| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 20.0, | |
| "eval_steps": 500, | |
| "global_step": 2440, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 4.245405197143555, | |
| "learning_rate": 4.75e-05, | |
| "loss": 0.5417, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7368421052631579, | |
| "eval_f1": 0.6531440162271805, | |
| "eval_loss": 0.4691583514213562, | |
| "eval_precision": 0.6762575228471654, | |
| "eval_recall": 0.6437988725222767, | |
| "eval_runtime": 5.2679, | |
| "eval_samples_per_second": 75.742, | |
| "eval_steps_per_second": 9.491, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 3.8213655948638916, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.4301, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7769423558897243, | |
| "eval_f1": 0.7593078346448687, | |
| "eval_loss": 0.4378258287906647, | |
| "eval_precision": 0.7546743295019157, | |
| "eval_recall": 0.8021913075104565, | |
| "eval_runtime": 5.0861, | |
| "eval_samples_per_second": 78.449, | |
| "eval_steps_per_second": 9.831, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 3.3787574768066406, | |
| "learning_rate": 4.25e-05, | |
| "loss": 0.3347, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8446115288220551, | |
| "eval_f1": 0.8086360542112754, | |
| "eval_loss": 0.34514203667640686, | |
| "eval_precision": 0.8158019614046188, | |
| "eval_recall": 0.8025550100018185, | |
| "eval_runtime": 5.1159, | |
| "eval_samples_per_second": 77.992, | |
| "eval_steps_per_second": 9.773, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.8603061437606812, | |
| "learning_rate": 4e-05, | |
| "loss": 0.2954, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8646616541353384, | |
| "eval_f1": 0.8359175094431583, | |
| "eval_loss": 0.33369535207748413, | |
| "eval_precision": 0.8376607470912432, | |
| "eval_recall": 0.8342425895617385, | |
| "eval_runtime": 5.1268, | |
| "eval_samples_per_second": 77.826, | |
| "eval_steps_per_second": 9.753, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 2.168339967727661, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 0.2632, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8571428571428571, | |
| "eval_f1": 0.8321363988633428, | |
| "eval_loss": 0.33563923835754395, | |
| "eval_precision": 0.8247520756457565, | |
| "eval_recall": 0.8414257137661394, | |
| "eval_runtime": 5.117, | |
| "eval_samples_per_second": 77.976, | |
| "eval_steps_per_second": 9.771, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 1.1380066871643066, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.2492, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8446115288220551, | |
| "eval_f1": 0.8231484129253646, | |
| "eval_loss": 0.32611119747161865, | |
| "eval_precision": 0.8109975961538461, | |
| "eval_recall": 0.84506273867976, | |
| "eval_runtime": 5.0938, | |
| "eval_samples_per_second": 78.331, | |
| "eval_steps_per_second": 9.816, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.41156822443008423, | |
| "learning_rate": 3.2500000000000004e-05, | |
| "loss": 0.227, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8796992481203008, | |
| "eval_f1": 0.8602043795620438, | |
| "eval_loss": 0.2977831959724426, | |
| "eval_precision": 0.849624060150376, | |
| "eval_recall": 0.8748863429714493, | |
| "eval_runtime": 5.1152, | |
| "eval_samples_per_second": 78.003, | |
| "eval_steps_per_second": 9.775, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 1.2771873474121094, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2189, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8947368421052632, | |
| "eval_f1": 0.8703663593044124, | |
| "eval_loss": 0.2742370069026947, | |
| "eval_precision": 0.8789149003479912, | |
| "eval_recall": 0.8630205491907619, | |
| "eval_runtime": 5.0805, | |
| "eval_samples_per_second": 78.536, | |
| "eval_steps_per_second": 9.842, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 7.185235023498535, | |
| "learning_rate": 2.7500000000000004e-05, | |
| "loss": 0.2068, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.8922305764411027, | |
| "eval_f1": 0.8715803025426456, | |
| "eval_loss": 0.28745463490486145, | |
| "eval_precision": 0.8673433153814287, | |
| "eval_recall": 0.8762502273140571, | |
| "eval_runtime": 5.257, | |
| "eval_samples_per_second": 75.898, | |
| "eval_steps_per_second": 9.511, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 6.480859756469727, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.1935, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.9072681704260651, | |
| "eval_f1": 0.8872855539522206, | |
| "eval_loss": 0.2693336009979248, | |
| "eval_precision": 0.8903508771929824, | |
| "eval_recall": 0.8843880705582834, | |
| "eval_runtime": 5.0952, | |
| "eval_samples_per_second": 78.309, | |
| "eval_steps_per_second": 9.813, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 1.1324069499969482, | |
| "learning_rate": 2.25e-05, | |
| "loss": 0.1729, | |
| "step": 1342 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.9072681704260651, | |
| "eval_f1": 0.8900228699985846, | |
| "eval_loss": 0.27149420976638794, | |
| "eval_precision": 0.8840175953079179, | |
| "eval_recall": 0.8968903436988543, | |
| "eval_runtime": 5.1005, | |
| "eval_samples_per_second": 78.228, | |
| "eval_steps_per_second": 9.803, | |
| "step": 1342 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 1.9420219659805298, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1639, | |
| "step": 1464 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.899749373433584, | |
| "eval_f1": 0.882467302933899, | |
| "eval_loss": 0.2754858136177063, | |
| "eval_precision": 0.8732988802756245, | |
| "eval_recall": 0.8940716493907983, | |
| "eval_runtime": 5.1018, | |
| "eval_samples_per_second": 78.208, | |
| "eval_steps_per_second": 9.801, | |
| "step": 1464 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 3.3185982704162598, | |
| "learning_rate": 1.75e-05, | |
| "loss": 0.1564, | |
| "step": 1586 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.9022556390977443, | |
| "eval_f1": 0.8817957385392532, | |
| "eval_loss": 0.2662343382835388, | |
| "eval_precision": 0.8827677592299257, | |
| "eval_recall": 0.8808419712675032, | |
| "eval_runtime": 5.116, | |
| "eval_samples_per_second": 77.991, | |
| "eval_steps_per_second": 9.773, | |
| "step": 1586 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 6.053642272949219, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.1495, | |
| "step": 1708 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.899749373433584, | |
| "eval_f1": 0.8835036496350365, | |
| "eval_loss": 0.29733341932296753, | |
| "eval_precision": 0.8721804511278195, | |
| "eval_recall": 0.8990725586470267, | |
| "eval_runtime": 5.3105, | |
| "eval_samples_per_second": 75.134, | |
| "eval_steps_per_second": 9.415, | |
| "step": 1708 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 0.5649229884147644, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.1487, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.9097744360902256, | |
| "eval_f1": 0.8932457339913193, | |
| "eval_loss": 0.27316734194755554, | |
| "eval_precision": 0.8864525547445254, | |
| "eval_recall": 0.9011638479723586, | |
| "eval_runtime": 5.0557, | |
| "eval_samples_per_second": 78.921, | |
| "eval_steps_per_second": 9.89, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 1.219386100769043, | |
| "learning_rate": 1e-05, | |
| "loss": 0.141, | |
| "step": 1952 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.9047619047619048, | |
| "eval_f1": 0.8888416422287391, | |
| "eval_loss": 0.2841818034648895, | |
| "eval_precision": 0.8783752990771334, | |
| "eval_recall": 0.9026186579378068, | |
| "eval_runtime": 5.1317, | |
| "eval_samples_per_second": 77.752, | |
| "eval_steps_per_second": 9.743, | |
| "step": 1952 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 3.3799021244049072, | |
| "learning_rate": 7.5e-06, | |
| "loss": 0.1276, | |
| "step": 2074 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.9047619047619048, | |
| "eval_f1": 0.8878351186601172, | |
| "eval_loss": 0.2794151306152344, | |
| "eval_precision": 0.879776516905975, | |
| "eval_recall": 0.8976177486815784, | |
| "eval_runtime": 5.1403, | |
| "eval_samples_per_second": 77.623, | |
| "eval_steps_per_second": 9.727, | |
| "step": 2074 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 4.4008965492248535, | |
| "learning_rate": 5e-06, | |
| "loss": 0.1383, | |
| "step": 2196 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.9072681704260651, | |
| "eval_f1": 0.8910359080340997, | |
| "eval_loss": 0.2787146270275116, | |
| "eval_precision": 0.8822647601476015, | |
| "eval_recall": 0.9018912529550827, | |
| "eval_runtime": 5.1249, | |
| "eval_samples_per_second": 77.855, | |
| "eval_steps_per_second": 9.756, | |
| "step": 2196 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 3.290771245956421, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.1371, | |
| "step": 2318 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.9022556390977443, | |
| "eval_f1": 0.8856624319419237, | |
| "eval_loss": 0.2780250012874603, | |
| "eval_precision": 0.8758364312267658, | |
| "eval_recall": 0.8983451536643026, | |
| "eval_runtime": 5.1132, | |
| "eval_samples_per_second": 78.033, | |
| "eval_steps_per_second": 9.779, | |
| "step": 2318 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 2.2898199558258057, | |
| "learning_rate": 0.0, | |
| "loss": 0.1248, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.9022556390977443, | |
| "eval_f1": 0.8856624319419237, | |
| "eval_loss": 0.2808963358402252, | |
| "eval_precision": 0.8758364312267658, | |
| "eval_recall": 0.8983451536643026, | |
| "eval_runtime": 5.079, | |
| "eval_samples_per_second": 78.559, | |
| "eval_steps_per_second": 9.845, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 2440, | |
| "total_flos": 8444128359504000.0, | |
| "train_loss": 0.22104478507745462, | |
| "train_runtime": 1939.3334, | |
| "train_samples_per_second": 37.518, | |
| "train_steps_per_second": 1.258 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2440, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 8444128359504000.0, | |
| "train_batch_size": 30, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |