| { | |
| "best_metric": 0.06633109599351883, | |
| "best_model_checkpoint": "flan-t5-base-v3-edos_labelled_aggregated/checkpoint-3500", | |
| "epoch": 5.0, | |
| "global_step": 8750, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.714285714285714e-05, | |
| "loss": 0.1362, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 4.428571428571428e-05, | |
| "loss": 0.09, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 4.1428571428571437e-05, | |
| "loss": 0.0822, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_gen_len": 4.805, | |
| "eval_loss": 0.06878727674484253, | |
| "eval_rouge1": 95.8375, | |
| "eval_rouge2": 71.9, | |
| "eval_rougeL": 95.8417, | |
| "eval_rougeLsum": 95.8333, | |
| "eval_runtime": 126.0351, | |
| "eval_samples_per_second": 31.737, | |
| "eval_steps_per_second": 3.967, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "learning_rate": 3.857142857142858e-05, | |
| "loss": 0.0751, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "learning_rate": 3.571428571428572e-05, | |
| "loss": 0.068, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "learning_rate": 3.285714285714286e-05, | |
| "loss": 0.073, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0663, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_gen_len": 4.79025, | |
| "eval_loss": 0.06633109599351883, | |
| "eval_rouge1": 95.8792, | |
| "eval_rouge2": 71.25, | |
| "eval_rougeL": 95.875, | |
| "eval_rougeLsum": 95.8792, | |
| "eval_runtime": 124.8268, | |
| "eval_samples_per_second": 32.044, | |
| "eval_steps_per_second": 4.006, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "learning_rate": 2.714285714285714e-05, | |
| "loss": 0.0571, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "learning_rate": 2.4285714285714288e-05, | |
| "loss": 0.058, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "learning_rate": 2.1428571428571428e-05, | |
| "loss": 0.0536, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_gen_len": 4.758, | |
| "eval_loss": 0.07638958841562271, | |
| "eval_rouge1": 95.65, | |
| "eval_rouge2": 69.3, | |
| "eval_rougeL": 95.65, | |
| "eval_rougeLsum": 95.6667, | |
| "eval_runtime": 124.7009, | |
| "eval_samples_per_second": 32.077, | |
| "eval_steps_per_second": 4.01, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "learning_rate": 1.8571428571428572e-05, | |
| "loss": 0.0552, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 3.43, | |
| "learning_rate": 1.5714285714285715e-05, | |
| "loss": 0.049, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "learning_rate": 1.2857142857142857e-05, | |
| "loss": 0.0468, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0486, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_gen_len": 4.76575, | |
| "eval_loss": 0.08493922650814056, | |
| "eval_rouge1": 95.8333, | |
| "eval_rouge2": 69.925, | |
| "eval_rougeL": 95.8167, | |
| "eval_rougeLsum": 95.8333, | |
| "eval_runtime": 124.2728, | |
| "eval_samples_per_second": 32.187, | |
| "eval_steps_per_second": 4.023, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 4.29, | |
| "learning_rate": 7.142857142857143e-06, | |
| "loss": 0.0429, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 4.57, | |
| "learning_rate": 4.285714285714286e-06, | |
| "loss": 0.043, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 4.86, | |
| "learning_rate": 1.4285714285714286e-06, | |
| "loss": 0.0467, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_gen_len": 4.7635, | |
| "eval_loss": 0.09458264708518982, | |
| "eval_rouge1": 95.825, | |
| "eval_rouge2": 69.8, | |
| "eval_rougeL": 95.8083, | |
| "eval_rougeLsum": 95.825, | |
| "eval_runtime": 124.9433, | |
| "eval_samples_per_second": 32.015, | |
| "eval_steps_per_second": 4.002, | |
| "step": 8750 | |
| } | |
| ], | |
| "max_steps": 8750, | |
| "num_train_epochs": 5, | |
| "total_flos": 1.198325366784e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |