| { | |
| "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": 6.385986447654567, | |
| "learning_rate": 8.695652173913044e-07, | |
| "loss": 1.0298, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.060759493670886074, | |
| "grad_norm": 6.560514136224517, | |
| "learning_rate": 1.7391304347826088e-06, | |
| "loss": 1.0557, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.09113924050632911, | |
| "grad_norm": 6.397314014677567, | |
| "learning_rate": 2.6086956521739132e-06, | |
| "loss": 1.0417, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.12151898734177215, | |
| "grad_norm": 5.996984377828155, | |
| "learning_rate": 3.4782608695652175e-06, | |
| "loss": 1.0524, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.1518987341772152, | |
| "grad_norm": 4.56512095391756, | |
| "learning_rate": 4.347826086956522e-06, | |
| "loss": 0.9949, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.18227848101265823, | |
| "grad_norm": 2.7350945902058412, | |
| "learning_rate": 5.2173913043478265e-06, | |
| "loss": 0.9499, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.21265822784810126, | |
| "grad_norm": 2.525415425377413, | |
| "learning_rate": 6.086956521739132e-06, | |
| "loss": 0.9546, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.2430379746835443, | |
| "grad_norm": 3.9707412715781483, | |
| "learning_rate": 6.956521739130435e-06, | |
| "loss": 0.9243, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.27341772151898736, | |
| "grad_norm": 4.230606555198297, | |
| "learning_rate": 7.82608695652174e-06, | |
| "loss": 0.968, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.3037974683544304, | |
| "grad_norm": 3.7259600436834495, | |
| "learning_rate": 8.695652173913044e-06, | |
| "loss": 0.9321, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3341772151898734, | |
| "grad_norm": 3.5814920721103864, | |
| "learning_rate": 9.565217391304349e-06, | |
| "loss": 0.9101, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.36455696202531646, | |
| "grad_norm": 2.936858744919314, | |
| "learning_rate": 1.0434782608695653e-05, | |
| "loss": 0.8518, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.3949367088607595, | |
| "grad_norm": 2.187994658787228, | |
| "learning_rate": 1.1304347826086957e-05, | |
| "loss": 0.8832, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.4253164556962025, | |
| "grad_norm": 1.4912592034082488, | |
| "learning_rate": 1.2173913043478263e-05, | |
| "loss": 0.8515, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.45569620253164556, | |
| "grad_norm": 1.5325213632015104, | |
| "learning_rate": 1.3043478260869566e-05, | |
| "loss": 0.8264, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.4860759493670886, | |
| "grad_norm": 1.3349730535552873, | |
| "learning_rate": 1.391304347826087e-05, | |
| "loss": 0.8349, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.5164556962025316, | |
| "grad_norm": 1.0822853683485523, | |
| "learning_rate": 1.4782608695652174e-05, | |
| "loss": 0.7923, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.5468354430379747, | |
| "grad_norm": 1.1408012520798394, | |
| "learning_rate": 1.565217391304348e-05, | |
| "loss": 0.7978, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.5772151898734177, | |
| "grad_norm": 0.9572581185901935, | |
| "learning_rate": 1.6521739130434785e-05, | |
| "loss": 0.7885, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.6075949367088608, | |
| "grad_norm": 0.9735174496656804, | |
| "learning_rate": 1.739130434782609e-05, | |
| "loss": 0.7607, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.6379746835443038, | |
| "grad_norm": 0.9644255893601036, | |
| "learning_rate": 1.8260869565217393e-05, | |
| "loss": 0.7635, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.6683544303797468, | |
| "grad_norm": 0.906068396318747, | |
| "learning_rate": 1.9130434782608697e-05, | |
| "loss": 0.7405, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.6987341772151898, | |
| "grad_norm": 0.7672254803056492, | |
| "learning_rate": 2e-05, | |
| "loss": 0.752, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.7291139240506329, | |
| "grad_norm": 0.9193718585902947, | |
| "learning_rate": 1.999877856940653e-05, | |
| "loss": 0.7552, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.759493670886076, | |
| "grad_norm": 0.814783279935625, | |
| "learning_rate": 1.999511457600466e-05, | |
| "loss": 0.7324, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.789873417721519, | |
| "grad_norm": 0.8493121941492644, | |
| "learning_rate": 1.9989008914857115e-05, | |
| "loss": 0.7267, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.8202531645569621, | |
| "grad_norm": 0.7345291387866795, | |
| "learning_rate": 1.998046307749216e-05, | |
| "loss": 0.7243, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.850632911392405, | |
| "grad_norm": 0.760270238715357, | |
| "learning_rate": 1.9969479151539238e-05, | |
| "loss": 0.7354, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.8810126582278481, | |
| "grad_norm": 0.6681927581945148, | |
| "learning_rate": 1.9956059820218982e-05, | |
| "loss": 0.7318, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.9113924050632911, | |
| "grad_norm": 0.873320522945356, | |
| "learning_rate": 1.9940208361687762e-05, | |
| "loss": 0.7298, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.9417721518987342, | |
| "grad_norm": 0.6705461225314656, | |
| "learning_rate": 1.9921928648236855e-05, | |
| "loss": 0.7307, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.9721518987341772, | |
| "grad_norm": 0.7721800578911916, | |
| "learning_rate": 1.990122514534651e-05, | |
| "loss": 0.6923, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.010126582278481, | |
| "grad_norm": 0.6891251395321004, | |
| "learning_rate": 1.9878102910595097e-05, | |
| "loss": 0.7127, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.040506329113924, | |
| "grad_norm": 0.7404774579783256, | |
| "learning_rate": 1.985256759242359e-05, | |
| "loss": 0.6873, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.070886075949367, | |
| "grad_norm": 0.586272696667338, | |
| "learning_rate": 1.982462542875576e-05, | |
| "loss": 0.6717, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.1012658227848102, | |
| "grad_norm": 0.8052811987450509, | |
| "learning_rate": 1.979428324547432e-05, | |
| "loss": 0.6758, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.1316455696202532, | |
| "grad_norm": 0.7543672982841814, | |
| "learning_rate": 1.9761548454753455e-05, | |
| "loss": 0.6753, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.1620253164556962, | |
| "grad_norm": 0.6515297334515208, | |
| "learning_rate": 1.972642905324813e-05, | |
| "loss": 0.6708, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.1924050632911392, | |
| "grad_norm": 0.6322402783261385, | |
| "learning_rate": 1.9688933620140638e-05, | |
| "loss": 0.6812, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.2227848101265824, | |
| "grad_norm": 0.6564153902599098, | |
| "learning_rate": 1.96490713150448e-05, | |
| "loss": 0.6687, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.2531645569620253, | |
| "grad_norm": 0.6276143707710902, | |
| "learning_rate": 1.9606851875768404e-05, | |
| "loss": 0.6727, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.2835443037974683, | |
| "grad_norm": 0.553446544120281, | |
| "learning_rate": 1.956228561593441e-05, | |
| "loss": 0.667, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.3139240506329113, | |
| "grad_norm": 0.6447931327666168, | |
| "learning_rate": 1.9515383422461457e-05, | |
| "loss": 0.6691, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.3443037974683545, | |
| "grad_norm": 0.5542900121228235, | |
| "learning_rate": 1.9466156752904344e-05, | |
| "loss": 0.6638, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.3746835443037975, | |
| "grad_norm": 0.5032361502402146, | |
| "learning_rate": 1.9414617632655114e-05, | |
| "loss": 0.6594, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.4050632911392404, | |
| "grad_norm": 0.538039736473092, | |
| "learning_rate": 1.9360778652005416e-05, | |
| "loss": 0.6508, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.4354430379746836, | |
| "grad_norm": 0.4951148606993244, | |
| "learning_rate": 1.9304652963070868e-05, | |
| "loss": 0.654, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.4658227848101266, | |
| "grad_norm": 0.589760329539897, | |
| "learning_rate": 1.9246254276578175e-05, | |
| "loss": 0.6548, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.4962025316455696, | |
| "grad_norm": 0.5243656891175175, | |
| "learning_rate": 1.9185596858515797e-05, | |
| "loss": 0.647, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.5265822784810128, | |
| "grad_norm": 0.5443765124602752, | |
| "learning_rate": 1.9122695526648968e-05, | |
| "loss": 0.655, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.5569620253164556, | |
| "grad_norm": 0.45690559798624136, | |
| "learning_rate": 1.905756564689991e-05, | |
| "loss": 0.6598, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.5873417721518988, | |
| "grad_norm": 0.5209287671122564, | |
| "learning_rate": 1.8990223129594146e-05, | |
| "loss": 0.6321, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.6177215189873417, | |
| "grad_norm": 0.47599825235265614, | |
| "learning_rate": 1.8920684425573865e-05, | |
| "loss": 0.6339, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.6481012658227847, | |
| "grad_norm": 0.5338766044704407, | |
| "learning_rate": 1.884896652217917e-05, | |
| "loss": 0.6452, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.678481012658228, | |
| "grad_norm": 0.45246705857603065, | |
| "learning_rate": 1.877508693909831e-05, | |
| "loss": 0.6345, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.7088607594936709, | |
| "grad_norm": 0.5283369634176305, | |
| "learning_rate": 1.8699063724087905e-05, | |
| "loss": 0.6502, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.7392405063291139, | |
| "grad_norm": 0.5488855982163836, | |
| "learning_rate": 1.862091544856407e-05, | |
| "loss": 0.6368, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.769620253164557, | |
| "grad_norm": 0.48456179249124, | |
| "learning_rate": 1.854066120306571e-05, | |
| "loss": 0.6375, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.5804476831386609, | |
| "learning_rate": 1.8458320592590976e-05, | |
| "loss": 0.6403, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.830379746835443, | |
| "grad_norm": 0.4799042491510617, | |
| "learning_rate": 1.837391373180801e-05, | |
| "loss": 0.6289, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.8607594936708862, | |
| "grad_norm": 0.5294745099492968, | |
| "learning_rate": 1.8287461240141217e-05, | |
| "loss": 0.6469, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.891139240506329, | |
| "grad_norm": 0.4126498401667685, | |
| "learning_rate": 1.8198984236734246e-05, | |
| "loss": 0.6313, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.9215189873417722, | |
| "grad_norm": 0.4872755757296018, | |
| "learning_rate": 1.8108504335290852e-05, | |
| "loss": 0.6384, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.9518987341772152, | |
| "grad_norm": 0.4589047618723706, | |
| "learning_rate": 1.8016043638794975e-05, | |
| "loss": 0.6386, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.9822784810126581, | |
| "grad_norm": 0.5517643851409115, | |
| "learning_rate": 1.7921624734111292e-05, | |
| "loss": 0.6432, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.020253164556962, | |
| "grad_norm": 0.5525594613838342, | |
| "learning_rate": 1.7825270686467567e-05, | |
| "loss": 0.6084, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.050632911392405, | |
| "grad_norm": 0.6073805475810474, | |
| "learning_rate": 1.7727005033820117e-05, | |
| "loss": 0.599, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.081012658227848, | |
| "grad_norm": 0.524623046354894, | |
| "learning_rate": 1.762685178110382e-05, | |
| "loss": 0.5678, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.1113924050632913, | |
| "grad_norm": 0.7133946111348964, | |
| "learning_rate": 1.752483539436807e-05, | |
| "loss": 0.5934, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.141772151898734, | |
| "grad_norm": 0.6255957727496456, | |
| "learning_rate": 1.7420980794800013e-05, | |
| "loss": 0.5612, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.1721518987341772, | |
| "grad_norm": 0.5201955507195181, | |
| "learning_rate": 1.731531335263669e-05, | |
| "loss": 0.5544, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.2025316455696204, | |
| "grad_norm": 0.5548469491746606, | |
| "learning_rate": 1.720785888096743e-05, | |
| "loss": 0.5732, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.232911392405063, | |
| "grad_norm": 0.5412031478636811, | |
| "learning_rate": 1.7098643629428035e-05, | |
| "loss": 0.578, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.2632911392405064, | |
| "grad_norm": 0.5596902846678067, | |
| "learning_rate": 1.698769427778842e-05, | |
| "loss": 0.597, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.293670886075949, | |
| "grad_norm": 0.4642928265318802, | |
| "learning_rate": 1.687503792943506e-05, | |
| "loss": 0.5728, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.3240506329113924, | |
| "grad_norm": 0.5684530184364521, | |
| "learning_rate": 1.6760702104750046e-05, | |
| "loss": 0.5612, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 2.3544303797468356, | |
| "grad_norm": 0.587761347353498, | |
| "learning_rate": 1.664471473438822e-05, | |
| "loss": 0.5835, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 2.3848101265822783, | |
| "grad_norm": 0.5676453892862477, | |
| "learning_rate": 1.6527104152454096e-05, | |
| "loss": 0.5732, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 2.4151898734177215, | |
| "grad_norm": 0.5811532287172524, | |
| "learning_rate": 1.6407899089580263e-05, | |
| "loss": 0.5803, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 2.4455696202531647, | |
| "grad_norm": 0.6165999812962104, | |
| "learning_rate": 1.628712866590885e-05, | |
| "loss": 0.5703, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.4759493670886075, | |
| "grad_norm": 0.4785039990428188, | |
| "learning_rate": 1.6164822383977912e-05, | |
| "loss": 0.5549, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 2.5063291139240507, | |
| "grad_norm": 0.6298659701274344, | |
| "learning_rate": 1.604101012151436e-05, | |
| "loss": 0.5672, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 2.536708860759494, | |
| "grad_norm": 0.5644321954515753, | |
| "learning_rate": 1.5915722124135227e-05, | |
| "loss": 0.5662, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 2.5670886075949366, | |
| "grad_norm": 0.6218709037297634, | |
| "learning_rate": 1.5788988997959115e-05, | |
| "loss": 0.5711, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 2.59746835443038, | |
| "grad_norm": 0.483470485767467, | |
| "learning_rate": 1.5660841702129533e-05, | |
| "loss": 0.549, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 2.6278481012658226, | |
| "grad_norm": 0.6260945484819205, | |
| "learning_rate": 1.5531311541251995e-05, | |
| "loss": 0.5727, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 2.6582278481012658, | |
| "grad_norm": 0.46819641721178673, | |
| "learning_rate": 1.540043015774676e-05, | |
| "loss": 0.5587, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 2.688607594936709, | |
| "grad_norm": 0.5850385934592891, | |
| "learning_rate": 1.5268229524119007e-05, | |
| "loss": 0.5906, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 2.7189873417721517, | |
| "grad_norm": 0.5233339618760808, | |
| "learning_rate": 1.513474193514842e-05, | |
| "loss": 0.589, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 2.749367088607595, | |
| "grad_norm": 0.5616509166348814, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 0.5632, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 2.779746835443038, | |
| "grad_norm": 0.45245126213981257, | |
| "learning_rate": 1.4864036634258112e-05, | |
| "loss": 0.5686, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 2.810126582278481, | |
| "grad_norm": 0.5439835967359893, | |
| "learning_rate": 1.4726885051885654e-05, | |
| "loss": 0.5581, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 2.840506329113924, | |
| "grad_norm": 0.44165711379036954, | |
| "learning_rate": 1.4588578757110359e-05, | |
| "loss": 0.5644, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 2.8708860759493673, | |
| "grad_norm": 0.5958290443054869, | |
| "learning_rate": 1.4449151536240167e-05, | |
| "loss": 0.5776, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 2.90126582278481, | |
| "grad_norm": 0.5029863306187736, | |
| "learning_rate": 1.4308637449409705e-05, | |
| "loss": 0.5757, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 2.9316455696202532, | |
| "grad_norm": 0.4846327994381222, | |
| "learning_rate": 1.4167070822259868e-05, | |
| "loss": 0.5633, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.962025316455696, | |
| "grad_norm": 0.48833140699321187, | |
| "learning_rate": 1.402448623755254e-05, | |
| "loss": 0.5617, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 2.992405063291139, | |
| "grad_norm": 0.40860856061511125, | |
| "learning_rate": 1.3880918526722497e-05, | |
| "loss": 0.562, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 3.030379746835443, | |
| "grad_norm": 0.6299291846160827, | |
| "learning_rate": 1.3736402761368597e-05, | |
| "loss": 0.5019, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 3.060759493670886, | |
| "grad_norm": 0.43019937274303754, | |
| "learning_rate": 1.3590974244686248e-05, | |
| "loss": 0.5174, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.091139240506329, | |
| "grad_norm": 0.6203019431275842, | |
| "learning_rate": 1.344466850284333e-05, | |
| "loss": 0.5131, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 3.1215189873417724, | |
| "grad_norm": 0.5079899053313864, | |
| "learning_rate": 1.3297521276301666e-05, | |
| "loss": 0.5194, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 3.151898734177215, | |
| "grad_norm": 0.5648554886592537, | |
| "learning_rate": 1.3149568511086104e-05, | |
| "loss": 0.5194, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 3.1822784810126583, | |
| "grad_norm": 0.4167717449048031, | |
| "learning_rate": 1.300084635000341e-05, | |
| "loss": 0.5005, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 3.212658227848101, | |
| "grad_norm": 0.4493463744421846, | |
| "learning_rate": 1.2851391123813075e-05, | |
| "loss": 0.4774, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 3.2430379746835443, | |
| "grad_norm": 0.5341842554146895, | |
| "learning_rate": 1.2701239342352223e-05, | |
| "loss": 0.518, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 3.2734177215189875, | |
| "grad_norm": 0.4104371320674546, | |
| "learning_rate": 1.2550427685616767e-05, | |
| "loss": 0.5063, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 3.3037974683544302, | |
| "grad_norm": 0.5119453432507418, | |
| "learning_rate": 1.239899299480098e-05, | |
| "loss": 0.5125, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 3.3341772151898734, | |
| "grad_norm": 0.44961357092490245, | |
| "learning_rate": 1.2246972263297718e-05, | |
| "loss": 0.5069, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 3.3645569620253166, | |
| "grad_norm": 0.4654434075198753, | |
| "learning_rate": 1.2094402627661447e-05, | |
| "loss": 0.5072, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 3.3949367088607594, | |
| "grad_norm": 0.5289460185994969, | |
| "learning_rate": 1.1941321358536278e-05, | |
| "loss": 0.5035, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 3.4253164556962026, | |
| "grad_norm": 0.5029468147498375, | |
| "learning_rate": 1.1787765851551296e-05, | |
| "loss": 0.4948, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 3.4556962025316453, | |
| "grad_norm": 0.4814829762362206, | |
| "learning_rate": 1.1633773618185302e-05, | |
| "loss": 0.5072, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 3.4860759493670885, | |
| "grad_norm": 0.44434543507405055, | |
| "learning_rate": 1.14793822766033e-05, | |
| "loss": 0.517, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 3.5164556962025317, | |
| "grad_norm": 0.405485709565862, | |
| "learning_rate": 1.132462954246688e-05, | |
| "loss": 0.5096, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 3.546835443037975, | |
| "grad_norm": 0.39124685823947064, | |
| "learning_rate": 1.1169553219720828e-05, | |
| "loss": 0.5104, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 3.5772151898734177, | |
| "grad_norm": 0.4053741810907782, | |
| "learning_rate": 1.1014191191358118e-05, | |
| "loss": 0.5047, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 3.607594936708861, | |
| "grad_norm": 0.3841931882034226, | |
| "learning_rate": 1.085858141016566e-05, | |
| "loss": 0.4963, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 3.6379746835443036, | |
| "grad_norm": 0.4578499293042073, | |
| "learning_rate": 1.070276188945293e-05, | |
| "loss": 0.4875, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 3.668354430379747, | |
| "grad_norm": 0.36892934290968277, | |
| "learning_rate": 1.0546770693765859e-05, | |
| "loss": 0.5234, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 3.69873417721519, | |
| "grad_norm": 0.40610301048783426, | |
| "learning_rate": 1.0390645929588197e-05, | |
| "loss": 0.5135, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 3.729113924050633, | |
| "grad_norm": 0.4712758211471604, | |
| "learning_rate": 1.0234425736032607e-05, | |
| "loss": 0.4929, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 3.759493670886076, | |
| "grad_norm": 0.4259049508092099, | |
| "learning_rate": 1.007814827552384e-05, | |
| "loss": 0.4924, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 3.7898734177215188, | |
| "grad_norm": 0.47609046485055867, | |
| "learning_rate": 9.92185172447616e-06, | |
| "loss": 0.5194, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 3.820253164556962, | |
| "grad_norm": 0.39069761757795607, | |
| "learning_rate": 9.765574263967397e-06, | |
| "loss": 0.5005, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 3.850632911392405, | |
| "grad_norm": 0.4522604147588625, | |
| "learning_rate": 9.609354070411807e-06, | |
| "loss": 0.5244, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 3.8810126582278484, | |
| "grad_norm": 0.35949402870378555, | |
| "learning_rate": 9.453229306234143e-06, | |
| "loss": 0.5034, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 3.911392405063291, | |
| "grad_norm": 0.4470166980321721, | |
| "learning_rate": 9.297238110547075e-06, | |
| "loss": 0.5124, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 3.9417721518987343, | |
| "grad_norm": 0.3756150762174879, | |
| "learning_rate": 9.14141858983434e-06, | |
| "loss": 0.5013, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 3.972151898734177, | |
| "grad_norm": 0.3472045448111381, | |
| "learning_rate": 8.985808808641883e-06, | |
| "loss": 0.4974, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 4.010126582278481, | |
| "grad_norm": 0.4147288900857289, | |
| "learning_rate": 8.830446780279175e-06, | |
| "loss": 0.4917, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 4.040506329113924, | |
| "grad_norm": 0.4949640947041675, | |
| "learning_rate": 8.675370457533122e-06, | |
| "loss": 0.447, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 4.0708860759493675, | |
| "grad_norm": 0.41192343210057647, | |
| "learning_rate": 8.520617723396702e-06, | |
| "loss": 0.4599, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 4.10126582278481, | |
| "grad_norm": 0.6456464532807268, | |
| "learning_rate": 8.366226381814698e-06, | |
| "loss": 0.4597, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 4.131645569620253, | |
| "grad_norm": 0.42226990451827806, | |
| "learning_rate": 8.212234148448708e-06, | |
| "loss": 0.4433, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 4.162025316455696, | |
| "grad_norm": 0.5357450855193673, | |
| "learning_rate": 8.058678641463724e-06, | |
| "loss": 0.4462, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 4.192405063291139, | |
| "grad_norm": 0.47766968502977414, | |
| "learning_rate": 7.905597372338558e-06, | |
| "loss": 0.4453, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 4.222784810126583, | |
| "grad_norm": 0.4058986041542239, | |
| "learning_rate": 7.753027736702283e-06, | |
| "loss": 0.4525, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 4.253164556962025, | |
| "grad_norm": 0.45106346693188104, | |
| "learning_rate": 7.601007005199022e-06, | |
| "loss": 0.4589, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 4.283544303797468, | |
| "grad_norm": 0.4960166489310215, | |
| "learning_rate": 7.449572314383237e-06, | |
| "loss": 0.4394, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 4.313924050632911, | |
| "grad_norm": 0.4123398601141748, | |
| "learning_rate": 7.298760657647779e-06, | |
| "loss": 0.4652, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 4.3443037974683545, | |
| "grad_norm": 0.42897403506101073, | |
| "learning_rate": 7.148608876186931e-06, | |
| "loss": 0.4633, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 4.374683544303798, | |
| "grad_norm": 0.4599117336005473, | |
| "learning_rate": 6.999153649996595e-06, | |
| "loss": 0.4589, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 4.405063291139241, | |
| "grad_norm": 0.42110484450736085, | |
| "learning_rate": 6.8504314889138956e-06, | |
| "loss": 0.4422, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 4.435443037974683, | |
| "grad_norm": 0.3834427101666007, | |
| "learning_rate": 6.702478723698336e-06, | |
| "loss": 0.4531, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 4.465822784810126, | |
| "grad_norm": 0.4720883116815651, | |
| "learning_rate": 6.555331497156671e-06, | |
| "loss": 0.4459, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 4.49620253164557, | |
| "grad_norm": 0.38634051586781215, | |
| "learning_rate": 6.4090257553137566e-06, | |
| "loss": 0.4687, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 4.526582278481013, | |
| "grad_norm": 0.38410612525070503, | |
| "learning_rate": 6.263597238631405e-06, | |
| "loss": 0.4512, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 4.556962025316456, | |
| "grad_norm": 0.3928743944118678, | |
| "learning_rate": 6.119081473277502e-06, | |
| "loss": 0.4546, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 4.587341772151898, | |
| "grad_norm": 0.38293233090199696, | |
| "learning_rate": 5.975513762447465e-06, | |
| "loss": 0.4436, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 4.6177215189873415, | |
| "grad_norm": 0.33030474913019137, | |
| "learning_rate": 5.832929177740134e-06, | |
| "loss": 0.4356, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 4.648101265822785, | |
| "grad_norm": 0.39933773263322075, | |
| "learning_rate": 5.6913625505902966e-06, | |
| "loss": 0.4443, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 4.678481012658228, | |
| "grad_norm": 0.40292474443009385, | |
| "learning_rate": 5.550848463759835e-06, | |
| "loss": 0.4448, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 4.708860759493671, | |
| "grad_norm": 0.35167578898849355, | |
| "learning_rate": 5.411421242889643e-06, | |
| "loss": 0.4504, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 4.739240506329114, | |
| "grad_norm": 0.33025760749545424, | |
| "learning_rate": 5.273114948114346e-06, | |
| "loss": 0.4668, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 4.769620253164557, | |
| "grad_norm": 0.38459442154779877, | |
| "learning_rate": 5.135963365741892e-06, | |
| "loss": 0.4616, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.3604910960328784, | |
| "learning_rate": 5.000000000000003e-06, | |
| "loss": 0.4553, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 4.830379746835443, | |
| "grad_norm": 0.3430106954484483, | |
| "learning_rate": 4.865258064851579e-06, | |
| "loss": 0.4611, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 4.860759493670886, | |
| "grad_norm": 0.4262931441090022, | |
| "learning_rate": 4.731770475880995e-06, | |
| "loss": 0.4517, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 4.891139240506329, | |
| "grad_norm": 0.33536151186260815, | |
| "learning_rate": 4.599569842253244e-06, | |
| "loss": 0.4536, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 4.921518987341772, | |
| "grad_norm": 0.28974166895444037, | |
| "learning_rate": 4.468688458748006e-06, | |
| "loss": 0.4701, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 4.951898734177215, | |
| "grad_norm": 0.35301091353554587, | |
| "learning_rate": 4.339158297870469e-06, | |
| "loss": 0.4598, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 4.982278481012658, | |
| "grad_norm": 0.3375144467587323, | |
| "learning_rate": 4.211011002040885e-06, | |
| "loss": 0.4527, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 5.020253164556962, | |
| "grad_norm": 0.37769144548243655, | |
| "learning_rate": 4.084277875864776e-06, | |
| "loss": 0.4285, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 5.050632911392405, | |
| "grad_norm": 0.3768520536113485, | |
| "learning_rate": 3.958989878485644e-06, | |
| "loss": 0.4124, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 5.0810126582278485, | |
| "grad_norm": 0.3429124461235311, | |
| "learning_rate": 3.83517761602209e-06, | |
| "loss": 0.4333, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 5.111392405063291, | |
| "grad_norm": 0.4202768233679485, | |
| "learning_rate": 3.712871334091154e-06, | |
| "loss": 0.4124, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 5.141772151898734, | |
| "grad_norm": 0.4301228370217613, | |
| "learning_rate": 3.592100910419738e-06, | |
| "loss": 0.417, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 5.172151898734177, | |
| "grad_norm": 0.33370420405313156, | |
| "learning_rate": 3.4728958475459052e-06, | |
| "loss": 0.4097, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 5.2025316455696204, | |
| "grad_norm": 0.4000947970665193, | |
| "learning_rate": 3.355285265611784e-06, | |
| "loss": 0.3978, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 5.232911392405064, | |
| "grad_norm": 0.34075486092649226, | |
| "learning_rate": 3.2392978952499553e-06, | |
| "loss": 0.4204, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 5.263291139240506, | |
| "grad_norm": 0.30953181651305767, | |
| "learning_rate": 3.1249620705649417e-06, | |
| "loss": 0.4159, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 5.293670886075949, | |
| "grad_norm": 0.30676017789087784, | |
| "learning_rate": 3.0123057222115835e-06, | |
| "loss": 0.4389, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 5.324050632911392, | |
| "grad_norm": 0.3494822270129837, | |
| "learning_rate": 2.9013563705719673e-06, | |
| "loss": 0.4134, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 5.3544303797468356, | |
| "grad_norm": 0.3387258993173036, | |
| "learning_rate": 2.7921411190325753e-06, | |
| "loss": 0.4184, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 5.384810126582279, | |
| "grad_norm": 0.306357467583021, | |
| "learning_rate": 2.6846866473633126e-06, | |
| "loss": 0.4191, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 5.415189873417722, | |
| "grad_norm": 0.30602145401870245, | |
| "learning_rate": 2.579019205199992e-06, | |
| "loss": 0.4117, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 5.445569620253164, | |
| "grad_norm": 0.31473156390439483, | |
| "learning_rate": 2.4751646056319334e-06, | |
| "loss": 0.4099, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 5.4759493670886075, | |
| "grad_norm": 0.3378024057016722, | |
| "learning_rate": 2.373148218896182e-06, | |
| "loss": 0.4034, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 5.506329113924051, | |
| "grad_norm": 0.28129515492668866, | |
| "learning_rate": 2.2729949661798876e-06, | |
| "loss": 0.4107, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 5.536708860759494, | |
| "grad_norm": 0.2837627779454437, | |
| "learning_rate": 2.174729313532433e-06, | |
| "loss": 0.4222, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 5.567088607594937, | |
| "grad_norm": 0.29098892496742607, | |
| "learning_rate": 2.078375265888707e-06, | |
| "loss": 0.4115, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 5.597468354430379, | |
| "grad_norm": 0.2984167198433826, | |
| "learning_rate": 1.9839563612050273e-06, | |
| "loss": 0.4143, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 5.627848101265823, | |
| "grad_norm": 0.3057434815558622, | |
| "learning_rate": 1.8914956647091497e-06, | |
| "loss": 0.3894, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 5.658227848101266, | |
| "grad_norm": 0.2926421836839038, | |
| "learning_rate": 1.8010157632657544e-06, | |
| "loss": 0.4062, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 5.688607594936709, | |
| "grad_norm": 0.3180222545785948, | |
| "learning_rate": 1.7125387598587862e-06, | |
| "loss": 0.41, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 5.718987341772152, | |
| "grad_norm": 0.2742323493317311, | |
| "learning_rate": 1.6260862681919965e-06, | |
| "loss": 0.4318, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 5.749367088607595, | |
| "grad_norm": 0.3627767121940232, | |
| "learning_rate": 1.5416794074090258e-06, | |
| "loss": 0.4242, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 5.779746835443038, | |
| "grad_norm": 0.2921954756930859, | |
| "learning_rate": 1.459338796934293e-06, | |
| "loss": 0.4055, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 5.810126582278481, | |
| "grad_norm": 0.26712626210105345, | |
| "learning_rate": 1.3790845514359363e-06, | |
| "loss": 0.4309, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 5.840506329113924, | |
| "grad_norm": 0.2701186183955813, | |
| "learning_rate": 1.300936275912098e-06, | |
| "loss": 0.425, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 5.870886075949367, | |
| "grad_norm": 0.2627917149246142, | |
| "learning_rate": 1.224913060901688e-06, | |
| "loss": 0.4174, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 5.9012658227848105, | |
| "grad_norm": 0.2610414568051843, | |
| "learning_rate": 1.1510334778208332e-06, | |
| "loss": 0.4205, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 5.931645569620253, | |
| "grad_norm": 0.25945995273662487, | |
| "learning_rate": 1.0793155744261352e-06, | |
| "loss": 0.4291, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 5.962025316455696, | |
| "grad_norm": 0.26816187987441314, | |
| "learning_rate": 1.0097768704058542e-06, | |
| "loss": 0.4253, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 5.992405063291139, | |
| "grad_norm": 0.26354021725379706, | |
| "learning_rate": 9.424343531000968e-07, | |
| "loss": 0.4334, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 6.030379746835443, | |
| "grad_norm": 0.30157984436726143, | |
| "learning_rate": 8.773044733510338e-07, | |
| "loss": 0.3926, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 6.060759493670886, | |
| "grad_norm": 0.2885651709754447, | |
| "learning_rate": 8.144031414842012e-07, | |
| "loss": 0.4068, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 6.091139240506329, | |
| "grad_norm": 0.28226203364443786, | |
| "learning_rate": 7.537457234218271e-07, | |
| "loss": 0.386, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 6.121518987341772, | |
| "grad_norm": 0.27561219835716777, | |
| "learning_rate": 6.953470369291349e-07, | |
| "loss": 0.4037, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 6.151898734177215, | |
| "grad_norm": 0.2612622644180863, | |
| "learning_rate": 6.392213479945852e-07, | |
| "loss": 0.4137, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 6.182278481012658, | |
| "grad_norm": 0.2746383708702675, | |
| "learning_rate": 5.853823673448877e-07, | |
| "loss": 0.3923, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 6.2126582278481015, | |
| "grad_norm": 0.282890057435792, | |
| "learning_rate": 5.33843247095659e-07, | |
| "loss": 0.4049, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 6.243037974683545, | |
| "grad_norm": 0.282662224026126, | |
| "learning_rate": 4.846165775385459e-07, | |
| "loss": 0.4005, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 6.273417721518987, | |
| "grad_norm": 0.2636697818963676, | |
| "learning_rate": 4.3771438406559173e-07, | |
| "loss": 0.4, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 6.30379746835443, | |
| "grad_norm": 0.26818472120808856, | |
| "learning_rate": 3.931481242315993e-07, | |
| "loss": 0.4224, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 6.334177215189873, | |
| "grad_norm": 0.26508236316113604, | |
| "learning_rate": 3.5092868495520294e-07, | |
| "loss": 0.3959, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 6.364556962025317, | |
| "grad_norm": 0.26453184297201304, | |
| "learning_rate": 3.110663798593616e-07, | |
| "loss": 0.4101, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 6.39493670886076, | |
| "grad_norm": 0.250489489623135, | |
| "learning_rate": 2.735709467518699e-07, | |
| "loss": 0.3912, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 6.425316455696202, | |
| "grad_norm": 0.2603953509889015, | |
| "learning_rate": 2.384515452465475e-07, | |
| "loss": 0.4224, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 6.455696202531645, | |
| "grad_norm": 0.2809355408045964, | |
| "learning_rate": 2.0571675452567997e-07, | |
| "loss": 0.385, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 6.4860759493670885, | |
| "grad_norm": 0.2524655182414268, | |
| "learning_rate": 1.7537457124423896e-07, | |
| "loss": 0.4014, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 6.516455696202532, | |
| "grad_norm": 0.2583570571577369, | |
| "learning_rate": 1.474324075764111e-07, | |
| "loss": 0.3878, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 6.546835443037975, | |
| "grad_norm": 0.2525481155477391, | |
| "learning_rate": 1.2189708940490653e-07, | |
| "loss": 0.4035, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 6.577215189873417, | |
| "grad_norm": 0.25710848829507454, | |
| "learning_rate": 9.877485465349057e-08, | |
| "loss": 0.4061, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 6.6075949367088604, | |
| "grad_norm": 0.261556719366437, | |
| "learning_rate": 7.807135176314707e-08, | |
| "loss": 0.401, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 6.637974683544304, | |
| "grad_norm": 0.27008961374708734, | |
| "learning_rate": 5.979163831223988e-08, | |
| "loss": 0.382, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 6.668354430379747, | |
| "grad_norm": 0.2600499779083102, | |
| "learning_rate": 4.394017978101905e-08, | |
| "loss": 0.3885, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 6.69873417721519, | |
| "grad_norm": 0.24694462505095843, | |
| "learning_rate": 3.0520848460765525e-08, | |
| "loss": 0.4156, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 6.729113924050633, | |
| "grad_norm": 0.2550150521416793, | |
| "learning_rate": 1.9536922507841227e-08, | |
| "loss": 0.4179, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 6.759493670886076, | |
| "grad_norm": 0.2468813851627281, | |
| "learning_rate": 1.099108514288627e-08, | |
| "loss": 0.4069, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 6.789873417721519, | |
| "grad_norm": 0.24759797511768317, | |
| "learning_rate": 4.885423995341088e-09, | |
| "loss": 0.41, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 6.820253164556962, | |
| "grad_norm": 0.26649360169852315, | |
| "learning_rate": 1.2214305934699078e-09, | |
| "loss": 0.3975, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 6.850632911392405, | |
| "grad_norm": 0.2493142178698674, | |
| "learning_rate": 0.0, | |
| "loss": 0.4103, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 6.850632911392405, | |
| "step": 224, | |
| "total_flos": 9.023420808004895e+17, | |
| "train_loss": 0.5525622120393174, | |
| "train_runtime": 27715.8163, | |
| "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": 9.023420808004895e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |