| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9838509316770185, | |
| "eval_steps": 500, | |
| "global_step": 402, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.007453416149068323, | |
| "grad_norm": 0.9149440636536349, | |
| "learning_rate": 2.3809523809523808e-06, | |
| "loss": 0.5267, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.014906832298136646, | |
| "grad_norm": 0.813081054678093, | |
| "learning_rate": 4.7619047619047615e-06, | |
| "loss": 0.501, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.02236024844720497, | |
| "grad_norm": 0.7030352003534567, | |
| "learning_rate": 7.142857142857143e-06, | |
| "loss": 0.4786, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.02981366459627329, | |
| "grad_norm": 0.744974289676234, | |
| "learning_rate": 9.523809523809523e-06, | |
| "loss": 0.5096, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.037267080745341616, | |
| "grad_norm": 0.5298082392046644, | |
| "learning_rate": 1.1904761904761905e-05, | |
| "loss": 0.4947, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.04472049689440994, | |
| "grad_norm": 0.6276792078680746, | |
| "learning_rate": 1.4285714285714285e-05, | |
| "loss": 0.4889, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.05217391304347826, | |
| "grad_norm": 0.9938211412587247, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 0.4966, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.05962732919254658, | |
| "grad_norm": 1.0688888961766796, | |
| "learning_rate": 1.9047619047619046e-05, | |
| "loss": 0.5423, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0670807453416149, | |
| "grad_norm": 0.7660208255317018, | |
| "learning_rate": 2.1428571428571428e-05, | |
| "loss": 0.5054, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.07453416149068323, | |
| "grad_norm": 0.6232752556675165, | |
| "learning_rate": 2.380952380952381e-05, | |
| "loss": 0.516, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.08198757763975155, | |
| "grad_norm": 0.5696191527866266, | |
| "learning_rate": 2.6190476190476192e-05, | |
| "loss": 0.4902, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.08944099378881988, | |
| "grad_norm": 0.5574577130188229, | |
| "learning_rate": 2.857142857142857e-05, | |
| "loss": 0.5054, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.0968944099378882, | |
| "grad_norm": 0.5992840639020023, | |
| "learning_rate": 3.095238095238095e-05, | |
| "loss": 0.5008, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.10434782608695652, | |
| "grad_norm": 0.5954549641006874, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.5023, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.11180124223602485, | |
| "grad_norm": 0.576921385529578, | |
| "learning_rate": 3.571428571428572e-05, | |
| "loss": 0.5236, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.11925465838509317, | |
| "grad_norm": 0.5706806482628634, | |
| "learning_rate": 3.809523809523809e-05, | |
| "loss": 0.5017, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.1267080745341615, | |
| "grad_norm": 0.7086057812964052, | |
| "learning_rate": 4.047619047619048e-05, | |
| "loss": 0.5267, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.1341614906832298, | |
| "grad_norm": 0.5808757241964403, | |
| "learning_rate": 4.2857142857142856e-05, | |
| "loss": 0.4904, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.14161490683229813, | |
| "grad_norm": 0.6317506464382385, | |
| "learning_rate": 4.523809523809524e-05, | |
| "loss": 0.5043, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.14906832298136646, | |
| "grad_norm": 0.6359839242618298, | |
| "learning_rate": 4.761904761904762e-05, | |
| "loss": 0.5055, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1565217391304348, | |
| "grad_norm": 0.5752065827486615, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5084, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.1639751552795031, | |
| "grad_norm": 0.574109201187312, | |
| "learning_rate": 4.999923510846293e-05, | |
| "loss": 0.4981, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.17142857142857143, | |
| "grad_norm": 0.5996134853952727, | |
| "learning_rate": 4.999694048585699e-05, | |
| "loss": 0.4889, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.17888198757763976, | |
| "grad_norm": 0.6523805926616552, | |
| "learning_rate": 4.999311628819437e-05, | |
| "loss": 0.4927, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.18633540372670807, | |
| "grad_norm": 0.7101608078052835, | |
| "learning_rate": 4.9987762775483666e-05, | |
| "loss": 0.5078, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.1937888198757764, | |
| "grad_norm": 0.7396307593577484, | |
| "learning_rate": 4.99808803117121e-05, | |
| "loss": 0.4943, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.20124223602484473, | |
| "grad_norm": 0.653880275554884, | |
| "learning_rate": 4.9972469364820877e-05, | |
| "loss": 0.4981, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.20869565217391303, | |
| "grad_norm": 0.9242693837414091, | |
| "learning_rate": 4.996253050667329e-05, | |
| "loss": 0.4875, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.21614906832298136, | |
| "grad_norm": 1.2095187282407587, | |
| "learning_rate": 4.995106441301587e-05, | |
| "loss": 0.5071, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.2236024844720497, | |
| "grad_norm": 0.7405451123886392, | |
| "learning_rate": 4.993807186343243e-05, | |
| "loss": 0.5168, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.231055900621118, | |
| "grad_norm": 0.6770762633940498, | |
| "learning_rate": 4.992355374129109e-05, | |
| "loss": 0.4956, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.23850931677018633, | |
| "grad_norm": 0.8790648152676741, | |
| "learning_rate": 4.990751103368418e-05, | |
| "loss": 0.4914, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.24596273291925466, | |
| "grad_norm": 0.8761927581585033, | |
| "learning_rate": 4.988994483136115e-05, | |
| "loss": 0.5028, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.253416149068323, | |
| "grad_norm": 1.0155828765501556, | |
| "learning_rate": 4.98708563286544e-05, | |
| "loss": 0.497, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.2608695652173913, | |
| "grad_norm": 0.7164731439245322, | |
| "learning_rate": 4.985024682339807e-05, | |
| "loss": 0.4782, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.2683229813664596, | |
| "grad_norm": 0.7789231779550446, | |
| "learning_rate": 4.982811771683982e-05, | |
| "loss": 0.5117, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.27577639751552796, | |
| "grad_norm": 0.7333852855548396, | |
| "learning_rate": 4.980447051354555e-05, | |
| "loss": 0.5147, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.28322981366459626, | |
| "grad_norm": 0.6356599304692622, | |
| "learning_rate": 4.977930682129711e-05, | |
| "loss": 0.4986, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.2906832298136646, | |
| "grad_norm": 0.6005754618090398, | |
| "learning_rate": 4.975262835098295e-05, | |
| "loss": 0.4853, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.2981366459627329, | |
| "grad_norm": 0.6269613620581922, | |
| "learning_rate": 4.9724436916481836e-05, | |
| "loss": 0.4842, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.30559006211180123, | |
| "grad_norm": 0.4930016834035119, | |
| "learning_rate": 4.969473443453953e-05, | |
| "loss": 0.5001, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.3130434782608696, | |
| "grad_norm": 0.6087046909213417, | |
| "learning_rate": 4.966352292463842e-05, | |
| "loss": 0.4903, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.3204968944099379, | |
| "grad_norm": 0.6680768802723248, | |
| "learning_rate": 4.96308045088603e-05, | |
| "loss": 0.4864, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.3279503105590062, | |
| "grad_norm": 0.5024028870492778, | |
| "learning_rate": 4.959658141174196e-05, | |
| "loss": 0.4928, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.33540372670807456, | |
| "grad_norm": 0.5691045158483018, | |
| "learning_rate": 4.956085596012407e-05, | |
| "loss": 0.4895, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.34285714285714286, | |
| "grad_norm": 0.5960753346949116, | |
| "learning_rate": 4.952363058299291e-05, | |
| "loss": 0.5037, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.35031055900621116, | |
| "grad_norm": 0.5809783536316744, | |
| "learning_rate": 4.948490781131523e-05, | |
| "loss": 0.4955, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.3577639751552795, | |
| "grad_norm": 0.4638121725312366, | |
| "learning_rate": 4.944469027786616e-05, | |
| "loss": 0.4828, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.3652173913043478, | |
| "grad_norm": 0.6519884038596414, | |
| "learning_rate": 4.940298071705024e-05, | |
| "loss": 0.4869, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.37267080745341613, | |
| "grad_norm": 0.725573010046233, | |
| "learning_rate": 4.935978196471548e-05, | |
| "loss": 0.4983, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3801242236024845, | |
| "grad_norm": 0.7723414020223717, | |
| "learning_rate": 4.931509695796055e-05, | |
| "loss": 0.5097, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.3875776397515528, | |
| "grad_norm": 0.6749832378781871, | |
| "learning_rate": 4.926892873493509e-05, | |
| "loss": 0.5009, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.3950310559006211, | |
| "grad_norm": 0.5287344971699504, | |
| "learning_rate": 4.922128043463316e-05, | |
| "loss": 0.5033, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.40248447204968946, | |
| "grad_norm": 0.7590242284157045, | |
| "learning_rate": 4.917215529667979e-05, | |
| "loss": 0.5041, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.40993788819875776, | |
| "grad_norm": 0.9634027388127688, | |
| "learning_rate": 4.9121556661110734e-05, | |
| "loss": 0.5084, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.41739130434782606, | |
| "grad_norm": 0.8942287814733841, | |
| "learning_rate": 4.906948796814539e-05, | |
| "loss": 0.4982, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.4248447204968944, | |
| "grad_norm": 0.7332394854708344, | |
| "learning_rate": 4.901595275795287e-05, | |
| "loss": 0.4893, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.4322981366459627, | |
| "grad_norm": 0.6811910538390739, | |
| "learning_rate": 4.896095467041133e-05, | |
| "loss": 0.498, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.43975155279503103, | |
| "grad_norm": 1.0815817510677823, | |
| "learning_rate": 4.8904497444860495e-05, | |
| "loss": 0.4951, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.4472049689440994, | |
| "grad_norm": 0.5430186184806628, | |
| "learning_rate": 4.884658491984735e-05, | |
| "loss": 0.4974, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.4546583850931677, | |
| "grad_norm": 0.7264258373815644, | |
| "learning_rate": 4.878722103286529e-05, | |
| "loss": 0.5017, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.462111801242236, | |
| "grad_norm": 0.7414704944298521, | |
| "learning_rate": 4.8726409820086274e-05, | |
| "loss": 0.502, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.46956521739130436, | |
| "grad_norm": 0.523437918394872, | |
| "learning_rate": 4.8664155416086495e-05, | |
| "loss": 0.4957, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.47701863354037266, | |
| "grad_norm": 0.6159908506967338, | |
| "learning_rate": 4.8600462053565204e-05, | |
| "loss": 0.4956, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.484472049689441, | |
| "grad_norm": 0.5860033924334783, | |
| "learning_rate": 4.8535334063056984e-05, | |
| "loss": 0.5036, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.4919254658385093, | |
| "grad_norm": 0.510141066708458, | |
| "learning_rate": 4.846877587263728e-05, | |
| "loss": 0.4862, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.4993788819875776, | |
| "grad_norm": 0.5214414519844613, | |
| "learning_rate": 4.840079200762135e-05, | |
| "loss": 0.4728, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.506832298136646, | |
| "grad_norm": 0.5627288353125691, | |
| "learning_rate": 4.833138709025656e-05, | |
| "loss": 0.49, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.5142857142857142, | |
| "grad_norm": 0.6512461745013858, | |
| "learning_rate": 4.826056583940815e-05, | |
| "loss": 0.4763, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.5217391304347826, | |
| "grad_norm": 0.501857537261379, | |
| "learning_rate": 4.818833307023839e-05, | |
| "loss": 0.4801, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.529192546583851, | |
| "grad_norm": 0.5401175474298149, | |
| "learning_rate": 4.811469369387917e-05, | |
| "loss": 0.5099, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.5366459627329192, | |
| "grad_norm": 0.6056173499301607, | |
| "learning_rate": 4.803965271709811e-05, | |
| "loss": 0.4755, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.5440993788819876, | |
| "grad_norm": 0.5300538292613007, | |
| "learning_rate": 4.7963215241958156e-05, | |
| "loss": 0.4914, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.5515527950310559, | |
| "grad_norm": 0.49477384366828114, | |
| "learning_rate": 4.7885386465470675e-05, | |
| "loss": 0.4942, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.5590062111801242, | |
| "grad_norm": 0.5832215591212782, | |
| "learning_rate": 4.780617167924209e-05, | |
| "loss": 0.4821, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.5664596273291925, | |
| "grad_norm": 0.5825547831853878, | |
| "learning_rate": 4.772557626911417e-05, | |
| "loss": 0.4991, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.5739130434782609, | |
| "grad_norm": 0.6260775830743696, | |
| "learning_rate": 4.7643605714797754e-05, | |
| "loss": 0.5178, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.5813664596273292, | |
| "grad_norm": 0.5002047281833832, | |
| "learning_rate": 4.756026558950025e-05, | |
| "loss": 0.4788, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.5888198757763975, | |
| "grad_norm": 0.658164441734046, | |
| "learning_rate": 4.747556155954668e-05, | |
| "loss": 0.4924, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.5962732919254659, | |
| "grad_norm": 0.5642052228011414, | |
| "learning_rate": 4.738949938399445e-05, | |
| "loss": 0.4793, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.6037267080745342, | |
| "grad_norm": 0.6800799125862594, | |
| "learning_rate": 4.730208491424174e-05, | |
| "loss": 0.4972, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.6111801242236025, | |
| "grad_norm": 0.6022489226977471, | |
| "learning_rate": 4.721332409362973e-05, | |
| "loss": 0.4839, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.6186335403726708, | |
| "grad_norm": 0.5548484325825994, | |
| "learning_rate": 4.712322295703846e-05, | |
| "loss": 0.4786, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.6260869565217392, | |
| "grad_norm": 0.5717126912005164, | |
| "learning_rate": 4.7031787630476544e-05, | |
| "loss": 0.4938, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.6335403726708074, | |
| "grad_norm": 0.5272588715397946, | |
| "learning_rate": 4.693902433066464e-05, | |
| "loss": 0.4793, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.6409937888198758, | |
| "grad_norm": 0.5463413109564844, | |
| "learning_rate": 4.6844939364612796e-05, | |
| "loss": 0.4925, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.6484472049689441, | |
| "grad_norm": 0.6075766167098049, | |
| "learning_rate": 4.674953912919161e-05, | |
| "loss": 0.5025, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.6559006211180124, | |
| "grad_norm": 0.6310158980679005, | |
| "learning_rate": 4.665283011069733e-05, | |
| "loss": 0.4911, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.6633540372670808, | |
| "grad_norm": 0.50469934840197, | |
| "learning_rate": 4.655481888441083e-05, | |
| "loss": 0.4764, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.6708074534161491, | |
| "grad_norm": 0.488371832002747, | |
| "learning_rate": 4.6455512114150546e-05, | |
| "loss": 0.4937, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.6782608695652174, | |
| "grad_norm": 0.4770981226634187, | |
| "learning_rate": 4.635491655181943e-05, | |
| "loss": 0.4884, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.6857142857142857, | |
| "grad_norm": 0.5775420955722163, | |
| "learning_rate": 4.625303903694586e-05, | |
| "loss": 0.4853, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.6931677018633541, | |
| "grad_norm": 0.47430309169555146, | |
| "learning_rate": 4.614988649621864e-05, | |
| "loss": 0.5139, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.7006211180124223, | |
| "grad_norm": 0.6638176198018706, | |
| "learning_rate": 4.604546594301602e-05, | |
| "loss": 0.4953, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.7080745341614907, | |
| "grad_norm": 0.5170361990821142, | |
| "learning_rate": 4.5939784476928894e-05, | |
| "loss": 0.4714, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.715527950310559, | |
| "grad_norm": 0.5989923841003543, | |
| "learning_rate": 4.583284928327805e-05, | |
| "loss": 0.4826, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.7229813664596273, | |
| "grad_norm": 0.5119408388808395, | |
| "learning_rate": 4.572466763262569e-05, | |
| "loss": 0.4876, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.7304347826086957, | |
| "grad_norm": 0.4719741454775879, | |
| "learning_rate": 4.5615246880281066e-05, | |
| "loss": 0.4746, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.737888198757764, | |
| "grad_norm": 0.5792116907988438, | |
| "learning_rate": 4.550459446580039e-05, | |
| "loss": 0.4778, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.7453416149068323, | |
| "grad_norm": 0.5650326179874091, | |
| "learning_rate": 4.5392717912481025e-05, | |
| "loss": 0.5068, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.7527950310559006, | |
| "grad_norm": 0.49349361289777244, | |
| "learning_rate": 4.527962482684998e-05, | |
| "loss": 0.4787, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.760248447204969, | |
| "grad_norm": 0.590606894561532, | |
| "learning_rate": 4.516532289814674e-05, | |
| "loss": 0.4827, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.7677018633540372, | |
| "grad_norm": 0.6325065592193169, | |
| "learning_rate": 4.5049819897800445e-05, | |
| "loss": 0.4822, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.7751552795031056, | |
| "grad_norm": 0.48212618053402273, | |
| "learning_rate": 4.493312367890154e-05, | |
| "loss": 0.4862, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.782608695652174, | |
| "grad_norm": 0.6659769851227831, | |
| "learning_rate": 4.481524217566783e-05, | |
| "loss": 0.4763, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.7900621118012422, | |
| "grad_norm": 0.5100428848300065, | |
| "learning_rate": 4.4696183402905005e-05, | |
| "loss": 0.4861, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.7975155279503106, | |
| "grad_norm": 0.4930498732601481, | |
| "learning_rate": 4.4575955455461764e-05, | |
| "loss": 0.4709, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.8049689440993789, | |
| "grad_norm": 0.5557882232422525, | |
| "learning_rate": 4.4454566507679395e-05, | |
| "loss": 0.4823, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.8124223602484472, | |
| "grad_norm": 0.4779459811311798, | |
| "learning_rate": 4.4332024812836026e-05, | |
| "loss": 0.4902, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.8198757763975155, | |
| "grad_norm": 0.4951096303075202, | |
| "learning_rate": 4.420833870258544e-05, | |
| "loss": 0.4799, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.8273291925465839, | |
| "grad_norm": 0.4972554983396596, | |
| "learning_rate": 4.4083516586390694e-05, | |
| "loss": 0.4757, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.8347826086956521, | |
| "grad_norm": 0.6922157872179194, | |
| "learning_rate": 4.395756695095222e-05, | |
| "loss": 0.4871, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.8422360248447205, | |
| "grad_norm": 0.614533156193526, | |
| "learning_rate": 4.383049835963095e-05, | |
| "loss": 0.4858, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.8496894409937888, | |
| "grad_norm": 0.4998865414356172, | |
| "learning_rate": 4.370231945186601e-05, | |
| "loss": 0.5084, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 0.5297858251800804, | |
| "learning_rate": 4.357303894258733e-05, | |
| "loss": 0.4844, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.8645962732919255, | |
| "grad_norm": 0.490173761162275, | |
| "learning_rate": 4.344266562162313e-05, | |
| "loss": 0.4775, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.8720496894409938, | |
| "grad_norm": 0.5188647432432411, | |
| "learning_rate": 4.331120835310228e-05, | |
| "loss": 0.4731, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.8795031055900621, | |
| "grad_norm": 0.5006594475547541, | |
| "learning_rate": 4.3178676074851646e-05, | |
| "loss": 0.4834, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.8869565217391304, | |
| "grad_norm": 0.6966182509694935, | |
| "learning_rate": 4.3045077797788386e-05, | |
| "loss": 0.482, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.8944099378881988, | |
| "grad_norm": 0.630693615513719, | |
| "learning_rate": 4.29104226053073e-05, | |
| "loss": 0.4806, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.901863354037267, | |
| "grad_norm": 0.4855641171083717, | |
| "learning_rate": 4.277471965266325e-05, | |
| "loss": 0.4881, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.9093167701863354, | |
| "grad_norm": 0.6046785410123138, | |
| "learning_rate": 4.2637978166348666e-05, | |
| "loss": 0.4985, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.9167701863354037, | |
| "grad_norm": 0.6094452228967235, | |
| "learning_rate": 4.250020744346629e-05, | |
| "loss": 0.488, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.924223602484472, | |
| "grad_norm": 0.5207025073876308, | |
| "learning_rate": 4.2361416851097e-05, | |
| "loss": 0.4839, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.9316770186335404, | |
| "grad_norm": 0.5910466009416144, | |
| "learning_rate": 4.222161582566299e-05, | |
| "loss": 0.4816, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.9391304347826087, | |
| "grad_norm": 0.6533302840950863, | |
| "learning_rate": 4.208081387228612e-05, | |
| "loss": 0.4848, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.9465838509316771, | |
| "grad_norm": 0.5615426924165942, | |
| "learning_rate": 4.193902056414175e-05, | |
| "loss": 0.4952, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.9540372670807453, | |
| "grad_norm": 0.5584154804823008, | |
| "learning_rate": 4.179624554180778e-05, | |
| "loss": 0.4998, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.9614906832298137, | |
| "grad_norm": 0.6235842837428364, | |
| "learning_rate": 4.165249851260921e-05, | |
| "loss": 0.4918, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.968944099378882, | |
| "grad_norm": 0.5109516402930713, | |
| "learning_rate": 4.1507789249958134e-05, | |
| "loss": 0.4812, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.9763975155279503, | |
| "grad_norm": 0.5909881028917956, | |
| "learning_rate": 4.136212759268926e-05, | |
| "loss": 0.4713, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.9838509316770186, | |
| "grad_norm": 0.8317014634966002, | |
| "learning_rate": 4.121552344439093e-05, | |
| "loss": 0.5024, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.991304347826087, | |
| "grad_norm": 0.49067977677281605, | |
| "learning_rate": 4.1067986772731795e-05, | |
| "loss": 0.4816, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.9987577639751553, | |
| "grad_norm": 0.6766338838361023, | |
| "learning_rate": 4.0919527608783105e-05, | |
| "loss": 0.477, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.6766338838361023, | |
| "learning_rate": 4.077015604633669e-05, | |
| "loss": 0.4422, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.0074534161490682, | |
| "grad_norm": 1.3875151975518343, | |
| "learning_rate": 4.0619882241218684e-05, | |
| "loss": 0.3755, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.0149068322981367, | |
| "grad_norm": 0.7353931983255126, | |
| "learning_rate": 4.046871641059903e-05, | |
| "loss": 0.3713, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.022360248447205, | |
| "grad_norm": 0.7602599418526752, | |
| "learning_rate": 4.031666883229678e-05, | |
| "loss": 0.3719, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.0298136645962732, | |
| "grad_norm": 0.5465117063921796, | |
| "learning_rate": 4.016374984408137e-05, | |
| "loss": 0.3602, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.0372670807453417, | |
| "grad_norm": 0.5515849784244337, | |
| "learning_rate": 4.000996984296967e-05, | |
| "loss": 0.3669, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.04472049689441, | |
| "grad_norm": 0.547418960473341, | |
| "learning_rate": 3.985533928451914e-05, | |
| "loss": 0.3565, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.0521739130434782, | |
| "grad_norm": 0.43709687343518044, | |
| "learning_rate": 3.969986868211693e-05, | |
| "loss": 0.3592, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.0596273291925467, | |
| "grad_norm": 0.510533171129714, | |
| "learning_rate": 3.954356860626509e-05, | |
| "loss": 0.3548, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.067080745341615, | |
| "grad_norm": 0.7014866699955652, | |
| "learning_rate": 3.938644968386188e-05, | |
| "loss": 0.3458, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.0745341614906831, | |
| "grad_norm": 0.6054199711863072, | |
| "learning_rate": 3.922852259747921e-05, | |
| "loss": 0.3717, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.0819875776397516, | |
| "grad_norm": 0.5885498856638683, | |
| "learning_rate": 3.9069798084636357e-05, | |
| "loss": 0.3563, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.0894409937888199, | |
| "grad_norm": 0.6718059622184547, | |
| "learning_rate": 3.8910286937069894e-05, | |
| "loss": 0.3396, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.0968944099378881, | |
| "grad_norm": 0.6261142587490304, | |
| "learning_rate": 3.875e-05, | |
| "loss": 0.3675, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.1043478260869566, | |
| "grad_norm": 0.7988169272624631, | |
| "learning_rate": 3.858894817139304e-05, | |
| "loss": 0.3475, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.1118012422360248, | |
| "grad_norm": 0.5931174308761531, | |
| "learning_rate": 3.8427142401220634e-05, | |
| "loss": 0.3475, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.119254658385093, | |
| "grad_norm": 0.6442383924397598, | |
| "learning_rate": 3.8264593690715165e-05, | |
| "loss": 0.3497, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.1267080745341616, | |
| "grad_norm": 0.5725732820991489, | |
| "learning_rate": 3.810131309162178e-05, | |
| "loss": 0.3559, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.1341614906832298, | |
| "grad_norm": 0.5564507237104445, | |
| "learning_rate": 3.7937311705447016e-05, | |
| "loss": 0.3626, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.141614906832298, | |
| "grad_norm": 0.6381547420777212, | |
| "learning_rate": 3.777260068270396e-05, | |
| "loss": 0.3405, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.1490683229813665, | |
| "grad_norm": 0.5057416039471847, | |
| "learning_rate": 3.760719122215416e-05, | |
| "loss": 0.3524, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.1565217391304348, | |
| "grad_norm": 0.5412142450588553, | |
| "learning_rate": 3.74410945700462e-05, | |
| "loss": 0.3513, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.163975155279503, | |
| "grad_norm": 0.5764227357077137, | |
| "learning_rate": 3.727432201935107e-05, | |
| "loss": 0.349, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.1714285714285715, | |
| "grad_norm": 0.43438798608153883, | |
| "learning_rate": 3.7106884908994314e-05, | |
| "loss": 0.3526, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.1788819875776397, | |
| "grad_norm": 0.4757449865488938, | |
| "learning_rate": 3.693879462308516e-05, | |
| "loss": 0.3554, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.186335403726708, | |
| "grad_norm": 0.5422686029816232, | |
| "learning_rate": 3.677006259014247e-05, | |
| "loss": 0.3449, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.1937888198757765, | |
| "grad_norm": 0.4828269978461137, | |
| "learning_rate": 3.6600700282317704e-05, | |
| "loss": 0.3442, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.2012422360248447, | |
| "grad_norm": 0.6064830533601335, | |
| "learning_rate": 3.643071921461497e-05, | |
| "loss": 0.3602, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.208695652173913, | |
| "grad_norm": 0.37720957669313665, | |
| "learning_rate": 3.626013094410803e-05, | |
| "loss": 0.3632, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.2161490683229814, | |
| "grad_norm": 0.5304204188881894, | |
| "learning_rate": 3.6088947069154624e-05, | |
| "loss": 0.3555, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.2236024844720497, | |
| "grad_norm": 0.38241688724581824, | |
| "learning_rate": 3.591717922860785e-05, | |
| "loss": 0.3461, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.231055900621118, | |
| "grad_norm": 0.4132634739723132, | |
| "learning_rate": 3.574483910102481e-05, | |
| "loss": 0.3473, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.2385093167701864, | |
| "grad_norm": 0.41939353269128443, | |
| "learning_rate": 3.557193840387263e-05, | |
| "loss": 0.3592, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.2459627329192546, | |
| "grad_norm": 0.4076888356556453, | |
| "learning_rate": 3.539848889273175e-05, | |
| "loss": 0.3503, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.253416149068323, | |
| "grad_norm": 0.4270944790896845, | |
| "learning_rate": 3.522450236049668e-05, | |
| "loss": 0.3541, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.2608695652173914, | |
| "grad_norm": 0.463757861659065, | |
| "learning_rate": 3.504999063657418e-05, | |
| "loss": 0.3542, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.2683229813664596, | |
| "grad_norm": 0.3981674010803556, | |
| "learning_rate": 3.487496558607898e-05, | |
| "loss": 0.3487, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.275776397515528, | |
| "grad_norm": 0.41590787708330634, | |
| "learning_rate": 3.4699439109027074e-05, | |
| "loss": 0.3498, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.2832298136645963, | |
| "grad_norm": 0.44072247076963306, | |
| "learning_rate": 3.452342313952662e-05, | |
| "loss": 0.3525, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.2906832298136646, | |
| "grad_norm": 0.38745116662865536, | |
| "learning_rate": 3.4346929644966564e-05, | |
| "loss": 0.3546, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.298136645962733, | |
| "grad_norm": 0.40585109654853857, | |
| "learning_rate": 3.416997062520292e-05, | |
| "loss": 0.3508, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.3055900621118013, | |
| "grad_norm": 0.4410011484738261, | |
| "learning_rate": 3.399255811174295e-05, | |
| "loss": 0.351, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.3130434782608695, | |
| "grad_norm": 0.4196810442218097, | |
| "learning_rate": 3.38147041669271e-05, | |
| "loss": 0.3678, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.320496894409938, | |
| "grad_norm": 0.3579720168253587, | |
| "learning_rate": 3.36364208831089e-05, | |
| "loss": 0.3585, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.3279503105590063, | |
| "grad_norm": 0.428787604095252, | |
| "learning_rate": 3.345772038183281e-05, | |
| "loss": 0.3481, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.3354037267080745, | |
| "grad_norm": 0.3731847145526466, | |
| "learning_rate": 3.3278614813010034e-05, | |
| "loss": 0.3529, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.342857142857143, | |
| "grad_norm": 0.4426288735756756, | |
| "learning_rate": 3.309911635409246e-05, | |
| "loss": 0.3661, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.3503105590062112, | |
| "grad_norm": 0.3937030484260621, | |
| "learning_rate": 3.291923720924473e-05, | |
| "loss": 0.3501, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.3577639751552795, | |
| "grad_norm": 0.40353386330372387, | |
| "learning_rate": 3.273898960851443e-05, | |
| "loss": 0.3537, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.365217391304348, | |
| "grad_norm": 0.3721891152141779, | |
| "learning_rate": 3.2558385807000654e-05, | |
| "loss": 0.3596, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.3726708074534162, | |
| "grad_norm": 0.36845786723670104, | |
| "learning_rate": 3.2377438084020644e-05, | |
| "loss": 0.3532, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.3801242236024844, | |
| "grad_norm": 0.43575242497786837, | |
| "learning_rate": 3.219615874227504e-05, | |
| "loss": 0.3581, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.387577639751553, | |
| "grad_norm": 0.3972116108444874, | |
| "learning_rate": 3.2014560107011324e-05, | |
| "loss": 0.3544, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.3950310559006212, | |
| "grad_norm": 0.3668661244872624, | |
| "learning_rate": 3.18326545251859e-05, | |
| "loss": 0.3551, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.4024844720496894, | |
| "grad_norm": 0.35145343781990773, | |
| "learning_rate": 3.165045436462451e-05, | |
| "loss": 0.3502, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.4099378881987579, | |
| "grad_norm": 0.3698602702933315, | |
| "learning_rate": 3.146797201318149e-05, | |
| "loss": 0.3473, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.4173913043478261, | |
| "grad_norm": 0.40228974580805166, | |
| "learning_rate": 3.128521987789738e-05, | |
| "loss": 0.36, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.4248447204968944, | |
| "grad_norm": 0.36657081099285116, | |
| "learning_rate": 3.110221038415545e-05, | |
| "loss": 0.3364, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.4322981366459628, | |
| "grad_norm": 0.4544371717792479, | |
| "learning_rate": 3.0918955974836846e-05, | |
| "loss": 0.3521, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.439751552795031, | |
| "grad_norm": 0.41609260719942065, | |
| "learning_rate": 3.073546910947461e-05, | |
| "loss": 0.3447, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.4472049689440993, | |
| "grad_norm": 0.42225853754054415, | |
| "learning_rate": 3.0551762263406576e-05, | |
| "loss": 0.3579, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.4546583850931678, | |
| "grad_norm": 0.4344358503994225, | |
| "learning_rate": 3.0367847926927134e-05, | |
| "loss": 0.3537, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.462111801242236, | |
| "grad_norm": 0.4160571014028236, | |
| "learning_rate": 3.0183738604438006e-05, | |
| "loss": 0.3529, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.4695652173913043, | |
| "grad_norm": 0.38345984224367013, | |
| "learning_rate": 2.999944681359811e-05, | |
| "loss": 0.3689, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.4770186335403728, | |
| "grad_norm": 0.40710896458852786, | |
| "learning_rate": 2.9814985084472418e-05, | |
| "loss": 0.3596, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.484472049689441, | |
| "grad_norm": 0.36671518555061516, | |
| "learning_rate": 2.9630365958680107e-05, | |
| "loss": 0.3622, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.4919254658385093, | |
| "grad_norm": 0.4044241295664962, | |
| "learning_rate": 2.9445601988541782e-05, | |
| "loss": 0.3423, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.4993788819875777, | |
| "grad_norm": 0.4084788856944912, | |
| "learning_rate": 2.9260705736226075e-05, | |
| "loss": 0.3523, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.506832298136646, | |
| "grad_norm": 0.37362508421237023, | |
| "learning_rate": 2.9075689772895538e-05, | |
| "loss": 0.342, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.5142857142857142, | |
| "grad_norm": 0.3760546446289586, | |
| "learning_rate": 2.88905666778519e-05, | |
| "loss": 0.3472, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.5217391304347827, | |
| "grad_norm": 0.34422522904634617, | |
| "learning_rate": 2.870534903768082e-05, | |
| "loss": 0.3435, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.529192546583851, | |
| "grad_norm": 0.36637523532571264, | |
| "learning_rate": 2.852004944539614e-05, | |
| "loss": 0.3576, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.5366459627329192, | |
| "grad_norm": 0.4326264584972339, | |
| "learning_rate": 2.8334680499583617e-05, | |
| "loss": 0.3509, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.5440993788819877, | |
| "grad_norm": 0.36854187428297824, | |
| "learning_rate": 2.814925480354441e-05, | |
| "loss": 0.3598, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.551552795031056, | |
| "grad_norm": 0.43487065325981833, | |
| "learning_rate": 2.7963784964438122e-05, | |
| "loss": 0.3453, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.5590062111801242, | |
| "grad_norm": 0.37541698341479846, | |
| "learning_rate": 2.777828359242567e-05, | |
| "loss": 0.3498, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.5664596273291926, | |
| "grad_norm": 0.4124753900399903, | |
| "learning_rate": 2.759276329981191e-05, | |
| "loss": 0.3532, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.5739130434782609, | |
| "grad_norm": 0.37293434521740576, | |
| "learning_rate": 2.74072367001881e-05, | |
| "loss": 0.3485, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.5813664596273291, | |
| "grad_norm": 0.4288037014823596, | |
| "learning_rate": 2.722171640757434e-05, | |
| "loss": 0.3532, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.5888198757763976, | |
| "grad_norm": 0.36312626249643093, | |
| "learning_rate": 2.703621503556189e-05, | |
| "loss": 0.3652, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.5962732919254659, | |
| "grad_norm": 0.47318947905865477, | |
| "learning_rate": 2.6850745196455594e-05, | |
| "loss": 0.3546, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.603726708074534, | |
| "grad_norm": 0.3525081596484065, | |
| "learning_rate": 2.6665319500416385e-05, | |
| "loss": 0.3503, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.6111801242236026, | |
| "grad_norm": 0.45691496651133917, | |
| "learning_rate": 2.6479950554603862e-05, | |
| "loss": 0.3483, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.6186335403726708, | |
| "grad_norm": 0.31443317345218114, | |
| "learning_rate": 2.6294650962319177e-05, | |
| "loss": 0.3639, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.626086956521739, | |
| "grad_norm": 0.3939981518889869, | |
| "learning_rate": 2.6109433322148112e-05, | |
| "loss": 0.3591, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.6335403726708075, | |
| "grad_norm": 0.34489071813729033, | |
| "learning_rate": 2.592431022710447e-05, | |
| "loss": 0.3524, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.6409937888198758, | |
| "grad_norm": 0.35520259178429653, | |
| "learning_rate": 2.5739294263773933e-05, | |
| "loss": 0.3516, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.648447204968944, | |
| "grad_norm": 0.3388748252085296, | |
| "learning_rate": 2.555439801145823e-05, | |
| "loss": 0.3586, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.6559006211180125, | |
| "grad_norm": 0.34569372832787504, | |
| "learning_rate": 2.5369634041319895e-05, | |
| "loss": 0.353, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.6633540372670808, | |
| "grad_norm": 0.3945538297720147, | |
| "learning_rate": 2.5185014915527587e-05, | |
| "loss": 0.3527, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.670807453416149, | |
| "grad_norm": 0.3762612151126058, | |
| "learning_rate": 2.50005531864019e-05, | |
| "loss": 0.3489, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.6782608695652175, | |
| "grad_norm": 0.44476348389377207, | |
| "learning_rate": 2.4816261395562003e-05, | |
| "loss": 0.3563, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.6857142857142857, | |
| "grad_norm": 0.3646157307605566, | |
| "learning_rate": 2.4632152073072878e-05, | |
| "loss": 0.3532, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.693167701863354, | |
| "grad_norm": 0.39063038331746, | |
| "learning_rate": 2.4448237736593422e-05, | |
| "loss": 0.3518, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.7006211180124224, | |
| "grad_norm": 0.36030888961041857, | |
| "learning_rate": 2.4264530890525395e-05, | |
| "loss": 0.3514, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.7080745341614907, | |
| "grad_norm": 0.34694110159996094, | |
| "learning_rate": 2.408104402516317e-05, | |
| "loss": 0.3537, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.715527950310559, | |
| "grad_norm": 0.35170013263514505, | |
| "learning_rate": 2.3897789615844557e-05, | |
| "loss": 0.3556, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.7229813664596274, | |
| "grad_norm": 0.339176023939733, | |
| "learning_rate": 2.3714780122102626e-05, | |
| "loss": 0.342, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.7304347826086957, | |
| "grad_norm": 0.35440772273510307, | |
| "learning_rate": 2.3532027986818518e-05, | |
| "loss": 0.3431, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.737888198757764, | |
| "grad_norm": 0.3624908891100353, | |
| "learning_rate": 2.3349545635375498e-05, | |
| "loss": 0.3549, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.7453416149068324, | |
| "grad_norm": 0.3198798531907377, | |
| "learning_rate": 2.3167345474814118e-05, | |
| "loss": 0.358, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.7527950310559006, | |
| "grad_norm": 0.33748379039276893, | |
| "learning_rate": 2.298543989298867e-05, | |
| "loss": 0.352, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.7602484472049689, | |
| "grad_norm": 0.35787480655458953, | |
| "learning_rate": 2.2803841257724962e-05, | |
| "loss": 0.3517, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.7677018633540373, | |
| "grad_norm": 0.3263237153649648, | |
| "learning_rate": 2.2622561915979358e-05, | |
| "loss": 0.3554, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.7751552795031056, | |
| "grad_norm": 0.381086077893375, | |
| "learning_rate": 2.2441614192999354e-05, | |
| "loss": 0.3614, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.7826086956521738, | |
| "grad_norm": 0.3615867584131201, | |
| "learning_rate": 2.226101039148557e-05, | |
| "loss": 0.3521, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.7900621118012423, | |
| "grad_norm": 0.33249983999528465, | |
| "learning_rate": 2.2080762790755282e-05, | |
| "loss": 0.35, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.7975155279503106, | |
| "grad_norm": 0.3466313876235434, | |
| "learning_rate": 2.1900883645907545e-05, | |
| "loss": 0.3478, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.8049689440993788, | |
| "grad_norm": 0.3545791130868414, | |
| "learning_rate": 2.1721385186989978e-05, | |
| "loss": 0.3383, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.8124223602484473, | |
| "grad_norm": 0.3433493456392416, | |
| "learning_rate": 2.1542279618167194e-05, | |
| "loss": 0.357, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.8198757763975155, | |
| "grad_norm": 0.35215043297702653, | |
| "learning_rate": 2.13635791168911e-05, | |
| "loss": 0.3584, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.8273291925465838, | |
| "grad_norm": 0.3306587366051955, | |
| "learning_rate": 2.1185295833072914e-05, | |
| "loss": 0.3536, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.8347826086956522, | |
| "grad_norm": 0.37995802267618, | |
| "learning_rate": 2.1007441888257055e-05, | |
| "loss": 0.3555, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.8422360248447205, | |
| "grad_norm": 0.3211766118761742, | |
| "learning_rate": 2.0830029374797085e-05, | |
| "loss": 0.3302, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.8496894409937887, | |
| "grad_norm": 0.34496506016293466, | |
| "learning_rate": 2.0653070355033438e-05, | |
| "loss": 0.3331, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.8571428571428572, | |
| "grad_norm": 0.35158994392589726, | |
| "learning_rate": 2.047657686047338e-05, | |
| "loss": 0.3522, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.8645962732919255, | |
| "grad_norm": 0.2997777930152674, | |
| "learning_rate": 2.030056089097293e-05, | |
| "loss": 0.345, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.8720496894409937, | |
| "grad_norm": 0.36933604277512955, | |
| "learning_rate": 2.0125034413921024e-05, | |
| "loss": 0.3553, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.8795031055900622, | |
| "grad_norm": 0.30278033326279763, | |
| "learning_rate": 1.9950009363425827e-05, | |
| "loss": 0.3504, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.8869565217391304, | |
| "grad_norm": 0.3317258855093456, | |
| "learning_rate": 1.9775497639503325e-05, | |
| "loss": 0.3637, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.8944099378881987, | |
| "grad_norm": 0.32228882411814175, | |
| "learning_rate": 1.9601511107268255e-05, | |
| "loss": 0.3553, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.9018633540372671, | |
| "grad_norm": 0.31378115419623065, | |
| "learning_rate": 1.9428061596127383e-05, | |
| "loss": 0.347, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.9093167701863354, | |
| "grad_norm": 0.32150232133903023, | |
| "learning_rate": 1.9255160898975195e-05, | |
| "loss": 0.3416, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.9167701863354036, | |
| "grad_norm": 0.30317071268510376, | |
| "learning_rate": 1.9082820771392157e-05, | |
| "loss": 0.3488, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.924223602484472, | |
| "grad_norm": 0.3160493816316982, | |
| "learning_rate": 1.891105293084538e-05, | |
| "loss": 0.3524, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.9316770186335404, | |
| "grad_norm": 0.31367068123245806, | |
| "learning_rate": 1.8739869055891972e-05, | |
| "loss": 0.3532, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.9391304347826086, | |
| "grad_norm": 0.32732364352375504, | |
| "learning_rate": 1.8569280785385046e-05, | |
| "loss": 0.3437, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.946583850931677, | |
| "grad_norm": 0.3044343110689431, | |
| "learning_rate": 1.8399299717682304e-05, | |
| "loss": 0.351, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.9540372670807453, | |
| "grad_norm": 0.37461713163454335, | |
| "learning_rate": 1.822993740985754e-05, | |
| "loss": 0.3491, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.9614906832298136, | |
| "grad_norm": 0.31668154420249334, | |
| "learning_rate": 1.806120537691485e-05, | |
| "loss": 0.3473, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.968944099378882, | |
| "grad_norm": 0.3012343742566849, | |
| "learning_rate": 1.7893115091005695e-05, | |
| "loss": 0.3473, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.9763975155279503, | |
| "grad_norm": 0.39737978489918535, | |
| "learning_rate": 1.772567798064894e-05, | |
| "loss": 0.3494, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.9838509316770185, | |
| "grad_norm": 0.33805655625614117, | |
| "learning_rate": 1.7558905429953805e-05, | |
| "loss": 0.3475, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.991304347826087, | |
| "grad_norm": 0.3322232086747, | |
| "learning_rate": 1.739280877784584e-05, | |
| "loss": 0.3308, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.9987577639751553, | |
| "grad_norm": 0.38072556291830295, | |
| "learning_rate": 1.7227399317296043e-05, | |
| "loss": 0.3425, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.38072556291830295, | |
| "learning_rate": 1.7062688294552992e-05, | |
| "loss": 0.3282, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.0074534161490685, | |
| "grad_norm": 0.7460455385201656, | |
| "learning_rate": 1.6898686908378227e-05, | |
| "loss": 0.2385, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.0149068322981365, | |
| "grad_norm": 0.4584394574079776, | |
| "learning_rate": 1.6735406309284847e-05, | |
| "loss": 0.2325, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.022360248447205, | |
| "grad_norm": 0.35402810903371174, | |
| "learning_rate": 1.657285759877937e-05, | |
| "loss": 0.2289, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.0298136645962734, | |
| "grad_norm": 0.41508856204183064, | |
| "learning_rate": 1.6411051828606964e-05, | |
| "loss": 0.2289, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.0372670807453415, | |
| "grad_norm": 0.5703640626631827, | |
| "learning_rate": 1.6250000000000005e-05, | |
| "loss": 0.251, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.04472049689441, | |
| "grad_norm": 0.44030030337444215, | |
| "learning_rate": 1.6089713062930108e-05, | |
| "loss": 0.2393, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.0521739130434784, | |
| "grad_norm": 0.3634716004644788, | |
| "learning_rate": 1.5930201915363652e-05, | |
| "loss": 0.2208, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.0596273291925464, | |
| "grad_norm": 0.40263713578890487, | |
| "learning_rate": 1.5771477402520797e-05, | |
| "loss": 0.2251, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.067080745341615, | |
| "grad_norm": 0.38573706299136595, | |
| "learning_rate": 1.5613550316138116e-05, | |
| "loss": 0.2373, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.0745341614906834, | |
| "grad_norm": 0.3886650442753063, | |
| "learning_rate": 1.5456431393734912e-05, | |
| "loss": 0.2328, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.0819875776397514, | |
| "grad_norm": 0.3353566159721077, | |
| "learning_rate": 1.530013131788308e-05, | |
| "loss": 0.2368, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.08944099378882, | |
| "grad_norm": 0.32235589864048775, | |
| "learning_rate": 1.5144660715480877e-05, | |
| "loss": 0.2372, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.0968944099378883, | |
| "grad_norm": 0.3449572995224796, | |
| "learning_rate": 1.4990030157030343e-05, | |
| "loss": 0.2302, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.1043478260869564, | |
| "grad_norm": 0.35370627629200513, | |
| "learning_rate": 1.4836250155918632e-05, | |
| "loss": 0.2254, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.111801242236025, | |
| "grad_norm": 0.32255576283733056, | |
| "learning_rate": 1.4683331167703218e-05, | |
| "loss": 0.2304, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.1192546583850933, | |
| "grad_norm": 0.33160244908908576, | |
| "learning_rate": 1.453128358940098e-05, | |
| "loss": 0.229, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.1267080745341613, | |
| "grad_norm": 0.3342512523189852, | |
| "learning_rate": 1.4380117758781318e-05, | |
| "loss": 0.2232, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.13416149068323, | |
| "grad_norm": 0.3154712145125721, | |
| "learning_rate": 1.4229843953663313e-05, | |
| "loss": 0.2295, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.1416149068322983, | |
| "grad_norm": 0.2998522567104169, | |
| "learning_rate": 1.4080472391216898e-05, | |
| "loss": 0.2318, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 2.1490683229813663, | |
| "grad_norm": 0.3120711414761453, | |
| "learning_rate": 1.3932013227268208e-05, | |
| "loss": 0.2199, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.1565217391304348, | |
| "grad_norm": 0.31545849948494736, | |
| "learning_rate": 1.3784476555609077e-05, | |
| "loss": 0.2258, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.1639751552795032, | |
| "grad_norm": 0.31288349011845973, | |
| "learning_rate": 1.3637872407310748e-05, | |
| "loss": 0.2372, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.1714285714285713, | |
| "grad_norm": 0.30783876364368057, | |
| "learning_rate": 1.3492210750041873e-05, | |
| "loss": 0.2273, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.1788819875776397, | |
| "grad_norm": 0.29777744790391, | |
| "learning_rate": 1.3347501487390801e-05, | |
| "loss": 0.2298, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.186335403726708, | |
| "grad_norm": 0.30475333543232624, | |
| "learning_rate": 1.3203754458192225e-05, | |
| "loss": 0.234, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.1937888198757762, | |
| "grad_norm": 0.29158802455098537, | |
| "learning_rate": 1.3060979435858259e-05, | |
| "loss": 0.2177, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.2012422360248447, | |
| "grad_norm": 0.2910747993340585, | |
| "learning_rate": 1.2919186127713885e-05, | |
| "loss": 0.2257, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.208695652173913, | |
| "grad_norm": 0.29580086796673544, | |
| "learning_rate": 1.2778384174337025e-05, | |
| "loss": 0.2256, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.216149068322981, | |
| "grad_norm": 0.28921712308433817, | |
| "learning_rate": 1.2638583148903008e-05, | |
| "loss": 0.2321, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 2.2236024844720497, | |
| "grad_norm": 0.28971064806152724, | |
| "learning_rate": 1.2499792556533716e-05, | |
| "loss": 0.2319, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.231055900621118, | |
| "grad_norm": 0.2855654258873247, | |
| "learning_rate": 1.236202183365134e-05, | |
| "loss": 0.2255, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 2.238509316770186, | |
| "grad_norm": 0.291778227302567, | |
| "learning_rate": 1.2225280347336763e-05, | |
| "loss": 0.2329, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 2.2459627329192546, | |
| "grad_norm": 0.28470732682114813, | |
| "learning_rate": 1.20895773946927e-05, | |
| "loss": 0.2263, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 2.253416149068323, | |
| "grad_norm": 0.27214743085323334, | |
| "learning_rate": 1.1954922202211615e-05, | |
| "loss": 0.2262, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 2.260869565217391, | |
| "grad_norm": 0.2783027932885126, | |
| "learning_rate": 1.1821323925148358e-05, | |
| "loss": 0.2278, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.2683229813664596, | |
| "grad_norm": 0.2940915504230925, | |
| "learning_rate": 1.1688791646897726e-05, | |
| "loss": 0.2254, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 2.275776397515528, | |
| "grad_norm": 0.29241392964362756, | |
| "learning_rate": 1.1557334378376882e-05, | |
| "loss": 0.2274, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 2.283229813664596, | |
| "grad_norm": 0.2859155808493438, | |
| "learning_rate": 1.1426961057412672e-05, | |
| "loss": 0.2242, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.2906832298136646, | |
| "grad_norm": 0.2863824212664396, | |
| "learning_rate": 1.1297680548133993e-05, | |
| "loss": 0.2211, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.298136645962733, | |
| "grad_norm": 0.27218685640719575, | |
| "learning_rate": 1.1169501640369051e-05, | |
| "loss": 0.2166, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.305590062111801, | |
| "grad_norm": 0.2739565076313759, | |
| "learning_rate": 1.1042433049047781e-05, | |
| "loss": 0.2288, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.3130434782608695, | |
| "grad_norm": 0.2858456970822226, | |
| "learning_rate": 1.0916483413609315e-05, | |
| "loss": 0.2252, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.320496894409938, | |
| "grad_norm": 0.2730982695340446, | |
| "learning_rate": 1.079166129741455e-05, | |
| "loss": 0.2289, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 2.327950310559006, | |
| "grad_norm": 0.2858131409653359, | |
| "learning_rate": 1.0667975187163976e-05, | |
| "loss": 0.2308, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 2.3354037267080745, | |
| "grad_norm": 0.2666152133579397, | |
| "learning_rate": 1.0545433492320603e-05, | |
| "loss": 0.2278, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 2.342857142857143, | |
| "grad_norm": 0.2770908273244185, | |
| "learning_rate": 1.042404454453824e-05, | |
| "loss": 0.2168, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 2.350310559006211, | |
| "grad_norm": 0.2738838241569534, | |
| "learning_rate": 1.0303816597095004e-05, | |
| "loss": 0.2381, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 2.3577639751552795, | |
| "grad_norm": 0.27892490339684844, | |
| "learning_rate": 1.0184757824332187e-05, | |
| "loss": 0.2365, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 2.365217391304348, | |
| "grad_norm": 0.27327486293785713, | |
| "learning_rate": 1.0066876321098467e-05, | |
| "loss": 0.2262, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 2.372670807453416, | |
| "grad_norm": 0.2691733028937572, | |
| "learning_rate": 9.950180102199564e-06, | |
| "loss": 0.2287, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.3801242236024844, | |
| "grad_norm": 0.2706087500883035, | |
| "learning_rate": 9.834677101853265e-06, | |
| "loss": 0.2244, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 2.387577639751553, | |
| "grad_norm": 0.2855306323449237, | |
| "learning_rate": 9.720375173150024e-06, | |
| "loss": 0.2349, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 2.395031055900621, | |
| "grad_norm": 0.26777926603611185, | |
| "learning_rate": 9.607282087518984e-06, | |
| "loss": 0.2154, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 2.4024844720496894, | |
| "grad_norm": 0.2821902505187217, | |
| "learning_rate": 9.495405534199617e-06, | |
| "loss": 0.2325, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 2.409937888198758, | |
| "grad_norm": 0.27147857463244984, | |
| "learning_rate": 9.384753119718937e-06, | |
| "loss": 0.2228, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.417391304347826, | |
| "grad_norm": 0.2699371891385757, | |
| "learning_rate": 9.27533236737431e-06, | |
| "loss": 0.2264, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 2.4248447204968944, | |
| "grad_norm": 0.2762402133611527, | |
| "learning_rate": 9.167150716721954e-06, | |
| "loss": 0.2208, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 2.432298136645963, | |
| "grad_norm": 0.27110209952512, | |
| "learning_rate": 9.060215523071117e-06, | |
| "loss": 0.2253, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 2.439751552795031, | |
| "grad_norm": 0.27014197741120893, | |
| "learning_rate": 8.954534056983984e-06, | |
| "loss": 0.2293, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 2.4472049689440993, | |
| "grad_norm": 0.27495397443241254, | |
| "learning_rate": 8.850113503781367e-06, | |
| "loss": 0.2209, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.454658385093168, | |
| "grad_norm": 0.27468805910530714, | |
| "learning_rate": 8.746960963054145e-06, | |
| "loss": 0.2349, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 2.462111801242236, | |
| "grad_norm": 0.2770381414499631, | |
| "learning_rate": 8.645083448180574e-06, | |
| "loss": 0.2238, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 2.4695652173913043, | |
| "grad_norm": 0.27376387230241217, | |
| "learning_rate": 8.54448788584946e-06, | |
| "loss": 0.2207, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 2.4770186335403728, | |
| "grad_norm": 0.27580353431299076, | |
| "learning_rate": 8.445181115589179e-06, | |
| "loss": 0.2219, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 2.4844720496894412, | |
| "grad_norm": 0.26703692348696817, | |
| "learning_rate": 8.34716988930267e-06, | |
| "loss": 0.2187, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 2.4919254658385093, | |
| "grad_norm": 0.2651902683216356, | |
| "learning_rate": 8.250460870808394e-06, | |
| "loss": 0.2238, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 2.4993788819875777, | |
| "grad_norm": 0.2602841629540144, | |
| "learning_rate": 8.155060635387206e-06, | |
| "loss": 0.2237, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 2.506832298136646, | |
| "grad_norm": 0.26591023651045076, | |
| "learning_rate": 8.060975669335365e-06, | |
| "loss": 0.2283, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 2.5142857142857142, | |
| "grad_norm": 0.26252263621744065, | |
| "learning_rate": 7.968212369523462e-06, | |
| "loss": 0.2245, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 2.5217391304347827, | |
| "grad_norm": 0.2748374618397896, | |
| "learning_rate": 7.876777042961544e-06, | |
| "loss": 0.2228, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.529192546583851, | |
| "grad_norm": 0.2597274246070528, | |
| "learning_rate": 7.786675906370278e-06, | |
| "loss": 0.2283, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 2.536645962732919, | |
| "grad_norm": 0.2687998792747167, | |
| "learning_rate": 7.697915085758266e-06, | |
| "loss": 0.2205, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 2.5440993788819877, | |
| "grad_norm": 0.26277761372100056, | |
| "learning_rate": 7.610500616005556e-06, | |
| "loss": 0.2201, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 2.551552795031056, | |
| "grad_norm": 0.273978880458127, | |
| "learning_rate": 7.524438440453323e-06, | |
| "loss": 0.2332, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 2.559006211180124, | |
| "grad_norm": 0.25462514116789337, | |
| "learning_rate": 7.439734410499752e-06, | |
| "loss": 0.2188, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 2.5664596273291926, | |
| "grad_norm": 0.271284156037868, | |
| "learning_rate": 7.356394285202248e-06, | |
| "loss": 0.2252, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 2.573913043478261, | |
| "grad_norm": 0.2709719202368051, | |
| "learning_rate": 7.274423730885835e-06, | |
| "loss": 0.2377, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 2.581366459627329, | |
| "grad_norm": 0.2687629580156147, | |
| "learning_rate": 7.193828320757909e-06, | |
| "loss": 0.2212, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 2.5888198757763976, | |
| "grad_norm": 0.2723242179612203, | |
| "learning_rate": 7.114613534529333e-06, | |
| "loss": 0.2169, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 2.596273291925466, | |
| "grad_norm": 0.25830008261692217, | |
| "learning_rate": 7.036784758041846e-06, | |
| "loss": 0.2151, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.603726708074534, | |
| "grad_norm": 0.27058413732950065, | |
| "learning_rate": 6.960347282901894e-06, | |
| "loss": 0.226, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 2.6111801242236026, | |
| "grad_norm": 0.2602830736394775, | |
| "learning_rate": 6.885306306120837e-06, | |
| "loss": 0.2173, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 2.618633540372671, | |
| "grad_norm": 0.2884065006859397, | |
| "learning_rate": 6.811666929761612e-06, | |
| "loss": 0.223, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 2.626086956521739, | |
| "grad_norm": 0.2678949940900833, | |
| "learning_rate": 6.739434160591852e-06, | |
| "loss": 0.2232, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 2.6335403726708075, | |
| "grad_norm": 0.26613900004744706, | |
| "learning_rate": 6.668612909743448e-06, | |
| "loss": 0.2109, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 2.640993788819876, | |
| "grad_norm": 0.2600134334991607, | |
| "learning_rate": 6.599207992378657e-06, | |
| "loss": 0.2216, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 2.648447204968944, | |
| "grad_norm": 0.26265543365401706, | |
| "learning_rate": 6.531224127362726e-06, | |
| "loss": 0.223, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 2.6559006211180125, | |
| "grad_norm": 0.26558869901544513, | |
| "learning_rate": 6.464665936943023e-06, | |
| "loss": 0.2172, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 2.663354037267081, | |
| "grad_norm": 0.2693433022214806, | |
| "learning_rate": 6.399537946434801e-06, | |
| "loss": 0.2198, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 2.670807453416149, | |
| "grad_norm": 0.2604150700405178, | |
| "learning_rate": 6.335844583913515e-06, | |
| "loss": 0.22, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.6782608695652175, | |
| "grad_norm": 0.26795928209419323, | |
| "learning_rate": 6.27359017991373e-06, | |
| "loss": 0.2254, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 2.685714285714286, | |
| "grad_norm": 0.26825259134506735, | |
| "learning_rate": 6.212778967134715e-06, | |
| "loss": 0.2286, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 2.693167701863354, | |
| "grad_norm": 0.2624757714546698, | |
| "learning_rate": 6.153415080152655e-06, | |
| "loss": 0.2224, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 2.7006211180124224, | |
| "grad_norm": 0.2538413882658132, | |
| "learning_rate": 6.095502555139516e-06, | |
| "loss": 0.2116, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 2.708074534161491, | |
| "grad_norm": 0.26097610705351, | |
| "learning_rate": 6.039045329588671e-06, | |
| "loss": 0.2135, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 2.715527950310559, | |
| "grad_norm": 0.26841385627861275, | |
| "learning_rate": 5.984047242047134e-06, | |
| "loss": 0.2126, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 2.7229813664596274, | |
| "grad_norm": 0.26691120572518234, | |
| "learning_rate": 5.930512031854617e-06, | |
| "loss": 0.2231, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 2.730434782608696, | |
| "grad_norm": 0.2595939680410351, | |
| "learning_rate": 5.8784433388892726e-06, | |
| "loss": 0.2187, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 2.737888198757764, | |
| "grad_norm": 0.2620978338502642, | |
| "learning_rate": 5.827844703320216e-06, | |
| "loss": 0.2275, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 2.7453416149068324, | |
| "grad_norm": 0.27293115847294985, | |
| "learning_rate": 5.778719565366846e-06, | |
| "loss": 0.2216, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 2.752795031055901, | |
| "grad_norm": 0.26642579246201126, | |
| "learning_rate": 5.731071265064913e-06, | |
| "loss": 0.2253, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 2.760248447204969, | |
| "grad_norm": 0.2673788170594958, | |
| "learning_rate": 5.684903042039452e-06, | |
| "loss": 0.2271, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 2.7677018633540373, | |
| "grad_norm": 0.2554915331836976, | |
| "learning_rate": 5.640218035284521e-06, | |
| "loss": 0.2295, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 2.775155279503106, | |
| "grad_norm": 0.2678452830442339, | |
| "learning_rate": 5.59701928294976e-06, | |
| "loss": 0.2215, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "grad_norm": 0.2815341168477551, | |
| "learning_rate": 5.555309722133842e-06, | |
| "loss": 0.2281, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 2.7900621118012423, | |
| "grad_norm": 0.2624788732767209, | |
| "learning_rate": 5.515092188684775e-06, | |
| "loss": 0.2247, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 2.7975155279503108, | |
| "grad_norm": 0.27489059149241923, | |
| "learning_rate": 5.476369417007091e-06, | |
| "loss": 0.2327, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 2.804968944099379, | |
| "grad_norm": 0.28772902740092976, | |
| "learning_rate": 5.439144039875931e-06, | |
| "loss": 0.2167, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 2.8124223602484473, | |
| "grad_norm": 0.26111982505416115, | |
| "learning_rate": 5.403418588258045e-06, | |
| "loss": 0.2197, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 2.8198757763975157, | |
| "grad_norm": 0.267283694967123, | |
| "learning_rate": 5.369195491139709e-06, | |
| "loss": 0.2262, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 2.8273291925465838, | |
| "grad_norm": 0.27325385624236986, | |
| "learning_rate": 5.336477075361577e-06, | |
| "loss": 0.2172, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 2.8347826086956522, | |
| "grad_norm": 0.2647115488846027, | |
| "learning_rate": 5.305265565460477e-06, | |
| "loss": 0.2244, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 2.8422360248447207, | |
| "grad_norm": 0.26432835544386557, | |
| "learning_rate": 5.275563083518169e-06, | |
| "loss": 0.216, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 2.8496894409937887, | |
| "grad_norm": 0.2574806933503353, | |
| "learning_rate": 5.247371649017059e-06, | |
| "loss": 0.2309, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 0.26132718645601316, | |
| "learning_rate": 5.220693178702895e-06, | |
| "loss": 0.2142, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 2.8645962732919257, | |
| "grad_norm": 0.2778266622426319, | |
| "learning_rate": 5.195529486454448e-06, | |
| "loss": 0.2246, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 2.8720496894409937, | |
| "grad_norm": 0.2548174424842807, | |
| "learning_rate": 5.171882283160185e-06, | |
| "loss": 0.2207, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 2.879503105590062, | |
| "grad_norm": 0.26469737507644614, | |
| "learning_rate": 5.149753176601936e-06, | |
| "loss": 0.2245, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 2.8869565217391306, | |
| "grad_norm": 0.2660467641695125, | |
| "learning_rate": 5.129143671345609e-06, | |
| "loss": 0.227, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 2.8944099378881987, | |
| "grad_norm": 0.253693975774761, | |
| "learning_rate": 5.110055168638854e-06, | |
| "loss": 0.219, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 2.901863354037267, | |
| "grad_norm": 0.2795604845969123, | |
| "learning_rate": 5.09248896631582e-06, | |
| "loss": 0.2321, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 2.9093167701863356, | |
| "grad_norm": 0.7496235586126184, | |
| "learning_rate": 5.07644625870891e-06, | |
| "loss": 0.2366, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 2.9167701863354036, | |
| "grad_norm": 0.2618794070766183, | |
| "learning_rate": 5.06192813656757e-06, | |
| "loss": 0.2283, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 2.924223602484472, | |
| "grad_norm": 0.2616116501289447, | |
| "learning_rate": 5.048935586984133e-06, | |
| "loss": 0.2177, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 2.9316770186335406, | |
| "grad_norm": 0.2719537845754615, | |
| "learning_rate": 5.0374694933267114e-06, | |
| "loss": 0.2273, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 2.9391304347826086, | |
| "grad_norm": 0.26226288303787226, | |
| "learning_rate": 5.027530635179121e-06, | |
| "loss": 0.2132, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 2.946583850931677, | |
| "grad_norm": 0.24929011431595574, | |
| "learning_rate": 5.019119688287901e-06, | |
| "loss": 0.2238, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 2.9540372670807455, | |
| "grad_norm": 0.2606373590411471, | |
| "learning_rate": 5.012237224516342e-06, | |
| "loss": 0.2216, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 2.9614906832298136, | |
| "grad_norm": 0.29892861469452736, | |
| "learning_rate": 5.00688371180563e-06, | |
| "loss": 0.2362, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 2.968944099378882, | |
| "grad_norm": 0.2665719018831708, | |
| "learning_rate": 5.003059514143014e-06, | |
| "loss": 0.221, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.9763975155279505, | |
| "grad_norm": 0.25860725999612966, | |
| "learning_rate": 5.000764891537067e-06, | |
| "loss": 0.2204, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 2.9838509316770185, | |
| "grad_norm": 0.25161372246359515, | |
| "learning_rate": 5e-06, | |
| "loss": 0.2245, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 2.9838509316770185, | |
| "step": 402, | |
| "total_flos": 8.574354709534474e+17, | |
| "train_loss": 0.3579157315083404, | |
| "train_runtime": 11555.4492, | |
| "train_samples_per_second": 6.688, | |
| "train_steps_per_second": 0.035 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 402, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 8.574354709534474e+17, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |