| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.014222222222222223, | |
| "eval_steps": 5, | |
| "global_step": 20, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0007111111111111111, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.0, | |
| "loss": 3.9473, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0007111111111111111, | |
| "eval_loss": 4.628085136413574, | |
| "eval_runtime": 44.9571, | |
| "eval_samples_per_second": 52.695, | |
| "eval_steps_per_second": 6.606, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0014222222222222223, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.0, | |
| "loss": 4.0139, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0021333333333333334, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.0, | |
| "loss": 3.4741, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0028444444444444446, | |
| "grad_norm": 10.534650802612305, | |
| "learning_rate": 2e-05, | |
| "loss": 4.8927, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0035555555555555557, | |
| "grad_norm": 6.7862019538879395, | |
| "learning_rate": 4e-05, | |
| "loss": 3.526, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0035555555555555557, | |
| "eval_loss": 4.499617576599121, | |
| "eval_runtime": 39.947, | |
| "eval_samples_per_second": 59.304, | |
| "eval_steps_per_second": 7.435, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.004266666666666667, | |
| "grad_norm": 7.032112121582031, | |
| "learning_rate": 6e-05, | |
| "loss": 3.8045, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.004977777777777778, | |
| "grad_norm": 6.311283111572266, | |
| "learning_rate": 8e-05, | |
| "loss": 4.0029, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.005688888888888889, | |
| "grad_norm": 5.887501239776611, | |
| "learning_rate": 0.0001, | |
| "loss": 3.8592, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0064, | |
| "grad_norm": 4.442018508911133, | |
| "learning_rate": 0.00012, | |
| "loss": 3.252, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0071111111111111115, | |
| "grad_norm": 4.024985313415527, | |
| "learning_rate": 0.00014, | |
| "loss": 3.4748, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0071111111111111115, | |
| "eval_loss": 3.1893694400787354, | |
| "eval_runtime": 37.4378, | |
| "eval_samples_per_second": 63.278, | |
| "eval_steps_per_second": 7.933, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.007822222222222222, | |
| "grad_norm": 3.4165191650390625, | |
| "learning_rate": 0.00016, | |
| "loss": 2.8739, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.008533333333333334, | |
| "grad_norm": 3.099381446838379, | |
| "learning_rate": 0.00018, | |
| "loss": 2.9354, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.009244444444444444, | |
| "grad_norm": 2.8003978729248047, | |
| "learning_rate": 0.0002, | |
| "loss": 2.7117, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.009955555555555556, | |
| "grad_norm": 3.4195311069488525, | |
| "learning_rate": 0.00019510565162951537, | |
| "loss": 2.9435, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.010666666666666666, | |
| "grad_norm": 4.078439712524414, | |
| "learning_rate": 0.00018090169943749476, | |
| "loss": 2.9743, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.010666666666666666, | |
| "eval_loss": 2.8235130310058594, | |
| "eval_runtime": 37.5377, | |
| "eval_samples_per_second": 63.11, | |
| "eval_steps_per_second": 7.912, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.011377777777777778, | |
| "grad_norm": 3.4342968463897705, | |
| "learning_rate": 0.00015877852522924732, | |
| "loss": 2.9536, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.012088888888888889, | |
| "grad_norm": 3.3408138751983643, | |
| "learning_rate": 0.00013090169943749476, | |
| "loss": 2.9875, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0128, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.00013090169943749476, | |
| "loss": 2.8077, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.013511111111111111, | |
| "grad_norm": 3.445453643798828, | |
| "learning_rate": 0.0001, | |
| "loss": 2.9706, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.014222222222222223, | |
| "grad_norm": 3.3325419425964355, | |
| "learning_rate": 6.909830056250527e-05, | |
| "loss": 2.9389, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.014222222222222223, | |
| "eval_loss": 2.74452543258667, | |
| "eval_runtime": 39.7359, | |
| "eval_samples_per_second": 59.619, | |
| "eval_steps_per_second": 7.474, | |
| "step": 20 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 20, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 5, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5906018252881920.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |