{ "best_global_step": 471, "best_metric": 0.6097560975609756, "best_model_checkpoint": "/linkhome/rech/genini01/udd26kf/scratch/weborganizer/models/runs/answerdotai--ModernBERT-base_FormatAnnotations-Llama-3.1-8B_bsz512_lr1e-4_epochs5_warmup0.1_url1_FormatAnnotations-Llama-3.1-405B-FP8_bsz512_lr1e-4_epochs5_warmup0.1_url1/checkpoint-471", "epoch": 4.9728, "eval_steps": 500, "global_step": 780, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.64, "grad_norm": 23.25, "learning_rate": 9.686609686609687e-05, "loss": 3.0414, "num_input_tokens_seen": 249204064, "step": 100 }, { "epoch": 1.0, "eval_validation.parquet_accuracy": 0.8204, "eval_validation.parquet_accuracy__0": 0.8175675675675675, "eval_validation.parquet_accuracy__1": 0.7605633802816901, "eval_validation.parquet_accuracy__10": 0.9124797406807131, "eval_validation.parquet_accuracy__11": 0.6826722338204593, "eval_validation.parquet_accuracy__12": 0.7278481012658228, "eval_validation.parquet_accuracy__13": 0.794392523364486, "eval_validation.parquet_accuracy__14": 0.8229166666666666, "eval_validation.parquet_accuracy__15": 0.8694379391100703, "eval_validation.parquet_accuracy__16": 0.8818443804034583, "eval_validation.parquet_accuracy__17": 0.7916666666666666, "eval_validation.parquet_accuracy__18": 0.76, "eval_validation.parquet_accuracy__19": 0.7697368421052632, "eval_validation.parquet_accuracy__2": 0.7435897435897436, "eval_validation.parquet_accuracy__20": 0.5609756097560976, "eval_validation.parquet_accuracy__21": 0.696969696969697, "eval_validation.parquet_accuracy__22": 0.7434402332361516, "eval_validation.parquet_accuracy__23": 0.8166666666666667, "eval_validation.parquet_accuracy__3": 0.751937984496124, "eval_validation.parquet_accuracy__4": 0.8727544910179641, "eval_validation.parquet_accuracy__5": 0.78125, "eval_validation.parquet_accuracy__6": 0.6387832699619772, "eval_validation.parquet_accuracy__7": 0.6719367588932806, "eval_validation.parquet_accuracy__8": 0.8928571428571429, "eval_validation.parquet_accuracy__9": 0.7727272727272727, "eval_validation.parquet_accuracy_conf50": 0.8353854112778065, "eval_validation.parquet_accuracy_conf50__0": 0.8394160583941606, "eval_validation.parquet_accuracy_conf50__1": 0.7776141384388807, "eval_validation.parquet_accuracy_conf50__10": 0.9185855263157895, "eval_validation.parquet_accuracy_conf50__11": 0.7041942604856513, "eval_validation.parquet_accuracy_conf50__12": 0.7582781456953642, "eval_validation.parquet_accuracy_conf50__13": 0.810077519379845, "eval_validation.parquet_accuracy_conf50__14": 0.8432432432432433, "eval_validation.parquet_accuracy_conf50__15": 0.8792792792792793, "eval_validation.parquet_accuracy_conf50__16": 0.8890855457227138, "eval_validation.parquet_accuracy_conf50__17": 0.7916666666666666, "eval_validation.parquet_accuracy_conf50__18": 0.7699530516431925, "eval_validation.parquet_accuracy_conf50__19": 0.7903780068728522, "eval_validation.parquet_accuracy_conf50__2": 0.76, "eval_validation.parquet_accuracy_conf50__20": 0.6052631578947368, "eval_validation.parquet_accuracy_conf50__21": 0.7040816326530612, "eval_validation.parquet_accuracy_conf50__22": 0.7716049382716049, "eval_validation.parquet_accuracy_conf50__23": 0.8258426966292135, "eval_validation.parquet_accuracy_conf50__3": 0.775, "eval_validation.parquet_accuracy_conf50__4": 0.8803030303030303, "eval_validation.parquet_accuracy_conf50__5": 0.78125, "eval_validation.parquet_accuracy_conf50__6": 0.6680672268907563, "eval_validation.parquet_accuracy_conf50__7": 0.7205240174672489, "eval_validation.parquet_accuracy_conf50__8": 0.8928571428571429, "eval_validation.parquet_accuracy_conf50__9": 0.7810650887573964, "eval_validation.parquet_accuracy_conf75": 0.8959486663311524, "eval_validation.parquet_accuracy_conf75__0": 0.8859649122807017, "eval_validation.parquet_accuracy_conf75__1": 0.8411552346570397, "eval_validation.parquet_accuracy_conf75__10": 0.9511070110701108, "eval_validation.parquet_accuracy_conf75__11": 0.8146853146853147, "eval_validation.parquet_accuracy_conf75__12": 0.8326848249027238, "eval_validation.parquet_accuracy_conf75__13": 0.8713592233009708, "eval_validation.parquet_accuracy_conf75__14": 0.8823529411764706, "eval_validation.parquet_accuracy_conf75__15": 0.9323636363636364, "eval_validation.parquet_accuracy_conf75__16": 0.9233844103930713, "eval_validation.parquet_accuracy_conf75__17": 0.9113924050632911, "eval_validation.parquet_accuracy_conf75__18": 0.8301886792452831, "eval_validation.parquet_accuracy_conf75__19": 0.8878923766816144, "eval_validation.parquet_accuracy_conf75__2": 0.7796610169491526, "eval_validation.parquet_accuracy_conf75__20": 0.7407407407407407, "eval_validation.parquet_accuracy_conf75__21": 0.8266666666666667, "eval_validation.parquet_accuracy_conf75__22": 0.8582375478927203, "eval_validation.parquet_accuracy_conf75__23": 0.8656716417910447, "eval_validation.parquet_accuracy_conf75__3": 0.8390804597701149, "eval_validation.parquet_accuracy_conf75__4": 0.9332096474953617, "eval_validation.parquet_accuracy_conf75__5": 0.8518518518518519, "eval_validation.parquet_accuracy_conf75__6": 0.7719298245614035, "eval_validation.parquet_accuracy_conf75__7": 0.7790697674418605, "eval_validation.parquet_accuracy_conf75__8": 0.94, "eval_validation.parquet_accuracy_conf75__9": 0.8863636363636364, "eval_validation.parquet_accuracy_label_average": 0.7722922880043742, "eval_validation.parquet_accuracy_label_average_conf50": 0.7890679322442429, "eval_validation.parquet_accuracy_label_average_conf75": 0.8598755738060326, "eval_validation.parquet_accuracy_label_min": 0.5609756097560976, "eval_validation.parquet_accuracy_label_min_conf50": 0.6052631578947368, "eval_validation.parquet_accuracy_label_min_conf75": 0.7407407407407407, "eval_validation.parquet_loss": 0.6205468773841858, "eval_validation.parquet_proportion_conf50": 0.9665, "eval_validation.parquet_proportion_conf75": 0.7948, "eval_validation.parquet_runtime": 10.5515, "eval_validation.parquet_samples_per_second": 947.73, "eval_validation.parquet_steps_per_second": 29.664, "num_input_tokens_seen": 390215936, "step": 157 }, { "epoch": 1.2752, "grad_norm": 11.8125, "learning_rate": 8.262108262108262e-05, "loss": 2.4648, "num_input_tokens_seen": 499147424, "step": 200 }, { "epoch": 1.9152, "grad_norm": 15.375, "learning_rate": 6.837606837606838e-05, "loss": 2.3634, "num_input_tokens_seen": 751160992, "step": 300 }, { "epoch": 2.0, "eval_validation.parquet_accuracy": 0.824, "eval_validation.parquet_accuracy__0": 0.777027027027027, "eval_validation.parquet_accuracy__1": 0.780281690140845, "eval_validation.parquet_accuracy__10": 0.9149108589951378, "eval_validation.parquet_accuracy__11": 0.8225469728601252, "eval_validation.parquet_accuracy__12": 0.7436708860759493, "eval_validation.parquet_accuracy__13": 0.7158878504672898, "eval_validation.parquet_accuracy__14": 0.8229166666666666, "eval_validation.parquet_accuracy__15": 0.8852459016393442, "eval_validation.parquet_accuracy__16": 0.8818443804034583, "eval_validation.parquet_accuracy__17": 0.84375, "eval_validation.parquet_accuracy__18": 0.6711111111111111, "eval_validation.parquet_accuracy__19": 0.756578947368421, "eval_validation.parquet_accuracy__2": 0.717948717948718, "eval_validation.parquet_accuracy__20": 0.6097560975609756, "eval_validation.parquet_accuracy__21": 0.7777777777777778, "eval_validation.parquet_accuracy__22": 0.7725947521865889, "eval_validation.parquet_accuracy__23": 0.7888888888888889, "eval_validation.parquet_accuracy__3": 0.7364341085271318, "eval_validation.parquet_accuracy__4": 0.8622754491017964, "eval_validation.parquet_accuracy__5": 0.78125, "eval_validation.parquet_accuracy__6": 0.5513307984790875, "eval_validation.parquet_accuracy__7": 0.6640316205533597, "eval_validation.parquet_accuracy__8": 0.8928571428571429, "eval_validation.parquet_accuracy__9": 0.8181818181818182, "eval_validation.parquet_accuracy_conf50": 0.8392136575271598, "eval_validation.parquet_accuracy_conf50__0": 0.8102189781021898, "eval_validation.parquet_accuracy_conf50__1": 0.7997054491899853, "eval_validation.parquet_accuracy_conf50__10": 0.9210526315789473, "eval_validation.parquet_accuracy_conf50__11": 0.8410596026490066, "eval_validation.parquet_accuracy_conf50__12": 0.7715231788079471, "eval_validation.parquet_accuracy_conf50__13": 0.7344961240310077, "eval_validation.parquet_accuracy_conf50__14": 0.8486486486486486, "eval_validation.parquet_accuracy_conf50__15": 0.8942942942942943, "eval_validation.parquet_accuracy_conf50__16": 0.8896755162241888, "eval_validation.parquet_accuracy_conf50__17": 0.84375, "eval_validation.parquet_accuracy_conf50__18": 0.6854460093896714, "eval_validation.parquet_accuracy_conf50__19": 0.7766323024054983, "eval_validation.parquet_accuracy_conf50__2": 0.7333333333333333, "eval_validation.parquet_accuracy_conf50__20": 0.6578947368421053, "eval_validation.parquet_accuracy_conf50__21": 0.7857142857142857, "eval_validation.parquet_accuracy_conf50__22": 0.7932098765432098, "eval_validation.parquet_accuracy_conf50__23": 0.797752808988764, "eval_validation.parquet_accuracy_conf50__3": 0.7666666666666667, "eval_validation.parquet_accuracy_conf50__4": 0.8696969696969697, "eval_validation.parquet_accuracy_conf50__5": 0.78125, "eval_validation.parquet_accuracy_conf50__6": 0.5756302521008403, "eval_validation.parquet_accuracy_conf50__7": 0.7161572052401747, "eval_validation.parquet_accuracy_conf50__8": 0.8928571428571429, "eval_validation.parquet_accuracy_conf50__9": 0.8224852071005917, "eval_validation.parquet_accuracy_conf75": 0.8993457473578259, "eval_validation.parquet_accuracy_conf75__0": 0.8771929824561403, "eval_validation.parquet_accuracy_conf75__1": 0.8646209386281588, "eval_validation.parquet_accuracy_conf75__10": 0.9566420664206642, "eval_validation.parquet_accuracy_conf75__11": 0.9405594405594405, "eval_validation.parquet_accuracy_conf75__12": 0.8326848249027238, "eval_validation.parquet_accuracy_conf75__13": 0.8106796116504854, "eval_validation.parquet_accuracy_conf75__14": 0.8823529411764706, "eval_validation.parquet_accuracy_conf75__15": 0.944, "eval_validation.parquet_accuracy_conf75__16": 0.9227181878747501, "eval_validation.parquet_accuracy_conf75__17": 0.9493670886075949, "eval_validation.parquet_accuracy_conf75__18": 0.7672955974842768, "eval_validation.parquet_accuracy_conf75__19": 0.8834080717488789, "eval_validation.parquet_accuracy_conf75__2": 0.7627118644067796, "eval_validation.parquet_accuracy_conf75__20": 0.7777777777777778, "eval_validation.parquet_accuracy_conf75__21": 0.8933333333333333, "eval_validation.parquet_accuracy_conf75__22": 0.8735632183908046, "eval_validation.parquet_accuracy_conf75__23": 0.8432835820895522, "eval_validation.parquet_accuracy_conf75__3": 0.8390804597701149, "eval_validation.parquet_accuracy_conf75__4": 0.9294990723562152, "eval_validation.parquet_accuracy_conf75__5": 0.8518518518518519, "eval_validation.parquet_accuracy_conf75__6": 0.6900584795321637, "eval_validation.parquet_accuracy_conf75__7": 0.7674418604651163, "eval_validation.parquet_accuracy_conf75__8": 0.94, "eval_validation.parquet_accuracy_conf75__9": 0.9090909090909091, "eval_validation.parquet_accuracy_label_average": 0.7745458110341109, "eval_validation.parquet_accuracy_label_average_conf50": 0.7920479675168944, "eval_validation.parquet_accuracy_label_average_conf75": 0.8628839233572584, "eval_validation.parquet_accuracy_label_min": 0.5513307984790875, "eval_validation.parquet_accuracy_label_min_conf50": 0.5756302521008403, "eval_validation.parquet_accuracy_label_min_conf75": 0.6900584795321637, "eval_validation.parquet_loss": 0.6139147281646729, "eval_validation.parquet_proportion_conf50": 0.9665, "eval_validation.parquet_proportion_conf75": 0.7948, "eval_validation.parquet_runtime": 8.3891, "eval_validation.parquet_samples_per_second": 1192.017, "eval_validation.parquet_steps_per_second": 37.31, "num_input_tokens_seen": 783399104, "step": 314 }, { "epoch": 2.5504, "grad_norm": 9.25, "learning_rate": 5.413105413105414e-05, "loss": 2.2093, "num_input_tokens_seen": 999700736, "step": 400 }, { "epoch": 3.0, "eval_validation.parquet_accuracy": 0.8279, "eval_validation.parquet_accuracy__0": 0.831081081081081, "eval_validation.parquet_accuracy__1": 0.7971830985915493, "eval_validation.parquet_accuracy__10": 0.9108589951377634, "eval_validation.parquet_accuracy__11": 0.7390396659707724, "eval_validation.parquet_accuracy__12": 0.8164556962025317, "eval_validation.parquet_accuracy__13": 0.7906542056074767, "eval_validation.parquet_accuracy__14": 0.875, "eval_validation.parquet_accuracy__15": 0.8706088992974239, "eval_validation.parquet_accuracy__16": 0.8806916426512968, "eval_validation.parquet_accuracy__17": 0.8229166666666666, "eval_validation.parquet_accuracy__18": 0.7022222222222222, "eval_validation.parquet_accuracy__19": 0.7072368421052632, "eval_validation.parquet_accuracy__2": 0.717948717948718, "eval_validation.parquet_accuracy__20": 0.6097560975609756, "eval_validation.parquet_accuracy__21": 0.7575757575757576, "eval_validation.parquet_accuracy__22": 0.7551020408163265, "eval_validation.parquet_accuracy__23": 0.8111111111111111, "eval_validation.parquet_accuracy__3": 0.751937984496124, "eval_validation.parquet_accuracy__4": 0.874251497005988, "eval_validation.parquet_accuracy__5": 0.8125, "eval_validation.parquet_accuracy__6": 0.623574144486692, "eval_validation.parquet_accuracy__7": 0.6719367588932806, "eval_validation.parquet_accuracy__8": 0.8928571428571429, "eval_validation.parquet_accuracy__9": 0.8181818181818182, "eval_validation.parquet_accuracy_conf50": 0.8428349715468184, "eval_validation.parquet_accuracy_conf50__0": 0.8540145985401459, "eval_validation.parquet_accuracy_conf50__1": 0.8159057437407953, "eval_validation.parquet_accuracy_conf50__10": 0.9169407894736842, "eval_validation.parquet_accuracy_conf50__11": 0.7593818984547461, "eval_validation.parquet_accuracy_conf50__12": 0.8377483443708609, "eval_validation.parquet_accuracy_conf50__13": 0.8062015503875969, "eval_validation.parquet_accuracy_conf50__14": 0.8918918918918919, "eval_validation.parquet_accuracy_conf50__15": 0.8822822822822823, "eval_validation.parquet_accuracy_conf50__16": 0.887905604719764, "eval_validation.parquet_accuracy_conf50__17": 0.8229166666666666, "eval_validation.parquet_accuracy_conf50__18": 0.7183098591549296, "eval_validation.parquet_accuracy_conf50__19": 0.7353951890034365, "eval_validation.parquet_accuracy_conf50__2": 0.7333333333333333, "eval_validation.parquet_accuracy_conf50__20": 0.6578947368421053, "eval_validation.parquet_accuracy_conf50__21": 0.7653061224489796, "eval_validation.parquet_accuracy_conf50__22": 0.7808641975308642, "eval_validation.parquet_accuracy_conf50__23": 0.8202247191011236, "eval_validation.parquet_accuracy_conf50__3": 0.775, "eval_validation.parquet_accuracy_conf50__4": 0.8803030303030303, "eval_validation.parquet_accuracy_conf50__5": 0.8125, "eval_validation.parquet_accuracy_conf50__6": 0.6470588235294118, "eval_validation.parquet_accuracy_conf50__7": 0.7205240174672489, "eval_validation.parquet_accuracy_conf50__8": 0.8928571428571429, "eval_validation.parquet_accuracy_conf50__9": 0.8224852071005917, "eval_validation.parquet_accuracy_conf75": 0.9034977352793155, "eval_validation.parquet_accuracy_conf75__0": 0.8947368421052632, "eval_validation.parquet_accuracy_conf75__1": 0.8790613718411552, "eval_validation.parquet_accuracy_conf75__10": 0.9492619926199262, "eval_validation.parquet_accuracy_conf75__11": 0.8811188811188811, "eval_validation.parquet_accuracy_conf75__12": 0.8793774319066148, "eval_validation.parquet_accuracy_conf75__13": 0.8762135922330098, "eval_validation.parquet_accuracy_conf75__14": 0.9235294117647059, "eval_validation.parquet_accuracy_conf75__15": 0.9338181818181818, "eval_validation.parquet_accuracy_conf75__16": 0.9240506329113924, "eval_validation.parquet_accuracy_conf75__17": 0.9240506329113924, "eval_validation.parquet_accuracy_conf75__18": 0.7861635220125787, "eval_validation.parquet_accuracy_conf75__19": 0.8385650224215246, "eval_validation.parquet_accuracy_conf75__2": 0.7627118644067796, "eval_validation.parquet_accuracy_conf75__20": 0.7777777777777778, "eval_validation.parquet_accuracy_conf75__21": 0.88, "eval_validation.parquet_accuracy_conf75__22": 0.8735632183908046, "eval_validation.parquet_accuracy_conf75__23": 0.8656716417910447, "eval_validation.parquet_accuracy_conf75__3": 0.8390804597701149, "eval_validation.parquet_accuracy_conf75__4": 0.9387755102040817, "eval_validation.parquet_accuracy_conf75__5": 0.8888888888888888, "eval_validation.parquet_accuracy_conf75__6": 0.7543859649122807, "eval_validation.parquet_accuracy_conf75__7": 0.7965116279069767, "eval_validation.parquet_accuracy_conf75__8": 0.94, "eval_validation.parquet_accuracy_conf75__9": 0.9015151515151515, "eval_validation.parquet_accuracy_label_average": 0.7850284202694991, "eval_validation.parquet_accuracy_label_average_conf50": 0.801551906216693, "eval_validation.parquet_accuracy_label_average_conf75": 0.8712012342178551, "eval_validation.parquet_accuracy_label_min": 0.6097560975609756, "eval_validation.parquet_accuracy_label_min_conf50": 0.6470588235294118, "eval_validation.parquet_accuracy_label_min_conf75": 0.7543859649122807, "eval_validation.parquet_loss": 0.6012852787971497, "eval_validation.parquet_proportion_conf50": 0.9665, "eval_validation.parquet_proportion_conf75": 0.7948, "eval_validation.parquet_runtime": 8.4045, "eval_validation.parquet_samples_per_second": 1189.832, "eval_validation.parquet_steps_per_second": 37.242, "num_input_tokens_seen": 1176307328, "step": 471 }, { "epoch": 3.1856, "grad_norm": 11.125, "learning_rate": 3.988603988603989e-05, "loss": 2.1824, "num_input_tokens_seen": 1250925472, "step": 500 }, { "epoch": 3.8256, "grad_norm": 10.3125, "learning_rate": 2.564102564102564e-05, "loss": 2.1124, "num_input_tokens_seen": 1499507040, "step": 600 }, { "epoch": 4.0, "eval_validation.parquet_accuracy": 0.8296, "eval_validation.parquet_accuracy__0": 0.8378378378378378, "eval_validation.parquet_accuracy__1": 0.8140845070422535, "eval_validation.parquet_accuracy__10": 0.9181523500810372, "eval_validation.parquet_accuracy__11": 0.7536534446764092, "eval_validation.parquet_accuracy__12": 0.8037974683544303, "eval_validation.parquet_accuracy__13": 0.7738317757009345, "eval_validation.parquet_accuracy__14": 0.8645833333333334, "eval_validation.parquet_accuracy__15": 0.8600702576112412, "eval_validation.parquet_accuracy__16": 0.8737752161383285, "eval_validation.parquet_accuracy__17": 0.8229166666666666, "eval_validation.parquet_accuracy__18": 0.7466666666666667, "eval_validation.parquet_accuracy__19": 0.7138157894736842, "eval_validation.parquet_accuracy__2": 0.782051282051282, "eval_validation.parquet_accuracy__20": 0.6097560975609756, "eval_validation.parquet_accuracy__21": 0.7474747474747475, "eval_validation.parquet_accuracy__22": 0.7900874635568513, "eval_validation.parquet_accuracy__23": 0.7722222222222223, "eval_validation.parquet_accuracy__3": 0.751937984496124, "eval_validation.parquet_accuracy__4": 0.8937125748502994, "eval_validation.parquet_accuracy__5": 0.78125, "eval_validation.parquet_accuracy__6": 0.596958174904943, "eval_validation.parquet_accuracy__7": 0.6996047430830039, "eval_validation.parquet_accuracy__8": 0.8928571428571429, "eval_validation.parquet_accuracy__9": 0.8181818181818182, "eval_validation.parquet_accuracy_conf50": 0.8443869632695292, "eval_validation.parquet_accuracy_conf50__0": 0.8540145985401459, "eval_validation.parquet_accuracy_conf50__1": 0.8306332842415317, "eval_validation.parquet_accuracy_conf50__10": 0.9243421052631579, "eval_validation.parquet_accuracy_conf50__11": 0.7748344370860927, "eval_validation.parquet_accuracy_conf50__12": 0.8278145695364238, "eval_validation.parquet_accuracy_conf50__13": 0.7906976744186046, "eval_validation.parquet_accuracy_conf50__14": 0.8864864864864865, "eval_validation.parquet_accuracy_conf50__15": 0.8714714714714715, "eval_validation.parquet_accuracy_conf50__16": 0.8825958702064897, "eval_validation.parquet_accuracy_conf50__17": 0.8229166666666666, "eval_validation.parquet_accuracy_conf50__18": 0.755868544600939, "eval_validation.parquet_accuracy_conf50__19": 0.738831615120275, "eval_validation.parquet_accuracy_conf50__2": 0.8, "eval_validation.parquet_accuracy_conf50__20": 0.6578947368421053, "eval_validation.parquet_accuracy_conf50__21": 0.7551020408163265, "eval_validation.parquet_accuracy_conf50__22": 0.808641975308642, "eval_validation.parquet_accuracy_conf50__23": 0.7808988764044944, "eval_validation.parquet_accuracy_conf50__3": 0.775, "eval_validation.parquet_accuracy_conf50__4": 0.9, "eval_validation.parquet_accuracy_conf50__5": 0.78125, "eval_validation.parquet_accuracy_conf50__6": 0.6176470588235294, "eval_validation.parquet_accuracy_conf50__7": 0.7510917030567685, "eval_validation.parquet_accuracy_conf50__8": 0.8928571428571429, "eval_validation.parquet_accuracy_conf50__9": 0.8284023668639053, "eval_validation.parquet_accuracy_conf75": 0.903749370910921, "eval_validation.parquet_accuracy_conf75__0": 0.9035087719298246, "eval_validation.parquet_accuracy_conf75__1": 0.8916967509025271, "eval_validation.parquet_accuracy_conf75__10": 0.9529520295202952, "eval_validation.parquet_accuracy_conf75__11": 0.8811188811188811, "eval_validation.parquet_accuracy_conf75__12": 0.8793774319066148, "eval_validation.parquet_accuracy_conf75__13": 0.8640776699029126, "eval_validation.parquet_accuracy_conf75__14": 0.9176470588235294, "eval_validation.parquet_accuracy_conf75__15": 0.928, "eval_validation.parquet_accuracy_conf75__16": 0.9187208527648234, "eval_validation.parquet_accuracy_conf75__17": 0.9240506329113924, "eval_validation.parquet_accuracy_conf75__18": 0.8238993710691824, "eval_validation.parquet_accuracy_conf75__19": 0.8385650224215246, "eval_validation.parquet_accuracy_conf75__2": 0.8305084745762712, "eval_validation.parquet_accuracy_conf75__20": 0.7777777777777778, "eval_validation.parquet_accuracy_conf75__21": 0.8666666666666667, "eval_validation.parquet_accuracy_conf75__22": 0.8850574712643678, "eval_validation.parquet_accuracy_conf75__23": 0.835820895522388, "eval_validation.parquet_accuracy_conf75__3": 0.8390804597701149, "eval_validation.parquet_accuracy_conf75__4": 0.948051948051948, "eval_validation.parquet_accuracy_conf75__5": 0.8518518518518519, "eval_validation.parquet_accuracy_conf75__6": 0.7309941520467836, "eval_validation.parquet_accuracy_conf75__7": 0.813953488372093, "eval_validation.parquet_accuracy_conf75__8": 0.94, "eval_validation.parquet_accuracy_conf75__9": 0.9090909090909091, "eval_validation.parquet_accuracy_label_average": 0.7883033152009263, "eval_validation.parquet_accuracy_label_average_conf50": 0.8045538843588002, "eval_validation.parquet_accuracy_label_average_conf75": 0.8730195236776117, "eval_validation.parquet_accuracy_label_min": 0.596958174904943, "eval_validation.parquet_accuracy_label_min_conf50": 0.6176470588235294, "eval_validation.parquet_accuracy_label_min_conf75": 0.7309941520467836, "eval_validation.parquet_loss": 0.5987765789031982, "eval_validation.parquet_proportion_conf50": 0.9665, "eval_validation.parquet_proportion_conf75": 0.7948, "eval_validation.parquet_runtime": 8.4305, "eval_validation.parquet_samples_per_second": 1186.174, "eval_validation.parquet_steps_per_second": 37.127, "num_input_tokens_seen": 1566401088, "step": 628 }, { "epoch": 4.4608, "grad_norm": 7.875, "learning_rate": 1.1396011396011397e-05, "loss": 2.0751, "num_input_tokens_seen": 1745927840, "step": 700 }, { "epoch": 4.9728, "eval_validation.parquet_accuracy": 0.8297, "eval_validation.parquet_accuracy__0": 0.8378378378378378, "eval_validation.parquet_accuracy__1": 0.8014084507042254, "eval_validation.parquet_accuracy__10": 0.9173419773095624, "eval_validation.parquet_accuracy__11": 0.755741127348643, "eval_validation.parquet_accuracy__12": 0.8006329113924051, "eval_validation.parquet_accuracy__13": 0.7700934579439253, "eval_validation.parquet_accuracy__14": 0.8489583333333334, "eval_validation.parquet_accuracy__15": 0.8676814988290398, "eval_validation.parquet_accuracy__16": 0.8731988472622478, "eval_validation.parquet_accuracy__17": 0.8229166666666666, "eval_validation.parquet_accuracy__18": 0.72, "eval_validation.parquet_accuracy__19": 0.743421052631579, "eval_validation.parquet_accuracy__2": 0.7564102564102564, "eval_validation.parquet_accuracy__20": 0.6341463414634146, "eval_validation.parquet_accuracy__21": 0.7575757575757576, "eval_validation.parquet_accuracy__22": 0.7871720116618076, "eval_validation.parquet_accuracy__23": 0.7944444444444444, "eval_validation.parquet_accuracy__3": 0.751937984496124, "eval_validation.parquet_accuracy__4": 0.8907185628742516, "eval_validation.parquet_accuracy__5": 0.8125, "eval_validation.parquet_accuracy__6": 0.596958174904943, "eval_validation.parquet_accuracy__7": 0.6996047430830039, "eval_validation.parquet_accuracy__8": 0.8928571428571429, "eval_validation.parquet_accuracy__9": 0.8125, "eval_validation.parquet_accuracy_conf50": 0.844593895499224, "eval_validation.parquet_accuracy_conf50__0": 0.8540145985401459, "eval_validation.parquet_accuracy_conf50__1": 0.8188512518409425, "eval_validation.parquet_accuracy_conf50__10": 0.9235197368421053, "eval_validation.parquet_accuracy_conf50__11": 0.7770419426048565, "eval_validation.parquet_accuracy_conf50__12": 0.8278145695364238, "eval_validation.parquet_accuracy_conf50__13": 0.7868217054263565, "eval_validation.parquet_accuracy_conf50__14": 0.8702702702702703, "eval_validation.parquet_accuracy_conf50__15": 0.8786786786786787, "eval_validation.parquet_accuracy_conf50__16": 0.8814159292035398, "eval_validation.parquet_accuracy_conf50__17": 0.8229166666666666, "eval_validation.parquet_accuracy_conf50__18": 0.7370892018779343, "eval_validation.parquet_accuracy_conf50__19": 0.7697594501718213, "eval_validation.parquet_accuracy_conf50__2": 0.7733333333333333, "eval_validation.parquet_accuracy_conf50__20": 0.6578947368421053, "eval_validation.parquet_accuracy_conf50__21": 0.7653061224489796, "eval_validation.parquet_accuracy_conf50__22": 0.8055555555555556, "eval_validation.parquet_accuracy_conf50__23": 0.8033707865168539, "eval_validation.parquet_accuracy_conf50__3": 0.775, "eval_validation.parquet_accuracy_conf50__4": 0.896969696969697, "eval_validation.parquet_accuracy_conf50__5": 0.8125, "eval_validation.parquet_accuracy_conf50__6": 0.6176470588235294, "eval_validation.parquet_accuracy_conf50__7": 0.7510917030567685, "eval_validation.parquet_accuracy_conf50__8": 0.8928571428571429, "eval_validation.parquet_accuracy_conf50__9": 0.8224852071005917, "eval_validation.parquet_accuracy_conf75": 0.903749370910921, "eval_validation.parquet_accuracy_conf75__0": 0.9035087719298246, "eval_validation.parquet_accuracy_conf75__1": 0.8844765342960289, "eval_validation.parquet_accuracy_conf75__10": 0.9520295202952029, "eval_validation.parquet_accuracy_conf75__11": 0.8846153846153846, "eval_validation.parquet_accuracy_conf75__12": 0.8793774319066148, "eval_validation.parquet_accuracy_conf75__13": 0.8640776699029126, "eval_validation.parquet_accuracy_conf75__14": 0.9058823529411765, "eval_validation.parquet_accuracy_conf75__15": 0.9330909090909091, "eval_validation.parquet_accuracy_conf75__16": 0.9173884077281812, "eval_validation.parquet_accuracy_conf75__17": 0.9240506329113924, "eval_validation.parquet_accuracy_conf75__18": 0.8050314465408805, "eval_validation.parquet_accuracy_conf75__19": 0.8609865470852018, "eval_validation.parquet_accuracy_conf75__2": 0.7966101694915254, "eval_validation.parquet_accuracy_conf75__20": 0.7777777777777778, "eval_validation.parquet_accuracy_conf75__21": 0.88, "eval_validation.parquet_accuracy_conf75__22": 0.8812260536398467, "eval_validation.parquet_accuracy_conf75__23": 0.8507462686567164, "eval_validation.parquet_accuracy_conf75__3": 0.8390804597701149, "eval_validation.parquet_accuracy_conf75__4": 0.9461966604823747, "eval_validation.parquet_accuracy_conf75__5": 0.8888888888888888, "eval_validation.parquet_accuracy_conf75__6": 0.7309941520467836, "eval_validation.parquet_accuracy_conf75__7": 0.8081395348837209, "eval_validation.parquet_accuracy_conf75__8": 0.94, "eval_validation.parquet_accuracy_conf75__9": 0.9090909090909091, "eval_validation.parquet_accuracy_label_average": 0.7894190658762755, "eval_validation.parquet_accuracy_label_average_conf50": 0.805091889381846, "eval_validation.parquet_accuracy_label_average_conf75": 0.873469436832182, "eval_validation.parquet_accuracy_label_min": 0.596958174904943, "eval_validation.parquet_accuracy_label_min_conf50": 0.6176470588235294, "eval_validation.parquet_accuracy_label_min_conf75": 0.7309941520467836, "eval_validation.parquet_loss": 0.5989147424697876, "eval_validation.parquet_proportion_conf50": 0.9665, "eval_validation.parquet_proportion_conf75": 0.7948, "eval_validation.parquet_runtime": 8.5755, "eval_validation.parquet_samples_per_second": 1166.109, "eval_validation.parquet_steps_per_second": 36.499, "num_input_tokens_seen": 1949274656, "step": 780 }, { "epoch": 4.9728, "num_input_tokens_seen": 1949274656, "step": 780, "total_flos": 1.297523316772307e+18, "train_loss": 2.327354235526843, "train_runtime": 577.2472, "train_samples_per_second": 692.944, "train_steps_per_second": 1.351 } ], "logging_steps": 100, "max_steps": 780, "num_input_tokens_seen": 1949274656, "num_train_epochs": 5, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 1.297523316772307e+18, "train_batch_size": 32, "trial_name": null, "trial_params": null }