad_norm": 0.16399383544921875, + "learning_rate": 4.9140698015263984e-05, + "loss": 3.2604, + "step": 5684 + }, + { + "epoch": 0.17536215678487288, + "grad_norm": 0.1751384437084198, + "learning_rate": 4.9139998165945825e-05, + "loss": 3.2457, + "step": 5685 + }, + { + "epoch": 0.17539300325044632, + "grad_norm": 0.16357460618019104, + "learning_rate": 4.913929803673888e-05, + "loss": 3.2387, + "step": 5686 + }, + { + "epoch": 0.17542384971601974, + "grad_norm": 0.16695651412010193, + "learning_rate": 4.913859762765128e-05, + "loss": 3.2088, + "step": 5687 + }, + { + "epoch": 0.17545469618159315, + "grad_norm": 0.16421334445476532, + "learning_rate": 4.913789693869115e-05, + "loss": 3.2579, + "step": 5688 + }, + { + "epoch": 0.17548554264716656, + "grad_norm": 0.1609276384115219, + "learning_rate": 4.913719596986661e-05, + "loss": 3.2458, + "step": 5689 + }, + { + "epoch": 0.17551638911273998, + "grad_norm": 0.15886656939983368, + "learning_rate": 4.913649472118579e-05, + "loss": 3.2184, + "step": 5690 + }, + { + "epoch": 0.1755472355783134, + "grad_norm": 0.8230074048042297, + "learning_rate": 4.913579319265681e-05, + "loss": 3.2729, + "step": 5691 + }, + { + "epoch": 0.1755780820438868, + "grad_norm": 0.16901808977127075, + "learning_rate": 4.913509138428782e-05, + "loss": 3.2577, + "step": 5692 + }, + { + "epoch": 0.17560892850946022, + "grad_norm": 0.1710551381111145, + "learning_rate": 4.9134389296086945e-05, + "loss": 3.2208, + "step": 5693 + }, + { + "epoch": 0.17563977497503364, + "grad_norm": 0.16632448136806488, + "learning_rate": 4.913368692806233e-05, + "loss": 3.2252, + "step": 5694 + }, + { + "epoch": 0.17567062144060705, + "grad_norm": 0.17842072248458862, + "learning_rate": 4.9132984280222116e-05, + "loss": 3.2825, + "step": 5695 + }, + { + "epoch": 0.17570146790618046, + "grad_norm": 0.24512025713920593, + "learning_rate": 4.9132281352574454e-05, + "loss": 3.2338, + "step": 5696 + }, + { + "epoch": 0.17573231437175388, + "grad_norm": 0.17266690731048584, + "learning_rate": 4.91315781451275e-05, + "loss": 3.221, + "step": 5697 + }, + { + "epoch": 0.17576316083732732, + "grad_norm": 0.1723242551088333, + "learning_rate": 4.913087465788938e-05, + "loss": 3.2318, + "step": 5698 + }, + { + "epoch": 0.17579400730290073, + "grad_norm": 0.187696635723114, + "learning_rate": 4.913017089086828e-05, + "loss": 3.2228, + "step": 5699 + }, + { + "epoch": 0.17582485376847415, + "grad_norm": 0.1770828515291214, + "learning_rate": 4.912946684407235e-05, + "loss": 3.2425, + "step": 5700 + }, + { + "epoch": 0.17585570023404756, + "grad_norm": 0.17437750101089478, + "learning_rate": 4.9128762517509754e-05, + "loss": 3.27, + "step": 5701 + }, + { + "epoch": 0.17588654669962098, + "grad_norm": 0.2229948192834854, + "learning_rate": 4.912805791118865e-05, + "loss": 3.2301, + "step": 5702 + }, + { + "epoch": 0.1759173931651944, + "grad_norm": 0.1589549481868744, + "learning_rate": 4.9127353025117215e-05, + "loss": 3.2593, + "step": 5703 + }, + { + "epoch": 0.1759482396307678, + "grad_norm": 0.16855798661708832, + "learning_rate": 4.912664785930362e-05, + "loss": 3.2397, + "step": 5704 + }, + { + "epoch": 0.17597908609634122, + "grad_norm": 0.17196980118751526, + "learning_rate": 4.9125942413756043e-05, + "loss": 3.2724, + "step": 5705 + }, + { + "epoch": 0.17600993256191463, + "grad_norm": 0.16570019721984863, + "learning_rate": 4.9125236688482664e-05, + "loss": 3.2105, + "step": 5706 + }, + { + "epoch": 0.17604077902748805, + "grad_norm": 0.1722661405801773, + "learning_rate": 4.912453068349166e-05, + "loss": 3.2429, + "step": 5707 + }, + { + "epoch": 0.17607162549306146, + "grad_norm": 0.16572482883930206, + "learning_rate": 4.912382439879122e-05, + "loss": 3.2343, + "step": 5708 + }, + { + "epoch": 0.17610247195863488, + "grad_norm": 0.17256003618240356, + "learning_rate": 4.9123117834389534e-05, + "loss": 3.2538, + "step": 5709 + }, + { + "epoch": 0.17613331842420832, + "grad_norm": 0.15473000705242157, + "learning_rate": 4.912241099029479e-05, + "loss": 3.291, + "step": 5710 + }, + { + "epoch": 0.17616416488978173, + "grad_norm": 0.19195546209812164, + "learning_rate": 4.912170386651519e-05, + "loss": 3.2648, + "step": 5711 + }, + { + "epoch": 0.17619501135535515, + "grad_norm": 0.17019899189472198, + "learning_rate": 4.9120996463058935e-05, + "loss": 3.2558, + "step": 5712 + }, + { + "epoch": 0.17622585782092856, + "grad_norm": 0.1578134298324585, + "learning_rate": 4.912028877993421e-05, + "loss": 3.2296, + "step": 5713 + }, + { + "epoch": 0.17625670428650198, + "grad_norm": 0.16616804897785187, + "learning_rate": 4.9119580817149234e-05, + "loss": 3.2336, + "step": 5714 + }, + { + "epoch": 0.1762875507520754, + "grad_norm": 0.16951599717140198, + "learning_rate": 4.911887257471222e-05, + "loss": 3.2526, + "step": 5715 + }, + { + "epoch": 0.1763183972176488, + "grad_norm": 0.16005420684814453, + "learning_rate": 4.911816405263136e-05, + "loss": 3.2285, + "step": 5716 + }, + { + "epoch": 0.17634924368322222, + "grad_norm": 0.16145475208759308, + "learning_rate": 4.911745525091489e-05, + "loss": 3.2476, + "step": 5717 + }, + { + "epoch": 0.17638009014879563, + "grad_norm": 0.17069247364997864, + "learning_rate": 4.911674616957102e-05, + "loss": 3.2444, + "step": 5718 + }, + { + "epoch": 0.17641093661436905, + "grad_norm": 0.170532688498497, + "learning_rate": 4.911603680860797e-05, + "loss": 3.2439, + "step": 5719 + }, + { + "epoch": 0.17644178307994246, + "grad_norm": 0.16335774958133698, + "learning_rate": 4.911532716803396e-05, + "loss": 3.2292, + "step": 5720 + }, + { + "epoch": 0.17647262954551587, + "grad_norm": 0.16207283735275269, + "learning_rate": 4.911461724785723e-05, + "loss": 3.2295, + "step": 5721 + }, + { + "epoch": 0.17650347601108932, + "grad_norm": 0.17208975553512573, + "learning_rate": 4.9113907048085996e-05, + "loss": 3.2275, + "step": 5722 + }, + { + "epoch": 0.17653432247666273, + "grad_norm": 0.16152454912662506, + "learning_rate": 4.91131965687285e-05, + "loss": 3.2289, + "step": 5723 + }, + { + "epoch": 0.17656516894223614, + "grad_norm": 0.15700168907642365, + "learning_rate": 4.911248580979299e-05, + "loss": 3.2363, + "step": 5724 + }, + { + "epoch": 0.17659601540780956, + "grad_norm": 0.1616014838218689, + "learning_rate": 4.911177477128769e-05, + "loss": 3.2315, + "step": 5725 + }, + { + "epoch": 0.17662686187338297, + "grad_norm": 0.17593953013420105, + "learning_rate": 4.911106345322085e-05, + "loss": 3.2549, + "step": 5726 + }, + { + "epoch": 0.1766577083389564, + "grad_norm": 0.15879233181476593, + "learning_rate": 4.911035185560072e-05, + "loss": 3.2445, + "step": 5727 + }, + { + "epoch": 0.1766885548045298, + "grad_norm": 0.16273759305477142, + "learning_rate": 4.910963997843555e-05, + "loss": 3.2213, + "step": 5728 + }, + { + "epoch": 0.17671940127010322, + "grad_norm": 0.16145943105220795, + "learning_rate": 4.91089278217336e-05, + "loss": 3.2277, + "step": 5729 + }, + { + "epoch": 0.17675024773567663, + "grad_norm": 0.1687065064907074, + "learning_rate": 4.910821538550311e-05, + "loss": 3.2027, + "step": 5730 + }, + { + "epoch": 0.17678109420125004, + "grad_norm": 0.82180255651474, + "learning_rate": 4.910750266975235e-05, + "loss": 3.2658, + "step": 5731 + }, + { + "epoch": 0.17681194066682346, + "grad_norm": 0.16572093963623047, + "learning_rate": 4.9106789674489583e-05, + "loss": 3.2333, + "step": 5732 + }, + { + "epoch": 0.17684278713239687, + "grad_norm": 0.17173241078853607, + "learning_rate": 4.910607639972308e-05, + "loss": 3.2558, + "step": 5733 + }, + { + "epoch": 0.17687363359797031, + "grad_norm": 0.15733902156352997, + "learning_rate": 4.9105362845461114e-05, + "loss": 3.244, + "step": 5734 + }, + { + "epoch": 0.17690448006354373, + "grad_norm": 0.1670921891927719, + "learning_rate": 4.9104649011711936e-05, + "loss": 3.2501, + "step": 5735 + }, + { + "epoch": 0.17693532652911714, + "grad_norm": 0.1614975929260254, + "learning_rate": 4.910393489848385e-05, + "loss": 3.2339, + "step": 5736 + }, + { + "epoch": 0.17696617299469056, + "grad_norm": 0.16981469094753265, + "learning_rate": 4.910322050578512e-05, + "loss": 3.2521, + "step": 5737 + }, + { + "epoch": 0.17699701946026397, + "grad_norm": 0.16618824005126953, + "learning_rate": 4.910250583362403e-05, + "loss": 3.2448, + "step": 5738 + }, + { + "epoch": 0.17702786592583739, + "grad_norm": 0.16570109128952026, + "learning_rate": 4.9101790882008866e-05, + "loss": 3.2572, + "step": 5739 + }, + { + "epoch": 0.1770587123914108, + "grad_norm": 0.17198839783668518, + "learning_rate": 4.910107565094793e-05, + "loss": 3.2627, + "step": 5740 + }, + { + "epoch": 0.1770895588569842, + "grad_norm": 0.5115522146224976, + "learning_rate": 4.91003601404495e-05, + "loss": 3.2296, + "step": 5741 + }, + { + "epoch": 0.17712040532255763, + "grad_norm": 0.17962709069252014, + "learning_rate": 4.909964435052188e-05, + "loss": 3.265, + "step": 5742 + }, + { + "epoch": 0.17715125178813104, + "grad_norm": 0.17606720328330994, + "learning_rate": 4.909892828117336e-05, + "loss": 3.2265, + "step": 5743 + }, + { + "epoch": 0.17718209825370446, + "grad_norm": 0.1800462305545807, + "learning_rate": 4.909821193241225e-05, + "loss": 3.2992, + "step": 5744 + }, + { + "epoch": 0.17721294471927787, + "grad_norm": 0.17704539000988007, + "learning_rate": 4.909749530424686e-05, + "loss": 3.2576, + "step": 5745 + }, + { + "epoch": 0.1772437911848513, + "grad_norm": 0.1615956574678421, + "learning_rate": 4.9096778396685486e-05, + "loss": 3.2869, + "step": 5746 + }, + { + "epoch": 0.17727463765042473, + "grad_norm": 0.17652453482151031, + "learning_rate": 4.909606120973644e-05, + "loss": 3.2336, + "step": 5747 + }, + { + "epoch": 0.17730548411599814, + "grad_norm": 0.1786748468875885, + "learning_rate": 4.9095343743408056e-05, + "loss": 3.2319, + "step": 5748 + }, + { + "epoch": 0.17733633058157156, + "grad_norm": 0.16755332052707672, + "learning_rate": 4.909462599770864e-05, + "loss": 3.2494, + "step": 5749 + }, + { + "epoch": 0.17736717704714497, + "grad_norm": 0.1615372598171234, + "learning_rate": 4.909390797264651e-05, + "loss": 3.2557, + "step": 5750 + }, + { + "epoch": 0.17739802351271838, + "grad_norm": 0.16053596138954163, + "learning_rate": 4.909318966823e-05, + "loss": 3.243, + "step": 5751 + }, + { + "epoch": 0.1774288699782918, + "grad_norm": 0.1589948683977127, + "learning_rate": 4.9092471084467425e-05, + "loss": 3.2574, + "step": 5752 + }, + { + "epoch": 0.1774597164438652, + "grad_norm": 0.15608687698841095, + "learning_rate": 4.909175222136714e-05, + "loss": 3.2487, + "step": 5753 + }, + { + "epoch": 0.17749056290943863, + "grad_norm": 0.16133840382099152, + "learning_rate": 4.909103307893746e-05, + "loss": 3.2345, + "step": 5754 + }, + { + "epoch": 0.17752140937501204, + "grad_norm": 0.16168776154518127, + "learning_rate": 4.909031365718672e-05, + "loss": 3.2388, + "step": 5755 + }, + { + "epoch": 0.17755225584058545, + "grad_norm": 0.15604053437709808, + "learning_rate": 4.908959395612328e-05, + "loss": 3.2696, + "step": 5756 + }, + { + "epoch": 0.17758310230615887, + "grad_norm": 0.16208188235759735, + "learning_rate": 4.908887397575548e-05, + "loss": 3.2831, + "step": 5757 + }, + { + "epoch": 0.1776139487717323, + "grad_norm": 0.15296867489814758, + "learning_rate": 4.908815371609164e-05, + "loss": 3.2604, + "step": 5758 + }, + { + "epoch": 0.17764479523730572, + "grad_norm": 0.15773379802703857, + "learning_rate": 4.908743317714015e-05, + "loss": 3.2606, + "step": 5759 + }, + { + "epoch": 0.17767564170287914, + "grad_norm": 0.15486373007297516, + "learning_rate": 4.9086712358909346e-05, + "loss": 3.2473, + "step": 5760 + }, + { + "epoch": 0.17770648816845255, + "grad_norm": 0.15441088378429413, + "learning_rate": 4.908599126140758e-05, + "loss": 3.2409, + "step": 5761 + }, + { + "epoch": 0.17773733463402597, + "grad_norm": 0.16147638857364655, + "learning_rate": 4.908526988464323e-05, + "loss": 3.2401, + "step": 5762 + }, + { + "epoch": 0.17776818109959938, + "grad_norm": 0.15785421431064606, + "learning_rate": 4.908454822862464e-05, + "loss": 3.2497, + "step": 5763 + }, + { + "epoch": 0.1777990275651728, + "grad_norm": 0.17181307077407837, + "learning_rate": 4.908382629336019e-05, + "loss": 3.2372, + "step": 5764 + }, + { + "epoch": 0.1778298740307462, + "grad_norm": 0.17942284047603607, + "learning_rate": 4.908310407885824e-05, + "loss": 3.227, + "step": 5765 + }, + { + "epoch": 0.17786072049631962, + "grad_norm": 0.1671600639820099, + "learning_rate": 4.908238158512718e-05, + "loss": 3.2537, + "step": 5766 + }, + { + "epoch": 0.17789156696189304, + "grad_norm": 0.1678330898284912, + "learning_rate": 4.908165881217537e-05, + "loss": 3.2819, + "step": 5767 + }, + { + "epoch": 0.17792241342746645, + "grad_norm": 0.15895573794841766, + "learning_rate": 4.9080935760011203e-05, + "loss": 3.2855, + "step": 5768 + }, + { + "epoch": 0.17795325989303987, + "grad_norm": 0.19115884602069855, + "learning_rate": 4.908021242864306e-05, + "loss": 3.284, + "step": 5769 + }, + { + "epoch": 0.1779841063586133, + "grad_norm": 0.15844665467739105, + "learning_rate": 4.907948881807932e-05, + "loss": 3.2334, + "step": 5770 + }, + { + "epoch": 0.17801495282418672, + "grad_norm": 0.16797713935375214, + "learning_rate": 4.9078764928328374e-05, + "loss": 3.2662, + "step": 5771 + }, + { + "epoch": 0.17804579928976014, + "grad_norm": 0.15544752776622772, + "learning_rate": 4.907804075939862e-05, + "loss": 3.2572, + "step": 5772 + }, + { + "epoch": 0.17807664575533355, + "grad_norm": 0.1649674028158188, + "learning_rate": 4.9077316311298454e-05, + "loss": 3.2158, + "step": 5773 + }, + { + "epoch": 0.17810749222090697, + "grad_norm": 0.15267616510391235, + "learning_rate": 4.907659158403628e-05, + "loss": 3.2556, + "step": 5774 + }, + { + "epoch": 0.17813833868648038, + "grad_norm": 0.1821231245994568, + "learning_rate": 4.907586657762048e-05, + "loss": 3.2335, + "step": 5775 + }, + { + "epoch": 0.1781691851520538, + "grad_norm": 0.15625642240047455, + "learning_rate": 4.907514129205949e-05, + "loss": 3.2592, + "step": 5776 + }, + { + "epoch": 0.1782000316176272, + "grad_norm": 0.1749194860458374, + "learning_rate": 4.907441572736169e-05, + "loss": 3.2351, + "step": 5777 + }, + { + "epoch": 0.17823087808320062, + "grad_norm": 0.1674068123102188, + "learning_rate": 4.9073689883535514e-05, + "loss": 3.244, + "step": 5778 + }, + { + "epoch": 0.17826172454877404, + "grad_norm": 0.15594986081123352, + "learning_rate": 4.9072963760589374e-05, + "loss": 3.2082, + "step": 5779 + }, + { + "epoch": 0.17829257101434745, + "grad_norm": 0.16211222112178802, + "learning_rate": 4.907223735853168e-05, + "loss": 3.2346, + "step": 5780 + }, + { + "epoch": 0.1783234174799209, + "grad_norm": 0.1629471331834793, + "learning_rate": 4.9071510677370866e-05, + "loss": 3.2524, + "step": 5781 + }, + { + "epoch": 0.1783542639454943, + "grad_norm": 0.15907984972000122, + "learning_rate": 4.9070783717115343e-05, + "loss": 3.2432, + "step": 5782 + }, + { + "epoch": 0.17838511041106772, + "grad_norm": 0.16543777287006378, + "learning_rate": 4.907005647777355e-05, + "loss": 3.2352, + "step": 5783 + }, + { + "epoch": 0.17841595687664114, + "grad_norm": 0.1673869490623474, + "learning_rate": 4.906932895935392e-05, + "loss": 3.2392, + "step": 5784 + }, + { + "epoch": 0.17844680334221455, + "grad_norm": 0.16502781212329865, + "learning_rate": 4.906860116186488e-05, + "loss": 3.2591, + "step": 5785 + }, + { + "epoch": 0.17847764980778796, + "grad_norm": 0.15622098743915558, + "learning_rate": 4.906787308531488e-05, + "loss": 3.2217, + "step": 5786 + }, + { + "epoch": 0.17850849627336138, + "grad_norm": 0.17292076349258423, + "learning_rate": 4.906714472971235e-05, + "loss": 3.2473, + "step": 5787 + }, + { + "epoch": 0.1785393427389348, + "grad_norm": 0.15657617151737213, + "learning_rate": 4.906641609506574e-05, + "loss": 3.2359, + "step": 5788 + }, + { + "epoch": 0.1785701892045082, + "grad_norm": 0.1546202152967453, + "learning_rate": 4.90656871813835e-05, + "loss": 3.2434, + "step": 5789 + }, + { + "epoch": 0.17860103567008162, + "grad_norm": 0.1511954963207245, + "learning_rate": 4.9064957988674076e-05, + "loss": 3.2427, + "step": 5790 + }, + { + "epoch": 0.17863188213565503, + "grad_norm": 0.15414033830165863, + "learning_rate": 4.906422851694592e-05, + "loss": 3.2305, + "step": 5791 + }, + { + "epoch": 0.17866272860122845, + "grad_norm": 0.15113013982772827, + "learning_rate": 4.906349876620751e-05, + "loss": 3.2453, + "step": 5792 + }, + { + "epoch": 0.1786935750668019, + "grad_norm": 0.15890143811702728, + "learning_rate": 4.9062768736467277e-05, + "loss": 3.2591, + "step": 5793 + }, + { + "epoch": 0.1787244215323753, + "grad_norm": 0.15499287843704224, + "learning_rate": 4.906203842773371e-05, + "loss": 3.2306, + "step": 5794 + }, + { + "epoch": 0.17875526799794872, + "grad_norm": 0.5991644859313965, + "learning_rate": 4.9061307840015255e-05, + "loss": 3.2159, + "step": 5795 + }, + { + "epoch": 0.17878611446352213, + "grad_norm": 0.16097687184810638, + "learning_rate": 4.906057697332041e-05, + "loss": 3.248, + "step": 5796 + }, + { + "epoch": 0.17881696092909555, + "grad_norm": 0.15929265320301056, + "learning_rate": 4.905984582765762e-05, + "loss": 3.2376, + "step": 5797 + }, + { + "epoch": 0.17884780739466896, + "grad_norm": 0.15910230576992035, + "learning_rate": 4.905911440303539e-05, + "loss": 3.2208, + "step": 5798 + }, + { + "epoch": 0.17887865386024238, + "grad_norm": 0.16551871597766876, + "learning_rate": 4.905838269946218e-05, + "loss": 3.2154, + "step": 5799 + }, + { + "epoch": 0.1789095003258158, + "grad_norm": 0.1580893099308014, + "learning_rate": 4.905765071694648e-05, + "loss": 3.219, + "step": 5800 + }, + { + "epoch": 0.1789403467913892, + "grad_norm": 0.1610344499349594, + "learning_rate": 4.905691845549678e-05, + "loss": 3.2551, + "step": 5801 + }, + { + "epoch": 0.17897119325696262, + "grad_norm": 0.15711748600006104, + "learning_rate": 4.905618591512156e-05, + "loss": 3.2301, + "step": 5802 + }, + { + "epoch": 0.17900203972253603, + "grad_norm": 0.16189530491828918, + "learning_rate": 4.9055453095829326e-05, + "loss": 3.2345, + "step": 5803 + }, + { + "epoch": 0.17903288618810945, + "grad_norm": 0.15755486488342285, + "learning_rate": 4.905471999762857e-05, + "loss": 3.2265, + "step": 5804 + }, + { + "epoch": 0.1790637326536829, + "grad_norm": 0.15550534427165985, + "learning_rate": 4.905398662052778e-05, + "loss": 3.2165, + "step": 5805 + }, + { + "epoch": 0.1790945791192563, + "grad_norm": 0.16115613281726837, + "learning_rate": 4.905325296453548e-05, + "loss": 3.2573, + "step": 5806 + }, + { + "epoch": 0.17912542558482972, + "grad_norm": 0.16664433479309082, + "learning_rate": 4.905251902966017e-05, + "loss": 3.2543, + "step": 5807 + }, + { + "epoch": 0.17915627205040313, + "grad_norm": 0.15964671969413757, + "learning_rate": 4.905178481591035e-05, + "loss": 3.231, + "step": 5808 + }, + { + "epoch": 0.17918711851597655, + "grad_norm": 0.16683773696422577, + "learning_rate": 4.905105032329453e-05, + "loss": 3.2734, + "step": 5809 + }, + { + "epoch": 0.17921796498154996, + "grad_norm": 0.16444897651672363, + "learning_rate": 4.9050315551821245e-05, + "loss": 3.2634, + "step": 5810 + }, + { + "epoch": 0.17924881144712337, + "grad_norm": 0.1607344001531601, + "learning_rate": 4.9049580501499007e-05, + "loss": 3.2173, + "step": 5811 + }, + { + "epoch": 0.1792796579126968, + "grad_norm": 0.15594501793384552, + "learning_rate": 4.904884517233633e-05, + "loss": 3.2377, + "step": 5812 + }, + { + "epoch": 0.1793105043782702, + "grad_norm": 0.15971195697784424, + "learning_rate": 4.904810956434174e-05, + "loss": 3.2134, + "step": 5813 + }, + { + "epoch": 0.17934135084384362, + "grad_norm": 0.15213340520858765, + "learning_rate": 4.904737367752378e-05, + "loss": 3.2113, + "step": 5814 + }, + { + "epoch": 0.17937219730941703, + "grad_norm": 0.1652430295944214, + "learning_rate": 4.904663751189097e-05, + "loss": 3.2839, + "step": 5815 + }, + { + "epoch": 0.17940304377499044, + "grad_norm": 0.15952174365520477, + "learning_rate": 4.9045901067451836e-05, + "loss": 3.2567, + "step": 5816 + }, + { + "epoch": 0.1794338902405639, + "grad_norm": 0.1609531193971634, + "learning_rate": 4.904516434421494e-05, + "loss": 3.2514, + "step": 5817 + }, + { + "epoch": 0.1794647367061373, + "grad_norm": 0.15743622183799744, + "learning_rate": 4.904442734218882e-05, + "loss": 3.2328, + "step": 5818 + }, + { + "epoch": 0.17949558317171072, + "grad_norm": 0.16162540018558502, + "learning_rate": 4.904369006138201e-05, + "loss": 3.246, + "step": 5819 + }, + { + "epoch": 0.17952642963728413, + "grad_norm": 0.15932491421699524, + "learning_rate": 4.904295250180305e-05, + "loss": 3.2489, + "step": 5820 + }, + { + "epoch": 0.17955727610285754, + "grad_norm": 0.15788675844669342, + "learning_rate": 4.904221466346052e-05, + "loss": 3.2643, + "step": 5821 + }, + { + "epoch": 0.17958812256843096, + "grad_norm": 0.18445134162902832, + "learning_rate": 4.904147654636295e-05, + "loss": 3.2039, + "step": 5822 + }, + { + "epoch": 0.17961896903400437, + "grad_norm": 0.15957249701023102, + "learning_rate": 4.9040738150518903e-05, + "loss": 3.259, + "step": 5823 + }, + { + "epoch": 0.1796498154995778, + "grad_norm": 0.16504018008708954, + "learning_rate": 4.903999947593695e-05, + "loss": 3.2245, + "step": 5824 + }, + { + "epoch": 0.1796806619651512, + "grad_norm": 0.17490670084953308, + "learning_rate": 4.903926052262565e-05, + "loss": 3.2268, + "step": 5825 + }, + { + "epoch": 0.17971150843072461, + "grad_norm": 0.15866898000240326, + "learning_rate": 4.903852129059358e-05, + "loss": 3.248, + "step": 5826 + }, + { + "epoch": 0.17974235489629803, + "grad_norm": 0.2040978968143463, + "learning_rate": 4.903778177984928e-05, + "loss": 3.2506, + "step": 5827 + }, + { + "epoch": 0.17977320136187144, + "grad_norm": 0.21190978586673737, + "learning_rate": 4.9037041990401355e-05, + "loss": 3.2493, + "step": 5828 + }, + { + "epoch": 0.17980404782744489, + "grad_norm": 0.1769857257604599, + "learning_rate": 4.9036301922258376e-05, + "loss": 3.2334, + "step": 5829 + }, + { + "epoch": 0.1798348942930183, + "grad_norm": 0.15789951384067535, + "learning_rate": 4.903556157542892e-05, + "loss": 3.2313, + "step": 5830 + }, + { + "epoch": 0.1798657407585917, + "grad_norm": 0.17798656225204468, + "learning_rate": 4.9034820949921567e-05, + "loss": 3.2333, + "step": 5831 + }, + { + "epoch": 0.17989658722416513, + "grad_norm": 0.15706691145896912, + "learning_rate": 4.903408004574491e-05, + "loss": 3.2453, + "step": 5832 + }, + { + "epoch": 0.17992743368973854, + "grad_norm": 0.16660039126873016, + "learning_rate": 4.903333886290754e-05, + "loss": 3.2407, + "step": 5833 + }, + { + "epoch": 0.17995828015531196, + "grad_norm": 0.16585859656333923, + "learning_rate": 4.903259740141805e-05, + "loss": 3.2762, + "step": 5834 + }, + { + "epoch": 0.17998912662088537, + "grad_norm": 0.7879397869110107, + "learning_rate": 4.903185566128502e-05, + "loss": 3.2282, + "step": 5835 + }, + { + "epoch": 0.18001997308645878, + "grad_norm": 0.1704307198524475, + "learning_rate": 4.903111364251708e-05, + "loss": 3.2226, + "step": 5836 + }, + { + "epoch": 0.1800508195520322, + "grad_norm": 0.16064564883708954, + "learning_rate": 4.9030371345122813e-05, + "loss": 3.2102, + "step": 5837 + }, + { + "epoch": 0.1800816660176056, + "grad_norm": 0.17783097922801971, + "learning_rate": 4.902962876911083e-05, + "loss": 3.2554, + "step": 5838 + }, + { + "epoch": 0.18011251248317903, + "grad_norm": 0.16096583008766174, + "learning_rate": 4.902888591448974e-05, + "loss": 3.2134, + "step": 5839 + }, + { + "epoch": 0.18014335894875244, + "grad_norm": 0.16803182661533356, + "learning_rate": 4.9028142781268165e-05, + "loss": 3.2491, + "step": 5840 + }, + { + "epoch": 0.18017420541432588, + "grad_norm": 0.3177025616168976, + "learning_rate": 4.9027399369454706e-05, + "loss": 3.2244, + "step": 5841 + }, + { + "epoch": 0.1802050518798993, + "grad_norm": 0.1671120822429657, + "learning_rate": 4.9026655679058e-05, + "loss": 3.2345, + "step": 5842 + }, + { + "epoch": 0.1802358983454727, + "grad_norm": 0.16299962997436523, + "learning_rate": 4.902591171008665e-05, + "loss": 3.2579, + "step": 5843 + }, + { + "epoch": 0.18026674481104613, + "grad_norm": 0.1700609028339386, + "learning_rate": 4.90251674625493e-05, + "loss": 3.2609, + "step": 5844 + }, + { + "epoch": 0.18029759127661954, + "grad_norm": 0.17615678906440735, + "learning_rate": 4.902442293645456e-05, + "loss": 3.2237, + "step": 5845 + }, + { + "epoch": 0.18032843774219295, + "grad_norm": 0.16443490982055664, + "learning_rate": 4.902367813181108e-05, + "loss": 3.2507, + "step": 5846 + }, + { + "epoch": 0.18035928420776637, + "grad_norm": 0.15967823565006256, + "learning_rate": 4.9022933048627496e-05, + "loss": 3.2483, + "step": 5847 + }, + { + "epoch": 0.18039013067333978, + "grad_norm": 0.16224339604377747, + "learning_rate": 4.9022187686912425e-05, + "loss": 3.2543, + "step": 5848 + }, + { + "epoch": 0.1804209771389132, + "grad_norm": 0.1671667844057083, + "learning_rate": 4.902144204667453e-05, + "loss": 3.2399, + "step": 5849 + }, + { + "epoch": 0.1804518236044866, + "grad_norm": 0.15739057958126068, + "learning_rate": 4.902069612792245e-05, + "loss": 3.2498, + "step": 5850 + }, + { + "epoch": 0.18048267007006003, + "grad_norm": 0.17022375762462616, + "learning_rate": 4.9019949930664836e-05, + "loss": 3.233, + "step": 5851 + }, + { + "epoch": 0.18051351653563344, + "grad_norm": 0.152833953499794, + "learning_rate": 4.901920345491033e-05, + "loss": 3.2355, + "step": 5852 + }, + { + "epoch": 0.18054436300120688, + "grad_norm": 0.1652706116437912, + "learning_rate": 4.90184567006676e-05, + "loss": 3.2347, + "step": 5853 + }, + { + "epoch": 0.1805752094667803, + "grad_norm": 0.15783245861530304, + "learning_rate": 4.901770966794529e-05, + "loss": 3.226, + "step": 5854 + }, + { + "epoch": 0.1806060559323537, + "grad_norm": 0.1649070382118225, + "learning_rate": 4.9016962356752074e-05, + "loss": 3.2674, + "step": 5855 + }, + { + "epoch": 0.18063690239792712, + "grad_norm": 0.15876127779483795, + "learning_rate": 4.901621476709661e-05, + "loss": 3.2631, + "step": 5856 + }, + { + "epoch": 0.18066774886350054, + "grad_norm": 0.17141282558441162, + "learning_rate": 4.9015466898987575e-05, + "loss": 3.239, + "step": 5857 + }, + { + "epoch": 0.18069859532907395, + "grad_norm": 0.16794131696224213, + "learning_rate": 4.901471875243363e-05, + "loss": 3.228, + "step": 5858 + }, + { + "epoch": 0.18072944179464737, + "grad_norm": 0.16415679454803467, + "learning_rate": 4.901397032744345e-05, + "loss": 3.251, + "step": 5859 + }, + { + "epoch": 0.18076028826022078, + "grad_norm": 0.1570754498243332, + "learning_rate": 4.901322162402571e-05, + "loss": 3.2272, + "step": 5860 + }, + { + "epoch": 0.1807911347257942, + "grad_norm": 0.1787242740392685, + "learning_rate": 4.90124726421891e-05, + "loss": 3.2484, + "step": 5861 + }, + { + "epoch": 0.1808219811913676, + "grad_norm": 0.1627543568611145, + "learning_rate": 4.90117233819423e-05, + "loss": 3.1702, + "step": 5862 + }, + { + "epoch": 0.18085282765694102, + "grad_norm": 0.15425211191177368, + "learning_rate": 4.9010973843294e-05, + "loss": 3.2732, + "step": 5863 + }, + { + "epoch": 0.18088367412251444, + "grad_norm": 0.16327430307865143, + "learning_rate": 4.9010224026252884e-05, + "loss": 3.2416, + "step": 5864 + }, + { + "epoch": 0.18091452058808788, + "grad_norm": 0.15373843908309937, + "learning_rate": 4.900947393082765e-05, + "loss": 3.2227, + "step": 5865 + }, + { + "epoch": 0.1809453670536613, + "grad_norm": 0.1637030839920044, + "learning_rate": 4.900872355702699e-05, + "loss": 3.271, + "step": 5866 + }, + { + "epoch": 0.1809762135192347, + "grad_norm": 0.16262613236904144, + "learning_rate": 4.9007972904859615e-05, + "loss": 3.2319, + "step": 5867 + }, + { + "epoch": 0.18100705998480812, + "grad_norm": 0.16027508676052094, + "learning_rate": 4.900722197433422e-05, + "loss": 3.2715, + "step": 5868 + }, + { + "epoch": 0.18103790645038154, + "grad_norm": 0.1645079255104065, + "learning_rate": 4.9006470765459504e-05, + "loss": 3.2207, + "step": 5869 + }, + { + "epoch": 0.18106875291595495, + "grad_norm": 0.1648932546377182, + "learning_rate": 4.9005719278244193e-05, + "loss": 3.2265, + "step": 5870 + }, + { + "epoch": 0.18109959938152836, + "grad_norm": 0.1564738005399704, + "learning_rate": 4.9004967512696994e-05, + "loss": 3.2278, + "step": 5871 + }, + { + "epoch": 0.18113044584710178, + "grad_norm": 0.15038412809371948, + "learning_rate": 4.900421546882662e-05, + "loss": 3.208, + "step": 5872 + }, + { + "epoch": 0.1811612923126752, + "grad_norm": 0.15090681612491608, + "learning_rate": 4.900346314664179e-05, + "loss": 3.2409, + "step": 5873 + }, + { + "epoch": 0.1811921387782486, + "grad_norm": 0.15602907538414001, + "learning_rate": 4.900271054615123e-05, + "loss": 3.2239, + "step": 5874 + }, + { + "epoch": 0.18122298524382202, + "grad_norm": 0.15503664314746857, + "learning_rate": 4.9001957667363664e-05, + "loss": 3.2746, + "step": 5875 + }, + { + "epoch": 0.18125383170939544, + "grad_norm": 0.15883347392082214, + "learning_rate": 4.900120451028782e-05, + "loss": 3.2293, + "step": 5876 + }, + { + "epoch": 0.18128467817496888, + "grad_norm": 0.15964584052562714, + "learning_rate": 4.900045107493244e-05, + "loss": 3.2125, + "step": 5877 + }, + { + "epoch": 0.1813155246405423, + "grad_norm": 0.15178696811199188, + "learning_rate": 4.899969736130625e-05, + "loss": 3.2247, + "step": 5878 + }, + { + "epoch": 0.1813463711061157, + "grad_norm": 0.1651269793510437, + "learning_rate": 4.899894336941799e-05, + "loss": 3.2549, + "step": 5879 + }, + { + "epoch": 0.18137721757168912, + "grad_norm": 0.15920819342136383, + "learning_rate": 4.8998189099276406e-05, + "loss": 3.2505, + "step": 5880 + }, + { + "epoch": 0.18140806403726253, + "grad_norm": 0.15720762312412262, + "learning_rate": 4.8997434550890245e-05, + "loss": 3.2431, + "step": 5881 + }, + { + "epoch": 0.18143891050283595, + "grad_norm": 0.166233092546463, + "learning_rate": 4.899667972426824e-05, + "loss": 3.231, + "step": 5882 + }, + { + "epoch": 0.18146975696840936, + "grad_norm": 0.1561093032360077, + "learning_rate": 4.899592461941916e-05, + "loss": 3.2566, + "step": 5883 + }, + { + "epoch": 0.18150060343398278, + "grad_norm": 0.16476796567440033, + "learning_rate": 4.8995169236351754e-05, + "loss": 3.1975, + "step": 5884 + }, + { + "epoch": 0.1815314498995562, + "grad_norm": 0.1561373621225357, + "learning_rate": 4.899441357507478e-05, + "loss": 3.2396, + "step": 5885 + }, + { + "epoch": 0.1815622963651296, + "grad_norm": 0.16169746220111847, + "learning_rate": 4.8993657635597e-05, + "loss": 3.2313, + "step": 5886 + }, + { + "epoch": 0.18159314283070302, + "grad_norm": 0.15113389492034912, + "learning_rate": 4.8992901417927173e-05, + "loss": 3.2404, + "step": 5887 + }, + { + "epoch": 0.18162398929627643, + "grad_norm": 0.16293533146381378, + "learning_rate": 4.899214492207408e-05, + "loss": 3.2965, + "step": 5888 + }, + { + "epoch": 0.18165483576184988, + "grad_norm": 0.1568426936864853, + "learning_rate": 4.899138814804648e-05, + "loss": 3.2354, + "step": 5889 + }, + { + "epoch": 0.1816856822274233, + "grad_norm": 0.15887461602687836, + "learning_rate": 4.8990631095853145e-05, + "loss": 3.2535, + "step": 5890 + }, + { + "epoch": 0.1817165286929967, + "grad_norm": 0.15947306156158447, + "learning_rate": 4.898987376550287e-05, + "loss": 3.2263, + "step": 5891 + }, + { + "epoch": 0.18174737515857012, + "grad_norm": 0.1553950309753418, + "learning_rate": 4.898911615700442e-05, + "loss": 3.2351, + "step": 5892 + }, + { + "epoch": 0.18177822162414353, + "grad_norm": 0.15972460806369781, + "learning_rate": 4.8988358270366586e-05, + "loss": 3.2361, + "step": 5893 + }, + { + "epoch": 0.18180906808971695, + "grad_norm": 0.16340075433254242, + "learning_rate": 4.898760010559815e-05, + "loss": 3.2599, + "step": 5894 + }, + { + "epoch": 0.18183991455529036, + "grad_norm": 0.16354119777679443, + "learning_rate": 4.898684166270791e-05, + "loss": 3.2192, + "step": 5895 + }, + { + "epoch": 0.18187076102086377, + "grad_norm": 0.15514978766441345, + "learning_rate": 4.898608294170465e-05, + "loss": 3.2192, + "step": 5896 + }, + { + "epoch": 0.1819016074864372, + "grad_norm": 0.15985320508480072, + "learning_rate": 4.898532394259717e-05, + "loss": 3.2245, + "step": 5897 + }, + { + "epoch": 0.1819324539520106, + "grad_norm": 0.19051218032836914, + "learning_rate": 4.898456466539428e-05, + "loss": 3.2385, + "step": 5898 + }, + { + "epoch": 0.18196330041758402, + "grad_norm": 0.792879581451416, + "learning_rate": 4.898380511010477e-05, + "loss": 3.2263, + "step": 5899 + }, + { + "epoch": 0.18199414688315743, + "grad_norm": 0.16416959464550018, + "learning_rate": 4.898304527673746e-05, + "loss": 3.2726, + "step": 5900 + }, + { + "epoch": 0.18202499334873087, + "grad_norm": 0.17577005922794342, + "learning_rate": 4.898228516530115e-05, + "loss": 3.2294, + "step": 5901 + }, + { + "epoch": 0.1820558398143043, + "grad_norm": 0.17447449266910553, + "learning_rate": 4.898152477580466e-05, + "loss": 3.2389, + "step": 5902 + }, + { + "epoch": 0.1820866862798777, + "grad_norm": 0.16527603566646576, + "learning_rate": 4.8980764108256786e-05, + "loss": 3.237, + "step": 5903 + }, + { + "epoch": 0.18211753274545112, + "grad_norm": 0.1705876886844635, + "learning_rate": 4.898000316266638e-05, + "loss": 3.216, + "step": 5904 + }, + { + "epoch": 0.18214837921102453, + "grad_norm": 0.16958869993686676, + "learning_rate": 4.897924193904223e-05, + "loss": 3.22, + "step": 5905 + }, + { + "epoch": 0.18217922567659794, + "grad_norm": 0.16765563189983368, + "learning_rate": 4.8978480437393195e-05, + "loss": 3.2555, + "step": 5906 + }, + { + "epoch": 0.18221007214217136, + "grad_norm": 0.16220003366470337, + "learning_rate": 4.8977718657728076e-05, + "loss": 3.2177, + "step": 5907 + }, + { + "epoch": 0.18224091860774477, + "grad_norm": 0.1666293889284134, + "learning_rate": 4.8976956600055726e-05, + "loss": 3.2277, + "step": 5908 + }, + { + "epoch": 0.1822717650733182, + "grad_norm": 0.1649230271577835, + "learning_rate": 4.897619426438497e-05, + "loss": 3.255, + "step": 5909 + }, + { + "epoch": 0.1823026115388916, + "grad_norm": 0.1800260841846466, + "learning_rate": 4.897543165072465e-05, + "loss": 3.2697, + "step": 5910 + }, + { + "epoch": 0.18233345800446502, + "grad_norm": 0.18074850738048553, + "learning_rate": 4.897466875908361e-05, + "loss": 3.2923, + "step": 5911 + }, + { + "epoch": 0.18236430447003843, + "grad_norm": 0.16770701110363007, + "learning_rate": 4.8973905589470695e-05, + "loss": 3.2392, + "step": 5912 + }, + { + "epoch": 0.18239515093561187, + "grad_norm": 0.1722884178161621, + "learning_rate": 4.8973142141894744e-05, + "loss": 3.2507, + "step": 5913 + }, + { + "epoch": 0.18242599740118529, + "grad_norm": 0.16344356536865234, + "learning_rate": 4.8972378416364614e-05, + "loss": 3.2172, + "step": 5914 + }, + { + "epoch": 0.1824568438667587, + "grad_norm": 0.1624731570482254, + "learning_rate": 4.8971614412889174e-05, + "loss": 3.2438, + "step": 5915 + }, + { + "epoch": 0.18248769033233211, + "grad_norm": 0.17133978009223938, + "learning_rate": 4.8970850131477266e-05, + "loss": 3.2362, + "step": 5916 + }, + { + "epoch": 0.18251853679790553, + "grad_norm": 0.1636885702610016, + "learning_rate": 4.897008557213775e-05, + "loss": 3.2206, + "step": 5917 + }, + { + "epoch": 0.18254938326347894, + "grad_norm": 0.16041919589042664, + "learning_rate": 4.89693207348795e-05, + "loss": 3.2456, + "step": 5918 + }, + { + "epoch": 0.18258022972905236, + "grad_norm": 0.16842082142829895, + "learning_rate": 4.8968555619711375e-05, + "loss": 3.2569, + "step": 5919 + }, + { + "epoch": 0.18261107619462577, + "grad_norm": 0.15709520876407623, + "learning_rate": 4.8967790226642254e-05, + "loss": 3.2387, + "step": 5920 + }, + { + "epoch": 0.18264192266019919, + "grad_norm": 0.174279123544693, + "learning_rate": 4.896702455568101e-05, + "loss": 3.2184, + "step": 5921 + }, + { + "epoch": 0.1826727691257726, + "grad_norm": 0.15676353871822357, + "learning_rate": 4.8966258606836525e-05, + "loss": 3.2632, + "step": 5922 + }, + { + "epoch": 0.182703615591346, + "grad_norm": 0.17005515098571777, + "learning_rate": 4.896549238011766e-05, + "loss": 3.2723, + "step": 5923 + }, + { + "epoch": 0.18273446205691943, + "grad_norm": 0.19637127220630646, + "learning_rate": 4.8964725875533326e-05, + "loss": 3.2627, + "step": 5924 + }, + { + "epoch": 0.18276530852249287, + "grad_norm": 0.15759532153606415, + "learning_rate": 4.896395909309238e-05, + "loss": 3.2427, + "step": 5925 + }, + { + "epoch": 0.18279615498806628, + "grad_norm": 0.15581586956977844, + "learning_rate": 4.896319203280375e-05, + "loss": 3.2075, + "step": 5926 + }, + { + "epoch": 0.1828270014536397, + "grad_norm": 0.180036261677742, + "learning_rate": 4.896242469467629e-05, + "loss": 3.2516, + "step": 5927 + }, + { + "epoch": 0.1828578479192131, + "grad_norm": 0.15938803553581238, + "learning_rate": 4.8961657078718924e-05, + "loss": 3.2572, + "step": 5928 + }, + { + "epoch": 0.18288869438478653, + "grad_norm": 0.1713617593050003, + "learning_rate": 4.896088918494054e-05, + "loss": 3.214, + "step": 5929 + }, + { + "epoch": 0.18291954085035994, + "grad_norm": 0.15745866298675537, + "learning_rate": 4.896012101335005e-05, + "loss": 3.2741, + "step": 5930 + }, + { + "epoch": 0.18295038731593335, + "grad_norm": 0.16407376527786255, + "learning_rate": 4.895935256395635e-05, + "loss": 3.2177, + "step": 5931 + }, + { + "epoch": 0.18298123378150677, + "grad_norm": 0.1563345193862915, + "learning_rate": 4.8958583836768354e-05, + "loss": 3.2179, + "step": 5932 + }, + { + "epoch": 0.18301208024708018, + "grad_norm": 0.18722812831401825, + "learning_rate": 4.895781483179498e-05, + "loss": 3.2313, + "step": 5933 + }, + { + "epoch": 0.1830429267126536, + "grad_norm": 0.1530526578426361, + "learning_rate": 4.8957045549045146e-05, + "loss": 3.2523, + "step": 5934 + }, + { + "epoch": 0.183073773178227, + "grad_norm": 0.1624165177345276, + "learning_rate": 4.895627598852776e-05, + "loss": 3.2528, + "step": 5935 + }, + { + "epoch": 0.18310461964380043, + "grad_norm": 0.15636280179023743, + "learning_rate": 4.895550615025174e-05, + "loss": 3.2576, + "step": 5936 + }, + { + "epoch": 0.18313546610937387, + "grad_norm": 0.15900950133800507, + "learning_rate": 4.895473603422603e-05, + "loss": 3.259, + "step": 5937 + }, + { + "epoch": 0.18316631257494728, + "grad_norm": 0.1566111445426941, + "learning_rate": 4.895396564045955e-05, + "loss": 3.2422, + "step": 5938 + }, + { + "epoch": 0.1831971590405207, + "grad_norm": 0.1598825454711914, + "learning_rate": 4.8953194968961234e-05, + "loss": 3.2539, + "step": 5939 + }, + { + "epoch": 0.1832280055060941, + "grad_norm": 0.15618231892585754, + "learning_rate": 4.8952424019740016e-05, + "loss": 3.2572, + "step": 5940 + }, + { + "epoch": 0.18325885197166752, + "grad_norm": 0.14942671358585358, + "learning_rate": 4.895165279280484e-05, + "loss": 3.242, + "step": 5941 + }, + { + "epoch": 0.18328969843724094, + "grad_norm": 0.1546318233013153, + "learning_rate": 4.895088128816464e-05, + "loss": 3.2637, + "step": 5942 + }, + { + "epoch": 0.18332054490281435, + "grad_norm": 0.15114276111125946, + "learning_rate": 4.895010950582836e-05, + "loss": 3.2418, + "step": 5943 + }, + { + "epoch": 0.18335139136838777, + "grad_norm": 0.17386700212955475, + "learning_rate": 4.894933744580496e-05, + "loss": 3.2171, + "step": 5944 + }, + { + "epoch": 0.18338223783396118, + "grad_norm": 0.15410786867141724, + "learning_rate": 4.894856510810337e-05, + "loss": 3.231, + "step": 5945 + }, + { + "epoch": 0.1834130842995346, + "grad_norm": 0.16696098446846008, + "learning_rate": 4.8947792492732575e-05, + "loss": 3.274, + "step": 5946 + }, + { + "epoch": 0.183443930765108, + "grad_norm": 0.15828385949134827, + "learning_rate": 4.894701959970151e-05, + "loss": 3.2424, + "step": 5947 + }, + { + "epoch": 0.18347477723068142, + "grad_norm": 0.3704298436641693, + "learning_rate": 4.894624642901915e-05, + "loss": 3.2368, + "step": 5948 + }, + { + "epoch": 0.18350562369625487, + "grad_norm": 0.15852297842502594, + "learning_rate": 4.894547298069444e-05, + "loss": 3.2458, + "step": 5949 + }, + { + "epoch": 0.18353647016182828, + "grad_norm": 0.2020682394504547, + "learning_rate": 4.894469925473637e-05, + "loss": 3.2614, + "step": 5950 + }, + { + "epoch": 0.1835673166274017, + "grad_norm": 0.16819822788238525, + "learning_rate": 4.89439252511539e-05, + "loss": 3.2217, + "step": 5951 + }, + { + "epoch": 0.1835981630929751, + "grad_norm": 0.16165007650852203, + "learning_rate": 4.8943150969956e-05, + "loss": 3.1915, + "step": 5952 + }, + { + "epoch": 0.18362900955854852, + "grad_norm": 0.16050288081169128, + "learning_rate": 4.894237641115166e-05, + "loss": 3.2091, + "step": 5953 + }, + { + "epoch": 0.18365985602412194, + "grad_norm": 0.16002970933914185, + "learning_rate": 4.8941601574749844e-05, + "loss": 3.2504, + "step": 5954 + }, + { + "epoch": 0.18369070248969535, + "grad_norm": 0.16077707707881927, + "learning_rate": 4.894082646075955e-05, + "loss": 3.2506, + "step": 5955 + }, + { + "epoch": 0.18372154895526877, + "grad_norm": 0.15751241147518158, + "learning_rate": 4.8940051069189766e-05, + "loss": 3.2323, + "step": 5956 + }, + { + "epoch": 0.18375239542084218, + "grad_norm": 0.15731124579906464, + "learning_rate": 4.8939275400049466e-05, + "loss": 3.2123, + "step": 5957 + }, + { + "epoch": 0.1837832418864156, + "grad_norm": 0.15663231909275055, + "learning_rate": 4.893849945334766e-05, + "loss": 3.2529, + "step": 5958 + }, + { + "epoch": 0.183814088351989, + "grad_norm": 0.14898426830768585, + "learning_rate": 4.8937723229093336e-05, + "loss": 3.2114, + "step": 5959 + }, + { + "epoch": 0.18384493481756242, + "grad_norm": 0.15485572814941406, + "learning_rate": 4.89369467272955e-05, + "loss": 3.2423, + "step": 5960 + }, + { + "epoch": 0.18387578128313586, + "grad_norm": 0.1572944074869156, + "learning_rate": 4.8936169947963146e-05, + "loss": 3.215, + "step": 5961 + }, + { + "epoch": 0.18390662774870928, + "grad_norm": 0.15715086460113525, + "learning_rate": 4.893539289110529e-05, + "loss": 3.2341, + "step": 5962 + }, + { + "epoch": 0.1839374742142827, + "grad_norm": 0.15593765676021576, + "learning_rate": 4.893461555673093e-05, + "loss": 3.2578, + "step": 5963 + }, + { + "epoch": 0.1839683206798561, + "grad_norm": 0.16091574728488922, + "learning_rate": 4.8933837944849095e-05, + "loss": 3.2524, + "step": 5964 + }, + { + "epoch": 0.18399916714542952, + "grad_norm": 0.1541256606578827, + "learning_rate": 4.8933060055468784e-05, + "loss": 3.242, + "step": 5965 + }, + { + "epoch": 0.18403001361100293, + "grad_norm": 0.17450103163719177, + "learning_rate": 4.8932281888599016e-05, + "loss": 3.235, + "step": 5966 + }, + { + "epoch": 0.18406086007657635, + "grad_norm": 0.16258089244365692, + "learning_rate": 4.893150344424884e-05, + "loss": 3.2546, + "step": 5967 + }, + { + "epoch": 0.18409170654214976, + "grad_norm": 0.16626092791557312, + "learning_rate": 4.893072472242725e-05, + "loss": 3.2316, + "step": 5968 + }, + { + "epoch": 0.18412255300772318, + "grad_norm": 0.16182675957679749, + "learning_rate": 4.892994572314329e-05, + "loss": 3.2773, + "step": 5969 + }, + { + "epoch": 0.1841533994732966, + "grad_norm": 0.16955910623073578, + "learning_rate": 4.8929166446405993e-05, + "loss": 3.2616, + "step": 5970 + }, + { + "epoch": 0.18418424593887, + "grad_norm": 0.15882565081119537, + "learning_rate": 4.892838689222439e-05, + "loss": 3.282, + "step": 5971 + }, + { + "epoch": 0.18421509240444342, + "grad_norm": 0.15788230299949646, + "learning_rate": 4.892760706060752e-05, + "loss": 3.2365, + "step": 5972 + }, + { + "epoch": 0.18424593887001686, + "grad_norm": 0.1541747897863388, + "learning_rate": 4.892682695156442e-05, + "loss": 3.251, + "step": 5973 + }, + { + "epoch": 0.18427678533559028, + "grad_norm": 0.16114670038223267, + "learning_rate": 4.892604656510415e-05, + "loss": 3.2624, + "step": 5974 + }, + { + "epoch": 0.1843076318011637, + "grad_norm": 0.1546025276184082, + "learning_rate": 4.892526590123574e-05, + "loss": 3.2261, + "step": 5975 + }, + { + "epoch": 0.1843384782667371, + "grad_norm": 0.1596580296754837, + "learning_rate": 4.892448495996825e-05, + "loss": 3.2471, + "step": 5976 + }, + { + "epoch": 0.18436932473231052, + "grad_norm": 0.1571500599384308, + "learning_rate": 4.8923703741310735e-05, + "loss": 3.2447, + "step": 5977 + }, + { + "epoch": 0.18440017119788393, + "grad_norm": 0.15692737698554993, + "learning_rate": 4.892292224527225e-05, + "loss": 3.239, + "step": 5978 + }, + { + "epoch": 0.18443101766345735, + "grad_norm": 0.15622921288013458, + "learning_rate": 4.8922140471861855e-05, + "loss": 3.2217, + "step": 5979 + }, + { + "epoch": 0.18446186412903076, + "grad_norm": 0.16062813997268677, + "learning_rate": 4.892135842108863e-05, + "loss": 3.2644, + "step": 5980 + }, + { + "epoch": 0.18449271059460418, + "grad_norm": 0.15709832310676575, + "learning_rate": 4.892057609296161e-05, + "loss": 3.257, + "step": 5981 + }, + { + "epoch": 0.1845235570601776, + "grad_norm": 0.28035470843315125, + "learning_rate": 4.89197934874899e-05, + "loss": 3.2659, + "step": 5982 + }, + { + "epoch": 0.184554403525751, + "grad_norm": 0.16174913942813873, + "learning_rate": 4.8919010604682555e-05, + "loss": 3.2334, + "step": 5983 + }, + { + "epoch": 0.18458524999132442, + "grad_norm": 0.16494303941726685, + "learning_rate": 4.891822744454865e-05, + "loss": 3.2527, + "step": 5984 + }, + { + "epoch": 0.18461609645689786, + "grad_norm": 0.16157102584838867, + "learning_rate": 4.8917444007097276e-05, + "loss": 3.2332, + "step": 5985 + }, + { + "epoch": 0.18464694292247127, + "grad_norm": 0.16297194361686707, + "learning_rate": 4.8916660292337514e-05, + "loss": 3.2696, + "step": 5986 + }, + { + "epoch": 0.1846777893880447, + "grad_norm": 0.1605558544397354, + "learning_rate": 4.891587630027844e-05, + "loss": 3.2513, + "step": 5987 + }, + { + "epoch": 0.1847086358536181, + "grad_norm": 0.16437652707099915, + "learning_rate": 4.891509203092916e-05, + "loss": 3.2176, + "step": 5988 + }, + { + "epoch": 0.18473948231919152, + "grad_norm": 0.15422561764717102, + "learning_rate": 4.891430748429876e-05, + "loss": 3.2425, + "step": 5989 + }, + { + "epoch": 0.18477032878476493, + "grad_norm": 0.17042642831802368, + "learning_rate": 4.8913522660396324e-05, + "loss": 3.2261, + "step": 5990 + }, + { + "epoch": 0.18480117525033835, + "grad_norm": 0.1613352745771408, + "learning_rate": 4.891273755923097e-05, + "loss": 3.2233, + "step": 5991 + }, + { + "epoch": 0.18483202171591176, + "grad_norm": 0.15756858885288239, + "learning_rate": 4.891195218081179e-05, + "loss": 3.2355, + "step": 5992 + }, + { + "epoch": 0.18486286818148517, + "grad_norm": 0.1545894891023636, + "learning_rate": 4.8911166525147905e-05, + "loss": 3.2075, + "step": 5993 + }, + { + "epoch": 0.1848937146470586, + "grad_norm": 0.158486008644104, + "learning_rate": 4.891038059224841e-05, + "loss": 3.2605, + "step": 5994 + }, + { + "epoch": 0.184924561112632, + "grad_norm": 0.16084027290344238, + "learning_rate": 4.8909594382122414e-05, + "loss": 3.2329, + "step": 5995 + }, + { + "epoch": 0.18495540757820542, + "grad_norm": 0.225813090801239, + "learning_rate": 4.8908807894779043e-05, + "loss": 3.2321, + "step": 5996 + }, + { + "epoch": 0.18498625404377886, + "grad_norm": 0.5254908800125122, + "learning_rate": 4.890802113022741e-05, + "loss": 3.2518, + "step": 5997 + }, + { + "epoch": 0.18501710050935227, + "grad_norm": 0.15519019961357117, + "learning_rate": 4.890723408847664e-05, + "loss": 3.2586, + "step": 5998 + }, + { + "epoch": 0.1850479469749257, + "grad_norm": 0.15535609424114227, + "learning_rate": 4.890644676953586e-05, + "loss": 3.2284, + "step": 5999 + }, + { + "epoch": 0.1850787934404991, + "grad_norm": 0.16168494522571564, + "learning_rate": 4.890565917341419e-05, + "loss": 3.2583, + "step": 6000 + }, + { + "epoch": 0.18510963990607252, + "grad_norm": 0.15538431704044342, + "learning_rate": 4.890487130012077e-05, + "loss": 3.2768, + "step": 6001 + }, + { + "epoch": 0.18514048637164593, + "grad_norm": 0.15997786819934845, + "learning_rate": 4.890408314966474e-05, + "loss": 3.2352, + "step": 6002 + }, + { + "epoch": 0.18517133283721934, + "grad_norm": 0.1606704145669937, + "learning_rate": 4.8903294722055215e-05, + "loss": 3.2269, + "step": 6003 + }, + { + "epoch": 0.18520217930279276, + "grad_norm": 0.16704124212265015, + "learning_rate": 4.8902506017301364e-05, + "loss": 3.2218, + "step": 6004 + }, + { + "epoch": 0.18523302576836617, + "grad_norm": 0.1746658980846405, + "learning_rate": 4.890171703541231e-05, + "loss": 3.2534, + "step": 6005 + }, + { + "epoch": 0.18526387223393959, + "grad_norm": 0.19808685779571533, + "learning_rate": 4.890092777639722e-05, + "loss": 3.2428, + "step": 6006 + }, + { + "epoch": 0.185294718699513, + "grad_norm": 0.16401612758636475, + "learning_rate": 4.890013824026524e-05, + "loss": 3.2191, + "step": 6007 + }, + { + "epoch": 0.18532556516508641, + "grad_norm": 0.17441092431545258, + "learning_rate": 4.889934842702551e-05, + "loss": 3.2176, + "step": 6008 + }, + { + "epoch": 0.18535641163065986, + "grad_norm": 0.1584986448287964, + "learning_rate": 4.889855833668719e-05, + "loss": 3.1979, + "step": 6009 + }, + { + "epoch": 0.18538725809623327, + "grad_norm": 0.1625232994556427, + "learning_rate": 4.889776796925946e-05, + "loss": 3.2255, + "step": 6010 + }, + { + "epoch": 0.18541810456180668, + "grad_norm": 0.15631835162639618, + "learning_rate": 4.889697732475146e-05, + "loss": 3.2334, + "step": 6011 + }, + { + "epoch": 0.1854489510273801, + "grad_norm": 0.16614660620689392, + "learning_rate": 4.889618640317238e-05, + "loss": 3.2201, + "step": 6012 + }, + { + "epoch": 0.1854797974929535, + "grad_norm": 0.15326835215091705, + "learning_rate": 4.889539520453137e-05, + "loss": 3.2332, + "step": 6013 + }, + { + "epoch": 0.18551064395852693, + "grad_norm": 0.1716015487909317, + "learning_rate": 4.889460372883762e-05, + "loss": 3.241, + "step": 6014 + }, + { + "epoch": 0.18554149042410034, + "grad_norm": 0.16324900090694427, + "learning_rate": 4.889381197610029e-05, + "loss": 3.2328, + "step": 6015 + }, + { + "epoch": 0.18557233688967376, + "grad_norm": 0.17172645032405853, + "learning_rate": 4.889301994632858e-05, + "loss": 3.2363, + "step": 6016 + }, + { + "epoch": 0.18560318335524717, + "grad_norm": 0.1547732949256897, + "learning_rate": 4.8892227639531654e-05, + "loss": 3.2204, + "step": 6017 + }, + { + "epoch": 0.18563402982082058, + "grad_norm": 0.16318701207637787, + "learning_rate": 4.88914350557187e-05, + "loss": 3.2245, + "step": 6018 + }, + { + "epoch": 0.185664876286394, + "grad_norm": 0.16063128411769867, + "learning_rate": 4.8890642194898924e-05, + "loss": 3.2602, + "step": 6019 + }, + { + "epoch": 0.1856957227519674, + "grad_norm": 0.1609872430562973, + "learning_rate": 4.8889849057081504e-05, + "loss": 3.2445, + "step": 6020 + }, + { + "epoch": 0.18572656921754085, + "grad_norm": 0.16039235889911652, + "learning_rate": 4.888905564227564e-05, + "loss": 3.2651, + "step": 6021 + }, + { + "epoch": 0.18575741568311427, + "grad_norm": 0.16731169819831848, + "learning_rate": 4.888826195049053e-05, + "loss": 3.2627, + "step": 6022 + }, + { + "epoch": 0.18578826214868768, + "grad_norm": 0.16445545852184296, + "learning_rate": 4.888746798173538e-05, + "loss": 3.2533, + "step": 6023 + }, + { + "epoch": 0.1858191086142611, + "grad_norm": 0.1738460808992386, + "learning_rate": 4.888667373601939e-05, + "loss": 3.2151, + "step": 6024 + }, + { + "epoch": 0.1858499550798345, + "grad_norm": 0.1624167114496231, + "learning_rate": 4.8885879213351785e-05, + "loss": 3.2138, + "step": 6025 + }, + { + "epoch": 0.18588080154540793, + "grad_norm": 0.16950969398021698, + "learning_rate": 4.8885084413741754e-05, + "loss": 3.2809, + "step": 6026 + }, + { + "epoch": 0.18591164801098134, + "grad_norm": 0.16658611595630646, + "learning_rate": 4.888428933719852e-05, + "loss": 3.2337, + "step": 6027 + }, + { + "epoch": 0.18594249447655475, + "grad_norm": 0.1752244085073471, + "learning_rate": 4.8883493983731307e-05, + "loss": 3.2411, + "step": 6028 + }, + { + "epoch": 0.18597334094212817, + "grad_norm": 0.17151294648647308, + "learning_rate": 4.8882698353349335e-05, + "loss": 3.2559, + "step": 6029 + }, + { + "epoch": 0.18600418740770158, + "grad_norm": 0.18705949187278748, + "learning_rate": 4.888190244606184e-05, + "loss": 3.2417, + "step": 6030 + }, + { + "epoch": 0.186035033873275, + "grad_norm": 0.1672133207321167, + "learning_rate": 4.8881106261878015e-05, + "loss": 3.2068, + "step": 6031 + }, + { + "epoch": 0.1860658803388484, + "grad_norm": 0.1647646576166153, + "learning_rate": 4.8880309800807125e-05, + "loss": 3.2421, + "step": 6032 + }, + { + "epoch": 0.18609672680442185, + "grad_norm": 0.16919410228729248, + "learning_rate": 4.887951306285839e-05, + "loss": 3.2677, + "step": 6033 + }, + { + "epoch": 0.18612757326999527, + "grad_norm": 0.1582053005695343, + "learning_rate": 4.887871604804105e-05, + "loss": 3.2557, + "step": 6034 + }, + { + "epoch": 0.18615841973556868, + "grad_norm": 0.15723252296447754, + "learning_rate": 4.887791875636435e-05, + "loss": 3.2429, + "step": 6035 + }, + { + "epoch": 0.1861892662011421, + "grad_norm": 0.1622195839881897, + "learning_rate": 4.887712118783753e-05, + "loss": 3.2575, + "step": 6036 + }, + { + "epoch": 0.1862201126667155, + "grad_norm": 0.17079941928386688, + "learning_rate": 4.887632334246984e-05, + "loss": 3.2522, + "step": 6037 + }, + { + "epoch": 0.18625095913228892, + "grad_norm": 0.16479776799678802, + "learning_rate": 4.887552522027052e-05, + "loss": 3.2566, + "step": 6038 + }, + { + "epoch": 0.18628180559786234, + "grad_norm": 0.16514192521572113, + "learning_rate": 4.8874726821248843e-05, + "loss": 3.2501, + "step": 6039 + }, + { + "epoch": 0.18631265206343575, + "grad_norm": 0.15174098312854767, + "learning_rate": 4.887392814541405e-05, + "loss": 3.2237, + "step": 6040 + }, + { + "epoch": 0.18634349852900917, + "grad_norm": 0.16451461613178253, + "learning_rate": 4.887312919277541e-05, + "loss": 3.2365, + "step": 6041 + }, + { + "epoch": 0.18637434499458258, + "grad_norm": 0.15703947842121124, + "learning_rate": 4.887232996334219e-05, + "loss": 3.2429, + "step": 6042 + }, + { + "epoch": 0.186405191460156, + "grad_norm": 0.1551700085401535, + "learning_rate": 4.887153045712364e-05, + "loss": 3.201, + "step": 6043 + }, + { + "epoch": 0.1864360379257294, + "grad_norm": 0.16830284893512726, + "learning_rate": 4.887073067412904e-05, + "loss": 3.2374, + "step": 6044 + }, + { + "epoch": 0.18646688439130285, + "grad_norm": 0.16250954568386078, + "learning_rate": 4.886993061436767e-05, + "loss": 3.26, + "step": 6045 + }, + { + "epoch": 0.18649773085687626, + "grad_norm": 0.16182902455329895, + "learning_rate": 4.886913027784879e-05, + "loss": 3.2526, + "step": 6046 + }, + { + "epoch": 0.18652857732244968, + "grad_norm": 0.16552108526229858, + "learning_rate": 4.8868329664581685e-05, + "loss": 3.228, + "step": 6047 + }, + { + "epoch": 0.1865594237880231, + "grad_norm": 0.15619157254695892, + "learning_rate": 4.886752877457565e-05, + "loss": 3.2686, + "step": 6048 + }, + { + "epoch": 0.1865902702535965, + "grad_norm": 0.1560971885919571, + "learning_rate": 4.8866727607839955e-05, + "loss": 3.2189, + "step": 6049 + }, + { + "epoch": 0.18662111671916992, + "grad_norm": 0.1707444041967392, + "learning_rate": 4.88659261643839e-05, + "loss": 3.2418, + "step": 6050 + }, + { + "epoch": 0.18665196318474334, + "grad_norm": 0.1600903570652008, + "learning_rate": 4.886512444421677e-05, + "loss": 3.2781, + "step": 6051 + }, + { + "epoch": 0.18668280965031675, + "grad_norm": 0.15959696471691132, + "learning_rate": 4.886432244734787e-05, + "loss": 3.2394, + "step": 6052 + }, + { + "epoch": 0.18671365611589016, + "grad_norm": 0.15281204879283905, + "learning_rate": 4.886352017378648e-05, + "loss": 3.2451, + "step": 6053 + }, + { + "epoch": 0.18674450258146358, + "grad_norm": 0.1585131138563156, + "learning_rate": 4.886271762354192e-05, + "loss": 3.2671, + "step": 6054 + }, + { + "epoch": 0.186775349047037, + "grad_norm": 0.1627529263496399, + "learning_rate": 4.8861914796623495e-05, + "loss": 3.2406, + "step": 6055 + }, + { + "epoch": 0.18680619551261043, + "grad_norm": 0.15568335354328156, + "learning_rate": 4.88611116930405e-05, + "loss": 3.2089, + "step": 6056 + }, + { + "epoch": 0.18683704197818385, + "grad_norm": 0.43050721287727356, + "learning_rate": 4.886030831280225e-05, + "loss": 3.2376, + "step": 6057 + }, + { + "epoch": 0.18686788844375726, + "grad_norm": 0.1574643850326538, + "learning_rate": 4.8859504655918075e-05, + "loss": 3.245, + "step": 6058 + }, + { + "epoch": 0.18689873490933068, + "grad_norm": 0.17525024712085724, + "learning_rate": 4.8858700722397274e-05, + "loss": 3.2604, + "step": 6059 + }, + { + "epoch": 0.1869295813749041, + "grad_norm": 0.15850631892681122, + "learning_rate": 4.885789651224918e-05, + "loss": 3.2423, + "step": 6060 + }, + { + "epoch": 0.1869604278404775, + "grad_norm": 0.18843333423137665, + "learning_rate": 4.885709202548311e-05, + "loss": 3.196, + "step": 6061 + }, + { + "epoch": 0.18699127430605092, + "grad_norm": 0.16003353893756866, + "learning_rate": 4.885628726210839e-05, + "loss": 3.2607, + "step": 6062 + }, + { + "epoch": 0.18702212077162433, + "grad_norm": 0.16914694011211395, + "learning_rate": 4.885548222213436e-05, + "loss": 3.2171, + "step": 6063 + }, + { + "epoch": 0.18705296723719775, + "grad_norm": 0.1690816730260849, + "learning_rate": 4.885467690557035e-05, + "loss": 3.2155, + "step": 6064 + }, + { + "epoch": 0.18708381370277116, + "grad_norm": 0.1511792242527008, + "learning_rate": 4.8853871312425694e-05, + "loss": 3.2484, + "step": 6065 + }, + { + "epoch": 0.18711466016834458, + "grad_norm": 0.1633671671152115, + "learning_rate": 4.885306544270974e-05, + "loss": 3.2351, + "step": 6066 + }, + { + "epoch": 0.187145506633918, + "grad_norm": 0.16696405410766602, + "learning_rate": 4.8852259296431825e-05, + "loss": 3.238, + "step": 6067 + }, + { + "epoch": 0.18717635309949143, + "grad_norm": 0.1529703140258789, + "learning_rate": 4.885145287360129e-05, + "loss": 3.2457, + "step": 6068 + }, + { + "epoch": 0.18720719956506485, + "grad_norm": 0.16445069015026093, + "learning_rate": 4.8850646174227496e-05, + "loss": 3.2548, + "step": 6069 + }, + { + "epoch": 0.18723804603063826, + "grad_norm": 0.1568254828453064, + "learning_rate": 4.884983919831979e-05, + "loss": 3.249, + "step": 6070 + }, + { + "epoch": 0.18726889249621168, + "grad_norm": 0.33624809980392456, + "learning_rate": 4.884903194588754e-05, + "loss": 3.2385, + "step": 6071 + }, + { + "epoch": 0.1872997389617851, + "grad_norm": 0.15229809284210205, + "learning_rate": 4.884822441694008e-05, + "loss": 3.2118, + "step": 6072 + }, + { + "epoch": 0.1873305854273585, + "grad_norm": 0.16259215772151947, + "learning_rate": 4.88474166114868e-05, + "loss": 3.2385, + "step": 6073 + }, + { + "epoch": 0.18736143189293192, + "grad_norm": 0.16566313803195953, + "learning_rate": 4.884660852953705e-05, + "loss": 3.2181, + "step": 6074 + }, + { + "epoch": 0.18739227835850533, + "grad_norm": 0.15590974688529968, + "learning_rate": 4.884580017110021e-05, + "loss": 3.2541, + "step": 6075 + }, + { + "epoch": 0.18742312482407875, + "grad_norm": 0.15884369611740112, + "learning_rate": 4.8844991536185645e-05, + "loss": 3.2751, + "step": 6076 + }, + { + "epoch": 0.18745397128965216, + "grad_norm": 0.15809917449951172, + "learning_rate": 4.8844182624802733e-05, + "loss": 3.2395, + "step": 6077 + }, + { + "epoch": 0.18748481775522557, + "grad_norm": 0.1636354774236679, + "learning_rate": 4.884337343696085e-05, + "loss": 3.2366, + "step": 6078 + }, + { + "epoch": 0.187515664220799, + "grad_norm": 0.15766721963882446, + "learning_rate": 4.884256397266938e-05, + "loss": 3.2105, + "step": 6079 + }, + { + "epoch": 0.18754651068637243, + "grad_norm": 0.15489202737808228, + "learning_rate": 4.884175423193771e-05, + "loss": 3.2131, + "step": 6080 + }, + { + "epoch": 0.18757735715194584, + "grad_norm": 0.15795092284679413, + "learning_rate": 4.884094421477522e-05, + "loss": 3.2555, + "step": 6081 + }, + { + "epoch": 0.18760820361751926, + "grad_norm": 0.1795729249715805, + "learning_rate": 4.8840133921191316e-05, + "loss": 3.221, + "step": 6082 + }, + { + "epoch": 0.18763905008309267, + "grad_norm": 0.1577662229537964, + "learning_rate": 4.8839323351195386e-05, + "loss": 3.2372, + "step": 6083 + }, + { + "epoch": 0.1876698965486661, + "grad_norm": 0.1572243869304657, + "learning_rate": 4.883851250479682e-05, + "loss": 3.225, + "step": 6084 + }, + { + "epoch": 0.1877007430142395, + "grad_norm": 0.1579417735338211, + "learning_rate": 4.883770138200503e-05, + "loss": 3.2052, + "step": 6085 + }, + { + "epoch": 0.18773158947981292, + "grad_norm": 0.1573963463306427, + "learning_rate": 4.883688998282943e-05, + "loss": 3.2429, + "step": 6086 + }, + { + "epoch": 0.18776243594538633, + "grad_norm": 0.15711437165737152, + "learning_rate": 4.88360783072794e-05, + "loss": 3.235, + "step": 6087 + }, + { + "epoch": 0.18779328241095974, + "grad_norm": 0.15172933042049408, + "learning_rate": 4.8835266355364376e-05, + "loss": 3.2313, + "step": 6088 + }, + { + "epoch": 0.18782412887653316, + "grad_norm": 0.15097083151340485, + "learning_rate": 4.8834454127093745e-05, + "loss": 3.2732, + "step": 6089 + }, + { + "epoch": 0.18785497534210657, + "grad_norm": 0.15426486730575562, + "learning_rate": 4.8833641622476956e-05, + "loss": 3.2329, + "step": 6090 + }, + { + "epoch": 0.18788582180768, + "grad_norm": 0.1559959501028061, + "learning_rate": 4.883282884152341e-05, + "loss": 3.2542, + "step": 6091 + }, + { + "epoch": 0.18791666827325343, + "grad_norm": 0.15420852601528168, + "learning_rate": 4.8832015784242534e-05, + "loss": 3.2213, + "step": 6092 + }, + { + "epoch": 0.18794751473882684, + "grad_norm": 0.15043163299560547, + "learning_rate": 4.883120245064377e-05, + "loss": 3.2034, + "step": 6093 + }, + { + "epoch": 0.18797836120440026, + "grad_norm": 0.1489507257938385, + "learning_rate": 4.883038884073652e-05, + "loss": 3.2164, + "step": 6094 + }, + { + "epoch": 0.18800920766997367, + "grad_norm": 0.14916037023067474, + "learning_rate": 4.882957495453023e-05, + "loss": 3.222, + "step": 6095 + }, + { + "epoch": 0.18804005413554709, + "grad_norm": 0.15060414373874664, + "learning_rate": 4.8828760792034346e-05, + "loss": 3.2241, + "step": 6096 + }, + { + "epoch": 0.1880709006011205, + "grad_norm": 0.15295365452766418, + "learning_rate": 4.8827946353258296e-05, + "loss": 3.2455, + "step": 6097 + }, + { + "epoch": 0.18810174706669391, + "grad_norm": 0.1536731719970703, + "learning_rate": 4.882713163821153e-05, + "loss": 3.2091, + "step": 6098 + }, + { + "epoch": 0.18813259353226733, + "grad_norm": 0.16124123334884644, + "learning_rate": 4.882631664690349e-05, + "loss": 3.2515, + "step": 6099 + }, + { + "epoch": 0.18816343999784074, + "grad_norm": 0.1587848961353302, + "learning_rate": 4.882550137934362e-05, + "loss": 3.2419, + "step": 6100 + }, + { + "epoch": 0.18819428646341416, + "grad_norm": 0.16169613599777222, + "learning_rate": 4.882468583554139e-05, + "loss": 3.2161, + "step": 6101 + }, + { + "epoch": 0.18822513292898757, + "grad_norm": 0.1614069789648056, + "learning_rate": 4.882387001550624e-05, + "loss": 3.245, + "step": 6102 + }, + { + "epoch": 0.18825597939456098, + "grad_norm": 0.15244098007678986, + "learning_rate": 4.882305391924763e-05, + "loss": 3.2274, + "step": 6103 + }, + { + "epoch": 0.18828682586013443, + "grad_norm": 0.15093785524368286, + "learning_rate": 4.882223754677503e-05, + "loss": 3.2415, + "step": 6104 + }, + { + "epoch": 0.18831767232570784, + "grad_norm": 0.16859857738018036, + "learning_rate": 4.88214208980979e-05, + "loss": 3.238, + "step": 6105 + }, + { + "epoch": 0.18834851879128126, + "grad_norm": 0.15106602013111115, + "learning_rate": 4.882060397322571e-05, + "loss": 3.2327, + "step": 6106 + }, + { + "epoch": 0.18837936525685467, + "grad_norm": 0.17525619268417358, + "learning_rate": 4.881978677216793e-05, + "loss": 3.209, + "step": 6107 + }, + { + "epoch": 0.18841021172242808, + "grad_norm": 0.15807032585144043, + "learning_rate": 4.881896929493404e-05, + "loss": 3.212, + "step": 6108 + }, + { + "epoch": 0.1884410581880015, + "grad_norm": 0.16771851480007172, + "learning_rate": 4.881815154153351e-05, + "loss": 3.2498, + "step": 6109 + }, + { + "epoch": 0.1884719046535749, + "grad_norm": 0.14987532794475555, + "learning_rate": 4.881733351197583e-05, + "loss": 3.2484, + "step": 6110 + }, + { + "epoch": 0.18850275111914833, + "grad_norm": 0.16520267724990845, + "learning_rate": 4.881651520627048e-05, + "loss": 3.2164, + "step": 6111 + }, + { + "epoch": 0.18853359758472174, + "grad_norm": 0.15412241220474243, + "learning_rate": 4.8815696624426946e-05, + "loss": 3.2524, + "step": 6112 + }, + { + "epoch": 0.18856444405029515, + "grad_norm": 0.16035529971122742, + "learning_rate": 4.8814877766454726e-05, + "loss": 3.2251, + "step": 6113 + }, + { + "epoch": 0.18859529051586857, + "grad_norm": 0.1573476642370224, + "learning_rate": 4.8814058632363305e-05, + "loss": 3.2496, + "step": 6114 + }, + { + "epoch": 0.18862613698144198, + "grad_norm": 0.1827724426984787, + "learning_rate": 4.881323922216219e-05, + "loss": 3.2411, + "step": 6115 + }, + { + "epoch": 0.18865698344701542, + "grad_norm": 0.16047921776771545, + "learning_rate": 4.8812419535860865e-05, + "loss": 3.2661, + "step": 6116 + }, + { + "epoch": 0.18868782991258884, + "grad_norm": 0.170062854886055, + "learning_rate": 4.8811599573468853e-05, + "loss": 3.2426, + "step": 6117 + }, + { + "epoch": 0.18871867637816225, + "grad_norm": 0.1700005829334259, + "learning_rate": 4.881077933499565e-05, + "loss": 3.2182, + "step": 6118 + }, + { + "epoch": 0.18874952284373567, + "grad_norm": 0.16152751445770264, + "learning_rate": 4.880995882045078e-05, + "loss": 3.2237, + "step": 6119 + }, + { + "epoch": 0.18878036930930908, + "grad_norm": 0.1594395786523819, + "learning_rate": 4.8809138029843726e-05, + "loss": 3.2234, + "step": 6120 + }, + { + "epoch": 0.1888112157748825, + "grad_norm": 0.1559724062681198, + "learning_rate": 4.880831696318404e-05, + "loss": 3.2642, + "step": 6121 + }, + { + "epoch": 0.1888420622404559, + "grad_norm": 0.1573238968849182, + "learning_rate": 4.880749562048122e-05, + "loss": 3.2296, + "step": 6122 + }, + { + "epoch": 0.18887290870602932, + "grad_norm": 0.15993179380893707, + "learning_rate": 4.88066740017448e-05, + "loss": 3.2435, + "step": 6123 + }, + { + "epoch": 0.18890375517160274, + "grad_norm": 0.16138045489788055, + "learning_rate": 4.880585210698429e-05, + "loss": 3.2515, + "step": 6124 + }, + { + "epoch": 0.18893460163717615, + "grad_norm": 0.16185319423675537, + "learning_rate": 4.880502993620924e-05, + "loss": 3.2076, + "step": 6125 + }, + { + "epoch": 0.18896544810274957, + "grad_norm": 0.16151922941207886, + "learning_rate": 4.8804207489429164e-05, + "loss": 3.2218, + "step": 6126 + }, + { + "epoch": 0.18899629456832298, + "grad_norm": 0.1551419496536255, + "learning_rate": 4.880338476665362e-05, + "loss": 3.219, + "step": 6127 + }, + { + "epoch": 0.18902714103389642, + "grad_norm": 0.16464808583259583, + "learning_rate": 4.8802561767892124e-05, + "loss": 3.2162, + "step": 6128 + }, + { + "epoch": 0.18905798749946984, + "grad_norm": 0.15367919206619263, + "learning_rate": 4.880173849315423e-05, + "loss": 3.2333, + "step": 6129 + }, + { + "epoch": 0.18908883396504325, + "grad_norm": 0.16632257401943207, + "learning_rate": 4.880091494244948e-05, + "loss": 3.2541, + "step": 6130 + }, + { + "epoch": 0.18911968043061667, + "grad_norm": 0.1648884117603302, + "learning_rate": 4.880009111578743e-05, + "loss": 3.2298, + "step": 6131 + }, + { + "epoch": 0.18915052689619008, + "grad_norm": 0.15953028202056885, + "learning_rate": 4.879926701317762e-05, + "loss": 3.2268, + "step": 6132 + }, + { + "epoch": 0.1891813733617635, + "grad_norm": 0.15638943016529083, + "learning_rate": 4.8798442634629615e-05, + "loss": 3.2221, + "step": 6133 + }, + { + "epoch": 0.1892122198273369, + "grad_norm": 0.1606038510799408, + "learning_rate": 4.8797617980152965e-05, + "loss": 3.2371, + "step": 6134 + }, + { + "epoch": 0.18924306629291032, + "grad_norm": 0.15749245882034302, + "learning_rate": 4.879679304975724e-05, + "loss": 3.2286, + "step": 6135 + }, + { + "epoch": 0.18927391275848374, + "grad_norm": 0.1598139852285385, + "learning_rate": 4.8795967843452e-05, + "loss": 3.2477, + "step": 6136 + }, + { + "epoch": 0.18930475922405715, + "grad_norm": 0.16248652338981628, + "learning_rate": 4.8795142361246804e-05, + "loss": 3.2437, + "step": 6137 + }, + { + "epoch": 0.18933560568963057, + "grad_norm": 0.15393821895122528, + "learning_rate": 4.8794316603151235e-05, + "loss": 3.2172, + "step": 6138 + }, + { + "epoch": 0.18936645215520398, + "grad_norm": 0.1616666167974472, + "learning_rate": 4.879349056917487e-05, + "loss": 3.236, + "step": 6139 + }, + { + "epoch": 0.18939729862077742, + "grad_norm": 0.16241222620010376, + "learning_rate": 4.879266425932727e-05, + "loss": 3.2328, + "step": 6140 + }, + { + "epoch": 0.18942814508635084, + "grad_norm": 0.17272332310676575, + "learning_rate": 4.879183767361803e-05, + "loss": 3.2256, + "step": 6141 + }, + { + "epoch": 0.18945899155192425, + "grad_norm": 0.15801957249641418, + "learning_rate": 4.8791010812056736e-05, + "loss": 3.2332, + "step": 6142 + }, + { + "epoch": 0.18948983801749766, + "grad_norm": 0.16489280760288239, + "learning_rate": 4.879018367465295e-05, + "loss": 3.2325, + "step": 6143 + }, + { + "epoch": 0.18952068448307108, + "grad_norm": 0.16383247077465057, + "learning_rate": 4.87893562614163e-05, + "loss": 3.2606, + "step": 6144 + }, + { + "epoch": 0.1895515309486445, + "grad_norm": 0.1588422805070877, + "learning_rate": 4.8788528572356356e-05, + "loss": 3.2252, + "step": 6145 + }, + { + "epoch": 0.1895823774142179, + "grad_norm": 0.1612585335969925, + "learning_rate": 4.878770060748272e-05, + "loss": 3.245, + "step": 6146 + }, + { + "epoch": 0.18961322387979132, + "grad_norm": 0.16314376890659332, + "learning_rate": 4.8786872366804984e-05, + "loss": 3.2264, + "step": 6147 + }, + { + "epoch": 0.18964407034536473, + "grad_norm": 0.15986962616443634, + "learning_rate": 4.878604385033276e-05, + "loss": 3.2388, + "step": 6148 + }, + { + "epoch": 0.18967491681093815, + "grad_norm": 0.15550343692302704, + "learning_rate": 4.8785215058075645e-05, + "loss": 3.2084, + "step": 6149 + }, + { + "epoch": 0.18970576327651156, + "grad_norm": 0.16478365659713745, + "learning_rate": 4.8784385990043265e-05, + "loss": 3.2431, + "step": 6150 + }, + { + "epoch": 0.18973660974208498, + "grad_norm": 0.16340480744838715, + "learning_rate": 4.8783556646245215e-05, + "loss": 3.2554, + "step": 6151 + }, + { + "epoch": 0.18976745620765842, + "grad_norm": 0.17266933619976044, + "learning_rate": 4.878272702669112e-05, + "loss": 3.2252, + "step": 6152 + }, + { + "epoch": 0.18979830267323183, + "grad_norm": 0.15964913368225098, + "learning_rate": 4.87818971313906e-05, + "loss": 3.2285, + "step": 6153 + }, + { + "epoch": 0.18982914913880525, + "grad_norm": 0.15429723262786865, + "learning_rate": 4.8781066960353264e-05, + "loss": 3.2825, + "step": 6154 + }, + { + "epoch": 0.18985999560437866, + "grad_norm": 0.16399453580379486, + "learning_rate": 4.8780236513588754e-05, + "loss": 3.2553, + "step": 6155 + }, + { + "epoch": 0.18989084206995208, + "grad_norm": 0.1508844494819641, + "learning_rate": 4.877940579110669e-05, + "loss": 3.2672, + "step": 6156 + }, + { + "epoch": 0.1899216885355255, + "grad_norm": 0.17073951661586761, + "learning_rate": 4.8778574792916707e-05, + "loss": 3.2879, + "step": 6157 + }, + { + "epoch": 0.1899525350010989, + "grad_norm": 0.15765662491321564, + "learning_rate": 4.877774351902845e-05, + "loss": 3.2496, + "step": 6158 + }, + { + "epoch": 0.18998338146667232, + "grad_norm": 0.15842951834201813, + "learning_rate": 4.877691196945153e-05, + "loss": 3.2253, + "step": 6159 + }, + { + "epoch": 0.19001422793224573, + "grad_norm": 0.1572420448064804, + "learning_rate": 4.877608014419561e-05, + "loss": 3.2265, + "step": 6160 + }, + { + "epoch": 0.19004507439781915, + "grad_norm": 0.15526476502418518, + "learning_rate": 4.877524804327033e-05, + "loss": 3.2588, + "step": 6161 + }, + { + "epoch": 0.19007592086339256, + "grad_norm": 0.16067425906658173, + "learning_rate": 4.877441566668533e-05, + "loss": 3.245, + "step": 6162 + }, + { + "epoch": 0.19010676732896598, + "grad_norm": 0.15239831805229187, + "learning_rate": 4.877358301445028e-05, + "loss": 3.2701, + "step": 6163 + }, + { + "epoch": 0.19013761379453942, + "grad_norm": 0.16387666761875153, + "learning_rate": 4.87727500865748e-05, + "loss": 3.2135, + "step": 6164 + }, + { + "epoch": 0.19016846026011283, + "grad_norm": 0.16066960990428925, + "learning_rate": 4.8771916883068585e-05, + "loss": 3.2186, + "step": 6165 + }, + { + "epoch": 0.19019930672568625, + "grad_norm": 0.15917424857616425, + "learning_rate": 4.8771083403941276e-05, + "loss": 3.2381, + "step": 6166 + }, + { + "epoch": 0.19023015319125966, + "grad_norm": 0.15044836699962616, + "learning_rate": 4.877024964920254e-05, + "loss": 3.2395, + "step": 6167 + }, + { + "epoch": 0.19026099965683307, + "grad_norm": 0.1557704210281372, + "learning_rate": 4.876941561886203e-05, + "loss": 3.2731, + "step": 6168 + }, + { + "epoch": 0.1902918461224065, + "grad_norm": 0.1573891043663025, + "learning_rate": 4.876858131292945e-05, + "loss": 3.224, + "step": 6169 + }, + { + "epoch": 0.1903226925879799, + "grad_norm": 0.157740518450737, + "learning_rate": 4.8767746731414444e-05, + "loss": 3.2119, + "step": 6170 + }, + { + "epoch": 0.19035353905355332, + "grad_norm": 0.16172797977924347, + "learning_rate": 4.87669118743267e-05, + "loss": 3.2435, + "step": 6171 + }, + { + "epoch": 0.19038438551912673, + "grad_norm": 0.15136629343032837, + "learning_rate": 4.8766076741675894e-05, + "loss": 3.233, + "step": 6172 + }, + { + "epoch": 0.19041523198470015, + "grad_norm": 0.1511928141117096, + "learning_rate": 4.8765241333471705e-05, + "loss": 3.1928, + "step": 6173 + }, + { + "epoch": 0.19044607845027356, + "grad_norm": 0.16926440596580505, + "learning_rate": 4.876440564972383e-05, + "loss": 3.2312, + "step": 6174 + }, + { + "epoch": 0.19047692491584697, + "grad_norm": 0.15834958851337433, + "learning_rate": 4.876356969044194e-05, + "loss": 3.2423, + "step": 6175 + }, + { + "epoch": 0.19050777138142042, + "grad_norm": 0.16788898408412933, + "learning_rate": 4.8762733455635756e-05, + "loss": 3.2367, + "step": 6176 + }, + { + "epoch": 0.19053861784699383, + "grad_norm": 0.1602586954832077, + "learning_rate": 4.876189694531495e-05, + "loss": 3.2293, + "step": 6177 + }, + { + "epoch": 0.19056946431256724, + "grad_norm": 0.16592544317245483, + "learning_rate": 4.876106015948924e-05, + "loss": 3.2703, + "step": 6178 + }, + { + "epoch": 0.19060031077814066, + "grad_norm": 0.15766368806362152, + "learning_rate": 4.8760223098168304e-05, + "loss": 3.2199, + "step": 6179 + }, + { + "epoch": 0.19063115724371407, + "grad_norm": 0.15526574850082397, + "learning_rate": 4.875938576136185e-05, + "loss": 3.2466, + "step": 6180 + }, + { + "epoch": 0.1906620037092875, + "grad_norm": 0.16050532460212708, + "learning_rate": 4.875854814907961e-05, + "loss": 3.2468, + "step": 6181 + }, + { + "epoch": 0.1906928501748609, + "grad_norm": 0.15633049607276917, + "learning_rate": 4.875771026133128e-05, + "loss": 3.1927, + "step": 6182 + }, + { + "epoch": 0.19072369664043431, + "grad_norm": 0.16153205931186676, + "learning_rate": 4.875687209812657e-05, + "loss": 3.2288, + "step": 6183 + }, + { + "epoch": 0.19075454310600773, + "grad_norm": 0.15347881615161896, + "learning_rate": 4.875603365947521e-05, + "loss": 3.2493, + "step": 6184 + }, + { + "epoch": 0.19078538957158114, + "grad_norm": 0.15180538594722748, + "learning_rate": 4.875519494538692e-05, + "loss": 3.2262, + "step": 6185 + }, + { + "epoch": 0.19081623603715456, + "grad_norm": 0.15619748830795288, + "learning_rate": 4.8754355955871415e-05, + "loss": 3.2337, + "step": 6186 + }, + { + "epoch": 0.19084708250272797, + "grad_norm": 0.15815778076648712, + "learning_rate": 4.875351669093842e-05, + "loss": 3.2152, + "step": 6187 + }, + { + "epoch": 0.1908779289683014, + "grad_norm": 0.15643107891082764, + "learning_rate": 4.875267715059768e-05, + "loss": 3.2517, + "step": 6188 + }, + { + "epoch": 0.19090877543387483, + "grad_norm": 0.16229692101478577, + "learning_rate": 4.875183733485892e-05, + "loss": 3.2457, + "step": 6189 + }, + { + "epoch": 0.19093962189944824, + "grad_norm": 0.16087572276592255, + "learning_rate": 4.875099724373188e-05, + "loss": 3.2515, + "step": 6190 + }, + { + "epoch": 0.19097046836502166, + "grad_norm": 0.15625247359275818, + "learning_rate": 4.87501568772263e-05, + "loss": 3.2306, + "step": 6191 + }, + { + "epoch": 0.19100131483059507, + "grad_norm": 0.16335241496562958, + "learning_rate": 4.874931623535193e-05, + "loss": 3.2393, + "step": 6192 + }, + { + "epoch": 0.19103216129616848, + "grad_norm": 0.15984423458576202, + "learning_rate": 4.8748475318118505e-05, + "loss": 3.2137, + "step": 6193 + }, + { + "epoch": 0.1910630077617419, + "grad_norm": 0.16108006238937378, + "learning_rate": 4.874763412553577e-05, + "loss": 3.2383, + "step": 6194 + }, + { + "epoch": 0.1910938542273153, + "grad_norm": 0.15490470826625824, + "learning_rate": 4.874679265761351e-05, + "loss": 3.2074, + "step": 6195 + }, + { + "epoch": 0.19112470069288873, + "grad_norm": 0.15596936643123627, + "learning_rate": 4.8745950914361435e-05, + "loss": 3.2373, + "step": 6196 + }, + { + "epoch": 0.19115554715846214, + "grad_norm": 0.15281757712364197, + "learning_rate": 4.874510889578934e-05, + "loss": 3.2346, + "step": 6197 + }, + { + "epoch": 0.19118639362403556, + "grad_norm": 0.15200237929821014, + "learning_rate": 4.874426660190697e-05, + "loss": 3.2071, + "step": 6198 + }, + { + "epoch": 0.19121724008960897, + "grad_norm": 0.15063923597335815, + "learning_rate": 4.874342403272411e-05, + "loss": 3.2238, + "step": 6199 + }, + { + "epoch": 0.1912480865551824, + "grad_norm": 0.15348318219184875, + "learning_rate": 4.8742581188250516e-05, + "loss": 3.2238, + "step": 6200 + }, + { + "epoch": 0.19127893302075583, + "grad_norm": 0.15257467329502106, + "learning_rate": 4.874173806849595e-05, + "loss": 3.2098, + "step": 6201 + }, + { + "epoch": 0.19130977948632924, + "grad_norm": 0.15351712703704834, + "learning_rate": 4.874089467347021e-05, + "loss": 3.2213, + "step": 6202 + }, + { + "epoch": 0.19134062595190265, + "grad_norm": 0.15509463846683502, + "learning_rate": 4.874005100318305e-05, + "loss": 3.2497, + "step": 6203 + }, + { + "epoch": 0.19137147241747607, + "grad_norm": 0.1524890810251236, + "learning_rate": 4.8739207057644266e-05, + "loss": 3.2125, + "step": 6204 + }, + { + "epoch": 0.19140231888304948, + "grad_norm": 0.156911700963974, + "learning_rate": 4.873836283686365e-05, + "loss": 3.2474, + "step": 6205 + }, + { + "epoch": 0.1914331653486229, + "grad_norm": 0.16440778970718384, + "learning_rate": 4.873751834085097e-05, + "loss": 3.2482, + "step": 6206 + }, + { + "epoch": 0.1914640118141963, + "grad_norm": 0.1799805611371994, + "learning_rate": 4.873667356961604e-05, + "loss": 3.232, + "step": 6207 + }, + { + "epoch": 0.19149485827976973, + "grad_norm": 0.48116540908813477, + "learning_rate": 4.8735828523168645e-05, + "loss": 3.2592, + "step": 6208 + }, + { + "epoch": 0.19152570474534314, + "grad_norm": 0.1613529473543167, + "learning_rate": 4.873498320151858e-05, + "loss": 3.2648, + "step": 6209 + }, + { + "epoch": 0.19155655121091655, + "grad_norm": 0.15901337563991547, + "learning_rate": 4.8734137604675645e-05, + "loss": 3.242, + "step": 6210 + }, + { + "epoch": 0.19158739767648997, + "grad_norm": 0.16051101684570312, + "learning_rate": 4.873329173264965e-05, + "loss": 3.206, + "step": 6211 + }, + { + "epoch": 0.1916182441420634, + "grad_norm": 0.1586272120475769, + "learning_rate": 4.87324455854504e-05, + "loss": 3.2579, + "step": 6212 + }, + { + "epoch": 0.19164909060763682, + "grad_norm": 0.15820303559303284, + "learning_rate": 4.873159916308771e-05, + "loss": 3.2334, + "step": 6213 + }, + { + "epoch": 0.19167993707321024, + "grad_norm": 0.16201430559158325, + "learning_rate": 4.8730752465571375e-05, + "loss": 3.2506, + "step": 6214 + }, + { + "epoch": 0.19171078353878365, + "grad_norm": 0.15476928651332855, + "learning_rate": 4.872990549291123e-05, + "loss": 3.2424, + "step": 6215 + }, + { + "epoch": 0.19174163000435707, + "grad_norm": 0.15661296248435974, + "learning_rate": 4.872905824511709e-05, + "loss": 3.2302, + "step": 6216 + }, + { + "epoch": 0.19177247646993048, + "grad_norm": 0.15568751096725464, + "learning_rate": 4.872821072219878e-05, + "loss": 3.254, + "step": 6217 + }, + { + "epoch": 0.1918033229355039, + "grad_norm": 0.15689770877361298, + "learning_rate": 4.872736292416613e-05, + "loss": 3.2308, + "step": 6218 + }, + { + "epoch": 0.1918341694010773, + "grad_norm": 0.1539836972951889, + "learning_rate": 4.872651485102897e-05, + "loss": 3.2131, + "step": 6219 + }, + { + "epoch": 0.19186501586665072, + "grad_norm": 0.16221430897712708, + "learning_rate": 4.872566650279712e-05, + "loss": 3.26, + "step": 6220 + }, + { + "epoch": 0.19189586233222414, + "grad_norm": 0.1633770763874054, + "learning_rate": 4.872481787948042e-05, + "loss": 3.2117, + "step": 6221 + }, + { + "epoch": 0.19192670879779755, + "grad_norm": 0.2154778093099594, + "learning_rate": 4.872396898108873e-05, + "loss": 3.2505, + "step": 6222 + }, + { + "epoch": 0.19195755526337097, + "grad_norm": 0.16221822798252106, + "learning_rate": 4.8723119807631854e-05, + "loss": 3.2166, + "step": 6223 + }, + { + "epoch": 0.1919884017289444, + "grad_norm": 0.1752185970544815, + "learning_rate": 4.872227035911967e-05, + "loss": 3.2244, + "step": 6224 + }, + { + "epoch": 0.19201924819451782, + "grad_norm": 0.16990555822849274, + "learning_rate": 4.872142063556202e-05, + "loss": 3.195, + "step": 6225 + }, + { + "epoch": 0.19205009466009124, + "grad_norm": 0.15366683900356293, + "learning_rate": 4.8720570636968756e-05, + "loss": 3.2049, + "step": 6226 + }, + { + "epoch": 0.19208094112566465, + "grad_norm": 0.16328921914100647, + "learning_rate": 4.8719720363349726e-05, + "loss": 3.2446, + "step": 6227 + }, + { + "epoch": 0.19211178759123806, + "grad_norm": 0.15476307272911072, + "learning_rate": 4.871886981471479e-05, + "loss": 3.2576, + "step": 6228 + }, + { + "epoch": 0.19214263405681148, + "grad_norm": 0.15561333298683167, + "learning_rate": 4.871801899107381e-05, + "loss": 3.2515, + "step": 6229 + }, + { + "epoch": 0.1921734805223849, + "grad_norm": 0.2764318287372589, + "learning_rate": 4.8717167892436656e-05, + "loss": 3.2181, + "step": 6230 + }, + { + "epoch": 0.1922043269879583, + "grad_norm": 0.15828147530555725, + "learning_rate": 4.871631651881319e-05, + "loss": 3.2563, + "step": 6231 + }, + { + "epoch": 0.19223517345353172, + "grad_norm": 0.16792063415050507, + "learning_rate": 4.871546487021329e-05, + "loss": 3.2299, + "step": 6232 + }, + { + "epoch": 0.19226601991910514, + "grad_norm": 0.15625569224357605, + "learning_rate": 4.871461294664682e-05, + "loss": 3.2507, + "step": 6233 + }, + { + "epoch": 0.19229686638467855, + "grad_norm": 0.27225261926651, + "learning_rate": 4.871376074812367e-05, + "loss": 3.2305, + "step": 6234 + }, + { + "epoch": 0.19232771285025196, + "grad_norm": 0.15470930933952332, + "learning_rate": 4.8712908274653714e-05, + "loss": 3.205, + "step": 6235 + }, + { + "epoch": 0.1923585593158254, + "grad_norm": 0.15069548785686493, + "learning_rate": 4.871205552624683e-05, + "loss": 3.2318, + "step": 6236 + }, + { + "epoch": 0.19238940578139882, + "grad_norm": 0.15797556936740875, + "learning_rate": 4.8711202502912925e-05, + "loss": 3.2427, + "step": 6237 + }, + { + "epoch": 0.19242025224697223, + "grad_norm": 0.15897372364997864, + "learning_rate": 4.8710349204661863e-05, + "loss": 3.24, + "step": 6238 + }, + { + "epoch": 0.19245109871254565, + "grad_norm": 0.15797758102416992, + "learning_rate": 4.8709495631503555e-05, + "loss": 3.2299, + "step": 6239 + }, + { + "epoch": 0.19248194517811906, + "grad_norm": 0.15660859644412994, + "learning_rate": 4.8708641783447894e-05, + "loss": 3.2257, + "step": 6240 + }, + { + "epoch": 0.19251279164369248, + "grad_norm": 0.1626482903957367, + "learning_rate": 4.870778766050478e-05, + "loss": 3.2388, + "step": 6241 + }, + { + "epoch": 0.1925436381092659, + "grad_norm": 0.16315652430057526, + "learning_rate": 4.8706933262684104e-05, + "loss": 3.2221, + "step": 6242 + }, + { + "epoch": 0.1925744845748393, + "grad_norm": 0.159071147441864, + "learning_rate": 4.8706078589995796e-05, + "loss": 3.1974, + "step": 6243 + }, + { + "epoch": 0.19260533104041272, + "grad_norm": 0.17011213302612305, + "learning_rate": 4.870522364244975e-05, + "loss": 3.252, + "step": 6244 + }, + { + "epoch": 0.19263617750598613, + "grad_norm": 0.1610531061887741, + "learning_rate": 4.8704368420055876e-05, + "loss": 3.2441, + "step": 6245 + }, + { + "epoch": 0.19266702397155955, + "grad_norm": 0.17719531059265137, + "learning_rate": 4.870351292282409e-05, + "loss": 3.2522, + "step": 6246 + }, + { + "epoch": 0.19269787043713296, + "grad_norm": 0.16323237121105194, + "learning_rate": 4.870265715076433e-05, + "loss": 3.2291, + "step": 6247 + }, + { + "epoch": 0.1927287169027064, + "grad_norm": 0.16203688085079193, + "learning_rate": 4.870180110388649e-05, + "loss": 3.2198, + "step": 6248 + }, + { + "epoch": 0.19275956336827982, + "grad_norm": 0.1565237045288086, + "learning_rate": 4.870094478220052e-05, + "loss": 3.2455, + "step": 6249 + }, + { + "epoch": 0.19279040983385323, + "grad_norm": 0.1742282509803772, + "learning_rate": 4.870008818571633e-05, + "loss": 3.2722, + "step": 6250 + }, + { + "epoch": 0.19282125629942665, + "grad_norm": 0.15272752940654755, + "learning_rate": 4.8699231314443866e-05, + "loss": 3.2192, + "step": 6251 + }, + { + "epoch": 0.19285210276500006, + "grad_norm": 0.17575524747371674, + "learning_rate": 4.869837416839306e-05, + "loss": 3.2686, + "step": 6252 + }, + { + "epoch": 0.19288294923057347, + "grad_norm": 0.15969720482826233, + "learning_rate": 4.8697516747573833e-05, + "loss": 3.2094, + "step": 6253 + }, + { + "epoch": 0.1929137956961469, + "grad_norm": 0.17099697887897491, + "learning_rate": 4.869665905199615e-05, + "loss": 3.2489, + "step": 6254 + }, + { + "epoch": 0.1929446421617203, + "grad_norm": 0.16168563067913055, + "learning_rate": 4.8695801081669946e-05, + "loss": 3.2507, + "step": 6255 + }, + { + "epoch": 0.19297548862729372, + "grad_norm": 0.19853250682353973, + "learning_rate": 4.869494283660516e-05, + "loss": 3.2475, + "step": 6256 + }, + { + "epoch": 0.19300633509286713, + "grad_norm": 0.1601652354001999, + "learning_rate": 4.8694084316811764e-05, + "loss": 3.2381, + "step": 6257 + }, + { + "epoch": 0.19303718155844055, + "grad_norm": 0.4778505861759186, + "learning_rate": 4.8693225522299684e-05, + "loss": 3.2322, + "step": 6258 + }, + { + "epoch": 0.19306802802401396, + "grad_norm": 0.1573527306318283, + "learning_rate": 4.8692366453078894e-05, + "loss": 3.2381, + "step": 6259 + }, + { + "epoch": 0.1930988744895874, + "grad_norm": 0.17634528875350952, + "learning_rate": 4.8691507109159364e-05, + "loss": 3.239, + "step": 6260 + }, + { + "epoch": 0.19312972095516082, + "grad_norm": 0.15891271829605103, + "learning_rate": 4.869064749055103e-05, + "loss": 3.2533, + "step": 6261 + }, + { + "epoch": 0.19316056742073423, + "grad_norm": 0.16367243230342865, + "learning_rate": 4.868978759726389e-05, + "loss": 3.255, + "step": 6262 + }, + { + "epoch": 0.19319141388630764, + "grad_norm": 0.16580770909786224, + "learning_rate": 4.868892742930789e-05, + "loss": 3.2078, + "step": 6263 + }, + { + "epoch": 0.19322226035188106, + "grad_norm": 0.16802170872688293, + "learning_rate": 4.868806698669301e-05, + "loss": 3.2526, + "step": 6264 + }, + { + "epoch": 0.19325310681745447, + "grad_norm": 0.16180266439914703, + "learning_rate": 4.868720626942924e-05, + "loss": 3.2469, + "step": 6265 + }, + { + "epoch": 0.1932839532830279, + "grad_norm": 0.15735091269016266, + "learning_rate": 4.8686345277526535e-05, + "loss": 3.218, + "step": 6266 + }, + { + "epoch": 0.1933147997486013, + "grad_norm": 0.16451391577720642, + "learning_rate": 4.86854840109949e-05, + "loss": 3.1877, + "step": 6267 + }, + { + "epoch": 0.19334564621417472, + "grad_norm": 0.16067738831043243, + "learning_rate": 4.86846224698443e-05, + "loss": 3.2065, + "step": 6268 + }, + { + "epoch": 0.19337649267974813, + "grad_norm": 0.16770203411579132, + "learning_rate": 4.868376065408475e-05, + "loss": 3.1969, + "step": 6269 + }, + { + "epoch": 0.19340733914532154, + "grad_norm": 0.16290228068828583, + "learning_rate": 4.868289856372622e-05, + "loss": 3.2371, + "step": 6270 + }, + { + "epoch": 0.19343818561089496, + "grad_norm": 0.1645880937576294, + "learning_rate": 4.868203619877871e-05, + "loss": 3.2014, + "step": 6271 + }, + { + "epoch": 0.1934690320764684, + "grad_norm": 0.16627942025661469, + "learning_rate": 4.8681173559252226e-05, + "loss": 3.2287, + "step": 6272 + }, + { + "epoch": 0.19349987854204181, + "grad_norm": 0.18438825011253357, + "learning_rate": 4.8680310645156755e-05, + "loss": 3.2335, + "step": 6273 + }, + { + "epoch": 0.19353072500761523, + "grad_norm": 0.16280561685562134, + "learning_rate": 4.8679447456502324e-05, + "loss": 3.2548, + "step": 6274 + }, + { + "epoch": 0.19356157147318864, + "grad_norm": 0.15931087732315063, + "learning_rate": 4.867858399329893e-05, + "loss": 3.2184, + "step": 6275 + }, + { + "epoch": 0.19359241793876206, + "grad_norm": 0.15627074241638184, + "learning_rate": 4.867772025555658e-05, + "loss": 3.2254, + "step": 6276 + }, + { + "epoch": 0.19362326440433547, + "grad_norm": 0.1551595777273178, + "learning_rate": 4.8676856243285285e-05, + "loss": 3.2345, + "step": 6277 + }, + { + "epoch": 0.19365411086990889, + "grad_norm": 0.16023243963718414, + "learning_rate": 4.8675991956495084e-05, + "loss": 3.2375, + "step": 6278 + }, + { + "epoch": 0.1936849573354823, + "grad_norm": 0.16284650564193726, + "learning_rate": 4.867512739519597e-05, + "loss": 3.2596, + "step": 6279 + }, + { + "epoch": 0.1937158038010557, + "grad_norm": 0.15737439692020416, + "learning_rate": 4.8674262559397985e-05, + "loss": 3.2397, + "step": 6280 + }, + { + "epoch": 0.19374665026662913, + "grad_norm": 0.15681138634681702, + "learning_rate": 4.867339744911116e-05, + "loss": 3.2183, + "step": 6281 + }, + { + "epoch": 0.19377749673220254, + "grad_norm": 0.16412296891212463, + "learning_rate": 4.867253206434551e-05, + "loss": 3.2246, + "step": 6282 + }, + { + "epoch": 0.19380834319777596, + "grad_norm": 0.1628301441669464, + "learning_rate": 4.867166640511108e-05, + "loss": 3.2061, + "step": 6283 + }, + { + "epoch": 0.1938391896633494, + "grad_norm": 0.1611282229423523, + "learning_rate": 4.86708004714179e-05, + "loss": 3.2845, + "step": 6284 + }, + { + "epoch": 0.1938700361289228, + "grad_norm": 0.16253407299518585, + "learning_rate": 4.866993426327602e-05, + "loss": 3.2752, + "step": 6285 + }, + { + "epoch": 0.19390088259449623, + "grad_norm": 0.1663486659526825, + "learning_rate": 4.866906778069547e-05, + "loss": 3.2611, + "step": 6286 + }, + { + "epoch": 0.19393172906006964, + "grad_norm": 0.16764219105243683, + "learning_rate": 4.86682010236863e-05, + "loss": 3.27, + "step": 6287 + }, + { + "epoch": 0.19396257552564306, + "grad_norm": 0.15933097898960114, + "learning_rate": 4.8667333992258567e-05, + "loss": 3.2629, + "step": 6288 + }, + { + "epoch": 0.19399342199121647, + "grad_norm": 0.16463248431682587, + "learning_rate": 4.866646668642232e-05, + "loss": 3.2274, + "step": 6289 + }, + { + "epoch": 0.19402426845678988, + "grad_norm": 0.15903915464878082, + "learning_rate": 4.866559910618761e-05, + "loss": 3.2177, + "step": 6290 + }, + { + "epoch": 0.1940551149223633, + "grad_norm": 0.1611415594816208, + "learning_rate": 4.86647312515645e-05, + "loss": 3.1916, + "step": 6291 + }, + { + "epoch": 0.1940859613879367, + "grad_norm": 0.20914439857006073, + "learning_rate": 4.866386312256306e-05, + "loss": 3.2097, + "step": 6292 + }, + { + "epoch": 0.19411680785351013, + "grad_norm": 0.15664710104465485, + "learning_rate": 4.8662994719193336e-05, + "loss": 3.2452, + "step": 6293 + }, + { + "epoch": 0.19414765431908354, + "grad_norm": 0.15948699414730072, + "learning_rate": 4.8662126041465414e-05, + "loss": 3.2564, + "step": 6294 + }, + { + "epoch": 0.19417850078465695, + "grad_norm": 0.1578395813703537, + "learning_rate": 4.8661257089389356e-05, + "loss": 3.2292, + "step": 6295 + }, + { + "epoch": 0.1942093472502304, + "grad_norm": 0.15557460486888885, + "learning_rate": 4.866038786297524e-05, + "loss": 3.2221, + "step": 6296 + }, + { + "epoch": 0.1942401937158038, + "grad_norm": 0.15756040811538696, + "learning_rate": 4.8659518362233145e-05, + "loss": 3.2212, + "step": 6297 + }, + { + "epoch": 0.19427104018137722, + "grad_norm": 0.1598091572523117, + "learning_rate": 4.865864858717316e-05, + "loss": 3.2448, + "step": 6298 + }, + { + "epoch": 0.19430188664695064, + "grad_norm": 0.15854336321353912, + "learning_rate": 4.8657778537805355e-05, + "loss": 3.2203, + "step": 6299 + }, + { + "epoch": 0.19433273311252405, + "grad_norm": 0.5970571041107178, + "learning_rate": 4.865690821413983e-05, + "loss": 3.2113, + "step": 6300 + }, + { + "epoch": 0.19436357957809747, + "grad_norm": 0.16230955719947815, + "learning_rate": 4.8656037616186665e-05, + "loss": 3.2116, + "step": 6301 + }, + { + "epoch": 0.19439442604367088, + "grad_norm": 0.16065680980682373, + "learning_rate": 4.865516674395596e-05, + "loss": 3.2483, + "step": 6302 + }, + { + "epoch": 0.1944252725092443, + "grad_norm": 0.15820541977882385, + "learning_rate": 4.865429559745782e-05, + "loss": 3.26, + "step": 6303 + }, + { + "epoch": 0.1944561189748177, + "grad_norm": 0.1538958102464676, + "learning_rate": 4.8653424176702325e-05, + "loss": 3.2348, + "step": 6304 + }, + { + "epoch": 0.19448696544039112, + "grad_norm": 0.1540842354297638, + "learning_rate": 4.86525524816996e-05, + "loss": 3.2258, + "step": 6305 + }, + { + "epoch": 0.19451781190596454, + "grad_norm": 0.165103942155838, + "learning_rate": 4.8651680512459736e-05, + "loss": 3.2786, + "step": 6306 + }, + { + "epoch": 0.19454865837153795, + "grad_norm": 0.159452885389328, + "learning_rate": 4.8650808268992857e-05, + "loss": 3.21, + "step": 6307 + }, + { + "epoch": 0.1945795048371114, + "grad_norm": 0.21691779792308807, + "learning_rate": 4.864993575130906e-05, + "loss": 3.266, + "step": 6308 + }, + { + "epoch": 0.1946103513026848, + "grad_norm": 0.15247730910778046, + "learning_rate": 4.8649062959418476e-05, + "loss": 3.2212, + "step": 6309 + }, + { + "epoch": 0.19464119776825822, + "grad_norm": 0.1617688089609146, + "learning_rate": 4.864818989333121e-05, + "loss": 3.2051, + "step": 6310 + }, + { + "epoch": 0.19467204423383164, + "grad_norm": 0.1621054708957672, + "learning_rate": 4.86473165530574e-05, + "loss": 3.2056, + "step": 6311 + }, + { + "epoch": 0.19470289069940505, + "grad_norm": 0.16406211256980896, + "learning_rate": 4.864644293860716e-05, + "loss": 3.2384, + "step": 6312 + }, + { + "epoch": 0.19473373716497847, + "grad_norm": 0.1578780710697174, + "learning_rate": 4.8645569049990633e-05, + "loss": 3.2319, + "step": 6313 + }, + { + "epoch": 0.19476458363055188, + "grad_norm": 0.16770827770233154, + "learning_rate": 4.8644694887217934e-05, + "loss": 3.2715, + "step": 6314 + }, + { + "epoch": 0.1947954300961253, + "grad_norm": 0.1608401983976364, + "learning_rate": 4.864382045029921e-05, + "loss": 3.2317, + "step": 6315 + }, + { + "epoch": 0.1948262765616987, + "grad_norm": 0.17721830308437347, + "learning_rate": 4.86429457392446e-05, + "loss": 3.2168, + "step": 6316 + }, + { + "epoch": 0.19485712302727212, + "grad_norm": 0.1826719343662262, + "learning_rate": 4.864207075406424e-05, + "loss": 3.2568, + "step": 6317 + }, + { + "epoch": 0.19488796949284554, + "grad_norm": 0.1609838604927063, + "learning_rate": 4.864119549476827e-05, + "loss": 3.2351, + "step": 6318 + }, + { + "epoch": 0.19491881595841895, + "grad_norm": 0.1606369912624359, + "learning_rate": 4.864031996136685e-05, + "loss": 3.22, + "step": 6319 + }, + { + "epoch": 0.1949496624239924, + "grad_norm": 0.171094611287117, + "learning_rate": 4.863944415387013e-05, + "loss": 3.2501, + "step": 6320 + }, + { + "epoch": 0.1949805088895658, + "grad_norm": 0.20012015104293823, + "learning_rate": 4.8638568072288256e-05, + "loss": 3.2218, + "step": 6321 + }, + { + "epoch": 0.19501135535513922, + "grad_norm": 0.1613694280385971, + "learning_rate": 4.863769171663139e-05, + "loss": 3.2229, + "step": 6322 + }, + { + "epoch": 0.19504220182071264, + "grad_norm": 0.16335085034370422, + "learning_rate": 4.863681508690969e-05, + "loss": 3.2411, + "step": 6323 + }, + { + "epoch": 0.19507304828628605, + "grad_norm": 0.16505518555641174, + "learning_rate": 4.863593818313333e-05, + "loss": 3.2747, + "step": 6324 + }, + { + "epoch": 0.19510389475185946, + "grad_norm": 0.15975108742713928, + "learning_rate": 4.8635061005312467e-05, + "loss": 3.2359, + "step": 6325 + }, + { + "epoch": 0.19513474121743288, + "grad_norm": 0.1663978397846222, + "learning_rate": 4.863418355345727e-05, + "loss": 3.2383, + "step": 6326 + }, + { + "epoch": 0.1951655876830063, + "grad_norm": 0.16037334501743317, + "learning_rate": 4.863330582757792e-05, + "loss": 3.2501, + "step": 6327 + }, + { + "epoch": 0.1951964341485797, + "grad_norm": 0.15563009679317474, + "learning_rate": 4.86324278276846e-05, + "loss": 3.2177, + "step": 6328 + }, + { + "epoch": 0.19522728061415312, + "grad_norm": 0.18250449001789093, + "learning_rate": 4.8631549553787475e-05, + "loss": 3.2316, + "step": 6329 + }, + { + "epoch": 0.19525812707972653, + "grad_norm": 0.1537615954875946, + "learning_rate": 4.8630671005896734e-05, + "loss": 3.2113, + "step": 6330 + }, + { + "epoch": 0.19528897354529998, + "grad_norm": 0.16999086737632751, + "learning_rate": 4.862979218402256e-05, + "loss": 3.2473, + "step": 6331 + }, + { + "epoch": 0.1953198200108734, + "grad_norm": 0.16265787184238434, + "learning_rate": 4.862891308817515e-05, + "loss": 3.2508, + "step": 6332 + }, + { + "epoch": 0.1953506664764468, + "grad_norm": 0.1619262397289276, + "learning_rate": 4.8628033718364685e-05, + "loss": 3.2025, + "step": 6333 + }, + { + "epoch": 0.19538151294202022, + "grad_norm": 0.3742295801639557, + "learning_rate": 4.862715407460138e-05, + "loss": 3.2286, + "step": 6334 + }, + { + "epoch": 0.19541235940759363, + "grad_norm": 0.1705268919467926, + "learning_rate": 4.862627415689541e-05, + "loss": 3.2541, + "step": 6335 + }, + { + "epoch": 0.19544320587316705, + "grad_norm": 0.16588322818279266, + "learning_rate": 4.862539396525699e-05, + "loss": 3.2129, + "step": 6336 + }, + { + "epoch": 0.19547405233874046, + "grad_norm": 0.17778748273849487, + "learning_rate": 4.862451349969633e-05, + "loss": 3.24, + "step": 6337 + }, + { + "epoch": 0.19550489880431388, + "grad_norm": 0.16756470501422882, + "learning_rate": 4.8623632760223624e-05, + "loss": 3.2398, + "step": 6338 + }, + { + "epoch": 0.1955357452698873, + "grad_norm": 0.1777217537164688, + "learning_rate": 4.8622751746849106e-05, + "loss": 3.2441, + "step": 6339 + }, + { + "epoch": 0.1955665917354607, + "grad_norm": 0.16737470030784607, + "learning_rate": 4.862187045958296e-05, + "loss": 3.2341, + "step": 6340 + }, + { + "epoch": 0.19559743820103412, + "grad_norm": 0.17728359997272491, + "learning_rate": 4.862098889843544e-05, + "loss": 3.2179, + "step": 6341 + }, + { + "epoch": 0.19562828466660753, + "grad_norm": 0.16514191031455994, + "learning_rate": 4.862010706341674e-05, + "loss": 3.2225, + "step": 6342 + }, + { + "epoch": 0.19565913113218097, + "grad_norm": 0.17432332038879395, + "learning_rate": 4.861922495453709e-05, + "loss": 3.2212, + "step": 6343 + }, + { + "epoch": 0.1956899775977544, + "grad_norm": 0.1640642136335373, + "learning_rate": 4.861834257180672e-05, + "loss": 3.2321, + "step": 6344 + }, + { + "epoch": 0.1957208240633278, + "grad_norm": 0.1606902778148651, + "learning_rate": 4.8617459915235865e-05, + "loss": 3.2378, + "step": 6345 + }, + { + "epoch": 0.19575167052890122, + "grad_norm": 0.16161993145942688, + "learning_rate": 4.861657698483475e-05, + "loss": 3.2004, + "step": 6346 + }, + { + "epoch": 0.19578251699447463, + "grad_norm": 0.15922614932060242, + "learning_rate": 4.8615693780613615e-05, + "loss": 3.1865, + "step": 6347 + }, + { + "epoch": 0.19581336346004805, + "grad_norm": 0.16144372522830963, + "learning_rate": 4.861481030258271e-05, + "loss": 3.2288, + "step": 6348 + }, + { + "epoch": 0.19584420992562146, + "grad_norm": 0.17148558795452118, + "learning_rate": 4.8613926550752264e-05, + "loss": 3.2101, + "step": 6349 + }, + { + "epoch": 0.19587505639119487, + "grad_norm": 0.1573246717453003, + "learning_rate": 4.8613042525132535e-05, + "loss": 3.2104, + "step": 6350 + }, + { + "epoch": 0.1959059028567683, + "grad_norm": 0.16256824135780334, + "learning_rate": 4.861215822573376e-05, + "loss": 3.2177, + "step": 6351 + }, + { + "epoch": 0.1959367493223417, + "grad_norm": 0.1539076566696167, + "learning_rate": 4.86112736525662e-05, + "loss": 3.2153, + "step": 6352 + }, + { + "epoch": 0.19596759578791512, + "grad_norm": 0.1626991182565689, + "learning_rate": 4.861038880564012e-05, + "loss": 3.2201, + "step": 6353 + }, + { + "epoch": 0.19599844225348853, + "grad_norm": 0.16456621885299683, + "learning_rate": 4.860950368496576e-05, + "loss": 3.2489, + "step": 6354 + }, + { + "epoch": 0.19602928871906197, + "grad_norm": 0.1562565565109253, + "learning_rate": 4.8608618290553406e-05, + "loss": 3.2382, + "step": 6355 + }, + { + "epoch": 0.1960601351846354, + "grad_norm": 0.15474748611450195, + "learning_rate": 4.86077326224133e-05, + "loss": 3.2436, + "step": 6356 + }, + { + "epoch": 0.1960909816502088, + "grad_norm": 0.16356772184371948, + "learning_rate": 4.860684668055572e-05, + "loss": 3.2298, + "step": 6357 + }, + { + "epoch": 0.19612182811578222, + "grad_norm": 0.15254366397857666, + "learning_rate": 4.860596046499094e-05, + "loss": 3.2237, + "step": 6358 + }, + { + "epoch": 0.19615267458135563, + "grad_norm": 0.15286009013652802, + "learning_rate": 4.8605073975729235e-05, + "loss": 3.2073, + "step": 6359 + }, + { + "epoch": 0.19618352104692904, + "grad_norm": 0.15897861123085022, + "learning_rate": 4.8604187212780886e-05, + "loss": 3.245, + "step": 6360 + }, + { + "epoch": 0.19621436751250246, + "grad_norm": 0.15341193974018097, + "learning_rate": 4.8603300176156164e-05, + "loss": 3.2404, + "step": 6361 + }, + { + "epoch": 0.19624521397807587, + "grad_norm": 0.16666768491268158, + "learning_rate": 4.860241286586536e-05, + "loss": 3.2408, + "step": 6362 + }, + { + "epoch": 0.1962760604436493, + "grad_norm": 0.15089689195156097, + "learning_rate": 4.860152528191877e-05, + "loss": 3.2304, + "step": 6363 + }, + { + "epoch": 0.1963069069092227, + "grad_norm": 0.1602046936750412, + "learning_rate": 4.8600637424326676e-05, + "loss": 3.2897, + "step": 6364 + }, + { + "epoch": 0.19633775337479611, + "grad_norm": 0.14868196845054626, + "learning_rate": 4.8599749293099375e-05, + "loss": 3.1784, + "step": 6365 + }, + { + "epoch": 0.19636859984036953, + "grad_norm": 0.16238713264465332, + "learning_rate": 4.8598860888247155e-05, + "loss": 3.2393, + "step": 6366 + }, + { + "epoch": 0.19639944630594297, + "grad_norm": 0.15031126141548157, + "learning_rate": 4.859797220978033e-05, + "loss": 3.2636, + "step": 6367 + }, + { + "epoch": 0.19643029277151638, + "grad_norm": 0.1815951019525528, + "learning_rate": 4.8597083257709194e-05, + "loss": 3.2189, + "step": 6368 + }, + { + "epoch": 0.1964611392370898, + "grad_norm": 0.1574449986219406, + "learning_rate": 4.8596194032044054e-05, + "loss": 3.2229, + "step": 6369 + }, + { + "epoch": 0.1964919857026632, + "grad_norm": 0.16446347534656525, + "learning_rate": 4.859530453279523e-05, + "loss": 3.2067, + "step": 6370 + }, + { + "epoch": 0.19652283216823663, + "grad_norm": 0.1562635749578476, + "learning_rate": 4.8594414759973026e-05, + "loss": 3.257, + "step": 6371 + }, + { + "epoch": 0.19655367863381004, + "grad_norm": 0.1662917286157608, + "learning_rate": 4.8593524713587766e-05, + "loss": 3.2127, + "step": 6372 + }, + { + "epoch": 0.19658452509938346, + "grad_norm": 0.15281158685684204, + "learning_rate": 4.859263439364976e-05, + "loss": 3.2553, + "step": 6373 + }, + { + "epoch": 0.19661537156495687, + "grad_norm": 0.15733930468559265, + "learning_rate": 4.859174380016934e-05, + "loss": 3.2447, + "step": 6374 + }, + { + "epoch": 0.19664621803053028, + "grad_norm": 0.15580353140830994, + "learning_rate": 4.859085293315683e-05, + "loss": 3.2463, + "step": 6375 + }, + { + "epoch": 0.1966770644961037, + "grad_norm": 0.16457289457321167, + "learning_rate": 4.858996179262255e-05, + "loss": 3.241, + "step": 6376 + }, + { + "epoch": 0.1967079109616771, + "grad_norm": 0.15739849209785461, + "learning_rate": 4.858907037857684e-05, + "loss": 3.203, + "step": 6377 + }, + { + "epoch": 0.19673875742725053, + "grad_norm": 0.15542960166931152, + "learning_rate": 4.858817869103004e-05, + "loss": 3.2063, + "step": 6378 + }, + { + "epoch": 0.19676960389282397, + "grad_norm": 0.16106288135051727, + "learning_rate": 4.858728672999248e-05, + "loss": 3.2362, + "step": 6379 + }, + { + "epoch": 0.19680045035839738, + "grad_norm": 0.15600630640983582, + "learning_rate": 4.85863944954745e-05, + "loss": 3.2132, + "step": 6380 + }, + { + "epoch": 0.1968312968239708, + "grad_norm": 0.18949009478092194, + "learning_rate": 4.858550198748646e-05, + "loss": 3.2806, + "step": 6381 + }, + { + "epoch": 0.1968621432895442, + "grad_norm": 0.15722382068634033, + "learning_rate": 4.858460920603869e-05, + "loss": 3.2265, + "step": 6382 + }, + { + "epoch": 0.19689298975511763, + "grad_norm": 0.16147850453853607, + "learning_rate": 4.858371615114155e-05, + "loss": 3.2753, + "step": 6383 + }, + { + "epoch": 0.19692383622069104, + "grad_norm": 0.156454399228096, + "learning_rate": 4.858282282280539e-05, + "loss": 3.1969, + "step": 6384 + }, + { + "epoch": 0.19695468268626445, + "grad_norm": 0.15246723592281342, + "learning_rate": 4.8581929221040575e-05, + "loss": 3.2466, + "step": 6385 + }, + { + "epoch": 0.19698552915183787, + "grad_norm": 0.15443414449691772, + "learning_rate": 4.858103534585746e-05, + "loss": 3.2431, + "step": 6386 + }, + { + "epoch": 0.19701637561741128, + "grad_norm": 0.15773402154445648, + "learning_rate": 4.8580141197266416e-05, + "loss": 3.1914, + "step": 6387 + }, + { + "epoch": 0.1970472220829847, + "grad_norm": 0.15590700507164001, + "learning_rate": 4.85792467752778e-05, + "loss": 3.2034, + "step": 6388 + }, + { + "epoch": 0.1970780685485581, + "grad_norm": 0.15366096794605255, + "learning_rate": 4.8578352079901984e-05, + "loss": 3.1818, + "step": 6389 + }, + { + "epoch": 0.19710891501413152, + "grad_norm": 0.157017782330513, + "learning_rate": 4.8577457111149347e-05, + "loss": 3.2028, + "step": 6390 + }, + { + "epoch": 0.19713976147970497, + "grad_norm": 0.17005649209022522, + "learning_rate": 4.857656186903027e-05, + "loss": 3.2253, + "step": 6391 + }, + { + "epoch": 0.19717060794527838, + "grad_norm": 0.151948481798172, + "learning_rate": 4.857566635355511e-05, + "loss": 3.2595, + "step": 6392 + }, + { + "epoch": 0.1972014544108518, + "grad_norm": 0.16534948348999023, + "learning_rate": 4.857477056473429e-05, + "loss": 3.2293, + "step": 6393 + }, + { + "epoch": 0.1972323008764252, + "grad_norm": 0.15714791417121887, + "learning_rate": 4.857387450257815e-05, + "loss": 3.2262, + "step": 6394 + }, + { + "epoch": 0.19726314734199862, + "grad_norm": 0.15879876911640167, + "learning_rate": 4.8572978167097114e-05, + "loss": 3.2103, + "step": 6395 + }, + { + "epoch": 0.19729399380757204, + "grad_norm": 0.16144120693206787, + "learning_rate": 4.857208155830155e-05, + "loss": 3.2259, + "step": 6396 + }, + { + "epoch": 0.19732484027314545, + "grad_norm": 0.16038373112678528, + "learning_rate": 4.8571184676201875e-05, + "loss": 3.235, + "step": 6397 + }, + { + "epoch": 0.19735568673871887, + "grad_norm": 0.1515282839536667, + "learning_rate": 4.8570287520808476e-05, + "loss": 3.1853, + "step": 6398 + }, + { + "epoch": 0.19738653320429228, + "grad_norm": 0.1643432080745697, + "learning_rate": 4.856939009213176e-05, + "loss": 3.2308, + "step": 6399 + }, + { + "epoch": 0.1974173796698657, + "grad_norm": 0.16019132733345032, + "learning_rate": 4.856849239018213e-05, + "loss": 3.2691, + "step": 6400 + }, + { + "epoch": 0.1974482261354391, + "grad_norm": 0.1642768234014511, + "learning_rate": 4.856759441496999e-05, + "loss": 3.239, + "step": 6401 + }, + { + "epoch": 0.19747907260101252, + "grad_norm": 0.16083738207817078, + "learning_rate": 4.856669616650575e-05, + "loss": 3.2194, + "step": 6402 + }, + { + "epoch": 0.19750991906658596, + "grad_norm": 0.16868573427200317, + "learning_rate": 4.856579764479984e-05, + "loss": 3.254, + "step": 6403 + }, + { + "epoch": 0.19754076553215938, + "grad_norm": 0.16946867108345032, + "learning_rate": 4.856489884986267e-05, + "loss": 3.1992, + "step": 6404 + }, + { + "epoch": 0.1975716119977328, + "grad_norm": 0.1643933802843094, + "learning_rate": 4.856399978170465e-05, + "loss": 3.2248, + "step": 6405 + }, + { + "epoch": 0.1976024584633062, + "grad_norm": 0.16780655086040497, + "learning_rate": 4.856310044033622e-05, + "loss": 3.2204, + "step": 6406 + }, + { + "epoch": 0.19763330492887962, + "grad_norm": 0.1588951051235199, + "learning_rate": 4.856220082576779e-05, + "loss": 3.2127, + "step": 6407 + }, + { + "epoch": 0.19766415139445304, + "grad_norm": 0.17033828794956207, + "learning_rate": 4.856130093800981e-05, + "loss": 3.2789, + "step": 6408 + }, + { + "epoch": 0.19769499786002645, + "grad_norm": 0.15991568565368652, + "learning_rate": 4.85604007770727e-05, + "loss": 3.2415, + "step": 6409 + }, + { + "epoch": 0.19772584432559986, + "grad_norm": 0.16313137114048004, + "learning_rate": 4.85595003429669e-05, + "loss": 3.2381, + "step": 6410 + }, + { + "epoch": 0.19775669079117328, + "grad_norm": 0.1656457781791687, + "learning_rate": 4.855859963570285e-05, + "loss": 3.2003, + "step": 6411 + }, + { + "epoch": 0.1977875372567467, + "grad_norm": 0.16440406441688538, + "learning_rate": 4.8557698655291e-05, + "loss": 3.2298, + "step": 6412 + }, + { + "epoch": 0.1978183837223201, + "grad_norm": 0.16313908994197845, + "learning_rate": 4.8556797401741785e-05, + "loss": 3.2743, + "step": 6413 + }, + { + "epoch": 0.19784923018789352, + "grad_norm": 0.17556889355182648, + "learning_rate": 4.8555895875065665e-05, + "loss": 3.2334, + "step": 6414 + }, + { + "epoch": 0.19788007665346696, + "grad_norm": 0.1792946755886078, + "learning_rate": 4.855499407527308e-05, + "loss": 3.2329, + "step": 6415 + }, + { + "epoch": 0.19791092311904038, + "grad_norm": 0.9593802690505981, + "learning_rate": 4.8554092002374506e-05, + "loss": 3.2502, + "step": 6416 + }, + { + "epoch": 0.1979417695846138, + "grad_norm": 0.16303162276744843, + "learning_rate": 4.8553189656380384e-05, + "loss": 3.2419, + "step": 6417 + }, + { + "epoch": 0.1979726160501872, + "grad_norm": 0.18196478486061096, + "learning_rate": 4.855228703730118e-05, + "loss": 3.219, + "step": 6418 + }, + { + "epoch": 0.19800346251576062, + "grad_norm": 0.3215218186378479, + "learning_rate": 4.855138414514736e-05, + "loss": 3.2279, + "step": 6419 + }, + { + "epoch": 0.19803430898133403, + "grad_norm": 0.18470388650894165, + "learning_rate": 4.8550480979929395e-05, + "loss": 3.2323, + "step": 6420 + }, + { + "epoch": 0.19806515544690745, + "grad_norm": 0.1815059781074524, + "learning_rate": 4.854957754165775e-05, + "loss": 3.2723, + "step": 6421 + }, + { + "epoch": 0.19809600191248086, + "grad_norm": 0.16710612177848816, + "learning_rate": 4.854867383034292e-05, + "loss": 3.2081, + "step": 6422 + }, + { + "epoch": 0.19812684837805428, + "grad_norm": 0.1944444328546524, + "learning_rate": 4.854776984599536e-05, + "loss": 3.2512, + "step": 6423 + }, + { + "epoch": 0.1981576948436277, + "grad_norm": 0.17580349743366241, + "learning_rate": 4.8546865588625555e-05, + "loss": 3.2271, + "step": 6424 + }, + { + "epoch": 0.1981885413092011, + "grad_norm": 0.19669866561889648, + "learning_rate": 4.8545961058244e-05, + "loss": 3.2654, + "step": 6425 + }, + { + "epoch": 0.19821938777477452, + "grad_norm": 0.1697361022233963, + "learning_rate": 4.8545056254861165e-05, + "loss": 3.2021, + "step": 6426 + }, + { + "epoch": 0.19825023424034796, + "grad_norm": 0.19285820424556732, + "learning_rate": 4.854415117848756e-05, + "loss": 3.2395, + "step": 6427 + }, + { + "epoch": 0.19828108070592138, + "grad_norm": 0.16701872646808624, + "learning_rate": 4.8543245829133673e-05, + "loss": 3.2516, + "step": 6428 + }, + { + "epoch": 0.1983119271714948, + "grad_norm": 0.18615247309207916, + "learning_rate": 4.854234020680999e-05, + "loss": 3.2332, + "step": 6429 + }, + { + "epoch": 0.1983427736370682, + "grad_norm": 0.40054386854171753, + "learning_rate": 4.854143431152703e-05, + "loss": 3.216, + "step": 6430 + }, + { + "epoch": 0.19837362010264162, + "grad_norm": 0.18075698614120483, + "learning_rate": 4.854052814329528e-05, + "loss": 3.23, + "step": 6431 + }, + { + "epoch": 0.19840446656821503, + "grad_norm": 0.1695702075958252, + "learning_rate": 4.853962170212525e-05, + "loss": 3.2167, + "step": 6432 + }, + { + "epoch": 0.19843531303378845, + "grad_norm": 0.17950765788555145, + "learning_rate": 4.853871498802746e-05, + "loss": 3.2376, + "step": 6433 + }, + { + "epoch": 0.19846615949936186, + "grad_norm": 0.16762639582157135, + "learning_rate": 4.853780800101241e-05, + "loss": 3.2446, + "step": 6434 + }, + { + "epoch": 0.19849700596493527, + "grad_norm": 0.1963077038526535, + "learning_rate": 4.853690074109062e-05, + "loss": 3.2207, + "step": 6435 + }, + { + "epoch": 0.1985278524305087, + "grad_norm": 0.15718235075473785, + "learning_rate": 4.8535993208272606e-05, + "loss": 3.2359, + "step": 6436 + }, + { + "epoch": 0.1985586988960821, + "grad_norm": 0.16830037534236908, + "learning_rate": 4.85350854025689e-05, + "loss": 3.218, + "step": 6437 + }, + { + "epoch": 0.19858954536165552, + "grad_norm": 0.19004017114639282, + "learning_rate": 4.853417732399002e-05, + "loss": 3.2367, + "step": 6438 + }, + { + "epoch": 0.19862039182722896, + "grad_norm": 0.17146702110767365, + "learning_rate": 4.85332689725465e-05, + "loss": 3.221, + "step": 6439 + }, + { + "epoch": 0.19865123829280237, + "grad_norm": 0.1760735958814621, + "learning_rate": 4.853236034824886e-05, + "loss": 3.2326, + "step": 6440 + }, + { + "epoch": 0.1986820847583758, + "grad_norm": 0.18121102452278137, + "learning_rate": 4.853145145110765e-05, + "loss": 3.2235, + "step": 6441 + }, + { + "epoch": 0.1987129312239492, + "grad_norm": 0.16586342453956604, + "learning_rate": 4.8530542281133393e-05, + "loss": 3.2201, + "step": 6442 + }, + { + "epoch": 0.19874377768952262, + "grad_norm": 0.1713094711303711, + "learning_rate": 4.852963283833665e-05, + "loss": 3.2428, + "step": 6443 + }, + { + "epoch": 0.19877462415509603, + "grad_norm": 0.17065949738025665, + "learning_rate": 4.852872312272795e-05, + "loss": 3.2594, + "step": 6444 + }, + { + "epoch": 0.19880547062066944, + "grad_norm": 0.18839627504348755, + "learning_rate": 4.8527813134317845e-05, + "loss": 3.1714, + "step": 6445 + }, + { + "epoch": 0.19883631708624286, + "grad_norm": 0.18016591668128967, + "learning_rate": 4.852690287311688e-05, + "loss": 3.2605, + "step": 6446 + }, + { + "epoch": 0.19886716355181627, + "grad_norm": 0.1755504012107849, + "learning_rate": 4.8525992339135615e-05, + "loss": 3.2432, + "step": 6447 + }, + { + "epoch": 0.1988980100173897, + "grad_norm": 0.16802546381950378, + "learning_rate": 4.852508153238461e-05, + "loss": 3.215, + "step": 6448 + }, + { + "epoch": 0.1989288564829631, + "grad_norm": 0.16885466873645782, + "learning_rate": 4.852417045287442e-05, + "loss": 3.2416, + "step": 6449 + }, + { + "epoch": 0.19895970294853652, + "grad_norm": 0.16882963478565216, + "learning_rate": 4.852325910061561e-05, + "loss": 3.2623, + "step": 6450 + }, + { + "epoch": 0.19899054941410996, + "grad_norm": 0.16470003128051758, + "learning_rate": 4.8522347475618754e-05, + "loss": 3.2423, + "step": 6451 + }, + { + "epoch": 0.19902139587968337, + "grad_norm": 0.15868955850601196, + "learning_rate": 4.85214355778944e-05, + "loss": 3.2415, + "step": 6452 + }, + { + "epoch": 0.19905224234525679, + "grad_norm": 0.16405178606510162, + "learning_rate": 4.8520523407453145e-05, + "loss": 3.2334, + "step": 6453 + }, + { + "epoch": 0.1990830888108302, + "grad_norm": 0.16700199246406555, + "learning_rate": 4.851961096430555e-05, + "loss": 3.2505, + "step": 6454 + }, + { + "epoch": 0.19911393527640361, + "grad_norm": 0.15450480580329895, + "learning_rate": 4.85186982484622e-05, + "loss": 3.2323, + "step": 6455 + }, + { + "epoch": 0.19914478174197703, + "grad_norm": 0.16525062918663025, + "learning_rate": 4.851778525993369e-05, + "loss": 3.2226, + "step": 6456 + }, + { + "epoch": 0.19917562820755044, + "grad_norm": 0.15733012557029724, + "learning_rate": 4.8516871998730575e-05, + "loss": 3.2422, + "step": 6457 + }, + { + "epoch": 0.19920647467312386, + "grad_norm": 0.15703779458999634, + "learning_rate": 4.851595846486347e-05, + "loss": 3.2469, + "step": 6458 + }, + { + "epoch": 0.19923732113869727, + "grad_norm": 0.15462908148765564, + "learning_rate": 4.8515044658342954e-05, + "loss": 3.2419, + "step": 6459 + }, + { + "epoch": 0.19926816760427069, + "grad_norm": 0.16249839961528778, + "learning_rate": 4.851413057917963e-05, + "loss": 3.2567, + "step": 6460 + }, + { + "epoch": 0.1992990140698441, + "grad_norm": 0.16501620411872864, + "learning_rate": 4.851321622738409e-05, + "loss": 3.2765, + "step": 6461 + }, + { + "epoch": 0.1993298605354175, + "grad_norm": 0.16653405129909515, + "learning_rate": 4.851230160296694e-05, + "loss": 3.2371, + "step": 6462 + }, + { + "epoch": 0.19936070700099096, + "grad_norm": 0.15852417051792145, + "learning_rate": 4.851138670593878e-05, + "loss": 3.2104, + "step": 6463 + }, + { + "epoch": 0.19939155346656437, + "grad_norm": 0.16038310527801514, + "learning_rate": 4.851047153631022e-05, + "loss": 3.2037, + "step": 6464 + }, + { + "epoch": 0.19942239993213778, + "grad_norm": 0.15985028445720673, + "learning_rate": 4.850955609409187e-05, + "loss": 3.2172, + "step": 6465 + }, + { + "epoch": 0.1994532463977112, + "grad_norm": 0.16941657662391663, + "learning_rate": 4.850864037929435e-05, + "loss": 3.2349, + "step": 6466 + }, + { + "epoch": 0.1994840928632846, + "grad_norm": 0.15613029897212982, + "learning_rate": 4.850772439192826e-05, + "loss": 3.2568, + "step": 6467 + }, + { + "epoch": 0.19951493932885803, + "grad_norm": 0.18076281249523163, + "learning_rate": 4.850680813200424e-05, + "loss": 3.2336, + "step": 6468 + }, + { + "epoch": 0.19954578579443144, + "grad_norm": 0.16303682327270508, + "learning_rate": 4.85058915995329e-05, + "loss": 3.1919, + "step": 6469 + }, + { + "epoch": 0.19957663226000485, + "grad_norm": 0.17326028645038605, + "learning_rate": 4.850497479452488e-05, + "loss": 3.2279, + "step": 6470 + }, + { + "epoch": 0.19960747872557827, + "grad_norm": 1.0040403604507446, + "learning_rate": 4.85040577169908e-05, + "loss": 3.2789, + "step": 6471 + }, + { + "epoch": 0.19963832519115168, + "grad_norm": 0.20331917703151703, + "learning_rate": 4.8503140366941296e-05, + "loss": 3.253, + "step": 6472 + }, + { + "epoch": 0.1996691716567251, + "grad_norm": 0.1868678331375122, + "learning_rate": 4.8502222744387006e-05, + "loss": 3.2204, + "step": 6473 + }, + { + "epoch": 0.1997000181222985, + "grad_norm": 0.20319108664989471, + "learning_rate": 4.850130484933856e-05, + "loss": 3.2412, + "step": 6474 + }, + { + "epoch": 0.19973086458787195, + "grad_norm": 0.17033179104328156, + "learning_rate": 4.850038668180661e-05, + "loss": 3.2124, + "step": 6475 + }, + { + "epoch": 0.19976171105344537, + "grad_norm": 0.19404080510139465, + "learning_rate": 4.8499468241801795e-05, + "loss": 3.2482, + "step": 6476 + }, + { + "epoch": 0.19979255751901878, + "grad_norm": 0.16843917965888977, + "learning_rate": 4.849854952933477e-05, + "loss": 3.2707, + "step": 6477 + }, + { + "epoch": 0.1998234039845922, + "grad_norm": 0.1700013130903244, + "learning_rate": 4.8497630544416187e-05, + "loss": 3.2176, + "step": 6478 + }, + { + "epoch": 0.1998542504501656, + "grad_norm": 0.1702960878610611, + "learning_rate": 4.849671128705669e-05, + "loss": 3.2334, + "step": 6479 + }, + { + "epoch": 0.19988509691573902, + "grad_norm": 0.17561054229736328, + "learning_rate": 4.849579175726695e-05, + "loss": 3.2605, + "step": 6480 + }, + { + "epoch": 0.19991594338131244, + "grad_norm": 0.17086827754974365, + "learning_rate": 4.849487195505762e-05, + "loss": 3.2215, + "step": 6481 + }, + { + "epoch": 0.19994678984688585, + "grad_norm": 0.17181213200092316, + "learning_rate": 4.8493951880439384e-05, + "loss": 3.2108, + "step": 6482 + }, + { + "epoch": 0.19997763631245927, + "grad_norm": 0.16571970283985138, + "learning_rate": 4.849303153342288e-05, + "loss": 3.216, + "step": 6483 + }, + { + "epoch": 0.20000848277803268, + "grad_norm": 0.16418464481830597, + "learning_rate": 4.8492110914018796e-05, + "loss": 3.2668, + "step": 6484 + }, + { + "epoch": 0.2000393292436061, + "grad_norm": 0.16187334060668945, + "learning_rate": 4.84911900222378e-05, + "loss": 3.2253, + "step": 6485 + }, + { + "epoch": 0.2000701757091795, + "grad_norm": 0.1663745641708374, + "learning_rate": 4.8490268858090574e-05, + "loss": 3.2327, + "step": 6486 + }, + { + "epoch": 0.20010102217475295, + "grad_norm": 0.16570883989334106, + "learning_rate": 4.8489347421587805e-05, + "loss": 3.2357, + "step": 6487 + }, + { + "epoch": 0.20013186864032637, + "grad_norm": 0.15694017708301544, + "learning_rate": 4.848842571274016e-05, + "loss": 3.2155, + "step": 6488 + }, + { + "epoch": 0.20016271510589978, + "grad_norm": 0.15333524346351624, + "learning_rate": 4.8487503731558336e-05, + "loss": 3.2229, + "step": 6489 + }, + { + "epoch": 0.2001935615714732, + "grad_norm": 0.1568560153245926, + "learning_rate": 4.8486581478053014e-05, + "loss": 3.2248, + "step": 6490 + }, + { + "epoch": 0.2002244080370466, + "grad_norm": 0.46121394634246826, + "learning_rate": 4.84856589522349e-05, + "loss": 3.214, + "step": 6491 + }, + { + "epoch": 0.20025525450262002, + "grad_norm": 0.17758961021900177, + "learning_rate": 4.848473615411468e-05, + "loss": 3.2256, + "step": 6492 + }, + { + "epoch": 0.20028610096819344, + "grad_norm": 0.1730838268995285, + "learning_rate": 4.8483813083703064e-05, + "loss": 3.2416, + "step": 6493 + }, + { + "epoch": 0.20031694743376685, + "grad_norm": 0.15601786971092224, + "learning_rate": 4.848288974101075e-05, + "loss": 3.228, + "step": 6494 + }, + { + "epoch": 0.20034779389934027, + "grad_norm": 0.17401641607284546, + "learning_rate": 4.848196612604843e-05, + "loss": 3.2628, + "step": 6495 + }, + { + "epoch": 0.20037864036491368, + "grad_norm": 0.15621456503868103, + "learning_rate": 4.848104223882682e-05, + "loss": 3.2019, + "step": 6496 + }, + { + "epoch": 0.2004094868304871, + "grad_norm": 0.16250933706760406, + "learning_rate": 4.848011807935665e-05, + "loss": 3.2151, + "step": 6497 + }, + { + "epoch": 0.2004403332960605, + "grad_norm": 0.15945222973823547, + "learning_rate": 4.847919364764861e-05, + "loss": 3.2121, + "step": 6498 + }, + { + "epoch": 0.20047117976163395, + "grad_norm": 0.16292144358158112, + "learning_rate": 4.847826894371344e-05, + "loss": 3.2203, + "step": 6499 + }, + { + "epoch": 0.20050202622720736, + "grad_norm": 0.16230152547359467, + "learning_rate": 4.847734396756185e-05, + "loss": 3.2189, + "step": 6500 + }, + { + "epoch": 0.20053287269278078, + "grad_norm": 0.16609640419483185, + "learning_rate": 4.8476418719204554e-05, + "loss": 3.2264, + "step": 6501 + }, + { + "epoch": 0.2005637191583542, + "grad_norm": 0.16403299570083618, + "learning_rate": 4.8475493198652294e-05, + "loss": 3.2477, + "step": 6502 + }, + { + "epoch": 0.2005945656239276, + "grad_norm": 0.20002563297748566, + "learning_rate": 4.84745674059158e-05, + "loss": 3.2111, + "step": 6503 + }, + { + "epoch": 0.20062541208950102, + "grad_norm": 0.15794682502746582, + "learning_rate": 4.84736413410058e-05, + "loss": 3.2438, + "step": 6504 + }, + { + "epoch": 0.20065625855507443, + "grad_norm": 0.1721830517053604, + "learning_rate": 4.847271500393304e-05, + "loss": 3.203, + "step": 6505 + }, + { + "epoch": 0.20068710502064785, + "grad_norm": 0.1595253199338913, + "learning_rate": 4.847178839470825e-05, + "loss": 3.2398, + "step": 6506 + }, + { + "epoch": 0.20071795148622126, + "grad_norm": 0.16656529903411865, + "learning_rate": 4.847086151334218e-05, + "loss": 3.2191, + "step": 6507 + }, + { + "epoch": 0.20074879795179468, + "grad_norm": 0.1531352698802948, + "learning_rate": 4.8469934359845584e-05, + "loss": 3.1868, + "step": 6508 + }, + { + "epoch": 0.2007796444173681, + "grad_norm": 0.15947528183460236, + "learning_rate": 4.846900693422919e-05, + "loss": 3.2209, + "step": 6509 + }, + { + "epoch": 0.2008104908829415, + "grad_norm": 0.16022346913814545, + "learning_rate": 4.846807923650377e-05, + "loss": 3.2095, + "step": 6510 + }, + { + "epoch": 0.20084133734851495, + "grad_norm": 0.18861690163612366, + "learning_rate": 4.8467151266680065e-05, + "loss": 3.2091, + "step": 6511 + }, + { + "epoch": 0.20087218381408836, + "grad_norm": 0.15976861119270325, + "learning_rate": 4.846622302476885e-05, + "loss": 3.226, + "step": 6512 + }, + { + "epoch": 0.20090303027966178, + "grad_norm": 0.1660216748714447, + "learning_rate": 4.846529451078089e-05, + "loss": 3.2369, + "step": 6513 + }, + { + "epoch": 0.2009338767452352, + "grad_norm": 0.3099530041217804, + "learning_rate": 4.8464365724726927e-05, + "loss": 3.2539, + "step": 6514 + }, + { + "epoch": 0.2009647232108086, + "grad_norm": 0.16463243961334229, + "learning_rate": 4.846343666661775e-05, + "loss": 3.223, + "step": 6515 + }, + { + "epoch": 0.20099556967638202, + "grad_norm": 0.1630508005619049, + "learning_rate": 4.8462507336464124e-05, + "loss": 3.2287, + "step": 6516 + }, + { + "epoch": 0.20102641614195543, + "grad_norm": 0.16960623860359192, + "learning_rate": 4.846157773427683e-05, + "loss": 3.2355, + "step": 6517 + }, + { + "epoch": 0.20105726260752885, + "grad_norm": 0.16360020637512207, + "learning_rate": 4.846064786006663e-05, + "loss": 3.2169, + "step": 6518 + }, + { + "epoch": 0.20108810907310226, + "grad_norm": 0.18138203024864197, + "learning_rate": 4.845971771384432e-05, + "loss": 3.2098, + "step": 6519 + }, + { + "epoch": 0.20111895553867568, + "grad_norm": 0.16586343944072723, + "learning_rate": 4.845878729562069e-05, + "loss": 3.2042, + "step": 6520 + }, + { + "epoch": 0.2011498020042491, + "grad_norm": 0.15084338188171387, + "learning_rate": 4.84578566054065e-05, + "loss": 3.2003, + "step": 6521 + }, + { + "epoch": 0.2011806484698225, + "grad_norm": 0.16542640328407288, + "learning_rate": 4.845692564321258e-05, + "loss": 3.247, + "step": 6522 + }, + { + "epoch": 0.20121149493539595, + "grad_norm": 0.14987045526504517, + "learning_rate": 4.8455994409049695e-05, + "loss": 3.2171, + "step": 6523 + }, + { + "epoch": 0.20124234140096936, + "grad_norm": 0.15958529710769653, + "learning_rate": 4.845506290292865e-05, + "loss": 3.24, + "step": 6524 + }, + { + "epoch": 0.20127318786654277, + "grad_norm": 0.16024647653102875, + "learning_rate": 4.8454131124860245e-05, + "loss": 3.1975, + "step": 6525 + }, + { + "epoch": 0.2013040343321162, + "grad_norm": 0.15081866085529327, + "learning_rate": 4.845319907485528e-05, + "loss": 3.1895, + "step": 6526 + }, + { + "epoch": 0.2013348807976896, + "grad_norm": 0.1636318862438202, + "learning_rate": 4.845226675292457e-05, + "loss": 3.2348, + "step": 6527 + }, + { + "epoch": 0.20136572726326302, + "grad_norm": 0.15266580879688263, + "learning_rate": 4.8451334159078914e-05, + "loss": 3.236, + "step": 6528 + }, + { + "epoch": 0.20139657372883643, + "grad_norm": 0.16640044748783112, + "learning_rate": 4.845040129332914e-05, + "loss": 3.2105, + "step": 6529 + }, + { + "epoch": 0.20142742019440985, + "grad_norm": 0.16617582738399506, + "learning_rate": 4.844946815568605e-05, + "loss": 3.2566, + "step": 6530 + }, + { + "epoch": 0.20145826665998326, + "grad_norm": 0.16860876977443695, + "learning_rate": 4.8448534746160466e-05, + "loss": 3.2227, + "step": 6531 + }, + { + "epoch": 0.20148911312555667, + "grad_norm": 0.17527711391448975, + "learning_rate": 4.8447601064763214e-05, + "loss": 3.2376, + "step": 6532 + }, + { + "epoch": 0.2015199595911301, + "grad_norm": 0.15665467083454132, + "learning_rate": 4.8446667111505126e-05, + "loss": 3.2627, + "step": 6533 + }, + { + "epoch": 0.2015508060567035, + "grad_norm": 0.16094139218330383, + "learning_rate": 4.844573288639701e-05, + "loss": 3.2356, + "step": 6534 + }, + { + "epoch": 0.20158165252227694, + "grad_norm": 0.15514014661312103, + "learning_rate": 4.844479838944972e-05, + "loss": 3.2396, + "step": 6535 + }, + { + "epoch": 0.20161249898785036, + "grad_norm": 0.16961172223091125, + "learning_rate": 4.844386362067408e-05, + "loss": 3.199, + "step": 6536 + }, + { + "epoch": 0.20164334545342377, + "grad_norm": 0.16028113663196564, + "learning_rate": 4.8442928580080925e-05, + "loss": 3.2595, + "step": 6537 + }, + { + "epoch": 0.2016741919189972, + "grad_norm": 0.15364614129066467, + "learning_rate": 4.844199326768111e-05, + "loss": 3.2135, + "step": 6538 + }, + { + "epoch": 0.2017050383845706, + "grad_norm": 0.1714380979537964, + "learning_rate": 4.8441057683485456e-05, + "loss": 3.1864, + "step": 6539 + }, + { + "epoch": 0.20173588485014401, + "grad_norm": 0.15909691154956818, + "learning_rate": 4.8440121827504834e-05, + "loss": 3.2153, + "step": 6540 + }, + { + "epoch": 0.20176673131571743, + "grad_norm": 0.15867818892002106, + "learning_rate": 4.843918569975009e-05, + "loss": 3.249, + "step": 6541 + }, + { + "epoch": 0.20179757778129084, + "grad_norm": 0.1648842841386795, + "learning_rate": 4.8438249300232055e-05, + "loss": 3.2128, + "step": 6542 + }, + { + "epoch": 0.20182842424686426, + "grad_norm": 0.16505882143974304, + "learning_rate": 4.8437312628961615e-05, + "loss": 3.1777, + "step": 6543 + }, + { + "epoch": 0.20185927071243767, + "grad_norm": 0.16140252351760864, + "learning_rate": 4.843637568594962e-05, + "loss": 3.2511, + "step": 6544 + }, + { + "epoch": 0.20189011717801109, + "grad_norm": 0.16834039986133575, + "learning_rate": 4.843543847120694e-05, + "loss": 3.2589, + "step": 6545 + }, + { + "epoch": 0.2019209636435845, + "grad_norm": 0.1629181206226349, + "learning_rate": 4.843450098474442e-05, + "loss": 3.2196, + "step": 6546 + }, + { + "epoch": 0.20195181010915794, + "grad_norm": 0.16220884025096893, + "learning_rate": 4.8433563226572954e-05, + "loss": 3.2087, + "step": 6547 + }, + { + "epoch": 0.20198265657473136, + "grad_norm": 0.15391497313976288, + "learning_rate": 4.8432625196703395e-05, + "loss": 3.2747, + "step": 6548 + }, + { + "epoch": 0.20201350304030477, + "grad_norm": 0.162793830037117, + "learning_rate": 4.8431686895146635e-05, + "loss": 3.2277, + "step": 6549 + }, + { + "epoch": 0.20204434950587818, + "grad_norm": 0.452547162771225, + "learning_rate": 4.8430748321913545e-05, + "loss": 3.2563, + "step": 6550 + }, + { + "epoch": 0.2020751959714516, + "grad_norm": 0.16646701097488403, + "learning_rate": 4.8429809477015006e-05, + "loss": 3.2321, + "step": 6551 + }, + { + "epoch": 0.202106042437025, + "grad_norm": 0.15204904973506927, + "learning_rate": 4.842887036046191e-05, + "loss": 3.2423, + "step": 6552 + }, + { + "epoch": 0.20213688890259843, + "grad_norm": 0.15759263932704926, + "learning_rate": 4.842793097226514e-05, + "loss": 3.2333, + "step": 6553 + }, + { + "epoch": 0.20216773536817184, + "grad_norm": 0.6423621773719788, + "learning_rate": 4.842699131243559e-05, + "loss": 3.2435, + "step": 6554 + }, + { + "epoch": 0.20219858183374526, + "grad_norm": 0.16394084692001343, + "learning_rate": 4.842605138098415e-05, + "loss": 3.2232, + "step": 6555 + }, + { + "epoch": 0.20222942829931867, + "grad_norm": 0.166731595993042, + "learning_rate": 4.8425111177921725e-05, + "loss": 3.2249, + "step": 6556 + }, + { + "epoch": 0.20226027476489208, + "grad_norm": 0.16669856011867523, + "learning_rate": 4.842417070325921e-05, + "loss": 3.2332, + "step": 6557 + }, + { + "epoch": 0.2022911212304655, + "grad_norm": 0.16745048761367798, + "learning_rate": 4.842322995700751e-05, + "loss": 3.2138, + "step": 6558 + }, + { + "epoch": 0.20232196769603894, + "grad_norm": 0.1659344881772995, + "learning_rate": 4.842228893917754e-05, + "loss": 3.1934, + "step": 6559 + }, + { + "epoch": 0.20235281416161235, + "grad_norm": 0.18391898274421692, + "learning_rate": 4.84213476497802e-05, + "loss": 3.2312, + "step": 6560 + }, + { + "epoch": 0.20238366062718577, + "grad_norm": 0.15847095847129822, + "learning_rate": 4.842040608882641e-05, + "loss": 3.1914, + "step": 6561 + }, + { + "epoch": 0.20241450709275918, + "grad_norm": 0.28790420293807983, + "learning_rate": 4.841946425632709e-05, + "loss": 3.235, + "step": 6562 + }, + { + "epoch": 0.2024453535583326, + "grad_norm": 0.16514962911605835, + "learning_rate": 4.8418522152293154e-05, + "loss": 3.2096, + "step": 6563 + }, + { + "epoch": 0.202476200023906, + "grad_norm": 0.20657339692115784, + "learning_rate": 4.841757977673552e-05, + "loss": 3.2257, + "step": 6564 + }, + { + "epoch": 0.20250704648947943, + "grad_norm": 0.1709422469139099, + "learning_rate": 4.841663712966513e-05, + "loss": 3.2381, + "step": 6565 + }, + { + "epoch": 0.20253789295505284, + "grad_norm": 0.15806391835212708, + "learning_rate": 4.84156942110929e-05, + "loss": 3.2183, + "step": 6566 + }, + { + "epoch": 0.20256873942062625, + "grad_norm": 0.18050739169120789, + "learning_rate": 4.841475102102976e-05, + "loss": 3.2522, + "step": 6567 + }, + { + "epoch": 0.20259958588619967, + "grad_norm": 0.1631564497947693, + "learning_rate": 4.841380755948666e-05, + "loss": 3.2559, + "step": 6568 + }, + { + "epoch": 0.20263043235177308, + "grad_norm": 0.15375621616840363, + "learning_rate": 4.841286382647453e-05, + "loss": 3.1974, + "step": 6569 + }, + { + "epoch": 0.2026612788173465, + "grad_norm": 0.16101978719234467, + "learning_rate": 4.841191982200431e-05, + "loss": 3.2575, + "step": 6570 + }, + { + "epoch": 0.20269212528291994, + "grad_norm": 0.15489932894706726, + "learning_rate": 4.841097554608696e-05, + "loss": 3.2152, + "step": 6571 + }, + { + "epoch": 0.20272297174849335, + "grad_norm": 0.17813114821910858, + "learning_rate": 4.8410030998733405e-05, + "loss": 3.264, + "step": 6572 + }, + { + "epoch": 0.20275381821406677, + "grad_norm": 0.1602289378643036, + "learning_rate": 4.8409086179954615e-05, + "loss": 3.2124, + "step": 6573 + }, + { + "epoch": 0.20278466467964018, + "grad_norm": 0.17092503607273102, + "learning_rate": 4.840814108976154e-05, + "loss": 3.2622, + "step": 6574 + }, + { + "epoch": 0.2028155111452136, + "grad_norm": 0.16183127462863922, + "learning_rate": 4.840719572816513e-05, + "loss": 3.2486, + "step": 6575 + }, + { + "epoch": 0.202846357610787, + "grad_norm": 0.15780340135097504, + "learning_rate": 4.840625009517635e-05, + "loss": 3.2072, + "step": 6576 + }, + { + "epoch": 0.20287720407636042, + "grad_norm": 0.1803041398525238, + "learning_rate": 4.8405304190806165e-05, + "loss": 3.1974, + "step": 6577 + }, + { + "epoch": 0.20290805054193384, + "grad_norm": 0.16249951720237732, + "learning_rate": 4.840435801506556e-05, + "loss": 3.2348, + "step": 6578 + }, + { + "epoch": 0.20293889700750725, + "grad_norm": 0.1571265012025833, + "learning_rate": 4.840341156796546e-05, + "loss": 3.2177, + "step": 6579 + }, + { + "epoch": 0.20296974347308067, + "grad_norm": 0.1525992602109909, + "learning_rate": 4.8402464849516886e-05, + "loss": 3.2265, + "step": 6580 + }, + { + "epoch": 0.20300058993865408, + "grad_norm": 0.15929733216762543, + "learning_rate": 4.840151785973079e-05, + "loss": 3.2403, + "step": 6581 + }, + { + "epoch": 0.2030314364042275, + "grad_norm": 0.22055421769618988, + "learning_rate": 4.840057059861816e-05, + "loss": 3.231, + "step": 6582 + }, + { + "epoch": 0.20306228286980094, + "grad_norm": 0.16424502432346344, + "learning_rate": 4.839962306618997e-05, + "loss": 3.2497, + "step": 6583 + }, + { + "epoch": 0.20309312933537435, + "grad_norm": 0.15415465831756592, + "learning_rate": 4.839867526245722e-05, + "loss": 3.2517, + "step": 6584 + }, + { + "epoch": 0.20312397580094776, + "grad_norm": 0.16488446295261383, + "learning_rate": 4.839772718743089e-05, + "loss": 3.2162, + "step": 6585 + }, + { + "epoch": 0.20315482226652118, + "grad_norm": 0.15682287514209747, + "learning_rate": 4.839677884112196e-05, + "loss": 3.2472, + "step": 6586 + }, + { + "epoch": 0.2031856687320946, + "grad_norm": 0.17241019010543823, + "learning_rate": 4.839583022354146e-05, + "loss": 3.2037, + "step": 6587 + }, + { + "epoch": 0.203216515197668, + "grad_norm": 0.15831680595874786, + "learning_rate": 4.8394881334700354e-05, + "loss": 3.2413, + "step": 6588 + }, + { + "epoch": 0.20324736166324142, + "grad_norm": 0.16300565004348755, + "learning_rate": 4.839393217460966e-05, + "loss": 3.2143, + "step": 6589 + }, + { + "epoch": 0.20327820812881484, + "grad_norm": 0.24737827479839325, + "learning_rate": 4.839298274328038e-05, + "loss": 3.236, + "step": 6590 + }, + { + "epoch": 0.20330905459438825, + "grad_norm": 0.1667342483997345, + "learning_rate": 4.8392033040723525e-05, + "loss": 3.2457, + "step": 6591 + }, + { + "epoch": 0.20333990105996166, + "grad_norm": 0.1836925446987152, + "learning_rate": 4.8391083066950106e-05, + "loss": 3.2291, + "step": 6592 + }, + { + "epoch": 0.20337074752553508, + "grad_norm": 0.15960946679115295, + "learning_rate": 4.8390132821971137e-05, + "loss": 3.22, + "step": 6593 + }, + { + "epoch": 0.2034015939911085, + "grad_norm": 0.1717323511838913, + "learning_rate": 4.838918230579763e-05, + "loss": 3.2943, + "step": 6594 + }, + { + "epoch": 0.20343244045668193, + "grad_norm": 0.1606885939836502, + "learning_rate": 4.838823151844061e-05, + "loss": 3.194, + "step": 6595 + }, + { + "epoch": 0.20346328692225535, + "grad_norm": 0.17204564809799194, + "learning_rate": 4.8387280459911096e-05, + "loss": 3.2448, + "step": 6596 + }, + { + "epoch": 0.20349413338782876, + "grad_norm": 0.1615964025259018, + "learning_rate": 4.838632913022012e-05, + "loss": 3.235, + "step": 6597 + }, + { + "epoch": 0.20352497985340218, + "grad_norm": 0.1505090594291687, + "learning_rate": 4.8385377529378715e-05, + "loss": 3.2493, + "step": 6598 + }, + { + "epoch": 0.2035558263189756, + "grad_norm": 0.16892923414707184, + "learning_rate": 4.838442565739791e-05, + "loss": 3.2408, + "step": 6599 + }, + { + "epoch": 0.203586672784549, + "grad_norm": 0.16104717552661896, + "learning_rate": 4.8383473514288734e-05, + "loss": 3.2469, + "step": 6600 + }, + { + "epoch": 0.20361751925012242, + "grad_norm": 0.191755011677742, + "learning_rate": 4.838252110006224e-05, + "loss": 3.2311, + "step": 6601 + }, + { + "epoch": 0.20364836571569583, + "grad_norm": 0.1703275740146637, + "learning_rate": 4.838156841472947e-05, + "loss": 3.2421, + "step": 6602 + }, + { + "epoch": 0.20367921218126925, + "grad_norm": 0.15765529870986938, + "learning_rate": 4.8380615458301464e-05, + "loss": 3.2331, + "step": 6603 + }, + { + "epoch": 0.20371005864684266, + "grad_norm": 0.203220933675766, + "learning_rate": 4.837966223078927e-05, + "loss": 3.2376, + "step": 6604 + }, + { + "epoch": 0.20374090511241608, + "grad_norm": 0.15686537325382233, + "learning_rate": 4.8378708732203936e-05, + "loss": 3.2292, + "step": 6605 + }, + { + "epoch": 0.20377175157798952, + "grad_norm": 0.18510331213474274, + "learning_rate": 4.837775496255653e-05, + "loss": 3.2518, + "step": 6606 + }, + { + "epoch": 0.20380259804356293, + "grad_norm": 0.2933347523212433, + "learning_rate": 4.8376800921858104e-05, + "loss": 3.2222, + "step": 6607 + }, + { + "epoch": 0.20383344450913635, + "grad_norm": 0.16884326934814453, + "learning_rate": 4.837584661011972e-05, + "loss": 3.2431, + "step": 6608 + }, + { + "epoch": 0.20386429097470976, + "grad_norm": 0.16122391819953918, + "learning_rate": 4.837489202735244e-05, + "loss": 3.2399, + "step": 6609 + }, + { + "epoch": 0.20389513744028318, + "grad_norm": 0.16556118428707123, + "learning_rate": 4.837393717356733e-05, + "loss": 3.2021, + "step": 6610 + }, + { + "epoch": 0.2039259839058566, + "grad_norm": 0.16014249622821808, + "learning_rate": 4.837298204877547e-05, + "loss": 3.2609, + "step": 6611 + }, + { + "epoch": 0.20395683037143, + "grad_norm": 0.16131503880023956, + "learning_rate": 4.8372026652987926e-05, + "loss": 3.29, + "step": 6612 + }, + { + "epoch": 0.20398767683700342, + "grad_norm": 0.17249755561351776, + "learning_rate": 4.8371070986215775e-05, + "loss": 3.2056, + "step": 6613 + }, + { + "epoch": 0.20401852330257683, + "grad_norm": 0.17031456530094147, + "learning_rate": 4.83701150484701e-05, + "loss": 3.2241, + "step": 6614 + }, + { + "epoch": 0.20404936976815025, + "grad_norm": 0.1513546109199524, + "learning_rate": 4.836915883976199e-05, + "loss": 3.2156, + "step": 6615 + }, + { + "epoch": 0.20408021623372366, + "grad_norm": 0.16940781474113464, + "learning_rate": 4.8368202360102526e-05, + "loss": 3.2246, + "step": 6616 + }, + { + "epoch": 0.20411106269929707, + "grad_norm": 0.15418459475040436, + "learning_rate": 4.83672456095028e-05, + "loss": 3.2349, + "step": 6617 + }, + { + "epoch": 0.20414190916487052, + "grad_norm": 0.16928336024284363, + "learning_rate": 4.8366288587973896e-05, + "loss": 3.254, + "step": 6618 + }, + { + "epoch": 0.20417275563044393, + "grad_norm": 0.1503811478614807, + "learning_rate": 4.836533129552692e-05, + "loss": 3.2031, + "step": 6619 + }, + { + "epoch": 0.20420360209601734, + "grad_norm": 0.1510644257068634, + "learning_rate": 4.8364373732172976e-05, + "loss": 3.2116, + "step": 6620 + }, + { + "epoch": 0.20423444856159076, + "grad_norm": 0.16332662105560303, + "learning_rate": 4.836341589792315e-05, + "loss": 3.2434, + "step": 6621 + }, + { + "epoch": 0.20426529502716417, + "grad_norm": 0.15938878059387207, + "learning_rate": 4.8362457792788556e-05, + "loss": 3.2252, + "step": 6622 + }, + { + "epoch": 0.2042961414927376, + "grad_norm": 0.15768863260746002, + "learning_rate": 4.836149941678031e-05, + "loss": 3.2121, + "step": 6623 + }, + { + "epoch": 0.204326987958311, + "grad_norm": 0.16147926449775696, + "learning_rate": 4.8360540769909515e-05, + "loss": 3.2125, + "step": 6624 + }, + { + "epoch": 0.20435783442388442, + "grad_norm": 0.15517571568489075, + "learning_rate": 4.835958185218728e-05, + "loss": 3.1925, + "step": 6625 + }, + { + "epoch": 0.20438868088945783, + "grad_norm": 0.16673888266086578, + "learning_rate": 4.835862266362474e-05, + "loss": 3.2388, + "step": 6626 + }, + { + "epoch": 0.20441952735503124, + "grad_norm": 0.16950243711471558, + "learning_rate": 4.8357663204233005e-05, + "loss": 3.261, + "step": 6627 + }, + { + "epoch": 0.20445037382060466, + "grad_norm": 0.14973850548267365, + "learning_rate": 4.8356703474023204e-05, + "loss": 3.2068, + "step": 6628 + }, + { + "epoch": 0.20448122028617807, + "grad_norm": 0.1640504151582718, + "learning_rate": 4.835574347300645e-05, + "loss": 3.2484, + "step": 6629 + }, + { + "epoch": 0.20451206675175151, + "grad_norm": 0.17433059215545654, + "learning_rate": 4.83547832011939e-05, + "loss": 3.2475, + "step": 6630 + }, + { + "epoch": 0.20454291321732493, + "grad_norm": 0.16270656883716583, + "learning_rate": 4.835382265859666e-05, + "loss": 3.2671, + "step": 6631 + }, + { + "epoch": 0.20457375968289834, + "grad_norm": 0.16464757919311523, + "learning_rate": 4.8352861845225896e-05, + "loss": 3.2313, + "step": 6632 + }, + { + "epoch": 0.20460460614847176, + "grad_norm": 0.15497590601444244, + "learning_rate": 4.835190076109272e-05, + "loss": 3.224, + "step": 6633 + }, + { + "epoch": 0.20463545261404517, + "grad_norm": 0.1632273644208908, + "learning_rate": 4.8350939406208295e-05, + "loss": 3.247, + "step": 6634 + }, + { + "epoch": 0.20466629907961859, + "grad_norm": 0.16400288045406342, + "learning_rate": 4.8349977780583755e-05, + "loss": 3.2078, + "step": 6635 + }, + { + "epoch": 0.204697145545192, + "grad_norm": 0.15890078246593475, + "learning_rate": 4.834901588423025e-05, + "loss": 3.2239, + "step": 6636 + }, + { + "epoch": 0.2047279920107654, + "grad_norm": 0.16659146547317505, + "learning_rate": 4.834805371715895e-05, + "loss": 3.2011, + "step": 6637 + }, + { + "epoch": 0.20475883847633883, + "grad_norm": 0.16199500858783722, + "learning_rate": 4.834709127938099e-05, + "loss": 3.2572, + "step": 6638 + }, + { + "epoch": 0.20478968494191224, + "grad_norm": 0.16170549392700195, + "learning_rate": 4.8346128570907536e-05, + "loss": 3.2448, + "step": 6639 + }, + { + "epoch": 0.20482053140748566, + "grad_norm": 0.1601780503988266, + "learning_rate": 4.834516559174975e-05, + "loss": 3.216, + "step": 6640 + }, + { + "epoch": 0.20485137787305907, + "grad_norm": 0.15701612830162048, + "learning_rate": 4.83442023419188e-05, + "loss": 3.2362, + "step": 6641 + }, + { + "epoch": 0.2048822243386325, + "grad_norm": 0.15294885635375977, + "learning_rate": 4.834323882142585e-05, + "loss": 3.249, + "step": 6642 + }, + { + "epoch": 0.20491307080420593, + "grad_norm": 0.16175852715969086, + "learning_rate": 4.834227503028208e-05, + "loss": 3.2098, + "step": 6643 + }, + { + "epoch": 0.20494391726977934, + "grad_norm": 0.16371886432170868, + "learning_rate": 4.8341310968498656e-05, + "loss": 3.2502, + "step": 6644 + }, + { + "epoch": 0.20497476373535276, + "grad_norm": 0.17220759391784668, + "learning_rate": 4.834034663608675e-05, + "loss": 3.2141, + "step": 6645 + }, + { + "epoch": 0.20500561020092617, + "grad_norm": 0.15911409258842468, + "learning_rate": 4.833938203305756e-05, + "loss": 3.2503, + "step": 6646 + }, + { + "epoch": 0.20503645666649958, + "grad_norm": 0.16347265243530273, + "learning_rate": 4.833841715942226e-05, + "loss": 3.242, + "step": 6647 + }, + { + "epoch": 0.205067303132073, + "grad_norm": 0.1631903499364853, + "learning_rate": 4.833745201519203e-05, + "loss": 3.2491, + "step": 6648 + }, + { + "epoch": 0.2050981495976464, + "grad_norm": 0.16364380717277527, + "learning_rate": 4.833648660037807e-05, + "loss": 3.2604, + "step": 6649 + }, + { + "epoch": 0.20512899606321983, + "grad_norm": 0.15702077746391296, + "learning_rate": 4.833552091499157e-05, + "loss": 3.2136, + "step": 6650 + }, + { + "epoch": 0.20515984252879324, + "grad_norm": 0.15335828065872192, + "learning_rate": 4.8334554959043734e-05, + "loss": 3.2123, + "step": 6651 + }, + { + "epoch": 0.20519068899436665, + "grad_norm": 0.15967117249965668, + "learning_rate": 4.833358873254575e-05, + "loss": 3.2259, + "step": 6652 + }, + { + "epoch": 0.20522153545994007, + "grad_norm": 0.15382008254528046, + "learning_rate": 4.8332622235508836e-05, + "loss": 3.2426, + "step": 6653 + }, + { + "epoch": 0.2052523819255135, + "grad_norm": 0.24016468226909637, + "learning_rate": 4.833165546794418e-05, + "loss": 3.229, + "step": 6654 + }, + { + "epoch": 0.20528322839108692, + "grad_norm": 0.16108158230781555, + "learning_rate": 4.8330688429863e-05, + "loss": 3.1909, + "step": 6655 + }, + { + "epoch": 0.20531407485666034, + "grad_norm": 0.15487277507781982, + "learning_rate": 4.8329721121276514e-05, + "loss": 3.2007, + "step": 6656 + }, + { + "epoch": 0.20534492132223375, + "grad_norm": 0.1680505871772766, + "learning_rate": 4.832875354219593e-05, + "loss": 3.2549, + "step": 6657 + }, + { + "epoch": 0.20537576778780717, + "grad_norm": 0.15876485407352448, + "learning_rate": 4.832778569263246e-05, + "loss": 3.209, + "step": 6658 + }, + { + "epoch": 0.20540661425338058, + "grad_norm": 0.16081808507442474, + "learning_rate": 4.832681757259734e-05, + "loss": 3.2358, + "step": 6659 + }, + { + "epoch": 0.205437460718954, + "grad_norm": 0.16664451360702515, + "learning_rate": 4.832584918210178e-05, + "loss": 3.227, + "step": 6660 + }, + { + "epoch": 0.2054683071845274, + "grad_norm": 0.23149517178535461, + "learning_rate": 4.8324880521157025e-05, + "loss": 3.2103, + "step": 6661 + }, + { + "epoch": 0.20549915365010082, + "grad_norm": 0.16572809219360352, + "learning_rate": 4.832391158977429e-05, + "loss": 3.1999, + "step": 6662 + }, + { + "epoch": 0.20553000011567424, + "grad_norm": 0.16459478437900543, + "learning_rate": 4.832294238796482e-05, + "loss": 3.2382, + "step": 6663 + }, + { + "epoch": 0.20556084658124765, + "grad_norm": 0.1601380556821823, + "learning_rate": 4.8321972915739855e-05, + "loss": 3.2001, + "step": 6664 + }, + { + "epoch": 0.20559169304682107, + "grad_norm": 0.1652655005455017, + "learning_rate": 4.832100317311062e-05, + "loss": 3.2556, + "step": 6665 + }, + { + "epoch": 0.2056225395123945, + "grad_norm": 0.16145732998847961, + "learning_rate": 4.832003316008837e-05, + "loss": 3.215, + "step": 6666 + }, + { + "epoch": 0.20565338597796792, + "grad_norm": 0.16371500492095947, + "learning_rate": 4.831906287668435e-05, + "loss": 3.2195, + "step": 6667 + }, + { + "epoch": 0.20568423244354134, + "grad_norm": 0.1616748869419098, + "learning_rate": 4.83180923229098e-05, + "loss": 3.2055, + "step": 6668 + }, + { + "epoch": 0.20571507890911475, + "grad_norm": 0.1783760040998459, + "learning_rate": 4.8317121498776e-05, + "loss": 3.2595, + "step": 6669 + }, + { + "epoch": 0.20574592537468817, + "grad_norm": 0.15294957160949707, + "learning_rate": 4.831615040429417e-05, + "loss": 3.2152, + "step": 6670 + }, + { + "epoch": 0.20577677184026158, + "grad_norm": 0.16874445974826813, + "learning_rate": 4.83151790394756e-05, + "loss": 3.2758, + "step": 6671 + }, + { + "epoch": 0.205807618305835, + "grad_norm": 0.15704818069934845, + "learning_rate": 4.8314207404331535e-05, + "loss": 3.2508, + "step": 6672 + }, + { + "epoch": 0.2058384647714084, + "grad_norm": 0.16266773641109467, + "learning_rate": 4.831323549887325e-05, + "loss": 3.2239, + "step": 6673 + }, + { + "epoch": 0.20586931123698182, + "grad_norm": 0.15190376341342926, + "learning_rate": 4.8312263323112e-05, + "loss": 3.2367, + "step": 6674 + }, + { + "epoch": 0.20590015770255524, + "grad_norm": 0.1664484143257141, + "learning_rate": 4.8311290877059065e-05, + "loss": 3.1981, + "step": 6675 + }, + { + "epoch": 0.20593100416812865, + "grad_norm": 0.15734754502773285, + "learning_rate": 4.8310318160725735e-05, + "loss": 3.2556, + "step": 6676 + }, + { + "epoch": 0.20596185063370206, + "grad_norm": 0.16164758801460266, + "learning_rate": 4.830934517412327e-05, + "loss": 3.2212, + "step": 6677 + }, + { + "epoch": 0.2059926970992755, + "grad_norm": 0.15481093525886536, + "learning_rate": 4.830837191726294e-05, + "loss": 3.2585, + "step": 6678 + }, + { + "epoch": 0.20602354356484892, + "grad_norm": 0.16731947660446167, + "learning_rate": 4.830739839015605e-05, + "loss": 3.2374, + "step": 6679 + }, + { + "epoch": 0.20605439003042234, + "grad_norm": 0.16265663504600525, + "learning_rate": 4.830642459281389e-05, + "loss": 3.2428, + "step": 6680 + }, + { + "epoch": 0.20608523649599575, + "grad_norm": 0.15969839692115784, + "learning_rate": 4.830545052524774e-05, + "loss": 3.2313, + "step": 6681 + }, + { + "epoch": 0.20611608296156916, + "grad_norm": 0.16790111362934113, + "learning_rate": 4.8304476187468895e-05, + "loss": 3.242, + "step": 6682 + }, + { + "epoch": 0.20614692942714258, + "grad_norm": 0.16449375450611115, + "learning_rate": 4.830350157948865e-05, + "loss": 3.227, + "step": 6683 + }, + { + "epoch": 0.206177775892716, + "grad_norm": 0.1738618016242981, + "learning_rate": 4.8302526701318315e-05, + "loss": 3.2334, + "step": 6684 + }, + { + "epoch": 0.2062086223582894, + "grad_norm": 0.1566760540008545, + "learning_rate": 4.830155155296918e-05, + "loss": 3.2458, + "step": 6685 + }, + { + "epoch": 0.20623946882386282, + "grad_norm": 0.15901099145412445, + "learning_rate": 4.8300576134452566e-05, + "loss": 3.2062, + "step": 6686 + }, + { + "epoch": 0.20627031528943623, + "grad_norm": 0.1584770232439041, + "learning_rate": 4.8299600445779766e-05, + "loss": 3.2546, + "step": 6687 + }, + { + "epoch": 0.20630116175500965, + "grad_norm": 0.15831436216831207, + "learning_rate": 4.82986244869621e-05, + "loss": 3.2197, + "step": 6688 + }, + { + "epoch": 0.20633200822058306, + "grad_norm": 0.15972130000591278, + "learning_rate": 4.8297648258010885e-05, + "loss": 3.1993, + "step": 6689 + }, + { + "epoch": 0.2063628546861565, + "grad_norm": 0.15415826439857483, + "learning_rate": 4.8296671758937437e-05, + "loss": 3.2267, + "step": 6690 + }, + { + "epoch": 0.20639370115172992, + "grad_norm": 0.1538253277540207, + "learning_rate": 4.829569498975308e-05, + "loss": 3.2162, + "step": 6691 + }, + { + "epoch": 0.20642454761730333, + "grad_norm": 0.15112033486366272, + "learning_rate": 4.8294717950469145e-05, + "loss": 3.2261, + "step": 6692 + }, + { + "epoch": 0.20645539408287675, + "grad_norm": 0.15633296966552734, + "learning_rate": 4.829374064109695e-05, + "loss": 3.1697, + "step": 6693 + }, + { + "epoch": 0.20648624054845016, + "grad_norm": 0.17423486709594727, + "learning_rate": 4.829276306164783e-05, + "loss": 3.2193, + "step": 6694 + }, + { + "epoch": 0.20651708701402358, + "grad_norm": 0.15229275822639465, + "learning_rate": 4.829178521213311e-05, + "loss": 3.2392, + "step": 6695 + }, + { + "epoch": 0.206547933479597, + "grad_norm": 0.16614645719528198, + "learning_rate": 4.829080709256415e-05, + "loss": 3.2117, + "step": 6696 + }, + { + "epoch": 0.2065787799451704, + "grad_norm": 0.161565363407135, + "learning_rate": 4.828982870295227e-05, + "loss": 3.2418, + "step": 6697 + }, + { + "epoch": 0.20660962641074382, + "grad_norm": 0.15730901062488556, + "learning_rate": 4.828885004330883e-05, + "loss": 3.2394, + "step": 6698 + }, + { + "epoch": 0.20664047287631723, + "grad_norm": 0.15993823111057281, + "learning_rate": 4.828787111364517e-05, + "loss": 3.2436, + "step": 6699 + }, + { + "epoch": 0.20667131934189065, + "grad_norm": 0.16129063069820404, + "learning_rate": 4.828689191397263e-05, + "loss": 3.2557, + "step": 6700 + }, + { + "epoch": 0.20670216580746406, + "grad_norm": 0.1509927660226822, + "learning_rate": 4.828591244430257e-05, + "loss": 3.2052, + "step": 6701 + }, + { + "epoch": 0.2067330122730375, + "grad_norm": 0.15459759533405304, + "learning_rate": 4.828493270464636e-05, + "loss": 3.249, + "step": 6702 + }, + { + "epoch": 0.20676385873861092, + "grad_norm": 0.15473082661628723, + "learning_rate": 4.828395269501534e-05, + "loss": 3.2326, + "step": 6703 + }, + { + "epoch": 0.20679470520418433, + "grad_norm": 0.15292058885097504, + "learning_rate": 4.828297241542088e-05, + "loss": 3.2256, + "step": 6704 + }, + { + "epoch": 0.20682555166975775, + "grad_norm": 0.16029092669487, + "learning_rate": 4.8281991865874346e-05, + "loss": 3.2043, + "step": 6705 + }, + { + "epoch": 0.20685639813533116, + "grad_norm": 0.15560252964496613, + "learning_rate": 4.828101104638711e-05, + "loss": 3.2372, + "step": 6706 + }, + { + "epoch": 0.20688724460090457, + "grad_norm": 0.15304628014564514, + "learning_rate": 4.828002995697054e-05, + "loss": 3.2333, + "step": 6707 + }, + { + "epoch": 0.206918091066478, + "grad_norm": 0.1565280705690384, + "learning_rate": 4.827904859763601e-05, + "loss": 3.2164, + "step": 6708 + }, + { + "epoch": 0.2069489375320514, + "grad_norm": 0.15617600083351135, + "learning_rate": 4.827806696839491e-05, + "loss": 3.2349, + "step": 6709 + }, + { + "epoch": 0.20697978399762482, + "grad_norm": 0.15503937005996704, + "learning_rate": 4.82770850692586e-05, + "loss": 3.2352, + "step": 6710 + }, + { + "epoch": 0.20701063046319823, + "grad_norm": 0.15744592249393463, + "learning_rate": 4.8276102900238484e-05, + "loss": 3.2216, + "step": 6711 + }, + { + "epoch": 0.20704147692877164, + "grad_norm": 0.15468443930149078, + "learning_rate": 4.8275120461345944e-05, + "loss": 3.2356, + "step": 6712 + }, + { + "epoch": 0.20707232339434506, + "grad_norm": 0.155927374958992, + "learning_rate": 4.8274137752592365e-05, + "loss": 3.2001, + "step": 6713 + }, + { + "epoch": 0.2071031698599185, + "grad_norm": 0.1508437693119049, + "learning_rate": 4.827315477398914e-05, + "loss": 3.2228, + "step": 6714 + }, + { + "epoch": 0.20713401632549192, + "grad_norm": 0.15470018982887268, + "learning_rate": 4.827217152554768e-05, + "loss": 3.2416, + "step": 6715 + }, + { + "epoch": 0.20716486279106533, + "grad_norm": 0.16480712592601776, + "learning_rate": 4.8271188007279374e-05, + "loss": 3.22, + "step": 6716 + }, + { + "epoch": 0.20719570925663874, + "grad_norm": 0.15424123406410217, + "learning_rate": 4.827020421919563e-05, + "loss": 3.2545, + "step": 6717 + }, + { + "epoch": 0.20722655572221216, + "grad_norm": 0.15449084341526031, + "learning_rate": 4.8269220161307847e-05, + "loss": 3.2209, + "step": 6718 + }, + { + "epoch": 0.20725740218778557, + "grad_norm": 0.1591401845216751, + "learning_rate": 4.826823583362743e-05, + "loss": 3.2369, + "step": 6719 + }, + { + "epoch": 0.207288248653359, + "grad_norm": 0.1479128748178482, + "learning_rate": 4.8267251236165814e-05, + "loss": 3.1976, + "step": 6720 + }, + { + "epoch": 0.2073190951189324, + "grad_norm": 0.15487755835056305, + "learning_rate": 4.82662663689344e-05, + "loss": 3.2407, + "step": 6721 + }, + { + "epoch": 0.20734994158450581, + "grad_norm": 0.1661437302827835, + "learning_rate": 4.826528123194461e-05, + "loss": 3.2023, + "step": 6722 + }, + { + "epoch": 0.20738078805007923, + "grad_norm": 0.1581302285194397, + "learning_rate": 4.826429582520786e-05, + "loss": 3.2049, + "step": 6723 + }, + { + "epoch": 0.20741163451565264, + "grad_norm": 0.1530691683292389, + "learning_rate": 4.826331014873559e-05, + "loss": 3.2212, + "step": 6724 + }, + { + "epoch": 0.20744248098122606, + "grad_norm": 0.16353736817836761, + "learning_rate": 4.8262324202539204e-05, + "loss": 3.2286, + "step": 6725 + }, + { + "epoch": 0.2074733274467995, + "grad_norm": 0.1630285531282425, + "learning_rate": 4.8261337986630156e-05, + "loss": 3.2173, + "step": 6726 + }, + { + "epoch": 0.2075041739123729, + "grad_norm": 0.15705493092536926, + "learning_rate": 4.8260351501019875e-05, + "loss": 3.2255, + "step": 6727 + }, + { + "epoch": 0.20753502037794633, + "grad_norm": 0.163082093000412, + "learning_rate": 4.825936474571979e-05, + "loss": 3.2363, + "step": 6728 + }, + { + "epoch": 0.20756586684351974, + "grad_norm": 0.15316076576709747, + "learning_rate": 4.8258377720741354e-05, + "loss": 3.2227, + "step": 6729 + }, + { + "epoch": 0.20759671330909316, + "grad_norm": 0.16413632035255432, + "learning_rate": 4.8257390426095996e-05, + "loss": 3.2295, + "step": 6730 + }, + { + "epoch": 0.20762755977466657, + "grad_norm": 0.15752343833446503, + "learning_rate": 4.825640286179518e-05, + "loss": 3.2286, + "step": 6731 + }, + { + "epoch": 0.20765840624023998, + "grad_norm": 0.15892119705677032, + "learning_rate": 4.825541502785035e-05, + "loss": 3.1959, + "step": 6732 + }, + { + "epoch": 0.2076892527058134, + "grad_norm": 0.16394920647144318, + "learning_rate": 4.825442692427295e-05, + "loss": 3.214, + "step": 6733 + }, + { + "epoch": 0.2077200991713868, + "grad_norm": 0.16242167353630066, + "learning_rate": 4.825343855107445e-05, + "loss": 3.2135, + "step": 6734 + }, + { + "epoch": 0.20775094563696023, + "grad_norm": 0.16622541844844818, + "learning_rate": 4.82524499082663e-05, + "loss": 3.2441, + "step": 6735 + }, + { + "epoch": 0.20778179210253364, + "grad_norm": 0.1721966713666916, + "learning_rate": 4.8251460995859964e-05, + "loss": 3.2606, + "step": 6736 + }, + { + "epoch": 0.20781263856810706, + "grad_norm": 0.16389617323875427, + "learning_rate": 4.8250471813866914e-05, + "loss": 3.2406, + "step": 6737 + }, + { + "epoch": 0.2078434850336805, + "grad_norm": 0.16146504878997803, + "learning_rate": 4.824948236229862e-05, + "loss": 3.2261, + "step": 6738 + }, + { + "epoch": 0.2078743314992539, + "grad_norm": 0.1744224578142166, + "learning_rate": 4.8248492641166545e-05, + "loss": 3.2333, + "step": 6739 + }, + { + "epoch": 0.20790517796482733, + "grad_norm": 0.15002577006816864, + "learning_rate": 4.824750265048217e-05, + "loss": 3.2107, + "step": 6740 + }, + { + "epoch": 0.20793602443040074, + "grad_norm": 0.17433449625968933, + "learning_rate": 4.824651239025697e-05, + "loss": 3.2399, + "step": 6741 + }, + { + "epoch": 0.20796687089597415, + "grad_norm": 0.15340639650821686, + "learning_rate": 4.8245521860502426e-05, + "loss": 3.2472, + "step": 6742 + }, + { + "epoch": 0.20799771736154757, + "grad_norm": 0.1644487828016281, + "learning_rate": 4.824453106123003e-05, + "loss": 3.2249, + "step": 6743 + }, + { + "epoch": 0.20802856382712098, + "grad_norm": 0.15600162744522095, + "learning_rate": 4.824353999245126e-05, + "loss": 3.2534, + "step": 6744 + }, + { + "epoch": 0.2080594102926944, + "grad_norm": 0.1671910583972931, + "learning_rate": 4.8242548654177615e-05, + "loss": 3.2551, + "step": 6745 + }, + { + "epoch": 0.2080902567582678, + "grad_norm": 0.18042127788066864, + "learning_rate": 4.824155704642058e-05, + "loss": 3.2614, + "step": 6746 + }, + { + "epoch": 0.20812110322384123, + "grad_norm": 0.16130955517292023, + "learning_rate": 4.8240565169191664e-05, + "loss": 3.21, + "step": 6747 + }, + { + "epoch": 0.20815194968941464, + "grad_norm": 0.16385617852210999, + "learning_rate": 4.823957302250236e-05, + "loss": 3.1928, + "step": 6748 + }, + { + "epoch": 0.20818279615498805, + "grad_norm": 0.1578904092311859, + "learning_rate": 4.823858060636417e-05, + "loss": 3.2601, + "step": 6749 + }, + { + "epoch": 0.2082136426205615, + "grad_norm": 0.15892331302165985, + "learning_rate": 4.823758792078861e-05, + "loss": 3.1912, + "step": 6750 + }, + { + "epoch": 0.2082444890861349, + "grad_norm": 0.1562507152557373, + "learning_rate": 4.823659496578717e-05, + "loss": 3.2266, + "step": 6751 + }, + { + "epoch": 0.20827533555170832, + "grad_norm": 0.15610018372535706, + "learning_rate": 4.8235601741371386e-05, + "loss": 3.2174, + "step": 6752 + }, + { + "epoch": 0.20830618201728174, + "grad_norm": 0.1566774845123291, + "learning_rate": 4.8234608247552754e-05, + "loss": 3.2064, + "step": 6753 + }, + { + "epoch": 0.20833702848285515, + "grad_norm": 0.15581010282039642, + "learning_rate": 4.82336144843428e-05, + "loss": 3.2542, + "step": 6754 + }, + { + "epoch": 0.20836787494842857, + "grad_norm": 0.14990226924419403, + "learning_rate": 4.823262045175306e-05, + "loss": 3.2189, + "step": 6755 + }, + { + "epoch": 0.20839872141400198, + "grad_norm": 0.1559952348470688, + "learning_rate": 4.823162614979504e-05, + "loss": 3.212, + "step": 6756 + }, + { + "epoch": 0.2084295678795754, + "grad_norm": 0.18678723275661469, + "learning_rate": 4.823063157848028e-05, + "loss": 3.211, + "step": 6757 + }, + { + "epoch": 0.2084604143451488, + "grad_norm": 0.15804806351661682, + "learning_rate": 4.8229636737820295e-05, + "loss": 3.2289, + "step": 6758 + }, + { + "epoch": 0.20849126081072222, + "grad_norm": 0.14792245626449585, + "learning_rate": 4.822864162782664e-05, + "loss": 3.1903, + "step": 6759 + }, + { + "epoch": 0.20852210727629564, + "grad_norm": 0.1500328779220581, + "learning_rate": 4.8227646248510845e-05, + "loss": 3.2201, + "step": 6760 + }, + { + "epoch": 0.20855295374186905, + "grad_norm": 0.15218617022037506, + "learning_rate": 4.822665059988445e-05, + "loss": 3.2022, + "step": 6761 + }, + { + "epoch": 0.2085838002074425, + "grad_norm": 0.15275385975837708, + "learning_rate": 4.8225654681958995e-05, + "loss": 3.2339, + "step": 6762 + }, + { + "epoch": 0.2086146466730159, + "grad_norm": 0.15098856389522552, + "learning_rate": 4.822465849474603e-05, + "loss": 3.2222, + "step": 6763 + }, + { + "epoch": 0.20864549313858932, + "grad_norm": 0.20638374984264374, + "learning_rate": 4.822366203825711e-05, + "loss": 3.2261, + "step": 6764 + }, + { + "epoch": 0.20867633960416274, + "grad_norm": 0.16799885034561157, + "learning_rate": 4.8222665312503786e-05, + "loss": 3.1945, + "step": 6765 + }, + { + "epoch": 0.20870718606973615, + "grad_norm": 0.1520565152168274, + "learning_rate": 4.822166831749761e-05, + "loss": 3.2249, + "step": 6766 + }, + { + "epoch": 0.20873803253530956, + "grad_norm": 0.15549132227897644, + "learning_rate": 4.822067105325014e-05, + "loss": 3.2542, + "step": 6767 + }, + { + "epoch": 0.20876887900088298, + "grad_norm": 0.15516076982021332, + "learning_rate": 4.8219673519772953e-05, + "loss": 3.2458, + "step": 6768 + }, + { + "epoch": 0.2087997254664564, + "grad_norm": 0.1598573625087738, + "learning_rate": 4.82186757170776e-05, + "loss": 3.2259, + "step": 6769 + }, + { + "epoch": 0.2088305719320298, + "grad_norm": 0.15984244644641876, + "learning_rate": 4.821767764517565e-05, + "loss": 3.2062, + "step": 6770 + }, + { + "epoch": 0.20886141839760322, + "grad_norm": 0.15884144604206085, + "learning_rate": 4.821667930407868e-05, + "loss": 3.2316, + "step": 6771 + }, + { + "epoch": 0.20889226486317664, + "grad_norm": 0.16458557546138763, + "learning_rate": 4.8215680693798275e-05, + "loss": 3.2223, + "step": 6772 + }, + { + "epoch": 0.20892311132875005, + "grad_norm": 0.15410158038139343, + "learning_rate": 4.8214681814346e-05, + "loss": 3.2016, + "step": 6773 + }, + { + "epoch": 0.2089539577943235, + "grad_norm": 0.18271993100643158, + "learning_rate": 4.821368266573344e-05, + "loss": 3.2285, + "step": 6774 + }, + { + "epoch": 0.2089848042598969, + "grad_norm": 0.1635541468858719, + "learning_rate": 4.8212683247972175e-05, + "loss": 3.2337, + "step": 6775 + }, + { + "epoch": 0.20901565072547032, + "grad_norm": 0.14922617375850677, + "learning_rate": 4.8211683561073796e-05, + "loss": 3.2149, + "step": 6776 + }, + { + "epoch": 0.20904649719104373, + "grad_norm": 0.5658249855041504, + "learning_rate": 4.82106836050499e-05, + "loss": 3.2097, + "step": 6777 + }, + { + "epoch": 0.20907734365661715, + "grad_norm": 0.15426014363765717, + "learning_rate": 4.820968337991207e-05, + "loss": 3.201, + "step": 6778 + }, + { + "epoch": 0.20910819012219056, + "grad_norm": 0.15606272220611572, + "learning_rate": 4.8208682885671916e-05, + "loss": 3.2243, + "step": 6779 + }, + { + "epoch": 0.20913903658776398, + "grad_norm": 0.16782517731189728, + "learning_rate": 4.820768212234103e-05, + "loss": 3.2102, + "step": 6780 + }, + { + "epoch": 0.2091698830533374, + "grad_norm": 0.15528935194015503, + "learning_rate": 4.820668108993101e-05, + "loss": 3.2441, + "step": 6781 + }, + { + "epoch": 0.2092007295189108, + "grad_norm": 0.16488590836524963, + "learning_rate": 4.8205679788453464e-05, + "loss": 3.2207, + "step": 6782 + }, + { + "epoch": 0.20923157598448422, + "grad_norm": 0.21375896036624908, + "learning_rate": 4.820467821792002e-05, + "loss": 3.252, + "step": 6783 + }, + { + "epoch": 0.20926242245005763, + "grad_norm": 0.16357798874378204, + "learning_rate": 4.8203676378342263e-05, + "loss": 3.2245, + "step": 6784 + }, + { + "epoch": 0.20929326891563105, + "grad_norm": 0.16058769822120667, + "learning_rate": 4.820267426973183e-05, + "loss": 3.2478, + "step": 6785 + }, + { + "epoch": 0.2093241153812045, + "grad_norm": 0.1639024168252945, + "learning_rate": 4.820167189210033e-05, + "loss": 3.216, + "step": 6786 + }, + { + "epoch": 0.2093549618467779, + "grad_norm": 0.1680074781179428, + "learning_rate": 4.820066924545939e-05, + "loss": 3.2007, + "step": 6787 + }, + { + "epoch": 0.20938580831235132, + "grad_norm": 0.16216690838336945, + "learning_rate": 4.819966632982062e-05, + "loss": 3.211, + "step": 6788 + }, + { + "epoch": 0.20941665477792473, + "grad_norm": 0.1896141916513443, + "learning_rate": 4.819866314519568e-05, + "loss": 3.1979, + "step": 6789 + }, + { + "epoch": 0.20944750124349815, + "grad_norm": 0.1732195019721985, + "learning_rate": 4.819765969159617e-05, + "loss": 3.2252, + "step": 6790 + }, + { + "epoch": 0.20947834770907156, + "grad_norm": 0.16481323540210724, + "learning_rate": 4.819665596903374e-05, + "loss": 3.2155, + "step": 6791 + }, + { + "epoch": 0.20950919417464497, + "grad_norm": 0.16315200924873352, + "learning_rate": 4.8195651977520015e-05, + "loss": 3.218, + "step": 6792 + }, + { + "epoch": 0.2095400406402184, + "grad_norm": 0.15824578702449799, + "learning_rate": 4.8194647717066654e-05, + "loss": 3.2189, + "step": 6793 + }, + { + "epoch": 0.2095708871057918, + "grad_norm": 0.16330747306346893, + "learning_rate": 4.8193643187685286e-05, + "loss": 3.2095, + "step": 6794 + }, + { + "epoch": 0.20960173357136522, + "grad_norm": 0.15581220388412476, + "learning_rate": 4.8192638389387565e-05, + "loss": 3.2284, + "step": 6795 + }, + { + "epoch": 0.20963258003693863, + "grad_norm": 0.1525992751121521, + "learning_rate": 4.8191633322185134e-05, + "loss": 3.2048, + "step": 6796 + }, + { + "epoch": 0.20966342650251205, + "grad_norm": 0.1640002280473709, + "learning_rate": 4.819062798608967e-05, + "loss": 3.2012, + "step": 6797 + }, + { + "epoch": 0.2096942729680855, + "grad_norm": 0.15152327716350555, + "learning_rate": 4.818962238111279e-05, + "loss": 3.2794, + "step": 6798 + }, + { + "epoch": 0.2097251194336589, + "grad_norm": 0.3921668827533722, + "learning_rate": 4.818861650726618e-05, + "loss": 3.2856, + "step": 6799 + }, + { + "epoch": 0.20975596589923232, + "grad_norm": 0.15819118916988373, + "learning_rate": 4.8187610364561496e-05, + "loss": 3.2336, + "step": 6800 + }, + { + "epoch": 0.20978681236480573, + "grad_norm": 0.1653851419687271, + "learning_rate": 4.81866039530104e-05, + "loss": 3.2287, + "step": 6801 + }, + { + "epoch": 0.20981765883037914, + "grad_norm": 0.16474780440330505, + "learning_rate": 4.8185597272624574e-05, + "loss": 3.1829, + "step": 6802 + }, + { + "epoch": 0.20984850529595256, + "grad_norm": 0.16394755244255066, + "learning_rate": 4.818459032341567e-05, + "loss": 3.2168, + "step": 6803 + }, + { + "epoch": 0.20987935176152597, + "grad_norm": 0.15859940648078918, + "learning_rate": 4.818358310539538e-05, + "loss": 3.2276, + "step": 6804 + }, + { + "epoch": 0.2099101982270994, + "grad_norm": 0.1577334851026535, + "learning_rate": 4.8182575618575375e-05, + "loss": 3.2396, + "step": 6805 + }, + { + "epoch": 0.2099410446926728, + "grad_norm": 0.15577787160873413, + "learning_rate": 4.8181567862967334e-05, + "loss": 3.2094, + "step": 6806 + }, + { + "epoch": 0.20997189115824622, + "grad_norm": 0.16389568150043488, + "learning_rate": 4.818055983858295e-05, + "loss": 3.2397, + "step": 6807 + }, + { + "epoch": 0.21000273762381963, + "grad_norm": 0.16053485870361328, + "learning_rate": 4.81795515454339e-05, + "loss": 3.195, + "step": 6808 + }, + { + "epoch": 0.21003358408939304, + "grad_norm": 0.1529531478881836, + "learning_rate": 4.817854298353188e-05, + "loss": 3.2302, + "step": 6809 + }, + { + "epoch": 0.21006443055496649, + "grad_norm": 0.15775339305400848, + "learning_rate": 4.8177534152888583e-05, + "loss": 3.2335, + "step": 6810 + }, + { + "epoch": 0.2100952770205399, + "grad_norm": 0.168767511844635, + "learning_rate": 4.817652505351571e-05, + "loss": 3.2228, + "step": 6811 + }, + { + "epoch": 0.21012612348611331, + "grad_norm": 0.1622428447008133, + "learning_rate": 4.817551568542495e-05, + "loss": 3.1885, + "step": 6812 + }, + { + "epoch": 0.21015696995168673, + "grad_norm": 0.1926865130662918, + "learning_rate": 4.8174506048628006e-05, + "loss": 3.2772, + "step": 6813 + }, + { + "epoch": 0.21018781641726014, + "grad_norm": 0.16019582748413086, + "learning_rate": 4.817349614313661e-05, + "loss": 3.2332, + "step": 6814 + }, + { + "epoch": 0.21021866288283356, + "grad_norm": 0.1627427488565445, + "learning_rate": 4.817248596896244e-05, + "loss": 3.2269, + "step": 6815 + }, + { + "epoch": 0.21024950934840697, + "grad_norm": 0.1598995327949524, + "learning_rate": 4.817147552611722e-05, + "loss": 3.2421, + "step": 6816 + }, + { + "epoch": 0.21028035581398039, + "grad_norm": 0.1532595008611679, + "learning_rate": 4.817046481461267e-05, + "loss": 3.2245, + "step": 6817 + }, + { + "epoch": 0.2103112022795538, + "grad_norm": 0.15431822836399078, + "learning_rate": 4.8169453834460496e-05, + "loss": 3.2232, + "step": 6818 + }, + { + "epoch": 0.2103420487451272, + "grad_norm": 0.17158812284469604, + "learning_rate": 4.816844258567243e-05, + "loss": 3.231, + "step": 6819 + }, + { + "epoch": 0.21037289521070063, + "grad_norm": 0.1583348512649536, + "learning_rate": 4.8167431068260204e-05, + "loss": 3.2275, + "step": 6820 + }, + { + "epoch": 0.21040374167627404, + "grad_norm": 0.16978201270103455, + "learning_rate": 4.816641928223553e-05, + "loss": 3.2345, + "step": 6821 + }, + { + "epoch": 0.21043458814184748, + "grad_norm": 0.16278241574764252, + "learning_rate": 4.816540722761015e-05, + "loss": 3.2242, + "step": 6822 + }, + { + "epoch": 0.2104654346074209, + "grad_norm": 0.15273739397525787, + "learning_rate": 4.8164394904395786e-05, + "loss": 3.2556, + "step": 6823 + }, + { + "epoch": 0.2104962810729943, + "grad_norm": 0.164208322763443, + "learning_rate": 4.816338231260419e-05, + "loss": 3.2303, + "step": 6824 + }, + { + "epoch": 0.21052712753856773, + "grad_norm": 0.15657271444797516, + "learning_rate": 4.816236945224709e-05, + "loss": 3.2627, + "step": 6825 + }, + { + "epoch": 0.21055797400414114, + "grad_norm": 0.16548331081867218, + "learning_rate": 4.816135632333624e-05, + "loss": 3.2101, + "step": 6826 + }, + { + "epoch": 0.21058882046971455, + "grad_norm": 0.15425272285938263, + "learning_rate": 4.816034292588338e-05, + "loss": 3.2321, + "step": 6827 + }, + { + "epoch": 0.21061966693528797, + "grad_norm": 0.1545306146144867, + "learning_rate": 4.815932925990027e-05, + "loss": 3.2485, + "step": 6828 + }, + { + "epoch": 0.21065051340086138, + "grad_norm": 0.1603478193283081, + "learning_rate": 4.815831532539864e-05, + "loss": 3.2374, + "step": 6829 + }, + { + "epoch": 0.2106813598664348, + "grad_norm": 0.165154829621315, + "learning_rate": 4.815730112239027e-05, + "loss": 3.2389, + "step": 6830 + }, + { + "epoch": 0.2107122063320082, + "grad_norm": 0.151805117726326, + "learning_rate": 4.81562866508869e-05, + "loss": 3.2303, + "step": 6831 + }, + { + "epoch": 0.21074305279758163, + "grad_norm": 0.16346444189548492, + "learning_rate": 4.815527191090031e-05, + "loss": 3.2666, + "step": 6832 + }, + { + "epoch": 0.21077389926315504, + "grad_norm": 0.16216757893562317, + "learning_rate": 4.815425690244226e-05, + "loss": 3.241, + "step": 6833 + }, + { + "epoch": 0.21080474572872848, + "grad_norm": 0.15913552045822144, + "learning_rate": 4.815324162552451e-05, + "loss": 3.2216, + "step": 6834 + }, + { + "epoch": 0.2108355921943019, + "grad_norm": 0.15449585020542145, + "learning_rate": 4.8152226080158835e-05, + "loss": 3.2019, + "step": 6835 + }, + { + "epoch": 0.2108664386598753, + "grad_norm": 0.1695776879787445, + "learning_rate": 4.815121026635701e-05, + "loss": 3.2022, + "step": 6836 + }, + { + "epoch": 0.21089728512544872, + "grad_norm": 0.1667395979166031, + "learning_rate": 4.815019418413082e-05, + "loss": 3.2551, + "step": 6837 + }, + { + "epoch": 0.21092813159102214, + "grad_norm": 0.159615620970726, + "learning_rate": 4.8149177833492035e-05, + "loss": 3.2076, + "step": 6838 + }, + { + "epoch": 0.21095897805659555, + "grad_norm": 0.1518634408712387, + "learning_rate": 4.8148161214452446e-05, + "loss": 3.2487, + "step": 6839 + }, + { + "epoch": 0.21098982452216897, + "grad_norm": 0.15319910645484924, + "learning_rate": 4.814714432702384e-05, + "loss": 3.2755, + "step": 6840 + }, + { + "epoch": 0.21102067098774238, + "grad_norm": 0.16190485656261444, + "learning_rate": 4.814612717121801e-05, + "loss": 3.2296, + "step": 6841 + }, + { + "epoch": 0.2110515174533158, + "grad_norm": 0.1569053679704666, + "learning_rate": 4.814510974704674e-05, + "loss": 3.2002, + "step": 6842 + }, + { + "epoch": 0.2110823639188892, + "grad_norm": 0.15669281780719757, + "learning_rate": 4.814409205452183e-05, + "loss": 3.2175, + "step": 6843 + }, + { + "epoch": 0.21111321038446262, + "grad_norm": 0.1651158183813095, + "learning_rate": 4.8143074093655076e-05, + "loss": 3.2214, + "step": 6844 + }, + { + "epoch": 0.21114405685003604, + "grad_norm": 0.16371624171733856, + "learning_rate": 4.814205586445829e-05, + "loss": 3.2634, + "step": 6845 + }, + { + "epoch": 0.21117490331560948, + "grad_norm": 0.16722749173641205, + "learning_rate": 4.814103736694328e-05, + "loss": 3.2219, + "step": 6846 + }, + { + "epoch": 0.2112057497811829, + "grad_norm": 0.16359737515449524, + "learning_rate": 4.814001860112183e-05, + "loss": 3.2426, + "step": 6847 + }, + { + "epoch": 0.2112365962467563, + "grad_norm": 0.15458615124225616, + "learning_rate": 4.8138999567005784e-05, + "loss": 3.2037, + "step": 6848 + }, + { + "epoch": 0.21126744271232972, + "grad_norm": 0.16432982683181763, + "learning_rate": 4.8137980264606944e-05, + "loss": 3.2341, + "step": 6849 + }, + { + "epoch": 0.21129828917790314, + "grad_norm": 0.1608530730009079, + "learning_rate": 4.8136960693937116e-05, + "loss": 3.2742, + "step": 6850 + }, + { + "epoch": 0.21132913564347655, + "grad_norm": 0.1565203070640564, + "learning_rate": 4.8135940855008146e-05, + "loss": 3.2084, + "step": 6851 + }, + { + "epoch": 0.21135998210904997, + "grad_norm": 0.17942345142364502, + "learning_rate": 4.813492074783184e-05, + "loss": 3.2444, + "step": 6852 + }, + { + "epoch": 0.21139082857462338, + "grad_norm": 0.15676265954971313, + "learning_rate": 4.813390037242002e-05, + "loss": 3.2184, + "step": 6853 + }, + { + "epoch": 0.2114216750401968, + "grad_norm": 0.15006332099437714, + "learning_rate": 4.813287972878454e-05, + "loss": 3.1958, + "step": 6854 + }, + { + "epoch": 0.2114525215057702, + "grad_norm": 0.16700707376003265, + "learning_rate": 4.8131858816937214e-05, + "loss": 3.2319, + "step": 6855 + }, + { + "epoch": 0.21148336797134362, + "grad_norm": 0.15832695364952087, + "learning_rate": 4.813083763688989e-05, + "loss": 3.2311, + "step": 6856 + }, + { + "epoch": 0.21151421443691704, + "grad_norm": 0.1531638205051422, + "learning_rate": 4.8129816188654395e-05, + "loss": 3.257, + "step": 6857 + }, + { + "epoch": 0.21154506090249048, + "grad_norm": 0.15675131976604462, + "learning_rate": 4.812879447224259e-05, + "loss": 3.2065, + "step": 6858 + }, + { + "epoch": 0.2115759073680639, + "grad_norm": 0.15328016877174377, + "learning_rate": 4.812777248766631e-05, + "loss": 3.2277, + "step": 6859 + }, + { + "epoch": 0.2116067538336373, + "grad_norm": 0.15203586220741272, + "learning_rate": 4.8126750234937407e-05, + "loss": 3.1942, + "step": 6860 + }, + { + "epoch": 0.21163760029921072, + "grad_norm": 0.1536298543214798, + "learning_rate": 4.812572771406773e-05, + "loss": 3.2209, + "step": 6861 + }, + { + "epoch": 0.21166844676478414, + "grad_norm": 0.15473301708698273, + "learning_rate": 4.812470492506914e-05, + "loss": 3.2103, + "step": 6862 + }, + { + "epoch": 0.21169929323035755, + "grad_norm": 0.16073206067085266, + "learning_rate": 4.8123681867953485e-05, + "loss": 3.2147, + "step": 6863 + }, + { + "epoch": 0.21173013969593096, + "grad_norm": 0.15226882696151733, + "learning_rate": 4.8122658542732645e-05, + "loss": 3.2058, + "step": 6864 + }, + { + "epoch": 0.21176098616150438, + "grad_norm": 0.155996635556221, + "learning_rate": 4.812163494941847e-05, + "loss": 3.1885, + "step": 6865 + }, + { + "epoch": 0.2117918326270778, + "grad_norm": 0.15151600539684296, + "learning_rate": 4.812061108802283e-05, + "loss": 3.2317, + "step": 6866 + }, + { + "epoch": 0.2118226790926512, + "grad_norm": 0.15039218962192535, + "learning_rate": 4.81195869585576e-05, + "loss": 3.1837, + "step": 6867 + }, + { + "epoch": 0.21185352555822462, + "grad_norm": 0.16717156767845154, + "learning_rate": 4.811856256103465e-05, + "loss": 3.2179, + "step": 6868 + }, + { + "epoch": 0.21188437202379803, + "grad_norm": 0.14571617543697357, + "learning_rate": 4.811753789546587e-05, + "loss": 3.1959, + "step": 6869 + }, + { + "epoch": 0.21191521848937148, + "grad_norm": 0.15923696756362915, + "learning_rate": 4.8116512961863115e-05, + "loss": 3.2083, + "step": 6870 + }, + { + "epoch": 0.2119460649549449, + "grad_norm": 0.15510256588459015, + "learning_rate": 4.8115487760238296e-05, + "loss": 3.2258, + "step": 6871 + }, + { + "epoch": 0.2119769114205183, + "grad_norm": 0.1614210307598114, + "learning_rate": 4.811446229060328e-05, + "loss": 3.2406, + "step": 6872 + }, + { + "epoch": 0.21200775788609172, + "grad_norm": 0.1585838794708252, + "learning_rate": 4.811343655296997e-05, + "loss": 3.2118, + "step": 6873 + }, + { + "epoch": 0.21203860435166513, + "grad_norm": 0.1541222184896469, + "learning_rate": 4.811241054735024e-05, + "loss": 3.187, + "step": 6874 + }, + { + "epoch": 0.21206945081723855, + "grad_norm": 0.15135420858860016, + "learning_rate": 4.811138427375601e-05, + "loss": 3.2167, + "step": 6875 + }, + { + "epoch": 0.21210029728281196, + "grad_norm": 0.15292978286743164, + "learning_rate": 4.811035773219916e-05, + "loss": 3.2344, + "step": 6876 + }, + { + "epoch": 0.21213114374838538, + "grad_norm": 0.15573836863040924, + "learning_rate": 4.810933092269161e-05, + "loss": 3.2601, + "step": 6877 + }, + { + "epoch": 0.2121619902139588, + "grad_norm": 0.1579369306564331, + "learning_rate": 4.810830384524525e-05, + "loss": 3.2272, + "step": 6878 + }, + { + "epoch": 0.2121928366795322, + "grad_norm": 0.1473424881696701, + "learning_rate": 4.8107276499871995e-05, + "loss": 3.2396, + "step": 6879 + }, + { + "epoch": 0.21222368314510562, + "grad_norm": 0.16815809905529022, + "learning_rate": 4.810624888658375e-05, + "loss": 3.2428, + "step": 6880 + }, + { + "epoch": 0.21225452961067906, + "grad_norm": 0.1538865715265274, + "learning_rate": 4.810522100539244e-05, + "loss": 3.1945, + "step": 6881 + }, + { + "epoch": 0.21228537607625247, + "grad_norm": 0.15196765959262848, + "learning_rate": 4.810419285630997e-05, + "loss": 3.1772, + "step": 6882 + }, + { + "epoch": 0.2123162225418259, + "grad_norm": 0.1578824669122696, + "learning_rate": 4.8103164439348284e-05, + "loss": 3.1969, + "step": 6883 + }, + { + "epoch": 0.2123470690073993, + "grad_norm": 0.17320744693279266, + "learning_rate": 4.8102135754519274e-05, + "loss": 3.2087, + "step": 6884 + }, + { + "epoch": 0.21237791547297272, + "grad_norm": 0.1542360633611679, + "learning_rate": 4.8101106801834885e-05, + "loss": 3.2255, + "step": 6885 + }, + { + "epoch": 0.21240876193854613, + "grad_norm": 0.1482093185186386, + "learning_rate": 4.8100077581307044e-05, + "loss": 3.2047, + "step": 6886 + }, + { + "epoch": 0.21243960840411955, + "grad_norm": 0.14969629049301147, + "learning_rate": 4.809904809294769e-05, + "loss": 3.238, + "step": 6887 + }, + { + "epoch": 0.21247045486969296, + "grad_norm": 0.15623001754283905, + "learning_rate": 4.8098018336768755e-05, + "loss": 3.2556, + "step": 6888 + }, + { + "epoch": 0.21250130133526637, + "grad_norm": 0.1530306190252304, + "learning_rate": 4.8096988312782174e-05, + "loss": 3.2078, + "step": 6889 + }, + { + "epoch": 0.2125321478008398, + "grad_norm": 0.15761953592300415, + "learning_rate": 4.8095958020999885e-05, + "loss": 3.1903, + "step": 6890 + }, + { + "epoch": 0.2125629942664132, + "grad_norm": 0.1524045765399933, + "learning_rate": 4.8094927461433856e-05, + "loss": 3.2426, + "step": 6891 + }, + { + "epoch": 0.21259384073198662, + "grad_norm": 0.15663231909275055, + "learning_rate": 4.809389663409601e-05, + "loss": 3.1994, + "step": 6892 + }, + { + "epoch": 0.21262468719756006, + "grad_norm": 0.15573172271251678, + "learning_rate": 4.809286553899832e-05, + "loss": 3.2347, + "step": 6893 + }, + { + "epoch": 0.21265553366313347, + "grad_norm": 0.1522114872932434, + "learning_rate": 4.8091834176152725e-05, + "loss": 3.2529, + "step": 6894 + }, + { + "epoch": 0.2126863801287069, + "grad_norm": 0.1659248173236847, + "learning_rate": 4.80908025455712e-05, + "loss": 3.2224, + "step": 6895 + }, + { + "epoch": 0.2127172265942803, + "grad_norm": 0.14749743044376373, + "learning_rate": 4.808977064726569e-05, + "loss": 3.1947, + "step": 6896 + }, + { + "epoch": 0.21274807305985372, + "grad_norm": 0.15272800624370575, + "learning_rate": 4.8088738481248155e-05, + "loss": 3.2249, + "step": 6897 + }, + { + "epoch": 0.21277891952542713, + "grad_norm": 0.15644928812980652, + "learning_rate": 4.8087706047530576e-05, + "loss": 3.2524, + "step": 6898 + }, + { + "epoch": 0.21280976599100054, + "grad_norm": 0.1502804309129715, + "learning_rate": 4.808667334612493e-05, + "loss": 3.2489, + "step": 6899 + }, + { + "epoch": 0.21284061245657396, + "grad_norm": 0.15710456669330597, + "learning_rate": 4.808564037704317e-05, + "loss": 3.1909, + "step": 6900 + }, + { + "epoch": 0.21287145892214737, + "grad_norm": 0.1726612001657486, + "learning_rate": 4.808460714029729e-05, + "loss": 3.2162, + "step": 6901 + }, + { + "epoch": 0.21290230538772079, + "grad_norm": 0.15713410079479218, + "learning_rate": 4.808357363589925e-05, + "loss": 3.2234, + "step": 6902 + }, + { + "epoch": 0.2129331518532942, + "grad_norm": 0.14887037873268127, + "learning_rate": 4.808253986386106e-05, + "loss": 3.2044, + "step": 6903 + }, + { + "epoch": 0.21296399831886761, + "grad_norm": 0.15474705398082733, + "learning_rate": 4.8081505824194687e-05, + "loss": 3.2365, + "step": 6904 + }, + { + "epoch": 0.21299484478444106, + "grad_norm": 0.15326422452926636, + "learning_rate": 4.808047151691212e-05, + "loss": 3.2659, + "step": 6905 + }, + { + "epoch": 0.21302569125001447, + "grad_norm": 0.1557539701461792, + "learning_rate": 4.807943694202536e-05, + "loss": 3.2089, + "step": 6906 + }, + { + "epoch": 0.21305653771558788, + "grad_norm": 0.1602124720811844, + "learning_rate": 4.807840209954639e-05, + "loss": 3.2332, + "step": 6907 + }, + { + "epoch": 0.2130873841811613, + "grad_norm": 0.16489411890506744, + "learning_rate": 4.807736698948722e-05, + "loss": 3.1823, + "step": 6908 + }, + { + "epoch": 0.2131182306467347, + "grad_norm": 0.16108350455760956, + "learning_rate": 4.807633161185985e-05, + "loss": 3.2169, + "step": 6909 + }, + { + "epoch": 0.21314907711230813, + "grad_norm": 0.3500319719314575, + "learning_rate": 4.807529596667628e-05, + "loss": 3.2332, + "step": 6910 + }, + { + "epoch": 0.21317992357788154, + "grad_norm": 0.15268771350383759, + "learning_rate": 4.807426005394852e-05, + "loss": 3.2355, + "step": 6911 + }, + { + "epoch": 0.21321077004345496, + "grad_norm": 0.14916884899139404, + "learning_rate": 4.807322387368858e-05, + "loss": 3.24, + "step": 6912 + }, + { + "epoch": 0.21324161650902837, + "grad_norm": 0.1683206856250763, + "learning_rate": 4.8072187425908476e-05, + "loss": 3.2138, + "step": 6913 + }, + { + "epoch": 0.21327246297460178, + "grad_norm": 0.15187770128250122, + "learning_rate": 4.807115071062023e-05, + "loss": 3.193, + "step": 6914 + }, + { + "epoch": 0.2133033094401752, + "grad_norm": 0.15513989329338074, + "learning_rate": 4.807011372783585e-05, + "loss": 3.2083, + "step": 6915 + }, + { + "epoch": 0.2133341559057486, + "grad_norm": 0.16079995036125183, + "learning_rate": 4.8069076477567365e-05, + "loss": 3.1999, + "step": 6916 + }, + { + "epoch": 0.21336500237132205, + "grad_norm": 0.16570693254470825, + "learning_rate": 4.80680389598268e-05, + "loss": 3.2365, + "step": 6917 + }, + { + "epoch": 0.21339584883689547, + "grad_norm": 0.1553819328546524, + "learning_rate": 4.806700117462618e-05, + "loss": 3.2142, + "step": 6918 + }, + { + "epoch": 0.21342669530246888, + "grad_norm": 0.6294457912445068, + "learning_rate": 4.806596312197755e-05, + "loss": 3.2346, + "step": 6919 + }, + { + "epoch": 0.2134575417680423, + "grad_norm": 0.1584113985300064, + "learning_rate": 4.806492480189294e-05, + "loss": 3.2155, + "step": 6920 + }, + { + "epoch": 0.2134883882336157, + "grad_norm": 0.1685524880886078, + "learning_rate": 4.8063886214384385e-05, + "loss": 3.2045, + "step": 6921 + }, + { + "epoch": 0.21351923469918913, + "grad_norm": 0.16666196286678314, + "learning_rate": 4.806284735946393e-05, + "loss": 3.225, + "step": 6922 + }, + { + "epoch": 0.21355008116476254, + "grad_norm": 0.16409660875797272, + "learning_rate": 4.806180823714362e-05, + "loss": 3.2141, + "step": 6923 + }, + { + "epoch": 0.21358092763033595, + "grad_norm": 0.16609396040439606, + "learning_rate": 4.80607688474355e-05, + "loss": 3.2381, + "step": 6924 + }, + { + "epoch": 0.21361177409590937, + "grad_norm": 0.15878812968730927, + "learning_rate": 4.805972919035162e-05, + "loss": 3.2428, + "step": 6925 + }, + { + "epoch": 0.21364262056148278, + "grad_norm": 0.16535213589668274, + "learning_rate": 4.805868926590404e-05, + "loss": 3.2324, + "step": 6926 + }, + { + "epoch": 0.2136734670270562, + "grad_norm": 0.15831956267356873, + "learning_rate": 4.805764907410482e-05, + "loss": 3.2003, + "step": 6927 + }, + { + "epoch": 0.2137043134926296, + "grad_norm": 0.16547273099422455, + "learning_rate": 4.805660861496601e-05, + "loss": 3.2317, + "step": 6928 + }, + { + "epoch": 0.21373515995820305, + "grad_norm": 0.1629832237958908, + "learning_rate": 4.805556788849967e-05, + "loss": 3.2415, + "step": 6929 + }, + { + "epoch": 0.21376600642377647, + "grad_norm": 0.3016655147075653, + "learning_rate": 4.805452689471789e-05, + "loss": 3.2544, + "step": 6930 + }, + { + "epoch": 0.21379685288934988, + "grad_norm": 0.16451796889305115, + "learning_rate": 4.805348563363272e-05, + "loss": 3.2638, + "step": 6931 + }, + { + "epoch": 0.2138276993549233, + "grad_norm": 0.1604529172182083, + "learning_rate": 4.805244410525623e-05, + "loss": 3.2154, + "step": 6932 + }, + { + "epoch": 0.2138585458204967, + "grad_norm": 0.1587255597114563, + "learning_rate": 4.8051402309600514e-05, + "loss": 3.197, + "step": 6933 + }, + { + "epoch": 0.21388939228607012, + "grad_norm": 0.15545272827148438, + "learning_rate": 4.805036024667764e-05, + "loss": 3.2051, + "step": 6934 + }, + { + "epoch": 0.21392023875164354, + "grad_norm": 0.16109824180603027, + "learning_rate": 4.8049317916499685e-05, + "loss": 3.2464, + "step": 6935 + }, + { + "epoch": 0.21395108521721695, + "grad_norm": 0.1666591912508011, + "learning_rate": 4.804827531907874e-05, + "loss": 3.2466, + "step": 6936 + }, + { + "epoch": 0.21398193168279037, + "grad_norm": 0.1561024934053421, + "learning_rate": 4.804723245442689e-05, + "loss": 3.2368, + "step": 6937 + }, + { + "epoch": 0.21401277814836378, + "grad_norm": 0.1583617478609085, + "learning_rate": 4.804618932255624e-05, + "loss": 3.2208, + "step": 6938 + }, + { + "epoch": 0.2140436246139372, + "grad_norm": 0.16773316264152527, + "learning_rate": 4.804514592347887e-05, + "loss": 3.2293, + "step": 6939 + }, + { + "epoch": 0.2140744710795106, + "grad_norm": 0.16082081198692322, + "learning_rate": 4.804410225720688e-05, + "loss": 3.248, + "step": 6940 + }, + { + "epoch": 0.21410531754508405, + "grad_norm": 0.16377946734428406, + "learning_rate": 4.804305832375237e-05, + "loss": 3.2301, + "step": 6941 + }, + { + "epoch": 0.21413616401065746, + "grad_norm": 0.1834593415260315, + "learning_rate": 4.804201412312746e-05, + "loss": 3.2137, + "step": 6942 + }, + { + "epoch": 0.21416701047623088, + "grad_norm": 0.16250073909759521, + "learning_rate": 4.804096965534422e-05, + "loss": 3.2279, + "step": 6943 + }, + { + "epoch": 0.2141978569418043, + "grad_norm": 0.15206634998321533, + "learning_rate": 4.803992492041479e-05, + "loss": 3.216, + "step": 6944 + }, + { + "epoch": 0.2142287034073777, + "grad_norm": 0.15786056220531464, + "learning_rate": 4.8038879918351286e-05, + "loss": 3.2243, + "step": 6945 + }, + { + "epoch": 0.21425954987295112, + "grad_norm": 0.15185874700546265, + "learning_rate": 4.80378346491658e-05, + "loss": 3.1966, + "step": 6946 + }, + { + "epoch": 0.21429039633852454, + "grad_norm": 0.15631596744060516, + "learning_rate": 4.8036789112870485e-05, + "loss": 3.2167, + "step": 6947 + }, + { + "epoch": 0.21432124280409795, + "grad_norm": 0.1625301092863083, + "learning_rate": 4.803574330947743e-05, + "loss": 3.2573, + "step": 6948 + }, + { + "epoch": 0.21435208926967136, + "grad_norm": 0.15227101743221283, + "learning_rate": 4.803469723899877e-05, + "loss": 3.2232, + "step": 6949 + }, + { + "epoch": 0.21438293573524478, + "grad_norm": 0.16204197704792023, + "learning_rate": 4.8033650901446646e-05, + "loss": 3.229, + "step": 6950 + }, + { + "epoch": 0.2144137822008182, + "grad_norm": 0.1547716110944748, + "learning_rate": 4.8032604296833185e-05, + "loss": 3.2079, + "step": 6951 + }, + { + "epoch": 0.2144446286663916, + "grad_norm": 0.170948326587677, + "learning_rate": 4.803155742517051e-05, + "loss": 3.233, + "step": 6952 + }, + { + "epoch": 0.21447547513196505, + "grad_norm": 0.17003649473190308, + "learning_rate": 4.8030510286470765e-05, + "loss": 3.2354, + "step": 6953 + }, + { + "epoch": 0.21450632159753846, + "grad_norm": 0.21793732047080994, + "learning_rate": 4.80294628807461e-05, + "loss": 3.2226, + "step": 6954 + }, + { + "epoch": 0.21453716806311188, + "grad_norm": 0.16064734756946564, + "learning_rate": 4.802841520800865e-05, + "loss": 3.2452, + "step": 6955 + }, + { + "epoch": 0.2145680145286853, + "grad_norm": 0.16778863966464996, + "learning_rate": 4.8027367268270564e-05, + "loss": 3.2194, + "step": 6956 + }, + { + "epoch": 0.2145988609942587, + "grad_norm": 0.16324812173843384, + "learning_rate": 4.802631906154399e-05, + "loss": 3.2305, + "step": 6957 + }, + { + "epoch": 0.21462970745983212, + "grad_norm": 0.14822465181350708, + "learning_rate": 4.802527058784109e-05, + "loss": 3.1834, + "step": 6958 + }, + { + "epoch": 0.21466055392540553, + "grad_norm": 0.1615249216556549, + "learning_rate": 4.8024221847174006e-05, + "loss": 3.2372, + "step": 6959 + }, + { + "epoch": 0.21469140039097895, + "grad_norm": 0.23460274934768677, + "learning_rate": 4.802317283955491e-05, + "loss": 3.2515, + "step": 6960 + }, + { + "epoch": 0.21472224685655236, + "grad_norm": 0.1670861393213272, + "learning_rate": 4.802212356499596e-05, + "loss": 3.1956, + "step": 6961 + }, + { + "epoch": 0.21475309332212578, + "grad_norm": 0.15342974662780762, + "learning_rate": 4.802107402350932e-05, + "loss": 3.2356, + "step": 6962 + }, + { + "epoch": 0.2147839397876992, + "grad_norm": 0.15587283670902252, + "learning_rate": 4.8020024215107165e-05, + "loss": 3.2429, + "step": 6963 + }, + { + "epoch": 0.2148147862532726, + "grad_norm": 0.16355790197849274, + "learning_rate": 4.801897413980166e-05, + "loss": 3.2276, + "step": 6964 + }, + { + "epoch": 0.21484563271884605, + "grad_norm": 0.1597791463136673, + "learning_rate": 4.801792379760498e-05, + "loss": 3.1776, + "step": 6965 + }, + { + "epoch": 0.21487647918441946, + "grad_norm": 0.1681433767080307, + "learning_rate": 4.801687318852931e-05, + "loss": 3.2456, + "step": 6966 + }, + { + "epoch": 0.21490732564999288, + "grad_norm": 0.15313468873500824, + "learning_rate": 4.8015822312586825e-05, + "loss": 3.2579, + "step": 6967 + }, + { + "epoch": 0.2149381721155663, + "grad_norm": 0.16034498810768127, + "learning_rate": 4.8014771169789706e-05, + "loss": 3.2224, + "step": 6968 + }, + { + "epoch": 0.2149690185811397, + "grad_norm": 0.15175484120845795, + "learning_rate": 4.801371976015016e-05, + "loss": 3.2195, + "step": 6969 + }, + { + "epoch": 0.21499986504671312, + "grad_norm": 0.16179941594600677, + "learning_rate": 4.801266808368035e-05, + "loss": 3.2503, + "step": 6970 + }, + { + "epoch": 0.21503071151228653, + "grad_norm": 0.1483241617679596, + "learning_rate": 4.801161614039249e-05, + "loss": 3.234, + "step": 6971 + }, + { + "epoch": 0.21506155797785995, + "grad_norm": 0.1582636684179306, + "learning_rate": 4.801056393029876e-05, + "loss": 3.2664, + "step": 6972 + }, + { + "epoch": 0.21509240444343336, + "grad_norm": 0.15272462368011475, + "learning_rate": 4.800951145341138e-05, + "loss": 3.1675, + "step": 6973 + }, + { + "epoch": 0.21512325090900677, + "grad_norm": 0.15670569241046906, + "learning_rate": 4.800845870974253e-05, + "loss": 3.2411, + "step": 6974 + }, + { + "epoch": 0.2151540973745802, + "grad_norm": 0.16305162012577057, + "learning_rate": 4.800740569930444e-05, + "loss": 3.2323, + "step": 6975 + }, + { + "epoch": 0.2151849438401536, + "grad_norm": 0.15680500864982605, + "learning_rate": 4.8006352422109305e-05, + "loss": 3.2084, + "step": 6976 + }, + { + "epoch": 0.21521579030572704, + "grad_norm": 0.15911288559436798, + "learning_rate": 4.800529887816933e-05, + "loss": 3.2222, + "step": 6977 + }, + { + "epoch": 0.21524663677130046, + "grad_norm": 0.15536537766456604, + "learning_rate": 4.800424506749674e-05, + "loss": 3.2219, + "step": 6978 + }, + { + "epoch": 0.21527748323687387, + "grad_norm": 0.1816033571958542, + "learning_rate": 4.8003190990103765e-05, + "loss": 3.2474, + "step": 6979 + }, + { + "epoch": 0.2153083297024473, + "grad_norm": 0.16416659951210022, + "learning_rate": 4.8002136646002604e-05, + "loss": 3.2298, + "step": 6980 + }, + { + "epoch": 0.2153391761680207, + "grad_norm": 0.1622118055820465, + "learning_rate": 4.800108203520549e-05, + "loss": 3.2307, + "step": 6981 + }, + { + "epoch": 0.21537002263359412, + "grad_norm": 0.16158533096313477, + "learning_rate": 4.800002715772466e-05, + "loss": 3.1884, + "step": 6982 + }, + { + "epoch": 0.21540086909916753, + "grad_norm": 0.15604552626609802, + "learning_rate": 4.799897201357232e-05, + "loss": 3.2125, + "step": 6983 + }, + { + "epoch": 0.21543171556474094, + "grad_norm": 0.15530312061309814, + "learning_rate": 4.799791660276074e-05, + "loss": 3.2366, + "step": 6984 + }, + { + "epoch": 0.21546256203031436, + "grad_norm": 0.16218316555023193, + "learning_rate": 4.7996860925302124e-05, + "loss": 3.2144, + "step": 6985 + }, + { + "epoch": 0.21549340849588777, + "grad_norm": 0.15335984528064728, + "learning_rate": 4.799580498120873e-05, + "loss": 3.2302, + "step": 6986 + }, + { + "epoch": 0.2155242549614612, + "grad_norm": 0.1617809236049652, + "learning_rate": 4.7994748770492795e-05, + "loss": 3.223, + "step": 6987 + }, + { + "epoch": 0.2155551014270346, + "grad_norm": 0.1641528308391571, + "learning_rate": 4.799369229316657e-05, + "loss": 3.2012, + "step": 6988 + }, + { + "epoch": 0.21558594789260804, + "grad_norm": 0.16799384355545044, + "learning_rate": 4.799263554924229e-05, + "loss": 3.1836, + "step": 6989 + }, + { + "epoch": 0.21561679435818146, + "grad_norm": 0.16173897683620453, + "learning_rate": 4.799157853873223e-05, + "loss": 3.2091, + "step": 6990 + }, + { + "epoch": 0.21564764082375487, + "grad_norm": 0.167301744222641, + "learning_rate": 4.7990521261648625e-05, + "loss": 3.2493, + "step": 6991 + }, + { + "epoch": 0.21567848728932829, + "grad_norm": 0.15229755640029907, + "learning_rate": 4.7989463718003745e-05, + "loss": 3.2213, + "step": 6992 + }, + { + "epoch": 0.2157093337549017, + "grad_norm": 0.16411225497722626, + "learning_rate": 4.7988405907809855e-05, + "loss": 3.2498, + "step": 6993 + }, + { + "epoch": 0.21574018022047511, + "grad_norm": 0.1526447832584381, + "learning_rate": 4.79873478310792e-05, + "loss": 3.2257, + "step": 6994 + }, + { + "epoch": 0.21577102668604853, + "grad_norm": 0.16710712015628815, + "learning_rate": 4.7986289487824064e-05, + "loss": 3.2261, + "step": 6995 + }, + { + "epoch": 0.21580187315162194, + "grad_norm": 0.15185001492500305, + "learning_rate": 4.798523087805672e-05, + "loss": 3.247, + "step": 6996 + }, + { + "epoch": 0.21583271961719536, + "grad_norm": 0.15883225202560425, + "learning_rate": 4.798417200178943e-05, + "loss": 3.2335, + "step": 6997 + }, + { + "epoch": 0.21586356608276877, + "grad_norm": 0.15039125084877014, + "learning_rate": 4.798311285903449e-05, + "loss": 3.2411, + "step": 6998 + }, + { + "epoch": 0.21589441254834218, + "grad_norm": 0.4463953673839569, + "learning_rate": 4.7982053449804155e-05, + "loss": 3.2007, + "step": 6999 + }, + { + "epoch": 0.2159252590139156, + "grad_norm": 0.159379780292511, + "learning_rate": 4.798099377411073e-05, + "loss": 3.2009, + "step": 7000 + }, + { + "epoch": 0.21595610547948904, + "grad_norm": 0.15279541909694672, + "learning_rate": 4.797993383196648e-05, + "loss": 3.2497, + "step": 7001 + }, + { + "epoch": 0.21598695194506246, + "grad_norm": 0.15661220252513885, + "learning_rate": 4.797887362338371e-05, + "loss": 3.2191, + "step": 7002 + }, + { + "epoch": 0.21601779841063587, + "grad_norm": 0.16129229962825775, + "learning_rate": 4.7977813148374716e-05, + "loss": 3.2624, + "step": 7003 + }, + { + "epoch": 0.21604864487620928, + "grad_norm": 0.15442818403244019, + "learning_rate": 4.797675240695179e-05, + "loss": 3.2154, + "step": 7004 + }, + { + "epoch": 0.2160794913417827, + "grad_norm": 0.16045229136943817, + "learning_rate": 4.7975691399127214e-05, + "loss": 3.2525, + "step": 7005 + }, + { + "epoch": 0.2161103378073561, + "grad_norm": 0.1586962640285492, + "learning_rate": 4.797463012491331e-05, + "loss": 3.2358, + "step": 7006 + }, + { + "epoch": 0.21614118427292953, + "grad_norm": 0.1555679589509964, + "learning_rate": 4.7973568584322375e-05, + "loss": 3.2282, + "step": 7007 + }, + { + "epoch": 0.21617203073850294, + "grad_norm": 0.16889074444770813, + "learning_rate": 4.797250677736672e-05, + "loss": 3.2595, + "step": 7008 + }, + { + "epoch": 0.21620287720407635, + "grad_norm": 0.15538744628429413, + "learning_rate": 4.797144470405866e-05, + "loss": 3.2302, + "step": 7009 + }, + { + "epoch": 0.21623372366964977, + "grad_norm": 0.15710316598415375, + "learning_rate": 4.797038236441048e-05, + "loss": 3.1973, + "step": 7010 + }, + { + "epoch": 0.21626457013522318, + "grad_norm": 0.15435686707496643, + "learning_rate": 4.796931975843454e-05, + "loss": 3.2072, + "step": 7011 + }, + { + "epoch": 0.2162954166007966, + "grad_norm": 0.1542200744152069, + "learning_rate": 4.7968256886143134e-05, + "loss": 3.2122, + "step": 7012 + }, + { + "epoch": 0.21632626306637004, + "grad_norm": 0.1634698510169983, + "learning_rate": 4.7967193747548593e-05, + "loss": 3.231, + "step": 7013 + }, + { + "epoch": 0.21635710953194345, + "grad_norm": 0.15877307951450348, + "learning_rate": 4.796613034266324e-05, + "loss": 3.2171, + "step": 7014 + }, + { + "epoch": 0.21638795599751687, + "grad_norm": 0.14931657910346985, + "learning_rate": 4.7965066671499416e-05, + "loss": 3.2004, + "step": 7015 + }, + { + "epoch": 0.21641880246309028, + "grad_norm": 0.15879926085472107, + "learning_rate": 4.796400273406944e-05, + "loss": 3.2182, + "step": 7016 + }, + { + "epoch": 0.2164496489286637, + "grad_norm": 0.14557726681232452, + "learning_rate": 4.796293853038564e-05, + "loss": 3.2602, + "step": 7017 + }, + { + "epoch": 0.2164804953942371, + "grad_norm": 0.16934162378311157, + "learning_rate": 4.796187406046038e-05, + "loss": 3.2038, + "step": 7018 + }, + { + "epoch": 0.21651134185981052, + "grad_norm": 0.15360932052135468, + "learning_rate": 4.7960809324305986e-05, + "loss": 3.226, + "step": 7019 + }, + { + "epoch": 0.21654218832538394, + "grad_norm": 0.14916881918907166, + "learning_rate": 4.79597443219348e-05, + "loss": 3.2007, + "step": 7020 + }, + { + "epoch": 0.21657303479095735, + "grad_norm": 0.15892872214317322, + "learning_rate": 4.795867905335918e-05, + "loss": 3.2051, + "step": 7021 + }, + { + "epoch": 0.21660388125653077, + "grad_norm": 0.16280068457126617, + "learning_rate": 4.7957613518591475e-05, + "loss": 3.2648, + "step": 7022 + }, + { + "epoch": 0.21663472772210418, + "grad_norm": 0.1576167643070221, + "learning_rate": 4.795654771764403e-05, + "loss": 3.231, + "step": 7023 + }, + { + "epoch": 0.2166655741876776, + "grad_norm": 0.15895231068134308, + "learning_rate": 4.7955481650529225e-05, + "loss": 3.255, + "step": 7024 + }, + { + "epoch": 0.21669642065325104, + "grad_norm": 0.15620644390583038, + "learning_rate": 4.795441531725939e-05, + "loss": 3.2497, + "step": 7025 + }, + { + "epoch": 0.21672726711882445, + "grad_norm": 0.15548281371593475, + "learning_rate": 4.795334871784691e-05, + "loss": 3.2026, + "step": 7026 + }, + { + "epoch": 0.21675811358439787, + "grad_norm": 0.16667649149894714, + "learning_rate": 4.795228185230414e-05, + "loss": 3.1884, + "step": 7027 + }, + { + "epoch": 0.21678896004997128, + "grad_norm": 0.16211910545825958, + "learning_rate": 4.7951214720643455e-05, + "loss": 3.2159, + "step": 7028 + }, + { + "epoch": 0.2168198065155447, + "grad_norm": 0.15561100840568542, + "learning_rate": 4.795014732287724e-05, + "loss": 3.2255, + "step": 7029 + }, + { + "epoch": 0.2168506529811181, + "grad_norm": 0.1624460518360138, + "learning_rate": 4.7949079659017844e-05, + "loss": 3.165, + "step": 7030 + }, + { + "epoch": 0.21688149944669152, + "grad_norm": 0.15692731738090515, + "learning_rate": 4.7948011729077666e-05, + "loss": 3.2015, + "step": 7031 + }, + { + "epoch": 0.21691234591226494, + "grad_norm": 0.16406142711639404, + "learning_rate": 4.794694353306908e-05, + "loss": 3.1989, + "step": 7032 + }, + { + "epoch": 0.21694319237783835, + "grad_norm": 0.15949755907058716, + "learning_rate": 4.794587507100448e-05, + "loss": 3.1907, + "step": 7033 + }, + { + "epoch": 0.21697403884341177, + "grad_norm": 0.3907559812068939, + "learning_rate": 4.794480634289623e-05, + "loss": 3.2263, + "step": 7034 + }, + { + "epoch": 0.21700488530898518, + "grad_norm": 0.15561428666114807, + "learning_rate": 4.794373734875676e-05, + "loss": 3.2369, + "step": 7035 + }, + { + "epoch": 0.2170357317745586, + "grad_norm": 0.16460493206977844, + "learning_rate": 4.794266808859843e-05, + "loss": 3.2243, + "step": 7036 + }, + { + "epoch": 0.21706657824013204, + "grad_norm": 0.16190071403980255, + "learning_rate": 4.7941598562433654e-05, + "loss": 3.2436, + "step": 7037 + }, + { + "epoch": 0.21709742470570545, + "grad_norm": 0.1650281548500061, + "learning_rate": 4.794052877027483e-05, + "loss": 3.1839, + "step": 7038 + }, + { + "epoch": 0.21712827117127886, + "grad_norm": 0.1579875648021698, + "learning_rate": 4.7939458712134354e-05, + "loss": 3.2353, + "step": 7039 + }, + { + "epoch": 0.21715911763685228, + "grad_norm": 0.15676958858966827, + "learning_rate": 4.793838838802464e-05, + "loss": 3.2096, + "step": 7040 + }, + { + "epoch": 0.2171899641024257, + "grad_norm": 0.15305161476135254, + "learning_rate": 4.79373177979581e-05, + "loss": 3.202, + "step": 7041 + }, + { + "epoch": 0.2172208105679991, + "grad_norm": 0.15691004693508148, + "learning_rate": 4.793624694194715e-05, + "loss": 3.2007, + "step": 7042 + }, + { + "epoch": 0.21725165703357252, + "grad_norm": 0.15793155133724213, + "learning_rate": 4.793517582000419e-05, + "loss": 3.245, + "step": 7043 + }, + { + "epoch": 0.21728250349914593, + "grad_norm": 0.1509120613336563, + "learning_rate": 4.793410443214166e-05, + "loss": 3.2033, + "step": 7044 + }, + { + "epoch": 0.21731334996471935, + "grad_norm": 0.16293534636497498, + "learning_rate": 4.7933032778371966e-05, + "loss": 3.1977, + "step": 7045 + }, + { + "epoch": 0.21734419643029276, + "grad_norm": 0.1704811006784439, + "learning_rate": 4.793196085870754e-05, + "loss": 3.2184, + "step": 7046 + }, + { + "epoch": 0.21737504289586618, + "grad_norm": 0.15642966330051422, + "learning_rate": 4.79308886731608e-05, + "loss": 3.2269, + "step": 7047 + }, + { + "epoch": 0.2174058893614396, + "grad_norm": 0.1668517291545868, + "learning_rate": 4.792981622174419e-05, + "loss": 3.203, + "step": 7048 + }, + { + "epoch": 0.21743673582701303, + "grad_norm": 0.15275241434574127, + "learning_rate": 4.792874350447014e-05, + "loss": 3.2161, + "step": 7049 + }, + { + "epoch": 0.21746758229258645, + "grad_norm": 0.16676212847232819, + "learning_rate": 4.792767052135109e-05, + "loss": 3.2302, + "step": 7050 + }, + { + "epoch": 0.21749842875815986, + "grad_norm": 0.15067850053310394, + "learning_rate": 4.7926597272399475e-05, + "loss": 3.229, + "step": 7051 + }, + { + "epoch": 0.21752927522373328, + "grad_norm": 0.1636485457420349, + "learning_rate": 4.792552375762774e-05, + "loss": 3.2362, + "step": 7052 + }, + { + "epoch": 0.2175601216893067, + "grad_norm": 0.1579631119966507, + "learning_rate": 4.792444997704835e-05, + "loss": 3.2363, + "step": 7053 + }, + { + "epoch": 0.2175909681548801, + "grad_norm": 0.1617586463689804, + "learning_rate": 4.792337593067372e-05, + "loss": 3.2272, + "step": 7054 + }, + { + "epoch": 0.21762181462045352, + "grad_norm": 0.15500354766845703, + "learning_rate": 4.792230161851633e-05, + "loss": 3.2305, + "step": 7055 + }, + { + "epoch": 0.21765266108602693, + "grad_norm": 0.15969429910182953, + "learning_rate": 4.7921227040588625e-05, + "loss": 3.229, + "step": 7056 + }, + { + "epoch": 0.21768350755160035, + "grad_norm": 0.1666918247938156, + "learning_rate": 4.7920152196903066e-05, + "loss": 3.2159, + "step": 7057 + }, + { + "epoch": 0.21771435401717376, + "grad_norm": 0.1597302258014679, + "learning_rate": 4.7919077087472114e-05, + "loss": 3.2482, + "step": 7058 + }, + { + "epoch": 0.21774520048274718, + "grad_norm": 0.15432104468345642, + "learning_rate": 4.791800171230824e-05, + "loss": 3.2018, + "step": 7059 + }, + { + "epoch": 0.2177760469483206, + "grad_norm": 0.16419336199760437, + "learning_rate": 4.7916926071423905e-05, + "loss": 3.2264, + "step": 7060 + }, + { + "epoch": 0.21780689341389403, + "grad_norm": 0.15434694290161133, + "learning_rate": 4.7915850164831585e-05, + "loss": 3.2247, + "step": 7061 + }, + { + "epoch": 0.21783773987946745, + "grad_norm": 0.16151335835456848, + "learning_rate": 4.791477399254375e-05, + "loss": 3.2187, + "step": 7062 + }, + { + "epoch": 0.21786858634504086, + "grad_norm": 0.15799580514431, + "learning_rate": 4.791369755457288e-05, + "loss": 3.1993, + "step": 7063 + }, + { + "epoch": 0.21789943281061427, + "grad_norm": 0.15251825749874115, + "learning_rate": 4.791262085093147e-05, + "loss": 3.2082, + "step": 7064 + }, + { + "epoch": 0.2179302792761877, + "grad_norm": 0.15538471937179565, + "learning_rate": 4.7911543881631974e-05, + "loss": 3.2763, + "step": 7065 + }, + { + "epoch": 0.2179611257417611, + "grad_norm": 0.15322646498680115, + "learning_rate": 4.79104666466869e-05, + "loss": 3.2121, + "step": 7066 + }, + { + "epoch": 0.21799197220733452, + "grad_norm": 0.15618057548999786, + "learning_rate": 4.790938914610874e-05, + "loss": 3.1814, + "step": 7067 + }, + { + "epoch": 0.21802281867290793, + "grad_norm": 0.15460728108882904, + "learning_rate": 4.790831137990997e-05, + "loss": 3.237, + "step": 7068 + }, + { + "epoch": 0.21805366513848135, + "grad_norm": 0.162848100066185, + "learning_rate": 4.7907233348103095e-05, + "loss": 3.2218, + "step": 7069 + }, + { + "epoch": 0.21808451160405476, + "grad_norm": 0.1471758484840393, + "learning_rate": 4.790615505070062e-05, + "loss": 3.2154, + "step": 7070 + }, + { + "epoch": 0.21811535806962817, + "grad_norm": 0.15456975996494293, + "learning_rate": 4.790507648771505e-05, + "loss": 3.224, + "step": 7071 + }, + { + "epoch": 0.2181462045352016, + "grad_norm": 0.15357688069343567, + "learning_rate": 4.790399765915887e-05, + "loss": 3.2189, + "step": 7072 + }, + { + "epoch": 0.21817705100077503, + "grad_norm": 0.15185607969760895, + "learning_rate": 4.790291856504461e-05, + "loss": 3.215, + "step": 7073 + }, + { + "epoch": 0.21820789746634844, + "grad_norm": 0.15361277759075165, + "learning_rate": 4.7901839205384764e-05, + "loss": 3.2305, + "step": 7074 + }, + { + "epoch": 0.21823874393192186, + "grad_norm": 0.15206317603588104, + "learning_rate": 4.7900759580191855e-05, + "loss": 3.2119, + "step": 7075 + }, + { + "epoch": 0.21826959039749527, + "grad_norm": 0.2043314427137375, + "learning_rate": 4.78996796894784e-05, + "loss": 3.2317, + "step": 7076 + }, + { + "epoch": 0.2183004368630687, + "grad_norm": 0.15866005420684814, + "learning_rate": 4.789859953325693e-05, + "loss": 3.243, + "step": 7077 + }, + { + "epoch": 0.2183312833286421, + "grad_norm": 0.15387198328971863, + "learning_rate": 4.7897519111539953e-05, + "loss": 3.2431, + "step": 7078 + }, + { + "epoch": 0.21836212979421551, + "grad_norm": 0.1589370220899582, + "learning_rate": 4.789643842434e-05, + "loss": 3.2457, + "step": 7079 + }, + { + "epoch": 0.21839297625978893, + "grad_norm": 0.15749873220920563, + "learning_rate": 4.78953574716696e-05, + "loss": 3.2298, + "step": 7080 + }, + { + "epoch": 0.21842382272536234, + "grad_norm": 0.16611143946647644, + "learning_rate": 4.7894276253541296e-05, + "loss": 3.265, + "step": 7081 + }, + { + "epoch": 0.21845466919093576, + "grad_norm": 0.15658661723136902, + "learning_rate": 4.7893194769967614e-05, + "loss": 3.2202, + "step": 7082 + }, + { + "epoch": 0.21848551565650917, + "grad_norm": 0.15587438642978668, + "learning_rate": 4.7892113020961095e-05, + "loss": 3.2512, + "step": 7083 + }, + { + "epoch": 0.21851636212208259, + "grad_norm": 0.1646602898836136, + "learning_rate": 4.789103100653428e-05, + "loss": 3.1974, + "step": 7084 + }, + { + "epoch": 0.21854720858765603, + "grad_norm": 0.16669654846191406, + "learning_rate": 4.788994872669972e-05, + "loss": 3.1867, + "step": 7085 + }, + { + "epoch": 0.21857805505322944, + "grad_norm": 0.16991426050662994, + "learning_rate": 4.788886618146996e-05, + "loss": 3.1905, + "step": 7086 + }, + { + "epoch": 0.21860890151880286, + "grad_norm": 0.1560596376657486, + "learning_rate": 4.788778337085755e-05, + "loss": 3.2494, + "step": 7087 + }, + { + "epoch": 0.21863974798437627, + "grad_norm": 0.16607283055782318, + "learning_rate": 4.788670029487504e-05, + "loss": 3.2048, + "step": 7088 + }, + { + "epoch": 0.21867059444994968, + "grad_norm": 0.1585831344127655, + "learning_rate": 4.7885616953535006e-05, + "loss": 3.2262, + "step": 7089 + }, + { + "epoch": 0.2187014409155231, + "grad_norm": 0.16241830587387085, + "learning_rate": 4.7884533346849986e-05, + "loss": 3.2169, + "step": 7090 + }, + { + "epoch": 0.2187322873810965, + "grad_norm": 0.1893136352300644, + "learning_rate": 4.788344947483256e-05, + "loss": 3.2491, + "step": 7091 + }, + { + "epoch": 0.21876313384666993, + "grad_norm": 0.16759338974952698, + "learning_rate": 4.788236533749528e-05, + "loss": 3.2388, + "step": 7092 + }, + { + "epoch": 0.21879398031224334, + "grad_norm": 0.17091578245162964, + "learning_rate": 4.788128093485074e-05, + "loss": 3.247, + "step": 7093 + }, + { + "epoch": 0.21882482677781676, + "grad_norm": 0.17038792371749878, + "learning_rate": 4.788019626691148e-05, + "loss": 3.2603, + "step": 7094 + }, + { + "epoch": 0.21885567324339017, + "grad_norm": 0.15502235293388367, + "learning_rate": 4.7879111333690104e-05, + "loss": 3.2132, + "step": 7095 + }, + { + "epoch": 0.21888651970896358, + "grad_norm": 0.1590675413608551, + "learning_rate": 4.787802613519918e-05, + "loss": 3.2125, + "step": 7096 + }, + { + "epoch": 0.21891736617453703, + "grad_norm": 0.15515241026878357, + "learning_rate": 4.787694067145129e-05, + "loss": 3.2158, + "step": 7097 + }, + { + "epoch": 0.21894821264011044, + "grad_norm": 0.17632433772087097, + "learning_rate": 4.787585494245902e-05, + "loss": 3.2288, + "step": 7098 + }, + { + "epoch": 0.21897905910568385, + "grad_norm": 0.1492018699645996, + "learning_rate": 4.787476894823496e-05, + "loss": 3.2508, + "step": 7099 + }, + { + "epoch": 0.21900990557125727, + "grad_norm": 0.15761740505695343, + "learning_rate": 4.7873682688791695e-05, + "loss": 3.1945, + "step": 7100 + }, + { + "epoch": 0.21904075203683068, + "grad_norm": 0.15432560443878174, + "learning_rate": 4.787259616414183e-05, + "loss": 3.234, + "step": 7101 + }, + { + "epoch": 0.2190715985024041, + "grad_norm": 0.15932178497314453, + "learning_rate": 4.787150937429795e-05, + "loss": 3.2042, + "step": 7102 + }, + { + "epoch": 0.2191024449679775, + "grad_norm": 0.15109337866306305, + "learning_rate": 4.7870422319272666e-05, + "loss": 3.2089, + "step": 7103 + }, + { + "epoch": 0.21913329143355093, + "grad_norm": 0.15263287723064423, + "learning_rate": 4.7869334999078586e-05, + "loss": 3.2396, + "step": 7104 + }, + { + "epoch": 0.21916413789912434, + "grad_norm": 0.15535330772399902, + "learning_rate": 4.786824741372831e-05, + "loss": 3.2102, + "step": 7105 + }, + { + "epoch": 0.21919498436469775, + "grad_norm": 0.16002865135669708, + "learning_rate": 4.786715956323444e-05, + "loss": 3.2253, + "step": 7106 + }, + { + "epoch": 0.21922583083027117, + "grad_norm": 0.15313221514225006, + "learning_rate": 4.7866071447609594e-05, + "loss": 3.2001, + "step": 7107 + }, + { + "epoch": 0.21925667729584458, + "grad_norm": 0.1529688984155655, + "learning_rate": 4.78649830668664e-05, + "loss": 3.2291, + "step": 7108 + }, + { + "epoch": 0.21928752376141802, + "grad_norm": 0.22345948219299316, + "learning_rate": 4.786389442101746e-05, + "loss": 3.2331, + "step": 7109 + }, + { + "epoch": 0.21931837022699144, + "grad_norm": 0.16051095724105835, + "learning_rate": 4.7862805510075406e-05, + "loss": 3.2835, + "step": 7110 + }, + { + "epoch": 0.21934921669256485, + "grad_norm": 0.15679779648780823, + "learning_rate": 4.786171633405287e-05, + "loss": 3.2639, + "step": 7111 + }, + { + "epoch": 0.21938006315813827, + "grad_norm": 0.16132254898548126, + "learning_rate": 4.786062689296246e-05, + "loss": 3.2151, + "step": 7112 + }, + { + "epoch": 0.21941090962371168, + "grad_norm": 0.1593177616596222, + "learning_rate": 4.7859537186816816e-05, + "loss": 3.2247, + "step": 7113 + }, + { + "epoch": 0.2194417560892851, + "grad_norm": 0.16023509204387665, + "learning_rate": 4.7858447215628585e-05, + "loss": 3.2385, + "step": 7114 + }, + { + "epoch": 0.2194726025548585, + "grad_norm": 0.15876160562038422, + "learning_rate": 4.785735697941039e-05, + "loss": 3.2292, + "step": 7115 + }, + { + "epoch": 0.21950344902043192, + "grad_norm": 0.15857045352458954, + "learning_rate": 4.785626647817487e-05, + "loss": 3.2172, + "step": 7116 + }, + { + "epoch": 0.21953429548600534, + "grad_norm": 0.1573854386806488, + "learning_rate": 4.785517571193468e-05, + "loss": 3.2426, + "step": 7117 + }, + { + "epoch": 0.21956514195157875, + "grad_norm": 0.16220863163471222, + "learning_rate": 4.785408468070247e-05, + "loss": 3.2146, + "step": 7118 + }, + { + "epoch": 0.21959598841715217, + "grad_norm": 0.1619555801153183, + "learning_rate": 4.785299338449088e-05, + "loss": 3.2735, + "step": 7119 + }, + { + "epoch": 0.21962683488272558, + "grad_norm": 0.17095091938972473, + "learning_rate": 4.785190182331256e-05, + "loss": 3.2054, + "step": 7120 + }, + { + "epoch": 0.21965768134829902, + "grad_norm": 0.15477792918682098, + "learning_rate": 4.7850809997180166e-05, + "loss": 3.2131, + "step": 7121 + }, + { + "epoch": 0.21968852781387244, + "grad_norm": 0.1909385770559311, + "learning_rate": 4.784971790610637e-05, + "loss": 3.2413, + "step": 7122 + }, + { + "epoch": 0.21971937427944585, + "grad_norm": 0.1757558435201645, + "learning_rate": 4.784862555010383e-05, + "loss": 3.2644, + "step": 7123 + }, + { + "epoch": 0.21975022074501926, + "grad_norm": 0.15933017432689667, + "learning_rate": 4.78475329291852e-05, + "loss": 3.2237, + "step": 7124 + }, + { + "epoch": 0.21978106721059268, + "grad_norm": 0.15823602676391602, + "learning_rate": 4.784644004336315e-05, + "loss": 3.2177, + "step": 7125 + }, + { + "epoch": 0.2198119136761661, + "grad_norm": 0.16513487696647644, + "learning_rate": 4.784534689265037e-05, + "loss": 3.2259, + "step": 7126 + }, + { + "epoch": 0.2198427601417395, + "grad_norm": 0.3247344195842743, + "learning_rate": 4.784425347705952e-05, + "loss": 3.2619, + "step": 7127 + }, + { + "epoch": 0.21987360660731292, + "grad_norm": 0.16202548146247864, + "learning_rate": 4.7843159796603266e-05, + "loss": 3.1717, + "step": 7128 + }, + { + "epoch": 0.21990445307288634, + "grad_norm": 0.1597333550453186, + "learning_rate": 4.784206585129431e-05, + "loss": 3.2298, + "step": 7129 + }, + { + "epoch": 0.21993529953845975, + "grad_norm": 0.1608469933271408, + "learning_rate": 4.7840971641145325e-05, + "loss": 3.244, + "step": 7130 + }, + { + "epoch": 0.21996614600403316, + "grad_norm": 0.16188134253025055, + "learning_rate": 4.7839877166169e-05, + "loss": 3.2585, + "step": 7131 + }, + { + "epoch": 0.21999699246960658, + "grad_norm": 0.15783391892910004, + "learning_rate": 4.7838782426378026e-05, + "loss": 3.2131, + "step": 7132 + }, + { + "epoch": 0.22002783893518002, + "grad_norm": 0.1593771129846573, + "learning_rate": 4.783768742178509e-05, + "loss": 3.2275, + "step": 7133 + }, + { + "epoch": 0.22005868540075343, + "grad_norm": 0.15632575750350952, + "learning_rate": 4.783659215240289e-05, + "loss": 3.212, + "step": 7134 + }, + { + "epoch": 0.22008953186632685, + "grad_norm": 0.15698237717151642, + "learning_rate": 4.783549661824413e-05, + "loss": 3.2507, + "step": 7135 + }, + { + "epoch": 0.22012037833190026, + "grad_norm": 0.16435544192790985, + "learning_rate": 4.783440081932152e-05, + "loss": 3.2196, + "step": 7136 + }, + { + "epoch": 0.22015122479747368, + "grad_norm": 0.15014784038066864, + "learning_rate": 4.783330475564775e-05, + "loss": 3.2261, + "step": 7137 + }, + { + "epoch": 0.2201820712630471, + "grad_norm": 0.1619347631931305, + "learning_rate": 4.7832208427235525e-05, + "loss": 3.1989, + "step": 7138 + }, + { + "epoch": 0.2202129177286205, + "grad_norm": 0.15470445156097412, + "learning_rate": 4.783111183409756e-05, + "loss": 3.1828, + "step": 7139 + }, + { + "epoch": 0.22024376419419392, + "grad_norm": 0.16378329694271088, + "learning_rate": 4.783001497624658e-05, + "loss": 3.2324, + "step": 7140 + }, + { + "epoch": 0.22027461065976733, + "grad_norm": 0.15569479763507843, + "learning_rate": 4.782891785369531e-05, + "loss": 3.2182, + "step": 7141 + }, + { + "epoch": 0.22030545712534075, + "grad_norm": 0.20637449622154236, + "learning_rate": 4.7827820466456434e-05, + "loss": 3.2378, + "step": 7142 + }, + { + "epoch": 0.22033630359091416, + "grad_norm": 0.16108179092407227, + "learning_rate": 4.782672281454271e-05, + "loss": 3.2379, + "step": 7143 + }, + { + "epoch": 0.22036715005648758, + "grad_norm": 0.15124207735061646, + "learning_rate": 4.7825624897966845e-05, + "loss": 3.2253, + "step": 7144 + }, + { + "epoch": 0.22039799652206102, + "grad_norm": 0.1621723473072052, + "learning_rate": 4.782452671674158e-05, + "loss": 3.2541, + "step": 7145 + }, + { + "epoch": 0.22042884298763443, + "grad_norm": 0.1633932888507843, + "learning_rate": 4.782342827087964e-05, + "loss": 3.2229, + "step": 7146 + }, + { + "epoch": 0.22045968945320785, + "grad_norm": 0.1603749692440033, + "learning_rate": 4.782232956039377e-05, + "loss": 3.2141, + "step": 7147 + }, + { + "epoch": 0.22049053591878126, + "grad_norm": 0.16096819937229156, + "learning_rate": 4.78212305852967e-05, + "loss": 3.2222, + "step": 7148 + }, + { + "epoch": 0.22052138238435467, + "grad_norm": 0.1521308422088623, + "learning_rate": 4.782013134560117e-05, + "loss": 3.2078, + "step": 7149 + }, + { + "epoch": 0.2205522288499281, + "grad_norm": 0.17069105803966522, + "learning_rate": 4.7819031841319936e-05, + "loss": 3.1985, + "step": 7150 + }, + { + "epoch": 0.2205830753155015, + "grad_norm": 0.1515609622001648, + "learning_rate": 4.7817932072465745e-05, + "loss": 3.1985, + "step": 7151 + }, + { + "epoch": 0.22061392178107492, + "grad_norm": 0.15751929581165314, + "learning_rate": 4.781683203905134e-05, + "loss": 3.2403, + "step": 7152 + }, + { + "epoch": 0.22064476824664833, + "grad_norm": 0.1560620665550232, + "learning_rate": 4.7815731741089476e-05, + "loss": 3.1923, + "step": 7153 + }, + { + "epoch": 0.22067561471222175, + "grad_norm": 0.1531316488981247, + "learning_rate": 4.781463117859292e-05, + "loss": 3.1904, + "step": 7154 + }, + { + "epoch": 0.22070646117779516, + "grad_norm": 0.3465610444545746, + "learning_rate": 4.781353035157442e-05, + "loss": 3.2372, + "step": 7155 + }, + { + "epoch": 0.2207373076433686, + "grad_norm": 0.1533123105764389, + "learning_rate": 4.781242926004674e-05, + "loss": 3.1544, + "step": 7156 + }, + { + "epoch": 0.22076815410894202, + "grad_norm": 0.1559496521949768, + "learning_rate": 4.781132790402266e-05, + "loss": 3.2474, + "step": 7157 + }, + { + "epoch": 0.22079900057451543, + "grad_norm": 0.15853388607501984, + "learning_rate": 4.7810226283514934e-05, + "loss": 3.2307, + "step": 7158 + }, + { + "epoch": 0.22082984704008884, + "grad_norm": 0.15606804192066193, + "learning_rate": 4.7809124398536356e-05, + "loss": 3.2078, + "step": 7159 + }, + { + "epoch": 0.22086069350566226, + "grad_norm": 0.15871171653270721, + "learning_rate": 4.780802224909967e-05, + "loss": 3.2257, + "step": 7160 + }, + { + "epoch": 0.22089153997123567, + "grad_norm": 0.15376056730747223, + "learning_rate": 4.780691983521768e-05, + "loss": 3.2126, + "step": 7161 + }, + { + "epoch": 0.2209223864368091, + "grad_norm": 0.16019316017627716, + "learning_rate": 4.780581715690315e-05, + "loss": 3.2499, + "step": 7162 + }, + { + "epoch": 0.2209532329023825, + "grad_norm": 0.15108108520507812, + "learning_rate": 4.780471421416889e-05, + "loss": 3.2029, + "step": 7163 + }, + { + "epoch": 0.22098407936795592, + "grad_norm": 0.16086003184318542, + "learning_rate": 4.780361100702767e-05, + "loss": 3.2154, + "step": 7164 + }, + { + "epoch": 0.22101492583352933, + "grad_norm": 0.1552462875843048, + "learning_rate": 4.780250753549228e-05, + "loss": 3.1969, + "step": 7165 + }, + { + "epoch": 0.22104577229910274, + "grad_norm": 0.15413635969161987, + "learning_rate": 4.7801403799575515e-05, + "loss": 3.2216, + "step": 7166 + }, + { + "epoch": 0.22107661876467616, + "grad_norm": 0.15160366892814636, + "learning_rate": 4.780029979929018e-05, + "loss": 3.2312, + "step": 7167 + }, + { + "epoch": 0.2211074652302496, + "grad_norm": 0.15160788595676422, + "learning_rate": 4.779919553464907e-05, + "loss": 3.2134, + "step": 7168 + }, + { + "epoch": 0.22113831169582301, + "grad_norm": 0.15983031690120697, + "learning_rate": 4.779809100566498e-05, + "loss": 3.251, + "step": 7169 + }, + { + "epoch": 0.22116915816139643, + "grad_norm": 0.1522379219532013, + "learning_rate": 4.779698621235074e-05, + "loss": 3.2305, + "step": 7170 + }, + { + "epoch": 0.22120000462696984, + "grad_norm": 0.15407980978488922, + "learning_rate": 4.779588115471913e-05, + "loss": 3.2277, + "step": 7171 + }, + { + "epoch": 0.22123085109254326, + "grad_norm": 0.17400915920734406, + "learning_rate": 4.779477583278298e-05, + "loss": 3.1825, + "step": 7172 + }, + { + "epoch": 0.22126169755811667, + "grad_norm": 0.15421105921268463, + "learning_rate": 4.77936702465551e-05, + "loss": 3.2495, + "step": 7173 + }, + { + "epoch": 0.22129254402369009, + "grad_norm": 0.1569949984550476, + "learning_rate": 4.779256439604832e-05, + "loss": 3.1951, + "step": 7174 + }, + { + "epoch": 0.2213233904892635, + "grad_norm": 0.15320290625095367, + "learning_rate": 4.779145828127545e-05, + "loss": 3.2185, + "step": 7175 + }, + { + "epoch": 0.2213542369548369, + "grad_norm": 0.16154798865318298, + "learning_rate": 4.779035190224932e-05, + "loss": 3.2506, + "step": 7176 + }, + { + "epoch": 0.22138508342041033, + "grad_norm": 0.15501974523067474, + "learning_rate": 4.7789245258982746e-05, + "loss": 3.2449, + "step": 7177 + }, + { + "epoch": 0.22141592988598374, + "grad_norm": 0.15037648379802704, + "learning_rate": 4.7788138351488574e-05, + "loss": 3.2307, + "step": 7178 + }, + { + "epoch": 0.22144677635155716, + "grad_norm": 0.15990521013736725, + "learning_rate": 4.7787031179779626e-05, + "loss": 3.1996, + "step": 7179 + }, + { + "epoch": 0.2214776228171306, + "grad_norm": 0.15522749722003937, + "learning_rate": 4.7785923743868755e-05, + "loss": 3.2473, + "step": 7180 + }, + { + "epoch": 0.221508469282704, + "grad_norm": 0.16706615686416626, + "learning_rate": 4.778481604376878e-05, + "loss": 3.2219, + "step": 7181 + }, + { + "epoch": 0.22153931574827743, + "grad_norm": 0.1528584212064743, + "learning_rate": 4.778370807949256e-05, + "loss": 3.2556, + "step": 7182 + }, + { + "epoch": 0.22157016221385084, + "grad_norm": 0.1524006873369217, + "learning_rate": 4.778259985105294e-05, + "loss": 3.2596, + "step": 7183 + }, + { + "epoch": 0.22160100867942426, + "grad_norm": 0.16828356683254242, + "learning_rate": 4.778149135846276e-05, + "loss": 3.199, + "step": 7184 + }, + { + "epoch": 0.22163185514499767, + "grad_norm": 0.1537332981824875, + "learning_rate": 4.778038260173489e-05, + "loss": 3.2274, + "step": 7185 + }, + { + "epoch": 0.22166270161057108, + "grad_norm": 0.1613777130842209, + "learning_rate": 4.7779273580882156e-05, + "loss": 3.2303, + "step": 7186 + }, + { + "epoch": 0.2216935480761445, + "grad_norm": 0.15996482968330383, + "learning_rate": 4.7778164295917436e-05, + "loss": 3.2364, + "step": 7187 + }, + { + "epoch": 0.2217243945417179, + "grad_norm": 0.15850120782852173, + "learning_rate": 4.77770547468536e-05, + "loss": 3.1836, + "step": 7188 + }, + { + "epoch": 0.22175524100729133, + "grad_norm": 0.1528140753507614, + "learning_rate": 4.777594493370349e-05, + "loss": 3.2072, + "step": 7189 + }, + { + "epoch": 0.22178608747286474, + "grad_norm": 0.15362930297851562, + "learning_rate": 4.777483485647999e-05, + "loss": 3.2038, + "step": 7190 + }, + { + "epoch": 0.22181693393843815, + "grad_norm": 0.15194763243198395, + "learning_rate": 4.777372451519597e-05, + "loss": 3.2167, + "step": 7191 + }, + { + "epoch": 0.2218477804040116, + "grad_norm": 0.1517990678548813, + "learning_rate": 4.777261390986429e-05, + "loss": 3.2145, + "step": 7192 + }, + { + "epoch": 0.221878626869585, + "grad_norm": 0.15141665935516357, + "learning_rate": 4.777150304049785e-05, + "loss": 3.2177, + "step": 7193 + }, + { + "epoch": 0.22190947333515842, + "grad_norm": 0.15079578757286072, + "learning_rate": 4.77703919071095e-05, + "loss": 3.2418, + "step": 7194 + }, + { + "epoch": 0.22194031980073184, + "grad_norm": 0.14997541904449463, + "learning_rate": 4.7769280509712154e-05, + "loss": 3.223, + "step": 7195 + }, + { + "epoch": 0.22197116626630525, + "grad_norm": 0.15763793885707855, + "learning_rate": 4.7768168848318676e-05, + "loss": 3.2233, + "step": 7196 + }, + { + "epoch": 0.22200201273187867, + "grad_norm": 0.14814795553684235, + "learning_rate": 4.7767056922941965e-05, + "loss": 3.2168, + "step": 7197 + }, + { + "epoch": 0.22203285919745208, + "grad_norm": 0.18048718571662903, + "learning_rate": 4.776594473359491e-05, + "loss": 3.2428, + "step": 7198 + }, + { + "epoch": 0.2220637056630255, + "grad_norm": 0.15936367213726044, + "learning_rate": 4.776483228029041e-05, + "loss": 3.2309, + "step": 7199 + }, + { + "epoch": 0.2220945521285989, + "grad_norm": 0.1600257009267807, + "learning_rate": 4.776371956304136e-05, + "loss": 3.2111, + "step": 7200 + }, + { + "epoch": 0.22212539859417232, + "grad_norm": 0.15534619987010956, + "learning_rate": 4.776260658186065e-05, + "loss": 3.2161, + "step": 7201 + }, + { + "epoch": 0.22215624505974574, + "grad_norm": 0.1617128998041153, + "learning_rate": 4.77614933367612e-05, + "loss": 3.2364, + "step": 7202 + }, + { + "epoch": 0.22218709152531915, + "grad_norm": 0.15263518691062927, + "learning_rate": 4.776037982775591e-05, + "loss": 3.2183, + "step": 7203 + }, + { + "epoch": 0.2222179379908926, + "grad_norm": 0.1625336855649948, + "learning_rate": 4.77592660548577e-05, + "loss": 3.2068, + "step": 7204 + }, + { + "epoch": 0.222248784456466, + "grad_norm": 0.29098600149154663, + "learning_rate": 4.775815201807947e-05, + "loss": 3.2109, + "step": 7205 + }, + { + "epoch": 0.22227963092203942, + "grad_norm": 0.16015931963920593, + "learning_rate": 4.7757037717434147e-05, + "loss": 3.223, + "step": 7206 + }, + { + "epoch": 0.22231047738761284, + "grad_norm": 0.16198112070560455, + "learning_rate": 4.7755923152934656e-05, + "loss": 3.2214, + "step": 7207 + }, + { + "epoch": 0.22234132385318625, + "grad_norm": 0.1574772298336029, + "learning_rate": 4.77548083245939e-05, + "loss": 3.2365, + "step": 7208 + }, + { + "epoch": 0.22237217031875967, + "grad_norm": 0.1697663962841034, + "learning_rate": 4.775369323242482e-05, + "loss": 3.221, + "step": 7209 + }, + { + "epoch": 0.22240301678433308, + "grad_norm": 0.1574689745903015, + "learning_rate": 4.7752577876440343e-05, + "loss": 3.2155, + "step": 7210 + }, + { + "epoch": 0.2224338632499065, + "grad_norm": 0.15904776751995087, + "learning_rate": 4.775146225665339e-05, + "loss": 3.2164, + "step": 7211 + }, + { + "epoch": 0.2224647097154799, + "grad_norm": 0.16665883362293243, + "learning_rate": 4.7750346373076903e-05, + "loss": 3.2611, + "step": 7212 + }, + { + "epoch": 0.22249555618105332, + "grad_norm": 0.15871545672416687, + "learning_rate": 4.774923022572383e-05, + "loss": 3.2389, + "step": 7213 + }, + { + "epoch": 0.22252640264662674, + "grad_norm": 0.16132865846157074, + "learning_rate": 4.7748113814607096e-05, + "loss": 3.1839, + "step": 7214 + }, + { + "epoch": 0.22255724911220015, + "grad_norm": 0.1582946628332138, + "learning_rate": 4.774699713973966e-05, + "loss": 3.2292, + "step": 7215 + }, + { + "epoch": 0.2225880955777736, + "grad_norm": 0.15449075400829315, + "learning_rate": 4.774588020113446e-05, + "loss": 3.189, + "step": 7216 + }, + { + "epoch": 0.222618942043347, + "grad_norm": 0.15899984538555145, + "learning_rate": 4.7744762998804446e-05, + "loss": 3.2256, + "step": 7217 + }, + { + "epoch": 0.22264978850892042, + "grad_norm": 0.16481570899486542, + "learning_rate": 4.774364553276257e-05, + "loss": 3.2361, + "step": 7218 + }, + { + "epoch": 0.22268063497449384, + "grad_norm": 0.15978088974952698, + "learning_rate": 4.77425278030218e-05, + "loss": 3.2012, + "step": 7219 + }, + { + "epoch": 0.22271148144006725, + "grad_norm": 0.16240336000919342, + "learning_rate": 4.7741409809595075e-05, + "loss": 3.2258, + "step": 7220 + }, + { + "epoch": 0.22274232790564066, + "grad_norm": 0.16315124928951263, + "learning_rate": 4.7740291552495376e-05, + "loss": 3.2328, + "step": 7221 + }, + { + "epoch": 0.22277317437121408, + "grad_norm": 0.16106252372264862, + "learning_rate": 4.773917303173566e-05, + "loss": 3.2472, + "step": 7222 + }, + { + "epoch": 0.2228040208367875, + "grad_norm": 0.16022560000419617, + "learning_rate": 4.77380542473289e-05, + "loss": 3.2501, + "step": 7223 + }, + { + "epoch": 0.2228348673023609, + "grad_norm": 0.1599309742450714, + "learning_rate": 4.773693519928806e-05, + "loss": 3.2169, + "step": 7224 + }, + { + "epoch": 0.22286571376793432, + "grad_norm": 0.18278324604034424, + "learning_rate": 4.7735815887626126e-05, + "loss": 3.2162, + "step": 7225 + }, + { + "epoch": 0.22289656023350773, + "grad_norm": 0.16192127764225006, + "learning_rate": 4.773469631235606e-05, + "loss": 3.24, + "step": 7226 + }, + { + "epoch": 0.22292740669908115, + "grad_norm": 0.2057877779006958, + "learning_rate": 4.773357647349086e-05, + "loss": 3.2091, + "step": 7227 + }, + { + "epoch": 0.2229582531646546, + "grad_norm": 0.16257384419441223, + "learning_rate": 4.77324563710435e-05, + "loss": 3.2314, + "step": 7228 + }, + { + "epoch": 0.222989099630228, + "grad_norm": 0.15021122992038727, + "learning_rate": 4.773133600502697e-05, + "loss": 3.2349, + "step": 7229 + }, + { + "epoch": 0.22301994609580142, + "grad_norm": 0.15484769642353058, + "learning_rate": 4.773021537545426e-05, + "loss": 3.2035, + "step": 7230 + }, + { + "epoch": 0.22305079256137483, + "grad_norm": 0.15410442650318146, + "learning_rate": 4.772909448233835e-05, + "loss": 3.2232, + "step": 7231 + }, + { + "epoch": 0.22308163902694825, + "grad_norm": 0.1570543646812439, + "learning_rate": 4.7727973325692266e-05, + "loss": 3.2284, + "step": 7232 + }, + { + "epoch": 0.22311248549252166, + "grad_norm": 0.15323962271213531, + "learning_rate": 4.772685190552898e-05, + "loss": 3.208, + "step": 7233 + }, + { + "epoch": 0.22314333195809508, + "grad_norm": 0.15227656066417694, + "learning_rate": 4.772573022186151e-05, + "loss": 3.2191, + "step": 7234 + }, + { + "epoch": 0.2231741784236685, + "grad_norm": 0.15872690081596375, + "learning_rate": 4.772460827470285e-05, + "loss": 3.1959, + "step": 7235 + }, + { + "epoch": 0.2232050248892419, + "grad_norm": 0.16116419434547424, + "learning_rate": 4.772348606406601e-05, + "loss": 3.2122, + "step": 7236 + }, + { + "epoch": 0.22323587135481532, + "grad_norm": 0.2150016874074936, + "learning_rate": 4.7722363589964005e-05, + "loss": 3.232, + "step": 7237 + }, + { + "epoch": 0.22326671782038873, + "grad_norm": 0.1614050716161728, + "learning_rate": 4.7721240852409854e-05, + "loss": 3.2586, + "step": 7238 + }, + { + "epoch": 0.22329756428596215, + "grad_norm": 0.15865285694599152, + "learning_rate": 4.772011785141657e-05, + "loss": 3.2319, + "step": 7239 + }, + { + "epoch": 0.2233284107515356, + "grad_norm": 0.1597650796175003, + "learning_rate": 4.7718994586997165e-05, + "loss": 3.1971, + "step": 7240 + }, + { + "epoch": 0.223359257217109, + "grad_norm": 0.16710549592971802, + "learning_rate": 4.771787105916467e-05, + "loss": 3.2399, + "step": 7241 + }, + { + "epoch": 0.22339010368268242, + "grad_norm": 0.1652577668428421, + "learning_rate": 4.771674726793212e-05, + "loss": 3.2347, + "step": 7242 + }, + { + "epoch": 0.22342095014825583, + "grad_norm": 0.15690737962722778, + "learning_rate": 4.771562321331253e-05, + "loss": 3.2257, + "step": 7243 + }, + { + "epoch": 0.22345179661382925, + "grad_norm": 0.15705008804798126, + "learning_rate": 4.7714498895318934e-05, + "loss": 3.1996, + "step": 7244 + }, + { + "epoch": 0.22348264307940266, + "grad_norm": 0.16013750433921814, + "learning_rate": 4.771337431396439e-05, + "loss": 3.2292, + "step": 7245 + }, + { + "epoch": 0.22351348954497607, + "grad_norm": 0.15174081921577454, + "learning_rate": 4.7712249469261905e-05, + "loss": 3.2175, + "step": 7246 + }, + { + "epoch": 0.2235443360105495, + "grad_norm": 0.17588302493095398, + "learning_rate": 4.771112436122455e-05, + "loss": 3.2201, + "step": 7247 + }, + { + "epoch": 0.2235751824761229, + "grad_norm": 0.15712039172649384, + "learning_rate": 4.7709998989865345e-05, + "loss": 3.2308, + "step": 7248 + }, + { + "epoch": 0.22360602894169632, + "grad_norm": 0.19179478287696838, + "learning_rate": 4.770887335519735e-05, + "loss": 3.2279, + "step": 7249 + }, + { + "epoch": 0.22363687540726973, + "grad_norm": 0.1526995748281479, + "learning_rate": 4.770774745723362e-05, + "loss": 3.2403, + "step": 7250 + }, + { + "epoch": 0.22366772187284314, + "grad_norm": 0.17461104691028595, + "learning_rate": 4.77066212959872e-05, + "loss": 3.265, + "step": 7251 + }, + { + "epoch": 0.2236985683384166, + "grad_norm": 0.15557216107845306, + "learning_rate": 4.7705494871471145e-05, + "loss": 3.205, + "step": 7252 + }, + { + "epoch": 0.22372941480399, + "grad_norm": 0.2129773199558258, + "learning_rate": 4.770436818369853e-05, + "loss": 3.2049, + "step": 7253 + }, + { + "epoch": 0.22376026126956342, + "grad_norm": 0.15124356746673584, + "learning_rate": 4.77032412326824e-05, + "loss": 3.2035, + "step": 7254 + }, + { + "epoch": 0.22379110773513683, + "grad_norm": 0.3262666165828705, + "learning_rate": 4.7702114018435845e-05, + "loss": 3.2308, + "step": 7255 + }, + { + "epoch": 0.22382195420071024, + "grad_norm": 0.15441501140594482, + "learning_rate": 4.7700986540971915e-05, + "loss": 3.2109, + "step": 7256 + }, + { + "epoch": 0.22385280066628366, + "grad_norm": 0.16302110254764557, + "learning_rate": 4.769985880030369e-05, + "loss": 3.2301, + "step": 7257 + }, + { + "epoch": 0.22388364713185707, + "grad_norm": 0.1576215922832489, + "learning_rate": 4.7698730796444236e-05, + "loss": 3.2356, + "step": 7258 + }, + { + "epoch": 0.2239144935974305, + "grad_norm": 0.17025253176689148, + "learning_rate": 4.7697602529406645e-05, + "loss": 3.2399, + "step": 7259 + }, + { + "epoch": 0.2239453400630039, + "grad_norm": 0.15384414792060852, + "learning_rate": 4.769647399920399e-05, + "loss": 3.2338, + "step": 7260 + }, + { + "epoch": 0.22397618652857731, + "grad_norm": 0.16437511146068573, + "learning_rate": 4.769534520584935e-05, + "loss": 3.247, + "step": 7261 + }, + { + "epoch": 0.22400703299415073, + "grad_norm": 0.33131691813468933, + "learning_rate": 4.769421614935583e-05, + "loss": 3.219, + "step": 7262 + }, + { + "epoch": 0.22403787945972414, + "grad_norm": 0.16388210654258728, + "learning_rate": 4.769308682973651e-05, + "loss": 3.2255, + "step": 7263 + }, + { + "epoch": 0.22406872592529758, + "grad_norm": 0.1620943248271942, + "learning_rate": 4.769195724700449e-05, + "loss": 3.2443, + "step": 7264 + }, + { + "epoch": 0.224099572390871, + "grad_norm": 0.18211784958839417, + "learning_rate": 4.769082740117286e-05, + "loss": 3.2053, + "step": 7265 + }, + { + "epoch": 0.2241304188564444, + "grad_norm": 0.17286035418510437, + "learning_rate": 4.7689697292254716e-05, + "loss": 3.2269, + "step": 7266 + }, + { + "epoch": 0.22416126532201783, + "grad_norm": 0.15400874614715576, + "learning_rate": 4.7688566920263164e-05, + "loss": 3.1868, + "step": 7267 + }, + { + "epoch": 0.22419211178759124, + "grad_norm": 0.16602759063243866, + "learning_rate": 4.768743628521132e-05, + "loss": 3.2366, + "step": 7268 + }, + { + "epoch": 0.22422295825316466, + "grad_norm": 0.15602707862854004, + "learning_rate": 4.768630538711228e-05, + "loss": 3.2022, + "step": 7269 + }, + { + "epoch": 0.22425380471873807, + "grad_norm": 0.17141804099082947, + "learning_rate": 4.768517422597917e-05, + "loss": 3.2124, + "step": 7270 + }, + { + "epoch": 0.22428465118431148, + "grad_norm": 0.15144114196300507, + "learning_rate": 4.76840428018251e-05, + "loss": 3.2267, + "step": 7271 + }, + { + "epoch": 0.2243154976498849, + "grad_norm": 0.1644364595413208, + "learning_rate": 4.7682911114663176e-05, + "loss": 3.217, + "step": 7272 + }, + { + "epoch": 0.2243463441154583, + "grad_norm": 0.1517351120710373, + "learning_rate": 4.7681779164506535e-05, + "loss": 3.2062, + "step": 7273 + }, + { + "epoch": 0.22437719058103173, + "grad_norm": 0.17476381361484528, + "learning_rate": 4.768064695136829e-05, + "loss": 3.21, + "step": 7274 + }, + { + "epoch": 0.22440803704660514, + "grad_norm": 0.15955117344856262, + "learning_rate": 4.767951447526157e-05, + "loss": 3.2385, + "step": 7275 + }, + { + "epoch": 0.22443888351217858, + "grad_norm": 0.20263542234897614, + "learning_rate": 4.7678381736199515e-05, + "loss": 3.2103, + "step": 7276 + }, + { + "epoch": 0.224469729977752, + "grad_norm": 0.16027706861495972, + "learning_rate": 4.7677248734195246e-05, + "loss": 3.2138, + "step": 7277 + }, + { + "epoch": 0.2245005764433254, + "grad_norm": 0.15769696235656738, + "learning_rate": 4.767611546926191e-05, + "loss": 3.2141, + "step": 7278 + }, + { + "epoch": 0.22453142290889883, + "grad_norm": 0.1562851071357727, + "learning_rate": 4.7674981941412646e-05, + "loss": 3.2243, + "step": 7279 + }, + { + "epoch": 0.22456226937447224, + "grad_norm": 0.15628257393836975, + "learning_rate": 4.767384815066058e-05, + "loss": 3.2064, + "step": 7280 + }, + { + "epoch": 0.22459311584004565, + "grad_norm": 0.1602841019630432, + "learning_rate": 4.7672714097018876e-05, + "loss": 3.2121, + "step": 7281 + }, + { + "epoch": 0.22462396230561907, + "grad_norm": 0.15321923792362213, + "learning_rate": 4.767157978050068e-05, + "loss": 3.2051, + "step": 7282 + }, + { + "epoch": 0.22465480877119248, + "grad_norm": 0.14996500313282013, + "learning_rate": 4.7670445201119144e-05, + "loss": 3.2036, + "step": 7283 + }, + { + "epoch": 0.2246856552367659, + "grad_norm": 0.37372279167175293, + "learning_rate": 4.766931035888741e-05, + "loss": 3.2462, + "step": 7284 + }, + { + "epoch": 0.2247165017023393, + "grad_norm": 0.16111019253730774, + "learning_rate": 4.766817525381865e-05, + "loss": 3.247, + "step": 7285 + }, + { + "epoch": 0.22474734816791272, + "grad_norm": 0.16644783318042755, + "learning_rate": 4.766703988592601e-05, + "loss": 3.1987, + "step": 7286 + }, + { + "epoch": 0.22477819463348614, + "grad_norm": 0.15417148172855377, + "learning_rate": 4.7665904255222673e-05, + "loss": 3.2307, + "step": 7287 + }, + { + "epoch": 0.22480904109905958, + "grad_norm": 0.1583309769630432, + "learning_rate": 4.76647683617218e-05, + "loss": 3.2358, + "step": 7288 + }, + { + "epoch": 0.224839887564633, + "grad_norm": 0.1630466878414154, + "learning_rate": 4.766363220543656e-05, + "loss": 3.2455, + "step": 7289 + }, + { + "epoch": 0.2248707340302064, + "grad_norm": 0.16090843081474304, + "learning_rate": 4.7662495786380116e-05, + "loss": 3.2533, + "step": 7290 + }, + { + "epoch": 0.22490158049577982, + "grad_norm": 0.15113915503025055, + "learning_rate": 4.7661359104565653e-05, + "loss": 3.1967, + "step": 7291 + }, + { + "epoch": 0.22493242696135324, + "grad_norm": 0.16634972393512726, + "learning_rate": 4.766022216000635e-05, + "loss": 3.222, + "step": 7292 + }, + { + "epoch": 0.22496327342692665, + "grad_norm": 0.15141525864601135, + "learning_rate": 4.765908495271538e-05, + "loss": 3.1935, + "step": 7293 + }, + { + "epoch": 0.22499411989250007, + "grad_norm": 0.15747851133346558, + "learning_rate": 4.765794748270595e-05, + "loss": 3.2203, + "step": 7294 + }, + { + "epoch": 0.22502496635807348, + "grad_norm": 0.14990265667438507, + "learning_rate": 4.765680974999123e-05, + "loss": 3.2171, + "step": 7295 + }, + { + "epoch": 0.2250558128236469, + "grad_norm": 0.15696056187152863, + "learning_rate": 4.7655671754584416e-05, + "loss": 3.2095, + "step": 7296 + }, + { + "epoch": 0.2250866592892203, + "grad_norm": 0.15683303773403168, + "learning_rate": 4.76545334964987e-05, + "loss": 3.2495, + "step": 7297 + }, + { + "epoch": 0.22511750575479372, + "grad_norm": 0.1541966050863266, + "learning_rate": 4.7653394975747285e-05, + "loss": 3.2093, + "step": 7298 + }, + { + "epoch": 0.22514835222036714, + "grad_norm": 0.1611153483390808, + "learning_rate": 4.765225619234337e-05, + "loss": 3.2392, + "step": 7299 + }, + { + "epoch": 0.22517919868594058, + "grad_norm": 0.1709214150905609, + "learning_rate": 4.765111714630015e-05, + "loss": 3.2018, + "step": 7300 + }, + { + "epoch": 0.225210045151514, + "grad_norm": 0.1550784856081009, + "learning_rate": 4.7649977837630844e-05, + "loss": 3.1924, + "step": 7301 + }, + { + "epoch": 0.2252408916170874, + "grad_norm": 0.16324472427368164, + "learning_rate": 4.764883826634865e-05, + "loss": 3.2503, + "step": 7302 + }, + { + "epoch": 0.22527173808266082, + "grad_norm": 0.1546812653541565, + "learning_rate": 4.764769843246679e-05, + "loss": 3.2181, + "step": 7303 + }, + { + "epoch": 0.22530258454823424, + "grad_norm": 0.15927548706531525, + "learning_rate": 4.764655833599848e-05, + "loss": 3.1982, + "step": 7304 + }, + { + "epoch": 0.22533343101380765, + "grad_norm": 0.1607206016778946, + "learning_rate": 4.7645417976956935e-05, + "loss": 3.1919, + "step": 7305 + }, + { + "epoch": 0.22536427747938106, + "grad_norm": 0.15304118394851685, + "learning_rate": 4.764427735535537e-05, + "loss": 3.1803, + "step": 7306 + }, + { + "epoch": 0.22539512394495448, + "grad_norm": 0.16508866846561432, + "learning_rate": 4.764313647120703e-05, + "loss": 3.2068, + "step": 7307 + }, + { + "epoch": 0.2254259704105279, + "grad_norm": 0.15070345997810364, + "learning_rate": 4.764199532452512e-05, + "loss": 3.2371, + "step": 7308 + }, + { + "epoch": 0.2254568168761013, + "grad_norm": 0.15863800048828125, + "learning_rate": 4.764085391532288e-05, + "loss": 3.2007, + "step": 7309 + }, + { + "epoch": 0.22548766334167472, + "grad_norm": 0.15277834236621857, + "learning_rate": 4.7639712243613543e-05, + "loss": 3.255, + "step": 7310 + }, + { + "epoch": 0.22551850980724814, + "grad_norm": 0.15984943509101868, + "learning_rate": 4.7638570309410354e-05, + "loss": 3.2455, + "step": 7311 + }, + { + "epoch": 0.22554935627282158, + "grad_norm": 0.15743787586688995, + "learning_rate": 4.763742811272654e-05, + "loss": 3.2137, + "step": 7312 + }, + { + "epoch": 0.225580202738395, + "grad_norm": 0.150337815284729, + "learning_rate": 4.7636285653575354e-05, + "loss": 3.214, + "step": 7313 + }, + { + "epoch": 0.2256110492039684, + "grad_norm": 0.16118057072162628, + "learning_rate": 4.763514293197003e-05, + "loss": 3.228, + "step": 7314 + }, + { + "epoch": 0.22564189566954182, + "grad_norm": 0.1552322655916214, + "learning_rate": 4.7633999947923843e-05, + "loss": 3.2014, + "step": 7315 + }, + { + "epoch": 0.22567274213511523, + "grad_norm": 0.15755011141300201, + "learning_rate": 4.7632856701450016e-05, + "loss": 3.1916, + "step": 7316 + }, + { + "epoch": 0.22570358860068865, + "grad_norm": 0.15408310294151306, + "learning_rate": 4.763171319256182e-05, + "loss": 3.2112, + "step": 7317 + }, + { + "epoch": 0.22573443506626206, + "grad_norm": 0.1563519835472107, + "learning_rate": 4.763056942127251e-05, + "loss": 3.2025, + "step": 7318 + }, + { + "epoch": 0.22576528153183548, + "grad_norm": 0.1617918759584427, + "learning_rate": 4.762942538759534e-05, + "loss": 3.2205, + "step": 7319 + }, + { + "epoch": 0.2257961279974089, + "grad_norm": 0.16446086764335632, + "learning_rate": 4.7628281091543595e-05, + "loss": 3.2357, + "step": 7320 + }, + { + "epoch": 0.2258269744629823, + "grad_norm": 0.1535269021987915, + "learning_rate": 4.7627136533130515e-05, + "loss": 3.2457, + "step": 7321 + }, + { + "epoch": 0.22585782092855572, + "grad_norm": 0.16005127131938934, + "learning_rate": 4.762599171236939e-05, + "loss": 3.2212, + "step": 7322 + }, + { + "epoch": 0.22588866739412913, + "grad_norm": 0.16749820113182068, + "learning_rate": 4.7624846629273486e-05, + "loss": 3.2037, + "step": 7323 + }, + { + "epoch": 0.22591951385970258, + "grad_norm": 0.18742366135120392, + "learning_rate": 4.762370128385609e-05, + "loss": 3.1962, + "step": 7324 + }, + { + "epoch": 0.225950360325276, + "grad_norm": 0.16136878728866577, + "learning_rate": 4.762255567613047e-05, + "loss": 3.242, + "step": 7325 + }, + { + "epoch": 0.2259812067908494, + "grad_norm": 0.15368077158927917, + "learning_rate": 4.7621409806109895e-05, + "loss": 3.2225, + "step": 7326 + }, + { + "epoch": 0.22601205325642282, + "grad_norm": 0.15892040729522705, + "learning_rate": 4.7620263673807683e-05, + "loss": 3.2144, + "step": 7327 + }, + { + "epoch": 0.22604289972199623, + "grad_norm": 0.15302333235740662, + "learning_rate": 4.76191172792371e-05, + "loss": 3.238, + "step": 7328 + }, + { + "epoch": 0.22607374618756965, + "grad_norm": 0.15293189883232117, + "learning_rate": 4.761797062241145e-05, + "loss": 3.2392, + "step": 7329 + }, + { + "epoch": 0.22610459265314306, + "grad_norm": 0.1565956026315689, + "learning_rate": 4.7616823703344024e-05, + "loss": 3.1715, + "step": 7330 + }, + { + "epoch": 0.22613543911871647, + "grad_norm": 0.41940754652023315, + "learning_rate": 4.7615676522048116e-05, + "loss": 3.2364, + "step": 7331 + }, + { + "epoch": 0.2261662855842899, + "grad_norm": 0.1541851907968521, + "learning_rate": 4.7614529078537026e-05, + "loss": 3.2097, + "step": 7332 + }, + { + "epoch": 0.2261971320498633, + "grad_norm": 0.15477554500102997, + "learning_rate": 4.761338137282406e-05, + "loss": 3.205, + "step": 7333 + }, + { + "epoch": 0.22622797851543672, + "grad_norm": 0.16056624054908752, + "learning_rate": 4.761223340492254e-05, + "loss": 3.1791, + "step": 7334 + }, + { + "epoch": 0.22625882498101013, + "grad_norm": 0.2358560413122177, + "learning_rate": 4.761108517484574e-05, + "loss": 3.2235, + "step": 7335 + }, + { + "epoch": 0.22628967144658357, + "grad_norm": 0.1525174081325531, + "learning_rate": 4.760993668260701e-05, + "loss": 3.2102, + "step": 7336 + }, + { + "epoch": 0.226320517912157, + "grad_norm": 0.15760205686092377, + "learning_rate": 4.760878792821965e-05, + "loss": 3.2262, + "step": 7337 + }, + { + "epoch": 0.2263513643777304, + "grad_norm": 0.15766632556915283, + "learning_rate": 4.760763891169697e-05, + "loss": 3.2114, + "step": 7338 + }, + { + "epoch": 0.22638221084330382, + "grad_norm": 0.1631905734539032, + "learning_rate": 4.7606489633052315e-05, + "loss": 3.19, + "step": 7339 + }, + { + "epoch": 0.22641305730887723, + "grad_norm": 0.15529710054397583, + "learning_rate": 4.760534009229899e-05, + "loss": 3.1907, + "step": 7340 + }, + { + "epoch": 0.22644390377445064, + "grad_norm": 0.15922683477401733, + "learning_rate": 4.7604190289450335e-05, + "loss": 3.2102, + "step": 7341 + }, + { + "epoch": 0.22647475024002406, + "grad_norm": 0.15232880413532257, + "learning_rate": 4.760304022451968e-05, + "loss": 3.203, + "step": 7342 + }, + { + "epoch": 0.22650559670559747, + "grad_norm": 0.2491171658039093, + "learning_rate": 4.7601889897520345e-05, + "loss": 3.2024, + "step": 7343 + }, + { + "epoch": 0.2265364431711709, + "grad_norm": 0.1614120751619339, + "learning_rate": 4.760073930846569e-05, + "loss": 3.1882, + "step": 7344 + }, + { + "epoch": 0.2265672896367443, + "grad_norm": 0.1644321084022522, + "learning_rate": 4.759958845736904e-05, + "loss": 3.214, + "step": 7345 + }, + { + "epoch": 0.22659813610231772, + "grad_norm": 0.15186740458011627, + "learning_rate": 4.759843734424374e-05, + "loss": 3.2143, + "step": 7346 + }, + { + "epoch": 0.22662898256789113, + "grad_norm": 0.16318967938423157, + "learning_rate": 4.7597285969103146e-05, + "loss": 3.2249, + "step": 7347 + }, + { + "epoch": 0.22665982903346457, + "grad_norm": 0.15949782729148865, + "learning_rate": 4.75961343319606e-05, + "loss": 3.2125, + "step": 7348 + }, + { + "epoch": 0.22669067549903799, + "grad_norm": 0.1636711061000824, + "learning_rate": 4.759498243282945e-05, + "loss": 3.2108, + "step": 7349 + }, + { + "epoch": 0.2267215219646114, + "grad_norm": 0.1559343934059143, + "learning_rate": 4.7593830271723054e-05, + "loss": 3.1914, + "step": 7350 + }, + { + "epoch": 0.22675236843018481, + "grad_norm": 0.157964289188385, + "learning_rate": 4.759267784865478e-05, + "loss": 3.2094, + "step": 7351 + }, + { + "epoch": 0.22678321489575823, + "grad_norm": 0.15204420685768127, + "learning_rate": 4.759152516363798e-05, + "loss": 3.2105, + "step": 7352 + }, + { + "epoch": 0.22681406136133164, + "grad_norm": 0.14513273537158966, + "learning_rate": 4.759037221668602e-05, + "loss": 3.2125, + "step": 7353 + }, + { + "epoch": 0.22684490782690506, + "grad_norm": 0.1566774845123291, + "learning_rate": 4.7589219007812266e-05, + "loss": 3.2439, + "step": 7354 + }, + { + "epoch": 0.22687575429247847, + "grad_norm": 0.15039785206317902, + "learning_rate": 4.75880655370301e-05, + "loss": 3.2157, + "step": 7355 + }, + { + "epoch": 0.22690660075805189, + "grad_norm": 0.4398164451122284, + "learning_rate": 4.758691180435288e-05, + "loss": 3.211, + "step": 7356 + }, + { + "epoch": 0.2269374472236253, + "grad_norm": 0.18854045867919922, + "learning_rate": 4.758575780979401e-05, + "loss": 3.2269, + "step": 7357 + }, + { + "epoch": 0.2269682936891987, + "grad_norm": 0.16645821928977966, + "learning_rate": 4.7584603553366834e-05, + "loss": 3.2438, + "step": 7358 + }, + { + "epoch": 0.22699914015477213, + "grad_norm": 0.14902439713478088, + "learning_rate": 4.758344903508475e-05, + "loss": 3.1882, + "step": 7359 + }, + { + "epoch": 0.22702998662034557, + "grad_norm": 0.15753266215324402, + "learning_rate": 4.7582294254961156e-05, + "loss": 3.1964, + "step": 7360 + }, + { + "epoch": 0.22706083308591898, + "grad_norm": 0.15352371335029602, + "learning_rate": 4.758113921300942e-05, + "loss": 3.247, + "step": 7361 + }, + { + "epoch": 0.2270916795514924, + "grad_norm": 0.14808423817157745, + "learning_rate": 4.757998390924295e-05, + "loss": 3.176, + "step": 7362 + }, + { + "epoch": 0.2271225260170658, + "grad_norm": 0.1575491726398468, + "learning_rate": 4.757882834367514e-05, + "loss": 3.2265, + "step": 7363 + }, + { + "epoch": 0.22715337248263923, + "grad_norm": 0.16390612721443176, + "learning_rate": 4.757767251631937e-05, + "loss": 3.223, + "step": 7364 + }, + { + "epoch": 0.22718421894821264, + "grad_norm": 0.15313328802585602, + "learning_rate": 4.7576516427189064e-05, + "loss": 3.2325, + "step": 7365 + }, + { + "epoch": 0.22721506541378605, + "grad_norm": 0.14823031425476074, + "learning_rate": 4.757536007629762e-05, + "loss": 3.1976, + "step": 7366 + }, + { + "epoch": 0.22724591187935947, + "grad_norm": 0.15460267663002014, + "learning_rate": 4.757420346365844e-05, + "loss": 3.2126, + "step": 7367 + }, + { + "epoch": 0.22727675834493288, + "grad_norm": 0.16656272113323212, + "learning_rate": 4.7573046589284934e-05, + "loss": 3.2119, + "step": 7368 + }, + { + "epoch": 0.2273076048105063, + "grad_norm": 0.16114118695259094, + "learning_rate": 4.757188945319052e-05, + "loss": 3.2168, + "step": 7369 + }, + { + "epoch": 0.2273384512760797, + "grad_norm": 0.15766963362693787, + "learning_rate": 4.7570732055388605e-05, + "loss": 3.2052, + "step": 7370 + }, + { + "epoch": 0.22736929774165313, + "grad_norm": 0.14849978685379028, + "learning_rate": 4.7569574395892623e-05, + "loss": 3.2345, + "step": 7371 + }, + { + "epoch": 0.22740014420722657, + "grad_norm": 0.15717411041259766, + "learning_rate": 4.7568416474715984e-05, + "loss": 3.1759, + "step": 7372 + }, + { + "epoch": 0.22743099067279998, + "grad_norm": 0.1566469818353653, + "learning_rate": 4.756725829187212e-05, + "loss": 3.186, + "step": 7373 + }, + { + "epoch": 0.2274618371383734, + "grad_norm": 0.1514650285243988, + "learning_rate": 4.756609984737446e-05, + "loss": 3.2421, + "step": 7374 + }, + { + "epoch": 0.2274926836039468, + "grad_norm": 0.155424565076828, + "learning_rate": 4.756494114123643e-05, + "loss": 3.1876, + "step": 7375 + }, + { + "epoch": 0.22752353006952022, + "grad_norm": 0.15160556137561798, + "learning_rate": 4.7563782173471464e-05, + "loss": 3.2315, + "step": 7376 + }, + { + "epoch": 0.22755437653509364, + "grad_norm": 0.15193940699100494, + "learning_rate": 4.756262294409301e-05, + "loss": 3.2269, + "step": 7377 + }, + { + "epoch": 0.22758522300066705, + "grad_norm": 0.1452501267194748, + "learning_rate": 4.7561463453114496e-05, + "loss": 3.2207, + "step": 7378 + }, + { + "epoch": 0.22761606946624047, + "grad_norm": 0.15555667877197266, + "learning_rate": 4.7560303700549374e-05, + "loss": 3.2378, + "step": 7379 + }, + { + "epoch": 0.22764691593181388, + "grad_norm": 0.1434764862060547, + "learning_rate": 4.755914368641109e-05, + "loss": 3.2371, + "step": 7380 + }, + { + "epoch": 0.2276777623973873, + "grad_norm": 0.15224528312683105, + "learning_rate": 4.7557983410713084e-05, + "loss": 3.1794, + "step": 7381 + }, + { + "epoch": 0.2277086088629607, + "grad_norm": 0.14538715779781342, + "learning_rate": 4.755682287346882e-05, + "loss": 3.2337, + "step": 7382 + }, + { + "epoch": 0.22773945532853412, + "grad_norm": 0.1521819680929184, + "learning_rate": 4.755566207469175e-05, + "loss": 3.2211, + "step": 7383 + }, + { + "epoch": 0.22777030179410757, + "grad_norm": 0.16098849475383759, + "learning_rate": 4.7554501014395335e-05, + "loss": 3.2191, + "step": 7384 + }, + { + "epoch": 0.22780114825968098, + "grad_norm": 0.15866492688655853, + "learning_rate": 4.7553339692593034e-05, + "loss": 3.2011, + "step": 7385 + }, + { + "epoch": 0.2278319947252544, + "grad_norm": 0.1557365506887436, + "learning_rate": 4.7552178109298314e-05, + "loss": 3.2139, + "step": 7386 + }, + { + "epoch": 0.2278628411908278, + "grad_norm": 0.15535984933376312, + "learning_rate": 4.755101626452464e-05, + "loss": 3.2733, + "step": 7387 + }, + { + "epoch": 0.22789368765640122, + "grad_norm": 0.1582169234752655, + "learning_rate": 4.754985415828548e-05, + "loss": 3.2045, + "step": 7388 + }, + { + "epoch": 0.22792453412197464, + "grad_norm": 0.15147487819194794, + "learning_rate": 4.754869179059431e-05, + "loss": 3.1986, + "step": 7389 + }, + { + "epoch": 0.22795538058754805, + "grad_norm": 0.15107963979244232, + "learning_rate": 4.754752916146461e-05, + "loss": 3.2221, + "step": 7390 + }, + { + "epoch": 0.22798622705312147, + "grad_norm": 0.1684526950120926, + "learning_rate": 4.754636627090987e-05, + "loss": 3.2377, + "step": 7391 + }, + { + "epoch": 0.22801707351869488, + "grad_norm": 0.15275974571704865, + "learning_rate": 4.7545203118943546e-05, + "loss": 3.1879, + "step": 7392 + }, + { + "epoch": 0.2280479199842683, + "grad_norm": 0.15325003862380981, + "learning_rate": 4.7544039705579154e-05, + "loss": 3.2303, + "step": 7393 + }, + { + "epoch": 0.2280787664498417, + "grad_norm": 0.15289604663848877, + "learning_rate": 4.7542876030830166e-05, + "loss": 3.1882, + "step": 7394 + }, + { + "epoch": 0.22810961291541512, + "grad_norm": 0.16803979873657227, + "learning_rate": 4.7541712094710066e-05, + "loss": 3.2602, + "step": 7395 + }, + { + "epoch": 0.22814045938098856, + "grad_norm": 0.16331109404563904, + "learning_rate": 4.754054789723237e-05, + "loss": 3.2544, + "step": 7396 + }, + { + "epoch": 0.22817130584656198, + "grad_norm": 0.15493562817573547, + "learning_rate": 4.7539383438410567e-05, + "loss": 3.2161, + "step": 7397 + }, + { + "epoch": 0.2282021523121354, + "grad_norm": 0.1515084058046341, + "learning_rate": 4.7538218718258156e-05, + "loss": 3.2173, + "step": 7398 + }, + { + "epoch": 0.2282329987777088, + "grad_norm": 0.17178227007389069, + "learning_rate": 4.753705373678864e-05, + "loss": 3.225, + "step": 7399 + }, + { + "epoch": 0.22826384524328222, + "grad_norm": 0.16131237149238586, + "learning_rate": 4.753588849401553e-05, + "loss": 3.2137, + "step": 7400 + }, + { + "epoch": 0.22829469170885563, + "grad_norm": 0.1593797355890274, + "learning_rate": 4.753472298995233e-05, + "loss": 3.184, + "step": 7401 + }, + { + "epoch": 0.22832553817442905, + "grad_norm": 0.1572284698486328, + "learning_rate": 4.753355722461257e-05, + "loss": 3.2413, + "step": 7402 + }, + { + "epoch": 0.22835638464000246, + "grad_norm": 0.16020876169204712, + "learning_rate": 4.753239119800975e-05, + "loss": 3.1801, + "step": 7403 + }, + { + "epoch": 0.22838723110557588, + "grad_norm": 0.14872850477695465, + "learning_rate": 4.753122491015739e-05, + "loss": 3.2189, + "step": 7404 + }, + { + "epoch": 0.2284180775711493, + "grad_norm": 0.15000833570957184, + "learning_rate": 4.753005836106902e-05, + "loss": 3.2057, + "step": 7405 + }, + { + "epoch": 0.2284489240367227, + "grad_norm": 0.1547221839427948, + "learning_rate": 4.752889155075817e-05, + "loss": 3.2003, + "step": 7406 + }, + { + "epoch": 0.22847977050229612, + "grad_norm": 0.15839378535747528, + "learning_rate": 4.7527724479238354e-05, + "loss": 3.2233, + "step": 7407 + }, + { + "epoch": 0.22851061696786956, + "grad_norm": 0.15268483757972717, + "learning_rate": 4.752655714652311e-05, + "loss": 3.2259, + "step": 7408 + }, + { + "epoch": 0.22854146343344298, + "grad_norm": 0.15694551169872284, + "learning_rate": 4.752538955262598e-05, + "loss": 3.2336, + "step": 7409 + }, + { + "epoch": 0.2285723098990164, + "grad_norm": 0.15254759788513184, + "learning_rate": 4.752422169756048e-05, + "loss": 3.2358, + "step": 7410 + }, + { + "epoch": 0.2286031563645898, + "grad_norm": 0.1543646901845932, + "learning_rate": 4.752305358134017e-05, + "loss": 3.2125, + "step": 7411 + }, + { + "epoch": 0.22863400283016322, + "grad_norm": 0.14953362941741943, + "learning_rate": 4.7521885203978594e-05, + "loss": 3.2245, + "step": 7412 + }, + { + "epoch": 0.22866484929573663, + "grad_norm": 1.2319700717926025, + "learning_rate": 4.752071656548929e-05, + "loss": 3.2801, + "step": 7413 + }, + { + "epoch": 0.22869569576131005, + "grad_norm": 0.16513153910636902, + "learning_rate": 4.751954766588581e-05, + "loss": 3.1844, + "step": 7414 + }, + { + "epoch": 0.22872654222688346, + "grad_norm": 0.16062186658382416, + "learning_rate": 4.751837850518171e-05, + "loss": 3.1818, + "step": 7415 + }, + { + "epoch": 0.22875738869245688, + "grad_norm": 0.16388684511184692, + "learning_rate": 4.751720908339055e-05, + "loss": 3.2038, + "step": 7416 + }, + { + "epoch": 0.2287882351580303, + "grad_norm": 0.1606515496969223, + "learning_rate": 4.751603940052587e-05, + "loss": 3.2503, + "step": 7417 + }, + { + "epoch": 0.2288190816236037, + "grad_norm": 0.15740199387073517, + "learning_rate": 4.751486945660124e-05, + "loss": 3.2241, + "step": 7418 + }, + { + "epoch": 0.22884992808917712, + "grad_norm": 0.15086615085601807, + "learning_rate": 4.751369925163024e-05, + "loss": 3.1926, + "step": 7419 + }, + { + "epoch": 0.22888077455475056, + "grad_norm": 0.15300318598747253, + "learning_rate": 4.751252878562642e-05, + "loss": 3.1832, + "step": 7420 + }, + { + "epoch": 0.22891162102032397, + "grad_norm": 0.1535147875547409, + "learning_rate": 4.751135805860336e-05, + "loss": 3.214, + "step": 7421 + }, + { + "epoch": 0.2289424674858974, + "grad_norm": 0.15841132402420044, + "learning_rate": 4.751018707057463e-05, + "loss": 3.2306, + "step": 7422 + }, + { + "epoch": 0.2289733139514708, + "grad_norm": 0.15389271080493927, + "learning_rate": 4.7509015821553806e-05, + "loss": 3.2199, + "step": 7423 + }, + { + "epoch": 0.22900416041704422, + "grad_norm": 0.157559335231781, + "learning_rate": 4.750784431155448e-05, + "loss": 3.2145, + "step": 7424 + }, + { + "epoch": 0.22903500688261763, + "grad_norm": 0.1597459614276886, + "learning_rate": 4.750667254059021e-05, + "loss": 3.2388, + "step": 7425 + }, + { + "epoch": 0.22906585334819105, + "grad_norm": 0.15070931613445282, + "learning_rate": 4.7505500508674606e-05, + "loss": 3.212, + "step": 7426 + }, + { + "epoch": 0.22909669981376446, + "grad_norm": 0.16762769222259521, + "learning_rate": 4.7504328215821246e-05, + "loss": 3.2401, + "step": 7427 + }, + { + "epoch": 0.22912754627933787, + "grad_norm": 0.15410268306732178, + "learning_rate": 4.750315566204372e-05, + "loss": 3.2356, + "step": 7428 + }, + { + "epoch": 0.2291583927449113, + "grad_norm": 0.15735255181789398, + "learning_rate": 4.750198284735564e-05, + "loss": 3.2028, + "step": 7429 + }, + { + "epoch": 0.2291892392104847, + "grad_norm": 0.15850384533405304, + "learning_rate": 4.750080977177058e-05, + "loss": 3.1787, + "step": 7430 + }, + { + "epoch": 0.22922008567605814, + "grad_norm": 0.16016298532485962, + "learning_rate": 4.749963643530215e-05, + "loss": 3.229, + "step": 7431 + }, + { + "epoch": 0.22925093214163156, + "grad_norm": 0.16459183394908905, + "learning_rate": 4.7498462837963964e-05, + "loss": 3.2059, + "step": 7432 + }, + { + "epoch": 0.22928177860720497, + "grad_norm": 0.16598965227603912, + "learning_rate": 4.7497288979769614e-05, + "loss": 3.197, + "step": 7433 + }, + { + "epoch": 0.2293126250727784, + "grad_norm": 0.1496271789073944, + "learning_rate": 4.749611486073272e-05, + "loss": 3.1919, + "step": 7434 + }, + { + "epoch": 0.2293434715383518, + "grad_norm": 0.1727103739976883, + "learning_rate": 4.74949404808669e-05, + "loss": 3.2373, + "step": 7435 + }, + { + "epoch": 0.22937431800392521, + "grad_norm": 0.16900977492332458, + "learning_rate": 4.749376584018576e-05, + "loss": 3.2255, + "step": 7436 + }, + { + "epoch": 0.22940516446949863, + "grad_norm": 0.1553669273853302, + "learning_rate": 4.749259093870291e-05, + "loss": 3.2419, + "step": 7437 + }, + { + "epoch": 0.22943601093507204, + "grad_norm": 0.16264423727989197, + "learning_rate": 4.7491415776432e-05, + "loss": 3.1885, + "step": 7438 + }, + { + "epoch": 0.22946685740064546, + "grad_norm": 0.1556665301322937, + "learning_rate": 4.749024035338663e-05, + "loss": 3.2097, + "step": 7439 + }, + { + "epoch": 0.22949770386621887, + "grad_norm": 0.1575068086385727, + "learning_rate": 4.748906466958044e-05, + "loss": 3.249, + "step": 7440 + }, + { + "epoch": 0.22952855033179229, + "grad_norm": 0.15326131880283356, + "learning_rate": 4.7487888725027064e-05, + "loss": 3.2155, + "step": 7441 + }, + { + "epoch": 0.2295593967973657, + "grad_norm": 0.18895314633846283, + "learning_rate": 4.748671251974013e-05, + "loss": 3.2342, + "step": 7442 + }, + { + "epoch": 0.22959024326293914, + "grad_norm": 0.1538863480091095, + "learning_rate": 4.748553605373327e-05, + "loss": 3.2014, + "step": 7443 + }, + { + "epoch": 0.22962108972851256, + "grad_norm": 0.16659271717071533, + "learning_rate": 4.7484359327020146e-05, + "loss": 3.2278, + "step": 7444 + }, + { + "epoch": 0.22965193619408597, + "grad_norm": 0.15507586300373077, + "learning_rate": 4.7483182339614374e-05, + "loss": 3.188, + "step": 7445 + }, + { + "epoch": 0.22968278265965938, + "grad_norm": 0.1599980592727661, + "learning_rate": 4.7482005091529625e-05, + "loss": 3.1793, + "step": 7446 + }, + { + "epoch": 0.2297136291252328, + "grad_norm": 0.1750478446483612, + "learning_rate": 4.748082758277953e-05, + "loss": 3.1975, + "step": 7447 + }, + { + "epoch": 0.2297444755908062, + "grad_norm": 0.1675185114145279, + "learning_rate": 4.7479649813377746e-05, + "loss": 3.2118, + "step": 7448 + }, + { + "epoch": 0.22977532205637963, + "grad_norm": 0.15587693452835083, + "learning_rate": 4.747847178333794e-05, + "loss": 3.2272, + "step": 7449 + }, + { + "epoch": 0.22980616852195304, + "grad_norm": 0.17162729799747467, + "learning_rate": 4.7477293492673754e-05, + "loss": 3.1896, + "step": 7450 + }, + { + "epoch": 0.22983701498752646, + "grad_norm": 0.15090322494506836, + "learning_rate": 4.747611494139885e-05, + "loss": 3.2072, + "step": 7451 + }, + { + "epoch": 0.22986786145309987, + "grad_norm": 0.16815711557865143, + "learning_rate": 4.7474936129526917e-05, + "loss": 3.245, + "step": 7452 + }, + { + "epoch": 0.22989870791867328, + "grad_norm": 0.15410779416561127, + "learning_rate": 4.74737570570716e-05, + "loss": 3.1993, + "step": 7453 + }, + { + "epoch": 0.2299295543842467, + "grad_norm": 0.1542227417230606, + "learning_rate": 4.747257772404657e-05, + "loss": 3.2176, + "step": 7454 + }, + { + "epoch": 0.22996040084982014, + "grad_norm": 0.1542416512966156, + "learning_rate": 4.7471398130465505e-05, + "loss": 3.2561, + "step": 7455 + }, + { + "epoch": 0.22999124731539355, + "grad_norm": 0.1616598665714264, + "learning_rate": 4.747021827634209e-05, + "loss": 3.2039, + "step": 7456 + }, + { + "epoch": 0.23002209378096697, + "grad_norm": 0.1570194661617279, + "learning_rate": 4.746903816169e-05, + "loss": 3.2121, + "step": 7457 + }, + { + "epoch": 0.23005294024654038, + "grad_norm": 0.15935996174812317, + "learning_rate": 4.7467857786522895e-05, + "loss": 3.2077, + "step": 7458 + }, + { + "epoch": 0.2300837867121138, + "grad_norm": 0.15539270639419556, + "learning_rate": 4.74666771508545e-05, + "loss": 3.1935, + "step": 7459 + }, + { + "epoch": 0.2301146331776872, + "grad_norm": 0.15727849304676056, + "learning_rate": 4.7465496254698474e-05, + "loss": 3.2312, + "step": 7460 + }, + { + "epoch": 0.23014547964326063, + "grad_norm": 0.1612820327281952, + "learning_rate": 4.7464315098068526e-05, + "loss": 3.2214, + "step": 7461 + }, + { + "epoch": 0.23017632610883404, + "grad_norm": 0.1528947353363037, + "learning_rate": 4.746313368097834e-05, + "loss": 3.2251, + "step": 7462 + }, + { + "epoch": 0.23020717257440745, + "grad_norm": 0.17802657186985016, + "learning_rate": 4.746195200344162e-05, + "loss": 3.1861, + "step": 7463 + }, + { + "epoch": 0.23023801903998087, + "grad_norm": 0.15062496066093445, + "learning_rate": 4.746077006547206e-05, + "loss": 3.1829, + "step": 7464 + }, + { + "epoch": 0.23026886550555428, + "grad_norm": 0.175985187292099, + "learning_rate": 4.7459587867083376e-05, + "loss": 3.2181, + "step": 7465 + }, + { + "epoch": 0.2302997119711277, + "grad_norm": 0.16071341931819916, + "learning_rate": 4.745840540828926e-05, + "loss": 3.2338, + "step": 7466 + }, + { + "epoch": 0.23033055843670114, + "grad_norm": 0.1632901281118393, + "learning_rate": 4.7457222689103434e-05, + "loss": 3.2253, + "step": 7467 + }, + { + "epoch": 0.23036140490227455, + "grad_norm": 0.15288560092449188, + "learning_rate": 4.74560397095396e-05, + "loss": 3.2101, + "step": 7468 + }, + { + "epoch": 0.23039225136784797, + "grad_norm": 0.1793062686920166, + "learning_rate": 4.745485646961148e-05, + "loss": 3.2423, + "step": 7469 + }, + { + "epoch": 0.23042309783342138, + "grad_norm": 0.14916054904460907, + "learning_rate": 4.745367296933281e-05, + "loss": 3.1808, + "step": 7470 + }, + { + "epoch": 0.2304539442989948, + "grad_norm": 0.16285845637321472, + "learning_rate": 4.745248920871728e-05, + "loss": 3.2287, + "step": 7471 + }, + { + "epoch": 0.2304847907645682, + "grad_norm": 0.15058588981628418, + "learning_rate": 4.745130518777863e-05, + "loss": 3.2072, + "step": 7472 + }, + { + "epoch": 0.23051563723014162, + "grad_norm": 0.1632135957479477, + "learning_rate": 4.745012090653059e-05, + "loss": 3.2238, + "step": 7473 + }, + { + "epoch": 0.23054648369571504, + "grad_norm": 0.1544756144285202, + "learning_rate": 4.744893636498689e-05, + "loss": 3.19, + "step": 7474 + }, + { + "epoch": 0.23057733016128845, + "grad_norm": 0.152268648147583, + "learning_rate": 4.744775156316126e-05, + "loss": 3.2359, + "step": 7475 + }, + { + "epoch": 0.23060817662686187, + "grad_norm": 0.15186725556850433, + "learning_rate": 4.7446566501067444e-05, + "loss": 3.2028, + "step": 7476 + }, + { + "epoch": 0.23063902309243528, + "grad_norm": 0.16962067782878876, + "learning_rate": 4.744538117871918e-05, + "loss": 3.2221, + "step": 7477 + }, + { + "epoch": 0.2306698695580087, + "grad_norm": 0.15110796689987183, + "learning_rate": 4.74441955961302e-05, + "loss": 3.1928, + "step": 7478 + }, + { + "epoch": 0.23070071602358214, + "grad_norm": 0.1532507687807083, + "learning_rate": 4.744300975331427e-05, + "loss": 3.2038, + "step": 7479 + }, + { + "epoch": 0.23073156248915555, + "grad_norm": 0.1534733921289444, + "learning_rate": 4.7441823650285124e-05, + "loss": 3.2033, + "step": 7480 + }, + { + "epoch": 0.23076240895472896, + "grad_norm": 0.16148842871189117, + "learning_rate": 4.744063728705652e-05, + "loss": 3.201, + "step": 7481 + }, + { + "epoch": 0.23079325542030238, + "grad_norm": 0.14628523588180542, + "learning_rate": 4.743945066364222e-05, + "loss": 3.2059, + "step": 7482 + }, + { + "epoch": 0.2308241018858758, + "grad_norm": 0.15599384903907776, + "learning_rate": 4.743826378005596e-05, + "loss": 3.2154, + "step": 7483 + }, + { + "epoch": 0.2308549483514492, + "grad_norm": 0.1571965366601944, + "learning_rate": 4.7437076636311514e-05, + "loss": 3.2267, + "step": 7484 + }, + { + "epoch": 0.23088579481702262, + "grad_norm": 0.1518636792898178, + "learning_rate": 4.743588923242266e-05, + "loss": 3.225, + "step": 7485 + }, + { + "epoch": 0.23091664128259604, + "grad_norm": 0.15988728404045105, + "learning_rate": 4.7434701568403144e-05, + "loss": 3.2261, + "step": 7486 + }, + { + "epoch": 0.23094748774816945, + "grad_norm": 0.1525585651397705, + "learning_rate": 4.7433513644266746e-05, + "loss": 3.1962, + "step": 7487 + }, + { + "epoch": 0.23097833421374286, + "grad_norm": 0.1561451107263565, + "learning_rate": 4.743232546002724e-05, + "loss": 3.2262, + "step": 7488 + }, + { + "epoch": 0.23100918067931628, + "grad_norm": 0.1562286913394928, + "learning_rate": 4.74311370156984e-05, + "loss": 3.2408, + "step": 7489 + }, + { + "epoch": 0.2310400271448897, + "grad_norm": 0.160136416554451, + "learning_rate": 4.742994831129401e-05, + "loss": 3.2537, + "step": 7490 + }, + { + "epoch": 0.23107087361046313, + "grad_norm": 0.15398792922496796, + "learning_rate": 4.742875934682784e-05, + "loss": 3.2095, + "step": 7491 + }, + { + "epoch": 0.23110172007603655, + "grad_norm": 0.15791691839694977, + "learning_rate": 4.742757012231369e-05, + "loss": 3.2547, + "step": 7492 + }, + { + "epoch": 0.23113256654160996, + "grad_norm": 0.15382148325443268, + "learning_rate": 4.742638063776534e-05, + "loss": 3.1893, + "step": 7493 + }, + { + "epoch": 0.23116341300718338, + "grad_norm": 0.1495608389377594, + "learning_rate": 4.742519089319658e-05, + "loss": 3.2381, + "step": 7494 + }, + { + "epoch": 0.2311942594727568, + "grad_norm": 0.1572500616312027, + "learning_rate": 4.742400088862121e-05, + "loss": 3.1981, + "step": 7495 + }, + { + "epoch": 0.2312251059383302, + "grad_norm": 0.15395483374595642, + "learning_rate": 4.7422810624053025e-05, + "loss": 3.2403, + "step": 7496 + }, + { + "epoch": 0.23125595240390362, + "grad_norm": 0.16124437749385834, + "learning_rate": 4.7421620099505834e-05, + "loss": 3.2072, + "step": 7497 + }, + { + "epoch": 0.23128679886947703, + "grad_norm": 0.1618873029947281, + "learning_rate": 4.7420429314993417e-05, + "loss": 3.1937, + "step": 7498 + }, + { + "epoch": 0.23131764533505045, + "grad_norm": 0.15285536646842957, + "learning_rate": 4.741923827052961e-05, + "loss": 3.1892, + "step": 7499 + }, + { + "epoch": 0.23134849180062386, + "grad_norm": 0.1547931432723999, + "learning_rate": 4.741804696612819e-05, + "loss": 3.2105, + "step": 7500 + }, + { + "epoch": 0.23137933826619728, + "grad_norm": 0.15650376677513123, + "learning_rate": 4.7416855401803e-05, + "loss": 3.1948, + "step": 7501 + }, + { + "epoch": 0.2314101847317707, + "grad_norm": 0.15674173831939697, + "learning_rate": 4.741566357756784e-05, + "loss": 3.2066, + "step": 7502 + }, + { + "epoch": 0.23144103119734413, + "grad_norm": 0.1507578343153, + "learning_rate": 4.741447149343653e-05, + "loss": 3.1995, + "step": 7503 + }, + { + "epoch": 0.23147187766291755, + "grad_norm": 0.1499904990196228, + "learning_rate": 4.74132791494229e-05, + "loss": 3.2029, + "step": 7504 + }, + { + "epoch": 0.23150272412849096, + "grad_norm": 0.15675987303256989, + "learning_rate": 4.741208654554076e-05, + "loss": 3.1996, + "step": 7505 + }, + { + "epoch": 0.23153357059406438, + "grad_norm": 0.15635940432548523, + "learning_rate": 4.7410893681803936e-05, + "loss": 3.198, + "step": 7506 + }, + { + "epoch": 0.2315644170596378, + "grad_norm": 0.1551559865474701, + "learning_rate": 4.740970055822628e-05, + "loss": 3.2131, + "step": 7507 + }, + { + "epoch": 0.2315952635252112, + "grad_norm": 0.15538813173770905, + "learning_rate": 4.74085071748216e-05, + "loss": 3.243, + "step": 7508 + }, + { + "epoch": 0.23162610999078462, + "grad_norm": 0.16149236261844635, + "learning_rate": 4.740731353160376e-05, + "loss": 3.2389, + "step": 7509 + }, + { + "epoch": 0.23165695645635803, + "grad_norm": 0.17732056975364685, + "learning_rate": 4.740611962858658e-05, + "loss": 3.208, + "step": 7510 + }, + { + "epoch": 0.23168780292193145, + "grad_norm": 0.15516507625579834, + "learning_rate": 4.7404925465783905e-05, + "loss": 3.181, + "step": 7511 + }, + { + "epoch": 0.23171864938750486, + "grad_norm": 0.16302071511745453, + "learning_rate": 4.740373104320959e-05, + "loss": 3.1983, + "step": 7512 + }, + { + "epoch": 0.23174949585307827, + "grad_norm": 0.1529977023601532, + "learning_rate": 4.740253636087746e-05, + "loss": 3.2284, + "step": 7513 + }, + { + "epoch": 0.2317803423186517, + "grad_norm": 0.16047371923923492, + "learning_rate": 4.7401341418801395e-05, + "loss": 3.2119, + "step": 7514 + }, + { + "epoch": 0.23181118878422513, + "grad_norm": 0.15952549874782562, + "learning_rate": 4.740014621699524e-05, + "loss": 3.1853, + "step": 7515 + }, + { + "epoch": 0.23184203524979854, + "grad_norm": 0.171355739235878, + "learning_rate": 4.739895075547285e-05, + "loss": 3.2398, + "step": 7516 + }, + { + "epoch": 0.23187288171537196, + "grad_norm": 0.15464287996292114, + "learning_rate": 4.739775503424808e-05, + "loss": 3.211, + "step": 7517 + }, + { + "epoch": 0.23190372818094537, + "grad_norm": 0.15976788103580475, + "learning_rate": 4.7396559053334806e-05, + "loss": 3.207, + "step": 7518 + }, + { + "epoch": 0.2319345746465188, + "grad_norm": 0.15209387242794037, + "learning_rate": 4.739536281274688e-05, + "loss": 3.2163, + "step": 7519 + }, + { + "epoch": 0.2319654211120922, + "grad_norm": 0.18677987158298492, + "learning_rate": 4.739416631249819e-05, + "loss": 3.2365, + "step": 7520 + }, + { + "epoch": 0.23199626757766562, + "grad_norm": 0.15300817787647247, + "learning_rate": 4.7392969552602594e-05, + "loss": 3.1853, + "step": 7521 + }, + { + "epoch": 0.23202711404323903, + "grad_norm": 0.1585652083158493, + "learning_rate": 4.739177253307398e-05, + "loss": 3.1885, + "step": 7522 + }, + { + "epoch": 0.23205796050881244, + "grad_norm": 0.15271718800067902, + "learning_rate": 4.7390575253926205e-05, + "loss": 3.2706, + "step": 7523 + }, + { + "epoch": 0.23208880697438586, + "grad_norm": 0.16179601848125458, + "learning_rate": 4.7389377715173175e-05, + "loss": 3.257, + "step": 7524 + }, + { + "epoch": 0.23211965343995927, + "grad_norm": 0.1551317572593689, + "learning_rate": 4.738817991682877e-05, + "loss": 3.2335, + "step": 7525 + }, + { + "epoch": 0.2321504999055327, + "grad_norm": 0.16138914227485657, + "learning_rate": 4.738698185890686e-05, + "loss": 3.2168, + "step": 7526 + }, + { + "epoch": 0.23218134637110613, + "grad_norm": 0.14899279177188873, + "learning_rate": 4.7385783541421356e-05, + "loss": 3.2061, + "step": 7527 + }, + { + "epoch": 0.23221219283667954, + "grad_norm": 0.15656781196594238, + "learning_rate": 4.738458496438613e-05, + "loss": 3.2094, + "step": 7528 + }, + { + "epoch": 0.23224303930225296, + "grad_norm": 0.16296295821666718, + "learning_rate": 4.7383386127815116e-05, + "loss": 3.2267, + "step": 7529 + }, + { + "epoch": 0.23227388576782637, + "grad_norm": 0.16133363544940948, + "learning_rate": 4.738218703172218e-05, + "loss": 3.2207, + "step": 7530 + }, + { + "epoch": 0.23230473223339979, + "grad_norm": 0.15276066958904266, + "learning_rate": 4.738098767612124e-05, + "loss": 3.2264, + "step": 7531 + }, + { + "epoch": 0.2323355786989732, + "grad_norm": 0.16405148804187775, + "learning_rate": 4.7379788061026187e-05, + "loss": 3.2242, + "step": 7532 + }, + { + "epoch": 0.23236642516454661, + "grad_norm": 0.16512954235076904, + "learning_rate": 4.737858818645095e-05, + "loss": 3.2099, + "step": 7533 + }, + { + "epoch": 0.23239727163012003, + "grad_norm": 0.15584422647953033, + "learning_rate": 4.7377388052409424e-05, + "loss": 3.1898, + "step": 7534 + }, + { + "epoch": 0.23242811809569344, + "grad_norm": 0.15690279006958008, + "learning_rate": 4.737618765891553e-05, + "loss": 3.1813, + "step": 7535 + }, + { + "epoch": 0.23245896456126686, + "grad_norm": 0.15813516080379486, + "learning_rate": 4.737498700598319e-05, + "loss": 3.2216, + "step": 7536 + }, + { + "epoch": 0.23248981102684027, + "grad_norm": 0.159438818693161, + "learning_rate": 4.737378609362632e-05, + "loss": 3.2378, + "step": 7537 + }, + { + "epoch": 0.23252065749241368, + "grad_norm": 0.16081735491752625, + "learning_rate": 4.737258492185885e-05, + "loss": 3.2157, + "step": 7538 + }, + { + "epoch": 0.23255150395798713, + "grad_norm": 0.16085730493068695, + "learning_rate": 4.7371383490694696e-05, + "loss": 3.2378, + "step": 7539 + }, + { + "epoch": 0.23258235042356054, + "grad_norm": 0.17821452021598816, + "learning_rate": 4.73701818001478e-05, + "loss": 3.226, + "step": 7540 + }, + { + "epoch": 0.23261319688913396, + "grad_norm": 0.15920624136924744, + "learning_rate": 4.7368979850232085e-05, + "loss": 3.2179, + "step": 7541 + }, + { + "epoch": 0.23264404335470737, + "grad_norm": 0.16718989610671997, + "learning_rate": 4.7367777640961494e-05, + "loss": 3.2151, + "step": 7542 + }, + { + "epoch": 0.23267488982028078, + "grad_norm": 0.15648317337036133, + "learning_rate": 4.736657517234996e-05, + "loss": 3.1948, + "step": 7543 + }, + { + "epoch": 0.2327057362858542, + "grad_norm": 0.159878209233284, + "learning_rate": 4.7365372444411424e-05, + "loss": 3.2199, + "step": 7544 + }, + { + "epoch": 0.2327365827514276, + "grad_norm": 0.1598787158727646, + "learning_rate": 4.736416945715984e-05, + "loss": 3.2398, + "step": 7545 + }, + { + "epoch": 0.23276742921700103, + "grad_norm": 0.1568092554807663, + "learning_rate": 4.736296621060915e-05, + "loss": 3.1944, + "step": 7546 + }, + { + "epoch": 0.23279827568257444, + "grad_norm": 0.15157043933868408, + "learning_rate": 4.7361762704773303e-05, + "loss": 3.2089, + "step": 7547 + }, + { + "epoch": 0.23282912214814785, + "grad_norm": 0.21121609210968018, + "learning_rate": 4.736055893966625e-05, + "loss": 3.2444, + "step": 7548 + }, + { + "epoch": 0.23285996861372127, + "grad_norm": 0.15564818680286407, + "learning_rate": 4.7359354915301967e-05, + "loss": 3.2123, + "step": 7549 + }, + { + "epoch": 0.23289081507929468, + "grad_norm": 0.25048401951789856, + "learning_rate": 4.73581506316944e-05, + "loss": 3.2252, + "step": 7550 + }, + { + "epoch": 0.23292166154486812, + "grad_norm": 0.15582245588302612, + "learning_rate": 4.73569460888575e-05, + "loss": 3.2297, + "step": 7551 + }, + { + "epoch": 0.23295250801044154, + "grad_norm": 0.1607741117477417, + "learning_rate": 4.735574128680525e-05, + "loss": 3.248, + "step": 7552 + }, + { + "epoch": 0.23298335447601495, + "grad_norm": 0.15040865540504456, + "learning_rate": 4.735453622555161e-05, + "loss": 3.2117, + "step": 7553 + }, + { + "epoch": 0.23301420094158837, + "grad_norm": 0.15669775009155273, + "learning_rate": 4.735333090511056e-05, + "loss": 3.2212, + "step": 7554 + }, + { + "epoch": 0.23304504740716178, + "grad_norm": 0.1637960970401764, + "learning_rate": 4.735212532549608e-05, + "loss": 3.1969, + "step": 7555 + }, + { + "epoch": 0.2330758938727352, + "grad_norm": 0.15644237399101257, + "learning_rate": 4.7350919486722126e-05, + "loss": 3.2259, + "step": 7556 + }, + { + "epoch": 0.2331067403383086, + "grad_norm": 0.1623886227607727, + "learning_rate": 4.73497133888027e-05, + "loss": 3.248, + "step": 7557 + }, + { + "epoch": 0.23313758680388202, + "grad_norm": 0.15919585525989532, + "learning_rate": 4.734850703175177e-05, + "loss": 3.2249, + "step": 7558 + }, + { + "epoch": 0.23316843326945544, + "grad_norm": 0.16893193125724792, + "learning_rate": 4.734730041558334e-05, + "loss": 3.2224, + "step": 7559 + }, + { + "epoch": 0.23319927973502885, + "grad_norm": 0.14955763518810272, + "learning_rate": 4.734609354031138e-05, + "loss": 3.1762, + "step": 7560 + }, + { + "epoch": 0.23323012620060227, + "grad_norm": 0.15592841804027557, + "learning_rate": 4.73448864059499e-05, + "loss": 3.1999, + "step": 7561 + }, + { + "epoch": 0.23326097266617568, + "grad_norm": 0.15588966012001038, + "learning_rate": 4.734367901251289e-05, + "loss": 3.1996, + "step": 7562 + }, + { + "epoch": 0.23329181913174912, + "grad_norm": 0.15927524864673615, + "learning_rate": 4.734247136001435e-05, + "loss": 3.2073, + "step": 7563 + }, + { + "epoch": 0.23332266559732254, + "grad_norm": 0.15897776186466217, + "learning_rate": 4.7341263448468274e-05, + "loss": 3.1998, + "step": 7564 + }, + { + "epoch": 0.23335351206289595, + "grad_norm": 0.15545079112052917, + "learning_rate": 4.734005527788868e-05, + "loss": 3.1987, + "step": 7565 + }, + { + "epoch": 0.23338435852846937, + "grad_norm": 0.1530350148677826, + "learning_rate": 4.733884684828957e-05, + "loss": 3.1743, + "step": 7566 + }, + { + "epoch": 0.23341520499404278, + "grad_norm": 0.15217165648937225, + "learning_rate": 4.733763815968496e-05, + "loss": 3.2435, + "step": 7567 + }, + { + "epoch": 0.2334460514596162, + "grad_norm": 0.17709952592849731, + "learning_rate": 4.733642921208885e-05, + "loss": 3.2606, + "step": 7568 + }, + { + "epoch": 0.2334768979251896, + "grad_norm": 0.16262632608413696, + "learning_rate": 4.733522000551527e-05, + "loss": 3.2198, + "step": 7569 + }, + { + "epoch": 0.23350774439076302, + "grad_norm": 0.15548869967460632, + "learning_rate": 4.733401053997824e-05, + "loss": 3.2342, + "step": 7570 + }, + { + "epoch": 0.23353859085633644, + "grad_norm": 0.16248342394828796, + "learning_rate": 4.733280081549176e-05, + "loss": 3.2057, + "step": 7571 + }, + { + "epoch": 0.23356943732190985, + "grad_norm": 0.15035851299762726, + "learning_rate": 4.733159083206989e-05, + "loss": 3.2432, + "step": 7572 + }, + { + "epoch": 0.23360028378748326, + "grad_norm": 0.15480662882328033, + "learning_rate": 4.733038058972664e-05, + "loss": 3.1985, + "step": 7573 + }, + { + "epoch": 0.23363113025305668, + "grad_norm": 0.15190432965755463, + "learning_rate": 4.732917008847605e-05, + "loss": 3.2611, + "step": 7574 + }, + { + "epoch": 0.23366197671863012, + "grad_norm": 0.15531308948993683, + "learning_rate": 4.732795932833215e-05, + "loss": 3.2422, + "step": 7575 + }, + { + "epoch": 0.23369282318420354, + "grad_norm": 0.15016376972198486, + "learning_rate": 4.732674830930898e-05, + "loss": 3.2318, + "step": 7576 + }, + { + "epoch": 0.23372366964977695, + "grad_norm": 0.14934708178043365, + "learning_rate": 4.732553703142058e-05, + "loss": 3.2112, + "step": 7577 + }, + { + "epoch": 0.23375451611535036, + "grad_norm": 0.15365122258663177, + "learning_rate": 4.7324325494680996e-05, + "loss": 3.1977, + "step": 7578 + }, + { + "epoch": 0.23378536258092378, + "grad_norm": 0.14802855253219604, + "learning_rate": 4.732311369910427e-05, + "loss": 3.228, + "step": 7579 + }, + { + "epoch": 0.2338162090464972, + "grad_norm": 0.15286247432231903, + "learning_rate": 4.732190164470446e-05, + "loss": 3.2076, + "step": 7580 + }, + { + "epoch": 0.2338470555120706, + "grad_norm": 0.1592823714017868, + "learning_rate": 4.7320689331495606e-05, + "loss": 3.2145, + "step": 7581 + }, + { + "epoch": 0.23387790197764402, + "grad_norm": 0.1552189290523529, + "learning_rate": 4.7319476759491775e-05, + "loss": 3.1982, + "step": 7582 + }, + { + "epoch": 0.23390874844321743, + "grad_norm": 0.15338526666164398, + "learning_rate": 4.7318263928707025e-05, + "loss": 3.2159, + "step": 7583 + }, + { + "epoch": 0.23393959490879085, + "grad_norm": 0.15282557904720306, + "learning_rate": 4.731705083915542e-05, + "loss": 3.1921, + "step": 7584 + }, + { + "epoch": 0.23397044137436426, + "grad_norm": 0.15604521334171295, + "learning_rate": 4.731583749085101e-05, + "loss": 3.226, + "step": 7585 + }, + { + "epoch": 0.23400128783993768, + "grad_norm": 0.16400788724422455, + "learning_rate": 4.7314623883807884e-05, + "loss": 3.2399, + "step": 7586 + }, + { + "epoch": 0.23403213430551112, + "grad_norm": 0.16693058609962463, + "learning_rate": 4.731341001804011e-05, + "loss": 3.1723, + "step": 7587 + }, + { + "epoch": 0.23406298077108453, + "grad_norm": 0.1461304873228073, + "learning_rate": 4.731219589356174e-05, + "loss": 3.2331, + "step": 7588 + }, + { + "epoch": 0.23409382723665795, + "grad_norm": 0.15229055285453796, + "learning_rate": 4.7310981510386875e-05, + "loss": 3.175, + "step": 7589 + }, + { + "epoch": 0.23412467370223136, + "grad_norm": 0.14998239278793335, + "learning_rate": 4.7309766868529586e-05, + "loss": 3.1989, + "step": 7590 + }, + { + "epoch": 0.23415552016780478, + "grad_norm": 0.14661931991577148, + "learning_rate": 4.730855196800396e-05, + "loss": 3.1828, + "step": 7591 + }, + { + "epoch": 0.2341863666333782, + "grad_norm": 0.15121614933013916, + "learning_rate": 4.7307336808824076e-05, + "loss": 3.2236, + "step": 7592 + }, + { + "epoch": 0.2342172130989516, + "grad_norm": 0.1526281237602234, + "learning_rate": 4.7306121391004024e-05, + "loss": 3.2445, + "step": 7593 + }, + { + "epoch": 0.23424805956452502, + "grad_norm": 0.1522611826658249, + "learning_rate": 4.7304905714557904e-05, + "loss": 3.2065, + "step": 7594 + }, + { + "epoch": 0.23427890603009843, + "grad_norm": 0.14967568218708038, + "learning_rate": 4.7303689779499806e-05, + "loss": 3.2158, + "step": 7595 + }, + { + "epoch": 0.23430975249567185, + "grad_norm": 0.158911794424057, + "learning_rate": 4.730247358584382e-05, + "loss": 3.2127, + "step": 7596 + }, + { + "epoch": 0.23434059896124526, + "grad_norm": 0.15512751042842865, + "learning_rate": 4.730125713360407e-05, + "loss": 3.2023, + "step": 7597 + }, + { + "epoch": 0.23437144542681868, + "grad_norm": 0.15593233704566956, + "learning_rate": 4.730004042279463e-05, + "loss": 3.2356, + "step": 7598 + }, + { + "epoch": 0.23440229189239212, + "grad_norm": 0.18395549058914185, + "learning_rate": 4.729882345342963e-05, + "loss": 3.1918, + "step": 7599 + }, + { + "epoch": 0.23443313835796553, + "grad_norm": 0.15306580066680908, + "learning_rate": 4.7297606225523175e-05, + "loss": 3.193, + "step": 7600 + }, + { + "epoch": 0.23446398482353895, + "grad_norm": 0.16904932260513306, + "learning_rate": 4.729638873908936e-05, + "loss": 3.2268, + "step": 7601 + }, + { + "epoch": 0.23449483128911236, + "grad_norm": 0.15174192190170288, + "learning_rate": 4.7295170994142335e-05, + "loss": 3.2111, + "step": 7602 + }, + { + "epoch": 0.23452567775468577, + "grad_norm": 0.16767841577529907, + "learning_rate": 4.7293952990696186e-05, + "loss": 3.2728, + "step": 7603 + }, + { + "epoch": 0.2345565242202592, + "grad_norm": 0.16863428056240082, + "learning_rate": 4.729273472876506e-05, + "loss": 3.2344, + "step": 7604 + }, + { + "epoch": 0.2345873706858326, + "grad_norm": 0.15290972590446472, + "learning_rate": 4.729151620836306e-05, + "loss": 3.2036, + "step": 7605 + }, + { + "epoch": 0.23461821715140602, + "grad_norm": 0.163411483168602, + "learning_rate": 4.7290297429504335e-05, + "loss": 3.221, + "step": 7606 + }, + { + "epoch": 0.23464906361697943, + "grad_norm": 0.15825004875659943, + "learning_rate": 4.7289078392203e-05, + "loss": 3.2066, + "step": 7607 + }, + { + "epoch": 0.23467991008255285, + "grad_norm": 0.19931595027446747, + "learning_rate": 4.7287859096473197e-05, + "loss": 3.1715, + "step": 7608 + }, + { + "epoch": 0.23471075654812626, + "grad_norm": 0.1526259481906891, + "learning_rate": 4.728663954232906e-05, + "loss": 3.216, + "step": 7609 + }, + { + "epoch": 0.23474160301369967, + "grad_norm": 0.16393768787384033, + "learning_rate": 4.728541972978474e-05, + "loss": 3.2284, + "step": 7610 + }, + { + "epoch": 0.23477244947927312, + "grad_norm": 0.1545591950416565, + "learning_rate": 4.728419965885436e-05, + "loss": 3.2319, + "step": 7611 + }, + { + "epoch": 0.23480329594484653, + "grad_norm": 0.1617555022239685, + "learning_rate": 4.728297932955208e-05, + "loss": 3.2285, + "step": 7612 + }, + { + "epoch": 0.23483414241041994, + "grad_norm": 0.15771640837192535, + "learning_rate": 4.7281758741892045e-05, + "loss": 3.1919, + "step": 7613 + }, + { + "epoch": 0.23486498887599336, + "grad_norm": 0.17590869963169098, + "learning_rate": 4.72805378958884e-05, + "loss": 3.2168, + "step": 7614 + }, + { + "epoch": 0.23489583534156677, + "grad_norm": 0.1545969396829605, + "learning_rate": 4.7279316791555315e-05, + "loss": 3.2101, + "step": 7615 + }, + { + "epoch": 0.2349266818071402, + "grad_norm": 0.16130924224853516, + "learning_rate": 4.727809542890694e-05, + "loss": 3.2402, + "step": 7616 + }, + { + "epoch": 0.2349575282727136, + "grad_norm": 0.1608303040266037, + "learning_rate": 4.7276873807957434e-05, + "loss": 3.1961, + "step": 7617 + }, + { + "epoch": 0.23498837473828701, + "grad_norm": 0.15284520387649536, + "learning_rate": 4.7275651928720965e-05, + "loss": 3.235, + "step": 7618 + }, + { + "epoch": 0.23501922120386043, + "grad_norm": 0.17105911672115326, + "learning_rate": 4.7274429791211696e-05, + "loss": 3.2111, + "step": 7619 + }, + { + "epoch": 0.23505006766943384, + "grad_norm": 0.15208065509796143, + "learning_rate": 4.7273207395443795e-05, + "loss": 3.2082, + "step": 7620 + }, + { + "epoch": 0.23508091413500726, + "grad_norm": 0.17615418136119843, + "learning_rate": 4.7271984741431446e-05, + "loss": 3.2148, + "step": 7621 + }, + { + "epoch": 0.23511176060058067, + "grad_norm": 0.1511700451374054, + "learning_rate": 4.727076182918882e-05, + "loss": 3.213, + "step": 7622 + }, + { + "epoch": 0.2351426070661541, + "grad_norm": 0.1803281307220459, + "learning_rate": 4.7269538658730085e-05, + "loss": 3.2163, + "step": 7623 + }, + { + "epoch": 0.23517345353172753, + "grad_norm": 0.14854736626148224, + "learning_rate": 4.726831523006944e-05, + "loss": 3.2278, + "step": 7624 + }, + { + "epoch": 0.23520429999730094, + "grad_norm": 0.16615162789821625, + "learning_rate": 4.7267091543221056e-05, + "loss": 3.191, + "step": 7625 + }, + { + "epoch": 0.23523514646287436, + "grad_norm": 0.15854512155056, + "learning_rate": 4.7265867598199124e-05, + "loss": 3.2343, + "step": 7626 + }, + { + "epoch": 0.23526599292844777, + "grad_norm": 0.15315167605876923, + "learning_rate": 4.726464339501784e-05, + "loss": 3.2153, + "step": 7627 + }, + { + "epoch": 0.23529683939402118, + "grad_norm": 0.15803106129169464, + "learning_rate": 4.726341893369139e-05, + "loss": 3.2323, + "step": 7628 + }, + { + "epoch": 0.2353276858595946, + "grad_norm": 0.1541263312101364, + "learning_rate": 4.726219421423399e-05, + "loss": 3.2148, + "step": 7629 + }, + { + "epoch": 0.235358532325168, + "grad_norm": 0.16104544699192047, + "learning_rate": 4.726096923665982e-05, + "loss": 3.2223, + "step": 7630 + }, + { + "epoch": 0.23538937879074143, + "grad_norm": 0.1601909101009369, + "learning_rate": 4.7259744000983085e-05, + "loss": 3.245, + "step": 7631 + }, + { + "epoch": 0.23542022525631484, + "grad_norm": 0.1573406606912613, + "learning_rate": 4.7258518507218e-05, + "loss": 3.2178, + "step": 7632 + }, + { + "epoch": 0.23545107172188826, + "grad_norm": 0.15416549146175385, + "learning_rate": 4.725729275537877e-05, + "loss": 3.2067, + "step": 7633 + }, + { + "epoch": 0.23548191818746167, + "grad_norm": 0.14735297858715057, + "learning_rate": 4.72560667454796e-05, + "loss": 3.2042, + "step": 7634 + }, + { + "epoch": 0.2355127646530351, + "grad_norm": 0.15228746831417084, + "learning_rate": 4.725484047753471e-05, + "loss": 3.2153, + "step": 7635 + }, + { + "epoch": 0.23554361111860853, + "grad_norm": 0.15118113160133362, + "learning_rate": 4.725361395155832e-05, + "loss": 3.2073, + "step": 7636 + }, + { + "epoch": 0.23557445758418194, + "grad_norm": 0.1483575403690338, + "learning_rate": 4.725238716756465e-05, + "loss": 3.1983, + "step": 7637 + }, + { + "epoch": 0.23560530404975535, + "grad_norm": 0.14805275201797485, + "learning_rate": 4.725116012556792e-05, + "loss": 3.2353, + "step": 7638 + }, + { + "epoch": 0.23563615051532877, + "grad_norm": 0.15001560747623444, + "learning_rate": 4.7249932825582364e-05, + "loss": 3.2434, + "step": 7639 + }, + { + "epoch": 0.23566699698090218, + "grad_norm": 0.1848364621400833, + "learning_rate": 4.724870526762221e-05, + "loss": 3.2156, + "step": 7640 + }, + { + "epoch": 0.2356978434464756, + "grad_norm": 0.15167103707790375, + "learning_rate": 4.724747745170168e-05, + "loss": 3.2005, + "step": 7641 + }, + { + "epoch": 0.235728689912049, + "grad_norm": 0.16249625384807587, + "learning_rate": 4.724624937783503e-05, + "loss": 3.1486, + "step": 7642 + }, + { + "epoch": 0.23575953637762243, + "grad_norm": 0.14625827968120575, + "learning_rate": 4.7245021046036474e-05, + "loss": 3.2684, + "step": 7643 + }, + { + "epoch": 0.23579038284319584, + "grad_norm": 0.16686001420021057, + "learning_rate": 4.7243792456320265e-05, + "loss": 3.2384, + "step": 7644 + }, + { + "epoch": 0.23582122930876925, + "grad_norm": 0.14617714285850525, + "learning_rate": 4.7242563608700664e-05, + "loss": 3.2145, + "step": 7645 + }, + { + "epoch": 0.23585207577434267, + "grad_norm": 0.16376246511936188, + "learning_rate": 4.7241334503191893e-05, + "loss": 3.199, + "step": 7646 + }, + { + "epoch": 0.2358829222399161, + "grad_norm": 0.15029974281787872, + "learning_rate": 4.7240105139808225e-05, + "loss": 3.191, + "step": 7647 + }, + { + "epoch": 0.23591376870548952, + "grad_norm": 0.1582373082637787, + "learning_rate": 4.723887551856389e-05, + "loss": 3.2119, + "step": 7648 + }, + { + "epoch": 0.23594461517106294, + "grad_norm": 0.15777547657489777, + "learning_rate": 4.723764563947316e-05, + "loss": 3.2151, + "step": 7649 + }, + { + "epoch": 0.23597546163663635, + "grad_norm": 0.16140949726104736, + "learning_rate": 4.7236415502550294e-05, + "loss": 3.2054, + "step": 7650 + }, + { + "epoch": 0.23600630810220977, + "grad_norm": 0.15849769115447998, + "learning_rate": 4.723518510780955e-05, + "loss": 3.183, + "step": 7651 + }, + { + "epoch": 0.23603715456778318, + "grad_norm": 0.16104158759117126, + "learning_rate": 4.72339544552652e-05, + "loss": 3.2408, + "step": 7652 + }, + { + "epoch": 0.2360680010333566, + "grad_norm": 0.1624433994293213, + "learning_rate": 4.7232723544931513e-05, + "loss": 3.2287, + "step": 7653 + }, + { + "epoch": 0.23609884749893, + "grad_norm": 0.15382260084152222, + "learning_rate": 4.7231492376822754e-05, + "loss": 3.2009, + "step": 7654 + }, + { + "epoch": 0.23612969396450342, + "grad_norm": 0.15964174270629883, + "learning_rate": 4.7230260950953195e-05, + "loss": 3.1908, + "step": 7655 + }, + { + "epoch": 0.23616054043007684, + "grad_norm": 0.15820112824440002, + "learning_rate": 4.722902926733713e-05, + "loss": 3.2215, + "step": 7656 + }, + { + "epoch": 0.23619138689565025, + "grad_norm": 0.15496376156806946, + "learning_rate": 4.722779732598882e-05, + "loss": 3.201, + "step": 7657 + }, + { + "epoch": 0.23622223336122367, + "grad_norm": 0.15715236961841583, + "learning_rate": 4.7226565126922565e-05, + "loss": 3.2115, + "step": 7658 + }, + { + "epoch": 0.2362530798267971, + "grad_norm": 0.157684788107872, + "learning_rate": 4.722533267015263e-05, + "loss": 3.2044, + "step": 7659 + }, + { + "epoch": 0.23628392629237052, + "grad_norm": 0.16601231694221497, + "learning_rate": 4.7224099955693334e-05, + "loss": 3.2095, + "step": 7660 + }, + { + "epoch": 0.23631477275794394, + "grad_norm": 0.14698071777820587, + "learning_rate": 4.722286698355894e-05, + "loss": 3.2335, + "step": 7661 + }, + { + "epoch": 0.23634561922351735, + "grad_norm": 0.15313293039798737, + "learning_rate": 4.722163375376377e-05, + "loss": 3.2191, + "step": 7662 + }, + { + "epoch": 0.23637646568909076, + "grad_norm": 0.15852327644824982, + "learning_rate": 4.7220400266322106e-05, + "loss": 3.2007, + "step": 7663 + }, + { + "epoch": 0.23640731215466418, + "grad_norm": 0.1506379395723343, + "learning_rate": 4.721916652124825e-05, + "loss": 3.1984, + "step": 7664 + }, + { + "epoch": 0.2364381586202376, + "grad_norm": 0.16755405068397522, + "learning_rate": 4.721793251855652e-05, + "loss": 3.2406, + "step": 7665 + }, + { + "epoch": 0.236469005085811, + "grad_norm": 0.1534949243068695, + "learning_rate": 4.7216698258261205e-05, + "loss": 3.2172, + "step": 7666 + }, + { + "epoch": 0.23649985155138442, + "grad_norm": 0.16609159111976624, + "learning_rate": 4.721546374037662e-05, + "loss": 3.2463, + "step": 7667 + }, + { + "epoch": 0.23653069801695784, + "grad_norm": 0.15099860727787018, + "learning_rate": 4.7214228964917093e-05, + "loss": 3.2265, + "step": 7668 + }, + { + "epoch": 0.23656154448253125, + "grad_norm": 0.16485275328159332, + "learning_rate": 4.7212993931896925e-05, + "loss": 3.1964, + "step": 7669 + }, + { + "epoch": 0.23659239094810466, + "grad_norm": 0.1541009396314621, + "learning_rate": 4.7211758641330446e-05, + "loss": 3.187, + "step": 7670 + }, + { + "epoch": 0.2366232374136781, + "grad_norm": 0.15868230164051056, + "learning_rate": 4.721052309323197e-05, + "loss": 3.2271, + "step": 7671 + }, + { + "epoch": 0.23665408387925152, + "grad_norm": 0.15944023430347443, + "learning_rate": 4.7209287287615824e-05, + "loss": 3.2015, + "step": 7672 + }, + { + "epoch": 0.23668493034482493, + "grad_norm": 0.1560000330209732, + "learning_rate": 4.7208051224496344e-05, + "loss": 3.222, + "step": 7673 + }, + { + "epoch": 0.23671577681039835, + "grad_norm": 0.23340849578380585, + "learning_rate": 4.720681490388785e-05, + "loss": 3.1772, + "step": 7674 + }, + { + "epoch": 0.23674662327597176, + "grad_norm": 0.17542679607868195, + "learning_rate": 4.720557832580468e-05, + "loss": 3.2062, + "step": 7675 + }, + { + "epoch": 0.23677746974154518, + "grad_norm": 0.16697867214679718, + "learning_rate": 4.720434149026117e-05, + "loss": 3.1765, + "step": 7676 + }, + { + "epoch": 0.2368083162071186, + "grad_norm": 0.19594626128673553, + "learning_rate": 4.720310439727168e-05, + "loss": 3.2202, + "step": 7677 + }, + { + "epoch": 0.236839162672692, + "grad_norm": 0.16684599220752716, + "learning_rate": 4.720186704685052e-05, + "loss": 3.2021, + "step": 7678 + }, + { + "epoch": 0.23687000913826542, + "grad_norm": 0.17270785570144653, + "learning_rate": 4.720062943901207e-05, + "loss": 3.2576, + "step": 7679 + }, + { + "epoch": 0.23690085560383883, + "grad_norm": 0.17385107278823853, + "learning_rate": 4.719939157377065e-05, + "loss": 3.2424, + "step": 7680 + }, + { + "epoch": 0.23693170206941225, + "grad_norm": 0.1573985368013382, + "learning_rate": 4.7198153451140627e-05, + "loss": 3.2029, + "step": 7681 + }, + { + "epoch": 0.23696254853498566, + "grad_norm": 0.17969320714473724, + "learning_rate": 4.719691507113636e-05, + "loss": 3.2058, + "step": 7682 + }, + { + "epoch": 0.2369933950005591, + "grad_norm": 0.1587412804365158, + "learning_rate": 4.719567643377219e-05, + "loss": 3.2016, + "step": 7683 + }, + { + "epoch": 0.23702424146613252, + "grad_norm": 0.16994328796863556, + "learning_rate": 4.71944375390625e-05, + "loss": 3.2302, + "step": 7684 + }, + { + "epoch": 0.23705508793170593, + "grad_norm": 0.16472773253917694, + "learning_rate": 4.719319838702164e-05, + "loss": 3.2084, + "step": 7685 + }, + { + "epoch": 0.23708593439727935, + "grad_norm": 0.1682339757680893, + "learning_rate": 4.719195897766398e-05, + "loss": 3.2286, + "step": 7686 + }, + { + "epoch": 0.23711678086285276, + "grad_norm": 0.1662929505109787, + "learning_rate": 4.7190719311003895e-05, + "loss": 3.2272, + "step": 7687 + }, + { + "epoch": 0.23714762732842617, + "grad_norm": 0.1572415828704834, + "learning_rate": 4.718947938705576e-05, + "loss": 3.2181, + "step": 7688 + }, + { + "epoch": 0.2371784737939996, + "grad_norm": 0.15755386650562286, + "learning_rate": 4.7188239205833936e-05, + "loss": 3.2482, + "step": 7689 + }, + { + "epoch": 0.237209320259573, + "grad_norm": 0.1603802591562271, + "learning_rate": 4.718699876735281e-05, + "loss": 3.2162, + "step": 7690 + }, + { + "epoch": 0.23724016672514642, + "grad_norm": 0.15712785720825195, + "learning_rate": 4.7185758071626776e-05, + "loss": 3.2317, + "step": 7691 + }, + { + "epoch": 0.23727101319071983, + "grad_norm": 0.14732612669467926, + "learning_rate": 4.71845171186702e-05, + "loss": 3.2068, + "step": 7692 + }, + { + "epoch": 0.23730185965629325, + "grad_norm": 0.15585605800151825, + "learning_rate": 4.7183275908497484e-05, + "loss": 3.1903, + "step": 7693 + }, + { + "epoch": 0.23733270612186666, + "grad_norm": 0.16560354828834534, + "learning_rate": 4.718203444112301e-05, + "loss": 3.2138, + "step": 7694 + }, + { + "epoch": 0.2373635525874401, + "grad_norm": 0.1612272709608078, + "learning_rate": 4.718079271656118e-05, + "loss": 3.2408, + "step": 7695 + }, + { + "epoch": 0.23739439905301352, + "grad_norm": 0.15608692169189453, + "learning_rate": 4.717955073482639e-05, + "loss": 3.2175, + "step": 7696 + }, + { + "epoch": 0.23742524551858693, + "grad_norm": 0.15971308946609497, + "learning_rate": 4.7178308495933044e-05, + "loss": 3.1986, + "step": 7697 + }, + { + "epoch": 0.23745609198416034, + "grad_norm": 0.15758097171783447, + "learning_rate": 4.717706599989553e-05, + "loss": 3.2253, + "step": 7698 + }, + { + "epoch": 0.23748693844973376, + "grad_norm": 2.4596121311187744, + "learning_rate": 4.7175823246728256e-05, + "loss": 3.2053, + "step": 7699 + }, + { + "epoch": 0.23751778491530717, + "grad_norm": 0.17388944327831268, + "learning_rate": 4.7174580236445655e-05, + "loss": 3.2465, + "step": 7700 + }, + { + "epoch": 0.2375486313808806, + "grad_norm": 0.16209568083286285, + "learning_rate": 4.717333696906211e-05, + "loss": 3.2109, + "step": 7701 + }, + { + "epoch": 0.237579477846454, + "grad_norm": 0.16490626335144043, + "learning_rate": 4.7172093444592045e-05, + "loss": 3.2071, + "step": 7702 + }, + { + "epoch": 0.23761032431202742, + "grad_norm": 0.16172561049461365, + "learning_rate": 4.717084966304988e-05, + "loss": 3.202, + "step": 7703 + }, + { + "epoch": 0.23764117077760083, + "grad_norm": 0.16585856676101685, + "learning_rate": 4.716960562445004e-05, + "loss": 3.2067, + "step": 7704 + }, + { + "epoch": 0.23767201724317424, + "grad_norm": 0.16012513637542725, + "learning_rate": 4.716836132880695e-05, + "loss": 3.1762, + "step": 7705 + }, + { + "epoch": 0.23770286370874769, + "grad_norm": 0.21138419210910797, + "learning_rate": 4.7167116776135024e-05, + "loss": 3.2406, + "step": 7706 + }, + { + "epoch": 0.2377337101743211, + "grad_norm": 0.1570005565881729, + "learning_rate": 4.7165871966448705e-05, + "loss": 3.2352, + "step": 7707 + }, + { + "epoch": 0.23776455663989451, + "grad_norm": 0.16032643616199493, + "learning_rate": 4.716462689976242e-05, + "loss": 3.2412, + "step": 7708 + }, + { + "epoch": 0.23779540310546793, + "grad_norm": 0.15872062742710114, + "learning_rate": 4.7163381576090605e-05, + "loss": 3.1745, + "step": 7709 + }, + { + "epoch": 0.23782624957104134, + "grad_norm": 0.1623506247997284, + "learning_rate": 4.71621359954477e-05, + "loss": 3.2276, + "step": 7710 + }, + { + "epoch": 0.23785709603661476, + "grad_norm": 0.15497057139873505, + "learning_rate": 4.716089015784815e-05, + "loss": 3.2277, + "step": 7711 + }, + { + "epoch": 0.23788794250218817, + "grad_norm": 0.1604350358247757, + "learning_rate": 4.7159644063306384e-05, + "loss": 3.2285, + "step": 7712 + }, + { + "epoch": 0.23791878896776159, + "grad_norm": 0.1544296145439148, + "learning_rate": 4.7158397711836864e-05, + "loss": 3.1922, + "step": 7713 + }, + { + "epoch": 0.237949635433335, + "grad_norm": 0.15891984105110168, + "learning_rate": 4.7157151103454045e-05, + "loss": 3.261, + "step": 7714 + }, + { + "epoch": 0.2379804818989084, + "grad_norm": 0.1597582846879959, + "learning_rate": 4.7155904238172374e-05, + "loss": 3.2319, + "step": 7715 + }, + { + "epoch": 0.23801132836448183, + "grad_norm": 0.155729278922081, + "learning_rate": 4.7154657116006305e-05, + "loss": 3.2502, + "step": 7716 + }, + { + "epoch": 0.23804217483005524, + "grad_norm": 0.16136986017227173, + "learning_rate": 4.71534097369703e-05, + "loss": 3.2342, + "step": 7717 + }, + { + "epoch": 0.23807302129562868, + "grad_norm": 0.15176846086978912, + "learning_rate": 4.715216210107882e-05, + "loss": 3.2063, + "step": 7718 + }, + { + "epoch": 0.2381038677612021, + "grad_norm": 0.15271282196044922, + "learning_rate": 4.7150914208346335e-05, + "loss": 3.1577, + "step": 7719 + }, + { + "epoch": 0.2381347142267755, + "grad_norm": 0.15840519964694977, + "learning_rate": 4.714966605878731e-05, + "loss": 3.2125, + "step": 7720 + }, + { + "epoch": 0.23816556069234893, + "grad_norm": 0.15578272938728333, + "learning_rate": 4.714841765241622e-05, + "loss": 3.248, + "step": 7721 + }, + { + "epoch": 0.23819640715792234, + "grad_norm": 0.16599604487419128, + "learning_rate": 4.714716898924754e-05, + "loss": 3.1875, + "step": 7722 + }, + { + "epoch": 0.23822725362349575, + "grad_norm": 0.1578877568244934, + "learning_rate": 4.714592006929573e-05, + "loss": 3.2251, + "step": 7723 + }, + { + "epoch": 0.23825810008906917, + "grad_norm": 0.15311042964458466, + "learning_rate": 4.7144670892575294e-05, + "loss": 3.1944, + "step": 7724 + }, + { + "epoch": 0.23828894655464258, + "grad_norm": 0.1754216104745865, + "learning_rate": 4.7143421459100705e-05, + "loss": 3.2363, + "step": 7725 + }, + { + "epoch": 0.238319793020216, + "grad_norm": 0.15276360511779785, + "learning_rate": 4.714217176888646e-05, + "loss": 3.1886, + "step": 7726 + }, + { + "epoch": 0.2383506394857894, + "grad_norm": 0.24610061943531036, + "learning_rate": 4.714092182194703e-05, + "loss": 3.2081, + "step": 7727 + }, + { + "epoch": 0.23838148595136283, + "grad_norm": 0.1630566567182541, + "learning_rate": 4.713967161829691e-05, + "loss": 3.1853, + "step": 7728 + }, + { + "epoch": 0.23841233241693624, + "grad_norm": 0.15684989094734192, + "learning_rate": 4.713842115795061e-05, + "loss": 3.2055, + "step": 7729 + }, + { + "epoch": 0.23844317888250968, + "grad_norm": 0.16395385563373566, + "learning_rate": 4.7137170440922626e-05, + "loss": 3.1976, + "step": 7730 + }, + { + "epoch": 0.2384740253480831, + "grad_norm": 0.15214233100414276, + "learning_rate": 4.713591946722744e-05, + "loss": 3.2026, + "step": 7731 + }, + { + "epoch": 0.2385048718136565, + "grad_norm": 0.15612739324569702, + "learning_rate": 4.713466823687958e-05, + "loss": 3.2319, + "step": 7732 + }, + { + "epoch": 0.23853571827922992, + "grad_norm": 0.16143806278705597, + "learning_rate": 4.7133416749893544e-05, + "loss": 3.2088, + "step": 7733 + }, + { + "epoch": 0.23856656474480334, + "grad_norm": 0.15412484109401703, + "learning_rate": 4.713216500628383e-05, + "loss": 3.2066, + "step": 7734 + }, + { + "epoch": 0.23859741121037675, + "grad_norm": 0.1500757783651352, + "learning_rate": 4.713091300606498e-05, + "loss": 3.2031, + "step": 7735 + }, + { + "epoch": 0.23862825767595017, + "grad_norm": 0.15870916843414307, + "learning_rate": 4.712966074925148e-05, + "loss": 3.2272, + "step": 7736 + }, + { + "epoch": 0.23865910414152358, + "grad_norm": 0.15177099406719208, + "learning_rate": 4.7128408235857865e-05, + "loss": 3.2351, + "step": 7737 + }, + { + "epoch": 0.238689950607097, + "grad_norm": 0.15433648228645325, + "learning_rate": 4.712715546589866e-05, + "loss": 3.2117, + "step": 7738 + }, + { + "epoch": 0.2387207970726704, + "grad_norm": 0.15830330550670624, + "learning_rate": 4.7125902439388375e-05, + "loss": 3.2344, + "step": 7739 + }, + { + "epoch": 0.23875164353824382, + "grad_norm": 0.152256041765213, + "learning_rate": 4.712464915634155e-05, + "loss": 3.1886, + "step": 7740 + }, + { + "epoch": 0.23878249000381724, + "grad_norm": 0.1593243032693863, + "learning_rate": 4.7123395616772715e-05, + "loss": 3.2328, + "step": 7741 + }, + { + "epoch": 0.23881333646939068, + "grad_norm": 0.15664632618427277, + "learning_rate": 4.712214182069641e-05, + "loss": 3.2115, + "step": 7742 + }, + { + "epoch": 0.2388441829349641, + "grad_norm": 0.16783928871154785, + "learning_rate": 4.7120887768127155e-05, + "loss": 3.2124, + "step": 7743 + }, + { + "epoch": 0.2388750294005375, + "grad_norm": 0.1545751988887787, + "learning_rate": 4.7119633459079496e-05, + "loss": 3.211, + "step": 7744 + }, + { + "epoch": 0.23890587586611092, + "grad_norm": 0.16359375417232513, + "learning_rate": 4.711837889356798e-05, + "loss": 3.2224, + "step": 7745 + }, + { + "epoch": 0.23893672233168434, + "grad_norm": 0.15769033133983612, + "learning_rate": 4.7117124071607165e-05, + "loss": 3.1773, + "step": 7746 + }, + { + "epoch": 0.23896756879725775, + "grad_norm": 0.1674419641494751, + "learning_rate": 4.7115868993211576e-05, + "loss": 3.1956, + "step": 7747 + }, + { + "epoch": 0.23899841526283117, + "grad_norm": 0.23551174998283386, + "learning_rate": 4.7114613658395776e-05, + "loss": 3.2235, + "step": 7748 + }, + { + "epoch": 0.23902926172840458, + "grad_norm": 0.1629827916622162, + "learning_rate": 4.711335806717433e-05, + "loss": 3.2125, + "step": 7749 + }, + { + "epoch": 0.239060108193978, + "grad_norm": 0.16417936980724335, + "learning_rate": 4.711210221956178e-05, + "loss": 3.2156, + "step": 7750 + }, + { + "epoch": 0.2390909546595514, + "grad_norm": 0.15712569653987885, + "learning_rate": 4.7110846115572694e-05, + "loss": 3.1831, + "step": 7751 + }, + { + "epoch": 0.23912180112512482, + "grad_norm": 0.1550266295671463, + "learning_rate": 4.710958975522163e-05, + "loss": 3.22, + "step": 7752 + }, + { + "epoch": 0.23915264759069824, + "grad_norm": 0.16219930350780487, + "learning_rate": 4.710833313852316e-05, + "loss": 3.2124, + "step": 7753 + }, + { + "epoch": 0.23918349405627168, + "grad_norm": 0.16330890357494354, + "learning_rate": 4.7107076265491855e-05, + "loss": 3.2155, + "step": 7754 + }, + { + "epoch": 0.2392143405218451, + "grad_norm": 0.15630964934825897, + "learning_rate": 4.710581913614228e-05, + "loss": 3.2033, + "step": 7755 + }, + { + "epoch": 0.2392451869874185, + "grad_norm": 0.15139690041542053, + "learning_rate": 4.7104561750489016e-05, + "loss": 3.2249, + "step": 7756 + }, + { + "epoch": 0.23927603345299192, + "grad_norm": 0.15732939541339874, + "learning_rate": 4.710330410854665e-05, + "loss": 3.246, + "step": 7757 + }, + { + "epoch": 0.23930687991856534, + "grad_norm": 0.14959700405597687, + "learning_rate": 4.7102046210329756e-05, + "loss": 3.1881, + "step": 7758 + }, + { + "epoch": 0.23933772638413875, + "grad_norm": 0.15313848853111267, + "learning_rate": 4.7100788055852914e-05, + "loss": 3.2082, + "step": 7759 + }, + { + "epoch": 0.23936857284971216, + "grad_norm": 0.1624545305967331, + "learning_rate": 4.7099529645130716e-05, + "loss": 3.2413, + "step": 7760 + }, + { + "epoch": 0.23939941931528558, + "grad_norm": 0.1465131789445877, + "learning_rate": 4.709827097817775e-05, + "loss": 3.191, + "step": 7761 + }, + { + "epoch": 0.239430265780859, + "grad_norm": 0.1587386578321457, + "learning_rate": 4.709701205500861e-05, + "loss": 3.2348, + "step": 7762 + }, + { + "epoch": 0.2394611122464324, + "grad_norm": 0.15735282003879547, + "learning_rate": 4.70957528756379e-05, + "loss": 3.1609, + "step": 7763 + }, + { + "epoch": 0.23949195871200582, + "grad_norm": 0.15294450521469116, + "learning_rate": 4.709449344008021e-05, + "loss": 3.2139, + "step": 7764 + }, + { + "epoch": 0.23952280517757923, + "grad_norm": 0.15808551013469696, + "learning_rate": 4.709323374835014e-05, + "loss": 3.23, + "step": 7765 + }, + { + "epoch": 0.23955365164315268, + "grad_norm": 0.1574871689081192, + "learning_rate": 4.709197380046231e-05, + "loss": 3.213, + "step": 7766 + }, + { + "epoch": 0.2395844981087261, + "grad_norm": 0.1537720412015915, + "learning_rate": 4.709071359643132e-05, + "loss": 3.1861, + "step": 7767 + }, + { + "epoch": 0.2396153445742995, + "grad_norm": 0.16025961935520172, + "learning_rate": 4.708945313627178e-05, + "loss": 3.2001, + "step": 7768 + }, + { + "epoch": 0.23964619103987292, + "grad_norm": 0.1524161845445633, + "learning_rate": 4.70881924199983e-05, + "loss": 3.2039, + "step": 7769 + }, + { + "epoch": 0.23967703750544633, + "grad_norm": 0.1619616150856018, + "learning_rate": 4.70869314476255e-05, + "loss": 3.2204, + "step": 7770 + }, + { + "epoch": 0.23970788397101975, + "grad_norm": 0.15363222360610962, + "learning_rate": 4.7085670219168007e-05, + "loss": 3.1906, + "step": 7771 + }, + { + "epoch": 0.23973873043659316, + "grad_norm": 0.16000649333000183, + "learning_rate": 4.7084408734640445e-05, + "loss": 3.2047, + "step": 7772 + }, + { + "epoch": 0.23976957690216658, + "grad_norm": 0.16406314074993134, + "learning_rate": 4.708314699405743e-05, + "loss": 3.2332, + "step": 7773 + }, + { + "epoch": 0.23980042336774, + "grad_norm": 0.15913395583629608, + "learning_rate": 4.70818849974336e-05, + "loss": 3.2178, + "step": 7774 + }, + { + "epoch": 0.2398312698333134, + "grad_norm": 0.1741543412208557, + "learning_rate": 4.708062274478357e-05, + "loss": 3.2076, + "step": 7775 + }, + { + "epoch": 0.23986211629888682, + "grad_norm": 0.14734691381454468, + "learning_rate": 4.707936023612201e-05, + "loss": 3.2157, + "step": 7776 + }, + { + "epoch": 0.23989296276446023, + "grad_norm": 0.15894490480422974, + "learning_rate": 4.7078097471463517e-05, + "loss": 3.2469, + "step": 7777 + }, + { + "epoch": 0.23992380923003367, + "grad_norm": 0.15656429529190063, + "learning_rate": 4.707683445082276e-05, + "loss": 3.2483, + "step": 7778 + }, + { + "epoch": 0.2399546556956071, + "grad_norm": 0.1477328985929489, + "learning_rate": 4.707557117421437e-05, + "loss": 3.1925, + "step": 7779 + }, + { + "epoch": 0.2399855021611805, + "grad_norm": 0.16008178889751434, + "learning_rate": 4.7074307641653e-05, + "loss": 3.2176, + "step": 7780 + }, + { + "epoch": 0.24001634862675392, + "grad_norm": 0.14881032705307007, + "learning_rate": 4.7073043853153295e-05, + "loss": 3.2346, + "step": 7781 + }, + { + "epoch": 0.24004719509232733, + "grad_norm": 0.23760615289211273, + "learning_rate": 4.707177980872992e-05, + "loss": 3.2281, + "step": 7782 + }, + { + "epoch": 0.24007804155790075, + "grad_norm": 0.15314257144927979, + "learning_rate": 4.7070515508397514e-05, + "loss": 3.2487, + "step": 7783 + }, + { + "epoch": 0.24010888802347416, + "grad_norm": 0.16435304284095764, + "learning_rate": 4.7069250952170746e-05, + "loss": 3.2469, + "step": 7784 + }, + { + "epoch": 0.24013973448904757, + "grad_norm": 0.15601199865341187, + "learning_rate": 4.7067986140064276e-05, + "loss": 3.2455, + "step": 7785 + }, + { + "epoch": 0.240170580954621, + "grad_norm": 0.15692198276519775, + "learning_rate": 4.706672107209277e-05, + "loss": 3.2368, + "step": 7786 + }, + { + "epoch": 0.2402014274201944, + "grad_norm": 0.15326566994190216, + "learning_rate": 4.7065455748270886e-05, + "loss": 3.2256, + "step": 7787 + }, + { + "epoch": 0.24023227388576782, + "grad_norm": 0.15755297243595123, + "learning_rate": 4.706419016861331e-05, + "loss": 3.205, + "step": 7788 + }, + { + "epoch": 0.24026312035134123, + "grad_norm": 0.15866884589195251, + "learning_rate": 4.70629243331347e-05, + "loss": 3.2104, + "step": 7789 + }, + { + "epoch": 0.24029396681691467, + "grad_norm": 0.15391016006469727, + "learning_rate": 4.706165824184975e-05, + "loss": 3.1854, + "step": 7790 + }, + { + "epoch": 0.2403248132824881, + "grad_norm": 0.1524178385734558, + "learning_rate": 4.706039189477312e-05, + "loss": 3.2255, + "step": 7791 + }, + { + "epoch": 0.2403556597480615, + "grad_norm": 0.17367897927761078, + "learning_rate": 4.705912529191952e-05, + "loss": 3.2082, + "step": 7792 + }, + { + "epoch": 0.24038650621363492, + "grad_norm": 0.15099595487117767, + "learning_rate": 4.7057858433303594e-05, + "loss": 3.1832, + "step": 7793 + }, + { + "epoch": 0.24041735267920833, + "grad_norm": 0.15326382219791412, + "learning_rate": 4.705659131894007e-05, + "loss": 3.2347, + "step": 7794 + }, + { + "epoch": 0.24044819914478174, + "grad_norm": 0.15003108978271484, + "learning_rate": 4.7055323948843624e-05, + "loss": 3.2408, + "step": 7795 + }, + { + "epoch": 0.24047904561035516, + "grad_norm": 0.1470668613910675, + "learning_rate": 4.705405632302895e-05, + "loss": 3.1716, + "step": 7796 + }, + { + "epoch": 0.24050989207592857, + "grad_norm": 0.16880175471305847, + "learning_rate": 4.7052788441510745e-05, + "loss": 3.2467, + "step": 7797 + }, + { + "epoch": 0.24054073854150199, + "grad_norm": 0.1467769891023636, + "learning_rate": 4.705152030430371e-05, + "loss": 3.2235, + "step": 7798 + }, + { + "epoch": 0.2405715850070754, + "grad_norm": 0.1723811775445938, + "learning_rate": 4.705025191142254e-05, + "loss": 3.1998, + "step": 7799 + }, + { + "epoch": 0.24060243147264881, + "grad_norm": 0.15236477553844452, + "learning_rate": 4.7048983262881964e-05, + "loss": 3.2005, + "step": 7800 + }, + { + "epoch": 0.24063327793822223, + "grad_norm": 0.16089601814746857, + "learning_rate": 4.704771435869667e-05, + "loss": 3.2325, + "step": 7801 + }, + { + "epoch": 0.24066412440379567, + "grad_norm": 0.16078904271125793, + "learning_rate": 4.704644519888137e-05, + "loss": 3.2147, + "step": 7802 + }, + { + "epoch": 0.24069497086936908, + "grad_norm": 0.15803228318691254, + "learning_rate": 4.70451757834508e-05, + "loss": 3.2052, + "step": 7803 + }, + { + "epoch": 0.2407258173349425, + "grad_norm": 0.15281210839748383, + "learning_rate": 4.704390611241966e-05, + "loss": 3.2397, + "step": 7804 + }, + { + "epoch": 0.2407566638005159, + "grad_norm": 0.15909947454929352, + "learning_rate": 4.704263618580267e-05, + "loss": 3.2153, + "step": 7805 + }, + { + "epoch": 0.24078751026608933, + "grad_norm": 0.15470977127552032, + "learning_rate": 4.704136600361456e-05, + "loss": 3.1828, + "step": 7806 + }, + { + "epoch": 0.24081835673166274, + "grad_norm": 0.15349622070789337, + "learning_rate": 4.704009556587006e-05, + "loss": 3.2203, + "step": 7807 + }, + { + "epoch": 0.24084920319723616, + "grad_norm": 0.15612149238586426, + "learning_rate": 4.70388248725839e-05, + "loss": 3.2263, + "step": 7808 + }, + { + "epoch": 0.24088004966280957, + "grad_norm": 0.14466406404972076, + "learning_rate": 4.703755392377081e-05, + "loss": 3.2091, + "step": 7809 + }, + { + "epoch": 0.24091089612838298, + "grad_norm": 0.15855775773525238, + "learning_rate": 4.703628271944552e-05, + "loss": 3.2358, + "step": 7810 + }, + { + "epoch": 0.2409417425939564, + "grad_norm": 0.1502097100019455, + "learning_rate": 4.703501125962278e-05, + "loss": 3.1955, + "step": 7811 + }, + { + "epoch": 0.2409725890595298, + "grad_norm": 0.17394207417964935, + "learning_rate": 4.7033739544317314e-05, + "loss": 3.2372, + "step": 7812 + }, + { + "epoch": 0.24100343552510323, + "grad_norm": 0.20627328753471375, + "learning_rate": 4.703246757354388e-05, + "loss": 3.2206, + "step": 7813 + }, + { + "epoch": 0.24103428199067667, + "grad_norm": 0.16185839474201202, + "learning_rate": 4.703119534731724e-05, + "loss": 3.2102, + "step": 7814 + }, + { + "epoch": 0.24106512845625008, + "grad_norm": 0.15435487031936646, + "learning_rate": 4.702992286565211e-05, + "loss": 3.2315, + "step": 7815 + }, + { + "epoch": 0.2410959749218235, + "grad_norm": 0.1552567034959793, + "learning_rate": 4.702865012856328e-05, + "loss": 3.1744, + "step": 7816 + }, + { + "epoch": 0.2411268213873969, + "grad_norm": 0.21293358504772186, + "learning_rate": 4.7027377136065475e-05, + "loss": 3.2323, + "step": 7817 + }, + { + "epoch": 0.24115766785297033, + "grad_norm": 0.15532124042510986, + "learning_rate": 4.702610388817348e-05, + "loss": 3.2069, + "step": 7818 + }, + { + "epoch": 0.24118851431854374, + "grad_norm": 0.252130925655365, + "learning_rate": 4.7024830384902044e-05, + "loss": 3.2482, + "step": 7819 + }, + { + "epoch": 0.24121936078411715, + "grad_norm": 0.15114158391952515, + "learning_rate": 4.702355662626593e-05, + "loss": 3.2324, + "step": 7820 + }, + { + "epoch": 0.24125020724969057, + "grad_norm": 0.16532981395721436, + "learning_rate": 4.702228261227991e-05, + "loss": 3.1798, + "step": 7821 + }, + { + "epoch": 0.24128105371526398, + "grad_norm": 0.16280026733875275, + "learning_rate": 4.7021008342958766e-05, + "loss": 3.2233, + "step": 7822 + }, + { + "epoch": 0.2413119001808374, + "grad_norm": 0.15721526741981506, + "learning_rate": 4.701973381831726e-05, + "loss": 3.2229, + "step": 7823 + }, + { + "epoch": 0.2413427466464108, + "grad_norm": 0.16374845802783966, + "learning_rate": 4.7018459038370164e-05, + "loss": 3.2511, + "step": 7824 + }, + { + "epoch": 0.24137359311198422, + "grad_norm": 0.15426097810268402, + "learning_rate": 4.701718400313228e-05, + "loss": 3.2122, + "step": 7825 + }, + { + "epoch": 0.24140443957755767, + "grad_norm": 0.1656549870967865, + "learning_rate": 4.701590871261837e-05, + "loss": 3.2529, + "step": 7826 + }, + { + "epoch": 0.24143528604313108, + "grad_norm": 0.15512537956237793, + "learning_rate": 4.7014633166843236e-05, + "loss": 3.2201, + "step": 7827 + }, + { + "epoch": 0.2414661325087045, + "grad_norm": 0.15729457139968872, + "learning_rate": 4.701335736582165e-05, + "loss": 3.1982, + "step": 7828 + }, + { + "epoch": 0.2414969789742779, + "grad_norm": 0.15913410484790802, + "learning_rate": 4.701208130956842e-05, + "loss": 3.2281, + "step": 7829 + }, + { + "epoch": 0.24152782543985132, + "grad_norm": 0.1582784354686737, + "learning_rate": 4.701080499809834e-05, + "loss": 3.2099, + "step": 7830 + }, + { + "epoch": 0.24155867190542474, + "grad_norm": 0.15850302577018738, + "learning_rate": 4.70095284314262e-05, + "loss": 3.1874, + "step": 7831 + }, + { + "epoch": 0.24158951837099815, + "grad_norm": 0.1593150943517685, + "learning_rate": 4.70082516095668e-05, + "loss": 3.2222, + "step": 7832 + }, + { + "epoch": 0.24162036483657157, + "grad_norm": 0.15574322640895844, + "learning_rate": 4.700697453253494e-05, + "loss": 3.2301, + "step": 7833 + }, + { + "epoch": 0.24165121130214498, + "grad_norm": 0.15599077939987183, + "learning_rate": 4.700569720034545e-05, + "loss": 3.1877, + "step": 7834 + }, + { + "epoch": 0.2416820577677184, + "grad_norm": 0.15658462047576904, + "learning_rate": 4.700441961301312e-05, + "loss": 3.1994, + "step": 7835 + }, + { + "epoch": 0.2417129042332918, + "grad_norm": 0.16297610104084015, + "learning_rate": 4.700314177055276e-05, + "loss": 3.2489, + "step": 7836 + }, + { + "epoch": 0.24174375069886522, + "grad_norm": 0.15424256026744843, + "learning_rate": 4.700186367297921e-05, + "loss": 3.1783, + "step": 7837 + }, + { + "epoch": 0.24177459716443866, + "grad_norm": 0.15660689771175385, + "learning_rate": 4.7000585320307255e-05, + "loss": 3.2088, + "step": 7838 + }, + { + "epoch": 0.24180544363001208, + "grad_norm": 0.1573457419872284, + "learning_rate": 4.699930671255174e-05, + "loss": 3.1979, + "step": 7839 + }, + { + "epoch": 0.2418362900955855, + "grad_norm": 0.1554226577281952, + "learning_rate": 4.699802784972749e-05, + "loss": 3.2176, + "step": 7840 + }, + { + "epoch": 0.2418671365611589, + "grad_norm": 0.15484850108623505, + "learning_rate": 4.699674873184932e-05, + "loss": 3.2034, + "step": 7841 + }, + { + "epoch": 0.24189798302673232, + "grad_norm": 0.15695948898792267, + "learning_rate": 4.6995469358932064e-05, + "loss": 3.1776, + "step": 7842 + }, + { + "epoch": 0.24192882949230574, + "grad_norm": 0.1834382563829422, + "learning_rate": 4.6994189730990566e-05, + "loss": 3.2526, + "step": 7843 + }, + { + "epoch": 0.24195967595787915, + "grad_norm": 0.15233615040779114, + "learning_rate": 4.6992909848039656e-05, + "loss": 3.2377, + "step": 7844 + }, + { + "epoch": 0.24199052242345256, + "grad_norm": 0.14755086600780487, + "learning_rate": 4.6991629710094166e-05, + "loss": 3.2121, + "step": 7845 + }, + { + "epoch": 0.24202136888902598, + "grad_norm": 0.16686639189720154, + "learning_rate": 4.699034931716895e-05, + "loss": 3.1952, + "step": 7846 + }, + { + "epoch": 0.2420522153545994, + "grad_norm": 0.16689275205135345, + "learning_rate": 4.698906866927885e-05, + "loss": 3.2076, + "step": 7847 + }, + { + "epoch": 0.2420830618201728, + "grad_norm": 0.17658860981464386, + "learning_rate": 4.69877877664387e-05, + "loss": 3.1944, + "step": 7848 + }, + { + "epoch": 0.24211390828574622, + "grad_norm": 0.15065139532089233, + "learning_rate": 4.698650660866337e-05, + "loss": 3.189, + "step": 7849 + }, + { + "epoch": 0.24214475475131966, + "grad_norm": 0.18421584367752075, + "learning_rate": 4.698522519596772e-05, + "loss": 3.2066, + "step": 7850 + }, + { + "epoch": 0.24217560121689308, + "grad_norm": 0.15276385843753815, + "learning_rate": 4.698394352836658e-05, + "loss": 3.1918, + "step": 7851 + }, + { + "epoch": 0.2422064476824665, + "grad_norm": 0.1675918996334076, + "learning_rate": 4.6982661605874834e-05, + "loss": 3.2407, + "step": 7852 + }, + { + "epoch": 0.2422372941480399, + "grad_norm": 0.1520320326089859, + "learning_rate": 4.6981379428507334e-05, + "loss": 3.1963, + "step": 7853 + }, + { + "epoch": 0.24226814061361332, + "grad_norm": 0.16165202856063843, + "learning_rate": 4.698009699627895e-05, + "loss": 3.2293, + "step": 7854 + }, + { + "epoch": 0.24229898707918673, + "grad_norm": 0.15693777799606323, + "learning_rate": 4.697881430920454e-05, + "loss": 3.2229, + "step": 7855 + }, + { + "epoch": 0.24232983354476015, + "grad_norm": 0.14860938489437103, + "learning_rate": 4.6977531367299e-05, + "loss": 3.1472, + "step": 7856 + }, + { + "epoch": 0.24236068001033356, + "grad_norm": 0.17492567002773285, + "learning_rate": 4.6976248170577186e-05, + "loss": 3.2369, + "step": 7857 + }, + { + "epoch": 0.24239152647590698, + "grad_norm": 0.15390832722187042, + "learning_rate": 4.6974964719053985e-05, + "loss": 3.215, + "step": 7858 + }, + { + "epoch": 0.2424223729414804, + "grad_norm": 0.16216939687728882, + "learning_rate": 4.697368101274426e-05, + "loss": 3.233, + "step": 7859 + }, + { + "epoch": 0.2424532194070538, + "grad_norm": 0.1541181057691574, + "learning_rate": 4.697239705166292e-05, + "loss": 3.1915, + "step": 7860 + }, + { + "epoch": 0.24248406587262722, + "grad_norm": 0.1626705676317215, + "learning_rate": 4.697111283582484e-05, + "loss": 3.2106, + "step": 7861 + }, + { + "epoch": 0.24251491233820066, + "grad_norm": 0.15664586424827576, + "learning_rate": 4.6969828365244905e-05, + "loss": 3.2241, + "step": 7862 + }, + { + "epoch": 0.24254575880377408, + "grad_norm": 0.1575997918844223, + "learning_rate": 4.696854363993802e-05, + "loss": 3.2227, + "step": 7863 + }, + { + "epoch": 0.2425766052693475, + "grad_norm": 0.17069493234157562, + "learning_rate": 4.6967258659919065e-05, + "loss": 3.2156, + "step": 7864 + }, + { + "epoch": 0.2426074517349209, + "grad_norm": 0.15399940311908722, + "learning_rate": 4.696597342520295e-05, + "loss": 3.226, + "step": 7865 + }, + { + "epoch": 0.24263829820049432, + "grad_norm": 0.15050077438354492, + "learning_rate": 4.6964687935804574e-05, + "loss": 3.1981, + "step": 7866 + }, + { + "epoch": 0.24266914466606773, + "grad_norm": 0.15784050524234772, + "learning_rate": 4.6963402191738834e-05, + "loss": 3.2031, + "step": 7867 + }, + { + "epoch": 0.24269999113164115, + "grad_norm": 0.15402735769748688, + "learning_rate": 4.696211619302065e-05, + "loss": 3.2253, + "step": 7868 + }, + { + "epoch": 0.24273083759721456, + "grad_norm": 0.1606055051088333, + "learning_rate": 4.696082993966493e-05, + "loss": 3.1691, + "step": 7869 + }, + { + "epoch": 0.24276168406278797, + "grad_norm": 0.15905621647834778, + "learning_rate": 4.6959543431686575e-05, + "loss": 3.2182, + "step": 7870 + }, + { + "epoch": 0.2427925305283614, + "grad_norm": 0.14805737137794495, + "learning_rate": 4.695825666910052e-05, + "loss": 3.1924, + "step": 7871 + }, + { + "epoch": 0.2428233769939348, + "grad_norm": 0.1541564166545868, + "learning_rate": 4.695696965192166e-05, + "loss": 3.2119, + "step": 7872 + }, + { + "epoch": 0.24285422345950822, + "grad_norm": 0.15238900482654572, + "learning_rate": 4.695568238016495e-05, + "loss": 3.2091, + "step": 7873 + }, + { + "epoch": 0.24288506992508166, + "grad_norm": 0.15709704160690308, + "learning_rate": 4.695439485384528e-05, + "loss": 3.2098, + "step": 7874 + }, + { + "epoch": 0.24291591639065507, + "grad_norm": 0.1474246233701706, + "learning_rate": 4.6953107072977596e-05, + "loss": 3.1872, + "step": 7875 + }, + { + "epoch": 0.2429467628562285, + "grad_norm": 0.15571051836013794, + "learning_rate": 4.695181903757684e-05, + "loss": 3.2001, + "step": 7876 + }, + { + "epoch": 0.2429776093218019, + "grad_norm": 0.15205693244934082, + "learning_rate": 4.695053074765792e-05, + "loss": 3.2228, + "step": 7877 + }, + { + "epoch": 0.24300845578737532, + "grad_norm": 0.15063951909542084, + "learning_rate": 4.69492422032358e-05, + "loss": 3.2019, + "step": 7878 + }, + { + "epoch": 0.24303930225294873, + "grad_norm": 0.16404566168785095, + "learning_rate": 4.6947953404325394e-05, + "loss": 3.2293, + "step": 7879 + }, + { + "epoch": 0.24307014871852214, + "grad_norm": 0.15023136138916016, + "learning_rate": 4.6946664350941665e-05, + "loss": 3.2081, + "step": 7880 + }, + { + "epoch": 0.24310099518409556, + "grad_norm": 0.16420413553714752, + "learning_rate": 4.694537504309955e-05, + "loss": 3.2046, + "step": 7881 + }, + { + "epoch": 0.24313184164966897, + "grad_norm": 0.15803460776805878, + "learning_rate": 4.6944085480814006e-05, + "loss": 3.2014, + "step": 7882 + }, + { + "epoch": 0.2431626881152424, + "grad_norm": 0.15262702107429504, + "learning_rate": 4.694279566409997e-05, + "loss": 3.209, + "step": 7883 + }, + { + "epoch": 0.2431935345808158, + "grad_norm": 0.18360695242881775, + "learning_rate": 4.69415055929724e-05, + "loss": 3.2019, + "step": 7884 + }, + { + "epoch": 0.24322438104638922, + "grad_norm": 0.15580499172210693, + "learning_rate": 4.6940215267446265e-05, + "loss": 3.239, + "step": 7885 + }, + { + "epoch": 0.24325522751196266, + "grad_norm": 0.1528800129890442, + "learning_rate": 4.693892468753651e-05, + "loss": 3.2059, + "step": 7886 + }, + { + "epoch": 0.24328607397753607, + "grad_norm": 0.14709119498729706, + "learning_rate": 4.693763385325812e-05, + "loss": 3.2043, + "step": 7887 + }, + { + "epoch": 0.24331692044310949, + "grad_norm": 0.16804082691669464, + "learning_rate": 4.693634276462604e-05, + "loss": 3.226, + "step": 7888 + }, + { + "epoch": 0.2433477669086829, + "grad_norm": 0.18765267729759216, + "learning_rate": 4.6935051421655254e-05, + "loss": 3.2192, + "step": 7889 + }, + { + "epoch": 0.24337861337425631, + "grad_norm": 0.15338855981826782, + "learning_rate": 4.6933759824360724e-05, + "loss": 3.2393, + "step": 7890 + }, + { + "epoch": 0.24340945983982973, + "grad_norm": 0.1601540446281433, + "learning_rate": 4.6932467972757426e-05, + "loss": 3.1998, + "step": 7891 + }, + { + "epoch": 0.24344030630540314, + "grad_norm": 0.15109945833683014, + "learning_rate": 4.693117586686034e-05, + "loss": 3.2157, + "step": 7892 + }, + { + "epoch": 0.24347115277097656, + "grad_norm": 0.15302592515945435, + "learning_rate": 4.6929883506684456e-05, + "loss": 3.1908, + "step": 7893 + }, + { + "epoch": 0.24350199923654997, + "grad_norm": 0.1566639244556427, + "learning_rate": 4.692859089224475e-05, + "loss": 3.2209, + "step": 7894 + }, + { + "epoch": 0.24353284570212339, + "grad_norm": 0.15463578701019287, + "learning_rate": 4.692729802355621e-05, + "loss": 3.2495, + "step": 7895 + }, + { + "epoch": 0.2435636921676968, + "grad_norm": 0.16140058636665344, + "learning_rate": 4.6926004900633824e-05, + "loss": 3.2345, + "step": 7896 + }, + { + "epoch": 0.2435945386332702, + "grad_norm": 0.1468927413225174, + "learning_rate": 4.692471152349259e-05, + "loss": 3.2278, + "step": 7897 + }, + { + "epoch": 0.24362538509884366, + "grad_norm": 0.1596388965845108, + "learning_rate": 4.6923417892147506e-05, + "loss": 3.2071, + "step": 7898 + }, + { + "epoch": 0.24365623156441707, + "grad_norm": 0.15116415917873383, + "learning_rate": 4.6922124006613554e-05, + "loss": 3.2027, + "step": 7899 + }, + { + "epoch": 0.24368707802999048, + "grad_norm": 0.16679681837558746, + "learning_rate": 4.6920829866905756e-05, + "loss": 3.218, + "step": 7900 + }, + { + "epoch": 0.2437179244955639, + "grad_norm": 0.15536463260650635, + "learning_rate": 4.691953547303911e-05, + "loss": 3.2392, + "step": 7901 + }, + { + "epoch": 0.2437487709611373, + "grad_norm": 0.15627148747444153, + "learning_rate": 4.6918240825028614e-05, + "loss": 3.2207, + "step": 7902 + }, + { + "epoch": 0.24377961742671073, + "grad_norm": 0.15085960924625397, + "learning_rate": 4.691694592288929e-05, + "loss": 3.2149, + "step": 7903 + }, + { + "epoch": 0.24381046389228414, + "grad_norm": 0.16203206777572632, + "learning_rate": 4.6915650766636156e-05, + "loss": 3.1654, + "step": 7904 + }, + { + "epoch": 0.24384131035785755, + "grad_norm": 0.16722847521305084, + "learning_rate": 4.691435535628421e-05, + "loss": 3.2182, + "step": 7905 + }, + { + "epoch": 0.24387215682343097, + "grad_norm": 0.16075651347637177, + "learning_rate": 4.6913059691848486e-05, + "loss": 3.2045, + "step": 7906 + }, + { + "epoch": 0.24390300328900438, + "grad_norm": 0.17034676671028137, + "learning_rate": 4.691176377334401e-05, + "loss": 3.2286, + "step": 7907 + }, + { + "epoch": 0.2439338497545778, + "grad_norm": 0.1601322740316391, + "learning_rate": 4.691046760078579e-05, + "loss": 3.2253, + "step": 7908 + }, + { + "epoch": 0.2439646962201512, + "grad_norm": 0.15168625116348267, + "learning_rate": 4.690917117418887e-05, + "loss": 3.2059, + "step": 7909 + }, + { + "epoch": 0.24399554268572465, + "grad_norm": 0.15378299355506897, + "learning_rate": 4.690787449356828e-05, + "loss": 3.2158, + "step": 7910 + }, + { + "epoch": 0.24402638915129807, + "grad_norm": 0.15447089076042175, + "learning_rate": 4.6906577558939044e-05, + "loss": 3.2234, + "step": 7911 + }, + { + "epoch": 0.24405723561687148, + "grad_norm": 0.15731562674045563, + "learning_rate": 4.690528037031621e-05, + "loss": 3.2289, + "step": 7912 + }, + { + "epoch": 0.2440880820824449, + "grad_norm": 0.1477171927690506, + "learning_rate": 4.6903982927714815e-05, + "loss": 3.2158, + "step": 7913 + }, + { + "epoch": 0.2441189285480183, + "grad_norm": 0.1638932228088379, + "learning_rate": 4.69026852311499e-05, + "loss": 3.2292, + "step": 7914 + }, + { + "epoch": 0.24414977501359172, + "grad_norm": 0.154495969414711, + "learning_rate": 4.69013872806365e-05, + "loss": 3.232, + "step": 7915 + }, + { + "epoch": 0.24418062147916514, + "grad_norm": 0.15534575283527374, + "learning_rate": 4.690008907618969e-05, + "loss": 3.1748, + "step": 7916 + }, + { + "epoch": 0.24421146794473855, + "grad_norm": 0.15230083465576172, + "learning_rate": 4.689879061782451e-05, + "loss": 3.2061, + "step": 7917 + }, + { + "epoch": 0.24424231441031197, + "grad_norm": 0.1485542207956314, + "learning_rate": 4.6897491905556e-05, + "loss": 3.1791, + "step": 7918 + }, + { + "epoch": 0.24427316087588538, + "grad_norm": 0.15663330256938934, + "learning_rate": 4.689619293939923e-05, + "loss": 3.2047, + "step": 7919 + }, + { + "epoch": 0.2443040073414588, + "grad_norm": 0.15627555549144745, + "learning_rate": 4.689489371936927e-05, + "loss": 3.2045, + "step": 7920 + }, + { + "epoch": 0.2443348538070322, + "grad_norm": 0.15571311116218567, + "learning_rate": 4.6893594245481164e-05, + "loss": 3.1863, + "step": 7921 + }, + { + "epoch": 0.24436570027260565, + "grad_norm": 0.15139859914779663, + "learning_rate": 4.689229451775e-05, + "loss": 3.1923, + "step": 7922 + }, + { + "epoch": 0.24439654673817907, + "grad_norm": 0.1457633078098297, + "learning_rate": 4.6890994536190835e-05, + "loss": 3.235, + "step": 7923 + }, + { + "epoch": 0.24442739320375248, + "grad_norm": 0.14834216237068176, + "learning_rate": 4.688969430081874e-05, + "loss": 3.1731, + "step": 7924 + }, + { + "epoch": 0.2444582396693259, + "grad_norm": 0.15652139484882355, + "learning_rate": 4.6888393811648794e-05, + "loss": 3.2375, + "step": 7925 + }, + { + "epoch": 0.2444890861348993, + "grad_norm": 0.15138059854507446, + "learning_rate": 4.6887093068696074e-05, + "loss": 3.1919, + "step": 7926 + }, + { + "epoch": 0.24451993260047272, + "grad_norm": 0.147497296333313, + "learning_rate": 4.6885792071975665e-05, + "loss": 3.2341, + "step": 7927 + }, + { + "epoch": 0.24455077906604614, + "grad_norm": 0.18025194108486176, + "learning_rate": 4.688449082150265e-05, + "loss": 3.2035, + "step": 7928 + }, + { + "epoch": 0.24458162553161955, + "grad_norm": 0.15894530713558197, + "learning_rate": 4.688318931729211e-05, + "loss": 3.2359, + "step": 7929 + }, + { + "epoch": 0.24461247199719297, + "grad_norm": 0.16434314846992493, + "learning_rate": 4.688188755935915e-05, + "loss": 3.2729, + "step": 7930 + }, + { + "epoch": 0.24464331846276638, + "grad_norm": 0.14916004240512848, + "learning_rate": 4.6880585547718845e-05, + "loss": 3.2337, + "step": 7931 + }, + { + "epoch": 0.2446741649283398, + "grad_norm": 0.16166280210018158, + "learning_rate": 4.68792832823863e-05, + "loss": 3.1858, + "step": 7932 + }, + { + "epoch": 0.2447050113939132, + "grad_norm": 0.16047179698944092, + "learning_rate": 4.687798076337662e-05, + "loss": 3.2059, + "step": 7933 + }, + { + "epoch": 0.24473585785948665, + "grad_norm": 0.1546086221933365, + "learning_rate": 4.687667799070489e-05, + "loss": 3.218, + "step": 7934 + }, + { + "epoch": 0.24476670432506006, + "grad_norm": 0.1630944013595581, + "learning_rate": 4.687537496438623e-05, + "loss": 3.2171, + "step": 7935 + }, + { + "epoch": 0.24479755079063348, + "grad_norm": 0.1516592651605606, + "learning_rate": 4.687407168443574e-05, + "loss": 3.2571, + "step": 7936 + }, + { + "epoch": 0.2448283972562069, + "grad_norm": 0.1522257924079895, + "learning_rate": 4.687276815086854e-05, + "loss": 3.1914, + "step": 7937 + }, + { + "epoch": 0.2448592437217803, + "grad_norm": 0.15387186408042908, + "learning_rate": 4.687146436369973e-05, + "loss": 3.2369, + "step": 7938 + }, + { + "epoch": 0.24489009018735372, + "grad_norm": 0.16493354737758636, + "learning_rate": 4.687016032294443e-05, + "loss": 3.2868, + "step": 7939 + }, + { + "epoch": 0.24492093665292713, + "grad_norm": 0.15825197100639343, + "learning_rate": 4.686885602861778e-05, + "loss": 3.1962, + "step": 7940 + }, + { + "epoch": 0.24495178311850055, + "grad_norm": 0.1579744815826416, + "learning_rate": 4.6867551480734875e-05, + "loss": 3.2098, + "step": 7941 + }, + { + "epoch": 0.24498262958407396, + "grad_norm": 0.17978361248970032, + "learning_rate": 4.686624667931086e-05, + "loss": 3.2174, + "step": 7942 + }, + { + "epoch": 0.24501347604964738, + "grad_norm": 0.1508253663778305, + "learning_rate": 4.686494162436084e-05, + "loss": 3.2225, + "step": 7943 + }, + { + "epoch": 0.2450443225152208, + "grad_norm": 0.16134899854660034, + "learning_rate": 4.686363631589997e-05, + "loss": 3.2155, + "step": 7944 + }, + { + "epoch": 0.2450751689807942, + "grad_norm": 0.16203096508979797, + "learning_rate": 4.686233075394337e-05, + "loss": 3.2141, + "step": 7945 + }, + { + "epoch": 0.24510601544636765, + "grad_norm": 0.16260847449302673, + "learning_rate": 4.686102493850619e-05, + "loss": 3.2177, + "step": 7946 + }, + { + "epoch": 0.24513686191194106, + "grad_norm": 0.16274109482765198, + "learning_rate": 4.685971886960357e-05, + "loss": 3.1993, + "step": 7947 + }, + { + "epoch": 0.24516770837751448, + "grad_norm": 0.166073277592659, + "learning_rate": 4.685841254725063e-05, + "loss": 3.2381, + "step": 7948 + }, + { + "epoch": 0.2451985548430879, + "grad_norm": 0.1546248495578766, + "learning_rate": 4.685710597146253e-05, + "loss": 3.2111, + "step": 7949 + }, + { + "epoch": 0.2452294013086613, + "grad_norm": 0.1514422446489334, + "learning_rate": 4.685579914225444e-05, + "loss": 3.1793, + "step": 7950 + }, + { + "epoch": 0.24526024777423472, + "grad_norm": 0.1604173481464386, + "learning_rate": 4.6854492059641475e-05, + "loss": 3.2254, + "step": 7951 + }, + { + "epoch": 0.24529109423980813, + "grad_norm": 0.16909116506576538, + "learning_rate": 4.685318472363882e-05, + "loss": 3.2077, + "step": 7952 + }, + { + "epoch": 0.24532194070538155, + "grad_norm": 0.14923204481601715, + "learning_rate": 4.6851877134261615e-05, + "loss": 3.2385, + "step": 7953 + }, + { + "epoch": 0.24535278717095496, + "grad_norm": 0.16285395622253418, + "learning_rate": 4.6850569291525014e-05, + "loss": 3.2233, + "step": 7954 + }, + { + "epoch": 0.24538363363652838, + "grad_norm": 0.16888393461704254, + "learning_rate": 4.6849261195444205e-05, + "loss": 3.22, + "step": 7955 + }, + { + "epoch": 0.2454144801021018, + "grad_norm": 0.1487773209810257, + "learning_rate": 4.684795284603435e-05, + "loss": 3.2076, + "step": 7956 + }, + { + "epoch": 0.2454453265676752, + "grad_norm": 0.17695985734462738, + "learning_rate": 4.68466442433106e-05, + "loss": 3.1641, + "step": 7957 + }, + { + "epoch": 0.24547617303324865, + "grad_norm": 0.22558113932609558, + "learning_rate": 4.6845335387288135e-05, + "loss": 3.1691, + "step": 7958 + }, + { + "epoch": 0.24550701949882206, + "grad_norm": 0.1546066403388977, + "learning_rate": 4.6844026277982144e-05, + "loss": 3.1897, + "step": 7959 + }, + { + "epoch": 0.24553786596439547, + "grad_norm": 0.15677376091480255, + "learning_rate": 4.684271691540779e-05, + "loss": 3.1806, + "step": 7960 + }, + { + "epoch": 0.2455687124299689, + "grad_norm": 0.16059564054012299, + "learning_rate": 4.684140729958026e-05, + "loss": 3.198, + "step": 7961 + }, + { + "epoch": 0.2455995588955423, + "grad_norm": 0.16604898869991302, + "learning_rate": 4.684009743051473e-05, + "loss": 3.1738, + "step": 7962 + }, + { + "epoch": 0.24563040536111572, + "grad_norm": 0.15408207476139069, + "learning_rate": 4.6838787308226406e-05, + "loss": 3.186, + "step": 7963 + }, + { + "epoch": 0.24566125182668913, + "grad_norm": 0.15658724308013916, + "learning_rate": 4.683747693273046e-05, + "loss": 3.1914, + "step": 7964 + }, + { + "epoch": 0.24569209829226255, + "grad_norm": 0.1555565595626831, + "learning_rate": 4.6836166304042094e-05, + "loss": 3.2001, + "step": 7965 + }, + { + "epoch": 0.24572294475783596, + "grad_norm": 0.1619795262813568, + "learning_rate": 4.6834855422176506e-05, + "loss": 3.2253, + "step": 7966 + }, + { + "epoch": 0.24575379122340937, + "grad_norm": 0.1462593823671341, + "learning_rate": 4.683354428714889e-05, + "loss": 3.2203, + "step": 7967 + }, + { + "epoch": 0.2457846376889828, + "grad_norm": 0.15878795087337494, + "learning_rate": 4.683223289897444e-05, + "loss": 3.1818, + "step": 7968 + }, + { + "epoch": 0.2458154841545562, + "grad_norm": 0.15460160374641418, + "learning_rate": 4.683092125766838e-05, + "loss": 3.195, + "step": 7969 + }, + { + "epoch": 0.24584633062012964, + "grad_norm": 0.15216444432735443, + "learning_rate": 4.68296093632459e-05, + "loss": 3.2285, + "step": 7970 + }, + { + "epoch": 0.24587717708570306, + "grad_norm": 0.16430382430553436, + "learning_rate": 4.6828297215722214e-05, + "loss": 3.1872, + "step": 7971 + }, + { + "epoch": 0.24590802355127647, + "grad_norm": 0.15261493623256683, + "learning_rate": 4.682698481511255e-05, + "loss": 3.2231, + "step": 7972 + }, + { + "epoch": 0.2459388700168499, + "grad_norm": 0.16051149368286133, + "learning_rate": 4.68256721614321e-05, + "loss": 3.2177, + "step": 7973 + }, + { + "epoch": 0.2459697164824233, + "grad_norm": 0.1580025553703308, + "learning_rate": 4.6824359254696105e-05, + "loss": 3.2024, + "step": 7974 + }, + { + "epoch": 0.24600056294799671, + "grad_norm": 0.15123115479946136, + "learning_rate": 4.682304609491978e-05, + "loss": 3.2475, + "step": 7975 + }, + { + "epoch": 0.24603140941357013, + "grad_norm": 0.16256311535835266, + "learning_rate": 4.682173268211835e-05, + "loss": 3.1901, + "step": 7976 + }, + { + "epoch": 0.24606225587914354, + "grad_norm": 0.15147408843040466, + "learning_rate": 4.682041901630704e-05, + "loss": 3.2002, + "step": 7977 + }, + { + "epoch": 0.24609310234471696, + "grad_norm": 0.15263725817203522, + "learning_rate": 4.681910509750109e-05, + "loss": 3.2181, + "step": 7978 + }, + { + "epoch": 0.24612394881029037, + "grad_norm": 0.1627875119447708, + "learning_rate": 4.681779092571572e-05, + "loss": 3.1521, + "step": 7979 + }, + { + "epoch": 0.24615479527586379, + "grad_norm": 0.17285461723804474, + "learning_rate": 4.681647650096618e-05, + "loss": 3.2088, + "step": 7980 + }, + { + "epoch": 0.24618564174143723, + "grad_norm": 0.15100598335266113, + "learning_rate": 4.681516182326771e-05, + "loss": 3.1915, + "step": 7981 + }, + { + "epoch": 0.24621648820701064, + "grad_norm": 0.16130584478378296, + "learning_rate": 4.6813846892635536e-05, + "loss": 3.2044, + "step": 7982 + }, + { + "epoch": 0.24624733467258406, + "grad_norm": 0.15016861259937286, + "learning_rate": 4.6812531709084915e-05, + "loss": 3.1953, + "step": 7983 + }, + { + "epoch": 0.24627818113815747, + "grad_norm": 0.18097670376300812, + "learning_rate": 4.681121627263111e-05, + "loss": 3.2263, + "step": 7984 + }, + { + "epoch": 0.24630902760373088, + "grad_norm": 0.15332987904548645, + "learning_rate": 4.680990058328935e-05, + "loss": 3.2352, + "step": 7985 + }, + { + "epoch": 0.2463398740693043, + "grad_norm": 0.14770491421222687, + "learning_rate": 4.68085846410749e-05, + "loss": 3.1959, + "step": 7986 + }, + { + "epoch": 0.2463707205348777, + "grad_norm": 0.1532505452632904, + "learning_rate": 4.680726844600302e-05, + "loss": 3.2188, + "step": 7987 + }, + { + "epoch": 0.24640156700045113, + "grad_norm": 0.16710351407527924, + "learning_rate": 4.6805951998088966e-05, + "loss": 3.1797, + "step": 7988 + }, + { + "epoch": 0.24643241346602454, + "grad_norm": 0.16139386594295502, + "learning_rate": 4.6804635297348e-05, + "loss": 3.2255, + "step": 7989 + }, + { + "epoch": 0.24646325993159796, + "grad_norm": 0.15358705818653107, + "learning_rate": 4.680331834379539e-05, + "loss": 3.2035, + "step": 7990 + }, + { + "epoch": 0.24649410639717137, + "grad_norm": 0.1621120721101761, + "learning_rate": 4.680200113744641e-05, + "loss": 3.2382, + "step": 7991 + }, + { + "epoch": 0.24652495286274478, + "grad_norm": 0.14941321313381195, + "learning_rate": 4.6800683678316325e-05, + "loss": 3.2163, + "step": 7992 + }, + { + "epoch": 0.24655579932831823, + "grad_norm": 0.1628745198249817, + "learning_rate": 4.6799365966420414e-05, + "loss": 3.2191, + "step": 7993 + }, + { + "epoch": 0.24658664579389164, + "grad_norm": 0.15720655024051666, + "learning_rate": 4.6798048001773964e-05, + "loss": 3.2227, + "step": 7994 + }, + { + "epoch": 0.24661749225946505, + "grad_norm": 0.16162818670272827, + "learning_rate": 4.6796729784392236e-05, + "loss": 3.2274, + "step": 7995 + }, + { + "epoch": 0.24664833872503847, + "grad_norm": 0.1613074690103531, + "learning_rate": 4.6795411314290524e-05, + "loss": 3.2105, + "step": 7996 + }, + { + "epoch": 0.24667918519061188, + "grad_norm": 0.16726931929588318, + "learning_rate": 4.6794092591484115e-05, + "loss": 3.184, + "step": 7997 + }, + { + "epoch": 0.2467100316561853, + "grad_norm": 0.16175729036331177, + "learning_rate": 4.679277361598831e-05, + "loss": 3.1997, + "step": 7998 + }, + { + "epoch": 0.2467408781217587, + "grad_norm": 0.15802346169948578, + "learning_rate": 4.679145438781838e-05, + "loss": 3.1909, + "step": 7999 + }, + { + "epoch": 0.24677172458733213, + "grad_norm": 0.18149082362651825, + "learning_rate": 4.6790134906989637e-05, + "loss": 3.2505, + "step": 8000 + }, + { + "epoch": 0.24680257105290554, + "grad_norm": 0.15485021471977234, + "learning_rate": 4.6788815173517365e-05, + "loss": 3.2386, + "step": 8001 + }, + { + "epoch": 0.24683341751847895, + "grad_norm": 0.16757364571094513, + "learning_rate": 4.678749518741689e-05, + "loss": 3.185, + "step": 8002 + }, + { + "epoch": 0.24686426398405237, + "grad_norm": 0.1476469188928604, + "learning_rate": 4.678617494870349e-05, + "loss": 3.1959, + "step": 8003 + }, + { + "epoch": 0.24689511044962578, + "grad_norm": 0.16354873776435852, + "learning_rate": 4.678485445739248e-05, + "loss": 3.2331, + "step": 8004 + }, + { + "epoch": 0.24692595691519922, + "grad_norm": 0.16802382469177246, + "learning_rate": 4.6783533713499186e-05, + "loss": 3.2503, + "step": 8005 + }, + { + "epoch": 0.24695680338077264, + "grad_norm": 0.15069809556007385, + "learning_rate": 4.6782212717038896e-05, + "loss": 3.1818, + "step": 8006 + }, + { + "epoch": 0.24698764984634605, + "grad_norm": 0.15415164828300476, + "learning_rate": 4.6780891468026944e-05, + "loss": 3.214, + "step": 8007 + }, + { + "epoch": 0.24701849631191947, + "grad_norm": 0.16799625754356384, + "learning_rate": 4.677956996647865e-05, + "loss": 3.2213, + "step": 8008 + }, + { + "epoch": 0.24704934277749288, + "grad_norm": 0.1504306048154831, + "learning_rate": 4.6778248212409324e-05, + "loss": 3.2132, + "step": 8009 + }, + { + "epoch": 0.2470801892430663, + "grad_norm": 0.16654027998447418, + "learning_rate": 4.6776926205834306e-05, + "loss": 3.2417, + "step": 8010 + }, + { + "epoch": 0.2471110357086397, + "grad_norm": 0.20035040378570557, + "learning_rate": 4.67756039467689e-05, + "loss": 3.1829, + "step": 8011 + }, + { + "epoch": 0.24714188217421312, + "grad_norm": 0.15089814364910126, + "learning_rate": 4.6774281435228464e-05, + "loss": 3.2307, + "step": 8012 + }, + { + "epoch": 0.24717272863978654, + "grad_norm": 0.15627232193946838, + "learning_rate": 4.677295867122832e-05, + "loss": 3.2005, + "step": 8013 + }, + { + "epoch": 0.24720357510535995, + "grad_norm": 0.1531578153371811, + "learning_rate": 4.67716356547838e-05, + "loss": 3.2392, + "step": 8014 + }, + { + "epoch": 0.24723442157093337, + "grad_norm": 0.16171178221702576, + "learning_rate": 4.6770312385910245e-05, + "loss": 3.1707, + "step": 8015 + }, + { + "epoch": 0.24726526803650678, + "grad_norm": 0.17567189037799835, + "learning_rate": 4.6768988864622995e-05, + "loss": 3.2349, + "step": 8016 + }, + { + "epoch": 0.24729611450208022, + "grad_norm": 0.15031632781028748, + "learning_rate": 4.676766509093742e-05, + "loss": 3.2468, + "step": 8017 + }, + { + "epoch": 0.24732696096765364, + "grad_norm": 0.16067369282245636, + "learning_rate": 4.676634106486884e-05, + "loss": 3.2265, + "step": 8018 + }, + { + "epoch": 0.24735780743322705, + "grad_norm": 0.15530942380428314, + "learning_rate": 4.6765016786432606e-05, + "loss": 3.2151, + "step": 8019 + }, + { + "epoch": 0.24738865389880046, + "grad_norm": 0.15598584711551666, + "learning_rate": 4.676369225564409e-05, + "loss": 3.2168, + "step": 8020 + }, + { + "epoch": 0.24741950036437388, + "grad_norm": 0.15685993432998657, + "learning_rate": 4.676236747251864e-05, + "loss": 3.2436, + "step": 8021 + }, + { + "epoch": 0.2474503468299473, + "grad_norm": 0.15399238467216492, + "learning_rate": 4.676104243707161e-05, + "loss": 3.2142, + "step": 8022 + }, + { + "epoch": 0.2474811932955207, + "grad_norm": 0.1577618569135666, + "learning_rate": 4.675971714931838e-05, + "loss": 3.2147, + "step": 8023 + }, + { + "epoch": 0.24751203976109412, + "grad_norm": 0.15811964869499207, + "learning_rate": 4.67583916092743e-05, + "loss": 3.1858, + "step": 8024 + }, + { + "epoch": 0.24754288622666754, + "grad_norm": 0.15538403391838074, + "learning_rate": 4.675706581695475e-05, + "loss": 3.2062, + "step": 8025 + }, + { + "epoch": 0.24757373269224095, + "grad_norm": 0.15081390738487244, + "learning_rate": 4.6755739772375094e-05, + "loss": 3.1855, + "step": 8026 + }, + { + "epoch": 0.24760457915781436, + "grad_norm": 0.15815818309783936, + "learning_rate": 4.675441347555071e-05, + "loss": 3.2047, + "step": 8027 + }, + { + "epoch": 0.24763542562338778, + "grad_norm": 0.1531575620174408, + "learning_rate": 4.675308692649697e-05, + "loss": 3.2424, + "step": 8028 + }, + { + "epoch": 0.24766627208896122, + "grad_norm": 0.15705962479114532, + "learning_rate": 4.675176012522926e-05, + "loss": 3.1626, + "step": 8029 + }, + { + "epoch": 0.24769711855453463, + "grad_norm": 0.15273158252239227, + "learning_rate": 4.6750433071762965e-05, + "loss": 3.2185, + "step": 8030 + }, + { + "epoch": 0.24772796502010805, + "grad_norm": 0.15327796339988708, + "learning_rate": 4.674910576611347e-05, + "loss": 3.2179, + "step": 8031 + }, + { + "epoch": 0.24775881148568146, + "grad_norm": 0.15219347178936005, + "learning_rate": 4.674777820829616e-05, + "loss": 3.2042, + "step": 8032 + }, + { + "epoch": 0.24778965795125488, + "grad_norm": 0.1531471461057663, + "learning_rate": 4.674645039832643e-05, + "loss": 3.2031, + "step": 8033 + }, + { + "epoch": 0.2478205044168283, + "grad_norm": 0.15417012572288513, + "learning_rate": 4.674512233621968e-05, + "loss": 3.2188, + "step": 8034 + }, + { + "epoch": 0.2478513508824017, + "grad_norm": 0.15078958868980408, + "learning_rate": 4.674379402199131e-05, + "loss": 3.226, + "step": 8035 + }, + { + "epoch": 0.24788219734797512, + "grad_norm": 0.15789221227169037, + "learning_rate": 4.6742465455656705e-05, + "loss": 3.2114, + "step": 8036 + }, + { + "epoch": 0.24791304381354853, + "grad_norm": 0.14891815185546875, + "learning_rate": 4.674113663723127e-05, + "loss": 3.1835, + "step": 8037 + }, + { + "epoch": 0.24794389027912195, + "grad_norm": 0.15065351128578186, + "learning_rate": 4.6739807566730434e-05, + "loss": 3.1883, + "step": 8038 + }, + { + "epoch": 0.24797473674469536, + "grad_norm": 0.14119720458984375, + "learning_rate": 4.673847824416959e-05, + "loss": 3.1658, + "step": 8039 + }, + { + "epoch": 0.24800558321026878, + "grad_norm": 0.156471848487854, + "learning_rate": 4.673714866956416e-05, + "loss": 3.2032, + "step": 8040 + }, + { + "epoch": 0.24803642967584222, + "grad_norm": 0.14586234092712402, + "learning_rate": 4.6735818842929543e-05, + "loss": 3.217, + "step": 8041 + }, + { + "epoch": 0.24806727614141563, + "grad_norm": 0.15464095771312714, + "learning_rate": 4.6734488764281174e-05, + "loss": 3.2172, + "step": 8042 + }, + { + "epoch": 0.24809812260698905, + "grad_norm": 0.14820829033851624, + "learning_rate": 4.673315843363447e-05, + "loss": 3.1757, + "step": 8043 + }, + { + "epoch": 0.24812896907256246, + "grad_norm": 0.1503593474626541, + "learning_rate": 4.673182785100485e-05, + "loss": 3.1893, + "step": 8044 + }, + { + "epoch": 0.24815981553813588, + "grad_norm": 0.15720611810684204, + "learning_rate": 4.673049701640775e-05, + "loss": 3.1882, + "step": 8045 + }, + { + "epoch": 0.2481906620037093, + "grad_norm": 0.15018635988235474, + "learning_rate": 4.672916592985859e-05, + "loss": 3.1963, + "step": 8046 + }, + { + "epoch": 0.2482215084692827, + "grad_norm": 0.16440951824188232, + "learning_rate": 4.672783459137281e-05, + "loss": 3.2122, + "step": 8047 + }, + { + "epoch": 0.24825235493485612, + "grad_norm": 0.163935124874115, + "learning_rate": 4.672650300096585e-05, + "loss": 3.2176, + "step": 8048 + }, + { + "epoch": 0.24828320140042953, + "grad_norm": 0.15799857676029205, + "learning_rate": 4.672517115865314e-05, + "loss": 3.2209, + "step": 8049 + }, + { + "epoch": 0.24831404786600295, + "grad_norm": 0.15298864245414734, + "learning_rate": 4.672383906445013e-05, + "loss": 3.1764, + "step": 8050 + }, + { + "epoch": 0.24834489433157636, + "grad_norm": 0.15278927981853485, + "learning_rate": 4.6722506718372256e-05, + "loss": 3.1825, + "step": 8051 + }, + { + "epoch": 0.24837574079714977, + "grad_norm": 0.16184815764427185, + "learning_rate": 4.672117412043497e-05, + "loss": 3.2014, + "step": 8052 + }, + { + "epoch": 0.24840658726272322, + "grad_norm": 0.1598578840494156, + "learning_rate": 4.6719841270653723e-05, + "loss": 3.1862, + "step": 8053 + }, + { + "epoch": 0.24843743372829663, + "grad_norm": 0.15590356290340424, + "learning_rate": 4.671850816904397e-05, + "loss": 3.1922, + "step": 8054 + }, + { + "epoch": 0.24846828019387004, + "grad_norm": 0.1563921570777893, + "learning_rate": 4.671717481562117e-05, + "loss": 3.2137, + "step": 8055 + }, + { + "epoch": 0.24849912665944346, + "grad_norm": 0.1500825583934784, + "learning_rate": 4.671584121040077e-05, + "loss": 3.1947, + "step": 8056 + }, + { + "epoch": 0.24852997312501687, + "grad_norm": 0.15658052265644073, + "learning_rate": 4.6714507353398247e-05, + "loss": 3.2003, + "step": 8057 + }, + { + "epoch": 0.2485608195905903, + "grad_norm": 0.15345418453216553, + "learning_rate": 4.671317324462906e-05, + "loss": 3.2186, + "step": 8058 + }, + { + "epoch": 0.2485916660561637, + "grad_norm": 0.15314586460590363, + "learning_rate": 4.671183888410867e-05, + "loss": 3.218, + "step": 8059 + }, + { + "epoch": 0.24862251252173712, + "grad_norm": 0.15705737471580505, + "learning_rate": 4.6710504271852565e-05, + "loss": 3.2077, + "step": 8060 + }, + { + "epoch": 0.24865335898731053, + "grad_norm": 0.1512116640806198, + "learning_rate": 4.6709169407876205e-05, + "loss": 3.2159, + "step": 8061 + }, + { + "epoch": 0.24868420545288394, + "grad_norm": 0.1558438241481781, + "learning_rate": 4.6707834292195074e-05, + "loss": 3.2009, + "step": 8062 + }, + { + "epoch": 0.24871505191845736, + "grad_norm": 0.15148837864398956, + "learning_rate": 4.670649892482465e-05, + "loss": 3.2103, + "step": 8063 + }, + { + "epoch": 0.24874589838403077, + "grad_norm": 0.15568284690380096, + "learning_rate": 4.6705163305780405e-05, + "loss": 3.2211, + "step": 8064 + }, + { + "epoch": 0.24877674484960421, + "grad_norm": 0.14920826256275177, + "learning_rate": 4.670382743507784e-05, + "loss": 3.2259, + "step": 8065 + }, + { + "epoch": 0.24880759131517763, + "grad_norm": 0.15357990562915802, + "learning_rate": 4.670249131273244e-05, + "loss": 3.1864, + "step": 8066 + }, + { + "epoch": 0.24883843778075104, + "grad_norm": 0.15510477125644684, + "learning_rate": 4.67011549387597e-05, + "loss": 3.2136, + "step": 8067 + }, + { + "epoch": 0.24886928424632446, + "grad_norm": 0.14937162399291992, + "learning_rate": 4.66998183131751e-05, + "loss": 3.2527, + "step": 8068 + }, + { + "epoch": 0.24890013071189787, + "grad_norm": 0.15330901741981506, + "learning_rate": 4.669848143599416e-05, + "loss": 3.2361, + "step": 8069 + }, + { + "epoch": 0.24893097717747129, + "grad_norm": 0.15958784520626068, + "learning_rate": 4.669714430723235e-05, + "loss": 3.2225, + "step": 8070 + }, + { + "epoch": 0.2489618236430447, + "grad_norm": 0.14677804708480835, + "learning_rate": 4.6695806926905204e-05, + "loss": 3.2295, + "step": 8071 + }, + { + "epoch": 0.2489926701086181, + "grad_norm": 0.14962567389011383, + "learning_rate": 4.669446929502821e-05, + "loss": 3.2142, + "step": 8072 + }, + { + "epoch": 0.24902351657419153, + "grad_norm": 0.14874745905399323, + "learning_rate": 4.6693131411616874e-05, + "loss": 3.1772, + "step": 8073 + }, + { + "epoch": 0.24905436303976494, + "grad_norm": 0.17959950864315033, + "learning_rate": 4.669179327668672e-05, + "loss": 3.1825, + "step": 8074 + }, + { + "epoch": 0.24908520950533836, + "grad_norm": 0.14904336631298065, + "learning_rate": 4.669045489025327e-05, + "loss": 3.1542, + "step": 8075 + }, + { + "epoch": 0.24911605597091177, + "grad_norm": 0.14738920331001282, + "learning_rate": 4.668911625233201e-05, + "loss": 3.2285, + "step": 8076 + }, + { + "epoch": 0.2491469024364852, + "grad_norm": 0.15036270022392273, + "learning_rate": 4.6687777362938494e-05, + "loss": 3.2098, + "step": 8077 + }, + { + "epoch": 0.24917774890205863, + "grad_norm": 0.15088213980197906, + "learning_rate": 4.668643822208823e-05, + "loss": 3.2244, + "step": 8078 + }, + { + "epoch": 0.24920859536763204, + "grad_norm": 0.14769601821899414, + "learning_rate": 4.668509882979675e-05, + "loss": 3.2081, + "step": 8079 + }, + { + "epoch": 0.24923944183320546, + "grad_norm": 0.1488528996706009, + "learning_rate": 4.6683759186079575e-05, + "loss": 3.2112, + "step": 8080 + }, + { + "epoch": 0.24927028829877887, + "grad_norm": 0.15312817692756653, + "learning_rate": 4.668241929095224e-05, + "loss": 3.1729, + "step": 8081 + }, + { + "epoch": 0.24930113476435228, + "grad_norm": 0.14998382329940796, + "learning_rate": 4.668107914443029e-05, + "loss": 3.2235, + "step": 8082 + }, + { + "epoch": 0.2493319812299257, + "grad_norm": 0.15605953335762024, + "learning_rate": 4.667973874652925e-05, + "loss": 3.1982, + "step": 8083 + }, + { + "epoch": 0.2493628276954991, + "grad_norm": 0.15071548521518707, + "learning_rate": 4.6678398097264676e-05, + "loss": 3.2104, + "step": 8084 + }, + { + "epoch": 0.24939367416107253, + "grad_norm": 0.16248518228530884, + "learning_rate": 4.66770571966521e-05, + "loss": 3.2144, + "step": 8085 + }, + { + "epoch": 0.24942452062664594, + "grad_norm": 0.14863274991512299, + "learning_rate": 4.6675716044707065e-05, + "loss": 3.181, + "step": 8086 + }, + { + "epoch": 0.24945536709221935, + "grad_norm": 0.1525280475616455, + "learning_rate": 4.6674374641445126e-05, + "loss": 3.2477, + "step": 8087 + }, + { + "epoch": 0.24948621355779277, + "grad_norm": 0.16260847449302673, + "learning_rate": 4.667303298688185e-05, + "loss": 3.2203, + "step": 8088 + }, + { + "epoch": 0.2495170600233662, + "grad_norm": 0.15162628889083862, + "learning_rate": 4.6671691081032773e-05, + "loss": 3.1962, + "step": 8089 + }, + { + "epoch": 0.24954790648893962, + "grad_norm": 0.2027571201324463, + "learning_rate": 4.667034892391346e-05, + "loss": 3.2162, + "step": 8090 + }, + { + "epoch": 0.24957875295451304, + "grad_norm": 0.14839285612106323, + "learning_rate": 4.666900651553948e-05, + "loss": 3.2142, + "step": 8091 + }, + { + "epoch": 0.24960959942008645, + "grad_norm": 0.14170928299427032, + "learning_rate": 4.666766385592639e-05, + "loss": 3.2008, + "step": 8092 + }, + { + "epoch": 0.24964044588565987, + "grad_norm": 0.15780553221702576, + "learning_rate": 4.6666320945089756e-05, + "loss": 3.2572, + "step": 8093 + }, + { + "epoch": 0.24967129235123328, + "grad_norm": 0.15267299115657806, + "learning_rate": 4.666497778304515e-05, + "loss": 3.2305, + "step": 8094 + }, + { + "epoch": 0.2497021388168067, + "grad_norm": 0.1554846167564392, + "learning_rate": 4.666363436980814e-05, + "loss": 3.2497, + "step": 8095 + }, + { + "epoch": 0.2497329852823801, + "grad_norm": 0.176071897149086, + "learning_rate": 4.6662290705394315e-05, + "loss": 3.2262, + "step": 8096 + }, + { + "epoch": 0.24976383174795352, + "grad_norm": 0.14859262108802795, + "learning_rate": 4.666094678981925e-05, + "loss": 3.2199, + "step": 8097 + }, + { + "epoch": 0.24979467821352694, + "grad_norm": 0.15115149319171906, + "learning_rate": 4.665960262309852e-05, + "loss": 3.1746, + "step": 8098 + }, + { + "epoch": 0.24982552467910035, + "grad_norm": 0.14760559797286987, + "learning_rate": 4.665825820524771e-05, + "loss": 3.2041, + "step": 8099 + }, + { + "epoch": 0.24985637114467377, + "grad_norm": 0.15316930413246155, + "learning_rate": 4.665691353628242e-05, + "loss": 3.1988, + "step": 8100 + }, + { + "epoch": 0.2498872176102472, + "grad_norm": 0.15168200433254242, + "learning_rate": 4.6655568616218224e-05, + "loss": 3.246, + "step": 8101 + }, + { + "epoch": 0.24991806407582062, + "grad_norm": 0.1593446284532547, + "learning_rate": 4.665422344507072e-05, + "loss": 3.2174, + "step": 8102 + }, + { + "epoch": 0.24994891054139404, + "grad_norm": 0.15573076903820038, + "learning_rate": 4.665287802285552e-05, + "loss": 3.1931, + "step": 8103 + }, + { + "epoch": 0.24997975700696745, + "grad_norm": 0.1786538064479828, + "learning_rate": 4.66515323495882e-05, + "loss": 3.2204, + "step": 8104 + }, + { + "epoch": 0.25001060347254084, + "grad_norm": 0.1548336297273636, + "learning_rate": 4.665018642528438e-05, + "loss": 3.2228, + "step": 8105 + }, + { + "epoch": 0.2500414499381143, + "grad_norm": 0.15390223264694214, + "learning_rate": 4.6648840249959656e-05, + "loss": 3.2229, + "step": 8106 + }, + { + "epoch": 0.2500722964036877, + "grad_norm": 0.15522347390651703, + "learning_rate": 4.664749382362964e-05, + "loss": 3.2271, + "step": 8107 + }, + { + "epoch": 0.2501031428692611, + "grad_norm": 0.15038804709911346, + "learning_rate": 4.6646147146309935e-05, + "loss": 3.1896, + "step": 8108 + }, + { + "epoch": 0.25013398933483455, + "grad_norm": 0.15087071061134338, + "learning_rate": 4.664480021801617e-05, + "loss": 3.1915, + "step": 8109 + }, + { + "epoch": 0.25016483580040794, + "grad_norm": 0.1519342064857483, + "learning_rate": 4.6643453038763954e-05, + "loss": 3.2147, + "step": 8110 + }, + { + "epoch": 0.2501956822659814, + "grad_norm": 0.1526372879743576, + "learning_rate": 4.6642105608568906e-05, + "loss": 3.1967, + "step": 8111 + }, + { + "epoch": 0.25022652873155476, + "grad_norm": 0.15046687424182892, + "learning_rate": 4.664075792744664e-05, + "loss": 3.2353, + "step": 8112 + }, + { + "epoch": 0.2502573751971282, + "grad_norm": 0.14885826408863068, + "learning_rate": 4.66394099954128e-05, + "loss": 3.2058, + "step": 8113 + }, + { + "epoch": 0.2502882216627016, + "grad_norm": 0.1534653604030609, + "learning_rate": 4.6638061812483005e-05, + "loss": 3.2147, + "step": 8114 + }, + { + "epoch": 0.25031906812827504, + "grad_norm": 0.1577036827802658, + "learning_rate": 4.663671337867288e-05, + "loss": 3.2301, + "step": 8115 + }, + { + "epoch": 0.2503499145938484, + "grad_norm": 0.14946840703487396, + "learning_rate": 4.6635364693998074e-05, + "loss": 3.222, + "step": 8116 + }, + { + "epoch": 0.25038076105942186, + "grad_norm": 0.15825453400611877, + "learning_rate": 4.6634015758474205e-05, + "loss": 3.2254, + "step": 8117 + }, + { + "epoch": 0.25041160752499525, + "grad_norm": 0.1497325897216797, + "learning_rate": 4.663266657211693e-05, + "loss": 3.2503, + "step": 8118 + }, + { + "epoch": 0.2504424539905687, + "grad_norm": 0.15683205425739288, + "learning_rate": 4.6631317134941884e-05, + "loss": 3.1973, + "step": 8119 + }, + { + "epoch": 0.25047330045614213, + "grad_norm": 0.16163600981235504, + "learning_rate": 4.662996744696472e-05, + "loss": 3.2154, + "step": 8120 + }, + { + "epoch": 0.2505041469217155, + "grad_norm": 0.16832469403743744, + "learning_rate": 4.6628617508201075e-05, + "loss": 3.2245, + "step": 8121 + }, + { + "epoch": 0.25053499338728896, + "grad_norm": 0.1578255593776703, + "learning_rate": 4.6627267318666605e-05, + "loss": 3.2058, + "step": 8122 + }, + { + "epoch": 0.25056583985286235, + "grad_norm": 0.1510142982006073, + "learning_rate": 4.6625916878376975e-05, + "loss": 3.2029, + "step": 8123 + }, + { + "epoch": 0.2505966863184358, + "grad_norm": 0.1613849401473999, + "learning_rate": 4.6624566187347826e-05, + "loss": 3.1741, + "step": 8124 + }, + { + "epoch": 0.2506275327840092, + "grad_norm": 0.14932307600975037, + "learning_rate": 4.662321524559483e-05, + "loss": 3.2002, + "step": 8125 + }, + { + "epoch": 0.2506583792495826, + "grad_norm": 0.15059474110603333, + "learning_rate": 4.6621864053133655e-05, + "loss": 3.2092, + "step": 8126 + }, + { + "epoch": 0.250689225715156, + "grad_norm": 0.15390260517597198, + "learning_rate": 4.6620512609979947e-05, + "loss": 3.1944, + "step": 8127 + }, + { + "epoch": 0.25072007218072945, + "grad_norm": 0.14836448431015015, + "learning_rate": 4.66191609161494e-05, + "loss": 3.2126, + "step": 8128 + }, + { + "epoch": 0.25075091864630283, + "grad_norm": 0.15907427668571472, + "learning_rate": 4.6617808971657665e-05, + "loss": 3.2239, + "step": 8129 + }, + { + "epoch": 0.2507817651118763, + "grad_norm": 0.153529092669487, + "learning_rate": 4.661645677652043e-05, + "loss": 3.2394, + "step": 8130 + }, + { + "epoch": 0.2508126115774497, + "grad_norm": 0.15440520644187927, + "learning_rate": 4.661510433075337e-05, + "loss": 3.2299, + "step": 8131 + }, + { + "epoch": 0.2508434580430231, + "grad_norm": 0.14726416766643524, + "learning_rate": 4.661375163437216e-05, + "loss": 3.2296, + "step": 8132 + }, + { + "epoch": 0.25087430450859655, + "grad_norm": 0.15818777680397034, + "learning_rate": 4.661239868739249e-05, + "loss": 3.2398, + "step": 8133 + }, + { + "epoch": 0.25090515097416993, + "grad_norm": 0.15405219793319702, + "learning_rate": 4.661104548983004e-05, + "loss": 3.1975, + "step": 8134 + }, + { + "epoch": 0.2509359974397434, + "grad_norm": 0.15222322940826416, + "learning_rate": 4.6609692041700516e-05, + "loss": 3.2023, + "step": 8135 + }, + { + "epoch": 0.25096684390531676, + "grad_norm": 0.1551714390516281, + "learning_rate": 4.660833834301959e-05, + "loss": 3.1955, + "step": 8136 + }, + { + "epoch": 0.2509976903708902, + "grad_norm": 0.15912796556949615, + "learning_rate": 4.6606984393802963e-05, + "loss": 3.2292, + "step": 8137 + }, + { + "epoch": 0.2510285368364636, + "grad_norm": 0.15203018486499786, + "learning_rate": 4.660563019406634e-05, + "loss": 3.2219, + "step": 8138 + }, + { + "epoch": 0.25105938330203703, + "grad_norm": 0.1510217785835266, + "learning_rate": 4.660427574382542e-05, + "loss": 3.1775, + "step": 8139 + }, + { + "epoch": 0.2510902297676104, + "grad_norm": 0.14883197844028473, + "learning_rate": 4.66029210430959e-05, + "loss": 3.2154, + "step": 8140 + }, + { + "epoch": 0.25112107623318386, + "grad_norm": 0.17022688686847687, + "learning_rate": 4.6601566091893505e-05, + "loss": 3.2185, + "step": 8141 + }, + { + "epoch": 0.25115192269875725, + "grad_norm": 0.15642523765563965, + "learning_rate": 4.660021089023393e-05, + "loss": 3.2288, + "step": 8142 + }, + { + "epoch": 0.2511827691643307, + "grad_norm": 0.15636557340621948, + "learning_rate": 4.6598855438132876e-05, + "loss": 3.2258, + "step": 8143 + }, + { + "epoch": 0.25121361562990413, + "grad_norm": 0.16258274018764496, + "learning_rate": 4.659749973560609e-05, + "loss": 3.197, + "step": 8144 + }, + { + "epoch": 0.2512444620954775, + "grad_norm": 0.16668730974197388, + "learning_rate": 4.6596143782669255e-05, + "loss": 3.2203, + "step": 8145 + }, + { + "epoch": 0.25127530856105096, + "grad_norm": 0.1531555950641632, + "learning_rate": 4.6594787579338126e-05, + "loss": 3.2096, + "step": 8146 + }, + { + "epoch": 0.25130615502662434, + "grad_norm": 0.16852165758609772, + "learning_rate": 4.659343112562841e-05, + "loss": 3.1939, + "step": 8147 + }, + { + "epoch": 0.2513370014921978, + "grad_norm": 0.14572419226169586, + "learning_rate": 4.659207442155582e-05, + "loss": 3.2411, + "step": 8148 + }, + { + "epoch": 0.2513678479577712, + "grad_norm": 0.15658943355083466, + "learning_rate": 4.6590717467136117e-05, + "loss": 3.2144, + "step": 8149 + }, + { + "epoch": 0.2513986944233446, + "grad_norm": 0.14975012838840485, + "learning_rate": 4.658936026238502e-05, + "loss": 3.2246, + "step": 8150 + }, + { + "epoch": 0.251429540888918, + "grad_norm": 0.16312292218208313, + "learning_rate": 4.6588002807318256e-05, + "loss": 3.2285, + "step": 8151 + }, + { + "epoch": 0.25146038735449144, + "grad_norm": 0.15146052837371826, + "learning_rate": 4.658664510195158e-05, + "loss": 3.2028, + "step": 8152 + }, + { + "epoch": 0.25149123382006483, + "grad_norm": 0.15204676985740662, + "learning_rate": 4.6585287146300724e-05, + "loss": 3.1895, + "step": 8153 + }, + { + "epoch": 0.25152208028563827, + "grad_norm": 0.15357711911201477, + "learning_rate": 4.6583928940381436e-05, + "loss": 3.2294, + "step": 8154 + }, + { + "epoch": 0.2515529267512117, + "grad_norm": 0.1519845426082611, + "learning_rate": 4.658257048420946e-05, + "loss": 3.2055, + "step": 8155 + }, + { + "epoch": 0.2515837732167851, + "grad_norm": 0.1568540781736374, + "learning_rate": 4.658121177780055e-05, + "loss": 3.1876, + "step": 8156 + }, + { + "epoch": 0.25161461968235854, + "grad_norm": 0.15902365744113922, + "learning_rate": 4.6579852821170456e-05, + "loss": 3.2017, + "step": 8157 + }, + { + "epoch": 0.25164546614793193, + "grad_norm": 0.1571044623851776, + "learning_rate": 4.657849361433494e-05, + "loss": 3.243, + "step": 8158 + }, + { + "epoch": 0.25167631261350537, + "grad_norm": 0.15624484419822693, + "learning_rate": 4.657713415730976e-05, + "loss": 3.2258, + "step": 8159 + }, + { + "epoch": 0.25170715907907876, + "grad_norm": 0.15840084850788116, + "learning_rate": 4.657577445011066e-05, + "loss": 3.2139, + "step": 8160 + }, + { + "epoch": 0.2517380055446522, + "grad_norm": 0.15004608035087585, + "learning_rate": 4.657441449275344e-05, + "loss": 3.2249, + "step": 8161 + }, + { + "epoch": 0.2517688520102256, + "grad_norm": 0.1542922407388687, + "learning_rate": 4.657305428525384e-05, + "loss": 3.1873, + "step": 8162 + }, + { + "epoch": 0.251799698475799, + "grad_norm": 0.16167831420898438, + "learning_rate": 4.657169382762764e-05, + "loss": 3.2066, + "step": 8163 + }, + { + "epoch": 0.2518305449413724, + "grad_norm": 0.151265949010849, + "learning_rate": 4.6570333119890616e-05, + "loss": 3.2006, + "step": 8164 + }, + { + "epoch": 0.25186139140694586, + "grad_norm": 0.1567380279302597, + "learning_rate": 4.6568972162058536e-05, + "loss": 3.2046, + "step": 8165 + }, + { + "epoch": 0.25189223787251924, + "grad_norm": 0.15921075642108917, + "learning_rate": 4.656761095414719e-05, + "loss": 3.2279, + "step": 8166 + }, + { + "epoch": 0.2519230843380927, + "grad_norm": 0.15243828296661377, + "learning_rate": 4.656624949617235e-05, + "loss": 3.198, + "step": 8167 + }, + { + "epoch": 0.2519539308036661, + "grad_norm": 0.16135108470916748, + "learning_rate": 4.6564887788149815e-05, + "loss": 3.2012, + "step": 8168 + }, + { + "epoch": 0.2519847772692395, + "grad_norm": 0.14944276213645935, + "learning_rate": 4.656352583009537e-05, + "loss": 3.2435, + "step": 8169 + }, + { + "epoch": 0.25201562373481295, + "grad_norm": 0.15257655084133148, + "learning_rate": 4.656216362202478e-05, + "loss": 3.2288, + "step": 8170 + }, + { + "epoch": 0.25204647020038634, + "grad_norm": 0.14982697367668152, + "learning_rate": 4.6560801163953885e-05, + "loss": 3.1798, + "step": 8171 + }, + { + "epoch": 0.2520773166659598, + "grad_norm": 0.15411806106567383, + "learning_rate": 4.655943845589844e-05, + "loss": 3.1988, + "step": 8172 + }, + { + "epoch": 0.25210816313153317, + "grad_norm": 0.15328197181224823, + "learning_rate": 4.6558075497874266e-05, + "loss": 3.227, + "step": 8173 + }, + { + "epoch": 0.2521390095971066, + "grad_norm": 0.1541500985622406, + "learning_rate": 4.655671228989716e-05, + "loss": 3.2121, + "step": 8174 + }, + { + "epoch": 0.25216985606268, + "grad_norm": 0.15589740872383118, + "learning_rate": 4.6555348831982936e-05, + "loss": 3.2054, + "step": 8175 + }, + { + "epoch": 0.25220070252825344, + "grad_norm": 0.1571151465177536, + "learning_rate": 4.6553985124147384e-05, + "loss": 3.2119, + "step": 8176 + }, + { + "epoch": 0.2522315489938268, + "grad_norm": 0.15492679178714752, + "learning_rate": 4.655262116640634e-05, + "loss": 3.2248, + "step": 8177 + }, + { + "epoch": 0.25226239545940027, + "grad_norm": 0.1674169898033142, + "learning_rate": 4.6551256958775596e-05, + "loss": 3.203, + "step": 8178 + }, + { + "epoch": 0.2522932419249737, + "grad_norm": 0.17281511425971985, + "learning_rate": 4.654989250127098e-05, + "loss": 3.2031, + "step": 8179 + }, + { + "epoch": 0.2523240883905471, + "grad_norm": 0.1574636697769165, + "learning_rate": 4.654852779390831e-05, + "loss": 3.2453, + "step": 8180 + }, + { + "epoch": 0.25235493485612054, + "grad_norm": 0.15716217458248138, + "learning_rate": 4.654716283670342e-05, + "loss": 3.2345, + "step": 8181 + }, + { + "epoch": 0.2523857813216939, + "grad_norm": 0.16166697442531586, + "learning_rate": 4.654579762967211e-05, + "loss": 3.2028, + "step": 8182 + }, + { + "epoch": 0.25241662778726737, + "grad_norm": 0.16348853707313538, + "learning_rate": 4.654443217283023e-05, + "loss": 3.2331, + "step": 8183 + }, + { + "epoch": 0.25244747425284075, + "grad_norm": 0.16162937879562378, + "learning_rate": 4.654306646619361e-05, + "loss": 3.2015, + "step": 8184 + }, + { + "epoch": 0.2524783207184142, + "grad_norm": 0.16736361384391785, + "learning_rate": 4.6541700509778074e-05, + "loss": 3.2275, + "step": 8185 + }, + { + "epoch": 0.2525091671839876, + "grad_norm": 0.1502310335636139, + "learning_rate": 4.654033430359946e-05, + "loss": 3.1985, + "step": 8186 + }, + { + "epoch": 0.252540013649561, + "grad_norm": 0.16558785736560822, + "learning_rate": 4.653896784767363e-05, + "loss": 3.2213, + "step": 8187 + }, + { + "epoch": 0.2525708601151344, + "grad_norm": 0.14571741223335266, + "learning_rate": 4.65376011420164e-05, + "loss": 3.2113, + "step": 8188 + }, + { + "epoch": 0.25260170658070785, + "grad_norm": 0.16456033289432526, + "learning_rate": 4.653623418664363e-05, + "loss": 3.2189, + "step": 8189 + }, + { + "epoch": 0.25263255304628124, + "grad_norm": 0.14978404343128204, + "learning_rate": 4.653486698157117e-05, + "loss": 3.2105, + "step": 8190 + }, + { + "epoch": 0.2526633995118547, + "grad_norm": 0.1534702628850937, + "learning_rate": 4.653349952681487e-05, + "loss": 3.2022, + "step": 8191 + }, + { + "epoch": 0.2526942459774281, + "grad_norm": 0.15905684232711792, + "learning_rate": 4.6532131822390576e-05, + "loss": 3.2104, + "step": 8192 + }, + { + "epoch": 0.2527250924430015, + "grad_norm": 0.1444462090730667, + "learning_rate": 4.653076386831415e-05, + "loss": 3.2071, + "step": 8193 + }, + { + "epoch": 0.25275593890857495, + "grad_norm": 0.1622656285762787, + "learning_rate": 4.652939566460146e-05, + "loss": 3.212, + "step": 8194 + }, + { + "epoch": 0.25278678537414834, + "grad_norm": 0.14968538284301758, + "learning_rate": 4.652802721126837e-05, + "loss": 3.1978, + "step": 8195 + }, + { + "epoch": 0.2528176318397218, + "grad_norm": 0.1559765785932541, + "learning_rate": 4.652665850833073e-05, + "loss": 3.2457, + "step": 8196 + }, + { + "epoch": 0.25284847830529517, + "grad_norm": 0.15054355561733246, + "learning_rate": 4.652528955580443e-05, + "loss": 3.1837, + "step": 8197 + }, + { + "epoch": 0.2528793247708686, + "grad_norm": 0.14406970143318176, + "learning_rate": 4.652392035370533e-05, + "loss": 3.2037, + "step": 8198 + }, + { + "epoch": 0.252910171236442, + "grad_norm": 0.1487010419368744, + "learning_rate": 4.652255090204931e-05, + "loss": 3.171, + "step": 8199 + }, + { + "epoch": 0.25294101770201544, + "grad_norm": 0.150107279419899, + "learning_rate": 4.6521181200852246e-05, + "loss": 3.184, + "step": 8200 + }, + { + "epoch": 0.2529718641675888, + "grad_norm": 0.14830058813095093, + "learning_rate": 4.651981125013001e-05, + "loss": 3.1765, + "step": 8201 + }, + { + "epoch": 0.25300271063316226, + "grad_norm": 0.15368035435676575, + "learning_rate": 4.651844104989851e-05, + "loss": 3.2124, + "step": 8202 + }, + { + "epoch": 0.2530335570987357, + "grad_norm": 0.1490894854068756, + "learning_rate": 4.6517070600173604e-05, + "loss": 3.2385, + "step": 8203 + }, + { + "epoch": 0.2530644035643091, + "grad_norm": 0.1564416140317917, + "learning_rate": 4.65156999009712e-05, + "loss": 3.1866, + "step": 8204 + }, + { + "epoch": 0.25309525002988253, + "grad_norm": 0.1672132909297943, + "learning_rate": 4.651432895230718e-05, + "loss": 3.2153, + "step": 8205 + }, + { + "epoch": 0.2531260964954559, + "grad_norm": 0.14919413626194, + "learning_rate": 4.651295775419745e-05, + "loss": 3.2112, + "step": 8206 + }, + { + "epoch": 0.25315694296102936, + "grad_norm": 0.16022935509681702, + "learning_rate": 4.65115863066579e-05, + "loss": 3.2358, + "step": 8207 + }, + { + "epoch": 0.25318778942660275, + "grad_norm": 0.146981343626976, + "learning_rate": 4.651021460970444e-05, + "loss": 3.1944, + "step": 8208 + }, + { + "epoch": 0.2532186358921762, + "grad_norm": 0.14962030947208405, + "learning_rate": 4.650884266335296e-05, + "loss": 3.1763, + "step": 8209 + }, + { + "epoch": 0.2532494823577496, + "grad_norm": 0.15152396261692047, + "learning_rate": 4.650747046761937e-05, + "loss": 3.2018, + "step": 8210 + }, + { + "epoch": 0.253280328823323, + "grad_norm": 0.15006865561008453, + "learning_rate": 4.65060980225196e-05, + "loss": 3.219, + "step": 8211 + }, + { + "epoch": 0.2533111752888964, + "grad_norm": 0.15608809888362885, + "learning_rate": 4.650472532806953e-05, + "loss": 3.2037, + "step": 8212 + }, + { + "epoch": 0.25334202175446985, + "grad_norm": 0.15099428594112396, + "learning_rate": 4.65033523842851e-05, + "loss": 3.2149, + "step": 8213 + }, + { + "epoch": 0.25337286822004323, + "grad_norm": 0.1490844041109085, + "learning_rate": 4.650197919118223e-05, + "loss": 3.1606, + "step": 8214 + }, + { + "epoch": 0.2534037146856167, + "grad_norm": 0.14874525368213654, + "learning_rate": 4.650060574877683e-05, + "loss": 3.1897, + "step": 8215 + }, + { + "epoch": 0.2534345611511901, + "grad_norm": 0.15042513608932495, + "learning_rate": 4.649923205708482e-05, + "loss": 3.217, + "step": 8216 + }, + { + "epoch": 0.2534654076167635, + "grad_norm": 0.1537875086069107, + "learning_rate": 4.649785811612213e-05, + "loss": 3.233, + "step": 8217 + }, + { + "epoch": 0.25349625408233695, + "grad_norm": 0.1464274674654007, + "learning_rate": 4.6496483925904696e-05, + "loss": 3.1896, + "step": 8218 + }, + { + "epoch": 0.25352710054791033, + "grad_norm": 0.15633028745651245, + "learning_rate": 4.649510948644846e-05, + "loss": 3.2378, + "step": 8219 + }, + { + "epoch": 0.2535579470134838, + "grad_norm": 0.14682786166667938, + "learning_rate": 4.649373479776933e-05, + "loss": 3.191, + "step": 8220 + }, + { + "epoch": 0.25358879347905716, + "grad_norm": 0.14798006415367126, + "learning_rate": 4.6492359859883274e-05, + "loss": 3.1794, + "step": 8221 + }, + { + "epoch": 0.2536196399446306, + "grad_norm": 0.15305669605731964, + "learning_rate": 4.6490984672806214e-05, + "loss": 3.2031, + "step": 8222 + }, + { + "epoch": 0.253650486410204, + "grad_norm": 0.14963610470294952, + "learning_rate": 4.6489609236554104e-05, + "loss": 3.2065, + "step": 8223 + }, + { + "epoch": 0.25368133287577743, + "grad_norm": 0.15105266869068146, + "learning_rate": 4.648823355114289e-05, + "loss": 3.2333, + "step": 8224 + }, + { + "epoch": 0.2537121793413508, + "grad_norm": 0.15723848342895508, + "learning_rate": 4.648685761658852e-05, + "loss": 3.2269, + "step": 8225 + }, + { + "epoch": 0.25374302580692426, + "grad_norm": 0.1486719399690628, + "learning_rate": 4.648548143290695e-05, + "loss": 3.2307, + "step": 8226 + }, + { + "epoch": 0.2537738722724977, + "grad_norm": 0.14626938104629517, + "learning_rate": 4.648410500011413e-05, + "loss": 3.1839, + "step": 8227 + }, + { + "epoch": 0.2538047187380711, + "grad_norm": 0.15523529052734375, + "learning_rate": 4.648272831822602e-05, + "loss": 3.1782, + "step": 8228 + }, + { + "epoch": 0.25383556520364453, + "grad_norm": 0.16043876111507416, + "learning_rate": 4.6481351387258596e-05, + "loss": 3.1934, + "step": 8229 + }, + { + "epoch": 0.2538664116692179, + "grad_norm": 0.17149588465690613, + "learning_rate": 4.64799742072278e-05, + "loss": 3.2304, + "step": 8230 + }, + { + "epoch": 0.25389725813479136, + "grad_norm": 0.15613476932048798, + "learning_rate": 4.647859677814962e-05, + "loss": 3.246, + "step": 8231 + }, + { + "epoch": 0.25392810460036475, + "grad_norm": 0.16666439175605774, + "learning_rate": 4.647721910004001e-05, + "loss": 3.1834, + "step": 8232 + }, + { + "epoch": 0.2539589510659382, + "grad_norm": 0.1483202427625656, + "learning_rate": 4.647584117291496e-05, + "loss": 3.2288, + "step": 8233 + }, + { + "epoch": 0.2539897975315116, + "grad_norm": 0.1741354614496231, + "learning_rate": 4.6474462996790425e-05, + "loss": 3.1801, + "step": 8234 + }, + { + "epoch": 0.254020643997085, + "grad_norm": 0.1781657338142395, + "learning_rate": 4.64730845716824e-05, + "loss": 3.1791, + "step": 8235 + }, + { + "epoch": 0.2540514904626584, + "grad_norm": 0.15558557212352753, + "learning_rate": 4.647170589760686e-05, + "loss": 3.1954, + "step": 8236 + }, + { + "epoch": 0.25408233692823184, + "grad_norm": 0.15370416641235352, + "learning_rate": 4.6470326974579804e-05, + "loss": 3.2374, + "step": 8237 + }, + { + "epoch": 0.25411318339380523, + "grad_norm": 0.15681611001491547, + "learning_rate": 4.646894780261721e-05, + "loss": 3.2047, + "step": 8238 + }, + { + "epoch": 0.2541440298593787, + "grad_norm": 0.15381892025470734, + "learning_rate": 4.6467568381735064e-05, + "loss": 3.2216, + "step": 8239 + }, + { + "epoch": 0.2541748763249521, + "grad_norm": 0.1517125368118286, + "learning_rate": 4.646618871194935e-05, + "loss": 3.1972, + "step": 8240 + }, + { + "epoch": 0.2542057227905255, + "grad_norm": 0.153535395860672, + "learning_rate": 4.6464808793276094e-05, + "loss": 3.1802, + "step": 8241 + }, + { + "epoch": 0.25423656925609894, + "grad_norm": 0.15435276925563812, + "learning_rate": 4.646342862573127e-05, + "loss": 3.197, + "step": 8242 + }, + { + "epoch": 0.25426741572167233, + "grad_norm": 0.4436056613922119, + "learning_rate": 4.6462048209330886e-05, + "loss": 3.1816, + "step": 8243 + }, + { + "epoch": 0.25429826218724577, + "grad_norm": 0.14810095727443695, + "learning_rate": 4.646066754409096e-05, + "loss": 3.2191, + "step": 8244 + }, + { + "epoch": 0.25432910865281916, + "grad_norm": 0.1598743200302124, + "learning_rate": 4.645928663002749e-05, + "loss": 3.2078, + "step": 8245 + }, + { + "epoch": 0.2543599551183926, + "grad_norm": 0.18781085312366486, + "learning_rate": 4.645790546715648e-05, + "loss": 3.2056, + "step": 8246 + }, + { + "epoch": 0.254390801583966, + "grad_norm": 0.14926880598068237, + "learning_rate": 4.645652405549396e-05, + "loss": 3.2125, + "step": 8247 + }, + { + "epoch": 0.25442164804953943, + "grad_norm": 0.16191859543323517, + "learning_rate": 4.6455142395055926e-05, + "loss": 3.1837, + "step": 8248 + }, + { + "epoch": 0.2544524945151128, + "grad_norm": 0.14401184022426605, + "learning_rate": 4.645376048585841e-05, + "loss": 3.1692, + "step": 8249 + }, + { + "epoch": 0.25448334098068626, + "grad_norm": 0.14797164499759674, + "learning_rate": 4.6452378327917444e-05, + "loss": 3.2186, + "step": 8250 + }, + { + "epoch": 0.2545141874462597, + "grad_norm": 0.15679173171520233, + "learning_rate": 4.645099592124904e-05, + "loss": 3.1841, + "step": 8251 + }, + { + "epoch": 0.2545450339118331, + "grad_norm": 0.1492667943239212, + "learning_rate": 4.644961326586922e-05, + "loss": 3.1909, + "step": 8252 + }, + { + "epoch": 0.2545758803774065, + "grad_norm": 0.151566281914711, + "learning_rate": 4.644823036179403e-05, + "loss": 3.1912, + "step": 8253 + }, + { + "epoch": 0.2546067268429799, + "grad_norm": 0.1536017507314682, + "learning_rate": 4.6446847209039504e-05, + "loss": 3.1833, + "step": 8254 + }, + { + "epoch": 0.25463757330855336, + "grad_norm": 0.14888574182987213, + "learning_rate": 4.6445463807621656e-05, + "loss": 3.1805, + "step": 8255 + }, + { + "epoch": 0.25466841977412674, + "grad_norm": 0.14888761937618256, + "learning_rate": 4.6444080157556554e-05, + "loss": 3.235, + "step": 8256 + }, + { + "epoch": 0.2546992662397002, + "grad_norm": 0.15110594034194946, + "learning_rate": 4.644269625886023e-05, + "loss": 3.2284, + "step": 8257 + }, + { + "epoch": 0.25473011270527357, + "grad_norm": 0.14706359803676605, + "learning_rate": 4.644131211154872e-05, + "loss": 3.1951, + "step": 8258 + }, + { + "epoch": 0.254760959170847, + "grad_norm": 0.1680697798728943, + "learning_rate": 4.643992771563809e-05, + "loss": 3.2213, + "step": 8259 + }, + { + "epoch": 0.2547918056364204, + "grad_norm": 0.14919202029705048, + "learning_rate": 4.643854307114438e-05, + "loss": 3.1985, + "step": 8260 + }, + { + "epoch": 0.25482265210199384, + "grad_norm": 0.1493794471025467, + "learning_rate": 4.6437158178083644e-05, + "loss": 3.1941, + "step": 8261 + }, + { + "epoch": 0.2548534985675673, + "grad_norm": 0.15195909142494202, + "learning_rate": 4.6435773036471944e-05, + "loss": 3.22, + "step": 8262 + }, + { + "epoch": 0.25488434503314067, + "grad_norm": 0.1567079871892929, + "learning_rate": 4.6434387646325334e-05, + "loss": 3.2178, + "step": 8263 + }, + { + "epoch": 0.2549151914987141, + "grad_norm": 0.15670333802700043, + "learning_rate": 4.6433002007659874e-05, + "loss": 3.2185, + "step": 8264 + }, + { + "epoch": 0.2549460379642875, + "grad_norm": 0.1607249528169632, + "learning_rate": 4.643161612049165e-05, + "loss": 3.1715, + "step": 8265 + }, + { + "epoch": 0.25497688442986094, + "grad_norm": 0.15071094036102295, + "learning_rate": 4.6430229984836704e-05, + "loss": 3.2364, + "step": 8266 + }, + { + "epoch": 0.2550077308954343, + "grad_norm": 0.15291909873485565, + "learning_rate": 4.642884360071113e-05, + "loss": 3.2339, + "step": 8267 + }, + { + "epoch": 0.25503857736100777, + "grad_norm": 0.15150229632854462, + "learning_rate": 4.642745696813098e-05, + "loss": 3.2097, + "step": 8268 + }, + { + "epoch": 0.25506942382658115, + "grad_norm": 0.16091258823871613, + "learning_rate": 4.642607008711235e-05, + "loss": 3.1955, + "step": 8269 + }, + { + "epoch": 0.2551002702921546, + "grad_norm": 0.17256228625774384, + "learning_rate": 4.642468295767131e-05, + "loss": 3.2097, + "step": 8270 + }, + { + "epoch": 0.255131116757728, + "grad_norm": 0.15698929131031036, + "learning_rate": 4.642329557982394e-05, + "loss": 3.2216, + "step": 8271 + }, + { + "epoch": 0.2551619632233014, + "grad_norm": 0.15410886704921722, + "learning_rate": 4.642190795358634e-05, + "loss": 3.2015, + "step": 8272 + }, + { + "epoch": 0.2551928096888748, + "grad_norm": 0.15486842393875122, + "learning_rate": 4.6420520078974596e-05, + "loss": 3.2049, + "step": 8273 + }, + { + "epoch": 0.25522365615444825, + "grad_norm": 0.1593635529279709, + "learning_rate": 4.641913195600478e-05, + "loss": 3.2089, + "step": 8274 + }, + { + "epoch": 0.2552545026200217, + "grad_norm": 0.15437722206115723, + "learning_rate": 4.6417743584693005e-05, + "loss": 3.2098, + "step": 8275 + }, + { + "epoch": 0.2552853490855951, + "grad_norm": 0.1653943508863449, + "learning_rate": 4.641635496505537e-05, + "loss": 3.1988, + "step": 8276 + }, + { + "epoch": 0.2553161955511685, + "grad_norm": 0.1589290052652359, + "learning_rate": 4.6414966097107966e-05, + "loss": 3.2096, + "step": 8277 + }, + { + "epoch": 0.2553470420167419, + "grad_norm": 0.15730398893356323, + "learning_rate": 4.64135769808669e-05, + "loss": 3.2282, + "step": 8278 + }, + { + "epoch": 0.25537788848231535, + "grad_norm": 0.15683776140213013, + "learning_rate": 4.641218761634827e-05, + "loss": 3.2347, + "step": 8279 + }, + { + "epoch": 0.25540873494788874, + "grad_norm": 0.1532643735408783, + "learning_rate": 4.64107980035682e-05, + "loss": 3.2041, + "step": 8280 + }, + { + "epoch": 0.2554395814134622, + "grad_norm": 0.18439115583896637, + "learning_rate": 4.640940814254279e-05, + "loss": 3.2007, + "step": 8281 + }, + { + "epoch": 0.25547042787903557, + "grad_norm": 0.164102703332901, + "learning_rate": 4.640801803328816e-05, + "loss": 3.215, + "step": 8282 + }, + { + "epoch": 0.255501274344609, + "grad_norm": 0.15337638556957245, + "learning_rate": 4.640662767582042e-05, + "loss": 3.2094, + "step": 8283 + }, + { + "epoch": 0.2555321208101824, + "grad_norm": 0.1831592172384262, + "learning_rate": 4.64052370701557e-05, + "loss": 3.1808, + "step": 8284 + }, + { + "epoch": 0.25556296727575584, + "grad_norm": 0.1515939086675644, + "learning_rate": 4.640384621631012e-05, + "loss": 3.1902, + "step": 8285 + }, + { + "epoch": 0.2555938137413293, + "grad_norm": 0.17319080233573914, + "learning_rate": 4.6402455114299804e-05, + "loss": 3.2038, + "step": 8286 + }, + { + "epoch": 0.25562466020690267, + "grad_norm": 0.15840831398963928, + "learning_rate": 4.6401063764140884e-05, + "loss": 3.2288, + "step": 8287 + }, + { + "epoch": 0.2556555066724761, + "grad_norm": 0.14678914844989777, + "learning_rate": 4.6399672165849484e-05, + "loss": 3.1979, + "step": 8288 + }, + { + "epoch": 0.2556863531380495, + "grad_norm": 0.15671689808368683, + "learning_rate": 4.639828031944176e-05, + "loss": 3.1915, + "step": 8289 + }, + { + "epoch": 0.25571719960362294, + "grad_norm": 0.15376779437065125, + "learning_rate": 4.639688822493382e-05, + "loss": 3.1722, + "step": 8290 + }, + { + "epoch": 0.2557480460691963, + "grad_norm": 0.24981176853179932, + "learning_rate": 4.639549588234182e-05, + "loss": 3.2235, + "step": 8291 + }, + { + "epoch": 0.25577889253476976, + "grad_norm": 0.14881663024425507, + "learning_rate": 4.6394103291681906e-05, + "loss": 3.237, + "step": 8292 + }, + { + "epoch": 0.25580973900034315, + "grad_norm": 0.15740425884723663, + "learning_rate": 4.639271045297022e-05, + "loss": 3.2042, + "step": 8293 + }, + { + "epoch": 0.2558405854659166, + "grad_norm": 0.15653036534786224, + "learning_rate": 4.639131736622291e-05, + "loss": 3.1698, + "step": 8294 + }, + { + "epoch": 0.25587143193149, + "grad_norm": 0.16097286343574524, + "learning_rate": 4.6389924031456134e-05, + "loss": 3.2412, + "step": 8295 + }, + { + "epoch": 0.2559022783970634, + "grad_norm": 0.14912071824073792, + "learning_rate": 4.638853044868604e-05, + "loss": 3.2143, + "step": 8296 + }, + { + "epoch": 0.2559331248626368, + "grad_norm": 0.1579957753419876, + "learning_rate": 4.63871366179288e-05, + "loss": 3.188, + "step": 8297 + }, + { + "epoch": 0.25596397132821025, + "grad_norm": 0.150410458445549, + "learning_rate": 4.638574253920054e-05, + "loss": 3.239, + "step": 8298 + }, + { + "epoch": 0.2559948177937837, + "grad_norm": 0.14966848492622375, + "learning_rate": 4.6384348212517457e-05, + "loss": 3.1939, + "step": 8299 + }, + { + "epoch": 0.2560256642593571, + "grad_norm": 0.1549370139837265, + "learning_rate": 4.638295363789571e-05, + "loss": 3.2121, + "step": 8300 + }, + { + "epoch": 0.2560565107249305, + "grad_norm": 0.14978782832622528, + "learning_rate": 4.6381558815351464e-05, + "loss": 3.2157, + "step": 8301 + }, + { + "epoch": 0.2560873571905039, + "grad_norm": 0.15570038557052612, + "learning_rate": 4.63801637449009e-05, + "loss": 3.1973, + "step": 8302 + }, + { + "epoch": 0.25611820365607735, + "grad_norm": 0.14426499605178833, + "learning_rate": 4.637876842656018e-05, + "loss": 3.2151, + "step": 8303 + }, + { + "epoch": 0.25614905012165073, + "grad_norm": 0.15340836346149445, + "learning_rate": 4.6377372860345484e-05, + "loss": 3.1894, + "step": 8304 + }, + { + "epoch": 0.2561798965872242, + "grad_norm": 0.1532222032546997, + "learning_rate": 4.6375977046273e-05, + "loss": 3.1772, + "step": 8305 + }, + { + "epoch": 0.25621074305279756, + "grad_norm": 0.15277498960494995, + "learning_rate": 4.6374580984358896e-05, + "loss": 3.1923, + "step": 8306 + }, + { + "epoch": 0.256241589518371, + "grad_norm": 0.15260587632656097, + "learning_rate": 4.6373184674619385e-05, + "loss": 3.1739, + "step": 8307 + }, + { + "epoch": 0.2562724359839444, + "grad_norm": 0.1542501300573349, + "learning_rate": 4.6371788117070635e-05, + "loss": 3.1482, + "step": 8308 + }, + { + "epoch": 0.25630328244951783, + "grad_norm": 0.153610497713089, + "learning_rate": 4.637039131172885e-05, + "loss": 3.191, + "step": 8309 + }, + { + "epoch": 0.2563341289150913, + "grad_norm": 0.1465749442577362, + "learning_rate": 4.636899425861021e-05, + "loss": 3.2257, + "step": 8310 + }, + { + "epoch": 0.25636497538066466, + "grad_norm": 0.14924700558185577, + "learning_rate": 4.636759695773093e-05, + "loss": 3.1826, + "step": 8311 + }, + { + "epoch": 0.2563958218462381, + "grad_norm": 0.15352152287960052, + "learning_rate": 4.63661994091072e-05, + "loss": 3.2011, + "step": 8312 + }, + { + "epoch": 0.2564266683118115, + "grad_norm": 0.14613580703735352, + "learning_rate": 4.6364801612755225e-05, + "loss": 3.2023, + "step": 8313 + }, + { + "epoch": 0.25645751477738493, + "grad_norm": 0.15350206196308136, + "learning_rate": 4.6363403568691225e-05, + "loss": 3.2098, + "step": 8314 + }, + { + "epoch": 0.2564883612429583, + "grad_norm": 0.15510910749435425, + "learning_rate": 4.6362005276931386e-05, + "loss": 3.2264, + "step": 8315 + }, + { + "epoch": 0.25651920770853176, + "grad_norm": 0.14770793914794922, + "learning_rate": 4.6360606737491944e-05, + "loss": 3.1909, + "step": 8316 + }, + { + "epoch": 0.25655005417410515, + "grad_norm": 0.1479172557592392, + "learning_rate": 4.6359207950389094e-05, + "loss": 3.2248, + "step": 8317 + }, + { + "epoch": 0.2565809006396786, + "grad_norm": 0.1467176377773285, + "learning_rate": 4.6357808915639066e-05, + "loss": 3.1963, + "step": 8318 + }, + { + "epoch": 0.256611747105252, + "grad_norm": 0.14952300488948822, + "learning_rate": 4.635640963325808e-05, + "loss": 3.2047, + "step": 8319 + }, + { + "epoch": 0.2566425935708254, + "grad_norm": 0.15596462786197662, + "learning_rate": 4.6355010103262356e-05, + "loss": 3.1775, + "step": 8320 + }, + { + "epoch": 0.2566734400363988, + "grad_norm": 0.15388123691082, + "learning_rate": 4.6353610325668126e-05, + "loss": 3.2104, + "step": 8321 + }, + { + "epoch": 0.25670428650197225, + "grad_norm": 0.14876748621463776, + "learning_rate": 4.635221030049161e-05, + "loss": 3.1912, + "step": 8322 + }, + { + "epoch": 0.2567351329675457, + "grad_norm": 0.15071450173854828, + "learning_rate": 4.635081002774905e-05, + "loss": 3.2556, + "step": 8323 + }, + { + "epoch": 0.2567659794331191, + "grad_norm": 0.14963644742965698, + "learning_rate": 4.634940950745668e-05, + "loss": 3.2385, + "step": 8324 + }, + { + "epoch": 0.2567968258986925, + "grad_norm": 0.15126147866249084, + "learning_rate": 4.6348008739630734e-05, + "loss": 3.2254, + "step": 8325 + }, + { + "epoch": 0.2568276723642659, + "grad_norm": 0.14387303590774536, + "learning_rate": 4.634660772428746e-05, + "loss": 3.2169, + "step": 8326 + }, + { + "epoch": 0.25685851882983934, + "grad_norm": 0.1529635488986969, + "learning_rate": 4.634520646144309e-05, + "loss": 3.2102, + "step": 8327 + }, + { + "epoch": 0.25688936529541273, + "grad_norm": 0.14490939676761627, + "learning_rate": 4.634380495111388e-05, + "loss": 3.1773, + "step": 8328 + }, + { + "epoch": 0.2569202117609862, + "grad_norm": 0.21116140484809875, + "learning_rate": 4.634240319331609e-05, + "loss": 3.1849, + "step": 8329 + }, + { + "epoch": 0.25695105822655956, + "grad_norm": 0.15092626214027405, + "learning_rate": 4.6341001188065944e-05, + "loss": 3.2212, + "step": 8330 + }, + { + "epoch": 0.256981904692133, + "grad_norm": 0.15221618115901947, + "learning_rate": 4.633959893537972e-05, + "loss": 3.2062, + "step": 8331 + }, + { + "epoch": 0.2570127511577064, + "grad_norm": 0.16571082174777985, + "learning_rate": 4.633819643527367e-05, + "loss": 3.2194, + "step": 8332 + }, + { + "epoch": 0.25704359762327983, + "grad_norm": 0.14935603737831116, + "learning_rate": 4.633679368776405e-05, + "loss": 3.2069, + "step": 8333 + }, + { + "epoch": 0.25707444408885327, + "grad_norm": 0.15150418877601624, + "learning_rate": 4.633539069286713e-05, + "loss": 3.2344, + "step": 8334 + }, + { + "epoch": 0.25710529055442666, + "grad_norm": 0.14790423214435577, + "learning_rate": 4.6333987450599185e-05, + "loss": 3.1999, + "step": 8335 + }, + { + "epoch": 0.2571361370200001, + "grad_norm": 0.14956696331501007, + "learning_rate": 4.6332583960976474e-05, + "loss": 3.2078, + "step": 8336 + }, + { + "epoch": 0.2571669834855735, + "grad_norm": 0.15116074681282043, + "learning_rate": 4.6331180224015264e-05, + "loss": 3.1903, + "step": 8337 + }, + { + "epoch": 0.25719782995114693, + "grad_norm": 0.1570831537246704, + "learning_rate": 4.6329776239731845e-05, + "loss": 3.1902, + "step": 8338 + }, + { + "epoch": 0.2572286764167203, + "grad_norm": 0.1474999636411667, + "learning_rate": 4.6328372008142486e-05, + "loss": 3.2362, + "step": 8339 + }, + { + "epoch": 0.25725952288229376, + "grad_norm": 0.15514367818832397, + "learning_rate": 4.6326967529263464e-05, + "loss": 3.2316, + "step": 8340 + }, + { + "epoch": 0.25729036934786714, + "grad_norm": 0.15118557214736938, + "learning_rate": 4.6325562803111075e-05, + "loss": 3.219, + "step": 8341 + }, + { + "epoch": 0.2573212158134406, + "grad_norm": 0.15499798953533173, + "learning_rate": 4.63241578297016e-05, + "loss": 3.2219, + "step": 8342 + }, + { + "epoch": 0.25735206227901397, + "grad_norm": 0.1527244746685028, + "learning_rate": 4.632275260905133e-05, + "loss": 3.1812, + "step": 8343 + }, + { + "epoch": 0.2573829087445874, + "grad_norm": 0.20884154736995697, + "learning_rate": 4.6321347141176555e-05, + "loss": 3.2315, + "step": 8344 + }, + { + "epoch": 0.2574137552101608, + "grad_norm": 0.14822979271411896, + "learning_rate": 4.631994142609357e-05, + "loss": 3.2095, + "step": 8345 + }, + { + "epoch": 0.25744460167573424, + "grad_norm": 0.15417319536209106, + "learning_rate": 4.6318535463818684e-05, + "loss": 3.22, + "step": 8346 + }, + { + "epoch": 0.2574754481413077, + "grad_norm": 0.1530783325433731, + "learning_rate": 4.6317129254368185e-05, + "loss": 3.2389, + "step": 8347 + }, + { + "epoch": 0.25750629460688107, + "grad_norm": 0.15902411937713623, + "learning_rate": 4.6315722797758386e-05, + "loss": 3.2294, + "step": 8348 + }, + { + "epoch": 0.2575371410724545, + "grad_norm": 0.1500316858291626, + "learning_rate": 4.631431609400559e-05, + "loss": 3.1753, + "step": 8349 + }, + { + "epoch": 0.2575679875380279, + "grad_norm": 0.1479903757572174, + "learning_rate": 4.6312909143126104e-05, + "loss": 3.2131, + "step": 8350 + }, + { + "epoch": 0.25759883400360134, + "grad_norm": 0.1560288667678833, + "learning_rate": 4.631150194513625e-05, + "loss": 3.2308, + "step": 8351 + }, + { + "epoch": 0.2576296804691747, + "grad_norm": 0.1559136062860489, + "learning_rate": 4.6310094500052324e-05, + "loss": 3.1716, + "step": 8352 + }, + { + "epoch": 0.25766052693474817, + "grad_norm": 0.15402711927890778, + "learning_rate": 4.6308686807890675e-05, + "loss": 3.2158, + "step": 8353 + }, + { + "epoch": 0.25769137340032156, + "grad_norm": 0.15207859873771667, + "learning_rate": 4.6307278868667595e-05, + "loss": 3.2263, + "step": 8354 + }, + { + "epoch": 0.257722219865895, + "grad_norm": 0.16308742761611938, + "learning_rate": 4.630587068239943e-05, + "loss": 3.232, + "step": 8355 + }, + { + "epoch": 0.2577530663314684, + "grad_norm": 0.15343652665615082, + "learning_rate": 4.630446224910249e-05, + "loss": 3.2141, + "step": 8356 + }, + { + "epoch": 0.2577839127970418, + "grad_norm": 0.151804119348526, + "learning_rate": 4.630305356879311e-05, + "loss": 3.1829, + "step": 8357 + }, + { + "epoch": 0.25781475926261527, + "grad_norm": 0.15385444462299347, + "learning_rate": 4.630164464148764e-05, + "loss": 3.2198, + "step": 8358 + }, + { + "epoch": 0.25784560572818865, + "grad_norm": 0.14954565465450287, + "learning_rate": 4.630023546720239e-05, + "loss": 3.2076, + "step": 8359 + }, + { + "epoch": 0.2578764521937621, + "grad_norm": 0.15815649926662445, + "learning_rate": 4.6298826045953715e-05, + "loss": 3.2209, + "step": 8360 + }, + { + "epoch": 0.2579072986593355, + "grad_norm": 0.16498097777366638, + "learning_rate": 4.629741637775794e-05, + "loss": 3.216, + "step": 8361 + }, + { + "epoch": 0.2579381451249089, + "grad_norm": 0.14945657551288605, + "learning_rate": 4.6296006462631434e-05, + "loss": 3.2231, + "step": 8362 + }, + { + "epoch": 0.2579689915904823, + "grad_norm": 0.14613813161849976, + "learning_rate": 4.6294596300590524e-05, + "loss": 3.237, + "step": 8363 + }, + { + "epoch": 0.25799983805605575, + "grad_norm": 0.15609128773212433, + "learning_rate": 4.629318589165157e-05, + "loss": 3.2096, + "step": 8364 + }, + { + "epoch": 0.25803068452162914, + "grad_norm": 0.1539527028799057, + "learning_rate": 4.6291775235830914e-05, + "loss": 3.2339, + "step": 8365 + }, + { + "epoch": 0.2580615309872026, + "grad_norm": 0.15891924500465393, + "learning_rate": 4.629036433314493e-05, + "loss": 3.25, + "step": 8366 + }, + { + "epoch": 0.25809237745277597, + "grad_norm": 0.14738526940345764, + "learning_rate": 4.628895318360996e-05, + "loss": 3.1793, + "step": 8367 + }, + { + "epoch": 0.2581232239183494, + "grad_norm": 0.16207638382911682, + "learning_rate": 4.628754178724237e-05, + "loss": 3.1953, + "step": 8368 + }, + { + "epoch": 0.2581540703839228, + "grad_norm": 0.15054091811180115, + "learning_rate": 4.6286130144058525e-05, + "loss": 3.1962, + "step": 8369 + }, + { + "epoch": 0.25818491684949624, + "grad_norm": 0.14920377731323242, + "learning_rate": 4.628471825407479e-05, + "loss": 3.2142, + "step": 8370 + }, + { + "epoch": 0.2582157633150697, + "grad_norm": 0.15339572727680206, + "learning_rate": 4.6283306117307545e-05, + "loss": 3.1818, + "step": 8371 + }, + { + "epoch": 0.25824660978064307, + "grad_norm": 0.15233027935028076, + "learning_rate": 4.628189373377315e-05, + "loss": 3.1869, + "step": 8372 + }, + { + "epoch": 0.2582774562462165, + "grad_norm": 0.15020841360092163, + "learning_rate": 4.6280481103487985e-05, + "loss": 3.208, + "step": 8373 + }, + { + "epoch": 0.2583083027117899, + "grad_norm": 0.15306800603866577, + "learning_rate": 4.6279068226468436e-05, + "loss": 3.2084, + "step": 8374 + }, + { + "epoch": 0.25833914917736334, + "grad_norm": 0.15106111764907837, + "learning_rate": 4.627765510273087e-05, + "loss": 3.1765, + "step": 8375 + }, + { + "epoch": 0.2583699956429367, + "grad_norm": 0.15130098164081573, + "learning_rate": 4.6276241732291684e-05, + "loss": 3.2319, + "step": 8376 + }, + { + "epoch": 0.25840084210851016, + "grad_norm": 0.15738581120967865, + "learning_rate": 4.6274828115167264e-05, + "loss": 3.2216, + "step": 8377 + }, + { + "epoch": 0.25843168857408355, + "grad_norm": 0.1604832112789154, + "learning_rate": 4.627341425137399e-05, + "loss": 3.2068, + "step": 8378 + }, + { + "epoch": 0.258462535039657, + "grad_norm": 0.15039539337158203, + "learning_rate": 4.6272000140928264e-05, + "loss": 3.1719, + "step": 8379 + }, + { + "epoch": 0.2584933815052304, + "grad_norm": 0.16046558320522308, + "learning_rate": 4.627058578384648e-05, + "loss": 3.2148, + "step": 8380 + }, + { + "epoch": 0.2585242279708038, + "grad_norm": 0.16329418122768402, + "learning_rate": 4.626917118014504e-05, + "loss": 3.1785, + "step": 8381 + }, + { + "epoch": 0.25855507443637726, + "grad_norm": 0.14757736027240753, + "learning_rate": 4.626775632984033e-05, + "loss": 3.1946, + "step": 8382 + }, + { + "epoch": 0.25858592090195065, + "grad_norm": 0.15842357277870178, + "learning_rate": 4.626634123294878e-05, + "loss": 3.1828, + "step": 8383 + }, + { + "epoch": 0.2586167673675241, + "grad_norm": 0.15093573927879333, + "learning_rate": 4.6264925889486774e-05, + "loss": 3.1968, + "step": 8384 + }, + { + "epoch": 0.2586476138330975, + "grad_norm": 0.16064977645874023, + "learning_rate": 4.626351029947074e-05, + "loss": 3.188, + "step": 8385 + }, + { + "epoch": 0.2586784602986709, + "grad_norm": 0.15370061993598938, + "learning_rate": 4.6262094462917074e-05, + "loss": 3.2203, + "step": 8386 + }, + { + "epoch": 0.2587093067642443, + "grad_norm": 0.1626019924879074, + "learning_rate": 4.626067837984219e-05, + "loss": 3.1884, + "step": 8387 + }, + { + "epoch": 0.25874015322981775, + "grad_norm": 0.15251962840557098, + "learning_rate": 4.625926205026253e-05, + "loss": 3.205, + "step": 8388 + }, + { + "epoch": 0.25877099969539114, + "grad_norm": 0.17414021492004395, + "learning_rate": 4.62578454741945e-05, + "loss": 3.1994, + "step": 8389 + }, + { + "epoch": 0.2588018461609646, + "grad_norm": 0.15203151106834412, + "learning_rate": 4.625642865165453e-05, + "loss": 3.1909, + "step": 8390 + }, + { + "epoch": 0.25883269262653796, + "grad_norm": 0.1601545214653015, + "learning_rate": 4.6255011582659037e-05, + "loss": 3.1837, + "step": 8391 + }, + { + "epoch": 0.2588635390921114, + "grad_norm": 0.15406891703605652, + "learning_rate": 4.625359426722445e-05, + "loss": 3.2391, + "step": 8392 + }, + { + "epoch": 0.2588943855576848, + "grad_norm": 0.1610194891691208, + "learning_rate": 4.625217670536722e-05, + "loss": 3.1884, + "step": 8393 + }, + { + "epoch": 0.25892523202325823, + "grad_norm": 0.15359239280223846, + "learning_rate": 4.6250758897103775e-05, + "loss": 3.1879, + "step": 8394 + }, + { + "epoch": 0.2589560784888317, + "grad_norm": 0.1549641489982605, + "learning_rate": 4.624934084245055e-05, + "loss": 3.2071, + "step": 8395 + }, + { + "epoch": 0.25898692495440506, + "grad_norm": 0.14728423953056335, + "learning_rate": 4.624792254142398e-05, + "loss": 3.1566, + "step": 8396 + }, + { + "epoch": 0.2590177714199785, + "grad_norm": 0.14969997107982635, + "learning_rate": 4.624650399404052e-05, + "loss": 3.2118, + "step": 8397 + }, + { + "epoch": 0.2590486178855519, + "grad_norm": 0.14970441162586212, + "learning_rate": 4.624508520031662e-05, + "loss": 3.2135, + "step": 8398 + }, + { + "epoch": 0.25907946435112533, + "grad_norm": 0.15578195452690125, + "learning_rate": 4.6243666160268716e-05, + "loss": 3.1993, + "step": 8399 + }, + { + "epoch": 0.2591103108166987, + "grad_norm": 0.15501897037029266, + "learning_rate": 4.624224687391327e-05, + "loss": 3.2121, + "step": 8400 + }, + { + "epoch": 0.25914115728227216, + "grad_norm": 0.15246137976646423, + "learning_rate": 4.6240827341266735e-05, + "loss": 3.1898, + "step": 8401 + }, + { + "epoch": 0.25917200374784555, + "grad_norm": 0.16083668172359467, + "learning_rate": 4.6239407562345576e-05, + "loss": 3.2225, + "step": 8402 + }, + { + "epoch": 0.259202850213419, + "grad_norm": 0.15598176419734955, + "learning_rate": 4.623798753716625e-05, + "loss": 3.2192, + "step": 8403 + }, + { + "epoch": 0.2592336966789924, + "grad_norm": 0.15915323793888092, + "learning_rate": 4.623656726574522e-05, + "loss": 3.2594, + "step": 8404 + }, + { + "epoch": 0.2592645431445658, + "grad_norm": 0.15303783118724823, + "learning_rate": 4.623514674809894e-05, + "loss": 3.2266, + "step": 8405 + }, + { + "epoch": 0.25929538961013926, + "grad_norm": 0.1533498913049698, + "learning_rate": 4.6233725984243915e-05, + "loss": 3.2108, + "step": 8406 + }, + { + "epoch": 0.25932623607571265, + "grad_norm": 0.15047229826450348, + "learning_rate": 4.623230497419659e-05, + "loss": 3.2092, + "step": 8407 + }, + { + "epoch": 0.2593570825412861, + "grad_norm": 0.17429116368293762, + "learning_rate": 4.623088371797345e-05, + "loss": 3.18, + "step": 8408 + }, + { + "epoch": 0.2593879290068595, + "grad_norm": 0.15053221583366394, + "learning_rate": 4.622946221559097e-05, + "loss": 3.2024, + "step": 8409 + }, + { + "epoch": 0.2594187754724329, + "grad_norm": 0.16096118092536926, + "learning_rate": 4.622804046706563e-05, + "loss": 3.1716, + "step": 8410 + }, + { + "epoch": 0.2594496219380063, + "grad_norm": 0.15963967144489288, + "learning_rate": 4.6226618472413915e-05, + "loss": 3.2041, + "step": 8411 + }, + { + "epoch": 0.25948046840357974, + "grad_norm": 0.15982158482074738, + "learning_rate": 4.6225196231652325e-05, + "loss": 3.2368, + "step": 8412 + }, + { + "epoch": 0.25951131486915313, + "grad_norm": 0.1518370509147644, + "learning_rate": 4.6223773744797326e-05, + "loss": 3.221, + "step": 8413 + }, + { + "epoch": 0.2595421613347266, + "grad_norm": 0.1577429324388504, + "learning_rate": 4.622235101186544e-05, + "loss": 3.1994, + "step": 8414 + }, + { + "epoch": 0.25957300780029996, + "grad_norm": 0.15836423635482788, + "learning_rate": 4.622092803287313e-05, + "loss": 3.2264, + "step": 8415 + }, + { + "epoch": 0.2596038542658734, + "grad_norm": 0.1538601666688919, + "learning_rate": 4.621950480783693e-05, + "loss": 3.1984, + "step": 8416 + }, + { + "epoch": 0.2596347007314468, + "grad_norm": 0.16694411635398865, + "learning_rate": 4.621808133677331e-05, + "loss": 3.1779, + "step": 8417 + }, + { + "epoch": 0.25966554719702023, + "grad_norm": 0.15823893249034882, + "learning_rate": 4.621665761969879e-05, + "loss": 3.2029, + "step": 8418 + }, + { + "epoch": 0.25969639366259367, + "grad_norm": 0.16437411308288574, + "learning_rate": 4.6215233656629876e-05, + "loss": 3.2093, + "step": 8419 + }, + { + "epoch": 0.25972724012816706, + "grad_norm": 0.1519143283367157, + "learning_rate": 4.621380944758308e-05, + "loss": 3.164, + "step": 8420 + }, + { + "epoch": 0.2597580865937405, + "grad_norm": 0.1629306823015213, + "learning_rate": 4.6212384992574906e-05, + "loss": 3.217, + "step": 8421 + }, + { + "epoch": 0.2597889330593139, + "grad_norm": 0.1490415334701538, + "learning_rate": 4.621096029162188e-05, + "loss": 3.1952, + "step": 8422 + }, + { + "epoch": 0.25981977952488733, + "grad_norm": 0.15319734811782837, + "learning_rate": 4.620953534474051e-05, + "loss": 3.2134, + "step": 8423 + }, + { + "epoch": 0.2598506259904607, + "grad_norm": 0.16263505816459656, + "learning_rate": 4.6208110151947326e-05, + "loss": 3.2516, + "step": 8424 + }, + { + "epoch": 0.25988147245603416, + "grad_norm": 0.1513398438692093, + "learning_rate": 4.6206684713258854e-05, + "loss": 3.2084, + "step": 8425 + }, + { + "epoch": 0.25991231892160754, + "grad_norm": 0.16006116569042206, + "learning_rate": 4.6205259028691605e-05, + "loss": 3.2044, + "step": 8426 + }, + { + "epoch": 0.259943165387181, + "grad_norm": 0.15463405847549438, + "learning_rate": 4.620383309826213e-05, + "loss": 3.2039, + "step": 8427 + }, + { + "epoch": 0.25997401185275437, + "grad_norm": 0.1640108823776245, + "learning_rate": 4.6202406921986953e-05, + "loss": 3.2142, + "step": 8428 + }, + { + "epoch": 0.2600048583183278, + "grad_norm": 0.1531175673007965, + "learning_rate": 4.6200980499882605e-05, + "loss": 3.2035, + "step": 8429 + }, + { + "epoch": 0.26003570478390126, + "grad_norm": 0.15906736254692078, + "learning_rate": 4.619955383196562e-05, + "loss": 3.2283, + "step": 8430 + }, + { + "epoch": 0.26006655124947464, + "grad_norm": 0.1514875292778015, + "learning_rate": 4.619812691825257e-05, + "loss": 3.1868, + "step": 8431 + }, + { + "epoch": 0.2600973977150481, + "grad_norm": 0.15540029108524323, + "learning_rate": 4.619669975875996e-05, + "loss": 3.1863, + "step": 8432 + }, + { + "epoch": 0.26012824418062147, + "grad_norm": 0.15802229940891266, + "learning_rate": 4.619527235350436e-05, + "loss": 3.1999, + "step": 8433 + }, + { + "epoch": 0.2601590906461949, + "grad_norm": 0.15068615972995758, + "learning_rate": 4.619384470250231e-05, + "loss": 3.207, + "step": 8434 + }, + { + "epoch": 0.2601899371117683, + "grad_norm": 0.15216626226902008, + "learning_rate": 4.6192416805770366e-05, + "loss": 3.2055, + "step": 8435 + }, + { + "epoch": 0.26022078357734174, + "grad_norm": 0.1526876538991928, + "learning_rate": 4.619098866332509e-05, + "loss": 3.1999, + "step": 8436 + }, + { + "epoch": 0.2602516300429151, + "grad_norm": 0.15553687512874603, + "learning_rate": 4.6189560275183034e-05, + "loss": 3.2715, + "step": 8437 + }, + { + "epoch": 0.26028247650848857, + "grad_norm": 0.16788694262504578, + "learning_rate": 4.6188131641360755e-05, + "loss": 3.1938, + "step": 8438 + }, + { + "epoch": 0.26031332297406196, + "grad_norm": 0.14809398353099823, + "learning_rate": 4.6186702761874825e-05, + "loss": 3.1845, + "step": 8439 + }, + { + "epoch": 0.2603441694396354, + "grad_norm": 0.15904858708381653, + "learning_rate": 4.6185273636741805e-05, + "loss": 3.2284, + "step": 8440 + }, + { + "epoch": 0.2603750159052088, + "grad_norm": 0.18355484306812286, + "learning_rate": 4.618384426597827e-05, + "loss": 3.2162, + "step": 8441 + }, + { + "epoch": 0.2604058623707822, + "grad_norm": 0.1560346484184265, + "learning_rate": 4.6182414649600794e-05, + "loss": 3.1854, + "step": 8442 + }, + { + "epoch": 0.26043670883635567, + "grad_norm": 0.14962074160575867, + "learning_rate": 4.6180984787625946e-05, + "loss": 3.2256, + "step": 8443 + }, + { + "epoch": 0.26046755530192905, + "grad_norm": 0.15468266606330872, + "learning_rate": 4.6179554680070314e-05, + "loss": 3.2193, + "step": 8444 + }, + { + "epoch": 0.2604984017675025, + "grad_norm": 0.15393564105033875, + "learning_rate": 4.617812432695046e-05, + "loss": 3.2041, + "step": 8445 + }, + { + "epoch": 0.2605292482330759, + "grad_norm": 0.15022753179073334, + "learning_rate": 4.6176693728282996e-05, + "loss": 3.1737, + "step": 8446 + }, + { + "epoch": 0.2605600946986493, + "grad_norm": 0.23609283566474915, + "learning_rate": 4.617526288408448e-05, + "loss": 3.2229, + "step": 8447 + }, + { + "epoch": 0.2605909411642227, + "grad_norm": 0.15167362987995148, + "learning_rate": 4.6173831794371526e-05, + "loss": 3.2208, + "step": 8448 + }, + { + "epoch": 0.26062178762979615, + "grad_norm": 0.15219895541667938, + "learning_rate": 4.6172400459160716e-05, + "loss": 3.1811, + "step": 8449 + }, + { + "epoch": 0.26065263409536954, + "grad_norm": 0.15346749126911163, + "learning_rate": 4.617096887846863e-05, + "loss": 3.2262, + "step": 8450 + }, + { + "epoch": 0.260683480560943, + "grad_norm": 0.1525389403104782, + "learning_rate": 4.616953705231191e-05, + "loss": 3.2147, + "step": 8451 + }, + { + "epoch": 0.26071432702651637, + "grad_norm": 0.1587955355644226, + "learning_rate": 4.6168104980707107e-05, + "loss": 3.2259, + "step": 8452 + }, + { + "epoch": 0.2607451734920898, + "grad_norm": 0.15178227424621582, + "learning_rate": 4.616667266367085e-05, + "loss": 3.204, + "step": 8453 + }, + { + "epoch": 0.26077601995766325, + "grad_norm": 0.15781232714653015, + "learning_rate": 4.616524010121975e-05, + "loss": 3.2076, + "step": 8454 + }, + { + "epoch": 0.26080686642323664, + "grad_norm": 0.15433411300182343, + "learning_rate": 4.616380729337041e-05, + "loss": 3.1976, + "step": 8455 + }, + { + "epoch": 0.2608377128888101, + "grad_norm": 0.16547127068042755, + "learning_rate": 4.616237424013944e-05, + "loss": 3.1924, + "step": 8456 + }, + { + "epoch": 0.26086855935438347, + "grad_norm": 0.15809717774391174, + "learning_rate": 4.616094094154346e-05, + "loss": 3.17, + "step": 8457 + }, + { + "epoch": 0.2608994058199569, + "grad_norm": 0.16654406487941742, + "learning_rate": 4.615950739759908e-05, + "loss": 3.1866, + "step": 8458 + }, + { + "epoch": 0.2609302522855303, + "grad_norm": 0.16205671429634094, + "learning_rate": 4.615807360832293e-05, + "loss": 3.2288, + "step": 8459 + }, + { + "epoch": 0.26096109875110374, + "grad_norm": 0.1650528758764267, + "learning_rate": 4.615663957373163e-05, + "loss": 3.2317, + "step": 8460 + }, + { + "epoch": 0.2609919452166771, + "grad_norm": 0.16525177657604218, + "learning_rate": 4.615520529384181e-05, + "loss": 3.1886, + "step": 8461 + }, + { + "epoch": 0.26102279168225057, + "grad_norm": 0.15325073897838593, + "learning_rate": 4.61537707686701e-05, + "loss": 3.1638, + "step": 8462 + }, + { + "epoch": 0.26105363814782395, + "grad_norm": 0.1613917499780655, + "learning_rate": 4.6152335998233124e-05, + "loss": 3.2029, + "step": 8463 + }, + { + "epoch": 0.2610844846133974, + "grad_norm": 0.15551286935806274, + "learning_rate": 4.615090098254753e-05, + "loss": 3.2092, + "step": 8464 + }, + { + "epoch": 0.2611153310789708, + "grad_norm": 0.1516495943069458, + "learning_rate": 4.6149465721629936e-05, + "loss": 3.192, + "step": 8465 + }, + { + "epoch": 0.2611461775445442, + "grad_norm": 0.16028070449829102, + "learning_rate": 4.6148030215497015e-05, + "loss": 3.2165, + "step": 8466 + }, + { + "epoch": 0.26117702401011766, + "grad_norm": 0.15480343997478485, + "learning_rate": 4.6146594464165375e-05, + "loss": 3.1986, + "step": 8467 + }, + { + "epoch": 0.26120787047569105, + "grad_norm": 0.15441323816776276, + "learning_rate": 4.6145158467651686e-05, + "loss": 3.2056, + "step": 8468 + }, + { + "epoch": 0.2612387169412645, + "grad_norm": 0.14559341967105865, + "learning_rate": 4.61437222259726e-05, + "loss": 3.1578, + "step": 8469 + }, + { + "epoch": 0.2612695634068379, + "grad_norm": 0.1553826779127121, + "learning_rate": 4.6142285739144754e-05, + "loss": 3.2101, + "step": 8470 + }, + { + "epoch": 0.2613004098724113, + "grad_norm": 0.15496116876602173, + "learning_rate": 4.61408490071848e-05, + "loss": 3.196, + "step": 8471 + }, + { + "epoch": 0.2613312563379847, + "grad_norm": 0.15121155977249146, + "learning_rate": 4.613941203010942e-05, + "loss": 3.1884, + "step": 8472 + }, + { + "epoch": 0.26136210280355815, + "grad_norm": 0.15508045256137848, + "learning_rate": 4.6137974807935255e-05, + "loss": 3.1738, + "step": 8473 + }, + { + "epoch": 0.26139294926913154, + "grad_norm": 0.15433694422245026, + "learning_rate": 4.613653734067898e-05, + "loss": 3.2172, + "step": 8474 + }, + { + "epoch": 0.261423795734705, + "grad_norm": 0.1557684689760208, + "learning_rate": 4.6135099628357245e-05, + "loss": 3.1807, + "step": 8475 + }, + { + "epoch": 0.26145464220027836, + "grad_norm": 0.15793898701667786, + "learning_rate": 4.613366167098674e-05, + "loss": 3.1795, + "step": 8476 + }, + { + "epoch": 0.2614854886658518, + "grad_norm": 0.15638625621795654, + "learning_rate": 4.613222346858413e-05, + "loss": 3.1791, + "step": 8477 + }, + { + "epoch": 0.26151633513142525, + "grad_norm": 0.3167325556278229, + "learning_rate": 4.613078502116608e-05, + "loss": 3.2212, + "step": 8478 + }, + { + "epoch": 0.26154718159699863, + "grad_norm": 0.15675199031829834, + "learning_rate": 4.612934632874928e-05, + "loss": 3.2384, + "step": 8479 + }, + { + "epoch": 0.2615780280625721, + "grad_norm": 0.30604925751686096, + "learning_rate": 4.6127907391350406e-05, + "loss": 3.228, + "step": 8480 + }, + { + "epoch": 0.26160887452814546, + "grad_norm": 0.15465986728668213, + "learning_rate": 4.612646820898614e-05, + "loss": 3.1981, + "step": 8481 + }, + { + "epoch": 0.2616397209937189, + "grad_norm": 0.15727701783180237, + "learning_rate": 4.6125028781673174e-05, + "loss": 3.1848, + "step": 8482 + }, + { + "epoch": 0.2616705674592923, + "grad_norm": 0.1538253277540207, + "learning_rate": 4.612358910942819e-05, + "loss": 3.2006, + "step": 8483 + }, + { + "epoch": 0.26170141392486573, + "grad_norm": 0.15676866471767426, + "learning_rate": 4.6122149192267884e-05, + "loss": 3.1965, + "step": 8484 + }, + { + "epoch": 0.2617322603904391, + "grad_norm": 0.16662636399269104, + "learning_rate": 4.6120709030208954e-05, + "loss": 3.2001, + "step": 8485 + }, + { + "epoch": 0.26176310685601256, + "grad_norm": 0.1559368520975113, + "learning_rate": 4.6119268623268096e-05, + "loss": 3.2008, + "step": 8486 + }, + { + "epoch": 0.26179395332158595, + "grad_norm": 0.15696272253990173, + "learning_rate": 4.611782797146201e-05, + "loss": 3.2028, + "step": 8487 + }, + { + "epoch": 0.2618247997871594, + "grad_norm": 0.17010565102100372, + "learning_rate": 4.61163870748074e-05, + "loss": 3.2217, + "step": 8488 + }, + { + "epoch": 0.2618556462527328, + "grad_norm": 0.1551000028848648, + "learning_rate": 4.6114945933320964e-05, + "loss": 3.1707, + "step": 8489 + }, + { + "epoch": 0.2618864927183062, + "grad_norm": 0.16965270042419434, + "learning_rate": 4.611350454701942e-05, + "loss": 3.1863, + "step": 8490 + }, + { + "epoch": 0.26191733918387966, + "grad_norm": 0.15913020074367523, + "learning_rate": 4.6112062915919485e-05, + "loss": 3.1967, + "step": 8491 + }, + { + "epoch": 0.26194818564945305, + "grad_norm": 0.16241614520549774, + "learning_rate": 4.611062104003786e-05, + "loss": 3.2188, + "step": 8492 + }, + { + "epoch": 0.2619790321150265, + "grad_norm": 0.16189627349376678, + "learning_rate": 4.6109178919391274e-05, + "loss": 3.2288, + "step": 8493 + }, + { + "epoch": 0.2620098785805999, + "grad_norm": 0.16083656251430511, + "learning_rate": 4.610773655399644e-05, + "loss": 3.2061, + "step": 8494 + }, + { + "epoch": 0.2620407250461733, + "grad_norm": 0.16020752489566803, + "learning_rate": 4.6106293943870094e-05, + "loss": 3.2395, + "step": 8495 + }, + { + "epoch": 0.2620715715117467, + "grad_norm": 0.15986937284469604, + "learning_rate": 4.610485108902894e-05, + "loss": 3.2029, + "step": 8496 + }, + { + "epoch": 0.26210241797732015, + "grad_norm": 0.15208731591701508, + "learning_rate": 4.6103407989489736e-05, + "loss": 3.1885, + "step": 8497 + }, + { + "epoch": 0.26213326444289353, + "grad_norm": 0.15344034135341644, + "learning_rate": 4.6101964645269186e-05, + "loss": 3.242, + "step": 8498 + }, + { + "epoch": 0.262164110908467, + "grad_norm": 0.15283659100532532, + "learning_rate": 4.6100521056384037e-05, + "loss": 3.1754, + "step": 8499 + }, + { + "epoch": 0.26219495737404036, + "grad_norm": 0.1497875303030014, + "learning_rate": 4.6099077222851025e-05, + "loss": 3.1718, + "step": 8500 + }, + { + "epoch": 0.2622258038396138, + "grad_norm": 0.182483971118927, + "learning_rate": 4.60976331446869e-05, + "loss": 3.1512, + "step": 8501 + }, + { + "epoch": 0.26225665030518724, + "grad_norm": 0.1632506549358368, + "learning_rate": 4.609618882190839e-05, + "loss": 3.2264, + "step": 8502 + }, + { + "epoch": 0.26228749677076063, + "grad_norm": 0.14597222208976746, + "learning_rate": 4.609474425453226e-05, + "loss": 3.2239, + "step": 8503 + }, + { + "epoch": 0.2623183432363341, + "grad_norm": 0.1662517637014389, + "learning_rate": 4.6093299442575235e-05, + "loss": 3.1998, + "step": 8504 + }, + { + "epoch": 0.26234918970190746, + "grad_norm": 0.17382203042507172, + "learning_rate": 4.6091854386054076e-05, + "loss": 3.2211, + "step": 8505 + }, + { + "epoch": 0.2623800361674809, + "grad_norm": 0.15007257461547852, + "learning_rate": 4.6090409084985546e-05, + "loss": 3.2116, + "step": 8506 + }, + { + "epoch": 0.2624108826330543, + "grad_norm": 0.15081611275672913, + "learning_rate": 4.60889635393864e-05, + "loss": 3.2247, + "step": 8507 + }, + { + "epoch": 0.26244172909862773, + "grad_norm": 0.16498622298240662, + "learning_rate": 4.608751774927339e-05, + "loss": 3.2183, + "step": 8508 + }, + { + "epoch": 0.2624725755642011, + "grad_norm": 0.1542593389749527, + "learning_rate": 4.608607171466328e-05, + "loss": 3.208, + "step": 8509 + }, + { + "epoch": 0.26250342202977456, + "grad_norm": 0.15499445796012878, + "learning_rate": 4.608462543557285e-05, + "loss": 3.2171, + "step": 8510 + }, + { + "epoch": 0.26253426849534794, + "grad_norm": 0.16163191199302673, + "learning_rate": 4.608317891201885e-05, + "loss": 3.2194, + "step": 8511 + }, + { + "epoch": 0.2625651149609214, + "grad_norm": 0.14382855594158173, + "learning_rate": 4.608173214401806e-05, + "loss": 3.1877, + "step": 8512 + }, + { + "epoch": 0.2625959614264948, + "grad_norm": 0.15511338412761688, + "learning_rate": 4.6080285131587265e-05, + "loss": 3.2106, + "step": 8513 + }, + { + "epoch": 0.2626268078920682, + "grad_norm": 0.1515176147222519, + "learning_rate": 4.607883787474322e-05, + "loss": 3.1837, + "step": 8514 + }, + { + "epoch": 0.26265765435764166, + "grad_norm": 0.1530701369047165, + "learning_rate": 4.6077390373502715e-05, + "loss": 3.2055, + "step": 8515 + }, + { + "epoch": 0.26268850082321504, + "grad_norm": 0.15130175650119781, + "learning_rate": 4.6075942627882544e-05, + "loss": 3.1964, + "step": 8516 + }, + { + "epoch": 0.2627193472887885, + "grad_norm": 0.15753482282161713, + "learning_rate": 4.607449463789948e-05, + "loss": 3.2262, + "step": 8517 + }, + { + "epoch": 0.26275019375436187, + "grad_norm": 0.15850354731082916, + "learning_rate": 4.607304640357032e-05, + "loss": 3.2255, + "step": 8518 + }, + { + "epoch": 0.2627810402199353, + "grad_norm": 0.1536390483379364, + "learning_rate": 4.607159792491184e-05, + "loss": 3.2149, + "step": 8519 + }, + { + "epoch": 0.2628118866855087, + "grad_norm": 0.15579575300216675, + "learning_rate": 4.6070149201940846e-05, + "loss": 3.1674, + "step": 8520 + }, + { + "epoch": 0.26284273315108214, + "grad_norm": 0.14711984992027283, + "learning_rate": 4.6068700234674146e-05, + "loss": 3.2108, + "step": 8521 + }, + { + "epoch": 0.26287357961665553, + "grad_norm": 0.161619633436203, + "learning_rate": 4.606725102312852e-05, + "loss": 3.2276, + "step": 8522 + }, + { + "epoch": 0.26290442608222897, + "grad_norm": 0.147486612200737, + "learning_rate": 4.606580156732077e-05, + "loss": 3.1702, + "step": 8523 + }, + { + "epoch": 0.26293527254780236, + "grad_norm": 0.1479700207710266, + "learning_rate": 4.6064351867267716e-05, + "loss": 3.1835, + "step": 8524 + }, + { + "epoch": 0.2629661190133758, + "grad_norm": 0.15211519598960876, + "learning_rate": 4.6062901922986166e-05, + "loss": 3.2022, + "step": 8525 + }, + { + "epoch": 0.26299696547894924, + "grad_norm": 0.14972755312919617, + "learning_rate": 4.606145173449292e-05, + "loss": 3.216, + "step": 8526 + }, + { + "epoch": 0.2630278119445226, + "grad_norm": 0.1531565934419632, + "learning_rate": 4.60600013018048e-05, + "loss": 3.2364, + "step": 8527 + }, + { + "epoch": 0.26305865841009607, + "grad_norm": 0.14900067448616028, + "learning_rate": 4.605855062493862e-05, + "loss": 3.2134, + "step": 8528 + }, + { + "epoch": 0.26308950487566946, + "grad_norm": 0.15201863646507263, + "learning_rate": 4.60570997039112e-05, + "loss": 3.1968, + "step": 8529 + }, + { + "epoch": 0.2631203513412429, + "grad_norm": 0.14815525710582733, + "learning_rate": 4.605564853873936e-05, + "loss": 3.185, + "step": 8530 + }, + { + "epoch": 0.2631511978068163, + "grad_norm": 0.1521361917257309, + "learning_rate": 4.6054197129439925e-05, + "loss": 3.2088, + "step": 8531 + }, + { + "epoch": 0.2631820442723897, + "grad_norm": 0.1560104787349701, + "learning_rate": 4.6052745476029735e-05, + "loss": 3.2223, + "step": 8532 + }, + { + "epoch": 0.2632128907379631, + "grad_norm": 0.15418298542499542, + "learning_rate": 4.605129357852561e-05, + "loss": 3.2392, + "step": 8533 + }, + { + "epoch": 0.26324373720353655, + "grad_norm": 0.15328426659107208, + "learning_rate": 4.6049841436944385e-05, + "loss": 3.2265, + "step": 8534 + }, + { + "epoch": 0.26327458366910994, + "grad_norm": 0.14640267193317413, + "learning_rate": 4.6048389051302896e-05, + "loss": 3.1681, + "step": 8535 + }, + { + "epoch": 0.2633054301346834, + "grad_norm": 0.14752735197544098, + "learning_rate": 4.6046936421617995e-05, + "loss": 3.2275, + "step": 8536 + }, + { + "epoch": 0.2633362766002568, + "grad_norm": 0.146574467420578, + "learning_rate": 4.60454835479065e-05, + "loss": 3.2095, + "step": 8537 + }, + { + "epoch": 0.2633671230658302, + "grad_norm": 0.15525583922863007, + "learning_rate": 4.6044030430185276e-05, + "loss": 3.192, + "step": 8538 + }, + { + "epoch": 0.26339796953140365, + "grad_norm": 0.15273983776569366, + "learning_rate": 4.6042577068471164e-05, + "loss": 3.1622, + "step": 8539 + }, + { + "epoch": 0.26342881599697704, + "grad_norm": 0.15353913605213165, + "learning_rate": 4.6041123462781016e-05, + "loss": 3.1626, + "step": 8540 + }, + { + "epoch": 0.2634596624625505, + "grad_norm": 0.15142402052879333, + "learning_rate": 4.6039669613131684e-05, + "loss": 3.1795, + "step": 8541 + }, + { + "epoch": 0.26349050892812387, + "grad_norm": 0.1516169309616089, + "learning_rate": 4.603821551954003e-05, + "loss": 3.1902, + "step": 8542 + }, + { + "epoch": 0.2635213553936973, + "grad_norm": 0.15108460187911987, + "learning_rate": 4.6036761182022914e-05, + "loss": 3.214, + "step": 8543 + }, + { + "epoch": 0.2635522018592707, + "grad_norm": 0.14931826293468475, + "learning_rate": 4.603530660059718e-05, + "loss": 3.2391, + "step": 8544 + }, + { + "epoch": 0.26358304832484414, + "grad_norm": 0.15390680730342865, + "learning_rate": 4.603385177527972e-05, + "loss": 3.2041, + "step": 8545 + }, + { + "epoch": 0.2636138947904175, + "grad_norm": 0.15381570160388947, + "learning_rate": 4.603239670608739e-05, + "loss": 3.208, + "step": 8546 + }, + { + "epoch": 0.26364474125599097, + "grad_norm": 0.17367564141750336, + "learning_rate": 4.6030941393037056e-05, + "loss": 3.2086, + "step": 8547 + }, + { + "epoch": 0.26367558772156435, + "grad_norm": 0.14984333515167236, + "learning_rate": 4.602948583614558e-05, + "loss": 3.2124, + "step": 8548 + }, + { + "epoch": 0.2637064341871378, + "grad_norm": 0.1553225964307785, + "learning_rate": 4.602803003542987e-05, + "loss": 3.2115, + "step": 8549 + }, + { + "epoch": 0.26373728065271124, + "grad_norm": 0.1494092494249344, + "learning_rate": 4.602657399090678e-05, + "loss": 3.1914, + "step": 8550 + }, + { + "epoch": 0.2637681271182846, + "grad_norm": 0.1579565405845642, + "learning_rate": 4.6025117702593203e-05, + "loss": 3.1797, + "step": 8551 + }, + { + "epoch": 0.26379897358385807, + "grad_norm": 0.14749379456043243, + "learning_rate": 4.602366117050602e-05, + "loss": 3.1944, + "step": 8552 + }, + { + "epoch": 0.26382982004943145, + "grad_norm": 0.15712685883045197, + "learning_rate": 4.602220439466212e-05, + "loss": 3.1909, + "step": 8553 + }, + { + "epoch": 0.2638606665150049, + "grad_norm": 0.154163658618927, + "learning_rate": 4.602074737507839e-05, + "loss": 3.2095, + "step": 8554 + }, + { + "epoch": 0.2638915129805783, + "grad_norm": 0.15377219021320343, + "learning_rate": 4.601929011177173e-05, + "loss": 3.1832, + "step": 8555 + }, + { + "epoch": 0.2639223594461517, + "grad_norm": 0.16151489317417145, + "learning_rate": 4.601783260475903e-05, + "loss": 3.2135, + "step": 8556 + }, + { + "epoch": 0.2639532059117251, + "grad_norm": 0.1500660628080368, + "learning_rate": 4.60163748540572e-05, + "loss": 3.2262, + "step": 8557 + }, + { + "epoch": 0.26398405237729855, + "grad_norm": 0.16312889754772186, + "learning_rate": 4.601491685968312e-05, + "loss": 3.1977, + "step": 8558 + }, + { + "epoch": 0.26401489884287194, + "grad_norm": 0.14616593718528748, + "learning_rate": 4.601345862165372e-05, + "loss": 3.2239, + "step": 8559 + }, + { + "epoch": 0.2640457453084454, + "grad_norm": 0.24125567078590393, + "learning_rate": 4.601200013998588e-05, + "loss": 3.2078, + "step": 8560 + }, + { + "epoch": 0.2640765917740188, + "grad_norm": 0.15966369211673737, + "learning_rate": 4.601054141469654e-05, + "loss": 3.229, + "step": 8561 + }, + { + "epoch": 0.2641074382395922, + "grad_norm": 0.15026165544986725, + "learning_rate": 4.6009082445802585e-05, + "loss": 3.2115, + "step": 8562 + }, + { + "epoch": 0.26413828470516565, + "grad_norm": 0.14719119668006897, + "learning_rate": 4.600762323332095e-05, + "loss": 3.1838, + "step": 8563 + }, + { + "epoch": 0.26416913117073904, + "grad_norm": 0.15563669800758362, + "learning_rate": 4.6006163777268555e-05, + "loss": 3.2136, + "step": 8564 + }, + { + "epoch": 0.2641999776363125, + "grad_norm": 0.15092380344867706, + "learning_rate": 4.60047040776623e-05, + "loss": 3.2455, + "step": 8565 + }, + { + "epoch": 0.26423082410188586, + "grad_norm": 0.15347743034362793, + "learning_rate": 4.600324413451913e-05, + "loss": 3.2308, + "step": 8566 + }, + { + "epoch": 0.2642616705674593, + "grad_norm": 0.15492664277553558, + "learning_rate": 4.600178394785597e-05, + "loss": 3.1983, + "step": 8567 + }, + { + "epoch": 0.2642925170330327, + "grad_norm": 0.14918559789657593, + "learning_rate": 4.600032351768974e-05, + "loss": 3.1832, + "step": 8568 + }, + { + "epoch": 0.26432336349860613, + "grad_norm": 0.1702941358089447, + "learning_rate": 4.5998862844037385e-05, + "loss": 3.2051, + "step": 8569 + }, + { + "epoch": 0.2643542099641795, + "grad_norm": 0.1513349562883377, + "learning_rate": 4.599740192691583e-05, + "loss": 3.2127, + "step": 8570 + }, + { + "epoch": 0.26438505642975296, + "grad_norm": 0.1526581048965454, + "learning_rate": 4.5995940766342014e-05, + "loss": 3.183, + "step": 8571 + }, + { + "epoch": 0.26441590289532635, + "grad_norm": 0.15027105808258057, + "learning_rate": 4.599447936233289e-05, + "loss": 3.1656, + "step": 8572 + }, + { + "epoch": 0.2644467493608998, + "grad_norm": 0.15312163531780243, + "learning_rate": 4.599301771490539e-05, + "loss": 3.166, + "step": 8573 + }, + { + "epoch": 0.26447759582647323, + "grad_norm": 0.15350276231765747, + "learning_rate": 4.599155582407646e-05, + "loss": 3.2051, + "step": 8574 + }, + { + "epoch": 0.2645084422920466, + "grad_norm": 0.156635120511055, + "learning_rate": 4.5990093689863056e-05, + "loss": 3.2122, + "step": 8575 + }, + { + "epoch": 0.26453928875762006, + "grad_norm": 0.15048690140247345, + "learning_rate": 4.5988631312282135e-05, + "loss": 3.2283, + "step": 8576 + }, + { + "epoch": 0.26457013522319345, + "grad_norm": 0.15733256936073303, + "learning_rate": 4.598716869135065e-05, + "loss": 3.2008, + "step": 8577 + }, + { + "epoch": 0.2646009816887669, + "grad_norm": 0.15265290439128876, + "learning_rate": 4.598570582708555e-05, + "loss": 3.2362, + "step": 8578 + }, + { + "epoch": 0.2646318281543403, + "grad_norm": 0.1506376564502716, + "learning_rate": 4.5984242719503796e-05, + "loss": 3.2133, + "step": 8579 + }, + { + "epoch": 0.2646626746199137, + "grad_norm": 0.14865054190158844, + "learning_rate": 4.598277936862236e-05, + "loss": 3.2071, + "step": 8580 + }, + { + "epoch": 0.2646935210854871, + "grad_norm": 0.14987491071224213, + "learning_rate": 4.598131577445821e-05, + "loss": 3.2015, + "step": 8581 + }, + { + "epoch": 0.26472436755106055, + "grad_norm": 0.1459757536649704, + "learning_rate": 4.597985193702832e-05, + "loss": 3.2249, + "step": 8582 + }, + { + "epoch": 0.26475521401663393, + "grad_norm": 0.14823441207408905, + "learning_rate": 4.5978387856349643e-05, + "loss": 3.1882, + "step": 8583 + }, + { + "epoch": 0.2647860604822074, + "grad_norm": 0.15207017958164215, + "learning_rate": 4.5976923532439164e-05, + "loss": 3.2077, + "step": 8584 + }, + { + "epoch": 0.2648169069477808, + "grad_norm": 0.15731236338615417, + "learning_rate": 4.597545896531387e-05, + "loss": 3.2324, + "step": 8585 + }, + { + "epoch": 0.2648477534133542, + "grad_norm": 0.1627463847398758, + "learning_rate": 4.5973994154990727e-05, + "loss": 3.2427, + "step": 8586 + }, + { + "epoch": 0.26487859987892765, + "grad_norm": 0.14394879341125488, + "learning_rate": 4.597252910148673e-05, + "loss": 3.1735, + "step": 8587 + }, + { + "epoch": 0.26490944634450103, + "grad_norm": 0.15253563225269318, + "learning_rate": 4.597106380481887e-05, + "loss": 3.2195, + "step": 8588 + }, + { + "epoch": 0.2649402928100745, + "grad_norm": 0.14354990422725677, + "learning_rate": 4.596959826500411e-05, + "loss": 3.2179, + "step": 8589 + }, + { + "epoch": 0.26497113927564786, + "grad_norm": 0.14804844558238983, + "learning_rate": 4.5968132482059466e-05, + "loss": 3.2345, + "step": 8590 + }, + { + "epoch": 0.2650019857412213, + "grad_norm": 0.15312643349170685, + "learning_rate": 4.596666645600193e-05, + "loss": 3.2143, + "step": 8591 + }, + { + "epoch": 0.2650328322067947, + "grad_norm": 0.1507696658372879, + "learning_rate": 4.596520018684849e-05, + "loss": 3.1609, + "step": 8592 + }, + { + "epoch": 0.26506367867236813, + "grad_norm": 0.15381057560443878, + "learning_rate": 4.5963733674616164e-05, + "loss": 3.2036, + "step": 8593 + }, + { + "epoch": 0.2650945251379415, + "grad_norm": 0.1556359976530075, + "learning_rate": 4.596226691932193e-05, + "loss": 3.2189, + "step": 8594 + }, + { + "epoch": 0.26512537160351496, + "grad_norm": 0.15378186106681824, + "learning_rate": 4.596079992098281e-05, + "loss": 3.2208, + "step": 8595 + }, + { + "epoch": 0.26515621806908835, + "grad_norm": 0.15462203323841095, + "learning_rate": 4.595933267961582e-05, + "loss": 3.1979, + "step": 8596 + }, + { + "epoch": 0.2651870645346618, + "grad_norm": 0.14734914898872375, + "learning_rate": 4.595786519523796e-05, + "loss": 3.2379, + "step": 8597 + }, + { + "epoch": 0.26521791100023523, + "grad_norm": 0.1552390605211258, + "learning_rate": 4.595639746786624e-05, + "loss": 3.2095, + "step": 8598 + }, + { + "epoch": 0.2652487574658086, + "grad_norm": 0.16522446274757385, + "learning_rate": 4.5954929497517685e-05, + "loss": 3.1715, + "step": 8599 + }, + { + "epoch": 0.26527960393138206, + "grad_norm": 0.15520334243774414, + "learning_rate": 4.595346128420932e-05, + "loss": 3.2105, + "step": 8600 + }, + { + "epoch": 0.26531045039695544, + "grad_norm": 0.15699142217636108, + "learning_rate": 4.5951992827958166e-05, + "loss": 3.2165, + "step": 8601 + }, + { + "epoch": 0.2653412968625289, + "grad_norm": 0.15423522889614105, + "learning_rate": 4.595052412878124e-05, + "loss": 3.1785, + "step": 8602 + }, + { + "epoch": 0.2653721433281023, + "grad_norm": 0.1581113189458847, + "learning_rate": 4.5949055186695576e-05, + "loss": 3.1877, + "step": 8603 + }, + { + "epoch": 0.2654029897936757, + "grad_norm": 0.1558380275964737, + "learning_rate": 4.594758600171821e-05, + "loss": 3.1813, + "step": 8604 + }, + { + "epoch": 0.2654338362592491, + "grad_norm": 0.146762415766716, + "learning_rate": 4.594611657386617e-05, + "loss": 3.1871, + "step": 8605 + }, + { + "epoch": 0.26546468272482254, + "grad_norm": 0.15639320015907288, + "learning_rate": 4.5944646903156496e-05, + "loss": 3.2196, + "step": 8606 + }, + { + "epoch": 0.26549552919039593, + "grad_norm": 0.14967334270477295, + "learning_rate": 4.594317698960623e-05, + "loss": 3.2002, + "step": 8607 + }, + { + "epoch": 0.26552637565596937, + "grad_norm": 0.14819340407848358, + "learning_rate": 4.594170683323241e-05, + "loss": 3.2409, + "step": 8608 + }, + { + "epoch": 0.2655572221215428, + "grad_norm": 0.15705439448356628, + "learning_rate": 4.594023643405208e-05, + "loss": 3.2317, + "step": 8609 + }, + { + "epoch": 0.2655880685871162, + "grad_norm": 0.14460134506225586, + "learning_rate": 4.59387657920823e-05, + "loss": 3.1956, + "step": 8610 + }, + { + "epoch": 0.26561891505268964, + "grad_norm": 0.1540375053882599, + "learning_rate": 4.593729490734011e-05, + "loss": 3.202, + "step": 8611 + }, + { + "epoch": 0.26564976151826303, + "grad_norm": 0.1474883109331131, + "learning_rate": 4.5935823779842565e-05, + "loss": 3.2039, + "step": 8612 + }, + { + "epoch": 0.26568060798383647, + "grad_norm": 0.14976926147937775, + "learning_rate": 4.593435240960673e-05, + "loss": 3.1976, + "step": 8613 + }, + { + "epoch": 0.26571145444940986, + "grad_norm": 0.16688773036003113, + "learning_rate": 4.5932880796649646e-05, + "loss": 3.2001, + "step": 8614 + }, + { + "epoch": 0.2657423009149833, + "grad_norm": 0.1507244110107422, + "learning_rate": 4.5931408940988404e-05, + "loss": 3.2212, + "step": 8615 + }, + { + "epoch": 0.2657731473805567, + "grad_norm": 0.15111933648586273, + "learning_rate": 4.5929936842640045e-05, + "loss": 3.237, + "step": 8616 + }, + { + "epoch": 0.2658039938461301, + "grad_norm": 0.15489225089550018, + "learning_rate": 4.592846450162165e-05, + "loss": 3.156, + "step": 8617 + }, + { + "epoch": 0.2658348403117035, + "grad_norm": 0.15201883018016815, + "learning_rate": 4.592699191795028e-05, + "loss": 3.198, + "step": 8618 + }, + { + "epoch": 0.26586568677727695, + "grad_norm": 0.15482152998447418, + "learning_rate": 4.592551909164302e-05, + "loss": 3.2103, + "step": 8619 + }, + { + "epoch": 0.26589653324285034, + "grad_norm": 0.157907173037529, + "learning_rate": 4.592404602271694e-05, + "loss": 3.2113, + "step": 8620 + }, + { + "epoch": 0.2659273797084238, + "grad_norm": 0.15655672550201416, + "learning_rate": 4.592257271118911e-05, + "loss": 3.2044, + "step": 8621 + }, + { + "epoch": 0.2659582261739972, + "grad_norm": 0.1585032343864441, + "learning_rate": 4.5921099157076636e-05, + "loss": 3.2229, + "step": 8622 + }, + { + "epoch": 0.2659890726395706, + "grad_norm": 0.15848958492279053, + "learning_rate": 4.591962536039658e-05, + "loss": 3.1998, + "step": 8623 + }, + { + "epoch": 0.26601991910514405, + "grad_norm": 0.15522901713848114, + "learning_rate": 4.591815132116605e-05, + "loss": 3.2203, + "step": 8624 + }, + { + "epoch": 0.26605076557071744, + "grad_norm": 0.1581430286169052, + "learning_rate": 4.591667703940212e-05, + "loss": 3.226, + "step": 8625 + }, + { + "epoch": 0.2660816120362909, + "grad_norm": 0.1514531373977661, + "learning_rate": 4.591520251512188e-05, + "loss": 3.1934, + "step": 8626 + }, + { + "epoch": 0.26611245850186427, + "grad_norm": 0.1538553237915039, + "learning_rate": 4.591372774834245e-05, + "loss": 3.2208, + "step": 8627 + }, + { + "epoch": 0.2661433049674377, + "grad_norm": 0.15239274501800537, + "learning_rate": 4.591225273908091e-05, + "loss": 3.2045, + "step": 8628 + }, + { + "epoch": 0.2661741514330111, + "grad_norm": 0.15304900705814362, + "learning_rate": 4.591077748735436e-05, + "loss": 3.194, + "step": 8629 + }, + { + "epoch": 0.26620499789858454, + "grad_norm": 0.15745848417282104, + "learning_rate": 4.590930199317992e-05, + "loss": 3.2186, + "step": 8630 + }, + { + "epoch": 0.2662358443641579, + "grad_norm": 0.16479939222335815, + "learning_rate": 4.590782625657468e-05, + "loss": 3.1965, + "step": 8631 + }, + { + "epoch": 0.26626669082973137, + "grad_norm": 0.15884606540203094, + "learning_rate": 4.5906350277555764e-05, + "loss": 3.2165, + "step": 8632 + }, + { + "epoch": 0.2662975372953048, + "grad_norm": 0.15448950231075287, + "learning_rate": 4.590487405614028e-05, + "loss": 3.1953, + "step": 8633 + }, + { + "epoch": 0.2663283837608782, + "grad_norm": 0.16453491151332855, + "learning_rate": 4.590339759234534e-05, + "loss": 3.1931, + "step": 8634 + }, + { + "epoch": 0.26635923022645164, + "grad_norm": 0.16206619143486023, + "learning_rate": 4.590192088618807e-05, + "loss": 3.2076, + "step": 8635 + }, + { + "epoch": 0.266390076692025, + "grad_norm": 0.15697070956230164, + "learning_rate": 4.590044393768559e-05, + "loss": 3.2145, + "step": 8636 + }, + { + "epoch": 0.26642092315759847, + "grad_norm": 0.16497653722763062, + "learning_rate": 4.5898966746855014e-05, + "loss": 3.2193, + "step": 8637 + }, + { + "epoch": 0.26645176962317185, + "grad_norm": 0.1502954661846161, + "learning_rate": 4.589748931371348e-05, + "loss": 3.2281, + "step": 8638 + }, + { + "epoch": 0.2664826160887453, + "grad_norm": 0.15444190800189972, + "learning_rate": 4.5896011638278126e-05, + "loss": 3.1665, + "step": 8639 + }, + { + "epoch": 0.2665134625543187, + "grad_norm": 0.15235595405101776, + "learning_rate": 4.589453372056606e-05, + "loss": 3.1986, + "step": 8640 + }, + { + "epoch": 0.2665443090198921, + "grad_norm": 0.15341119468212128, + "learning_rate": 4.589305556059444e-05, + "loss": 3.2082, + "step": 8641 + }, + { + "epoch": 0.2665751554854655, + "grad_norm": 0.16401168704032898, + "learning_rate": 4.589157715838039e-05, + "loss": 3.1591, + "step": 8642 + }, + { + "epoch": 0.26660600195103895, + "grad_norm": 0.15508438646793365, + "learning_rate": 4.589009851394106e-05, + "loss": 3.1832, + "step": 8643 + }, + { + "epoch": 0.26663684841661234, + "grad_norm": 0.1582082062959671, + "learning_rate": 4.588861962729359e-05, + "loss": 3.1788, + "step": 8644 + }, + { + "epoch": 0.2666676948821858, + "grad_norm": 0.1521252989768982, + "learning_rate": 4.588714049845513e-05, + "loss": 3.1884, + "step": 8645 + }, + { + "epoch": 0.2666985413477592, + "grad_norm": 0.15385447442531586, + "learning_rate": 4.588566112744283e-05, + "loss": 3.2162, + "step": 8646 + }, + { + "epoch": 0.2667293878133326, + "grad_norm": 0.1495153307914734, + "learning_rate": 4.5884181514273836e-05, + "loss": 3.1931, + "step": 8647 + }, + { + "epoch": 0.26676023427890605, + "grad_norm": 0.1568268984556198, + "learning_rate": 4.588270165896531e-05, + "loss": 3.2317, + "step": 8648 + }, + { + "epoch": 0.26679108074447944, + "grad_norm": 0.15663443505764008, + "learning_rate": 4.58812215615344e-05, + "loss": 3.2012, + "step": 8649 + }, + { + "epoch": 0.2668219272100529, + "grad_norm": 0.16209334135055542, + "learning_rate": 4.587974122199828e-05, + "loss": 3.2114, + "step": 8650 + }, + { + "epoch": 0.26685277367562626, + "grad_norm": 0.16200105845928192, + "learning_rate": 4.5878260640374114e-05, + "loss": 3.176, + "step": 8651 + }, + { + "epoch": 0.2668836201411997, + "grad_norm": 0.16383084654808044, + "learning_rate": 4.587677981667906e-05, + "loss": 3.193, + "step": 8652 + }, + { + "epoch": 0.2669144666067731, + "grad_norm": 0.14811047911643982, + "learning_rate": 4.587529875093028e-05, + "loss": 3.1879, + "step": 8653 + }, + { + "epoch": 0.26694531307234654, + "grad_norm": 0.16984207928180695, + "learning_rate": 4.587381744314496e-05, + "loss": 3.2108, + "step": 8654 + }, + { + "epoch": 0.2669761595379199, + "grad_norm": 0.147295281291008, + "learning_rate": 4.587233589334028e-05, + "loss": 3.2048, + "step": 8655 + }, + { + "epoch": 0.26700700600349336, + "grad_norm": 0.16776466369628906, + "learning_rate": 4.58708541015334e-05, + "loss": 3.2236, + "step": 8656 + }, + { + "epoch": 0.2670378524690668, + "grad_norm": 0.15426690876483917, + "learning_rate": 4.586937206774151e-05, + "loss": 3.1967, + "step": 8657 + }, + { + "epoch": 0.2670686989346402, + "grad_norm": 0.1531650274991989, + "learning_rate": 4.586788979198179e-05, + "loss": 3.2004, + "step": 8658 + }, + { + "epoch": 0.26709954540021363, + "grad_norm": 0.15131835639476776, + "learning_rate": 4.586640727427143e-05, + "loss": 3.2112, + "step": 8659 + }, + { + "epoch": 0.267130391865787, + "grad_norm": 0.36842092871665955, + "learning_rate": 4.586492451462761e-05, + "loss": 3.2178, + "step": 8660 + }, + { + "epoch": 0.26716123833136046, + "grad_norm": 0.14869241416454315, + "learning_rate": 4.586344151306753e-05, + "loss": 3.1907, + "step": 8661 + }, + { + "epoch": 0.26719208479693385, + "grad_norm": 0.14660929143428802, + "learning_rate": 4.586195826960839e-05, + "loss": 3.1863, + "step": 8662 + }, + { + "epoch": 0.2672229312625073, + "grad_norm": 0.15071897208690643, + "learning_rate": 4.586047478426738e-05, + "loss": 3.2141, + "step": 8663 + }, + { + "epoch": 0.2672537777280807, + "grad_norm": 0.17954804003238678, + "learning_rate": 4.5858991057061694e-05, + "loss": 3.2014, + "step": 8664 + }, + { + "epoch": 0.2672846241936541, + "grad_norm": 0.158795565366745, + "learning_rate": 4.585750708800855e-05, + "loss": 3.2063, + "step": 8665 + }, + { + "epoch": 0.2673154706592275, + "grad_norm": 0.14811120927333832, + "learning_rate": 4.585602287712514e-05, + "loss": 3.2056, + "step": 8666 + }, + { + "epoch": 0.26734631712480095, + "grad_norm": 0.15910698473453522, + "learning_rate": 4.585453842442868e-05, + "loss": 3.2007, + "step": 8667 + }, + { + "epoch": 0.26737716359037433, + "grad_norm": 0.15271832048892975, + "learning_rate": 4.5853053729936376e-05, + "loss": 3.2032, + "step": 8668 + }, + { + "epoch": 0.2674080100559478, + "grad_norm": 0.15352529287338257, + "learning_rate": 4.5851568793665446e-05, + "loss": 3.2089, + "step": 8669 + }, + { + "epoch": 0.2674388565215212, + "grad_norm": 0.1496545374393463, + "learning_rate": 4.585008361563311e-05, + "loss": 3.2036, + "step": 8670 + }, + { + "epoch": 0.2674697029870946, + "grad_norm": 0.16172489523887634, + "learning_rate": 4.584859819585658e-05, + "loss": 3.2268, + "step": 8671 + }, + { + "epoch": 0.26750054945266805, + "grad_norm": 0.15261931717395782, + "learning_rate": 4.584711253435309e-05, + "loss": 3.1966, + "step": 8672 + }, + { + "epoch": 0.26753139591824143, + "grad_norm": 0.15048865973949432, + "learning_rate": 4.5845626631139846e-05, + "loss": 3.2294, + "step": 8673 + }, + { + "epoch": 0.2675622423838149, + "grad_norm": 0.148284450173378, + "learning_rate": 4.5844140486234086e-05, + "loss": 3.1832, + "step": 8674 + }, + { + "epoch": 0.26759308884938826, + "grad_norm": 0.15467579662799835, + "learning_rate": 4.584265409965306e-05, + "loss": 3.2124, + "step": 8675 + }, + { + "epoch": 0.2676239353149617, + "grad_norm": 0.15430638194084167, + "learning_rate": 4.584116747141397e-05, + "loss": 3.2307, + "step": 8676 + }, + { + "epoch": 0.2676547817805351, + "grad_norm": 0.15083159506320953, + "learning_rate": 4.5839680601534074e-05, + "loss": 3.2115, + "step": 8677 + }, + { + "epoch": 0.26768562824610853, + "grad_norm": 0.15853305160999298, + "learning_rate": 4.58381934900306e-05, + "loss": 3.224, + "step": 8678 + }, + { + "epoch": 0.2677164747116819, + "grad_norm": 0.1539057344198227, + "learning_rate": 4.583670613692081e-05, + "loss": 3.2239, + "step": 8679 + }, + { + "epoch": 0.26774732117725536, + "grad_norm": 0.1494709551334381, + "learning_rate": 4.5835218542221914e-05, + "loss": 3.1716, + "step": 8680 + }, + { + "epoch": 0.2677781676428288, + "grad_norm": 0.16589507460594177, + "learning_rate": 4.5833730705951184e-05, + "loss": 3.2078, + "step": 8681 + }, + { + "epoch": 0.2678090141084022, + "grad_norm": 0.15570437908172607, + "learning_rate": 4.583224262812587e-05, + "loss": 3.2096, + "step": 8682 + }, + { + "epoch": 0.26783986057397563, + "grad_norm": 0.15483178198337555, + "learning_rate": 4.583075430876322e-05, + "loss": 3.1832, + "step": 8683 + }, + { + "epoch": 0.267870707039549, + "grad_norm": 0.14590157568454742, + "learning_rate": 4.582926574788049e-05, + "loss": 3.1815, + "step": 8684 + }, + { + "epoch": 0.26790155350512246, + "grad_norm": 0.14914363622665405, + "learning_rate": 4.582777694549495e-05, + "loss": 3.1549, + "step": 8685 + }, + { + "epoch": 0.26793239997069584, + "grad_norm": 0.1504165232181549, + "learning_rate": 4.5826287901623844e-05, + "loss": 3.2211, + "step": 8686 + }, + { + "epoch": 0.2679632464362693, + "grad_norm": 0.15040788054466248, + "learning_rate": 4.582479861628444e-05, + "loss": 3.2221, + "step": 8687 + }, + { + "epoch": 0.2679940929018427, + "grad_norm": 0.1519264429807663, + "learning_rate": 4.5823309089494015e-05, + "loss": 3.2157, + "step": 8688 + }, + { + "epoch": 0.2680249393674161, + "grad_norm": 0.1545119732618332, + "learning_rate": 4.582181932126984e-05, + "loss": 3.2188, + "step": 8689 + }, + { + "epoch": 0.2680557858329895, + "grad_norm": 0.1507766991853714, + "learning_rate": 4.582032931162917e-05, + "loss": 3.2413, + "step": 8690 + }, + { + "epoch": 0.26808663229856294, + "grad_norm": 0.15159527957439423, + "learning_rate": 4.5818839060589306e-05, + "loss": 3.2425, + "step": 8691 + }, + { + "epoch": 0.26811747876413633, + "grad_norm": 0.14981809258460999, + "learning_rate": 4.5817348568167506e-05, + "loss": 3.1872, + "step": 8692 + }, + { + "epoch": 0.26814832522970977, + "grad_norm": 0.14829829335212708, + "learning_rate": 4.5815857834381065e-05, + "loss": 3.1945, + "step": 8693 + }, + { + "epoch": 0.2681791716952832, + "grad_norm": 0.15805283188819885, + "learning_rate": 4.581436685924726e-05, + "loss": 3.1873, + "step": 8694 + }, + { + "epoch": 0.2682100181608566, + "grad_norm": 0.1477857232093811, + "learning_rate": 4.5812875642783374e-05, + "loss": 3.179, + "step": 8695 + }, + { + "epoch": 0.26824086462643004, + "grad_norm": 0.14949913322925568, + "learning_rate": 4.581138418500671e-05, + "loss": 3.1936, + "step": 8696 + }, + { + "epoch": 0.26827171109200343, + "grad_norm": 0.14775292575359344, + "learning_rate": 4.580989248593455e-05, + "loss": 3.1908, + "step": 8697 + }, + { + "epoch": 0.26830255755757687, + "grad_norm": 0.14615832269191742, + "learning_rate": 4.580840054558419e-05, + "loss": 3.1884, + "step": 8698 + }, + { + "epoch": 0.26833340402315026, + "grad_norm": 0.1478612869977951, + "learning_rate": 4.580690836397293e-05, + "loss": 3.1796, + "step": 8699 + }, + { + "epoch": 0.2683642504887237, + "grad_norm": 0.14542047679424286, + "learning_rate": 4.580541594111807e-05, + "loss": 3.2145, + "step": 8700 + }, + { + "epoch": 0.2683950969542971, + "grad_norm": 0.14839255809783936, + "learning_rate": 4.580392327703691e-05, + "loss": 3.2148, + "step": 8701 + }, + { + "epoch": 0.2684259434198705, + "grad_norm": 0.14987801015377045, + "learning_rate": 4.580243037174677e-05, + "loss": 3.1929, + "step": 8702 + }, + { + "epoch": 0.2684567898854439, + "grad_norm": 0.14902479946613312, + "learning_rate": 4.580093722526495e-05, + "loss": 3.2215, + "step": 8703 + }, + { + "epoch": 0.26848763635101736, + "grad_norm": 0.14644703269004822, + "learning_rate": 4.5799443837608755e-05, + "loss": 3.1993, + "step": 8704 + }, + { + "epoch": 0.2685184828165908, + "grad_norm": 0.1527256965637207, + "learning_rate": 4.579795020879552e-05, + "loss": 3.215, + "step": 8705 + }, + { + "epoch": 0.2685493292821642, + "grad_norm": 0.2685375213623047, + "learning_rate": 4.579645633884254e-05, + "loss": 3.222, + "step": 8706 + }, + { + "epoch": 0.2685801757477376, + "grad_norm": 0.1536417156457901, + "learning_rate": 4.579496222776716e-05, + "loss": 3.2156, + "step": 8707 + }, + { + "epoch": 0.268611022213311, + "grad_norm": 0.14926879107952118, + "learning_rate": 4.579346787558667e-05, + "loss": 3.1988, + "step": 8708 + }, + { + "epoch": 0.26864186867888445, + "grad_norm": 0.14884215593338013, + "learning_rate": 4.579197328231842e-05, + "loss": 3.1993, + "step": 8709 + }, + { + "epoch": 0.26867271514445784, + "grad_norm": 0.15060827136039734, + "learning_rate": 4.579047844797975e-05, + "loss": 3.1955, + "step": 8710 + }, + { + "epoch": 0.2687035616100313, + "grad_norm": 0.16786418855190277, + "learning_rate": 4.5788983372587965e-05, + "loss": 3.2329, + "step": 8711 + }, + { + "epoch": 0.26873440807560467, + "grad_norm": 0.15034009516239166, + "learning_rate": 4.578748805616041e-05, + "loss": 3.1828, + "step": 8712 + }, + { + "epoch": 0.2687652545411781, + "grad_norm": 0.15027156472206116, + "learning_rate": 4.578599249871442e-05, + "loss": 3.2412, + "step": 8713 + }, + { + "epoch": 0.2687961010067515, + "grad_norm": 0.14805425703525543, + "learning_rate": 4.578449670026734e-05, + "loss": 3.1909, + "step": 8714 + }, + { + "epoch": 0.26882694747232494, + "grad_norm": 0.15554119646549225, + "learning_rate": 4.578300066083651e-05, + "loss": 3.2278, + "step": 8715 + }, + { + "epoch": 0.2688577939378983, + "grad_norm": 0.14708201587200165, + "learning_rate": 4.5781504380439285e-05, + "loss": 3.2198, + "step": 8716 + }, + { + "epoch": 0.26888864040347177, + "grad_norm": 0.15052548050880432, + "learning_rate": 4.5780007859092993e-05, + "loss": 3.2226, + "step": 8717 + }, + { + "epoch": 0.2689194868690452, + "grad_norm": 0.15036337077617645, + "learning_rate": 4.5778511096815004e-05, + "loss": 3.217, + "step": 8718 + }, + { + "epoch": 0.2689503333346186, + "grad_norm": 0.14815761148929596, + "learning_rate": 4.5777014093622664e-05, + "loss": 3.1992, + "step": 8719 + }, + { + "epoch": 0.26898117980019204, + "grad_norm": 0.14497824013233185, + "learning_rate": 4.577551684953333e-05, + "loss": 3.1857, + "step": 8720 + }, + { + "epoch": 0.2690120262657654, + "grad_norm": 0.15512944757938385, + "learning_rate": 4.577401936456436e-05, + "loss": 3.2035, + "step": 8721 + }, + { + "epoch": 0.26904287273133887, + "grad_norm": 0.14867529273033142, + "learning_rate": 4.5772521638733126e-05, + "loss": 3.2127, + "step": 8722 + }, + { + "epoch": 0.26907371919691225, + "grad_norm": 0.14241889119148254, + "learning_rate": 4.577102367205698e-05, + "loss": 3.1944, + "step": 8723 + }, + { + "epoch": 0.2691045656624857, + "grad_norm": 0.16301469504833221, + "learning_rate": 4.576952546455331e-05, + "loss": 3.1956, + "step": 8724 + }, + { + "epoch": 0.2691354121280591, + "grad_norm": 0.1479911208152771, + "learning_rate": 4.576802701623946e-05, + "loss": 3.2174, + "step": 8725 + }, + { + "epoch": 0.2691662585936325, + "grad_norm": 0.14658337831497192, + "learning_rate": 4.576652832713282e-05, + "loss": 3.204, + "step": 8726 + }, + { + "epoch": 0.2691971050592059, + "grad_norm": 0.1499091237783432, + "learning_rate": 4.576502939725076e-05, + "loss": 3.1993, + "step": 8727 + }, + { + "epoch": 0.26922795152477935, + "grad_norm": 0.1470576375722885, + "learning_rate": 4.576353022661067e-05, + "loss": 3.166, + "step": 8728 + }, + { + "epoch": 0.2692587979903528, + "grad_norm": 0.15416228771209717, + "learning_rate": 4.576203081522992e-05, + "loss": 3.2116, + "step": 8729 + }, + { + "epoch": 0.2692896444559262, + "grad_norm": 0.1501166969537735, + "learning_rate": 4.5760531163125896e-05, + "loss": 3.2139, + "step": 8730 + }, + { + "epoch": 0.2693204909214996, + "grad_norm": 0.1732725203037262, + "learning_rate": 4.575903127031599e-05, + "loss": 3.2058, + "step": 8731 + }, + { + "epoch": 0.269351337387073, + "grad_norm": 0.15132898092269897, + "learning_rate": 4.57575311368176e-05, + "loss": 3.2363, + "step": 8732 + }, + { + "epoch": 0.26938218385264645, + "grad_norm": 0.14974044263362885, + "learning_rate": 4.57560307626481e-05, + "loss": 3.1565, + "step": 8733 + }, + { + "epoch": 0.26941303031821984, + "grad_norm": 0.14540068805217743, + "learning_rate": 4.57545301478249e-05, + "loss": 3.1977, + "step": 8734 + }, + { + "epoch": 0.2694438767837933, + "grad_norm": 0.14883273839950562, + "learning_rate": 4.5753029292365395e-05, + "loss": 3.1859, + "step": 8735 + }, + { + "epoch": 0.26947472324936667, + "grad_norm": 0.14986342191696167, + "learning_rate": 4.5751528196286984e-05, + "loss": 3.183, + "step": 8736 + }, + { + "epoch": 0.2695055697149401, + "grad_norm": 0.14949360489845276, + "learning_rate": 4.5750026859607075e-05, + "loss": 3.2201, + "step": 8737 + }, + { + "epoch": 0.2695364161805135, + "grad_norm": 0.1505255401134491, + "learning_rate": 4.574852528234308e-05, + "loss": 3.2029, + "step": 8738 + }, + { + "epoch": 0.26956726264608694, + "grad_norm": 0.1489306092262268, + "learning_rate": 4.57470234645124e-05, + "loss": 3.1945, + "step": 8739 + }, + { + "epoch": 0.2695981091116603, + "grad_norm": 0.14747847616672516, + "learning_rate": 4.574552140613244e-05, + "loss": 3.153, + "step": 8740 + }, + { + "epoch": 0.26962895557723376, + "grad_norm": 0.14712412655353546, + "learning_rate": 4.5744019107220635e-05, + "loss": 3.2013, + "step": 8741 + }, + { + "epoch": 0.2696598020428072, + "grad_norm": 0.15842796862125397, + "learning_rate": 4.5742516567794396e-05, + "loss": 3.1803, + "step": 8742 + }, + { + "epoch": 0.2696906485083806, + "grad_norm": 0.15397758781909943, + "learning_rate": 4.574101378787114e-05, + "loss": 3.2292, + "step": 8743 + }, + { + "epoch": 0.26972149497395403, + "grad_norm": 0.15826007723808289, + "learning_rate": 4.5739510767468295e-05, + "loss": 3.1957, + "step": 8744 + }, + { + "epoch": 0.2697523414395274, + "grad_norm": 0.15232431888580322, + "learning_rate": 4.573800750660328e-05, + "loss": 3.1792, + "step": 8745 + }, + { + "epoch": 0.26978318790510086, + "grad_norm": 0.15676020085811615, + "learning_rate": 4.573650400529353e-05, + "loss": 3.1974, + "step": 8746 + }, + { + "epoch": 0.26981403437067425, + "grad_norm": 0.15828917920589447, + "learning_rate": 4.5735000263556485e-05, + "loss": 3.2249, + "step": 8747 + }, + { + "epoch": 0.2698448808362477, + "grad_norm": 0.15104810893535614, + "learning_rate": 4.573349628140957e-05, + "loss": 3.2327, + "step": 8748 + }, + { + "epoch": 0.2698757273018211, + "grad_norm": 0.17297805845737457, + "learning_rate": 4.573199205887022e-05, + "loss": 3.1835, + "step": 8749 + }, + { + "epoch": 0.2699065737673945, + "grad_norm": 0.15572655200958252, + "learning_rate": 4.5730487595955885e-05, + "loss": 3.1681, + "step": 8750 + }, + { + "epoch": 0.2699374202329679, + "grad_norm": 0.15829943120479584, + "learning_rate": 4.5728982892684004e-05, + "loss": 3.1777, + "step": 8751 + }, + { + "epoch": 0.26996826669854135, + "grad_norm": 0.16015079617500305, + "learning_rate": 4.5727477949072014e-05, + "loss": 3.2456, + "step": 8752 + }, + { + "epoch": 0.2699991131641148, + "grad_norm": 0.1492297500371933, + "learning_rate": 4.572597276513738e-05, + "loss": 3.251, + "step": 8753 + }, + { + "epoch": 0.2700299596296882, + "grad_norm": 0.15780745446681976, + "learning_rate": 4.5724467340897546e-05, + "loss": 3.21, + "step": 8754 + }, + { + "epoch": 0.2700608060952616, + "grad_norm": 0.15250839293003082, + "learning_rate": 4.5722961676369966e-05, + "loss": 3.2096, + "step": 8755 + }, + { + "epoch": 0.270091652560835, + "grad_norm": 0.15214350819587708, + "learning_rate": 4.5721455771572106e-05, + "loss": 3.2264, + "step": 8756 + }, + { + "epoch": 0.27012249902640845, + "grad_norm": 0.14990894496440887, + "learning_rate": 4.5719949626521414e-05, + "loss": 3.1944, + "step": 8757 + }, + { + "epoch": 0.27015334549198183, + "grad_norm": 0.15423338115215302, + "learning_rate": 4.571844324123535e-05, + "loss": 3.2195, + "step": 8758 + }, + { + "epoch": 0.2701841919575553, + "grad_norm": 0.16203026473522186, + "learning_rate": 4.57169366157314e-05, + "loss": 3.1909, + "step": 8759 + }, + { + "epoch": 0.27021503842312866, + "grad_norm": 0.28030380606651306, + "learning_rate": 4.571542975002701e-05, + "loss": 3.2111, + "step": 8760 + }, + { + "epoch": 0.2702458848887021, + "grad_norm": 0.15138033032417297, + "learning_rate": 4.571392264413966e-05, + "loss": 3.1912, + "step": 8761 + }, + { + "epoch": 0.2702767313542755, + "grad_norm": 0.16126975417137146, + "learning_rate": 4.571241529808683e-05, + "loss": 3.1928, + "step": 8762 + }, + { + "epoch": 0.27030757781984893, + "grad_norm": 0.15699848532676697, + "learning_rate": 4.571090771188598e-05, + "loss": 3.2276, + "step": 8763 + }, + { + "epoch": 0.2703384242854223, + "grad_norm": 0.15293057262897491, + "learning_rate": 4.5709399885554614e-05, + "loss": 3.2113, + "step": 8764 + }, + { + "epoch": 0.27036927075099576, + "grad_norm": 0.15394416451454163, + "learning_rate": 4.5707891819110207e-05, + "loss": 3.1769, + "step": 8765 + }, + { + "epoch": 0.2704001172165692, + "grad_norm": 0.15244311094284058, + "learning_rate": 4.5706383512570225e-05, + "loss": 3.1832, + "step": 8766 + }, + { + "epoch": 0.2704309636821426, + "grad_norm": 0.1501310169696808, + "learning_rate": 4.570487496595217e-05, + "loss": 3.1546, + "step": 8767 + }, + { + "epoch": 0.27046181014771603, + "grad_norm": 0.16108426451683044, + "learning_rate": 4.5703366179273534e-05, + "loss": 3.1889, + "step": 8768 + }, + { + "epoch": 0.2704926566132894, + "grad_norm": 0.15189896523952484, + "learning_rate": 4.5701857152551805e-05, + "loss": 3.1907, + "step": 8769 + }, + { + "epoch": 0.27052350307886286, + "grad_norm": 0.16339914500713348, + "learning_rate": 4.5700347885804493e-05, + "loss": 3.1803, + "step": 8770 + }, + { + "epoch": 0.27055434954443625, + "grad_norm": 0.15242943167686462, + "learning_rate": 4.569883837904908e-05, + "loss": 3.2186, + "step": 8771 + }, + { + "epoch": 0.2705851960100097, + "grad_norm": 0.15325325727462769, + "learning_rate": 4.569732863230307e-05, + "loss": 3.2037, + "step": 8772 + }, + { + "epoch": 0.2706160424755831, + "grad_norm": 0.15196797251701355, + "learning_rate": 4.569581864558398e-05, + "loss": 3.1975, + "step": 8773 + }, + { + "epoch": 0.2706468889411565, + "grad_norm": 0.15063069760799408, + "learning_rate": 4.569430841890931e-05, + "loss": 3.2138, + "step": 8774 + }, + { + "epoch": 0.2706777354067299, + "grad_norm": 0.15847232937812805, + "learning_rate": 4.569279795229656e-05, + "loss": 3.1973, + "step": 8775 + }, + { + "epoch": 0.27070858187230334, + "grad_norm": 0.14717832207679749, + "learning_rate": 4.569128724576327e-05, + "loss": 3.2317, + "step": 8776 + }, + { + "epoch": 0.2707394283378768, + "grad_norm": 0.1664283126592636, + "learning_rate": 4.568977629932693e-05, + "loss": 3.217, + "step": 8777 + }, + { + "epoch": 0.2707702748034502, + "grad_norm": 0.14965948462486267, + "learning_rate": 4.5688265113005056e-05, + "loss": 3.189, + "step": 8778 + }, + { + "epoch": 0.2708011212690236, + "grad_norm": 0.15220220386981964, + "learning_rate": 4.5686753686815184e-05, + "loss": 3.1816, + "step": 8779 + }, + { + "epoch": 0.270831967734597, + "grad_norm": 0.14381735026836395, + "learning_rate": 4.568524202077484e-05, + "loss": 3.2008, + "step": 8780 + }, + { + "epoch": 0.27086281420017044, + "grad_norm": 0.16161632537841797, + "learning_rate": 4.5683730114901544e-05, + "loss": 3.1794, + "step": 8781 + }, + { + "epoch": 0.27089366066574383, + "grad_norm": 0.14745166897773743, + "learning_rate": 4.568221796921283e-05, + "loss": 3.2027, + "step": 8782 + }, + { + "epoch": 0.27092450713131727, + "grad_norm": 0.156106635928154, + "learning_rate": 4.568070558372622e-05, + "loss": 3.1979, + "step": 8783 + }, + { + "epoch": 0.27095535359689066, + "grad_norm": 0.15569014847278595, + "learning_rate": 4.5679192958459264e-05, + "loss": 3.2185, + "step": 8784 + }, + { + "epoch": 0.2709862000624641, + "grad_norm": 0.15145641565322876, + "learning_rate": 4.5677680093429486e-05, + "loss": 3.1914, + "step": 8785 + }, + { + "epoch": 0.2710170465280375, + "grad_norm": 0.1534300446510315, + "learning_rate": 4.567616698865444e-05, + "loss": 3.2358, + "step": 8786 + }, + { + "epoch": 0.27104789299361093, + "grad_norm": 0.15218228101730347, + "learning_rate": 4.567465364415165e-05, + "loss": 3.2078, + "step": 8787 + }, + { + "epoch": 0.2710787394591843, + "grad_norm": 0.14868931472301483, + "learning_rate": 4.567314005993869e-05, + "loss": 3.1901, + "step": 8788 + }, + { + "epoch": 0.27110958592475776, + "grad_norm": 0.1481381058692932, + "learning_rate": 4.5671626236033086e-05, + "loss": 3.2093, + "step": 8789 + }, + { + "epoch": 0.2711404323903312, + "grad_norm": 0.15244083106517792, + "learning_rate": 4.5670112172452406e-05, + "loss": 3.1965, + "step": 8790 + }, + { + "epoch": 0.2711712788559046, + "grad_norm": 0.14697670936584473, + "learning_rate": 4.566859786921419e-05, + "loss": 3.1907, + "step": 8791 + }, + { + "epoch": 0.271202125321478, + "grad_norm": 0.15316201746463776, + "learning_rate": 4.5667083326336e-05, + "loss": 3.1995, + "step": 8792 + }, + { + "epoch": 0.2712329717870514, + "grad_norm": 0.17691873013973236, + "learning_rate": 4.56655685438354e-05, + "loss": 3.2007, + "step": 8793 + }, + { + "epoch": 0.27126381825262486, + "grad_norm": 0.1522255390882492, + "learning_rate": 4.566405352172996e-05, + "loss": 3.2553, + "step": 8794 + }, + { + "epoch": 0.27129466471819824, + "grad_norm": 0.15504808723926544, + "learning_rate": 4.566253826003723e-05, + "loss": 3.2002, + "step": 8795 + }, + { + "epoch": 0.2713255111837717, + "grad_norm": 0.1479077786207199, + "learning_rate": 4.566102275877477e-05, + "loss": 3.2133, + "step": 8796 + }, + { + "epoch": 0.27135635764934507, + "grad_norm": 0.15038259327411652, + "learning_rate": 4.565950701796019e-05, + "loss": 3.1993, + "step": 8797 + }, + { + "epoch": 0.2713872041149185, + "grad_norm": 0.1550457626581192, + "learning_rate": 4.565799103761104e-05, + "loss": 3.2146, + "step": 8798 + }, + { + "epoch": 0.2714180505804919, + "grad_norm": 0.1501491218805313, + "learning_rate": 4.565647481774489e-05, + "loss": 3.1842, + "step": 8799 + }, + { + "epoch": 0.27144889704606534, + "grad_norm": 0.14941492676734924, + "learning_rate": 4.5654958358379316e-05, + "loss": 3.2187, + "step": 8800 + }, + { + "epoch": 0.2714797435116388, + "grad_norm": 0.14940008521080017, + "learning_rate": 4.565344165953193e-05, + "loss": 3.1736, + "step": 8801 + }, + { + "epoch": 0.27151058997721217, + "grad_norm": 0.15570853650569916, + "learning_rate": 4.56519247212203e-05, + "loss": 3.1591, + "step": 8802 + }, + { + "epoch": 0.2715414364427856, + "grad_norm": 0.15143583714962006, + "learning_rate": 4.5650407543462e-05, + "loss": 3.2187, + "step": 8803 + }, + { + "epoch": 0.271572282908359, + "grad_norm": 0.28475597500801086, + "learning_rate": 4.5648890126274644e-05, + "loss": 3.2122, + "step": 8804 + }, + { + "epoch": 0.27160312937393244, + "grad_norm": 0.15476791560649872, + "learning_rate": 4.564737246967581e-05, + "loss": 3.2004, + "step": 8805 + }, + { + "epoch": 0.2716339758395058, + "grad_norm": 0.1569722443819046, + "learning_rate": 4.56458545736831e-05, + "loss": 3.2079, + "step": 8806 + }, + { + "epoch": 0.27166482230507927, + "grad_norm": 0.1468389630317688, + "learning_rate": 4.564433643831412e-05, + "loss": 3.1639, + "step": 8807 + }, + { + "epoch": 0.27169566877065265, + "grad_norm": 0.14694616198539734, + "learning_rate": 4.564281806358646e-05, + "loss": 3.1761, + "step": 8808 + }, + { + "epoch": 0.2717265152362261, + "grad_norm": 0.1530999094247818, + "learning_rate": 4.5641299449517726e-05, + "loss": 3.2279, + "step": 8809 + }, + { + "epoch": 0.2717573617017995, + "grad_norm": 0.16314998269081116, + "learning_rate": 4.5639780596125534e-05, + "loss": 3.1958, + "step": 8810 + }, + { + "epoch": 0.2717882081673729, + "grad_norm": 0.1644883155822754, + "learning_rate": 4.5638261503427486e-05, + "loss": 3.2137, + "step": 8811 + }, + { + "epoch": 0.27181905463294637, + "grad_norm": 0.15290948748588562, + "learning_rate": 4.56367421714412e-05, + "loss": 3.1803, + "step": 8812 + }, + { + "epoch": 0.27184990109851975, + "grad_norm": 0.15498469769954681, + "learning_rate": 4.563522260018429e-05, + "loss": 3.2249, + "step": 8813 + }, + { + "epoch": 0.2718807475640932, + "grad_norm": 0.16023951768875122, + "learning_rate": 4.563370278967437e-05, + "loss": 3.1924, + "step": 8814 + }, + { + "epoch": 0.2719115940296666, + "grad_norm": 0.15302908420562744, + "learning_rate": 4.563218273992908e-05, + "loss": 3.2588, + "step": 8815 + }, + { + "epoch": 0.27194244049524, + "grad_norm": 0.15115714073181152, + "learning_rate": 4.563066245096601e-05, + "loss": 3.2175, + "step": 8816 + }, + { + "epoch": 0.2719732869608134, + "grad_norm": 0.14661645889282227, + "learning_rate": 4.5629141922802815e-05, + "loss": 3.199, + "step": 8817 + }, + { + "epoch": 0.27200413342638685, + "grad_norm": 0.15182141959667206, + "learning_rate": 4.5627621155457116e-05, + "loss": 3.1975, + "step": 8818 + }, + { + "epoch": 0.27203497989196024, + "grad_norm": 0.14821632206439972, + "learning_rate": 4.562610014894655e-05, + "loss": 3.2099, + "step": 8819 + }, + { + "epoch": 0.2720658263575337, + "grad_norm": 0.14420603215694427, + "learning_rate": 4.562457890328874e-05, + "loss": 3.1989, + "step": 8820 + }, + { + "epoch": 0.27209667282310707, + "grad_norm": 0.15237806737422943, + "learning_rate": 4.5623057418501337e-05, + "loss": 3.1616, + "step": 8821 + }, + { + "epoch": 0.2721275192886805, + "grad_norm": 0.14605027437210083, + "learning_rate": 4.562153569460197e-05, + "loss": 3.2229, + "step": 8822 + }, + { + "epoch": 0.2721583657542539, + "grad_norm": 0.1627635806798935, + "learning_rate": 4.562001373160829e-05, + "loss": 3.1792, + "step": 8823 + }, + { + "epoch": 0.27218921221982734, + "grad_norm": 0.15509256720542908, + "learning_rate": 4.561849152953795e-05, + "loss": 3.2134, + "step": 8824 + }, + { + "epoch": 0.2722200586854008, + "grad_norm": 0.14807020127773285, + "learning_rate": 4.5616969088408585e-05, + "loss": 3.1981, + "step": 8825 + }, + { + "epoch": 0.27225090515097417, + "grad_norm": 0.15113697946071625, + "learning_rate": 4.561544640823785e-05, + "loss": 3.1877, + "step": 8826 + }, + { + "epoch": 0.2722817516165476, + "grad_norm": 0.15351267158985138, + "learning_rate": 4.561392348904341e-05, + "loss": 3.2142, + "step": 8827 + }, + { + "epoch": 0.272312598082121, + "grad_norm": 0.15112130343914032, + "learning_rate": 4.561240033084291e-05, + "loss": 3.205, + "step": 8828 + }, + { + "epoch": 0.27234344454769444, + "grad_norm": 0.1507541537284851, + "learning_rate": 4.5610876933654014e-05, + "loss": 3.1808, + "step": 8829 + }, + { + "epoch": 0.2723742910132678, + "grad_norm": 0.14211219549179077, + "learning_rate": 4.5609353297494385e-05, + "loss": 3.1862, + "step": 8830 + }, + { + "epoch": 0.27240513747884126, + "grad_norm": 0.14725437760353088, + "learning_rate": 4.560782942238169e-05, + "loss": 3.1995, + "step": 8831 + }, + { + "epoch": 0.27243598394441465, + "grad_norm": 0.16525299847126007, + "learning_rate": 4.5606305308333604e-05, + "loss": 3.1916, + "step": 8832 + }, + { + "epoch": 0.2724668304099881, + "grad_norm": 0.14378894865512848, + "learning_rate": 4.560478095536778e-05, + "loss": 3.2092, + "step": 8833 + }, + { + "epoch": 0.2724976768755615, + "grad_norm": 0.14821162819862366, + "learning_rate": 4.560325636350191e-05, + "loss": 3.2302, + "step": 8834 + }, + { + "epoch": 0.2725285233411349, + "grad_norm": 0.17014120519161224, + "learning_rate": 4.560173153275366e-05, + "loss": 3.2269, + "step": 8835 + }, + { + "epoch": 0.27255936980670836, + "grad_norm": 0.16017086803913116, + "learning_rate": 4.5600206463140716e-05, + "loss": 3.2364, + "step": 8836 + }, + { + "epoch": 0.27259021627228175, + "grad_norm": 0.15308301150798798, + "learning_rate": 4.5598681154680755e-05, + "loss": 3.1615, + "step": 8837 + }, + { + "epoch": 0.2726210627378552, + "grad_norm": 0.22953753173351288, + "learning_rate": 4.5597155607391465e-05, + "loss": 3.2116, + "step": 8838 + }, + { + "epoch": 0.2726519092034286, + "grad_norm": 0.15038461983203888, + "learning_rate": 4.559562982129053e-05, + "loss": 3.2224, + "step": 8839 + }, + { + "epoch": 0.272682755669002, + "grad_norm": 1.602156400680542, + "learning_rate": 4.559410379639565e-05, + "loss": 3.1996, + "step": 8840 + }, + { + "epoch": 0.2727136021345754, + "grad_norm": 0.17283578217029572, + "learning_rate": 4.559257753272451e-05, + "loss": 3.2211, + "step": 8841 + }, + { + "epoch": 0.27274444860014885, + "grad_norm": 0.17634183168411255, + "learning_rate": 4.559105103029481e-05, + "loss": 3.1995, + "step": 8842 + }, + { + "epoch": 0.27277529506572223, + "grad_norm": 0.1797764003276825, + "learning_rate": 4.558952428912424e-05, + "loss": 3.2027, + "step": 8843 + }, + { + "epoch": 0.2728061415312957, + "grad_norm": 0.16126057505607605, + "learning_rate": 4.5587997309230516e-05, + "loss": 3.1936, + "step": 8844 + }, + { + "epoch": 0.27283698799686906, + "grad_norm": 0.18033991754055023, + "learning_rate": 4.5586470090631325e-05, + "loss": 3.227, + "step": 8845 + }, + { + "epoch": 0.2728678344624425, + "grad_norm": 0.1548478752374649, + "learning_rate": 4.5584942633344396e-05, + "loss": 3.1695, + "step": 8846 + }, + { + "epoch": 0.2728986809280159, + "grad_norm": 0.17416569590568542, + "learning_rate": 4.558341493738743e-05, + "loss": 3.2086, + "step": 8847 + }, + { + "epoch": 0.27292952739358933, + "grad_norm": 0.15314799547195435, + "learning_rate": 4.558188700277812e-05, + "loss": 3.1922, + "step": 8848 + }, + { + "epoch": 0.2729603738591628, + "grad_norm": 0.17969632148742676, + "learning_rate": 4.558035882953421e-05, + "loss": 3.1736, + "step": 8849 + }, + { + "epoch": 0.27299122032473616, + "grad_norm": 0.18871545791625977, + "learning_rate": 4.557883041767341e-05, + "loss": 3.1841, + "step": 8850 + }, + { + "epoch": 0.2730220667903096, + "grad_norm": 0.18068023025989532, + "learning_rate": 4.557730176721343e-05, + "loss": 3.1891, + "step": 8851 + }, + { + "epoch": 0.273052913255883, + "grad_norm": 0.17490136623382568, + "learning_rate": 4.557577287817201e-05, + "loss": 3.2038, + "step": 8852 + }, + { + "epoch": 0.27308375972145643, + "grad_norm": 0.1634497195482254, + "learning_rate": 4.557424375056686e-05, + "loss": 3.2155, + "step": 8853 + }, + { + "epoch": 0.2731146061870298, + "grad_norm": 0.16129016876220703, + "learning_rate": 4.5572714384415715e-05, + "loss": 3.2169, + "step": 8854 + }, + { + "epoch": 0.27314545265260326, + "grad_norm": 0.16044676303863525, + "learning_rate": 4.557118477973632e-05, + "loss": 3.2209, + "step": 8855 + }, + { + "epoch": 0.27317629911817665, + "grad_norm": 0.17346449196338654, + "learning_rate": 4.556965493654639e-05, + "loss": 3.1861, + "step": 8856 + }, + { + "epoch": 0.2732071455837501, + "grad_norm": 0.167439267039299, + "learning_rate": 4.556812485486367e-05, + "loss": 3.171, + "step": 8857 + }, + { + "epoch": 0.2732379920493235, + "grad_norm": 0.19953082501888275, + "learning_rate": 4.5566594534705915e-05, + "loss": 3.1953, + "step": 8858 + }, + { + "epoch": 0.2732688385148969, + "grad_norm": 0.15907645225524902, + "learning_rate": 4.556506397609085e-05, + "loss": 3.2063, + "step": 8859 + }, + { + "epoch": 0.27329968498047036, + "grad_norm": 0.15834841132164001, + "learning_rate": 4.5563533179036224e-05, + "loss": 3.185, + "step": 8860 + }, + { + "epoch": 0.27333053144604375, + "grad_norm": 0.15420392155647278, + "learning_rate": 4.556200214355979e-05, + "loss": 3.1992, + "step": 8861 + }, + { + "epoch": 0.2733613779116172, + "grad_norm": 0.2091778963804245, + "learning_rate": 4.55604708696793e-05, + "loss": 3.1795, + "step": 8862 + }, + { + "epoch": 0.2733922243771906, + "grad_norm": 0.1699969619512558, + "learning_rate": 4.555893935741249e-05, + "loss": 3.1754, + "step": 8863 + }, + { + "epoch": 0.273423070842764, + "grad_norm": 0.15127956867218018, + "learning_rate": 4.5557407606777145e-05, + "loss": 3.1977, + "step": 8864 + }, + { + "epoch": 0.2734539173083374, + "grad_norm": 0.24374724924564362, + "learning_rate": 4.5555875617791016e-05, + "loss": 3.1817, + "step": 8865 + }, + { + "epoch": 0.27348476377391084, + "grad_norm": 0.16202133893966675, + "learning_rate": 4.555434339047186e-05, + "loss": 3.1802, + "step": 8866 + }, + { + "epoch": 0.27351561023948423, + "grad_norm": 0.17287424206733704, + "learning_rate": 4.555281092483744e-05, + "loss": 3.2009, + "step": 8867 + }, + { + "epoch": 0.2735464567050577, + "grad_norm": 0.16178366541862488, + "learning_rate": 4.555127822090553e-05, + "loss": 3.1902, + "step": 8868 + }, + { + "epoch": 0.27357730317063106, + "grad_norm": 0.16422684490680695, + "learning_rate": 4.55497452786939e-05, + "loss": 3.2377, + "step": 8869 + }, + { + "epoch": 0.2736081496362045, + "grad_norm": 0.1591896265745163, + "learning_rate": 4.554821209822032e-05, + "loss": 3.2009, + "step": 8870 + }, + { + "epoch": 0.2736389961017779, + "grad_norm": 0.1659175008535385, + "learning_rate": 4.554667867950257e-05, + "loss": 3.2137, + "step": 8871 + }, + { + "epoch": 0.27366984256735133, + "grad_norm": 0.156519815325737, + "learning_rate": 4.554514502255843e-05, + "loss": 3.2048, + "step": 8872 + }, + { + "epoch": 0.27370068903292477, + "grad_norm": 0.17127496004104614, + "learning_rate": 4.5543611127405674e-05, + "loss": 3.1678, + "step": 8873 + }, + { + "epoch": 0.27373153549849816, + "grad_norm": 0.15337976813316345, + "learning_rate": 4.55420769940621e-05, + "loss": 3.178, + "step": 8874 + }, + { + "epoch": 0.2737623819640716, + "grad_norm": 0.1852007955312729, + "learning_rate": 4.5540542622545476e-05, + "loss": 3.2099, + "step": 8875 + }, + { + "epoch": 0.273793228429645, + "grad_norm": 0.15128308534622192, + "learning_rate": 4.553900801287362e-05, + "loss": 3.1842, + "step": 8876 + }, + { + "epoch": 0.27382407489521843, + "grad_norm": 0.15490582585334778, + "learning_rate": 4.5537473165064294e-05, + "loss": 3.189, + "step": 8877 + }, + { + "epoch": 0.2738549213607918, + "grad_norm": 0.15576566755771637, + "learning_rate": 4.5535938079135316e-05, + "loss": 3.192, + "step": 8878 + }, + { + "epoch": 0.27388576782636526, + "grad_norm": 0.15510623157024384, + "learning_rate": 4.5534402755104476e-05, + "loss": 3.169, + "step": 8879 + }, + { + "epoch": 0.27391661429193864, + "grad_norm": 0.15967918932437897, + "learning_rate": 4.5532867192989574e-05, + "loss": 3.2318, + "step": 8880 + }, + { + "epoch": 0.2739474607575121, + "grad_norm": 0.15361616015434265, + "learning_rate": 4.553133139280842e-05, + "loss": 3.1939, + "step": 8881 + }, + { + "epoch": 0.27397830722308547, + "grad_norm": 0.15254418551921844, + "learning_rate": 4.5529795354578805e-05, + "loss": 3.1851, + "step": 8882 + }, + { + "epoch": 0.2740091536886589, + "grad_norm": 0.1512005627155304, + "learning_rate": 4.552825907831856e-05, + "loss": 3.2137, + "step": 8883 + }, + { + "epoch": 0.27404000015423235, + "grad_norm": 0.16337519884109497, + "learning_rate": 4.5526722564045486e-05, + "loss": 3.2187, + "step": 8884 + }, + { + "epoch": 0.27407084661980574, + "grad_norm": 0.1568254977464676, + "learning_rate": 4.55251858117774e-05, + "loss": 3.1748, + "step": 8885 + }, + { + "epoch": 0.2741016930853792, + "grad_norm": 0.1480860412120819, + "learning_rate": 4.5523648821532116e-05, + "loss": 3.2237, + "step": 8886 + }, + { + "epoch": 0.27413253955095257, + "grad_norm": 0.15014934539794922, + "learning_rate": 4.5522111593327456e-05, + "loss": 3.1628, + "step": 8887 + }, + { + "epoch": 0.274163386016526, + "grad_norm": 0.23566223680973053, + "learning_rate": 4.5520574127181246e-05, + "loss": 3.1836, + "step": 8888 + }, + { + "epoch": 0.2741942324820994, + "grad_norm": 0.169295534491539, + "learning_rate": 4.551903642311131e-05, + "loss": 3.2244, + "step": 8889 + }, + { + "epoch": 0.27422507894767284, + "grad_norm": 0.14810916781425476, + "learning_rate": 4.551749848113548e-05, + "loss": 3.1796, + "step": 8890 + }, + { + "epoch": 0.2742559254132462, + "grad_norm": 0.1558329164981842, + "learning_rate": 4.5515960301271574e-05, + "loss": 3.1891, + "step": 8891 + }, + { + "epoch": 0.27428677187881967, + "grad_norm": 0.1568170040845871, + "learning_rate": 4.551442188353745e-05, + "loss": 3.2199, + "step": 8892 + }, + { + "epoch": 0.27431761834439305, + "grad_norm": 0.15808439254760742, + "learning_rate": 4.551288322795093e-05, + "loss": 3.1946, + "step": 8893 + }, + { + "epoch": 0.2743484648099665, + "grad_norm": 0.14907607436180115, + "learning_rate": 4.551134433452985e-05, + "loss": 3.2247, + "step": 8894 + }, + { + "epoch": 0.2743793112755399, + "grad_norm": 0.15144862234592438, + "learning_rate": 4.5509805203292064e-05, + "loss": 3.222, + "step": 8895 + }, + { + "epoch": 0.2744101577411133, + "grad_norm": 0.17847351729869843, + "learning_rate": 4.5508265834255405e-05, + "loss": 3.1928, + "step": 8896 + }, + { + "epoch": 0.27444100420668677, + "grad_norm": 0.145828515291214, + "learning_rate": 4.550672622743773e-05, + "loss": 3.2109, + "step": 8897 + }, + { + "epoch": 0.27447185067226015, + "grad_norm": 0.16096359491348267, + "learning_rate": 4.550518638285689e-05, + "loss": 3.2311, + "step": 8898 + }, + { + "epoch": 0.2745026971378336, + "grad_norm": 0.15301723778247833, + "learning_rate": 4.5503646300530734e-05, + "loss": 3.2125, + "step": 8899 + }, + { + "epoch": 0.274533543603407, + "grad_norm": 0.17200948297977448, + "learning_rate": 4.550210598047712e-05, + "loss": 3.1928, + "step": 8900 + }, + { + "epoch": 0.2745643900689804, + "grad_norm": 0.15366630256175995, + "learning_rate": 4.5500565422713914e-05, + "loss": 3.1775, + "step": 8901 + }, + { + "epoch": 0.2745952365345538, + "grad_norm": 0.15741519629955292, + "learning_rate": 4.5499024627258965e-05, + "loss": 3.2137, + "step": 8902 + }, + { + "epoch": 0.27462608300012725, + "grad_norm": 0.1657698154449463, + "learning_rate": 4.549748359413014e-05, + "loss": 3.1931, + "step": 8903 + }, + { + "epoch": 0.27465692946570064, + "grad_norm": 0.15590567886829376, + "learning_rate": 4.549594232334532e-05, + "loss": 3.1925, + "step": 8904 + }, + { + "epoch": 0.2746877759312741, + "grad_norm": 0.14672890305519104, + "learning_rate": 4.549440081492236e-05, + "loss": 3.1732, + "step": 8905 + }, + { + "epoch": 0.27471862239684747, + "grad_norm": 0.15777426958084106, + "learning_rate": 4.549285906887914e-05, + "loss": 3.1903, + "step": 8906 + }, + { + "epoch": 0.2747494688624209, + "grad_norm": 0.14682245254516602, + "learning_rate": 4.549131708523353e-05, + "loss": 3.1955, + "step": 8907 + }, + { + "epoch": 0.27478031532799435, + "grad_norm": 0.15751329064369202, + "learning_rate": 4.5489774864003423e-05, + "loss": 3.2314, + "step": 8908 + }, + { + "epoch": 0.27481116179356774, + "grad_norm": 0.1531192511320114, + "learning_rate": 4.5488232405206674e-05, + "loss": 3.2053, + "step": 8909 + }, + { + "epoch": 0.2748420082591412, + "grad_norm": 0.15544293820858002, + "learning_rate": 4.5486689708861195e-05, + "loss": 3.2372, + "step": 8910 + }, + { + "epoch": 0.27487285472471457, + "grad_norm": 0.16113176941871643, + "learning_rate": 4.5485146774984855e-05, + "loss": 3.2139, + "step": 8911 + }, + { + "epoch": 0.274903701190288, + "grad_norm": 0.1490170955657959, + "learning_rate": 4.548360360359555e-05, + "loss": 3.1994, + "step": 8912 + }, + { + "epoch": 0.2749345476558614, + "grad_norm": 0.1594710648059845, + "learning_rate": 4.5482060194711163e-05, + "loss": 3.2551, + "step": 8913 + }, + { + "epoch": 0.27496539412143484, + "grad_norm": 0.15006648004055023, + "learning_rate": 4.5480516548349604e-05, + "loss": 3.1828, + "step": 8914 + }, + { + "epoch": 0.2749962405870082, + "grad_norm": 0.15724240243434906, + "learning_rate": 4.547897266452876e-05, + "loss": 3.2466, + "step": 8915 + }, + { + "epoch": 0.27502708705258166, + "grad_norm": 0.16372907161712646, + "learning_rate": 4.547742854326654e-05, + "loss": 3.2164, + "step": 8916 + }, + { + "epoch": 0.27505793351815505, + "grad_norm": 0.16065599024295807, + "learning_rate": 4.547588418458083e-05, + "loss": 3.2099, + "step": 8917 + }, + { + "epoch": 0.2750887799837285, + "grad_norm": 0.15675999224185944, + "learning_rate": 4.5474339588489556e-05, + "loss": 3.1869, + "step": 8918 + }, + { + "epoch": 0.2751196264493019, + "grad_norm": 0.15819783508777618, + "learning_rate": 4.547279475501062e-05, + "loss": 3.2212, + "step": 8919 + }, + { + "epoch": 0.2751504729148753, + "grad_norm": 0.16577032208442688, + "learning_rate": 4.547124968416192e-05, + "loss": 3.1508, + "step": 8920 + }, + { + "epoch": 0.27518131938044876, + "grad_norm": 0.1594056934118271, + "learning_rate": 4.5469704375961395e-05, + "loss": 3.1841, + "step": 8921 + }, + { + "epoch": 0.27521216584602215, + "grad_norm": 0.15961000323295593, + "learning_rate": 4.5468158830426946e-05, + "loss": 3.205, + "step": 8922 + }, + { + "epoch": 0.2752430123115956, + "grad_norm": 0.154756560921669, + "learning_rate": 4.546661304757649e-05, + "loss": 3.2085, + "step": 8923 + }, + { + "epoch": 0.275273858777169, + "grad_norm": 0.15455256402492523, + "learning_rate": 4.546506702742795e-05, + "loss": 3.2106, + "step": 8924 + }, + { + "epoch": 0.2753047052427424, + "grad_norm": 0.17006626725196838, + "learning_rate": 4.546352076999927e-05, + "loss": 3.2314, + "step": 8925 + }, + { + "epoch": 0.2753355517083158, + "grad_norm": 0.15239839255809784, + "learning_rate": 4.5461974275308356e-05, + "loss": 3.1738, + "step": 8926 + }, + { + "epoch": 0.27536639817388925, + "grad_norm": 0.15588738024234772, + "learning_rate": 4.546042754337315e-05, + "loss": 3.2205, + "step": 8927 + }, + { + "epoch": 0.27539724463946263, + "grad_norm": 0.1667526811361313, + "learning_rate": 4.5458880574211575e-05, + "loss": 3.1902, + "step": 8928 + }, + { + "epoch": 0.2754280911050361, + "grad_norm": 0.1618822067975998, + "learning_rate": 4.545733336784158e-05, + "loss": 3.2089, + "step": 8929 + }, + { + "epoch": 0.27545893757060946, + "grad_norm": 0.15084867179393768, + "learning_rate": 4.54557859242811e-05, + "loss": 3.1568, + "step": 8930 + }, + { + "epoch": 0.2754897840361829, + "grad_norm": 0.16106411814689636, + "learning_rate": 4.545423824354806e-05, + "loss": 3.2074, + "step": 8931 + }, + { + "epoch": 0.27552063050175635, + "grad_norm": 0.16936008632183075, + "learning_rate": 4.545269032566044e-05, + "loss": 3.2033, + "step": 8932 + }, + { + "epoch": 0.27555147696732973, + "grad_norm": 0.16299208998680115, + "learning_rate": 4.545114217063616e-05, + "loss": 3.2112, + "step": 8933 + }, + { + "epoch": 0.2755823234329032, + "grad_norm": 0.15601949393749237, + "learning_rate": 4.544959377849317e-05, + "loss": 3.2133, + "step": 8934 + }, + { + "epoch": 0.27561316989847656, + "grad_norm": 0.15890197455883026, + "learning_rate": 4.544804514924943e-05, + "loss": 3.2168, + "step": 8935 + }, + { + "epoch": 0.27564401636405, + "grad_norm": 0.15576112270355225, + "learning_rate": 4.54464962829229e-05, + "loss": 3.1928, + "step": 8936 + }, + { + "epoch": 0.2756748628296234, + "grad_norm": 0.15877988934516907, + "learning_rate": 4.5444947179531524e-05, + "loss": 3.1736, + "step": 8937 + }, + { + "epoch": 0.27570570929519683, + "grad_norm": 0.1465292125940323, + "learning_rate": 4.544339783909328e-05, + "loss": 3.1772, + "step": 8938 + }, + { + "epoch": 0.2757365557607702, + "grad_norm": 0.1589493751525879, + "learning_rate": 4.5441848261626116e-05, + "loss": 3.1781, + "step": 8939 + }, + { + "epoch": 0.27576740222634366, + "grad_norm": 0.15900661051273346, + "learning_rate": 4.544029844714801e-05, + "loss": 3.2152, + "step": 8940 + }, + { + "epoch": 0.27579824869191705, + "grad_norm": 0.1820458471775055, + "learning_rate": 4.543874839567692e-05, + "loss": 3.2062, + "step": 8941 + }, + { + "epoch": 0.2758290951574905, + "grad_norm": 0.1485781967639923, + "learning_rate": 4.5437198107230834e-05, + "loss": 3.2025, + "step": 8942 + }, + { + "epoch": 0.2758599416230639, + "grad_norm": 0.15455611050128937, + "learning_rate": 4.543564758182771e-05, + "loss": 3.1807, + "step": 8943 + }, + { + "epoch": 0.2758907880886373, + "grad_norm": 0.16227193176746368, + "learning_rate": 4.543409681948554e-05, + "loss": 3.2012, + "step": 8944 + }, + { + "epoch": 0.27592163455421076, + "grad_norm": 0.14893808960914612, + "learning_rate": 4.543254582022229e-05, + "loss": 3.2107, + "step": 8945 + }, + { + "epoch": 0.27595248101978415, + "grad_norm": 0.1600847691297531, + "learning_rate": 4.543099458405595e-05, + "loss": 3.2124, + "step": 8946 + }, + { + "epoch": 0.2759833274853576, + "grad_norm": 0.14221185445785522, + "learning_rate": 4.542944311100451e-05, + "loss": 3.1897, + "step": 8947 + }, + { + "epoch": 0.276014173950931, + "grad_norm": 0.1567074954509735, + "learning_rate": 4.542789140108595e-05, + "loss": 3.1963, + "step": 8948 + }, + { + "epoch": 0.2760450204165044, + "grad_norm": 0.14692729711532593, + "learning_rate": 4.542633945431827e-05, + "loss": 3.1632, + "step": 8949 + }, + { + "epoch": 0.2760758668820778, + "grad_norm": 0.147823765873909, + "learning_rate": 4.542478727071946e-05, + "loss": 3.2002, + "step": 8950 + }, + { + "epoch": 0.27610671334765124, + "grad_norm": 0.1500146985054016, + "learning_rate": 4.5423234850307504e-05, + "loss": 3.2095, + "step": 8951 + }, + { + "epoch": 0.27613755981322463, + "grad_norm": 0.1482539027929306, + "learning_rate": 4.542168219310042e-05, + "loss": 3.2191, + "step": 8952 + }, + { + "epoch": 0.2761684062787981, + "grad_norm": 0.15058672428131104, + "learning_rate": 4.54201292991162e-05, + "loss": 3.1889, + "step": 8953 + }, + { + "epoch": 0.27619925274437146, + "grad_norm": 0.1609315574169159, + "learning_rate": 4.5418576168372864e-05, + "loss": 3.2234, + "step": 8954 + }, + { + "epoch": 0.2762300992099449, + "grad_norm": 0.1529136747121811, + "learning_rate": 4.54170228008884e-05, + "loss": 3.1962, + "step": 8955 + }, + { + "epoch": 0.27626094567551834, + "grad_norm": 0.1539437472820282, + "learning_rate": 4.5415469196680814e-05, + "loss": 3.1831, + "step": 8956 + }, + { + "epoch": 0.27629179214109173, + "grad_norm": 0.14843972027301788, + "learning_rate": 4.5413915355768146e-05, + "loss": 3.2377, + "step": 8957 + }, + { + "epoch": 0.27632263860666517, + "grad_norm": 0.1473524272441864, + "learning_rate": 4.541236127816839e-05, + "loss": 3.2422, + "step": 8958 + }, + { + "epoch": 0.27635348507223856, + "grad_norm": 0.1930496245622635, + "learning_rate": 4.541080696389958e-05, + "loss": 3.1994, + "step": 8959 + }, + { + "epoch": 0.276384331537812, + "grad_norm": 0.14927922189235687, + "learning_rate": 4.540925241297972e-05, + "loss": 3.1612, + "step": 8960 + }, + { + "epoch": 0.2764151780033854, + "grad_norm": 0.14999857544898987, + "learning_rate": 4.5407697625426835e-05, + "loss": 3.1869, + "step": 8961 + }, + { + "epoch": 0.27644602446895883, + "grad_norm": 0.1513872742652893, + "learning_rate": 4.540614260125897e-05, + "loss": 3.1716, + "step": 8962 + }, + { + "epoch": 0.2764768709345322, + "grad_norm": 0.1513299196958542, + "learning_rate": 4.540458734049415e-05, + "loss": 3.2125, + "step": 8963 + }, + { + "epoch": 0.27650771740010566, + "grad_norm": 0.1532604694366455, + "learning_rate": 4.540303184315039e-05, + "loss": 3.2183, + "step": 8964 + }, + { + "epoch": 0.27653856386567904, + "grad_norm": 0.15080633759498596, + "learning_rate": 4.5401476109245746e-05, + "loss": 3.2164, + "step": 8965 + }, + { + "epoch": 0.2765694103312525, + "grad_norm": 0.1501089185476303, + "learning_rate": 4.539992013879825e-05, + "loss": 3.2044, + "step": 8966 + }, + { + "epoch": 0.27660025679682587, + "grad_norm": 0.15132209658622742, + "learning_rate": 4.539836393182592e-05, + "loss": 3.2035, + "step": 8967 + }, + { + "epoch": 0.2766311032623993, + "grad_norm": 0.15688416361808777, + "learning_rate": 4.5396807488346835e-05, + "loss": 3.207, + "step": 8968 + }, + { + "epoch": 0.27666194972797276, + "grad_norm": 0.14940989017486572, + "learning_rate": 4.539525080837901e-05, + "loss": 3.1741, + "step": 8969 + }, + { + "epoch": 0.27669279619354614, + "grad_norm": 0.1508972942829132, + "learning_rate": 4.539369389194052e-05, + "loss": 3.1961, + "step": 8970 + }, + { + "epoch": 0.2767236426591196, + "grad_norm": 0.14566971361637115, + "learning_rate": 4.53921367390494e-05, + "loss": 3.2208, + "step": 8971 + }, + { + "epoch": 0.27675448912469297, + "grad_norm": 0.15951259434223175, + "learning_rate": 4.539057934972371e-05, + "loss": 3.2108, + "step": 8972 + }, + { + "epoch": 0.2767853355902664, + "grad_norm": 0.15308082103729248, + "learning_rate": 4.538902172398151e-05, + "loss": 3.1971, + "step": 8973 + }, + { + "epoch": 0.2768161820558398, + "grad_norm": 0.15299367904663086, + "learning_rate": 4.538746386184085e-05, + "loss": 3.1946, + "step": 8974 + }, + { + "epoch": 0.27684702852141324, + "grad_norm": 0.17336201667785645, + "learning_rate": 4.5385905763319796e-05, + "loss": 3.1977, + "step": 8975 + }, + { + "epoch": 0.2768778749869866, + "grad_norm": 0.15281826257705688, + "learning_rate": 4.538434742843642e-05, + "loss": 3.2052, + "step": 8976 + }, + { + "epoch": 0.27690872145256007, + "grad_norm": 0.17441557347774506, + "learning_rate": 4.5382788857208783e-05, + "loss": 3.1962, + "step": 8977 + }, + { + "epoch": 0.27693956791813346, + "grad_norm": 0.16343268752098083, + "learning_rate": 4.5381230049654956e-05, + "loss": 3.2071, + "step": 8978 + }, + { + "epoch": 0.2769704143837069, + "grad_norm": 0.188322976231575, + "learning_rate": 4.537967100579302e-05, + "loss": 3.1919, + "step": 8979 + }, + { + "epoch": 0.27700126084928034, + "grad_norm": 0.16096127033233643, + "learning_rate": 4.537811172564104e-05, + "loss": 3.2035, + "step": 8980 + }, + { + "epoch": 0.2770321073148537, + "grad_norm": 0.15315064787864685, + "learning_rate": 4.53765522092171e-05, + "loss": 3.2057, + "step": 8981 + }, + { + "epoch": 0.27706295378042717, + "grad_norm": 0.15441463887691498, + "learning_rate": 4.537499245653928e-05, + "loss": 3.2067, + "step": 8982 + }, + { + "epoch": 0.27709380024600055, + "grad_norm": 0.16241098940372467, + "learning_rate": 4.5373432467625676e-05, + "loss": 3.1876, + "step": 8983 + }, + { + "epoch": 0.277124646711574, + "grad_norm": 0.16089129447937012, + "learning_rate": 4.5371872242494354e-05, + "loss": 3.2351, + "step": 8984 + }, + { + "epoch": 0.2771554931771474, + "grad_norm": 0.15253449976444244, + "learning_rate": 4.537031178116343e-05, + "loss": 3.218, + "step": 8985 + }, + { + "epoch": 0.2771863396427208, + "grad_norm": 0.15461651980876923, + "learning_rate": 4.5368751083650966e-05, + "loss": 3.1901, + "step": 8986 + }, + { + "epoch": 0.2772171861082942, + "grad_norm": 0.14664995670318604, + "learning_rate": 4.536719014997508e-05, + "loss": 3.1594, + "step": 8987 + }, + { + "epoch": 0.27724803257386765, + "grad_norm": 0.15330763161182404, + "learning_rate": 4.5365628980153866e-05, + "loss": 3.2134, + "step": 8988 + }, + { + "epoch": 0.27727887903944104, + "grad_norm": 0.15070964395999908, + "learning_rate": 4.536406757420542e-05, + "loss": 3.2049, + "step": 8989 + }, + { + "epoch": 0.2773097255050145, + "grad_norm": 0.1469525247812271, + "learning_rate": 4.5362505932147844e-05, + "loss": 3.2132, + "step": 8990 + }, + { + "epoch": 0.27734057197058787, + "grad_norm": 0.1657923460006714, + "learning_rate": 4.536094405399926e-05, + "loss": 3.2548, + "step": 8991 + }, + { + "epoch": 0.2773714184361613, + "grad_norm": 0.15019667148590088, + "learning_rate": 4.5359381939777755e-05, + "loss": 3.2213, + "step": 8992 + }, + { + "epoch": 0.27740226490173475, + "grad_norm": 0.20791023969650269, + "learning_rate": 4.5357819589501457e-05, + "loss": 3.1782, + "step": 8993 + }, + { + "epoch": 0.27743311136730814, + "grad_norm": 0.16427212953567505, + "learning_rate": 4.5356257003188476e-05, + "loss": 3.2073, + "step": 8994 + }, + { + "epoch": 0.2774639578328816, + "grad_norm": 0.16130203008651733, + "learning_rate": 4.5354694180856915e-05, + "loss": 3.1772, + "step": 8995 + }, + { + "epoch": 0.27749480429845497, + "grad_norm": 0.17893566191196442, + "learning_rate": 4.535313112252492e-05, + "loss": 3.2086, + "step": 8996 + }, + { + "epoch": 0.2775256507640284, + "grad_norm": 0.16451627016067505, + "learning_rate": 4.535156782821059e-05, + "loss": 3.1772, + "step": 8997 + }, + { + "epoch": 0.2775564972296018, + "grad_norm": 0.16345839202404022, + "learning_rate": 4.535000429793207e-05, + "loss": 3.1876, + "step": 8998 + }, + { + "epoch": 0.27758734369517524, + "grad_norm": 0.15129581093788147, + "learning_rate": 4.534844053170747e-05, + "loss": 3.1999, + "step": 8999 + }, + { + "epoch": 0.2776181901607486, + "grad_norm": 0.16938406229019165, + "learning_rate": 4.5346876529554935e-05, + "loss": 3.2093, + "step": 9000 + }, + { + "epoch": 0.27764903662632207, + "grad_norm": 0.14907024800777435, + "learning_rate": 4.53453122914926e-05, + "loss": 3.2256, + "step": 9001 + }, + { + "epoch": 0.27767988309189545, + "grad_norm": 0.16386070847511292, + "learning_rate": 4.534374781753858e-05, + "loss": 3.205, + "step": 9002 + }, + { + "epoch": 0.2777107295574689, + "grad_norm": 0.1565655916929245, + "learning_rate": 4.534218310771104e-05, + "loss": 3.1773, + "step": 9003 + }, + { + "epoch": 0.27774157602304234, + "grad_norm": 0.17230331897735596, + "learning_rate": 4.534061816202811e-05, + "loss": 3.2256, + "step": 9004 + }, + { + "epoch": 0.2777724224886157, + "grad_norm": 0.15718787908554077, + "learning_rate": 4.533905298050794e-05, + "loss": 3.2175, + "step": 9005 + }, + { + "epoch": 0.27780326895418916, + "grad_norm": 0.18553142249584198, + "learning_rate": 4.533748756316867e-05, + "loss": 3.1951, + "step": 9006 + }, + { + "epoch": 0.27783411541976255, + "grad_norm": 0.14815859496593475, + "learning_rate": 4.5335921910028446e-05, + "loss": 3.2098, + "step": 9007 + }, + { + "epoch": 0.277864961885336, + "grad_norm": 0.16027426719665527, + "learning_rate": 4.533435602110544e-05, + "loss": 3.1885, + "step": 9008 + }, + { + "epoch": 0.2778958083509094, + "grad_norm": 0.16323047876358032, + "learning_rate": 4.5332789896417784e-05, + "loss": 3.1969, + "step": 9009 + }, + { + "epoch": 0.2779266548164828, + "grad_norm": 0.1544392704963684, + "learning_rate": 4.533122353598365e-05, + "loss": 3.2237, + "step": 9010 + }, + { + "epoch": 0.2779575012820562, + "grad_norm": 0.149229496717453, + "learning_rate": 4.53296569398212e-05, + "loss": 3.1985, + "step": 9011 + }, + { + "epoch": 0.27798834774762965, + "grad_norm": 0.15454699099063873, + "learning_rate": 4.532809010794859e-05, + "loss": 3.1776, + "step": 9012 + }, + { + "epoch": 0.27801919421320304, + "grad_norm": 0.15016904473304749, + "learning_rate": 4.5326523040384e-05, + "loss": 3.2005, + "step": 9013 + }, + { + "epoch": 0.2780500406787765, + "grad_norm": 0.1573253870010376, + "learning_rate": 4.532495573714558e-05, + "loss": 3.2031, + "step": 9014 + }, + { + "epoch": 0.27808088714434986, + "grad_norm": 0.23725102841854095, + "learning_rate": 4.532338819825152e-05, + "loss": 3.1717, + "step": 9015 + }, + { + "epoch": 0.2781117336099233, + "grad_norm": 0.1674616038799286, + "learning_rate": 4.5321820423719983e-05, + "loss": 3.2094, + "step": 9016 + }, + { + "epoch": 0.27814258007549675, + "grad_norm": 0.15693628787994385, + "learning_rate": 4.532025241356915e-05, + "loss": 3.2078, + "step": 9017 + }, + { + "epoch": 0.27817342654107013, + "grad_norm": 0.15623535215854645, + "learning_rate": 4.53186841678172e-05, + "loss": 3.192, + "step": 9018 + }, + { + "epoch": 0.2782042730066436, + "grad_norm": 0.15441995859146118, + "learning_rate": 4.5317115686482326e-05, + "loss": 3.2147, + "step": 9019 + }, + { + "epoch": 0.27823511947221696, + "grad_norm": 0.1515919715166092, + "learning_rate": 4.531554696958269e-05, + "loss": 3.1727, + "step": 9020 + }, + { + "epoch": 0.2782659659377904, + "grad_norm": 0.17019051313400269, + "learning_rate": 4.5313978017136506e-05, + "loss": 3.1811, + "step": 9021 + }, + { + "epoch": 0.2782968124033638, + "grad_norm": 0.15632446110248566, + "learning_rate": 4.531240882916195e-05, + "loss": 3.224, + "step": 9022 + }, + { + "epoch": 0.27832765886893723, + "grad_norm": 0.15628884732723236, + "learning_rate": 4.531083940567723e-05, + "loss": 3.2106, + "step": 9023 + }, + { + "epoch": 0.2783585053345106, + "grad_norm": 0.15531609952449799, + "learning_rate": 4.530926974670052e-05, + "loss": 3.2445, + "step": 9024 + }, + { + "epoch": 0.27838935180008406, + "grad_norm": 0.1509828120470047, + "learning_rate": 4.530769985225003e-05, + "loss": 3.2036, + "step": 9025 + }, + { + "epoch": 0.27842019826565745, + "grad_norm": 0.15637905895709991, + "learning_rate": 4.5306129722343974e-05, + "loss": 3.1839, + "step": 9026 + }, + { + "epoch": 0.2784510447312309, + "grad_norm": 0.1510355919599533, + "learning_rate": 4.530455935700054e-05, + "loss": 3.2165, + "step": 9027 + }, + { + "epoch": 0.27848189119680433, + "grad_norm": 0.15414686501026154, + "learning_rate": 4.530298875623794e-05, + "loss": 3.2106, + "step": 9028 + }, + { + "epoch": 0.2785127376623777, + "grad_norm": 0.15434589982032776, + "learning_rate": 4.5301417920074393e-05, + "loss": 3.1821, + "step": 9029 + }, + { + "epoch": 0.27854358412795116, + "grad_norm": 0.15900246798992157, + "learning_rate": 4.5299846848528105e-05, + "loss": 3.2145, + "step": 9030 + }, + { + "epoch": 0.27857443059352455, + "grad_norm": 0.14973683655261993, + "learning_rate": 4.5298275541617286e-05, + "loss": 3.1913, + "step": 9031 + }, + { + "epoch": 0.278605277059098, + "grad_norm": 0.15711502730846405, + "learning_rate": 4.529670399936016e-05, + "loss": 3.2079, + "step": 9032 + }, + { + "epoch": 0.2786361235246714, + "grad_norm": 0.14671668410301208, + "learning_rate": 4.529513222177495e-05, + "loss": 3.205, + "step": 9033 + }, + { + "epoch": 0.2786669699902448, + "grad_norm": 0.16120846569538116, + "learning_rate": 4.529356020887988e-05, + "loss": 3.2001, + "step": 9034 + }, + { + "epoch": 0.2786978164558182, + "grad_norm": 0.15803538262844086, + "learning_rate": 4.529198796069317e-05, + "loss": 3.1763, + "step": 9035 + }, + { + "epoch": 0.27872866292139165, + "grad_norm": 0.15237291157245636, + "learning_rate": 4.529041547723306e-05, + "loss": 3.2457, + "step": 9036 + }, + { + "epoch": 0.27875950938696503, + "grad_norm": 0.1461673229932785, + "learning_rate": 4.528884275851777e-05, + "loss": 3.1944, + "step": 9037 + }, + { + "epoch": 0.2787903558525385, + "grad_norm": 0.15005435049533844, + "learning_rate": 4.528726980456554e-05, + "loss": 3.2024, + "step": 9038 + }, + { + "epoch": 0.27882120231811186, + "grad_norm": 0.15442773699760437, + "learning_rate": 4.528569661539461e-05, + "loss": 3.2126, + "step": 9039 + }, + { + "epoch": 0.2788520487836853, + "grad_norm": 0.15658500790596008, + "learning_rate": 4.528412319102322e-05, + "loss": 3.2176, + "step": 9040 + }, + { + "epoch": 0.27888289524925874, + "grad_norm": 0.15132518112659454, + "learning_rate": 4.528254953146961e-05, + "loss": 3.2218, + "step": 9041 + }, + { + "epoch": 0.27891374171483213, + "grad_norm": 0.14652837812900543, + "learning_rate": 4.528097563675202e-05, + "loss": 3.2086, + "step": 9042 + }, + { + "epoch": 0.2789445881804056, + "grad_norm": 0.1444806456565857, + "learning_rate": 4.527940150688871e-05, + "loss": 3.1679, + "step": 9043 + }, + { + "epoch": 0.27897543464597896, + "grad_norm": 0.16656248271465302, + "learning_rate": 4.527782714189793e-05, + "loss": 3.2363, + "step": 9044 + }, + { + "epoch": 0.2790062811115524, + "grad_norm": 0.15684069693088531, + "learning_rate": 4.5276252541797924e-05, + "loss": 3.2609, + "step": 9045 + }, + { + "epoch": 0.2790371275771258, + "grad_norm": 0.1467301845550537, + "learning_rate": 4.527467770660696e-05, + "loss": 3.2221, + "step": 9046 + }, + { + "epoch": 0.27906797404269923, + "grad_norm": 0.1523360162973404, + "learning_rate": 4.5273102636343284e-05, + "loss": 3.1927, + "step": 9047 + }, + { + "epoch": 0.2790988205082726, + "grad_norm": 0.1497986763715744, + "learning_rate": 4.5271527331025174e-05, + "loss": 3.2476, + "step": 9048 + }, + { + "epoch": 0.27912966697384606, + "grad_norm": 0.1541588455438614, + "learning_rate": 4.5269951790670885e-05, + "loss": 3.2087, + "step": 9049 + }, + { + "epoch": 0.27916051343941944, + "grad_norm": 0.14949633181095123, + "learning_rate": 4.526837601529867e-05, + "loss": 3.1901, + "step": 9050 + }, + { + "epoch": 0.2791913599049929, + "grad_norm": 0.15288865566253662, + "learning_rate": 4.5266800004926837e-05, + "loss": 3.2157, + "step": 9051 + }, + { + "epoch": 0.27922220637056633, + "grad_norm": 0.1508794128894806, + "learning_rate": 4.526522375957363e-05, + "loss": 3.2275, + "step": 9052 + }, + { + "epoch": 0.2792530528361397, + "grad_norm": 0.17540043592453003, + "learning_rate": 4.526364727925732e-05, + "loss": 3.1914, + "step": 9053 + }, + { + "epoch": 0.27928389930171316, + "grad_norm": 0.1651729792356491, + "learning_rate": 4.526207056399621e-05, + "loss": 3.2213, + "step": 9054 + }, + { + "epoch": 0.27931474576728654, + "grad_norm": 0.1507999449968338, + "learning_rate": 4.5260493613808563e-05, + "loss": 3.176, + "step": 9055 + }, + { + "epoch": 0.27934559223286, + "grad_norm": 0.16641384363174438, + "learning_rate": 4.5258916428712674e-05, + "loss": 3.2121, + "step": 9056 + }, + { + "epoch": 0.27937643869843337, + "grad_norm": 0.1507343351840973, + "learning_rate": 4.5257339008726816e-05, + "loss": 3.1923, + "step": 9057 + }, + { + "epoch": 0.2794072851640068, + "grad_norm": 0.15025384724140167, + "learning_rate": 4.5255761353869286e-05, + "loss": 3.1718, + "step": 9058 + }, + { + "epoch": 0.2794381316295802, + "grad_norm": 0.14908207952976227, + "learning_rate": 4.525418346415838e-05, + "loss": 3.191, + "step": 9059 + }, + { + "epoch": 0.27946897809515364, + "grad_norm": 0.15217360854148865, + "learning_rate": 4.5252605339612386e-05, + "loss": 3.1919, + "step": 9060 + }, + { + "epoch": 0.27949982456072703, + "grad_norm": 0.1506180614233017, + "learning_rate": 4.52510269802496e-05, + "loss": 3.2051, + "step": 9061 + }, + { + "epoch": 0.27953067102630047, + "grad_norm": 0.1444859802722931, + "learning_rate": 4.524944838608833e-05, + "loss": 3.1965, + "step": 9062 + }, + { + "epoch": 0.27956151749187386, + "grad_norm": 0.15504631400108337, + "learning_rate": 4.524786955714687e-05, + "loss": 3.2281, + "step": 9063 + }, + { + "epoch": 0.2795923639574473, + "grad_norm": 0.1469048410654068, + "learning_rate": 4.524629049344353e-05, + "loss": 3.202, + "step": 9064 + }, + { + "epoch": 0.27962321042302074, + "grad_norm": 0.1836327761411667, + "learning_rate": 4.524471119499663e-05, + "loss": 3.205, + "step": 9065 + }, + { + "epoch": 0.2796540568885941, + "grad_norm": 0.14949072897434235, + "learning_rate": 4.5243131661824455e-05, + "loss": 3.2017, + "step": 9066 + }, + { + "epoch": 0.27968490335416757, + "grad_norm": 0.15040229260921478, + "learning_rate": 4.524155189394534e-05, + "loss": 3.2033, + "step": 9067 + }, + { + "epoch": 0.27971574981974096, + "grad_norm": 0.1461566984653473, + "learning_rate": 4.523997189137759e-05, + "loss": 3.1668, + "step": 9068 + }, + { + "epoch": 0.2797465962853144, + "grad_norm": 0.1490936577320099, + "learning_rate": 4.5238391654139536e-05, + "loss": 3.1716, + "step": 9069 + }, + { + "epoch": 0.2797774427508878, + "grad_norm": 0.14666257798671722, + "learning_rate": 4.523681118224948e-05, + "loss": 3.1917, + "step": 9070 + }, + { + "epoch": 0.2798082892164612, + "grad_norm": 0.15333309769630432, + "learning_rate": 4.5235230475725774e-05, + "loss": 3.2314, + "step": 9071 + }, + { + "epoch": 0.2798391356820346, + "grad_norm": 0.1498071253299713, + "learning_rate": 4.523364953458672e-05, + "loss": 3.2076, + "step": 9072 + }, + { + "epoch": 0.27986998214760805, + "grad_norm": 0.15614114701747894, + "learning_rate": 4.523206835885067e-05, + "loss": 3.1864, + "step": 9073 + }, + { + "epoch": 0.27990082861318144, + "grad_norm": 0.1469532698392868, + "learning_rate": 4.5230486948535936e-05, + "loss": 3.2135, + "step": 9074 + }, + { + "epoch": 0.2799316750787549, + "grad_norm": 0.15118548274040222, + "learning_rate": 4.522890530366086e-05, + "loss": 3.2352, + "step": 9075 + }, + { + "epoch": 0.2799625215443283, + "grad_norm": 0.15898297727108002, + "learning_rate": 4.522732342424379e-05, + "loss": 3.2204, + "step": 9076 + }, + { + "epoch": 0.2799933680099017, + "grad_norm": 0.16159872710704803, + "learning_rate": 4.522574131030306e-05, + "loss": 3.1975, + "step": 9077 + }, + { + "epoch": 0.28002421447547515, + "grad_norm": 0.16076016426086426, + "learning_rate": 4.522415896185701e-05, + "loss": 3.1673, + "step": 9078 + }, + { + "epoch": 0.28005506094104854, + "grad_norm": 0.16125863790512085, + "learning_rate": 4.5222576378923995e-05, + "loss": 3.21, + "step": 9079 + }, + { + "epoch": 0.280085907406622, + "grad_norm": 0.17094168066978455, + "learning_rate": 4.522099356152236e-05, + "loss": 3.2471, + "step": 9080 + }, + { + "epoch": 0.28011675387219537, + "grad_norm": 0.152724489569664, + "learning_rate": 4.521941050967045e-05, + "loss": 3.2288, + "step": 9081 + }, + { + "epoch": 0.2801476003377688, + "grad_norm": 0.14825496077537537, + "learning_rate": 4.521782722338663e-05, + "loss": 3.1931, + "step": 9082 + }, + { + "epoch": 0.2801784468033422, + "grad_norm": 0.15649771690368652, + "learning_rate": 4.521624370268925e-05, + "loss": 3.1992, + "step": 9083 + }, + { + "epoch": 0.28020929326891564, + "grad_norm": 0.14914172887802124, + "learning_rate": 4.521465994759667e-05, + "loss": 3.1757, + "step": 9084 + }, + { + "epoch": 0.280240139734489, + "grad_norm": 0.14795495569705963, + "learning_rate": 4.521307595812726e-05, + "loss": 3.1805, + "step": 9085 + }, + { + "epoch": 0.28027098620006247, + "grad_norm": 0.15759272873401642, + "learning_rate": 4.521149173429938e-05, + "loss": 3.1989, + "step": 9086 + }, + { + "epoch": 0.2803018326656359, + "grad_norm": 0.15612095594406128, + "learning_rate": 4.52099072761314e-05, + "loss": 3.2079, + "step": 9087 + }, + { + "epoch": 0.2803326791312093, + "grad_norm": 0.15855556726455688, + "learning_rate": 4.5208322583641695e-05, + "loss": 3.206, + "step": 9088 + }, + { + "epoch": 0.28036352559678274, + "grad_norm": 0.15692132711410522, + "learning_rate": 4.5206737656848616e-05, + "loss": 3.189, + "step": 9089 + }, + { + "epoch": 0.2803943720623561, + "grad_norm": 0.15383830666542053, + "learning_rate": 4.5205152495770574e-05, + "loss": 3.2008, + "step": 9090 + }, + { + "epoch": 0.28042521852792957, + "grad_norm": 0.16599592566490173, + "learning_rate": 4.5203567100425925e-05, + "loss": 3.2189, + "step": 9091 + }, + { + "epoch": 0.28045606499350295, + "grad_norm": 0.19931398332118988, + "learning_rate": 4.5201981470833055e-05, + "loss": 3.1975, + "step": 9092 + }, + { + "epoch": 0.2804869114590764, + "grad_norm": 0.15846051275730133, + "learning_rate": 4.5200395607010354e-05, + "loss": 3.1932, + "step": 9093 + }, + { + "epoch": 0.2805177579246498, + "grad_norm": 0.15502820909023285, + "learning_rate": 4.5198809508976206e-05, + "loss": 3.171, + "step": 9094 + }, + { + "epoch": 0.2805486043902232, + "grad_norm": 0.15591032803058624, + "learning_rate": 4.5197223176748994e-05, + "loss": 3.1781, + "step": 9095 + }, + { + "epoch": 0.2805794508557966, + "grad_norm": 0.15184153616428375, + "learning_rate": 4.519563661034712e-05, + "loss": 3.2023, + "step": 9096 + }, + { + "epoch": 0.28061029732137005, + "grad_norm": 0.15607154369354248, + "learning_rate": 4.519404980978897e-05, + "loss": 3.1951, + "step": 9097 + }, + { + "epoch": 0.28064114378694344, + "grad_norm": 0.1627911478281021, + "learning_rate": 4.5192462775092956e-05, + "loss": 3.1936, + "step": 9098 + }, + { + "epoch": 0.2806719902525169, + "grad_norm": 0.1451161503791809, + "learning_rate": 4.519087550627746e-05, + "loss": 3.2087, + "step": 9099 + }, + { + "epoch": 0.2807028367180903, + "grad_norm": 0.15394984185695648, + "learning_rate": 4.51892880033609e-05, + "loss": 3.2328, + "step": 9100 + }, + { + "epoch": 0.2807336831836637, + "grad_norm": 0.15696585178375244, + "learning_rate": 4.518770026636169e-05, + "loss": 3.1954, + "step": 9101 + }, + { + "epoch": 0.28076452964923715, + "grad_norm": 0.15222837030887604, + "learning_rate": 4.5186112295298214e-05, + "loss": 3.1864, + "step": 9102 + }, + { + "epoch": 0.28079537611481054, + "grad_norm": 0.1518321931362152, + "learning_rate": 4.51845240901889e-05, + "loss": 3.1965, + "step": 9103 + }, + { + "epoch": 0.280826222580384, + "grad_norm": 0.14891241490840912, + "learning_rate": 4.5182935651052156e-05, + "loss": 3.2197, + "step": 9104 + }, + { + "epoch": 0.28085706904595736, + "grad_norm": 0.14864902198314667, + "learning_rate": 4.51813469779064e-05, + "loss": 3.2054, + "step": 9105 + }, + { + "epoch": 0.2808879155115308, + "grad_norm": 0.14873909950256348, + "learning_rate": 4.517975807077006e-05, + "loss": 3.1866, + "step": 9106 + }, + { + "epoch": 0.2809187619771042, + "grad_norm": 0.14492204785346985, + "learning_rate": 4.5178168929661546e-05, + "loss": 3.1814, + "step": 9107 + }, + { + "epoch": 0.28094960844267763, + "grad_norm": 0.14830511808395386, + "learning_rate": 4.5176579554599294e-05, + "loss": 3.1888, + "step": 9108 + }, + { + "epoch": 0.280980454908251, + "grad_norm": 0.1493295133113861, + "learning_rate": 4.517498994560172e-05, + "loss": 3.2046, + "step": 9109 + }, + { + "epoch": 0.28101130137382446, + "grad_norm": 0.1918649971485138, + "learning_rate": 4.517340010268727e-05, + "loss": 3.1862, + "step": 9110 + }, + { + "epoch": 0.2810421478393979, + "grad_norm": 0.1507672518491745, + "learning_rate": 4.5171810025874364e-05, + "loss": 3.2136, + "step": 9111 + }, + { + "epoch": 0.2810729943049713, + "grad_norm": 0.16488516330718994, + "learning_rate": 4.5170219715181436e-05, + "loss": 3.1643, + "step": 9112 + }, + { + "epoch": 0.28110384077054473, + "grad_norm": 0.15182048082351685, + "learning_rate": 4.516862917062694e-05, + "loss": 3.1818, + "step": 9113 + }, + { + "epoch": 0.2811346872361181, + "grad_norm": 0.15096665918827057, + "learning_rate": 4.516703839222931e-05, + "loss": 3.1929, + "step": 9114 + }, + { + "epoch": 0.28116553370169156, + "grad_norm": 0.15018503367900848, + "learning_rate": 4.516544738000699e-05, + "loss": 3.191, + "step": 9115 + }, + { + "epoch": 0.28119638016726495, + "grad_norm": 0.16335603594779968, + "learning_rate": 4.516385613397842e-05, + "loss": 3.1902, + "step": 9116 + }, + { + "epoch": 0.2812272266328384, + "grad_norm": 0.14952340722084045, + "learning_rate": 4.5162264654162056e-05, + "loss": 3.2029, + "step": 9117 + }, + { + "epoch": 0.2812580730984118, + "grad_norm": 0.1533374786376953, + "learning_rate": 4.516067294057635e-05, + "loss": 3.1889, + "step": 9118 + }, + { + "epoch": 0.2812889195639852, + "grad_norm": 0.3863016366958618, + "learning_rate": 4.5159080993239756e-05, + "loss": 3.1883, + "step": 9119 + }, + { + "epoch": 0.2813197660295586, + "grad_norm": 0.1499297171831131, + "learning_rate": 4.5157488812170734e-05, + "loss": 3.2128, + "step": 9120 + }, + { + "epoch": 0.28135061249513205, + "grad_norm": 0.15110936760902405, + "learning_rate": 4.515589639738774e-05, + "loss": 3.2224, + "step": 9121 + }, + { + "epoch": 0.28138145896070543, + "grad_norm": 0.15038910508155823, + "learning_rate": 4.515430374890924e-05, + "loss": 3.2228, + "step": 9122 + }, + { + "epoch": 0.2814123054262789, + "grad_norm": 0.151884526014328, + "learning_rate": 4.5152710866753703e-05, + "loss": 3.1869, + "step": 9123 + }, + { + "epoch": 0.2814431518918523, + "grad_norm": 0.14619320631027222, + "learning_rate": 4.515111775093959e-05, + "loss": 3.1836, + "step": 9124 + }, + { + "epoch": 0.2814739983574257, + "grad_norm": 0.15143509209156036, + "learning_rate": 4.5149524401485375e-05, + "loss": 3.2302, + "step": 9125 + }, + { + "epoch": 0.28150484482299915, + "grad_norm": 0.1450996845960617, + "learning_rate": 4.514793081840953e-05, + "loss": 3.184, + "step": 9126 + }, + { + "epoch": 0.28153569128857253, + "grad_norm": 0.15161268413066864, + "learning_rate": 4.514633700173054e-05, + "loss": 3.216, + "step": 9127 + }, + { + "epoch": 0.281566537754146, + "grad_norm": 0.15518254041671753, + "learning_rate": 4.514474295146688e-05, + "loss": 3.1754, + "step": 9128 + }, + { + "epoch": 0.28159738421971936, + "grad_norm": 0.14759819209575653, + "learning_rate": 4.514314866763703e-05, + "loss": 3.1984, + "step": 9129 + }, + { + "epoch": 0.2816282306852928, + "grad_norm": 0.1537170261144638, + "learning_rate": 4.514155415025948e-05, + "loss": 3.1751, + "step": 9130 + }, + { + "epoch": 0.2816590771508662, + "grad_norm": 0.14975108206272125, + "learning_rate": 4.51399593993527e-05, + "loss": 3.2138, + "step": 9131 + }, + { + "epoch": 0.28168992361643963, + "grad_norm": 0.14774321019649506, + "learning_rate": 4.5138364414935206e-05, + "loss": 3.1504, + "step": 9132 + }, + { + "epoch": 0.281720770082013, + "grad_norm": 0.17121058702468872, + "learning_rate": 4.5136769197025474e-05, + "loss": 3.198, + "step": 9133 + }, + { + "epoch": 0.28175161654758646, + "grad_norm": 0.1508641391992569, + "learning_rate": 4.5135173745642e-05, + "loss": 3.1834, + "step": 9134 + }, + { + "epoch": 0.2817824630131599, + "grad_norm": 0.1572738140821457, + "learning_rate": 4.5133578060803295e-05, + "loss": 3.1933, + "step": 9135 + }, + { + "epoch": 0.2818133094787333, + "grad_norm": 0.15857413411140442, + "learning_rate": 4.5131982142527854e-05, + "loss": 3.183, + "step": 9136 + }, + { + "epoch": 0.28184415594430673, + "grad_norm": 0.1512751430273056, + "learning_rate": 4.513038599083417e-05, + "loss": 3.1512, + "step": 9137 + }, + { + "epoch": 0.2818750024098801, + "grad_norm": 0.16079746186733246, + "learning_rate": 4.512878960574075e-05, + "loss": 3.1731, + "step": 9138 + }, + { + "epoch": 0.28190584887545356, + "grad_norm": 0.16001875698566437, + "learning_rate": 4.512719298726612e-05, + "loss": 3.1778, + "step": 9139 + }, + { + "epoch": 0.28193669534102694, + "grad_norm": 0.15773692727088928, + "learning_rate": 4.512559613542878e-05, + "loss": 3.2334, + "step": 9140 + }, + { + "epoch": 0.2819675418066004, + "grad_norm": 0.1656002402305603, + "learning_rate": 4.5123999050247256e-05, + "loss": 3.1713, + "step": 9141 + }, + { + "epoch": 0.28199838827217377, + "grad_norm": 0.1545095592737198, + "learning_rate": 4.512240173174004e-05, + "loss": 3.211, + "step": 9142 + }, + { + "epoch": 0.2820292347377472, + "grad_norm": 0.17617185413837433, + "learning_rate": 4.512080417992568e-05, + "loss": 3.2201, + "step": 9143 + }, + { + "epoch": 0.2820600812033206, + "grad_norm": 0.14932601153850555, + "learning_rate": 4.511920639482268e-05, + "loss": 3.1613, + "step": 9144 + }, + { + "epoch": 0.28209092766889404, + "grad_norm": 0.15129415690898895, + "learning_rate": 4.511760837644958e-05, + "loss": 3.2313, + "step": 9145 + }, + { + "epoch": 0.28212177413446743, + "grad_norm": 0.1532592922449112, + "learning_rate": 4.5116010124824894e-05, + "loss": 3.2142, + "step": 9146 + }, + { + "epoch": 0.28215262060004087, + "grad_norm": 0.5121219754219055, + "learning_rate": 4.511441163996716e-05, + "loss": 3.2214, + "step": 9147 + }, + { + "epoch": 0.2821834670656143, + "grad_norm": 0.15371523797512054, + "learning_rate": 4.511281292189491e-05, + "loss": 3.2129, + "step": 9148 + }, + { + "epoch": 0.2822143135311877, + "grad_norm": 0.15853434801101685, + "learning_rate": 4.511121397062668e-05, + "loss": 3.1809, + "step": 9149 + }, + { + "epoch": 0.28224515999676114, + "grad_norm": 0.15099561214447021, + "learning_rate": 4.510961478618101e-05, + "loss": 3.1892, + "step": 9150 + }, + { + "epoch": 0.2822760064623345, + "grad_norm": 0.15859264135360718, + "learning_rate": 4.510801536857644e-05, + "loss": 3.1932, + "step": 9151 + }, + { + "epoch": 0.28230685292790797, + "grad_norm": 0.15632013976573944, + "learning_rate": 4.5106415717831515e-05, + "loss": 3.1802, + "step": 9152 + }, + { + "epoch": 0.28233769939348136, + "grad_norm": 0.30256935954093933, + "learning_rate": 4.5104815833964784e-05, + "loss": 3.1707, + "step": 9153 + }, + { + "epoch": 0.2823685458590548, + "grad_norm": 0.1558988094329834, + "learning_rate": 4.51032157169948e-05, + "loss": 3.2275, + "step": 9154 + }, + { + "epoch": 0.2823993923246282, + "grad_norm": 0.15446847677230835, + "learning_rate": 4.510161536694011e-05, + "loss": 3.2037, + "step": 9155 + }, + { + "epoch": 0.2824302387902016, + "grad_norm": 0.1642707884311676, + "learning_rate": 4.510001478381926e-05, + "loss": 3.221, + "step": 9156 + }, + { + "epoch": 0.282461085255775, + "grad_norm": 0.15761393308639526, + "learning_rate": 4.509841396765083e-05, + "loss": 3.1745, + "step": 9157 + }, + { + "epoch": 0.28249193172134845, + "grad_norm": 0.15543846786022186, + "learning_rate": 4.509681291845336e-05, + "loss": 3.172, + "step": 9158 + }, + { + "epoch": 0.2825227781869219, + "grad_norm": 0.1602693796157837, + "learning_rate": 4.5095211636245415e-05, + "loss": 3.1811, + "step": 9159 + }, + { + "epoch": 0.2825536246524953, + "grad_norm": 0.16749756038188934, + "learning_rate": 4.5093610121045574e-05, + "loss": 3.2005, + "step": 9160 + }, + { + "epoch": 0.2825844711180687, + "grad_norm": 0.1858697533607483, + "learning_rate": 4.50920083728724e-05, + "loss": 3.247, + "step": 9161 + }, + { + "epoch": 0.2826153175836421, + "grad_norm": 0.15471121668815613, + "learning_rate": 4.509040639174446e-05, + "loss": 3.1883, + "step": 9162 + }, + { + "epoch": 0.28264616404921555, + "grad_norm": 0.1626942902803421, + "learning_rate": 4.508880417768033e-05, + "loss": 3.2095, + "step": 9163 + }, + { + "epoch": 0.28267701051478894, + "grad_norm": 0.1540507674217224, + "learning_rate": 4.508720173069859e-05, + "loss": 3.2293, + "step": 9164 + }, + { + "epoch": 0.2827078569803624, + "grad_norm": 0.16053371131420135, + "learning_rate": 4.5085599050817814e-05, + "loss": 3.2357, + "step": 9165 + }, + { + "epoch": 0.28273870344593577, + "grad_norm": 0.15632739663124084, + "learning_rate": 4.508399613805659e-05, + "loss": 3.1937, + "step": 9166 + }, + { + "epoch": 0.2827695499115092, + "grad_norm": 0.15343564748764038, + "learning_rate": 4.50823929924335e-05, + "loss": 3.2178, + "step": 9167 + }, + { + "epoch": 0.2828003963770826, + "grad_norm": 0.1504717618227005, + "learning_rate": 4.508078961396713e-05, + "loss": 3.2251, + "step": 9168 + }, + { + "epoch": 0.28283124284265604, + "grad_norm": 0.15602344274520874, + "learning_rate": 4.5079186002676064e-05, + "loss": 3.1983, + "step": 9169 + }, + { + "epoch": 0.2828620893082294, + "grad_norm": 0.15943211317062378, + "learning_rate": 4.5077582158578914e-05, + "loss": 3.1951, + "step": 9170 + }, + { + "epoch": 0.28289293577380287, + "grad_norm": 0.15059493482112885, + "learning_rate": 4.507597808169426e-05, + "loss": 3.2134, + "step": 9171 + }, + { + "epoch": 0.2829237822393763, + "grad_norm": 0.1537509560585022, + "learning_rate": 4.5074373772040705e-05, + "loss": 3.2122, + "step": 9172 + }, + { + "epoch": 0.2829546287049497, + "grad_norm": 0.1538919359445572, + "learning_rate": 4.507276922963685e-05, + "loss": 3.1611, + "step": 9173 + }, + { + "epoch": 0.28298547517052314, + "grad_norm": 0.14629504084587097, + "learning_rate": 4.5071164454501296e-05, + "loss": 3.2053, + "step": 9174 + }, + { + "epoch": 0.2830163216360965, + "grad_norm": 0.15097962319850922, + "learning_rate": 4.506955944665265e-05, + "loss": 3.2114, + "step": 9175 + }, + { + "epoch": 0.28304716810166997, + "grad_norm": 0.1618650108575821, + "learning_rate": 4.5067954206109534e-05, + "loss": 3.2538, + "step": 9176 + }, + { + "epoch": 0.28307801456724335, + "grad_norm": 0.15050722658634186, + "learning_rate": 4.506634873289053e-05, + "loss": 3.1585, + "step": 9177 + }, + { + "epoch": 0.2831088610328168, + "grad_norm": 0.14526601135730743, + "learning_rate": 4.5064743027014286e-05, + "loss": 3.1589, + "step": 9178 + }, + { + "epoch": 0.2831397074983902, + "grad_norm": 0.1493675708770752, + "learning_rate": 4.50631370884994e-05, + "loss": 3.183, + "step": 9179 + }, + { + "epoch": 0.2831705539639636, + "grad_norm": 0.1542447805404663, + "learning_rate": 4.506153091736449e-05, + "loss": 3.2064, + "step": 9180 + }, + { + "epoch": 0.283201400429537, + "grad_norm": 0.14816972613334656, + "learning_rate": 4.505992451362819e-05, + "loss": 3.2009, + "step": 9181 + }, + { + "epoch": 0.28323224689511045, + "grad_norm": 0.14481455087661743, + "learning_rate": 4.505831787730912e-05, + "loss": 3.2058, + "step": 9182 + }, + { + "epoch": 0.2832630933606839, + "grad_norm": 0.156694695353508, + "learning_rate": 4.5056711008425906e-05, + "loss": 3.222, + "step": 9183 + }, + { + "epoch": 0.2832939398262573, + "grad_norm": 0.14291447401046753, + "learning_rate": 4.5055103906997184e-05, + "loss": 3.1883, + "step": 9184 + }, + { + "epoch": 0.2833247862918307, + "grad_norm": 0.1523924022912979, + "learning_rate": 4.505349657304158e-05, + "loss": 3.1334, + "step": 9185 + }, + { + "epoch": 0.2833556327574041, + "grad_norm": 0.1545708328485489, + "learning_rate": 4.505188900657774e-05, + "loss": 3.1905, + "step": 9186 + }, + { + "epoch": 0.28338647922297755, + "grad_norm": 0.14334219694137573, + "learning_rate": 4.505028120762429e-05, + "loss": 3.1793, + "step": 9187 + }, + { + "epoch": 0.28341732568855094, + "grad_norm": 0.1670122593641281, + "learning_rate": 4.504867317619988e-05, + "loss": 3.2568, + "step": 9188 + }, + { + "epoch": 0.2834481721541244, + "grad_norm": 0.14743824303150177, + "learning_rate": 4.504706491232315e-05, + "loss": 3.2285, + "step": 9189 + }, + { + "epoch": 0.28347901861969776, + "grad_norm": 0.15147852897644043, + "learning_rate": 4.504545641601275e-05, + "loss": 3.2147, + "step": 9190 + }, + { + "epoch": 0.2835098650852712, + "grad_norm": 0.1534496396780014, + "learning_rate": 4.504384768728733e-05, + "loss": 3.1919, + "step": 9191 + }, + { + "epoch": 0.2835407115508446, + "grad_norm": 0.16025781631469727, + "learning_rate": 4.5042238726165545e-05, + "loss": 3.227, + "step": 9192 + }, + { + "epoch": 0.28357155801641803, + "grad_norm": 0.1474176049232483, + "learning_rate": 4.5040629532666044e-05, + "loss": 3.2092, + "step": 9193 + }, + { + "epoch": 0.2836024044819914, + "grad_norm": 0.17471842467784882, + "learning_rate": 4.503902010680748e-05, + "loss": 3.1734, + "step": 9194 + }, + { + "epoch": 0.28363325094756486, + "grad_norm": 0.16009418666362762, + "learning_rate": 4.503741044860853e-05, + "loss": 3.1834, + "step": 9195 + }, + { + "epoch": 0.2836640974131383, + "grad_norm": 0.15263143181800842, + "learning_rate": 4.5035800558087844e-05, + "loss": 3.2203, + "step": 9196 + }, + { + "epoch": 0.2836949438787117, + "grad_norm": 0.15625840425491333, + "learning_rate": 4.503419043526409e-05, + "loss": 3.2293, + "step": 9197 + }, + { + "epoch": 0.28372579034428513, + "grad_norm": 0.14701692759990692, + "learning_rate": 4.503258008015593e-05, + "loss": 3.1948, + "step": 9198 + }, + { + "epoch": 0.2837566368098585, + "grad_norm": 0.15289486944675446, + "learning_rate": 4.503096949278205e-05, + "loss": 3.2199, + "step": 9199 + }, + { + "epoch": 0.28378748327543196, + "grad_norm": 0.15322346985340118, + "learning_rate": 4.5029358673161117e-05, + "loss": 3.2303, + "step": 9200 + }, + { + "epoch": 0.28381832974100535, + "grad_norm": 0.15215815603733063, + "learning_rate": 4.50277476213118e-05, + "loss": 3.22, + "step": 9201 + }, + { + "epoch": 0.2838491762065788, + "grad_norm": 0.1503061205148697, + "learning_rate": 4.5026136337252786e-05, + "loss": 3.198, + "step": 9202 + }, + { + "epoch": 0.2838800226721522, + "grad_norm": 0.155244842171669, + "learning_rate": 4.502452482100276e-05, + "loss": 3.1876, + "step": 9203 + }, + { + "epoch": 0.2839108691377256, + "grad_norm": 0.14742416143417358, + "learning_rate": 4.50229130725804e-05, + "loss": 3.2031, + "step": 9204 + }, + { + "epoch": 0.283941715603299, + "grad_norm": 0.15079684555530548, + "learning_rate": 4.50213010920044e-05, + "loss": 3.1753, + "step": 9205 + }, + { + "epoch": 0.28397256206887245, + "grad_norm": 0.15205542743206024, + "learning_rate": 4.501968887929343e-05, + "loss": 3.1614, + "step": 9206 + }, + { + "epoch": 0.2840034085344459, + "grad_norm": 0.154934823513031, + "learning_rate": 4.5018076434466214e-05, + "loss": 3.1957, + "step": 9207 + }, + { + "epoch": 0.2840342550000193, + "grad_norm": 0.1526223123073578, + "learning_rate": 4.5016463757541427e-05, + "loss": 3.2176, + "step": 9208 + }, + { + "epoch": 0.2840651014655927, + "grad_norm": 0.14885471761226654, + "learning_rate": 4.5014850848537764e-05, + "loss": 3.2048, + "step": 9209 + }, + { + "epoch": 0.2840959479311661, + "grad_norm": 0.14702670276165009, + "learning_rate": 4.501323770747394e-05, + "loss": 3.213, + "step": 9210 + }, + { + "epoch": 0.28412679439673955, + "grad_norm": 0.15023136138916016, + "learning_rate": 4.5011624334368655e-05, + "loss": 3.2045, + "step": 9211 + }, + { + "epoch": 0.28415764086231293, + "grad_norm": 0.15414856374263763, + "learning_rate": 4.501001072924061e-05, + "loss": 3.236, + "step": 9212 + }, + { + "epoch": 0.2841884873278864, + "grad_norm": 0.14853322505950928, + "learning_rate": 4.500839689210851e-05, + "loss": 3.1954, + "step": 9213 + }, + { + "epoch": 0.28421933379345976, + "grad_norm": 0.1451243907213211, + "learning_rate": 4.5006782822991064e-05, + "loss": 3.2083, + "step": 9214 + }, + { + "epoch": 0.2842501802590332, + "grad_norm": 0.14636999368667603, + "learning_rate": 4.500516852190701e-05, + "loss": 3.2127, + "step": 9215 + }, + { + "epoch": 0.2842810267246066, + "grad_norm": 0.1574976146221161, + "learning_rate": 4.5003553988875033e-05, + "loss": 3.1766, + "step": 9216 + }, + { + "epoch": 0.28431187319018003, + "grad_norm": 0.14904972910881042, + "learning_rate": 4.500193922391388e-05, + "loss": 3.1839, + "step": 9217 + }, + { + "epoch": 0.2843427196557534, + "grad_norm": 0.1558409333229065, + "learning_rate": 4.500032422704226e-05, + "loss": 3.2143, + "step": 9218 + }, + { + "epoch": 0.28437356612132686, + "grad_norm": 0.16326908767223358, + "learning_rate": 4.4998708998278895e-05, + "loss": 3.1959, + "step": 9219 + }, + { + "epoch": 0.2844044125869003, + "grad_norm": 0.15282244980335236, + "learning_rate": 4.499709353764252e-05, + "loss": 3.2084, + "step": 9220 + }, + { + "epoch": 0.2844352590524737, + "grad_norm": 0.14814844727516174, + "learning_rate": 4.499547784515186e-05, + "loss": 3.2131, + "step": 9221 + }, + { + "epoch": 0.28446610551804713, + "grad_norm": 0.15469719469547272, + "learning_rate": 4.4993861920825654e-05, + "loss": 3.2151, + "step": 9222 + }, + { + "epoch": 0.2844969519836205, + "grad_norm": 0.14885041117668152, + "learning_rate": 4.499224576468263e-05, + "loss": 3.2258, + "step": 9223 + }, + { + "epoch": 0.28452779844919396, + "grad_norm": 0.15663853287696838, + "learning_rate": 4.4990629376741534e-05, + "loss": 3.2031, + "step": 9224 + }, + { + "epoch": 0.28455864491476734, + "grad_norm": 0.15223951637744904, + "learning_rate": 4.49890127570211e-05, + "loss": 3.2004, + "step": 9225 + }, + { + "epoch": 0.2845894913803408, + "grad_norm": 0.1443093717098236, + "learning_rate": 4.498739590554008e-05, + "loss": 3.177, + "step": 9226 + }, + { + "epoch": 0.2846203378459142, + "grad_norm": 0.15652288496494293, + "learning_rate": 4.4985778822317216e-05, + "loss": 3.2143, + "step": 9227 + }, + { + "epoch": 0.2846511843114876, + "grad_norm": 0.14587706327438354, + "learning_rate": 4.498416150737125e-05, + "loss": 3.1843, + "step": 9228 + }, + { + "epoch": 0.284682030777061, + "grad_norm": 0.14987482130527496, + "learning_rate": 4.4982543960720946e-05, + "loss": 3.2049, + "step": 9229 + }, + { + "epoch": 0.28471287724263444, + "grad_norm": 0.14634962379932404, + "learning_rate": 4.4980926182385044e-05, + "loss": 3.2203, + "step": 9230 + }, + { + "epoch": 0.2847437237082079, + "grad_norm": 0.1471215933561325, + "learning_rate": 4.497930817238232e-05, + "loss": 3.2081, + "step": 9231 + }, + { + "epoch": 0.28477457017378127, + "grad_norm": 0.14970295131206512, + "learning_rate": 4.497768993073152e-05, + "loss": 3.171, + "step": 9232 + }, + { + "epoch": 0.2848054166393547, + "grad_norm": 0.1497335582971573, + "learning_rate": 4.497607145745141e-05, + "loss": 3.2023, + "step": 9233 + }, + { + "epoch": 0.2848362631049281, + "grad_norm": 0.14638440310955048, + "learning_rate": 4.497445275256076e-05, + "loss": 3.2062, + "step": 9234 + }, + { + "epoch": 0.28486710957050154, + "grad_norm": 0.15926916897296906, + "learning_rate": 4.497283381607833e-05, + "loss": 3.1545, + "step": 9235 + }, + { + "epoch": 0.28489795603607493, + "grad_norm": 0.14825139939785004, + "learning_rate": 4.497121464802289e-05, + "loss": 3.2268, + "step": 9236 + }, + { + "epoch": 0.28492880250164837, + "grad_norm": 0.14274124801158905, + "learning_rate": 4.496959524841322e-05, + "loss": 3.207, + "step": 9237 + }, + { + "epoch": 0.28495964896722176, + "grad_norm": 0.14529570937156677, + "learning_rate": 4.4967975617268096e-05, + "loss": 3.2, + "step": 9238 + }, + { + "epoch": 0.2849904954327952, + "grad_norm": 0.14220310747623444, + "learning_rate": 4.496635575460629e-05, + "loss": 3.2237, + "step": 9239 + }, + { + "epoch": 0.2850213418983686, + "grad_norm": 0.15484856069087982, + "learning_rate": 4.4964735660446594e-05, + "loss": 3.2056, + "step": 9240 + }, + { + "epoch": 0.285052188363942, + "grad_norm": 0.1513386219739914, + "learning_rate": 4.496311533480778e-05, + "loss": 3.2048, + "step": 9241 + }, + { + "epoch": 0.2850830348295154, + "grad_norm": 0.15024502575397491, + "learning_rate": 4.4961494777708644e-05, + "loss": 3.2059, + "step": 9242 + }, + { + "epoch": 0.28511388129508886, + "grad_norm": 0.14278925955295563, + "learning_rate": 4.4959873989167965e-05, + "loss": 3.2293, + "step": 9243 + }, + { + "epoch": 0.2851447277606623, + "grad_norm": 0.15910659730434418, + "learning_rate": 4.495825296920455e-05, + "loss": 3.183, + "step": 9244 + }, + { + "epoch": 0.2851755742262357, + "grad_norm": 0.1479422152042389, + "learning_rate": 4.495663171783718e-05, + "loss": 3.1971, + "step": 9245 + }, + { + "epoch": 0.2852064206918091, + "grad_norm": 0.14994221925735474, + "learning_rate": 4.495501023508466e-05, + "loss": 3.2171, + "step": 9246 + }, + { + "epoch": 0.2852372671573825, + "grad_norm": 0.14996403455734253, + "learning_rate": 4.495338852096579e-05, + "loss": 3.1952, + "step": 9247 + }, + { + "epoch": 0.28526811362295595, + "grad_norm": 0.14869672060012817, + "learning_rate": 4.4951766575499364e-05, + "loss": 3.2009, + "step": 9248 + }, + { + "epoch": 0.28529896008852934, + "grad_norm": 0.14816917479038239, + "learning_rate": 4.4950144398704204e-05, + "loss": 3.2129, + "step": 9249 + }, + { + "epoch": 0.2853298065541028, + "grad_norm": 0.15257370471954346, + "learning_rate": 4.4948521990599105e-05, + "loss": 3.1738, + "step": 9250 + }, + { + "epoch": 0.28536065301967617, + "grad_norm": 0.14614728093147278, + "learning_rate": 4.494689935120288e-05, + "loss": 3.1716, + "step": 9251 + }, + { + "epoch": 0.2853914994852496, + "grad_norm": 0.15831951797008514, + "learning_rate": 4.494527648053434e-05, + "loss": 3.1771, + "step": 9252 + }, + { + "epoch": 0.285422345950823, + "grad_norm": 0.1468699872493744, + "learning_rate": 4.49436533786123e-05, + "loss": 3.1979, + "step": 9253 + }, + { + "epoch": 0.28545319241639644, + "grad_norm": 0.15107543766498566, + "learning_rate": 4.4942030045455596e-05, + "loss": 3.2018, + "step": 9254 + }, + { + "epoch": 0.2854840388819699, + "grad_norm": 0.15303660929203033, + "learning_rate": 4.494040648108303e-05, + "loss": 3.1856, + "step": 9255 + }, + { + "epoch": 0.28551488534754327, + "grad_norm": 0.17057733237743378, + "learning_rate": 4.4938782685513435e-05, + "loss": 3.1927, + "step": 9256 + }, + { + "epoch": 0.2855457318131167, + "grad_norm": 0.1448042094707489, + "learning_rate": 4.4937158658765644e-05, + "loss": 3.2251, + "step": 9257 + }, + { + "epoch": 0.2855765782786901, + "grad_norm": 0.15451060235500336, + "learning_rate": 4.493553440085847e-05, + "loss": 3.1711, + "step": 9258 + }, + { + "epoch": 0.28560742474426354, + "grad_norm": 0.15215842425823212, + "learning_rate": 4.493390991181076e-05, + "loss": 3.2647, + "step": 9259 + }, + { + "epoch": 0.2856382712098369, + "grad_norm": 0.15934167802333832, + "learning_rate": 4.4932285191641335e-05, + "loss": 3.227, + "step": 9260 + }, + { + "epoch": 0.28566911767541037, + "grad_norm": 0.15428873896598816, + "learning_rate": 4.493066024036905e-05, + "loss": 3.2125, + "step": 9261 + }, + { + "epoch": 0.28569996414098375, + "grad_norm": 0.1541547179222107, + "learning_rate": 4.492903505801273e-05, + "loss": 3.1646, + "step": 9262 + }, + { + "epoch": 0.2857308106065572, + "grad_norm": 0.15651027858257294, + "learning_rate": 4.492740964459123e-05, + "loss": 3.2091, + "step": 9263 + }, + { + "epoch": 0.2857616570721306, + "grad_norm": 0.16051122546195984, + "learning_rate": 4.492578400012338e-05, + "loss": 3.196, + "step": 9264 + }, + { + "epoch": 0.285792503537704, + "grad_norm": 0.1587219536304474, + "learning_rate": 4.4924158124628055e-05, + "loss": 3.2104, + "step": 9265 + }, + { + "epoch": 0.2858233500032774, + "grad_norm": 0.15447412431240082, + "learning_rate": 4.4922532018124076e-05, + "loss": 3.2394, + "step": 9266 + }, + { + "epoch": 0.28585419646885085, + "grad_norm": 0.1728278398513794, + "learning_rate": 4.492090568063032e-05, + "loss": 3.1979, + "step": 9267 + }, + { + "epoch": 0.2858850429344243, + "grad_norm": 0.15104831755161285, + "learning_rate": 4.4919279112165625e-05, + "loss": 3.1716, + "step": 9268 + }, + { + "epoch": 0.2859158893999977, + "grad_norm": 0.15372078120708466, + "learning_rate": 4.4917652312748864e-05, + "loss": 3.202, + "step": 9269 + }, + { + "epoch": 0.2859467358655711, + "grad_norm": 0.14890731871128082, + "learning_rate": 4.491602528239889e-05, + "loss": 3.2269, + "step": 9270 + }, + { + "epoch": 0.2859775823311445, + "grad_norm": 0.14693765342235565, + "learning_rate": 4.4914398021134576e-05, + "loss": 3.1907, + "step": 9271 + }, + { + "epoch": 0.28600842879671795, + "grad_norm": 0.14990700781345367, + "learning_rate": 4.491277052897478e-05, + "loss": 3.1639, + "step": 9272 + }, + { + "epoch": 0.28603927526229134, + "grad_norm": 0.15094609558582306, + "learning_rate": 4.491114280593838e-05, + "loss": 3.1918, + "step": 9273 + }, + { + "epoch": 0.2860701217278648, + "grad_norm": 0.15066389739513397, + "learning_rate": 4.4909514852044244e-05, + "loss": 3.1722, + "step": 9274 + }, + { + "epoch": 0.28610096819343817, + "grad_norm": 0.14426285028457642, + "learning_rate": 4.490788666731124e-05, + "loss": 3.2236, + "step": 9275 + }, + { + "epoch": 0.2861318146590116, + "grad_norm": 0.14986884593963623, + "learning_rate": 4.490625825175826e-05, + "loss": 3.1661, + "step": 9276 + }, + { + "epoch": 0.286162661124585, + "grad_norm": 0.15011878311634064, + "learning_rate": 4.490462960540418e-05, + "loss": 3.1771, + "step": 9277 + }, + { + "epoch": 0.28619350759015844, + "grad_norm": 0.15093213319778442, + "learning_rate": 4.4903000728267876e-05, + "loss": 3.1837, + "step": 9278 + }, + { + "epoch": 0.2862243540557319, + "grad_norm": 0.16199246048927307, + "learning_rate": 4.4901371620368236e-05, + "loss": 3.1818, + "step": 9279 + }, + { + "epoch": 0.28625520052130526, + "grad_norm": 0.1555112898349762, + "learning_rate": 4.489974228172416e-05, + "loss": 3.2017, + "step": 9280 + }, + { + "epoch": 0.2862860469868787, + "grad_norm": 0.15068869292736053, + "learning_rate": 4.489811271235452e-05, + "loss": 3.2107, + "step": 9281 + }, + { + "epoch": 0.2863168934524521, + "grad_norm": 0.14717315137386322, + "learning_rate": 4.489648291227824e-05, + "loss": 3.2184, + "step": 9282 + }, + { + "epoch": 0.28634773991802553, + "grad_norm": 0.14502742886543274, + "learning_rate": 4.489485288151418e-05, + "loss": 3.1841, + "step": 9283 + }, + { + "epoch": 0.2863785863835989, + "grad_norm": 0.15263110399246216, + "learning_rate": 4.4893222620081256e-05, + "loss": 3.1858, + "step": 9284 + }, + { + "epoch": 0.28640943284917236, + "grad_norm": 0.15128666162490845, + "learning_rate": 4.4891592127998375e-05, + "loss": 3.1797, + "step": 9285 + }, + { + "epoch": 0.28644027931474575, + "grad_norm": 0.1579216867685318, + "learning_rate": 4.4889961405284434e-05, + "loss": 3.2077, + "step": 9286 + }, + { + "epoch": 0.2864711257803192, + "grad_norm": 0.1469510793685913, + "learning_rate": 4.4888330451958346e-05, + "loss": 3.1901, + "step": 9287 + }, + { + "epoch": 0.2865019722458926, + "grad_norm": 0.1542111039161682, + "learning_rate": 4.488669926803902e-05, + "loss": 3.1924, + "step": 9288 + }, + { + "epoch": 0.286532818711466, + "grad_norm": 0.14558808505535126, + "learning_rate": 4.488506785354536e-05, + "loss": 3.1972, + "step": 9289 + }, + { + "epoch": 0.2865636651770394, + "grad_norm": 0.1594768762588501, + "learning_rate": 4.488343620849629e-05, + "loss": 3.1604, + "step": 9290 + }, + { + "epoch": 0.28659451164261285, + "grad_norm": 0.16414916515350342, + "learning_rate": 4.488180433291073e-05, + "loss": 3.1781, + "step": 9291 + }, + { + "epoch": 0.2866253581081863, + "grad_norm": 0.15272486209869385, + "learning_rate": 4.4880172226807594e-05, + "loss": 3.208, + "step": 9292 + }, + { + "epoch": 0.2866562045737597, + "grad_norm": 0.16135847568511963, + "learning_rate": 4.4878539890205804e-05, + "loss": 3.2026, + "step": 9293 + }, + { + "epoch": 0.2866870510393331, + "grad_norm": 0.14499397575855255, + "learning_rate": 4.4876907323124295e-05, + "loss": 3.1994, + "step": 9294 + }, + { + "epoch": 0.2867178975049065, + "grad_norm": 0.15638238191604614, + "learning_rate": 4.487527452558199e-05, + "loss": 3.1921, + "step": 9295 + }, + { + "epoch": 0.28674874397047995, + "grad_norm": 0.14903734624385834, + "learning_rate": 4.4873641497597815e-05, + "loss": 3.2186, + "step": 9296 + }, + { + "epoch": 0.28677959043605333, + "grad_norm": 0.1466771513223648, + "learning_rate": 4.487200823919071e-05, + "loss": 3.2051, + "step": 9297 + }, + { + "epoch": 0.2868104369016268, + "grad_norm": 0.1545642465353012, + "learning_rate": 4.4870374750379616e-05, + "loss": 3.2075, + "step": 9298 + }, + { + "epoch": 0.28684128336720016, + "grad_norm": 0.15657907724380493, + "learning_rate": 4.486874103118346e-05, + "loss": 3.2323, + "step": 9299 + }, + { + "epoch": 0.2868721298327736, + "grad_norm": 0.14658212661743164, + "learning_rate": 4.48671070816212e-05, + "loss": 3.1734, + "step": 9300 + }, + { + "epoch": 0.286902976298347, + "grad_norm": 0.1504320651292801, + "learning_rate": 4.4865472901711766e-05, + "loss": 3.1305, + "step": 9301 + }, + { + "epoch": 0.28693382276392043, + "grad_norm": 0.15347056090831757, + "learning_rate": 4.4863838491474116e-05, + "loss": 3.2144, + "step": 9302 + }, + { + "epoch": 0.2869646692294939, + "grad_norm": 0.1551641821861267, + "learning_rate": 4.48622038509272e-05, + "loss": 3.211, + "step": 9303 + }, + { + "epoch": 0.28699551569506726, + "grad_norm": 0.15015603601932526, + "learning_rate": 4.486056898008996e-05, + "loss": 3.218, + "step": 9304 + }, + { + "epoch": 0.2870263621606407, + "grad_norm": 0.16946415603160858, + "learning_rate": 4.485893387898136e-05, + "loss": 3.1899, + "step": 9305 + }, + { + "epoch": 0.2870572086262141, + "grad_norm": 0.14614853262901306, + "learning_rate": 4.485729854762035e-05, + "loss": 3.2122, + "step": 9306 + }, + { + "epoch": 0.28708805509178753, + "grad_norm": 0.15841981768608093, + "learning_rate": 4.48556629860259e-05, + "loss": 3.2068, + "step": 9307 + }, + { + "epoch": 0.2871189015573609, + "grad_norm": 0.15198367834091187, + "learning_rate": 4.4854027194216965e-05, + "loss": 3.2066, + "step": 9308 + }, + { + "epoch": 0.28714974802293436, + "grad_norm": 0.15612325072288513, + "learning_rate": 4.485239117221252e-05, + "loss": 3.2014, + "step": 9309 + }, + { + "epoch": 0.28718059448850775, + "grad_norm": 0.14872276782989502, + "learning_rate": 4.485075492003153e-05, + "loss": 3.1861, + "step": 9310 + }, + { + "epoch": 0.2872114409540812, + "grad_norm": 0.1667512208223343, + "learning_rate": 4.484911843769296e-05, + "loss": 3.1956, + "step": 9311 + }, + { + "epoch": 0.2872422874196546, + "grad_norm": 0.15341496467590332, + "learning_rate": 4.4847481725215795e-05, + "loss": 3.1987, + "step": 9312 + }, + { + "epoch": 0.287273133885228, + "grad_norm": 0.14973287284374237, + "learning_rate": 4.484584478261901e-05, + "loss": 3.1716, + "step": 9313 + }, + { + "epoch": 0.2873039803508014, + "grad_norm": 0.18124933540821075, + "learning_rate": 4.484420760992158e-05, + "loss": 3.1963, + "step": 9314 + }, + { + "epoch": 0.28733482681637484, + "grad_norm": 0.1480063647031784, + "learning_rate": 4.484257020714248e-05, + "loss": 3.1628, + "step": 9315 + }, + { + "epoch": 0.2873656732819483, + "grad_norm": 0.15378116071224213, + "learning_rate": 4.48409325743007e-05, + "loss": 3.2224, + "step": 9316 + }, + { + "epoch": 0.2873965197475217, + "grad_norm": 0.1552547812461853, + "learning_rate": 4.4839294711415234e-05, + "loss": 3.1647, + "step": 9317 + }, + { + "epoch": 0.2874273662130951, + "grad_norm": 0.1901940107345581, + "learning_rate": 4.4837656618505077e-05, + "loss": 3.1934, + "step": 9318 + }, + { + "epoch": 0.2874582126786685, + "grad_norm": 0.15458449721336365, + "learning_rate": 4.48360182955892e-05, + "loss": 3.2159, + "step": 9319 + }, + { + "epoch": 0.28748905914424194, + "grad_norm": 0.15117418766021729, + "learning_rate": 4.4834379742686614e-05, + "loss": 3.1963, + "step": 9320 + }, + { + "epoch": 0.28751990560981533, + "grad_norm": 0.15417417883872986, + "learning_rate": 4.4832740959816324e-05, + "loss": 3.2176, + "step": 9321 + }, + { + "epoch": 0.28755075207538877, + "grad_norm": 0.31230542063713074, + "learning_rate": 4.483110194699731e-05, + "loss": 3.1939, + "step": 9322 + }, + { + "epoch": 0.28758159854096216, + "grad_norm": 0.15285053849220276, + "learning_rate": 4.482946270424859e-05, + "loss": 3.1839, + "step": 9323 + }, + { + "epoch": 0.2876124450065356, + "grad_norm": 0.16416260600090027, + "learning_rate": 4.482782323158916e-05, + "loss": 3.2096, + "step": 9324 + }, + { + "epoch": 0.287643291472109, + "grad_norm": 0.14886368811130524, + "learning_rate": 4.4826183529038044e-05, + "loss": 3.1856, + "step": 9325 + }, + { + "epoch": 0.28767413793768243, + "grad_norm": 0.15082448720932007, + "learning_rate": 4.482454359661424e-05, + "loss": 3.1813, + "step": 9326 + }, + { + "epoch": 0.28770498440325587, + "grad_norm": 0.1553199738264084, + "learning_rate": 4.482290343433676e-05, + "loss": 3.1971, + "step": 9327 + }, + { + "epoch": 0.28773583086882926, + "grad_norm": 0.14688825607299805, + "learning_rate": 4.4821263042224636e-05, + "loss": 3.2074, + "step": 9328 + }, + { + "epoch": 0.2877666773344027, + "grad_norm": 0.15377214550971985, + "learning_rate": 4.4819622420296875e-05, + "loss": 3.2361, + "step": 9329 + }, + { + "epoch": 0.2877975237999761, + "grad_norm": 0.16249239444732666, + "learning_rate": 4.4817981568572504e-05, + "loss": 3.1694, + "step": 9330 + }, + { + "epoch": 0.2878283702655495, + "grad_norm": 0.1525731235742569, + "learning_rate": 4.481634048707054e-05, + "loss": 3.2142, + "step": 9331 + }, + { + "epoch": 0.2878592167311229, + "grad_norm": 0.1482783704996109, + "learning_rate": 4.481469917581002e-05, + "loss": 3.1802, + "step": 9332 + }, + { + "epoch": 0.28789006319669636, + "grad_norm": 0.1539820283651352, + "learning_rate": 4.4813057634809976e-05, + "loss": 3.2092, + "step": 9333 + }, + { + "epoch": 0.28792090966226974, + "grad_norm": 0.14931297302246094, + "learning_rate": 4.481141586408943e-05, + "loss": 3.1562, + "step": 9334 + }, + { + "epoch": 0.2879517561278432, + "grad_norm": 0.14806371927261353, + "learning_rate": 4.4809773863667426e-05, + "loss": 3.1849, + "step": 9335 + }, + { + "epoch": 0.28798260259341657, + "grad_norm": 0.17300790548324585, + "learning_rate": 4.4808131633562986e-05, + "loss": 3.2204, + "step": 9336 + }, + { + "epoch": 0.28801344905899, + "grad_norm": 0.15198983252048492, + "learning_rate": 4.480648917379518e-05, + "loss": 3.2231, + "step": 9337 + }, + { + "epoch": 0.2880442955245634, + "grad_norm": 0.1518932580947876, + "learning_rate": 4.480484648438302e-05, + "loss": 3.1699, + "step": 9338 + }, + { + "epoch": 0.28807514199013684, + "grad_norm": 0.1544426530599594, + "learning_rate": 4.4803203565345574e-05, + "loss": 3.2396, + "step": 9339 + }, + { + "epoch": 0.2881059884557103, + "grad_norm": 0.15488137304782867, + "learning_rate": 4.480156041670188e-05, + "loss": 3.1967, + "step": 9340 + }, + { + "epoch": 0.28813683492128367, + "grad_norm": 0.1520427018404007, + "learning_rate": 4.4799917038470994e-05, + "loss": 3.1931, + "step": 9341 + }, + { + "epoch": 0.2881676813868571, + "grad_norm": 0.15367065370082855, + "learning_rate": 4.4798273430671965e-05, + "loss": 3.2121, + "step": 9342 + }, + { + "epoch": 0.2881985278524305, + "grad_norm": 0.14743672311306, + "learning_rate": 4.4796629593323855e-05, + "loss": 3.1979, + "step": 9343 + }, + { + "epoch": 0.28822937431800394, + "grad_norm": 0.17448405921459198, + "learning_rate": 4.479498552644573e-05, + "loss": 3.1987, + "step": 9344 + }, + { + "epoch": 0.2882602207835773, + "grad_norm": 0.1520143300294876, + "learning_rate": 4.479334123005663e-05, + "loss": 3.2169, + "step": 9345 + }, + { + "epoch": 0.28829106724915077, + "grad_norm": 0.15083779394626617, + "learning_rate": 4.479169670417564e-05, + "loss": 3.1611, + "step": 9346 + }, + { + "epoch": 0.28832191371472415, + "grad_norm": 0.15237975120544434, + "learning_rate": 4.479005194882182e-05, + "loss": 3.1973, + "step": 9347 + }, + { + "epoch": 0.2883527601802976, + "grad_norm": 0.15813879668712616, + "learning_rate": 4.478840696401424e-05, + "loss": 3.2109, + "step": 9348 + }, + { + "epoch": 0.288383606645871, + "grad_norm": 0.14770331978797913, + "learning_rate": 4.4786761749771965e-05, + "loss": 3.1714, + "step": 9349 + }, + { + "epoch": 0.2884144531114444, + "grad_norm": 0.15580973029136658, + "learning_rate": 4.4785116306114085e-05, + "loss": 3.1465, + "step": 9350 + }, + { + "epoch": 0.28844529957701787, + "grad_norm": 0.17020367085933685, + "learning_rate": 4.478347063305967e-05, + "loss": 3.18, + "step": 9351 + }, + { + "epoch": 0.28847614604259125, + "grad_norm": 0.1512719988822937, + "learning_rate": 4.47818247306278e-05, + "loss": 3.1899, + "step": 9352 + }, + { + "epoch": 0.2885069925081647, + "grad_norm": 0.15193438529968262, + "learning_rate": 4.4780178598837564e-05, + "loss": 3.2078, + "step": 9353 + }, + { + "epoch": 0.2885378389737381, + "grad_norm": 0.1557133048772812, + "learning_rate": 4.477853223770804e-05, + "loss": 3.2076, + "step": 9354 + }, + { + "epoch": 0.2885686854393115, + "grad_norm": 0.14841420948505402, + "learning_rate": 4.477688564725832e-05, + "loss": 3.1896, + "step": 9355 + }, + { + "epoch": 0.2885995319048849, + "grad_norm": 0.14701348543167114, + "learning_rate": 4.477523882750749e-05, + "loss": 3.1855, + "step": 9356 + }, + { + "epoch": 0.28863037837045835, + "grad_norm": 0.15191248059272766, + "learning_rate": 4.4773591778474666e-05, + "loss": 3.1936, + "step": 9357 + }, + { + "epoch": 0.28866122483603174, + "grad_norm": 0.15294130146503448, + "learning_rate": 4.4771944500178916e-05, + "loss": 3.2032, + "step": 9358 + }, + { + "epoch": 0.2886920713016052, + "grad_norm": 0.15891999006271362, + "learning_rate": 4.477029699263935e-05, + "loss": 3.2122, + "step": 9359 + }, + { + "epoch": 0.28872291776717857, + "grad_norm": 0.15183936059474945, + "learning_rate": 4.476864925587507e-05, + "loss": 3.2406, + "step": 9360 + }, + { + "epoch": 0.288753764232752, + "grad_norm": 0.15586164593696594, + "learning_rate": 4.476700128990519e-05, + "loss": 3.1858, + "step": 9361 + }, + { + "epoch": 0.28878461069832545, + "grad_norm": 0.15003284811973572, + "learning_rate": 4.476535309474881e-05, + "loss": 3.221, + "step": 9362 + }, + { + "epoch": 0.28881545716389884, + "grad_norm": 0.14607690274715424, + "learning_rate": 4.476370467042503e-05, + "loss": 3.1988, + "step": 9363 + }, + { + "epoch": 0.2888463036294723, + "grad_norm": 0.1497257500886917, + "learning_rate": 4.476205601695297e-05, + "loss": 3.2238, + "step": 9364 + }, + { + "epoch": 0.28887715009504566, + "grad_norm": 0.15133731067180634, + "learning_rate": 4.476040713435174e-05, + "loss": 3.2297, + "step": 9365 + }, + { + "epoch": 0.2889079965606191, + "grad_norm": 0.1527027189731598, + "learning_rate": 4.475875802264048e-05, + "loss": 3.2007, + "step": 9366 + }, + { + "epoch": 0.2889388430261925, + "grad_norm": 0.15133638679981232, + "learning_rate": 4.475710868183829e-05, + "loss": 3.2091, + "step": 9367 + }, + { + "epoch": 0.28896968949176594, + "grad_norm": 0.14837147295475006, + "learning_rate": 4.4755459111964294e-05, + "loss": 3.167, + "step": 9368 + }, + { + "epoch": 0.2890005359573393, + "grad_norm": 0.15099719166755676, + "learning_rate": 4.475380931303762e-05, + "loss": 3.2054, + "step": 9369 + }, + { + "epoch": 0.28903138242291276, + "grad_norm": 0.1528233289718628, + "learning_rate": 4.47521592850774e-05, + "loss": 3.1909, + "step": 9370 + }, + { + "epoch": 0.28906222888848615, + "grad_norm": 0.1464589238166809, + "learning_rate": 4.475050902810277e-05, + "loss": 3.2162, + "step": 9371 + }, + { + "epoch": 0.2890930753540596, + "grad_norm": 0.15863345563411713, + "learning_rate": 4.4748858542132846e-05, + "loss": 3.1939, + "step": 9372 + }, + { + "epoch": 0.289123921819633, + "grad_norm": 0.1511395126581192, + "learning_rate": 4.474720782718678e-05, + "loss": 3.2011, + "step": 9373 + }, + { + "epoch": 0.2891547682852064, + "grad_norm": 0.15180666744709015, + "learning_rate": 4.47455568832837e-05, + "loss": 3.2183, + "step": 9374 + }, + { + "epoch": 0.28918561475077986, + "grad_norm": 0.1401868462562561, + "learning_rate": 4.4743905710442756e-05, + "loss": 3.1588, + "step": 9375 + }, + { + "epoch": 0.28921646121635325, + "grad_norm": 0.15439490973949432, + "learning_rate": 4.474225430868309e-05, + "loss": 3.1582, + "step": 9376 + }, + { + "epoch": 0.2892473076819267, + "grad_norm": 0.14322294294834137, + "learning_rate": 4.4740602678023855e-05, + "loss": 3.1752, + "step": 9377 + }, + { + "epoch": 0.2892781541475001, + "grad_norm": 0.1495647132396698, + "learning_rate": 4.473895081848419e-05, + "loss": 3.1896, + "step": 9378 + }, + { + "epoch": 0.2893090006130735, + "grad_norm": 0.14433959126472473, + "learning_rate": 4.4737298730083255e-05, + "loss": 3.2021, + "step": 9379 + }, + { + "epoch": 0.2893398470786469, + "grad_norm": 0.1424720585346222, + "learning_rate": 4.47356464128402e-05, + "loss": 3.1779, + "step": 9380 + }, + { + "epoch": 0.28937069354422035, + "grad_norm": 0.15000423789024353, + "learning_rate": 4.4733993866774185e-05, + "loss": 3.2187, + "step": 9381 + }, + { + "epoch": 0.28940154000979373, + "grad_norm": 0.14813733100891113, + "learning_rate": 4.4732341091904366e-05, + "loss": 3.1745, + "step": 9382 + }, + { + "epoch": 0.2894323864753672, + "grad_norm": 0.14512322843074799, + "learning_rate": 4.473068808824991e-05, + "loss": 3.1802, + "step": 9383 + }, + { + "epoch": 0.28946323294094056, + "grad_norm": 0.1589820683002472, + "learning_rate": 4.4729034855829985e-05, + "loss": 3.2087, + "step": 9384 + }, + { + "epoch": 0.289494079406514, + "grad_norm": 0.15062277019023895, + "learning_rate": 4.4727381394663756e-05, + "loss": 3.1827, + "step": 9385 + }, + { + "epoch": 0.28952492587208745, + "grad_norm": 0.15750353038311005, + "learning_rate": 4.47257277047704e-05, + "loss": 3.2165, + "step": 9386 + }, + { + "epoch": 0.28955577233766083, + "grad_norm": 0.14824767410755157, + "learning_rate": 4.472407378616908e-05, + "loss": 3.2015, + "step": 9387 + }, + { + "epoch": 0.2895866188032343, + "grad_norm": 0.17753228545188904, + "learning_rate": 4.4722419638878974e-05, + "loss": 3.201, + "step": 9388 + }, + { + "epoch": 0.28961746526880766, + "grad_norm": 0.15725542604923248, + "learning_rate": 4.472076526291927e-05, + "loss": 3.1891, + "step": 9389 + }, + { + "epoch": 0.2896483117343811, + "grad_norm": 0.15571871399879456, + "learning_rate": 4.471911065830914e-05, + "loss": 3.1475, + "step": 9390 + }, + { + "epoch": 0.2896791581999545, + "grad_norm": 0.15079155564308167, + "learning_rate": 4.4717455825067774e-05, + "loss": 3.1769, + "step": 9391 + }, + { + "epoch": 0.28971000466552793, + "grad_norm": 0.1525374799966812, + "learning_rate": 4.471580076321436e-05, + "loss": 3.2227, + "step": 9392 + }, + { + "epoch": 0.2897408511311013, + "grad_norm": 0.14721879363059998, + "learning_rate": 4.4714145472768074e-05, + "loss": 3.2184, + "step": 9393 + }, + { + "epoch": 0.28977169759667476, + "grad_norm": 0.15433959662914276, + "learning_rate": 4.471248995374812e-05, + "loss": 3.1974, + "step": 9394 + }, + { + "epoch": 0.28980254406224815, + "grad_norm": 0.15040577948093414, + "learning_rate": 4.471083420617369e-05, + "loss": 3.2014, + "step": 9395 + }, + { + "epoch": 0.2898333905278216, + "grad_norm": 0.31198006868362427, + "learning_rate": 4.470917823006399e-05, + "loss": 3.2177, + "step": 9396 + }, + { + "epoch": 0.289864236993395, + "grad_norm": 0.14876176416873932, + "learning_rate": 4.470752202543821e-05, + "loss": 3.1985, + "step": 9397 + }, + { + "epoch": 0.2898950834589684, + "grad_norm": 0.1608709692955017, + "learning_rate": 4.470586559231556e-05, + "loss": 3.205, + "step": 9398 + }, + { + "epoch": 0.28992592992454186, + "grad_norm": 0.15416325628757477, + "learning_rate": 4.470420893071523e-05, + "loss": 3.2138, + "step": 9399 + }, + { + "epoch": 0.28995677639011525, + "grad_norm": 0.16942842304706573, + "learning_rate": 4.470255204065644e-05, + "loss": 3.1931, + "step": 9400 + }, + { + "epoch": 0.2899876228556887, + "grad_norm": 0.16555088758468628, + "learning_rate": 4.47008949221584e-05, + "loss": 3.2034, + "step": 9401 + }, + { + "epoch": 0.2900184693212621, + "grad_norm": 0.14769117534160614, + "learning_rate": 4.4699237575240324e-05, + "loss": 3.197, + "step": 9402 + }, + { + "epoch": 0.2900493157868355, + "grad_norm": 0.16381852328777313, + "learning_rate": 4.4697579999921425e-05, + "loss": 3.1637, + "step": 9403 + }, + { + "epoch": 0.2900801622524089, + "grad_norm": 0.1491103619337082, + "learning_rate": 4.469592219622093e-05, + "loss": 3.1751, + "step": 9404 + }, + { + "epoch": 0.29011100871798234, + "grad_norm": 0.17765046656131744, + "learning_rate": 4.469426416415804e-05, + "loss": 3.2012, + "step": 9405 + }, + { + "epoch": 0.29014185518355573, + "grad_norm": 0.1531924158334732, + "learning_rate": 4.4692605903752e-05, + "loss": 3.177, + "step": 9406 + }, + { + "epoch": 0.29017270164912917, + "grad_norm": 0.15786518156528473, + "learning_rate": 4.4690947415022025e-05, + "loss": 3.2191, + "step": 9407 + }, + { + "epoch": 0.29020354811470256, + "grad_norm": 0.1607697457075119, + "learning_rate": 4.4689288697987356e-05, + "loss": 3.1998, + "step": 9408 + }, + { + "epoch": 0.290234394580276, + "grad_norm": 0.15728174149990082, + "learning_rate": 4.4687629752667204e-05, + "loss": 3.1967, + "step": 9409 + }, + { + "epoch": 0.29026524104584944, + "grad_norm": 0.15422925353050232, + "learning_rate": 4.468597057908083e-05, + "loss": 3.1903, + "step": 9410 + }, + { + "epoch": 0.29029608751142283, + "grad_norm": 0.14798897504806519, + "learning_rate": 4.468431117724744e-05, + "loss": 3.2151, + "step": 9411 + }, + { + "epoch": 0.29032693397699627, + "grad_norm": 0.15742595493793488, + "learning_rate": 4.46826515471863e-05, + "loss": 3.1984, + "step": 9412 + }, + { + "epoch": 0.29035778044256966, + "grad_norm": 0.15954338014125824, + "learning_rate": 4.468099168891664e-05, + "loss": 3.2058, + "step": 9413 + }, + { + "epoch": 0.2903886269081431, + "grad_norm": 0.15310628712177277, + "learning_rate": 4.4679331602457716e-05, + "loss": 3.2264, + "step": 9414 + }, + { + "epoch": 0.2904194733737165, + "grad_norm": 0.15362587571144104, + "learning_rate": 4.4677671287828756e-05, + "loss": 3.1799, + "step": 9415 + }, + { + "epoch": 0.2904503198392899, + "grad_norm": 0.15845990180969238, + "learning_rate": 4.467601074504903e-05, + "loss": 3.1848, + "step": 9416 + }, + { + "epoch": 0.2904811663048633, + "grad_norm": 0.141227588057518, + "learning_rate": 4.4674349974137784e-05, + "loss": 3.1921, + "step": 9417 + }, + { + "epoch": 0.29051201277043676, + "grad_norm": 0.15206366777420044, + "learning_rate": 4.4672688975114274e-05, + "loss": 3.1856, + "step": 9418 + }, + { + "epoch": 0.29054285923601014, + "grad_norm": 0.1448068767786026, + "learning_rate": 4.4671027747997754e-05, + "loss": 3.1594, + "step": 9419 + }, + { + "epoch": 0.2905737057015836, + "grad_norm": 0.14523258805274963, + "learning_rate": 4.466936629280749e-05, + "loss": 3.1882, + "step": 9420 + }, + { + "epoch": 0.29060455216715697, + "grad_norm": 0.1518716812133789, + "learning_rate": 4.4667704609562744e-05, + "loss": 3.1931, + "step": 9421 + }, + { + "epoch": 0.2906353986327304, + "grad_norm": 0.14369389414787292, + "learning_rate": 4.466604269828278e-05, + "loss": 3.2041, + "step": 9422 + }, + { + "epoch": 0.29066624509830385, + "grad_norm": 0.1551714390516281, + "learning_rate": 4.466438055898687e-05, + "loss": 3.2147, + "step": 9423 + }, + { + "epoch": 0.29069709156387724, + "grad_norm": 0.14528732001781464, + "learning_rate": 4.466271819169429e-05, + "loss": 3.1936, + "step": 9424 + }, + { + "epoch": 0.2907279380294507, + "grad_norm": 0.150056391954422, + "learning_rate": 4.466105559642429e-05, + "loss": 3.1923, + "step": 9425 + }, + { + "epoch": 0.29075878449502407, + "grad_norm": 0.153924822807312, + "learning_rate": 4.4659392773196183e-05, + "loss": 3.2169, + "step": 9426 + }, + { + "epoch": 0.2907896309605975, + "grad_norm": 0.1638214886188507, + "learning_rate": 4.4657729722029225e-05, + "loss": 3.1811, + "step": 9427 + }, + { + "epoch": 0.2908204774261709, + "grad_norm": 0.15387378633022308, + "learning_rate": 4.46560664429427e-05, + "loss": 3.1931, + "step": 9428 + }, + { + "epoch": 0.29085132389174434, + "grad_norm": 0.15828923881053925, + "learning_rate": 4.4654402935955905e-05, + "loss": 3.1998, + "step": 9429 + }, + { + "epoch": 0.2908821703573177, + "grad_norm": 0.15796880424022675, + "learning_rate": 4.465273920108811e-05, + "loss": 3.2113, + "step": 9430 + }, + { + "epoch": 0.29091301682289117, + "grad_norm": 0.14787471294403076, + "learning_rate": 4.465107523835862e-05, + "loss": 3.162, + "step": 9431 + }, + { + "epoch": 0.29094386328846455, + "grad_norm": 0.15320834517478943, + "learning_rate": 4.464941104778672e-05, + "loss": 3.1827, + "step": 9432 + }, + { + "epoch": 0.290974709754038, + "grad_norm": 0.15315021574497223, + "learning_rate": 4.464774662939171e-05, + "loss": 3.1977, + "step": 9433 + }, + { + "epoch": 0.29100555621961144, + "grad_norm": 0.155333012342453, + "learning_rate": 4.464608198319288e-05, + "loss": 3.184, + "step": 9434 + }, + { + "epoch": 0.2910364026851848, + "grad_norm": 0.15434707701206207, + "learning_rate": 4.464441710920953e-05, + "loss": 3.197, + "step": 9435 + }, + { + "epoch": 0.29106724915075827, + "grad_norm": 0.1500234752893448, + "learning_rate": 4.464275200746098e-05, + "loss": 3.2214, + "step": 9436 + }, + { + "epoch": 0.29109809561633165, + "grad_norm": 0.15008701384067535, + "learning_rate": 4.464108667796652e-05, + "loss": 3.1995, + "step": 9437 + }, + { + "epoch": 0.2911289420819051, + "grad_norm": 0.14878059923648834, + "learning_rate": 4.4639421120745464e-05, + "loss": 3.2024, + "step": 9438 + }, + { + "epoch": 0.2911597885474785, + "grad_norm": 0.1522700935602188, + "learning_rate": 4.463775533581712e-05, + "loss": 3.2016, + "step": 9439 + }, + { + "epoch": 0.2911906350130519, + "grad_norm": 0.15076836943626404, + "learning_rate": 4.4636089323200804e-05, + "loss": 3.1666, + "step": 9440 + }, + { + "epoch": 0.2912214814786253, + "grad_norm": 0.1497928947210312, + "learning_rate": 4.4634423082915835e-05, + "loss": 3.1802, + "step": 9441 + }, + { + "epoch": 0.29125232794419875, + "grad_norm": 0.15656408667564392, + "learning_rate": 4.4632756614981526e-05, + "loss": 3.2027, + "step": 9442 + }, + { + "epoch": 0.29128317440977214, + "grad_norm": 0.15744338929653168, + "learning_rate": 4.46310899194172e-05, + "loss": 3.1994, + "step": 9443 + }, + { + "epoch": 0.2913140208753456, + "grad_norm": 0.14855125546455383, + "learning_rate": 4.462942299624219e-05, + "loss": 3.2217, + "step": 9444 + }, + { + "epoch": 0.29134486734091897, + "grad_norm": 0.1494584083557129, + "learning_rate": 4.462775584547581e-05, + "loss": 3.2042, + "step": 9445 + }, + { + "epoch": 0.2913757138064924, + "grad_norm": 0.16584371030330658, + "learning_rate": 4.46260884671374e-05, + "loss": 3.2228, + "step": 9446 + }, + { + "epoch": 0.29140656027206585, + "grad_norm": 0.1515832245349884, + "learning_rate": 4.462442086124628e-05, + "loss": 3.2202, + "step": 9447 + }, + { + "epoch": 0.29143740673763924, + "grad_norm": 0.1518627107143402, + "learning_rate": 4.462275302782181e-05, + "loss": 3.1989, + "step": 9448 + }, + { + "epoch": 0.2914682532032127, + "grad_norm": 0.1642414629459381, + "learning_rate": 4.46210849668833e-05, + "loss": 3.1819, + "step": 9449 + }, + { + "epoch": 0.29149909966878607, + "grad_norm": 0.1629331111907959, + "learning_rate": 4.46194166784501e-05, + "loss": 3.2024, + "step": 9450 + }, + { + "epoch": 0.2915299461343595, + "grad_norm": 0.15364597737789154, + "learning_rate": 4.461774816254155e-05, + "loss": 3.2706, + "step": 9451 + }, + { + "epoch": 0.2915607925999329, + "grad_norm": 0.15582574903964996, + "learning_rate": 4.4616079419177006e-05, + "loss": 3.2072, + "step": 9452 + }, + { + "epoch": 0.29159163906550634, + "grad_norm": 0.15351371467113495, + "learning_rate": 4.46144104483758e-05, + "loss": 3.1979, + "step": 9453 + }, + { + "epoch": 0.2916224855310797, + "grad_norm": 0.14959649741649628, + "learning_rate": 4.46127412501573e-05, + "loss": 3.2134, + "step": 9454 + }, + { + "epoch": 0.29165333199665316, + "grad_norm": 0.15370531380176544, + "learning_rate": 4.461107182454085e-05, + "loss": 3.234, + "step": 9455 + }, + { + "epoch": 0.29168417846222655, + "grad_norm": 0.1500413417816162, + "learning_rate": 4.460940217154581e-05, + "loss": 3.1855, + "step": 9456 + }, + { + "epoch": 0.2917150249278, + "grad_norm": 0.15559376776218414, + "learning_rate": 4.460773229119153e-05, + "loss": 3.1641, + "step": 9457 + }, + { + "epoch": 0.29174587139337343, + "grad_norm": 0.15381434559822083, + "learning_rate": 4.460606218349738e-05, + "loss": 3.2112, + "step": 9458 + }, + { + "epoch": 0.2917767178589468, + "grad_norm": 0.15020400285720825, + "learning_rate": 4.460439184848272e-05, + "loss": 3.1799, + "step": 9459 + }, + { + "epoch": 0.29180756432452026, + "grad_norm": 0.14930827915668488, + "learning_rate": 4.460272128616691e-05, + "loss": 3.179, + "step": 9460 + }, + { + "epoch": 0.29183841079009365, + "grad_norm": 0.14782999455928802, + "learning_rate": 4.4601050496569344e-05, + "loss": 3.2242, + "step": 9461 + }, + { + "epoch": 0.2918692572556671, + "grad_norm": 0.15303777158260345, + "learning_rate": 4.459937947970936e-05, + "loss": 3.2077, + "step": 9462 + }, + { + "epoch": 0.2919001037212405, + "grad_norm": 0.14805707335472107, + "learning_rate": 4.459770823560636e-05, + "loss": 3.1862, + "step": 9463 + }, + { + "epoch": 0.2919309501868139, + "grad_norm": 0.14888451993465424, + "learning_rate": 4.45960367642797e-05, + "loss": 3.2278, + "step": 9464 + }, + { + "epoch": 0.2919617966523873, + "grad_norm": 0.14765797555446625, + "learning_rate": 4.459436506574877e-05, + "loss": 3.1754, + "step": 9465 + }, + { + "epoch": 0.29199264311796075, + "grad_norm": 0.1492813527584076, + "learning_rate": 4.459269314003296e-05, + "loss": 3.1612, + "step": 9466 + }, + { + "epoch": 0.29202348958353413, + "grad_norm": 0.15844669938087463, + "learning_rate": 4.4591020987151646e-05, + "loss": 3.2046, + "step": 9467 + }, + { + "epoch": 0.2920543360491076, + "grad_norm": 0.14515820145606995, + "learning_rate": 4.458934860712421e-05, + "loss": 3.1829, + "step": 9468 + }, + { + "epoch": 0.29208518251468096, + "grad_norm": 0.14790669083595276, + "learning_rate": 4.458767599997005e-05, + "loss": 3.2153, + "step": 9469 + }, + { + "epoch": 0.2921160289802544, + "grad_norm": 0.1503608673810959, + "learning_rate": 4.4586003165708563e-05, + "loss": 3.193, + "step": 9470 + }, + { + "epoch": 0.29214687544582785, + "grad_norm": 0.15484483540058136, + "learning_rate": 4.4584330104359136e-05, + "loss": 3.2436, + "step": 9471 + }, + { + "epoch": 0.29217772191140123, + "grad_norm": 0.6781267523765564, + "learning_rate": 4.458265681594117e-05, + "loss": 3.1611, + "step": 9472 + }, + { + "epoch": 0.2922085683769747, + "grad_norm": 0.15409447252750397, + "learning_rate": 4.458098330047407e-05, + "loss": 3.1952, + "step": 9473 + }, + { + "epoch": 0.29223941484254806, + "grad_norm": 0.1491982787847519, + "learning_rate": 4.457930955797723e-05, + "loss": 3.2002, + "step": 9474 + }, + { + "epoch": 0.2922702613081215, + "grad_norm": 0.15125679969787598, + "learning_rate": 4.4577635588470056e-05, + "loss": 3.2057, + "step": 9475 + }, + { + "epoch": 0.2923011077736949, + "grad_norm": 0.15212203562259674, + "learning_rate": 4.457596139197198e-05, + "loss": 3.2002, + "step": 9476 + }, + { + "epoch": 0.29233195423926833, + "grad_norm": 0.15256313979625702, + "learning_rate": 4.457428696850238e-05, + "loss": 3.1899, + "step": 9477 + }, + { + "epoch": 0.2923628007048417, + "grad_norm": 0.1502588987350464, + "learning_rate": 4.4572612318080686e-05, + "loss": 3.1747, + "step": 9478 + }, + { + "epoch": 0.29239364717041516, + "grad_norm": 0.1507606953382492, + "learning_rate": 4.4570937440726325e-05, + "loss": 3.1893, + "step": 9479 + }, + { + "epoch": 0.29242449363598855, + "grad_norm": 0.163375586271286, + "learning_rate": 4.4569262336458695e-05, + "loss": 3.1907, + "step": 9480 + }, + { + "epoch": 0.292455340101562, + "grad_norm": 0.15906551480293274, + "learning_rate": 4.456758700529723e-05, + "loss": 3.1635, + "step": 9481 + }, + { + "epoch": 0.29248618656713543, + "grad_norm": 0.15154936909675598, + "learning_rate": 4.456591144726136e-05, + "loss": 3.1991, + "step": 9482 + }, + { + "epoch": 0.2925170330327088, + "grad_norm": 0.23972754180431366, + "learning_rate": 4.456423566237049e-05, + "loss": 3.1936, + "step": 9483 + }, + { + "epoch": 0.29254787949828226, + "grad_norm": 0.15688857436180115, + "learning_rate": 4.4562559650644076e-05, + "loss": 3.2103, + "step": 9484 + }, + { + "epoch": 0.29257872596385565, + "grad_norm": 0.15655717253684998, + "learning_rate": 4.4560883412101535e-05, + "loss": 3.186, + "step": 9485 + }, + { + "epoch": 0.2926095724294291, + "grad_norm": 0.1552780717611313, + "learning_rate": 4.4559206946762296e-05, + "loss": 3.1967, + "step": 9486 + }, + { + "epoch": 0.2926404188950025, + "grad_norm": 0.15236754715442657, + "learning_rate": 4.4557530254645816e-05, + "loss": 3.2067, + "step": 9487 + }, + { + "epoch": 0.2926712653605759, + "grad_norm": 0.16064129769802094, + "learning_rate": 4.4555853335771534e-05, + "loss": 3.1901, + "step": 9488 + }, + { + "epoch": 0.2927021118261493, + "grad_norm": 0.14829504489898682, + "learning_rate": 4.455417619015887e-05, + "loss": 3.1971, + "step": 9489 + }, + { + "epoch": 0.29273295829172274, + "grad_norm": 0.14737044274806976, + "learning_rate": 4.455249881782729e-05, + "loss": 3.1844, + "step": 9490 + }, + { + "epoch": 0.29276380475729613, + "grad_norm": 0.14735543727874756, + "learning_rate": 4.455082121879624e-05, + "loss": 3.2158, + "step": 9491 + }, + { + "epoch": 0.2927946512228696, + "grad_norm": 0.15830855071544647, + "learning_rate": 4.4549143393085156e-05, + "loss": 3.2075, + "step": 9492 + }, + { + "epoch": 0.29282549768844296, + "grad_norm": 0.15003496408462524, + "learning_rate": 4.454746534071351e-05, + "loss": 3.2062, + "step": 9493 + }, + { + "epoch": 0.2928563441540164, + "grad_norm": 0.14638657867908478, + "learning_rate": 4.454578706170075e-05, + "loss": 3.178, + "step": 9494 + }, + { + "epoch": 0.29288719061958984, + "grad_norm": 0.16389037668704987, + "learning_rate": 4.454410855606633e-05, + "loss": 3.1773, + "step": 9495 + }, + { + "epoch": 0.29291803708516323, + "grad_norm": 0.16448509693145752, + "learning_rate": 4.4542429823829715e-05, + "loss": 3.1786, + "step": 9496 + }, + { + "epoch": 0.29294888355073667, + "grad_norm": 0.15114407241344452, + "learning_rate": 4.454075086501038e-05, + "loss": 3.2231, + "step": 9497 + }, + { + "epoch": 0.29297973001631006, + "grad_norm": 0.1508348435163498, + "learning_rate": 4.453907167962778e-05, + "loss": 3.2153, + "step": 9498 + }, + { + "epoch": 0.2930105764818835, + "grad_norm": 0.14739283919334412, + "learning_rate": 4.453739226770137e-05, + "loss": 3.2243, + "step": 9499 + }, + { + "epoch": 0.2930414229474569, + "grad_norm": 0.14635825157165527, + "learning_rate": 4.453571262925065e-05, + "loss": 3.2163, + "step": 9500 + }, + { + "epoch": 3.084646557341652e-05, + "grad_norm": 0.14976362884044647, + "learning_rate": 4.453403276429507e-05, + "loss": 3.2216, + "step": 9501 + }, + { + "epoch": 6.169293114683304e-05, + "grad_norm": 0.15046991407871246, + "learning_rate": 4.4532352672854136e-05, + "loss": 3.1621, + "step": 9502 + }, + { + "epoch": 9.253939672024955e-05, + "grad_norm": 0.14611366391181946, + "learning_rate": 4.45306723549473e-05, + "loss": 3.1716, + "step": 9503 + }, + { + "epoch": 0.00012338586229366608, + "grad_norm": 0.14780190587043762, + "learning_rate": 4.452899181059406e-05, + "loss": 3.1491, + "step": 9504 + }, + { + "epoch": 0.00015423232786708257, + "grad_norm": 0.14668314158916473, + "learning_rate": 4.452731103981389e-05, + "loss": 3.1836, + "step": 9505 + }, + { + "epoch": 0.0001850787934404991, + "grad_norm": 0.14298176765441895, + "learning_rate": 4.4525630042626284e-05, + "loss": 3.2357, + "step": 9506 + }, + { + "epoch": 0.00021592525901391563, + "grad_norm": 0.1461256593465805, + "learning_rate": 4.452394881905073e-05, + "loss": 3.179, + "step": 9507 + }, + { + "epoch": 0.00024677172458733215, + "grad_norm": 0.14447714388370514, + "learning_rate": 4.452226736910673e-05, + "loss": 3.2001, + "step": 9508 + }, + { + "epoch": 0.0002776181901607486, + "grad_norm": 0.1459803432226181, + "learning_rate": 4.452058569281377e-05, + "loss": 3.1663, + "step": 9509 + }, + { + "epoch": 0.00030846465573416515, + "grad_norm": 0.14628803730010986, + "learning_rate": 4.451890379019134e-05, + "loss": 3.2139, + "step": 9510 + }, + { + "epoch": 0.0003393111213075817, + "grad_norm": 0.1484789252281189, + "learning_rate": 4.451722166125895e-05, + "loss": 3.2028, + "step": 9511 + }, + { + "epoch": 0.0003701575868809982, + "grad_norm": 0.14878231287002563, + "learning_rate": 4.451553930603611e-05, + "loss": 3.1909, + "step": 9512 + }, + { + "epoch": 0.0004010040524544147, + "grad_norm": 0.15265117585659027, + "learning_rate": 4.451385672454232e-05, + "loss": 3.2254, + "step": 9513 + }, + { + "epoch": 0.00043185051802783125, + "grad_norm": 0.14584046602249146, + "learning_rate": 4.4512173916797085e-05, + "loss": 3.1782, + "step": 9514 + }, + { + "epoch": 0.0004626969836012477, + "grad_norm": 0.16618448495864868, + "learning_rate": 4.451049088281992e-05, + "loss": 3.1966, + "step": 9515 + }, + { + "epoch": 0.0004935434491746643, + "grad_norm": 0.14405973255634308, + "learning_rate": 4.4508807622630334e-05, + "loss": 3.1794, + "step": 9516 + }, + { + "epoch": 0.0005243899147480808, + "grad_norm": 0.14851243793964386, + "learning_rate": 4.450712413624785e-05, + "loss": 3.2092, + "step": 9517 + }, + { + "epoch": 0.0005552363803214972, + "grad_norm": 0.14587368071079254, + "learning_rate": 4.4505440423691984e-05, + "loss": 3.1921, + "step": 9518 + }, + { + "epoch": 0.0005860828458949138, + "grad_norm": 0.15253448486328125, + "learning_rate": 4.4503756484982265e-05, + "loss": 3.2065, + "step": 9519 + }, + { + "epoch": 0.0006169293114683303, + "grad_norm": 0.14939557015895844, + "learning_rate": 4.45020723201382e-05, + "loss": 3.2228, + "step": 9520 + }, + { + "epoch": 0.0006477757770417469, + "grad_norm": 0.1482362300157547, + "learning_rate": 4.4500387929179323e-05, + "loss": 3.189, + "step": 9521 + }, + { + "epoch": 0.0006786222426151633, + "grad_norm": 0.15925435721874237, + "learning_rate": 4.449870331212517e-05, + "loss": 3.235, + "step": 9522 + }, + { + "epoch": 0.0007094687081885798, + "grad_norm": 0.14296773076057434, + "learning_rate": 4.449701846899528e-05, + "loss": 3.1967, + "step": 9523 + }, + { + "epoch": 0.0007403151737619964, + "grad_norm": 0.14382599294185638, + "learning_rate": 4.4495333399809164e-05, + "loss": 3.1648, + "step": 9524 + }, + { + "epoch": 0.0007711616393354129, + "grad_norm": 0.15028826892375946, + "learning_rate": 4.449364810458637e-05, + "loss": 3.1809, + "step": 9525 + }, + { + "epoch": 0.0008020081049088294, + "grad_norm": 0.15004971623420715, + "learning_rate": 4.449196258334646e-05, + "loss": 3.1893, + "step": 9526 + }, + { + "epoch": 0.0008328545704822459, + "grad_norm": 0.14541654288768768, + "learning_rate": 4.449027683610894e-05, + "loss": 3.1634, + "step": 9527 + }, + { + "epoch": 0.0008637010360556625, + "grad_norm": 0.14775657653808594, + "learning_rate": 4.4488590862893374e-05, + "loss": 3.2284, + "step": 9528 + }, + { + "epoch": 0.000894547501629079, + "grad_norm": 0.1451207995414734, + "learning_rate": 4.448690466371932e-05, + "loss": 3.2334, + "step": 9529 + }, + { + "epoch": 0.0009253939672024954, + "grad_norm": 0.14722919464111328, + "learning_rate": 4.4485218238606296e-05, + "loss": 3.1997, + "step": 9530 + }, + { + "epoch": 0.000956240432775912, + "grad_norm": 0.14939375221729279, + "learning_rate": 4.448353158757389e-05, + "loss": 3.182, + "step": 9531 + }, + { + "epoch": 0.0009870868983493286, + "grad_norm": 0.1462886482477188, + "learning_rate": 4.448184471064165e-05, + "loss": 3.2063, + "step": 9532 + }, + { + "epoch": 0.001017933363922745, + "grad_norm": 0.1464189887046814, + "learning_rate": 4.448015760782911e-05, + "loss": 3.1922, + "step": 9533 + }, + { + "epoch": 0.0010487798294961615, + "grad_norm": 0.15543191134929657, + "learning_rate": 4.447847027915586e-05, + "loss": 3.2211, + "step": 9534 + }, + { + "epoch": 0.0010796262950695781, + "grad_norm": 0.24357016384601593, + "learning_rate": 4.4476782724641455e-05, + "loss": 3.1591, + "step": 9535 + }, + { + "epoch": 0.0011104727606429945, + "grad_norm": 0.14802850782871246, + "learning_rate": 4.447509494430545e-05, + "loss": 3.1689, + "step": 9536 + }, + { + "epoch": 0.001141319226216411, + "grad_norm": 0.14711792767047882, + "learning_rate": 4.447340693816743e-05, + "loss": 3.1966, + "step": 9537 + }, + { + "epoch": 0.0011721656917898276, + "grad_norm": 0.1497952938079834, + "learning_rate": 4.447171870624695e-05, + "loss": 3.1814, + "step": 9538 + }, + { + "epoch": 0.0012030121573632442, + "grad_norm": 0.15760329365730286, + "learning_rate": 4.44700302485636e-05, + "loss": 3.201, + "step": 9539 + }, + { + "epoch": 0.0012338586229366606, + "grad_norm": 0.15420536696910858, + "learning_rate": 4.446834156513695e-05, + "loss": 3.1925, + "step": 9540 + }, + { + "epoch": 0.0012647050885100772, + "grad_norm": 0.1477322280406952, + "learning_rate": 4.446665265598657e-05, + "loss": 3.2359, + "step": 9541 + }, + { + "epoch": 0.0012955515540834938, + "grad_norm": 0.14771223068237305, + "learning_rate": 4.4464963521132055e-05, + "loss": 3.171, + "step": 9542 + }, + { + "epoch": 0.0013263980196569101, + "grad_norm": 0.1525636464357376, + "learning_rate": 4.446327416059299e-05, + "loss": 3.1813, + "step": 9543 + }, + { + "epoch": 0.0013572444852303267, + "grad_norm": 0.15574850142002106, + "learning_rate": 4.446158457438895e-05, + "loss": 3.1943, + "step": 9544 + }, + { + "epoch": 0.0013880909508037433, + "grad_norm": 0.15952672064304352, + "learning_rate": 4.4459894762539535e-05, + "loss": 3.1802, + "step": 9545 + }, + { + "epoch": 0.0014189374163771596, + "grad_norm": 0.1510176956653595, + "learning_rate": 4.4458204725064335e-05, + "loss": 3.1882, + "step": 9546 + }, + { + "epoch": 0.0014497838819505762, + "grad_norm": 0.15518996119499207, + "learning_rate": 4.4456514461982945e-05, + "loss": 3.2051, + "step": 9547 + }, + { + "epoch": 0.0014806303475239928, + "grad_norm": 0.15284240245819092, + "learning_rate": 4.4454823973314965e-05, + "loss": 3.1927, + "step": 9548 + }, + { + "epoch": 0.0015114768130974094, + "grad_norm": 0.15223923325538635, + "learning_rate": 4.445313325907998e-05, + "loss": 3.1766, + "step": 9549 + }, + { + "epoch": 0.0015423232786708257, + "grad_norm": 0.14681193232536316, + "learning_rate": 4.445144231929761e-05, + "loss": 3.2272, + "step": 9550 + }, + { + "epoch": 0.0015731697442442423, + "grad_norm": 0.15134665369987488, + "learning_rate": 4.444975115398746e-05, + "loss": 3.1771, + "step": 9551 + }, + { + "epoch": 0.001604016209817659, + "grad_norm": 0.15655680000782013, + "learning_rate": 4.444805976316914e-05, + "loss": 3.195, + "step": 9552 + }, + { + "epoch": 0.0016348626753910753, + "grad_norm": 0.14966794848442078, + "learning_rate": 4.444636814686224e-05, + "loss": 3.1889, + "step": 9553 + }, + { + "epoch": 0.0016657091409644918, + "grad_norm": 0.17978490889072418, + "learning_rate": 4.444467630508639e-05, + "loss": 3.192, + "step": 9554 + }, + { + "epoch": 0.0016965556065379084, + "grad_norm": 0.14848046004772186, + "learning_rate": 4.444298423786121e-05, + "loss": 3.1719, + "step": 9555 + }, + { + "epoch": 0.001727402072111325, + "grad_norm": 0.14903123676776886, + "learning_rate": 4.4441291945206295e-05, + "loss": 3.1739, + "step": 9556 + }, + { + "epoch": 0.0017582485376847414, + "grad_norm": 0.1566578596830368, + "learning_rate": 4.44395994271413e-05, + "loss": 3.1602, + "step": 9557 + }, + { + "epoch": 0.001789095003258158, + "grad_norm": 0.15414494276046753, + "learning_rate": 4.443790668368583e-05, + "loss": 3.1914, + "step": 9558 + }, + { + "epoch": 0.0018199414688315745, + "grad_norm": 0.166900634765625, + "learning_rate": 4.443621371485951e-05, + "loss": 3.1778, + "step": 9559 + }, + { + "epoch": 0.0018507879344049909, + "grad_norm": 0.1547260284423828, + "learning_rate": 4.4434520520681965e-05, + "loss": 3.2156, + "step": 9560 + }, + { + "epoch": 0.0018816343999784075, + "grad_norm": 0.15384769439697266, + "learning_rate": 4.443282710117284e-05, + "loss": 3.1998, + "step": 9561 + }, + { + "epoch": 0.001912480865551824, + "grad_norm": 0.16572882235050201, + "learning_rate": 4.443113345635176e-05, + "loss": 3.199, + "step": 9562 + }, + { + "epoch": 0.0019433273311252404, + "grad_norm": 0.16769905388355255, + "learning_rate": 4.442943958623837e-05, + "loss": 3.2078, + "step": 9563 + }, + { + "epoch": 0.001974173796698657, + "grad_norm": 0.16239489614963531, + "learning_rate": 4.4427745490852304e-05, + "loss": 3.2003, + "step": 9564 + }, + { + "epoch": 0.0020050202622720736, + "grad_norm": 0.16426977515220642, + "learning_rate": 4.442605117021319e-05, + "loss": 3.2087, + "step": 9565 + }, + { + "epoch": 0.00203586672784549, + "grad_norm": 0.15452024340629578, + "learning_rate": 4.44243566243407e-05, + "loss": 3.213, + "step": 9566 + }, + { + "epoch": 0.0020667131934189067, + "grad_norm": 0.1705000400543213, + "learning_rate": 4.442266185325446e-05, + "loss": 3.2035, + "step": 9567 + }, + { + "epoch": 0.002097559658992323, + "grad_norm": 0.16140305995941162, + "learning_rate": 4.442096685697413e-05, + "loss": 3.1995, + "step": 9568 + }, + { + "epoch": 0.0021284061245657395, + "grad_norm": 0.15974625945091248, + "learning_rate": 4.4419271635519365e-05, + "loss": 3.2062, + "step": 9569 + }, + { + "epoch": 0.0021592525901391563, + "grad_norm": 0.1540258228778839, + "learning_rate": 4.4417576188909815e-05, + "loss": 3.1816, + "step": 9570 + }, + { + "epoch": 0.0021900990557125726, + "grad_norm": 0.15210044384002686, + "learning_rate": 4.441588051716513e-05, + "loss": 3.2216, + "step": 9571 + }, + { + "epoch": 0.002220945521285989, + "grad_norm": 0.15188603103160858, + "learning_rate": 4.4414184620304984e-05, + "loss": 3.2168, + "step": 9572 + }, + { + "epoch": 0.0022517919868594058, + "grad_norm": 0.566066324710846, + "learning_rate": 4.4412488498349025e-05, + "loss": 3.173, + "step": 9573 + }, + { + "epoch": 0.002282638452432822, + "grad_norm": 0.15783318877220154, + "learning_rate": 4.441079215131694e-05, + "loss": 3.1766, + "step": 9574 + }, + { + "epoch": 0.0023134849180062385, + "grad_norm": 0.17796504497528076, + "learning_rate": 4.440909557922838e-05, + "loss": 3.1597, + "step": 9575 + }, + { + "epoch": 0.0023443313835796553, + "grad_norm": 0.1523173451423645, + "learning_rate": 4.440739878210302e-05, + "loss": 3.2079, + "step": 9576 + }, + { + "epoch": 0.0023751778491530717, + "grad_norm": 0.1589062213897705, + "learning_rate": 4.4405701759960525e-05, + "loss": 3.189, + "step": 9577 + }, + { + "epoch": 0.0024060243147264885, + "grad_norm": 0.15728266537189484, + "learning_rate": 4.440400451282059e-05, + "loss": 3.1828, + "step": 9578 + }, + { + "epoch": 0.002436870780299905, + "grad_norm": 0.15867361426353455, + "learning_rate": 4.440230704070289e-05, + "loss": 3.1758, + "step": 9579 + }, + { + "epoch": 0.002467717245873321, + "grad_norm": 0.15037474036216736, + "learning_rate": 4.440060934362708e-05, + "loss": 3.1809, + "step": 9580 + }, + { + "epoch": 0.002498563711446738, + "grad_norm": 0.15328183770179749, + "learning_rate": 4.439891142161287e-05, + "loss": 3.2235, + "step": 9581 + }, + { + "epoch": 0.0025294101770201543, + "grad_norm": 0.16748474538326263, + "learning_rate": 4.439721327467994e-05, + "loss": 3.1838, + "step": 9582 + }, + { + "epoch": 0.0025602566425935707, + "grad_norm": 0.14823231101036072, + "learning_rate": 4.439551490284798e-05, + "loss": 3.1931, + "step": 9583 + }, + { + "epoch": 0.0025911031081669875, + "grad_norm": 0.14651349186897278, + "learning_rate": 4.439381630613668e-05, + "loss": 3.2075, + "step": 9584 + }, + { + "epoch": 0.002621949573740404, + "grad_norm": 0.15178966522216797, + "learning_rate": 4.439211748456574e-05, + "loss": 3.187, + "step": 9585 + }, + { + "epoch": 0.0026527960393138202, + "grad_norm": 0.15272389352321625, + "learning_rate": 4.4390418438154847e-05, + "loss": 3.1964, + "step": 9586 + }, + { + "epoch": 0.002683642504887237, + "grad_norm": 0.14831461012363434, + "learning_rate": 4.4388719166923695e-05, + "loss": 3.1924, + "step": 9587 + }, + { + "epoch": 0.0027144889704606534, + "grad_norm": 0.15537485480308533, + "learning_rate": 4.4387019670892004e-05, + "loss": 3.1601, + "step": 9588 + }, + { + "epoch": 0.0027453354360340698, + "grad_norm": 0.14821258187294006, + "learning_rate": 4.4385319950079474e-05, + "loss": 3.1777, + "step": 9589 + }, + { + "epoch": 0.0027761819016074865, + "grad_norm": 0.15665802359580994, + "learning_rate": 4.43836200045058e-05, + "loss": 3.1525, + "step": 9590 + }, + { + "epoch": 0.002807028367180903, + "grad_norm": 0.19506748020648956, + "learning_rate": 4.438191983419071e-05, + "loss": 3.1952, + "step": 9591 + }, + { + "epoch": 0.0028378748327543193, + "grad_norm": 0.14922022819519043, + "learning_rate": 4.4380219439153906e-05, + "loss": 3.1823, + "step": 9592 + }, + { + "epoch": 0.002868721298327736, + "grad_norm": 0.15515105426311493, + "learning_rate": 4.437851881941509e-05, + "loss": 3.2064, + "step": 9593 + }, + { + "epoch": 0.0028995677639011524, + "grad_norm": 0.15572813153266907, + "learning_rate": 4.4376817974994e-05, + "loss": 3.2008, + "step": 9594 + }, + { + "epoch": 0.0029304142294745692, + "grad_norm": 0.17082658410072327, + "learning_rate": 4.4375116905910355e-05, + "loss": 3.2087, + "step": 9595 + }, + { + "epoch": 0.0029612606950479856, + "grad_norm": 0.14694589376449585, + "learning_rate": 4.437341561218387e-05, + "loss": 3.1819, + "step": 9596 + }, + { + "epoch": 0.002992107160621402, + "grad_norm": 0.15920482575893402, + "learning_rate": 4.437171409383427e-05, + "loss": 3.1726, + "step": 9597 + }, + { + "epoch": 0.0030229536261948188, + "grad_norm": 0.15928776562213898, + "learning_rate": 4.4370012350881294e-05, + "loss": 3.2137, + "step": 9598 + }, + { + "epoch": 0.003053800091768235, + "grad_norm": 0.15259508788585663, + "learning_rate": 4.4368310383344656e-05, + "loss": 3.1245, + "step": 9599 + }, + { + "epoch": 0.0030846465573416515, + "grad_norm": 0.1632494181394577, + "learning_rate": 4.43666081912441e-05, + "loss": 3.1985, + "step": 9600 + }, + { + "epoch": 0.0031154930229150683, + "grad_norm": 0.16191212832927704, + "learning_rate": 4.4364905774599365e-05, + "loss": 3.2228, + "step": 9601 + }, + { + "epoch": 0.0031463394884884846, + "grad_norm": 0.15751530230045319, + "learning_rate": 4.4363203133430184e-05, + "loss": 3.197, + "step": 9602 + }, + { + "epoch": 0.003177185954061901, + "grad_norm": 0.15522250533103943, + "learning_rate": 4.4361500267756295e-05, + "loss": 3.187, + "step": 9603 + }, + { + "epoch": 0.003208032419635318, + "grad_norm": 0.15772660076618195, + "learning_rate": 4.4359797177597446e-05, + "loss": 3.2009, + "step": 9604 + }, + { + "epoch": 0.003238878885208734, + "grad_norm": 0.15080879628658295, + "learning_rate": 4.435809386297338e-05, + "loss": 3.2016, + "step": 9605 + }, + { + "epoch": 0.0032697253507821505, + "grad_norm": 0.17637848854064941, + "learning_rate": 4.4356390323903854e-05, + "loss": 3.2099, + "step": 9606 + }, + { + "epoch": 0.0033005718163555673, + "grad_norm": 0.151162788271904, + "learning_rate": 4.435468656040861e-05, + "loss": 3.1876, + "step": 9607 + }, + { + "epoch": 0.0033314182819289837, + "grad_norm": 0.15080702304840088, + "learning_rate": 4.4352982572507405e-05, + "loss": 3.2016, + "step": 9608 + }, + { + "epoch": 0.0033622647475024, + "grad_norm": 0.158397376537323, + "learning_rate": 4.435127836022001e-05, + "loss": 3.1861, + "step": 9609 + }, + { + "epoch": 0.003393111213075817, + "grad_norm": 0.15517708659172058, + "learning_rate": 4.4349573923566154e-05, + "loss": 3.244, + "step": 9610 + }, + { + "epoch": 0.003423957678649233, + "grad_norm": 0.14959760010242462, + "learning_rate": 4.434786926256562e-05, + "loss": 3.1909, + "step": 9611 + }, + { + "epoch": 0.00345480414422265, + "grad_norm": 0.15918323397636414, + "learning_rate": 4.434616437723817e-05, + "loss": 3.2172, + "step": 9612 + }, + { + "epoch": 0.0034856506097960664, + "grad_norm": 0.1926119476556778, + "learning_rate": 4.4344459267603576e-05, + "loss": 3.234, + "step": 9613 + }, + { + "epoch": 0.0035164970753694827, + "grad_norm": 0.15368768572807312, + "learning_rate": 4.434275393368159e-05, + "loss": 3.193, + "step": 9614 + }, + { + "epoch": 0.0035473435409428995, + "grad_norm": 0.15261447429656982, + "learning_rate": 4.434104837549201e-05, + "loss": 3.1927, + "step": 9615 + }, + { + "epoch": 0.003578190006516316, + "grad_norm": 0.16707895696163177, + "learning_rate": 4.433934259305459e-05, + "loss": 3.1957, + "step": 9616 + }, + { + "epoch": 0.0036090364720897323, + "grad_norm": 0.15798193216323853, + "learning_rate": 4.433763658638911e-05, + "loss": 3.1913, + "step": 9617 + }, + { + "epoch": 0.003639882937663149, + "grad_norm": 0.15128672122955322, + "learning_rate": 4.433593035551535e-05, + "loss": 3.1837, + "step": 9618 + }, + { + "epoch": 0.0036707294032365654, + "grad_norm": 0.15994873642921448, + "learning_rate": 4.433422390045311e-05, + "loss": 3.1893, + "step": 9619 + }, + { + "epoch": 0.0037015758688099818, + "grad_norm": 0.1478361338376999, + "learning_rate": 4.433251722122216e-05, + "loss": 3.1901, + "step": 9620 + }, + { + "epoch": 0.0037324223343833986, + "grad_norm": 0.15659984946250916, + "learning_rate": 4.433081031784229e-05, + "loss": 3.1888, + "step": 9621 + }, + { + "epoch": 0.003763268799956815, + "grad_norm": 0.1705477386713028, + "learning_rate": 4.432910319033328e-05, + "loss": 3.2169, + "step": 9622 + }, + { + "epoch": 0.0037941152655302313, + "grad_norm": 0.2048177421092987, + "learning_rate": 4.4327395838714946e-05, + "loss": 3.2131, + "step": 9623 + }, + { + "epoch": 0.003824961731103648, + "grad_norm": 0.16327659785747528, + "learning_rate": 4.432568826300707e-05, + "loss": 3.2332, + "step": 9624 + }, + { + "epoch": 0.0038558081966770645, + "grad_norm": 0.1544979214668274, + "learning_rate": 4.432398046322944e-05, + "loss": 3.1905, + "step": 9625 + }, + { + "epoch": 0.003886654662250481, + "grad_norm": 0.15373282134532928, + "learning_rate": 4.432227243940188e-05, + "loss": 3.1951, + "step": 9626 + }, + { + "epoch": 0.003917501127823898, + "grad_norm": 0.1536756455898285, + "learning_rate": 4.432056419154419e-05, + "loss": 3.1852, + "step": 9627 + }, + { + "epoch": 0.003948347593397314, + "grad_norm": 0.16834330558776855, + "learning_rate": 4.4318855719676154e-05, + "loss": 3.1789, + "step": 9628 + }, + { + "epoch": 0.00397919405897073, + "grad_norm": 0.15993379056453705, + "learning_rate": 4.431714702381761e-05, + "loss": 3.1808, + "step": 9629 + }, + { + "epoch": 0.004010040524544147, + "grad_norm": 0.15790876746177673, + "learning_rate": 4.431543810398834e-05, + "loss": 3.227, + "step": 9630 + }, + { + "epoch": 0.004040886990117564, + "grad_norm": 0.15569108724594116, + "learning_rate": 4.4313728960208176e-05, + "loss": 3.21, + "step": 9631 + }, + { + "epoch": 0.00407173345569098, + "grad_norm": 0.14982794225215912, + "learning_rate": 4.4312019592496935e-05, + "loss": 3.1985, + "step": 9632 + }, + { + "epoch": 0.004102579921264397, + "grad_norm": 0.1550399661064148, + "learning_rate": 4.4310310000874425e-05, + "loss": 3.2254, + "step": 9633 + }, + { + "epoch": 0.0041334263868378135, + "grad_norm": 0.15944316983222961, + "learning_rate": 4.430860018536048e-05, + "loss": 3.2145, + "step": 9634 + }, + { + "epoch": 0.004164272852411229, + "grad_norm": 0.14665178954601288, + "learning_rate": 4.430689014597492e-05, + "loss": 3.2038, + "step": 9635 + }, + { + "epoch": 0.004195119317984646, + "grad_norm": 0.15359511971473694, + "learning_rate": 4.430517988273757e-05, + "loss": 3.2019, + "step": 9636 + }, + { + "epoch": 0.004225965783558063, + "grad_norm": 0.14822423458099365, + "learning_rate": 4.4303469395668255e-05, + "loss": 3.1475, + "step": 9637 + }, + { + "epoch": 0.004256812249131479, + "grad_norm": 0.15572860836982727, + "learning_rate": 4.4301758684786824e-05, + "loss": 3.2214, + "step": 9638 + }, + { + "epoch": 0.004287658714704896, + "grad_norm": 0.15109458565711975, + "learning_rate": 4.430004775011309e-05, + "loss": 3.2153, + "step": 9639 + }, + { + "epoch": 0.0043185051802783125, + "grad_norm": 0.14540573954582214, + "learning_rate": 4.42983365916669e-05, + "loss": 3.1745, + "step": 9640 + }, + { + "epoch": 0.004349351645851728, + "grad_norm": 0.147481307387352, + "learning_rate": 4.4296625209468104e-05, + "loss": 3.1784, + "step": 9641 + }, + { + "epoch": 0.004380198111425145, + "grad_norm": 0.15355660021305084, + "learning_rate": 4.4294913603536525e-05, + "loss": 3.2139, + "step": 9642 + }, + { + "epoch": 0.004411044576998562, + "grad_norm": 0.15636317431926727, + "learning_rate": 4.429320177389202e-05, + "loss": 3.2193, + "step": 9643 + }, + { + "epoch": 0.004441891042571978, + "grad_norm": 0.14524033665657043, + "learning_rate": 4.429148972055443e-05, + "loss": 3.1668, + "step": 9644 + }, + { + "epoch": 0.004472737508145395, + "grad_norm": 0.14915604889392853, + "learning_rate": 4.428977744354362e-05, + "loss": 3.165, + "step": 9645 + }, + { + "epoch": 0.0045035839737188115, + "grad_norm": 0.15099938213825226, + "learning_rate": 4.428806494287942e-05, + "loss": 3.2098, + "step": 9646 + }, + { + "epoch": 0.0045344304392922275, + "grad_norm": 0.14689554274082184, + "learning_rate": 4.42863522185817e-05, + "loss": 3.188, + "step": 9647 + }, + { + "epoch": 0.004565276904865644, + "grad_norm": 0.14495165646076202, + "learning_rate": 4.4284639270670326e-05, + "loss": 3.1795, + "step": 9648 + }, + { + "epoch": 0.004596123370439061, + "grad_norm": 0.15187180042266846, + "learning_rate": 4.428292609916514e-05, + "loss": 3.2099, + "step": 9649 + }, + { + "epoch": 0.004626969836012477, + "grad_norm": 0.14844506978988647, + "learning_rate": 4.428121270408602e-05, + "loss": 3.1874, + "step": 9650 + }, + { + "epoch": 0.004657816301585894, + "grad_norm": 0.16075162589550018, + "learning_rate": 4.427949908545282e-05, + "loss": 3.2118, + "step": 9651 + }, + { + "epoch": 0.004688662767159311, + "grad_norm": 0.1458684802055359, + "learning_rate": 4.427778524328542e-05, + "loss": 3.2229, + "step": 9652 + }, + { + "epoch": 0.0047195092327327265, + "grad_norm": 0.15335123240947723, + "learning_rate": 4.427607117760368e-05, + "loss": 3.1834, + "step": 9653 + }, + { + "epoch": 0.004750355698306143, + "grad_norm": 0.16287894546985626, + "learning_rate": 4.427435688842748e-05, + "loss": 3.1679, + "step": 9654 + }, + { + "epoch": 0.00478120216387956, + "grad_norm": 0.16354499757289886, + "learning_rate": 4.427264237577668e-05, + "loss": 3.1944, + "step": 9655 + }, + { + "epoch": 0.004812048629452977, + "grad_norm": 0.15121491253376007, + "learning_rate": 4.4270927639671195e-05, + "loss": 3.1988, + "step": 9656 + }, + { + "epoch": 0.004842895095026393, + "grad_norm": 0.14894698560237885, + "learning_rate": 4.426921268013088e-05, + "loss": 3.2025, + "step": 9657 + }, + { + "epoch": 0.00487374156059981, + "grad_norm": 0.15849372744560242, + "learning_rate": 4.426749749717561e-05, + "loss": 3.2029, + "step": 9658 + }, + { + "epoch": 0.0049045880261732264, + "grad_norm": 0.1608913242816925, + "learning_rate": 4.42657820908253e-05, + "loss": 3.226, + "step": 9659 + }, + { + "epoch": 0.004935434491746642, + "grad_norm": 0.1553320586681366, + "learning_rate": 4.4264066461099816e-05, + "loss": 3.2004, + "step": 9660 + }, + { + "epoch": 0.004966280957320059, + "grad_norm": 0.15581662952899933, + "learning_rate": 4.426235060801906e-05, + "loss": 3.1974, + "step": 9661 + }, + { + "epoch": 0.004997127422893476, + "grad_norm": 0.17718295753002167, + "learning_rate": 4.426063453160293e-05, + "loss": 3.2212, + "step": 9662 + }, + { + "epoch": 0.005027973888466892, + "grad_norm": 0.15454886853694916, + "learning_rate": 4.4258918231871305e-05, + "loss": 3.1796, + "step": 9663 + }, + { + "epoch": 0.005058820354040309, + "grad_norm": 0.16867980360984802, + "learning_rate": 4.42572017088441e-05, + "loss": 3.2076, + "step": 9664 + }, + { + "epoch": 0.0050896668196137255, + "grad_norm": 0.15796226263046265, + "learning_rate": 4.425548496254122e-05, + "loss": 3.2014, + "step": 9665 + }, + { + "epoch": 0.005120513285187141, + "grad_norm": 0.17452013492584229, + "learning_rate": 4.4253767992982554e-05, + "loss": 3.2055, + "step": 9666 + }, + { + "epoch": 0.005151359750760558, + "grad_norm": 0.14474062621593475, + "learning_rate": 4.425205080018802e-05, + "loss": 3.2099, + "step": 9667 + }, + { + "epoch": 0.005182206216333975, + "grad_norm": 0.1661698967218399, + "learning_rate": 4.4250333384177535e-05, + "loss": 3.1821, + "step": 9668 + }, + { + "epoch": 0.005213052681907391, + "grad_norm": 0.15512943267822266, + "learning_rate": 4.424861574497099e-05, + "loss": 3.1695, + "step": 9669 + }, + { + "epoch": 0.005243899147480808, + "grad_norm": 0.15047626197338104, + "learning_rate": 4.4246897882588324e-05, + "loss": 3.1807, + "step": 9670 + }, + { + "epoch": 0.0052747456130542245, + "grad_norm": 0.180130273103714, + "learning_rate": 4.4245179797049435e-05, + "loss": 3.2121, + "step": 9671 + }, + { + "epoch": 0.0053055920786276405, + "grad_norm": 0.1532960832118988, + "learning_rate": 4.424346148837426e-05, + "loss": 3.168, + "step": 9672 + }, + { + "epoch": 0.005336438544201057, + "grad_norm": 0.16171926259994507, + "learning_rate": 4.42417429565827e-05, + "loss": 3.2172, + "step": 9673 + }, + { + "epoch": 0.005367285009774474, + "grad_norm": 0.15362626314163208, + "learning_rate": 4.42400242016947e-05, + "loss": 3.1792, + "step": 9674 + }, + { + "epoch": 0.00539813147534789, + "grad_norm": 0.16848857700824738, + "learning_rate": 4.423830522373018e-05, + "loss": 3.2, + "step": 9675 + }, + { + "epoch": 0.005428977940921307, + "grad_norm": 0.1583082377910614, + "learning_rate": 4.423658602270908e-05, + "loss": 3.2005, + "step": 9676 + }, + { + "epoch": 0.005459824406494724, + "grad_norm": 0.16087177395820618, + "learning_rate": 4.4234866598651316e-05, + "loss": 3.2079, + "step": 9677 + }, + { + "epoch": 0.0054906708720681395, + "grad_norm": 0.15253198146820068, + "learning_rate": 4.423314695157683e-05, + "loss": 3.1822, + "step": 9678 + }, + { + "epoch": 0.005521517337641556, + "grad_norm": 0.17214909195899963, + "learning_rate": 4.4231427081505575e-05, + "loss": 3.2045, + "step": 9679 + }, + { + "epoch": 0.005552363803214973, + "grad_norm": 0.14934109151363373, + "learning_rate": 4.422970698845747e-05, + "loss": 3.2268, + "step": 9680 + }, + { + "epoch": 0.005583210268788389, + "grad_norm": 0.15787357091903687, + "learning_rate": 4.422798667245247e-05, + "loss": 3.2069, + "step": 9681 + }, + { + "epoch": 0.005614056734361806, + "grad_norm": 0.15986502170562744, + "learning_rate": 4.422626613351052e-05, + "loss": 3.2298, + "step": 9682 + }, + { + "epoch": 0.005644903199935223, + "grad_norm": 0.15432004630565643, + "learning_rate": 4.422454537165157e-05, + "loss": 3.1843, + "step": 9683 + }, + { + "epoch": 0.0056757496655086385, + "grad_norm": 0.14931388199329376, + "learning_rate": 4.4222824386895565e-05, + "loss": 3.1834, + "step": 9684 + }, + { + "epoch": 0.005706596131082055, + "grad_norm": 0.15634548664093018, + "learning_rate": 4.4221103179262466e-05, + "loss": 3.2042, + "step": 9685 + }, + { + "epoch": 0.005737442596655472, + "grad_norm": 0.15200021862983704, + "learning_rate": 4.4219381748772226e-05, + "loss": 3.18, + "step": 9686 + }, + { + "epoch": 0.005768289062228888, + "grad_norm": 0.16793349385261536, + "learning_rate": 4.42176600954448e-05, + "loss": 3.2201, + "step": 9687 + }, + { + "epoch": 0.005799135527802305, + "grad_norm": 0.1558707058429718, + "learning_rate": 4.421593821930016e-05, + "loss": 3.1876, + "step": 9688 + }, + { + "epoch": 0.005829981993375722, + "grad_norm": 0.15435375273227692, + "learning_rate": 4.421421612035826e-05, + "loss": 3.1777, + "step": 9689 + }, + { + "epoch": 0.0058608284589491385, + "grad_norm": 0.15401573479175568, + "learning_rate": 4.421249379863907e-05, + "loss": 3.155, + "step": 9690 + }, + { + "epoch": 0.005891674924522554, + "grad_norm": 0.14740988612174988, + "learning_rate": 4.4210771254162564e-05, + "loss": 3.1721, + "step": 9691 + }, + { + "epoch": 0.005922521390095971, + "grad_norm": 0.15258143842220306, + "learning_rate": 4.42090484869487e-05, + "loss": 3.1578, + "step": 9692 + }, + { + "epoch": 0.005953367855669388, + "grad_norm": 0.1568569540977478, + "learning_rate": 4.4207325497017474e-05, + "loss": 3.184, + "step": 9693 + }, + { + "epoch": 0.005984214321242804, + "grad_norm": 0.15857060253620148, + "learning_rate": 4.420560228438885e-05, + "loss": 3.1896, + "step": 9694 + }, + { + "epoch": 0.006015060786816221, + "grad_norm": 0.14421173930168152, + "learning_rate": 4.42038788490828e-05, + "loss": 3.1905, + "step": 9695 + }, + { + "epoch": 0.0060459072523896375, + "grad_norm": 0.14680889248847961, + "learning_rate": 4.420215519111932e-05, + "loss": 3.1973, + "step": 9696 + }, + { + "epoch": 0.0060767537179630534, + "grad_norm": 0.14790616929531097, + "learning_rate": 4.4200431310518385e-05, + "loss": 3.1739, + "step": 9697 + }, + { + "epoch": 0.00610760018353647, + "grad_norm": 0.1565980166196823, + "learning_rate": 4.41987072073e-05, + "loss": 3.2, + "step": 9698 + }, + { + "epoch": 0.006138446649109887, + "grad_norm": 0.14396759867668152, + "learning_rate": 4.419698288148413e-05, + "loss": 3.1878, + "step": 9699 + }, + { + "epoch": 0.006169293114683303, + "grad_norm": 0.14868199825286865, + "learning_rate": 4.419525833309078e-05, + "loss": 3.1647, + "step": 9700 + }, + { + "epoch": 0.00620013958025672, + "grad_norm": 0.14710502326488495, + "learning_rate": 4.419353356213994e-05, + "loss": 3.162, + "step": 9701 + }, + { + "epoch": 0.0062309860458301366, + "grad_norm": 0.15035022795200348, + "learning_rate": 4.4191808568651627e-05, + "loss": 3.1987, + "step": 9702 + }, + { + "epoch": 0.0062618325114035525, + "grad_norm": 0.1498863846063614, + "learning_rate": 4.419008335264582e-05, + "loss": 3.1799, + "step": 9703 + }, + { + "epoch": 0.006292678976976969, + "grad_norm": 0.1498921662569046, + "learning_rate": 4.418835791414253e-05, + "loss": 3.1785, + "step": 9704 + }, + { + "epoch": 0.006323525442550386, + "grad_norm": 0.15576742589473724, + "learning_rate": 4.418663225316175e-05, + "loss": 3.1895, + "step": 9705 + }, + { + "epoch": 0.006354371908123802, + "grad_norm": 0.15460312366485596, + "learning_rate": 4.418490636972351e-05, + "loss": 3.1687, + "step": 9706 + }, + { + "epoch": 0.006385218373697219, + "grad_norm": 0.14749513566493988, + "learning_rate": 4.4183180263847815e-05, + "loss": 3.1903, + "step": 9707 + }, + { + "epoch": 0.006416064839270636, + "grad_norm": 0.2273758500814438, + "learning_rate": 4.4181453935554655e-05, + "loss": 3.2048, + "step": 9708 + }, + { + "epoch": 0.0064469113048440515, + "grad_norm": 0.14856405556201935, + "learning_rate": 4.417972738486408e-05, + "loss": 3.1786, + "step": 9709 + }, + { + "epoch": 0.006477757770417468, + "grad_norm": 0.1740439236164093, + "learning_rate": 4.417800061179608e-05, + "loss": 3.2028, + "step": 9710 + }, + { + "epoch": 0.006508604235990885, + "grad_norm": 0.1709778904914856, + "learning_rate": 4.417627361637069e-05, + "loss": 3.1579, + "step": 9711 + }, + { + "epoch": 0.006539450701564301, + "grad_norm": 0.148243248462677, + "learning_rate": 4.417454639860793e-05, + "loss": 3.1712, + "step": 9712 + }, + { + "epoch": 0.006570297167137718, + "grad_norm": 0.15241177380084991, + "learning_rate": 4.4172818958527834e-05, + "loss": 3.1773, + "step": 9713 + }, + { + "epoch": 0.006601143632711135, + "grad_norm": 0.18947112560272217, + "learning_rate": 4.4171091296150416e-05, + "loss": 3.2345, + "step": 9714 + }, + { + "epoch": 0.006631990098284551, + "grad_norm": 0.15650935471057892, + "learning_rate": 4.4169363411495724e-05, + "loss": 3.1883, + "step": 9715 + }, + { + "epoch": 0.006662836563857967, + "grad_norm": 0.14956487715244293, + "learning_rate": 4.4167635304583774e-05, + "loss": 3.2139, + "step": 9716 + }, + { + "epoch": 0.006693683029431384, + "grad_norm": 0.14704911410808563, + "learning_rate": 4.416590697543461e-05, + "loss": 3.1619, + "step": 9717 + }, + { + "epoch": 0.0067245294950048, + "grad_norm": 0.15186721086502075, + "learning_rate": 4.416417842406828e-05, + "loss": 3.1849, + "step": 9718 + }, + { + "epoch": 0.006755375960578217, + "grad_norm": 0.15510475635528564, + "learning_rate": 4.416244965050482e-05, + "loss": 3.1967, + "step": 9719 + }, + { + "epoch": 0.006786222426151634, + "grad_norm": 0.15022428333759308, + "learning_rate": 4.416072065476427e-05, + "loss": 3.2025, + "step": 9720 + }, + { + "epoch": 0.00681706889172505, + "grad_norm": 0.15177878737449646, + "learning_rate": 4.4158991436866666e-05, + "loss": 3.2003, + "step": 9721 + }, + { + "epoch": 0.006847915357298466, + "grad_norm": 0.1716996729373932, + "learning_rate": 4.415726199683208e-05, + "loss": 3.2067, + "step": 9722 + }, + { + "epoch": 0.006878761822871883, + "grad_norm": 0.14679521322250366, + "learning_rate": 4.415553233468055e-05, + "loss": 3.1934, + "step": 9723 + }, + { + "epoch": 0.0069096082884453, + "grad_norm": 0.18413183093070984, + "learning_rate": 4.415380245043213e-05, + "loss": 3.1907, + "step": 9724 + }, + { + "epoch": 0.006940454754018716, + "grad_norm": 0.15176640450954437, + "learning_rate": 4.415207234410689e-05, + "loss": 3.1859, + "step": 9725 + }, + { + "epoch": 0.006971301219592133, + "grad_norm": 0.1649112105369568, + "learning_rate": 4.415034201572487e-05, + "loss": 3.179, + "step": 9726 + }, + { + "epoch": 0.0070021476851655495, + "grad_norm": 0.15307246148586273, + "learning_rate": 4.414861146530615e-05, + "loss": 3.1893, + "step": 9727 + }, + { + "epoch": 0.0070329941507389655, + "grad_norm": 0.15551456809043884, + "learning_rate": 4.4146880692870784e-05, + "loss": 3.1921, + "step": 9728 + }, + { + "epoch": 0.007063840616312382, + "grad_norm": 0.15900245308876038, + "learning_rate": 4.4145149698438844e-05, + "loss": 3.1894, + "step": 9729 + }, + { + "epoch": 0.007094687081885799, + "grad_norm": 0.1575513631105423, + "learning_rate": 4.4143418482030395e-05, + "loss": 3.2079, + "step": 9730 + }, + { + "epoch": 0.007125533547459215, + "grad_norm": 0.1509416252374649, + "learning_rate": 4.414168704366551e-05, + "loss": 3.1541, + "step": 9731 + }, + { + "epoch": 0.007156380013032632, + "grad_norm": 0.15777289867401123, + "learning_rate": 4.413995538336427e-05, + "loss": 3.2177, + "step": 9732 + }, + { + "epoch": 0.007187226478606049, + "grad_norm": 0.15098631381988525, + "learning_rate": 4.413822350114676e-05, + "loss": 3.1755, + "step": 9733 + }, + { + "epoch": 0.0072180729441794645, + "grad_norm": 0.1484590768814087, + "learning_rate": 4.413649139703303e-05, + "loss": 3.173, + "step": 9734 + }, + { + "epoch": 0.007248919409752881, + "grad_norm": 0.15287896990776062, + "learning_rate": 4.41347590710432e-05, + "loss": 3.2077, + "step": 9735 + }, + { + "epoch": 0.007279765875326298, + "grad_norm": 0.15565034747123718, + "learning_rate": 4.413302652319733e-05, + "loss": 3.1996, + "step": 9736 + }, + { + "epoch": 0.007310612340899714, + "grad_norm": 0.15482689440250397, + "learning_rate": 4.4131293753515505e-05, + "loss": 3.1846, + "step": 9737 + }, + { + "epoch": 0.007341458806473131, + "grad_norm": 0.1430514007806778, + "learning_rate": 4.412956076201784e-05, + "loss": 3.1918, + "step": 9738 + }, + { + "epoch": 0.007372305272046548, + "grad_norm": 0.15770350396633148, + "learning_rate": 4.4127827548724405e-05, + "loss": 3.1983, + "step": 9739 + }, + { + "epoch": 0.0074031517376199635, + "grad_norm": 0.22073689103126526, + "learning_rate": 4.41260941136553e-05, + "loss": 3.2193, + "step": 9740 + }, + { + "epoch": 0.00743399820319338, + "grad_norm": 0.1590922474861145, + "learning_rate": 4.4124360456830636e-05, + "loss": 3.1975, + "step": 9741 + }, + { + "epoch": 0.007464844668766797, + "grad_norm": 0.1605476289987564, + "learning_rate": 4.412262657827051e-05, + "loss": 3.1922, + "step": 9742 + }, + { + "epoch": 0.007495691134340213, + "grad_norm": 0.1505478024482727, + "learning_rate": 4.4120892477995015e-05, + "loss": 3.1709, + "step": 9743 + }, + { + "epoch": 0.00752653759991363, + "grad_norm": 0.15907344222068787, + "learning_rate": 4.4119158156024256e-05, + "loss": 3.1756, + "step": 9744 + }, + { + "epoch": 0.007557384065487047, + "grad_norm": 0.15226569771766663, + "learning_rate": 4.411742361237835e-05, + "loss": 3.177, + "step": 9745 + }, + { + "epoch": 0.007588230531060463, + "grad_norm": 0.15865443646907806, + "learning_rate": 4.411568884707741e-05, + "loss": 3.1499, + "step": 9746 + }, + { + "epoch": 0.007619076996633879, + "grad_norm": 0.152632474899292, + "learning_rate": 4.4113953860141544e-05, + "loss": 3.213, + "step": 9747 + }, + { + "epoch": 0.007649923462207296, + "grad_norm": 0.16804306209087372, + "learning_rate": 4.4112218651590866e-05, + "loss": 3.2146, + "step": 9748 + }, + { + "epoch": 0.007680769927780712, + "grad_norm": 0.16031868755817413, + "learning_rate": 4.41104832214455e-05, + "loss": 3.1898, + "step": 9749 + }, + { + "epoch": 0.007711616393354129, + "grad_norm": 0.16311122477054596, + "learning_rate": 4.4108747569725563e-05, + "loss": 3.1945, + "step": 9750 + }, + { + "epoch": 0.007742462858927546, + "grad_norm": 0.15576501190662384, + "learning_rate": 4.410701169645118e-05, + "loss": 3.1948, + "step": 9751 + }, + { + "epoch": 0.007773309324500962, + "grad_norm": 0.15559446811676025, + "learning_rate": 4.410527560164248e-05, + "loss": 3.1881, + "step": 9752 + }, + { + "epoch": 0.0078041557900743784, + "grad_norm": 0.15022413432598114, + "learning_rate": 4.410353928531959e-05, + "loss": 3.1942, + "step": 9753 + }, + { + "epoch": 0.007835002255647795, + "grad_norm": 0.1793787032365799, + "learning_rate": 4.410180274750264e-05, + "loss": 3.1819, + "step": 9754 + }, + { + "epoch": 0.007865848721221211, + "grad_norm": 0.1553964763879776, + "learning_rate": 4.410006598821177e-05, + "loss": 3.203, + "step": 9755 + }, + { + "epoch": 0.007896695186794629, + "grad_norm": 0.16162823140621185, + "learning_rate": 4.409832900746711e-05, + "loss": 3.1926, + "step": 9756 + }, + { + "epoch": 0.007927541652368045, + "grad_norm": 0.1462961733341217, + "learning_rate": 4.409659180528881e-05, + "loss": 3.2033, + "step": 9757 + }, + { + "epoch": 0.00795838811794146, + "grad_norm": 0.1514253169298172, + "learning_rate": 4.409485438169699e-05, + "loss": 3.2045, + "step": 9758 + }, + { + "epoch": 0.007989234583514878, + "grad_norm": 0.14858798682689667, + "learning_rate": 4.4093116736711815e-05, + "loss": 3.1692, + "step": 9759 + }, + { + "epoch": 0.008020081049088294, + "grad_norm": 0.14779072999954224, + "learning_rate": 4.409137887035343e-05, + "loss": 3.1789, + "step": 9760 + }, + { + "epoch": 0.00805092751466171, + "grad_norm": 0.1519584357738495, + "learning_rate": 4.408964078264197e-05, + "loss": 3.1803, + "step": 9761 + }, + { + "epoch": 0.008081773980235128, + "grad_norm": 0.1567167490720749, + "learning_rate": 4.40879024735976e-05, + "loss": 3.2071, + "step": 9762 + }, + { + "epoch": 0.008112620445808544, + "grad_norm": 0.15370003879070282, + "learning_rate": 4.408616394324048e-05, + "loss": 3.17, + "step": 9763 + }, + { + "epoch": 0.00814346691138196, + "grad_norm": 0.15790975093841553, + "learning_rate": 4.408442519159075e-05, + "loss": 3.1944, + "step": 9764 + }, + { + "epoch": 0.008174313376955377, + "grad_norm": 0.15137068927288055, + "learning_rate": 4.408268621866858e-05, + "loss": 3.2063, + "step": 9765 + }, + { + "epoch": 0.008205159842528793, + "grad_norm": 0.16249825060367584, + "learning_rate": 4.408094702449412e-05, + "loss": 3.2058, + "step": 9766 + }, + { + "epoch": 0.00823600630810221, + "grad_norm": 0.1628628522157669, + "learning_rate": 4.407920760908756e-05, + "loss": 3.1962, + "step": 9767 + }, + { + "epoch": 0.008266852773675627, + "grad_norm": 0.15595337748527527, + "learning_rate": 4.407746797246905e-05, + "loss": 3.1883, + "step": 9768 + }, + { + "epoch": 0.008297699239249043, + "grad_norm": 0.15086597204208374, + "learning_rate": 4.407572811465876e-05, + "loss": 3.2008, + "step": 9769 + }, + { + "epoch": 0.008328545704822459, + "grad_norm": 0.15129947662353516, + "learning_rate": 4.407398803567687e-05, + "loss": 3.1867, + "step": 9770 + }, + { + "epoch": 0.008359392170395876, + "grad_norm": 0.15855316817760468, + "learning_rate": 4.407224773554355e-05, + "loss": 3.1725, + "step": 9771 + }, + { + "epoch": 0.008390238635969292, + "grad_norm": 0.1498861312866211, + "learning_rate": 4.4070507214278976e-05, + "loss": 3.195, + "step": 9772 + }, + { + "epoch": 0.008421085101542708, + "grad_norm": 0.151406392455101, + "learning_rate": 4.406876647190333e-05, + "loss": 3.1948, + "step": 9773 + }, + { + "epoch": 0.008451931567116126, + "grad_norm": 0.14833621680736542, + "learning_rate": 4.40670255084368e-05, + "loss": 3.2289, + "step": 9774 + }, + { + "epoch": 0.008482778032689542, + "grad_norm": 0.14842906594276428, + "learning_rate": 4.406528432389956e-05, + "loss": 3.2054, + "step": 9775 + }, + { + "epoch": 0.008513624498262958, + "grad_norm": 0.15722180902957916, + "learning_rate": 4.406354291831182e-05, + "loss": 3.2039, + "step": 9776 + }, + { + "epoch": 0.008544470963836375, + "grad_norm": 0.1605103462934494, + "learning_rate": 4.406180129169374e-05, + "loss": 3.194, + "step": 9777 + }, + { + "epoch": 0.008575317429409791, + "grad_norm": 0.1621682047843933, + "learning_rate": 4.406005944406553e-05, + "loss": 3.1862, + "step": 9778 + }, + { + "epoch": 0.008606163894983207, + "grad_norm": 0.15079393982887268, + "learning_rate": 4.405831737544739e-05, + "loss": 3.1634, + "step": 9779 + }, + { + "epoch": 0.008637010360556625, + "grad_norm": 0.14763841032981873, + "learning_rate": 4.4056575085859514e-05, + "loss": 3.2166, + "step": 9780 + }, + { + "epoch": 0.008667856826130041, + "grad_norm": 0.15355858206748962, + "learning_rate": 4.40548325753221e-05, + "loss": 3.1999, + "step": 9781 + }, + { + "epoch": 0.008698703291703457, + "grad_norm": 0.1468457728624344, + "learning_rate": 4.405308984385536e-05, + "loss": 3.2097, + "step": 9782 + }, + { + "epoch": 0.008729549757276875, + "grad_norm": 0.1439363956451416, + "learning_rate": 4.405134689147948e-05, + "loss": 3.205, + "step": 9783 + }, + { + "epoch": 0.00876039622285029, + "grad_norm": 0.155225470662117, + "learning_rate": 4.404960371821468e-05, + "loss": 3.1945, + "step": 9784 + }, + { + "epoch": 0.008791242688423706, + "grad_norm": 0.1466435343027115, + "learning_rate": 4.404786032408118e-05, + "loss": 3.2323, + "step": 9785 + }, + { + "epoch": 0.008822089153997124, + "grad_norm": 0.15702782571315765, + "learning_rate": 4.4046116709099184e-05, + "loss": 3.1838, + "step": 9786 + }, + { + "epoch": 0.00885293561957054, + "grad_norm": 0.155763179063797, + "learning_rate": 4.4044372873288913e-05, + "loss": 3.2313, + "step": 9787 + }, + { + "epoch": 0.008883782085143956, + "grad_norm": 0.14973482489585876, + "learning_rate": 4.4042628816670584e-05, + "loss": 3.2141, + "step": 9788 + }, + { + "epoch": 0.008914628550717374, + "grad_norm": 0.1497003436088562, + "learning_rate": 4.404088453926442e-05, + "loss": 3.189, + "step": 9789 + }, + { + "epoch": 0.00894547501629079, + "grad_norm": 0.1569911241531372, + "learning_rate": 4.403914004109063e-05, + "loss": 3.2135, + "step": 9790 + }, + { + "epoch": 0.008976321481864205, + "grad_norm": 0.15750008821487427, + "learning_rate": 4.4037395322169455e-05, + "loss": 3.1634, + "step": 9791 + }, + { + "epoch": 0.009007167947437623, + "grad_norm": 0.15590205788612366, + "learning_rate": 4.403565038252113e-05, + "loss": 3.1694, + "step": 9792 + }, + { + "epoch": 0.009038014413011039, + "grad_norm": 0.14669813215732574, + "learning_rate": 4.403390522216587e-05, + "loss": 3.2192, + "step": 9793 + }, + { + "epoch": 0.009068860878584455, + "grad_norm": 0.15884193778038025, + "learning_rate": 4.403215984112392e-05, + "loss": 3.2012, + "step": 9794 + }, + { + "epoch": 0.009099707344157873, + "grad_norm": 0.15315212309360504, + "learning_rate": 4.403041423941551e-05, + "loss": 3.1961, + "step": 9795 + }, + { + "epoch": 0.009130553809731289, + "grad_norm": 0.1470518261194229, + "learning_rate": 4.402866841706089e-05, + "loss": 3.1657, + "step": 9796 + }, + { + "epoch": 0.009161400275304704, + "grad_norm": 0.15922687947750092, + "learning_rate": 4.4026922374080294e-05, + "loss": 3.1495, + "step": 9797 + }, + { + "epoch": 0.009192246740878122, + "grad_norm": 0.15750660002231598, + "learning_rate": 4.402517611049396e-05, + "loss": 3.1991, + "step": 9798 + }, + { + "epoch": 0.009223093206451538, + "grad_norm": 0.1543986201286316, + "learning_rate": 4.4023429626322143e-05, + "loss": 3.2145, + "step": 9799 + }, + { + "epoch": 0.009253939672024954, + "grad_norm": 0.15900395810604095, + "learning_rate": 4.40216829215851e-05, + "loss": 3.1861, + "step": 9800 + }, + { + "epoch": 0.009284786137598372, + "grad_norm": 0.1638905555009842, + "learning_rate": 4.401993599630306e-05, + "loss": 3.2107, + "step": 9801 + }, + { + "epoch": 0.009315632603171788, + "grad_norm": 0.18782155215740204, + "learning_rate": 4.40181888504963e-05, + "loss": 3.1922, + "step": 9802 + }, + { + "epoch": 0.009346479068745204, + "grad_norm": 0.1609722077846527, + "learning_rate": 4.4016441484185066e-05, + "loss": 3.2064, + "step": 9803 + }, + { + "epoch": 0.009377325534318621, + "grad_norm": 0.15365126729011536, + "learning_rate": 4.401469389738962e-05, + "loss": 3.1572, + "step": 9804 + }, + { + "epoch": 0.009408171999892037, + "grad_norm": 0.15898288786411285, + "learning_rate": 4.4012946090130234e-05, + "loss": 3.1786, + "step": 9805 + }, + { + "epoch": 0.009439018465465453, + "grad_norm": 0.1493692249059677, + "learning_rate": 4.401119806242716e-05, + "loss": 3.2137, + "step": 9806 + }, + { + "epoch": 0.00946986493103887, + "grad_norm": 0.15456415712833405, + "learning_rate": 4.400944981430066e-05, + "loss": 3.1964, + "step": 9807 + }, + { + "epoch": 0.009500711396612287, + "grad_norm": 0.155643031001091, + "learning_rate": 4.4007701345771024e-05, + "loss": 3.1942, + "step": 9808 + }, + { + "epoch": 0.009531557862185703, + "grad_norm": 0.16246679425239563, + "learning_rate": 4.4005952656858506e-05, + "loss": 3.1748, + "step": 9809 + }, + { + "epoch": 0.00956240432775912, + "grad_norm": 0.15065635740756989, + "learning_rate": 4.4004203747583395e-05, + "loss": 3.1741, + "step": 9810 + }, + { + "epoch": 0.009593250793332536, + "grad_norm": 0.1481941044330597, + "learning_rate": 4.4002454617965956e-05, + "loss": 3.1989, + "step": 9811 + }, + { + "epoch": 0.009624097258905954, + "grad_norm": 0.1499778926372528, + "learning_rate": 4.4000705268026474e-05, + "loss": 3.1951, + "step": 9812 + }, + { + "epoch": 0.00965494372447937, + "grad_norm": 0.14504516124725342, + "learning_rate": 4.399895569778524e-05, + "loss": 3.1705, + "step": 9813 + }, + { + "epoch": 0.009685790190052786, + "grad_norm": 0.14920194447040558, + "learning_rate": 4.3997205907262526e-05, + "loss": 3.1713, + "step": 9814 + }, + { + "epoch": 0.009716636655626203, + "grad_norm": 0.15208910405635834, + "learning_rate": 4.3995455896478624e-05, + "loss": 3.2065, + "step": 9815 + }, + { + "epoch": 0.00974748312119962, + "grad_norm": 0.14788983762264252, + "learning_rate": 4.399370566545383e-05, + "loss": 3.1696, + "step": 9816 + }, + { + "epoch": 0.009778329586773035, + "grad_norm": 0.15008017420768738, + "learning_rate": 4.3991955214208424e-05, + "loss": 3.1629, + "step": 9817 + }, + { + "epoch": 0.009809176052346453, + "grad_norm": 0.14652079343795776, + "learning_rate": 4.399020454276272e-05, + "loss": 3.193, + "step": 9818 + }, + { + "epoch": 0.009840022517919869, + "grad_norm": 0.14991910755634308, + "learning_rate": 4.3988453651137e-05, + "loss": 3.213, + "step": 9819 + }, + { + "epoch": 0.009870868983493285, + "grad_norm": 0.15804634988307953, + "learning_rate": 4.398670253935158e-05, + "loss": 3.1735, + "step": 9820 + }, + { + "epoch": 0.009901715449066702, + "grad_norm": 0.1575479954481125, + "learning_rate": 4.398495120742674e-05, + "loss": 3.2025, + "step": 9821 + }, + { + "epoch": 0.009932561914640118, + "grad_norm": 0.15228165686130524, + "learning_rate": 4.398319965538281e-05, + "loss": 3.1884, + "step": 9822 + }, + { + "epoch": 0.009963408380213534, + "grad_norm": 0.148068368434906, + "learning_rate": 4.3981447883240076e-05, + "loss": 3.2204, + "step": 9823 + }, + { + "epoch": 0.009994254845786952, + "grad_norm": 0.1548304706811905, + "learning_rate": 4.3979695891018875e-05, + "loss": 3.2191, + "step": 9824 + }, + { + "epoch": 0.010025101311360368, + "grad_norm": 0.15757305920124054, + "learning_rate": 4.39779436787395e-05, + "loss": 3.2161, + "step": 9825 + }, + { + "epoch": 0.010055947776933784, + "grad_norm": 0.14942990243434906, + "learning_rate": 4.397619124642227e-05, + "loss": 3.1652, + "step": 9826 + }, + { + "epoch": 0.010086794242507201, + "grad_norm": 0.14867264032363892, + "learning_rate": 4.397443859408751e-05, + "loss": 3.192, + "step": 9827 + }, + { + "epoch": 0.010117640708080617, + "grad_norm": 0.16530495882034302, + "learning_rate": 4.397268572175553e-05, + "loss": 3.1994, + "step": 9828 + }, + { + "epoch": 0.010148487173654033, + "grad_norm": 0.1516425609588623, + "learning_rate": 4.397093262944667e-05, + "loss": 3.2118, + "step": 9829 + }, + { + "epoch": 0.010179333639227451, + "grad_norm": 0.1558454930782318, + "learning_rate": 4.396917931718123e-05, + "loss": 3.1943, + "step": 9830 + }, + { + "epoch": 0.010210180104800867, + "grad_norm": 0.1575126349925995, + "learning_rate": 4.396742578497957e-05, + "loss": 3.1739, + "step": 9831 + }, + { + "epoch": 0.010241026570374283, + "grad_norm": 0.15738272666931152, + "learning_rate": 4.396567203286201e-05, + "loss": 3.2271, + "step": 9832 + }, + { + "epoch": 0.0102718730359477, + "grad_norm": 0.1583673357963562, + "learning_rate": 4.3963918060848865e-05, + "loss": 3.2267, + "step": 9833 + }, + { + "epoch": 0.010302719501521116, + "grad_norm": 0.15272639691829681, + "learning_rate": 4.396216386896049e-05, + "loss": 3.2117, + "step": 9834 + }, + { + "epoch": 0.010333565967094532, + "grad_norm": 0.1516740322113037, + "learning_rate": 4.396040945721722e-05, + "loss": 3.1971, + "step": 9835 + }, + { + "epoch": 0.01036441243266795, + "grad_norm": 0.14648553729057312, + "learning_rate": 4.39586548256394e-05, + "loss": 3.1691, + "step": 9836 + }, + { + "epoch": 0.010395258898241366, + "grad_norm": 0.15533487498760223, + "learning_rate": 4.3956899974247365e-05, + "loss": 3.2039, + "step": 9837 + }, + { + "epoch": 0.010426105363814782, + "grad_norm": 0.14941616356372833, + "learning_rate": 4.3955144903061475e-05, + "loss": 3.1829, + "step": 9838 + }, + { + "epoch": 0.0104569518293882, + "grad_norm": 0.15625859797000885, + "learning_rate": 4.395338961210207e-05, + "loss": 3.1987, + "step": 9839 + }, + { + "epoch": 0.010487798294961615, + "grad_norm": 0.16057048738002777, + "learning_rate": 4.395163410138949e-05, + "loss": 3.1902, + "step": 9840 + }, + { + "epoch": 0.010518644760535031, + "grad_norm": 0.14992791414260864, + "learning_rate": 4.394987837094412e-05, + "loss": 3.171, + "step": 9841 + }, + { + "epoch": 0.010549491226108449, + "grad_norm": 0.15704990923404694, + "learning_rate": 4.3948122420786284e-05, + "loss": 3.1993, + "step": 9842 + }, + { + "epoch": 0.010580337691681865, + "grad_norm": 0.16028746962547302, + "learning_rate": 4.3946366250936354e-05, + "loss": 3.2039, + "step": 9843 + }, + { + "epoch": 0.010611184157255281, + "grad_norm": 0.15378250181674957, + "learning_rate": 4.3944609861414697e-05, + "loss": 3.2067, + "step": 9844 + }, + { + "epoch": 0.010642030622828699, + "grad_norm": 0.1802055984735489, + "learning_rate": 4.3942853252241676e-05, + "loss": 3.156, + "step": 9845 + }, + { + "epoch": 0.010672877088402115, + "grad_norm": 0.1542234718799591, + "learning_rate": 4.394109642343765e-05, + "loss": 3.176, + "step": 9846 + }, + { + "epoch": 0.01070372355397553, + "grad_norm": 0.15915855765342712, + "learning_rate": 4.3939339375023e-05, + "loss": 3.165, + "step": 9847 + }, + { + "epoch": 0.010734570019548948, + "grad_norm": 0.1615767627954483, + "learning_rate": 4.393758210701808e-05, + "loss": 3.2011, + "step": 9848 + }, + { + "epoch": 0.010765416485122364, + "grad_norm": 0.17452839016914368, + "learning_rate": 4.393582461944329e-05, + "loss": 3.1666, + "step": 9849 + }, + { + "epoch": 0.01079626295069578, + "grad_norm": 0.1641661375761032, + "learning_rate": 4.393406691231898e-05, + "loss": 3.1717, + "step": 9850 + }, + { + "epoch": 0.010827109416269198, + "grad_norm": 0.575663685798645, + "learning_rate": 4.393230898566555e-05, + "loss": 3.2009, + "step": 9851 + }, + { + "epoch": 0.010857955881842614, + "grad_norm": 0.1555711328983307, + "learning_rate": 4.3930550839503365e-05, + "loss": 3.2005, + "step": 9852 + }, + { + "epoch": 0.01088880234741603, + "grad_norm": 0.1525668203830719, + "learning_rate": 4.3928792473852826e-05, + "loss": 3.1865, + "step": 9853 + }, + { + "epoch": 0.010919648812989447, + "grad_norm": 0.16874100267887115, + "learning_rate": 4.3927033888734316e-05, + "loss": 3.195, + "step": 9854 + }, + { + "epoch": 0.010950495278562863, + "grad_norm": 0.17599420249462128, + "learning_rate": 4.3925275084168217e-05, + "loss": 3.1918, + "step": 9855 + }, + { + "epoch": 0.010981341744136279, + "grad_norm": 0.18972589075565338, + "learning_rate": 4.3923516060174926e-05, + "loss": 3.2171, + "step": 9856 + }, + { + "epoch": 0.011012188209709697, + "grad_norm": 0.1701868176460266, + "learning_rate": 4.392175681677484e-05, + "loss": 3.1745, + "step": 9857 + }, + { + "epoch": 0.011043034675283113, + "grad_norm": 0.1556263566017151, + "learning_rate": 4.391999735398835e-05, + "loss": 3.2076, + "step": 9858 + }, + { + "epoch": 0.011073881140856529, + "grad_norm": 0.15682367980480194, + "learning_rate": 4.391823767183587e-05, + "loss": 3.1795, + "step": 9859 + }, + { + "epoch": 0.011104727606429946, + "grad_norm": 0.16331705451011658, + "learning_rate": 4.391647777033778e-05, + "loss": 3.1876, + "step": 9860 + }, + { + "epoch": 0.011135574072003362, + "grad_norm": 0.1552308350801468, + "learning_rate": 4.39147176495145e-05, + "loss": 3.208, + "step": 9861 + }, + { + "epoch": 0.011166420537576778, + "grad_norm": 0.15876562893390656, + "learning_rate": 4.391295730938645e-05, + "loss": 3.1465, + "step": 9862 + }, + { + "epoch": 0.011197267003150196, + "grad_norm": 0.1709706336259842, + "learning_rate": 4.391119674997401e-05, + "loss": 3.1886, + "step": 9863 + }, + { + "epoch": 0.011228113468723612, + "grad_norm": 0.15424536168575287, + "learning_rate": 4.390943597129761e-05, + "loss": 3.2145, + "step": 9864 + }, + { + "epoch": 0.011258959934297028, + "grad_norm": 0.15467852354049683, + "learning_rate": 4.390767497337767e-05, + "loss": 3.183, + "step": 9865 + }, + { + "epoch": 0.011289806399870445, + "grad_norm": 0.15203578770160675, + "learning_rate": 4.39059137562346e-05, + "loss": 3.1838, + "step": 9866 + }, + { + "epoch": 0.011320652865443861, + "grad_norm": 0.15967658162117004, + "learning_rate": 4.390415231988882e-05, + "loss": 3.1998, + "step": 9867 + }, + { + "epoch": 0.011351499331017277, + "grad_norm": 0.16309532523155212, + "learning_rate": 4.3902390664360745e-05, + "loss": 3.194, + "step": 9868 + }, + { + "epoch": 0.011382345796590695, + "grad_norm": 0.15586745738983154, + "learning_rate": 4.390062878967082e-05, + "loss": 3.1805, + "step": 9869 + }, + { + "epoch": 0.01141319226216411, + "grad_norm": 0.15432487428188324, + "learning_rate": 4.389886669583946e-05, + "loss": 3.18, + "step": 9870 + }, + { + "epoch": 0.011444038727737527, + "grad_norm": 0.17678473889827728, + "learning_rate": 4.3897104382887096e-05, + "loss": 3.2019, + "step": 9871 + }, + { + "epoch": 0.011474885193310944, + "grad_norm": 0.14661887288093567, + "learning_rate": 4.389534185083416e-05, + "loss": 3.1738, + "step": 9872 + }, + { + "epoch": 0.01150573165888436, + "grad_norm": 0.16934992372989655, + "learning_rate": 4.3893579099701096e-05, + "loss": 3.1577, + "step": 9873 + }, + { + "epoch": 0.011536578124457776, + "grad_norm": 0.15686126053333282, + "learning_rate": 4.389181612950833e-05, + "loss": 3.1916, + "step": 9874 + }, + { + "epoch": 0.011567424590031194, + "grad_norm": 0.15056651830673218, + "learning_rate": 4.389005294027631e-05, + "loss": 3.2047, + "step": 9875 + }, + { + "epoch": 0.01159827105560461, + "grad_norm": 0.1635604351758957, + "learning_rate": 4.388828953202548e-05, + "loss": 3.1789, + "step": 9876 + }, + { + "epoch": 0.011629117521178026, + "grad_norm": 0.15551987290382385, + "learning_rate": 4.3886525904776283e-05, + "loss": 3.1704, + "step": 9877 + }, + { + "epoch": 0.011659963986751443, + "grad_norm": 0.15508976578712463, + "learning_rate": 4.388476205854917e-05, + "loss": 3.1704, + "step": 9878 + }, + { + "epoch": 0.01169081045232486, + "grad_norm": 0.16143320500850677, + "learning_rate": 4.388299799336458e-05, + "loss": 3.2222, + "step": 9879 + }, + { + "epoch": 0.011721656917898277, + "grad_norm": 0.15069961547851562, + "learning_rate": 4.388123370924299e-05, + "loss": 3.2293, + "step": 9880 + }, + { + "epoch": 0.011752503383471693, + "grad_norm": 0.15734970569610596, + "learning_rate": 4.3879469206204825e-05, + "loss": 3.1935, + "step": 9881 + }, + { + "epoch": 0.011783349849045109, + "grad_norm": 0.1524030566215515, + "learning_rate": 4.3877704484270575e-05, + "loss": 3.2062, + "step": 9882 + }, + { + "epoch": 0.011814196314618526, + "grad_norm": 0.19731181859970093, + "learning_rate": 4.3875939543460674e-05, + "loss": 3.1828, + "step": 9883 + }, + { + "epoch": 0.011845042780191942, + "grad_norm": 0.15367256104946136, + "learning_rate": 4.3874174383795605e-05, + "loss": 3.1816, + "step": 9884 + }, + { + "epoch": 0.011875889245765358, + "grad_norm": 0.15342660248279572, + "learning_rate": 4.387240900529582e-05, + "loss": 3.1852, + "step": 9885 + }, + { + "epoch": 0.011906735711338776, + "grad_norm": 0.15411187708377838, + "learning_rate": 4.38706434079818e-05, + "loss": 3.1692, + "step": 9886 + }, + { + "epoch": 0.011937582176912192, + "grad_norm": 0.16224293410778046, + "learning_rate": 4.3868877591874e-05, + "loss": 3.1824, + "step": 9887 + }, + { + "epoch": 0.011968428642485608, + "grad_norm": 0.15418227016925812, + "learning_rate": 4.386711155699291e-05, + "loss": 3.1696, + "step": 9888 + }, + { + "epoch": 0.011999275108059025, + "grad_norm": 0.15639343857765198, + "learning_rate": 4.386534530335901e-05, + "loss": 3.175, + "step": 9889 + }, + { + "epoch": 0.012030121573632441, + "grad_norm": 0.15530472993850708, + "learning_rate": 4.386357883099275e-05, + "loss": 3.1891, + "step": 9890 + }, + { + "epoch": 0.012060968039205857, + "grad_norm": 0.15457680821418762, + "learning_rate": 4.386181213991464e-05, + "loss": 3.1889, + "step": 9891 + }, + { + "epoch": 0.012091814504779275, + "grad_norm": 0.1536475569009781, + "learning_rate": 4.386004523014514e-05, + "loss": 3.1911, + "step": 9892 + }, + { + "epoch": 0.012122660970352691, + "grad_norm": 0.16413508355617523, + "learning_rate": 4.3858278101704764e-05, + "loss": 3.1985, + "step": 9893 + }, + { + "epoch": 0.012153507435926107, + "grad_norm": 0.1683902144432068, + "learning_rate": 4.385651075461398e-05, + "loss": 3.1799, + "step": 9894 + }, + { + "epoch": 0.012184353901499525, + "grad_norm": 0.1503283530473709, + "learning_rate": 4.385474318889329e-05, + "loss": 3.2063, + "step": 9895 + }, + { + "epoch": 0.01221520036707294, + "grad_norm": 0.1487453430891037, + "learning_rate": 4.3852975404563176e-05, + "loss": 3.2095, + "step": 9896 + }, + { + "epoch": 0.012246046832646356, + "grad_norm": 0.16660140454769135, + "learning_rate": 4.385120740164414e-05, + "loss": 3.1766, + "step": 9897 + }, + { + "epoch": 0.012276893298219774, + "grad_norm": 0.15295575559139252, + "learning_rate": 4.3849439180156685e-05, + "loss": 3.2066, + "step": 9898 + }, + { + "epoch": 0.01230773976379319, + "grad_norm": 0.1506565511226654, + "learning_rate": 4.384767074012132e-05, + "loss": 3.2067, + "step": 9899 + }, + { + "epoch": 0.012338586229366606, + "grad_norm": 0.15306252241134644, + "learning_rate": 4.3845902081558524e-05, + "loss": 3.1688, + "step": 9900 + }, + { + "epoch": 0.012369432694940024, + "grad_norm": 0.149606391787529, + "learning_rate": 4.384413320448882e-05, + "loss": 3.1681, + "step": 9901 + }, + { + "epoch": 0.01240027916051344, + "grad_norm": 0.19714824855327606, + "learning_rate": 4.384236410893272e-05, + "loss": 3.1923, + "step": 9902 + }, + { + "epoch": 0.012431125626086855, + "grad_norm": 0.14689095318317413, + "learning_rate": 4.384059479491073e-05, + "loss": 3.1973, + "step": 9903 + }, + { + "epoch": 0.012461972091660273, + "grad_norm": 0.15034158527851105, + "learning_rate": 4.383882526244337e-05, + "loss": 3.2121, + "step": 9904 + }, + { + "epoch": 0.012492818557233689, + "grad_norm": 0.1528887003660202, + "learning_rate": 4.383705551155115e-05, + "loss": 3.1957, + "step": 9905 + }, + { + "epoch": 0.012523665022807105, + "grad_norm": 0.16250428557395935, + "learning_rate": 4.383528554225459e-05, + "loss": 3.2148, + "step": 9906 + }, + { + "epoch": 0.012554511488380523, + "grad_norm": 0.14475861191749573, + "learning_rate": 4.383351535457421e-05, + "loss": 3.144, + "step": 9907 + }, + { + "epoch": 0.012585357953953939, + "grad_norm": 0.15661996603012085, + "learning_rate": 4.3831744948530537e-05, + "loss": 3.2063, + "step": 9908 + }, + { + "epoch": 0.012616204419527354, + "grad_norm": 0.14942391216754913, + "learning_rate": 4.38299743241441e-05, + "loss": 3.1967, + "step": 9909 + }, + { + "epoch": 0.012647050885100772, + "grad_norm": 0.1791783571243286, + "learning_rate": 4.3828203481435424e-05, + "loss": 3.2069, + "step": 9910 + }, + { + "epoch": 0.012677897350674188, + "grad_norm": 0.15564122796058655, + "learning_rate": 4.3826432420425046e-05, + "loss": 3.1473, + "step": 9911 + }, + { + "epoch": 0.012708743816247604, + "grad_norm": 0.15223829448223114, + "learning_rate": 4.38246611411335e-05, + "loss": 3.1798, + "step": 9912 + }, + { + "epoch": 0.012739590281821022, + "grad_norm": 0.1562739610671997, + "learning_rate": 4.382288964358132e-05, + "loss": 3.2003, + "step": 9913 + }, + { + "epoch": 0.012770436747394438, + "grad_norm": 0.15043431520462036, + "learning_rate": 4.382111792778904e-05, + "loss": 3.2172, + "step": 9914 + }, + { + "epoch": 0.012801283212967854, + "grad_norm": 0.14790217578411102, + "learning_rate": 4.381934599377721e-05, + "loss": 3.1781, + "step": 9915 + }, + { + "epoch": 0.012832129678541271, + "grad_norm": 0.15516825020313263, + "learning_rate": 4.381757384156637e-05, + "loss": 3.182, + "step": 9916 + }, + { + "epoch": 0.012862976144114687, + "grad_norm": 0.1621103584766388, + "learning_rate": 4.381580147117708e-05, + "loss": 3.2031, + "step": 9917 + }, + { + "epoch": 0.012893822609688103, + "grad_norm": 0.15685558319091797, + "learning_rate": 4.381402888262987e-05, + "loss": 3.1809, + "step": 9918 + }, + { + "epoch": 0.01292466907526152, + "grad_norm": 0.15483270585536957, + "learning_rate": 4.38122560759453e-05, + "loss": 3.1945, + "step": 9919 + }, + { + "epoch": 0.012955515540834937, + "grad_norm": 0.1482497900724411, + "learning_rate": 4.381048305114393e-05, + "loss": 3.191, + "step": 9920 + }, + { + "epoch": 0.012986362006408353, + "grad_norm": 0.18660099804401398, + "learning_rate": 4.380870980824631e-05, + "loss": 3.1602, + "step": 9921 + }, + { + "epoch": 0.01301720847198177, + "grad_norm": 0.14576470851898193, + "learning_rate": 4.3806936347273006e-05, + "loss": 3.2241, + "step": 9922 + }, + { + "epoch": 0.013048054937555186, + "grad_norm": 0.15078052878379822, + "learning_rate": 4.380516266824457e-05, + "loss": 3.2368, + "step": 9923 + }, + { + "epoch": 0.013078901403128602, + "grad_norm": 0.14931528270244598, + "learning_rate": 4.380338877118158e-05, + "loss": 3.1825, + "step": 9924 + }, + { + "epoch": 0.01310974786870202, + "grad_norm": 0.16028819978237152, + "learning_rate": 4.3801614656104595e-05, + "loss": 3.186, + "step": 9925 + }, + { + "epoch": 0.013140594334275436, + "grad_norm": 0.15303142368793488, + "learning_rate": 4.3799840323034185e-05, + "loss": 3.1733, + "step": 9926 + }, + { + "epoch": 0.013171440799848852, + "grad_norm": 0.14937084913253784, + "learning_rate": 4.3798065771990925e-05, + "loss": 3.1993, + "step": 9927 + }, + { + "epoch": 0.01320228726542227, + "grad_norm": 1.0477674007415771, + "learning_rate": 4.379629100299538e-05, + "loss": 3.2147, + "step": 9928 + }, + { + "epoch": 0.013233133730995685, + "grad_norm": 0.15169298648834229, + "learning_rate": 4.379451601606815e-05, + "loss": 3.1906, + "step": 9929 + }, + { + "epoch": 0.013263980196569101, + "grad_norm": 0.1653587520122528, + "learning_rate": 4.3792740811229794e-05, + "loss": 3.1854, + "step": 9930 + }, + { + "epoch": 0.013294826662142519, + "grad_norm": 0.17475207149982452, + "learning_rate": 4.37909653885009e-05, + "loss": 3.1896, + "step": 9931 + }, + { + "epoch": 0.013325673127715935, + "grad_norm": 0.15551596879959106, + "learning_rate": 4.378918974790206e-05, + "loss": 3.2073, + "step": 9932 + }, + { + "epoch": 0.01335651959328935, + "grad_norm": 0.15061922371387482, + "learning_rate": 4.378741388945386e-05, + "loss": 3.1861, + "step": 9933 + }, + { + "epoch": 0.013387366058862768, + "grad_norm": 0.15200349688529968, + "learning_rate": 4.378563781317687e-05, + "loss": 3.2124, + "step": 9934 + }, + { + "epoch": 0.013418212524436184, + "grad_norm": 0.15013693273067474, + "learning_rate": 4.3783861519091713e-05, + "loss": 3.1957, + "step": 9935 + }, + { + "epoch": 0.0134490589900096, + "grad_norm": 0.1523367017507553, + "learning_rate": 4.3782085007218965e-05, + "loss": 3.1847, + "step": 9936 + }, + { + "epoch": 0.013479905455583018, + "grad_norm": 0.15669547021389008, + "learning_rate": 4.378030827757922e-05, + "loss": 3.1703, + "step": 9937 + }, + { + "epoch": 0.013510751921156434, + "grad_norm": 0.15379972755908966, + "learning_rate": 4.377853133019309e-05, + "loss": 3.196, + "step": 9938 + }, + { + "epoch": 0.01354159838672985, + "grad_norm": 0.15864889323711395, + "learning_rate": 4.3776754165081183e-05, + "loss": 3.1779, + "step": 9939 + }, + { + "epoch": 0.013572444852303267, + "grad_norm": 0.15594907104969025, + "learning_rate": 4.3774976782264086e-05, + "loss": 3.2152, + "step": 9940 + }, + { + "epoch": 0.013603291317876683, + "grad_norm": 0.15065404772758484, + "learning_rate": 4.377319918176243e-05, + "loss": 3.2038, + "step": 9941 + }, + { + "epoch": 0.0136341377834501, + "grad_norm": 0.1544085443019867, + "learning_rate": 4.377142136359679e-05, + "loss": 3.2089, + "step": 9942 + }, + { + "epoch": 0.013664984249023517, + "grad_norm": 0.3395777642726898, + "learning_rate": 4.376964332778781e-05, + "loss": 3.2195, + "step": 9943 + }, + { + "epoch": 0.013695830714596933, + "grad_norm": 0.16523492336273193, + "learning_rate": 4.37678650743561e-05, + "loss": 3.2114, + "step": 9944 + }, + { + "epoch": 0.013726677180170349, + "grad_norm": 0.15789949893951416, + "learning_rate": 4.3766086603322265e-05, + "loss": 3.178, + "step": 9945 + }, + { + "epoch": 0.013757523645743766, + "grad_norm": 0.17302103340625763, + "learning_rate": 4.376430791470693e-05, + "loss": 3.2179, + "step": 9946 + }, + { + "epoch": 0.013788370111317182, + "grad_norm": 0.15268810093402863, + "learning_rate": 4.3762529008530725e-05, + "loss": 3.1705, + "step": 9947 + }, + { + "epoch": 0.0138192165768906, + "grad_norm": 0.1682732105255127, + "learning_rate": 4.376074988481427e-05, + "loss": 3.1895, + "step": 9948 + }, + { + "epoch": 0.013850063042464016, + "grad_norm": 0.1600591242313385, + "learning_rate": 4.37589705435782e-05, + "loss": 3.1988, + "step": 9949 + }, + { + "epoch": 0.013880909508037432, + "grad_norm": 0.15717241168022156, + "learning_rate": 4.3757190984843135e-05, + "loss": 3.1743, + "step": 9950 + }, + { + "epoch": 0.01391175597361085, + "grad_norm": 0.15379256010055542, + "learning_rate": 4.375541120862971e-05, + "loss": 3.2188, + "step": 9951 + }, + { + "epoch": 0.013942602439184265, + "grad_norm": 0.1516760289669037, + "learning_rate": 4.375363121495856e-05, + "loss": 3.1888, + "step": 9952 + }, + { + "epoch": 0.013973448904757681, + "grad_norm": 0.1474515050649643, + "learning_rate": 4.375185100385033e-05, + "loss": 3.2064, + "step": 9953 + }, + { + "epoch": 0.014004295370331099, + "grad_norm": 0.1442224383354187, + "learning_rate": 4.375007057532567e-05, + "loss": 3.1915, + "step": 9954 + }, + { + "epoch": 0.014035141835904515, + "grad_norm": 0.15276281535625458, + "learning_rate": 4.3748289929405186e-05, + "loss": 3.1679, + "step": 9955 + }, + { + "epoch": 0.014065988301477931, + "grad_norm": 0.14833304286003113, + "learning_rate": 4.3746509066109556e-05, + "loss": 3.2211, + "step": 9956 + }, + { + "epoch": 0.014096834767051349, + "grad_norm": 0.15719731152057648, + "learning_rate": 4.3744727985459424e-05, + "loss": 3.1691, + "step": 9957 + }, + { + "epoch": 0.014127681232624765, + "grad_norm": 0.15252068638801575, + "learning_rate": 4.374294668747543e-05, + "loss": 3.1857, + "step": 9958 + }, + { + "epoch": 0.01415852769819818, + "grad_norm": 0.15538376569747925, + "learning_rate": 4.374116517217823e-05, + "loss": 3.2104, + "step": 9959 + }, + { + "epoch": 0.014189374163771598, + "grad_norm": 0.19414693117141724, + "learning_rate": 4.373938343958848e-05, + "loss": 3.212, + "step": 9960 + }, + { + "epoch": 0.014220220629345014, + "grad_norm": 0.15688499808311462, + "learning_rate": 4.3737601489726846e-05, + "loss": 3.1893, + "step": 9961 + }, + { + "epoch": 0.01425106709491843, + "grad_norm": 0.15280228853225708, + "learning_rate": 4.373581932261398e-05, + "loss": 3.1489, + "step": 9962 + }, + { + "epoch": 0.014281913560491848, + "grad_norm": 0.1524985283613205, + "learning_rate": 4.373403693827055e-05, + "loss": 3.1998, + "step": 9963 + }, + { + "epoch": 0.014312760026065264, + "grad_norm": 0.15536603331565857, + "learning_rate": 4.373225433671722e-05, + "loss": 3.1807, + "step": 9964 + }, + { + "epoch": 0.01434360649163868, + "grad_norm": 0.15804751217365265, + "learning_rate": 4.373047151797466e-05, + "loss": 3.174, + "step": 9965 + }, + { + "epoch": 0.014374452957212097, + "grad_norm": 0.14933468401432037, + "learning_rate": 4.3728688482063534e-05, + "loss": 3.2155, + "step": 9966 + }, + { + "epoch": 0.014405299422785513, + "grad_norm": 0.14879639446735382, + "learning_rate": 4.372690522900452e-05, + "loss": 3.2166, + "step": 9967 + }, + { + "epoch": 0.014436145888358929, + "grad_norm": 0.15270866453647614, + "learning_rate": 4.37251217588183e-05, + "loss": 3.1693, + "step": 9968 + }, + { + "epoch": 0.014466992353932347, + "grad_norm": 0.15351969003677368, + "learning_rate": 4.3723338071525535e-05, + "loss": 3.226, + "step": 9969 + }, + { + "epoch": 0.014497838819505763, + "grad_norm": 0.15284991264343262, + "learning_rate": 4.372155416714692e-05, + "loss": 3.1617, + "step": 9970 + }, + { + "epoch": 0.014528685285079179, + "grad_norm": 0.1619681864976883, + "learning_rate": 4.371977004570313e-05, + "loss": 3.2133, + "step": 9971 + }, + { + "epoch": 0.014559531750652596, + "grad_norm": 0.1459193229675293, + "learning_rate": 4.371798570721487e-05, + "loss": 3.1902, + "step": 9972 + }, + { + "epoch": 0.014590378216226012, + "grad_norm": 0.15211352705955505, + "learning_rate": 4.3716201151702804e-05, + "loss": 3.1928, + "step": 9973 + }, + { + "epoch": 0.014621224681799428, + "grad_norm": 0.15227355062961578, + "learning_rate": 4.371441637918764e-05, + "loss": 3.1708, + "step": 9974 + }, + { + "epoch": 0.014652071147372846, + "grad_norm": 0.1587970107793808, + "learning_rate": 4.3712631389690054e-05, + "loss": 3.229, + "step": 9975 + }, + { + "epoch": 0.014682917612946262, + "grad_norm": 0.14869268238544464, + "learning_rate": 4.371084618323076e-05, + "loss": 3.1709, + "step": 9976 + }, + { + "epoch": 0.014713764078519678, + "grad_norm": 0.14702759683132172, + "learning_rate": 4.3709060759830446e-05, + "loss": 3.163, + "step": 9977 + }, + { + "epoch": 0.014744610544093095, + "grad_norm": 0.15396186709403992, + "learning_rate": 4.37072751195098e-05, + "loss": 3.1986, + "step": 9978 + }, + { + "epoch": 0.014775457009666511, + "grad_norm": 0.15598943829536438, + "learning_rate": 4.3705489262289566e-05, + "loss": 3.2096, + "step": 9979 + }, + { + "epoch": 0.014806303475239927, + "grad_norm": 0.18143780529499054, + "learning_rate": 4.3703703188190414e-05, + "loss": 3.2274, + "step": 9980 + }, + { + "epoch": 0.014837149940813345, + "grad_norm": 0.15126769244670868, + "learning_rate": 4.3701916897233066e-05, + "loss": 3.1978, + "step": 9981 + }, + { + "epoch": 0.01486799640638676, + "grad_norm": 0.15317678451538086, + "learning_rate": 4.370013038943822e-05, + "loss": 3.1861, + "step": 9982 + }, + { + "epoch": 0.014898842871960177, + "grad_norm": 0.15283700823783875, + "learning_rate": 4.369834366482661e-05, + "loss": 3.2008, + "step": 9983 + }, + { + "epoch": 0.014929689337533594, + "grad_norm": 0.15204741060733795, + "learning_rate": 4.3696556723418935e-05, + "loss": 3.2054, + "step": 9984 + }, + { + "epoch": 0.01496053580310701, + "grad_norm": 0.16205282509326935, + "learning_rate": 4.369476956523593e-05, + "loss": 3.1588, + "step": 9985 + }, + { + "epoch": 0.014991382268680426, + "grad_norm": 0.14966309070587158, + "learning_rate": 4.369298219029829e-05, + "loss": 3.2024, + "step": 9986 + }, + { + "epoch": 0.015022228734253844, + "grad_norm": 0.16148914396762848, + "learning_rate": 4.369119459862677e-05, + "loss": 3.171, + "step": 9987 + }, + { + "epoch": 0.01505307519982726, + "grad_norm": 0.1700209230184555, + "learning_rate": 4.368940679024207e-05, + "loss": 3.1925, + "step": 9988 + }, + { + "epoch": 0.015083921665400676, + "grad_norm": 0.16088876128196716, + "learning_rate": 4.368761876516494e-05, + "loss": 3.1873, + "step": 9989 + }, + { + "epoch": 0.015114768130974093, + "grad_norm": 0.15702341496944427, + "learning_rate": 4.368583052341609e-05, + "loss": 3.1644, + "step": 9990 + }, + { + "epoch": 0.01514561459654751, + "grad_norm": 0.15390844643115997, + "learning_rate": 4.3684042065016276e-05, + "loss": 3.168, + "step": 9991 + }, + { + "epoch": 0.015176461062120925, + "grad_norm": 0.1636146605014801, + "learning_rate": 4.368225338998621e-05, + "loss": 3.1687, + "step": 9992 + }, + { + "epoch": 0.015207307527694343, + "grad_norm": 0.15281377732753754, + "learning_rate": 4.368046449834665e-05, + "loss": 3.1769, + "step": 9993 + }, + { + "epoch": 0.015238153993267759, + "grad_norm": 0.15778806805610657, + "learning_rate": 4.367867539011833e-05, + "loss": 3.1531, + "step": 9994 + }, + { + "epoch": 0.015269000458841175, + "grad_norm": 0.15494650602340698, + "learning_rate": 4.3676886065321985e-05, + "loss": 3.2006, + "step": 9995 + }, + { + "epoch": 0.015299846924414592, + "grad_norm": 0.16416819393634796, + "learning_rate": 4.3675096523978376e-05, + "loss": 3.1526, + "step": 9996 + }, + { + "epoch": 0.015330693389988008, + "grad_norm": 0.20134076476097107, + "learning_rate": 4.367330676610824e-05, + "loss": 3.1927, + "step": 9997 + }, + { + "epoch": 0.015361539855561424, + "grad_norm": 0.16006310284137726, + "learning_rate": 4.367151679173235e-05, + "loss": 3.2071, + "step": 9998 + }, + { + "epoch": 0.015392386321134842, + "grad_norm": 0.15256069600582123, + "learning_rate": 4.366972660087142e-05, + "loss": 3.1947, + "step": 9999 + }, + { + "epoch": 0.015423232786708258, + "grad_norm": 0.15350748598575592, + "learning_rate": 4.366793619354624e-05, + "loss": 3.1843, + "step": 10000 + }, + { + "epoch": 0.015454079252281674, + "grad_norm": 0.15814776718616486, + "learning_rate": 4.3666145569777564e-05, + "loss": 3.1638, + "step": 10001 + }, + { + "epoch": 0.015484925717855091, + "grad_norm": 0.16135616600513458, + "learning_rate": 4.366435472958614e-05, + "loss": 3.1858, + "step": 10002 + }, + { + "epoch": 0.015515772183428507, + "grad_norm": 0.15492412447929382, + "learning_rate": 4.366256367299274e-05, + "loss": 3.2202, + "step": 10003 + }, + { + "epoch": 0.015546618649001923, + "grad_norm": 0.15579882264137268, + "learning_rate": 4.366077240001813e-05, + "loss": 3.1984, + "step": 10004 + }, + { + "epoch": 0.015577465114575341, + "grad_norm": 0.1533801555633545, + "learning_rate": 4.365898091068308e-05, + "loss": 3.1761, + "step": 10005 + }, + { + "epoch": 0.015608311580148757, + "grad_norm": 0.15483956038951874, + "learning_rate": 4.3657189205008356e-05, + "loss": 3.2101, + "step": 10006 + }, + { + "epoch": 0.015639158045722173, + "grad_norm": 0.15258799493312836, + "learning_rate": 4.3655397283014735e-05, + "loss": 3.1969, + "step": 10007 + }, + { + "epoch": 0.01567000451129559, + "grad_norm": 0.15757954120635986, + "learning_rate": 4.3653605144722984e-05, + "loss": 3.2205, + "step": 10008 + }, + { + "epoch": 0.015700850976869008, + "grad_norm": 0.1551590859889984, + "learning_rate": 4.36518127901539e-05, + "loss": 3.2195, + "step": 10009 + }, + { + "epoch": 0.015731697442442422, + "grad_norm": 0.15718920528888702, + "learning_rate": 4.365002021932826e-05, + "loss": 3.1841, + "step": 10010 + }, + { + "epoch": 0.01576254390801584, + "grad_norm": 0.15736541152000427, + "learning_rate": 4.364822743226684e-05, + "loss": 3.2145, + "step": 10011 + }, + { + "epoch": 0.015793390373589258, + "grad_norm": 0.15764909982681274, + "learning_rate": 4.364643442899042e-05, + "loss": 3.2011, + "step": 10012 + }, + { + "epoch": 0.015824236839162672, + "grad_norm": 0.15986879169940948, + "learning_rate": 4.364464120951981e-05, + "loss": 3.1638, + "step": 10013 + }, + { + "epoch": 0.01585508330473609, + "grad_norm": 0.14914613962173462, + "learning_rate": 4.3642847773875786e-05, + "loss": 3.1587, + "step": 10014 + }, + { + "epoch": 0.015885929770309507, + "grad_norm": 0.15553148090839386, + "learning_rate": 4.364105412207914e-05, + "loss": 3.2162, + "step": 10015 + }, + { + "epoch": 0.01591677623588292, + "grad_norm": 0.1496581882238388, + "learning_rate": 4.363926025415068e-05, + "loss": 3.2115, + "step": 10016 + }, + { + "epoch": 0.01594762270145634, + "grad_norm": 0.1473352015018463, + "learning_rate": 4.3637466170111196e-05, + "loss": 3.1979, + "step": 10017 + }, + { + "epoch": 0.015978469167029757, + "grad_norm": 0.15889757871627808, + "learning_rate": 4.363567186998149e-05, + "loss": 3.2074, + "step": 10018 + }, + { + "epoch": 0.01600931563260317, + "grad_norm": 0.15363053977489471, + "learning_rate": 4.363387735378237e-05, + "loss": 3.1774, + "step": 10019 + }, + { + "epoch": 0.01604016209817659, + "grad_norm": 0.15521010756492615, + "learning_rate": 4.363208262153464e-05, + "loss": 3.1677, + "step": 10020 + }, + { + "epoch": 0.016071008563750006, + "grad_norm": 0.16146208345890045, + "learning_rate": 4.363028767325911e-05, + "loss": 3.2244, + "step": 10021 + }, + { + "epoch": 0.01610185502932342, + "grad_norm": 0.15268364548683167, + "learning_rate": 4.3628492508976584e-05, + "loss": 3.1847, + "step": 10022 + }, + { + "epoch": 0.016132701494896838, + "grad_norm": 0.16154341399669647, + "learning_rate": 4.362669712870788e-05, + "loss": 3.2067, + "step": 10023 + }, + { + "epoch": 0.016163547960470256, + "grad_norm": 0.1483083814382553, + "learning_rate": 4.3624901532473824e-05, + "loss": 3.2243, + "step": 10024 + }, + { + "epoch": 0.01619439442604367, + "grad_norm": 0.1596055030822754, + "learning_rate": 4.362310572029522e-05, + "loss": 3.1829, + "step": 10025 + }, + { + "epoch": 0.016225240891617088, + "grad_norm": 0.18385925889015198, + "learning_rate": 4.3621309692192905e-05, + "loss": 3.1736, + "step": 10026 + }, + { + "epoch": 0.016256087357190505, + "grad_norm": 0.15222549438476562, + "learning_rate": 4.3619513448187684e-05, + "loss": 3.22, + "step": 10027 + }, + { + "epoch": 0.01628693382276392, + "grad_norm": 0.1571342945098877, + "learning_rate": 4.3617716988300395e-05, + "loss": 3.2352, + "step": 10028 + }, + { + "epoch": 0.016317780288337337, + "grad_norm": 0.15527907013893127, + "learning_rate": 4.361592031255187e-05, + "loss": 3.2218, + "step": 10029 + }, + { + "epoch": 0.016348626753910755, + "grad_norm": 0.15859410166740417, + "learning_rate": 4.3614123420962936e-05, + "loss": 3.1955, + "step": 10030 + }, + { + "epoch": 0.01637947321948417, + "grad_norm": 0.16392004489898682, + "learning_rate": 4.361232631355443e-05, + "loss": 3.2017, + "step": 10031 + }, + { + "epoch": 0.016410319685057587, + "grad_norm": 0.16983619332313538, + "learning_rate": 4.3610528990347177e-05, + "loss": 3.1841, + "step": 10032 + }, + { + "epoch": 0.016441166150631004, + "grad_norm": 0.1544618159532547, + "learning_rate": 4.360873145136203e-05, + "loss": 3.2116, + "step": 10033 + }, + { + "epoch": 0.01647201261620442, + "grad_norm": 0.14957019686698914, + "learning_rate": 4.3606933696619826e-05, + "loss": 3.2015, + "step": 10034 + }, + { + "epoch": 0.016502859081777836, + "grad_norm": 0.2337682694196701, + "learning_rate": 4.36051357261414e-05, + "loss": 3.2022, + "step": 10035 + }, + { + "epoch": 0.016533705547351254, + "grad_norm": 0.1469356268644333, + "learning_rate": 4.360333753994761e-05, + "loss": 3.1954, + "step": 10036 + }, + { + "epoch": 0.016564552012924668, + "grad_norm": 0.14602814614772797, + "learning_rate": 4.36015391380593e-05, + "loss": 3.1925, + "step": 10037 + }, + { + "epoch": 0.016595398478498086, + "grad_norm": 0.1530967503786087, + "learning_rate": 4.359974052049732e-05, + "loss": 3.2136, + "step": 10038 + }, + { + "epoch": 0.016626244944071503, + "grad_norm": 0.14801926910877228, + "learning_rate": 4.359794168728253e-05, + "loss": 3.2285, + "step": 10039 + }, + { + "epoch": 0.016657091409644918, + "grad_norm": 0.15516935288906097, + "learning_rate": 4.359614263843578e-05, + "loss": 3.1913, + "step": 10040 + }, + { + "epoch": 0.016687937875218335, + "grad_norm": 0.16299304366111755, + "learning_rate": 4.359434337397792e-05, + "loss": 3.1994, + "step": 10041 + }, + { + "epoch": 0.016718784340791753, + "grad_norm": 0.151241734623909, + "learning_rate": 4.359254389392984e-05, + "loss": 3.1882, + "step": 10042 + }, + { + "epoch": 0.016749630806365167, + "grad_norm": 0.16141445934772491, + "learning_rate": 4.359074419831238e-05, + "loss": 3.1993, + "step": 10043 + }, + { + "epoch": 0.016780477271938585, + "grad_norm": 0.19165079295635223, + "learning_rate": 4.358894428714641e-05, + "loss": 3.2042, + "step": 10044 + }, + { + "epoch": 0.016811323737512002, + "grad_norm": 0.14184211194515228, + "learning_rate": 4.35871441604528e-05, + "loss": 3.1754, + "step": 10045 + }, + { + "epoch": 0.016842170203085417, + "grad_norm": 0.1514093279838562, + "learning_rate": 4.358534381825243e-05, + "loss": 3.1838, + "step": 10046 + }, + { + "epoch": 0.016873016668658834, + "grad_norm": 0.1552286446094513, + "learning_rate": 4.3583543260566164e-05, + "loss": 3.2131, + "step": 10047 + }, + { + "epoch": 0.016903863134232252, + "grad_norm": 0.14515990018844604, + "learning_rate": 4.3581742487414875e-05, + "loss": 3.1705, + "step": 10048 + }, + { + "epoch": 0.016934709599805666, + "grad_norm": 0.15558293461799622, + "learning_rate": 4.357994149881946e-05, + "loss": 3.195, + "step": 10049 + }, + { + "epoch": 0.016965556065379084, + "grad_norm": 0.1445617824792862, + "learning_rate": 4.357814029480078e-05, + "loss": 3.1875, + "step": 10050 + }, + { + "epoch": 0.0169964025309525, + "grad_norm": 0.1532091498374939, + "learning_rate": 4.3576338875379734e-05, + "loss": 3.2032, + "step": 10051 + }, + { + "epoch": 0.017027248996525916, + "grad_norm": 0.15031228959560394, + "learning_rate": 4.357453724057719e-05, + "loss": 3.1866, + "step": 10052 + }, + { + "epoch": 0.017058095462099333, + "grad_norm": 0.14705541729927063, + "learning_rate": 4.3572735390414066e-05, + "loss": 3.2225, + "step": 10053 + }, + { + "epoch": 0.01708894192767275, + "grad_norm": 0.18985939025878906, + "learning_rate": 4.3570933324911224e-05, + "loss": 3.1866, + "step": 10054 + }, + { + "epoch": 0.017119788393246165, + "grad_norm": 0.15213347971439362, + "learning_rate": 4.356913104408957e-05, + "loss": 3.2161, + "step": 10055 + }, + { + "epoch": 0.017150634858819583, + "grad_norm": 0.15651708841323853, + "learning_rate": 4.356732854797001e-05, + "loss": 3.2053, + "step": 10056 + }, + { + "epoch": 0.017181481324393, + "grad_norm": 0.15079320967197418, + "learning_rate": 4.356552583657343e-05, + "loss": 3.1777, + "step": 10057 + }, + { + "epoch": 0.017212327789966415, + "grad_norm": 0.1580631136894226, + "learning_rate": 4.356372290992072e-05, + "loss": 3.1905, + "step": 10058 + }, + { + "epoch": 0.017243174255539832, + "grad_norm": 0.21343708038330078, + "learning_rate": 4.356191976803281e-05, + "loss": 3.1497, + "step": 10059 + }, + { + "epoch": 0.01727402072111325, + "grad_norm": 0.14883673191070557, + "learning_rate": 4.3560116410930594e-05, + "loss": 3.1778, + "step": 10060 + }, + { + "epoch": 0.017304867186686664, + "grad_norm": 0.16157639026641846, + "learning_rate": 4.355831283863499e-05, + "loss": 3.1935, + "step": 10061 + }, + { + "epoch": 0.017335713652260082, + "grad_norm": 0.1492876261472702, + "learning_rate": 4.355650905116689e-05, + "loss": 3.234, + "step": 10062 + }, + { + "epoch": 0.0173665601178335, + "grad_norm": 0.1477501392364502, + "learning_rate": 4.355470504854722e-05, + "loss": 3.1919, + "step": 10063 + }, + { + "epoch": 0.017397406583406914, + "grad_norm": 0.15787167847156525, + "learning_rate": 4.355290083079689e-05, + "loss": 3.2025, + "step": 10064 + }, + { + "epoch": 0.01742825304898033, + "grad_norm": 0.1476512998342514, + "learning_rate": 4.355109639793683e-05, + "loss": 3.2195, + "step": 10065 + }, + { + "epoch": 0.01745909951455375, + "grad_norm": 0.15140186250209808, + "learning_rate": 4.354929174998796e-05, + "loss": 3.2057, + "step": 10066 + }, + { + "epoch": 0.017489945980127163, + "grad_norm": 0.1556055098772049, + "learning_rate": 4.354748688697119e-05, + "loss": 3.161, + "step": 10067 + }, + { + "epoch": 0.01752079244570058, + "grad_norm": 0.1535610556602478, + "learning_rate": 4.354568180890746e-05, + "loss": 3.1813, + "step": 10068 + }, + { + "epoch": 0.017551638911274, + "grad_norm": 0.15258745849132538, + "learning_rate": 4.35438765158177e-05, + "loss": 3.1625, + "step": 10069 + }, + { + "epoch": 0.017582485376847413, + "grad_norm": 0.1573907434940338, + "learning_rate": 4.354207100772283e-05, + "loss": 3.1867, + "step": 10070 + }, + { + "epoch": 0.01761333184242083, + "grad_norm": 0.17032389342784882, + "learning_rate": 4.354026528464378e-05, + "loss": 3.2144, + "step": 10071 + }, + { + "epoch": 0.017644178307994248, + "grad_norm": 0.16190817952156067, + "learning_rate": 4.353845934660151e-05, + "loss": 3.16, + "step": 10072 + }, + { + "epoch": 0.017675024773567662, + "grad_norm": 0.16037145256996155, + "learning_rate": 4.353665319361693e-05, + "loss": 3.2073, + "step": 10073 + }, + { + "epoch": 0.01770587123914108, + "grad_norm": 0.21095611155033112, + "learning_rate": 4.353484682571101e-05, + "loss": 3.2285, + "step": 10074 + }, + { + "epoch": 0.017736717704714498, + "grad_norm": 0.161722332239151, + "learning_rate": 4.353304024290466e-05, + "loss": 3.1932, + "step": 10075 + }, + { + "epoch": 0.017767564170287912, + "grad_norm": 0.16141095757484436, + "learning_rate": 4.3531233445218865e-05, + "loss": 3.1902, + "step": 10076 + }, + { + "epoch": 0.01779841063586133, + "grad_norm": 0.20066221058368683, + "learning_rate": 4.352942643267455e-05, + "loss": 3.2087, + "step": 10077 + }, + { + "epoch": 0.017829257101434747, + "grad_norm": 0.16726860404014587, + "learning_rate": 4.352761920529267e-05, + "loss": 3.1639, + "step": 10078 + }, + { + "epoch": 0.01786010356700816, + "grad_norm": 0.16031162440776825, + "learning_rate": 4.352581176309417e-05, + "loss": 3.2136, + "step": 10079 + }, + { + "epoch": 0.01789095003258158, + "grad_norm": 0.1615462601184845, + "learning_rate": 4.352400410610003e-05, + "loss": 3.2277, + "step": 10080 + }, + { + "epoch": 0.017921796498154997, + "grad_norm": 0.14864520728588104, + "learning_rate": 4.352219623433118e-05, + "loss": 3.1694, + "step": 10081 + }, + { + "epoch": 0.01795264296372841, + "grad_norm": 0.16052012145519257, + "learning_rate": 4.35203881478086e-05, + "loss": 3.208, + "step": 10082 + }, + { + "epoch": 0.01798348942930183, + "grad_norm": 0.1481272578239441, + "learning_rate": 4.351857984655325e-05, + "loss": 3.1774, + "step": 10083 + }, + { + "epoch": 0.018014335894875246, + "grad_norm": 0.17478129267692566, + "learning_rate": 4.3516771330586094e-05, + "loss": 3.1709, + "step": 10084 + }, + { + "epoch": 0.01804518236044866, + "grad_norm": 0.15608905255794525, + "learning_rate": 4.35149625999281e-05, + "loss": 3.2357, + "step": 10085 + }, + { + "epoch": 0.018076028826022078, + "grad_norm": 0.15021100640296936, + "learning_rate": 4.351315365460025e-05, + "loss": 3.2194, + "step": 10086 + }, + { + "epoch": 0.018106875291595496, + "grad_norm": 0.15317273139953613, + "learning_rate": 4.35113444946235e-05, + "loss": 3.1947, + "step": 10087 + }, + { + "epoch": 0.01813772175716891, + "grad_norm": 0.15586934983730316, + "learning_rate": 4.350953512001883e-05, + "loss": 3.192, + "step": 10088 + }, + { + "epoch": 0.018168568222742328, + "grad_norm": 0.14789816737174988, + "learning_rate": 4.3507725530807235e-05, + "loss": 3.1559, + "step": 10089 + }, + { + "epoch": 0.018199414688315745, + "grad_norm": 0.15000945329666138, + "learning_rate": 4.350591572700968e-05, + "loss": 3.2106, + "step": 10090 + }, + { + "epoch": 0.01823026115388916, + "grad_norm": 0.17615625262260437, + "learning_rate": 4.350410570864715e-05, + "loss": 3.1837, + "step": 10091 + }, + { + "epoch": 0.018261107619462577, + "grad_norm": 0.1712397336959839, + "learning_rate": 4.350229547574064e-05, + "loss": 3.1551, + "step": 10092 + }, + { + "epoch": 0.018291954085035995, + "grad_norm": 0.1579994261264801, + "learning_rate": 4.350048502831113e-05, + "loss": 3.2029, + "step": 10093 + }, + { + "epoch": 0.01832280055060941, + "grad_norm": 0.1562415361404419, + "learning_rate": 4.349867436637961e-05, + "loss": 3.1756, + "step": 10094 + }, + { + "epoch": 0.018353647016182827, + "grad_norm": 0.1524234414100647, + "learning_rate": 4.3496863489967075e-05, + "loss": 3.1756, + "step": 10095 + }, + { + "epoch": 0.018384493481756244, + "grad_norm": 0.15296058356761932, + "learning_rate": 4.349505239909453e-05, + "loss": 3.1822, + "step": 10096 + }, + { + "epoch": 0.01841533994732966, + "grad_norm": 0.15513963997364044, + "learning_rate": 4.3493241093782964e-05, + "loss": 3.1703, + "step": 10097 + }, + { + "epoch": 0.018446186412903076, + "grad_norm": 0.1524975299835205, + "learning_rate": 4.3491429574053386e-05, + "loss": 3.1907, + "step": 10098 + }, + { + "epoch": 0.018477032878476494, + "grad_norm": 0.1618032604455948, + "learning_rate": 4.3489617839926786e-05, + "loss": 3.1804, + "step": 10099 + }, + { + "epoch": 0.018507879344049908, + "grad_norm": 0.14827501773834229, + "learning_rate": 4.3487805891424176e-05, + "loss": 3.2354, + "step": 10100 + }, + { + "epoch": 0.018538725809623326, + "grad_norm": 0.16332155466079712, + "learning_rate": 4.348599372856658e-05, + "loss": 3.1653, + "step": 10101 + }, + { + "epoch": 0.018569572275196743, + "grad_norm": 0.15121614933013916, + "learning_rate": 4.348418135137498e-05, + "loss": 3.218, + "step": 10102 + }, + { + "epoch": 0.018600418740770158, + "grad_norm": 0.16060401499271393, + "learning_rate": 4.348236875987041e-05, + "loss": 3.1985, + "step": 10103 + }, + { + "epoch": 0.018631265206343575, + "grad_norm": 0.15467962622642517, + "learning_rate": 4.3480555954073885e-05, + "loss": 3.1447, + "step": 10104 + }, + { + "epoch": 0.018662111671916993, + "grad_norm": 1.0527417659759521, + "learning_rate": 4.3478742934006415e-05, + "loss": 3.224, + "step": 10105 + }, + { + "epoch": 0.018692958137490407, + "grad_norm": 0.15247999131679535, + "learning_rate": 4.347692969968903e-05, + "loss": 3.2108, + "step": 10106 + }, + { + "epoch": 0.018723804603063825, + "grad_norm": 0.1573486030101776, + "learning_rate": 4.3475116251142745e-05, + "loss": 3.1955, + "step": 10107 + }, + { + "epoch": 0.018754651068637242, + "grad_norm": 0.1529846340417862, + "learning_rate": 4.3473302588388585e-05, + "loss": 3.1819, + "step": 10108 + }, + { + "epoch": 0.018785497534210657, + "grad_norm": 0.15578106045722961, + "learning_rate": 4.3471488711447597e-05, + "loss": 3.2272, + "step": 10109 + }, + { + "epoch": 0.018816343999784074, + "grad_norm": 0.1503053456544876, + "learning_rate": 4.3469674620340785e-05, + "loss": 3.191, + "step": 10110 + }, + { + "epoch": 0.018847190465357492, + "grad_norm": 0.15555992722511292, + "learning_rate": 4.34678603150892e-05, + "loss": 3.2015, + "step": 10111 + }, + { + "epoch": 0.018878036930930906, + "grad_norm": 0.15795612335205078, + "learning_rate": 4.3466045795713874e-05, + "loss": 3.1799, + "step": 10112 + }, + { + "epoch": 0.018908883396504324, + "grad_norm": 0.14635244011878967, + "learning_rate": 4.3464231062235835e-05, + "loss": 3.1619, + "step": 10113 + }, + { + "epoch": 0.01893972986207774, + "grad_norm": 0.14249810576438904, + "learning_rate": 4.346241611467614e-05, + "loss": 3.1905, + "step": 10114 + }, + { + "epoch": 0.018970576327651156, + "grad_norm": 0.15111328661441803, + "learning_rate": 4.346060095305582e-05, + "loss": 3.2078, + "step": 10115 + }, + { + "epoch": 0.019001422793224573, + "grad_norm": 0.154251366853714, + "learning_rate": 4.345878557739593e-05, + "loss": 3.2069, + "step": 10116 + }, + { + "epoch": 0.01903226925879799, + "grad_norm": 0.1563742309808731, + "learning_rate": 4.345696998771751e-05, + "loss": 3.1929, + "step": 10117 + }, + { + "epoch": 0.019063115724371405, + "grad_norm": 0.1571069210767746, + "learning_rate": 4.3455154184041616e-05, + "loss": 3.2078, + "step": 10118 + }, + { + "epoch": 0.019093962189944823, + "grad_norm": 0.15311264991760254, + "learning_rate": 4.34533381663893e-05, + "loss": 3.1863, + "step": 10119 + }, + { + "epoch": 0.01912480865551824, + "grad_norm": 0.15355122089385986, + "learning_rate": 4.3451521934781617e-05, + "loss": 3.2023, + "step": 10120 + }, + { + "epoch": 0.019155655121091658, + "grad_norm": 0.14984643459320068, + "learning_rate": 4.344970548923962e-05, + "loss": 3.2051, + "step": 10121 + }, + { + "epoch": 0.019186501586665072, + "grad_norm": 0.14821496605873108, + "learning_rate": 4.3447888829784383e-05, + "loss": 3.1613, + "step": 10122 + }, + { + "epoch": 0.01921734805223849, + "grad_norm": 0.15874285995960236, + "learning_rate": 4.344607195643696e-05, + "loss": 3.1718, + "step": 10123 + }, + { + "epoch": 0.019248194517811908, + "grad_norm": 0.150995135307312, + "learning_rate": 4.3444254869218415e-05, + "loss": 3.1452, + "step": 10124 + }, + { + "epoch": 0.019279040983385322, + "grad_norm": 0.156389981508255, + "learning_rate": 4.344243756814982e-05, + "loss": 3.2037, + "step": 10125 + }, + { + "epoch": 0.01930988744895874, + "grad_norm": 0.1627093255519867, + "learning_rate": 4.344062005325224e-05, + "loss": 3.1722, + "step": 10126 + }, + { + "epoch": 0.019340733914532157, + "grad_norm": 0.15060190856456757, + "learning_rate": 4.343880232454676e-05, + "loss": 3.165, + "step": 10127 + }, + { + "epoch": 0.01937158038010557, + "grad_norm": 0.1447807103395462, + "learning_rate": 4.3436984382054435e-05, + "loss": 3.1692, + "step": 10128 + }, + { + "epoch": 0.01940242684567899, + "grad_norm": 0.1592789590358734, + "learning_rate": 4.3435166225796365e-05, + "loss": 3.1772, + "step": 10129 + }, + { + "epoch": 0.019433273311252407, + "grad_norm": 0.14783084392547607, + "learning_rate": 4.343334785579362e-05, + "loss": 3.2049, + "step": 10130 + }, + { + "epoch": 0.01946411977682582, + "grad_norm": 0.15129241347312927, + "learning_rate": 4.343152927206728e-05, + "loss": 3.1976, + "step": 10131 + }, + { + "epoch": 0.01949496624239924, + "grad_norm": 0.1460898369550705, + "learning_rate": 4.342971047463844e-05, + "loss": 3.1728, + "step": 10132 + }, + { + "epoch": 0.019525812707972656, + "grad_norm": 0.15967720746994019, + "learning_rate": 4.342789146352817e-05, + "loss": 3.1918, + "step": 10133 + }, + { + "epoch": 0.01955665917354607, + "grad_norm": 0.14519336819648743, + "learning_rate": 4.3426072238757586e-05, + "loss": 3.1708, + "step": 10134 + }, + { + "epoch": 0.019587505639119488, + "grad_norm": 0.15219712257385254, + "learning_rate": 4.342425280034776e-05, + "loss": 3.1792, + "step": 10135 + }, + { + "epoch": 0.019618352104692906, + "grad_norm": 0.157237246632576, + "learning_rate": 4.3422433148319795e-05, + "loss": 3.1862, + "step": 10136 + }, + { + "epoch": 0.01964919857026632, + "grad_norm": 0.14863230288028717, + "learning_rate": 4.342061328269479e-05, + "loss": 3.1457, + "step": 10137 + }, + { + "epoch": 0.019680045035839738, + "grad_norm": 0.15647754073143005, + "learning_rate": 4.3418793203493844e-05, + "loss": 3.1845, + "step": 10138 + }, + { + "epoch": 0.019710891501413155, + "grad_norm": 0.1463407427072525, + "learning_rate": 4.341697291073806e-05, + "loss": 3.1751, + "step": 10139 + }, + { + "epoch": 0.01974173796698657, + "grad_norm": 0.1499798595905304, + "learning_rate": 4.341515240444854e-05, + "loss": 3.1676, + "step": 10140 + }, + { + "epoch": 0.019772584432559987, + "grad_norm": 0.15098418295383453, + "learning_rate": 4.3413331684646395e-05, + "loss": 3.1742, + "step": 10141 + }, + { + "epoch": 0.019803430898133405, + "grad_norm": 0.14956672489643097, + "learning_rate": 4.341151075135273e-05, + "loss": 3.2372, + "step": 10142 + }, + { + "epoch": 0.01983427736370682, + "grad_norm": 0.1468300223350525, + "learning_rate": 4.340968960458867e-05, + "loss": 3.2273, + "step": 10143 + }, + { + "epoch": 0.019865123829280237, + "grad_norm": 0.14648373425006866, + "learning_rate": 4.3407868244375315e-05, + "loss": 3.2074, + "step": 10144 + }, + { + "epoch": 0.019895970294853654, + "grad_norm": 0.15039144456386566, + "learning_rate": 4.3406046670733797e-05, + "loss": 3.2046, + "step": 10145 + }, + { + "epoch": 0.01992681676042707, + "grad_norm": 0.1519792377948761, + "learning_rate": 4.340422488368522e-05, + "loss": 3.2322, + "step": 10146 + }, + { + "epoch": 0.019957663226000486, + "grad_norm": 0.14617259800434113, + "learning_rate": 4.340240288325072e-05, + "loss": 3.1865, + "step": 10147 + }, + { + "epoch": 0.019988509691573904, + "grad_norm": 0.1410588175058365, + "learning_rate": 4.340058066945142e-05, + "loss": 3.1773, + "step": 10148 + }, + { + "epoch": 0.020019356157147318, + "grad_norm": 0.1514255255460739, + "learning_rate": 4.339875824230843e-05, + "loss": 3.1662, + "step": 10149 + }, + { + "epoch": 0.020050202622720736, + "grad_norm": 0.1504877209663391, + "learning_rate": 4.3396935601842917e-05, + "loss": 3.2047, + "step": 10150 + }, + { + "epoch": 0.020081049088294153, + "grad_norm": 0.14653845131397247, + "learning_rate": 4.339511274807598e-05, + "loss": 3.1714, + "step": 10151 + }, + { + "epoch": 0.020111895553867568, + "grad_norm": 0.1579580307006836, + "learning_rate": 4.339328968102876e-05, + "loss": 3.2004, + "step": 10152 + }, + { + "epoch": 0.020142742019440985, + "grad_norm": 0.1516551673412323, + "learning_rate": 4.3391466400722404e-05, + "loss": 3.195, + "step": 10153 + }, + { + "epoch": 0.020173588485014403, + "grad_norm": 0.14900994300842285, + "learning_rate": 4.338964290717804e-05, + "loss": 3.1634, + "step": 10154 + }, + { + "epoch": 0.020204434950587817, + "grad_norm": 0.1473160982131958, + "learning_rate": 4.338781920041683e-05, + "loss": 3.2106, + "step": 10155 + }, + { + "epoch": 0.020235281416161235, + "grad_norm": 0.14464066922664642, + "learning_rate": 4.33859952804599e-05, + "loss": 3.1877, + "step": 10156 + }, + { + "epoch": 0.020266127881734652, + "grad_norm": 0.1463935822248459, + "learning_rate": 4.33841711473284e-05, + "loss": 3.1747, + "step": 10157 + }, + { + "epoch": 0.020296974347308067, + "grad_norm": 0.14922700822353363, + "learning_rate": 4.3382346801043495e-05, + "loss": 3.1979, + "step": 10158 + }, + { + "epoch": 0.020327820812881484, + "grad_norm": 0.14710308611392975, + "learning_rate": 4.338052224162631e-05, + "loss": 3.2196, + "step": 10159 + }, + { + "epoch": 0.020358667278454902, + "grad_norm": 0.15487217903137207, + "learning_rate": 4.3378697469098024e-05, + "loss": 3.1878, + "step": 10160 + }, + { + "epoch": 0.020389513744028316, + "grad_norm": 0.13976897299289703, + "learning_rate": 4.337687248347978e-05, + "loss": 3.2001, + "step": 10161 + }, + { + "epoch": 0.020420360209601734, + "grad_norm": 0.15215472877025604, + "learning_rate": 4.3375047284792744e-05, + "loss": 3.2133, + "step": 10162 + }, + { + "epoch": 0.02045120667517515, + "grad_norm": 0.14701692759990692, + "learning_rate": 4.3373221873058086e-05, + "loss": 3.193, + "step": 10163 + }, + { + "epoch": 0.020482053140748566, + "grad_norm": 0.1492406278848648, + "learning_rate": 4.337139624829695e-05, + "loss": 3.1602, + "step": 10164 + }, + { + "epoch": 0.020512899606321983, + "grad_norm": 0.14998112618923187, + "learning_rate": 4.336957041053053e-05, + "loss": 3.1566, + "step": 10165 + }, + { + "epoch": 0.0205437460718954, + "grad_norm": 0.14624455571174622, + "learning_rate": 4.3367744359779964e-05, + "loss": 3.1967, + "step": 10166 + }, + { + "epoch": 0.020574592537468815, + "grad_norm": 0.15134583413600922, + "learning_rate": 4.336591809606644e-05, + "loss": 3.1941, + "step": 10167 + }, + { + "epoch": 0.020605439003042233, + "grad_norm": 0.1470748782157898, + "learning_rate": 4.336409161941114e-05, + "loss": 3.2011, + "step": 10168 + }, + { + "epoch": 0.02063628546861565, + "grad_norm": 0.14727000892162323, + "learning_rate": 4.336226492983523e-05, + "loss": 3.1659, + "step": 10169 + }, + { + "epoch": 0.020667131934189065, + "grad_norm": 0.14621904492378235, + "learning_rate": 4.336043802735989e-05, + "loss": 3.2134, + "step": 10170 + }, + { + "epoch": 0.020697978399762482, + "grad_norm": 0.1501152515411377, + "learning_rate": 4.3358610912006306e-05, + "loss": 3.1877, + "step": 10171 + }, + { + "epoch": 0.0207288248653359, + "grad_norm": 0.1455530822277069, + "learning_rate": 4.3356783583795665e-05, + "loss": 3.2045, + "step": 10172 + }, + { + "epoch": 0.020759671330909314, + "grad_norm": 0.14636896550655365, + "learning_rate": 4.3354956042749143e-05, + "loss": 3.2104, + "step": 10173 + }, + { + "epoch": 0.020790517796482732, + "grad_norm": 0.14748252928256989, + "learning_rate": 4.335312828888794e-05, + "loss": 3.1703, + "step": 10174 + }, + { + "epoch": 0.02082136426205615, + "grad_norm": 0.15168800950050354, + "learning_rate": 4.335130032223324e-05, + "loss": 3.1938, + "step": 10175 + }, + { + "epoch": 0.020852210727629564, + "grad_norm": 0.14441828429698944, + "learning_rate": 4.3349472142806244e-05, + "loss": 3.2053, + "step": 10176 + }, + { + "epoch": 0.02088305719320298, + "grad_norm": 0.16809889674186707, + "learning_rate": 4.334764375062814e-05, + "loss": 3.2222, + "step": 10177 + }, + { + "epoch": 0.0209139036587764, + "grad_norm": 0.14760880172252655, + "learning_rate": 4.334581514572014e-05, + "loss": 3.1934, + "step": 10178 + }, + { + "epoch": 0.020944750124349813, + "grad_norm": 0.15311112999916077, + "learning_rate": 4.3343986328103435e-05, + "loss": 3.2045, + "step": 10179 + }, + { + "epoch": 0.02097559658992323, + "grad_norm": 0.1475490927696228, + "learning_rate": 4.3342157297799236e-05, + "loss": 3.1892, + "step": 10180 + }, + { + "epoch": 0.02100644305549665, + "grad_norm": 0.15204250812530518, + "learning_rate": 4.334032805482874e-05, + "loss": 3.199, + "step": 10181 + }, + { + "epoch": 0.021037289521070063, + "grad_norm": 0.1484163999557495, + "learning_rate": 4.333849859921316e-05, + "loss": 3.1683, + "step": 10182 + }, + { + "epoch": 0.02106813598664348, + "grad_norm": 0.1580631285905838, + "learning_rate": 4.333666893097371e-05, + "loss": 3.2115, + "step": 10183 + }, + { + "epoch": 0.021098982452216898, + "grad_norm": 0.150120347738266, + "learning_rate": 4.33348390501316e-05, + "loss": 3.1771, + "step": 10184 + }, + { + "epoch": 0.021129828917790312, + "grad_norm": 0.14955252408981323, + "learning_rate": 4.333300895670806e-05, + "loss": 3.1877, + "step": 10185 + }, + { + "epoch": 0.02116067538336373, + "grad_norm": 0.15064507722854614, + "learning_rate": 4.333117865072429e-05, + "loss": 3.1722, + "step": 10186 + }, + { + "epoch": 0.021191521848937148, + "grad_norm": 0.14857646822929382, + "learning_rate": 4.332934813220152e-05, + "loss": 3.2314, + "step": 10187 + }, + { + "epoch": 0.021222368314510562, + "grad_norm": 0.15311409533023834, + "learning_rate": 4.332751740116098e-05, + "loss": 3.225, + "step": 10188 + }, + { + "epoch": 0.02125321478008398, + "grad_norm": 0.14809317886829376, + "learning_rate": 4.3325686457623874e-05, + "loss": 3.2128, + "step": 10189 + }, + { + "epoch": 0.021284061245657397, + "grad_norm": 0.1599239557981491, + "learning_rate": 4.332385530161146e-05, + "loss": 3.1815, + "step": 10190 + }, + { + "epoch": 0.02131490771123081, + "grad_norm": 0.1497308611869812, + "learning_rate": 4.332202393314495e-05, + "loss": 3.182, + "step": 10191 + }, + { + "epoch": 0.02134575417680423, + "grad_norm": 0.14971664547920227, + "learning_rate": 4.332019235224558e-05, + "loss": 3.1919, + "step": 10192 + }, + { + "epoch": 0.021376600642377647, + "grad_norm": 0.15299320220947266, + "learning_rate": 4.331836055893459e-05, + "loss": 3.2276, + "step": 10193 + }, + { + "epoch": 0.02140744710795106, + "grad_norm": 0.15210887789726257, + "learning_rate": 4.331652855323323e-05, + "loss": 3.2082, + "step": 10194 + }, + { + "epoch": 0.02143829357352448, + "grad_norm": 0.14934934675693512, + "learning_rate": 4.331469633516271e-05, + "loss": 3.1942, + "step": 10195 + }, + { + "epoch": 0.021469140039097896, + "grad_norm": 0.15716935694217682, + "learning_rate": 4.33128639047443e-05, + "loss": 3.1752, + "step": 10196 + }, + { + "epoch": 0.02149998650467131, + "grad_norm": 0.14600156247615814, + "learning_rate": 4.3311031261999234e-05, + "loss": 3.2164, + "step": 10197 + }, + { + "epoch": 0.021530832970244728, + "grad_norm": 0.16072893142700195, + "learning_rate": 4.330919840694877e-05, + "loss": 3.1987, + "step": 10198 + }, + { + "epoch": 0.021561679435818146, + "grad_norm": 0.1482946127653122, + "learning_rate": 4.330736533961415e-05, + "loss": 3.1895, + "step": 10199 + }, + { + "epoch": 0.02159252590139156, + "grad_norm": 0.1521674394607544, + "learning_rate": 4.330553206001663e-05, + "loss": 3.1513, + "step": 10200 + }, + { + "epoch": 0.021623372366964978, + "grad_norm": 0.15202021598815918, + "learning_rate": 4.330369856817746e-05, + "loss": 3.1776, + "step": 10201 + }, + { + "epoch": 0.021654218832538395, + "grad_norm": 0.16194459795951843, + "learning_rate": 4.3301864864117915e-05, + "loss": 3.1933, + "step": 10202 + }, + { + "epoch": 0.02168506529811181, + "grad_norm": 0.15499122440814972, + "learning_rate": 4.330003094785924e-05, + "loss": 3.1764, + "step": 10203 + }, + { + "epoch": 0.021715911763685227, + "grad_norm": 0.15250466763973236, + "learning_rate": 4.329819681942271e-05, + "loss": 3.2183, + "step": 10204 + }, + { + "epoch": 0.021746758229258645, + "grad_norm": 0.17005112767219543, + "learning_rate": 4.329636247882958e-05, + "loss": 3.2392, + "step": 10205 + }, + { + "epoch": 0.02177760469483206, + "grad_norm": 0.15297254920005798, + "learning_rate": 4.329452792610111e-05, + "loss": 3.1985, + "step": 10206 + }, + { + "epoch": 0.021808451160405477, + "grad_norm": 0.15077157318592072, + "learning_rate": 4.32926931612586e-05, + "loss": 3.2254, + "step": 10207 + }, + { + "epoch": 0.021839297625978894, + "grad_norm": 0.15185630321502686, + "learning_rate": 4.3290858184323295e-05, + "loss": 3.2007, + "step": 10208 + }, + { + "epoch": 0.02187014409155231, + "grad_norm": 0.14383965730667114, + "learning_rate": 4.3289022995316485e-05, + "loss": 3.1954, + "step": 10209 + }, + { + "epoch": 0.021900990557125726, + "grad_norm": 0.154164120554924, + "learning_rate": 4.328718759425944e-05, + "loss": 3.21, + "step": 10210 + }, + { + "epoch": 0.021931837022699144, + "grad_norm": 0.14712204039096832, + "learning_rate": 4.328535198117345e-05, + "loss": 3.1759, + "step": 10211 + }, + { + "epoch": 0.021962683488272558, + "grad_norm": 0.14271283149719238, + "learning_rate": 4.328351615607978e-05, + "loss": 3.1727, + "step": 10212 + }, + { + "epoch": 0.021993529953845976, + "grad_norm": 0.15888847410678864, + "learning_rate": 4.328168011899975e-05, + "loss": 3.2239, + "step": 10213 + }, + { + "epoch": 0.022024376419419393, + "grad_norm": 0.14444617927074432, + "learning_rate": 4.3279843869954604e-05, + "loss": 3.162, + "step": 10214 + }, + { + "epoch": 0.022055222884992808, + "grad_norm": 0.14927636086940765, + "learning_rate": 4.327800740896566e-05, + "loss": 3.1835, + "step": 10215 + }, + { + "epoch": 0.022086069350566225, + "grad_norm": 0.14897093176841736, + "learning_rate": 4.3276170736054205e-05, + "loss": 3.1947, + "step": 10216 + }, + { + "epoch": 0.022116915816139643, + "grad_norm": 0.15954282879829407, + "learning_rate": 4.327433385124153e-05, + "loss": 3.1642, + "step": 10217 + }, + { + "epoch": 0.022147762281713057, + "grad_norm": 0.15213891863822937, + "learning_rate": 4.327249675454894e-05, + "loss": 3.1819, + "step": 10218 + }, + { + "epoch": 0.022178608747286475, + "grad_norm": 0.1745249480009079, + "learning_rate": 4.327065944599773e-05, + "loss": 3.1876, + "step": 10219 + }, + { + "epoch": 0.022209455212859892, + "grad_norm": 0.14954344928264618, + "learning_rate": 4.3268821925609204e-05, + "loss": 3.221, + "step": 10220 + }, + { + "epoch": 0.022240301678433307, + "grad_norm": 0.14626263082027435, + "learning_rate": 4.326698419340466e-05, + "loss": 3.1763, + "step": 10221 + }, + { + "epoch": 0.022271148144006724, + "grad_norm": 0.14708013832569122, + "learning_rate": 4.326514624940542e-05, + "loss": 3.1842, + "step": 10222 + }, + { + "epoch": 0.022301994609580142, + "grad_norm": 0.15738366544246674, + "learning_rate": 4.3263308093632775e-05, + "loss": 3.1532, + "step": 10223 + }, + { + "epoch": 0.022332841075153556, + "grad_norm": 0.14508484303951263, + "learning_rate": 4.326146972610805e-05, + "loss": 3.2145, + "step": 10224 + }, + { + "epoch": 0.022363687540726974, + "grad_norm": 0.15244483947753906, + "learning_rate": 4.325963114685256e-05, + "loss": 3.192, + "step": 10225 + }, + { + "epoch": 0.02239453400630039, + "grad_norm": 0.143249973654747, + "learning_rate": 4.3257792355887624e-05, + "loss": 3.1998, + "step": 10226 + }, + { + "epoch": 0.022425380471873806, + "grad_norm": 0.1467970460653305, + "learning_rate": 4.325595335323455e-05, + "loss": 3.2027, + "step": 10227 + }, + { + "epoch": 0.022456226937447223, + "grad_norm": 0.14963914453983307, + "learning_rate": 4.325411413891467e-05, + "loss": 3.1975, + "step": 10228 + }, + { + "epoch": 0.02248707340302064, + "grad_norm": 0.14852046966552734, + "learning_rate": 4.3252274712949314e-05, + "loss": 3.1834, + "step": 10229 + }, + { + "epoch": 0.022517919868594055, + "grad_norm": 0.15135963261127472, + "learning_rate": 4.325043507535979e-05, + "loss": 3.2212, + "step": 10230 + }, + { + "epoch": 0.022548766334167473, + "grad_norm": 0.1475364863872528, + "learning_rate": 4.3248595226167444e-05, + "loss": 3.234, + "step": 10231 + }, + { + "epoch": 0.02257961279974089, + "grad_norm": 0.14173801243305206, + "learning_rate": 4.3246755165393596e-05, + "loss": 3.1824, + "step": 10232 + }, + { + "epoch": 0.022610459265314305, + "grad_norm": 0.17514583468437195, + "learning_rate": 4.3244914893059594e-05, + "loss": 3.1909, + "step": 10233 + }, + { + "epoch": 0.022641305730887722, + "grad_norm": 0.15257738530635834, + "learning_rate": 4.3243074409186765e-05, + "loss": 3.2035, + "step": 10234 + }, + { + "epoch": 0.02267215219646114, + "grad_norm": 0.15154214203357697, + "learning_rate": 4.324123371379645e-05, + "loss": 3.1561, + "step": 10235 + }, + { + "epoch": 0.022702998662034554, + "grad_norm": 0.15453487634658813, + "learning_rate": 4.3239392806909996e-05, + "loss": 3.2011, + "step": 10236 + }, + { + "epoch": 0.022733845127607972, + "grad_norm": 0.14901858568191528, + "learning_rate": 4.323755168854874e-05, + "loss": 3.1911, + "step": 10237 + }, + { + "epoch": 0.02276469159318139, + "grad_norm": 0.15682350099086761, + "learning_rate": 4.3235710358734036e-05, + "loss": 3.1747, + "step": 10238 + }, + { + "epoch": 0.022795538058754804, + "grad_norm": 0.1535535454750061, + "learning_rate": 4.3233868817487224e-05, + "loss": 3.2044, + "step": 10239 + }, + { + "epoch": 0.02282638452432822, + "grad_norm": 0.15095756947994232, + "learning_rate": 4.323202706482966e-05, + "loss": 3.2068, + "step": 10240 + }, + { + "epoch": 0.02285723098990164, + "grad_norm": 0.1445164978504181, + "learning_rate": 4.3230185100782694e-05, + "loss": 3.2149, + "step": 10241 + }, + { + "epoch": 0.022888077455475053, + "grad_norm": 0.15359492599964142, + "learning_rate": 4.3228342925367695e-05, + "loss": 3.1538, + "step": 10242 + }, + { + "epoch": 0.02291892392104847, + "grad_norm": 0.15284065902233124, + "learning_rate": 4.322650053860601e-05, + "loss": 3.1968, + "step": 10243 + }, + { + "epoch": 0.02294977038662189, + "grad_norm": 0.15493467450141907, + "learning_rate": 4.322465794051901e-05, + "loss": 3.1877, + "step": 10244 + }, + { + "epoch": 0.022980616852195303, + "grad_norm": 0.15904498100280762, + "learning_rate": 4.322281513112805e-05, + "loss": 3.1611, + "step": 10245 + }, + { + "epoch": 0.02301146331776872, + "grad_norm": 0.14895139634609222, + "learning_rate": 4.3220972110454494e-05, + "loss": 3.1682, + "step": 10246 + }, + { + "epoch": 0.023042309783342138, + "grad_norm": 0.15339650213718414, + "learning_rate": 4.3219128878519714e-05, + "loss": 3.209, + "step": 10247 + }, + { + "epoch": 0.023073156248915552, + "grad_norm": 0.14849425852298737, + "learning_rate": 4.321728543534509e-05, + "loss": 3.2261, + "step": 10248 + }, + { + "epoch": 0.02310400271448897, + "grad_norm": 0.17704880237579346, + "learning_rate": 4.321544178095199e-05, + "loss": 3.1893, + "step": 10249 + }, + { + "epoch": 0.023134849180062388, + "grad_norm": 0.14845260977745056, + "learning_rate": 4.321359791536178e-05, + "loss": 3.185, + "step": 10250 + }, + { + "epoch": 0.023165695645635802, + "grad_norm": 0.14481167495250702, + "learning_rate": 4.321175383859586e-05, + "loss": 3.1967, + "step": 10251 + }, + { + "epoch": 0.02319654211120922, + "grad_norm": 0.14754290878772736, + "learning_rate": 4.3209909550675586e-05, + "loss": 3.2033, + "step": 10252 + }, + { + "epoch": 0.023227388576782637, + "grad_norm": 0.145193412899971, + "learning_rate": 4.320806505162236e-05, + "loss": 3.1785, + "step": 10253 + }, + { + "epoch": 0.02325823504235605, + "grad_norm": 0.16752029955387115, + "learning_rate": 4.320622034145756e-05, + "loss": 3.1744, + "step": 10254 + }, + { + "epoch": 0.02328908150792947, + "grad_norm": 0.14147025346755981, + "learning_rate": 4.320437542020258e-05, + "loss": 3.1711, + "step": 10255 + }, + { + "epoch": 0.023319927973502887, + "grad_norm": 0.14700038731098175, + "learning_rate": 4.32025302878788e-05, + "loss": 3.1976, + "step": 10256 + }, + { + "epoch": 0.023350774439076304, + "grad_norm": 0.14041486382484436, + "learning_rate": 4.320068494450762e-05, + "loss": 3.2174, + "step": 10257 + }, + { + "epoch": 0.02338162090464972, + "grad_norm": 0.14986349642276764, + "learning_rate": 4.319883939011043e-05, + "loss": 3.1888, + "step": 10258 + }, + { + "epoch": 0.023412467370223136, + "grad_norm": 0.1493278443813324, + "learning_rate": 4.3196993624708646e-05, + "loss": 3.1865, + "step": 10259 + }, + { + "epoch": 0.023443313835796554, + "grad_norm": 0.15336807072162628, + "learning_rate": 4.319514764832365e-05, + "loss": 3.2033, + "step": 10260 + }, + { + "epoch": 0.023474160301369968, + "grad_norm": 0.17061077058315277, + "learning_rate": 4.319330146097685e-05, + "loss": 3.2045, + "step": 10261 + }, + { + "epoch": 0.023505006766943386, + "grad_norm": 0.1530742347240448, + "learning_rate": 4.319145506268966e-05, + "loss": 3.1636, + "step": 10262 + }, + { + "epoch": 0.023535853232516803, + "grad_norm": 0.1446112096309662, + "learning_rate": 4.3189608453483475e-05, + "loss": 3.1646, + "step": 10263 + }, + { + "epoch": 0.023566699698090218, + "grad_norm": 0.14969506859779358, + "learning_rate": 4.3187761633379715e-05, + "loss": 3.1629, + "step": 10264 + }, + { + "epoch": 0.023597546163663635, + "grad_norm": 0.14911110699176788, + "learning_rate": 4.318591460239978e-05, + "loss": 3.178, + "step": 10265 + }, + { + "epoch": 0.023628392629237053, + "grad_norm": 0.15699036419391632, + "learning_rate": 4.31840673605651e-05, + "loss": 3.1748, + "step": 10266 + }, + { + "epoch": 0.023659239094810467, + "grad_norm": 0.21353916823863983, + "learning_rate": 4.318221990789708e-05, + "loss": 3.1887, + "step": 10267 + }, + { + "epoch": 0.023690085560383885, + "grad_norm": 0.14444172382354736, + "learning_rate": 4.318037224441716e-05, + "loss": 3.1846, + "step": 10268 + }, + { + "epoch": 0.023720932025957302, + "grad_norm": 0.1454557180404663, + "learning_rate": 4.3178524370146736e-05, + "loss": 3.1907, + "step": 10269 + }, + { + "epoch": 0.023751778491530717, + "grad_norm": 0.14672957360744476, + "learning_rate": 4.317667628510725e-05, + "loss": 3.177, + "step": 10270 + }, + { + "epoch": 0.023782624957104134, + "grad_norm": 0.1460006684064865, + "learning_rate": 4.317482798932012e-05, + "loss": 3.1635, + "step": 10271 + }, + { + "epoch": 0.023813471422677552, + "grad_norm": 0.15259096026420593, + "learning_rate": 4.317297948280679e-05, + "loss": 3.1948, + "step": 10272 + }, + { + "epoch": 0.023844317888250966, + "grad_norm": 0.1472521424293518, + "learning_rate": 4.317113076558868e-05, + "loss": 3.1753, + "step": 10273 + }, + { + "epoch": 0.023875164353824384, + "grad_norm": 0.1447562575340271, + "learning_rate": 4.316928183768723e-05, + "loss": 3.1816, + "step": 10274 + }, + { + "epoch": 0.0239060108193978, + "grad_norm": 0.15078742802143097, + "learning_rate": 4.316743269912387e-05, + "loss": 3.2087, + "step": 10275 + }, + { + "epoch": 0.023936857284971216, + "grad_norm": 0.14803259074687958, + "learning_rate": 4.316558334992005e-05, + "loss": 3.1779, + "step": 10276 + }, + { + "epoch": 0.023967703750544633, + "grad_norm": 0.16638366878032684, + "learning_rate": 4.316373379009721e-05, + "loss": 3.1513, + "step": 10277 + }, + { + "epoch": 0.02399855021611805, + "grad_norm": 0.15282709896564484, + "learning_rate": 4.3161884019676786e-05, + "loss": 3.1817, + "step": 10278 + }, + { + "epoch": 0.024029396681691465, + "grad_norm": 0.14677295088768005, + "learning_rate": 4.3160034038680234e-05, + "loss": 3.2196, + "step": 10279 + }, + { + "epoch": 0.024060243147264883, + "grad_norm": 0.14993394911289215, + "learning_rate": 4.3158183847129e-05, + "loss": 3.2001, + "step": 10280 + }, + { + "epoch": 0.0240910896128383, + "grad_norm": 0.14894022047519684, + "learning_rate": 4.315633344504453e-05, + "loss": 3.1672, + "step": 10281 + }, + { + "epoch": 0.024121936078411715, + "grad_norm": 0.1468077301979065, + "learning_rate": 4.315448283244829e-05, + "loss": 3.1965, + "step": 10282 + }, + { + "epoch": 0.024152782543985132, + "grad_norm": 0.15272970497608185, + "learning_rate": 4.315263200936173e-05, + "loss": 3.178, + "step": 10283 + }, + { + "epoch": 0.02418362900955855, + "grad_norm": 0.23701220750808716, + "learning_rate": 4.3150780975806315e-05, + "loss": 3.1683, + "step": 10284 + }, + { + "epoch": 0.024214475475131964, + "grad_norm": 0.14616680145263672, + "learning_rate": 4.3148929731803495e-05, + "loss": 3.1952, + "step": 10285 + }, + { + "epoch": 0.024245321940705382, + "grad_norm": 0.15108542144298553, + "learning_rate": 4.314707827737474e-05, + "loss": 3.1959, + "step": 10286 + }, + { + "epoch": 0.0242761684062788, + "grad_norm": 0.1521303802728653, + "learning_rate": 4.3145226612541525e-05, + "loss": 3.2323, + "step": 10287 + }, + { + "epoch": 0.024307014871852214, + "grad_norm": 0.1533684879541397, + "learning_rate": 4.314337473732531e-05, + "loss": 3.1739, + "step": 10288 + }, + { + "epoch": 0.02433786133742563, + "grad_norm": 0.1523647904396057, + "learning_rate": 4.314152265174757e-05, + "loss": 3.1832, + "step": 10289 + }, + { + "epoch": 0.02436870780299905, + "grad_norm": 0.15345124900341034, + "learning_rate": 4.313967035582977e-05, + "loss": 3.2205, + "step": 10290 + }, + { + "epoch": 0.024399554268572463, + "grad_norm": 0.1467706263065338, + "learning_rate": 4.313781784959339e-05, + "loss": 3.2088, + "step": 10291 + }, + { + "epoch": 0.02443040073414588, + "grad_norm": 0.16067148745059967, + "learning_rate": 4.313596513305992e-05, + "loss": 3.1477, + "step": 10292 + }, + { + "epoch": 0.0244612471997193, + "grad_norm": 0.15047918260097504, + "learning_rate": 4.313411220625083e-05, + "loss": 3.1922, + "step": 10293 + }, + { + "epoch": 0.024492093665292713, + "grad_norm": 0.14866100251674652, + "learning_rate": 4.313225906918761e-05, + "loss": 3.205, + "step": 10294 + }, + { + "epoch": 0.02452294013086613, + "grad_norm": 0.14360319077968597, + "learning_rate": 4.3130405721891734e-05, + "loss": 3.1852, + "step": 10295 + }, + { + "epoch": 0.024553786596439548, + "grad_norm": 0.14766563475131989, + "learning_rate": 4.312855216438471e-05, + "loss": 3.1784, + "step": 10296 + }, + { + "epoch": 0.024584633062012962, + "grad_norm": 0.14956825971603394, + "learning_rate": 4.3126698396688e-05, + "loss": 3.1902, + "step": 10297 + }, + { + "epoch": 0.02461547952758638, + "grad_norm": 0.14411704242229462, + "learning_rate": 4.312484441882313e-05, + "loss": 3.1926, + "step": 10298 + }, + { + "epoch": 0.024646325993159798, + "grad_norm": 0.14666664600372314, + "learning_rate": 4.312299023081158e-05, + "loss": 3.2149, + "step": 10299 + }, + { + "epoch": 0.024677172458733212, + "grad_norm": 0.14959481358528137, + "learning_rate": 4.312113583267484e-05, + "loss": 3.1573, + "step": 10300 + }, + { + "epoch": 0.02470801892430663, + "grad_norm": 0.14742574095726013, + "learning_rate": 4.3119281224434424e-05, + "loss": 3.1803, + "step": 10301 + }, + { + "epoch": 0.024738865389880047, + "grad_norm": 0.15307189524173737, + "learning_rate": 4.311742640611183e-05, + "loss": 3.2296, + "step": 10302 + }, + { + "epoch": 0.02476971185545346, + "grad_norm": 0.1470736265182495, + "learning_rate": 4.311557137772857e-05, + "loss": 3.1844, + "step": 10303 + }, + { + "epoch": 0.02480055832102688, + "grad_norm": 0.1576118767261505, + "learning_rate": 4.311371613930614e-05, + "loss": 3.1683, + "step": 10304 + }, + { + "epoch": 0.024831404786600297, + "grad_norm": 0.14961443841457367, + "learning_rate": 4.311186069086606e-05, + "loss": 3.2148, + "step": 10305 + }, + { + "epoch": 0.02486225125217371, + "grad_norm": 0.14396820962429047, + "learning_rate": 4.3110005032429835e-05, + "loss": 3.1666, + "step": 10306 + }, + { + "epoch": 0.02489309771774713, + "grad_norm": 0.14706282317638397, + "learning_rate": 4.310814916401899e-05, + "loss": 3.1756, + "step": 10307 + }, + { + "epoch": 0.024923944183320546, + "grad_norm": 0.14449955523014069, + "learning_rate": 4.310629308565503e-05, + "loss": 3.172, + "step": 10308 + }, + { + "epoch": 0.02495479064889396, + "grad_norm": 0.14438341557979584, + "learning_rate": 4.3104436797359485e-05, + "loss": 3.1938, + "step": 10309 + }, + { + "epoch": 0.024985637114467378, + "grad_norm": 0.14617721736431122, + "learning_rate": 4.310258029915388e-05, + "loss": 3.1571, + "step": 10310 + }, + { + "epoch": 0.025016483580040796, + "grad_norm": 0.1515883207321167, + "learning_rate": 4.3100723591059735e-05, + "loss": 3.1961, + "step": 10311 + }, + { + "epoch": 0.02504733004561421, + "grad_norm": 0.1543775349855423, + "learning_rate": 4.309886667309857e-05, + "loss": 3.1813, + "step": 10312 + }, + { + "epoch": 0.025078176511187628, + "grad_norm": 0.14414578676223755, + "learning_rate": 4.309700954529192e-05, + "loss": 3.1577, + "step": 10313 + }, + { + "epoch": 0.025109022976761045, + "grad_norm": 0.1465790718793869, + "learning_rate": 4.3095152207661324e-05, + "loss": 3.199, + "step": 10314 + }, + { + "epoch": 0.02513986944233446, + "grad_norm": 0.15428149700164795, + "learning_rate": 4.3093294660228314e-05, + "loss": 3.1746, + "step": 10315 + }, + { + "epoch": 0.025170715907907877, + "grad_norm": 0.15255847573280334, + "learning_rate": 4.309143690301443e-05, + "loss": 3.1823, + "step": 10316 + }, + { + "epoch": 0.025201562373481295, + "grad_norm": 0.14524418115615845, + "learning_rate": 4.30895789360412e-05, + "loss": 3.1655, + "step": 10317 + }, + { + "epoch": 0.02523240883905471, + "grad_norm": 0.1425996571779251, + "learning_rate": 4.308772075933017e-05, + "loss": 3.1849, + "step": 10318 + }, + { + "epoch": 0.025263255304628127, + "grad_norm": 0.15152941644191742, + "learning_rate": 4.308586237290289e-05, + "loss": 3.1545, + "step": 10319 + }, + { + "epoch": 0.025294101770201544, + "grad_norm": 0.1554560512304306, + "learning_rate": 4.3084003776780906e-05, + "loss": 3.174, + "step": 10320 + }, + { + "epoch": 0.02532494823577496, + "grad_norm": 0.14950497448444366, + "learning_rate": 4.308214497098577e-05, + "loss": 3.1925, + "step": 10321 + }, + { + "epoch": 0.025355794701348376, + "grad_norm": 0.15633822977542877, + "learning_rate": 4.308028595553902e-05, + "loss": 3.2131, + "step": 10322 + }, + { + "epoch": 0.025386641166921794, + "grad_norm": 0.15710347890853882, + "learning_rate": 4.3078426730462215e-05, + "loss": 3.1697, + "step": 10323 + }, + { + "epoch": 0.025417487632495208, + "grad_norm": 0.17522694170475006, + "learning_rate": 4.307656729577693e-05, + "loss": 3.1974, + "step": 10324 + }, + { + "epoch": 0.025448334098068626, + "grad_norm": 0.15540330111980438, + "learning_rate": 4.3074707651504704e-05, + "loss": 3.2181, + "step": 10325 + }, + { + "epoch": 0.025479180563642043, + "grad_norm": 0.4407714605331421, + "learning_rate": 4.30728477976671e-05, + "loss": 3.1678, + "step": 10326 + }, + { + "epoch": 0.025510027029215458, + "grad_norm": 0.15629610419273376, + "learning_rate": 4.307098773428569e-05, + "loss": 3.2309, + "step": 10327 + }, + { + "epoch": 0.025540873494788875, + "grad_norm": 0.15317417681217194, + "learning_rate": 4.306912746138204e-05, + "loss": 3.1739, + "step": 10328 + }, + { + "epoch": 0.025571719960362293, + "grad_norm": 0.18401184678077698, + "learning_rate": 4.3067266978977716e-05, + "loss": 3.1956, + "step": 10329 + }, + { + "epoch": 0.025602566425935707, + "grad_norm": 0.15656425058841705, + "learning_rate": 4.306540628709428e-05, + "loss": 3.1892, + "step": 10330 + }, + { + "epoch": 0.025633412891509125, + "grad_norm": 0.1504710465669632, + "learning_rate": 4.306354538575332e-05, + "loss": 3.1715, + "step": 10331 + }, + { + "epoch": 0.025664259357082542, + "grad_norm": 0.15078634023666382, + "learning_rate": 4.30616842749764e-05, + "loss": 3.1615, + "step": 10332 + }, + { + "epoch": 0.025695105822655957, + "grad_norm": 0.14811427891254425, + "learning_rate": 4.305982295478512e-05, + "loss": 3.1918, + "step": 10333 + }, + { + "epoch": 0.025725952288229374, + "grad_norm": 0.15630465745925903, + "learning_rate": 4.305796142520103e-05, + "loss": 3.1745, + "step": 10334 + }, + { + "epoch": 0.025756798753802792, + "grad_norm": 0.15863923728466034, + "learning_rate": 4.305609968624573e-05, + "loss": 3.1969, + "step": 10335 + }, + { + "epoch": 0.025787645219376206, + "grad_norm": 0.2106775939464569, + "learning_rate": 4.3054237737940816e-05, + "loss": 3.1909, + "step": 10336 + }, + { + "epoch": 0.025818491684949624, + "grad_norm": 0.15268470346927643, + "learning_rate": 4.3052375580307855e-05, + "loss": 3.1894, + "step": 10337 + }, + { + "epoch": 0.02584933815052304, + "grad_norm": 0.14518381655216217, + "learning_rate": 4.3050513213368446e-05, + "loss": 3.1469, + "step": 10338 + }, + { + "epoch": 0.025880184616096456, + "grad_norm": 0.15645982325077057, + "learning_rate": 4.30486506371442e-05, + "loss": 3.1606, + "step": 10339 + }, + { + "epoch": 0.025911031081669873, + "grad_norm": 0.16315686702728271, + "learning_rate": 4.304678785165668e-05, + "loss": 3.1958, + "step": 10340 + }, + { + "epoch": 0.02594187754724329, + "grad_norm": 0.15355338156223297, + "learning_rate": 4.30449248569275e-05, + "loss": 3.1971, + "step": 10341 + }, + { + "epoch": 0.025972724012816705, + "grad_norm": 0.14389897882938385, + "learning_rate": 4.304306165297827e-05, + "loss": 3.1886, + "step": 10342 + }, + { + "epoch": 0.026003570478390123, + "grad_norm": 0.14787781238555908, + "learning_rate": 4.3041198239830574e-05, + "loss": 3.1962, + "step": 10343 + }, + { + "epoch": 0.02603441694396354, + "grad_norm": 0.14817017316818237, + "learning_rate": 4.303933461750602e-05, + "loss": 3.203, + "step": 10344 + }, + { + "epoch": 0.026065263409536955, + "grad_norm": 0.15479277074337006, + "learning_rate": 4.303747078602623e-05, + "loss": 3.2203, + "step": 10345 + }, + { + "epoch": 0.026096109875110372, + "grad_norm": 0.14518216252326965, + "learning_rate": 4.303560674541281e-05, + "loss": 3.1836, + "step": 10346 + }, + { + "epoch": 0.02612695634068379, + "grad_norm": 0.16376061737537384, + "learning_rate": 4.3033742495687356e-05, + "loss": 3.1982, + "step": 10347 + }, + { + "epoch": 0.026157802806257204, + "grad_norm": 0.1526668518781662, + "learning_rate": 4.303187803687151e-05, + "loss": 3.2123, + "step": 10348 + }, + { + "epoch": 0.026188649271830622, + "grad_norm": 0.14279359579086304, + "learning_rate": 4.303001336898686e-05, + "loss": 3.1922, + "step": 10349 + }, + { + "epoch": 0.02621949573740404, + "grad_norm": 0.16057433187961578, + "learning_rate": 4.302814849205504e-05, + "loss": 3.1871, + "step": 10350 + }, + { + "epoch": 0.026250342202977454, + "grad_norm": 0.14982552826404572, + "learning_rate": 4.302628340609769e-05, + "loss": 3.1622, + "step": 10351 + }, + { + "epoch": 0.02628118866855087, + "grad_norm": 0.1434125155210495, + "learning_rate": 4.302441811113639e-05, + "loss": 3.1732, + "step": 10352 + }, + { + "epoch": 0.02631203513412429, + "grad_norm": 0.1500854194164276, + "learning_rate": 4.302255260719281e-05, + "loss": 3.2042, + "step": 10353 + }, + { + "epoch": 0.026342881599697703, + "grad_norm": 0.1566811501979828, + "learning_rate": 4.3020686894288564e-05, + "loss": 3.2056, + "step": 10354 + }, + { + "epoch": 0.02637372806527112, + "grad_norm": 0.1478518694639206, + "learning_rate": 4.301882097244527e-05, + "loss": 3.1717, + "step": 10355 + }, + { + "epoch": 0.02640457453084454, + "grad_norm": 0.15149520337581635, + "learning_rate": 4.3016954841684584e-05, + "loss": 3.1657, + "step": 10356 + }, + { + "epoch": 0.026435420996417953, + "grad_norm": 0.14513163268566132, + "learning_rate": 4.301508850202814e-05, + "loss": 3.2029, + "step": 10357 + }, + { + "epoch": 0.02646626746199137, + "grad_norm": 0.14448530972003937, + "learning_rate": 4.301322195349756e-05, + "loss": 3.2003, + "step": 10358 + }, + { + "epoch": 0.026497113927564788, + "grad_norm": 0.15563781559467316, + "learning_rate": 4.30113551961145e-05, + "loss": 3.1731, + "step": 10359 + }, + { + "epoch": 0.026527960393138202, + "grad_norm": 0.19404926896095276, + "learning_rate": 4.30094882299006e-05, + "loss": 3.1678, + "step": 10360 + }, + { + "epoch": 0.02655880685871162, + "grad_norm": 0.1584646850824356, + "learning_rate": 4.300762105487751e-05, + "loss": 3.206, + "step": 10361 + }, + { + "epoch": 0.026589653324285038, + "grad_norm": 0.15734072029590607, + "learning_rate": 4.300575367106687e-05, + "loss": 3.1836, + "step": 10362 + }, + { + "epoch": 0.026620499789858452, + "grad_norm": 0.15352804958820343, + "learning_rate": 4.300388607849034e-05, + "loss": 3.194, + "step": 10363 + }, + { + "epoch": 0.02665134625543187, + "grad_norm": 0.14821501076221466, + "learning_rate": 4.3002018277169565e-05, + "loss": 3.1999, + "step": 10364 + }, + { + "epoch": 0.026682192721005287, + "grad_norm": 0.1522713005542755, + "learning_rate": 4.3000150267126204e-05, + "loss": 3.1961, + "step": 10365 + }, + { + "epoch": 0.0267130391865787, + "grad_norm": 0.15482571721076965, + "learning_rate": 4.299828204838193e-05, + "loss": 3.187, + "step": 10366 + }, + { + "epoch": 0.02674388565215212, + "grad_norm": 0.14754784107208252, + "learning_rate": 4.299641362095838e-05, + "loss": 3.1822, + "step": 10367 + }, + { + "epoch": 0.026774732117725537, + "grad_norm": 0.14988821744918823, + "learning_rate": 4.2994544984877236e-05, + "loss": 3.1606, + "step": 10368 + }, + { + "epoch": 0.02680557858329895, + "grad_norm": 0.14497607946395874, + "learning_rate": 4.299267614016015e-05, + "loss": 3.1483, + "step": 10369 + }, + { + "epoch": 0.02683642504887237, + "grad_norm": 0.15224121510982513, + "learning_rate": 4.29908070868288e-05, + "loss": 3.2073, + "step": 10370 + }, + { + "epoch": 0.026867271514445786, + "grad_norm": 0.1474587768316269, + "learning_rate": 4.2988937824904854e-05, + "loss": 3.1918, + "step": 10371 + }, + { + "epoch": 0.0268981179800192, + "grad_norm": 0.14986619353294373, + "learning_rate": 4.2987068354409976e-05, + "loss": 3.1781, + "step": 10372 + }, + { + "epoch": 0.026928964445592618, + "grad_norm": 0.15367980301380157, + "learning_rate": 4.2985198675365855e-05, + "loss": 3.2109, + "step": 10373 + }, + { + "epoch": 0.026959810911166036, + "grad_norm": 0.15337149798870087, + "learning_rate": 4.298332878779416e-05, + "loss": 3.1787, + "step": 10374 + }, + { + "epoch": 0.02699065737673945, + "grad_norm": 0.15607500076293945, + "learning_rate": 4.298145869171658e-05, + "loss": 3.1522, + "step": 10375 + }, + { + "epoch": 0.027021503842312868, + "grad_norm": 0.1486983597278595, + "learning_rate": 4.29795883871548e-05, + "loss": 3.1874, + "step": 10376 + }, + { + "epoch": 0.027052350307886285, + "grad_norm": 0.16323639452457428, + "learning_rate": 4.297771787413048e-05, + "loss": 3.1958, + "step": 10377 + }, + { + "epoch": 0.0270831967734597, + "grad_norm": 0.1473522186279297, + "learning_rate": 4.2975847152665334e-05, + "loss": 3.1782, + "step": 10378 + }, + { + "epoch": 0.027114043239033117, + "grad_norm": 0.15967419743537903, + "learning_rate": 4.297397622278104e-05, + "loss": 3.1838, + "step": 10379 + }, + { + "epoch": 0.027144889704606535, + "grad_norm": 0.15453998744487762, + "learning_rate": 4.297210508449929e-05, + "loss": 3.1974, + "step": 10380 + }, + { + "epoch": 0.02717573617017995, + "grad_norm": 0.1441251039505005, + "learning_rate": 4.297023373784179e-05, + "loss": 3.2328, + "step": 10381 + }, + { + "epoch": 0.027206582635753367, + "grad_norm": 0.15049509704113007, + "learning_rate": 4.296836218283022e-05, + "loss": 3.206, + "step": 10382 + }, + { + "epoch": 0.027237429101326784, + "grad_norm": 0.14704197645187378, + "learning_rate": 4.296649041948629e-05, + "loss": 3.1367, + "step": 10383 + }, + { + "epoch": 0.0272682755669002, + "grad_norm": 0.15122868120670319, + "learning_rate": 4.296461844783171e-05, + "loss": 3.2107, + "step": 10384 + }, + { + "epoch": 0.027299122032473616, + "grad_norm": 0.14594677090644836, + "learning_rate": 4.296274626788816e-05, + "loss": 3.1301, + "step": 10385 + }, + { + "epoch": 0.027329968498047034, + "grad_norm": 0.1515302211046219, + "learning_rate": 4.296087387967736e-05, + "loss": 3.1964, + "step": 10386 + }, + { + "epoch": 0.027360814963620448, + "grad_norm": 0.14751966297626495, + "learning_rate": 4.2959001283221034e-05, + "loss": 3.1759, + "step": 10387 + }, + { + "epoch": 0.027391661429193866, + "grad_norm": 0.14637593924999237, + "learning_rate": 4.295712847854088e-05, + "loss": 3.2268, + "step": 10388 + }, + { + "epoch": 0.027422507894767283, + "grad_norm": 0.15664072334766388, + "learning_rate": 4.295525546565861e-05, + "loss": 3.1545, + "step": 10389 + }, + { + "epoch": 0.027453354360340698, + "grad_norm": 0.14736059308052063, + "learning_rate": 4.2953382244595934e-05, + "loss": 3.1813, + "step": 10390 + }, + { + "epoch": 0.027484200825914115, + "grad_norm": 0.14976215362548828, + "learning_rate": 4.295150881537459e-05, + "loss": 3.1983, + "step": 10391 + }, + { + "epoch": 0.027515047291487533, + "grad_norm": 0.14482228457927704, + "learning_rate": 4.294963517801628e-05, + "loss": 3.2199, + "step": 10392 + }, + { + "epoch": 0.027545893757060947, + "grad_norm": 0.1498156487941742, + "learning_rate": 4.294776133254274e-05, + "loss": 3.1834, + "step": 10393 + }, + { + "epoch": 0.027576740222634365, + "grad_norm": 0.1473272144794464, + "learning_rate": 4.294588727897569e-05, + "loss": 3.1873, + "step": 10394 + }, + { + "epoch": 0.027607586688207782, + "grad_norm": 0.14840058982372284, + "learning_rate": 4.294401301733687e-05, + "loss": 3.1715, + "step": 10395 + }, + { + "epoch": 0.0276384331537812, + "grad_norm": 0.2524338662624359, + "learning_rate": 4.2942138547647994e-05, + "loss": 3.1931, + "step": 10396 + }, + { + "epoch": 0.027669279619354614, + "grad_norm": 0.14971022307872772, + "learning_rate": 4.294026386993081e-05, + "loss": 3.1741, + "step": 10397 + }, + { + "epoch": 0.027700126084928032, + "grad_norm": 0.15312029421329498, + "learning_rate": 4.293838898420704e-05, + "loss": 3.1546, + "step": 10398 + }, + { + "epoch": 0.02773097255050145, + "grad_norm": 0.15523992478847504, + "learning_rate": 4.293651389049843e-05, + "loss": 3.1699, + "step": 10399 + }, + { + "epoch": 0.027761819016074864, + "grad_norm": 0.1533287763595581, + "learning_rate": 4.293463858882673e-05, + "loss": 3.1654, + "step": 10400 + }, + { + "epoch": 0.02779266548164828, + "grad_norm": 0.15568163990974426, + "learning_rate": 4.293276307921367e-05, + "loss": 3.1952, + "step": 10401 + }, + { + "epoch": 0.0278235119472217, + "grad_norm": 0.15110169351100922, + "learning_rate": 4.2930887361680985e-05, + "loss": 3.1764, + "step": 10402 + }, + { + "epoch": 0.027854358412795113, + "grad_norm": 0.1574011594057083, + "learning_rate": 4.292901143625045e-05, + "loss": 3.1548, + "step": 10403 + }, + { + "epoch": 0.02788520487836853, + "grad_norm": 0.19988468289375305, + "learning_rate": 4.29271353029438e-05, + "loss": 3.1837, + "step": 10404 + }, + { + "epoch": 0.02791605134394195, + "grad_norm": 0.14632783830165863, + "learning_rate": 4.292525896178278e-05, + "loss": 3.2356, + "step": 10405 + }, + { + "epoch": 0.027946897809515363, + "grad_norm": 0.1543598473072052, + "learning_rate": 4.292338241278916e-05, + "loss": 3.1615, + "step": 10406 + }, + { + "epoch": 0.02797774427508878, + "grad_norm": 0.15168078243732452, + "learning_rate": 4.292150565598469e-05, + "loss": 3.1695, + "step": 10407 + }, + { + "epoch": 0.028008590740662198, + "grad_norm": 0.1578170359134674, + "learning_rate": 4.291962869139114e-05, + "loss": 3.1884, + "step": 10408 + }, + { + "epoch": 0.028039437206235612, + "grad_norm": 0.15120290219783783, + "learning_rate": 4.291775151903026e-05, + "loss": 3.1638, + "step": 10409 + }, + { + "epoch": 0.02807028367180903, + "grad_norm": 0.15579871833324432, + "learning_rate": 4.2915874138923814e-05, + "loss": 3.1933, + "step": 10410 + }, + { + "epoch": 0.028101130137382448, + "grad_norm": 0.14915788173675537, + "learning_rate": 4.291399655109357e-05, + "loss": 3.2139, + "step": 10411 + }, + { + "epoch": 0.028131976602955862, + "grad_norm": 0.15339119732379913, + "learning_rate": 4.291211875556131e-05, + "loss": 3.2004, + "step": 10412 + }, + { + "epoch": 0.02816282306852928, + "grad_norm": 0.1503378301858902, + "learning_rate": 4.2910240752348796e-05, + "loss": 3.1707, + "step": 10413 + }, + { + "epoch": 0.028193669534102697, + "grad_norm": 0.15063506364822388, + "learning_rate": 4.29083625414778e-05, + "loss": 3.1966, + "step": 10414 + }, + { + "epoch": 0.02822451599967611, + "grad_norm": 0.15331155061721802, + "learning_rate": 4.2906484122970105e-05, + "loss": 3.1779, + "step": 10415 + }, + { + "epoch": 0.02825536246524953, + "grad_norm": 0.14468523859977722, + "learning_rate": 4.290460549684748e-05, + "loss": 3.1895, + "step": 10416 + }, + { + "epoch": 0.028286208930822947, + "grad_norm": 0.15037895739078522, + "learning_rate": 4.290272666313172e-05, + "loss": 3.1958, + "step": 10417 + }, + { + "epoch": 0.02831705539639636, + "grad_norm": 0.1490205079317093, + "learning_rate": 4.2900847621844606e-05, + "loss": 3.1894, + "step": 10418 + }, + { + "epoch": 0.02834790186196978, + "grad_norm": 0.14920221269130707, + "learning_rate": 4.289896837300791e-05, + "loss": 3.2024, + "step": 10419 + }, + { + "epoch": 0.028378748327543196, + "grad_norm": 0.14795035123825073, + "learning_rate": 4.289708891664344e-05, + "loss": 3.18, + "step": 10420 + }, + { + "epoch": 0.02840959479311661, + "grad_norm": 0.15794387459754944, + "learning_rate": 4.2895209252772975e-05, + "loss": 3.1935, + "step": 10421 + }, + { + "epoch": 0.028440441258690028, + "grad_norm": 0.1514630764722824, + "learning_rate": 4.289332938141831e-05, + "loss": 3.1845, + "step": 10422 + }, + { + "epoch": 0.028471287724263446, + "grad_norm": 0.14881831407546997, + "learning_rate": 4.2891449302601255e-05, + "loss": 3.2204, + "step": 10423 + }, + { + "epoch": 0.02850213418983686, + "grad_norm": 0.15540166199207306, + "learning_rate": 4.288956901634359e-05, + "loss": 3.1739, + "step": 10424 + }, + { + "epoch": 0.028532980655410278, + "grad_norm": 0.15182732045650482, + "learning_rate": 4.288768852266712e-05, + "loss": 3.1821, + "step": 10425 + }, + { + "epoch": 0.028563827120983695, + "grad_norm": 0.1481882631778717, + "learning_rate": 4.2885807821593655e-05, + "loss": 3.1735, + "step": 10426 + }, + { + "epoch": 0.02859467358655711, + "grad_norm": 0.1528109908103943, + "learning_rate": 4.288392691314499e-05, + "loss": 3.1583, + "step": 10427 + }, + { + "epoch": 0.028625520052130527, + "grad_norm": 0.14860543608665466, + "learning_rate": 4.288204579734295e-05, + "loss": 3.2033, + "step": 10428 + }, + { + "epoch": 0.028656366517703945, + "grad_norm": 0.14800041913986206, + "learning_rate": 4.288016447420933e-05, + "loss": 3.1778, + "step": 10429 + }, + { + "epoch": 0.02868721298327736, + "grad_norm": 0.15900352597236633, + "learning_rate": 4.287828294376594e-05, + "loss": 3.1946, + "step": 10430 + }, + { + "epoch": 0.028718059448850777, + "grad_norm": 0.148664653301239, + "learning_rate": 4.287640120603461e-05, + "loss": 3.2239, + "step": 10431 + }, + { + "epoch": 0.028748905914424194, + "grad_norm": 0.1489771008491516, + "learning_rate": 4.2874519261037146e-05, + "loss": 3.2057, + "step": 10432 + }, + { + "epoch": 0.02877975237999761, + "grad_norm": 0.14998963475227356, + "learning_rate": 4.287263710879538e-05, + "loss": 3.1843, + "step": 10433 + }, + { + "epoch": 0.028810598845571026, + "grad_norm": 0.16890330612659454, + "learning_rate": 4.287075474933112e-05, + "loss": 3.1841, + "step": 10434 + }, + { + "epoch": 0.028841445311144444, + "grad_norm": 0.1458589881658554, + "learning_rate": 4.286887218266621e-05, + "loss": 3.1325, + "step": 10435 + }, + { + "epoch": 0.028872291776717858, + "grad_norm": 0.1463206708431244, + "learning_rate": 4.2866989408822456e-05, + "loss": 3.1554, + "step": 10436 + }, + { + "epoch": 0.028903138242291276, + "grad_norm": 0.1431787759065628, + "learning_rate": 4.286510642782169e-05, + "loss": 3.1702, + "step": 10437 + }, + { + "epoch": 0.028933984707864693, + "grad_norm": 0.14556272327899933, + "learning_rate": 4.2863223239685756e-05, + "loss": 3.1912, + "step": 10438 + }, + { + "epoch": 0.028964831173438108, + "grad_norm": 0.1468978375196457, + "learning_rate": 4.286133984443649e-05, + "loss": 3.1824, + "step": 10439 + }, + { + "epoch": 0.028995677639011525, + "grad_norm": 0.14408433437347412, + "learning_rate": 4.2859456242095716e-05, + "loss": 3.2079, + "step": 10440 + }, + { + "epoch": 0.029026524104584943, + "grad_norm": 0.14719654619693756, + "learning_rate": 4.285757243268528e-05, + "loss": 3.1721, + "step": 10441 + }, + { + "epoch": 0.029057370570158357, + "grad_norm": 0.15264219045639038, + "learning_rate": 4.285568841622702e-05, + "loss": 3.194, + "step": 10442 + }, + { + "epoch": 0.029088217035731775, + "grad_norm": 0.1508413404226303, + "learning_rate": 4.2853804192742785e-05, + "loss": 3.1508, + "step": 10443 + }, + { + "epoch": 0.029119063501305192, + "grad_norm": 0.15724113583564758, + "learning_rate": 4.285191976225442e-05, + "loss": 3.1969, + "step": 10444 + }, + { + "epoch": 0.029149909966878607, + "grad_norm": 0.14894792437553406, + "learning_rate": 4.285003512478377e-05, + "loss": 3.1883, + "step": 10445 + }, + { + "epoch": 0.029180756432452024, + "grad_norm": 0.14393265545368195, + "learning_rate": 4.28481502803527e-05, + "loss": 3.2041, + "step": 10446 + }, + { + "epoch": 0.029211602898025442, + "grad_norm": 0.14926904439926147, + "learning_rate": 4.284626522898304e-05, + "loss": 3.2224, + "step": 10447 + }, + { + "epoch": 0.029242449363598856, + "grad_norm": 0.14934761822223663, + "learning_rate": 4.284437997069666e-05, + "loss": 3.1665, + "step": 10448 + }, + { + "epoch": 0.029273295829172274, + "grad_norm": 0.14290031790733337, + "learning_rate": 4.284249450551542e-05, + "loss": 3.2024, + "step": 10449 + }, + { + "epoch": 0.02930414229474569, + "grad_norm": 0.15198656916618347, + "learning_rate": 4.284060883346118e-05, + "loss": 3.1947, + "step": 10450 + }, + { + "epoch": 0.029334988760319106, + "grad_norm": 0.14793707430362701, + "learning_rate": 4.2838722954555805e-05, + "loss": 3.1836, + "step": 10451 + }, + { + "epoch": 0.029365835225892523, + "grad_norm": 0.15009988844394684, + "learning_rate": 4.283683686882115e-05, + "loss": 3.1846, + "step": 10452 + }, + { + "epoch": 0.02939668169146594, + "grad_norm": 0.14759869873523712, + "learning_rate": 4.28349505762791e-05, + "loss": 3.1828, + "step": 10453 + }, + { + "epoch": 0.029427528157039355, + "grad_norm": 0.14655789732933044, + "learning_rate": 4.28330640769515e-05, + "loss": 3.195, + "step": 10454 + }, + { + "epoch": 0.029458374622612773, + "grad_norm": 0.15287867188453674, + "learning_rate": 4.283117737086025e-05, + "loss": 3.1697, + "step": 10455 + }, + { + "epoch": 0.02948922108818619, + "grad_norm": 0.14482331275939941, + "learning_rate": 4.282929045802722e-05, + "loss": 3.2017, + "step": 10456 + }, + { + "epoch": 0.029520067553759605, + "grad_norm": 0.15257741510868073, + "learning_rate": 4.282740333847426e-05, + "loss": 3.2166, + "step": 10457 + }, + { + "epoch": 0.029550914019333022, + "grad_norm": 0.15013203024864197, + "learning_rate": 4.282551601222329e-05, + "loss": 3.187, + "step": 10458 + }, + { + "epoch": 0.02958176048490644, + "grad_norm": 0.1484224945306778, + "learning_rate": 4.2823628479296166e-05, + "loss": 3.2122, + "step": 10459 + }, + { + "epoch": 0.029612606950479854, + "grad_norm": 0.1470772922039032, + "learning_rate": 4.282174073971478e-05, + "loss": 3.1881, + "step": 10460 + }, + { + "epoch": 0.029643453416053272, + "grad_norm": 0.1773788332939148, + "learning_rate": 4.281985279350102e-05, + "loss": 3.1763, + "step": 10461 + }, + { + "epoch": 0.02967429988162669, + "grad_norm": 0.14278988540172577, + "learning_rate": 4.2817964640676786e-05, + "loss": 3.1546, + "step": 10462 + }, + { + "epoch": 0.029705146347200104, + "grad_norm": 0.1498429775238037, + "learning_rate": 4.281607628126395e-05, + "loss": 3.1671, + "step": 10463 + }, + { + "epoch": 0.02973599281277352, + "grad_norm": 0.14673863351345062, + "learning_rate": 4.281418771528442e-05, + "loss": 3.192, + "step": 10464 + }, + { + "epoch": 0.02976683927834694, + "grad_norm": 0.1445137858390808, + "learning_rate": 4.2812298942760076e-05, + "loss": 3.179, + "step": 10465 + }, + { + "epoch": 0.029797685743920353, + "grad_norm": 0.146891251206398, + "learning_rate": 4.281040996371284e-05, + "loss": 3.155, + "step": 10466 + }, + { + "epoch": 0.02982853220949377, + "grad_norm": 0.1941874474287033, + "learning_rate": 4.2808520778164605e-05, + "loss": 3.2031, + "step": 10467 + }, + { + "epoch": 0.02985937867506719, + "grad_norm": 0.1424001008272171, + "learning_rate": 4.280663138613727e-05, + "loss": 3.1755, + "step": 10468 + }, + { + "epoch": 0.029890225140640603, + "grad_norm": 0.14740660786628723, + "learning_rate": 4.280474178765274e-05, + "loss": 3.1823, + "step": 10469 + }, + { + "epoch": 0.02992107160621402, + "grad_norm": 0.1466917097568512, + "learning_rate": 4.280285198273294e-05, + "loss": 3.1876, + "step": 10470 + }, + { + "epoch": 0.029951918071787438, + "grad_norm": 0.14738033711910248, + "learning_rate": 4.280096197139976e-05, + "loss": 3.1677, + "step": 10471 + }, + { + "epoch": 0.029982764537360852, + "grad_norm": 0.15253295004367828, + "learning_rate": 4.2799071753675126e-05, + "loss": 3.2219, + "step": 10472 + }, + { + "epoch": 0.03001361100293427, + "grad_norm": 0.15584124624729156, + "learning_rate": 4.2797181329580945e-05, + "loss": 3.1943, + "step": 10473 + }, + { + "epoch": 0.030044457468507688, + "grad_norm": 0.3552749752998352, + "learning_rate": 4.2795290699139144e-05, + "loss": 3.2089, + "step": 10474 + }, + { + "epoch": 0.030075303934081102, + "grad_norm": 0.1556408554315567, + "learning_rate": 4.279339986237165e-05, + "loss": 3.1958, + "step": 10475 + }, + { + "epoch": 0.03010615039965452, + "grad_norm": 0.15783889591693878, + "learning_rate": 4.279150881930038e-05, + "loss": 3.1923, + "step": 10476 + }, + { + "epoch": 0.030136996865227937, + "grad_norm": 0.14784924685955048, + "learning_rate": 4.278961756994724e-05, + "loss": 3.2172, + "step": 10477 + }, + { + "epoch": 0.03016784333080135, + "grad_norm": 0.15851324796676636, + "learning_rate": 4.278772611433418e-05, + "loss": 3.1801, + "step": 10478 + }, + { + "epoch": 0.03019868979637477, + "grad_norm": 0.14815284311771393, + "learning_rate": 4.278583445248313e-05, + "loss": 3.1532, + "step": 10479 + }, + { + "epoch": 0.030229536261948187, + "grad_norm": 0.14865177869796753, + "learning_rate": 4.278394258441602e-05, + "loss": 3.1902, + "step": 10480 + }, + { + "epoch": 0.0302603827275216, + "grad_norm": 0.14865586161613464, + "learning_rate": 4.278205051015477e-05, + "loss": 3.1725, + "step": 10481 + }, + { + "epoch": 0.03029122919309502, + "grad_norm": 0.14876163005828857, + "learning_rate": 4.278015822972134e-05, + "loss": 3.1591, + "step": 10482 + }, + { + "epoch": 0.030322075658668436, + "grad_norm": 0.15731990337371826, + "learning_rate": 4.2778265743137656e-05, + "loss": 3.2287, + "step": 10483 + }, + { + "epoch": 0.03035292212424185, + "grad_norm": 0.14723822474479675, + "learning_rate": 4.2776373050425666e-05, + "loss": 3.1864, + "step": 10484 + }, + { + "epoch": 0.030383768589815268, + "grad_norm": 0.150101900100708, + "learning_rate": 4.277448015160731e-05, + "loss": 3.1533, + "step": 10485 + }, + { + "epoch": 0.030414615055388686, + "grad_norm": 0.18026217818260193, + "learning_rate": 4.277258704670455e-05, + "loss": 3.1841, + "step": 10486 + }, + { + "epoch": 0.0304454615209621, + "grad_norm": 0.17494387924671173, + "learning_rate": 4.277069373573931e-05, + "loss": 3.187, + "step": 10487 + }, + { + "epoch": 0.030476307986535518, + "grad_norm": 0.15561063587665558, + "learning_rate": 4.2768800218733556e-05, + "loss": 3.169, + "step": 10488 + }, + { + "epoch": 0.030507154452108935, + "grad_norm": 0.1553724855184555, + "learning_rate": 4.276690649570925e-05, + "loss": 3.1883, + "step": 10489 + }, + { + "epoch": 0.03053800091768235, + "grad_norm": 0.14783452451229095, + "learning_rate": 4.276501256668833e-05, + "loss": 3.1714, + "step": 10490 + }, + { + "epoch": 0.030568847383255767, + "grad_norm": 0.15297669172286987, + "learning_rate": 4.2763118431692774e-05, + "loss": 3.1833, + "step": 10491 + }, + { + "epoch": 0.030599693848829185, + "grad_norm": 0.14780713617801666, + "learning_rate": 4.276122409074452e-05, + "loss": 3.1698, + "step": 10492 + }, + { + "epoch": 0.0306305403144026, + "grad_norm": 0.14414846897125244, + "learning_rate": 4.275932954386556e-05, + "loss": 3.1499, + "step": 10493 + }, + { + "epoch": 0.030661386779976017, + "grad_norm": 0.16759158670902252, + "learning_rate": 4.275743479107785e-05, + "loss": 3.2148, + "step": 10494 + }, + { + "epoch": 0.030692233245549434, + "grad_norm": 0.15577569603919983, + "learning_rate": 4.2755539832403334e-05, + "loss": 3.1913, + "step": 10495 + }, + { + "epoch": 0.03072307971112285, + "grad_norm": 0.15336304903030396, + "learning_rate": 4.2753644667864016e-05, + "loss": 3.1906, + "step": 10496 + }, + { + "epoch": 0.030753926176696266, + "grad_norm": 0.15854792296886444, + "learning_rate": 4.275174929748186e-05, + "loss": 3.1753, + "step": 10497 + }, + { + "epoch": 0.030784772642269684, + "grad_norm": 0.15394997596740723, + "learning_rate": 4.274985372127883e-05, + "loss": 3.2316, + "step": 10498 + }, + { + "epoch": 0.030815619107843098, + "grad_norm": 0.14863580465316772, + "learning_rate": 4.2747957939276916e-05, + "loss": 3.1828, + "step": 10499 + }, + { + "epoch": 0.030846465573416516, + "grad_norm": 0.16270999610424042, + "learning_rate": 4.27460619514981e-05, + "loss": 3.2026, + "step": 10500 + }, + { + "epoch": 0.030877312038989933, + "grad_norm": 0.15332649648189545, + "learning_rate": 4.274416575796436e-05, + "loss": 3.2225, + "step": 10501 + }, + { + "epoch": 0.030908158504563348, + "grad_norm": 0.1505642831325531, + "learning_rate": 4.274226935869767e-05, + "loss": 3.2143, + "step": 10502 + }, + { + "epoch": 0.030939004970136765, + "grad_norm": 0.16288788616657257, + "learning_rate": 4.2740372753720035e-05, + "loss": 3.1886, + "step": 10503 + }, + { + "epoch": 0.030969851435710183, + "grad_norm": 0.14907194674015045, + "learning_rate": 4.273847594305344e-05, + "loss": 3.2362, + "step": 10504 + }, + { + "epoch": 0.031000697901283597, + "grad_norm": 0.1638263463973999, + "learning_rate": 4.2736578926719876e-05, + "loss": 3.2, + "step": 10505 + }, + { + "epoch": 0.031031544366857015, + "grad_norm": 0.15068049728870392, + "learning_rate": 4.273468170474134e-05, + "loss": 3.1612, + "step": 10506 + }, + { + "epoch": 0.031062390832430432, + "grad_norm": 0.15395981073379517, + "learning_rate": 4.273278427713982e-05, + "loss": 3.2023, + "step": 10507 + }, + { + "epoch": 0.031093237298003847, + "grad_norm": 0.14462408423423767, + "learning_rate": 4.2730886643937326e-05, + "loss": 3.1817, + "step": 10508 + }, + { + "epoch": 0.031124083763577264, + "grad_norm": 0.14851614832878113, + "learning_rate": 4.272898880515586e-05, + "loss": 3.2045, + "step": 10509 + }, + { + "epoch": 0.031154930229150682, + "grad_norm": 0.1421746015548706, + "learning_rate": 4.272709076081741e-05, + "loss": 3.1275, + "step": 10510 + }, + { + "epoch": 0.031185776694724096, + "grad_norm": 0.1494644582271576, + "learning_rate": 4.272519251094401e-05, + "loss": 3.1969, + "step": 10511 + }, + { + "epoch": 0.031216623160297514, + "grad_norm": 0.14457465708255768, + "learning_rate": 4.272329405555765e-05, + "loss": 3.2073, + "step": 10512 + }, + { + "epoch": 0.03124746962587093, + "grad_norm": 0.15185289084911346, + "learning_rate": 4.272139539468034e-05, + "loss": 3.1752, + "step": 10513 + }, + { + "epoch": 0.031278316091444346, + "grad_norm": 0.15025854110717773, + "learning_rate": 4.2719496528334116e-05, + "loss": 3.2349, + "step": 10514 + }, + { + "epoch": 0.03130916255701777, + "grad_norm": 0.14804871380329132, + "learning_rate": 4.271759745654096e-05, + "loss": 3.1494, + "step": 10515 + }, + { + "epoch": 0.03134000902259118, + "grad_norm": 0.15027688443660736, + "learning_rate": 4.2715698179322915e-05, + "loss": 3.1925, + "step": 10516 + }, + { + "epoch": 0.031370855488164595, + "grad_norm": 0.14397287368774414, + "learning_rate": 4.271379869670199e-05, + "loss": 3.1788, + "step": 10517 + }, + { + "epoch": 0.031401701953738016, + "grad_norm": 0.15085352957248688, + "learning_rate": 4.271189900870023e-05, + "loss": 3.2328, + "step": 10518 + }, + { + "epoch": 0.03143254841931143, + "grad_norm": 0.14755068719387054, + "learning_rate": 4.270999911533963e-05, + "loss": 3.2141, + "step": 10519 + }, + { + "epoch": 0.031463394884884845, + "grad_norm": 0.1456715166568756, + "learning_rate": 4.2708099016642235e-05, + "loss": 3.2119, + "step": 10520 + }, + { + "epoch": 0.031494241350458266, + "grad_norm": 0.15842121839523315, + "learning_rate": 4.270619871263007e-05, + "loss": 3.1583, + "step": 10521 + }, + { + "epoch": 0.03152508781603168, + "grad_norm": 0.1502211093902588, + "learning_rate": 4.270429820332518e-05, + "loss": 3.1902, + "step": 10522 + }, + { + "epoch": 0.031555934281605094, + "grad_norm": 0.15582695603370667, + "learning_rate": 4.270239748874959e-05, + "loss": 3.1545, + "step": 10523 + }, + { + "epoch": 0.031586780747178515, + "grad_norm": 0.14471125602722168, + "learning_rate": 4.270049656892534e-05, + "loss": 3.1652, + "step": 10524 + }, + { + "epoch": 0.03161762721275193, + "grad_norm": 0.14980807900428772, + "learning_rate": 4.269859544387446e-05, + "loss": 3.1451, + "step": 10525 + }, + { + "epoch": 0.031648473678325344, + "grad_norm": 0.15120336413383484, + "learning_rate": 4.269669411361901e-05, + "loss": 3.1615, + "step": 10526 + }, + { + "epoch": 0.031679320143898765, + "grad_norm": 0.152284175157547, + "learning_rate": 4.269479257818102e-05, + "loss": 3.1619, + "step": 10527 + }, + { + "epoch": 0.03171016660947218, + "grad_norm": 0.1580820232629776, + "learning_rate": 4.269289083758255e-05, + "loss": 3.1527, + "step": 10528 + }, + { + "epoch": 0.03174101307504559, + "grad_norm": 0.15157045423984528, + "learning_rate": 4.269098889184564e-05, + "loss": 3.1654, + "step": 10529 + }, + { + "epoch": 0.031771859540619014, + "grad_norm": 0.1486513763666153, + "learning_rate": 4.268908674099234e-05, + "loss": 3.1583, + "step": 10530 + }, + { + "epoch": 0.03180270600619243, + "grad_norm": 0.14370737969875336, + "learning_rate": 4.268718438504472e-05, + "loss": 3.1651, + "step": 10531 + }, + { + "epoch": 0.03183355247176584, + "grad_norm": 0.15106487274169922, + "learning_rate": 4.268528182402481e-05, + "loss": 3.1828, + "step": 10532 + }, + { + "epoch": 0.031864398937339264, + "grad_norm": 0.1488841027021408, + "learning_rate": 4.26833790579547e-05, + "loss": 3.1927, + "step": 10533 + }, + { + "epoch": 0.03189524540291268, + "grad_norm": 0.14519625902175903, + "learning_rate": 4.2681476086856434e-05, + "loss": 3.1881, + "step": 10534 + }, + { + "epoch": 0.03192609186848609, + "grad_norm": 0.1461745798587799, + "learning_rate": 4.267957291075207e-05, + "loss": 3.1691, + "step": 10535 + }, + { + "epoch": 0.03195693833405951, + "grad_norm": 0.14841721951961517, + "learning_rate": 4.267766952966369e-05, + "loss": 3.1864, + "step": 10536 + }, + { + "epoch": 0.03198778479963293, + "grad_norm": 0.1428438276052475, + "learning_rate": 4.267576594361336e-05, + "loss": 3.1826, + "step": 10537 + }, + { + "epoch": 0.03201863126520634, + "grad_norm": 0.14577652513980865, + "learning_rate": 4.267386215262313e-05, + "loss": 3.1995, + "step": 10538 + }, + { + "epoch": 0.03204947773077976, + "grad_norm": 0.14947931468486786, + "learning_rate": 4.2671958156715094e-05, + "loss": 3.1753, + "step": 10539 + }, + { + "epoch": 0.03208032419635318, + "grad_norm": 0.15127232670783997, + "learning_rate": 4.267005395591133e-05, + "loss": 3.1715, + "step": 10540 + }, + { + "epoch": 0.03211117066192659, + "grad_norm": 0.15558210015296936, + "learning_rate": 4.26681495502339e-05, + "loss": 3.208, + "step": 10541 + }, + { + "epoch": 0.03214201712750001, + "grad_norm": 0.1490904837846756, + "learning_rate": 4.26662449397049e-05, + "loss": 3.1736, + "step": 10542 + }, + { + "epoch": 0.03217286359307343, + "grad_norm": 0.1511223018169403, + "learning_rate": 4.26643401243464e-05, + "loss": 3.2087, + "step": 10543 + }, + { + "epoch": 0.03220371005864684, + "grad_norm": 0.1510717272758484, + "learning_rate": 4.26624351041805e-05, + "loss": 3.192, + "step": 10544 + }, + { + "epoch": 0.03223455652422026, + "grad_norm": 0.15527918934822083, + "learning_rate": 4.2660529879229275e-05, + "loss": 3.1731, + "step": 10545 + }, + { + "epoch": 0.032265402989793676, + "grad_norm": 0.14414002001285553, + "learning_rate": 4.265862444951482e-05, + "loss": 3.1912, + "step": 10546 + }, + { + "epoch": 0.03229624945536709, + "grad_norm": 0.1600710153579712, + "learning_rate": 4.265671881505922e-05, + "loss": 3.182, + "step": 10547 + }, + { + "epoch": 0.03232709592094051, + "grad_norm": 0.14080271124839783, + "learning_rate": 4.2654812975884585e-05, + "loss": 3.1766, + "step": 10548 + }, + { + "epoch": 0.032357942386513926, + "grad_norm": 0.5160145163536072, + "learning_rate": 4.2652906932012994e-05, + "loss": 3.1967, + "step": 10549 + }, + { + "epoch": 0.03238878885208734, + "grad_norm": 0.14265283942222595, + "learning_rate": 4.265100068346656e-05, + "loss": 3.1825, + "step": 10550 + }, + { + "epoch": 0.03241963531766076, + "grad_norm": 0.15013369917869568, + "learning_rate": 4.264909423026738e-05, + "loss": 3.1974, + "step": 10551 + }, + { + "epoch": 0.032450481783234175, + "grad_norm": 0.14802470803260803, + "learning_rate": 4.264718757243755e-05, + "loss": 3.2202, + "step": 10552 + }, + { + "epoch": 0.03248132824880759, + "grad_norm": 0.14710307121276855, + "learning_rate": 4.264528070999919e-05, + "loss": 3.1293, + "step": 10553 + }, + { + "epoch": 0.03251217471438101, + "grad_norm": 0.14903269708156586, + "learning_rate": 4.264337364297441e-05, + "loss": 3.184, + "step": 10554 + }, + { + "epoch": 0.032543021179954425, + "grad_norm": 0.15360228717327118, + "learning_rate": 4.264146637138531e-05, + "loss": 3.1626, + "step": 10555 + }, + { + "epoch": 0.03257386764552784, + "grad_norm": 0.13891927897930145, + "learning_rate": 4.2639558895254014e-05, + "loss": 3.1835, + "step": 10556 + }, + { + "epoch": 0.03260471411110126, + "grad_norm": 0.151344433426857, + "learning_rate": 4.263765121460263e-05, + "loss": 3.1506, + "step": 10557 + }, + { + "epoch": 0.032635560576674674, + "grad_norm": 0.14757852256298065, + "learning_rate": 4.2635743329453275e-05, + "loss": 3.1523, + "step": 10558 + }, + { + "epoch": 0.03266640704224809, + "grad_norm": 0.14840900897979736, + "learning_rate": 4.2633835239828076e-05, + "loss": 3.1632, + "step": 10559 + }, + { + "epoch": 0.03269725350782151, + "grad_norm": 0.14628438651561737, + "learning_rate": 4.2631926945749154e-05, + "loss": 3.1608, + "step": 10560 + }, + { + "epoch": 0.032728099973394924, + "grad_norm": 0.1468905359506607, + "learning_rate": 4.263001844723863e-05, + "loss": 3.1572, + "step": 10561 + }, + { + "epoch": 0.03275894643896834, + "grad_norm": 0.15069471299648285, + "learning_rate": 4.2628109744318646e-05, + "loss": 3.1744, + "step": 10562 + }, + { + "epoch": 0.03278979290454176, + "grad_norm": 0.15239185094833374, + "learning_rate": 4.262620083701132e-05, + "loss": 3.2314, + "step": 10563 + }, + { + "epoch": 0.03282063937011517, + "grad_norm": 0.14851726591587067, + "learning_rate": 4.262429172533878e-05, + "loss": 3.193, + "step": 10564 + }, + { + "epoch": 0.03285148583568859, + "grad_norm": 0.15102393925189972, + "learning_rate": 4.262238240932318e-05, + "loss": 3.178, + "step": 10565 + }, + { + "epoch": 0.03288233230126201, + "grad_norm": 0.15254628658294678, + "learning_rate": 4.262047288898664e-05, + "loss": 3.1987, + "step": 10566 + }, + { + "epoch": 0.03291317876683542, + "grad_norm": 0.14589793980121613, + "learning_rate": 4.26185631643513e-05, + "loss": 3.1599, + "step": 10567 + }, + { + "epoch": 0.03294402523240884, + "grad_norm": 0.1493920087814331, + "learning_rate": 4.261665323543932e-05, + "loss": 3.1964, + "step": 10568 + }, + { + "epoch": 0.03297487169798226, + "grad_norm": 0.14796669781208038, + "learning_rate": 4.2614743102272826e-05, + "loss": 3.1814, + "step": 10569 + }, + { + "epoch": 0.03300571816355567, + "grad_norm": 0.15041512250900269, + "learning_rate": 4.2612832764873964e-05, + "loss": 3.2138, + "step": 10570 + }, + { + "epoch": 0.03303656462912909, + "grad_norm": 0.14554686844348907, + "learning_rate": 4.26109222232649e-05, + "loss": 3.1674, + "step": 10571 + }, + { + "epoch": 0.03306741109470251, + "grad_norm": 0.14698369801044464, + "learning_rate": 4.260901147746778e-05, + "loss": 3.177, + "step": 10572 + }, + { + "epoch": 0.03309825756027592, + "grad_norm": 0.14885300397872925, + "learning_rate": 4.260710052750474e-05, + "loss": 3.1817, + "step": 10573 + }, + { + "epoch": 0.033129104025849336, + "grad_norm": 0.1476966142654419, + "learning_rate": 4.2605189373397954e-05, + "loss": 3.1895, + "step": 10574 + }, + { + "epoch": 0.03315995049142276, + "grad_norm": 0.15139031410217285, + "learning_rate": 4.260327801516958e-05, + "loss": 3.199, + "step": 10575 + }, + { + "epoch": 0.03319079695699617, + "grad_norm": 0.15499107539653778, + "learning_rate": 4.260136645284177e-05, + "loss": 3.2042, + "step": 10576 + }, + { + "epoch": 0.033221643422569586, + "grad_norm": 0.1495019942522049, + "learning_rate": 4.25994546864367e-05, + "loss": 3.1853, + "step": 10577 + }, + { + "epoch": 0.03325248988814301, + "grad_norm": 0.15032270550727844, + "learning_rate": 4.259754271597652e-05, + "loss": 3.1853, + "step": 10578 + }, + { + "epoch": 0.03328333635371642, + "grad_norm": 0.1438450962305069, + "learning_rate": 4.259563054148342e-05, + "loss": 3.2059, + "step": 10579 + }, + { + "epoch": 0.033314182819289835, + "grad_norm": 0.14335693418979645, + "learning_rate": 4.259371816297955e-05, + "loss": 3.194, + "step": 10580 + }, + { + "epoch": 0.033345029284863256, + "grad_norm": 0.15535134077072144, + "learning_rate": 4.2591805580487096e-05, + "loss": 3.2066, + "step": 10581 + }, + { + "epoch": 0.03337587575043667, + "grad_norm": 0.15279629826545715, + "learning_rate": 4.2589892794028216e-05, + "loss": 3.2029, + "step": 10582 + }, + { + "epoch": 0.033406722216010085, + "grad_norm": 0.15576986968517303, + "learning_rate": 4.258797980362511e-05, + "loss": 3.1708, + "step": 10583 + }, + { + "epoch": 0.033437568681583506, + "grad_norm": 0.15125009417533875, + "learning_rate": 4.2586066609299946e-05, + "loss": 3.1275, + "step": 10584 + }, + { + "epoch": 0.03346841514715692, + "grad_norm": 0.1506127417087555, + "learning_rate": 4.258415321107491e-05, + "loss": 3.1747, + "step": 10585 + }, + { + "epoch": 0.033499261612730334, + "grad_norm": 0.15430012345314026, + "learning_rate": 4.2582239608972176e-05, + "loss": 3.2009, + "step": 10586 + }, + { + "epoch": 0.033530108078303755, + "grad_norm": 0.1530386507511139, + "learning_rate": 4.2580325803013944e-05, + "loss": 3.189, + "step": 10587 + }, + { + "epoch": 0.03356095454387717, + "grad_norm": 0.15084707736968994, + "learning_rate": 4.2578411793222396e-05, + "loss": 3.2056, + "step": 10588 + }, + { + "epoch": 0.033591801009450584, + "grad_norm": 0.14894329011440277, + "learning_rate": 4.257649757961973e-05, + "loss": 3.1551, + "step": 10589 + }, + { + "epoch": 0.033622647475024005, + "grad_norm": 0.14441430568695068, + "learning_rate": 4.257458316222813e-05, + "loss": 3.1612, + "step": 10590 + }, + { + "epoch": 0.03365349394059742, + "grad_norm": 0.14474664628505707, + "learning_rate": 4.257266854106981e-05, + "loss": 3.2099, + "step": 10591 + }, + { + "epoch": 0.03368434040617083, + "grad_norm": 0.15418830513954163, + "learning_rate": 4.2570753716166954e-05, + "loss": 3.2528, + "step": 10592 + }, + { + "epoch": 0.033715186871744254, + "grad_norm": 0.14332543313503265, + "learning_rate": 4.256883868754177e-05, + "loss": 3.2194, + "step": 10593 + }, + { + "epoch": 0.03374603333731767, + "grad_norm": 0.15779314935207367, + "learning_rate": 4.256692345521645e-05, + "loss": 3.1857, + "step": 10594 + }, + { + "epoch": 0.03377687980289108, + "grad_norm": 0.1439167857170105, + "learning_rate": 4.256500801921321e-05, + "loss": 3.2188, + "step": 10595 + }, + { + "epoch": 0.033807726268464504, + "grad_norm": 0.14676937460899353, + "learning_rate": 4.256309237955427e-05, + "loss": 3.1726, + "step": 10596 + }, + { + "epoch": 0.03383857273403792, + "grad_norm": 0.1511286050081253, + "learning_rate": 4.256117653626182e-05, + "loss": 3.1956, + "step": 10597 + }, + { + "epoch": 0.03386941919961133, + "grad_norm": 0.1444367617368698, + "learning_rate": 4.255926048935808e-05, + "loss": 3.1856, + "step": 10598 + }, + { + "epoch": 0.03390026566518475, + "grad_norm": 0.1501750499010086, + "learning_rate": 4.255734423886527e-05, + "loss": 3.1966, + "step": 10599 + }, + { + "epoch": 0.03393111213075817, + "grad_norm": 0.15699507296085358, + "learning_rate": 4.25554277848056e-05, + "loss": 3.1901, + "step": 10600 + }, + { + "epoch": 0.03396195859633158, + "grad_norm": 0.1518840342760086, + "learning_rate": 4.25535111272013e-05, + "loss": 3.1949, + "step": 10601 + }, + { + "epoch": 0.033992805061905, + "grad_norm": 0.1484074741601944, + "learning_rate": 4.255159426607458e-05, + "loss": 3.1619, + "step": 10602 + }, + { + "epoch": 0.03402365152747842, + "grad_norm": 0.1474076360464096, + "learning_rate": 4.254967720144768e-05, + "loss": 3.1776, + "step": 10603 + }, + { + "epoch": 0.03405449799305183, + "grad_norm": 0.1547335982322693, + "learning_rate": 4.2547759933342813e-05, + "loss": 3.1848, + "step": 10604 + }, + { + "epoch": 0.03408534445862525, + "grad_norm": 0.16637735068798065, + "learning_rate": 4.2545842461782214e-05, + "loss": 3.1598, + "step": 10605 + }, + { + "epoch": 0.03411619092419867, + "grad_norm": 0.14374633133411407, + "learning_rate": 4.254392478678812e-05, + "loss": 3.1928, + "step": 10606 + }, + { + "epoch": 0.03414703738977208, + "grad_norm": 0.14577871561050415, + "learning_rate": 4.254200690838276e-05, + "loss": 3.2134, + "step": 10607 + }, + { + "epoch": 0.0341778838553455, + "grad_norm": 0.15385395288467407, + "learning_rate": 4.254008882658837e-05, + "loss": 3.2015, + "step": 10608 + }, + { + "epoch": 0.034208730320918916, + "grad_norm": 0.15593399107456207, + "learning_rate": 4.2538170541427183e-05, + "loss": 3.2044, + "step": 10609 + }, + { + "epoch": 0.03423957678649233, + "grad_norm": 0.14926190674304962, + "learning_rate": 4.253625205292146e-05, + "loss": 3.191, + "step": 10610 + }, + { + "epoch": 0.03427042325206575, + "grad_norm": 0.14556966722011566, + "learning_rate": 4.2534333361093426e-05, + "loss": 3.1921, + "step": 10611 + }, + { + "epoch": 0.034301269717639166, + "grad_norm": 0.14794187247753143, + "learning_rate": 4.253241446596533e-05, + "loss": 3.1845, + "step": 10612 + }, + { + "epoch": 0.03433211618321258, + "grad_norm": 0.14793133735656738, + "learning_rate": 4.253049536755943e-05, + "loss": 3.1769, + "step": 10613 + }, + { + "epoch": 0.034362962648786, + "grad_norm": 0.1437925398349762, + "learning_rate": 4.2528576065897976e-05, + "loss": 3.1886, + "step": 10614 + }, + { + "epoch": 0.034393809114359415, + "grad_norm": 0.15872275829315186, + "learning_rate": 4.252665656100321e-05, + "loss": 3.174, + "step": 10615 + }, + { + "epoch": 0.03442465557993283, + "grad_norm": 0.1473238170146942, + "learning_rate": 4.252473685289739e-05, + "loss": 3.1799, + "step": 10616 + }, + { + "epoch": 0.03445550204550625, + "grad_norm": 0.1539502888917923, + "learning_rate": 4.252281694160278e-05, + "loss": 3.149, + "step": 10617 + }, + { + "epoch": 0.034486348511079665, + "grad_norm": 0.15521323680877686, + "learning_rate": 4.252089682714164e-05, + "loss": 3.1986, + "step": 10618 + }, + { + "epoch": 0.03451719497665308, + "grad_norm": 0.14656329154968262, + "learning_rate": 4.251897650953623e-05, + "loss": 3.1525, + "step": 10619 + }, + { + "epoch": 0.0345480414422265, + "grad_norm": 0.14594018459320068, + "learning_rate": 4.251705598880881e-05, + "loss": 3.2011, + "step": 10620 + }, + { + "epoch": 0.034578887907799914, + "grad_norm": 0.15950244665145874, + "learning_rate": 4.251513526498165e-05, + "loss": 3.1363, + "step": 10621 + }, + { + "epoch": 0.03460973437337333, + "grad_norm": 0.146334707736969, + "learning_rate": 4.2513214338077036e-05, + "loss": 3.216, + "step": 10622 + }, + { + "epoch": 0.03464058083894675, + "grad_norm": 0.16095200181007385, + "learning_rate": 4.2511293208117216e-05, + "loss": 3.1461, + "step": 10623 + }, + { + "epoch": 0.034671427304520164, + "grad_norm": 0.15544192492961884, + "learning_rate": 4.2509371875124474e-05, + "loss": 3.1538, + "step": 10624 + }, + { + "epoch": 0.03470227377009358, + "grad_norm": 0.15086692571640015, + "learning_rate": 4.250745033912109e-05, + "loss": 3.182, + "step": 10625 + }, + { + "epoch": 0.034733120235667, + "grad_norm": 0.1441364884376526, + "learning_rate": 4.250552860012934e-05, + "loss": 3.2134, + "step": 10626 + }, + { + "epoch": 0.03476396670124041, + "grad_norm": 0.14760953187942505, + "learning_rate": 4.25036066581715e-05, + "loss": 3.2078, + "step": 10627 + }, + { + "epoch": 0.03479481316681383, + "grad_norm": 0.149004265666008, + "learning_rate": 4.250168451326987e-05, + "loss": 3.1767, + "step": 10628 + }, + { + "epoch": 0.03482565963238725, + "grad_norm": 0.14588592946529388, + "learning_rate": 4.2499762165446716e-05, + "loss": 3.1783, + "step": 10629 + }, + { + "epoch": 0.03485650609796066, + "grad_norm": 0.15194468200206757, + "learning_rate": 4.249783961472434e-05, + "loss": 3.2154, + "step": 10630 + }, + { + "epoch": 0.03488735256353408, + "grad_norm": 0.15093660354614258, + "learning_rate": 4.249591686112503e-05, + "loss": 3.1853, + "step": 10631 + }, + { + "epoch": 0.0349181990291075, + "grad_norm": 0.15906962752342224, + "learning_rate": 4.249399390467107e-05, + "loss": 3.1871, + "step": 10632 + }, + { + "epoch": 0.03494904549468091, + "grad_norm": 0.1479213684797287, + "learning_rate": 4.249207074538477e-05, + "loss": 3.1443, + "step": 10633 + }, + { + "epoch": 0.034979891960254327, + "grad_norm": 0.14857012033462524, + "learning_rate": 4.249014738328842e-05, + "loss": 3.1955, + "step": 10634 + }, + { + "epoch": 0.03501073842582775, + "grad_norm": 0.14783070981502533, + "learning_rate": 4.248822381840432e-05, + "loss": 3.1797, + "step": 10635 + }, + { + "epoch": 0.03504158489140116, + "grad_norm": 0.16061167418956757, + "learning_rate": 4.248630005075478e-05, + "loss": 3.2113, + "step": 10636 + }, + { + "epoch": 0.035072431356974576, + "grad_norm": 0.14942415058612823, + "learning_rate": 4.248437608036209e-05, + "loss": 3.1846, + "step": 10637 + }, + { + "epoch": 0.035103277822548, + "grad_norm": 0.16740168631076813, + "learning_rate": 4.248245190724857e-05, + "loss": 3.1684, + "step": 10638 + }, + { + "epoch": 0.03513412428812141, + "grad_norm": 0.15220151841640472, + "learning_rate": 4.248052753143653e-05, + "loss": 3.1834, + "step": 10639 + }, + { + "epoch": 0.035164970753694826, + "grad_norm": 0.15067510306835175, + "learning_rate": 4.2478602952948274e-05, + "loss": 3.1482, + "step": 10640 + }, + { + "epoch": 0.03519581721926825, + "grad_norm": 0.15762214362621307, + "learning_rate": 4.247667817180612e-05, + "loss": 3.1801, + "step": 10641 + }, + { + "epoch": 0.03522666368484166, + "grad_norm": 0.14965054392814636, + "learning_rate": 4.247475318803238e-05, + "loss": 3.1835, + "step": 10642 + }, + { + "epoch": 0.035257510150415075, + "grad_norm": 0.1528022140264511, + "learning_rate": 4.247282800164939e-05, + "loss": 3.1786, + "step": 10643 + }, + { + "epoch": 0.035288356615988496, + "grad_norm": 0.15002849698066711, + "learning_rate": 4.2470902612679455e-05, + "loss": 3.1616, + "step": 10644 + }, + { + "epoch": 0.03531920308156191, + "grad_norm": 0.16947415471076965, + "learning_rate": 4.24689770211449e-05, + "loss": 3.2146, + "step": 10645 + }, + { + "epoch": 0.035350049547135325, + "grad_norm": 0.1505000740289688, + "learning_rate": 4.246705122706805e-05, + "loss": 3.1864, + "step": 10646 + }, + { + "epoch": 0.035380896012708746, + "grad_norm": 0.163343146443367, + "learning_rate": 4.2465125230471244e-05, + "loss": 3.1697, + "step": 10647 + }, + { + "epoch": 0.03541174247828216, + "grad_norm": 0.15515394508838654, + "learning_rate": 4.2463199031376805e-05, + "loss": 3.2312, + "step": 10648 + }, + { + "epoch": 0.035442588943855574, + "grad_norm": 0.16431252658367157, + "learning_rate": 4.246127262980706e-05, + "loss": 3.18, + "step": 10649 + }, + { + "epoch": 0.035473435409428995, + "grad_norm": 0.18705102801322937, + "learning_rate": 4.245934602578436e-05, + "loss": 3.1857, + "step": 10650 + }, + { + "epoch": 0.03550428187500241, + "grad_norm": 0.16022977232933044, + "learning_rate": 4.245741921933104e-05, + "loss": 3.1564, + "step": 10651 + }, + { + "epoch": 0.035535128340575824, + "grad_norm": 0.15534088015556335, + "learning_rate": 4.245549221046943e-05, + "loss": 3.1889, + "step": 10652 + }, + { + "epoch": 0.035565974806149245, + "grad_norm": 0.14766636490821838, + "learning_rate": 4.2453564999221875e-05, + "loss": 3.2122, + "step": 10653 + }, + { + "epoch": 0.03559682127172266, + "grad_norm": 0.1668514907360077, + "learning_rate": 4.2451637585610716e-05, + "loss": 3.1918, + "step": 10654 + }, + { + "epoch": 0.03562766773729607, + "grad_norm": 0.14448878169059753, + "learning_rate": 4.244970996965831e-05, + "loss": 3.1478, + "step": 10655 + }, + { + "epoch": 0.035658514202869494, + "grad_norm": 0.15355095267295837, + "learning_rate": 4.244778215138701e-05, + "loss": 3.199, + "step": 10656 + }, + { + "epoch": 0.03568936066844291, + "grad_norm": 0.1503198742866516, + "learning_rate": 4.244585413081916e-05, + "loss": 3.1923, + "step": 10657 + }, + { + "epoch": 0.03572020713401632, + "grad_norm": 0.14549167454242706, + "learning_rate": 4.2443925907977114e-05, + "loss": 3.2236, + "step": 10658 + }, + { + "epoch": 0.035751053599589744, + "grad_norm": 0.14978289604187012, + "learning_rate": 4.244199748288322e-05, + "loss": 3.1867, + "step": 10659 + }, + { + "epoch": 0.03578190006516316, + "grad_norm": 0.1477821320295334, + "learning_rate": 4.244006885555986e-05, + "loss": 3.1755, + "step": 10660 + }, + { + "epoch": 0.03581274653073657, + "grad_norm": 0.14678487181663513, + "learning_rate": 4.243814002602937e-05, + "loss": 3.1719, + "step": 10661 + }, + { + "epoch": 0.03584359299630999, + "grad_norm": 0.14300303161144257, + "learning_rate": 4.2436210994314134e-05, + "loss": 3.1784, + "step": 10662 + }, + { + "epoch": 0.03587443946188341, + "grad_norm": 0.16302375495433807, + "learning_rate": 4.243428176043651e-05, + "loss": 3.1781, + "step": 10663 + }, + { + "epoch": 0.03590528592745682, + "grad_norm": 0.15256460011005402, + "learning_rate": 4.2432352324418856e-05, + "loss": 3.1776, + "step": 10664 + }, + { + "epoch": 0.03593613239303024, + "grad_norm": 0.15112660825252533, + "learning_rate": 4.243042268628357e-05, + "loss": 3.1604, + "step": 10665 + }, + { + "epoch": 0.03596697885860366, + "grad_norm": 0.14576901495456696, + "learning_rate": 4.242849284605299e-05, + "loss": 3.1836, + "step": 10666 + }, + { + "epoch": 0.03599782532417707, + "grad_norm": 0.14832843840122223, + "learning_rate": 4.242656280374952e-05, + "loss": 3.1941, + "step": 10667 + }, + { + "epoch": 0.03602867178975049, + "grad_norm": 0.15699124336242676, + "learning_rate": 4.2424632559395515e-05, + "loss": 3.1872, + "step": 10668 + }, + { + "epoch": 0.03605951825532391, + "grad_norm": 0.16753613948822021, + "learning_rate": 4.242270211301338e-05, + "loss": 3.1928, + "step": 10669 + }, + { + "epoch": 0.03609036472089732, + "grad_norm": 0.15215714275836945, + "learning_rate": 4.2420771464625476e-05, + "loss": 3.1658, + "step": 10670 + }, + { + "epoch": 0.03612121118647074, + "grad_norm": 0.1458410769701004, + "learning_rate": 4.2418840614254196e-05, + "loss": 3.1744, + "step": 10671 + }, + { + "epoch": 0.036152057652044156, + "grad_norm": 0.1462574601173401, + "learning_rate": 4.241690956192192e-05, + "loss": 3.1727, + "step": 10672 + }, + { + "epoch": 0.03618290411761757, + "grad_norm": 0.14691239595413208, + "learning_rate": 4.241497830765105e-05, + "loss": 3.1785, + "step": 10673 + }, + { + "epoch": 0.03621375058319099, + "grad_norm": 0.1471043974161148, + "learning_rate": 4.241304685146398e-05, + "loss": 3.1382, + "step": 10674 + }, + { + "epoch": 0.036244597048764406, + "grad_norm": 0.14657138288021088, + "learning_rate": 4.241111519338309e-05, + "loss": 3.1962, + "step": 10675 + }, + { + "epoch": 0.03627544351433782, + "grad_norm": 0.1420779824256897, + "learning_rate": 4.240918333343078e-05, + "loss": 3.1809, + "step": 10676 + }, + { + "epoch": 0.03630628997991124, + "grad_norm": 0.14866653084754944, + "learning_rate": 4.240725127162945e-05, + "loss": 3.1903, + "step": 10677 + }, + { + "epoch": 0.036337136445484655, + "grad_norm": 0.15255242586135864, + "learning_rate": 4.240531900800151e-05, + "loss": 3.1354, + "step": 10678 + }, + { + "epoch": 0.03636798291105807, + "grad_norm": 0.15100742876529694, + "learning_rate": 4.240338654256934e-05, + "loss": 3.2282, + "step": 10679 + }, + { + "epoch": 0.03639882937663149, + "grad_norm": 0.15455418825149536, + "learning_rate": 4.240145387535538e-05, + "loss": 3.2312, + "step": 10680 + }, + { + "epoch": 0.036429675842204905, + "grad_norm": 0.14393192529678345, + "learning_rate": 4.2399521006382004e-05, + "loss": 3.146, + "step": 10681 + }, + { + "epoch": 0.03646052230777832, + "grad_norm": 0.15062753856182098, + "learning_rate": 4.2397587935671645e-05, + "loss": 3.1829, + "step": 10682 + }, + { + "epoch": 0.03649136877335174, + "grad_norm": 0.14483846724033356, + "learning_rate": 4.239565466324671e-05, + "loss": 3.1825, + "step": 10683 + }, + { + "epoch": 0.036522215238925154, + "grad_norm": 0.14144350588321686, + "learning_rate": 4.239372118912961e-05, + "loss": 3.1533, + "step": 10684 + }, + { + "epoch": 0.03655306170449857, + "grad_norm": 0.14502891898155212, + "learning_rate": 4.239178751334276e-05, + "loss": 3.1976, + "step": 10685 + }, + { + "epoch": 0.03658390817007199, + "grad_norm": 0.154914990067482, + "learning_rate": 4.238985363590859e-05, + "loss": 3.1873, + "step": 10686 + }, + { + "epoch": 0.036614754635645404, + "grad_norm": 0.15127743780612946, + "learning_rate": 4.238791955684952e-05, + "loss": 3.1817, + "step": 10687 + }, + { + "epoch": 0.03664560110121882, + "grad_norm": 0.14106886088848114, + "learning_rate": 4.238598527618797e-05, + "loss": 3.1771, + "step": 10688 + }, + { + "epoch": 0.03667644756679224, + "grad_norm": 0.14045557379722595, + "learning_rate": 4.238405079394637e-05, + "loss": 3.1511, + "step": 10689 + }, + { + "epoch": 0.03670729403236565, + "grad_norm": 0.1520015299320221, + "learning_rate": 4.238211611014715e-05, + "loss": 3.1793, + "step": 10690 + }, + { + "epoch": 0.03673814049793907, + "grad_norm": 0.1554996818304062, + "learning_rate": 4.2380181224812735e-05, + "loss": 3.1718, + "step": 10691 + }, + { + "epoch": 0.03676898696351249, + "grad_norm": 0.1535162776708603, + "learning_rate": 4.237824613796556e-05, + "loss": 3.1505, + "step": 10692 + }, + { + "epoch": 0.0367998334290859, + "grad_norm": 0.1471935361623764, + "learning_rate": 4.237631084962806e-05, + "loss": 3.1853, + "step": 10693 + }, + { + "epoch": 0.03683067989465932, + "grad_norm": 0.14715726673603058, + "learning_rate": 4.237437535982269e-05, + "loss": 3.166, + "step": 10694 + }, + { + "epoch": 0.03686152636023274, + "grad_norm": 0.14734363555908203, + "learning_rate": 4.2372439668571875e-05, + "loss": 3.1739, + "step": 10695 + }, + { + "epoch": 0.03689237282580615, + "grad_norm": 0.14653567969799042, + "learning_rate": 4.2370503775898055e-05, + "loss": 3.1688, + "step": 10696 + }, + { + "epoch": 0.036923219291379566, + "grad_norm": 0.15031351149082184, + "learning_rate": 4.2368567681823687e-05, + "loss": 3.1784, + "step": 10697 + }, + { + "epoch": 0.03695406575695299, + "grad_norm": 0.14848068356513977, + "learning_rate": 4.236663138637121e-05, + "loss": 3.1077, + "step": 10698 + }, + { + "epoch": 0.0369849122225264, + "grad_norm": 0.15077349543571472, + "learning_rate": 4.236469488956309e-05, + "loss": 3.2095, + "step": 10699 + }, + { + "epoch": 0.037015758688099816, + "grad_norm": 0.14989307522773743, + "learning_rate": 4.236275819142176e-05, + "loss": 3.1737, + "step": 10700 + }, + { + "epoch": 0.03704660515367324, + "grad_norm": 0.14450500905513763, + "learning_rate": 4.2360821291969675e-05, + "loss": 3.1775, + "step": 10701 + }, + { + "epoch": 0.03707745161924665, + "grad_norm": 0.15394145250320435, + "learning_rate": 4.235888419122931e-05, + "loss": 3.1808, + "step": 10702 + }, + { + "epoch": 0.037108298084820066, + "grad_norm": 0.14846129715442657, + "learning_rate": 4.2356946889223116e-05, + "loss": 3.1951, + "step": 10703 + }, + { + "epoch": 0.03713914455039349, + "grad_norm": 0.14562077820301056, + "learning_rate": 4.235500938597354e-05, + "loss": 3.1547, + "step": 10704 + }, + { + "epoch": 0.0371699910159669, + "grad_norm": 0.18552321195602417, + "learning_rate": 4.235307168150308e-05, + "loss": 3.1656, + "step": 10705 + }, + { + "epoch": 0.037200837481540315, + "grad_norm": 0.15186619758605957, + "learning_rate": 4.235113377583416e-05, + "loss": 3.1856, + "step": 10706 + }, + { + "epoch": 0.037231683947113736, + "grad_norm": 0.14443929493427277, + "learning_rate": 4.234919566898928e-05, + "loss": 3.1803, + "step": 10707 + }, + { + "epoch": 0.03726253041268715, + "grad_norm": 0.14363348484039307, + "learning_rate": 4.234725736099091e-05, + "loss": 3.1465, + "step": 10708 + }, + { + "epoch": 0.037293376878260565, + "grad_norm": 0.14356806874275208, + "learning_rate": 4.23453188518615e-05, + "loss": 3.1759, + "step": 10709 + }, + { + "epoch": 0.037324223343833986, + "grad_norm": 0.14590971171855927, + "learning_rate": 4.2343380141623554e-05, + "loss": 3.179, + "step": 10710 + }, + { + "epoch": 0.0373550698094074, + "grad_norm": 0.15173298120498657, + "learning_rate": 4.234144123029953e-05, + "loss": 3.1842, + "step": 10711 + }, + { + "epoch": 0.037385916274980814, + "grad_norm": 0.1552470475435257, + "learning_rate": 4.2339502117911915e-05, + "loss": 3.1863, + "step": 10712 + }, + { + "epoch": 0.037416762740554235, + "grad_norm": 0.1467876434326172, + "learning_rate": 4.23375628044832e-05, + "loss": 3.1816, + "step": 10713 + }, + { + "epoch": 0.03744760920612765, + "grad_norm": 0.14859570562839508, + "learning_rate": 4.2335623290035856e-05, + "loss": 3.1523, + "step": 10714 + }, + { + "epoch": 0.037478455671701064, + "grad_norm": 0.15321429073810577, + "learning_rate": 4.233368357459238e-05, + "loss": 3.1693, + "step": 10715 + }, + { + "epoch": 0.037509302137274485, + "grad_norm": 0.14630725979804993, + "learning_rate": 4.233174365817526e-05, + "loss": 3.1869, + "step": 10716 + }, + { + "epoch": 0.0375401486028479, + "grad_norm": 0.1559990495443344, + "learning_rate": 4.232980354080699e-05, + "loss": 3.2025, + "step": 10717 + }, + { + "epoch": 0.03757099506842131, + "grad_norm": 0.15009333193302155, + "learning_rate": 4.232786322251006e-05, + "loss": 3.1768, + "step": 10718 + }, + { + "epoch": 0.037601841533994734, + "grad_norm": 0.15579968690872192, + "learning_rate": 4.232592270330696e-05, + "loss": 3.1689, + "step": 10719 + }, + { + "epoch": 0.03763268799956815, + "grad_norm": 0.14854633808135986, + "learning_rate": 4.2323981983220204e-05, + "loss": 3.1272, + "step": 10720 + }, + { + "epoch": 0.03766353446514156, + "grad_norm": 0.16027618944644928, + "learning_rate": 4.2322041062272295e-05, + "loss": 3.2145, + "step": 10721 + }, + { + "epoch": 0.037694380930714984, + "grad_norm": 0.15638317167758942, + "learning_rate": 4.2320099940485725e-05, + "loss": 3.1785, + "step": 10722 + }, + { + "epoch": 0.0377252273962884, + "grad_norm": 0.15372760593891144, + "learning_rate": 4.2318158617883e-05, + "loss": 3.1643, + "step": 10723 + }, + { + "epoch": 0.03775607386186181, + "grad_norm": 0.14831635355949402, + "learning_rate": 4.231621709448664e-05, + "loss": 3.1876, + "step": 10724 + }, + { + "epoch": 0.03778692032743523, + "grad_norm": 0.14719609916210175, + "learning_rate": 4.2314275370319146e-05, + "loss": 3.1596, + "step": 10725 + }, + { + "epoch": 0.03781776679300865, + "grad_norm": 0.14975576102733612, + "learning_rate": 4.231233344540303e-05, + "loss": 3.1974, + "step": 10726 + }, + { + "epoch": 0.03784861325858206, + "grad_norm": 0.1510949730873108, + "learning_rate": 4.231039131976081e-05, + "loss": 3.1477, + "step": 10727 + }, + { + "epoch": 0.03787945972415548, + "grad_norm": 0.14580030739307404, + "learning_rate": 4.230844899341501e-05, + "loss": 3.1424, + "step": 10728 + }, + { + "epoch": 0.0379103061897289, + "grad_norm": 0.1395585834980011, + "learning_rate": 4.230650646638814e-05, + "loss": 3.1433, + "step": 10729 + }, + { + "epoch": 0.03794115265530231, + "grad_norm": 0.146394744515419, + "learning_rate": 4.230456373870273e-05, + "loss": 3.1794, + "step": 10730 + }, + { + "epoch": 0.03797199912087573, + "grad_norm": 0.17698760330677032, + "learning_rate": 4.230262081038131e-05, + "loss": 3.2008, + "step": 10731 + }, + { + "epoch": 0.03800284558644915, + "grad_norm": 0.1434270143508911, + "learning_rate": 4.2300677681446385e-05, + "loss": 3.1819, + "step": 10732 + }, + { + "epoch": 0.03803369205202256, + "grad_norm": 0.14660947024822235, + "learning_rate": 4.229873435192051e-05, + "loss": 3.2104, + "step": 10733 + }, + { + "epoch": 0.03806453851759598, + "grad_norm": 0.15143749117851257, + "learning_rate": 4.2296790821826206e-05, + "loss": 3.1873, + "step": 10734 + }, + { + "epoch": 0.038095384983169396, + "grad_norm": 0.16825859248638153, + "learning_rate": 4.2294847091186e-05, + "loss": 3.2278, + "step": 10735 + }, + { + "epoch": 0.03812623144874281, + "grad_norm": 0.15706785023212433, + "learning_rate": 4.229290316002244e-05, + "loss": 3.1891, + "step": 10736 + }, + { + "epoch": 0.03815707791431623, + "grad_norm": 0.14667083323001862, + "learning_rate": 4.229095902835805e-05, + "loss": 3.1279, + "step": 10737 + }, + { + "epoch": 0.038187924379889646, + "grad_norm": 0.1441805511713028, + "learning_rate": 4.228901469621539e-05, + "loss": 3.1944, + "step": 10738 + }, + { + "epoch": 0.03821877084546307, + "grad_norm": 0.14340899884700775, + "learning_rate": 4.228707016361699e-05, + "loss": 3.148, + "step": 10739 + }, + { + "epoch": 0.03824961731103648, + "grad_norm": 0.1611872911453247, + "learning_rate": 4.2285125430585405e-05, + "loss": 3.1883, + "step": 10740 + }, + { + "epoch": 0.038280463776609895, + "grad_norm": 0.1474999040365219, + "learning_rate": 4.228318049714317e-05, + "loss": 3.1979, + "step": 10741 + }, + { + "epoch": 0.038311310242183316, + "grad_norm": 0.14490070939064026, + "learning_rate": 4.228123536331285e-05, + "loss": 3.1921, + "step": 10742 + }, + { + "epoch": 0.03834215670775673, + "grad_norm": 0.15430380403995514, + "learning_rate": 4.2279290029116994e-05, + "loss": 3.1703, + "step": 10743 + }, + { + "epoch": 0.038373003173330145, + "grad_norm": 0.14439865946769714, + "learning_rate": 4.227734449457814e-05, + "loss": 3.1394, + "step": 10744 + }, + { + "epoch": 0.038403849638903566, + "grad_norm": 0.1491665095090866, + "learning_rate": 4.2275398759718866e-05, + "loss": 3.1749, + "step": 10745 + }, + { + "epoch": 0.03843469610447698, + "grad_norm": 0.14265117049217224, + "learning_rate": 4.2273452824561736e-05, + "loss": 3.169, + "step": 10746 + }, + { + "epoch": 0.038465542570050394, + "grad_norm": 0.15719915926456451, + "learning_rate": 4.227150668912928e-05, + "loss": 3.1592, + "step": 10747 + }, + { + "epoch": 0.038496389035623815, + "grad_norm": 0.15207785367965698, + "learning_rate": 4.226956035344409e-05, + "loss": 3.185, + "step": 10748 + }, + { + "epoch": 0.03852723550119723, + "grad_norm": 0.14291241765022278, + "learning_rate": 4.2267613817528734e-05, + "loss": 3.1743, + "step": 10749 + }, + { + "epoch": 0.038558081966770644, + "grad_norm": 0.14787913858890533, + "learning_rate": 4.226566708140577e-05, + "loss": 3.16, + "step": 10750 + }, + { + "epoch": 0.038588928432344065, + "grad_norm": 0.28958234190940857, + "learning_rate": 4.226372014509776e-05, + "loss": 3.2191, + "step": 10751 + }, + { + "epoch": 0.03861977489791748, + "grad_norm": 0.15393947064876556, + "learning_rate": 4.2261773008627304e-05, + "loss": 3.1575, + "step": 10752 + }, + { + "epoch": 0.03865062136349089, + "grad_norm": 0.14862190186977386, + "learning_rate": 4.225982567201695e-05, + "loss": 3.1833, + "step": 10753 + }, + { + "epoch": 0.038681467829064314, + "grad_norm": 0.14911694824695587, + "learning_rate": 4.22578781352893e-05, + "loss": 3.1997, + "step": 10754 + }, + { + "epoch": 0.03871231429463773, + "grad_norm": 0.15093378722667694, + "learning_rate": 4.225593039846692e-05, + "loss": 3.1886, + "step": 10755 + }, + { + "epoch": 0.03874316076021114, + "grad_norm": 0.1474570333957672, + "learning_rate": 4.22539824615724e-05, + "loss": 3.2194, + "step": 10756 + }, + { + "epoch": 0.038774007225784564, + "grad_norm": 0.1501884162425995, + "learning_rate": 4.225203432462832e-05, + "loss": 3.1484, + "step": 10757 + }, + { + "epoch": 0.03880485369135798, + "grad_norm": 0.14999422430992126, + "learning_rate": 4.2250085987657265e-05, + "loss": 3.2033, + "step": 10758 + }, + { + "epoch": 0.03883570015693139, + "grad_norm": 0.15922392904758453, + "learning_rate": 4.224813745068184e-05, + "loss": 3.1882, + "step": 10759 + }, + { + "epoch": 0.03886654662250481, + "grad_norm": 0.15151700377464294, + "learning_rate": 4.224618871372461e-05, + "loss": 3.1956, + "step": 10760 + }, + { + "epoch": 0.03889739308807823, + "grad_norm": 0.14431381225585938, + "learning_rate": 4.2244239776808195e-05, + "loss": 3.1386, + "step": 10761 + }, + { + "epoch": 0.03892823955365164, + "grad_norm": 0.15029188990592957, + "learning_rate": 4.2242290639955184e-05, + "loss": 3.1683, + "step": 10762 + }, + { + "epoch": 0.03895908601922506, + "grad_norm": 0.1481025218963623, + "learning_rate": 4.224034130318817e-05, + "loss": 3.1712, + "step": 10763 + }, + { + "epoch": 0.03898993248479848, + "grad_norm": 0.1562187373638153, + "learning_rate": 4.2238391766529764e-05, + "loss": 3.1566, + "step": 10764 + }, + { + "epoch": 0.03902077895037189, + "grad_norm": 0.1452781856060028, + "learning_rate": 4.223644203000255e-05, + "loss": 3.1854, + "step": 10765 + }, + { + "epoch": 0.03905162541594531, + "grad_norm": 0.14622685313224792, + "learning_rate": 4.223449209362916e-05, + "loss": 3.1679, + "step": 10766 + }, + { + "epoch": 0.03908247188151873, + "grad_norm": 0.15313485264778137, + "learning_rate": 4.223254195743219e-05, + "loss": 3.1636, + "step": 10767 + }, + { + "epoch": 0.03911331834709214, + "grad_norm": 0.14202311635017395, + "learning_rate": 4.223059162143426e-05, + "loss": 3.1888, + "step": 10768 + }, + { + "epoch": 0.03914416481266556, + "grad_norm": 0.14503376185894012, + "learning_rate": 4.2228641085657965e-05, + "loss": 3.1895, + "step": 10769 + }, + { + "epoch": 0.039175011278238976, + "grad_norm": 0.1396801918745041, + "learning_rate": 4.2226690350125934e-05, + "loss": 3.2028, + "step": 10770 + }, + { + "epoch": 0.03920585774381239, + "grad_norm": 0.14530189335346222, + "learning_rate": 4.222473941486077e-05, + "loss": 3.1694, + "step": 10771 + }, + { + "epoch": 0.03923670420938581, + "grad_norm": 0.14740689098834991, + "learning_rate": 4.2222788279885116e-05, + "loss": 3.1938, + "step": 10772 + }, + { + "epoch": 0.039267550674959226, + "grad_norm": 0.1413148045539856, + "learning_rate": 4.2220836945221575e-05, + "loss": 3.1911, + "step": 10773 + }, + { + "epoch": 0.03929839714053264, + "grad_norm": 0.14511825144290924, + "learning_rate": 4.2218885410892785e-05, + "loss": 3.2014, + "step": 10774 + }, + { + "epoch": 0.03932924360610606, + "grad_norm": 0.14736883342266083, + "learning_rate": 4.221693367692136e-05, + "loss": 3.1626, + "step": 10775 + }, + { + "epoch": 0.039360090071679475, + "grad_norm": 0.15525837242603302, + "learning_rate": 4.221498174332993e-05, + "loss": 3.2217, + "step": 10776 + }, + { + "epoch": 0.03939093653725289, + "grad_norm": 0.14592891931533813, + "learning_rate": 4.221302961014114e-05, + "loss": 3.2, + "step": 10777 + }, + { + "epoch": 0.03942178300282631, + "grad_norm": 0.143689826130867, + "learning_rate": 4.221107727737762e-05, + "loss": 3.214, + "step": 10778 + }, + { + "epoch": 0.039452629468399725, + "grad_norm": 0.14632625877857208, + "learning_rate": 4.220912474506199e-05, + "loss": 3.1885, + "step": 10779 + }, + { + "epoch": 0.03948347593397314, + "grad_norm": 0.14319206774234772, + "learning_rate": 4.22071720132169e-05, + "loss": 3.1919, + "step": 10780 + }, + { + "epoch": 0.03951432239954656, + "grad_norm": 0.146508127450943, + "learning_rate": 4.2205219081865e-05, + "loss": 3.2183, + "step": 10781 + }, + { + "epoch": 0.039545168865119974, + "grad_norm": 0.14330531656742096, + "learning_rate": 4.220326595102892e-05, + "loss": 3.1675, + "step": 10782 + }, + { + "epoch": 0.03957601533069339, + "grad_norm": 0.14661762118339539, + "learning_rate": 4.220131262073131e-05, + "loss": 3.2148, + "step": 10783 + }, + { + "epoch": 0.03960686179626681, + "grad_norm": 0.15436001121997833, + "learning_rate": 4.2199359090994814e-05, + "loss": 3.1832, + "step": 10784 + }, + { + "epoch": 0.039637708261840224, + "grad_norm": 0.14577800035476685, + "learning_rate": 4.2197405361842084e-05, + "loss": 3.1812, + "step": 10785 + }, + { + "epoch": 0.03966855472741364, + "grad_norm": 0.14338326454162598, + "learning_rate": 4.219545143329578e-05, + "loss": 3.1872, + "step": 10786 + }, + { + "epoch": 0.03969940119298706, + "grad_norm": 0.1407981961965561, + "learning_rate": 4.2193497305378546e-05, + "loss": 3.1543, + "step": 10787 + }, + { + "epoch": 0.03973024765856047, + "grad_norm": 0.1438210904598236, + "learning_rate": 4.219154297811304e-05, + "loss": 3.1599, + "step": 10788 + }, + { + "epoch": 0.03976109412413389, + "grad_norm": 0.2975313365459442, + "learning_rate": 4.2189588451521925e-05, + "loss": 3.2142, + "step": 10789 + }, + { + "epoch": 0.03979194058970731, + "grad_norm": 0.1471734195947647, + "learning_rate": 4.218763372562787e-05, + "loss": 3.1754, + "step": 10790 + }, + { + "epoch": 0.03982278705528072, + "grad_norm": 0.1610306054353714, + "learning_rate": 4.218567880045352e-05, + "loss": 3.1811, + "step": 10791 + }, + { + "epoch": 0.03985363352085414, + "grad_norm": 0.14581897854804993, + "learning_rate": 4.218372367602156e-05, + "loss": 3.1844, + "step": 10792 + }, + { + "epoch": 0.03988447998642756, + "grad_norm": 0.1527847945690155, + "learning_rate": 4.218176835235464e-05, + "loss": 3.1836, + "step": 10793 + }, + { + "epoch": 0.03991532645200097, + "grad_norm": 0.1483609974384308, + "learning_rate": 4.217981282947545e-05, + "loss": 3.1823, + "step": 10794 + }, + { + "epoch": 0.03994617291757439, + "grad_norm": 0.1516396403312683, + "learning_rate": 4.217785710740665e-05, + "loss": 3.183, + "step": 10795 + }, + { + "epoch": 0.03997701938314781, + "grad_norm": 0.15055736899375916, + "learning_rate": 4.2175901186170916e-05, + "loss": 3.1839, + "step": 10796 + }, + { + "epoch": 0.04000786584872122, + "grad_norm": 0.14935612678527832, + "learning_rate": 4.2173945065790935e-05, + "loss": 3.1723, + "step": 10797 + }, + { + "epoch": 0.040038712314294636, + "grad_norm": 0.14996019005775452, + "learning_rate": 4.217198874628938e-05, + "loss": 3.1839, + "step": 10798 + }, + { + "epoch": 0.04006955877986806, + "grad_norm": 0.14491043984889984, + "learning_rate": 4.217003222768894e-05, + "loss": 3.1733, + "step": 10799 + }, + { + "epoch": 0.04010040524544147, + "grad_norm": 0.149569571018219, + "learning_rate": 4.2168075510012285e-05, + "loss": 3.1349, + "step": 10800 + }, + { + "epoch": 0.040131251711014886, + "grad_norm": 0.14517107605934143, + "learning_rate": 4.2166118593282116e-05, + "loss": 3.1736, + "step": 10801 + }, + { + "epoch": 0.04016209817658831, + "grad_norm": 0.15411047637462616, + "learning_rate": 4.216416147752112e-05, + "loss": 3.1971, + "step": 10802 + }, + { + "epoch": 0.04019294464216172, + "grad_norm": 0.14541830122470856, + "learning_rate": 4.2162204162751985e-05, + "loss": 3.1502, + "step": 10803 + }, + { + "epoch": 0.040223791107735135, + "grad_norm": 0.14954374730587006, + "learning_rate": 4.2160246648997406e-05, + "loss": 3.2049, + "step": 10804 + }, + { + "epoch": 0.040254637573308556, + "grad_norm": 0.1511160433292389, + "learning_rate": 4.215828893628008e-05, + "loss": 3.1842, + "step": 10805 + }, + { + "epoch": 0.04028548403888197, + "grad_norm": 0.1504942625761032, + "learning_rate": 4.21563310246227e-05, + "loss": 3.1828, + "step": 10806 + }, + { + "epoch": 0.040316330504455385, + "grad_norm": 0.1476057767868042, + "learning_rate": 4.2154372914047966e-05, + "loss": 3.1414, + "step": 10807 + }, + { + "epoch": 0.040347176970028806, + "grad_norm": 0.15106439590454102, + "learning_rate": 4.2152414604578594e-05, + "loss": 3.2022, + "step": 10808 + }, + { + "epoch": 0.04037802343560222, + "grad_norm": 0.14425374567508698, + "learning_rate": 4.2150456096237283e-05, + "loss": 3.1804, + "step": 10809 + }, + { + "epoch": 0.040408869901175634, + "grad_norm": 0.14435729384422302, + "learning_rate": 4.2148497389046735e-05, + "loss": 3.1863, + "step": 10810 + }, + { + "epoch": 0.040439716366749055, + "grad_norm": 0.15235444903373718, + "learning_rate": 4.214653848302968e-05, + "loss": 3.1668, + "step": 10811 + }, + { + "epoch": 0.04047056283232247, + "grad_norm": 0.14916440844535828, + "learning_rate": 4.2144579378208795e-05, + "loss": 3.1799, + "step": 10812 + }, + { + "epoch": 0.040501409297895884, + "grad_norm": 0.14686319231987, + "learning_rate": 4.2142620074606825e-05, + "loss": 3.1889, + "step": 10813 + }, + { + "epoch": 0.040532255763469305, + "grad_norm": 0.17818482220172882, + "learning_rate": 4.214066057224647e-05, + "loss": 3.1569, + "step": 10814 + }, + { + "epoch": 0.04056310222904272, + "grad_norm": 0.14086082577705383, + "learning_rate": 4.213870087115047e-05, + "loss": 3.1879, + "step": 10815 + }, + { + "epoch": 0.04059394869461613, + "grad_norm": 0.15623921155929565, + "learning_rate": 4.2136740971341526e-05, + "loss": 3.1531, + "step": 10816 + }, + { + "epoch": 0.040624795160189554, + "grad_norm": 0.15520484745502472, + "learning_rate": 4.213478087284236e-05, + "loss": 3.1961, + "step": 10817 + }, + { + "epoch": 0.04065564162576297, + "grad_norm": 0.14494948089122772, + "learning_rate": 4.2132820575675716e-05, + "loss": 3.1745, + "step": 10818 + }, + { + "epoch": 0.04068648809133638, + "grad_norm": 0.15506580471992493, + "learning_rate": 4.213086007986431e-05, + "loss": 3.1376, + "step": 10819 + }, + { + "epoch": 0.040717334556909804, + "grad_norm": 0.15077848732471466, + "learning_rate": 4.212889938543088e-05, + "loss": 3.1887, + "step": 10820 + }, + { + "epoch": 0.04074818102248322, + "grad_norm": 0.15263532102108002, + "learning_rate": 4.212693849239815e-05, + "loss": 3.1761, + "step": 10821 + }, + { + "epoch": 0.04077902748805663, + "grad_norm": 0.1524679958820343, + "learning_rate": 4.2124977400788856e-05, + "loss": 3.166, + "step": 10822 + }, + { + "epoch": 0.04080987395363005, + "grad_norm": 0.15290692448616028, + "learning_rate": 4.212301611062575e-05, + "loss": 3.169, + "step": 10823 + }, + { + "epoch": 0.04084072041920347, + "grad_norm": 0.15148085355758667, + "learning_rate": 4.2121054621931556e-05, + "loss": 3.1804, + "step": 10824 + }, + { + "epoch": 0.04087156688477688, + "grad_norm": 0.14505328238010406, + "learning_rate": 4.211909293472902e-05, + "loss": 3.1666, + "step": 10825 + }, + { + "epoch": 0.0409024133503503, + "grad_norm": 0.15015071630477905, + "learning_rate": 4.211713104904089e-05, + "loss": 3.1885, + "step": 10826 + }, + { + "epoch": 0.04093325981592372, + "grad_norm": 0.1497279405593872, + "learning_rate": 4.2115168964889917e-05, + "loss": 3.1981, + "step": 10827 + }, + { + "epoch": 0.04096410628149713, + "grad_norm": 0.15082399547100067, + "learning_rate": 4.211320668229885e-05, + "loss": 3.1511, + "step": 10828 + }, + { + "epoch": 0.04099495274707055, + "grad_norm": 0.1494419276714325, + "learning_rate": 4.211124420129042e-05, + "loss": 3.2064, + "step": 10829 + }, + { + "epoch": 0.04102579921264397, + "grad_norm": 0.14564019441604614, + "learning_rate": 4.21092815218874e-05, + "loss": 3.1797, + "step": 10830 + }, + { + "epoch": 0.04105664567821738, + "grad_norm": 0.15497401356697083, + "learning_rate": 4.210731864411255e-05, + "loss": 3.1654, + "step": 10831 + }, + { + "epoch": 0.0410874921437908, + "grad_norm": 0.15108051896095276, + "learning_rate": 4.210535556798861e-05, + "loss": 3.1749, + "step": 10832 + }, + { + "epoch": 0.041118338609364216, + "grad_norm": 0.1456734538078308, + "learning_rate": 4.210339229353836e-05, + "loss": 3.1598, + "step": 10833 + }, + { + "epoch": 0.04114918507493763, + "grad_norm": 0.14573177695274353, + "learning_rate": 4.210142882078455e-05, + "loss": 3.1901, + "step": 10834 + }, + { + "epoch": 0.04118003154051105, + "grad_norm": 0.15040963888168335, + "learning_rate": 4.209946514974995e-05, + "loss": 3.2116, + "step": 10835 + }, + { + "epoch": 0.041210878006084466, + "grad_norm": 0.16909192502498627, + "learning_rate": 4.209750128045733e-05, + "loss": 3.1898, + "step": 10836 + }, + { + "epoch": 0.04124172447165788, + "grad_norm": 0.15054234862327576, + "learning_rate": 4.209553721292946e-05, + "loss": 3.173, + "step": 10837 + }, + { + "epoch": 0.0412725709372313, + "grad_norm": 0.15001192688941956, + "learning_rate": 4.2093572947189094e-05, + "loss": 3.1464, + "step": 10838 + }, + { + "epoch": 0.041303417402804715, + "grad_norm": 0.1458168476819992, + "learning_rate": 4.209160848325904e-05, + "loss": 3.1546, + "step": 10839 + }, + { + "epoch": 0.04133426386837813, + "grad_norm": 0.14897586405277252, + "learning_rate": 4.208964382116205e-05, + "loss": 3.2241, + "step": 10840 + }, + { + "epoch": 0.04136511033395155, + "grad_norm": 0.15335458517074585, + "learning_rate": 4.2087678960920905e-05, + "loss": 3.1674, + "step": 10841 + }, + { + "epoch": 0.041395956799524965, + "grad_norm": 0.14663563668727875, + "learning_rate": 4.20857139025584e-05, + "loss": 3.1906, + "step": 10842 + }, + { + "epoch": 0.04142680326509838, + "grad_norm": 0.15038883686065674, + "learning_rate": 4.20837486460973e-05, + "loss": 3.181, + "step": 10843 + }, + { + "epoch": 0.0414576497306718, + "grad_norm": 0.15992482006549835, + "learning_rate": 4.2081783191560405e-05, + "loss": 3.2085, + "step": 10844 + }, + { + "epoch": 0.041488496196245214, + "grad_norm": 0.14262399077415466, + "learning_rate": 4.20798175389705e-05, + "loss": 3.1485, + "step": 10845 + }, + { + "epoch": 0.04151934266181863, + "grad_norm": 0.17691166698932648, + "learning_rate": 4.207785168835037e-05, + "loss": 3.1994, + "step": 10846 + }, + { + "epoch": 0.04155018912739205, + "grad_norm": 0.14854148030281067, + "learning_rate": 4.207588563972281e-05, + "loss": 3.2359, + "step": 10847 + }, + { + "epoch": 0.041581035592965464, + "grad_norm": 0.14941799640655518, + "learning_rate": 4.207391939311063e-05, + "loss": 3.1584, + "step": 10848 + }, + { + "epoch": 0.04161188205853888, + "grad_norm": 0.14825814962387085, + "learning_rate": 4.207195294853661e-05, + "loss": 3.156, + "step": 10849 + }, + { + "epoch": 0.0416427285241123, + "grad_norm": 0.15235315263271332, + "learning_rate": 4.206998630602355e-05, + "loss": 3.2039, + "step": 10850 + }, + { + "epoch": 0.04167357498968571, + "grad_norm": 0.14282000064849854, + "learning_rate": 4.2068019465594255e-05, + "loss": 3.1927, + "step": 10851 + }, + { + "epoch": 0.04170442145525913, + "grad_norm": 0.13932135701179504, + "learning_rate": 4.206605242727154e-05, + "loss": 3.1809, + "step": 10852 + }, + { + "epoch": 0.04173526792083255, + "grad_norm": 0.14851655066013336, + "learning_rate": 4.206408519107819e-05, + "loss": 3.1917, + "step": 10853 + }, + { + "epoch": 0.04176611438640596, + "grad_norm": 0.14925092458724976, + "learning_rate": 4.2062117757037036e-05, + "loss": 3.1513, + "step": 10854 + }, + { + "epoch": 0.04179696085197938, + "grad_norm": 0.14549414813518524, + "learning_rate": 4.206015012517088e-05, + "loss": 3.198, + "step": 10855 + }, + { + "epoch": 0.0418278073175528, + "grad_norm": 0.14805178344249725, + "learning_rate": 4.2058182295502535e-05, + "loss": 3.1773, + "step": 10856 + }, + { + "epoch": 0.04185865378312621, + "grad_norm": 0.14362548291683197, + "learning_rate": 4.2056214268054815e-05, + "loss": 3.1595, + "step": 10857 + }, + { + "epoch": 0.041889500248699627, + "grad_norm": 0.15890195965766907, + "learning_rate": 4.2054246042850543e-05, + "loss": 3.1542, + "step": 10858 + }, + { + "epoch": 0.04192034671427305, + "grad_norm": 0.15443824231624603, + "learning_rate": 4.205227761991253e-05, + "loss": 3.2107, + "step": 10859 + }, + { + "epoch": 0.04195119317984646, + "grad_norm": 0.1489051729440689, + "learning_rate": 4.205030899926361e-05, + "loss": 3.1975, + "step": 10860 + }, + { + "epoch": 0.041982039645419876, + "grad_norm": 0.14245210587978363, + "learning_rate": 4.2048340180926595e-05, + "loss": 3.1529, + "step": 10861 + }, + { + "epoch": 0.0420128861109933, + "grad_norm": 0.14598199725151062, + "learning_rate": 4.2046371164924333e-05, + "loss": 3.1557, + "step": 10862 + }, + { + "epoch": 0.04204373257656671, + "grad_norm": 0.14794321358203888, + "learning_rate": 4.204440195127963e-05, + "loss": 3.2003, + "step": 10863 + }, + { + "epoch": 0.042074579042140126, + "grad_norm": 0.14738339185714722, + "learning_rate": 4.2042432540015335e-05, + "loss": 3.1813, + "step": 10864 + }, + { + "epoch": 0.04210542550771355, + "grad_norm": 0.15284325182437897, + "learning_rate": 4.204046293115428e-05, + "loss": 3.2212, + "step": 10865 + }, + { + "epoch": 0.04213627197328696, + "grad_norm": 0.1528407335281372, + "learning_rate": 4.2038493124719294e-05, + "loss": 3.1665, + "step": 10866 + }, + { + "epoch": 0.042167118438860375, + "grad_norm": 0.1455480009317398, + "learning_rate": 4.2036523120733216e-05, + "loss": 3.1906, + "step": 10867 + }, + { + "epoch": 0.042197964904433796, + "grad_norm": 0.17431522905826569, + "learning_rate": 4.203455291921889e-05, + "loss": 3.1651, + "step": 10868 + }, + { + "epoch": 0.04222881137000721, + "grad_norm": 0.27550435066223145, + "learning_rate": 4.2032582520199164e-05, + "loss": 3.2148, + "step": 10869 + }, + { + "epoch": 0.042259657835580625, + "grad_norm": 0.15060117840766907, + "learning_rate": 4.203061192369687e-05, + "loss": 3.1874, + "step": 10870 + }, + { + "epoch": 0.042290504301154046, + "grad_norm": 0.15483418107032776, + "learning_rate": 4.202864112973488e-05, + "loss": 3.1717, + "step": 10871 + }, + { + "epoch": 0.04232135076672746, + "grad_norm": 0.15080678462982178, + "learning_rate": 4.2026670138336025e-05, + "loss": 3.1709, + "step": 10872 + }, + { + "epoch": 0.042352197232300874, + "grad_norm": 0.15124064683914185, + "learning_rate": 4.2024698949523153e-05, + "loss": 3.2283, + "step": 10873 + }, + { + "epoch": 0.042383043697874295, + "grad_norm": 0.14581650495529175, + "learning_rate": 4.202272756331913e-05, + "loss": 3.1668, + "step": 10874 + }, + { + "epoch": 0.04241389016344771, + "grad_norm": 0.14567580819129944, + "learning_rate": 4.2020755979746814e-05, + "loss": 3.1559, + "step": 10875 + }, + { + "epoch": 0.042444736629021124, + "grad_norm": 0.15681466460227966, + "learning_rate": 4.201878419882905e-05, + "loss": 3.1721, + "step": 10876 + }, + { + "epoch": 0.042475583094594545, + "grad_norm": 0.15868058800697327, + "learning_rate": 4.2016812220588725e-05, + "loss": 3.1848, + "step": 10877 + }, + { + "epoch": 0.04250642956016796, + "grad_norm": 0.14680856466293335, + "learning_rate": 4.201484004504869e-05, + "loss": 3.1535, + "step": 10878 + }, + { + "epoch": 0.04253727602574137, + "grad_norm": 0.14470861852169037, + "learning_rate": 4.201286767223179e-05, + "loss": 3.1667, + "step": 10879 + }, + { + "epoch": 0.042568122491314794, + "grad_norm": 0.14722053706645966, + "learning_rate": 4.201089510216093e-05, + "loss": 3.1933, + "step": 10880 + }, + { + "epoch": 0.04259896895688821, + "grad_norm": 0.1439506709575653, + "learning_rate": 4.2008922334858955e-05, + "loss": 3.1668, + "step": 10881 + }, + { + "epoch": 0.04262981542246162, + "grad_norm": 0.14933477342128754, + "learning_rate": 4.2006949370348744e-05, + "loss": 3.1891, + "step": 10882 + }, + { + "epoch": 0.042660661888035044, + "grad_norm": 0.15303826332092285, + "learning_rate": 4.200497620865319e-05, + "loss": 3.1905, + "step": 10883 + }, + { + "epoch": 0.04269150835360846, + "grad_norm": 0.14956265687942505, + "learning_rate": 4.2003002849795145e-05, + "loss": 3.2024, + "step": 10884 + }, + { + "epoch": 0.04272235481918187, + "grad_norm": 0.14659248292446136, + "learning_rate": 4.20010292937975e-05, + "loss": 3.2073, + "step": 10885 + }, + { + "epoch": 0.04275320128475529, + "grad_norm": 0.15603427588939667, + "learning_rate": 4.199905554068314e-05, + "loss": 3.1892, + "step": 10886 + }, + { + "epoch": 0.04278404775032871, + "grad_norm": 0.14094310998916626, + "learning_rate": 4.199708159047495e-05, + "loss": 3.1938, + "step": 10887 + }, + { + "epoch": 0.04281489421590212, + "grad_norm": 0.15104281902313232, + "learning_rate": 4.19951074431958e-05, + "loss": 3.1845, + "step": 10888 + }, + { + "epoch": 0.04284574068147554, + "grad_norm": 0.1488516479730606, + "learning_rate": 4.19931330988686e-05, + "loss": 3.1822, + "step": 10889 + }, + { + "epoch": 0.04287658714704896, + "grad_norm": 0.14341287314891815, + "learning_rate": 4.199115855751623e-05, + "loss": 3.1709, + "step": 10890 + }, + { + "epoch": 0.04290743361262237, + "grad_norm": 0.15636999905109406, + "learning_rate": 4.1989183819161594e-05, + "loss": 3.1894, + "step": 10891 + }, + { + "epoch": 0.04293828007819579, + "grad_norm": 0.1690208911895752, + "learning_rate": 4.1987208883827576e-05, + "loss": 3.2103, + "step": 10892 + }, + { + "epoch": 0.04296912654376921, + "grad_norm": 0.15530890226364136, + "learning_rate": 4.198523375153708e-05, + "loss": 3.1815, + "step": 10893 + }, + { + "epoch": 0.04299997300934262, + "grad_norm": 0.15234088897705078, + "learning_rate": 4.1983258422313e-05, + "loss": 3.1781, + "step": 10894 + }, + { + "epoch": 0.04303081947491604, + "grad_norm": 0.16739793121814728, + "learning_rate": 4.1981282896178254e-05, + "loss": 3.2011, + "step": 10895 + }, + { + "epoch": 0.043061665940489456, + "grad_norm": 0.15079258382320404, + "learning_rate": 4.197930717315573e-05, + "loss": 3.2054, + "step": 10896 + }, + { + "epoch": 0.04309251240606287, + "grad_norm": 0.14597755670547485, + "learning_rate": 4.197733125326835e-05, + "loss": 3.1555, + "step": 10897 + }, + { + "epoch": 0.04312335887163629, + "grad_norm": 0.14465922117233276, + "learning_rate": 4.1975355136539016e-05, + "loss": 3.2022, + "step": 10898 + }, + { + "epoch": 0.043154205337209706, + "grad_norm": 0.14673565328121185, + "learning_rate": 4.1973378822990635e-05, + "loss": 3.1703, + "step": 10899 + }, + { + "epoch": 0.04318505180278312, + "grad_norm": 0.14306512475013733, + "learning_rate": 4.1971402312646124e-05, + "loss": 3.1592, + "step": 10900 + }, + { + "epoch": 0.04321589826835654, + "grad_norm": 0.1567057967185974, + "learning_rate": 4.1969425605528404e-05, + "loss": 3.1542, + "step": 10901 + }, + { + "epoch": 0.043246744733929955, + "grad_norm": 0.15127646923065186, + "learning_rate": 4.1967448701660395e-05, + "loss": 3.1898, + "step": 10902 + }, + { + "epoch": 0.04327759119950337, + "grad_norm": 0.14481636881828308, + "learning_rate": 4.196547160106501e-05, + "loss": 3.1527, + "step": 10903 + }, + { + "epoch": 0.04330843766507679, + "grad_norm": 0.14705409109592438, + "learning_rate": 4.196349430376518e-05, + "loss": 3.1688, + "step": 10904 + }, + { + "epoch": 0.043339284130650205, + "grad_norm": 0.14955122768878937, + "learning_rate": 4.196151680978383e-05, + "loss": 3.1673, + "step": 10905 + }, + { + "epoch": 0.04337013059622362, + "grad_norm": 0.14893266558647156, + "learning_rate": 4.195953911914388e-05, + "loss": 3.1807, + "step": 10906 + }, + { + "epoch": 0.04340097706179704, + "grad_norm": 0.14085082709789276, + "learning_rate": 4.195756123186827e-05, + "loss": 3.1631, + "step": 10907 + }, + { + "epoch": 0.043431823527370454, + "grad_norm": 0.1667957901954651, + "learning_rate": 4.1955583147979924e-05, + "loss": 3.1762, + "step": 10908 + }, + { + "epoch": 0.04346266999294387, + "grad_norm": 0.146100252866745, + "learning_rate": 4.1953604867501785e-05, + "loss": 3.164, + "step": 10909 + }, + { + "epoch": 0.04349351645851729, + "grad_norm": 0.1431294083595276, + "learning_rate": 4.1951626390456775e-05, + "loss": 3.1969, + "step": 10910 + }, + { + "epoch": 0.043524362924090704, + "grad_norm": 0.15446780622005463, + "learning_rate": 4.194964771686786e-05, + "loss": 3.1658, + "step": 10911 + }, + { + "epoch": 0.04355520938966412, + "grad_norm": 0.15006300806999207, + "learning_rate": 4.194766884675796e-05, + "loss": 3.172, + "step": 10912 + }, + { + "epoch": 0.04358605585523754, + "grad_norm": 0.15177562832832336, + "learning_rate": 4.194568978015001e-05, + "loss": 3.2205, + "step": 10913 + }, + { + "epoch": 0.04361690232081095, + "grad_norm": 0.1550137847661972, + "learning_rate": 4.1943710517066984e-05, + "loss": 3.1937, + "step": 10914 + }, + { + "epoch": 0.04364774878638437, + "grad_norm": 0.14452050626277924, + "learning_rate": 4.1941731057531816e-05, + "loss": 3.2116, + "step": 10915 + }, + { + "epoch": 0.04367859525195779, + "grad_norm": 0.15439684689044952, + "learning_rate": 4.193975140156745e-05, + "loss": 3.203, + "step": 10916 + }, + { + "epoch": 0.0437094417175312, + "grad_norm": 0.15297457575798035, + "learning_rate": 4.1937771549196855e-05, + "loss": 3.1886, + "step": 10917 + }, + { + "epoch": 0.04374028818310462, + "grad_norm": 0.146813303232193, + "learning_rate": 4.1935791500442964e-05, + "loss": 3.181, + "step": 10918 + }, + { + "epoch": 0.04377113464867804, + "grad_norm": 0.15006078779697418, + "learning_rate": 4.193381125532876e-05, + "loss": 3.1797, + "step": 10919 + }, + { + "epoch": 0.04380198111425145, + "grad_norm": 0.1467643529176712, + "learning_rate": 4.193183081387719e-05, + "loss": 3.1576, + "step": 10920 + }, + { + "epoch": 0.043832827579824866, + "grad_norm": 0.14808425307273865, + "learning_rate": 4.192985017611121e-05, + "loss": 3.2129, + "step": 10921 + }, + { + "epoch": 0.04386367404539829, + "grad_norm": 0.1545167863368988, + "learning_rate": 4.19278693420538e-05, + "loss": 3.211, + "step": 10922 + }, + { + "epoch": 0.0438945205109717, + "grad_norm": 0.14809489250183105, + "learning_rate": 4.192588831172791e-05, + "loss": 3.1506, + "step": 10923 + }, + { + "epoch": 0.043925366976545116, + "grad_norm": 0.14150270819664001, + "learning_rate": 4.1923907085156524e-05, + "loss": 3.2182, + "step": 10924 + }, + { + "epoch": 0.04395621344211854, + "grad_norm": 0.1512877494096756, + "learning_rate": 4.192192566236259e-05, + "loss": 3.1556, + "step": 10925 + }, + { + "epoch": 0.04398705990769195, + "grad_norm": 0.1446048766374588, + "learning_rate": 4.191994404336911e-05, + "loss": 3.2032, + "step": 10926 + }, + { + "epoch": 0.044017906373265366, + "grad_norm": 0.14572463929653168, + "learning_rate": 4.191796222819905e-05, + "loss": 3.1895, + "step": 10927 + }, + { + "epoch": 0.04404875283883879, + "grad_norm": 0.1466648429632187, + "learning_rate": 4.1915980216875375e-05, + "loss": 3.1801, + "step": 10928 + }, + { + "epoch": 0.0440795993044122, + "grad_norm": 0.1404353380203247, + "learning_rate": 4.191399800942107e-05, + "loss": 3.1631, + "step": 10929 + }, + { + "epoch": 0.044110445769985615, + "grad_norm": 0.14676786959171295, + "learning_rate": 4.1912015605859135e-05, + "loss": 3.1853, + "step": 10930 + }, + { + "epoch": 0.044141292235559036, + "grad_norm": 0.14867964386940002, + "learning_rate": 4.191003300621253e-05, + "loss": 3.1799, + "step": 10931 + }, + { + "epoch": 0.04417213870113245, + "grad_norm": 0.15318186581134796, + "learning_rate": 4.1908050210504265e-05, + "loss": 3.1625, + "step": 10932 + }, + { + "epoch": 0.044202985166705865, + "grad_norm": 0.14701415598392487, + "learning_rate": 4.1906067218757316e-05, + "loss": 3.1834, + "step": 10933 + }, + { + "epoch": 0.044233831632279286, + "grad_norm": 0.14497576653957367, + "learning_rate": 4.190408403099467e-05, + "loss": 3.1553, + "step": 10934 + }, + { + "epoch": 0.0442646780978527, + "grad_norm": 0.14112961292266846, + "learning_rate": 4.1902100647239335e-05, + "loss": 3.2033, + "step": 10935 + }, + { + "epoch": 0.044295524563426114, + "grad_norm": 0.1438886821269989, + "learning_rate": 4.19001170675143e-05, + "loss": 3.2083, + "step": 10936 + }, + { + "epoch": 0.044326371028999535, + "grad_norm": 0.1535692811012268, + "learning_rate": 4.189813329184256e-05, + "loss": 3.1932, + "step": 10937 + }, + { + "epoch": 0.04435721749457295, + "grad_norm": 0.14359867572784424, + "learning_rate": 4.189614932024712e-05, + "loss": 3.1427, + "step": 10938 + }, + { + "epoch": 0.044388063960146364, + "grad_norm": 0.15483152866363525, + "learning_rate": 4.189416515275099e-05, + "loss": 3.2073, + "step": 10939 + }, + { + "epoch": 0.044418910425719785, + "grad_norm": 0.14773094654083252, + "learning_rate": 4.189218078937715e-05, + "loss": 3.1739, + "step": 10940 + }, + { + "epoch": 0.0444497568912932, + "grad_norm": 0.1486710160970688, + "learning_rate": 4.189019623014864e-05, + "loss": 3.1605, + "step": 10941 + }, + { + "epoch": 0.04448060335686661, + "grad_norm": 0.14695093035697937, + "learning_rate": 4.188821147508845e-05, + "loss": 3.1752, + "step": 10942 + }, + { + "epoch": 0.044511449822440034, + "grad_norm": 0.1486148089170456, + "learning_rate": 4.1886226524219594e-05, + "loss": 3.1918, + "step": 10943 + }, + { + "epoch": 0.04454229628801345, + "grad_norm": 0.14910022914409637, + "learning_rate": 4.1884241377565085e-05, + "loss": 3.18, + "step": 10944 + }, + { + "epoch": 0.04457314275358686, + "grad_norm": 0.1548784077167511, + "learning_rate": 4.188225603514795e-05, + "loss": 3.1685, + "step": 10945 + }, + { + "epoch": 0.044603989219160284, + "grad_norm": 0.1493455320596695, + "learning_rate": 4.1880270496991206e-05, + "loss": 3.1839, + "step": 10946 + }, + { + "epoch": 0.0446348356847337, + "grad_norm": 0.14990487694740295, + "learning_rate": 4.187828476311786e-05, + "loss": 3.1718, + "step": 10947 + }, + { + "epoch": 0.04466568215030711, + "grad_norm": 0.14557482302188873, + "learning_rate": 4.1876298833550946e-05, + "loss": 3.1672, + "step": 10948 + }, + { + "epoch": 0.04469652861588053, + "grad_norm": 0.1446310430765152, + "learning_rate": 4.187431270831349e-05, + "loss": 3.1905, + "step": 10949 + }, + { + "epoch": 0.04472737508145395, + "grad_norm": 0.14499549567699432, + "learning_rate": 4.187232638742852e-05, + "loss": 3.209, + "step": 10950 + }, + { + "epoch": 0.04475822154702736, + "grad_norm": 0.14356876909732819, + "learning_rate": 4.1870339870919064e-05, + "loss": 3.2016, + "step": 10951 + }, + { + "epoch": 0.04478906801260078, + "grad_norm": 0.15120674669742584, + "learning_rate": 4.186835315880816e-05, + "loss": 3.2112, + "step": 10952 + }, + { + "epoch": 0.0448199144781742, + "grad_norm": 0.1454544961452484, + "learning_rate": 4.1866366251118824e-05, + "loss": 3.1681, + "step": 10953 + }, + { + "epoch": 0.04485076094374761, + "grad_norm": 0.1455458551645279, + "learning_rate": 4.1864379147874125e-05, + "loss": 3.1443, + "step": 10954 + }, + { + "epoch": 0.04488160740932103, + "grad_norm": 0.15202559530735016, + "learning_rate": 4.186239184909707e-05, + "loss": 3.1584, + "step": 10955 + }, + { + "epoch": 0.04491245387489445, + "grad_norm": 0.14520753920078278, + "learning_rate": 4.1860404354810725e-05, + "loss": 3.1587, + "step": 10956 + }, + { + "epoch": 0.04494330034046786, + "grad_norm": 0.14434534311294556, + "learning_rate": 4.1858416665038116e-05, + "loss": 3.1821, + "step": 10957 + }, + { + "epoch": 0.04497414680604128, + "grad_norm": 0.1471737027168274, + "learning_rate": 4.1856428779802293e-05, + "loss": 3.1591, + "step": 10958 + }, + { + "epoch": 0.045004993271614696, + "grad_norm": 0.1462526172399521, + "learning_rate": 4.185444069912632e-05, + "loss": 3.1895, + "step": 10959 + }, + { + "epoch": 0.04503583973718811, + "grad_norm": 0.14664073288440704, + "learning_rate": 4.1852452423033226e-05, + "loss": 3.1782, + "step": 10960 + }, + { + "epoch": 0.04506668620276153, + "grad_norm": 0.14206886291503906, + "learning_rate": 4.185046395154607e-05, + "loss": 3.1863, + "step": 10961 + }, + { + "epoch": 0.045097532668334946, + "grad_norm": 0.15042991936206818, + "learning_rate": 4.184847528468792e-05, + "loss": 3.1677, + "step": 10962 + }, + { + "epoch": 0.04512837913390836, + "grad_norm": 0.14676077663898468, + "learning_rate": 4.184648642248182e-05, + "loss": 3.1785, + "step": 10963 + }, + { + "epoch": 0.04515922559948178, + "grad_norm": 0.14718681573867798, + "learning_rate": 4.184449736495083e-05, + "loss": 3.1757, + "step": 10964 + }, + { + "epoch": 0.045190072065055195, + "grad_norm": 0.1474734991788864, + "learning_rate": 4.184250811211802e-05, + "loss": 3.1715, + "step": 10965 + }, + { + "epoch": 0.04522091853062861, + "grad_norm": 0.15071764588356018, + "learning_rate": 4.184051866400645e-05, + "loss": 3.2182, + "step": 10966 + }, + { + "epoch": 0.04525176499620203, + "grad_norm": 0.16880473494529724, + "learning_rate": 4.1838529020639186e-05, + "loss": 3.1834, + "step": 10967 + }, + { + "epoch": 0.045282611461775445, + "grad_norm": 0.1630345731973648, + "learning_rate": 4.183653918203929e-05, + "loss": 3.2063, + "step": 10968 + }, + { + "epoch": 0.04531345792734886, + "grad_norm": 0.1551659107208252, + "learning_rate": 4.183454914822985e-05, + "loss": 3.1602, + "step": 10969 + }, + { + "epoch": 0.04534430439292228, + "grad_norm": 0.178705632686615, + "learning_rate": 4.183255891923392e-05, + "loss": 3.1589, + "step": 10970 + }, + { + "epoch": 0.045375150858495694, + "grad_norm": 0.1527281403541565, + "learning_rate": 4.1830568495074594e-05, + "loss": 3.1642, + "step": 10971 + }, + { + "epoch": 0.04540599732406911, + "grad_norm": 0.1489914208650589, + "learning_rate": 4.1828577875774935e-05, + "loss": 3.1661, + "step": 10972 + }, + { + "epoch": 0.04543684378964253, + "grad_norm": 0.15323033928871155, + "learning_rate": 4.1826587061358025e-05, + "loss": 3.1832, + "step": 10973 + }, + { + "epoch": 0.045467690255215944, + "grad_norm": 0.17978909611701965, + "learning_rate": 4.1824596051846956e-05, + "loss": 3.1801, + "step": 10974 + }, + { + "epoch": 0.04549853672078936, + "grad_norm": 0.14642105996608734, + "learning_rate": 4.18226048472648e-05, + "loss": 3.2033, + "step": 10975 + }, + { + "epoch": 0.04552938318636278, + "grad_norm": 0.14922000467777252, + "learning_rate": 4.182061344763466e-05, + "loss": 3.2055, + "step": 10976 + }, + { + "epoch": 0.04556022965193619, + "grad_norm": 0.1487666368484497, + "learning_rate": 4.1818621852979604e-05, + "loss": 3.1563, + "step": 10977 + }, + { + "epoch": 0.04559107611750961, + "grad_norm": 0.1453012079000473, + "learning_rate": 4.1816630063322744e-05, + "loss": 3.1822, + "step": 10978 + }, + { + "epoch": 0.04562192258308303, + "grad_norm": 0.147248774766922, + "learning_rate": 4.1814638078687153e-05, + "loss": 3.1614, + "step": 10979 + }, + { + "epoch": 0.04565276904865644, + "grad_norm": 0.14978042244911194, + "learning_rate": 4.1812645899095946e-05, + "loss": 3.1892, + "step": 10980 + }, + { + "epoch": 0.04568361551422986, + "grad_norm": 0.1579016149044037, + "learning_rate": 4.1810653524572215e-05, + "loss": 3.1519, + "step": 10981 + }, + { + "epoch": 0.04571446197980328, + "grad_norm": 0.14403152465820312, + "learning_rate": 4.1808660955139046e-05, + "loss": 3.2028, + "step": 10982 + }, + { + "epoch": 0.04574530844537669, + "grad_norm": 0.15448269248008728, + "learning_rate": 4.180666819081956e-05, + "loss": 3.1805, + "step": 10983 + }, + { + "epoch": 0.045776154910950106, + "grad_norm": 0.14878734946250916, + "learning_rate": 4.180467523163686e-05, + "loss": 3.1743, + "step": 10984 + }, + { + "epoch": 0.04580700137652353, + "grad_norm": 0.14528393745422363, + "learning_rate": 4.180268207761405e-05, + "loss": 3.1888, + "step": 10985 + }, + { + "epoch": 0.04583784784209694, + "grad_norm": 0.1588563621044159, + "learning_rate": 4.180068872877424e-05, + "loss": 3.1923, + "step": 10986 + }, + { + "epoch": 0.045868694307670356, + "grad_norm": 0.1453743427991867, + "learning_rate": 4.1798695185140535e-05, + "loss": 3.1802, + "step": 10987 + }, + { + "epoch": 0.04589954077324378, + "grad_norm": 0.14930735528469086, + "learning_rate": 4.1796701446736054e-05, + "loss": 3.1985, + "step": 10988 + }, + { + "epoch": 0.04593038723881719, + "grad_norm": 0.14785653352737427, + "learning_rate": 4.1794707513583913e-05, + "loss": 3.1635, + "step": 10989 + }, + { + "epoch": 0.045961233704390606, + "grad_norm": 0.14702239632606506, + "learning_rate": 4.179271338570723e-05, + "loss": 3.1803, + "step": 10990 + }, + { + "epoch": 0.04599208016996403, + "grad_norm": 0.1475820541381836, + "learning_rate": 4.179071906312913e-05, + "loss": 3.1735, + "step": 10991 + }, + { + "epoch": 0.04602292663553744, + "grad_norm": 0.14528925716876984, + "learning_rate": 4.1788724545872735e-05, + "loss": 3.1578, + "step": 10992 + }, + { + "epoch": 0.046053773101110855, + "grad_norm": 0.17976011335849762, + "learning_rate": 4.1786729833961166e-05, + "loss": 3.1967, + "step": 10993 + }, + { + "epoch": 0.046084619566684276, + "grad_norm": 0.14552642405033112, + "learning_rate": 4.1784734927417546e-05, + "loss": 3.2021, + "step": 10994 + }, + { + "epoch": 0.04611546603225769, + "grad_norm": 0.14077326655387878, + "learning_rate": 4.1782739826265016e-05, + "loss": 3.1684, + "step": 10995 + }, + { + "epoch": 0.046146312497831105, + "grad_norm": 0.1462620198726654, + "learning_rate": 4.1780744530526705e-05, + "loss": 3.171, + "step": 10996 + }, + { + "epoch": 0.046177158963404526, + "grad_norm": 0.15760529041290283, + "learning_rate": 4.177874904022574e-05, + "loss": 3.1962, + "step": 10997 + }, + { + "epoch": 0.04620800542897794, + "grad_norm": 0.14475151896476746, + "learning_rate": 4.1776753355385264e-05, + "loss": 3.1768, + "step": 10998 + }, + { + "epoch": 0.046238851894551354, + "grad_norm": 0.1452018916606903, + "learning_rate": 4.177475747602842e-05, + "loss": 3.193, + "step": 10999 + }, + { + "epoch": 0.046269698360124775, + "grad_norm": 0.16047653555870056, + "learning_rate": 4.177276140217834e-05, + "loss": 3.1917, + "step": 11000 + }, + { + "epoch": 0.04630054482569819, + "grad_norm": 0.14751550555229187, + "learning_rate": 4.177076513385817e-05, + "loss": 3.1411, + "step": 11001 + }, + { + "epoch": 0.046331391291271604, + "grad_norm": 0.14540286362171173, + "learning_rate": 4.176876867109105e-05, + "loss": 3.1877, + "step": 11002 + }, + { + "epoch": 0.046362237756845025, + "grad_norm": 0.1751050353050232, + "learning_rate": 4.176677201390015e-05, + "loss": 3.1411, + "step": 11003 + }, + { + "epoch": 0.04639308422241844, + "grad_norm": 0.1564645618200302, + "learning_rate": 4.176477516230859e-05, + "loss": 3.1463, + "step": 11004 + }, + { + "epoch": 0.04642393068799185, + "grad_norm": 0.16018690168857574, + "learning_rate": 4.176277811633954e-05, + "loss": 3.1522, + "step": 11005 + }, + { + "epoch": 0.046454777153565274, + "grad_norm": 0.15257340669631958, + "learning_rate": 4.176078087601615e-05, + "loss": 3.1618, + "step": 11006 + }, + { + "epoch": 0.04648562361913869, + "grad_norm": 0.15093854069709778, + "learning_rate": 4.1758783441361585e-05, + "loss": 3.1846, + "step": 11007 + }, + { + "epoch": 0.0465164700847121, + "grad_norm": 0.1507740020751953, + "learning_rate": 4.175678581239899e-05, + "loss": 3.1532, + "step": 11008 + }, + { + "epoch": 0.046547316550285524, + "grad_norm": 0.1583373099565506, + "learning_rate": 4.175478798915154e-05, + "loss": 3.1808, + "step": 11009 + }, + { + "epoch": 0.04657816301585894, + "grad_norm": 0.15408477187156677, + "learning_rate": 4.1752789971642384e-05, + "loss": 3.1585, + "step": 11010 + }, + { + "epoch": 0.04660900948143235, + "grad_norm": 0.14693617820739746, + "learning_rate": 4.175079175989469e-05, + "loss": 3.178, + "step": 11011 + }, + { + "epoch": 0.04663985594700577, + "grad_norm": 0.14954660832881927, + "learning_rate": 4.174879335393165e-05, + "loss": 3.1566, + "step": 11012 + }, + { + "epoch": 0.04667070241257919, + "grad_norm": 0.14909185469150543, + "learning_rate": 4.1746794753776404e-05, + "loss": 3.1933, + "step": 11013 + }, + { + "epoch": 0.04670154887815261, + "grad_norm": 0.16575855016708374, + "learning_rate": 4.174479595945213e-05, + "loss": 3.1801, + "step": 11014 + }, + { + "epoch": 0.04673239534372602, + "grad_norm": 0.14933359622955322, + "learning_rate": 4.174279697098202e-05, + "loss": 3.1655, + "step": 11015 + }, + { + "epoch": 0.04676324180929944, + "grad_norm": 0.14802242815494537, + "learning_rate": 4.174079778838924e-05, + "loss": 3.1812, + "step": 11016 + }, + { + "epoch": 0.04679408827487286, + "grad_norm": 0.1651923507452011, + "learning_rate": 4.1738798411696964e-05, + "loss": 3.1747, + "step": 11017 + }, + { + "epoch": 0.04682493474044627, + "grad_norm": 0.14930318295955658, + "learning_rate": 4.1736798840928385e-05, + "loss": 3.1684, + "step": 11018 + }, + { + "epoch": 0.04685578120601969, + "grad_norm": 0.1485980749130249, + "learning_rate": 4.1734799076106686e-05, + "loss": 3.1813, + "step": 11019 + }, + { + "epoch": 0.04688662767159311, + "grad_norm": 0.1460142284631729, + "learning_rate": 4.173279911725504e-05, + "loss": 3.178, + "step": 11020 + }, + { + "epoch": 0.04691747413716652, + "grad_norm": 0.1495753675699234, + "learning_rate": 4.1730798964396645e-05, + "loss": 3.1812, + "step": 11021 + }, + { + "epoch": 0.046948320602739936, + "grad_norm": 0.14689038693904877, + "learning_rate": 4.172879861755469e-05, + "loss": 3.2007, + "step": 11022 + }, + { + "epoch": 0.04697916706831336, + "grad_norm": 0.149704709649086, + "learning_rate": 4.172679807675237e-05, + "loss": 3.195, + "step": 11023 + }, + { + "epoch": 0.04701001353388677, + "grad_norm": 0.15123291313648224, + "learning_rate": 4.1724797342012865e-05, + "loss": 3.1698, + "step": 11024 + }, + { + "epoch": 0.047040859999460186, + "grad_norm": 0.15782934427261353, + "learning_rate": 4.17227964133594e-05, + "loss": 3.192, + "step": 11025 + }, + { + "epoch": 0.04707170646503361, + "grad_norm": 0.1485937535762787, + "learning_rate": 4.172079529081516e-05, + "loss": 3.1883, + "step": 11026 + }, + { + "epoch": 0.04710255293060702, + "grad_norm": 0.1537921130657196, + "learning_rate": 4.1718793974403337e-05, + "loss": 3.1982, + "step": 11027 + }, + { + "epoch": 0.047133399396180435, + "grad_norm": 0.14525316655635834, + "learning_rate": 4.171679246414715e-05, + "loss": 3.1517, + "step": 11028 + }, + { + "epoch": 0.047164245861753856, + "grad_norm": 0.14803291857242584, + "learning_rate": 4.17147907600698e-05, + "loss": 3.1825, + "step": 11029 + }, + { + "epoch": 0.04719509232732727, + "grad_norm": 0.1469743549823761, + "learning_rate": 4.17127888621945e-05, + "loss": 3.1626, + "step": 11030 + }, + { + "epoch": 0.047225938792900685, + "grad_norm": 0.16364595293998718, + "learning_rate": 4.171078677054445e-05, + "loss": 3.1849, + "step": 11031 + }, + { + "epoch": 0.047256785258474106, + "grad_norm": 0.14633339643478394, + "learning_rate": 4.170878448514287e-05, + "loss": 3.162, + "step": 11032 + }, + { + "epoch": 0.04728763172404752, + "grad_norm": 0.14484412968158722, + "learning_rate": 4.170678200601298e-05, + "loss": 3.146, + "step": 11033 + }, + { + "epoch": 0.047318478189620934, + "grad_norm": 0.14491881430149078, + "learning_rate": 4.170477933317799e-05, + "loss": 3.1807, + "step": 11034 + }, + { + "epoch": 0.047349324655194355, + "grad_norm": 0.14686693251132965, + "learning_rate": 4.170277646666112e-05, + "loss": 3.2053, + "step": 11035 + }, + { + "epoch": 0.04738017112076777, + "grad_norm": 0.14913873374462128, + "learning_rate": 4.1700773406485594e-05, + "loss": 3.2073, + "step": 11036 + }, + { + "epoch": 0.047411017586341184, + "grad_norm": 0.14426930248737335, + "learning_rate": 4.169877015267463e-05, + "loss": 3.1731, + "step": 11037 + }, + { + "epoch": 0.047441864051914605, + "grad_norm": 0.15392200648784637, + "learning_rate": 4.169676670525147e-05, + "loss": 3.1675, + "step": 11038 + }, + { + "epoch": 0.04747271051748802, + "grad_norm": 0.1533963680267334, + "learning_rate": 4.169476306423933e-05, + "loss": 3.1378, + "step": 11039 + }, + { + "epoch": 0.04750355698306143, + "grad_norm": 0.14561615884304047, + "learning_rate": 4.1692759229661447e-05, + "loss": 3.1791, + "step": 11040 + }, + { + "epoch": 0.047534403448634854, + "grad_norm": 0.15414665639400482, + "learning_rate": 4.1690755201541055e-05, + "loss": 3.1957, + "step": 11041 + }, + { + "epoch": 0.04756524991420827, + "grad_norm": 0.16475550830364227, + "learning_rate": 4.168875097990138e-05, + "loss": 3.2093, + "step": 11042 + }, + { + "epoch": 0.04759609637978168, + "grad_norm": 0.1534699648618698, + "learning_rate": 4.168674656476567e-05, + "loss": 3.1729, + "step": 11043 + }, + { + "epoch": 0.047626942845355104, + "grad_norm": 0.15373486280441284, + "learning_rate": 4.168474195615716e-05, + "loss": 3.1707, + "step": 11044 + }, + { + "epoch": 0.04765778931092852, + "grad_norm": 0.15135246515274048, + "learning_rate": 4.168273715409909e-05, + "loss": 3.1568, + "step": 11045 + }, + { + "epoch": 0.04768863577650193, + "grad_norm": 0.15691520273685455, + "learning_rate": 4.168073215861471e-05, + "loss": 3.2084, + "step": 11046 + }, + { + "epoch": 0.04771948224207535, + "grad_norm": 0.162602961063385, + "learning_rate": 4.1678726969727265e-05, + "loss": 3.2152, + "step": 11047 + }, + { + "epoch": 0.04775032870764877, + "grad_norm": 0.1635720431804657, + "learning_rate": 4.167672158746e-05, + "loss": 3.1362, + "step": 11048 + }, + { + "epoch": 0.04778117517322218, + "grad_norm": 0.17514978349208832, + "learning_rate": 4.167471601183618e-05, + "loss": 3.1868, + "step": 11049 + }, + { + "epoch": 0.0478120216387956, + "grad_norm": 0.14783693850040436, + "learning_rate": 4.167271024287904e-05, + "loss": 3.1768, + "step": 11050 + }, + { + "epoch": 0.04784286810436902, + "grad_norm": 0.1787538379430771, + "learning_rate": 4.167070428061184e-05, + "loss": 3.1301, + "step": 11051 + }, + { + "epoch": 0.04787371456994243, + "grad_norm": 0.14904801547527313, + "learning_rate": 4.166869812505785e-05, + "loss": 3.1832, + "step": 11052 + }, + { + "epoch": 0.04790456103551585, + "grad_norm": 0.14252322912216187, + "learning_rate": 4.166669177624032e-05, + "loss": 3.1995, + "step": 11053 + }, + { + "epoch": 0.04793540750108927, + "grad_norm": 0.15728293359279633, + "learning_rate": 4.166468523418251e-05, + "loss": 3.1973, + "step": 11054 + }, + { + "epoch": 0.04796625396666268, + "grad_norm": 0.14264515042304993, + "learning_rate": 4.16626784989077e-05, + "loss": 3.1551, + "step": 11055 + }, + { + "epoch": 0.0479971004322361, + "grad_norm": 0.14822915196418762, + "learning_rate": 4.1660671570439136e-05, + "loss": 3.1602, + "step": 11056 + }, + { + "epoch": 0.048027946897809516, + "grad_norm": 0.1510394960641861, + "learning_rate": 4.16586644488001e-05, + "loss": 3.1595, + "step": 11057 + }, + { + "epoch": 0.04805879336338293, + "grad_norm": 0.14952382445335388, + "learning_rate": 4.1656657134013866e-05, + "loss": 3.2092, + "step": 11058 + }, + { + "epoch": 0.04808963982895635, + "grad_norm": 0.1671806126832962, + "learning_rate": 4.1654649626103695e-05, + "loss": 3.2069, + "step": 11059 + }, + { + "epoch": 0.048120486294529766, + "grad_norm": 0.14386360347270966, + "learning_rate": 4.165264192509287e-05, + "loss": 3.1714, + "step": 11060 + }, + { + "epoch": 0.04815133276010318, + "grad_norm": 0.15409421920776367, + "learning_rate": 4.165063403100466e-05, + "loss": 3.1497, + "step": 11061 + }, + { + "epoch": 0.0481821792256766, + "grad_norm": 0.15921810269355774, + "learning_rate": 4.1648625943862374e-05, + "loss": 3.2023, + "step": 11062 + }, + { + "epoch": 0.048213025691250015, + "grad_norm": 0.15243908762931824, + "learning_rate": 4.1646617663689266e-05, + "loss": 3.1809, + "step": 11063 + }, + { + "epoch": 0.04824387215682343, + "grad_norm": 0.17065311968326569, + "learning_rate": 4.164460919050862e-05, + "loss": 3.2044, + "step": 11064 + }, + { + "epoch": 0.04827471862239685, + "grad_norm": 0.14720241725444794, + "learning_rate": 4.1642600524343745e-05, + "loss": 3.2007, + "step": 11065 + }, + { + "epoch": 0.048305565087970265, + "grad_norm": 0.14836354553699493, + "learning_rate": 4.164059166521792e-05, + "loss": 3.2035, + "step": 11066 + }, + { + "epoch": 0.04833641155354368, + "grad_norm": 0.1606786698102951, + "learning_rate": 4.163858261315443e-05, + "loss": 3.1669, + "step": 11067 + }, + { + "epoch": 0.0483672580191171, + "grad_norm": 0.1516532599925995, + "learning_rate": 4.1636573368176565e-05, + "loss": 3.1822, + "step": 11068 + }, + { + "epoch": 0.048398104484690514, + "grad_norm": 0.15840673446655273, + "learning_rate": 4.163456393030764e-05, + "loss": 3.1707, + "step": 11069 + }, + { + "epoch": 0.04842895095026393, + "grad_norm": 0.15387937426567078, + "learning_rate": 4.1632554299570935e-05, + "loss": 3.1862, + "step": 11070 + }, + { + "epoch": 0.04845979741583735, + "grad_norm": 0.14943771064281464, + "learning_rate": 4.1630544475989766e-05, + "loss": 3.1909, + "step": 11071 + }, + { + "epoch": 0.048490643881410764, + "grad_norm": 0.14873749017715454, + "learning_rate": 4.162853445958742e-05, + "loss": 3.1919, + "step": 11072 + }, + { + "epoch": 0.04852149034698418, + "grad_norm": 0.14645026624202728, + "learning_rate": 4.162652425038721e-05, + "loss": 3.1592, + "step": 11073 + }, + { + "epoch": 0.0485523368125576, + "grad_norm": 0.14365018904209137, + "learning_rate": 4.162451384841245e-05, + "loss": 3.1746, + "step": 11074 + }, + { + "epoch": 0.04858318327813101, + "grad_norm": 0.1477659046649933, + "learning_rate": 4.162250325368644e-05, + "loss": 3.1568, + "step": 11075 + }, + { + "epoch": 0.04861402974370443, + "grad_norm": 0.15217985212802887, + "learning_rate": 4.1620492466232485e-05, + "loss": 3.1591, + "step": 11076 + }, + { + "epoch": 0.04864487620927785, + "grad_norm": 0.1501341462135315, + "learning_rate": 4.161848148607392e-05, + "loss": 3.1997, + "step": 11077 + }, + { + "epoch": 0.04867572267485126, + "grad_norm": 0.1507384330034256, + "learning_rate": 4.1616470313234046e-05, + "loss": 3.1541, + "step": 11078 + }, + { + "epoch": 0.04870656914042468, + "grad_norm": 0.14918005466461182, + "learning_rate": 4.161445894773618e-05, + "loss": 3.1627, + "step": 11079 + }, + { + "epoch": 0.0487374156059981, + "grad_norm": 0.1438746154308319, + "learning_rate": 4.161244738960364e-05, + "loss": 3.1854, + "step": 11080 + }, + { + "epoch": 0.04876826207157151, + "grad_norm": 0.14902809262275696, + "learning_rate": 4.161043563885977e-05, + "loss": 3.1711, + "step": 11081 + }, + { + "epoch": 0.048799108537144927, + "grad_norm": 0.1469445377588272, + "learning_rate": 4.1608423695527866e-05, + "loss": 3.1931, + "step": 11082 + }, + { + "epoch": 0.04882995500271835, + "grad_norm": 0.150088369846344, + "learning_rate": 4.160641155963128e-05, + "loss": 3.1695, + "step": 11083 + }, + { + "epoch": 0.04886080146829176, + "grad_norm": 0.14624787867069244, + "learning_rate": 4.160439923119333e-05, + "loss": 3.1975, + "step": 11084 + }, + { + "epoch": 0.048891647933865176, + "grad_norm": 0.15156470239162445, + "learning_rate": 4.160238671023734e-05, + "loss": 3.1716, + "step": 11085 + }, + { + "epoch": 0.0489224943994386, + "grad_norm": 0.17135557532310486, + "learning_rate": 4.1600373996786667e-05, + "loss": 3.2156, + "step": 11086 + }, + { + "epoch": 0.04895334086501201, + "grad_norm": 0.1459627002477646, + "learning_rate": 4.1598361090864624e-05, + "loss": 3.1478, + "step": 11087 + }, + { + "epoch": 0.048984187330585426, + "grad_norm": 0.14889618754386902, + "learning_rate": 4.159634799249456e-05, + "loss": 3.1759, + "step": 11088 + }, + { + "epoch": 0.04901503379615885, + "grad_norm": 0.1480228304862976, + "learning_rate": 4.159433470169982e-05, + "loss": 3.1441, + "step": 11089 + }, + { + "epoch": 0.04904588026173226, + "grad_norm": 0.1498022824525833, + "learning_rate": 4.159232121850374e-05, + "loss": 3.1758, + "step": 11090 + }, + { + "epoch": 0.049076726727305675, + "grad_norm": 0.15037544071674347, + "learning_rate": 4.1590307542929654e-05, + "loss": 3.1986, + "step": 11091 + }, + { + "epoch": 0.049107573192879096, + "grad_norm": 0.15250714123249054, + "learning_rate": 4.158829367500093e-05, + "loss": 3.1857, + "step": 11092 + }, + { + "epoch": 0.04913841965845251, + "grad_norm": 0.14984306693077087, + "learning_rate": 4.158627961474091e-05, + "loss": 3.1857, + "step": 11093 + }, + { + "epoch": 0.049169266124025925, + "grad_norm": 0.14562509953975677, + "learning_rate": 4.158426536217295e-05, + "loss": 3.1755, + "step": 11094 + }, + { + "epoch": 0.049200112589599346, + "grad_norm": 0.15335732698440552, + "learning_rate": 4.158225091732039e-05, + "loss": 3.2099, + "step": 11095 + }, + { + "epoch": 0.04923095905517276, + "grad_norm": 0.14995840191841125, + "learning_rate": 4.15802362802066e-05, + "loss": 3.1916, + "step": 11096 + }, + { + "epoch": 0.049261805520746174, + "grad_norm": 0.14697274565696716, + "learning_rate": 4.1578221450854935e-05, + "loss": 3.1456, + "step": 11097 + }, + { + "epoch": 0.049292651986319595, + "grad_norm": 0.14896361529827118, + "learning_rate": 4.157620642928875e-05, + "loss": 3.2049, + "step": 11098 + }, + { + "epoch": 0.04932349845189301, + "grad_norm": 0.14620444178581238, + "learning_rate": 4.1574191215531424e-05, + "loss": 3.1698, + "step": 11099 + }, + { + "epoch": 0.049354344917466424, + "grad_norm": 0.14619730412960052, + "learning_rate": 4.1572175809606295e-05, + "loss": 3.1836, + "step": 11100 + }, + { + "epoch": 0.049385191383039845, + "grad_norm": 0.14443038403987885, + "learning_rate": 4.157016021153675e-05, + "loss": 3.1838, + "step": 11101 + }, + { + "epoch": 0.04941603784861326, + "grad_norm": 0.14393161237239838, + "learning_rate": 4.1568144421346165e-05, + "loss": 3.1859, + "step": 11102 + }, + { + "epoch": 0.04944688431418667, + "grad_norm": 0.14598298072814941, + "learning_rate": 4.156612843905789e-05, + "loss": 3.1703, + "step": 11103 + }, + { + "epoch": 0.049477730779760094, + "grad_norm": 0.14640861749649048, + "learning_rate": 4.156411226469532e-05, + "loss": 3.1591, + "step": 11104 + }, + { + "epoch": 0.04950857724533351, + "grad_norm": 0.1466016322374344, + "learning_rate": 4.1562095898281815e-05, + "loss": 3.2004, + "step": 11105 + }, + { + "epoch": 0.04953942371090692, + "grad_norm": 0.15036749839782715, + "learning_rate": 4.156007933984076e-05, + "loss": 3.1632, + "step": 11106 + }, + { + "epoch": 0.049570270176480344, + "grad_norm": 0.14706705510616302, + "learning_rate": 4.1558062589395544e-05, + "loss": 3.1681, + "step": 11107 + }, + { + "epoch": 0.04960111664205376, + "grad_norm": 0.1504717469215393, + "learning_rate": 4.155604564696954e-05, + "loss": 3.1779, + "step": 11108 + }, + { + "epoch": 0.04963196310762717, + "grad_norm": 0.14369259774684906, + "learning_rate": 4.155402851258613e-05, + "loss": 3.1744, + "step": 11109 + }, + { + "epoch": 0.04966280957320059, + "grad_norm": 0.1535508781671524, + "learning_rate": 4.155201118626871e-05, + "loss": 3.1638, + "step": 11110 + }, + { + "epoch": 0.04969365603877401, + "grad_norm": 0.1558820605278015, + "learning_rate": 4.154999366804067e-05, + "loss": 3.1563, + "step": 11111 + }, + { + "epoch": 0.04972450250434742, + "grad_norm": 0.14411135017871857, + "learning_rate": 4.15479759579254e-05, + "loss": 3.1996, + "step": 11112 + }, + { + "epoch": 0.04975534896992084, + "grad_norm": 0.15327408909797668, + "learning_rate": 4.1545958055946286e-05, + "loss": 3.2266, + "step": 11113 + }, + { + "epoch": 0.04978619543549426, + "grad_norm": 0.15147529542446136, + "learning_rate": 4.154393996212674e-05, + "loss": 3.1429, + "step": 11114 + }, + { + "epoch": 0.04981704190106767, + "grad_norm": 0.14991815388202667, + "learning_rate": 4.154192167649015e-05, + "loss": 3.1958, + "step": 11115 + }, + { + "epoch": 0.04984788836664109, + "grad_norm": 0.14477211236953735, + "learning_rate": 4.153990319905991e-05, + "loss": 3.157, + "step": 11116 + }, + { + "epoch": 0.04987873483221451, + "grad_norm": 0.15295127034187317, + "learning_rate": 4.153788452985944e-05, + "loss": 3.1661, + "step": 11117 + }, + { + "epoch": 0.04990958129778792, + "grad_norm": 0.15152452886104584, + "learning_rate": 4.153586566891213e-05, + "loss": 3.169, + "step": 11118 + }, + { + "epoch": 0.04994042776336134, + "grad_norm": 0.14705154299736023, + "learning_rate": 4.1533846616241404e-05, + "loss": 3.1643, + "step": 11119 + }, + { + "epoch": 0.049971274228934756, + "grad_norm": 0.15718650817871094, + "learning_rate": 4.153182737187066e-05, + "loss": 3.1908, + "step": 11120 + }, + { + "epoch": 0.05000212069450817, + "grad_norm": 0.14991243183612823, + "learning_rate": 4.152980793582331e-05, + "loss": 3.1634, + "step": 11121 + }, + { + "epoch": 0.05003296716008159, + "grad_norm": 0.15753066539764404, + "learning_rate": 4.152778830812277e-05, + "loss": 3.1518, + "step": 11122 + }, + { + "epoch": 0.050063813625655006, + "grad_norm": 0.15059266984462738, + "learning_rate": 4.152576848879245e-05, + "loss": 3.1604, + "step": 11123 + }, + { + "epoch": 0.05009466009122842, + "grad_norm": 0.15389469265937805, + "learning_rate": 4.152374847785579e-05, + "loss": 3.1677, + "step": 11124 + }, + { + "epoch": 0.05012550655680184, + "grad_norm": 0.14940617978572845, + "learning_rate": 4.152172827533619e-05, + "loss": 3.19, + "step": 11125 + }, + { + "epoch": 0.050156353022375255, + "grad_norm": 0.16701021790504456, + "learning_rate": 4.151970788125708e-05, + "loss": 3.169, + "step": 11126 + }, + { + "epoch": 0.05018719948794867, + "grad_norm": 0.1608363687992096, + "learning_rate": 4.151768729564187e-05, + "loss": 3.2142, + "step": 11127 + }, + { + "epoch": 0.05021804595352209, + "grad_norm": 0.14643506705760956, + "learning_rate": 4.151566651851402e-05, + "loss": 3.1801, + "step": 11128 + }, + { + "epoch": 0.050248892419095505, + "grad_norm": 0.1522902548313141, + "learning_rate": 4.151364554989693e-05, + "loss": 3.1953, + "step": 11129 + }, + { + "epoch": 0.05027973888466892, + "grad_norm": 0.14270909130573273, + "learning_rate": 4.1511624389814045e-05, + "loss": 3.1688, + "step": 11130 + }, + { + "epoch": 0.05031058535024234, + "grad_norm": 0.1512869894504547, + "learning_rate": 4.1509603038288804e-05, + "loss": 3.1716, + "step": 11131 + }, + { + "epoch": 0.050341431815815754, + "grad_norm": 0.14775709807872772, + "learning_rate": 4.150758149534463e-05, + "loss": 3.1871, + "step": 11132 + }, + { + "epoch": 0.05037227828138917, + "grad_norm": 0.14431816339492798, + "learning_rate": 4.150555976100497e-05, + "loss": 3.2092, + "step": 11133 + }, + { + "epoch": 0.05040312474696259, + "grad_norm": 0.1473536193370819, + "learning_rate": 4.150353783529327e-05, + "loss": 3.1527, + "step": 11134 + }, + { + "epoch": 0.050433971212536004, + "grad_norm": 0.14507418870925903, + "learning_rate": 4.150151571823296e-05, + "loss": 3.2079, + "step": 11135 + }, + { + "epoch": 0.05046481767810942, + "grad_norm": 0.14805097877979279, + "learning_rate": 4.149949340984749e-05, + "loss": 3.1783, + "step": 11136 + }, + { + "epoch": 0.05049566414368284, + "grad_norm": 0.14621078968048096, + "learning_rate": 4.149747091016032e-05, + "loss": 3.1737, + "step": 11137 + }, + { + "epoch": 0.05052651060925625, + "grad_norm": 0.14321652054786682, + "learning_rate": 4.1495448219194875e-05, + "loss": 3.1817, + "step": 11138 + }, + { + "epoch": 0.05055735707482967, + "grad_norm": 0.14967963099479675, + "learning_rate": 4.149342533697462e-05, + "loss": 3.1728, + "step": 11139 + }, + { + "epoch": 0.05058820354040309, + "grad_norm": 0.14274685084819794, + "learning_rate": 4.149140226352302e-05, + "loss": 3.1678, + "step": 11140 + }, + { + "epoch": 0.0506190500059765, + "grad_norm": 0.14852526783943176, + "learning_rate": 4.148937899886352e-05, + "loss": 3.2029, + "step": 11141 + }, + { + "epoch": 0.05064989647154992, + "grad_norm": 0.15255625545978546, + "learning_rate": 4.148735554301958e-05, + "loss": 3.1885, + "step": 11142 + }, + { + "epoch": 0.05068074293712334, + "grad_norm": 0.14264748990535736, + "learning_rate": 4.148533189601465e-05, + "loss": 3.1264, + "step": 11143 + }, + { + "epoch": 0.05071158940269675, + "grad_norm": 0.14717063307762146, + "learning_rate": 4.14833080578722e-05, + "loss": 3.1696, + "step": 11144 + }, + { + "epoch": 0.050742435868270166, + "grad_norm": 0.1538676917552948, + "learning_rate": 4.148128402861572e-05, + "loss": 3.1939, + "step": 11145 + }, + { + "epoch": 0.05077328233384359, + "grad_norm": 0.14538685977458954, + "learning_rate": 4.1479259808268636e-05, + "loss": 3.1694, + "step": 11146 + }, + { + "epoch": 0.050804128799417, + "grad_norm": 0.15424562990665436, + "learning_rate": 4.1477235396854444e-05, + "loss": 3.1786, + "step": 11147 + }, + { + "epoch": 0.050834975264990416, + "grad_norm": 0.1507306545972824, + "learning_rate": 4.147521079439661e-05, + "loss": 3.1611, + "step": 11148 + }, + { + "epoch": 0.05086582173056384, + "grad_norm": 0.14004822075366974, + "learning_rate": 4.1473186000918614e-05, + "loss": 3.148, + "step": 11149 + }, + { + "epoch": 0.05089666819613725, + "grad_norm": 0.149087056517601, + "learning_rate": 4.1471161016443915e-05, + "loss": 3.1977, + "step": 11150 + }, + { + "epoch": 0.050927514661710666, + "grad_norm": 0.14731085300445557, + "learning_rate": 4.146913584099601e-05, + "loss": 3.2022, + "step": 11151 + }, + { + "epoch": 0.05095836112728409, + "grad_norm": 0.15368255972862244, + "learning_rate": 4.146711047459837e-05, + "loss": 3.1576, + "step": 11152 + }, + { + "epoch": 0.0509892075928575, + "grad_norm": 0.14426545798778534, + "learning_rate": 4.146508491727447e-05, + "loss": 3.1845, + "step": 11153 + }, + { + "epoch": 0.051020054058430915, + "grad_norm": 0.14877773821353912, + "learning_rate": 4.1463059169047814e-05, + "loss": 3.1653, + "step": 11154 + }, + { + "epoch": 0.051050900524004336, + "grad_norm": 0.14742955565452576, + "learning_rate": 4.1461033229941885e-05, + "loss": 3.2382, + "step": 11155 + }, + { + "epoch": 0.05108174698957775, + "grad_norm": 0.15454117953777313, + "learning_rate": 4.1459007099980154e-05, + "loss": 3.1817, + "step": 11156 + }, + { + "epoch": 0.051112593455151165, + "grad_norm": 0.1562579870223999, + "learning_rate": 4.1456980779186135e-05, + "loss": 3.1754, + "step": 11157 + }, + { + "epoch": 0.051143439920724586, + "grad_norm": 0.1617935299873352, + "learning_rate": 4.1454954267583315e-05, + "loss": 3.2057, + "step": 11158 + }, + { + "epoch": 0.051174286386298, + "grad_norm": 0.15260685980319977, + "learning_rate": 4.145292756519517e-05, + "loss": 3.1629, + "step": 11159 + }, + { + "epoch": 0.051205132851871414, + "grad_norm": 0.1621655970811844, + "learning_rate": 4.1450900672045244e-05, + "loss": 3.1922, + "step": 11160 + }, + { + "epoch": 0.051235979317444835, + "grad_norm": 0.14721401035785675, + "learning_rate": 4.144887358815699e-05, + "loss": 3.1824, + "step": 11161 + }, + { + "epoch": 0.05126682578301825, + "grad_norm": 0.1595020443201065, + "learning_rate": 4.144684631355394e-05, + "loss": 3.1681, + "step": 11162 + }, + { + "epoch": 0.051297672248591664, + "grad_norm": 0.16177430748939514, + "learning_rate": 4.1444818848259596e-05, + "loss": 3.1786, + "step": 11163 + }, + { + "epoch": 0.051328518714165085, + "grad_norm": 0.14564107358455658, + "learning_rate": 4.1442791192297456e-05, + "loss": 3.1735, + "step": 11164 + }, + { + "epoch": 0.0513593651797385, + "grad_norm": 0.31918516755104065, + "learning_rate": 4.144076334569102e-05, + "loss": 3.1898, + "step": 11165 + }, + { + "epoch": 0.05139021164531191, + "grad_norm": 0.18058931827545166, + "learning_rate": 4.1438735308463835e-05, + "loss": 3.2129, + "step": 11166 + }, + { + "epoch": 0.051421058110885334, + "grad_norm": 0.16914717853069305, + "learning_rate": 4.143670708063938e-05, + "loss": 3.1912, + "step": 11167 + }, + { + "epoch": 0.05145190457645875, + "grad_norm": 0.1735004037618637, + "learning_rate": 4.143467866224118e-05, + "loss": 3.1933, + "step": 11168 + }, + { + "epoch": 0.05148275104203216, + "grad_norm": 0.1558932363986969, + "learning_rate": 4.143265005329276e-05, + "loss": 3.1846, + "step": 11169 + }, + { + "epoch": 0.051513597507605584, + "grad_norm": 0.20767757296562195, + "learning_rate": 4.1430621253817644e-05, + "loss": 3.1938, + "step": 11170 + }, + { + "epoch": 0.051544443973179, + "grad_norm": 0.15072569251060486, + "learning_rate": 4.142859226383934e-05, + "loss": 3.1884, + "step": 11171 + }, + { + "epoch": 0.05157529043875241, + "grad_norm": 0.15076445043087006, + "learning_rate": 4.1426563083381394e-05, + "loss": 3.1631, + "step": 11172 + }, + { + "epoch": 0.05160613690432583, + "grad_norm": 0.14991778135299683, + "learning_rate": 4.14245337124673e-05, + "loss": 3.1398, + "step": 11173 + }, + { + "epoch": 0.05163698336989925, + "grad_norm": 0.1572091430425644, + "learning_rate": 4.142250415112062e-05, + "loss": 3.1409, + "step": 11174 + }, + { + "epoch": 0.05166782983547266, + "grad_norm": 0.14744779467582703, + "learning_rate": 4.1420474399364884e-05, + "loss": 3.1906, + "step": 11175 + }, + { + "epoch": 0.05169867630104608, + "grad_norm": 0.15225954353809357, + "learning_rate": 4.14184444572236e-05, + "loss": 3.1482, + "step": 11176 + }, + { + "epoch": 0.0517295227666195, + "grad_norm": 0.149088054895401, + "learning_rate": 4.141641432472032e-05, + "loss": 3.1369, + "step": 11177 + }, + { + "epoch": 0.05176036923219291, + "grad_norm": 0.18245135247707367, + "learning_rate": 4.141438400187858e-05, + "loss": 3.1998, + "step": 11178 + }, + { + "epoch": 0.05179121569776633, + "grad_norm": 0.14620521664619446, + "learning_rate": 4.1412353488721925e-05, + "loss": 3.1616, + "step": 11179 + }, + { + "epoch": 0.05182206216333975, + "grad_norm": 0.14589917659759521, + "learning_rate": 4.141032278527389e-05, + "loss": 3.1776, + "step": 11180 + }, + { + "epoch": 0.05185290862891316, + "grad_norm": 0.1461264044046402, + "learning_rate": 4.140829189155803e-05, + "loss": 3.1673, + "step": 11181 + }, + { + "epoch": 0.05188375509448658, + "grad_norm": 0.14713023602962494, + "learning_rate": 4.1406260807597886e-05, + "loss": 3.186, + "step": 11182 + }, + { + "epoch": 0.051914601560059996, + "grad_norm": 0.16004444658756256, + "learning_rate": 4.1404229533417005e-05, + "loss": 3.2125, + "step": 11183 + }, + { + "epoch": 0.05194544802563341, + "grad_norm": 0.15593405067920685, + "learning_rate": 4.1402198069038934e-05, + "loss": 3.1812, + "step": 11184 + }, + { + "epoch": 0.05197629449120683, + "grad_norm": 0.1472742110490799, + "learning_rate": 4.140016641448724e-05, + "loss": 3.1601, + "step": 11185 + }, + { + "epoch": 0.052007140956780246, + "grad_norm": 0.14258360862731934, + "learning_rate": 4.1398134569785476e-05, + "loss": 3.1666, + "step": 11186 + }, + { + "epoch": 0.05203798742235366, + "grad_norm": 0.1499461829662323, + "learning_rate": 4.1396102534957184e-05, + "loss": 3.1698, + "step": 11187 + }, + { + "epoch": 0.05206883388792708, + "grad_norm": 0.154301717877388, + "learning_rate": 4.139407031002594e-05, + "loss": 3.1839, + "step": 11188 + }, + { + "epoch": 0.052099680353500495, + "grad_norm": 0.14578059315681458, + "learning_rate": 4.13920378950153e-05, + "loss": 3.142, + "step": 11189 + }, + { + "epoch": 0.05213052681907391, + "grad_norm": 0.14973020553588867, + "learning_rate": 4.139000528994884e-05, + "loss": 3.1875, + "step": 11190 + }, + { + "epoch": 0.05216137328464733, + "grad_norm": 0.1485022008419037, + "learning_rate": 4.138797249485011e-05, + "loss": 3.1548, + "step": 11191 + }, + { + "epoch": 0.052192219750220745, + "grad_norm": 0.14705903828144073, + "learning_rate": 4.138593950974269e-05, + "loss": 3.1828, + "step": 11192 + }, + { + "epoch": 0.05222306621579416, + "grad_norm": 0.14844976365566254, + "learning_rate": 4.138390633465015e-05, + "loss": 3.1829, + "step": 11193 + }, + { + "epoch": 0.05225391268136758, + "grad_norm": 0.148152157664299, + "learning_rate": 4.138187296959606e-05, + "loss": 3.1862, + "step": 11194 + }, + { + "epoch": 0.052284759146940994, + "grad_norm": 0.15059351921081543, + "learning_rate": 4.1379839414604e-05, + "loss": 3.1703, + "step": 11195 + }, + { + "epoch": 0.05231560561251441, + "grad_norm": 0.15600991249084473, + "learning_rate": 4.137780566969754e-05, + "loss": 3.1964, + "step": 11196 + }, + { + "epoch": 0.05234645207808783, + "grad_norm": 0.14462697505950928, + "learning_rate": 4.1375771734900264e-05, + "loss": 3.1751, + "step": 11197 + }, + { + "epoch": 0.052377298543661244, + "grad_norm": 0.15039187669754028, + "learning_rate": 4.137373761023576e-05, + "loss": 3.1728, + "step": 11198 + }, + { + "epoch": 0.05240814500923466, + "grad_norm": 0.14681291580200195, + "learning_rate": 4.137170329572761e-05, + "loss": 3.1454, + "step": 11199 + }, + { + "epoch": 0.05243899147480808, + "grad_norm": 0.14374297857284546, + "learning_rate": 4.1369668791399393e-05, + "loss": 3.1809, + "step": 11200 + }, + { + "epoch": 0.05246983794038149, + "grad_norm": 0.14717012643814087, + "learning_rate": 4.136763409727471e-05, + "loss": 3.1622, + "step": 11201 + }, + { + "epoch": 0.05250068440595491, + "grad_norm": 0.14239773154258728, + "learning_rate": 4.136559921337714e-05, + "loss": 3.1456, + "step": 11202 + }, + { + "epoch": 0.05253153087152833, + "grad_norm": 0.15662071108818054, + "learning_rate": 4.136356413973028e-05, + "loss": 3.2007, + "step": 11203 + }, + { + "epoch": 0.05256237733710174, + "grad_norm": 0.1497691124677658, + "learning_rate": 4.136152887635773e-05, + "loss": 3.1849, + "step": 11204 + }, + { + "epoch": 0.05259322380267516, + "grad_norm": 0.14706666767597198, + "learning_rate": 4.135949342328308e-05, + "loss": 3.1962, + "step": 11205 + }, + { + "epoch": 0.05262407026824858, + "grad_norm": 0.15165264904499054, + "learning_rate": 4.135745778052994e-05, + "loss": 3.155, + "step": 11206 + }, + { + "epoch": 0.05265491673382199, + "grad_norm": 0.14746586978435516, + "learning_rate": 4.135542194812191e-05, + "loss": 3.1527, + "step": 11207 + }, + { + "epoch": 0.052685763199395406, + "grad_norm": 0.14842161536216736, + "learning_rate": 4.135338592608259e-05, + "loss": 3.1354, + "step": 11208 + }, + { + "epoch": 0.05271660966496883, + "grad_norm": 0.14643479883670807, + "learning_rate": 4.135134971443558e-05, + "loss": 3.1724, + "step": 11209 + }, + { + "epoch": 0.05274745613054224, + "grad_norm": 0.14163252711296082, + "learning_rate": 4.13493133132045e-05, + "loss": 3.1776, + "step": 11210 + }, + { + "epoch": 0.052778302596115656, + "grad_norm": 0.14424635469913483, + "learning_rate": 4.134727672241296e-05, + "loss": 3.1285, + "step": 11211 + }, + { + "epoch": 0.05280914906168908, + "grad_norm": 0.1432003676891327, + "learning_rate": 4.134523994208457e-05, + "loss": 3.2039, + "step": 11212 + }, + { + "epoch": 0.05283999552726249, + "grad_norm": 0.15384328365325928, + "learning_rate": 4.1343202972242936e-05, + "loss": 3.1433, + "step": 11213 + }, + { + "epoch": 0.052870841992835906, + "grad_norm": 0.14366263151168823, + "learning_rate": 4.134116581291169e-05, + "loss": 3.1355, + "step": 11214 + }, + { + "epoch": 0.05290168845840933, + "grad_norm": 0.15260049700737, + "learning_rate": 4.133912846411445e-05, + "loss": 3.1958, + "step": 11215 + }, + { + "epoch": 0.05293253492398274, + "grad_norm": 0.15102390944957733, + "learning_rate": 4.1337090925874835e-05, + "loss": 3.1737, + "step": 11216 + }, + { + "epoch": 0.052963381389556155, + "grad_norm": 0.1478048712015152, + "learning_rate": 4.1335053198216465e-05, + "loss": 3.1778, + "step": 11217 + }, + { + "epoch": 0.052994227855129576, + "grad_norm": 0.15411481261253357, + "learning_rate": 4.133301528116297e-05, + "loss": 3.1348, + "step": 11218 + }, + { + "epoch": 0.05302507432070299, + "grad_norm": 0.15556679666042328, + "learning_rate": 4.1330977174737976e-05, + "loss": 3.1791, + "step": 11219 + }, + { + "epoch": 0.053055920786276405, + "grad_norm": 0.15188336372375488, + "learning_rate": 4.132893887896512e-05, + "loss": 3.1729, + "step": 11220 + }, + { + "epoch": 0.053086767251849826, + "grad_norm": 0.15529125928878784, + "learning_rate": 4.132690039386802e-05, + "loss": 3.1817, + "step": 11221 + }, + { + "epoch": 0.05311761371742324, + "grad_norm": 0.16070424020290375, + "learning_rate": 4.132486171947033e-05, + "loss": 3.1602, + "step": 11222 + }, + { + "epoch": 0.053148460182996654, + "grad_norm": 0.15381808578968048, + "learning_rate": 4.1322822855795676e-05, + "loss": 3.1832, + "step": 11223 + }, + { + "epoch": 0.053179306648570075, + "grad_norm": 0.14264680445194244, + "learning_rate": 4.13207838028677e-05, + "loss": 3.1499, + "step": 11224 + }, + { + "epoch": 0.05321015311414349, + "grad_norm": 0.15399269759655, + "learning_rate": 4.131874456071005e-05, + "loss": 3.1648, + "step": 11225 + }, + { + "epoch": 0.053240999579716904, + "grad_norm": 0.14460642635822296, + "learning_rate": 4.131670512934635e-05, + "loss": 3.1458, + "step": 11226 + }, + { + "epoch": 0.053271846045290325, + "grad_norm": 0.1517481505870819, + "learning_rate": 4.131466550880027e-05, + "loss": 3.1839, + "step": 11227 + }, + { + "epoch": 0.05330269251086374, + "grad_norm": 0.1450367122888565, + "learning_rate": 4.131262569909544e-05, + "loss": 3.1754, + "step": 11228 + }, + { + "epoch": 0.05333353897643715, + "grad_norm": 0.15681497752666473, + "learning_rate": 4.1310585700255523e-05, + "loss": 3.1762, + "step": 11229 + }, + { + "epoch": 0.053364385442010574, + "grad_norm": 0.1495969295501709, + "learning_rate": 4.1308545512304164e-05, + "loss": 3.2006, + "step": 11230 + }, + { + "epoch": 0.05339523190758399, + "grad_norm": 0.15018180012702942, + "learning_rate": 4.1306505135265024e-05, + "loss": 3.1601, + "step": 11231 + }, + { + "epoch": 0.0534260783731574, + "grad_norm": 0.1471392810344696, + "learning_rate": 4.130446456916175e-05, + "loss": 3.1697, + "step": 11232 + }, + { + "epoch": 0.053456924838730824, + "grad_norm": 0.1601645052433014, + "learning_rate": 4.130242381401801e-05, + "loss": 3.2238, + "step": 11233 + }, + { + "epoch": 0.05348777130430424, + "grad_norm": 0.14847244322299957, + "learning_rate": 4.130038286985747e-05, + "loss": 3.184, + "step": 11234 + }, + { + "epoch": 0.05351861776987765, + "grad_norm": 0.1484220027923584, + "learning_rate": 4.129834173670377e-05, + "loss": 3.1889, + "step": 11235 + }, + { + "epoch": 0.05354946423545107, + "grad_norm": 0.15048234164714813, + "learning_rate": 4.12963004145806e-05, + "loss": 3.2003, + "step": 11236 + }, + { + "epoch": 0.05358031070102449, + "grad_norm": 0.14027680456638336, + "learning_rate": 4.129425890351162e-05, + "loss": 3.1474, + "step": 11237 + }, + { + "epoch": 0.0536111571665979, + "grad_norm": 0.1527215838432312, + "learning_rate": 4.12922172035205e-05, + "loss": 3.2061, + "step": 11238 + }, + { + "epoch": 0.05364200363217132, + "grad_norm": 0.1470610499382019, + "learning_rate": 4.129017531463092e-05, + "loss": 3.189, + "step": 11239 + }, + { + "epoch": 0.05367285009774474, + "grad_norm": 0.1442156881093979, + "learning_rate": 4.1288133236866535e-05, + "loss": 3.1783, + "step": 11240 + }, + { + "epoch": 0.05370369656331815, + "grad_norm": 0.1523507535457611, + "learning_rate": 4.128609097025104e-05, + "loss": 3.1706, + "step": 11241 + }, + { + "epoch": 0.05373454302889157, + "grad_norm": 0.14583630859851837, + "learning_rate": 4.12840485148081e-05, + "loss": 3.1701, + "step": 11242 + }, + { + "epoch": 0.05376538949446499, + "grad_norm": 0.15631891787052155, + "learning_rate": 4.1282005870561404e-05, + "loss": 3.1917, + "step": 11243 + }, + { + "epoch": 0.0537962359600384, + "grad_norm": 0.14706967771053314, + "learning_rate": 4.1279963037534635e-05, + "loss": 3.178, + "step": 11244 + }, + { + "epoch": 0.05382708242561182, + "grad_norm": 0.15194793045520782, + "learning_rate": 4.127792001575148e-05, + "loss": 3.166, + "step": 11245 + }, + { + "epoch": 0.053857928891185236, + "grad_norm": 0.1549869030714035, + "learning_rate": 4.127587680523562e-05, + "loss": 3.1622, + "step": 11246 + }, + { + "epoch": 0.05388877535675865, + "grad_norm": 0.14462910592556, + "learning_rate": 4.127383340601076e-05, + "loss": 3.1616, + "step": 11247 + }, + { + "epoch": 0.05391962182233207, + "grad_norm": 0.1470925360918045, + "learning_rate": 4.127178981810057e-05, + "loss": 3.189, + "step": 11248 + }, + { + "epoch": 0.053950468287905486, + "grad_norm": 0.17916050553321838, + "learning_rate": 4.1269746041528766e-05, + "loss": 3.166, + "step": 11249 + }, + { + "epoch": 0.0539813147534789, + "grad_norm": 0.1597147136926651, + "learning_rate": 4.126770207631903e-05, + "loss": 3.2095, + "step": 11250 + }, + { + "epoch": 0.05401216121905232, + "grad_norm": 0.1536785066127777, + "learning_rate": 4.126565792249505e-05, + "loss": 3.164, + "step": 11251 + }, + { + "epoch": 0.054043007684625735, + "grad_norm": 0.1552533358335495, + "learning_rate": 4.126361358008056e-05, + "loss": 3.2165, + "step": 11252 + }, + { + "epoch": 0.05407385415019915, + "grad_norm": 0.1526089459657669, + "learning_rate": 4.126156904909923e-05, + "loss": 3.1534, + "step": 11253 + }, + { + "epoch": 0.05410470061577257, + "grad_norm": 0.16266195476055145, + "learning_rate": 4.1259524329574786e-05, + "loss": 3.1893, + "step": 11254 + }, + { + "epoch": 0.054135547081345985, + "grad_norm": 0.15138578414916992, + "learning_rate": 4.125747942153093e-05, + "loss": 3.1743, + "step": 11255 + }, + { + "epoch": 0.0541663935469194, + "grad_norm": 0.1532023847103119, + "learning_rate": 4.125543432499136e-05, + "loss": 3.1856, + "step": 11256 + }, + { + "epoch": 0.05419724001249282, + "grad_norm": 0.14363721013069153, + "learning_rate": 4.12533890399798e-05, + "loss": 3.15, + "step": 11257 + }, + { + "epoch": 0.054228086478066234, + "grad_norm": 0.14811143279075623, + "learning_rate": 4.1251343566519976e-05, + "loss": 3.1832, + "step": 11258 + }, + { + "epoch": 0.05425893294363965, + "grad_norm": 0.1508772224187851, + "learning_rate": 4.1249297904635575e-05, + "loss": 3.1751, + "step": 11259 + }, + { + "epoch": 0.05428977940921307, + "grad_norm": 0.15141461789608002, + "learning_rate": 4.124725205435033e-05, + "loss": 3.1957, + "step": 11260 + }, + { + "epoch": 0.054320625874786484, + "grad_norm": 0.14683638513088226, + "learning_rate": 4.1245206015687964e-05, + "loss": 3.1691, + "step": 11261 + }, + { + "epoch": 0.0543514723403599, + "grad_norm": 0.14305022358894348, + "learning_rate": 4.12431597886722e-05, + "loss": 3.165, + "step": 11262 + }, + { + "epoch": 0.05438231880593332, + "grad_norm": 0.15135721862316132, + "learning_rate": 4.124111337332676e-05, + "loss": 3.1592, + "step": 11263 + }, + { + "epoch": 0.05441316527150673, + "grad_norm": 0.1438494324684143, + "learning_rate": 4.123906676967536e-05, + "loss": 3.1574, + "step": 11264 + }, + { + "epoch": 0.05444401173708015, + "grad_norm": 0.1551428586244583, + "learning_rate": 4.123701997774174e-05, + "loss": 3.2065, + "step": 11265 + }, + { + "epoch": 0.05447485820265357, + "grad_norm": 0.15208245813846588, + "learning_rate": 4.123497299754964e-05, + "loss": 3.1634, + "step": 11266 + }, + { + "epoch": 0.05450570466822698, + "grad_norm": 0.1423480361700058, + "learning_rate": 4.123292582912278e-05, + "loss": 3.1357, + "step": 11267 + }, + { + "epoch": 0.0545365511338004, + "grad_norm": 0.14994840323925018, + "learning_rate": 4.1230878472484904e-05, + "loss": 3.1614, + "step": 11268 + }, + { + "epoch": 0.05456739759937382, + "grad_norm": 0.1468484252691269, + "learning_rate": 4.122883092765974e-05, + "loss": 3.2081, + "step": 11269 + }, + { + "epoch": 0.05459824406494723, + "grad_norm": 0.1490316092967987, + "learning_rate": 4.122678319467104e-05, + "loss": 3.173, + "step": 11270 + }, + { + "epoch": 0.054629090530520646, + "grad_norm": 0.14929205179214478, + "learning_rate": 4.122473527354253e-05, + "loss": 3.1566, + "step": 11271 + }, + { + "epoch": 0.05465993699609407, + "grad_norm": 0.14941047132015228, + "learning_rate": 4.122268716429797e-05, + "loss": 3.1536, + "step": 11272 + }, + { + "epoch": 0.05469078346166748, + "grad_norm": 0.15637117624282837, + "learning_rate": 4.1220638866961105e-05, + "loss": 3.1662, + "step": 11273 + }, + { + "epoch": 0.054721629927240896, + "grad_norm": 0.14833542704582214, + "learning_rate": 4.121859038155567e-05, + "loss": 3.1469, + "step": 11274 + }, + { + "epoch": 0.05475247639281432, + "grad_norm": 0.14192749559879303, + "learning_rate": 4.1216541708105435e-05, + "loss": 3.1742, + "step": 11275 + }, + { + "epoch": 0.05478332285838773, + "grad_norm": 0.14945010840892792, + "learning_rate": 4.121449284663414e-05, + "loss": 3.1987, + "step": 11276 + }, + { + "epoch": 0.054814169323961145, + "grad_norm": 0.1492098718881607, + "learning_rate": 4.1212443797165546e-05, + "loss": 3.1993, + "step": 11277 + }, + { + "epoch": 0.05484501578953457, + "grad_norm": 0.147098571062088, + "learning_rate": 4.121039455972341e-05, + "loss": 3.1658, + "step": 11278 + }, + { + "epoch": 0.05487586225510798, + "grad_norm": 0.14740096032619476, + "learning_rate": 4.120834513433148e-05, + "loss": 3.1632, + "step": 11279 + }, + { + "epoch": 0.054906708720681395, + "grad_norm": 0.1514090895652771, + "learning_rate": 4.1206295521013535e-05, + "loss": 3.147, + "step": 11280 + }, + { + "epoch": 0.054937555186254816, + "grad_norm": 0.14672629535198212, + "learning_rate": 4.1204245719793336e-05, + "loss": 3.1934, + "step": 11281 + }, + { + "epoch": 0.05496840165182823, + "grad_norm": 0.14488659799098969, + "learning_rate": 4.120219573069464e-05, + "loss": 3.171, + "step": 11282 + }, + { + "epoch": 0.054999248117401645, + "grad_norm": 0.14393803477287292, + "learning_rate": 4.1200145553741225e-05, + "loss": 3.163, + "step": 11283 + }, + { + "epoch": 0.055030094582975066, + "grad_norm": 0.15363916754722595, + "learning_rate": 4.1198095188956856e-05, + "loss": 3.2178, + "step": 11284 + }, + { + "epoch": 0.05506094104854848, + "grad_norm": 0.1470312774181366, + "learning_rate": 4.1196044636365306e-05, + "loss": 3.1649, + "step": 11285 + }, + { + "epoch": 0.055091787514121894, + "grad_norm": 0.1463627815246582, + "learning_rate": 4.119399389599036e-05, + "loss": 3.1699, + "step": 11286 + }, + { + "epoch": 0.055122633979695315, + "grad_norm": 0.1465693861246109, + "learning_rate": 4.1191942967855776e-05, + "loss": 3.196, + "step": 11287 + }, + { + "epoch": 0.05515348044526873, + "grad_norm": 0.14947621524333954, + "learning_rate": 4.118989185198534e-05, + "loss": 3.1317, + "step": 11288 + }, + { + "epoch": 0.05518432691084215, + "grad_norm": 0.14770428836345673, + "learning_rate": 4.1187840548402844e-05, + "loss": 3.2053, + "step": 11289 + }, + { + "epoch": 0.055215173376415565, + "grad_norm": 0.14380937814712524, + "learning_rate": 4.1185789057132065e-05, + "loss": 3.2093, + "step": 11290 + }, + { + "epoch": 0.05524601984198898, + "grad_norm": 0.15020346641540527, + "learning_rate": 4.118373737819679e-05, + "loss": 3.2133, + "step": 11291 + }, + { + "epoch": 0.0552768663075624, + "grad_norm": 0.14866437017917633, + "learning_rate": 4.118168551162081e-05, + "loss": 3.1698, + "step": 11292 + }, + { + "epoch": 0.055307712773135814, + "grad_norm": 0.14716894924640656, + "learning_rate": 4.1179633457427897e-05, + "loss": 3.2055, + "step": 11293 + }, + { + "epoch": 0.05533855923870923, + "grad_norm": 0.14619019627571106, + "learning_rate": 4.117758121564186e-05, + "loss": 3.1567, + "step": 11294 + }, + { + "epoch": 0.05536940570428265, + "grad_norm": 0.14943474531173706, + "learning_rate": 4.11755287862865e-05, + "loss": 3.1792, + "step": 11295 + }, + { + "epoch": 0.055400252169856064, + "grad_norm": 0.143184632062912, + "learning_rate": 4.117347616938559e-05, + "loss": 3.1468, + "step": 11296 + }, + { + "epoch": 0.05543109863542948, + "grad_norm": 0.15401722490787506, + "learning_rate": 4.117142336496295e-05, + "loss": 3.1616, + "step": 11297 + }, + { + "epoch": 0.0554619451010029, + "grad_norm": 0.14861242473125458, + "learning_rate": 4.1169370373042374e-05, + "loss": 3.194, + "step": 11298 + }, + { + "epoch": 0.05549279156657631, + "grad_norm": 0.14177149534225464, + "learning_rate": 4.116731719364767e-05, + "loss": 3.1584, + "step": 11299 + }, + { + "epoch": 0.05552363803214973, + "grad_norm": 0.15398180484771729, + "learning_rate": 4.116526382680264e-05, + "loss": 3.1355, + "step": 11300 + }, + { + "epoch": 0.05555448449772315, + "grad_norm": 0.14809000492095947, + "learning_rate": 4.1163210272531075e-05, + "loss": 3.1925, + "step": 11301 + }, + { + "epoch": 0.05558533096329656, + "grad_norm": 0.1450880765914917, + "learning_rate": 4.116115653085681e-05, + "loss": 3.166, + "step": 11302 + }, + { + "epoch": 0.05561617742886998, + "grad_norm": 0.14581114053726196, + "learning_rate": 4.1159102601803644e-05, + "loss": 3.1821, + "step": 11303 + }, + { + "epoch": 0.0556470238944434, + "grad_norm": 0.15935778617858887, + "learning_rate": 4.11570484853954e-05, + "loss": 3.1631, + "step": 11304 + }, + { + "epoch": 0.05567787036001681, + "grad_norm": 0.14423640072345734, + "learning_rate": 4.1154994181655884e-05, + "loss": 3.1686, + "step": 11305 + }, + { + "epoch": 0.055708716825590227, + "grad_norm": 0.14736154675483704, + "learning_rate": 4.115293969060892e-05, + "loss": 3.1599, + "step": 11306 + }, + { + "epoch": 0.05573956329116365, + "grad_norm": 0.21962666511535645, + "learning_rate": 4.1150885012278324e-05, + "loss": 3.1704, + "step": 11307 + }, + { + "epoch": 0.05577040975673706, + "grad_norm": 0.14811967313289642, + "learning_rate": 4.114883014668792e-05, + "loss": 3.2271, + "step": 11308 + }, + { + "epoch": 0.055801256222310476, + "grad_norm": 0.14685389399528503, + "learning_rate": 4.1146775093861546e-05, + "loss": 3.1414, + "step": 11309 + }, + { + "epoch": 0.0558321026878839, + "grad_norm": 0.14990873634815216, + "learning_rate": 4.1144719853823005e-05, + "loss": 3.2165, + "step": 11310 + }, + { + "epoch": 0.05586294915345731, + "grad_norm": 0.14301833510398865, + "learning_rate": 4.1142664426596146e-05, + "loss": 3.164, + "step": 11311 + }, + { + "epoch": 0.055893795619030726, + "grad_norm": 0.14887216687202454, + "learning_rate": 4.114060881220478e-05, + "loss": 3.1733, + "step": 11312 + }, + { + "epoch": 0.05592464208460415, + "grad_norm": 0.14182059466838837, + "learning_rate": 4.1138553010672775e-05, + "loss": 3.1567, + "step": 11313 + }, + { + "epoch": 0.05595548855017756, + "grad_norm": 0.14883233606815338, + "learning_rate": 4.1136497022023934e-05, + "loss": 3.1617, + "step": 11314 + }, + { + "epoch": 0.055986335015750975, + "grad_norm": 0.14957527816295624, + "learning_rate": 4.113444084628211e-05, + "loss": 3.2098, + "step": 11315 + }, + { + "epoch": 0.056017181481324396, + "grad_norm": 0.14799237251281738, + "learning_rate": 4.113238448347114e-05, + "loss": 3.1817, + "step": 11316 + }, + { + "epoch": 0.05604802794689781, + "grad_norm": 0.14606358110904694, + "learning_rate": 4.113032793361487e-05, + "loss": 3.2052, + "step": 11317 + }, + { + "epoch": 0.056078874412471225, + "grad_norm": 0.15971721708774567, + "learning_rate": 4.1128271196737135e-05, + "loss": 3.1999, + "step": 11318 + }, + { + "epoch": 0.056109720878044646, + "grad_norm": 0.19133636355400085, + "learning_rate": 4.112621427286179e-05, + "loss": 3.2078, + "step": 11319 + }, + { + "epoch": 0.05614056734361806, + "grad_norm": 0.15580230951309204, + "learning_rate": 4.112415716201268e-05, + "loss": 3.1534, + "step": 11320 + }, + { + "epoch": 0.056171413809191474, + "grad_norm": 0.14945630729198456, + "learning_rate": 4.1122099864213656e-05, + "loss": 3.1623, + "step": 11321 + }, + { + "epoch": 0.056202260274764895, + "grad_norm": 0.14722713828086853, + "learning_rate": 4.1120042379488576e-05, + "loss": 3.1791, + "step": 11322 + }, + { + "epoch": 0.05623310674033831, + "grad_norm": 0.15630318224430084, + "learning_rate": 4.111798470786129e-05, + "loss": 3.1677, + "step": 11323 + }, + { + "epoch": 0.056263953205911724, + "grad_norm": 0.1653730720281601, + "learning_rate": 4.1115926849355656e-05, + "loss": 3.1883, + "step": 11324 + }, + { + "epoch": 0.056294799671485145, + "grad_norm": 0.15580283105373383, + "learning_rate": 4.1113868803995534e-05, + "loss": 3.2158, + "step": 11325 + }, + { + "epoch": 0.05632564613705856, + "grad_norm": 0.14916029572486877, + "learning_rate": 4.1111810571804786e-05, + "loss": 3.1507, + "step": 11326 + }, + { + "epoch": 0.05635649260263197, + "grad_norm": 0.15829645097255707, + "learning_rate": 4.110975215280728e-05, + "loss": 3.1975, + "step": 11327 + }, + { + "epoch": 0.056387339068205394, + "grad_norm": 0.15711399912834167, + "learning_rate": 4.110769354702688e-05, + "loss": 3.2207, + "step": 11328 + }, + { + "epoch": 0.05641818553377881, + "grad_norm": 0.1450992077589035, + "learning_rate": 4.110563475448745e-05, + "loss": 3.1649, + "step": 11329 + }, + { + "epoch": 0.05644903199935222, + "grad_norm": 0.15424250066280365, + "learning_rate": 4.110357577521286e-05, + "loss": 3.1233, + "step": 11330 + }, + { + "epoch": 0.056479878464925644, + "grad_norm": 0.14657717943191528, + "learning_rate": 4.1101516609226994e-05, + "loss": 3.1846, + "step": 11331 + }, + { + "epoch": 0.05651072493049906, + "grad_norm": 0.1512654423713684, + "learning_rate": 4.109945725655371e-05, + "loss": 3.1418, + "step": 11332 + }, + { + "epoch": 0.05654157139607247, + "grad_norm": 0.1483934223651886, + "learning_rate": 4.1097397717216904e-05, + "loss": 3.167, + "step": 11333 + }, + { + "epoch": 0.05657241786164589, + "grad_norm": 0.15051275491714478, + "learning_rate": 4.1095337991240436e-05, + "loss": 3.1749, + "step": 11334 + }, + { + "epoch": 0.05660326432721931, + "grad_norm": 0.14264138042926788, + "learning_rate": 4.10932780786482e-05, + "loss": 3.1379, + "step": 11335 + }, + { + "epoch": 0.05663411079279272, + "grad_norm": 0.15152306854724884, + "learning_rate": 4.109121797946407e-05, + "loss": 3.1828, + "step": 11336 + }, + { + "epoch": 0.05666495725836614, + "grad_norm": 0.1465742588043213, + "learning_rate": 4.108915769371194e-05, + "loss": 3.1602, + "step": 11337 + }, + { + "epoch": 0.05669580372393956, + "grad_norm": 0.14601586759090424, + "learning_rate": 4.108709722141571e-05, + "loss": 3.1973, + "step": 11338 + }, + { + "epoch": 0.05672665018951297, + "grad_norm": 0.1480490267276764, + "learning_rate": 4.108503656259923e-05, + "loss": 3.1849, + "step": 11339 + }, + { + "epoch": 0.05675749665508639, + "grad_norm": 0.14752961695194244, + "learning_rate": 4.108297571728643e-05, + "loss": 3.1523, + "step": 11340 + }, + { + "epoch": 0.05678834312065981, + "grad_norm": 0.14969545602798462, + "learning_rate": 4.1080914685501196e-05, + "loss": 3.1501, + "step": 11341 + }, + { + "epoch": 0.05681918958623322, + "grad_norm": 0.14391307532787323, + "learning_rate": 4.1078853467267415e-05, + "loss": 3.1557, + "step": 11342 + }, + { + "epoch": 0.05685003605180664, + "grad_norm": 0.14579786360263824, + "learning_rate": 4.107679206260899e-05, + "loss": 3.1806, + "step": 11343 + }, + { + "epoch": 0.056880882517380056, + "grad_norm": 0.15438112616539001, + "learning_rate": 4.107473047154982e-05, + "loss": 3.1762, + "step": 11344 + }, + { + "epoch": 0.05691172898295347, + "grad_norm": 0.14683127403259277, + "learning_rate": 4.107266869411381e-05, + "loss": 3.1927, + "step": 11345 + }, + { + "epoch": 0.05694257544852689, + "grad_norm": 0.14706926047801971, + "learning_rate": 4.107060673032487e-05, + "loss": 3.1546, + "step": 11346 + }, + { + "epoch": 0.056973421914100306, + "grad_norm": 0.14745749533176422, + "learning_rate": 4.106854458020689e-05, + "loss": 3.1757, + "step": 11347 + }, + { + "epoch": 0.05700426837967372, + "grad_norm": 0.15202999114990234, + "learning_rate": 4.10664822437838e-05, + "loss": 3.1772, + "step": 11348 + }, + { + "epoch": 0.05703511484524714, + "grad_norm": 0.14710162580013275, + "learning_rate": 4.106441972107951e-05, + "loss": 3.1468, + "step": 11349 + }, + { + "epoch": 0.057065961310820555, + "grad_norm": 0.15225693583488464, + "learning_rate": 4.106235701211792e-05, + "loss": 3.1795, + "step": 11350 + }, + { + "epoch": 0.05709680777639397, + "grad_norm": 0.15374982357025146, + "learning_rate": 4.106029411692295e-05, + "loss": 3.1988, + "step": 11351 + }, + { + "epoch": 0.05712765424196739, + "grad_norm": 0.5071514248847961, + "learning_rate": 4.105823103551853e-05, + "loss": 3.1518, + "step": 11352 + }, + { + "epoch": 0.057158500707540805, + "grad_norm": 0.16586934030056, + "learning_rate": 4.105616776792856e-05, + "loss": 3.1745, + "step": 11353 + }, + { + "epoch": 0.05718934717311422, + "grad_norm": 0.14621363580226898, + "learning_rate": 4.1054104314176976e-05, + "loss": 3.1554, + "step": 11354 + }, + { + "epoch": 0.05722019363868764, + "grad_norm": 0.1530809849500656, + "learning_rate": 4.105204067428771e-05, + "loss": 3.1605, + "step": 11355 + }, + { + "epoch": 0.057251040104261054, + "grad_norm": 0.1512451469898224, + "learning_rate": 4.104997684828467e-05, + "loss": 3.1782, + "step": 11356 + }, + { + "epoch": 0.05728188656983447, + "grad_norm": 0.15001524984836578, + "learning_rate": 4.1047912836191796e-05, + "loss": 3.1737, + "step": 11357 + }, + { + "epoch": 0.05731273303540789, + "grad_norm": 0.14948107302188873, + "learning_rate": 4.104584863803301e-05, + "loss": 3.1888, + "step": 11358 + }, + { + "epoch": 0.057343579500981304, + "grad_norm": 0.15915873646736145, + "learning_rate": 4.104378425383226e-05, + "loss": 3.178, + "step": 11359 + }, + { + "epoch": 0.05737442596655472, + "grad_norm": 0.15471546351909637, + "learning_rate": 4.104171968361347e-05, + "loss": 3.1823, + "step": 11360 + }, + { + "epoch": 0.05740527243212814, + "grad_norm": 0.14894674718379974, + "learning_rate": 4.103965492740057e-05, + "loss": 3.1479, + "step": 11361 + }, + { + "epoch": 0.05743611889770155, + "grad_norm": 0.1583547443151474, + "learning_rate": 4.103758998521752e-05, + "loss": 3.1987, + "step": 11362 + }, + { + "epoch": 0.05746696536327497, + "grad_norm": 0.1446244865655899, + "learning_rate": 4.1035524857088254e-05, + "loss": 3.1384, + "step": 11363 + }, + { + "epoch": 0.05749781182884839, + "grad_norm": 0.16870221495628357, + "learning_rate": 4.103345954303671e-05, + "loss": 3.1789, + "step": 11364 + }, + { + "epoch": 0.0575286582944218, + "grad_norm": 0.14518491923809052, + "learning_rate": 4.1031394043086837e-05, + "loss": 3.1736, + "step": 11365 + }, + { + "epoch": 0.05755950475999522, + "grad_norm": 0.15411826968193054, + "learning_rate": 4.102932835726259e-05, + "loss": 3.1632, + "step": 11366 + }, + { + "epoch": 0.05759035122556864, + "grad_norm": 0.1534607857465744, + "learning_rate": 4.1027262485587904e-05, + "loss": 3.1778, + "step": 11367 + }, + { + "epoch": 0.05762119769114205, + "grad_norm": 0.14918139576911926, + "learning_rate": 4.102519642808674e-05, + "loss": 3.1638, + "step": 11368 + }, + { + "epoch": 0.057652044156715467, + "grad_norm": 0.1554303914308548, + "learning_rate": 4.102313018478306e-05, + "loss": 3.1634, + "step": 11369 + }, + { + "epoch": 0.05768289062228889, + "grad_norm": 0.1449422538280487, + "learning_rate": 4.102106375570081e-05, + "loss": 3.2111, + "step": 11370 + }, + { + "epoch": 0.0577137370878623, + "grad_norm": 0.18268653750419617, + "learning_rate": 4.101899714086396e-05, + "loss": 3.2074, + "step": 11371 + }, + { + "epoch": 0.057744583553435716, + "grad_norm": 0.14829511940479279, + "learning_rate": 4.1016930340296455e-05, + "loss": 3.1925, + "step": 11372 + }, + { + "epoch": 0.05777543001900914, + "grad_norm": 0.16593365371227264, + "learning_rate": 4.101486335402228e-05, + "loss": 3.1589, + "step": 11373 + }, + { + "epoch": 0.05780627648458255, + "grad_norm": 0.20952218770980835, + "learning_rate": 4.101279618206537e-05, + "loss": 3.1468, + "step": 11374 + }, + { + "epoch": 0.057837122950155966, + "grad_norm": 0.7697446346282959, + "learning_rate": 4.101072882444973e-05, + "loss": 3.172, + "step": 11375 + }, + { + "epoch": 0.05786796941572939, + "grad_norm": 0.15970918536186218, + "learning_rate": 4.10086612811993e-05, + "loss": 3.1636, + "step": 11376 + }, + { + "epoch": 0.0578988158813028, + "grad_norm": 0.14900043606758118, + "learning_rate": 4.100659355233807e-05, + "loss": 3.2019, + "step": 11377 + }, + { + "epoch": 0.057929662346876215, + "grad_norm": 0.16122446954250336, + "learning_rate": 4.100452563789e-05, + "loss": 3.1879, + "step": 11378 + }, + { + "epoch": 0.057960508812449636, + "grad_norm": 0.15149517357349396, + "learning_rate": 4.100245753787907e-05, + "loss": 3.1451, + "step": 11379 + }, + { + "epoch": 0.05799135527802305, + "grad_norm": 0.15142783522605896, + "learning_rate": 4.100038925232926e-05, + "loss": 3.1461, + "step": 11380 + }, + { + "epoch": 0.058022201743596465, + "grad_norm": 0.15399111807346344, + "learning_rate": 4.0998320781264564e-05, + "loss": 3.1429, + "step": 11381 + }, + { + "epoch": 0.058053048209169886, + "grad_norm": 0.16039276123046875, + "learning_rate": 4.099625212470895e-05, + "loss": 3.2018, + "step": 11382 + }, + { + "epoch": 0.0580838946747433, + "grad_norm": 0.15595005452632904, + "learning_rate": 4.099418328268639e-05, + "loss": 3.159, + "step": 11383 + }, + { + "epoch": 0.058114741140316714, + "grad_norm": 0.15264897048473358, + "learning_rate": 4.09921142552209e-05, + "loss": 3.1716, + "step": 11384 + }, + { + "epoch": 0.058145587605890135, + "grad_norm": 0.16265295445919037, + "learning_rate": 4.099004504233646e-05, + "loss": 3.151, + "step": 11385 + }, + { + "epoch": 0.05817643407146355, + "grad_norm": 0.15065297484397888, + "learning_rate": 4.098797564405704e-05, + "loss": 3.1863, + "step": 11386 + }, + { + "epoch": 0.058207280537036964, + "grad_norm": 0.15145260095596313, + "learning_rate": 4.098590606040666e-05, + "loss": 3.1936, + "step": 11387 + }, + { + "epoch": 0.058238127002610385, + "grad_norm": 0.15065591037273407, + "learning_rate": 4.09838362914093e-05, + "loss": 3.1735, + "step": 11388 + }, + { + "epoch": 0.0582689734681838, + "grad_norm": 0.156698539853096, + "learning_rate": 4.098176633708898e-05, + "loss": 3.1693, + "step": 11389 + }, + { + "epoch": 0.05829981993375721, + "grad_norm": 0.1507943719625473, + "learning_rate": 4.097969619746967e-05, + "loss": 3.2086, + "step": 11390 + }, + { + "epoch": 0.058330666399330634, + "grad_norm": 0.15533633530139923, + "learning_rate": 4.097762587257539e-05, + "loss": 3.1667, + "step": 11391 + }, + { + "epoch": 0.05836151286490405, + "grad_norm": 0.1487014889717102, + "learning_rate": 4.097555536243013e-05, + "loss": 3.1558, + "step": 11392 + }, + { + "epoch": 0.05839235933047746, + "grad_norm": 0.15089762210845947, + "learning_rate": 4.0973484667057914e-05, + "loss": 3.1785, + "step": 11393 + }, + { + "epoch": 0.058423205796050884, + "grad_norm": 0.15358488261699677, + "learning_rate": 4.097141378648274e-05, + "loss": 3.1855, + "step": 11394 + }, + { + "epoch": 0.0584540522616243, + "grad_norm": 0.15645095705986023, + "learning_rate": 4.096934272072862e-05, + "loss": 3.1988, + "step": 11395 + }, + { + "epoch": 0.05848489872719771, + "grad_norm": 0.1462259292602539, + "learning_rate": 4.096727146981957e-05, + "loss": 3.16, + "step": 11396 + }, + { + "epoch": 0.05851574519277113, + "grad_norm": 0.15930452942848206, + "learning_rate": 4.09652000337796e-05, + "loss": 3.1473, + "step": 11397 + }, + { + "epoch": 0.05854659165834455, + "grad_norm": 0.14798735082149506, + "learning_rate": 4.096312841263273e-05, + "loss": 3.1578, + "step": 11398 + }, + { + "epoch": 0.05857743812391796, + "grad_norm": 0.15708258748054504, + "learning_rate": 4.096105660640298e-05, + "loss": 3.1757, + "step": 11399 + }, + { + "epoch": 0.05860828458949138, + "grad_norm": 0.15208198130130768, + "learning_rate": 4.095898461511437e-05, + "loss": 3.1505, + "step": 11400 + }, + { + "epoch": 0.0586391310550648, + "grad_norm": 0.1614701896905899, + "learning_rate": 4.095691243879092e-05, + "loss": 3.1613, + "step": 11401 + }, + { + "epoch": 0.05866997752063821, + "grad_norm": 0.1527896523475647, + "learning_rate": 4.095484007745667e-05, + "loss": 3.1586, + "step": 11402 + }, + { + "epoch": 0.05870082398621163, + "grad_norm": 0.15269190073013306, + "learning_rate": 4.095276753113563e-05, + "loss": 3.1742, + "step": 11403 + }, + { + "epoch": 0.05873167045178505, + "grad_norm": 0.15494988858699799, + "learning_rate": 4.095069479985183e-05, + "loss": 3.1732, + "step": 11404 + }, + { + "epoch": 0.05876251691735846, + "grad_norm": 0.15432555973529816, + "learning_rate": 4.094862188362932e-05, + "loss": 3.1623, + "step": 11405 + }, + { + "epoch": 0.05879336338293188, + "grad_norm": 0.17096200585365295, + "learning_rate": 4.094654878249211e-05, + "loss": 3.1389, + "step": 11406 + }, + { + "epoch": 0.058824209848505296, + "grad_norm": 0.14930197596549988, + "learning_rate": 4.0944475496464265e-05, + "loss": 3.1945, + "step": 11407 + }, + { + "epoch": 0.05885505631407871, + "grad_norm": 0.16032253205776215, + "learning_rate": 4.0942402025569805e-05, + "loss": 3.1867, + "step": 11408 + }, + { + "epoch": 0.05888590277965213, + "grad_norm": 0.14803312718868256, + "learning_rate": 4.094032836983277e-05, + "loss": 3.1685, + "step": 11409 + }, + { + "epoch": 0.058916749245225546, + "grad_norm": 0.15659120678901672, + "learning_rate": 4.09382545292772e-05, + "loss": 3.1448, + "step": 11410 + }, + { + "epoch": 0.05894759571079896, + "grad_norm": 0.1419268399477005, + "learning_rate": 4.0936180503927156e-05, + "loss": 3.1337, + "step": 11411 + }, + { + "epoch": 0.05897844217637238, + "grad_norm": 0.15584145486354828, + "learning_rate": 4.0934106293806674e-05, + "loss": 3.1661, + "step": 11412 + }, + { + "epoch": 0.059009288641945795, + "grad_norm": 0.1557578444480896, + "learning_rate": 4.09320318989398e-05, + "loss": 3.1406, + "step": 11413 + }, + { + "epoch": 0.05904013510751921, + "grad_norm": 0.15765073895454407, + "learning_rate": 4.0929957319350596e-05, + "loss": 3.1683, + "step": 11414 + }, + { + "epoch": 0.05907098157309263, + "grad_norm": 0.14904166758060455, + "learning_rate": 4.0927882555063105e-05, + "loss": 3.1857, + "step": 11415 + }, + { + "epoch": 0.059101828038666045, + "grad_norm": 0.1558101624250412, + "learning_rate": 4.09258076061014e-05, + "loss": 3.2096, + "step": 11416 + }, + { + "epoch": 0.05913267450423946, + "grad_norm": 0.1475760042667389, + "learning_rate": 4.092373247248951e-05, + "loss": 3.1817, + "step": 11417 + }, + { + "epoch": 0.05916352096981288, + "grad_norm": 0.1546998769044876, + "learning_rate": 4.092165715425151e-05, + "loss": 3.1755, + "step": 11418 + }, + { + "epoch": 0.059194367435386294, + "grad_norm": 0.15528808534145355, + "learning_rate": 4.0919581651411476e-05, + "loss": 3.1611, + "step": 11419 + }, + { + "epoch": 0.05922521390095971, + "grad_norm": 0.15420489013195038, + "learning_rate": 4.0917505963993455e-05, + "loss": 3.1546, + "step": 11420 + }, + { + "epoch": 0.05925606036653313, + "grad_norm": 0.14655347168445587, + "learning_rate": 4.091543009202151e-05, + "loss": 3.1594, + "step": 11421 + }, + { + "epoch": 0.059286906832106544, + "grad_norm": 0.1500406712293625, + "learning_rate": 4.091335403551973e-05, + "loss": 3.1809, + "step": 11422 + }, + { + "epoch": 0.05931775329767996, + "grad_norm": 0.18388038873672485, + "learning_rate": 4.091127779451216e-05, + "loss": 3.1883, + "step": 11423 + }, + { + "epoch": 0.05934859976325338, + "grad_norm": 0.1508694440126419, + "learning_rate": 4.09092013690229e-05, + "loss": 3.1914, + "step": 11424 + }, + { + "epoch": 0.05937944622882679, + "grad_norm": 0.15050269663333893, + "learning_rate": 4.0907124759076e-05, + "loss": 3.1823, + "step": 11425 + }, + { + "epoch": 0.05941029269440021, + "grad_norm": 0.1484415978193283, + "learning_rate": 4.0905047964695545e-05, + "loss": 3.1386, + "step": 11426 + }, + { + "epoch": 0.05944113915997363, + "grad_norm": 0.15323589742183685, + "learning_rate": 4.090297098590562e-05, + "loss": 3.1719, + "step": 11427 + }, + { + "epoch": 0.05947198562554704, + "grad_norm": 0.15962715446949005, + "learning_rate": 4.0900893822730306e-05, + "loss": 3.1631, + "step": 11428 + }, + { + "epoch": 0.05950283209112046, + "grad_norm": 0.1538861095905304, + "learning_rate": 4.089881647519368e-05, + "loss": 3.1967, + "step": 11429 + }, + { + "epoch": 0.05953367855669388, + "grad_norm": 0.1549607217311859, + "learning_rate": 4.089673894331983e-05, + "loss": 3.1624, + "step": 11430 + }, + { + "epoch": 0.05956452502226729, + "grad_norm": 0.15102042257785797, + "learning_rate": 4.089466122713286e-05, + "loss": 3.2083, + "step": 11431 + }, + { + "epoch": 0.059595371487840706, + "grad_norm": 0.15019695460796356, + "learning_rate": 4.089258332665682e-05, + "loss": 3.139, + "step": 11432 + }, + { + "epoch": 0.05962621795341413, + "grad_norm": 0.1468128263950348, + "learning_rate": 4.089050524191585e-05, + "loss": 3.1375, + "step": 11433 + }, + { + "epoch": 0.05965706441898754, + "grad_norm": 0.15548236668109894, + "learning_rate": 4.0888426972934004e-05, + "loss": 3.1662, + "step": 11434 + }, + { + "epoch": 0.059687910884560956, + "grad_norm": 0.15353117883205414, + "learning_rate": 4.08863485197354e-05, + "loss": 3.1837, + "step": 11435 + }, + { + "epoch": 0.05971875735013438, + "grad_norm": 0.14749611914157867, + "learning_rate": 4.088426988234413e-05, + "loss": 3.1745, + "step": 11436 + }, + { + "epoch": 0.05974960381570779, + "grad_norm": 0.15487752854824066, + "learning_rate": 4.0882191060784306e-05, + "loss": 3.2135, + "step": 11437 + }, + { + "epoch": 0.059780450281281206, + "grad_norm": 0.14225752651691437, + "learning_rate": 4.088011205508001e-05, + "loss": 3.1789, + "step": 11438 + }, + { + "epoch": 0.05981129674685463, + "grad_norm": 0.15330356359481812, + "learning_rate": 4.087803286525536e-05, + "loss": 3.1942, + "step": 11439 + }, + { + "epoch": 0.05984214321242804, + "grad_norm": 0.1470959186553955, + "learning_rate": 4.087595349133446e-05, + "loss": 3.1547, + "step": 11440 + }, + { + "epoch": 0.059872989678001455, + "grad_norm": 0.1465694010257721, + "learning_rate": 4.0873873933341425e-05, + "loss": 3.1677, + "step": 11441 + }, + { + "epoch": 0.059903836143574876, + "grad_norm": 0.14224611222743988, + "learning_rate": 4.087179419130035e-05, + "loss": 3.1845, + "step": 11442 + }, + { + "epoch": 0.05993468260914829, + "grad_norm": 0.14809437096118927, + "learning_rate": 4.086971426523537e-05, + "loss": 3.1442, + "step": 11443 + }, + { + "epoch": 0.059965529074721705, + "grad_norm": 0.1473465859889984, + "learning_rate": 4.086763415517059e-05, + "loss": 3.1866, + "step": 11444 + }, + { + "epoch": 0.059996375540295126, + "grad_norm": 0.1472526490688324, + "learning_rate": 4.086555386113011e-05, + "loss": 3.1732, + "step": 11445 + }, + { + "epoch": 0.06002722200586854, + "grad_norm": 0.1496272087097168, + "learning_rate": 4.0863473383138087e-05, + "loss": 3.1866, + "step": 11446 + }, + { + "epoch": 0.060058068471441954, + "grad_norm": 0.14836910367012024, + "learning_rate": 4.086139272121862e-05, + "loss": 3.1904, + "step": 11447 + }, + { + "epoch": 0.060088914937015375, + "grad_norm": 0.1438712179660797, + "learning_rate": 4.0859311875395826e-05, + "loss": 3.1741, + "step": 11448 + }, + { + "epoch": 0.06011976140258879, + "grad_norm": 0.1546114832162857, + "learning_rate": 4.085723084569385e-05, + "loss": 3.1899, + "step": 11449 + }, + { + "epoch": 0.060150607868162204, + "grad_norm": 0.1472843438386917, + "learning_rate": 4.08551496321368e-05, + "loss": 3.2049, + "step": 11450 + }, + { + "epoch": 0.060181454333735625, + "grad_norm": 0.14826852083206177, + "learning_rate": 4.0853068234748835e-05, + "loss": 3.178, + "step": 11451 + }, + { + "epoch": 0.06021230079930904, + "grad_norm": 0.14812706410884857, + "learning_rate": 4.0850986653554055e-05, + "loss": 3.1813, + "step": 11452 + }, + { + "epoch": 0.06024314726488245, + "grad_norm": 0.14514589309692383, + "learning_rate": 4.0848904888576626e-05, + "loss": 3.1722, + "step": 11453 + }, + { + "epoch": 0.060273993730455874, + "grad_norm": 0.1456688940525055, + "learning_rate": 4.084682293984066e-05, + "loss": 3.2221, + "step": 11454 + }, + { + "epoch": 0.06030484019602929, + "grad_norm": 0.14750058948993683, + "learning_rate": 4.08447408073703e-05, + "loss": 3.1644, + "step": 11455 + }, + { + "epoch": 0.0603356866616027, + "grad_norm": 0.14764176309108734, + "learning_rate": 4.0842658491189696e-05, + "loss": 3.1832, + "step": 11456 + }, + { + "epoch": 0.060366533127176124, + "grad_norm": 0.1562431901693344, + "learning_rate": 4.0840575991322985e-05, + "loss": 3.1689, + "step": 11457 + }, + { + "epoch": 0.06039737959274954, + "grad_norm": 0.1605505794286728, + "learning_rate": 4.0838493307794324e-05, + "loss": 3.1854, + "step": 11458 + }, + { + "epoch": 0.06042822605832295, + "grad_norm": 0.14617453515529633, + "learning_rate": 4.0836410440627835e-05, + "loss": 3.1641, + "step": 11459 + }, + { + "epoch": 0.06045907252389637, + "grad_norm": 0.16960017383098602, + "learning_rate": 4.083432738984769e-05, + "loss": 3.1743, + "step": 11460 + }, + { + "epoch": 0.06048991898946979, + "grad_norm": 0.17066410183906555, + "learning_rate": 4.083224415547804e-05, + "loss": 3.1879, + "step": 11461 + }, + { + "epoch": 0.0605207654550432, + "grad_norm": 0.16434553265571594, + "learning_rate": 4.0830160737543034e-05, + "loss": 3.1464, + "step": 11462 + }, + { + "epoch": 0.06055161192061662, + "grad_norm": 0.15913742780685425, + "learning_rate": 4.0828077136066824e-05, + "loss": 3.143, + "step": 11463 + }, + { + "epoch": 0.06058245838619004, + "grad_norm": 0.15745389461517334, + "learning_rate": 4.082599335107357e-05, + "loss": 3.1801, + "step": 11464 + }, + { + "epoch": 0.06061330485176345, + "grad_norm": 0.16611707210540771, + "learning_rate": 4.082390938258743e-05, + "loss": 3.1503, + "step": 11465 + }, + { + "epoch": 0.06064415131733687, + "grad_norm": 0.14812880754470825, + "learning_rate": 4.082182523063257e-05, + "loss": 3.1302, + "step": 11466 + }, + { + "epoch": 0.06067499778291029, + "grad_norm": 0.16124217212200165, + "learning_rate": 4.081974089523316e-05, + "loss": 3.1767, + "step": 11467 + }, + { + "epoch": 0.0607058442484837, + "grad_norm": 0.16519366204738617, + "learning_rate": 4.081765637641336e-05, + "loss": 3.2027, + "step": 11468 + }, + { + "epoch": 0.06073669071405712, + "grad_norm": 0.1541912704706192, + "learning_rate": 4.081557167419734e-05, + "loss": 3.2062, + "step": 11469 + }, + { + "epoch": 0.060767537179630536, + "grad_norm": 0.15642201900482178, + "learning_rate": 4.081348678860926e-05, + "loss": 3.1595, + "step": 11470 + }, + { + "epoch": 0.06079838364520395, + "grad_norm": 0.14572159945964813, + "learning_rate": 4.0811401719673314e-05, + "loss": 3.1983, + "step": 11471 + }, + { + "epoch": 0.06082923011077737, + "grad_norm": 0.15831038355827332, + "learning_rate": 4.080931646741367e-05, + "loss": 3.1779, + "step": 11472 + }, + { + "epoch": 0.060860076576350786, + "grad_norm": 0.5133817791938782, + "learning_rate": 4.080723103185449e-05, + "loss": 3.1338, + "step": 11473 + }, + { + "epoch": 0.0608909230419242, + "grad_norm": 0.148798868060112, + "learning_rate": 4.080514541301998e-05, + "loss": 3.169, + "step": 11474 + }, + { + "epoch": 0.06092176950749762, + "grad_norm": 0.1527075469493866, + "learning_rate": 4.0803059610934294e-05, + "loss": 3.1487, + "step": 11475 + }, + { + "epoch": 0.060952615973071035, + "grad_norm": 0.16381141543388367, + "learning_rate": 4.080097362562163e-05, + "loss": 3.195, + "step": 11476 + }, + { + "epoch": 0.06098346243864445, + "grad_norm": 0.15208296477794647, + "learning_rate": 4.079888745710617e-05, + "loss": 3.1537, + "step": 11477 + }, + { + "epoch": 0.06101430890421787, + "grad_norm": 0.15943194925785065, + "learning_rate": 4.079680110541212e-05, + "loss": 3.1649, + "step": 11478 + }, + { + "epoch": 0.061045155369791285, + "grad_norm": 0.14640262722969055, + "learning_rate": 4.079471457056364e-05, + "loss": 3.0958, + "step": 11479 + }, + { + "epoch": 0.0610760018353647, + "grad_norm": 0.14800383150577545, + "learning_rate": 4.079262785258494e-05, + "loss": 3.1819, + "step": 11480 + }, + { + "epoch": 0.06110684830093812, + "grad_norm": 0.16663266718387604, + "learning_rate": 4.079054095150021e-05, + "loss": 3.1787, + "step": 11481 + }, + { + "epoch": 0.061137694766511534, + "grad_norm": 0.1504441201686859, + "learning_rate": 4.078845386733364e-05, + "loss": 3.1806, + "step": 11482 + }, + { + "epoch": 0.06116854123208495, + "grad_norm": 0.15599939227104187, + "learning_rate": 4.078636660010944e-05, + "loss": 3.1852, + "step": 11483 + }, + { + "epoch": 0.06119938769765837, + "grad_norm": 0.17132878303527832, + "learning_rate": 4.0784279149851815e-05, + "loss": 3.171, + "step": 11484 + }, + { + "epoch": 0.061230234163231784, + "grad_norm": 0.16284626722335815, + "learning_rate": 4.078219151658494e-05, + "loss": 3.1697, + "step": 11485 + }, + { + "epoch": 0.0612610806288052, + "grad_norm": 0.168400377035141, + "learning_rate": 4.078010370033305e-05, + "loss": 3.163, + "step": 11486 + }, + { + "epoch": 0.06129192709437862, + "grad_norm": 0.1541835516691208, + "learning_rate": 4.0778015701120344e-05, + "loss": 3.1968, + "step": 11487 + }, + { + "epoch": 0.06132277355995203, + "grad_norm": 0.15140876173973083, + "learning_rate": 4.077592751897101e-05, + "loss": 3.1663, + "step": 11488 + }, + { + "epoch": 0.06135362002552545, + "grad_norm": 0.15404462814331055, + "learning_rate": 4.07738391539093e-05, + "loss": 3.1472, + "step": 11489 + }, + { + "epoch": 0.06138446649109887, + "grad_norm": 0.14840026199817657, + "learning_rate": 4.077175060595939e-05, + "loss": 3.1571, + "step": 11490 + }, + { + "epoch": 0.06141531295667228, + "grad_norm": 0.1545998752117157, + "learning_rate": 4.0769661875145514e-05, + "loss": 3.1636, + "step": 11491 + }, + { + "epoch": 0.0614461594222457, + "grad_norm": 0.1436830461025238, + "learning_rate": 4.076757296149188e-05, + "loss": 3.1463, + "step": 11492 + }, + { + "epoch": 0.06147700588781912, + "grad_norm": 0.15075498819351196, + "learning_rate": 4.0765483865022716e-05, + "loss": 3.1645, + "step": 11493 + }, + { + "epoch": 0.06150785235339253, + "grad_norm": 0.14827483892440796, + "learning_rate": 4.076339458576224e-05, + "loss": 3.1944, + "step": 11494 + }, + { + "epoch": 0.061538698818965946, + "grad_norm": 0.1488434374332428, + "learning_rate": 4.076130512373468e-05, + "loss": 3.1356, + "step": 11495 + }, + { + "epoch": 0.06156954528453937, + "grad_norm": 0.1544807404279709, + "learning_rate": 4.0759215478964255e-05, + "loss": 3.1529, + "step": 11496 + }, + { + "epoch": 0.06160039175011278, + "grad_norm": 0.14630401134490967, + "learning_rate": 4.0757125651475196e-05, + "loss": 3.1618, + "step": 11497 + }, + { + "epoch": 0.061631238215686196, + "grad_norm": 0.19059012830257416, + "learning_rate": 4.075503564129174e-05, + "loss": 3.2166, + "step": 11498 + }, + { + "epoch": 0.06166208468125962, + "grad_norm": 0.14968648552894592, + "learning_rate": 4.0752945448438105e-05, + "loss": 3.1755, + "step": 11499 + }, + { + "epoch": 0.06169293114683303, + "grad_norm": 0.15082043409347534, + "learning_rate": 4.0750855072938536e-05, + "loss": 3.1377, + "step": 11500 + }, + { + "epoch": 0.061723777612406446, + "grad_norm": 0.21304459869861603, + "learning_rate": 4.0748764514817274e-05, + "loss": 3.2021, + "step": 11501 + }, + { + "epoch": 0.06175462407797987, + "grad_norm": 0.14819350838661194, + "learning_rate": 4.0746673774098545e-05, + "loss": 3.1972, + "step": 11502 + }, + { + "epoch": 0.06178547054355328, + "grad_norm": 0.15111114084720612, + "learning_rate": 4.07445828508066e-05, + "loss": 3.1717, + "step": 11503 + }, + { + "epoch": 0.061816317009126695, + "grad_norm": 0.15132620930671692, + "learning_rate": 4.0742491744965674e-05, + "loss": 3.1677, + "step": 11504 + }, + { + "epoch": 0.061847163474700116, + "grad_norm": 0.15626421570777893, + "learning_rate": 4.074040045660002e-05, + "loss": 3.18, + "step": 11505 + }, + { + "epoch": 0.06187800994027353, + "grad_norm": 0.14558321237564087, + "learning_rate": 4.073830898573387e-05, + "loss": 3.1398, + "step": 11506 + }, + { + "epoch": 0.061908856405846945, + "grad_norm": 0.1548008769750595, + "learning_rate": 4.07362173323915e-05, + "loss": 3.1807, + "step": 11507 + }, + { + "epoch": 0.061939702871420366, + "grad_norm": 0.15951664745807648, + "learning_rate": 4.073412549659715e-05, + "loss": 3.1531, + "step": 11508 + }, + { + "epoch": 0.06197054933699378, + "grad_norm": 0.15591643750667572, + "learning_rate": 4.0732033478375056e-05, + "loss": 3.2108, + "step": 11509 + }, + { + "epoch": 0.062001395802567194, + "grad_norm": 0.14976684749126434, + "learning_rate": 4.072994127774949e-05, + "loss": 3.1891, + "step": 11510 + }, + { + "epoch": 0.062032242268140615, + "grad_norm": 0.15353800356388092, + "learning_rate": 4.072784889474471e-05, + "loss": 3.193, + "step": 11511 + }, + { + "epoch": 0.06206308873371403, + "grad_norm": 0.1487264335155487, + "learning_rate": 4.072575632938498e-05, + "loss": 3.1824, + "step": 11512 + }, + { + "epoch": 0.062093935199287444, + "grad_norm": 0.14636501669883728, + "learning_rate": 4.072366358169455e-05, + "loss": 3.1909, + "step": 11513 + }, + { + "epoch": 0.062124781664860865, + "grad_norm": 0.1536000519990921, + "learning_rate": 4.072157065169768e-05, + "loss": 3.1645, + "step": 11514 + }, + { + "epoch": 0.06215562813043428, + "grad_norm": 0.1439344584941864, + "learning_rate": 4.071947753941866e-05, + "loss": 3.1706, + "step": 11515 + }, + { + "epoch": 0.06218647459600769, + "grad_norm": 0.1523779183626175, + "learning_rate": 4.071738424488174e-05, + "loss": 3.1634, + "step": 11516 + }, + { + "epoch": 0.062217321061581114, + "grad_norm": 0.9157229065895081, + "learning_rate": 4.0715290768111193e-05, + "loss": 3.2059, + "step": 11517 + }, + { + "epoch": 0.06224816752715453, + "grad_norm": 0.14693118631839752, + "learning_rate": 4.07131971091313e-05, + "loss": 3.1875, + "step": 11518 + }, + { + "epoch": 0.06227901399272794, + "grad_norm": 0.15524671971797943, + "learning_rate": 4.071110326796632e-05, + "loss": 3.186, + "step": 11519 + }, + { + "epoch": 0.062309860458301364, + "grad_norm": 0.1521158218383789, + "learning_rate": 4.070900924464054e-05, + "loss": 3.143, + "step": 11520 + }, + { + "epoch": 0.06234070692387478, + "grad_norm": 0.14877739548683167, + "learning_rate": 4.070691503917824e-05, + "loss": 3.1597, + "step": 11521 + }, + { + "epoch": 0.06237155338944819, + "grad_norm": 0.16409531235694885, + "learning_rate": 4.07048206516037e-05, + "loss": 3.1726, + "step": 11522 + }, + { + "epoch": 0.06240239985502161, + "grad_norm": 0.15673910081386566, + "learning_rate": 4.07027260819412e-05, + "loss": 3.15, + "step": 11523 + }, + { + "epoch": 0.06243324632059503, + "grad_norm": 0.15023505687713623, + "learning_rate": 4.070063133021503e-05, + "loss": 3.1715, + "step": 11524 + }, + { + "epoch": 0.06246409278616844, + "grad_norm": 0.16342219710350037, + "learning_rate": 4.069853639644947e-05, + "loss": 3.1949, + "step": 11525 + }, + { + "epoch": 0.06249493925174186, + "grad_norm": 0.1531844437122345, + "learning_rate": 4.069644128066882e-05, + "loss": 3.2356, + "step": 11526 + }, + { + "epoch": 0.06252578571731528, + "grad_norm": 0.17718270421028137, + "learning_rate": 4.069434598289736e-05, + "loss": 3.1828, + "step": 11527 + }, + { + "epoch": 0.06255663218288869, + "grad_norm": 0.15465442836284637, + "learning_rate": 4.069225050315939e-05, + "loss": 3.2055, + "step": 11528 + }, + { + "epoch": 0.0625874786484621, + "grad_norm": 0.14768192172050476, + "learning_rate": 4.069015484147921e-05, + "loss": 3.1542, + "step": 11529 + }, + { + "epoch": 0.06261832511403553, + "grad_norm": 0.15921516716480255, + "learning_rate": 4.068805899788111e-05, + "loss": 3.1586, + "step": 11530 + }, + { + "epoch": 0.06264917157960895, + "grad_norm": 0.1546725481748581, + "learning_rate": 4.06859629723894e-05, + "loss": 3.1629, + "step": 11531 + }, + { + "epoch": 0.06268001804518236, + "grad_norm": 0.15371203422546387, + "learning_rate": 4.0683866765028366e-05, + "loss": 3.1977, + "step": 11532 + }, + { + "epoch": 0.06271086451075578, + "grad_norm": 0.14443327486515045, + "learning_rate": 4.0681770375822324e-05, + "loss": 3.1557, + "step": 11533 + }, + { + "epoch": 0.06274171097632919, + "grad_norm": 0.15225434303283691, + "learning_rate": 4.0679673804795584e-05, + "loss": 3.1612, + "step": 11534 + }, + { + "epoch": 0.0627725574419026, + "grad_norm": 0.15437054634094238, + "learning_rate": 4.067757705197244e-05, + "loss": 3.1564, + "step": 11535 + }, + { + "epoch": 0.06280340390747603, + "grad_norm": 0.15082091093063354, + "learning_rate": 4.067548011737722e-05, + "loss": 3.2197, + "step": 11536 + }, + { + "epoch": 0.06283425037304945, + "grad_norm": 0.1463136076927185, + "learning_rate": 4.067338300103422e-05, + "loss": 3.1663, + "step": 11537 + }, + { + "epoch": 0.06286509683862286, + "grad_norm": 0.15112341940402985, + "learning_rate": 4.0671285702967765e-05, + "loss": 3.1819, + "step": 11538 + }, + { + "epoch": 0.06289594330419628, + "grad_norm": 0.1435430347919464, + "learning_rate": 4.0669188223202174e-05, + "loss": 3.1186, + "step": 11539 + }, + { + "epoch": 0.06292678976976969, + "grad_norm": 0.15769998729228973, + "learning_rate": 4.066709056176175e-05, + "loss": 3.1749, + "step": 11540 + }, + { + "epoch": 0.0629576362353431, + "grad_norm": 0.14720971882343292, + "learning_rate": 4.066499271867084e-05, + "loss": 3.1561, + "step": 11541 + }, + { + "epoch": 0.06298848270091653, + "grad_norm": 0.14966793358325958, + "learning_rate": 4.066289469395375e-05, + "loss": 3.1806, + "step": 11542 + }, + { + "epoch": 0.06301932916648995, + "grad_norm": 0.14443618059158325, + "learning_rate": 4.0660796487634805e-05, + "loss": 3.1797, + "step": 11543 + }, + { + "epoch": 0.06305017563206336, + "grad_norm": 0.14664511382579803, + "learning_rate": 4.065869809973833e-05, + "loss": 3.136, + "step": 11544 + }, + { + "epoch": 0.06308102209763677, + "grad_norm": 0.14449195563793182, + "learning_rate": 4.065659953028867e-05, + "loss": 3.179, + "step": 11545 + }, + { + "epoch": 0.06311186856321019, + "grad_norm": 0.15222321450710297, + "learning_rate": 4.0654500779310144e-05, + "loss": 3.1578, + "step": 11546 + }, + { + "epoch": 0.0631427150287836, + "grad_norm": 0.14588072896003723, + "learning_rate": 4.065240184682709e-05, + "loss": 3.1754, + "step": 11547 + }, + { + "epoch": 0.06317356149435703, + "grad_norm": 0.14792807400226593, + "learning_rate": 4.065030273286384e-05, + "loss": 3.1835, + "step": 11548 + }, + { + "epoch": 0.06320440795993044, + "grad_norm": 0.1462620347738266, + "learning_rate": 4.064820343744474e-05, + "loss": 3.1807, + "step": 11549 + }, + { + "epoch": 0.06323525442550386, + "grad_norm": 0.15994806587696075, + "learning_rate": 4.064610396059412e-05, + "loss": 3.1973, + "step": 11550 + }, + { + "epoch": 0.06326610089107727, + "grad_norm": 0.15459182858467102, + "learning_rate": 4.0644004302336325e-05, + "loss": 3.1691, + "step": 11551 + }, + { + "epoch": 0.06329694735665069, + "grad_norm": 0.14809776842594147, + "learning_rate": 4.0641904462695704e-05, + "loss": 3.1783, + "step": 11552 + }, + { + "epoch": 0.0633277938222241, + "grad_norm": 0.14879176020622253, + "learning_rate": 4.06398044416966e-05, + "loss": 3.1638, + "step": 11553 + }, + { + "epoch": 0.06335864028779753, + "grad_norm": 0.16999073326587677, + "learning_rate": 4.063770423936336e-05, + "loss": 3.2057, + "step": 11554 + }, + { + "epoch": 0.06338948675337094, + "grad_norm": 0.14472422003746033, + "learning_rate": 4.063560385572034e-05, + "loss": 3.1787, + "step": 11555 + }, + { + "epoch": 0.06342033321894436, + "grad_norm": 0.14682620763778687, + "learning_rate": 4.0633503290791884e-05, + "loss": 3.1867, + "step": 11556 + }, + { + "epoch": 0.06345117968451777, + "grad_norm": 0.18183033168315887, + "learning_rate": 4.063140254460236e-05, + "loss": 3.1805, + "step": 11557 + }, + { + "epoch": 0.06348202615009119, + "grad_norm": 0.14608852565288544, + "learning_rate": 4.062930161717611e-05, + "loss": 3.1513, + "step": 11558 + }, + { + "epoch": 0.0635128726156646, + "grad_norm": 0.15930654108524323, + "learning_rate": 4.06272005085375e-05, + "loss": 3.1688, + "step": 11559 + }, + { + "epoch": 0.06354371908123803, + "grad_norm": 0.15116962790489197, + "learning_rate": 4.0625099218710896e-05, + "loss": 3.1372, + "step": 11560 + }, + { + "epoch": 0.06357456554681144, + "grad_norm": 0.14622612297534943, + "learning_rate": 4.062299774772065e-05, + "loss": 3.1538, + "step": 11561 + }, + { + "epoch": 0.06360541201238486, + "grad_norm": 0.1477515995502472, + "learning_rate": 4.062089609559113e-05, + "loss": 3.1601, + "step": 11562 + }, + { + "epoch": 0.06363625847795827, + "grad_norm": 0.13730941712856293, + "learning_rate": 4.061879426234672e-05, + "loss": 3.1553, + "step": 11563 + }, + { + "epoch": 0.06366710494353169, + "grad_norm": 0.15117569267749786, + "learning_rate": 4.061669224801177e-05, + "loss": 3.1501, + "step": 11564 + }, + { + "epoch": 0.0636979514091051, + "grad_norm": 0.14416560530662537, + "learning_rate": 4.0614590052610655e-05, + "loss": 3.1546, + "step": 11565 + }, + { + "epoch": 0.06372879787467853, + "grad_norm": 0.14769743382930756, + "learning_rate": 4.0612487676167756e-05, + "loss": 3.1822, + "step": 11566 + }, + { + "epoch": 0.06375964434025194, + "grad_norm": 0.14146751165390015, + "learning_rate": 4.061038511870744e-05, + "loss": 3.171, + "step": 11567 + }, + { + "epoch": 0.06379049080582536, + "grad_norm": 0.14824728667736053, + "learning_rate": 4.060828238025409e-05, + "loss": 3.1939, + "step": 11568 + }, + { + "epoch": 0.06382133727139877, + "grad_norm": 0.1440262347459793, + "learning_rate": 4.06061794608321e-05, + "loss": 3.1637, + "step": 11569 + }, + { + "epoch": 0.06385218373697218, + "grad_norm": 0.15315034985542297, + "learning_rate": 4.0604076360465815e-05, + "loss": 3.1563, + "step": 11570 + }, + { + "epoch": 0.0638830302025456, + "grad_norm": 0.1475367397069931, + "learning_rate": 4.060197307917966e-05, + "loss": 3.1826, + "step": 11571 + }, + { + "epoch": 0.06391387666811903, + "grad_norm": 0.15132726728916168, + "learning_rate": 4.059986961699799e-05, + "loss": 3.1787, + "step": 11572 + }, + { + "epoch": 0.06394472313369244, + "grad_norm": 0.1500992625951767, + "learning_rate": 4.0597765973945205e-05, + "loss": 3.1547, + "step": 11573 + }, + { + "epoch": 0.06397556959926586, + "grad_norm": 0.14875267446041107, + "learning_rate": 4.0595662150045704e-05, + "loss": 3.1328, + "step": 11574 + }, + { + "epoch": 0.06400641606483927, + "grad_norm": 0.15408477187156677, + "learning_rate": 4.059355814532387e-05, + "loss": 3.1872, + "step": 11575 + }, + { + "epoch": 0.06403726253041268, + "grad_norm": 0.1438366323709488, + "learning_rate": 4.05914539598041e-05, + "loss": 3.1939, + "step": 11576 + }, + { + "epoch": 0.0640681089959861, + "grad_norm": 0.15628236532211304, + "learning_rate": 4.058934959351079e-05, + "loss": 3.1943, + "step": 11577 + }, + { + "epoch": 0.06409895546155953, + "grad_norm": 0.15028390288352966, + "learning_rate": 4.058724504646834e-05, + "loss": 3.1473, + "step": 11578 + }, + { + "epoch": 0.06412980192713294, + "grad_norm": 0.14830999076366425, + "learning_rate": 4.0585140318701156e-05, + "loss": 3.171, + "step": 11579 + }, + { + "epoch": 0.06416064839270635, + "grad_norm": 0.15089017152786255, + "learning_rate": 4.0583035410233625e-05, + "loss": 3.1922, + "step": 11580 + }, + { + "epoch": 0.06419149485827977, + "grad_norm": 0.14765942096710205, + "learning_rate": 4.058093032109017e-05, + "loss": 3.1888, + "step": 11581 + }, + { + "epoch": 0.06422234132385318, + "grad_norm": 0.1491975337266922, + "learning_rate": 4.057882505129519e-05, + "loss": 3.1497, + "step": 11582 + }, + { + "epoch": 0.0642531877894266, + "grad_norm": 0.1521112322807312, + "learning_rate": 4.057671960087309e-05, + "loss": 3.1286, + "step": 11583 + }, + { + "epoch": 0.06428403425500002, + "grad_norm": 0.14686505496501923, + "learning_rate": 4.05746139698483e-05, + "loss": 3.1756, + "step": 11584 + }, + { + "epoch": 0.06431488072057344, + "grad_norm": 0.14470726251602173, + "learning_rate": 4.05725081582452e-05, + "loss": 3.1814, + "step": 11585 + }, + { + "epoch": 0.06434572718614685, + "grad_norm": 0.14739060401916504, + "learning_rate": 4.0570402166088246e-05, + "loss": 3.1369, + "step": 11586 + }, + { + "epoch": 0.06437657365172027, + "grad_norm": 0.1431828737258911, + "learning_rate": 4.056829599340183e-05, + "loss": 3.1602, + "step": 11587 + }, + { + "epoch": 0.06440742011729368, + "grad_norm": 0.15220828354358673, + "learning_rate": 4.0566189640210366e-05, + "loss": 3.1722, + "step": 11588 + }, + { + "epoch": 0.0644382665828671, + "grad_norm": 0.14569035172462463, + "learning_rate": 4.05640831065383e-05, + "loss": 3.1587, + "step": 11589 + }, + { + "epoch": 0.06446911304844052, + "grad_norm": 0.1470562368631363, + "learning_rate": 4.056197639241004e-05, + "loss": 3.2001, + "step": 11590 + }, + { + "epoch": 0.06449995951401394, + "grad_norm": 0.14617189764976501, + "learning_rate": 4.055986949785001e-05, + "loss": 3.1355, + "step": 11591 + }, + { + "epoch": 0.06453080597958735, + "grad_norm": 0.1463479846715927, + "learning_rate": 4.055776242288265e-05, + "loss": 3.1708, + "step": 11592 + }, + { + "epoch": 0.06456165244516077, + "grad_norm": 0.15240859985351562, + "learning_rate": 4.055565516753238e-05, + "loss": 3.1613, + "step": 11593 + }, + { + "epoch": 0.06459249891073418, + "grad_norm": 0.14498764276504517, + "learning_rate": 4.055354773182364e-05, + "loss": 3.1859, + "step": 11594 + }, + { + "epoch": 0.0646233453763076, + "grad_norm": 0.14434337615966797, + "learning_rate": 4.055144011578086e-05, + "loss": 3.1549, + "step": 11595 + }, + { + "epoch": 0.06465419184188102, + "grad_norm": 0.15395241975784302, + "learning_rate": 4.054933231942848e-05, + "loss": 3.1791, + "step": 11596 + }, + { + "epoch": 0.06468503830745444, + "grad_norm": 0.15445871651172638, + "learning_rate": 4.054722434279092e-05, + "loss": 3.1552, + "step": 11597 + }, + { + "epoch": 0.06471588477302785, + "grad_norm": 0.14354734122753143, + "learning_rate": 4.054511618589265e-05, + "loss": 3.1631, + "step": 11598 + }, + { + "epoch": 0.06474673123860127, + "grad_norm": 0.15145458281040192, + "learning_rate": 4.054300784875811e-05, + "loss": 3.2014, + "step": 11599 + }, + { + "epoch": 0.06477757770417468, + "grad_norm": 0.14881856739521027, + "learning_rate": 4.054089933141171e-05, + "loss": 3.1926, + "step": 11600 + }, + { + "epoch": 0.0648084241697481, + "grad_norm": 0.14152903854846954, + "learning_rate": 4.053879063387794e-05, + "loss": 3.1765, + "step": 11601 + }, + { + "epoch": 0.06483927063532152, + "grad_norm": 0.14109928905963898, + "learning_rate": 4.053668175618122e-05, + "loss": 3.1681, + "step": 11602 + }, + { + "epoch": 0.06487011710089494, + "grad_norm": 0.1487313061952591, + "learning_rate": 4.053457269834602e-05, + "loss": 3.1696, + "step": 11603 + }, + { + "epoch": 0.06490096356646835, + "grad_norm": 0.14749738574028015, + "learning_rate": 4.0532463460396773e-05, + "loss": 3.1742, + "step": 11604 + }, + { + "epoch": 0.06493181003204176, + "grad_norm": 0.144117072224617, + "learning_rate": 4.053035404235795e-05, + "loss": 3.185, + "step": 11605 + }, + { + "epoch": 0.06496265649761518, + "grad_norm": 0.1608632653951645, + "learning_rate": 4.052824444425402e-05, + "loss": 3.1492, + "step": 11606 + }, + { + "epoch": 0.06499350296318859, + "grad_norm": 0.18826518952846527, + "learning_rate": 4.0526134666109404e-05, + "loss": 3.146, + "step": 11607 + }, + { + "epoch": 0.06502434942876202, + "grad_norm": 0.15177324414253235, + "learning_rate": 4.0524024707948605e-05, + "loss": 3.1682, + "step": 11608 + }, + { + "epoch": 0.06505519589433544, + "grad_norm": 0.1469840109348297, + "learning_rate": 4.052191456979605e-05, + "loss": 3.1796, + "step": 11609 + }, + { + "epoch": 0.06508604235990885, + "grad_norm": 0.14926715195178986, + "learning_rate": 4.051980425167624e-05, + "loss": 3.1647, + "step": 11610 + }, + { + "epoch": 0.06511688882548226, + "grad_norm": 0.1483141928911209, + "learning_rate": 4.051769375361362e-05, + "loss": 3.1715, + "step": 11611 + }, + { + "epoch": 0.06514773529105568, + "grad_norm": 0.15065351128578186, + "learning_rate": 4.0515583075632665e-05, + "loss": 3.1315, + "step": 11612 + }, + { + "epoch": 0.06517858175662909, + "grad_norm": 0.14643889665603638, + "learning_rate": 4.051347221775785e-05, + "loss": 3.1565, + "step": 11613 + }, + { + "epoch": 0.06520942822220252, + "grad_norm": 0.1567612737417221, + "learning_rate": 4.051136118001364e-05, + "loss": 3.1778, + "step": 11614 + }, + { + "epoch": 0.06524027468777593, + "grad_norm": 0.1526242047548294, + "learning_rate": 4.050924996242452e-05, + "loss": 3.1324, + "step": 11615 + }, + { + "epoch": 0.06527112115334935, + "grad_norm": 0.1459278166294098, + "learning_rate": 4.050713856501498e-05, + "loss": 3.1732, + "step": 11616 + }, + { + "epoch": 0.06530196761892276, + "grad_norm": 0.14741097390651703, + "learning_rate": 4.0505026987809475e-05, + "loss": 3.2005, + "step": 11617 + }, + { + "epoch": 0.06533281408449618, + "grad_norm": 0.15635833144187927, + "learning_rate": 4.0502915230832506e-05, + "loss": 3.1785, + "step": 11618 + }, + { + "epoch": 0.06536366055006959, + "grad_norm": 0.14454200863838196, + "learning_rate": 4.050080329410855e-05, + "loss": 3.1442, + "step": 11619 + }, + { + "epoch": 0.06539450701564302, + "grad_norm": 0.1506233662366867, + "learning_rate": 4.0498691177662085e-05, + "loss": 3.1572, + "step": 11620 + }, + { + "epoch": 0.06542535348121643, + "grad_norm": 0.1448906511068344, + "learning_rate": 4.049657888151762e-05, + "loss": 3.1662, + "step": 11621 + }, + { + "epoch": 0.06545619994678985, + "grad_norm": 0.14172910153865814, + "learning_rate": 4.049446640569963e-05, + "loss": 3.1418, + "step": 11622 + }, + { + "epoch": 0.06548704641236326, + "grad_norm": 0.15178221464157104, + "learning_rate": 4.0492353750232616e-05, + "loss": 3.1254, + "step": 11623 + }, + { + "epoch": 0.06551789287793668, + "grad_norm": 0.16243548691272736, + "learning_rate": 4.049024091514107e-05, + "loss": 3.1336, + "step": 11624 + }, + { + "epoch": 0.06554873934351009, + "grad_norm": 0.1426752209663391, + "learning_rate": 4.048812790044949e-05, + "loss": 3.1435, + "step": 11625 + }, + { + "epoch": 0.06557958580908352, + "grad_norm": 0.14376600086688995, + "learning_rate": 4.048601470618236e-05, + "loss": 3.1817, + "step": 11626 + }, + { + "epoch": 0.06561043227465693, + "grad_norm": 0.14624282717704773, + "learning_rate": 4.048390133236422e-05, + "loss": 3.1621, + "step": 11627 + }, + { + "epoch": 0.06564127874023035, + "grad_norm": 0.1472344845533371, + "learning_rate": 4.048178777901953e-05, + "loss": 3.1584, + "step": 11628 + }, + { + "epoch": 0.06567212520580376, + "grad_norm": 0.14640815556049347, + "learning_rate": 4.047967404617282e-05, + "loss": 3.1734, + "step": 11629 + }, + { + "epoch": 0.06570297167137717, + "grad_norm": 0.14953987300395966, + "learning_rate": 4.047756013384859e-05, + "loss": 3.2031, + "step": 11630 + }, + { + "epoch": 0.06573381813695059, + "grad_norm": 0.14887163043022156, + "learning_rate": 4.0475446042071355e-05, + "loss": 3.1639, + "step": 11631 + }, + { + "epoch": 0.06576466460252402, + "grad_norm": 0.14995215833187103, + "learning_rate": 4.0473331770865626e-05, + "loss": 3.1126, + "step": 11632 + }, + { + "epoch": 0.06579551106809743, + "grad_norm": 0.16405490040779114, + "learning_rate": 4.0471217320255906e-05, + "loss": 3.1743, + "step": 11633 + }, + { + "epoch": 0.06582635753367085, + "grad_norm": 0.1445521116256714, + "learning_rate": 4.0469102690266726e-05, + "loss": 3.1758, + "step": 11634 + }, + { + "epoch": 0.06585720399924426, + "grad_norm": 0.3671128749847412, + "learning_rate": 4.046698788092259e-05, + "loss": 3.1561, + "step": 11635 + }, + { + "epoch": 0.06588805046481767, + "grad_norm": 0.14986608922481537, + "learning_rate": 4.046487289224802e-05, + "loss": 3.1954, + "step": 11636 + }, + { + "epoch": 0.06591889693039109, + "grad_norm": 0.16334599256515503, + "learning_rate": 4.0462757724267553e-05, + "loss": 3.1791, + "step": 11637 + }, + { + "epoch": 0.06594974339596452, + "grad_norm": 0.14693418145179749, + "learning_rate": 4.0460642377005696e-05, + "loss": 3.1728, + "step": 11638 + }, + { + "epoch": 0.06598058986153793, + "grad_norm": 0.15964628756046295, + "learning_rate": 4.045852685048698e-05, + "loss": 3.1486, + "step": 11639 + }, + { + "epoch": 0.06601143632711134, + "grad_norm": 0.15065114200115204, + "learning_rate": 4.045641114473594e-05, + "loss": 3.1844, + "step": 11640 + }, + { + "epoch": 0.06604228279268476, + "grad_norm": 0.1558459848165512, + "learning_rate": 4.04542952597771e-05, + "loss": 3.1906, + "step": 11641 + }, + { + "epoch": 0.06607312925825817, + "grad_norm": 0.16925384104251862, + "learning_rate": 4.0452179195634986e-05, + "loss": 3.1662, + "step": 11642 + }, + { + "epoch": 0.06610397572383159, + "grad_norm": 0.15066556632518768, + "learning_rate": 4.045006295233414e-05, + "loss": 3.1829, + "step": 11643 + }, + { + "epoch": 0.06613482218940502, + "grad_norm": 0.14758241176605225, + "learning_rate": 4.044794652989911e-05, + "loss": 3.1186, + "step": 11644 + }, + { + "epoch": 0.06616566865497843, + "grad_norm": 0.1507830023765564, + "learning_rate": 4.0445829928354416e-05, + "loss": 3.1681, + "step": 11645 + }, + { + "epoch": 0.06619651512055184, + "grad_norm": 0.1524401605129242, + "learning_rate": 4.044371314772461e-05, + "loss": 3.2068, + "step": 11646 + }, + { + "epoch": 0.06622736158612526, + "grad_norm": 0.14559590816497803, + "learning_rate": 4.0441596188034225e-05, + "loss": 3.1843, + "step": 11647 + }, + { + "epoch": 0.06625820805169867, + "grad_norm": 0.15517529845237732, + "learning_rate": 4.0439479049307815e-05, + "loss": 3.1612, + "step": 11648 + }, + { + "epoch": 0.06628905451727209, + "grad_norm": 0.15106956660747528, + "learning_rate": 4.043736173156992e-05, + "loss": 3.1562, + "step": 11649 + }, + { + "epoch": 0.06631990098284551, + "grad_norm": 0.15036509931087494, + "learning_rate": 4.043524423484509e-05, + "loss": 3.1678, + "step": 11650 + }, + { + "epoch": 0.06635074744841893, + "grad_norm": 0.1449194997549057, + "learning_rate": 4.043312655915789e-05, + "loss": 3.1767, + "step": 11651 + }, + { + "epoch": 0.06638159391399234, + "grad_norm": 0.14780637621879578, + "learning_rate": 4.0431008704532856e-05, + "loss": 3.1666, + "step": 11652 + }, + { + "epoch": 0.06641244037956576, + "grad_norm": 0.6639805436134338, + "learning_rate": 4.0428890670994545e-05, + "loss": 3.148, + "step": 11653 + }, + { + "epoch": 0.06644328684513917, + "grad_norm": 0.1499292105436325, + "learning_rate": 4.0426772458567524e-05, + "loss": 3.2073, + "step": 11654 + }, + { + "epoch": 0.06647413331071259, + "grad_norm": 0.15838296711444855, + "learning_rate": 4.0424654067276344e-05, + "loss": 3.1596, + "step": 11655 + }, + { + "epoch": 0.06650497977628601, + "grad_norm": 0.1528780311346054, + "learning_rate": 4.042253549714557e-05, + "loss": 3.1716, + "step": 11656 + }, + { + "epoch": 0.06653582624185943, + "grad_norm": 0.15614338219165802, + "learning_rate": 4.042041674819976e-05, + "loss": 3.1492, + "step": 11657 + }, + { + "epoch": 0.06656667270743284, + "grad_norm": 0.15657266974449158, + "learning_rate": 4.04182978204635e-05, + "loss": 3.1583, + "step": 11658 + }, + { + "epoch": 0.06659751917300626, + "grad_norm": 0.1563463807106018, + "learning_rate": 4.041617871396132e-05, + "loss": 3.2, + "step": 11659 + }, + { + "epoch": 0.06662836563857967, + "grad_norm": 0.166629821062088, + "learning_rate": 4.041405942871783e-05, + "loss": 3.1918, + "step": 11660 + }, + { + "epoch": 0.06665921210415308, + "grad_norm": 0.14575469493865967, + "learning_rate": 4.041193996475757e-05, + "loss": 3.1739, + "step": 11661 + }, + { + "epoch": 0.06669005856972651, + "grad_norm": 0.15231850743293762, + "learning_rate": 4.040982032210514e-05, + "loss": 3.1706, + "step": 11662 + }, + { + "epoch": 0.06672090503529993, + "grad_norm": 0.15155643224716187, + "learning_rate": 4.040770050078509e-05, + "loss": 3.138, + "step": 11663 + }, + { + "epoch": 0.06675175150087334, + "grad_norm": 0.1702064573764801, + "learning_rate": 4.040558050082203e-05, + "loss": 3.1443, + "step": 11664 + }, + { + "epoch": 0.06678259796644676, + "grad_norm": 0.15021663904190063, + "learning_rate": 4.04034603222405e-05, + "loss": 3.1661, + "step": 11665 + }, + { + "epoch": 0.06681344443202017, + "grad_norm": 0.15412981808185577, + "learning_rate": 4.040133996506512e-05, + "loss": 3.1549, + "step": 11666 + }, + { + "epoch": 0.06684429089759358, + "grad_norm": 0.15156082808971405, + "learning_rate": 4.0399219429320455e-05, + "loss": 3.1433, + "step": 11667 + }, + { + "epoch": 0.06687513736316701, + "grad_norm": 0.15518224239349365, + "learning_rate": 4.039709871503109e-05, + "loss": 3.1579, + "step": 11668 + }, + { + "epoch": 0.06690598382874043, + "grad_norm": 0.15328356623649597, + "learning_rate": 4.039497782222162e-05, + "loss": 3.192, + "step": 11669 + }, + { + "epoch": 0.06693683029431384, + "grad_norm": 0.14678862690925598, + "learning_rate": 4.039285675091663e-05, + "loss": 3.1949, + "step": 11670 + }, + { + "epoch": 0.06696767675988725, + "grad_norm": 0.15256322920322418, + "learning_rate": 4.039073550114072e-05, + "loss": 3.1979, + "step": 11671 + }, + { + "epoch": 0.06699852322546067, + "grad_norm": 0.14585822820663452, + "learning_rate": 4.0388614072918484e-05, + "loss": 3.1681, + "step": 11672 + }, + { + "epoch": 0.06702936969103408, + "grad_norm": 0.14739924669265747, + "learning_rate": 4.0386492466274516e-05, + "loss": 3.1952, + "step": 11673 + }, + { + "epoch": 0.06706021615660751, + "grad_norm": 0.15530888736248016, + "learning_rate": 4.038437068123341e-05, + "loss": 3.1377, + "step": 11674 + }, + { + "epoch": 0.06709106262218092, + "grad_norm": 0.15336835384368896, + "learning_rate": 4.038224871781977e-05, + "loss": 3.1496, + "step": 11675 + }, + { + "epoch": 0.06712190908775434, + "grad_norm": 0.1455618143081665, + "learning_rate": 4.03801265760582e-05, + "loss": 3.1106, + "step": 11676 + }, + { + "epoch": 0.06715275555332775, + "grad_norm": 0.14603087306022644, + "learning_rate": 4.0378004255973305e-05, + "loss": 3.154, + "step": 11677 + }, + { + "epoch": 0.06718360201890117, + "grad_norm": 0.1677592396736145, + "learning_rate": 4.037588175758969e-05, + "loss": 3.1856, + "step": 11678 + }, + { + "epoch": 0.06721444848447458, + "grad_norm": 0.14778263866901398, + "learning_rate": 4.037375908093197e-05, + "loss": 3.1862, + "step": 11679 + }, + { + "epoch": 0.06724529495004801, + "grad_norm": 0.14966894686222076, + "learning_rate": 4.0371636226024755e-05, + "loss": 3.1746, + "step": 11680 + }, + { + "epoch": 0.06727614141562142, + "grad_norm": 0.1490064114332199, + "learning_rate": 4.036951319289265e-05, + "loss": 3.1566, + "step": 11681 + }, + { + "epoch": 0.06730698788119484, + "grad_norm": 0.1446569710969925, + "learning_rate": 4.036738998156027e-05, + "loss": 3.1768, + "step": 11682 + }, + { + "epoch": 0.06733783434676825, + "grad_norm": 0.14363335072994232, + "learning_rate": 4.036526659205225e-05, + "loss": 3.1712, + "step": 11683 + }, + { + "epoch": 0.06736868081234167, + "grad_norm": 0.14508138597011566, + "learning_rate": 4.036314302439319e-05, + "loss": 3.1841, + "step": 11684 + }, + { + "epoch": 0.06739952727791508, + "grad_norm": 0.14851996302604675, + "learning_rate": 4.036101927860772e-05, + "loss": 3.1568, + "step": 11685 + }, + { + "epoch": 0.06743037374348851, + "grad_norm": 0.1493758112192154, + "learning_rate": 4.035889535472046e-05, + "loss": 3.153, + "step": 11686 + }, + { + "epoch": 0.06746122020906192, + "grad_norm": 0.14475731551647186, + "learning_rate": 4.035677125275604e-05, + "loss": 3.1645, + "step": 11687 + }, + { + "epoch": 0.06749206667463534, + "grad_norm": 0.14519663155078888, + "learning_rate": 4.035464697273909e-05, + "loss": 3.179, + "step": 11688 + }, + { + "epoch": 0.06752291314020875, + "grad_norm": 0.14614298939704895, + "learning_rate": 4.035252251469423e-05, + "loss": 3.1808, + "step": 11689 + }, + { + "epoch": 0.06755375960578217, + "grad_norm": 0.14166559278964996, + "learning_rate": 4.03503978786461e-05, + "loss": 3.1755, + "step": 11690 + }, + { + "epoch": 0.06758460607135558, + "grad_norm": 0.15487481653690338, + "learning_rate": 4.0348273064619324e-05, + "loss": 3.156, + "step": 11691 + }, + { + "epoch": 0.06761545253692901, + "grad_norm": 0.1435340940952301, + "learning_rate": 4.034614807263856e-05, + "loss": 3.1626, + "step": 11692 + }, + { + "epoch": 0.06764629900250242, + "grad_norm": 0.1569274514913559, + "learning_rate": 4.034402290272842e-05, + "loss": 3.179, + "step": 11693 + }, + { + "epoch": 0.06767714546807584, + "grad_norm": 0.15384408831596375, + "learning_rate": 4.0341897554913557e-05, + "loss": 3.192, + "step": 11694 + }, + { + "epoch": 0.06770799193364925, + "grad_norm": 0.14062300324440002, + "learning_rate": 4.0339772029218604e-05, + "loss": 3.1621, + "step": 11695 + }, + { + "epoch": 0.06773883839922266, + "grad_norm": 0.14771996438503265, + "learning_rate": 4.033764632566822e-05, + "loss": 3.1932, + "step": 11696 + }, + { + "epoch": 0.06776968486479608, + "grad_norm": 0.1452936977148056, + "learning_rate": 4.0335520444287045e-05, + "loss": 3.1883, + "step": 11697 + }, + { + "epoch": 0.0678005313303695, + "grad_norm": 0.15978631377220154, + "learning_rate": 4.033339438509973e-05, + "loss": 3.1819, + "step": 11698 + }, + { + "epoch": 0.06783137779594292, + "grad_norm": 0.14421223104000092, + "learning_rate": 4.0331268148130915e-05, + "loss": 3.1894, + "step": 11699 + }, + { + "epoch": 0.06786222426151634, + "grad_norm": 0.154568150639534, + "learning_rate": 4.032914173340526e-05, + "loss": 3.1229, + "step": 11700 + }, + { + "epoch": 0.06789307072708975, + "grad_norm": 0.15651175379753113, + "learning_rate": 4.032701514094742e-05, + "loss": 3.153, + "step": 11701 + }, + { + "epoch": 0.06792391719266316, + "grad_norm": 0.14788556098937988, + "learning_rate": 4.0324888370782043e-05, + "loss": 3.1899, + "step": 11702 + }, + { + "epoch": 0.06795476365823659, + "grad_norm": 0.23043134808540344, + "learning_rate": 4.0322761422933796e-05, + "loss": 3.1425, + "step": 11703 + }, + { + "epoch": 0.06798561012381, + "grad_norm": 0.14422990381717682, + "learning_rate": 4.032063429742735e-05, + "loss": 3.1754, + "step": 11704 + }, + { + "epoch": 0.06801645658938342, + "grad_norm": 0.1473957896232605, + "learning_rate": 4.031850699428734e-05, + "loss": 3.1827, + "step": 11705 + }, + { + "epoch": 0.06804730305495683, + "grad_norm": 0.15429769456386566, + "learning_rate": 4.031637951353846e-05, + "loss": 3.1281, + "step": 11706 + }, + { + "epoch": 0.06807814952053025, + "grad_norm": 0.15346316993236542, + "learning_rate": 4.031425185520535e-05, + "loss": 3.1575, + "step": 11707 + }, + { + "epoch": 0.06810899598610366, + "grad_norm": 0.15392418205738068, + "learning_rate": 4.0312124019312706e-05, + "loss": 3.1706, + "step": 11708 + }, + { + "epoch": 0.06813984245167709, + "grad_norm": 0.15134009718894958, + "learning_rate": 4.030999600588517e-05, + "loss": 3.1608, + "step": 11709 + }, + { + "epoch": 0.0681706889172505, + "grad_norm": 1.1355273723602295, + "learning_rate": 4.030786781494745e-05, + "loss": 3.1742, + "step": 11710 + }, + { + "epoch": 0.06820153538282392, + "grad_norm": 0.15409262478351593, + "learning_rate": 4.0305739446524195e-05, + "loss": 3.1493, + "step": 11711 + }, + { + "epoch": 0.06823238184839733, + "grad_norm": 0.15430307388305664, + "learning_rate": 4.030361090064008e-05, + "loss": 3.1664, + "step": 11712 + }, + { + "epoch": 0.06826322831397075, + "grad_norm": 0.1602199226617813, + "learning_rate": 4.0301482177319806e-05, + "loss": 3.1447, + "step": 11713 + }, + { + "epoch": 0.06829407477954416, + "grad_norm": 0.15887390077114105, + "learning_rate": 4.029935327658803e-05, + "loss": 3.1686, + "step": 11714 + }, + { + "epoch": 0.06832492124511759, + "grad_norm": 0.15687309205532074, + "learning_rate": 4.029722419846945e-05, + "loss": 3.166, + "step": 11715 + }, + { + "epoch": 0.068355767710691, + "grad_norm": 0.16323551535606384, + "learning_rate": 4.029509494298875e-05, + "loss": 3.1599, + "step": 11716 + }, + { + "epoch": 0.06838661417626442, + "grad_norm": 0.15788815915584564, + "learning_rate": 4.0292965510170614e-05, + "loss": 3.1816, + "step": 11717 + }, + { + "epoch": 0.06841746064183783, + "grad_norm": 0.15840773284435272, + "learning_rate": 4.029083590003973e-05, + "loss": 3.1801, + "step": 11718 + }, + { + "epoch": 0.06844830710741125, + "grad_norm": 0.15174879133701324, + "learning_rate": 4.02887061126208e-05, + "loss": 3.1704, + "step": 11719 + }, + { + "epoch": 0.06847915357298466, + "grad_norm": 0.15559250116348267, + "learning_rate": 4.02865761479385e-05, + "loss": 3.1564, + "step": 11720 + }, + { + "epoch": 0.06851000003855809, + "grad_norm": 0.16039006412029266, + "learning_rate": 4.028444600601755e-05, + "loss": 3.156, + "step": 11721 + }, + { + "epoch": 0.0685408465041315, + "grad_norm": 0.1521182805299759, + "learning_rate": 4.028231568688262e-05, + "loss": 3.1782, + "step": 11722 + }, + { + "epoch": 0.06857169296970492, + "grad_norm": 0.1840634047985077, + "learning_rate": 4.0280185190558425e-05, + "loss": 3.1437, + "step": 11723 + }, + { + "epoch": 0.06860253943527833, + "grad_norm": 0.159534752368927, + "learning_rate": 4.0278054517069674e-05, + "loss": 3.1809, + "step": 11724 + }, + { + "epoch": 0.06863338590085175, + "grad_norm": 0.14995351433753967, + "learning_rate": 4.027592366644106e-05, + "loss": 3.1603, + "step": 11725 + }, + { + "epoch": 0.06866423236642516, + "grad_norm": 0.15074199438095093, + "learning_rate": 4.0273792638697285e-05, + "loss": 3.1489, + "step": 11726 + }, + { + "epoch": 0.06869507883199859, + "grad_norm": 0.1561887562274933, + "learning_rate": 4.027166143386306e-05, + "loss": 3.1658, + "step": 11727 + }, + { + "epoch": 0.068725925297572, + "grad_norm": 0.14531880617141724, + "learning_rate": 4.0269530051963106e-05, + "loss": 3.1621, + "step": 11728 + }, + { + "epoch": 0.06875677176314542, + "grad_norm": 0.14242933690547943, + "learning_rate": 4.026739849302212e-05, + "loss": 3.1722, + "step": 11729 + }, + { + "epoch": 0.06878761822871883, + "grad_norm": 0.15275083482265472, + "learning_rate": 4.026526675706484e-05, + "loss": 3.1778, + "step": 11730 + }, + { + "epoch": 0.06881846469429224, + "grad_norm": 0.16034279763698578, + "learning_rate": 4.0263134844115946e-05, + "loss": 3.2142, + "step": 11731 + }, + { + "epoch": 0.06884931115986566, + "grad_norm": 0.1466527283191681, + "learning_rate": 4.026100275420018e-05, + "loss": 3.1674, + "step": 11732 + }, + { + "epoch": 0.06888015762543909, + "grad_norm": 0.14914405345916748, + "learning_rate": 4.025887048734226e-05, + "loss": 3.1626, + "step": 11733 + }, + { + "epoch": 0.0689110040910125, + "grad_norm": 0.14642713963985443, + "learning_rate": 4.0256738043566903e-05, + "loss": 3.1671, + "step": 11734 + }, + { + "epoch": 0.06894185055658592, + "grad_norm": 0.152680441737175, + "learning_rate": 4.025460542289884e-05, + "loss": 3.1856, + "step": 11735 + }, + { + "epoch": 0.06897269702215933, + "grad_norm": 0.15504437685012817, + "learning_rate": 4.025247262536279e-05, + "loss": 3.1653, + "step": 11736 + }, + { + "epoch": 0.06900354348773274, + "grad_norm": 0.15097057819366455, + "learning_rate": 4.0250339650983495e-05, + "loss": 3.1847, + "step": 11737 + }, + { + "epoch": 0.06903438995330616, + "grad_norm": 0.14939604699611664, + "learning_rate": 4.0248206499785667e-05, + "loss": 3.1536, + "step": 11738 + }, + { + "epoch": 0.06906523641887959, + "grad_norm": 0.14170803129673004, + "learning_rate": 4.024607317179405e-05, + "loss": 3.1637, + "step": 11739 + }, + { + "epoch": 0.069096082884453, + "grad_norm": 0.1565011441707611, + "learning_rate": 4.024393966703338e-05, + "loss": 3.1872, + "step": 11740 + }, + { + "epoch": 0.06912692935002641, + "grad_norm": 0.14462994039058685, + "learning_rate": 4.024180598552838e-05, + "loss": 3.1552, + "step": 11741 + }, + { + "epoch": 0.06915777581559983, + "grad_norm": 0.14230725169181824, + "learning_rate": 4.023967212730381e-05, + "loss": 3.16, + "step": 11742 + }, + { + "epoch": 0.06918862228117324, + "grad_norm": 0.1458989977836609, + "learning_rate": 4.02375380923844e-05, + "loss": 3.1637, + "step": 11743 + }, + { + "epoch": 0.06921946874674666, + "grad_norm": 0.14278742671012878, + "learning_rate": 4.023540388079489e-05, + "loss": 3.1697, + "step": 11744 + }, + { + "epoch": 0.06925031521232008, + "grad_norm": 0.14817160367965698, + "learning_rate": 4.023326949256002e-05, + "loss": 3.1749, + "step": 11745 + }, + { + "epoch": 0.0692811616778935, + "grad_norm": 0.15352784097194672, + "learning_rate": 4.0231134927704556e-05, + "loss": 3.1991, + "step": 11746 + }, + { + "epoch": 0.06931200814346691, + "grad_norm": 0.14961881935596466, + "learning_rate": 4.022900018625323e-05, + "loss": 3.1787, + "step": 11747 + }, + { + "epoch": 0.06934285460904033, + "grad_norm": 0.14419247210025787, + "learning_rate": 4.0226865268230794e-05, + "loss": 3.1651, + "step": 11748 + }, + { + "epoch": 0.06937370107461374, + "grad_norm": 0.14211702346801758, + "learning_rate": 4.022473017366201e-05, + "loss": 3.1345, + "step": 11749 + }, + { + "epoch": 0.06940454754018716, + "grad_norm": 0.1551240235567093, + "learning_rate": 4.022259490257163e-05, + "loss": 3.1853, + "step": 11750 + }, + { + "epoch": 0.06943539400576058, + "grad_norm": 0.14813299477100372, + "learning_rate": 4.0220459454984405e-05, + "loss": 3.1608, + "step": 11751 + }, + { + "epoch": 0.069466240471334, + "grad_norm": 0.1561623513698578, + "learning_rate": 4.021832383092511e-05, + "loss": 3.1583, + "step": 11752 + }, + { + "epoch": 0.06949708693690741, + "grad_norm": 0.15094567835330963, + "learning_rate": 4.021618803041848e-05, + "loss": 3.1224, + "step": 11753 + }, + { + "epoch": 0.06952793340248083, + "grad_norm": 0.15407326817512512, + "learning_rate": 4.0214052053489304e-05, + "loss": 3.1363, + "step": 11754 + }, + { + "epoch": 0.06955877986805424, + "grad_norm": 0.1590680480003357, + "learning_rate": 4.021191590016233e-05, + "loss": 3.1611, + "step": 11755 + }, + { + "epoch": 0.06958962633362765, + "grad_norm": 0.15361779928207397, + "learning_rate": 4.0209779570462336e-05, + "loss": 3.188, + "step": 11756 + }, + { + "epoch": 0.06962047279920108, + "grad_norm": 0.14506292343139648, + "learning_rate": 4.020764306441409e-05, + "loss": 3.1531, + "step": 11757 + }, + { + "epoch": 0.0696513192647745, + "grad_norm": 0.1594819724559784, + "learning_rate": 4.020550638204236e-05, + "loss": 3.1535, + "step": 11758 + }, + { + "epoch": 0.06968216573034791, + "grad_norm": 0.20404116809368134, + "learning_rate": 4.020336952337192e-05, + "loss": 3.1567, + "step": 11759 + }, + { + "epoch": 0.06971301219592133, + "grad_norm": 0.15393637120723724, + "learning_rate": 4.020123248842755e-05, + "loss": 3.177, + "step": 11760 + }, + { + "epoch": 0.06974385866149474, + "grad_norm": 0.16881807148456573, + "learning_rate": 4.0199095277234025e-05, + "loss": 3.1824, + "step": 11761 + }, + { + "epoch": 0.06977470512706815, + "grad_norm": 0.14597831666469574, + "learning_rate": 4.019695788981611e-05, + "loss": 3.1536, + "step": 11762 + }, + { + "epoch": 0.06980555159264158, + "grad_norm": 0.1526559293270111, + "learning_rate": 4.019482032619862e-05, + "loss": 3.1648, + "step": 11763 + }, + { + "epoch": 0.069836398058215, + "grad_norm": 0.158670574426651, + "learning_rate": 4.0192682586406306e-05, + "loss": 3.1619, + "step": 11764 + }, + { + "epoch": 0.06986724452378841, + "grad_norm": 0.19919368624687195, + "learning_rate": 4.019054467046397e-05, + "loss": 3.1912, + "step": 11765 + }, + { + "epoch": 0.06989809098936182, + "grad_norm": 0.1666477620601654, + "learning_rate": 4.01884065783964e-05, + "loss": 3.1482, + "step": 11766 + }, + { + "epoch": 0.06992893745493524, + "grad_norm": 0.1488250195980072, + "learning_rate": 4.018626831022838e-05, + "loss": 3.1425, + "step": 11767 + }, + { + "epoch": 0.06995978392050865, + "grad_norm": 0.16196392476558685, + "learning_rate": 4.018412986598471e-05, + "loss": 3.1274, + "step": 11768 + }, + { + "epoch": 0.06999063038608208, + "grad_norm": 0.1488756388425827, + "learning_rate": 4.018199124569018e-05, + "loss": 3.1419, + "step": 11769 + }, + { + "epoch": 0.0700214768516555, + "grad_norm": 0.15442027151584625, + "learning_rate": 4.017985244936957e-05, + "loss": 3.1431, + "step": 11770 + }, + { + "epoch": 0.07005232331722891, + "grad_norm": 0.14774125814437866, + "learning_rate": 4.0177713477047704e-05, + "loss": 3.1662, + "step": 11771 + }, + { + "epoch": 0.07008316978280232, + "grad_norm": 0.14652900397777557, + "learning_rate": 4.0175574328749375e-05, + "loss": 3.176, + "step": 11772 + }, + { + "epoch": 0.07011401624837574, + "grad_norm": 0.14855022728443146, + "learning_rate": 4.017343500449937e-05, + "loss": 3.1448, + "step": 11773 + }, + { + "epoch": 0.07014486271394915, + "grad_norm": 0.14556941390037537, + "learning_rate": 4.017129550432251e-05, + "loss": 3.1808, + "step": 11774 + }, + { + "epoch": 0.07017570917952258, + "grad_norm": 0.15503740310668945, + "learning_rate": 4.01691558282436e-05, + "loss": 3.1537, + "step": 11775 + }, + { + "epoch": 0.070206555645096, + "grad_norm": 0.14603111147880554, + "learning_rate": 4.0167015976287435e-05, + "loss": 3.1861, + "step": 11776 + }, + { + "epoch": 0.07023740211066941, + "grad_norm": 0.1458377242088318, + "learning_rate": 4.016487594847884e-05, + "loss": 3.1572, + "step": 11777 + }, + { + "epoch": 0.07026824857624282, + "grad_norm": 0.14365902543067932, + "learning_rate": 4.016273574484262e-05, + "loss": 3.1694, + "step": 11778 + }, + { + "epoch": 0.07029909504181624, + "grad_norm": 0.14608101546764374, + "learning_rate": 4.0160595365403585e-05, + "loss": 3.1397, + "step": 11779 + }, + { + "epoch": 0.07032994150738965, + "grad_norm": 0.19157761335372925, + "learning_rate": 4.0158454810186566e-05, + "loss": 3.1643, + "step": 11780 + }, + { + "epoch": 0.07036078797296308, + "grad_norm": 0.14479336142539978, + "learning_rate": 4.015631407921637e-05, + "loss": 3.187, + "step": 11781 + }, + { + "epoch": 0.0703916344385365, + "grad_norm": 0.14744655787944794, + "learning_rate": 4.015417317251782e-05, + "loss": 3.1709, + "step": 11782 + }, + { + "epoch": 0.07042248090410991, + "grad_norm": 0.14978794753551483, + "learning_rate": 4.015203209011574e-05, + "loss": 3.1476, + "step": 11783 + }, + { + "epoch": 0.07045332736968332, + "grad_norm": 0.1425250619649887, + "learning_rate": 4.0149890832034954e-05, + "loss": 3.1621, + "step": 11784 + }, + { + "epoch": 0.07048417383525674, + "grad_norm": 0.14373183250427246, + "learning_rate": 4.014774939830028e-05, + "loss": 3.1572, + "step": 11785 + }, + { + "epoch": 0.07051502030083015, + "grad_norm": 0.147437185049057, + "learning_rate": 4.014560778893656e-05, + "loss": 3.1646, + "step": 11786 + }, + { + "epoch": 0.07054586676640358, + "grad_norm": 0.1449986845254898, + "learning_rate": 4.014346600396862e-05, + "loss": 3.1468, + "step": 11787 + }, + { + "epoch": 0.07057671323197699, + "grad_norm": 0.1473894715309143, + "learning_rate": 4.0141324043421294e-05, + "loss": 3.1688, + "step": 11788 + }, + { + "epoch": 0.0706075596975504, + "grad_norm": 0.14858457446098328, + "learning_rate": 4.0139181907319404e-05, + "loss": 3.1708, + "step": 11789 + }, + { + "epoch": 0.07063840616312382, + "grad_norm": 0.1480007767677307, + "learning_rate": 4.013703959568781e-05, + "loss": 3.1704, + "step": 11790 + }, + { + "epoch": 0.07066925262869724, + "grad_norm": 0.1512412428855896, + "learning_rate": 4.013489710855133e-05, + "loss": 3.1874, + "step": 11791 + }, + { + "epoch": 0.07070009909427065, + "grad_norm": 0.1401321440935135, + "learning_rate": 4.013275444593482e-05, + "loss": 3.1587, + "step": 11792 + }, + { + "epoch": 0.07073094555984408, + "grad_norm": 0.14502422511577606, + "learning_rate": 4.013061160786311e-05, + "loss": 3.1703, + "step": 11793 + }, + { + "epoch": 0.07076179202541749, + "grad_norm": 0.14687718451023102, + "learning_rate": 4.012846859436105e-05, + "loss": 3.1685, + "step": 11794 + }, + { + "epoch": 0.0707926384909909, + "grad_norm": 0.1431635171175003, + "learning_rate": 4.012632540545349e-05, + "loss": 3.1972, + "step": 11795 + }, + { + "epoch": 0.07082348495656432, + "grad_norm": 0.1532827615737915, + "learning_rate": 4.012418204116528e-05, + "loss": 3.1509, + "step": 11796 + }, + { + "epoch": 0.07085433142213773, + "grad_norm": 0.1491946578025818, + "learning_rate": 4.0122038501521266e-05, + "loss": 3.1533, + "step": 11797 + }, + { + "epoch": 0.07088517788771115, + "grad_norm": 0.1520770937204361, + "learning_rate": 4.0119894786546296e-05, + "loss": 3.1729, + "step": 11798 + }, + { + "epoch": 0.07091602435328458, + "grad_norm": 0.14371779561042786, + "learning_rate": 4.011775089626525e-05, + "loss": 3.2059, + "step": 11799 + }, + { + "epoch": 0.07094687081885799, + "grad_norm": 0.15152618288993835, + "learning_rate": 4.011560683070295e-05, + "loss": 3.1795, + "step": 11800 + }, + { + "epoch": 0.0709777172844314, + "grad_norm": 0.33709144592285156, + "learning_rate": 4.011346258988428e-05, + "loss": 3.1649, + "step": 11801 + }, + { + "epoch": 0.07100856375000482, + "grad_norm": 0.1428249627351761, + "learning_rate": 4.011131817383409e-05, + "loss": 3.1736, + "step": 11802 + }, + { + "epoch": 0.07103941021557823, + "grad_norm": 0.15872763097286224, + "learning_rate": 4.0109173582577245e-05, + "loss": 3.1643, + "step": 11803 + }, + { + "epoch": 0.07107025668115165, + "grad_norm": 0.1417926549911499, + "learning_rate": 4.010702881613861e-05, + "loss": 3.1701, + "step": 11804 + }, + { + "epoch": 0.07110110314672508, + "grad_norm": 0.1609552949666977, + "learning_rate": 4.010488387454306e-05, + "loss": 3.16, + "step": 11805 + }, + { + "epoch": 0.07113194961229849, + "grad_norm": 0.14309971034526825, + "learning_rate": 4.010273875781546e-05, + "loss": 3.1877, + "step": 11806 + }, + { + "epoch": 0.0711627960778719, + "grad_norm": 0.18360304832458496, + "learning_rate": 4.0100593465980676e-05, + "loss": 3.1792, + "step": 11807 + }, + { + "epoch": 0.07119364254344532, + "grad_norm": 0.16597242653369904, + "learning_rate": 4.009844799906358e-05, + "loss": 3.1697, + "step": 11808 + }, + { + "epoch": 0.07122448900901873, + "grad_norm": 0.14606373012065887, + "learning_rate": 4.0096302357089056e-05, + "loss": 3.2021, + "step": 11809 + }, + { + "epoch": 0.07125533547459215, + "grad_norm": 0.14357510209083557, + "learning_rate": 4.009415654008198e-05, + "loss": 3.1715, + "step": 11810 + }, + { + "epoch": 0.07128618194016557, + "grad_norm": 0.14426608383655548, + "learning_rate": 4.009201054806723e-05, + "loss": 3.1786, + "step": 11811 + }, + { + "epoch": 0.07131702840573899, + "grad_norm": 0.14274980127811432, + "learning_rate": 4.0089864381069686e-05, + "loss": 3.1554, + "step": 11812 + }, + { + "epoch": 0.0713478748713124, + "grad_norm": 0.1486242413520813, + "learning_rate": 4.0087718039114235e-05, + "loss": 3.1398, + "step": 11813 + }, + { + "epoch": 0.07137872133688582, + "grad_norm": 0.15553520619869232, + "learning_rate": 4.0085571522225765e-05, + "loss": 3.2105, + "step": 11814 + }, + { + "epoch": 0.07140956780245923, + "grad_norm": 0.14317598938941956, + "learning_rate": 4.0083424830429145e-05, + "loss": 3.1734, + "step": 11815 + }, + { + "epoch": 0.07144041426803265, + "grad_norm": 0.1468994915485382, + "learning_rate": 4.0081277963749285e-05, + "loss": 3.2168, + "step": 11816 + }, + { + "epoch": 0.07147126073360607, + "grad_norm": 0.1418161690235138, + "learning_rate": 4.0079130922211075e-05, + "loss": 3.1559, + "step": 11817 + }, + { + "epoch": 0.07150210719917949, + "grad_norm": 0.14928217232227325, + "learning_rate": 4.007698370583939e-05, + "loss": 3.1784, + "step": 11818 + }, + { + "epoch": 0.0715329536647529, + "grad_norm": 0.1455998569726944, + "learning_rate": 4.0074836314659157e-05, + "loss": 3.2126, + "step": 11819 + }, + { + "epoch": 0.07156380013032632, + "grad_norm": 0.15511727333068848, + "learning_rate": 4.007268874869524e-05, + "loss": 3.1115, + "step": 11820 + }, + { + "epoch": 0.07159464659589973, + "grad_norm": 0.14607951045036316, + "learning_rate": 4.007054100797256e-05, + "loss": 3.1543, + "step": 11821 + }, + { + "epoch": 0.07162549306147314, + "grad_norm": 0.6089773774147034, + "learning_rate": 4.006839309251601e-05, + "loss": 3.1725, + "step": 11822 + }, + { + "epoch": 0.07165633952704657, + "grad_norm": 0.1446574330329895, + "learning_rate": 4.00662450023505e-05, + "loss": 3.1569, + "step": 11823 + }, + { + "epoch": 0.07168718599261999, + "grad_norm": 0.1477476805448532, + "learning_rate": 4.006409673750094e-05, + "loss": 3.1238, + "step": 11824 + }, + { + "epoch": 0.0717180324581934, + "grad_norm": 0.14739441871643066, + "learning_rate": 4.006194829799222e-05, + "loss": 3.1959, + "step": 11825 + }, + { + "epoch": 0.07174887892376682, + "grad_norm": 0.14948022365570068, + "learning_rate": 4.005979968384926e-05, + "loss": 3.15, + "step": 11826 + }, + { + "epoch": 0.07177972538934023, + "grad_norm": 0.311659574508667, + "learning_rate": 4.005765089509699e-05, + "loss": 3.1727, + "step": 11827 + }, + { + "epoch": 0.07181057185491364, + "grad_norm": 0.1466073989868164, + "learning_rate": 4.005550193176029e-05, + "loss": 3.1526, + "step": 11828 + }, + { + "epoch": 0.07184141832048707, + "grad_norm": 0.15548254549503326, + "learning_rate": 4.005335279386409e-05, + "loss": 3.1903, + "step": 11829 + }, + { + "epoch": 0.07187226478606049, + "grad_norm": 0.14565548300743103, + "learning_rate": 4.005120348143333e-05, + "loss": 3.168, + "step": 11830 + }, + { + "epoch": 0.0719031112516339, + "grad_norm": 0.14981408417224884, + "learning_rate": 4.004905399449289e-05, + "loss": 3.1478, + "step": 11831 + }, + { + "epoch": 0.07193395771720731, + "grad_norm": 0.14962203800678253, + "learning_rate": 4.004690433306772e-05, + "loss": 3.1284, + "step": 11832 + }, + { + "epoch": 0.07196480418278073, + "grad_norm": 0.1455821394920349, + "learning_rate": 4.004475449718274e-05, + "loss": 3.1559, + "step": 11833 + }, + { + "epoch": 0.07199565064835414, + "grad_norm": 0.15584930777549744, + "learning_rate": 4.004260448686287e-05, + "loss": 3.1261, + "step": 11834 + }, + { + "epoch": 0.07202649711392757, + "grad_norm": 0.14084406197071075, + "learning_rate": 4.004045430213303e-05, + "loss": 3.1296, + "step": 11835 + }, + { + "epoch": 0.07205734357950098, + "grad_norm": 0.14498230814933777, + "learning_rate": 4.003830394301818e-05, + "loss": 3.1671, + "step": 11836 + }, + { + "epoch": 0.0720881900450744, + "grad_norm": 0.14855846762657166, + "learning_rate": 4.003615340954322e-05, + "loss": 3.1362, + "step": 11837 + }, + { + "epoch": 0.07211903651064781, + "grad_norm": 0.15041877329349518, + "learning_rate": 4.00340027017331e-05, + "loss": 3.1773, + "step": 11838 + }, + { + "epoch": 0.07214988297622123, + "grad_norm": 0.14652130007743835, + "learning_rate": 4.0031851819612754e-05, + "loss": 3.1491, + "step": 11839 + }, + { + "epoch": 0.07218072944179464, + "grad_norm": 0.1486414223909378, + "learning_rate": 4.0029700763207125e-05, + "loss": 3.1634, + "step": 11840 + }, + { + "epoch": 0.07221157590736807, + "grad_norm": 0.15160280466079712, + "learning_rate": 4.002754953254113e-05, + "loss": 3.1762, + "step": 11841 + }, + { + "epoch": 0.07224242237294148, + "grad_norm": 0.15478084981441498, + "learning_rate": 4.0025398127639746e-05, + "loss": 3.1714, + "step": 11842 + }, + { + "epoch": 0.0722732688385149, + "grad_norm": 0.14596900343894958, + "learning_rate": 4.002324654852789e-05, + "loss": 3.1686, + "step": 11843 + }, + { + "epoch": 0.07230411530408831, + "grad_norm": 0.14873766899108887, + "learning_rate": 4.002109479523052e-05, + "loss": 3.1454, + "step": 11844 + }, + { + "epoch": 0.07233496176966173, + "grad_norm": 0.1490432769060135, + "learning_rate": 4.001894286777259e-05, + "loss": 3.1593, + "step": 11845 + }, + { + "epoch": 0.07236580823523514, + "grad_norm": 0.14743174612522125, + "learning_rate": 4.0016790766179037e-05, + "loss": 3.161, + "step": 11846 + }, + { + "epoch": 0.07239665470080857, + "grad_norm": 0.1442510187625885, + "learning_rate": 4.0014638490474806e-05, + "loss": 3.1394, + "step": 11847 + }, + { + "epoch": 0.07242750116638198, + "grad_norm": 0.1557513028383255, + "learning_rate": 4.001248604068488e-05, + "loss": 3.1954, + "step": 11848 + }, + { + "epoch": 0.0724583476319554, + "grad_norm": 0.15510843694210052, + "learning_rate": 4.00103334168342e-05, + "loss": 3.1689, + "step": 11849 + }, + { + "epoch": 0.07248919409752881, + "grad_norm": 0.15135319530963898, + "learning_rate": 4.000818061894772e-05, + "loss": 3.1849, + "step": 11850 + }, + { + "epoch": 0.07252004056310223, + "grad_norm": 0.1426791399717331, + "learning_rate": 4.000602764705041e-05, + "loss": 3.142, + "step": 11851 + }, + { + "epoch": 0.07255088702867564, + "grad_norm": 0.14907197654247284, + "learning_rate": 4.0003874501167223e-05, + "loss": 3.1536, + "step": 11852 + }, + { + "epoch": 0.07258173349424907, + "grad_norm": 0.15063263475894928, + "learning_rate": 4.000172118132313e-05, + "loss": 3.1907, + "step": 11853 + }, + { + "epoch": 0.07261257995982248, + "grad_norm": 0.14683349430561066, + "learning_rate": 3.9999567687543086e-05, + "loss": 3.1563, + "step": 11854 + }, + { + "epoch": 0.0726434264253959, + "grad_norm": 0.14338700473308563, + "learning_rate": 3.9997414019852076e-05, + "loss": 3.1616, + "step": 11855 + }, + { + "epoch": 0.07267427289096931, + "grad_norm": 0.15298128128051758, + "learning_rate": 3.999526017827506e-05, + "loss": 3.1853, + "step": 11856 + }, + { + "epoch": 0.07270511935654272, + "grad_norm": 0.14598655700683594, + "learning_rate": 3.9993106162837014e-05, + "loss": 3.1668, + "step": 11857 + }, + { + "epoch": 0.07273596582211614, + "grad_norm": 0.1493327021598816, + "learning_rate": 3.9990951973562915e-05, + "loss": 3.1908, + "step": 11858 + }, + { + "epoch": 0.07276681228768957, + "grad_norm": 0.15319550037384033, + "learning_rate": 3.998879761047773e-05, + "loss": 3.1843, + "step": 11859 + }, + { + "epoch": 0.07279765875326298, + "grad_norm": 0.1535869985818863, + "learning_rate": 3.998664307360645e-05, + "loss": 3.1667, + "step": 11860 + }, + { + "epoch": 0.0728285052188364, + "grad_norm": 0.15534810721874237, + "learning_rate": 3.9984488362974045e-05, + "loss": 3.171, + "step": 11861 + }, + { + "epoch": 0.07285935168440981, + "grad_norm": 0.15699131786823273, + "learning_rate": 3.9982333478605506e-05, + "loss": 3.1442, + "step": 11862 + }, + { + "epoch": 0.07289019814998322, + "grad_norm": 0.15391065180301666, + "learning_rate": 3.998017842052582e-05, + "loss": 3.1562, + "step": 11863 + }, + { + "epoch": 0.07292104461555664, + "grad_norm": 0.14083610475063324, + "learning_rate": 3.997802318875995e-05, + "loss": 3.1583, + "step": 11864 + }, + { + "epoch": 0.07295189108113007, + "grad_norm": 0.14633245766162872, + "learning_rate": 3.997586778333291e-05, + "loss": 3.1673, + "step": 11865 + }, + { + "epoch": 0.07298273754670348, + "grad_norm": 0.1444370448589325, + "learning_rate": 3.9973712204269686e-05, + "loss": 3.1514, + "step": 11866 + }, + { + "epoch": 0.0730135840122769, + "grad_norm": 0.14843915402889252, + "learning_rate": 3.997155645159526e-05, + "loss": 3.1505, + "step": 11867 + }, + { + "epoch": 0.07304443047785031, + "grad_norm": 0.1550016850233078, + "learning_rate": 3.996940052533464e-05, + "loss": 3.1655, + "step": 11868 + }, + { + "epoch": 0.07307527694342372, + "grad_norm": 0.14386451244354248, + "learning_rate": 3.9967244425512815e-05, + "loss": 3.1709, + "step": 11869 + }, + { + "epoch": 0.07310612340899714, + "grad_norm": 0.15287421643733978, + "learning_rate": 3.996508815215479e-05, + "loss": 3.152, + "step": 11870 + }, + { + "epoch": 0.07313696987457056, + "grad_norm": 0.14842891693115234, + "learning_rate": 3.9962931705285556e-05, + "loss": 3.1393, + "step": 11871 + }, + { + "epoch": 0.07316781634014398, + "grad_norm": 0.1567426472902298, + "learning_rate": 3.996077508493013e-05, + "loss": 3.1678, + "step": 11872 + }, + { + "epoch": 0.0731986628057174, + "grad_norm": 0.16326004266738892, + "learning_rate": 3.9958618291113484e-05, + "loss": 3.1395, + "step": 11873 + }, + { + "epoch": 0.07322950927129081, + "grad_norm": 0.15595798194408417, + "learning_rate": 3.995646132386066e-05, + "loss": 3.167, + "step": 11874 + }, + { + "epoch": 0.07326035573686422, + "grad_norm": 0.15174920856952667, + "learning_rate": 3.995430418319666e-05, + "loss": 3.1632, + "step": 11875 + }, + { + "epoch": 0.07329120220243764, + "grad_norm": 0.18931283056735992, + "learning_rate": 3.995214686914648e-05, + "loss": 3.194, + "step": 11876 + }, + { + "epoch": 0.07332204866801106, + "grad_norm": 0.15182940661907196, + "learning_rate": 3.9949989381735156e-05, + "loss": 3.1586, + "step": 11877 + }, + { + "epoch": 0.07335289513358448, + "grad_norm": 0.14939960837364197, + "learning_rate": 3.9947831720987684e-05, + "loss": 3.1174, + "step": 11878 + }, + { + "epoch": 0.07338374159915789, + "grad_norm": 0.15009094774723053, + "learning_rate": 3.994567388692908e-05, + "loss": 3.1767, + "step": 11879 + }, + { + "epoch": 0.0734145880647313, + "grad_norm": 0.14926967024803162, + "learning_rate": 3.994351587958437e-05, + "loss": 3.1853, + "step": 11880 + }, + { + "epoch": 0.07344543453030472, + "grad_norm": 0.14789046347141266, + "learning_rate": 3.9941357698978574e-05, + "loss": 3.1813, + "step": 11881 + }, + { + "epoch": 0.07347628099587813, + "grad_norm": 0.14964108169078827, + "learning_rate": 3.9939199345136716e-05, + "loss": 3.1687, + "step": 11882 + }, + { + "epoch": 0.07350712746145156, + "grad_norm": 0.1461850255727768, + "learning_rate": 3.993704081808382e-05, + "loss": 3.1916, + "step": 11883 + }, + { + "epoch": 0.07353797392702498, + "grad_norm": 0.14604735374450684, + "learning_rate": 3.993488211784491e-05, + "loss": 3.1724, + "step": 11884 + }, + { + "epoch": 0.07356882039259839, + "grad_norm": 0.14693868160247803, + "learning_rate": 3.9932723244445016e-05, + "loss": 3.1769, + "step": 11885 + }, + { + "epoch": 0.0735996668581718, + "grad_norm": 0.14719203114509583, + "learning_rate": 3.993056419790917e-05, + "loss": 3.145, + "step": 11886 + }, + { + "epoch": 0.07363051332374522, + "grad_norm": 0.14706715941429138, + "learning_rate": 3.9928404978262406e-05, + "loss": 3.1641, + "step": 11887 + }, + { + "epoch": 0.07366135978931863, + "grad_norm": 0.1456660032272339, + "learning_rate": 3.9926245585529753e-05, + "loss": 3.1807, + "step": 11888 + }, + { + "epoch": 0.07369220625489206, + "grad_norm": 0.14502492547035217, + "learning_rate": 3.992408601973626e-05, + "loss": 3.1674, + "step": 11889 + }, + { + "epoch": 0.07372305272046548, + "grad_norm": 0.1437671184539795, + "learning_rate": 3.992192628090695e-05, + "loss": 3.1615, + "step": 11890 + }, + { + "epoch": 0.07375389918603889, + "grad_norm": 0.1456734538078308, + "learning_rate": 3.9919766369066875e-05, + "loss": 3.1566, + "step": 11891 + }, + { + "epoch": 0.0737847456516123, + "grad_norm": 0.15752439200878143, + "learning_rate": 3.991760628424107e-05, + "loss": 3.1808, + "step": 11892 + }, + { + "epoch": 0.07381559211718572, + "grad_norm": 0.14813625812530518, + "learning_rate": 3.9915446026454594e-05, + "loss": 3.1958, + "step": 11893 + }, + { + "epoch": 0.07384643858275913, + "grad_norm": 0.13897094130516052, + "learning_rate": 3.991328559573248e-05, + "loss": 3.1267, + "step": 11894 + }, + { + "epoch": 0.07387728504833256, + "grad_norm": 0.1436348706483841, + "learning_rate": 3.991112499209978e-05, + "loss": 3.1428, + "step": 11895 + }, + { + "epoch": 0.07390813151390598, + "grad_norm": 0.27737361192703247, + "learning_rate": 3.990896421558155e-05, + "loss": 3.1572, + "step": 11896 + }, + { + "epoch": 0.07393897797947939, + "grad_norm": 0.14459003508090973, + "learning_rate": 3.990680326620283e-05, + "loss": 3.1665, + "step": 11897 + }, + { + "epoch": 0.0739698244450528, + "grad_norm": 0.1512211561203003, + "learning_rate": 3.990464214398869e-05, + "loss": 3.136, + "step": 11898 + }, + { + "epoch": 0.07400067091062622, + "grad_norm": 0.14391222596168518, + "learning_rate": 3.990248084896419e-05, + "loss": 3.1942, + "step": 11899 + }, + { + "epoch": 0.07403151737619963, + "grad_norm": 0.1538868546485901, + "learning_rate": 3.9900319381154364e-05, + "loss": 3.1309, + "step": 11900 + }, + { + "epoch": 0.07406236384177306, + "grad_norm": 0.15256322920322418, + "learning_rate": 3.98981577405843e-05, + "loss": 3.1646, + "step": 11901 + }, + { + "epoch": 0.07409321030734647, + "grad_norm": 0.1478494554758072, + "learning_rate": 3.9895995927279045e-05, + "loss": 3.1631, + "step": 11902 + }, + { + "epoch": 0.07412405677291989, + "grad_norm": 0.15493758022785187, + "learning_rate": 3.989383394126367e-05, + "loss": 3.1947, + "step": 11903 + }, + { + "epoch": 0.0741549032384933, + "grad_norm": 0.14749978482723236, + "learning_rate": 3.9891671782563234e-05, + "loss": 3.2007, + "step": 11904 + }, + { + "epoch": 0.07418574970406672, + "grad_norm": 0.15063810348510742, + "learning_rate": 3.988950945120282e-05, + "loss": 3.1751, + "step": 11905 + }, + { + "epoch": 0.07421659616964013, + "grad_norm": 0.14900127053260803, + "learning_rate": 3.988734694720749e-05, + "loss": 3.2083, + "step": 11906 + }, + { + "epoch": 0.07424744263521356, + "grad_norm": 0.14685988426208496, + "learning_rate": 3.988518427060232e-05, + "loss": 3.1604, + "step": 11907 + }, + { + "epoch": 0.07427828910078697, + "grad_norm": 0.15100333094596863, + "learning_rate": 3.988302142141238e-05, + "loss": 3.1692, + "step": 11908 + }, + { + "epoch": 0.07430913556636039, + "grad_norm": 0.14235474169254303, + "learning_rate": 3.988085839966275e-05, + "loss": 3.1543, + "step": 11909 + }, + { + "epoch": 0.0743399820319338, + "grad_norm": 0.1529495120048523, + "learning_rate": 3.9878695205378515e-05, + "loss": 3.1432, + "step": 11910 + }, + { + "epoch": 0.07437082849750722, + "grad_norm": 0.14803078770637512, + "learning_rate": 3.9876531838584744e-05, + "loss": 3.1564, + "step": 11911 + }, + { + "epoch": 0.07440167496308063, + "grad_norm": 0.1451490968465805, + "learning_rate": 3.987436829930653e-05, + "loss": 3.178, + "step": 11912 + }, + { + "epoch": 0.07443252142865406, + "grad_norm": 0.1400495022535324, + "learning_rate": 3.987220458756895e-05, + "loss": 3.1684, + "step": 11913 + }, + { + "epoch": 0.07446336789422747, + "grad_norm": 0.17477715015411377, + "learning_rate": 3.98700407033971e-05, + "loss": 3.1329, + "step": 11914 + }, + { + "epoch": 0.07449421435980089, + "grad_norm": 0.1445254385471344, + "learning_rate": 3.986787664681606e-05, + "loss": 3.1936, + "step": 11915 + }, + { + "epoch": 0.0745250608253743, + "grad_norm": 0.14494265615940094, + "learning_rate": 3.9865712417850934e-05, + "loss": 3.1826, + "step": 11916 + }, + { + "epoch": 0.07455590729094771, + "grad_norm": 0.16903559863567352, + "learning_rate": 3.986354801652679e-05, + "loss": 3.1785, + "step": 11917 + }, + { + "epoch": 0.07458675375652113, + "grad_norm": 0.1398439258337021, + "learning_rate": 3.986138344286875e-05, + "loss": 3.1762, + "step": 11918 + }, + { + "epoch": 0.07461760022209456, + "grad_norm": 0.14134135842323303, + "learning_rate": 3.9859218696901896e-05, + "loss": 3.1469, + "step": 11919 + }, + { + "epoch": 0.07464844668766797, + "grad_norm": 0.1527274250984192, + "learning_rate": 3.985705377865133e-05, + "loss": 3.1577, + "step": 11920 + }, + { + "epoch": 0.07467929315324139, + "grad_norm": 0.14207379519939423, + "learning_rate": 3.9854888688142154e-05, + "loss": 3.1274, + "step": 11921 + }, + { + "epoch": 0.0747101396188148, + "grad_norm": 0.14512032270431519, + "learning_rate": 3.985272342539948e-05, + "loss": 3.1773, + "step": 11922 + }, + { + "epoch": 0.07474098608438821, + "grad_norm": 0.14740945398807526, + "learning_rate": 3.985055799044838e-05, + "loss": 3.1624, + "step": 11923 + }, + { + "epoch": 0.07477183254996163, + "grad_norm": 0.17663025856018066, + "learning_rate": 3.9848392383314e-05, + "loss": 3.167, + "step": 11924 + }, + { + "epoch": 0.07480267901553506, + "grad_norm": 0.14766629040241241, + "learning_rate": 3.9846226604021433e-05, + "loss": 3.1526, + "step": 11925 + }, + { + "epoch": 0.07483352548110847, + "grad_norm": 0.14954420924186707, + "learning_rate": 3.9844060652595786e-05, + "loss": 3.1499, + "step": 11926 + }, + { + "epoch": 0.07486437194668188, + "grad_norm": 0.14682596921920776, + "learning_rate": 3.9841894529062185e-05, + "loss": 3.1315, + "step": 11927 + }, + { + "epoch": 0.0748952184122553, + "grad_norm": 0.15372976660728455, + "learning_rate": 3.983972823344573e-05, + "loss": 3.166, + "step": 11928 + }, + { + "epoch": 0.07492606487782871, + "grad_norm": 0.15672817826271057, + "learning_rate": 3.983756176577153e-05, + "loss": 3.1498, + "step": 11929 + }, + { + "epoch": 0.07495691134340213, + "grad_norm": 0.14864137768745422, + "learning_rate": 3.983539512606474e-05, + "loss": 3.1532, + "step": 11930 + }, + { + "epoch": 0.07498775780897556, + "grad_norm": 0.14689353108406067, + "learning_rate": 3.9833228314350454e-05, + "loss": 3.1676, + "step": 11931 + }, + { + "epoch": 0.07501860427454897, + "grad_norm": 0.15202303230762482, + "learning_rate": 3.9831061330653784e-05, + "loss": 3.1954, + "step": 11932 + }, + { + "epoch": 0.07504945074012238, + "grad_norm": 0.1499643623828888, + "learning_rate": 3.9828894174999884e-05, + "loss": 3.1622, + "step": 11933 + }, + { + "epoch": 0.0750802972056958, + "grad_norm": 0.14987196028232574, + "learning_rate": 3.982672684741387e-05, + "loss": 3.1542, + "step": 11934 + }, + { + "epoch": 0.07511114367126921, + "grad_norm": 0.1564767211675644, + "learning_rate": 3.9824559347920856e-05, + "loss": 3.155, + "step": 11935 + }, + { + "epoch": 0.07514199013684263, + "grad_norm": 0.15317422151565552, + "learning_rate": 3.9822391676546e-05, + "loss": 3.174, + "step": 11936 + }, + { + "epoch": 0.07517283660241605, + "grad_norm": 0.14520595967769623, + "learning_rate": 3.982022383331441e-05, + "loss": 3.1755, + "step": 11937 + }, + { + "epoch": 0.07520368306798947, + "grad_norm": 0.15202349424362183, + "learning_rate": 3.981805581825123e-05, + "loss": 3.1602, + "step": 11938 + }, + { + "epoch": 0.07523452953356288, + "grad_norm": 0.14579667150974274, + "learning_rate": 3.98158876313816e-05, + "loss": 3.1693, + "step": 11939 + }, + { + "epoch": 0.0752653759991363, + "grad_norm": 0.14961272478103638, + "learning_rate": 3.981371927273066e-05, + "loss": 3.109, + "step": 11940 + }, + { + "epoch": 0.07529622246470971, + "grad_norm": 0.15152508020401, + "learning_rate": 3.9811550742323534e-05, + "loss": 3.1916, + "step": 11941 + }, + { + "epoch": 0.07532706893028313, + "grad_norm": 0.15406520664691925, + "learning_rate": 3.980938204018538e-05, + "loss": 3.153, + "step": 11942 + }, + { + "epoch": 0.07535791539585655, + "grad_norm": 0.15544483065605164, + "learning_rate": 3.980721316634135e-05, + "loss": 3.1952, + "step": 11943 + }, + { + "epoch": 0.07538876186142997, + "grad_norm": 0.14945250749588013, + "learning_rate": 3.980504412081658e-05, + "loss": 3.1678, + "step": 11944 + }, + { + "epoch": 0.07541960832700338, + "grad_norm": 0.1558958739042282, + "learning_rate": 3.980287490363622e-05, + "loss": 3.1913, + "step": 11945 + }, + { + "epoch": 0.0754504547925768, + "grad_norm": 0.15428215265274048, + "learning_rate": 3.980070551482542e-05, + "loss": 3.1682, + "step": 11946 + }, + { + "epoch": 0.07548130125815021, + "grad_norm": 0.15026918053627014, + "learning_rate": 3.979853595440933e-05, + "loss": 3.1737, + "step": 11947 + }, + { + "epoch": 0.07551214772372362, + "grad_norm": 0.15234488248825073, + "learning_rate": 3.979636622241311e-05, + "loss": 3.187, + "step": 11948 + }, + { + "epoch": 0.07554299418929705, + "grad_norm": 0.15031757950782776, + "learning_rate": 3.979419631886192e-05, + "loss": 3.1411, + "step": 11949 + }, + { + "epoch": 0.07557384065487047, + "grad_norm": 0.14722934365272522, + "learning_rate": 3.979202624378091e-05, + "loss": 3.1652, + "step": 11950 + }, + { + "epoch": 0.07560468712044388, + "grad_norm": 0.14666782319545746, + "learning_rate": 3.9789855997195247e-05, + "loss": 3.1804, + "step": 11951 + }, + { + "epoch": 0.0756355335860173, + "grad_norm": 0.1529426872730255, + "learning_rate": 3.978768557913009e-05, + "loss": 3.1774, + "step": 11952 + }, + { + "epoch": 0.07566638005159071, + "grad_norm": 0.14651522040367126, + "learning_rate": 3.978551498961061e-05, + "loss": 3.1727, + "step": 11953 + }, + { + "epoch": 0.07569722651716412, + "grad_norm": 0.14844520390033722, + "learning_rate": 3.9783344228661954e-05, + "loss": 3.1398, + "step": 11954 + }, + { + "epoch": 0.07572807298273755, + "grad_norm": 0.1491088718175888, + "learning_rate": 3.978117329630932e-05, + "loss": 3.1985, + "step": 11955 + }, + { + "epoch": 0.07575891944831097, + "grad_norm": 0.15020184218883514, + "learning_rate": 3.977900219257785e-05, + "loss": 3.1882, + "step": 11956 + }, + { + "epoch": 0.07578976591388438, + "grad_norm": 0.14879873394966125, + "learning_rate": 3.977683091749275e-05, + "loss": 3.1636, + "step": 11957 + }, + { + "epoch": 0.0758206123794578, + "grad_norm": 0.14837013185024261, + "learning_rate": 3.977465947107917e-05, + "loss": 3.1334, + "step": 11958 + }, + { + "epoch": 0.07585145884503121, + "grad_norm": 0.15355397760868073, + "learning_rate": 3.977248785336228e-05, + "loss": 3.1678, + "step": 11959 + }, + { + "epoch": 0.07588230531060462, + "grad_norm": 0.14605656266212463, + "learning_rate": 3.977031606436728e-05, + "loss": 3.1904, + "step": 11960 + }, + { + "epoch": 0.07591315177617805, + "grad_norm": 0.1407882422208786, + "learning_rate": 3.976814410411934e-05, + "loss": 3.1394, + "step": 11961 + }, + { + "epoch": 0.07594399824175146, + "grad_norm": 0.15933029353618622, + "learning_rate": 3.976597197264364e-05, + "loss": 3.1752, + "step": 11962 + }, + { + "epoch": 0.07597484470732488, + "grad_norm": 0.15821129083633423, + "learning_rate": 3.9763799669965376e-05, + "loss": 3.1475, + "step": 11963 + }, + { + "epoch": 0.0760056911728983, + "grad_norm": 0.16250839829444885, + "learning_rate": 3.976162719610972e-05, + "loss": 3.1436, + "step": 11964 + }, + { + "epoch": 0.07603653763847171, + "grad_norm": 0.15734180808067322, + "learning_rate": 3.975945455110187e-05, + "loss": 3.1599, + "step": 11965 + }, + { + "epoch": 0.07606738410404512, + "grad_norm": 0.16273939609527588, + "learning_rate": 3.975728173496701e-05, + "loss": 3.1758, + "step": 11966 + }, + { + "epoch": 0.07609823056961855, + "grad_norm": 0.15107591450214386, + "learning_rate": 3.975510874773034e-05, + "loss": 3.1183, + "step": 11967 + }, + { + "epoch": 0.07612907703519196, + "grad_norm": 0.1612277776002884, + "learning_rate": 3.975293558941705e-05, + "loss": 3.18, + "step": 11968 + }, + { + "epoch": 0.07615992350076538, + "grad_norm": 0.1570100486278534, + "learning_rate": 3.975076226005234e-05, + "loss": 3.1441, + "step": 11969 + }, + { + "epoch": 0.07619076996633879, + "grad_norm": 0.15001501142978668, + "learning_rate": 3.97485887596614e-05, + "loss": 3.1512, + "step": 11970 + }, + { + "epoch": 0.0762216164319122, + "grad_norm": 0.15038219094276428, + "learning_rate": 3.974641508826944e-05, + "loss": 3.1296, + "step": 11971 + }, + { + "epoch": 0.07625246289748562, + "grad_norm": 0.15038347244262695, + "learning_rate": 3.974424124590166e-05, + "loss": 3.1974, + "step": 11972 + }, + { + "epoch": 0.07628330936305905, + "grad_norm": 0.14887486398220062, + "learning_rate": 3.974206723258326e-05, + "loss": 3.1509, + "step": 11973 + }, + { + "epoch": 0.07631415582863246, + "grad_norm": 0.15115614235401154, + "learning_rate": 3.973989304833945e-05, + "loss": 3.165, + "step": 11974 + }, + { + "epoch": 0.07634500229420588, + "grad_norm": 0.14017610251903534, + "learning_rate": 3.973771869319543e-05, + "loss": 3.1071, + "step": 11975 + }, + { + "epoch": 0.07637584875977929, + "grad_norm": 0.15414562821388245, + "learning_rate": 3.973554416717643e-05, + "loss": 3.1556, + "step": 11976 + }, + { + "epoch": 0.0764066952253527, + "grad_norm": 0.14623452723026276, + "learning_rate": 3.9733369470307654e-05, + "loss": 3.1457, + "step": 11977 + }, + { + "epoch": 0.07643754169092613, + "grad_norm": 0.14785131812095642, + "learning_rate": 3.97311946026143e-05, + "loss": 3.1888, + "step": 11978 + }, + { + "epoch": 0.07646838815649955, + "grad_norm": 0.14812499284744263, + "learning_rate": 3.97290195641216e-05, + "loss": 3.1653, + "step": 11979 + }, + { + "epoch": 0.07649923462207296, + "grad_norm": 0.1447620689868927, + "learning_rate": 3.972684435485478e-05, + "loss": 3.1381, + "step": 11980 + }, + { + "epoch": 0.07653008108764638, + "grad_norm": 0.14914605021476746, + "learning_rate": 3.972466897483904e-05, + "loss": 3.1543, + "step": 11981 + }, + { + "epoch": 0.07656092755321979, + "grad_norm": 0.14873558282852173, + "learning_rate": 3.972249342409961e-05, + "loss": 3.1653, + "step": 11982 + }, + { + "epoch": 0.0765917740187932, + "grad_norm": 0.14600461721420288, + "learning_rate": 3.9720317702661724e-05, + "loss": 3.1706, + "step": 11983 + }, + { + "epoch": 0.07662262048436663, + "grad_norm": 0.1460447609424591, + "learning_rate": 3.971814181055059e-05, + "loss": 3.1786, + "step": 11984 + }, + { + "epoch": 0.07665346694994005, + "grad_norm": 0.15419307351112366, + "learning_rate": 3.971596574779145e-05, + "loss": 3.1495, + "step": 11985 + }, + { + "epoch": 0.07668431341551346, + "grad_norm": 0.14279410243034363, + "learning_rate": 3.9713789514409535e-05, + "loss": 3.1694, + "step": 11986 + }, + { + "epoch": 0.07671515988108688, + "grad_norm": 0.1525508165359497, + "learning_rate": 3.971161311043007e-05, + "loss": 3.1949, + "step": 11987 + }, + { + "epoch": 0.07674600634666029, + "grad_norm": 0.15380647778511047, + "learning_rate": 3.970943653587829e-05, + "loss": 3.1262, + "step": 11988 + }, + { + "epoch": 0.0767768528122337, + "grad_norm": 0.14169135689735413, + "learning_rate": 3.9707259790779436e-05, + "loss": 3.1667, + "step": 11989 + }, + { + "epoch": 0.07680769927780713, + "grad_norm": 0.1504848599433899, + "learning_rate": 3.970508287515874e-05, + "loss": 3.1723, + "step": 11990 + }, + { + "epoch": 0.07683854574338055, + "grad_norm": 0.14269639551639557, + "learning_rate": 3.970290578904144e-05, + "loss": 3.1394, + "step": 11991 + }, + { + "epoch": 0.07686939220895396, + "grad_norm": 0.14707127213478088, + "learning_rate": 3.970072853245279e-05, + "loss": 3.135, + "step": 11992 + }, + { + "epoch": 0.07690023867452737, + "grad_norm": 0.14690423011779785, + "learning_rate": 3.9698551105418027e-05, + "loss": 3.1173, + "step": 11993 + }, + { + "epoch": 0.07693108514010079, + "grad_norm": 0.14644254744052887, + "learning_rate": 3.9696373507962393e-05, + "loss": 3.1895, + "step": 11994 + }, + { + "epoch": 0.0769619316056742, + "grad_norm": 0.15062718093395233, + "learning_rate": 3.969419574011114e-05, + "loss": 3.154, + "step": 11995 + }, + { + "epoch": 0.07699277807124763, + "grad_norm": 0.14359869062900543, + "learning_rate": 3.9692017801889525e-05, + "loss": 3.1406, + "step": 11996 + }, + { + "epoch": 0.07702362453682104, + "grad_norm": 0.1522141844034195, + "learning_rate": 3.968983969332278e-05, + "loss": 3.1451, + "step": 11997 + }, + { + "epoch": 0.07705447100239446, + "grad_norm": 0.15086273849010468, + "learning_rate": 3.968766141443618e-05, + "loss": 3.2126, + "step": 11998 + }, + { + "epoch": 0.07708531746796787, + "grad_norm": 0.15069307386875153, + "learning_rate": 3.968548296525497e-05, + "loss": 3.1591, + "step": 11999 + }, + { + "epoch": 0.07711616393354129, + "grad_norm": 0.14525409042835236, + "learning_rate": 3.968330434580441e-05, + "loss": 3.1963, + "step": 12000 + }, + { + "epoch": 0.0771470103991147, + "grad_norm": 0.1510707437992096, + "learning_rate": 3.968112555610976e-05, + "loss": 3.1284, + "step": 12001 + }, + { + "epoch": 0.07717785686468813, + "grad_norm": 0.14906440675258636, + "learning_rate": 3.9678946596196286e-05, + "loss": 3.136, + "step": 12002 + }, + { + "epoch": 0.07720870333026154, + "grad_norm": 0.2292294204235077, + "learning_rate": 3.967676746608924e-05, + "loss": 3.1302, + "step": 12003 + }, + { + "epoch": 0.07723954979583496, + "grad_norm": 0.15643076598644257, + "learning_rate": 3.96745881658139e-05, + "loss": 3.1566, + "step": 12004 + }, + { + "epoch": 0.07727039626140837, + "grad_norm": 0.1475272923707962, + "learning_rate": 3.967240869539552e-05, + "loss": 3.1852, + "step": 12005 + }, + { + "epoch": 0.07730124272698179, + "grad_norm": 0.14588633179664612, + "learning_rate": 3.967022905485939e-05, + "loss": 3.165, + "step": 12006 + }, + { + "epoch": 0.0773320891925552, + "grad_norm": 0.14824502170085907, + "learning_rate": 3.966804924423076e-05, + "loss": 3.1419, + "step": 12007 + }, + { + "epoch": 0.07736293565812863, + "grad_norm": 0.1468818187713623, + "learning_rate": 3.966586926353492e-05, + "loss": 3.1371, + "step": 12008 + }, + { + "epoch": 0.07739378212370204, + "grad_norm": 0.14832599461078644, + "learning_rate": 3.966368911279713e-05, + "loss": 3.173, + "step": 12009 + }, + { + "epoch": 0.07742462858927546, + "grad_norm": 0.1472969651222229, + "learning_rate": 3.966150879204269e-05, + "loss": 3.1358, + "step": 12010 + }, + { + "epoch": 0.07745547505484887, + "grad_norm": 0.14777004718780518, + "learning_rate": 3.965932830129685e-05, + "loss": 3.1633, + "step": 12011 + }, + { + "epoch": 0.07748632152042229, + "grad_norm": 0.1443253606557846, + "learning_rate": 3.965714764058492e-05, + "loss": 3.1692, + "step": 12012 + }, + { + "epoch": 0.0775171679859957, + "grad_norm": 0.14414815604686737, + "learning_rate": 3.965496680993217e-05, + "loss": 3.19, + "step": 12013 + }, + { + "epoch": 0.07754801445156913, + "grad_norm": 0.1583535522222519, + "learning_rate": 3.965278580936388e-05, + "loss": 3.1899, + "step": 12014 + }, + { + "epoch": 0.07757886091714254, + "grad_norm": 0.14575733244419098, + "learning_rate": 3.965060463890535e-05, + "loss": 3.172, + "step": 12015 + }, + { + "epoch": 0.07760970738271596, + "grad_norm": 0.1561920940876007, + "learning_rate": 3.9648423298581866e-05, + "loss": 3.1583, + "step": 12016 + }, + { + "epoch": 0.07764055384828937, + "grad_norm": 0.15950903296470642, + "learning_rate": 3.9646241788418707e-05, + "loss": 3.1499, + "step": 12017 + }, + { + "epoch": 0.07767140031386278, + "grad_norm": 0.14806251227855682, + "learning_rate": 3.964406010844118e-05, + "loss": 3.1766, + "step": 12018 + }, + { + "epoch": 0.0777022467794362, + "grad_norm": 0.16380344331264496, + "learning_rate": 3.964187825867458e-05, + "loss": 3.1511, + "step": 12019 + }, + { + "epoch": 0.07773309324500963, + "grad_norm": 0.15704286098480225, + "learning_rate": 3.9639696239144196e-05, + "loss": 3.1638, + "step": 12020 + }, + { + "epoch": 0.07776393971058304, + "grad_norm": 0.15306781232357025, + "learning_rate": 3.963751404987533e-05, + "loss": 3.2068, + "step": 12021 + }, + { + "epoch": 0.07779478617615646, + "grad_norm": 0.15539780259132385, + "learning_rate": 3.963533169089329e-05, + "loss": 3.1765, + "step": 12022 + }, + { + "epoch": 0.07782563264172987, + "grad_norm": 0.15586483478546143, + "learning_rate": 3.963314916222337e-05, + "loss": 3.1137, + "step": 12023 + }, + { + "epoch": 0.07785647910730328, + "grad_norm": 0.15320336818695068, + "learning_rate": 3.963096646389088e-05, + "loss": 3.1396, + "step": 12024 + }, + { + "epoch": 0.0778873255728767, + "grad_norm": 0.15242324769496918, + "learning_rate": 3.962878359592113e-05, + "loss": 3.1537, + "step": 12025 + }, + { + "epoch": 0.07791817203845013, + "grad_norm": 0.14725859463214874, + "learning_rate": 3.9626600558339424e-05, + "loss": 3.1694, + "step": 12026 + }, + { + "epoch": 0.07794901850402354, + "grad_norm": 0.1622498482465744, + "learning_rate": 3.9624417351171064e-05, + "loss": 3.1677, + "step": 12027 + }, + { + "epoch": 0.07797986496959695, + "grad_norm": 0.1567217856645584, + "learning_rate": 3.962223397444138e-05, + "loss": 3.1605, + "step": 12028 + }, + { + "epoch": 0.07801071143517037, + "grad_norm": 0.15211187303066254, + "learning_rate": 3.9620050428175684e-05, + "loss": 3.1764, + "step": 12029 + }, + { + "epoch": 0.07804155790074378, + "grad_norm": 0.14875145256519318, + "learning_rate": 3.961786671239929e-05, + "loss": 3.1645, + "step": 12030 + }, + { + "epoch": 0.0780724043663172, + "grad_norm": 0.15275190770626068, + "learning_rate": 3.9615682827137504e-05, + "loss": 3.1918, + "step": 12031 + }, + { + "epoch": 0.07810325083189062, + "grad_norm": 0.14616864919662476, + "learning_rate": 3.961349877241567e-05, + "loss": 3.1669, + "step": 12032 + }, + { + "epoch": 0.07813409729746404, + "grad_norm": 0.15462492406368256, + "learning_rate": 3.961131454825909e-05, + "loss": 3.1532, + "step": 12033 + }, + { + "epoch": 0.07816494376303745, + "grad_norm": 0.14820800721645355, + "learning_rate": 3.960913015469311e-05, + "loss": 3.149, + "step": 12034 + }, + { + "epoch": 0.07819579022861087, + "grad_norm": 0.14669781923294067, + "learning_rate": 3.9606945591743036e-05, + "loss": 3.1662, + "step": 12035 + }, + { + "epoch": 0.07822663669418428, + "grad_norm": 0.14761456847190857, + "learning_rate": 3.960476085943421e-05, + "loss": 3.1193, + "step": 12036 + }, + { + "epoch": 0.0782574831597577, + "grad_norm": 0.14881329238414764, + "learning_rate": 3.9602575957791955e-05, + "loss": 3.1864, + "step": 12037 + }, + { + "epoch": 0.07828832962533112, + "grad_norm": 0.14501358568668365, + "learning_rate": 3.960039088684161e-05, + "loss": 3.1271, + "step": 12038 + }, + { + "epoch": 0.07831917609090454, + "grad_norm": 0.1487099528312683, + "learning_rate": 3.95982056466085e-05, + "loss": 3.2005, + "step": 12039 + }, + { + "epoch": 0.07835002255647795, + "grad_norm": 0.15469013154506683, + "learning_rate": 3.9596020237117976e-05, + "loss": 3.2261, + "step": 12040 + }, + { + "epoch": 0.07838086902205137, + "grad_norm": 0.14209797978401184, + "learning_rate": 3.959383465839536e-05, + "loss": 3.1351, + "step": 12041 + }, + { + "epoch": 0.07841171548762478, + "grad_norm": 0.14855502545833588, + "learning_rate": 3.9591648910466005e-05, + "loss": 3.146, + "step": 12042 + }, + { + "epoch": 0.0784425619531982, + "grad_norm": 0.14543548226356506, + "learning_rate": 3.958946299335526e-05, + "loss": 3.1444, + "step": 12043 + }, + { + "epoch": 0.07847340841877162, + "grad_norm": 0.149138405919075, + "learning_rate": 3.9587276907088444e-05, + "loss": 3.1789, + "step": 12044 + }, + { + "epoch": 0.07850425488434504, + "grad_norm": 0.14944396913051605, + "learning_rate": 3.958509065169093e-05, + "loss": 3.1698, + "step": 12045 + }, + { + "epoch": 0.07853510134991845, + "grad_norm": 0.14778709411621094, + "learning_rate": 3.958290422718805e-05, + "loss": 3.1752, + "step": 12046 + }, + { + "epoch": 0.07856594781549187, + "grad_norm": 0.15389375388622284, + "learning_rate": 3.9580717633605146e-05, + "loss": 3.1251, + "step": 12047 + }, + { + "epoch": 0.07859679428106528, + "grad_norm": 0.16429324448108673, + "learning_rate": 3.95785308709676e-05, + "loss": 3.192, + "step": 12048 + }, + { + "epoch": 0.0786276407466387, + "grad_norm": 0.1473710834980011, + "learning_rate": 3.957634393930074e-05, + "loss": 3.1628, + "step": 12049 + }, + { + "epoch": 0.07865848721221212, + "grad_norm": 0.15711656212806702, + "learning_rate": 3.957415683862994e-05, + "loss": 3.1668, + "step": 12050 + }, + { + "epoch": 0.07868933367778554, + "grad_norm": 0.15186043083667755, + "learning_rate": 3.957196956898054e-05, + "loss": 3.174, + "step": 12051 + }, + { + "epoch": 0.07872018014335895, + "grad_norm": 0.14707864820957184, + "learning_rate": 3.956978213037792e-05, + "loss": 3.1531, + "step": 12052 + }, + { + "epoch": 0.07875102660893236, + "grad_norm": 0.15176816284656525, + "learning_rate": 3.956759452284742e-05, + "loss": 3.1829, + "step": 12053 + }, + { + "epoch": 0.07878187307450578, + "grad_norm": 0.15250353515148163, + "learning_rate": 3.956540674641443e-05, + "loss": 3.162, + "step": 12054 + }, + { + "epoch": 0.07881271954007919, + "grad_norm": 0.1488851010799408, + "learning_rate": 3.956321880110428e-05, + "loss": 3.1367, + "step": 12055 + }, + { + "epoch": 0.07884356600565262, + "grad_norm": 0.14841192960739136, + "learning_rate": 3.956103068694238e-05, + "loss": 3.1899, + "step": 12056 + }, + { + "epoch": 0.07887441247122604, + "grad_norm": 0.14414732158184052, + "learning_rate": 3.9558842403954064e-05, + "loss": 3.128, + "step": 12057 + }, + { + "epoch": 0.07890525893679945, + "grad_norm": 0.14380432665348053, + "learning_rate": 3.955665395216473e-05, + "loss": 3.1513, + "step": 12058 + }, + { + "epoch": 0.07893610540237286, + "grad_norm": 0.14729620516300201, + "learning_rate": 3.955446533159973e-05, + "loss": 3.1499, + "step": 12059 + }, + { + "epoch": 0.07896695186794628, + "grad_norm": 0.1473388522863388, + "learning_rate": 3.9552276542284464e-05, + "loss": 3.1326, + "step": 12060 + }, + { + "epoch": 0.07899779833351969, + "grad_norm": 0.14970432221889496, + "learning_rate": 3.9550087584244286e-05, + "loss": 3.1617, + "step": 12061 + }, + { + "epoch": 0.07902864479909312, + "grad_norm": 0.15215617418289185, + "learning_rate": 3.9547898457504595e-05, + "loss": 3.1842, + "step": 12062 + }, + { + "epoch": 0.07905949126466653, + "grad_norm": 0.14575207233428955, + "learning_rate": 3.954570916209076e-05, + "loss": 3.1465, + "step": 12063 + }, + { + "epoch": 0.07909033773023995, + "grad_norm": 0.1514611840248108, + "learning_rate": 3.9543519698028167e-05, + "loss": 3.1833, + "step": 12064 + }, + { + "epoch": 0.07912118419581336, + "grad_norm": 0.14941641688346863, + "learning_rate": 3.9541330065342204e-05, + "loss": 3.1308, + "step": 12065 + }, + { + "epoch": 0.07915203066138678, + "grad_norm": 0.1476513147354126, + "learning_rate": 3.9539140264058264e-05, + "loss": 3.1797, + "step": 12066 + }, + { + "epoch": 0.07918287712696019, + "grad_norm": 0.1457676887512207, + "learning_rate": 3.9536950294201716e-05, + "loss": 3.1515, + "step": 12067 + }, + { + "epoch": 0.07921372359253362, + "grad_norm": 0.1517801135778427, + "learning_rate": 3.9534760155797965e-05, + "loss": 3.1834, + "step": 12068 + }, + { + "epoch": 0.07924457005810703, + "grad_norm": 0.14777958393096924, + "learning_rate": 3.9532569848872415e-05, + "loss": 3.1299, + "step": 12069 + }, + { + "epoch": 0.07927541652368045, + "grad_norm": 0.1553187072277069, + "learning_rate": 3.953037937345044e-05, + "loss": 3.1567, + "step": 12070 + }, + { + "epoch": 0.07930626298925386, + "grad_norm": 0.15105018019676208, + "learning_rate": 3.952818872955746e-05, + "loss": 3.1727, + "step": 12071 + }, + { + "epoch": 0.07933710945482728, + "grad_norm": 0.14818020164966583, + "learning_rate": 3.952599791721886e-05, + "loss": 3.1431, + "step": 12072 + }, + { + "epoch": 0.07936795592040069, + "grad_norm": 0.1485530585050583, + "learning_rate": 3.9523806936460035e-05, + "loss": 3.1795, + "step": 12073 + }, + { + "epoch": 0.07939880238597412, + "grad_norm": 0.145654559135437, + "learning_rate": 3.9521615787306397e-05, + "loss": 3.1384, + "step": 12074 + }, + { + "epoch": 0.07942964885154753, + "grad_norm": 0.15179958939552307, + "learning_rate": 3.9519424469783364e-05, + "loss": 3.1604, + "step": 12075 + }, + { + "epoch": 0.07946049531712095, + "grad_norm": 0.14454813301563263, + "learning_rate": 3.9517232983916313e-05, + "loss": 3.1795, + "step": 12076 + }, + { + "epoch": 0.07949134178269436, + "grad_norm": 0.14933879673480988, + "learning_rate": 3.951504132973068e-05, + "loss": 3.1643, + "step": 12077 + }, + { + "epoch": 0.07952218824826777, + "grad_norm": 0.14711976051330566, + "learning_rate": 3.951284950725186e-05, + "loss": 3.1568, + "step": 12078 + }, + { + "epoch": 0.07955303471384119, + "grad_norm": 0.14365342259407043, + "learning_rate": 3.9510657516505265e-05, + "loss": 3.1795, + "step": 12079 + }, + { + "epoch": 0.07958388117941462, + "grad_norm": 0.15282712876796722, + "learning_rate": 3.950846535751633e-05, + "loss": 3.1759, + "step": 12080 + }, + { + "epoch": 0.07961472764498803, + "grad_norm": 0.1471206098794937, + "learning_rate": 3.9506273030310455e-05, + "loss": 3.1599, + "step": 12081 + }, + { + "epoch": 0.07964557411056145, + "grad_norm": 0.15111353993415833, + "learning_rate": 3.9504080534913055e-05, + "loss": 3.1791, + "step": 12082 + }, + { + "epoch": 0.07967642057613486, + "grad_norm": 0.14382661879062653, + "learning_rate": 3.950188787134956e-05, + "loss": 3.1481, + "step": 12083 + }, + { + "epoch": 0.07970726704170827, + "grad_norm": 0.15144596993923187, + "learning_rate": 3.949969503964539e-05, + "loss": 3.1507, + "step": 12084 + }, + { + "epoch": 0.07973811350728169, + "grad_norm": 0.15294410288333893, + "learning_rate": 3.949750203982597e-05, + "loss": 3.2037, + "step": 12085 + }, + { + "epoch": 0.07976895997285512, + "grad_norm": 0.15026798844337463, + "learning_rate": 3.9495308871916725e-05, + "loss": 3.1637, + "step": 12086 + }, + { + "epoch": 0.07979980643842853, + "grad_norm": 0.14350321888923645, + "learning_rate": 3.949311553594308e-05, + "loss": 3.1511, + "step": 12087 + }, + { + "epoch": 0.07983065290400194, + "grad_norm": 0.14833179116249084, + "learning_rate": 3.949092203193048e-05, + "loss": 3.132, + "step": 12088 + }, + { + "epoch": 0.07986149936957536, + "grad_norm": 0.1504022479057312, + "learning_rate": 3.948872835990434e-05, + "loss": 3.1696, + "step": 12089 + }, + { + "epoch": 0.07989234583514877, + "grad_norm": 0.1446722149848938, + "learning_rate": 3.94865345198901e-05, + "loss": 3.1518, + "step": 12090 + }, + { + "epoch": 0.07992319230072219, + "grad_norm": 0.14825940132141113, + "learning_rate": 3.94843405119132e-05, + "loss": 3.1238, + "step": 12091 + }, + { + "epoch": 0.07995403876629562, + "grad_norm": 0.15441644191741943, + "learning_rate": 3.948214633599907e-05, + "loss": 3.1635, + "step": 12092 + }, + { + "epoch": 0.07998488523186903, + "grad_norm": 0.14441175758838654, + "learning_rate": 3.947995199217317e-05, + "loss": 3.1878, + "step": 12093 + }, + { + "epoch": 0.08001573169744244, + "grad_norm": 0.14950615167617798, + "learning_rate": 3.947775748046092e-05, + "loss": 3.1294, + "step": 12094 + }, + { + "epoch": 0.08004657816301586, + "grad_norm": 0.14137707650661469, + "learning_rate": 3.9475562800887766e-05, + "loss": 3.1493, + "step": 12095 + }, + { + "epoch": 0.08007742462858927, + "grad_norm": 0.14568723738193512, + "learning_rate": 3.947336795347917e-05, + "loss": 3.1594, + "step": 12096 + }, + { + "epoch": 0.08010827109416269, + "grad_norm": 0.14871999621391296, + "learning_rate": 3.947117293826056e-05, + "loss": 3.1267, + "step": 12097 + }, + { + "epoch": 0.08013911755973611, + "grad_norm": 0.14440980553627014, + "learning_rate": 3.946897775525741e-05, + "loss": 3.1556, + "step": 12098 + }, + { + "epoch": 0.08016996402530953, + "grad_norm": 0.14166291058063507, + "learning_rate": 3.946678240449515e-05, + "loss": 3.1575, + "step": 12099 + }, + { + "epoch": 0.08020081049088294, + "grad_norm": 0.14283011853694916, + "learning_rate": 3.946458688599924e-05, + "loss": 3.1635, + "step": 12100 + }, + { + "epoch": 0.08023165695645636, + "grad_norm": 0.14745251834392548, + "learning_rate": 3.9462391199795135e-05, + "loss": 3.1862, + "step": 12101 + }, + { + "epoch": 0.08026250342202977, + "grad_norm": 0.14434009790420532, + "learning_rate": 3.9460195345908305e-05, + "loss": 3.1734, + "step": 12102 + }, + { + "epoch": 0.08029334988760319, + "grad_norm": 0.15322166681289673, + "learning_rate": 3.945799932436419e-05, + "loss": 3.1697, + "step": 12103 + }, + { + "epoch": 0.08032419635317661, + "grad_norm": 0.14770382642745972, + "learning_rate": 3.9455803135188265e-05, + "loss": 3.1632, + "step": 12104 + }, + { + "epoch": 0.08035504281875003, + "grad_norm": 0.15428918600082397, + "learning_rate": 3.945360677840598e-05, + "loss": 3.1557, + "step": 12105 + }, + { + "epoch": 0.08038588928432344, + "grad_norm": 0.1440151482820511, + "learning_rate": 3.945141025404282e-05, + "loss": 3.1624, + "step": 12106 + }, + { + "epoch": 0.08041673574989686, + "grad_norm": 0.14973993599414825, + "learning_rate": 3.9449213562124244e-05, + "loss": 3.1428, + "step": 12107 + }, + { + "epoch": 0.08044758221547027, + "grad_norm": 0.15061447024345398, + "learning_rate": 3.944701670267572e-05, + "loss": 3.1808, + "step": 12108 + }, + { + "epoch": 0.08047842868104368, + "grad_norm": 0.15329121053218842, + "learning_rate": 3.944481967572271e-05, + "loss": 3.1571, + "step": 12109 + }, + { + "epoch": 0.08050927514661711, + "grad_norm": 0.1513589471578598, + "learning_rate": 3.944262248129071e-05, + "loss": 3.1434, + "step": 12110 + }, + { + "epoch": 0.08054012161219053, + "grad_norm": 0.15060997009277344, + "learning_rate": 3.944042511940517e-05, + "loss": 3.1373, + "step": 12111 + }, + { + "epoch": 0.08057096807776394, + "grad_norm": 0.145879328250885, + "learning_rate": 3.943822759009158e-05, + "loss": 3.1675, + "step": 12112 + }, + { + "epoch": 0.08060181454333736, + "grad_norm": 0.15799987316131592, + "learning_rate": 3.943602989337543e-05, + "loss": 3.1789, + "step": 12113 + }, + { + "epoch": 0.08063266100891077, + "grad_norm": 0.1505163460969925, + "learning_rate": 3.943383202928217e-05, + "loss": 3.1298, + "step": 12114 + }, + { + "epoch": 0.08066350747448418, + "grad_norm": 0.15374761819839478, + "learning_rate": 3.943163399783731e-05, + "loss": 3.1629, + "step": 12115 + }, + { + "epoch": 0.08069435394005761, + "grad_norm": 0.15057331323623657, + "learning_rate": 3.942943579906633e-05, + "loss": 3.1757, + "step": 12116 + }, + { + "epoch": 0.08072520040563103, + "grad_norm": 0.15918543934822083, + "learning_rate": 3.94272374329947e-05, + "loss": 3.1756, + "step": 12117 + }, + { + "epoch": 0.08075604687120444, + "grad_norm": 0.1481984406709671, + "learning_rate": 3.942503889964793e-05, + "loss": 3.135, + "step": 12118 + }, + { + "epoch": 0.08078689333677785, + "grad_norm": 0.14514173567295074, + "learning_rate": 3.94228401990515e-05, + "loss": 3.1557, + "step": 12119 + }, + { + "epoch": 0.08081773980235127, + "grad_norm": 0.1470140963792801, + "learning_rate": 3.942064133123091e-05, + "loss": 3.1703, + "step": 12120 + }, + { + "epoch": 0.08084858626792468, + "grad_norm": 0.1562604010105133, + "learning_rate": 3.941844229621165e-05, + "loss": 3.171, + "step": 12121 + }, + { + "epoch": 0.08087943273349811, + "grad_norm": 0.14730344712734222, + "learning_rate": 3.941624309401921e-05, + "loss": 3.1798, + "step": 12122 + }, + { + "epoch": 0.08091027919907152, + "grad_norm": 0.14471140503883362, + "learning_rate": 3.9414043724679096e-05, + "loss": 3.1798, + "step": 12123 + }, + { + "epoch": 0.08094112566464494, + "grad_norm": 0.15131933987140656, + "learning_rate": 3.9411844188216814e-05, + "loss": 3.1714, + "step": 12124 + }, + { + "epoch": 0.08097197213021835, + "grad_norm": 0.24471771717071533, + "learning_rate": 3.940964448465785e-05, + "loss": 3.1467, + "step": 12125 + }, + { + "epoch": 0.08100281859579177, + "grad_norm": 0.1553565263748169, + "learning_rate": 3.940744461402772e-05, + "loss": 3.1649, + "step": 12126 + }, + { + "epoch": 0.08103366506136518, + "grad_norm": 0.14674822986125946, + "learning_rate": 3.940524457635193e-05, + "loss": 3.1791, + "step": 12127 + }, + { + "epoch": 0.08106451152693861, + "grad_norm": 0.15298700332641602, + "learning_rate": 3.940304437165598e-05, + "loss": 3.1667, + "step": 12128 + }, + { + "epoch": 0.08109535799251202, + "grad_norm": 0.1438097208738327, + "learning_rate": 3.940084399996539e-05, + "loss": 3.1675, + "step": 12129 + }, + { + "epoch": 0.08112620445808544, + "grad_norm": 0.14508363604545593, + "learning_rate": 3.939864346130567e-05, + "loss": 3.1999, + "step": 12130 + }, + { + "epoch": 0.08115705092365885, + "grad_norm": 0.16537851095199585, + "learning_rate": 3.9396442755702334e-05, + "loss": 3.1636, + "step": 12131 + }, + { + "epoch": 0.08118789738923227, + "grad_norm": 0.14015845954418182, + "learning_rate": 3.9394241883180887e-05, + "loss": 3.1569, + "step": 12132 + }, + { + "epoch": 0.08121874385480568, + "grad_norm": 0.14442379772663116, + "learning_rate": 3.9392040843766864e-05, + "loss": 3.1367, + "step": 12133 + }, + { + "epoch": 0.08124959032037911, + "grad_norm": 0.15087077021598816, + "learning_rate": 3.938983963748577e-05, + "loss": 3.1837, + "step": 12134 + }, + { + "epoch": 0.08128043678595252, + "grad_norm": 0.14773418009281158, + "learning_rate": 3.938763826436313e-05, + "loss": 3.1215, + "step": 12135 + }, + { + "epoch": 0.08131128325152594, + "grad_norm": 0.1463237851858139, + "learning_rate": 3.938543672442448e-05, + "loss": 3.167, + "step": 12136 + }, + { + "epoch": 0.08134212971709935, + "grad_norm": 0.14903485774993896, + "learning_rate": 3.938323501769533e-05, + "loss": 3.1682, + "step": 12137 + }, + { + "epoch": 0.08137297618267277, + "grad_norm": 0.15358203649520874, + "learning_rate": 3.9381033144201205e-05, + "loss": 3.1427, + "step": 12138 + }, + { + "epoch": 0.08140382264824618, + "grad_norm": 0.14634060859680176, + "learning_rate": 3.937883110396765e-05, + "loss": 3.1417, + "step": 12139 + }, + { + "epoch": 0.08143466911381961, + "grad_norm": 0.14667248725891113, + "learning_rate": 3.93766288970202e-05, + "loss": 3.147, + "step": 12140 + }, + { + "epoch": 0.08146551557939302, + "grad_norm": 0.143588125705719, + "learning_rate": 3.937442652338436e-05, + "loss": 3.159, + "step": 12141 + }, + { + "epoch": 0.08149636204496644, + "grad_norm": 0.1456041783094406, + "learning_rate": 3.937222398308569e-05, + "loss": 3.1684, + "step": 12142 + }, + { + "epoch": 0.08152720851053985, + "grad_norm": 0.14549501240253448, + "learning_rate": 3.9370021276149724e-05, + "loss": 3.1647, + "step": 12143 + }, + { + "epoch": 0.08155805497611326, + "grad_norm": 0.14447425305843353, + "learning_rate": 3.9367818402601985e-05, + "loss": 3.1164, + "step": 12144 + }, + { + "epoch": 0.08158890144168668, + "grad_norm": 0.14355722069740295, + "learning_rate": 3.936561536246803e-05, + "loss": 3.1517, + "step": 12145 + }, + { + "epoch": 0.0816197479072601, + "grad_norm": 0.14409196376800537, + "learning_rate": 3.936341215577341e-05, + "loss": 3.179, + "step": 12146 + }, + { + "epoch": 0.08165059437283352, + "grad_norm": 0.15455450117588043, + "learning_rate": 3.936120878254365e-05, + "loss": 3.1311, + "step": 12147 + }, + { + "epoch": 0.08168144083840694, + "grad_norm": 0.14603158831596375, + "learning_rate": 3.935900524280429e-05, + "loss": 3.1537, + "step": 12148 + }, + { + "epoch": 0.08171228730398035, + "grad_norm": 0.16121560335159302, + "learning_rate": 3.935680153658091e-05, + "loss": 3.1711, + "step": 12149 + }, + { + "epoch": 0.08174313376955376, + "grad_norm": 0.14780129492282867, + "learning_rate": 3.935459766389904e-05, + "loss": 3.1418, + "step": 12150 + }, + { + "epoch": 0.08177398023512718, + "grad_norm": 0.15801234543323517, + "learning_rate": 3.9352393624784235e-05, + "loss": 3.1701, + "step": 12151 + }, + { + "epoch": 0.0818048267007006, + "grad_norm": 0.14722715318202972, + "learning_rate": 3.935018941926205e-05, + "loss": 3.1535, + "step": 12152 + }, + { + "epoch": 0.08183567316627402, + "grad_norm": 0.14892035722732544, + "learning_rate": 3.934798504735804e-05, + "loss": 3.1416, + "step": 12153 + }, + { + "epoch": 0.08186651963184743, + "grad_norm": 0.15781083703041077, + "learning_rate": 3.934578050909777e-05, + "loss": 3.1571, + "step": 12154 + }, + { + "epoch": 0.08189736609742085, + "grad_norm": 0.1512671411037445, + "learning_rate": 3.9343575804506796e-05, + "loss": 3.1571, + "step": 12155 + }, + { + "epoch": 0.08192821256299426, + "grad_norm": 0.18097084760665894, + "learning_rate": 3.934137093361068e-05, + "loss": 3.1458, + "step": 12156 + }, + { + "epoch": 0.08195905902856768, + "grad_norm": 0.1524074226617813, + "learning_rate": 3.933916589643498e-05, + "loss": 3.1521, + "step": 12157 + }, + { + "epoch": 0.0819899054941411, + "grad_norm": 0.1615990847349167, + "learning_rate": 3.9336960693005275e-05, + "loss": 3.1545, + "step": 12158 + }, + { + "epoch": 0.08202075195971452, + "grad_norm": 0.16539336740970612, + "learning_rate": 3.9334755323347115e-05, + "loss": 3.1832, + "step": 12159 + }, + { + "epoch": 0.08205159842528793, + "grad_norm": 0.15375690162181854, + "learning_rate": 3.9332549787486095e-05, + "loss": 3.1616, + "step": 12160 + }, + { + "epoch": 0.08208244489086135, + "grad_norm": 0.1543123424053192, + "learning_rate": 3.933034408544777e-05, + "loss": 3.1595, + "step": 12161 + }, + { + "epoch": 0.08211329135643476, + "grad_norm": 0.1480153650045395, + "learning_rate": 3.9328138217257714e-05, + "loss": 3.1526, + "step": 12162 + }, + { + "epoch": 0.08214413782200818, + "grad_norm": 0.1627649962902069, + "learning_rate": 3.932593218294151e-05, + "loss": 3.1846, + "step": 12163 + }, + { + "epoch": 0.0821749842875816, + "grad_norm": 0.31659895181655884, + "learning_rate": 3.932372598252472e-05, + "loss": 3.1516, + "step": 12164 + }, + { + "epoch": 0.08220583075315502, + "grad_norm": 0.16519151628017426, + "learning_rate": 3.9321519616032946e-05, + "loss": 3.2025, + "step": 12165 + }, + { + "epoch": 0.08223667721872843, + "grad_norm": 0.16081486642360687, + "learning_rate": 3.931931308349176e-05, + "loss": 3.1716, + "step": 12166 + }, + { + "epoch": 0.08226752368430185, + "grad_norm": 0.17209474742412567, + "learning_rate": 3.931710638492673e-05, + "loss": 3.1416, + "step": 12167 + }, + { + "epoch": 0.08229837014987526, + "grad_norm": 0.16761356592178345, + "learning_rate": 3.9314899520363465e-05, + "loss": 3.1936, + "step": 12168 + }, + { + "epoch": 0.08232921661544867, + "grad_norm": 0.15123409032821655, + "learning_rate": 3.931269248982754e-05, + "loss": 3.1796, + "step": 12169 + }, + { + "epoch": 0.0823600630810221, + "grad_norm": 0.15599805116653442, + "learning_rate": 3.931048529334455e-05, + "loss": 3.1617, + "step": 12170 + }, + { + "epoch": 0.08239090954659552, + "grad_norm": 0.15977127850055695, + "learning_rate": 3.930827793094008e-05, + "loss": 3.2073, + "step": 12171 + }, + { + "epoch": 0.08242175601216893, + "grad_norm": 0.15579372644424438, + "learning_rate": 3.9306070402639725e-05, + "loss": 3.1557, + "step": 12172 + }, + { + "epoch": 0.08245260247774235, + "grad_norm": 0.15540526807308197, + "learning_rate": 3.930386270846908e-05, + "loss": 3.1506, + "step": 12173 + }, + { + "epoch": 0.08248344894331576, + "grad_norm": 0.14918319880962372, + "learning_rate": 3.9301654848453744e-05, + "loss": 3.1692, + "step": 12174 + }, + { + "epoch": 0.08251429540888917, + "grad_norm": 0.1503492295742035, + "learning_rate": 3.9299446822619314e-05, + "loss": 3.1652, + "step": 12175 + }, + { + "epoch": 0.0825451418744626, + "grad_norm": 0.1507677584886551, + "learning_rate": 3.929723863099139e-05, + "loss": 3.1653, + "step": 12176 + }, + { + "epoch": 0.08257598834003602, + "grad_norm": 0.14915476739406586, + "learning_rate": 3.929503027359558e-05, + "loss": 3.1436, + "step": 12177 + }, + { + "epoch": 0.08260683480560943, + "grad_norm": 0.14653736352920532, + "learning_rate": 3.9292821750457485e-05, + "loss": 3.1359, + "step": 12178 + }, + { + "epoch": 0.08263768127118284, + "grad_norm": 0.1622752994298935, + "learning_rate": 3.92906130616027e-05, + "loss": 3.1793, + "step": 12179 + }, + { + "epoch": 0.08266852773675626, + "grad_norm": 0.15391482412815094, + "learning_rate": 3.928840420705685e-05, + "loss": 3.1634, + "step": 12180 + }, + { + "epoch": 0.08269937420232967, + "grad_norm": 0.15357938408851624, + "learning_rate": 3.9286195186845544e-05, + "loss": 3.1593, + "step": 12181 + }, + { + "epoch": 0.0827302206679031, + "grad_norm": 0.16409966349601746, + "learning_rate": 3.928398600099439e-05, + "loss": 3.1648, + "step": 12182 + }, + { + "epoch": 0.08276106713347652, + "grad_norm": 0.14688444137573242, + "learning_rate": 3.9281776649528996e-05, + "loss": 3.146, + "step": 12183 + }, + { + "epoch": 0.08279191359904993, + "grad_norm": 0.1452619731426239, + "learning_rate": 3.927956713247499e-05, + "loss": 3.1795, + "step": 12184 + }, + { + "epoch": 0.08282276006462334, + "grad_norm": 0.1498345583677292, + "learning_rate": 3.927735744985798e-05, + "loss": 3.1665, + "step": 12185 + }, + { + "epoch": 0.08285360653019676, + "grad_norm": 0.1432408094406128, + "learning_rate": 3.927514760170359e-05, + "loss": 3.1548, + "step": 12186 + }, + { + "epoch": 0.08288445299577017, + "grad_norm": 0.15074552595615387, + "learning_rate": 3.927293758803744e-05, + "loss": 3.1591, + "step": 12187 + }, + { + "epoch": 0.0829152994613436, + "grad_norm": 0.14384648203849792, + "learning_rate": 3.9270727408885156e-05, + "loss": 3.1372, + "step": 12188 + }, + { + "epoch": 0.08294614592691701, + "grad_norm": 0.15208275616168976, + "learning_rate": 3.926851706427237e-05, + "loss": 3.1798, + "step": 12189 + }, + { + "epoch": 0.08297699239249043, + "grad_norm": 0.14751465618610382, + "learning_rate": 3.92663065542247e-05, + "loss": 3.1534, + "step": 12190 + }, + { + "epoch": 0.08300783885806384, + "grad_norm": 0.1488429307937622, + "learning_rate": 3.926409587876777e-05, + "loss": 3.1566, + "step": 12191 + }, + { + "epoch": 0.08303868532363726, + "grad_norm": 0.14858390390872955, + "learning_rate": 3.926188503792723e-05, + "loss": 3.1792, + "step": 12192 + }, + { + "epoch": 0.08306953178921067, + "grad_norm": 0.1566113978624344, + "learning_rate": 3.9259674031728704e-05, + "loss": 3.1699, + "step": 12193 + }, + { + "epoch": 0.0831003782547841, + "grad_norm": 0.15523730218410492, + "learning_rate": 3.925746286019782e-05, + "loss": 3.1403, + "step": 12194 + }, + { + "epoch": 0.08313122472035751, + "grad_norm": 0.1457427442073822, + "learning_rate": 3.925525152336023e-05, + "loss": 3.1366, + "step": 12195 + }, + { + "epoch": 0.08316207118593093, + "grad_norm": 0.15276560187339783, + "learning_rate": 3.925304002124156e-05, + "loss": 3.1575, + "step": 12196 + }, + { + "epoch": 0.08319291765150434, + "grad_norm": 0.14750823378562927, + "learning_rate": 3.925082835386745e-05, + "loss": 3.1635, + "step": 12197 + }, + { + "epoch": 0.08322376411707776, + "grad_norm": 0.2639191746711731, + "learning_rate": 3.924861652126356e-05, + "loss": 3.1279, + "step": 12198 + }, + { + "epoch": 0.08325461058265117, + "grad_norm": 0.1518690586090088, + "learning_rate": 3.924640452345552e-05, + "loss": 3.1614, + "step": 12199 + }, + { + "epoch": 0.0832854570482246, + "grad_norm": 0.15292580425739288, + "learning_rate": 3.9244192360468976e-05, + "loss": 3.1638, + "step": 12200 + }, + { + "epoch": 0.08331630351379801, + "grad_norm": 0.14802752435207367, + "learning_rate": 3.924198003232958e-05, + "loss": 3.1382, + "step": 12201 + }, + { + "epoch": 0.08334714997937143, + "grad_norm": 0.14868274331092834, + "learning_rate": 3.9239767539063e-05, + "loss": 3.1511, + "step": 12202 + }, + { + "epoch": 0.08337799644494484, + "grad_norm": 0.16218288242816925, + "learning_rate": 3.923755488069486e-05, + "loss": 3.1389, + "step": 12203 + }, + { + "epoch": 0.08340884291051825, + "grad_norm": 0.15044429898262024, + "learning_rate": 3.923534205725082e-05, + "loss": 3.148, + "step": 12204 + }, + { + "epoch": 0.08343968937609167, + "grad_norm": 0.14387641847133636, + "learning_rate": 3.923312906875656e-05, + "loss": 3.1753, + "step": 12205 + }, + { + "epoch": 0.0834705358416651, + "grad_norm": 0.14956147968769073, + "learning_rate": 3.923091591523771e-05, + "loss": 3.1499, + "step": 12206 + }, + { + "epoch": 0.08350138230723851, + "grad_norm": 0.1513216197490692, + "learning_rate": 3.9228702596719944e-05, + "loss": 3.1774, + "step": 12207 + }, + { + "epoch": 0.08353222877281193, + "grad_norm": 0.1452919840812683, + "learning_rate": 3.922648911322892e-05, + "loss": 3.1401, + "step": 12208 + }, + { + "epoch": 0.08356307523838534, + "grad_norm": 0.14804762601852417, + "learning_rate": 3.922427546479031e-05, + "loss": 3.1213, + "step": 12209 + }, + { + "epoch": 0.08359392170395875, + "grad_norm": 0.15121646225452423, + "learning_rate": 3.9222061651429775e-05, + "loss": 3.166, + "step": 12210 + }, + { + "epoch": 0.08362476816953217, + "grad_norm": 0.1569754183292389, + "learning_rate": 3.921984767317297e-05, + "loss": 3.1695, + "step": 12211 + }, + { + "epoch": 0.0836556146351056, + "grad_norm": 0.1490923911333084, + "learning_rate": 3.921763353004559e-05, + "loss": 3.1743, + "step": 12212 + }, + { + "epoch": 0.08368646110067901, + "grad_norm": 0.15261787176132202, + "learning_rate": 3.921541922207329e-05, + "loss": 3.1576, + "step": 12213 + }, + { + "epoch": 0.08371730756625242, + "grad_norm": 0.1435520201921463, + "learning_rate": 3.921320474928174e-05, + "loss": 3.1675, + "step": 12214 + }, + { + "epoch": 0.08374815403182584, + "grad_norm": 0.15125608444213867, + "learning_rate": 3.921099011169662e-05, + "loss": 3.1483, + "step": 12215 + }, + { + "epoch": 0.08377900049739925, + "grad_norm": 0.1497819870710373, + "learning_rate": 3.920877530934362e-05, + "loss": 3.1758, + "step": 12216 + }, + { + "epoch": 0.08380984696297267, + "grad_norm": 0.1483977735042572, + "learning_rate": 3.92065603422484e-05, + "loss": 3.1893, + "step": 12217 + }, + { + "epoch": 0.0838406934285461, + "grad_norm": 0.14840328693389893, + "learning_rate": 3.920434521043665e-05, + "loss": 3.1916, + "step": 12218 + }, + { + "epoch": 0.08387153989411951, + "grad_norm": 0.1533944010734558, + "learning_rate": 3.9202129913934064e-05, + "loss": 3.1198, + "step": 12219 + }, + { + "epoch": 0.08390238635969292, + "grad_norm": 0.15801455080509186, + "learning_rate": 3.9199914452766306e-05, + "loss": 3.14, + "step": 12220 + }, + { + "epoch": 0.08393323282526634, + "grad_norm": 0.15174169838428497, + "learning_rate": 3.919769882695908e-05, + "loss": 3.1554, + "step": 12221 + }, + { + "epoch": 0.08396407929083975, + "grad_norm": 0.1552533209323883, + "learning_rate": 3.919548303653806e-05, + "loss": 3.1917, + "step": 12222 + }, + { + "epoch": 0.08399492575641317, + "grad_norm": 0.14865723252296448, + "learning_rate": 3.9193267081528953e-05, + "loss": 3.1567, + "step": 12223 + }, + { + "epoch": 0.0840257722219866, + "grad_norm": 0.150033101439476, + "learning_rate": 3.919105096195744e-05, + "loss": 3.1772, + "step": 12224 + }, + { + "epoch": 0.08405661868756001, + "grad_norm": 0.1540224254131317, + "learning_rate": 3.9188834677849215e-05, + "loss": 3.1365, + "step": 12225 + }, + { + "epoch": 0.08408746515313342, + "grad_norm": 0.15257763862609863, + "learning_rate": 3.9186618229229976e-05, + "loss": 3.1778, + "step": 12226 + }, + { + "epoch": 0.08411831161870684, + "grad_norm": 0.1483117938041687, + "learning_rate": 3.918440161612543e-05, + "loss": 3.1521, + "step": 12227 + }, + { + "epoch": 0.08414915808428025, + "grad_norm": 0.14636549353599548, + "learning_rate": 3.918218483856128e-05, + "loss": 3.1481, + "step": 12228 + }, + { + "epoch": 0.08418000454985367, + "grad_norm": 0.1531909704208374, + "learning_rate": 3.91799678965632e-05, + "loss": 3.1668, + "step": 12229 + }, + { + "epoch": 0.0842108510154271, + "grad_norm": 0.14819639921188354, + "learning_rate": 3.9177750790156934e-05, + "loss": 3.1572, + "step": 12230 + }, + { + "epoch": 0.08424169748100051, + "grad_norm": 0.15883497893810272, + "learning_rate": 3.9175533519368146e-05, + "loss": 3.1582, + "step": 12231 + }, + { + "epoch": 0.08427254394657392, + "grad_norm": 0.14760933816432953, + "learning_rate": 3.917331608422258e-05, + "loss": 3.1533, + "step": 12232 + }, + { + "epoch": 0.08430339041214734, + "grad_norm": 0.14551804959774017, + "learning_rate": 3.917109848474593e-05, + "loss": 3.1347, + "step": 12233 + }, + { + "epoch": 0.08433423687772075, + "grad_norm": 0.15621426701545715, + "learning_rate": 3.91688807209639e-05, + "loss": 3.164, + "step": 12234 + }, + { + "epoch": 0.08436508334329416, + "grad_norm": 0.15155450999736786, + "learning_rate": 3.916666279290222e-05, + "loss": 3.182, + "step": 12235 + }, + { + "epoch": 0.08439592980886759, + "grad_norm": 0.14691247045993805, + "learning_rate": 3.9164444700586596e-05, + "loss": 3.1601, + "step": 12236 + }, + { + "epoch": 0.084426776274441, + "grad_norm": 0.1599595993757248, + "learning_rate": 3.916222644404275e-05, + "loss": 3.1422, + "step": 12237 + }, + { + "epoch": 0.08445762274001442, + "grad_norm": 0.14596259593963623, + "learning_rate": 3.916000802329639e-05, + "loss": 3.1291, + "step": 12238 + }, + { + "epoch": 0.08448846920558784, + "grad_norm": 0.15292181074619293, + "learning_rate": 3.915778943837325e-05, + "loss": 3.189, + "step": 12239 + }, + { + "epoch": 0.08451931567116125, + "grad_norm": 0.14362284541130066, + "learning_rate": 3.915557068929906e-05, + "loss": 3.1413, + "step": 12240 + }, + { + "epoch": 0.08455016213673466, + "grad_norm": 0.14388705790042877, + "learning_rate": 3.9153351776099524e-05, + "loss": 3.1594, + "step": 12241 + }, + { + "epoch": 0.08458100860230809, + "grad_norm": 0.1550915390253067, + "learning_rate": 3.9151132698800385e-05, + "loss": 3.1381, + "step": 12242 + }, + { + "epoch": 0.0846118550678815, + "grad_norm": 0.15239523351192474, + "learning_rate": 3.914891345742736e-05, + "loss": 3.162, + "step": 12243 + }, + { + "epoch": 0.08464270153345492, + "grad_norm": 0.1508013755083084, + "learning_rate": 3.914669405200619e-05, + "loss": 3.1669, + "step": 12244 + }, + { + "epoch": 0.08467354799902833, + "grad_norm": 0.15478262305259705, + "learning_rate": 3.91444744825626e-05, + "loss": 3.1313, + "step": 12245 + }, + { + "epoch": 0.08470439446460175, + "grad_norm": 0.16727037727832794, + "learning_rate": 3.9142254749122336e-05, + "loss": 3.1639, + "step": 12246 + }, + { + "epoch": 0.08473524093017516, + "grad_norm": 0.1662338227033615, + "learning_rate": 3.914003485171112e-05, + "loss": 3.1611, + "step": 12247 + }, + { + "epoch": 0.08476608739574859, + "grad_norm": 0.149321049451828, + "learning_rate": 3.91378147903547e-05, + "loss": 3.1249, + "step": 12248 + }, + { + "epoch": 0.084796933861322, + "grad_norm": 0.1541149616241455, + "learning_rate": 3.9135594565078815e-05, + "loss": 3.1531, + "step": 12249 + }, + { + "epoch": 0.08482778032689542, + "grad_norm": 0.15243464708328247, + "learning_rate": 3.9133374175909196e-05, + "loss": 3.1508, + "step": 12250 + }, + { + "epoch": 0.08485862679246883, + "grad_norm": 0.1560843586921692, + "learning_rate": 3.913115362287161e-05, + "loss": 3.1521, + "step": 12251 + }, + { + "epoch": 0.08488947325804225, + "grad_norm": 0.14978738129138947, + "learning_rate": 3.912893290599178e-05, + "loss": 3.1562, + "step": 12252 + }, + { + "epoch": 0.08492031972361568, + "grad_norm": 0.15581052005290985, + "learning_rate": 3.9126712025295465e-05, + "loss": 3.1844, + "step": 12253 + }, + { + "epoch": 0.08495116618918909, + "grad_norm": 0.14795242249965668, + "learning_rate": 3.912449098080841e-05, + "loss": 3.1522, + "step": 12254 + }, + { + "epoch": 0.0849820126547625, + "grad_norm": 0.15356099605560303, + "learning_rate": 3.9122269772556374e-05, + "loss": 3.1645, + "step": 12255 + }, + { + "epoch": 0.08501285912033592, + "grad_norm": 0.15339332818984985, + "learning_rate": 3.912004840056511e-05, + "loss": 3.1686, + "step": 12256 + }, + { + "epoch": 0.08504370558590933, + "grad_norm": 0.1448194980621338, + "learning_rate": 3.9117826864860366e-05, + "loss": 3.1184, + "step": 12257 + }, + { + "epoch": 0.08507455205148275, + "grad_norm": 0.15686824917793274, + "learning_rate": 3.911560516546791e-05, + "loss": 3.1674, + "step": 12258 + }, + { + "epoch": 0.08510539851705617, + "grad_norm": 0.14080828428268433, + "learning_rate": 3.911338330241348e-05, + "loss": 3.1591, + "step": 12259 + }, + { + "epoch": 0.08513624498262959, + "grad_norm": 0.14686445891857147, + "learning_rate": 3.9111161275722864e-05, + "loss": 3.151, + "step": 12260 + }, + { + "epoch": 0.085167091448203, + "grad_norm": 0.156638041138649, + "learning_rate": 3.910893908542181e-05, + "loss": 3.1666, + "step": 12261 + }, + { + "epoch": 0.08519793791377642, + "grad_norm": 0.15784090757369995, + "learning_rate": 3.910671673153609e-05, + "loss": 3.1627, + "step": 12262 + }, + { + "epoch": 0.08522878437934983, + "grad_norm": 0.17145942151546478, + "learning_rate": 3.9104494214091455e-05, + "loss": 3.1585, + "step": 12263 + }, + { + "epoch": 0.08525963084492325, + "grad_norm": 0.15334852039813995, + "learning_rate": 3.910227153311369e-05, + "loss": 3.1634, + "step": 12264 + }, + { + "epoch": 0.08529047731049667, + "grad_norm": 0.1549917608499527, + "learning_rate": 3.910004868862856e-05, + "loss": 3.1622, + "step": 12265 + }, + { + "epoch": 0.08532132377607009, + "grad_norm": 0.15995492041110992, + "learning_rate": 3.909782568066185e-05, + "loss": 3.1695, + "step": 12266 + }, + { + "epoch": 0.0853521702416435, + "grad_norm": 0.15492329001426697, + "learning_rate": 3.9095602509239316e-05, + "loss": 3.1922, + "step": 12267 + }, + { + "epoch": 0.08538301670721692, + "grad_norm": 0.1463651806116104, + "learning_rate": 3.909337917438673e-05, + "loss": 3.1648, + "step": 12268 + }, + { + "epoch": 0.08541386317279033, + "grad_norm": 0.15530070662498474, + "learning_rate": 3.90911556761299e-05, + "loss": 3.1638, + "step": 12269 + }, + { + "epoch": 0.08544470963836374, + "grad_norm": 0.15767298638820648, + "learning_rate": 3.9088932014494575e-05, + "loss": 3.1724, + "step": 12270 + }, + { + "epoch": 0.08547555610393717, + "grad_norm": 0.14927135407924652, + "learning_rate": 3.908670818950656e-05, + "loss": 3.1724, + "step": 12271 + }, + { + "epoch": 0.08550640256951059, + "grad_norm": 0.15365976095199585, + "learning_rate": 3.908448420119162e-05, + "loss": 3.1943, + "step": 12272 + }, + { + "epoch": 0.085537249035084, + "grad_norm": 0.14711715281009674, + "learning_rate": 3.9082260049575545e-05, + "loss": 3.1568, + "step": 12273 + }, + { + "epoch": 0.08556809550065742, + "grad_norm": 0.14835816621780396, + "learning_rate": 3.908003573468414e-05, + "loss": 3.1557, + "step": 12274 + }, + { + "epoch": 0.08559894196623083, + "grad_norm": 0.14426539838314056, + "learning_rate": 3.907781125654318e-05, + "loss": 3.126, + "step": 12275 + }, + { + "epoch": 0.08562978843180424, + "grad_norm": 0.144505575299263, + "learning_rate": 3.907558661517845e-05, + "loss": 3.1381, + "step": 12276 + }, + { + "epoch": 0.08566063489737767, + "grad_norm": 0.15372279286384583, + "learning_rate": 3.907336181061576e-05, + "loss": 3.1174, + "step": 12277 + }, + { + "epoch": 0.08569148136295109, + "grad_norm": 0.14298786222934723, + "learning_rate": 3.90711368428809e-05, + "loss": 3.1436, + "step": 12278 + }, + { + "epoch": 0.0857223278285245, + "grad_norm": 0.1462901085615158, + "learning_rate": 3.9068911711999644e-05, + "loss": 3.1549, + "step": 12279 + }, + { + "epoch": 0.08575317429409791, + "grad_norm": 0.15332892537117004, + "learning_rate": 3.9066686417997835e-05, + "loss": 3.1557, + "step": 12280 + }, + { + "epoch": 0.08578402075967133, + "grad_norm": 0.15398123860359192, + "learning_rate": 3.906446096090124e-05, + "loss": 3.0992, + "step": 12281 + }, + { + "epoch": 0.08581486722524474, + "grad_norm": 0.14632870256900787, + "learning_rate": 3.906223534073566e-05, + "loss": 3.1136, + "step": 12282 + }, + { + "epoch": 0.08584571369081817, + "grad_norm": 0.15203562378883362, + "learning_rate": 3.906000955752693e-05, + "loss": 3.1758, + "step": 12283 + }, + { + "epoch": 0.08587656015639158, + "grad_norm": 0.15501007437705994, + "learning_rate": 3.9057783611300836e-05, + "loss": 3.1301, + "step": 12284 + }, + { + "epoch": 0.085907406621965, + "grad_norm": 0.1517382115125656, + "learning_rate": 3.905555750208318e-05, + "loss": 3.184, + "step": 12285 + }, + { + "epoch": 0.08593825308753841, + "grad_norm": 0.14256703853607178, + "learning_rate": 3.9053331229899784e-05, + "loss": 3.1396, + "step": 12286 + }, + { + "epoch": 0.08596909955311183, + "grad_norm": 0.15107384324073792, + "learning_rate": 3.905110479477646e-05, + "loss": 3.1443, + "step": 12287 + }, + { + "epoch": 0.08599994601868524, + "grad_norm": 0.15302547812461853, + "learning_rate": 3.904887819673901e-05, + "loss": 3.1768, + "step": 12288 + }, + { + "epoch": 0.08603079248425867, + "grad_norm": 0.14822211861610413, + "learning_rate": 3.904665143581328e-05, + "loss": 3.1701, + "step": 12289 + }, + { + "epoch": 0.08606163894983208, + "grad_norm": 0.1432514488697052, + "learning_rate": 3.9044424512025054e-05, + "loss": 3.137, + "step": 12290 + }, + { + "epoch": 0.0860924854154055, + "grad_norm": 0.15344879031181335, + "learning_rate": 3.904219742540016e-05, + "loss": 3.1169, + "step": 12291 + }, + { + "epoch": 0.08612333188097891, + "grad_norm": 0.16776172816753387, + "learning_rate": 3.903997017596444e-05, + "loss": 3.1389, + "step": 12292 + }, + { + "epoch": 0.08615417834655233, + "grad_norm": 0.8937952518463135, + "learning_rate": 3.903774276374369e-05, + "loss": 3.1373, + "step": 12293 + }, + { + "epoch": 0.08618502481212574, + "grad_norm": 0.15324635803699493, + "learning_rate": 3.9035515188763756e-05, + "loss": 3.1481, + "step": 12294 + }, + { + "epoch": 0.08621587127769917, + "grad_norm": 0.175282821059227, + "learning_rate": 3.9033287451050454e-05, + "loss": 3.1586, + "step": 12295 + }, + { + "epoch": 0.08624671774327258, + "grad_norm": 0.14941421151161194, + "learning_rate": 3.9031059550629625e-05, + "loss": 3.1725, + "step": 12296 + }, + { + "epoch": 0.086277564208846, + "grad_norm": 0.15598273277282715, + "learning_rate": 3.902883148752708e-05, + "loss": 3.162, + "step": 12297 + }, + { + "epoch": 0.08630841067441941, + "grad_norm": 0.1487947553396225, + "learning_rate": 3.9026603261768666e-05, + "loss": 3.1227, + "step": 12298 + }, + { + "epoch": 0.08633925713999283, + "grad_norm": 0.16773205995559692, + "learning_rate": 3.9024374873380226e-05, + "loss": 3.1495, + "step": 12299 + }, + { + "epoch": 0.08637010360556624, + "grad_norm": 0.15757058560848236, + "learning_rate": 3.9022146322387576e-05, + "loss": 3.1275, + "step": 12300 + }, + { + "epoch": 0.08640095007113967, + "grad_norm": 0.1564418524503708, + "learning_rate": 3.9019917608816575e-05, + "loss": 3.1616, + "step": 12301 + }, + { + "epoch": 0.08643179653671308, + "grad_norm": 0.1663464456796646, + "learning_rate": 3.901768873269305e-05, + "loss": 3.175, + "step": 12302 + }, + { + "epoch": 0.0864626430022865, + "grad_norm": 0.15103696286678314, + "learning_rate": 3.9015459694042846e-05, + "loss": 3.1603, + "step": 12303 + }, + { + "epoch": 0.08649348946785991, + "grad_norm": 0.1683070957660675, + "learning_rate": 3.9013230492891816e-05, + "loss": 3.1665, + "step": 12304 + }, + { + "epoch": 0.08652433593343332, + "grad_norm": 0.15201516449451447, + "learning_rate": 3.901100112926579e-05, + "loss": 3.1542, + "step": 12305 + }, + { + "epoch": 0.08655518239900674, + "grad_norm": 0.1678793728351593, + "learning_rate": 3.900877160319063e-05, + "loss": 3.1407, + "step": 12306 + }, + { + "epoch": 0.08658602886458017, + "grad_norm": 0.14741788804531097, + "learning_rate": 3.900654191469218e-05, + "loss": 3.117, + "step": 12307 + }, + { + "epoch": 0.08661687533015358, + "grad_norm": 0.15932424366474152, + "learning_rate": 3.90043120637963e-05, + "loss": 3.1656, + "step": 12308 + }, + { + "epoch": 0.086647721795727, + "grad_norm": 0.1505887359380722, + "learning_rate": 3.900208205052883e-05, + "loss": 3.1754, + "step": 12309 + }, + { + "epoch": 0.08667856826130041, + "grad_norm": 0.1666785180568695, + "learning_rate": 3.899985187491564e-05, + "loss": 3.1525, + "step": 12310 + }, + { + "epoch": 0.08670941472687382, + "grad_norm": 0.16716121137142181, + "learning_rate": 3.899762153698257e-05, + "loss": 3.1579, + "step": 12311 + }, + { + "epoch": 0.08674026119244724, + "grad_norm": 0.15099574625492096, + "learning_rate": 3.899539103675549e-05, + "loss": 3.1563, + "step": 12312 + }, + { + "epoch": 0.08677110765802067, + "grad_norm": 0.1497052162885666, + "learning_rate": 3.8993160374260276e-05, + "loss": 3.1838, + "step": 12313 + }, + { + "epoch": 0.08680195412359408, + "grad_norm": 0.15523332357406616, + "learning_rate": 3.899092954952276e-05, + "loss": 3.1479, + "step": 12314 + }, + { + "epoch": 0.0868328005891675, + "grad_norm": 0.15555301308631897, + "learning_rate": 3.8988698562568834e-05, + "loss": 3.1679, + "step": 12315 + }, + { + "epoch": 0.08686364705474091, + "grad_norm": 0.15029415488243103, + "learning_rate": 3.8986467413424354e-05, + "loss": 3.1565, + "step": 12316 + }, + { + "epoch": 0.08689449352031432, + "grad_norm": 0.1475684642791748, + "learning_rate": 3.898423610211518e-05, + "loss": 3.1493, + "step": 12317 + }, + { + "epoch": 0.08692533998588774, + "grad_norm": 0.15113164484500885, + "learning_rate": 3.89820046286672e-05, + "loss": 3.1242, + "step": 12318 + }, + { + "epoch": 0.08695618645146116, + "grad_norm": 0.148415669798851, + "learning_rate": 3.897977299310628e-05, + "loss": 3.1699, + "step": 12319 + }, + { + "epoch": 0.08698703291703458, + "grad_norm": 0.15903963148593903, + "learning_rate": 3.897754119545828e-05, + "loss": 3.1274, + "step": 12320 + }, + { + "epoch": 0.087017879382608, + "grad_norm": 0.15406931936740875, + "learning_rate": 3.89753092357491e-05, + "loss": 3.1816, + "step": 12321 + }, + { + "epoch": 0.08704872584818141, + "grad_norm": 0.15030232071876526, + "learning_rate": 3.8973077114004616e-05, + "loss": 3.177, + "step": 12322 + }, + { + "epoch": 0.08707957231375482, + "grad_norm": 0.15820656716823578, + "learning_rate": 3.897084483025069e-05, + "loss": 3.1513, + "step": 12323 + }, + { + "epoch": 0.08711041877932824, + "grad_norm": 0.1451093554496765, + "learning_rate": 3.896861238451321e-05, + "loss": 3.1575, + "step": 12324 + }, + { + "epoch": 0.08714126524490166, + "grad_norm": 0.1465972512960434, + "learning_rate": 3.8966379776818066e-05, + "loss": 3.1463, + "step": 12325 + }, + { + "epoch": 0.08717211171047508, + "grad_norm": 0.15534500777721405, + "learning_rate": 3.8964147007191144e-05, + "loss": 3.1219, + "step": 12326 + }, + { + "epoch": 0.08720295817604849, + "grad_norm": 0.15441656112670898, + "learning_rate": 3.896191407565833e-05, + "loss": 3.1117, + "step": 12327 + }, + { + "epoch": 0.0872338046416219, + "grad_norm": 0.1501666009426117, + "learning_rate": 3.895968098224551e-05, + "loss": 3.1577, + "step": 12328 + }, + { + "epoch": 0.08726465110719532, + "grad_norm": 0.1606348603963852, + "learning_rate": 3.895744772697858e-05, + "loss": 3.1282, + "step": 12329 + }, + { + "epoch": 0.08729549757276873, + "grad_norm": 0.15193474292755127, + "learning_rate": 3.895521430988343e-05, + "loss": 3.1647, + "step": 12330 + }, + { + "epoch": 0.08732634403834216, + "grad_norm": 0.15520797669887543, + "learning_rate": 3.8952980730985954e-05, + "loss": 3.1447, + "step": 12331 + }, + { + "epoch": 0.08735719050391558, + "grad_norm": 0.15503185987472534, + "learning_rate": 3.8950746990312045e-05, + "loss": 3.1373, + "step": 12332 + }, + { + "epoch": 0.08738803696948899, + "grad_norm": 0.1590525209903717, + "learning_rate": 3.894851308788762e-05, + "loss": 3.1812, + "step": 12333 + }, + { + "epoch": 0.0874188834350624, + "grad_norm": 0.1541002243757248, + "learning_rate": 3.8946279023738565e-05, + "loss": 3.1223, + "step": 12334 + }, + { + "epoch": 0.08744972990063582, + "grad_norm": 0.14641745388507843, + "learning_rate": 3.8944044797890785e-05, + "loss": 3.1747, + "step": 12335 + }, + { + "epoch": 0.08748057636620923, + "grad_norm": 0.1492997407913208, + "learning_rate": 3.8941810410370185e-05, + "loss": 3.1478, + "step": 12336 + }, + { + "epoch": 0.08751142283178266, + "grad_norm": 0.14848065376281738, + "learning_rate": 3.893957586120267e-05, + "loss": 3.1584, + "step": 12337 + }, + { + "epoch": 0.08754226929735608, + "grad_norm": 0.1467387080192566, + "learning_rate": 3.893734115041415e-05, + "loss": 3.1385, + "step": 12338 + }, + { + "epoch": 0.08757311576292949, + "grad_norm": 0.1434127539396286, + "learning_rate": 3.893510627803053e-05, + "loss": 3.133, + "step": 12339 + }, + { + "epoch": 0.0876039622285029, + "grad_norm": 0.1421196013689041, + "learning_rate": 3.8932871244077736e-05, + "loss": 3.1155, + "step": 12340 + }, + { + "epoch": 0.08763480869407632, + "grad_norm": 0.15113843977451324, + "learning_rate": 3.8930636048581664e-05, + "loss": 3.1869, + "step": 12341 + }, + { + "epoch": 0.08766565515964973, + "grad_norm": 0.15152092278003693, + "learning_rate": 3.892840069156825e-05, + "loss": 3.1725, + "step": 12342 + }, + { + "epoch": 0.08769650162522316, + "grad_norm": 0.14359845221042633, + "learning_rate": 3.892616517306339e-05, + "loss": 3.1528, + "step": 12343 + }, + { + "epoch": 0.08772734809079658, + "grad_norm": 0.1680515706539154, + "learning_rate": 3.892392949309301e-05, + "loss": 3.1616, + "step": 12344 + }, + { + "epoch": 0.08775819455636999, + "grad_norm": 0.14615944027900696, + "learning_rate": 3.892169365168304e-05, + "loss": 3.1566, + "step": 12345 + }, + { + "epoch": 0.0877890410219434, + "grad_norm": 0.15234333276748657, + "learning_rate": 3.8919457648859406e-05, + "loss": 3.1262, + "step": 12346 + }, + { + "epoch": 0.08781988748751682, + "grad_norm": 0.15712088346481323, + "learning_rate": 3.891722148464801e-05, + "loss": 3.1505, + "step": 12347 + }, + { + "epoch": 0.08785073395309023, + "grad_norm": 0.15166111290454865, + "learning_rate": 3.89149851590748e-05, + "loss": 3.1794, + "step": 12348 + }, + { + "epoch": 0.08788158041866366, + "grad_norm": 0.1600721925497055, + "learning_rate": 3.891274867216571e-05, + "loss": 3.1383, + "step": 12349 + }, + { + "epoch": 0.08791242688423707, + "grad_norm": 0.18535389006137848, + "learning_rate": 3.8910512023946644e-05, + "loss": 3.1464, + "step": 12350 + }, + { + "epoch": 0.08794327334981049, + "grad_norm": 0.14779949188232422, + "learning_rate": 3.8908275214443555e-05, + "loss": 3.1417, + "step": 12351 + }, + { + "epoch": 0.0879741198153839, + "grad_norm": 0.15157055854797363, + "learning_rate": 3.890603824368238e-05, + "loss": 3.1533, + "step": 12352 + }, + { + "epoch": 0.08800496628095732, + "grad_norm": 0.15228161215782166, + "learning_rate": 3.8903801111689034e-05, + "loss": 3.1677, + "step": 12353 + }, + { + "epoch": 0.08803581274653073, + "grad_norm": 0.15183979272842407, + "learning_rate": 3.890156381848948e-05, + "loss": 3.1682, + "step": 12354 + }, + { + "epoch": 0.08806665921210416, + "grad_norm": 0.14790470898151398, + "learning_rate": 3.889932636410965e-05, + "loss": 3.1631, + "step": 12355 + }, + { + "epoch": 0.08809750567767757, + "grad_norm": 0.1519516408443451, + "learning_rate": 3.8897088748575475e-05, + "loss": 3.1119, + "step": 12356 + }, + { + "epoch": 0.08812835214325099, + "grad_norm": 0.1455385833978653, + "learning_rate": 3.8894850971912915e-05, + "loss": 3.155, + "step": 12357 + }, + { + "epoch": 0.0881591986088244, + "grad_norm": 0.14876008033752441, + "learning_rate": 3.8892613034147906e-05, + "loss": 3.1619, + "step": 12358 + }, + { + "epoch": 0.08819004507439782, + "grad_norm": 0.15336713194847107, + "learning_rate": 3.889037493530639e-05, + "loss": 3.1609, + "step": 12359 + }, + { + "epoch": 0.08822089153997123, + "grad_norm": 0.14653384685516357, + "learning_rate": 3.888813667541433e-05, + "loss": 3.146, + "step": 12360 + }, + { + "epoch": 0.08825173800554466, + "grad_norm": 0.1551414430141449, + "learning_rate": 3.8885898254497664e-05, + "loss": 3.1326, + "step": 12361 + }, + { + "epoch": 0.08828258447111807, + "grad_norm": 0.14732372760772705, + "learning_rate": 3.888365967258236e-05, + "loss": 3.1212, + "step": 12362 + }, + { + "epoch": 0.08831343093669149, + "grad_norm": 0.1537230908870697, + "learning_rate": 3.888142092969437e-05, + "loss": 3.158, + "step": 12363 + }, + { + "epoch": 0.0883442774022649, + "grad_norm": 0.15245303511619568, + "learning_rate": 3.887918202585964e-05, + "loss": 3.1631, + "step": 12364 + }, + { + "epoch": 0.08837512386783831, + "grad_norm": 0.1420481950044632, + "learning_rate": 3.887694296110414e-05, + "loss": 3.1545, + "step": 12365 + }, + { + "epoch": 0.08840597033341173, + "grad_norm": 0.1484014242887497, + "learning_rate": 3.887470373545381e-05, + "loss": 3.1793, + "step": 12366 + }, + { + "epoch": 0.08843681679898516, + "grad_norm": 0.17964127659797668, + "learning_rate": 3.887246434893463e-05, + "loss": 3.1646, + "step": 12367 + }, + { + "epoch": 0.08846766326455857, + "grad_norm": 0.1494363248348236, + "learning_rate": 3.887022480157258e-05, + "loss": 3.1631, + "step": 12368 + }, + { + "epoch": 0.08849850973013199, + "grad_norm": 0.1498682051897049, + "learning_rate": 3.88679850933936e-05, + "loss": 3.1252, + "step": 12369 + }, + { + "epoch": 0.0885293561957054, + "grad_norm": 0.14606691896915436, + "learning_rate": 3.886574522442366e-05, + "loss": 3.1871, + "step": 12370 + }, + { + "epoch": 0.08856020266127881, + "grad_norm": 0.15282292664051056, + "learning_rate": 3.8863505194688746e-05, + "loss": 3.1603, + "step": 12371 + }, + { + "epoch": 0.08859104912685223, + "grad_norm": 0.14957021176815033, + "learning_rate": 3.886126500421481e-05, + "loss": 3.1603, + "step": 12372 + }, + { + "epoch": 0.08862189559242566, + "grad_norm": 0.15332460403442383, + "learning_rate": 3.8859024653027844e-05, + "loss": 3.1523, + "step": 12373 + }, + { + "epoch": 0.08865274205799907, + "grad_norm": 0.1499529331922531, + "learning_rate": 3.885678414115381e-05, + "loss": 3.1698, + "step": 12374 + }, + { + "epoch": 0.08868358852357248, + "grad_norm": 0.15909989178180695, + "learning_rate": 3.88545434686187e-05, + "loss": 3.1463, + "step": 12375 + }, + { + "epoch": 0.0887144349891459, + "grad_norm": 0.14318320155143738, + "learning_rate": 3.885230263544848e-05, + "loss": 3.1446, + "step": 12376 + }, + { + "epoch": 0.08874528145471931, + "grad_norm": 0.15737903118133545, + "learning_rate": 3.8850061641669134e-05, + "loss": 3.185, + "step": 12377 + }, + { + "epoch": 0.08877612792029273, + "grad_norm": 0.14539892971515656, + "learning_rate": 3.884782048730664e-05, + "loss": 3.1619, + "step": 12378 + }, + { + "epoch": 0.08880697438586616, + "grad_norm": 0.14936856925487518, + "learning_rate": 3.884557917238699e-05, + "loss": 3.1369, + "step": 12379 + }, + { + "epoch": 0.08883782085143957, + "grad_norm": 0.14727632701396942, + "learning_rate": 3.884333769693618e-05, + "loss": 3.1359, + "step": 12380 + }, + { + "epoch": 0.08886866731701298, + "grad_norm": 0.15094831585884094, + "learning_rate": 3.8841096060980184e-05, + "loss": 3.1322, + "step": 12381 + }, + { + "epoch": 0.0888995137825864, + "grad_norm": 0.15008987486362457, + "learning_rate": 3.883885426454499e-05, + "loss": 3.1205, + "step": 12382 + }, + { + "epoch": 0.08893036024815981, + "grad_norm": 0.15003585815429688, + "learning_rate": 3.88366123076566e-05, + "loss": 3.1829, + "step": 12383 + }, + { + "epoch": 0.08896120671373323, + "grad_norm": 0.15321287512779236, + "learning_rate": 3.8834370190341016e-05, + "loss": 3.1905, + "step": 12384 + }, + { + "epoch": 0.08899205317930665, + "grad_norm": 0.1496441811323166, + "learning_rate": 3.883212791262421e-05, + "loss": 3.1778, + "step": 12385 + }, + { + "epoch": 0.08902289964488007, + "grad_norm": 0.15178407728672028, + "learning_rate": 3.8829885474532195e-05, + "loss": 3.1986, + "step": 12386 + }, + { + "epoch": 0.08905374611045348, + "grad_norm": 0.142987921833992, + "learning_rate": 3.882764287609098e-05, + "loss": 3.1443, + "step": 12387 + }, + { + "epoch": 0.0890845925760269, + "grad_norm": 0.14911913871765137, + "learning_rate": 3.882540011732654e-05, + "loss": 3.1737, + "step": 12388 + }, + { + "epoch": 0.08911543904160031, + "grad_norm": 0.15498247742652893, + "learning_rate": 3.88231571982649e-05, + "loss": 3.1539, + "step": 12389 + }, + { + "epoch": 0.08914628550717373, + "grad_norm": 0.14496628940105438, + "learning_rate": 3.8820914118932063e-05, + "loss": 3.1438, + "step": 12390 + }, + { + "epoch": 0.08917713197274715, + "grad_norm": 0.14738601446151733, + "learning_rate": 3.881867087935402e-05, + "loss": 3.1616, + "step": 12391 + }, + { + "epoch": 0.08920797843832057, + "grad_norm": 0.1449604034423828, + "learning_rate": 3.881642747955681e-05, + "loss": 3.1562, + "step": 12392 + }, + { + "epoch": 0.08923882490389398, + "grad_norm": 0.14671577513217926, + "learning_rate": 3.881418391956642e-05, + "loss": 3.1538, + "step": 12393 + }, + { + "epoch": 0.0892696713694674, + "grad_norm": 0.14239320158958435, + "learning_rate": 3.881194019940886e-05, + "loss": 3.1425, + "step": 12394 + }, + { + "epoch": 0.08930051783504081, + "grad_norm": 0.14817914366722107, + "learning_rate": 3.8809696319110154e-05, + "loss": 3.12, + "step": 12395 + }, + { + "epoch": 0.08933136430061422, + "grad_norm": 0.1455736756324768, + "learning_rate": 3.880745227869633e-05, + "loss": 3.1229, + "step": 12396 + }, + { + "epoch": 0.08936221076618765, + "grad_norm": 0.1489296406507492, + "learning_rate": 3.880520807819339e-05, + "loss": 3.1474, + "step": 12397 + }, + { + "epoch": 0.08939305723176107, + "grad_norm": 0.14840972423553467, + "learning_rate": 3.880296371762735e-05, + "loss": 3.17, + "step": 12398 + }, + { + "epoch": 0.08942390369733448, + "grad_norm": 0.14373771846294403, + "learning_rate": 3.880071919702425e-05, + "loss": 3.1572, + "step": 12399 + }, + { + "epoch": 0.0894547501629079, + "grad_norm": 0.14759404957294464, + "learning_rate": 3.8798474516410096e-05, + "loss": 3.1448, + "step": 12400 + }, + { + "epoch": 0.08948559662848131, + "grad_norm": 0.16649584472179413, + "learning_rate": 3.879622967581092e-05, + "loss": 3.1529, + "step": 12401 + }, + { + "epoch": 0.08951644309405472, + "grad_norm": 0.1450618952512741, + "learning_rate": 3.879398467525276e-05, + "loss": 3.1494, + "step": 12402 + }, + { + "epoch": 0.08954728955962815, + "grad_norm": 0.152757465839386, + "learning_rate": 3.879173951476163e-05, + "loss": 3.1708, + "step": 12403 + }, + { + "epoch": 0.08957813602520157, + "grad_norm": 0.15321414172649384, + "learning_rate": 3.878949419436357e-05, + "loss": 3.1384, + "step": 12404 + }, + { + "epoch": 0.08960898249077498, + "grad_norm": 0.14977294206619263, + "learning_rate": 3.8787248714084626e-05, + "loss": 3.1612, + "step": 12405 + }, + { + "epoch": 0.0896398289563484, + "grad_norm": 0.1524350345134735, + "learning_rate": 3.87850030739508e-05, + "loss": 3.1922, + "step": 12406 + }, + { + "epoch": 0.08967067542192181, + "grad_norm": 0.14820969104766846, + "learning_rate": 3.8782757273988155e-05, + "loss": 3.1711, + "step": 12407 + }, + { + "epoch": 0.08970152188749522, + "grad_norm": 0.14553014934062958, + "learning_rate": 3.878051131422272e-05, + "loss": 3.1835, + "step": 12408 + }, + { + "epoch": 0.08973236835306865, + "grad_norm": 0.15080861747264862, + "learning_rate": 3.8778265194680527e-05, + "loss": 3.1594, + "step": 12409 + }, + { + "epoch": 0.08976321481864206, + "grad_norm": 0.14350950717926025, + "learning_rate": 3.8776018915387644e-05, + "loss": 3.172, + "step": 12410 + }, + { + "epoch": 0.08979406128421548, + "grad_norm": 0.14250285923480988, + "learning_rate": 3.87737724763701e-05, + "loss": 3.1265, + "step": 12411 + }, + { + "epoch": 0.0898249077497889, + "grad_norm": 0.1537894606590271, + "learning_rate": 3.8771525877653935e-05, + "loss": 3.1667, + "step": 12412 + }, + { + "epoch": 0.08985575421536231, + "grad_norm": 0.1500389277935028, + "learning_rate": 3.8769279119265204e-05, + "loss": 3.1407, + "step": 12413 + }, + { + "epoch": 0.08988660068093572, + "grad_norm": 0.1477043181657791, + "learning_rate": 3.876703220122996e-05, + "loss": 3.1671, + "step": 12414 + }, + { + "epoch": 0.08991744714650915, + "grad_norm": 0.1458318531513214, + "learning_rate": 3.8764785123574244e-05, + "loss": 3.1481, + "step": 12415 + }, + { + "epoch": 0.08994829361208256, + "grad_norm": 0.15072645246982574, + "learning_rate": 3.8762537886324124e-05, + "loss": 3.1776, + "step": 12416 + }, + { + "epoch": 0.08997914007765598, + "grad_norm": 0.1426822990179062, + "learning_rate": 3.8760290489505634e-05, + "loss": 3.1496, + "step": 12417 + }, + { + "epoch": 0.09000998654322939, + "grad_norm": 0.16471660137176514, + "learning_rate": 3.8758042933144857e-05, + "loss": 3.1685, + "step": 12418 + }, + { + "epoch": 0.0900408330088028, + "grad_norm": 0.14875924587249756, + "learning_rate": 3.875579521726784e-05, + "loss": 3.1487, + "step": 12419 + }, + { + "epoch": 0.09007167947437622, + "grad_norm": 0.14716176688671112, + "learning_rate": 3.875354734190063e-05, + "loss": 3.1591, + "step": 12420 + }, + { + "epoch": 0.09010252593994965, + "grad_norm": 0.14775343239307404, + "learning_rate": 3.8751299307069325e-05, + "loss": 3.1689, + "step": 12421 + }, + { + "epoch": 0.09013337240552306, + "grad_norm": 0.14992131292819977, + "learning_rate": 3.874905111279995e-05, + "loss": 3.1384, + "step": 12422 + }, + { + "epoch": 0.09016421887109648, + "grad_norm": 0.1461847424507141, + "learning_rate": 3.8746802759118597e-05, + "loss": 3.1551, + "step": 12423 + }, + { + "epoch": 0.09019506533666989, + "grad_norm": 0.14966803789138794, + "learning_rate": 3.874455424605133e-05, + "loss": 3.1482, + "step": 12424 + }, + { + "epoch": 0.0902259118022433, + "grad_norm": 0.1621818244457245, + "learning_rate": 3.874230557362421e-05, + "loss": 3.1353, + "step": 12425 + }, + { + "epoch": 0.09025675826781672, + "grad_norm": 0.14613257348537445, + "learning_rate": 3.8740056741863315e-05, + "loss": 3.1407, + "step": 12426 + }, + { + "epoch": 0.09028760473339015, + "grad_norm": 0.15872794389724731, + "learning_rate": 3.8737807750794723e-05, + "loss": 3.1859, + "step": 12427 + }, + { + "epoch": 0.09031845119896356, + "grad_norm": 0.14724385738372803, + "learning_rate": 3.8735558600444506e-05, + "loss": 3.1349, + "step": 12428 + }, + { + "epoch": 0.09034929766453698, + "grad_norm": 0.17150117456912994, + "learning_rate": 3.873330929083874e-05, + "loss": 3.1728, + "step": 12429 + }, + { + "epoch": 0.09038014413011039, + "grad_norm": 0.14665767550468445, + "learning_rate": 3.8731059822003504e-05, + "loss": 3.1536, + "step": 12430 + }, + { + "epoch": 0.0904109905956838, + "grad_norm": 0.15809960663318634, + "learning_rate": 3.872881019396488e-05, + "loss": 3.1602, + "step": 12431 + }, + { + "epoch": 0.09044183706125722, + "grad_norm": 0.15988275408744812, + "learning_rate": 3.8726560406748955e-05, + "loss": 3.1791, + "step": 12432 + }, + { + "epoch": 0.09047268352683065, + "grad_norm": 0.1487780660390854, + "learning_rate": 3.8724310460381806e-05, + "loss": 3.1384, + "step": 12433 + }, + { + "epoch": 0.09050352999240406, + "grad_norm": 0.16122840344905853, + "learning_rate": 3.8722060354889536e-05, + "loss": 3.1551, + "step": 12434 + }, + { + "epoch": 0.09053437645797748, + "grad_norm": 0.14587822556495667, + "learning_rate": 3.87198100902982e-05, + "loss": 3.1192, + "step": 12435 + }, + { + "epoch": 0.09056522292355089, + "grad_norm": 0.15239091217517853, + "learning_rate": 3.871755966663393e-05, + "loss": 3.1654, + "step": 12436 + }, + { + "epoch": 0.0905960693891243, + "grad_norm": 0.1561124324798584, + "learning_rate": 3.8715309083922794e-05, + "loss": 3.1202, + "step": 12437 + }, + { + "epoch": 0.09062691585469772, + "grad_norm": 0.14951522648334503, + "learning_rate": 3.871305834219089e-05, + "loss": 3.1745, + "step": 12438 + }, + { + "epoch": 0.09065776232027115, + "grad_norm": 0.1447712779045105, + "learning_rate": 3.871080744146432e-05, + "loss": 3.1473, + "step": 12439 + }, + { + "epoch": 0.09068860878584456, + "grad_norm": 0.14946769177913666, + "learning_rate": 3.8708556381769164e-05, + "loss": 3.1784, + "step": 12440 + }, + { + "epoch": 0.09071945525141797, + "grad_norm": 0.14656716585159302, + "learning_rate": 3.8706305163131545e-05, + "loss": 3.1431, + "step": 12441 + }, + { + "epoch": 0.09075030171699139, + "grad_norm": 0.1464308798313141, + "learning_rate": 3.8704053785577555e-05, + "loss": 3.1547, + "step": 12442 + }, + { + "epoch": 0.0907811481825648, + "grad_norm": 0.15272046625614166, + "learning_rate": 3.8701802249133296e-05, + "loss": 3.1402, + "step": 12443 + }, + { + "epoch": 0.09081199464813822, + "grad_norm": 0.1473115235567093, + "learning_rate": 3.869955055382487e-05, + "loss": 3.1421, + "step": 12444 + }, + { + "epoch": 0.09084284111371164, + "grad_norm": 0.1566738784313202, + "learning_rate": 3.869729869967839e-05, + "loss": 3.1533, + "step": 12445 + }, + { + "epoch": 0.09087368757928506, + "grad_norm": 0.14991408586502075, + "learning_rate": 3.869504668671996e-05, + "loss": 3.1759, + "step": 12446 + }, + { + "epoch": 0.09090453404485847, + "grad_norm": 0.14784061908721924, + "learning_rate": 3.8692794514975697e-05, + "loss": 3.162, + "step": 12447 + }, + { + "epoch": 0.09093538051043189, + "grad_norm": 0.14859293401241302, + "learning_rate": 3.869054218447171e-05, + "loss": 3.1645, + "step": 12448 + }, + { + "epoch": 0.0909662269760053, + "grad_norm": 0.15636129677295685, + "learning_rate": 3.8688289695234116e-05, + "loss": 3.1725, + "step": 12449 + }, + { + "epoch": 0.09099707344157872, + "grad_norm": 0.15469998121261597, + "learning_rate": 3.868603704728901e-05, + "loss": 3.1266, + "step": 12450 + }, + { + "epoch": 0.09102791990715214, + "grad_norm": 0.1497073620557785, + "learning_rate": 3.868378424066255e-05, + "loss": 3.1272, + "step": 12451 + }, + { + "epoch": 0.09105876637272556, + "grad_norm": 0.14739441871643066, + "learning_rate": 3.868153127538084e-05, + "loss": 3.1228, + "step": 12452 + }, + { + "epoch": 0.09108961283829897, + "grad_norm": 0.15501052141189575, + "learning_rate": 3.867927815146997e-05, + "loss": 3.148, + "step": 12453 + }, + { + "epoch": 0.09112045930387239, + "grad_norm": 0.154035285115242, + "learning_rate": 3.867702486895611e-05, + "loss": 3.1378, + "step": 12454 + }, + { + "epoch": 0.0911513057694458, + "grad_norm": 0.15404033660888672, + "learning_rate": 3.867477142786536e-05, + "loss": 3.1514, + "step": 12455 + }, + { + "epoch": 0.09118215223501921, + "grad_norm": 0.14858154952526093, + "learning_rate": 3.8672517828223855e-05, + "loss": 3.1949, + "step": 12456 + }, + { + "epoch": 0.09121299870059264, + "grad_norm": 0.15043950080871582, + "learning_rate": 3.867026407005772e-05, + "loss": 3.1443, + "step": 12457 + }, + { + "epoch": 0.09124384516616606, + "grad_norm": 0.1486435830593109, + "learning_rate": 3.866801015339308e-05, + "loss": 3.1355, + "step": 12458 + }, + { + "epoch": 0.09127469163173947, + "grad_norm": 0.14907221496105194, + "learning_rate": 3.866575607825609e-05, + "loss": 3.1558, + "step": 12459 + }, + { + "epoch": 0.09130553809731289, + "grad_norm": 0.14764772355556488, + "learning_rate": 3.8663501844672866e-05, + "loss": 3.1418, + "step": 12460 + }, + { + "epoch": 0.0913363845628863, + "grad_norm": 0.15522155165672302, + "learning_rate": 3.866124745266955e-05, + "loss": 3.181, + "step": 12461 + }, + { + "epoch": 0.09136723102845971, + "grad_norm": 0.1499738246202469, + "learning_rate": 3.865899290227227e-05, + "loss": 3.1229, + "step": 12462 + }, + { + "epoch": 0.09139807749403314, + "grad_norm": 0.15838633477687836, + "learning_rate": 3.865673819350719e-05, + "loss": 3.1303, + "step": 12463 + }, + { + "epoch": 0.09142892395960656, + "grad_norm": 0.1486433893442154, + "learning_rate": 3.865448332640043e-05, + "loss": 3.1544, + "step": 12464 + }, + { + "epoch": 0.09145977042517997, + "grad_norm": 0.15124157071113586, + "learning_rate": 3.8652228300978134e-05, + "loss": 3.1571, + "step": 12465 + }, + { + "epoch": 0.09149061689075338, + "grad_norm": 0.14443181455135345, + "learning_rate": 3.8649973117266464e-05, + "loss": 3.143, + "step": 12466 + }, + { + "epoch": 0.0915214633563268, + "grad_norm": 0.15993528068065643, + "learning_rate": 3.8647717775291546e-05, + "loss": 3.1736, + "step": 12467 + }, + { + "epoch": 0.09155230982190021, + "grad_norm": 0.20738641917705536, + "learning_rate": 3.864546227507956e-05, + "loss": 3.1398, + "step": 12468 + }, + { + "epoch": 0.09158315628747364, + "grad_norm": 0.1438748687505722, + "learning_rate": 3.864320661665663e-05, + "loss": 3.153, + "step": 12469 + }, + { + "epoch": 0.09161400275304706, + "grad_norm": 0.14746001362800598, + "learning_rate": 3.864095080004891e-05, + "loss": 3.1606, + "step": 12470 + }, + { + "epoch": 0.09164484921862047, + "grad_norm": 0.16209712624549866, + "learning_rate": 3.863869482528258e-05, + "loss": 3.1526, + "step": 12471 + }, + { + "epoch": 0.09167569568419388, + "grad_norm": 0.14650602638721466, + "learning_rate": 3.863643869238377e-05, + "loss": 3.1622, + "step": 12472 + }, + { + "epoch": 0.0917065421497673, + "grad_norm": 0.14887481927871704, + "learning_rate": 3.863418240137865e-05, + "loss": 3.1692, + "step": 12473 + }, + { + "epoch": 0.09173738861534071, + "grad_norm": 0.1494794338941574, + "learning_rate": 3.863192595229337e-05, + "loss": 3.1553, + "step": 12474 + }, + { + "epoch": 0.09176823508091414, + "grad_norm": 0.1437661200761795, + "learning_rate": 3.862966934515411e-05, + "loss": 3.1348, + "step": 12475 + }, + { + "epoch": 0.09179908154648755, + "grad_norm": 0.15657886862754822, + "learning_rate": 3.862741257998702e-05, + "loss": 3.1645, + "step": 12476 + }, + { + "epoch": 0.09182992801206097, + "grad_norm": 0.24821127951145172, + "learning_rate": 3.8625155656818276e-05, + "loss": 3.1603, + "step": 12477 + }, + { + "epoch": 0.09186077447763438, + "grad_norm": 0.14989587664604187, + "learning_rate": 3.862289857567404e-05, + "loss": 3.1747, + "step": 12478 + }, + { + "epoch": 0.0918916209432078, + "grad_norm": 0.14584271609783173, + "learning_rate": 3.8620641336580474e-05, + "loss": 3.1226, + "step": 12479 + }, + { + "epoch": 0.09192246740878121, + "grad_norm": 0.15261033177375793, + "learning_rate": 3.8618383939563765e-05, + "loss": 3.1608, + "step": 12480 + }, + { + "epoch": 0.09195331387435464, + "grad_norm": 0.1522604525089264, + "learning_rate": 3.8616126384650064e-05, + "loss": 3.1518, + "step": 12481 + }, + { + "epoch": 0.09198416033992805, + "grad_norm": 0.1536988615989685, + "learning_rate": 3.861386867186557e-05, + "loss": 3.1729, + "step": 12482 + }, + { + "epoch": 0.09201500680550147, + "grad_norm": 0.15182951092720032, + "learning_rate": 3.861161080123645e-05, + "loss": 3.1727, + "step": 12483 + }, + { + "epoch": 0.09204585327107488, + "grad_norm": 0.16002076864242554, + "learning_rate": 3.860935277278888e-05, + "loss": 3.1403, + "step": 12484 + }, + { + "epoch": 0.0920766997366483, + "grad_norm": 0.1519380807876587, + "learning_rate": 3.860709458654904e-05, + "loss": 3.1504, + "step": 12485 + }, + { + "epoch": 0.09210754620222171, + "grad_norm": 0.15692685544490814, + "learning_rate": 3.860483624254312e-05, + "loss": 3.1651, + "step": 12486 + }, + { + "epoch": 0.09213839266779514, + "grad_norm": 0.14805251359939575, + "learning_rate": 3.86025777407973e-05, + "loss": 3.1838, + "step": 12487 + }, + { + "epoch": 0.09216923913336855, + "grad_norm": 0.15399859845638275, + "learning_rate": 3.8600319081337764e-05, + "loss": 3.1593, + "step": 12488 + }, + { + "epoch": 0.09220008559894197, + "grad_norm": 0.1593552529811859, + "learning_rate": 3.85980602641907e-05, + "loss": 3.1506, + "step": 12489 + }, + { + "epoch": 0.09223093206451538, + "grad_norm": 0.15438689291477203, + "learning_rate": 3.8595801289382305e-05, + "loss": 3.1651, + "step": 12490 + }, + { + "epoch": 0.0922617785300888, + "grad_norm": 0.15248854458332062, + "learning_rate": 3.8593542156938756e-05, + "loss": 3.1562, + "step": 12491 + }, + { + "epoch": 0.09229262499566221, + "grad_norm": 0.1610730141401291, + "learning_rate": 3.8591282866886256e-05, + "loss": 3.1668, + "step": 12492 + }, + { + "epoch": 0.09232347146123564, + "grad_norm": 0.15712828934192657, + "learning_rate": 3.8589023419251e-05, + "loss": 3.1472, + "step": 12493 + }, + { + "epoch": 0.09235431792680905, + "grad_norm": 0.1463104635477066, + "learning_rate": 3.858676381405918e-05, + "loss": 3.162, + "step": 12494 + }, + { + "epoch": 0.09238516439238247, + "grad_norm": 0.1508655548095703, + "learning_rate": 3.8584504051337e-05, + "loss": 3.1468, + "step": 12495 + }, + { + "epoch": 0.09241601085795588, + "grad_norm": 0.1454014778137207, + "learning_rate": 3.858224413111067e-05, + "loss": 3.1176, + "step": 12496 + }, + { + "epoch": 0.0924468573235293, + "grad_norm": 0.14719560742378235, + "learning_rate": 3.8579984053406365e-05, + "loss": 3.1289, + "step": 12497 + }, + { + "epoch": 0.09247770378910271, + "grad_norm": 0.14817532896995544, + "learning_rate": 3.857772381825031e-05, + "loss": 3.1475, + "step": 12498 + }, + { + "epoch": 0.09250855025467614, + "grad_norm": 0.1577799916267395, + "learning_rate": 3.857546342566871e-05, + "loss": 3.1508, + "step": 12499 + }, + { + "epoch": 0.09253939672024955, + "grad_norm": 0.1478809118270874, + "learning_rate": 3.8573202875687764e-05, + "loss": 3.1768, + "step": 12500 + }, + { + "epoch": 0.09257024318582296, + "grad_norm": 0.14885355532169342, + "learning_rate": 3.85709421683337e-05, + "loss": 3.1516, + "step": 12501 + }, + { + "epoch": 0.09260108965139638, + "grad_norm": 0.1553841084241867, + "learning_rate": 3.8568681303632706e-05, + "loss": 3.1418, + "step": 12502 + }, + { + "epoch": 0.09263193611696979, + "grad_norm": 0.15342025458812714, + "learning_rate": 3.8566420281611e-05, + "loss": 3.1739, + "step": 12503 + }, + { + "epoch": 0.09266278258254321, + "grad_norm": 0.1432815045118332, + "learning_rate": 3.8564159102294814e-05, + "loss": 3.152, + "step": 12504 + }, + { + "epoch": 0.09269362904811664, + "grad_norm": 0.1460948884487152, + "learning_rate": 3.856189776571035e-05, + "loss": 3.1284, + "step": 12505 + }, + { + "epoch": 0.09272447551369005, + "grad_norm": 0.14740249514579773, + "learning_rate": 3.8559636271883835e-05, + "loss": 3.1263, + "step": 12506 + }, + { + "epoch": 0.09275532197926346, + "grad_norm": 0.14945659041404724, + "learning_rate": 3.855737462084149e-05, + "loss": 3.1592, + "step": 12507 + }, + { + "epoch": 0.09278616844483688, + "grad_norm": 0.17667776346206665, + "learning_rate": 3.855511281260953e-05, + "loss": 3.1698, + "step": 12508 + }, + { + "epoch": 0.09281701491041029, + "grad_norm": 0.15103456377983093, + "learning_rate": 3.855285084721417e-05, + "loss": 3.1389, + "step": 12509 + }, + { + "epoch": 0.0928478613759837, + "grad_norm": 0.15905940532684326, + "learning_rate": 3.8550588724681664e-05, + "loss": 3.1459, + "step": 12510 + }, + { + "epoch": 0.09287870784155713, + "grad_norm": 0.14893710613250732, + "learning_rate": 3.854832644503822e-05, + "loss": 3.1562, + "step": 12511 + }, + { + "epoch": 0.09290955430713055, + "grad_norm": 0.15349836647510529, + "learning_rate": 3.8546064008310066e-05, + "loss": 3.1916, + "step": 12512 + }, + { + "epoch": 0.09294040077270396, + "grad_norm": 0.15900471806526184, + "learning_rate": 3.854380141452345e-05, + "loss": 3.1566, + "step": 12513 + }, + { + "epoch": 0.09297124723827738, + "grad_norm": 0.15625210106372833, + "learning_rate": 3.854153866370459e-05, + "loss": 3.1519, + "step": 12514 + }, + { + "epoch": 0.09300209370385079, + "grad_norm": 0.1502264142036438, + "learning_rate": 3.8539275755879724e-05, + "loss": 3.1391, + "step": 12515 + }, + { + "epoch": 0.0930329401694242, + "grad_norm": 0.1541593223810196, + "learning_rate": 3.85370126910751e-05, + "loss": 3.1549, + "step": 12516 + }, + { + "epoch": 0.09306378663499763, + "grad_norm": 0.14852024614810944, + "learning_rate": 3.853474946931694e-05, + "loss": 3.1574, + "step": 12517 + }, + { + "epoch": 0.09309463310057105, + "grad_norm": 0.15268439054489136, + "learning_rate": 3.85324860906315e-05, + "loss": 3.1968, + "step": 12518 + }, + { + "epoch": 0.09312547956614446, + "grad_norm": 0.16086603701114655, + "learning_rate": 3.8530222555045016e-05, + "loss": 3.1481, + "step": 12519 + }, + { + "epoch": 0.09315632603171788, + "grad_norm": 0.154897540807724, + "learning_rate": 3.8527958862583726e-05, + "loss": 3.1613, + "step": 12520 + }, + { + "epoch": 0.09318717249729129, + "grad_norm": 0.1531940996646881, + "learning_rate": 3.852569501327389e-05, + "loss": 3.1669, + "step": 12521 + }, + { + "epoch": 0.0932180189628647, + "grad_norm": 0.15477024018764496, + "learning_rate": 3.8523431007141744e-05, + "loss": 3.1416, + "step": 12522 + }, + { + "epoch": 0.09324886542843813, + "grad_norm": 0.14881478250026703, + "learning_rate": 3.852116684421354e-05, + "loss": 3.1643, + "step": 12523 + }, + { + "epoch": 0.09327971189401155, + "grad_norm": 0.16556881368160248, + "learning_rate": 3.851890252451553e-05, + "loss": 3.1379, + "step": 12524 + }, + { + "epoch": 0.09331055835958496, + "grad_norm": 0.16501455008983612, + "learning_rate": 3.851663804807398e-05, + "loss": 3.1697, + "step": 12525 + }, + { + "epoch": 0.09334140482515838, + "grad_norm": 0.15746380388736725, + "learning_rate": 3.851437341491512e-05, + "loss": 3.1796, + "step": 12526 + }, + { + "epoch": 0.09337225129073179, + "grad_norm": 0.14774903655052185, + "learning_rate": 3.8512108625065235e-05, + "loss": 3.1346, + "step": 12527 + }, + { + "epoch": 0.09340309775630522, + "grad_norm": 0.15462897717952728, + "learning_rate": 3.850984367855056e-05, + "loss": 3.1705, + "step": 12528 + }, + { + "epoch": 0.09343394422187863, + "grad_norm": 0.14558812975883484, + "learning_rate": 3.850757857539737e-05, + "loss": 3.1471, + "step": 12529 + }, + { + "epoch": 0.09346479068745205, + "grad_norm": 0.1508655846118927, + "learning_rate": 3.850531331563193e-05, + "loss": 3.1434, + "step": 12530 + }, + { + "epoch": 0.09349563715302546, + "grad_norm": 0.15131144225597382, + "learning_rate": 3.850304789928049e-05, + "loss": 3.1571, + "step": 12531 + }, + { + "epoch": 0.09352648361859887, + "grad_norm": 0.14289860427379608, + "learning_rate": 3.850078232636932e-05, + "loss": 3.1569, + "step": 12532 + }, + { + "epoch": 0.09355733008417229, + "grad_norm": 0.15620829164981842, + "learning_rate": 3.8498516596924705e-05, + "loss": 3.1821, + "step": 12533 + }, + { + "epoch": 0.09358817654974572, + "grad_norm": 0.15069328248500824, + "learning_rate": 3.849625071097289e-05, + "loss": 3.174, + "step": 12534 + }, + { + "epoch": 0.09361902301531913, + "grad_norm": 0.14792220294475555, + "learning_rate": 3.8493984668540165e-05, + "loss": 3.1539, + "step": 12535 + }, + { + "epoch": 0.09364986948089254, + "grad_norm": 0.1442490667104721, + "learning_rate": 3.849171846965279e-05, + "loss": 3.1408, + "step": 12536 + }, + { + "epoch": 0.09368071594646596, + "grad_norm": 0.15109966695308685, + "learning_rate": 3.8489452114337054e-05, + "loss": 3.1168, + "step": 12537 + }, + { + "epoch": 0.09371156241203937, + "grad_norm": 0.14903397858142853, + "learning_rate": 3.8487185602619224e-05, + "loss": 3.1512, + "step": 12538 + }, + { + "epoch": 0.09374240887761279, + "grad_norm": 0.15107686817646027, + "learning_rate": 3.848491893452558e-05, + "loss": 3.1344, + "step": 12539 + }, + { + "epoch": 0.09377325534318622, + "grad_norm": 0.15397940576076508, + "learning_rate": 3.848265211008242e-05, + "loss": 3.1419, + "step": 12540 + }, + { + "epoch": 0.09380410180875963, + "grad_norm": 0.15150074660778046, + "learning_rate": 3.848038512931599e-05, + "loss": 3.1745, + "step": 12541 + }, + { + "epoch": 0.09383494827433304, + "grad_norm": 0.15291796624660492, + "learning_rate": 3.8478117992252606e-05, + "loss": 3.1554, + "step": 12542 + }, + { + "epoch": 0.09386579473990646, + "grad_norm": 0.14840815961360931, + "learning_rate": 3.8475850698918535e-05, + "loss": 3.1442, + "step": 12543 + }, + { + "epoch": 0.09389664120547987, + "grad_norm": 0.14720304310321808, + "learning_rate": 3.8473583249340076e-05, + "loss": 3.1802, + "step": 12544 + }, + { + "epoch": 0.09392748767105329, + "grad_norm": 0.14907307922840118, + "learning_rate": 3.847131564354352e-05, + "loss": 3.1681, + "step": 12545 + }, + { + "epoch": 0.09395833413662671, + "grad_norm": 0.15827055275440216, + "learning_rate": 3.846904788155515e-05, + "loss": 3.1408, + "step": 12546 + }, + { + "epoch": 0.09398918060220013, + "grad_norm": 0.1971842646598816, + "learning_rate": 3.8466779963401266e-05, + "loss": 3.1451, + "step": 12547 + }, + { + "epoch": 0.09402002706777354, + "grad_norm": 0.1467488408088684, + "learning_rate": 3.8464511889108165e-05, + "loss": 3.1719, + "step": 12548 + }, + { + "epoch": 0.09405087353334696, + "grad_norm": 0.15287770330905914, + "learning_rate": 3.8462243658702134e-05, + "loss": 3.1502, + "step": 12549 + }, + { + "epoch": 0.09408171999892037, + "grad_norm": 0.16651871800422668, + "learning_rate": 3.845997527220948e-05, + "loss": 3.1301, + "step": 12550 + }, + { + "epoch": 0.09411256646449379, + "grad_norm": 0.15667033195495605, + "learning_rate": 3.845770672965651e-05, + "loss": 3.1537, + "step": 12551 + }, + { + "epoch": 0.09414341293006721, + "grad_norm": 0.15500085055828094, + "learning_rate": 3.84554380310695e-05, + "loss": 3.1597, + "step": 12552 + }, + { + "epoch": 0.09417425939564063, + "grad_norm": 0.15643593668937683, + "learning_rate": 3.845316917647478e-05, + "loss": 3.1817, + "step": 12553 + }, + { + "epoch": 0.09420510586121404, + "grad_norm": 0.14938579499721527, + "learning_rate": 3.845090016589865e-05, + "loss": 3.178, + "step": 12554 + }, + { + "epoch": 0.09423595232678746, + "grad_norm": 0.15404821932315826, + "learning_rate": 3.844863099936742e-05, + "loss": 3.1594, + "step": 12555 + }, + { + "epoch": 0.09426679879236087, + "grad_norm": 0.14980576932430267, + "learning_rate": 3.844636167690738e-05, + "loss": 3.129, + "step": 12556 + }, + { + "epoch": 0.09429764525793428, + "grad_norm": 0.16578347980976105, + "learning_rate": 3.8444092198544867e-05, + "loss": 3.1776, + "step": 12557 + }, + { + "epoch": 0.09432849172350771, + "grad_norm": 0.14460626244544983, + "learning_rate": 3.844182256430619e-05, + "loss": 3.1724, + "step": 12558 + }, + { + "epoch": 0.09435933818908113, + "grad_norm": 0.14994142949581146, + "learning_rate": 3.8439552774217644e-05, + "loss": 3.1361, + "step": 12559 + }, + { + "epoch": 0.09439018465465454, + "grad_norm": 0.1527656465768814, + "learning_rate": 3.843728282830557e-05, + "loss": 3.0922, + "step": 12560 + }, + { + "epoch": 0.09442103112022796, + "grad_norm": 0.14125937223434448, + "learning_rate": 3.8435012726596267e-05, + "loss": 3.1413, + "step": 12561 + }, + { + "epoch": 0.09445187758580137, + "grad_norm": 0.15882781147956848, + "learning_rate": 3.8432742469116076e-05, + "loss": 3.1098, + "step": 12562 + }, + { + "epoch": 0.09448272405137478, + "grad_norm": 0.14874249696731567, + "learning_rate": 3.84304720558913e-05, + "loss": 3.1391, + "step": 12563 + }, + { + "epoch": 0.09451357051694821, + "grad_norm": 0.14718782901763916, + "learning_rate": 3.842820148694828e-05, + "loss": 3.139, + "step": 12564 + }, + { + "epoch": 0.09454441698252163, + "grad_norm": 0.14501650631427765, + "learning_rate": 3.8425930762313325e-05, + "loss": 3.1102, + "step": 12565 + }, + { + "epoch": 0.09457526344809504, + "grad_norm": 0.14481492340564728, + "learning_rate": 3.8423659882012776e-05, + "loss": 3.1548, + "step": 12566 + }, + { + "epoch": 0.09460610991366845, + "grad_norm": 0.16309590637683868, + "learning_rate": 3.8421388846072945e-05, + "loss": 3.1219, + "step": 12567 + }, + { + "epoch": 0.09463695637924187, + "grad_norm": 0.16389897465705872, + "learning_rate": 3.841911765452019e-05, + "loss": 3.1342, + "step": 12568 + }, + { + "epoch": 0.09466780284481528, + "grad_norm": 0.15333113074302673, + "learning_rate": 3.841684630738083e-05, + "loss": 3.1213, + "step": 12569 + }, + { + "epoch": 0.09469864931038871, + "grad_norm": 0.14018677175045013, + "learning_rate": 3.841457480468119e-05, + "loss": 3.1536, + "step": 12570 + }, + { + "epoch": 0.09472949577596212, + "grad_norm": 0.15262095630168915, + "learning_rate": 3.841230314644762e-05, + "loss": 3.1456, + "step": 12571 + }, + { + "epoch": 0.09476034224153554, + "grad_norm": 0.15574726462364197, + "learning_rate": 3.8410031332706456e-05, + "loss": 3.1651, + "step": 12572 + }, + { + "epoch": 0.09479118870710895, + "grad_norm": 0.15855908393859863, + "learning_rate": 3.840775936348404e-05, + "loss": 3.1716, + "step": 12573 + }, + { + "epoch": 0.09482203517268237, + "grad_norm": 0.15594834089279175, + "learning_rate": 3.840548723880671e-05, + "loss": 3.1466, + "step": 12574 + }, + { + "epoch": 0.09485288163825578, + "grad_norm": 0.15382793545722961, + "learning_rate": 3.8403214958700814e-05, + "loss": 3.1416, + "step": 12575 + }, + { + "epoch": 0.09488372810382921, + "grad_norm": 0.14891520142555237, + "learning_rate": 3.8400942523192683e-05, + "loss": 3.1284, + "step": 12576 + }, + { + "epoch": 0.09491457456940262, + "grad_norm": 0.15104977786540985, + "learning_rate": 3.8398669932308684e-05, + "loss": 3.1389, + "step": 12577 + }, + { + "epoch": 0.09494542103497604, + "grad_norm": 0.1524331271648407, + "learning_rate": 3.8396397186075163e-05, + "loss": 3.1687, + "step": 12578 + }, + { + "epoch": 0.09497626750054945, + "grad_norm": 0.1476842164993286, + "learning_rate": 3.839412428451846e-05, + "loss": 3.1277, + "step": 12579 + }, + { + "epoch": 0.09500711396612287, + "grad_norm": 0.15126287937164307, + "learning_rate": 3.8391851227664946e-05, + "loss": 3.147, + "step": 12580 + }, + { + "epoch": 0.09503796043169628, + "grad_norm": 0.1662503331899643, + "learning_rate": 3.838957801554095e-05, + "loss": 3.1447, + "step": 12581 + }, + { + "epoch": 0.09506880689726971, + "grad_norm": 0.16605308651924133, + "learning_rate": 3.838730464817284e-05, + "loss": 3.1568, + "step": 12582 + }, + { + "epoch": 0.09509965336284312, + "grad_norm": 0.15883521735668182, + "learning_rate": 3.8385031125587e-05, + "loss": 3.1799, + "step": 12583 + }, + { + "epoch": 0.09513049982841654, + "grad_norm": 0.15356343984603882, + "learning_rate": 3.8382757447809755e-05, + "loss": 3.1507, + "step": 12584 + }, + { + "epoch": 0.09516134629398995, + "grad_norm": 0.15428900718688965, + "learning_rate": 3.8380483614867474e-05, + "loss": 3.1953, + "step": 12585 + }, + { + "epoch": 0.09519219275956337, + "grad_norm": 0.14965294301509857, + "learning_rate": 3.8378209626786535e-05, + "loss": 3.1305, + "step": 12586 + }, + { + "epoch": 0.09522303922513678, + "grad_norm": 0.15119293332099915, + "learning_rate": 3.8375935483593284e-05, + "loss": 3.1586, + "step": 12587 + }, + { + "epoch": 0.09525388569071021, + "grad_norm": 0.15200911462306976, + "learning_rate": 3.837366118531411e-05, + "loss": 3.1494, + "step": 12588 + }, + { + "epoch": 0.09528473215628362, + "grad_norm": 0.1565874218940735, + "learning_rate": 3.8371386731975365e-05, + "loss": 3.1563, + "step": 12589 + }, + { + "epoch": 0.09531557862185704, + "grad_norm": 0.14644084870815277, + "learning_rate": 3.8369112123603436e-05, + "loss": 3.153, + "step": 12590 + }, + { + "epoch": 0.09534642508743045, + "grad_norm": 0.1683824211359024, + "learning_rate": 3.8366837360224675e-05, + "loss": 3.162, + "step": 12591 + }, + { + "epoch": 0.09537727155300386, + "grad_norm": 0.15215449035167694, + "learning_rate": 3.836456244186547e-05, + "loss": 3.1641, + "step": 12592 + }, + { + "epoch": 0.09540811801857728, + "grad_norm": 0.15132509171962738, + "learning_rate": 3.8362287368552195e-05, + "loss": 3.1587, + "step": 12593 + }, + { + "epoch": 0.0954389644841507, + "grad_norm": 0.14266306161880493, + "learning_rate": 3.836001214031122e-05, + "loss": 3.1502, + "step": 12594 + }, + { + "epoch": 0.09546981094972412, + "grad_norm": 0.14316625893115997, + "learning_rate": 3.835773675716895e-05, + "loss": 3.1291, + "step": 12595 + }, + { + "epoch": 0.09550065741529754, + "grad_norm": 0.1459498554468155, + "learning_rate": 3.835546121915174e-05, + "loss": 3.1359, + "step": 12596 + }, + { + "epoch": 0.09553150388087095, + "grad_norm": 0.14919961988925934, + "learning_rate": 3.835318552628598e-05, + "loss": 3.1598, + "step": 12597 + }, + { + "epoch": 0.09556235034644436, + "grad_norm": 0.158364936709404, + "learning_rate": 3.8350909678598064e-05, + "loss": 3.142, + "step": 12598 + }, + { + "epoch": 0.09559319681201778, + "grad_norm": 0.15607303380966187, + "learning_rate": 3.834863367611437e-05, + "loss": 3.1648, + "step": 12599 + }, + { + "epoch": 0.0956240432775912, + "grad_norm": 0.14841103553771973, + "learning_rate": 3.8346357518861286e-05, + "loss": 3.1627, + "step": 12600 + }, + { + "epoch": 0.09565488974316462, + "grad_norm": 0.16173812747001648, + "learning_rate": 3.834408120686521e-05, + "loss": 3.1536, + "step": 12601 + }, + { + "epoch": 0.09568573620873803, + "grad_norm": 0.14996004104614258, + "learning_rate": 3.834180474015254e-05, + "loss": 3.1303, + "step": 12602 + }, + { + "epoch": 0.09571658267431145, + "grad_norm": 0.1507776528596878, + "learning_rate": 3.833952811874965e-05, + "loss": 3.1315, + "step": 12603 + }, + { + "epoch": 0.09574742913988486, + "grad_norm": 0.14740394055843353, + "learning_rate": 3.833725134268295e-05, + "loss": 3.131, + "step": 12604 + }, + { + "epoch": 0.09577827560545828, + "grad_norm": 0.162022665143013, + "learning_rate": 3.833497441197884e-05, + "loss": 3.1564, + "step": 12605 + }, + { + "epoch": 0.0958091220710317, + "grad_norm": 0.14602576196193695, + "learning_rate": 3.833269732666371e-05, + "loss": 3.1874, + "step": 12606 + }, + { + "epoch": 0.09583996853660512, + "grad_norm": 0.1546902060508728, + "learning_rate": 3.8330420086763966e-05, + "loss": 3.1489, + "step": 12607 + }, + { + "epoch": 0.09587081500217853, + "grad_norm": 0.1542103886604309, + "learning_rate": 3.8328142692306026e-05, + "loss": 3.1479, + "step": 12608 + }, + { + "epoch": 0.09590166146775195, + "grad_norm": 0.32484444975852966, + "learning_rate": 3.832586514331626e-05, + "loss": 3.1356, + "step": 12609 + }, + { + "epoch": 0.09593250793332536, + "grad_norm": 0.1508627086877823, + "learning_rate": 3.832358743982111e-05, + "loss": 3.1521, + "step": 12610 + }, + { + "epoch": 0.09596335439889878, + "grad_norm": 0.1488562971353531, + "learning_rate": 3.8321309581846955e-05, + "loss": 3.1549, + "step": 12611 + }, + { + "epoch": 0.0959942008644722, + "grad_norm": 0.1554533690214157, + "learning_rate": 3.831903156942023e-05, + "loss": 3.1595, + "step": 12612 + }, + { + "epoch": 0.09602504733004562, + "grad_norm": 0.15213382244110107, + "learning_rate": 3.831675340256735e-05, + "loss": 3.1882, + "step": 12613 + }, + { + "epoch": 0.09605589379561903, + "grad_norm": 0.976231038570404, + "learning_rate": 3.8314475081314694e-05, + "loss": 3.1788, + "step": 12614 + }, + { + "epoch": 0.09608674026119245, + "grad_norm": 0.15258052945137024, + "learning_rate": 3.831219660568871e-05, + "loss": 3.1398, + "step": 12615 + }, + { + "epoch": 0.09611758672676586, + "grad_norm": 0.16195376217365265, + "learning_rate": 3.8309917975715816e-05, + "loss": 3.1128, + "step": 12616 + }, + { + "epoch": 0.09614843319233927, + "grad_norm": 0.4269203543663025, + "learning_rate": 3.8307639191422405e-05, + "loss": 3.1671, + "step": 12617 + }, + { + "epoch": 0.0961792796579127, + "grad_norm": 0.1554064005613327, + "learning_rate": 3.830536025283493e-05, + "loss": 3.0955, + "step": 12618 + }, + { + "epoch": 0.09621012612348612, + "grad_norm": 0.18014097213745117, + "learning_rate": 3.8303081159979794e-05, + "loss": 3.145, + "step": 12619 + }, + { + "epoch": 0.09624097258905953, + "grad_norm": 0.16219961643218994, + "learning_rate": 3.830080191288342e-05, + "loss": 3.1505, + "step": 12620 + }, + { + "epoch": 0.09627181905463295, + "grad_norm": 0.16947384178638458, + "learning_rate": 3.8298522511572236e-05, + "loss": 3.1657, + "step": 12621 + }, + { + "epoch": 0.09630266552020636, + "grad_norm": 0.15975187718868256, + "learning_rate": 3.829624295607269e-05, + "loss": 3.1804, + "step": 12622 + }, + { + "epoch": 0.09633351198577977, + "grad_norm": 0.1824568212032318, + "learning_rate": 3.829396324641118e-05, + "loss": 3.1466, + "step": 12623 + }, + { + "epoch": 0.0963643584513532, + "grad_norm": 0.15631359815597534, + "learning_rate": 3.829168338261417e-05, + "loss": 3.1298, + "step": 12624 + }, + { + "epoch": 0.09639520491692662, + "grad_norm": 0.17808720469474792, + "learning_rate": 3.8289403364708075e-05, + "loss": 3.1523, + "step": 12625 + }, + { + "epoch": 0.09642605138250003, + "grad_norm": 0.1967097967863083, + "learning_rate": 3.828712319271933e-05, + "loss": 3.1667, + "step": 12626 + }, + { + "epoch": 0.09645689784807344, + "grad_norm": 0.16806913912296295, + "learning_rate": 3.828484286667437e-05, + "loss": 3.1447, + "step": 12627 + }, + { + "epoch": 0.09648774431364686, + "grad_norm": 0.1587931215763092, + "learning_rate": 3.828256238659966e-05, + "loss": 3.1483, + "step": 12628 + }, + { + "epoch": 0.09651859077922027, + "grad_norm": 0.5351662635803223, + "learning_rate": 3.828028175252161e-05, + "loss": 3.1482, + "step": 12629 + }, + { + "epoch": 0.0965494372447937, + "grad_norm": 0.16330543160438538, + "learning_rate": 3.8278000964466674e-05, + "loss": 3.1641, + "step": 12630 + }, + { + "epoch": 0.09658028371036712, + "grad_norm": 0.16426074504852295, + "learning_rate": 3.827572002246129e-05, + "loss": 3.1716, + "step": 12631 + }, + { + "epoch": 0.09661113017594053, + "grad_norm": 0.15280604362487793, + "learning_rate": 3.827343892653191e-05, + "loss": 3.1655, + "step": 12632 + }, + { + "epoch": 0.09664197664151394, + "grad_norm": 0.16682404279708862, + "learning_rate": 3.827115767670499e-05, + "loss": 3.1641, + "step": 12633 + }, + { + "epoch": 0.09667282310708736, + "grad_norm": 0.18762479722499847, + "learning_rate": 3.8268876273006976e-05, + "loss": 3.1594, + "step": 12634 + }, + { + "epoch": 0.09670366957266077, + "grad_norm": 0.15487173199653625, + "learning_rate": 3.826659471546431e-05, + "loss": 3.1894, + "step": 12635 + }, + { + "epoch": 0.0967345160382342, + "grad_norm": 0.16700276732444763, + "learning_rate": 3.826431300410345e-05, + "loss": 3.1436, + "step": 12636 + }, + { + "epoch": 0.09676536250380761, + "grad_norm": 0.15980955958366394, + "learning_rate": 3.826203113895085e-05, + "loss": 3.161, + "step": 12637 + }, + { + "epoch": 0.09679620896938103, + "grad_norm": 0.15719780325889587, + "learning_rate": 3.825974912003296e-05, + "loss": 3.1486, + "step": 12638 + }, + { + "epoch": 0.09682705543495444, + "grad_norm": 0.16452427208423615, + "learning_rate": 3.825746694737626e-05, + "loss": 3.1371, + "step": 12639 + }, + { + "epoch": 0.09685790190052786, + "grad_norm": 0.16199591755867004, + "learning_rate": 3.82551846210072e-05, + "loss": 3.165, + "step": 12640 + }, + { + "epoch": 0.09688874836610127, + "grad_norm": 0.16150183975696564, + "learning_rate": 3.825290214095223e-05, + "loss": 3.1588, + "step": 12641 + }, + { + "epoch": 0.0969195948316747, + "grad_norm": 0.15167033672332764, + "learning_rate": 3.825061950723783e-05, + "loss": 3.1434, + "step": 12642 + }, + { + "epoch": 0.09695044129724811, + "grad_norm": 0.15835846960544586, + "learning_rate": 3.824833671989045e-05, + "loss": 3.1346, + "step": 12643 + }, + { + "epoch": 0.09698128776282153, + "grad_norm": 0.15980644524097443, + "learning_rate": 3.824605377893657e-05, + "loss": 3.1472, + "step": 12644 + }, + { + "epoch": 0.09701213422839494, + "grad_norm": 0.15771232545375824, + "learning_rate": 3.824377068440267e-05, + "loss": 3.1346, + "step": 12645 + }, + { + "epoch": 0.09704298069396836, + "grad_norm": 0.15027858316898346, + "learning_rate": 3.8241487436315196e-05, + "loss": 3.1301, + "step": 12646 + }, + { + "epoch": 0.09707382715954177, + "grad_norm": 0.15019576251506805, + "learning_rate": 3.823920403470064e-05, + "loss": 3.1294, + "step": 12647 + }, + { + "epoch": 0.0971046736251152, + "grad_norm": 0.15422378480434418, + "learning_rate": 3.823692047958546e-05, + "loss": 3.1659, + "step": 12648 + }, + { + "epoch": 0.09713552009068861, + "grad_norm": 0.15325747430324554, + "learning_rate": 3.823463677099614e-05, + "loss": 3.1248, + "step": 12649 + }, + { + "epoch": 0.09716636655626203, + "grad_norm": 0.15223194658756256, + "learning_rate": 3.8232352908959166e-05, + "loss": 3.1424, + "step": 12650 + }, + { + "epoch": 0.09719721302183544, + "grad_norm": 0.15896816551685333, + "learning_rate": 3.8230068893501016e-05, + "loss": 3.1506, + "step": 12651 + }, + { + "epoch": 0.09722805948740885, + "grad_norm": 0.15044035017490387, + "learning_rate": 3.8227784724648155e-05, + "loss": 3.1375, + "step": 12652 + }, + { + "epoch": 0.09725890595298227, + "grad_norm": 0.15937615931034088, + "learning_rate": 3.822550040242708e-05, + "loss": 3.1475, + "step": 12653 + }, + { + "epoch": 0.0972897524185557, + "grad_norm": 0.1486966758966446, + "learning_rate": 3.822321592686429e-05, + "loss": 3.1565, + "step": 12654 + }, + { + "epoch": 0.09732059888412911, + "grad_norm": 0.15648868680000305, + "learning_rate": 3.822093129798625e-05, + "loss": 3.1226, + "step": 12655 + }, + { + "epoch": 0.09735144534970253, + "grad_norm": 0.15133407711982727, + "learning_rate": 3.821864651581945e-05, + "loss": 3.1637, + "step": 12656 + }, + { + "epoch": 0.09738229181527594, + "grad_norm": 0.15082867443561554, + "learning_rate": 3.821636158039039e-05, + "loss": 3.1375, + "step": 12657 + }, + { + "epoch": 0.09741313828084935, + "grad_norm": 0.1663205623626709, + "learning_rate": 3.821407649172557e-05, + "loss": 3.1567, + "step": 12658 + }, + { + "epoch": 0.09744398474642277, + "grad_norm": 0.14632612466812134, + "learning_rate": 3.8211791249851464e-05, + "loss": 3.1469, + "step": 12659 + }, + { + "epoch": 0.0974748312119962, + "grad_norm": 0.22336016595363617, + "learning_rate": 3.820950585479458e-05, + "loss": 3.1668, + "step": 12660 + }, + { + "epoch": 0.09750567767756961, + "grad_norm": 0.1505894958972931, + "learning_rate": 3.8207220306581404e-05, + "loss": 3.1228, + "step": 12661 + }, + { + "epoch": 0.09753652414314302, + "grad_norm": 0.32364562153816223, + "learning_rate": 3.8204934605238465e-05, + "loss": 3.1402, + "step": 12662 + }, + { + "epoch": 0.09756737060871644, + "grad_norm": 0.1750403791666031, + "learning_rate": 3.8202648750792234e-05, + "loss": 3.1339, + "step": 12663 + }, + { + "epoch": 0.09759821707428985, + "grad_norm": 0.15285935997962952, + "learning_rate": 3.820036274326922e-05, + "loss": 3.1437, + "step": 12664 + }, + { + "epoch": 0.09762906353986327, + "grad_norm": 0.14940477907657623, + "learning_rate": 3.819807658269594e-05, + "loss": 3.1619, + "step": 12665 + }, + { + "epoch": 0.0976599100054367, + "grad_norm": 0.1565280705690384, + "learning_rate": 3.819579026909889e-05, + "loss": 3.1646, + "step": 12666 + }, + { + "epoch": 0.09769075647101011, + "grad_norm": 0.1497875601053238, + "learning_rate": 3.8193503802504575e-05, + "loss": 3.1805, + "step": 12667 + }, + { + "epoch": 0.09772160293658352, + "grad_norm": 0.15254054963588715, + "learning_rate": 3.819121718293952e-05, + "loss": 3.1438, + "step": 12668 + }, + { + "epoch": 0.09775244940215694, + "grad_norm": 0.1793551743030548, + "learning_rate": 3.818893041043023e-05, + "loss": 3.1464, + "step": 12669 + }, + { + "epoch": 0.09778329586773035, + "grad_norm": 0.15199705958366394, + "learning_rate": 3.8186643485003214e-05, + "loss": 3.1809, + "step": 12670 + }, + { + "epoch": 0.09781414233330377, + "grad_norm": 0.16798673570156097, + "learning_rate": 3.818435640668499e-05, + "loss": 3.1373, + "step": 12671 + }, + { + "epoch": 0.0978449887988772, + "grad_norm": 0.1575695276260376, + "learning_rate": 3.818206917550208e-05, + "loss": 3.169, + "step": 12672 + }, + { + "epoch": 0.09787583526445061, + "grad_norm": 0.15848460793495178, + "learning_rate": 3.8179781791480994e-05, + "loss": 3.1154, + "step": 12673 + }, + { + "epoch": 0.09790668173002402, + "grad_norm": 0.1502181738615036, + "learning_rate": 3.817749425464827e-05, + "loss": 3.1371, + "step": 12674 + }, + { + "epoch": 0.09793752819559744, + "grad_norm": 0.16575853526592255, + "learning_rate": 3.817520656503041e-05, + "loss": 3.1492, + "step": 12675 + }, + { + "epoch": 0.09796837466117085, + "grad_norm": 0.15008647739887238, + "learning_rate": 3.817291872265394e-05, + "loss": 3.1539, + "step": 12676 + }, + { + "epoch": 0.09799922112674427, + "grad_norm": 0.14615492522716522, + "learning_rate": 3.81706307275454e-05, + "loss": 3.1172, + "step": 12677 + }, + { + "epoch": 0.0980300675923177, + "grad_norm": 0.1588008552789688, + "learning_rate": 3.816834257973132e-05, + "loss": 3.1495, + "step": 12678 + }, + { + "epoch": 0.09806091405789111, + "grad_norm": 0.14953340590000153, + "learning_rate": 3.816605427923821e-05, + "loss": 3.1726, + "step": 12679 + }, + { + "epoch": 0.09809176052346452, + "grad_norm": 0.15749286115169525, + "learning_rate": 3.8163765826092626e-05, + "loss": 3.1504, + "step": 12680 + }, + { + "epoch": 0.09812260698903794, + "grad_norm": 0.15971393883228302, + "learning_rate": 3.816147722032108e-05, + "loss": 3.1508, + "step": 12681 + }, + { + "epoch": 0.09815345345461135, + "grad_norm": 0.15388980507850647, + "learning_rate": 3.8159188461950105e-05, + "loss": 3.1242, + "step": 12682 + }, + { + "epoch": 0.09818429992018476, + "grad_norm": 0.14575690031051636, + "learning_rate": 3.815689955100625e-05, + "loss": 3.1224, + "step": 12683 + }, + { + "epoch": 0.09821514638575819, + "grad_norm": 0.14946934580802917, + "learning_rate": 3.8154610487516063e-05, + "loss": 3.1351, + "step": 12684 + }, + { + "epoch": 0.0982459928513316, + "grad_norm": 0.16079649329185486, + "learning_rate": 3.8152321271506065e-05, + "loss": 3.1662, + "step": 12685 + }, + { + "epoch": 0.09827683931690502, + "grad_norm": 0.15284323692321777, + "learning_rate": 3.81500319030028e-05, + "loss": 3.1491, + "step": 12686 + }, + { + "epoch": 0.09830768578247844, + "grad_norm": 0.15194052457809448, + "learning_rate": 3.814774238203283e-05, + "loss": 3.1518, + "step": 12687 + }, + { + "epoch": 0.09833853224805185, + "grad_norm": 0.1501188427209854, + "learning_rate": 3.814545270862267e-05, + "loss": 3.1197, + "step": 12688 + }, + { + "epoch": 0.09836937871362526, + "grad_norm": 0.14471228420734406, + "learning_rate": 3.814316288279891e-05, + "loss": 3.0895, + "step": 12689 + }, + { + "epoch": 0.09840022517919869, + "grad_norm": 0.1486758589744568, + "learning_rate": 3.814087290458806e-05, + "loss": 3.1442, + "step": 12690 + }, + { + "epoch": 0.0984310716447721, + "grad_norm": 0.14636163413524628, + "learning_rate": 3.8138582774016686e-05, + "loss": 3.1359, + "step": 12691 + }, + { + "epoch": 0.09846191811034552, + "grad_norm": 0.14381544291973114, + "learning_rate": 3.8136292491111344e-05, + "loss": 3.1633, + "step": 12692 + }, + { + "epoch": 0.09849276457591893, + "grad_norm": 0.151360422372818, + "learning_rate": 3.8134002055898574e-05, + "loss": 3.1436, + "step": 12693 + }, + { + "epoch": 0.09852361104149235, + "grad_norm": 0.14753705263137817, + "learning_rate": 3.8131711468404954e-05, + "loss": 3.1294, + "step": 12694 + }, + { + "epoch": 0.09855445750706576, + "grad_norm": 0.15061263740062714, + "learning_rate": 3.8129420728657026e-05, + "loss": 3.144, + "step": 12695 + }, + { + "epoch": 0.09858530397263919, + "grad_norm": 0.17685852944850922, + "learning_rate": 3.812712983668136e-05, + "loss": 3.1197, + "step": 12696 + }, + { + "epoch": 0.0986161504382126, + "grad_norm": 0.19684353470802307, + "learning_rate": 3.8124838792504505e-05, + "loss": 3.1687, + "step": 12697 + }, + { + "epoch": 0.09864699690378602, + "grad_norm": 0.15483823418617249, + "learning_rate": 3.812254759615304e-05, + "loss": 3.149, + "step": 12698 + }, + { + "epoch": 0.09867784336935943, + "grad_norm": 0.1470405012369156, + "learning_rate": 3.812025624765351e-05, + "loss": 3.134, + "step": 12699 + }, + { + "epoch": 0.09870868983493285, + "grad_norm": 0.1598513126373291, + "learning_rate": 3.81179647470325e-05, + "loss": 3.1772, + "step": 12700 + }, + { + "epoch": 0.09873953630050626, + "grad_norm": 0.15791456401348114, + "learning_rate": 3.811567309431657e-05, + "loss": 3.1737, + "step": 12701 + }, + { + "epoch": 0.09877038276607969, + "grad_norm": 0.1477932631969452, + "learning_rate": 3.811338128953229e-05, + "loss": 3.1684, + "step": 12702 + }, + { + "epoch": 0.0988012292316531, + "grad_norm": 0.16814535856246948, + "learning_rate": 3.811108933270623e-05, + "loss": 3.1719, + "step": 12703 + }, + { + "epoch": 0.09883207569722652, + "grad_norm": 0.14944979548454285, + "learning_rate": 3.8108797223864976e-05, + "loss": 3.1239, + "step": 12704 + }, + { + "epoch": 0.09886292216279993, + "grad_norm": 0.15837693214416504, + "learning_rate": 3.8106504963035095e-05, + "loss": 3.1534, + "step": 12705 + }, + { + "epoch": 0.09889376862837335, + "grad_norm": 0.14655359089374542, + "learning_rate": 3.8104212550243166e-05, + "loss": 3.1283, + "step": 12706 + }, + { + "epoch": 0.09892461509394676, + "grad_norm": 0.1522364616394043, + "learning_rate": 3.810191998551577e-05, + "loss": 3.1629, + "step": 12707 + }, + { + "epoch": 0.09895546155952019, + "grad_norm": 0.1521579623222351, + "learning_rate": 3.809962726887948e-05, + "loss": 3.1647, + "step": 12708 + }, + { + "epoch": 0.0989863080250936, + "grad_norm": 0.14463616907596588, + "learning_rate": 3.809733440036088e-05, + "loss": 3.1423, + "step": 12709 + }, + { + "epoch": 0.09901715449066702, + "grad_norm": 0.14454615116119385, + "learning_rate": 3.809504137998656e-05, + "loss": 3.1331, + "step": 12710 + }, + { + "epoch": 0.09904800095624043, + "grad_norm": 0.14872391521930695, + "learning_rate": 3.8092748207783104e-05, + "loss": 3.1539, + "step": 12711 + }, + { + "epoch": 0.09907884742181385, + "grad_norm": 0.1410403996706009, + "learning_rate": 3.8090454883777106e-05, + "loss": 3.1412, + "step": 12712 + }, + { + "epoch": 0.09910969388738726, + "grad_norm": 0.146319180727005, + "learning_rate": 3.808816140799515e-05, + "loss": 3.1378, + "step": 12713 + }, + { + "epoch": 0.09914054035296069, + "grad_norm": 0.15277181565761566, + "learning_rate": 3.8085867780463825e-05, + "loss": 3.1229, + "step": 12714 + }, + { + "epoch": 0.0991713868185341, + "grad_norm": 0.14709296822547913, + "learning_rate": 3.8083574001209724e-05, + "loss": 3.1173, + "step": 12715 + }, + { + "epoch": 0.09920223328410752, + "grad_norm": 0.14660526812076569, + "learning_rate": 3.808128007025945e-05, + "loss": 3.1125, + "step": 12716 + }, + { + "epoch": 0.09923307974968093, + "grad_norm": 0.15535427629947662, + "learning_rate": 3.8078985987639595e-05, + "loss": 3.1398, + "step": 12717 + }, + { + "epoch": 0.09926392621525434, + "grad_norm": 0.15304021537303925, + "learning_rate": 3.807669175337675e-05, + "loss": 3.1509, + "step": 12718 + }, + { + "epoch": 0.09929477268082776, + "grad_norm": 0.14875119924545288, + "learning_rate": 3.807439736749753e-05, + "loss": 3.126, + "step": 12719 + }, + { + "epoch": 0.09932561914640119, + "grad_norm": 0.1529451608657837, + "learning_rate": 3.807210283002852e-05, + "loss": 3.167, + "step": 12720 + }, + { + "epoch": 0.0993564656119746, + "grad_norm": 0.14838279783725739, + "learning_rate": 3.806980814099634e-05, + "loss": 3.1842, + "step": 12721 + }, + { + "epoch": 0.09938731207754802, + "grad_norm": 0.15091174840927124, + "learning_rate": 3.806751330042759e-05, + "loss": 3.1796, + "step": 12722 + }, + { + "epoch": 0.09941815854312143, + "grad_norm": 0.14969830214977264, + "learning_rate": 3.806521830834887e-05, + "loss": 3.1436, + "step": 12723 + }, + { + "epoch": 0.09944900500869484, + "grad_norm": 0.1570502668619156, + "learning_rate": 3.80629231647868e-05, + "loss": 3.1352, + "step": 12724 + }, + { + "epoch": 0.09947985147426826, + "grad_norm": 0.16368094086647034, + "learning_rate": 3.806062786976798e-05, + "loss": 3.1544, + "step": 12725 + }, + { + "epoch": 0.09951069793984169, + "grad_norm": 0.1493438482284546, + "learning_rate": 3.8058332423319036e-05, + "loss": 3.1531, + "step": 12726 + }, + { + "epoch": 0.0995415444054151, + "grad_norm": 0.15491105616092682, + "learning_rate": 3.805603682546657e-05, + "loss": 3.1746, + "step": 12727 + }, + { + "epoch": 0.09957239087098851, + "grad_norm": 0.1513214409351349, + "learning_rate": 3.8053741076237204e-05, + "loss": 3.1847, + "step": 12728 + }, + { + "epoch": 0.09960323733656193, + "grad_norm": 0.14993661642074585, + "learning_rate": 3.805144517565755e-05, + "loss": 3.1367, + "step": 12729 + }, + { + "epoch": 0.09963408380213534, + "grad_norm": 0.14844922721385956, + "learning_rate": 3.8049149123754246e-05, + "loss": 3.1599, + "step": 12730 + }, + { + "epoch": 0.09966493026770876, + "grad_norm": 0.1508026123046875, + "learning_rate": 3.804685292055389e-05, + "loss": 3.1632, + "step": 12731 + }, + { + "epoch": 0.09969577673328218, + "grad_norm": 0.14853504300117493, + "learning_rate": 3.8044556566083114e-05, + "loss": 3.1253, + "step": 12732 + }, + { + "epoch": 0.0997266231988556, + "grad_norm": 0.15722082555294037, + "learning_rate": 3.804226006036854e-05, + "loss": 3.1495, + "step": 12733 + }, + { + "epoch": 0.09975746966442901, + "grad_norm": 0.15029260516166687, + "learning_rate": 3.8039963403436806e-05, + "loss": 3.1287, + "step": 12734 + }, + { + "epoch": 0.09978831613000243, + "grad_norm": 0.14503952860832214, + "learning_rate": 3.8037666595314525e-05, + "loss": 3.1467, + "step": 12735 + }, + { + "epoch": 0.09981916259557584, + "grad_norm": 0.15562863647937775, + "learning_rate": 3.803536963602834e-05, + "loss": 3.1576, + "step": 12736 + }, + { + "epoch": 0.09985000906114926, + "grad_norm": 0.1943366527557373, + "learning_rate": 3.803307252560487e-05, + "loss": 3.1338, + "step": 12737 + }, + { + "epoch": 0.09988085552672268, + "grad_norm": 0.16004951298236847, + "learning_rate": 3.803077526407076e-05, + "loss": 3.1669, + "step": 12738 + }, + { + "epoch": 0.0999117019922961, + "grad_norm": 0.14387840032577515, + "learning_rate": 3.802847785145265e-05, + "loss": 3.147, + "step": 12739 + }, + { + "epoch": 0.09994254845786951, + "grad_norm": 0.1530381292104721, + "learning_rate": 3.802618028777717e-05, + "loss": 3.1413, + "step": 12740 + }, + { + "epoch": 0.09997339492344293, + "grad_norm": 0.15696561336517334, + "learning_rate": 3.8023882573070946e-05, + "loss": 3.166, + "step": 12741 + }, + { + "epoch": 0.10000424138901634, + "grad_norm": 0.1560179889202118, + "learning_rate": 3.802158470736063e-05, + "loss": 3.1655, + "step": 12742 + }, + { + "epoch": 0.10003508785458975, + "grad_norm": 0.15336909890174866, + "learning_rate": 3.8019286690672866e-05, + "loss": 3.1711, + "step": 12743 + }, + { + "epoch": 0.10006593432016318, + "grad_norm": 0.15097831189632416, + "learning_rate": 3.80169885230343e-05, + "loss": 3.1402, + "step": 12744 + }, + { + "epoch": 0.1000967807857366, + "grad_norm": 0.1475093960762024, + "learning_rate": 3.801469020447157e-05, + "loss": 3.1436, + "step": 12745 + }, + { + "epoch": 0.10012762725131001, + "grad_norm": 0.15076568722724915, + "learning_rate": 3.801239173501133e-05, + "loss": 3.1702, + "step": 12746 + }, + { + "epoch": 0.10015847371688343, + "grad_norm": 0.175540030002594, + "learning_rate": 3.801009311468023e-05, + "loss": 3.1191, + "step": 12747 + }, + { + "epoch": 0.10018932018245684, + "grad_norm": 0.15419645607471466, + "learning_rate": 3.800779434350491e-05, + "loss": 3.1265, + "step": 12748 + }, + { + "epoch": 0.10022016664803025, + "grad_norm": 0.15237455070018768, + "learning_rate": 3.800549542151204e-05, + "loss": 3.164, + "step": 12749 + }, + { + "epoch": 0.10025101311360368, + "grad_norm": 0.14682596921920776, + "learning_rate": 3.800319634872827e-05, + "loss": 3.1547, + "step": 12750 + }, + { + "epoch": 0.1002818595791771, + "grad_norm": 0.14996635913848877, + "learning_rate": 3.8000897125180244e-05, + "loss": 3.1252, + "step": 12751 + }, + { + "epoch": 0.10031270604475051, + "grad_norm": 0.15924735367298126, + "learning_rate": 3.799859775089463e-05, + "loss": 3.1395, + "step": 12752 + }, + { + "epoch": 0.10034355251032392, + "grad_norm": 0.14617277681827545, + "learning_rate": 3.7996298225898085e-05, + "loss": 3.0982, + "step": 12753 + }, + { + "epoch": 0.10037439897589734, + "grad_norm": 0.14897514879703522, + "learning_rate": 3.799399855021727e-05, + "loss": 3.1049, + "step": 12754 + }, + { + "epoch": 0.10040524544147075, + "grad_norm": 0.1499677449464798, + "learning_rate": 3.799169872387886e-05, + "loss": 3.1609, + "step": 12755 + }, + { + "epoch": 0.10043609190704418, + "grad_norm": 0.17728663980960846, + "learning_rate": 3.79893987469095e-05, + "loss": 3.1423, + "step": 12756 + }, + { + "epoch": 0.1004669383726176, + "grad_norm": 0.2025107741355896, + "learning_rate": 3.798709861933588e-05, + "loss": 3.145, + "step": 12757 + }, + { + "epoch": 0.10049778483819101, + "grad_norm": 0.15890972316265106, + "learning_rate": 3.7984798341184645e-05, + "loss": 3.1642, + "step": 12758 + }, + { + "epoch": 0.10052863130376442, + "grad_norm": 0.1520875096321106, + "learning_rate": 3.798249791248248e-05, + "loss": 3.1487, + "step": 12759 + }, + { + "epoch": 0.10055947776933784, + "grad_norm": 0.1502283364534378, + "learning_rate": 3.798019733325605e-05, + "loss": 3.1504, + "step": 12760 + }, + { + "epoch": 0.10059032423491125, + "grad_norm": 0.14673756062984467, + "learning_rate": 3.7977896603532025e-05, + "loss": 3.1302, + "step": 12761 + }, + { + "epoch": 0.10062117070048468, + "grad_norm": 0.15281610190868378, + "learning_rate": 3.79755957233371e-05, + "loss": 3.1648, + "step": 12762 + }, + { + "epoch": 0.1006520171660581, + "grad_norm": 0.14690393209457397, + "learning_rate": 3.797329469269795e-05, + "loss": 3.1294, + "step": 12763 + }, + { + "epoch": 0.10068286363163151, + "grad_norm": 0.1475202888250351, + "learning_rate": 3.7970993511641226e-05, + "loss": 3.1463, + "step": 12764 + }, + { + "epoch": 0.10071371009720492, + "grad_norm": 0.15320000052452087, + "learning_rate": 3.796869218019362e-05, + "loss": 3.1651, + "step": 12765 + }, + { + "epoch": 0.10074455656277834, + "grad_norm": 0.14761510491371155, + "learning_rate": 3.796639069838184e-05, + "loss": 3.1168, + "step": 12766 + }, + { + "epoch": 0.10077540302835175, + "grad_norm": 0.15335699915885925, + "learning_rate": 3.796408906623254e-05, + "loss": 3.1443, + "step": 12767 + }, + { + "epoch": 0.10080624949392518, + "grad_norm": 0.1542024165391922, + "learning_rate": 3.796178728377243e-05, + "loss": 3.1264, + "step": 12768 + }, + { + "epoch": 0.1008370959594986, + "grad_norm": 0.14532917737960815, + "learning_rate": 3.795948535102818e-05, + "loss": 3.1304, + "step": 12769 + }, + { + "epoch": 0.10086794242507201, + "grad_norm": 0.16020016372203827, + "learning_rate": 3.795718326802648e-05, + "loss": 3.1461, + "step": 12770 + }, + { + "epoch": 0.10089878889064542, + "grad_norm": 0.15399287641048431, + "learning_rate": 3.795488103479403e-05, + "loss": 3.1349, + "step": 12771 + }, + { + "epoch": 0.10092963535621884, + "grad_norm": 0.1557427942752838, + "learning_rate": 3.7952578651357514e-05, + "loss": 3.1063, + "step": 12772 + }, + { + "epoch": 0.10096048182179225, + "grad_norm": 0.15214653313159943, + "learning_rate": 3.795027611774365e-05, + "loss": 3.1605, + "step": 12773 + }, + { + "epoch": 0.10099132828736568, + "grad_norm": 0.15146225690841675, + "learning_rate": 3.79479734339791e-05, + "loss": 3.1529, + "step": 12774 + }, + { + "epoch": 0.10102217475293909, + "grad_norm": 0.1433069109916687, + "learning_rate": 3.7945670600090584e-05, + "loss": 3.1415, + "step": 12775 + }, + { + "epoch": 0.1010530212185125, + "grad_norm": 0.1462544947862625, + "learning_rate": 3.794336761610479e-05, + "loss": 3.1761, + "step": 12776 + }, + { + "epoch": 0.10108386768408592, + "grad_norm": 0.15828542411327362, + "learning_rate": 3.794106448204843e-05, + "loss": 3.1527, + "step": 12777 + }, + { + "epoch": 0.10111471414965933, + "grad_norm": 0.1521095484495163, + "learning_rate": 3.7938761197948215e-05, + "loss": 3.1474, + "step": 12778 + }, + { + "epoch": 0.10114556061523275, + "grad_norm": 0.143866166472435, + "learning_rate": 3.7936457763830826e-05, + "loss": 3.1627, + "step": 12779 + }, + { + "epoch": 0.10117640708080618, + "grad_norm": 0.1509791612625122, + "learning_rate": 3.793415417972299e-05, + "loss": 3.1402, + "step": 12780 + }, + { + "epoch": 0.10120725354637959, + "grad_norm": 0.14340023696422577, + "learning_rate": 3.793185044565141e-05, + "loss": 3.1611, + "step": 12781 + }, + { + "epoch": 0.101238100011953, + "grad_norm": 0.14886273443698883, + "learning_rate": 3.792954656164278e-05, + "loss": 3.1306, + "step": 12782 + }, + { + "epoch": 0.10126894647752642, + "grad_norm": 0.14465057849884033, + "learning_rate": 3.792724252772385e-05, + "loss": 3.1399, + "step": 12783 + }, + { + "epoch": 0.10129979294309983, + "grad_norm": 0.14774399995803833, + "learning_rate": 3.792493834392129e-05, + "loss": 3.1296, + "step": 12784 + }, + { + "epoch": 0.10133063940867325, + "grad_norm": 0.1556495577096939, + "learning_rate": 3.7922634010261846e-05, + "loss": 3.1167, + "step": 12785 + }, + { + "epoch": 0.10136148587424668, + "grad_norm": 0.15989536046981812, + "learning_rate": 3.792032952677224e-05, + "loss": 3.1271, + "step": 12786 + }, + { + "epoch": 0.10139233233982009, + "grad_norm": 0.14487148821353912, + "learning_rate": 3.7918024893479156e-05, + "loss": 3.1534, + "step": 12787 + }, + { + "epoch": 0.1014231788053935, + "grad_norm": 0.15199556946754456, + "learning_rate": 3.791572011040935e-05, + "loss": 3.1458, + "step": 12788 + }, + { + "epoch": 0.10145402527096692, + "grad_norm": 0.1519445925951004, + "learning_rate": 3.791341517758952e-05, + "loss": 3.1775, + "step": 12789 + }, + { + "epoch": 0.10148487173654033, + "grad_norm": 0.15398241579532623, + "learning_rate": 3.791111009504641e-05, + "loss": 3.1395, + "step": 12790 + }, + { + "epoch": 0.10151571820211375, + "grad_norm": 0.14764875173568726, + "learning_rate": 3.7908804862806737e-05, + "loss": 3.1597, + "step": 12791 + }, + { + "epoch": 0.10154656466768718, + "grad_norm": 0.15142476558685303, + "learning_rate": 3.790649948089722e-05, + "loss": 3.1415, + "step": 12792 + }, + { + "epoch": 0.10157741113326059, + "grad_norm": 0.15666238963603973, + "learning_rate": 3.79041939493446e-05, + "loss": 3.1404, + "step": 12793 + }, + { + "epoch": 0.101608257598834, + "grad_norm": 0.14580561220645905, + "learning_rate": 3.790188826817561e-05, + "loss": 3.0796, + "step": 12794 + }, + { + "epoch": 0.10163910406440742, + "grad_norm": 0.14780861139297485, + "learning_rate": 3.789958243741698e-05, + "loss": 3.1323, + "step": 12795 + }, + { + "epoch": 0.10166995052998083, + "grad_norm": 0.19618375599384308, + "learning_rate": 3.789727645709544e-05, + "loss": 3.1685, + "step": 12796 + }, + { + "epoch": 0.10170079699555425, + "grad_norm": 0.15348029136657715, + "learning_rate": 3.7894970327237735e-05, + "loss": 3.1297, + "step": 12797 + }, + { + "epoch": 0.10173164346112767, + "grad_norm": 0.15671736001968384, + "learning_rate": 3.789266404787059e-05, + "loss": 3.1506, + "step": 12798 + }, + { + "epoch": 0.10176248992670109, + "grad_norm": 0.1535063534975052, + "learning_rate": 3.789035761902075e-05, + "loss": 3.1422, + "step": 12799 + }, + { + "epoch": 0.1017933363922745, + "grad_norm": 0.1504669487476349, + "learning_rate": 3.788805104071497e-05, + "loss": 3.1458, + "step": 12800 + }, + { + "epoch": 0.10182418285784792, + "grad_norm": 0.14829809963703156, + "learning_rate": 3.7885744312979984e-05, + "loss": 3.1374, + "step": 12801 + }, + { + "epoch": 0.10185502932342133, + "grad_norm": 0.15366443991661072, + "learning_rate": 3.7883437435842536e-05, + "loss": 3.1383, + "step": 12802 + }, + { + "epoch": 0.10188587578899476, + "grad_norm": 0.24892227351665497, + "learning_rate": 3.788113040932937e-05, + "loss": 3.1479, + "step": 12803 + }, + { + "epoch": 0.10191672225456817, + "grad_norm": 0.15293124318122864, + "learning_rate": 3.7878823233467234e-05, + "loss": 3.1443, + "step": 12804 + }, + { + "epoch": 0.10194756872014159, + "grad_norm": 0.14619788527488708, + "learning_rate": 3.7876515908282886e-05, + "loss": 3.1338, + "step": 12805 + }, + { + "epoch": 0.101978415185715, + "grad_norm": 0.1512361615896225, + "learning_rate": 3.7874208433803065e-05, + "loss": 3.1338, + "step": 12806 + }, + { + "epoch": 0.10200926165128842, + "grad_norm": 0.15174606442451477, + "learning_rate": 3.787190081005454e-05, + "loss": 3.1631, + "step": 12807 + }, + { + "epoch": 0.10204010811686183, + "grad_norm": 0.15615727007389069, + "learning_rate": 3.786959303706406e-05, + "loss": 3.1318, + "step": 12808 + }, + { + "epoch": 0.10207095458243526, + "grad_norm": 0.14979691803455353, + "learning_rate": 3.786728511485839e-05, + "loss": 3.1373, + "step": 12809 + }, + { + "epoch": 0.10210180104800867, + "grad_norm": 0.1515773981809616, + "learning_rate": 3.7864977043464264e-05, + "loss": 3.1508, + "step": 12810 + }, + { + "epoch": 0.10213264751358209, + "grad_norm": 0.14873896539211273, + "learning_rate": 3.786266882290848e-05, + "loss": 3.1479, + "step": 12811 + }, + { + "epoch": 0.1021634939791555, + "grad_norm": 0.1476590782403946, + "learning_rate": 3.786036045321777e-05, + "loss": 3.1336, + "step": 12812 + }, + { + "epoch": 0.10219434044472891, + "grad_norm": 0.1521352380514145, + "learning_rate": 3.7858051934418906e-05, + "loss": 3.133, + "step": 12813 + }, + { + "epoch": 0.10222518691030233, + "grad_norm": 0.14387117326259613, + "learning_rate": 3.7855743266538665e-05, + "loss": 3.1546, + "step": 12814 + }, + { + "epoch": 0.10225603337587576, + "grad_norm": 0.14859530329704285, + "learning_rate": 3.78534344496038e-05, + "loss": 3.181, + "step": 12815 + }, + { + "epoch": 0.10228687984144917, + "grad_norm": 0.15157832205295563, + "learning_rate": 3.7851125483641084e-05, + "loss": 3.1571, + "step": 12816 + }, + { + "epoch": 0.10231772630702259, + "grad_norm": 0.15201734006404877, + "learning_rate": 3.78488163686773e-05, + "loss": 3.1245, + "step": 12817 + }, + { + "epoch": 0.102348572772596, + "grad_norm": 0.15884065628051758, + "learning_rate": 3.7846507104739205e-05, + "loss": 3.1336, + "step": 12818 + }, + { + "epoch": 0.10237941923816941, + "grad_norm": 0.1531878411769867, + "learning_rate": 3.784419769185358e-05, + "loss": 3.1462, + "step": 12819 + }, + { + "epoch": 0.10241026570374283, + "grad_norm": 0.14258383214473724, + "learning_rate": 3.7841888130047206e-05, + "loss": 3.1696, + "step": 12820 + }, + { + "epoch": 0.10244111216931626, + "grad_norm": 0.15058931708335876, + "learning_rate": 3.783957841934685e-05, + "loss": 3.1468, + "step": 12821 + }, + { + "epoch": 0.10247195863488967, + "grad_norm": 0.14779400825500488, + "learning_rate": 3.7837268559779304e-05, + "loss": 3.1455, + "step": 12822 + }, + { + "epoch": 0.10250280510046308, + "grad_norm": 0.1453303098678589, + "learning_rate": 3.783495855137134e-05, + "loss": 3.1375, + "step": 12823 + }, + { + "epoch": 0.1025336515660365, + "grad_norm": 0.14772000908851624, + "learning_rate": 3.7832648394149745e-05, + "loss": 3.1255, + "step": 12824 + }, + { + "epoch": 0.10256449803160991, + "grad_norm": 0.16030409932136536, + "learning_rate": 3.7830338088141307e-05, + "loss": 3.1376, + "step": 12825 + }, + { + "epoch": 0.10259534449718333, + "grad_norm": 0.1451767385005951, + "learning_rate": 3.78280276333728e-05, + "loss": 3.133, + "step": 12826 + }, + { + "epoch": 0.10262619096275676, + "grad_norm": 0.154126837849617, + "learning_rate": 3.782571702987102e-05, + "loss": 3.1679, + "step": 12827 + }, + { + "epoch": 0.10265703742833017, + "grad_norm": 0.15438833832740784, + "learning_rate": 3.782340627766278e-05, + "loss": 3.1922, + "step": 12828 + }, + { + "epoch": 0.10268788389390358, + "grad_norm": 0.1477796584367752, + "learning_rate": 3.7821095376774826e-05, + "loss": 3.1332, + "step": 12829 + }, + { + "epoch": 0.102718730359477, + "grad_norm": 0.15186971426010132, + "learning_rate": 3.781878432723399e-05, + "loss": 3.164, + "step": 12830 + }, + { + "epoch": 0.10274957682505041, + "grad_norm": 0.15183913707733154, + "learning_rate": 3.7816473129067044e-05, + "loss": 3.1397, + "step": 12831 + }, + { + "epoch": 0.10278042329062383, + "grad_norm": 0.15586495399475098, + "learning_rate": 3.781416178230079e-05, + "loss": 3.118, + "step": 12832 + }, + { + "epoch": 0.10281126975619725, + "grad_norm": 0.14526565372943878, + "learning_rate": 3.781185028696205e-05, + "loss": 3.1162, + "step": 12833 + }, + { + "epoch": 0.10284211622177067, + "grad_norm": 0.15042191743850708, + "learning_rate": 3.780953864307759e-05, + "loss": 3.1601, + "step": 12834 + }, + { + "epoch": 0.10287296268734408, + "grad_norm": 0.1508873552083969, + "learning_rate": 3.7807226850674224e-05, + "loss": 3.172, + "step": 12835 + }, + { + "epoch": 0.1029038091529175, + "grad_norm": 0.15335366129875183, + "learning_rate": 3.780491490977877e-05, + "loss": 3.1338, + "step": 12836 + }, + { + "epoch": 0.10293465561849091, + "grad_norm": 0.14483007788658142, + "learning_rate": 3.780260282041802e-05, + "loss": 3.1421, + "step": 12837 + }, + { + "epoch": 0.10296550208406433, + "grad_norm": 0.14386536180973053, + "learning_rate": 3.7800290582618776e-05, + "loss": 3.1353, + "step": 12838 + }, + { + "epoch": 0.10299634854963775, + "grad_norm": 0.15060758590698242, + "learning_rate": 3.7797978196407865e-05, + "loss": 3.1591, + "step": 12839 + }, + { + "epoch": 0.10302719501521117, + "grad_norm": 0.14678971469402313, + "learning_rate": 3.779566566181208e-05, + "loss": 3.1616, + "step": 12840 + }, + { + "epoch": 0.10305804148078458, + "grad_norm": 0.1457720547914505, + "learning_rate": 3.779335297885824e-05, + "loss": 3.1334, + "step": 12841 + }, + { + "epoch": 0.103088887946358, + "grad_norm": 0.14963965117931366, + "learning_rate": 3.779104014757317e-05, + "loss": 3.1854, + "step": 12842 + }, + { + "epoch": 0.10311973441193141, + "grad_norm": 0.15002663433551788, + "learning_rate": 3.778872716798367e-05, + "loss": 3.1448, + "step": 12843 + }, + { + "epoch": 0.10315058087750482, + "grad_norm": 0.14805659651756287, + "learning_rate": 3.778641404011656e-05, + "loss": 3.1261, + "step": 12844 + }, + { + "epoch": 0.10318142734307825, + "grad_norm": 0.14716976881027222, + "learning_rate": 3.778410076399866e-05, + "loss": 3.1376, + "step": 12845 + }, + { + "epoch": 0.10321227380865167, + "grad_norm": 0.157673180103302, + "learning_rate": 3.778178733965681e-05, + "loss": 3.1331, + "step": 12846 + }, + { + "epoch": 0.10324312027422508, + "grad_norm": 0.15285883843898773, + "learning_rate": 3.77794737671178e-05, + "loss": 3.1057, + "step": 12847 + }, + { + "epoch": 0.1032739667397985, + "grad_norm": 0.1471565067768097, + "learning_rate": 3.777716004640848e-05, + "loss": 3.1557, + "step": 12848 + }, + { + "epoch": 0.10330481320537191, + "grad_norm": 0.1455836147069931, + "learning_rate": 3.7774846177555665e-05, + "loss": 3.1369, + "step": 12849 + }, + { + "epoch": 0.10333565967094532, + "grad_norm": 0.15042372047901154, + "learning_rate": 3.777253216058619e-05, + "loss": 3.1415, + "step": 12850 + }, + { + "epoch": 0.10336650613651875, + "grad_norm": 0.14670266211032867, + "learning_rate": 3.7770217995526875e-05, + "loss": 3.1621, + "step": 12851 + }, + { + "epoch": 0.10339735260209217, + "grad_norm": 0.14800520241260529, + "learning_rate": 3.776790368240456e-05, + "loss": 3.1268, + "step": 12852 + }, + { + "epoch": 0.10342819906766558, + "grad_norm": 0.14097808301448822, + "learning_rate": 3.776558922124607e-05, + "loss": 3.1223, + "step": 12853 + }, + { + "epoch": 0.103459045533239, + "grad_norm": 0.14709003269672394, + "learning_rate": 3.7763274612078246e-05, + "loss": 3.1682, + "step": 12854 + }, + { + "epoch": 0.10348989199881241, + "grad_norm": 0.14686131477355957, + "learning_rate": 3.7760959854927926e-05, + "loss": 3.1543, + "step": 12855 + }, + { + "epoch": 0.10352073846438582, + "grad_norm": 0.15402662754058838, + "learning_rate": 3.7758644949821935e-05, + "loss": 3.1522, + "step": 12856 + }, + { + "epoch": 0.10355158492995925, + "grad_norm": 0.14824609458446503, + "learning_rate": 3.7756329896787136e-05, + "loss": 3.1705, + "step": 12857 + }, + { + "epoch": 0.10358243139553266, + "grad_norm": 0.15466569364070892, + "learning_rate": 3.775401469585035e-05, + "loss": 3.1126, + "step": 12858 + }, + { + "epoch": 0.10361327786110608, + "grad_norm": 0.15274667739868164, + "learning_rate": 3.775169934703843e-05, + "loss": 3.1514, + "step": 12859 + }, + { + "epoch": 0.1036441243266795, + "grad_norm": 0.14871960878372192, + "learning_rate": 3.7749383850378214e-05, + "loss": 3.1432, + "step": 12860 + }, + { + "epoch": 0.10367497079225291, + "grad_norm": 0.15686988830566406, + "learning_rate": 3.7747068205896554e-05, + "loss": 3.134, + "step": 12861 + }, + { + "epoch": 0.10370581725782632, + "grad_norm": 0.15053048729896545, + "learning_rate": 3.77447524136203e-05, + "loss": 3.1706, + "step": 12862 + }, + { + "epoch": 0.10373666372339975, + "grad_norm": 0.15536853671073914, + "learning_rate": 3.774243647357631e-05, + "loss": 3.1859, + "step": 12863 + }, + { + "epoch": 0.10376751018897316, + "grad_norm": 0.1456563025712967, + "learning_rate": 3.774012038579142e-05, + "loss": 3.1217, + "step": 12864 + }, + { + "epoch": 0.10379835665454658, + "grad_norm": 0.1531732678413391, + "learning_rate": 3.7737804150292485e-05, + "loss": 3.1862, + "step": 12865 + }, + { + "epoch": 0.10382920312011999, + "grad_norm": 0.14167460799217224, + "learning_rate": 3.773548776710637e-05, + "loss": 3.1236, + "step": 12866 + }, + { + "epoch": 0.1038600495856934, + "grad_norm": 0.14869894087314606, + "learning_rate": 3.7733171236259926e-05, + "loss": 3.127, + "step": 12867 + }, + { + "epoch": 0.10389089605126682, + "grad_norm": 0.14917361736297607, + "learning_rate": 3.7730854557780014e-05, + "loss": 3.1665, + "step": 12868 + }, + { + "epoch": 0.10392174251684025, + "grad_norm": 0.14612939953804016, + "learning_rate": 3.772853773169349e-05, + "loss": 3.148, + "step": 12869 + }, + { + "epoch": 0.10395258898241366, + "grad_norm": 0.14714835584163666, + "learning_rate": 3.772622075802723e-05, + "loss": 3.1483, + "step": 12870 + }, + { + "epoch": 0.10398343544798708, + "grad_norm": 0.1528109610080719, + "learning_rate": 3.772390363680808e-05, + "loss": 3.1401, + "step": 12871 + }, + { + "epoch": 0.10401428191356049, + "grad_norm": 0.14890523254871368, + "learning_rate": 3.772158636806291e-05, + "loss": 3.1542, + "step": 12872 + }, + { + "epoch": 0.1040451283791339, + "grad_norm": 0.15302078425884247, + "learning_rate": 3.7719268951818596e-05, + "loss": 3.1697, + "step": 12873 + }, + { + "epoch": 0.10407597484470732, + "grad_norm": 0.14964498579502106, + "learning_rate": 3.7716951388102e-05, + "loss": 3.1367, + "step": 12874 + }, + { + "epoch": 0.10410682131028075, + "grad_norm": 0.1641964614391327, + "learning_rate": 3.771463367694e-05, + "loss": 3.1709, + "step": 12875 + }, + { + "epoch": 0.10413766777585416, + "grad_norm": 0.14870932698249817, + "learning_rate": 3.771231581835946e-05, + "loss": 3.1131, + "step": 12876 + }, + { + "epoch": 0.10416851424142758, + "grad_norm": 0.15399983525276184, + "learning_rate": 3.7709997812387255e-05, + "loss": 3.156, + "step": 12877 + }, + { + "epoch": 0.10419936070700099, + "grad_norm": 0.15243090689182281, + "learning_rate": 3.770767965905027e-05, + "loss": 3.1419, + "step": 12878 + }, + { + "epoch": 0.1042302071725744, + "grad_norm": 0.15598490834236145, + "learning_rate": 3.770536135837537e-05, + "loss": 3.123, + "step": 12879 + }, + { + "epoch": 0.10426105363814782, + "grad_norm": 0.16360315680503845, + "learning_rate": 3.770304291038944e-05, + "loss": 3.1645, + "step": 12880 + }, + { + "epoch": 0.10429190010372125, + "grad_norm": 0.15878994762897491, + "learning_rate": 3.7700724315119366e-05, + "loss": 3.1163, + "step": 12881 + }, + { + "epoch": 0.10432274656929466, + "grad_norm": 0.18282261490821838, + "learning_rate": 3.769840557259201e-05, + "loss": 3.1495, + "step": 12882 + }, + { + "epoch": 0.10435359303486808, + "grad_norm": 0.14762236177921295, + "learning_rate": 3.769608668283429e-05, + "loss": 3.1368, + "step": 12883 + }, + { + "epoch": 0.10438443950044149, + "grad_norm": 0.3376985490322113, + "learning_rate": 3.769376764587307e-05, + "loss": 3.1712, + "step": 12884 + }, + { + "epoch": 0.1044152859660149, + "grad_norm": 0.15019066631793976, + "learning_rate": 3.769144846173524e-05, + "loss": 3.1678, + "step": 12885 + }, + { + "epoch": 0.10444613243158832, + "grad_norm": 0.1544419229030609, + "learning_rate": 3.76891291304477e-05, + "loss": 3.1261, + "step": 12886 + }, + { + "epoch": 0.10447697889716175, + "grad_norm": 0.1526632010936737, + "learning_rate": 3.7686809652037314e-05, + "loss": 3.1308, + "step": 12887 + }, + { + "epoch": 0.10450782536273516, + "grad_norm": 0.1583424061536789, + "learning_rate": 3.7684490026531004e-05, + "loss": 3.1519, + "step": 12888 + }, + { + "epoch": 0.10453867182830857, + "grad_norm": 0.15495140850543976, + "learning_rate": 3.7682170253955664e-05, + "loss": 3.1451, + "step": 12889 + }, + { + "epoch": 0.10456951829388199, + "grad_norm": 0.14875270426273346, + "learning_rate": 3.7679850334338177e-05, + "loss": 3.1436, + "step": 12890 + }, + { + "epoch": 0.1046003647594554, + "grad_norm": 0.16163843870162964, + "learning_rate": 3.7677530267705436e-05, + "loss": 3.1643, + "step": 12891 + }, + { + "epoch": 0.10463121122502882, + "grad_norm": 0.14517003297805786, + "learning_rate": 3.767521005408435e-05, + "loss": 3.1264, + "step": 12892 + }, + { + "epoch": 0.10466205769060224, + "grad_norm": 0.15269799530506134, + "learning_rate": 3.767288969350183e-05, + "loss": 3.1695, + "step": 12893 + }, + { + "epoch": 0.10469290415617566, + "grad_norm": 0.1603257656097412, + "learning_rate": 3.767056918598476e-05, + "loss": 3.114, + "step": 12894 + }, + { + "epoch": 0.10472375062174907, + "grad_norm": 0.15136195719242096, + "learning_rate": 3.7668248531560064e-05, + "loss": 3.1435, + "step": 12895 + }, + { + "epoch": 0.10475459708732249, + "grad_norm": 0.15501807630062103, + "learning_rate": 3.766592773025464e-05, + "loss": 3.1446, + "step": 12896 + }, + { + "epoch": 0.1047854435528959, + "grad_norm": 0.15777815878391266, + "learning_rate": 3.766360678209538e-05, + "loss": 3.1458, + "step": 12897 + }, + { + "epoch": 0.10481629001846932, + "grad_norm": 0.15737184882164001, + "learning_rate": 3.766128568710922e-05, + "loss": 3.1811, + "step": 12898 + }, + { + "epoch": 0.10484713648404274, + "grad_norm": 0.14823825657367706, + "learning_rate": 3.765896444532307e-05, + "loss": 3.1601, + "step": 12899 + }, + { + "epoch": 0.10487798294961616, + "grad_norm": 0.1496589034795761, + "learning_rate": 3.765664305676381e-05, + "loss": 3.1744, + "step": 12900 + }, + { + "epoch": 0.10490882941518957, + "grad_norm": 0.15963058173656464, + "learning_rate": 3.76543215214584e-05, + "loss": 3.1644, + "step": 12901 + }, + { + "epoch": 0.10493967588076299, + "grad_norm": 0.14945156872272491, + "learning_rate": 3.7651999839433736e-05, + "loss": 3.1423, + "step": 12902 + }, + { + "epoch": 0.1049705223463364, + "grad_norm": 0.14327943325042725, + "learning_rate": 3.764967801071673e-05, + "loss": 3.1607, + "step": 12903 + }, + { + "epoch": 0.10500136881190981, + "grad_norm": 0.15931998193264008, + "learning_rate": 3.7647356035334306e-05, + "loss": 3.1368, + "step": 12904 + }, + { + "epoch": 0.10503221527748324, + "grad_norm": 0.15334799885749817, + "learning_rate": 3.764503391331339e-05, + "loss": 3.0997, + "step": 12905 + }, + { + "epoch": 0.10506306174305666, + "grad_norm": 0.14898188412189484, + "learning_rate": 3.7642711644680904e-05, + "loss": 3.1323, + "step": 12906 + }, + { + "epoch": 0.10509390820863007, + "grad_norm": 0.15386274456977844, + "learning_rate": 3.764038922946378e-05, + "loss": 3.1267, + "step": 12907 + }, + { + "epoch": 0.10512475467420349, + "grad_norm": 0.14598554372787476, + "learning_rate": 3.763806666768894e-05, + "loss": 3.1407, + "step": 12908 + }, + { + "epoch": 0.1051556011397769, + "grad_norm": 0.1543588787317276, + "learning_rate": 3.763574395938331e-05, + "loss": 3.1459, + "step": 12909 + }, + { + "epoch": 0.10518644760535031, + "grad_norm": 0.15527719259262085, + "learning_rate": 3.7633421104573816e-05, + "loss": 3.1517, + "step": 12910 + }, + { + "epoch": 0.10521729407092374, + "grad_norm": 0.1490974873304367, + "learning_rate": 3.76310981032874e-05, + "loss": 3.1644, + "step": 12911 + }, + { + "epoch": 0.10524814053649716, + "grad_norm": 0.1935824304819107, + "learning_rate": 3.7628774955551e-05, + "loss": 3.1723, + "step": 12912 + }, + { + "epoch": 0.10527898700207057, + "grad_norm": 0.15448173880577087, + "learning_rate": 3.7626451661391524e-05, + "loss": 3.1348, + "step": 12913 + }, + { + "epoch": 0.10530983346764398, + "grad_norm": 0.15071555972099304, + "learning_rate": 3.7624128220835944e-05, + "loss": 3.1824, + "step": 12914 + }, + { + "epoch": 0.1053406799332174, + "grad_norm": 0.1544743776321411, + "learning_rate": 3.7621804633911174e-05, + "loss": 3.1781, + "step": 12915 + }, + { + "epoch": 0.10537152639879081, + "grad_norm": 0.1477002203464508, + "learning_rate": 3.7619480900644176e-05, + "loss": 3.1293, + "step": 12916 + }, + { + "epoch": 0.10540237286436424, + "grad_norm": 0.1489584892988205, + "learning_rate": 3.7617157021061866e-05, + "loss": 3.1539, + "step": 12917 + }, + { + "epoch": 0.10543321932993766, + "grad_norm": 0.15239529311656952, + "learning_rate": 3.761483299519122e-05, + "loss": 3.1666, + "step": 12918 + }, + { + "epoch": 0.10546406579551107, + "grad_norm": 0.15155723690986633, + "learning_rate": 3.761250882305916e-05, + "loss": 3.1597, + "step": 12919 + }, + { + "epoch": 0.10549491226108448, + "grad_norm": 0.16056425869464874, + "learning_rate": 3.7610184504692626e-05, + "loss": 3.1126, + "step": 12920 + }, + { + "epoch": 0.1055257587266579, + "grad_norm": 0.16633786261081696, + "learning_rate": 3.760786004011859e-05, + "loss": 3.133, + "step": 12921 + }, + { + "epoch": 0.10555660519223131, + "grad_norm": 0.14905159175395966, + "learning_rate": 3.760553542936399e-05, + "loss": 3.1417, + "step": 12922 + }, + { + "epoch": 0.10558745165780474, + "grad_norm": 0.14855390787124634, + "learning_rate": 3.760321067245579e-05, + "loss": 3.1236, + "step": 12923 + }, + { + "epoch": 0.10561829812337815, + "grad_norm": 0.14557652175426483, + "learning_rate": 3.760088576942094e-05, + "loss": 3.1491, + "step": 12924 + }, + { + "epoch": 0.10564914458895157, + "grad_norm": 0.14751505851745605, + "learning_rate": 3.759856072028638e-05, + "loss": 3.089, + "step": 12925 + }, + { + "epoch": 0.10567999105452498, + "grad_norm": 0.16998152434825897, + "learning_rate": 3.759623552507908e-05, + "loss": 3.1428, + "step": 12926 + }, + { + "epoch": 0.1057108375200984, + "grad_norm": 0.14910948276519775, + "learning_rate": 3.7593910183826e-05, + "loss": 3.1523, + "step": 12927 + }, + { + "epoch": 0.10574168398567181, + "grad_norm": 0.17082205414772034, + "learning_rate": 3.759158469655409e-05, + "loss": 3.1243, + "step": 12928 + }, + { + "epoch": 0.10577253045124524, + "grad_norm": 0.17128412425518036, + "learning_rate": 3.758925906329034e-05, + "loss": 3.1007, + "step": 12929 + }, + { + "epoch": 0.10580337691681865, + "grad_norm": 0.15106850862503052, + "learning_rate": 3.758693328406169e-05, + "loss": 3.1417, + "step": 12930 + }, + { + "epoch": 0.10583422338239207, + "grad_norm": 0.15784721076488495, + "learning_rate": 3.758460735889511e-05, + "loss": 3.1453, + "step": 12931 + }, + { + "epoch": 0.10586506984796548, + "grad_norm": 0.1752839833498001, + "learning_rate": 3.758228128781757e-05, + "loss": 3.1596, + "step": 12932 + }, + { + "epoch": 0.1058959163135389, + "grad_norm": 0.15505871176719666, + "learning_rate": 3.757995507085604e-05, + "loss": 3.1488, + "step": 12933 + }, + { + "epoch": 0.10592676277911231, + "grad_norm": 0.1435016244649887, + "learning_rate": 3.757762870803749e-05, + "loss": 3.1498, + "step": 12934 + }, + { + "epoch": 0.10595760924468574, + "grad_norm": 0.14794713258743286, + "learning_rate": 3.75753021993889e-05, + "loss": 3.1232, + "step": 12935 + }, + { + "epoch": 0.10598845571025915, + "grad_norm": 0.14707614481449127, + "learning_rate": 3.757297554493723e-05, + "loss": 3.161, + "step": 12936 + }, + { + "epoch": 0.10601930217583257, + "grad_norm": 0.1484859436750412, + "learning_rate": 3.757064874470947e-05, + "loss": 3.1282, + "step": 12937 + }, + { + "epoch": 0.10605014864140598, + "grad_norm": 0.155892476439476, + "learning_rate": 3.756832179873259e-05, + "loss": 3.1567, + "step": 12938 + }, + { + "epoch": 0.1060809951069794, + "grad_norm": 0.14796745777130127, + "learning_rate": 3.756599470703356e-05, + "loss": 3.1586, + "step": 12939 + }, + { + "epoch": 0.10611184157255281, + "grad_norm": 0.14966309070587158, + "learning_rate": 3.7563667469639393e-05, + "loss": 3.1201, + "step": 12940 + }, + { + "epoch": 0.10614268803812624, + "grad_norm": 0.14423970878124237, + "learning_rate": 3.756134008657703e-05, + "loss": 3.1421, + "step": 12941 + }, + { + "epoch": 0.10617353450369965, + "grad_norm": 0.14577344059944153, + "learning_rate": 3.7559012557873494e-05, + "loss": 3.1515, + "step": 12942 + }, + { + "epoch": 0.10620438096927307, + "grad_norm": 0.14485585689544678, + "learning_rate": 3.755668488355575e-05, + "loss": 3.1501, + "step": 12943 + }, + { + "epoch": 0.10623522743484648, + "grad_norm": 0.15096233785152435, + "learning_rate": 3.755435706365079e-05, + "loss": 3.1205, + "step": 12944 + }, + { + "epoch": 0.1062660739004199, + "grad_norm": 0.16434434056282043, + "learning_rate": 3.75520290981856e-05, + "loss": 3.1494, + "step": 12945 + }, + { + "epoch": 0.10629692036599331, + "grad_norm": 0.14945831894874573, + "learning_rate": 3.754970098718718e-05, + "loss": 3.1626, + "step": 12946 + }, + { + "epoch": 0.10632776683156674, + "grad_norm": 0.14391754567623138, + "learning_rate": 3.754737273068252e-05, + "loss": 3.1518, + "step": 12947 + }, + { + "epoch": 0.10635861329714015, + "grad_norm": 0.15093103051185608, + "learning_rate": 3.754504432869861e-05, + "loss": 3.152, + "step": 12948 + }, + { + "epoch": 0.10638945976271356, + "grad_norm": 0.15008097887039185, + "learning_rate": 3.754271578126245e-05, + "loss": 3.1132, + "step": 12949 + }, + { + "epoch": 0.10642030622828698, + "grad_norm": 0.1477268487215042, + "learning_rate": 3.754038708840104e-05, + "loss": 3.1104, + "step": 12950 + }, + { + "epoch": 0.10645115269386039, + "grad_norm": 0.1507493257522583, + "learning_rate": 3.753805825014137e-05, + "loss": 3.1458, + "step": 12951 + }, + { + "epoch": 0.10648199915943381, + "grad_norm": 0.15145672857761383, + "learning_rate": 3.753572926651047e-05, + "loss": 3.1359, + "step": 12952 + }, + { + "epoch": 0.10651284562500724, + "grad_norm": 0.15376605093479156, + "learning_rate": 3.75334001375353e-05, + "loss": 3.1529, + "step": 12953 + }, + { + "epoch": 0.10654369209058065, + "grad_norm": 0.15858164429664612, + "learning_rate": 3.753107086324289e-05, + "loss": 3.1252, + "step": 12954 + }, + { + "epoch": 0.10657453855615406, + "grad_norm": 0.15196119248867035, + "learning_rate": 3.7528741443660256e-05, + "loss": 3.1669, + "step": 12955 + }, + { + "epoch": 0.10660538502172748, + "grad_norm": 0.1447891891002655, + "learning_rate": 3.752641187881438e-05, + "loss": 3.1371, + "step": 12956 + }, + { + "epoch": 0.10663623148730089, + "grad_norm": 0.15335425734519958, + "learning_rate": 3.7524082168732297e-05, + "loss": 3.1386, + "step": 12957 + }, + { + "epoch": 0.1066670779528743, + "grad_norm": 0.14571866393089294, + "learning_rate": 3.7521752313441e-05, + "loss": 3.1259, + "step": 12958 + }, + { + "epoch": 0.10669792441844773, + "grad_norm": 0.15914779901504517, + "learning_rate": 3.751942231296751e-05, + "loss": 3.1888, + "step": 12959 + }, + { + "epoch": 0.10672877088402115, + "grad_norm": 0.14806509017944336, + "learning_rate": 3.751709216733884e-05, + "loss": 3.1464, + "step": 12960 + }, + { + "epoch": 0.10675961734959456, + "grad_norm": 0.14524053037166595, + "learning_rate": 3.751476187658201e-05, + "loss": 3.1605, + "step": 12961 + }, + { + "epoch": 0.10679046381516798, + "grad_norm": 0.1509716957807541, + "learning_rate": 3.751243144072404e-05, + "loss": 3.1299, + "step": 12962 + }, + { + "epoch": 0.10682131028074139, + "grad_norm": 0.14829225838184357, + "learning_rate": 3.751010085979195e-05, + "loss": 3.1315, + "step": 12963 + }, + { + "epoch": 0.1068521567463148, + "grad_norm": 0.1476079523563385, + "learning_rate": 3.7507770133812745e-05, + "loss": 3.1478, + "step": 12964 + }, + { + "epoch": 0.10688300321188823, + "grad_norm": 0.14621898531913757, + "learning_rate": 3.750543926281347e-05, + "loss": 3.1363, + "step": 12965 + }, + { + "epoch": 0.10691384967746165, + "grad_norm": 0.14422456920146942, + "learning_rate": 3.7503108246821144e-05, + "loss": 3.1359, + "step": 12966 + }, + { + "epoch": 0.10694469614303506, + "grad_norm": 0.14485995471477509, + "learning_rate": 3.750077708586279e-05, + "loss": 3.1342, + "step": 12967 + }, + { + "epoch": 0.10697554260860848, + "grad_norm": 0.14604292809963226, + "learning_rate": 3.749844577996544e-05, + "loss": 3.1376, + "step": 12968 + }, + { + "epoch": 0.10700638907418189, + "grad_norm": 0.15647734701633453, + "learning_rate": 3.7496114329156116e-05, + "loss": 3.1258, + "step": 12969 + }, + { + "epoch": 0.1070372355397553, + "grad_norm": 0.15013349056243896, + "learning_rate": 3.7493782733461856e-05, + "loss": 3.1126, + "step": 12970 + }, + { + "epoch": 0.10706808200532873, + "grad_norm": 0.14614780247211456, + "learning_rate": 3.7491450992909695e-05, + "loss": 3.1341, + "step": 12971 + }, + { + "epoch": 0.10709892847090215, + "grad_norm": 0.14219114184379578, + "learning_rate": 3.7489119107526664e-05, + "loss": 3.1416, + "step": 12972 + }, + { + "epoch": 0.10712977493647556, + "grad_norm": 0.15557341277599335, + "learning_rate": 3.74867870773398e-05, + "loss": 3.1921, + "step": 12973 + }, + { + "epoch": 0.10716062140204898, + "grad_norm": 0.14838843047618866, + "learning_rate": 3.748445490237615e-05, + "loss": 3.1508, + "step": 12974 + }, + { + "epoch": 0.10719146786762239, + "grad_norm": 0.15321682393550873, + "learning_rate": 3.748212258266275e-05, + "loss": 3.1478, + "step": 12975 + }, + { + "epoch": 0.1072223143331958, + "grad_norm": 0.16673921048641205, + "learning_rate": 3.7479790118226635e-05, + "loss": 3.1388, + "step": 12976 + }, + { + "epoch": 0.10725316079876923, + "grad_norm": 0.16196592152118683, + "learning_rate": 3.7477457509094845e-05, + "loss": 3.1305, + "step": 12977 + }, + { + "epoch": 0.10728400726434265, + "grad_norm": 0.14712122082710266, + "learning_rate": 3.747512475529445e-05, + "loss": 3.1397, + "step": 12978 + }, + { + "epoch": 0.10731485372991606, + "grad_norm": 0.15990136563777924, + "learning_rate": 3.747279185685247e-05, + "loss": 3.1609, + "step": 12979 + }, + { + "epoch": 0.10734570019548947, + "grad_norm": 0.153654545545578, + "learning_rate": 3.747045881379596e-05, + "loss": 3.1755, + "step": 12980 + }, + { + "epoch": 0.10737654666106289, + "grad_norm": 0.1536266803741455, + "learning_rate": 3.7468125626151986e-05, + "loss": 3.1497, + "step": 12981 + }, + { + "epoch": 0.1074073931266363, + "grad_norm": 0.16520357131958008, + "learning_rate": 3.746579229394757e-05, + "loss": 3.1561, + "step": 12982 + }, + { + "epoch": 0.10743823959220973, + "grad_norm": 0.15531271696090698, + "learning_rate": 3.74634588172098e-05, + "loss": 3.1752, + "step": 12983 + }, + { + "epoch": 0.10746908605778314, + "grad_norm": 0.1489976942539215, + "learning_rate": 3.7461125195965715e-05, + "loss": 3.1382, + "step": 12984 + }, + { + "epoch": 0.10749993252335656, + "grad_norm": 0.14954204857349396, + "learning_rate": 3.745879143024236e-05, + "loss": 3.1525, + "step": 12985 + }, + { + "epoch": 0.10753077898892997, + "grad_norm": 0.15238547325134277, + "learning_rate": 3.745645752006681e-05, + "loss": 3.1539, + "step": 12986 + }, + { + "epoch": 0.10756162545450339, + "grad_norm": 0.15654093027114868, + "learning_rate": 3.7454123465466127e-05, + "loss": 3.1139, + "step": 12987 + }, + { + "epoch": 0.1075924719200768, + "grad_norm": 0.1457214057445526, + "learning_rate": 3.745178926646735e-05, + "loss": 3.1142, + "step": 12988 + }, + { + "epoch": 0.10762331838565023, + "grad_norm": 0.14885707199573517, + "learning_rate": 3.744945492309758e-05, + "loss": 3.1447, + "step": 12989 + }, + { + "epoch": 0.10765416485122364, + "grad_norm": 0.14666718244552612, + "learning_rate": 3.744712043538385e-05, + "loss": 3.1331, + "step": 12990 + }, + { + "epoch": 0.10768501131679706, + "grad_norm": 0.14755719900131226, + "learning_rate": 3.744478580335323e-05, + "loss": 3.1461, + "step": 12991 + }, + { + "epoch": 0.10771585778237047, + "grad_norm": 0.15174002945423126, + "learning_rate": 3.744245102703282e-05, + "loss": 3.1438, + "step": 12992 + }, + { + "epoch": 0.10774670424794389, + "grad_norm": 0.15283870697021484, + "learning_rate": 3.7440116106449644e-05, + "loss": 3.1444, + "step": 12993 + }, + { + "epoch": 0.1077775507135173, + "grad_norm": 0.1516609787940979, + "learning_rate": 3.743778104163081e-05, + "loss": 3.1521, + "step": 12994 + }, + { + "epoch": 0.10780839717909073, + "grad_norm": 0.14600752294063568, + "learning_rate": 3.7435445832603376e-05, + "loss": 3.1081, + "step": 12995 + }, + { + "epoch": 0.10783924364466414, + "grad_norm": 0.15473110973834991, + "learning_rate": 3.7433110479394424e-05, + "loss": 3.1366, + "step": 12996 + }, + { + "epoch": 0.10787009011023756, + "grad_norm": 0.1570134311914444, + "learning_rate": 3.743077498203102e-05, + "loss": 3.1262, + "step": 12997 + }, + { + "epoch": 0.10790093657581097, + "grad_norm": 0.1511765718460083, + "learning_rate": 3.7428439340540254e-05, + "loss": 3.1373, + "step": 12998 + }, + { + "epoch": 0.10793178304138439, + "grad_norm": 0.14946985244750977, + "learning_rate": 3.7426103554949206e-05, + "loss": 3.1144, + "step": 12999 + }, + { + "epoch": 0.1079626295069578, + "grad_norm": 0.15988025069236755, + "learning_rate": 3.742376762528495e-05, + "loss": 3.1413, + "step": 13000 + }, + { + "epoch": 0.10799347597253123, + "grad_norm": 0.15355493128299713, + "learning_rate": 3.742143155157458e-05, + "loss": 3.1218, + "step": 13001 + }, + { + "epoch": 0.10802432243810464, + "grad_norm": 0.16174958646297455, + "learning_rate": 3.741909533384517e-05, + "loss": 3.1236, + "step": 13002 + }, + { + "epoch": 0.10805516890367806, + "grad_norm": 0.15269066393375397, + "learning_rate": 3.741675897212381e-05, + "loss": 3.1415, + "step": 13003 + }, + { + "epoch": 0.10808601536925147, + "grad_norm": 0.15593312680721283, + "learning_rate": 3.741442246643759e-05, + "loss": 3.1483, + "step": 13004 + }, + { + "epoch": 0.10811686183482488, + "grad_norm": 0.15437646210193634, + "learning_rate": 3.7412085816813617e-05, + "loss": 3.1033, + "step": 13005 + }, + { + "epoch": 0.1081477083003983, + "grad_norm": 0.48509347438812256, + "learning_rate": 3.740974902327895e-05, + "loss": 3.1171, + "step": 13006 + }, + { + "epoch": 0.10817855476597173, + "grad_norm": 0.1525733917951584, + "learning_rate": 3.740741208586072e-05, + "loss": 3.1298, + "step": 13007 + }, + { + "epoch": 0.10820940123154514, + "grad_norm": 0.15109695494174957, + "learning_rate": 3.740507500458599e-05, + "loss": 3.1206, + "step": 13008 + }, + { + "epoch": 0.10824024769711856, + "grad_norm": 0.1579938679933548, + "learning_rate": 3.740273777948187e-05, + "loss": 3.1144, + "step": 13009 + }, + { + "epoch": 0.10827109416269197, + "grad_norm": 0.1590903103351593, + "learning_rate": 3.740040041057545e-05, + "loss": 3.1451, + "step": 13010 + }, + { + "epoch": 0.10830194062826538, + "grad_norm": 0.1537369340658188, + "learning_rate": 3.7398062897893847e-05, + "loss": 3.0852, + "step": 13011 + }, + { + "epoch": 0.1083327870938388, + "grad_norm": 0.16072028875350952, + "learning_rate": 3.7395725241464164e-05, + "loss": 3.164, + "step": 13012 + }, + { + "epoch": 0.10836363355941223, + "grad_norm": 0.15283165872097015, + "learning_rate": 3.739338744131349e-05, + "loss": 3.1467, + "step": 13013 + }, + { + "epoch": 0.10839448002498564, + "grad_norm": 0.15148229897022247, + "learning_rate": 3.739104949746893e-05, + "loss": 3.1901, + "step": 13014 + }, + { + "epoch": 0.10842532649055905, + "grad_norm": 0.1489398330450058, + "learning_rate": 3.73887114099576e-05, + "loss": 3.1905, + "step": 13015 + }, + { + "epoch": 0.10845617295613247, + "grad_norm": 0.15315070748329163, + "learning_rate": 3.738637317880661e-05, + "loss": 3.1757, + "step": 13016 + }, + { + "epoch": 0.10848701942170588, + "grad_norm": 0.1488935798406601, + "learning_rate": 3.7384034804043063e-05, + "loss": 3.1319, + "step": 13017 + }, + { + "epoch": 0.1085178658872793, + "grad_norm": 0.1537526696920395, + "learning_rate": 3.738169628569408e-05, + "loss": 3.111, + "step": 13018 + }, + { + "epoch": 0.10854871235285272, + "grad_norm": 0.15679208934307098, + "learning_rate": 3.737935762378677e-05, + "loss": 3.1562, + "step": 13019 + }, + { + "epoch": 0.10857955881842614, + "grad_norm": 0.1580590158700943, + "learning_rate": 3.737701881834823e-05, + "loss": 3.1085, + "step": 13020 + }, + { + "epoch": 0.10861040528399955, + "grad_norm": 0.1622033566236496, + "learning_rate": 3.7374679869405604e-05, + "loss": 3.1726, + "step": 13021 + }, + { + "epoch": 0.10864125174957297, + "grad_norm": 0.15981274843215942, + "learning_rate": 3.737234077698601e-05, + "loss": 3.1192, + "step": 13022 + }, + { + "epoch": 0.10867209821514638, + "grad_norm": 0.16518805921077728, + "learning_rate": 3.737000154111655e-05, + "loss": 3.1237, + "step": 13023 + }, + { + "epoch": 0.1087029446807198, + "grad_norm": 0.15980446338653564, + "learning_rate": 3.7367662161824365e-05, + "loss": 3.1279, + "step": 13024 + }, + { + "epoch": 0.10873379114629322, + "grad_norm": 0.15801502764225006, + "learning_rate": 3.7365322639136564e-05, + "loss": 3.1264, + "step": 13025 + }, + { + "epoch": 0.10876463761186664, + "grad_norm": 0.146766796708107, + "learning_rate": 3.736298297308027e-05, + "loss": 3.1502, + "step": 13026 + }, + { + "epoch": 0.10879548407744005, + "grad_norm": 0.16061407327651978, + "learning_rate": 3.736064316368263e-05, + "loss": 3.1487, + "step": 13027 + }, + { + "epoch": 0.10882633054301347, + "grad_norm": 0.17074479162693024, + "learning_rate": 3.735830321097075e-05, + "loss": 3.1566, + "step": 13028 + }, + { + "epoch": 0.10885717700858688, + "grad_norm": 0.15955576300621033, + "learning_rate": 3.735596311497178e-05, + "loss": 3.1727, + "step": 13029 + }, + { + "epoch": 0.1088880234741603, + "grad_norm": 0.15999633073806763, + "learning_rate": 3.735362287571284e-05, + "loss": 3.1158, + "step": 13030 + }, + { + "epoch": 0.10891886993973372, + "grad_norm": 0.15228760242462158, + "learning_rate": 3.7351282493221065e-05, + "loss": 3.1595, + "step": 13031 + }, + { + "epoch": 0.10894971640530714, + "grad_norm": 0.14481689035892487, + "learning_rate": 3.734894196752359e-05, + "loss": 3.137, + "step": 13032 + }, + { + "epoch": 0.10898056287088055, + "grad_norm": 0.15554197132587433, + "learning_rate": 3.7346601298647556e-05, + "loss": 3.1128, + "step": 13033 + }, + { + "epoch": 0.10901140933645397, + "grad_norm": 0.15047979354858398, + "learning_rate": 3.7344260486620094e-05, + "loss": 3.1627, + "step": 13034 + }, + { + "epoch": 0.10904225580202738, + "grad_norm": 0.15300266444683075, + "learning_rate": 3.734191953146836e-05, + "loss": 3.1447, + "step": 13035 + }, + { + "epoch": 0.1090731022676008, + "grad_norm": 0.16803058981895447, + "learning_rate": 3.733957843321948e-05, + "loss": 3.1497, + "step": 13036 + }, + { + "epoch": 0.10910394873317422, + "grad_norm": 0.14827901124954224, + "learning_rate": 3.73372371919006e-05, + "loss": 3.1532, + "step": 13037 + }, + { + "epoch": 0.10913479519874764, + "grad_norm": 0.1560729742050171, + "learning_rate": 3.733489580753887e-05, + "loss": 3.1789, + "step": 13038 + }, + { + "epoch": 0.10916564166432105, + "grad_norm": 0.15057921409606934, + "learning_rate": 3.733255428016144e-05, + "loss": 3.1312, + "step": 13039 + }, + { + "epoch": 0.10919648812989446, + "grad_norm": 0.1525551825761795, + "learning_rate": 3.7330212609795455e-05, + "loss": 3.1663, + "step": 13040 + }, + { + "epoch": 0.10922733459546788, + "grad_norm": 0.1435917168855667, + "learning_rate": 3.732787079646806e-05, + "loss": 3.1814, + "step": 13041 + }, + { + "epoch": 0.10925818106104129, + "grad_norm": 0.15038026869297028, + "learning_rate": 3.7325528840206414e-05, + "loss": 3.1119, + "step": 13042 + }, + { + "epoch": 0.10928902752661472, + "grad_norm": 0.14868082106113434, + "learning_rate": 3.732318674103766e-05, + "loss": 3.148, + "step": 13043 + }, + { + "epoch": 0.10931987399218814, + "grad_norm": 0.1506836861371994, + "learning_rate": 3.732084449898897e-05, + "loss": 3.1005, + "step": 13044 + }, + { + "epoch": 0.10935072045776155, + "grad_norm": 0.14259566366672516, + "learning_rate": 3.7318502114087495e-05, + "loss": 3.1485, + "step": 13045 + }, + { + "epoch": 0.10938156692333496, + "grad_norm": 0.1487756371498108, + "learning_rate": 3.731615958636038e-05, + "loss": 3.1608, + "step": 13046 + }, + { + "epoch": 0.10941241338890838, + "grad_norm": 0.1487146019935608, + "learning_rate": 3.73138169158348e-05, + "loss": 3.1296, + "step": 13047 + }, + { + "epoch": 0.10944325985448179, + "grad_norm": 0.14627978205680847, + "learning_rate": 3.731147410253791e-05, + "loss": 3.1243, + "step": 13048 + }, + { + "epoch": 0.10947410632005522, + "grad_norm": 0.14920973777770996, + "learning_rate": 3.7309131146496886e-05, + "loss": 3.1636, + "step": 13049 + }, + { + "epoch": 0.10950495278562863, + "grad_norm": 0.15539012849330902, + "learning_rate": 3.730678804773887e-05, + "loss": 3.1352, + "step": 13050 + }, + { + "epoch": 0.10953579925120205, + "grad_norm": 0.14970581233501434, + "learning_rate": 3.730444480629106e-05, + "loss": 3.1545, + "step": 13051 + }, + { + "epoch": 0.10956664571677546, + "grad_norm": 0.15305498242378235, + "learning_rate": 3.7302101422180593e-05, + "loss": 3.1173, + "step": 13052 + }, + { + "epoch": 0.10959749218234888, + "grad_norm": 0.16990076005458832, + "learning_rate": 3.7299757895434653e-05, + "loss": 3.1552, + "step": 13053 + }, + { + "epoch": 0.10962833864792229, + "grad_norm": 0.19287467002868652, + "learning_rate": 3.729741422608041e-05, + "loss": 3.1419, + "step": 13054 + }, + { + "epoch": 0.10965918511349572, + "grad_norm": 0.1672748327255249, + "learning_rate": 3.729507041414505e-05, + "loss": 3.137, + "step": 13055 + }, + { + "epoch": 0.10969003157906913, + "grad_norm": 0.16223423182964325, + "learning_rate": 3.7292726459655725e-05, + "loss": 3.1473, + "step": 13056 + }, + { + "epoch": 0.10972087804464255, + "grad_norm": 0.15928350389003754, + "learning_rate": 3.729038236263963e-05, + "loss": 3.1366, + "step": 13057 + }, + { + "epoch": 0.10975172451021596, + "grad_norm": 0.15090951323509216, + "learning_rate": 3.728803812312393e-05, + "loss": 3.177, + "step": 13058 + }, + { + "epoch": 0.10978257097578938, + "grad_norm": 0.14782248437404633, + "learning_rate": 3.7285693741135826e-05, + "loss": 3.1342, + "step": 13059 + }, + { + "epoch": 0.10981341744136279, + "grad_norm": 0.18640559911727905, + "learning_rate": 3.728334921670247e-05, + "loss": 3.1429, + "step": 13060 + }, + { + "epoch": 0.10984426390693622, + "grad_norm": 0.15493501722812653, + "learning_rate": 3.7281004549851074e-05, + "loss": 3.1522, + "step": 13061 + }, + { + "epoch": 0.10987511037250963, + "grad_norm": 0.17452484369277954, + "learning_rate": 3.72786597406088e-05, + "loss": 3.1475, + "step": 13062 + }, + { + "epoch": 0.10990595683808305, + "grad_norm": 0.1576615273952484, + "learning_rate": 3.727631478900286e-05, + "loss": 3.1051, + "step": 13063 + }, + { + "epoch": 0.10993680330365646, + "grad_norm": 0.17551246285438538, + "learning_rate": 3.7273969695060415e-05, + "loss": 3.1594, + "step": 13064 + }, + { + "epoch": 0.10996764976922987, + "grad_norm": 0.15313720703125, + "learning_rate": 3.727162445880866e-05, + "loss": 3.1526, + "step": 13065 + }, + { + "epoch": 0.10999849623480329, + "grad_norm": 0.16661134362220764, + "learning_rate": 3.726927908027481e-05, + "loss": 3.1693, + "step": 13066 + }, + { + "epoch": 0.11002934270037672, + "grad_norm": 0.16590963304042816, + "learning_rate": 3.7266933559486035e-05, + "loss": 3.1287, + "step": 13067 + }, + { + "epoch": 0.11006018916595013, + "grad_norm": 0.15567226707935333, + "learning_rate": 3.726458789646954e-05, + "loss": 3.1043, + "step": 13068 + }, + { + "epoch": 0.11009103563152355, + "grad_norm": 0.17117738723754883, + "learning_rate": 3.726224209125252e-05, + "loss": 3.1412, + "step": 13069 + }, + { + "epoch": 0.11012188209709696, + "grad_norm": 0.1547655463218689, + "learning_rate": 3.7259896143862164e-05, + "loss": 3.151, + "step": 13070 + }, + { + "epoch": 0.11015272856267037, + "grad_norm": 0.16077259182929993, + "learning_rate": 3.725755005432568e-05, + "loss": 3.1279, + "step": 13071 + }, + { + "epoch": 0.11018357502824379, + "grad_norm": 0.15353639423847198, + "learning_rate": 3.725520382267027e-05, + "loss": 3.1697, + "step": 13072 + }, + { + "epoch": 0.11021442149381722, + "grad_norm": 0.15002009272575378, + "learning_rate": 3.725285744892314e-05, + "loss": 3.1154, + "step": 13073 + }, + { + "epoch": 0.11024526795939063, + "grad_norm": 0.15905512869358063, + "learning_rate": 3.725051093311149e-05, + "loss": 3.1629, + "step": 13074 + }, + { + "epoch": 0.11027611442496404, + "grad_norm": 0.1542019546031952, + "learning_rate": 3.7248164275262535e-05, + "loss": 3.1263, + "step": 13075 + }, + { + "epoch": 0.11030696089053746, + "grad_norm": 0.15246888995170593, + "learning_rate": 3.7245817475403464e-05, + "loss": 3.106, + "step": 13076 + }, + { + "epoch": 0.11033780735611087, + "grad_norm": 0.15792734920978546, + "learning_rate": 3.72434705335615e-05, + "loss": 3.1493, + "step": 13077 + }, + { + "epoch": 0.1103686538216843, + "grad_norm": 0.15166029334068298, + "learning_rate": 3.724112344976385e-05, + "loss": 3.154, + "step": 13078 + }, + { + "epoch": 0.11039950028725772, + "grad_norm": 0.15638317167758942, + "learning_rate": 3.7238776224037735e-05, + "loss": 3.1378, + "step": 13079 + }, + { + "epoch": 0.11043034675283113, + "grad_norm": 0.14754177629947662, + "learning_rate": 3.7236428856410366e-05, + "loss": 3.1768, + "step": 13080 + }, + { + "epoch": 0.11046119321840454, + "grad_norm": 0.15352660417556763, + "learning_rate": 3.723408134690895e-05, + "loss": 3.1317, + "step": 13081 + }, + { + "epoch": 0.11049203968397796, + "grad_norm": 0.15244634449481964, + "learning_rate": 3.723173369556072e-05, + "loss": 3.1349, + "step": 13082 + }, + { + "epoch": 0.11052288614955137, + "grad_norm": 0.15818962454795837, + "learning_rate": 3.7229385902392885e-05, + "loss": 3.1516, + "step": 13083 + }, + { + "epoch": 0.1105537326151248, + "grad_norm": 0.15617747604846954, + "learning_rate": 3.722703796743267e-05, + "loss": 3.1523, + "step": 13084 + }, + { + "epoch": 0.11058457908069821, + "grad_norm": 0.15853407979011536, + "learning_rate": 3.72246898907073e-05, + "loss": 3.1448, + "step": 13085 + }, + { + "epoch": 0.11061542554627163, + "grad_norm": 0.15620553493499756, + "learning_rate": 3.7222341672243995e-05, + "loss": 3.1587, + "step": 13086 + }, + { + "epoch": 0.11064627201184504, + "grad_norm": 0.14900456368923187, + "learning_rate": 3.721999331206998e-05, + "loss": 3.1382, + "step": 13087 + }, + { + "epoch": 0.11067711847741846, + "grad_norm": 0.1702354997396469, + "learning_rate": 3.721764481021249e-05, + "loss": 3.1857, + "step": 13088 + }, + { + "epoch": 0.11070796494299187, + "grad_norm": 0.16034331917762756, + "learning_rate": 3.7215296166698745e-05, + "loss": 3.1531, + "step": 13089 + }, + { + "epoch": 0.1107388114085653, + "grad_norm": 0.14912615716457367, + "learning_rate": 3.721294738155599e-05, + "loss": 3.0717, + "step": 13090 + }, + { + "epoch": 0.11076965787413871, + "grad_norm": 0.3456840515136719, + "learning_rate": 3.721059845481144e-05, + "loss": 3.1458, + "step": 13091 + }, + { + "epoch": 0.11080050433971213, + "grad_norm": 0.14918828010559082, + "learning_rate": 3.7208249386492346e-05, + "loss": 3.1482, + "step": 13092 + }, + { + "epoch": 0.11083135080528554, + "grad_norm": 0.16380609571933746, + "learning_rate": 3.720590017662593e-05, + "loss": 3.1274, + "step": 13093 + }, + { + "epoch": 0.11086219727085896, + "grad_norm": 0.16327762603759766, + "learning_rate": 3.720355082523944e-05, + "loss": 3.1383, + "step": 13094 + }, + { + "epoch": 0.11089304373643237, + "grad_norm": 0.14840762317180634, + "learning_rate": 3.720120133236011e-05, + "loss": 3.1053, + "step": 13095 + }, + { + "epoch": 0.1109238902020058, + "grad_norm": 0.17130383849143982, + "learning_rate": 3.719885169801519e-05, + "loss": 3.1531, + "step": 13096 + }, + { + "epoch": 0.11095473666757921, + "grad_norm": 0.14718881249427795, + "learning_rate": 3.719650192223191e-05, + "loss": 3.1048, + "step": 13097 + }, + { + "epoch": 0.11098558313315263, + "grad_norm": 0.16304758191108704, + "learning_rate": 3.719415200503752e-05, + "loss": 3.1732, + "step": 13098 + }, + { + "epoch": 0.11101642959872604, + "grad_norm": 0.15570159256458282, + "learning_rate": 3.719180194645926e-05, + "loss": 3.1359, + "step": 13099 + }, + { + "epoch": 0.11104727606429945, + "grad_norm": 0.16401228308677673, + "learning_rate": 3.718945174652439e-05, + "loss": 3.1508, + "step": 13100 + }, + { + "epoch": 0.11107812252987287, + "grad_norm": 0.16033495962619781, + "learning_rate": 3.718710140526015e-05, + "loss": 3.1355, + "step": 13101 + }, + { + "epoch": 0.1111089689954463, + "grad_norm": 0.15676894783973694, + "learning_rate": 3.718475092269379e-05, + "loss": 3.1606, + "step": 13102 + }, + { + "epoch": 0.11113981546101971, + "grad_norm": 0.16385473310947418, + "learning_rate": 3.718240029885257e-05, + "loss": 3.1494, + "step": 13103 + }, + { + "epoch": 0.11117066192659313, + "grad_norm": 0.14822913706302643, + "learning_rate": 3.7180049533763735e-05, + "loss": 3.1057, + "step": 13104 + }, + { + "epoch": 0.11120150839216654, + "grad_norm": 0.15399128198623657, + "learning_rate": 3.7177698627454546e-05, + "loss": 3.1448, + "step": 13105 + }, + { + "epoch": 0.11123235485773995, + "grad_norm": 0.15810047090053558, + "learning_rate": 3.7175347579952256e-05, + "loss": 3.1345, + "step": 13106 + }, + { + "epoch": 0.11126320132331337, + "grad_norm": 0.15368303656578064, + "learning_rate": 3.717299639128414e-05, + "loss": 3.1278, + "step": 13107 + }, + { + "epoch": 0.1112940477888868, + "grad_norm": 0.17456524074077606, + "learning_rate": 3.717064506147744e-05, + "loss": 3.1294, + "step": 13108 + }, + { + "epoch": 0.11132489425446021, + "grad_norm": 0.1498059332370758, + "learning_rate": 3.7168293590559416e-05, + "loss": 3.1382, + "step": 13109 + }, + { + "epoch": 0.11135574072003362, + "grad_norm": 0.16597341001033783, + "learning_rate": 3.7165941978557345e-05, + "loss": 3.1873, + "step": 13110 + }, + { + "epoch": 0.11138658718560704, + "grad_norm": 0.1815870851278305, + "learning_rate": 3.7163590225498484e-05, + "loss": 3.1485, + "step": 13111 + }, + { + "epoch": 0.11141743365118045, + "grad_norm": 0.1622888147830963, + "learning_rate": 3.716123833141011e-05, + "loss": 3.1844, + "step": 13112 + }, + { + "epoch": 0.11144828011675387, + "grad_norm": 0.14902852475643158, + "learning_rate": 3.715888629631949e-05, + "loss": 3.1563, + "step": 13113 + }, + { + "epoch": 0.1114791265823273, + "grad_norm": 0.1485525369644165, + "learning_rate": 3.7156534120253886e-05, + "loss": 3.1436, + "step": 13114 + }, + { + "epoch": 0.11150997304790071, + "grad_norm": 0.17033667862415314, + "learning_rate": 3.7154181803240563e-05, + "loss": 3.1415, + "step": 13115 + }, + { + "epoch": 0.11154081951347412, + "grad_norm": 0.14625494182109833, + "learning_rate": 3.7151829345306816e-05, + "loss": 3.1561, + "step": 13116 + }, + { + "epoch": 0.11157166597904754, + "grad_norm": 0.15181927382946014, + "learning_rate": 3.7149476746479906e-05, + "loss": 3.1131, + "step": 13117 + }, + { + "epoch": 0.11160251244462095, + "grad_norm": 0.17534399032592773, + "learning_rate": 3.714712400678712e-05, + "loss": 3.1668, + "step": 13118 + }, + { + "epoch": 0.11163335891019437, + "grad_norm": 0.15259188413619995, + "learning_rate": 3.714477112625573e-05, + "loss": 3.1437, + "step": 13119 + }, + { + "epoch": 0.1116642053757678, + "grad_norm": 0.15233513712882996, + "learning_rate": 3.7142418104913014e-05, + "loss": 3.0999, + "step": 13120 + }, + { + "epoch": 0.11169505184134121, + "grad_norm": 0.161477729678154, + "learning_rate": 3.7140064942786256e-05, + "loss": 3.1115, + "step": 13121 + }, + { + "epoch": 0.11172589830691462, + "grad_norm": 0.16196522116661072, + "learning_rate": 3.713771163990275e-05, + "loss": 3.1459, + "step": 13122 + }, + { + "epoch": 0.11175674477248804, + "grad_norm": 0.1497841626405716, + "learning_rate": 3.713535819628976e-05, + "loss": 3.1724, + "step": 13123 + }, + { + "epoch": 0.11178759123806145, + "grad_norm": 0.15410590171813965, + "learning_rate": 3.7133004611974585e-05, + "loss": 3.1265, + "step": 13124 + }, + { + "epoch": 0.11181843770363487, + "grad_norm": 0.159048393368721, + "learning_rate": 3.713065088698452e-05, + "loss": 3.104, + "step": 13125 + }, + { + "epoch": 0.1118492841692083, + "grad_norm": 0.1456906944513321, + "learning_rate": 3.712829702134683e-05, + "loss": 3.1655, + "step": 13126 + }, + { + "epoch": 0.11188013063478171, + "grad_norm": 0.15535040199756622, + "learning_rate": 3.712594301508884e-05, + "loss": 3.1304, + "step": 13127 + }, + { + "epoch": 0.11191097710035512, + "grad_norm": 0.5166134238243103, + "learning_rate": 3.712358886823783e-05, + "loss": 3.1169, + "step": 13128 + }, + { + "epoch": 0.11194182356592854, + "grad_norm": 0.15423816442489624, + "learning_rate": 3.712123458082108e-05, + "loss": 3.1301, + "step": 13129 + }, + { + "epoch": 0.11197267003150195, + "grad_norm": 0.15134918689727783, + "learning_rate": 3.7118880152865906e-05, + "loss": 3.15, + "step": 13130 + }, + { + "epoch": 0.11200351649707536, + "grad_norm": 0.14693774282932281, + "learning_rate": 3.7116525584399606e-05, + "loss": 3.121, + "step": 13131 + }, + { + "epoch": 0.11203436296264879, + "grad_norm": 0.15223880112171173, + "learning_rate": 3.711417087544946e-05, + "loss": 3.197, + "step": 13132 + }, + { + "epoch": 0.1120652094282222, + "grad_norm": 0.1511206328868866, + "learning_rate": 3.711181602604279e-05, + "loss": 3.1402, + "step": 13133 + }, + { + "epoch": 0.11209605589379562, + "grad_norm": 0.14674586057662964, + "learning_rate": 3.710946103620689e-05, + "loss": 3.1616, + "step": 13134 + }, + { + "epoch": 0.11212690235936904, + "grad_norm": 0.14659611880779266, + "learning_rate": 3.710710590596907e-05, + "loss": 3.1456, + "step": 13135 + }, + { + "epoch": 0.11215774882494245, + "grad_norm": 0.1450577825307846, + "learning_rate": 3.7104750635356636e-05, + "loss": 3.1099, + "step": 13136 + }, + { + "epoch": 0.11218859529051586, + "grad_norm": 0.1770365983247757, + "learning_rate": 3.710239522439688e-05, + "loss": 3.1619, + "step": 13137 + }, + { + "epoch": 0.11221944175608929, + "grad_norm": 0.15332838892936707, + "learning_rate": 3.710003967311713e-05, + "loss": 3.147, + "step": 13138 + }, + { + "epoch": 0.1122502882216627, + "grad_norm": 0.15036195516586304, + "learning_rate": 3.709768398154469e-05, + "loss": 3.1285, + "step": 13139 + }, + { + "epoch": 0.11228113468723612, + "grad_norm": 0.14771127700805664, + "learning_rate": 3.7095328149706873e-05, + "loss": 3.1097, + "step": 13140 + }, + { + "epoch": 0.11231198115280953, + "grad_norm": 0.16044747829437256, + "learning_rate": 3.7092972177631e-05, + "loss": 3.1278, + "step": 13141 + }, + { + "epoch": 0.11234282761838295, + "grad_norm": 0.148961141705513, + "learning_rate": 3.709061606534438e-05, + "loss": 3.1316, + "step": 13142 + }, + { + "epoch": 0.11237367408395636, + "grad_norm": 0.15536010265350342, + "learning_rate": 3.708825981287433e-05, + "loss": 3.1421, + "step": 13143 + }, + { + "epoch": 0.11240452054952979, + "grad_norm": 0.14778383076190948, + "learning_rate": 3.708590342024817e-05, + "loss": 3.1425, + "step": 13144 + }, + { + "epoch": 0.1124353670151032, + "grad_norm": 0.14840546250343323, + "learning_rate": 3.708354688749323e-05, + "loss": 3.1537, + "step": 13145 + }, + { + "epoch": 0.11246621348067662, + "grad_norm": 0.1503094583749771, + "learning_rate": 3.708119021463682e-05, + "loss": 3.147, + "step": 13146 + }, + { + "epoch": 0.11249705994625003, + "grad_norm": 0.15528331696987152, + "learning_rate": 3.707883340170626e-05, + "loss": 3.1195, + "step": 13147 + }, + { + "epoch": 0.11252790641182345, + "grad_norm": 0.15297766029834747, + "learning_rate": 3.7076476448728895e-05, + "loss": 3.1283, + "step": 13148 + }, + { + "epoch": 0.11255875287739686, + "grad_norm": 0.1559963971376419, + "learning_rate": 3.707411935573205e-05, + "loss": 3.1525, + "step": 13149 + }, + { + "epoch": 0.11258959934297029, + "grad_norm": 0.14722959697246552, + "learning_rate": 3.707176212274303e-05, + "loss": 3.1398, + "step": 13150 + }, + { + "epoch": 0.1126204458085437, + "grad_norm": 0.1506386697292328, + "learning_rate": 3.706940474978919e-05, + "loss": 3.1304, + "step": 13151 + }, + { + "epoch": 0.11265129227411712, + "grad_norm": 0.14643269777297974, + "learning_rate": 3.706704723689787e-05, + "loss": 3.1531, + "step": 13152 + }, + { + "epoch": 0.11268213873969053, + "grad_norm": 0.15510809421539307, + "learning_rate": 3.706468958409637e-05, + "loss": 3.1669, + "step": 13153 + }, + { + "epoch": 0.11271298520526395, + "grad_norm": 0.14306022226810455, + "learning_rate": 3.7062331791412045e-05, + "loss": 3.1266, + "step": 13154 + }, + { + "epoch": 0.11274383167083736, + "grad_norm": 0.14765708148479462, + "learning_rate": 3.705997385887224e-05, + "loss": 3.1282, + "step": 13155 + }, + { + "epoch": 0.11277467813641079, + "grad_norm": 0.1509227752685547, + "learning_rate": 3.705761578650428e-05, + "loss": 3.1533, + "step": 13156 + }, + { + "epoch": 0.1128055246019842, + "grad_norm": 0.15731321275234222, + "learning_rate": 3.7055257574335514e-05, + "loss": 3.1761, + "step": 13157 + }, + { + "epoch": 0.11283637106755762, + "grad_norm": 0.14891837537288666, + "learning_rate": 3.705289922239328e-05, + "loss": 3.1612, + "step": 13158 + }, + { + "epoch": 0.11286721753313103, + "grad_norm": 0.14732401072978973, + "learning_rate": 3.705054073070492e-05, + "loss": 3.1087, + "step": 13159 + }, + { + "epoch": 0.11289806399870445, + "grad_norm": 0.18119971454143524, + "learning_rate": 3.704818209929778e-05, + "loss": 3.1506, + "step": 13160 + }, + { + "epoch": 0.11292891046427786, + "grad_norm": 0.15071918070316315, + "learning_rate": 3.7045823328199205e-05, + "loss": 3.1492, + "step": 13161 + }, + { + "epoch": 0.11295975692985129, + "grad_norm": 0.15103955566883087, + "learning_rate": 3.704346441743656e-05, + "loss": 3.1248, + "step": 13162 + }, + { + "epoch": 0.1129906033954247, + "grad_norm": 0.14946551620960236, + "learning_rate": 3.704110536703718e-05, + "loss": 3.1659, + "step": 13163 + }, + { + "epoch": 0.11302144986099812, + "grad_norm": 0.15252427756786346, + "learning_rate": 3.7038746177028416e-05, + "loss": 3.1667, + "step": 13164 + }, + { + "epoch": 0.11305229632657153, + "grad_norm": 0.17006082832813263, + "learning_rate": 3.7036386847437624e-05, + "loss": 3.1327, + "step": 13165 + }, + { + "epoch": 0.11308314279214494, + "grad_norm": 0.14766772091388702, + "learning_rate": 3.703402737829216e-05, + "loss": 3.1124, + "step": 13166 + }, + { + "epoch": 0.11311398925771836, + "grad_norm": 0.15290839970111847, + "learning_rate": 3.703166776961937e-05, + "loss": 3.1427, + "step": 13167 + }, + { + "epoch": 0.11314483572329179, + "grad_norm": 0.15444907546043396, + "learning_rate": 3.702930802144664e-05, + "loss": 3.1431, + "step": 13168 + }, + { + "epoch": 0.1131756821888652, + "grad_norm": 0.1486058235168457, + "learning_rate": 3.70269481338013e-05, + "loss": 3.143, + "step": 13169 + }, + { + "epoch": 0.11320652865443862, + "grad_norm": 0.16966190934181213, + "learning_rate": 3.7024588106710734e-05, + "loss": 3.13, + "step": 13170 + }, + { + "epoch": 0.11323737512001203, + "grad_norm": 0.163986936211586, + "learning_rate": 3.702222794020228e-05, + "loss": 3.1589, + "step": 13171 + }, + { + "epoch": 0.11326822158558544, + "grad_norm": 0.15166479349136353, + "learning_rate": 3.701986763430334e-05, + "loss": 3.1315, + "step": 13172 + }, + { + "epoch": 0.11329906805115886, + "grad_norm": 0.15086697041988373, + "learning_rate": 3.7017507189041235e-05, + "loss": 3.1466, + "step": 13173 + }, + { + "epoch": 0.11332991451673229, + "grad_norm": 0.14787191152572632, + "learning_rate": 3.7015146604443365e-05, + "loss": 3.1094, + "step": 13174 + }, + { + "epoch": 0.1133607609823057, + "grad_norm": 0.15776999294757843, + "learning_rate": 3.7012785880537096e-05, + "loss": 3.1457, + "step": 13175 + }, + { + "epoch": 0.11339160744787911, + "grad_norm": 0.1488649845123291, + "learning_rate": 3.701042501734978e-05, + "loss": 3.1165, + "step": 13176 + }, + { + "epoch": 0.11342245391345253, + "grad_norm": 0.14855758845806122, + "learning_rate": 3.7008064014908814e-05, + "loss": 3.1335, + "step": 13177 + }, + { + "epoch": 0.11345330037902594, + "grad_norm": 0.15423201024532318, + "learning_rate": 3.700570287324156e-05, + "loss": 3.1649, + "step": 13178 + }, + { + "epoch": 0.11348414684459936, + "grad_norm": 0.14998042583465576, + "learning_rate": 3.7003341592375396e-05, + "loss": 3.1827, + "step": 13179 + }, + { + "epoch": 0.11351499331017278, + "grad_norm": 0.14670751988887787, + "learning_rate": 3.70009801723377e-05, + "loss": 3.1496, + "step": 13180 + }, + { + "epoch": 0.1135458397757462, + "grad_norm": 0.14793753623962402, + "learning_rate": 3.6998618613155845e-05, + "loss": 3.1215, + "step": 13181 + }, + { + "epoch": 0.11357668624131961, + "grad_norm": 0.1625836342573166, + "learning_rate": 3.6996256914857216e-05, + "loss": 3.1911, + "step": 13182 + }, + { + "epoch": 0.11360753270689303, + "grad_norm": 0.1794232279062271, + "learning_rate": 3.699389507746921e-05, + "loss": 3.136, + "step": 13183 + }, + { + "epoch": 0.11363837917246644, + "grad_norm": 0.1555781364440918, + "learning_rate": 3.699153310101919e-05, + "loss": 3.1433, + "step": 13184 + }, + { + "epoch": 0.11366922563803986, + "grad_norm": 0.1498391479253769, + "learning_rate": 3.698917098553455e-05, + "loss": 3.1362, + "step": 13185 + }, + { + "epoch": 0.11370007210361328, + "grad_norm": 0.15176646411418915, + "learning_rate": 3.698680873104268e-05, + "loss": 3.1335, + "step": 13186 + }, + { + "epoch": 0.1137309185691867, + "grad_norm": 0.16711343824863434, + "learning_rate": 3.698444633757096e-05, + "loss": 3.1477, + "step": 13187 + }, + { + "epoch": 0.11376176503476011, + "grad_norm": 0.15314048528671265, + "learning_rate": 3.698208380514679e-05, + "loss": 3.1639, + "step": 13188 + }, + { + "epoch": 0.11379261150033353, + "grad_norm": 0.14804552495479584, + "learning_rate": 3.697972113379756e-05, + "loss": 3.1456, + "step": 13189 + }, + { + "epoch": 0.11382345796590694, + "grad_norm": 0.15726153552532196, + "learning_rate": 3.697735832355066e-05, + "loss": 3.1476, + "step": 13190 + }, + { + "epoch": 0.11385430443148035, + "grad_norm": 0.1511615365743637, + "learning_rate": 3.697499537443349e-05, + "loss": 3.0896, + "step": 13191 + }, + { + "epoch": 0.11388515089705378, + "grad_norm": 0.14989417791366577, + "learning_rate": 3.697263228647345e-05, + "loss": 3.1357, + "step": 13192 + }, + { + "epoch": 0.1139159973626272, + "grad_norm": 0.1551560014486313, + "learning_rate": 3.697026905969792e-05, + "loss": 3.1098, + "step": 13193 + }, + { + "epoch": 0.11394684382820061, + "grad_norm": 0.14561684429645538, + "learning_rate": 3.696790569413432e-05, + "loss": 3.1652, + "step": 13194 + }, + { + "epoch": 0.11397769029377403, + "grad_norm": 0.16314950585365295, + "learning_rate": 3.6965542189810046e-05, + "loss": 3.1042, + "step": 13195 + }, + { + "epoch": 0.11400853675934744, + "grad_norm": 0.1520436704158783, + "learning_rate": 3.696317854675251e-05, + "loss": 3.1184, + "step": 13196 + }, + { + "epoch": 0.11403938322492085, + "grad_norm": 0.16105304658412933, + "learning_rate": 3.696081476498909e-05, + "loss": 3.1457, + "step": 13197 + }, + { + "epoch": 0.11407022969049428, + "grad_norm": 0.1463477462530136, + "learning_rate": 3.695845084454722e-05, + "loss": 3.1386, + "step": 13198 + }, + { + "epoch": 0.1141010761560677, + "grad_norm": 0.1587296575307846, + "learning_rate": 3.6956086785454294e-05, + "loss": 3.1531, + "step": 13199 + }, + { + "epoch": 0.11413192262164111, + "grad_norm": 0.15055620670318604, + "learning_rate": 3.695372258773773e-05, + "loss": 3.1378, + "step": 13200 + }, + { + "epoch": 0.11416276908721452, + "grad_norm": 0.14320354163646698, + "learning_rate": 3.695135825142494e-05, + "loss": 3.1548, + "step": 13201 + }, + { + "epoch": 0.11419361555278794, + "grad_norm": 0.16090922057628632, + "learning_rate": 3.6948993776543335e-05, + "loss": 3.1612, + "step": 13202 + }, + { + "epoch": 0.11422446201836135, + "grad_norm": 0.1549932360649109, + "learning_rate": 3.694662916312032e-05, + "loss": 3.1559, + "step": 13203 + }, + { + "epoch": 0.11425530848393478, + "grad_norm": 0.15526644885540009, + "learning_rate": 3.6944264411183314e-05, + "loss": 3.1215, + "step": 13204 + }, + { + "epoch": 0.1142861549495082, + "grad_norm": 0.1571066677570343, + "learning_rate": 3.694189952075975e-05, + "loss": 3.1037, + "step": 13205 + }, + { + "epoch": 0.11431700141508161, + "grad_norm": 0.14633134007453918, + "learning_rate": 3.6939534491877034e-05, + "loss": 3.104, + "step": 13206 + }, + { + "epoch": 0.11434784788065502, + "grad_norm": 0.147661030292511, + "learning_rate": 3.693716932456259e-05, + "loss": 3.1207, + "step": 13207 + }, + { + "epoch": 0.11437869434622844, + "grad_norm": 0.14739462733268738, + "learning_rate": 3.693480401884384e-05, + "loss": 3.1348, + "step": 13208 + }, + { + "epoch": 0.11440954081180185, + "grad_norm": 0.14800049364566803, + "learning_rate": 3.6932438574748206e-05, + "loss": 3.1288, + "step": 13209 + }, + { + "epoch": 0.11444038727737528, + "grad_norm": 0.243785098195076, + "learning_rate": 3.693007299230313e-05, + "loss": 3.1104, + "step": 13210 + }, + { + "epoch": 0.1144712337429487, + "grad_norm": 0.17844656109809875, + "learning_rate": 3.692770727153601e-05, + "loss": 3.1389, + "step": 13211 + }, + { + "epoch": 0.11450208020852211, + "grad_norm": 0.16832700371742249, + "learning_rate": 3.6925341412474304e-05, + "loss": 3.1206, + "step": 13212 + }, + { + "epoch": 0.11453292667409552, + "grad_norm": 0.15224792063236237, + "learning_rate": 3.6922975415145434e-05, + "loss": 3.1546, + "step": 13213 + }, + { + "epoch": 0.11456377313966894, + "grad_norm": 0.14983907341957092, + "learning_rate": 3.6920609279576816e-05, + "loss": 3.0949, + "step": 13214 + }, + { + "epoch": 0.11459461960524235, + "grad_norm": 0.15790969133377075, + "learning_rate": 3.69182430057959e-05, + "loss": 3.1664, + "step": 13215 + }, + { + "epoch": 0.11462546607081578, + "grad_norm": 0.1497974842786789, + "learning_rate": 3.6915876593830125e-05, + "loss": 3.1576, + "step": 13216 + }, + { + "epoch": 0.1146563125363892, + "grad_norm": 0.15178467333316803, + "learning_rate": 3.691351004370692e-05, + "loss": 3.1383, + "step": 13217 + }, + { + "epoch": 0.11468715900196261, + "grad_norm": 0.14729313552379608, + "learning_rate": 3.691114335545372e-05, + "loss": 3.1609, + "step": 13218 + }, + { + "epoch": 0.11471800546753602, + "grad_norm": 0.1538219004869461, + "learning_rate": 3.6908776529097974e-05, + "loss": 3.1617, + "step": 13219 + }, + { + "epoch": 0.11474885193310944, + "grad_norm": 0.15006782114505768, + "learning_rate": 3.690640956466712e-05, + "loss": 3.0847, + "step": 13220 + }, + { + "epoch": 0.11477969839868285, + "grad_norm": 0.1509031504392624, + "learning_rate": 3.69040424621886e-05, + "loss": 3.1488, + "step": 13221 + }, + { + "epoch": 0.11481054486425628, + "grad_norm": 0.16188842058181763, + "learning_rate": 3.690167522168987e-05, + "loss": 3.1586, + "step": 13222 + }, + { + "epoch": 0.11484139132982969, + "grad_norm": 0.1665712147951126, + "learning_rate": 3.6899307843198355e-05, + "loss": 3.1304, + "step": 13223 + }, + { + "epoch": 0.1148722377954031, + "grad_norm": 0.1480393409729004, + "learning_rate": 3.689694032674153e-05, + "loss": 3.142, + "step": 13224 + }, + { + "epoch": 0.11490308426097652, + "grad_norm": 0.1745680570602417, + "learning_rate": 3.689457267234682e-05, + "loss": 3.128, + "step": 13225 + }, + { + "epoch": 0.11493393072654993, + "grad_norm": 0.15959985554218292, + "learning_rate": 3.689220488004169e-05, + "loss": 3.1284, + "step": 13226 + }, + { + "epoch": 0.11496477719212335, + "grad_norm": 0.17370936274528503, + "learning_rate": 3.688983694985359e-05, + "loss": 3.1276, + "step": 13227 + }, + { + "epoch": 0.11499562365769678, + "grad_norm": 0.1600739061832428, + "learning_rate": 3.688746888180998e-05, + "loss": 3.1292, + "step": 13228 + }, + { + "epoch": 0.11502647012327019, + "grad_norm": 0.15587666630744934, + "learning_rate": 3.6885100675938306e-05, + "loss": 3.1343, + "step": 13229 + }, + { + "epoch": 0.1150573165888436, + "grad_norm": 0.15810997784137726, + "learning_rate": 3.688273233226604e-05, + "loss": 3.1315, + "step": 13230 + }, + { + "epoch": 0.11508816305441702, + "grad_norm": 0.14867641031742096, + "learning_rate": 3.688036385082063e-05, + "loss": 3.1342, + "step": 13231 + }, + { + "epoch": 0.11511900951999043, + "grad_norm": 0.15391060709953308, + "learning_rate": 3.687799523162953e-05, + "loss": 3.1218, + "step": 13232 + }, + { + "epoch": 0.11514985598556385, + "grad_norm": 0.16017532348632812, + "learning_rate": 3.6875626474720226e-05, + "loss": 3.1425, + "step": 13233 + }, + { + "epoch": 0.11518070245113728, + "grad_norm": 0.1470223218202591, + "learning_rate": 3.687325758012017e-05, + "loss": 3.1189, + "step": 13234 + }, + { + "epoch": 0.11521154891671069, + "grad_norm": 0.1500832885503769, + "learning_rate": 3.687088854785682e-05, + "loss": 3.1437, + "step": 13235 + }, + { + "epoch": 0.1152423953822841, + "grad_norm": 0.1626804918050766, + "learning_rate": 3.6868519377957654e-05, + "loss": 3.1315, + "step": 13236 + }, + { + "epoch": 0.11527324184785752, + "grad_norm": 0.15039636194705963, + "learning_rate": 3.6866150070450135e-05, + "loss": 3.1388, + "step": 13237 + }, + { + "epoch": 0.11530408831343093, + "grad_norm": 0.16528072953224182, + "learning_rate": 3.686378062536174e-05, + "loss": 3.1472, + "step": 13238 + }, + { + "epoch": 0.11533493477900435, + "grad_norm": 0.16525417566299438, + "learning_rate": 3.686141104271994e-05, + "loss": 3.1432, + "step": 13239 + }, + { + "epoch": 0.11536578124457778, + "grad_norm": 0.1708260029554367, + "learning_rate": 3.6859041322552206e-05, + "loss": 3.1358, + "step": 13240 + }, + { + "epoch": 0.11539662771015119, + "grad_norm": 0.14683088660240173, + "learning_rate": 3.6856671464886006e-05, + "loss": 3.1687, + "step": 13241 + }, + { + "epoch": 0.1154274741757246, + "grad_norm": 0.15537269413471222, + "learning_rate": 3.685430146974884e-05, + "loss": 3.1482, + "step": 13242 + }, + { + "epoch": 0.11545832064129802, + "grad_norm": 0.1604653298854828, + "learning_rate": 3.685193133716816e-05, + "loss": 3.1223, + "step": 13243 + }, + { + "epoch": 0.11548916710687143, + "grad_norm": 0.1533491462469101, + "learning_rate": 3.6849561067171455e-05, + "loss": 3.1304, + "step": 13244 + }, + { + "epoch": 0.11552001357244485, + "grad_norm": 0.15351852774620056, + "learning_rate": 3.6847190659786226e-05, + "loss": 3.1353, + "step": 13245 + }, + { + "epoch": 0.11555086003801827, + "grad_norm": 0.16270920634269714, + "learning_rate": 3.6844820115039934e-05, + "loss": 3.1109, + "step": 13246 + }, + { + "epoch": 0.11558170650359169, + "grad_norm": 0.15834112465381622, + "learning_rate": 3.684244943296006e-05, + "loss": 3.1638, + "step": 13247 + }, + { + "epoch": 0.1156125529691651, + "grad_norm": 0.15509900450706482, + "learning_rate": 3.684007861357411e-05, + "loss": 3.1505, + "step": 13248 + }, + { + "epoch": 0.11564339943473852, + "grad_norm": 0.15391463041305542, + "learning_rate": 3.683770765690957e-05, + "loss": 3.1417, + "step": 13249 + }, + { + "epoch": 0.11567424590031193, + "grad_norm": 0.15448912978172302, + "learning_rate": 3.6835336562993914e-05, + "loss": 3.1222, + "step": 13250 + }, + { + "epoch": 0.11570509236588535, + "grad_norm": 0.16496223211288452, + "learning_rate": 3.683296533185465e-05, + "loss": 3.1453, + "step": 13251 + }, + { + "epoch": 0.11573593883145877, + "grad_norm": 0.1584901213645935, + "learning_rate": 3.683059396351926e-05, + "loss": 3.1299, + "step": 13252 + }, + { + "epoch": 0.11576678529703219, + "grad_norm": 0.16865739226341248, + "learning_rate": 3.682822245801524e-05, + "loss": 3.1383, + "step": 13253 + }, + { + "epoch": 0.1157976317626056, + "grad_norm": 0.15867742896080017, + "learning_rate": 3.682585081537009e-05, + "loss": 3.1338, + "step": 13254 + }, + { + "epoch": 0.11582847822817902, + "grad_norm": 0.15120336413383484, + "learning_rate": 3.6823479035611305e-05, + "loss": 3.1477, + "step": 13255 + }, + { + "epoch": 0.11585932469375243, + "grad_norm": 0.15144705772399902, + "learning_rate": 3.6821107118766384e-05, + "loss": 3.1273, + "step": 13256 + }, + { + "epoch": 0.11589017115932584, + "grad_norm": 0.14837725460529327, + "learning_rate": 3.681873506486284e-05, + "loss": 3.1356, + "step": 13257 + }, + { + "epoch": 0.11592101762489927, + "grad_norm": 0.1477588564157486, + "learning_rate": 3.681636287392816e-05, + "loss": 3.1183, + "step": 13258 + }, + { + "epoch": 0.11595186409047269, + "grad_norm": 0.15742239356040955, + "learning_rate": 3.681399054598984e-05, + "loss": 3.16, + "step": 13259 + }, + { + "epoch": 0.1159827105560461, + "grad_norm": 0.15022705495357513, + "learning_rate": 3.681161808107542e-05, + "loss": 3.1403, + "step": 13260 + }, + { + "epoch": 0.11601355702161951, + "grad_norm": 0.15601930022239685, + "learning_rate": 3.680924547921237e-05, + "loss": 3.1164, + "step": 13261 + }, + { + "epoch": 0.11604440348719293, + "grad_norm": 0.15240757167339325, + "learning_rate": 3.6806872740428225e-05, + "loss": 3.1497, + "step": 13262 + }, + { + "epoch": 0.11607524995276634, + "grad_norm": 0.15885132551193237, + "learning_rate": 3.6804499864750485e-05, + "loss": 3.1403, + "step": 13263 + }, + { + "epoch": 0.11610609641833977, + "grad_norm": 0.15759791433811188, + "learning_rate": 3.680212685220666e-05, + "loss": 3.163, + "step": 13264 + }, + { + "epoch": 0.11613694288391319, + "grad_norm": 0.1591295450925827, + "learning_rate": 3.679975370282426e-05, + "loss": 3.1538, + "step": 13265 + }, + { + "epoch": 0.1161677893494866, + "grad_norm": 0.1640598624944687, + "learning_rate": 3.6797380416630814e-05, + "loss": 3.1358, + "step": 13266 + }, + { + "epoch": 0.11619863581506001, + "grad_norm": 0.15252330899238586, + "learning_rate": 3.679500699365383e-05, + "loss": 3.135, + "step": 13267 + }, + { + "epoch": 0.11622948228063343, + "grad_norm": 0.14646793901920319, + "learning_rate": 3.679263343392083e-05, + "loss": 3.1157, + "step": 13268 + }, + { + "epoch": 0.11626032874620684, + "grad_norm": 0.15455412864685059, + "learning_rate": 3.6790259737459336e-05, + "loss": 3.144, + "step": 13269 + }, + { + "epoch": 0.11629117521178027, + "grad_norm": 0.15247926115989685, + "learning_rate": 3.678788590429686e-05, + "loss": 3.1415, + "step": 13270 + }, + { + "epoch": 0.11632202167735368, + "grad_norm": 0.15098266303539276, + "learning_rate": 3.678551193446093e-05, + "loss": 3.0944, + "step": 13271 + }, + { + "epoch": 0.1163528681429271, + "grad_norm": 0.36959052085876465, + "learning_rate": 3.678313782797907e-05, + "loss": 3.1264, + "step": 13272 + }, + { + "epoch": 0.11638371460850051, + "grad_norm": 0.15641987323760986, + "learning_rate": 3.678076358487881e-05, + "loss": 3.1375, + "step": 13273 + }, + { + "epoch": 0.11641456107407393, + "grad_norm": 0.16255803406238556, + "learning_rate": 3.677838920518768e-05, + "loss": 3.1123, + "step": 13274 + }, + { + "epoch": 0.11644540753964734, + "grad_norm": 0.15556076169013977, + "learning_rate": 3.6776014688933204e-05, + "loss": 3.1594, + "step": 13275 + }, + { + "epoch": 0.11647625400522077, + "grad_norm": 0.16316325962543488, + "learning_rate": 3.6773640036142895e-05, + "loss": 3.1315, + "step": 13276 + }, + { + "epoch": 0.11650710047079418, + "grad_norm": 0.1528576761484146, + "learning_rate": 3.677126524684433e-05, + "loss": 3.1481, + "step": 13277 + }, + { + "epoch": 0.1165379469363676, + "grad_norm": 0.15486973524093628, + "learning_rate": 3.6768890321065005e-05, + "loss": 3.1391, + "step": 13278 + }, + { + "epoch": 0.11656879340194101, + "grad_norm": 0.15153436362743378, + "learning_rate": 3.6766515258832474e-05, + "loss": 3.1295, + "step": 13279 + }, + { + "epoch": 0.11659963986751443, + "grad_norm": 0.15896327793598175, + "learning_rate": 3.676414006017427e-05, + "loss": 3.127, + "step": 13280 + }, + { + "epoch": 0.11663048633308784, + "grad_norm": 0.15239004790782928, + "learning_rate": 3.676176472511793e-05, + "loss": 3.1502, + "step": 13281 + }, + { + "epoch": 0.11666133279866127, + "grad_norm": 0.15227000415325165, + "learning_rate": 3.675938925369099e-05, + "loss": 3.103, + "step": 13282 + }, + { + "epoch": 0.11669217926423468, + "grad_norm": 0.1647123545408249, + "learning_rate": 3.675701364592101e-05, + "loss": 3.166, + "step": 13283 + }, + { + "epoch": 0.1167230257298081, + "grad_norm": 0.15153443813323975, + "learning_rate": 3.675463790183552e-05, + "loss": 3.1416, + "step": 13284 + }, + { + "epoch": 0.11675387219538151, + "grad_norm": 0.15572749078273773, + "learning_rate": 3.675226202146206e-05, + "loss": 3.1276, + "step": 13285 + }, + { + "epoch": 0.11678471866095493, + "grad_norm": 0.16306191682815552, + "learning_rate": 3.674988600482819e-05, + "loss": 3.1503, + "step": 13286 + }, + { + "epoch": 0.11681556512652834, + "grad_norm": 0.16092899441719055, + "learning_rate": 3.674750985196145e-05, + "loss": 3.1528, + "step": 13287 + }, + { + "epoch": 0.11684641159210177, + "grad_norm": 0.16090045869350433, + "learning_rate": 3.674513356288939e-05, + "loss": 3.1478, + "step": 13288 + }, + { + "epoch": 0.11687725805767518, + "grad_norm": 0.14968864619731903, + "learning_rate": 3.6742757137639566e-05, + "loss": 3.1239, + "step": 13289 + }, + { + "epoch": 0.1169081045232486, + "grad_norm": 0.15326406061649323, + "learning_rate": 3.674038057623953e-05, + "loss": 3.0947, + "step": 13290 + }, + { + "epoch": 0.11693895098882201, + "grad_norm": 0.17971162497997284, + "learning_rate": 3.673800387871684e-05, + "loss": 3.1652, + "step": 13291 + }, + { + "epoch": 0.11696979745439542, + "grad_norm": 0.15854628384113312, + "learning_rate": 3.6735627045099044e-05, + "loss": 3.1434, + "step": 13292 + }, + { + "epoch": 0.11700064391996884, + "grad_norm": 0.14910495281219482, + "learning_rate": 3.673325007541371e-05, + "loss": 3.1323, + "step": 13293 + }, + { + "epoch": 0.11703149038554227, + "grad_norm": 0.15268437564373016, + "learning_rate": 3.673087296968838e-05, + "loss": 3.1337, + "step": 13294 + }, + { + "epoch": 0.11706233685111568, + "grad_norm": 0.15463300049304962, + "learning_rate": 3.672849572795064e-05, + "loss": 3.1391, + "step": 13295 + }, + { + "epoch": 0.1170931833166891, + "grad_norm": 0.15297235548496246, + "learning_rate": 3.672611835022804e-05, + "loss": 3.1296, + "step": 13296 + }, + { + "epoch": 0.11712402978226251, + "grad_norm": 0.15658049285411835, + "learning_rate": 3.6723740836548134e-05, + "loss": 3.1196, + "step": 13297 + }, + { + "epoch": 0.11715487624783592, + "grad_norm": 0.14926467835903168, + "learning_rate": 3.672136318693851e-05, + "loss": 3.1375, + "step": 13298 + }, + { + "epoch": 0.11718572271340934, + "grad_norm": 0.17449615895748138, + "learning_rate": 3.671898540142671e-05, + "loss": 3.1596, + "step": 13299 + }, + { + "epoch": 0.11721656917898277, + "grad_norm": 0.20287296175956726, + "learning_rate": 3.671660748004033e-05, + "loss": 3.1491, + "step": 13300 + }, + { + "epoch": 0.11724741564455618, + "grad_norm": 0.149725079536438, + "learning_rate": 3.6714229422806925e-05, + "loss": 3.1237, + "step": 13301 + }, + { + "epoch": 0.1172782621101296, + "grad_norm": 0.15153005719184875, + "learning_rate": 3.671185122975407e-05, + "loss": 3.1253, + "step": 13302 + }, + { + "epoch": 0.11730910857570301, + "grad_norm": 0.15853656828403473, + "learning_rate": 3.6709472900909326e-05, + "loss": 3.1137, + "step": 13303 + }, + { + "epoch": 0.11733995504127642, + "grad_norm": 0.15120871365070343, + "learning_rate": 3.670709443630029e-05, + "loss": 3.1464, + "step": 13304 + }, + { + "epoch": 0.11737080150684984, + "grad_norm": 0.16132892668247223, + "learning_rate": 3.670471583595453e-05, + "loss": 3.1421, + "step": 13305 + }, + { + "epoch": 0.11740164797242326, + "grad_norm": 0.15355169773101807, + "learning_rate": 3.670233709989961e-05, + "loss": 3.1277, + "step": 13306 + }, + { + "epoch": 0.11743249443799668, + "grad_norm": 0.15464600920677185, + "learning_rate": 3.669995822816315e-05, + "loss": 3.156, + "step": 13307 + }, + { + "epoch": 0.1174633409035701, + "grad_norm": 0.15305139124393463, + "learning_rate": 3.669757922077268e-05, + "loss": 3.1104, + "step": 13308 + }, + { + "epoch": 0.11749418736914351, + "grad_norm": 0.1506323665380478, + "learning_rate": 3.6695200077755816e-05, + "loss": 3.1463, + "step": 13309 + }, + { + "epoch": 0.11752503383471692, + "grad_norm": 0.15755920112133026, + "learning_rate": 3.6692820799140136e-05, + "loss": 3.1306, + "step": 13310 + }, + { + "epoch": 0.11755588030029034, + "grad_norm": 0.14507056772708893, + "learning_rate": 3.669044138495323e-05, + "loss": 3.1555, + "step": 13311 + }, + { + "epoch": 0.11758672676586376, + "grad_norm": 0.1508718580007553, + "learning_rate": 3.668806183522268e-05, + "loss": 3.1505, + "step": 13312 + }, + { + "epoch": 0.11761757323143718, + "grad_norm": 0.16185598075389862, + "learning_rate": 3.668568214997608e-05, + "loss": 3.1448, + "step": 13313 + }, + { + "epoch": 0.11764841969701059, + "grad_norm": 0.14626628160476685, + "learning_rate": 3.668330232924101e-05, + "loss": 3.1215, + "step": 13314 + }, + { + "epoch": 0.117679266162584, + "grad_norm": 0.16746729612350464, + "learning_rate": 3.668092237304507e-05, + "loss": 3.1232, + "step": 13315 + }, + { + "epoch": 0.11771011262815742, + "grad_norm": 0.1508706957101822, + "learning_rate": 3.6678542281415854e-05, + "loss": 3.1126, + "step": 13316 + }, + { + "epoch": 0.11774095909373083, + "grad_norm": 0.14684216678142548, + "learning_rate": 3.667616205438096e-05, + "loss": 3.1294, + "step": 13317 + }, + { + "epoch": 0.11777180555930426, + "grad_norm": 0.16076143085956573, + "learning_rate": 3.667378169196799e-05, + "loss": 3.1234, + "step": 13318 + }, + { + "epoch": 0.11780265202487768, + "grad_norm": 0.14667882025241852, + "learning_rate": 3.6671401194204535e-05, + "loss": 3.1236, + "step": 13319 + }, + { + "epoch": 0.11783349849045109, + "grad_norm": 0.15314604341983795, + "learning_rate": 3.666902056111819e-05, + "loss": 3.1668, + "step": 13320 + }, + { + "epoch": 0.1178643449560245, + "grad_norm": 0.1509569138288498, + "learning_rate": 3.666663979273656e-05, + "loss": 3.1388, + "step": 13321 + }, + { + "epoch": 0.11789519142159792, + "grad_norm": 0.14868691563606262, + "learning_rate": 3.6664258889087265e-05, + "loss": 3.12, + "step": 13322 + }, + { + "epoch": 0.11792603788717133, + "grad_norm": 0.14961464703083038, + "learning_rate": 3.666187785019789e-05, + "loss": 3.1436, + "step": 13323 + }, + { + "epoch": 0.11795688435274476, + "grad_norm": 0.16590607166290283, + "learning_rate": 3.665949667609605e-05, + "loss": 3.1528, + "step": 13324 + }, + { + "epoch": 0.11798773081831818, + "grad_norm": 0.14793561398983002, + "learning_rate": 3.6657115366809366e-05, + "loss": 3.1027, + "step": 13325 + }, + { + "epoch": 0.11801857728389159, + "grad_norm": 0.1457539051771164, + "learning_rate": 3.665473392236541e-05, + "loss": 3.1361, + "step": 13326 + }, + { + "epoch": 0.118049423749465, + "grad_norm": 0.15739396214485168, + "learning_rate": 3.665235234279183e-05, + "loss": 3.1103, + "step": 13327 + }, + { + "epoch": 0.11808027021503842, + "grad_norm": 0.15059731900691986, + "learning_rate": 3.664997062811624e-05, + "loss": 3.1189, + "step": 13328 + }, + { + "epoch": 0.11811111668061183, + "grad_norm": 0.1457037329673767, + "learning_rate": 3.664758877836622e-05, + "loss": 3.1198, + "step": 13329 + }, + { + "epoch": 0.11814196314618526, + "grad_norm": 0.1585550755262375, + "learning_rate": 3.664520679356942e-05, + "loss": 3.1615, + "step": 13330 + }, + { + "epoch": 0.11817280961175868, + "grad_norm": 0.1497560441493988, + "learning_rate": 3.664282467375344e-05, + "loss": 3.1423, + "step": 13331 + }, + { + "epoch": 0.11820365607733209, + "grad_norm": 0.14857196807861328, + "learning_rate": 3.66404424189459e-05, + "loss": 3.1536, + "step": 13332 + }, + { + "epoch": 0.1182345025429055, + "grad_norm": 0.15324576199054718, + "learning_rate": 3.663806002917444e-05, + "loss": 3.1195, + "step": 13333 + }, + { + "epoch": 0.11826534900847892, + "grad_norm": 0.1478334367275238, + "learning_rate": 3.6635677504466656e-05, + "loss": 3.1213, + "step": 13334 + }, + { + "epoch": 0.11829619547405233, + "grad_norm": 0.14774614572525024, + "learning_rate": 3.663329484485019e-05, + "loss": 3.1129, + "step": 13335 + }, + { + "epoch": 0.11832704193962576, + "grad_norm": 0.15708640217781067, + "learning_rate": 3.663091205035265e-05, + "loss": 3.1241, + "step": 13336 + }, + { + "epoch": 0.11835788840519917, + "grad_norm": 0.1477155089378357, + "learning_rate": 3.662852912100168e-05, + "loss": 3.1273, + "step": 13337 + }, + { + "epoch": 0.11838873487077259, + "grad_norm": 0.16638703644275665, + "learning_rate": 3.6626146056824903e-05, + "loss": 3.1553, + "step": 13338 + }, + { + "epoch": 0.118419581336346, + "grad_norm": 0.15404126048088074, + "learning_rate": 3.6623762857849954e-05, + "loss": 3.1152, + "step": 13339 + }, + { + "epoch": 0.11845042780191942, + "grad_norm": 0.1557295173406601, + "learning_rate": 3.662137952410446e-05, + "loss": 3.1282, + "step": 13340 + }, + { + "epoch": 0.11848127426749283, + "grad_norm": 0.2581477463245392, + "learning_rate": 3.6618996055616046e-05, + "loss": 3.1285, + "step": 13341 + }, + { + "epoch": 0.11851212073306626, + "grad_norm": 0.1641445904970169, + "learning_rate": 3.6616612452412356e-05, + "loss": 3.0989, + "step": 13342 + }, + { + "epoch": 0.11854296719863967, + "grad_norm": 0.19060057401657104, + "learning_rate": 3.661422871452103e-05, + "loss": 3.1493, + "step": 13343 + }, + { + "epoch": 0.11857381366421309, + "grad_norm": 0.15571585297584534, + "learning_rate": 3.66118448419697e-05, + "loss": 3.1283, + "step": 13344 + }, + { + "epoch": 0.1186046601297865, + "grad_norm": 0.15640586614608765, + "learning_rate": 3.660946083478601e-05, + "loss": 3.1303, + "step": 13345 + }, + { + "epoch": 0.11863550659535992, + "grad_norm": 0.1538161039352417, + "learning_rate": 3.660707669299759e-05, + "loss": 3.1432, + "step": 13346 + }, + { + "epoch": 0.11866635306093333, + "grad_norm": 0.15552227199077606, + "learning_rate": 3.66046924166321e-05, + "loss": 3.1332, + "step": 13347 + }, + { + "epoch": 0.11869719952650676, + "grad_norm": 0.1768275648355484, + "learning_rate": 3.660230800571717e-05, + "loss": 3.1384, + "step": 13348 + }, + { + "epoch": 0.11872804599208017, + "grad_norm": 0.16705961525440216, + "learning_rate": 3.659992346028045e-05, + "loss": 3.1174, + "step": 13349 + }, + { + "epoch": 0.11875889245765359, + "grad_norm": 0.1565156877040863, + "learning_rate": 3.659753878034959e-05, + "loss": 3.0858, + "step": 13350 + }, + { + "epoch": 0.118789738923227, + "grad_norm": 0.15138979256153107, + "learning_rate": 3.6595153965952236e-05, + "loss": 3.1509, + "step": 13351 + }, + { + "epoch": 0.11882058538880041, + "grad_norm": 0.15460312366485596, + "learning_rate": 3.6592769017116044e-05, + "loss": 3.159, + "step": 13352 + }, + { + "epoch": 0.11885143185437384, + "grad_norm": 0.1624716967344284, + "learning_rate": 3.659038393386865e-05, + "loss": 3.123, + "step": 13353 + }, + { + "epoch": 0.11888227831994726, + "grad_norm": 0.16617323458194733, + "learning_rate": 3.658799871623773e-05, + "loss": 3.1236, + "step": 13354 + }, + { + "epoch": 0.11891312478552067, + "grad_norm": 0.15035270154476166, + "learning_rate": 3.6585613364250927e-05, + "loss": 3.1588, + "step": 13355 + }, + { + "epoch": 0.11894397125109409, + "grad_norm": 0.15220999717712402, + "learning_rate": 3.65832278779359e-05, + "loss": 3.1554, + "step": 13356 + }, + { + "epoch": 0.1189748177166675, + "grad_norm": 0.15027913451194763, + "learning_rate": 3.6580842257320304e-05, + "loss": 3.1424, + "step": 13357 + }, + { + "epoch": 0.11900566418224091, + "grad_norm": 0.15459081530570984, + "learning_rate": 3.6578456502431804e-05, + "loss": 3.1465, + "step": 13358 + }, + { + "epoch": 0.11903651064781434, + "grad_norm": 0.1488700807094574, + "learning_rate": 3.657607061329805e-05, + "loss": 3.1137, + "step": 13359 + }, + { + "epoch": 0.11906735711338776, + "grad_norm": 0.15944117307662964, + "learning_rate": 3.657368458994672e-05, + "loss": 3.138, + "step": 13360 + }, + { + "epoch": 0.11909820357896117, + "grad_norm": 0.1521780639886856, + "learning_rate": 3.6571298432405476e-05, + "loss": 3.1403, + "step": 13361 + }, + { + "epoch": 0.11912905004453458, + "grad_norm": 0.14995841681957245, + "learning_rate": 3.656891214070198e-05, + "loss": 3.1314, + "step": 13362 + }, + { + "epoch": 0.119159896510108, + "grad_norm": 0.15211161971092224, + "learning_rate": 3.6566525714863896e-05, + "loss": 3.188, + "step": 13363 + }, + { + "epoch": 0.11919074297568141, + "grad_norm": 0.15092463791370392, + "learning_rate": 3.6564139154918895e-05, + "loss": 3.1744, + "step": 13364 + }, + { + "epoch": 0.11922158944125484, + "grad_norm": 0.15954507887363434, + "learning_rate": 3.656175246089465e-05, + "loss": 3.1317, + "step": 13365 + }, + { + "epoch": 0.11925243590682826, + "grad_norm": 0.14715664088726044, + "learning_rate": 3.6559365632818835e-05, + "loss": 3.0727, + "step": 13366 + }, + { + "epoch": 0.11928328237240167, + "grad_norm": 0.151030033826828, + "learning_rate": 3.6556978670719125e-05, + "loss": 3.1705, + "step": 13367 + }, + { + "epoch": 0.11931412883797508, + "grad_norm": 0.1466217041015625, + "learning_rate": 3.655459157462319e-05, + "loss": 3.1345, + "step": 13368 + }, + { + "epoch": 0.1193449753035485, + "grad_norm": 0.14927946031093597, + "learning_rate": 3.655220434455871e-05, + "loss": 3.1281, + "step": 13369 + }, + { + "epoch": 0.11937582176912191, + "grad_norm": 0.143205925822258, + "learning_rate": 3.654981698055336e-05, + "loss": 3.1037, + "step": 13370 + }, + { + "epoch": 0.11940666823469534, + "grad_norm": 0.15388928353786469, + "learning_rate": 3.654742948263482e-05, + "loss": 3.1389, + "step": 13371 + }, + { + "epoch": 0.11943751470026875, + "grad_norm": 0.1461993008852005, + "learning_rate": 3.654504185083078e-05, + "loss": 3.1081, + "step": 13372 + }, + { + "epoch": 0.11946836116584217, + "grad_norm": 0.17580316960811615, + "learning_rate": 3.654265408516892e-05, + "loss": 3.1627, + "step": 13373 + }, + { + "epoch": 0.11949920763141558, + "grad_norm": 0.15028546750545502, + "learning_rate": 3.6540266185676913e-05, + "loss": 3.1716, + "step": 13374 + }, + { + "epoch": 0.119530054096989, + "grad_norm": 0.14751189947128296, + "learning_rate": 3.653787815238247e-05, + "loss": 3.1191, + "step": 13375 + }, + { + "epoch": 0.11956090056256241, + "grad_norm": 0.15017102658748627, + "learning_rate": 3.653548998531324e-05, + "loss": 3.1274, + "step": 13376 + }, + { + "epoch": 0.11959174702813584, + "grad_norm": 0.14939020574092865, + "learning_rate": 3.6533101684496954e-05, + "loss": 3.1296, + "step": 13377 + }, + { + "epoch": 0.11962259349370925, + "grad_norm": 0.15018422901630402, + "learning_rate": 3.653071324996128e-05, + "loss": 3.1305, + "step": 13378 + }, + { + "epoch": 0.11965343995928267, + "grad_norm": 0.14610789716243744, + "learning_rate": 3.6528324681733905e-05, + "loss": 3.1269, + "step": 13379 + }, + { + "epoch": 0.11968428642485608, + "grad_norm": 0.16475513577461243, + "learning_rate": 3.652593597984255e-05, + "loss": 3.1806, + "step": 13380 + }, + { + "epoch": 0.1197151328904295, + "grad_norm": 0.16121207177639008, + "learning_rate": 3.652354714431488e-05, + "loss": 3.1225, + "step": 13381 + }, + { + "epoch": 0.11974597935600291, + "grad_norm": 0.15502969920635223, + "learning_rate": 3.652115817517861e-05, + "loss": 3.1437, + "step": 13382 + }, + { + "epoch": 0.11977682582157634, + "grad_norm": 0.16771593689918518, + "learning_rate": 3.651876907246144e-05, + "loss": 3.1239, + "step": 13383 + }, + { + "epoch": 0.11980767228714975, + "grad_norm": 0.153351828455925, + "learning_rate": 3.651637983619106e-05, + "loss": 3.1205, + "step": 13384 + }, + { + "epoch": 0.11983851875272317, + "grad_norm": 0.15955336391925812, + "learning_rate": 3.6513990466395174e-05, + "loss": 3.1341, + "step": 13385 + }, + { + "epoch": 0.11986936521829658, + "grad_norm": 0.1573714315891266, + "learning_rate": 3.65116009631015e-05, + "loss": 3.1399, + "step": 13386 + }, + { + "epoch": 0.11990021168387, + "grad_norm": 0.161021888256073, + "learning_rate": 3.650921132633771e-05, + "loss": 3.1338, + "step": 13387 + }, + { + "epoch": 0.11993105814944341, + "grad_norm": 0.15957216918468475, + "learning_rate": 3.650682155613154e-05, + "loss": 3.1353, + "step": 13388 + }, + { + "epoch": 0.11996190461501684, + "grad_norm": 0.1559787392616272, + "learning_rate": 3.650443165251069e-05, + "loss": 3.1458, + "step": 13389 + }, + { + "epoch": 0.11999275108059025, + "grad_norm": 0.15908733010292053, + "learning_rate": 3.650204161550288e-05, + "loss": 3.0881, + "step": 13390 + }, + { + "epoch": 0.12002359754616367, + "grad_norm": 0.16553151607513428, + "learning_rate": 3.649965144513579e-05, + "loss": 3.1395, + "step": 13391 + }, + { + "epoch": 0.12005444401173708, + "grad_norm": 0.15114688873291016, + "learning_rate": 3.649726114143716e-05, + "loss": 3.1465, + "step": 13392 + }, + { + "epoch": 0.1200852904773105, + "grad_norm": 0.16591915488243103, + "learning_rate": 3.64948707044347e-05, + "loss": 3.1167, + "step": 13393 + }, + { + "epoch": 0.12011613694288391, + "grad_norm": 0.15341390669345856, + "learning_rate": 3.6492480134156116e-05, + "loss": 3.1364, + "step": 13394 + }, + { + "epoch": 0.12014698340845734, + "grad_norm": 0.16661730408668518, + "learning_rate": 3.649008943062913e-05, + "loss": 3.1308, + "step": 13395 + }, + { + "epoch": 0.12017782987403075, + "grad_norm": 0.1598353087902069, + "learning_rate": 3.648769859388147e-05, + "loss": 3.1676, + "step": 13396 + }, + { + "epoch": 0.12020867633960416, + "grad_norm": 0.16176925599575043, + "learning_rate": 3.648530762394084e-05, + "loss": 3.1159, + "step": 13397 + }, + { + "epoch": 0.12023952280517758, + "grad_norm": 0.16644752025604248, + "learning_rate": 3.6482916520834975e-05, + "loss": 3.0896, + "step": 13398 + }, + { + "epoch": 0.12027036927075099, + "grad_norm": 0.1486928015947342, + "learning_rate": 3.648052528459159e-05, + "loss": 3.1429, + "step": 13399 + }, + { + "epoch": 0.12030121573632441, + "grad_norm": 0.17257803678512573, + "learning_rate": 3.647813391523841e-05, + "loss": 3.1623, + "step": 13400 + }, + { + "epoch": 0.12033206220189784, + "grad_norm": 0.1645934283733368, + "learning_rate": 3.6475742412803176e-05, + "loss": 3.1582, + "step": 13401 + }, + { + "epoch": 0.12036290866747125, + "grad_norm": 0.16093552112579346, + "learning_rate": 3.6473350777313596e-05, + "loss": 3.17, + "step": 13402 + }, + { + "epoch": 0.12039375513304466, + "grad_norm": 0.16495263576507568, + "learning_rate": 3.647095900879741e-05, + "loss": 3.1371, + "step": 13403 + }, + { + "epoch": 0.12042460159861808, + "grad_norm": 0.15923835337162018, + "learning_rate": 3.646856710728235e-05, + "loss": 3.1506, + "step": 13404 + }, + { + "epoch": 0.12045544806419149, + "grad_norm": 0.1714833527803421, + "learning_rate": 3.646617507279615e-05, + "loss": 3.1935, + "step": 13405 + }, + { + "epoch": 0.1204862945297649, + "grad_norm": 0.1520104557275772, + "learning_rate": 3.6463782905366526e-05, + "loss": 3.1422, + "step": 13406 + }, + { + "epoch": 0.12051714099533833, + "grad_norm": 0.16361048817634583, + "learning_rate": 3.646139060502124e-05, + "loss": 3.1588, + "step": 13407 + }, + { + "epoch": 0.12054798746091175, + "grad_norm": 0.15652133524417877, + "learning_rate": 3.645899817178802e-05, + "loss": 3.1462, + "step": 13408 + }, + { + "epoch": 0.12057883392648516, + "grad_norm": 0.15811370313167572, + "learning_rate": 3.645660560569459e-05, + "loss": 3.164, + "step": 13409 + }, + { + "epoch": 0.12060968039205858, + "grad_norm": 0.1628277599811554, + "learning_rate": 3.6454212906768706e-05, + "loss": 3.1291, + "step": 13410 + }, + { + "epoch": 0.12064052685763199, + "grad_norm": 0.16058112680912018, + "learning_rate": 3.645182007503811e-05, + "loss": 3.1295, + "step": 13411 + }, + { + "epoch": 0.1206713733232054, + "grad_norm": 0.1588064432144165, + "learning_rate": 3.644942711053054e-05, + "loss": 3.149, + "step": 13412 + }, + { + "epoch": 0.12070221978877883, + "grad_norm": 0.1528451144695282, + "learning_rate": 3.644703401327375e-05, + "loss": 3.1456, + "step": 13413 + }, + { + "epoch": 0.12073306625435225, + "grad_norm": 0.17153578996658325, + "learning_rate": 3.6444640783295476e-05, + "loss": 3.1399, + "step": 13414 + }, + { + "epoch": 0.12076391271992566, + "grad_norm": 0.15631628036499023, + "learning_rate": 3.6442247420623467e-05, + "loss": 3.1697, + "step": 13415 + }, + { + "epoch": 0.12079475918549908, + "grad_norm": 0.15314404666423798, + "learning_rate": 3.643985392528548e-05, + "loss": 3.1216, + "step": 13416 + }, + { + "epoch": 0.12082560565107249, + "grad_norm": 0.16622354090213776, + "learning_rate": 3.643746029730925e-05, + "loss": 3.1519, + "step": 13417 + }, + { + "epoch": 0.1208564521166459, + "grad_norm": 0.1510835886001587, + "learning_rate": 3.643506653672255e-05, + "loss": 3.1524, + "step": 13418 + }, + { + "epoch": 0.12088729858221933, + "grad_norm": 0.15835627913475037, + "learning_rate": 3.643267264355313e-05, + "loss": 3.1399, + "step": 13419 + }, + { + "epoch": 0.12091814504779275, + "grad_norm": 0.15712812542915344, + "learning_rate": 3.643027861782873e-05, + "loss": 3.1574, + "step": 13420 + }, + { + "epoch": 0.12094899151336616, + "grad_norm": 0.1510452926158905, + "learning_rate": 3.642788445957712e-05, + "loss": 3.1216, + "step": 13421 + }, + { + "epoch": 0.12097983797893958, + "grad_norm": 0.15265297889709473, + "learning_rate": 3.642549016882606e-05, + "loss": 3.1466, + "step": 13422 + }, + { + "epoch": 0.12101068444451299, + "grad_norm": 0.16062112152576447, + "learning_rate": 3.6423095745603306e-05, + "loss": 3.1263, + "step": 13423 + }, + { + "epoch": 0.1210415309100864, + "grad_norm": 0.1540553718805313, + "learning_rate": 3.6420701189936626e-05, + "loss": 3.154, + "step": 13424 + }, + { + "epoch": 0.12107237737565983, + "grad_norm": 0.1500631719827652, + "learning_rate": 3.641830650185377e-05, + "loss": 3.1029, + "step": 13425 + }, + { + "epoch": 0.12110322384123325, + "grad_norm": 0.14811071753501892, + "learning_rate": 3.641591168138252e-05, + "loss": 3.1211, + "step": 13426 + }, + { + "epoch": 0.12113407030680666, + "grad_norm": 0.15590178966522217, + "learning_rate": 3.641351672855063e-05, + "loss": 3.1612, + "step": 13427 + }, + { + "epoch": 0.12116491677238007, + "grad_norm": 0.15045984089374542, + "learning_rate": 3.6411121643385876e-05, + "loss": 3.1219, + "step": 13428 + }, + { + "epoch": 0.12119576323795349, + "grad_norm": 0.14639140665531158, + "learning_rate": 3.640872642591602e-05, + "loss": 3.1322, + "step": 13429 + }, + { + "epoch": 0.1212266097035269, + "grad_norm": 0.14987026154994965, + "learning_rate": 3.6406331076168844e-05, + "loss": 3.1085, + "step": 13430 + }, + { + "epoch": 0.12125745616910033, + "grad_norm": 0.15742546319961548, + "learning_rate": 3.64039355941721e-05, + "loss": 3.1414, + "step": 13431 + }, + { + "epoch": 0.12128830263467374, + "grad_norm": 0.16418279707431793, + "learning_rate": 3.640153997995358e-05, + "loss": 3.1425, + "step": 13432 + }, + { + "epoch": 0.12131914910024716, + "grad_norm": 0.1440618634223938, + "learning_rate": 3.639914423354106e-05, + "loss": 3.1485, + "step": 13433 + }, + { + "epoch": 0.12134999556582057, + "grad_norm": 0.1525173783302307, + "learning_rate": 3.639674835496232e-05, + "loss": 3.0822, + "step": 13434 + }, + { + "epoch": 0.12138084203139399, + "grad_norm": 0.164683997631073, + "learning_rate": 3.639435234424512e-05, + "loss": 3.0985, + "step": 13435 + }, + { + "epoch": 0.1214116884969674, + "grad_norm": 0.15015827119350433, + "learning_rate": 3.639195620141725e-05, + "loss": 3.1116, + "step": 13436 + }, + { + "epoch": 0.12144253496254083, + "grad_norm": 0.15511631965637207, + "learning_rate": 3.63895599265065e-05, + "loss": 3.1331, + "step": 13437 + }, + { + "epoch": 0.12147338142811424, + "grad_norm": 0.14661860466003418, + "learning_rate": 3.638716351954064e-05, + "loss": 3.1621, + "step": 13438 + }, + { + "epoch": 0.12150422789368766, + "grad_norm": 0.1454824060201645, + "learning_rate": 3.6384766980547476e-05, + "loss": 3.119, + "step": 13439 + }, + { + "epoch": 0.12153507435926107, + "grad_norm": 0.1454760879278183, + "learning_rate": 3.638237030955477e-05, + "loss": 3.1001, + "step": 13440 + }, + { + "epoch": 0.12156592082483449, + "grad_norm": 0.14459270238876343, + "learning_rate": 3.637997350659033e-05, + "loss": 3.1258, + "step": 13441 + }, + { + "epoch": 0.1215967672904079, + "grad_norm": 0.14988191425800323, + "learning_rate": 3.637757657168191e-05, + "loss": 3.1387, + "step": 13442 + }, + { + "epoch": 0.12162761375598133, + "grad_norm": 0.14385296404361725, + "learning_rate": 3.637517950485735e-05, + "loss": 3.1247, + "step": 13443 + }, + { + "epoch": 0.12165846022155474, + "grad_norm": 0.14682455360889435, + "learning_rate": 3.6372782306144415e-05, + "loss": 3.1521, + "step": 13444 + }, + { + "epoch": 0.12168930668712816, + "grad_norm": 0.14838427305221558, + "learning_rate": 3.6370384975570905e-05, + "loss": 3.1931, + "step": 13445 + }, + { + "epoch": 0.12172015315270157, + "grad_norm": 0.20021742582321167, + "learning_rate": 3.636798751316462e-05, + "loss": 3.1508, + "step": 13446 + }, + { + "epoch": 0.12175099961827499, + "grad_norm": 0.1497557908296585, + "learning_rate": 3.6365589918953335e-05, + "loss": 3.1751, + "step": 13447 + }, + { + "epoch": 0.1217818460838484, + "grad_norm": 0.15284831821918488, + "learning_rate": 3.636319219296487e-05, + "loss": 3.1455, + "step": 13448 + }, + { + "epoch": 0.12181269254942183, + "grad_norm": 0.1523789018392563, + "learning_rate": 3.636079433522702e-05, + "loss": 3.1529, + "step": 13449 + }, + { + "epoch": 0.12184353901499524, + "grad_norm": 0.15809087455272675, + "learning_rate": 3.6358396345767594e-05, + "loss": 3.1049, + "step": 13450 + }, + { + "epoch": 0.12187438548056866, + "grad_norm": 0.17567318677902222, + "learning_rate": 3.6355998224614384e-05, + "loss": 3.1352, + "step": 13451 + }, + { + "epoch": 0.12190523194614207, + "grad_norm": 0.15594299137592316, + "learning_rate": 3.63535999717952e-05, + "loss": 3.1737, + "step": 13452 + }, + { + "epoch": 0.12193607841171548, + "grad_norm": 0.15383009612560272, + "learning_rate": 3.635120158733784e-05, + "loss": 3.1875, + "step": 13453 + }, + { + "epoch": 0.1219669248772889, + "grad_norm": 0.15296250581741333, + "learning_rate": 3.634880307127012e-05, + "loss": 3.1344, + "step": 13454 + }, + { + "epoch": 0.12199777134286233, + "grad_norm": 0.1579822599887848, + "learning_rate": 3.634640442361986e-05, + "loss": 3.108, + "step": 13455 + }, + { + "epoch": 0.12202861780843574, + "grad_norm": 0.15046927332878113, + "learning_rate": 3.634400564441485e-05, + "loss": 3.1382, + "step": 13456 + }, + { + "epoch": 0.12205946427400916, + "grad_norm": 0.2805477976799011, + "learning_rate": 3.634160673368291e-05, + "loss": 3.1321, + "step": 13457 + }, + { + "epoch": 0.12209031073958257, + "grad_norm": 0.1574847251176834, + "learning_rate": 3.633920769145186e-05, + "loss": 3.1344, + "step": 13458 + }, + { + "epoch": 0.12212115720515598, + "grad_norm": 0.16174612939357758, + "learning_rate": 3.63368085177495e-05, + "loss": 3.1525, + "step": 13459 + }, + { + "epoch": 0.1221520036707294, + "grad_norm": 0.16464264690876007, + "learning_rate": 3.633440921260367e-05, + "loss": 3.1287, + "step": 13460 + }, + { + "epoch": 0.12218285013630283, + "grad_norm": 0.15947787463665009, + "learning_rate": 3.633200977604217e-05, + "loss": 3.1528, + "step": 13461 + }, + { + "epoch": 0.12221369660187624, + "grad_norm": 0.16318179666996002, + "learning_rate": 3.632961020809283e-05, + "loss": 3.153, + "step": 13462 + }, + { + "epoch": 0.12224454306744965, + "grad_norm": 0.1565307229757309, + "learning_rate": 3.632721050878347e-05, + "loss": 3.126, + "step": 13463 + }, + { + "epoch": 0.12227538953302307, + "grad_norm": 0.15383125841617584, + "learning_rate": 3.6324810678141905e-05, + "loss": 3.1029, + "step": 13464 + }, + { + "epoch": 0.12230623599859648, + "grad_norm": 0.15830424427986145, + "learning_rate": 3.632241071619597e-05, + "loss": 3.1441, + "step": 13465 + }, + { + "epoch": 0.1223370824641699, + "grad_norm": 0.1549808830022812, + "learning_rate": 3.632001062297349e-05, + "loss": 3.1362, + "step": 13466 + }, + { + "epoch": 0.12236792892974332, + "grad_norm": 0.14940513670444489, + "learning_rate": 3.6317610398502276e-05, + "loss": 3.1114, + "step": 13467 + }, + { + "epoch": 0.12239877539531674, + "grad_norm": 0.16325414180755615, + "learning_rate": 3.6315210042810185e-05, + "loss": 3.1146, + "step": 13468 + }, + { + "epoch": 0.12242962186089015, + "grad_norm": 0.18168291449546814, + "learning_rate": 3.631280955592503e-05, + "loss": 3.1098, + "step": 13469 + }, + { + "epoch": 0.12246046832646357, + "grad_norm": 0.17632469534873962, + "learning_rate": 3.631040893787464e-05, + "loss": 3.1555, + "step": 13470 + }, + { + "epoch": 0.12249131479203698, + "grad_norm": 0.15514463186264038, + "learning_rate": 3.630800818868686e-05, + "loss": 3.1355, + "step": 13471 + }, + { + "epoch": 0.1225221612576104, + "grad_norm": 0.15850712358951569, + "learning_rate": 3.630560730838952e-05, + "loss": 3.1602, + "step": 13472 + }, + { + "epoch": 0.12255300772318382, + "grad_norm": 0.1479278802871704, + "learning_rate": 3.6303206297010454e-05, + "loss": 3.1424, + "step": 13473 + }, + { + "epoch": 0.12258385418875724, + "grad_norm": 0.1585034281015396, + "learning_rate": 3.630080515457751e-05, + "loss": 3.1121, + "step": 13474 + }, + { + "epoch": 0.12261470065433065, + "grad_norm": 0.15740451216697693, + "learning_rate": 3.6298403881118515e-05, + "loss": 3.1428, + "step": 13475 + }, + { + "epoch": 0.12264554711990407, + "grad_norm": 0.15189261734485626, + "learning_rate": 3.629600247666132e-05, + "loss": 3.1176, + "step": 13476 + }, + { + "epoch": 0.12267639358547748, + "grad_norm": 0.15069498121738434, + "learning_rate": 3.6293600941233766e-05, + "loss": 3.1276, + "step": 13477 + }, + { + "epoch": 0.1227072400510509, + "grad_norm": 0.15540559589862823, + "learning_rate": 3.6291199274863695e-05, + "loss": 3.1472, + "step": 13478 + }, + { + "epoch": 0.12273808651662432, + "grad_norm": 0.15331634879112244, + "learning_rate": 3.628879747757895e-05, + "loss": 3.1686, + "step": 13479 + }, + { + "epoch": 0.12276893298219774, + "grad_norm": 0.15903142094612122, + "learning_rate": 3.628639554940738e-05, + "loss": 3.1027, + "step": 13480 + }, + { + "epoch": 0.12279977944777115, + "grad_norm": 0.15724502503871918, + "learning_rate": 3.6283993490376844e-05, + "loss": 3.1313, + "step": 13481 + }, + { + "epoch": 0.12283062591334457, + "grad_norm": 0.15630491077899933, + "learning_rate": 3.628159130051517e-05, + "loss": 3.1836, + "step": 13482 + }, + { + "epoch": 0.12286147237891798, + "grad_norm": 0.16120442748069763, + "learning_rate": 3.6279188979850245e-05, + "loss": 3.1387, + "step": 13483 + }, + { + "epoch": 0.1228923188444914, + "grad_norm": 0.14484642446041107, + "learning_rate": 3.627678652840989e-05, + "loss": 3.1156, + "step": 13484 + }, + { + "epoch": 0.12292316531006482, + "grad_norm": 0.16781774163246155, + "learning_rate": 3.627438394622197e-05, + "loss": 3.147, + "step": 13485 + }, + { + "epoch": 0.12295401177563824, + "grad_norm": 0.15476000308990479, + "learning_rate": 3.6271981233314354e-05, + "loss": 3.1389, + "step": 13486 + }, + { + "epoch": 0.12298485824121165, + "grad_norm": 0.15091215074062347, + "learning_rate": 3.6269578389714874e-05, + "loss": 3.1368, + "step": 13487 + }, + { + "epoch": 0.12301570470678506, + "grad_norm": 0.15116867423057556, + "learning_rate": 3.626717541545141e-05, + "loss": 3.1242, + "step": 13488 + }, + { + "epoch": 0.12304655117235848, + "grad_norm": 0.15303939580917358, + "learning_rate": 3.626477231055182e-05, + "loss": 3.117, + "step": 13489 + }, + { + "epoch": 0.12307739763793189, + "grad_norm": 0.15608079731464386, + "learning_rate": 3.626236907504397e-05, + "loss": 3.1309, + "step": 13490 + }, + { + "epoch": 0.12310824410350532, + "grad_norm": 0.17300185561180115, + "learning_rate": 3.6259965708955714e-05, + "loss": 3.157, + "step": 13491 + }, + { + "epoch": 0.12313909056907874, + "grad_norm": 0.15559640526771545, + "learning_rate": 3.625756221231491e-05, + "loss": 3.1245, + "step": 13492 + }, + { + "epoch": 0.12316993703465215, + "grad_norm": 0.14905664324760437, + "learning_rate": 3.625515858514946e-05, + "loss": 3.1698, + "step": 13493 + }, + { + "epoch": 0.12320078350022556, + "grad_norm": 0.16632568836212158, + "learning_rate": 3.625275482748719e-05, + "loss": 3.1187, + "step": 13494 + }, + { + "epoch": 0.12323162996579898, + "grad_norm": 0.15090802311897278, + "learning_rate": 3.6250350939356006e-05, + "loss": 3.13, + "step": 13495 + }, + { + "epoch": 0.12326247643137239, + "grad_norm": 0.1642681509256363, + "learning_rate": 3.6247946920783755e-05, + "loss": 3.1602, + "step": 13496 + }, + { + "epoch": 0.12329332289694582, + "grad_norm": 0.1497718244791031, + "learning_rate": 3.6245542771798314e-05, + "loss": 3.1092, + "step": 13497 + }, + { + "epoch": 0.12332416936251923, + "grad_norm": 0.149012953042984, + "learning_rate": 3.624313849242757e-05, + "loss": 3.1182, + "step": 13498 + }, + { + "epoch": 0.12335501582809265, + "grad_norm": 0.14952386915683746, + "learning_rate": 3.624073408269939e-05, + "loss": 3.1328, + "step": 13499 + }, + { + "epoch": 0.12338586229366606, + "grad_norm": 0.15578943490982056, + "learning_rate": 3.623832954264165e-05, + "loss": 3.1309, + "step": 13500 + }, + { + "epoch": 0.12341670875923948, + "grad_norm": 0.1606902927160263, + "learning_rate": 3.6235924872282236e-05, + "loss": 3.1385, + "step": 13501 + }, + { + "epoch": 0.12344755522481289, + "grad_norm": 0.15229180455207825, + "learning_rate": 3.6233520071649026e-05, + "loss": 3.1376, + "step": 13502 + }, + { + "epoch": 0.12347840169038632, + "grad_norm": 0.14824777841567993, + "learning_rate": 3.62311151407699e-05, + "loss": 3.1442, + "step": 13503 + }, + { + "epoch": 0.12350924815595973, + "grad_norm": 0.15643766522407532, + "learning_rate": 3.6228710079672734e-05, + "loss": 3.1293, + "step": 13504 + }, + { + "epoch": 0.12354009462153315, + "grad_norm": 0.15311270952224731, + "learning_rate": 3.6226304888385435e-05, + "loss": 3.1511, + "step": 13505 + }, + { + "epoch": 0.12357094108710656, + "grad_norm": 0.18600037693977356, + "learning_rate": 3.622389956693587e-05, + "loss": 3.1425, + "step": 13506 + }, + { + "epoch": 0.12360178755267998, + "grad_norm": 0.15165065228939056, + "learning_rate": 3.622149411535194e-05, + "loss": 3.1262, + "step": 13507 + }, + { + "epoch": 0.12363263401825339, + "grad_norm": 0.15785817801952362, + "learning_rate": 3.621908853366152e-05, + "loss": 3.1279, + "step": 13508 + }, + { + "epoch": 0.12366348048382682, + "grad_norm": 0.1595030277967453, + "learning_rate": 3.6216682821892514e-05, + "loss": 3.1822, + "step": 13509 + }, + { + "epoch": 0.12369432694940023, + "grad_norm": 0.15410763025283813, + "learning_rate": 3.621427698007281e-05, + "loss": 3.1736, + "step": 13510 + }, + { + "epoch": 0.12372517341497365, + "grad_norm": 0.1558404266834259, + "learning_rate": 3.621187100823031e-05, + "loss": 3.1207, + "step": 13511 + }, + { + "epoch": 0.12375601988054706, + "grad_norm": 0.15279018878936768, + "learning_rate": 3.62094649063929e-05, + "loss": 3.145, + "step": 13512 + }, + { + "epoch": 0.12378686634612047, + "grad_norm": 0.16133591532707214, + "learning_rate": 3.620705867458849e-05, + "loss": 3.145, + "step": 13513 + }, + { + "epoch": 0.12381771281169389, + "grad_norm": 0.14925873279571533, + "learning_rate": 3.6204652312844956e-05, + "loss": 3.1747, + "step": 13514 + }, + { + "epoch": 0.12384855927726732, + "grad_norm": 0.16250677406787872, + "learning_rate": 3.620224582119021e-05, + "loss": 3.1508, + "step": 13515 + }, + { + "epoch": 0.12387940574284073, + "grad_norm": 0.15354213118553162, + "learning_rate": 3.619983919965216e-05, + "loss": 3.1301, + "step": 13516 + }, + { + "epoch": 0.12391025220841415, + "grad_norm": 0.15433970093727112, + "learning_rate": 3.619743244825871e-05, + "loss": 3.1136, + "step": 13517 + }, + { + "epoch": 0.12394109867398756, + "grad_norm": 0.15996478497982025, + "learning_rate": 3.619502556703775e-05, + "loss": 3.1675, + "step": 13518 + }, + { + "epoch": 0.12397194513956097, + "grad_norm": 0.15790019929409027, + "learning_rate": 3.619261855601721e-05, + "loss": 3.1334, + "step": 13519 + }, + { + "epoch": 0.12400279160513439, + "grad_norm": 0.15528811514377594, + "learning_rate": 3.619021141522497e-05, + "loss": 3.1251, + "step": 13520 + }, + { + "epoch": 0.12403363807070782, + "grad_norm": 0.15353892743587494, + "learning_rate": 3.618780414468896e-05, + "loss": 3.1516, + "step": 13521 + }, + { + "epoch": 0.12406448453628123, + "grad_norm": 0.15388956665992737, + "learning_rate": 3.618539674443708e-05, + "loss": 3.1223, + "step": 13522 + }, + { + "epoch": 0.12409533100185464, + "grad_norm": 0.15498502552509308, + "learning_rate": 3.618298921449725e-05, + "loss": 3.152, + "step": 13523 + }, + { + "epoch": 0.12412617746742806, + "grad_norm": 0.15072152018547058, + "learning_rate": 3.618058155489738e-05, + "loss": 3.1111, + "step": 13524 + }, + { + "epoch": 0.12415702393300147, + "grad_norm": 0.1524551957845688, + "learning_rate": 3.617817376566538e-05, + "loss": 3.1709, + "step": 13525 + }, + { + "epoch": 0.12418787039857489, + "grad_norm": 0.1483800709247589, + "learning_rate": 3.617576584682918e-05, + "loss": 3.1418, + "step": 13526 + }, + { + "epoch": 0.12421871686414832, + "grad_norm": 0.15148456394672394, + "learning_rate": 3.617335779841668e-05, + "loss": 3.1308, + "step": 13527 + }, + { + "epoch": 0.12424956332972173, + "grad_norm": 0.1541120558977127, + "learning_rate": 3.617094962045583e-05, + "loss": 3.1779, + "step": 13528 + }, + { + "epoch": 0.12428040979529514, + "grad_norm": 0.16197851300239563, + "learning_rate": 3.616854131297452e-05, + "loss": 3.0954, + "step": 13529 + }, + { + "epoch": 0.12431125626086856, + "grad_norm": 0.14837311208248138, + "learning_rate": 3.616613287600069e-05, + "loss": 3.1547, + "step": 13530 + }, + { + "epoch": 0.12434210272644197, + "grad_norm": 0.14672936499118805, + "learning_rate": 3.6163724309562244e-05, + "loss": 3.1333, + "step": 13531 + }, + { + "epoch": 0.12437294919201539, + "grad_norm": 0.1474795788526535, + "learning_rate": 3.616131561368714e-05, + "loss": 3.1096, + "step": 13532 + }, + { + "epoch": 0.12440379565758881, + "grad_norm": 0.15061748027801514, + "learning_rate": 3.615890678840328e-05, + "loss": 3.1656, + "step": 13533 + }, + { + "epoch": 0.12443464212316223, + "grad_norm": 0.15208019316196442, + "learning_rate": 3.615649783373861e-05, + "loss": 3.1148, + "step": 13534 + }, + { + "epoch": 0.12446548858873564, + "grad_norm": 0.15255144238471985, + "learning_rate": 3.6154088749721046e-05, + "loss": 3.1544, + "step": 13535 + }, + { + "epoch": 0.12449633505430906, + "grad_norm": 0.15118631720542908, + "learning_rate": 3.615167953637852e-05, + "loss": 3.1446, + "step": 13536 + }, + { + "epoch": 0.12452718151988247, + "grad_norm": 0.15786588191986084, + "learning_rate": 3.614927019373898e-05, + "loss": 3.1226, + "step": 13537 + }, + { + "epoch": 0.12455802798545589, + "grad_norm": 0.1571701169013977, + "learning_rate": 3.6146860721830346e-05, + "loss": 3.1433, + "step": 13538 + }, + { + "epoch": 0.12458887445102931, + "grad_norm": 0.14590729773044586, + "learning_rate": 3.6144451120680564e-05, + "loss": 3.102, + "step": 13539 + }, + { + "epoch": 0.12461972091660273, + "grad_norm": 0.15546530485153198, + "learning_rate": 3.614204139031757e-05, + "loss": 3.1257, + "step": 13540 + }, + { + "epoch": 0.12465056738217614, + "grad_norm": 0.1488177478313446, + "learning_rate": 3.6139631530769303e-05, + "loss": 3.1177, + "step": 13541 + }, + { + "epoch": 0.12468141384774956, + "grad_norm": 0.1496141254901886, + "learning_rate": 3.6137221542063696e-05, + "loss": 3.1379, + "step": 13542 + }, + { + "epoch": 0.12471226031332297, + "grad_norm": 0.1529044508934021, + "learning_rate": 3.61348114242287e-05, + "loss": 3.1276, + "step": 13543 + }, + { + "epoch": 0.12474310677889638, + "grad_norm": 0.1790861040353775, + "learning_rate": 3.613240117729226e-05, + "loss": 3.1316, + "step": 13544 + }, + { + "epoch": 0.12477395324446981, + "grad_norm": 0.15678457915782928, + "learning_rate": 3.612999080128231e-05, + "loss": 3.1218, + "step": 13545 + }, + { + "epoch": 0.12480479971004323, + "grad_norm": 0.15279683470726013, + "learning_rate": 3.6127580296226824e-05, + "loss": 3.1283, + "step": 13546 + }, + { + "epoch": 0.12483564617561664, + "grad_norm": 0.1518900990486145, + "learning_rate": 3.6125169662153715e-05, + "loss": 3.1498, + "step": 13547 + }, + { + "epoch": 0.12486649264119005, + "grad_norm": 0.18482986092567444, + "learning_rate": 3.6122758899090946e-05, + "loss": 3.1609, + "step": 13548 + }, + { + "epoch": 0.12489733910676347, + "grad_norm": 0.15327247977256775, + "learning_rate": 3.612034800706648e-05, + "loss": 3.1235, + "step": 13549 + }, + { + "epoch": 0.12492818557233688, + "grad_norm": 0.14757809042930603, + "learning_rate": 3.6117936986108256e-05, + "loss": 3.114, + "step": 13550 + }, + { + "epoch": 0.12495903203791031, + "grad_norm": 0.15504735708236694, + "learning_rate": 3.6115525836244246e-05, + "loss": 3.1606, + "step": 13551 + }, + { + "epoch": 0.12498987850348373, + "grad_norm": 0.14884310960769653, + "learning_rate": 3.6113114557502385e-05, + "loss": 3.1435, + "step": 13552 + }, + { + "epoch": 0.12502072496905714, + "grad_norm": 0.15588028728961945, + "learning_rate": 3.611070314991064e-05, + "loss": 3.139, + "step": 13553 + }, + { + "epoch": 0.12505157143463055, + "grad_norm": 0.14820602536201477, + "learning_rate": 3.610829161349696e-05, + "loss": 3.1515, + "step": 13554 + }, + { + "epoch": 0.12508241790020397, + "grad_norm": 0.15508608520030975, + "learning_rate": 3.610587994828933e-05, + "loss": 3.1592, + "step": 13555 + }, + { + "epoch": 0.12511326436577738, + "grad_norm": 0.15382125973701477, + "learning_rate": 3.610346815431568e-05, + "loss": 3.1855, + "step": 13556 + }, + { + "epoch": 0.1251441108313508, + "grad_norm": 0.15647603571414948, + "learning_rate": 3.6101056231604e-05, + "loss": 3.1575, + "step": 13557 + }, + { + "epoch": 0.1251749572969242, + "grad_norm": 0.15354856848716736, + "learning_rate": 3.609864418018224e-05, + "loss": 3.1223, + "step": 13558 + }, + { + "epoch": 0.12520580376249762, + "grad_norm": 0.14973676204681396, + "learning_rate": 3.609623200007837e-05, + "loss": 3.1274, + "step": 13559 + }, + { + "epoch": 0.12523665022807107, + "grad_norm": 0.1514693945646286, + "learning_rate": 3.6093819691320355e-05, + "loss": 3.1036, + "step": 13560 + }, + { + "epoch": 0.12526749669364448, + "grad_norm": 0.15432442724704742, + "learning_rate": 3.609140725393617e-05, + "loss": 3.1057, + "step": 13561 + }, + { + "epoch": 0.1252983431592179, + "grad_norm": 0.15596376359462738, + "learning_rate": 3.608899468795379e-05, + "loss": 3.1184, + "step": 13562 + }, + { + "epoch": 0.1253291896247913, + "grad_norm": 0.15672333538532257, + "learning_rate": 3.608658199340118e-05, + "loss": 3.1239, + "step": 13563 + }, + { + "epoch": 0.12536003609036472, + "grad_norm": 0.15411663055419922, + "learning_rate": 3.6084169170306306e-05, + "loss": 3.1614, + "step": 13564 + }, + { + "epoch": 0.12539088255593814, + "grad_norm": 0.1508455127477646, + "learning_rate": 3.608175621869715e-05, + "loss": 3.0968, + "step": 13565 + }, + { + "epoch": 0.12542172902151155, + "grad_norm": 0.1552790105342865, + "learning_rate": 3.60793431386017e-05, + "loss": 3.1487, + "step": 13566 + }, + { + "epoch": 0.12545257548708497, + "grad_norm": 0.16340358555316925, + "learning_rate": 3.607692993004791e-05, + "loss": 3.1398, + "step": 13567 + }, + { + "epoch": 0.12548342195265838, + "grad_norm": 0.15127605199813843, + "learning_rate": 3.6074516593063784e-05, + "loss": 3.1488, + "step": 13568 + }, + { + "epoch": 0.1255142684182318, + "grad_norm": 0.15246520936489105, + "learning_rate": 3.607210312767729e-05, + "loss": 3.1321, + "step": 13569 + }, + { + "epoch": 0.1255451148838052, + "grad_norm": 0.15187810361385345, + "learning_rate": 3.606968953391642e-05, + "loss": 3.1193, + "step": 13570 + }, + { + "epoch": 0.12557596134937862, + "grad_norm": 0.14919573068618774, + "learning_rate": 3.606727581180914e-05, + "loss": 3.1564, + "step": 13571 + }, + { + "epoch": 0.12560680781495207, + "grad_norm": 0.1497623324394226, + "learning_rate": 3.606486196138345e-05, + "loss": 3.1311, + "step": 13572 + }, + { + "epoch": 0.12563765428052548, + "grad_norm": 0.1434279829263687, + "learning_rate": 3.606244798266735e-05, + "loss": 3.1116, + "step": 13573 + }, + { + "epoch": 0.1256685007460989, + "grad_norm": 0.15724065899848938, + "learning_rate": 3.6060033875688804e-05, + "loss": 3.1398, + "step": 13574 + }, + { + "epoch": 0.1256993472116723, + "grad_norm": 0.15426911413669586, + "learning_rate": 3.60576196404758e-05, + "loss": 3.1494, + "step": 13575 + }, + { + "epoch": 0.12573019367724572, + "grad_norm": 0.14787760376930237, + "learning_rate": 3.605520527705635e-05, + "loss": 3.1074, + "step": 13576 + }, + { + "epoch": 0.12576104014281914, + "grad_norm": 0.14922833442687988, + "learning_rate": 3.6052790785458444e-05, + "loss": 3.0873, + "step": 13577 + }, + { + "epoch": 0.12579188660839255, + "grad_norm": 0.1713053584098816, + "learning_rate": 3.605037616571007e-05, + "loss": 3.1417, + "step": 13578 + }, + { + "epoch": 0.12582273307396596, + "grad_norm": 0.15109224617481232, + "learning_rate": 3.604796141783922e-05, + "loss": 3.1387, + "step": 13579 + }, + { + "epoch": 0.12585357953953938, + "grad_norm": 0.15360520780086517, + "learning_rate": 3.60455465418739e-05, + "loss": 3.1313, + "step": 13580 + }, + { + "epoch": 0.1258844260051128, + "grad_norm": 0.1500856727361679, + "learning_rate": 3.60431315378421e-05, + "loss": 3.1112, + "step": 13581 + }, + { + "epoch": 0.1259152724706862, + "grad_norm": 0.15343303978443146, + "learning_rate": 3.6040716405771824e-05, + "loss": 3.1669, + "step": 13582 + }, + { + "epoch": 0.12594611893625962, + "grad_norm": 0.1622234433889389, + "learning_rate": 3.603830114569109e-05, + "loss": 3.1458, + "step": 13583 + }, + { + "epoch": 0.12597696540183306, + "grad_norm": 0.14796003699302673, + "learning_rate": 3.603588575762788e-05, + "loss": 3.1229, + "step": 13584 + }, + { + "epoch": 0.12600781186740648, + "grad_norm": 0.15389248728752136, + "learning_rate": 3.60334702416102e-05, + "loss": 3.1469, + "step": 13585 + }, + { + "epoch": 0.1260386583329799, + "grad_norm": 0.15249022841453552, + "learning_rate": 3.603105459766607e-05, + "loss": 3.1419, + "step": 13586 + }, + { + "epoch": 0.1260695047985533, + "grad_norm": 0.14742538332939148, + "learning_rate": 3.6028638825823486e-05, + "loss": 3.1034, + "step": 13587 + }, + { + "epoch": 0.12610035126412672, + "grad_norm": 0.1488286405801773, + "learning_rate": 3.6026222926110464e-05, + "loss": 3.0896, + "step": 13588 + }, + { + "epoch": 0.12613119772970013, + "grad_norm": 0.14663191139698029, + "learning_rate": 3.6023806898555016e-05, + "loss": 3.1331, + "step": 13589 + }, + { + "epoch": 0.12616204419527355, + "grad_norm": 0.14989544451236725, + "learning_rate": 3.602139074318515e-05, + "loss": 3.1272, + "step": 13590 + }, + { + "epoch": 0.12619289066084696, + "grad_norm": 0.15143965184688568, + "learning_rate": 3.601897446002888e-05, + "loss": 3.1119, + "step": 13591 + }, + { + "epoch": 0.12622373712642038, + "grad_norm": 0.14985817670822144, + "learning_rate": 3.6016558049114225e-05, + "loss": 3.1244, + "step": 13592 + }, + { + "epoch": 0.1262545835919938, + "grad_norm": 0.156367689371109, + "learning_rate": 3.60141415104692e-05, + "loss": 3.1065, + "step": 13593 + }, + { + "epoch": 0.1262854300575672, + "grad_norm": 0.1483597457408905, + "learning_rate": 3.601172484412182e-05, + "loss": 3.1611, + "step": 13594 + }, + { + "epoch": 0.12631627652314062, + "grad_norm": 0.150534525513649, + "learning_rate": 3.600930805010011e-05, + "loss": 3.1029, + "step": 13595 + }, + { + "epoch": 0.12634712298871406, + "grad_norm": 0.1520792841911316, + "learning_rate": 3.6006891128432093e-05, + "loss": 3.1604, + "step": 13596 + }, + { + "epoch": 0.12637796945428748, + "grad_norm": 0.15102159976959229, + "learning_rate": 3.6004474079145776e-05, + "loss": 3.1106, + "step": 13597 + }, + { + "epoch": 0.1264088159198609, + "grad_norm": 0.7679414749145508, + "learning_rate": 3.60020569022692e-05, + "loss": 3.1538, + "step": 13598 + }, + { + "epoch": 0.1264396623854343, + "grad_norm": 0.15147793292999268, + "learning_rate": 3.599963959783039e-05, + "loss": 3.1325, + "step": 13599 + }, + { + "epoch": 0.12647050885100772, + "grad_norm": 0.15615756809711456, + "learning_rate": 3.599722216585736e-05, + "loss": 3.1578, + "step": 13600 + }, + { + "epoch": 0.12650135531658113, + "grad_norm": 0.14826208353042603, + "learning_rate": 3.599480460637816e-05, + "loss": 3.1259, + "step": 13601 + }, + { + "epoch": 0.12653220178215455, + "grad_norm": 0.16043221950531006, + "learning_rate": 3.599238691942081e-05, + "loss": 3.1395, + "step": 13602 + }, + { + "epoch": 0.12656304824772796, + "grad_norm": 0.2829633951187134, + "learning_rate": 3.598996910501332e-05, + "loss": 3.1374, + "step": 13603 + }, + { + "epoch": 0.12659389471330137, + "grad_norm": 0.1577787846326828, + "learning_rate": 3.598755116318375e-05, + "loss": 3.1709, + "step": 13604 + }, + { + "epoch": 0.1266247411788748, + "grad_norm": 0.1710679680109024, + "learning_rate": 3.598513309396014e-05, + "loss": 3.1399, + "step": 13605 + }, + { + "epoch": 0.1266555876444482, + "grad_norm": 0.4648360311985016, + "learning_rate": 3.59827148973705e-05, + "loss": 3.1529, + "step": 13606 + }, + { + "epoch": 0.12668643411002162, + "grad_norm": 0.17034687101840973, + "learning_rate": 3.5980296573442885e-05, + "loss": 3.1441, + "step": 13607 + }, + { + "epoch": 0.12671728057559506, + "grad_norm": 0.2158721387386322, + "learning_rate": 3.597787812220532e-05, + "loss": 3.0959, + "step": 13608 + }, + { + "epoch": 0.12674812704116847, + "grad_norm": 0.31970080733299255, + "learning_rate": 3.597545954368586e-05, + "loss": 3.0948, + "step": 13609 + }, + { + "epoch": 0.1267789735067419, + "grad_norm": 0.16222423315048218, + "learning_rate": 3.5973040837912556e-05, + "loss": 3.1533, + "step": 13610 + }, + { + "epoch": 0.1268098199723153, + "grad_norm": 0.19539974629878998, + "learning_rate": 3.5970622004913426e-05, + "loss": 3.1275, + "step": 13611 + }, + { + "epoch": 0.12684066643788872, + "grad_norm": 0.19122113287448883, + "learning_rate": 3.596820304471653e-05, + "loss": 3.1399, + "step": 13612 + }, + { + "epoch": 0.12687151290346213, + "grad_norm": 0.176155686378479, + "learning_rate": 3.596578395734991e-05, + "loss": 3.1503, + "step": 13613 + }, + { + "epoch": 0.12690235936903554, + "grad_norm": 0.160533145070076, + "learning_rate": 3.596336474284161e-05, + "loss": 3.1484, + "step": 13614 + }, + { + "epoch": 0.12693320583460896, + "grad_norm": 0.21467535197734833, + "learning_rate": 3.5960945401219686e-05, + "loss": 3.121, + "step": 13615 + }, + { + "epoch": 0.12696405230018237, + "grad_norm": 0.16054654121398926, + "learning_rate": 3.5958525932512185e-05, + "loss": 3.119, + "step": 13616 + }, + { + "epoch": 0.1269948987657558, + "grad_norm": 0.16598989069461823, + "learning_rate": 3.5956106336747167e-05, + "loss": 3.1368, + "step": 13617 + }, + { + "epoch": 0.1270257452313292, + "grad_norm": 0.16108468174934387, + "learning_rate": 3.595368661395268e-05, + "loss": 3.1519, + "step": 13618 + }, + { + "epoch": 0.12705659169690262, + "grad_norm": 0.16090033948421478, + "learning_rate": 3.595126676415678e-05, + "loss": 3.1356, + "step": 13619 + }, + { + "epoch": 0.12708743816247606, + "grad_norm": 0.16522160172462463, + "learning_rate": 3.594884678738751e-05, + "loss": 3.124, + "step": 13620 + }, + { + "epoch": 0.12711828462804947, + "grad_norm": 0.16369624435901642, + "learning_rate": 3.594642668367295e-05, + "loss": 3.1374, + "step": 13621 + }, + { + "epoch": 0.12714913109362289, + "grad_norm": 0.1620732694864273, + "learning_rate": 3.594400645304116e-05, + "loss": 3.1312, + "step": 13622 + }, + { + "epoch": 0.1271799775591963, + "grad_norm": 0.16236884891986847, + "learning_rate": 3.594158609552017e-05, + "loss": 3.0648, + "step": 13623 + }, + { + "epoch": 0.12721082402476971, + "grad_norm": 0.15817001461982727, + "learning_rate": 3.593916561113808e-05, + "loss": 3.1457, + "step": 13624 + }, + { + "epoch": 0.12724167049034313, + "grad_norm": 0.1529187262058258, + "learning_rate": 3.593674499992294e-05, + "loss": 3.1409, + "step": 13625 + }, + { + "epoch": 0.12727251695591654, + "grad_norm": 0.15738998353481293, + "learning_rate": 3.593432426190281e-05, + "loss": 3.1043, + "step": 13626 + }, + { + "epoch": 0.12730336342148996, + "grad_norm": 0.15650293231010437, + "learning_rate": 3.593190339710576e-05, + "loss": 3.1862, + "step": 13627 + }, + { + "epoch": 0.12733420988706337, + "grad_norm": 0.15469413995742798, + "learning_rate": 3.5929482405559856e-05, + "loss": 3.0951, + "step": 13628 + }, + { + "epoch": 0.12736505635263679, + "grad_norm": 0.1525096595287323, + "learning_rate": 3.5927061287293176e-05, + "loss": 3.1141, + "step": 13629 + }, + { + "epoch": 0.1273959028182102, + "grad_norm": 0.15656083822250366, + "learning_rate": 3.592464004233379e-05, + "loss": 3.1219, + "step": 13630 + }, + { + "epoch": 0.12742674928378364, + "grad_norm": 0.15321554243564606, + "learning_rate": 3.592221867070976e-05, + "loss": 3.1637, + "step": 13631 + }, + { + "epoch": 0.12745759574935706, + "grad_norm": 0.15256401896476746, + "learning_rate": 3.5919797172449165e-05, + "loss": 3.142, + "step": 13632 + }, + { + "epoch": 0.12748844221493047, + "grad_norm": 0.14747552573680878, + "learning_rate": 3.5917375547580096e-05, + "loss": 3.103, + "step": 13633 + }, + { + "epoch": 0.12751928868050388, + "grad_norm": 0.15534399449825287, + "learning_rate": 3.591495379613061e-05, + "loss": 3.1545, + "step": 13634 + }, + { + "epoch": 0.1275501351460773, + "grad_norm": 0.153951957821846, + "learning_rate": 3.59125319181288e-05, + "loss": 3.1935, + "step": 13635 + }, + { + "epoch": 0.1275809816116507, + "grad_norm": 0.1581033319234848, + "learning_rate": 3.5910109913602735e-05, + "loss": 3.1095, + "step": 13636 + }, + { + "epoch": 0.12761182807722413, + "grad_norm": 0.1642233282327652, + "learning_rate": 3.59076877825805e-05, + "loss": 3.1227, + "step": 13637 + }, + { + "epoch": 0.12764267454279754, + "grad_norm": 0.14397059381008148, + "learning_rate": 3.590526552509018e-05, + "loss": 3.1357, + "step": 13638 + }, + { + "epoch": 0.12767352100837095, + "grad_norm": 0.15764744579792023, + "learning_rate": 3.590284314115987e-05, + "loss": 3.1263, + "step": 13639 + }, + { + "epoch": 0.12770436747394437, + "grad_norm": 0.15576499700546265, + "learning_rate": 3.590042063081764e-05, + "loss": 3.0915, + "step": 13640 + }, + { + "epoch": 0.12773521393951778, + "grad_norm": 0.1658191829919815, + "learning_rate": 3.5897997994091575e-05, + "loss": 3.1338, + "step": 13641 + }, + { + "epoch": 0.1277660604050912, + "grad_norm": 0.15371370315551758, + "learning_rate": 3.589557523100978e-05, + "loss": 3.1371, + "step": 13642 + }, + { + "epoch": 0.12779690687066464, + "grad_norm": 0.15242747962474823, + "learning_rate": 3.589315234160034e-05, + "loss": 3.1245, + "step": 13643 + }, + { + "epoch": 0.12782775333623805, + "grad_norm": 0.15593592822551727, + "learning_rate": 3.589072932589134e-05, + "loss": 3.1328, + "step": 13644 + }, + { + "epoch": 0.12785859980181147, + "grad_norm": 0.15136733651161194, + "learning_rate": 3.588830618391087e-05, + "loss": 3.1358, + "step": 13645 + }, + { + "epoch": 0.12788944626738488, + "grad_norm": 0.15440967679023743, + "learning_rate": 3.5885882915687054e-05, + "loss": 3.1628, + "step": 13646 + }, + { + "epoch": 0.1279202927329583, + "grad_norm": 0.14979521930217743, + "learning_rate": 3.5883459521247946e-05, + "loss": 3.1253, + "step": 13647 + }, + { + "epoch": 0.1279511391985317, + "grad_norm": 0.14583423733711243, + "learning_rate": 3.588103600062167e-05, + "loss": 3.1143, + "step": 13648 + }, + { + "epoch": 0.12798198566410512, + "grad_norm": 0.14952352643013, + "learning_rate": 3.587861235383633e-05, + "loss": 3.1075, + "step": 13649 + }, + { + "epoch": 0.12801283212967854, + "grad_norm": 0.14962933957576752, + "learning_rate": 3.587618858092001e-05, + "loss": 3.1268, + "step": 13650 + }, + { + "epoch": 0.12804367859525195, + "grad_norm": 0.164512500166893, + "learning_rate": 3.587376468190082e-05, + "loss": 3.1547, + "step": 13651 + }, + { + "epoch": 0.12807452506082537, + "grad_norm": 0.14880980551242828, + "learning_rate": 3.587134065680687e-05, + "loss": 3.1682, + "step": 13652 + }, + { + "epoch": 0.12810537152639878, + "grad_norm": 0.15145769715309143, + "learning_rate": 3.586891650566625e-05, + "loss": 3.1078, + "step": 13653 + }, + { + "epoch": 0.1281362179919722, + "grad_norm": 0.14561480283737183, + "learning_rate": 3.586649222850707e-05, + "loss": 3.1269, + "step": 13654 + }, + { + "epoch": 0.12816706445754564, + "grad_norm": 0.15017272531986237, + "learning_rate": 3.586406782535745e-05, + "loss": 3.1706, + "step": 13655 + }, + { + "epoch": 0.12819791092311905, + "grad_norm": 0.15527048707008362, + "learning_rate": 3.586164329624549e-05, + "loss": 3.1594, + "step": 13656 + }, + { + "epoch": 0.12822875738869247, + "grad_norm": 0.15192916989326477, + "learning_rate": 3.585921864119931e-05, + "loss": 3.1333, + "step": 13657 + }, + { + "epoch": 0.12825960385426588, + "grad_norm": 0.15404585003852844, + "learning_rate": 3.5856793860247006e-05, + "loss": 3.1165, + "step": 13658 + }, + { + "epoch": 0.1282904503198393, + "grad_norm": 0.15513119101524353, + "learning_rate": 3.58543689534167e-05, + "loss": 3.1694, + "step": 13659 + }, + { + "epoch": 0.1283212967854127, + "grad_norm": 0.1497264802455902, + "learning_rate": 3.585194392073652e-05, + "loss": 3.1359, + "step": 13660 + }, + { + "epoch": 0.12835214325098612, + "grad_norm": 0.15821272134780884, + "learning_rate": 3.584951876223456e-05, + "loss": 3.1146, + "step": 13661 + }, + { + "epoch": 0.12838298971655954, + "grad_norm": 0.1587495654821396, + "learning_rate": 3.5847093477938956e-05, + "loss": 3.1239, + "step": 13662 + }, + { + "epoch": 0.12841383618213295, + "grad_norm": 0.16625644266605377, + "learning_rate": 3.5844668067877824e-05, + "loss": 3.1504, + "step": 13663 + }, + { + "epoch": 0.12844468264770637, + "grad_norm": 0.15849916636943817, + "learning_rate": 3.584224253207927e-05, + "loss": 3.1502, + "step": 13664 + }, + { + "epoch": 0.12847552911327978, + "grad_norm": 0.1585327535867691, + "learning_rate": 3.583981687057144e-05, + "loss": 3.111, + "step": 13665 + }, + { + "epoch": 0.1285063755788532, + "grad_norm": 0.17136922478675842, + "learning_rate": 3.583739108338245e-05, + "loss": 3.0854, + "step": 13666 + }, + { + "epoch": 0.12853722204442664, + "grad_norm": 0.15844905376434326, + "learning_rate": 3.583496517054041e-05, + "loss": 3.1202, + "step": 13667 + }, + { + "epoch": 0.12856806851000005, + "grad_norm": 0.16439232230186462, + "learning_rate": 3.583253913207347e-05, + "loss": 3.1168, + "step": 13668 + }, + { + "epoch": 0.12859891497557346, + "grad_norm": 0.1651669293642044, + "learning_rate": 3.583011296800976e-05, + "loss": 3.1531, + "step": 13669 + }, + { + "epoch": 0.12862976144114688, + "grad_norm": 0.1657915562391281, + "learning_rate": 3.5827686678377375e-05, + "loss": 3.1535, + "step": 13670 + }, + { + "epoch": 0.1286606079067203, + "grad_norm": 0.16880656778812408, + "learning_rate": 3.582526026320448e-05, + "loss": 3.1309, + "step": 13671 + }, + { + "epoch": 0.1286914543722937, + "grad_norm": 0.1653713583946228, + "learning_rate": 3.58228337225192e-05, + "loss": 3.1396, + "step": 13672 + }, + { + "epoch": 0.12872230083786712, + "grad_norm": 0.15655644237995148, + "learning_rate": 3.582040705634966e-05, + "loss": 3.1533, + "step": 13673 + }, + { + "epoch": 0.12875314730344053, + "grad_norm": 0.1573583483695984, + "learning_rate": 3.581798026472401e-05, + "loss": 3.1045, + "step": 13674 + }, + { + "epoch": 0.12878399376901395, + "grad_norm": 0.15631836652755737, + "learning_rate": 3.581555334767037e-05, + "loss": 3.14, + "step": 13675 + }, + { + "epoch": 0.12881484023458736, + "grad_norm": 0.15269765257835388, + "learning_rate": 3.581312630521689e-05, + "loss": 3.1372, + "step": 13676 + }, + { + "epoch": 0.12884568670016078, + "grad_norm": 0.1534358412027359, + "learning_rate": 3.5810699137391715e-05, + "loss": 3.1314, + "step": 13677 + }, + { + "epoch": 0.1288765331657342, + "grad_norm": 0.1474297046661377, + "learning_rate": 3.580827184422299e-05, + "loss": 3.1177, + "step": 13678 + }, + { + "epoch": 0.12890737963130763, + "grad_norm": 0.14994072914123535, + "learning_rate": 3.5805844425738826e-05, + "loss": 3.1338, + "step": 13679 + }, + { + "epoch": 0.12893822609688105, + "grad_norm": 0.14902102947235107, + "learning_rate": 3.58034168819674e-05, + "loss": 3.1281, + "step": 13680 + }, + { + "epoch": 0.12896907256245446, + "grad_norm": 0.1551084667444229, + "learning_rate": 3.580098921293685e-05, + "loss": 3.1074, + "step": 13681 + }, + { + "epoch": 0.12899991902802788, + "grad_norm": 0.15559394657611847, + "learning_rate": 3.579856141867531e-05, + "loss": 3.1494, + "step": 13682 + }, + { + "epoch": 0.1290307654936013, + "grad_norm": 0.14643514156341553, + "learning_rate": 3.579613349921095e-05, + "loss": 3.1226, + "step": 13683 + }, + { + "epoch": 0.1290616119591747, + "grad_norm": 0.15395782887935638, + "learning_rate": 3.579370545457191e-05, + "loss": 3.1271, + "step": 13684 + }, + { + "epoch": 0.12909245842474812, + "grad_norm": 0.15910428762435913, + "learning_rate": 3.5791277284786335e-05, + "loss": 3.1342, + "step": 13685 + }, + { + "epoch": 0.12912330489032153, + "grad_norm": 0.14818412065505981, + "learning_rate": 3.578884898988239e-05, + "loss": 3.134, + "step": 13686 + }, + { + "epoch": 0.12915415135589495, + "grad_norm": 0.15093408524990082, + "learning_rate": 3.578642056988822e-05, + "loss": 3.1192, + "step": 13687 + }, + { + "epoch": 0.12918499782146836, + "grad_norm": 0.1489500105381012, + "learning_rate": 3.5783992024831984e-05, + "loss": 3.1485, + "step": 13688 + }, + { + "epoch": 0.12921584428704178, + "grad_norm": 0.14537203311920166, + "learning_rate": 3.5781563354741845e-05, + "loss": 3.1401, + "step": 13689 + }, + { + "epoch": 0.1292466907526152, + "grad_norm": 0.15227435529232025, + "learning_rate": 3.577913455964596e-05, + "loss": 3.1249, + "step": 13690 + }, + { + "epoch": 0.12927753721818863, + "grad_norm": 0.15788790583610535, + "learning_rate": 3.577670563957248e-05, + "loss": 3.1481, + "step": 13691 + }, + { + "epoch": 0.12930838368376205, + "grad_norm": 0.15346309542655945, + "learning_rate": 3.577427659454958e-05, + "loss": 3.1069, + "step": 13692 + }, + { + "epoch": 0.12933923014933546, + "grad_norm": 0.16912202537059784, + "learning_rate": 3.577184742460542e-05, + "loss": 3.1269, + "step": 13693 + }, + { + "epoch": 0.12937007661490887, + "grad_norm": 0.15400901436805725, + "learning_rate": 3.576941812976815e-05, + "loss": 3.1575, + "step": 13694 + }, + { + "epoch": 0.1294009230804823, + "grad_norm": 0.1522047519683838, + "learning_rate": 3.576698871006596e-05, + "loss": 3.1441, + "step": 13695 + }, + { + "epoch": 0.1294317695460557, + "grad_norm": 0.15890242159366608, + "learning_rate": 3.5764559165527e-05, + "loss": 3.1305, + "step": 13696 + }, + { + "epoch": 0.12946261601162912, + "grad_norm": 0.149008646607399, + "learning_rate": 3.576212949617945e-05, + "loss": 3.1335, + "step": 13697 + }, + { + "epoch": 0.12949346247720253, + "grad_norm": 0.14524893462657928, + "learning_rate": 3.5759699702051474e-05, + "loss": 3.1016, + "step": 13698 + }, + { + "epoch": 0.12952430894277595, + "grad_norm": 0.1546190083026886, + "learning_rate": 3.575726978317125e-05, + "loss": 3.1125, + "step": 13699 + }, + { + "epoch": 0.12955515540834936, + "grad_norm": 0.14895157516002655, + "learning_rate": 3.5754839739566935e-05, + "loss": 3.1219, + "step": 13700 + }, + { + "epoch": 0.12958600187392277, + "grad_norm": 0.14976443350315094, + "learning_rate": 3.575240957126673e-05, + "loss": 3.1606, + "step": 13701 + }, + { + "epoch": 0.1296168483394962, + "grad_norm": 0.14850565791130066, + "learning_rate": 3.57499792782988e-05, + "loss": 3.1219, + "step": 13702 + }, + { + "epoch": 0.12964769480506963, + "grad_norm": 0.14813010394573212, + "learning_rate": 3.574754886069131e-05, + "loss": 3.1339, + "step": 13703 + }, + { + "epoch": 0.12967854127064304, + "grad_norm": 0.1485918015241623, + "learning_rate": 3.574511831847244e-05, + "loss": 3.1169, + "step": 13704 + }, + { + "epoch": 0.12970938773621646, + "grad_norm": 0.1594172716140747, + "learning_rate": 3.57426876516704e-05, + "loss": 3.1205, + "step": 13705 + }, + { + "epoch": 0.12974023420178987, + "grad_norm": 0.15212541818618774, + "learning_rate": 3.574025686031334e-05, + "loss": 3.1522, + "step": 13706 + }, + { + "epoch": 0.1297710806673633, + "grad_norm": 0.17579136788845062, + "learning_rate": 3.573782594442946e-05, + "loss": 3.1559, + "step": 13707 + }, + { + "epoch": 0.1298019271329367, + "grad_norm": 0.15099306404590607, + "learning_rate": 3.573539490404694e-05, + "loss": 3.1234, + "step": 13708 + }, + { + "epoch": 0.12983277359851012, + "grad_norm": 0.15171048045158386, + "learning_rate": 3.573296373919396e-05, + "loss": 3.1284, + "step": 13709 + }, + { + "epoch": 0.12986362006408353, + "grad_norm": 0.16494455933570862, + "learning_rate": 3.573053244989873e-05, + "loss": 3.1452, + "step": 13710 + }, + { + "epoch": 0.12989446652965694, + "grad_norm": 0.15527725219726562, + "learning_rate": 3.572810103618941e-05, + "loss": 3.1414, + "step": 13711 + }, + { + "epoch": 0.12992531299523036, + "grad_norm": 0.149014413356781, + "learning_rate": 3.572566949809422e-05, + "loss": 3.14, + "step": 13712 + }, + { + "epoch": 0.12995615946080377, + "grad_norm": 0.174447163939476, + "learning_rate": 3.5723237835641335e-05, + "loss": 3.1336, + "step": 13713 + }, + { + "epoch": 0.12998700592637719, + "grad_norm": 0.15797969698905945, + "learning_rate": 3.572080604885894e-05, + "loss": 3.1166, + "step": 13714 + }, + { + "epoch": 0.13001785239195063, + "grad_norm": 0.15991993248462677, + "learning_rate": 3.5718374137775244e-05, + "loss": 3.1597, + "step": 13715 + }, + { + "epoch": 0.13004869885752404, + "grad_norm": 0.15132935345172882, + "learning_rate": 3.571594210241845e-05, + "loss": 3.1476, + "step": 13716 + }, + { + "epoch": 0.13007954532309746, + "grad_norm": 0.17168046534061432, + "learning_rate": 3.5713509942816735e-05, + "loss": 3.137, + "step": 13717 + }, + { + "epoch": 0.13011039178867087, + "grad_norm": 0.15074747800827026, + "learning_rate": 3.571107765899832e-05, + "loss": 3.1344, + "step": 13718 + }, + { + "epoch": 0.13014123825424428, + "grad_norm": 0.18474379181861877, + "learning_rate": 3.5708645250991396e-05, + "loss": 3.121, + "step": 13719 + }, + { + "epoch": 0.1301720847198177, + "grad_norm": 0.1494545042514801, + "learning_rate": 3.5706212718824156e-05, + "loss": 3.0949, + "step": 13720 + }, + { + "epoch": 0.1302029311853911, + "grad_norm": 0.1450686901807785, + "learning_rate": 3.570378006252482e-05, + "loss": 3.1326, + "step": 13721 + }, + { + "epoch": 0.13023377765096453, + "grad_norm": 0.1553851217031479, + "learning_rate": 3.570134728212159e-05, + "loss": 3.1342, + "step": 13722 + }, + { + "epoch": 0.13026462411653794, + "grad_norm": 0.15105579793453217, + "learning_rate": 3.569891437764266e-05, + "loss": 3.0914, + "step": 13723 + }, + { + "epoch": 0.13029547058211136, + "grad_norm": 0.14782913029193878, + "learning_rate": 3.569648134911626e-05, + "loss": 3.1101, + "step": 13724 + }, + { + "epoch": 0.13032631704768477, + "grad_norm": 0.1510557234287262, + "learning_rate": 3.5694048196570576e-05, + "loss": 3.1059, + "step": 13725 + }, + { + "epoch": 0.13035716351325818, + "grad_norm": 0.15964719653129578, + "learning_rate": 3.5691614920033825e-05, + "loss": 3.1964, + "step": 13726 + }, + { + "epoch": 0.13038800997883163, + "grad_norm": 0.15032216906547546, + "learning_rate": 3.5689181519534236e-05, + "loss": 3.1497, + "step": 13727 + }, + { + "epoch": 0.13041885644440504, + "grad_norm": 0.15273712575435638, + "learning_rate": 3.568674799510001e-05, + "loss": 3.1392, + "step": 13728 + }, + { + "epoch": 0.13044970290997845, + "grad_norm": 0.16052155196666718, + "learning_rate": 3.5684314346759356e-05, + "loss": 3.1536, + "step": 13729 + }, + { + "epoch": 0.13048054937555187, + "grad_norm": 0.18672634661197662, + "learning_rate": 3.56818805745405e-05, + "loss": 3.163, + "step": 13730 + }, + { + "epoch": 0.13051139584112528, + "grad_norm": 0.14993882179260254, + "learning_rate": 3.567944667847166e-05, + "loss": 3.1327, + "step": 13731 + }, + { + "epoch": 0.1305422423066987, + "grad_norm": 0.1496400237083435, + "learning_rate": 3.567701265858104e-05, + "loss": 3.1189, + "step": 13732 + }, + { + "epoch": 0.1305730887722721, + "grad_norm": 0.1521245390176773, + "learning_rate": 3.5674578514896894e-05, + "loss": 3.1384, + "step": 13733 + }, + { + "epoch": 0.13060393523784553, + "grad_norm": 0.15267230570316315, + "learning_rate": 3.567214424744742e-05, + "loss": 3.1162, + "step": 13734 + }, + { + "epoch": 0.13063478170341894, + "grad_norm": 0.15177813172340393, + "learning_rate": 3.566970985626084e-05, + "loss": 3.1367, + "step": 13735 + }, + { + "epoch": 0.13066562816899235, + "grad_norm": 0.15561772882938385, + "learning_rate": 3.566727534136539e-05, + "loss": 3.1822, + "step": 13736 + }, + { + "epoch": 0.13069647463456577, + "grad_norm": 0.1518823653459549, + "learning_rate": 3.5664840702789295e-05, + "loss": 3.1442, + "step": 13737 + }, + { + "epoch": 0.13072732110013918, + "grad_norm": 0.15689373016357422, + "learning_rate": 3.566240594056077e-05, + "loss": 3.1104, + "step": 13738 + }, + { + "epoch": 0.13075816756571262, + "grad_norm": 0.14814388751983643, + "learning_rate": 3.565997105470807e-05, + "loss": 3.1507, + "step": 13739 + }, + { + "epoch": 0.13078901403128604, + "grad_norm": 0.1505560725927353, + "learning_rate": 3.565753604525941e-05, + "loss": 3.1141, + "step": 13740 + }, + { + "epoch": 0.13081986049685945, + "grad_norm": 0.14823633432388306, + "learning_rate": 3.5655100912243016e-05, + "loss": 3.1263, + "step": 13741 + }, + { + "epoch": 0.13085070696243287, + "grad_norm": 0.1527334749698639, + "learning_rate": 3.565266565568713e-05, + "loss": 3.1464, + "step": 13742 + }, + { + "epoch": 0.13088155342800628, + "grad_norm": 0.1474340409040451, + "learning_rate": 3.565023027561999e-05, + "loss": 3.1179, + "step": 13743 + }, + { + "epoch": 0.1309123998935797, + "grad_norm": 0.1522047221660614, + "learning_rate": 3.564779477206983e-05, + "loss": 3.1103, + "step": 13744 + }, + { + "epoch": 0.1309432463591531, + "grad_norm": 0.14718681573867798, + "learning_rate": 3.564535914506489e-05, + "loss": 3.0974, + "step": 13745 + }, + { + "epoch": 0.13097409282472652, + "grad_norm": 0.19776059687137604, + "learning_rate": 3.564292339463341e-05, + "loss": 3.0982, + "step": 13746 + }, + { + "epoch": 0.13100493929029994, + "grad_norm": 0.14683674275875092, + "learning_rate": 3.5640487520803625e-05, + "loss": 3.1128, + "step": 13747 + }, + { + "epoch": 0.13103578575587335, + "grad_norm": 0.14729203283786774, + "learning_rate": 3.563805152360378e-05, + "loss": 3.1465, + "step": 13748 + }, + { + "epoch": 0.13106663222144677, + "grad_norm": 0.15150892734527588, + "learning_rate": 3.563561540306212e-05, + "loss": 3.1421, + "step": 13749 + }, + { + "epoch": 0.13109747868702018, + "grad_norm": 0.14837943017482758, + "learning_rate": 3.5633179159206895e-05, + "loss": 3.1509, + "step": 13750 + }, + { + "epoch": 0.13112832515259362, + "grad_norm": 0.14921347796916962, + "learning_rate": 3.563074279206634e-05, + "loss": 3.1555, + "step": 13751 + }, + { + "epoch": 0.13115917161816704, + "grad_norm": 0.14798063039779663, + "learning_rate": 3.5628306301668726e-05, + "loss": 3.1436, + "step": 13752 + }, + { + "epoch": 0.13119001808374045, + "grad_norm": 0.14682038128376007, + "learning_rate": 3.562586968804227e-05, + "loss": 3.1428, + "step": 13753 + }, + { + "epoch": 0.13122086454931386, + "grad_norm": 0.14932911098003387, + "learning_rate": 3.5623432951215245e-05, + "loss": 3.1336, + "step": 13754 + }, + { + "epoch": 0.13125171101488728, + "grad_norm": 0.1463804543018341, + "learning_rate": 3.56209960912159e-05, + "loss": 3.1363, + "step": 13755 + }, + { + "epoch": 0.1312825574804607, + "grad_norm": 0.1467554122209549, + "learning_rate": 3.561855910807249e-05, + "loss": 3.1614, + "step": 13756 + }, + { + "epoch": 0.1313134039460341, + "grad_norm": 0.1541484296321869, + "learning_rate": 3.561612200181327e-05, + "loss": 3.1115, + "step": 13757 + }, + { + "epoch": 0.13134425041160752, + "grad_norm": 0.14871053397655487, + "learning_rate": 3.5613684772466486e-05, + "loss": 3.1263, + "step": 13758 + }, + { + "epoch": 0.13137509687718094, + "grad_norm": 0.15042823553085327, + "learning_rate": 3.5611247420060414e-05, + "loss": 3.1432, + "step": 13759 + }, + { + "epoch": 0.13140594334275435, + "grad_norm": 0.1436496376991272, + "learning_rate": 3.56088099446233e-05, + "loss": 3.1467, + "step": 13760 + }, + { + "epoch": 0.13143678980832776, + "grad_norm": 0.15927831828594208, + "learning_rate": 3.560637234618341e-05, + "loss": 3.1489, + "step": 13761 + }, + { + "epoch": 0.13146763627390118, + "grad_norm": 0.14668980240821838, + "learning_rate": 3.5603934624769014e-05, + "loss": 3.1297, + "step": 13762 + }, + { + "epoch": 0.13149848273947462, + "grad_norm": 0.14635537564754486, + "learning_rate": 3.560149678040836e-05, + "loss": 3.1389, + "step": 13763 + }, + { + "epoch": 0.13152932920504803, + "grad_norm": 0.14482998847961426, + "learning_rate": 3.559905881312973e-05, + "loss": 3.1373, + "step": 13764 + }, + { + "epoch": 0.13156017567062145, + "grad_norm": 0.15165716409683228, + "learning_rate": 3.559662072296137e-05, + "loss": 3.1187, + "step": 13765 + }, + { + "epoch": 0.13159102213619486, + "grad_norm": 0.14515189826488495, + "learning_rate": 3.559418250993157e-05, + "loss": 3.1056, + "step": 13766 + }, + { + "epoch": 0.13162186860176828, + "grad_norm": 0.14647254347801208, + "learning_rate": 3.559174417406858e-05, + "loss": 3.1352, + "step": 13767 + }, + { + "epoch": 0.1316527150673417, + "grad_norm": 0.1458253562450409, + "learning_rate": 3.5589305715400695e-05, + "loss": 3.0845, + "step": 13768 + }, + { + "epoch": 0.1316835615329151, + "grad_norm": 0.15053364634513855, + "learning_rate": 3.558686713395617e-05, + "loss": 3.1371, + "step": 13769 + }, + { + "epoch": 0.13171440799848852, + "grad_norm": 0.14979112148284912, + "learning_rate": 3.5584428429763275e-05, + "loss": 3.1459, + "step": 13770 + }, + { + "epoch": 0.13174525446406193, + "grad_norm": 0.14695613086223602, + "learning_rate": 3.55819896028503e-05, + "loss": 3.1296, + "step": 13771 + }, + { + "epoch": 0.13177610092963535, + "grad_norm": 0.1488562375307083, + "learning_rate": 3.557955065324551e-05, + "loss": 3.1129, + "step": 13772 + }, + { + "epoch": 0.13180694739520876, + "grad_norm": 0.1473706215620041, + "learning_rate": 3.557711158097719e-05, + "loss": 3.1436, + "step": 13773 + }, + { + "epoch": 0.13183779386078218, + "grad_norm": 0.15079276263713837, + "learning_rate": 3.557467238607361e-05, + "loss": 3.1246, + "step": 13774 + }, + { + "epoch": 0.13186864032635562, + "grad_norm": 0.15287254750728607, + "learning_rate": 3.557223306856306e-05, + "loss": 3.176, + "step": 13775 + }, + { + "epoch": 0.13189948679192903, + "grad_norm": 0.1460389494895935, + "learning_rate": 3.5569793628473836e-05, + "loss": 3.1373, + "step": 13776 + }, + { + "epoch": 0.13193033325750245, + "grad_norm": 0.146465465426445, + "learning_rate": 3.5567354065834194e-05, + "loss": 3.1324, + "step": 13777 + }, + { + "epoch": 0.13196117972307586, + "grad_norm": 0.14862023293972015, + "learning_rate": 3.5564914380672436e-05, + "loss": 3.1406, + "step": 13778 + }, + { + "epoch": 0.13199202618864928, + "grad_norm": 0.15482299029827118, + "learning_rate": 3.5562474573016846e-05, + "loss": 3.1531, + "step": 13779 + }, + { + "epoch": 0.1320228726542227, + "grad_norm": 0.15195101499557495, + "learning_rate": 3.556003464289571e-05, + "loss": 3.1424, + "step": 13780 + }, + { + "epoch": 0.1320537191197961, + "grad_norm": 0.14359626173973083, + "learning_rate": 3.555759459033731e-05, + "loss": 3.0844, + "step": 13781 + }, + { + "epoch": 0.13208456558536952, + "grad_norm": 0.16259045898914337, + "learning_rate": 3.555515441536995e-05, + "loss": 3.1244, + "step": 13782 + }, + { + "epoch": 0.13211541205094293, + "grad_norm": 0.14865529537200928, + "learning_rate": 3.555271411802192e-05, + "loss": 3.1853, + "step": 13783 + }, + { + "epoch": 0.13214625851651635, + "grad_norm": 0.14882250130176544, + "learning_rate": 3.555027369832151e-05, + "loss": 3.1131, + "step": 13784 + }, + { + "epoch": 0.13217710498208976, + "grad_norm": 0.15168195962905884, + "learning_rate": 3.5547833156297015e-05, + "loss": 3.1631, + "step": 13785 + }, + { + "epoch": 0.13220795144766317, + "grad_norm": 0.1577095240354538, + "learning_rate": 3.554539249197674e-05, + "loss": 3.0981, + "step": 13786 + }, + { + "epoch": 0.13223879791323662, + "grad_norm": 0.15304894745349884, + "learning_rate": 3.554295170538897e-05, + "loss": 3.1486, + "step": 13787 + }, + { + "epoch": 0.13226964437881003, + "grad_norm": 0.15387262403964996, + "learning_rate": 3.5540510796562005e-05, + "loss": 3.1118, + "step": 13788 + }, + { + "epoch": 0.13230049084438344, + "grad_norm": 0.1512194573879242, + "learning_rate": 3.553806976552416e-05, + "loss": 3.1499, + "step": 13789 + }, + { + "epoch": 0.13233133730995686, + "grad_norm": 0.1437089443206787, + "learning_rate": 3.553562861230373e-05, + "loss": 3.1376, + "step": 13790 + }, + { + "epoch": 0.13236218377553027, + "grad_norm": 0.15398675203323364, + "learning_rate": 3.553318733692901e-05, + "loss": 3.1085, + "step": 13791 + }, + { + "epoch": 0.1323930302411037, + "grad_norm": 0.15692737698554993, + "learning_rate": 3.553074593942831e-05, + "loss": 3.14, + "step": 13792 + }, + { + "epoch": 0.1324238767066771, + "grad_norm": 0.15575934946537018, + "learning_rate": 3.552830441982995e-05, + "loss": 3.1053, + "step": 13793 + }, + { + "epoch": 0.13245472317225052, + "grad_norm": 0.15120598673820496, + "learning_rate": 3.552586277816222e-05, + "loss": 3.158, + "step": 13794 + }, + { + "epoch": 0.13248556963782393, + "grad_norm": 0.1479341983795166, + "learning_rate": 3.552342101445344e-05, + "loss": 3.1397, + "step": 13795 + }, + { + "epoch": 0.13251641610339734, + "grad_norm": 0.15360628068447113, + "learning_rate": 3.5520979128731915e-05, + "loss": 3.1562, + "step": 13796 + }, + { + "epoch": 0.13254726256897076, + "grad_norm": 0.15190528333187103, + "learning_rate": 3.5518537121025955e-05, + "loss": 3.1319, + "step": 13797 + }, + { + "epoch": 0.13257810903454417, + "grad_norm": 0.15589599311351776, + "learning_rate": 3.5516094991363886e-05, + "loss": 3.1773, + "step": 13798 + }, + { + "epoch": 0.13260895550011761, + "grad_norm": 0.15404534339904785, + "learning_rate": 3.551365273977401e-05, + "loss": 3.1182, + "step": 13799 + }, + { + "epoch": 0.13263980196569103, + "grad_norm": 0.146166130900383, + "learning_rate": 3.551121036628465e-05, + "loss": 3.1209, + "step": 13800 + }, + { + "epoch": 0.13267064843126444, + "grad_norm": 0.14874795079231262, + "learning_rate": 3.5508767870924126e-05, + "loss": 3.108, + "step": 13801 + }, + { + "epoch": 0.13270149489683786, + "grad_norm": 0.14999356865882874, + "learning_rate": 3.550632525372075e-05, + "loss": 3.1284, + "step": 13802 + }, + { + "epoch": 0.13273234136241127, + "grad_norm": 0.15808823704719543, + "learning_rate": 3.550388251470284e-05, + "loss": 3.1276, + "step": 13803 + }, + { + "epoch": 0.13276318782798469, + "grad_norm": 0.1643029749393463, + "learning_rate": 3.550143965389874e-05, + "loss": 3.1317, + "step": 13804 + }, + { + "epoch": 0.1327940342935581, + "grad_norm": 0.14776961505413055, + "learning_rate": 3.549899667133675e-05, + "loss": 3.1121, + "step": 13805 + }, + { + "epoch": 0.13282488075913151, + "grad_norm": 0.16709722578525543, + "learning_rate": 3.5496553567045197e-05, + "loss": 3.1392, + "step": 13806 + }, + { + "epoch": 0.13285572722470493, + "grad_norm": 0.1582997888326645, + "learning_rate": 3.549411034105242e-05, + "loss": 3.156, + "step": 13807 + }, + { + "epoch": 0.13288657369027834, + "grad_norm": 0.15311433374881744, + "learning_rate": 3.5491666993386733e-05, + "loss": 3.1244, + "step": 13808 + }, + { + "epoch": 0.13291742015585176, + "grad_norm": 0.1688113808631897, + "learning_rate": 3.548922352407648e-05, + "loss": 3.1447, + "step": 13809 + }, + { + "epoch": 0.13294826662142517, + "grad_norm": 0.15343767404556274, + "learning_rate": 3.548677993314998e-05, + "loss": 3.1043, + "step": 13810 + }, + { + "epoch": 0.1329791130869986, + "grad_norm": 0.16718600690364838, + "learning_rate": 3.548433622063556e-05, + "loss": 3.1055, + "step": 13811 + }, + { + "epoch": 0.13300995955257203, + "grad_norm": 0.15453007817268372, + "learning_rate": 3.548189238656158e-05, + "loss": 3.1453, + "step": 13812 + }, + { + "epoch": 0.13304080601814544, + "grad_norm": 0.15548966825008392, + "learning_rate": 3.5479448430956346e-05, + "loss": 3.1448, + "step": 13813 + }, + { + "epoch": 0.13307165248371886, + "grad_norm": 0.16038718819618225, + "learning_rate": 3.54770043538482e-05, + "loss": 3.1446, + "step": 13814 + }, + { + "epoch": 0.13310249894929227, + "grad_norm": 0.1510186344385147, + "learning_rate": 3.547456015526549e-05, + "loss": 3.1356, + "step": 13815 + }, + { + "epoch": 0.13313334541486568, + "grad_norm": 0.14561094343662262, + "learning_rate": 3.547211583523655e-05, + "loss": 3.1226, + "step": 13816 + }, + { + "epoch": 0.1331641918804391, + "grad_norm": 0.15091748535633087, + "learning_rate": 3.546967139378972e-05, + "loss": 3.1377, + "step": 13817 + }, + { + "epoch": 0.1331950383460125, + "grad_norm": 0.15671755373477936, + "learning_rate": 3.5467226830953344e-05, + "loss": 3.1451, + "step": 13818 + }, + { + "epoch": 0.13322588481158593, + "grad_norm": 0.1446581780910492, + "learning_rate": 3.546478214675576e-05, + "loss": 3.1325, + "step": 13819 + }, + { + "epoch": 0.13325673127715934, + "grad_norm": 0.14813022315502167, + "learning_rate": 3.546233734122531e-05, + "loss": 3.1265, + "step": 13820 + }, + { + "epoch": 0.13328757774273275, + "grad_norm": 1.6891908645629883, + "learning_rate": 3.545989241439036e-05, + "loss": 3.1374, + "step": 13821 + }, + { + "epoch": 0.13331842420830617, + "grad_norm": 0.15429557859897614, + "learning_rate": 3.545744736627923e-05, + "loss": 3.1438, + "step": 13822 + }, + { + "epoch": 0.1333492706738796, + "grad_norm": 0.3475809693336487, + "learning_rate": 3.5455002196920284e-05, + "loss": 3.1384, + "step": 13823 + }, + { + "epoch": 0.13338011713945302, + "grad_norm": 0.15697503089904785, + "learning_rate": 3.5452556906341875e-05, + "loss": 3.1339, + "step": 13824 + }, + { + "epoch": 0.13341096360502644, + "grad_norm": 0.16475909948349, + "learning_rate": 3.5450111494572344e-05, + "loss": 3.1207, + "step": 13825 + }, + { + "epoch": 0.13344181007059985, + "grad_norm": 0.2653828263282776, + "learning_rate": 3.5447665961640056e-05, + "loss": 3.1445, + "step": 13826 + }, + { + "epoch": 0.13347265653617327, + "grad_norm": 0.1612945795059204, + "learning_rate": 3.5445220307573356e-05, + "loss": 3.1313, + "step": 13827 + }, + { + "epoch": 0.13350350300174668, + "grad_norm": 0.15795408189296722, + "learning_rate": 3.5442774532400615e-05, + "loss": 3.1042, + "step": 13828 + }, + { + "epoch": 0.1335343494673201, + "grad_norm": 0.16197024285793304, + "learning_rate": 3.544032863615016e-05, + "loss": 3.1281, + "step": 13829 + }, + { + "epoch": 0.1335651959328935, + "grad_norm": 0.1707334965467453, + "learning_rate": 3.543788261885038e-05, + "loss": 3.1654, + "step": 13830 + }, + { + "epoch": 0.13359604239846692, + "grad_norm": 0.15776053071022034, + "learning_rate": 3.543543648052962e-05, + "loss": 3.1537, + "step": 13831 + }, + { + "epoch": 0.13362688886404034, + "grad_norm": 0.1528630554676056, + "learning_rate": 3.5432990221216244e-05, + "loss": 3.1312, + "step": 13832 + }, + { + "epoch": 0.13365773532961375, + "grad_norm": 0.1611369103193283, + "learning_rate": 3.5430543840938616e-05, + "loss": 3.1344, + "step": 13833 + }, + { + "epoch": 0.13368858179518717, + "grad_norm": 0.160429447889328, + "learning_rate": 3.542809733972511e-05, + "loss": 3.1432, + "step": 13834 + }, + { + "epoch": 0.1337194282607606, + "grad_norm": 0.15917281806468964, + "learning_rate": 3.542565071760406e-05, + "loss": 3.1491, + "step": 13835 + }, + { + "epoch": 0.13375027472633402, + "grad_norm": 0.1487332582473755, + "learning_rate": 3.542320397460388e-05, + "loss": 3.0606, + "step": 13836 + }, + { + "epoch": 0.13378112119190744, + "grad_norm": 0.15074695646762848, + "learning_rate": 3.542075711075289e-05, + "loss": 3.144, + "step": 13837 + }, + { + "epoch": 0.13381196765748085, + "grad_norm": 0.15653125941753387, + "learning_rate": 3.54183101260795e-05, + "loss": 3.1426, + "step": 13838 + }, + { + "epoch": 0.13384281412305427, + "grad_norm": 0.15793487429618835, + "learning_rate": 3.541586302061206e-05, + "loss": 3.1478, + "step": 13839 + }, + { + "epoch": 0.13387366058862768, + "grad_norm": 0.15630772709846497, + "learning_rate": 3.541341579437895e-05, + "loss": 3.1519, + "step": 13840 + }, + { + "epoch": 0.1339045070542011, + "grad_norm": 0.15391427278518677, + "learning_rate": 3.541096844740854e-05, + "loss": 3.1411, + "step": 13841 + }, + { + "epoch": 0.1339353535197745, + "grad_norm": 0.15784993767738342, + "learning_rate": 3.54085209797292e-05, + "loss": 3.1395, + "step": 13842 + }, + { + "epoch": 0.13396619998534792, + "grad_norm": 0.15202219784259796, + "learning_rate": 3.540607339136933e-05, + "loss": 3.1458, + "step": 13843 + }, + { + "epoch": 0.13399704645092134, + "grad_norm": 0.151865616440773, + "learning_rate": 3.540362568235728e-05, + "loss": 3.118, + "step": 13844 + }, + { + "epoch": 0.13402789291649475, + "grad_norm": 0.15340472757816315, + "learning_rate": 3.540117785272144e-05, + "loss": 3.1449, + "step": 13845 + }, + { + "epoch": 0.13405873938206816, + "grad_norm": 0.1517673283815384, + "learning_rate": 3.53987299024902e-05, + "loss": 3.1674, + "step": 13846 + }, + { + "epoch": 0.1340895858476416, + "grad_norm": 0.15338470041751862, + "learning_rate": 3.5396281831691934e-05, + "loss": 3.121, + "step": 13847 + }, + { + "epoch": 0.13412043231321502, + "grad_norm": 0.1468110978603363, + "learning_rate": 3.539383364035502e-05, + "loss": 3.1566, + "step": 13848 + }, + { + "epoch": 0.13415127877878844, + "grad_norm": 0.14981935918331146, + "learning_rate": 3.5391385328507865e-05, + "loss": 3.097, + "step": 13849 + }, + { + "epoch": 0.13418212524436185, + "grad_norm": 0.1500837355852127, + "learning_rate": 3.538893689617883e-05, + "loss": 3.1298, + "step": 13850 + }, + { + "epoch": 0.13421297170993526, + "grad_norm": 0.15063293278217316, + "learning_rate": 3.538648834339632e-05, + "loss": 3.1174, + "step": 13851 + }, + { + "epoch": 0.13424381817550868, + "grad_norm": 0.15329061448574066, + "learning_rate": 3.538403967018872e-05, + "loss": 3.1353, + "step": 13852 + }, + { + "epoch": 0.1342746646410821, + "grad_norm": 0.14948202669620514, + "learning_rate": 3.5381590876584415e-05, + "loss": 3.1447, + "step": 13853 + }, + { + "epoch": 0.1343055111066555, + "grad_norm": 0.15335744619369507, + "learning_rate": 3.537914196261181e-05, + "loss": 3.1131, + "step": 13854 + }, + { + "epoch": 0.13433635757222892, + "grad_norm": 0.36719053983688354, + "learning_rate": 3.537669292829929e-05, + "loss": 3.1261, + "step": 13855 + }, + { + "epoch": 0.13436720403780233, + "grad_norm": 0.1555035561323166, + "learning_rate": 3.537424377367524e-05, + "loss": 3.0884, + "step": 13856 + }, + { + "epoch": 0.13439805050337575, + "grad_norm": 0.1587074249982834, + "learning_rate": 3.537179449876808e-05, + "loss": 3.1717, + "step": 13857 + }, + { + "epoch": 0.13442889696894916, + "grad_norm": 0.16234692931175232, + "learning_rate": 3.5369345103606194e-05, + "loss": 3.1176, + "step": 13858 + }, + { + "epoch": 0.1344597434345226, + "grad_norm": 0.16291023790836334, + "learning_rate": 3.536689558821797e-05, + "loss": 3.162, + "step": 13859 + }, + { + "epoch": 0.13449058990009602, + "grad_norm": 0.1548725664615631, + "learning_rate": 3.536444595263184e-05, + "loss": 3.1214, + "step": 13860 + }, + { + "epoch": 0.13452143636566943, + "grad_norm": 0.16476492583751678, + "learning_rate": 3.536199619687618e-05, + "loss": 3.1409, + "step": 13861 + }, + { + "epoch": 0.13455228283124285, + "grad_norm": 0.15146704018115997, + "learning_rate": 3.53595463209794e-05, + "loss": 3.1138, + "step": 13862 + }, + { + "epoch": 0.13458312929681626, + "grad_norm": 0.15778988599777222, + "learning_rate": 3.5357096324969906e-05, + "loss": 3.1237, + "step": 13863 + }, + { + "epoch": 0.13461397576238968, + "grad_norm": 0.15741974115371704, + "learning_rate": 3.53546462088761e-05, + "loss": 3.1397, + "step": 13864 + }, + { + "epoch": 0.1346448222279631, + "grad_norm": 0.14911024272441864, + "learning_rate": 3.53521959727264e-05, + "loss": 3.0826, + "step": 13865 + }, + { + "epoch": 0.1346756686935365, + "grad_norm": 0.15921297669410706, + "learning_rate": 3.5349745616549205e-05, + "loss": 3.118, + "step": 13866 + }, + { + "epoch": 0.13470651515910992, + "grad_norm": 0.15102213621139526, + "learning_rate": 3.534729514037293e-05, + "loss": 3.1272, + "step": 13867 + }, + { + "epoch": 0.13473736162468333, + "grad_norm": 0.14488482475280762, + "learning_rate": 3.534484454422599e-05, + "loss": 3.1573, + "step": 13868 + }, + { + "epoch": 0.13476820809025675, + "grad_norm": 0.15358825027942657, + "learning_rate": 3.534239382813679e-05, + "loss": 3.1477, + "step": 13869 + }, + { + "epoch": 0.13479905455583016, + "grad_norm": 0.15394432842731476, + "learning_rate": 3.533994299213375e-05, + "loss": 3.1765, + "step": 13870 + }, + { + "epoch": 0.1348299010214036, + "grad_norm": 0.1496194303035736, + "learning_rate": 3.533749203624528e-05, + "loss": 3.1273, + "step": 13871 + }, + { + "epoch": 0.13486074748697702, + "grad_norm": 0.16061931848526, + "learning_rate": 3.53350409604998e-05, + "loss": 3.1406, + "step": 13872 + }, + { + "epoch": 0.13489159395255043, + "grad_norm": 0.16072145104408264, + "learning_rate": 3.533258976492574e-05, + "loss": 3.1125, + "step": 13873 + }, + { + "epoch": 0.13492244041812385, + "grad_norm": 0.151466965675354, + "learning_rate": 3.533013844955151e-05, + "loss": 3.1231, + "step": 13874 + }, + { + "epoch": 0.13495328688369726, + "grad_norm": 0.1504213958978653, + "learning_rate": 3.5327687014405525e-05, + "loss": 3.1296, + "step": 13875 + }, + { + "epoch": 0.13498413334927067, + "grad_norm": 0.14592541754245758, + "learning_rate": 3.5325235459516215e-05, + "loss": 3.1533, + "step": 13876 + }, + { + "epoch": 0.1350149798148441, + "grad_norm": 0.15875598788261414, + "learning_rate": 3.532278378491202e-05, + "loss": 3.1413, + "step": 13877 + }, + { + "epoch": 0.1350458262804175, + "grad_norm": 0.1664288491010666, + "learning_rate": 3.5320331990621335e-05, + "loss": 3.1425, + "step": 13878 + }, + { + "epoch": 0.13507667274599092, + "grad_norm": 0.15198400616645813, + "learning_rate": 3.5317880076672604e-05, + "loss": 3.0815, + "step": 13879 + }, + { + "epoch": 0.13510751921156433, + "grad_norm": 0.145096093416214, + "learning_rate": 3.531542804309426e-05, + "loss": 3.1375, + "step": 13880 + }, + { + "epoch": 0.13513836567713775, + "grad_norm": 0.15605680644512177, + "learning_rate": 3.5312975889914726e-05, + "loss": 3.1782, + "step": 13881 + }, + { + "epoch": 0.13516921214271116, + "grad_norm": 0.1521175354719162, + "learning_rate": 3.5310523617162425e-05, + "loss": 3.1375, + "step": 13882 + }, + { + "epoch": 0.1352000586082846, + "grad_norm": 0.1647539585828781, + "learning_rate": 3.530807122486581e-05, + "loss": 3.1579, + "step": 13883 + }, + { + "epoch": 0.13523090507385802, + "grad_norm": 0.15908017754554749, + "learning_rate": 3.530561871305329e-05, + "loss": 3.1369, + "step": 13884 + }, + { + "epoch": 0.13526175153943143, + "grad_norm": 0.15591803193092346, + "learning_rate": 3.530316608175333e-05, + "loss": 3.1427, + "step": 13885 + }, + { + "epoch": 0.13529259800500484, + "grad_norm": 0.1479579359292984, + "learning_rate": 3.530071333099434e-05, + "loss": 3.0811, + "step": 13886 + }, + { + "epoch": 0.13532344447057826, + "grad_norm": 0.16030879318714142, + "learning_rate": 3.529826046080477e-05, + "loss": 3.1516, + "step": 13887 + }, + { + "epoch": 0.13535429093615167, + "grad_norm": 0.15512463450431824, + "learning_rate": 3.529580747121306e-05, + "loss": 3.1434, + "step": 13888 + }, + { + "epoch": 0.1353851374017251, + "grad_norm": 0.15776973962783813, + "learning_rate": 3.529335436224766e-05, + "loss": 3.1267, + "step": 13889 + }, + { + "epoch": 0.1354159838672985, + "grad_norm": 0.15667328238487244, + "learning_rate": 3.5290901133936984e-05, + "loss": 3.1529, + "step": 13890 + }, + { + "epoch": 0.13544683033287191, + "grad_norm": 0.15486733615398407, + "learning_rate": 3.5288447786309495e-05, + "loss": 3.122, + "step": 13891 + }, + { + "epoch": 0.13547767679844533, + "grad_norm": 0.1548655927181244, + "learning_rate": 3.5285994319393645e-05, + "loss": 3.1321, + "step": 13892 + }, + { + "epoch": 0.13550852326401874, + "grad_norm": 0.1589503139257431, + "learning_rate": 3.5283540733217873e-05, + "loss": 3.1747, + "step": 13893 + }, + { + "epoch": 0.13553936972959216, + "grad_norm": 0.1478351652622223, + "learning_rate": 3.528108702781062e-05, + "loss": 3.1027, + "step": 13894 + }, + { + "epoch": 0.1355702161951656, + "grad_norm": 0.1523900330066681, + "learning_rate": 3.5278633203200346e-05, + "loss": 3.1082, + "step": 13895 + }, + { + "epoch": 0.135601062660739, + "grad_norm": 0.1471651941537857, + "learning_rate": 3.52761792594155e-05, + "loss": 3.135, + "step": 13896 + }, + { + "epoch": 0.13563190912631243, + "grad_norm": 0.1601828634738922, + "learning_rate": 3.527372519648452e-05, + "loss": 3.1239, + "step": 13897 + }, + { + "epoch": 0.13566275559188584, + "grad_norm": 0.15599636733531952, + "learning_rate": 3.527127101443587e-05, + "loss": 3.1474, + "step": 13898 + }, + { + "epoch": 0.13569360205745926, + "grad_norm": 0.14944346249103546, + "learning_rate": 3.526881671329801e-05, + "loss": 3.1462, + "step": 13899 + }, + { + "epoch": 0.13572444852303267, + "grad_norm": 0.15220481157302856, + "learning_rate": 3.5266362293099395e-05, + "loss": 3.1203, + "step": 13900 + }, + { + "epoch": 0.13575529498860608, + "grad_norm": 0.15996037423610687, + "learning_rate": 3.5263907753868477e-05, + "loss": 3.137, + "step": 13901 + }, + { + "epoch": 0.1357861414541795, + "grad_norm": 0.16290399432182312, + "learning_rate": 3.526145309563372e-05, + "loss": 3.1495, + "step": 13902 + }, + { + "epoch": 0.1358169879197529, + "grad_norm": 0.14873839914798737, + "learning_rate": 3.525899831842356e-05, + "loss": 3.1231, + "step": 13903 + }, + { + "epoch": 0.13584783438532633, + "grad_norm": 0.1513223648071289, + "learning_rate": 3.5256543422266505e-05, + "loss": 3.1212, + "step": 13904 + }, + { + "epoch": 0.13587868085089974, + "grad_norm": 0.1531175971031189, + "learning_rate": 3.525408840719098e-05, + "loss": 3.1099, + "step": 13905 + }, + { + "epoch": 0.13590952731647318, + "grad_norm": 0.14603394269943237, + "learning_rate": 3.525163327322546e-05, + "loss": 3.1344, + "step": 13906 + }, + { + "epoch": 0.1359403737820466, + "grad_norm": 0.1562367081642151, + "learning_rate": 3.524917802039842e-05, + "loss": 3.1244, + "step": 13907 + }, + { + "epoch": 0.13597122024762, + "grad_norm": 0.15462686121463776, + "learning_rate": 3.5246722648738315e-05, + "loss": 3.0995, + "step": 13908 + }, + { + "epoch": 0.13600206671319343, + "grad_norm": 0.15441736578941345, + "learning_rate": 3.524426715827362e-05, + "loss": 3.11, + "step": 13909 + }, + { + "epoch": 0.13603291317876684, + "grad_norm": 0.14553432166576385, + "learning_rate": 3.5241811549032806e-05, + "loss": 3.1168, + "step": 13910 + }, + { + "epoch": 0.13606375964434025, + "grad_norm": 0.15270423889160156, + "learning_rate": 3.5239355821044334e-05, + "loss": 3.118, + "step": 13911 + }, + { + "epoch": 0.13609460610991367, + "grad_norm": 0.1503232717514038, + "learning_rate": 3.523689997433669e-05, + "loss": 3.1427, + "step": 13912 + }, + { + "epoch": 0.13612545257548708, + "grad_norm": 0.1478797197341919, + "learning_rate": 3.523444400893835e-05, + "loss": 3.1061, + "step": 13913 + }, + { + "epoch": 0.1361562990410605, + "grad_norm": 0.14714445173740387, + "learning_rate": 3.523198792487777e-05, + "loss": 3.1257, + "step": 13914 + }, + { + "epoch": 0.1361871455066339, + "grad_norm": 0.14675548672676086, + "learning_rate": 3.522953172218344e-05, + "loss": 3.1674, + "step": 13915 + }, + { + "epoch": 0.13621799197220733, + "grad_norm": 0.15265227854251862, + "learning_rate": 3.522707540088384e-05, + "loss": 3.1375, + "step": 13916 + }, + { + "epoch": 0.13624883843778074, + "grad_norm": 0.14508718252182007, + "learning_rate": 3.522461896100744e-05, + "loss": 3.1467, + "step": 13917 + }, + { + "epoch": 0.13627968490335418, + "grad_norm": 0.1470165252685547, + "learning_rate": 3.5222162402582745e-05, + "loss": 3.1076, + "step": 13918 + }, + { + "epoch": 0.1363105313689276, + "grad_norm": 0.1505529284477234, + "learning_rate": 3.521970572563821e-05, + "loss": 3.1042, + "step": 13919 + }, + { + "epoch": 0.136341377834501, + "grad_norm": 0.15487468242645264, + "learning_rate": 3.5217248930202316e-05, + "loss": 3.0974, + "step": 13920 + }, + { + "epoch": 0.13637222430007442, + "grad_norm": 0.15683011710643768, + "learning_rate": 3.5214792016303573e-05, + "loss": 3.1127, + "step": 13921 + }, + { + "epoch": 0.13640307076564784, + "grad_norm": 0.1476069539785385, + "learning_rate": 3.5212334983970454e-05, + "loss": 3.0896, + "step": 13922 + }, + { + "epoch": 0.13643391723122125, + "grad_norm": 0.14777961373329163, + "learning_rate": 3.520987783323145e-05, + "loss": 3.1125, + "step": 13923 + }, + { + "epoch": 0.13646476369679467, + "grad_norm": 0.15850991010665894, + "learning_rate": 3.5207420564115045e-05, + "loss": 3.1537, + "step": 13924 + }, + { + "epoch": 0.13649561016236808, + "grad_norm": 0.15628354251384735, + "learning_rate": 3.520496317664973e-05, + "loss": 3.1309, + "step": 13925 + }, + { + "epoch": 0.1365264566279415, + "grad_norm": 0.14741185307502747, + "learning_rate": 3.520250567086399e-05, + "loss": 3.1375, + "step": 13926 + }, + { + "epoch": 0.1365573030935149, + "grad_norm": 0.1517927348613739, + "learning_rate": 3.520004804678635e-05, + "loss": 3.1396, + "step": 13927 + }, + { + "epoch": 0.13658814955908832, + "grad_norm": 0.15009504556655884, + "learning_rate": 3.5197590304445274e-05, + "loss": 3.122, + "step": 13928 + }, + { + "epoch": 0.13661899602466174, + "grad_norm": 0.15104655921459198, + "learning_rate": 3.519513244386926e-05, + "loss": 3.1349, + "step": 13929 + }, + { + "epoch": 0.13664984249023518, + "grad_norm": 0.15113763511180878, + "learning_rate": 3.519267446508682e-05, + "loss": 3.1267, + "step": 13930 + }, + { + "epoch": 0.1366806889558086, + "grad_norm": 0.15957531332969666, + "learning_rate": 3.519021636812644e-05, + "loss": 3.1061, + "step": 13931 + }, + { + "epoch": 0.136711535421382, + "grad_norm": 0.17637336254119873, + "learning_rate": 3.518775815301662e-05, + "loss": 3.1138, + "step": 13932 + }, + { + "epoch": 0.13674238188695542, + "grad_norm": 0.15172962844371796, + "learning_rate": 3.518529981978588e-05, + "loss": 3.1313, + "step": 13933 + }, + { + "epoch": 0.13677322835252884, + "grad_norm": 0.15780757367610931, + "learning_rate": 3.51828413684627e-05, + "loss": 3.1359, + "step": 13934 + }, + { + "epoch": 0.13680407481810225, + "grad_norm": 0.15034903585910797, + "learning_rate": 3.518038279907559e-05, + "loss": 3.1305, + "step": 13935 + }, + { + "epoch": 0.13683492128367566, + "grad_norm": 0.5129814147949219, + "learning_rate": 3.5177924111653064e-05, + "loss": 3.1636, + "step": 13936 + }, + { + "epoch": 0.13686576774924908, + "grad_norm": 0.1719985008239746, + "learning_rate": 3.517546530622363e-05, + "loss": 3.1424, + "step": 13937 + }, + { + "epoch": 0.1368966142148225, + "grad_norm": 0.15942925214767456, + "learning_rate": 3.5173006382815775e-05, + "loss": 3.1554, + "step": 13938 + }, + { + "epoch": 0.1369274606803959, + "grad_norm": 0.15512526035308838, + "learning_rate": 3.5170547341458036e-05, + "loss": 3.1151, + "step": 13939 + }, + { + "epoch": 0.13695830714596932, + "grad_norm": 0.16892123222351074, + "learning_rate": 3.516808818217892e-05, + "loss": 3.1324, + "step": 13940 + }, + { + "epoch": 0.13698915361154274, + "grad_norm": 0.1519538164138794, + "learning_rate": 3.516562890500692e-05, + "loss": 3.1218, + "step": 13941 + }, + { + "epoch": 0.13702000007711618, + "grad_norm": 0.1779552847146988, + "learning_rate": 3.5163169509970556e-05, + "loss": 3.1227, + "step": 13942 + }, + { + "epoch": 0.1370508465426896, + "grad_norm": 0.16031818091869354, + "learning_rate": 3.516070999709836e-05, + "loss": 3.1547, + "step": 13943 + }, + { + "epoch": 0.137081693008263, + "grad_norm": 0.1484771966934204, + "learning_rate": 3.515825036641883e-05, + "loss": 3.1439, + "step": 13944 + }, + { + "epoch": 0.13711253947383642, + "grad_norm": 0.17232045531272888, + "learning_rate": 3.51557906179605e-05, + "loss": 3.1293, + "step": 13945 + }, + { + "epoch": 0.13714338593940983, + "grad_norm": 0.15609872341156006, + "learning_rate": 3.515333075175188e-05, + "loss": 3.1396, + "step": 13946 + }, + { + "epoch": 0.13717423240498325, + "grad_norm": 0.15544714033603668, + "learning_rate": 3.515087076782149e-05, + "loss": 3.0851, + "step": 13947 + }, + { + "epoch": 0.13720507887055666, + "grad_norm": 0.16002367436885834, + "learning_rate": 3.514841066619784e-05, + "loss": 3.1675, + "step": 13948 + }, + { + "epoch": 0.13723592533613008, + "grad_norm": 0.15025345981121063, + "learning_rate": 3.514595044690948e-05, + "loss": 3.1393, + "step": 13949 + }, + { + "epoch": 0.1372667718017035, + "grad_norm": 0.15131433308124542, + "learning_rate": 3.5143490109984923e-05, + "loss": 3.1516, + "step": 13950 + }, + { + "epoch": 0.1372976182672769, + "grad_norm": 0.15773087739944458, + "learning_rate": 3.514102965545269e-05, + "loss": 3.1268, + "step": 13951 + }, + { + "epoch": 0.13732846473285032, + "grad_norm": 0.15301965177059174, + "learning_rate": 3.513856908334132e-05, + "loss": 3.1321, + "step": 13952 + }, + { + "epoch": 0.13735931119842373, + "grad_norm": 0.14918123185634613, + "learning_rate": 3.513610839367931e-05, + "loss": 3.1259, + "step": 13953 + }, + { + "epoch": 0.13739015766399718, + "grad_norm": 0.15178395807743073, + "learning_rate": 3.5133647586495235e-05, + "loss": 3.108, + "step": 13954 + }, + { + "epoch": 0.1374210041295706, + "grad_norm": 0.16135157644748688, + "learning_rate": 3.513118666181761e-05, + "loss": 3.1035, + "step": 13955 + }, + { + "epoch": 0.137451850595144, + "grad_norm": 0.15978868305683136, + "learning_rate": 3.512872561967495e-05, + "loss": 3.1363, + "step": 13956 + }, + { + "epoch": 0.13748269706071742, + "grad_norm": 0.15135148167610168, + "learning_rate": 3.5126264460095806e-05, + "loss": 3.105, + "step": 13957 + }, + { + "epoch": 0.13751354352629083, + "grad_norm": 0.14946375787258148, + "learning_rate": 3.5123803183108715e-05, + "loss": 3.1524, + "step": 13958 + }, + { + "epoch": 0.13754438999186425, + "grad_norm": 0.14943861961364746, + "learning_rate": 3.5121341788742204e-05, + "loss": 3.1355, + "step": 13959 + }, + { + "epoch": 0.13757523645743766, + "grad_norm": 0.1580139696598053, + "learning_rate": 3.511888027702482e-05, + "loss": 3.133, + "step": 13960 + }, + { + "epoch": 0.13760608292301107, + "grad_norm": 0.15706081688404083, + "learning_rate": 3.511641864798511e-05, + "loss": 3.1619, + "step": 13961 + }, + { + "epoch": 0.1376369293885845, + "grad_norm": 0.14430497586727142, + "learning_rate": 3.511395690165159e-05, + "loss": 3.1241, + "step": 13962 + }, + { + "epoch": 0.1376677758541579, + "grad_norm": 0.15966998040676117, + "learning_rate": 3.511149503805283e-05, + "loss": 3.1449, + "step": 13963 + }, + { + "epoch": 0.13769862231973132, + "grad_norm": 0.15710866451263428, + "learning_rate": 3.510903305721735e-05, + "loss": 3.144, + "step": 13964 + }, + { + "epoch": 0.13772946878530473, + "grad_norm": 0.14617066085338593, + "learning_rate": 3.5106570959173704e-05, + "loss": 3.1361, + "step": 13965 + }, + { + "epoch": 0.13776031525087817, + "grad_norm": 0.1504903882741928, + "learning_rate": 3.510410874395045e-05, + "loss": 3.1346, + "step": 13966 + }, + { + "epoch": 0.1377911617164516, + "grad_norm": 0.14911261200904846, + "learning_rate": 3.510164641157613e-05, + "loss": 3.0956, + "step": 13967 + }, + { + "epoch": 0.137822008182025, + "grad_norm": 0.1508685052394867, + "learning_rate": 3.5099183962079284e-05, + "loss": 3.1538, + "step": 13968 + }, + { + "epoch": 0.13785285464759842, + "grad_norm": 0.1505390852689743, + "learning_rate": 3.5096721395488475e-05, + "loss": 3.1237, + "step": 13969 + }, + { + "epoch": 0.13788370111317183, + "grad_norm": 0.15858040750026703, + "learning_rate": 3.509425871183224e-05, + "loss": 3.1124, + "step": 13970 + }, + { + "epoch": 0.13791454757874524, + "grad_norm": 0.1524621546268463, + "learning_rate": 3.5091795911139146e-05, + "loss": 3.0867, + "step": 13971 + }, + { + "epoch": 0.13794539404431866, + "grad_norm": 0.16299620270729065, + "learning_rate": 3.5089332993437744e-05, + "loss": 3.1425, + "step": 13972 + }, + { + "epoch": 0.13797624050989207, + "grad_norm": 0.14969108998775482, + "learning_rate": 3.5086869958756586e-05, + "loss": 3.1087, + "step": 13973 + }, + { + "epoch": 0.1380070869754655, + "grad_norm": 0.1567852646112442, + "learning_rate": 3.508440680712424e-05, + "loss": 3.1545, + "step": 13974 + }, + { + "epoch": 0.1380379334410389, + "grad_norm": 0.15643157064914703, + "learning_rate": 3.508194353856925e-05, + "loss": 3.1518, + "step": 13975 + }, + { + "epoch": 0.13806877990661232, + "grad_norm": 0.1500614732503891, + "learning_rate": 3.507948015312017e-05, + "loss": 3.1433, + "step": 13976 + }, + { + "epoch": 0.13809962637218573, + "grad_norm": 0.15044540166854858, + "learning_rate": 3.5077016650805595e-05, + "loss": 3.1246, + "step": 13977 + }, + { + "epoch": 0.13813047283775917, + "grad_norm": 0.15052801370620728, + "learning_rate": 3.5074553031654064e-05, + "loss": 3.1219, + "step": 13978 + }, + { + "epoch": 0.13816131930333259, + "grad_norm": 0.15144404768943787, + "learning_rate": 3.507208929569414e-05, + "loss": 3.156, + "step": 13979 + }, + { + "epoch": 0.138192165768906, + "grad_norm": 0.14386069774627686, + "learning_rate": 3.50696254429544e-05, + "loss": 3.0902, + "step": 13980 + }, + { + "epoch": 0.13822301223447941, + "grad_norm": 0.1564427614212036, + "learning_rate": 3.506716147346339e-05, + "loss": 3.1405, + "step": 13981 + }, + { + "epoch": 0.13825385870005283, + "grad_norm": 0.14910274744033813, + "learning_rate": 3.506469738724971e-05, + "loss": 3.1291, + "step": 13982 + }, + { + "epoch": 0.13828470516562624, + "grad_norm": 0.15084387362003326, + "learning_rate": 3.5062233184341896e-05, + "loss": 3.1718, + "step": 13983 + }, + { + "epoch": 0.13831555163119966, + "grad_norm": 0.1551840454339981, + "learning_rate": 3.505976886476855e-05, + "loss": 3.1538, + "step": 13984 + }, + { + "epoch": 0.13834639809677307, + "grad_norm": 0.15305761992931366, + "learning_rate": 3.505730442855822e-05, + "loss": 3.1145, + "step": 13985 + }, + { + "epoch": 0.13837724456234649, + "grad_norm": 0.1458127647638321, + "learning_rate": 3.5054839875739485e-05, + "loss": 3.1305, + "step": 13986 + }, + { + "epoch": 0.1384080910279199, + "grad_norm": 0.14757780730724335, + "learning_rate": 3.5052375206340934e-05, + "loss": 3.1244, + "step": 13987 + }, + { + "epoch": 0.1384389374934933, + "grad_norm": 0.14915139973163605, + "learning_rate": 3.504991042039113e-05, + "loss": 3.1206, + "step": 13988 + }, + { + "epoch": 0.13846978395906673, + "grad_norm": 0.14517685770988464, + "learning_rate": 3.504744551791865e-05, + "loss": 3.1181, + "step": 13989 + }, + { + "epoch": 0.13850063042464017, + "grad_norm": 0.15253548324108124, + "learning_rate": 3.504498049895208e-05, + "loss": 3.1243, + "step": 13990 + }, + { + "epoch": 0.13853147689021358, + "grad_norm": 0.14606578648090363, + "learning_rate": 3.504251536352e-05, + "loss": 3.1281, + "step": 13991 + }, + { + "epoch": 0.138562323355787, + "grad_norm": 0.839918851852417, + "learning_rate": 3.5040050111650984e-05, + "loss": 3.1257, + "step": 13992 + }, + { + "epoch": 0.1385931698213604, + "grad_norm": 0.15040381252765656, + "learning_rate": 3.5037584743373625e-05, + "loss": 3.1054, + "step": 13993 + }, + { + "epoch": 0.13862401628693383, + "grad_norm": 0.18834806978702545, + "learning_rate": 3.5035119258716495e-05, + "loss": 3.1159, + "step": 13994 + }, + { + "epoch": 0.13865486275250724, + "grad_norm": 0.17843031883239746, + "learning_rate": 3.50326536577082e-05, + "loss": 3.1374, + "step": 13995 + }, + { + "epoch": 0.13868570921808065, + "grad_norm": 0.2813800871372223, + "learning_rate": 3.5030187940377306e-05, + "loss": 3.1428, + "step": 13996 + }, + { + "epoch": 0.13871655568365407, + "grad_norm": 0.19516292214393616, + "learning_rate": 3.502772210675241e-05, + "loss": 3.1211, + "step": 13997 + }, + { + "epoch": 0.13874740214922748, + "grad_norm": 0.23833423852920532, + "learning_rate": 3.502525615686211e-05, + "loss": 3.1115, + "step": 13998 + }, + { + "epoch": 0.1387782486148009, + "grad_norm": 0.23581665754318237, + "learning_rate": 3.502279009073498e-05, + "loss": 3.1252, + "step": 13999 + }, + { + "epoch": 0.1388090950803743, + "grad_norm": 0.1905311942100525, + "learning_rate": 3.502032390839962e-05, + "loss": 3.123, + "step": 14000 + }, + { + "epoch": 0.13883994154594773, + "grad_norm": 0.18303200602531433, + "learning_rate": 3.501785760988464e-05, + "loss": 3.1373, + "step": 14001 + }, + { + "epoch": 0.13887078801152117, + "grad_norm": 0.2088242471218109, + "learning_rate": 3.5015391195218616e-05, + "loss": 3.1774, + "step": 14002 + }, + { + "epoch": 0.13890163447709458, + "grad_norm": 0.1946306973695755, + "learning_rate": 3.501292466443014e-05, + "loss": 3.1376, + "step": 14003 + }, + { + "epoch": 0.138932480942668, + "grad_norm": 0.17969758808612823, + "learning_rate": 3.5010458017547824e-05, + "loss": 3.1263, + "step": 14004 + }, + { + "epoch": 0.1389633274082414, + "grad_norm": 0.17559440433979034, + "learning_rate": 3.500799125460027e-05, + "loss": 3.1057, + "step": 14005 + }, + { + "epoch": 0.13899417387381482, + "grad_norm": 0.16293178498744965, + "learning_rate": 3.5005524375616064e-05, + "loss": 3.1224, + "step": 14006 + }, + { + "epoch": 0.13902502033938824, + "grad_norm": 0.171814426779747, + "learning_rate": 3.5003057380623816e-05, + "loss": 3.1552, + "step": 14007 + }, + { + "epoch": 0.13905586680496165, + "grad_norm": 0.1701282411813736, + "learning_rate": 3.500059026965213e-05, + "loss": 3.1012, + "step": 14008 + }, + { + "epoch": 0.13908671327053507, + "grad_norm": 0.16369566321372986, + "learning_rate": 3.49981230427296e-05, + "loss": 3.1109, + "step": 14009 + }, + { + "epoch": 0.13911755973610848, + "grad_norm": 0.5691606998443604, + "learning_rate": 3.4995655699884864e-05, + "loss": 3.1286, + "step": 14010 + }, + { + "epoch": 0.1391484062016819, + "grad_norm": 0.1566886603832245, + "learning_rate": 3.499318824114649e-05, + "loss": 3.1271, + "step": 14011 + }, + { + "epoch": 0.1391792526672553, + "grad_norm": 0.15792782604694366, + "learning_rate": 3.4990720666543106e-05, + "loss": 3.1436, + "step": 14012 + }, + { + "epoch": 0.13921009913282872, + "grad_norm": 0.1671229898929596, + "learning_rate": 3.498825297610333e-05, + "loss": 3.1221, + "step": 14013 + }, + { + "epoch": 0.13924094559840217, + "grad_norm": 0.15326659381389618, + "learning_rate": 3.4985785169855746e-05, + "loss": 3.1523, + "step": 14014 + }, + { + "epoch": 0.13927179206397558, + "grad_norm": 0.16604556143283844, + "learning_rate": 3.4983317247828987e-05, + "loss": 3.1564, + "step": 14015 + }, + { + "epoch": 0.139302638529549, + "grad_norm": 0.1694176346063614, + "learning_rate": 3.498084921005167e-05, + "loss": 3.1466, + "step": 14016 + }, + { + "epoch": 0.1393334849951224, + "grad_norm": 0.15708708763122559, + "learning_rate": 3.4978381056552404e-05, + "loss": 3.1448, + "step": 14017 + }, + { + "epoch": 0.13936433146069582, + "grad_norm": 0.15959204733371735, + "learning_rate": 3.497591278735981e-05, + "loss": 3.1454, + "step": 14018 + }, + { + "epoch": 0.13939517792626924, + "grad_norm": 0.15512004494667053, + "learning_rate": 3.4973444402502494e-05, + "loss": 3.1194, + "step": 14019 + }, + { + "epoch": 0.13942602439184265, + "grad_norm": 0.15701501071453094, + "learning_rate": 3.4970975902009085e-05, + "loss": 3.1301, + "step": 14020 + }, + { + "epoch": 0.13945687085741607, + "grad_norm": 0.15687043964862823, + "learning_rate": 3.49685072859082e-05, + "loss": 3.1379, + "step": 14021 + }, + { + "epoch": 0.13948771732298948, + "grad_norm": 0.1478731781244278, + "learning_rate": 3.496603855422848e-05, + "loss": 3.1278, + "step": 14022 + }, + { + "epoch": 0.1395185637885629, + "grad_norm": 0.15953248739242554, + "learning_rate": 3.496356970699851e-05, + "loss": 3.1462, + "step": 14023 + }, + { + "epoch": 0.1395494102541363, + "grad_norm": 0.15292780101299286, + "learning_rate": 3.496110074424695e-05, + "loss": 3.1092, + "step": 14024 + }, + { + "epoch": 0.13958025671970972, + "grad_norm": 0.15353582799434662, + "learning_rate": 3.495863166600241e-05, + "loss": 3.1276, + "step": 14025 + }, + { + "epoch": 0.13961110318528316, + "grad_norm": 0.15880104899406433, + "learning_rate": 3.4956162472293515e-05, + "loss": 3.1297, + "step": 14026 + }, + { + "epoch": 0.13964194965085658, + "grad_norm": 0.1498136967420578, + "learning_rate": 3.4953693163148904e-05, + "loss": 3.1186, + "step": 14027 + }, + { + "epoch": 0.13967279611643, + "grad_norm": 0.16075558960437775, + "learning_rate": 3.495122373859721e-05, + "loss": 3.1302, + "step": 14028 + }, + { + "epoch": 0.1397036425820034, + "grad_norm": 0.15063884854316711, + "learning_rate": 3.494875419866704e-05, + "loss": 3.1431, + "step": 14029 + }, + { + "epoch": 0.13973448904757682, + "grad_norm": 0.1545974314212799, + "learning_rate": 3.4946284543387056e-05, + "loss": 3.1487, + "step": 14030 + }, + { + "epoch": 0.13976533551315024, + "grad_norm": 0.15961630642414093, + "learning_rate": 3.494381477278587e-05, + "loss": 3.1339, + "step": 14031 + }, + { + "epoch": 0.13979618197872365, + "grad_norm": 0.15343891084194183, + "learning_rate": 3.494134488689213e-05, + "loss": 3.1154, + "step": 14032 + }, + { + "epoch": 0.13982702844429706, + "grad_norm": 0.15104876458644867, + "learning_rate": 3.493887488573447e-05, + "loss": 3.1509, + "step": 14033 + }, + { + "epoch": 0.13985787490987048, + "grad_norm": 0.15632738173007965, + "learning_rate": 3.493640476934154e-05, + "loss": 3.1579, + "step": 14034 + }, + { + "epoch": 0.1398887213754439, + "grad_norm": 0.15467506647109985, + "learning_rate": 3.493393453774195e-05, + "loss": 3.141, + "step": 14035 + }, + { + "epoch": 0.1399195678410173, + "grad_norm": 0.14794406294822693, + "learning_rate": 3.493146419096437e-05, + "loss": 3.1264, + "step": 14036 + }, + { + "epoch": 0.13995041430659072, + "grad_norm": 0.14981332421302795, + "learning_rate": 3.492899372903743e-05, + "loss": 3.1385, + "step": 14037 + }, + { + "epoch": 0.13998126077216416, + "grad_norm": 0.15835727751255035, + "learning_rate": 3.4926523151989764e-05, + "loss": 3.1517, + "step": 14038 + }, + { + "epoch": 0.14001210723773758, + "grad_norm": 0.15425704419612885, + "learning_rate": 3.492405245985004e-05, + "loss": 3.1205, + "step": 14039 + }, + { + "epoch": 0.140042953703311, + "grad_norm": 0.15243931114673615, + "learning_rate": 3.492158165264689e-05, + "loss": 3.1196, + "step": 14040 + }, + { + "epoch": 0.1400738001688844, + "grad_norm": 0.169507697224617, + "learning_rate": 3.491911073040896e-05, + "loss": 3.1249, + "step": 14041 + }, + { + "epoch": 0.14010464663445782, + "grad_norm": 0.20157364010810852, + "learning_rate": 3.4916639693164896e-05, + "loss": 3.1501, + "step": 14042 + }, + { + "epoch": 0.14013549310003123, + "grad_norm": 0.18346120417118073, + "learning_rate": 3.4914168540943366e-05, + "loss": 3.0968, + "step": 14043 + }, + { + "epoch": 0.14016633956560465, + "grad_norm": 0.15407729148864746, + "learning_rate": 3.4911697273773e-05, + "loss": 3.0732, + "step": 14044 + }, + { + "epoch": 0.14019718603117806, + "grad_norm": 0.1551281362771988, + "learning_rate": 3.490922589168246e-05, + "loss": 3.1377, + "step": 14045 + }, + { + "epoch": 0.14022803249675148, + "grad_norm": 0.16066627204418182, + "learning_rate": 3.4906754394700416e-05, + "loss": 3.1344, + "step": 14046 + }, + { + "epoch": 0.1402588789623249, + "grad_norm": 0.15041185915470123, + "learning_rate": 3.490428278285549e-05, + "loss": 3.1112, + "step": 14047 + }, + { + "epoch": 0.1402897254278983, + "grad_norm": 0.1651945412158966, + "learning_rate": 3.490181105617636e-05, + "loss": 3.1569, + "step": 14048 + }, + { + "epoch": 0.14032057189347172, + "grad_norm": 0.15437428653240204, + "learning_rate": 3.489933921469168e-05, + "loss": 3.1242, + "step": 14049 + }, + { + "epoch": 0.14035141835904516, + "grad_norm": 0.1488858312368393, + "learning_rate": 3.489686725843012e-05, + "loss": 3.0911, + "step": 14050 + }, + { + "epoch": 0.14038226482461857, + "grad_norm": 0.17051450908184052, + "learning_rate": 3.4894395187420326e-05, + "loss": 3.0924, + "step": 14051 + }, + { + "epoch": 0.140413111290192, + "grad_norm": 0.15379121899604797, + "learning_rate": 3.489192300169097e-05, + "loss": 3.0971, + "step": 14052 + }, + { + "epoch": 0.1404439577557654, + "grad_norm": 0.1524086892604828, + "learning_rate": 3.48894507012707e-05, + "loss": 3.1521, + "step": 14053 + }, + { + "epoch": 0.14047480422133882, + "grad_norm": 0.15569420158863068, + "learning_rate": 3.48869782861882e-05, + "loss": 3.1638, + "step": 14054 + }, + { + "epoch": 0.14050565068691223, + "grad_norm": 0.18807154893875122, + "learning_rate": 3.488450575647213e-05, + "loss": 3.1169, + "step": 14055 + }, + { + "epoch": 0.14053649715248565, + "grad_norm": 0.15402153134346008, + "learning_rate": 3.488203311215115e-05, + "loss": 3.1162, + "step": 14056 + }, + { + "epoch": 0.14056734361805906, + "grad_norm": 0.14723345637321472, + "learning_rate": 3.487956035325394e-05, + "loss": 3.1269, + "step": 14057 + }, + { + "epoch": 0.14059819008363247, + "grad_norm": 0.17337074875831604, + "learning_rate": 3.4877087479809155e-05, + "loss": 3.1089, + "step": 14058 + }, + { + "epoch": 0.1406290365492059, + "grad_norm": 0.14967429637908936, + "learning_rate": 3.487461449184549e-05, + "loss": 3.1129, + "step": 14059 + }, + { + "epoch": 0.1406598830147793, + "grad_norm": 0.151399627327919, + "learning_rate": 3.4872141389391596e-05, + "loss": 3.1131, + "step": 14060 + }, + { + "epoch": 0.14069072948035272, + "grad_norm": 0.15706922113895416, + "learning_rate": 3.486966817247615e-05, + "loss": 3.1348, + "step": 14061 + }, + { + "epoch": 0.14072157594592616, + "grad_norm": 0.15518932044506073, + "learning_rate": 3.486719484112784e-05, + "loss": 3.1264, + "step": 14062 + }, + { + "epoch": 0.14075242241149957, + "grad_norm": 0.15431618690490723, + "learning_rate": 3.486472139537534e-05, + "loss": 3.1758, + "step": 14063 + }, + { + "epoch": 0.140783268877073, + "grad_norm": 0.1576337367296219, + "learning_rate": 3.486224783524731e-05, + "loss": 3.1383, + "step": 14064 + }, + { + "epoch": 0.1408141153426464, + "grad_norm": 0.1537024825811386, + "learning_rate": 3.4859774160772454e-05, + "loss": 3.0656, + "step": 14065 + }, + { + "epoch": 0.14084496180821982, + "grad_norm": 0.16098006069660187, + "learning_rate": 3.485730037197944e-05, + "loss": 3.1143, + "step": 14066 + }, + { + "epoch": 0.14087580827379323, + "grad_norm": 0.16429436206817627, + "learning_rate": 3.485482646889694e-05, + "loss": 3.1227, + "step": 14067 + }, + { + "epoch": 0.14090665473936664, + "grad_norm": 0.1655241847038269, + "learning_rate": 3.4852352451553665e-05, + "loss": 3.1229, + "step": 14068 + }, + { + "epoch": 0.14093750120494006, + "grad_norm": 0.15101425349712372, + "learning_rate": 3.4849878319978276e-05, + "loss": 3.176, + "step": 14069 + }, + { + "epoch": 0.14096834767051347, + "grad_norm": 0.18704964220523834, + "learning_rate": 3.484740407419946e-05, + "loss": 3.1594, + "step": 14070 + }, + { + "epoch": 0.14099919413608689, + "grad_norm": 0.16454675793647766, + "learning_rate": 3.4844929714245925e-05, + "loss": 3.1386, + "step": 14071 + }, + { + "epoch": 0.1410300406016603, + "grad_norm": 0.15248017013072968, + "learning_rate": 3.4842455240146344e-05, + "loss": 3.1287, + "step": 14072 + }, + { + "epoch": 0.14106088706723371, + "grad_norm": 0.14644744992256165, + "learning_rate": 3.483998065192941e-05, + "loss": 3.1093, + "step": 14073 + }, + { + "epoch": 0.14109173353280716, + "grad_norm": 0.15674196183681488, + "learning_rate": 3.48375059496238e-05, + "loss": 3.1461, + "step": 14074 + }, + { + "epoch": 0.14112257999838057, + "grad_norm": 0.15600784122943878, + "learning_rate": 3.483503113325823e-05, + "loss": 3.1308, + "step": 14075 + }, + { + "epoch": 0.14115342646395398, + "grad_norm": 0.14992940425872803, + "learning_rate": 3.483255620286139e-05, + "loss": 3.1168, + "step": 14076 + }, + { + "epoch": 0.1411842729295274, + "grad_norm": 0.15350382030010223, + "learning_rate": 3.483008115846197e-05, + "loss": 3.1585, + "step": 14077 + }, + { + "epoch": 0.1412151193951008, + "grad_norm": 0.1522352397441864, + "learning_rate": 3.4827606000088655e-05, + "loss": 3.1111, + "step": 14078 + }, + { + "epoch": 0.14124596586067423, + "grad_norm": 0.15892112255096436, + "learning_rate": 3.4825130727770154e-05, + "loss": 3.1352, + "step": 14079 + }, + { + "epoch": 0.14127681232624764, + "grad_norm": 0.16775377094745636, + "learning_rate": 3.4822655341535175e-05, + "loss": 3.1272, + "step": 14080 + }, + { + "epoch": 0.14130765879182106, + "grad_norm": 0.1556280106306076, + "learning_rate": 3.48201798414124e-05, + "loss": 3.0998, + "step": 14081 + }, + { + "epoch": 0.14133850525739447, + "grad_norm": 0.15333491563796997, + "learning_rate": 3.481770422743055e-05, + "loss": 3.1095, + "step": 14082 + }, + { + "epoch": 0.14136935172296788, + "grad_norm": 0.15314793586730957, + "learning_rate": 3.481522849961832e-05, + "loss": 3.155, + "step": 14083 + }, + { + "epoch": 0.1414001981885413, + "grad_norm": 0.14759159088134766, + "learning_rate": 3.481275265800441e-05, + "loss": 3.1318, + "step": 14084 + }, + { + "epoch": 0.1414310446541147, + "grad_norm": 0.15068702399730682, + "learning_rate": 3.4810276702617525e-05, + "loss": 3.1083, + "step": 14085 + }, + { + "epoch": 0.14146189111968815, + "grad_norm": 0.1479915976524353, + "learning_rate": 3.480780063348638e-05, + "loss": 3.1204, + "step": 14086 + }, + { + "epoch": 0.14149273758526157, + "grad_norm": 0.14973516762256622, + "learning_rate": 3.4805324450639684e-05, + "loss": 3.1508, + "step": 14087 + }, + { + "epoch": 0.14152358405083498, + "grad_norm": 0.15149393677711487, + "learning_rate": 3.480284815410614e-05, + "loss": 3.1698, + "step": 14088 + }, + { + "epoch": 0.1415544305164084, + "grad_norm": 0.18189890682697296, + "learning_rate": 3.4800371743914465e-05, + "loss": 3.1343, + "step": 14089 + }, + { + "epoch": 0.1415852769819818, + "grad_norm": 0.15561531484127045, + "learning_rate": 3.4797895220093374e-05, + "loss": 3.1292, + "step": 14090 + }, + { + "epoch": 0.14161612344755523, + "grad_norm": 0.26796191930770874, + "learning_rate": 3.479541858267157e-05, + "loss": 3.1084, + "step": 14091 + }, + { + "epoch": 0.14164696991312864, + "grad_norm": 0.1527218520641327, + "learning_rate": 3.4792941831677774e-05, + "loss": 3.1499, + "step": 14092 + }, + { + "epoch": 0.14167781637870205, + "grad_norm": 0.15719251334667206, + "learning_rate": 3.47904649671407e-05, + "loss": 3.1146, + "step": 14093 + }, + { + "epoch": 0.14170866284427547, + "grad_norm": 0.1472730189561844, + "learning_rate": 3.4787987989089064e-05, + "loss": 3.0966, + "step": 14094 + }, + { + "epoch": 0.14173950930984888, + "grad_norm": 0.1515449732542038, + "learning_rate": 3.47855108975516e-05, + "loss": 3.1203, + "step": 14095 + }, + { + "epoch": 0.1417703557754223, + "grad_norm": 0.18099303543567657, + "learning_rate": 3.4783033692557013e-05, + "loss": 3.156, + "step": 14096 + }, + { + "epoch": 0.1418012022409957, + "grad_norm": 0.15529921650886536, + "learning_rate": 3.478055637413403e-05, + "loss": 3.145, + "step": 14097 + }, + { + "epoch": 0.14183204870656915, + "grad_norm": 0.14523352682590485, + "learning_rate": 3.477807894231137e-05, + "loss": 3.1197, + "step": 14098 + }, + { + "epoch": 0.14186289517214257, + "grad_norm": 0.15478260815143585, + "learning_rate": 3.4775601397117766e-05, + "loss": 3.1147, + "step": 14099 + }, + { + "epoch": 0.14189374163771598, + "grad_norm": 0.15786808729171753, + "learning_rate": 3.477312373858193e-05, + "loss": 3.1435, + "step": 14100 + }, + { + "epoch": 0.1419245881032894, + "grad_norm": 0.16994446516036987, + "learning_rate": 3.4770645966732606e-05, + "loss": 3.1182, + "step": 14101 + }, + { + "epoch": 0.1419554345688628, + "grad_norm": 0.14735794067382812, + "learning_rate": 3.476816808159851e-05, + "loss": 3.1379, + "step": 14102 + }, + { + "epoch": 0.14198628103443622, + "grad_norm": 0.15752331912517548, + "learning_rate": 3.476569008320837e-05, + "loss": 3.0777, + "step": 14103 + }, + { + "epoch": 0.14201712750000964, + "grad_norm": 0.15444384515285492, + "learning_rate": 3.4763211971590934e-05, + "loss": 3.0597, + "step": 14104 + }, + { + "epoch": 0.14204797396558305, + "grad_norm": 0.14824950695037842, + "learning_rate": 3.476073374677491e-05, + "loss": 3.1729, + "step": 14105 + }, + { + "epoch": 0.14207882043115647, + "grad_norm": 0.14428365230560303, + "learning_rate": 3.475825540878905e-05, + "loss": 3.1039, + "step": 14106 + }, + { + "epoch": 0.14210966689672988, + "grad_norm": 0.15921618044376373, + "learning_rate": 3.475577695766208e-05, + "loss": 3.1728, + "step": 14107 + }, + { + "epoch": 0.1421405133623033, + "grad_norm": 0.14821329712867737, + "learning_rate": 3.475329839342274e-05, + "loss": 3.1572, + "step": 14108 + }, + { + "epoch": 0.1421713598278767, + "grad_norm": 0.14663252234458923, + "learning_rate": 3.475081971609977e-05, + "loss": 3.106, + "step": 14109 + }, + { + "epoch": 0.14220220629345015, + "grad_norm": 0.1465284824371338, + "learning_rate": 3.47483409257219e-05, + "loss": 3.1203, + "step": 14110 + }, + { + "epoch": 0.14223305275902356, + "grad_norm": 0.1736268252134323, + "learning_rate": 3.474586202231788e-05, + "loss": 3.1383, + "step": 14111 + }, + { + "epoch": 0.14226389922459698, + "grad_norm": 0.15074659883975983, + "learning_rate": 3.474338300591644e-05, + "loss": 3.1261, + "step": 14112 + }, + { + "epoch": 0.1422947456901704, + "grad_norm": 0.1508418768644333, + "learning_rate": 3.474090387654633e-05, + "loss": 3.1421, + "step": 14113 + }, + { + "epoch": 0.1423255921557438, + "grad_norm": 0.15088587999343872, + "learning_rate": 3.47384246342363e-05, + "loss": 3.1456, + "step": 14114 + }, + { + "epoch": 0.14235643862131722, + "grad_norm": 0.1470162570476532, + "learning_rate": 3.473594527901508e-05, + "loss": 3.1107, + "step": 14115 + }, + { + "epoch": 0.14238728508689064, + "grad_norm": 0.15252317488193512, + "learning_rate": 3.473346581091143e-05, + "loss": 3.1206, + "step": 14116 + }, + { + "epoch": 0.14241813155246405, + "grad_norm": 0.15214990079402924, + "learning_rate": 3.473098622995409e-05, + "loss": 3.1388, + "step": 14117 + }, + { + "epoch": 0.14244897801803746, + "grad_norm": 0.15344969928264618, + "learning_rate": 3.472850653617181e-05, + "loss": 3.1287, + "step": 14118 + }, + { + "epoch": 0.14247982448361088, + "grad_norm": 0.1493418961763382, + "learning_rate": 3.4726026729593344e-05, + "loss": 3.1286, + "step": 14119 + }, + { + "epoch": 0.1425106709491843, + "grad_norm": 0.15598151087760925, + "learning_rate": 3.472354681024745e-05, + "loss": 3.1469, + "step": 14120 + }, + { + "epoch": 0.1425415174147577, + "grad_norm": 0.1500251293182373, + "learning_rate": 3.472106677816287e-05, + "loss": 3.137, + "step": 14121 + }, + { + "epoch": 0.14257236388033115, + "grad_norm": 0.1447773575782776, + "learning_rate": 3.471858663336837e-05, + "loss": 3.1148, + "step": 14122 + }, + { + "epoch": 0.14260321034590456, + "grad_norm": 0.1502034068107605, + "learning_rate": 3.4716106375892684e-05, + "loss": 3.1365, + "step": 14123 + }, + { + "epoch": 0.14263405681147798, + "grad_norm": 0.15329918265342712, + "learning_rate": 3.4713626005764585e-05, + "loss": 3.1375, + "step": 14124 + }, + { + "epoch": 0.1426649032770514, + "grad_norm": 0.157113716006279, + "learning_rate": 3.4711145523012836e-05, + "loss": 3.1025, + "step": 14125 + }, + { + "epoch": 0.1426957497426248, + "grad_norm": 0.15949946641921997, + "learning_rate": 3.470866492766619e-05, + "loss": 3.12, + "step": 14126 + }, + { + "epoch": 0.14272659620819822, + "grad_norm": 0.15257200598716736, + "learning_rate": 3.470618421975341e-05, + "loss": 3.1173, + "step": 14127 + }, + { + "epoch": 0.14275744267377163, + "grad_norm": 0.15145301818847656, + "learning_rate": 3.470370339930326e-05, + "loss": 3.115, + "step": 14128 + }, + { + "epoch": 0.14278828913934505, + "grad_norm": 0.15392978489398956, + "learning_rate": 3.4701222466344485e-05, + "loss": 3.1465, + "step": 14129 + }, + { + "epoch": 0.14281913560491846, + "grad_norm": 0.1582440882921219, + "learning_rate": 3.4698741420905877e-05, + "loss": 3.1648, + "step": 14130 + }, + { + "epoch": 0.14284998207049188, + "grad_norm": 0.1483912318944931, + "learning_rate": 3.4696260263016184e-05, + "loss": 3.1235, + "step": 14131 + }, + { + "epoch": 0.1428808285360653, + "grad_norm": 0.1479939967393875, + "learning_rate": 3.469377899270418e-05, + "loss": 3.1122, + "step": 14132 + }, + { + "epoch": 0.1429116750016387, + "grad_norm": 0.14610259234905243, + "learning_rate": 3.469129760999864e-05, + "loss": 3.1074, + "step": 14133 + }, + { + "epoch": 0.14294252146721215, + "grad_norm": 0.14953544735908508, + "learning_rate": 3.4688816114928327e-05, + "loss": 3.1377, + "step": 14134 + }, + { + "epoch": 0.14297336793278556, + "grad_norm": 0.17382732033729553, + "learning_rate": 3.468633450752201e-05, + "loss": 3.1264, + "step": 14135 + }, + { + "epoch": 0.14300421439835898, + "grad_norm": 0.14929409325122833, + "learning_rate": 3.468385278780847e-05, + "loss": 3.145, + "step": 14136 + }, + { + "epoch": 0.1430350608639324, + "grad_norm": 0.14714454114437103, + "learning_rate": 3.4681370955816475e-05, + "loss": 3.1369, + "step": 14137 + }, + { + "epoch": 0.1430659073295058, + "grad_norm": 0.14886514842510223, + "learning_rate": 3.4678889011574794e-05, + "loss": 3.1264, + "step": 14138 + }, + { + "epoch": 0.14309675379507922, + "grad_norm": 0.14333312213420868, + "learning_rate": 3.467640695511222e-05, + "loss": 3.0946, + "step": 14139 + }, + { + "epoch": 0.14312760026065263, + "grad_norm": 0.1496620923280716, + "learning_rate": 3.467392478645752e-05, + "loss": 3.0979, + "step": 14140 + }, + { + "epoch": 0.14315844672622605, + "grad_norm": 0.15080420672893524, + "learning_rate": 3.467144250563947e-05, + "loss": 3.1132, + "step": 14141 + }, + { + "epoch": 0.14318929319179946, + "grad_norm": 0.15110339224338531, + "learning_rate": 3.4668960112686865e-05, + "loss": 3.1518, + "step": 14142 + }, + { + "epoch": 0.14322013965737287, + "grad_norm": 0.14480237662792206, + "learning_rate": 3.466647760762848e-05, + "loss": 3.1603, + "step": 14143 + }, + { + "epoch": 0.1432509861229463, + "grad_norm": 0.16257579624652863, + "learning_rate": 3.4663994990493084e-05, + "loss": 3.1396, + "step": 14144 + }, + { + "epoch": 0.1432818325885197, + "grad_norm": 0.1514471471309662, + "learning_rate": 3.466151226130948e-05, + "loss": 3.1197, + "step": 14145 + }, + { + "epoch": 0.14331267905409315, + "grad_norm": 0.1532699167728424, + "learning_rate": 3.465902942010645e-05, + "loss": 3.1134, + "step": 14146 + }, + { + "epoch": 0.14334352551966656, + "grad_norm": 0.15799580514431, + "learning_rate": 3.465654646691277e-05, + "loss": 3.1517, + "step": 14147 + }, + { + "epoch": 0.14337437198523997, + "grad_norm": 0.1526700258255005, + "learning_rate": 3.4654063401757244e-05, + "loss": 3.133, + "step": 14148 + }, + { + "epoch": 0.1434052184508134, + "grad_norm": 0.15201297402381897, + "learning_rate": 3.465158022466865e-05, + "loss": 3.1204, + "step": 14149 + }, + { + "epoch": 0.1434360649163868, + "grad_norm": 0.1483450084924698, + "learning_rate": 3.464909693567578e-05, + "loss": 3.1022, + "step": 14150 + }, + { + "epoch": 0.14346691138196022, + "grad_norm": 0.14819331467151642, + "learning_rate": 3.464661353480744e-05, + "loss": 3.126, + "step": 14151 + }, + { + "epoch": 0.14349775784753363, + "grad_norm": 0.1452333778142929, + "learning_rate": 3.46441300220924e-05, + "loss": 3.1089, + "step": 14152 + }, + { + "epoch": 0.14352860431310704, + "grad_norm": 0.15765629708766937, + "learning_rate": 3.464164639755948e-05, + "loss": 3.1442, + "step": 14153 + }, + { + "epoch": 0.14355945077868046, + "grad_norm": 0.14838647842407227, + "learning_rate": 3.463916266123746e-05, + "loss": 3.1411, + "step": 14154 + }, + { + "epoch": 0.14359029724425387, + "grad_norm": 0.15053771436214447, + "learning_rate": 3.4636678813155146e-05, + "loss": 3.1081, + "step": 14155 + }, + { + "epoch": 0.1436211437098273, + "grad_norm": 0.15032726526260376, + "learning_rate": 3.4634194853341315e-05, + "loss": 3.1238, + "step": 14156 + }, + { + "epoch": 0.1436519901754007, + "grad_norm": 0.15890099108219147, + "learning_rate": 3.4631710781824803e-05, + "loss": 3.1195, + "step": 14157 + }, + { + "epoch": 0.14368283664097414, + "grad_norm": 0.15303082764148712, + "learning_rate": 3.4629226598634384e-05, + "loss": 3.1246, + "step": 14158 + }, + { + "epoch": 0.14371368310654756, + "grad_norm": 0.1561734676361084, + "learning_rate": 3.462674230379886e-05, + "loss": 3.1519, + "step": 14159 + }, + { + "epoch": 0.14374452957212097, + "grad_norm": 0.1484808325767517, + "learning_rate": 3.4624257897347066e-05, + "loss": 3.1302, + "step": 14160 + }, + { + "epoch": 0.14377537603769439, + "grad_norm": 0.15465278923511505, + "learning_rate": 3.462177337930777e-05, + "loss": 3.1123, + "step": 14161 + }, + { + "epoch": 0.1438062225032678, + "grad_norm": 0.17695528268814087, + "learning_rate": 3.46192887497098e-05, + "loss": 3.0797, + "step": 14162 + }, + { + "epoch": 0.14383706896884121, + "grad_norm": 0.15544649958610535, + "learning_rate": 3.461680400858196e-05, + "loss": 3.135, + "step": 14163 + }, + { + "epoch": 0.14386791543441463, + "grad_norm": 0.1518397331237793, + "learning_rate": 3.461431915595305e-05, + "loss": 3.1001, + "step": 14164 + }, + { + "epoch": 0.14389876189998804, + "grad_norm": 0.154733344912529, + "learning_rate": 3.4611834191851884e-05, + "loss": 3.1409, + "step": 14165 + }, + { + "epoch": 0.14392960836556146, + "grad_norm": 0.1563221514225006, + "learning_rate": 3.460934911630728e-05, + "loss": 3.1106, + "step": 14166 + }, + { + "epoch": 0.14396045483113487, + "grad_norm": 0.15032391250133514, + "learning_rate": 3.4606863929348044e-05, + "loss": 3.1036, + "step": 14167 + }, + { + "epoch": 0.14399130129670829, + "grad_norm": 0.1515321284532547, + "learning_rate": 3.4604378631003005e-05, + "loss": 3.151, + "step": 14168 + }, + { + "epoch": 0.1440221477622817, + "grad_norm": 0.15255969762802124, + "learning_rate": 3.4601893221300956e-05, + "loss": 3.1536, + "step": 14169 + }, + { + "epoch": 0.14405299422785514, + "grad_norm": 0.15475189685821533, + "learning_rate": 3.4599407700270725e-05, + "loss": 3.1212, + "step": 14170 + }, + { + "epoch": 0.14408384069342856, + "grad_norm": 0.15427368879318237, + "learning_rate": 3.459692206794114e-05, + "loss": 3.0951, + "step": 14171 + }, + { + "epoch": 0.14411468715900197, + "grad_norm": 0.15081238746643066, + "learning_rate": 3.4594436324341005e-05, + "loss": 3.1446, + "step": 14172 + }, + { + "epoch": 0.14414553362457538, + "grad_norm": 0.1518552601337433, + "learning_rate": 3.459195046949915e-05, + "loss": 3.1406, + "step": 14173 + }, + { + "epoch": 0.1441763800901488, + "grad_norm": 0.1524316370487213, + "learning_rate": 3.458946450344439e-05, + "loss": 3.1485, + "step": 14174 + }, + { + "epoch": 0.1442072265557222, + "grad_norm": 0.14598305523395538, + "learning_rate": 3.458697842620555e-05, + "loss": 3.1216, + "step": 14175 + }, + { + "epoch": 0.14423807302129563, + "grad_norm": 0.15567167103290558, + "learning_rate": 3.458449223781146e-05, + "loss": 3.1008, + "step": 14176 + }, + { + "epoch": 0.14426891948686904, + "grad_norm": 0.14990682899951935, + "learning_rate": 3.458200593829094e-05, + "loss": 3.1139, + "step": 14177 + }, + { + "epoch": 0.14429976595244245, + "grad_norm": 0.14688071608543396, + "learning_rate": 3.4579519527672824e-05, + "loss": 3.106, + "step": 14178 + }, + { + "epoch": 0.14433061241801587, + "grad_norm": 0.15230955183506012, + "learning_rate": 3.457703300598593e-05, + "loss": 3.1197, + "step": 14179 + }, + { + "epoch": 0.14436145888358928, + "grad_norm": 0.15571489930152893, + "learning_rate": 3.457454637325909e-05, + "loss": 3.1164, + "step": 14180 + }, + { + "epoch": 0.14439230534916273, + "grad_norm": 0.15507708489894867, + "learning_rate": 3.457205962952115e-05, + "loss": 3.1331, + "step": 14181 + }, + { + "epoch": 0.14442315181473614, + "grad_norm": 0.16084066033363342, + "learning_rate": 3.456957277480092e-05, + "loss": 3.1068, + "step": 14182 + }, + { + "epoch": 0.14445399828030955, + "grad_norm": 0.15723851323127747, + "learning_rate": 3.456708580912725e-05, + "loss": 3.1046, + "step": 14183 + }, + { + "epoch": 0.14448484474588297, + "grad_norm": 0.15175041556358337, + "learning_rate": 3.4564598732528966e-05, + "loss": 3.1323, + "step": 14184 + }, + { + "epoch": 0.14451569121145638, + "grad_norm": 0.15720100700855255, + "learning_rate": 3.4562111545034906e-05, + "loss": 3.158, + "step": 14185 + }, + { + "epoch": 0.1445465376770298, + "grad_norm": 0.16118969023227692, + "learning_rate": 3.455962424667391e-05, + "loss": 3.1385, + "step": 14186 + }, + { + "epoch": 0.1445773841426032, + "grad_norm": 0.14818230271339417, + "learning_rate": 3.4557136837474824e-05, + "loss": 3.1314, + "step": 14187 + }, + { + "epoch": 0.14460823060817662, + "grad_norm": 0.15348631143569946, + "learning_rate": 3.455464931746647e-05, + "loss": 3.1496, + "step": 14188 + }, + { + "epoch": 0.14463907707375004, + "grad_norm": 0.1567564159631729, + "learning_rate": 3.4552161686677695e-05, + "loss": 3.1368, + "step": 14189 + }, + { + "epoch": 0.14466992353932345, + "grad_norm": 0.15566599369049072, + "learning_rate": 3.454967394513736e-05, + "loss": 3.1527, + "step": 14190 + }, + { + "epoch": 0.14470077000489687, + "grad_norm": 0.15218912065029144, + "learning_rate": 3.4547186092874285e-05, + "loss": 3.0957, + "step": 14191 + }, + { + "epoch": 0.14473161647047028, + "grad_norm": 0.14961963891983032, + "learning_rate": 3.4544698129917327e-05, + "loss": 3.1251, + "step": 14192 + }, + { + "epoch": 0.14476246293604372, + "grad_norm": 0.15283580124378204, + "learning_rate": 3.4542210056295324e-05, + "loss": 3.1125, + "step": 14193 + }, + { + "epoch": 0.14479330940161714, + "grad_norm": 0.15510593354701996, + "learning_rate": 3.453972187203713e-05, + "loss": 3.1081, + "step": 14194 + }, + { + "epoch": 0.14482415586719055, + "grad_norm": 0.14604425430297852, + "learning_rate": 3.453723357717161e-05, + "loss": 3.1355, + "step": 14195 + }, + { + "epoch": 0.14485500233276397, + "grad_norm": 0.1478433907032013, + "learning_rate": 3.4534745171727585e-05, + "loss": 3.1152, + "step": 14196 + }, + { + "epoch": 0.14488584879833738, + "grad_norm": 0.16277265548706055, + "learning_rate": 3.453225665573392e-05, + "loss": 3.1264, + "step": 14197 + }, + { + "epoch": 0.1449166952639108, + "grad_norm": 0.15137867629528046, + "learning_rate": 3.452976802921946e-05, + "loss": 3.1351, + "step": 14198 + }, + { + "epoch": 0.1449475417294842, + "grad_norm": 0.15462812781333923, + "learning_rate": 3.452727929221308e-05, + "loss": 3.1155, + "step": 14199 + }, + { + "epoch": 0.14497838819505762, + "grad_norm": 0.15599733591079712, + "learning_rate": 3.452479044474361e-05, + "loss": 3.1439, + "step": 14200 + }, + { + "epoch": 0.14500923466063104, + "grad_norm": 0.1540561318397522, + "learning_rate": 3.452230148683993e-05, + "loss": 3.136, + "step": 14201 + }, + { + "epoch": 0.14504008112620445, + "grad_norm": 0.15239743888378143, + "learning_rate": 3.451981241853087e-05, + "loss": 3.1451, + "step": 14202 + }, + { + "epoch": 0.14507092759177787, + "grad_norm": 0.1577468067407608, + "learning_rate": 3.4517323239845326e-05, + "loss": 3.0837, + "step": 14203 + }, + { + "epoch": 0.14510177405735128, + "grad_norm": 0.1522655040025711, + "learning_rate": 3.451483395081212e-05, + "loss": 3.139, + "step": 14204 + }, + { + "epoch": 0.14513262052292472, + "grad_norm": 0.15030606091022491, + "learning_rate": 3.4512344551460155e-05, + "loss": 3.1318, + "step": 14205 + }, + { + "epoch": 0.14516346698849814, + "grad_norm": 0.157120481133461, + "learning_rate": 3.450985504181825e-05, + "loss": 3.1503, + "step": 14206 + }, + { + "epoch": 0.14519431345407155, + "grad_norm": 0.15401232242584229, + "learning_rate": 3.45073654219153e-05, + "loss": 3.1772, + "step": 14207 + }, + { + "epoch": 0.14522515991964496, + "grad_norm": 0.1522357314825058, + "learning_rate": 3.450487569178015e-05, + "loss": 3.1357, + "step": 14208 + }, + { + "epoch": 0.14525600638521838, + "grad_norm": 0.15283837914466858, + "learning_rate": 3.450238585144169e-05, + "loss": 3.1184, + "step": 14209 + }, + { + "epoch": 0.1452868528507918, + "grad_norm": 0.15219831466674805, + "learning_rate": 3.449989590092878e-05, + "loss": 3.1318, + "step": 14210 + }, + { + "epoch": 0.1453176993163652, + "grad_norm": 0.15455934405326843, + "learning_rate": 3.449740584027027e-05, + "loss": 3.1409, + "step": 14211 + }, + { + "epoch": 0.14534854578193862, + "grad_norm": 0.15280769765377045, + "learning_rate": 3.449491566949506e-05, + "loss": 3.1403, + "step": 14212 + }, + { + "epoch": 0.14537939224751203, + "grad_norm": 0.1549694985151291, + "learning_rate": 3.4492425388632e-05, + "loss": 3.1291, + "step": 14213 + }, + { + "epoch": 0.14541023871308545, + "grad_norm": 0.15125268697738647, + "learning_rate": 3.448993499770998e-05, + "loss": 3.1486, + "step": 14214 + }, + { + "epoch": 0.14544108517865886, + "grad_norm": 0.15287275612354279, + "learning_rate": 3.448744449675786e-05, + "loss": 3.1441, + "step": 14215 + }, + { + "epoch": 0.14547193164423228, + "grad_norm": 0.15595737099647522, + "learning_rate": 3.448495388580453e-05, + "loss": 3.1395, + "step": 14216 + }, + { + "epoch": 0.14550277810980572, + "grad_norm": 0.15099172294139862, + "learning_rate": 3.4482463164878845e-05, + "loss": 3.1208, + "step": 14217 + }, + { + "epoch": 0.14553362457537913, + "grad_norm": 0.15118570625782013, + "learning_rate": 3.4479972334009715e-05, + "loss": 3.1129, + "step": 14218 + }, + { + "epoch": 0.14556447104095255, + "grad_norm": 0.15098904073238373, + "learning_rate": 3.447748139322599e-05, + "loss": 3.0989, + "step": 14219 + }, + { + "epoch": 0.14559531750652596, + "grad_norm": 0.15700720250606537, + "learning_rate": 3.447499034255657e-05, + "loss": 3.1319, + "step": 14220 + }, + { + "epoch": 0.14562616397209938, + "grad_norm": 0.15167316794395447, + "learning_rate": 3.4472499182030324e-05, + "loss": 3.1389, + "step": 14221 + }, + { + "epoch": 0.1456570104376728, + "grad_norm": 0.15754778683185577, + "learning_rate": 3.447000791167616e-05, + "loss": 3.1354, + "step": 14222 + }, + { + "epoch": 0.1456878569032462, + "grad_norm": 0.15081410109996796, + "learning_rate": 3.4467516531522924e-05, + "loss": 3.1233, + "step": 14223 + }, + { + "epoch": 0.14571870336881962, + "grad_norm": 0.15368464589118958, + "learning_rate": 3.4465025041599534e-05, + "loss": 3.113, + "step": 14224 + }, + { + "epoch": 0.14574954983439303, + "grad_norm": 0.1587638407945633, + "learning_rate": 3.446253344193486e-05, + "loss": 3.1218, + "step": 14225 + }, + { + "epoch": 0.14578039629996645, + "grad_norm": 0.15035219490528107, + "learning_rate": 3.4460041732557804e-05, + "loss": 3.1154, + "step": 14226 + }, + { + "epoch": 0.14581124276553986, + "grad_norm": 0.15445931255817413, + "learning_rate": 3.4457549913497245e-05, + "loss": 3.1191, + "step": 14227 + }, + { + "epoch": 0.14584208923111328, + "grad_norm": 0.15241765975952148, + "learning_rate": 3.4455057984782084e-05, + "loss": 3.1271, + "step": 14228 + }, + { + "epoch": 0.14587293569668672, + "grad_norm": 0.1524672508239746, + "learning_rate": 3.44525659464412e-05, + "loss": 3.1514, + "step": 14229 + }, + { + "epoch": 0.14590378216226013, + "grad_norm": 0.15077659487724304, + "learning_rate": 3.44500737985035e-05, + "loss": 3.1581, + "step": 14230 + }, + { + "epoch": 0.14593462862783355, + "grad_norm": 0.1472020298242569, + "learning_rate": 3.444758154099787e-05, + "loss": 3.1152, + "step": 14231 + }, + { + "epoch": 0.14596547509340696, + "grad_norm": 0.1517813354730606, + "learning_rate": 3.44450891739532e-05, + "loss": 3.1256, + "step": 14232 + }, + { + "epoch": 0.14599632155898037, + "grad_norm": 0.15079547464847565, + "learning_rate": 3.4442596697398414e-05, + "loss": 3.1195, + "step": 14233 + }, + { + "epoch": 0.1460271680245538, + "grad_norm": 0.1669587790966034, + "learning_rate": 3.444010411136239e-05, + "loss": 3.1269, + "step": 14234 + }, + { + "epoch": 0.1460580144901272, + "grad_norm": 0.1541866809129715, + "learning_rate": 3.443761141587403e-05, + "loss": 3.1221, + "step": 14235 + }, + { + "epoch": 0.14608886095570062, + "grad_norm": 0.15057243406772614, + "learning_rate": 3.443511861096222e-05, + "loss": 3.1104, + "step": 14236 + }, + { + "epoch": 0.14611970742127403, + "grad_norm": 0.1498246043920517, + "learning_rate": 3.44326256966559e-05, + "loss": 3.1254, + "step": 14237 + }, + { + "epoch": 0.14615055388684745, + "grad_norm": 0.15267518162727356, + "learning_rate": 3.443013267298395e-05, + "loss": 3.1167, + "step": 14238 + }, + { + "epoch": 0.14618140035242086, + "grad_norm": 0.1488199084997177, + "learning_rate": 3.442763953997527e-05, + "loss": 3.0976, + "step": 14239 + }, + { + "epoch": 0.14621224681799427, + "grad_norm": 0.14909933507442474, + "learning_rate": 3.442514629765879e-05, + "loss": 3.1421, + "step": 14240 + }, + { + "epoch": 0.14624309328356772, + "grad_norm": 0.1556067317724228, + "learning_rate": 3.442265294606339e-05, + "loss": 3.1214, + "step": 14241 + }, + { + "epoch": 0.14627393974914113, + "grad_norm": 0.1495363861322403, + "learning_rate": 3.442015948521799e-05, + "loss": 3.1424, + "step": 14242 + }, + { + "epoch": 0.14630478621471454, + "grad_norm": 0.1635206937789917, + "learning_rate": 3.441766591515151e-05, + "loss": 3.1295, + "step": 14243 + }, + { + "epoch": 0.14633563268028796, + "grad_norm": 0.1557466834783554, + "learning_rate": 3.4415172235892854e-05, + "loss": 3.1605, + "step": 14244 + }, + { + "epoch": 0.14636647914586137, + "grad_norm": 0.14855767786502838, + "learning_rate": 3.441267844747093e-05, + "loss": 3.0947, + "step": 14245 + }, + { + "epoch": 0.1463973256114348, + "grad_norm": 0.19353534281253815, + "learning_rate": 3.441018454991466e-05, + "loss": 3.1542, + "step": 14246 + }, + { + "epoch": 0.1464281720770082, + "grad_norm": 0.18961982429027557, + "learning_rate": 3.440769054325295e-05, + "loss": 3.1054, + "step": 14247 + }, + { + "epoch": 0.14645901854258161, + "grad_norm": 0.1550689935684204, + "learning_rate": 3.4405196427514716e-05, + "loss": 3.1088, + "step": 14248 + }, + { + "epoch": 0.14648986500815503, + "grad_norm": 0.15565747022628784, + "learning_rate": 3.4402702202728887e-05, + "loss": 3.1337, + "step": 14249 + }, + { + "epoch": 0.14652071147372844, + "grad_norm": 0.14843817055225372, + "learning_rate": 3.4400207868924374e-05, + "loss": 3.1101, + "step": 14250 + }, + { + "epoch": 0.14655155793930186, + "grad_norm": 0.16278159618377686, + "learning_rate": 3.4397713426130105e-05, + "loss": 3.1216, + "step": 14251 + }, + { + "epoch": 0.14658240440487527, + "grad_norm": 0.15731540322303772, + "learning_rate": 3.439521887437499e-05, + "loss": 3.1384, + "step": 14252 + }, + { + "epoch": 0.1466132508704487, + "grad_norm": 0.15943825244903564, + "learning_rate": 3.439272421368796e-05, + "loss": 3.1377, + "step": 14253 + }, + { + "epoch": 0.14664409733602213, + "grad_norm": 0.15924346446990967, + "learning_rate": 3.439022944409794e-05, + "loss": 3.0528, + "step": 14254 + }, + { + "epoch": 0.14667494380159554, + "grad_norm": 0.15247270464897156, + "learning_rate": 3.438773456563385e-05, + "loss": 3.122, + "step": 14255 + }, + { + "epoch": 0.14670579026716896, + "grad_norm": 0.15903885662555695, + "learning_rate": 3.438523957832462e-05, + "loss": 3.1255, + "step": 14256 + }, + { + "epoch": 0.14673663673274237, + "grad_norm": 0.14961554110050201, + "learning_rate": 3.438274448219918e-05, + "loss": 3.1205, + "step": 14257 + }, + { + "epoch": 0.14676748319831578, + "grad_norm": 0.15786980092525482, + "learning_rate": 3.438024927728645e-05, + "loss": 3.1139, + "step": 14258 + }, + { + "epoch": 0.1467983296638892, + "grad_norm": 0.17361463606357574, + "learning_rate": 3.437775396361537e-05, + "loss": 3.1128, + "step": 14259 + }, + { + "epoch": 0.1468291761294626, + "grad_norm": 0.15468448400497437, + "learning_rate": 3.437525854121487e-05, + "loss": 3.1598, + "step": 14260 + }, + { + "epoch": 0.14686002259503603, + "grad_norm": 0.15626442432403564, + "learning_rate": 3.4372763010113874e-05, + "loss": 3.0866, + "step": 14261 + }, + { + "epoch": 0.14689086906060944, + "grad_norm": 0.16039760410785675, + "learning_rate": 3.437026737034133e-05, + "loss": 3.1391, + "step": 14262 + }, + { + "epoch": 0.14692171552618286, + "grad_norm": 0.1600562185049057, + "learning_rate": 3.436777162192617e-05, + "loss": 3.1339, + "step": 14263 + }, + { + "epoch": 0.14695256199175627, + "grad_norm": 0.172490656375885, + "learning_rate": 3.436527576489732e-05, + "loss": 3.1368, + "step": 14264 + }, + { + "epoch": 0.1469834084573297, + "grad_norm": 0.1657240241765976, + "learning_rate": 3.4362779799283726e-05, + "loss": 3.1286, + "step": 14265 + }, + { + "epoch": 0.14701425492290313, + "grad_norm": 0.17240886390209198, + "learning_rate": 3.436028372511433e-05, + "loss": 3.1349, + "step": 14266 + }, + { + "epoch": 0.14704510138847654, + "grad_norm": 0.1627047061920166, + "learning_rate": 3.435778754241806e-05, + "loss": 3.1248, + "step": 14267 + }, + { + "epoch": 0.14707594785404995, + "grad_norm": 0.16981931030750275, + "learning_rate": 3.4355291251223875e-05, + "loss": 3.1512, + "step": 14268 + }, + { + "epoch": 0.14710679431962337, + "grad_norm": 0.17557229101657867, + "learning_rate": 3.435279485156071e-05, + "loss": 3.127, + "step": 14269 + }, + { + "epoch": 0.14713764078519678, + "grad_norm": 0.15510208904743195, + "learning_rate": 3.4350298343457505e-05, + "loss": 3.1583, + "step": 14270 + }, + { + "epoch": 0.1471684872507702, + "grad_norm": 0.1644531786441803, + "learning_rate": 3.434780172694321e-05, + "loss": 3.1474, + "step": 14271 + }, + { + "epoch": 0.1471993337163436, + "grad_norm": 0.16715744137763977, + "learning_rate": 3.434530500204678e-05, + "loss": 3.1121, + "step": 14272 + }, + { + "epoch": 0.14723018018191703, + "grad_norm": 0.1559322476387024, + "learning_rate": 3.434280816879715e-05, + "loss": 3.1482, + "step": 14273 + }, + { + "epoch": 0.14726102664749044, + "grad_norm": 0.16084684431552887, + "learning_rate": 3.4340311227223273e-05, + "loss": 3.1016, + "step": 14274 + }, + { + "epoch": 0.14729187311306385, + "grad_norm": 0.15309381484985352, + "learning_rate": 3.4337814177354096e-05, + "loss": 3.1063, + "step": 14275 + }, + { + "epoch": 0.14732271957863727, + "grad_norm": 0.15383228659629822, + "learning_rate": 3.433531701921857e-05, + "loss": 3.1238, + "step": 14276 + }, + { + "epoch": 0.1473535660442107, + "grad_norm": 0.16400329768657684, + "learning_rate": 3.433281975284567e-05, + "loss": 3.1248, + "step": 14277 + }, + { + "epoch": 0.14738441250978412, + "grad_norm": 0.1545610874891281, + "learning_rate": 3.433032237826432e-05, + "loss": 3.1173, + "step": 14278 + }, + { + "epoch": 0.14741525897535754, + "grad_norm": 0.16071772575378418, + "learning_rate": 3.43278248955035e-05, + "loss": 3.1468, + "step": 14279 + }, + { + "epoch": 0.14744610544093095, + "grad_norm": 0.1622474044561386, + "learning_rate": 3.432532730459215e-05, + "loss": 3.1086, + "step": 14280 + }, + { + "epoch": 0.14747695190650437, + "grad_norm": 0.15534469485282898, + "learning_rate": 3.4322829605559225e-05, + "loss": 3.0996, + "step": 14281 + }, + { + "epoch": 0.14750779837207778, + "grad_norm": 0.1660146862268448, + "learning_rate": 3.43203317984337e-05, + "loss": 3.1703, + "step": 14282 + }, + { + "epoch": 0.1475386448376512, + "grad_norm": 0.16300657391548157, + "learning_rate": 3.431783388324453e-05, + "loss": 3.1067, + "step": 14283 + }, + { + "epoch": 0.1475694913032246, + "grad_norm": 0.1609652191400528, + "learning_rate": 3.431533586002068e-05, + "loss": 3.1036, + "step": 14284 + }, + { + "epoch": 0.14760033776879802, + "grad_norm": 0.1552424430847168, + "learning_rate": 3.43128377287911e-05, + "loss": 3.138, + "step": 14285 + }, + { + "epoch": 0.14763118423437144, + "grad_norm": 0.1613544076681137, + "learning_rate": 3.431033948958476e-05, + "loss": 3.1276, + "step": 14286 + }, + { + "epoch": 0.14766203069994485, + "grad_norm": 0.1735762357711792, + "learning_rate": 3.430784114243064e-05, + "loss": 3.1327, + "step": 14287 + }, + { + "epoch": 0.14769287716551827, + "grad_norm": 0.1548198163509369, + "learning_rate": 3.430534268735769e-05, + "loss": 3.1441, + "step": 14288 + }, + { + "epoch": 0.1477237236310917, + "grad_norm": 0.20528551936149597, + "learning_rate": 3.4302844124394886e-05, + "loss": 3.1196, + "step": 14289 + }, + { + "epoch": 0.14775457009666512, + "grad_norm": 0.16077911853790283, + "learning_rate": 3.43003454535712e-05, + "loss": 3.1272, + "step": 14290 + }, + { + "epoch": 0.14778541656223854, + "grad_norm": 0.15406955778598785, + "learning_rate": 3.429784667491558e-05, + "loss": 3.1198, + "step": 14291 + }, + { + "epoch": 0.14781626302781195, + "grad_norm": 0.1682729423046112, + "learning_rate": 3.4295347788457026e-05, + "loss": 3.1345, + "step": 14292 + }, + { + "epoch": 0.14784710949338536, + "grad_norm": 0.1570587307214737, + "learning_rate": 3.4292848794224505e-05, + "loss": 3.1515, + "step": 14293 + }, + { + "epoch": 0.14787795595895878, + "grad_norm": 0.15306498110294342, + "learning_rate": 3.429034969224697e-05, + "loss": 3.0812, + "step": 14294 + }, + { + "epoch": 0.1479088024245322, + "grad_norm": 0.1530916541814804, + "learning_rate": 3.428785048255343e-05, + "loss": 3.1131, + "step": 14295 + }, + { + "epoch": 0.1479396488901056, + "grad_norm": 0.15321089327335358, + "learning_rate": 3.428535116517284e-05, + "loss": 3.0769, + "step": 14296 + }, + { + "epoch": 0.14797049535567902, + "grad_norm": 0.1592373549938202, + "learning_rate": 3.428285174013418e-05, + "loss": 3.1206, + "step": 14297 + }, + { + "epoch": 0.14800134182125244, + "grad_norm": 0.15209881961345673, + "learning_rate": 3.4280352207466436e-05, + "loss": 3.1309, + "step": 14298 + }, + { + "epoch": 0.14803218828682585, + "grad_norm": 0.1581445038318634, + "learning_rate": 3.4277852567198586e-05, + "loss": 3.1135, + "step": 14299 + }, + { + "epoch": 0.14806303475239926, + "grad_norm": 0.1579919010400772, + "learning_rate": 3.4275352819359604e-05, + "loss": 3.1211, + "step": 14300 + }, + { + "epoch": 0.1480938812179727, + "grad_norm": 0.15331484377384186, + "learning_rate": 3.427285296397849e-05, + "loss": 3.1194, + "step": 14301 + }, + { + "epoch": 0.14812472768354612, + "grad_norm": 0.15709288418293, + "learning_rate": 3.427035300108421e-05, + "loss": 3.1341, + "step": 14302 + }, + { + "epoch": 0.14815557414911953, + "grad_norm": 0.17552463710308075, + "learning_rate": 3.4267852930705755e-05, + "loss": 3.1643, + "step": 14303 + }, + { + "epoch": 0.14818642061469295, + "grad_norm": 0.15395578742027283, + "learning_rate": 3.4265352752872124e-05, + "loss": 3.115, + "step": 14304 + }, + { + "epoch": 0.14821726708026636, + "grad_norm": 0.15769365429878235, + "learning_rate": 3.426285246761229e-05, + "loss": 3.1318, + "step": 14305 + }, + { + "epoch": 0.14824811354583978, + "grad_norm": 0.1525404304265976, + "learning_rate": 3.426035207495525e-05, + "loss": 3.1423, + "step": 14306 + }, + { + "epoch": 0.1482789600114132, + "grad_norm": 0.15569335222244263, + "learning_rate": 3.425785157492999e-05, + "loss": 3.1194, + "step": 14307 + }, + { + "epoch": 0.1483098064769866, + "grad_norm": 0.14865660667419434, + "learning_rate": 3.4255350967565506e-05, + "loss": 3.1406, + "step": 14308 + }, + { + "epoch": 0.14834065294256002, + "grad_norm": 0.1551600992679596, + "learning_rate": 3.425285025289079e-05, + "loss": 3.1584, + "step": 14309 + }, + { + "epoch": 0.14837149940813343, + "grad_norm": 0.1525895744562149, + "learning_rate": 3.4250349430934836e-05, + "loss": 3.0958, + "step": 14310 + }, + { + "epoch": 0.14840234587370685, + "grad_norm": 0.1456896960735321, + "learning_rate": 3.4247848501726636e-05, + "loss": 3.13, + "step": 14311 + }, + { + "epoch": 0.14843319233928026, + "grad_norm": 0.14533838629722595, + "learning_rate": 3.42453474652952e-05, + "loss": 3.0757, + "step": 14312 + }, + { + "epoch": 0.1484640388048537, + "grad_norm": 0.15083551406860352, + "learning_rate": 3.4242846321669506e-05, + "loss": 3.1247, + "step": 14313 + }, + { + "epoch": 0.14849488527042712, + "grad_norm": 0.1477842479944229, + "learning_rate": 3.424034507087855e-05, + "loss": 3.1117, + "step": 14314 + }, + { + "epoch": 0.14852573173600053, + "grad_norm": 0.15286073088645935, + "learning_rate": 3.4237843712951376e-05, + "loss": 3.13, + "step": 14315 + }, + { + "epoch": 0.14855657820157395, + "grad_norm": 0.1459997445344925, + "learning_rate": 3.423534224791694e-05, + "loss": 3.1207, + "step": 14316 + }, + { + "epoch": 0.14858742466714736, + "grad_norm": 0.14590884745121002, + "learning_rate": 3.423284067580426e-05, + "loss": 3.1439, + "step": 14317 + }, + { + "epoch": 0.14861827113272078, + "grad_norm": 0.14769425988197327, + "learning_rate": 3.4230338996642344e-05, + "loss": 3.1217, + "step": 14318 + }, + { + "epoch": 0.1486491175982942, + "grad_norm": 0.1511652022600174, + "learning_rate": 3.422783721046019e-05, + "loss": 3.1476, + "step": 14319 + }, + { + "epoch": 0.1486799640638676, + "grad_norm": 0.14994683861732483, + "learning_rate": 3.422533531728681e-05, + "loss": 3.1369, + "step": 14320 + }, + { + "epoch": 0.14871081052944102, + "grad_norm": 0.1505090296268463, + "learning_rate": 3.422283331715121e-05, + "loss": 3.1096, + "step": 14321 + }, + { + "epoch": 0.14874165699501443, + "grad_norm": 0.14680403470993042, + "learning_rate": 3.42203312100824e-05, + "loss": 3.1084, + "step": 14322 + }, + { + "epoch": 0.14877250346058785, + "grad_norm": 0.15250149369239807, + "learning_rate": 3.421782899610939e-05, + "loss": 3.1386, + "step": 14323 + }, + { + "epoch": 0.14880334992616126, + "grad_norm": 0.14635342359542847, + "learning_rate": 3.42153266752612e-05, + "loss": 3.1291, + "step": 14324 + }, + { + "epoch": 0.1488341963917347, + "grad_norm": 0.1460188627243042, + "learning_rate": 3.421282424756683e-05, + "loss": 3.1101, + "step": 14325 + }, + { + "epoch": 0.14886504285730812, + "grad_norm": 0.1478513479232788, + "learning_rate": 3.421032171305529e-05, + "loss": 3.1324, + "step": 14326 + }, + { + "epoch": 0.14889588932288153, + "grad_norm": 0.15808089077472687, + "learning_rate": 3.420781907175562e-05, + "loss": 3.1462, + "step": 14327 + }, + { + "epoch": 0.14892673578845494, + "grad_norm": 0.15044675767421722, + "learning_rate": 3.420531632369681e-05, + "loss": 3.1253, + "step": 14328 + }, + { + "epoch": 0.14895758225402836, + "grad_norm": 0.15279249846935272, + "learning_rate": 3.420281346890789e-05, + "loss": 3.1528, + "step": 14329 + }, + { + "epoch": 0.14898842871960177, + "grad_norm": 0.1493755728006363, + "learning_rate": 3.4200310507417885e-05, + "loss": 3.1226, + "step": 14330 + }, + { + "epoch": 0.1490192751851752, + "grad_norm": 0.15523816645145416, + "learning_rate": 3.41978074392558e-05, + "loss": 3.1411, + "step": 14331 + }, + { + "epoch": 0.1490501216507486, + "grad_norm": 0.15175509452819824, + "learning_rate": 3.4195304264450664e-05, + "loss": 3.117, + "step": 14332 + }, + { + "epoch": 0.14908096811632202, + "grad_norm": 0.16216784715652466, + "learning_rate": 3.4192800983031516e-05, + "loss": 3.1497, + "step": 14333 + }, + { + "epoch": 0.14911181458189543, + "grad_norm": 0.1527019739151001, + "learning_rate": 3.419029759502735e-05, + "loss": 3.1081, + "step": 14334 + }, + { + "epoch": 0.14914266104746884, + "grad_norm": 0.1514088660478592, + "learning_rate": 3.418779410046721e-05, + "loss": 3.1452, + "step": 14335 + }, + { + "epoch": 0.14917350751304226, + "grad_norm": 0.15192337334156036, + "learning_rate": 3.418529049938012e-05, + "loss": 3.1235, + "step": 14336 + }, + { + "epoch": 0.1492043539786157, + "grad_norm": 0.1515233963727951, + "learning_rate": 3.4182786791795106e-05, + "loss": 3.1318, + "step": 14337 + }, + { + "epoch": 0.14923520044418911, + "grad_norm": 0.1488068699836731, + "learning_rate": 3.4180282977741196e-05, + "loss": 3.1485, + "step": 14338 + }, + { + "epoch": 0.14926604690976253, + "grad_norm": 0.15954430401325226, + "learning_rate": 3.4177779057247426e-05, + "loss": 3.1337, + "step": 14339 + }, + { + "epoch": 0.14929689337533594, + "grad_norm": 0.1548590362071991, + "learning_rate": 3.4175275030342824e-05, + "loss": 3.1118, + "step": 14340 + }, + { + "epoch": 0.14932773984090936, + "grad_norm": 0.15344631671905518, + "learning_rate": 3.417277089705641e-05, + "loss": 3.1468, + "step": 14341 + }, + { + "epoch": 0.14935858630648277, + "grad_norm": 0.15081898868083954, + "learning_rate": 3.417026665741724e-05, + "loss": 3.1141, + "step": 14342 + }, + { + "epoch": 0.14938943277205619, + "grad_norm": 0.16229650378227234, + "learning_rate": 3.416776231145433e-05, + "loss": 3.1661, + "step": 14343 + }, + { + "epoch": 0.1494202792376296, + "grad_norm": 0.15028229355812073, + "learning_rate": 3.416525785919673e-05, + "loss": 3.1709, + "step": 14344 + }, + { + "epoch": 0.149451125703203, + "grad_norm": 0.15175019204616547, + "learning_rate": 3.4162753300673476e-05, + "loss": 3.0904, + "step": 14345 + }, + { + "epoch": 0.14948197216877643, + "grad_norm": 0.1526900976896286, + "learning_rate": 3.416024863591361e-05, + "loss": 3.1096, + "step": 14346 + }, + { + "epoch": 0.14951281863434984, + "grad_norm": 0.15307928621768951, + "learning_rate": 3.4157743864946153e-05, + "loss": 3.1234, + "step": 14347 + }, + { + "epoch": 0.14954366509992326, + "grad_norm": 0.14771665632724762, + "learning_rate": 3.415523898780016e-05, + "loss": 3.1119, + "step": 14348 + }, + { + "epoch": 0.1495745115654967, + "grad_norm": 0.14774687588214874, + "learning_rate": 3.415273400450468e-05, + "loss": 3.1289, + "step": 14349 + }, + { + "epoch": 0.1496053580310701, + "grad_norm": 0.14865455031394958, + "learning_rate": 3.4150228915088745e-05, + "loss": 3.1251, + "step": 14350 + }, + { + "epoch": 0.14963620449664353, + "grad_norm": 0.14966049790382385, + "learning_rate": 3.4147723719581405e-05, + "loss": 3.1322, + "step": 14351 + }, + { + "epoch": 0.14966705096221694, + "grad_norm": 0.15008734166622162, + "learning_rate": 3.4145218418011704e-05, + "loss": 3.167, + "step": 14352 + }, + { + "epoch": 0.14969789742779036, + "grad_norm": 0.149118110537529, + "learning_rate": 3.414271301040869e-05, + "loss": 3.1048, + "step": 14353 + }, + { + "epoch": 0.14972874389336377, + "grad_norm": 0.15109407901763916, + "learning_rate": 3.414020749680141e-05, + "loss": 3.1239, + "step": 14354 + }, + { + "epoch": 0.14975959035893718, + "grad_norm": 0.1551581621170044, + "learning_rate": 3.413770187721893e-05, + "loss": 3.1747, + "step": 14355 + }, + { + "epoch": 0.1497904368245106, + "grad_norm": 0.1500406414270401, + "learning_rate": 3.413519615169027e-05, + "loss": 3.1108, + "step": 14356 + }, + { + "epoch": 0.149821283290084, + "grad_norm": 0.15341171622276306, + "learning_rate": 3.413269032024452e-05, + "loss": 3.1233, + "step": 14357 + }, + { + "epoch": 0.14985212975565743, + "grad_norm": 0.14424796402454376, + "learning_rate": 3.41301843829107e-05, + "loss": 3.0879, + "step": 14358 + }, + { + "epoch": 0.14988297622123084, + "grad_norm": 0.15167748928070068, + "learning_rate": 3.4127678339717886e-05, + "loss": 3.1512, + "step": 14359 + }, + { + "epoch": 0.14991382268680425, + "grad_norm": 0.148590087890625, + "learning_rate": 3.412517219069512e-05, + "loss": 3.1262, + "step": 14360 + }, + { + "epoch": 0.1499446691523777, + "grad_norm": 0.15286067128181458, + "learning_rate": 3.4122665935871466e-05, + "loss": 3.1527, + "step": 14361 + }, + { + "epoch": 0.1499755156179511, + "grad_norm": 0.14918862283229828, + "learning_rate": 3.412015957527599e-05, + "loss": 3.1284, + "step": 14362 + }, + { + "epoch": 0.15000636208352452, + "grad_norm": 0.1545591652393341, + "learning_rate": 3.411765310893774e-05, + "loss": 3.1435, + "step": 14363 + }, + { + "epoch": 0.15003720854909794, + "grad_norm": 0.15429681539535522, + "learning_rate": 3.4115146536885776e-05, + "loss": 3.125, + "step": 14364 + }, + { + "epoch": 0.15006805501467135, + "grad_norm": 0.15234768390655518, + "learning_rate": 3.4112639859149174e-05, + "loss": 3.1093, + "step": 14365 + }, + { + "epoch": 0.15009890148024477, + "grad_norm": 0.15722928941249847, + "learning_rate": 3.411013307575699e-05, + "loss": 3.1095, + "step": 14366 + }, + { + "epoch": 0.15012974794581818, + "grad_norm": 0.1516907811164856, + "learning_rate": 3.4107626186738276e-05, + "loss": 3.1473, + "step": 14367 + }, + { + "epoch": 0.1501605944113916, + "grad_norm": 0.15861757099628448, + "learning_rate": 3.410511919212212e-05, + "loss": 3.1287, + "step": 14368 + }, + { + "epoch": 0.150191440876965, + "grad_norm": 0.16109821200370789, + "learning_rate": 3.410261209193757e-05, + "loss": 3.1236, + "step": 14369 + }, + { + "epoch": 0.15022228734253842, + "grad_norm": 0.15286584198474884, + "learning_rate": 3.410010488621371e-05, + "loss": 3.1518, + "step": 14370 + }, + { + "epoch": 0.15025313380811184, + "grad_norm": 0.15222632884979248, + "learning_rate": 3.40975975749796e-05, + "loss": 3.1145, + "step": 14371 + }, + { + "epoch": 0.15028398027368525, + "grad_norm": 0.14845193922519684, + "learning_rate": 3.409509015826432e-05, + "loss": 3.1459, + "step": 14372 + }, + { + "epoch": 0.1503148267392587, + "grad_norm": 0.1649554967880249, + "learning_rate": 3.4092582636096926e-05, + "loss": 3.1186, + "step": 14373 + }, + { + "epoch": 0.1503456732048321, + "grad_norm": 0.15961673855781555, + "learning_rate": 3.409007500850651e-05, + "loss": 3.1487, + "step": 14374 + }, + { + "epoch": 0.15037651967040552, + "grad_norm": 0.15651005506515503, + "learning_rate": 3.408756727552212e-05, + "loss": 3.1276, + "step": 14375 + }, + { + "epoch": 0.15040736613597894, + "grad_norm": 0.14987778663635254, + "learning_rate": 3.408505943717286e-05, + "loss": 3.0948, + "step": 14376 + }, + { + "epoch": 0.15043821260155235, + "grad_norm": 0.16121138632297516, + "learning_rate": 3.408255149348779e-05, + "loss": 3.1312, + "step": 14377 + }, + { + "epoch": 0.15046905906712577, + "grad_norm": 0.15326139330863953, + "learning_rate": 3.4080043444496e-05, + "loss": 3.1269, + "step": 14378 + }, + { + "epoch": 0.15049990553269918, + "grad_norm": 0.15768927335739136, + "learning_rate": 3.407753529022656e-05, + "loss": 3.1486, + "step": 14379 + }, + { + "epoch": 0.1505307519982726, + "grad_norm": 0.15362009406089783, + "learning_rate": 3.407502703070856e-05, + "loss": 3.1419, + "step": 14380 + }, + { + "epoch": 0.150561598463846, + "grad_norm": 0.15716472268104553, + "learning_rate": 3.407251866597107e-05, + "loss": 3.1573, + "step": 14381 + }, + { + "epoch": 0.15059244492941942, + "grad_norm": 0.1486995965242386, + "learning_rate": 3.4070010196043165e-05, + "loss": 3.116, + "step": 14382 + }, + { + "epoch": 0.15062329139499284, + "grad_norm": 0.14882387220859528, + "learning_rate": 3.406750162095396e-05, + "loss": 3.1511, + "step": 14383 + }, + { + "epoch": 0.15065413786056625, + "grad_norm": 0.16442030668258667, + "learning_rate": 3.4064992940732516e-05, + "loss": 3.1262, + "step": 14384 + }, + { + "epoch": 0.1506849843261397, + "grad_norm": 0.15172161161899567, + "learning_rate": 3.4062484155407924e-05, + "loss": 3.1297, + "step": 14385 + }, + { + "epoch": 0.1507158307917131, + "grad_norm": 0.15946964919567108, + "learning_rate": 3.405997526500927e-05, + "loss": 3.1345, + "step": 14386 + }, + { + "epoch": 0.15074667725728652, + "grad_norm": 0.1504390686750412, + "learning_rate": 3.405746626956566e-05, + "loss": 3.1124, + "step": 14387 + }, + { + "epoch": 0.15077752372285994, + "grad_norm": 0.14840489625930786, + "learning_rate": 3.405495716910616e-05, + "loss": 3.1272, + "step": 14388 + }, + { + "epoch": 0.15080837018843335, + "grad_norm": 0.16619710624217987, + "learning_rate": 3.4052447963659885e-05, + "loss": 3.1384, + "step": 14389 + }, + { + "epoch": 0.15083921665400676, + "grad_norm": 0.15337945520877838, + "learning_rate": 3.404993865325591e-05, + "loss": 3.0992, + "step": 14390 + }, + { + "epoch": 0.15087006311958018, + "grad_norm": 0.15584193170070648, + "learning_rate": 3.404742923792333e-05, + "loss": 3.0961, + "step": 14391 + }, + { + "epoch": 0.1509009095851536, + "grad_norm": 0.14520631730556488, + "learning_rate": 3.404491971769124e-05, + "loss": 3.1182, + "step": 14392 + }, + { + "epoch": 0.150931756050727, + "grad_norm": 0.14943049848079681, + "learning_rate": 3.404241009258876e-05, + "loss": 3.1138, + "step": 14393 + }, + { + "epoch": 0.15096260251630042, + "grad_norm": 0.16084666550159454, + "learning_rate": 3.403990036264495e-05, + "loss": 3.1103, + "step": 14394 + }, + { + "epoch": 0.15099344898187383, + "grad_norm": 0.14947883784770966, + "learning_rate": 3.4037390527888944e-05, + "loss": 3.1321, + "step": 14395 + }, + { + "epoch": 0.15102429544744725, + "grad_norm": 0.154388427734375, + "learning_rate": 3.403488058834982e-05, + "loss": 3.129, + "step": 14396 + }, + { + "epoch": 0.1510551419130207, + "grad_norm": 0.15855903923511505, + "learning_rate": 3.4032370544056676e-05, + "loss": 3.1214, + "step": 14397 + }, + { + "epoch": 0.1510859883785941, + "grad_norm": 0.15648293495178223, + "learning_rate": 3.4029860395038624e-05, + "loss": 3.1215, + "step": 14398 + }, + { + "epoch": 0.15111683484416752, + "grad_norm": 0.1519925743341446, + "learning_rate": 3.402735014132478e-05, + "loss": 3.1188, + "step": 14399 + }, + { + "epoch": 0.15114768130974093, + "grad_norm": 0.16440191864967346, + "learning_rate": 3.402483978294423e-05, + "loss": 3.1354, + "step": 14400 + }, + { + "epoch": 0.15117852777531435, + "grad_norm": 0.1757269948720932, + "learning_rate": 3.4022329319926086e-05, + "loss": 3.106, + "step": 14401 + }, + { + "epoch": 0.15120937424088776, + "grad_norm": 0.16899482905864716, + "learning_rate": 3.4019818752299446e-05, + "loss": 3.0979, + "step": 14402 + }, + { + "epoch": 0.15124022070646118, + "grad_norm": 0.16525068879127502, + "learning_rate": 3.4017308080093435e-05, + "loss": 3.1078, + "step": 14403 + }, + { + "epoch": 0.1512710671720346, + "grad_norm": 0.16424071788787842, + "learning_rate": 3.4014797303337166e-05, + "loss": 3.136, + "step": 14404 + }, + { + "epoch": 0.151301913637608, + "grad_norm": 0.15917906165122986, + "learning_rate": 3.401228642205973e-05, + "loss": 3.1137, + "step": 14405 + }, + { + "epoch": 0.15133276010318142, + "grad_norm": 0.1686783730983734, + "learning_rate": 3.4009775436290245e-05, + "loss": 3.1356, + "step": 14406 + }, + { + "epoch": 0.15136360656875483, + "grad_norm": 0.15446697175502777, + "learning_rate": 3.4007264346057835e-05, + "loss": 3.1409, + "step": 14407 + }, + { + "epoch": 0.15139445303432825, + "grad_norm": 0.15839309990406036, + "learning_rate": 3.40047531513916e-05, + "loss": 3.1241, + "step": 14408 + }, + { + "epoch": 0.1514252994999017, + "grad_norm": 0.1643240749835968, + "learning_rate": 3.400224185232067e-05, + "loss": 3.1159, + "step": 14409 + }, + { + "epoch": 0.1514561459654751, + "grad_norm": 0.15134121477603912, + "learning_rate": 3.3999730448874154e-05, + "loss": 3.1359, + "step": 14410 + }, + { + "epoch": 0.15148699243104852, + "grad_norm": 0.16482311487197876, + "learning_rate": 3.3997218941081166e-05, + "loss": 3.1355, + "step": 14411 + }, + { + "epoch": 0.15151783889662193, + "grad_norm": 0.14956507086753845, + "learning_rate": 3.399470732897084e-05, + "loss": 3.1515, + "step": 14412 + }, + { + "epoch": 0.15154868536219535, + "grad_norm": 0.16675011813640594, + "learning_rate": 3.3992195612572285e-05, + "loss": 3.1014, + "step": 14413 + }, + { + "epoch": 0.15157953182776876, + "grad_norm": 0.1497851461172104, + "learning_rate": 3.398968379191462e-05, + "loss": 3.1305, + "step": 14414 + }, + { + "epoch": 0.15161037829334217, + "grad_norm": 0.15767797827720642, + "learning_rate": 3.3987171867026976e-05, + "loss": 3.137, + "step": 14415 + }, + { + "epoch": 0.1516412247589156, + "grad_norm": 0.1533249318599701, + "learning_rate": 3.3984659837938485e-05, + "loss": 3.1351, + "step": 14416 + }, + { + "epoch": 0.151672071224489, + "grad_norm": 0.14517845213413239, + "learning_rate": 3.3982147704678255e-05, + "loss": 3.0946, + "step": 14417 + }, + { + "epoch": 0.15170291769006242, + "grad_norm": 0.396704763174057, + "learning_rate": 3.397963546727542e-05, + "loss": 3.1341, + "step": 14418 + }, + { + "epoch": 0.15173376415563583, + "grad_norm": 0.15602941811084747, + "learning_rate": 3.3977123125759104e-05, + "loss": 3.1408, + "step": 14419 + }, + { + "epoch": 0.15176461062120924, + "grad_norm": 0.15350472927093506, + "learning_rate": 3.3974610680158436e-05, + "loss": 3.1255, + "step": 14420 + }, + { + "epoch": 0.1517954570867827, + "grad_norm": 0.15011106431484222, + "learning_rate": 3.3972098130502563e-05, + "loss": 3.1352, + "step": 14421 + }, + { + "epoch": 0.1518263035523561, + "grad_norm": 0.16287590563297272, + "learning_rate": 3.3969585476820596e-05, + "loss": 3.127, + "step": 14422 + }, + { + "epoch": 0.15185715001792952, + "grad_norm": 0.14876480400562286, + "learning_rate": 3.396707271914168e-05, + "loss": 3.1271, + "step": 14423 + }, + { + "epoch": 0.15188799648350293, + "grad_norm": 0.17175129055976868, + "learning_rate": 3.396455985749495e-05, + "loss": 3.1265, + "step": 14424 + }, + { + "epoch": 0.15191884294907634, + "grad_norm": 0.15303902328014374, + "learning_rate": 3.396204689190952e-05, + "loss": 3.1447, + "step": 14425 + }, + { + "epoch": 0.15194968941464976, + "grad_norm": 0.15294769406318665, + "learning_rate": 3.395953382241455e-05, + "loss": 3.1252, + "step": 14426 + }, + { + "epoch": 0.15198053588022317, + "grad_norm": 0.16472962498664856, + "learning_rate": 3.395702064903917e-05, + "loss": 3.1709, + "step": 14427 + }, + { + "epoch": 0.1520113823457966, + "grad_norm": 0.1622852087020874, + "learning_rate": 3.395450737181252e-05, + "loss": 3.0785, + "step": 14428 + }, + { + "epoch": 0.15204222881137, + "grad_norm": 0.16223233938217163, + "learning_rate": 3.395199399076373e-05, + "loss": 3.1417, + "step": 14429 + }, + { + "epoch": 0.15207307527694341, + "grad_norm": 0.15643396973609924, + "learning_rate": 3.394948050592196e-05, + "loss": 3.112, + "step": 14430 + }, + { + "epoch": 0.15210392174251683, + "grad_norm": 0.15487389266490936, + "learning_rate": 3.394696691731633e-05, + "loss": 3.1272, + "step": 14431 + }, + { + "epoch": 0.15213476820809024, + "grad_norm": 0.15283678472042084, + "learning_rate": 3.3944453224976e-05, + "loss": 3.1278, + "step": 14432 + }, + { + "epoch": 0.15216561467366368, + "grad_norm": 0.15362700819969177, + "learning_rate": 3.394193942893011e-05, + "loss": 3.1291, + "step": 14433 + }, + { + "epoch": 0.1521964611392371, + "grad_norm": 0.15414537489414215, + "learning_rate": 3.393942552920781e-05, + "loss": 3.1433, + "step": 14434 + }, + { + "epoch": 0.1522273076048105, + "grad_norm": 0.15184490382671356, + "learning_rate": 3.393691152583824e-05, + "loss": 3.1338, + "step": 14435 + }, + { + "epoch": 0.15225815407038393, + "grad_norm": 0.16222555935382843, + "learning_rate": 3.393439741885055e-05, + "loss": 3.1356, + "step": 14436 + }, + { + "epoch": 0.15228900053595734, + "grad_norm": 0.15001210570335388, + "learning_rate": 3.393188320827389e-05, + "loss": 3.1306, + "step": 14437 + }, + { + "epoch": 0.15231984700153076, + "grad_norm": 0.1459551453590393, + "learning_rate": 3.392936889413741e-05, + "loss": 3.1335, + "step": 14438 + }, + { + "epoch": 0.15235069346710417, + "grad_norm": 0.16193588078022003, + "learning_rate": 3.392685447647027e-05, + "loss": 3.1199, + "step": 14439 + }, + { + "epoch": 0.15238153993267758, + "grad_norm": 0.149070605635643, + "learning_rate": 3.392433995530162e-05, + "loss": 3.0967, + "step": 14440 + }, + { + "epoch": 0.152412386398251, + "grad_norm": 0.15533018112182617, + "learning_rate": 3.39218253306606e-05, + "loss": 3.1603, + "step": 14441 + }, + { + "epoch": 0.1524432328638244, + "grad_norm": 0.15138204395771027, + "learning_rate": 3.391931060257638e-05, + "loss": 3.138, + "step": 14442 + }, + { + "epoch": 0.15247407932939783, + "grad_norm": 0.15256235003471375, + "learning_rate": 3.391679577107812e-05, + "loss": 3.1434, + "step": 14443 + }, + { + "epoch": 0.15250492579497124, + "grad_norm": 0.17431075870990753, + "learning_rate": 3.3914280836194964e-05, + "loss": 3.1308, + "step": 14444 + }, + { + "epoch": 0.15253577226054468, + "grad_norm": 0.15079356729984283, + "learning_rate": 3.391176579795608e-05, + "loss": 3.1398, + "step": 14445 + }, + { + "epoch": 0.1525666187261181, + "grad_norm": 0.15355415642261505, + "learning_rate": 3.390925065639063e-05, + "loss": 3.1227, + "step": 14446 + }, + { + "epoch": 0.1525974651916915, + "grad_norm": 0.15626363456249237, + "learning_rate": 3.390673541152776e-05, + "loss": 3.1003, + "step": 14447 + }, + { + "epoch": 0.15262831165726493, + "grad_norm": 0.1501031517982483, + "learning_rate": 3.390422006339666e-05, + "loss": 3.1397, + "step": 14448 + }, + { + "epoch": 0.15265915812283834, + "grad_norm": 0.16085933148860931, + "learning_rate": 3.390170461202648e-05, + "loss": 3.1006, + "step": 14449 + }, + { + "epoch": 0.15269000458841175, + "grad_norm": 0.150247722864151, + "learning_rate": 3.3899189057446376e-05, + "loss": 3.1447, + "step": 14450 + }, + { + "epoch": 0.15272085105398517, + "grad_norm": 0.1517074555158615, + "learning_rate": 3.389667339968553e-05, + "loss": 3.1529, + "step": 14451 + }, + { + "epoch": 0.15275169751955858, + "grad_norm": 0.15204598009586334, + "learning_rate": 3.389415763877309e-05, + "loss": 3.1351, + "step": 14452 + }, + { + "epoch": 0.152782543985132, + "grad_norm": 0.1478978842496872, + "learning_rate": 3.389164177473825e-05, + "loss": 3.0893, + "step": 14453 + }, + { + "epoch": 0.1528133904507054, + "grad_norm": 0.15067356824874878, + "learning_rate": 3.388912580761016e-05, + "loss": 3.1381, + "step": 14454 + }, + { + "epoch": 0.15284423691627883, + "grad_norm": 0.15112578868865967, + "learning_rate": 3.388660973741801e-05, + "loss": 3.1219, + "step": 14455 + }, + { + "epoch": 0.15287508338185227, + "grad_norm": 0.15023581683635712, + "learning_rate": 3.3884093564190944e-05, + "loss": 3.1343, + "step": 14456 + }, + { + "epoch": 0.15290592984742568, + "grad_norm": 0.15827304124832153, + "learning_rate": 3.3881577287958156e-05, + "loss": 3.123, + "step": 14457 + }, + { + "epoch": 0.1529367763129991, + "grad_norm": 0.1500084400177002, + "learning_rate": 3.3879060908748814e-05, + "loss": 3.1074, + "step": 14458 + }, + { + "epoch": 0.1529676227785725, + "grad_norm": 0.15364110469818115, + "learning_rate": 3.3876544426592105e-05, + "loss": 3.1481, + "step": 14459 + }, + { + "epoch": 0.15299846924414592, + "grad_norm": 0.1550489366054535, + "learning_rate": 3.387402784151719e-05, + "loss": 3.1327, + "step": 14460 + }, + { + "epoch": 0.15302931570971934, + "grad_norm": 0.1520398110151291, + "learning_rate": 3.387151115355326e-05, + "loss": 3.1247, + "step": 14461 + }, + { + "epoch": 0.15306016217529275, + "grad_norm": 0.15406301617622375, + "learning_rate": 3.386899436272949e-05, + "loss": 3.1036, + "step": 14462 + }, + { + "epoch": 0.15309100864086617, + "grad_norm": 0.15240007638931274, + "learning_rate": 3.386647746907506e-05, + "loss": 3.1745, + "step": 14463 + }, + { + "epoch": 0.15312185510643958, + "grad_norm": 0.1524921953678131, + "learning_rate": 3.386396047261914e-05, + "loss": 3.1537, + "step": 14464 + }, + { + "epoch": 0.153152701572013, + "grad_norm": 0.17888693511486053, + "learning_rate": 3.386144337339094e-05, + "loss": 3.1684, + "step": 14465 + }, + { + "epoch": 0.1531835480375864, + "grad_norm": 0.16585251688957214, + "learning_rate": 3.385892617141963e-05, + "loss": 3.1233, + "step": 14466 + }, + { + "epoch": 0.15321439450315982, + "grad_norm": 0.1487908661365509, + "learning_rate": 3.3856408866734376e-05, + "loss": 3.1352, + "step": 14467 + }, + { + "epoch": 0.15324524096873327, + "grad_norm": 0.1656837910413742, + "learning_rate": 3.3853891459364405e-05, + "loss": 3.1285, + "step": 14468 + }, + { + "epoch": 0.15327608743430668, + "grad_norm": 0.23677551746368408, + "learning_rate": 3.385137394933887e-05, + "loss": 3.1447, + "step": 14469 + }, + { + "epoch": 0.1533069338998801, + "grad_norm": 0.16433420777320862, + "learning_rate": 3.384885633668697e-05, + "loss": 3.1063, + "step": 14470 + }, + { + "epoch": 0.1533377803654535, + "grad_norm": 0.16880054771900177, + "learning_rate": 3.384633862143791e-05, + "loss": 3.1176, + "step": 14471 + }, + { + "epoch": 0.15336862683102692, + "grad_norm": 0.16202537715435028, + "learning_rate": 3.3843820803620866e-05, + "loss": 3.1436, + "step": 14472 + }, + { + "epoch": 0.15339947329660034, + "grad_norm": 0.1619219332933426, + "learning_rate": 3.3841302883265035e-05, + "loss": 3.1602, + "step": 14473 + }, + { + "epoch": 0.15343031976217375, + "grad_norm": 0.16416683793067932, + "learning_rate": 3.383878486039961e-05, + "loss": 3.1641, + "step": 14474 + }, + { + "epoch": 0.15346116622774716, + "grad_norm": 0.16116517782211304, + "learning_rate": 3.3836266735053776e-05, + "loss": 3.1378, + "step": 14475 + }, + { + "epoch": 0.15349201269332058, + "grad_norm": 0.1558535248041153, + "learning_rate": 3.383374850725675e-05, + "loss": 3.1058, + "step": 14476 + }, + { + "epoch": 0.153522859158894, + "grad_norm": 0.14854957163333893, + "learning_rate": 3.3831230177037724e-05, + "loss": 3.1286, + "step": 14477 + }, + { + "epoch": 0.1535537056244674, + "grad_norm": 0.15633195638656616, + "learning_rate": 3.382871174442588e-05, + "loss": 3.1033, + "step": 14478 + }, + { + "epoch": 0.15358455209004082, + "grad_norm": 0.16270850598812103, + "learning_rate": 3.382619320945043e-05, + "loss": 3.118, + "step": 14479 + }, + { + "epoch": 0.15361539855561426, + "grad_norm": 0.15475775301456451, + "learning_rate": 3.3823674572140576e-05, + "loss": 3.1175, + "step": 14480 + }, + { + "epoch": 0.15364624502118768, + "grad_norm": 0.1543283313512802, + "learning_rate": 3.382115583252552e-05, + "loss": 3.107, + "step": 14481 + }, + { + "epoch": 0.1536770914867611, + "grad_norm": 0.16432911157608032, + "learning_rate": 3.3818636990634455e-05, + "loss": 3.1316, + "step": 14482 + }, + { + "epoch": 0.1537079379523345, + "grad_norm": 0.15349802374839783, + "learning_rate": 3.3816118046496605e-05, + "loss": 3.1505, + "step": 14483 + }, + { + "epoch": 0.15373878441790792, + "grad_norm": 0.15087366104125977, + "learning_rate": 3.381359900014116e-05, + "loss": 3.1255, + "step": 14484 + }, + { + "epoch": 0.15376963088348133, + "grad_norm": 0.15513727068901062, + "learning_rate": 3.3811079851597325e-05, + "loss": 3.109, + "step": 14485 + }, + { + "epoch": 0.15380047734905475, + "grad_norm": 0.15212179720401764, + "learning_rate": 3.380856060089431e-05, + "loss": 3.1098, + "step": 14486 + }, + { + "epoch": 0.15383132381462816, + "grad_norm": 0.15495885908603668, + "learning_rate": 3.3806041248061344e-05, + "loss": 3.1132, + "step": 14487 + }, + { + "epoch": 0.15386217028020158, + "grad_norm": 0.15144848823547363, + "learning_rate": 3.380352179312761e-05, + "loss": 3.1284, + "step": 14488 + }, + { + "epoch": 0.153893016745775, + "grad_norm": 0.15450598299503326, + "learning_rate": 3.3801002236122336e-05, + "loss": 3.1191, + "step": 14489 + }, + { + "epoch": 0.1539238632113484, + "grad_norm": 0.15504097938537598, + "learning_rate": 3.3798482577074734e-05, + "loss": 3.1545, + "step": 14490 + }, + { + "epoch": 0.15395470967692182, + "grad_norm": 0.15639784932136536, + "learning_rate": 3.379596281601401e-05, + "loss": 3.1338, + "step": 14491 + }, + { + "epoch": 0.15398555614249526, + "grad_norm": 0.15990835428237915, + "learning_rate": 3.3793442952969375e-05, + "loss": 3.1576, + "step": 14492 + }, + { + "epoch": 0.15401640260806868, + "grad_norm": 0.15596437454223633, + "learning_rate": 3.379092298797006e-05, + "loss": 3.1189, + "step": 14493 + }, + { + "epoch": 0.1540472490736421, + "grad_norm": 0.16558443009853363, + "learning_rate": 3.378840292104527e-05, + "loss": 3.1323, + "step": 14494 + }, + { + "epoch": 0.1540780955392155, + "grad_norm": 0.15294864773750305, + "learning_rate": 3.378588275222424e-05, + "loss": 3.1048, + "step": 14495 + }, + { + "epoch": 0.15410894200478892, + "grad_norm": 0.15888287127017975, + "learning_rate": 3.378336248153617e-05, + "loss": 3.1464, + "step": 14496 + }, + { + "epoch": 0.15413978847036233, + "grad_norm": 0.17102456092834473, + "learning_rate": 3.378084210901029e-05, + "loss": 3.1087, + "step": 14497 + }, + { + "epoch": 0.15417063493593575, + "grad_norm": 0.14998286962509155, + "learning_rate": 3.377832163467583e-05, + "loss": 3.1104, + "step": 14498 + }, + { + "epoch": 0.15420148140150916, + "grad_norm": 0.15606722235679626, + "learning_rate": 3.377580105856201e-05, + "loss": 3.0913, + "step": 14499 + }, + { + "epoch": 0.15423232786708257, + "grad_norm": 0.16190850734710693, + "learning_rate": 3.377328038069803e-05, + "loss": 3.1104, + "step": 14500 + }, + { + "epoch": 0.154263174332656, + "grad_norm": 0.18828275799751282, + "learning_rate": 3.377075960111315e-05, + "loss": 3.1398, + "step": 14501 + }, + { + "epoch": 0.1542940207982294, + "grad_norm": 0.24736233055591583, + "learning_rate": 3.376823871983659e-05, + "loss": 3.1235, + "step": 14502 + }, + { + "epoch": 0.15432486726380282, + "grad_norm": 0.15175378322601318, + "learning_rate": 3.376571773689755e-05, + "loss": 3.1031, + "step": 14503 + }, + { + "epoch": 0.15435571372937626, + "grad_norm": 0.15027469396591187, + "learning_rate": 3.37631966523253e-05, + "loss": 3.1435, + "step": 14504 + }, + { + "epoch": 0.15438656019494967, + "grad_norm": 0.15743036568164825, + "learning_rate": 3.376067546614905e-05, + "loss": 3.1324, + "step": 14505 + }, + { + "epoch": 0.1544174066605231, + "grad_norm": 0.17283381521701813, + "learning_rate": 3.375815417839802e-05, + "loss": 3.1308, + "step": 14506 + }, + { + "epoch": 0.1544482531260965, + "grad_norm": 0.15420863032341003, + "learning_rate": 3.3755632789101464e-05, + "loss": 3.1182, + "step": 14507 + }, + { + "epoch": 0.15447909959166992, + "grad_norm": 0.1543309986591339, + "learning_rate": 3.3753111298288606e-05, + "loss": 3.1477, + "step": 14508 + }, + { + "epoch": 0.15450994605724333, + "grad_norm": 0.15375734865665436, + "learning_rate": 3.375058970598868e-05, + "loss": 3.1381, + "step": 14509 + }, + { + "epoch": 0.15454079252281674, + "grad_norm": 0.15412570536136627, + "learning_rate": 3.374806801223093e-05, + "loss": 3.1248, + "step": 14510 + }, + { + "epoch": 0.15457163898839016, + "grad_norm": 0.1492650806903839, + "learning_rate": 3.374554621704458e-05, + "loss": 3.1139, + "step": 14511 + }, + { + "epoch": 0.15460248545396357, + "grad_norm": 0.152025505900383, + "learning_rate": 3.374302432045888e-05, + "loss": 3.1333, + "step": 14512 + }, + { + "epoch": 0.154633331919537, + "grad_norm": 0.15480941534042358, + "learning_rate": 3.374050232250308e-05, + "loss": 3.14, + "step": 14513 + }, + { + "epoch": 0.1546641783851104, + "grad_norm": 0.15775303542613983, + "learning_rate": 3.373798022320638e-05, + "loss": 3.1532, + "step": 14514 + }, + { + "epoch": 0.15469502485068382, + "grad_norm": 0.18803822994232178, + "learning_rate": 3.3735458022598074e-05, + "loss": 3.1154, + "step": 14515 + }, + { + "epoch": 0.15472587131625726, + "grad_norm": 0.15503552556037903, + "learning_rate": 3.3732935720707374e-05, + "loss": 3.1561, + "step": 14516 + }, + { + "epoch": 0.15475671778183067, + "grad_norm": 0.1595814824104309, + "learning_rate": 3.3730413317563535e-05, + "loss": 3.096, + "step": 14517 + }, + { + "epoch": 0.15478756424740409, + "grad_norm": 0.1496323198080063, + "learning_rate": 3.37278908131958e-05, + "loss": 3.0919, + "step": 14518 + }, + { + "epoch": 0.1548184107129775, + "grad_norm": 0.15542753040790558, + "learning_rate": 3.372536820763341e-05, + "loss": 3.1348, + "step": 14519 + }, + { + "epoch": 0.15484925717855091, + "grad_norm": 0.15353935956954956, + "learning_rate": 3.372284550090562e-05, + "loss": 3.1512, + "step": 14520 + }, + { + "epoch": 0.15488010364412433, + "grad_norm": 0.1508731096982956, + "learning_rate": 3.372032269304168e-05, + "loss": 3.134, + "step": 14521 + }, + { + "epoch": 0.15491095010969774, + "grad_norm": 0.14953787624835968, + "learning_rate": 3.371779978407085e-05, + "loss": 3.0911, + "step": 14522 + }, + { + "epoch": 0.15494179657527116, + "grad_norm": 0.14710089564323425, + "learning_rate": 3.371527677402235e-05, + "loss": 3.1184, + "step": 14523 + }, + { + "epoch": 0.15497264304084457, + "grad_norm": 0.15119454264640808, + "learning_rate": 3.371275366292547e-05, + "loss": 3.102, + "step": 14524 + }, + { + "epoch": 0.15500348950641799, + "grad_norm": 0.14743205904960632, + "learning_rate": 3.371023045080944e-05, + "loss": 3.1084, + "step": 14525 + }, + { + "epoch": 0.1550343359719914, + "grad_norm": 0.1504286378622055, + "learning_rate": 3.370770713770352e-05, + "loss": 3.1145, + "step": 14526 + }, + { + "epoch": 0.1550651824375648, + "grad_norm": 0.15939562022686005, + "learning_rate": 3.370518372363697e-05, + "loss": 3.1493, + "step": 14527 + }, + { + "epoch": 0.15509602890313826, + "grad_norm": 0.1558113992214203, + "learning_rate": 3.370266020863905e-05, + "loss": 3.1174, + "step": 14528 + }, + { + "epoch": 0.15512687536871167, + "grad_norm": 0.15717807412147522, + "learning_rate": 3.3700136592739004e-05, + "loss": 3.1106, + "step": 14529 + }, + { + "epoch": 0.15515772183428508, + "grad_norm": 0.15313155949115753, + "learning_rate": 3.369761287596611e-05, + "loss": 3.0863, + "step": 14530 + }, + { + "epoch": 0.1551885682998585, + "grad_norm": 0.14554254710674286, + "learning_rate": 3.3695089058349625e-05, + "loss": 3.1025, + "step": 14531 + }, + { + "epoch": 0.1552194147654319, + "grad_norm": 0.1503293663263321, + "learning_rate": 3.3692565139918805e-05, + "loss": 3.1491, + "step": 14532 + }, + { + "epoch": 0.15525026123100533, + "grad_norm": 0.15383794903755188, + "learning_rate": 3.3690041120702916e-05, + "loss": 3.1418, + "step": 14533 + }, + { + "epoch": 0.15528110769657874, + "grad_norm": 0.14868608117103577, + "learning_rate": 3.3687517000731225e-05, + "loss": 3.1126, + "step": 14534 + }, + { + "epoch": 0.15531195416215215, + "grad_norm": 0.1530807912349701, + "learning_rate": 3.3684992780032985e-05, + "loss": 3.0899, + "step": 14535 + }, + { + "epoch": 0.15534280062772557, + "grad_norm": 0.15858574211597443, + "learning_rate": 3.3682468458637476e-05, + "loss": 3.126, + "step": 14536 + }, + { + "epoch": 0.15537364709329898, + "grad_norm": 0.1493063122034073, + "learning_rate": 3.367994403657397e-05, + "loss": 3.11, + "step": 14537 + }, + { + "epoch": 0.1554044935588724, + "grad_norm": 0.1640312224626541, + "learning_rate": 3.367741951387172e-05, + "loss": 3.1486, + "step": 14538 + }, + { + "epoch": 0.1554353400244458, + "grad_norm": 0.15648187696933746, + "learning_rate": 3.367489489056001e-05, + "loss": 3.0991, + "step": 14539 + }, + { + "epoch": 0.15546618649001925, + "grad_norm": 0.1524624228477478, + "learning_rate": 3.367237016666811e-05, + "loss": 3.144, + "step": 14540 + }, + { + "epoch": 0.15549703295559267, + "grad_norm": 0.15636710822582245, + "learning_rate": 3.366984534222528e-05, + "loss": 3.1356, + "step": 14541 + }, + { + "epoch": 0.15552787942116608, + "grad_norm": 0.15960748493671417, + "learning_rate": 3.366732041726081e-05, + "loss": 3.1261, + "step": 14542 + }, + { + "epoch": 0.1555587258867395, + "grad_norm": 0.15335574746131897, + "learning_rate": 3.366479539180396e-05, + "loss": 3.1201, + "step": 14543 + }, + { + "epoch": 0.1555895723523129, + "grad_norm": 0.1513136625289917, + "learning_rate": 3.3662270265884024e-05, + "loss": 3.1375, + "step": 14544 + }, + { + "epoch": 0.15562041881788632, + "grad_norm": 0.15285363793373108, + "learning_rate": 3.365974503953027e-05, + "loss": 3.1135, + "step": 14545 + }, + { + "epoch": 0.15565126528345974, + "grad_norm": 0.15285131335258484, + "learning_rate": 3.365721971277198e-05, + "loss": 3.1338, + "step": 14546 + }, + { + "epoch": 0.15568211174903315, + "grad_norm": 0.14945544302463531, + "learning_rate": 3.3654694285638414e-05, + "loss": 3.1341, + "step": 14547 + }, + { + "epoch": 0.15571295821460657, + "grad_norm": 0.1488707959651947, + "learning_rate": 3.365216875815888e-05, + "loss": 3.1052, + "step": 14548 + }, + { + "epoch": 0.15574380468017998, + "grad_norm": 0.15263177454471588, + "learning_rate": 3.3649643130362655e-05, + "loss": 3.1286, + "step": 14549 + }, + { + "epoch": 0.1557746511457534, + "grad_norm": 0.1516878455877304, + "learning_rate": 3.3647117402279e-05, + "loss": 3.1441, + "step": 14550 + }, + { + "epoch": 0.1558054976113268, + "grad_norm": 0.15256232023239136, + "learning_rate": 3.364459157393722e-05, + "loss": 3.1319, + "step": 14551 + }, + { + "epoch": 0.15583634407690025, + "grad_norm": 0.1514604389667511, + "learning_rate": 3.36420656453666e-05, + "loss": 3.091, + "step": 14552 + }, + { + "epoch": 0.15586719054247367, + "grad_norm": 0.14881932735443115, + "learning_rate": 3.3639539616596416e-05, + "loss": 3.1156, + "step": 14553 + }, + { + "epoch": 0.15589803700804708, + "grad_norm": 0.147516667842865, + "learning_rate": 3.363701348765597e-05, + "loss": 3.0969, + "step": 14554 + }, + { + "epoch": 0.1559288834736205, + "grad_norm": 0.1546689122915268, + "learning_rate": 3.3634487258574536e-05, + "loss": 3.116, + "step": 14555 + }, + { + "epoch": 0.1559597299391939, + "grad_norm": 0.166395902633667, + "learning_rate": 3.363196092938141e-05, + "loss": 3.144, + "step": 14556 + }, + { + "epoch": 0.15599057640476732, + "grad_norm": 0.15302659571170807, + "learning_rate": 3.36294345001059e-05, + "loss": 3.1023, + "step": 14557 + }, + { + "epoch": 0.15602142287034074, + "grad_norm": 0.15164536237716675, + "learning_rate": 3.362690797077726e-05, + "loss": 3.1181, + "step": 14558 + }, + { + "epoch": 0.15605226933591415, + "grad_norm": 0.15314160287380219, + "learning_rate": 3.3624381341424816e-05, + "loss": 3.1342, + "step": 14559 + }, + { + "epoch": 0.15608311580148757, + "grad_norm": 0.1516169309616089, + "learning_rate": 3.362185461207785e-05, + "loss": 3.1244, + "step": 14560 + }, + { + "epoch": 0.15611396226706098, + "grad_norm": 0.15280233323574066, + "learning_rate": 3.361932778276566e-05, + "loss": 3.0977, + "step": 14561 + }, + { + "epoch": 0.1561448087326344, + "grad_norm": 0.15457479655742645, + "learning_rate": 3.361680085351756e-05, + "loss": 3.1107, + "step": 14562 + }, + { + "epoch": 0.1561756551982078, + "grad_norm": 0.15045137703418732, + "learning_rate": 3.3614273824362816e-05, + "loss": 3.1299, + "step": 14563 + }, + { + "epoch": 0.15620650166378125, + "grad_norm": 0.15107572078704834, + "learning_rate": 3.361174669533075e-05, + "loss": 3.1536, + "step": 14564 + }, + { + "epoch": 0.15623734812935466, + "grad_norm": 0.15003202855587006, + "learning_rate": 3.360921946645065e-05, + "loss": 3.1248, + "step": 14565 + }, + { + "epoch": 0.15626819459492808, + "grad_norm": 0.14615486562252045, + "learning_rate": 3.360669213775183e-05, + "loss": 3.1041, + "step": 14566 + }, + { + "epoch": 0.1562990410605015, + "grad_norm": 0.15351992845535278, + "learning_rate": 3.360416470926359e-05, + "loss": 3.1353, + "step": 14567 + }, + { + "epoch": 0.1563298875260749, + "grad_norm": 0.15287736058235168, + "learning_rate": 3.3601637181015225e-05, + "loss": 3.1038, + "step": 14568 + }, + { + "epoch": 0.15636073399164832, + "grad_norm": 0.1514740288257599, + "learning_rate": 3.3599109553036044e-05, + "loss": 3.1145, + "step": 14569 + }, + { + "epoch": 0.15639158045722173, + "grad_norm": 0.1521814912557602, + "learning_rate": 3.359658182535535e-05, + "loss": 3.1083, + "step": 14570 + }, + { + "epoch": 0.15642242692279515, + "grad_norm": 0.1610611081123352, + "learning_rate": 3.3594053998002464e-05, + "loss": 3.1658, + "step": 14571 + }, + { + "epoch": 0.15645327338836856, + "grad_norm": 0.15102382004261017, + "learning_rate": 3.3591526071006694e-05, + "loss": 3.1144, + "step": 14572 + }, + { + "epoch": 0.15648411985394198, + "grad_norm": 0.14808818697929382, + "learning_rate": 3.358899804439733e-05, + "loss": 3.1156, + "step": 14573 + }, + { + "epoch": 0.1565149663195154, + "grad_norm": 0.16010819375514984, + "learning_rate": 3.35864699182037e-05, + "loss": 3.1322, + "step": 14574 + }, + { + "epoch": 0.1565458127850888, + "grad_norm": 0.15422289073467255, + "learning_rate": 3.3583941692455114e-05, + "loss": 3.1384, + "step": 14575 + }, + { + "epoch": 0.15657665925066225, + "grad_norm": 0.15740641951560974, + "learning_rate": 3.358141336718087e-05, + "loss": 3.1393, + "step": 14576 + }, + { + "epoch": 0.15660750571623566, + "grad_norm": 0.15663979947566986, + "learning_rate": 3.35788849424103e-05, + "loss": 3.1356, + "step": 14577 + }, + { + "epoch": 0.15663835218180908, + "grad_norm": 0.14897847175598145, + "learning_rate": 3.3576356418172725e-05, + "loss": 3.1262, + "step": 14578 + }, + { + "epoch": 0.1566691986473825, + "grad_norm": 0.14973801374435425, + "learning_rate": 3.3573827794497444e-05, + "loss": 3.0873, + "step": 14579 + }, + { + "epoch": 0.1567000451129559, + "grad_norm": 0.15964575111865997, + "learning_rate": 3.3571299071413774e-05, + "loss": 3.0922, + "step": 14580 + }, + { + "epoch": 0.15673089157852932, + "grad_norm": 0.14967843890190125, + "learning_rate": 3.3568770248951055e-05, + "loss": 3.1057, + "step": 14581 + }, + { + "epoch": 0.15676173804410273, + "grad_norm": 0.152933269739151, + "learning_rate": 3.356624132713859e-05, + "loss": 3.1175, + "step": 14582 + }, + { + "epoch": 0.15679258450967615, + "grad_norm": 0.15250886976718903, + "learning_rate": 3.35637123060057e-05, + "loss": 3.1399, + "step": 14583 + }, + { + "epoch": 0.15682343097524956, + "grad_norm": 0.1536293476819992, + "learning_rate": 3.356118318558173e-05, + "loss": 3.1403, + "step": 14584 + }, + { + "epoch": 0.15685427744082298, + "grad_norm": 0.15388214588165283, + "learning_rate": 3.355865396589596e-05, + "loss": 3.112, + "step": 14585 + }, + { + "epoch": 0.1568851239063964, + "grad_norm": 0.14711564779281616, + "learning_rate": 3.3556124646977746e-05, + "loss": 3.1371, + "step": 14586 + }, + { + "epoch": 0.1569159703719698, + "grad_norm": 0.15091413259506226, + "learning_rate": 3.355359522885642e-05, + "loss": 3.1163, + "step": 14587 + }, + { + "epoch": 0.15694681683754325, + "grad_norm": 0.14724023640155792, + "learning_rate": 3.355106571156129e-05, + "loss": 3.1375, + "step": 14588 + }, + { + "epoch": 0.15697766330311666, + "grad_norm": 0.14681844413280487, + "learning_rate": 3.3548536095121696e-05, + "loss": 3.1013, + "step": 14589 + }, + { + "epoch": 0.15700850976869007, + "grad_norm": 0.1535087674856186, + "learning_rate": 3.3546006379566966e-05, + "loss": 3.1453, + "step": 14590 + }, + { + "epoch": 0.1570393562342635, + "grad_norm": 0.14814527332782745, + "learning_rate": 3.354347656492641e-05, + "loss": 3.1053, + "step": 14591 + }, + { + "epoch": 0.1570702026998369, + "grad_norm": 0.14970383048057556, + "learning_rate": 3.354094665122939e-05, + "loss": 3.1193, + "step": 14592 + }, + { + "epoch": 0.15710104916541032, + "grad_norm": 0.1465257853269577, + "learning_rate": 3.353841663850523e-05, + "loss": 3.112, + "step": 14593 + }, + { + "epoch": 0.15713189563098373, + "grad_norm": 0.14907430112361908, + "learning_rate": 3.353588652678325e-05, + "loss": 3.1211, + "step": 14594 + }, + { + "epoch": 0.15716274209655715, + "grad_norm": 0.1558019071817398, + "learning_rate": 3.35333563160928e-05, + "loss": 3.1144, + "step": 14595 + }, + { + "epoch": 0.15719358856213056, + "grad_norm": 0.15433703362941742, + "learning_rate": 3.353082600646322e-05, + "loss": 3.1221, + "step": 14596 + }, + { + "epoch": 0.15722443502770397, + "grad_norm": 0.14775371551513672, + "learning_rate": 3.352829559792382e-05, + "loss": 3.1244, + "step": 14597 + }, + { + "epoch": 0.1572552814932774, + "grad_norm": 0.14501318335533142, + "learning_rate": 3.352576509050397e-05, + "loss": 3.1326, + "step": 14598 + }, + { + "epoch": 0.1572861279588508, + "grad_norm": 0.1515725553035736, + "learning_rate": 3.3523234484233e-05, + "loss": 3.1273, + "step": 14599 + }, + { + "epoch": 0.15731697442442424, + "grad_norm": 0.15282264351844788, + "learning_rate": 3.352070377914024e-05, + "loss": 3.1268, + "step": 14600 + }, + { + "epoch": 0.15734782088999766, + "grad_norm": 0.15185096859931946, + "learning_rate": 3.3518172975255044e-05, + "loss": 3.1298, + "step": 14601 + }, + { + "epoch": 0.15737866735557107, + "grad_norm": 0.1629389077425003, + "learning_rate": 3.3515642072606746e-05, + "loss": 3.1469, + "step": 14602 + }, + { + "epoch": 0.1574095138211445, + "grad_norm": 0.15446491539478302, + "learning_rate": 3.35131110712247e-05, + "loss": 3.1321, + "step": 14603 + }, + { + "epoch": 0.1574403602867179, + "grad_norm": 0.15840020775794983, + "learning_rate": 3.351057997113825e-05, + "loss": 3.1336, + "step": 14604 + }, + { + "epoch": 0.15747120675229132, + "grad_norm": 0.16289971768856049, + "learning_rate": 3.350804877237674e-05, + "loss": 3.1579, + "step": 14605 + }, + { + "epoch": 0.15750205321786473, + "grad_norm": 0.14495033025741577, + "learning_rate": 3.350551747496951e-05, + "loss": 3.1573, + "step": 14606 + }, + { + "epoch": 0.15753289968343814, + "grad_norm": 0.14780466258525848, + "learning_rate": 3.350298607894592e-05, + "loss": 3.1215, + "step": 14607 + }, + { + "epoch": 0.15756374614901156, + "grad_norm": 0.14547577500343323, + "learning_rate": 3.350045458433532e-05, + "loss": 3.1263, + "step": 14608 + }, + { + "epoch": 0.15759459261458497, + "grad_norm": 0.1645391434431076, + "learning_rate": 3.3497922991167045e-05, + "loss": 3.1079, + "step": 14609 + }, + { + "epoch": 0.15762543908015839, + "grad_norm": 0.15161645412445068, + "learning_rate": 3.349539129947047e-05, + "loss": 3.1066, + "step": 14610 + }, + { + "epoch": 0.1576562855457318, + "grad_norm": 0.1793285459280014, + "learning_rate": 3.3492859509274935e-05, + "loss": 3.1312, + "step": 14611 + }, + { + "epoch": 0.15768713201130524, + "grad_norm": 0.15432067215442657, + "learning_rate": 3.34903276206098e-05, + "loss": 3.119, + "step": 14612 + }, + { + "epoch": 0.15771797847687866, + "grad_norm": 0.15402229130268097, + "learning_rate": 3.348779563350441e-05, + "loss": 3.0951, + "step": 14613 + }, + { + "epoch": 0.15774882494245207, + "grad_norm": 0.15593314170837402, + "learning_rate": 3.3485263547988136e-05, + "loss": 3.1367, + "step": 14614 + }, + { + "epoch": 0.15777967140802548, + "grad_norm": 0.1555114984512329, + "learning_rate": 3.348273136409034e-05, + "loss": 3.1171, + "step": 14615 + }, + { + "epoch": 0.1578105178735989, + "grad_norm": 0.1594313681125641, + "learning_rate": 3.348019908184036e-05, + "loss": 3.1199, + "step": 14616 + }, + { + "epoch": 0.1578413643391723, + "grad_norm": 0.15810352563858032, + "learning_rate": 3.347766670126757e-05, + "loss": 3.137, + "step": 14617 + }, + { + "epoch": 0.15787221080474573, + "grad_norm": 0.154951810836792, + "learning_rate": 3.347513422240133e-05, + "loss": 3.117, + "step": 14618 + }, + { + "epoch": 0.15790305727031914, + "grad_norm": 0.15270958840847015, + "learning_rate": 3.3472601645271e-05, + "loss": 3.1319, + "step": 14619 + }, + { + "epoch": 0.15793390373589256, + "grad_norm": 0.16076143085956573, + "learning_rate": 3.347006896990594e-05, + "loss": 3.1253, + "step": 14620 + }, + { + "epoch": 0.15796475020146597, + "grad_norm": 0.15982070565223694, + "learning_rate": 3.346753619633553e-05, + "loss": 3.1189, + "step": 14621 + }, + { + "epoch": 0.15799559666703938, + "grad_norm": 0.1619911640882492, + "learning_rate": 3.3465003324589125e-05, + "loss": 3.1101, + "step": 14622 + }, + { + "epoch": 0.1580264431326128, + "grad_norm": 0.15451255440711975, + "learning_rate": 3.346247035469608e-05, + "loss": 3.1052, + "step": 14623 + }, + { + "epoch": 0.15805728959818624, + "grad_norm": 0.16400481760501862, + "learning_rate": 3.34599372866858e-05, + "loss": 3.1306, + "step": 14624 + }, + { + "epoch": 0.15808813606375965, + "grad_norm": 0.15511909127235413, + "learning_rate": 3.3457404120587615e-05, + "loss": 3.1044, + "step": 14625 + }, + { + "epoch": 0.15811898252933307, + "grad_norm": 0.1539638340473175, + "learning_rate": 3.345487085643091e-05, + "loss": 3.097, + "step": 14626 + }, + { + "epoch": 0.15814982899490648, + "grad_norm": 0.15275314450263977, + "learning_rate": 3.3452337494245066e-05, + "loss": 3.1343, + "step": 14627 + }, + { + "epoch": 0.1581806754604799, + "grad_norm": 0.15739476680755615, + "learning_rate": 3.344980403405945e-05, + "loss": 3.1365, + "step": 14628 + }, + { + "epoch": 0.1582115219260533, + "grad_norm": 0.1602359563112259, + "learning_rate": 3.344727047590343e-05, + "loss": 3.1455, + "step": 14629 + }, + { + "epoch": 0.15824236839162673, + "grad_norm": 0.15497860312461853, + "learning_rate": 3.3444736819806375e-05, + "loss": 3.1113, + "step": 14630 + }, + { + "epoch": 0.15827321485720014, + "grad_norm": 0.15253318846225739, + "learning_rate": 3.344220306579769e-05, + "loss": 3.11, + "step": 14631 + }, + { + "epoch": 0.15830406132277355, + "grad_norm": 0.1718081533908844, + "learning_rate": 3.343966921390672e-05, + "loss": 3.1008, + "step": 14632 + }, + { + "epoch": 0.15833490778834697, + "grad_norm": 0.15047061443328857, + "learning_rate": 3.343713526416286e-05, + "loss": 3.1289, + "step": 14633 + }, + { + "epoch": 0.15836575425392038, + "grad_norm": 0.1501670479774475, + "learning_rate": 3.343460121659549e-05, + "loss": 3.1258, + "step": 14634 + }, + { + "epoch": 0.1583966007194938, + "grad_norm": 0.1510409265756607, + "learning_rate": 3.3432067071233985e-05, + "loss": 3.1551, + "step": 14635 + }, + { + "epoch": 0.15842744718506724, + "grad_norm": 0.14770732820034027, + "learning_rate": 3.342953282810772e-05, + "loss": 3.0981, + "step": 14636 + }, + { + "epoch": 0.15845829365064065, + "grad_norm": 0.1518569141626358, + "learning_rate": 3.3426998487246106e-05, + "loss": 3.1165, + "step": 14637 + }, + { + "epoch": 0.15848914011621407, + "grad_norm": 0.14886383712291718, + "learning_rate": 3.3424464048678495e-05, + "loss": 3.1121, + "step": 14638 + }, + { + "epoch": 0.15851998658178748, + "grad_norm": 0.15699243545532227, + "learning_rate": 3.34219295124343e-05, + "loss": 3.0823, + "step": 14639 + }, + { + "epoch": 0.1585508330473609, + "grad_norm": 0.15349091589450836, + "learning_rate": 3.341939487854288e-05, + "loss": 3.1227, + "step": 14640 + }, + { + "epoch": 0.1585816795129343, + "grad_norm": 0.15122170746326447, + "learning_rate": 3.341686014703364e-05, + "loss": 3.1604, + "step": 14641 + }, + { + "epoch": 0.15861252597850772, + "grad_norm": 0.1737399697303772, + "learning_rate": 3.3414325317935956e-05, + "loss": 3.1028, + "step": 14642 + }, + { + "epoch": 0.15864337244408114, + "grad_norm": 0.15401366353034973, + "learning_rate": 3.3411790391279243e-05, + "loss": 3.0977, + "step": 14643 + }, + { + "epoch": 0.15867421890965455, + "grad_norm": 0.150917649269104, + "learning_rate": 3.340925536709286e-05, + "loss": 3.0966, + "step": 14644 + }, + { + "epoch": 0.15870506537522797, + "grad_norm": 0.15028171241283417, + "learning_rate": 3.340672024540623e-05, + "loss": 3.1331, + "step": 14645 + }, + { + "epoch": 0.15873591184080138, + "grad_norm": 0.15365809202194214, + "learning_rate": 3.3404185026248724e-05, + "loss": 3.145, + "step": 14646 + }, + { + "epoch": 0.1587667583063748, + "grad_norm": 0.15471959114074707, + "learning_rate": 3.340164970964973e-05, + "loss": 3.1252, + "step": 14647 + }, + { + "epoch": 0.15879760477194824, + "grad_norm": 0.1490897834300995, + "learning_rate": 3.339911429563867e-05, + "loss": 3.0957, + "step": 14648 + }, + { + "epoch": 0.15882845123752165, + "grad_norm": 0.14841829240322113, + "learning_rate": 3.339657878424493e-05, + "loss": 3.1274, + "step": 14649 + }, + { + "epoch": 0.15885929770309506, + "grad_norm": 0.15247265994548798, + "learning_rate": 3.33940431754979e-05, + "loss": 3.0835, + "step": 14650 + }, + { + "epoch": 0.15889014416866848, + "grad_norm": 0.16310827434062958, + "learning_rate": 3.339150746942698e-05, + "loss": 3.1205, + "step": 14651 + }, + { + "epoch": 0.1589209906342419, + "grad_norm": 0.14943446218967438, + "learning_rate": 3.3388971666061575e-05, + "loss": 3.119, + "step": 14652 + }, + { + "epoch": 0.1589518370998153, + "grad_norm": 0.15276820957660675, + "learning_rate": 3.3386435765431086e-05, + "loss": 3.1207, + "step": 14653 + }, + { + "epoch": 0.15898268356538872, + "grad_norm": 0.1492909938097, + "learning_rate": 3.3383899767564914e-05, + "loss": 3.1006, + "step": 14654 + }, + { + "epoch": 0.15901353003096214, + "grad_norm": 0.15436460077762604, + "learning_rate": 3.338136367249247e-05, + "loss": 3.1591, + "step": 14655 + }, + { + "epoch": 0.15904437649653555, + "grad_norm": 0.15351007878780365, + "learning_rate": 3.337882748024314e-05, + "loss": 3.1393, + "step": 14656 + }, + { + "epoch": 0.15907522296210896, + "grad_norm": 0.1458900421857834, + "learning_rate": 3.3376291190846345e-05, + "loss": 3.1521, + "step": 14657 + }, + { + "epoch": 0.15910606942768238, + "grad_norm": 0.146723210811615, + "learning_rate": 3.337375480433148e-05, + "loss": 3.134, + "step": 14658 + }, + { + "epoch": 0.1591369158932558, + "grad_norm": 0.15166181325912476, + "learning_rate": 3.337121832072797e-05, + "loss": 3.0988, + "step": 14659 + }, + { + "epoch": 0.15916776235882923, + "grad_norm": 0.15081921219825745, + "learning_rate": 3.336868174006521e-05, + "loss": 3.1307, + "step": 14660 + }, + { + "epoch": 0.15919860882440265, + "grad_norm": 0.15270419418811798, + "learning_rate": 3.336614506237261e-05, + "loss": 3.1023, + "step": 14661 + }, + { + "epoch": 0.15922945528997606, + "grad_norm": 0.14992672204971313, + "learning_rate": 3.336360828767959e-05, + "loss": 3.1162, + "step": 14662 + }, + { + "epoch": 0.15926030175554948, + "grad_norm": 0.14750945568084717, + "learning_rate": 3.336107141601555e-05, + "loss": 3.1106, + "step": 14663 + }, + { + "epoch": 0.1592911482211229, + "grad_norm": 0.15479546785354614, + "learning_rate": 3.335853444740992e-05, + "loss": 3.1353, + "step": 14664 + }, + { + "epoch": 0.1593219946866963, + "grad_norm": 0.15340247750282288, + "learning_rate": 3.335599738189211e-05, + "loss": 3.1085, + "step": 14665 + }, + { + "epoch": 0.15935284115226972, + "grad_norm": 0.14787791669368744, + "learning_rate": 3.335346021949153e-05, + "loss": 3.0747, + "step": 14666 + }, + { + "epoch": 0.15938368761784313, + "grad_norm": 0.1545981913805008, + "learning_rate": 3.335092296023759e-05, + "loss": 3.098, + "step": 14667 + }, + { + "epoch": 0.15941453408341655, + "grad_norm": 0.14816713333129883, + "learning_rate": 3.334838560415973e-05, + "loss": 3.1496, + "step": 14668 + }, + { + "epoch": 0.15944538054898996, + "grad_norm": 0.15236693620681763, + "learning_rate": 3.334584815128734e-05, + "loss": 3.141, + "step": 14669 + }, + { + "epoch": 0.15947622701456338, + "grad_norm": 0.16072839498519897, + "learning_rate": 3.3343310601649864e-05, + "loss": 3.123, + "step": 14670 + }, + { + "epoch": 0.1595070734801368, + "grad_norm": 0.149248406291008, + "learning_rate": 3.334077295527672e-05, + "loss": 3.1068, + "step": 14671 + }, + { + "epoch": 0.15953791994571023, + "grad_norm": 0.19940851628780365, + "learning_rate": 3.333823521219732e-05, + "loss": 3.0904, + "step": 14672 + }, + { + "epoch": 0.15956876641128365, + "grad_norm": 0.155377596616745, + "learning_rate": 3.333569737244109e-05, + "loss": 3.1363, + "step": 14673 + }, + { + "epoch": 0.15959961287685706, + "grad_norm": 0.15303078293800354, + "learning_rate": 3.3333159436037466e-05, + "loss": 3.1164, + "step": 14674 + }, + { + "epoch": 0.15963045934243048, + "grad_norm": 0.1502324342727661, + "learning_rate": 3.333062140301586e-05, + "loss": 3.1528, + "step": 14675 + }, + { + "epoch": 0.1596613058080039, + "grad_norm": 0.14983665943145752, + "learning_rate": 3.3328083273405705e-05, + "loss": 3.1024, + "step": 14676 + }, + { + "epoch": 0.1596921522735773, + "grad_norm": 0.1505708247423172, + "learning_rate": 3.332554504723643e-05, + "loss": 3.1133, + "step": 14677 + }, + { + "epoch": 0.15972299873915072, + "grad_norm": 0.15189987421035767, + "learning_rate": 3.332300672453747e-05, + "loss": 3.1216, + "step": 14678 + }, + { + "epoch": 0.15975384520472413, + "grad_norm": 0.15449753403663635, + "learning_rate": 3.332046830533824e-05, + "loss": 3.1082, + "step": 14679 + }, + { + "epoch": 0.15978469167029755, + "grad_norm": 0.1442776471376419, + "learning_rate": 3.331792978966818e-05, + "loss": 3.1032, + "step": 14680 + }, + { + "epoch": 0.15981553813587096, + "grad_norm": 0.15738429129123688, + "learning_rate": 3.3315391177556725e-05, + "loss": 3.1213, + "step": 14681 + }, + { + "epoch": 0.15984638460144437, + "grad_norm": 0.15365219116210938, + "learning_rate": 3.3312852469033306e-05, + "loss": 3.1227, + "step": 14682 + }, + { + "epoch": 0.1598772310670178, + "grad_norm": 0.15116707980632782, + "learning_rate": 3.3310313664127364e-05, + "loss": 3.1226, + "step": 14683 + }, + { + "epoch": 0.15990807753259123, + "grad_norm": 0.15895682573318481, + "learning_rate": 3.330777476286833e-05, + "loss": 3.1254, + "step": 14684 + }, + { + "epoch": 0.15993892399816464, + "grad_norm": 0.15434925258159637, + "learning_rate": 3.3305235765285625e-05, + "loss": 3.1173, + "step": 14685 + }, + { + "epoch": 0.15996977046373806, + "grad_norm": 0.15156270563602448, + "learning_rate": 3.3302696671408714e-05, + "loss": 3.1234, + "step": 14686 + }, + { + "epoch": 0.16000061692931147, + "grad_norm": 0.15143580734729767, + "learning_rate": 3.3300157481267016e-05, + "loss": 3.1468, + "step": 14687 + }, + { + "epoch": 0.1600314633948849, + "grad_norm": 0.15260684490203857, + "learning_rate": 3.329761819488999e-05, + "loss": 3.0856, + "step": 14688 + }, + { + "epoch": 0.1600623098604583, + "grad_norm": 0.15100154280662537, + "learning_rate": 3.329507881230706e-05, + "loss": 3.1085, + "step": 14689 + }, + { + "epoch": 0.16009315632603172, + "grad_norm": 0.1460326611995697, + "learning_rate": 3.3292539333547683e-05, + "loss": 3.1211, + "step": 14690 + }, + { + "epoch": 0.16012400279160513, + "grad_norm": 0.14801530539989471, + "learning_rate": 3.328999975864129e-05, + "loss": 3.1247, + "step": 14691 + }, + { + "epoch": 0.16015484925717854, + "grad_norm": 0.15008343756198883, + "learning_rate": 3.3287460087617325e-05, + "loss": 3.1327, + "step": 14692 + }, + { + "epoch": 0.16018569572275196, + "grad_norm": 0.15026788413524628, + "learning_rate": 3.328492032050525e-05, + "loss": 3.0957, + "step": 14693 + }, + { + "epoch": 0.16021654218832537, + "grad_norm": 0.1472223699092865, + "learning_rate": 3.3282380457334505e-05, + "loss": 3.0977, + "step": 14694 + }, + { + "epoch": 0.1602473886538988, + "grad_norm": 0.15327118337154388, + "learning_rate": 3.3279840498134536e-05, + "loss": 3.1127, + "step": 14695 + }, + { + "epoch": 0.16027823511947223, + "grad_norm": 0.1535491943359375, + "learning_rate": 3.3277300442934786e-05, + "loss": 3.1156, + "step": 14696 + }, + { + "epoch": 0.16030908158504564, + "grad_norm": 0.1633671075105667, + "learning_rate": 3.3274760291764706e-05, + "loss": 3.1035, + "step": 14697 + }, + { + "epoch": 0.16033992805061906, + "grad_norm": 0.49527689814567566, + "learning_rate": 3.3272220044653765e-05, + "loss": 3.0857, + "step": 14698 + }, + { + "epoch": 0.16037077451619247, + "grad_norm": 0.15773868560791016, + "learning_rate": 3.32696797016314e-05, + "loss": 3.1398, + "step": 14699 + }, + { + "epoch": 0.16040162098176589, + "grad_norm": 0.16226641833782196, + "learning_rate": 3.326713926272707e-05, + "loss": 3.0995, + "step": 14700 + }, + { + "epoch": 0.1604324674473393, + "grad_norm": 0.17207054793834686, + "learning_rate": 3.326459872797022e-05, + "loss": 3.1516, + "step": 14701 + }, + { + "epoch": 0.16046331391291271, + "grad_norm": 0.15288479626178741, + "learning_rate": 3.3262058097390317e-05, + "loss": 3.1368, + "step": 14702 + }, + { + "epoch": 0.16049416037848613, + "grad_norm": 0.1514563411474228, + "learning_rate": 3.325951737101682e-05, + "loss": 3.1221, + "step": 14703 + }, + { + "epoch": 0.16052500684405954, + "grad_norm": 0.1602700650691986, + "learning_rate": 3.3256976548879184e-05, + "loss": 3.1171, + "step": 14704 + }, + { + "epoch": 0.16055585330963296, + "grad_norm": 0.16363942623138428, + "learning_rate": 3.3254435631006864e-05, + "loss": 3.1218, + "step": 14705 + }, + { + "epoch": 0.16058669977520637, + "grad_norm": 0.1522882878780365, + "learning_rate": 3.325189461742932e-05, + "loss": 3.1021, + "step": 14706 + }, + { + "epoch": 0.16061754624077978, + "grad_norm": 0.15053711831569672, + "learning_rate": 3.324935350817602e-05, + "loss": 3.1199, + "step": 14707 + }, + { + "epoch": 0.16064839270635323, + "grad_norm": 0.15606412291526794, + "learning_rate": 3.324681230327642e-05, + "loss": 3.1395, + "step": 14708 + }, + { + "epoch": 0.16067923917192664, + "grad_norm": 0.1520317792892456, + "learning_rate": 3.324427100275998e-05, + "loss": 3.122, + "step": 14709 + }, + { + "epoch": 0.16071008563750006, + "grad_norm": 0.15636903047561646, + "learning_rate": 3.3241729606656194e-05, + "loss": 3.1292, + "step": 14710 + }, + { + "epoch": 0.16074093210307347, + "grad_norm": 0.15005502104759216, + "learning_rate": 3.3239188114994486e-05, + "loss": 3.0996, + "step": 14711 + }, + { + "epoch": 0.16077177856864688, + "grad_norm": 0.15496574342250824, + "learning_rate": 3.323664652780435e-05, + "loss": 3.1305, + "step": 14712 + }, + { + "epoch": 0.1608026250342203, + "grad_norm": 0.15389421582221985, + "learning_rate": 3.323410484511524e-05, + "loss": 3.1189, + "step": 14713 + }, + { + "epoch": 0.1608334714997937, + "grad_norm": 0.15670372545719147, + "learning_rate": 3.323156306695665e-05, + "loss": 3.1501, + "step": 14714 + }, + { + "epoch": 0.16086431796536713, + "grad_norm": 0.15199369192123413, + "learning_rate": 3.322902119335802e-05, + "loss": 3.0882, + "step": 14715 + }, + { + "epoch": 0.16089516443094054, + "grad_norm": 0.15639549493789673, + "learning_rate": 3.322647922434884e-05, + "loss": 3.1391, + "step": 14716 + }, + { + "epoch": 0.16092601089651395, + "grad_norm": 0.15695126354694366, + "learning_rate": 3.322393715995857e-05, + "loss": 3.1244, + "step": 14717 + }, + { + "epoch": 0.16095685736208737, + "grad_norm": 0.1582154482603073, + "learning_rate": 3.32213950002167e-05, + "loss": 3.1182, + "step": 14718 + }, + { + "epoch": 0.16098770382766078, + "grad_norm": 0.2088017463684082, + "learning_rate": 3.321885274515268e-05, + "loss": 3.1334, + "step": 14719 + }, + { + "epoch": 0.16101855029323422, + "grad_norm": 0.1577482372522354, + "learning_rate": 3.3216310394796015e-05, + "loss": 3.1478, + "step": 14720 + }, + { + "epoch": 0.16104939675880764, + "grad_norm": 0.15632866322994232, + "learning_rate": 3.321376794917617e-05, + "loss": 3.0753, + "step": 14721 + }, + { + "epoch": 0.16108024322438105, + "grad_norm": 0.16285894811153412, + "learning_rate": 3.321122540832261e-05, + "loss": 3.1279, + "step": 14722 + }, + { + "epoch": 0.16111108968995447, + "grad_norm": 0.1501881182193756, + "learning_rate": 3.320868277226483e-05, + "loss": 3.1465, + "step": 14723 + }, + { + "epoch": 0.16114193615552788, + "grad_norm": 0.15829336643218994, + "learning_rate": 3.320614004103231e-05, + "loss": 3.1346, + "step": 14724 + }, + { + "epoch": 0.1611727826211013, + "grad_norm": 0.15534624457359314, + "learning_rate": 3.320359721465453e-05, + "loss": 3.1176, + "step": 14725 + }, + { + "epoch": 0.1612036290866747, + "grad_norm": 0.15635395050048828, + "learning_rate": 3.320105429316096e-05, + "loss": 3.1221, + "step": 14726 + }, + { + "epoch": 0.16123447555224812, + "grad_norm": 0.1566358059644699, + "learning_rate": 3.31985112765811e-05, + "loss": 3.1067, + "step": 14727 + }, + { + "epoch": 0.16126532201782154, + "grad_norm": 0.1538948267698288, + "learning_rate": 3.319596816494444e-05, + "loss": 3.0992, + "step": 14728 + }, + { + "epoch": 0.16129616848339495, + "grad_norm": 0.1537388563156128, + "learning_rate": 3.319342495828043e-05, + "loss": 3.0936, + "step": 14729 + }, + { + "epoch": 0.16132701494896837, + "grad_norm": 0.15256468951702118, + "learning_rate": 3.3190881656618595e-05, + "loss": 3.1266, + "step": 14730 + }, + { + "epoch": 0.1613578614145418, + "grad_norm": 0.17820674180984497, + "learning_rate": 3.3188338259988414e-05, + "loss": 3.1201, + "step": 14731 + }, + { + "epoch": 0.16138870788011522, + "grad_norm": 0.15918061137199402, + "learning_rate": 3.318579476841936e-05, + "loss": 3.1163, + "step": 14732 + }, + { + "epoch": 0.16141955434568864, + "grad_norm": 0.1514243185520172, + "learning_rate": 3.3183251181940946e-05, + "loss": 3.1055, + "step": 14733 + }, + { + "epoch": 0.16145040081126205, + "grad_norm": 0.16051924228668213, + "learning_rate": 3.318070750058264e-05, + "loss": 3.1475, + "step": 14734 + }, + { + "epoch": 0.16148124727683547, + "grad_norm": 0.15425238013267517, + "learning_rate": 3.3178163724373956e-05, + "loss": 3.1364, + "step": 14735 + }, + { + "epoch": 0.16151209374240888, + "grad_norm": 0.15107260644435883, + "learning_rate": 3.317561985334437e-05, + "loss": 3.1353, + "step": 14736 + }, + { + "epoch": 0.1615429402079823, + "grad_norm": 0.17855115234851837, + "learning_rate": 3.317307588752339e-05, + "loss": 3.1217, + "step": 14737 + }, + { + "epoch": 0.1615737866735557, + "grad_norm": 0.15590959787368774, + "learning_rate": 3.317053182694051e-05, + "loss": 3.1321, + "step": 14738 + }, + { + "epoch": 0.16160463313912912, + "grad_norm": 0.18501828610897064, + "learning_rate": 3.316798767162522e-05, + "loss": 3.1295, + "step": 14739 + }, + { + "epoch": 0.16163547960470254, + "grad_norm": 0.30381155014038086, + "learning_rate": 3.316544342160702e-05, + "loss": 3.15, + "step": 14740 + }, + { + "epoch": 0.16166632607027595, + "grad_norm": 0.15026600658893585, + "learning_rate": 3.316289907691541e-05, + "loss": 3.1221, + "step": 14741 + }, + { + "epoch": 0.16169717253584937, + "grad_norm": 0.15891103446483612, + "learning_rate": 3.3160354637579896e-05, + "loss": 3.1174, + "step": 14742 + }, + { + "epoch": 0.1617280190014228, + "grad_norm": 0.1523505598306656, + "learning_rate": 3.3157810103629965e-05, + "loss": 3.1192, + "step": 14743 + }, + { + "epoch": 0.16175886546699622, + "grad_norm": 0.1524875909090042, + "learning_rate": 3.315526547509513e-05, + "loss": 3.1447, + "step": 14744 + }, + { + "epoch": 0.16178971193256964, + "grad_norm": 0.15894818305969238, + "learning_rate": 3.3152720752004904e-05, + "loss": 3.1388, + "step": 14745 + }, + { + "epoch": 0.16182055839814305, + "grad_norm": 0.15460993349552155, + "learning_rate": 3.315017593438877e-05, + "loss": 3.1411, + "step": 14746 + }, + { + "epoch": 0.16185140486371646, + "grad_norm": 0.23004525899887085, + "learning_rate": 3.314763102227624e-05, + "loss": 3.1187, + "step": 14747 + }, + { + "epoch": 0.16188225132928988, + "grad_norm": 0.1568644642829895, + "learning_rate": 3.314508601569683e-05, + "loss": 3.1364, + "step": 14748 + }, + { + "epoch": 0.1619130977948633, + "grad_norm": 0.15687207877635956, + "learning_rate": 3.314254091468004e-05, + "loss": 3.1281, + "step": 14749 + }, + { + "epoch": 0.1619439442604367, + "grad_norm": 0.15608616173267365, + "learning_rate": 3.313999571925538e-05, + "loss": 3.1095, + "step": 14750 + }, + { + "epoch": 0.16197479072601012, + "grad_norm": 0.1562786102294922, + "learning_rate": 3.3137450429452374e-05, + "loss": 3.087, + "step": 14751 + }, + { + "epoch": 0.16200563719158353, + "grad_norm": 0.16266438364982605, + "learning_rate": 3.313490504530051e-05, + "loss": 3.1035, + "step": 14752 + }, + { + "epoch": 0.16203648365715695, + "grad_norm": 0.15558773279190063, + "learning_rate": 3.313235956682931e-05, + "loss": 3.0983, + "step": 14753 + }, + { + "epoch": 0.16206733012273036, + "grad_norm": 0.15820375084877014, + "learning_rate": 3.312981399406829e-05, + "loss": 3.1181, + "step": 14754 + }, + { + "epoch": 0.1620981765883038, + "grad_norm": 0.16598571836948395, + "learning_rate": 3.312726832704696e-05, + "loss": 3.1068, + "step": 14755 + }, + { + "epoch": 0.16212902305387722, + "grad_norm": 0.16842524707317352, + "learning_rate": 3.312472256579484e-05, + "loss": 3.115, + "step": 14756 + }, + { + "epoch": 0.16215986951945063, + "grad_norm": 0.17437537014484406, + "learning_rate": 3.312217671034144e-05, + "loss": 3.1352, + "step": 14757 + }, + { + "epoch": 0.16219071598502405, + "grad_norm": 0.15943196415901184, + "learning_rate": 3.311963076071629e-05, + "loss": 3.1028, + "step": 14758 + }, + { + "epoch": 0.16222156245059746, + "grad_norm": 0.15450376272201538, + "learning_rate": 3.3117084716948885e-05, + "loss": 3.1218, + "step": 14759 + }, + { + "epoch": 0.16225240891617088, + "grad_norm": 0.18261587619781494, + "learning_rate": 3.3114538579068775e-05, + "loss": 3.1159, + "step": 14760 + }, + { + "epoch": 0.1622832553817443, + "grad_norm": 0.1665201485157013, + "learning_rate": 3.311199234710546e-05, + "loss": 3.1411, + "step": 14761 + }, + { + "epoch": 0.1623141018473177, + "grad_norm": 0.15674564242362976, + "learning_rate": 3.3109446021088475e-05, + "loss": 3.1316, + "step": 14762 + }, + { + "epoch": 0.16234494831289112, + "grad_norm": 0.15555627644062042, + "learning_rate": 3.310689960104733e-05, + "loss": 3.11, + "step": 14763 + }, + { + "epoch": 0.16237579477846453, + "grad_norm": 0.1601734608411789, + "learning_rate": 3.310435308701156e-05, + "loss": 3.1162, + "step": 14764 + }, + { + "epoch": 0.16240664124403795, + "grad_norm": 0.15104614198207855, + "learning_rate": 3.310180647901068e-05, + "loss": 3.1068, + "step": 14765 + }, + { + "epoch": 0.16243748770961136, + "grad_norm": 0.15682850778102875, + "learning_rate": 3.309925977707423e-05, + "loss": 3.1189, + "step": 14766 + }, + { + "epoch": 0.1624683341751848, + "grad_norm": 0.154885932803154, + "learning_rate": 3.309671298123173e-05, + "loss": 3.162, + "step": 14767 + }, + { + "epoch": 0.16249918064075822, + "grad_norm": 0.16927827894687653, + "learning_rate": 3.309416609151271e-05, + "loss": 3.1063, + "step": 14768 + }, + { + "epoch": 0.16253002710633163, + "grad_norm": 0.15560173988342285, + "learning_rate": 3.3091619107946684e-05, + "loss": 3.1291, + "step": 14769 + }, + { + "epoch": 0.16256087357190505, + "grad_norm": 0.16420383751392365, + "learning_rate": 3.30890720305632e-05, + "loss": 3.0941, + "step": 14770 + }, + { + "epoch": 0.16259172003747846, + "grad_norm": 0.16146539151668549, + "learning_rate": 3.30865248593918e-05, + "loss": 3.1442, + "step": 14771 + }, + { + "epoch": 0.16262256650305187, + "grad_norm": 0.15225860476493835, + "learning_rate": 3.3083977594462e-05, + "loss": 3.1264, + "step": 14772 + }, + { + "epoch": 0.1626534129686253, + "grad_norm": 0.15725082159042358, + "learning_rate": 3.3081430235803334e-05, + "loss": 3.1361, + "step": 14773 + }, + { + "epoch": 0.1626842594341987, + "grad_norm": 0.1524428278207779, + "learning_rate": 3.307888278344535e-05, + "loss": 3.0955, + "step": 14774 + }, + { + "epoch": 0.16271510589977212, + "grad_norm": 0.150509312748909, + "learning_rate": 3.307633523741756e-05, + "loss": 3.1605, + "step": 14775 + }, + { + "epoch": 0.16274595236534553, + "grad_norm": 0.14813624322414398, + "learning_rate": 3.307378759774952e-05, + "loss": 3.1054, + "step": 14776 + }, + { + "epoch": 0.16277679883091895, + "grad_norm": 0.15841910243034363, + "learning_rate": 3.307123986447077e-05, + "loss": 3.1364, + "step": 14777 + }, + { + "epoch": 0.16280764529649236, + "grad_norm": 0.1519535928964615, + "learning_rate": 3.306869203761084e-05, + "loss": 3.135, + "step": 14778 + }, + { + "epoch": 0.1628384917620658, + "grad_norm": 0.14739510416984558, + "learning_rate": 3.306614411719928e-05, + "loss": 3.0718, + "step": 14779 + }, + { + "epoch": 0.16286933822763922, + "grad_norm": 0.1531229466199875, + "learning_rate": 3.306359610326562e-05, + "loss": 3.1062, + "step": 14780 + }, + { + "epoch": 0.16290018469321263, + "grad_norm": 0.15691670775413513, + "learning_rate": 3.3061047995839414e-05, + "loss": 3.0882, + "step": 14781 + }, + { + "epoch": 0.16293103115878604, + "grad_norm": 0.15308327972888947, + "learning_rate": 3.305849979495019e-05, + "loss": 3.1051, + "step": 14782 + }, + { + "epoch": 0.16296187762435946, + "grad_norm": 0.15308451652526855, + "learning_rate": 3.3055951500627517e-05, + "loss": 3.1196, + "step": 14783 + }, + { + "epoch": 0.16299272408993287, + "grad_norm": 0.15700696408748627, + "learning_rate": 3.305340311290092e-05, + "loss": 3.094, + "step": 14784 + }, + { + "epoch": 0.1630235705555063, + "grad_norm": 0.15101468563079834, + "learning_rate": 3.305085463179996e-05, + "loss": 3.1107, + "step": 14785 + }, + { + "epoch": 0.1630544170210797, + "grad_norm": 0.15933577716350555, + "learning_rate": 3.304830605735417e-05, + "loss": 3.1135, + "step": 14786 + }, + { + "epoch": 0.16308526348665311, + "grad_norm": 0.16211968660354614, + "learning_rate": 3.304575738959311e-05, + "loss": 3.1206, + "step": 14787 + }, + { + "epoch": 0.16311610995222653, + "grad_norm": 0.16014011204242706, + "learning_rate": 3.304320862854633e-05, + "loss": 3.1467, + "step": 14788 + }, + { + "epoch": 0.16314695641779994, + "grad_norm": 0.15460239350795746, + "learning_rate": 3.304065977424338e-05, + "loss": 3.1122, + "step": 14789 + }, + { + "epoch": 0.16317780288337336, + "grad_norm": 0.15650016069412231, + "learning_rate": 3.303811082671381e-05, + "loss": 3.1215, + "step": 14790 + }, + { + "epoch": 0.1632086493489468, + "grad_norm": 0.15184137225151062, + "learning_rate": 3.3035561785987174e-05, + "loss": 3.1148, + "step": 14791 + }, + { + "epoch": 0.1632394958145202, + "grad_norm": 0.15594740211963654, + "learning_rate": 3.3033012652093024e-05, + "loss": 3.1239, + "step": 14792 + }, + { + "epoch": 0.16327034228009363, + "grad_norm": 0.15204259753227234, + "learning_rate": 3.3030463425060934e-05, + "loss": 3.1234, + "step": 14793 + }, + { + "epoch": 0.16330118874566704, + "grad_norm": 0.1636558324098587, + "learning_rate": 3.3027914104920434e-05, + "loss": 3.1141, + "step": 14794 + }, + { + "epoch": 0.16333203521124046, + "grad_norm": 0.15233373641967773, + "learning_rate": 3.3025364691701096e-05, + "loss": 3.121, + "step": 14795 + }, + { + "epoch": 0.16336288167681387, + "grad_norm": 0.15805377066135406, + "learning_rate": 3.302281518543248e-05, + "loss": 3.1063, + "step": 14796 + }, + { + "epoch": 0.16339372814238728, + "grad_norm": 0.15034246444702148, + "learning_rate": 3.302026558614414e-05, + "loss": 3.1299, + "step": 14797 + }, + { + "epoch": 0.1634245746079607, + "grad_norm": 0.15509100258350372, + "learning_rate": 3.3017715893865645e-05, + "loss": 3.1325, + "step": 14798 + }, + { + "epoch": 0.1634554210735341, + "grad_norm": 0.1481751799583435, + "learning_rate": 3.301516610862655e-05, + "loss": 3.1106, + "step": 14799 + }, + { + "epoch": 0.16348626753910753, + "grad_norm": 0.15452344715595245, + "learning_rate": 3.3012616230456415e-05, + "loss": 3.1331, + "step": 14800 + }, + { + "epoch": 0.16351711400468094, + "grad_norm": 0.1541140228509903, + "learning_rate": 3.301006625938482e-05, + "loss": 3.1169, + "step": 14801 + }, + { + "epoch": 0.16354796047025436, + "grad_norm": 0.17738422751426697, + "learning_rate": 3.3007516195441315e-05, + "loss": 3.1363, + "step": 14802 + }, + { + "epoch": 0.1635788069358278, + "grad_norm": 0.1629900485277176, + "learning_rate": 3.300496603865547e-05, + "loss": 3.1133, + "step": 14803 + }, + { + "epoch": 0.1636096534014012, + "grad_norm": 0.14862899482250214, + "learning_rate": 3.300241578905685e-05, + "loss": 3.1254, + "step": 14804 + }, + { + "epoch": 0.16364049986697463, + "grad_norm": 0.16247688233852386, + "learning_rate": 3.299986544667504e-05, + "loss": 3.0904, + "step": 14805 + }, + { + "epoch": 0.16367134633254804, + "grad_norm": 0.1589481085538864, + "learning_rate": 3.299731501153959e-05, + "loss": 3.1057, + "step": 14806 + }, + { + "epoch": 0.16370219279812145, + "grad_norm": 0.1538524329662323, + "learning_rate": 3.299476448368008e-05, + "loss": 3.1038, + "step": 14807 + }, + { + "epoch": 0.16373303926369487, + "grad_norm": 0.1590852588415146, + "learning_rate": 3.299221386312607e-05, + "loss": 3.1222, + "step": 14808 + }, + { + "epoch": 0.16376388572926828, + "grad_norm": 0.15204288065433502, + "learning_rate": 3.298966314990716e-05, + "loss": 3.1364, + "step": 14809 + }, + { + "epoch": 0.1637947321948417, + "grad_norm": 0.1539229452610016, + "learning_rate": 3.29871123440529e-05, + "loss": 3.1204, + "step": 14810 + }, + { + "epoch": 0.1638255786604151, + "grad_norm": 0.23836536705493927, + "learning_rate": 3.2984561445592865e-05, + "loss": 3.1378, + "step": 14811 + }, + { + "epoch": 0.16385642512598853, + "grad_norm": 0.15898443758487701, + "learning_rate": 3.298201045455665e-05, + "loss": 3.1308, + "step": 14812 + }, + { + "epoch": 0.16388727159156194, + "grad_norm": 0.15714168548583984, + "learning_rate": 3.2979459370973806e-05, + "loss": 3.0879, + "step": 14813 + }, + { + "epoch": 0.16391811805713535, + "grad_norm": 0.14891541004180908, + "learning_rate": 3.297690819487394e-05, + "loss": 3.1159, + "step": 14814 + }, + { + "epoch": 0.1639489645227088, + "grad_norm": 0.16491873562335968, + "learning_rate": 3.2974356926286615e-05, + "loss": 3.1583, + "step": 14815 + }, + { + "epoch": 0.1639798109882822, + "grad_norm": 0.16368615627288818, + "learning_rate": 3.297180556524141e-05, + "loss": 3.1586, + "step": 14816 + }, + { + "epoch": 0.16401065745385562, + "grad_norm": 0.14842231571674347, + "learning_rate": 3.296925411176791e-05, + "loss": 3.095, + "step": 14817 + }, + { + "epoch": 0.16404150391942904, + "grad_norm": 0.16669994592666626, + "learning_rate": 3.29667025658957e-05, + "loss": 3.1258, + "step": 14818 + }, + { + "epoch": 0.16407235038500245, + "grad_norm": 0.15475738048553467, + "learning_rate": 3.296415092765436e-05, + "loss": 3.1246, + "step": 14819 + }, + { + "epoch": 0.16410319685057587, + "grad_norm": 0.1557496190071106, + "learning_rate": 3.296159919707347e-05, + "loss": 3.1276, + "step": 14820 + }, + { + "epoch": 0.16413404331614928, + "grad_norm": 0.16347385942935944, + "learning_rate": 3.295904737418263e-05, + "loss": 3.088, + "step": 14821 + }, + { + "epoch": 0.1641648897817227, + "grad_norm": 0.1538783609867096, + "learning_rate": 3.295649545901142e-05, + "loss": 3.0945, + "step": 14822 + }, + { + "epoch": 0.1641957362472961, + "grad_norm": 0.16015231609344482, + "learning_rate": 3.2953943451589416e-05, + "loss": 3.1219, + "step": 14823 + }, + { + "epoch": 0.16422658271286952, + "grad_norm": 0.15822923183441162, + "learning_rate": 3.295139135194623e-05, + "loss": 3.1483, + "step": 14824 + }, + { + "epoch": 0.16425742917844294, + "grad_norm": 0.1554798036813736, + "learning_rate": 3.2948839160111425e-05, + "loss": 3.088, + "step": 14825 + }, + { + "epoch": 0.16428827564401635, + "grad_norm": 0.1611587256193161, + "learning_rate": 3.2946286876114605e-05, + "loss": 3.1342, + "step": 14826 + }, + { + "epoch": 0.1643191221095898, + "grad_norm": 0.1555228978395462, + "learning_rate": 3.294373449998538e-05, + "loss": 3.0895, + "step": 14827 + }, + { + "epoch": 0.1643499685751632, + "grad_norm": 0.15429936349391937, + "learning_rate": 3.294118203175333e-05, + "loss": 3.1546, + "step": 14828 + }, + { + "epoch": 0.16438081504073662, + "grad_norm": 0.16934113204479218, + "learning_rate": 3.293862947144803e-05, + "loss": 3.1362, + "step": 14829 + }, + { + "epoch": 0.16441166150631004, + "grad_norm": 0.15988236665725708, + "learning_rate": 3.2936076819099094e-05, + "loss": 3.0913, + "step": 14830 + }, + { + "epoch": 0.16444250797188345, + "grad_norm": 0.16625551879405975, + "learning_rate": 3.293352407473612e-05, + "loss": 3.1437, + "step": 14831 + }, + { + "epoch": 0.16447335443745686, + "grad_norm": 0.1565983146429062, + "learning_rate": 3.29309712383887e-05, + "loss": 3.135, + "step": 14832 + }, + { + "epoch": 0.16450420090303028, + "grad_norm": 0.1484745740890503, + "learning_rate": 3.292841831008644e-05, + "loss": 3.1439, + "step": 14833 + }, + { + "epoch": 0.1645350473686037, + "grad_norm": 0.1578214466571808, + "learning_rate": 3.292586528985894e-05, + "loss": 3.1379, + "step": 14834 + }, + { + "epoch": 0.1645658938341771, + "grad_norm": 0.17382119596004486, + "learning_rate": 3.292331217773578e-05, + "loss": 3.1343, + "step": 14835 + }, + { + "epoch": 0.16459674029975052, + "grad_norm": 0.15809695422649384, + "learning_rate": 3.292075897374658e-05, + "loss": 3.1114, + "step": 14836 + }, + { + "epoch": 0.16462758676532394, + "grad_norm": 0.15391957759857178, + "learning_rate": 3.291820567792094e-05, + "loss": 3.1309, + "step": 14837 + }, + { + "epoch": 0.16465843323089735, + "grad_norm": 0.15918394923210144, + "learning_rate": 3.2915652290288467e-05, + "loss": 3.1485, + "step": 14838 + }, + { + "epoch": 0.1646892796964708, + "grad_norm": 0.15220724046230316, + "learning_rate": 3.2913098810878765e-05, + "loss": 3.1248, + "step": 14839 + }, + { + "epoch": 0.1647201261620442, + "grad_norm": 0.16145971417427063, + "learning_rate": 3.2910545239721436e-05, + "loss": 3.1047, + "step": 14840 + }, + { + "epoch": 0.16475097262761762, + "grad_norm": 0.16136138141155243, + "learning_rate": 3.290799157684608e-05, + "loss": 3.1096, + "step": 14841 + }, + { + "epoch": 0.16478181909319103, + "grad_norm": 0.1575385481119156, + "learning_rate": 3.290543782228232e-05, + "loss": 3.1028, + "step": 14842 + }, + { + "epoch": 0.16481266555876445, + "grad_norm": 0.15673041343688965, + "learning_rate": 3.290288397605976e-05, + "loss": 3.08, + "step": 14843 + }, + { + "epoch": 0.16484351202433786, + "grad_norm": 0.16279789805412292, + "learning_rate": 3.2900330038208006e-05, + "loss": 3.1599, + "step": 14844 + }, + { + "epoch": 0.16487435848991128, + "grad_norm": 0.16048194468021393, + "learning_rate": 3.2897776008756684e-05, + "loss": 3.1619, + "step": 14845 + }, + { + "epoch": 0.1649052049554847, + "grad_norm": 0.16092319786548615, + "learning_rate": 3.2895221887735375e-05, + "loss": 3.1505, + "step": 14846 + }, + { + "epoch": 0.1649360514210581, + "grad_norm": 0.16614411771297455, + "learning_rate": 3.289266767517373e-05, + "loss": 3.1174, + "step": 14847 + }, + { + "epoch": 0.16496689788663152, + "grad_norm": 0.1556704193353653, + "learning_rate": 3.289011337110134e-05, + "loss": 3.0967, + "step": 14848 + }, + { + "epoch": 0.16499774435220493, + "grad_norm": 0.15822647511959076, + "learning_rate": 3.288755897554783e-05, + "loss": 3.1722, + "step": 14849 + }, + { + "epoch": 0.16502859081777835, + "grad_norm": 0.16317680478096008, + "learning_rate": 3.2885004488542805e-05, + "loss": 3.121, + "step": 14850 + }, + { + "epoch": 0.1650594372833518, + "grad_norm": 0.14622808992862701, + "learning_rate": 3.28824499101159e-05, + "loss": 3.1027, + "step": 14851 + }, + { + "epoch": 0.1650902837489252, + "grad_norm": 0.16147704422473907, + "learning_rate": 3.287989524029672e-05, + "loss": 3.1459, + "step": 14852 + }, + { + "epoch": 0.16512113021449862, + "grad_norm": 0.15283912420272827, + "learning_rate": 3.287734047911489e-05, + "loss": 3.13, + "step": 14853 + }, + { + "epoch": 0.16515197668007203, + "grad_norm": 0.1522965431213379, + "learning_rate": 3.2874785626600034e-05, + "loss": 3.1242, + "step": 14854 + }, + { + "epoch": 0.16518282314564545, + "grad_norm": 0.15522846579551697, + "learning_rate": 3.2872230682781775e-05, + "loss": 3.1182, + "step": 14855 + }, + { + "epoch": 0.16521366961121886, + "grad_norm": 0.15701603889465332, + "learning_rate": 3.2869675647689726e-05, + "loss": 3.1025, + "step": 14856 + }, + { + "epoch": 0.16524451607679227, + "grad_norm": 0.14974001049995422, + "learning_rate": 3.286712052135352e-05, + "loss": 3.1243, + "step": 14857 + }, + { + "epoch": 0.1652753625423657, + "grad_norm": 0.1616886854171753, + "learning_rate": 3.2864565303802774e-05, + "loss": 3.1431, + "step": 14858 + }, + { + "epoch": 0.1653062090079391, + "grad_norm": 0.15290534496307373, + "learning_rate": 3.2862009995067116e-05, + "loss": 3.1551, + "step": 14859 + }, + { + "epoch": 0.16533705547351252, + "grad_norm": 0.15451689064502716, + "learning_rate": 3.2859454595176185e-05, + "loss": 3.1346, + "step": 14860 + }, + { + "epoch": 0.16536790193908593, + "grad_norm": 0.16412554681301117, + "learning_rate": 3.28568991041596e-05, + "loss": 3.1169, + "step": 14861 + }, + { + "epoch": 0.16539874840465935, + "grad_norm": 0.15382970869541168, + "learning_rate": 3.285434352204699e-05, + "loss": 3.1419, + "step": 14862 + }, + { + "epoch": 0.1654295948702328, + "grad_norm": 0.16754958033561707, + "learning_rate": 3.285178784886799e-05, + "loss": 3.1236, + "step": 14863 + }, + { + "epoch": 0.1654604413358062, + "grad_norm": 0.15199105441570282, + "learning_rate": 3.284923208465221e-05, + "loss": 3.1264, + "step": 14864 + }, + { + "epoch": 0.16549128780137962, + "grad_norm": 0.15444821119308472, + "learning_rate": 3.284667622942933e-05, + "loss": 3.1354, + "step": 14865 + }, + { + "epoch": 0.16552213426695303, + "grad_norm": 0.15312707424163818, + "learning_rate": 3.284412028322894e-05, + "loss": 3.0721, + "step": 14866 + }, + { + "epoch": 0.16555298073252644, + "grad_norm": 0.21701239049434662, + "learning_rate": 3.2841564246080685e-05, + "loss": 3.101, + "step": 14867 + }, + { + "epoch": 0.16558382719809986, + "grad_norm": 0.15800559520721436, + "learning_rate": 3.283900811801421e-05, + "loss": 3.0955, + "step": 14868 + }, + { + "epoch": 0.16561467366367327, + "grad_norm": 0.21737802028656006, + "learning_rate": 3.283645189905914e-05, + "loss": 3.1211, + "step": 14869 + }, + { + "epoch": 0.1656455201292467, + "grad_norm": 0.15973088145256042, + "learning_rate": 3.283389558924513e-05, + "loss": 3.1359, + "step": 14870 + }, + { + "epoch": 0.1656763665948201, + "grad_norm": 0.15362997353076935, + "learning_rate": 3.2831339188601805e-05, + "loss": 3.1191, + "step": 14871 + }, + { + "epoch": 0.16570721306039352, + "grad_norm": 0.15464642643928528, + "learning_rate": 3.282878269715881e-05, + "loss": 3.1464, + "step": 14872 + }, + { + "epoch": 0.16573805952596693, + "grad_norm": 0.16391493380069733, + "learning_rate": 3.282622611494578e-05, + "loss": 3.1423, + "step": 14873 + }, + { + "epoch": 0.16576890599154034, + "grad_norm": 0.15121707320213318, + "learning_rate": 3.282366944199236e-05, + "loss": 3.0821, + "step": 14874 + }, + { + "epoch": 0.16579975245711379, + "grad_norm": 0.15134692192077637, + "learning_rate": 3.28211126783282e-05, + "loss": 3.1102, + "step": 14875 + }, + { + "epoch": 0.1658305989226872, + "grad_norm": 0.15035924315452576, + "learning_rate": 3.2818555823982926e-05, + "loss": 3.0803, + "step": 14876 + }, + { + "epoch": 0.16586144538826061, + "grad_norm": 0.15532682836055756, + "learning_rate": 3.281599887898621e-05, + "loss": 3.0899, + "step": 14877 + }, + { + "epoch": 0.16589229185383403, + "grad_norm": 0.15939143300056458, + "learning_rate": 3.281344184336768e-05, + "loss": 3.1181, + "step": 14878 + }, + { + "epoch": 0.16592313831940744, + "grad_norm": 0.15784470736980438, + "learning_rate": 3.281088471715699e-05, + "loss": 3.1161, + "step": 14879 + }, + { + "epoch": 0.16595398478498086, + "grad_norm": 0.14830446243286133, + "learning_rate": 3.2808327500383775e-05, + "loss": 3.0949, + "step": 14880 + }, + { + "epoch": 0.16598483125055427, + "grad_norm": 0.1542525589466095, + "learning_rate": 3.28057701930777e-05, + "loss": 3.1299, + "step": 14881 + }, + { + "epoch": 0.16601567771612769, + "grad_norm": 0.1486072540283203, + "learning_rate": 3.280321279526841e-05, + "loss": 3.1185, + "step": 14882 + }, + { + "epoch": 0.1660465241817011, + "grad_norm": 0.158078134059906, + "learning_rate": 3.280065530698556e-05, + "loss": 3.1028, + "step": 14883 + }, + { + "epoch": 0.1660773706472745, + "grad_norm": 0.16008710861206055, + "learning_rate": 3.27980977282588e-05, + "loss": 3.1393, + "step": 14884 + }, + { + "epoch": 0.16610821711284793, + "grad_norm": 0.14585094153881073, + "learning_rate": 3.279554005911778e-05, + "loss": 3.0744, + "step": 14885 + }, + { + "epoch": 0.16613906357842134, + "grad_norm": 0.15186119079589844, + "learning_rate": 3.279298229959215e-05, + "loss": 3.1241, + "step": 14886 + }, + { + "epoch": 0.16616991004399478, + "grad_norm": 0.15650390088558197, + "learning_rate": 3.279042444971159e-05, + "loss": 3.1217, + "step": 14887 + }, + { + "epoch": 0.1662007565095682, + "grad_norm": 0.15355144441127777, + "learning_rate": 3.2787866509505726e-05, + "loss": 3.11, + "step": 14888 + }, + { + "epoch": 0.1662316029751416, + "grad_norm": 0.16014941036701202, + "learning_rate": 3.278530847900424e-05, + "loss": 3.0919, + "step": 14889 + }, + { + "epoch": 0.16626244944071503, + "grad_norm": 0.14858378469944, + "learning_rate": 3.278275035823677e-05, + "loss": 3.1205, + "step": 14890 + }, + { + "epoch": 0.16629329590628844, + "grad_norm": 0.15217430889606476, + "learning_rate": 3.2780192147232995e-05, + "loss": 3.1276, + "step": 14891 + }, + { + "epoch": 0.16632414237186186, + "grad_norm": 0.1487867683172226, + "learning_rate": 3.277763384602256e-05, + "loss": 3.1198, + "step": 14892 + }, + { + "epoch": 0.16635498883743527, + "grad_norm": 0.15392231941223145, + "learning_rate": 3.277507545463514e-05, + "loss": 3.097, + "step": 14893 + }, + { + "epoch": 0.16638583530300868, + "grad_norm": 0.14804378151893616, + "learning_rate": 3.2772516973100393e-05, + "loss": 3.094, + "step": 14894 + }, + { + "epoch": 0.1664166817685821, + "grad_norm": 0.1549673080444336, + "learning_rate": 3.2769958401447984e-05, + "loss": 3.1289, + "step": 14895 + }, + { + "epoch": 0.1664475282341555, + "grad_norm": 0.15265421569347382, + "learning_rate": 3.276739973970757e-05, + "loss": 3.1053, + "step": 14896 + }, + { + "epoch": 0.16647837469972893, + "grad_norm": 0.15940704941749573, + "learning_rate": 3.276484098790882e-05, + "loss": 3.1408, + "step": 14897 + }, + { + "epoch": 0.16650922116530234, + "grad_norm": 0.15028351545333862, + "learning_rate": 3.276228214608142e-05, + "loss": 3.1043, + "step": 14898 + }, + { + "epoch": 0.16654006763087578, + "grad_norm": 0.367511510848999, + "learning_rate": 3.275972321425502e-05, + "loss": 3.1513, + "step": 14899 + }, + { + "epoch": 0.1665709140964492, + "grad_norm": 0.15301071107387543, + "learning_rate": 3.2757164192459276e-05, + "loss": 3.14, + "step": 14900 + }, + { + "epoch": 0.1666017605620226, + "grad_norm": 0.15762586891651154, + "learning_rate": 3.2754605080723893e-05, + "loss": 3.1258, + "step": 14901 + }, + { + "epoch": 0.16663260702759602, + "grad_norm": 0.15783847868442535, + "learning_rate": 3.275204587907851e-05, + "loss": 3.1278, + "step": 14902 + }, + { + "epoch": 0.16666345349316944, + "grad_norm": 0.15790584683418274, + "learning_rate": 3.2749486587552824e-05, + "loss": 3.0771, + "step": 14903 + }, + { + "epoch": 0.16669429995874285, + "grad_norm": 0.15109769999980927, + "learning_rate": 3.274692720617649e-05, + "loss": 3.1454, + "step": 14904 + }, + { + "epoch": 0.16672514642431627, + "grad_norm": 0.1594090610742569, + "learning_rate": 3.2744367734979203e-05, + "loss": 3.1388, + "step": 14905 + }, + { + "epoch": 0.16675599288988968, + "grad_norm": 0.15712691843509674, + "learning_rate": 3.274180817399062e-05, + "loss": 3.1274, + "step": 14906 + }, + { + "epoch": 0.1667868393554631, + "grad_norm": 0.15476509928703308, + "learning_rate": 3.273924852324042e-05, + "loss": 3.1107, + "step": 14907 + }, + { + "epoch": 0.1668176858210365, + "grad_norm": 0.1518135815858841, + "learning_rate": 3.273668878275827e-05, + "loss": 3.1221, + "step": 14908 + }, + { + "epoch": 0.16684853228660992, + "grad_norm": 0.15330660343170166, + "learning_rate": 3.273412895257388e-05, + "loss": 3.0951, + "step": 14909 + }, + { + "epoch": 0.16687937875218334, + "grad_norm": 0.15193626284599304, + "learning_rate": 3.273156903271692e-05, + "loss": 3.1471, + "step": 14910 + }, + { + "epoch": 0.16691022521775678, + "grad_norm": 0.16432659327983856, + "learning_rate": 3.272900902321705e-05, + "loss": 3.0981, + "step": 14911 + }, + { + "epoch": 0.1669410716833302, + "grad_norm": 0.16357199847698212, + "learning_rate": 3.272644892410398e-05, + "loss": 3.1439, + "step": 14912 + }, + { + "epoch": 0.1669719181489036, + "grad_norm": 0.14933915436267853, + "learning_rate": 3.272388873540736e-05, + "loss": 3.1091, + "step": 14913 + }, + { + "epoch": 0.16700276461447702, + "grad_norm": 0.14654700458049774, + "learning_rate": 3.272132845715689e-05, + "loss": 3.1183, + "step": 14914 + }, + { + "epoch": 0.16703361108005044, + "grad_norm": 0.15516497194766998, + "learning_rate": 3.2718768089382266e-05, + "loss": 3.121, + "step": 14915 + }, + { + "epoch": 0.16706445754562385, + "grad_norm": 1.3575286865234375, + "learning_rate": 3.271620763211317e-05, + "loss": 3.1303, + "step": 14916 + }, + { + "epoch": 0.16709530401119727, + "grad_norm": 0.1519567221403122, + "learning_rate": 3.2713647085379274e-05, + "loss": 3.1463, + "step": 14917 + }, + { + "epoch": 0.16712615047677068, + "grad_norm": 0.15840773284435272, + "learning_rate": 3.2711086449210284e-05, + "loss": 3.1304, + "step": 14918 + }, + { + "epoch": 0.1671569969423441, + "grad_norm": 0.14823496341705322, + "learning_rate": 3.270852572363587e-05, + "loss": 3.1045, + "step": 14919 + }, + { + "epoch": 0.1671878434079175, + "grad_norm": 0.15626294910907745, + "learning_rate": 3.2705964908685734e-05, + "loss": 3.1132, + "step": 14920 + }, + { + "epoch": 0.16721868987349092, + "grad_norm": 0.15253199636936188, + "learning_rate": 3.270340400438957e-05, + "loss": 3.1271, + "step": 14921 + }, + { + "epoch": 0.16724953633906434, + "grad_norm": 0.15213419497013092, + "learning_rate": 3.2700843010777066e-05, + "loss": 3.1254, + "step": 14922 + }, + { + "epoch": 0.16728038280463778, + "grad_norm": 0.1543392688035965, + "learning_rate": 3.269828192787791e-05, + "loss": 3.1168, + "step": 14923 + }, + { + "epoch": 0.1673112292702112, + "grad_norm": 0.16275081038475037, + "learning_rate": 3.269572075572181e-05, + "loss": 3.1114, + "step": 14924 + }, + { + "epoch": 0.1673420757357846, + "grad_norm": 0.1505821794271469, + "learning_rate": 3.2693159494338445e-05, + "loss": 3.1139, + "step": 14925 + }, + { + "epoch": 0.16737292220135802, + "grad_norm": 0.15374498069286346, + "learning_rate": 3.269059814375752e-05, + "loss": 3.111, + "step": 14926 + }, + { + "epoch": 0.16740376866693144, + "grad_norm": 0.15231335163116455, + "learning_rate": 3.268803670400873e-05, + "loss": 3.1318, + "step": 14927 + }, + { + "epoch": 0.16743461513250485, + "grad_norm": 0.15207883715629578, + "learning_rate": 3.268547517512178e-05, + "loss": 3.0874, + "step": 14928 + }, + { + "epoch": 0.16746546159807826, + "grad_norm": 0.15504981577396393, + "learning_rate": 3.268291355712636e-05, + "loss": 3.1009, + "step": 14929 + }, + { + "epoch": 0.16749630806365168, + "grad_norm": 0.15609215199947357, + "learning_rate": 3.268035185005217e-05, + "loss": 3.1088, + "step": 14930 + }, + { + "epoch": 0.1675271545292251, + "grad_norm": 0.152186319231987, + "learning_rate": 3.267779005392892e-05, + "loss": 3.1417, + "step": 14931 + }, + { + "epoch": 0.1675580009947985, + "grad_norm": 0.15757328271865845, + "learning_rate": 3.267522816878631e-05, + "loss": 3.1556, + "step": 14932 + }, + { + "epoch": 0.16758884746037192, + "grad_norm": 0.1528790295124054, + "learning_rate": 3.2672666194654036e-05, + "loss": 3.1387, + "step": 14933 + }, + { + "epoch": 0.16761969392594533, + "grad_norm": 0.16147634387016296, + "learning_rate": 3.2670104131561816e-05, + "loss": 3.1118, + "step": 14934 + }, + { + "epoch": 0.16765054039151878, + "grad_norm": 0.14994822442531586, + "learning_rate": 3.266754197953934e-05, + "loss": 3.1221, + "step": 14935 + }, + { + "epoch": 0.1676813868570922, + "grad_norm": 0.16126517951488495, + "learning_rate": 3.2664979738616324e-05, + "loss": 3.124, + "step": 14936 + }, + { + "epoch": 0.1677122333226656, + "grad_norm": 0.16621540486812592, + "learning_rate": 3.266241740882248e-05, + "loss": 3.1342, + "step": 14937 + }, + { + "epoch": 0.16774307978823902, + "grad_norm": 0.16323767602443695, + "learning_rate": 3.26598549901875e-05, + "loss": 3.1095, + "step": 14938 + }, + { + "epoch": 0.16777392625381243, + "grad_norm": 0.15525884926319122, + "learning_rate": 3.265729248274112e-05, + "loss": 3.1111, + "step": 14939 + }, + { + "epoch": 0.16780477271938585, + "grad_norm": 0.15120425820350647, + "learning_rate": 3.265472988651303e-05, + "loss": 3.103, + "step": 14940 + }, + { + "epoch": 0.16783561918495926, + "grad_norm": 0.15561345219612122, + "learning_rate": 3.265216720153293e-05, + "loss": 3.1177, + "step": 14941 + }, + { + "epoch": 0.16786646565053268, + "grad_norm": 0.1540018767118454, + "learning_rate": 3.2649604427830566e-05, + "loss": 3.1019, + "step": 14942 + }, + { + "epoch": 0.1678973121161061, + "grad_norm": 0.15111514925956726, + "learning_rate": 3.264704156543564e-05, + "loss": 3.1202, + "step": 14943 + }, + { + "epoch": 0.1679281585816795, + "grad_norm": 0.1555365025997162, + "learning_rate": 3.264447861437785e-05, + "loss": 3.149, + "step": 14944 + }, + { + "epoch": 0.16795900504725292, + "grad_norm": 0.2029934823513031, + "learning_rate": 3.264191557468693e-05, + "loss": 3.064, + "step": 14945 + }, + { + "epoch": 0.16798985151282633, + "grad_norm": 0.1536126285791397, + "learning_rate": 3.263935244639259e-05, + "loss": 3.1281, + "step": 14946 + }, + { + "epoch": 0.16802069797839977, + "grad_norm": 0.15261726081371307, + "learning_rate": 3.2636789229524545e-05, + "loss": 3.1054, + "step": 14947 + }, + { + "epoch": 0.1680515444439732, + "grad_norm": 0.1549898386001587, + "learning_rate": 3.263422592411253e-05, + "loss": 3.1254, + "step": 14948 + }, + { + "epoch": 0.1680823909095466, + "grad_norm": 0.15322300791740417, + "learning_rate": 3.2631662530186244e-05, + "loss": 3.1162, + "step": 14949 + }, + { + "epoch": 0.16811323737512002, + "grad_norm": 0.16250132024288177, + "learning_rate": 3.2629099047775416e-05, + "loss": 3.0826, + "step": 14950 + }, + { + "epoch": 0.16814408384069343, + "grad_norm": 0.1512865275144577, + "learning_rate": 3.262653547690977e-05, + "loss": 3.1534, + "step": 14951 + }, + { + "epoch": 0.16817493030626685, + "grad_norm": 0.16975459456443787, + "learning_rate": 3.262397181761903e-05, + "loss": 3.1641, + "step": 14952 + }, + { + "epoch": 0.16820577677184026, + "grad_norm": 0.16341419517993927, + "learning_rate": 3.262140806993291e-05, + "loss": 3.1339, + "step": 14953 + }, + { + "epoch": 0.16823662323741367, + "grad_norm": 0.15432099997997284, + "learning_rate": 3.2618844233881166e-05, + "loss": 3.1693, + "step": 14954 + }, + { + "epoch": 0.1682674697029871, + "grad_norm": 0.15562650561332703, + "learning_rate": 3.2616280309493486e-05, + "loss": 3.0877, + "step": 14955 + }, + { + "epoch": 0.1682983161685605, + "grad_norm": 0.15102142095565796, + "learning_rate": 3.2613716296799606e-05, + "loss": 3.1552, + "step": 14956 + }, + { + "epoch": 0.16832916263413392, + "grad_norm": 0.1660197675228119, + "learning_rate": 3.261115219582927e-05, + "loss": 3.1383, + "step": 14957 + }, + { + "epoch": 0.16836000909970733, + "grad_norm": 0.15697211027145386, + "learning_rate": 3.2608588006612195e-05, + "loss": 3.1338, + "step": 14958 + }, + { + "epoch": 0.16839085556528077, + "grad_norm": 0.1522553414106369, + "learning_rate": 3.260602372917811e-05, + "loss": 3.0993, + "step": 14959 + }, + { + "epoch": 0.1684217020308542, + "grad_norm": 0.15018559992313385, + "learning_rate": 3.260345936355677e-05, + "loss": 3.1247, + "step": 14960 + }, + { + "epoch": 0.1684525484964276, + "grad_norm": 0.16380566358566284, + "learning_rate": 3.260089490977787e-05, + "loss": 3.1556, + "step": 14961 + }, + { + "epoch": 0.16848339496200102, + "grad_norm": 0.1492166966199875, + "learning_rate": 3.2598330367871164e-05, + "loss": 3.1066, + "step": 14962 + }, + { + "epoch": 0.16851424142757443, + "grad_norm": 0.16046299040317535, + "learning_rate": 3.259576573786638e-05, + "loss": 3.126, + "step": 14963 + }, + { + "epoch": 0.16854508789314784, + "grad_norm": 0.19902296364307404, + "learning_rate": 3.2593201019793264e-05, + "loss": 3.0999, + "step": 14964 + }, + { + "epoch": 0.16857593435872126, + "grad_norm": 0.15725335478782654, + "learning_rate": 3.259063621368154e-05, + "loss": 3.1295, + "step": 14965 + }, + { + "epoch": 0.16860678082429467, + "grad_norm": 0.14861728250980377, + "learning_rate": 3.258807131956096e-05, + "loss": 3.0931, + "step": 14966 + }, + { + "epoch": 0.1686376272898681, + "grad_norm": 0.15077851712703705, + "learning_rate": 3.2585506337461245e-05, + "loss": 3.0832, + "step": 14967 + }, + { + "epoch": 0.1686684737554415, + "grad_norm": 0.1545781046152115, + "learning_rate": 3.2582941267412145e-05, + "loss": 3.1473, + "step": 14968 + }, + { + "epoch": 0.16869932022101491, + "grad_norm": 0.15227089822292328, + "learning_rate": 3.258037610944339e-05, + "loss": 3.1292, + "step": 14969 + }, + { + "epoch": 0.16873016668658833, + "grad_norm": 0.15088066458702087, + "learning_rate": 3.2577810863584736e-05, + "loss": 3.1345, + "step": 14970 + }, + { + "epoch": 0.16876101315216177, + "grad_norm": 0.15379394590854645, + "learning_rate": 3.2575245529865915e-05, + "loss": 3.1233, + "step": 14971 + }, + { + "epoch": 0.16879185961773518, + "grad_norm": 0.15580005943775177, + "learning_rate": 3.257268010831668e-05, + "loss": 3.1354, + "step": 14972 + }, + { + "epoch": 0.1688227060833086, + "grad_norm": 0.19144542515277863, + "learning_rate": 3.257011459896677e-05, + "loss": 3.1027, + "step": 14973 + }, + { + "epoch": 0.168853552548882, + "grad_norm": 0.15034925937652588, + "learning_rate": 3.256754900184593e-05, + "loss": 3.1365, + "step": 14974 + }, + { + "epoch": 0.16888439901445543, + "grad_norm": 0.15501722693443298, + "learning_rate": 3.2564983316983904e-05, + "loss": 3.0954, + "step": 14975 + }, + { + "epoch": 0.16891524548002884, + "grad_norm": 0.1585722118616104, + "learning_rate": 3.256241754441044e-05, + "loss": 3.1117, + "step": 14976 + }, + { + "epoch": 0.16894609194560226, + "grad_norm": 0.15569907426834106, + "learning_rate": 3.2559851684155294e-05, + "loss": 3.129, + "step": 14977 + }, + { + "epoch": 0.16897693841117567, + "grad_norm": 0.15186361968517303, + "learning_rate": 3.2557285736248214e-05, + "loss": 3.11, + "step": 14978 + }, + { + "epoch": 0.16900778487674908, + "grad_norm": 0.16429193317890167, + "learning_rate": 3.2554719700718936e-05, + "loss": 3.0938, + "step": 14979 + }, + { + "epoch": 0.1690386313423225, + "grad_norm": 0.1464606672525406, + "learning_rate": 3.255215357759723e-05, + "loss": 3.1518, + "step": 14980 + }, + { + "epoch": 0.1690694778078959, + "grad_norm": 0.15896958112716675, + "learning_rate": 3.254958736691284e-05, + "loss": 3.125, + "step": 14981 + }, + { + "epoch": 0.16910032427346933, + "grad_norm": 0.15051916241645813, + "learning_rate": 3.254702106869552e-05, + "loss": 3.1611, + "step": 14982 + }, + { + "epoch": 0.16913117073904277, + "grad_norm": 0.15013828873634338, + "learning_rate": 3.254445468297503e-05, + "loss": 3.1669, + "step": 14983 + }, + { + "epoch": 0.16916201720461618, + "grad_norm": 0.14838837087154388, + "learning_rate": 3.254188820978112e-05, + "loss": 3.1039, + "step": 14984 + }, + { + "epoch": 0.1691928636701896, + "grad_norm": 0.15502455830574036, + "learning_rate": 3.253932164914355e-05, + "loss": 3.1243, + "step": 14985 + }, + { + "epoch": 0.169223710135763, + "grad_norm": 0.15225021541118622, + "learning_rate": 3.253675500109207e-05, + "loss": 3.1065, + "step": 14986 + }, + { + "epoch": 0.16925455660133643, + "grad_norm": 0.15069329738616943, + "learning_rate": 3.253418826565645e-05, + "loss": 3.1047, + "step": 14987 + }, + { + "epoch": 0.16928540306690984, + "grad_norm": 0.14966140687465668, + "learning_rate": 3.253162144286644e-05, + "loss": 3.1006, + "step": 14988 + }, + { + "epoch": 0.16931624953248325, + "grad_norm": 0.15528032183647156, + "learning_rate": 3.2529054532751815e-05, + "loss": 3.1229, + "step": 14989 + }, + { + "epoch": 0.16934709599805667, + "grad_norm": 0.14982563257217407, + "learning_rate": 3.252648753534232e-05, + "loss": 3.1205, + "step": 14990 + }, + { + "epoch": 0.16937794246363008, + "grad_norm": 0.151223286986351, + "learning_rate": 3.2523920450667725e-05, + "loss": 3.1186, + "step": 14991 + }, + { + "epoch": 0.1694087889292035, + "grad_norm": 0.15416420996189117, + "learning_rate": 3.252135327875779e-05, + "loss": 3.1185, + "step": 14992 + }, + { + "epoch": 0.1694396353947769, + "grad_norm": 0.1565183401107788, + "learning_rate": 3.2518786019642294e-05, + "loss": 3.1098, + "step": 14993 + }, + { + "epoch": 0.16947048186035032, + "grad_norm": 0.15283465385437012, + "learning_rate": 3.2516218673350984e-05, + "loss": 3.1078, + "step": 14994 + }, + { + "epoch": 0.16950132832592377, + "grad_norm": 0.15224997699260712, + "learning_rate": 3.251365123991364e-05, + "loss": 3.1134, + "step": 14995 + }, + { + "epoch": 0.16953217479149718, + "grad_norm": 0.16373807191848755, + "learning_rate": 3.251108371936002e-05, + "loss": 3.1294, + "step": 14996 + }, + { + "epoch": 0.1695630212570706, + "grad_norm": 0.14761602878570557, + "learning_rate": 3.25085161117199e-05, + "loss": 3.0971, + "step": 14997 + }, + { + "epoch": 0.169593867722644, + "grad_norm": 0.1516892910003662, + "learning_rate": 3.2505948417023046e-05, + "loss": 3.1496, + "step": 14998 + }, + { + "epoch": 0.16962471418821742, + "grad_norm": 0.15630382299423218, + "learning_rate": 3.2503380635299244e-05, + "loss": 3.0751, + "step": 14999 + }, + { + "epoch": 0.16965556065379084, + "grad_norm": 0.15486642718315125, + "learning_rate": 3.250081276657823e-05, + "loss": 3.1039, + "step": 15000 + }, + { + "epoch": 0.16968640711936425, + "grad_norm": 0.16510280966758728, + "learning_rate": 3.2498244810889815e-05, + "loss": 3.117, + "step": 15001 + }, + { + "epoch": 0.16971725358493767, + "grad_norm": 0.15035320818424225, + "learning_rate": 3.249567676826375e-05, + "loss": 3.1184, + "step": 15002 + }, + { + "epoch": 0.16974810005051108, + "grad_norm": 0.1500387340784073, + "learning_rate": 3.2493108638729816e-05, + "loss": 3.1079, + "step": 15003 + }, + { + "epoch": 0.1697789465160845, + "grad_norm": 0.16171978414058685, + "learning_rate": 3.2490540422317806e-05, + "loss": 3.0719, + "step": 15004 + }, + { + "epoch": 0.1698097929816579, + "grad_norm": 0.15388131141662598, + "learning_rate": 3.248797211905747e-05, + "loss": 3.1324, + "step": 15005 + }, + { + "epoch": 0.16984063944723135, + "grad_norm": 0.1548752784729004, + "learning_rate": 3.248540372897859e-05, + "loss": 3.1062, + "step": 15006 + }, + { + "epoch": 0.16987148591280476, + "grad_norm": 0.14878258109092712, + "learning_rate": 3.2482835252110964e-05, + "loss": 3.0843, + "step": 15007 + }, + { + "epoch": 0.16990233237837818, + "grad_norm": 0.15306450426578522, + "learning_rate": 3.248026668848435e-05, + "loss": 3.1317, + "step": 15008 + }, + { + "epoch": 0.1699331788439516, + "grad_norm": 0.1579195261001587, + "learning_rate": 3.247769803812855e-05, + "loss": 3.1117, + "step": 15009 + }, + { + "epoch": 0.169964025309525, + "grad_norm": 0.15657679736614227, + "learning_rate": 3.2475129301073324e-05, + "loss": 3.1349, + "step": 15010 + }, + { + "epoch": 0.16999487177509842, + "grad_norm": 0.1512119472026825, + "learning_rate": 3.247256047734847e-05, + "loss": 3.1218, + "step": 15011 + }, + { + "epoch": 0.17002571824067184, + "grad_norm": 0.14845424890518188, + "learning_rate": 3.246999156698377e-05, + "loss": 3.1033, + "step": 15012 + }, + { + "epoch": 0.17005656470624525, + "grad_norm": 0.15397436916828156, + "learning_rate": 3.2467422570009e-05, + "loss": 3.1261, + "step": 15013 + }, + { + "epoch": 0.17008741117181866, + "grad_norm": 0.15172722935676575, + "learning_rate": 3.2464853486453955e-05, + "loss": 3.1157, + "step": 15014 + }, + { + "epoch": 0.17011825763739208, + "grad_norm": 0.15071290731430054, + "learning_rate": 3.246228431634842e-05, + "loss": 3.15, + "step": 15015 + }, + { + "epoch": 0.1701491041029655, + "grad_norm": 0.14673464000225067, + "learning_rate": 3.24597150597222e-05, + "loss": 3.0732, + "step": 15016 + }, + { + "epoch": 0.1701799505685389, + "grad_norm": 0.15278032422065735, + "learning_rate": 3.245714571660504e-05, + "loss": 3.1252, + "step": 15017 + }, + { + "epoch": 0.17021079703411235, + "grad_norm": 0.14818407595157623, + "learning_rate": 3.245457628702677e-05, + "loss": 3.0992, + "step": 15018 + }, + { + "epoch": 0.17024164349968576, + "grad_norm": 0.14650489389896393, + "learning_rate": 3.2452006771017165e-05, + "loss": 3.1167, + "step": 15019 + }, + { + "epoch": 0.17027248996525918, + "grad_norm": 0.1499580293893814, + "learning_rate": 3.244943716860602e-05, + "loss": 3.126, + "step": 15020 + }, + { + "epoch": 0.1703033364308326, + "grad_norm": 0.154172882437706, + "learning_rate": 3.244686747982313e-05, + "loss": 3.1314, + "step": 15021 + }, + { + "epoch": 0.170334182896406, + "grad_norm": 0.1516495943069458, + "learning_rate": 3.244429770469829e-05, + "loss": 3.1129, + "step": 15022 + }, + { + "epoch": 0.17036502936197942, + "grad_norm": 0.15904222428798676, + "learning_rate": 3.2441727843261284e-05, + "loss": 3.0963, + "step": 15023 + }, + { + "epoch": 0.17039587582755283, + "grad_norm": 0.14912037551403046, + "learning_rate": 3.2439157895541916e-05, + "loss": 3.1199, + "step": 15024 + }, + { + "epoch": 0.17042672229312625, + "grad_norm": 0.1503729373216629, + "learning_rate": 3.243658786156998e-05, + "loss": 3.1789, + "step": 15025 + }, + { + "epoch": 0.17045756875869966, + "grad_norm": 0.14990617334842682, + "learning_rate": 3.243401774137528e-05, + "loss": 3.1299, + "step": 15026 + }, + { + "epoch": 0.17048841522427308, + "grad_norm": 0.15477152168750763, + "learning_rate": 3.243144753498761e-05, + "loss": 3.0958, + "step": 15027 + }, + { + "epoch": 0.1705192616898465, + "grad_norm": 0.16676510870456696, + "learning_rate": 3.242887724243678e-05, + "loss": 3.1401, + "step": 15028 + }, + { + "epoch": 0.1705501081554199, + "grad_norm": 0.14993855357170105, + "learning_rate": 3.242630686375257e-05, + "loss": 3.0754, + "step": 15029 + }, + { + "epoch": 0.17058095462099335, + "grad_norm": 0.1568634957075119, + "learning_rate": 3.242373639896481e-05, + "loss": 3.1111, + "step": 15030 + }, + { + "epoch": 0.17061180108656676, + "grad_norm": 0.14809638261795044, + "learning_rate": 3.2421165848103277e-05, + "loss": 3.1, + "step": 15031 + }, + { + "epoch": 0.17064264755214018, + "grad_norm": 0.15420997142791748, + "learning_rate": 3.241859521119779e-05, + "loss": 3.122, + "step": 15032 + }, + { + "epoch": 0.1706734940177136, + "grad_norm": 0.1763772964477539, + "learning_rate": 3.241602448827815e-05, + "loss": 3.0844, + "step": 15033 + }, + { + "epoch": 0.170704340483287, + "grad_norm": 0.15452656149864197, + "learning_rate": 3.241345367937416e-05, + "loss": 3.1143, + "step": 15034 + }, + { + "epoch": 0.17073518694886042, + "grad_norm": 0.1631760448217392, + "learning_rate": 3.241088278451563e-05, + "loss": 3.0943, + "step": 15035 + }, + { + "epoch": 0.17076603341443383, + "grad_norm": 0.15824757516384125, + "learning_rate": 3.240831180373236e-05, + "loss": 3.1031, + "step": 15036 + }, + { + "epoch": 0.17079687988000725, + "grad_norm": 0.15488208830356598, + "learning_rate": 3.240574073705417e-05, + "loss": 3.128, + "step": 15037 + }, + { + "epoch": 0.17082772634558066, + "grad_norm": 0.16604574024677277, + "learning_rate": 3.240316958451087e-05, + "loss": 3.1085, + "step": 15038 + }, + { + "epoch": 0.17085857281115407, + "grad_norm": 0.21592843532562256, + "learning_rate": 3.240059834613227e-05, + "loss": 3.0958, + "step": 15039 + }, + { + "epoch": 0.1708894192767275, + "grad_norm": 0.15658491849899292, + "learning_rate": 3.239802702194819e-05, + "loss": 3.1294, + "step": 15040 + }, + { + "epoch": 0.1709202657423009, + "grad_norm": 0.152015820145607, + "learning_rate": 3.23954556119884e-05, + "loss": 3.0895, + "step": 15041 + }, + { + "epoch": 0.17095111220787435, + "grad_norm": 0.1551504284143448, + "learning_rate": 3.239288411628277e-05, + "loss": 3.1521, + "step": 15042 + }, + { + "epoch": 0.17098195867344776, + "grad_norm": 0.1558615118265152, + "learning_rate": 3.239031253486109e-05, + "loss": 3.1163, + "step": 15043 + }, + { + "epoch": 0.17101280513902117, + "grad_norm": 0.1540578305721283, + "learning_rate": 3.238774086775317e-05, + "loss": 3.1186, + "step": 15044 + }, + { + "epoch": 0.1710436516045946, + "grad_norm": 0.1543923169374466, + "learning_rate": 3.238516911498885e-05, + "loss": 3.1181, + "step": 15045 + }, + { + "epoch": 0.171074498070168, + "grad_norm": 0.1529865562915802, + "learning_rate": 3.238259727659791e-05, + "loss": 3.0885, + "step": 15046 + }, + { + "epoch": 0.17110534453574142, + "grad_norm": 0.15644311904907227, + "learning_rate": 3.23800253526102e-05, + "loss": 3.1291, + "step": 15047 + }, + { + "epoch": 0.17113619100131483, + "grad_norm": 0.17673377692699432, + "learning_rate": 3.237745334305554e-05, + "loss": 3.1251, + "step": 15048 + }, + { + "epoch": 0.17116703746688824, + "grad_norm": 0.1467052698135376, + "learning_rate": 3.237488124796373e-05, + "loss": 3.1066, + "step": 15049 + }, + { + "epoch": 0.17119788393246166, + "grad_norm": 0.15062759816646576, + "learning_rate": 3.2372309067364604e-05, + "loss": 3.1313, + "step": 15050 + }, + { + "epoch": 0.17122873039803507, + "grad_norm": 0.18321149051189423, + "learning_rate": 3.236973680128799e-05, + "loss": 3.118, + "step": 15051 + }, + { + "epoch": 0.1712595768636085, + "grad_norm": 0.1523728221654892, + "learning_rate": 3.2367164449763706e-05, + "loss": 3.1012, + "step": 15052 + }, + { + "epoch": 0.1712904233291819, + "grad_norm": 0.15576419234275818, + "learning_rate": 3.236459201282157e-05, + "loss": 3.1157, + "step": 15053 + }, + { + "epoch": 0.17132126979475534, + "grad_norm": 0.1526104062795639, + "learning_rate": 3.236201949049142e-05, + "loss": 3.0961, + "step": 15054 + }, + { + "epoch": 0.17135211626032876, + "grad_norm": 0.15072958171367645, + "learning_rate": 3.235944688280309e-05, + "loss": 3.106, + "step": 15055 + }, + { + "epoch": 0.17138296272590217, + "grad_norm": 0.15966956317424774, + "learning_rate": 3.235687418978638e-05, + "loss": 3.0894, + "step": 15056 + }, + { + "epoch": 0.17141380919147559, + "grad_norm": 0.14833849668502808, + "learning_rate": 3.235430141147114e-05, + "loss": 3.0866, + "step": 15057 + }, + { + "epoch": 0.171444655657049, + "grad_norm": 0.15100504457950592, + "learning_rate": 3.2351728547887196e-05, + "loss": 3.0949, + "step": 15058 + }, + { + "epoch": 0.17147550212262241, + "grad_norm": 0.15181130170822144, + "learning_rate": 3.234915559906437e-05, + "loss": 3.1379, + "step": 15059 + }, + { + "epoch": 0.17150634858819583, + "grad_norm": 0.15071214735507965, + "learning_rate": 3.2346582565032515e-05, + "loss": 3.1182, + "step": 15060 + }, + { + "epoch": 0.17153719505376924, + "grad_norm": 0.15173104405403137, + "learning_rate": 3.234400944582144e-05, + "loss": 3.13, + "step": 15061 + }, + { + "epoch": 0.17156804151934266, + "grad_norm": 0.14420416951179504, + "learning_rate": 3.234143624146099e-05, + "loss": 3.121, + "step": 15062 + }, + { + "epoch": 0.17159888798491607, + "grad_norm": 0.14964115619659424, + "learning_rate": 3.2338862951981e-05, + "loss": 3.0918, + "step": 15063 + }, + { + "epoch": 0.17162973445048949, + "grad_norm": 0.15157529711723328, + "learning_rate": 3.23362895774113e-05, + "loss": 3.1147, + "step": 15064 + }, + { + "epoch": 0.1716605809160629, + "grad_norm": 0.14896605908870697, + "learning_rate": 3.2333716117781736e-05, + "loss": 3.1152, + "step": 15065 + }, + { + "epoch": 0.17169142738163634, + "grad_norm": 0.1592099815607071, + "learning_rate": 3.2331142573122144e-05, + "loss": 3.1613, + "step": 15066 + }, + { + "epoch": 0.17172227384720976, + "grad_norm": 0.154774010181427, + "learning_rate": 3.232856894346235e-05, + "loss": 3.1528, + "step": 15067 + }, + { + "epoch": 0.17175312031278317, + "grad_norm": 0.15378637611865997, + "learning_rate": 3.232599522883221e-05, + "loss": 3.0807, + "step": 15068 + }, + { + "epoch": 0.17178396677835658, + "grad_norm": 0.15924888849258423, + "learning_rate": 3.232342142926155e-05, + "loss": 3.103, + "step": 15069 + }, + { + "epoch": 0.17181481324393, + "grad_norm": 0.1519232839345932, + "learning_rate": 3.2320847544780224e-05, + "loss": 3.0849, + "step": 15070 + }, + { + "epoch": 0.1718456597095034, + "grad_norm": 0.15485075116157532, + "learning_rate": 3.2318273575418075e-05, + "loss": 3.1424, + "step": 15071 + }, + { + "epoch": 0.17187650617507683, + "grad_norm": 0.1542532593011856, + "learning_rate": 3.231569952120494e-05, + "loss": 3.1058, + "step": 15072 + }, + { + "epoch": 0.17190735264065024, + "grad_norm": 0.1525961458683014, + "learning_rate": 3.231312538217066e-05, + "loss": 3.1174, + "step": 15073 + }, + { + "epoch": 0.17193819910622365, + "grad_norm": 0.19556018710136414, + "learning_rate": 3.231055115834508e-05, + "loss": 3.1154, + "step": 15074 + }, + { + "epoch": 0.17196904557179707, + "grad_norm": 0.16130323708057404, + "learning_rate": 3.230797684975807e-05, + "loss": 3.1288, + "step": 15075 + }, + { + "epoch": 0.17199989203737048, + "grad_norm": 0.15849000215530396, + "learning_rate": 3.230540245643945e-05, + "loss": 3.1263, + "step": 15076 + }, + { + "epoch": 0.1720307385029439, + "grad_norm": 0.15459977090358734, + "learning_rate": 3.230282797841908e-05, + "loss": 3.1161, + "step": 15077 + }, + { + "epoch": 0.17206158496851734, + "grad_norm": 0.1577487289905548, + "learning_rate": 3.230025341572681e-05, + "loss": 3.1379, + "step": 15078 + }, + { + "epoch": 0.17209243143409075, + "grad_norm": 0.15388499200344086, + "learning_rate": 3.229767876839248e-05, + "loss": 3.1044, + "step": 15079 + }, + { + "epoch": 0.17212327789966417, + "grad_norm": 0.15750496089458466, + "learning_rate": 3.229510403644596e-05, + "loss": 3.155, + "step": 15080 + }, + { + "epoch": 0.17215412436523758, + "grad_norm": 0.1539364457130432, + "learning_rate": 3.229252921991709e-05, + "loss": 3.1202, + "step": 15081 + }, + { + "epoch": 0.172184970830811, + "grad_norm": 0.15116411447525024, + "learning_rate": 3.228995431883572e-05, + "loss": 3.133, + "step": 15082 + }, + { + "epoch": 0.1722158172963844, + "grad_norm": 0.1524934023618698, + "learning_rate": 3.2287379333231714e-05, + "loss": 3.1258, + "step": 15083 + }, + { + "epoch": 0.17224666376195782, + "grad_norm": 0.15749605000019073, + "learning_rate": 3.228480426313493e-05, + "loss": 3.1175, + "step": 15084 + }, + { + "epoch": 0.17227751022753124, + "grad_norm": 0.15723344683647156, + "learning_rate": 3.228222910857521e-05, + "loss": 3.1465, + "step": 15085 + }, + { + "epoch": 0.17230835669310465, + "grad_norm": 0.15198083221912384, + "learning_rate": 3.227965386958242e-05, + "loss": 3.1085, + "step": 15086 + }, + { + "epoch": 0.17233920315867807, + "grad_norm": 0.1503995954990387, + "learning_rate": 3.2277078546186424e-05, + "loss": 3.1145, + "step": 15087 + }, + { + "epoch": 0.17237004962425148, + "grad_norm": 0.15728464722633362, + "learning_rate": 3.2274503138417075e-05, + "loss": 3.1144, + "step": 15088 + }, + { + "epoch": 0.1724008960898249, + "grad_norm": 0.16656243801116943, + "learning_rate": 3.227192764630423e-05, + "loss": 3.0966, + "step": 15089 + }, + { + "epoch": 0.17243174255539834, + "grad_norm": 0.15662500262260437, + "learning_rate": 3.226935206987775e-05, + "loss": 3.1686, + "step": 15090 + }, + { + "epoch": 0.17246258902097175, + "grad_norm": 0.1559283435344696, + "learning_rate": 3.226677640916751e-05, + "loss": 3.1451, + "step": 15091 + }, + { + "epoch": 0.17249343548654517, + "grad_norm": 0.14762820303440094, + "learning_rate": 3.226420066420336e-05, + "loss": 3.1262, + "step": 15092 + }, + { + "epoch": 0.17252428195211858, + "grad_norm": 0.15521955490112305, + "learning_rate": 3.226162483501517e-05, + "loss": 3.1168, + "step": 15093 + }, + { + "epoch": 0.172555128417692, + "grad_norm": 0.15332213044166565, + "learning_rate": 3.22590489216328e-05, + "loss": 3.0794, + "step": 15094 + }, + { + "epoch": 0.1725859748832654, + "grad_norm": 0.15130427479743958, + "learning_rate": 3.225647292408612e-05, + "loss": 3.1473, + "step": 15095 + }, + { + "epoch": 0.17261682134883882, + "grad_norm": 0.15462346374988556, + "learning_rate": 3.2253896842405e-05, + "loss": 3.128, + "step": 15096 + }, + { + "epoch": 0.17264766781441224, + "grad_norm": 0.15053431689739227, + "learning_rate": 3.22513206766193e-05, + "loss": 3.1223, + "step": 15097 + }, + { + "epoch": 0.17267851427998565, + "grad_norm": 0.18554286658763885, + "learning_rate": 3.2248744426758904e-05, + "loss": 3.1048, + "step": 15098 + }, + { + "epoch": 0.17270936074555907, + "grad_norm": 0.15735022723674774, + "learning_rate": 3.2246168092853665e-05, + "loss": 3.1089, + "step": 15099 + }, + { + "epoch": 0.17274020721113248, + "grad_norm": 0.15195544064044952, + "learning_rate": 3.224359167493346e-05, + "loss": 3.106, + "step": 15100 + }, + { + "epoch": 0.1727710536767059, + "grad_norm": 0.14813660085201263, + "learning_rate": 3.2241015173028177e-05, + "loss": 3.1179, + "step": 15101 + }, + { + "epoch": 0.17280190014227934, + "grad_norm": 0.15728524327278137, + "learning_rate": 3.223843858716766e-05, + "loss": 3.1389, + "step": 15102 + }, + { + "epoch": 0.17283274660785275, + "grad_norm": 0.15100723505020142, + "learning_rate": 3.22358619173818e-05, + "loss": 3.0748, + "step": 15103 + }, + { + "epoch": 0.17286359307342616, + "grad_norm": 0.1622324138879776, + "learning_rate": 3.223328516370047e-05, + "loss": 3.1177, + "step": 15104 + }, + { + "epoch": 0.17289443953899958, + "grad_norm": 0.15261785686016083, + "learning_rate": 3.2230708326153545e-05, + "loss": 3.1328, + "step": 15105 + }, + { + "epoch": 0.172925286004573, + "grad_norm": 0.15163013339042664, + "learning_rate": 3.22281314047709e-05, + "loss": 3.1476, + "step": 15106 + }, + { + "epoch": 0.1729561324701464, + "grad_norm": 0.15957637131214142, + "learning_rate": 3.2225554399582425e-05, + "loss": 3.1121, + "step": 15107 + }, + { + "epoch": 0.17298697893571982, + "grad_norm": 0.17045988142490387, + "learning_rate": 3.2222977310617976e-05, + "loss": 3.1135, + "step": 15108 + }, + { + "epoch": 0.17301782540129323, + "grad_norm": 0.16662250459194183, + "learning_rate": 3.222040013790745e-05, + "loss": 3.1295, + "step": 15109 + }, + { + "epoch": 0.17304867186686665, + "grad_norm": 0.1635906994342804, + "learning_rate": 3.221782288148072e-05, + "loss": 3.1031, + "step": 15110 + }, + { + "epoch": 0.17307951833244006, + "grad_norm": 0.1554432362318039, + "learning_rate": 3.221524554136768e-05, + "loss": 3.0751, + "step": 15111 + }, + { + "epoch": 0.17311036479801348, + "grad_norm": 0.16507549583911896, + "learning_rate": 3.22126681175982e-05, + "loss": 3.101, + "step": 15112 + }, + { + "epoch": 0.1731412112635869, + "grad_norm": 0.15555591881275177, + "learning_rate": 3.221009061020217e-05, + "loss": 3.1427, + "step": 15113 + }, + { + "epoch": 0.17317205772916033, + "grad_norm": 0.16011500358581543, + "learning_rate": 3.2207513019209455e-05, + "loss": 3.1312, + "step": 15114 + }, + { + "epoch": 0.17320290419473375, + "grad_norm": 0.16356614232063293, + "learning_rate": 3.220493534464998e-05, + "loss": 3.1333, + "step": 15115 + }, + { + "epoch": 0.17323375066030716, + "grad_norm": 0.16170555353164673, + "learning_rate": 3.2202357586553596e-05, + "loss": 3.1335, + "step": 15116 + }, + { + "epoch": 0.17326459712588058, + "grad_norm": 0.15427404642105103, + "learning_rate": 3.2199779744950216e-05, + "loss": 3.1227, + "step": 15117 + }, + { + "epoch": 0.173295443591454, + "grad_norm": 0.1597643792629242, + "learning_rate": 3.219720181986971e-05, + "loss": 3.1002, + "step": 15118 + }, + { + "epoch": 0.1733262900570274, + "grad_norm": 0.15494516491889954, + "learning_rate": 3.219462381134197e-05, + "loss": 3.1264, + "step": 15119 + }, + { + "epoch": 0.17335713652260082, + "grad_norm": 0.16064751148223877, + "learning_rate": 3.219204571939689e-05, + "loss": 3.145, + "step": 15120 + }, + { + "epoch": 0.17338798298817423, + "grad_norm": 0.1607142686843872, + "learning_rate": 3.218946754406436e-05, + "loss": 3.1067, + "step": 15121 + }, + { + "epoch": 0.17341882945374765, + "grad_norm": 0.1517169326543808, + "learning_rate": 3.218688928537429e-05, + "loss": 3.1114, + "step": 15122 + }, + { + "epoch": 0.17344967591932106, + "grad_norm": 0.16505961120128632, + "learning_rate": 3.218431094335654e-05, + "loss": 3.1179, + "step": 15123 + }, + { + "epoch": 0.17348052238489448, + "grad_norm": 0.15634563565254211, + "learning_rate": 3.218173251804103e-05, + "loss": 3.1247, + "step": 15124 + }, + { + "epoch": 0.1735113688504679, + "grad_norm": 0.15648411214351654, + "learning_rate": 3.2179154009457644e-05, + "loss": 3.1269, + "step": 15125 + }, + { + "epoch": 0.17354221531604133, + "grad_norm": 0.15962955355644226, + "learning_rate": 3.217657541763629e-05, + "loss": 3.1369, + "step": 15126 + }, + { + "epoch": 0.17357306178161475, + "grad_norm": 0.15191268920898438, + "learning_rate": 3.2173996742606846e-05, + "loss": 3.1281, + "step": 15127 + }, + { + "epoch": 0.17360390824718816, + "grad_norm": 0.16662637889385223, + "learning_rate": 3.217141798439923e-05, + "loss": 3.1622, + "step": 15128 + }, + { + "epoch": 0.17363475471276157, + "grad_norm": 0.15941289067268372, + "learning_rate": 3.216883914304333e-05, + "loss": 3.0846, + "step": 15129 + }, + { + "epoch": 0.173665601178335, + "grad_norm": 0.161531001329422, + "learning_rate": 3.216626021856904e-05, + "loss": 3.1153, + "step": 15130 + }, + { + "epoch": 0.1736964476439084, + "grad_norm": 0.2696567177772522, + "learning_rate": 3.2163681211006275e-05, + "loss": 3.1059, + "step": 15131 + }, + { + "epoch": 0.17372729410948182, + "grad_norm": 0.16509507596492767, + "learning_rate": 3.216110212038493e-05, + "loss": 3.1163, + "step": 15132 + }, + { + "epoch": 0.17375814057505523, + "grad_norm": 0.15868981182575226, + "learning_rate": 3.2158522946734906e-05, + "loss": 3.1377, + "step": 15133 + }, + { + "epoch": 0.17378898704062865, + "grad_norm": 0.16053606569766998, + "learning_rate": 3.2155943690086124e-05, + "loss": 3.158, + "step": 15134 + }, + { + "epoch": 0.17381983350620206, + "grad_norm": 0.157362163066864, + "learning_rate": 3.215336435046846e-05, + "loss": 3.0786, + "step": 15135 + }, + { + "epoch": 0.17385067997177547, + "grad_norm": 0.15881022810935974, + "learning_rate": 3.2150784927911834e-05, + "loss": 3.0978, + "step": 15136 + }, + { + "epoch": 0.1738815264373489, + "grad_norm": 0.15657292306423187, + "learning_rate": 3.214820542244617e-05, + "loss": 3.0832, + "step": 15137 + }, + { + "epoch": 0.17391237290292233, + "grad_norm": 0.15157859027385712, + "learning_rate": 3.214562583410134e-05, + "loss": 3.1153, + "step": 15138 + }, + { + "epoch": 0.17394321936849574, + "grad_norm": 0.1634196937084198, + "learning_rate": 3.2143046162907285e-05, + "loss": 3.1293, + "step": 15139 + }, + { + "epoch": 0.17397406583406916, + "grad_norm": 0.14911514520645142, + "learning_rate": 3.21404664088939e-05, + "loss": 3.1143, + "step": 15140 + }, + { + "epoch": 0.17400491229964257, + "grad_norm": 0.1521507352590561, + "learning_rate": 3.213788657209109e-05, + "loss": 3.08, + "step": 15141 + }, + { + "epoch": 0.174035758765216, + "grad_norm": 0.16057340800762177, + "learning_rate": 3.2135306652528785e-05, + "loss": 3.1315, + "step": 15142 + }, + { + "epoch": 0.1740666052307894, + "grad_norm": 0.15645819902420044, + "learning_rate": 3.2132726650236886e-05, + "loss": 3.1297, + "step": 15143 + }, + { + "epoch": 0.17409745169636281, + "grad_norm": 0.15025533735752106, + "learning_rate": 3.2130146565245304e-05, + "loss": 3.0859, + "step": 15144 + }, + { + "epoch": 0.17412829816193623, + "grad_norm": 0.1523531973361969, + "learning_rate": 3.2127566397583964e-05, + "loss": 3.0815, + "step": 15145 + }, + { + "epoch": 0.17415914462750964, + "grad_norm": 0.16968534886837006, + "learning_rate": 3.212498614728276e-05, + "loss": 3.0984, + "step": 15146 + }, + { + "epoch": 0.17418999109308306, + "grad_norm": 0.14792682230472565, + "learning_rate": 3.212240581437164e-05, + "loss": 3.1144, + "step": 15147 + }, + { + "epoch": 0.17422083755865647, + "grad_norm": 0.1544799953699112, + "learning_rate": 3.21198253988805e-05, + "loss": 3.1299, + "step": 15148 + }, + { + "epoch": 0.17425168402422989, + "grad_norm": 0.15560103952884674, + "learning_rate": 3.2117244900839264e-05, + "loss": 3.1263, + "step": 15149 + }, + { + "epoch": 0.17428253048980333, + "grad_norm": 0.14953328669071198, + "learning_rate": 3.211466432027784e-05, + "loss": 3.1358, + "step": 15150 + }, + { + "epoch": 0.17431337695537674, + "grad_norm": 0.15330669283866882, + "learning_rate": 3.211208365722617e-05, + "loss": 3.098, + "step": 15151 + }, + { + "epoch": 0.17434422342095016, + "grad_norm": 0.1484685093164444, + "learning_rate": 3.210950291171416e-05, + "loss": 3.1062, + "step": 15152 + }, + { + "epoch": 0.17437506988652357, + "grad_norm": 0.14717155694961548, + "learning_rate": 3.210692208377173e-05, + "loss": 3.1306, + "step": 15153 + }, + { + "epoch": 0.17440591635209698, + "grad_norm": 0.15082405507564545, + "learning_rate": 3.210434117342882e-05, + "loss": 3.1358, + "step": 15154 + }, + { + "epoch": 0.1744367628176704, + "grad_norm": 0.15282423794269562, + "learning_rate": 3.210176018071534e-05, + "loss": 3.1348, + "step": 15155 + }, + { + "epoch": 0.1744676092832438, + "grad_norm": 0.15032725036144257, + "learning_rate": 3.2099179105661216e-05, + "loss": 3.1618, + "step": 15156 + }, + { + "epoch": 0.17449845574881723, + "grad_norm": 0.16357460618019104, + "learning_rate": 3.209659794829638e-05, + "loss": 3.1174, + "step": 15157 + }, + { + "epoch": 0.17452930221439064, + "grad_norm": 0.1513413041830063, + "learning_rate": 3.209401670865074e-05, + "loss": 3.0802, + "step": 15158 + }, + { + "epoch": 0.17456014867996406, + "grad_norm": 0.15788030624389648, + "learning_rate": 3.209143538675425e-05, + "loss": 3.1471, + "step": 15159 + }, + { + "epoch": 0.17459099514553747, + "grad_norm": 0.1541554480791092, + "learning_rate": 3.208885398263683e-05, + "loss": 3.1482, + "step": 15160 + }, + { + "epoch": 0.17462184161111088, + "grad_norm": 0.15597382187843323, + "learning_rate": 3.208627249632841e-05, + "loss": 3.1445, + "step": 15161 + }, + { + "epoch": 0.17465268807668433, + "grad_norm": 0.1516299843788147, + "learning_rate": 3.208369092785892e-05, + "loss": 3.1404, + "step": 15162 + }, + { + "epoch": 0.17468353454225774, + "grad_norm": 0.15021273493766785, + "learning_rate": 3.208110927725829e-05, + "loss": 3.1338, + "step": 15163 + }, + { + "epoch": 0.17471438100783115, + "grad_norm": 0.1502426266670227, + "learning_rate": 3.207852754455644e-05, + "loss": 3.1159, + "step": 15164 + }, + { + "epoch": 0.17474522747340457, + "grad_norm": 0.15089359879493713, + "learning_rate": 3.207594572978333e-05, + "loss": 3.1039, + "step": 15165 + }, + { + "epoch": 0.17477607393897798, + "grad_norm": 0.1585555225610733, + "learning_rate": 3.207336383296888e-05, + "loss": 3.1015, + "step": 15166 + }, + { + "epoch": 0.1748069204045514, + "grad_norm": 0.1489323228597641, + "learning_rate": 3.207078185414303e-05, + "loss": 3.1305, + "step": 15167 + }, + { + "epoch": 0.1748377668701248, + "grad_norm": 0.15965615212917328, + "learning_rate": 3.206819979333571e-05, + "loss": 3.1218, + "step": 15168 + }, + { + "epoch": 0.17486861333569823, + "grad_norm": 0.1506299376487732, + "learning_rate": 3.206561765057685e-05, + "loss": 3.1158, + "step": 15169 + }, + { + "epoch": 0.17489945980127164, + "grad_norm": 0.15733233094215393, + "learning_rate": 3.206303542589641e-05, + "loss": 3.1141, + "step": 15170 + }, + { + "epoch": 0.17493030626684505, + "grad_norm": 0.14857715368270874, + "learning_rate": 3.206045311932432e-05, + "loss": 3.0756, + "step": 15171 + }, + { + "epoch": 0.17496115273241847, + "grad_norm": 0.15494239330291748, + "learning_rate": 3.205787073089052e-05, + "loss": 3.1304, + "step": 15172 + }, + { + "epoch": 0.17499199919799188, + "grad_norm": 0.15143710374832153, + "learning_rate": 3.205528826062494e-05, + "loss": 3.1318, + "step": 15173 + }, + { + "epoch": 0.17502284566356532, + "grad_norm": 0.15522275865077972, + "learning_rate": 3.205270570855754e-05, + "loss": 3.146, + "step": 15174 + }, + { + "epoch": 0.17505369212913874, + "grad_norm": 0.1608215719461441, + "learning_rate": 3.2050123074718256e-05, + "loss": 3.1399, + "step": 15175 + }, + { + "epoch": 0.17508453859471215, + "grad_norm": 0.14845508337020874, + "learning_rate": 3.204754035913703e-05, + "loss": 3.1171, + "step": 15176 + }, + { + "epoch": 0.17511538506028557, + "grad_norm": 0.1521775722503662, + "learning_rate": 3.2044957561843816e-05, + "loss": 3.1157, + "step": 15177 + }, + { + "epoch": 0.17514623152585898, + "grad_norm": 0.1560697853565216, + "learning_rate": 3.204237468286854e-05, + "loss": 3.1144, + "step": 15178 + }, + { + "epoch": 0.1751770779914324, + "grad_norm": 0.15598200261592865, + "learning_rate": 3.203979172224116e-05, + "loss": 3.1575, + "step": 15179 + }, + { + "epoch": 0.1752079244570058, + "grad_norm": 0.15543919801712036, + "learning_rate": 3.203720867999164e-05, + "loss": 3.1457, + "step": 15180 + }, + { + "epoch": 0.17523877092257922, + "grad_norm": 0.15241210162639618, + "learning_rate": 3.203462555614991e-05, + "loss": 3.1257, + "step": 15181 + }, + { + "epoch": 0.17526961738815264, + "grad_norm": 0.1539706140756607, + "learning_rate": 3.203204235074591e-05, + "loss": 3.1148, + "step": 15182 + }, + { + "epoch": 0.17530046385372605, + "grad_norm": 0.1484224796295166, + "learning_rate": 3.202945906380962e-05, + "loss": 3.1078, + "step": 15183 + }, + { + "epoch": 0.17533131031929947, + "grad_norm": 0.15594610571861267, + "learning_rate": 3.2026875695370975e-05, + "loss": 3.1286, + "step": 15184 + }, + { + "epoch": 0.17536215678487288, + "grad_norm": 0.15242981910705566, + "learning_rate": 3.202429224545992e-05, + "loss": 3.1174, + "step": 15185 + }, + { + "epoch": 0.17539300325044632, + "grad_norm": 0.15528294444084167, + "learning_rate": 3.202170871410643e-05, + "loss": 3.1084, + "step": 15186 + }, + { + "epoch": 0.17542384971601974, + "grad_norm": 0.15060079097747803, + "learning_rate": 3.201912510134043e-05, + "loss": 3.084, + "step": 15187 + }, + { + "epoch": 0.17545469618159315, + "grad_norm": 0.15548409521579742, + "learning_rate": 3.2016541407191905e-05, + "loss": 3.1272, + "step": 15188 + }, + { + "epoch": 0.17548554264716656, + "grad_norm": 0.15543274581432343, + "learning_rate": 3.20139576316908e-05, + "loss": 3.1189, + "step": 15189 + }, + { + "epoch": 0.17551638911273998, + "grad_norm": 0.1503191590309143, + "learning_rate": 3.201137377486707e-05, + "loss": 3.0923, + "step": 15190 + }, + { + "epoch": 0.1755472355783134, + "grad_norm": 0.16511449217796326, + "learning_rate": 3.2008789836750666e-05, + "loss": 3.1243, + "step": 15191 + }, + { + "epoch": 0.1755780820438868, + "grad_norm": 0.16098207235336304, + "learning_rate": 3.200620581737156e-05, + "loss": 3.1237, + "step": 15192 + }, + { + "epoch": 0.17560892850946022, + "grad_norm": 0.15275442600250244, + "learning_rate": 3.200362171675971e-05, + "loss": 3.0966, + "step": 15193 + }, + { + "epoch": 0.17563977497503364, + "grad_norm": 0.15825071930885315, + "learning_rate": 3.2001037534945076e-05, + "loss": 3.1005, + "step": 15194 + }, + { + "epoch": 0.17567062144060705, + "grad_norm": 0.1808285415172577, + "learning_rate": 3.199845327195762e-05, + "loss": 3.1455, + "step": 15195 + }, + { + "epoch": 0.17570146790618046, + "grad_norm": 0.1568116396665573, + "learning_rate": 3.19958689278273e-05, + "loss": 3.1095, + "step": 15196 + }, + { + "epoch": 0.17573231437175388, + "grad_norm": 0.15760578215122223, + "learning_rate": 3.199328450258408e-05, + "loss": 3.094, + "step": 15197 + }, + { + "epoch": 0.17576316083732732, + "grad_norm": 0.15580891072750092, + "learning_rate": 3.1990699996257936e-05, + "loss": 3.0999, + "step": 15198 + }, + { + "epoch": 0.17579400730290073, + "grad_norm": 0.1692148894071579, + "learning_rate": 3.198811540887883e-05, + "loss": 3.093, + "step": 15199 + }, + { + "epoch": 0.17582485376847415, + "grad_norm": 0.1502346396446228, + "learning_rate": 3.1985530740476713e-05, + "loss": 3.1145, + "step": 15200 + }, + { + "epoch": 0.17585570023404756, + "grad_norm": 0.16046801209449768, + "learning_rate": 3.198294599108157e-05, + "loss": 3.139, + "step": 15201 + }, + { + "epoch": 0.17588654669962098, + "grad_norm": 0.1574101597070694, + "learning_rate": 3.1980361160723366e-05, + "loss": 3.0992, + "step": 15202 + }, + { + "epoch": 0.1759173931651944, + "grad_norm": 0.16302376985549927, + "learning_rate": 3.1977776249432065e-05, + "loss": 3.1344, + "step": 15203 + }, + { + "epoch": 0.1759482396307678, + "grad_norm": 0.15018852055072784, + "learning_rate": 3.1975191257237645e-05, + "loss": 3.1153, + "step": 15204 + }, + { + "epoch": 0.17597908609634122, + "grad_norm": 0.16974253952503204, + "learning_rate": 3.197260618417008e-05, + "loss": 3.142, + "step": 15205 + }, + { + "epoch": 0.17600993256191463, + "grad_norm": 0.15717072784900665, + "learning_rate": 3.197002103025933e-05, + "loss": 3.0891, + "step": 15206 + }, + { + "epoch": 0.17604077902748805, + "grad_norm": 0.15981332957744598, + "learning_rate": 3.1967435795535374e-05, + "loss": 3.118, + "step": 15207 + }, + { + "epoch": 0.17607162549306146, + "grad_norm": 0.15256759524345398, + "learning_rate": 3.196485048002819e-05, + "loss": 3.1016, + "step": 15208 + }, + { + "epoch": 0.17610247195863488, + "grad_norm": 0.1648041009902954, + "learning_rate": 3.196226508376774e-05, + "loss": 3.1259, + "step": 15209 + }, + { + "epoch": 0.17613331842420832, + "grad_norm": 0.15103285014629364, + "learning_rate": 3.195967960678402e-05, + "loss": 3.1645, + "step": 15210 + }, + { + "epoch": 0.17616416488978173, + "grad_norm": 0.1485898196697235, + "learning_rate": 3.1957094049106996e-05, + "loss": 3.1353, + "step": 15211 + }, + { + "epoch": 0.17619501135535515, + "grad_norm": 0.15452618896961212, + "learning_rate": 3.195450841076665e-05, + "loss": 3.1272, + "step": 15212 + }, + { + "epoch": 0.17622585782092856, + "grad_norm": 0.15611374378204346, + "learning_rate": 3.195192269179295e-05, + "loss": 3.097, + "step": 15213 + }, + { + "epoch": 0.17625670428650198, + "grad_norm": 0.14800743758678436, + "learning_rate": 3.194933689221589e-05, + "loss": 3.1031, + "step": 15214 + }, + { + "epoch": 0.1762875507520754, + "grad_norm": 0.15845906734466553, + "learning_rate": 3.1946751012065445e-05, + "loss": 3.1225, + "step": 15215 + }, + { + "epoch": 0.1763183972176488, + "grad_norm": 0.14938902854919434, + "learning_rate": 3.19441650513716e-05, + "loss": 3.1016, + "step": 15216 + }, + { + "epoch": 0.17634924368322222, + "grad_norm": 0.15095487236976624, + "learning_rate": 3.194157901016432e-05, + "loss": 3.1198, + "step": 15217 + }, + { + "epoch": 0.17638009014879563, + "grad_norm": 0.15783533453941345, + "learning_rate": 3.193899288847362e-05, + "loss": 3.1154, + "step": 15218 + }, + { + "epoch": 0.17641093661436905, + "grad_norm": 0.15458990633487701, + "learning_rate": 3.1936406686329456e-05, + "loss": 3.1148, + "step": 15219 + }, + { + "epoch": 0.17644178307994246, + "grad_norm": 0.16756665706634521, + "learning_rate": 3.1933820403761826e-05, + "loss": 3.1027, + "step": 15220 + }, + { + "epoch": 0.17647262954551587, + "grad_norm": 0.15160657465457916, + "learning_rate": 3.193123404080072e-05, + "loss": 3.104, + "step": 15221 + }, + { + "epoch": 0.17650347601108932, + "grad_norm": 0.14763937890529633, + "learning_rate": 3.1928647597476116e-05, + "loss": 3.098, + "step": 15222 + }, + { + "epoch": 0.17653432247666273, + "grad_norm": 0.15938851237297058, + "learning_rate": 3.1926061073818006e-05, + "loss": 3.1002, + "step": 15223 + }, + { + "epoch": 0.17656516894223614, + "grad_norm": 0.16910052299499512, + "learning_rate": 3.192347446985638e-05, + "loss": 3.11, + "step": 15224 + }, + { + "epoch": 0.17659601540780956, + "grad_norm": 0.15191493928432465, + "learning_rate": 3.1920887785621235e-05, + "loss": 3.098, + "step": 15225 + }, + { + "epoch": 0.17662686187338297, + "grad_norm": 0.15525612235069275, + "learning_rate": 3.191830102114255e-05, + "loss": 3.1249, + "step": 15226 + }, + { + "epoch": 0.1766577083389564, + "grad_norm": 0.14769592881202698, + "learning_rate": 3.191571417645033e-05, + "loss": 3.1184, + "step": 15227 + }, + { + "epoch": 0.1766885548045298, + "grad_norm": 0.1529129296541214, + "learning_rate": 3.191312725157455e-05, + "loss": 3.0964, + "step": 15228 + }, + { + "epoch": 0.17671940127010322, + "grad_norm": 0.15515995025634766, + "learning_rate": 3.191054024654522e-05, + "loss": 3.1049, + "step": 15229 + }, + { + "epoch": 0.17675024773567663, + "grad_norm": 0.14907504618167877, + "learning_rate": 3.190795316139232e-05, + "loss": 3.0712, + "step": 15230 + }, + { + "epoch": 0.17678109420125004, + "grad_norm": 0.35374611616134644, + "learning_rate": 3.190536599614586e-05, + "loss": 3.1534, + "step": 15231 + }, + { + "epoch": 0.17681194066682346, + "grad_norm": 0.14896027743816376, + "learning_rate": 3.1902778750835824e-05, + "loss": 3.1038, + "step": 15232 + }, + { + "epoch": 0.17684278713239687, + "grad_norm": 0.16238446533679962, + "learning_rate": 3.1900191425492234e-05, + "loss": 3.1242, + "step": 15233 + }, + { + "epoch": 0.17687363359797031, + "grad_norm": 0.16123369336128235, + "learning_rate": 3.1897604020145054e-05, + "loss": 3.1203, + "step": 15234 + }, + { + "epoch": 0.17690448006354373, + "grad_norm": 0.15522074699401855, + "learning_rate": 3.189501653482431e-05, + "loss": 3.1186, + "step": 15235 + }, + { + "epoch": 0.17693532652911714, + "grad_norm": 0.15428397059440613, + "learning_rate": 3.1892428969559986e-05, + "loss": 3.1047, + "step": 15236 + }, + { + "epoch": 0.17696617299469056, + "grad_norm": 0.16683493554592133, + "learning_rate": 3.188984132438209e-05, + "loss": 3.1214, + "step": 15237 + }, + { + "epoch": 0.17699701946026397, + "grad_norm": 0.15081317722797394, + "learning_rate": 3.188725359932063e-05, + "loss": 3.114, + "step": 15238 + }, + { + "epoch": 0.17702786592583739, + "grad_norm": 0.1521141082048416, + "learning_rate": 3.18846657944056e-05, + "loss": 3.1225, + "step": 15239 + }, + { + "epoch": 0.1770587123914108, + "grad_norm": 0.15037159621715546, + "learning_rate": 3.188207790966701e-05, + "loss": 3.1339, + "step": 15240 + }, + { + "epoch": 0.1770895588569842, + "grad_norm": 0.1490737944841385, + "learning_rate": 3.1879489945134855e-05, + "loss": 3.1013, + "step": 15241 + }, + { + "epoch": 0.17712040532255763, + "grad_norm": 0.15018866956233978, + "learning_rate": 3.187690190083915e-05, + "loss": 3.1363, + "step": 15242 + }, + { + "epoch": 0.17715125178813104, + "grad_norm": 0.15352657437324524, + "learning_rate": 3.1874313776809914e-05, + "loss": 3.0928, + "step": 15243 + }, + { + "epoch": 0.17718209825370446, + "grad_norm": 0.1615821123123169, + "learning_rate": 3.187172557307712e-05, + "loss": 3.1714, + "step": 15244 + }, + { + "epoch": 0.17721294471927787, + "grad_norm": 0.1976843774318695, + "learning_rate": 3.186913728967081e-05, + "loss": 3.1309, + "step": 15245 + }, + { + "epoch": 0.1772437911848513, + "grad_norm": 0.17000596225261688, + "learning_rate": 3.186654892662098e-05, + "loss": 3.1599, + "step": 15246 + }, + { + "epoch": 0.17727463765042473, + "grad_norm": 0.15737436711788177, + "learning_rate": 3.186396048395763e-05, + "loss": 3.1111, + "step": 15247 + }, + { + "epoch": 0.17730548411599814, + "grad_norm": 0.15743644535541534, + "learning_rate": 3.18613719617108e-05, + "loss": 3.1068, + "step": 15248 + }, + { + "epoch": 0.17733633058157156, + "grad_norm": 0.15535353124141693, + "learning_rate": 3.185878335991048e-05, + "loss": 3.1216, + "step": 15249 + }, + { + "epoch": 0.17736717704714497, + "grad_norm": 0.17870062589645386, + "learning_rate": 3.185619467858668e-05, + "loss": 3.1262, + "step": 15250 + }, + { + "epoch": 0.17739802351271838, + "grad_norm": 0.16040712594985962, + "learning_rate": 3.185360591776944e-05, + "loss": 3.1177, + "step": 15251 + }, + { + "epoch": 0.1774288699782918, + "grad_norm": 0.1556777060031891, + "learning_rate": 3.1851017077488746e-05, + "loss": 3.1348, + "step": 15252 + }, + { + "epoch": 0.1774597164438652, + "grad_norm": 0.15507441759109497, + "learning_rate": 3.184842815777462e-05, + "loss": 3.1119, + "step": 15253 + }, + { + "epoch": 0.17749056290943863, + "grad_norm": 0.1714223027229309, + "learning_rate": 3.184583915865709e-05, + "loss": 3.1082, + "step": 15254 + }, + { + "epoch": 0.17752140937501204, + "grad_norm": 0.15343542397022247, + "learning_rate": 3.184325008016618e-05, + "loss": 3.1143, + "step": 15255 + }, + { + "epoch": 0.17755225584058545, + "grad_norm": 0.15614378452301025, + "learning_rate": 3.1840660922331885e-05, + "loss": 3.1446, + "step": 15256 + }, + { + "epoch": 0.17758310230615887, + "grad_norm": 0.16169525682926178, + "learning_rate": 3.183807168518423e-05, + "loss": 3.1598, + "step": 15257 + }, + { + "epoch": 0.1776139487717323, + "grad_norm": 0.1548771858215332, + "learning_rate": 3.1835482368753254e-05, + "loss": 3.136, + "step": 15258 + }, + { + "epoch": 0.17764479523730572, + "grad_norm": 0.1596042364835739, + "learning_rate": 3.183289297306897e-05, + "loss": 3.1347, + "step": 15259 + }, + { + "epoch": 0.17767564170287914, + "grad_norm": 0.15183424949645996, + "learning_rate": 3.183030349816141e-05, + "loss": 3.1208, + "step": 15260 + }, + { + "epoch": 0.17770648816845255, + "grad_norm": 0.15435849130153656, + "learning_rate": 3.182771394406056e-05, + "loss": 3.1105, + "step": 15261 + }, + { + "epoch": 0.17773733463402597, + "grad_norm": 0.15423528850078583, + "learning_rate": 3.182512431079648e-05, + "loss": 3.1149, + "step": 15262 + }, + { + "epoch": 0.17776818109959938, + "grad_norm": 0.1553199142217636, + "learning_rate": 3.182253459839919e-05, + "loss": 3.1249, + "step": 15263 + }, + { + "epoch": 0.1777990275651728, + "grad_norm": 0.15030135214328766, + "learning_rate": 3.181994480689871e-05, + "loss": 3.1066, + "step": 15264 + }, + { + "epoch": 0.1778298740307462, + "grad_norm": 0.14936836063861847, + "learning_rate": 3.181735493632506e-05, + "loss": 3.0981, + "step": 15265 + }, + { + "epoch": 0.17786072049631962, + "grad_norm": 0.1556192934513092, + "learning_rate": 3.181476498670828e-05, + "loss": 3.1241, + "step": 15266 + }, + { + "epoch": 0.17789156696189304, + "grad_norm": 0.15682871639728546, + "learning_rate": 3.18121749580784e-05, + "loss": 3.1507, + "step": 15267 + }, + { + "epoch": 0.17792241342746645, + "grad_norm": 0.15590247511863708, + "learning_rate": 3.180958485046545e-05, + "loss": 3.1575, + "step": 15268 + }, + { + "epoch": 0.17795325989303987, + "grad_norm": 0.15707343816757202, + "learning_rate": 3.180699466389944e-05, + "loss": 3.1532, + "step": 15269 + }, + { + "epoch": 0.1779841063586133, + "grad_norm": 0.15145133435726166, + "learning_rate": 3.1804404398410425e-05, + "loss": 3.1038, + "step": 15270 + }, + { + "epoch": 0.17801495282418672, + "grad_norm": 0.1535886973142624, + "learning_rate": 3.1801814054028426e-05, + "loss": 3.1378, + "step": 15271 + }, + { + "epoch": 0.17804579928976014, + "grad_norm": 0.15404808521270752, + "learning_rate": 3.179922363078349e-05, + "loss": 3.1323, + "step": 15272 + }, + { + "epoch": 0.17807664575533355, + "grad_norm": 0.16341596841812134, + "learning_rate": 3.179663312870563e-05, + "loss": 3.0897, + "step": 15273 + }, + { + "epoch": 0.17810749222090697, + "grad_norm": 0.15426529943943024, + "learning_rate": 3.1794042547824904e-05, + "loss": 3.1299, + "step": 15274 + }, + { + "epoch": 0.17813833868648038, + "grad_norm": 0.15237602591514587, + "learning_rate": 3.179145188817133e-05, + "loss": 3.1074, + "step": 15275 + }, + { + "epoch": 0.1781691851520538, + "grad_norm": 0.1538711041212082, + "learning_rate": 3.178886114977496e-05, + "loss": 3.1318, + "step": 15276 + }, + { + "epoch": 0.1782000316176272, + "grad_norm": 0.19711296260356903, + "learning_rate": 3.1786270332665816e-05, + "loss": 3.1104, + "step": 15277 + }, + { + "epoch": 0.17823087808320062, + "grad_norm": 0.15398578345775604, + "learning_rate": 3.1783679436873954e-05, + "loss": 3.1155, + "step": 15278 + }, + { + "epoch": 0.17826172454877404, + "grad_norm": 0.15704266726970673, + "learning_rate": 3.17810884624294e-05, + "loss": 3.0846, + "step": 15279 + }, + { + "epoch": 0.17829257101434745, + "grad_norm": 0.1503816395998001, + "learning_rate": 3.177849740936219e-05, + "loss": 3.1032, + "step": 15280 + }, + { + "epoch": 0.1783234174799209, + "grad_norm": 0.15365266799926758, + "learning_rate": 3.17759062777024e-05, + "loss": 3.1301, + "step": 15281 + }, + { + "epoch": 0.1783542639454943, + "grad_norm": 0.15561869740486145, + "learning_rate": 3.177331506748003e-05, + "loss": 3.1096, + "step": 15282 + }, + { + "epoch": 0.17838511041106772, + "grad_norm": 0.14870548248291016, + "learning_rate": 3.1770723778725154e-05, + "loss": 3.1079, + "step": 15283 + }, + { + "epoch": 0.17841595687664114, + "grad_norm": 0.14888857305049896, + "learning_rate": 3.17681324114678e-05, + "loss": 3.1122, + "step": 15284 + }, + { + "epoch": 0.17844680334221455, + "grad_norm": 0.1507096141576767, + "learning_rate": 3.176554096573801e-05, + "loss": 3.1324, + "step": 15285 + }, + { + "epoch": 0.17847764980778796, + "grad_norm": 0.15177211165428162, + "learning_rate": 3.1762949441565846e-05, + "loss": 3.0951, + "step": 15286 + }, + { + "epoch": 0.17850849627336138, + "grad_norm": 0.163002148270607, + "learning_rate": 3.176035783898135e-05, + "loss": 3.1133, + "step": 15287 + }, + { + "epoch": 0.1785393427389348, + "grad_norm": 0.1532658338546753, + "learning_rate": 3.1757766158014555e-05, + "loss": 3.1128, + "step": 15288 + }, + { + "epoch": 0.1785701892045082, + "grad_norm": 0.16106700897216797, + "learning_rate": 3.1755174398695534e-05, + "loss": 3.1146, + "step": 15289 + }, + { + "epoch": 0.17860103567008162, + "grad_norm": 0.15691015124320984, + "learning_rate": 3.1752582561054325e-05, + "loss": 3.1158, + "step": 15290 + }, + { + "epoch": 0.17863188213565503, + "grad_norm": 0.15440340340137482, + "learning_rate": 3.174999064512097e-05, + "loss": 3.1085, + "step": 15291 + }, + { + "epoch": 0.17866272860122845, + "grad_norm": 0.15369915962219238, + "learning_rate": 3.1747398650925536e-05, + "loss": 3.1175, + "step": 15292 + }, + { + "epoch": 0.1786935750668019, + "grad_norm": 0.15344196557998657, + "learning_rate": 3.174480657849807e-05, + "loss": 3.133, + "step": 15293 + }, + { + "epoch": 0.1787244215323753, + "grad_norm": 0.1491473764181137, + "learning_rate": 3.1742214427868625e-05, + "loss": 3.1071, + "step": 15294 + }, + { + "epoch": 0.17875526799794872, + "grad_norm": 0.16004033386707306, + "learning_rate": 3.1739622199067254e-05, + "loss": 3.073, + "step": 15295 + }, + { + "epoch": 0.17878611446352213, + "grad_norm": 0.154813751578331, + "learning_rate": 3.173702989212401e-05, + "loss": 3.1239, + "step": 15296 + }, + { + "epoch": 0.17881696092909555, + "grad_norm": 0.1470956653356552, + "learning_rate": 3.1734437507068956e-05, + "loss": 3.1129, + "step": 15297 + }, + { + "epoch": 0.17884780739466896, + "grad_norm": 0.1544150561094284, + "learning_rate": 3.173184504393215e-05, + "loss": 3.0956, + "step": 15298 + }, + { + "epoch": 0.17887865386024238, + "grad_norm": 0.15178775787353516, + "learning_rate": 3.172925250274364e-05, + "loss": 3.0887, + "step": 15299 + }, + { + "epoch": 0.1789095003258158, + "grad_norm": 0.15112213790416718, + "learning_rate": 3.172665988353349e-05, + "loss": 3.0977, + "step": 15300 + }, + { + "epoch": 0.1789403467913892, + "grad_norm": 0.15652847290039062, + "learning_rate": 3.172406718633177e-05, + "loss": 3.1292, + "step": 15301 + }, + { + "epoch": 0.17897119325696262, + "grad_norm": 0.1603630930185318, + "learning_rate": 3.172147441116853e-05, + "loss": 3.1001, + "step": 15302 + }, + { + "epoch": 0.17900203972253603, + "grad_norm": 0.15871554613113403, + "learning_rate": 3.1718881558073824e-05, + "loss": 3.1036, + "step": 15303 + }, + { + "epoch": 0.17903288618810945, + "grad_norm": 0.1550714373588562, + "learning_rate": 3.171628862707773e-05, + "loss": 3.0975, + "step": 15304 + }, + { + "epoch": 0.1790637326536829, + "grad_norm": 0.1517011821269989, + "learning_rate": 3.171369561821031e-05, + "loss": 3.0867, + "step": 15305 + }, + { + "epoch": 0.1790945791192563, + "grad_norm": 0.1640353947877884, + "learning_rate": 3.1711102531501615e-05, + "loss": 3.1316, + "step": 15306 + }, + { + "epoch": 0.17912542558482972, + "grad_norm": 0.1543441116809845, + "learning_rate": 3.170850936698172e-05, + "loss": 3.1267, + "step": 15307 + }, + { + "epoch": 0.17915627205040313, + "grad_norm": 0.15556494891643524, + "learning_rate": 3.170591612468069e-05, + "loss": 3.1045, + "step": 15308 + }, + { + "epoch": 0.17918711851597655, + "grad_norm": 0.15185286104679108, + "learning_rate": 3.17033228046286e-05, + "loss": 3.1459, + "step": 15309 + }, + { + "epoch": 0.17921796498154996, + "grad_norm": 0.15215399861335754, + "learning_rate": 3.1700729406855504e-05, + "loss": 3.1366, + "step": 15310 + }, + { + "epoch": 0.17924881144712337, + "grad_norm": 0.15495191514492035, + "learning_rate": 3.1698135931391476e-05, + "loss": 3.0928, + "step": 15311 + }, + { + "epoch": 0.1792796579126968, + "grad_norm": 0.15296411514282227, + "learning_rate": 3.1695542378266595e-05, + "loss": 3.1146, + "step": 15312 + }, + { + "epoch": 0.1793105043782702, + "grad_norm": 0.14926299452781677, + "learning_rate": 3.1692948747510913e-05, + "loss": 3.0878, + "step": 15313 + }, + { + "epoch": 0.17934135084384362, + "grad_norm": 0.1476403921842575, + "learning_rate": 3.169035503915452e-05, + "loss": 3.0883, + "step": 15314 + }, + { + "epoch": 0.17937219730941703, + "grad_norm": 0.16518940031528473, + "learning_rate": 3.168776125322748e-05, + "loss": 3.1569, + "step": 15315 + }, + { + "epoch": 0.17940304377499044, + "grad_norm": 0.15068213641643524, + "learning_rate": 3.168516738975987e-05, + "loss": 3.1308, + "step": 15316 + }, + { + "epoch": 0.1794338902405639, + "grad_norm": 0.15088686347007751, + "learning_rate": 3.168257344878175e-05, + "loss": 3.1243, + "step": 15317 + }, + { + "epoch": 0.1794647367061373, + "grad_norm": 0.15001803636550903, + "learning_rate": 3.1679979430323214e-05, + "loss": 3.1119, + "step": 15318 + }, + { + "epoch": 0.17949558317171072, + "grad_norm": 0.15488551557064056, + "learning_rate": 3.1677385334414335e-05, + "loss": 3.1235, + "step": 15319 + }, + { + "epoch": 0.17952642963728413, + "grad_norm": 0.15561851859092712, + "learning_rate": 3.1674791161085174e-05, + "loss": 3.1222, + "step": 15320 + }, + { + "epoch": 0.17955727610285754, + "grad_norm": 0.15347415208816528, + "learning_rate": 3.167219691036583e-05, + "loss": 3.1357, + "step": 15321 + }, + { + "epoch": 0.17958812256843096, + "grad_norm": 0.1486154943704605, + "learning_rate": 3.1669602582286374e-05, + "loss": 3.0709, + "step": 15322 + }, + { + "epoch": 0.17961896903400437, + "grad_norm": 0.16415561735630035, + "learning_rate": 3.166700817687688e-05, + "loss": 3.1325, + "step": 15323 + }, + { + "epoch": 0.1796498154995778, + "grad_norm": 0.15331794321537018, + "learning_rate": 3.1664413694167424e-05, + "loss": 3.0973, + "step": 15324 + }, + { + "epoch": 0.1796806619651512, + "grad_norm": 0.15425606071949005, + "learning_rate": 3.1661819134188106e-05, + "loss": 3.0975, + "step": 15325 + }, + { + "epoch": 0.17971150843072461, + "grad_norm": 0.1488591730594635, + "learning_rate": 3.1659224496969e-05, + "loss": 3.1221, + "step": 15326 + }, + { + "epoch": 0.17974235489629803, + "grad_norm": 0.1598309427499771, + "learning_rate": 3.165662978254018e-05, + "loss": 3.1249, + "step": 15327 + }, + { + "epoch": 0.17977320136187144, + "grad_norm": 0.15699084103107452, + "learning_rate": 3.165403499093175e-05, + "loss": 3.1236, + "step": 15328 + }, + { + "epoch": 0.17980404782744489, + "grad_norm": 0.15537911653518677, + "learning_rate": 3.165144012217377e-05, + "loss": 3.1072, + "step": 15329 + }, + { + "epoch": 0.1798348942930183, + "grad_norm": 0.1498335748910904, + "learning_rate": 3.1648845176296346e-05, + "loss": 3.1091, + "step": 15330 + }, + { + "epoch": 0.1798657407585917, + "grad_norm": 0.1507221907377243, + "learning_rate": 3.164625015332956e-05, + "loss": 3.1053, + "step": 15331 + }, + { + "epoch": 0.17989658722416513, + "grad_norm": 0.15328647196292877, + "learning_rate": 3.164365505330349e-05, + "loss": 3.1198, + "step": 15332 + }, + { + "epoch": 0.17992743368973854, + "grad_norm": 0.17778752744197845, + "learning_rate": 3.164105987624824e-05, + "loss": 3.1088, + "step": 15333 + }, + { + "epoch": 0.17995828015531196, + "grad_norm": 0.16383709013462067, + "learning_rate": 3.1638464622193886e-05, + "loss": 3.1436, + "step": 15334 + }, + { + "epoch": 0.17998912662088537, + "grad_norm": 0.1578681468963623, + "learning_rate": 3.1635869291170525e-05, + "loss": 3.0948, + "step": 15335 + }, + { + "epoch": 0.18001997308645878, + "grad_norm": 0.14598482847213745, + "learning_rate": 3.163327388320824e-05, + "loss": 3.0973, + "step": 15336 + }, + { + "epoch": 0.1800508195520322, + "grad_norm": 0.15233926475048065, + "learning_rate": 3.163067839833714e-05, + "loss": 3.0917, + "step": 15337 + }, + { + "epoch": 0.1800816660176056, + "grad_norm": 0.1539946347475052, + "learning_rate": 3.1628082836587303e-05, + "loss": 3.129, + "step": 15338 + }, + { + "epoch": 0.18011251248317903, + "grad_norm": 0.15157853066921234, + "learning_rate": 3.162548719798884e-05, + "loss": 3.0875, + "step": 15339 + }, + { + "epoch": 0.18014335894875244, + "grad_norm": 0.15122178196907043, + "learning_rate": 3.162289148257182e-05, + "loss": 3.1272, + "step": 15340 + }, + { + "epoch": 0.18017420541432588, + "grad_norm": 0.14741142094135284, + "learning_rate": 3.1620295690366356e-05, + "loss": 3.0966, + "step": 15341 + }, + { + "epoch": 0.1802050518798993, + "grad_norm": 0.14741109311580658, + "learning_rate": 3.161769982140255e-05, + "loss": 3.1042, + "step": 15342 + }, + { + "epoch": 0.1802358983454727, + "grad_norm": 0.15593934059143066, + "learning_rate": 3.161510387571049e-05, + "loss": 3.1279, + "step": 15343 + }, + { + "epoch": 0.18026674481104613, + "grad_norm": 0.1472696214914322, + "learning_rate": 3.1612507853320265e-05, + "loss": 3.1352, + "step": 15344 + }, + { + "epoch": 0.18029759127661954, + "grad_norm": 0.15411360561847687, + "learning_rate": 3.1609911754261996e-05, + "loss": 3.0954, + "step": 15345 + }, + { + "epoch": 0.18032843774219295, + "grad_norm": 0.4910973012447357, + "learning_rate": 3.1607315578565766e-05, + "loss": 3.1219, + "step": 15346 + }, + { + "epoch": 0.18035928420776637, + "grad_norm": 0.1519385129213333, + "learning_rate": 3.1604719326261687e-05, + "loss": 3.1203, + "step": 15347 + }, + { + "epoch": 0.18039013067333978, + "grad_norm": 0.1606736034154892, + "learning_rate": 3.160212299737985e-05, + "loss": 3.1294, + "step": 15348 + }, + { + "epoch": 0.1804209771389132, + "grad_norm": 0.1625196486711502, + "learning_rate": 3.159952659195038e-05, + "loss": 3.1066, + "step": 15349 + }, + { + "epoch": 0.1804518236044866, + "grad_norm": 0.15880395472049713, + "learning_rate": 3.159693011000334e-05, + "loss": 3.1207, + "step": 15350 + }, + { + "epoch": 0.18048267007006003, + "grad_norm": 0.17539231479167938, + "learning_rate": 3.159433355156888e-05, + "loss": 3.1049, + "step": 15351 + }, + { + "epoch": 0.18051351653563344, + "grad_norm": 0.15668152272701263, + "learning_rate": 3.1591736916677077e-05, + "loss": 3.1091, + "step": 15352 + }, + { + "epoch": 0.18054436300120688, + "grad_norm": 0.15857477486133575, + "learning_rate": 3.158914020535805e-05, + "loss": 3.1033, + "step": 15353 + }, + { + "epoch": 0.1805752094667803, + "grad_norm": 0.15326052904129028, + "learning_rate": 3.15865434176419e-05, + "loss": 3.0963, + "step": 15354 + }, + { + "epoch": 0.1806060559323537, + "grad_norm": 0.15609264373779297, + "learning_rate": 3.158394655355874e-05, + "loss": 3.1408, + "step": 15355 + }, + { + "epoch": 0.18063690239792712, + "grad_norm": 0.16123607754707336, + "learning_rate": 3.1581349613138666e-05, + "loss": 3.1388, + "step": 15356 + }, + { + "epoch": 0.18066774886350054, + "grad_norm": 0.15876100957393646, + "learning_rate": 3.15787525964118e-05, + "loss": 3.1095, + "step": 15357 + }, + { + "epoch": 0.18069859532907395, + "grad_norm": 0.16524599492549896, + "learning_rate": 3.1576155503408257e-05, + "loss": 3.1059, + "step": 15358 + }, + { + "epoch": 0.18072944179464737, + "grad_norm": 0.15982557833194733, + "learning_rate": 3.157355833415814e-05, + "loss": 3.1217, + "step": 15359 + }, + { + "epoch": 0.18076028826022078, + "grad_norm": 0.15451757609844208, + "learning_rate": 3.157096108869157e-05, + "loss": 3.1056, + "step": 15360 + }, + { + "epoch": 0.1807911347257942, + "grad_norm": 0.1598253697156906, + "learning_rate": 3.156836376703865e-05, + "loss": 3.1167, + "step": 15361 + }, + { + "epoch": 0.1808219811913676, + "grad_norm": 0.14786948263645172, + "learning_rate": 3.15657663692295e-05, + "loss": 3.0444, + "step": 15362 + }, + { + "epoch": 0.18085282765694102, + "grad_norm": 0.1552755981683731, + "learning_rate": 3.156316889529423e-05, + "loss": 3.1531, + "step": 15363 + }, + { + "epoch": 0.18088367412251444, + "grad_norm": 0.15315620601177216, + "learning_rate": 3.156057134526296e-05, + "loss": 3.1166, + "step": 15364 + }, + { + "epoch": 0.18091452058808788, + "grad_norm": 0.15432730317115784, + "learning_rate": 3.155797371916582e-05, + "loss": 3.101, + "step": 15365 + }, + { + "epoch": 0.1809453670536613, + "grad_norm": 0.15721583366394043, + "learning_rate": 3.155537601703291e-05, + "loss": 3.1447, + "step": 15366 + }, + { + "epoch": 0.1809762135192347, + "grad_norm": 0.14917168021202087, + "learning_rate": 3.1552778238894354e-05, + "loss": 3.1072, + "step": 15367 + }, + { + "epoch": 0.18100705998480812, + "grad_norm": 0.15522022545337677, + "learning_rate": 3.155018038478027e-05, + "loss": 3.1436, + "step": 15368 + }, + { + "epoch": 0.18103790645038154, + "grad_norm": 0.15921443700790405, + "learning_rate": 3.154758245472078e-05, + "loss": 3.0921, + "step": 15369 + }, + { + "epoch": 0.18106875291595495, + "grad_norm": 0.15720117092132568, + "learning_rate": 3.154498444874601e-05, + "loss": 3.0997, + "step": 15370 + }, + { + "epoch": 0.18109959938152836, + "grad_norm": 0.15311913192272186, + "learning_rate": 3.154238636688607e-05, + "loss": 3.1091, + "step": 15371 + }, + { + "epoch": 0.18113044584710178, + "grad_norm": 0.15492688119411469, + "learning_rate": 3.15397882091711e-05, + "loss": 3.0853, + "step": 15372 + }, + { + "epoch": 0.1811612923126752, + "grad_norm": 0.1631796956062317, + "learning_rate": 3.153718997563121e-05, + "loss": 3.1162, + "step": 15373 + }, + { + "epoch": 0.1811921387782486, + "grad_norm": 0.1521017998456955, + "learning_rate": 3.153459166629653e-05, + "loss": 3.0921, + "step": 15374 + }, + { + "epoch": 0.18122298524382202, + "grad_norm": 0.15436804294586182, + "learning_rate": 3.1531993281197195e-05, + "loss": 3.1478, + "step": 15375 + }, + { + "epoch": 0.18125383170939544, + "grad_norm": 0.15006661415100098, + "learning_rate": 3.1529394820363316e-05, + "loss": 3.0995, + "step": 15376 + }, + { + "epoch": 0.18128467817496888, + "grad_norm": 0.16851280629634857, + "learning_rate": 3.1526796283825033e-05, + "loss": 3.081, + "step": 15377 + }, + { + "epoch": 0.1813155246405423, + "grad_norm": 0.1541167050600052, + "learning_rate": 3.152419767161247e-05, + "loss": 3.0976, + "step": 15378 + }, + { + "epoch": 0.1813463711061157, + "grad_norm": 0.18482467532157898, + "learning_rate": 3.152159898375575e-05, + "loss": 3.1286, + "step": 15379 + }, + { + "epoch": 0.18137721757168912, + "grad_norm": 0.15834523737430573, + "learning_rate": 3.1519000220285014e-05, + "loss": 3.1258, + "step": 15380 + }, + { + "epoch": 0.18140806403726253, + "grad_norm": 0.15442107617855072, + "learning_rate": 3.1516401381230385e-05, + "loss": 3.1237, + "step": 15381 + }, + { + "epoch": 0.18143891050283595, + "grad_norm": 0.1548844873905182, + "learning_rate": 3.151380246662199e-05, + "loss": 3.1085, + "step": 15382 + }, + { + "epoch": 0.18146975696840936, + "grad_norm": 0.15201440453529358, + "learning_rate": 3.151120347648998e-05, + "loss": 3.1361, + "step": 15383 + }, + { + "epoch": 0.18150060343398278, + "grad_norm": 0.14712119102478027, + "learning_rate": 3.150860441086448e-05, + "loss": 3.069, + "step": 15384 + }, + { + "epoch": 0.1815314498995562, + "grad_norm": 0.21777966618537903, + "learning_rate": 3.1506005269775616e-05, + "loss": 3.1142, + "step": 15385 + }, + { + "epoch": 0.1815622963651296, + "grad_norm": 0.15133412182331085, + "learning_rate": 3.1503406053253534e-05, + "loss": 3.1033, + "step": 15386 + }, + { + "epoch": 0.18159314283070302, + "grad_norm": 0.15338467061519623, + "learning_rate": 3.150080676132837e-05, + "loss": 3.1114, + "step": 15387 + }, + { + "epoch": 0.18162398929627643, + "grad_norm": 0.15191635489463806, + "learning_rate": 3.149820739403025e-05, + "loss": 3.1705, + "step": 15388 + }, + { + "epoch": 0.18165483576184988, + "grad_norm": 0.14790309965610504, + "learning_rate": 3.149560795138933e-05, + "loss": 3.1118, + "step": 15389 + }, + { + "epoch": 0.1816856822274233, + "grad_norm": 0.14994224905967712, + "learning_rate": 3.149300843343573e-05, + "loss": 3.1284, + "step": 15390 + }, + { + "epoch": 0.1817165286929967, + "grad_norm": 0.16749638319015503, + "learning_rate": 3.1490408840199595e-05, + "loss": 3.1026, + "step": 15391 + }, + { + "epoch": 0.18174737515857012, + "grad_norm": 0.15149064362049103, + "learning_rate": 3.148780917171108e-05, + "loss": 3.109, + "step": 15392 + }, + { + "epoch": 0.18177822162414353, + "grad_norm": 0.14944115281105042, + "learning_rate": 3.148520942800032e-05, + "loss": 3.1044, + "step": 15393 + }, + { + "epoch": 0.18180906808971695, + "grad_norm": 0.15578986704349518, + "learning_rate": 3.148260960909745e-05, + "loss": 3.1335, + "step": 15394 + }, + { + "epoch": 0.18183991455529036, + "grad_norm": 0.15852603316307068, + "learning_rate": 3.148000971503261e-05, + "loss": 3.095, + "step": 15395 + }, + { + "epoch": 0.18187076102086377, + "grad_norm": 0.15733452141284943, + "learning_rate": 3.147740974583596e-05, + "loss": 3.091, + "step": 15396 + }, + { + "epoch": 0.1819016074864372, + "grad_norm": 0.15052610635757446, + "learning_rate": 3.147480970153763e-05, + "loss": 3.1043, + "step": 15397 + }, + { + "epoch": 0.1819324539520106, + "grad_norm": 0.16015625, + "learning_rate": 3.1472209582167775e-05, + "loss": 3.1157, + "step": 15398 + }, + { + "epoch": 0.18196330041758402, + "grad_norm": 0.151716411113739, + "learning_rate": 3.1469609387756544e-05, + "loss": 3.0885, + "step": 15399 + }, + { + "epoch": 0.18199414688315743, + "grad_norm": 0.15379977226257324, + "learning_rate": 3.1467009118334074e-05, + "loss": 3.1459, + "step": 15400 + }, + { + "epoch": 0.18202499334873087, + "grad_norm": 0.15021884441375732, + "learning_rate": 3.146440877393052e-05, + "loss": 3.1076, + "step": 15401 + }, + { + "epoch": 0.1820558398143043, + "grad_norm": 0.15283356606960297, + "learning_rate": 3.146180835457603e-05, + "loss": 3.1099, + "step": 15402 + }, + { + "epoch": 0.1820866862798777, + "grad_norm": 0.1577957719564438, + "learning_rate": 3.1459207860300754e-05, + "loss": 3.1134, + "step": 15403 + }, + { + "epoch": 0.18211753274545112, + "grad_norm": 0.14856278896331787, + "learning_rate": 3.145660729113485e-05, + "loss": 3.096, + "step": 15404 + }, + { + "epoch": 0.18214837921102453, + "grad_norm": 0.15066087245941162, + "learning_rate": 3.145400664710847e-05, + "loss": 3.0922, + "step": 15405 + }, + { + "epoch": 0.18217922567659794, + "grad_norm": 0.15096111595630646, + "learning_rate": 3.145140592825175e-05, + "loss": 3.1268, + "step": 15406 + }, + { + "epoch": 0.18221007214217136, + "grad_norm": 0.1552966684103012, + "learning_rate": 3.144880513459485e-05, + "loss": 3.0933, + "step": 15407 + }, + { + "epoch": 0.18224091860774477, + "grad_norm": 0.16310589015483856, + "learning_rate": 3.144620426616794e-05, + "loss": 3.0991, + "step": 15408 + }, + { + "epoch": 0.1822717650733182, + "grad_norm": 0.15786467492580414, + "learning_rate": 3.144360332300116e-05, + "loss": 3.1293, + "step": 15409 + }, + { + "epoch": 0.1823026115388916, + "grad_norm": 0.16228102147579193, + "learning_rate": 3.1441002305124684e-05, + "loss": 3.146, + "step": 15410 + }, + { + "epoch": 0.18233345800446502, + "grad_norm": 0.1514253318309784, + "learning_rate": 3.143840121256865e-05, + "loss": 3.1661, + "step": 15411 + }, + { + "epoch": 0.18236430447003843, + "grad_norm": 0.1695687621831894, + "learning_rate": 3.143580004536322e-05, + "loss": 3.1145, + "step": 15412 + }, + { + "epoch": 0.18239515093561187, + "grad_norm": 0.15756317973136902, + "learning_rate": 3.1433198803538555e-05, + "loss": 3.1298, + "step": 15413 + }, + { + "epoch": 0.18242599740118529, + "grad_norm": 0.16633090376853943, + "learning_rate": 3.1430597487124815e-05, + "loss": 3.0796, + "step": 15414 + }, + { + "epoch": 0.1824568438667587, + "grad_norm": 0.15830452740192413, + "learning_rate": 3.1427996096152165e-05, + "loss": 3.1203, + "step": 15415 + }, + { + "epoch": 0.18248769033233211, + "grad_norm": 0.1551935374736786, + "learning_rate": 3.142539463065076e-05, + "loss": 3.1063, + "step": 15416 + }, + { + "epoch": 0.18251853679790553, + "grad_norm": 0.15329445898532867, + "learning_rate": 3.142279309065077e-05, + "loss": 3.0939, + "step": 15417 + }, + { + "epoch": 0.18254938326347894, + "grad_norm": 0.15048664808273315, + "learning_rate": 3.1420191476182356e-05, + "loss": 3.1172, + "step": 15418 + }, + { + "epoch": 0.18258022972905236, + "grad_norm": 0.1558474898338318, + "learning_rate": 3.141758978727567e-05, + "loss": 3.1278, + "step": 15419 + }, + { + "epoch": 0.18261107619462577, + "grad_norm": 0.1616695076227188, + "learning_rate": 3.141498802396089e-05, + "loss": 3.1064, + "step": 15420 + }, + { + "epoch": 0.18264192266019919, + "grad_norm": 0.15212127566337585, + "learning_rate": 3.141238618626818e-05, + "loss": 3.0919, + "step": 15421 + }, + { + "epoch": 0.1826727691257726, + "grad_norm": 0.1569114476442337, + "learning_rate": 3.140978427422772e-05, + "loss": 3.1342, + "step": 15422 + }, + { + "epoch": 0.182703615591346, + "grad_norm": 0.15361319482326508, + "learning_rate": 3.140718228786964e-05, + "loss": 3.1457, + "step": 15423 + }, + { + "epoch": 0.18273446205691943, + "grad_norm": 0.16792340576648712, + "learning_rate": 3.1404580227224136e-05, + "loss": 3.1269, + "step": 15424 + }, + { + "epoch": 0.18276530852249287, + "grad_norm": 0.16300618648529053, + "learning_rate": 3.140197809232138e-05, + "loss": 3.1131, + "step": 15425 + }, + { + "epoch": 0.18279615498806628, + "grad_norm": 0.15915457904338837, + "learning_rate": 3.139937588319153e-05, + "loss": 3.08, + "step": 15426 + }, + { + "epoch": 0.1828270014536397, + "grad_norm": 0.15299208462238312, + "learning_rate": 3.139677359986477e-05, + "loss": 3.13, + "step": 15427 + }, + { + "epoch": 0.1828578479192131, + "grad_norm": 0.15207462012767792, + "learning_rate": 3.139417124237127e-05, + "loss": 3.1245, + "step": 15428 + }, + { + "epoch": 0.18288869438478653, + "grad_norm": 0.15540625154972076, + "learning_rate": 3.1391568810741175e-05, + "loss": 3.0887, + "step": 15429 + }, + { + "epoch": 0.18291954085035994, + "grad_norm": 0.15207041800022125, + "learning_rate": 3.13889663050047e-05, + "loss": 3.1461, + "step": 15430 + }, + { + "epoch": 0.18295038731593335, + "grad_norm": 0.18501682579517365, + "learning_rate": 3.1386363725191995e-05, + "loss": 3.0953, + "step": 15431 + }, + { + "epoch": 0.18298123378150677, + "grad_norm": 0.15081365406513214, + "learning_rate": 3.138376107133323e-05, + "loss": 3.0876, + "step": 15432 + }, + { + "epoch": 0.18301208024708018, + "grad_norm": 0.15648822486400604, + "learning_rate": 3.138115834345861e-05, + "loss": 3.1029, + "step": 15433 + }, + { + "epoch": 0.1830429267126536, + "grad_norm": 0.15466441214084625, + "learning_rate": 3.1378555541598276e-05, + "loss": 3.127, + "step": 15434 + }, + { + "epoch": 0.183073773178227, + "grad_norm": 0.15177910029888153, + "learning_rate": 3.1375952665782426e-05, + "loss": 3.1281, + "step": 15435 + }, + { + "epoch": 0.18310461964380043, + "grad_norm": 0.14956353604793549, + "learning_rate": 3.137334971604125e-05, + "loss": 3.1352, + "step": 15436 + }, + { + "epoch": 0.18313546610937387, + "grad_norm": 0.15880145132541656, + "learning_rate": 3.13707466924049e-05, + "loss": 3.1355, + "step": 15437 + }, + { + "epoch": 0.18316631257494728, + "grad_norm": 0.1514071226119995, + "learning_rate": 3.136814359490357e-05, + "loss": 3.1162, + "step": 15438 + }, + { + "epoch": 0.1831971590405207, + "grad_norm": 0.1512114554643631, + "learning_rate": 3.136554042356745e-05, + "loss": 3.1284, + "step": 15439 + }, + { + "epoch": 0.1832280055060941, + "grad_norm": 0.14828206598758698, + "learning_rate": 3.136293717842671e-05, + "loss": 3.1342, + "step": 15440 + }, + { + "epoch": 0.18325885197166752, + "grad_norm": 0.15208514034748077, + "learning_rate": 3.1360333859511534e-05, + "loss": 3.1165, + "step": 15441 + }, + { + "epoch": 0.18328969843724094, + "grad_norm": 0.15599152445793152, + "learning_rate": 3.135773046685212e-05, + "loss": 3.1391, + "step": 15442 + }, + { + "epoch": 0.18332054490281435, + "grad_norm": 0.15334421396255493, + "learning_rate": 3.1355127000478636e-05, + "loss": 3.1142, + "step": 15443 + }, + { + "epoch": 0.18335139136838777, + "grad_norm": 0.148986354470253, + "learning_rate": 3.135252346042127e-05, + "loss": 3.0916, + "step": 15444 + }, + { + "epoch": 0.18338223783396118, + "grad_norm": 0.1518632471561432, + "learning_rate": 3.1349919846710216e-05, + "loss": 3.1031, + "step": 15445 + }, + { + "epoch": 0.1834130842995346, + "grad_norm": 0.1502721607685089, + "learning_rate": 3.1347316159375654e-05, + "loss": 3.1472, + "step": 15446 + }, + { + "epoch": 0.183443930765108, + "grad_norm": 0.16670837998390198, + "learning_rate": 3.134471239844777e-05, + "loss": 3.1221, + "step": 15447 + }, + { + "epoch": 0.18347477723068142, + "grad_norm": 0.15536299347877502, + "learning_rate": 3.1342108563956774e-05, + "loss": 3.1071, + "step": 15448 + }, + { + "epoch": 0.18350562369625487, + "grad_norm": 0.1518263965845108, + "learning_rate": 3.133950465593283e-05, + "loss": 3.1181, + "step": 15449 + }, + { + "epoch": 0.18353647016182828, + "grad_norm": 0.1534506380558014, + "learning_rate": 3.133690067440614e-05, + "loss": 3.1294, + "step": 15450 + }, + { + "epoch": 0.1835673166274017, + "grad_norm": 0.1517329216003418, + "learning_rate": 3.13342966194069e-05, + "loss": 3.1013, + "step": 15451 + }, + { + "epoch": 0.1835981630929751, + "grad_norm": 0.15166155993938446, + "learning_rate": 3.133169249096529e-05, + "loss": 3.0706, + "step": 15452 + }, + { + "epoch": 0.18362900955854852, + "grad_norm": 0.1497327983379364, + "learning_rate": 3.132908828911151e-05, + "loss": 3.084, + "step": 15453 + }, + { + "epoch": 0.18365985602412194, + "grad_norm": 0.15679588913917542, + "learning_rate": 3.1326484013875764e-05, + "loss": 3.1275, + "step": 15454 + }, + { + "epoch": 0.18369070248969535, + "grad_norm": 0.1599162071943283, + "learning_rate": 3.1323879665288235e-05, + "loss": 3.1235, + "step": 15455 + }, + { + "epoch": 0.18372154895526877, + "grad_norm": 0.16225633025169373, + "learning_rate": 3.1321275243379115e-05, + "loss": 3.1077, + "step": 15456 + }, + { + "epoch": 0.18375239542084218, + "grad_norm": 0.15404091775417328, + "learning_rate": 3.1318670748178605e-05, + "loss": 3.0878, + "step": 15457 + }, + { + "epoch": 0.1837832418864156, + "grad_norm": 0.15109431743621826, + "learning_rate": 3.131606617971691e-05, + "loss": 3.1288, + "step": 15458 + }, + { + "epoch": 0.183814088351989, + "grad_norm": 0.1494722068309784, + "learning_rate": 3.1313461538024226e-05, + "loss": 3.0929, + "step": 15459 + }, + { + "epoch": 0.18384493481756242, + "grad_norm": 0.1516558825969696, + "learning_rate": 3.1310856823130746e-05, + "loss": 3.1255, + "step": 15460 + }, + { + "epoch": 0.18387578128313586, + "grad_norm": 0.15613117814064026, + "learning_rate": 3.130825203506668e-05, + "loss": 3.0889, + "step": 15461 + }, + { + "epoch": 0.18390662774870928, + "grad_norm": 0.1495523750782013, + "learning_rate": 3.1305647173862216e-05, + "loss": 3.1126, + "step": 15462 + }, + { + "epoch": 0.1839374742142827, + "grad_norm": 0.1598450392484665, + "learning_rate": 3.130304223954756e-05, + "loss": 3.1344, + "step": 15463 + }, + { + "epoch": 0.1839683206798561, + "grad_norm": 0.15223988890647888, + "learning_rate": 3.130043723215291e-05, + "loss": 3.1258, + "step": 15464 + }, + { + "epoch": 0.18399916714542952, + "grad_norm": 0.15212176740169525, + "learning_rate": 3.12978321517085e-05, + "loss": 3.1188, + "step": 15465 + }, + { + "epoch": 0.18403001361100293, + "grad_norm": 0.15729936957359314, + "learning_rate": 3.129522699824449e-05, + "loss": 3.1037, + "step": 15466 + }, + { + "epoch": 0.18406086007657635, + "grad_norm": 0.16792325675487518, + "learning_rate": 3.1292621771791116e-05, + "loss": 3.1326, + "step": 15467 + }, + { + "epoch": 0.18409170654214976, + "grad_norm": 0.15942394733428955, + "learning_rate": 3.129001647237857e-05, + "loss": 3.1069, + "step": 15468 + }, + { + "epoch": 0.18412255300772318, + "grad_norm": 0.1545361876487732, + "learning_rate": 3.128741110003706e-05, + "loss": 3.1558, + "step": 15469 + }, + { + "epoch": 0.1841533994732966, + "grad_norm": 0.16093434393405914, + "learning_rate": 3.12848056547968e-05, + "loss": 3.1378, + "step": 15470 + }, + { + "epoch": 0.18418424593887, + "grad_norm": 0.1663791686296463, + "learning_rate": 3.1282200136688e-05, + "loss": 3.1502, + "step": 15471 + }, + { + "epoch": 0.18421509240444342, + "grad_norm": 0.15937037765979767, + "learning_rate": 3.1279594545740856e-05, + "loss": 3.1139, + "step": 15472 + }, + { + "epoch": 0.18424593887001686, + "grad_norm": 0.15150444209575653, + "learning_rate": 3.127698888198559e-05, + "loss": 3.1273, + "step": 15473 + }, + { + "epoch": 0.18427678533559028, + "grad_norm": 0.1558447629213333, + "learning_rate": 3.127438314545241e-05, + "loss": 3.137, + "step": 15474 + }, + { + "epoch": 0.1843076318011637, + "grad_norm": 0.158274844288826, + "learning_rate": 3.127177733617153e-05, + "loss": 3.1021, + "step": 15475 + }, + { + "epoch": 0.1843384782667371, + "grad_norm": 0.15689195692539215, + "learning_rate": 3.126917145417316e-05, + "loss": 3.1155, + "step": 15476 + }, + { + "epoch": 0.18436932473231052, + "grad_norm": 0.15271392464637756, + "learning_rate": 3.126656549948752e-05, + "loss": 3.1167, + "step": 15477 + }, + { + "epoch": 0.18440017119788393, + "grad_norm": 0.16024188697338104, + "learning_rate": 3.126395947214481e-05, + "loss": 3.1147, + "step": 15478 + }, + { + "epoch": 0.18443101766345735, + "grad_norm": 0.1528458297252655, + "learning_rate": 3.1261353372175255e-05, + "loss": 3.0939, + "step": 15479 + }, + { + "epoch": 0.18446186412903076, + "grad_norm": 0.15132872760295868, + "learning_rate": 3.1258747199609075e-05, + "loss": 3.1373, + "step": 15480 + }, + { + "epoch": 0.18449271059460418, + "grad_norm": 0.18356941640377045, + "learning_rate": 3.125614095447648e-05, + "loss": 3.1262, + "step": 15481 + }, + { + "epoch": 0.1845235570601776, + "grad_norm": 0.15734361112117767, + "learning_rate": 3.1253534636807685e-05, + "loss": 3.128, + "step": 15482 + }, + { + "epoch": 0.184554403525751, + "grad_norm": 0.16319406032562256, + "learning_rate": 3.125092824663292e-05, + "loss": 3.1127, + "step": 15483 + }, + { + "epoch": 0.18458524999132442, + "grad_norm": 0.1691102683544159, + "learning_rate": 3.12483217839824e-05, + "loss": 3.1299, + "step": 15484 + }, + { + "epoch": 0.18461609645689786, + "grad_norm": 0.15210281312465668, + "learning_rate": 3.124571524888634e-05, + "loss": 3.1119, + "step": 15485 + }, + { + "epoch": 0.18464694292247127, + "grad_norm": 0.15688969194889069, + "learning_rate": 3.1243108641374955e-05, + "loss": 3.145, + "step": 15486 + }, + { + "epoch": 0.1846777893880447, + "grad_norm": 0.1537049561738968, + "learning_rate": 3.12405019614785e-05, + "loss": 3.1219, + "step": 15487 + }, + { + "epoch": 0.1847086358536181, + "grad_norm": 0.15835605561733246, + "learning_rate": 3.1237895209227154e-05, + "loss": 3.0894, + "step": 15488 + }, + { + "epoch": 0.18473948231919152, + "grad_norm": 0.15605975687503815, + "learning_rate": 3.1235288384651176e-05, + "loss": 3.1215, + "step": 15489 + }, + { + "epoch": 0.18477032878476493, + "grad_norm": 0.15691566467285156, + "learning_rate": 3.123268148778076e-05, + "loss": 3.1043, + "step": 15490 + }, + { + "epoch": 0.18480117525033835, + "grad_norm": 0.15597130358219147, + "learning_rate": 3.1230074518646154e-05, + "loss": 3.0967, + "step": 15491 + }, + { + "epoch": 0.18483202171591176, + "grad_norm": 0.16025744378566742, + "learning_rate": 3.122746747727759e-05, + "loss": 3.1089, + "step": 15492 + }, + { + "epoch": 0.18486286818148517, + "grad_norm": 0.15969489514827728, + "learning_rate": 3.122486036370528e-05, + "loss": 3.084, + "step": 15493 + }, + { + "epoch": 0.1848937146470586, + "grad_norm": 0.1594599187374115, + "learning_rate": 3.1222253177959446e-05, + "loss": 3.1339, + "step": 15494 + }, + { + "epoch": 0.184924561112632, + "grad_norm": 0.1581120491027832, + "learning_rate": 3.1219645920070326e-05, + "loss": 3.1091, + "step": 15495 + }, + { + "epoch": 0.18495540757820542, + "grad_norm": 0.16015934944152832, + "learning_rate": 3.121703859006815e-05, + "loss": 3.108, + "step": 15496 + }, + { + "epoch": 0.18498625404377886, + "grad_norm": 0.15785694122314453, + "learning_rate": 3.121443118798314e-05, + "loss": 3.1225, + "step": 15497 + }, + { + "epoch": 0.18501710050935227, + "grad_norm": 0.15253812074661255, + "learning_rate": 3.121182371384555e-05, + "loss": 3.1377, + "step": 15498 + }, + { + "epoch": 0.1850479469749257, + "grad_norm": 0.15388807654380798, + "learning_rate": 3.12092161676856e-05, + "loss": 3.1031, + "step": 15499 + }, + { + "epoch": 0.1850787934404991, + "grad_norm": 0.16270364820957184, + "learning_rate": 3.1206608549533504e-05, + "loss": 3.1319, + "step": 15500 + }, + { + "epoch": 0.18510963990607252, + "grad_norm": 0.15778017044067383, + "learning_rate": 3.1204000859419525e-05, + "loss": 3.1514, + "step": 15501 + }, + { + "epoch": 0.18514048637164593, + "grad_norm": 0.16742488741874695, + "learning_rate": 3.1201393097373874e-05, + "loss": 3.1113, + "step": 15502 + }, + { + "epoch": 0.18517133283721934, + "grad_norm": 0.1568191945552826, + "learning_rate": 3.1198785263426795e-05, + "loss": 3.0994, + "step": 15503 + }, + { + "epoch": 0.18520217930279276, + "grad_norm": 0.1532638967037201, + "learning_rate": 3.119617735760854e-05, + "loss": 3.103, + "step": 15504 + }, + { + "epoch": 0.18523302576836617, + "grad_norm": 0.15952278673648834, + "learning_rate": 3.1193569379949315e-05, + "loss": 3.1237, + "step": 15505 + }, + { + "epoch": 0.18526387223393959, + "grad_norm": 0.16910289227962494, + "learning_rate": 3.1190961330479384e-05, + "loss": 3.114, + "step": 15506 + }, + { + "epoch": 0.185294718699513, + "grad_norm": 0.15535835921764374, + "learning_rate": 3.118835320922897e-05, + "loss": 3.0867, + "step": 15507 + }, + { + "epoch": 0.18532556516508641, + "grad_norm": 0.15702733397483826, + "learning_rate": 3.118574501622832e-05, + "loss": 3.0893, + "step": 15508 + }, + { + "epoch": 0.18535641163065986, + "grad_norm": 0.15949386358261108, + "learning_rate": 3.118313675150768e-05, + "loss": 3.0702, + "step": 15509 + }, + { + "epoch": 0.18538725809623327, + "grad_norm": 0.1574220359325409, + "learning_rate": 3.118052841509729e-05, + "loss": 3.1016, + "step": 15510 + }, + { + "epoch": 0.18541810456180668, + "grad_norm": 0.15560148656368256, + "learning_rate": 3.1177920007027385e-05, + "loss": 3.1098, + "step": 15511 + }, + { + "epoch": 0.1854489510273801, + "grad_norm": 0.17169024050235748, + "learning_rate": 3.1175311527328204e-05, + "loss": 3.0919, + "step": 15512 + }, + { + "epoch": 0.1854797974929535, + "grad_norm": 0.1527387797832489, + "learning_rate": 3.117270297603e-05, + "loss": 3.1079, + "step": 15513 + }, + { + "epoch": 0.18551064395852693, + "grad_norm": 0.15411429107189178, + "learning_rate": 3.1170094353163007e-05, + "loss": 3.1129, + "step": 15514 + }, + { + "epoch": 0.18554149042410034, + "grad_norm": 0.16238179802894592, + "learning_rate": 3.116748565875749e-05, + "loss": 3.1081, + "step": 15515 + }, + { + "epoch": 0.18557233688967376, + "grad_norm": 0.15232057869434357, + "learning_rate": 3.116487689284367e-05, + "loss": 3.1128, + "step": 15516 + }, + { + "epoch": 0.18560318335524717, + "grad_norm": 0.15357837080955505, + "learning_rate": 3.1162268055451815e-05, + "loss": 3.0946, + "step": 15517 + }, + { + "epoch": 0.18563402982082058, + "grad_norm": 0.1669347733259201, + "learning_rate": 3.115965914661217e-05, + "loss": 3.1015, + "step": 15518 + }, + { + "epoch": 0.185664876286394, + "grad_norm": 0.15687549114227295, + "learning_rate": 3.115705016635497e-05, + "loss": 3.1323, + "step": 15519 + }, + { + "epoch": 0.1856957227519674, + "grad_norm": 0.17387516796588898, + "learning_rate": 3.1154441114710474e-05, + "loss": 3.1205, + "step": 15520 + }, + { + "epoch": 0.18572656921754085, + "grad_norm": 0.1482793688774109, + "learning_rate": 3.115183199170893e-05, + "loss": 3.1364, + "step": 15521 + }, + { + "epoch": 0.18575741568311427, + "grad_norm": 0.16848637163639069, + "learning_rate": 3.11492227973806e-05, + "loss": 3.1321, + "step": 15522 + }, + { + "epoch": 0.18578826214868768, + "grad_norm": 0.1599651724100113, + "learning_rate": 3.114661353175572e-05, + "loss": 3.12, + "step": 15523 + }, + { + "epoch": 0.1858191086142611, + "grad_norm": 0.1735348403453827, + "learning_rate": 3.1144004194864554e-05, + "loss": 3.0819, + "step": 15524 + }, + { + "epoch": 0.1858499550798345, + "grad_norm": 0.1512186974287033, + "learning_rate": 3.1141394786737344e-05, + "loss": 3.0838, + "step": 15525 + }, + { + "epoch": 0.18588080154540793, + "grad_norm": 0.16364184021949768, + "learning_rate": 3.113878530740436e-05, + "loss": 3.1568, + "step": 15526 + }, + { + "epoch": 0.18591164801098134, + "grad_norm": 0.15541410446166992, + "learning_rate": 3.113617575689585e-05, + "loss": 3.106, + "step": 15527 + }, + { + "epoch": 0.18594249447655475, + "grad_norm": 0.15544891357421875, + "learning_rate": 3.113356613524207e-05, + "loss": 3.1152, + "step": 15528 + }, + { + "epoch": 0.18597334094212817, + "grad_norm": 0.15123656392097473, + "learning_rate": 3.1130956442473264e-05, + "loss": 3.129, + "step": 15529 + }, + { + "epoch": 0.18600418740770158, + "grad_norm": 0.16385383903980255, + "learning_rate": 3.112834667861971e-05, + "loss": 3.1141, + "step": 15530 + }, + { + "epoch": 0.186035033873275, + "grad_norm": 0.14877668023109436, + "learning_rate": 3.1125736843711654e-05, + "loss": 3.084, + "step": 15531 + }, + { + "epoch": 0.1860658803388484, + "grad_norm": 0.16254545748233795, + "learning_rate": 3.1123126937779366e-05, + "loss": 3.1154, + "step": 15532 + }, + { + "epoch": 0.18609672680442185, + "grad_norm": 0.16011391580104828, + "learning_rate": 3.11205169608531e-05, + "loss": 3.1353, + "step": 15533 + }, + { + "epoch": 0.18612757326999527, + "grad_norm": 0.15968534350395203, + "learning_rate": 3.1117906912963124e-05, + "loss": 3.1345, + "step": 15534 + }, + { + "epoch": 0.18615841973556868, + "grad_norm": 0.16925707459449768, + "learning_rate": 3.111529679413969e-05, + "loss": 3.1175, + "step": 15535 + }, + { + "epoch": 0.1861892662011421, + "grad_norm": 0.15397344529628754, + "learning_rate": 3.111268660441306e-05, + "loss": 3.1267, + "step": 15536 + }, + { + "epoch": 0.1862201126667155, + "grad_norm": 0.1681673675775528, + "learning_rate": 3.1110076343813506e-05, + "loss": 3.1246, + "step": 15537 + }, + { + "epoch": 0.18625095913228892, + "grad_norm": 0.1551407128572464, + "learning_rate": 3.110746601237128e-05, + "loss": 3.1287, + "step": 15538 + }, + { + "epoch": 0.18628180559786234, + "grad_norm": 0.16335266828536987, + "learning_rate": 3.110485561011667e-05, + "loss": 3.1269, + "step": 15539 + }, + { + "epoch": 0.18631265206343575, + "grad_norm": 0.16841275990009308, + "learning_rate": 3.1102245137079914e-05, + "loss": 3.103, + "step": 15540 + }, + { + "epoch": 0.18634349852900917, + "grad_norm": 0.1531207412481308, + "learning_rate": 3.10996345932913e-05, + "loss": 3.1098, + "step": 15541 + }, + { + "epoch": 0.18637434499458258, + "grad_norm": 0.16676019132137299, + "learning_rate": 3.109702397878109e-05, + "loss": 3.1139, + "step": 15542 + }, + { + "epoch": 0.186405191460156, + "grad_norm": 0.1561344861984253, + "learning_rate": 3.1094413293579547e-05, + "loss": 3.0755, + "step": 15543 + }, + { + "epoch": 0.1864360379257294, + "grad_norm": 0.1616479456424713, + "learning_rate": 3.1091802537716943e-05, + "loss": 3.1144, + "step": 15544 + }, + { + "epoch": 0.18646688439130285, + "grad_norm": 0.15679219365119934, + "learning_rate": 3.1089191711223556e-05, + "loss": 3.1316, + "step": 15545 + }, + { + "epoch": 0.18649773085687626, + "grad_norm": 0.16976332664489746, + "learning_rate": 3.108658081412965e-05, + "loss": 3.1309, + "step": 15546 + }, + { + "epoch": 0.18652857732244968, + "grad_norm": 0.15700499713420868, + "learning_rate": 3.108396984646549e-05, + "loss": 3.0985, + "step": 15547 + }, + { + "epoch": 0.1865594237880231, + "grad_norm": 0.1509494185447693, + "learning_rate": 3.108135880826136e-05, + "loss": 3.1495, + "step": 15548 + }, + { + "epoch": 0.1865902702535965, + "grad_norm": 0.15266159176826477, + "learning_rate": 3.107874769954753e-05, + "loss": 3.0939, + "step": 15549 + }, + { + "epoch": 0.18662111671916992, + "grad_norm": 0.16657766699790955, + "learning_rate": 3.107613652035427e-05, + "loss": 3.1106, + "step": 15550 + }, + { + "epoch": 0.18665196318474334, + "grad_norm": 0.15759965777397156, + "learning_rate": 3.107352527071186e-05, + "loss": 3.1524, + "step": 15551 + }, + { + "epoch": 0.18668280965031675, + "grad_norm": 0.15654894709587097, + "learning_rate": 3.107091395065057e-05, + "loss": 3.1144, + "step": 15552 + }, + { + "epoch": 0.18671365611589016, + "grad_norm": 0.1577141135931015, + "learning_rate": 3.1068302560200684e-05, + "loss": 3.1259, + "step": 15553 + }, + { + "epoch": 0.18674450258146358, + "grad_norm": 0.15629930794239044, + "learning_rate": 3.106569109939248e-05, + "loss": 3.1435, + "step": 15554 + }, + { + "epoch": 0.186775349047037, + "grad_norm": 0.15088582038879395, + "learning_rate": 3.106307956825624e-05, + "loss": 3.1159, + "step": 15555 + }, + { + "epoch": 0.18680619551261043, + "grad_norm": 0.1537705808877945, + "learning_rate": 3.106046796682222e-05, + "loss": 3.0845, + "step": 15556 + }, + { + "epoch": 0.18683704197818385, + "grad_norm": 0.6230531930923462, + "learning_rate": 3.105785629512072e-05, + "loss": 3.1099, + "step": 15557 + }, + { + "epoch": 0.18686788844375726, + "grad_norm": 0.15124623477458954, + "learning_rate": 3.105524455318201e-05, + "loss": 3.1204, + "step": 15558 + }, + { + "epoch": 0.18689873490933068, + "grad_norm": 0.15482831001281738, + "learning_rate": 3.105263274103639e-05, + "loss": 3.1361, + "step": 15559 + }, + { + "epoch": 0.1869295813749041, + "grad_norm": 0.15251770615577698, + "learning_rate": 3.105002085871413e-05, + "loss": 3.1177, + "step": 15560 + }, + { + "epoch": 0.1869604278404775, + "grad_norm": 0.15690238773822784, + "learning_rate": 3.104740890624551e-05, + "loss": 3.0655, + "step": 15561 + }, + { + "epoch": 0.18699127430605092, + "grad_norm": 0.15629565715789795, + "learning_rate": 3.104479688366082e-05, + "loss": 3.1349, + "step": 15562 + }, + { + "epoch": 0.18702212077162433, + "grad_norm": 0.1556919366121292, + "learning_rate": 3.104218479099034e-05, + "loss": 3.0909, + "step": 15563 + }, + { + "epoch": 0.18705296723719775, + "grad_norm": 0.15768170356750488, + "learning_rate": 3.103957262826436e-05, + "loss": 3.0918, + "step": 15564 + }, + { + "epoch": 0.18708381370277116, + "grad_norm": 0.15806786715984344, + "learning_rate": 3.103696039551316e-05, + "loss": 3.1235, + "step": 15565 + }, + { + "epoch": 0.18711466016834458, + "grad_norm": 0.16050253808498383, + "learning_rate": 3.103434809276705e-05, + "loss": 3.1084, + "step": 15566 + }, + { + "epoch": 0.187145506633918, + "grad_norm": 0.15391772985458374, + "learning_rate": 3.103173572005628e-05, + "loss": 3.1175, + "step": 15567 + }, + { + "epoch": 0.18717635309949143, + "grad_norm": 0.1694839745759964, + "learning_rate": 3.102912327741117e-05, + "loss": 3.1251, + "step": 15568 + }, + { + "epoch": 0.18720719956506485, + "grad_norm": 0.1555030643939972, + "learning_rate": 3.102651076486199e-05, + "loss": 3.1288, + "step": 15569 + }, + { + "epoch": 0.18723804603063826, + "grad_norm": 0.16340819001197815, + "learning_rate": 3.1023898182439046e-05, + "loss": 3.1256, + "step": 15570 + }, + { + "epoch": 0.18726889249621168, + "grad_norm": 0.15376710891723633, + "learning_rate": 3.102128553017262e-05, + "loss": 3.1057, + "step": 15571 + }, + { + "epoch": 0.1872997389617851, + "grad_norm": 0.15938544273376465, + "learning_rate": 3.1018672808093005e-05, + "loss": 3.0894, + "step": 15572 + }, + { + "epoch": 0.1873305854273585, + "grad_norm": 0.16607466340065002, + "learning_rate": 3.10160600162305e-05, + "loss": 3.117, + "step": 15573 + }, + { + "epoch": 0.18736143189293192, + "grad_norm": 0.15889191627502441, + "learning_rate": 3.101344715461539e-05, + "loss": 3.0927, + "step": 15574 + }, + { + "epoch": 0.18739227835850533, + "grad_norm": 0.1641755849123001, + "learning_rate": 3.101083422327797e-05, + "loss": 3.1326, + "step": 15575 + }, + { + "epoch": 0.18742312482407875, + "grad_norm": 0.1596289426088333, + "learning_rate": 3.1008221222248547e-05, + "loss": 3.1489, + "step": 15576 + }, + { + "epoch": 0.18745397128965216, + "grad_norm": 0.15971966087818146, + "learning_rate": 3.1005608151557405e-05, + "loss": 3.1165, + "step": 15577 + }, + { + "epoch": 0.18748481775522557, + "grad_norm": 0.15304511785507202, + "learning_rate": 3.1002995011234856e-05, + "loss": 3.1086, + "step": 15578 + }, + { + "epoch": 0.187515664220799, + "grad_norm": 0.15082013607025146, + "learning_rate": 3.1000381801311176e-05, + "loss": 3.0876, + "step": 15579 + }, + { + "epoch": 0.18754651068637243, + "grad_norm": 0.1539013534784317, + "learning_rate": 3.0997768521816676e-05, + "loss": 3.0913, + "step": 15580 + }, + { + "epoch": 0.18757735715194584, + "grad_norm": 0.153211310505867, + "learning_rate": 3.099515517278166e-05, + "loss": 3.124, + "step": 15581 + }, + { + "epoch": 0.18760820361751926, + "grad_norm": 0.1925879865884781, + "learning_rate": 3.099254175423641e-05, + "loss": 3.1001, + "step": 15582 + }, + { + "epoch": 0.18763905008309267, + "grad_norm": 0.1495426744222641, + "learning_rate": 3.098992826621126e-05, + "loss": 3.1132, + "step": 15583 + }, + { + "epoch": 0.1876698965486661, + "grad_norm": 0.1525510996580124, + "learning_rate": 3.0987314708736474e-05, + "loss": 3.1007, + "step": 15584 + }, + { + "epoch": 0.1877007430142395, + "grad_norm": 0.14778777956962585, + "learning_rate": 3.0984701081842374e-05, + "loss": 3.0833, + "step": 15585 + }, + { + "epoch": 0.18773158947981292, + "grad_norm": 0.152093306183815, + "learning_rate": 3.098208738555927e-05, + "loss": 3.1107, + "step": 15586 + }, + { + "epoch": 0.18776243594538633, + "grad_norm": 0.14721068739891052, + "learning_rate": 3.097947361991745e-05, + "loss": 3.1111, + "step": 15587 + }, + { + "epoch": 0.18779328241095974, + "grad_norm": 0.15510615706443787, + "learning_rate": 3.097685978494723e-05, + "loss": 3.106, + "step": 15588 + }, + { + "epoch": 0.18782412887653316, + "grad_norm": 0.15303659439086914, + "learning_rate": 3.097424588067891e-05, + "loss": 3.1508, + "step": 15589 + }, + { + "epoch": 0.18785497534210657, + "grad_norm": 0.15772747993469238, + "learning_rate": 3.0971631907142794e-05, + "loss": 3.1091, + "step": 15590 + }, + { + "epoch": 0.18788582180768, + "grad_norm": 0.15511594712734222, + "learning_rate": 3.09690178643692e-05, + "loss": 3.1277, + "step": 15591 + }, + { + "epoch": 0.18791666827325343, + "grad_norm": 0.14928938448429108, + "learning_rate": 3.096640375238844e-05, + "loss": 3.0955, + "step": 15592 + }, + { + "epoch": 0.18794751473882684, + "grad_norm": 0.14825163781642914, + "learning_rate": 3.0963789571230807e-05, + "loss": 3.0813, + "step": 15593 + }, + { + "epoch": 0.18797836120440026, + "grad_norm": 0.14678546786308289, + "learning_rate": 3.096117532092661e-05, + "loss": 3.0899, + "step": 15594 + }, + { + "epoch": 0.18800920766997367, + "grad_norm": 0.15207891166210175, + "learning_rate": 3.095856100150617e-05, + "loss": 3.0943, + "step": 15595 + }, + { + "epoch": 0.18804005413554709, + "grad_norm": 0.14844076335430145, + "learning_rate": 3.09559466129998e-05, + "loss": 3.0982, + "step": 15596 + }, + { + "epoch": 0.1880709006011205, + "grad_norm": 0.15209569036960602, + "learning_rate": 3.095333215543781e-05, + "loss": 3.1193, + "step": 15597 + }, + { + "epoch": 0.18810174706669391, + "grad_norm": 0.14909516274929047, + "learning_rate": 3.095071762885051e-05, + "loss": 3.0869, + "step": 15598 + }, + { + "epoch": 0.18813259353226733, + "grad_norm": 0.1587282419204712, + "learning_rate": 3.094810303326822e-05, + "loss": 3.1275, + "step": 15599 + }, + { + "epoch": 0.18816343999784074, + "grad_norm": 0.15357251465320587, + "learning_rate": 3.094548836872124e-05, + "loss": 3.1203, + "step": 15600 + }, + { + "epoch": 0.18819428646341416, + "grad_norm": 0.15823781490325928, + "learning_rate": 3.09428736352399e-05, + "loss": 3.0967, + "step": 15601 + }, + { + "epoch": 0.18822513292898757, + "grad_norm": 0.1526394933462143, + "learning_rate": 3.094025883285451e-05, + "loss": 3.1193, + "step": 15602 + }, + { + "epoch": 0.18825597939456098, + "grad_norm": 0.16784606873989105, + "learning_rate": 3.093764396159538e-05, + "loss": 3.1038, + "step": 15603 + }, + { + "epoch": 0.18828682586013443, + "grad_norm": 0.1483873426914215, + "learning_rate": 3.093502902149285e-05, + "loss": 3.1204, + "step": 15604 + }, + { + "epoch": 0.18831767232570784, + "grad_norm": 0.1528395563364029, + "learning_rate": 3.093241401257722e-05, + "loss": 3.1151, + "step": 15605 + }, + { + "epoch": 0.18834851879128126, + "grad_norm": 0.196846604347229, + "learning_rate": 3.092979893487882e-05, + "loss": 3.1089, + "step": 15606 + }, + { + "epoch": 0.18837936525685467, + "grad_norm": 0.1492672711610794, + "learning_rate": 3.092718378842795e-05, + "loss": 3.0848, + "step": 15607 + }, + { + "epoch": 0.18841021172242808, + "grad_norm": 0.1517173796892166, + "learning_rate": 3.0924568573254956e-05, + "loss": 3.0895, + "step": 15608 + }, + { + "epoch": 0.1884410581880015, + "grad_norm": 0.1508607268333435, + "learning_rate": 3.0921953289390146e-05, + "loss": 3.1167, + "step": 15609 + }, + { + "epoch": 0.1884719046535749, + "grad_norm": 0.15261611342430115, + "learning_rate": 3.0919337936863856e-05, + "loss": 3.1227, + "step": 15610 + }, + { + "epoch": 0.18850275111914833, + "grad_norm": 0.15116649866104126, + "learning_rate": 3.091672251570639e-05, + "loss": 3.0942, + "step": 15611 + }, + { + "epoch": 0.18853359758472174, + "grad_norm": 0.14947883784770966, + "learning_rate": 3.0914107025948084e-05, + "loss": 3.1261, + "step": 15612 + }, + { + "epoch": 0.18856444405029515, + "grad_norm": 0.14946305751800537, + "learning_rate": 3.091149146761926e-05, + "loss": 3.0966, + "step": 15613 + }, + { + "epoch": 0.18859529051586857, + "grad_norm": 0.14608454704284668, + "learning_rate": 3.090887584075024e-05, + "loss": 3.1233, + "step": 15614 + }, + { + "epoch": 0.18862613698144198, + "grad_norm": 0.1572827398777008, + "learning_rate": 3.0906260145371375e-05, + "loss": 3.1114, + "step": 15615 + }, + { + "epoch": 0.18865698344701542, + "grad_norm": 0.1565873622894287, + "learning_rate": 3.090364438151296e-05, + "loss": 3.1347, + "step": 15616 + }, + { + "epoch": 0.18868782991258884, + "grad_norm": 0.15484926104545593, + "learning_rate": 3.090102854920533e-05, + "loss": 3.1124, + "step": 15617 + }, + { + "epoch": 0.18871867637816225, + "grad_norm": 0.16226692497730255, + "learning_rate": 3.089841264847883e-05, + "loss": 3.0918, + "step": 15618 + }, + { + "epoch": 0.18874952284373567, + "grad_norm": 0.1553654670715332, + "learning_rate": 3.089579667936378e-05, + "loss": 3.092, + "step": 15619 + }, + { + "epoch": 0.18878036930930908, + "grad_norm": 0.15947329998016357, + "learning_rate": 3.08931806418905e-05, + "loss": 3.0942, + "step": 15620 + }, + { + "epoch": 0.1888112157748825, + "grad_norm": 0.16170048713684082, + "learning_rate": 3.089056453608934e-05, + "loss": 3.1421, + "step": 15621 + }, + { + "epoch": 0.1888420622404559, + "grad_norm": 0.15229229629039764, + "learning_rate": 3.088794836199063e-05, + "loss": 3.1088, + "step": 15622 + }, + { + "epoch": 0.18887290870602932, + "grad_norm": 0.1649838536977768, + "learning_rate": 3.088533211962468e-05, + "loss": 3.1208, + "step": 15623 + }, + { + "epoch": 0.18890375517160274, + "grad_norm": 0.1531371772289276, + "learning_rate": 3.088271580902185e-05, + "loss": 3.129, + "step": 15624 + }, + { + "epoch": 0.18893460163717615, + "grad_norm": 0.15087608993053436, + "learning_rate": 3.0880099430212466e-05, + "loss": 3.0818, + "step": 15625 + }, + { + "epoch": 0.18896544810274957, + "grad_norm": 0.15641185641288757, + "learning_rate": 3.087748298322686e-05, + "loss": 3.1017, + "step": 15626 + }, + { + "epoch": 0.18899629456832298, + "grad_norm": 0.14941337704658508, + "learning_rate": 3.0874866468095376e-05, + "loss": 3.0925, + "step": 15627 + }, + { + "epoch": 0.18902714103389642, + "grad_norm": 0.1555749624967575, + "learning_rate": 3.0872249884848347e-05, + "loss": 3.0972, + "step": 15628 + }, + { + "epoch": 0.18905798749946984, + "grad_norm": 0.15987016260623932, + "learning_rate": 3.08696332335161e-05, + "loss": 3.1088, + "step": 15629 + }, + { + "epoch": 0.18908883396504325, + "grad_norm": 0.15103527903556824, + "learning_rate": 3.086701651412899e-05, + "loss": 3.1285, + "step": 15630 + }, + { + "epoch": 0.18911968043061667, + "grad_norm": 0.14737997949123383, + "learning_rate": 3.0864399726717344e-05, + "loss": 3.1072, + "step": 15631 + }, + { + "epoch": 0.18915052689619008, + "grad_norm": 0.14847347140312195, + "learning_rate": 3.086178287131151e-05, + "loss": 3.1036, + "step": 15632 + }, + { + "epoch": 0.1891813733617635, + "grad_norm": 0.15840037167072296, + "learning_rate": 3.085916594794183e-05, + "loss": 3.097, + "step": 15633 + }, + { + "epoch": 0.1892122198273369, + "grad_norm": 0.15524275600910187, + "learning_rate": 3.085654895663864e-05, + "loss": 3.1154, + "step": 15634 + }, + { + "epoch": 0.18924306629291032, + "grad_norm": 0.15441939234733582, + "learning_rate": 3.085393189743227e-05, + "loss": 3.1054, + "step": 15635 + }, + { + "epoch": 0.18927391275848374, + "grad_norm": 0.15613719820976257, + "learning_rate": 3.085131477035309e-05, + "loss": 3.1295, + "step": 15636 + }, + { + "epoch": 0.18930475922405715, + "grad_norm": 0.15003125369548798, + "learning_rate": 3.084869757543144e-05, + "loss": 3.1257, + "step": 15637 + }, + { + "epoch": 0.18933560568963057, + "grad_norm": 0.1535581648349762, + "learning_rate": 3.084608031269764e-05, + "loss": 3.0923, + "step": 15638 + }, + { + "epoch": 0.18936645215520398, + "grad_norm": 0.15980607271194458, + "learning_rate": 3.084346298218207e-05, + "loss": 3.1128, + "step": 15639 + }, + { + "epoch": 0.18939729862077742, + "grad_norm": 0.15733107924461365, + "learning_rate": 3.084084558391504e-05, + "loss": 3.1083, + "step": 15640 + }, + { + "epoch": 0.18942814508635084, + "grad_norm": 0.1551043689250946, + "learning_rate": 3.083822811792692e-05, + "loss": 3.1007, + "step": 15641 + }, + { + "epoch": 0.18945899155192425, + "grad_norm": 0.15553179383277893, + "learning_rate": 3.0835610584248064e-05, + "loss": 3.107, + "step": 15642 + }, + { + "epoch": 0.18948983801749766, + "grad_norm": 0.15579761564731598, + "learning_rate": 3.08329929829088e-05, + "loss": 3.1018, + "step": 15643 + }, + { + "epoch": 0.18952068448307108, + "grad_norm": 0.16089892387390137, + "learning_rate": 3.0830375313939494e-05, + "loss": 3.1333, + "step": 15644 + }, + { + "epoch": 0.1895515309486445, + "grad_norm": 0.1495816856622696, + "learning_rate": 3.082775757737049e-05, + "loss": 3.1039, + "step": 15645 + }, + { + "epoch": 0.1895823774142179, + "grad_norm": 0.1813967376947403, + "learning_rate": 3.082513977323212e-05, + "loss": 3.1178, + "step": 15646 + }, + { + "epoch": 0.18961322387979132, + "grad_norm": 0.15674926340579987, + "learning_rate": 3.082252190155477e-05, + "loss": 3.1078, + "step": 15647 + }, + { + "epoch": 0.18964407034536473, + "grad_norm": 0.15338057279586792, + "learning_rate": 3.081990396236878e-05, + "loss": 3.118, + "step": 15648 + }, + { + "epoch": 0.18967491681093815, + "grad_norm": 0.15412423014640808, + "learning_rate": 3.08172859557045e-05, + "loss": 3.0809, + "step": 15649 + }, + { + "epoch": 0.18970576327651156, + "grad_norm": 0.15340125560760498, + "learning_rate": 3.081466788159228e-05, + "loss": 3.1213, + "step": 15650 + }, + { + "epoch": 0.18973660974208498, + "grad_norm": 0.16117163002490997, + "learning_rate": 3.0812049740062494e-05, + "loss": 3.1258, + "step": 15651 + }, + { + "epoch": 0.18976745620765842, + "grad_norm": 0.1534065157175064, + "learning_rate": 3.080943153114547e-05, + "loss": 3.1026, + "step": 15652 + }, + { + "epoch": 0.18979830267323183, + "grad_norm": 0.15346919000148773, + "learning_rate": 3.0806813254871584e-05, + "loss": 3.1061, + "step": 15653 + }, + { + "epoch": 0.18982914913880525, + "grad_norm": 0.16860255599021912, + "learning_rate": 3.080419491127118e-05, + "loss": 3.1503, + "step": 15654 + }, + { + "epoch": 0.18985999560437866, + "grad_norm": 0.15479888021945953, + "learning_rate": 3.080157650037463e-05, + "loss": 3.1279, + "step": 15655 + }, + { + "epoch": 0.18989084206995208, + "grad_norm": 0.16674934327602386, + "learning_rate": 3.0798958022212287e-05, + "loss": 3.1409, + "step": 15656 + }, + { + "epoch": 0.1899216885355255, + "grad_norm": 0.15201090276241302, + "learning_rate": 3.0796339476814506e-05, + "loss": 3.1645, + "step": 15657 + }, + { + "epoch": 0.1899525350010989, + "grad_norm": 0.15959420800209045, + "learning_rate": 3.0793720864211654e-05, + "loss": 3.1277, + "step": 15658 + }, + { + "epoch": 0.18998338146667232, + "grad_norm": 0.1554781198501587, + "learning_rate": 3.0791102184434095e-05, + "loss": 3.1035, + "step": 15659 + }, + { + "epoch": 0.19001422793224573, + "grad_norm": 0.16010348498821259, + "learning_rate": 3.078848343751218e-05, + "loss": 3.1052, + "step": 15660 + }, + { + "epoch": 0.19004507439781915, + "grad_norm": 0.15173651278018951, + "learning_rate": 3.078586462347629e-05, + "loss": 3.1339, + "step": 15661 + }, + { + "epoch": 0.19007592086339256, + "grad_norm": 0.15808314085006714, + "learning_rate": 3.078324574235676e-05, + "loss": 3.1143, + "step": 15662 + }, + { + "epoch": 0.19010676732896598, + "grad_norm": 0.15753412246704102, + "learning_rate": 3.078062679418398e-05, + "loss": 3.1487, + "step": 15663 + }, + { + "epoch": 0.19013761379453942, + "grad_norm": 0.15282543003559113, + "learning_rate": 3.0778007778988305e-05, + "loss": 3.0976, + "step": 15664 + }, + { + "epoch": 0.19016846026011283, + "grad_norm": 0.15573887526988983, + "learning_rate": 3.07753886968001e-05, + "loss": 3.0909, + "step": 15665 + }, + { + "epoch": 0.19019930672568625, + "grad_norm": 0.1602376401424408, + "learning_rate": 3.077276954764974e-05, + "loss": 3.1139, + "step": 15666 + }, + { + "epoch": 0.19023015319125966, + "grad_norm": 0.1517569124698639, + "learning_rate": 3.077015033156758e-05, + "loss": 3.1198, + "step": 15667 + }, + { + "epoch": 0.19026099965683307, + "grad_norm": 0.1602158099412918, + "learning_rate": 3.076753104858399e-05, + "loss": 3.1517, + "step": 15668 + }, + { + "epoch": 0.1902918461224065, + "grad_norm": 0.15365487337112427, + "learning_rate": 3.0764911698729355e-05, + "loss": 3.102, + "step": 15669 + }, + { + "epoch": 0.1903226925879799, + "grad_norm": 0.17117299139499664, + "learning_rate": 3.076229228203402e-05, + "loss": 3.0892, + "step": 15670 + }, + { + "epoch": 0.19035353905355332, + "grad_norm": 0.15564332902431488, + "learning_rate": 3.0759672798528384e-05, + "loss": 3.1211, + "step": 15671 + }, + { + "epoch": 0.19038438551912673, + "grad_norm": 0.14902791380882263, + "learning_rate": 3.0757053248242794e-05, + "loss": 3.1123, + "step": 15672 + }, + { + "epoch": 0.19041523198470015, + "grad_norm": 0.15533077716827393, + "learning_rate": 3.075443363120763e-05, + "loss": 3.0728, + "step": 15673 + }, + { + "epoch": 0.19044607845027356, + "grad_norm": 0.15814334154129028, + "learning_rate": 3.0751813947453265e-05, + "loss": 3.1045, + "step": 15674 + }, + { + "epoch": 0.19047692491584697, + "grad_norm": 0.1518603414297104, + "learning_rate": 3.074919419701008e-05, + "loss": 3.1196, + "step": 15675 + }, + { + "epoch": 0.19050777138142042, + "grad_norm": 0.15805453062057495, + "learning_rate": 3.074657437990844e-05, + "loss": 3.1126, + "step": 15676 + }, + { + "epoch": 0.19053861784699383, + "grad_norm": 0.1550818532705307, + "learning_rate": 3.074395449617873e-05, + "loss": 3.1036, + "step": 15677 + }, + { + "epoch": 0.19056946431256724, + "grad_norm": 0.15935471653938293, + "learning_rate": 3.074133454585131e-05, + "loss": 3.1438, + "step": 15678 + }, + { + "epoch": 0.19060031077814066, + "grad_norm": 0.1547548472881317, + "learning_rate": 3.073871452895657e-05, + "loss": 3.0948, + "step": 15679 + }, + { + "epoch": 0.19063115724371407, + "grad_norm": 0.15110087394714355, + "learning_rate": 3.0736094445524886e-05, + "loss": 3.1239, + "step": 15680 + }, + { + "epoch": 0.1906620037092875, + "grad_norm": 0.1467285007238388, + "learning_rate": 3.073347429558663e-05, + "loss": 3.119, + "step": 15681 + }, + { + "epoch": 0.1906928501748609, + "grad_norm": 0.15263517200946808, + "learning_rate": 3.073085407917218e-05, + "loss": 3.0717, + "step": 15682 + }, + { + "epoch": 0.19072369664043431, + "grad_norm": 0.1544337272644043, + "learning_rate": 3.072823379631193e-05, + "loss": 3.1042, + "step": 15683 + }, + { + "epoch": 0.19075454310600773, + "grad_norm": 0.16277414560317993, + "learning_rate": 3.072561344703625e-05, + "loss": 3.1295, + "step": 15684 + }, + { + "epoch": 0.19078538957158114, + "grad_norm": 0.1598019003868103, + "learning_rate": 3.0722993031375515e-05, + "loss": 3.1048, + "step": 15685 + }, + { + "epoch": 0.19081623603715456, + "grad_norm": 0.14961522817611694, + "learning_rate": 3.072037254936012e-05, + "loss": 3.1105, + "step": 15686 + }, + { + "epoch": 0.19084708250272797, + "grad_norm": 0.1514594554901123, + "learning_rate": 3.071775200102045e-05, + "loss": 3.0913, + "step": 15687 + }, + { + "epoch": 0.1908779289683014, + "grad_norm": 0.1534411758184433, + "learning_rate": 3.071513138638687e-05, + "loss": 3.1249, + "step": 15688 + }, + { + "epoch": 0.19090877543387483, + "grad_norm": 0.1535721719264984, + "learning_rate": 3.0712510705489786e-05, + "loss": 3.1195, + "step": 15689 + }, + { + "epoch": 0.19093962189944824, + "grad_norm": 0.15423858165740967, + "learning_rate": 3.0709889958359564e-05, + "loss": 3.1251, + "step": 15690 + }, + { + "epoch": 0.19097046836502166, + "grad_norm": 0.15132646262645721, + "learning_rate": 3.0707269145026594e-05, + "loss": 3.1107, + "step": 15691 + }, + { + "epoch": 0.19100131483059507, + "grad_norm": 0.15075956284999847, + "learning_rate": 3.0704648265521284e-05, + "loss": 3.1138, + "step": 15692 + }, + { + "epoch": 0.19103216129616848, + "grad_norm": 0.14770840108394623, + "learning_rate": 3.0702027319874e-05, + "loss": 3.0949, + "step": 15693 + }, + { + "epoch": 0.1910630077617419, + "grad_norm": 0.15073084831237793, + "learning_rate": 3.069940630811513e-05, + "loss": 3.1139, + "step": 15694 + }, + { + "epoch": 0.1910938542273153, + "grad_norm": 0.15764333307743073, + "learning_rate": 3.0696785230275075e-05, + "loss": 3.0804, + "step": 15695 + }, + { + "epoch": 0.19112470069288873, + "grad_norm": 0.1522337943315506, + "learning_rate": 3.069416408638422e-05, + "loss": 3.1138, + "step": 15696 + }, + { + "epoch": 0.19115554715846214, + "grad_norm": 0.15270522236824036, + "learning_rate": 3.0691542876472943e-05, + "loss": 3.1082, + "step": 15697 + }, + { + "epoch": 0.19118639362403556, + "grad_norm": 0.15200045704841614, + "learning_rate": 3.068892160057166e-05, + "loss": 3.0866, + "step": 15698 + }, + { + "epoch": 0.19121724008960897, + "grad_norm": 0.14912474155426025, + "learning_rate": 3.068630025871075e-05, + "loss": 3.0995, + "step": 15699 + }, + { + "epoch": 0.1912480865551824, + "grad_norm": 0.15077757835388184, + "learning_rate": 3.06836788509206e-05, + "loss": 3.0974, + "step": 15700 + }, + { + "epoch": 0.19127893302075583, + "grad_norm": 0.15921947360038757, + "learning_rate": 3.06810573772316e-05, + "loss": 3.0881, + "step": 15701 + }, + { + "epoch": 0.19130977948632924, + "grad_norm": 0.16371969878673553, + "learning_rate": 3.067843583767417e-05, + "loss": 3.0987, + "step": 15702 + }, + { + "epoch": 0.19134062595190265, + "grad_norm": 0.16212265193462372, + "learning_rate": 3.0675814232278674e-05, + "loss": 3.1273, + "step": 15703 + }, + { + "epoch": 0.19137147241747607, + "grad_norm": 0.15206243097782135, + "learning_rate": 3.0673192561075533e-05, + "loss": 3.0848, + "step": 15704 + }, + { + "epoch": 0.19140231888304948, + "grad_norm": 0.15153652429580688, + "learning_rate": 3.0670570824095134e-05, + "loss": 3.1198, + "step": 15705 + }, + { + "epoch": 0.1914331653486229, + "grad_norm": 0.15374499559402466, + "learning_rate": 3.066794902136787e-05, + "loss": 3.1228, + "step": 15706 + }, + { + "epoch": 0.1914640118141963, + "grad_norm": 0.16293984651565552, + "learning_rate": 3.066532715292414e-05, + "loss": 3.1116, + "step": 15707 + }, + { + "epoch": 0.19149485827976973, + "grad_norm": 0.4427826404571533, + "learning_rate": 3.0662705218794344e-05, + "loss": 3.1247, + "step": 15708 + }, + { + "epoch": 0.19152570474534314, + "grad_norm": 0.15285935997962952, + "learning_rate": 3.0660083219008895e-05, + "loss": 3.1427, + "step": 15709 + }, + { + "epoch": 0.19155655121091655, + "grad_norm": 0.15362991392612457, + "learning_rate": 3.065746115359818e-05, + "loss": 3.1152, + "step": 15710 + }, + { + "epoch": 0.19158739767648997, + "grad_norm": 0.156197652220726, + "learning_rate": 3.0654839022592595e-05, + "loss": 3.0815, + "step": 15711 + }, + { + "epoch": 0.1916182441420634, + "grad_norm": 0.15233944356441498, + "learning_rate": 3.0652216826022554e-05, + "loss": 3.1347, + "step": 15712 + }, + { + "epoch": 0.19164909060763682, + "grad_norm": 0.15762966871261597, + "learning_rate": 3.064959456391846e-05, + "loss": 3.1141, + "step": 15713 + }, + { + "epoch": 0.19167993707321024, + "grad_norm": 0.15446385741233826, + "learning_rate": 3.06469722363107e-05, + "loss": 3.1256, + "step": 15714 + }, + { + "epoch": 0.19171078353878365, + "grad_norm": 0.15897493064403534, + "learning_rate": 3.06443498432297e-05, + "loss": 3.1212, + "step": 15715 + }, + { + "epoch": 0.19174163000435707, + "grad_norm": 0.15359701216220856, + "learning_rate": 3.064172738470586e-05, + "loss": 3.1078, + "step": 15716 + }, + { + "epoch": 0.19177247646993048, + "grad_norm": 0.14901459217071533, + "learning_rate": 3.063910486076957e-05, + "loss": 3.1253, + "step": 15717 + }, + { + "epoch": 0.1918033229355039, + "grad_norm": 0.16077210009098053, + "learning_rate": 3.063648227145125e-05, + "loss": 3.1045, + "step": 15718 + }, + { + "epoch": 0.1918341694010773, + "grad_norm": 0.15157809853553772, + "learning_rate": 3.063385961678131e-05, + "loss": 3.0948, + "step": 15719 + }, + { + "epoch": 0.19186501586665072, + "grad_norm": 0.16151708364486694, + "learning_rate": 3.063123689679015e-05, + "loss": 3.1315, + "step": 15720 + }, + { + "epoch": 0.19189586233222414, + "grad_norm": 0.15925413370132446, + "learning_rate": 3.0628614111508186e-05, + "loss": 3.088, + "step": 15721 + }, + { + "epoch": 0.19192670879779755, + "grad_norm": 0.15019753575325012, + "learning_rate": 3.0625991260965826e-05, + "loss": 3.1297, + "step": 15722 + }, + { + "epoch": 0.19195755526337097, + "grad_norm": 0.15657691657543182, + "learning_rate": 3.062336834519347e-05, + "loss": 3.0956, + "step": 15723 + }, + { + "epoch": 0.1919884017289444, + "grad_norm": 0.15249475836753845, + "learning_rate": 3.062074536422154e-05, + "loss": 3.105, + "step": 15724 + }, + { + "epoch": 0.19201924819451782, + "grad_norm": 0.15132665634155273, + "learning_rate": 3.061812231808045e-05, + "loss": 3.0692, + "step": 15725 + }, + { + "epoch": 0.19205009466009124, + "grad_norm": 0.16112734377384186, + "learning_rate": 3.0615499206800596e-05, + "loss": 3.0809, + "step": 15726 + }, + { + "epoch": 0.19208094112566465, + "grad_norm": 0.15259157121181488, + "learning_rate": 3.061287603041242e-05, + "loss": 3.1178, + "step": 15727 + }, + { + "epoch": 0.19211178759123806, + "grad_norm": 0.1542876809835434, + "learning_rate": 3.061025278894631e-05, + "loss": 3.1327, + "step": 15728 + }, + { + "epoch": 0.19214263405681148, + "grad_norm": 0.15297076106071472, + "learning_rate": 3.0607629482432685e-05, + "loss": 3.13, + "step": 15729 + }, + { + "epoch": 0.1921734805223849, + "grad_norm": 0.1556204855442047, + "learning_rate": 3.060500611090197e-05, + "loss": 3.0858, + "step": 15730 + }, + { + "epoch": 0.1922043269879583, + "grad_norm": 0.15992504358291626, + "learning_rate": 3.060238267438458e-05, + "loss": 3.1304, + "step": 15731 + }, + { + "epoch": 0.19223517345353172, + "grad_norm": 0.1551571488380432, + "learning_rate": 3.059975917291093e-05, + "loss": 3.1022, + "step": 15732 + }, + { + "epoch": 0.19226601991910514, + "grad_norm": 0.14987941086292267, + "learning_rate": 3.059713560651144e-05, + "loss": 3.1297, + "step": 15733 + }, + { + "epoch": 0.19229686638467855, + "grad_norm": 0.14912700653076172, + "learning_rate": 3.059451197521652e-05, + "loss": 3.1106, + "step": 15734 + }, + { + "epoch": 0.19232771285025196, + "grad_norm": 0.16047292947769165, + "learning_rate": 3.059188827905659e-05, + "loss": 3.082, + "step": 15735 + }, + { + "epoch": 0.1923585593158254, + "grad_norm": 0.15429024398326874, + "learning_rate": 3.058926451806209e-05, + "loss": 3.1072, + "step": 15736 + }, + { + "epoch": 0.19238940578139882, + "grad_norm": 0.1559903621673584, + "learning_rate": 3.058664069226342e-05, + "loss": 3.1188, + "step": 15737 + }, + { + "epoch": 0.19242025224697223, + "grad_norm": 0.16014187037944794, + "learning_rate": 3.0584016801691004e-05, + "loss": 3.1156, + "step": 15738 + }, + { + "epoch": 0.19245109871254565, + "grad_norm": 0.15866833925247192, + "learning_rate": 3.058139284637528e-05, + "loss": 3.1041, + "step": 15739 + }, + { + "epoch": 0.19248194517811906, + "grad_norm": 0.15592612326145172, + "learning_rate": 3.0578768826346646e-05, + "loss": 3.0967, + "step": 15740 + }, + { + "epoch": 0.19251279164369248, + "grad_norm": 0.15986354649066925, + "learning_rate": 3.057614474163555e-05, + "loss": 3.115, + "step": 15741 + }, + { + "epoch": 0.1925436381092659, + "grad_norm": 0.15178178250789642, + "learning_rate": 3.05735205922724e-05, + "loss": 3.0986, + "step": 15742 + }, + { + "epoch": 0.1925744845748393, + "grad_norm": 0.15334832668304443, + "learning_rate": 3.0570896378287636e-05, + "loss": 3.0742, + "step": 15743 + }, + { + "epoch": 0.19260533104041272, + "grad_norm": 0.16013357043266296, + "learning_rate": 3.056827209971167e-05, + "loss": 3.1278, + "step": 15744 + }, + { + "epoch": 0.19263617750598613, + "grad_norm": 0.15004663169384003, + "learning_rate": 3.056564775657494e-05, + "loss": 3.1202, + "step": 15745 + }, + { + "epoch": 0.19266702397155955, + "grad_norm": 0.16065002977848053, + "learning_rate": 3.056302334890786e-05, + "loss": 3.1224, + "step": 15746 + }, + { + "epoch": 0.19269787043713296, + "grad_norm": 0.15445555746555328, + "learning_rate": 3.0560398876740874e-05, + "loss": 3.106, + "step": 15747 + }, + { + "epoch": 0.1927287169027064, + "grad_norm": 0.15433204174041748, + "learning_rate": 3.0557774340104406e-05, + "loss": 3.0932, + "step": 15748 + }, + { + "epoch": 0.19275956336827982, + "grad_norm": 0.1567688137292862, + "learning_rate": 3.055514973902888e-05, + "loss": 3.121, + "step": 15749 + }, + { + "epoch": 0.19279040983385323, + "grad_norm": 0.1578507125377655, + "learning_rate": 3.055252507354474e-05, + "loss": 3.1403, + "step": 15750 + }, + { + "epoch": 0.19282125629942665, + "grad_norm": 0.156483456492424, + "learning_rate": 3.0549900343682396e-05, + "loss": 3.0996, + "step": 15751 + }, + { + "epoch": 0.19285210276500006, + "grad_norm": 0.15846548974514008, + "learning_rate": 3.0547275549472295e-05, + "loss": 3.1464, + "step": 15752 + }, + { + "epoch": 0.19288294923057347, + "grad_norm": 0.15512992441654205, + "learning_rate": 3.054465069094487e-05, + "loss": 3.0808, + "step": 15753 + }, + { + "epoch": 0.1929137956961469, + "grad_norm": 0.16046734154224396, + "learning_rate": 3.0542025768130556e-05, + "loss": 3.122, + "step": 15754 + }, + { + "epoch": 0.1929446421617203, + "grad_norm": 0.15678438544273376, + "learning_rate": 3.053940078105979e-05, + "loss": 3.1215, + "step": 15755 + }, + { + "epoch": 0.19297548862729372, + "grad_norm": 0.17447783052921295, + "learning_rate": 3.053677572976298e-05, + "loss": 3.1191, + "step": 15756 + }, + { + "epoch": 0.19300633509286713, + "grad_norm": 0.2006862461566925, + "learning_rate": 3.053415061427059e-05, + "loss": 3.1101, + "step": 15757 + }, + { + "epoch": 0.19303718155844055, + "grad_norm": 0.159057155251503, + "learning_rate": 3.053152543461305e-05, + "loss": 3.0996, + "step": 15758 + }, + { + "epoch": 0.19306802802401396, + "grad_norm": 0.15693555772304535, + "learning_rate": 3.05289001908208e-05, + "loss": 3.1166, + "step": 15759 + }, + { + "epoch": 0.1930988744895874, + "grad_norm": 0.1609862893819809, + "learning_rate": 3.052627488292428e-05, + "loss": 3.1115, + "step": 15760 + }, + { + "epoch": 0.19312972095516082, + "grad_norm": 0.16022533178329468, + "learning_rate": 3.052364951095392e-05, + "loss": 3.1325, + "step": 15761 + }, + { + "epoch": 0.19316056742073423, + "grad_norm": 0.15905451774597168, + "learning_rate": 3.052102407494016e-05, + "loss": 3.1319, + "step": 15762 + }, + { + "epoch": 0.19319141388630764, + "grad_norm": 0.16053710877895355, + "learning_rate": 3.0518398574913445e-05, + "loss": 3.0824, + "step": 15763 + }, + { + "epoch": 0.19322226035188106, + "grad_norm": 0.153938889503479, + "learning_rate": 3.0515773010904207e-05, + "loss": 3.1302, + "step": 15764 + }, + { + "epoch": 0.19325310681745447, + "grad_norm": 0.15282942354679108, + "learning_rate": 3.0513147382942907e-05, + "loss": 3.123, + "step": 15765 + }, + { + "epoch": 0.1932839532830279, + "grad_norm": 0.16917909681797028, + "learning_rate": 3.0510521691059978e-05, + "loss": 3.0992, + "step": 15766 + }, + { + "epoch": 0.1933147997486013, + "grad_norm": 0.15465696156024933, + "learning_rate": 3.050789593528585e-05, + "loss": 3.0624, + "step": 15767 + }, + { + "epoch": 0.19334564621417472, + "grad_norm": 0.1547827571630478, + "learning_rate": 3.0505270115650987e-05, + "loss": 3.0863, + "step": 15768 + }, + { + "epoch": 0.19337649267974813, + "grad_norm": 0.17086485028266907, + "learning_rate": 3.0502644232185824e-05, + "loss": 3.0741, + "step": 15769 + }, + { + "epoch": 0.19340733914532154, + "grad_norm": 0.16020427644252777, + "learning_rate": 3.05000182849208e-05, + "loss": 3.1107, + "step": 15770 + }, + { + "epoch": 0.19343818561089496, + "grad_norm": 0.15918143093585968, + "learning_rate": 3.0497392273886377e-05, + "loss": 3.0772, + "step": 15771 + }, + { + "epoch": 0.1934690320764684, + "grad_norm": 0.16486681997776031, + "learning_rate": 3.0494766199112996e-05, + "loss": 3.1067, + "step": 15772 + }, + { + "epoch": 0.19349987854204181, + "grad_norm": 0.1626054048538208, + "learning_rate": 3.0492140060631098e-05, + "loss": 3.1051, + "step": 15773 + }, + { + "epoch": 0.19353072500761523, + "grad_norm": 0.1590595245361328, + "learning_rate": 3.0489513858471137e-05, + "loss": 3.132, + "step": 15774 + }, + { + "epoch": 0.19356157147318864, + "grad_norm": 0.1625828742980957, + "learning_rate": 3.048688759266356e-05, + "loss": 3.0928, + "step": 15775 + }, + { + "epoch": 0.19359241793876206, + "grad_norm": 0.15907321870326996, + "learning_rate": 3.048426126323882e-05, + "loss": 3.1016, + "step": 15776 + }, + { + "epoch": 0.19362326440433547, + "grad_norm": 0.16569480299949646, + "learning_rate": 3.0481634870227372e-05, + "loss": 3.1055, + "step": 15777 + }, + { + "epoch": 0.19365411086990889, + "grad_norm": 0.16672495007514954, + "learning_rate": 3.047900841365966e-05, + "loss": 3.1159, + "step": 15778 + }, + { + "epoch": 0.1936849573354823, + "grad_norm": 0.16029399633407593, + "learning_rate": 3.0476381893566132e-05, + "loss": 3.1406, + "step": 15779 + }, + { + "epoch": 0.1937158038010557, + "grad_norm": 0.1595541536808014, + "learning_rate": 3.0473755309977246e-05, + "loss": 3.11, + "step": 15780 + }, + { + "epoch": 0.19374665026662913, + "grad_norm": 0.1619035005569458, + "learning_rate": 3.0471128662923465e-05, + "loss": 3.0966, + "step": 15781 + }, + { + "epoch": 0.19377749673220254, + "grad_norm": 0.1677611768245697, + "learning_rate": 3.0468501952435224e-05, + "loss": 3.0899, + "step": 15782 + }, + { + "epoch": 0.19380834319777596, + "grad_norm": 0.1535625010728836, + "learning_rate": 3.0465875178543003e-05, + "loss": 3.0828, + "step": 15783 + }, + { + "epoch": 0.1938391896633494, + "grad_norm": 0.1656786948442459, + "learning_rate": 3.0463248341277233e-05, + "loss": 3.1552, + "step": 15784 + }, + { + "epoch": 0.1938700361289228, + "grad_norm": 0.15564732253551483, + "learning_rate": 3.046062144066838e-05, + "loss": 3.1465, + "step": 15785 + }, + { + "epoch": 0.19390088259449623, + "grad_norm": 0.157375305891037, + "learning_rate": 3.045799447674691e-05, + "loss": 3.1384, + "step": 15786 + }, + { + "epoch": 0.19393172906006964, + "grad_norm": 0.1673966944217682, + "learning_rate": 3.045536744954327e-05, + "loss": 3.1427, + "step": 15787 + }, + { + "epoch": 0.19396257552564306, + "grad_norm": 0.1635211557149887, + "learning_rate": 3.045274035908793e-05, + "loss": 3.14, + "step": 15788 + }, + { + "epoch": 0.19399342199121647, + "grad_norm": 0.15361528098583221, + "learning_rate": 3.0450113205411335e-05, + "loss": 3.1074, + "step": 15789 + }, + { + "epoch": 0.19402426845678988, + "grad_norm": 0.1622699797153473, + "learning_rate": 3.0447485988543952e-05, + "loss": 3.0982, + "step": 15790 + }, + { + "epoch": 0.1940551149223633, + "grad_norm": 0.18278507888317108, + "learning_rate": 3.044485870851624e-05, + "loss": 3.073, + "step": 15791 + }, + { + "epoch": 0.1940859613879367, + "grad_norm": 0.15957500040531158, + "learning_rate": 3.044223136535867e-05, + "loss": 3.0811, + "step": 15792 + }, + { + "epoch": 0.19411680785351013, + "grad_norm": 0.16249285638332367, + "learning_rate": 3.04396039591017e-05, + "loss": 3.1232, + "step": 15793 + }, + { + "epoch": 0.19414765431908354, + "grad_norm": 0.16595859825611115, + "learning_rate": 3.0436976489775783e-05, + "loss": 3.1315, + "step": 15794 + }, + { + "epoch": 0.19417850078465695, + "grad_norm": 0.18723130226135254, + "learning_rate": 3.0434348957411395e-05, + "loss": 3.1063, + "step": 15795 + }, + { + "epoch": 0.1942093472502304, + "grad_norm": 0.16215097904205322, + "learning_rate": 3.0431721362038995e-05, + "loss": 3.1004, + "step": 15796 + }, + { + "epoch": 0.1942401937158038, + "grad_norm": 0.16292640566825867, + "learning_rate": 3.0429093703689047e-05, + "loss": 3.0974, + "step": 15797 + }, + { + "epoch": 0.19427104018137722, + "grad_norm": 0.1635582596063614, + "learning_rate": 3.0426465982392026e-05, + "loss": 3.1258, + "step": 15798 + }, + { + "epoch": 0.19430188664695064, + "grad_norm": 0.15318550169467926, + "learning_rate": 3.0423838198178393e-05, + "loss": 3.0976, + "step": 15799 + }, + { + "epoch": 0.19433273311252405, + "grad_norm": 0.15742012858390808, + "learning_rate": 3.042121035107861e-05, + "loss": 3.084, + "step": 15800 + }, + { + "epoch": 0.19436357957809747, + "grad_norm": 0.16022007167339325, + "learning_rate": 3.0418582441123156e-05, + "loss": 3.0902, + "step": 15801 + }, + { + "epoch": 0.19439442604367088, + "grad_norm": 0.1614435911178589, + "learning_rate": 3.041595446834249e-05, + "loss": 3.126, + "step": 15802 + }, + { + "epoch": 0.1944252725092443, + "grad_norm": 0.16468101739883423, + "learning_rate": 3.0413326432767087e-05, + "loss": 3.1342, + "step": 15803 + }, + { + "epoch": 0.1944561189748177, + "grad_norm": 0.15047267079353333, + "learning_rate": 3.0410698334427423e-05, + "loss": 3.1161, + "step": 15804 + }, + { + "epoch": 0.19448696544039112, + "grad_norm": 0.15708518028259277, + "learning_rate": 3.0408070173353965e-05, + "loss": 3.0999, + "step": 15805 + }, + { + "epoch": 0.19451781190596454, + "grad_norm": 0.16406914591789246, + "learning_rate": 3.0405441949577172e-05, + "loss": 3.1487, + "step": 15806 + }, + { + "epoch": 0.19454865837153795, + "grad_norm": 0.15888839960098267, + "learning_rate": 3.0402813663127534e-05, + "loss": 3.0843, + "step": 15807 + }, + { + "epoch": 0.1945795048371114, + "grad_norm": 0.1650659143924713, + "learning_rate": 3.0400185314035513e-05, + "loss": 3.1365, + "step": 15808 + }, + { + "epoch": 0.1946103513026848, + "grad_norm": 0.15930570662021637, + "learning_rate": 3.0397556902331593e-05, + "loss": 3.092, + "step": 15809 + }, + { + "epoch": 0.19464119776825822, + "grad_norm": 0.16652359068393707, + "learning_rate": 3.039492842804625e-05, + "loss": 3.0793, + "step": 15810 + }, + { + "epoch": 0.19467204423383164, + "grad_norm": 0.17300111055374146, + "learning_rate": 3.0392299891209948e-05, + "loss": 3.0805, + "step": 15811 + }, + { + "epoch": 0.19470289069940505, + "grad_norm": 0.15139934420585632, + "learning_rate": 3.0389671291853166e-05, + "loss": 3.1193, + "step": 15812 + }, + { + "epoch": 0.19473373716497847, + "grad_norm": 0.16850751638412476, + "learning_rate": 3.038704263000639e-05, + "loss": 3.1106, + "step": 15813 + }, + { + "epoch": 0.19476458363055188, + "grad_norm": 0.16041871905326843, + "learning_rate": 3.038441390570008e-05, + "loss": 3.1465, + "step": 15814 + }, + { + "epoch": 0.1947954300961253, + "grad_norm": 0.17196647822856903, + "learning_rate": 3.0381785118964744e-05, + "loss": 3.1029, + "step": 15815 + }, + { + "epoch": 0.1948262765616987, + "grad_norm": 0.1532813459634781, + "learning_rate": 3.0379156269830838e-05, + "loss": 3.0931, + "step": 15816 + }, + { + "epoch": 0.19485712302727212, + "grad_norm": 0.16388802230358124, + "learning_rate": 3.0376527358328842e-05, + "loss": 3.1313, + "step": 15817 + }, + { + "epoch": 0.19488796949284554, + "grad_norm": 0.16996803879737854, + "learning_rate": 3.0373898384489243e-05, + "loss": 3.1189, + "step": 15818 + }, + { + "epoch": 0.19491881595841895, + "grad_norm": 0.15838433802127838, + "learning_rate": 3.0371269348342523e-05, + "loss": 3.0925, + "step": 15819 + }, + { + "epoch": 0.1949496624239924, + "grad_norm": 0.1635330617427826, + "learning_rate": 3.0368640249919167e-05, + "loss": 3.1245, + "step": 15820 + }, + { + "epoch": 0.1949805088895658, + "grad_norm": 0.15189708769321442, + "learning_rate": 3.0366011089249653e-05, + "loss": 3.0999, + "step": 15821 + }, + { + "epoch": 0.19501135535513922, + "grad_norm": 0.15230487287044525, + "learning_rate": 3.036338186636446e-05, + "loss": 3.0996, + "step": 15822 + }, + { + "epoch": 0.19504220182071264, + "grad_norm": 0.15953291952610016, + "learning_rate": 3.0360752581294078e-05, + "loss": 3.1131, + "step": 15823 + }, + { + "epoch": 0.19507304828628605, + "grad_norm": 0.15287628769874573, + "learning_rate": 3.0358123234068993e-05, + "loss": 3.1509, + "step": 15824 + }, + { + "epoch": 0.19510389475185946, + "grad_norm": 0.15117742121219635, + "learning_rate": 3.035549382471969e-05, + "loss": 3.1178, + "step": 15825 + }, + { + "epoch": 0.19513474121743288, + "grad_norm": 0.15828461945056915, + "learning_rate": 3.035286435327665e-05, + "loss": 3.1128, + "step": 15826 + }, + { + "epoch": 0.1951655876830063, + "grad_norm": 0.15084269642829895, + "learning_rate": 3.0350234819770368e-05, + "loss": 3.1235, + "step": 15827 + }, + { + "epoch": 0.1951964341485797, + "grad_norm": 0.15816877782344818, + "learning_rate": 3.0347605224231333e-05, + "loss": 3.0955, + "step": 15828 + }, + { + "epoch": 0.19522728061415312, + "grad_norm": 0.15419024229049683, + "learning_rate": 3.034497556669002e-05, + "loss": 3.1072, + "step": 15829 + }, + { + "epoch": 0.19525812707972653, + "grad_norm": 0.1496540755033493, + "learning_rate": 3.0342345847176928e-05, + "loss": 3.0867, + "step": 15830 + }, + { + "epoch": 0.19528897354529998, + "grad_norm": 0.1531912088394165, + "learning_rate": 3.033971606572255e-05, + "loss": 3.1225, + "step": 15831 + }, + { + "epoch": 0.1953198200108734, + "grad_norm": 0.16536864638328552, + "learning_rate": 3.0337086222357373e-05, + "loss": 3.136, + "step": 15832 + }, + { + "epoch": 0.1953506664764468, + "grad_norm": 0.16519413888454437, + "learning_rate": 3.0334456317111892e-05, + "loss": 3.0739, + "step": 15833 + }, + { + "epoch": 0.19538151294202022, + "grad_norm": 0.15858563780784607, + "learning_rate": 3.033182635001659e-05, + "loss": 3.1077, + "step": 15834 + }, + { + "epoch": 0.19541235940759363, + "grad_norm": 0.1663825660943985, + "learning_rate": 3.032919632110196e-05, + "loss": 3.1334, + "step": 15835 + }, + { + "epoch": 0.19544320587316705, + "grad_norm": 0.16089783608913422, + "learning_rate": 3.0326566230398514e-05, + "loss": 3.0932, + "step": 15836 + }, + { + "epoch": 0.19547405233874046, + "grad_norm": 0.16176460683345795, + "learning_rate": 3.0323936077936726e-05, + "loss": 3.1143, + "step": 15837 + }, + { + "epoch": 0.19550489880431388, + "grad_norm": 0.16644442081451416, + "learning_rate": 3.03213058637471e-05, + "loss": 3.1208, + "step": 15838 + }, + { + "epoch": 0.1955357452698873, + "grad_norm": 0.16324296593666077, + "learning_rate": 3.0318675587860134e-05, + "loss": 3.1212, + "step": 15839 + }, + { + "epoch": 0.1955665917354607, + "grad_norm": 0.15596938133239746, + "learning_rate": 3.0316045250306312e-05, + "loss": 3.1037, + "step": 15840 + }, + { + "epoch": 0.19559743820103412, + "grad_norm": 0.15445081889629364, + "learning_rate": 3.0313414851116146e-05, + "loss": 3.0939, + "step": 15841 + }, + { + "epoch": 0.19562828466660753, + "grad_norm": 0.1559808850288391, + "learning_rate": 3.031078439032013e-05, + "loss": 3.1015, + "step": 15842 + }, + { + "epoch": 0.19565913113218097, + "grad_norm": 0.15318728983402252, + "learning_rate": 3.0308153867948756e-05, + "loss": 3.0988, + "step": 15843 + }, + { + "epoch": 0.1956899775977544, + "grad_norm": 0.18644513189792633, + "learning_rate": 3.030552328403253e-05, + "loss": 3.1097, + "step": 15844 + }, + { + "epoch": 0.1957208240633278, + "grad_norm": 0.1589105874300003, + "learning_rate": 3.030289263860195e-05, + "loss": 3.1152, + "step": 15845 + }, + { + "epoch": 0.19575167052890122, + "grad_norm": 0.14965634047985077, + "learning_rate": 3.0300261931687512e-05, + "loss": 3.0801, + "step": 15846 + }, + { + "epoch": 0.19578251699447463, + "grad_norm": 0.1527303159236908, + "learning_rate": 3.0297631163319723e-05, + "loss": 3.0642, + "step": 15847 + }, + { + "epoch": 0.19581336346004805, + "grad_norm": 0.1702928990125656, + "learning_rate": 3.029500033352909e-05, + "loss": 3.1054, + "step": 15848 + }, + { + "epoch": 0.19584420992562146, + "grad_norm": 0.16480372846126556, + "learning_rate": 3.0292369442346108e-05, + "loss": 3.0894, + "step": 15849 + }, + { + "epoch": 0.19587505639119487, + "grad_norm": 0.1513206511735916, + "learning_rate": 3.0289738489801277e-05, + "loss": 3.0877, + "step": 15850 + }, + { + "epoch": 0.1959059028567683, + "grad_norm": 0.15699215233325958, + "learning_rate": 3.028710747592511e-05, + "loss": 3.0999, + "step": 15851 + }, + { + "epoch": 0.1959367493223417, + "grad_norm": 0.15578311681747437, + "learning_rate": 3.028447640074811e-05, + "loss": 3.0927, + "step": 15852 + }, + { + "epoch": 0.19596759578791512, + "grad_norm": 0.158676877617836, + "learning_rate": 3.028184526430078e-05, + "loss": 3.0947, + "step": 15853 + }, + { + "epoch": 0.19599844225348853, + "grad_norm": 0.1663709580898285, + "learning_rate": 3.027921406661363e-05, + "loss": 3.1234, + "step": 15854 + }, + { + "epoch": 0.19602928871906197, + "grad_norm": 0.15912073850631714, + "learning_rate": 3.027658280771717e-05, + "loss": 3.1081, + "step": 15855 + }, + { + "epoch": 0.1960601351846354, + "grad_norm": 0.1613406389951706, + "learning_rate": 3.0273951487641893e-05, + "loss": 3.1172, + "step": 15856 + }, + { + "epoch": 0.1960909816502088, + "grad_norm": 0.1545582264661789, + "learning_rate": 3.027132010641832e-05, + "loss": 3.1104, + "step": 15857 + }, + { + "epoch": 0.19612182811578222, + "grad_norm": 0.16096258163452148, + "learning_rate": 3.0268688664076962e-05, + "loss": 3.1022, + "step": 15858 + }, + { + "epoch": 0.19615267458135563, + "grad_norm": 0.1586107760667801, + "learning_rate": 3.0266057160648327e-05, + "loss": 3.0889, + "step": 15859 + }, + { + "epoch": 0.19618352104692904, + "grad_norm": 0.16033008694648743, + "learning_rate": 3.0263425596162924e-05, + "loss": 3.1198, + "step": 15860 + }, + { + "epoch": 0.19621436751250246, + "grad_norm": 0.15716701745986938, + "learning_rate": 3.0260793970651257e-05, + "loss": 3.1182, + "step": 15861 + }, + { + "epoch": 0.19624521397807587, + "grad_norm": 0.14969190955162048, + "learning_rate": 3.0258162284143854e-05, + "loss": 3.1192, + "step": 15862 + }, + { + "epoch": 0.1962760604436493, + "grad_norm": 0.15372483432292938, + "learning_rate": 3.0255530536671213e-05, + "loss": 3.1059, + "step": 15863 + }, + { + "epoch": 0.1963069069092227, + "grad_norm": 0.15874247252941132, + "learning_rate": 3.0252898728263852e-05, + "loss": 3.1667, + "step": 15864 + }, + { + "epoch": 0.19633775337479611, + "grad_norm": 0.14678291976451874, + "learning_rate": 3.025026685895229e-05, + "loss": 3.0546, + "step": 15865 + }, + { + "epoch": 0.19636859984036953, + "grad_norm": 0.16372056305408478, + "learning_rate": 3.0247634928767044e-05, + "loss": 3.1183, + "step": 15866 + }, + { + "epoch": 0.19639944630594297, + "grad_norm": 0.17786721885204315, + "learning_rate": 3.024500293773862e-05, + "loss": 3.1433, + "step": 15867 + }, + { + "epoch": 0.19643029277151638, + "grad_norm": 0.16037365794181824, + "learning_rate": 3.024237088589753e-05, + "loss": 3.1004, + "step": 15868 + }, + { + "epoch": 0.1964611392370898, + "grad_norm": 0.1530480533838272, + "learning_rate": 3.0239738773274313e-05, + "loss": 3.0941, + "step": 15869 + }, + { + "epoch": 0.1964919857026632, + "grad_norm": 0.17044736444950104, + "learning_rate": 3.023710659989947e-05, + "loss": 3.0787, + "step": 15870 + }, + { + "epoch": 0.19652283216823663, + "grad_norm": 0.15540458261966705, + "learning_rate": 3.0234474365803523e-05, + "loss": 3.1314, + "step": 15871 + }, + { + "epoch": 0.19655367863381004, + "grad_norm": 0.16025811433792114, + "learning_rate": 3.0231842071016993e-05, + "loss": 3.0796, + "step": 15872 + }, + { + "epoch": 0.19658452509938346, + "grad_norm": 0.15968644618988037, + "learning_rate": 3.0229209715570388e-05, + "loss": 3.1319, + "step": 15873 + }, + { + "epoch": 0.19661537156495687, + "grad_norm": 0.15290191769599915, + "learning_rate": 3.0226577299494245e-05, + "loss": 3.1232, + "step": 15874 + }, + { + "epoch": 0.19664621803053028, + "grad_norm": 0.15161827206611633, + "learning_rate": 3.0223944822819085e-05, + "loss": 3.1232, + "step": 15875 + }, + { + "epoch": 0.1966770644961037, + "grad_norm": 0.15198130905628204, + "learning_rate": 3.0221312285575414e-05, + "loss": 3.1161, + "step": 15876 + }, + { + "epoch": 0.1967079109616771, + "grad_norm": 0.14866404235363007, + "learning_rate": 3.021867968779377e-05, + "loss": 3.083, + "step": 15877 + }, + { + "epoch": 0.19673875742725053, + "grad_norm": 0.16047579050064087, + "learning_rate": 3.021604702950468e-05, + "loss": 3.0826, + "step": 15878 + }, + { + "epoch": 0.19676960389282397, + "grad_norm": 0.15447485446929932, + "learning_rate": 3.0213414310738647e-05, + "loss": 3.1115, + "step": 15879 + }, + { + "epoch": 0.19680045035839738, + "grad_norm": 0.1601797640323639, + "learning_rate": 3.0210781531526205e-05, + "loss": 3.0893, + "step": 15880 + }, + { + "epoch": 0.1968312968239708, + "grad_norm": 0.1632867157459259, + "learning_rate": 3.020814869189789e-05, + "loss": 3.1569, + "step": 15881 + }, + { + "epoch": 0.1968621432895442, + "grad_norm": 0.1541447639465332, + "learning_rate": 3.020551579188422e-05, + "loss": 3.1032, + "step": 15882 + }, + { + "epoch": 0.19689298975511763, + "grad_norm": 0.1544661670923233, + "learning_rate": 3.0202882831515723e-05, + "loss": 3.1512, + "step": 15883 + }, + { + "epoch": 0.19692383622069104, + "grad_norm": 0.1504252851009369, + "learning_rate": 3.0200249810822922e-05, + "loss": 3.0728, + "step": 15884 + }, + { + "epoch": 0.19695468268626445, + "grad_norm": 0.15706771612167358, + "learning_rate": 3.0197616729836355e-05, + "loss": 3.1209, + "step": 15885 + }, + { + "epoch": 0.19698552915183787, + "grad_norm": 0.15836098790168762, + "learning_rate": 3.019498358858654e-05, + "loss": 3.1173, + "step": 15886 + }, + { + "epoch": 0.19701637561741128, + "grad_norm": 0.1597580760717392, + "learning_rate": 3.0192350387104017e-05, + "loss": 3.0679, + "step": 15887 + }, + { + "epoch": 0.1970472220829847, + "grad_norm": 0.15602202713489532, + "learning_rate": 3.0189717125419308e-05, + "loss": 3.0787, + "step": 15888 + }, + { + "epoch": 0.1970780685485581, + "grad_norm": 0.15204788744449615, + "learning_rate": 3.0187083803562953e-05, + "loss": 3.0556, + "step": 15889 + }, + { + "epoch": 0.19710891501413152, + "grad_norm": 0.1511426866054535, + "learning_rate": 3.0184450421565467e-05, + "loss": 3.082, + "step": 15890 + }, + { + "epoch": 0.19713976147970497, + "grad_norm": 0.15631213784217834, + "learning_rate": 3.01818169794574e-05, + "loss": 3.1011, + "step": 15891 + }, + { + "epoch": 0.19717060794527838, + "grad_norm": 0.15227767825126648, + "learning_rate": 3.0179183477269284e-05, + "loss": 3.1387, + "step": 15892 + }, + { + "epoch": 0.1972014544108518, + "grad_norm": 0.15124927461147308, + "learning_rate": 3.017654991503165e-05, + "loss": 3.1029, + "step": 15893 + }, + { + "epoch": 0.1972323008764252, + "grad_norm": 0.1539772003889084, + "learning_rate": 3.0173916292775017e-05, + "loss": 3.1047, + "step": 15894 + }, + { + "epoch": 0.19726314734199862, + "grad_norm": 0.16208875179290771, + "learning_rate": 3.0171282610529945e-05, + "loss": 3.0879, + "step": 15895 + }, + { + "epoch": 0.19729399380757204, + "grad_norm": 0.1822236031293869, + "learning_rate": 3.0168648868326955e-05, + "loss": 3.1056, + "step": 15896 + }, + { + "epoch": 0.19732484027314545, + "grad_norm": 0.15369857847690582, + "learning_rate": 3.0166015066196584e-05, + "loss": 3.1128, + "step": 15897 + }, + { + "epoch": 0.19735568673871887, + "grad_norm": 0.14948280155658722, + "learning_rate": 3.016338120416938e-05, + "loss": 3.0626, + "step": 15898 + }, + { + "epoch": 0.19738653320429228, + "grad_norm": 0.15433581173419952, + "learning_rate": 3.016074728227587e-05, + "loss": 3.1103, + "step": 15899 + }, + { + "epoch": 0.1974173796698657, + "grad_norm": 0.1560877412557602, + "learning_rate": 3.0158113300546593e-05, + "loss": 3.1477, + "step": 15900 + }, + { + "epoch": 0.1974482261354391, + "grad_norm": 0.15779832005500793, + "learning_rate": 3.0155479259012093e-05, + "loss": 3.1131, + "step": 15901 + }, + { + "epoch": 0.19747907260101252, + "grad_norm": 0.15157295763492584, + "learning_rate": 3.0152845157702915e-05, + "loss": 3.0952, + "step": 15902 + }, + { + "epoch": 0.19750991906658596, + "grad_norm": 0.16070625185966492, + "learning_rate": 3.0150210996649586e-05, + "loss": 3.1279, + "step": 15903 + }, + { + "epoch": 0.19754076553215938, + "grad_norm": 0.15539848804473877, + "learning_rate": 3.0147576775882664e-05, + "loss": 3.0724, + "step": 15904 + }, + { + "epoch": 0.1975716119977328, + "grad_norm": 0.16761668026447296, + "learning_rate": 3.0144942495432683e-05, + "loss": 3.0942, + "step": 15905 + }, + { + "epoch": 0.1976024584633062, + "grad_norm": 0.16171248257160187, + "learning_rate": 3.0142308155330172e-05, + "loss": 3.0947, + "step": 15906 + }, + { + "epoch": 0.19763330492887962, + "grad_norm": 0.15444546937942505, + "learning_rate": 3.0139673755605697e-05, + "loss": 3.0885, + "step": 15907 + }, + { + "epoch": 0.19766415139445304, + "grad_norm": 0.16121406853199005, + "learning_rate": 3.0137039296289794e-05, + "loss": 3.1554, + "step": 15908 + }, + { + "epoch": 0.19769499786002645, + "grad_norm": 0.15932439267635345, + "learning_rate": 3.0134404777413007e-05, + "loss": 3.1198, + "step": 15909 + }, + { + "epoch": 0.19772584432559986, + "grad_norm": 0.1647924929857254, + "learning_rate": 3.013177019900588e-05, + "loss": 3.1167, + "step": 15910 + }, + { + "epoch": 0.19775669079117328, + "grad_norm": 0.16101042926311493, + "learning_rate": 3.0129135561098963e-05, + "loss": 3.0784, + "step": 15911 + }, + { + "epoch": 0.1977875372567467, + "grad_norm": 0.16322346031665802, + "learning_rate": 3.0126500863722807e-05, + "loss": 3.108, + "step": 15912 + }, + { + "epoch": 0.1978183837223201, + "grad_norm": 0.15458396077156067, + "learning_rate": 3.0123866106907943e-05, + "loss": 3.1514, + "step": 15913 + }, + { + "epoch": 0.19784923018789352, + "grad_norm": 0.1568254828453064, + "learning_rate": 3.0121231290684937e-05, + "loss": 3.1107, + "step": 15914 + }, + { + "epoch": 0.19788007665346696, + "grad_norm": 0.15882088243961334, + "learning_rate": 3.0118596415084334e-05, + "loss": 3.1108, + "step": 15915 + }, + { + "epoch": 0.19791092311904038, + "grad_norm": 0.16014763712882996, + "learning_rate": 3.011596148013668e-05, + "loss": 3.1046, + "step": 15916 + }, + { + "epoch": 0.1979417695846138, + "grad_norm": 0.1669819951057434, + "learning_rate": 3.011332648587252e-05, + "loss": 3.122, + "step": 15917 + }, + { + "epoch": 0.1979726160501872, + "grad_norm": 0.15584056079387665, + "learning_rate": 3.0110691432322413e-05, + "loss": 3.1008, + "step": 15918 + }, + { + "epoch": 0.19800346251576062, + "grad_norm": 0.16959881782531738, + "learning_rate": 3.0108056319516918e-05, + "loss": 3.1031, + "step": 15919 + }, + { + "epoch": 0.19803430898133403, + "grad_norm": 0.16314363479614258, + "learning_rate": 3.010542114748657e-05, + "loss": 3.1103, + "step": 15920 + }, + { + "epoch": 0.19806515544690745, + "grad_norm": 0.1561867892742157, + "learning_rate": 3.0102785916261944e-05, + "loss": 3.1437, + "step": 15921 + }, + { + "epoch": 0.19809600191248086, + "grad_norm": 0.19892463088035583, + "learning_rate": 3.010015062587358e-05, + "loss": 3.0868, + "step": 15922 + }, + { + "epoch": 0.19812684837805428, + "grad_norm": 0.17320676147937775, + "learning_rate": 3.009751527635203e-05, + "loss": 3.1253, + "step": 15923 + }, + { + "epoch": 0.1981576948436277, + "grad_norm": 0.16845183074474335, + "learning_rate": 3.0094879867727847e-05, + "loss": 3.1021, + "step": 15924 + }, + { + "epoch": 0.1981885413092011, + "grad_norm": 0.4907933175563812, + "learning_rate": 3.0092244400031604e-05, + "loss": 3.1452, + "step": 15925 + }, + { + "epoch": 0.19821938777477452, + "grad_norm": 0.16564211249351501, + "learning_rate": 3.008960887329384e-05, + "loss": 3.077, + "step": 15926 + }, + { + "epoch": 0.19825023424034796, + "grad_norm": 0.16910354793071747, + "learning_rate": 3.008697328754513e-05, + "loss": 3.1136, + "step": 15927 + }, + { + "epoch": 0.19828108070592138, + "grad_norm": 0.16070139408111572, + "learning_rate": 3.008433764281601e-05, + "loss": 3.1267, + "step": 15928 + }, + { + "epoch": 0.1983119271714948, + "grad_norm": 0.1913096308708191, + "learning_rate": 3.0081701939137057e-05, + "loss": 3.11, + "step": 15929 + }, + { + "epoch": 0.1983427736370682, + "grad_norm": 0.17245756089687347, + "learning_rate": 3.007906617653882e-05, + "loss": 3.089, + "step": 15930 + }, + { + "epoch": 0.19837362010264162, + "grad_norm": 0.15494844317436218, + "learning_rate": 3.0076430355051865e-05, + "loss": 3.1096, + "step": 15931 + }, + { + "epoch": 0.19840446656821503, + "grad_norm": 0.16217140853405, + "learning_rate": 3.007379447470675e-05, + "loss": 3.0942, + "step": 15932 + }, + { + "epoch": 0.19843531303378845, + "grad_norm": 0.16479681432247162, + "learning_rate": 3.007115853553404e-05, + "loss": 3.1132, + "step": 15933 + }, + { + "epoch": 0.19846615949936186, + "grad_norm": 0.1646004319190979, + "learning_rate": 3.0068522537564285e-05, + "loss": 3.1169, + "step": 15934 + }, + { + "epoch": 0.19849700596493527, + "grad_norm": 0.16615861654281616, + "learning_rate": 3.0065886480828066e-05, + "loss": 3.0956, + "step": 15935 + }, + { + "epoch": 0.1985278524305087, + "grad_norm": 0.16217276453971863, + "learning_rate": 3.0063250365355933e-05, + "loss": 3.1133, + "step": 15936 + }, + { + "epoch": 0.1985586988960821, + "grad_norm": 0.1671789586544037, + "learning_rate": 3.006061419117846e-05, + "loss": 3.0924, + "step": 15937 + }, + { + "epoch": 0.19858954536165552, + "grad_norm": 0.15712870657444, + "learning_rate": 3.00579779583262e-05, + "loss": 3.1064, + "step": 15938 + }, + { + "epoch": 0.19862039182722896, + "grad_norm": 0.16644026339054108, + "learning_rate": 3.0055341666829733e-05, + "loss": 3.1006, + "step": 15939 + }, + { + "epoch": 0.19865123829280237, + "grad_norm": 0.22833435237407684, + "learning_rate": 3.005270531671961e-05, + "loss": 3.1064, + "step": 15940 + }, + { + "epoch": 0.1986820847583758, + "grad_norm": 0.1607903242111206, + "learning_rate": 3.005006890802641e-05, + "loss": 3.1025, + "step": 15941 + }, + { + "epoch": 0.1987129312239492, + "grad_norm": 0.15674103796482086, + "learning_rate": 3.0047432440780693e-05, + "loss": 3.0986, + "step": 15942 + }, + { + "epoch": 0.19874377768952262, + "grad_norm": 0.156593918800354, + "learning_rate": 3.004479591501303e-05, + "loss": 3.118, + "step": 15943 + }, + { + "epoch": 0.19877462415509603, + "grad_norm": 0.16385692358016968, + "learning_rate": 3.004215933075399e-05, + "loss": 3.1378, + "step": 15944 + }, + { + "epoch": 0.19880547062066944, + "grad_norm": 0.15573188662528992, + "learning_rate": 3.0039522688034142e-05, + "loss": 3.0483, + "step": 15945 + }, + { + "epoch": 0.19883631708624286, + "grad_norm": 0.15860307216644287, + "learning_rate": 3.0036885986884056e-05, + "loss": 3.138, + "step": 15946 + }, + { + "epoch": 0.19886716355181627, + "grad_norm": 0.15988482534885406, + "learning_rate": 3.0034249227334304e-05, + "loss": 3.1204, + "step": 15947 + }, + { + "epoch": 0.1988980100173897, + "grad_norm": 0.15627039968967438, + "learning_rate": 3.003161240941546e-05, + "loss": 3.0974, + "step": 15948 + }, + { + "epoch": 0.1989288564829631, + "grad_norm": 0.15051671862602234, + "learning_rate": 3.00289755331581e-05, + "loss": 3.1241, + "step": 15949 + }, + { + "epoch": 0.19895970294853652, + "grad_norm": 0.15856370329856873, + "learning_rate": 3.0026338598592778e-05, + "loss": 3.1364, + "step": 15950 + }, + { + "epoch": 0.19899054941410996, + "grad_norm": 0.15969502925872803, + "learning_rate": 3.002370160575008e-05, + "loss": 3.115, + "step": 15951 + }, + { + "epoch": 0.19902139587968337, + "grad_norm": 0.15248677134513855, + "learning_rate": 3.0021064554660594e-05, + "loss": 3.1215, + "step": 15952 + }, + { + "epoch": 0.19905224234525679, + "grad_norm": 0.1537630558013916, + "learning_rate": 3.0018427445354868e-05, + "loss": 3.1048, + "step": 15953 + }, + { + "epoch": 0.1990830888108302, + "grad_norm": 0.1623348891735077, + "learning_rate": 3.0015790277863504e-05, + "loss": 3.1263, + "step": 15954 + }, + { + "epoch": 0.19911393527640361, + "grad_norm": 0.1586964726448059, + "learning_rate": 3.0013153052217064e-05, + "loss": 3.1075, + "step": 15955 + }, + { + "epoch": 0.19914478174197703, + "grad_norm": 0.15919743478298187, + "learning_rate": 3.001051576844612e-05, + "loss": 3.098, + "step": 15956 + }, + { + "epoch": 0.19917562820755044, + "grad_norm": 0.15762296319007874, + "learning_rate": 3.0007878426581253e-05, + "loss": 3.1199, + "step": 15957 + }, + { + "epoch": 0.19920647467312386, + "grad_norm": 0.1582716405391693, + "learning_rate": 3.000524102665305e-05, + "loss": 3.1267, + "step": 15958 + }, + { + "epoch": 0.19923732113869727, + "grad_norm": 0.153951033949852, + "learning_rate": 3.0002603568692088e-05, + "loss": 3.122, + "step": 15959 + }, + { + "epoch": 0.19926816760427069, + "grad_norm": 0.1543404906988144, + "learning_rate": 2.9999966052728944e-05, + "loss": 3.1324, + "step": 15960 + }, + { + "epoch": 0.1992990140698441, + "grad_norm": 0.15086780488491058, + "learning_rate": 2.9997328478794194e-05, + "loss": 3.153, + "step": 15961 + }, + { + "epoch": 0.1993298605354175, + "grad_norm": 0.1595766693353653, + "learning_rate": 2.9994690846918434e-05, + "loss": 3.1121, + "step": 15962 + }, + { + "epoch": 0.19936070700099096, + "grad_norm": 0.15324072539806366, + "learning_rate": 2.9992053157132223e-05, + "loss": 3.0855, + "step": 15963 + }, + { + "epoch": 0.19939155346656437, + "grad_norm": 0.15878427028656006, + "learning_rate": 2.998941540946616e-05, + "loss": 3.0795, + "step": 15964 + }, + { + "epoch": 0.19942239993213778, + "grad_norm": 0.1542593389749527, + "learning_rate": 2.998677760395082e-05, + "loss": 3.0929, + "step": 15965 + }, + { + "epoch": 0.1994532463977112, + "grad_norm": 0.1542208045721054, + "learning_rate": 2.99841397406168e-05, + "loss": 3.1109, + "step": 15966 + }, + { + "epoch": 0.1994840928632846, + "grad_norm": 0.15498769283294678, + "learning_rate": 2.9981501819494666e-05, + "loss": 3.1408, + "step": 15967 + }, + { + "epoch": 0.19951493932885803, + "grad_norm": 0.15802806615829468, + "learning_rate": 2.9978863840615013e-05, + "loss": 3.105, + "step": 15968 + }, + { + "epoch": 0.19954578579443144, + "grad_norm": 0.2944446802139282, + "learning_rate": 2.9976225804008428e-05, + "loss": 3.0729, + "step": 15969 + }, + { + "epoch": 0.19957663226000485, + "grad_norm": 0.1541878581047058, + "learning_rate": 2.9973587709705493e-05, + "loss": 3.098, + "step": 15970 + }, + { + "epoch": 0.19960747872557827, + "grad_norm": 0.3370220959186554, + "learning_rate": 2.9970949557736798e-05, + "loss": 3.1338, + "step": 15971 + }, + { + "epoch": 0.19963832519115168, + "grad_norm": 0.15110023319721222, + "learning_rate": 2.996831134813294e-05, + "loss": 3.132, + "step": 15972 + }, + { + "epoch": 0.1996691716567251, + "grad_norm": 0.15801692008972168, + "learning_rate": 2.9965673080924484e-05, + "loss": 3.0969, + "step": 15973 + }, + { + "epoch": 0.1997000181222985, + "grad_norm": 0.152312234044075, + "learning_rate": 2.9963034756142027e-05, + "loss": 3.1204, + "step": 15974 + }, + { + "epoch": 0.19973086458787195, + "grad_norm": 0.14453765749931335, + "learning_rate": 2.996039637381618e-05, + "loss": 3.0903, + "step": 15975 + }, + { + "epoch": 0.19976171105344537, + "grad_norm": 0.16712048649787903, + "learning_rate": 2.995775793397751e-05, + "loss": 3.1239, + "step": 15976 + }, + { + "epoch": 0.19979255751901878, + "grad_norm": 0.1500294953584671, + "learning_rate": 2.995511943665662e-05, + "loss": 3.1502, + "step": 15977 + }, + { + "epoch": 0.1998234039845922, + "grad_norm": 0.15517067909240723, + "learning_rate": 2.9952480881884098e-05, + "loss": 3.0908, + "step": 15978 + }, + { + "epoch": 0.1998542504501656, + "grad_norm": 0.21110238134860992, + "learning_rate": 2.994984226969053e-05, + "loss": 3.1061, + "step": 15979 + }, + { + "epoch": 0.19988509691573902, + "grad_norm": 0.16422557830810547, + "learning_rate": 2.994720360010652e-05, + "loss": 3.1371, + "step": 15980 + }, + { + "epoch": 0.19991594338131244, + "grad_norm": 0.15005014836788177, + "learning_rate": 2.9944564873162663e-05, + "loss": 3.0982, + "step": 15981 + }, + { + "epoch": 0.19994678984688585, + "grad_norm": 0.15238474309444427, + "learning_rate": 2.9941926088889538e-05, + "loss": 3.0913, + "step": 15982 + }, + { + "epoch": 0.19997763631245927, + "grad_norm": 0.15392890572547913, + "learning_rate": 2.9939287247317755e-05, + "loss": 3.0935, + "step": 15983 + }, + { + "epoch": 0.20000848277803268, + "grad_norm": 0.1628565490245819, + "learning_rate": 2.9936648348477898e-05, + "loss": 3.1436, + "step": 15984 + }, + { + "epoch": 0.2000393292436061, + "grad_norm": 0.1590094119310379, + "learning_rate": 2.9934009392400576e-05, + "loss": 3.1064, + "step": 15985 + }, + { + "epoch": 0.2000701757091795, + "grad_norm": 0.1701040416955948, + "learning_rate": 2.993137037911638e-05, + "loss": 3.1098, + "step": 15986 + }, + { + "epoch": 0.20010102217475295, + "grad_norm": 0.15449537336826324, + "learning_rate": 2.9928731308655904e-05, + "loss": 3.1108, + "step": 15987 + }, + { + "epoch": 0.20013186864032637, + "grad_norm": 0.14812012016773224, + "learning_rate": 2.992609218104975e-05, + "loss": 3.0963, + "step": 15988 + }, + { + "epoch": 0.20016271510589978, + "grad_norm": 0.15277665853500366, + "learning_rate": 2.9923452996328526e-05, + "loss": 3.0969, + "step": 15989 + }, + { + "epoch": 0.2001935615714732, + "grad_norm": 0.1577366441488266, + "learning_rate": 2.9920813754522813e-05, + "loss": 3.1024, + "step": 15990 + }, + { + "epoch": 0.2002244080370466, + "grad_norm": 0.15060190856456757, + "learning_rate": 2.9918174455663227e-05, + "loss": 3.0916, + "step": 15991 + }, + { + "epoch": 0.20025525450262002, + "grad_norm": 0.1601070910692215, + "learning_rate": 2.9915535099780367e-05, + "loss": 3.107, + "step": 15992 + }, + { + "epoch": 0.20028610096819344, + "grad_norm": 0.1519109308719635, + "learning_rate": 2.991289568690483e-05, + "loss": 3.1156, + "step": 15993 + }, + { + "epoch": 0.20031694743376685, + "grad_norm": 0.15216684341430664, + "learning_rate": 2.9910256217067213e-05, + "loss": 3.1078, + "step": 15994 + }, + { + "epoch": 0.20034779389934027, + "grad_norm": 0.15646445751190186, + "learning_rate": 2.990761669029813e-05, + "loss": 3.1372, + "step": 15995 + }, + { + "epoch": 0.20037864036491368, + "grad_norm": 0.1678955852985382, + "learning_rate": 2.990497710662818e-05, + "loss": 3.0786, + "step": 15996 + }, + { + "epoch": 0.2004094868304871, + "grad_norm": 0.15313921868801117, + "learning_rate": 2.9902337466087965e-05, + "loss": 3.0892, + "step": 15997 + }, + { + "epoch": 0.2004403332960605, + "grad_norm": 0.15502531826496124, + "learning_rate": 2.98996977687081e-05, + "loss": 3.0885, + "step": 15998 + }, + { + "epoch": 0.20047117976163395, + "grad_norm": 0.15550793707370758, + "learning_rate": 2.989705801451918e-05, + "loss": 3.0971, + "step": 15999 + }, + { + "epoch": 0.20050202622720736, + "grad_norm": 0.15661659836769104, + "learning_rate": 2.9894418203551815e-05, + "loss": 3.0908, + "step": 16000 + }, + { + "epoch": 0.20053287269278078, + "grad_norm": 0.17222319543361664, + "learning_rate": 2.9891778335836607e-05, + "loss": 3.1017, + "step": 16001 + }, + { + "epoch": 0.2005637191583542, + "grad_norm": 0.16046728193759918, + "learning_rate": 2.9889138411404182e-05, + "loss": 3.1207, + "step": 16002 + }, + { + "epoch": 0.2005945656239276, + "grad_norm": 0.14537709951400757, + "learning_rate": 2.988649843028512e-05, + "loss": 3.0905, + "step": 16003 + }, + { + "epoch": 0.20062541208950102, + "grad_norm": 0.1588934361934662, + "learning_rate": 2.9883858392510055e-05, + "loss": 3.1211, + "step": 16004 + }, + { + "epoch": 0.20065625855507443, + "grad_norm": 0.1509387493133545, + "learning_rate": 2.9881218298109594e-05, + "loss": 3.076, + "step": 16005 + }, + { + "epoch": 0.20068710502064785, + "grad_norm": 0.15937963128089905, + "learning_rate": 2.987857814711432e-05, + "loss": 3.1178, + "step": 16006 + }, + { + "epoch": 0.20071795148622126, + "grad_norm": 0.1869591325521469, + "learning_rate": 2.9875937939554875e-05, + "loss": 3.0989, + "step": 16007 + }, + { + "epoch": 0.20074879795179468, + "grad_norm": 0.1505911648273468, + "learning_rate": 2.987329767546186e-05, + "loss": 3.068, + "step": 16008 + }, + { + "epoch": 0.2007796444173681, + "grad_norm": 0.1685837358236313, + "learning_rate": 2.9870657354865887e-05, + "loss": 3.0976, + "step": 16009 + }, + { + "epoch": 0.2008104908829415, + "grad_norm": 0.1603696495294571, + "learning_rate": 2.9868016977797568e-05, + "loss": 3.0848, + "step": 16010 + }, + { + "epoch": 0.20084133734851495, + "grad_norm": 0.16010715067386627, + "learning_rate": 2.9865376544287517e-05, + "loss": 3.0866, + "step": 16011 + }, + { + "epoch": 0.20087218381408836, + "grad_norm": 0.15625542402267456, + "learning_rate": 2.9862736054366352e-05, + "loss": 3.1029, + "step": 16012 + }, + { + "epoch": 0.20090303027966178, + "grad_norm": 0.16222508251667023, + "learning_rate": 2.986009550806468e-05, + "loss": 3.1103, + "step": 16013 + }, + { + "epoch": 0.2009338767452352, + "grad_norm": 0.19887368381023407, + "learning_rate": 2.985745490541312e-05, + "loss": 3.1284, + "step": 16014 + }, + { + "epoch": 0.2009647232108086, + "grad_norm": 0.16025404632091522, + "learning_rate": 2.9854814246442297e-05, + "loss": 3.1018, + "step": 16015 + }, + { + "epoch": 0.20099556967638202, + "grad_norm": 0.17413990199565887, + "learning_rate": 2.9852173531182816e-05, + "loss": 3.1021, + "step": 16016 + }, + { + "epoch": 0.20102641614195543, + "grad_norm": 0.19243872165679932, + "learning_rate": 2.98495327596653e-05, + "loss": 3.1115, + "step": 16017 + }, + { + "epoch": 0.20105726260752885, + "grad_norm": 0.16040828824043274, + "learning_rate": 2.9846891931920356e-05, + "loss": 3.0895, + "step": 16018 + }, + { + "epoch": 0.20108810907310226, + "grad_norm": 0.15058571100234985, + "learning_rate": 2.9844251047978627e-05, + "loss": 3.0923, + "step": 16019 + }, + { + "epoch": 0.20111895553867568, + "grad_norm": 0.15287195146083832, + "learning_rate": 2.984161010787071e-05, + "loss": 3.0839, + "step": 16020 + }, + { + "epoch": 0.2011498020042491, + "grad_norm": 0.1553541123867035, + "learning_rate": 2.9838969111627235e-05, + "loss": 3.0779, + "step": 16021 + }, + { + "epoch": 0.2011806484698225, + "grad_norm": 0.1534089744091034, + "learning_rate": 2.983632805927883e-05, + "loss": 3.1189, + "step": 16022 + }, + { + "epoch": 0.20121149493539595, + "grad_norm": 0.15728013217449188, + "learning_rate": 2.9833686950856093e-05, + "loss": 3.0917, + "step": 16023 + }, + { + "epoch": 0.20124234140096936, + "grad_norm": 0.16582120954990387, + "learning_rate": 2.983104578638966e-05, + "loss": 3.1172, + "step": 16024 + }, + { + "epoch": 0.20127318786654277, + "grad_norm": 0.14986275136470795, + "learning_rate": 2.9828404565910167e-05, + "loss": 3.0762, + "step": 16025 + }, + { + "epoch": 0.2013040343321162, + "grad_norm": 0.15821897983551025, + "learning_rate": 2.9825763289448218e-05, + "loss": 3.0734, + "step": 16026 + }, + { + "epoch": 0.2013348807976896, + "grad_norm": 0.16852962970733643, + "learning_rate": 2.9823121957034443e-05, + "loss": 3.1053, + "step": 16027 + }, + { + "epoch": 0.20136572726326302, + "grad_norm": 0.16565535962581635, + "learning_rate": 2.982048056869947e-05, + "loss": 3.1138, + "step": 16028 + }, + { + "epoch": 0.20139657372883643, + "grad_norm": 0.1557822972536087, + "learning_rate": 2.981783912447392e-05, + "loss": 3.0876, + "step": 16029 + }, + { + "epoch": 0.20142742019440985, + "grad_norm": 0.15926538407802582, + "learning_rate": 2.9815197624388413e-05, + "loss": 3.1286, + "step": 16030 + }, + { + "epoch": 0.20145826665998326, + "grad_norm": 0.15637320280075073, + "learning_rate": 2.9812556068473597e-05, + "loss": 3.101, + "step": 16031 + }, + { + "epoch": 0.20148911312555667, + "grad_norm": 0.1508069783449173, + "learning_rate": 2.9809914456760073e-05, + "loss": 3.1123, + "step": 16032 + }, + { + "epoch": 0.2015199595911301, + "grad_norm": 0.1734161525964737, + "learning_rate": 2.980727278927849e-05, + "loss": 3.1379, + "step": 16033 + }, + { + "epoch": 0.2015508060567035, + "grad_norm": 0.16487130522727966, + "learning_rate": 2.9804631066059463e-05, + "loss": 3.1159, + "step": 16034 + }, + { + "epoch": 0.20158165252227694, + "grad_norm": 0.1578454226255417, + "learning_rate": 2.9801989287133625e-05, + "loss": 3.1193, + "step": 16035 + }, + { + "epoch": 0.20161249898785036, + "grad_norm": 0.16151063144207, + "learning_rate": 2.979934745253161e-05, + "loss": 3.0772, + "step": 16036 + }, + { + "epoch": 0.20164334545342377, + "grad_norm": 0.15524424612522125, + "learning_rate": 2.979670556228405e-05, + "loss": 3.142, + "step": 16037 + }, + { + "epoch": 0.2016741919189972, + "grad_norm": 0.1499778926372528, + "learning_rate": 2.9794063616421565e-05, + "loss": 3.0901, + "step": 16038 + }, + { + "epoch": 0.2017050383845706, + "grad_norm": 0.16010501980781555, + "learning_rate": 2.9791421614974795e-05, + "loss": 3.0599, + "step": 16039 + }, + { + "epoch": 0.20173588485014401, + "grad_norm": 0.14978821575641632, + "learning_rate": 2.9788779557974373e-05, + "loss": 3.093, + "step": 16040 + }, + { + "epoch": 0.20176673131571743, + "grad_norm": 0.1502017378807068, + "learning_rate": 2.9786137445450924e-05, + "loss": 3.128, + "step": 16041 + }, + { + "epoch": 0.20179757778129084, + "grad_norm": 0.15096865594387054, + "learning_rate": 2.978349527743509e-05, + "loss": 3.092, + "step": 16042 + }, + { + "epoch": 0.20182842424686426, + "grad_norm": 0.14782695472240448, + "learning_rate": 2.978085305395751e-05, + "loss": 3.0565, + "step": 16043 + }, + { + "epoch": 0.20185927071243767, + "grad_norm": 0.15162959694862366, + "learning_rate": 2.9778210775048804e-05, + "loss": 3.1318, + "step": 16044 + }, + { + "epoch": 0.20189011717801109, + "grad_norm": 0.15676888823509216, + "learning_rate": 2.9775568440739614e-05, + "loss": 3.137, + "step": 16045 + }, + { + "epoch": 0.2019209636435845, + "grad_norm": 0.1525530070066452, + "learning_rate": 2.9772926051060578e-05, + "loss": 3.1, + "step": 16046 + }, + { + "epoch": 0.20195181010915794, + "grad_norm": 0.15019075572490692, + "learning_rate": 2.977028360604234e-05, + "loss": 3.0847, + "step": 16047 + }, + { + "epoch": 0.20198265657473136, + "grad_norm": 0.15782617032527924, + "learning_rate": 2.9767641105715527e-05, + "loss": 3.1533, + "step": 16048 + }, + { + "epoch": 0.20201350304030477, + "grad_norm": 0.161017045378685, + "learning_rate": 2.9764998550110777e-05, + "loss": 3.1119, + "step": 16049 + }, + { + "epoch": 0.20204434950587818, + "grad_norm": 0.15169303119182587, + "learning_rate": 2.9762355939258733e-05, + "loss": 3.1317, + "step": 16050 + }, + { + "epoch": 0.2020751959714516, + "grad_norm": 0.15554280579090118, + "learning_rate": 2.975971327319003e-05, + "loss": 3.104, + "step": 16051 + }, + { + "epoch": 0.202106042437025, + "grad_norm": 0.15654048323631287, + "learning_rate": 2.9757070551935323e-05, + "loss": 3.1218, + "step": 16052 + }, + { + "epoch": 0.20213688890259843, + "grad_norm": 0.16435745358467102, + "learning_rate": 2.9754427775525228e-05, + "loss": 3.1122, + "step": 16053 + }, + { + "epoch": 0.20216773536817184, + "grad_norm": 0.15620207786560059, + "learning_rate": 2.975178494399041e-05, + "loss": 3.1172, + "step": 16054 + }, + { + "epoch": 0.20219858183374526, + "grad_norm": 0.15012024343013763, + "learning_rate": 2.9749142057361503e-05, + "loss": 3.1008, + "step": 16055 + }, + { + "epoch": 0.20222942829931867, + "grad_norm": 0.19000466167926788, + "learning_rate": 2.974649911566914e-05, + "loss": 3.0992, + "step": 16056 + }, + { + "epoch": 0.20226027476489208, + "grad_norm": 0.15615376830101013, + "learning_rate": 2.9743856118943976e-05, + "loss": 3.1133, + "step": 16057 + }, + { + "epoch": 0.2022911212304655, + "grad_norm": 0.15852102637290955, + "learning_rate": 2.9741213067216643e-05, + "loss": 3.0888, + "step": 16058 + }, + { + "epoch": 0.20232196769603894, + "grad_norm": 0.15237951278686523, + "learning_rate": 2.9738569960517804e-05, + "loss": 3.0702, + "step": 16059 + }, + { + "epoch": 0.20235281416161235, + "grad_norm": 0.15717914700508118, + "learning_rate": 2.9735926798878093e-05, + "loss": 3.1015, + "step": 16060 + }, + { + "epoch": 0.20238366062718577, + "grad_norm": 0.1602219045162201, + "learning_rate": 2.9733283582328154e-05, + "loss": 3.0712, + "step": 16061 + }, + { + "epoch": 0.20241450709275918, + "grad_norm": 0.159909188747406, + "learning_rate": 2.9730640310898638e-05, + "loss": 3.1191, + "step": 16062 + }, + { + "epoch": 0.2024453535583326, + "grad_norm": 0.15303701162338257, + "learning_rate": 2.9727996984620188e-05, + "loss": 3.0812, + "step": 16063 + }, + { + "epoch": 0.202476200023906, + "grad_norm": 0.15960311889648438, + "learning_rate": 2.9725353603523453e-05, + "loss": 3.1, + "step": 16064 + }, + { + "epoch": 0.20250704648947943, + "grad_norm": 0.1501045674085617, + "learning_rate": 2.9722710167639086e-05, + "loss": 3.1135, + "step": 16065 + }, + { + "epoch": 0.20253789295505284, + "grad_norm": 0.1579052358865738, + "learning_rate": 2.972006667699773e-05, + "loss": 3.0988, + "step": 16066 + }, + { + "epoch": 0.20256873942062625, + "grad_norm": 0.16051684319972992, + "learning_rate": 2.971742313163004e-05, + "loss": 3.1331, + "step": 16067 + }, + { + "epoch": 0.20259958588619967, + "grad_norm": 0.16287249326705933, + "learning_rate": 2.971477953156666e-05, + "loss": 3.129, + "step": 16068 + }, + { + "epoch": 0.20263043235177308, + "grad_norm": 0.16095079481601715, + "learning_rate": 2.971213587683825e-05, + "loss": 3.0786, + "step": 16069 + }, + { + "epoch": 0.2026612788173465, + "grad_norm": 0.1558690369129181, + "learning_rate": 2.970949216747545e-05, + "loss": 3.1348, + "step": 16070 + }, + { + "epoch": 0.20269212528291994, + "grad_norm": 0.18465758860111237, + "learning_rate": 2.9706848403508925e-05, + "loss": 3.0874, + "step": 16071 + }, + { + "epoch": 0.20272297174849335, + "grad_norm": 0.15472833812236786, + "learning_rate": 2.970420458496932e-05, + "loss": 3.1384, + "step": 16072 + }, + { + "epoch": 0.20275381821406677, + "grad_norm": 0.1558108925819397, + "learning_rate": 2.9701560711887283e-05, + "loss": 3.0883, + "step": 16073 + }, + { + "epoch": 0.20278466467964018, + "grad_norm": 0.15693560242652893, + "learning_rate": 2.969891678429348e-05, + "loss": 3.1415, + "step": 16074 + }, + { + "epoch": 0.2028155111452136, + "grad_norm": 0.1640184372663498, + "learning_rate": 2.969627280221856e-05, + "loss": 3.124, + "step": 16075 + }, + { + "epoch": 0.202846357610787, + "grad_norm": 0.16647689044475555, + "learning_rate": 2.9693628765693176e-05, + "loss": 3.0831, + "step": 16076 + }, + { + "epoch": 0.20287720407636042, + "grad_norm": 0.2246648371219635, + "learning_rate": 2.9690984674747995e-05, + "loss": 3.0739, + "step": 16077 + }, + { + "epoch": 0.20290805054193384, + "grad_norm": 0.1578930914402008, + "learning_rate": 2.9688340529413662e-05, + "loss": 3.1117, + "step": 16078 + }, + { + "epoch": 0.20293889700750725, + "grad_norm": 0.1587049663066864, + "learning_rate": 2.9685696329720834e-05, + "loss": 3.097, + "step": 16079 + }, + { + "epoch": 0.20296974347308067, + "grad_norm": 0.1719014197587967, + "learning_rate": 2.9683052075700174e-05, + "loss": 3.1068, + "step": 16080 + }, + { + "epoch": 0.20300058993865408, + "grad_norm": 0.16420264542102814, + "learning_rate": 2.968040776738234e-05, + "loss": 3.1092, + "step": 16081 + }, + { + "epoch": 0.2030314364042275, + "grad_norm": 0.17876674234867096, + "learning_rate": 2.967776340479799e-05, + "loss": 3.1087, + "step": 16082 + }, + { + "epoch": 0.20306228286980094, + "grad_norm": 0.16549856960773468, + "learning_rate": 2.967511898797779e-05, + "loss": 3.1244, + "step": 16083 + }, + { + "epoch": 0.20309312933537435, + "grad_norm": 0.16289325058460236, + "learning_rate": 2.9672474516952386e-05, + "loss": 3.1312, + "step": 16084 + }, + { + "epoch": 0.20312397580094776, + "grad_norm": 0.158726766705513, + "learning_rate": 2.966982999175245e-05, + "loss": 3.0936, + "step": 16085 + }, + { + "epoch": 0.20315482226652118, + "grad_norm": 0.15859250724315643, + "learning_rate": 2.9667185412408645e-05, + "loss": 3.1216, + "step": 16086 + }, + { + "epoch": 0.2031856687320946, + "grad_norm": 0.1507055014371872, + "learning_rate": 2.9664540778951634e-05, + "loss": 3.0799, + "step": 16087 + }, + { + "epoch": 0.203216515197668, + "grad_norm": 0.16538585722446442, + "learning_rate": 2.9661896091412067e-05, + "loss": 3.1165, + "step": 16088 + }, + { + "epoch": 0.20324736166324142, + "grad_norm": 0.15683747828006744, + "learning_rate": 2.965925134982062e-05, + "loss": 3.0989, + "step": 16089 + }, + { + "epoch": 0.20327820812881484, + "grad_norm": 0.15491226315498352, + "learning_rate": 2.9656606554207948e-05, + "loss": 3.1102, + "step": 16090 + }, + { + "epoch": 0.20330905459438825, + "grad_norm": 0.15495218336582184, + "learning_rate": 2.9653961704604727e-05, + "loss": 3.1269, + "step": 16091 + }, + { + "epoch": 0.20333990105996166, + "grad_norm": 0.15314766764640808, + "learning_rate": 2.965131680104162e-05, + "loss": 3.1088, + "step": 16092 + }, + { + "epoch": 0.20337074752553508, + "grad_norm": 0.16174595057964325, + "learning_rate": 2.9648671843549287e-05, + "loss": 3.0952, + "step": 16093 + }, + { + "epoch": 0.2034015939911085, + "grad_norm": 0.16178885102272034, + "learning_rate": 2.964602683215839e-05, + "loss": 3.1673, + "step": 16094 + }, + { + "epoch": 0.20343244045668193, + "grad_norm": 0.15734589099884033, + "learning_rate": 2.9643381766899615e-05, + "loss": 3.0731, + "step": 16095 + }, + { + "epoch": 0.20346328692225535, + "grad_norm": 0.16477559506893158, + "learning_rate": 2.9640736647803613e-05, + "loss": 3.1178, + "step": 16096 + }, + { + "epoch": 0.20349413338782876, + "grad_norm": 0.1501559317111969, + "learning_rate": 2.9638091474901052e-05, + "loss": 3.1136, + "step": 16097 + }, + { + "epoch": 0.20352497985340218, + "grad_norm": 0.15143626928329468, + "learning_rate": 2.963544624822262e-05, + "loss": 3.1321, + "step": 16098 + }, + { + "epoch": 0.2035558263189756, + "grad_norm": 0.15980324149131775, + "learning_rate": 2.9632800967798973e-05, + "loss": 3.1157, + "step": 16099 + }, + { + "epoch": 0.203586672784549, + "grad_norm": 0.1544918268918991, + "learning_rate": 2.9630155633660774e-05, + "loss": 3.1226, + "step": 16100 + }, + { + "epoch": 0.20361751925012242, + "grad_norm": 0.15945996344089508, + "learning_rate": 2.9627510245838707e-05, + "loss": 3.1111, + "step": 16101 + }, + { + "epoch": 0.20364836571569583, + "grad_norm": 0.1515236496925354, + "learning_rate": 2.962486480436344e-05, + "loss": 3.1191, + "step": 16102 + }, + { + "epoch": 0.20367921218126925, + "grad_norm": 0.15353864431381226, + "learning_rate": 2.9622219309265642e-05, + "loss": 3.1173, + "step": 16103 + }, + { + "epoch": 0.20371005864684266, + "grad_norm": 0.152169868350029, + "learning_rate": 2.9619573760575996e-05, + "loss": 3.1049, + "step": 16104 + }, + { + "epoch": 0.20374090511241608, + "grad_norm": 0.14998891949653625, + "learning_rate": 2.961692815832517e-05, + "loss": 3.1135, + "step": 16105 + }, + { + "epoch": 0.20377175157798952, + "grad_norm": 0.2742878198623657, + "learning_rate": 2.961428250254382e-05, + "loss": 3.1213, + "step": 16106 + }, + { + "epoch": 0.20380259804356293, + "grad_norm": 0.14999642968177795, + "learning_rate": 2.9611636793262647e-05, + "loss": 3.0899, + "step": 16107 + }, + { + "epoch": 0.20383344450913635, + "grad_norm": 0.15048080682754517, + "learning_rate": 2.9608991030512318e-05, + "loss": 3.1207, + "step": 16108 + }, + { + "epoch": 0.20386429097470976, + "grad_norm": 0.14959712326526642, + "learning_rate": 2.960634521432351e-05, + "loss": 3.1229, + "step": 16109 + }, + { + "epoch": 0.20389513744028318, + "grad_norm": 0.15316303074359894, + "learning_rate": 2.96036993447269e-05, + "loss": 3.0818, + "step": 16110 + }, + { + "epoch": 0.2039259839058566, + "grad_norm": 0.15532495081424713, + "learning_rate": 2.960105342175315e-05, + "loss": 3.1392, + "step": 16111 + }, + { + "epoch": 0.20395683037143, + "grad_norm": 0.14841794967651367, + "learning_rate": 2.9598407445432963e-05, + "loss": 3.1668, + "step": 16112 + }, + { + "epoch": 0.20398767683700342, + "grad_norm": 0.17082586884498596, + "learning_rate": 2.9595761415796997e-05, + "loss": 3.083, + "step": 16113 + }, + { + "epoch": 0.20401852330257683, + "grad_norm": 0.1569250375032425, + "learning_rate": 2.9593115332875936e-05, + "loss": 3.1015, + "step": 16114 + }, + { + "epoch": 0.20404936976815025, + "grad_norm": 0.15173625946044922, + "learning_rate": 2.959046919670047e-05, + "loss": 3.0913, + "step": 16115 + }, + { + "epoch": 0.20408021623372366, + "grad_norm": 0.1590459942817688, + "learning_rate": 2.9587823007301273e-05, + "loss": 3.105, + "step": 16116 + }, + { + "epoch": 0.20411106269929707, + "grad_norm": 0.16619566082954407, + "learning_rate": 2.9585176764709023e-05, + "loss": 3.1169, + "step": 16117 + }, + { + "epoch": 0.20414190916487052, + "grad_norm": 0.1540401428937912, + "learning_rate": 2.95825304689544e-05, + "loss": 3.1316, + "step": 16118 + }, + { + "epoch": 0.20417275563044393, + "grad_norm": 0.15120929479599, + "learning_rate": 2.9579884120068097e-05, + "loss": 3.0843, + "step": 16119 + }, + { + "epoch": 0.20420360209601734, + "grad_norm": 0.1546134054660797, + "learning_rate": 2.957723771808078e-05, + "loss": 3.0914, + "step": 16120 + }, + { + "epoch": 0.20423444856159076, + "grad_norm": 0.15703853964805603, + "learning_rate": 2.9574591263023156e-05, + "loss": 3.1216, + "step": 16121 + }, + { + "epoch": 0.20426529502716417, + "grad_norm": 0.1524684876203537, + "learning_rate": 2.957194475492589e-05, + "loss": 3.1044, + "step": 16122 + }, + { + "epoch": 0.2042961414927376, + "grad_norm": 0.16186480224132538, + "learning_rate": 2.9569298193819667e-05, + "loss": 3.0863, + "step": 16123 + }, + { + "epoch": 0.204326987958311, + "grad_norm": 0.148681178689003, + "learning_rate": 2.956665157973518e-05, + "loss": 3.0933, + "step": 16124 + }, + { + "epoch": 0.20435783442388442, + "grad_norm": 0.1512828767299652, + "learning_rate": 2.9564004912703115e-05, + "loss": 3.0731, + "step": 16125 + }, + { + "epoch": 0.20438868088945783, + "grad_norm": 0.17368657886981964, + "learning_rate": 2.9561358192754148e-05, + "loss": 3.12, + "step": 16126 + }, + { + "epoch": 0.20441952735503124, + "grad_norm": 0.15310581028461456, + "learning_rate": 2.9558711419918985e-05, + "loss": 3.1352, + "step": 16127 + }, + { + "epoch": 0.20445037382060466, + "grad_norm": 0.16189372539520264, + "learning_rate": 2.9556064594228295e-05, + "loss": 3.0894, + "step": 16128 + }, + { + "epoch": 0.20448122028617807, + "grad_norm": 0.18671108782291412, + "learning_rate": 2.955341771571277e-05, + "loss": 3.1186, + "step": 16129 + }, + { + "epoch": 0.20451206675175151, + "grad_norm": 0.1487981230020523, + "learning_rate": 2.955077078440311e-05, + "loss": 3.128, + "step": 16130 + }, + { + "epoch": 0.20454291321732493, + "grad_norm": 0.17057020962238312, + "learning_rate": 2.9548123800329996e-05, + "loss": 3.1463, + "step": 16131 + }, + { + "epoch": 0.20457375968289834, + "grad_norm": 0.1585511416196823, + "learning_rate": 2.9545476763524116e-05, + "loss": 3.1111, + "step": 16132 + }, + { + "epoch": 0.20460460614847176, + "grad_norm": 0.1505124717950821, + "learning_rate": 2.9542829674016172e-05, + "loss": 3.1052, + "step": 16133 + }, + { + "epoch": 0.20463545261404517, + "grad_norm": 0.16435743868350983, + "learning_rate": 2.9540182531836842e-05, + "loss": 3.1255, + "step": 16134 + }, + { + "epoch": 0.20466629907961859, + "grad_norm": 0.15343047678470612, + "learning_rate": 2.953753533701682e-05, + "loss": 3.0876, + "step": 16135 + }, + { + "epoch": 0.204697145545192, + "grad_norm": 0.15951313078403473, + "learning_rate": 2.953488808958681e-05, + "loss": 3.0991, + "step": 16136 + }, + { + "epoch": 0.2047279920107654, + "grad_norm": 0.15696021914482117, + "learning_rate": 2.9532240789577497e-05, + "loss": 3.0759, + "step": 16137 + }, + { + "epoch": 0.20475883847633883, + "grad_norm": 0.17350925505161285, + "learning_rate": 2.9529593437019575e-05, + "loss": 3.1359, + "step": 16138 + }, + { + "epoch": 0.20478968494191224, + "grad_norm": 0.15698254108428955, + "learning_rate": 2.9526946031943743e-05, + "loss": 3.1233, + "step": 16139 + }, + { + "epoch": 0.20482053140748566, + "grad_norm": 0.15327611565589905, + "learning_rate": 2.952429857438068e-05, + "loss": 3.0945, + "step": 16140 + }, + { + "epoch": 0.20485137787305907, + "grad_norm": 0.15326493978500366, + "learning_rate": 2.95216510643611e-05, + "loss": 3.1149, + "step": 16141 + }, + { + "epoch": 0.2048822243386325, + "grad_norm": 0.1533735990524292, + "learning_rate": 2.95190035019157e-05, + "loss": 3.1266, + "step": 16142 + }, + { + "epoch": 0.20491307080420593, + "grad_norm": 0.15500620007514954, + "learning_rate": 2.9516355887075164e-05, + "loss": 3.0819, + "step": 16143 + }, + { + "epoch": 0.20494391726977934, + "grad_norm": 0.16029293835163116, + "learning_rate": 2.9513708219870194e-05, + "loss": 3.118, + "step": 16144 + }, + { + "epoch": 0.20497476373535276, + "grad_norm": 0.1514861285686493, + "learning_rate": 2.9511060500331493e-05, + "loss": 3.0874, + "step": 16145 + }, + { + "epoch": 0.20500561020092617, + "grad_norm": 0.15283674001693726, + "learning_rate": 2.9508412728489754e-05, + "loss": 3.1288, + "step": 16146 + }, + { + "epoch": 0.20503645666649958, + "grad_norm": 0.15623688697814941, + "learning_rate": 2.9505764904375678e-05, + "loss": 3.1234, + "step": 16147 + }, + { + "epoch": 0.205067303132073, + "grad_norm": 0.15283216536045074, + "learning_rate": 2.9503117028019967e-05, + "loss": 3.1284, + "step": 16148 + }, + { + "epoch": 0.2050981495976464, + "grad_norm": 0.15997551381587982, + "learning_rate": 2.9500469099453325e-05, + "loss": 3.1346, + "step": 16149 + }, + { + "epoch": 0.20512899606321983, + "grad_norm": 0.15681718289852142, + "learning_rate": 2.949782111870644e-05, + "loss": 3.0903, + "step": 16150 + }, + { + "epoch": 0.20515984252879324, + "grad_norm": 0.1537199318408966, + "learning_rate": 2.9495173085810022e-05, + "loss": 3.0926, + "step": 16151 + }, + { + "epoch": 0.20519068899436665, + "grad_norm": 0.15928786993026733, + "learning_rate": 2.949252500079478e-05, + "loss": 3.1101, + "step": 16152 + }, + { + "epoch": 0.20522153545994007, + "grad_norm": 0.1542109251022339, + "learning_rate": 2.9489876863691407e-05, + "loss": 3.1168, + "step": 16153 + }, + { + "epoch": 0.2052523819255135, + "grad_norm": 0.15714581310749054, + "learning_rate": 2.948722867453061e-05, + "loss": 3.1081, + "step": 16154 + }, + { + "epoch": 0.20528322839108692, + "grad_norm": 0.15186813473701477, + "learning_rate": 2.9484580433343097e-05, + "loss": 3.0679, + "step": 16155 + }, + { + "epoch": 0.20531407485666034, + "grad_norm": 0.16084004938602448, + "learning_rate": 2.9481932140159562e-05, + "loss": 3.0844, + "step": 16156 + }, + { + "epoch": 0.20534492132223375, + "grad_norm": 0.15825578570365906, + "learning_rate": 2.947928379501072e-05, + "loss": 3.1325, + "step": 16157 + }, + { + "epoch": 0.20537576778780717, + "grad_norm": 0.15613232553005219, + "learning_rate": 2.947663539792727e-05, + "loss": 3.0849, + "step": 16158 + }, + { + "epoch": 0.20540661425338058, + "grad_norm": 0.1528795063495636, + "learning_rate": 2.947398694893993e-05, + "loss": 3.1151, + "step": 16159 + }, + { + "epoch": 0.205437460718954, + "grad_norm": 0.15143704414367676, + "learning_rate": 2.94713384480794e-05, + "loss": 3.102, + "step": 16160 + }, + { + "epoch": 0.2054683071845274, + "grad_norm": 0.15095025300979614, + "learning_rate": 2.9468689895376378e-05, + "loss": 3.0863, + "step": 16161 + }, + { + "epoch": 0.20549915365010082, + "grad_norm": 0.15050630271434784, + "learning_rate": 2.9466041290861594e-05, + "loss": 3.073, + "step": 16162 + }, + { + "epoch": 0.20553000011567424, + "grad_norm": 0.1545991748571396, + "learning_rate": 2.9463392634565735e-05, + "loss": 3.1126, + "step": 16163 + }, + { + "epoch": 0.20556084658124765, + "grad_norm": 0.15235182642936707, + "learning_rate": 2.9460743926519524e-05, + "loss": 3.0815, + "step": 16164 + }, + { + "epoch": 0.20559169304682107, + "grad_norm": 0.15803271532058716, + "learning_rate": 2.9458095166753674e-05, + "loss": 3.1335, + "step": 16165 + }, + { + "epoch": 0.2056225395123945, + "grad_norm": 0.15248215198516846, + "learning_rate": 2.945544635529888e-05, + "loss": 3.097, + "step": 16166 + }, + { + "epoch": 0.20565338597796792, + "grad_norm": 0.1581910252571106, + "learning_rate": 2.9452797492185867e-05, + "loss": 3.0947, + "step": 16167 + }, + { + "epoch": 0.20568423244354134, + "grad_norm": 0.15737858414649963, + "learning_rate": 2.945014857744534e-05, + "loss": 3.0825, + "step": 16168 + }, + { + "epoch": 0.20571507890911475, + "grad_norm": 0.15974809229373932, + "learning_rate": 2.944749961110802e-05, + "loss": 3.1354, + "step": 16169 + }, + { + "epoch": 0.20574592537468817, + "grad_norm": 0.14967834949493408, + "learning_rate": 2.9444850593204613e-05, + "loss": 3.0947, + "step": 16170 + }, + { + "epoch": 0.20577677184026158, + "grad_norm": 0.1522214412689209, + "learning_rate": 2.9442201523765832e-05, + "loss": 3.1572, + "step": 16171 + }, + { + "epoch": 0.205807618305835, + "grad_norm": 0.15311923623085022, + "learning_rate": 2.9439552402822402e-05, + "loss": 3.1223, + "step": 16172 + }, + { + "epoch": 0.2058384647714084, + "grad_norm": 0.15006095170974731, + "learning_rate": 2.943690323040502e-05, + "loss": 3.1006, + "step": 16173 + }, + { + "epoch": 0.20586931123698182, + "grad_norm": 0.19364207983016968, + "learning_rate": 2.9434254006544415e-05, + "loss": 3.1159, + "step": 16174 + }, + { + "epoch": 0.20590015770255524, + "grad_norm": 0.15246716141700745, + "learning_rate": 2.94316047312713e-05, + "loss": 3.075, + "step": 16175 + }, + { + "epoch": 0.20593100416812865, + "grad_norm": 0.15598483383655548, + "learning_rate": 2.9428955404616386e-05, + "loss": 3.1311, + "step": 16176 + }, + { + "epoch": 0.20596185063370206, + "grad_norm": 0.15629014372825623, + "learning_rate": 2.9426306026610407e-05, + "loss": 3.0981, + "step": 16177 + }, + { + "epoch": 0.2059926970992755, + "grad_norm": 0.15906913578510284, + "learning_rate": 2.942365659728406e-05, + "loss": 3.1342, + "step": 16178 + }, + { + "epoch": 0.20602354356484892, + "grad_norm": 0.16472788155078888, + "learning_rate": 2.942100711666807e-05, + "loss": 3.1103, + "step": 16179 + }, + { + "epoch": 0.20605439003042234, + "grad_norm": 0.15834788978099823, + "learning_rate": 2.9418357584793165e-05, + "loss": 3.117, + "step": 16180 + }, + { + "epoch": 0.20608523649599575, + "grad_norm": 0.16102197766304016, + "learning_rate": 2.9415708001690063e-05, + "loss": 3.1092, + "step": 16181 + }, + { + "epoch": 0.20611608296156916, + "grad_norm": 0.16403387486934662, + "learning_rate": 2.941305836738947e-05, + "loss": 3.1266, + "step": 16182 + }, + { + "epoch": 0.20614692942714258, + "grad_norm": 0.1499263048171997, + "learning_rate": 2.941040868192213e-05, + "loss": 3.108, + "step": 16183 + }, + { + "epoch": 0.206177775892716, + "grad_norm": 0.15920214354991913, + "learning_rate": 2.9407758945318742e-05, + "loss": 3.1087, + "step": 16184 + }, + { + "epoch": 0.2062086223582894, + "grad_norm": 0.16421782970428467, + "learning_rate": 2.9405109157610035e-05, + "loss": 3.1257, + "step": 16185 + }, + { + "epoch": 0.20623946882386282, + "grad_norm": 0.1555650532245636, + "learning_rate": 2.940245931882674e-05, + "loss": 3.0879, + "step": 16186 + }, + { + "epoch": 0.20627031528943623, + "grad_norm": 0.16868318617343903, + "learning_rate": 2.939980942899958e-05, + "loss": 3.1331, + "step": 16187 + }, + { + "epoch": 0.20630116175500965, + "grad_norm": 0.2999184727668762, + "learning_rate": 2.9397159488159265e-05, + "loss": 3.1, + "step": 16188 + }, + { + "epoch": 0.20633200822058306, + "grad_norm": 0.15331678092479706, + "learning_rate": 2.9394509496336536e-05, + "loss": 3.0751, + "step": 16189 + }, + { + "epoch": 0.2063628546861565, + "grad_norm": 0.15838384628295898, + "learning_rate": 2.9391859453562105e-05, + "loss": 3.1058, + "step": 16190 + }, + { + "epoch": 0.20639370115172992, + "grad_norm": 0.1670714169740677, + "learning_rate": 2.93892093598667e-05, + "loss": 3.0913, + "step": 16191 + }, + { + "epoch": 0.20642454761730333, + "grad_norm": 0.162473663687706, + "learning_rate": 2.9386559215281056e-05, + "loss": 3.1071, + "step": 16192 + }, + { + "epoch": 0.20645539408287675, + "grad_norm": 0.16599100828170776, + "learning_rate": 2.9383909019835892e-05, + "loss": 3.0496, + "step": 16193 + }, + { + "epoch": 0.20648624054845016, + "grad_norm": 0.17181473970413208, + "learning_rate": 2.9381258773561937e-05, + "loss": 3.096, + "step": 16194 + }, + { + "epoch": 0.20651708701402358, + "grad_norm": 0.16256314516067505, + "learning_rate": 2.9378608476489923e-05, + "loss": 3.1199, + "step": 16195 + }, + { + "epoch": 0.206547933479597, + "grad_norm": 0.15219692885875702, + "learning_rate": 2.9375958128650566e-05, + "loss": 3.0934, + "step": 16196 + }, + { + "epoch": 0.2065787799451704, + "grad_norm": 0.16080474853515625, + "learning_rate": 2.9373307730074605e-05, + "loss": 3.1147, + "step": 16197 + }, + { + "epoch": 0.20660962641074382, + "grad_norm": 0.17052984237670898, + "learning_rate": 2.937065728079278e-05, + "loss": 3.1184, + "step": 16198 + }, + { + "epoch": 0.20664047287631723, + "grad_norm": 0.15688125789165497, + "learning_rate": 2.9368006780835805e-05, + "loss": 3.1265, + "step": 16199 + }, + { + "epoch": 0.20667131934189065, + "grad_norm": 0.17671197652816772, + "learning_rate": 2.936535623023441e-05, + "loss": 3.1312, + "step": 16200 + }, + { + "epoch": 0.20670216580746406, + "grad_norm": 0.15888181328773499, + "learning_rate": 2.9362705629019338e-05, + "loss": 3.0852, + "step": 16201 + }, + { + "epoch": 0.2067330122730375, + "grad_norm": 0.1595323085784912, + "learning_rate": 2.936005497722132e-05, + "loss": 3.1296, + "step": 16202 + }, + { + "epoch": 0.20676385873861092, + "grad_norm": 0.16320498287677765, + "learning_rate": 2.935740427487107e-05, + "loss": 3.1078, + "step": 16203 + }, + { + "epoch": 0.20679470520418433, + "grad_norm": 0.1558026820421219, + "learning_rate": 2.935475352199935e-05, + "loss": 3.107, + "step": 16204 + }, + { + "epoch": 0.20682555166975775, + "grad_norm": 0.15320567786693573, + "learning_rate": 2.9352102718636883e-05, + "loss": 3.0835, + "step": 16205 + }, + { + "epoch": 0.20685639813533116, + "grad_norm": 0.16554968059062958, + "learning_rate": 2.9349451864814388e-05, + "loss": 3.1161, + "step": 16206 + }, + { + "epoch": 0.20688724460090457, + "grad_norm": 0.15751738846302032, + "learning_rate": 2.9346800960562615e-05, + "loss": 3.1114, + "step": 16207 + }, + { + "epoch": 0.206918091066478, + "grad_norm": 0.16031867265701294, + "learning_rate": 2.9344150005912296e-05, + "loss": 3.0994, + "step": 16208 + }, + { + "epoch": 0.2069489375320514, + "grad_norm": 0.16557176411151886, + "learning_rate": 2.9341499000894173e-05, + "loss": 3.1151, + "step": 16209 + }, + { + "epoch": 0.20697978399762482, + "grad_norm": 0.15388113260269165, + "learning_rate": 2.9338847945538982e-05, + "loss": 3.119, + "step": 16210 + }, + { + "epoch": 0.20701063046319823, + "grad_norm": 0.16234801709651947, + "learning_rate": 2.933619683987745e-05, + "loss": 3.1067, + "step": 16211 + }, + { + "epoch": 0.20704147692877164, + "grad_norm": 0.1543450653553009, + "learning_rate": 2.9333545683940324e-05, + "loss": 3.1122, + "step": 16212 + }, + { + "epoch": 0.20707232339434506, + "grad_norm": 0.15528717637062073, + "learning_rate": 2.9330894477758337e-05, + "loss": 3.0801, + "step": 16213 + }, + { + "epoch": 0.2071031698599185, + "grad_norm": 0.16361157596111298, + "learning_rate": 2.9328243221362228e-05, + "loss": 3.0982, + "step": 16214 + }, + { + "epoch": 0.20713401632549192, + "grad_norm": 0.1530025452375412, + "learning_rate": 2.9325591914782747e-05, + "loss": 3.1247, + "step": 16215 + }, + { + "epoch": 0.20716486279106533, + "grad_norm": 0.1492006480693817, + "learning_rate": 2.9322940558050628e-05, + "loss": 3.0991, + "step": 16216 + }, + { + "epoch": 0.20719570925663874, + "grad_norm": 0.1539086103439331, + "learning_rate": 2.9320289151196602e-05, + "loss": 3.1356, + "step": 16217 + }, + { + "epoch": 0.20722655572221216, + "grad_norm": 0.15499761700630188, + "learning_rate": 2.9317637694251422e-05, + "loss": 3.1031, + "step": 16218 + }, + { + "epoch": 0.20725740218778557, + "grad_norm": 0.15901589393615723, + "learning_rate": 2.9314986187245837e-05, + "loss": 3.114, + "step": 16219 + }, + { + "epoch": 0.207288248653359, + "grad_norm": 0.1594237983226776, + "learning_rate": 2.931233463021057e-05, + "loss": 3.0791, + "step": 16220 + }, + { + "epoch": 0.2073190951189324, + "grad_norm": 0.15260612964630127, + "learning_rate": 2.9309683023176382e-05, + "loss": 3.1208, + "step": 16221 + }, + { + "epoch": 0.20734994158450581, + "grad_norm": 0.15575987100601196, + "learning_rate": 2.9307031366174004e-05, + "loss": 3.0808, + "step": 16222 + }, + { + "epoch": 0.20738078805007923, + "grad_norm": 0.15624675154685974, + "learning_rate": 2.9304379659234182e-05, + "loss": 3.0876, + "step": 16223 + }, + { + "epoch": 0.20741163451565264, + "grad_norm": 0.15684440732002258, + "learning_rate": 2.930172790238767e-05, + "loss": 3.0993, + "step": 16224 + }, + { + "epoch": 0.20744248098122606, + "grad_norm": 0.16210801899433136, + "learning_rate": 2.9299076095665212e-05, + "loss": 3.1022, + "step": 16225 + }, + { + "epoch": 0.2074733274467995, + "grad_norm": 0.16288352012634277, + "learning_rate": 2.9296424239097542e-05, + "loss": 3.0965, + "step": 16226 + }, + { + "epoch": 0.2075041739123729, + "grad_norm": 0.15890894830226898, + "learning_rate": 2.929377233271542e-05, + "loss": 3.1019, + "step": 16227 + }, + { + "epoch": 0.20753502037794633, + "grad_norm": 0.15963312983512878, + "learning_rate": 2.9291120376549587e-05, + "loss": 3.1135, + "step": 16228 + }, + { + "epoch": 0.20756586684351974, + "grad_norm": 0.15985704958438873, + "learning_rate": 2.9288468370630784e-05, + "loss": 3.0992, + "step": 16229 + }, + { + "epoch": 0.20759671330909316, + "grad_norm": 0.16197824478149414, + "learning_rate": 2.9285816314989778e-05, + "loss": 3.1094, + "step": 16230 + }, + { + "epoch": 0.20762755977466657, + "grad_norm": 0.1535755693912506, + "learning_rate": 2.9283164209657303e-05, + "loss": 3.1105, + "step": 16231 + }, + { + "epoch": 0.20765840624023998, + "grad_norm": 0.16361470520496368, + "learning_rate": 2.928051205466411e-05, + "loss": 3.0742, + "step": 16232 + }, + { + "epoch": 0.2076892527058134, + "grad_norm": 0.15497344732284546, + "learning_rate": 2.9277859850040962e-05, + "loss": 3.0974, + "step": 16233 + }, + { + "epoch": 0.2077200991713868, + "grad_norm": 0.15265217423439026, + "learning_rate": 2.9275207595818587e-05, + "loss": 3.0933, + "step": 16234 + }, + { + "epoch": 0.20775094563696023, + "grad_norm": 0.15299862623214722, + "learning_rate": 2.9272555292027753e-05, + "loss": 3.1261, + "step": 16235 + }, + { + "epoch": 0.20778179210253364, + "grad_norm": 0.152146115899086, + "learning_rate": 2.9269902938699217e-05, + "loss": 3.1413, + "step": 16236 + }, + { + "epoch": 0.20781263856810706, + "grad_norm": 0.15482109785079956, + "learning_rate": 2.9267250535863715e-05, + "loss": 3.118, + "step": 16237 + }, + { + "epoch": 0.2078434850336805, + "grad_norm": 0.15176717936992645, + "learning_rate": 2.926459808355201e-05, + "loss": 3.1051, + "step": 16238 + }, + { + "epoch": 0.2078743314992539, + "grad_norm": 0.15790955722332, + "learning_rate": 2.9261945581794854e-05, + "loss": 3.1115, + "step": 16239 + }, + { + "epoch": 0.20790517796482733, + "grad_norm": 0.15392927825450897, + "learning_rate": 2.9259293030622993e-05, + "loss": 3.0918, + "step": 16240 + }, + { + "epoch": 0.20793602443040074, + "grad_norm": 0.164358451962471, + "learning_rate": 2.925664043006719e-05, + "loss": 3.1173, + "step": 16241 + }, + { + "epoch": 0.20796687089597415, + "grad_norm": 0.15807199478149414, + "learning_rate": 2.9253987780158204e-05, + "loss": 3.1278, + "step": 16242 + }, + { + "epoch": 0.20799771736154757, + "grad_norm": 0.17153306305408478, + "learning_rate": 2.9251335080926783e-05, + "loss": 3.1048, + "step": 16243 + }, + { + "epoch": 0.20802856382712098, + "grad_norm": 0.1508130133152008, + "learning_rate": 2.9248682332403688e-05, + "loss": 3.1386, + "step": 16244 + }, + { + "epoch": 0.2080594102926944, + "grad_norm": 0.16188892722129822, + "learning_rate": 2.9246029534619678e-05, + "loss": 3.1351, + "step": 16245 + }, + { + "epoch": 0.2080902567582678, + "grad_norm": 0.16581696271896362, + "learning_rate": 2.92433766876055e-05, + "loss": 3.1333, + "step": 16246 + }, + { + "epoch": 0.20812110322384123, + "grad_norm": 0.1627786010503769, + "learning_rate": 2.9240723791391918e-05, + "loss": 3.0853, + "step": 16247 + }, + { + "epoch": 0.20815194968941464, + "grad_norm": 0.1579022854566574, + "learning_rate": 2.9238070846009703e-05, + "loss": 3.075, + "step": 16248 + }, + { + "epoch": 0.20818279615498805, + "grad_norm": 0.15569444000720978, + "learning_rate": 2.92354178514896e-05, + "loss": 3.143, + "step": 16249 + }, + { + "epoch": 0.2082136426205615, + "grad_norm": 0.15408727526664734, + "learning_rate": 2.9232764807862363e-05, + "loss": 3.0789, + "step": 16250 + }, + { + "epoch": 0.2082444890861349, + "grad_norm": 0.156840518116951, + "learning_rate": 2.9230111715158764e-05, + "loss": 3.103, + "step": 16251 + }, + { + "epoch": 0.20827533555170832, + "grad_norm": 0.15254177153110504, + "learning_rate": 2.9227458573409565e-05, + "loss": 3.0966, + "step": 16252 + }, + { + "epoch": 0.20830618201728174, + "grad_norm": 0.1548490971326828, + "learning_rate": 2.9224805382645527e-05, + "loss": 3.0858, + "step": 16253 + }, + { + "epoch": 0.20833702848285515, + "grad_norm": 0.15524998307228088, + "learning_rate": 2.9222152142897407e-05, + "loss": 3.1363, + "step": 16254 + }, + { + "epoch": 0.20836787494842857, + "grad_norm": 0.1585993617773056, + "learning_rate": 2.9219498854195974e-05, + "loss": 3.0967, + "step": 16255 + }, + { + "epoch": 0.20839872141400198, + "grad_norm": 0.15647925436496735, + "learning_rate": 2.921684551657198e-05, + "loss": 3.0951, + "step": 16256 + }, + { + "epoch": 0.2084295678795754, + "grad_norm": 0.14796584844589233, + "learning_rate": 2.9214192130056194e-05, + "loss": 3.0911, + "step": 16257 + }, + { + "epoch": 0.2084604143451488, + "grad_norm": 0.1529460996389389, + "learning_rate": 2.921153869467939e-05, + "loss": 3.1064, + "step": 16258 + }, + { + "epoch": 0.20849126081072222, + "grad_norm": 0.15149983763694763, + "learning_rate": 2.9208885210472325e-05, + "loss": 3.073, + "step": 16259 + }, + { + "epoch": 0.20852210727629564, + "grad_norm": 0.15252451598644257, + "learning_rate": 2.9206231677465767e-05, + "loss": 3.0991, + "step": 16260 + }, + { + "epoch": 0.20855295374186905, + "grad_norm": 0.15320971608161926, + "learning_rate": 2.9203578095690477e-05, + "loss": 3.0802, + "step": 16261 + }, + { + "epoch": 0.2085838002074425, + "grad_norm": 0.163032665848732, + "learning_rate": 2.920092446517722e-05, + "loss": 3.1151, + "step": 16262 + }, + { + "epoch": 0.2086146466730159, + "grad_norm": 0.15063731372356415, + "learning_rate": 2.919827078595678e-05, + "loss": 3.1094, + "step": 16263 + }, + { + "epoch": 0.20864549313858932, + "grad_norm": 0.6318920254707336, + "learning_rate": 2.9195617058059906e-05, + "loss": 3.1019, + "step": 16264 + }, + { + "epoch": 0.20867633960416274, + "grad_norm": 0.1527027040719986, + "learning_rate": 2.919296328151738e-05, + "loss": 3.0763, + "step": 16265 + }, + { + "epoch": 0.20870718606973615, + "grad_norm": 0.15272651612758636, + "learning_rate": 2.9190309456359955e-05, + "loss": 3.1036, + "step": 16266 + }, + { + "epoch": 0.20873803253530956, + "grad_norm": 0.15518216788768768, + "learning_rate": 2.918765558261841e-05, + "loss": 3.136, + "step": 16267 + }, + { + "epoch": 0.20876887900088298, + "grad_norm": 0.15336601436138153, + "learning_rate": 2.918500166032352e-05, + "loss": 3.1278, + "step": 16268 + }, + { + "epoch": 0.2087997254664564, + "grad_norm": 0.15369245409965515, + "learning_rate": 2.918234768950605e-05, + "loss": 3.1021, + "step": 16269 + }, + { + "epoch": 0.2088305719320298, + "grad_norm": 0.16048848628997803, + "learning_rate": 2.9179693670196773e-05, + "loss": 3.0807, + "step": 16270 + }, + { + "epoch": 0.20886141839760322, + "grad_norm": 0.1647234559059143, + "learning_rate": 2.917703960242646e-05, + "loss": 3.1058, + "step": 16271 + }, + { + "epoch": 0.20889226486317664, + "grad_norm": 0.15877960622310638, + "learning_rate": 2.917438548622588e-05, + "loss": 3.1002, + "step": 16272 + }, + { + "epoch": 0.20892311132875005, + "grad_norm": 0.15489622950553894, + "learning_rate": 2.9171731321625805e-05, + "loss": 3.0808, + "step": 16273 + }, + { + "epoch": 0.2089539577943235, + "grad_norm": 0.1545211523771286, + "learning_rate": 2.9169077108657016e-05, + "loss": 3.1123, + "step": 16274 + }, + { + "epoch": 0.2089848042598969, + "grad_norm": 0.17973878979682922, + "learning_rate": 2.916642284735029e-05, + "loss": 3.1141, + "step": 16275 + }, + { + "epoch": 0.20901565072547032, + "grad_norm": 0.15689432621002197, + "learning_rate": 2.9163768537736386e-05, + "loss": 3.0983, + "step": 16276 + }, + { + "epoch": 0.20904649719104373, + "grad_norm": 0.6372639536857605, + "learning_rate": 2.9161114179846094e-05, + "loss": 3.0925, + "step": 16277 + }, + { + "epoch": 0.20907734365661715, + "grad_norm": 0.15452206134796143, + "learning_rate": 2.9158459773710184e-05, + "loss": 3.0806, + "step": 16278 + }, + { + "epoch": 0.20910819012219056, + "grad_norm": 0.165225088596344, + "learning_rate": 2.915580531935942e-05, + "loss": 3.1019, + "step": 16279 + }, + { + "epoch": 0.20913903658776398, + "grad_norm": 0.18006563186645508, + "learning_rate": 2.9153150816824605e-05, + "loss": 3.0889, + "step": 16280 + }, + { + "epoch": 0.2091698830533374, + "grad_norm": 0.1657118946313858, + "learning_rate": 2.9150496266136496e-05, + "loss": 3.1246, + "step": 16281 + }, + { + "epoch": 0.2092007295189108, + "grad_norm": 0.15960435569286346, + "learning_rate": 2.9147841667325876e-05, + "loss": 3.1038, + "step": 16282 + }, + { + "epoch": 0.20923157598448422, + "grad_norm": 0.16683097183704376, + "learning_rate": 2.9145187020423524e-05, + "loss": 3.1312, + "step": 16283 + }, + { + "epoch": 0.20926242245005763, + "grad_norm": 0.15234145522117615, + "learning_rate": 2.914253232546022e-05, + "loss": 3.1069, + "step": 16284 + }, + { + "epoch": 0.20929326891563105, + "grad_norm": 0.15503308176994324, + "learning_rate": 2.9139877582466742e-05, + "loss": 3.1298, + "step": 16285 + }, + { + "epoch": 0.2093241153812045, + "grad_norm": 0.16077733039855957, + "learning_rate": 2.9137222791473874e-05, + "loss": 3.0967, + "step": 16286 + }, + { + "epoch": 0.2093549618467779, + "grad_norm": 0.15832506120204926, + "learning_rate": 2.91345679525124e-05, + "loss": 3.0827, + "step": 16287 + }, + { + "epoch": 0.20938580831235132, + "grad_norm": 0.1600627601146698, + "learning_rate": 2.913191306561308e-05, + "loss": 3.0931, + "step": 16288 + }, + { + "epoch": 0.20941665477792473, + "grad_norm": 0.6631948351860046, + "learning_rate": 2.9129258130806726e-05, + "loss": 3.0836, + "step": 16289 + }, + { + "epoch": 0.20944750124349815, + "grad_norm": 0.15538138151168823, + "learning_rate": 2.91266031481241e-05, + "loss": 3.1046, + "step": 16290 + }, + { + "epoch": 0.20947834770907156, + "grad_norm": 0.15863382816314697, + "learning_rate": 2.9123948117595985e-05, + "loss": 3.0969, + "step": 16291 + }, + { + "epoch": 0.20950919417464497, + "grad_norm": 0.16394954919815063, + "learning_rate": 2.9121293039253173e-05, + "loss": 3.0929, + "step": 16292 + }, + { + "epoch": 0.2095400406402184, + "grad_norm": 0.16012230515480042, + "learning_rate": 2.911863791312645e-05, + "loss": 3.0957, + "step": 16293 + }, + { + "epoch": 0.2095708871057918, + "grad_norm": 0.15922164916992188, + "learning_rate": 2.911598273924659e-05, + "loss": 3.0965, + "step": 16294 + }, + { + "epoch": 0.20960173357136522, + "grad_norm": 0.15748219192028046, + "learning_rate": 2.9113327517644384e-05, + "loss": 3.105, + "step": 16295 + }, + { + "epoch": 0.20963258003693863, + "grad_norm": 0.15715819597244263, + "learning_rate": 2.9110672248350618e-05, + "loss": 3.0868, + "step": 16296 + }, + { + "epoch": 0.20966342650251205, + "grad_norm": 0.15806251764297485, + "learning_rate": 2.910801693139607e-05, + "loss": 3.083, + "step": 16297 + }, + { + "epoch": 0.2096942729680855, + "grad_norm": 0.1510976254940033, + "learning_rate": 2.9105361566811545e-05, + "loss": 3.157, + "step": 16298 + }, + { + "epoch": 0.2097251194336589, + "grad_norm": 0.15677779912948608, + "learning_rate": 2.9102706154627824e-05, + "loss": 3.1241, + "step": 16299 + }, + { + "epoch": 0.20975596589923232, + "grad_norm": 0.15422949194908142, + "learning_rate": 2.910005069487568e-05, + "loss": 3.1112, + "step": 16300 + }, + { + "epoch": 0.20978681236480573, + "grad_norm": 0.15281222760677338, + "learning_rate": 2.909739518758591e-05, + "loss": 3.1099, + "step": 16301 + }, + { + "epoch": 0.20981765883037914, + "grad_norm": 0.16823089122772217, + "learning_rate": 2.9094739632789314e-05, + "loss": 3.0656, + "step": 16302 + }, + { + "epoch": 0.20984850529595256, + "grad_norm": 0.1530686467885971, + "learning_rate": 2.9092084030516664e-05, + "loss": 3.1013, + "step": 16303 + }, + { + "epoch": 0.20987935176152597, + "grad_norm": 0.1516103893518448, + "learning_rate": 2.9089428380798765e-05, + "loss": 3.1123, + "step": 16304 + }, + { + "epoch": 0.2099101982270994, + "grad_norm": 0.15485015511512756, + "learning_rate": 2.90867726836664e-05, + "loss": 3.1143, + "step": 16305 + }, + { + "epoch": 0.2099410446926728, + "grad_norm": 0.15392208099365234, + "learning_rate": 2.9084116939150364e-05, + "loss": 3.0925, + "step": 16306 + }, + { + "epoch": 0.20997189115824622, + "grad_norm": 0.15029315650463104, + "learning_rate": 2.908146114728144e-05, + "loss": 3.1194, + "step": 16307 + }, + { + "epoch": 0.21000273762381963, + "grad_norm": 0.15649691224098206, + "learning_rate": 2.907880530809043e-05, + "loss": 3.0752, + "step": 16308 + }, + { + "epoch": 0.21003358408939304, + "grad_norm": 0.15527872741222382, + "learning_rate": 2.9076149421608124e-05, + "loss": 3.1097, + "step": 16309 + }, + { + "epoch": 0.21006443055496649, + "grad_norm": 0.2678958773612976, + "learning_rate": 2.9073493487865315e-05, + "loss": 3.114, + "step": 16310 + }, + { + "epoch": 0.2100952770205399, + "grad_norm": 0.15444403886795044, + "learning_rate": 2.9070837506892794e-05, + "loss": 3.1048, + "step": 16311 + }, + { + "epoch": 0.21012612348611331, + "grad_norm": 0.157699853181839, + "learning_rate": 2.9068181478721363e-05, + "loss": 3.0679, + "step": 16312 + }, + { + "epoch": 0.21015696995168673, + "grad_norm": 0.1611166149377823, + "learning_rate": 2.906552540338181e-05, + "loss": 3.1483, + "step": 16313 + }, + { + "epoch": 0.21018781641726014, + "grad_norm": 0.1522762030363083, + "learning_rate": 2.9062869280904932e-05, + "loss": 3.109, + "step": 16314 + }, + { + "epoch": 0.21021866288283356, + "grad_norm": 0.2274654507637024, + "learning_rate": 2.906021311132153e-05, + "loss": 3.1119, + "step": 16315 + }, + { + "epoch": 0.21024950934840697, + "grad_norm": 0.15292151272296906, + "learning_rate": 2.9057556894662397e-05, + "loss": 3.1255, + "step": 16316 + }, + { + "epoch": 0.21028035581398039, + "grad_norm": 0.1524301916360855, + "learning_rate": 2.9054900630958327e-05, + "loss": 3.1047, + "step": 16317 + }, + { + "epoch": 0.2103112022795538, + "grad_norm": 0.15517164766788483, + "learning_rate": 2.9052244320240118e-05, + "loss": 3.1062, + "step": 16318 + }, + { + "epoch": 0.2103420487451272, + "grad_norm": 0.16051746904850006, + "learning_rate": 2.9049587962538576e-05, + "loss": 3.1047, + "step": 16319 + }, + { + "epoch": 0.21037289521070063, + "grad_norm": 0.1585921198129654, + "learning_rate": 2.9046931557884494e-05, + "loss": 3.1074, + "step": 16320 + }, + { + "epoch": 0.21040374167627404, + "grad_norm": 0.16058722138404846, + "learning_rate": 2.9044275106308677e-05, + "loss": 3.1074, + "step": 16321 + }, + { + "epoch": 0.21043458814184748, + "grad_norm": 0.1560528427362442, + "learning_rate": 2.9041618607841918e-05, + "loss": 3.1054, + "step": 16322 + }, + { + "epoch": 0.2104654346074209, + "grad_norm": 0.16439111530780792, + "learning_rate": 2.9038962062515013e-05, + "loss": 3.1301, + "step": 16323 + }, + { + "epoch": 0.2104962810729943, + "grad_norm": 0.1720680147409439, + "learning_rate": 2.903630547035877e-05, + "loss": 3.112, + "step": 16324 + }, + { + "epoch": 0.21052712753856773, + "grad_norm": 0.16075938940048218, + "learning_rate": 2.9033648831404004e-05, + "loss": 3.1427, + "step": 16325 + }, + { + "epoch": 0.21055797400414114, + "grad_norm": 0.18220357596874237, + "learning_rate": 2.903099214568149e-05, + "loss": 3.0902, + "step": 16326 + }, + { + "epoch": 0.21058882046971455, + "grad_norm": 0.1576649397611618, + "learning_rate": 2.9028335413222052e-05, + "loss": 3.1159, + "step": 16327 + }, + { + "epoch": 0.21061966693528797, + "grad_norm": 0.15329144895076752, + "learning_rate": 2.9025678634056486e-05, + "loss": 3.1288, + "step": 16328 + }, + { + "epoch": 0.21065051340086138, + "grad_norm": 0.16277526319026947, + "learning_rate": 2.902302180821559e-05, + "loss": 3.1129, + "step": 16329 + }, + { + "epoch": 0.2106813598664348, + "grad_norm": 0.1538301706314087, + "learning_rate": 2.9020364935730178e-05, + "loss": 3.1209, + "step": 16330 + }, + { + "epoch": 0.2107122063320082, + "grad_norm": 0.15728653967380524, + "learning_rate": 2.9017708016631052e-05, + "loss": 3.1097, + "step": 16331 + }, + { + "epoch": 0.21074305279758163, + "grad_norm": 0.1642770618200302, + "learning_rate": 2.9015051050949003e-05, + "loss": 3.1477, + "step": 16332 + }, + { + "epoch": 0.21077389926315504, + "grad_norm": 0.1506776362657547, + "learning_rate": 2.9012394038714868e-05, + "loss": 3.1236, + "step": 16333 + }, + { + "epoch": 0.21080474572872848, + "grad_norm": 0.1663881242275238, + "learning_rate": 2.9009736979959423e-05, + "loss": 3.1005, + "step": 16334 + }, + { + "epoch": 0.2108355921943019, + "grad_norm": 0.14974406361579895, + "learning_rate": 2.9007079874713487e-05, + "loss": 3.0855, + "step": 16335 + }, + { + "epoch": 0.2108664386598753, + "grad_norm": 0.14921541512012482, + "learning_rate": 2.900442272300788e-05, + "loss": 3.0828, + "step": 16336 + }, + { + "epoch": 0.21089728512544872, + "grad_norm": 0.17251314222812653, + "learning_rate": 2.9001765524873393e-05, + "loss": 3.1332, + "step": 16337 + }, + { + "epoch": 0.21092813159102214, + "grad_norm": 0.15643401443958282, + "learning_rate": 2.899910828034083e-05, + "loss": 3.0858, + "step": 16338 + }, + { + "epoch": 0.21095897805659555, + "grad_norm": 0.14898884296417236, + "learning_rate": 2.899645098944102e-05, + "loss": 3.1326, + "step": 16339 + }, + { + "epoch": 0.21098982452216897, + "grad_norm": 0.2102389931678772, + "learning_rate": 2.8993793652204755e-05, + "loss": 3.1583, + "step": 16340 + }, + { + "epoch": 0.21102067098774238, + "grad_norm": 0.15409818291664124, + "learning_rate": 2.8991136268662855e-05, + "loss": 3.1088, + "step": 16341 + }, + { + "epoch": 0.2110515174533158, + "grad_norm": 0.27129513025283813, + "learning_rate": 2.8988478838846135e-05, + "loss": 3.0819, + "step": 16342 + }, + { + "epoch": 0.2110823639188892, + "grad_norm": 0.15289771556854248, + "learning_rate": 2.8985821362785394e-05, + "loss": 3.1022, + "step": 16343 + }, + { + "epoch": 0.21111321038446262, + "grad_norm": 0.1570659875869751, + "learning_rate": 2.8983163840511447e-05, + "loss": 3.1038, + "step": 16344 + }, + { + "epoch": 0.21114405685003604, + "grad_norm": 0.17380332946777344, + "learning_rate": 2.8980506272055112e-05, + "loss": 3.1431, + "step": 16345 + }, + { + "epoch": 0.21117490331560948, + "grad_norm": 0.16785776615142822, + "learning_rate": 2.8977848657447197e-05, + "loss": 3.1021, + "step": 16346 + }, + { + "epoch": 0.2112057497811829, + "grad_norm": 0.16091427206993103, + "learning_rate": 2.8975190996718515e-05, + "loss": 3.1227, + "step": 16347 + }, + { + "epoch": 0.2112365962467563, + "grad_norm": 0.15445677936077118, + "learning_rate": 2.897253328989989e-05, + "loss": 3.0831, + "step": 16348 + }, + { + "epoch": 0.21126744271232972, + "grad_norm": 0.16426847875118256, + "learning_rate": 2.8969875537022128e-05, + "loss": 3.1099, + "step": 16349 + }, + { + "epoch": 0.21129828917790314, + "grad_norm": 0.1604573130607605, + "learning_rate": 2.8967217738116036e-05, + "loss": 3.1489, + "step": 16350 + }, + { + "epoch": 0.21132913564347655, + "grad_norm": 0.16174574196338654, + "learning_rate": 2.8964559893212447e-05, + "loss": 3.0835, + "step": 16351 + }, + { + "epoch": 0.21135998210904997, + "grad_norm": 0.15192902088165283, + "learning_rate": 2.896190200234216e-05, + "loss": 3.124, + "step": 16352 + }, + { + "epoch": 0.21139082857462338, + "grad_norm": 0.1547171026468277, + "learning_rate": 2.8959244065536006e-05, + "loss": 3.0992, + "step": 16353 + }, + { + "epoch": 0.2114216750401968, + "grad_norm": 0.15845313668251038, + "learning_rate": 2.89565860828248e-05, + "loss": 3.0767, + "step": 16354 + }, + { + "epoch": 0.2114525215057702, + "grad_norm": 0.16321811079978943, + "learning_rate": 2.8953928054239356e-05, + "loss": 3.1138, + "step": 16355 + }, + { + "epoch": 0.21148336797134362, + "grad_norm": 0.1566973626613617, + "learning_rate": 2.895126997981048e-05, + "loss": 3.1131, + "step": 16356 + }, + { + "epoch": 0.21151421443691704, + "grad_norm": 0.1535649448633194, + "learning_rate": 2.8948611859569007e-05, + "loss": 3.1354, + "step": 16357 + }, + { + "epoch": 0.21154506090249048, + "grad_norm": 0.17251858115196228, + "learning_rate": 2.8945953693545758e-05, + "loss": 3.0833, + "step": 16358 + }, + { + "epoch": 0.2115759073680639, + "grad_norm": 0.15615767240524292, + "learning_rate": 2.8943295481771548e-05, + "loss": 3.1093, + "step": 16359 + }, + { + "epoch": 0.2116067538336373, + "grad_norm": 0.14979392290115356, + "learning_rate": 2.8940637224277194e-05, + "loss": 3.0802, + "step": 16360 + }, + { + "epoch": 0.21163760029921072, + "grad_norm": 0.1567467600107193, + "learning_rate": 2.8937978921093517e-05, + "loss": 3.1018, + "step": 16361 + }, + { + "epoch": 0.21166844676478414, + "grad_norm": 0.15375050902366638, + "learning_rate": 2.893532057225134e-05, + "loss": 3.0908, + "step": 16362 + }, + { + "epoch": 0.21169929323035755, + "grad_norm": 0.16260214149951935, + "learning_rate": 2.8932662177781493e-05, + "loss": 3.0952, + "step": 16363 + }, + { + "epoch": 0.21173013969593096, + "grad_norm": 0.1591493785381317, + "learning_rate": 2.8930003737714784e-05, + "loss": 3.0893, + "step": 16364 + }, + { + "epoch": 0.21176098616150438, + "grad_norm": 0.1512654572725296, + "learning_rate": 2.892734525208205e-05, + "loss": 3.0685, + "step": 16365 + }, + { + "epoch": 0.2117918326270778, + "grad_norm": 0.15874549746513367, + "learning_rate": 2.8924686720914108e-05, + "loss": 3.1104, + "step": 16366 + }, + { + "epoch": 0.2118226790926512, + "grad_norm": 0.1571781188249588, + "learning_rate": 2.8922028144241774e-05, + "loss": 3.0653, + "step": 16367 + }, + { + "epoch": 0.21185352555822462, + "grad_norm": 0.15494537353515625, + "learning_rate": 2.8919369522095884e-05, + "loss": 3.0997, + "step": 16368 + }, + { + "epoch": 0.21188437202379803, + "grad_norm": 0.15067562460899353, + "learning_rate": 2.8916710854507262e-05, + "loss": 3.0753, + "step": 16369 + }, + { + "epoch": 0.21191521848937148, + "grad_norm": 0.15704375505447388, + "learning_rate": 2.8914052141506727e-05, + "loss": 3.0838, + "step": 16370 + }, + { + "epoch": 0.2119460649549449, + "grad_norm": 0.15451188385486603, + "learning_rate": 2.891139338312511e-05, + "loss": 3.1061, + "step": 16371 + }, + { + "epoch": 0.2119769114205183, + "grad_norm": 0.15607254207134247, + "learning_rate": 2.890873457939325e-05, + "loss": 3.1195, + "step": 16372 + }, + { + "epoch": 0.21200775788609172, + "grad_norm": 0.15343062579631805, + "learning_rate": 2.8906075730341947e-05, + "loss": 3.0914, + "step": 16373 + }, + { + "epoch": 0.21203860435166513, + "grad_norm": 0.15195965766906738, + "learning_rate": 2.8903416836002046e-05, + "loss": 3.0695, + "step": 16374 + }, + { + "epoch": 0.21206945081723855, + "grad_norm": 0.16823436319828033, + "learning_rate": 2.8900757896404372e-05, + "loss": 3.0927, + "step": 16375 + }, + { + "epoch": 0.21210029728281196, + "grad_norm": 0.15869386494159698, + "learning_rate": 2.8898098911579753e-05, + "loss": 3.1126, + "step": 16376 + }, + { + "epoch": 0.21213114374838538, + "grad_norm": 0.15309792757034302, + "learning_rate": 2.8895439881559028e-05, + "loss": 3.1388, + "step": 16377 + }, + { + "epoch": 0.2121619902139588, + "grad_norm": 0.1656261831521988, + "learning_rate": 2.8892780806373015e-05, + "loss": 3.1053, + "step": 16378 + }, + { + "epoch": 0.2121928366795322, + "grad_norm": 0.15858899056911469, + "learning_rate": 2.889012168605254e-05, + "loss": 3.1174, + "step": 16379 + }, + { + "epoch": 0.21222368314510562, + "grad_norm": 0.16877800226211548, + "learning_rate": 2.888746252062845e-05, + "loss": 3.1177, + "step": 16380 + }, + { + "epoch": 0.21225452961067906, + "grad_norm": 0.16327369213104248, + "learning_rate": 2.8884803310131562e-05, + "loss": 3.0743, + "step": 16381 + }, + { + "epoch": 0.21228537607625247, + "grad_norm": 0.15609124302864075, + "learning_rate": 2.888214405459272e-05, + "loss": 3.0532, + "step": 16382 + }, + { + "epoch": 0.2123162225418259, + "grad_norm": 0.15761396288871765, + "learning_rate": 2.8879484754042753e-05, + "loss": 3.0772, + "step": 16383 + }, + { + "epoch": 0.2123470690073993, + "grad_norm": 0.1618070751428604, + "learning_rate": 2.887682540851248e-05, + "loss": 3.0882, + "step": 16384 + }, + { + "epoch": 0.21237791547297272, + "grad_norm": 0.15648387372493744, + "learning_rate": 2.887416601803275e-05, + "loss": 3.1021, + "step": 16385 + }, + { + "epoch": 0.21240876193854613, + "grad_norm": 0.1519419550895691, + "learning_rate": 2.88715065826344e-05, + "loss": 3.0893, + "step": 16386 + }, + { + "epoch": 0.21243960840411955, + "grad_norm": 0.15257969498634338, + "learning_rate": 2.8868847102348257e-05, + "loss": 3.1171, + "step": 16387 + }, + { + "epoch": 0.21247045486969296, + "grad_norm": 0.1563636064529419, + "learning_rate": 2.886618757720515e-05, + "loss": 3.1401, + "step": 16388 + }, + { + "epoch": 0.21250130133526637, + "grad_norm": 0.151624396443367, + "learning_rate": 2.886352800723593e-05, + "loss": 3.0873, + "step": 16389 + }, + { + "epoch": 0.2125321478008398, + "grad_norm": 0.1663355678319931, + "learning_rate": 2.8860868392471418e-05, + "loss": 3.065, + "step": 16390 + }, + { + "epoch": 0.2125629942664132, + "grad_norm": 0.16735981404781342, + "learning_rate": 2.8858208732942454e-05, + "loss": 3.1269, + "step": 16391 + }, + { + "epoch": 0.21259384073198662, + "grad_norm": 0.1519905924797058, + "learning_rate": 2.8855549028679884e-05, + "loss": 3.0821, + "step": 16392 + }, + { + "epoch": 0.21262468719756006, + "grad_norm": 0.15652169287204742, + "learning_rate": 2.8852889279714544e-05, + "loss": 3.1148, + "step": 16393 + }, + { + "epoch": 0.21265553366313347, + "grad_norm": 0.14761099219322205, + "learning_rate": 2.8850229486077256e-05, + "loss": 3.1299, + "step": 16394 + }, + { + "epoch": 0.2126863801287069, + "grad_norm": 0.14949525892734528, + "learning_rate": 2.884756964779888e-05, + "loss": 3.102, + "step": 16395 + }, + { + "epoch": 0.2127172265942803, + "grad_norm": 0.14844675362110138, + "learning_rate": 2.8844909764910244e-05, + "loss": 3.0757, + "step": 16396 + }, + { + "epoch": 0.21274807305985372, + "grad_norm": 0.15538401901721954, + "learning_rate": 2.8842249837442187e-05, + "loss": 3.1068, + "step": 16397 + }, + { + "epoch": 0.21277891952542713, + "grad_norm": 0.1532151997089386, + "learning_rate": 2.883958986542556e-05, + "loss": 3.1281, + "step": 16398 + }, + { + "epoch": 0.21280976599100054, + "grad_norm": 0.15126967430114746, + "learning_rate": 2.8836929848891193e-05, + "loss": 3.1284, + "step": 16399 + }, + { + "epoch": 0.21284061245657396, + "grad_norm": 0.14796669781208038, + "learning_rate": 2.8834269787869923e-05, + "loss": 3.071, + "step": 16400 + }, + { + "epoch": 0.21287145892214737, + "grad_norm": 0.15009759366512299, + "learning_rate": 2.88316096823926e-05, + "loss": 3.0962, + "step": 16401 + }, + { + "epoch": 0.21290230538772079, + "grad_norm": 0.15060743689537048, + "learning_rate": 2.8828949532490073e-05, + "loss": 3.1068, + "step": 16402 + }, + { + "epoch": 0.2129331518532942, + "grad_norm": 0.17924457788467407, + "learning_rate": 2.882628933819317e-05, + "loss": 3.0854, + "step": 16403 + }, + { + "epoch": 0.21296399831886761, + "grad_norm": 0.15321657061576843, + "learning_rate": 2.8823629099532747e-05, + "loss": 3.1127, + "step": 16404 + }, + { + "epoch": 0.21299484478444106, + "grad_norm": 0.15162523090839386, + "learning_rate": 2.882096881653964e-05, + "loss": 3.1457, + "step": 16405 + }, + { + "epoch": 0.21302569125001447, + "grad_norm": 0.15081240236759186, + "learning_rate": 2.8818308489244696e-05, + "loss": 3.0912, + "step": 16406 + }, + { + "epoch": 0.21305653771558788, + "grad_norm": 0.15018825232982635, + "learning_rate": 2.8815648117678755e-05, + "loss": 3.1198, + "step": 16407 + }, + { + "epoch": 0.2130873841811613, + "grad_norm": 0.14876513183116913, + "learning_rate": 2.8812987701872667e-05, + "loss": 3.0667, + "step": 16408 + }, + { + "epoch": 0.2131182306467347, + "grad_norm": 0.1548740118741989, + "learning_rate": 2.8810327241857283e-05, + "loss": 3.0969, + "step": 16409 + }, + { + "epoch": 0.21314907711230813, + "grad_norm": 0.26255688071250916, + "learning_rate": 2.8807666737663446e-05, + "loss": 3.1169, + "step": 16410 + }, + { + "epoch": 0.21317992357788154, + "grad_norm": 0.15408742427825928, + "learning_rate": 2.8805006189321997e-05, + "loss": 3.1126, + "step": 16411 + }, + { + "epoch": 0.21321077004345496, + "grad_norm": 0.16123759746551514, + "learning_rate": 2.880234559686379e-05, + "loss": 3.1199, + "step": 16412 + }, + { + "epoch": 0.21324161650902837, + "grad_norm": 0.14964653551578522, + "learning_rate": 2.879968496031967e-05, + "loss": 3.0947, + "step": 16413 + }, + { + "epoch": 0.21327246297460178, + "grad_norm": 0.1656094342470169, + "learning_rate": 2.8797024279720486e-05, + "loss": 3.0697, + "step": 16414 + }, + { + "epoch": 0.2133033094401752, + "grad_norm": 0.15846200287342072, + "learning_rate": 2.8794363555097088e-05, + "loss": 3.087, + "step": 16415 + }, + { + "epoch": 0.2133341559057486, + "grad_norm": 0.154975026845932, + "learning_rate": 2.879170278648033e-05, + "loss": 3.0782, + "step": 16416 + }, + { + "epoch": 0.21336500237132205, + "grad_norm": 0.16405576467514038, + "learning_rate": 2.8789041973901045e-05, + "loss": 3.1151, + "step": 16417 + }, + { + "epoch": 0.21339584883689547, + "grad_norm": 0.14954660832881927, + "learning_rate": 2.87863811173901e-05, + "loss": 3.0971, + "step": 16418 + }, + { + "epoch": 0.21342669530246888, + "grad_norm": 0.15508049726486206, + "learning_rate": 2.8783720216978343e-05, + "loss": 3.1155, + "step": 16419 + }, + { + "epoch": 0.2134575417680423, + "grad_norm": 0.16059966385364532, + "learning_rate": 2.878105927269662e-05, + "loss": 3.0947, + "step": 16420 + }, + { + "epoch": 0.2134883882336157, + "grad_norm": 0.1616126298904419, + "learning_rate": 2.8778398284575796e-05, + "loss": 3.081, + "step": 16421 + }, + { + "epoch": 0.21351923469918913, + "grad_norm": 0.16004683077335358, + "learning_rate": 2.877573725264671e-05, + "loss": 3.1056, + "step": 16422 + }, + { + "epoch": 0.21355008116476254, + "grad_norm": 0.15385499596595764, + "learning_rate": 2.8773076176940217e-05, + "loss": 3.093, + "step": 16423 + }, + { + "epoch": 0.21358092763033595, + "grad_norm": 0.1576298177242279, + "learning_rate": 2.877041505748717e-05, + "loss": 3.1152, + "step": 16424 + }, + { + "epoch": 0.21361177409590937, + "grad_norm": 0.15236113965511322, + "learning_rate": 2.8767753894318433e-05, + "loss": 3.1251, + "step": 16425 + }, + { + "epoch": 0.21364262056148278, + "grad_norm": 0.16265065968036652, + "learning_rate": 2.876509268746485e-05, + "loss": 3.1103, + "step": 16426 + }, + { + "epoch": 0.2136734670270562, + "grad_norm": 0.15248709917068481, + "learning_rate": 2.8762431436957283e-05, + "loss": 3.0812, + "step": 16427 + }, + { + "epoch": 0.2137043134926296, + "grad_norm": 0.15874899923801422, + "learning_rate": 2.8759770142826587e-05, + "loss": 3.1028, + "step": 16428 + }, + { + "epoch": 0.21373515995820305, + "grad_norm": 0.16302385926246643, + "learning_rate": 2.8757108805103605e-05, + "loss": 3.1159, + "step": 16429 + }, + { + "epoch": 0.21376600642377647, + "grad_norm": 0.16589391231536865, + "learning_rate": 2.875444742381921e-05, + "loss": 3.133, + "step": 16430 + }, + { + "epoch": 0.21379685288934988, + "grad_norm": 0.16954068839550018, + "learning_rate": 2.8751785999004255e-05, + "loss": 3.1392, + "step": 16431 + }, + { + "epoch": 0.2138276993549233, + "grad_norm": 0.1518782377243042, + "learning_rate": 2.8749124530689592e-05, + "loss": 3.0945, + "step": 16432 + }, + { + "epoch": 0.2138585458204967, + "grad_norm": 0.16990156471729279, + "learning_rate": 2.874646301890609e-05, + "loss": 3.0752, + "step": 16433 + }, + { + "epoch": 0.21388939228607012, + "grad_norm": 0.1570417881011963, + "learning_rate": 2.8743801463684596e-05, + "loss": 3.0852, + "step": 16434 + }, + { + "epoch": 0.21392023875164354, + "grad_norm": 0.15905287861824036, + "learning_rate": 2.8741139865055967e-05, + "loss": 3.1247, + "step": 16435 + }, + { + "epoch": 0.21395108521721695, + "grad_norm": 0.15752637386322021, + "learning_rate": 2.873847822305108e-05, + "loss": 3.1229, + "step": 16436 + }, + { + "epoch": 0.21398193168279037, + "grad_norm": 0.15276449918746948, + "learning_rate": 2.873581653770078e-05, + "loss": 3.1152, + "step": 16437 + }, + { + "epoch": 0.21401277814836378, + "grad_norm": 0.1827034056186676, + "learning_rate": 2.8733154809035932e-05, + "loss": 3.0985, + "step": 16438 + }, + { + "epoch": 0.2140436246139372, + "grad_norm": 0.16931742429733276, + "learning_rate": 2.87304930370874e-05, + "loss": 3.1047, + "step": 16439 + }, + { + "epoch": 0.2140744710795106, + "grad_norm": 0.1740792691707611, + "learning_rate": 2.872783122188604e-05, + "loss": 3.1307, + "step": 16440 + }, + { + "epoch": 0.21410531754508405, + "grad_norm": 0.15570756793022156, + "learning_rate": 2.8725169363462718e-05, + "loss": 3.1136, + "step": 16441 + }, + { + "epoch": 0.21413616401065746, + "grad_norm": 0.19060492515563965, + "learning_rate": 2.8722507461848303e-05, + "loss": 3.0911, + "step": 16442 + }, + { + "epoch": 0.21416701047623088, + "grad_norm": 0.1536410003900528, + "learning_rate": 2.8719845517073647e-05, + "loss": 3.1095, + "step": 16443 + }, + { + "epoch": 0.2141978569418043, + "grad_norm": 0.15461067855358124, + "learning_rate": 2.871718352916961e-05, + "loss": 3.0959, + "step": 16444 + }, + { + "epoch": 0.2142287034073777, + "grad_norm": 0.15962375700473785, + "learning_rate": 2.8714521498167068e-05, + "loss": 3.1042, + "step": 16445 + }, + { + "epoch": 0.21425954987295112, + "grad_norm": 0.14973077178001404, + "learning_rate": 2.871185942409688e-05, + "loss": 3.0784, + "step": 16446 + }, + { + "epoch": 0.21429039633852454, + "grad_norm": 0.15239942073822021, + "learning_rate": 2.8709197306989915e-05, + "loss": 3.0932, + "step": 16447 + }, + { + "epoch": 0.21432124280409795, + "grad_norm": 0.1593387871980667, + "learning_rate": 2.870653514687704e-05, + "loss": 3.1372, + "step": 16448 + }, + { + "epoch": 0.21435208926967136, + "grad_norm": 0.15497435629367828, + "learning_rate": 2.8703872943789112e-05, + "loss": 3.1021, + "step": 16449 + }, + { + "epoch": 0.21438293573524478, + "grad_norm": 0.16118589043617249, + "learning_rate": 2.8701210697757003e-05, + "loss": 3.1123, + "step": 16450 + }, + { + "epoch": 0.2144137822008182, + "grad_norm": 0.16959460079669952, + "learning_rate": 2.8698548408811576e-05, + "loss": 3.093, + "step": 16451 + }, + { + "epoch": 0.2144446286663916, + "grad_norm": 0.14935968816280365, + "learning_rate": 2.8695886076983713e-05, + "loss": 3.1136, + "step": 16452 + }, + { + "epoch": 0.21447547513196505, + "grad_norm": 0.16622529923915863, + "learning_rate": 2.8693223702304268e-05, + "loss": 3.1115, + "step": 16453 + }, + { + "epoch": 0.21450632159753846, + "grad_norm": 0.1595056802034378, + "learning_rate": 2.869056128480411e-05, + "loss": 3.1017, + "step": 16454 + }, + { + "epoch": 0.21453716806311188, + "grad_norm": 0.15307600796222687, + "learning_rate": 2.868789882451412e-05, + "loss": 3.1229, + "step": 16455 + }, + { + "epoch": 0.2145680145286853, + "grad_norm": 0.15330113470554352, + "learning_rate": 2.8685236321465147e-05, + "loss": 3.1024, + "step": 16456 + }, + { + "epoch": 0.2145988609942587, + "grad_norm": 0.15201634168624878, + "learning_rate": 2.8682573775688082e-05, + "loss": 3.1093, + "step": 16457 + }, + { + "epoch": 0.21462970745983212, + "grad_norm": 0.14880773425102234, + "learning_rate": 2.8679911187213783e-05, + "loss": 3.0653, + "step": 16458 + }, + { + "epoch": 0.21466055392540553, + "grad_norm": 0.1522533893585205, + "learning_rate": 2.8677248556073127e-05, + "loss": 3.1169, + "step": 16459 + }, + { + "epoch": 0.21469140039097895, + "grad_norm": 0.15737585723400116, + "learning_rate": 2.867458588229699e-05, + "loss": 3.1286, + "step": 16460 + }, + { + "epoch": 0.21472224685655236, + "grad_norm": 0.15371470153331757, + "learning_rate": 2.8671923165916226e-05, + "loss": 3.0795, + "step": 16461 + }, + { + "epoch": 0.21475309332212578, + "grad_norm": 0.15359269082546234, + "learning_rate": 2.8669260406961724e-05, + "loss": 3.1176, + "step": 16462 + }, + { + "epoch": 0.2147839397876992, + "grad_norm": 0.15446990728378296, + "learning_rate": 2.8666597605464358e-05, + "loss": 3.1277, + "step": 16463 + }, + { + "epoch": 0.2148147862532726, + "grad_norm": 0.15288174152374268, + "learning_rate": 2.8663934761454987e-05, + "loss": 3.1116, + "step": 16464 + }, + { + "epoch": 0.21484563271884605, + "grad_norm": 0.1484336405992508, + "learning_rate": 2.8661271874964503e-05, + "loss": 3.0593, + "step": 16465 + }, + { + "epoch": 0.21487647918441946, + "grad_norm": 0.15556636452674866, + "learning_rate": 2.865860894602377e-05, + "loss": 3.1309, + "step": 16466 + }, + { + "epoch": 0.21490732564999288, + "grad_norm": 0.1673084944486618, + "learning_rate": 2.8655945974663663e-05, + "loss": 3.1364, + "step": 16467 + }, + { + "epoch": 0.2149381721155663, + "grad_norm": 0.15745577216148376, + "learning_rate": 2.865328296091505e-05, + "loss": 3.0999, + "step": 16468 + }, + { + "epoch": 0.2149690185811397, + "grad_norm": 0.14939522743225098, + "learning_rate": 2.865061990480883e-05, + "loss": 3.1026, + "step": 16469 + }, + { + "epoch": 0.21499986504671312, + "grad_norm": 0.16307201981544495, + "learning_rate": 2.8647956806375858e-05, + "loss": 3.1306, + "step": 16470 + }, + { + "epoch": 0.21503071151228653, + "grad_norm": 0.15355582535266876, + "learning_rate": 2.8645293665647027e-05, + "loss": 3.1156, + "step": 16471 + }, + { + "epoch": 0.21506155797785995, + "grad_norm": 0.156166210770607, + "learning_rate": 2.8642630482653205e-05, + "loss": 3.1429, + "step": 16472 + }, + { + "epoch": 0.21509240444343336, + "grad_norm": 0.1569298654794693, + "learning_rate": 2.8639967257425266e-05, + "loss": 3.0533, + "step": 16473 + }, + { + "epoch": 0.21512325090900677, + "grad_norm": 0.15349189937114716, + "learning_rate": 2.863730398999409e-05, + "loss": 3.1198, + "step": 16474 + }, + { + "epoch": 0.2151540973745802, + "grad_norm": 0.15770982205867767, + "learning_rate": 2.8634640680390573e-05, + "loss": 3.1109, + "step": 16475 + }, + { + "epoch": 0.2151849438401536, + "grad_norm": 0.15280845761299133, + "learning_rate": 2.863197732864557e-05, + "loss": 3.0885, + "step": 16476 + }, + { + "epoch": 0.21521579030572704, + "grad_norm": 0.15819379687309265, + "learning_rate": 2.862931393478998e-05, + "loss": 3.1071, + "step": 16477 + }, + { + "epoch": 0.21524663677130046, + "grad_norm": 0.15587987005710602, + "learning_rate": 2.8626650498854672e-05, + "loss": 3.0999, + "step": 16478 + }, + { + "epoch": 0.21527748323687387, + "grad_norm": 0.18253351747989655, + "learning_rate": 2.862398702087053e-05, + "loss": 3.1238, + "step": 16479 + }, + { + "epoch": 0.2153083297024473, + "grad_norm": 0.15422803163528442, + "learning_rate": 2.862132350086844e-05, + "loss": 3.1076, + "step": 16480 + }, + { + "epoch": 0.2153391761680207, + "grad_norm": 0.15750783681869507, + "learning_rate": 2.8618659938879283e-05, + "loss": 3.1101, + "step": 16481 + }, + { + "epoch": 0.21537002263359412, + "grad_norm": 0.15391357243061066, + "learning_rate": 2.861599633493393e-05, + "loss": 3.0743, + "step": 16482 + }, + { + "epoch": 0.21540086909916753, + "grad_norm": 0.1565800905227661, + "learning_rate": 2.861333268906328e-05, + "loss": 3.0983, + "step": 16483 + }, + { + "epoch": 0.21543171556474094, + "grad_norm": 0.1557079702615738, + "learning_rate": 2.8610669001298206e-05, + "loss": 3.1171, + "step": 16484 + }, + { + "epoch": 0.21546256203031436, + "grad_norm": 0.16025978326797485, + "learning_rate": 2.860800527166959e-05, + "loss": 3.0975, + "step": 16485 + }, + { + "epoch": 0.21549340849588777, + "grad_norm": 0.16127195954322815, + "learning_rate": 2.860534150020833e-05, + "loss": 3.1097, + "step": 16486 + }, + { + "epoch": 0.2155242549614612, + "grad_norm": 0.15227583050727844, + "learning_rate": 2.86026776869453e-05, + "loss": 3.1076, + "step": 16487 + }, + { + "epoch": 0.2155551014270346, + "grad_norm": 0.1624215841293335, + "learning_rate": 2.8600013831911376e-05, + "loss": 3.0768, + "step": 16488 + }, + { + "epoch": 0.21558594789260804, + "grad_norm": 0.15909422934055328, + "learning_rate": 2.8597349935137468e-05, + "loss": 3.0623, + "step": 16489 + }, + { + "epoch": 0.21561679435818146, + "grad_norm": 0.14778363704681396, + "learning_rate": 2.859468599665444e-05, + "loss": 3.0914, + "step": 16490 + }, + { + "epoch": 0.21564764082375487, + "grad_norm": 0.2217102199792862, + "learning_rate": 2.8592022016493192e-05, + "loss": 3.1273, + "step": 16491 + }, + { + "epoch": 0.21567848728932829, + "grad_norm": 0.1524970978498459, + "learning_rate": 2.8589357994684614e-05, + "loss": 3.1045, + "step": 16492 + }, + { + "epoch": 0.2157093337549017, + "grad_norm": 0.15339045226573944, + "learning_rate": 2.858669393125958e-05, + "loss": 3.1312, + "step": 16493 + }, + { + "epoch": 0.21574018022047511, + "grad_norm": 0.15862654149532318, + "learning_rate": 2.858402982624898e-05, + "loss": 3.1042, + "step": 16494 + }, + { + "epoch": 0.21577102668604853, + "grad_norm": 0.1633695513010025, + "learning_rate": 2.8581365679683714e-05, + "loss": 3.1055, + "step": 16495 + }, + { + "epoch": 0.21580187315162194, + "grad_norm": 0.16720961034297943, + "learning_rate": 2.8578701491594667e-05, + "loss": 3.1293, + "step": 16496 + }, + { + "epoch": 0.21583271961719536, + "grad_norm": 0.15521731972694397, + "learning_rate": 2.8576037262012722e-05, + "loss": 3.1164, + "step": 16497 + }, + { + "epoch": 0.21586356608276877, + "grad_norm": 0.16260719299316406, + "learning_rate": 2.857337299096878e-05, + "loss": 3.1171, + "step": 16498 + }, + { + "epoch": 0.21589441254834218, + "grad_norm": 0.15774600207805634, + "learning_rate": 2.857070867849373e-05, + "loss": 3.0813, + "step": 16499 + }, + { + "epoch": 0.2159252590139156, + "grad_norm": 0.15626537799835205, + "learning_rate": 2.8568044324618443e-05, + "loss": 3.0879, + "step": 16500 + } + ], + "logging_steps": 1.0, + "max_steps": 32418, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 2.753158166555394e+21, + "train_batch_size": 4, + "trial_name": null, + "trial_params": null +}