| { | |
| "best_metric": 5.173658792045899e-05, | |
| "best_model_checkpoint": "res/checkpoint-400", | |
| "epoch": 0.5633802816901409, | |
| "eval_steps": 200, | |
| "global_step": 400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 3.3899288177490234, | |
| "learning_rate": 5e-06, | |
| "loss": 0.6908, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.9440199136734009, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6992, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.45e-05, | |
| "loss": 0.6318, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 4.202402591705322, | |
| "learning_rate": 1.9e-05, | |
| "loss": 0.4382, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 21.559049606323242, | |
| "learning_rate": 2.4e-05, | |
| "loss": 0.0718, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.11590419709682465, | |
| "learning_rate": 2.9e-05, | |
| "loss": 0.0055, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.02931913733482361, | |
| "learning_rate": 3.35e-05, | |
| "loss": 0.0175, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 31.5627498626709, | |
| "learning_rate": 3.85e-05, | |
| "loss": 0.159, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.10093926638364792, | |
| "learning_rate": 4.35e-05, | |
| "loss": 0.0993, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.02363261766731739, | |
| "learning_rate": 4.85e-05, | |
| "loss": 0.0013, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.018094830214977264, | |
| "learning_rate": 4.9426229508196726e-05, | |
| "loss": 0.0008, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.010088359005749226, | |
| "learning_rate": 4.860655737704918e-05, | |
| "loss": 0.0005, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.00800824724137783, | |
| "learning_rate": 4.778688524590164e-05, | |
| "loss": 0.0004, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.03682546317577362, | |
| "learning_rate": 4.704918032786885e-05, | |
| "loss": 0.0664, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.006090231705456972, | |
| "learning_rate": 4.622950819672132e-05, | |
| "loss": 0.0002, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.00360495550557971, | |
| "learning_rate": 4.540983606557377e-05, | |
| "loss": 0.0002, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.004050845745950937, | |
| "learning_rate": 4.459016393442623e-05, | |
| "loss": 0.0002, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.0043782140128314495, | |
| "learning_rate": 4.377049180327869e-05, | |
| "loss": 0.0002, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.004006043076515198, | |
| "learning_rate": 4.295081967213115e-05, | |
| "loss": 0.0001, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.003426899667829275, | |
| "learning_rate": 4.213114754098361e-05, | |
| "loss": 0.1068, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "eval_accuracy": 0.9987674609695973, | |
| "eval_f1": 0.9987956643918106, | |
| "eval_loss": 0.002453433582559228, | |
| "eval_precision": 0.9983948635634029, | |
| "eval_recall": 0.9991967871485944, | |
| "eval_runtime": 335.4821, | |
| "eval_samples_per_second": 7.255, | |
| "eval_steps_per_second": 0.909, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.015531109645962715, | |
| "learning_rate": 4.131147540983607e-05, | |
| "loss": 0.062, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.02330106310546398, | |
| "learning_rate": 4.049180327868853e-05, | |
| "loss": 0.0089, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.007218244019895792, | |
| "learning_rate": 3.9672131147540983e-05, | |
| "loss": 0.0004, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.008732643909752369, | |
| "learning_rate": 3.8852459016393444e-05, | |
| "loss": 0.0003, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.007611180655658245, | |
| "learning_rate": 3.8032786885245905e-05, | |
| "loss": 0.0003, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.004713522270321846, | |
| "learning_rate": 3.721311475409836e-05, | |
| "loss": 0.0002, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.005065000616014004, | |
| "learning_rate": 3.6393442622950826e-05, | |
| "loss": 0.0002, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.0033236700110137463, | |
| "learning_rate": 3.557377049180328e-05, | |
| "loss": 0.0002, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.003254691371694207, | |
| "learning_rate": 3.475409836065574e-05, | |
| "loss": 0.0001, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.0026583385188132524, | |
| "learning_rate": 3.39344262295082e-05, | |
| "loss": 0.0001, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.0026992058847099543, | |
| "learning_rate": 3.3114754098360655e-05, | |
| "loss": 0.0001, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.002891330514103174, | |
| "learning_rate": 3.2295081967213116e-05, | |
| "loss": 0.0001, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.0022049755789339542, | |
| "learning_rate": 3.1475409836065576e-05, | |
| "loss": 0.0001, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.002381928265094757, | |
| "learning_rate": 3.065573770491804e-05, | |
| "loss": 0.0001, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.0018912949599325657, | |
| "learning_rate": 2.9836065573770494e-05, | |
| "loss": 0.0001, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.00272244680672884, | |
| "learning_rate": 2.901639344262295e-05, | |
| "loss": 0.0001, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.0019356077536940575, | |
| "learning_rate": 2.819672131147541e-05, | |
| "loss": 0.0001, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.002229230711236596, | |
| "learning_rate": 2.737704918032787e-05, | |
| "loss": 0.0001, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.002221511909738183, | |
| "learning_rate": 2.6557377049180327e-05, | |
| "loss": 0.0001, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.0020540908444672823, | |
| "learning_rate": 2.573770491803279e-05, | |
| "loss": 0.0001, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "eval_accuracy": 1.0, | |
| "eval_f1": 1.0, | |
| "eval_loss": 5.173658792045899e-05, | |
| "eval_precision": 1.0, | |
| "eval_recall": 1.0, | |
| "eval_runtime": 333.7573, | |
| "eval_samples_per_second": 7.293, | |
| "eval_steps_per_second": 0.914, | |
| "step": 400 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 710, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 400, | |
| "total_flos": 4203850314547200.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |