| { | |
| "best_metric": 0.74, | |
| "best_model_checkpoint": "distilhubert-finetuned-not-a-word2/run-15/checkpoint-144", | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 144, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.2651739120483398, | |
| "learning_rate": 3.8109672513839134e-06, | |
| "loss": 0.7021, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 1.1859484910964966, | |
| "learning_rate": 7.621934502767827e-06, | |
| "loss": 0.6876, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 1.4216609001159668, | |
| "learning_rate": 1.143290175415174e-05, | |
| "loss": 0.6866, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.9709990620613098, | |
| "learning_rate": 1.5243869005535653e-05, | |
| "loss": 0.67, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.9969441890716553, | |
| "learning_rate": 1.9054836256919568e-05, | |
| "loss": 0.647, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.8254842758178711, | |
| "learning_rate": 2.286580350830348e-05, | |
| "loss": 0.634, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 2.2362442016601562, | |
| "learning_rate": 2.667677075968739e-05, | |
| "loss": 0.5716, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 1.7431644201278687, | |
| "learning_rate": 3.0487738011071307e-05, | |
| "loss": 0.653, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 2.926987648010254, | |
| "learning_rate": 3.429870526245522e-05, | |
| "loss": 0.6223, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_f1": 0.7326732673267327, | |
| "eval_loss": 0.72772216796875, | |
| "eval_runtime": 1.3907, | |
| "eval_samples_per_second": 46.021, | |
| "eval_steps_per_second": 5.753, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 4.254884719848633, | |
| "learning_rate": 3.641590929100184e-05, | |
| "loss": 0.6903, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 0.8755391240119934, | |
| "learning_rate": 3.5992468485292515e-05, | |
| "loss": 0.6891, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 1.7239640951156616, | |
| "learning_rate": 3.556902767958319e-05, | |
| "loss": 0.4136, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 0.6247442960739136, | |
| "learning_rate": 3.514558687387387e-05, | |
| "loss": 0.5632, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 1.531494140625, | |
| "learning_rate": 3.4722146068164546e-05, | |
| "loss": 0.4834, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": Infinity, | |
| "learning_rate": 3.4383393423597085e-05, | |
| "loss": 0.7636, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 1.7097152471542358, | |
| "learning_rate": 3.395995261788776e-05, | |
| "loss": 0.6194, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 1.788198471069336, | |
| "learning_rate": 3.353651181217844e-05, | |
| "loss": 0.6013, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 1.9111418724060059, | |
| "learning_rate": 3.311307100646911e-05, | |
| "loss": 0.6008, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 1.2587237358093262, | |
| "learning_rate": 3.268963020075979e-05, | |
| "loss": 0.6136, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_f1": 0.7326732673267327, | |
| "eval_loss": 0.6896953582763672, | |
| "eval_runtime": 1.3733, | |
| "eval_samples_per_second": 46.605, | |
| "eval_steps_per_second": 5.826, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 0.9821504950523376, | |
| "learning_rate": 3.2266189395050464e-05, | |
| "loss": 0.5849, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "grad_norm": 1.001234769821167, | |
| "learning_rate": 3.184274858934114e-05, | |
| "loss": 0.4816, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "grad_norm": 1.6126900911331177, | |
| "learning_rate": 3.141930778363182e-05, | |
| "loss": 0.5824, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 1.0846352577209473, | |
| "learning_rate": 3.0995866977922495e-05, | |
| "loss": 0.5605, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": Infinity, | |
| "learning_rate": 3.0657114333355034e-05, | |
| "loss": 0.5292, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 3.342069625854492, | |
| "learning_rate": 3.023367352764571e-05, | |
| "loss": 0.4602, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "grad_norm": 13.714946746826172, | |
| "learning_rate": 2.9894920883078253e-05, | |
| "loss": 0.6429, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "grad_norm": 2.8365912437438965, | |
| "learning_rate": 2.947148007736893e-05, | |
| "loss": 0.38, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 3.0548512935638428, | |
| "learning_rate": 2.9048039271659608e-05, | |
| "loss": 0.3343, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_f1": 0.74, | |
| "eval_loss": 0.7983360290527344, | |
| "eval_runtime": 1.3751, | |
| "eval_samples_per_second": 46.543, | |
| "eval_steps_per_second": 5.818, | |
| "step": 144 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 480, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 2143403497915440.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": { | |
| "learning_rate": 3.658528561328557e-05, | |
| "per_device_train_batch_size": 4 | |
| } | |
| } | |