| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 375, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.7801982164382935, | |
| "learning_rate": 5.263157894736842e-06, | |
| "loss": 0.5021, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.8954057693481445, | |
| "learning_rate": 1.0526315789473684e-05, | |
| "loss": 0.4477, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.959135890007019, | |
| "learning_rate": 1.578947368421053e-05, | |
| "loss": 0.4523, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.6955620646476746, | |
| "learning_rate": 1.9998261969639324e-05, | |
| "loss": 0.453, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.48524388670921326, | |
| "learning_rate": 1.9937494319239112e-05, | |
| "loss": 0.4317, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.5131641030311584, | |
| "learning_rate": 1.979042835741503e-05, | |
| "loss": 0.4387, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.6963252425193787, | |
| "learning_rate": 1.9558341221417744e-05, | |
| "loss": 0.4648, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.5048828125, | |
| "learning_rate": 1.9243248381877605e-05, | |
| "loss": 0.4563, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.5360320210456848, | |
| "learning_rate": 1.8847886140232438e-05, | |
| "loss": 0.4356, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.5510643124580383, | |
| "learning_rate": 1.8375687866379988e-05, | |
| "loss": 0.4599, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.4948376715183258, | |
| "learning_rate": 1.7830754182909985e-05, | |
| "loss": 0.4438, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.43106070160865784, | |
| "learning_rate": 1.721781735483921e-05, | |
| "loss": 0.4399, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 0.44680503010749817, | |
| "learning_rate": 1.654220019409317e-05, | |
| "loss": 0.3948, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 0.8041722178459167, | |
| "learning_rate": 1.580976983561235e-05, | |
| "loss": 0.3355, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.4531875252723694, | |
| "learning_rate": 1.5026886786496624e-05, | |
| "loss": 0.3182, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.3980570137500763, | |
| "learning_rate": 1.4200349690650654e-05, | |
| "loss": 0.3274, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.3599999999999999, | |
| "grad_norm": 0.44365376234054565, | |
| "learning_rate": 1.3337336288600297e-05, | |
| "loss": 0.3524, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.3755331039428711, | |
| "learning_rate": 1.24453410851916e-05, | |
| "loss": 0.3114, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 0.4399131238460541, | |
| "learning_rate": 1.1532110266473026e-05, | |
| "loss": 0.3151, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.4257424473762512, | |
| "learning_rate": 1.0605574430949983e-05, | |
| "loss": 0.3391, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.6800000000000002, | |
| "grad_norm": 0.5761308073997498, | |
| "learning_rate": 9.673779719380967e-06, | |
| "loss": 0.3313, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 0.4401087462902069, | |
| "learning_rate": 8.744817941191862e-06, | |
| "loss": 0.3081, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.8399999999999999, | |
| "grad_norm": 0.37950068712234497, | |
| "learning_rate": 7.826756304298428e-06, | |
| "loss": 0.3324, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.4223324954509735, | |
| "learning_rate": 6.92756735857107e-06, | |
| "loss": 0.3175, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.4112967848777771, | |
| "learning_rate": 6.0550597613206205e-06, | |
| "loss": 0.3204, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 0.5121579766273499, | |
| "learning_rate": 5.216810466045448e-06, | |
| "loss": 0.2403, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 0.5039759874343872, | |
| "learning_rate": 4.420098923320378e-06, | |
| "loss": 0.2426, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.44253629446029663, | |
| "learning_rate": 3.671843865234238e-06, | |
| "loss": 0.2264, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 0.35495468974113464, | |
| "learning_rate": 2.978543222347076e-06, | |
| "loss": 0.2603, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.38393262028694153, | |
| "learning_rate": 2.346217694934847e-06, | |
| "loss": 0.2436, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 0.4627029299736023, | |
| "learning_rate": 1.7803584685552877e-06, | |
| "loss": 0.2437, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.40824005007743835, | |
| "learning_rate": 1.2858795279787517e-06, | |
| "loss": 0.2365, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 0.3864591419696808, | |
| "learning_rate": 8.670749835951964e-07, | |
| "loss": 0.246, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.7199999999999998, | |
| "grad_norm": 0.4451102316379547, | |
| "learning_rate": 5.275817808796013e-07, | |
| "loss": 0.2507, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 0.41204768419265747, | |
| "learning_rate": 2.703481167509281e-07, | |
| "loss": 0.225, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.43420955538749695, | |
| "learning_rate": 9.760783710056176e-08, | |
| "loss": 0.2448, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 0.40928831696510315, | |
| "learning_rate": 1.0861037824896337e-08, | |
| "loss": 0.2144, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 375, | |
| "total_flos": 74595673178112.0, | |
| "train_loss": 0.33862359380722046, | |
| "train_runtime": 6722.8271, | |
| "train_samples_per_second": 0.446, | |
| "train_steps_per_second": 0.056 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 375, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "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": 74595673178112.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |