| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 30, | |
| "global_step": 450, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.022222222222222223, | |
| "grad_norm": 0.8203227519989014, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.4543, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.044444444444444446, | |
| "grad_norm": 0.09747839719057083, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 0.0143, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06666666666666667, | |
| "grad_norm": 0.008741447702050209, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.0042, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.06666666666666667, | |
| "eval_loss": 0.004266222473233938, | |
| "eval_runtime": 117.2469, | |
| "eval_samples_per_second": 17.058, | |
| "eval_steps_per_second": 0.426, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.08888888888888889, | |
| "grad_norm": 0.014671124517917633, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 0.0042, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1111111111111111, | |
| "grad_norm": 0.008508998900651932, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.004, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.13333333333333333, | |
| "grad_norm": 0.009779859334230423, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 0.0039, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13333333333333333, | |
| "eval_loss": 0.0038896214682608843, | |
| "eval_runtime": 117.3001, | |
| "eval_samples_per_second": 17.05, | |
| "eval_steps_per_second": 0.426, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.15555555555555556, | |
| "grad_norm": 0.007619491778314114, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.0039, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.17777777777777778, | |
| "grad_norm": 0.006427408196032047, | |
| "learning_rate": 9.24024048078213e-05, | |
| "loss": 0.0039, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.006157474592328072, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.0039, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "eval_loss": 0.0038220968563109636, | |
| "eval_runtime": 117.37, | |
| "eval_samples_per_second": 17.04, | |
| "eval_steps_per_second": 0.426, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 0.03153691068291664, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 0.0039, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.24444444444444444, | |
| "grad_norm": 0.0021719736978411674, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 0.0038, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.26666666666666666, | |
| "grad_norm": 0.013368754647672176, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 0.0038, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.26666666666666666, | |
| "eval_loss": 0.0037406720221042633, | |
| "eval_runtime": 199.847, | |
| "eval_samples_per_second": 10.008, | |
| "eval_steps_per_second": 0.25, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.28888888888888886, | |
| "grad_norm": 0.014529922977089882, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 0.0037, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.3111111111111111, | |
| "grad_norm": 0.0757686048746109, | |
| "learning_rate": 7.795964517353735e-05, | |
| "loss": 0.0037, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 0.01011459156870842, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.0038, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "eval_loss": 0.0037290642503648996, | |
| "eval_runtime": 117.2836, | |
| "eval_samples_per_second": 17.053, | |
| "eval_steps_per_second": 0.426, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.35555555555555557, | |
| "grad_norm": 0.007393770385533571, | |
| "learning_rate": 7.191855733945387e-05, | |
| "loss": 0.0038, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.37777777777777777, | |
| "grad_norm": 0.007255127187818289, | |
| "learning_rate": 6.873032967079561e-05, | |
| "loss": 0.0037, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.029892487451434135, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 0.0038, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "eval_loss": 0.0037513442803174257, | |
| "eval_runtime": 117.2984, | |
| "eval_samples_per_second": 17.051, | |
| "eval_steps_per_second": 0.426, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.4222222222222222, | |
| "grad_norm": 0.013895494863390923, | |
| "learning_rate": 6.209609477998338e-05, | |
| "loss": 0.0037, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 0.009924459271132946, | |
| "learning_rate": 5.868240888334653e-05, | |
| "loss": 0.0036, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4666666666666667, | |
| "grad_norm": 0.01607881858944893, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 0.0037, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.4666666666666667, | |
| "eval_loss": 0.003465294372290373, | |
| "eval_runtime": 142.6159, | |
| "eval_samples_per_second": 14.024, | |
| "eval_steps_per_second": 0.351, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.4888888888888889, | |
| "grad_norm": 0.05860644206404686, | |
| "learning_rate": 5.174497483512506e-05, | |
| "loss": 0.0035, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.5111111111111111, | |
| "grad_norm": 0.07811647653579712, | |
| "learning_rate": 4.825502516487497e-05, | |
| "loss": 0.0036, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5333333333333333, | |
| "grad_norm": 0.019251324236392975, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 0.0032, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5333333333333333, | |
| "eval_loss": 0.0030640396289527416, | |
| "eval_runtime": 189.2243, | |
| "eval_samples_per_second": 10.569, | |
| "eval_steps_per_second": 0.264, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5555555555555556, | |
| "grad_norm": 0.019416367635130882, | |
| "learning_rate": 4.131759111665349e-05, | |
| "loss": 0.003, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5777777777777777, | |
| "grad_norm": 0.028752854093909264, | |
| "learning_rate": 3.790390522001662e-05, | |
| "loss": 0.0027, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.014850922860205173, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 0.0028, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "eval_loss": 0.002417274285107851, | |
| "eval_runtime": 117.247, | |
| "eval_samples_per_second": 17.058, | |
| "eval_steps_per_second": 0.426, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.6222222222222222, | |
| "grad_norm": 0.01956513710319996, | |
| "learning_rate": 3.12696703292044e-05, | |
| "loss": 0.0026, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6444444444444445, | |
| "grad_norm": 0.010545406490564346, | |
| "learning_rate": 2.8081442660546125e-05, | |
| "loss": 0.0026, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.05216179043054581, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 0.0025, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "eval_loss": 0.002407504478469491, | |
| "eval_runtime": 117.3038, | |
| "eval_samples_per_second": 17.05, | |
| "eval_steps_per_second": 0.426, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6888888888888889, | |
| "grad_norm": 0.017784912139177322, | |
| "learning_rate": 2.2040354826462668e-05, | |
| "loss": 0.0027, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.7111111111111111, | |
| "grad_norm": 0.02129705250263214, | |
| "learning_rate": 1.9216926233717085e-05, | |
| "loss": 0.0025, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.7333333333333333, | |
| "grad_norm": 0.011542588472366333, | |
| "learning_rate": 1.6543469682057106e-05, | |
| "loss": 0.0025, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7333333333333333, | |
| "eval_loss": 0.0024114707484841347, | |
| "eval_runtime": 211.1966, | |
| "eval_samples_per_second": 9.47, | |
| "eval_steps_per_second": 0.237, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7555555555555555, | |
| "grad_norm": 0.027737773954868317, | |
| "learning_rate": 1.4033009983067452e-05, | |
| "loss": 0.0024, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7777777777777778, | |
| "grad_norm": 0.02074122428894043, | |
| "learning_rate": 1.1697777844051105e-05, | |
| "loss": 0.0025, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.01723620295524597, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 0.0022, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "eval_loss": 0.0022669117897748947, | |
| "eval_runtime": 117.3791, | |
| "eval_samples_per_second": 17.039, | |
| "eval_steps_per_second": 0.426, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.8222222222222222, | |
| "grad_norm": 0.024718888103961945, | |
| "learning_rate": 7.597595192178702e-06, | |
| "loss": 0.0023, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.8444444444444444, | |
| "grad_norm": 0.010936993174254894, | |
| "learning_rate": 5.852620357053651e-06, | |
| "loss": 0.0023, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8666666666666667, | |
| "grad_norm": 0.0511295422911644, | |
| "learning_rate": 4.322727117869951e-06, | |
| "loss": 0.0023, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8666666666666667, | |
| "eval_loss": 0.002225763164460659, | |
| "eval_runtime": 117.4249, | |
| "eval_samples_per_second": 17.032, | |
| "eval_steps_per_second": 0.426, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 0.032784491777420044, | |
| "learning_rate": 3.0153689607045845e-06, | |
| "loss": 0.0021, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.9111111111111111, | |
| "grad_norm": 0.045701805502176285, | |
| "learning_rate": 1.9369152030840556e-06, | |
| "loss": 0.0022, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.9333333333333333, | |
| "grad_norm": 0.0103871775791049, | |
| "learning_rate": 1.0926199633097157e-06, | |
| "loss": 0.0022, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9333333333333333, | |
| "eval_loss": 0.0022246255539357662, | |
| "eval_runtime": 117.3984, | |
| "eval_samples_per_second": 17.036, | |
| "eval_steps_per_second": 0.426, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9555555555555556, | |
| "grad_norm": 0.03928959742188454, | |
| "learning_rate": 4.865965629214819e-07, | |
| "loss": 0.0024, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9777777777777777, | |
| "grad_norm": 0.03138284012675285, | |
| "learning_rate": 1.2179748700879012e-07, | |
| "loss": 0.0024, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.019241375848650932, | |
| "learning_rate": 0.0, | |
| "loss": 0.0021, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.0022388531360775232, | |
| "eval_runtime": 117.3729, | |
| "eval_samples_per_second": 17.04, | |
| "eval_steps_per_second": 0.426, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 450, | |
| "total_flos": 1.663912988085584e+18, | |
| "train_loss": 0.013403736940688558, | |
| "train_runtime": 5942.7183, | |
| "train_samples_per_second": 3.029, | |
| "train_steps_per_second": 0.076 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 450, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "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": 1.663912988085584e+18, | |
| "train_batch_size": 10, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |