| { |
| "best_metric": 0.4982830286026001, |
| "best_model_checkpoint": "./results/checkpoint-218", |
| "epoch": 12.0, |
| "eval_steps": 500, |
| "global_step": 327, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.37, |
| "grad_norm": 0.376583456993103, |
| "learning_rate": 0.0002, |
| "loss": 1.4678, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.73, |
| "grad_norm": 0.5780853629112244, |
| "learning_rate": 0.0002, |
| "loss": 1.473, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.99, |
| "eval_loss": 1.2861968278884888, |
| "eval_runtime": 122.192, |
| "eval_samples_per_second": 0.892, |
| "eval_steps_per_second": 0.115, |
| "step": 27 |
| }, |
| { |
| "epoch": 1.1, |
| "grad_norm": 0.5245345830917358, |
| "learning_rate": 0.0002, |
| "loss": 1.3456, |
| "step": 30 |
| }, |
| { |
| "epoch": 1.47, |
| "grad_norm": 0.6904464364051819, |
| "learning_rate": 0.0002, |
| "loss": 1.1644, |
| "step": 40 |
| }, |
| { |
| "epoch": 1.83, |
| "grad_norm": 0.6669853925704956, |
| "learning_rate": 0.0002, |
| "loss": 1.0884, |
| "step": 50 |
| }, |
| { |
| "epoch": 1.98, |
| "eval_loss": 1.0433109998703003, |
| "eval_runtime": 122.2126, |
| "eval_samples_per_second": 0.892, |
| "eval_steps_per_second": 0.115, |
| "step": 54 |
| }, |
| { |
| "epoch": 2.2, |
| "grad_norm": 0.6909930109977722, |
| "learning_rate": 0.0002, |
| "loss": 0.898, |
| "step": 60 |
| }, |
| { |
| "epoch": 2.57, |
| "grad_norm": 0.733932375907898, |
| "learning_rate": 0.0002, |
| "loss": 0.8057, |
| "step": 70 |
| }, |
| { |
| "epoch": 2.94, |
| "grad_norm": 0.8011370897293091, |
| "learning_rate": 0.0002, |
| "loss": 0.7697, |
| "step": 80 |
| }, |
| { |
| "epoch": 2.97, |
| "eval_loss": 0.8525019884109497, |
| "eval_runtime": 122.2258, |
| "eval_samples_per_second": 0.892, |
| "eval_steps_per_second": 0.115, |
| "step": 81 |
| }, |
| { |
| "epoch": 3.3, |
| "grad_norm": 0.7747021913528442, |
| "learning_rate": 0.0002, |
| "loss": 0.5784, |
| "step": 90 |
| }, |
| { |
| "epoch": 3.67, |
| "grad_norm": 0.6826988458633423, |
| "learning_rate": 0.0002, |
| "loss": 0.5246, |
| "step": 100 |
| }, |
| { |
| "epoch": 4.0, |
| "eval_loss": 0.7216689586639404, |
| "eval_runtime": 122.1337, |
| "eval_samples_per_second": 0.892, |
| "eval_steps_per_second": 0.115, |
| "step": 109 |
| }, |
| { |
| "epoch": 4.04, |
| "grad_norm": 0.8631507158279419, |
| "learning_rate": 0.0002, |
| "loss": 0.5263, |
| "step": 110 |
| }, |
| { |
| "epoch": 4.4, |
| "grad_norm": 0.7449190020561218, |
| "learning_rate": 0.0002, |
| "loss": 0.3803, |
| "step": 120 |
| }, |
| { |
| "epoch": 4.77, |
| "grad_norm": 0.7547981142997742, |
| "learning_rate": 0.0002, |
| "loss": 0.3664, |
| "step": 130 |
| }, |
| { |
| "epoch": 4.99, |
| "eval_loss": 0.636491596698761, |
| "eval_runtime": 122.2077, |
| "eval_samples_per_second": 0.892, |
| "eval_steps_per_second": 0.115, |
| "step": 136 |
| }, |
| { |
| "epoch": 5.14, |
| "grad_norm": 0.8107866644859314, |
| "learning_rate": 0.0002, |
| "loss": 0.3591, |
| "step": 140 |
| }, |
| { |
| "epoch": 5.5, |
| "grad_norm": 0.7747142314910889, |
| "learning_rate": 0.0002, |
| "loss": 0.2611, |
| "step": 150 |
| }, |
| { |
| "epoch": 5.87, |
| "grad_norm": 0.66695237159729, |
| "learning_rate": 0.0002, |
| "loss": 0.2933, |
| "step": 160 |
| }, |
| { |
| "epoch": 5.98, |
| "eval_loss": 0.5818299055099487, |
| "eval_runtime": 122.1387, |
| "eval_samples_per_second": 0.892, |
| "eval_steps_per_second": 0.115, |
| "step": 163 |
| }, |
| { |
| "epoch": 6.24, |
| "grad_norm": 0.6931923031806946, |
| "learning_rate": 0.0002, |
| "loss": 0.2187, |
| "step": 170 |
| }, |
| { |
| "epoch": 6.61, |
| "grad_norm": 0.6592116355895996, |
| "learning_rate": 0.0002, |
| "loss": 0.2054, |
| "step": 180 |
| }, |
| { |
| "epoch": 6.97, |
| "grad_norm": 0.5976551175117493, |
| "learning_rate": 0.0002, |
| "loss": 0.2359, |
| "step": 190 |
| }, |
| { |
| "epoch": 6.97, |
| "eval_loss": 0.5233004689216614, |
| "eval_runtime": 122.2093, |
| "eval_samples_per_second": 0.892, |
| "eval_steps_per_second": 0.115, |
| "step": 190 |
| }, |
| { |
| "epoch": 7.34, |
| "grad_norm": 0.7745209336280823, |
| "learning_rate": 0.0002, |
| "loss": 0.1512, |
| "step": 200 |
| }, |
| { |
| "epoch": 7.71, |
| "grad_norm": 0.6759055256843567, |
| "learning_rate": 0.0002, |
| "loss": 0.1816, |
| "step": 210 |
| }, |
| { |
| "epoch": 8.0, |
| "eval_loss": 0.4982830286026001, |
| "eval_runtime": 122.1774, |
| "eval_samples_per_second": 0.892, |
| "eval_steps_per_second": 0.115, |
| "step": 218 |
| }, |
| { |
| "epoch": 8.07, |
| "grad_norm": 0.6047687530517578, |
| "learning_rate": 0.0002, |
| "loss": 0.1495, |
| "step": 220 |
| }, |
| { |
| "epoch": 8.44, |
| "grad_norm": 0.8196636438369751, |
| "learning_rate": 0.0002, |
| "loss": 0.1141, |
| "step": 230 |
| }, |
| { |
| "epoch": 8.81, |
| "grad_norm": 0.5866973996162415, |
| "learning_rate": 0.0002, |
| "loss": 0.1403, |
| "step": 240 |
| }, |
| { |
| "epoch": 8.99, |
| "eval_loss": 0.5128377676010132, |
| "eval_runtime": 122.2482, |
| "eval_samples_per_second": 0.892, |
| "eval_steps_per_second": 0.115, |
| "step": 245 |
| }, |
| { |
| "epoch": 9.17, |
| "grad_norm": 0.537775456905365, |
| "learning_rate": 0.0002, |
| "loss": 0.1013, |
| "step": 250 |
| }, |
| { |
| "epoch": 9.54, |
| "grad_norm": 0.5751469135284424, |
| "learning_rate": 0.0002, |
| "loss": 0.0721, |
| "step": 260 |
| }, |
| { |
| "epoch": 9.91, |
| "grad_norm": 0.7250769138336182, |
| "learning_rate": 0.0002, |
| "loss": 0.1221, |
| "step": 270 |
| }, |
| { |
| "epoch": 9.98, |
| "eval_loss": 0.5187135338783264, |
| "eval_runtime": 122.1638, |
| "eval_samples_per_second": 0.892, |
| "eval_steps_per_second": 0.115, |
| "step": 272 |
| }, |
| { |
| "epoch": 10.28, |
| "grad_norm": 0.564387321472168, |
| "learning_rate": 0.0002, |
| "loss": 0.0788, |
| "step": 280 |
| }, |
| { |
| "epoch": 10.64, |
| "grad_norm": 0.6789527535438538, |
| "learning_rate": 0.0002, |
| "loss": 0.0895, |
| "step": 290 |
| }, |
| { |
| "epoch": 10.97, |
| "eval_loss": 0.5100827217102051, |
| "eval_runtime": 122.168, |
| "eval_samples_per_second": 0.892, |
| "eval_steps_per_second": 0.115, |
| "step": 299 |
| }, |
| { |
| "epoch": 11.01, |
| "grad_norm": 0.5745537281036377, |
| "learning_rate": 0.0002, |
| "loss": 0.081, |
| "step": 300 |
| }, |
| { |
| "epoch": 11.38, |
| "grad_norm": 0.6635628342628479, |
| "learning_rate": 0.0002, |
| "loss": 0.071, |
| "step": 310 |
| }, |
| { |
| "epoch": 11.74, |
| "grad_norm": 0.5656697154045105, |
| "learning_rate": 0.0002, |
| "loss": 0.076, |
| "step": 320 |
| }, |
| { |
| "epoch": 12.0, |
| "eval_loss": 0.5288968682289124, |
| "eval_runtime": 122.2102, |
| "eval_samples_per_second": 0.892, |
| "eval_steps_per_second": 0.115, |
| "step": 327 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 500, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 19, |
| "save_steps": 10, |
| "total_flos": 1.1527346189028557e+17, |
| "train_batch_size": 4, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
