| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.0, | |
| "eval_steps": 500, | |
| "global_step": 924, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.008658008658008658, | |
| "grad_norm": 5.4204916523303837e-05, | |
| "learning_rate": 4.255319148936171e-06, | |
| "loss": 2.1959, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.017316017316017316, | |
| "grad_norm": 4.602531771524809e-05, | |
| "learning_rate": 1.2765957446808511e-05, | |
| "loss": 1.982, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.025974025974025976, | |
| "grad_norm": 3.824224040727131e-05, | |
| "learning_rate": 2.1276595744680852e-05, | |
| "loss": 2.2121, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.03463203463203463, | |
| "grad_norm": 2.3534847059636377e-05, | |
| "learning_rate": 2.9787234042553192e-05, | |
| "loss": 2.0893, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.04329004329004329, | |
| "grad_norm": 1.8666707546799444e-05, | |
| "learning_rate": 3.829787234042553e-05, | |
| "loss": 1.8373, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.05194805194805195, | |
| "grad_norm": 1.5847288523218594e-05, | |
| "learning_rate": 4.680851063829788e-05, | |
| "loss": 2.0273, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.06060606060606061, | |
| "grad_norm": 1.0995895536325406e-05, | |
| "learning_rate": 5.531914893617022e-05, | |
| "loss": 1.8151, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.06926406926406926, | |
| "grad_norm": 1.6404985217377543e-05, | |
| "learning_rate": 6.382978723404256e-05, | |
| "loss": 2.0606, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.07792207792207792, | |
| "grad_norm": 1.4904575436958112e-05, | |
| "learning_rate": 7.23404255319149e-05, | |
| "loss": 1.9488, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.08658008658008658, | |
| "grad_norm": 1.3000194485357497e-05, | |
| "learning_rate": 8.085106382978723e-05, | |
| "loss": 2.6669, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.09523809523809523, | |
| "grad_norm": 1.2479895303840749e-05, | |
| "learning_rate": 8.936170212765958e-05, | |
| "loss": 1.742, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.1038961038961039, | |
| "grad_norm": 1.0467254469403997e-05, | |
| "learning_rate": 9.787234042553192e-05, | |
| "loss": 1.7994, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.11255411255411256, | |
| "grad_norm": 1.3622248843603302e-05, | |
| "learning_rate": 0.00010638297872340425, | |
| "loss": 1.9219, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.12121212121212122, | |
| "grad_norm": 1.3718491572944913e-05, | |
| "learning_rate": 0.00011489361702127661, | |
| "loss": 1.8505, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.12987012987012986, | |
| "grad_norm": 1.3184439012547955e-05, | |
| "learning_rate": 0.00012340425531914893, | |
| "loss": 1.7647, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.13852813852813853, | |
| "grad_norm": 1.2686439731623977e-05, | |
| "learning_rate": 0.00013191489361702127, | |
| "loss": 1.8865, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.1471861471861472, | |
| "grad_norm": 3.454366378718987e-05, | |
| "learning_rate": 0.00014042553191489363, | |
| "loss": 1.8743, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.15584415584415584, | |
| "grad_norm": 1.81365103344433e-05, | |
| "learning_rate": 0.00014893617021276596, | |
| "loss": 1.9598, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.1645021645021645, | |
| "grad_norm": 1.4871564417262562e-05, | |
| "learning_rate": 0.00015744680851063832, | |
| "loss": 1.7956, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.17316017316017315, | |
| "grad_norm": 1.3106016922392882e-05, | |
| "learning_rate": 0.00016595744680851065, | |
| "loss": 1.6608, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 1.4674634257971775e-05, | |
| "learning_rate": 0.00017446808510638298, | |
| "loss": 1.7878, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.19047619047619047, | |
| "grad_norm": 2.0664912881329656e-05, | |
| "learning_rate": 0.00018297872340425532, | |
| "loss": 1.8639, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.19913419913419914, | |
| "grad_norm": 1.741530650178902e-05, | |
| "learning_rate": 0.00019148936170212768, | |
| "loss": 1.5938, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.2077922077922078, | |
| "grad_norm": 1.3442709132505115e-05, | |
| "learning_rate": 0.0002, | |
| "loss": 1.6691, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.21645021645021645, | |
| "grad_norm": 2.1207506506470963e-05, | |
| "learning_rate": 0.00019999743357429378, | |
| "loss": 1.804, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.22510822510822512, | |
| "grad_norm": 3.287912477389909e-05, | |
| "learning_rate": 0.00019998973442890598, | |
| "loss": 1.8404, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.23376623376623376, | |
| "grad_norm": 2.17068190977443e-05, | |
| "learning_rate": 0.00019997690295902226, | |
| "loss": 1.7561, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.24242424242424243, | |
| "grad_norm": 1.806787622626871e-05, | |
| "learning_rate": 0.00019995893982326286, | |
| "loss": 1.8094, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.2510822510822511, | |
| "grad_norm": 1.9331200746819377e-05, | |
| "learning_rate": 0.00019993584594364894, | |
| "loss": 1.9647, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.2597402597402597, | |
| "grad_norm": 1.799209712771699e-05, | |
| "learning_rate": 0.00019990762250555495, | |
| "loss": 1.6961, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.2683982683982684, | |
| "grad_norm": 1.7485201169620268e-05, | |
| "learning_rate": 0.0001998742709576481, | |
| "loss": 1.8473, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.27705627705627706, | |
| "grad_norm": 1.571387838339433e-05, | |
| "learning_rate": 0.00019983579301181373, | |
| "loss": 1.5972, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 1.9293982404633425e-05, | |
| "learning_rate": 0.00019979219064306762, | |
| "loss": 1.6437, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.2943722943722944, | |
| "grad_norm": 1.444847475795541e-05, | |
| "learning_rate": 0.00019974346608945466, | |
| "loss": 1.6366, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.30303030303030304, | |
| "grad_norm": 1.3742756891588215e-05, | |
| "learning_rate": 0.00019968962185193365, | |
| "loss": 1.5994, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.3116883116883117, | |
| "grad_norm": 8.885600254870951e-05, | |
| "learning_rate": 0.00019963066069424943, | |
| "loss": 1.7003, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.3203463203463203, | |
| "grad_norm": 2.5485469450359233e-05, | |
| "learning_rate": 0.0001995665856427905, | |
| "loss": 1.6381, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.329004329004329, | |
| "grad_norm": 3.399766137590632e-05, | |
| "learning_rate": 0.00019949739998643414, | |
| "loss": 1.6196, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.33766233766233766, | |
| "grad_norm": 3.7044861528556794e-05, | |
| "learning_rate": 0.00019942310727637724, | |
| "loss": 1.667, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.3463203463203463, | |
| "grad_norm": 3.5706671042134985e-05, | |
| "learning_rate": 0.00019934371132595424, | |
| "loss": 1.8312, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.354978354978355, | |
| "grad_norm": 2.0404797396622598e-05, | |
| "learning_rate": 0.00019925921621044129, | |
| "loss": 1.6482, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 1.7347399989375845e-05, | |
| "learning_rate": 0.00019916962626684707, | |
| "loss": 1.6175, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.3722943722943723, | |
| "grad_norm": 3.3090138458646834e-05, | |
| "learning_rate": 0.00019907494609369035, | |
| "loss": 1.8201, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.38095238095238093, | |
| "grad_norm": 2.85672413156135e-05, | |
| "learning_rate": 0.0001989751805507637, | |
| "loss": 1.7632, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.38961038961038963, | |
| "grad_norm": 3.10399555019103e-05, | |
| "learning_rate": 0.00019887033475888419, | |
| "loss": 1.648, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.39826839826839827, | |
| "grad_norm": 2.8758275220752694e-05, | |
| "learning_rate": 0.00019876041409963056, | |
| "loss": 1.6615, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.4069264069264069, | |
| "grad_norm": 2.411049536021892e-05, | |
| "learning_rate": 0.00019864542421506686, | |
| "loss": 1.6684, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.4155844155844156, | |
| "grad_norm": 2.1932239178568125e-05, | |
| "learning_rate": 0.00019852537100745307, | |
| "loss": 1.5879, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.42424242424242425, | |
| "grad_norm": 2.419168049527798e-05, | |
| "learning_rate": 0.00019840026063894193, | |
| "loss": 1.6055, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.4329004329004329, | |
| "grad_norm": 3.5360833862796426e-05, | |
| "learning_rate": 0.00019827009953126275, | |
| "loss": 1.6233, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.44155844155844154, | |
| "grad_norm": 5.8022818848257884e-05, | |
| "learning_rate": 0.0001981348943653918, | |
| "loss": 1.5232, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.45021645021645024, | |
| "grad_norm": 3.986386946053244e-05, | |
| "learning_rate": 0.0001979946520812093, | |
| "loss": 1.7553, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.4588744588744589, | |
| "grad_norm": 5.403523027780466e-05, | |
| "learning_rate": 0.00019784937987714333, | |
| "loss": 1.6931, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.4675324675324675, | |
| "grad_norm": 2.1774114429717883e-05, | |
| "learning_rate": 0.00019769908520980034, | |
| "loss": 1.5427, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.47619047619047616, | |
| "grad_norm": 3.063875192310661e-05, | |
| "learning_rate": 0.0001975437757935822, | |
| "loss": 1.7336, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.48484848484848486, | |
| "grad_norm": 2.7840827897307463e-05, | |
| "learning_rate": 0.0001973834596002905, | |
| "loss": 1.6536, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.4935064935064935, | |
| "grad_norm": 2.68031708401395e-05, | |
| "learning_rate": 0.00019721814485871726, | |
| "loss": 1.7237, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.5021645021645021, | |
| "grad_norm": 4.067725967615843e-05, | |
| "learning_rate": 0.0001970478400542225, | |
| "loss": 1.6127, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.5108225108225108, | |
| "grad_norm": 3.361668132129125e-05, | |
| "learning_rate": 0.00019687255392829877, | |
| "loss": 1.5502, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.5194805194805194, | |
| "grad_norm": 3.693327744258568e-05, | |
| "learning_rate": 0.00019669229547812249, | |
| "loss": 1.6483, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.5281385281385281, | |
| "grad_norm": 3.008819476235658e-05, | |
| "learning_rate": 0.00019650707395609204, | |
| "loss": 1.6436, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.5367965367965368, | |
| "grad_norm": 3.0448116376646794e-05, | |
| "learning_rate": 0.00019631689886935298, | |
| "loss": 1.4893, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.5454545454545454, | |
| "grad_norm": 4.133440961595625e-05, | |
| "learning_rate": 0.00019612177997930987, | |
| "loss": 1.6093, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.5541125541125541, | |
| "grad_norm": 8.941220585256815e-05, | |
| "learning_rate": 0.00019592172730112544, | |
| "loss": 1.737, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.5627705627705628, | |
| "grad_norm": 4.473665831028484e-05, | |
| "learning_rate": 0.00019571675110320643, | |
| "loss": 1.6455, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 0.00013181190297473222, | |
| "learning_rate": 0.00019550686190667648, | |
| "loss": 1.5996, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.5800865800865801, | |
| "grad_norm": 9.064975165529177e-05, | |
| "learning_rate": 0.0001952920704848362, | |
| "loss": 1.6947, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.5887445887445888, | |
| "grad_norm": 3.242671664338559e-05, | |
| "learning_rate": 0.00019507238786261008, | |
| "loss": 1.6403, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.5974025974025974, | |
| "grad_norm": 0.0001139045343734324, | |
| "learning_rate": 0.00019484782531598073, | |
| "loss": 1.6445, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.6060606060606061, | |
| "grad_norm": 0.0001039160051732324, | |
| "learning_rate": 0.00019461839437141004, | |
| "loss": 1.7483, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.6147186147186147, | |
| "grad_norm": 8.791654545348138e-05, | |
| "learning_rate": 0.0001943841068052474, | |
| "loss": 1.5357, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.6233766233766234, | |
| "grad_norm": 4.35874389950186e-05, | |
| "learning_rate": 0.0001941449746431255, | |
| "loss": 1.7572, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.6320346320346321, | |
| "grad_norm": 4.0783703298075125e-05, | |
| "learning_rate": 0.0001939010101593429, | |
| "loss": 1.7342, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.6406926406926406, | |
| "grad_norm": 0.00014383271627593786, | |
| "learning_rate": 0.00019365222587623405, | |
| "loss": 1.6868, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.6493506493506493, | |
| "grad_norm": 9.058301657205448e-05, | |
| "learning_rate": 0.00019339863456352657, | |
| "loss": 1.7322, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.658008658008658, | |
| "grad_norm": 4.600944521371275e-05, | |
| "learning_rate": 0.0001931402492376857, | |
| "loss": 1.7196, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.00011446730059105903, | |
| "learning_rate": 0.0001928770831612463, | |
| "loss": 1.6174, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.6753246753246753, | |
| "grad_norm": 8.582652662880719e-05, | |
| "learning_rate": 0.00019260914984213203, | |
| "loss": 1.6256, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.683982683982684, | |
| "grad_norm": 4.044194065500051e-05, | |
| "learning_rate": 0.00019233646303296205, | |
| "loss": 1.6417, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.6926406926406926, | |
| "grad_norm": 4.7330380766652524e-05, | |
| "learning_rate": 0.0001920590367303451, | |
| "loss": 1.6737, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.7012987012987013, | |
| "grad_norm": 5.3459320042748004e-05, | |
| "learning_rate": 0.00019177688517416105, | |
| "loss": 1.7177, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.70995670995671, | |
| "grad_norm": 9.026160842040554e-05, | |
| "learning_rate": 0.00019149002284683008, | |
| "loss": 1.7326, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.7186147186147186, | |
| "grad_norm": 6.937263970030472e-05, | |
| "learning_rate": 0.0001911984644725692, | |
| "loss": 1.7971, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 8.01292626420036e-05, | |
| "learning_rate": 0.0001909022250166365, | |
| "loss": 1.6087, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.7359307359307359, | |
| "grad_norm": 7.882928184699267e-05, | |
| "learning_rate": 0.00019060131968456312, | |
| "loss": 1.6477, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.7445887445887446, | |
| "grad_norm": 0.00010625456343404949, | |
| "learning_rate": 0.00019029576392137263, | |
| "loss": 1.7002, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.7532467532467533, | |
| "grad_norm": 0.00021590011601801962, | |
| "learning_rate": 0.00018998557341078835, | |
| "loss": 1.7764, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.7619047619047619, | |
| "grad_norm": 6.014715472701937e-05, | |
| "learning_rate": 0.00018967076407442829, | |
| "loss": 1.6852, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.7705627705627706, | |
| "grad_norm": 8.015201456146315e-05, | |
| "learning_rate": 0.00018935135207098785, | |
| "loss": 1.6184, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.7792207792207793, | |
| "grad_norm": 6.944081542314962e-05, | |
| "learning_rate": 0.00018902735379541064, | |
| "loss": 1.6294, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.7878787878787878, | |
| "grad_norm": 6.165856029838324e-05, | |
| "learning_rate": 0.0001886987858780467, | |
| "loss": 1.778, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.7965367965367965, | |
| "grad_norm": 8.64676694618538e-05, | |
| "learning_rate": 0.000188365665183799, | |
| "loss": 1.5531, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.8051948051948052, | |
| "grad_norm": 9.756162035046145e-05, | |
| "learning_rate": 0.00018802800881125784, | |
| "loss": 1.7175, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.8138528138528138, | |
| "grad_norm": 0.00021862791618332267, | |
| "learning_rate": 0.00018768583409182305, | |
| "loss": 1.8573, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.8225108225108225, | |
| "grad_norm": 0.0001369621604681015, | |
| "learning_rate": 0.00018733915858881462, | |
| "loss": 1.7662, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.8311688311688312, | |
| "grad_norm": 0.00031745131127536297, | |
| "learning_rate": 0.00018698800009657094, | |
| "loss": 2.0091, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.8398268398268398, | |
| "grad_norm": 0.00010971837764373049, | |
| "learning_rate": 0.00018663237663953567, | |
| "loss": 1.7376, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.8484848484848485, | |
| "grad_norm": 0.00029516848735511303, | |
| "learning_rate": 0.0001862723064713324, | |
| "loss": 1.7701, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 0.00034297676756978035, | |
| "learning_rate": 0.0001859078080738279, | |
| "loss": 1.9447, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.8658008658008658, | |
| "grad_norm": 0.0002530592610128224, | |
| "learning_rate": 0.00018553890015618333, | |
| "loss": 1.8612, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.8744588744588745, | |
| "grad_norm": 0.00010490286513231695, | |
| "learning_rate": 0.00018516560165389388, | |
| "loss": 1.9922, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.8831168831168831, | |
| "grad_norm": 0.00020270211098250002, | |
| "learning_rate": 0.00018478793172781708, | |
| "loss": 1.9509, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.8917748917748918, | |
| "grad_norm": 9.309218876296654e-05, | |
| "learning_rate": 0.000184405909763189, | |
| "loss": 1.909, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.9004329004329005, | |
| "grad_norm": 7.871213892940432e-05, | |
| "learning_rate": 0.00018401955536862948, | |
| "loss": 2.0207, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 8.371711010113358e-05, | |
| "learning_rate": 0.00018362888837513547, | |
| "loss": 2.0631, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.9177489177489178, | |
| "grad_norm": 0.00014524892321787775, | |
| "learning_rate": 0.00018323392883506335, | |
| "loss": 2.2097, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.9264069264069265, | |
| "grad_norm": 0.00010290888894814998, | |
| "learning_rate": 0.00018283469702109936, | |
| "loss": 2.1094, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.935064935064935, | |
| "grad_norm": 9.359593968838453e-05, | |
| "learning_rate": 0.00018243121342521935, | |
| "loss": 2.4721, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.9437229437229437, | |
| "grad_norm": 9.659567876951769e-05, | |
| "learning_rate": 0.0001820234987576368, | |
| "loss": 2.6673, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.9523809523809523, | |
| "grad_norm": 0.00016237075033131987, | |
| "learning_rate": 0.0001816115739457397, | |
| "loss": 2.9459, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.961038961038961, | |
| "grad_norm": 0.0001531827583676204, | |
| "learning_rate": 0.00018119546013301664, | |
| "loss": 3.2397, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.9696969696969697, | |
| "grad_norm": 0.0001731283700792119, | |
| "learning_rate": 0.0001807751786779713, | |
| "loss": 3.62, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.9783549783549783, | |
| "grad_norm": 8.82472813827917e-05, | |
| "learning_rate": 0.00018035075115302633, | |
| "loss": 3.9451, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.987012987012987, | |
| "grad_norm": 0.00010907312389463186, | |
| "learning_rate": 0.0001799221993434159, | |
| "loss": 4.6026, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.9956709956709957, | |
| "grad_norm": 0.00013694152585230768, | |
| "learning_rate": 0.00017948954524606763, | |
| "loss": 5.3581, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.0043290043290043, | |
| "grad_norm": 0.00015375320799648762, | |
| "learning_rate": 0.00017905281106847344, | |
| "loss": 5.3862, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.0129870129870129, | |
| "grad_norm": 0.000532141828443855, | |
| "learning_rate": 0.00017861201922754979, | |
| "loss": 6.0815, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.0216450216450217, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.00017839010874560733, | |
| "loss": 0.0, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.0303030303030303, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.00017839010874560733, | |
| "loss": 0.0, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.0389610389610389, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.00017839010874560733, | |
| "loss": 0.0, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.0476190476190477, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.00017839010874560733, | |
| "loss": 0.0, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.0562770562770563, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.00017839010874560733, | |
| "loss": 0.0, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.0649350649350648, | |
| "grad_norm": 0.1457211822271347, | |
| "learning_rate": 0.0001781671923484869, | |
| "loss": 5.7907, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.0735930735930737, | |
| "grad_norm": 0.4493274390697479, | |
| "learning_rate": 0.00017771835326358743, | |
| "loss": 7.5356, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.0822510822510822, | |
| "grad_norm": 3.0114400386810303, | |
| "learning_rate": 0.00017726552501109478, | |
| "loss": 7.336, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 0.23199492692947388, | |
| "learning_rate": 0.0001768087308340103, | |
| "loss": 8.1303, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.0995670995670996, | |
| "grad_norm": 0.23555096983909607, | |
| "learning_rate": 0.00017634799417890035, | |
| "loss": 9.2992, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.1082251082251082, | |
| "grad_norm": 0.17370502650737762, | |
| "learning_rate": 0.0001758833386946928, | |
| "loss": 9.3926, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.1168831168831168, | |
| "grad_norm": 0.25211918354034424, | |
| "learning_rate": 0.00017541478823146327, | |
| "loss": 9.3047, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.1255411255411256, | |
| "grad_norm": 0.18103384971618652, | |
| "learning_rate": 0.00017494236683921084, | |
| "loss": 9.1649, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.1341991341991342, | |
| "grad_norm": 0.05794016644358635, | |
| "learning_rate": 0.00017446609876662356, | |
| "loss": 8.9708, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 0.08229361474514008, | |
| "learning_rate": 0.000173986008459834, | |
| "loss": 9.7143, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.1515151515151516, | |
| "grad_norm": 0.1594569981098175, | |
| "learning_rate": 0.00017350212056116418, | |
| "loss": 10.6715, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.1601731601731602, | |
| "grad_norm": 0.19106224179267883, | |
| "learning_rate": 0.00017301445990786102, | |
| "loss": 11.408, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.1688311688311688, | |
| "grad_norm": 0.2994789183139801, | |
| "learning_rate": 0.00017252305153082114, | |
| "loss": 12.2243, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.1774891774891776, | |
| "grad_norm": 0.21744585037231445, | |
| "learning_rate": 0.00017202792065330646, | |
| "loss": 12.9086, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.1861471861471862, | |
| "grad_norm": 0.16616225242614746, | |
| "learning_rate": 0.00017152909268964916, | |
| "loss": 13.5227, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.1948051948051948, | |
| "grad_norm": 0.15657858550548553, | |
| "learning_rate": 0.00017102659324394747, | |
| "loss": 13.9467, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.2034632034632033, | |
| "grad_norm": 0.16317234933376312, | |
| "learning_rate": 0.00017052044810875126, | |
| "loss": 14.6343, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.2121212121212122, | |
| "grad_norm": 0.21492412686347961, | |
| "learning_rate": 0.00017001068326373827, | |
| "loss": 15.0887, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.2207792207792207, | |
| "grad_norm": 0.1301400512456894, | |
| "learning_rate": 0.00016949732487438047, | |
| "loss": 15.6847, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.2294372294372296, | |
| "grad_norm": 0.20310911536216736, | |
| "learning_rate": 0.00016898039929060129, | |
| "loss": 16.6419, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.2380952380952381, | |
| "grad_norm": 0.05390779674053192, | |
| "learning_rate": 0.00016845993304542283, | |
| "loss": 15.3988, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.2467532467532467, | |
| "grad_norm": 0.16172261536121368, | |
| "learning_rate": 0.0001679359528536041, | |
| "loss": 15.9208, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.2554112554112553, | |
| "grad_norm": 0.13750512897968292, | |
| "learning_rate": 0.0001674084856102698, | |
| "loss": 16.2978, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.2640692640692641, | |
| "grad_norm": 0.07984600216150284, | |
| "learning_rate": 0.00016687755838952972, | |
| "loss": 15.957, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.2727272727272727, | |
| "grad_norm": 0.13223163783550262, | |
| "learning_rate": 0.00016634319844308925, | |
| "loss": 16.2106, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.2813852813852815, | |
| "grad_norm": 0.08291322737932205, | |
| "learning_rate": 0.00016580543319885048, | |
| "loss": 16.14, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.29004329004329, | |
| "grad_norm": 0.052721720188856125, | |
| "learning_rate": 0.00016526429025950424, | |
| "loss": 15.9756, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.2987012987012987, | |
| "grad_norm": 0.08016930520534515, | |
| "learning_rate": 0.00016471979740111366, | |
| "loss": 16.2873, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.3073593073593073, | |
| "grad_norm": 0.09582065045833588, | |
| "learning_rate": 0.00016417198257168803, | |
| "loss": 16.4316, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.316017316017316, | |
| "grad_norm": 0.07704820483922958, | |
| "learning_rate": 0.00016362087388974863, | |
| "loss": 16.7252, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.3246753246753247, | |
| "grad_norm": 0.06254423409700394, | |
| "learning_rate": 0.00016306649964288516, | |
| "loss": 16.1704, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.07229200750589371, | |
| "learning_rate": 0.000162508888286304, | |
| "loss": 16.0079, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.341991341991342, | |
| "grad_norm": 0.052998676896095276, | |
| "learning_rate": 0.00016194806844136754, | |
| "loss": 16.0897, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.3506493506493507, | |
| "grad_norm": 0.10584064573049545, | |
| "learning_rate": 0.00016138406889412512, | |
| "loss": 15.995, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.3593073593073592, | |
| "grad_norm": 0.11045046895742416, | |
| "learning_rate": 0.00016081691859383545, | |
| "loss": 16.0464, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 1.3679653679653678, | |
| "grad_norm": 0.06835480034351349, | |
| "learning_rate": 0.00016024664665148077, | |
| "loss": 15.842, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 1.3766233766233766, | |
| "grad_norm": 0.05876036360859871, | |
| "learning_rate": 0.00015967328233827249, | |
| "loss": 15.5758, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.3852813852813852, | |
| "grad_norm": 0.05429469048976898, | |
| "learning_rate": 0.00015909685508414884, | |
| "loss": 15.6389, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.393939393939394, | |
| "grad_norm": 0.08564180880784988, | |
| "learning_rate": 0.00015851739447626434, | |
| "loss": 15.5351, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 1.4025974025974026, | |
| "grad_norm": 0.08055119216442108, | |
| "learning_rate": 0.00015793493025747092, | |
| "loss": 15.4569, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 1.4112554112554112, | |
| "grad_norm": 0.08411835134029388, | |
| "learning_rate": 0.00015734949232479152, | |
| "loss": 15.5636, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 1.4199134199134198, | |
| "grad_norm": 0.15642018616199493, | |
| "learning_rate": 0.00015676111072788527, | |
| "loss": 15.7577, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 0.17614306509494781, | |
| "learning_rate": 0.00015616981566750538, | |
| "loss": 15.3891, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.4372294372294372, | |
| "grad_norm": 0.10892828553915024, | |
| "learning_rate": 0.00015557563749394858, | |
| "loss": 15.0548, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 1.445887445887446, | |
| "grad_norm": 0.09866315126419067, | |
| "learning_rate": 0.00015497860670549772, | |
| "loss": 14.8649, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 0.1259247213602066, | |
| "learning_rate": 0.00015437875394685606, | |
| "loss": 14.9433, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 1.4632034632034632, | |
| "grad_norm": 0.1163773462176323, | |
| "learning_rate": 0.0001537761100075744, | |
| "loss": 15.0705, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 1.4718614718614718, | |
| "grad_norm": 0.10914402455091476, | |
| "learning_rate": 0.00015317070582047065, | |
| "loss": 15.2118, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.4805194805194806, | |
| "grad_norm": 0.15627121925354004, | |
| "learning_rate": 0.00015256257246004217, | |
| "loss": 15.6857, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 1.4891774891774892, | |
| "grad_norm": 0.12600019574165344, | |
| "learning_rate": 0.00015195174114087078, | |
| "loss": 15.2044, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 1.497835497835498, | |
| "grad_norm": 0.1444402039051056, | |
| "learning_rate": 0.00015133824321602045, | |
| "loss": 15.1016, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 1.5064935064935066, | |
| "grad_norm": 0.07496843487024307, | |
| "learning_rate": 0.00015072211017542813, | |
| "loss": 14.8805, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 1.5151515151515151, | |
| "grad_norm": 0.14399290084838867, | |
| "learning_rate": 0.0001501033736442872, | |
| "loss": 14.8005, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.5238095238095237, | |
| "grad_norm": 0.10512608289718628, | |
| "learning_rate": 0.00014948206538142457, | |
| "loss": 14.4348, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 1.5324675324675323, | |
| "grad_norm": 0.05682501569390297, | |
| "learning_rate": 0.00014885821727767006, | |
| "loss": 14.274, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 1.5411255411255411, | |
| "grad_norm": 0.09160462766885757, | |
| "learning_rate": 0.00014823186135421994, | |
| "loss": 14.1816, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 1.54978354978355, | |
| "grad_norm": 0.03531830757856369, | |
| "learning_rate": 0.00014760302976099304, | |
| "loss": 14.0882, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 1.5584415584415585, | |
| "grad_norm": 0.039006441831588745, | |
| "learning_rate": 0.00014697175477498074, | |
| "loss": 14.1806, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.567099567099567, | |
| "grad_norm": 0.07611044496297836, | |
| "learning_rate": 0.00014633806879859, | |
| "loss": 14.2338, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 1.5757575757575757, | |
| "grad_norm": 0.05836552381515503, | |
| "learning_rate": 0.00014570200435798044, | |
| "loss": 14.2683, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 1.5844155844155843, | |
| "grad_norm": 0.01268097199499607, | |
| "learning_rate": 0.0001450635941013947, | |
| "loss": 14.2289, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 1.593073593073593, | |
| "grad_norm": 0.07083582133054733, | |
| "learning_rate": 0.00014442287079748263, | |
| "loss": 14.3293, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 1.601731601731602, | |
| "grad_norm": 0.07286886125802994, | |
| "learning_rate": 0.0001437798673336194, | |
| "loss": 14.4253, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.6103896103896105, | |
| "grad_norm": 0.014986686408519745, | |
| "learning_rate": 0.00014313461671421735, | |
| "loss": 14.2353, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 1.619047619047619, | |
| "grad_norm": 0.06591885536909103, | |
| "learning_rate": 0.00014248715205903204, | |
| "loss": 14.1536, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 1.6277056277056277, | |
| "grad_norm": 0.07969338446855545, | |
| "learning_rate": 0.0001418375066014622, | |
| "loss": 14.6147, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 1.6363636363636362, | |
| "grad_norm": 0.06318385899066925, | |
| "learning_rate": 0.00014118571368684383, | |
| "loss": 14.345, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 1.645021645021645, | |
| "grad_norm": 0.05836130306124687, | |
| "learning_rate": 0.00014053180677073876, | |
| "loss": 14.2053, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.6536796536796536, | |
| "grad_norm": 0.06260073184967041, | |
| "learning_rate": 0.0001398758194172174, | |
| "loss": 14.1105, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 1.6623376623376624, | |
| "grad_norm": 0.060454513877630234, | |
| "learning_rate": 0.00013921778529713582, | |
| "loss": 14.0394, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 1.670995670995671, | |
| "grad_norm": 0.07355903834104538, | |
| "learning_rate": 0.00013855773818640773, | |
| "loss": 14.025, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 1.6796536796536796, | |
| "grad_norm": 0.066501684486866, | |
| "learning_rate": 0.00013789571196427055, | |
| "loss": 14.6727, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 1.6883116883116882, | |
| "grad_norm": 0.07008553296327591, | |
| "learning_rate": 0.0001372317406115465, | |
| "loss": 14.0043, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.696969696969697, | |
| "grad_norm": 0.010356806218624115, | |
| "learning_rate": 0.00013656585820889867, | |
| "loss": 14.0629, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 1.7056277056277056, | |
| "grad_norm": 0.06225749850273132, | |
| "learning_rate": 0.00013589809893508128, | |
| "loss": 13.9668, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "grad_norm": 0.07188686728477478, | |
| "learning_rate": 0.00013522849706518566, | |
| "loss": 13.8812, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 1.722943722943723, | |
| "grad_norm": 0.06185540184378624, | |
| "learning_rate": 0.00013455708696888085, | |
| "loss": 14.1862, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 1.7316017316017316, | |
| "grad_norm": 0.0615684911608696, | |
| "learning_rate": 0.00013388390310864945, | |
| "loss": 14.4132, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.7402597402597402, | |
| "grad_norm": 0.07077145576477051, | |
| "learning_rate": 0.00013320898003801879, | |
| "loss": 14.4692, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 1.7489177489177488, | |
| "grad_norm": 0.05472411587834358, | |
| "learning_rate": 0.00013253235239978715, | |
| "loss": 13.9146, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 1.7575757575757576, | |
| "grad_norm": 0.06190333142876625, | |
| "learning_rate": 0.00013185405492424588, | |
| "loss": 14.0298, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 1.7662337662337664, | |
| "grad_norm": 0.06982716172933578, | |
| "learning_rate": 0.00013117412242739655, | |
| "loss": 14.0501, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 1.774891774891775, | |
| "grad_norm": 0.05709117650985718, | |
| "learning_rate": 0.00013049258980916387, | |
| "loss": 14.0339, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.7835497835497836, | |
| "grad_norm": 0.1394786387681961, | |
| "learning_rate": 0.00012980949205160448, | |
| "loss": 14.2314, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 1.7922077922077921, | |
| "grad_norm": 0.09034628421068192, | |
| "learning_rate": 0.00012912486421711128, | |
| "loss": 14.1102, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 1.8008658008658007, | |
| "grad_norm": 0.06254860013723373, | |
| "learning_rate": 0.00012843874144661372, | |
| "loss": 14.0066, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 1.8095238095238095, | |
| "grad_norm": 0.09414557367563248, | |
| "learning_rate": 0.00012775115895777417, | |
| "loss": 14.0197, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 0.021693430840969086, | |
| "learning_rate": 0.00012706215204318007, | |
| "loss": 14.9639, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.826839826839827, | |
| "grad_norm": 0.058936670422554016, | |
| "learning_rate": 0.00012637175606853264, | |
| "loss": 13.7909, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 1.8354978354978355, | |
| "grad_norm": 0.04964934289455414, | |
| "learning_rate": 0.0001256800064708313, | |
| "loss": 13.8304, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 1.844155844155844, | |
| "grad_norm": 0.04889138787984848, | |
| "learning_rate": 0.00012498693875655516, | |
| "loss": 13.9045, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 1.8528138528138527, | |
| "grad_norm": 0.015483302064239979, | |
| "learning_rate": 0.00012429258849984014, | |
| "loss": 13.9397, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 1.8614718614718615, | |
| "grad_norm": 0.009186900220811367, | |
| "learning_rate": 0.00012359699134065314, | |
| "loss": 13.7447, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.87012987012987, | |
| "grad_norm": 0.05520382896065712, | |
| "learning_rate": 0.00012290018298296285, | |
| "loss": 13.8049, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 1.878787878787879, | |
| "grad_norm": 0.05643709748983383, | |
| "learning_rate": 0.00012220219919290687, | |
| "loss": 14.3662, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 1.8874458874458875, | |
| "grad_norm": 0.05961998179554939, | |
| "learning_rate": 0.00012150307579695601, | |
| "loss": 13.9983, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 1.896103896103896, | |
| "grad_norm": 0.006795608904212713, | |
| "learning_rate": 0.00012080284868007541, | |
| "loss": 13.9398, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 1.9047619047619047, | |
| "grad_norm": 0.007627868093550205, | |
| "learning_rate": 0.00012010155378388253, | |
| "loss": 13.9321, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.9134199134199135, | |
| "grad_norm": 0.027596835047006607, | |
| "learning_rate": 0.00011939922710480229, | |
| "loss": 14.0445, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 1.922077922077922, | |
| "grad_norm": 0.0442458800971508, | |
| "learning_rate": 0.00011869590469221965, | |
| "loss": 13.999, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 1.9307359307359309, | |
| "grad_norm": 0.0031651495955884457, | |
| "learning_rate": 0.0001179916226466289, | |
| "loss": 14.3455, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 1.9393939393939394, | |
| "grad_norm": 0.0655621886253357, | |
| "learning_rate": 0.00011728641711778103, | |
| "loss": 14.0114, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 1.948051948051948, | |
| "grad_norm": 0.05245671793818474, | |
| "learning_rate": 0.000116580324302828, | |
| "loss": 14.1412, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.9567099567099566, | |
| "grad_norm": 0.017913660034537315, | |
| "learning_rate": 0.00011587338044446476, | |
| "loss": 14.0516, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 1.9653679653679652, | |
| "grad_norm": 0.005986363161355257, | |
| "learning_rate": 0.00011516562182906922, | |
| "loss": 14.0402, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 1.974025974025974, | |
| "grad_norm": 0.06156434491276741, | |
| "learning_rate": 0.0001144570847848394, | |
| "loss": 13.9566, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 1.9826839826839828, | |
| "grad_norm": 0.009826838970184326, | |
| "learning_rate": 0.000113747805679929, | |
| "loss": 13.9418, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 1.9913419913419914, | |
| "grad_norm": 0.009458528831601143, | |
| "learning_rate": 0.00011303782092058061, | |
| "loss": 13.8423, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.22517594695091248, | |
| "learning_rate": 0.00011232716694925693, | |
| "loss": 13.9573, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 2.0086580086580086, | |
| "grad_norm": 0.003495636396110058, | |
| "learning_rate": 0.00011161588024277036, | |
| "loss": 14.1671, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 2.017316017316017, | |
| "grad_norm": 0.013556540943682194, | |
| "learning_rate": 0.00011090399731041072, | |
| "loss": 14.503, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 2.0259740259740258, | |
| "grad_norm": 0.05908797308802605, | |
| "learning_rate": 0.0001101915546920711, | |
| "loss": 14.613, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 2.034632034632035, | |
| "grad_norm": 0.002222011098638177, | |
| "learning_rate": 0.00010947858895637255, | |
| "loss": 14.9046, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 2.0432900432900434, | |
| "grad_norm": 0.005948877427726984, | |
| "learning_rate": 0.00010876513669878683, | |
| "loss": 15.0299, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 2.051948051948052, | |
| "grad_norm": 0.0032331624533981085, | |
| "learning_rate": 0.0001080512345397583, | |
| "loss": 15.1622, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 2.0606060606060606, | |
| "grad_norm": 0.04379121959209442, | |
| "learning_rate": 0.00010733691912282396, | |
| "loss": 15.2752, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 2.069264069264069, | |
| "grad_norm": 0.003626056481152773, | |
| "learning_rate": 0.00010662222711273279, | |
| "loss": 15.4877, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 2.0779220779220777, | |
| "grad_norm": 0.0034651593305170536, | |
| "learning_rate": 0.00010590719519356373, | |
| "loss": 15.7603, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 2.0865800865800868, | |
| "grad_norm": 0.007096354383975267, | |
| "learning_rate": 0.00010519186006684277, | |
| "loss": 15.4634, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 2.0952380952380953, | |
| "grad_norm": 0.005978535860776901, | |
| "learning_rate": 0.000104476258449659, | |
| "loss": 15.6005, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 2.103896103896104, | |
| "grad_norm": 0.0036010430194437504, | |
| "learning_rate": 0.0001037604270727802, | |
| "loss": 15.6865, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 2.1125541125541125, | |
| "grad_norm": 0.005017112474888563, | |
| "learning_rate": 0.00010304440267876727, | |
| "loss": 15.8028, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 2.121212121212121, | |
| "grad_norm": 0.002776139648631215, | |
| "learning_rate": 0.00010232822202008844, | |
| "loss": 15.662, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 2.1298701298701297, | |
| "grad_norm": 0.0025852976832538843, | |
| "learning_rate": 0.0001016119218572328, | |
| "loss": 16.5597, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 2.1385281385281387, | |
| "grad_norm": 0.0041742087341845036, | |
| "learning_rate": 0.0001008955389568233, | |
| "loss": 15.6964, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 2.1471861471861473, | |
| "grad_norm": 0.003886697581037879, | |
| "learning_rate": 0.00010017911008972982, | |
| "loss": 15.6254, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 2.155844155844156, | |
| "grad_norm": 0.06273438781499863, | |
| "learning_rate": 9.946267202918157e-05, | |
| "loss": 15.7853, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 2.1645021645021645, | |
| "grad_norm": 0.0052159507758915424, | |
| "learning_rate": 9.87462615488797e-05, | |
| "loss": 15.7882, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.173160173160173, | |
| "grad_norm": 0.003929893020540476, | |
| "learning_rate": 9.802991542110958e-05, | |
| "loss": 16.0684, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "grad_norm": 0.004246190190315247, | |
| "learning_rate": 9.731367041485359e-05, | |
| "loss": 15.7668, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 2.1904761904761907, | |
| "grad_norm": 0.0031269013416022062, | |
| "learning_rate": 9.659756329390367e-05, | |
| "loss": 15.7715, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 2.1991341991341993, | |
| "grad_norm": 0.001676430692896247, | |
| "learning_rate": 9.588163081497427e-05, | |
| "loss": 15.6645, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 2.207792207792208, | |
| "grad_norm": 0.003888419596478343, | |
| "learning_rate": 9.516590972581578e-05, | |
| "loss": 15.6698, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 2.2164502164502164, | |
| "grad_norm": 0.0032913736067712307, | |
| "learning_rate": 9.445043676332819e-05, | |
| "loss": 15.7315, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 2.225108225108225, | |
| "grad_norm": 0.0013064603554084897, | |
| "learning_rate": 9.373524865167555e-05, | |
| "loss": 15.6309, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 2.2337662337662336, | |
| "grad_norm": 0.0033070454373955727, | |
| "learning_rate": 9.302038210040099e-05, | |
| "loss": 15.8009, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 2.242424242424242, | |
| "grad_norm": 0.002767252502962947, | |
| "learning_rate": 9.230587380254237e-05, | |
| "loss": 15.7214, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 2.2510822510822512, | |
| "grad_norm": 0.0022966829128563404, | |
| "learning_rate": 9.159176043274895e-05, | |
| "loss": 15.6413, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 2.25974025974026, | |
| "grad_norm": 0.002309858100488782, | |
| "learning_rate": 9.087807864539897e-05, | |
| "loss": 15.7846, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 2.2683982683982684, | |
| "grad_norm": 0.0039366851560771465, | |
| "learning_rate": 9.016486507271803e-05, | |
| "loss": 15.8136, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 2.277056277056277, | |
| "grad_norm": 0.004376592580229044, | |
| "learning_rate": 8.945215632289912e-05, | |
| "loss": 15.7964, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 2.2857142857142856, | |
| "grad_norm": 0.002474917098879814, | |
| "learning_rate": 8.873998897822336e-05, | |
| "loss": 15.9371, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 2.2943722943722946, | |
| "grad_norm": 0.0036089515779167414, | |
| "learning_rate": 8.802839959318239e-05, | |
| "loss": 15.7513, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 2.303030303030303, | |
| "grad_norm": 0.0045991260558366776, | |
| "learning_rate": 8.731742469260201e-05, | |
| "loss": 15.7613, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 2.311688311688312, | |
| "grad_norm": 0.0072856624610722065, | |
| "learning_rate": 8.66071007697674e-05, | |
| "loss": 16.4219, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 2.3203463203463204, | |
| "grad_norm": 0.0017560477135702968, | |
| "learning_rate": 8.58974642845501e-05, | |
| "loss": 16.0309, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 2.329004329004329, | |
| "grad_norm": 0.004301704466342926, | |
| "learning_rate": 8.518855166153644e-05, | |
| "loss": 27.1619, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 2.3376623376623376, | |
| "grad_norm": 0.059041913598775864, | |
| "learning_rate": 8.448039928815804e-05, | |
| "loss": 16.0355, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.346320346320346, | |
| "grad_norm": 0.002861848333850503, | |
| "learning_rate": 8.377304351282399e-05, | |
| "loss": 15.8526, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 2.354978354978355, | |
| "grad_norm": 0.003043128876015544, | |
| "learning_rate": 8.306652064305517e-05, | |
| "loss": 15.7983, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 2.3636363636363638, | |
| "grad_norm": 0.0002984661259688437, | |
| "learning_rate": 8.23608669436207e-05, | |
| "loss": 15.8203, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 2.3722943722943723, | |
| "grad_norm": 0.0006706724525429308, | |
| "learning_rate": 8.165611863467644e-05, | |
| "loss": 15.7032, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 2.380952380952381, | |
| "grad_norm": 0.0023594857193529606, | |
| "learning_rate": 8.095231188990597e-05, | |
| "loss": 15.8033, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.3896103896103895, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.024948283466367e-05, | |
| "loss": 15.722, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 2.398268398268398, | |
| "grad_norm": 0.0003026532067451626, | |
| "learning_rate": 7.954766754412066e-05, | |
| "loss": 15.7503, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 2.4069264069264067, | |
| "grad_norm": 0.0020434176549315453, | |
| "learning_rate": 7.884690204141298e-05, | |
| "loss": 15.8143, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 2.4155844155844157, | |
| "grad_norm": 0.03126376122236252, | |
| "learning_rate": 7.814722229579264e-05, | |
| "loss": 16.6621, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 2.4242424242424243, | |
| "grad_norm": 0.0025495837908238173, | |
| "learning_rate": 7.744866422078133e-05, | |
| "loss": 15.757, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.432900432900433, | |
| "grad_norm": 0.0023321521002799273, | |
| "learning_rate": 7.67512636723271e-05, | |
| "loss": 15.7909, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 2.4415584415584415, | |
| "grad_norm": 0.0014553000219166279, | |
| "learning_rate": 7.605505644696387e-05, | |
| "loss": 15.7724, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 2.45021645021645, | |
| "grad_norm": 0.0026701318565756083, | |
| "learning_rate": 7.536007827997397e-05, | |
| "loss": 15.7491, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 2.458874458874459, | |
| "grad_norm": 0.0020363188814371824, | |
| "learning_rate": 7.46663648435541e-05, | |
| "loss": 15.7404, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 2.4675324675324677, | |
| "grad_norm": 0.005381477996706963, | |
| "learning_rate": 7.397395174498417e-05, | |
| "loss": 15.7682, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.4761904761904763, | |
| "grad_norm": 0.0010892607970163226, | |
| "learning_rate": 7.328287452479968e-05, | |
| "loss": 15.7591, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 2.484848484848485, | |
| "grad_norm": 0.0046081384643912315, | |
| "learning_rate": 7.259316865496757e-05, | |
| "loss": 16.2354, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 2.4935064935064934, | |
| "grad_norm": 0.0023402958177030087, | |
| "learning_rate": 7.19048695370652e-05, | |
| "loss": 15.8337, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 2.502164502164502, | |
| "grad_norm": 0.0027834863867610693, | |
| "learning_rate": 7.121801250046363e-05, | |
| "loss": 15.9034, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 2.5108225108225106, | |
| "grad_norm": 0.004065465647727251, | |
| "learning_rate": 7.053263280051394e-05, | |
| "loss": 15.717, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.5194805194805197, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.984876561673776e-05, | |
| "loss": 15.7805, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 2.5281385281385282, | |
| "grad_norm": 0.002813364379107952, | |
| "learning_rate": 6.91664460510215e-05, | |
| "loss": 15.7205, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 2.536796536796537, | |
| "grad_norm": 0.0021334670018404722, | |
| "learning_rate": 6.848570912581463e-05, | |
| "loss": 15.8153, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 2.5454545454545454, | |
| "grad_norm": 0.003558355150744319, | |
| "learning_rate": 6.780658978233199e-05, | |
| "loss": 15.7313, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 2.554112554112554, | |
| "grad_norm": 0.001838353113271296, | |
| "learning_rate": 6.71291228787604e-05, | |
| "loss": 15.6268, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.562770562770563, | |
| "grad_norm": 0.003100910922512412, | |
| "learning_rate": 6.64533431884694e-05, | |
| "loss": 15.7013, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 2.571428571428571, | |
| "grad_norm": 0.0012829442275688052, | |
| "learning_rate": 6.57792853982264e-05, | |
| "loss": 15.832, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 2.58008658008658, | |
| "grad_norm": 0.001535665593110025, | |
| "learning_rate": 6.51069841064162e-05, | |
| "loss": 15.7034, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 2.588744588744589, | |
| "grad_norm": 0.010908816941082478, | |
| "learning_rate": 6.443647382126509e-05, | |
| "loss": 15.8574, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 2.5974025974025974, | |
| "grad_norm": 0.0028520359192043543, | |
| "learning_rate": 6.376778895906976e-05, | |
| "loss": 15.8502, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.606060606060606, | |
| "grad_norm": 0.011398903094232082, | |
| "learning_rate": 6.310096384243061e-05, | |
| "loss": 15.9701, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 2.6147186147186146, | |
| "grad_norm": 0.0021338535007089376, | |
| "learning_rate": 6.243603269849003e-05, | |
| "loss": 15.7824, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 2.6233766233766236, | |
| "grad_norm": 0.004288196098059416, | |
| "learning_rate": 6.177302965717566e-05, | |
| "loss": 15.7025, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 2.632034632034632, | |
| "grad_norm": 0.0022099835332483053, | |
| "learning_rate": 6.111198874944845e-05, | |
| "loss": 15.8339, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 2.6406926406926408, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.045294390555598e-05, | |
| "loss": 15.6495, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.6493506493506493, | |
| "grad_norm": 0.06401393562555313, | |
| "learning_rate": 5.979592895329085e-05, | |
| "loss": 15.9235, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 2.658008658008658, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.914097761625428e-05, | |
| "loss": 15.7719, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 0.0014950314071029425, | |
| "learning_rate": 5.848812351212522e-05, | |
| "loss": 15.8033, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 2.675324675324675, | |
| "grad_norm": 0.003303313162177801, | |
| "learning_rate": 5.783740015093484e-05, | |
| "loss": 15.8201, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 2.683982683982684, | |
| "grad_norm": 0.0021313403267413378, | |
| "learning_rate": 5.718884093334627e-05, | |
| "loss": 15.6961, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.6926406926406927, | |
| "grad_norm": 0.004075351171195507, | |
| "learning_rate": 5.654247914894058e-05, | |
| "loss": 15.8514, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 2.7012987012987013, | |
| "grad_norm": 0.0010667102178558707, | |
| "learning_rate": 5.589834797450764e-05, | |
| "loss": 15.9098, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 2.70995670995671, | |
| "grad_norm": 0.0020862577948719263, | |
| "learning_rate": 5.525648047234364e-05, | |
| "loss": 15.8231, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 2.7186147186147185, | |
| "grad_norm": 0.0027077968697994947, | |
| "learning_rate": 5.4616909588553674e-05, | |
| "loss": 15.7527, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 2.7272727272727275, | |
| "grad_norm": 0.0037543477956205606, | |
| "learning_rate": 5.3979668151360905e-05, | |
| "loss": 15.8286, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.7359307359307357, | |
| "grad_norm": 0.0034395060501992702, | |
| "learning_rate": 5.33447888694214e-05, | |
| "loss": 15.8517, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 2.7445887445887447, | |
| "grad_norm": 0.0016386689385399222, | |
| "learning_rate": 5.271230433014542e-05, | |
| "loss": 15.6296, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 2.7532467532467533, | |
| "grad_norm": 0.002882522065192461, | |
| "learning_rate": 5.2082246998024485e-05, | |
| "loss": 15.9626, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 2.761904761904762, | |
| "grad_norm": 0.0016715782694518566, | |
| "learning_rate": 5.145464921296537e-05, | |
| "loss": 16.1011, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 2.7705627705627704, | |
| "grad_norm": 0.003780187340453267, | |
| "learning_rate": 5.082954318862978e-05, | |
| "loss": 15.7561, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.779220779220779, | |
| "grad_norm": 0.0015209164703264832, | |
| "learning_rate": 5.0206961010781085e-05, | |
| "loss": 15.7211, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 2.787878787878788, | |
| "grad_norm": 0.003561074612662196, | |
| "learning_rate": 4.958693463563748e-05, | |
| "loss": 15.9192, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 2.7965367965367967, | |
| "grad_norm": 0.0025618516374379396, | |
| "learning_rate": 4.8969495888231484e-05, | |
| "loss": 15.959, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 2.8051948051948052, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.835467646077656e-05, | |
| "loss": 15.6299, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 2.813852813852814, | |
| "grad_norm": 0.0032659226562827826, | |
| "learning_rate": 4.7742507911040325e-05, | |
| "loss": 15.8226, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.8225108225108224, | |
| "grad_norm": 0.0029562402050942183, | |
| "learning_rate": 4.713302166072492e-05, | |
| "loss": 16.5945, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 2.8311688311688314, | |
| "grad_norm": 0.003705563023686409, | |
| "learning_rate": 4.652624899385387e-05, | |
| "loss": 15.8928, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 2.8398268398268396, | |
| "grad_norm": 0.002291543409228325, | |
| "learning_rate": 4.5922221055166656e-05, | |
| "loss": 15.6579, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 2.8484848484848486, | |
| "grad_norm": 0.00401789927855134, | |
| "learning_rate": 4.532096884851978e-05, | |
| "loss": 15.9641, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 0.003257931210100651, | |
| "learning_rate": 4.4722523235295745e-05, | |
| "loss": 15.6643, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.865800865800866, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.41269149328185e-05, | |
| "loss": 15.7199, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 2.8744588744588744, | |
| "grad_norm": 0.0021371468901634216, | |
| "learning_rate": 4.3534174512777324e-05, | |
| "loss": 15.8272, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 2.883116883116883, | |
| "grad_norm": 0.0009620334021747112, | |
| "learning_rate": 4.2944332399657184e-05, | |
| "loss": 16.1842, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 2.891774891774892, | |
| "grad_norm": 0.0020075475331395864, | |
| "learning_rate": 4.2357418869177354e-05, | |
| "loss": 15.6193, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 2.9004329004329006, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.1773464046737276e-05, | |
| "loss": 15.6615, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "grad_norm": 0.000680964847560972, | |
| "learning_rate": 4.1192497905870276e-05, | |
| "loss": 15.7517, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 2.9177489177489178, | |
| "grad_norm": 0.004122753627598286, | |
| "learning_rate": 4.061455026670509e-05, | |
| "loss": 16.6327, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 2.9264069264069263, | |
| "grad_norm": 0.002312893746420741, | |
| "learning_rate": 4.0039650794435344e-05, | |
| "loss": 15.8981, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 2.935064935064935, | |
| "grad_norm": 0.024043424054980278, | |
| "learning_rate": 3.946782899779667e-05, | |
| "loss": 15.8999, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 2.9437229437229435, | |
| "grad_norm": 0.0012768743326887488, | |
| "learning_rate": 3.889911422755231e-05, | |
| "loss": 15.6874, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.9523809523809526, | |
| "grad_norm": 0.0023466376587748528, | |
| "learning_rate": 3.8333535674986275e-05, | |
| "loss": 15.7274, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 2.961038961038961, | |
| "grad_norm": 0.0034092010464519262, | |
| "learning_rate": 3.777112237040537e-05, | |
| "loss": 15.7464, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 2.9696969696969697, | |
| "grad_norm": 0.0017824557144194841, | |
| "learning_rate": 3.721190318164877e-05, | |
| "loss": 15.7556, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 2.9783549783549783, | |
| "grad_norm": 0.0014683044282719493, | |
| "learning_rate": 3.665590681260658e-05, | |
| "loss": 16.0187, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 2.987012987012987, | |
| "grad_norm": 0.00010432133422000334, | |
| "learning_rate": 3.610316180174622e-05, | |
| "loss": 16.4009, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.995670995670996, | |
| "grad_norm": 0.003945178352296352, | |
| "learning_rate": 3.555369652064787e-05, | |
| "loss": 15.7738, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 3.0043290043290045, | |
| "grad_norm": 0.0048090131022036076, | |
| "learning_rate": 3.500753917254787e-05, | |
| "loss": 15.9384, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 3.012987012987013, | |
| "grad_norm": 0.002933148993179202, | |
| "learning_rate": 3.446471779089144e-05, | |
| "loss": 15.8317, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 3.0216450216450217, | |
| "grad_norm": 0.003420398337766528, | |
| "learning_rate": 3.392526023789349e-05, | |
| "loss": 15.536, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 3.0303030303030303, | |
| "grad_norm": 0.0001099475848604925, | |
| "learning_rate": 3.338919420310871e-05, | |
| "loss": 15.5597, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.038961038961039, | |
| "grad_norm": 0.00021568694501183927, | |
| "learning_rate": 3.28565472020101e-05, | |
| "loss": 15.603, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 3.0476190476190474, | |
| "grad_norm": 0.003983495756983757, | |
| "learning_rate": 3.2327346574576753e-05, | |
| "loss": 15.5045, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 3.0562770562770565, | |
| "grad_norm": 0.00216678692959249, | |
| "learning_rate": 3.180161948389062e-05, | |
| "loss": 15.5247, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 3.064935064935065, | |
| "grad_norm": 0.031781021505594254, | |
| "learning_rate": 3.1279392914742046e-05, | |
| "loss": 15.4569, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 3.0735930735930737, | |
| "grad_norm": 0.00032434993772767484, | |
| "learning_rate": 3.076069367224486e-05, | |
| "loss": 16.0896, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 3.0822510822510822, | |
| "grad_norm": 0.0037430988159030676, | |
| "learning_rate": 3.0245548380460486e-05, | |
| "loss": 15.6038, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 3.090909090909091, | |
| "grad_norm": 0.0002255355502711609, | |
| "learning_rate": 2.9733983481031302e-05, | |
| "loss": 15.3307, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 3.0995670995670994, | |
| "grad_norm": 0.0028247262816876173, | |
| "learning_rate": 2.922602523182344e-05, | |
| "loss": 15.4397, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 3.108225108225108, | |
| "grad_norm": 0.018052997067570686, | |
| "learning_rate": 2.872169970557913e-05, | |
| "loss": 15.599, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 3.116883116883117, | |
| "grad_norm": 0.0025733679067343473, | |
| "learning_rate": 2.8221032788578205e-05, | |
| "loss": 15.4745, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 3.1255411255411256, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.7724050179309646e-05, | |
| "loss": 15.9485, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 3.134199134199134, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.7230777387152296e-05, | |
| "loss": 15.4227, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 3.142857142857143, | |
| "grad_norm": 0.000987619161605835, | |
| "learning_rate": 2.6741239731065647e-05, | |
| "loss": 15.5741, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 3.1515151515151514, | |
| "grad_norm": 0.0019787976052612066, | |
| "learning_rate": 2.625546233829016e-05, | |
| "loss": 16.6376, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 3.16017316017316, | |
| "grad_norm": 0.0037663017865270376, | |
| "learning_rate": 2.5773470143057655e-05, | |
| "loss": 15.5158, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 3.168831168831169, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.529528788531128e-05, | |
| "loss": 15.4871, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 3.1774891774891776, | |
| "grad_norm": 0.0023285530041903257, | |
| "learning_rate": 2.4820940109435885e-05, | |
| "loss": 15.4632, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 3.186147186147186, | |
| "grad_norm": 0.0020200940780341625, | |
| "learning_rate": 2.4350451162997877e-05, | |
| "loss": 15.5132, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 3.1948051948051948, | |
| "grad_norm": 0.003936352673918009, | |
| "learning_rate": 2.3883845195495878e-05, | |
| "loss": 15.5417, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 3.2034632034632033, | |
| "grad_norm": 0.0037546472158282995, | |
| "learning_rate": 2.342114615712081e-05, | |
| "loss": 15.9803, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 3.212121212121212, | |
| "grad_norm": 0.0005508167669177055, | |
| "learning_rate": 2.296237779752687e-05, | |
| "loss": 15.2634, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 3.220779220779221, | |
| "grad_norm": 0.0036611163523048162, | |
| "learning_rate": 2.2507563664612252e-05, | |
| "loss": 15.5582, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 3.2294372294372296, | |
| "grad_norm": 0.003752505173906684, | |
| "learning_rate": 2.205672710331059e-05, | |
| "loss": 15.5934, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 3.238095238095238, | |
| "grad_norm": 0.0008896394865587354, | |
| "learning_rate": 2.1609891254392678e-05, | |
| "loss": 15.4942, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 3.2467532467532467, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.1167079053278737e-05, | |
| "loss": 15.3308, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.2554112554112553, | |
| "grad_norm": 0.00010732792725320905, | |
| "learning_rate": 2.072831322886105e-05, | |
| "loss": 15.3753, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 3.264069264069264, | |
| "grad_norm": 0.003391178324818611, | |
| "learning_rate": 2.029361630233747e-05, | |
| "loss": 15.4389, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 3.2727272727272725, | |
| "grad_norm": 0.0003303180856164545, | |
| "learning_rate": 1.986301058605531e-05, | |
| "loss": 15.385, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 3.2813852813852815, | |
| "grad_norm": 0.0005520053091458976, | |
| "learning_rate": 1.9436518182366158e-05, | |
| "loss": 15.4051, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 3.29004329004329, | |
| "grad_norm": 0.003017255337908864, | |
| "learning_rate": 1.901416098249136e-05, | |
| "loss": 15.647, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 3.2987012987012987, | |
| "grad_norm": 0.021177958697080612, | |
| "learning_rate": 1.8595960665398458e-05, | |
| "loss": 15.6138, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 3.3073593073593073, | |
| "grad_norm": 0.0034116676542907953, | |
| "learning_rate": 1.8181938696688296e-05, | |
| "loss": 15.4569, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 3.316017316017316, | |
| "grad_norm": 0.00011423804244259372, | |
| "learning_rate": 1.7772116327493372e-05, | |
| "loss": 15.4194, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 3.324675324675325, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.736651459338695e-05, | |
| "loss": 15.4586, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 3.3333333333333335, | |
| "grad_norm": 0.001321481540799141, | |
| "learning_rate": 1.6965154313303368e-05, | |
| "loss": 15.613, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 3.341991341991342, | |
| "grad_norm": 0.0015613286523148417, | |
| "learning_rate": 1.6568056088469387e-05, | |
| "loss": 15.4395, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 3.3506493506493507, | |
| "grad_norm": 0.00010790528176585212, | |
| "learning_rate": 1.6175240301346906e-05, | |
| "loss": 15.4752, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 3.3593073593073592, | |
| "grad_norm": 0.003342802170664072, | |
| "learning_rate": 1.5786727114586586e-05, | |
| "loss": 15.4756, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 3.367965367965368, | |
| "grad_norm": 0.001664191484451294, | |
| "learning_rate": 1.540253646999299e-05, | |
| "loss": 15.4152, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 3.3766233766233764, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.5022688087501092e-05, | |
| "loss": 15.3462, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 3.3852813852813854, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.4647201464163906e-05, | |
| "loss": 15.4894, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 3.393939393939394, | |
| "grad_norm": 0.002181015908718109, | |
| "learning_rate": 1.4276095873151952e-05, | |
| "loss": 15.4537, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 3.4025974025974026, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.3909390362763752e-05, | |
| "loss": 15.4039, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 3.411255411255411, | |
| "grad_norm": 0.001633924781344831, | |
| "learning_rate": 1.3547103755448287e-05, | |
| "loss": 15.3927, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 3.41991341991342, | |
| "grad_norm": 0.0019662026315927505, | |
| "learning_rate": 1.3189254646838767e-05, | |
| "loss": 15.573, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 3.4285714285714284, | |
| "grad_norm": 0.0017789127305150032, | |
| "learning_rate": 1.2835861404798265e-05, | |
| "loss": 15.4386, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 3.4372294372294374, | |
| "grad_norm": 0.0037333201617002487, | |
| "learning_rate": 1.2486942168476756e-05, | |
| "loss": 15.4277, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 3.445887445887446, | |
| "grad_norm": 0.00033077617990784347, | |
| "learning_rate": 1.2142514847380237e-05, | |
| "loss": 15.43, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 3.4545454545454546, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.1802597120451286e-05, | |
| "loss": 15.5442, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 3.463203463203463, | |
| "grad_norm": 0.0014221479650586843, | |
| "learning_rate": 1.146720643516177e-05, | |
| "loss": 15.4446, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.4718614718614718, | |
| "grad_norm": 0.0004546408890746534, | |
| "learning_rate": 1.1136360006617185e-05, | |
| "loss": 16.2801, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 3.4805194805194803, | |
| "grad_norm": 0.0038407333195209503, | |
| "learning_rate": 1.0810074816673154e-05, | |
| "loss": 26.5808, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 3.4891774891774894, | |
| "grad_norm": 0.000989666790701449, | |
| "learning_rate": 1.048836761306361e-05, | |
| "loss": 15.379, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 3.497835497835498, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.0171254908541372e-05, | |
| "loss": 15.4584, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 3.5064935064935066, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.858752980030295e-06, | |
| "loss": 15.609, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 3.515151515151515, | |
| "grad_norm": 0.0013323862804099917, | |
| "learning_rate": 9.550877867790065e-06, | |
| "loss": 15.5316, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 3.5238095238095237, | |
| "grad_norm": 0.00010888870747294277, | |
| "learning_rate": 9.247645374592717e-06, | |
| "loss": 15.4492, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 3.5324675324675323, | |
| "grad_norm": 0.00010042625217465684, | |
| "learning_rate": 8.949071064911585e-06, | |
| "loss": 15.4255, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 3.541125541125541, | |
| "grad_norm": 0.0003212654555682093, | |
| "learning_rate": 8.655170264122303e-06, | |
| "loss": 15.9795, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 3.54978354978355, | |
| "grad_norm": 0.003979097120463848, | |
| "learning_rate": 8.365958057716338e-06, | |
| "loss": 15.5248, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 3.5584415584415585, | |
| "grad_norm": 0.0019642910920083523, | |
| "learning_rate": 8.081449290526432e-06, | |
| "loss": 15.4587, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 3.567099567099567, | |
| "grad_norm": 0.008268770761787891, | |
| "learning_rate": 7.80165856596492e-06, | |
| "loss": 15.7592, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 3.5757575757575757, | |
| "grad_norm": 0.0025506443344056606, | |
| "learning_rate": 7.526600245273918e-06, | |
| "loss": 15.8564, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 3.5844155844155843, | |
| "grad_norm": 0.0012161307968199253, | |
| "learning_rate": 7.256288446788362e-06, | |
| "loss": 15.3659, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 3.5930735930735933, | |
| "grad_norm": 0.004051461815834045, | |
| "learning_rate": 6.9907370452112046e-06, | |
| "loss": 15.9221, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 3.601731601731602, | |
| "grad_norm": 0.003108682343736291, | |
| "learning_rate": 6.729959670901309e-06, | |
| "loss": 15.5334, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 3.6103896103896105, | |
| "grad_norm": 0.0033409043680876493, | |
| "learning_rate": 6.4739697091738e-06, | |
| "loss": 15.426, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 3.619047619047619, | |
| "grad_norm": 0.0016489011468365788, | |
| "learning_rate": 6.222780299613074e-06, | |
| "loss": 15.3453, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 3.6277056277056277, | |
| "grad_norm": 0.00321365287527442, | |
| "learning_rate": 5.976404335398256e-06, | |
| "loss": 15.607, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 3.6363636363636362, | |
| "grad_norm": 0.00022275045921560377, | |
| "learning_rate": 5.734854462641548e-06, | |
| "loss": 15.4681, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 3.645021645021645, | |
| "grad_norm": 0.0018924333853647113, | |
| "learning_rate": 5.498143079738971e-06, | |
| "loss": 15.4828, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 3.653679653679654, | |
| "grad_norm": 0.0013136648340150714, | |
| "learning_rate": 5.2662823367340855e-06, | |
| "loss": 15.2643, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 3.6623376623376624, | |
| "grad_norm": 0.00011183915194123983, | |
| "learning_rate": 5.039284134694333e-06, | |
| "loss": 15.4385, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 3.670995670995671, | |
| "grad_norm": 0.0032093904446810484, | |
| "learning_rate": 4.817160125100106e-06, | |
| "loss": 15.4845, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 3.6796536796536796, | |
| "grad_norm": 0.001872088760137558, | |
| "learning_rate": 4.599921709246812e-06, | |
| "loss": 15.3745, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 3.688311688311688, | |
| "grad_norm": 0.002430541208013892, | |
| "learning_rate": 4.3875800376595e-06, | |
| "loss": 15.3891, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 3.6969696969696972, | |
| "grad_norm": 0.003143192734569311, | |
| "learning_rate": 4.180146009520702e-06, | |
| "loss": 15.6386, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 3.7056277056277054, | |
| "grad_norm": 0.003909118473529816, | |
| "learning_rate": 3.977630272110811e-06, | |
| "loss": 15.8103, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 3.7142857142857144, | |
| "grad_norm": 0.00011017678480129689, | |
| "learning_rate": 3.780043220261764e-06, | |
| "loss": 15.4262, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 3.722943722943723, | |
| "grad_norm": 0.002079217229038477, | |
| "learning_rate": 3.587394995823301e-06, | |
| "loss": 15.5448, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 3.7316017316017316, | |
| "grad_norm": 0.00011921657278435305, | |
| "learning_rate": 3.3996954871425845e-06, | |
| "loss": 15.3458, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 3.74025974025974, | |
| "grad_norm": 0.0003392777871340513, | |
| "learning_rate": 3.216954328556443e-06, | |
| "loss": 15.4115, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 3.7489177489177488, | |
| "grad_norm": 0.0011154324747622013, | |
| "learning_rate": 3.039180899897043e-06, | |
| "loss": 15.5333, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 3.757575757575758, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.8663843260103074e-06, | |
| "loss": 15.4002, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 3.7662337662337664, | |
| "grad_norm": 0.00021322118118405342, | |
| "learning_rate": 2.698573476287658e-06, | |
| "loss": 15.634, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 3.774891774891775, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.535756964210634e-06, | |
| "loss": 15.2512, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 3.7835497835497836, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.37794314690889e-06, | |
| "loss": 15.5646, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 3.792207792207792, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.225140124731151e-06, | |
| "loss": 15.436, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 3.8008658008658007, | |
| "grad_norm": 0.0018814082723110914, | |
| "learning_rate": 2.0773557408295343e-06, | |
| "loss": 15.5011, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 3.8095238095238093, | |
| "grad_norm": 0.0019670824985951185, | |
| "learning_rate": 1.9345975807568474e-06, | |
| "loss": 15.3244, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 3.8181818181818183, | |
| "grad_norm": 0.0035723568871617317, | |
| "learning_rate": 1.7968729720773459e-06, | |
| "loss": 15.6439, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 3.826839826839827, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.6641889839905445e-06, | |
| "loss": 15.3774, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 3.8354978354978355, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.536552426968396e-06, | |
| "loss": 15.3533, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 3.844155844155844, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.4139698524057165e-06, | |
| "loss": 16.2874, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 3.8528138528138527, | |
| "grad_norm": 0.00011134289525216445, | |
| "learning_rate": 1.2964475522839304e-06, | |
| "loss": 15.4719, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 3.8614718614718617, | |
| "grad_norm": 0.00011475420615170151, | |
| "learning_rate": 1.1839915588480743e-06, | |
| "loss": 15.2561, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 3.87012987012987, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.0766076442971895e-06, | |
| "loss": 15.4674, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 3.878787878787879, | |
| "grad_norm": 0.0022935173474252224, | |
| "learning_rate": 9.74301320488058e-07, | |
| "loss": 15.6186, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 3.8874458874458875, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.770778386522627e-07, | |
| "loss": 15.3455, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 3.896103896103896, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.849421891266584e-07, | |
| "loss": 15.4871, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.9047619047619047, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.978991010972547e-07, | |
| "loss": 16.203, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 3.9134199134199132, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.159530423563986e-07, | |
| "loss": 15.4409, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 3.9220779220779223, | |
| "grad_norm": 0.004353491589426994, | |
| "learning_rate": 5.391082190735252e-07, | |
| "loss": 15.4641, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 3.930735930735931, | |
| "grad_norm": 0.0022043841890990734, | |
| "learning_rate": 4.6736857557925227e-07, | |
| "loss": 15.527, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 3.9393939393939394, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.007377941628754e-07, | |
| "loss": 15.3882, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 3.948051948051948, | |
| "grad_norm": 0.0020064269192516804, | |
| "learning_rate": 3.392192948833861e-07, | |
| "loss": 15.5124, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 3.9567099567099566, | |
| "grad_norm": 0.00176339247263968, | |
| "learning_rate": 2.828162353939678e-07, | |
| "loss": 15.3776, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 3.965367965367965, | |
| "grad_norm": 0.0016750121721997857, | |
| "learning_rate": 2.315315107798366e-07, | |
| "loss": 15.4065, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 3.974025974025974, | |
| "grad_norm": 0.0027372916229069233, | |
| "learning_rate": 1.8536775340970425e-07, | |
| "loss": 15.5642, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 3.982683982683983, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.4432733280065335e-07, | |
| "loss": 15.3594, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 3.9913419913419914, | |
| "grad_norm": 0.0021978975273668766, | |
| "learning_rate": 1.0841235549648999e-07, | |
| "loss": 15.4857, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.25208330154418945, | |
| "learning_rate": 7.762466495964127e-08, | |
| "loss": 15.4448, | |
| "step": 924 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 924, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 200, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.3868198638796145e+18, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |