| { | |
| "best_metric": 0.04605380445718765, | |
| "best_model_checkpoint": "/data/wheld3/byt5-base-cstop_artificial/checkpoint-200", | |
| "epoch": 428.5019607843137, | |
| "global_step": 3000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 28.5, | |
| "learning_rate": 0.0009333333333333333, | |
| "loss": 0.2563, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 28.5, | |
| "eval_exact_match": 0.03756708407871199, | |
| "eval_loss": 0.04605380445718765, | |
| "eval_runtime": 27.8305, | |
| "eval_samples_per_second": 20.086, | |
| "eval_steps_per_second": 2.515, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 57.13, | |
| "learning_rate": 0.0008666666666666667, | |
| "loss": 0.0065, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 57.13, | |
| "eval_exact_match": 0.03756708407871199, | |
| "eval_loss": 0.05634024366736412, | |
| "eval_runtime": 28.5438, | |
| "eval_samples_per_second": 19.584, | |
| "eval_steps_per_second": 2.452, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 85.63, | |
| "learning_rate": 0.0008, | |
| "loss": 0.0021, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 85.63, | |
| "eval_exact_match": 0.03577817531305903, | |
| "eval_loss": 0.05924277380108833, | |
| "eval_runtime": 28.509, | |
| "eval_samples_per_second": 19.608, | |
| "eval_steps_per_second": 2.455, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 114.25, | |
| "learning_rate": 0.0007333333333333333, | |
| "loss": 0.0013, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 114.25, | |
| "eval_exact_match": 0.03756708407871199, | |
| "eval_loss": 0.05689763277769089, | |
| "eval_runtime": 28.1931, | |
| "eval_samples_per_second": 19.828, | |
| "eval_steps_per_second": 2.483, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 142.75, | |
| "learning_rate": 0.0006666666666666666, | |
| "loss": 0.0008, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 142.75, | |
| "eval_exact_match": 0.03577817531305903, | |
| "eval_loss": 0.06747602671384811, | |
| "eval_runtime": 28.4722, | |
| "eval_samples_per_second": 19.633, | |
| "eval_steps_per_second": 2.459, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 171.38, | |
| "learning_rate": 0.0006, | |
| "loss": 0.0007, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 171.38, | |
| "eval_exact_match": 0.03935599284436494, | |
| "eval_loss": 0.06273317337036133, | |
| "eval_runtime": 28.026, | |
| "eval_samples_per_second": 19.946, | |
| "eval_steps_per_second": 2.498, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 199.88, | |
| "learning_rate": 0.0005333333333333334, | |
| "loss": 0.0004, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 199.88, | |
| "eval_exact_match": 0.03577817531305903, | |
| "eval_loss": 0.06774432212114334, | |
| "eval_runtime": 29.2197, | |
| "eval_samples_per_second": 19.131, | |
| "eval_steps_per_second": 2.396, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 228.5, | |
| "learning_rate": 0.00046666666666666666, | |
| "loss": 0.0003, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 228.5, | |
| "eval_exact_match": 0.03756708407871199, | |
| "eval_loss": 0.06500059366226196, | |
| "eval_runtime": 27.5272, | |
| "eval_samples_per_second": 20.307, | |
| "eval_steps_per_second": 2.543, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 257.13, | |
| "learning_rate": 0.0004, | |
| "loss": 0.0002, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 257.13, | |
| "eval_exact_match": 0.03935599284436494, | |
| "eval_loss": 0.06925792992115021, | |
| "eval_runtime": 27.2422, | |
| "eval_samples_per_second": 20.52, | |
| "eval_steps_per_second": 2.57, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 285.63, | |
| "learning_rate": 0.0003333333333333333, | |
| "loss": 0.0002, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 285.63, | |
| "eval_exact_match": 0.03935599284436494, | |
| "eval_loss": 0.07205212116241455, | |
| "eval_runtime": 27.1898, | |
| "eval_samples_per_second": 20.559, | |
| "eval_steps_per_second": 2.574, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 314.25, | |
| "learning_rate": 0.0002666666666666667, | |
| "loss": 0.0002, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 314.25, | |
| "eval_exact_match": 0.03756708407871199, | |
| "eval_loss": 0.0713699460029602, | |
| "eval_runtime": 27.2794, | |
| "eval_samples_per_second": 20.492, | |
| "eval_steps_per_second": 2.566, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 342.75, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0002, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 342.75, | |
| "eval_exact_match": 0.03935599284436494, | |
| "eval_loss": 0.07009705901145935, | |
| "eval_runtime": 27.286, | |
| "eval_samples_per_second": 20.487, | |
| "eval_steps_per_second": 2.565, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 371.38, | |
| "learning_rate": 0.00013333333333333334, | |
| "loss": 0.0002, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 371.38, | |
| "eval_exact_match": 0.03935599284436494, | |
| "eval_loss": 0.07500004023313522, | |
| "eval_runtime": 27.1876, | |
| "eval_samples_per_second": 20.561, | |
| "eval_steps_per_second": 2.575, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 399.88, | |
| "learning_rate": 6.666666666666667e-05, | |
| "loss": 0.0001, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 399.88, | |
| "eval_exact_match": 0.03935599284436494, | |
| "eval_loss": 0.07391420006752014, | |
| "eval_runtime": 27.4245, | |
| "eval_samples_per_second": 20.383, | |
| "eval_steps_per_second": 2.552, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 428.5, | |
| "learning_rate": 0.0, | |
| "loss": 0.0001, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 428.5, | |
| "eval_exact_match": 0.03935599284436494, | |
| "eval_loss": 0.07452824711799622, | |
| "eval_runtime": 28.2025, | |
| "eval_samples_per_second": 19.821, | |
| "eval_steps_per_second": 2.482, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 428.5, | |
| "step": 3000, | |
| "total_flos": 3.592904346944225e+17, | |
| "train_loss": 0.017975078212097286, | |
| "train_runtime": 47263.2401, | |
| "train_samples_per_second": 32.499, | |
| "train_steps_per_second": 0.063 | |
| } | |
| ], | |
| "max_steps": 3000, | |
| "num_train_epochs": 429, | |
| "total_flos": 3.592904346944225e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |