| { | |
| "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": 4.053804397583008, | |
| "learning_rate": 4.75e-05, | |
| "loss": 0.4355, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8696741854636592, | |
| "eval_f1": 0.835906358747232, | |
| "eval_loss": 0.3243214786052704, | |
| "eval_precision": 0.853844109243139, | |
| "eval_recall": 0.8227859610838335, | |
| "eval_runtime": 4.9832, | |
| "eval_samples_per_second": 80.069, | |
| "eval_steps_per_second": 10.034, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 23.150257110595703, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.2295, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8897243107769424, | |
| "eval_f1": 0.8701248742380304, | |
| "eval_loss": 0.3046626150608063, | |
| "eval_precision": 0.8624507874015748, | |
| "eval_recall": 0.8794780869248955, | |
| "eval_runtime": 4.975, | |
| "eval_samples_per_second": 80.2, | |
| "eval_steps_per_second": 10.05, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.44390636682510376, | |
| "learning_rate": 4.25e-05, | |
| "loss": 0.1337, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.899749373433584, | |
| "eval_f1": 0.879667048676036, | |
| "eval_loss": 0.3747338354587555, | |
| "eval_precision": 0.8778361344537815, | |
| "eval_recall": 0.8815693762502272, | |
| "eval_runtime": 4.9774, | |
| "eval_samples_per_second": 80.162, | |
| "eval_steps_per_second": 10.045, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 15.949886322021484, | |
| "learning_rate": 4e-05, | |
| "loss": 0.1038, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8822055137844611, | |
| "eval_f1": 0.8651222336500356, | |
| "eval_loss": 0.41882890462875366, | |
| "eval_precision": 0.8518339768339769, | |
| "eval_recall": 0.8866612111292962, | |
| "eval_runtime": 5.0099, | |
| "eval_samples_per_second": 79.643, | |
| "eval_steps_per_second": 9.98, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 3.7783217430114746, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 0.072, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8872180451127819, | |
| "eval_f1": 0.8622036668943447, | |
| "eval_loss": 0.6270534992218018, | |
| "eval_precision": 0.8671602787456446, | |
| "eval_recall": 0.8577014002545917, | |
| "eval_runtime": 5.0204, | |
| "eval_samples_per_second": 79.476, | |
| "eval_steps_per_second": 9.959, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 44.57243347167969, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.0462, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8897243107769424, | |
| "eval_f1": 0.8695225637671682, | |
| "eval_loss": 0.6129250526428223, | |
| "eval_precision": 0.8631532846715328, | |
| "eval_recall": 0.8769776322967813, | |
| "eval_runtime": 4.9643, | |
| "eval_samples_per_second": 80.375, | |
| "eval_steps_per_second": 10.072, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.024074144661426544, | |
| "learning_rate": 3.2500000000000004e-05, | |
| "loss": 0.0459, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8897243107769424, | |
| "eval_f1": 0.8649122807017544, | |
| "eval_loss": 0.5890637636184692, | |
| "eval_precision": 0.8710116366366366, | |
| "eval_recall": 0.8594744498999818, | |
| "eval_runtime": 4.9832, | |
| "eval_samples_per_second": 80.069, | |
| "eval_steps_per_second": 10.034, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.022918157279491425, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0391, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8872180451127819, | |
| "eval_f1": 0.8680720368560659, | |
| "eval_loss": 0.5972921252250671, | |
| "eval_precision": 0.8587217615098657, | |
| "eval_recall": 0.8802054919076197, | |
| "eval_runtime": 5.0003, | |
| "eval_samples_per_second": 79.795, | |
| "eval_steps_per_second": 9.999, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 106.23094177246094, | |
| "learning_rate": 2.7500000000000004e-05, | |
| "loss": 0.0307, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.87468671679198, | |
| "eval_f1": 0.8584865509022812, | |
| "eval_loss": 0.7086873054504395, | |
| "eval_precision": 0.8441043083900227, | |
| "eval_recall": 0.8863429714493545, | |
| "eval_runtime": 4.9859, | |
| "eval_samples_per_second": 80.026, | |
| "eval_steps_per_second": 10.028, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.024997469037771225, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.0199, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.8972431077694235, | |
| "eval_f1": 0.8717112228173498, | |
| "eval_loss": 0.7264124155044556, | |
| "eval_precision": 0.8869295958279009, | |
| "eval_recall": 0.8597926895799237, | |
| "eval_runtime": 4.9651, | |
| "eval_samples_per_second": 80.361, | |
| "eval_steps_per_second": 10.07, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 0.004392046481370926, | |
| "learning_rate": 2.25e-05, | |
| "loss": 0.0105, | |
| "step": 1342 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.8972431077694235, | |
| "eval_f1": 0.8757339815412664, | |
| "eval_loss": 0.6738360524177551, | |
| "eval_precision": 0.8766906299500427, | |
| "eval_recall": 0.8747954173486088, | |
| "eval_runtime": 5.0179, | |
| "eval_samples_per_second": 79.516, | |
| "eval_steps_per_second": 9.964, | |
| "step": 1342 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.004026818089187145, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0131, | |
| "step": 1464 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.899749373433584, | |
| "eval_f1": 0.882467302933899, | |
| "eval_loss": 0.7488105297088623, | |
| "eval_precision": 0.8732988802756245, | |
| "eval_recall": 0.8940716493907983, | |
| "eval_runtime": 5.0007, | |
| "eval_samples_per_second": 79.788, | |
| "eval_steps_per_second": 9.999, | |
| "step": 1464 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 0.004543425515294075, | |
| "learning_rate": 1.75e-05, | |
| "loss": 0.0102, | |
| "step": 1586 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.8972431077694235, | |
| "eval_f1": 0.8792560061999484, | |
| "eval_loss": 0.7154756784439087, | |
| "eval_precision": 0.8707622232472325, | |
| "eval_recall": 0.889798145117294, | |
| "eval_runtime": 5.0136, | |
| "eval_samples_per_second": 79.584, | |
| "eval_steps_per_second": 9.973, | |
| "step": 1586 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 0.0037931231781840324, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.0061, | |
| "step": 1708 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.9072681704260651, | |
| "eval_f1": 0.8894993300948346, | |
| "eval_loss": 0.7196279168128967, | |
| "eval_precision": 0.8850535598035154, | |
| "eval_recall": 0.8943898890707401, | |
| "eval_runtime": 5.009, | |
| "eval_samples_per_second": 79.657, | |
| "eval_steps_per_second": 9.982, | |
| "step": 1708 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 0.0027608012314885855, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.0138, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.9022556390977443, | |
| "eval_f1": 0.884617951284618, | |
| "eval_loss": 0.7618029713630676, | |
| "eval_precision": 0.8772893772893773, | |
| "eval_recall": 0.8933442444080741, | |
| "eval_runtime": 5.0251, | |
| "eval_samples_per_second": 79.401, | |
| "eval_steps_per_second": 9.95, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.0022813216783106327, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0075, | |
| "step": 1952 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.9047619047619048, | |
| "eval_f1": 0.8873149414352814, | |
| "eval_loss": 0.7252941727638245, | |
| "eval_precision": 0.8806277372262774, | |
| "eval_recall": 0.8951172940534643, | |
| "eval_runtime": 5.0138, | |
| "eval_samples_per_second": 79.581, | |
| "eval_steps_per_second": 9.973, | |
| "step": 1952 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 15.638340950012207, | |
| "learning_rate": 7.5e-06, | |
| "loss": 0.0063, | |
| "step": 2074 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.9022556390977443, | |
| "eval_f1": 0.8840781602687784, | |
| "eval_loss": 0.7560042142868042, | |
| "eval_precision": 0.87816715542522, | |
| "eval_recall": 0.89084378977996, | |
| "eval_runtime": 5.0052, | |
| "eval_samples_per_second": 79.718, | |
| "eval_steps_per_second": 9.99, | |
| "step": 2074 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 0.002121408935636282, | |
| "learning_rate": 5e-06, | |
| "loss": 0.0066, | |
| "step": 2196 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.9022556390977443, | |
| "eval_f1": 0.8856624319419237, | |
| "eval_loss": 0.748332142829895, | |
| "eval_precision": 0.8758364312267658, | |
| "eval_recall": 0.8983451536643026, | |
| "eval_runtime": 4.9788, | |
| "eval_samples_per_second": 80.139, | |
| "eval_steps_per_second": 10.043, | |
| "step": 2196 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 0.004570267163217068, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.0023, | |
| "step": 2318 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.9022556390977443, | |
| "eval_f1": 0.884617951284618, | |
| "eval_loss": 0.7535205483436584, | |
| "eval_precision": 0.8772893772893773, | |
| "eval_recall": 0.8933442444080741, | |
| "eval_runtime": 5.0013, | |
| "eval_samples_per_second": 79.779, | |
| "eval_steps_per_second": 9.997, | |
| "step": 2318 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 0.0018295175395905972, | |
| "learning_rate": 0.0, | |
| "loss": 0.0021, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.9047619047619048, | |
| "eval_f1": 0.8878351186601172, | |
| "eval_loss": 0.7535876035690308, | |
| "eval_precision": 0.879776516905975, | |
| "eval_recall": 0.8976177486815784, | |
| "eval_runtime": 5.0187, | |
| "eval_samples_per_second": 79.503, | |
| "eval_steps_per_second": 9.963, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 2440, | |
| "total_flos": 7584162436176000.0, | |
| "train_loss": 0.06173487283655855, | |
| "train_runtime": 2765.1299, | |
| "train_samples_per_second": 26.313, | |
| "train_steps_per_second": 0.882 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2440, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 7584162436176000.0, | |
| "train_batch_size": 30, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |