| { | |
| "best_metric": 0.28952550888061523, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-100", | |
| "epoch": 2.9953488372093022, | |
| "eval_steps": 50, | |
| "global_step": 161, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.018604651162790697, | |
| "grad_norm": 1.01740562915802, | |
| "learning_rate": 7e-06, | |
| "loss": 1.7104, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.018604651162790697, | |
| "eval_loss": 0.6080694794654846, | |
| "eval_runtime": 10.5925, | |
| "eval_samples_per_second": 8.591, | |
| "eval_steps_per_second": 2.171, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.037209302325581395, | |
| "grad_norm": 1.2670235633850098, | |
| "learning_rate": 1.4e-05, | |
| "loss": 1.8752, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.05581395348837209, | |
| "grad_norm": 1.3142772912979126, | |
| "learning_rate": 2.1e-05, | |
| "loss": 1.7304, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.07441860465116279, | |
| "grad_norm": 1.2250651121139526, | |
| "learning_rate": 2.8e-05, | |
| "loss": 1.9433, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.09302325581395349, | |
| "grad_norm": 1.5665745735168457, | |
| "learning_rate": 3.5e-05, | |
| "loss": 2.1169, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.11162790697674418, | |
| "grad_norm": 1.5763330459594727, | |
| "learning_rate": 4.2e-05, | |
| "loss": 2.4289, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.13023255813953488, | |
| "grad_norm": 1.7122726440429688, | |
| "learning_rate": 4.899999999999999e-05, | |
| "loss": 2.4843, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.14883720930232558, | |
| "grad_norm": 1.7638782262802124, | |
| "learning_rate": 5.6e-05, | |
| "loss": 1.914, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.16744186046511628, | |
| "grad_norm": 2.0044069290161133, | |
| "learning_rate": 6.3e-05, | |
| "loss": 2.2168, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.18604651162790697, | |
| "grad_norm": 2.2012875080108643, | |
| "learning_rate": 7e-05, | |
| "loss": 2.3124, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.20465116279069767, | |
| "grad_norm": 2.3812570571899414, | |
| "learning_rate": 6.99924252512213e-05, | |
| "loss": 2.1676, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.22325581395348837, | |
| "grad_norm": 1.740301251411438, | |
| "learning_rate": 6.996970428356058e-05, | |
| "loss": 1.9273, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.24186046511627907, | |
| "grad_norm": 4.242422580718994, | |
| "learning_rate": 6.993184693162481e-05, | |
| "loss": 3.0044, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.26046511627906976, | |
| "grad_norm": 1.112087368965149, | |
| "learning_rate": 6.987886958169575e-05, | |
| "loss": 1.5393, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.27906976744186046, | |
| "grad_norm": 0.9684474468231201, | |
| "learning_rate": 6.981079516463718e-05, | |
| "loss": 1.1536, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.29767441860465116, | |
| "grad_norm": 0.8984664082527161, | |
| "learning_rate": 6.972765314596951e-05, | |
| "loss": 1.1399, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.31627906976744186, | |
| "grad_norm": 0.9321298599243164, | |
| "learning_rate": 6.962947951311591e-05, | |
| "loss": 1.4814, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.33488372093023255, | |
| "grad_norm": 0.8456665277481079, | |
| "learning_rate": 6.951631675982521e-05, | |
| "loss": 1.3894, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.35348837209302325, | |
| "grad_norm": 0.7932624220848083, | |
| "learning_rate": 6.938821386777899e-05, | |
| "loss": 1.4062, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.37209302325581395, | |
| "grad_norm": 1.1389163732528687, | |
| "learning_rate": 6.924522628539012e-05, | |
| "loss": 1.4328, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.39069767441860465, | |
| "grad_norm": 1.0435296297073364, | |
| "learning_rate": 6.908741590380231e-05, | |
| "loss": 1.496, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.40930232558139534, | |
| "grad_norm": 1.1448780298233032, | |
| "learning_rate": 6.891485103010098e-05, | |
| "loss": 1.2627, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.42790697674418604, | |
| "grad_norm": 1.52295982837677, | |
| "learning_rate": 6.87276063577471e-05, | |
| "loss": 1.4529, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.44651162790697674, | |
| "grad_norm": 1.8864622116088867, | |
| "learning_rate": 6.852576293424654e-05, | |
| "loss": 2.0397, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.46511627906976744, | |
| "grad_norm": 1.491450548171997, | |
| "learning_rate": 6.830940812606936e-05, | |
| "loss": 1.5753, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.48372093023255813, | |
| "grad_norm": 2.425402879714966, | |
| "learning_rate": 6.807863558083377e-05, | |
| "loss": 1.2114, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.5023255813953489, | |
| "grad_norm": 0.5629173517227173, | |
| "learning_rate": 6.78335451867715e-05, | |
| "loss": 1.6225, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.5209302325581395, | |
| "grad_norm": 0.7677317261695862, | |
| "learning_rate": 6.757424302949186e-05, | |
| "loss": 1.6867, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.5395348837209303, | |
| "grad_norm": 0.9299004673957825, | |
| "learning_rate": 6.73008413460634e-05, | |
| "loss": 1.288, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.5581395348837209, | |
| "grad_norm": 0.8549377918243408, | |
| "learning_rate": 6.701345847643282e-05, | |
| "loss": 1.1299, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.5767441860465117, | |
| "grad_norm": 1.0633025169372559, | |
| "learning_rate": 6.671221881220246e-05, | |
| "loss": 1.5311, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.5953488372093023, | |
| "grad_norm": 1.01902174949646, | |
| "learning_rate": 6.639725274278824e-05, | |
| "loss": 1.3447, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.6139534883720931, | |
| "grad_norm": 1.1506015062332153, | |
| "learning_rate": 6.606869659898156e-05, | |
| "loss": 1.3272, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.6325581395348837, | |
| "grad_norm": 1.0536060333251953, | |
| "learning_rate": 6.572669259393953e-05, | |
| "loss": 1.317, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.6511627906976745, | |
| "grad_norm": 1.214861512184143, | |
| "learning_rate": 6.537138876162896e-05, | |
| "loss": 1.2711, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.6697674418604651, | |
| "grad_norm": 1.3296098709106445, | |
| "learning_rate": 6.5002938892751e-05, | |
| "loss": 1.471, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.6883720930232559, | |
| "grad_norm": 1.5142545700073242, | |
| "learning_rate": 6.462150246817388e-05, | |
| "loss": 1.5411, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.7069767441860465, | |
| "grad_norm": 1.7185128927230835, | |
| "learning_rate": 6.422724458990284e-05, | |
| "loss": 1.1718, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.7255813953488373, | |
| "grad_norm": 2.9415876865386963, | |
| "learning_rate": 6.38203359096168e-05, | |
| "loss": 1.1685, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.7441860465116279, | |
| "grad_norm": 0.6370990872383118, | |
| "learning_rate": 6.340095255480317e-05, | |
| "loss": 1.2429, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.7627906976744186, | |
| "grad_norm": 0.7854994535446167, | |
| "learning_rate": 6.296927605252219e-05, | |
| "loss": 1.1277, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.7813953488372093, | |
| "grad_norm": 0.6397520899772644, | |
| "learning_rate": 6.252549325083437e-05, | |
| "loss": 0.9854, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.8113309741020203, | |
| "learning_rate": 6.206979623792457e-05, | |
| "loss": 1.5336, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.8186046511627907, | |
| "grad_norm": 0.9286227226257324, | |
| "learning_rate": 6.160238225895803e-05, | |
| "loss": 1.32, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.8372093023255814, | |
| "grad_norm": 0.9448747634887695, | |
| "learning_rate": 6.112345363070428e-05, | |
| "loss": 1.2975, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.8558139534883721, | |
| "grad_norm": 0.92212975025177, | |
| "learning_rate": 6.063321765396568e-05, | |
| "loss": 1.1939, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.8744186046511628, | |
| "grad_norm": 1.1105095148086548, | |
| "learning_rate": 6.013188652384889e-05, | |
| "loss": 1.0333, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.8930232558139535, | |
| "grad_norm": 1.1572589874267578, | |
| "learning_rate": 5.9619677237917655e-05, | |
| "loss": 1.2029, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.9116279069767442, | |
| "grad_norm": 1.4601818323135376, | |
| "learning_rate": 5.9096811502266993e-05, | |
| "loss": 1.3201, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.9302325581395349, | |
| "grad_norm": 1.729429006576538, | |
| "learning_rate": 5.8563515635559344e-05, | |
| "loss": 1.3054, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.9302325581395349, | |
| "eval_loss": 0.3099779188632965, | |
| "eval_runtime": 10.8877, | |
| "eval_samples_per_second": 8.358, | |
| "eval_steps_per_second": 2.112, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.9488372093023256, | |
| "grad_norm": 1.666062593460083, | |
| "learning_rate": 5.8020020471064145e-05, | |
| "loss": 1.1947, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.9674418604651163, | |
| "grad_norm": 2.129988431930542, | |
| "learning_rate": 5.746656125674329e-05, | |
| "loss": 1.003, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.986046511627907, | |
| "grad_norm": 0.9616814255714417, | |
| "learning_rate": 5.690337755342581e-05, | |
| "loss": 1.3023, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.0046511627906978, | |
| "grad_norm": 1.7374155521392822, | |
| "learning_rate": 5.633071313111564e-05, | |
| "loss": 1.3017, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.0232558139534884, | |
| "grad_norm": 0.5342385768890381, | |
| "learning_rate": 5.574881586347755e-05, | |
| "loss": 0.8393, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.041860465116279, | |
| "grad_norm": 0.7542828917503357, | |
| "learning_rate": 5.5157937620546804e-05, | |
| "loss": 0.9745, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.0604651162790697, | |
| "grad_norm": 0.7869417071342468, | |
| "learning_rate": 5.4558334159709074e-05, | |
| "loss": 1.1747, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.0790697674418606, | |
| "grad_norm": 0.773287296295166, | |
| "learning_rate": 5.3950265014997655e-05, | |
| "loss": 0.8018, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.0976744186046512, | |
| "grad_norm": 1.2457112073898315, | |
| "learning_rate": 5.3333993384756045e-05, | |
| "loss": 0.9032, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.1162790697674418, | |
| "grad_norm": 1.0725675821304321, | |
| "learning_rate": 5.270978601771444e-05, | |
| "loss": 0.9193, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.1348837209302325, | |
| "grad_norm": 0.763644814491272, | |
| "learning_rate": 5.207791309752948e-05, | |
| "loss": 0.7066, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.1534883720930234, | |
| "grad_norm": 0.8121412396430969, | |
| "learning_rate": 5.1438648125837204e-05, | |
| "loss": 0.7076, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.172093023255814, | |
| "grad_norm": 1.0602554082870483, | |
| "learning_rate": 5.07922678038698e-05, | |
| "loss": 0.7972, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.1906976744186046, | |
| "grad_norm": 1.2864611148834229, | |
| "learning_rate": 5.013905191268754e-05, | |
| "loss": 0.9987, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.2093023255813953, | |
| "grad_norm": 1.507265567779541, | |
| "learning_rate": 4.9479283192077487e-05, | |
| "loss": 0.9133, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.2279069767441861, | |
| "grad_norm": 1.0964022874832153, | |
| "learning_rate": 4.881324721817167e-05, | |
| "loss": 0.6193, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.2465116279069768, | |
| "grad_norm": 1.7399965524673462, | |
| "learning_rate": 4.8141232279837554e-05, | |
| "loss": 0.8116, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.2651162790697674, | |
| "grad_norm": 0.8102202415466309, | |
| "learning_rate": 4.746352925389414e-05, | |
| "loss": 1.1357, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.283720930232558, | |
| "grad_norm": 0.9682488441467285, | |
| "learning_rate": 4.678043147920822e-05, | |
| "loss": 1.0843, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.302325581395349, | |
| "grad_norm": 0.8420053720474243, | |
| "learning_rate": 4.60922346297246e-05, | |
| "loss": 1.052, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.3209302325581396, | |
| "grad_norm": 0.9388038516044617, | |
| "learning_rate": 4.539923658648585e-05, | |
| "loss": 0.8371, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.3395348837209302, | |
| "grad_norm": 0.8845089077949524, | |
| "learning_rate": 4.4701737308696665e-05, | |
| "loss": 0.7301, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.3581395348837209, | |
| "grad_norm": 1.6632778644561768, | |
| "learning_rate": 4.4000038703888604e-05, | |
| "loss": 1.0172, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.3767441860465115, | |
| "grad_norm": 1.184276819229126, | |
| "learning_rate": 4.3294444497241656e-05, | |
| "loss": 0.8308, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.3953488372093024, | |
| "grad_norm": 1.2161909341812134, | |
| "learning_rate": 4.2585260100118973e-05, | |
| "loss": 0.6478, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.413953488372093, | |
| "grad_norm": 1.1866018772125244, | |
| "learning_rate": 4.187279247787175e-05, | |
| "loss": 0.6407, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.4325581395348836, | |
| "grad_norm": 1.264217495918274, | |
| "learning_rate": 4.115735001697149e-05, | |
| "loss": 0.8542, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.4511627906976745, | |
| "grad_norm": 1.6465506553649902, | |
| "learning_rate": 4.0439242391527176e-05, | |
| "loss": 0.7179, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.4697674418604652, | |
| "grad_norm": 1.624822735786438, | |
| "learning_rate": 3.971878042924501e-05, | |
| "loss": 0.6409, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.4883720930232558, | |
| "grad_norm": 1.7168909311294556, | |
| "learning_rate": 3.899627597688895e-05, | |
| "loss": 0.7917, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.5069767441860464, | |
| "grad_norm": 0.8661062121391296, | |
| "learning_rate": 3.8272041765299995e-05, | |
| "loss": 0.9741, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.525581395348837, | |
| "grad_norm": 0.9207056164741516, | |
| "learning_rate": 3.754639127403304e-05, | |
| "loss": 1.2139, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.544186046511628, | |
| "grad_norm": 1.0560969114303589, | |
| "learning_rate": 3.6819638595669306e-05, | |
| "loss": 0.8169, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.5627906976744186, | |
| "grad_norm": 1.0724343061447144, | |
| "learning_rate": 3.609209829986387e-05, | |
| "loss": 0.9448, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.5813953488372094, | |
| "grad_norm": 1.2073761224746704, | |
| "learning_rate": 3.536408529718625e-05, | |
| "loss": 0.9814, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 1.2773185968399048, | |
| "learning_rate": 3.463591470281375e-05, | |
| "loss": 0.7854, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.6186046511627907, | |
| "grad_norm": 1.2945756912231445, | |
| "learning_rate": 3.390790170013612e-05, | |
| "loss": 0.8327, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.6372093023255814, | |
| "grad_norm": 1.2657415866851807, | |
| "learning_rate": 3.318036140433069e-05, | |
| "loss": 0.7456, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.655813953488372, | |
| "grad_norm": 1.308043360710144, | |
| "learning_rate": 3.245360872596697e-05, | |
| "loss": 0.5729, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.6744186046511627, | |
| "grad_norm": 1.3613492250442505, | |
| "learning_rate": 3.172795823469999e-05, | |
| "loss": 0.6906, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.6930232558139535, | |
| "grad_norm": 1.2736989259719849, | |
| "learning_rate": 3.100372402311106e-05, | |
| "loss": 0.3997, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.7116279069767442, | |
| "grad_norm": 1.814155101776123, | |
| "learning_rate": 3.0281219570754987e-05, | |
| "loss": 0.5502, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.730232558139535, | |
| "grad_norm": 1.9878065586090088, | |
| "learning_rate": 2.956075760847283e-05, | |
| "loss": 1.1875, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.7488372093023257, | |
| "grad_norm": 0.9894658923149109, | |
| "learning_rate": 2.8842649983028517e-05, | |
| "loss": 1.081, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.7674418604651163, | |
| "grad_norm": 0.9422222375869751, | |
| "learning_rate": 2.8127207522128252e-05, | |
| "loss": 0.6992, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.786046511627907, | |
| "grad_norm": 1.0437465906143188, | |
| "learning_rate": 2.7414739899881024e-05, | |
| "loss": 0.7245, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.8046511627906976, | |
| "grad_norm": 1.2193009853363037, | |
| "learning_rate": 2.670555550275834e-05, | |
| "loss": 0.7591, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.8232558139534882, | |
| "grad_norm": 1.1166754961013794, | |
| "learning_rate": 2.5999961296111394e-05, | |
| "loss": 0.9399, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.841860465116279, | |
| "grad_norm": 1.4289640188217163, | |
| "learning_rate": 2.5298262691303332e-05, | |
| "loss": 0.9413, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.8604651162790697, | |
| "grad_norm": 1.2337830066680908, | |
| "learning_rate": 2.4600763413514138e-05, | |
| "loss": 0.7486, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.8604651162790697, | |
| "eval_loss": 0.28952550888061523, | |
| "eval_runtime": 10.8771, | |
| "eval_samples_per_second": 8.366, | |
| "eval_steps_per_second": 2.115, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.8790697674418606, | |
| "grad_norm": 1.4255279302597046, | |
| "learning_rate": 2.3907765370275404e-05, | |
| "loss": 0.765, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.8976744186046512, | |
| "grad_norm": 1.7940795421600342, | |
| "learning_rate": 2.3219568520791783e-05, | |
| "loss": 0.6342, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.916279069767442, | |
| "grad_norm": 1.6805592775344849, | |
| "learning_rate": 2.2536470746105853e-05, | |
| "loss": 0.7094, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.9348837209302325, | |
| "grad_norm": 2.0769453048706055, | |
| "learning_rate": 2.1858767720162456e-05, | |
| "loss": 0.8721, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.9534883720930232, | |
| "grad_norm": 1.9850454330444336, | |
| "learning_rate": 2.1186752781828328e-05, | |
| "loss": 0.5382, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.9720930232558138, | |
| "grad_norm": 1.6767321825027466, | |
| "learning_rate": 2.0520716807922514e-05, | |
| "loss": 0.5035, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.9906976744186047, | |
| "grad_norm": 1.6948860883712769, | |
| "learning_rate": 1.986094808731247e-05, | |
| "loss": 0.9199, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 2.0093023255813955, | |
| "grad_norm": 1.2903045415878296, | |
| "learning_rate": 1.92077321961302e-05, | |
| "loss": 0.6539, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 2.027906976744186, | |
| "grad_norm": 0.6793351173400879, | |
| "learning_rate": 1.8561351874162783e-05, | |
| "loss": 0.6779, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 2.046511627906977, | |
| "grad_norm": 1.0761351585388184, | |
| "learning_rate": 1.7922086902470502e-05, | |
| "loss": 0.6817, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.0651162790697675, | |
| "grad_norm": 0.825526773929596, | |
| "learning_rate": 1.7290213982285564e-05, | |
| "loss": 0.5877, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 2.083720930232558, | |
| "grad_norm": 0.9566294550895691, | |
| "learning_rate": 1.666600661524396e-05, | |
| "loss": 0.5537, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 2.1023255813953488, | |
| "grad_norm": 0.9278555512428284, | |
| "learning_rate": 1.604973498500235e-05, | |
| "loss": 0.6015, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 2.1209302325581394, | |
| "grad_norm": 0.9701972007751465, | |
| "learning_rate": 1.5441665840290927e-05, | |
| "loss": 0.547, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 2.13953488372093, | |
| "grad_norm": 1.187429666519165, | |
| "learning_rate": 1.4842062379453195e-05, | |
| "loss": 0.5522, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.158139534883721, | |
| "grad_norm": 0.8903448581695557, | |
| "learning_rate": 1.4251184136522453e-05, | |
| "loss": 0.403, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 2.1767441860465118, | |
| "grad_norm": 0.9742254614830017, | |
| "learning_rate": 1.3669286868884356e-05, | |
| "loss": 0.279, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 2.1953488372093024, | |
| "grad_norm": 0.9774483442306519, | |
| "learning_rate": 1.3096622446574184e-05, | |
| "loss": 0.3753, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 2.213953488372093, | |
| "grad_norm": 1.1328277587890625, | |
| "learning_rate": 1.2533438743256703e-05, | |
| "loss": 0.2929, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 2.2325581395348837, | |
| "grad_norm": 1.241883397102356, | |
| "learning_rate": 1.1979979528935865e-05, | |
| "loss": 0.3254, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.2511627906976743, | |
| "grad_norm": 0.8701224327087402, | |
| "learning_rate": 1.1436484364440646e-05, | |
| "loss": 0.5831, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 2.269767441860465, | |
| "grad_norm": 0.9288952350616455, | |
| "learning_rate": 1.0903188497732999e-05, | |
| "loss": 0.8483, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.2883720930232556, | |
| "grad_norm": 0.7865915298461914, | |
| "learning_rate": 1.038032276208234e-05, | |
| "loss": 0.5125, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 2.3069767441860467, | |
| "grad_norm": 1.1094341278076172, | |
| "learning_rate": 9.86811347615111e-06, | |
| "loss": 0.453, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 2.3255813953488373, | |
| "grad_norm": 1.1128486394882202, | |
| "learning_rate": 9.36678234603432e-06, | |
| "loss": 0.4629, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.344186046511628, | |
| "grad_norm": 1.2837432622909546, | |
| "learning_rate": 8.876546369295717e-06, | |
| "loss": 0.6856, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.3627906976744186, | |
| "grad_norm": 1.148620367050171, | |
| "learning_rate": 8.397617741041962e-06, | |
| "loss": 0.5077, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.3813953488372093, | |
| "grad_norm": 1.2100828886032104, | |
| "learning_rate": 7.930203762075426e-06, | |
| "loss": 0.4271, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 1.403386116027832, | |
| "learning_rate": 7.47450674916562e-06, | |
| "loss": 0.3401, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.4186046511627906, | |
| "grad_norm": 1.1033246517181396, | |
| "learning_rate": 7.030723947477803e-06, | |
| "loss": 0.2728, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.4372093023255816, | |
| "grad_norm": 1.4235916137695312, | |
| "learning_rate": 6.599047445196836e-06, | |
| "loss": 0.3717, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.4558139534883723, | |
| "grad_norm": 1.2089860439300537, | |
| "learning_rate": 6.1796640903832005e-06, | |
| "loss": 0.3064, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.474418604651163, | |
| "grad_norm": 1.1210060119628906, | |
| "learning_rate": 5.772755410097166e-06, | |
| "loss": 0.2141, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.4930232558139536, | |
| "grad_norm": 1.133145809173584, | |
| "learning_rate": 5.378497531826109e-06, | |
| "loss": 0.5761, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.511627906976744, | |
| "grad_norm": 1.5591737031936646, | |
| "learning_rate": 4.997061107248992e-06, | |
| "loss": 0.8019, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.530232558139535, | |
| "grad_norm": 1.521730899810791, | |
| "learning_rate": 4.6286112383710245e-06, | |
| "loss": 0.58, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.5488372093023255, | |
| "grad_norm": 1.5007507801055908, | |
| "learning_rate": 4.273307406060465e-06, | |
| "loss": 0.6187, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.567441860465116, | |
| "grad_norm": 1.2669973373413086, | |
| "learning_rate": 3.931303401018427e-06, | |
| "loss": 0.5699, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.5860465116279068, | |
| "grad_norm": 1.4318617582321167, | |
| "learning_rate": 3.602747257211758e-06, | |
| "loss": 0.5587, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.604651162790698, | |
| "grad_norm": 1.6378320455551147, | |
| "learning_rate": 3.2877811877975422e-06, | |
| "loss": 0.4079, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.6232558139534885, | |
| "grad_norm": 1.452219009399414, | |
| "learning_rate": 2.9865415235671784e-06, | |
| "loss": 0.4239, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.641860465116279, | |
| "grad_norm": 1.3756489753723145, | |
| "learning_rate": 2.6991586539366018e-06, | |
| "loss": 0.4746, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.66046511627907, | |
| "grad_norm": 1.2296236753463745, | |
| "learning_rate": 2.425756970508135e-06, | |
| "loss": 0.3361, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.6790697674418604, | |
| "grad_norm": 1.558907389640808, | |
| "learning_rate": 2.166454813228505e-06, | |
| "loss": 0.4134, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.697674418604651, | |
| "grad_norm": 1.4058880805969238, | |
| "learning_rate": 1.9213644191662354e-06, | |
| "loss": 0.2047, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.7162790697674417, | |
| "grad_norm": 1.1016334295272827, | |
| "learning_rate": 1.6905918739306423e-06, | |
| "loss": 0.149, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.734883720930233, | |
| "grad_norm": 1.425627589225769, | |
| "learning_rate": 1.4742370657534552e-06, | |
| "loss": 0.9796, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.753488372093023, | |
| "grad_norm": 1.2110799551010132, | |
| "learning_rate": 1.2723936422528996e-06, | |
| "loss": 0.6505, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.772093023255814, | |
| "grad_norm": 1.069684386253357, | |
| "learning_rate": 1.0851489698990146e-06, | |
| "loss": 0.6269, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.7906976744186047, | |
| "grad_norm": 1.4680191278457642, | |
| "learning_rate": 9.125840961976982e-07, | |
| "loss": 0.8263, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.7906976744186047, | |
| "eval_loss": 0.3216976821422577, | |
| "eval_runtime": 10.869, | |
| "eval_samples_per_second": 8.372, | |
| "eval_steps_per_second": 2.116, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.8093023255813954, | |
| "grad_norm": 1.4344302415847778, | |
| "learning_rate": 7.547737146098809e-07, | |
| "loss": 0.768, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.827906976744186, | |
| "grad_norm": 1.6189731359481812, | |
| "learning_rate": 6.117861322210028e-07, | |
| "loss": 0.5337, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.8465116279069766, | |
| "grad_norm": 1.4234650135040283, | |
| "learning_rate": 4.836832401747892e-07, | |
| "loss": 0.5149, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.8651162790697673, | |
| "grad_norm": 1.1778002977371216, | |
| "learning_rate": 3.705204868840894e-07, | |
| "loss": 0.4272, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.883720930232558, | |
| "grad_norm": 1.35899019241333, | |
| "learning_rate": 2.7234685403047243e-07, | |
| "loss": 0.3179, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.902325581395349, | |
| "grad_norm": 1.4125027656555176, | |
| "learning_rate": 1.892048353628217e-07, | |
| "loss": 0.365, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.9209302325581397, | |
| "grad_norm": 1.287215232849121, | |
| "learning_rate": 1.211304183042494e-07, | |
| "loss": 0.3343, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.9395348837209303, | |
| "grad_norm": 1.398272156715393, | |
| "learning_rate": 6.81530683751752e-08, | |
| "loss": 0.3152, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.958139534883721, | |
| "grad_norm": 1.168246865272522, | |
| "learning_rate": 3.029571643941353e-08, | |
| "loss": 0.1939, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.9767441860465116, | |
| "grad_norm": 1.5252883434295654, | |
| "learning_rate": 7.574748778697015e-09, | |
| "loss": 0.4645, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.9953488372093022, | |
| "grad_norm": 1.418809413909912, | |
| "learning_rate": 0.0, | |
| "loss": 0.3772, | |
| "step": 161 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 161, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 4, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 1 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.532890163478528e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |