{ "best_metric": 0.5896586775779724, "best_model_checkpoint": "v21/checkpoint-177", "epoch": 1.0, "eval_steps": 500, "global_step": 177, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.03954802259887006, "grad_norm": 73.34220886230469, "learning_rate": 1.6853932584269663e-06, "loss": 3.1497, "step": 7 }, { "epoch": 0.07909604519774012, "grad_norm": 39.214630126953125, "learning_rate": 5.056179775280899e-06, "loss": 2.5072, "step": 14 }, { "epoch": 0.11864406779661017, "grad_norm": 43.25741195678711, "learning_rate": 8.98876404494382e-06, "loss": 1.752, "step": 21 }, { "epoch": 0.15819209039548024, "grad_norm": 47.583980560302734, "learning_rate": 1.2921348314606743e-05, "loss": 1.5765, "step": 28 }, { "epoch": 0.1977401129943503, "grad_norm": 21.266357421875, "learning_rate": 1.6853932584269665e-05, "loss": 1.287, "step": 35 }, { "epoch": 0.23728813559322035, "grad_norm": 15.081623077392578, "learning_rate": 2.0786516853932585e-05, "loss": 1.2136, "step": 42 }, { "epoch": 0.2768361581920904, "grad_norm": 9.609501838684082, "learning_rate": 2.4719101123595505e-05, "loss": 0.9708, "step": 49 }, { "epoch": 0.3163841807909605, "grad_norm": 10.065922737121582, "learning_rate": 2.8651685393258425e-05, "loss": 0.7196, "step": 56 }, { "epoch": 0.3559322033898305, "grad_norm": 7.166884899139404, "learning_rate": 3.258426966292135e-05, "loss": 0.7877, "step": 63 }, { "epoch": 0.3954802259887006, "grad_norm": 12.045182228088379, "learning_rate": 3.651685393258427e-05, "loss": 0.7109, "step": 70 }, { "epoch": 0.4350282485875706, "grad_norm": 9.929503440856934, "learning_rate": 4.044943820224719e-05, "loss": 0.7244, "step": 77 }, { "epoch": 0.4745762711864407, "grad_norm": 21.984752655029297, "learning_rate": 4.438202247191011e-05, "loss": 0.6535, "step": 84 }, { "epoch": 0.5141242937853108, "grad_norm": 21.552167892456055, "learning_rate": 4.831460674157304e-05, "loss": 0.6405, "step": 91 }, { "epoch": 0.5536723163841808, "grad_norm": 9.6996431350708, "learning_rate": 4.974874371859297e-05, "loss": 0.6403, "step": 98 }, { "epoch": 0.5932203389830508, "grad_norm": 26.493928909301758, "learning_rate": 4.930904522613066e-05, "loss": 0.7105, "step": 105 }, { "epoch": 0.632768361581921, "grad_norm": 7.643754005432129, "learning_rate": 4.8869346733668344e-05, "loss": 0.683, "step": 112 }, { "epoch": 0.672316384180791, "grad_norm": 8.287223815917969, "learning_rate": 4.8429648241206034e-05, "loss": 0.5715, "step": 119 }, { "epoch": 0.711864406779661, "grad_norm": 22.24546241760254, "learning_rate": 4.7989949748743725e-05, "loss": 0.6342, "step": 126 }, { "epoch": 0.751412429378531, "grad_norm": 11.908541679382324, "learning_rate": 4.755025125628141e-05, "loss": 0.5966, "step": 133 }, { "epoch": 0.7909604519774012, "grad_norm": 6.27039909362793, "learning_rate": 4.71105527638191e-05, "loss": 0.5414, "step": 140 }, { "epoch": 0.8305084745762712, "grad_norm": 16.940906524658203, "learning_rate": 4.6670854271356784e-05, "loss": 0.6909, "step": 147 }, { "epoch": 0.8700564971751412, "grad_norm": 10.0083589553833, "learning_rate": 4.6231155778894475e-05, "loss": 0.5056, "step": 154 }, { "epoch": 0.9096045197740112, "grad_norm": 10.436842918395996, "learning_rate": 4.5791457286432166e-05, "loss": 0.5116, "step": 161 }, { "epoch": 0.9491525423728814, "grad_norm": 10.915926933288574, "learning_rate": 4.535175879396985e-05, "loss": 0.7113, "step": 168 }, { "epoch": 0.9887005649717514, "grad_norm": 8.488564491271973, "learning_rate": 4.491206030150754e-05, "loss": 0.7555, "step": 175 }, { "epoch": 1.0, "eval_accuracy": 0.707936507936508, "eval_f1_macro": 0.6544353292567254, "eval_f1_micro": 0.707936507936508, "eval_f1_weighted": 0.693246083160111, "eval_loss": 0.5896586775779724, "eval_precision_macro": 0.7424193199047, "eval_precision_micro": 0.707936507936508, "eval_precision_weighted": 0.7052234495704283, "eval_recall_macro": 0.6125864829753165, "eval_recall_micro": 0.707936507936508, "eval_recall_weighted": 0.707936507936508, "eval_runtime": 0.843, "eval_samples_per_second": 373.658, "eval_steps_per_second": 23.724, "step": 177 } ], "logging_steps": 7, "max_steps": 885, "num_input_tokens_seen": 0, "num_train_epochs": 5, "save_steps": 500, "total_flos": 93076371521280.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }