| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 6.850632911392405, | |
| "eval_steps": 500, | |
| "global_step": 224, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.030379746835443037, | |
| "grad_norm": 6.3879851713105475, | |
| "learning_rate": 8.695652173913044e-07, | |
| "loss": 1.0292, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.060759493670886074, | |
| "grad_norm": 6.494020006454658, | |
| "learning_rate": 1.7391304347826088e-06, | |
| "loss": 1.0446, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.09113924050632911, | |
| "grad_norm": 6.43234729011982, | |
| "learning_rate": 2.6086956521739132e-06, | |
| "loss": 1.0392, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.12151898734177215, | |
| "grad_norm": 5.839565521716624, | |
| "learning_rate": 3.4782608695652175e-06, | |
| "loss": 1.022, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.1518987341772152, | |
| "grad_norm": 4.393068161057827, | |
| "learning_rate": 4.347826086956522e-06, | |
| "loss": 0.9732, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.18227848101265823, | |
| "grad_norm": 2.7840571673794052, | |
| "learning_rate": 5.2173913043478265e-06, | |
| "loss": 0.9403, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.21265822784810126, | |
| "grad_norm": 2.519822806102858, | |
| "learning_rate": 6.086956521739132e-06, | |
| "loss": 0.9503, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.2430379746835443, | |
| "grad_norm": 4.173680937704689, | |
| "learning_rate": 6.956521739130435e-06, | |
| "loss": 0.942, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.27341772151898736, | |
| "grad_norm": 4.325040025251528, | |
| "learning_rate": 7.82608695652174e-06, | |
| "loss": 0.9543, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.3037974683544304, | |
| "grad_norm": 3.752594166534657, | |
| "learning_rate": 8.695652173913044e-06, | |
| "loss": 0.9159, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3341772151898734, | |
| "grad_norm": 3.6156572533154887, | |
| "learning_rate": 9.565217391304349e-06, | |
| "loss": 0.8817, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.36455696202531646, | |
| "grad_norm": 3.150516119080985, | |
| "learning_rate": 1.0434782608695653e-05, | |
| "loss": 0.8814, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.3949367088607595, | |
| "grad_norm": 2.252685549710176, | |
| "learning_rate": 1.1304347826086957e-05, | |
| "loss": 0.8728, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.4253164556962025, | |
| "grad_norm": 1.4118327654199778, | |
| "learning_rate": 1.2173913043478263e-05, | |
| "loss": 0.8283, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.45569620253164556, | |
| "grad_norm": 1.5313179522473905, | |
| "learning_rate": 1.3043478260869566e-05, | |
| "loss": 0.8187, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.4860759493670886, | |
| "grad_norm": 1.3449715830939621, | |
| "learning_rate": 1.391304347826087e-05, | |
| "loss": 0.8071, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.5164556962025316, | |
| "grad_norm": 1.179170555148964, | |
| "learning_rate": 1.4782608695652174e-05, | |
| "loss": 0.7971, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.5468354430379747, | |
| "grad_norm": 1.049667040795823, | |
| "learning_rate": 1.565217391304348e-05, | |
| "loss": 0.7614, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.5772151898734177, | |
| "grad_norm": 1.0046444332216011, | |
| "learning_rate": 1.6521739130434785e-05, | |
| "loss": 0.7952, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.6075949367088608, | |
| "grad_norm": 1.0121384725810043, | |
| "learning_rate": 1.739130434782609e-05, | |
| "loss": 0.7646, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.6379746835443038, | |
| "grad_norm": 1.007166587838568, | |
| "learning_rate": 1.8260869565217393e-05, | |
| "loss": 0.7599, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.6683544303797468, | |
| "grad_norm": 0.8434480603962347, | |
| "learning_rate": 1.9130434782608697e-05, | |
| "loss": 0.7502, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.6987341772151898, | |
| "grad_norm": 0.7912625765581864, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7515, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.7291139240506329, | |
| "grad_norm": 0.8868245838464142, | |
| "learning_rate": 1.999877856940653e-05, | |
| "loss": 0.7523, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.759493670886076, | |
| "grad_norm": 0.7894735223611277, | |
| "learning_rate": 1.999511457600466e-05, | |
| "loss": 0.7354, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.789873417721519, | |
| "grad_norm": 0.8932695783756212, | |
| "learning_rate": 1.9989008914857115e-05, | |
| "loss": 0.7253, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.8202531645569621, | |
| "grad_norm": 0.6660004467974519, | |
| "learning_rate": 1.998046307749216e-05, | |
| "loss": 0.736, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.850632911392405, | |
| "grad_norm": 0.6771804784162709, | |
| "learning_rate": 1.9969479151539238e-05, | |
| "loss": 0.7104, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.8810126582278481, | |
| "grad_norm": 0.8055973142766584, | |
| "learning_rate": 1.9956059820218982e-05, | |
| "loss": 0.7019, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.9113924050632911, | |
| "grad_norm": 0.742505387009152, | |
| "learning_rate": 1.9940208361687762e-05, | |
| "loss": 0.7151, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.9417721518987342, | |
| "grad_norm": 0.6921719958642204, | |
| "learning_rate": 1.9921928648236855e-05, | |
| "loss": 0.7213, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.9721518987341772, | |
| "grad_norm": 0.8530995324298626, | |
| "learning_rate": 1.990122514534651e-05, | |
| "loss": 0.7109, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.010126582278481, | |
| "grad_norm": 0.7175968051416318, | |
| "learning_rate": 1.9878102910595097e-05, | |
| "loss": 0.7053, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.040506329113924, | |
| "grad_norm": 0.7977037069804985, | |
| "learning_rate": 1.985256759242359e-05, | |
| "loss": 0.6874, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.070886075949367, | |
| "grad_norm": 0.7295150625660252, | |
| "learning_rate": 1.982462542875576e-05, | |
| "loss": 0.6736, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.1012658227848102, | |
| "grad_norm": 0.5165378355051813, | |
| "learning_rate": 1.979428324547432e-05, | |
| "loss": 0.6712, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.1316455696202532, | |
| "grad_norm": 0.5982893458836316, | |
| "learning_rate": 1.9761548454753455e-05, | |
| "loss": 0.676, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.1620253164556962, | |
| "grad_norm": 0.5736345170850041, | |
| "learning_rate": 1.972642905324813e-05, | |
| "loss": 0.6575, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.1924050632911392, | |
| "grad_norm": 0.6017620216326773, | |
| "learning_rate": 1.9688933620140638e-05, | |
| "loss": 0.6665, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.2227848101265824, | |
| "grad_norm": 0.5138687390469493, | |
| "learning_rate": 1.96490713150448e-05, | |
| "loss": 0.6487, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.2531645569620253, | |
| "grad_norm": 0.5776927579394718, | |
| "learning_rate": 1.9606851875768404e-05, | |
| "loss": 0.6562, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.2835443037974683, | |
| "grad_norm": 0.5184748392250274, | |
| "learning_rate": 1.956228561593441e-05, | |
| "loss": 0.6471, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.3139240506329113, | |
| "grad_norm": 0.5811935023189696, | |
| "learning_rate": 1.9515383422461457e-05, | |
| "loss": 0.6409, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.3443037974683545, | |
| "grad_norm": 0.5690816743397383, | |
| "learning_rate": 1.9466156752904344e-05, | |
| "loss": 0.6534, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.3746835443037975, | |
| "grad_norm": 0.5413562774931343, | |
| "learning_rate": 1.9414617632655114e-05, | |
| "loss": 0.6515, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.4050632911392404, | |
| "grad_norm": 0.6000406960300322, | |
| "learning_rate": 1.9360778652005416e-05, | |
| "loss": 0.6503, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.4354430379746836, | |
| "grad_norm": 0.571191551256675, | |
| "learning_rate": 1.9304652963070868e-05, | |
| "loss": 0.6445, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.4658227848101266, | |
| "grad_norm": 0.6211512266562935, | |
| "learning_rate": 1.9246254276578175e-05, | |
| "loss": 0.6628, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.4962025316455696, | |
| "grad_norm": 0.5486784144759123, | |
| "learning_rate": 1.9185596858515797e-05, | |
| "loss": 0.6413, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.5265822784810128, | |
| "grad_norm": 0.6383497671810656, | |
| "learning_rate": 1.9122695526648968e-05, | |
| "loss": 0.6285, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.5569620253164556, | |
| "grad_norm": 0.6404954200236757, | |
| "learning_rate": 1.905756564689991e-05, | |
| "loss": 0.6388, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.5873417721518988, | |
| "grad_norm": 0.6966640989995635, | |
| "learning_rate": 1.8990223129594146e-05, | |
| "loss": 0.6339, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.6177215189873417, | |
| "grad_norm": 0.732880355582101, | |
| "learning_rate": 1.8920684425573865e-05, | |
| "loss": 0.6635, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.6481012658227847, | |
| "grad_norm": 0.8209623594160774, | |
| "learning_rate": 1.884896652217917e-05, | |
| "loss": 0.6554, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.678481012658228, | |
| "grad_norm": 0.6620551804668269, | |
| "learning_rate": 1.877508693909831e-05, | |
| "loss": 0.6307, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.7088607594936709, | |
| "grad_norm": 0.5569955983083782, | |
| "learning_rate": 1.8699063724087905e-05, | |
| "loss": 0.6392, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.7392405063291139, | |
| "grad_norm": 0.5694147652222962, | |
| "learning_rate": 1.862091544856407e-05, | |
| "loss": 0.6262, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.769620253164557, | |
| "grad_norm": 0.5131432523855696, | |
| "learning_rate": 1.854066120306571e-05, | |
| "loss": 0.6485, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.5195376658803109, | |
| "learning_rate": 1.8458320592590976e-05, | |
| "loss": 0.638, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.830379746835443, | |
| "grad_norm": 0.5005952823201074, | |
| "learning_rate": 1.837391373180801e-05, | |
| "loss": 0.6303, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.8607594936708862, | |
| "grad_norm": 0.5746334092893775, | |
| "learning_rate": 1.8287461240141217e-05, | |
| "loss": 0.6233, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.891139240506329, | |
| "grad_norm": 0.46342457766749023, | |
| "learning_rate": 1.8198984236734246e-05, | |
| "loss": 0.6241, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.9215189873417722, | |
| "grad_norm": 0.6196606384374721, | |
| "learning_rate": 1.8108504335290852e-05, | |
| "loss": 0.6203, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.9518987341772152, | |
| "grad_norm": 0.5348116446835595, | |
| "learning_rate": 1.8016043638794975e-05, | |
| "loss": 0.6352, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.9822784810126581, | |
| "grad_norm": 0.5546833496384462, | |
| "learning_rate": 1.7921624734111292e-05, | |
| "loss": 0.623, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.020253164556962, | |
| "grad_norm": 0.6265154008351655, | |
| "learning_rate": 1.7825270686467567e-05, | |
| "loss": 0.598, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.050632911392405, | |
| "grad_norm": 0.47478519113621515, | |
| "learning_rate": 1.7727005033820117e-05, | |
| "loss": 0.5791, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.081012658227848, | |
| "grad_norm": 0.5288742011981481, | |
| "learning_rate": 1.762685178110382e-05, | |
| "loss": 0.5774, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.1113924050632913, | |
| "grad_norm": 0.5670564749327208, | |
| "learning_rate": 1.752483539436807e-05, | |
| "loss": 0.5759, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.141772151898734, | |
| "grad_norm": 0.7142973097542259, | |
| "learning_rate": 1.7420980794800013e-05, | |
| "loss": 0.5556, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.1721518987341772, | |
| "grad_norm": 0.6203100900043612, | |
| "learning_rate": 1.731531335263669e-05, | |
| "loss": 0.5849, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.2025316455696204, | |
| "grad_norm": 0.4982900289934448, | |
| "learning_rate": 1.720785888096743e-05, | |
| "loss": 0.5605, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.232911392405063, | |
| "grad_norm": 0.6555831752576167, | |
| "learning_rate": 1.7098643629428035e-05, | |
| "loss": 0.5768, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.2632911392405064, | |
| "grad_norm": 0.5113940986823249, | |
| "learning_rate": 1.698769427778842e-05, | |
| "loss": 0.5613, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.293670886075949, | |
| "grad_norm": 0.6958247255087919, | |
| "learning_rate": 1.687503792943506e-05, | |
| "loss": 0.5733, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.3240506329113924, | |
| "grad_norm": 0.6723634553290758, | |
| "learning_rate": 1.6760702104750046e-05, | |
| "loss": 0.5568, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 2.3544303797468356, | |
| "grad_norm": 0.6180427342280245, | |
| "learning_rate": 1.664471473438822e-05, | |
| "loss": 0.5526, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 2.3848101265822783, | |
| "grad_norm": 0.6193388376642808, | |
| "learning_rate": 1.6527104152454096e-05, | |
| "loss": 0.5328, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 2.4151898734177215, | |
| "grad_norm": 0.6077926431780674, | |
| "learning_rate": 1.6407899089580263e-05, | |
| "loss": 0.5593, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 2.4455696202531647, | |
| "grad_norm": 0.5682666286294648, | |
| "learning_rate": 1.628712866590885e-05, | |
| "loss": 0.5646, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.4759493670886075, | |
| "grad_norm": 0.6108881127546718, | |
| "learning_rate": 1.6164822383977912e-05, | |
| "loss": 0.5627, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 2.5063291139240507, | |
| "grad_norm": 0.6479056654339819, | |
| "learning_rate": 1.604101012151436e-05, | |
| "loss": 0.5716, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 2.536708860759494, | |
| "grad_norm": 0.5620218128549114, | |
| "learning_rate": 1.5915722124135227e-05, | |
| "loss": 0.5819, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 2.5670886075949366, | |
| "grad_norm": 0.5135062245378966, | |
| "learning_rate": 1.5788988997959115e-05, | |
| "loss": 0.5514, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 2.59746835443038, | |
| "grad_norm": 0.4449261643380341, | |
| "learning_rate": 1.5660841702129533e-05, | |
| "loss": 0.5513, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 2.6278481012658226, | |
| "grad_norm": 0.536413720572977, | |
| "learning_rate": 1.5531311541251995e-05, | |
| "loss": 0.582, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 2.6582278481012658, | |
| "grad_norm": 0.467483307774275, | |
| "learning_rate": 1.540043015774676e-05, | |
| "loss": 0.5703, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 2.688607594936709, | |
| "grad_norm": 0.4541387632776588, | |
| "learning_rate": 1.5268229524119007e-05, | |
| "loss": 0.5541, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 2.7189873417721517, | |
| "grad_norm": 0.43565357696170937, | |
| "learning_rate": 1.513474193514842e-05, | |
| "loss": 0.5458, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 2.749367088607595, | |
| "grad_norm": 0.44260798700989695, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 0.5601, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 2.779746835443038, | |
| "grad_norm": 0.40267489470129286, | |
| "learning_rate": 1.4864036634258112e-05, | |
| "loss": 0.5718, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 2.810126582278481, | |
| "grad_norm": 0.4192345992600649, | |
| "learning_rate": 1.4726885051885654e-05, | |
| "loss": 0.5769, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 2.840506329113924, | |
| "grad_norm": 0.4026602216319346, | |
| "learning_rate": 1.4588578757110359e-05, | |
| "loss": 0.5654, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 2.8708860759493673, | |
| "grad_norm": 0.4231047726203865, | |
| "learning_rate": 1.4449151536240167e-05, | |
| "loss": 0.5607, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 2.90126582278481, | |
| "grad_norm": 0.4561926445783001, | |
| "learning_rate": 1.4308637449409705e-05, | |
| "loss": 0.5569, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 2.9316455696202532, | |
| "grad_norm": 0.4717803071311705, | |
| "learning_rate": 1.4167070822259868e-05, | |
| "loss": 0.5675, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.962025316455696, | |
| "grad_norm": 0.47933056765931364, | |
| "learning_rate": 1.402448623755254e-05, | |
| "loss": 0.5748, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 2.992405063291139, | |
| "grad_norm": 0.44783321584350805, | |
| "learning_rate": 1.3880918526722497e-05, | |
| "loss": 0.5695, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 3.030379746835443, | |
| "grad_norm": 0.6027698839810295, | |
| "learning_rate": 1.3736402761368597e-05, | |
| "loss": 0.5002, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 3.060759493670886, | |
| "grad_norm": 0.47122979379018803, | |
| "learning_rate": 1.3590974244686248e-05, | |
| "loss": 0.4881, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.091139240506329, | |
| "grad_norm": 0.7613490099199804, | |
| "learning_rate": 1.344466850284333e-05, | |
| "loss": 0.5085, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 3.1215189873417724, | |
| "grad_norm": 0.520928998078167, | |
| "learning_rate": 1.3297521276301666e-05, | |
| "loss": 0.5094, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 3.151898734177215, | |
| "grad_norm": 0.6134617798185299, | |
| "learning_rate": 1.3149568511086104e-05, | |
| "loss": 0.5194, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 3.1822784810126583, | |
| "grad_norm": 0.5700587181966504, | |
| "learning_rate": 1.300084635000341e-05, | |
| "loss": 0.4885, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 3.212658227848101, | |
| "grad_norm": 0.5416038804300697, | |
| "learning_rate": 1.2851391123813075e-05, | |
| "loss": 0.5012, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 3.2430379746835443, | |
| "grad_norm": 0.6040375356128672, | |
| "learning_rate": 1.2701239342352223e-05, | |
| "loss": 0.4982, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 3.2734177215189875, | |
| "grad_norm": 0.4286650388387901, | |
| "learning_rate": 1.2550427685616767e-05, | |
| "loss": 0.5037, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 3.3037974683544302, | |
| "grad_norm": 0.5392498559348753, | |
| "learning_rate": 1.239899299480098e-05, | |
| "loss": 0.5098, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 3.3341772151898734, | |
| "grad_norm": 0.3832056056453495, | |
| "learning_rate": 1.2246972263297718e-05, | |
| "loss": 0.4985, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 3.3645569620253166, | |
| "grad_norm": 0.449692481880837, | |
| "learning_rate": 1.2094402627661447e-05, | |
| "loss": 0.498, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 3.3949367088607594, | |
| "grad_norm": 0.5046176322673145, | |
| "learning_rate": 1.1941321358536278e-05, | |
| "loss": 0.4983, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 3.4253164556962026, | |
| "grad_norm": 0.45430661893828217, | |
| "learning_rate": 1.1787765851551296e-05, | |
| "loss": 0.5066, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 3.4556962025316453, | |
| "grad_norm": 0.4165469219387865, | |
| "learning_rate": 1.1633773618185302e-05, | |
| "loss": 0.4945, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 3.4860759493670885, | |
| "grad_norm": 0.41753251272697983, | |
| "learning_rate": 1.14793822766033e-05, | |
| "loss": 0.5074, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 3.5164556962025317, | |
| "grad_norm": 0.444995536317105, | |
| "learning_rate": 1.132462954246688e-05, | |
| "loss": 0.4973, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 3.546835443037975, | |
| "grad_norm": 0.42212881923423234, | |
| "learning_rate": 1.1169553219720828e-05, | |
| "loss": 0.4923, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 3.5772151898734177, | |
| "grad_norm": 0.40515660969534983, | |
| "learning_rate": 1.1014191191358118e-05, | |
| "loss": 0.4964, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 3.607594936708861, | |
| "grad_norm": 0.39114719722811153, | |
| "learning_rate": 1.085858141016566e-05, | |
| "loss": 0.4961, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 3.6379746835443036, | |
| "grad_norm": 0.381320402199478, | |
| "learning_rate": 1.070276188945293e-05, | |
| "loss": 0.5019, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 3.668354430379747, | |
| "grad_norm": 0.42482394074709817, | |
| "learning_rate": 1.0546770693765859e-05, | |
| "loss": 0.4872, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 3.69873417721519, | |
| "grad_norm": 0.37351415803234844, | |
| "learning_rate": 1.0390645929588197e-05, | |
| "loss": 0.488, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 3.729113924050633, | |
| "grad_norm": 0.38466472434979115, | |
| "learning_rate": 1.0234425736032607e-05, | |
| "loss": 0.4882, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 3.759493670886076, | |
| "grad_norm": 0.46027902283201394, | |
| "learning_rate": 1.007814827552384e-05, | |
| "loss": 0.5146, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 3.7898734177215188, | |
| "grad_norm": 0.3704780459009314, | |
| "learning_rate": 9.92185172447616e-06, | |
| "loss": 0.5085, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 3.820253164556962, | |
| "grad_norm": 0.37003880424132546, | |
| "learning_rate": 9.765574263967397e-06, | |
| "loss": 0.4964, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 3.850632911392405, | |
| "grad_norm": 0.399412471499315, | |
| "learning_rate": 9.609354070411807e-06, | |
| "loss": 0.5028, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 3.8810126582278484, | |
| "grad_norm": 0.4018865783854977, | |
| "learning_rate": 9.453229306234143e-06, | |
| "loss": 0.5165, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 3.911392405063291, | |
| "grad_norm": 0.3514320852494092, | |
| "learning_rate": 9.297238110547075e-06, | |
| "loss": 0.509, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 3.9417721518987343, | |
| "grad_norm": 0.3269049789045622, | |
| "learning_rate": 9.14141858983434e-06, | |
| "loss": 0.4985, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 3.972151898734177, | |
| "grad_norm": 0.36807340944942746, | |
| "learning_rate": 8.985808808641883e-06, | |
| "loss": 0.5018, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 4.010126582278481, | |
| "grad_norm": 0.4476848872633152, | |
| "learning_rate": 8.830446780279175e-06, | |
| "loss": 0.4854, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 4.040506329113924, | |
| "grad_norm": 0.47668151312796203, | |
| "learning_rate": 8.675370457533122e-06, | |
| "loss": 0.4629, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 4.0708860759493675, | |
| "grad_norm": 0.4005943826954106, | |
| "learning_rate": 8.520617723396702e-06, | |
| "loss": 0.4672, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 4.10126582278481, | |
| "grad_norm": 0.5348145819368196, | |
| "learning_rate": 8.366226381814698e-06, | |
| "loss": 0.4474, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 4.131645569620253, | |
| "grad_norm": 0.42181366124961445, | |
| "learning_rate": 8.212234148448708e-06, | |
| "loss": 0.4557, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 4.162025316455696, | |
| "grad_norm": 1.825786132148733, | |
| "learning_rate": 8.058678641463724e-06, | |
| "loss": 0.4482, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 4.192405063291139, | |
| "grad_norm": 0.5376789086736027, | |
| "learning_rate": 7.905597372338558e-06, | |
| "loss": 0.4431, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 4.222784810126583, | |
| "grad_norm": 0.41721076379719535, | |
| "learning_rate": 7.753027736702283e-06, | |
| "loss": 0.4622, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 4.253164556962025, | |
| "grad_norm": 0.434563369923113, | |
| "learning_rate": 7.601007005199022e-06, | |
| "loss": 0.4281, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 4.283544303797468, | |
| "grad_norm": 0.4918042827822312, | |
| "learning_rate": 7.449572314383237e-06, | |
| "loss": 0.4452, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 4.313924050632911, | |
| "grad_norm": 0.4336657726144659, | |
| "learning_rate": 7.298760657647779e-06, | |
| "loss": 0.4518, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 4.3443037974683545, | |
| "grad_norm": 0.41398522646014463, | |
| "learning_rate": 7.148608876186931e-06, | |
| "loss": 0.4558, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 4.374683544303798, | |
| "grad_norm": 0.43359735291126783, | |
| "learning_rate": 6.999153649996595e-06, | |
| "loss": 0.4536, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 4.405063291139241, | |
| "grad_norm": 0.4619128693249177, | |
| "learning_rate": 6.8504314889138956e-06, | |
| "loss": 0.4288, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 4.435443037974683, | |
| "grad_norm": 0.34784626922480777, | |
| "learning_rate": 6.702478723698336e-06, | |
| "loss": 0.4552, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 4.465822784810126, | |
| "grad_norm": 0.4553641272180987, | |
| "learning_rate": 6.555331497156671e-06, | |
| "loss": 0.4445, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 4.49620253164557, | |
| "grad_norm": 0.4469119907381613, | |
| "learning_rate": 6.4090257553137566e-06, | |
| "loss": 0.457, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 4.526582278481013, | |
| "grad_norm": 0.34422159170971595, | |
| "learning_rate": 6.263597238631405e-06, | |
| "loss": 0.4538, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 4.556962025316456, | |
| "grad_norm": 0.38035636093067965, | |
| "learning_rate": 6.119081473277502e-06, | |
| "loss": 0.4462, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 4.587341772151898, | |
| "grad_norm": 0.37029663778369426, | |
| "learning_rate": 5.975513762447465e-06, | |
| "loss": 0.4528, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 4.6177215189873415, | |
| "grad_norm": 0.3723435090574529, | |
| "learning_rate": 5.832929177740134e-06, | |
| "loss": 0.4256, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 4.648101265822785, | |
| "grad_norm": 0.3707032978854105, | |
| "learning_rate": 5.6913625505902966e-06, | |
| "loss": 0.4484, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 4.678481012658228, | |
| "grad_norm": 0.3954110701891792, | |
| "learning_rate": 5.550848463759835e-06, | |
| "loss": 0.4247, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 4.708860759493671, | |
| "grad_norm": 0.3510500965202149, | |
| "learning_rate": 5.411421242889643e-06, | |
| "loss": 0.4471, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 4.739240506329114, | |
| "grad_norm": 0.31995313210451937, | |
| "learning_rate": 5.273114948114346e-06, | |
| "loss": 0.4642, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 4.769620253164557, | |
| "grad_norm": 0.3553972423744872, | |
| "learning_rate": 5.135963365741892e-06, | |
| "loss": 0.4571, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.38384456978626086, | |
| "learning_rate": 5.000000000000003e-06, | |
| "loss": 0.4564, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 4.830379746835443, | |
| "grad_norm": 0.3372345334337055, | |
| "learning_rate": 4.865258064851579e-06, | |
| "loss": 0.4353, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 4.860759493670886, | |
| "grad_norm": 0.3907743908414007, | |
| "learning_rate": 4.731770475880995e-06, | |
| "loss": 0.4221, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 4.891139240506329, | |
| "grad_norm": 0.3552379653902961, | |
| "learning_rate": 4.599569842253244e-06, | |
| "loss": 0.4294, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 4.921518987341772, | |
| "grad_norm": 0.3229343524842094, | |
| "learning_rate": 4.468688458748006e-06, | |
| "loss": 0.4447, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 4.951898734177215, | |
| "grad_norm": 0.3271381778756471, | |
| "learning_rate": 4.339158297870469e-06, | |
| "loss": 0.4377, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 4.982278481012658, | |
| "grad_norm": 0.31530470042725844, | |
| "learning_rate": 4.211011002040885e-06, | |
| "loss": 0.4507, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 5.020253164556962, | |
| "grad_norm": 0.43102783314599225, | |
| "learning_rate": 4.084277875864776e-06, | |
| "loss": 0.436, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 5.050632911392405, | |
| "grad_norm": 0.45380321011962116, | |
| "learning_rate": 3.958989878485644e-06, | |
| "loss": 0.3982, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 5.0810126582278485, | |
| "grad_norm": 0.32251696306626093, | |
| "learning_rate": 3.83517761602209e-06, | |
| "loss": 0.4199, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 5.111392405063291, | |
| "grad_norm": 0.40082421825676645, | |
| "learning_rate": 3.712871334091154e-06, | |
| "loss": 0.3998, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 5.141772151898734, | |
| "grad_norm": 0.5395176017293357, | |
| "learning_rate": 3.592100910419738e-06, | |
| "loss": 0.4291, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 5.172151898734177, | |
| "grad_norm": 0.3771023471567734, | |
| "learning_rate": 3.4728958475459052e-06, | |
| "loss": 0.4099, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 5.2025316455696204, | |
| "grad_norm": 0.38370853829523194, | |
| "learning_rate": 3.355285265611784e-06, | |
| "loss": 0.4284, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 5.232911392405064, | |
| "grad_norm": 0.38204749183223125, | |
| "learning_rate": 3.2392978952499553e-06, | |
| "loss": 0.4154, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 5.263291139240506, | |
| "grad_norm": 0.4028658896509962, | |
| "learning_rate": 3.1249620705649417e-06, | |
| "loss": 0.3972, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 5.293670886075949, | |
| "grad_norm": 0.367728410023284, | |
| "learning_rate": 3.0123057222115835e-06, | |
| "loss": 0.434, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 5.324050632911392, | |
| "grad_norm": 0.3405378323258769, | |
| "learning_rate": 2.9013563705719673e-06, | |
| "loss": 0.4057, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 5.3544303797468356, | |
| "grad_norm": 0.38210954202756636, | |
| "learning_rate": 2.7921411190325753e-06, | |
| "loss": 0.4196, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 5.384810126582279, | |
| "grad_norm": 0.3804944421648198, | |
| "learning_rate": 2.6846866473633126e-06, | |
| "loss": 0.4104, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 5.415189873417722, | |
| "grad_norm": 0.38344342558303635, | |
| "learning_rate": 2.579019205199992e-06, | |
| "loss": 0.3989, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 5.445569620253164, | |
| "grad_norm": 0.3658539967565658, | |
| "learning_rate": 2.4751646056319334e-06, | |
| "loss": 0.4091, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 5.4759493670886075, | |
| "grad_norm": 0.32009110755042075, | |
| "learning_rate": 2.373148218896182e-06, | |
| "loss": 0.4089, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 5.506329113924051, | |
| "grad_norm": 0.2985470548233042, | |
| "learning_rate": 2.2729949661798876e-06, | |
| "loss": 0.4098, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 5.536708860759494, | |
| "grad_norm": 0.3642943319773281, | |
| "learning_rate": 2.174729313532433e-06, | |
| "loss": 0.4224, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 5.567088607594937, | |
| "grad_norm": 0.34493083178157297, | |
| "learning_rate": 2.078375265888707e-06, | |
| "loss": 0.3987, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 5.597468354430379, | |
| "grad_norm": 0.33513796513087296, | |
| "learning_rate": 1.9839563612050273e-06, | |
| "loss": 0.4084, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 5.627848101265823, | |
| "grad_norm": 0.3051378247635565, | |
| "learning_rate": 1.8914956647091497e-06, | |
| "loss": 0.4163, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 5.658227848101266, | |
| "grad_norm": 0.2900500971985913, | |
| "learning_rate": 1.8010157632657544e-06, | |
| "loss": 0.3717, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 5.688607594936709, | |
| "grad_norm": 0.31354238865151135, | |
| "learning_rate": 1.7125387598587862e-06, | |
| "loss": 0.4032, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 5.718987341772152, | |
| "grad_norm": 0.2989580092812493, | |
| "learning_rate": 1.6260862681919965e-06, | |
| "loss": 0.4064, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 5.749367088607595, | |
| "grad_norm": 0.31986705944057237, | |
| "learning_rate": 1.5416794074090258e-06, | |
| "loss": 0.4108, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 5.779746835443038, | |
| "grad_norm": 0.31293963729417645, | |
| "learning_rate": 1.459338796934293e-06, | |
| "loss": 0.4172, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 5.810126582278481, | |
| "grad_norm": 0.3008465072464133, | |
| "learning_rate": 1.3790845514359363e-06, | |
| "loss": 0.3864, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 5.840506329113924, | |
| "grad_norm": 0.2745798591649662, | |
| "learning_rate": 1.300936275912098e-06, | |
| "loss": 0.4131, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 5.870886075949367, | |
| "grad_norm": 0.28942117072123447, | |
| "learning_rate": 1.224913060901688e-06, | |
| "loss": 0.4339, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 5.9012658227848105, | |
| "grad_norm": 0.2906243234871826, | |
| "learning_rate": 1.1510334778208332e-06, | |
| "loss": 0.4208, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 5.931645569620253, | |
| "grad_norm": 0.2816553965543679, | |
| "learning_rate": 1.0793155744261352e-06, | |
| "loss": 0.4245, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 5.962025316455696, | |
| "grad_norm": 0.26782737908368254, | |
| "learning_rate": 1.0097768704058542e-06, | |
| "loss": 0.4121, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 5.992405063291139, | |
| "grad_norm": 0.2568223495211035, | |
| "learning_rate": 9.424343531000968e-07, | |
| "loss": 0.4195, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 6.030379746835443, | |
| "grad_norm": 0.33327305966975557, | |
| "learning_rate": 8.773044733510338e-07, | |
| "loss": 0.3984, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 6.060759493670886, | |
| "grad_norm": 0.3958974520652796, | |
| "learning_rate": 8.144031414842012e-07, | |
| "loss": 0.3928, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 6.091139240506329, | |
| "grad_norm": 0.30057511303828316, | |
| "learning_rate": 7.537457234218271e-07, | |
| "loss": 0.3936, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 6.121518987341772, | |
| "grad_norm": 0.25902413585106376, | |
| "learning_rate": 6.953470369291349e-07, | |
| "loss": 0.4059, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 6.151898734177215, | |
| "grad_norm": 0.25130559715597733, | |
| "learning_rate": 6.392213479945852e-07, | |
| "loss": 0.3897, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 6.182278481012658, | |
| "grad_norm": 0.29432310871264594, | |
| "learning_rate": 5.853823673448877e-07, | |
| "loss": 0.4025, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 6.2126582278481015, | |
| "grad_norm": 0.3081134707492938, | |
| "learning_rate": 5.33843247095659e-07, | |
| "loss": 0.4019, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 6.243037974683545, | |
| "grad_norm": 0.2878787524542973, | |
| "learning_rate": 4.846165775385459e-07, | |
| "loss": 0.3986, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 6.273417721518987, | |
| "grad_norm": 0.31287435718436796, | |
| "learning_rate": 4.3771438406559173e-07, | |
| "loss": 0.391, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 6.30379746835443, | |
| "grad_norm": 0.28354408405502884, | |
| "learning_rate": 3.931481242315993e-07, | |
| "loss": 0.3859, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 6.334177215189873, | |
| "grad_norm": 0.26164536123441023, | |
| "learning_rate": 3.5092868495520294e-07, | |
| "loss": 0.3886, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 6.364556962025317, | |
| "grad_norm": 0.2689329017551278, | |
| "learning_rate": 3.110663798593616e-07, | |
| "loss": 0.4003, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 6.39493670886076, | |
| "grad_norm": 0.2580536611966275, | |
| "learning_rate": 2.735709467518699e-07, | |
| "loss": 0.3985, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 6.425316455696202, | |
| "grad_norm": 0.2558832526897395, | |
| "learning_rate": 2.384515452465475e-07, | |
| "loss": 0.417, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 6.455696202531645, | |
| "grad_norm": 0.2534803170400261, | |
| "learning_rate": 2.0571675452567997e-07, | |
| "loss": 0.3801, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 6.4860759493670885, | |
| "grad_norm": 0.2541264043125379, | |
| "learning_rate": 1.7537457124423896e-07, | |
| "loss": 0.4162, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 6.516455696202532, | |
| "grad_norm": 0.25593449982611427, | |
| "learning_rate": 1.474324075764111e-07, | |
| "loss": 0.3942, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 6.546835443037975, | |
| "grad_norm": 0.2606703943702357, | |
| "learning_rate": 1.2189708940490653e-07, | |
| "loss": 0.3851, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 6.577215189873417, | |
| "grad_norm": 0.2574309323319613, | |
| "learning_rate": 9.877485465349057e-08, | |
| "loss": 0.3619, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 6.6075949367088604, | |
| "grad_norm": 0.261236633339879, | |
| "learning_rate": 7.807135176314707e-08, | |
| "loss": 0.3913, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 6.637974683544304, | |
| "grad_norm": 0.25009804468977803, | |
| "learning_rate": 5.979163831223988e-08, | |
| "loss": 0.4017, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 6.668354430379747, | |
| "grad_norm": 0.27347020181636067, | |
| "learning_rate": 4.394017978101905e-08, | |
| "loss": 0.3847, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 6.69873417721519, | |
| "grad_norm": 0.25723432049666295, | |
| "learning_rate": 3.0520848460765525e-08, | |
| "loss": 0.4028, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 6.729113924050633, | |
| "grad_norm": 0.2434414077583206, | |
| "learning_rate": 1.9536922507841227e-08, | |
| "loss": 0.3884, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 6.759493670886076, | |
| "grad_norm": 0.2455651597557325, | |
| "learning_rate": 1.099108514288627e-08, | |
| "loss": 0.3895, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 6.789873417721519, | |
| "grad_norm": 0.24137450253260526, | |
| "learning_rate": 4.885423995341088e-09, | |
| "loss": 0.4012, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 6.820253164556962, | |
| "grad_norm": 0.25657012607230134, | |
| "learning_rate": 1.2214305934699078e-09, | |
| "loss": 0.397, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 6.850632911392405, | |
| "grad_norm": 0.24651847198168722, | |
| "learning_rate": 0.0, | |
| "loss": 0.4046, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 6.850632911392405, | |
| "step": 224, | |
| "total_flos": 8.947189002629284e+17, | |
| "train_loss": 0.5461564424580762, | |
| "train_runtime": 27730.2744, | |
| "train_samples_per_second": 0.798, | |
| "train_steps_per_second": 0.008 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 224, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 7, | |
| "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": 8.947189002629284e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |