| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 20.0, | |
| "eval_steps": 500, | |
| "global_step": 10560, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.387133002281189, | |
| "learning_rate": 4.75e-05, | |
| "loss": 0.713, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8990360727417271, | |
| "eval_f1": 0.42582781456953644, | |
| "eval_loss": 0.3558189868927002, | |
| "eval_precision": 0.49499615088529636, | |
| "eval_recall": 0.3736199883788495, | |
| "eval_runtime": 4.4266, | |
| "eval_samples_per_second": 211.224, | |
| "eval_steps_per_second": 3.389, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.8080163598060608, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.2793, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9392328331511478, | |
| "eval_f1": 0.7174307174307173, | |
| "eval_loss": 0.19311487674713135, | |
| "eval_precision": 0.647196261682243, | |
| "eval_recall": 0.8047646717024985, | |
| "eval_runtime": 4.5574, | |
| "eval_samples_per_second": 205.159, | |
| "eval_steps_per_second": 3.291, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.451019287109375, | |
| "learning_rate": 4.25e-05, | |
| "loss": 0.1876, | |
| "step": 1584 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9461890092417768, | |
| "eval_f1": 0.7516120711890638, | |
| "eval_loss": 0.161931574344635, | |
| "eval_precision": 0.6757884972170687, | |
| "eval_recall": 0.8466008134805346, | |
| "eval_runtime": 4.5232, | |
| "eval_samples_per_second": 206.71, | |
| "eval_steps_per_second": 3.316, | |
| "step": 1584 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 2.173278570175171, | |
| "learning_rate": 4e-05, | |
| "loss": 0.1593, | |
| "step": 2112 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.95553015999205, | |
| "eval_f1": 0.7947551511907948, | |
| "eval_loss": 0.14164301753044128, | |
| "eval_precision": 0.7366071428571429, | |
| "eval_recall": 0.862870424171993, | |
| "eval_runtime": 4.5204, | |
| "eval_samples_per_second": 206.84, | |
| "eval_steps_per_second": 3.318, | |
| "step": 2112 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.1243175268173218, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 0.1412, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9559276557686575, | |
| "eval_f1": 0.7968959058067969, | |
| "eval_loss": 0.13503268361091614, | |
| "eval_precision": 0.7385912698412699, | |
| "eval_recall": 0.8651946542707728, | |
| "eval_runtime": 4.5021, | |
| "eval_samples_per_second": 207.683, | |
| "eval_steps_per_second": 3.332, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 1.4806779623031616, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.1325, | |
| "step": 3168 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.95553015999205, | |
| "eval_f1": 0.795219123505976, | |
| "eval_loss": 0.13609066605567932, | |
| "eval_precision": 0.7323874755381604, | |
| "eval_recall": 0.8698431144683324, | |
| "eval_runtime": 4.5099, | |
| "eval_samples_per_second": 207.323, | |
| "eval_steps_per_second": 3.326, | |
| "step": 3168 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 2.750013828277588, | |
| "learning_rate": 3.2500000000000004e-05, | |
| "loss": 0.126, | |
| "step": 3696 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9553314121037464, | |
| "eval_f1": 0.7943751658264792, | |
| "eval_loss": 0.13828907907009125, | |
| "eval_precision": 0.73095703125, | |
| "eval_recall": 0.8698431144683324, | |
| "eval_runtime": 4.5296, | |
| "eval_samples_per_second": 206.419, | |
| "eval_steps_per_second": 3.312, | |
| "step": 3696 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.7366927862167358, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1194, | |
| "step": 4224 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9583126304283017, | |
| "eval_f1": 0.8088274395107684, | |
| "eval_loss": 0.13490265607833862, | |
| "eval_precision": 0.7455882352941177, | |
| "eval_recall": 0.8837884950610111, | |
| "eval_runtime": 4.5389, | |
| "eval_samples_per_second": 205.997, | |
| "eval_steps_per_second": 3.305, | |
| "step": 4224 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 1.5051740407943726, | |
| "learning_rate": 2.7500000000000004e-05, | |
| "loss": 0.1137, | |
| "step": 4752 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.9583126304283017, | |
| "eval_f1": 0.8071869133816038, | |
| "eval_loss": 0.12994244694709778, | |
| "eval_precision": 0.7495019920318725, | |
| "eval_recall": 0.8744915746658919, | |
| "eval_runtime": 4.5117, | |
| "eval_samples_per_second": 207.237, | |
| "eval_steps_per_second": 3.325, | |
| "step": 4752 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 1.5836716890335083, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.1112, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.9579151346516943, | |
| "eval_f1": 0.8028962188254224, | |
| "eval_loss": 0.1285104751586914, | |
| "eval_precision": 0.7455179282868526, | |
| "eval_recall": 0.8698431144683324, | |
| "eval_runtime": 4.5077, | |
| "eval_samples_per_second": 207.421, | |
| "eval_steps_per_second": 3.328, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 2.5579893589019775, | |
| "learning_rate": 2.25e-05, | |
| "loss": 0.1065, | |
| "step": 5808 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.958710126204909, | |
| "eval_f1": 0.8118811881188118, | |
| "eval_loss": 0.13038235902786255, | |
| "eval_precision": 0.7524801587301587, | |
| "eval_recall": 0.8814642649622313, | |
| "eval_runtime": 4.5296, | |
| "eval_samples_per_second": 206.421, | |
| "eval_steps_per_second": 3.312, | |
| "step": 5808 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 1.483597993850708, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1044, | |
| "step": 6336 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.9577163867633907, | |
| "eval_f1": 0.8106080900080364, | |
| "eval_loss": 0.1328919380903244, | |
| "eval_precision": 0.7519880715705766, | |
| "eval_recall": 0.8791400348634515, | |
| "eval_runtime": 4.5066, | |
| "eval_samples_per_second": 207.475, | |
| "eval_steps_per_second": 3.328, | |
| "step": 6336 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 2.8680760860443115, | |
| "learning_rate": 1.75e-05, | |
| "loss": 0.1026, | |
| "step": 6864 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.9585113783166054, | |
| "eval_f1": 0.8076405703524348, | |
| "eval_loss": 0.1257127821445465, | |
| "eval_precision": 0.7520040080160321, | |
| "eval_recall": 0.8721673445671121, | |
| "eval_runtime": 4.5195, | |
| "eval_samples_per_second": 206.882, | |
| "eval_steps_per_second": 3.319, | |
| "step": 6864 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 0.6697184443473816, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.0989, | |
| "step": 7392 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.9599026135347312, | |
| "eval_f1": 0.8167341430499325, | |
| "eval_loss": 0.12645690143108368, | |
| "eval_precision": 0.7626008064516129, | |
| "eval_recall": 0.8791400348634515, | |
| "eval_runtime": 4.6991, | |
| "eval_samples_per_second": 198.976, | |
| "eval_steps_per_second": 3.192, | |
| "step": 7392 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 1.7948130369186401, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.0982, | |
| "step": 7920 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.9597038656464275, | |
| "eval_f1": 0.8180102453491508, | |
| "eval_loss": 0.1280616968870163, | |
| "eval_precision": 0.7630784708249497, | |
| "eval_recall": 0.8814642649622313, | |
| "eval_runtime": 4.812, | |
| "eval_samples_per_second": 194.304, | |
| "eval_steps_per_second": 3.117, | |
| "step": 7920 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 2.107656717300415, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0974, | |
| "step": 8448 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.9597038656464275, | |
| "eval_f1": 0.8093322606596943, | |
| "eval_loss": 0.12642435729503632, | |
| "eval_precision": 0.7514940239043825, | |
| "eval_recall": 0.8768158047646717, | |
| "eval_runtime": 4.5825, | |
| "eval_samples_per_second": 204.037, | |
| "eval_steps_per_second": 3.273, | |
| "step": 8448 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 0.980239748954773, | |
| "learning_rate": 7.5e-06, | |
| "loss": 0.0966, | |
| "step": 8976 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.9589088740932128, | |
| "eval_f1": 0.814021942734814, | |
| "eval_loss": 0.12821656465530396, | |
| "eval_precision": 0.7544642857142857, | |
| "eval_recall": 0.8837884950610111, | |
| "eval_runtime": 4.6007, | |
| "eval_samples_per_second": 203.232, | |
| "eval_steps_per_second": 3.26, | |
| "step": 8976 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 1.8313038349151611, | |
| "learning_rate": 5e-06, | |
| "loss": 0.095, | |
| "step": 9504 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.9589088740932128, | |
| "eval_f1": 0.8144966442953021, | |
| "eval_loss": 0.12922193109989166, | |
| "eval_precision": 0.7569860279441117, | |
| "eval_recall": 0.8814642649622313, | |
| "eval_runtime": 4.5707, | |
| "eval_samples_per_second": 204.565, | |
| "eval_steps_per_second": 3.282, | |
| "step": 9504 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 0.775888204574585, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.0941, | |
| "step": 10032 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.9595051177581239, | |
| "eval_f1": 0.8153722117710293, | |
| "eval_loss": 0.12684974074363708, | |
| "eval_precision": 0.7585, | |
| "eval_recall": 0.8814642649622313, | |
| "eval_runtime": 4.5518, | |
| "eval_samples_per_second": 205.415, | |
| "eval_steps_per_second": 3.295, | |
| "step": 10032 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 2.084799289703369, | |
| "learning_rate": 0.0, | |
| "loss": 0.0948, | |
| "step": 10560 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.9593063698698201, | |
| "eval_f1": 0.8162496637072908, | |
| "eval_loss": 0.12777453660964966, | |
| "eval_precision": 0.7600200400801603, | |
| "eval_recall": 0.8814642649622313, | |
| "eval_runtime": 4.5267, | |
| "eval_samples_per_second": 206.551, | |
| "eval_steps_per_second": 3.314, | |
| "step": 10560 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 10560, | |
| "total_flos": 4552961808488766.0, | |
| "train_loss": 0.15358480504064848, | |
| "train_runtime": 1231.7918, | |
| "train_samples_per_second": 136.987, | |
| "train_steps_per_second": 8.573 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 10560, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 4552961808488766.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |