| { |
| "best_metric": 0.92816, |
| "best_model_checkpoint": "/nvme1/yujiepan/token-dropping-logs/logs-final/train-imdb-baseline/0703-lr5e-05_warm0.1_epoch6/checkpoint-4692", |
| "epoch": 6.0, |
| "global_step": 4692, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.5, |
| "eval_accuracy": 0.91216, |
| "eval_loss": 0.4330781102180481, |
| "eval_runtime": 65.4808, |
| "eval_samples_per_second": 381.791, |
| "eval_steps_per_second": 5.971, |
| "step": 391 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 4.9680246328754146e-05, |
| "loss": 0.0385, |
| "step": 500 |
| }, |
| { |
| "epoch": 1.0, |
| "eval_accuracy": 0.91504, |
| "eval_loss": 0.3079942464828491, |
| "eval_runtime": 64.589, |
| "eval_samples_per_second": 387.063, |
| "eval_steps_per_second": 6.054, |
| "step": 782 |
| }, |
| { |
| "epoch": 1.28, |
| "learning_rate": 4.37588820464235e-05, |
| "loss": 0.0654, |
| "step": 1000 |
| }, |
| { |
| "epoch": 1.5, |
| "eval_accuracy": 0.91636, |
| "eval_loss": 0.3285704255104065, |
| "eval_runtime": 64.8695, |
| "eval_samples_per_second": 385.389, |
| "eval_steps_per_second": 6.027, |
| "step": 1173 |
| }, |
| { |
| "epoch": 1.92, |
| "learning_rate": 3.783751776409285e-05, |
| "loss": 0.0636, |
| "step": 1500 |
| }, |
| { |
| "epoch": 2.0, |
| "eval_accuracy": 0.9184, |
| "eval_loss": 0.42962637543678284, |
| "eval_runtime": 130.3196, |
| "eval_samples_per_second": 191.836, |
| "eval_steps_per_second": 3.0, |
| "step": 1564 |
| }, |
| { |
| "epoch": 2.5, |
| "eval_accuracy": 0.91704, |
| "eval_loss": 0.5365941524505615, |
| "eval_runtime": 133.611, |
| "eval_samples_per_second": 187.11, |
| "eval_steps_per_second": 2.926, |
| "step": 1955 |
| }, |
| { |
| "epoch": 2.56, |
| "learning_rate": 3.19161534817622e-05, |
| "loss": 0.0344, |
| "step": 2000 |
| }, |
| { |
| "epoch": 3.0, |
| "eval_accuracy": 0.91724, |
| "eval_loss": 0.45805788040161133, |
| "eval_runtime": 130.2627, |
| "eval_samples_per_second": 191.92, |
| "eval_steps_per_second": 3.002, |
| "step": 2346 |
| }, |
| { |
| "epoch": 3.2, |
| "learning_rate": 2.5994789199431553e-05, |
| "loss": 0.0285, |
| "step": 2500 |
| }, |
| { |
| "epoch": 3.5, |
| "eval_accuracy": 0.91644, |
| "eval_loss": 0.5552997589111328, |
| "eval_runtime": 130.5012, |
| "eval_samples_per_second": 191.569, |
| "eval_steps_per_second": 2.996, |
| "step": 2737 |
| }, |
| { |
| "epoch": 3.84, |
| "learning_rate": 2.008526764566556e-05, |
| "loss": 0.0163, |
| "step": 3000 |
| }, |
| { |
| "epoch": 4.0, |
| "eval_accuracy": 0.91936, |
| "eval_loss": 0.5054952502250671, |
| "eval_runtime": 132.4835, |
| "eval_samples_per_second": 188.703, |
| "eval_steps_per_second": 2.951, |
| "step": 3128 |
| }, |
| { |
| "epoch": 4.48, |
| "learning_rate": 1.4163903363334913e-05, |
| "loss": 0.0104, |
| "step": 3500 |
| }, |
| { |
| "epoch": 4.5, |
| "eval_accuracy": 0.92644, |
| "eval_loss": 0.5432029366493225, |
| "eval_runtime": 130.5202, |
| "eval_samples_per_second": 191.541, |
| "eval_steps_per_second": 2.996, |
| "step": 3519 |
| }, |
| { |
| "epoch": 5.0, |
| "eval_accuracy": 0.92684, |
| "eval_loss": 0.5357362627983093, |
| "eval_runtime": 130.5356, |
| "eval_samples_per_second": 191.519, |
| "eval_steps_per_second": 2.995, |
| "step": 3910 |
| }, |
| { |
| "epoch": 5.12, |
| "learning_rate": 8.242539081004264e-06, |
| "loss": 0.0034, |
| "step": 4000 |
| }, |
| { |
| "epoch": 5.5, |
| "eval_accuracy": 0.92488, |
| "eval_loss": 0.5456807613372803, |
| "eval_runtime": 132.2657, |
| "eval_samples_per_second": 189.014, |
| "eval_steps_per_second": 2.956, |
| "step": 4301 |
| }, |
| { |
| "epoch": 5.75, |
| "learning_rate": 2.3211747986736146e-06, |
| "loss": 0.0046, |
| "step": 4500 |
| }, |
| { |
| "epoch": 6.0, |
| "eval_accuracy": 0.92816, |
| "eval_loss": 0.5325844883918762, |
| "eval_runtime": 130.2637, |
| "eval_samples_per_second": 191.918, |
| "eval_steps_per_second": 3.002, |
| "step": 4692 |
| }, |
| { |
| "epoch": 6.0, |
| "step": 4692, |
| "total_flos": 2.959973061694464e+16, |
| "train_loss": 0.02827812901810002, |
| "train_runtime": 3645.2659, |
| "train_samples_per_second": 41.149, |
| "train_steps_per_second": 1.287 |
| } |
| ], |
| "max_steps": 4692, |
| "num_train_epochs": 6, |
| "total_flos": 2.959973061694464e+16, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
