| { | |
| "best_metric": 1.0, | |
| "best_model_checkpoint": "train_binary/05-13-2024_11:56:46/checkpoint-20", | |
| "epoch": 3.5, | |
| "eval_steps": 10, | |
| "global_step": 70, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 9.57083797454834, | |
| "learning_rate": 5e-05, | |
| "loss": 0.543, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_acc_product": 0.9359294458762887, | |
| "eval_loss": 0.07417334616184235, | |
| "eval_overall_acc": 0.9711111111111111, | |
| "eval_runtime": 2.1515, | |
| "eval_samples_per_second": 209.161, | |
| "eval_sarcasm_less_acc": 0.9921875, | |
| "eval_sarcasm_more_acc": 0.9432989690721649, | |
| "eval_steps_per_second": 1.859, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 7.6136794090271, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 0.0292, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_acc_product": 1.0, | |
| "eval_loss": 0.0033113760873675346, | |
| "eval_overall_acc": 1.0, | |
| "eval_runtime": 2.1413, | |
| "eval_samples_per_second": 210.153, | |
| "eval_sarcasm_less_acc": 1.0, | |
| "eval_sarcasm_more_acc": 1.0, | |
| "eval_steps_per_second": 1.868, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.8403103351593018, | |
| "learning_rate": 3.888888888888889e-05, | |
| "loss": 0.0208, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "eval_acc_product": 0.9948453608247423, | |
| "eval_loss": 0.01030806079506874, | |
| "eval_overall_acc": 0.9977777777777778, | |
| "eval_runtime": 2.1431, | |
| "eval_samples_per_second": 209.98, | |
| "eval_sarcasm_less_acc": 1.0, | |
| "eval_sarcasm_more_acc": 0.9948453608247423, | |
| "eval_steps_per_second": 1.866, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.03753209486603737, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.0177, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_acc_product": 0.9896907216494846, | |
| "eval_loss": 0.01667691580951214, | |
| "eval_overall_acc": 0.9955555555555555, | |
| "eval_runtime": 2.1482, | |
| "eval_samples_per_second": 209.475, | |
| "eval_sarcasm_less_acc": 1.0, | |
| "eval_sarcasm_more_acc": 0.9896907216494846, | |
| "eval_steps_per_second": 1.862, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.07835885137319565, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 0.0142, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "eval_acc_product": 1.0, | |
| "eval_loss": 0.0007662121206521988, | |
| "eval_overall_acc": 1.0, | |
| "eval_runtime": 2.1491, | |
| "eval_samples_per_second": 209.392, | |
| "eval_sarcasm_less_acc": 1.0, | |
| "eval_sarcasm_more_acc": 1.0, | |
| "eval_steps_per_second": 1.861, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.010010856203734875, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 0.0063, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_acc_product": 1.0, | |
| "eval_loss": 0.001220210688188672, | |
| "eval_overall_acc": 1.0, | |
| "eval_runtime": 2.1485, | |
| "eval_samples_per_second": 209.449, | |
| "eval_sarcasm_less_acc": 1.0, | |
| "eval_sarcasm_more_acc": 1.0, | |
| "eval_steps_per_second": 1.862, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 0.007230594754219055, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 0.0032, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "eval_acc_product": 1.0, | |
| "eval_loss": 0.0004007717070635408, | |
| "eval_overall_acc": 1.0, | |
| "eval_runtime": 2.1484, | |
| "eval_samples_per_second": 209.459, | |
| "eval_sarcasm_less_acc": 1.0, | |
| "eval_sarcasm_more_acc": 1.0, | |
| "eval_steps_per_second": 1.862, | |
| "step": 70 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 100, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 10, | |
| "total_flos": 2080536768629760.0, | |
| "train_batch_size": 128, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |