{ "best_metric": 0.6887156963348389, "best_model_checkpoint": "v18/checkpoint-264", "epoch": 2.0, "eval_steps": 500, "global_step": 264, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.05303030303030303, "grad_norm": Infinity, "learning_rate": 2.2727272727272728e-06, "loss": 3.4062, "step": 7 }, { "epoch": 0.10606060606060606, "grad_norm": 63.719356536865234, "learning_rate": 7.5757575757575764e-06, "loss": 2.4674, "step": 14 }, { "epoch": 0.1590909090909091, "grad_norm": 37.641502380371094, "learning_rate": 1.287878787878788e-05, "loss": 2.3778, "step": 21 }, { "epoch": 0.21212121212121213, "grad_norm": 63.14388656616211, "learning_rate": 1.8181818181818182e-05, "loss": 1.6267, "step": 28 }, { "epoch": 0.26515151515151514, "grad_norm": 33.05717086791992, "learning_rate": 2.3484848484848487e-05, "loss": 1.1217, "step": 35 }, { "epoch": 0.3181818181818182, "grad_norm": 19.658296585083008, "learning_rate": 2.878787878787879e-05, "loss": 0.9132, "step": 42 }, { "epoch": 0.3712121212121212, "grad_norm": 15.969008445739746, "learning_rate": 3.409090909090909e-05, "loss": 0.8221, "step": 49 }, { "epoch": 0.42424242424242425, "grad_norm": 22.162548065185547, "learning_rate": 3.939393939393939e-05, "loss": 0.9165, "step": 56 }, { "epoch": 0.4772727272727273, "grad_norm": 6.785209655761719, "learning_rate": 4.46969696969697e-05, "loss": 0.7287, "step": 63 }, { "epoch": 0.5303030303030303, "grad_norm": 10.331599235534668, "learning_rate": 5e-05, "loss": 0.7022, "step": 70 }, { "epoch": 0.5833333333333334, "grad_norm": 12.152400970458984, "learning_rate": 4.9410774410774415e-05, "loss": 0.8476, "step": 77 }, { "epoch": 0.6363636363636364, "grad_norm": 8.963262557983398, "learning_rate": 4.882154882154882e-05, "loss": 0.7803, "step": 84 }, { "epoch": 0.6893939393939394, "grad_norm": 11.988513946533203, "learning_rate": 4.823232323232323e-05, "loss": 0.6913, "step": 91 }, { "epoch": 0.7424242424242424, "grad_norm": 25.78940773010254, "learning_rate": 4.7643097643097646e-05, "loss": 1.0216, "step": 98 }, { "epoch": 0.7954545454545454, "grad_norm": 9.456551551818848, "learning_rate": 4.705387205387206e-05, "loss": 0.6587, "step": 105 }, { "epoch": 0.8484848484848485, "grad_norm": 8.910447120666504, "learning_rate": 4.6464646464646464e-05, "loss": 1.0787, "step": 112 }, { "epoch": 0.9015151515151515, "grad_norm": 6.479236125946045, "learning_rate": 4.5875420875420876e-05, "loss": 0.7669, "step": 119 }, { "epoch": 0.9545454545454546, "grad_norm": 10.731230735778809, "learning_rate": 4.528619528619529e-05, "loss": 0.73, "step": 126 }, { "epoch": 1.0, "eval_accuracy": 0.5206349206349207, "eval_f1_macro": 0.4747415759168441, "eval_f1_micro": 0.5206349206349207, "eval_f1_weighted": 0.5093092290259966, "eval_loss": 0.8705391883850098, "eval_precision_macro": 0.49376170524026586, "eval_precision_micro": 0.5206349206349207, "eval_precision_weighted": 0.6478416656117851, "eval_recall_macro": 0.7047796005921629, "eval_recall_micro": 0.5206349206349207, "eval_recall_weighted": 0.5206349206349207, "eval_runtime": 0.8297, "eval_samples_per_second": 379.673, "eval_steps_per_second": 24.106, "step": 132 }, { "epoch": 1.0075757575757576, "grad_norm": 4.0531535148620605, "learning_rate": 4.46969696969697e-05, "loss": 0.5614, "step": 133 }, { "epoch": 1.0606060606060606, "grad_norm": 34.37957763671875, "learning_rate": 4.410774410774411e-05, "loss": 0.5982, "step": 140 }, { "epoch": 1.1136363636363635, "grad_norm": 6.475064277648926, "learning_rate": 4.351851851851852e-05, "loss": 0.5497, "step": 147 }, { "epoch": 1.1666666666666667, "grad_norm": 10.707737922668457, "learning_rate": 4.292929292929293e-05, "loss": 0.5874, "step": 154 }, { "epoch": 1.2196969696969697, "grad_norm": 5.579798698425293, "learning_rate": 4.234006734006734e-05, "loss": 0.5648, "step": 161 }, { "epoch": 1.2727272727272727, "grad_norm": 8.199419975280762, "learning_rate": 4.175084175084175e-05, "loss": 0.4962, "step": 168 }, { "epoch": 1.3257575757575757, "grad_norm": 15.615039825439453, "learning_rate": 4.116161616161616e-05, "loss": 0.5891, "step": 175 }, { "epoch": 1.378787878787879, "grad_norm": 13.292581558227539, "learning_rate": 4.0572390572390575e-05, "loss": 0.5445, "step": 182 }, { "epoch": 1.4318181818181819, "grad_norm": 13.447976112365723, "learning_rate": 3.998316498316498e-05, "loss": 0.6877, "step": 189 }, { "epoch": 1.4848484848484849, "grad_norm": 7.894773960113525, "learning_rate": 3.939393939393939e-05, "loss": 0.5965, "step": 196 }, { "epoch": 1.5378787878787878, "grad_norm": 6.74298620223999, "learning_rate": 3.8804713804713806e-05, "loss": 0.4475, "step": 203 }, { "epoch": 1.5909090909090908, "grad_norm": 18.184062957763672, "learning_rate": 3.821548821548822e-05, "loss": 0.4569, "step": 210 }, { "epoch": 1.643939393939394, "grad_norm": 5.813370227813721, "learning_rate": 3.7626262626262624e-05, "loss": 0.5577, "step": 217 }, { "epoch": 1.696969696969697, "grad_norm": 24.790496826171875, "learning_rate": 3.7037037037037037e-05, "loss": 0.5338, "step": 224 }, { "epoch": 1.75, "grad_norm": 7.420760154724121, "learning_rate": 3.644781144781145e-05, "loss": 0.3236, "step": 231 }, { "epoch": 1.803030303030303, "grad_norm": 14.15146541595459, "learning_rate": 3.5858585858585855e-05, "loss": 0.5346, "step": 238 }, { "epoch": 1.856060606060606, "grad_norm": 18.612266540527344, "learning_rate": 3.526936026936027e-05, "loss": 0.6276, "step": 245 }, { "epoch": 1.9090909090909092, "grad_norm": 14.48696517944336, "learning_rate": 3.468013468013468e-05, "loss": 0.5348, "step": 252 }, { "epoch": 1.9621212121212122, "grad_norm": 20.534332275390625, "learning_rate": 3.409090909090909e-05, "loss": 0.522, "step": 259 }, { "epoch": 2.0, "eval_accuracy": 0.638095238095238, "eval_f1_macro": 0.6568419235085902, "eval_f1_micro": 0.638095238095238, "eval_f1_weighted": 0.6344955456066568, "eval_loss": 0.6887156963348389, "eval_precision_macro": 0.7028995456179922, "eval_precision_micro": 0.638095238095238, "eval_precision_weighted": 0.7468468005684834, "eval_recall_macro": 0.6885887791171937, "eval_recall_micro": 0.638095238095238, "eval_recall_weighted": 0.638095238095238, "eval_runtime": 0.834, "eval_samples_per_second": 377.711, "eval_steps_per_second": 23.982, "step": 264 } ], "logging_steps": 7, "max_steps": 660, "num_input_tokens_seen": 0, "num_train_epochs": 5, "save_steps": 500, "total_flos": 138134544307200.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }