| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 9.965426072808624, | |
| "eval_steps": 500, | |
| "global_step": 49000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1016880211511084, | |
| "grad_norm": 8.324226379394531, | |
| "learning_rate": 4.830519964748153e-05, | |
| "loss": 2.3588, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.2033760423022168, | |
| "grad_norm": 8.197102546691895, | |
| "learning_rate": 4.661039929496305e-05, | |
| "loss": 2.1044, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.3050640634533252, | |
| "grad_norm": 7.696329593658447, | |
| "learning_rate": 4.4915598942444584e-05, | |
| "loss": 2.0132, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.4067520846044336, | |
| "grad_norm": 7.981374263763428, | |
| "learning_rate": 4.322079858992611e-05, | |
| "loss": 1.9548, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.508440105755542, | |
| "grad_norm": 8.823938369750977, | |
| "learning_rate": 4.152599823740764e-05, | |
| "loss": 1.899, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.6101281269066504, | |
| "grad_norm": 8.75110149383545, | |
| "learning_rate": 3.983119788488916e-05, | |
| "loss": 1.8601, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.7118161480577588, | |
| "grad_norm": 11.115782737731934, | |
| "learning_rate": 3.8136397532370685e-05, | |
| "loss": 1.8438, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.8135041692088671, | |
| "grad_norm": 8.88609790802002, | |
| "learning_rate": 3.644159717985222e-05, | |
| "loss": 1.8221, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.9151921903599756, | |
| "grad_norm": 9.193504333496094, | |
| "learning_rate": 3.4746796827333746e-05, | |
| "loss": 1.8052, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.016880211511084, | |
| "grad_norm": 9.435020446777344, | |
| "learning_rate": 3.3051996474815266e-05, | |
| "loss": 1.7687, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.1185682326621924, | |
| "grad_norm": 8.944482803344727, | |
| "learning_rate": 3.135719612229679e-05, | |
| "loss": 1.7001, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.2202562538133008, | |
| "grad_norm": 9.68911075592041, | |
| "learning_rate": 2.9662395769778324e-05, | |
| "loss": 1.7166, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.3219442749644092, | |
| "grad_norm": 8.70807933807373, | |
| "learning_rate": 2.796759541725985e-05, | |
| "loss": 1.6879, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.4236322961155177, | |
| "grad_norm": 6.220630645751953, | |
| "learning_rate": 2.6272795064741374e-05, | |
| "loss": 1.6574, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.525320317266626, | |
| "grad_norm": 10.505532264709473, | |
| "learning_rate": 2.45779947122229e-05, | |
| "loss": 1.6609, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.6270083384177343, | |
| "grad_norm": 6.303832054138184, | |
| "learning_rate": 2.2883194359704428e-05, | |
| "loss": 1.6455, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.728696359568843, | |
| "grad_norm": 8.265469551086426, | |
| "learning_rate": 2.1188394007185955e-05, | |
| "loss": 1.6164, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.8303843807199511, | |
| "grad_norm": 7.970987319946289, | |
| "learning_rate": 1.949359365466748e-05, | |
| "loss": 1.6029, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.9320724018710596, | |
| "grad_norm": 9.984902381896973, | |
| "learning_rate": 1.779879330214901e-05, | |
| "loss": 1.6018, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 2.033760423022168, | |
| "grad_norm": 7.69148588180542, | |
| "learning_rate": 1.6103992949630533e-05, | |
| "loss": 1.6092, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 2.135448444173276, | |
| "grad_norm": 6.372833251953125, | |
| "learning_rate": 1.4409192597112062e-05, | |
| "loss": 1.5372, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 2.237136465324385, | |
| "grad_norm": 7.225351333618164, | |
| "learning_rate": 1.2714392244593587e-05, | |
| "loss": 1.5452, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 2.338824486475493, | |
| "grad_norm": 8.18689250946045, | |
| "learning_rate": 1.1019591892075114e-05, | |
| "loss": 1.539, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 2.4405125076266017, | |
| "grad_norm": 7.346330165863037, | |
| "learning_rate": 9.324791539556641e-06, | |
| "loss": 1.5227, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 2.54220052877771, | |
| "grad_norm": 9.162985801696777, | |
| "learning_rate": 7.629991187038167e-06, | |
| "loss": 1.5207, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 2.6438885499288185, | |
| "grad_norm": 7.583765029907227, | |
| "learning_rate": 5.935190834519694e-06, | |
| "loss": 1.5172, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 2.7455765710799267, | |
| "grad_norm": 7.970169544219971, | |
| "learning_rate": 4.240390482001221e-06, | |
| "loss": 1.5272, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 2.8472645922310353, | |
| "grad_norm": 9.05431079864502, | |
| "learning_rate": 2.5455901294827473e-06, | |
| "loss": 1.5014, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 2.9489526133821435, | |
| "grad_norm": 8.572937965393066, | |
| "learning_rate": 8.507897769642736e-07, | |
| "loss": 1.4757, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 3.050640634533252, | |
| "grad_norm": 8.990861892700195, | |
| "learning_rate": 3.0933496034167175e-05, | |
| "loss": 1.5368, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 3.1523286556843604, | |
| "grad_norm": 8.719560623168945, | |
| "learning_rate": 3.0297945901972747e-05, | |
| "loss": 1.5287, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 3.2540166768354686, | |
| "grad_norm": 8.392940521240234, | |
| "learning_rate": 2.9662395769778324e-05, | |
| "loss": 1.5461, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 3.3557046979865772, | |
| "grad_norm": 8.443445205688477, | |
| "learning_rate": 2.9026845637583893e-05, | |
| "loss": 1.5664, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 3.4573927191376854, | |
| "grad_norm": 7.7863850593566895, | |
| "learning_rate": 2.8391295505389465e-05, | |
| "loss": 1.5471, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 3.559080740288794, | |
| "grad_norm": 9.05781364440918, | |
| "learning_rate": 2.775574537319504e-05, | |
| "loss": 1.5505, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 3.6607687614399023, | |
| "grad_norm": 6.635441303253174, | |
| "learning_rate": 2.712019524100061e-05, | |
| "loss": 1.518, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 3.762456782591011, | |
| "grad_norm": 8.673697471618652, | |
| "learning_rate": 2.6484645108806183e-05, | |
| "loss": 1.5018, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 3.864144803742119, | |
| "grad_norm": 9.876930236816406, | |
| "learning_rate": 2.5849094976611753e-05, | |
| "loss": 1.5348, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 3.9658328248932273, | |
| "grad_norm": 8.771023750305176, | |
| "learning_rate": 2.5213544844417332e-05, | |
| "loss": 1.4903, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 4.067520846044336, | |
| "grad_norm": 7.6693501472473145, | |
| "learning_rate": 2.45779947122229e-05, | |
| "loss": 1.4806, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 4.169208867195445, | |
| "grad_norm": 6.399499416351318, | |
| "learning_rate": 2.3942444580028474e-05, | |
| "loss": 1.4807, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 4.270896888346552, | |
| "grad_norm": 8.920206069946289, | |
| "learning_rate": 2.3306894447834047e-05, | |
| "loss": 1.4578, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 4.372584909497661, | |
| "grad_norm": 8.65735912322998, | |
| "learning_rate": 2.267134431563962e-05, | |
| "loss": 1.468, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 4.47427293064877, | |
| "grad_norm": 5.02122688293457, | |
| "learning_rate": 2.2035794183445192e-05, | |
| "loss": 1.4579, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 4.575960951799878, | |
| "grad_norm": 8.269110679626465, | |
| "learning_rate": 2.1400244051250764e-05, | |
| "loss": 1.4484, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 4.677648972950986, | |
| "grad_norm": 7.763547897338867, | |
| "learning_rate": 2.0764693919056337e-05, | |
| "loss": 1.4554, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 4.779336994102095, | |
| "grad_norm": 9.097722053527832, | |
| "learning_rate": 2.012914378686191e-05, | |
| "loss": 1.4531, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 4.881025015253203, | |
| "grad_norm": 8.873741149902344, | |
| "learning_rate": 1.949359365466748e-05, | |
| "loss": 1.4677, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 4.982713036404312, | |
| "grad_norm": 8.350312232971191, | |
| "learning_rate": 1.8858043522473055e-05, | |
| "loss": 1.4306, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 5.08440105755542, | |
| "grad_norm": 5.5741143226623535, | |
| "learning_rate": 1.8222493390278624e-05, | |
| "loss": 1.4203, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 5.186089078706528, | |
| "grad_norm": 9.378081321716309, | |
| "learning_rate": 1.75869432580842e-05, | |
| "loss": 1.4165, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 5.287777099857637, | |
| "grad_norm": 8.990020751953125, | |
| "learning_rate": 1.6951393125889773e-05, | |
| "loss": 1.3941, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 5.389465121008746, | |
| "grad_norm": 8.84300708770752, | |
| "learning_rate": 1.6315842993695342e-05, | |
| "loss": 1.3888, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 5.491153142159853, | |
| "grad_norm": 7.583876132965088, | |
| "learning_rate": 1.5680292861500918e-05, | |
| "loss": 1.3947, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 5.592841163310962, | |
| "grad_norm": 7.263258934020996, | |
| "learning_rate": 1.5044742729306487e-05, | |
| "loss": 1.3716, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 5.694529184462071, | |
| "grad_norm": 9.45223331451416, | |
| "learning_rate": 1.4409192597112062e-05, | |
| "loss": 1.3853, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 5.796217205613178, | |
| "grad_norm": 6.165616989135742, | |
| "learning_rate": 1.3773642464917633e-05, | |
| "loss": 1.3963, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 5.897905226764287, | |
| "grad_norm": 7.537690162658691, | |
| "learning_rate": 1.3138092332723207e-05, | |
| "loss": 1.3644, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 5.999593247915396, | |
| "grad_norm": 9.683405876159668, | |
| "learning_rate": 1.2502542200528778e-05, | |
| "loss": 1.3892, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 6.101281269066504, | |
| "grad_norm": 8.821761131286621, | |
| "learning_rate": 1.186699206833435e-05, | |
| "loss": 1.3335, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 6.202969290217612, | |
| "grad_norm": 12.081774711608887, | |
| "learning_rate": 1.1231441936139923e-05, | |
| "loss": 1.3526, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 6.304657311368721, | |
| "grad_norm": 6.325799465179443, | |
| "learning_rate": 1.0595891803945496e-05, | |
| "loss": 1.3424, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 6.406345332519829, | |
| "grad_norm": 9.028151512145996, | |
| "learning_rate": 9.960341671751069e-06, | |
| "loss": 1.3693, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 6.508033353670937, | |
| "grad_norm": 6.508680820465088, | |
| "learning_rate": 9.324791539556641e-06, | |
| "loss": 1.3506, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 6.609721374822046, | |
| "grad_norm": 6.9301629066467285, | |
| "learning_rate": 8.689241407362212e-06, | |
| "loss": 1.3426, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 6.7114093959731544, | |
| "grad_norm": 8.2422456741333, | |
| "learning_rate": 8.053691275167785e-06, | |
| "loss": 1.3337, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 6.813097417124263, | |
| "grad_norm": 7.615689277648926, | |
| "learning_rate": 7.4181411429733575e-06, | |
| "loss": 1.3322, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 6.914785438275371, | |
| "grad_norm": 7.60098123550415, | |
| "learning_rate": 6.782591010778931e-06, | |
| "loss": 1.3182, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 7.0164734594264795, | |
| "grad_norm": 8.578591346740723, | |
| "learning_rate": 6.147040878584503e-06, | |
| "loss": 1.3333, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 7.118161480577588, | |
| "grad_norm": 8.674110412597656, | |
| "learning_rate": 5.511490746390076e-06, | |
| "loss": 1.2914, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 7.219849501728697, | |
| "grad_norm": 7.355325222015381, | |
| "learning_rate": 4.875940614195648e-06, | |
| "loss": 1.3241, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 7.3215375228798045, | |
| "grad_norm": 8.176190376281738, | |
| "learning_rate": 4.240390482001221e-06, | |
| "loss": 1.3184, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 7.423225544030913, | |
| "grad_norm": 8.22246265411377, | |
| "learning_rate": 3.604840349806793e-06, | |
| "loss": 1.3011, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 7.524913565182022, | |
| "grad_norm": 9.817005157470703, | |
| "learning_rate": 2.9692902176123655e-06, | |
| "loss": 1.3044, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 7.6266015863331305, | |
| "grad_norm": 8.255529403686523, | |
| "learning_rate": 2.3337400854179377e-06, | |
| "loss": 1.2964, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 7.728289607484238, | |
| "grad_norm": 6.6958160400390625, | |
| "learning_rate": 1.6981899532235102e-06, | |
| "loss": 1.3092, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 7.829977628635347, | |
| "grad_norm": 8.960187911987305, | |
| "learning_rate": 1.062639821029083e-06, | |
| "loss": 1.2976, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 7.9316656497864555, | |
| "grad_norm": 7.582452297210693, | |
| "learning_rate": 4.2708968883465534e-07, | |
| "loss": 1.2956, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 8.033353670937563, | |
| "grad_norm": 7.20064115524292, | |
| "learning_rate": 9.833231645312182e-06, | |
| "loss": 1.3157, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 8.135041692088672, | |
| "grad_norm": 7.36190128326416, | |
| "learning_rate": 9.324791539556641e-06, | |
| "loss": 1.3021, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 8.23672971323978, | |
| "grad_norm": 9.36536979675293, | |
| "learning_rate": 8.816351433801099e-06, | |
| "loss": 1.3073, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 8.33841773439089, | |
| "grad_norm": 7.667935848236084, | |
| "learning_rate": 8.307911328045556e-06, | |
| "loss": 1.2902, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 8.440105755541998, | |
| "grad_norm": 8.367385864257812, | |
| "learning_rate": 7.799471222290014e-06, | |
| "loss": 1.3164, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 8.541793776693105, | |
| "grad_norm": 8.879359245300293, | |
| "learning_rate": 7.291031116534472e-06, | |
| "loss": 1.2927, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 8.643481797844213, | |
| "grad_norm": 7.432185649871826, | |
| "learning_rate": 6.782591010778931e-06, | |
| "loss": 1.3027, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 8.745169818995322, | |
| "grad_norm": 7.75663423538208, | |
| "learning_rate": 6.274150905023389e-06, | |
| "loss": 1.3266, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 8.84685784014643, | |
| "grad_norm": 10.648377418518066, | |
| "learning_rate": 5.765710799267846e-06, | |
| "loss": 1.263, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 8.94854586129754, | |
| "grad_norm": 8.594388961791992, | |
| "learning_rate": 5.257270693512305e-06, | |
| "loss": 1.3049, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 9.050233882448648, | |
| "grad_norm": 8.348444938659668, | |
| "learning_rate": 4.748830587756762e-06, | |
| "loss": 1.2984, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 9.151921903599757, | |
| "grad_norm": 9.689435958862305, | |
| "learning_rate": 4.240390482001221e-06, | |
| "loss": 1.2609, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 9.253609924750865, | |
| "grad_norm": 7.50744104385376, | |
| "learning_rate": 3.7319503762456785e-06, | |
| "loss": 1.2705, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 9.355297945901972, | |
| "grad_norm": 7.884949684143066, | |
| "learning_rate": 3.2235102704901364e-06, | |
| "loss": 1.2949, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 9.45698596705308, | |
| "grad_norm": 7.593263626098633, | |
| "learning_rate": 2.7150701647345947e-06, | |
| "loss": 1.2635, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 9.55867398820419, | |
| "grad_norm": 6.893133640289307, | |
| "learning_rate": 2.2066300589790525e-06, | |
| "loss": 1.2811, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 9.660362009355298, | |
| "grad_norm": 9.131689071655273, | |
| "learning_rate": 1.6981899532235102e-06, | |
| "loss": 1.2854, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 9.762050030506407, | |
| "grad_norm": 6.920133113861084, | |
| "learning_rate": 1.1897498474679682e-06, | |
| "loss": 1.306, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 9.863738051657515, | |
| "grad_norm": 7.8755950927734375, | |
| "learning_rate": 6.813097417124263e-07, | |
| "loss": 1.2769, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 9.965426072808624, | |
| "grad_norm": 9.283025741577148, | |
| "learning_rate": 1.728696359568843e-07, | |
| "loss": 1.2504, | |
| "step": 49000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 49170, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.57905183894272e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |