{ "best_metric": null, "best_model_checkpoint": null, "epoch": 9.433962264150944, "eval_steps": 500, "global_step": 9000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.5241090146750524, "grad_norm": 14.80826473236084, "learning_rate": 4.97e-05, "loss": 1.8113, "step": 500 }, { "epoch": 0.5241090146750524, "eval_accuracy": 0.6428945988463556, "eval_f1_class_0": 0.632183908045977, "eval_f1_class_1": 0.9472222222222223, "eval_f1_class_2": 0.3657331136738056, "eval_f1_class_3": 0.5542168674698795, "eval_f1_class_4": 0.6165254237288136, "eval_f1_class_5": 0.6261808367071524, "eval_f1_class_6": 0.6833541927409262, "eval_f1_class_7": 0.5899053627760252, "eval_f1_class_8": 0.662280701754386, "eval_f1_class_9": 0.6956521739130435, "eval_f1_macro": 0.6373254803032232, "eval_f1_micro": 0.6428945988463556, "eval_f1_weighted": 0.6376303124409798, "eval_loss": 1.1190885305404663, "eval_recall_weighted": 0.6428945988463556, "eval_runtime": 18.5604, "eval_samples_per_second": 205.491, "eval_steps_per_second": 6.465, "step": 500 }, { "epoch": 1.0482180293501049, "grad_norm": 6.733876705169678, "learning_rate": 4.725110619469027e-05, "loss": 0.885, "step": 1000 }, { "epoch": 1.0482180293501049, "eval_accuracy": 0.8185631882538018, "eval_f1_class_0": 0.7974522292993631, "eval_f1_class_1": 0.9651324965132496, "eval_f1_class_2": 0.6306569343065694, "eval_f1_class_3": 0.7193229901269393, "eval_f1_class_4": 0.8341584158415841, "eval_f1_class_5": 0.807061790668348, "eval_f1_class_6": 0.8753246753246754, "eval_f1_class_7": 0.8368794326241134, "eval_f1_class_8": 0.8787446504992867, "eval_f1_class_9": 0.8255528255528255, "eval_f1_macro": 0.8170286440756953, "eval_f1_micro": 0.8185631882538018, "eval_f1_weighted": 0.8175319894578669, "eval_loss": 0.6289433836936951, "eval_recall_weighted": 0.8185631882538018, "eval_runtime": 18.5795, "eval_samples_per_second": 205.28, "eval_steps_per_second": 6.459, "step": 1000 }, { "epoch": 1.5723270440251573, "grad_norm": 21.78145408630371, "learning_rate": 4.449115044247788e-05, "loss": 0.4651, "step": 1500 }, { "epoch": 1.5723270440251573, "eval_accuracy": 0.8615626638699528, "eval_f1_class_0": 0.8236877523553162, "eval_f1_class_1": 0.9653259361997226, "eval_f1_class_2": 0.6986128625472887, "eval_f1_class_3": 0.8308457711442785, "eval_f1_class_4": 0.8667563930013459, "eval_f1_class_5": 0.8346456692913385, "eval_f1_class_6": 0.8898305084745762, "eval_f1_class_7": 0.9136420525657072, "eval_f1_class_8": 0.8668280871670703, "eval_f1_class_9": 0.9410150891632374, "eval_f1_macro": 0.8631190121909882, "eval_f1_micro": 0.8615626638699528, "eval_f1_weighted": 0.8631956720576096, "eval_loss": 0.5186640024185181, "eval_recall_weighted": 0.8615626638699528, "eval_runtime": 18.5317, "eval_samples_per_second": 205.809, "eval_steps_per_second": 6.475, "step": 1500 }, { "epoch": 2.0964360587002098, "grad_norm": 6.887778282165527, "learning_rate": 4.172566371681416e-05, "loss": 0.3292, "step": 2000 }, { "epoch": 2.0964360587002098, "eval_accuracy": 0.9003670686942842, "eval_f1_class_0": 0.8623188405797102, "eval_f1_class_1": 0.9615384615384616, "eval_f1_class_2": 0.7830045523520485, "eval_f1_class_3": 0.9002624671916011, "eval_f1_class_4": 0.8952618453865336, "eval_f1_class_5": 0.8938271604938272, "eval_f1_class_6": 0.9402173913043479, "eval_f1_class_7": 0.9341935483870967, "eval_f1_class_8": 0.9295039164490861, "eval_f1_class_9": 0.8950131233595802, "eval_f1_macro": 0.8995141307042294, "eval_f1_micro": 0.9003670686942842, "eval_f1_weighted": 0.899989630913346, "eval_loss": 0.37930119037628174, "eval_recall_weighted": 0.9003670686942842, "eval_runtime": 18.5937, "eval_samples_per_second": 205.123, "eval_steps_per_second": 6.454, "step": 2000 }, { "epoch": 2.620545073375262, "grad_norm": 18.899120330810547, "learning_rate": 3.896017699115044e-05, "loss": 0.2064, "step": 2500 }, { "epoch": 2.620545073375262, "eval_accuracy": 0.916885159937074, "eval_f1_class_0": 0.8660049627791564, "eval_f1_class_1": 0.9670329670329672, "eval_f1_class_2": 0.8037889039242219, "eval_f1_class_3": 0.9132653061224489, "eval_f1_class_4": 0.9252217997465145, "eval_f1_class_5": 0.896719319562576, "eval_f1_class_6": 0.9539295392953929, "eval_f1_class_7": 0.9450000000000001, "eval_f1_class_8": 0.9398601398601398, "eval_f1_class_9": 0.9660056657223797, "eval_f1_macro": 0.9176828604045797, "eval_f1_micro": 0.916885159937074, "eval_f1_weighted": 0.9176253934359663, "eval_loss": 0.34075525403022766, "eval_recall_weighted": 0.916885159937074, "eval_runtime": 18.4847, "eval_samples_per_second": 206.333, "eval_steps_per_second": 6.492, "step": 2500 }, { "epoch": 3.1446540880503147, "grad_norm": 1.2899887561798096, "learning_rate": 3.619469026548672e-05, "loss": 0.1562, "step": 3000 }, { "epoch": 3.1446540880503147, "eval_accuracy": 0.9326166754063975, "eval_f1_class_0": 0.8903061224489796, "eval_f1_class_1": 0.9594594594594594, "eval_f1_class_2": 0.8427128427128427, "eval_f1_class_3": 0.9363867684478372, "eval_f1_class_4": 0.9300000000000002, "eval_f1_class_5": 0.9108433734939759, "eval_f1_class_6": 0.9784366576819407, "eval_f1_class_7": 0.9619289340101522, "eval_f1_class_8": 0.9572192513368984, "eval_f1_class_9": 0.9567642956764295, "eval_f1_macro": 0.9324057705268516, "eval_f1_micro": 0.9326166754063975, "eval_f1_weighted": 0.9323951969454738, "eval_loss": 0.29545533657073975, "eval_recall_weighted": 0.9326166754063975, "eval_runtime": 18.4597, "eval_samples_per_second": 206.612, "eval_steps_per_second": 6.501, "step": 3000 }, { "epoch": 3.668763102725367, "grad_norm": 7.331942558288574, "learning_rate": 3.342920353982301e-05, "loss": 0.1097, "step": 3500 }, { "epoch": 3.668763102725367, "eval_accuracy": 0.936811746198217, "eval_f1_class_0": 0.8847497089639115, "eval_f1_class_1": 0.967391304347826, "eval_f1_class_2": 0.8646153846153847, "eval_f1_class_3": 0.9375830013280213, "eval_f1_class_4": 0.9284818067754077, "eval_f1_class_5": 0.9315068493150684, "eval_f1_class_6": 0.9733333333333333, "eval_f1_class_7": 0.9563046192259677, "eval_f1_class_8": 0.9731903485254693, "eval_f1_class_9": 0.9495225102319237, "eval_f1_macro": 0.9366678866662314, "eval_f1_micro": 0.936811746198217, "eval_f1_weighted": 0.9365912559521494, "eval_loss": 0.3290887176990509, "eval_recall_weighted": 0.936811746198217, "eval_runtime": 18.5309, "eval_samples_per_second": 205.818, "eval_steps_per_second": 6.476, "step": 3500 }, { "epoch": 4.1928721174004195, "grad_norm": 0.010240758769214153, "learning_rate": 3.066924778761062e-05, "loss": 0.1123, "step": 4000 }, { "epoch": 4.1928721174004195, "eval_accuracy": 0.9431043523859465, "eval_f1_class_0": 0.9055118110236221, "eval_f1_class_1": 0.9780821917808219, "eval_f1_class_2": 0.8661870503597122, "eval_f1_class_3": 0.9431524547803619, "eval_f1_class_4": 0.9369592088998764, "eval_f1_class_5": 0.9245742092457422, "eval_f1_class_6": 0.9734042553191489, "eval_f1_class_7": 0.958904109589041, "eval_f1_class_8": 0.9736842105263158, "eval_f1_class_9": 0.9680998613037449, "eval_f1_macro": 0.9428559362828388, "eval_f1_micro": 0.9431043523859465, "eval_f1_weighted": 0.9427303581875908, "eval_loss": 0.3067891001701355, "eval_recall_weighted": 0.9431043523859465, "eval_runtime": 18.589, "eval_samples_per_second": 205.175, "eval_steps_per_second": 6.455, "step": 4000 }, { "epoch": 4.716981132075472, "grad_norm": 0.4483562707901001, "learning_rate": 2.7903761061946903e-05, "loss": 0.0732, "step": 4500 }, { "epoch": 4.716981132075472, "eval_accuracy": 0.940744625065548, "eval_f1_class_0": 0.8982826948480845, "eval_f1_class_1": 0.9741496598639455, "eval_f1_class_2": 0.8587570621468926, "eval_f1_class_3": 0.940127388535032, "eval_f1_class_4": 0.9477707006369427, "eval_f1_class_5": 0.9290012033694345, "eval_f1_class_6": 0.972972972972973, "eval_f1_class_7": 0.9485294117647058, "eval_f1_class_8": 0.9693741677762983, "eval_f1_class_9": 0.9666666666666666, "eval_f1_macro": 0.9405631928580975, "eval_f1_micro": 0.940744625065548, "eval_f1_weighted": 0.9405207858122142, "eval_loss": 0.3208290636539459, "eval_recall_weighted": 0.940744625065548, "eval_runtime": 18.5774, "eval_samples_per_second": 205.303, "eval_steps_per_second": 6.459, "step": 4500 }, { "epoch": 5.241090146750524, "grad_norm": 0.00789484940469265, "learning_rate": 2.5138274336283185e-05, "loss": 0.0582, "step": 5000 }, { "epoch": 5.241090146750524, "eval_accuracy": 0.9444153120083901, "eval_f1_class_0": 0.8981132075471698, "eval_f1_class_1": 0.9780821917808219, "eval_f1_class_2": 0.8748241912798874, "eval_f1_class_3": 0.9479166666666665, "eval_f1_class_4": 0.9436795994993743, "eval_f1_class_5": 0.942189421894219, "eval_f1_class_6": 0.9613259668508287, "eval_f1_class_7": 0.9551122194513716, "eval_f1_class_8": 0.9713541666666666, "eval_f1_class_9": 0.9721448467966574, "eval_f1_macro": 0.9444742478433662, "eval_f1_micro": 0.9444153120083901, "eval_f1_weighted": 0.9443976005879655, "eval_loss": 0.3178301155567169, "eval_recall_weighted": 0.9444153120083901, "eval_runtime": 18.6189, "eval_samples_per_second": 204.845, "eval_steps_per_second": 6.445, "step": 5000 }, { "epoch": 5.765199161425577, "grad_norm": 0.2585061490535736, "learning_rate": 2.237278761061947e-05, "loss": 0.0533, "step": 5500 }, { "epoch": 5.765199161425577, "eval_accuracy": 0.951232302045097, "eval_f1_class_0": 0.9144316730523627, "eval_f1_class_1": 0.9754768392370572, "eval_f1_class_2": 0.8732782369146006, "eval_f1_class_3": 0.9662337662337663, "eval_f1_class_4": 0.958974358974359, "eval_f1_class_5": 0.9410288582183187, "eval_f1_class_6": 0.9840848806366048, "eval_f1_class_7": 0.9597989949748743, "eval_f1_class_8": 0.9712793733681462, "eval_f1_class_9": 0.9667590027700832, "eval_f1_macro": 0.9511345984380173, "eval_f1_micro": 0.9512323020450971, "eval_f1_weighted": 0.9512610025073973, "eval_loss": 0.30544909834861755, "eval_recall_weighted": 0.951232302045097, "eval_runtime": 18.5675, "eval_samples_per_second": 205.412, "eval_steps_per_second": 6.463, "step": 5500 }, { "epoch": 6.289308176100629, "grad_norm": 0.020399658009409904, "learning_rate": 1.9607300884955755e-05, "loss": 0.0424, "step": 6000 }, { "epoch": 6.289308176100629, "eval_accuracy": 0.9520188778185632, "eval_f1_class_0": 0.9041450777202072, "eval_f1_class_1": 0.9782016348773842, "eval_f1_class_2": 0.8834019204389574, "eval_f1_class_3": 0.9584415584415584, "eval_f1_class_4": 0.9491094147582698, "eval_f1_class_5": 0.9521472392638036, "eval_f1_class_6": 0.9824086603518267, "eval_f1_class_7": 0.9638854296388544, "eval_f1_class_8": 0.9814323607427056, "eval_f1_class_9": 0.9669421487603306, "eval_f1_macro": 0.9520115444993896, "eval_f1_micro": 0.9520188778185632, "eval_f1_weighted": 0.9520110205225185, "eval_loss": 0.30924680829048157, "eval_recall_weighted": 0.9520188778185632, "eval_runtime": 18.5678, "eval_samples_per_second": 205.41, "eval_steps_per_second": 6.463, "step": 6000 }, { "epoch": 6.813417190775681, "grad_norm": 3.3116018772125244, "learning_rate": 1.6841814159292034e-05, "loss": 0.0378, "step": 6500 }, { "epoch": 6.813417190775681, "eval_accuracy": 0.9530676455165181, "eval_f1_class_0": 0.9155107187894074, "eval_f1_class_1": 0.9768707482993196, "eval_f1_class_2": 0.8838526912181304, "eval_f1_class_3": 0.9488491048593349, "eval_f1_class_4": 0.951885565669701, "eval_f1_class_5": 0.9431680773881501, "eval_f1_class_6": 0.9813829787234043, "eval_f1_class_7": 0.9669211195928754, "eval_f1_class_8": 0.9868073878627968, "eval_f1_class_9": 0.9750000000000001, "eval_f1_macro": 0.953024839240312, "eval_f1_micro": 0.9530676455165181, "eval_f1_weighted": 0.952958033459228, "eval_loss": 0.30203908681869507, "eval_recall_weighted": 0.9530676455165181, "eval_runtime": 18.5121, "eval_samples_per_second": 206.028, "eval_steps_per_second": 6.482, "step": 6500 }, { "epoch": 7.337526205450734, "grad_norm": 0.005160727072507143, "learning_rate": 1.4081858407079645e-05, "loss": 0.0305, "step": 7000 }, { "epoch": 7.337526205450734, "eval_accuracy": 0.954116413214473, "eval_f1_class_0": 0.9171817058096415, "eval_f1_class_1": 0.9822646657571624, "eval_f1_class_2": 0.8885672937771346, "eval_f1_class_3": 0.95822454308094, "eval_f1_class_4": 0.952020202020202, "eval_f1_class_5": 0.9428918590522478, "eval_f1_class_6": 0.9761273209549071, "eval_f1_class_7": 0.9647355163727961, "eval_f1_class_8": 0.9866310160427807, "eval_f1_class_9": 0.9721448467966574, "eval_f1_macro": 0.9540788969664471, "eval_f1_micro": 0.954116413214473, "eval_f1_weighted": 0.9540130488123523, "eval_loss": 0.28602519631385803, "eval_recall_weighted": 0.954116413214473, "eval_runtime": 18.523, "eval_samples_per_second": 205.906, "eval_steps_per_second": 6.478, "step": 7000 }, { "epoch": 7.861635220125786, "grad_norm": 0.02748439833521843, "learning_rate": 1.131637168141593e-05, "loss": 0.0216, "step": 7500 }, { "epoch": 7.861635220125786, "eval_accuracy": 0.9562139486103828, "eval_f1_class_0": 0.9250317662007626, "eval_f1_class_1": 0.9794801641586868, "eval_f1_class_2": 0.8888888888888887, "eval_f1_class_3": 0.9527458492975734, "eval_f1_class_4": 0.9604086845466155, "eval_f1_class_5": 0.9476248477466506, "eval_f1_class_6": 0.9853137516688919, "eval_f1_class_7": 0.9682337992376113, "eval_f1_class_8": 0.9790575916230366, "eval_f1_class_9": 0.9736477115117891, "eval_f1_macro": 0.9560433054880507, "eval_f1_micro": 0.9562139486103828, "eval_f1_weighted": 0.956078385428201, "eval_loss": 0.303732693195343, "eval_recall_weighted": 0.9562139486103828, "eval_runtime": 18.6679, "eval_samples_per_second": 204.308, "eval_steps_per_second": 6.428, "step": 7500 }, { "epoch": 8.385744234800839, "grad_norm": 0.017909903079271317, "learning_rate": 8.550884955752212e-06, "loss": 0.0198, "step": 8000 }, { "epoch": 8.385744234800839, "eval_accuracy": 0.9562139486103828, "eval_f1_class_0": 0.9207547169811321, "eval_f1_class_1": 0.9795361527967258, "eval_f1_class_2": 0.8829337094499294, "eval_f1_class_3": 0.9623865110246432, "eval_f1_class_4": 0.9592875318066159, "eval_f1_class_5": 0.9451887941534713, "eval_f1_class_6": 0.9826897470039946, "eval_f1_class_7": 0.9707006369426752, "eval_f1_class_8": 0.9855072463768115, "eval_f1_class_9": 0.9721448467966574, "eval_f1_macro": 0.9561129893332657, "eval_f1_micro": 0.9562139486103828, "eval_f1_weighted": 0.9561806063383737, "eval_loss": 0.2868480682373047, "eval_recall_weighted": 0.9562139486103828, "eval_runtime": 18.5749, "eval_samples_per_second": 205.331, "eval_steps_per_second": 6.46, "step": 8000 }, { "epoch": 8.90985324947589, "grad_norm": 0.019022395834326744, "learning_rate": 5.785398230088496e-06, "loss": 0.0145, "step": 8500 }, { "epoch": 8.90985324947589, "eval_accuracy": 0.9580492920818039, "eval_f1_class_0": 0.9221260815822002, "eval_f1_class_1": 0.9754768392370572, "eval_f1_class_2": 0.8932584269662921, "eval_f1_class_3": 0.9621903520208606, "eval_f1_class_4": 0.9541984732824428, "eval_f1_class_5": 0.957920792079208, "eval_f1_class_6": 0.9879518072289156, "eval_f1_class_7": 0.9671717171717172, "eval_f1_class_8": 0.9867724867724869, "eval_f1_class_9": 0.9735006973500697, "eval_f1_macro": 0.9580567673691249, "eval_f1_micro": 0.9580492920818039, "eval_f1_weighted": 0.9581079328217166, "eval_loss": 0.29160091280937195, "eval_recall_weighted": 0.9580492920818039, "eval_runtime": 18.5713, "eval_samples_per_second": 205.371, "eval_steps_per_second": 6.462, "step": 8500 }, { "epoch": 9.433962264150944, "grad_norm": 0.0038110397290438414, "learning_rate": 3.019911504424779e-06, "loss": 0.014, "step": 9000 }, { "epoch": 9.433962264150944, "eval_accuracy": 0.959884635553225, "eval_f1_class_0": 0.9228855721393034, "eval_f1_class_1": 0.9809264305177112, "eval_f1_class_2": 0.9080459770114943, "eval_f1_class_3": 0.9571984435797665, "eval_f1_class_4": 0.9566326530612245, "eval_f1_class_5": 0.9585365853658537, "eval_f1_class_6": 0.9840848806366048, "eval_f1_class_7": 0.9745547073791349, "eval_f1_class_8": 0.9868073878627968, "eval_f1_class_9": 0.9680998613037449, "eval_f1_macro": 0.9597772498857635, "eval_f1_micro": 0.959884635553225, "eval_f1_weighted": 0.9597604965241316, "eval_loss": 0.28916341066360474, "eval_recall_weighted": 0.959884635553225, "eval_runtime": 18.5578, "eval_samples_per_second": 205.52, "eval_steps_per_second": 6.466, "step": 9000 } ], "logging_steps": 500, "max_steps": 9540, "num_input_tokens_seen": 0, "num_train_epochs": 10, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 3.787923857566925e+16, "train_batch_size": 16, "trial_name": null, "trial_params": null }