| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 325, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 5.738582378602418, | |
| "learning_rate": 6.060606060606061e-06, | |
| "loss": 2.1213, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.065556365217285, | |
| "learning_rate": 3.0303030303030306e-05, | |
| "loss": 2.0078, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 6.107029633427899, | |
| "learning_rate": 6.060606060606061e-05, | |
| "loss": 1.438, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.3973078608934906, | |
| "learning_rate": 9.090909090909092e-05, | |
| "loss": 0.172, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.049620937042406804, | |
| "learning_rate": 0.00012121212121212122, | |
| "loss": 0.003, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.008073111308502804, | |
| "learning_rate": 0.00015151515151515152, | |
| "loss": 0.0002, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.0061749583809672815, | |
| "learning_rate": 0.00018181818181818183, | |
| "loss": 0.0, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.015228286957835636, | |
| "learning_rate": 0.00019997685019798912, | |
| "loss": 0.0, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.9453136589525021, | |
| "learning_rate": 0.0001997165380022878, | |
| "loss": 0.0001, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.0010960399423771447, | |
| "learning_rate": 0.000199167731989929, | |
| "loss": 0.0, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.0006677902263119037, | |
| "learning_rate": 0.0001983320199330545, | |
| "loss": 0.0, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.0007278805069068889, | |
| "learning_rate": 0.00019721181966290613, | |
| "loss": 0.0, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.0004460552528072769, | |
| "learning_rate": 0.00019581037207470382, | |
| "loss": 0.0, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.00037234746610264403, | |
| "learning_rate": 0.00019413173175128473, | |
| "loss": 0.0, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.0003711556761775508, | |
| "learning_rate": 0.00019218075523263104, | |
| "loss": 0.0, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.0009620782234129658, | |
| "learning_rate": 0.00018996308696522433, | |
| "loss": 0.0298, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.0206702984018057, | |
| "learning_rate": 0.00018748514297187648, | |
| "loss": 0.0001, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.012769998574125136, | |
| "learning_rate": 0.00018475409228928312, | |
| "loss": 0.0001, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.00507995837089818, | |
| "learning_rate": 0.00018177783622700327, | |
| "loss": 0.0001, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.002287434094773598, | |
| "learning_rate": 0.00017856498550787144, | |
| "loss": 0.0, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.0016095121637470665, | |
| "learning_rate": 0.00017512483535597867, | |
| "loss": 0.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.0009581916701484739, | |
| "learning_rate": 0.00017146733860429612, | |
| "loss": 0.0, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.0007229980054986912, | |
| "learning_rate": 0.0001676030768997445, | |
| "loss": 0.0, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.0006396007111761404, | |
| "learning_rate": 0.00016354323008901776, | |
| "loss": 0.0, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.0004841518323758468, | |
| "learning_rate": 0.00015929954387373103, | |
| "loss": 0.0, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.0008910415688920396, | |
| "learning_rate": 0.00015488429582847192, | |
| "loss": 0.0205, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.00380673544543486, | |
| "learning_rate": 0.00015031025988006936, | |
| "loss": 0.0, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.004762201694169973, | |
| "learning_rate": 0.00014559066935084588, | |
| "loss": 0.0, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.001962425008336919, | |
| "learning_rate": 0.00014073917867277557, | |
| "loss": 0.0, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.0010385610431445224, | |
| "learning_rate": 0.0001357698238833126, | |
| "loss": 0.0, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.0007723000908395725, | |
| "learning_rate": 0.000130696982017182, | |
| "loss": 0.0, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.000566857138032962, | |
| "learning_rate": 0.0001255353295116187, | |
| "loss": 0.0, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.0005082431773701508, | |
| "learning_rate": 0.00012029979974539234, | |
| "loss": 0.0, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.0004675344623799639, | |
| "learning_rate": 0.00011500553983446527, | |
| "loss": 0.0, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.00036040960856775176, | |
| "learning_rate": 0.00010966786680927874, | |
| "loss": 0.0, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.0003729641796295395, | |
| "learning_rate": 0.00010430222330045304, | |
| "loss": 0.0, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.0003443172826396222, | |
| "learning_rate": 9.892413286110886e-05, | |
| "loss": 0.018, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.0005495491836887999, | |
| "learning_rate": 9.354915505506839e-05, | |
| "loss": 0.0, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.0008803367690701702, | |
| "learning_rate": 8.81928404408726e-05, | |
| "loss": 0.0, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.0011313961073884484, | |
| "learning_rate": 8.287068558185225e-05, | |
| "loss": 0.0, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.0057124676419671805, | |
| "learning_rate": 7.759808821241406e-05, | |
| "loss": 0.0079, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.01204973063281988, | |
| "learning_rate": 7.239030269025311e-05, | |
| "loss": 0.0, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.003624967374602045, | |
| "learning_rate": 6.726239586337408e-05, | |
| "loss": 0.0, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.0012067744182645878, | |
| "learning_rate": 6.22292034796035e-05, | |
| "loss": 0.0, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.0006712338568149837, | |
| "learning_rate": 5.730528726470792e-05, | |
| "loss": 0.0, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.00053984081321036, | |
| "learning_rate": 5.2504892793295e-05, | |
| "loss": 0.0, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.0004979282628494034, | |
| "learning_rate": 4.7841908274384616e-05, | |
| "loss": 0.0, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.000341740588014389, | |
| "learning_rate": 4.332982437088825e-05, | |
| "loss": 0.0, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.0003347620763231193, | |
| "learning_rate": 3.898169516924398e-05, | |
| "loss": 0.0, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.0003186862924366449, | |
| "learning_rate": 3.4810100412128747e-05, | |
| "loss": 0.0, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.00033256441192096515, | |
| "learning_rate": 3.0827109103512643e-05, | |
| "loss": 0.0, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.00029116215296989, | |
| "learning_rate": 2.7044244591351232e-05, | |
| "loss": 0.0, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.00029813168642378195, | |
| "learning_rate": 2.3472451228937253e-05, | |
| "loss": 0.0, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.0002542728444587161, | |
| "learning_rate": 2.0122062711363532e-05, | |
| "loss": 0.0, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.00030107064227284755, | |
| "learning_rate": 1.7002772178705716e-05, | |
| "loss": 0.0, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.00028827249953127903, | |
| "learning_rate": 1.4123604172419713e-05, | |
| "loss": 0.0, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.00028302069696532913, | |
| "learning_rate": 1.149288852608743e-05, | |
| "loss": 0.0, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.00028247863690175195, | |
| "learning_rate": 9.118236266049707e-06, | |
| "loss": 0.0, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.000255661505975097, | |
| "learning_rate": 7.0065175916482095e-06, | |
| "loss": 0.0, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.0002702855893991498, | |
| "learning_rate": 5.163841998782837e-06, | |
| "loss": 0.0, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.0002735846817383222, | |
| "learning_rate": 3.595540604290437e-06, | |
| "loss": 0.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.0002468116893628423, | |
| "learning_rate": 2.30615072228183e-06, | |
| "loss": 0.0, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.0002618777070782965, | |
| "learning_rate": 1.2994027370611173e-06, | |
| "loss": 0.0, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.0002639622475517554, | |
| "learning_rate": 5.782093106048159e-07, | |
| "loss": 0.0, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.0002841165380426929, | |
| "learning_rate": 1.446569558255395e-07, | |
| "loss": 0.0, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.0002936586206857914, | |
| "learning_rate": 0.0, | |
| "loss": 0.0, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": NaN, | |
| "eval_runtime": 744.3655, | |
| "eval_samples_per_second": 1.553, | |
| "eval_steps_per_second": 0.388, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 325, | |
| "total_flos": 1047256696487936.0, | |
| "train_loss": 0.05724426985613648, | |
| "train_runtime": 6023.5199, | |
| "train_samples_per_second": 1.725, | |
| "train_steps_per_second": 0.054 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 325, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 40, | |
| "total_flos": 1047256696487936.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |