| { | |
| "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": 4.483087062835693, | |
| "learning_rate": 4.75e-05, | |
| "loss": 0.5384, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7368421052631579, | |
| "eval_f1": 0.6617470147024391, | |
| "eval_loss": 0.49190446734428406, | |
| "eval_precision": 0.676996860830136, | |
| "eval_recall": 0.6538006910347336, | |
| "eval_runtime": 1.8289, | |
| "eval_samples_per_second": 218.165, | |
| "eval_steps_per_second": 27.339, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 3.5009255409240723, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.4212, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8245614035087719, | |
| "eval_f1": 0.804783605457392, | |
| "eval_loss": 0.41747915744781494, | |
| "eval_precision": 0.7930195663670017, | |
| "eval_recall": 0.8358792507728678, | |
| "eval_runtime": 1.8239, | |
| "eval_samples_per_second": 218.767, | |
| "eval_steps_per_second": 27.414, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 3.475321054458618, | |
| "learning_rate": 4.25e-05, | |
| "loss": 0.3413, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8446115288220551, | |
| "eval_f1": 0.8008725610148754, | |
| "eval_loss": 0.34032315015792847, | |
| "eval_precision": 0.825657894736842, | |
| "eval_recall": 0.7850518276050191, | |
| "eval_runtime": 1.8225, | |
| "eval_samples_per_second": 218.935, | |
| "eval_steps_per_second": 27.436, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 3.4994115829467773, | |
| "learning_rate": 4e-05, | |
| "loss": 0.2888, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8471177944862155, | |
| "eval_f1": 0.8151164115613962, | |
| "eval_loss": 0.3277858793735504, | |
| "eval_precision": 0.8159193371512123, | |
| "eval_recall": 0.8143298781596654, | |
| "eval_runtime": 1.824, | |
| "eval_samples_per_second": 218.745, | |
| "eval_steps_per_second": 27.412, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.4922258853912354, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 0.2577, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8596491228070176, | |
| "eval_f1": 0.8289650949173301, | |
| "eval_loss": 0.3102515935897827, | |
| "eval_precision": 0.8325081997648369, | |
| "eval_recall": 0.82569558101473, | |
| "eval_runtime": 1.824, | |
| "eval_samples_per_second": 218.754, | |
| "eval_steps_per_second": 27.413, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 2.708418846130371, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.2495, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8671679197994987, | |
| "eval_f1": 0.8368501045387564, | |
| "eval_loss": 0.30735519528388977, | |
| "eval_precision": 0.8435514312676942, | |
| "eval_recall": 0.8310147299509002, | |
| "eval_runtime": 1.8256, | |
| "eval_samples_per_second": 218.558, | |
| "eval_steps_per_second": 27.388, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.44828805327415466, | |
| "learning_rate": 3.2500000000000004e-05, | |
| "loss": 0.2391, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8671679197994987, | |
| "eval_f1": 0.8393634395533442, | |
| "eval_loss": 0.3005428612232208, | |
| "eval_precision": 0.8402278542707444, | |
| "eval_recall": 0.8385160938352427, | |
| "eval_runtime": 1.8233, | |
| "eval_samples_per_second": 218.828, | |
| "eval_steps_per_second": 27.422, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 2.012701988220215, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2177, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8696741854636592, | |
| "eval_f1": 0.8411818738518064, | |
| "eval_loss": 0.29794931411743164, | |
| "eval_precision": 0.8448542607834644, | |
| "eval_recall": 0.8377886888525186, | |
| "eval_runtime": 1.822, | |
| "eval_samples_per_second": 218.994, | |
| "eval_steps_per_second": 27.443, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 5.480318069458008, | |
| "learning_rate": 2.7500000000000004e-05, | |
| "loss": 0.2102, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.8796992481203008, | |
| "eval_f1": 0.8548827059465357, | |
| "eval_loss": 0.2961363196372986, | |
| "eval_precision": 0.8548827059465357, | |
| "eval_recall": 0.8548827059465357, | |
| "eval_runtime": 1.8216, | |
| "eval_samples_per_second": 219.038, | |
| "eval_steps_per_second": 27.448, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 5.2318220138549805, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.2029, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.8696741854636592, | |
| "eval_f1": 0.8339841249519908, | |
| "eval_loss": 0.3042662441730499, | |
| "eval_precision": 0.8579231241892538, | |
| "eval_recall": 0.8177850518276051, | |
| "eval_runtime": 1.8231, | |
| "eval_samples_per_second": 218.855, | |
| "eval_steps_per_second": 27.425, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 0.4476318061351776, | |
| "learning_rate": 2.25e-05, | |
| "loss": 0.1829, | |
| "step": 1342 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.8796992481203008, | |
| "eval_f1": 0.8533986527862829, | |
| "eval_loss": 0.30592578649520874, | |
| "eval_precision": 0.8572003218020917, | |
| "eval_recall": 0.8498817966903074, | |
| "eval_runtime": 1.8223, | |
| "eval_samples_per_second": 218.951, | |
| "eval_steps_per_second": 27.437, | |
| "step": 1342 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 6.3299665451049805, | |
| "learning_rate": 2e-05, | |
| "loss": 0.184, | |
| "step": 1464 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.8771929824561403, | |
| "eval_f1": 0.846679266293906, | |
| "eval_loss": 0.3001992702484131, | |
| "eval_precision": 0.8609191655801824, | |
| "eval_recall": 0.8356064739043463, | |
| "eval_runtime": 1.821, | |
| "eval_samples_per_second": 219.105, | |
| "eval_steps_per_second": 27.457, | |
| "step": 1464 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 0.40616801381111145, | |
| "learning_rate": 1.75e-05, | |
| "loss": 0.1802, | |
| "step": 1586 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.8897243107769424, | |
| "eval_f1": 0.8649122807017544, | |
| "eval_loss": 0.29539263248443604, | |
| "eval_precision": 0.8710116366366366, | |
| "eval_recall": 0.8594744498999818, | |
| "eval_runtime": 1.8246, | |
| "eval_samples_per_second": 218.68, | |
| "eval_steps_per_second": 27.404, | |
| "step": 1586 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 4.2211737632751465, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.1684, | |
| "step": 1708 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.8872180451127819, | |
| "eval_f1": 0.8642908431276217, | |
| "eval_loss": 0.3007645606994629, | |
| "eval_precision": 0.8633964654080464, | |
| "eval_recall": 0.8652027641389344, | |
| "eval_runtime": 1.8212, | |
| "eval_samples_per_second": 219.082, | |
| "eval_steps_per_second": 27.454, | |
| "step": 1708 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 2.8058817386627197, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.1627, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.8872180451127819, | |
| "eval_f1": 0.8622036668943447, | |
| "eval_loss": 0.3066510558128357, | |
| "eval_precision": 0.8671602787456446, | |
| "eval_recall": 0.8577014002545917, | |
| "eval_runtime": 1.8221, | |
| "eval_samples_per_second": 218.981, | |
| "eval_steps_per_second": 27.441, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 1.3792524337768555, | |
| "learning_rate": 1e-05, | |
| "loss": 0.1581, | |
| "step": 1952 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.8771929824561403, | |
| "eval_f1": 0.8522278069611882, | |
| "eval_loss": 0.31070253252983093, | |
| "eval_precision": 0.8513631702756499, | |
| "eval_recall": 0.8531096563011457, | |
| "eval_runtime": 1.8214, | |
| "eval_samples_per_second": 219.068, | |
| "eval_steps_per_second": 27.452, | |
| "step": 1952 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 0.5356613397598267, | |
| "learning_rate": 7.5e-06, | |
| "loss": 0.1468, | |
| "step": 2074 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.8771929824561403, | |
| "eval_f1": 0.8483536940081443, | |
| "eval_loss": 0.3229115903377533, | |
| "eval_precision": 0.8575792287132493, | |
| "eval_recall": 0.8406073831605747, | |
| "eval_runtime": 1.8217, | |
| "eval_samples_per_second": 219.021, | |
| "eval_steps_per_second": 27.446, | |
| "step": 2074 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 2.944960832595825, | |
| "learning_rate": 5e-06, | |
| "loss": 0.1433, | |
| "step": 2196 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.87468671679198, | |
| "eval_f1": 0.8456742372671576, | |
| "eval_loss": 0.32474786043167114, | |
| "eval_precision": 0.8536697247706422, | |
| "eval_recall": 0.8388343335151845, | |
| "eval_runtime": 1.8221, | |
| "eval_samples_per_second": 218.981, | |
| "eval_steps_per_second": 27.441, | |
| "step": 2196 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 4.548107147216797, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.1538, | |
| "step": 2318 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.87468671679198, | |
| "eval_f1": 0.8456742372671576, | |
| "eval_loss": 0.324598103761673, | |
| "eval_precision": 0.8536697247706422, | |
| "eval_recall": 0.8388343335151845, | |
| "eval_runtime": 1.8228, | |
| "eval_samples_per_second": 218.896, | |
| "eval_steps_per_second": 27.431, | |
| "step": 2318 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 3.463918924331665, | |
| "learning_rate": 0.0, | |
| "loss": 0.1412, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.87468671679198, | |
| "eval_f1": 0.8456742372671576, | |
| "eval_loss": 0.32347819209098816, | |
| "eval_precision": 0.8536697247706422, | |
| "eval_recall": 0.8388343335151845, | |
| "eval_runtime": 1.8216, | |
| "eval_samples_per_second": 219.041, | |
| "eval_steps_per_second": 27.449, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 2440, | |
| "total_flos": 8444128359504000.0, | |
| "train_loss": 0.2294086073265701, | |
| "train_runtime": 651.9313, | |
| "train_samples_per_second": 111.607, | |
| "train_steps_per_second": 3.743 | |
| } | |
| ], | |
| "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 | |
| } | |