| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.96, | |
| "eval_steps": 500, | |
| "global_step": 186, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.016, | |
| "grad_norm": 2.7748332351978715, | |
| "learning_rate": 5.263157894736843e-07, | |
| "loss": 0.9853, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.032, | |
| "grad_norm": 2.6630536083631458, | |
| "learning_rate": 1.0526315789473685e-06, | |
| "loss": 0.9283, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.048, | |
| "grad_norm": 2.927669610212838, | |
| "learning_rate": 1.5789473684210526e-06, | |
| "loss": 0.9314, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 3.374580493902324, | |
| "learning_rate": 2.105263157894737e-06, | |
| "loss": 0.8663, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 2.9896347917672914, | |
| "learning_rate": 2.631578947368421e-06, | |
| "loss": 0.9499, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 2.582577348790655, | |
| "learning_rate": 3.157894736842105e-06, | |
| "loss": 0.9735, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.112, | |
| "grad_norm": 1.8138211839586487, | |
| "learning_rate": 3.6842105263157896e-06, | |
| "loss": 0.8656, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 1.6587738524167377, | |
| "learning_rate": 4.210526315789474e-06, | |
| "loss": 0.8682, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.144, | |
| "grad_norm": 1.8109531448533194, | |
| "learning_rate": 4.736842105263158e-06, | |
| "loss": 0.8689, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.7363401935534606, | |
| "learning_rate": 5.263157894736842e-06, | |
| "loss": 0.8002, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.176, | |
| "grad_norm": 1.582046063748505, | |
| "learning_rate": 5.789473684210527e-06, | |
| "loss": 0.7844, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 1.63075244214434, | |
| "learning_rate": 6.31578947368421e-06, | |
| "loss": 0.8482, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.208, | |
| "grad_norm": 1.5695854938299345, | |
| "learning_rate": 6.842105263157896e-06, | |
| "loss": 0.815, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.224, | |
| "grad_norm": 1.6393959664209934, | |
| "learning_rate": 7.368421052631579e-06, | |
| "loss": 0.7927, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 1.427405122602582, | |
| "learning_rate": 7.894736842105265e-06, | |
| "loss": 0.7053, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 1.190798329611444, | |
| "learning_rate": 8.421052631578948e-06, | |
| "loss": 0.785, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.272, | |
| "grad_norm": 1.1237865106813079, | |
| "learning_rate": 8.947368421052632e-06, | |
| "loss": 0.7322, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 1.2850912169881596, | |
| "learning_rate": 9.473684210526315e-06, | |
| "loss": 0.7699, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.304, | |
| "grad_norm": 1.2404415303944918, | |
| "learning_rate": 1e-05, | |
| "loss": 0.7549, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 1.1432542313544858, | |
| "learning_rate": 9.999115304121459e-06, | |
| "loss": 0.7357, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.336, | |
| "grad_norm": 0.9982349176673481, | |
| "learning_rate": 9.996461529560553e-06, | |
| "loss": 0.6744, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.352, | |
| "grad_norm": 1.1965633471703596, | |
| "learning_rate": 9.992039615430648e-06, | |
| "loss": 0.6811, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.368, | |
| "grad_norm": 1.0999497983388995, | |
| "learning_rate": 9.985851126551428e-06, | |
| "loss": 0.7172, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 1.1136747331466013, | |
| "learning_rate": 9.977898252895133e-06, | |
| "loss": 0.7045, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 1.07004844918576, | |
| "learning_rate": 9.968183808811586e-06, | |
| "loss": 0.6501, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.416, | |
| "grad_norm": 1.0735349838789574, | |
| "learning_rate": 9.95671123203224e-06, | |
| "loss": 0.7066, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.432, | |
| "grad_norm": 1.084177002259536, | |
| "learning_rate": 9.943484582453653e-06, | |
| "loss": 0.6679, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 0.951431009849941, | |
| "learning_rate": 9.928508540700775e-06, | |
| "loss": 0.6399, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.464, | |
| "grad_norm": 0.9518294170466356, | |
| "learning_rate": 9.91178840647057e-06, | |
| "loss": 0.6646, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 1.0529394198621256, | |
| "learning_rate": 9.893330096656576e-06, | |
| "loss": 0.6284, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.496, | |
| "grad_norm": 0.9490390276661018, | |
| "learning_rate": 9.873140143255035e-06, | |
| "loss": 0.7074, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 1.0534268920005938, | |
| "learning_rate": 9.851225691053382e-06, | |
| "loss": 0.6585, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.528, | |
| "grad_norm": 1.0901867976699247, | |
| "learning_rate": 9.827594495101824e-06, | |
| "loss": 0.715, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.544, | |
| "grad_norm": 1.002965926619648, | |
| "learning_rate": 9.802254917969033e-06, | |
| "loss": 0.7039, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.7942186361723101, | |
| "learning_rate": 9.775215926782788e-06, | |
| "loss": 0.6193, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 1.042170177689746, | |
| "learning_rate": 9.746487090056712e-06, | |
| "loss": 0.6329, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.592, | |
| "grad_norm": 0.913970320031391, | |
| "learning_rate": 9.71607857430419e-06, | |
| "loss": 0.7023, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.608, | |
| "grad_norm": 0.8382364002272965, | |
| "learning_rate": 9.68400114044064e-06, | |
| "loss": 0.6785, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.624, | |
| "grad_norm": 0.8450560688480601, | |
| "learning_rate": 9.650266139975474e-06, | |
| "loss": 0.7239, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.8653852042607058, | |
| "learning_rate": 9.614885510995047e-06, | |
| "loss": 0.5876, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.656, | |
| "grad_norm": 1.072572316229336, | |
| "learning_rate": 9.577871773938013e-06, | |
| "loss": 0.6393, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.672, | |
| "grad_norm": 1.0153801864076717, | |
| "learning_rate": 9.539238027164618e-06, | |
| "loss": 0.6664, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.688, | |
| "grad_norm": 0.8319221486785371, | |
| "learning_rate": 9.498997942321484e-06, | |
| "loss": 0.6335, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 0.9932505623421072, | |
| "learning_rate": 9.457165759503492e-06, | |
| "loss": 0.667, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.9270065267290504, | |
| "learning_rate": 9.413756282214538e-06, | |
| "loss": 0.5831, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.736, | |
| "grad_norm": 0.9284213958952939, | |
| "learning_rate": 9.368784872128877e-06, | |
| "loss": 0.6141, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.752, | |
| "grad_norm": 0.9253794529400015, | |
| "learning_rate": 9.322267443654974e-06, | |
| "loss": 0.6159, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 1.0458282286459863, | |
| "learning_rate": 9.274220458303727e-06, | |
| "loss": 0.6229, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.784, | |
| "grad_norm": 0.8683602588823571, | |
| "learning_rate": 9.224660918863104e-06, | |
| "loss": 0.7051, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.8270896237543336, | |
| "learning_rate": 9.173606363381218e-06, | |
| "loss": 0.6346, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.816, | |
| "grad_norm": 0.9613549642577978, | |
| "learning_rate": 9.121074858959997e-06, | |
| "loss": 0.5842, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 0.8260856239263705, | |
| "learning_rate": 9.067084995361623e-06, | |
| "loss": 0.6479, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.848, | |
| "grad_norm": 0.9873032135511416, | |
| "learning_rate": 9.011655878430018e-06, | |
| "loss": 0.6301, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.864, | |
| "grad_norm": 0.9327430249002449, | |
| "learning_rate": 8.954807123329703e-06, | |
| "loss": 0.6817, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.9173609434575164, | |
| "learning_rate": 8.896558847604414e-06, | |
| "loss": 0.6299, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 0.9866716369750068, | |
| "learning_rate": 8.836931664057935e-06, | |
| "loss": 0.7217, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.912, | |
| "grad_norm": 0.8783809839243962, | |
| "learning_rate": 8.775946673459682e-06, | |
| "loss": 0.6144, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.928, | |
| "grad_norm": 0.9061115444020617, | |
| "learning_rate": 8.713625457077585e-06, | |
| "loss": 0.6225, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.944, | |
| "grad_norm": 0.8771983055721709, | |
| "learning_rate": 8.64999006904096e-06, | |
| "loss": 0.6165, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.9050842567508385, | |
| "learning_rate": 8.585063028536015e-06, | |
| "loss": 0.6065, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.976, | |
| "grad_norm": 0.9089868204229448, | |
| "learning_rate": 8.518867311836808e-06, | |
| "loss": 0.614, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.992, | |
| "grad_norm": 0.871550073504428, | |
| "learning_rate": 8.451426344174433e-06, | |
| "loss": 0.6314, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.871550073504428, | |
| "learning_rate": 8.382763991447344e-06, | |
| "loss": 0.5972, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.016, | |
| "grad_norm": 1.4237650455434694, | |
| "learning_rate": 8.312904551775731e-06, | |
| "loss": 0.6036, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.032, | |
| "grad_norm": 0.8124823867763251, | |
| "learning_rate": 8.241872746902934e-06, | |
| "loss": 0.6267, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.048, | |
| "grad_norm": 0.9141373183955689, | |
| "learning_rate": 8.16969371344696e-06, | |
| "loss": 0.5986, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.064, | |
| "grad_norm": 0.9354598699036616, | |
| "learning_rate": 8.096392994005177e-06, | |
| "loss": 0.5421, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 0.8024989779166635, | |
| "learning_rate": 8.021996528115335e-06, | |
| "loss": 0.5397, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.096, | |
| "grad_norm": 0.9253483651783203, | |
| "learning_rate": 7.946530643076138e-06, | |
| "loss": 0.4288, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.112, | |
| "grad_norm": 0.8802625033731535, | |
| "learning_rate": 7.870022044630569e-06, | |
| "loss": 0.547, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.1280000000000001, | |
| "grad_norm": 0.7822575611704824, | |
| "learning_rate": 7.792497807515317e-06, | |
| "loss": 0.4986, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.144, | |
| "grad_norm": 1.0261745477766617, | |
| "learning_rate": 7.713985365879607e-06, | |
| "loss": 0.4792, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 1.0116853901737157, | |
| "learning_rate": 7.63451250357685e-06, | |
| "loss": 0.541, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.176, | |
| "grad_norm": 0.8405415371571965, | |
| "learning_rate": 7.55410734433254e-06, | |
| "loss": 0.5147, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.192, | |
| "grad_norm": 1.0200272567701543, | |
| "learning_rate": 7.472798341791877e-06, | |
| "loss": 0.4497, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.208, | |
| "grad_norm": 0.8892024921601165, | |
| "learning_rate": 7.390614269450633e-06, | |
| "loss": 0.5398, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.224, | |
| "grad_norm": 0.9409235566252042, | |
| "learning_rate": 7.3075842104728445e-06, | |
| "loss": 0.4672, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 0.9995872856448315, | |
| "learning_rate": 7.223737547398898e-06, | |
| "loss": 0.4854, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.256, | |
| "grad_norm": 0.9570196272561535, | |
| "learning_rate": 7.139103951747694e-06, | |
| "loss": 0.5084, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.272, | |
| "grad_norm": 0.9773083133145911, | |
| "learning_rate": 7.053713373516538e-06, | |
| "loss": 0.5008, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.288, | |
| "grad_norm": 0.8973949006130701, | |
| "learning_rate": 6.9675960305824785e-06, | |
| "loss": 0.5378, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.304, | |
| "grad_norm": 0.8882666692574759, | |
| "learning_rate": 6.880782398008862e-06, | |
| "loss": 0.4967, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 0.8220336644812528, | |
| "learning_rate": 6.7933031972608644e-06, | |
| "loss": 0.5471, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.336, | |
| "grad_norm": 0.8535399851197573, | |
| "learning_rate": 6.70518938533383e-06, | |
| "loss": 0.5205, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.3519999999999999, | |
| "grad_norm": 0.8511047440970223, | |
| "learning_rate": 6.61647214379826e-06, | |
| "loss": 0.5774, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.3679999999999999, | |
| "grad_norm": 0.884532376967301, | |
| "learning_rate": 6.527182867765333e-06, | |
| "loss": 0.5073, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.384, | |
| "grad_norm": 0.8914241759089955, | |
| "learning_rate": 6.437353154776848e-06, | |
| "loss": 0.5466, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 0.9213389785402387, | |
| "learning_rate": 6.3470147936235485e-06, | |
| "loss": 0.5121, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.416, | |
| "grad_norm": 0.890133974816951, | |
| "learning_rate": 6.256199753095745e-06, | |
| "loss": 0.5628, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.432, | |
| "grad_norm": 0.8722854321596075, | |
| "learning_rate": 6.164940170670266e-06, | |
| "loss": 0.4931, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.448, | |
| "grad_norm": 1.0164650913815605, | |
| "learning_rate": 6.073268341137694e-06, | |
| "loss": 0.5104, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.464, | |
| "grad_norm": 0.9929986897473667, | |
| "learning_rate": 5.98121670517393e-06, | |
| "loss": 0.5363, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 0.8060891530527672, | |
| "learning_rate": 5.8888178378601565e-06, | |
| "loss": 0.526, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.496, | |
| "grad_norm": 0.8485720494109616, | |
| "learning_rate": 5.796104437155213e-06, | |
| "loss": 0.5064, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.512, | |
| "grad_norm": 0.8817067952044154, | |
| "learning_rate": 5.703109312324493e-06, | |
| "loss": 0.5216, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.528, | |
| "grad_norm": 0.9855641527444552, | |
| "learning_rate": 5.609865372329461e-06, | |
| "loss": 0.6199, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.544, | |
| "grad_norm": 0.9317605309596566, | |
| "learning_rate": 5.516405614181883e-06, | |
| "loss": 0.6119, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 0.7950706009561465, | |
| "learning_rate": 5.4227631112668955e-06, | |
| "loss": 0.6234, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.576, | |
| "grad_norm": 1.0154251250334365, | |
| "learning_rate": 5.328971001639054e-06, | |
| "loss": 0.562, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.592, | |
| "grad_norm": 0.8207053782635244, | |
| "learning_rate": 5.235062476295488e-06, | |
| "loss": 0.534, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.608, | |
| "grad_norm": 1.0080473039037783, | |
| "learning_rate": 5.141070767430331e-06, | |
| "loss": 0.5062, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.624, | |
| "grad_norm": 0.8101777785720703, | |
| "learning_rate": 5.047029136674563e-06, | |
| "loss": 0.5635, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.6400000000000001, | |
| "grad_norm": 0.9057216986339542, | |
| "learning_rate": 4.95297086332544e-06, | |
| "loss": 0.5301, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.6560000000000001, | |
| "grad_norm": 1.210413530518266, | |
| "learning_rate": 4.858929232569671e-06, | |
| "loss": 0.515, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.6720000000000002, | |
| "grad_norm": 0.8617855937822649, | |
| "learning_rate": 4.7649375237045135e-06, | |
| "loss": 0.5229, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.688, | |
| "grad_norm": 0.9101134842850922, | |
| "learning_rate": 4.671028998360947e-06, | |
| "loss": 0.5661, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.704, | |
| "grad_norm": 0.917819145086472, | |
| "learning_rate": 4.5772368887331044e-06, | |
| "loss": 0.5354, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 0.976693306348323, | |
| "learning_rate": 4.483594385818119e-06, | |
| "loss": 0.4019, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.736, | |
| "grad_norm": 1.009333803845179, | |
| "learning_rate": 4.39013462767054e-06, | |
| "loss": 0.5953, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.752, | |
| "grad_norm": 0.8937905687228984, | |
| "learning_rate": 4.29689068767551e-06, | |
| "loss": 0.4911, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.768, | |
| "grad_norm": 1.0231122186706687, | |
| "learning_rate": 4.203895562844789e-06, | |
| "loss": 0.5066, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.784, | |
| "grad_norm": 0.9796114354432495, | |
| "learning_rate": 4.111182162139844e-06, | |
| "loss": 0.5046, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.8425867656857713, | |
| "learning_rate": 4.018783294826071e-06, | |
| "loss": 0.4494, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.8159999999999998, | |
| "grad_norm": 0.8042379622677128, | |
| "learning_rate": 3.926731658862307e-06, | |
| "loss": 0.5334, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.8319999999999999, | |
| "grad_norm": 0.862253265081445, | |
| "learning_rate": 3.8350598293297345e-06, | |
| "loss": 0.5397, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.8479999999999999, | |
| "grad_norm": 0.8675485758847044, | |
| "learning_rate": 3.7438002469042567e-06, | |
| "loss": 0.4379, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.8639999999999999, | |
| "grad_norm": 0.8641674873660793, | |
| "learning_rate": 3.652985206376455e-06, | |
| "loss": 0.5083, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 0.7597156428333421, | |
| "learning_rate": 3.5626468452231534e-06, | |
| "loss": 0.5049, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.896, | |
| "grad_norm": 0.9093228924067716, | |
| "learning_rate": 3.472817132234669e-06, | |
| "loss": 0.5667, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.912, | |
| "grad_norm": 0.8475529841588751, | |
| "learning_rate": 3.3835278562017405e-06, | |
| "loss": 0.5066, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.928, | |
| "grad_norm": 0.8614156796252268, | |
| "learning_rate": 3.29481061466617e-06, | |
| "loss": 0.4796, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.944, | |
| "grad_norm": 0.9468246699152438, | |
| "learning_rate": 3.2066968027391377e-06, | |
| "loss": 0.4985, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 0.883456282712155, | |
| "learning_rate": 3.119217601991139e-06, | |
| "loss": 0.4587, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.976, | |
| "grad_norm": 0.8878398723281332, | |
| "learning_rate": 3.032403969417523e-06, | |
| "loss": 0.438, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.992, | |
| "grad_norm": 0.9784220145562418, | |
| "learning_rate": 2.946286626483463e-06, | |
| "loss": 0.4603, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.4073288270403166, | |
| "learning_rate": 2.8608960482523058e-06, | |
| "loss": 0.506, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.016, | |
| "grad_norm": 0.9595010032920018, | |
| "learning_rate": 2.776262452601104e-06, | |
| "loss": 0.4436, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.032, | |
| "grad_norm": 0.949658526999288, | |
| "learning_rate": 2.6924157895271563e-06, | |
| "loss": 0.462, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 0.9666333447707621, | |
| "learning_rate": 2.6093857305493666e-06, | |
| "loss": 0.4249, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.064, | |
| "grad_norm": 0.8225046423522504, | |
| "learning_rate": 2.5272016582081236e-06, | |
| "loss": 0.4569, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 0.937248675700986, | |
| "learning_rate": 2.445892655667462e-06, | |
| "loss": 0.4072, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.096, | |
| "grad_norm": 0.9530017720092098, | |
| "learning_rate": 2.365487496423152e-06, | |
| "loss": 0.4264, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.9384891289059365, | |
| "learning_rate": 2.2860146341203936e-06, | |
| "loss": 0.3988, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.128, | |
| "grad_norm": 0.9643383803092408, | |
| "learning_rate": 2.207502192484685e-06, | |
| "loss": 0.3552, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.144, | |
| "grad_norm": 1.0265863375174316, | |
| "learning_rate": 2.1299779553694323e-06, | |
| "loss": 0.4725, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 0.9898914299255372, | |
| "learning_rate": 2.053469356923865e-06, | |
| "loss": 0.5312, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.8576428031309623, | |
| "learning_rate": 1.9780034718846653e-06, | |
| "loss": 0.4683, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.192, | |
| "grad_norm": 0.8308846534609936, | |
| "learning_rate": 1.9036070059948253e-06, | |
| "loss": 0.4363, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.208, | |
| "grad_norm": 0.875002319404404, | |
| "learning_rate": 1.8303062865530407e-06, | |
| "loss": 0.4478, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.224, | |
| "grad_norm": 0.9133635393045865, | |
| "learning_rate": 1.7581272530970666e-06, | |
| "loss": 0.4478, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.9382288839507548, | |
| "learning_rate": 1.6870954482242707e-06, | |
| "loss": 0.4242, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.2560000000000002, | |
| "grad_norm": 0.886060967878929, | |
| "learning_rate": 1.6172360085526567e-06, | |
| "loss": 0.3774, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.2720000000000002, | |
| "grad_norm": 0.8890006388840977, | |
| "learning_rate": 1.54857365582557e-06, | |
| "loss": 0.3999, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.288, | |
| "grad_norm": 0.909611668899034, | |
| "learning_rate": 1.4811326881631937e-06, | |
| "loss": 0.44, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.7793418599904366, | |
| "learning_rate": 1.4149369714639856e-06, | |
| "loss": 0.4232, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 0.9300794602772559, | |
| "learning_rate": 1.3500099309590397e-06, | |
| "loss": 0.5263, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.336, | |
| "grad_norm": 0.948199719189033, | |
| "learning_rate": 1.2863745429224145e-06, | |
| "loss": 0.4198, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.352, | |
| "grad_norm": 0.8458186118990229, | |
| "learning_rate": 1.22405332654032e-06, | |
| "loss": 0.4926, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 1.0452361088055258, | |
| "learning_rate": 1.1630683359420653e-06, | |
| "loss": 0.4694, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.384, | |
| "grad_norm": 1.049356748132299, | |
| "learning_rate": 1.103441152395588e-06, | |
| "loss": 0.3825, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 1.618515362363599, | |
| "learning_rate": 1.045192876670298e-06, | |
| "loss": 0.3684, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.416, | |
| "grad_norm": 0.9620814913134381, | |
| "learning_rate": 9.883441215699824e-07, | |
| "loss": 0.4741, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 0.9389376512718313, | |
| "learning_rate": 9.329150046383773e-07, | |
| "loss": 0.4272, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.448, | |
| "grad_norm": 0.8869531536912516, | |
| "learning_rate": 8.789251410400024e-07, | |
| "loss": 0.3987, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.464, | |
| "grad_norm": 1.051749769079472, | |
| "learning_rate": 8.263936366187825e-07, | |
| "loss": 0.4149, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 0.9918649474981351, | |
| "learning_rate": 7.753390811368972e-07, | |
| "loss": 0.4184, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.9123394443383827, | |
| "learning_rate": 7.257795416962754e-07, | |
| "loss": 0.4235, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.512, | |
| "grad_norm": 0.9060968543805388, | |
| "learning_rate": 6.777325563450282e-07, | |
| "loss": 0.5057, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.528, | |
| "grad_norm": 0.9049457543528188, | |
| "learning_rate": 6.312151278711237e-07, | |
| "loss": 0.4302, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.544, | |
| "grad_norm": 1.0122827328498305, | |
| "learning_rate": 5.862437177854629e-07, | |
| "loss": 0.4337, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.88362396840638, | |
| "learning_rate": 5.428342404965076e-07, | |
| "loss": 0.4621, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.576, | |
| "grad_norm": 0.945109351083517, | |
| "learning_rate": 5.010020576785174e-07, | |
| "loss": 0.4236, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.592, | |
| "grad_norm": 0.9415002277707675, | |
| "learning_rate": 4.607619728353818e-07, | |
| "loss": 0.4458, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.608, | |
| "grad_norm": 0.8855807648934247, | |
| "learning_rate": 4.221282260619891e-07, | |
| "loss": 0.4533, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 0.9505495576527682, | |
| "learning_rate": 3.851144890049535e-07, | |
| "loss": 0.4772, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 0.9688417824831936, | |
| "learning_rate": 3.497338600245254e-07, | |
| "loss": 0.4147, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.656, | |
| "grad_norm": 0.8485799740116599, | |
| "learning_rate": 3.159988595593616e-07, | |
| "loss": 0.4236, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.672, | |
| "grad_norm": 1.0096938277725918, | |
| "learning_rate": 2.839214256958106e-07, | |
| "loss": 0.4818, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.878961396566758, | |
| "learning_rate": 2.5351290994328703e-07, | |
| "loss": 0.4575, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.7039999999999997, | |
| "grad_norm": 0.9770777710606354, | |
| "learning_rate": 2.2478407321721295e-07, | |
| "loss": 0.3594, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.7199999999999998, | |
| "grad_norm": 0.8978502130431599, | |
| "learning_rate": 1.9774508203096843e-07, | |
| "loss": 0.4001, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.7359999999999998, | |
| "grad_norm": 0.988105588760895, | |
| "learning_rate": 1.7240550489817652e-07, | |
| "loss": 0.3845, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.9622193305055735, | |
| "learning_rate": 1.4877430894662037e-07, | |
| "loss": 0.3896, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.768, | |
| "grad_norm": 0.880324062009391, | |
| "learning_rate": 1.268598567449647e-07, | |
| "loss": 0.5497, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.784, | |
| "grad_norm": 0.8932996189529179, | |
| "learning_rate": 1.0666990334342708e-07, | |
| "loss": 0.4089, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 0.8225085016946534, | |
| "learning_rate": 8.821159352943142e-08, | |
| "loss": 0.4339, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 0.928363541615099, | |
| "learning_rate": 7.149145929922607e-08, | |
| "loss": 0.4231, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.832, | |
| "grad_norm": 0.9461162298103971, | |
| "learning_rate": 5.651541754634726e-08, | |
| "loss": 0.3679, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.848, | |
| "grad_norm": 0.9108957411474731, | |
| "learning_rate": 4.328876796776071e-08, | |
| "loss": 0.438, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.864, | |
| "grad_norm": 0.8994084973669091, | |
| "learning_rate": 3.181619118841517e-08, | |
| "loss": 0.4638, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.8490400793528896, | |
| "learning_rate": 2.210174710486679e-08, | |
| "loss": 0.5289, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.896, | |
| "grad_norm": 0.838508239352245, | |
| "learning_rate": 1.4148873448573408e-08, | |
| "loss": 0.4228, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.912, | |
| "grad_norm": 0.9224265588390108, | |
| "learning_rate": 7.96038456935322e-09, | |
| "loss": 0.4551, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.928, | |
| "grad_norm": 0.8025957344080616, | |
| "learning_rate": 3.538470439448105e-09, | |
| "loss": 0.4526, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.90654826891552, | |
| "learning_rate": 8.846958785418969e-10, | |
| "loss": 0.3929, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 0.9314677968691474, | |
| "learning_rate": 0.0, | |
| "loss": 0.3891, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "step": 186, | |
| "total_flos": 14252361744384.0, | |
| "train_loss": 0.5579822720378957, | |
| "train_runtime": 2990.1449, | |
| "train_samples_per_second": 1.003, | |
| "train_steps_per_second": 0.062 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 186, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 14252361744384.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |