| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 9.6, | |
| "eval_steps": 75, | |
| "global_step": 6000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.6603230237960815, | |
| "learning_rate": 0.00020833333333333335, | |
| "loss": 2.7781, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 1.839109182357788, | |
| "eval_rouge1": 28.5024, | |
| "eval_rouge2": 11.2717, | |
| "eval_rougeL": 22.108, | |
| "eval_rougeLsum": 22.4361, | |
| "eval_runtime": 722.0927, | |
| "eval_samples_per_second": 13.849, | |
| "eval_steps_per_second": 0.866, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.18350088596344, | |
| "learning_rate": 0.0004166666666666667, | |
| "loss": 2.0622, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.7575578689575195, | |
| "eval_rouge1": 28.0245, | |
| "eval_rouge2": 10.6112, | |
| "eval_rougeL": 21.7353, | |
| "eval_rougeLsum": 22.0685, | |
| "eval_runtime": 728.2015, | |
| "eval_samples_per_second": 13.732, | |
| "eval_steps_per_second": 0.858, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.9635187983512878, | |
| "learning_rate": 0.000625, | |
| "loss": 1.8636, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 1.617906928062439, | |
| "eval_rouge1": 27.353, | |
| "eval_rouge2": 10.6238, | |
| "eval_rougeL": 21.4686, | |
| "eval_rougeLsum": 21.7512, | |
| "eval_runtime": 730.8407, | |
| "eval_samples_per_second": 13.683, | |
| "eval_steps_per_second": 0.855, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.1745034456253052, | |
| "learning_rate": 0.0008333333333333334, | |
| "loss": 1.7408, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 1.6142878532409668, | |
| "eval_rouge1": 28.0928, | |
| "eval_rouge2": 11.2857, | |
| "eval_rougeL": 22.06, | |
| "eval_rougeLsum": 22.3629, | |
| "eval_runtime": 802.3401, | |
| "eval_samples_per_second": 12.464, | |
| "eval_steps_per_second": 0.779, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.7837355136871338, | |
| "learning_rate": 0.0009615384615384616, | |
| "loss": 1.6492, | |
| "step": 3125 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 1.5411357879638672, | |
| "eval_rouge1": 27.8209, | |
| "eval_rouge2": 10.9184, | |
| "eval_rougeL": 21.6819, | |
| "eval_rougeLsum": 21.9773, | |
| "eval_runtime": 711.0964, | |
| "eval_samples_per_second": 14.063, | |
| "eval_steps_per_second": 0.879, | |
| "step": 3125 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.5993546843528748, | |
| "learning_rate": 0.0007692307692307693, | |
| "loss": 1.5448, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 1.4802035093307495, | |
| "eval_rouge1": 28.0433, | |
| "eval_rouge2": 11.4232, | |
| "eval_rougeL": 22.0696, | |
| "eval_rougeLsum": 22.373, | |
| "eval_runtime": 728.4308, | |
| "eval_samples_per_second": 13.728, | |
| "eval_steps_per_second": 0.858, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.7141011357307434, | |
| "learning_rate": 0.0005769230769230769, | |
| "loss": 1.4454, | |
| "step": 4375 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 1.462142825126648, | |
| "eval_rouge1": 27.8552, | |
| "eval_rouge2": 11.1708, | |
| "eval_rougeL": 21.8958, | |
| "eval_rougeLsum": 22.1949, | |
| "eval_runtime": 640.7723, | |
| "eval_samples_per_second": 15.606, | |
| "eval_steps_per_second": 0.975, | |
| "step": 4375 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.6400988101959229, | |
| "learning_rate": 0.00038461538461538467, | |
| "loss": 1.3636, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 1.4522408246994019, | |
| "eval_rouge1": 28.3264, | |
| "eval_rouge2": 11.7945, | |
| "eval_rougeL": 22.3563, | |
| "eval_rougeLsum": 22.6524, | |
| "eval_runtime": 715.9922, | |
| "eval_samples_per_second": 13.967, | |
| "eval_steps_per_second": 0.873, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 0.6958301067352295, | |
| "learning_rate": 0.00019230769230769233, | |
| "loss": 1.2978, | |
| "step": 5625 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 1.4346853494644165, | |
| "eval_rouge1": 28.444, | |
| "eval_rouge2": 11.9388, | |
| "eval_rougeL": 22.4279, | |
| "eval_rougeLsum": 22.7344, | |
| "eval_runtime": 652.829, | |
| "eval_samples_per_second": 15.318, | |
| "eval_steps_per_second": 0.957, | |
| "step": 5625 | |
| } | |
| ], | |
| "logging_steps": 625, | |
| "max_steps": 6250, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.075997032448e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |