| { | |
| "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": 6.37221868985156, | |
| "learning_rate": 8.695652173913044e-07, | |
| "loss": 1.0264, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.060759493670886074, | |
| "grad_norm": 6.5034802963479175, | |
| "learning_rate": 1.7391304347826088e-06, | |
| "loss": 1.0424, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.09113924050632911, | |
| "grad_norm": 6.380172477883636, | |
| "learning_rate": 2.6086956521739132e-06, | |
| "loss": 1.0317, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.12151898734177215, | |
| "grad_norm": 5.817576642017238, | |
| "learning_rate": 3.4782608695652175e-06, | |
| "loss": 1.0183, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.1518987341772152, | |
| "grad_norm": 4.401190148605733, | |
| "learning_rate": 4.347826086956522e-06, | |
| "loss": 0.9624, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.18227848101265823, | |
| "grad_norm": 2.7097647034322696, | |
| "learning_rate": 5.2173913043478265e-06, | |
| "loss": 0.959, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.21265822784810126, | |
| "grad_norm": 2.450867089297927, | |
| "learning_rate": 6.086956521739132e-06, | |
| "loss": 0.9543, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.2430379746835443, | |
| "grad_norm": 4.04136120197139, | |
| "learning_rate": 6.956521739130435e-06, | |
| "loss": 0.9325, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.27341772151898736, | |
| "grad_norm": 4.409279955403386, | |
| "learning_rate": 7.82608695652174e-06, | |
| "loss": 0.9357, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.3037974683544304, | |
| "grad_norm": 3.706439164950476, | |
| "learning_rate": 8.695652173913044e-06, | |
| "loss": 0.9084, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3341772151898734, | |
| "grad_norm": 3.5400094637960513, | |
| "learning_rate": 9.565217391304349e-06, | |
| "loss": 0.8941, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.36455696202531646, | |
| "grad_norm": 2.9424193684317865, | |
| "learning_rate": 1.0434782608695653e-05, | |
| "loss": 0.8536, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.3949367088607595, | |
| "grad_norm": 2.1302467308428508, | |
| "learning_rate": 1.1304347826086957e-05, | |
| "loss": 0.8482, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.4253164556962025, | |
| "grad_norm": 1.481291511968154, | |
| "learning_rate": 1.2173913043478263e-05, | |
| "loss": 0.8522, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.45569620253164556, | |
| "grad_norm": 1.5948371689967598, | |
| "learning_rate": 1.3043478260869566e-05, | |
| "loss": 0.8078, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.4860759493670886, | |
| "grad_norm": 1.3238320150665204, | |
| "learning_rate": 1.391304347826087e-05, | |
| "loss": 0.7868, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.5164556962025316, | |
| "grad_norm": 1.135690386872903, | |
| "learning_rate": 1.4782608695652174e-05, | |
| "loss": 0.7968, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.5468354430379747, | |
| "grad_norm": 1.1319959126772645, | |
| "learning_rate": 1.565217391304348e-05, | |
| "loss": 0.785, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.5772151898734177, | |
| "grad_norm": 0.9567822025036028, | |
| "learning_rate": 1.6521739130434785e-05, | |
| "loss": 0.7757, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.6075949367088608, | |
| "grad_norm": 1.0091484315875876, | |
| "learning_rate": 1.739130434782609e-05, | |
| "loss": 0.7549, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.6379746835443038, | |
| "grad_norm": 1.048072111801254, | |
| "learning_rate": 1.8260869565217393e-05, | |
| "loss": 0.7415, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.6683544303797468, | |
| "grad_norm": 0.9493693220268987, | |
| "learning_rate": 1.9130434782608697e-05, | |
| "loss": 0.7361, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.6987341772151898, | |
| "grad_norm": 0.9262684629013186, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7422, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.7291139240506329, | |
| "grad_norm": 0.8741096470542656, | |
| "learning_rate": 1.999877856940653e-05, | |
| "loss": 0.7486, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.759493670886076, | |
| "grad_norm": 0.7873818311147306, | |
| "learning_rate": 1.999511457600466e-05, | |
| "loss": 0.7235, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.789873417721519, | |
| "grad_norm": 0.6420798721747906, | |
| "learning_rate": 1.9989008914857115e-05, | |
| "loss": 0.7188, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.8202531645569621, | |
| "grad_norm": 0.7203837580063897, | |
| "learning_rate": 1.998046307749216e-05, | |
| "loss": 0.7193, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.850632911392405, | |
| "grad_norm": 0.8500834268635933, | |
| "learning_rate": 1.9969479151539238e-05, | |
| "loss": 0.7127, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.8810126582278481, | |
| "grad_norm": 0.6058389288197817, | |
| "learning_rate": 1.9956059820218982e-05, | |
| "loss": 0.7259, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.9113924050632911, | |
| "grad_norm": 0.7078168051247413, | |
| "learning_rate": 1.9940208361687762e-05, | |
| "loss": 0.6921, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.9417721518987342, | |
| "grad_norm": 0.7015805895369476, | |
| "learning_rate": 1.9921928648236855e-05, | |
| "loss": 0.7197, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.9721518987341772, | |
| "grad_norm": 0.7387848558163191, | |
| "learning_rate": 1.990122514534651e-05, | |
| "loss": 0.6959, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.010126582278481, | |
| "grad_norm": 0.5763144897662861, | |
| "learning_rate": 1.9878102910595097e-05, | |
| "loss": 0.6832, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.040506329113924, | |
| "grad_norm": 0.6437270212433482, | |
| "learning_rate": 1.985256759242359e-05, | |
| "loss": 0.6611, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.070886075949367, | |
| "grad_norm": 0.5953857969572114, | |
| "learning_rate": 1.982462542875576e-05, | |
| "loss": 0.6687, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.1012658227848102, | |
| "grad_norm": 0.6001397001527595, | |
| "learning_rate": 1.979428324547432e-05, | |
| "loss": 0.6662, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.1316455696202532, | |
| "grad_norm": 0.6353749730621574, | |
| "learning_rate": 1.9761548454753455e-05, | |
| "loss": 0.6533, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.1620253164556962, | |
| "grad_norm": 0.5856471519075721, | |
| "learning_rate": 1.972642905324813e-05, | |
| "loss": 0.646, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.1924050632911392, | |
| "grad_norm": 0.5360846001231565, | |
| "learning_rate": 1.9688933620140638e-05, | |
| "loss": 0.6501, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.2227848101265824, | |
| "grad_norm": 0.5361166895681458, | |
| "learning_rate": 1.96490713150448e-05, | |
| "loss": 0.6444, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.2531645569620253, | |
| "grad_norm": 0.43296430499453425, | |
| "learning_rate": 1.9606851875768404e-05, | |
| "loss": 0.6456, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.2835443037974683, | |
| "grad_norm": 0.526080735366251, | |
| "learning_rate": 1.956228561593441e-05, | |
| "loss": 0.6494, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.3139240506329113, | |
| "grad_norm": 0.4906847188578764, | |
| "learning_rate": 1.9515383422461457e-05, | |
| "loss": 0.6527, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.3443037974683545, | |
| "grad_norm": 0.5680631359585419, | |
| "learning_rate": 1.9466156752904344e-05, | |
| "loss": 0.6447, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.3746835443037975, | |
| "grad_norm": 0.5130133069861942, | |
| "learning_rate": 1.9414617632655114e-05, | |
| "loss": 0.6314, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.4050632911392404, | |
| "grad_norm": 0.47327015900025143, | |
| "learning_rate": 1.9360778652005416e-05, | |
| "loss": 0.6304, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.4354430379746836, | |
| "grad_norm": 0.6242281508405784, | |
| "learning_rate": 1.9304652963070868e-05, | |
| "loss": 0.6398, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.4658227848101266, | |
| "grad_norm": 0.5892904102401884, | |
| "learning_rate": 1.9246254276578175e-05, | |
| "loss": 0.6248, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.4962025316455696, | |
| "grad_norm": 0.6531942163155297, | |
| "learning_rate": 1.9185596858515797e-05, | |
| "loss": 0.6404, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.5265822784810128, | |
| "grad_norm": 0.6483567267386434, | |
| "learning_rate": 1.9122695526648968e-05, | |
| "loss": 0.6365, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.5569620253164556, | |
| "grad_norm": 0.6983956278308369, | |
| "learning_rate": 1.905756564689991e-05, | |
| "loss": 0.643, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.5873417721518988, | |
| "grad_norm": 0.7862995399359736, | |
| "learning_rate": 1.8990223129594146e-05, | |
| "loss": 0.6419, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.6177215189873417, | |
| "grad_norm": 0.5430925229786947, | |
| "learning_rate": 1.8920684425573865e-05, | |
| "loss": 0.6324, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.6481012658227847, | |
| "grad_norm": 0.885437054994622, | |
| "learning_rate": 1.884896652217917e-05, | |
| "loss": 0.6356, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.678481012658228, | |
| "grad_norm": 0.5143264601586393, | |
| "learning_rate": 1.877508693909831e-05, | |
| "loss": 0.6333, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.7088607594936709, | |
| "grad_norm": 0.8076187986221476, | |
| "learning_rate": 1.8699063724087905e-05, | |
| "loss": 0.6271, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.7392405063291139, | |
| "grad_norm": 0.6430936572112902, | |
| "learning_rate": 1.862091544856407e-05, | |
| "loss": 0.6277, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.769620253164557, | |
| "grad_norm": 0.4956450969395412, | |
| "learning_rate": 1.854066120306571e-05, | |
| "loss": 0.6035, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.4950456942853279, | |
| "learning_rate": 1.8458320592590976e-05, | |
| "loss": 0.6386, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.830379746835443, | |
| "grad_norm": 0.6371116080859314, | |
| "learning_rate": 1.837391373180801e-05, | |
| "loss": 0.6229, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.8607594936708862, | |
| "grad_norm": 0.43714150441878535, | |
| "learning_rate": 1.8287461240141217e-05, | |
| "loss": 0.6176, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.891139240506329, | |
| "grad_norm": 0.6453784990074324, | |
| "learning_rate": 1.8198984236734246e-05, | |
| "loss": 0.6077, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.9215189873417722, | |
| "grad_norm": 0.5570607723349542, | |
| "learning_rate": 1.8108504335290852e-05, | |
| "loss": 0.6352, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.9518987341772152, | |
| "grad_norm": 0.48245224757972355, | |
| "learning_rate": 1.8016043638794975e-05, | |
| "loss": 0.637, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.9822784810126581, | |
| "grad_norm": 0.5721748046329147, | |
| "learning_rate": 1.7921624734111292e-05, | |
| "loss": 0.6076, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.020253164556962, | |
| "grad_norm": 0.5141868421323431, | |
| "learning_rate": 1.7825270686467567e-05, | |
| "loss": 0.5894, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.050632911392405, | |
| "grad_norm": 0.5381266684759717, | |
| "learning_rate": 1.7727005033820117e-05, | |
| "loss": 0.5556, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.081012658227848, | |
| "grad_norm": 0.6016121815074985, | |
| "learning_rate": 1.762685178110382e-05, | |
| "loss": 0.5527, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.1113924050632913, | |
| "grad_norm": 0.6729531958813787, | |
| "learning_rate": 1.752483539436807e-05, | |
| "loss": 0.563, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.141772151898734, | |
| "grad_norm": 0.5102620941806387, | |
| "learning_rate": 1.7420980794800013e-05, | |
| "loss": 0.5718, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.1721518987341772, | |
| "grad_norm": 0.6042146458914103, | |
| "learning_rate": 1.731531335263669e-05, | |
| "loss": 0.5608, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.2025316455696204, | |
| "grad_norm": 0.5894071656765725, | |
| "learning_rate": 1.720785888096743e-05, | |
| "loss": 0.5699, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.232911392405063, | |
| "grad_norm": 0.5775536328227947, | |
| "learning_rate": 1.7098643629428035e-05, | |
| "loss": 0.5625, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.2632911392405064, | |
| "grad_norm": 0.606548495008496, | |
| "learning_rate": 1.698769427778842e-05, | |
| "loss": 0.5404, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.293670886075949, | |
| "grad_norm": 0.6622908431371424, | |
| "learning_rate": 1.687503792943506e-05, | |
| "loss": 0.5747, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.3240506329113924, | |
| "grad_norm": 0.4533766615425194, | |
| "learning_rate": 1.6760702104750046e-05, | |
| "loss": 0.5481, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 2.3544303797468356, | |
| "grad_norm": 0.4811067517469304, | |
| "learning_rate": 1.664471473438822e-05, | |
| "loss": 0.5524, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 2.3848101265822783, | |
| "grad_norm": 0.5210479774326526, | |
| "learning_rate": 1.6527104152454096e-05, | |
| "loss": 0.5372, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 2.4151898734177215, | |
| "grad_norm": 0.546680221442232, | |
| "learning_rate": 1.6407899089580263e-05, | |
| "loss": 0.5534, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 2.4455696202531647, | |
| "grad_norm": 0.4607636332203975, | |
| "learning_rate": 1.628712866590885e-05, | |
| "loss": 0.556, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.4759493670886075, | |
| "grad_norm": 0.5340360502609477, | |
| "learning_rate": 1.6164822383977912e-05, | |
| "loss": 0.5578, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 2.5063291139240507, | |
| "grad_norm": 0.4829659674501385, | |
| "learning_rate": 1.604101012151436e-05, | |
| "loss": 0.5586, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 2.536708860759494, | |
| "grad_norm": 0.5154687639314293, | |
| "learning_rate": 1.5915722124135227e-05, | |
| "loss": 0.5456, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 2.5670886075949366, | |
| "grad_norm": 0.4828149719193968, | |
| "learning_rate": 1.5788988997959115e-05, | |
| "loss": 0.5485, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 2.59746835443038, | |
| "grad_norm": 0.49279954192791947, | |
| "learning_rate": 1.5660841702129533e-05, | |
| "loss": 0.57, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 2.6278481012658226, | |
| "grad_norm": 0.4742602851060561, | |
| "learning_rate": 1.5531311541251995e-05, | |
| "loss": 0.5446, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 2.6582278481012658, | |
| "grad_norm": 0.46089627282095114, | |
| "learning_rate": 1.540043015774676e-05, | |
| "loss": 0.5432, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 2.688607594936709, | |
| "grad_norm": 0.5193521045258312, | |
| "learning_rate": 1.5268229524119007e-05, | |
| "loss": 0.5526, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 2.7189873417721517, | |
| "grad_norm": 0.599090537230937, | |
| "learning_rate": 1.513474193514842e-05, | |
| "loss": 0.5387, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 2.749367088607595, | |
| "grad_norm": 0.5392240932844506, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 0.5644, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 2.779746835443038, | |
| "grad_norm": 0.6484053666330277, | |
| "learning_rate": 1.4864036634258112e-05, | |
| "loss": 0.5529, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 2.810126582278481, | |
| "grad_norm": 0.49558973661776645, | |
| "learning_rate": 1.4726885051885654e-05, | |
| "loss": 0.5574, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 2.840506329113924, | |
| "grad_norm": 0.5477713161891372, | |
| "learning_rate": 1.4588578757110359e-05, | |
| "loss": 0.5342, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 2.8708860759493673, | |
| "grad_norm": 0.5294733295915529, | |
| "learning_rate": 1.4449151536240167e-05, | |
| "loss": 0.56, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 2.90126582278481, | |
| "grad_norm": 0.4969749891984834, | |
| "learning_rate": 1.4308637449409705e-05, | |
| "loss": 0.5594, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 2.9316455696202532, | |
| "grad_norm": 0.4381308375079077, | |
| "learning_rate": 1.4167070822259868e-05, | |
| "loss": 0.5439, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.962025316455696, | |
| "grad_norm": 0.4173526948473983, | |
| "learning_rate": 1.402448623755254e-05, | |
| "loss": 0.5461, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 2.992405063291139, | |
| "grad_norm": 0.4328409092651217, | |
| "learning_rate": 1.3880918526722497e-05, | |
| "loss": 0.5489, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 3.030379746835443, | |
| "grad_norm": 0.5284467996380835, | |
| "learning_rate": 1.3736402761368597e-05, | |
| "loss": 0.5022, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 3.060759493670886, | |
| "grad_norm": 0.4698391974283234, | |
| "learning_rate": 1.3590974244686248e-05, | |
| "loss": 0.505, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.091139240506329, | |
| "grad_norm": 0.611188398162724, | |
| "learning_rate": 1.344466850284333e-05, | |
| "loss": 0.506, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 3.1215189873417724, | |
| "grad_norm": 0.5181870143762257, | |
| "learning_rate": 1.3297521276301666e-05, | |
| "loss": 0.5025, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 3.151898734177215, | |
| "grad_norm": 0.4746726810844399, | |
| "learning_rate": 1.3149568511086104e-05, | |
| "loss": 0.4962, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 3.1822784810126583, | |
| "grad_norm": 0.5222598143525973, | |
| "learning_rate": 1.300084635000341e-05, | |
| "loss": 0.4861, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 3.212658227848101, | |
| "grad_norm": 0.43490644698600855, | |
| "learning_rate": 1.2851391123813075e-05, | |
| "loss": 0.4993, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 3.2430379746835443, | |
| "grad_norm": 0.4833019704949156, | |
| "learning_rate": 1.2701239342352223e-05, | |
| "loss": 0.4633, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 3.2734177215189875, | |
| "grad_norm": 0.5030493664394047, | |
| "learning_rate": 1.2550427685616767e-05, | |
| "loss": 0.4863, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 3.3037974683544302, | |
| "grad_norm": 0.4335113333888567, | |
| "learning_rate": 1.239899299480098e-05, | |
| "loss": 0.4799, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 3.3341772151898734, | |
| "grad_norm": 0.4771808496390477, | |
| "learning_rate": 1.2246972263297718e-05, | |
| "loss": 0.4739, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 3.3645569620253166, | |
| "grad_norm": 0.43288468354190207, | |
| "learning_rate": 1.2094402627661447e-05, | |
| "loss": 0.4955, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 3.3949367088607594, | |
| "grad_norm": 0.3851443193327133, | |
| "learning_rate": 1.1941321358536278e-05, | |
| "loss": 0.492, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 3.4253164556962026, | |
| "grad_norm": 0.45280439112604726, | |
| "learning_rate": 1.1787765851551296e-05, | |
| "loss": 0.4927, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 3.4556962025316453, | |
| "grad_norm": 0.3837500484870304, | |
| "learning_rate": 1.1633773618185302e-05, | |
| "loss": 0.4912, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 3.4860759493670885, | |
| "grad_norm": 0.46258069326812357, | |
| "learning_rate": 1.14793822766033e-05, | |
| "loss": 0.506, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 3.5164556962025317, | |
| "grad_norm": 0.4067964537702053, | |
| "learning_rate": 1.132462954246688e-05, | |
| "loss": 0.4854, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 3.546835443037975, | |
| "grad_norm": 0.42990922993772457, | |
| "learning_rate": 1.1169553219720828e-05, | |
| "loss": 0.4929, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 3.5772151898734177, | |
| "grad_norm": 0.38635348880117143, | |
| "learning_rate": 1.1014191191358118e-05, | |
| "loss": 0.4825, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 3.607594936708861, | |
| "grad_norm": 0.3870514422615681, | |
| "learning_rate": 1.085858141016566e-05, | |
| "loss": 0.4776, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 3.6379746835443036, | |
| "grad_norm": 0.3626145330339587, | |
| "learning_rate": 1.070276188945293e-05, | |
| "loss": 0.4771, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 3.668354430379747, | |
| "grad_norm": 0.43606645169013275, | |
| "learning_rate": 1.0546770693765859e-05, | |
| "loss": 0.4893, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 3.69873417721519, | |
| "grad_norm": 0.3799921068914223, | |
| "learning_rate": 1.0390645929588197e-05, | |
| "loss": 0.482, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 3.729113924050633, | |
| "grad_norm": 0.4170885304781762, | |
| "learning_rate": 1.0234425736032607e-05, | |
| "loss": 0.4903, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 3.759493670886076, | |
| "grad_norm": 0.6975619010340954, | |
| "learning_rate": 1.007814827552384e-05, | |
| "loss": 0.4634, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 3.7898734177215188, | |
| "grad_norm": 0.40226723349625243, | |
| "learning_rate": 9.92185172447616e-06, | |
| "loss": 0.4851, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 3.820253164556962, | |
| "grad_norm": 0.3667501915935181, | |
| "learning_rate": 9.765574263967397e-06, | |
| "loss": 0.481, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 3.850632911392405, | |
| "grad_norm": 0.3697874278619689, | |
| "learning_rate": 9.609354070411807e-06, | |
| "loss": 0.4876, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 3.8810126582278484, | |
| "grad_norm": 0.357801393234647, | |
| "learning_rate": 9.453229306234143e-06, | |
| "loss": 0.4911, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 3.911392405063291, | |
| "grad_norm": 0.39163201253823016, | |
| "learning_rate": 9.297238110547075e-06, | |
| "loss": 0.4842, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 3.9417721518987343, | |
| "grad_norm": 0.3996876113061803, | |
| "learning_rate": 9.14141858983434e-06, | |
| "loss": 0.4838, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 3.972151898734177, | |
| "grad_norm": 0.35412118410591714, | |
| "learning_rate": 8.985808808641883e-06, | |
| "loss": 0.4864, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 4.010126582278481, | |
| "grad_norm": 2.067828413495023, | |
| "learning_rate": 8.830446780279175e-06, | |
| "loss": 0.4708, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 4.040506329113924, | |
| "grad_norm": 0.9152902583926013, | |
| "learning_rate": 8.675370457533122e-06, | |
| "loss": 0.4187, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 4.0708860759493675, | |
| "grad_norm": 17.35244540703921, | |
| "learning_rate": 8.520617723396702e-06, | |
| "loss": 0.4409, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 4.10126582278481, | |
| "grad_norm": 1.6577060257645506, | |
| "learning_rate": 8.366226381814698e-06, | |
| "loss": 0.4372, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 4.131645569620253, | |
| "grad_norm": 0.5280317347872258, | |
| "learning_rate": 8.212234148448708e-06, | |
| "loss": 0.4433, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 4.162025316455696, | |
| "grad_norm": 0.6688184987684095, | |
| "learning_rate": 8.058678641463724e-06, | |
| "loss": 0.4361, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 4.192405063291139, | |
| "grad_norm": 1.0058311734202963, | |
| "learning_rate": 7.905597372338558e-06, | |
| "loss": 0.4521, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 4.222784810126583, | |
| "grad_norm": 0.4461031074464891, | |
| "learning_rate": 7.753027736702283e-06, | |
| "loss": 0.4341, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 4.253164556962025, | |
| "grad_norm": 0.5391239885653292, | |
| "learning_rate": 7.601007005199022e-06, | |
| "loss": 0.4425, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 4.283544303797468, | |
| "grad_norm": 0.5301098206506064, | |
| "learning_rate": 7.449572314383237e-06, | |
| "loss": 0.4245, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 4.313924050632911, | |
| "grad_norm": 0.46623949539660703, | |
| "learning_rate": 7.298760657647779e-06, | |
| "loss": 0.4214, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 4.3443037974683545, | |
| "grad_norm": 0.442414367812498, | |
| "learning_rate": 7.148608876186931e-06, | |
| "loss": 0.4384, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 4.374683544303798, | |
| "grad_norm": 0.46015866197369026, | |
| "learning_rate": 6.999153649996595e-06, | |
| "loss": 0.449, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 4.405063291139241, | |
| "grad_norm": 0.4708636486435107, | |
| "learning_rate": 6.8504314889138956e-06, | |
| "loss": 0.4205, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 4.435443037974683, | |
| "grad_norm": 0.3957689916265755, | |
| "learning_rate": 6.702478723698336e-06, | |
| "loss": 0.4371, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 4.465822784810126, | |
| "grad_norm": 0.42482708366220556, | |
| "learning_rate": 6.555331497156671e-06, | |
| "loss": 0.432, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 4.49620253164557, | |
| "grad_norm": 0.48059851550114396, | |
| "learning_rate": 6.4090257553137566e-06, | |
| "loss": 0.4462, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 4.526582278481013, | |
| "grad_norm": 0.3571362661917008, | |
| "learning_rate": 6.263597238631405e-06, | |
| "loss": 0.4353, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 4.556962025316456, | |
| "grad_norm": 0.4167219945081819, | |
| "learning_rate": 6.119081473277502e-06, | |
| "loss": 0.4327, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 4.587341772151898, | |
| "grad_norm": 0.41818988585067385, | |
| "learning_rate": 5.975513762447465e-06, | |
| "loss": 0.4427, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 4.6177215189873415, | |
| "grad_norm": 0.3925674797200369, | |
| "learning_rate": 5.832929177740134e-06, | |
| "loss": 0.4364, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 4.648101265822785, | |
| "grad_norm": 0.40374638803129476, | |
| "learning_rate": 5.6913625505902966e-06, | |
| "loss": 0.4108, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 4.678481012658228, | |
| "grad_norm": 0.388387880997993, | |
| "learning_rate": 5.550848463759835e-06, | |
| "loss": 0.427, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 4.708860759493671, | |
| "grad_norm": 0.3642305851572156, | |
| "learning_rate": 5.411421242889643e-06, | |
| "loss": 0.4269, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 4.739240506329114, | |
| "grad_norm": 0.4082645818703638, | |
| "learning_rate": 5.273114948114346e-06, | |
| "loss": 0.4323, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 4.769620253164557, | |
| "grad_norm": 0.33785963966215543, | |
| "learning_rate": 5.135963365741892e-06, | |
| "loss": 0.4384, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.37611991947395673, | |
| "learning_rate": 5.000000000000003e-06, | |
| "loss": 0.4337, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 4.830379746835443, | |
| "grad_norm": 0.35623502983585953, | |
| "learning_rate": 4.865258064851579e-06, | |
| "loss": 0.4338, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 4.860759493670886, | |
| "grad_norm": 0.3185465826203866, | |
| "learning_rate": 4.731770475880995e-06, | |
| "loss": 0.4268, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 4.891139240506329, | |
| "grad_norm": 0.3568918726471837, | |
| "learning_rate": 4.599569842253244e-06, | |
| "loss": 0.4289, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 4.921518987341772, | |
| "grad_norm": 0.3819530518334745, | |
| "learning_rate": 4.468688458748006e-06, | |
| "loss": 0.4339, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 4.951898734177215, | |
| "grad_norm": 0.3351594331992901, | |
| "learning_rate": 4.339158297870469e-06, | |
| "loss": 0.4331, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 4.982278481012658, | |
| "grad_norm": 0.3021429578703631, | |
| "learning_rate": 4.211011002040885e-06, | |
| "loss": 0.4234, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 5.020253164556962, | |
| "grad_norm": 0.4006915974335434, | |
| "learning_rate": 4.084277875864776e-06, | |
| "loss": 0.4001, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 5.050632911392405, | |
| "grad_norm": 0.36778820176808397, | |
| "learning_rate": 3.958989878485644e-06, | |
| "loss": 0.4168, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 5.0810126582278485, | |
| "grad_norm": 0.32130448433145914, | |
| "learning_rate": 3.83517761602209e-06, | |
| "loss": 0.4065, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 5.111392405063291, | |
| "grad_norm": 0.3809966372789622, | |
| "learning_rate": 3.712871334091154e-06, | |
| "loss": 0.4138, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 5.141772151898734, | |
| "grad_norm": 0.3772267964660737, | |
| "learning_rate": 3.592100910419738e-06, | |
| "loss": 0.4033, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 5.172151898734177, | |
| "grad_norm": 0.3689981910530363, | |
| "learning_rate": 3.4728958475459052e-06, | |
| "loss": 0.4068, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 5.2025316455696204, | |
| "grad_norm": 0.35198321548754313, | |
| "learning_rate": 3.355285265611784e-06, | |
| "loss": 0.3949, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 5.232911392405064, | |
| "grad_norm": 0.35769710232766566, | |
| "learning_rate": 3.2392978952499553e-06, | |
| "loss": 0.39, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 5.263291139240506, | |
| "grad_norm": 0.33611090570761076, | |
| "learning_rate": 3.1249620705649417e-06, | |
| "loss": 0.4065, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 5.293670886075949, | |
| "grad_norm": 0.4049939946491654, | |
| "learning_rate": 3.0123057222115835e-06, | |
| "loss": 0.3778, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 5.324050632911392, | |
| "grad_norm": 0.3210913930686345, | |
| "learning_rate": 2.9013563705719673e-06, | |
| "loss": 0.3921, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 5.3544303797468356, | |
| "grad_norm": 0.3659685382050678, | |
| "learning_rate": 2.7921411190325753e-06, | |
| "loss": 0.3979, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 5.384810126582279, | |
| "grad_norm": 0.36117783642633866, | |
| "learning_rate": 2.6846866473633126e-06, | |
| "loss": 0.4048, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 5.415189873417722, | |
| "grad_norm": 0.30599384631214477, | |
| "learning_rate": 2.579019205199992e-06, | |
| "loss": 0.3918, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 5.445569620253164, | |
| "grad_norm": 0.3158134459858677, | |
| "learning_rate": 2.4751646056319334e-06, | |
| "loss": 0.3977, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 5.4759493670886075, | |
| "grad_norm": 0.3415178405429152, | |
| "learning_rate": 2.373148218896182e-06, | |
| "loss": 0.3941, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 5.506329113924051, | |
| "grad_norm": 0.34056377810339455, | |
| "learning_rate": 2.2729949661798876e-06, | |
| "loss": 0.3989, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 5.536708860759494, | |
| "grad_norm": 0.32197225533734486, | |
| "learning_rate": 2.174729313532433e-06, | |
| "loss": 0.3915, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 5.567088607594937, | |
| "grad_norm": 0.32400566343568027, | |
| "learning_rate": 2.078375265888707e-06, | |
| "loss": 0.3834, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 5.597468354430379, | |
| "grad_norm": 0.2988437918726148, | |
| "learning_rate": 1.9839563612050273e-06, | |
| "loss": 0.4048, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 5.627848101265823, | |
| "grad_norm": 0.3142881945889849, | |
| "learning_rate": 1.8914956647091497e-06, | |
| "loss": 0.4044, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 5.658227848101266, | |
| "grad_norm": 0.2912800871138812, | |
| "learning_rate": 1.8010157632657544e-06, | |
| "loss": 0.3915, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 5.688607594936709, | |
| "grad_norm": 0.331383445209847, | |
| "learning_rate": 1.7125387598587862e-06, | |
| "loss": 0.3922, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 5.718987341772152, | |
| "grad_norm": 0.295340510609225, | |
| "learning_rate": 1.6260862681919965e-06, | |
| "loss": 0.3803, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 5.749367088607595, | |
| "grad_norm": 0.2861546159019479, | |
| "learning_rate": 1.5416794074090258e-06, | |
| "loss": 0.3846, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 5.779746835443038, | |
| "grad_norm": 0.3026296765652463, | |
| "learning_rate": 1.459338796934293e-06, | |
| "loss": 0.4177, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 5.810126582278481, | |
| "grad_norm": 0.2905750515452127, | |
| "learning_rate": 1.3790845514359363e-06, | |
| "loss": 0.4132, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 5.840506329113924, | |
| "grad_norm": 0.28908547248165223, | |
| "learning_rate": 1.300936275912098e-06, | |
| "loss": 0.3877, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 5.870886075949367, | |
| "grad_norm": 0.3120697859896662, | |
| "learning_rate": 1.224913060901688e-06, | |
| "loss": 0.3857, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 5.9012658227848105, | |
| "grad_norm": 0.2681050832298745, | |
| "learning_rate": 1.1510334778208332e-06, | |
| "loss": 0.3955, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 5.931645569620253, | |
| "grad_norm": 0.26796636762585163, | |
| "learning_rate": 1.0793155744261352e-06, | |
| "loss": 0.4222, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 5.962025316455696, | |
| "grad_norm": 0.28122294877877857, | |
| "learning_rate": 1.0097768704058542e-06, | |
| "loss": 0.3983, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 5.992405063291139, | |
| "grad_norm": 0.28084753163815557, | |
| "learning_rate": 9.424343531000968e-07, | |
| "loss": 0.3781, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 6.030379746835443, | |
| "grad_norm": 0.3035601235316243, | |
| "learning_rate": 8.773044733510338e-07, | |
| "loss": 0.3843, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 6.060759493670886, | |
| "grad_norm": 0.286177647003549, | |
| "learning_rate": 8.144031414842012e-07, | |
| "loss": 0.3553, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 6.091139240506329, | |
| "grad_norm": 0.27499257842697156, | |
| "learning_rate": 7.537457234218271e-07, | |
| "loss": 0.3878, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 6.121518987341772, | |
| "grad_norm": 0.287831290077759, | |
| "learning_rate": 6.953470369291349e-07, | |
| "loss": 0.3636, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 6.151898734177215, | |
| "grad_norm": 0.3270864734675839, | |
| "learning_rate": 6.392213479945852e-07, | |
| "loss": 0.3976, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 6.182278481012658, | |
| "grad_norm": 0.2850933541865749, | |
| "learning_rate": 5.853823673448877e-07, | |
| "loss": 0.4009, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 6.2126582278481015, | |
| "grad_norm": 0.29495009440796316, | |
| "learning_rate": 5.33843247095659e-07, | |
| "loss": 0.3956, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 6.243037974683545, | |
| "grad_norm": 0.26822212764266967, | |
| "learning_rate": 4.846165775385459e-07, | |
| "loss": 0.3896, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 6.273417721518987, | |
| "grad_norm": 0.26953773095218353, | |
| "learning_rate": 4.3771438406559173e-07, | |
| "loss": 0.3782, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 6.30379746835443, | |
| "grad_norm": 0.26810105671294315, | |
| "learning_rate": 3.931481242315993e-07, | |
| "loss": 0.3892, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 6.334177215189873, | |
| "grad_norm": 0.2950226684361474, | |
| "learning_rate": 3.5092868495520294e-07, | |
| "loss": 0.3807, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 6.364556962025317, | |
| "grad_norm": 0.2681133556168342, | |
| "learning_rate": 3.110663798593616e-07, | |
| "loss": 0.3988, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 6.39493670886076, | |
| "grad_norm": 0.2670123595381677, | |
| "learning_rate": 2.735709467518699e-07, | |
| "loss": 0.3737, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 6.425316455696202, | |
| "grad_norm": 0.2631719335348355, | |
| "learning_rate": 2.384515452465475e-07, | |
| "loss": 0.3674, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 6.455696202531645, | |
| "grad_norm": 0.2654377798436416, | |
| "learning_rate": 2.0571675452567997e-07, | |
| "loss": 0.3816, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 6.4860759493670885, | |
| "grad_norm": 0.2741124110494854, | |
| "learning_rate": 1.7537457124423896e-07, | |
| "loss": 0.3987, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 6.516455696202532, | |
| "grad_norm": 0.2558742486931433, | |
| "learning_rate": 1.474324075764111e-07, | |
| "loss": 0.3914, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 6.546835443037975, | |
| "grad_norm": 0.2626289255712206, | |
| "learning_rate": 1.2189708940490653e-07, | |
| "loss": 0.3719, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 6.577215189873417, | |
| "grad_norm": 0.2576213975257158, | |
| "learning_rate": 9.877485465349057e-08, | |
| "loss": 0.382, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 6.6075949367088604, | |
| "grad_norm": 0.25520510038953714, | |
| "learning_rate": 7.807135176314707e-08, | |
| "loss": 0.3755, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 6.637974683544304, | |
| "grad_norm": 0.26587236188019, | |
| "learning_rate": 5.979163831223988e-08, | |
| "loss": 0.4008, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 6.668354430379747, | |
| "grad_norm": 0.2643841861061346, | |
| "learning_rate": 4.394017978101905e-08, | |
| "loss": 0.3581, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 6.69873417721519, | |
| "grad_norm": 0.26560598833713933, | |
| "learning_rate": 3.0520848460765525e-08, | |
| "loss": 0.3832, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 6.729113924050633, | |
| "grad_norm": 0.2619146192545906, | |
| "learning_rate": 1.9536922507841227e-08, | |
| "loss": 0.3856, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 6.759493670886076, | |
| "grad_norm": 0.2779962742950057, | |
| "learning_rate": 1.099108514288627e-08, | |
| "loss": 0.3657, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 6.789873417721519, | |
| "grad_norm": 0.25682917386508564, | |
| "learning_rate": 4.885423995341088e-09, | |
| "loss": 0.3704, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 6.820253164556962, | |
| "grad_norm": 0.2684109452017263, | |
| "learning_rate": 1.2214305934699078e-09, | |
| "loss": 0.372, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 6.850632911392405, | |
| "grad_norm": 0.25032280703356524, | |
| "learning_rate": 0.0, | |
| "loss": 0.3908, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 6.850632911392405, | |
| "step": 224, | |
| "total_flos": 8.569304900394025e+17, | |
| "train_loss": 0.5346918581053615, | |
| "train_runtime": 27285.7344, | |
| "train_samples_per_second": 0.811, | |
| "train_steps_per_second": 0.008 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 224, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 7, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 8.569304900394025e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |