diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,73876 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 3.15357931251971, + "eval_steps": 187, + "global_step": 10000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.000315357931251971, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.1179, + "step": 1 + }, + { + "epoch": 0.000315357931251971, + "eval_loss": 2.6460962295532227, + "eval_runtime": 148.0893, + "eval_samples_per_second": 6.753, + "eval_steps_per_second": 6.753, + "step": 1 + }, + { + "epoch": 0.000315357931251971, + "mmlu_eval_accuracy": 0.6142877929094968, + "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182, + "mmlu_eval_accuracy_anatomy": 0.42857142857142855, + "mmlu_eval_accuracy_astronomy": 0.6875, + "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, + "mmlu_eval_accuracy_clinical_knowledge": 0.6896551724137931, + "mmlu_eval_accuracy_college_biology": 0.6875, + "mmlu_eval_accuracy_college_chemistry": 0.5, + "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, + "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727, + "mmlu_eval_accuracy_college_medicine": 0.7272727272727273, + "mmlu_eval_accuracy_college_physics": 0.7272727272727273, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.6153846153846154, + "mmlu_eval_accuracy_econometrics": 0.5, + "mmlu_eval_accuracy_electrical_engineering": 0.5625, + "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244, + "mmlu_eval_accuracy_formal_logic": 0.5, + "mmlu_eval_accuracy_global_facts": 0.3, + "mmlu_eval_accuracy_high_school_biology": 0.625, + "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.8888888888888888, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8571428571428571, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093, + "mmlu_eval_accuracy_high_school_mathematics": 0.3448275862068966, + "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, + "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354, + "mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333, + "mmlu_eval_accuracy_high_school_statistics": 0.4782608695652174, + "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, + "mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693, + "mmlu_eval_accuracy_human_aging": 0.6086956521739131, + "mmlu_eval_accuracy_human_sexuality": 0.6666666666666666, + "mmlu_eval_accuracy_international_law": 1.0, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.36363636363636365, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.8, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7906976744186046, + "mmlu_eval_accuracy_moral_disputes": 0.6052631578947368, + "mmlu_eval_accuracy_moral_scenarios": 0.34, + "mmlu_eval_accuracy_nutrition": 0.6060606060606061, + "mmlu_eval_accuracy_philosophy": 0.7352941176470589, + "mmlu_eval_accuracy_prehistory": 0.5428571428571428, + "mmlu_eval_accuracy_professional_accounting": 0.3225806451612903, + "mmlu_eval_accuracy_professional_law": 0.4470588235294118, + "mmlu_eval_accuracy_professional_medicine": 0.6774193548387096, + "mmlu_eval_accuracy_professional_psychology": 0.6521739130434783, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7407407407407407, + "mmlu_eval_accuracy_sociology": 0.9090909090909091, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.5, + "mmlu_eval_accuracy_world_religions": 0.8947368421052632, + "mmlu_loss": 7.409840339309324, + "step": 1 + }, + { + "epoch": 0.000630715862503942, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.1389, + "step": 2 + }, + { + "epoch": 0.000946073793755913, + "grad_norm": 0.1181640625, + "learning_rate": 0.0002, + "loss": 1.1834, + "step": 3 + }, + { + "epoch": 0.001261431725007884, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.2343, + "step": 4 + }, + { + "epoch": 0.001576789656259855, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.1631, + "step": 5 + }, + { + "epoch": 0.001892147587511826, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.2841, + "step": 6 + }, + { + "epoch": 0.002207505518763797, + "grad_norm": 0.150390625, + "learning_rate": 0.0002, + "loss": 1.2863, + "step": 7 + }, + { + "epoch": 0.002522863450015768, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.2222, + "step": 8 + }, + { + "epoch": 0.002838221381267739, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.1678, + "step": 9 + }, + { + "epoch": 0.00315357931251971, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 0.9816, + "step": 10 + }, + { + "epoch": 0.003468937243771681, + "grad_norm": 0.1357421875, + "learning_rate": 0.0002, + "loss": 1.0717, + "step": 11 + }, + { + "epoch": 0.003784295175023652, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.1686, + "step": 12 + }, + { + "epoch": 0.004099653106275623, + "grad_norm": 0.1240234375, + "learning_rate": 0.0002, + "loss": 1.1416, + "step": 13 + }, + { + "epoch": 0.004415011037527594, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.2273, + "step": 14 + }, + { + "epoch": 0.004730368968779565, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.2419, + "step": 15 + }, + { + "epoch": 0.005045726900031536, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 16 + }, + { + "epoch": 0.005361084831283507, + "grad_norm": 0.17578125, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 17 + }, + { + "epoch": 0.005676442762535478, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 18 + }, + { + "epoch": 0.005991800693787449, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 19 + }, + { + "epoch": 0.00630715862503942, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.2758, + "step": 20 + }, + { + "epoch": 0.006622516556291391, + "grad_norm": 0.1865234375, + "learning_rate": 0.0002, + "loss": 1.3935, + "step": 21 + }, + { + "epoch": 0.006937874487543362, + "grad_norm": 0.1865234375, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 22 + }, + { + "epoch": 0.007253232418795333, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.3766, + "step": 23 + }, + { + "epoch": 0.007568590350047304, + "grad_norm": 0.189453125, + "learning_rate": 0.0002, + "loss": 1.2477, + "step": 24 + }, + { + "epoch": 0.007883948281299276, + "grad_norm": 0.255859375, + "learning_rate": 0.0002, + "loss": 1.4829, + "step": 25 + }, + { + "epoch": 0.008199306212551246, + "grad_norm": 0.40234375, + "learning_rate": 0.0002, + "loss": 1.607, + "step": 26 + }, + { + "epoch": 0.008514664143803218, + "grad_norm": 0.416015625, + "learning_rate": 0.0002, + "loss": 1.8145, + "step": 27 + }, + { + "epoch": 0.008830022075055188, + "grad_norm": 0.51953125, + "learning_rate": 0.0002, + "loss": 1.7999, + "step": 28 + }, + { + "epoch": 0.00914538000630716, + "grad_norm": 0.392578125, + "learning_rate": 0.0002, + "loss": 1.8201, + "step": 29 + }, + { + "epoch": 0.00946073793755913, + "grad_norm": 0.392578125, + "learning_rate": 0.0002, + "loss": 1.8245, + "step": 30 + }, + { + "epoch": 0.009776095868811101, + "grad_norm": 0.9921875, + "learning_rate": 0.0002, + "loss": 2.3066, + "step": 31 + }, + { + "epoch": 0.010091453800063072, + "grad_norm": 0.396484375, + "learning_rate": 0.0002, + "loss": 2.0342, + "step": 32 + }, + { + "epoch": 0.010406811731315043, + "grad_norm": 0.498046875, + "learning_rate": 0.0002, + "loss": 2.0077, + "step": 33 + }, + { + "epoch": 0.010722169662567014, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 2.2356, + "step": 34 + }, + { + "epoch": 0.011037527593818985, + "grad_norm": 0.53515625, + "learning_rate": 0.0002, + "loss": 2.1424, + "step": 35 + }, + { + "epoch": 0.011352885525070956, + "grad_norm": 0.6484375, + "learning_rate": 0.0002, + "loss": 2.2481, + "step": 36 + }, + { + "epoch": 0.011668243456322927, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 2.1016, + "step": 37 + }, + { + "epoch": 0.011983601387574897, + "grad_norm": 0.8046875, + "learning_rate": 0.0002, + "loss": 2.9898, + "step": 38 + }, + { + "epoch": 0.01229895931882687, + "grad_norm": 1.21875, + "learning_rate": 0.0002, + "loss": 2.8473, + "step": 39 + }, + { + "epoch": 0.01261431725007884, + "grad_norm": 0.74609375, + "learning_rate": 0.0002, + "loss": 2.4889, + "step": 40 + }, + { + "epoch": 0.012929675181330811, + "grad_norm": 0.7265625, + "learning_rate": 0.0002, + "loss": 2.2115, + "step": 41 + }, + { + "epoch": 0.013245033112582781, + "grad_norm": 0.7734375, + "learning_rate": 0.0002, + "loss": 2.7112, + "step": 42 + }, + { + "epoch": 0.013560391043834753, + "grad_norm": 0.9765625, + "learning_rate": 0.0002, + "loss": 2.6663, + "step": 43 + }, + { + "epoch": 0.013875748975086723, + "grad_norm": 0.625, + "learning_rate": 0.0002, + "loss": 2.5093, + "step": 44 + }, + { + "epoch": 0.014191106906338695, + "grad_norm": 0.90625, + "learning_rate": 0.0002, + "loss": 2.4377, + "step": 45 + }, + { + "epoch": 0.014506464837590665, + "grad_norm": 0.9921875, + "learning_rate": 0.0002, + "loss": 2.488, + "step": 46 + }, + { + "epoch": 0.014821822768842637, + "grad_norm": 0.7578125, + "learning_rate": 0.0002, + "loss": 2.4603, + "step": 47 + }, + { + "epoch": 0.015137180700094607, + "grad_norm": 1.2734375, + "learning_rate": 0.0002, + "loss": 2.7226, + "step": 48 + }, + { + "epoch": 0.01545253863134658, + "grad_norm": 0.90625, + "learning_rate": 0.0002, + "loss": 2.3431, + "step": 49 + }, + { + "epoch": 0.01576789656259855, + "grad_norm": 1.0546875, + "learning_rate": 0.0002, + "loss": 2.9972, + "step": 50 + }, + { + "epoch": 0.01608325449385052, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 1.2168, + "step": 51 + }, + { + "epoch": 0.01639861242510249, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 1.3789, + "step": 52 + }, + { + "epoch": 0.01671397035635446, + "grad_norm": 0.193359375, + "learning_rate": 0.0002, + "loss": 1.1127, + "step": 53 + }, + { + "epoch": 0.017029328287606435, + "grad_norm": 0.17578125, + "learning_rate": 0.0002, + "loss": 1.1272, + "step": 54 + }, + { + "epoch": 0.017344686218858405, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.133, + "step": 55 + }, + { + "epoch": 0.017660044150110375, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.2203, + "step": 56 + }, + { + "epoch": 0.017975402081362345, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 0.9945, + "step": 57 + }, + { + "epoch": 0.01829076001261432, + "grad_norm": 0.138671875, + "learning_rate": 0.0002, + "loss": 1.1217, + "step": 58 + }, + { + "epoch": 0.01860611794386629, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.2388, + "step": 59 + }, + { + "epoch": 0.01892147587511826, + "grad_norm": 0.1552734375, + "learning_rate": 0.0002, + "loss": 1.1047, + "step": 60 + }, + { + "epoch": 0.01923683380637023, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.376, + "step": 61 + }, + { + "epoch": 0.019552191737622203, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.1604, + "step": 62 + }, + { + "epoch": 0.019867549668874173, + "grad_norm": 0.16015625, + "learning_rate": 0.0002, + "loss": 1.2511, + "step": 63 + }, + { + "epoch": 0.020182907600126143, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.0817, + "step": 64 + }, + { + "epoch": 0.020498265531378113, + "grad_norm": 0.1494140625, + "learning_rate": 0.0002, + "loss": 1.0093, + "step": 65 + }, + { + "epoch": 0.020813623462630087, + "grad_norm": 0.177734375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 66 + }, + { + "epoch": 0.021128981393882057, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.437, + "step": 67 + }, + { + "epoch": 0.021444339325134027, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.121, + "step": 68 + }, + { + "epoch": 0.021759697256385997, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 1.4741, + "step": 69 + }, + { + "epoch": 0.02207505518763797, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.2632, + "step": 70 + }, + { + "epoch": 0.02239041311888994, + "grad_norm": 0.193359375, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 71 + }, + { + "epoch": 0.02270577105014191, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.2291, + "step": 72 + }, + { + "epoch": 0.02302112898139388, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 1.4129, + "step": 73 + }, + { + "epoch": 0.023336486912645855, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 1.5363, + "step": 74 + }, + { + "epoch": 0.023651844843897825, + "grad_norm": 0.208984375, + "learning_rate": 0.0002, + "loss": 1.4946, + "step": 75 + }, + { + "epoch": 0.023967202775149795, + "grad_norm": 0.5546875, + "learning_rate": 0.0002, + "loss": 1.6625, + "step": 76 + }, + { + "epoch": 0.024282560706401765, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 1.4696, + "step": 77 + }, + { + "epoch": 0.02459791863765374, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 1.4485, + "step": 78 + }, + { + "epoch": 0.02491327656890571, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 1.4319, + "step": 79 + }, + { + "epoch": 0.02522863450015768, + "grad_norm": 0.41015625, + "learning_rate": 0.0002, + "loss": 1.7743, + "step": 80 + }, + { + "epoch": 0.02554399243140965, + "grad_norm": 0.427734375, + "learning_rate": 0.0002, + "loss": 1.8532, + "step": 81 + }, + { + "epoch": 0.025859350362661623, + "grad_norm": 0.396484375, + "learning_rate": 0.0002, + "loss": 1.6602, + "step": 82 + }, + { + "epoch": 0.026174708293913593, + "grad_norm": 0.45703125, + "learning_rate": 0.0002, + "loss": 2.0508, + "step": 83 + }, + { + "epoch": 0.026490066225165563, + "grad_norm": 0.46484375, + "learning_rate": 0.0002, + "loss": 2.044, + "step": 84 + }, + { + "epoch": 0.026805424156417533, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 2.3228, + "step": 85 + }, + { + "epoch": 0.027120782087669507, + "grad_norm": 0.515625, + "learning_rate": 0.0002, + "loss": 1.8849, + "step": 86 + }, + { + "epoch": 0.027436140018921477, + "grad_norm": 0.51953125, + "learning_rate": 0.0002, + "loss": 2.1106, + "step": 87 + }, + { + "epoch": 0.027751497950173447, + "grad_norm": 0.5390625, + "learning_rate": 0.0002, + "loss": 1.9036, + "step": 88 + }, + { + "epoch": 0.028066855881425417, + "grad_norm": 0.72265625, + "learning_rate": 0.0002, + "loss": 2.5189, + "step": 89 + }, + { + "epoch": 0.02838221381267739, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 2.2565, + "step": 90 + }, + { + "epoch": 0.02869757174392936, + "grad_norm": 0.734375, + "learning_rate": 0.0002, + "loss": 2.3621, + "step": 91 + }, + { + "epoch": 0.02901292967518133, + "grad_norm": 0.83203125, + "learning_rate": 0.0002, + "loss": 2.7675, + "step": 92 + }, + { + "epoch": 0.0293282876064333, + "grad_norm": 1.0, + "learning_rate": 0.0002, + "loss": 2.7243, + "step": 93 + }, + { + "epoch": 0.029643645537685274, + "grad_norm": 1.2265625, + "learning_rate": 0.0002, + "loss": 2.7591, + "step": 94 + }, + { + "epoch": 0.029959003468937245, + "grad_norm": 0.8515625, + "learning_rate": 0.0002, + "loss": 2.2846, + "step": 95 + }, + { + "epoch": 0.030274361400189215, + "grad_norm": 0.96484375, + "learning_rate": 0.0002, + "loss": 2.6086, + "step": 96 + }, + { + "epoch": 0.030589719331441185, + "grad_norm": 0.69921875, + "learning_rate": 0.0002, + "loss": 2.1553, + "step": 97 + }, + { + "epoch": 0.03090507726269316, + "grad_norm": 0.7890625, + "learning_rate": 0.0002, + "loss": 2.4491, + "step": 98 + }, + { + "epoch": 0.03122043519394513, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 2.0883, + "step": 99 + }, + { + "epoch": 0.0315357931251971, + "grad_norm": 1.796875, + "learning_rate": 0.0002, + "loss": 2.5, + "step": 100 + }, + { + "epoch": 0.03185115105644907, + "grad_norm": 0.408203125, + "learning_rate": 0.0002, + "loss": 1.209, + "step": 101 + }, + { + "epoch": 0.03216650898770104, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 102 + }, + { + "epoch": 0.03248186691895301, + "grad_norm": 0.2373046875, + "learning_rate": 0.0002, + "loss": 1.258, + "step": 103 + }, + { + "epoch": 0.03279722485020498, + "grad_norm": 0.1494140625, + "learning_rate": 0.0002, + "loss": 1.2613, + "step": 104 + }, + { + "epoch": 0.033112582781456956, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.1535, + "step": 105 + }, + { + "epoch": 0.03342794071270892, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.1236, + "step": 106 + }, + { + "epoch": 0.033743298643960896, + "grad_norm": 0.138671875, + "learning_rate": 0.0002, + "loss": 1.1012, + "step": 107 + }, + { + "epoch": 0.03405865657521287, + "grad_norm": 0.150390625, + "learning_rate": 0.0002, + "loss": 1.1971, + "step": 108 + }, + { + "epoch": 0.03437401450646484, + "grad_norm": 0.1513671875, + "learning_rate": 0.0002, + "loss": 1.1603, + "step": 109 + }, + { + "epoch": 0.03468937243771681, + "grad_norm": 0.1640625, + "learning_rate": 0.0002, + "loss": 1.2024, + "step": 110 + }, + { + "epoch": 0.03500473036896878, + "grad_norm": 0.138671875, + "learning_rate": 0.0002, + "loss": 1.1149, + "step": 111 + }, + { + "epoch": 0.03532008830022075, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.133, + "step": 112 + }, + { + "epoch": 0.035635446231472724, + "grad_norm": 0.1455078125, + "learning_rate": 0.0002, + "loss": 1.2023, + "step": 113 + }, + { + "epoch": 0.03595080416272469, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.2423, + "step": 114 + }, + { + "epoch": 0.036266162093976664, + "grad_norm": 0.1640625, + "learning_rate": 0.0002, + "loss": 1.359, + "step": 115 + }, + { + "epoch": 0.03658152002522864, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.2799, + "step": 116 + }, + { + "epoch": 0.036896877956480605, + "grad_norm": 0.166015625, + "learning_rate": 0.0002, + "loss": 1.282, + "step": 117 + }, + { + "epoch": 0.03721223588773258, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.2249, + "step": 118 + }, + { + "epoch": 0.037527593818984545, + "grad_norm": 0.177734375, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 119 + }, + { + "epoch": 0.03784295175023652, + "grad_norm": 0.177734375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 120 + }, + { + "epoch": 0.03815830968148849, + "grad_norm": 0.1796875, + "learning_rate": 0.0002, + "loss": 1.15, + "step": 121 + }, + { + "epoch": 0.03847366761274046, + "grad_norm": 0.2119140625, + "learning_rate": 0.0002, + "loss": 1.4206, + "step": 122 + }, + { + "epoch": 0.03878902554399243, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 1.3758, + "step": 123 + }, + { + "epoch": 0.039104383475244406, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 1.6466, + "step": 124 + }, + { + "epoch": 0.03941974140649637, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 1.4051, + "step": 125 + }, + { + "epoch": 0.039735099337748346, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.3627, + "step": 126 + }, + { + "epoch": 0.04005045726900031, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.4157, + "step": 127 + }, + { + "epoch": 0.040365815200252286, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.4072, + "step": 128 + }, + { + "epoch": 0.04068117313150426, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 1.574, + "step": 129 + }, + { + "epoch": 0.04099653106275623, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.5011, + "step": 130 + }, + { + "epoch": 0.0413118889940082, + "grad_norm": 0.87890625, + "learning_rate": 0.0002, + "loss": 1.8846, + "step": 131 + }, + { + "epoch": 0.041627246925260174, + "grad_norm": 0.45703125, + "learning_rate": 0.0002, + "loss": 1.8736, + "step": 132 + }, + { + "epoch": 0.04194260485651214, + "grad_norm": 0.458984375, + "learning_rate": 0.0002, + "loss": 1.8582, + "step": 133 + }, + { + "epoch": 0.042257962787764114, + "grad_norm": 1.2890625, + "learning_rate": 0.0002, + "loss": 2.2945, + "step": 134 + }, + { + "epoch": 0.04257332071901608, + "grad_norm": 0.63671875, + "learning_rate": 0.0002, + "loss": 2.1367, + "step": 135 + }, + { + "epoch": 0.042888678650268054, + "grad_norm": 0.4765625, + "learning_rate": 0.0002, + "loss": 1.9376, + "step": 136 + }, + { + "epoch": 0.04320403658152003, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 2.0184, + "step": 137 + }, + { + "epoch": 0.043519394512771994, + "grad_norm": 0.46484375, + "learning_rate": 0.0002, + "loss": 2.0211, + "step": 138 + }, + { + "epoch": 0.04383475244402397, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 2.2129, + "step": 139 + }, + { + "epoch": 0.04415011037527594, + "grad_norm": 0.703125, + "learning_rate": 0.0002, + "loss": 2.3542, + "step": 140 + }, + { + "epoch": 0.04446546830652791, + "grad_norm": 0.6796875, + "learning_rate": 0.0002, + "loss": 2.1137, + "step": 141 + }, + { + "epoch": 0.04478082623777988, + "grad_norm": 0.83984375, + "learning_rate": 0.0002, + "loss": 2.7357, + "step": 142 + }, + { + "epoch": 0.04509618416903185, + "grad_norm": 0.8359375, + "learning_rate": 0.0002, + "loss": 2.7374, + "step": 143 + }, + { + "epoch": 0.04541154210028382, + "grad_norm": 1.03125, + "learning_rate": 0.0002, + "loss": 3.4736, + "step": 144 + }, + { + "epoch": 0.045726900031535796, + "grad_norm": 0.7734375, + "learning_rate": 0.0002, + "loss": 2.4158, + "step": 145 + }, + { + "epoch": 0.04604225796278776, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 2.3138, + "step": 146 + }, + { + "epoch": 0.046357615894039736, + "grad_norm": 1.234375, + "learning_rate": 0.0002, + "loss": 3.3427, + "step": 147 + }, + { + "epoch": 0.04667297382529171, + "grad_norm": 0.96875, + "learning_rate": 0.0002, + "loss": 2.6367, + "step": 148 + }, + { + "epoch": 0.046988331756543676, + "grad_norm": 0.6640625, + "learning_rate": 0.0002, + "loss": 2.2514, + "step": 149 + }, + { + "epoch": 0.04730368968779565, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 2.3439, + "step": 150 + }, + { + "epoch": 0.047619047619047616, + "grad_norm": 0.416015625, + "learning_rate": 0.0002, + "loss": 1.1685, + "step": 151 + }, + { + "epoch": 0.04793440555029959, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 1.154, + "step": 152 + }, + { + "epoch": 0.048249763481551564, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 1.2132, + "step": 153 + }, + { + "epoch": 0.04856512141280353, + "grad_norm": 0.158203125, + "learning_rate": 0.0002, + "loss": 1.0973, + "step": 154 + }, + { + "epoch": 0.048880479344055504, + "grad_norm": 0.1552734375, + "learning_rate": 0.0002, + "loss": 1.1699, + "step": 155 + }, + { + "epoch": 0.04919583727530748, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.1844, + "step": 156 + }, + { + "epoch": 0.049511195206559444, + "grad_norm": 0.1455078125, + "learning_rate": 0.0002, + "loss": 1.2201, + "step": 157 + }, + { + "epoch": 0.04982655313781142, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.0949, + "step": 158 + }, + { + "epoch": 0.050141911069063384, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.2012, + "step": 159 + }, + { + "epoch": 0.05045726900031536, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.1745, + "step": 160 + }, + { + "epoch": 0.05077262693156733, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.135, + "step": 161 + }, + { + "epoch": 0.0510879848628193, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.1535, + "step": 162 + }, + { + "epoch": 0.05140334279407127, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 1.432, + "step": 163 + }, + { + "epoch": 0.051718700725323245, + "grad_norm": 0.166015625, + "learning_rate": 0.0002, + "loss": 1.2511, + "step": 164 + }, + { + "epoch": 0.05203405865657521, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.2163, + "step": 165 + }, + { + "epoch": 0.052349416587827186, + "grad_norm": 0.1767578125, + "learning_rate": 0.0002, + "loss": 1.1748, + "step": 166 + }, + { + "epoch": 0.05266477451907915, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.2973, + "step": 167 + }, + { + "epoch": 0.052980132450331126, + "grad_norm": 0.166015625, + "learning_rate": 0.0002, + "loss": 1.1683, + "step": 168 + }, + { + "epoch": 0.0532954903815831, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.0487, + "step": 169 + }, + { + "epoch": 0.053610848312835066, + "grad_norm": 0.1884765625, + "learning_rate": 0.0002, + "loss": 1.0222, + "step": 170 + }, + { + "epoch": 0.05392620624408704, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 171 + }, + { + "epoch": 0.05424156417533901, + "grad_norm": 0.1943359375, + "learning_rate": 0.0002, + "loss": 1.1896, + "step": 172 + }, + { + "epoch": 0.05455692210659098, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.2838, + "step": 173 + }, + { + "epoch": 0.05487228003784295, + "grad_norm": 0.2314453125, + "learning_rate": 0.0002, + "loss": 1.4516, + "step": 174 + }, + { + "epoch": 0.05518763796909492, + "grad_norm": 0.2412109375, + "learning_rate": 0.0002, + "loss": 1.3882, + "step": 175 + }, + { + "epoch": 0.055502995900346894, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 1.379, + "step": 176 + }, + { + "epoch": 0.05581835383159887, + "grad_norm": 0.76953125, + "learning_rate": 0.0002, + "loss": 1.761, + "step": 177 + }, + { + "epoch": 0.056133711762850834, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.642, + "step": 178 + }, + { + "epoch": 0.05644906969410281, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 1.6247, + "step": 179 + }, + { + "epoch": 0.05676442762535478, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 1.7785, + "step": 180 + }, + { + "epoch": 0.05707978555660675, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 1.9513, + "step": 181 + }, + { + "epoch": 0.05739514348785872, + "grad_norm": 0.37109375, + "learning_rate": 0.0002, + "loss": 1.5212, + "step": 182 + }, + { + "epoch": 0.05771050141911069, + "grad_norm": 0.408203125, + "learning_rate": 0.0002, + "loss": 1.7207, + "step": 183 + }, + { + "epoch": 0.05802585935036266, + "grad_norm": 0.4609375, + "learning_rate": 0.0002, + "loss": 2.0351, + "step": 184 + }, + { + "epoch": 0.058341217281614635, + "grad_norm": 0.640625, + "learning_rate": 0.0002, + "loss": 2.4458, + "step": 185 + }, + { + "epoch": 0.0586565752128666, + "grad_norm": 0.578125, + "learning_rate": 0.0002, + "loss": 2.1338, + "step": 186 + }, + { + "epoch": 0.058971933144118575, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 2.1519, + "step": 187 + }, + { + "epoch": 0.058971933144118575, + "eval_loss": 1.7583619356155396, + "eval_runtime": 148.5095, + "eval_samples_per_second": 6.734, + "eval_steps_per_second": 6.734, + "step": 187 + }, + { + "epoch": 0.058971933144118575, + "mmlu_eval_accuracy": 0.6091000989737743, + "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182, + "mmlu_eval_accuracy_anatomy": 0.5, + "mmlu_eval_accuracy_astronomy": 0.6875, + "mmlu_eval_accuracy_business_ethics": 0.45454545454545453, + "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138, + "mmlu_eval_accuracy_college_biology": 0.625, + "mmlu_eval_accuracy_college_chemistry": 0.375, + "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, + "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365, + "mmlu_eval_accuracy_college_medicine": 0.6363636363636364, + "mmlu_eval_accuracy_college_physics": 0.7272727272727273, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.6153846153846154, + "mmlu_eval_accuracy_econometrics": 0.5, + "mmlu_eval_accuracy_electrical_engineering": 0.5625, + "mmlu_eval_accuracy_elementary_mathematics": 0.4634146341463415, + "mmlu_eval_accuracy_formal_logic": 0.42857142857142855, + "mmlu_eval_accuracy_global_facts": 0.3, + "mmlu_eval_accuracy_high_school_biology": 0.625, + "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091, + "mmlu_eval_accuracy_high_school_computer_science": 0.7777777777777778, + "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.627906976744186, + "mmlu_eval_accuracy_high_school_mathematics": 0.3103448275862069, + "mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923, + "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, + "mmlu_eval_accuracy_high_school_psychology": 0.9, + "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, + "mmlu_eval_accuracy_high_school_us_history": 0.8636363636363636, + "mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693, + "mmlu_eval_accuracy_human_aging": 0.6521739130434783, + "mmlu_eval_accuracy_human_sexuality": 0.6666666666666666, + "mmlu_eval_accuracy_international_law": 0.9230769230769231, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.36363636363636365, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.8, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7790697674418605, + "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, + "mmlu_eval_accuracy_moral_scenarios": 0.4, + "mmlu_eval_accuracy_nutrition": 0.6666666666666666, + "mmlu_eval_accuracy_philosophy": 0.7352941176470589, + "mmlu_eval_accuracy_prehistory": 0.5714285714285714, + "mmlu_eval_accuracy_professional_accounting": 0.3225806451612903, + "mmlu_eval_accuracy_professional_law": 0.43529411764705883, + "mmlu_eval_accuracy_professional_medicine": 0.6774193548387096, + "mmlu_eval_accuracy_professional_psychology": 0.6666666666666666, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7037037037037037, + "mmlu_eval_accuracy_sociology": 0.9545454545454546, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.5, + "mmlu_eval_accuracy_world_religions": 0.8947368421052632, + "mmlu_loss": 5.9885171454845105, + "step": 187 + }, + { + "epoch": 0.05928729107537055, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 2.2734, + "step": 188 + }, + { + "epoch": 0.059602649006622516, + "grad_norm": 0.703125, + "learning_rate": 0.0002, + "loss": 2.4382, + "step": 189 + }, + { + "epoch": 0.05991800693787449, + "grad_norm": 0.7265625, + "learning_rate": 0.0002, + "loss": 2.5062, + "step": 190 + }, + { + "epoch": 0.060233364869126456, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 2.4622, + "step": 191 + }, + { + "epoch": 0.06054872280037843, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 2.1925, + "step": 192 + }, + { + "epoch": 0.0608640807316304, + "grad_norm": 0.7109375, + "learning_rate": 0.0002, + "loss": 2.3531, + "step": 193 + }, + { + "epoch": 0.06117943866288237, + "grad_norm": 0.73828125, + "learning_rate": 0.0002, + "loss": 2.6286, + "step": 194 + }, + { + "epoch": 0.06149479659413434, + "grad_norm": 0.6875, + "learning_rate": 0.0002, + "loss": 2.3818, + "step": 195 + }, + { + "epoch": 0.06181015452538632, + "grad_norm": 0.71875, + "learning_rate": 0.0002, + "loss": 2.375, + "step": 196 + }, + { + "epoch": 0.062125512456638284, + "grad_norm": 0.9296875, + "learning_rate": 0.0002, + "loss": 2.7019, + "step": 197 + }, + { + "epoch": 0.06244087038789026, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 1.9829, + "step": 198 + }, + { + "epoch": 0.06275622831914222, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 2.2847, + "step": 199 + }, + { + "epoch": 0.0630715862503942, + "grad_norm": 0.8984375, + "learning_rate": 0.0002, + "loss": 2.1117, + "step": 200 + }, + { + "epoch": 0.06338694418164617, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 1.099, + "step": 201 + }, + { + "epoch": 0.06370230211289814, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 1.1376, + "step": 202 + }, + { + "epoch": 0.0640176600441501, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.2963, + "step": 203 + }, + { + "epoch": 0.06433301797540208, + "grad_norm": 0.1689453125, + "learning_rate": 0.0002, + "loss": 1.1675, + "step": 204 + }, + { + "epoch": 0.06464837590665405, + "grad_norm": 0.1513671875, + "learning_rate": 0.0002, + "loss": 1.1398, + "step": 205 + }, + { + "epoch": 0.06496373383790602, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.1689, + "step": 206 + }, + { + "epoch": 0.065279091769158, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.179, + "step": 207 + }, + { + "epoch": 0.06559444970040997, + "grad_norm": 0.138671875, + "learning_rate": 0.0002, + "loss": 1.094, + "step": 208 + }, + { + "epoch": 0.06590980763166193, + "grad_norm": 0.158203125, + "learning_rate": 0.0002, + "loss": 1.0376, + "step": 209 + }, + { + "epoch": 0.06622516556291391, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 1.2896, + "step": 210 + }, + { + "epoch": 0.06654052349416588, + "grad_norm": 0.16015625, + "learning_rate": 0.0002, + "loss": 1.1933, + "step": 211 + }, + { + "epoch": 0.06685588142541785, + "grad_norm": 0.1474609375, + "learning_rate": 0.0002, + "loss": 1.1329, + "step": 212 + }, + { + "epoch": 0.06717123935666983, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.1273, + "step": 213 + }, + { + "epoch": 0.06748659728792179, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.2095, + "step": 214 + }, + { + "epoch": 0.06780195521917376, + "grad_norm": 0.189453125, + "learning_rate": 0.0002, + "loss": 1.2426, + "step": 215 + }, + { + "epoch": 0.06811731315042574, + "grad_norm": 0.18359375, + "learning_rate": 0.0002, + "loss": 1.1932, + "step": 216 + }, + { + "epoch": 0.0684326710816777, + "grad_norm": 0.169921875, + "learning_rate": 0.0002, + "loss": 1.2119, + "step": 217 + }, + { + "epoch": 0.06874802901292967, + "grad_norm": 0.2314453125, + "learning_rate": 0.0002, + "loss": 1.2809, + "step": 218 + }, + { + "epoch": 0.06906338694418164, + "grad_norm": 0.1962890625, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 219 + }, + { + "epoch": 0.06937874487543362, + "grad_norm": 0.2216796875, + "learning_rate": 0.0002, + "loss": 1.2066, + "step": 220 + }, + { + "epoch": 0.06969410280668559, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 1.2451, + "step": 221 + }, + { + "epoch": 0.07000946073793755, + "grad_norm": 0.423828125, + "learning_rate": 0.0002, + "loss": 1.3951, + "step": 222 + }, + { + "epoch": 0.07032481866918953, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 1.2939, + "step": 223 + }, + { + "epoch": 0.0706401766004415, + "grad_norm": 0.2470703125, + "learning_rate": 0.0002, + "loss": 1.5152, + "step": 224 + }, + { + "epoch": 0.07095553453169347, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.5395, + "step": 225 + }, + { + "epoch": 0.07127089246294545, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 1.6232, + "step": 226 + }, + { + "epoch": 0.07158625039419741, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.1741, + "step": 227 + }, + { + "epoch": 0.07190160832544938, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.4193, + "step": 228 + }, + { + "epoch": 0.07221696625670136, + "grad_norm": 0.353515625, + "learning_rate": 0.0002, + "loss": 1.6961, + "step": 229 + }, + { + "epoch": 0.07253232418795333, + "grad_norm": 0.408203125, + "learning_rate": 0.0002, + "loss": 1.5422, + "step": 230 + }, + { + "epoch": 0.0728476821192053, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 1.616, + "step": 231 + }, + { + "epoch": 0.07316304005045728, + "grad_norm": 0.455078125, + "learning_rate": 0.0002, + "loss": 1.9557, + "step": 232 + }, + { + "epoch": 0.07347839798170924, + "grad_norm": 0.39453125, + "learning_rate": 0.0002, + "loss": 1.654, + "step": 233 + }, + { + "epoch": 0.07379375591296121, + "grad_norm": 0.458984375, + "learning_rate": 0.0002, + "loss": 2.1714, + "step": 234 + }, + { + "epoch": 0.07410911384421318, + "grad_norm": 0.431640625, + "learning_rate": 0.0002, + "loss": 1.8316, + "step": 235 + }, + { + "epoch": 0.07442447177546516, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 2.1725, + "step": 236 + }, + { + "epoch": 0.07473982970671712, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 2.4451, + "step": 237 + }, + { + "epoch": 0.07505518763796909, + "grad_norm": 0.625, + "learning_rate": 0.0002, + "loss": 2.0518, + "step": 238 + }, + { + "epoch": 0.07537054556922107, + "grad_norm": 0.765625, + "learning_rate": 0.0002, + "loss": 2.7386, + "step": 239 + }, + { + "epoch": 0.07568590350047304, + "grad_norm": 1.125, + "learning_rate": 0.0002, + "loss": 2.5739, + "step": 240 + }, + { + "epoch": 0.076001261431725, + "grad_norm": 0.6875, + "learning_rate": 0.0002, + "loss": 2.5645, + "step": 241 + }, + { + "epoch": 0.07631661936297698, + "grad_norm": 0.68359375, + "learning_rate": 0.0002, + "loss": 2.3116, + "step": 242 + }, + { + "epoch": 0.07663197729422895, + "grad_norm": 0.8515625, + "learning_rate": 0.0002, + "loss": 3.1317, + "step": 243 + }, + { + "epoch": 0.07694733522548092, + "grad_norm": 0.7265625, + "learning_rate": 0.0002, + "loss": 2.5665, + "step": 244 + }, + { + "epoch": 0.0772626931567329, + "grad_norm": 0.7890625, + "learning_rate": 0.0002, + "loss": 2.6618, + "step": 245 + }, + { + "epoch": 0.07757805108798486, + "grad_norm": 0.8359375, + "learning_rate": 0.0002, + "loss": 2.7445, + "step": 246 + }, + { + "epoch": 0.07789340901923683, + "grad_norm": 0.9296875, + "learning_rate": 0.0002, + "loss": 3.012, + "step": 247 + }, + { + "epoch": 0.07820876695048881, + "grad_norm": 0.73046875, + "learning_rate": 0.0002, + "loss": 2.5823, + "step": 248 + }, + { + "epoch": 0.07852412488174078, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 1.7542, + "step": 249 + }, + { + "epoch": 0.07883948281299274, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 2.2869, + "step": 250 + }, + { + "epoch": 0.07915484074424471, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 1.2814, + "step": 251 + }, + { + "epoch": 0.07947019867549669, + "grad_norm": 0.248046875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 252 + }, + { + "epoch": 0.07978555660674866, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 1.1811, + "step": 253 + }, + { + "epoch": 0.08010091453800063, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.1484, + "step": 254 + }, + { + "epoch": 0.0804162724692526, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.1784, + "step": 255 + }, + { + "epoch": 0.08073163040050457, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.2706, + "step": 256 + }, + { + "epoch": 0.08104698833175654, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.2432, + "step": 257 + }, + { + "epoch": 0.08136234626300852, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.1425, + "step": 258 + }, + { + "epoch": 0.08167770419426049, + "grad_norm": 0.1865234375, + "learning_rate": 0.0002, + "loss": 1.1499, + "step": 259 + }, + { + "epoch": 0.08199306212551245, + "grad_norm": 0.1337890625, + "learning_rate": 0.0002, + "loss": 1.0245, + "step": 260 + }, + { + "epoch": 0.08230842005676443, + "grad_norm": 0.166015625, + "learning_rate": 0.0002, + "loss": 1.2836, + "step": 261 + }, + { + "epoch": 0.0826237779880164, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.0242, + "step": 262 + }, + { + "epoch": 0.08293913591926837, + "grad_norm": 0.1689453125, + "learning_rate": 0.0002, + "loss": 1.1379, + "step": 263 + }, + { + "epoch": 0.08325449385052035, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.1335, + "step": 264 + }, + { + "epoch": 0.08356985178177231, + "grad_norm": 0.169921875, + "learning_rate": 0.0002, + "loss": 1.2115, + "step": 265 + }, + { + "epoch": 0.08388520971302428, + "grad_norm": 0.17578125, + "learning_rate": 0.0002, + "loss": 1.1152, + "step": 266 + }, + { + "epoch": 0.08420056764427625, + "grad_norm": 0.169921875, + "learning_rate": 0.0002, + "loss": 1.2669, + "step": 267 + }, + { + "epoch": 0.08451592557552823, + "grad_norm": 0.1767578125, + "learning_rate": 0.0002, + "loss": 1.2378, + "step": 268 + }, + { + "epoch": 0.0848312835067802, + "grad_norm": 0.158203125, + "learning_rate": 0.0002, + "loss": 1.095, + "step": 269 + }, + { + "epoch": 0.08514664143803216, + "grad_norm": 0.1884765625, + "learning_rate": 0.0002, + "loss": 1.1293, + "step": 270 + }, + { + "epoch": 0.08546199936928414, + "grad_norm": 0.208984375, + "learning_rate": 0.0002, + "loss": 1.722, + "step": 271 + }, + { + "epoch": 0.08577735730053611, + "grad_norm": 0.1875, + "learning_rate": 0.0002, + "loss": 1.2574, + "step": 272 + }, + { + "epoch": 0.08609271523178808, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 1.1467, + "step": 273 + }, + { + "epoch": 0.08640807316304006, + "grad_norm": 0.2431640625, + "learning_rate": 0.0002, + "loss": 1.6219, + "step": 274 + }, + { + "epoch": 0.08672343109429202, + "grad_norm": 0.248046875, + "learning_rate": 0.0002, + "loss": 1.4198, + "step": 275 + }, + { + "epoch": 0.08703878902554399, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 1.5756, + "step": 276 + }, + { + "epoch": 0.08735414695679597, + "grad_norm": 0.2294921875, + "learning_rate": 0.0002, + "loss": 1.3784, + "step": 277 + }, + { + "epoch": 0.08766950488804794, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 1.6183, + "step": 278 + }, + { + "epoch": 0.0879848628192999, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 1.3687, + "step": 279 + }, + { + "epoch": 0.08830022075055188, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 1.462, + "step": 280 + }, + { + "epoch": 0.08861557868180385, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 1.2247, + "step": 281 + }, + { + "epoch": 0.08893093661305582, + "grad_norm": 0.353515625, + "learning_rate": 0.0002, + "loss": 1.6736, + "step": 282 + }, + { + "epoch": 0.08924629454430778, + "grad_norm": 0.462890625, + "learning_rate": 0.0002, + "loss": 1.9971, + "step": 283 + }, + { + "epoch": 0.08956165247555976, + "grad_norm": 0.408203125, + "learning_rate": 0.0002, + "loss": 1.993, + "step": 284 + }, + { + "epoch": 0.08987701040681173, + "grad_norm": 0.482421875, + "learning_rate": 0.0002, + "loss": 2.0232, + "step": 285 + }, + { + "epoch": 0.0901923683380637, + "grad_norm": 0.416015625, + "learning_rate": 0.0002, + "loss": 1.677, + "step": 286 + }, + { + "epoch": 0.09050772626931568, + "grad_norm": 0.482421875, + "learning_rate": 0.0002, + "loss": 2.0904, + "step": 287 + }, + { + "epoch": 0.09082308420056764, + "grad_norm": 0.734375, + "learning_rate": 0.0002, + "loss": 2.5839, + "step": 288 + }, + { + "epoch": 0.09113844213181961, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 2.2263, + "step": 289 + }, + { + "epoch": 0.09145380006307159, + "grad_norm": 0.62890625, + "learning_rate": 0.0002, + "loss": 2.5586, + "step": 290 + }, + { + "epoch": 0.09176915799432356, + "grad_norm": 0.62890625, + "learning_rate": 0.0002, + "loss": 2.2638, + "step": 291 + }, + { + "epoch": 0.09208451592557552, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 1.9463, + "step": 292 + }, + { + "epoch": 0.0923998738568275, + "grad_norm": 1.4765625, + "learning_rate": 0.0002, + "loss": 3.2001, + "step": 293 + }, + { + "epoch": 0.09271523178807947, + "grad_norm": 1.609375, + "learning_rate": 0.0002, + "loss": 2.8795, + "step": 294 + }, + { + "epoch": 0.09303058971933144, + "grad_norm": 0.94140625, + "learning_rate": 0.0002, + "loss": 2.4123, + "step": 295 + }, + { + "epoch": 0.09334594765058342, + "grad_norm": 1.0859375, + "learning_rate": 0.0002, + "loss": 3.2725, + "step": 296 + }, + { + "epoch": 0.09366130558183539, + "grad_norm": 0.73046875, + "learning_rate": 0.0002, + "loss": 2.0153, + "step": 297 + }, + { + "epoch": 0.09397666351308735, + "grad_norm": 0.625, + "learning_rate": 0.0002, + "loss": 1.8583, + "step": 298 + }, + { + "epoch": 0.09429202144433932, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 2.1395, + "step": 299 + }, + { + "epoch": 0.0946073793755913, + "grad_norm": 1.65625, + "learning_rate": 0.0002, + "loss": 3.2025, + "step": 300 + }, + { + "epoch": 0.09492273730684327, + "grad_norm": 0.2421875, + "learning_rate": 0.0002, + "loss": 1.303, + "step": 301 + }, + { + "epoch": 0.09523809523809523, + "grad_norm": 0.216796875, + "learning_rate": 0.0002, + "loss": 1.1385, + "step": 302 + }, + { + "epoch": 0.09555345316934721, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 1.1925, + "step": 303 + }, + { + "epoch": 0.09586881110059918, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.1681, + "step": 304 + }, + { + "epoch": 0.09618416903185115, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.1538, + "step": 305 + }, + { + "epoch": 0.09649952696310313, + "grad_norm": 0.1787109375, + "learning_rate": 0.0002, + "loss": 1.1933, + "step": 306 + }, + { + "epoch": 0.0968148848943551, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.2118, + "step": 307 + }, + { + "epoch": 0.09713024282560706, + "grad_norm": 0.142578125, + "learning_rate": 0.0002, + "loss": 0.9842, + "step": 308 + }, + { + "epoch": 0.09744560075685904, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.1814, + "step": 309 + }, + { + "epoch": 0.09776095868811101, + "grad_norm": 0.142578125, + "learning_rate": 0.0002, + "loss": 1.0759, + "step": 310 + }, + { + "epoch": 0.09807631661936297, + "grad_norm": 0.134765625, + "learning_rate": 0.0002, + "loss": 1.1069, + "step": 311 + }, + { + "epoch": 0.09839167455061495, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.1498, + "step": 312 + }, + { + "epoch": 0.09870703248186692, + "grad_norm": 0.150390625, + "learning_rate": 0.0002, + "loss": 1.2509, + "step": 313 + }, + { + "epoch": 0.09902239041311889, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.2325, + "step": 314 + }, + { + "epoch": 0.09933774834437085, + "grad_norm": 0.1494140625, + "learning_rate": 0.0002, + "loss": 1.1934, + "step": 315 + }, + { + "epoch": 0.09965310627562284, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.1264, + "step": 316 + }, + { + "epoch": 0.0999684642068748, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.3939, + "step": 317 + }, + { + "epoch": 0.10028382213812677, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.2418, + "step": 318 + }, + { + "epoch": 0.10059918006937875, + "grad_norm": 0.1943359375, + "learning_rate": 0.0002, + "loss": 1.1956, + "step": 319 + }, + { + "epoch": 0.10091453800063072, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.5603, + "step": 320 + }, + { + "epoch": 0.10122989593188268, + "grad_norm": 0.189453125, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 321 + }, + { + "epoch": 0.10154525386313466, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 1.4091, + "step": 322 + }, + { + "epoch": 0.10186061179438663, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 1.3891, + "step": 323 + }, + { + "epoch": 0.1021759697256386, + "grad_norm": 0.2119140625, + "learning_rate": 0.0002, + "loss": 1.3763, + "step": 324 + }, + { + "epoch": 0.10249132765689058, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 1.1734, + "step": 325 + }, + { + "epoch": 0.10280668558814254, + "grad_norm": 0.212890625, + "learning_rate": 0.0002, + "loss": 1.2596, + "step": 326 + }, + { + "epoch": 0.10312204351939451, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.5531, + "step": 327 + }, + { + "epoch": 0.10343740145064649, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.5241, + "step": 328 + }, + { + "epoch": 0.10375275938189846, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.6096, + "step": 329 + }, + { + "epoch": 0.10406811731315042, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 330 + }, + { + "epoch": 0.10438347524440239, + "grad_norm": 0.40234375, + "learning_rate": 0.0002, + "loss": 1.7788, + "step": 331 + }, + { + "epoch": 0.10469883317565437, + "grad_norm": 0.404296875, + "learning_rate": 0.0002, + "loss": 1.645, + "step": 332 + }, + { + "epoch": 0.10501419110690634, + "grad_norm": 0.390625, + "learning_rate": 0.0002, + "loss": 1.616, + "step": 333 + }, + { + "epoch": 0.1053295490381583, + "grad_norm": 0.5546875, + "learning_rate": 0.0002, + "loss": 2.0529, + "step": 334 + }, + { + "epoch": 0.10564490696941028, + "grad_norm": 0.498046875, + "learning_rate": 0.0002, + "loss": 1.9434, + "step": 335 + }, + { + "epoch": 0.10596026490066225, + "grad_norm": 0.52734375, + "learning_rate": 0.0002, + "loss": 2.0844, + "step": 336 + }, + { + "epoch": 0.10627562283191422, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 2.3603, + "step": 337 + }, + { + "epoch": 0.1065909807631662, + "grad_norm": 0.7578125, + "learning_rate": 0.0002, + "loss": 2.1956, + "step": 338 + }, + { + "epoch": 0.10690633869441817, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 2.4787, + "step": 339 + }, + { + "epoch": 0.10722169662567013, + "grad_norm": 0.5703125, + "learning_rate": 0.0002, + "loss": 2.2313, + "step": 340 + }, + { + "epoch": 0.10753705455692211, + "grad_norm": 0.6171875, + "learning_rate": 0.0002, + "loss": 2.2433, + "step": 341 + }, + { + "epoch": 0.10785241248817408, + "grad_norm": 0.6171875, + "learning_rate": 0.0002, + "loss": 2.6058, + "step": 342 + }, + { + "epoch": 0.10816777041942605, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 2.3355, + "step": 343 + }, + { + "epoch": 0.10848312835067803, + "grad_norm": 0.7109375, + "learning_rate": 0.0002, + "loss": 2.5664, + "step": 344 + }, + { + "epoch": 0.10879848628192999, + "grad_norm": 1.1640625, + "learning_rate": 0.0002, + "loss": 2.8838, + "step": 345 + }, + { + "epoch": 0.10911384421318196, + "grad_norm": 0.90625, + "learning_rate": 0.0002, + "loss": 2.4161, + "step": 346 + }, + { + "epoch": 0.10942920214443393, + "grad_norm": 0.921875, + "learning_rate": 0.0002, + "loss": 2.7561, + "step": 347 + }, + { + "epoch": 0.1097445600756859, + "grad_norm": 0.71484375, + "learning_rate": 0.0002, + "loss": 2.032, + "step": 348 + }, + { + "epoch": 0.11005991800693787, + "grad_norm": 0.75390625, + "learning_rate": 0.0002, + "loss": 1.9817, + "step": 349 + }, + { + "epoch": 0.11037527593818984, + "grad_norm": 1.359375, + "learning_rate": 0.0002, + "loss": 3.0408, + "step": 350 + }, + { + "epoch": 0.11069063386944182, + "grad_norm": 0.42578125, + "learning_rate": 0.0002, + "loss": 1.2797, + "step": 351 + }, + { + "epoch": 0.11100599180069379, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 1.2104, + "step": 352 + }, + { + "epoch": 0.11132134973194575, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.2535, + "step": 353 + }, + { + "epoch": 0.11163670766319773, + "grad_norm": 0.177734375, + "learning_rate": 0.0002, + "loss": 1.0378, + "step": 354 + }, + { + "epoch": 0.1119520655944497, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.1025, + "step": 355 + }, + { + "epoch": 0.11226742352570167, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.2201, + "step": 356 + }, + { + "epoch": 0.11258278145695365, + "grad_norm": 0.1474609375, + "learning_rate": 0.0002, + "loss": 1.2402, + "step": 357 + }, + { + "epoch": 0.11289813938820562, + "grad_norm": 0.1552734375, + "learning_rate": 0.0002, + "loss": 1.1844, + "step": 358 + }, + { + "epoch": 0.11321349731945758, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.1586, + "step": 359 + }, + { + "epoch": 0.11352885525070956, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.2167, + "step": 360 + }, + { + "epoch": 0.11384421318196153, + "grad_norm": 0.142578125, + "learning_rate": 0.0002, + "loss": 1.0358, + "step": 361 + }, + { + "epoch": 0.1141595711132135, + "grad_norm": 0.130859375, + "learning_rate": 0.0002, + "loss": 1.0204, + "step": 362 + }, + { + "epoch": 0.11447492904446546, + "grad_norm": 0.154296875, + "learning_rate": 0.0002, + "loss": 1.1835, + "step": 363 + }, + { + "epoch": 0.11479028697571744, + "grad_norm": 0.1591796875, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 364 + }, + { + "epoch": 0.11510564490696941, + "grad_norm": 0.1591796875, + "learning_rate": 0.0002, + "loss": 1.1104, + "step": 365 + }, + { + "epoch": 0.11542100283822138, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.2516, + "step": 366 + }, + { + "epoch": 0.11573636076947336, + "grad_norm": 0.17578125, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 367 + }, + { + "epoch": 0.11605171870072532, + "grad_norm": 0.177734375, + "learning_rate": 0.0002, + "loss": 1.1958, + "step": 368 + }, + { + "epoch": 0.11636707663197729, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 1.3831, + "step": 369 + }, + { + "epoch": 0.11668243456322927, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 1.4311, + "step": 370 + }, + { + "epoch": 0.11699779249448124, + "grad_norm": 0.5, + "learning_rate": 0.0002, + "loss": 1.4955, + "step": 371 + }, + { + "epoch": 0.1173131504257332, + "grad_norm": 0.2451171875, + "learning_rate": 0.0002, + "loss": 1.3997, + "step": 372 + }, + { + "epoch": 0.11762850835698518, + "grad_norm": 0.2373046875, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 373 + }, + { + "epoch": 0.11794386628823715, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 1.4299, + "step": 374 + }, + { + "epoch": 0.11794386628823715, + "eval_loss": 1.7538520097732544, + "eval_runtime": 148.5995, + "eval_samples_per_second": 6.729, + "eval_steps_per_second": 6.729, + "step": 374 + }, + { + "epoch": 0.11794386628823715, + "mmlu_eval_accuracy": 0.6234283651455597, + "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182, + "mmlu_eval_accuracy_anatomy": 0.5714285714285714, + "mmlu_eval_accuracy_astronomy": 0.75, + "mmlu_eval_accuracy_business_ethics": 0.45454545454545453, + "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138, + "mmlu_eval_accuracy_college_biology": 0.625, + "mmlu_eval_accuracy_college_chemistry": 0.5, + "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, + "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365, + "mmlu_eval_accuracy_college_medicine": 0.6818181818181818, + "mmlu_eval_accuracy_college_physics": 0.7272727272727273, + "mmlu_eval_accuracy_computer_security": 0.5454545454545454, + "mmlu_eval_accuracy_conceptual_physics": 0.6153846153846154, + "mmlu_eval_accuracy_econometrics": 0.5833333333333334, + "mmlu_eval_accuracy_electrical_engineering": 0.5625, + "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244, + "mmlu_eval_accuracy_formal_logic": 0.42857142857142855, + "mmlu_eval_accuracy_global_facts": 0.3, + "mmlu_eval_accuracy_high_school_biology": 0.6875, + "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453, + "mmlu_eval_accuracy_high_school_computer_science": 0.7777777777777778, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.6046511627906976, + "mmlu_eval_accuracy_high_school_mathematics": 0.3448275862068966, + "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, + "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354, + "mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333, + "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, + "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, + "mmlu_eval_accuracy_high_school_world_history": 0.8076923076923077, + "mmlu_eval_accuracy_human_aging": 0.6521739130434783, + "mmlu_eval_accuracy_human_sexuality": 0.75, + "mmlu_eval_accuracy_international_law": 0.9230769230769231, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.36363636363636365, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.8, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7558139534883721, + "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, + "mmlu_eval_accuracy_moral_scenarios": 0.37, + "mmlu_eval_accuracy_nutrition": 0.7575757575757576, + "mmlu_eval_accuracy_philosophy": 0.7647058823529411, + "mmlu_eval_accuracy_prehistory": 0.5714285714285714, + "mmlu_eval_accuracy_professional_accounting": 0.3870967741935484, + "mmlu_eval_accuracy_professional_law": 0.4117647058823529, + "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806, + "mmlu_eval_accuracy_professional_psychology": 0.6521739130434783, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7037037037037037, + "mmlu_eval_accuracy_sociology": 0.9545454545454546, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.5555555555555556, + "mmlu_eval_accuracy_world_religions": 0.8421052631578947, + "mmlu_loss": 6.131260786548617, + "step": 374 + }, + { + "epoch": 0.11825922421948912, + "grad_norm": 0.2490234375, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 375 + }, + { + "epoch": 0.1185745821507411, + "grad_norm": 0.2421875, + "learning_rate": 0.0002, + "loss": 1.2818, + "step": 376 + }, + { + "epoch": 0.11888994008199306, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 1.6989, + "step": 377 + }, + { + "epoch": 0.11920529801324503, + "grad_norm": 0.443359375, + "learning_rate": 0.0002, + "loss": 1.4423, + "step": 378 + }, + { + "epoch": 0.119520655944497, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 1.2524, + "step": 379 + }, + { + "epoch": 0.11983601387574898, + "grad_norm": 0.365234375, + "learning_rate": 0.0002, + "loss": 1.6477, + "step": 380 + }, + { + "epoch": 0.12015137180700095, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 1.5686, + "step": 381 + }, + { + "epoch": 0.12046672973825291, + "grad_norm": 0.478515625, + "learning_rate": 0.0002, + "loss": 1.6437, + "step": 382 + }, + { + "epoch": 0.12078208766950489, + "grad_norm": 0.482421875, + "learning_rate": 0.0002, + "loss": 2.0001, + "step": 383 + }, + { + "epoch": 0.12109744560075686, + "grad_norm": 0.443359375, + "learning_rate": 0.0002, + "loss": 1.782, + "step": 384 + }, + { + "epoch": 0.12141280353200883, + "grad_norm": 0.51953125, + "learning_rate": 0.0002, + "loss": 1.7598, + "step": 385 + }, + { + "epoch": 0.1217281614632608, + "grad_norm": 0.77734375, + "learning_rate": 0.0002, + "loss": 2.5275, + "step": 386 + }, + { + "epoch": 0.12204351939451277, + "grad_norm": 0.67578125, + "learning_rate": 0.0002, + "loss": 2.4165, + "step": 387 + }, + { + "epoch": 0.12235887732576474, + "grad_norm": 0.5390625, + "learning_rate": 0.0002, + "loss": 2.0436, + "step": 388 + }, + { + "epoch": 0.12267423525701672, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 2.6016, + "step": 389 + }, + { + "epoch": 0.12298959318826869, + "grad_norm": 0.7265625, + "learning_rate": 0.0002, + "loss": 2.3348, + "step": 390 + }, + { + "epoch": 0.12330495111952065, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 2.4088, + "step": 391 + }, + { + "epoch": 0.12362030905077263, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 2.306, + "step": 392 + }, + { + "epoch": 0.1239356669820246, + "grad_norm": 0.796875, + "learning_rate": 0.0002, + "loss": 2.673, + "step": 393 + }, + { + "epoch": 0.12425102491327657, + "grad_norm": 0.89453125, + "learning_rate": 0.0002, + "loss": 2.8786, + "step": 394 + }, + { + "epoch": 0.12456638284452853, + "grad_norm": 0.8515625, + "learning_rate": 0.0002, + "loss": 2.5925, + "step": 395 + }, + { + "epoch": 0.12488174077578051, + "grad_norm": 0.7890625, + "learning_rate": 0.0002, + "loss": 2.2632, + "step": 396 + }, + { + "epoch": 0.1251970987070325, + "grad_norm": 0.6875, + "learning_rate": 0.0002, + "loss": 2.4347, + "step": 397 + }, + { + "epoch": 0.12551245663828445, + "grad_norm": 0.79296875, + "learning_rate": 0.0002, + "loss": 2.2099, + "step": 398 + }, + { + "epoch": 0.12582781456953643, + "grad_norm": 0.71484375, + "learning_rate": 0.0002, + "loss": 2.2267, + "step": 399 + }, + { + "epoch": 0.1261431725007884, + "grad_norm": 0.828125, + "learning_rate": 0.0002, + "loss": 2.1834, + "step": 400 + }, + { + "epoch": 0.12645853043204036, + "grad_norm": 0.40234375, + "learning_rate": 0.0002, + "loss": 1.1324, + "step": 401 + }, + { + "epoch": 0.12677388836329234, + "grad_norm": 0.365234375, + "learning_rate": 0.0002, + "loss": 1.2114, + "step": 402 + }, + { + "epoch": 0.1270892462945443, + "grad_norm": 0.197265625, + "learning_rate": 0.0002, + "loss": 1.1506, + "step": 403 + }, + { + "epoch": 0.12740460422579628, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 1.1258, + "step": 404 + }, + { + "epoch": 0.12771996215704826, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.1558, + "step": 405 + }, + { + "epoch": 0.1280353200883002, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 1.1892, + "step": 406 + }, + { + "epoch": 0.1283506780195522, + "grad_norm": 0.158203125, + "learning_rate": 0.0002, + "loss": 1.0804, + "step": 407 + }, + { + "epoch": 0.12866603595080417, + "grad_norm": 0.1474609375, + "learning_rate": 0.0002, + "loss": 1.195, + "step": 408 + }, + { + "epoch": 0.12898139388205612, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.011, + "step": 409 + }, + { + "epoch": 0.1292967518133081, + "grad_norm": 0.1962890625, + "learning_rate": 0.0002, + "loss": 1.0291, + "step": 410 + }, + { + "epoch": 0.12961210974456008, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.2766, + "step": 411 + }, + { + "epoch": 0.12992746767581204, + "grad_norm": 0.1552734375, + "learning_rate": 0.0002, + "loss": 1.2474, + "step": 412 + }, + { + "epoch": 0.13024282560706402, + "grad_norm": 0.216796875, + "learning_rate": 0.0002, + "loss": 1.2832, + "step": 413 + }, + { + "epoch": 0.130558183538316, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.247, + "step": 414 + }, + { + "epoch": 0.13087354146956795, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.1659, + "step": 415 + }, + { + "epoch": 0.13118889940081993, + "grad_norm": 0.1728515625, + "learning_rate": 0.0002, + "loss": 1.2937, + "step": 416 + }, + { + "epoch": 0.1315042573320719, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 1.2618, + "step": 417 + }, + { + "epoch": 0.13181961526332386, + "grad_norm": 0.18359375, + "learning_rate": 0.0002, + "loss": 1.2106, + "step": 418 + }, + { + "epoch": 0.13213497319457584, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.0309, + "step": 419 + }, + { + "epoch": 0.13245033112582782, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 420 + }, + { + "epoch": 0.13276568905707978, + "grad_norm": 0.2060546875, + "learning_rate": 0.0002, + "loss": 1.212, + "step": 421 + }, + { + "epoch": 0.13308104698833176, + "grad_norm": 0.2314453125, + "learning_rate": 0.0002, + "loss": 1.2081, + "step": 422 + }, + { + "epoch": 0.13339640491958374, + "grad_norm": 0.2412109375, + "learning_rate": 0.0002, + "loss": 1.5565, + "step": 423 + }, + { + "epoch": 0.1337117628508357, + "grad_norm": 0.2392578125, + "learning_rate": 0.0002, + "loss": 1.4326, + "step": 424 + }, + { + "epoch": 0.13402712078208767, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 1.558, + "step": 425 + }, + { + "epoch": 0.13434247871333965, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 1.4221, + "step": 426 + }, + { + "epoch": 0.1346578366445916, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 1.4799, + "step": 427 + }, + { + "epoch": 0.13497319457584359, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 1.5228, + "step": 428 + }, + { + "epoch": 0.13528855250709557, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.5753, + "step": 429 + }, + { + "epoch": 0.13560391043834752, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 1.364, + "step": 430 + }, + { + "epoch": 0.1359192683695995, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 1.4184, + "step": 431 + }, + { + "epoch": 0.13623462630085148, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 1.5472, + "step": 432 + }, + { + "epoch": 0.13654998423210343, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 2.0288, + "step": 433 + }, + { + "epoch": 0.1368653421633554, + "grad_norm": 0.52734375, + "learning_rate": 0.0002, + "loss": 1.8391, + "step": 434 + }, + { + "epoch": 0.13718070009460737, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 1.6465, + "step": 435 + }, + { + "epoch": 0.13749605802585935, + "grad_norm": 0.43359375, + "learning_rate": 0.0002, + "loss": 2.0095, + "step": 436 + }, + { + "epoch": 0.13781141595711133, + "grad_norm": 0.45703125, + "learning_rate": 0.0002, + "loss": 1.6293, + "step": 437 + }, + { + "epoch": 0.13812677388836328, + "grad_norm": 0.60546875, + "learning_rate": 0.0002, + "loss": 2.4437, + "step": 438 + }, + { + "epoch": 0.13844213181961526, + "grad_norm": 0.72265625, + "learning_rate": 0.0002, + "loss": 2.4307, + "step": 439 + }, + { + "epoch": 0.13875748975086724, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 2.0603, + "step": 440 + }, + { + "epoch": 0.1390728476821192, + "grad_norm": 0.796875, + "learning_rate": 0.0002, + "loss": 2.2516, + "step": 441 + }, + { + "epoch": 0.13938820561337117, + "grad_norm": 0.83984375, + "learning_rate": 0.0002, + "loss": 2.7348, + "step": 442 + }, + { + "epoch": 0.13970356354462315, + "grad_norm": 0.83203125, + "learning_rate": 0.0002, + "loss": 2.5543, + "step": 443 + }, + { + "epoch": 0.1400189214758751, + "grad_norm": 0.91015625, + "learning_rate": 0.0002, + "loss": 2.4882, + "step": 444 + }, + { + "epoch": 0.1403342794071271, + "grad_norm": 0.765625, + "learning_rate": 0.0002, + "loss": 2.4923, + "step": 445 + }, + { + "epoch": 0.14064963733837907, + "grad_norm": 0.77734375, + "learning_rate": 0.0002, + "loss": 2.6614, + "step": 446 + }, + { + "epoch": 0.14096499526963102, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 2.6648, + "step": 447 + }, + { + "epoch": 0.141280353200883, + "grad_norm": 0.9453125, + "learning_rate": 0.0002, + "loss": 2.3619, + "step": 448 + }, + { + "epoch": 0.14159571113213498, + "grad_norm": 0.76953125, + "learning_rate": 0.0002, + "loss": 2.302, + "step": 449 + }, + { + "epoch": 0.14191106906338694, + "grad_norm": 1.3984375, + "learning_rate": 0.0002, + "loss": 3.227, + "step": 450 + }, + { + "epoch": 0.14222642699463892, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 1.185, + "step": 451 + }, + { + "epoch": 0.1425417849258909, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 1.2713, + "step": 452 + }, + { + "epoch": 0.14285714285714285, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 1.1483, + "step": 453 + }, + { + "epoch": 0.14317250078839483, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.1871, + "step": 454 + }, + { + "epoch": 0.1434878587196468, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.1451, + "step": 455 + }, + { + "epoch": 0.14380321665089876, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.2011, + "step": 456 + }, + { + "epoch": 0.14411857458215074, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.0529, + "step": 457 + }, + { + "epoch": 0.14443393251340272, + "grad_norm": 0.1796875, + "learning_rate": 0.0002, + "loss": 1.1386, + "step": 458 + }, + { + "epoch": 0.14474929044465468, + "grad_norm": 0.1767578125, + "learning_rate": 0.0002, + "loss": 1.1819, + "step": 459 + }, + { + "epoch": 0.14506464837590666, + "grad_norm": 0.1689453125, + "learning_rate": 0.0002, + "loss": 1.109, + "step": 460 + }, + { + "epoch": 0.14538000630715864, + "grad_norm": 0.16015625, + "learning_rate": 0.0002, + "loss": 1.2992, + "step": 461 + }, + { + "epoch": 0.1456953642384106, + "grad_norm": 0.166015625, + "learning_rate": 0.0002, + "loss": 1.1324, + "step": 462 + }, + { + "epoch": 0.14601072216966257, + "grad_norm": 0.1494140625, + "learning_rate": 0.0002, + "loss": 1.0809, + "step": 463 + }, + { + "epoch": 0.14632608010091455, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.1787, + "step": 464 + }, + { + "epoch": 0.1466414380321665, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.1544, + "step": 465 + }, + { + "epoch": 0.14695679596341849, + "grad_norm": 0.1630859375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 466 + }, + { + "epoch": 0.14727215389467044, + "grad_norm": 0.17578125, + "learning_rate": 0.0002, + "loss": 1.0754, + "step": 467 + }, + { + "epoch": 0.14758751182592242, + "grad_norm": 0.1865234375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 468 + }, + { + "epoch": 0.1479028697571744, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 1.4709, + "step": 469 + }, + { + "epoch": 0.14821822768842635, + "grad_norm": 0.2001953125, + "learning_rate": 0.0002, + "loss": 1.2783, + "step": 470 + }, + { + "epoch": 0.14853358561967833, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 1.3577, + "step": 471 + }, + { + "epoch": 0.1488489435509303, + "grad_norm": 0.2451171875, + "learning_rate": 0.0002, + "loss": 1.403, + "step": 472 + }, + { + "epoch": 0.14916430148218227, + "grad_norm": 0.2275390625, + "learning_rate": 0.0002, + "loss": 1.2275, + "step": 473 + }, + { + "epoch": 0.14947965941343425, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 1.1934, + "step": 474 + }, + { + "epoch": 0.14979501734468623, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.1884, + "step": 475 + }, + { + "epoch": 0.15011037527593818, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.4928, + "step": 476 + }, + { + "epoch": 0.15042573320719016, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.2725, + "step": 477 + }, + { + "epoch": 0.15074109113844214, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 1.6468, + "step": 478 + }, + { + "epoch": 0.1510564490696941, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 1.7393, + "step": 479 + }, + { + "epoch": 0.15137180700094607, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 1.5446, + "step": 480 + }, + { + "epoch": 0.15168716493219805, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 1.7185, + "step": 481 + }, + { + "epoch": 0.15200252286345, + "grad_norm": 0.4765625, + "learning_rate": 0.0002, + "loss": 1.8016, + "step": 482 + }, + { + "epoch": 0.152317880794702, + "grad_norm": 0.3671875, + "learning_rate": 0.0002, + "loss": 1.8335, + "step": 483 + }, + { + "epoch": 0.15263323872595397, + "grad_norm": 0.396484375, + "learning_rate": 0.0002, + "loss": 1.6763, + "step": 484 + }, + { + "epoch": 0.15294859665720592, + "grad_norm": 0.43359375, + "learning_rate": 0.0002, + "loss": 2.0235, + "step": 485 + }, + { + "epoch": 0.1532639545884579, + "grad_norm": 0.5546875, + "learning_rate": 0.0002, + "loss": 2.2547, + "step": 486 + }, + { + "epoch": 0.15357931251970988, + "grad_norm": 0.48828125, + "learning_rate": 0.0002, + "loss": 1.9937, + "step": 487 + }, + { + "epoch": 0.15389467045096183, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 1.9108, + "step": 488 + }, + { + "epoch": 0.15421002838221382, + "grad_norm": 0.60546875, + "learning_rate": 0.0002, + "loss": 2.3341, + "step": 489 + }, + { + "epoch": 0.1545253863134658, + "grad_norm": 0.671875, + "learning_rate": 0.0002, + "loss": 2.6328, + "step": 490 + }, + { + "epoch": 0.15484074424471775, + "grad_norm": 0.76171875, + "learning_rate": 0.0002, + "loss": 2.428, + "step": 491 + }, + { + "epoch": 0.15515610217596973, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 2.0921, + "step": 492 + }, + { + "epoch": 0.1554714601072217, + "grad_norm": 0.61328125, + "learning_rate": 0.0002, + "loss": 2.3378, + "step": 493 + }, + { + "epoch": 0.15578681803847366, + "grad_norm": 0.734375, + "learning_rate": 0.0002, + "loss": 2.1567, + "step": 494 + }, + { + "epoch": 0.15610217596972564, + "grad_norm": 0.486328125, + "learning_rate": 0.0002, + "loss": 2.0534, + "step": 495 + }, + { + "epoch": 0.15641753390097762, + "grad_norm": 0.67578125, + "learning_rate": 0.0002, + "loss": 2.4619, + "step": 496 + }, + { + "epoch": 0.15673289183222958, + "grad_norm": 0.89453125, + "learning_rate": 0.0002, + "loss": 2.3073, + "step": 497 + }, + { + "epoch": 0.15704824976348156, + "grad_norm": 0.859375, + "learning_rate": 0.0002, + "loss": 2.486, + "step": 498 + }, + { + "epoch": 0.1573636076947335, + "grad_norm": 0.85546875, + "learning_rate": 0.0002, + "loss": 2.3597, + "step": 499 + }, + { + "epoch": 0.1576789656259855, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 2.8373, + "step": 500 + }, + { + "epoch": 0.15799432355723747, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 1.138, + "step": 501 + }, + { + "epoch": 0.15830968148848942, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 1.1889, + "step": 502 + }, + { + "epoch": 0.1586250394197414, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 1.1594, + "step": 503 + }, + { + "epoch": 0.15894039735099338, + "grad_norm": 0.1728515625, + "learning_rate": 0.0002, + "loss": 1.1404, + "step": 504 + }, + { + "epoch": 0.15925575528224534, + "grad_norm": 0.1494140625, + "learning_rate": 0.0002, + "loss": 1.0898, + "step": 505 + }, + { + "epoch": 0.15957111321349732, + "grad_norm": 0.166015625, + "learning_rate": 0.0002, + "loss": 1.0774, + "step": 506 + }, + { + "epoch": 0.1598864711447493, + "grad_norm": 0.1689453125, + "learning_rate": 0.0002, + "loss": 1.1109, + "step": 507 + }, + { + "epoch": 0.16020182907600125, + "grad_norm": 0.1943359375, + "learning_rate": 0.0002, + "loss": 1.0499, + "step": 508 + }, + { + "epoch": 0.16051718700725323, + "grad_norm": 0.1865234375, + "learning_rate": 0.0002, + "loss": 1.1447, + "step": 509 + }, + { + "epoch": 0.1608325449385052, + "grad_norm": 0.1787109375, + "learning_rate": 0.0002, + "loss": 1.2169, + "step": 510 + }, + { + "epoch": 0.16114790286975716, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.2663, + "step": 511 + }, + { + "epoch": 0.16146326080100915, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 512 + }, + { + "epoch": 0.16177861873226113, + "grad_norm": 0.1884765625, + "learning_rate": 0.0002, + "loss": 1.1487, + "step": 513 + }, + { + "epoch": 0.16209397666351308, + "grad_norm": 0.17578125, + "learning_rate": 0.0002, + "loss": 1.1424, + "step": 514 + }, + { + "epoch": 0.16240933459476506, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 1.2152, + "step": 515 + }, + { + "epoch": 0.16272469252601704, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.4146, + "step": 516 + }, + { + "epoch": 0.163040050457269, + "grad_norm": 0.197265625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 517 + }, + { + "epoch": 0.16335540838852097, + "grad_norm": 0.244140625, + "learning_rate": 0.0002, + "loss": 1.4217, + "step": 518 + }, + { + "epoch": 0.16367076631977295, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 1.3938, + "step": 519 + }, + { + "epoch": 0.1639861242510249, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 1.2667, + "step": 520 + }, + { + "epoch": 0.1643014821822769, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.2861, + "step": 521 + }, + { + "epoch": 0.16461684011352887, + "grad_norm": 0.2451171875, + "learning_rate": 0.0002, + "loss": 1.2891, + "step": 522 + }, + { + "epoch": 0.16493219804478082, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.4774, + "step": 523 + }, + { + "epoch": 0.1652475559760328, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 1.3757, + "step": 524 + }, + { + "epoch": 0.16556291390728478, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 1.3653, + "step": 525 + }, + { + "epoch": 0.16587827183853673, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 1.4896, + "step": 526 + }, + { + "epoch": 0.16619362976978871, + "grad_norm": 0.353515625, + "learning_rate": 0.0002, + "loss": 1.544, + "step": 527 + }, + { + "epoch": 0.1665089877010407, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 1.441, + "step": 528 + }, + { + "epoch": 0.16682434563229265, + "grad_norm": 0.388671875, + "learning_rate": 0.0002, + "loss": 1.7373, + "step": 529 + }, + { + "epoch": 0.16713970356354463, + "grad_norm": 0.369140625, + "learning_rate": 0.0002, + "loss": 1.6602, + "step": 530 + }, + { + "epoch": 0.16745506149479658, + "grad_norm": 0.734375, + "learning_rate": 0.0002, + "loss": 1.837, + "step": 531 + }, + { + "epoch": 0.16777041942604856, + "grad_norm": 0.447265625, + "learning_rate": 0.0002, + "loss": 2.1153, + "step": 532 + }, + { + "epoch": 0.16808577735730054, + "grad_norm": 0.482421875, + "learning_rate": 0.0002, + "loss": 2.0156, + "step": 533 + }, + { + "epoch": 0.1684011352885525, + "grad_norm": 0.47265625, + "learning_rate": 0.0002, + "loss": 1.8938, + "step": 534 + }, + { + "epoch": 0.16871649321980448, + "grad_norm": 0.5703125, + "learning_rate": 0.0002, + "loss": 2.018, + "step": 535 + }, + { + "epoch": 0.16903185115105646, + "grad_norm": 0.62109375, + "learning_rate": 0.0002, + "loss": 2.2735, + "step": 536 + }, + { + "epoch": 0.1693472090823084, + "grad_norm": 0.984375, + "learning_rate": 0.0002, + "loss": 2.5888, + "step": 537 + }, + { + "epoch": 0.1696625670135604, + "grad_norm": 0.640625, + "learning_rate": 0.0002, + "loss": 2.1984, + "step": 538 + }, + { + "epoch": 0.16997792494481237, + "grad_norm": 0.7265625, + "learning_rate": 0.0002, + "loss": 2.637, + "step": 539 + }, + { + "epoch": 0.17029328287606432, + "grad_norm": 0.80859375, + "learning_rate": 0.0002, + "loss": 2.6064, + "step": 540 + }, + { + "epoch": 0.1706086408073163, + "grad_norm": 0.6171875, + "learning_rate": 0.0002, + "loss": 2.337, + "step": 541 + }, + { + "epoch": 0.17092399873856828, + "grad_norm": 0.8984375, + "learning_rate": 0.0002, + "loss": 2.8711, + "step": 542 + }, + { + "epoch": 0.17123935666982024, + "grad_norm": 0.7890625, + "learning_rate": 0.0002, + "loss": 2.3873, + "step": 543 + }, + { + "epoch": 0.17155471460107222, + "grad_norm": 0.6640625, + "learning_rate": 0.0002, + "loss": 2.6881, + "step": 544 + }, + { + "epoch": 0.1718700725323242, + "grad_norm": 0.6484375, + "learning_rate": 0.0002, + "loss": 2.2061, + "step": 545 + }, + { + "epoch": 0.17218543046357615, + "grad_norm": 0.703125, + "learning_rate": 0.0002, + "loss": 2.4399, + "step": 546 + }, + { + "epoch": 0.17250078839482813, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 2.3735, + "step": 547 + }, + { + "epoch": 0.1728161463260801, + "grad_norm": 0.84375, + "learning_rate": 0.0002, + "loss": 2.7653, + "step": 548 + }, + { + "epoch": 0.17313150425733206, + "grad_norm": 0.73828125, + "learning_rate": 0.0002, + "loss": 2.4699, + "step": 549 + }, + { + "epoch": 0.17344686218858404, + "grad_norm": 0.86328125, + "learning_rate": 0.0002, + "loss": 2.3798, + "step": 550 + }, + { + "epoch": 0.17376222011983603, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 551 + }, + { + "epoch": 0.17407757805108798, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 1.1799, + "step": 552 + }, + { + "epoch": 0.17439293598233996, + "grad_norm": 0.19921875, + "learning_rate": 0.0002, + "loss": 1.1911, + "step": 553 + }, + { + "epoch": 0.17470829391359194, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.1913, + "step": 554 + }, + { + "epoch": 0.1750236518448439, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.1085, + "step": 555 + }, + { + "epoch": 0.17533900977609587, + "grad_norm": 0.169921875, + "learning_rate": 0.0002, + "loss": 1.2169, + "step": 556 + }, + { + "epoch": 0.17565436770734785, + "grad_norm": 0.150390625, + "learning_rate": 0.0002, + "loss": 1.0878, + "step": 557 + }, + { + "epoch": 0.1759697256385998, + "grad_norm": 0.1689453125, + "learning_rate": 0.0002, + "loss": 1.221, + "step": 558 + }, + { + "epoch": 0.17628508356985179, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.14, + "step": 559 + }, + { + "epoch": 0.17660044150110377, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.2455, + "step": 560 + }, + { + "epoch": 0.17691579943235572, + "grad_norm": 0.1865234375, + "learning_rate": 0.0002, + "loss": 1.1554, + "step": 561 + }, + { + "epoch": 0.17691579943235572, + "eval_loss": 1.7082377672195435, + "eval_runtime": 149.0121, + "eval_samples_per_second": 6.711, + "eval_steps_per_second": 6.711, + "step": 561 + }, + { + "epoch": 0.17691579943235572, + "mmlu_eval_accuracy": 0.6134551584031235, + "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, + "mmlu_eval_accuracy_anatomy": 0.5714285714285714, + "mmlu_eval_accuracy_astronomy": 0.6875, + "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, + "mmlu_eval_accuracy_clinical_knowledge": 0.6896551724137931, + "mmlu_eval_accuracy_college_biology": 0.625, + "mmlu_eval_accuracy_college_chemistry": 0.25, + "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, + "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365, + "mmlu_eval_accuracy_college_medicine": 0.6818181818181818, + "mmlu_eval_accuracy_college_physics": 0.6363636363636364, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384, + "mmlu_eval_accuracy_econometrics": 0.5, + "mmlu_eval_accuracy_electrical_engineering": 0.5625, + "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244, + "mmlu_eval_accuracy_formal_logic": 0.42857142857142855, + "mmlu_eval_accuracy_global_facts": 0.3, + "mmlu_eval_accuracy_high_school_biology": 0.625, + "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.8888888888888888, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209, + "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276, + "mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923, + "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, + "mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333, + "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, + "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, + "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, + "mmlu_eval_accuracy_human_aging": 0.6956521739130435, + "mmlu_eval_accuracy_human_sexuality": 0.75, + "mmlu_eval_accuracy_international_law": 1.0, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.36363636363636365, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.84, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7674418604651163, + "mmlu_eval_accuracy_moral_disputes": 0.631578947368421, + "mmlu_eval_accuracy_moral_scenarios": 0.39, + "mmlu_eval_accuracy_nutrition": 0.6666666666666666, + "mmlu_eval_accuracy_philosophy": 0.7352941176470589, + "mmlu_eval_accuracy_prehistory": 0.6285714285714286, + "mmlu_eval_accuracy_professional_accounting": 0.3870967741935484, + "mmlu_eval_accuracy_professional_law": 0.4470588235294118, + "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516, + "mmlu_eval_accuracy_professional_psychology": 0.6811594202898551, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7777777777777778, + "mmlu_eval_accuracy_sociology": 0.8636363636363636, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.4444444444444444, + "mmlu_eval_accuracy_world_religions": 0.8421052631578947, + "mmlu_loss": 6.543176429868134, + "step": 561 + }, + { + "epoch": 0.1772311573636077, + "grad_norm": 0.1630859375, + "learning_rate": 0.0002, + "loss": 1.2309, + "step": 562 + }, + { + "epoch": 0.17754651529485965, + "grad_norm": 0.1630859375, + "learning_rate": 0.0002, + "loss": 1.2385, + "step": 563 + }, + { + "epoch": 0.17786187322611163, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.134, + "step": 564 + }, + { + "epoch": 0.1781772311573636, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.1367, + "step": 565 + }, + { + "epoch": 0.17849258908861557, + "grad_norm": 0.18359375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 566 + }, + { + "epoch": 0.17880794701986755, + "grad_norm": 0.2451171875, + "learning_rate": 0.0002, + "loss": 1.503, + "step": 567 + }, + { + "epoch": 0.17912330495111953, + "grad_norm": 0.2001953125, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 568 + }, + { + "epoch": 0.17943866288237148, + "grad_norm": 0.349609375, + "learning_rate": 0.0002, + "loss": 1.568, + "step": 569 + }, + { + "epoch": 0.17975402081362346, + "grad_norm": 0.201171875, + "learning_rate": 0.0002, + "loss": 1.1767, + "step": 570 + }, + { + "epoch": 0.18006937874487544, + "grad_norm": 0.216796875, + "learning_rate": 0.0002, + "loss": 1.4655, + "step": 571 + }, + { + "epoch": 0.1803847366761274, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 1.434, + "step": 572 + }, + { + "epoch": 0.18070009460737937, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.2709, + "step": 573 + }, + { + "epoch": 0.18101545253863136, + "grad_norm": 0.22265625, + "learning_rate": 0.0002, + "loss": 1.2449, + "step": 574 + }, + { + "epoch": 0.1813308104698833, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.483, + "step": 575 + }, + { + "epoch": 0.1816461684011353, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.2962, + "step": 576 + }, + { + "epoch": 0.18196152633238727, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.6482, + "step": 577 + }, + { + "epoch": 0.18227688426363922, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 1.2956, + "step": 578 + }, + { + "epoch": 0.1825922421948912, + "grad_norm": 0.2470703125, + "learning_rate": 0.0002, + "loss": 1.2494, + "step": 579 + }, + { + "epoch": 0.18290760012614318, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.3809, + "step": 580 + }, + { + "epoch": 0.18322295805739514, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 1.6735, + "step": 581 + }, + { + "epoch": 0.18353831598864712, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 1.7792, + "step": 582 + }, + { + "epoch": 0.1838536739198991, + "grad_norm": 0.3984375, + "learning_rate": 0.0002, + "loss": 1.7878, + "step": 583 + }, + { + "epoch": 0.18416903185115105, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 1.5659, + "step": 584 + }, + { + "epoch": 0.18448438978240303, + "grad_norm": 0.41015625, + "learning_rate": 0.0002, + "loss": 1.6678, + "step": 585 + }, + { + "epoch": 0.184799747713655, + "grad_norm": 0.421875, + "learning_rate": 0.0002, + "loss": 1.6379, + "step": 586 + }, + { + "epoch": 0.18511510564490696, + "grad_norm": 0.470703125, + "learning_rate": 0.0002, + "loss": 2.1127, + "step": 587 + }, + { + "epoch": 0.18543046357615894, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 2.4, + "step": 588 + }, + { + "epoch": 0.18574582150741092, + "grad_norm": 0.65625, + "learning_rate": 0.0002, + "loss": 2.4513, + "step": 589 + }, + { + "epoch": 0.18606117943866288, + "grad_norm": 0.6953125, + "learning_rate": 0.0002, + "loss": 2.4424, + "step": 590 + }, + { + "epoch": 0.18637653736991486, + "grad_norm": 0.53515625, + "learning_rate": 0.0002, + "loss": 2.0237, + "step": 591 + }, + { + "epoch": 0.18669189530116684, + "grad_norm": 0.796875, + "learning_rate": 0.0002, + "loss": 2.4976, + "step": 592 + }, + { + "epoch": 0.1870072532324188, + "grad_norm": 0.78125, + "learning_rate": 0.0002, + "loss": 2.5493, + "step": 593 + }, + { + "epoch": 0.18732261116367077, + "grad_norm": 0.85546875, + "learning_rate": 0.0002, + "loss": 2.7405, + "step": 594 + }, + { + "epoch": 0.18763796909492272, + "grad_norm": 0.7421875, + "learning_rate": 0.0002, + "loss": 2.7221, + "step": 595 + }, + { + "epoch": 0.1879533270261747, + "grad_norm": 0.7265625, + "learning_rate": 0.0002, + "loss": 2.7122, + "step": 596 + }, + { + "epoch": 0.18826868495742669, + "grad_norm": 0.859375, + "learning_rate": 0.0002, + "loss": 2.2485, + "step": 597 + }, + { + "epoch": 0.18858404288867864, + "grad_norm": 0.71484375, + "learning_rate": 0.0002, + "loss": 2.1899, + "step": 598 + }, + { + "epoch": 0.18889940081993062, + "grad_norm": 0.98046875, + "learning_rate": 0.0002, + "loss": 2.6472, + "step": 599 + }, + { + "epoch": 0.1892147587511826, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 1.9071, + "step": 600 + }, + { + "epoch": 0.18953011668243455, + "grad_norm": 0.408203125, + "learning_rate": 0.0002, + "loss": 1.2374, + "step": 601 + }, + { + "epoch": 0.18984547461368653, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 1.2096, + "step": 602 + }, + { + "epoch": 0.1901608325449385, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 1.1185, + "step": 603 + }, + { + "epoch": 0.19047619047619047, + "grad_norm": 0.2333984375, + "learning_rate": 0.0002, + "loss": 1.1972, + "step": 604 + }, + { + "epoch": 0.19079154840744245, + "grad_norm": 0.1865234375, + "learning_rate": 0.0002, + "loss": 1.0142, + "step": 605 + }, + { + "epoch": 0.19110690633869443, + "grad_norm": 0.150390625, + "learning_rate": 0.0002, + "loss": 1.0489, + "step": 606 + }, + { + "epoch": 0.19142226426994638, + "grad_norm": 0.1591796875, + "learning_rate": 0.0002, + "loss": 1.0872, + "step": 607 + }, + { + "epoch": 0.19173762220119836, + "grad_norm": 0.1689453125, + "learning_rate": 0.0002, + "loss": 1.0761, + "step": 608 + }, + { + "epoch": 0.19205298013245034, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.0207, + "step": 609 + }, + { + "epoch": 0.1923683380637023, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 1.1316, + "step": 610 + }, + { + "epoch": 0.19268369599495427, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 1.0935, + "step": 611 + }, + { + "epoch": 0.19299905392620625, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 1.2298, + "step": 612 + }, + { + "epoch": 0.1933144118574582, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.1772, + "step": 613 + }, + { + "epoch": 0.1936297697887102, + "grad_norm": 0.2138671875, + "learning_rate": 0.0002, + "loss": 1.4453, + "step": 614 + }, + { + "epoch": 0.19394512771996217, + "grad_norm": 0.2041015625, + "learning_rate": 0.0002, + "loss": 1.2676, + "step": 615 + }, + { + "epoch": 0.19426048565121412, + "grad_norm": 0.177734375, + "learning_rate": 0.0002, + "loss": 1.1025, + "step": 616 + }, + { + "epoch": 0.1945758435824661, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.3907, + "step": 617 + }, + { + "epoch": 0.19489120151371808, + "grad_norm": 0.205078125, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 618 + }, + { + "epoch": 0.19520655944497003, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 1.0974, + "step": 619 + }, + { + "epoch": 0.19552191737622202, + "grad_norm": 0.208984375, + "learning_rate": 0.0002, + "loss": 1.2624, + "step": 620 + }, + { + "epoch": 0.195837275307474, + "grad_norm": 0.216796875, + "learning_rate": 0.0002, + "loss": 1.2332, + "step": 621 + }, + { + "epoch": 0.19615263323872595, + "grad_norm": 0.2314453125, + "learning_rate": 0.0002, + "loss": 1.1706, + "step": 622 + }, + { + "epoch": 0.19646799116997793, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 1.5416, + "step": 623 + }, + { + "epoch": 0.1967833491012299, + "grad_norm": 0.2431640625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 624 + }, + { + "epoch": 0.19709870703248186, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 1.5739, + "step": 625 + }, + { + "epoch": 0.19741406496373384, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 626 + }, + { + "epoch": 0.1977294228949858, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.5725, + "step": 627 + }, + { + "epoch": 0.19804478082623778, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 1.396, + "step": 628 + }, + { + "epoch": 0.19836013875748976, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 1.4986, + "step": 629 + }, + { + "epoch": 0.1986754966887417, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 1.4116, + "step": 630 + }, + { + "epoch": 0.1989908546199937, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 1.4884, + "step": 631 + }, + { + "epoch": 0.19930621255124567, + "grad_norm": 0.40625, + "learning_rate": 0.0002, + "loss": 1.8563, + "step": 632 + }, + { + "epoch": 0.19962157048249762, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 1.767, + "step": 633 + }, + { + "epoch": 0.1999369284137496, + "grad_norm": 0.423828125, + "learning_rate": 0.0002, + "loss": 1.8918, + "step": 634 + }, + { + "epoch": 0.20025228634500158, + "grad_norm": 0.44140625, + "learning_rate": 0.0002, + "loss": 1.8326, + "step": 635 + }, + { + "epoch": 0.20056764427625354, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 2.259, + "step": 636 + }, + { + "epoch": 0.20088300220750552, + "grad_norm": 0.494140625, + "learning_rate": 0.0002, + "loss": 2.1315, + "step": 637 + }, + { + "epoch": 0.2011983601387575, + "grad_norm": 0.6171875, + "learning_rate": 0.0002, + "loss": 2.399, + "step": 638 + }, + { + "epoch": 0.20151371807000945, + "grad_norm": 0.87890625, + "learning_rate": 0.0002, + "loss": 3.172, + "step": 639 + }, + { + "epoch": 0.20182907600126143, + "grad_norm": 0.81640625, + "learning_rate": 0.0002, + "loss": 2.4783, + "step": 640 + }, + { + "epoch": 0.2021444339325134, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 2.4311, + "step": 641 + }, + { + "epoch": 0.20245979186376536, + "grad_norm": 0.84765625, + "learning_rate": 0.0002, + "loss": 2.5225, + "step": 642 + }, + { + "epoch": 0.20277514979501735, + "grad_norm": 0.73046875, + "learning_rate": 0.0002, + "loss": 2.566, + "step": 643 + }, + { + "epoch": 0.20309050772626933, + "grad_norm": 0.6484375, + "learning_rate": 0.0002, + "loss": 2.2736, + "step": 644 + }, + { + "epoch": 0.20340586565752128, + "grad_norm": 0.70703125, + "learning_rate": 0.0002, + "loss": 2.4273, + "step": 645 + }, + { + "epoch": 0.20372122358877326, + "grad_norm": 0.734375, + "learning_rate": 0.0002, + "loss": 2.6348, + "step": 646 + }, + { + "epoch": 0.20403658152002524, + "grad_norm": 0.7421875, + "learning_rate": 0.0002, + "loss": 2.4896, + "step": 647 + }, + { + "epoch": 0.2043519394512772, + "grad_norm": 0.77734375, + "learning_rate": 0.0002, + "loss": 2.5071, + "step": 648 + }, + { + "epoch": 0.20466729738252917, + "grad_norm": 1.2265625, + "learning_rate": 0.0002, + "loss": 2.4664, + "step": 649 + }, + { + "epoch": 0.20498265531378115, + "grad_norm": 0.71875, + "learning_rate": 0.0002, + "loss": 2.1502, + "step": 650 + }, + { + "epoch": 0.2052980132450331, + "grad_norm": 0.2333984375, + "learning_rate": 0.0002, + "loss": 1.1735, + "step": 651 + }, + { + "epoch": 0.2056133711762851, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.0923, + "step": 652 + }, + { + "epoch": 0.20592872910753707, + "grad_norm": 0.2060546875, + "learning_rate": 0.0002, + "loss": 1.0753, + "step": 653 + }, + { + "epoch": 0.20624408703878902, + "grad_norm": 0.1796875, + "learning_rate": 0.0002, + "loss": 1.2184, + "step": 654 + }, + { + "epoch": 0.206559444970041, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.1121, + "step": 655 + }, + { + "epoch": 0.20687480290129298, + "grad_norm": 0.1455078125, + "learning_rate": 0.0002, + "loss": 1.0778, + "step": 656 + }, + { + "epoch": 0.20719016083254493, + "grad_norm": 0.1552734375, + "learning_rate": 0.0002, + "loss": 1.1024, + "step": 657 + }, + { + "epoch": 0.20750551876379691, + "grad_norm": 0.1640625, + "learning_rate": 0.0002, + "loss": 1.0803, + "step": 658 + }, + { + "epoch": 0.20782087669504887, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.1052, + "step": 659 + }, + { + "epoch": 0.20813623462630085, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.201, + "step": 660 + }, + { + "epoch": 0.20845159255755283, + "grad_norm": 0.1591796875, + "learning_rate": 0.0002, + "loss": 1.1822, + "step": 661 + }, + { + "epoch": 0.20876695048880478, + "grad_norm": 0.1591796875, + "learning_rate": 0.0002, + "loss": 1.1272, + "step": 662 + }, + { + "epoch": 0.20908230842005676, + "grad_norm": 0.1640625, + "learning_rate": 0.0002, + "loss": 1.2106, + "step": 663 + }, + { + "epoch": 0.20939766635130874, + "grad_norm": 0.1552734375, + "learning_rate": 0.0002, + "loss": 1.1659, + "step": 664 + }, + { + "epoch": 0.2097130242825607, + "grad_norm": 0.1513671875, + "learning_rate": 0.0002, + "loss": 1.1266, + "step": 665 + }, + { + "epoch": 0.21002838221381268, + "grad_norm": 0.1640625, + "learning_rate": 0.0002, + "loss": 1.0795, + "step": 666 + }, + { + "epoch": 0.21034374014506466, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.1887, + "step": 667 + }, + { + "epoch": 0.2106590980763166, + "grad_norm": 0.1884765625, + "learning_rate": 0.0002, + "loss": 1.2369, + "step": 668 + }, + { + "epoch": 0.2109744560075686, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.2676, + "step": 669 + }, + { + "epoch": 0.21128981393882057, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 1.2743, + "step": 670 + }, + { + "epoch": 0.21160517187007252, + "grad_norm": 0.2265625, + "learning_rate": 0.0002, + "loss": 1.4672, + "step": 671 + }, + { + "epoch": 0.2119205298013245, + "grad_norm": 0.2001953125, + "learning_rate": 0.0002, + "loss": 1.3022, + "step": 672 + }, + { + "epoch": 0.21223588773257648, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 1.4927, + "step": 673 + }, + { + "epoch": 0.21255124566382844, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.5227, + "step": 674 + }, + { + "epoch": 0.21286660359508042, + "grad_norm": 0.244140625, + "learning_rate": 0.0002, + "loss": 1.4333, + "step": 675 + }, + { + "epoch": 0.2131819615263324, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 1.4918, + "step": 676 + }, + { + "epoch": 0.21349731945758435, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.6029, + "step": 677 + }, + { + "epoch": 0.21381267738883633, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 1.5445, + "step": 678 + }, + { + "epoch": 0.2141280353200883, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 1.422, + "step": 679 + }, + { + "epoch": 0.21444339325134026, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 1.3776, + "step": 680 + }, + { + "epoch": 0.21475875118259224, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 1.5445, + "step": 681 + }, + { + "epoch": 0.21507410911384423, + "grad_norm": 0.3828125, + "learning_rate": 0.0002, + "loss": 1.5625, + "step": 682 + }, + { + "epoch": 0.21538946704509618, + "grad_norm": 0.416015625, + "learning_rate": 0.0002, + "loss": 1.774, + "step": 683 + }, + { + "epoch": 0.21570482497634816, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 1.7947, + "step": 684 + }, + { + "epoch": 0.21602018290760014, + "grad_norm": 0.52734375, + "learning_rate": 0.0002, + "loss": 1.7772, + "step": 685 + }, + { + "epoch": 0.2163355408388521, + "grad_norm": 0.625, + "learning_rate": 0.0002, + "loss": 2.2918, + "step": 686 + }, + { + "epoch": 0.21665089877010407, + "grad_norm": 0.5390625, + "learning_rate": 0.0002, + "loss": 2.0711, + "step": 687 + }, + { + "epoch": 0.21696625670135605, + "grad_norm": 0.7734375, + "learning_rate": 0.0002, + "loss": 2.4848, + "step": 688 + }, + { + "epoch": 0.217281614632608, + "grad_norm": 0.65625, + "learning_rate": 0.0002, + "loss": 2.6606, + "step": 689 + }, + { + "epoch": 0.21759697256385999, + "grad_norm": 0.5546875, + "learning_rate": 0.0002, + "loss": 2.0709, + "step": 690 + }, + { + "epoch": 0.21791233049511194, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 2.351, + "step": 691 + }, + { + "epoch": 0.21822768842636392, + "grad_norm": 0.73828125, + "learning_rate": 0.0002, + "loss": 2.47, + "step": 692 + }, + { + "epoch": 0.2185430463576159, + "grad_norm": 0.80078125, + "learning_rate": 0.0002, + "loss": 2.6881, + "step": 693 + }, + { + "epoch": 0.21885840428886785, + "grad_norm": 0.87109375, + "learning_rate": 0.0002, + "loss": 2.862, + "step": 694 + }, + { + "epoch": 0.21917376222011983, + "grad_norm": 0.578125, + "learning_rate": 0.0002, + "loss": 2.1039, + "step": 695 + }, + { + "epoch": 0.2194891201513718, + "grad_norm": 0.92578125, + "learning_rate": 0.0002, + "loss": 2.3882, + "step": 696 + }, + { + "epoch": 0.21980447808262377, + "grad_norm": 0.71484375, + "learning_rate": 0.0002, + "loss": 2.3712, + "step": 697 + }, + { + "epoch": 0.22011983601387575, + "grad_norm": 0.6796875, + "learning_rate": 0.0002, + "loss": 1.9271, + "step": 698 + }, + { + "epoch": 0.22043519394512773, + "grad_norm": 0.72265625, + "learning_rate": 0.0002, + "loss": 2.0941, + "step": 699 + }, + { + "epoch": 0.22075055187637968, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 2.6467, + "step": 700 + }, + { + "epoch": 0.22106590980763166, + "grad_norm": 0.1962890625, + "learning_rate": 0.0002, + "loss": 1.0458, + "step": 701 + }, + { + "epoch": 0.22138126773888364, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.0881, + "step": 702 + }, + { + "epoch": 0.2216966256701356, + "grad_norm": 0.1884765625, + "learning_rate": 0.0002, + "loss": 1.2009, + "step": 703 + }, + { + "epoch": 0.22201198360138757, + "grad_norm": 0.169921875, + "learning_rate": 0.0002, + "loss": 1.2119, + "step": 704 + }, + { + "epoch": 0.22232734153263956, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.0177, + "step": 705 + }, + { + "epoch": 0.2226426994638915, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.1216, + "step": 706 + }, + { + "epoch": 0.2229580573951435, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.1341, + "step": 707 + }, + { + "epoch": 0.22327341532639547, + "grad_norm": 0.1494140625, + "learning_rate": 0.0002, + "loss": 1.1636, + "step": 708 + }, + { + "epoch": 0.22358877325764742, + "grad_norm": 0.150390625, + "learning_rate": 0.0002, + "loss": 1.1492, + "step": 709 + }, + { + "epoch": 0.2239041311888994, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 1.2792, + "step": 710 + }, + { + "epoch": 0.22421948912015138, + "grad_norm": 0.18359375, + "learning_rate": 0.0002, + "loss": 1.0699, + "step": 711 + }, + { + "epoch": 0.22453484705140334, + "grad_norm": 0.1630859375, + "learning_rate": 0.0002, + "loss": 1.1939, + "step": 712 + }, + { + "epoch": 0.22485020498265532, + "grad_norm": 0.16015625, + "learning_rate": 0.0002, + "loss": 1.2615, + "step": 713 + }, + { + "epoch": 0.2251655629139073, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.1169, + "step": 714 + }, + { + "epoch": 0.22548092084515925, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 1.0992, + "step": 715 + }, + { + "epoch": 0.22579627877641123, + "grad_norm": 0.1884765625, + "learning_rate": 0.0002, + "loss": 1.1699, + "step": 716 + }, + { + "epoch": 0.2261116367076632, + "grad_norm": 0.1865234375, + "learning_rate": 0.0002, + "loss": 1.2824, + "step": 717 + }, + { + "epoch": 0.22642699463891516, + "grad_norm": 0.224609375, + "learning_rate": 0.0002, + "loss": 1.5041, + "step": 718 + }, + { + "epoch": 0.22674235257016714, + "grad_norm": 0.220703125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 719 + }, + { + "epoch": 0.22705771050141912, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 1.2869, + "step": 720 + }, + { + "epoch": 0.22737306843267108, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 1.2664, + "step": 721 + }, + { + "epoch": 0.22768842636392306, + "grad_norm": 0.201171875, + "learning_rate": 0.0002, + "loss": 1.2009, + "step": 722 + }, + { + "epoch": 0.228003784295175, + "grad_norm": 0.2333984375, + "learning_rate": 0.0002, + "loss": 1.3978, + "step": 723 + }, + { + "epoch": 0.228319142226427, + "grad_norm": 0.2431640625, + "learning_rate": 0.0002, + "loss": 1.4338, + "step": 724 + }, + { + "epoch": 0.22863450015767897, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 1.5298, + "step": 725 + }, + { + "epoch": 0.22894985808893092, + "grad_norm": 0.2421875, + "learning_rate": 0.0002, + "loss": 1.4318, + "step": 726 + }, + { + "epoch": 0.2292652160201829, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 727 + }, + { + "epoch": 0.22958057395143489, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 1.6147, + "step": 728 + }, + { + "epoch": 0.22989593188268684, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 1.4533, + "step": 729 + }, + { + "epoch": 0.23021128981393882, + "grad_norm": 0.51953125, + "learning_rate": 0.0002, + "loss": 1.9862, + "step": 730 + }, + { + "epoch": 0.2305266477451908, + "grad_norm": 0.3671875, + "learning_rate": 0.0002, + "loss": 1.685, + "step": 731 + }, + { + "epoch": 0.23084200567644275, + "grad_norm": 0.5, + "learning_rate": 0.0002, + "loss": 1.9971, + "step": 732 + }, + { + "epoch": 0.23115736360769473, + "grad_norm": 0.41796875, + "learning_rate": 0.0002, + "loss": 1.6793, + "step": 733 + }, + { + "epoch": 0.2314727215389467, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 1.9588, + "step": 734 + }, + { + "epoch": 0.23178807947019867, + "grad_norm": 0.51953125, + "learning_rate": 0.0002, + "loss": 2.2252, + "step": 735 + }, + { + "epoch": 0.23210343740145065, + "grad_norm": 0.56640625, + "learning_rate": 0.0002, + "loss": 2.0365, + "step": 736 + }, + { + "epoch": 0.23241879533270263, + "grad_norm": 0.8984375, + "learning_rate": 0.0002, + "loss": 2.2283, + "step": 737 + }, + { + "epoch": 0.23273415326395458, + "grad_norm": 0.58203125, + "learning_rate": 0.0002, + "loss": 2.0691, + "step": 738 + }, + { + "epoch": 0.23304951119520656, + "grad_norm": 0.64453125, + "learning_rate": 0.0002, + "loss": 2.3542, + "step": 739 + }, + { + "epoch": 0.23336486912645854, + "grad_norm": 0.62109375, + "learning_rate": 0.0002, + "loss": 2.2661, + "step": 740 + }, + { + "epoch": 0.2336802270577105, + "grad_norm": 1.015625, + "learning_rate": 0.0002, + "loss": 3.1825, + "step": 741 + }, + { + "epoch": 0.23399558498896247, + "grad_norm": 0.7265625, + "learning_rate": 0.0002, + "loss": 2.2887, + "step": 742 + }, + { + "epoch": 0.23431094292021445, + "grad_norm": 0.6796875, + "learning_rate": 0.0002, + "loss": 2.3296, + "step": 743 + }, + { + "epoch": 0.2346263008514664, + "grad_norm": 0.7421875, + "learning_rate": 0.0002, + "loss": 2.5607, + "step": 744 + }, + { + "epoch": 0.2349416587827184, + "grad_norm": 0.68359375, + "learning_rate": 0.0002, + "loss": 2.4613, + "step": 745 + }, + { + "epoch": 0.23525701671397037, + "grad_norm": 0.8828125, + "learning_rate": 0.0002, + "loss": 2.687, + "step": 746 + }, + { + "epoch": 0.23557237464522232, + "grad_norm": 0.73046875, + "learning_rate": 0.0002, + "loss": 2.6331, + "step": 747 + }, + { + "epoch": 0.2358877325764743, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 2.086, + "step": 748 + }, + { + "epoch": 0.2358877325764743, + "eval_loss": 1.7445578575134277, + "eval_runtime": 148.6203, + "eval_samples_per_second": 6.729, + "eval_steps_per_second": 6.729, + "step": 748 + }, + { + "epoch": 0.2358877325764743, + "mmlu_eval_accuracy": 0.6271676009313613, + "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, + "mmlu_eval_accuracy_anatomy": 0.5714285714285714, + "mmlu_eval_accuracy_astronomy": 0.6875, + "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, + "mmlu_eval_accuracy_clinical_knowledge": 0.6896551724137931, + "mmlu_eval_accuracy_college_biology": 0.6875, + "mmlu_eval_accuracy_college_chemistry": 0.375, + "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, + "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727, + "mmlu_eval_accuracy_college_medicine": 0.6363636363636364, + "mmlu_eval_accuracy_college_physics": 0.7272727272727273, + "mmlu_eval_accuracy_computer_security": 0.5454545454545454, + "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384, + "mmlu_eval_accuracy_econometrics": 0.5, + "mmlu_eval_accuracy_electrical_engineering": 0.625, + "mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536, + "mmlu_eval_accuracy_formal_logic": 0.42857142857142855, + "mmlu_eval_accuracy_global_facts": 0.3, + "mmlu_eval_accuracy_high_school_biology": 0.625, + "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.6046511627906976, + "mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724, + "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, + "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354, + "mmlu_eval_accuracy_high_school_psychology": 0.9, + "mmlu_eval_accuracy_high_school_statistics": 0.4782608695652174, + "mmlu_eval_accuracy_high_school_us_history": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_world_history": 0.8076923076923077, + "mmlu_eval_accuracy_human_aging": 0.6521739130434783, + "mmlu_eval_accuracy_human_sexuality": 0.75, + "mmlu_eval_accuracy_international_law": 1.0, + "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.36363636363636365, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.84, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7790697674418605, + "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527, + "mmlu_eval_accuracy_moral_scenarios": 0.39, + "mmlu_eval_accuracy_nutrition": 0.7272727272727273, + "mmlu_eval_accuracy_philosophy": 0.7352941176470589, + "mmlu_eval_accuracy_prehistory": 0.6285714285714286, + "mmlu_eval_accuracy_professional_accounting": 0.3225806451612903, + "mmlu_eval_accuracy_professional_law": 0.43529411764705883, + "mmlu_eval_accuracy_professional_medicine": 0.6774193548387096, + "mmlu_eval_accuracy_professional_psychology": 0.6811594202898551, + "mmlu_eval_accuracy_public_relations": 0.5833333333333334, + "mmlu_eval_accuracy_security_studies": 0.7407407407407407, + "mmlu_eval_accuracy_sociology": 1.0, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.4444444444444444, + "mmlu_eval_accuracy_world_religions": 0.8421052631578947, + "mmlu_loss": 5.830414561178854, + "step": 748 + }, + { + "epoch": 0.23620309050772628, + "grad_norm": 0.82421875, + "learning_rate": 0.0002, + "loss": 2.705, + "step": 749 + }, + { + "epoch": 0.23651844843897823, + "grad_norm": 0.703125, + "learning_rate": 0.0002, + "loss": 2.2097, + "step": 750 + }, + { + "epoch": 0.23683380637023022, + "grad_norm": 0.2294921875, + "learning_rate": 0.0002, + "loss": 1.1459, + "step": 751 + }, + { + "epoch": 0.2371491643014822, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.1949, + "step": 752 + }, + { + "epoch": 0.23746452223273415, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.0845, + "step": 753 + }, + { + "epoch": 0.23777988016398613, + "grad_norm": 0.169921875, + "learning_rate": 0.0002, + "loss": 1.1417, + "step": 754 + }, + { + "epoch": 0.23809523809523808, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.1012, + "step": 755 + }, + { + "epoch": 0.23841059602649006, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.2046, + "step": 756 + }, + { + "epoch": 0.23872595395774204, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.1631, + "step": 757 + }, + { + "epoch": 0.239041311888994, + "grad_norm": 0.1640625, + "learning_rate": 0.0002, + "loss": 1.0701, + "step": 758 + }, + { + "epoch": 0.23935666982024598, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.1339, + "step": 759 + }, + { + "epoch": 0.23967202775149796, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 1.0661, + "step": 760 + }, + { + "epoch": 0.2399873856827499, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.2737, + "step": 761 + }, + { + "epoch": 0.2403027436140019, + "grad_norm": 0.177734375, + "learning_rate": 0.0002, + "loss": 1.1383, + "step": 762 + }, + { + "epoch": 0.24061810154525387, + "grad_norm": 0.1728515625, + "learning_rate": 0.0002, + "loss": 1.2849, + "step": 763 + }, + { + "epoch": 0.24093345947650582, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.175, + "step": 764 + }, + { + "epoch": 0.2412488174077578, + "grad_norm": 0.189453125, + "learning_rate": 0.0002, + "loss": 1.1083, + "step": 765 + }, + { + "epoch": 0.24156417533900978, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 1.2371, + "step": 766 + }, + { + "epoch": 0.24187953327026174, + "grad_norm": 0.205078125, + "learning_rate": 0.0002, + "loss": 1.2733, + "step": 767 + }, + { + "epoch": 0.24219489120151372, + "grad_norm": 0.2119140625, + "learning_rate": 0.0002, + "loss": 1.2408, + "step": 768 + }, + { + "epoch": 0.2425102491327657, + "grad_norm": 0.2138671875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 769 + }, + { + "epoch": 0.24282560706401765, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.2444, + "step": 770 + }, + { + "epoch": 0.24314096499526963, + "grad_norm": 0.2451171875, + "learning_rate": 0.0002, + "loss": 1.3825, + "step": 771 + }, + { + "epoch": 0.2434563229265216, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.7026, + "step": 772 + }, + { + "epoch": 0.24377168085777356, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 1.5019, + "step": 773 + }, + { + "epoch": 0.24408703878902555, + "grad_norm": 0.2314453125, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 774 + }, + { + "epoch": 0.24440239672027753, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.3606, + "step": 775 + }, + { + "epoch": 0.24471775465152948, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.6243, + "step": 776 + }, + { + "epoch": 0.24503311258278146, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.8294, + "step": 777 + }, + { + "epoch": 0.24534847051403344, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 1.0675, + "step": 778 + }, + { + "epoch": 0.2456638284452854, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 1.5829, + "step": 779 + }, + { + "epoch": 0.24597918637653737, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 780 + }, + { + "epoch": 0.24629454430778935, + "grad_norm": 0.384765625, + "learning_rate": 0.0002, + "loss": 1.7693, + "step": 781 + }, + { + "epoch": 0.2466099022390413, + "grad_norm": 0.41796875, + "learning_rate": 0.0002, + "loss": 1.7155, + "step": 782 + }, + { + "epoch": 0.2469252601702933, + "grad_norm": 0.421875, + "learning_rate": 0.0002, + "loss": 1.6858, + "step": 783 + }, + { + "epoch": 0.24724061810154527, + "grad_norm": 0.43359375, + "learning_rate": 0.0002, + "loss": 1.8823, + "step": 784 + }, + { + "epoch": 0.24755597603279722, + "grad_norm": 0.4609375, + "learning_rate": 0.0002, + "loss": 1.8646, + "step": 785 + }, + { + "epoch": 0.2478713339640492, + "grad_norm": 0.482421875, + "learning_rate": 0.0002, + "loss": 2.0899, + "step": 786 + }, + { + "epoch": 0.24818669189530115, + "grad_norm": 0.486328125, + "learning_rate": 0.0002, + "loss": 1.9487, + "step": 787 + }, + { + "epoch": 0.24850204982655313, + "grad_norm": 0.55859375, + "learning_rate": 0.0002, + "loss": 1.971, + "step": 788 + }, + { + "epoch": 0.24881740775780511, + "grad_norm": 0.5390625, + "learning_rate": 0.0002, + "loss": 2.1209, + "step": 789 + }, + { + "epoch": 0.24913276568905707, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 2.3409, + "step": 790 + }, + { + "epoch": 0.24944812362030905, + "grad_norm": 0.890625, + "learning_rate": 0.0002, + "loss": 2.5004, + "step": 791 + }, + { + "epoch": 0.24976348155156103, + "grad_norm": 0.984375, + "learning_rate": 0.0002, + "loss": 2.6374, + "step": 792 + }, + { + "epoch": 0.250078839482813, + "grad_norm": 0.828125, + "learning_rate": 0.0002, + "loss": 2.1839, + "step": 793 + }, + { + "epoch": 0.250394197414065, + "grad_norm": 0.6640625, + "learning_rate": 0.0002, + "loss": 2.1404, + "step": 794 + }, + { + "epoch": 0.2507095553453169, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 2.0325, + "step": 795 + }, + { + "epoch": 0.2510249132765689, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 2.8468, + "step": 796 + }, + { + "epoch": 0.2513402712078209, + "grad_norm": 0.85546875, + "learning_rate": 0.0002, + "loss": 2.9575, + "step": 797 + }, + { + "epoch": 0.25165562913907286, + "grad_norm": 0.9296875, + "learning_rate": 0.0002, + "loss": 2.674, + "step": 798 + }, + { + "epoch": 0.25197098707032484, + "grad_norm": 0.61328125, + "learning_rate": 0.0002, + "loss": 1.8077, + "step": 799 + }, + { + "epoch": 0.2522863450015768, + "grad_norm": 1.0546875, + "learning_rate": 0.0002, + "loss": 2.6639, + "step": 800 + }, + { + "epoch": 0.25260170293282874, + "grad_norm": 0.19921875, + "learning_rate": 0.0002, + "loss": 1.1997, + "step": 801 + }, + { + "epoch": 0.2529170608640807, + "grad_norm": 0.189453125, + "learning_rate": 0.0002, + "loss": 1.2422, + "step": 802 + }, + { + "epoch": 0.2532324187953327, + "grad_norm": 0.1640625, + "learning_rate": 0.0002, + "loss": 1.1724, + "step": 803 + }, + { + "epoch": 0.2535477767265847, + "grad_norm": 0.158203125, + "learning_rate": 0.0002, + "loss": 1.0646, + "step": 804 + }, + { + "epoch": 0.25386313465783666, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.1282, + "step": 805 + }, + { + "epoch": 0.2541784925890886, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.141, + "step": 806 + }, + { + "epoch": 0.25449385052034057, + "grad_norm": 0.142578125, + "learning_rate": 0.0002, + "loss": 1.0867, + "step": 807 + }, + { + "epoch": 0.25480920845159255, + "grad_norm": 0.1455078125, + "learning_rate": 0.0002, + "loss": 1.1101, + "step": 808 + }, + { + "epoch": 0.25512456638284453, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.0087, + "step": 809 + }, + { + "epoch": 0.2554399243140965, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.1012, + "step": 810 + }, + { + "epoch": 0.2557552822453485, + "grad_norm": 0.158203125, + "learning_rate": 0.0002, + "loss": 1.1718, + "step": 811 + }, + { + "epoch": 0.2560706401766004, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.2142, + "step": 812 + }, + { + "epoch": 0.2563859981078524, + "grad_norm": 0.205078125, + "learning_rate": 0.0002, + "loss": 1.2953, + "step": 813 + }, + { + "epoch": 0.2567013560391044, + "grad_norm": 0.17578125, + "learning_rate": 0.0002, + "loss": 1.1857, + "step": 814 + }, + { + "epoch": 0.25701671397035636, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.1544, + "step": 815 + }, + { + "epoch": 0.25733207190160834, + "grad_norm": 0.197265625, + "learning_rate": 0.0002, + "loss": 1.2379, + "step": 816 + }, + { + "epoch": 0.2576474298328603, + "grad_norm": 0.19140625, + "learning_rate": 0.0002, + "loss": 1.1594, + "step": 817 + }, + { + "epoch": 0.25796278776411224, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.2316, + "step": 818 + }, + { + "epoch": 0.2582781456953642, + "grad_norm": 0.197265625, + "learning_rate": 0.0002, + "loss": 1.2533, + "step": 819 + }, + { + "epoch": 0.2585935036266162, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 820 + }, + { + "epoch": 0.2589088615578682, + "grad_norm": 0.2275390625, + "learning_rate": 0.0002, + "loss": 1.2649, + "step": 821 + }, + { + "epoch": 0.25922421948912017, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.5033, + "step": 822 + }, + { + "epoch": 0.25953957742037215, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.2977, + "step": 823 + }, + { + "epoch": 0.2598549353516241, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.4805, + "step": 824 + }, + { + "epoch": 0.26017029328287605, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.4617, + "step": 825 + }, + { + "epoch": 0.26048565121412803, + "grad_norm": 0.37890625, + "learning_rate": 0.0002, + "loss": 1.5247, + "step": 826 + }, + { + "epoch": 0.26080100914538, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.2379, + "step": 827 + }, + { + "epoch": 0.261116367076632, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 1.5074, + "step": 828 + }, + { + "epoch": 0.261431725007884, + "grad_norm": 0.41015625, + "learning_rate": 0.0002, + "loss": 1.8643, + "step": 829 + }, + { + "epoch": 0.2617470829391359, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 1.5099, + "step": 830 + }, + { + "epoch": 0.2620624408703879, + "grad_norm": 0.455078125, + "learning_rate": 0.0002, + "loss": 1.7793, + "step": 831 + }, + { + "epoch": 0.26237779880163986, + "grad_norm": 0.408203125, + "learning_rate": 0.0002, + "loss": 1.7065, + "step": 832 + }, + { + "epoch": 0.26269315673289184, + "grad_norm": 0.458984375, + "learning_rate": 0.0002, + "loss": 1.832, + "step": 833 + }, + { + "epoch": 0.2630085146641438, + "grad_norm": 0.4609375, + "learning_rate": 0.0002, + "loss": 1.7504, + "step": 834 + }, + { + "epoch": 0.26332387259539575, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 2.2507, + "step": 835 + }, + { + "epoch": 0.2636392305266477, + "grad_norm": 0.453125, + "learning_rate": 0.0002, + "loss": 1.6805, + "step": 836 + }, + { + "epoch": 0.2639545884578997, + "grad_norm": 0.734375, + "learning_rate": 0.0002, + "loss": 2.2011, + "step": 837 + }, + { + "epoch": 0.2642699463891517, + "grad_norm": 0.7109375, + "learning_rate": 0.0002, + "loss": 2.322, + "step": 838 + }, + { + "epoch": 0.26458530432040367, + "grad_norm": 0.71484375, + "learning_rate": 0.0002, + "loss": 2.5868, + "step": 839 + }, + { + "epoch": 0.26490066225165565, + "grad_norm": 0.66796875, + "learning_rate": 0.0002, + "loss": 2.3252, + "step": 840 + }, + { + "epoch": 0.2652160201829076, + "grad_norm": 0.72265625, + "learning_rate": 0.0002, + "loss": 2.42, + "step": 841 + }, + { + "epoch": 0.26553137811415956, + "grad_norm": 1.0390625, + "learning_rate": 0.0002, + "loss": 2.4416, + "step": 842 + }, + { + "epoch": 0.26584673604541154, + "grad_norm": 0.61328125, + "learning_rate": 0.0002, + "loss": 2.1427, + "step": 843 + }, + { + "epoch": 0.2661620939766635, + "grad_norm": 0.70703125, + "learning_rate": 0.0002, + "loss": 2.5096, + "step": 844 + }, + { + "epoch": 0.2664774519079155, + "grad_norm": 0.92578125, + "learning_rate": 0.0002, + "loss": 2.1731, + "step": 845 + }, + { + "epoch": 0.2667928098391675, + "grad_norm": 0.96484375, + "learning_rate": 0.0002, + "loss": 2.76, + "step": 846 + }, + { + "epoch": 0.2671081677704194, + "grad_norm": 0.90625, + "learning_rate": 0.0002, + "loss": 2.7595, + "step": 847 + }, + { + "epoch": 0.2674235257016714, + "grad_norm": 0.8046875, + "learning_rate": 0.0002, + "loss": 2.4337, + "step": 848 + }, + { + "epoch": 0.26773888363292336, + "grad_norm": 0.796875, + "learning_rate": 0.0002, + "loss": 2.3485, + "step": 849 + }, + { + "epoch": 0.26805424156417534, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 3.069, + "step": 850 + }, + { + "epoch": 0.2683695994954273, + "grad_norm": 0.2119140625, + "learning_rate": 0.0002, + "loss": 1.0209, + "step": 851 + }, + { + "epoch": 0.2686849574266793, + "grad_norm": 0.23046875, + "learning_rate": 0.0002, + "loss": 1.09, + "step": 852 + }, + { + "epoch": 0.26900031535793123, + "grad_norm": 0.205078125, + "learning_rate": 0.0002, + "loss": 1.2078, + "step": 853 + }, + { + "epoch": 0.2693156732891832, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.1643, + "step": 854 + }, + { + "epoch": 0.2696310312204352, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 1.1017, + "step": 855 + }, + { + "epoch": 0.26994638915168717, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.1358, + "step": 856 + }, + { + "epoch": 0.27026174708293915, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.0739, + "step": 857 + }, + { + "epoch": 0.27057710501419113, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.2301, + "step": 858 + }, + { + "epoch": 0.27089246294544306, + "grad_norm": 0.158203125, + "learning_rate": 0.0002, + "loss": 1.2157, + "step": 859 + }, + { + "epoch": 0.27120782087669504, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.1419, + "step": 860 + }, + { + "epoch": 0.271523178807947, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.3721, + "step": 861 + }, + { + "epoch": 0.271838536739199, + "grad_norm": 0.1689453125, + "learning_rate": 0.0002, + "loss": 1.2935, + "step": 862 + }, + { + "epoch": 0.272153894670451, + "grad_norm": 0.169921875, + "learning_rate": 0.0002, + "loss": 1.1503, + "step": 863 + }, + { + "epoch": 0.27246925260170296, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.1327, + "step": 864 + }, + { + "epoch": 0.2727846105329549, + "grad_norm": 0.1787109375, + "learning_rate": 0.0002, + "loss": 1.2519, + "step": 865 + }, + { + "epoch": 0.27309996846420687, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 866 + }, + { + "epoch": 0.27341532639545885, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.2043, + "step": 867 + }, + { + "epoch": 0.2737306843267108, + "grad_norm": 0.201171875, + "learning_rate": 0.0002, + "loss": 1.1766, + "step": 868 + }, + { + "epoch": 0.2740460422579628, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 869 + }, + { + "epoch": 0.27436140018921473, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.2378, + "step": 870 + }, + { + "epoch": 0.2746767581204667, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 1.4105, + "step": 871 + }, + { + "epoch": 0.2749921160517187, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.1715, + "step": 872 + }, + { + "epoch": 0.2753074739829707, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 873 + }, + { + "epoch": 0.27562283191422265, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 1.4148, + "step": 874 + }, + { + "epoch": 0.27593818984547464, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.4874, + "step": 875 + }, + { + "epoch": 0.27625354777672656, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.2065, + "step": 876 + }, + { + "epoch": 0.27656890570797854, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.4556, + "step": 877 + }, + { + "epoch": 0.2768842636392305, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 1.2932, + "step": 878 + }, + { + "epoch": 0.2771996215704825, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 1.5853, + "step": 879 + }, + { + "epoch": 0.2775149795017345, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.3843, + "step": 880 + }, + { + "epoch": 0.27783033743298646, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 1.7441, + "step": 881 + }, + { + "epoch": 0.2781456953642384, + "grad_norm": 0.380859375, + "learning_rate": 0.0002, + "loss": 1.6013, + "step": 882 + }, + { + "epoch": 0.27846105329549037, + "grad_norm": 0.400390625, + "learning_rate": 0.0002, + "loss": 1.7603, + "step": 883 + }, + { + "epoch": 0.27877641122674235, + "grad_norm": 0.486328125, + "learning_rate": 0.0002, + "loss": 1.8322, + "step": 884 + }, + { + "epoch": 0.27909176915799433, + "grad_norm": 0.44140625, + "learning_rate": 0.0002, + "loss": 1.6026, + "step": 885 + }, + { + "epoch": 0.2794071270892463, + "grad_norm": 0.5546875, + "learning_rate": 0.0002, + "loss": 1.7468, + "step": 886 + }, + { + "epoch": 0.2797224850204983, + "grad_norm": 0.578125, + "learning_rate": 0.0002, + "loss": 2.063, + "step": 887 + }, + { + "epoch": 0.2800378429517502, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 2.5716, + "step": 888 + }, + { + "epoch": 0.2803532008830022, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 1.8904, + "step": 889 + }, + { + "epoch": 0.2806685588142542, + "grad_norm": 0.73828125, + "learning_rate": 0.0002, + "loss": 2.5135, + "step": 890 + }, + { + "epoch": 0.28098391674550616, + "grad_norm": 0.7109375, + "learning_rate": 0.0002, + "loss": 2.1598, + "step": 891 + }, + { + "epoch": 0.28129927467675814, + "grad_norm": 0.91015625, + "learning_rate": 0.0002, + "loss": 3.089, + "step": 892 + }, + { + "epoch": 0.2816146326080101, + "grad_norm": 0.8203125, + "learning_rate": 0.0002, + "loss": 2.6604, + "step": 893 + }, + { + "epoch": 0.28192999053926204, + "grad_norm": 0.73046875, + "learning_rate": 0.0002, + "loss": 2.431, + "step": 894 + }, + { + "epoch": 0.282245348470514, + "grad_norm": 0.734375, + "learning_rate": 0.0002, + "loss": 2.4977, + "step": 895 + }, + { + "epoch": 0.282560706401766, + "grad_norm": 0.71875, + "learning_rate": 0.0002, + "loss": 2.4133, + "step": 896 + }, + { + "epoch": 0.282876064333018, + "grad_norm": 0.6875, + "learning_rate": 0.0002, + "loss": 2.4329, + "step": 897 + }, + { + "epoch": 0.28319142226426997, + "grad_norm": 0.796875, + "learning_rate": 0.0002, + "loss": 2.6003, + "step": 898 + }, + { + "epoch": 0.2835067801955219, + "grad_norm": 0.734375, + "learning_rate": 0.0002, + "loss": 2.1453, + "step": 899 + }, + { + "epoch": 0.28382213812677387, + "grad_norm": 1.09375, + "learning_rate": 0.0002, + "loss": 2.8257, + "step": 900 + }, + { + "epoch": 0.28413749605802585, + "grad_norm": 0.2001953125, + "learning_rate": 0.0002, + "loss": 1.0886, + "step": 901 + }, + { + "epoch": 0.28445285398927783, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.1039, + "step": 902 + }, + { + "epoch": 0.2847682119205298, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 1.1888, + "step": 903 + }, + { + "epoch": 0.2850835698517818, + "grad_norm": 0.224609375, + "learning_rate": 0.0002, + "loss": 0.9995, + "step": 904 + }, + { + "epoch": 0.2853989277830337, + "grad_norm": 0.2060546875, + "learning_rate": 0.0002, + "loss": 1.1286, + "step": 905 + }, + { + "epoch": 0.2857142857142857, + "grad_norm": 0.212890625, + "learning_rate": 0.0002, + "loss": 1.1795, + "step": 906 + }, + { + "epoch": 0.2860296436455377, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 1.1063, + "step": 907 + }, + { + "epoch": 0.28634500157678966, + "grad_norm": 0.1884765625, + "learning_rate": 0.0002, + "loss": 1.2697, + "step": 908 + }, + { + "epoch": 0.28666035950804164, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.1432, + "step": 909 + }, + { + "epoch": 0.2869757174392936, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.1289, + "step": 910 + }, + { + "epoch": 0.28729107537054555, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.1086, + "step": 911 + }, + { + "epoch": 0.2876064333017975, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.2419, + "step": 912 + }, + { + "epoch": 0.2879217912330495, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.2254, + "step": 913 + }, + { + "epoch": 0.2882371491643015, + "grad_norm": 0.1767578125, + "learning_rate": 0.0002, + "loss": 1.1627, + "step": 914 + }, + { + "epoch": 0.28855250709555347, + "grad_norm": 0.2138671875, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 915 + }, + { + "epoch": 0.28886786502680545, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 1.1689, + "step": 916 + }, + { + "epoch": 0.2891832229580574, + "grad_norm": 0.1962890625, + "learning_rate": 0.0002, + "loss": 1.2676, + "step": 917 + }, + { + "epoch": 0.28949858088930935, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 918 + }, + { + "epoch": 0.28981393882056133, + "grad_norm": 0.205078125, + "learning_rate": 0.0002, + "loss": 1.2576, + "step": 919 + }, + { + "epoch": 0.2901292967518133, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.3535, + "step": 920 + }, + { + "epoch": 0.2904446546830653, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.4115, + "step": 921 + }, + { + "epoch": 0.2907600126143173, + "grad_norm": 0.2216796875, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 922 + }, + { + "epoch": 0.2910753705455692, + "grad_norm": 0.22265625, + "learning_rate": 0.0002, + "loss": 1.1922, + "step": 923 + }, + { + "epoch": 0.2913907284768212, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 1.1553, + "step": 924 + }, + { + "epoch": 0.29170608640807316, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 1.2675, + "step": 925 + }, + { + "epoch": 0.29202144433932514, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 926 + }, + { + "epoch": 0.2923368022705771, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.4806, + "step": 927 + }, + { + "epoch": 0.2926521602018291, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 1.6807, + "step": 928 + }, + { + "epoch": 0.29296751813308103, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 1.3947, + "step": 929 + }, + { + "epoch": 0.293282876064333, + "grad_norm": 0.451171875, + "learning_rate": 0.0002, + "loss": 1.6014, + "step": 930 + }, + { + "epoch": 0.293598233995585, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 931 + }, + { + "epoch": 0.29391359192683697, + "grad_norm": 0.396484375, + "learning_rate": 0.0002, + "loss": 1.8164, + "step": 932 + }, + { + "epoch": 0.29422894985808895, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 1.7173, + "step": 933 + }, + { + "epoch": 0.2945443077893409, + "grad_norm": 0.43359375, + "learning_rate": 0.0002, + "loss": 1.6882, + "step": 934 + }, + { + "epoch": 0.29485966572059286, + "grad_norm": 0.44140625, + "learning_rate": 0.0002, + "loss": 1.76, + "step": 935 + }, + { + "epoch": 0.29485966572059286, + "eval_loss": 1.6987413167953491, + "eval_runtime": 148.3427, + "eval_samples_per_second": 6.741, + "eval_steps_per_second": 6.741, + "step": 935 + }, + { + "epoch": 0.29485966572059286, + "mmlu_eval_accuracy": 0.6310679765974508, + "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, + "mmlu_eval_accuracy_anatomy": 0.6428571428571429, + "mmlu_eval_accuracy_astronomy": 0.6875, + "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, + "mmlu_eval_accuracy_clinical_knowledge": 0.6896551724137931, + "mmlu_eval_accuracy_college_biology": 0.8125, + "mmlu_eval_accuracy_college_chemistry": 0.5, + "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, + "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453, + "mmlu_eval_accuracy_college_medicine": 0.6818181818181818, + "mmlu_eval_accuracy_college_physics": 0.7272727272727273, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.6153846153846154, + "mmlu_eval_accuracy_econometrics": 0.4166666666666667, + "mmlu_eval_accuracy_electrical_engineering": 0.625, + "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634, + "mmlu_eval_accuracy_formal_logic": 0.5714285714285714, + "mmlu_eval_accuracy_global_facts": 0.3, + "mmlu_eval_accuracy_high_school_biology": 0.65625, + "mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182, + "mmlu_eval_accuracy_high_school_computer_science": 0.7777777777777778, + "mmlu_eval_accuracy_high_school_european_history": 0.8888888888888888, + "mmlu_eval_accuracy_high_school_geography": 0.9545454545454546, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.6046511627906976, + "mmlu_eval_accuracy_high_school_mathematics": 0.3103448275862069, + "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, + "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, + "mmlu_eval_accuracy_high_school_psychology": 0.9333333333333333, + "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, + "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, + "mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693, + "mmlu_eval_accuracy_human_aging": 0.6956521739130435, + "mmlu_eval_accuracy_human_sexuality": 0.6666666666666666, + "mmlu_eval_accuracy_international_law": 1.0, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.8, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279, + "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, + "mmlu_eval_accuracy_moral_scenarios": 0.39, + "mmlu_eval_accuracy_nutrition": 0.696969696969697, + "mmlu_eval_accuracy_philosophy": 0.7647058823529411, + "mmlu_eval_accuracy_prehistory": 0.5714285714285714, + "mmlu_eval_accuracy_professional_accounting": 0.2903225806451613, + "mmlu_eval_accuracy_professional_law": 0.4176470588235294, + "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806, + "mmlu_eval_accuracy_professional_psychology": 0.6811594202898551, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7777777777777778, + "mmlu_eval_accuracy_sociology": 1.0, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.4444444444444444, + "mmlu_eval_accuracy_world_religions": 0.8947368421052632, + "mmlu_loss": 6.151379668735975, + "step": 935 + }, + { + "epoch": 0.29517502365184484, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 1.8902, + "step": 936 + }, + { + "epoch": 0.2954903815830968, + "grad_norm": 0.640625, + "learning_rate": 0.0002, + "loss": 2.1017, + "step": 937 + }, + { + "epoch": 0.2958057395143488, + "grad_norm": 0.6953125, + "learning_rate": 0.0002, + "loss": 2.0972, + "step": 938 + }, + { + "epoch": 0.2961210974456008, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 1.8734, + "step": 939 + }, + { + "epoch": 0.2964364553768527, + "grad_norm": 0.6953125, + "learning_rate": 0.0002, + "loss": 2.1895, + "step": 940 + }, + { + "epoch": 0.2967518133081047, + "grad_norm": 0.69921875, + "learning_rate": 0.0002, + "loss": 2.407, + "step": 941 + }, + { + "epoch": 0.29706717123935666, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 2.3273, + "step": 942 + }, + { + "epoch": 0.29738252917060864, + "grad_norm": 0.9453125, + "learning_rate": 0.0002, + "loss": 2.7966, + "step": 943 + }, + { + "epoch": 0.2976978871018606, + "grad_norm": 0.875, + "learning_rate": 0.0002, + "loss": 2.6099, + "step": 944 + }, + { + "epoch": 0.2980132450331126, + "grad_norm": 0.75, + "learning_rate": 0.0002, + "loss": 2.4427, + "step": 945 + }, + { + "epoch": 0.29832860296436453, + "grad_norm": 1.0546875, + "learning_rate": 0.0002, + "loss": 3.5831, + "step": 946 + }, + { + "epoch": 0.2986439608956165, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 2.1715, + "step": 947 + }, + { + "epoch": 0.2989593188268685, + "grad_norm": 0.7734375, + "learning_rate": 0.0002, + "loss": 2.3845, + "step": 948 + }, + { + "epoch": 0.2992746767581205, + "grad_norm": 0.80078125, + "learning_rate": 0.0002, + "loss": 2.714, + "step": 949 + }, + { + "epoch": 0.29959003468937245, + "grad_norm": 1.2734375, + "learning_rate": 0.0002, + "loss": 2.8706, + "step": 950 + }, + { + "epoch": 0.29990539262062443, + "grad_norm": 0.1865234375, + "learning_rate": 0.0002, + "loss": 1.1656, + "step": 951 + }, + { + "epoch": 0.30022075055187636, + "grad_norm": 0.220703125, + "learning_rate": 0.0002, + "loss": 1.1554, + "step": 952 + }, + { + "epoch": 0.30053610848312834, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.0394, + "step": 953 + }, + { + "epoch": 0.3008514664143803, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.0364, + "step": 954 + }, + { + "epoch": 0.3011668243456323, + "grad_norm": 0.1728515625, + "learning_rate": 0.0002, + "loss": 1.1327, + "step": 955 + }, + { + "epoch": 0.3014821822768843, + "grad_norm": 0.169921875, + "learning_rate": 0.0002, + "loss": 1.1433, + "step": 956 + }, + { + "epoch": 0.30179754020813626, + "grad_norm": 0.1591796875, + "learning_rate": 0.0002, + "loss": 1.1342, + "step": 957 + }, + { + "epoch": 0.3021128981393882, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.0749, + "step": 958 + }, + { + "epoch": 0.30242825607064017, + "grad_norm": 0.1728515625, + "learning_rate": 0.0002, + "loss": 1.2273, + "step": 959 + }, + { + "epoch": 0.30274361400189215, + "grad_norm": 0.1728515625, + "learning_rate": 0.0002, + "loss": 1.2461, + "step": 960 + }, + { + "epoch": 0.30305897193314413, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.1182, + "step": 961 + }, + { + "epoch": 0.3033743298643961, + "grad_norm": 0.177734375, + "learning_rate": 0.0002, + "loss": 1.0462, + "step": 962 + }, + { + "epoch": 0.30368968779564803, + "grad_norm": 0.1728515625, + "learning_rate": 0.0002, + "loss": 1.2163, + "step": 963 + }, + { + "epoch": 0.3040050457269, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.1917, + "step": 964 + }, + { + "epoch": 0.304320403658152, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.062, + "step": 965 + }, + { + "epoch": 0.304635761589404, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.249, + "step": 966 + }, + { + "epoch": 0.30495111952065596, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.1723, + "step": 967 + }, + { + "epoch": 0.30526647745190794, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.2137, + "step": 968 + }, + { + "epoch": 0.30558183538315986, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 1.1369, + "step": 969 + }, + { + "epoch": 0.30589719331441184, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 970 + }, + { + "epoch": 0.3062125512456638, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.5046, + "step": 971 + }, + { + "epoch": 0.3065279091769158, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 972 + }, + { + "epoch": 0.3068432671081678, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.2975, + "step": 973 + }, + { + "epoch": 0.30715862503941976, + "grad_norm": 0.2412109375, + "learning_rate": 0.0002, + "loss": 1.2504, + "step": 974 + }, + { + "epoch": 0.3074739829706717, + "grad_norm": 0.2333984375, + "learning_rate": 0.0002, + "loss": 1.4024, + "step": 975 + }, + { + "epoch": 0.30778934090192367, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 1.4981, + "step": 976 + }, + { + "epoch": 0.30810469883317565, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 1.4808, + "step": 977 + }, + { + "epoch": 0.30842005676442763, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.256, + "step": 978 + }, + { + "epoch": 0.3087354146956796, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 979 + }, + { + "epoch": 0.3090507726269316, + "grad_norm": 0.41796875, + "learning_rate": 0.0002, + "loss": 1.5887, + "step": 980 + }, + { + "epoch": 0.3093661305581835, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 1.6923, + "step": 981 + }, + { + "epoch": 0.3096814884894355, + "grad_norm": 0.40234375, + "learning_rate": 0.0002, + "loss": 1.8262, + "step": 982 + }, + { + "epoch": 0.3099968464206875, + "grad_norm": 0.431640625, + "learning_rate": 0.0002, + "loss": 1.8143, + "step": 983 + }, + { + "epoch": 0.31031220435193946, + "grad_norm": 0.4140625, + "learning_rate": 0.0002, + "loss": 1.5581, + "step": 984 + }, + { + "epoch": 0.31062756228319144, + "grad_norm": 0.52734375, + "learning_rate": 0.0002, + "loss": 2.1587, + "step": 985 + }, + { + "epoch": 0.3109429202144434, + "grad_norm": 0.498046875, + "learning_rate": 0.0002, + "loss": 1.7537, + "step": 986 + }, + { + "epoch": 0.31125827814569534, + "grad_norm": 0.5546875, + "learning_rate": 0.0002, + "loss": 2.1277, + "step": 987 + }, + { + "epoch": 0.3115736360769473, + "grad_norm": 0.63671875, + "learning_rate": 0.0002, + "loss": 2.3359, + "step": 988 + }, + { + "epoch": 0.3118889940081993, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 2.4108, + "step": 989 + }, + { + "epoch": 0.3122043519394513, + "grad_norm": 0.609375, + "learning_rate": 0.0002, + "loss": 2.2087, + "step": 990 + }, + { + "epoch": 0.31251970987070327, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 2.0836, + "step": 991 + }, + { + "epoch": 0.31283506780195525, + "grad_norm": 0.63671875, + "learning_rate": 0.0002, + "loss": 2.1081, + "step": 992 + }, + { + "epoch": 0.31315042573320717, + "grad_norm": 0.7421875, + "learning_rate": 0.0002, + "loss": 2.3995, + "step": 993 + }, + { + "epoch": 0.31346578366445915, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 2.0366, + "step": 994 + }, + { + "epoch": 0.31378114159571113, + "grad_norm": 0.6640625, + "learning_rate": 0.0002, + "loss": 2.4537, + "step": 995 + }, + { + "epoch": 0.3140964995269631, + "grad_norm": 0.94921875, + "learning_rate": 0.0002, + "loss": 2.8334, + "step": 996 + }, + { + "epoch": 0.3144118574582151, + "grad_norm": 0.76171875, + "learning_rate": 0.0002, + "loss": 2.4439, + "step": 997 + }, + { + "epoch": 0.314727215389467, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 2.0968, + "step": 998 + }, + { + "epoch": 0.315042573320719, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 1.9517, + "step": 999 + }, + { + "epoch": 0.315357931251971, + "grad_norm": 0.83984375, + "learning_rate": 0.0002, + "loss": 2.4547, + "step": 1000 + }, + { + "epoch": 0.31567328918322296, + "grad_norm": 0.1796875, + "learning_rate": 0.0002, + "loss": 1.1035, + "step": 1001 + }, + { + "epoch": 0.31598864711447494, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.1006, + "step": 1002 + }, + { + "epoch": 0.3163040050457269, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.0528, + "step": 1003 + }, + { + "epoch": 0.31661936297697885, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.0408, + "step": 1004 + }, + { + "epoch": 0.3169347209082308, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.0877, + "step": 1005 + }, + { + "epoch": 0.3172500788394828, + "grad_norm": 0.1630859375, + "learning_rate": 0.0002, + "loss": 1.1418, + "step": 1006 + }, + { + "epoch": 0.3175654367707348, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.0568, + "step": 1007 + }, + { + "epoch": 0.31788079470198677, + "grad_norm": 0.1552734375, + "learning_rate": 0.0002, + "loss": 1.0988, + "step": 1008 + }, + { + "epoch": 0.31819615263323875, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.1228, + "step": 1009 + }, + { + "epoch": 0.3185115105644907, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.2198, + "step": 1010 + }, + { + "epoch": 0.31882686849574265, + "grad_norm": 0.1552734375, + "learning_rate": 0.0002, + "loss": 1.1115, + "step": 1011 + }, + { + "epoch": 0.31914222642699464, + "grad_norm": 0.1591796875, + "learning_rate": 0.0002, + "loss": 1.081, + "step": 1012 + }, + { + "epoch": 0.3194575843582466, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.1508, + "step": 1013 + }, + { + "epoch": 0.3197729422894986, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.1753, + "step": 1014 + }, + { + "epoch": 0.3200883002207506, + "grad_norm": 0.1787109375, + "learning_rate": 0.0002, + "loss": 1.1638, + "step": 1015 + }, + { + "epoch": 0.3204036581520025, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.1645, + "step": 1016 + }, + { + "epoch": 0.3207190160832545, + "grad_norm": 0.1689453125, + "learning_rate": 0.0002, + "loss": 1.1071, + "step": 1017 + }, + { + "epoch": 0.32103437401450646, + "grad_norm": 0.19140625, + "learning_rate": 0.0002, + "loss": 1.2655, + "step": 1018 + }, + { + "epoch": 0.32134973194575844, + "grad_norm": 0.185546875, + "learning_rate": 0.0002, + "loss": 1.2028, + "step": 1019 + }, + { + "epoch": 0.3216650898770104, + "grad_norm": 0.220703125, + "learning_rate": 0.0002, + "loss": 1.427, + "step": 1020 + }, + { + "epoch": 0.3219804478082624, + "grad_norm": 0.2451171875, + "learning_rate": 0.0002, + "loss": 1.2988, + "step": 1021 + }, + { + "epoch": 0.32229580573951433, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 1.2846, + "step": 1022 + }, + { + "epoch": 0.3226111636707663, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.2383, + "step": 1023 + }, + { + "epoch": 0.3229265216020183, + "grad_norm": 0.2490234375, + "learning_rate": 0.0002, + "loss": 1.3921, + "step": 1024 + }, + { + "epoch": 0.32324187953327027, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 1.5095, + "step": 1025 + }, + { + "epoch": 0.32355723746452225, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.2815, + "step": 1026 + }, + { + "epoch": 0.3238725953957742, + "grad_norm": 0.2470703125, + "learning_rate": 0.0002, + "loss": 1.3008, + "step": 1027 + }, + { + "epoch": 0.32418795332702616, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 1.3993, + "step": 1028 + }, + { + "epoch": 0.32450331125827814, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 1.5904, + "step": 1029 + }, + { + "epoch": 0.3248186691895301, + "grad_norm": 0.330078125, + "learning_rate": 0.0002, + "loss": 1.8316, + "step": 1030 + }, + { + "epoch": 0.3251340271207821, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 1.5232, + "step": 1031 + }, + { + "epoch": 0.3254493850520341, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 1.5904, + "step": 1032 + }, + { + "epoch": 0.325764742983286, + "grad_norm": 0.39453125, + "learning_rate": 0.0002, + "loss": 1.8687, + "step": 1033 + }, + { + "epoch": 0.326080100914538, + "grad_norm": 0.427734375, + "learning_rate": 0.0002, + "loss": 1.8961, + "step": 1034 + }, + { + "epoch": 0.32639545884578997, + "grad_norm": 0.44140625, + "learning_rate": 0.0002, + "loss": 1.9137, + "step": 1035 + }, + { + "epoch": 0.32671081677704195, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 2.0847, + "step": 1036 + }, + { + "epoch": 0.3270261747082939, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 1.9891, + "step": 1037 + }, + { + "epoch": 0.3273415326395459, + "grad_norm": 0.5546875, + "learning_rate": 0.0002, + "loss": 2.047, + "step": 1038 + }, + { + "epoch": 0.32765689057079783, + "grad_norm": 0.82421875, + "learning_rate": 0.0002, + "loss": 2.4104, + "step": 1039 + }, + { + "epoch": 0.3279722485020498, + "grad_norm": 0.52734375, + "learning_rate": 0.0002, + "loss": 2.2347, + "step": 1040 + }, + { + "epoch": 0.3282876064333018, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 2.0205, + "step": 1041 + }, + { + "epoch": 0.3286029643645538, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 2.0693, + "step": 1042 + }, + { + "epoch": 0.32891832229580575, + "grad_norm": 0.62109375, + "learning_rate": 0.0002, + "loss": 2.0299, + "step": 1043 + }, + { + "epoch": 0.32923368022705773, + "grad_norm": 0.75390625, + "learning_rate": 0.0002, + "loss": 2.4705, + "step": 1044 + }, + { + "epoch": 0.32954903815830966, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 2.1362, + "step": 1045 + }, + { + "epoch": 0.32986439608956164, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 2.5837, + "step": 1046 + }, + { + "epoch": 0.3301797540208136, + "grad_norm": 0.90234375, + "learning_rate": 0.0002, + "loss": 2.9068, + "step": 1047 + }, + { + "epoch": 0.3304951119520656, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 2.208, + "step": 1048 + }, + { + "epoch": 0.3308104698833176, + "grad_norm": 0.921875, + "learning_rate": 0.0002, + "loss": 2.5448, + "step": 1049 + }, + { + "epoch": 0.33112582781456956, + "grad_norm": 1.09375, + "learning_rate": 0.0002, + "loss": 3.0281, + "step": 1050 + }, + { + "epoch": 0.3314411857458215, + "grad_norm": 0.1552734375, + "learning_rate": 0.0002, + "loss": 1.1472, + "step": 1051 + }, + { + "epoch": 0.33175654367707347, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.0056, + "step": 1052 + }, + { + "epoch": 0.33207190160832545, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.208, + "step": 1053 + }, + { + "epoch": 0.33238725953957743, + "grad_norm": 0.169921875, + "learning_rate": 0.0002, + "loss": 1.1551, + "step": 1054 + }, + { + "epoch": 0.3327026174708294, + "grad_norm": 0.1591796875, + "learning_rate": 0.0002, + "loss": 1.2248, + "step": 1055 + }, + { + "epoch": 0.3330179754020814, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.1689, + "step": 1056 + }, + { + "epoch": 0.3333333333333333, + "grad_norm": 0.19140625, + "learning_rate": 0.0002, + "loss": 1.2391, + "step": 1057 + }, + { + "epoch": 0.3336486912645853, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.0559, + "step": 1058 + }, + { + "epoch": 0.3339640491958373, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.0931, + "step": 1059 + }, + { + "epoch": 0.33427940712708926, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.193, + "step": 1060 + }, + { + "epoch": 0.33459476505834124, + "grad_norm": 0.154296875, + "learning_rate": 0.0002, + "loss": 1.0445, + "step": 1061 + }, + { + "epoch": 0.33491012298959316, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.1887, + "step": 1062 + }, + { + "epoch": 0.33522548092084514, + "grad_norm": 0.1728515625, + "learning_rate": 0.0002, + "loss": 1.2642, + "step": 1063 + }, + { + "epoch": 0.3355408388520971, + "grad_norm": 0.1591796875, + "learning_rate": 0.0002, + "loss": 1.0987, + "step": 1064 + }, + { + "epoch": 0.3358561967833491, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.1857, + "step": 1065 + }, + { + "epoch": 0.3361715547146011, + "grad_norm": 0.19921875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 1066 + }, + { + "epoch": 0.33648691264585306, + "grad_norm": 0.1943359375, + "learning_rate": 0.0002, + "loss": 1.1557, + "step": 1067 + }, + { + "epoch": 0.336802270577105, + "grad_norm": 0.2138671875, + "learning_rate": 0.0002, + "loss": 1.3632, + "step": 1068 + }, + { + "epoch": 0.33711762850835697, + "grad_norm": 0.2392578125, + "learning_rate": 0.0002, + "loss": 1.2158, + "step": 1069 + }, + { + "epoch": 0.33743298643960895, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.2439, + "step": 1070 + }, + { + "epoch": 0.33774834437086093, + "grad_norm": 0.232421875, + "learning_rate": 0.0002, + "loss": 1.3798, + "step": 1071 + }, + { + "epoch": 0.3380637023021129, + "grad_norm": 0.216796875, + "learning_rate": 0.0002, + "loss": 1.2986, + "step": 1072 + }, + { + "epoch": 0.3383790602333649, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.398, + "step": 1073 + }, + { + "epoch": 0.3386944181646168, + "grad_norm": 0.255859375, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 1074 + }, + { + "epoch": 0.3390097760958688, + "grad_norm": 0.2412109375, + "learning_rate": 0.0002, + "loss": 1.1649, + "step": 1075 + }, + { + "epoch": 0.3393251340271208, + "grad_norm": 0.365234375, + "learning_rate": 0.0002, + "loss": 1.8379, + "step": 1076 + }, + { + "epoch": 0.33964049195837276, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.271, + "step": 1077 + }, + { + "epoch": 0.33995584988962474, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 1.5615, + "step": 1078 + }, + { + "epoch": 0.3402712078208767, + "grad_norm": 0.37890625, + "learning_rate": 0.0002, + "loss": 1.5432, + "step": 1079 + }, + { + "epoch": 0.34058656575212864, + "grad_norm": 0.384765625, + "learning_rate": 0.0002, + "loss": 1.4448, + "step": 1080 + }, + { + "epoch": 0.3409019236833806, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 1.7414, + "step": 1081 + }, + { + "epoch": 0.3412172816146326, + "grad_norm": 0.62890625, + "learning_rate": 0.0002, + "loss": 2.0819, + "step": 1082 + }, + { + "epoch": 0.3415326395458846, + "grad_norm": 0.4765625, + "learning_rate": 0.0002, + "loss": 2.2139, + "step": 1083 + }, + { + "epoch": 0.34184799747713657, + "grad_norm": 0.419921875, + "learning_rate": 0.0002, + "loss": 1.7557, + "step": 1084 + }, + { + "epoch": 0.34216335540838855, + "grad_norm": 0.453125, + "learning_rate": 0.0002, + "loss": 1.7697, + "step": 1085 + }, + { + "epoch": 0.3424787133396405, + "grad_norm": 0.58203125, + "learning_rate": 0.0002, + "loss": 2.1897, + "step": 1086 + }, + { + "epoch": 0.34279407127089245, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 1.9514, + "step": 1087 + }, + { + "epoch": 0.34310942920214443, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 2.6291, + "step": 1088 + }, + { + "epoch": 0.3434247871333964, + "grad_norm": 0.49609375, + "learning_rate": 0.0002, + "loss": 2.112, + "step": 1089 + }, + { + "epoch": 0.3437401450646484, + "grad_norm": 0.65625, + "learning_rate": 0.0002, + "loss": 2.3361, + "step": 1090 + }, + { + "epoch": 0.3440555029959003, + "grad_norm": 0.69921875, + "learning_rate": 0.0002, + "loss": 2.4282, + "step": 1091 + }, + { + "epoch": 0.3443708609271523, + "grad_norm": 0.6171875, + "learning_rate": 0.0002, + "loss": 2.1705, + "step": 1092 + }, + { + "epoch": 0.3446862188584043, + "grad_norm": 0.6796875, + "learning_rate": 0.0002, + "loss": 2.2899, + "step": 1093 + }, + { + "epoch": 0.34500157678965626, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 2.7504, + "step": 1094 + }, + { + "epoch": 0.34531693472090824, + "grad_norm": 0.828125, + "learning_rate": 0.0002, + "loss": 2.6721, + "step": 1095 + }, + { + "epoch": 0.3456322926521602, + "grad_norm": 0.984375, + "learning_rate": 0.0002, + "loss": 2.8837, + "step": 1096 + }, + { + "epoch": 0.34594765058341215, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 2.1712, + "step": 1097 + }, + { + "epoch": 0.34626300851466413, + "grad_norm": 0.7421875, + "learning_rate": 0.0002, + "loss": 2.5248, + "step": 1098 + }, + { + "epoch": 0.3465783664459161, + "grad_norm": 0.60546875, + "learning_rate": 0.0002, + "loss": 1.7639, + "step": 1099 + }, + { + "epoch": 0.3468937243771681, + "grad_norm": 0.8984375, + "learning_rate": 0.0002, + "loss": 2.3044, + "step": 1100 + }, + { + "epoch": 0.34720908230842007, + "grad_norm": 0.1875, + "learning_rate": 0.0002, + "loss": 1.0793, + "step": 1101 + }, + { + "epoch": 0.34752444023967205, + "grad_norm": 0.1455078125, + "learning_rate": 0.0002, + "loss": 1.1471, + "step": 1102 + }, + { + "epoch": 0.347839798170924, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.2097, + "step": 1103 + }, + { + "epoch": 0.34815515610217596, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.1721, + "step": 1104 + }, + { + "epoch": 0.34847051403342794, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.0995, + "step": 1105 + }, + { + "epoch": 0.3487858719646799, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.1137, + "step": 1106 + }, + { + "epoch": 0.3491012298959319, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.1784, + "step": 1107 + }, + { + "epoch": 0.3494165878271839, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.2189, + "step": 1108 + }, + { + "epoch": 0.3497319457584358, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.2041, + "step": 1109 + }, + { + "epoch": 0.3500473036896878, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 1.1634, + "step": 1110 + }, + { + "epoch": 0.35036266162093976, + "grad_norm": 0.189453125, + "learning_rate": 0.0002, + "loss": 1.1733, + "step": 1111 + }, + { + "epoch": 0.35067801955219174, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.0755, + "step": 1112 + }, + { + "epoch": 0.3509933774834437, + "grad_norm": 0.1865234375, + "learning_rate": 0.0002, + "loss": 1.1514, + "step": 1113 + }, + { + "epoch": 0.3513087354146957, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 1.2379, + "step": 1114 + }, + { + "epoch": 0.35162409334594763, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 1115 + }, + { + "epoch": 0.3519394512771996, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 1.1726, + "step": 1116 + }, + { + "epoch": 0.3522548092084516, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 1.6044, + "step": 1117 + }, + { + "epoch": 0.35257016713970357, + "grad_norm": 0.2060546875, + "learning_rate": 0.0002, + "loss": 1.1657, + "step": 1118 + }, + { + "epoch": 0.35288552507095555, + "grad_norm": 0.193359375, + "learning_rate": 0.0002, + "loss": 1.1784, + "step": 1119 + }, + { + "epoch": 0.35320088300220753, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 1.1464, + "step": 1120 + }, + { + "epoch": 0.35351624093345946, + "grad_norm": 0.2314453125, + "learning_rate": 0.0002, + "loss": 1.4982, + "step": 1121 + }, + { + "epoch": 0.35383159886471144, + "grad_norm": 0.2392578125, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 1122 + }, + { + "epoch": 0.35383159886471144, + "eval_loss": 1.6802332401275635, + "eval_runtime": 149.0234, + "eval_samples_per_second": 6.71, + "eval_steps_per_second": 6.71, + "step": 1122 + }, + { + "epoch": 0.35383159886471144, + "mmlu_eval_accuracy": 0.628051156056193, + "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182, + "mmlu_eval_accuracy_anatomy": 0.5, + "mmlu_eval_accuracy_astronomy": 0.75, + "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, + "mmlu_eval_accuracy_clinical_knowledge": 0.7241379310344828, + "mmlu_eval_accuracy_college_biology": 0.75, + "mmlu_eval_accuracy_college_chemistry": 0.5, + "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, + "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453, + "mmlu_eval_accuracy_college_medicine": 0.7272727272727273, + "mmlu_eval_accuracy_college_physics": 0.7272727272727273, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.6153846153846154, + "mmlu_eval_accuracy_econometrics": 0.5, + "mmlu_eval_accuracy_electrical_engineering": 0.6875, + "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634, + "mmlu_eval_accuracy_formal_logic": 0.5714285714285714, + "mmlu_eval_accuracy_global_facts": 0.3, + "mmlu_eval_accuracy_high_school_biology": 0.65625, + "mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.8888888888888888, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209, + "mmlu_eval_accuracy_high_school_mathematics": 0.3448275862068966, + "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, + "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, + "mmlu_eval_accuracy_high_school_psychology": 0.9166666666666666, + "mmlu_eval_accuracy_high_school_statistics": 0.4782608695652174, + "mmlu_eval_accuracy_high_school_us_history": 0.8636363636363636, + "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, + "mmlu_eval_accuracy_human_aging": 0.6956521739130435, + "mmlu_eval_accuracy_human_sexuality": 0.6666666666666666, + "mmlu_eval_accuracy_international_law": 1.0, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.8, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7558139534883721, + "mmlu_eval_accuracy_moral_disputes": 0.47368421052631576, + "mmlu_eval_accuracy_moral_scenarios": 0.38, + "mmlu_eval_accuracy_nutrition": 0.7272727272727273, + "mmlu_eval_accuracy_philosophy": 0.7352941176470589, + "mmlu_eval_accuracy_prehistory": 0.5714285714285714, + "mmlu_eval_accuracy_professional_accounting": 0.25806451612903225, + "mmlu_eval_accuracy_professional_law": 0.4411764705882353, + "mmlu_eval_accuracy_professional_medicine": 0.7096774193548387, + "mmlu_eval_accuracy_professional_psychology": 0.6666666666666666, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7777777777777778, + "mmlu_eval_accuracy_sociology": 0.9545454545454546, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.5, + "mmlu_eval_accuracy_world_religions": 0.8947368421052632, + "mmlu_loss": 6.426044149230618, + "step": 1122 + }, + { + "epoch": 0.3541469567959634, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 1.2494, + "step": 1123 + }, + { + "epoch": 0.3544623147272154, + "grad_norm": 0.248046875, + "learning_rate": 0.0002, + "loss": 1.291, + "step": 1124 + }, + { + "epoch": 0.3547776726584674, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.4583, + "step": 1125 + }, + { + "epoch": 0.3550930305897193, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 1.3038, + "step": 1126 + }, + { + "epoch": 0.3554083885209713, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 1127 + }, + { + "epoch": 0.35572374645222327, + "grad_norm": 0.349609375, + "learning_rate": 0.0002, + "loss": 1.601, + "step": 1128 + }, + { + "epoch": 0.35603910438347525, + "grad_norm": 0.38671875, + "learning_rate": 0.0002, + "loss": 1.6901, + "step": 1129 + }, + { + "epoch": 0.3563544623147272, + "grad_norm": 0.404296875, + "learning_rate": 0.0002, + "loss": 1.8294, + "step": 1130 + }, + { + "epoch": 0.3566698202459792, + "grad_norm": 0.494140625, + "learning_rate": 0.0002, + "loss": 1.5643, + "step": 1131 + }, + { + "epoch": 0.35698517817723113, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 1.6684, + "step": 1132 + }, + { + "epoch": 0.3573005361084831, + "grad_norm": 0.458984375, + "learning_rate": 0.0002, + "loss": 1.9068, + "step": 1133 + }, + { + "epoch": 0.3576158940397351, + "grad_norm": 0.435546875, + "learning_rate": 0.0002, + "loss": 1.9151, + "step": 1134 + }, + { + "epoch": 0.3579312519709871, + "grad_norm": 0.4921875, + "learning_rate": 0.0002, + "loss": 1.7406, + "step": 1135 + }, + { + "epoch": 0.35824660990223905, + "grad_norm": 0.60546875, + "learning_rate": 0.0002, + "loss": 1.9015, + "step": 1136 + }, + { + "epoch": 0.35856196783349104, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 2.1666, + "step": 1137 + }, + { + "epoch": 0.35887732576474296, + "grad_norm": 0.62890625, + "learning_rate": 0.0002, + "loss": 2.4426, + "step": 1138 + }, + { + "epoch": 0.35919268369599494, + "grad_norm": 0.58203125, + "learning_rate": 0.0002, + "loss": 2.2136, + "step": 1139 + }, + { + "epoch": 0.3595080416272469, + "grad_norm": 0.578125, + "learning_rate": 0.0002, + "loss": 2.33, + "step": 1140 + }, + { + "epoch": 0.3598233995584989, + "grad_norm": 0.80859375, + "learning_rate": 0.0002, + "loss": 2.7172, + "step": 1141 + }, + { + "epoch": 0.3601387574897509, + "grad_norm": 0.65625, + "learning_rate": 0.0002, + "loss": 2.3408, + "step": 1142 + }, + { + "epoch": 0.36045411542100286, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 1.9673, + "step": 1143 + }, + { + "epoch": 0.3607694733522548, + "grad_norm": 0.78125, + "learning_rate": 0.0002, + "loss": 2.4385, + "step": 1144 + }, + { + "epoch": 0.36108483128350677, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 2.1918, + "step": 1145 + }, + { + "epoch": 0.36140018921475875, + "grad_norm": 0.953125, + "learning_rate": 0.0002, + "loss": 2.4936, + "step": 1146 + }, + { + "epoch": 0.36171554714601073, + "grad_norm": 0.703125, + "learning_rate": 0.0002, + "loss": 2.3361, + "step": 1147 + }, + { + "epoch": 0.3620309050772627, + "grad_norm": 0.71875, + "learning_rate": 0.0002, + "loss": 2.1521, + "step": 1148 + }, + { + "epoch": 0.3623462630085147, + "grad_norm": 0.60546875, + "learning_rate": 0.0002, + "loss": 2.0559, + "step": 1149 + }, + { + "epoch": 0.3626616209397666, + "grad_norm": 0.99609375, + "learning_rate": 0.0002, + "loss": 2.676, + "step": 1150 + }, + { + "epoch": 0.3629769788710186, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.1279, + "step": 1151 + }, + { + "epoch": 0.3632923368022706, + "grad_norm": 0.154296875, + "learning_rate": 0.0002, + "loss": 1.0861, + "step": 1152 + }, + { + "epoch": 0.36360769473352256, + "grad_norm": 0.166015625, + "learning_rate": 0.0002, + "loss": 1.0893, + "step": 1153 + }, + { + "epoch": 0.36392305266477454, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.1863, + "step": 1154 + }, + { + "epoch": 0.36423841059602646, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.1304, + "step": 1155 + }, + { + "epoch": 0.36455376852727844, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.1886, + "step": 1156 + }, + { + "epoch": 0.3648691264585304, + "grad_norm": 0.150390625, + "learning_rate": 0.0002, + "loss": 1.0965, + "step": 1157 + }, + { + "epoch": 0.3651844843897824, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.1661, + "step": 1158 + }, + { + "epoch": 0.3654998423210344, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.22, + "step": 1159 + }, + { + "epoch": 0.36581520025228637, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.0527, + "step": 1160 + }, + { + "epoch": 0.3661305581835383, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.264, + "step": 1161 + }, + { + "epoch": 0.36644591611479027, + "grad_norm": 0.169921875, + "learning_rate": 0.0002, + "loss": 1.1003, + "step": 1162 + }, + { + "epoch": 0.36676127404604225, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.1573, + "step": 1163 + }, + { + "epoch": 0.36707663197729423, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 1.1603, + "step": 1164 + }, + { + "epoch": 0.3673919899085462, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.2216, + "step": 1165 + }, + { + "epoch": 0.3677073478397982, + "grad_norm": 0.19140625, + "learning_rate": 0.0002, + "loss": 1.1132, + "step": 1166 + }, + { + "epoch": 0.3680227057710501, + "grad_norm": 0.2275390625, + "learning_rate": 0.0002, + "loss": 1.4086, + "step": 1167 + }, + { + "epoch": 0.3683380637023021, + "grad_norm": 0.2001953125, + "learning_rate": 0.0002, + "loss": 1.2427, + "step": 1168 + }, + { + "epoch": 0.3686534216335541, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.1662, + "step": 1169 + }, + { + "epoch": 0.36896877956480606, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.2729, + "step": 1170 + }, + { + "epoch": 0.36928413749605804, + "grad_norm": 0.2431640625, + "learning_rate": 0.0002, + "loss": 1.2533, + "step": 1171 + }, + { + "epoch": 0.36959949542731, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 1172 + }, + { + "epoch": 0.36991485335856195, + "grad_norm": 0.240234375, + "learning_rate": 0.0002, + "loss": 1.4886, + "step": 1173 + }, + { + "epoch": 0.3702302112898139, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 1174 + }, + { + "epoch": 0.3705455692210659, + "grad_norm": 0.2265625, + "learning_rate": 0.0002, + "loss": 1.1536, + "step": 1175 + }, + { + "epoch": 0.3708609271523179, + "grad_norm": 0.458984375, + "learning_rate": 0.0002, + "loss": 1.5264, + "step": 1176 + }, + { + "epoch": 0.37117628508356987, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 1.374, + "step": 1177 + }, + { + "epoch": 0.37149164301482185, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 1.4885, + "step": 1178 + }, + { + "epoch": 0.3718070009460738, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 1.5624, + "step": 1179 + }, + { + "epoch": 0.37212235887732575, + "grad_norm": 0.3828125, + "learning_rate": 0.0002, + "loss": 1.6148, + "step": 1180 + }, + { + "epoch": 0.37243771680857773, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 1.5494, + "step": 1181 + }, + { + "epoch": 0.3727530747398297, + "grad_norm": 0.4375, + "learning_rate": 0.0002, + "loss": 1.6401, + "step": 1182 + }, + { + "epoch": 0.3730684326710817, + "grad_norm": 0.46484375, + "learning_rate": 0.0002, + "loss": 1.9916, + "step": 1183 + }, + { + "epoch": 0.3733837906023337, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 1.6848, + "step": 1184 + }, + { + "epoch": 0.3736991485335856, + "grad_norm": 0.48828125, + "learning_rate": 0.0002, + "loss": 2.0125, + "step": 1185 + }, + { + "epoch": 0.3740145064648376, + "grad_norm": 0.4453125, + "learning_rate": 0.0002, + "loss": 1.9648, + "step": 1186 + }, + { + "epoch": 0.37432986439608956, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 1.987, + "step": 1187 + }, + { + "epoch": 0.37464522232734154, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 2.1474, + "step": 1188 + }, + { + "epoch": 0.3749605802585935, + "grad_norm": 0.49609375, + "learning_rate": 0.0002, + "loss": 2.0543, + "step": 1189 + }, + { + "epoch": 0.37527593818984545, + "grad_norm": 0.52734375, + "learning_rate": 0.0002, + "loss": 2.0623, + "step": 1190 + }, + { + "epoch": 0.37559129612109743, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 2.4857, + "step": 1191 + }, + { + "epoch": 0.3759066540523494, + "grad_norm": 0.6640625, + "learning_rate": 0.0002, + "loss": 2.4226, + "step": 1192 + }, + { + "epoch": 0.3762220119836014, + "grad_norm": 0.70703125, + "learning_rate": 0.0002, + "loss": 2.5481, + "step": 1193 + }, + { + "epoch": 0.37653736991485337, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 2.3573, + "step": 1194 + }, + { + "epoch": 0.37685272784610535, + "grad_norm": 0.7421875, + "learning_rate": 0.0002, + "loss": 2.5515, + "step": 1195 + }, + { + "epoch": 0.3771680857773573, + "grad_norm": 0.67578125, + "learning_rate": 0.0002, + "loss": 2.0389, + "step": 1196 + }, + { + "epoch": 0.37748344370860926, + "grad_norm": 0.7578125, + "learning_rate": 0.0002, + "loss": 2.8312, + "step": 1197 + }, + { + "epoch": 0.37779880163986124, + "grad_norm": 0.7578125, + "learning_rate": 0.0002, + "loss": 2.3917, + "step": 1198 + }, + { + "epoch": 0.3781141595711132, + "grad_norm": 0.63671875, + "learning_rate": 0.0002, + "loss": 1.782, + "step": 1199 + }, + { + "epoch": 0.3784295175023652, + "grad_norm": 1.3046875, + "learning_rate": 0.0002, + "loss": 3.2281, + "step": 1200 + }, + { + "epoch": 0.3787448754336172, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.178, + "step": 1201 + }, + { + "epoch": 0.3790602333648691, + "grad_norm": 0.150390625, + "learning_rate": 0.0002, + "loss": 1.2577, + "step": 1202 + }, + { + "epoch": 0.3793755912961211, + "grad_norm": 0.1591796875, + "learning_rate": 0.0002, + "loss": 0.9912, + "step": 1203 + }, + { + "epoch": 0.37969094922737306, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.1346, + "step": 1204 + }, + { + "epoch": 0.38000630715862505, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.0881, + "step": 1205 + }, + { + "epoch": 0.380321665089877, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.0612, + "step": 1206 + }, + { + "epoch": 0.380637023021129, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.0697, + "step": 1207 + }, + { + "epoch": 0.38095238095238093, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.2301, + "step": 1208 + }, + { + "epoch": 0.3812677388836329, + "grad_norm": 0.1689453125, + "learning_rate": 0.0002, + "loss": 1.2001, + "step": 1209 + }, + { + "epoch": 0.3815830968148849, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.0266, + "step": 1210 + }, + { + "epoch": 0.3818984547461369, + "grad_norm": 0.1630859375, + "learning_rate": 0.0002, + "loss": 1.0862, + "step": 1211 + }, + { + "epoch": 0.38221381267738885, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.1868, + "step": 1212 + }, + { + "epoch": 0.38252917060864083, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.07, + "step": 1213 + }, + { + "epoch": 0.38284452853989276, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.0034, + "step": 1214 + }, + { + "epoch": 0.38315988647114474, + "grad_norm": 0.166015625, + "learning_rate": 0.0002, + "loss": 1.1299, + "step": 1215 + }, + { + "epoch": 0.3834752444023967, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.1073, + "step": 1216 + }, + { + "epoch": 0.3837906023336487, + "grad_norm": 0.205078125, + "learning_rate": 0.0002, + "loss": 1.1819, + "step": 1217 + }, + { + "epoch": 0.3841059602649007, + "grad_norm": 0.2041015625, + "learning_rate": 0.0002, + "loss": 1.1673, + "step": 1218 + }, + { + "epoch": 0.3844213181961526, + "grad_norm": 0.205078125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 1219 + }, + { + "epoch": 0.3847366761274046, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 1220 + }, + { + "epoch": 0.38505203405865657, + "grad_norm": 0.2490234375, + "learning_rate": 0.0002, + "loss": 1.5638, + "step": 1221 + }, + { + "epoch": 0.38536739198990855, + "grad_norm": 0.2412109375, + "learning_rate": 0.0002, + "loss": 1.5259, + "step": 1222 + }, + { + "epoch": 0.38568274992116053, + "grad_norm": 0.2314453125, + "learning_rate": 0.0002, + "loss": 1.403, + "step": 1223 + }, + { + "epoch": 0.3859981078524125, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 1224 + }, + { + "epoch": 0.38631346578366443, + "grad_norm": 0.2412109375, + "learning_rate": 0.0002, + "loss": 1.2978, + "step": 1225 + }, + { + "epoch": 0.3866288237149164, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 1.7124, + "step": 1226 + }, + { + "epoch": 0.3869441816461684, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 1.3756, + "step": 1227 + }, + { + "epoch": 0.3872595395774204, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 1228 + }, + { + "epoch": 0.38757489750867236, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 1.396, + "step": 1229 + }, + { + "epoch": 0.38789025543992434, + "grad_norm": 0.423828125, + "learning_rate": 0.0002, + "loss": 1.4756, + "step": 1230 + }, + { + "epoch": 0.38820561337117626, + "grad_norm": 0.369140625, + "learning_rate": 0.0002, + "loss": 1.518, + "step": 1231 + }, + { + "epoch": 0.38852097130242824, + "grad_norm": 0.455078125, + "learning_rate": 0.0002, + "loss": 1.8133, + "step": 1232 + }, + { + "epoch": 0.3888363292336802, + "grad_norm": 0.6484375, + "learning_rate": 0.0002, + "loss": 2.3535, + "step": 1233 + }, + { + "epoch": 0.3891516871649322, + "grad_norm": 0.4609375, + "learning_rate": 0.0002, + "loss": 1.8049, + "step": 1234 + }, + { + "epoch": 0.3894670450961842, + "grad_norm": 0.44140625, + "learning_rate": 0.0002, + "loss": 1.7056, + "step": 1235 + }, + { + "epoch": 0.38978240302743616, + "grad_norm": 0.6484375, + "learning_rate": 0.0002, + "loss": 2.3153, + "step": 1236 + }, + { + "epoch": 0.3900977609586881, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 2.3006, + "step": 1237 + }, + { + "epoch": 0.39041311888994007, + "grad_norm": 0.67578125, + "learning_rate": 0.0002, + "loss": 2.5698, + "step": 1238 + }, + { + "epoch": 0.39072847682119205, + "grad_norm": 0.6484375, + "learning_rate": 0.0002, + "loss": 2.0901, + "step": 1239 + }, + { + "epoch": 0.39104383475244403, + "grad_norm": 0.5390625, + "learning_rate": 0.0002, + "loss": 1.9414, + "step": 1240 + }, + { + "epoch": 0.391359192683696, + "grad_norm": 0.73046875, + "learning_rate": 0.0002, + "loss": 2.3068, + "step": 1241 + }, + { + "epoch": 0.391674550614948, + "grad_norm": 0.84375, + "learning_rate": 0.0002, + "loss": 2.4992, + "step": 1242 + }, + { + "epoch": 0.3919899085461999, + "grad_norm": 0.70703125, + "learning_rate": 0.0002, + "loss": 2.5158, + "step": 1243 + }, + { + "epoch": 0.3923052664774519, + "grad_norm": 0.68359375, + "learning_rate": 0.0002, + "loss": 2.3429, + "step": 1244 + }, + { + "epoch": 0.3926206244087039, + "grad_norm": 0.7734375, + "learning_rate": 0.0002, + "loss": 2.4593, + "step": 1245 + }, + { + "epoch": 0.39293598233995586, + "grad_norm": 0.9765625, + "learning_rate": 0.0002, + "loss": 2.5155, + "step": 1246 + }, + { + "epoch": 0.39325134027120784, + "grad_norm": 0.9140625, + "learning_rate": 0.0002, + "loss": 2.5943, + "step": 1247 + }, + { + "epoch": 0.3935666982024598, + "grad_norm": 0.90234375, + "learning_rate": 0.0002, + "loss": 2.5022, + "step": 1248 + }, + { + "epoch": 0.39388205613371174, + "grad_norm": 0.77734375, + "learning_rate": 0.0002, + "loss": 2.5139, + "step": 1249 + }, + { + "epoch": 0.3941974140649637, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 2.4624, + "step": 1250 + }, + { + "epoch": 0.3945127719962157, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.0206, + "step": 1251 + }, + { + "epoch": 0.3948281299274677, + "grad_norm": 0.1591796875, + "learning_rate": 0.0002, + "loss": 1.0781, + "step": 1252 + }, + { + "epoch": 0.39514348785871967, + "grad_norm": 0.1884765625, + "learning_rate": 0.0002, + "loss": 1.1263, + "step": 1253 + }, + { + "epoch": 0.3954588457899716, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 1.0753, + "step": 1254 + }, + { + "epoch": 0.39577420372122357, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.0686, + "step": 1255 + }, + { + "epoch": 0.39608956165247555, + "grad_norm": 0.208984375, + "learning_rate": 0.0002, + "loss": 1.1594, + "step": 1256 + }, + { + "epoch": 0.39640491958372753, + "grad_norm": 0.189453125, + "learning_rate": 0.0002, + "loss": 1.0714, + "step": 1257 + }, + { + "epoch": 0.3967202775149795, + "grad_norm": 0.197265625, + "learning_rate": 0.0002, + "loss": 1.1877, + "step": 1258 + }, + { + "epoch": 0.3970356354462315, + "grad_norm": 0.1728515625, + "learning_rate": 0.0002, + "loss": 1.0445, + "step": 1259 + }, + { + "epoch": 0.3973509933774834, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.0368, + "step": 1260 + }, + { + "epoch": 0.3976663513087354, + "grad_norm": 0.1728515625, + "learning_rate": 0.0002, + "loss": 1.152, + "step": 1261 + }, + { + "epoch": 0.3979817092399874, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.1305, + "step": 1262 + }, + { + "epoch": 0.39829706717123936, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 1.1996, + "step": 1263 + }, + { + "epoch": 0.39861242510249134, + "grad_norm": 0.1875, + "learning_rate": 0.0002, + "loss": 1.0567, + "step": 1264 + }, + { + "epoch": 0.3989277830337433, + "grad_norm": 0.1767578125, + "learning_rate": 0.0002, + "loss": 1.2615, + "step": 1265 + }, + { + "epoch": 0.39924314096499525, + "grad_norm": 0.177734375, + "learning_rate": 0.0002, + "loss": 1.0911, + "step": 1266 + }, + { + "epoch": 0.3995584988962472, + "grad_norm": 0.1884765625, + "learning_rate": 0.0002, + "loss": 1.1341, + "step": 1267 + }, + { + "epoch": 0.3998738568274992, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 1.0408, + "step": 1268 + }, + { + "epoch": 0.4001892147587512, + "grad_norm": 0.201171875, + "learning_rate": 0.0002, + "loss": 1.2169, + "step": 1269 + }, + { + "epoch": 0.40050457269000317, + "grad_norm": 0.2119140625, + "learning_rate": 0.0002, + "loss": 1.4306, + "step": 1270 + }, + { + "epoch": 0.40081993062125515, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.2931, + "step": 1271 + }, + { + "epoch": 0.4011352885525071, + "grad_norm": 0.2275390625, + "learning_rate": 0.0002, + "loss": 1.466, + "step": 1272 + }, + { + "epoch": 0.40145064648375905, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 1.1389, + "step": 1273 + }, + { + "epoch": 0.40176600441501104, + "grad_norm": 0.23828125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 1274 + }, + { + "epoch": 0.402081362346263, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 1.4038, + "step": 1275 + }, + { + "epoch": 0.402396720277515, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.4392, + "step": 1276 + }, + { + "epoch": 0.402712078208767, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 1.6015, + "step": 1277 + }, + { + "epoch": 0.4030274361400189, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 1.3932, + "step": 1278 + }, + { + "epoch": 0.4033427940712709, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 1.5321, + "step": 1279 + }, + { + "epoch": 0.40365815200252286, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.4835, + "step": 1280 + }, + { + "epoch": 0.40397350993377484, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 1.5031, + "step": 1281 + }, + { + "epoch": 0.4042888678650268, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 1.5292, + "step": 1282 + }, + { + "epoch": 0.40460422579627875, + "grad_norm": 0.416015625, + "learning_rate": 0.0002, + "loss": 1.7969, + "step": 1283 + }, + { + "epoch": 0.40491958372753073, + "grad_norm": 0.4140625, + "learning_rate": 0.0002, + "loss": 1.6952, + "step": 1284 + }, + { + "epoch": 0.4052349416587827, + "grad_norm": 0.484375, + "learning_rate": 0.0002, + "loss": 2.0422, + "step": 1285 + }, + { + "epoch": 0.4055502995900347, + "grad_norm": 0.4296875, + "learning_rate": 0.0002, + "loss": 1.7785, + "step": 1286 + }, + { + "epoch": 0.40586565752128667, + "grad_norm": 0.498046875, + "learning_rate": 0.0002, + "loss": 1.8974, + "step": 1287 + }, + { + "epoch": 0.40618101545253865, + "grad_norm": 0.4609375, + "learning_rate": 0.0002, + "loss": 1.7238, + "step": 1288 + }, + { + "epoch": 0.4064963733837906, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 2.3844, + "step": 1289 + }, + { + "epoch": 0.40681173131504256, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 1.9476, + "step": 1290 + }, + { + "epoch": 0.40712708924629454, + "grad_norm": 0.609375, + "learning_rate": 0.0002, + "loss": 1.9325, + "step": 1291 + }, + { + "epoch": 0.4074424471775465, + "grad_norm": 0.84765625, + "learning_rate": 0.0002, + "loss": 3.1132, + "step": 1292 + }, + { + "epoch": 0.4077578051087985, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 2.6591, + "step": 1293 + }, + { + "epoch": 0.4080731630400505, + "grad_norm": 0.96484375, + "learning_rate": 0.0002, + "loss": 2.7748, + "step": 1294 + }, + { + "epoch": 0.4083885209713024, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 2.0097, + "step": 1295 + }, + { + "epoch": 0.4087038789025544, + "grad_norm": 0.75390625, + "learning_rate": 0.0002, + "loss": 2.5158, + "step": 1296 + }, + { + "epoch": 0.40901923683380637, + "grad_norm": 0.7578125, + "learning_rate": 0.0002, + "loss": 2.548, + "step": 1297 + }, + { + "epoch": 0.40933459476505835, + "grad_norm": 0.6640625, + "learning_rate": 0.0002, + "loss": 1.9665, + "step": 1298 + }, + { + "epoch": 0.4096499526963103, + "grad_norm": 0.76953125, + "learning_rate": 0.0002, + "loss": 1.8554, + "step": 1299 + }, + { + "epoch": 0.4099653106275623, + "grad_norm": 1.1953125, + "learning_rate": 0.0002, + "loss": 3.3549, + "step": 1300 + }, + { + "epoch": 0.41028066855881423, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.1999, + "step": 1301 + }, + { + "epoch": 0.4105960264900662, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.1532, + "step": 1302 + }, + { + "epoch": 0.4109113844213182, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.1939, + "step": 1303 + }, + { + "epoch": 0.4112267423525702, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.0317, + "step": 1304 + }, + { + "epoch": 0.41154210028382215, + "grad_norm": 0.1513671875, + "learning_rate": 0.0002, + "loss": 1.0139, + "step": 1305 + }, + { + "epoch": 0.41185745821507413, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.1184, + "step": 1306 + }, + { + "epoch": 0.41217281614632606, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.1735, + "step": 1307 + }, + { + "epoch": 0.41248817407757804, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.0962, + "step": 1308 + }, + { + "epoch": 0.41280353200883, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.1828, + "step": 1309 + }, + { + "epoch": 0.41280353200883, + "eval_loss": 1.6930510997772217, + "eval_runtime": 148.9809, + "eval_samples_per_second": 6.712, + "eval_steps_per_second": 6.712, + "step": 1309 + }, + { + "epoch": 0.41280353200883, + "mmlu_eval_accuracy": 0.6235082419023665, + "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182, + "mmlu_eval_accuracy_anatomy": 0.6428571428571429, + "mmlu_eval_accuracy_astronomy": 0.6875, + "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, + "mmlu_eval_accuracy_clinical_knowledge": 0.6896551724137931, + "mmlu_eval_accuracy_college_biology": 0.75, + "mmlu_eval_accuracy_college_chemistry": 0.5, + "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, + "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454, + "mmlu_eval_accuracy_college_medicine": 0.7272727272727273, + "mmlu_eval_accuracy_college_physics": 0.6363636363636364, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.5769230769230769, + "mmlu_eval_accuracy_econometrics": 0.5, + "mmlu_eval_accuracy_electrical_engineering": 0.625, + "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244, + "mmlu_eval_accuracy_formal_logic": 0.5714285714285714, + "mmlu_eval_accuracy_global_facts": 0.3, + "mmlu_eval_accuracy_high_school_biology": 0.65625, + "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.8888888888888888, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325, + "mmlu_eval_accuracy_high_school_mathematics": 0.4482758620689655, + "mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923, + "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, + "mmlu_eval_accuracy_high_school_psychology": 0.9333333333333333, + "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, + "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727, + "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, + "mmlu_eval_accuracy_human_aging": 0.6086956521739131, + "mmlu_eval_accuracy_human_sexuality": 0.6666666666666666, + "mmlu_eval_accuracy_international_law": 0.8461538461538461, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.88, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7790697674418605, + "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842, + "mmlu_eval_accuracy_moral_scenarios": 0.37, + "mmlu_eval_accuracy_nutrition": 0.696969696969697, + "mmlu_eval_accuracy_philosophy": 0.7058823529411765, + "mmlu_eval_accuracy_prehistory": 0.6285714285714286, + "mmlu_eval_accuracy_professional_accounting": 0.25806451612903225, + "mmlu_eval_accuracy_professional_law": 0.4294117647058823, + "mmlu_eval_accuracy_professional_medicine": 0.7096774193548387, + "mmlu_eval_accuracy_professional_psychology": 0.6666666666666666, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.8148148148148148, + "mmlu_eval_accuracy_sociology": 0.9090909090909091, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.4444444444444444, + "mmlu_eval_accuracy_world_religions": 0.8421052631578947, + "mmlu_loss": 6.288536165836666, + "step": 1309 + }, + { + "epoch": 0.413118889940082, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.1791, + "step": 1310 + }, + { + "epoch": 0.413434247871334, + "grad_norm": 0.18359375, + "learning_rate": 0.0002, + "loss": 1.2469, + "step": 1311 + }, + { + "epoch": 0.41374960580258596, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.2931, + "step": 1312 + }, + { + "epoch": 0.4140649637338379, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.1663, + "step": 1313 + }, + { + "epoch": 0.41438032166508987, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.0712, + "step": 1314 + }, + { + "epoch": 0.41469567959634185, + "grad_norm": 0.1767578125, + "learning_rate": 0.0002, + "loss": 1.1509, + "step": 1315 + }, + { + "epoch": 0.41501103752759383, + "grad_norm": 0.1875, + "learning_rate": 0.0002, + "loss": 1.1602, + "step": 1316 + }, + { + "epoch": 0.4153263954588458, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 1.1802, + "step": 1317 + }, + { + "epoch": 0.41564175339009773, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 1.1211, + "step": 1318 + }, + { + "epoch": 0.4159571113213497, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 1.1406, + "step": 1319 + }, + { + "epoch": 0.4162724692526017, + "grad_norm": 0.208984375, + "learning_rate": 0.0002, + "loss": 1.2631, + "step": 1320 + }, + { + "epoch": 0.4165878271838537, + "grad_norm": 0.224609375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 1321 + }, + { + "epoch": 0.41690318511510566, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 1.208, + "step": 1322 + }, + { + "epoch": 0.41721854304635764, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.281, + "step": 1323 + }, + { + "epoch": 0.41753390097760956, + "grad_norm": 0.2431640625, + "learning_rate": 0.0002, + "loss": 1.545, + "step": 1324 + }, + { + "epoch": 0.41784925890886154, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 1.453, + "step": 1325 + }, + { + "epoch": 0.4181646168401135, + "grad_norm": 0.23828125, + "learning_rate": 0.0002, + "loss": 1.165, + "step": 1326 + }, + { + "epoch": 0.4184799747713655, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 1.386, + "step": 1327 + }, + { + "epoch": 0.4187953327026175, + "grad_norm": 0.2451171875, + "learning_rate": 0.0002, + "loss": 1.1359, + "step": 1328 + }, + { + "epoch": 0.41911069063386946, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 1.5637, + "step": 1329 + }, + { + "epoch": 0.4194260485651214, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.4331, + "step": 1330 + }, + { + "epoch": 0.41974140649637337, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 1331 + }, + { + "epoch": 0.42005676442762535, + "grad_norm": 0.42578125, + "learning_rate": 0.0002, + "loss": 1.533, + "step": 1332 + }, + { + "epoch": 0.42037212235887733, + "grad_norm": 0.37109375, + "learning_rate": 0.0002, + "loss": 1.6294, + "step": 1333 + }, + { + "epoch": 0.4206874802901293, + "grad_norm": 0.53515625, + "learning_rate": 0.0002, + "loss": 1.6004, + "step": 1334 + }, + { + "epoch": 0.4210028382213813, + "grad_norm": 0.470703125, + "learning_rate": 0.0002, + "loss": 1.9542, + "step": 1335 + }, + { + "epoch": 0.4213181961526332, + "grad_norm": 0.48828125, + "learning_rate": 0.0002, + "loss": 1.9487, + "step": 1336 + }, + { + "epoch": 0.4216335540838852, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 2.0763, + "step": 1337 + }, + { + "epoch": 0.4219489120151372, + "grad_norm": 0.49609375, + "learning_rate": 0.0002, + "loss": 2.0012, + "step": 1338 + }, + { + "epoch": 0.42226426994638916, + "grad_norm": 0.53515625, + "learning_rate": 0.0002, + "loss": 1.9694, + "step": 1339 + }, + { + "epoch": 0.42257962787764114, + "grad_norm": 0.50390625, + "learning_rate": 0.0002, + "loss": 2.0704, + "step": 1340 + }, + { + "epoch": 0.4228949858088931, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 1.8401, + "step": 1341 + }, + { + "epoch": 0.42321034374014505, + "grad_norm": 0.5546875, + "learning_rate": 0.0002, + "loss": 2.0445, + "step": 1342 + }, + { + "epoch": 0.423525701671397, + "grad_norm": 0.671875, + "learning_rate": 0.0002, + "loss": 2.2419, + "step": 1343 + }, + { + "epoch": 0.423841059602649, + "grad_norm": 0.53515625, + "learning_rate": 0.0002, + "loss": 1.6483, + "step": 1344 + }, + { + "epoch": 0.424156417533901, + "grad_norm": 0.68359375, + "learning_rate": 0.0002, + "loss": 2.2735, + "step": 1345 + }, + { + "epoch": 0.42447177546515297, + "grad_norm": 0.734375, + "learning_rate": 0.0002, + "loss": 2.3618, + "step": 1346 + }, + { + "epoch": 0.4247871333964049, + "grad_norm": 0.9140625, + "learning_rate": 0.0002, + "loss": 2.706, + "step": 1347 + }, + { + "epoch": 0.4251024913276569, + "grad_norm": 1.1328125, + "learning_rate": 0.0002, + "loss": 2.5566, + "step": 1348 + }, + { + "epoch": 0.42541784925890885, + "grad_norm": 0.765625, + "learning_rate": 0.0002, + "loss": 2.2643, + "step": 1349 + }, + { + "epoch": 0.42573320719016083, + "grad_norm": 0.89453125, + "learning_rate": 0.0002, + "loss": 2.5886, + "step": 1350 + }, + { + "epoch": 0.4260485651214128, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.1676, + "step": 1351 + }, + { + "epoch": 0.4263639230526648, + "grad_norm": 0.1474609375, + "learning_rate": 0.0002, + "loss": 1.1285, + "step": 1352 + }, + { + "epoch": 0.4266792809839167, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.1588, + "step": 1353 + }, + { + "epoch": 0.4269946389151687, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.0702, + "step": 1354 + }, + { + "epoch": 0.4273099968464207, + "grad_norm": 0.1591796875, + "learning_rate": 0.0002, + "loss": 1.1546, + "step": 1355 + }, + { + "epoch": 0.42762535477767266, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.0377, + "step": 1356 + }, + { + "epoch": 0.42794071270892464, + "grad_norm": 0.1513671875, + "learning_rate": 0.0002, + "loss": 1.0661, + "step": 1357 + }, + { + "epoch": 0.4282560706401766, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.1694, + "step": 1358 + }, + { + "epoch": 0.42857142857142855, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.1105, + "step": 1359 + }, + { + "epoch": 0.42888678650268053, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.1026, + "step": 1360 + }, + { + "epoch": 0.4292021444339325, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.2178, + "step": 1361 + }, + { + "epoch": 0.4295175023651845, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.0674, + "step": 1362 + }, + { + "epoch": 0.42983286029643647, + "grad_norm": 0.1640625, + "learning_rate": 0.0002, + "loss": 1.1313, + "step": 1363 + }, + { + "epoch": 0.43014821822768845, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.2503, + "step": 1364 + }, + { + "epoch": 0.4304635761589404, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 1.2021, + "step": 1365 + }, + { + "epoch": 0.43077893409019236, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 1.2452, + "step": 1366 + }, + { + "epoch": 0.43109429202144434, + "grad_norm": 0.1943359375, + "learning_rate": 0.0002, + "loss": 1.1961, + "step": 1367 + }, + { + "epoch": 0.4314096499526963, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 1368 + }, + { + "epoch": 0.4317250078839483, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 1.2794, + "step": 1369 + }, + { + "epoch": 0.4320403658152003, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 1.2121, + "step": 1370 + }, + { + "epoch": 0.4323557237464522, + "grad_norm": 0.23046875, + "learning_rate": 0.0002, + "loss": 1.5329, + "step": 1371 + }, + { + "epoch": 0.4326710816777042, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 1372 + }, + { + "epoch": 0.43298643960895616, + "grad_norm": 0.248046875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 1373 + }, + { + "epoch": 0.43330179754020814, + "grad_norm": 0.2412109375, + "learning_rate": 0.0002, + "loss": 1.2824, + "step": 1374 + }, + { + "epoch": 0.4336171554714601, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 1.2644, + "step": 1375 + }, + { + "epoch": 0.4339325134027121, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 1.4182, + "step": 1376 + }, + { + "epoch": 0.43424787133396403, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 1.4803, + "step": 1377 + }, + { + "epoch": 0.434563229265216, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 1.3918, + "step": 1378 + }, + { + "epoch": 0.434878587196468, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 1.5691, + "step": 1379 + }, + { + "epoch": 0.43519394512771997, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 1.5462, + "step": 1380 + }, + { + "epoch": 0.43550930305897195, + "grad_norm": 0.388671875, + "learning_rate": 0.0002, + "loss": 1.6539, + "step": 1381 + }, + { + "epoch": 0.4358246609902239, + "grad_norm": 0.392578125, + "learning_rate": 0.0002, + "loss": 1.4971, + "step": 1382 + }, + { + "epoch": 0.43614001892147586, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 1.5913, + "step": 1383 + }, + { + "epoch": 0.43645537685272784, + "grad_norm": 0.5, + "learning_rate": 0.0002, + "loss": 1.8715, + "step": 1384 + }, + { + "epoch": 0.4367707347839798, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 2.2714, + "step": 1385 + }, + { + "epoch": 0.4370860927152318, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 1.9471, + "step": 1386 + }, + { + "epoch": 0.4374014506464838, + "grad_norm": 0.546875, + "learning_rate": 0.0002, + "loss": 1.9286, + "step": 1387 + }, + { + "epoch": 0.4377168085777357, + "grad_norm": 0.46875, + "learning_rate": 0.0002, + "loss": 1.8581, + "step": 1388 + }, + { + "epoch": 0.4380321665089877, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 2.1911, + "step": 1389 + }, + { + "epoch": 0.43834752444023967, + "grad_norm": 0.55859375, + "learning_rate": 0.0002, + "loss": 2.0862, + "step": 1390 + }, + { + "epoch": 0.43866288237149165, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 1.954, + "step": 1391 + }, + { + "epoch": 0.4389782403027436, + "grad_norm": 0.58203125, + "learning_rate": 0.0002, + "loss": 2.0492, + "step": 1392 + }, + { + "epoch": 0.4392935982339956, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 2.2761, + "step": 1393 + }, + { + "epoch": 0.43960895616524753, + "grad_norm": 0.625, + "learning_rate": 0.0002, + "loss": 2.5093, + "step": 1394 + }, + { + "epoch": 0.4399243140964995, + "grad_norm": 0.56640625, + "learning_rate": 0.0002, + "loss": 2.142, + "step": 1395 + }, + { + "epoch": 0.4402396720277515, + "grad_norm": 0.71484375, + "learning_rate": 0.0002, + "loss": 2.3021, + "step": 1396 + }, + { + "epoch": 0.4405550299590035, + "grad_norm": 0.70703125, + "learning_rate": 0.0002, + "loss": 2.1005, + "step": 1397 + }, + { + "epoch": 0.44087038789025546, + "grad_norm": 0.86328125, + "learning_rate": 0.0002, + "loss": 2.4223, + "step": 1398 + }, + { + "epoch": 0.44118574582150744, + "grad_norm": 0.82421875, + "learning_rate": 0.0002, + "loss": 1.9826, + "step": 1399 + }, + { + "epoch": 0.44150110375275936, + "grad_norm": 1.296875, + "learning_rate": 0.0002, + "loss": 2.9162, + "step": 1400 + }, + { + "epoch": 0.44181646168401134, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.1777, + "step": 1401 + }, + { + "epoch": 0.4421318196152633, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.108, + "step": 1402 + }, + { + "epoch": 0.4424471775465153, + "grad_norm": 0.16015625, + "learning_rate": 0.0002, + "loss": 1.1363, + "step": 1403 + }, + { + "epoch": 0.4427625354777673, + "grad_norm": 0.169921875, + "learning_rate": 0.0002, + "loss": 1.1777, + "step": 1404 + }, + { + "epoch": 0.44307789340901926, + "grad_norm": 0.154296875, + "learning_rate": 0.0002, + "loss": 1.1214, + "step": 1405 + }, + { + "epoch": 0.4433932513402712, + "grad_norm": 0.177734375, + "learning_rate": 0.0002, + "loss": 1.127, + "step": 1406 + }, + { + "epoch": 0.44370860927152317, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 0.9766, + "step": 1407 + }, + { + "epoch": 0.44402396720277515, + "grad_norm": 0.19921875, + "learning_rate": 0.0002, + "loss": 1.1239, + "step": 1408 + }, + { + "epoch": 0.44433932513402713, + "grad_norm": 0.1728515625, + "learning_rate": 0.0002, + "loss": 1.0954, + "step": 1409 + }, + { + "epoch": 0.4446546830652791, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.0611, + "step": 1410 + }, + { + "epoch": 0.44497004099653104, + "grad_norm": 0.2060546875, + "learning_rate": 0.0002, + "loss": 1.2093, + "step": 1411 + }, + { + "epoch": 0.445285398927783, + "grad_norm": 0.197265625, + "learning_rate": 0.0002, + "loss": 1.1506, + "step": 1412 + }, + { + "epoch": 0.445600756859035, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.1957, + "step": 1413 + }, + { + "epoch": 0.445916114790287, + "grad_norm": 0.1875, + "learning_rate": 0.0002, + "loss": 1.1637, + "step": 1414 + }, + { + "epoch": 0.44623147272153896, + "grad_norm": 0.1875, + "learning_rate": 0.0002, + "loss": 1.1183, + "step": 1415 + }, + { + "epoch": 0.44654683065279094, + "grad_norm": 0.232421875, + "learning_rate": 0.0002, + "loss": 1.2993, + "step": 1416 + }, + { + "epoch": 0.44686218858404286, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.4423, + "step": 1417 + }, + { + "epoch": 0.44717754651529484, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.3696, + "step": 1418 + }, + { + "epoch": 0.4474929044465468, + "grad_norm": 0.208984375, + "learning_rate": 0.0002, + "loss": 1.286, + "step": 1419 + }, + { + "epoch": 0.4478082623777988, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 1.1325, + "step": 1420 + }, + { + "epoch": 0.4481236203090508, + "grad_norm": 0.2275390625, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 1421 + }, + { + "epoch": 0.44843897824030277, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 1.2979, + "step": 1422 + }, + { + "epoch": 0.4487543361715547, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 1.4237, + "step": 1423 + }, + { + "epoch": 0.44906969410280667, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 1424 + }, + { + "epoch": 0.44938505203405865, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 1425 + }, + { + "epoch": 0.44970040996531063, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 1.2826, + "step": 1426 + }, + { + "epoch": 0.4500157678965626, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 1.3908, + "step": 1427 + }, + { + "epoch": 0.4503311258278146, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 1.6365, + "step": 1428 + }, + { + "epoch": 0.4506464837590665, + "grad_norm": 0.38671875, + "learning_rate": 0.0002, + "loss": 1.5152, + "step": 1429 + }, + { + "epoch": 0.4509618416903185, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 1.3567, + "step": 1430 + }, + { + "epoch": 0.4512771996215705, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 1.5568, + "step": 1431 + }, + { + "epoch": 0.45159255755282246, + "grad_norm": 0.392578125, + "learning_rate": 0.0002, + "loss": 1.7576, + "step": 1432 + }, + { + "epoch": 0.45190791548407444, + "grad_norm": 0.455078125, + "learning_rate": 0.0002, + "loss": 1.7251, + "step": 1433 + }, + { + "epoch": 0.4522232734153264, + "grad_norm": 0.4453125, + "learning_rate": 0.0002, + "loss": 1.697, + "step": 1434 + }, + { + "epoch": 0.45253863134657835, + "grad_norm": 0.42578125, + "learning_rate": 0.0002, + "loss": 1.8103, + "step": 1435 + }, + { + "epoch": 0.4528539892778303, + "grad_norm": 0.7578125, + "learning_rate": 0.0002, + "loss": 1.7809, + "step": 1436 + }, + { + "epoch": 0.4531693472090823, + "grad_norm": 0.46484375, + "learning_rate": 0.0002, + "loss": 1.9225, + "step": 1437 + }, + { + "epoch": 0.4534847051403343, + "grad_norm": 0.4296875, + "learning_rate": 0.0002, + "loss": 1.439, + "step": 1438 + }, + { + "epoch": 0.45380006307158627, + "grad_norm": 0.734375, + "learning_rate": 0.0002, + "loss": 2.3981, + "step": 1439 + }, + { + "epoch": 0.45411542100283825, + "grad_norm": 0.67578125, + "learning_rate": 0.0002, + "loss": 2.7074, + "step": 1440 + }, + { + "epoch": 0.4544307789340902, + "grad_norm": 0.7578125, + "learning_rate": 0.0002, + "loss": 1.9703, + "step": 1441 + }, + { + "epoch": 0.45474613686534215, + "grad_norm": 0.7890625, + "learning_rate": 0.0002, + "loss": 2.7091, + "step": 1442 + }, + { + "epoch": 0.45506149479659413, + "grad_norm": 0.60546875, + "learning_rate": 0.0002, + "loss": 2.3233, + "step": 1443 + }, + { + "epoch": 0.4553768527278461, + "grad_norm": 0.51953125, + "learning_rate": 0.0002, + "loss": 1.8927, + "step": 1444 + }, + { + "epoch": 0.4556922106590981, + "grad_norm": 0.703125, + "learning_rate": 0.0002, + "loss": 2.6221, + "step": 1445 + }, + { + "epoch": 0.45600756859035, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 1.8218, + "step": 1446 + }, + { + "epoch": 0.456322926521602, + "grad_norm": 0.578125, + "learning_rate": 0.0002, + "loss": 2.0773, + "step": 1447 + }, + { + "epoch": 0.456638284452854, + "grad_norm": 0.796875, + "learning_rate": 0.0002, + "loss": 2.5574, + "step": 1448 + }, + { + "epoch": 0.45695364238410596, + "grad_norm": 0.67578125, + "learning_rate": 0.0002, + "loss": 2.152, + "step": 1449 + }, + { + "epoch": 0.45726900031535794, + "grad_norm": 0.6171875, + "learning_rate": 0.0002, + "loss": 1.7062, + "step": 1450 + }, + { + "epoch": 0.4575843582466099, + "grad_norm": 0.1513671875, + "learning_rate": 0.0002, + "loss": 1.1308, + "step": 1451 + }, + { + "epoch": 0.45789971617786185, + "grad_norm": 0.16015625, + "learning_rate": 0.0002, + "loss": 1.0095, + "step": 1452 + }, + { + "epoch": 0.45821507410911383, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.0393, + "step": 1453 + }, + { + "epoch": 0.4585304320403658, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.0471, + "step": 1454 + }, + { + "epoch": 0.4588457899716178, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.125, + "step": 1455 + }, + { + "epoch": 0.45916114790286977, + "grad_norm": 0.16015625, + "learning_rate": 0.0002, + "loss": 1.1433, + "step": 1456 + }, + { + "epoch": 0.45947650583412175, + "grad_norm": 0.169921875, + "learning_rate": 0.0002, + "loss": 1.1046, + "step": 1457 + }, + { + "epoch": 0.4597918637653737, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 1.2477, + "step": 1458 + }, + { + "epoch": 0.46010722169662566, + "grad_norm": 0.1552734375, + "learning_rate": 0.0002, + "loss": 1.0147, + "step": 1459 + }, + { + "epoch": 0.46042257962787764, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.2253, + "step": 1460 + }, + { + "epoch": 0.4607379375591296, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.0698, + "step": 1461 + }, + { + "epoch": 0.4610532954903816, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.1452, + "step": 1462 + }, + { + "epoch": 0.4613686534216336, + "grad_norm": 0.177734375, + "learning_rate": 0.0002, + "loss": 1.1991, + "step": 1463 + }, + { + "epoch": 0.4616840113528855, + "grad_norm": 0.1962890625, + "learning_rate": 0.0002, + "loss": 1.1183, + "step": 1464 + }, + { + "epoch": 0.4619993692841375, + "grad_norm": 0.17578125, + "learning_rate": 0.0002, + "loss": 1.1854, + "step": 1465 + }, + { + "epoch": 0.46231472721538946, + "grad_norm": 0.1884765625, + "learning_rate": 0.0002, + "loss": 1.284, + "step": 1466 + }, + { + "epoch": 0.46263008514664145, + "grad_norm": 0.1962890625, + "learning_rate": 0.0002, + "loss": 1.1647, + "step": 1467 + }, + { + "epoch": 0.4629454430778934, + "grad_norm": 0.1943359375, + "learning_rate": 0.0002, + "loss": 1.1893, + "step": 1468 + }, + { + "epoch": 0.4632608010091454, + "grad_norm": 0.22265625, + "learning_rate": 0.0002, + "loss": 1.2554, + "step": 1469 + }, + { + "epoch": 0.46357615894039733, + "grad_norm": 0.400390625, + "learning_rate": 0.0002, + "loss": 1.5348, + "step": 1470 + }, + { + "epoch": 0.4638915168716493, + "grad_norm": 0.232421875, + "learning_rate": 0.0002, + "loss": 1.3825, + "step": 1471 + }, + { + "epoch": 0.4642068748029013, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.274, + "step": 1472 + }, + { + "epoch": 0.4645222327341533, + "grad_norm": 0.255859375, + "learning_rate": 0.0002, + "loss": 1.4272, + "step": 1473 + }, + { + "epoch": 0.46483759066540525, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.5371, + "step": 1474 + }, + { + "epoch": 0.4651529485966572, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 1.4891, + "step": 1475 + }, + { + "epoch": 0.46546830652790916, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.4908, + "step": 1476 + }, + { + "epoch": 0.46578366445916114, + "grad_norm": 0.34765625, + "learning_rate": 0.0002, + "loss": 1.6436, + "step": 1477 + }, + { + "epoch": 0.4660990223904131, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 1.4538, + "step": 1478 + }, + { + "epoch": 0.4664143803216651, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 1.4062, + "step": 1479 + }, + { + "epoch": 0.4667297382529171, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 1.7682, + "step": 1480 + }, + { + "epoch": 0.467045096184169, + "grad_norm": 0.40234375, + "learning_rate": 0.0002, + "loss": 1.6432, + "step": 1481 + }, + { + "epoch": 0.467360454115421, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 1.6506, + "step": 1482 + }, + { + "epoch": 0.46767581204667297, + "grad_norm": 0.39453125, + "learning_rate": 0.0002, + "loss": 1.742, + "step": 1483 + }, + { + "epoch": 0.46799116997792495, + "grad_norm": 0.44140625, + "learning_rate": 0.0002, + "loss": 1.7477, + "step": 1484 + }, + { + "epoch": 0.46830652790917693, + "grad_norm": 0.46484375, + "learning_rate": 0.0002, + "loss": 2.0039, + "step": 1485 + }, + { + "epoch": 0.4686218858404289, + "grad_norm": 0.58984375, + "learning_rate": 0.0002, + "loss": 2.3235, + "step": 1486 + }, + { + "epoch": 0.46893724377168083, + "grad_norm": 0.7265625, + "learning_rate": 0.0002, + "loss": 2.2069, + "step": 1487 + }, + { + "epoch": 0.4692526017029328, + "grad_norm": 0.56640625, + "learning_rate": 0.0002, + "loss": 1.992, + "step": 1488 + }, + { + "epoch": 0.4695679596341848, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 1.6668, + "step": 1489 + }, + { + "epoch": 0.4698833175654368, + "grad_norm": 0.6484375, + "learning_rate": 0.0002, + "loss": 2.4284, + "step": 1490 + }, + { + "epoch": 0.47019867549668876, + "grad_norm": 0.859375, + "learning_rate": 0.0002, + "loss": 2.7102, + "step": 1491 + }, + { + "epoch": 0.47051403342794074, + "grad_norm": 0.73046875, + "learning_rate": 0.0002, + "loss": 2.5687, + "step": 1492 + }, + { + "epoch": 0.47082939135919266, + "grad_norm": 0.578125, + "learning_rate": 0.0002, + "loss": 2.1435, + "step": 1493 + }, + { + "epoch": 0.47114474929044464, + "grad_norm": 0.67578125, + "learning_rate": 0.0002, + "loss": 2.2262, + "step": 1494 + }, + { + "epoch": 0.4714601072216966, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 2.6321, + "step": 1495 + }, + { + "epoch": 0.4717754651529486, + "grad_norm": 0.82421875, + "learning_rate": 0.0002, + "loss": 2.0012, + "step": 1496 + }, + { + "epoch": 0.4717754651529486, + "eval_loss": 1.6620019674301147, + "eval_runtime": 148.9072, + "eval_samples_per_second": 6.716, + "eval_steps_per_second": 6.716, + "step": 1496 + }, + { + "epoch": 0.4717754651529486, + "mmlu_eval_accuracy": 0.6287323863319004, + "mmlu_eval_accuracy_abstract_algebra": 0.09090909090909091, + "mmlu_eval_accuracy_anatomy": 0.5714285714285714, + "mmlu_eval_accuracy_astronomy": 0.75, + "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, + "mmlu_eval_accuracy_clinical_knowledge": 0.7241379310344828, + "mmlu_eval_accuracy_college_biology": 0.8125, + "mmlu_eval_accuracy_college_chemistry": 0.375, + "mmlu_eval_accuracy_college_computer_science": 0.5454545454545454, + "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454, + "mmlu_eval_accuracy_college_medicine": 0.6818181818181818, + "mmlu_eval_accuracy_college_physics": 0.6363636363636364, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.6538461538461539, + "mmlu_eval_accuracy_econometrics": 0.5, + "mmlu_eval_accuracy_electrical_engineering": 0.6875, + "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244, + "mmlu_eval_accuracy_formal_logic": 0.6428571428571429, + "mmlu_eval_accuracy_global_facts": 0.3, + "mmlu_eval_accuracy_high_school_biology": 0.65625, + "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209, + "mmlu_eval_accuracy_high_school_mathematics": 0.3448275862068966, + "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769, + "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354, + "mmlu_eval_accuracy_high_school_psychology": 0.9166666666666666, + "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, + "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727, + "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, + "mmlu_eval_accuracy_human_aging": 0.6521739130434783, + "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334, + "mmlu_eval_accuracy_international_law": 0.9230769230769231, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.5454545454545454, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.8, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7790697674418605, + "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842, + "mmlu_eval_accuracy_moral_scenarios": 0.37, + "mmlu_eval_accuracy_nutrition": 0.7272727272727273, + "mmlu_eval_accuracy_philosophy": 0.7647058823529411, + "mmlu_eval_accuracy_prehistory": 0.6571428571428571, + "mmlu_eval_accuracy_professional_accounting": 0.3225806451612903, + "mmlu_eval_accuracy_professional_law": 0.4294117647058823, + "mmlu_eval_accuracy_professional_medicine": 0.7741935483870968, + "mmlu_eval_accuracy_professional_psychology": 0.6521739130434783, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.8148148148148148, + "mmlu_eval_accuracy_sociology": 0.9545454545454546, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.4444444444444444, + "mmlu_eval_accuracy_world_religions": 0.8947368421052632, + "mmlu_loss": 5.807670999087203, + "step": 1496 + }, + { + "epoch": 0.4720908230842006, + "grad_norm": 0.84375, + "learning_rate": 0.0002, + "loss": 2.7558, + "step": 1497 + }, + { + "epoch": 0.47240618101545256, + "grad_norm": 0.8359375, + "learning_rate": 0.0002, + "loss": 1.679, + "step": 1498 + }, + { + "epoch": 0.4727215389467045, + "grad_norm": 0.8359375, + "learning_rate": 0.0002, + "loss": 1.9733, + "step": 1499 + }, + { + "epoch": 0.47303689687795647, + "grad_norm": 0.99609375, + "learning_rate": 0.0002, + "loss": 2.3989, + "step": 1500 + }, + { + "epoch": 0.47335225480920845, + "grad_norm": 0.1796875, + "learning_rate": 0.0002, + "loss": 1.1048, + "step": 1501 + }, + { + "epoch": 0.47366761274046043, + "grad_norm": 0.154296875, + "learning_rate": 0.0002, + "loss": 1.0386, + "step": 1502 + }, + { + "epoch": 0.4739829706717124, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 1.1604, + "step": 1503 + }, + { + "epoch": 0.4742983286029644, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.1663, + "step": 1504 + }, + { + "epoch": 0.4746136865342163, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.0526, + "step": 1505 + }, + { + "epoch": 0.4749290444654683, + "grad_norm": 0.16015625, + "learning_rate": 0.0002, + "loss": 1.2042, + "step": 1506 + }, + { + "epoch": 0.4752444023967203, + "grad_norm": 0.1630859375, + "learning_rate": 0.0002, + "loss": 1.1011, + "step": 1507 + }, + { + "epoch": 0.47555976032797226, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.0863, + "step": 1508 + }, + { + "epoch": 0.47587511825922424, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.1886, + "step": 1509 + }, + { + "epoch": 0.47619047619047616, + "grad_norm": 0.1728515625, + "learning_rate": 0.0002, + "loss": 1.0856, + "step": 1510 + }, + { + "epoch": 0.47650583412172814, + "grad_norm": 0.166015625, + "learning_rate": 0.0002, + "loss": 1.1191, + "step": 1511 + }, + { + "epoch": 0.4768211920529801, + "grad_norm": 0.2060546875, + "learning_rate": 0.0002, + "loss": 1.153, + "step": 1512 + }, + { + "epoch": 0.4771365499842321, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 1.1367, + "step": 1513 + }, + { + "epoch": 0.4774519079154841, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.1545, + "step": 1514 + }, + { + "epoch": 0.47776726584673607, + "grad_norm": 0.2373046875, + "learning_rate": 0.0002, + "loss": 1.2826, + "step": 1515 + }, + { + "epoch": 0.478082623777988, + "grad_norm": 0.19921875, + "learning_rate": 0.0002, + "loss": 1.1954, + "step": 1516 + }, + { + "epoch": 0.47839798170923997, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 1.2282, + "step": 1517 + }, + { + "epoch": 0.47871333964049195, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.2813, + "step": 1518 + }, + { + "epoch": 0.47902869757174393, + "grad_norm": 0.2392578125, + "learning_rate": 0.0002, + "loss": 1.5134, + "step": 1519 + }, + { + "epoch": 0.4793440555029959, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 1.155, + "step": 1520 + }, + { + "epoch": 0.4796594134342479, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 1521 + }, + { + "epoch": 0.4799747713654998, + "grad_norm": 0.2333984375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 1522 + }, + { + "epoch": 0.4802901292967518, + "grad_norm": 0.22265625, + "learning_rate": 0.0002, + "loss": 1.2985, + "step": 1523 + }, + { + "epoch": 0.4806054872280038, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 1.3675, + "step": 1524 + }, + { + "epoch": 0.48092084515925576, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 1.5046, + "step": 1525 + }, + { + "epoch": 0.48123620309050774, + "grad_norm": 0.244140625, + "learning_rate": 0.0002, + "loss": 1.3787, + "step": 1526 + }, + { + "epoch": 0.4815515610217597, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 1527 + }, + { + "epoch": 0.48186691895301165, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.169, + "step": 1528 + }, + { + "epoch": 0.4821822768842636, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 1.4494, + "step": 1529 + }, + { + "epoch": 0.4824976348155156, + "grad_norm": 0.5, + "learning_rate": 0.0002, + "loss": 1.875, + "step": 1530 + }, + { + "epoch": 0.4828129927467676, + "grad_norm": 0.380859375, + "learning_rate": 0.0002, + "loss": 1.5422, + "step": 1531 + }, + { + "epoch": 0.48312835067801957, + "grad_norm": 0.484375, + "learning_rate": 0.0002, + "loss": 1.7576, + "step": 1532 + }, + { + "epoch": 0.48344370860927155, + "grad_norm": 0.47265625, + "learning_rate": 0.0002, + "loss": 1.9933, + "step": 1533 + }, + { + "epoch": 0.4837590665405235, + "grad_norm": 0.404296875, + "learning_rate": 0.0002, + "loss": 1.7391, + "step": 1534 + }, + { + "epoch": 0.48407442447177546, + "grad_norm": 0.76171875, + "learning_rate": 0.0002, + "loss": 2.5812, + "step": 1535 + }, + { + "epoch": 0.48438978240302744, + "grad_norm": 0.453125, + "learning_rate": 0.0002, + "loss": 2.0537, + "step": 1536 + }, + { + "epoch": 0.4847051403342794, + "grad_norm": 0.458984375, + "learning_rate": 0.0002, + "loss": 1.8938, + "step": 1537 + }, + { + "epoch": 0.4850204982655314, + "grad_norm": 0.65625, + "learning_rate": 0.0002, + "loss": 2.4587, + "step": 1538 + }, + { + "epoch": 0.4853358561967833, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 2.1757, + "step": 1539 + }, + { + "epoch": 0.4856512141280353, + "grad_norm": 0.68359375, + "learning_rate": 0.0002, + "loss": 2.5713, + "step": 1540 + }, + { + "epoch": 0.4859665720592873, + "grad_norm": 0.76171875, + "learning_rate": 0.0002, + "loss": 2.8009, + "step": 1541 + }, + { + "epoch": 0.48628192999053926, + "grad_norm": 0.7421875, + "learning_rate": 0.0002, + "loss": 2.5462, + "step": 1542 + }, + { + "epoch": 0.48659728792179124, + "grad_norm": 0.6796875, + "learning_rate": 0.0002, + "loss": 2.5609, + "step": 1543 + }, + { + "epoch": 0.4869126458530432, + "grad_norm": 0.7421875, + "learning_rate": 0.0002, + "loss": 2.6305, + "step": 1544 + }, + { + "epoch": 0.48722800378429515, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 2.217, + "step": 1545 + }, + { + "epoch": 0.48754336171554713, + "grad_norm": 0.828125, + "learning_rate": 0.0002, + "loss": 2.7251, + "step": 1546 + }, + { + "epoch": 0.4878587196467991, + "grad_norm": 0.75, + "learning_rate": 0.0002, + "loss": 2.3412, + "step": 1547 + }, + { + "epoch": 0.4881740775780511, + "grad_norm": 0.69921875, + "learning_rate": 0.0002, + "loss": 2.3776, + "step": 1548 + }, + { + "epoch": 0.48848943550930307, + "grad_norm": 0.828125, + "learning_rate": 0.0002, + "loss": 2.1408, + "step": 1549 + }, + { + "epoch": 0.48880479344055505, + "grad_norm": 1.09375, + "learning_rate": 0.0002, + "loss": 3.0509, + "step": 1550 + }, + { + "epoch": 0.489120151371807, + "grad_norm": 0.130859375, + "learning_rate": 0.0002, + "loss": 1.0765, + "step": 1551 + }, + { + "epoch": 0.48943550930305896, + "grad_norm": 0.1494140625, + "learning_rate": 0.0002, + "loss": 1.1173, + "step": 1552 + }, + { + "epoch": 0.48975086723431094, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.0106, + "step": 1553 + }, + { + "epoch": 0.4900662251655629, + "grad_norm": 0.150390625, + "learning_rate": 0.0002, + "loss": 1.0564, + "step": 1554 + }, + { + "epoch": 0.4903815830968149, + "grad_norm": 0.193359375, + "learning_rate": 0.0002, + "loss": 1.0376, + "step": 1555 + }, + { + "epoch": 0.4906969410280669, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 1.2456, + "step": 1556 + }, + { + "epoch": 0.4910122989593188, + "grad_norm": 0.18359375, + "learning_rate": 0.0002, + "loss": 1.1458, + "step": 1557 + }, + { + "epoch": 0.4913276568905708, + "grad_norm": 0.17578125, + "learning_rate": 0.0002, + "loss": 0.995, + "step": 1558 + }, + { + "epoch": 0.49164301482182277, + "grad_norm": 0.1591796875, + "learning_rate": 0.0002, + "loss": 1.0969, + "step": 1559 + }, + { + "epoch": 0.49195837275307475, + "grad_norm": 0.17578125, + "learning_rate": 0.0002, + "loss": 1.0726, + "step": 1560 + }, + { + "epoch": 0.4922737306843267, + "grad_norm": 0.1875, + "learning_rate": 0.0002, + "loss": 1.2417, + "step": 1561 + }, + { + "epoch": 0.4925890886155787, + "grad_norm": 0.1865234375, + "learning_rate": 0.0002, + "loss": 1.1711, + "step": 1562 + }, + { + "epoch": 0.49290444654683063, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.1959, + "step": 1563 + }, + { + "epoch": 0.4932198044780826, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.1697, + "step": 1564 + }, + { + "epoch": 0.4935351624093346, + "grad_norm": 0.1796875, + "learning_rate": 0.0002, + "loss": 1.1245, + "step": 1565 + }, + { + "epoch": 0.4938505203405866, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 1.2367, + "step": 1566 + }, + { + "epoch": 0.49416587827183855, + "grad_norm": 0.197265625, + "learning_rate": 0.0002, + "loss": 1.1817, + "step": 1567 + }, + { + "epoch": 0.49448123620309054, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 1568 + }, + { + "epoch": 0.49479659413434246, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 0.9918, + "step": 1569 + }, + { + "epoch": 0.49511195206559444, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 1.4618, + "step": 1570 + }, + { + "epoch": 0.4954273099968464, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 1.2336, + "step": 1571 + }, + { + "epoch": 0.4957426679280984, + "grad_norm": 0.2490234375, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 1572 + }, + { + "epoch": 0.4960580258593504, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.375, + "step": 1573 + }, + { + "epoch": 0.4963733837906023, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 1.2482, + "step": 1574 + }, + { + "epoch": 0.4966887417218543, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 1.5608, + "step": 1575 + }, + { + "epoch": 0.49700409965310627, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.4267, + "step": 1576 + }, + { + "epoch": 0.49731945758435825, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 1.3514, + "step": 1577 + }, + { + "epoch": 0.49763481551561023, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 1.5925, + "step": 1578 + }, + { + "epoch": 0.4979501734468622, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 1.4211, + "step": 1579 + }, + { + "epoch": 0.49826553137811413, + "grad_norm": 0.39453125, + "learning_rate": 0.0002, + "loss": 1.6391, + "step": 1580 + }, + { + "epoch": 0.4985808893093661, + "grad_norm": 0.46484375, + "learning_rate": 0.0002, + "loss": 1.7929, + "step": 1581 + }, + { + "epoch": 0.4988962472406181, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 1.4579, + "step": 1582 + }, + { + "epoch": 0.4992116051718701, + "grad_norm": 0.390625, + "learning_rate": 0.0002, + "loss": 1.8107, + "step": 1583 + }, + { + "epoch": 0.49952696310312206, + "grad_norm": 0.50390625, + "learning_rate": 0.0002, + "loss": 2.0452, + "step": 1584 + }, + { + "epoch": 0.49984232103437404, + "grad_norm": 0.46875, + "learning_rate": 0.0002, + "loss": 1.7161, + "step": 1585 + }, + { + "epoch": 0.500157678965626, + "grad_norm": 0.546875, + "learning_rate": 0.0002, + "loss": 1.9937, + "step": 1586 + }, + { + "epoch": 0.5004730368968779, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 2.5589, + "step": 1587 + }, + { + "epoch": 0.50078839482813, + "grad_norm": 0.5390625, + "learning_rate": 0.0002, + "loss": 2.0468, + "step": 1588 + }, + { + "epoch": 0.5011037527593819, + "grad_norm": 0.58203125, + "learning_rate": 0.0002, + "loss": 2.0495, + "step": 1589 + }, + { + "epoch": 0.5014191106906338, + "grad_norm": 0.62109375, + "learning_rate": 0.0002, + "loss": 2.0591, + "step": 1590 + }, + { + "epoch": 0.5017344686218859, + "grad_norm": 0.671875, + "learning_rate": 0.0002, + "loss": 2.414, + "step": 1591 + }, + { + "epoch": 0.5020498265531378, + "grad_norm": 0.7421875, + "learning_rate": 0.0002, + "loss": 2.3117, + "step": 1592 + }, + { + "epoch": 0.5023651844843898, + "grad_norm": 0.58984375, + "learning_rate": 0.0002, + "loss": 2.4007, + "step": 1593 + }, + { + "epoch": 0.5026805424156418, + "grad_norm": 0.6640625, + "learning_rate": 0.0002, + "loss": 2.4051, + "step": 1594 + }, + { + "epoch": 0.5029959003468937, + "grad_norm": 0.72265625, + "learning_rate": 0.0002, + "loss": 2.7339, + "step": 1595 + }, + { + "epoch": 0.5033112582781457, + "grad_norm": 0.6953125, + "learning_rate": 0.0002, + "loss": 2.0979, + "step": 1596 + }, + { + "epoch": 0.5036266162093976, + "grad_norm": 0.7890625, + "learning_rate": 0.0002, + "loss": 2.6717, + "step": 1597 + }, + { + "epoch": 0.5039419741406497, + "grad_norm": 0.9453125, + "learning_rate": 0.0002, + "loss": 2.4618, + "step": 1598 + }, + { + "epoch": 0.5042573320719016, + "grad_norm": 0.703125, + "learning_rate": 0.0002, + "loss": 2.2825, + "step": 1599 + }, + { + "epoch": 0.5045726900031536, + "grad_norm": 0.94921875, + "learning_rate": 0.0002, + "loss": 2.363, + "step": 1600 + }, + { + "epoch": 0.5048880479344056, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.1691, + "step": 1601 + }, + { + "epoch": 0.5052034058656575, + "grad_norm": 0.150390625, + "learning_rate": 0.0002, + "loss": 0.9723, + "step": 1602 + }, + { + "epoch": 0.5055187637969095, + "grad_norm": 0.1591796875, + "learning_rate": 0.0002, + "loss": 1.0768, + "step": 1603 + }, + { + "epoch": 0.5058341217281614, + "grad_norm": 0.1474609375, + "learning_rate": 0.0002, + "loss": 1.046, + "step": 1604 + }, + { + "epoch": 0.5061494796594135, + "grad_norm": 0.1494140625, + "learning_rate": 0.0002, + "loss": 1.0029, + "step": 1605 + }, + { + "epoch": 0.5064648375906654, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.138, + "step": 1606 + }, + { + "epoch": 0.5067801955219173, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.1166, + "step": 1607 + }, + { + "epoch": 0.5070955534531694, + "grad_norm": 0.169921875, + "learning_rate": 0.0002, + "loss": 1.1836, + "step": 1608 + }, + { + "epoch": 0.5074109113844213, + "grad_norm": 0.1640625, + "learning_rate": 0.0002, + "loss": 1.085, + "step": 1609 + }, + { + "epoch": 0.5077262693156733, + "grad_norm": 0.1796875, + "learning_rate": 0.0002, + "loss": 1.2903, + "step": 1610 + }, + { + "epoch": 0.5080416272469253, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.1975, + "step": 1611 + }, + { + "epoch": 0.5083569851781772, + "grad_norm": 0.1787109375, + "learning_rate": 0.0002, + "loss": 1.2729, + "step": 1612 + }, + { + "epoch": 0.5086723431094292, + "grad_norm": 0.1767578125, + "learning_rate": 0.0002, + "loss": 1.1057, + "step": 1613 + }, + { + "epoch": 0.5089877010406811, + "grad_norm": 0.18359375, + "learning_rate": 0.0002, + "loss": 1.1512, + "step": 1614 + }, + { + "epoch": 0.5093030589719332, + "grad_norm": 0.1689453125, + "learning_rate": 0.0002, + "loss": 1.0953, + "step": 1615 + }, + { + "epoch": 0.5096184169031851, + "grad_norm": 0.2001953125, + "learning_rate": 0.0002, + "loss": 1.2263, + "step": 1616 + }, + { + "epoch": 0.5099337748344371, + "grad_norm": 0.1865234375, + "learning_rate": 0.0002, + "loss": 1.0444, + "step": 1617 + }, + { + "epoch": 0.5102491327656891, + "grad_norm": 0.2060546875, + "learning_rate": 0.0002, + "loss": 1.3781, + "step": 1618 + }, + { + "epoch": 0.510564490696941, + "grad_norm": 0.2138671875, + "learning_rate": 0.0002, + "loss": 1.3437, + "step": 1619 + }, + { + "epoch": 0.510879848628193, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.105, + "step": 1620 + }, + { + "epoch": 0.511195206559445, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.4051, + "step": 1621 + }, + { + "epoch": 0.511510564490697, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.3608, + "step": 1622 + }, + { + "epoch": 0.5118259224219489, + "grad_norm": 0.2412109375, + "learning_rate": 0.0002, + "loss": 1.2831, + "step": 1623 + }, + { + "epoch": 0.5121412803532008, + "grad_norm": 0.2421875, + "learning_rate": 0.0002, + "loss": 1.3733, + "step": 1624 + }, + { + "epoch": 0.5124566382844529, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 1.4046, + "step": 1625 + }, + { + "epoch": 0.5127719962157048, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 1626 + }, + { + "epoch": 0.5130873541469568, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 1.4584, + "step": 1627 + }, + { + "epoch": 0.5134027120782088, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 1.4068, + "step": 1628 + }, + { + "epoch": 0.5137180700094608, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.3558, + "step": 1629 + }, + { + "epoch": 0.5140334279407127, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 1.4796, + "step": 1630 + }, + { + "epoch": 0.5143487858719646, + "grad_norm": 0.388671875, + "learning_rate": 0.0002, + "loss": 1.7971, + "step": 1631 + }, + { + "epoch": 0.5146641438032167, + "grad_norm": 0.44140625, + "learning_rate": 0.0002, + "loss": 1.5611, + "step": 1632 + }, + { + "epoch": 0.5149795017344686, + "grad_norm": 0.53515625, + "learning_rate": 0.0002, + "loss": 1.9699, + "step": 1633 + }, + { + "epoch": 0.5152948596657206, + "grad_norm": 0.494140625, + "learning_rate": 0.0002, + "loss": 2.0354, + "step": 1634 + }, + { + "epoch": 0.5156102175969726, + "grad_norm": 0.51953125, + "learning_rate": 0.0002, + "loss": 2.055, + "step": 1635 + }, + { + "epoch": 0.5159255755282245, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 1.9003, + "step": 1636 + }, + { + "epoch": 0.5162409334594765, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 2.0187, + "step": 1637 + }, + { + "epoch": 0.5165562913907285, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 2.2968, + "step": 1638 + }, + { + "epoch": 0.5168716493219805, + "grad_norm": 0.5703125, + "learning_rate": 0.0002, + "loss": 1.9446, + "step": 1639 + }, + { + "epoch": 0.5171870072532324, + "grad_norm": 0.6640625, + "learning_rate": 0.0002, + "loss": 2.3766, + "step": 1640 + }, + { + "epoch": 0.5175023651844843, + "grad_norm": 0.63671875, + "learning_rate": 0.0002, + "loss": 2.2852, + "step": 1641 + }, + { + "epoch": 0.5178177231157364, + "grad_norm": 0.734375, + "learning_rate": 0.0002, + "loss": 2.5066, + "step": 1642 + }, + { + "epoch": 0.5181330810469883, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 2.2416, + "step": 1643 + }, + { + "epoch": 0.5184484389782403, + "grad_norm": 0.86328125, + "learning_rate": 0.0002, + "loss": 2.7273, + "step": 1644 + }, + { + "epoch": 0.5187637969094923, + "grad_norm": 0.7734375, + "learning_rate": 0.0002, + "loss": 2.4836, + "step": 1645 + }, + { + "epoch": 0.5190791548407443, + "grad_norm": 0.72265625, + "learning_rate": 0.0002, + "loss": 2.3729, + "step": 1646 + }, + { + "epoch": 0.5193945127719962, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 2.1587, + "step": 1647 + }, + { + "epoch": 0.5197098707032481, + "grad_norm": 0.78125, + "learning_rate": 0.0002, + "loss": 2.3299, + "step": 1648 + }, + { + "epoch": 0.5200252286345002, + "grad_norm": 0.6796875, + "learning_rate": 0.0002, + "loss": 2.0784, + "step": 1649 + }, + { + "epoch": 0.5203405865657521, + "grad_norm": 1.2578125, + "learning_rate": 0.0002, + "loss": 2.5547, + "step": 1650 + }, + { + "epoch": 0.5206559444970041, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.0778, + "step": 1651 + }, + { + "epoch": 0.5209713024282561, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.0669, + "step": 1652 + }, + { + "epoch": 0.521286660359508, + "grad_norm": 0.154296875, + "learning_rate": 0.0002, + "loss": 1.0919, + "step": 1653 + }, + { + "epoch": 0.52160201829076, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 1.1069, + "step": 1654 + }, + { + "epoch": 0.521917376222012, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.0598, + "step": 1655 + }, + { + "epoch": 0.522232734153264, + "grad_norm": 0.1640625, + "learning_rate": 0.0002, + "loss": 1.0973, + "step": 1656 + }, + { + "epoch": 0.5225480920845159, + "grad_norm": 0.169921875, + "learning_rate": 0.0002, + "loss": 1.244, + "step": 1657 + }, + { + "epoch": 0.522863450015768, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.16, + "step": 1658 + }, + { + "epoch": 0.5231788079470199, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.0951, + "step": 1659 + }, + { + "epoch": 0.5234941658782718, + "grad_norm": 0.177734375, + "learning_rate": 0.0002, + "loss": 1.0146, + "step": 1660 + }, + { + "epoch": 0.5238095238095238, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.1252, + "step": 1661 + }, + { + "epoch": 0.5241248817407758, + "grad_norm": 0.1884765625, + "learning_rate": 0.0002, + "loss": 1.1998, + "step": 1662 + }, + { + "epoch": 0.5244402396720278, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.1351, + "step": 1663 + }, + { + "epoch": 0.5247555976032797, + "grad_norm": 0.2001953125, + "learning_rate": 0.0002, + "loss": 1.2812, + "step": 1664 + }, + { + "epoch": 0.5250709555345316, + "grad_norm": 0.208984375, + "learning_rate": 0.0002, + "loss": 1.1797, + "step": 1665 + }, + { + "epoch": 0.5253863134657837, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.2521, + "step": 1666 + }, + { + "epoch": 0.5257016713970356, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 1.3777, + "step": 1667 + }, + { + "epoch": 0.5260170293282876, + "grad_norm": 0.23046875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 1668 + }, + { + "epoch": 0.5263323872595396, + "grad_norm": 0.2119140625, + "learning_rate": 0.0002, + "loss": 1.2347, + "step": 1669 + }, + { + "epoch": 0.5266477451907915, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 1.2333, + "step": 1670 + }, + { + "epoch": 0.5269631031220435, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 1671 + }, + { + "epoch": 0.5272784610532955, + "grad_norm": 0.2431640625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 1672 + }, + { + "epoch": 0.5275938189845475, + "grad_norm": 0.255859375, + "learning_rate": 0.0002, + "loss": 1.2776, + "step": 1673 + }, + { + "epoch": 0.5279091769157994, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 1674 + }, + { + "epoch": 0.5282245348470515, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 1675 + }, + { + "epoch": 0.5285398927783034, + "grad_norm": 0.43359375, + "learning_rate": 0.0002, + "loss": 1.6241, + "step": 1676 + }, + { + "epoch": 0.5288552507095553, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 1.7141, + "step": 1677 + }, + { + "epoch": 0.5291706086408073, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 1.4854, + "step": 1678 + }, + { + "epoch": 0.5294859665720593, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 1.4564, + "step": 1679 + }, + { + "epoch": 0.5298013245033113, + "grad_norm": 0.421875, + "learning_rate": 0.0002, + "loss": 1.398, + "step": 1680 + }, + { + "epoch": 0.5301166824345632, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 1.6234, + "step": 1681 + }, + { + "epoch": 0.5304320403658151, + "grad_norm": 0.41015625, + "learning_rate": 0.0002, + "loss": 1.5113, + "step": 1682 + }, + { + "epoch": 0.5307473982970672, + "grad_norm": 0.462890625, + "learning_rate": 0.0002, + "loss": 1.6794, + "step": 1683 + }, + { + "epoch": 0.5307473982970672, + "eval_loss": 1.6657214164733887, + "eval_runtime": 148.8534, + "eval_samples_per_second": 6.718, + "eval_steps_per_second": 6.718, + "step": 1683 + }, + { + "epoch": 0.5307473982970672, + "mmlu_eval_accuracy": 0.6329342977054345, + "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182, + "mmlu_eval_accuracy_anatomy": 0.5714285714285714, + "mmlu_eval_accuracy_astronomy": 0.75, + "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, + "mmlu_eval_accuracy_clinical_knowledge": 0.7241379310344828, + "mmlu_eval_accuracy_college_biology": 0.8125, + "mmlu_eval_accuracy_college_chemistry": 0.375, + "mmlu_eval_accuracy_college_computer_science": 0.5454545454545454, + "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453, + "mmlu_eval_accuracy_college_medicine": 0.6818181818181818, + "mmlu_eval_accuracy_college_physics": 0.7272727272727273, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.6153846153846154, + "mmlu_eval_accuracy_econometrics": 0.5, + "mmlu_eval_accuracy_electrical_engineering": 0.5625, + "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634, + "mmlu_eval_accuracy_formal_logic": 0.5, + "mmlu_eval_accuracy_global_facts": 0.4, + "mmlu_eval_accuracy_high_school_biology": 0.65625, + "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.9545454545454546, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.4883720930232558, + "mmlu_eval_accuracy_high_school_mathematics": 0.3448275862068966, + "mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923, + "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, + "mmlu_eval_accuracy_high_school_psychology": 0.9166666666666666, + "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, + "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727, + "mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693, + "mmlu_eval_accuracy_human_aging": 0.6086956521739131, + "mmlu_eval_accuracy_human_sexuality": 0.75, + "mmlu_eval_accuracy_international_law": 0.9230769230769231, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.5454545454545454, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.84, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7674418604651163, + "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, + "mmlu_eval_accuracy_moral_scenarios": 0.41, + "mmlu_eval_accuracy_nutrition": 0.7272727272727273, + "mmlu_eval_accuracy_philosophy": 0.7058823529411765, + "mmlu_eval_accuracy_prehistory": 0.6285714285714286, + "mmlu_eval_accuracy_professional_accounting": 0.25806451612903225, + "mmlu_eval_accuracy_professional_law": 0.4117647058823529, + "mmlu_eval_accuracy_professional_medicine": 0.7741935483870968, + "mmlu_eval_accuracy_professional_psychology": 0.6666666666666666, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.8148148148148148, + "mmlu_eval_accuracy_sociology": 0.9545454545454546, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.5555555555555556, + "mmlu_eval_accuracy_world_religions": 0.8421052631578947, + "mmlu_loss": 5.812497261690206, + "step": 1683 + }, + { + "epoch": 0.5310627562283191, + "grad_norm": 0.498046875, + "learning_rate": 0.0002, + "loss": 1.8613, + "step": 1684 + }, + { + "epoch": 0.5313781141595711, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 1.7239, + "step": 1685 + }, + { + "epoch": 0.5316934720908231, + "grad_norm": 0.7734375, + "learning_rate": 0.0002, + "loss": 2.284, + "step": 1686 + }, + { + "epoch": 0.5320088300220751, + "grad_norm": 0.65625, + "learning_rate": 0.0002, + "loss": 2.4905, + "step": 1687 + }, + { + "epoch": 0.532324187953327, + "grad_norm": 0.56640625, + "learning_rate": 0.0002, + "loss": 2.1796, + "step": 1688 + }, + { + "epoch": 0.532639545884579, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 2.049, + "step": 1689 + }, + { + "epoch": 0.532954903815831, + "grad_norm": 0.60546875, + "learning_rate": 0.0002, + "loss": 2.0297, + "step": 1690 + }, + { + "epoch": 0.5332702617470829, + "grad_norm": 0.890625, + "learning_rate": 0.0002, + "loss": 2.3787, + "step": 1691 + }, + { + "epoch": 0.533585619678335, + "grad_norm": 0.6015625, + "learning_rate": 0.0002, + "loss": 2.1766, + "step": 1692 + }, + { + "epoch": 0.5339009776095869, + "grad_norm": 0.76953125, + "learning_rate": 0.0002, + "loss": 2.5972, + "step": 1693 + }, + { + "epoch": 0.5342163355408388, + "grad_norm": 0.65625, + "learning_rate": 0.0002, + "loss": 2.3372, + "step": 1694 + }, + { + "epoch": 0.5345316934720908, + "grad_norm": 1.0859375, + "learning_rate": 0.0002, + "loss": 2.8134, + "step": 1695 + }, + { + "epoch": 0.5348470514033428, + "grad_norm": 0.87109375, + "learning_rate": 0.0002, + "loss": 2.6178, + "step": 1696 + }, + { + "epoch": 0.5351624093345948, + "grad_norm": 0.7421875, + "learning_rate": 0.0002, + "loss": 2.2491, + "step": 1697 + }, + { + "epoch": 0.5354777672658467, + "grad_norm": 0.66796875, + "learning_rate": 0.0002, + "loss": 2.2375, + "step": 1698 + }, + { + "epoch": 0.5357931251970987, + "grad_norm": 0.875, + "learning_rate": 0.0002, + "loss": 2.3785, + "step": 1699 + }, + { + "epoch": 0.5361084831283507, + "grad_norm": 0.96484375, + "learning_rate": 0.0002, + "loss": 2.1515, + "step": 1700 + }, + { + "epoch": 0.5364238410596026, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.071, + "step": 1701 + }, + { + "epoch": 0.5367391989908546, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 0.9735, + "step": 1702 + }, + { + "epoch": 0.5370545569221066, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.0737, + "step": 1703 + }, + { + "epoch": 0.5373699148533586, + "grad_norm": 0.18359375, + "learning_rate": 0.0002, + "loss": 1.0965, + "step": 1704 + }, + { + "epoch": 0.5376852727846105, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.0565, + "step": 1705 + }, + { + "epoch": 0.5380006307158625, + "grad_norm": 0.1630859375, + "learning_rate": 0.0002, + "loss": 1.1959, + "step": 1706 + }, + { + "epoch": 0.5383159886471145, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.0961, + "step": 1707 + }, + { + "epoch": 0.5386313465783664, + "grad_norm": 0.1513671875, + "learning_rate": 0.0002, + "loss": 1.0157, + "step": 1708 + }, + { + "epoch": 0.5389467045096185, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.0784, + "step": 1709 + }, + { + "epoch": 0.5392620624408704, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.1409, + "step": 1710 + }, + { + "epoch": 0.5395774203721223, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.0879, + "step": 1711 + }, + { + "epoch": 0.5398927783033743, + "grad_norm": 0.1787109375, + "learning_rate": 0.0002, + "loss": 1.1668, + "step": 1712 + }, + { + "epoch": 0.5402081362346263, + "grad_norm": 0.1689453125, + "learning_rate": 0.0002, + "loss": 1.046, + "step": 1713 + }, + { + "epoch": 0.5405234941658783, + "grad_norm": 0.177734375, + "learning_rate": 0.0002, + "loss": 1.187, + "step": 1714 + }, + { + "epoch": 0.5408388520971302, + "grad_norm": 0.1796875, + "learning_rate": 0.0002, + "loss": 1.1896, + "step": 1715 + }, + { + "epoch": 0.5411542100283823, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 1.2013, + "step": 1716 + }, + { + "epoch": 0.5414695679596342, + "grad_norm": 0.201171875, + "learning_rate": 0.0002, + "loss": 1.2116, + "step": 1717 + }, + { + "epoch": 0.5417849258908861, + "grad_norm": 0.2392578125, + "learning_rate": 0.0002, + "loss": 1.301, + "step": 1718 + }, + { + "epoch": 0.5421002838221382, + "grad_norm": 0.201171875, + "learning_rate": 0.0002, + "loss": 1.1377, + "step": 1719 + }, + { + "epoch": 0.5424156417533901, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 1.2893, + "step": 1720 + }, + { + "epoch": 0.5427309996846421, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 1.3606, + "step": 1721 + }, + { + "epoch": 0.543046357615894, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 1.5369, + "step": 1722 + }, + { + "epoch": 0.543361715547146, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.2707, + "step": 1723 + }, + { + "epoch": 0.543677073478398, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.5335, + "step": 1724 + }, + { + "epoch": 0.5439924314096499, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 1.6759, + "step": 1725 + }, + { + "epoch": 0.544307789340902, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 1.4226, + "step": 1726 + }, + { + "epoch": 0.5446231472721539, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 1.2235, + "step": 1727 + }, + { + "epoch": 0.5449385052034059, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 1.4166, + "step": 1728 + }, + { + "epoch": 0.5452538631346578, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 1.4283, + "step": 1729 + }, + { + "epoch": 0.5455692210659098, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 1.529, + "step": 1730 + }, + { + "epoch": 0.5458845789971618, + "grad_norm": 0.40234375, + "learning_rate": 0.0002, + "loss": 1.7404, + "step": 1731 + }, + { + "epoch": 0.5461999369284137, + "grad_norm": 0.443359375, + "learning_rate": 0.0002, + "loss": 1.715, + "step": 1732 + }, + { + "epoch": 0.5465152948596658, + "grad_norm": 0.462890625, + "learning_rate": 0.0002, + "loss": 1.8795, + "step": 1733 + }, + { + "epoch": 0.5468306527909177, + "grad_norm": 0.451171875, + "learning_rate": 0.0002, + "loss": 1.995, + "step": 1734 + }, + { + "epoch": 0.5471460107221696, + "grad_norm": 0.625, + "learning_rate": 0.0002, + "loss": 2.333, + "step": 1735 + }, + { + "epoch": 0.5474613686534217, + "grad_norm": 0.48046875, + "learning_rate": 0.0002, + "loss": 1.8815, + "step": 1736 + }, + { + "epoch": 0.5477767265846736, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 2.0686, + "step": 1737 + }, + { + "epoch": 0.5480920845159256, + "grad_norm": 0.6640625, + "learning_rate": 0.0002, + "loss": 2.152, + "step": 1738 + }, + { + "epoch": 0.5484074424471775, + "grad_norm": 0.51953125, + "learning_rate": 0.0002, + "loss": 1.9459, + "step": 1739 + }, + { + "epoch": 0.5487228003784295, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 2.1803, + "step": 1740 + }, + { + "epoch": 0.5490381583096815, + "grad_norm": 0.81640625, + "learning_rate": 0.0002, + "loss": 2.2721, + "step": 1741 + }, + { + "epoch": 0.5493535162409334, + "grad_norm": 0.64453125, + "learning_rate": 0.0002, + "loss": 2.1391, + "step": 1742 + }, + { + "epoch": 0.5496688741721855, + "grad_norm": 0.6796875, + "learning_rate": 0.0002, + "loss": 2.2637, + "step": 1743 + }, + { + "epoch": 0.5499842321034374, + "grad_norm": 0.6484375, + "learning_rate": 0.0002, + "loss": 2.3057, + "step": 1744 + }, + { + "epoch": 0.5502995900346894, + "grad_norm": 0.80078125, + "learning_rate": 0.0002, + "loss": 2.4886, + "step": 1745 + }, + { + "epoch": 0.5506149479659413, + "grad_norm": 0.6875, + "learning_rate": 0.0002, + "loss": 2.5021, + "step": 1746 + }, + { + "epoch": 0.5509303058971933, + "grad_norm": 0.91015625, + "learning_rate": 0.0002, + "loss": 2.6263, + "step": 1747 + }, + { + "epoch": 0.5512456638284453, + "grad_norm": 0.83984375, + "learning_rate": 0.0002, + "loss": 2.6732, + "step": 1748 + }, + { + "epoch": 0.5515610217596972, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 1.9112, + "step": 1749 + }, + { + "epoch": 0.5518763796909493, + "grad_norm": 0.9140625, + "learning_rate": 0.0002, + "loss": 2.671, + "step": 1750 + }, + { + "epoch": 0.5521917376222012, + "grad_norm": 0.1591796875, + "learning_rate": 0.0002, + "loss": 1.1247, + "step": 1751 + }, + { + "epoch": 0.5525070955534531, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.105, + "step": 1752 + }, + { + "epoch": 0.5528224534847052, + "grad_norm": 0.1474609375, + "learning_rate": 0.0002, + "loss": 1.0308, + "step": 1753 + }, + { + "epoch": 0.5531378114159571, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.1295, + "step": 1754 + }, + { + "epoch": 0.5534531693472091, + "grad_norm": 0.1640625, + "learning_rate": 0.0002, + "loss": 1.1344, + "step": 1755 + }, + { + "epoch": 0.553768527278461, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.1351, + "step": 1756 + }, + { + "epoch": 0.5540838852097131, + "grad_norm": 0.1591796875, + "learning_rate": 0.0002, + "loss": 1.0449, + "step": 1757 + }, + { + "epoch": 0.554399243140965, + "grad_norm": 0.1640625, + "learning_rate": 0.0002, + "loss": 1.0401, + "step": 1758 + }, + { + "epoch": 0.5547146010722169, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.1616, + "step": 1759 + }, + { + "epoch": 0.555029959003469, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.1423, + "step": 1760 + }, + { + "epoch": 0.5553453169347209, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.2394, + "step": 1761 + }, + { + "epoch": 0.5556606748659729, + "grad_norm": 0.2041015625, + "learning_rate": 0.0002, + "loss": 1.2648, + "step": 1762 + }, + { + "epoch": 0.5559760327972249, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 0.9897, + "step": 1763 + }, + { + "epoch": 0.5562913907284768, + "grad_norm": 0.1884765625, + "learning_rate": 0.0002, + "loss": 1.2403, + "step": 1764 + }, + { + "epoch": 0.5566067486597288, + "grad_norm": 0.19921875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 1765 + }, + { + "epoch": 0.5569221065909807, + "grad_norm": 0.1787109375, + "learning_rate": 0.0002, + "loss": 1.1026, + "step": 1766 + }, + { + "epoch": 0.5572374645222328, + "grad_norm": 0.1796875, + "learning_rate": 0.0002, + "loss": 1.2304, + "step": 1767 + }, + { + "epoch": 0.5575528224534847, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.5004, + "step": 1768 + }, + { + "epoch": 0.5578681803847366, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.213, + "step": 1769 + }, + { + "epoch": 0.5581835383159887, + "grad_norm": 0.220703125, + "learning_rate": 0.0002, + "loss": 1.4348, + "step": 1770 + }, + { + "epoch": 0.5584988962472406, + "grad_norm": 0.220703125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 1771 + }, + { + "epoch": 0.5588142541784926, + "grad_norm": 0.2470703125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 1772 + }, + { + "epoch": 0.5591296121097445, + "grad_norm": 0.2431640625, + "learning_rate": 0.0002, + "loss": 1.3859, + "step": 1773 + }, + { + "epoch": 0.5594449700409966, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 1.4814, + "step": 1774 + }, + { + "epoch": 0.5597603279722485, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.5698, + "step": 1775 + }, + { + "epoch": 0.5600756859035004, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.2736, + "step": 1776 + }, + { + "epoch": 0.5603910438347525, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 1.5712, + "step": 1777 + }, + { + "epoch": 0.5607064017660044, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 1.597, + "step": 1778 + }, + { + "epoch": 0.5610217596972564, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 1.3767, + "step": 1779 + }, + { + "epoch": 0.5613371176285084, + "grad_norm": 0.3515625, + "learning_rate": 0.0002, + "loss": 1.5996, + "step": 1780 + }, + { + "epoch": 0.5616524755597603, + "grad_norm": 0.423828125, + "learning_rate": 0.0002, + "loss": 1.7394, + "step": 1781 + }, + { + "epoch": 0.5619678334910123, + "grad_norm": 0.37890625, + "learning_rate": 0.0002, + "loss": 1.5119, + "step": 1782 + }, + { + "epoch": 0.5622831914222642, + "grad_norm": 0.4609375, + "learning_rate": 0.0002, + "loss": 1.7704, + "step": 1783 + }, + { + "epoch": 0.5625985493535163, + "grad_norm": 0.44140625, + "learning_rate": 0.0002, + "loss": 1.5578, + "step": 1784 + }, + { + "epoch": 0.5629139072847682, + "grad_norm": 0.63671875, + "learning_rate": 0.0002, + "loss": 2.2498, + "step": 1785 + }, + { + "epoch": 0.5632292652160202, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 1.9366, + "step": 1786 + }, + { + "epoch": 0.5635446231472722, + "grad_norm": 0.56640625, + "learning_rate": 0.0002, + "loss": 2.1193, + "step": 1787 + }, + { + "epoch": 0.5638599810785241, + "grad_norm": 0.56640625, + "learning_rate": 0.0002, + "loss": 2.1479, + "step": 1788 + }, + { + "epoch": 0.5641753390097761, + "grad_norm": 0.71484375, + "learning_rate": 0.0002, + "loss": 2.1727, + "step": 1789 + }, + { + "epoch": 0.564490696941028, + "grad_norm": 0.63671875, + "learning_rate": 0.0002, + "loss": 2.3727, + "step": 1790 + }, + { + "epoch": 0.5648060548722801, + "grad_norm": 0.640625, + "learning_rate": 0.0002, + "loss": 2.0171, + "step": 1791 + }, + { + "epoch": 0.565121412803532, + "grad_norm": 0.703125, + "learning_rate": 0.0002, + "loss": 2.7968, + "step": 1792 + }, + { + "epoch": 0.5654367707347839, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 2.6643, + "step": 1793 + }, + { + "epoch": 0.565752128666036, + "grad_norm": 0.8984375, + "learning_rate": 0.0002, + "loss": 2.7223, + "step": 1794 + }, + { + "epoch": 0.5660674865972879, + "grad_norm": 0.71484375, + "learning_rate": 0.0002, + "loss": 2.5056, + "step": 1795 + }, + { + "epoch": 0.5663828445285399, + "grad_norm": 0.609375, + "learning_rate": 0.0002, + "loss": 1.6946, + "step": 1796 + }, + { + "epoch": 0.5666982024597919, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 1.8502, + "step": 1797 + }, + { + "epoch": 0.5670135603910438, + "grad_norm": 0.9765625, + "learning_rate": 0.0002, + "loss": 2.3839, + "step": 1798 + }, + { + "epoch": 0.5673289183222958, + "grad_norm": 0.74609375, + "learning_rate": 0.0002, + "loss": 2.1396, + "step": 1799 + }, + { + "epoch": 0.5676442762535477, + "grad_norm": 0.8984375, + "learning_rate": 0.0002, + "loss": 2.0341, + "step": 1800 + }, + { + "epoch": 0.5679596341847998, + "grad_norm": 0.16015625, + "learning_rate": 0.0002, + "loss": 1.058, + "step": 1801 + }, + { + "epoch": 0.5682749921160517, + "grad_norm": 0.2177734375, + "learning_rate": 0.0002, + "loss": 1.0242, + "step": 1802 + }, + { + "epoch": 0.5685903500473037, + "grad_norm": 0.142578125, + "learning_rate": 0.0002, + "loss": 1.057, + "step": 1803 + }, + { + "epoch": 0.5689057079785557, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.149, + "step": 1804 + }, + { + "epoch": 0.5692210659098076, + "grad_norm": 0.2294921875, + "learning_rate": 0.0002, + "loss": 1.1326, + "step": 1805 + }, + { + "epoch": 0.5695364238410596, + "grad_norm": 0.1689453125, + "learning_rate": 0.0002, + "loss": 1.1525, + "step": 1806 + }, + { + "epoch": 0.5698517817723115, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.0488, + "step": 1807 + }, + { + "epoch": 0.5701671397035636, + "grad_norm": 0.16015625, + "learning_rate": 0.0002, + "loss": 1.1572, + "step": 1808 + }, + { + "epoch": 0.5704824976348155, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.0773, + "step": 1809 + }, + { + "epoch": 0.5707978555660674, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.1059, + "step": 1810 + }, + { + "epoch": 0.5711132134973195, + "grad_norm": 0.1962890625, + "learning_rate": 0.0002, + "loss": 1.3008, + "step": 1811 + }, + { + "epoch": 0.5714285714285714, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.1358, + "step": 1812 + }, + { + "epoch": 0.5717439293598234, + "grad_norm": 0.1865234375, + "learning_rate": 0.0002, + "loss": 1.2395, + "step": 1813 + }, + { + "epoch": 0.5720592872910754, + "grad_norm": 0.193359375, + "learning_rate": 0.0002, + "loss": 1.0587, + "step": 1814 + }, + { + "epoch": 0.5723746452223274, + "grad_norm": 0.212890625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 1815 + }, + { + "epoch": 0.5726900031535793, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 1.1546, + "step": 1816 + }, + { + "epoch": 0.5730053610848312, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 1.191, + "step": 1817 + }, + { + "epoch": 0.5733207190160833, + "grad_norm": 0.1962890625, + "learning_rate": 0.0002, + "loss": 1.2124, + "step": 1818 + }, + { + "epoch": 0.5736360769473352, + "grad_norm": 0.216796875, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 1819 + }, + { + "epoch": 0.5739514348785872, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.1827, + "step": 1820 + }, + { + "epoch": 0.5742667928098392, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.245, + "step": 1821 + }, + { + "epoch": 0.5745821507410911, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 1.2666, + "step": 1822 + }, + { + "epoch": 0.5748975086723431, + "grad_norm": 0.2431640625, + "learning_rate": 0.0002, + "loss": 1.1943, + "step": 1823 + }, + { + "epoch": 0.575212866603595, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 1.3766, + "step": 1824 + }, + { + "epoch": 0.5755282245348471, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.2017, + "step": 1825 + }, + { + "epoch": 0.575843582466099, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.4419, + "step": 1826 + }, + { + "epoch": 0.5761589403973509, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 1.2472, + "step": 1827 + }, + { + "epoch": 0.576474298328603, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.4287, + "step": 1828 + }, + { + "epoch": 0.5767896562598549, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 1.4513, + "step": 1829 + }, + { + "epoch": 0.5771050141911069, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 1.4393, + "step": 1830 + }, + { + "epoch": 0.5774203721223589, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 1.3815, + "step": 1831 + }, + { + "epoch": 0.5777357300536109, + "grad_norm": 0.365234375, + "learning_rate": 0.0002, + "loss": 1.7161, + "step": 1832 + }, + { + "epoch": 0.5780510879848628, + "grad_norm": 0.419921875, + "learning_rate": 0.0002, + "loss": 1.7358, + "step": 1833 + }, + { + "epoch": 0.5783664459161147, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 1.9796, + "step": 1834 + }, + { + "epoch": 0.5786818038473668, + "grad_norm": 0.48828125, + "learning_rate": 0.0002, + "loss": 1.9401, + "step": 1835 + }, + { + "epoch": 0.5789971617786187, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 1.8664, + "step": 1836 + }, + { + "epoch": 0.5793125197098707, + "grad_norm": 0.65625, + "learning_rate": 0.0002, + "loss": 2.4499, + "step": 1837 + }, + { + "epoch": 0.5796278776411227, + "grad_norm": 0.490234375, + "learning_rate": 0.0002, + "loss": 1.9624, + "step": 1838 + }, + { + "epoch": 0.5799432355723746, + "grad_norm": 0.546875, + "learning_rate": 0.0002, + "loss": 2.2097, + "step": 1839 + }, + { + "epoch": 0.5802585935036266, + "grad_norm": 0.625, + "learning_rate": 0.0002, + "loss": 2.3008, + "step": 1840 + }, + { + "epoch": 0.5805739514348786, + "grad_norm": 0.7265625, + "learning_rate": 0.0002, + "loss": 2.6313, + "step": 1841 + }, + { + "epoch": 0.5808893093661306, + "grad_norm": 0.6640625, + "learning_rate": 0.0002, + "loss": 2.2295, + "step": 1842 + }, + { + "epoch": 0.5812046672973825, + "grad_norm": 0.67578125, + "learning_rate": 0.0002, + "loss": 2.1921, + "step": 1843 + }, + { + "epoch": 0.5815200252286346, + "grad_norm": 1.015625, + "learning_rate": 0.0002, + "loss": 2.7551, + "step": 1844 + }, + { + "epoch": 0.5818353831598865, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 2.017, + "step": 1845 + }, + { + "epoch": 0.5821507410911384, + "grad_norm": 0.89453125, + "learning_rate": 0.0002, + "loss": 2.4157, + "step": 1846 + }, + { + "epoch": 0.5824660990223904, + "grad_norm": 1.234375, + "learning_rate": 0.0002, + "loss": 2.2565, + "step": 1847 + }, + { + "epoch": 0.5827814569536424, + "grad_norm": 0.796875, + "learning_rate": 0.0002, + "loss": 2.54, + "step": 1848 + }, + { + "epoch": 0.5830968148848944, + "grad_norm": 0.80078125, + "learning_rate": 0.0002, + "loss": 2.3615, + "step": 1849 + }, + { + "epoch": 0.5834121728161463, + "grad_norm": 0.85546875, + "learning_rate": 0.0002, + "loss": 2.6269, + "step": 1850 + }, + { + "epoch": 0.5837275307473982, + "grad_norm": 0.1494140625, + "learning_rate": 0.0002, + "loss": 1.1035, + "step": 1851 + }, + { + "epoch": 0.5840428886786503, + "grad_norm": 0.1513671875, + "learning_rate": 0.0002, + "loss": 1.1124, + "step": 1852 + }, + { + "epoch": 0.5843582466099022, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.1618, + "step": 1853 + }, + { + "epoch": 0.5846736045411542, + "grad_norm": 0.166015625, + "learning_rate": 0.0002, + "loss": 1.1753, + "step": 1854 + }, + { + "epoch": 0.5849889624724062, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.1687, + "step": 1855 + }, + { + "epoch": 0.5853043204036582, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.0488, + "step": 1856 + }, + { + "epoch": 0.5856196783349101, + "grad_norm": 0.16015625, + "learning_rate": 0.0002, + "loss": 1.033, + "step": 1857 + }, + { + "epoch": 0.5859350362661621, + "grad_norm": 0.166015625, + "learning_rate": 0.0002, + "loss": 1.1537, + "step": 1858 + }, + { + "epoch": 0.5862503941974141, + "grad_norm": 0.1689453125, + "learning_rate": 0.0002, + "loss": 1.1068, + "step": 1859 + }, + { + "epoch": 0.586565752128666, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.1355, + "step": 1860 + }, + { + "epoch": 0.586881110059918, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 1.2588, + "step": 1861 + }, + { + "epoch": 0.58719646799117, + "grad_norm": 0.1787109375, + "learning_rate": 0.0002, + "loss": 1.1468, + "step": 1862 + }, + { + "epoch": 0.5875118259224219, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 1.1612, + "step": 1863 + }, + { + "epoch": 0.5878271838536739, + "grad_norm": 0.177734375, + "learning_rate": 0.0002, + "loss": 1.0973, + "step": 1864 + }, + { + "epoch": 0.5881425417849259, + "grad_norm": 0.205078125, + "learning_rate": 0.0002, + "loss": 1.2321, + "step": 1865 + }, + { + "epoch": 0.5884578997161779, + "grad_norm": 0.2451171875, + "learning_rate": 0.0002, + "loss": 1.4068, + "step": 1866 + }, + { + "epoch": 0.5887732576474298, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 1867 + }, + { + "epoch": 0.5890886155786818, + "grad_norm": 0.201171875, + "learning_rate": 0.0002, + "loss": 1.252, + "step": 1868 + }, + { + "epoch": 0.5894039735099338, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 1869 + }, + { + "epoch": 0.5897193314411857, + "grad_norm": 0.2333984375, + "learning_rate": 0.0002, + "loss": 1.2777, + "step": 1870 + }, + { + "epoch": 0.5897193314411857, + "eval_loss": 1.6543413400650024, + "eval_runtime": 148.9622, + "eval_samples_per_second": 6.713, + "eval_steps_per_second": 6.713, + "step": 1870 + }, + { + "epoch": 0.5897193314411857, + "mmlu_eval_accuracy": 0.6248503619812634, + "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, + "mmlu_eval_accuracy_anatomy": 0.5, + "mmlu_eval_accuracy_astronomy": 0.6875, + "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, + "mmlu_eval_accuracy_clinical_knowledge": 0.6896551724137931, + "mmlu_eval_accuracy_college_biology": 0.75, + "mmlu_eval_accuracy_college_chemistry": 0.625, + "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, + "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453, + "mmlu_eval_accuracy_college_medicine": 0.7272727272727273, + "mmlu_eval_accuracy_college_physics": 0.6363636363636364, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.5769230769230769, + "mmlu_eval_accuracy_econometrics": 0.5833333333333334, + "mmlu_eval_accuracy_electrical_engineering": 0.5625, + "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244, + "mmlu_eval_accuracy_formal_logic": 0.42857142857142855, + "mmlu_eval_accuracy_global_facts": 0.4, + "mmlu_eval_accuracy_high_school_biology": 0.625, + "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.9545454545454546, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442, + "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793, + "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, + "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, + "mmlu_eval_accuracy_high_school_psychology": 0.9166666666666666, + "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, + "mmlu_eval_accuracy_high_school_us_history": 0.8636363636363636, + "mmlu_eval_accuracy_high_school_world_history": 0.8076923076923077, + "mmlu_eval_accuracy_human_aging": 0.6521739130434783, + "mmlu_eval_accuracy_human_sexuality": 0.75, + "mmlu_eval_accuracy_international_law": 0.8461538461538461, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.88, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279, + "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, + "mmlu_eval_accuracy_moral_scenarios": 0.41, + "mmlu_eval_accuracy_nutrition": 0.696969696969697, + "mmlu_eval_accuracy_philosophy": 0.7352941176470589, + "mmlu_eval_accuracy_prehistory": 0.6, + "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194, + "mmlu_eval_accuracy_professional_law": 0.3941176470588235, + "mmlu_eval_accuracy_professional_medicine": 0.7741935483870968, + "mmlu_eval_accuracy_professional_psychology": 0.6376811594202898, + "mmlu_eval_accuracy_public_relations": 0.4166666666666667, + "mmlu_eval_accuracy_security_studies": 0.7777777777777778, + "mmlu_eval_accuracy_sociology": 0.9545454545454546, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.4444444444444444, + "mmlu_eval_accuracy_world_religions": 0.8421052631578947, + "mmlu_loss": 5.613258161239574, + "step": 1870 + }, + { + "epoch": 0.5900346893724377, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 1.3897, + "step": 1871 + }, + { + "epoch": 0.5903500473036897, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 1.519, + "step": 1872 + }, + { + "epoch": 0.5906654052349417, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 1.4295, + "step": 1873 + }, + { + "epoch": 0.5909807631661936, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.5358, + "step": 1874 + }, + { + "epoch": 0.5912961210974456, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 1.4393, + "step": 1875 + }, + { + "epoch": 0.5916114790286976, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.2858, + "step": 1876 + }, + { + "epoch": 0.5919268369599495, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.3779, + "step": 1877 + }, + { + "epoch": 0.5922421948912016, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 1.396, + "step": 1878 + }, + { + "epoch": 0.5925575528224535, + "grad_norm": 0.3984375, + "learning_rate": 0.0002, + "loss": 1.7037, + "step": 1879 + }, + { + "epoch": 0.5928729107537054, + "grad_norm": 0.3515625, + "learning_rate": 0.0002, + "loss": 1.5246, + "step": 1880 + }, + { + "epoch": 0.5931882686849574, + "grad_norm": 0.37109375, + "learning_rate": 0.0002, + "loss": 1.5004, + "step": 1881 + }, + { + "epoch": 0.5935036266162094, + "grad_norm": 0.52734375, + "learning_rate": 0.0002, + "loss": 2.2041, + "step": 1882 + }, + { + "epoch": 0.5938189845474614, + "grad_norm": 0.462890625, + "learning_rate": 0.0002, + "loss": 1.8278, + "step": 1883 + }, + { + "epoch": 0.5941343424787133, + "grad_norm": 0.44921875, + "learning_rate": 0.0002, + "loss": 1.6211, + "step": 1884 + }, + { + "epoch": 0.5944497004099654, + "grad_norm": 0.53515625, + "learning_rate": 0.0002, + "loss": 2.1753, + "step": 1885 + }, + { + "epoch": 0.5947650583412173, + "grad_norm": 0.546875, + "learning_rate": 0.0002, + "loss": 2.082, + "step": 1886 + }, + { + "epoch": 0.5950804162724692, + "grad_norm": 0.6953125, + "learning_rate": 0.0002, + "loss": 2.4544, + "step": 1887 + }, + { + "epoch": 0.5953957742037213, + "grad_norm": 0.6171875, + "learning_rate": 0.0002, + "loss": 2.1983, + "step": 1888 + }, + { + "epoch": 0.5957111321349732, + "grad_norm": 0.65625, + "learning_rate": 0.0002, + "loss": 2.2173, + "step": 1889 + }, + { + "epoch": 0.5960264900662252, + "grad_norm": 0.97265625, + "learning_rate": 0.0002, + "loss": 2.706, + "step": 1890 + }, + { + "epoch": 0.5963418479974771, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 2.1884, + "step": 1891 + }, + { + "epoch": 0.5966572059287291, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 2.2882, + "step": 1892 + }, + { + "epoch": 0.5969725638599811, + "grad_norm": 0.9296875, + "learning_rate": 0.0002, + "loss": 3.0429, + "step": 1893 + }, + { + "epoch": 0.597287921791233, + "grad_norm": 0.73828125, + "learning_rate": 0.0002, + "loss": 2.6665, + "step": 1894 + }, + { + "epoch": 0.5976032797224851, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 2.4846, + "step": 1895 + }, + { + "epoch": 0.597918637653737, + "grad_norm": 0.62890625, + "learning_rate": 0.0002, + "loss": 1.9601, + "step": 1896 + }, + { + "epoch": 0.5982339955849889, + "grad_norm": 0.7734375, + "learning_rate": 0.0002, + "loss": 2.4885, + "step": 1897 + }, + { + "epoch": 0.598549353516241, + "grad_norm": 0.9375, + "learning_rate": 0.0002, + "loss": 2.5466, + "step": 1898 + }, + { + "epoch": 0.5988647114474929, + "grad_norm": 0.64453125, + "learning_rate": 0.0002, + "loss": 2.0374, + "step": 1899 + }, + { + "epoch": 0.5991800693787449, + "grad_norm": 1.453125, + "learning_rate": 0.0002, + "loss": 3.3882, + "step": 1900 + }, + { + "epoch": 0.5994954273099968, + "grad_norm": 0.142578125, + "learning_rate": 0.0002, + "loss": 1.0917, + "step": 1901 + }, + { + "epoch": 0.5998107852412489, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.1062, + "step": 1902 + }, + { + "epoch": 0.6001261431725008, + "grad_norm": 0.1640625, + "learning_rate": 0.0002, + "loss": 1.0843, + "step": 1903 + }, + { + "epoch": 0.6004415011037527, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.1218, + "step": 1904 + }, + { + "epoch": 0.6007568590350048, + "grad_norm": 0.166015625, + "learning_rate": 0.0002, + "loss": 1.0871, + "step": 1905 + }, + { + "epoch": 0.6010722169662567, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 1.1474, + "step": 1906 + }, + { + "epoch": 0.6013875748975087, + "grad_norm": 0.16015625, + "learning_rate": 0.0002, + "loss": 1.1882, + "step": 1907 + }, + { + "epoch": 0.6017029328287606, + "grad_norm": 0.1591796875, + "learning_rate": 0.0002, + "loss": 1.1237, + "step": 1908 + }, + { + "epoch": 0.6020182907600126, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 1.0545, + "step": 1909 + }, + { + "epoch": 0.6023336486912646, + "grad_norm": 0.1865234375, + "learning_rate": 0.0002, + "loss": 1.0512, + "step": 1910 + }, + { + "epoch": 0.6026490066225165, + "grad_norm": 0.58984375, + "learning_rate": 0.0002, + "loss": 1.2776, + "step": 1911 + }, + { + "epoch": 0.6029643645537686, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.2092, + "step": 1912 + }, + { + "epoch": 0.6032797224850205, + "grad_norm": 0.2294921875, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 1913 + }, + { + "epoch": 0.6035950804162725, + "grad_norm": 0.2216796875, + "learning_rate": 0.0002, + "loss": 1.3709, + "step": 1914 + }, + { + "epoch": 0.6039104383475244, + "grad_norm": 0.2216796875, + "learning_rate": 0.0002, + "loss": 1.3005, + "step": 1915 + }, + { + "epoch": 0.6042257962787764, + "grad_norm": 0.22265625, + "learning_rate": 0.0002, + "loss": 1.2718, + "step": 1916 + }, + { + "epoch": 0.6045411542100284, + "grad_norm": 0.208984375, + "learning_rate": 0.0002, + "loss": 1.201, + "step": 1917 + }, + { + "epoch": 0.6048565121412803, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.2261, + "step": 1918 + }, + { + "epoch": 0.6051718700725324, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.3822, + "step": 1919 + }, + { + "epoch": 0.6054872280037843, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 1.2977, + "step": 1920 + }, + { + "epoch": 0.6058025859350362, + "grad_norm": 0.2451171875, + "learning_rate": 0.0002, + "loss": 1.2331, + "step": 1921 + }, + { + "epoch": 0.6061179438662883, + "grad_norm": 0.244140625, + "learning_rate": 0.0002, + "loss": 1.3, + "step": 1922 + }, + { + "epoch": 0.6064333017975402, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 1.4132, + "step": 1923 + }, + { + "epoch": 0.6067486597287922, + "grad_norm": 0.34765625, + "learning_rate": 0.0002, + "loss": 1.6919, + "step": 1924 + }, + { + "epoch": 0.6070640176600441, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 1.5168, + "step": 1925 + }, + { + "epoch": 0.6073793755912961, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 1.2471, + "step": 1926 + }, + { + "epoch": 0.6076947335225481, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 1.3042, + "step": 1927 + }, + { + "epoch": 0.6080100914538, + "grad_norm": 0.388671875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 1928 + }, + { + "epoch": 0.6083254493850521, + "grad_norm": 0.330078125, + "learning_rate": 0.0002, + "loss": 1.5074, + "step": 1929 + }, + { + "epoch": 0.608640807316304, + "grad_norm": 0.38671875, + "learning_rate": 0.0002, + "loss": 1.5218, + "step": 1930 + }, + { + "epoch": 0.608956165247556, + "grad_norm": 0.384765625, + "learning_rate": 0.0002, + "loss": 1.4689, + "step": 1931 + }, + { + "epoch": 0.609271523178808, + "grad_norm": 0.45703125, + "learning_rate": 0.0002, + "loss": 1.883, + "step": 1932 + }, + { + "epoch": 0.6095868811100599, + "grad_norm": 0.451171875, + "learning_rate": 0.0002, + "loss": 1.7069, + "step": 1933 + }, + { + "epoch": 0.6099022390413119, + "grad_norm": 0.494140625, + "learning_rate": 0.0002, + "loss": 1.7866, + "step": 1934 + }, + { + "epoch": 0.6102175969725638, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 1.5811, + "step": 1935 + }, + { + "epoch": 0.6105329549038159, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 2.3389, + "step": 1936 + }, + { + "epoch": 0.6108483128350678, + "grad_norm": 0.60546875, + "learning_rate": 0.0002, + "loss": 2.1602, + "step": 1937 + }, + { + "epoch": 0.6111636707663197, + "grad_norm": 0.5703125, + "learning_rate": 0.0002, + "loss": 1.9258, + "step": 1938 + }, + { + "epoch": 0.6114790286975718, + "grad_norm": 0.6953125, + "learning_rate": 0.0002, + "loss": 2.4463, + "step": 1939 + }, + { + "epoch": 0.6117943866288237, + "grad_norm": 0.66796875, + "learning_rate": 0.0002, + "loss": 2.4257, + "step": 1940 + }, + { + "epoch": 0.6121097445600757, + "grad_norm": 0.66796875, + "learning_rate": 0.0002, + "loss": 2.3827, + "step": 1941 + }, + { + "epoch": 0.6124251024913276, + "grad_norm": 0.625, + "learning_rate": 0.0002, + "loss": 2.2228, + "step": 1942 + }, + { + "epoch": 0.6127404604225797, + "grad_norm": 0.66796875, + "learning_rate": 0.0002, + "loss": 2.3855, + "step": 1943 + }, + { + "epoch": 0.6130558183538316, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 2.4568, + "step": 1944 + }, + { + "epoch": 0.6133711762850835, + "grad_norm": 0.7890625, + "learning_rate": 0.0002, + "loss": 2.8518, + "step": 1945 + }, + { + "epoch": 0.6136865342163356, + "grad_norm": 0.765625, + "learning_rate": 0.0002, + "loss": 2.5442, + "step": 1946 + }, + { + "epoch": 0.6140018921475875, + "grad_norm": 0.7578125, + "learning_rate": 0.0002, + "loss": 2.6823, + "step": 1947 + }, + { + "epoch": 0.6143172500788395, + "grad_norm": 0.7421875, + "learning_rate": 0.0002, + "loss": 2.1682, + "step": 1948 + }, + { + "epoch": 0.6146326080100915, + "grad_norm": 0.87109375, + "learning_rate": 0.0002, + "loss": 2.3595, + "step": 1949 + }, + { + "epoch": 0.6149479659413434, + "grad_norm": 1.078125, + "learning_rate": 0.0002, + "loss": 2.707, + "step": 1950 + }, + { + "epoch": 0.6152633238725954, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.0574, + "step": 1951 + }, + { + "epoch": 0.6155786818038473, + "grad_norm": 0.1591796875, + "learning_rate": 0.0002, + "loss": 1.138, + "step": 1952 + }, + { + "epoch": 0.6158940397350994, + "grad_norm": 0.177734375, + "learning_rate": 0.0002, + "loss": 1.1292, + "step": 1953 + }, + { + "epoch": 0.6162093976663513, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 0.9761, + "step": 1954 + }, + { + "epoch": 0.6165247555976032, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.0781, + "step": 1955 + }, + { + "epoch": 0.6168401135288553, + "grad_norm": 0.154296875, + "learning_rate": 0.0002, + "loss": 1.1821, + "step": 1956 + }, + { + "epoch": 0.6171554714601072, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.23, + "step": 1957 + }, + { + "epoch": 0.6174708293913592, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.2718, + "step": 1958 + }, + { + "epoch": 0.6177861873226111, + "grad_norm": 0.1689453125, + "learning_rate": 0.0002, + "loss": 1.1082, + "step": 1959 + }, + { + "epoch": 0.6181015452538632, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.2495, + "step": 1960 + }, + { + "epoch": 0.6184169031851151, + "grad_norm": 0.1787109375, + "learning_rate": 0.0002, + "loss": 1.1913, + "step": 1961 + }, + { + "epoch": 0.618732261116367, + "grad_norm": 0.18359375, + "learning_rate": 0.0002, + "loss": 1.1063, + "step": 1962 + }, + { + "epoch": 0.6190476190476191, + "grad_norm": 0.1875, + "learning_rate": 0.0002, + "loss": 1.1906, + "step": 1963 + }, + { + "epoch": 0.619362976978871, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.1672, + "step": 1964 + }, + { + "epoch": 0.619678334910123, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 1.2248, + "step": 1965 + }, + { + "epoch": 0.619993692841375, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.2185, + "step": 1966 + }, + { + "epoch": 0.6203090507726269, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.4492, + "step": 1967 + }, + { + "epoch": 0.6206244087038789, + "grad_norm": 0.1962890625, + "learning_rate": 0.0002, + "loss": 1.2657, + "step": 1968 + }, + { + "epoch": 0.6209397666351308, + "grad_norm": 0.2177734375, + "learning_rate": 0.0002, + "loss": 1.2467, + "step": 1969 + }, + { + "epoch": 0.6212551245663829, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 1.4009, + "step": 1970 + }, + { + "epoch": 0.6215704824976348, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.4323, + "step": 1971 + }, + { + "epoch": 0.6218858404288868, + "grad_norm": 0.2333984375, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 1972 + }, + { + "epoch": 0.6222011983601388, + "grad_norm": 0.2490234375, + "learning_rate": 0.0002, + "loss": 1.3045, + "step": 1973 + }, + { + "epoch": 0.6225165562913907, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.1842, + "step": 1974 + }, + { + "epoch": 0.6228319142226427, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 1975 + }, + { + "epoch": 0.6231472721538946, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 1976 + }, + { + "epoch": 0.6234626300851467, + "grad_norm": 0.3671875, + "learning_rate": 0.0002, + "loss": 1.4972, + "step": 1977 + }, + { + "epoch": 0.6237779880163986, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 1.4306, + "step": 1978 + }, + { + "epoch": 0.6240933459476505, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 1.3645, + "step": 1979 + }, + { + "epoch": 0.6244087038789026, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 1.3725, + "step": 1980 + }, + { + "epoch": 0.6247240618101545, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 1.4471, + "step": 1981 + }, + { + "epoch": 0.6250394197414065, + "grad_norm": 0.400390625, + "learning_rate": 0.0002, + "loss": 1.5261, + "step": 1982 + }, + { + "epoch": 0.6253547776726585, + "grad_norm": 0.5, + "learning_rate": 0.0002, + "loss": 1.9231, + "step": 1983 + }, + { + "epoch": 0.6256701356039105, + "grad_norm": 0.640625, + "learning_rate": 0.0002, + "loss": 1.9727, + "step": 1984 + }, + { + "epoch": 0.6259854935351624, + "grad_norm": 0.47265625, + "learning_rate": 0.0002, + "loss": 1.918, + "step": 1985 + }, + { + "epoch": 0.6263008514664143, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 2.0199, + "step": 1986 + }, + { + "epoch": 0.6266162093976664, + "grad_norm": 0.70703125, + "learning_rate": 0.0002, + "loss": 2.3791, + "step": 1987 + }, + { + "epoch": 0.6269315673289183, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 2.0079, + "step": 1988 + }, + { + "epoch": 0.6272469252601703, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 1.6909, + "step": 1989 + }, + { + "epoch": 0.6275622831914223, + "grad_norm": 0.671875, + "learning_rate": 0.0002, + "loss": 2.2142, + "step": 1990 + }, + { + "epoch": 0.6278776411226742, + "grad_norm": 0.6171875, + "learning_rate": 0.0002, + "loss": 2.5312, + "step": 1991 + }, + { + "epoch": 0.6281929990539262, + "grad_norm": 0.76171875, + "learning_rate": 0.0002, + "loss": 2.4839, + "step": 1992 + }, + { + "epoch": 0.6285083569851782, + "grad_norm": 0.66796875, + "learning_rate": 0.0002, + "loss": 2.311, + "step": 1993 + }, + { + "epoch": 0.6288237149164302, + "grad_norm": 0.625, + "learning_rate": 0.0002, + "loss": 2.0299, + "step": 1994 + }, + { + "epoch": 0.6291390728476821, + "grad_norm": 0.7890625, + "learning_rate": 0.0002, + "loss": 2.3085, + "step": 1995 + }, + { + "epoch": 0.629454430778934, + "grad_norm": 0.7734375, + "learning_rate": 0.0002, + "loss": 2.2304, + "step": 1996 + }, + { + "epoch": 0.6297697887101861, + "grad_norm": 0.73046875, + "learning_rate": 0.0002, + "loss": 2.2698, + "step": 1997 + }, + { + "epoch": 0.630085146641438, + "grad_norm": 0.80078125, + "learning_rate": 0.0002, + "loss": 2.3537, + "step": 1998 + }, + { + "epoch": 0.63040050457269, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 2.973, + "step": 1999 + }, + { + "epoch": 0.630715862503942, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 2.2126, + "step": 2000 + }, + { + "epoch": 0.631031220435194, + "grad_norm": 0.142578125, + "learning_rate": 0.0002, + "loss": 1.1315, + "step": 2001 + }, + { + "epoch": 0.6313465783664459, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.0522, + "step": 2002 + }, + { + "epoch": 0.6316619362976978, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.0685, + "step": 2003 + }, + { + "epoch": 0.6319772942289499, + "grad_norm": 0.154296875, + "learning_rate": 0.0002, + "loss": 1.0932, + "step": 2004 + }, + { + "epoch": 0.6322926521602018, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.267, + "step": 2005 + }, + { + "epoch": 0.6326080100914538, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.0974, + "step": 2006 + }, + { + "epoch": 0.6329233680227058, + "grad_norm": 0.166015625, + "learning_rate": 0.0002, + "loss": 0.961, + "step": 2007 + }, + { + "epoch": 0.6332387259539577, + "grad_norm": 0.1767578125, + "learning_rate": 0.0002, + "loss": 1.14, + "step": 2008 + }, + { + "epoch": 0.6335540838852097, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.2619, + "step": 2009 + }, + { + "epoch": 0.6338694418164617, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.1347, + "step": 2010 + }, + { + "epoch": 0.6341847997477137, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.1244, + "step": 2011 + }, + { + "epoch": 0.6345001576789656, + "grad_norm": 0.2138671875, + "learning_rate": 0.0002, + "loss": 1.2154, + "step": 2012 + }, + { + "epoch": 0.6348155156102177, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 1.2252, + "step": 2013 + }, + { + "epoch": 0.6351308735414696, + "grad_norm": 0.1884765625, + "learning_rate": 0.0002, + "loss": 1.1314, + "step": 2014 + }, + { + "epoch": 0.6354462314727215, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.1097, + "step": 2015 + }, + { + "epoch": 0.6357615894039735, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.2582, + "step": 2016 + }, + { + "epoch": 0.6360769473352255, + "grad_norm": 0.2265625, + "learning_rate": 0.0002, + "loss": 1.2, + "step": 2017 + }, + { + "epoch": 0.6363923052664775, + "grad_norm": 0.2216796875, + "learning_rate": 0.0002, + "loss": 1.2702, + "step": 2018 + }, + { + "epoch": 0.6367076631977294, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 2019 + }, + { + "epoch": 0.6370230211289813, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.5556, + "step": 2020 + }, + { + "epoch": 0.6373383790602334, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 1.4595, + "step": 2021 + }, + { + "epoch": 0.6376537369914853, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.2686, + "step": 2022 + }, + { + "epoch": 0.6379690949227373, + "grad_norm": 0.244140625, + "learning_rate": 0.0002, + "loss": 1.115, + "step": 2023 + }, + { + "epoch": 0.6382844528539893, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.6663, + "step": 2024 + }, + { + "epoch": 0.6385998107852412, + "grad_norm": 0.478515625, + "learning_rate": 0.0002, + "loss": 1.7185, + "step": 2025 + }, + { + "epoch": 0.6389151687164932, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 1.4392, + "step": 2026 + }, + { + "epoch": 0.6392305266477452, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 1.6242, + "step": 2027 + }, + { + "epoch": 0.6395458845789972, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 1.2242, + "step": 2028 + }, + { + "epoch": 0.6398612425102491, + "grad_norm": 0.380859375, + "learning_rate": 0.0002, + "loss": 1.4124, + "step": 2029 + }, + { + "epoch": 0.6401766004415012, + "grad_norm": 0.3828125, + "learning_rate": 0.0002, + "loss": 1.6297, + "step": 2030 + }, + { + "epoch": 0.6404919583727531, + "grad_norm": 0.4296875, + "learning_rate": 0.0002, + "loss": 1.8021, + "step": 2031 + }, + { + "epoch": 0.640807316304005, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 1.8951, + "step": 2032 + }, + { + "epoch": 0.641122674235257, + "grad_norm": 0.58203125, + "learning_rate": 0.0002, + "loss": 2.0222, + "step": 2033 + }, + { + "epoch": 0.641438032166509, + "grad_norm": 0.453125, + "learning_rate": 0.0002, + "loss": 1.7258, + "step": 2034 + }, + { + "epoch": 0.641753390097761, + "grad_norm": 0.5546875, + "learning_rate": 0.0002, + "loss": 2.1058, + "step": 2035 + }, + { + "epoch": 0.6420687480290129, + "grad_norm": 0.515625, + "learning_rate": 0.0002, + "loss": 1.9603, + "step": 2036 + }, + { + "epoch": 0.6423841059602649, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 2.1623, + "step": 2037 + }, + { + "epoch": 0.6426994638915169, + "grad_norm": 0.61328125, + "learning_rate": 0.0002, + "loss": 2.5245, + "step": 2038 + }, + { + "epoch": 0.6430148218227688, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 2.2672, + "step": 2039 + }, + { + "epoch": 0.6433301797540208, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 1.9878, + "step": 2040 + }, + { + "epoch": 0.6436455376852728, + "grad_norm": 0.7421875, + "learning_rate": 0.0002, + "loss": 2.8974, + "step": 2041 + }, + { + "epoch": 0.6439608956165248, + "grad_norm": 0.62890625, + "learning_rate": 0.0002, + "loss": 2.3302, + "step": 2042 + }, + { + "epoch": 0.6442762535477767, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 2.6995, + "step": 2043 + }, + { + "epoch": 0.6445916114790287, + "grad_norm": 0.80859375, + "learning_rate": 0.0002, + "loss": 2.4607, + "step": 2044 + }, + { + "epoch": 0.6449069694102807, + "grad_norm": 0.6875, + "learning_rate": 0.0002, + "loss": 2.4102, + "step": 2045 + }, + { + "epoch": 0.6452223273415326, + "grad_norm": 0.91796875, + "learning_rate": 0.0002, + "loss": 2.569, + "step": 2046 + }, + { + "epoch": 0.6455376852727847, + "grad_norm": 0.80078125, + "learning_rate": 0.0002, + "loss": 2.4496, + "step": 2047 + }, + { + "epoch": 0.6458530432040366, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 2.8678, + "step": 2048 + }, + { + "epoch": 0.6461684011352885, + "grad_norm": 0.76171875, + "learning_rate": 0.0002, + "loss": 1.9381, + "step": 2049 + }, + { + "epoch": 0.6464837590665405, + "grad_norm": 0.953125, + "learning_rate": 0.0002, + "loss": 2.4581, + "step": 2050 + }, + { + "epoch": 0.6467991169977925, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.0895, + "step": 2051 + }, + { + "epoch": 0.6471144749290445, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.1131, + "step": 2052 + }, + { + "epoch": 0.6474298328602964, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.1707, + "step": 2053 + }, + { + "epoch": 0.6477451907915484, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.1499, + "step": 2054 + }, + { + "epoch": 0.6480605487228004, + "grad_norm": 0.1630859375, + "learning_rate": 0.0002, + "loss": 1.0771, + "step": 2055 + }, + { + "epoch": 0.6483759066540523, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.088, + "step": 2056 + }, + { + "epoch": 0.6486912645853044, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.1958, + "step": 2057 + }, + { + "epoch": 0.6486912645853044, + "eval_loss": 1.6923353672027588, + "eval_runtime": 148.8146, + "eval_samples_per_second": 6.72, + "eval_steps_per_second": 6.72, + "step": 2057 + }, + { + "epoch": 0.6486912645853044, + "mmlu_eval_accuracy": 0.6088051824290741, + "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, + "mmlu_eval_accuracy_anatomy": 0.5714285714285714, + "mmlu_eval_accuracy_astronomy": 0.6875, + "mmlu_eval_accuracy_business_ethics": 0.45454545454545453, + "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034, + "mmlu_eval_accuracy_college_biology": 0.8125, + "mmlu_eval_accuracy_college_chemistry": 0.5, + "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365, + "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365, + "mmlu_eval_accuracy_college_medicine": 0.6818181818181818, + "mmlu_eval_accuracy_college_physics": 0.6363636363636364, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384, + "mmlu_eval_accuracy_econometrics": 0.4166666666666667, + "mmlu_eval_accuracy_electrical_engineering": 0.5, + "mmlu_eval_accuracy_elementary_mathematics": 0.5121951219512195, + "mmlu_eval_accuracy_formal_logic": 0.2857142857142857, + "mmlu_eval_accuracy_global_facts": 0.3, + "mmlu_eval_accuracy_high_school_biology": 0.59375, + "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209, + "mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724, + "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769, + "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, + "mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333, + "mmlu_eval_accuracy_high_school_statistics": 0.4782608695652174, + "mmlu_eval_accuracy_high_school_us_history": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, + "mmlu_eval_accuracy_human_aging": 0.6086956521739131, + "mmlu_eval_accuracy_human_sexuality": 0.75, + "mmlu_eval_accuracy_international_law": 0.8461538461538461, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.88, + "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182, + "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279, + "mmlu_eval_accuracy_moral_disputes": 0.631578947368421, + "mmlu_eval_accuracy_moral_scenarios": 0.37, + "mmlu_eval_accuracy_nutrition": 0.696969696969697, + "mmlu_eval_accuracy_philosophy": 0.7647058823529411, + "mmlu_eval_accuracy_prehistory": 0.6, + "mmlu_eval_accuracy_professional_accounting": 0.3870967741935484, + "mmlu_eval_accuracy_professional_law": 0.4235294117647059, + "mmlu_eval_accuracy_professional_medicine": 0.7419354838709677, + "mmlu_eval_accuracy_professional_psychology": 0.6521739130434783, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7777777777777778, + "mmlu_eval_accuracy_sociology": 0.9545454545454546, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.3888888888888889, + "mmlu_eval_accuracy_world_religions": 0.8421052631578947, + "mmlu_loss": 5.507620678626969, + "step": 2057 + }, + { + "epoch": 0.6490066225165563, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 1.1029, + "step": 2058 + }, + { + "epoch": 0.6493219804478083, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.1546, + "step": 2059 + }, + { + "epoch": 0.6496373383790602, + "grad_norm": 0.201171875, + "learning_rate": 0.0002, + "loss": 1.3017, + "step": 2060 + }, + { + "epoch": 0.6499526963103122, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.113, + "step": 2061 + }, + { + "epoch": 0.6502680542415642, + "grad_norm": 0.1796875, + "learning_rate": 0.0002, + "loss": 1.2013, + "step": 2062 + }, + { + "epoch": 0.6505834121728161, + "grad_norm": 0.189453125, + "learning_rate": 0.0002, + "loss": 1.0702, + "step": 2063 + }, + { + "epoch": 0.6508987701040682, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.0474, + "step": 2064 + }, + { + "epoch": 0.6512141280353201, + "grad_norm": 0.1884765625, + "learning_rate": 0.0002, + "loss": 1.0136, + "step": 2065 + }, + { + "epoch": 0.651529485966572, + "grad_norm": 0.201171875, + "learning_rate": 0.0002, + "loss": 1.2202, + "step": 2066 + }, + { + "epoch": 0.651844843897824, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.0158, + "step": 2067 + }, + { + "epoch": 0.652160201829076, + "grad_norm": 0.224609375, + "learning_rate": 0.0002, + "loss": 1.2681, + "step": 2068 + }, + { + "epoch": 0.652475559760328, + "grad_norm": 0.2421875, + "learning_rate": 0.0002, + "loss": 1.3029, + "step": 2069 + }, + { + "epoch": 0.6527909176915799, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.2063, + "step": 2070 + }, + { + "epoch": 0.653106275622832, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 1.4116, + "step": 2071 + }, + { + "epoch": 0.6534216335540839, + "grad_norm": 0.3515625, + "learning_rate": 0.0002, + "loss": 1.3682, + "step": 2072 + }, + { + "epoch": 0.6537369914853358, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 1.5731, + "step": 2073 + }, + { + "epoch": 0.6540523494165879, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 1.2912, + "step": 2074 + }, + { + "epoch": 0.6543677073478398, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 1.2181, + "step": 2075 + }, + { + "epoch": 0.6546830652790918, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.3784, + "step": 2076 + }, + { + "epoch": 0.6549984232103437, + "grad_norm": 0.255859375, + "learning_rate": 0.0002, + "loss": 1.1312, + "step": 2077 + }, + { + "epoch": 0.6553137811415957, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 1.2805, + "step": 2078 + }, + { + "epoch": 0.6556291390728477, + "grad_norm": 0.3515625, + "learning_rate": 0.0002, + "loss": 1.6554, + "step": 2079 + }, + { + "epoch": 0.6559444970040996, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.5079, + "step": 2080 + }, + { + "epoch": 0.6562598549353517, + "grad_norm": 0.365234375, + "learning_rate": 0.0002, + "loss": 1.4851, + "step": 2081 + }, + { + "epoch": 0.6565752128666036, + "grad_norm": 0.3515625, + "learning_rate": 0.0002, + "loss": 1.5551, + "step": 2082 + }, + { + "epoch": 0.6568905707978555, + "grad_norm": 0.4765625, + "learning_rate": 0.0002, + "loss": 1.4823, + "step": 2083 + }, + { + "epoch": 0.6572059287291075, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 2.0317, + "step": 2084 + }, + { + "epoch": 0.6575212866603595, + "grad_norm": 0.412109375, + "learning_rate": 0.0002, + "loss": 1.6203, + "step": 2085 + }, + { + "epoch": 0.6578366445916115, + "grad_norm": 0.46875, + "learning_rate": 0.0002, + "loss": 1.6667, + "step": 2086 + }, + { + "epoch": 0.6581520025228634, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 1.9333, + "step": 2087 + }, + { + "epoch": 0.6584673604541155, + "grad_norm": 0.66796875, + "learning_rate": 0.0002, + "loss": 2.6031, + "step": 2088 + }, + { + "epoch": 0.6587827183853674, + "grad_norm": 0.69921875, + "learning_rate": 0.0002, + "loss": 2.0656, + "step": 2089 + }, + { + "epoch": 0.6590980763166193, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 2.3802, + "step": 2090 + }, + { + "epoch": 0.6594134342478714, + "grad_norm": 0.73046875, + "learning_rate": 0.0002, + "loss": 2.4076, + "step": 2091 + }, + { + "epoch": 0.6597287921791233, + "grad_norm": 0.7109375, + "learning_rate": 0.0002, + "loss": 2.6082, + "step": 2092 + }, + { + "epoch": 0.6600441501103753, + "grad_norm": 0.671875, + "learning_rate": 0.0002, + "loss": 2.4313, + "step": 2093 + }, + { + "epoch": 0.6603595080416272, + "grad_norm": 0.6953125, + "learning_rate": 0.0002, + "loss": 2.3582, + "step": 2094 + }, + { + "epoch": 0.6606748659728792, + "grad_norm": 0.75, + "learning_rate": 0.0002, + "loss": 2.5801, + "step": 2095 + }, + { + "epoch": 0.6609902239041312, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 1.9778, + "step": 2096 + }, + { + "epoch": 0.6613055818353831, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 2.5477, + "step": 2097 + }, + { + "epoch": 0.6616209397666352, + "grad_norm": 0.74609375, + "learning_rate": 0.0002, + "loss": 2.2557, + "step": 2098 + }, + { + "epoch": 0.6619362976978871, + "grad_norm": 0.83984375, + "learning_rate": 0.0002, + "loss": 2.1798, + "step": 2099 + }, + { + "epoch": 0.6622516556291391, + "grad_norm": 1.28125, + "learning_rate": 0.0002, + "loss": 2.4501, + "step": 2100 + }, + { + "epoch": 0.662567013560391, + "grad_norm": 0.1455078125, + "learning_rate": 0.0002, + "loss": 1.2141, + "step": 2101 + }, + { + "epoch": 0.662882371491643, + "grad_norm": 0.154296875, + "learning_rate": 0.0002, + "loss": 1.126, + "step": 2102 + }, + { + "epoch": 0.663197729422895, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.0411, + "step": 2103 + }, + { + "epoch": 0.6635130873541469, + "grad_norm": 0.158203125, + "learning_rate": 0.0002, + "loss": 1.0412, + "step": 2104 + }, + { + "epoch": 0.663828445285399, + "grad_norm": 0.166015625, + "learning_rate": 0.0002, + "loss": 1.1951, + "step": 2105 + }, + { + "epoch": 0.6641438032166509, + "grad_norm": 0.158203125, + "learning_rate": 0.0002, + "loss": 0.9379, + "step": 2106 + }, + { + "epoch": 0.6644591611479028, + "grad_norm": 0.1689453125, + "learning_rate": 0.0002, + "loss": 1.0688, + "step": 2107 + }, + { + "epoch": 0.6647745190791549, + "grad_norm": 0.1796875, + "learning_rate": 0.0002, + "loss": 1.1157, + "step": 2108 + }, + { + "epoch": 0.6650898770104068, + "grad_norm": 0.1796875, + "learning_rate": 0.0002, + "loss": 1.051, + "step": 2109 + }, + { + "epoch": 0.6654052349416588, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 1.1092, + "step": 2110 + }, + { + "epoch": 0.6657205928729107, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.1075, + "step": 2111 + }, + { + "epoch": 0.6660359508041628, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.0262, + "step": 2112 + }, + { + "epoch": 0.6663513087354147, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.2075, + "step": 2113 + }, + { + "epoch": 0.6666666666666666, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.1218, + "step": 2114 + }, + { + "epoch": 0.6669820245979187, + "grad_norm": 0.201171875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 2115 + }, + { + "epoch": 0.6672973825291706, + "grad_norm": 0.197265625, + "learning_rate": 0.0002, + "loss": 1.1666, + "step": 2116 + }, + { + "epoch": 0.6676127404604226, + "grad_norm": 0.19921875, + "learning_rate": 0.0002, + "loss": 1.2708, + "step": 2117 + }, + { + "epoch": 0.6679280983916746, + "grad_norm": 0.19140625, + "learning_rate": 0.0002, + "loss": 1.0451, + "step": 2118 + }, + { + "epoch": 0.6682434563229265, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.1923, + "step": 2119 + }, + { + "epoch": 0.6685588142541785, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 1.222, + "step": 2120 + }, + { + "epoch": 0.6688741721854304, + "grad_norm": 0.2373046875, + "learning_rate": 0.0002, + "loss": 1.4059, + "step": 2121 + }, + { + "epoch": 0.6691895301166825, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.0895, + "step": 2122 + }, + { + "epoch": 0.6695048880479344, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.4483, + "step": 2123 + }, + { + "epoch": 0.6698202459791863, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.4941, + "step": 2124 + }, + { + "epoch": 0.6701356039104384, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.6128, + "step": 2125 + }, + { + "epoch": 0.6704509618416903, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.39, + "step": 2126 + }, + { + "epoch": 0.6707663197729423, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 1.3993, + "step": 2127 + }, + { + "epoch": 0.6710816777041942, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 2128 + }, + { + "epoch": 0.6713970356354463, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 1.5245, + "step": 2129 + }, + { + "epoch": 0.6717123935666982, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 1.5547, + "step": 2130 + }, + { + "epoch": 0.6720277514979501, + "grad_norm": 0.384765625, + "learning_rate": 0.0002, + "loss": 1.5897, + "step": 2131 + }, + { + "epoch": 0.6723431094292022, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 1.4799, + "step": 2132 + }, + { + "epoch": 0.6726584673604541, + "grad_norm": 0.41796875, + "learning_rate": 0.0002, + "loss": 1.7749, + "step": 2133 + }, + { + "epoch": 0.6729738252917061, + "grad_norm": 0.484375, + "learning_rate": 0.0002, + "loss": 1.6808, + "step": 2134 + }, + { + "epoch": 0.673289183222958, + "grad_norm": 0.52734375, + "learning_rate": 0.0002, + "loss": 1.9236, + "step": 2135 + }, + { + "epoch": 0.67360454115421, + "grad_norm": 0.46875, + "learning_rate": 0.0002, + "loss": 1.9407, + "step": 2136 + }, + { + "epoch": 0.673919899085462, + "grad_norm": 0.4765625, + "learning_rate": 0.0002, + "loss": 1.7339, + "step": 2137 + }, + { + "epoch": 0.6742352570167139, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 2.1535, + "step": 2138 + }, + { + "epoch": 0.674550614947966, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 2.3963, + "step": 2139 + }, + { + "epoch": 0.6748659728792179, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 2.3325, + "step": 2140 + }, + { + "epoch": 0.6751813308104699, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 1.9217, + "step": 2141 + }, + { + "epoch": 0.6754966887417219, + "grad_norm": 0.7890625, + "learning_rate": 0.0002, + "loss": 2.5751, + "step": 2142 + }, + { + "epoch": 0.6758120466729738, + "grad_norm": 0.68359375, + "learning_rate": 0.0002, + "loss": 2.4127, + "step": 2143 + }, + { + "epoch": 0.6761274046042258, + "grad_norm": 0.62890625, + "learning_rate": 0.0002, + "loss": 2.0031, + "step": 2144 + }, + { + "epoch": 0.6764427625354777, + "grad_norm": 0.70703125, + "learning_rate": 0.0002, + "loss": 2.6867, + "step": 2145 + }, + { + "epoch": 0.6767581204667298, + "grad_norm": 0.6484375, + "learning_rate": 0.0002, + "loss": 2.2455, + "step": 2146 + }, + { + "epoch": 0.6770734783979817, + "grad_norm": 0.76953125, + "learning_rate": 0.0002, + "loss": 2.3334, + "step": 2147 + }, + { + "epoch": 0.6773888363292336, + "grad_norm": 0.80859375, + "learning_rate": 0.0002, + "loss": 2.2331, + "step": 2148 + }, + { + "epoch": 0.6777041942604857, + "grad_norm": 0.68359375, + "learning_rate": 0.0002, + "loss": 1.9905, + "step": 2149 + }, + { + "epoch": 0.6780195521917376, + "grad_norm": 0.9375, + "learning_rate": 0.0002, + "loss": 2.7107, + "step": 2150 + }, + { + "epoch": 0.6783349101229896, + "grad_norm": 0.140625, + "learning_rate": 0.0002, + "loss": 1.0281, + "step": 2151 + }, + { + "epoch": 0.6786502680542416, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.038, + "step": 2152 + }, + { + "epoch": 0.6789656259854935, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.1353, + "step": 2153 + }, + { + "epoch": 0.6792809839167455, + "grad_norm": 0.150390625, + "learning_rate": 0.0002, + "loss": 1.1591, + "step": 2154 + }, + { + "epoch": 0.6795963418479974, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.0941, + "step": 2155 + }, + { + "epoch": 0.6799116997792495, + "grad_norm": 0.158203125, + "learning_rate": 0.0002, + "loss": 1.0702, + "step": 2156 + }, + { + "epoch": 0.6802270577105014, + "grad_norm": 0.1591796875, + "learning_rate": 0.0002, + "loss": 1.1518, + "step": 2157 + }, + { + "epoch": 0.6805424156417534, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.0299, + "step": 2158 + }, + { + "epoch": 0.6808577735730054, + "grad_norm": 0.166015625, + "learning_rate": 0.0002, + "loss": 1.1233, + "step": 2159 + }, + { + "epoch": 0.6811731315042573, + "grad_norm": 0.1591796875, + "learning_rate": 0.0002, + "loss": 1.0661, + "step": 2160 + }, + { + "epoch": 0.6814884894355093, + "grad_norm": 0.197265625, + "learning_rate": 0.0002, + "loss": 1.1626, + "step": 2161 + }, + { + "epoch": 0.6818038473667613, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.1739, + "step": 2162 + }, + { + "epoch": 0.6821192052980133, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.0233, + "step": 2163 + }, + { + "epoch": 0.6824345632292652, + "grad_norm": 0.2470703125, + "learning_rate": 0.0002, + "loss": 1.3025, + "step": 2164 + }, + { + "epoch": 0.6827499211605171, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 1.1062, + "step": 2165 + }, + { + "epoch": 0.6830652790917692, + "grad_norm": 0.2177734375, + "learning_rate": 0.0002, + "loss": 1.2644, + "step": 2166 + }, + { + "epoch": 0.6833806370230211, + "grad_norm": 0.2041015625, + "learning_rate": 0.0002, + "loss": 1.1519, + "step": 2167 + }, + { + "epoch": 0.6836959949542731, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 2168 + }, + { + "epoch": 0.6840113528855251, + "grad_norm": 0.2421875, + "learning_rate": 0.0002, + "loss": 1.3743, + "step": 2169 + }, + { + "epoch": 0.6843267108167771, + "grad_norm": 0.2412109375, + "learning_rate": 0.0002, + "loss": 1.287, + "step": 2170 + }, + { + "epoch": 0.684642068748029, + "grad_norm": 0.23828125, + "learning_rate": 0.0002, + "loss": 1.4858, + "step": 2171 + }, + { + "epoch": 0.684957426679281, + "grad_norm": 0.255859375, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 2172 + }, + { + "epoch": 0.685272784610533, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 1.2715, + "step": 2173 + }, + { + "epoch": 0.6855881425417849, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 1.2564, + "step": 2174 + }, + { + "epoch": 0.6859035004730369, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.2535, + "step": 2175 + }, + { + "epoch": 0.6862188584042889, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 2176 + }, + { + "epoch": 0.6865342163355408, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 1.2955, + "step": 2177 + }, + { + "epoch": 0.6868495742667928, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 1.3946, + "step": 2178 + }, + { + "epoch": 0.6871649321980448, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 1.5554, + "step": 2179 + }, + { + "epoch": 0.6874802901292968, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 1.7363, + "step": 2180 + }, + { + "epoch": 0.6877956480605487, + "grad_norm": 0.380859375, + "learning_rate": 0.0002, + "loss": 1.567, + "step": 2181 + }, + { + "epoch": 0.6881110059918006, + "grad_norm": 0.40625, + "learning_rate": 0.0002, + "loss": 1.5473, + "step": 2182 + }, + { + "epoch": 0.6884263639230527, + "grad_norm": 0.453125, + "learning_rate": 0.0002, + "loss": 1.9715, + "step": 2183 + }, + { + "epoch": 0.6887417218543046, + "grad_norm": 0.421875, + "learning_rate": 0.0002, + "loss": 1.6001, + "step": 2184 + }, + { + "epoch": 0.6890570797855566, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 1.9526, + "step": 2185 + }, + { + "epoch": 0.6893724377168086, + "grad_norm": 0.62890625, + "learning_rate": 0.0002, + "loss": 2.4701, + "step": 2186 + }, + { + "epoch": 0.6896877956480606, + "grad_norm": 0.65625, + "learning_rate": 0.0002, + "loss": 2.1295, + "step": 2187 + }, + { + "epoch": 0.6900031535793125, + "grad_norm": 0.75, + "learning_rate": 0.0002, + "loss": 2.4516, + "step": 2188 + }, + { + "epoch": 0.6903185115105644, + "grad_norm": 0.62109375, + "learning_rate": 0.0002, + "loss": 1.957, + "step": 2189 + }, + { + "epoch": 0.6906338694418165, + "grad_norm": 0.7421875, + "learning_rate": 0.0002, + "loss": 2.6286, + "step": 2190 + }, + { + "epoch": 0.6909492273730684, + "grad_norm": 0.73046875, + "learning_rate": 0.0002, + "loss": 2.302, + "step": 2191 + }, + { + "epoch": 0.6912645853043204, + "grad_norm": 0.66796875, + "learning_rate": 0.0002, + "loss": 2.4644, + "step": 2192 + }, + { + "epoch": 0.6915799432355724, + "grad_norm": 0.91796875, + "learning_rate": 0.0002, + "loss": 2.5786, + "step": 2193 + }, + { + "epoch": 0.6918953011668243, + "grad_norm": 0.75, + "learning_rate": 0.0002, + "loss": 2.1743, + "step": 2194 + }, + { + "epoch": 0.6922106590980763, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 2.26, + "step": 2195 + }, + { + "epoch": 0.6925260170293283, + "grad_norm": 0.72265625, + "learning_rate": 0.0002, + "loss": 2.126, + "step": 2196 + }, + { + "epoch": 0.6928413749605803, + "grad_norm": 0.67578125, + "learning_rate": 0.0002, + "loss": 1.8881, + "step": 2197 + }, + { + "epoch": 0.6931567328918322, + "grad_norm": 0.8515625, + "learning_rate": 0.0002, + "loss": 2.3897, + "step": 2198 + }, + { + "epoch": 0.6934720908230843, + "grad_norm": 0.80859375, + "learning_rate": 0.0002, + "loss": 2.5024, + "step": 2199 + }, + { + "epoch": 0.6937874487543362, + "grad_norm": 0.9140625, + "learning_rate": 0.0002, + "loss": 2.6745, + "step": 2200 + }, + { + "epoch": 0.6941028066855881, + "grad_norm": 0.142578125, + "learning_rate": 0.0002, + "loss": 1.0635, + "step": 2201 + }, + { + "epoch": 0.6944181646168401, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.0519, + "step": 2202 + }, + { + "epoch": 0.6947335225480921, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.025, + "step": 2203 + }, + { + "epoch": 0.6950488804793441, + "grad_norm": 0.154296875, + "learning_rate": 0.0002, + "loss": 1.0383, + "step": 2204 + }, + { + "epoch": 0.695364238410596, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.1018, + "step": 2205 + }, + { + "epoch": 0.695679596341848, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.1374, + "step": 2206 + }, + { + "epoch": 0.6959949542731, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.1281, + "step": 2207 + }, + { + "epoch": 0.6963103122043519, + "grad_norm": 0.1767578125, + "learning_rate": 0.0002, + "loss": 1.1078, + "step": 2208 + }, + { + "epoch": 0.696625670135604, + "grad_norm": 0.1787109375, + "learning_rate": 0.0002, + "loss": 1.1395, + "step": 2209 + }, + { + "epoch": 0.6969410280668559, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 1.0748, + "step": 2210 + }, + { + "epoch": 0.6972563859981078, + "grad_norm": 0.19921875, + "learning_rate": 0.0002, + "loss": 1.2449, + "step": 2211 + }, + { + "epoch": 0.6975717439293598, + "grad_norm": 0.220703125, + "learning_rate": 0.0002, + "loss": 1.2006, + "step": 2212 + }, + { + "epoch": 0.6978871018606118, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.2345, + "step": 2213 + }, + { + "epoch": 0.6982024597918638, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.1866, + "step": 2214 + }, + { + "epoch": 0.6985178177231157, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.2816, + "step": 2215 + }, + { + "epoch": 0.6988331756543678, + "grad_norm": 0.2138671875, + "learning_rate": 0.0002, + "loss": 1.2749, + "step": 2216 + }, + { + "epoch": 0.6991485335856197, + "grad_norm": 0.205078125, + "learning_rate": 0.0002, + "loss": 1.2188, + "step": 2217 + }, + { + "epoch": 0.6994638915168716, + "grad_norm": 0.248046875, + "learning_rate": 0.0002, + "loss": 1.3671, + "step": 2218 + }, + { + "epoch": 0.6997792494481236, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.1885, + "step": 2219 + }, + { + "epoch": 0.7000946073793756, + "grad_norm": 0.2490234375, + "learning_rate": 0.0002, + "loss": 1.3678, + "step": 2220 + }, + { + "epoch": 0.7004099653106276, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 2221 + }, + { + "epoch": 0.7007253232418795, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.3601, + "step": 2222 + }, + { + "epoch": 0.7010406811731315, + "grad_norm": 0.255859375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 2223 + }, + { + "epoch": 0.7013560391043835, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.3505, + "step": 2224 + }, + { + "epoch": 0.7016713970356354, + "grad_norm": 0.330078125, + "learning_rate": 0.0002, + "loss": 1.4835, + "step": 2225 + }, + { + "epoch": 0.7019867549668874, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 1.6749, + "step": 2226 + }, + { + "epoch": 0.7023021128981394, + "grad_norm": 0.330078125, + "learning_rate": 0.0002, + "loss": 1.6617, + "step": 2227 + }, + { + "epoch": 0.7026174708293914, + "grad_norm": 0.369140625, + "learning_rate": 0.0002, + "loss": 1.4087, + "step": 2228 + }, + { + "epoch": 0.7029328287606433, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 1.4045, + "step": 2229 + }, + { + "epoch": 0.7032481866918953, + "grad_norm": 0.380859375, + "learning_rate": 0.0002, + "loss": 1.7496, + "step": 2230 + }, + { + "epoch": 0.7035635446231473, + "grad_norm": 0.416015625, + "learning_rate": 0.0002, + "loss": 1.6548, + "step": 2231 + }, + { + "epoch": 0.7038789025543992, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 2.0735, + "step": 2232 + }, + { + "epoch": 0.7041942604856513, + "grad_norm": 0.416015625, + "learning_rate": 0.0002, + "loss": 1.5881, + "step": 2233 + }, + { + "epoch": 0.7045096184169032, + "grad_norm": 0.435546875, + "learning_rate": 0.0002, + "loss": 1.5738, + "step": 2234 + }, + { + "epoch": 0.7048249763481551, + "grad_norm": 0.46484375, + "learning_rate": 0.0002, + "loss": 1.7005, + "step": 2235 + }, + { + "epoch": 0.7051403342794071, + "grad_norm": 0.515625, + "learning_rate": 0.0002, + "loss": 1.616, + "step": 2236 + }, + { + "epoch": 0.7054556922106591, + "grad_norm": 0.703125, + "learning_rate": 0.0002, + "loss": 2.2448, + "step": 2237 + }, + { + "epoch": 0.7057710501419111, + "grad_norm": 0.6484375, + "learning_rate": 0.0002, + "loss": 2.4584, + "step": 2238 + }, + { + "epoch": 0.706086408073163, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 2.1733, + "step": 2239 + }, + { + "epoch": 0.7064017660044151, + "grad_norm": 0.6015625, + "learning_rate": 0.0002, + "loss": 2.2761, + "step": 2240 + }, + { + "epoch": 0.706717123935667, + "grad_norm": 0.71484375, + "learning_rate": 0.0002, + "loss": 2.4879, + "step": 2241 + }, + { + "epoch": 0.7070324818669189, + "grad_norm": 0.6796875, + "learning_rate": 0.0002, + "loss": 2.2553, + "step": 2242 + }, + { + "epoch": 0.707347839798171, + "grad_norm": 0.88671875, + "learning_rate": 0.0002, + "loss": 2.7137, + "step": 2243 + }, + { + "epoch": 0.7076631977294229, + "grad_norm": 0.81640625, + "learning_rate": 0.0002, + "loss": 2.4625, + "step": 2244 + }, + { + "epoch": 0.7076631977294229, + "eval_loss": 1.6584293842315674, + "eval_runtime": 148.7366, + "eval_samples_per_second": 6.723, + "eval_steps_per_second": 6.723, + "step": 2244 + }, + { + "epoch": 0.7076631977294229, + "mmlu_eval_accuracy": 0.6196221764912175, + "mmlu_eval_accuracy_abstract_algebra": 0.09090909090909091, + "mmlu_eval_accuracy_anatomy": 0.6428571428571429, + "mmlu_eval_accuracy_astronomy": 0.6875, + "mmlu_eval_accuracy_business_ethics": 0.45454545454545453, + "mmlu_eval_accuracy_clinical_knowledge": 0.6896551724137931, + "mmlu_eval_accuracy_college_biology": 0.75, + "mmlu_eval_accuracy_college_chemistry": 0.5, + "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, + "mmlu_eval_accuracy_college_mathematics": 0.6363636363636364, + "mmlu_eval_accuracy_college_medicine": 0.6363636363636364, + "mmlu_eval_accuracy_college_physics": 0.6363636363636364, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.6153846153846154, + "mmlu_eval_accuracy_econometrics": 0.4166666666666667, + "mmlu_eval_accuracy_electrical_engineering": 0.4375, + "mmlu_eval_accuracy_elementary_mathematics": 0.4878048780487805, + "mmlu_eval_accuracy_formal_logic": 0.42857142857142855, + "mmlu_eval_accuracy_global_facts": 0.2, + "mmlu_eval_accuracy_high_school_biology": 0.59375, + "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453, + "mmlu_eval_accuracy_high_school_computer_science": 0.7777777777777778, + "mmlu_eval_accuracy_high_school_european_history": 0.8888888888888888, + "mmlu_eval_accuracy_high_school_geography": 0.9545454545454546, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209, + "mmlu_eval_accuracy_high_school_mathematics": 0.3793103448275862, + "mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923, + "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354, + "mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667, + "mmlu_eval_accuracy_high_school_statistics": 0.4782608695652174, + "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727, + "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, + "mmlu_eval_accuracy_human_aging": 0.6086956521739131, + "mmlu_eval_accuracy_human_sexuality": 0.6666666666666666, + "mmlu_eval_accuracy_international_law": 0.8461538461538461, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.7777777777777778, + "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.88, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7558139534883721, + "mmlu_eval_accuracy_moral_disputes": 0.5, + "mmlu_eval_accuracy_moral_scenarios": 0.28, + "mmlu_eval_accuracy_nutrition": 0.696969696969697, + "mmlu_eval_accuracy_philosophy": 0.7352941176470589, + "mmlu_eval_accuracy_prehistory": 0.6, + "mmlu_eval_accuracy_professional_accounting": 0.3870967741935484, + "mmlu_eval_accuracy_professional_law": 0.4, + "mmlu_eval_accuracy_professional_medicine": 0.7419354838709677, + "mmlu_eval_accuracy_professional_psychology": 0.6521739130434783, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7777777777777778, + "mmlu_eval_accuracy_sociology": 0.9545454545454546, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.5555555555555556, + "mmlu_eval_accuracy_world_religions": 0.8947368421052632, + "mmlu_loss": 5.918850970065804, + "step": 2244 + }, + { + "epoch": 0.7079785556606749, + "grad_norm": 0.9609375, + "learning_rate": 0.0002, + "loss": 2.5099, + "step": 2245 + }, + { + "epoch": 0.7082939135919268, + "grad_norm": 0.84765625, + "learning_rate": 0.0002, + "loss": 2.1803, + "step": 2246 + }, + { + "epoch": 0.7086092715231788, + "grad_norm": 0.67578125, + "learning_rate": 0.0002, + "loss": 2.2514, + "step": 2247 + }, + { + "epoch": 0.7089246294544308, + "grad_norm": 0.7421875, + "learning_rate": 0.0002, + "loss": 2.1884, + "step": 2248 + }, + { + "epoch": 0.7092399873856827, + "grad_norm": 0.86328125, + "learning_rate": 0.0002, + "loss": 2.4679, + "step": 2249 + }, + { + "epoch": 0.7095553453169348, + "grad_norm": 1.109375, + "learning_rate": 0.0002, + "loss": 2.4068, + "step": 2250 + }, + { + "epoch": 0.7098707032481867, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 0.9175, + "step": 2251 + }, + { + "epoch": 0.7101860611794386, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.031, + "step": 2252 + }, + { + "epoch": 0.7105014191106906, + "grad_norm": 0.169921875, + "learning_rate": 0.0002, + "loss": 1.0981, + "step": 2253 + }, + { + "epoch": 0.7108167770419426, + "grad_norm": 0.166015625, + "learning_rate": 0.0002, + "loss": 1.1611, + "step": 2254 + }, + { + "epoch": 0.7111321349731946, + "grad_norm": 0.1640625, + "learning_rate": 0.0002, + "loss": 1.0597, + "step": 2255 + }, + { + "epoch": 0.7114474929044465, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 0.9787, + "step": 2256 + }, + { + "epoch": 0.7117628508356986, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.033, + "step": 2257 + }, + { + "epoch": 0.7120782087669505, + "grad_norm": 0.189453125, + "learning_rate": 0.0002, + "loss": 1.1799, + "step": 2258 + }, + { + "epoch": 0.7123935666982024, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.0086, + "step": 2259 + }, + { + "epoch": 0.7127089246294545, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 1.045, + "step": 2260 + }, + { + "epoch": 0.7130242825607064, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.0753, + "step": 2261 + }, + { + "epoch": 0.7133396404919584, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 0.9893, + "step": 2262 + }, + { + "epoch": 0.7136549984232103, + "grad_norm": 0.208984375, + "learning_rate": 0.0002, + "loss": 1.2705, + "step": 2263 + }, + { + "epoch": 0.7139703563544623, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 1.0788, + "step": 2264 + }, + { + "epoch": 0.7142857142857143, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.2451, + "step": 2265 + }, + { + "epoch": 0.7146010722169662, + "grad_norm": 0.220703125, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 2266 + }, + { + "epoch": 0.7149164301482183, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.142, + "step": 2267 + }, + { + "epoch": 0.7152317880794702, + "grad_norm": 0.2275390625, + "learning_rate": 0.0002, + "loss": 1.4338, + "step": 2268 + }, + { + "epoch": 0.7155471460107222, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 1.4273, + "step": 2269 + }, + { + "epoch": 0.7158625039419741, + "grad_norm": 0.255859375, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 2270 + }, + { + "epoch": 0.7161778618732261, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 1.2906, + "step": 2271 + }, + { + "epoch": 0.7164932198044781, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 2272 + }, + { + "epoch": 0.71680857773573, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.2166, + "step": 2273 + }, + { + "epoch": 0.7171239356669821, + "grad_norm": 0.255859375, + "learning_rate": 0.0002, + "loss": 1.3002, + "step": 2274 + }, + { + "epoch": 0.717439293598234, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.6173, + "step": 2275 + }, + { + "epoch": 0.7177546515294859, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 1.4244, + "step": 2276 + }, + { + "epoch": 0.718070009460738, + "grad_norm": 0.38671875, + "learning_rate": 0.0002, + "loss": 1.6029, + "step": 2277 + }, + { + "epoch": 0.7183853673919899, + "grad_norm": 0.34765625, + "learning_rate": 0.0002, + "loss": 1.3813, + "step": 2278 + }, + { + "epoch": 0.7187007253232419, + "grad_norm": 0.369140625, + "learning_rate": 0.0002, + "loss": 1.6347, + "step": 2279 + }, + { + "epoch": 0.7190160832544938, + "grad_norm": 0.353515625, + "learning_rate": 0.0002, + "loss": 1.4817, + "step": 2280 + }, + { + "epoch": 0.7193314411857458, + "grad_norm": 0.400390625, + "learning_rate": 0.0002, + "loss": 1.5689, + "step": 2281 + }, + { + "epoch": 0.7196467991169978, + "grad_norm": 0.4453125, + "learning_rate": 0.0002, + "loss": 1.8533, + "step": 2282 + }, + { + "epoch": 0.7199621570482497, + "grad_norm": 0.546875, + "learning_rate": 0.0002, + "loss": 2.0288, + "step": 2283 + }, + { + "epoch": 0.7202775149795018, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 2.0307, + "step": 2284 + }, + { + "epoch": 0.7205928729107537, + "grad_norm": 0.494140625, + "learning_rate": 0.0002, + "loss": 2.0386, + "step": 2285 + }, + { + "epoch": 0.7209082308420057, + "grad_norm": 0.60546875, + "learning_rate": 0.0002, + "loss": 1.9619, + "step": 2286 + }, + { + "epoch": 0.7212235887732577, + "grad_norm": 0.65625, + "learning_rate": 0.0002, + "loss": 1.9426, + "step": 2287 + }, + { + "epoch": 0.7215389467045096, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 2.0074, + "step": 2288 + }, + { + "epoch": 0.7218543046357616, + "grad_norm": 0.6484375, + "learning_rate": 0.0002, + "loss": 1.9221, + "step": 2289 + }, + { + "epoch": 0.7221696625670135, + "grad_norm": 0.56640625, + "learning_rate": 0.0002, + "loss": 1.9942, + "step": 2290 + }, + { + "epoch": 0.7224850204982656, + "grad_norm": 0.67578125, + "learning_rate": 0.0002, + "loss": 2.1429, + "step": 2291 + }, + { + "epoch": 0.7228003784295175, + "grad_norm": 0.609375, + "learning_rate": 0.0002, + "loss": 2.1159, + "step": 2292 + }, + { + "epoch": 0.7231157363607694, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 2.2236, + "step": 2293 + }, + { + "epoch": 0.7234310942920215, + "grad_norm": 0.84375, + "learning_rate": 0.0002, + "loss": 2.7805, + "step": 2294 + }, + { + "epoch": 0.7237464522232734, + "grad_norm": 0.6640625, + "learning_rate": 0.0002, + "loss": 2.3309, + "step": 2295 + }, + { + "epoch": 0.7240618101545254, + "grad_norm": 0.7109375, + "learning_rate": 0.0002, + "loss": 2.3308, + "step": 2296 + }, + { + "epoch": 0.7243771680857773, + "grad_norm": 0.91015625, + "learning_rate": 0.0002, + "loss": 2.401, + "step": 2297 + }, + { + "epoch": 0.7246925260170294, + "grad_norm": 0.98046875, + "learning_rate": 0.0002, + "loss": 3.0552, + "step": 2298 + }, + { + "epoch": 0.7250078839482813, + "grad_norm": 0.859375, + "learning_rate": 0.0002, + "loss": 2.1556, + "step": 2299 + }, + { + "epoch": 0.7253232418795332, + "grad_norm": 0.96484375, + "learning_rate": 0.0002, + "loss": 2.5763, + "step": 2300 + }, + { + "epoch": 0.7256385998107853, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.1506, + "step": 2301 + }, + { + "epoch": 0.7259539577420372, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.0655, + "step": 2302 + }, + { + "epoch": 0.7262693156732892, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 0.9869, + "step": 2303 + }, + { + "epoch": 0.7265846736045412, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.0582, + "step": 2304 + }, + { + "epoch": 0.7269000315357931, + "grad_norm": 0.16015625, + "learning_rate": 0.0002, + "loss": 1.2911, + "step": 2305 + }, + { + "epoch": 0.7272153894670451, + "grad_norm": 0.154296875, + "learning_rate": 0.0002, + "loss": 1.0662, + "step": 2306 + }, + { + "epoch": 0.727530747398297, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 1.18, + "step": 2307 + }, + { + "epoch": 0.7278461053295491, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.1408, + "step": 2308 + }, + { + "epoch": 0.728161463260801, + "grad_norm": 0.169921875, + "learning_rate": 0.0002, + "loss": 1.1356, + "step": 2309 + }, + { + "epoch": 0.7284768211920529, + "grad_norm": 0.1728515625, + "learning_rate": 0.0002, + "loss": 1.2075, + "step": 2310 + }, + { + "epoch": 0.728792179123305, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.0911, + "step": 2311 + }, + { + "epoch": 0.7291075370545569, + "grad_norm": 0.1796875, + "learning_rate": 0.0002, + "loss": 1.1514, + "step": 2312 + }, + { + "epoch": 0.7294228949858089, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 1.5668, + "step": 2313 + }, + { + "epoch": 0.7297382529170608, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.2313, + "step": 2314 + }, + { + "epoch": 0.7300536108483129, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.1403, + "step": 2315 + }, + { + "epoch": 0.7303689687795648, + "grad_norm": 0.208984375, + "learning_rate": 0.0002, + "loss": 1.2116, + "step": 2316 + }, + { + "epoch": 0.7306843267108167, + "grad_norm": 0.240234375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 2317 + }, + { + "epoch": 0.7309996846420688, + "grad_norm": 0.255859375, + "learning_rate": 0.0002, + "loss": 1.472, + "step": 2318 + }, + { + "epoch": 0.7313150425733207, + "grad_norm": 0.2275390625, + "learning_rate": 0.0002, + "loss": 1.2176, + "step": 2319 + }, + { + "epoch": 0.7316304005045727, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 1.4217, + "step": 2320 + }, + { + "epoch": 0.7319457584358247, + "grad_norm": 0.255859375, + "learning_rate": 0.0002, + "loss": 1.3985, + "step": 2321 + }, + { + "epoch": 0.7322611163670766, + "grad_norm": 0.2451171875, + "learning_rate": 0.0002, + "loss": 1.4537, + "step": 2322 + }, + { + "epoch": 0.7325764742983286, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 1.2652, + "step": 2323 + }, + { + "epoch": 0.7328918322295805, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 1.5209, + "step": 2324 + }, + { + "epoch": 0.7332071901608326, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.3817, + "step": 2325 + }, + { + "epoch": 0.7335225480920845, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.4293, + "step": 2326 + }, + { + "epoch": 0.7338379060233365, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 1.4475, + "step": 2327 + }, + { + "epoch": 0.7341532639545885, + "grad_norm": 0.40234375, + "learning_rate": 0.0002, + "loss": 1.439, + "step": 2328 + }, + { + "epoch": 0.7344686218858404, + "grad_norm": 0.498046875, + "learning_rate": 0.0002, + "loss": 1.593, + "step": 2329 + }, + { + "epoch": 0.7347839798170924, + "grad_norm": 0.46484375, + "learning_rate": 0.0002, + "loss": 1.7826, + "step": 2330 + }, + { + "epoch": 0.7350993377483444, + "grad_norm": 0.41015625, + "learning_rate": 0.0002, + "loss": 1.4615, + "step": 2331 + }, + { + "epoch": 0.7354146956795964, + "grad_norm": 0.462890625, + "learning_rate": 0.0002, + "loss": 1.8592, + "step": 2332 + }, + { + "epoch": 0.7357300536108483, + "grad_norm": 0.4765625, + "learning_rate": 0.0002, + "loss": 1.6412, + "step": 2333 + }, + { + "epoch": 0.7360454115421002, + "grad_norm": 0.484375, + "learning_rate": 0.0002, + "loss": 2.0341, + "step": 2334 + }, + { + "epoch": 0.7363607694733523, + "grad_norm": 0.4140625, + "learning_rate": 0.0002, + "loss": 1.6387, + "step": 2335 + }, + { + "epoch": 0.7366761274046042, + "grad_norm": 0.498046875, + "learning_rate": 0.0002, + "loss": 1.6487, + "step": 2336 + }, + { + "epoch": 0.7369914853358562, + "grad_norm": 0.5546875, + "learning_rate": 0.0002, + "loss": 1.9093, + "step": 2337 + }, + { + "epoch": 0.7373068432671082, + "grad_norm": 0.55859375, + "learning_rate": 0.0002, + "loss": 1.8578, + "step": 2338 + }, + { + "epoch": 0.7376222011983601, + "grad_norm": 0.53515625, + "learning_rate": 0.0002, + "loss": 1.8484, + "step": 2339 + }, + { + "epoch": 0.7379375591296121, + "grad_norm": 1.0859375, + "learning_rate": 0.0002, + "loss": 2.8105, + "step": 2340 + }, + { + "epoch": 0.738252917060864, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 1.958, + "step": 2341 + }, + { + "epoch": 0.7385682749921161, + "grad_norm": 0.6796875, + "learning_rate": 0.0002, + "loss": 2.4239, + "step": 2342 + }, + { + "epoch": 0.738883632923368, + "grad_norm": 0.6875, + "learning_rate": 0.0002, + "loss": 2.3226, + "step": 2343 + }, + { + "epoch": 0.73919899085462, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 2.015, + "step": 2344 + }, + { + "epoch": 0.739514348785872, + "grad_norm": 0.82421875, + "learning_rate": 0.0002, + "loss": 2.6535, + "step": 2345 + }, + { + "epoch": 0.7398297067171239, + "grad_norm": 0.625, + "learning_rate": 0.0002, + "loss": 1.8172, + "step": 2346 + }, + { + "epoch": 0.7401450646483759, + "grad_norm": 1.15625, + "learning_rate": 0.0002, + "loss": 2.8481, + "step": 2347 + }, + { + "epoch": 0.7404604225796279, + "grad_norm": 0.62109375, + "learning_rate": 0.0002, + "loss": 2.082, + "step": 2348 + }, + { + "epoch": 0.7407757805108799, + "grad_norm": 0.71484375, + "learning_rate": 0.0002, + "loss": 1.9481, + "step": 2349 + }, + { + "epoch": 0.7410911384421318, + "grad_norm": 1.0, + "learning_rate": 0.0002, + "loss": 2.868, + "step": 2350 + }, + { + "epoch": 0.7414064963733837, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.0133, + "step": 2351 + }, + { + "epoch": 0.7417218543046358, + "grad_norm": 0.140625, + "learning_rate": 0.0002, + "loss": 1.101, + "step": 2352 + }, + { + "epoch": 0.7420372122358877, + "grad_norm": 0.193359375, + "learning_rate": 0.0002, + "loss": 1.1722, + "step": 2353 + }, + { + "epoch": 0.7423525701671397, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.0873, + "step": 2354 + }, + { + "epoch": 0.7426679280983917, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.1675, + "step": 2355 + }, + { + "epoch": 0.7429832860296437, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.1388, + "step": 2356 + }, + { + "epoch": 0.7432986439608956, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 1.1368, + "step": 2357 + }, + { + "epoch": 0.7436140018921475, + "grad_norm": 0.1875, + "learning_rate": 0.0002, + "loss": 1.0733, + "step": 2358 + }, + { + "epoch": 0.7439293598233996, + "grad_norm": 0.1787109375, + "learning_rate": 0.0002, + "loss": 1.1256, + "step": 2359 + }, + { + "epoch": 0.7442447177546515, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.1148, + "step": 2360 + }, + { + "epoch": 0.7445600756859035, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.2142, + "step": 2361 + }, + { + "epoch": 0.7448754336171555, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.1064, + "step": 2362 + }, + { + "epoch": 0.7451907915484074, + "grad_norm": 0.201171875, + "learning_rate": 0.0002, + "loss": 1.1356, + "step": 2363 + }, + { + "epoch": 0.7455061494796594, + "grad_norm": 0.1884765625, + "learning_rate": 0.0002, + "loss": 1.0928, + "step": 2364 + }, + { + "epoch": 0.7458215074109114, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 1.1458, + "step": 2365 + }, + { + "epoch": 0.7461368653421634, + "grad_norm": 0.2060546875, + "learning_rate": 0.0002, + "loss": 1.1887, + "step": 2366 + }, + { + "epoch": 0.7464522232734153, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.1748, + "step": 2367 + }, + { + "epoch": 0.7467675812046674, + "grad_norm": 0.2119140625, + "learning_rate": 0.0002, + "loss": 1.2467, + "step": 2368 + }, + { + "epoch": 0.7470829391359193, + "grad_norm": 0.2314453125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 2369 + }, + { + "epoch": 0.7473982970671712, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.179, + "step": 2370 + }, + { + "epoch": 0.7477136549984232, + "grad_norm": 0.2421875, + "learning_rate": 0.0002, + "loss": 1.2221, + "step": 2371 + }, + { + "epoch": 0.7480290129296752, + "grad_norm": 0.248046875, + "learning_rate": 0.0002, + "loss": 1.2369, + "step": 2372 + }, + { + "epoch": 0.7483443708609272, + "grad_norm": 0.2275390625, + "learning_rate": 0.0002, + "loss": 1.1038, + "step": 2373 + }, + { + "epoch": 0.7486597287921791, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 1.2797, + "step": 2374 + }, + { + "epoch": 0.748975086723431, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 1.4203, + "step": 2375 + }, + { + "epoch": 0.7492904446546831, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 1.5561, + "step": 2376 + }, + { + "epoch": 0.749605802585935, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 1.4168, + "step": 2377 + }, + { + "epoch": 0.749921160517187, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 1.1675, + "step": 2378 + }, + { + "epoch": 0.750236518448439, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 1.231, + "step": 2379 + }, + { + "epoch": 0.7505518763796909, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 1.4097, + "step": 2380 + }, + { + "epoch": 0.7508672343109429, + "grad_norm": 0.3828125, + "learning_rate": 0.0002, + "loss": 1.6152, + "step": 2381 + }, + { + "epoch": 0.7511825922421949, + "grad_norm": 0.62890625, + "learning_rate": 0.0002, + "loss": 2.1051, + "step": 2382 + }, + { + "epoch": 0.7514979501734469, + "grad_norm": 0.455078125, + "learning_rate": 0.0002, + "loss": 1.6616, + "step": 2383 + }, + { + "epoch": 0.7518133081046988, + "grad_norm": 0.51953125, + "learning_rate": 0.0002, + "loss": 1.9241, + "step": 2384 + }, + { + "epoch": 0.7521286660359509, + "grad_norm": 0.58984375, + "learning_rate": 0.0002, + "loss": 2.0805, + "step": 2385 + }, + { + "epoch": 0.7524440239672028, + "grad_norm": 0.75390625, + "learning_rate": 0.0002, + "loss": 1.963, + "step": 2386 + }, + { + "epoch": 0.7527593818984547, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 1.9717, + "step": 2387 + }, + { + "epoch": 0.7530747398297067, + "grad_norm": 0.6875, + "learning_rate": 0.0002, + "loss": 2.364, + "step": 2388 + }, + { + "epoch": 0.7533900977609587, + "grad_norm": 0.68359375, + "learning_rate": 0.0002, + "loss": 2.5844, + "step": 2389 + }, + { + "epoch": 0.7537054556922107, + "grad_norm": 0.66796875, + "learning_rate": 0.0002, + "loss": 2.0953, + "step": 2390 + }, + { + "epoch": 0.7540208136234626, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 2.2897, + "step": 2391 + }, + { + "epoch": 0.7543361715547146, + "grad_norm": 0.7734375, + "learning_rate": 0.0002, + "loss": 2.408, + "step": 2392 + }, + { + "epoch": 0.7546515294859666, + "grad_norm": 0.7265625, + "learning_rate": 0.0002, + "loss": 2.3527, + "step": 2393 + }, + { + "epoch": 0.7549668874172185, + "grad_norm": 1.2734375, + "learning_rate": 0.0002, + "loss": 2.355, + "step": 2394 + }, + { + "epoch": 0.7552822453484705, + "grad_norm": 0.6640625, + "learning_rate": 0.0002, + "loss": 2.084, + "step": 2395 + }, + { + "epoch": 0.7555976032797225, + "grad_norm": 0.85546875, + "learning_rate": 0.0002, + "loss": 2.4546, + "step": 2396 + }, + { + "epoch": 0.7559129612109745, + "grad_norm": 0.9140625, + "learning_rate": 0.0002, + "loss": 2.6237, + "step": 2397 + }, + { + "epoch": 0.7562283191422264, + "grad_norm": 0.7578125, + "learning_rate": 0.0002, + "loss": 2.1944, + "step": 2398 + }, + { + "epoch": 0.7565436770734784, + "grad_norm": 0.953125, + "learning_rate": 0.0002, + "loss": 2.9971, + "step": 2399 + }, + { + "epoch": 0.7568590350047304, + "grad_norm": 0.7109375, + "learning_rate": 0.0002, + "loss": 2.0175, + "step": 2400 + }, + { + "epoch": 0.7571743929359823, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.0158, + "step": 2401 + }, + { + "epoch": 0.7574897508672344, + "grad_norm": 0.158203125, + "learning_rate": 0.0002, + "loss": 1.0823, + "step": 2402 + }, + { + "epoch": 0.7578051087984863, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.2644, + "step": 2403 + }, + { + "epoch": 0.7581204667297382, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.0344, + "step": 2404 + }, + { + "epoch": 0.7584358246609902, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.0655, + "step": 2405 + }, + { + "epoch": 0.7587511825922422, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.0761, + "step": 2406 + }, + { + "epoch": 0.7590665405234942, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.0328, + "step": 2407 + }, + { + "epoch": 0.7593818984547461, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 1.0721, + "step": 2408 + }, + { + "epoch": 0.759697256385998, + "grad_norm": 0.169921875, + "learning_rate": 0.0002, + "loss": 1.2265, + "step": 2409 + }, + { + "epoch": 0.7600126143172501, + "grad_norm": 0.212890625, + "learning_rate": 0.0002, + "loss": 1.1525, + "step": 2410 + }, + { + "epoch": 0.760327972248502, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.1471, + "step": 2411 + }, + { + "epoch": 0.760643330179754, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 1.0304, + "step": 2412 + }, + { + "epoch": 0.760958688111006, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.115, + "step": 2413 + }, + { + "epoch": 0.761274046042258, + "grad_norm": 0.18359375, + "learning_rate": 0.0002, + "loss": 1.246, + "step": 2414 + }, + { + "epoch": 0.7615894039735099, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 1.0757, + "step": 2415 + }, + { + "epoch": 0.7619047619047619, + "grad_norm": 0.197265625, + "learning_rate": 0.0002, + "loss": 1.1863, + "step": 2416 + }, + { + "epoch": 0.7622201198360139, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.1691, + "step": 2417 + }, + { + "epoch": 0.7625354777672658, + "grad_norm": 0.22265625, + "learning_rate": 0.0002, + "loss": 1.2293, + "step": 2418 + }, + { + "epoch": 0.7628508356985179, + "grad_norm": 0.244140625, + "learning_rate": 0.0002, + "loss": 1.203, + "step": 2419 + }, + { + "epoch": 0.7631661936297698, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.2441, + "step": 2420 + }, + { + "epoch": 0.7634815515610217, + "grad_norm": 0.2490234375, + "learning_rate": 0.0002, + "loss": 1.266, + "step": 2421 + }, + { + "epoch": 0.7637969094922737, + "grad_norm": 0.2470703125, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 2422 + }, + { + "epoch": 0.7641122674235257, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 1.6129, + "step": 2423 + }, + { + "epoch": 0.7644276253547777, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 1.4377, + "step": 2424 + }, + { + "epoch": 0.7647429832860296, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 2425 + }, + { + "epoch": 0.7650583412172817, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 2426 + }, + { + "epoch": 0.7653736991485336, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 1.3562, + "step": 2427 + }, + { + "epoch": 0.7656890570797855, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.4668, + "step": 2428 + }, + { + "epoch": 0.7660044150110376, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 2429 + }, + { + "epoch": 0.7663197729422895, + "grad_norm": 0.416015625, + "learning_rate": 0.0002, + "loss": 1.5911, + "step": 2430 + }, + { + "epoch": 0.7666351308735415, + "grad_norm": 0.404296875, + "learning_rate": 0.0002, + "loss": 1.7427, + "step": 2431 + }, + { + "epoch": 0.7666351308735415, + "eval_loss": 1.6458227634429932, + "eval_runtime": 148.9378, + "eval_samples_per_second": 6.714, + "eval_steps_per_second": 6.714, + "step": 2431 + }, + { + "epoch": 0.7666351308735415, + "mmlu_eval_accuracy": 0.6202766164701206, + "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182, + "mmlu_eval_accuracy_anatomy": 0.6428571428571429, + "mmlu_eval_accuracy_astronomy": 0.75, + "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, + "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034, + "mmlu_eval_accuracy_college_biology": 0.8125, + "mmlu_eval_accuracy_college_chemistry": 0.5, + "mmlu_eval_accuracy_college_computer_science": 0.5454545454545454, + "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727, + "mmlu_eval_accuracy_college_medicine": 0.6818181818181818, + "mmlu_eval_accuracy_college_physics": 0.6363636363636364, + "mmlu_eval_accuracy_computer_security": 0.5454545454545454, + "mmlu_eval_accuracy_conceptual_physics": 0.5, + "mmlu_eval_accuracy_econometrics": 0.5, + "mmlu_eval_accuracy_electrical_engineering": 0.5, + "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244, + "mmlu_eval_accuracy_formal_logic": 0.35714285714285715, + "mmlu_eval_accuracy_global_facts": 0.4, + "mmlu_eval_accuracy_high_school_biology": 0.65625, + "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8571428571428571, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442, + "mmlu_eval_accuracy_high_school_mathematics": 0.3448275862068966, + "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769, + "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354, + "mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667, + "mmlu_eval_accuracy_high_school_statistics": 0.5217391304347826, + "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, + "mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693, + "mmlu_eval_accuracy_human_aging": 0.6521739130434783, + "mmlu_eval_accuracy_human_sexuality": 0.75, + "mmlu_eval_accuracy_international_law": 0.8461538461538461, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.84, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7093023255813954, + "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527, + "mmlu_eval_accuracy_moral_scenarios": 0.36, + "mmlu_eval_accuracy_nutrition": 0.7272727272727273, + "mmlu_eval_accuracy_philosophy": 0.7352941176470589, + "mmlu_eval_accuracy_prehistory": 0.5714285714285714, + "mmlu_eval_accuracy_professional_accounting": 0.2903225806451613, + "mmlu_eval_accuracy_professional_law": 0.4, + "mmlu_eval_accuracy_professional_medicine": 0.7741935483870968, + "mmlu_eval_accuracy_professional_psychology": 0.6376811594202898, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7777777777777778, + "mmlu_eval_accuracy_sociology": 1.0, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.5, + "mmlu_eval_accuracy_world_religions": 0.8947368421052632, + "mmlu_loss": 6.140492279805209, + "step": 2431 + }, + { + "epoch": 0.7669504888047934, + "grad_norm": 0.3984375, + "learning_rate": 0.0002, + "loss": 1.7463, + "step": 2432 + }, + { + "epoch": 0.7672658467360454, + "grad_norm": 0.40625, + "learning_rate": 0.0002, + "loss": 1.6551, + "step": 2433 + }, + { + "epoch": 0.7675812046672974, + "grad_norm": 0.48046875, + "learning_rate": 0.0002, + "loss": 1.7776, + "step": 2434 + }, + { + "epoch": 0.7678965625985493, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 2.1326, + "step": 2435 + }, + { + "epoch": 0.7682119205298014, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 1.8893, + "step": 2436 + }, + { + "epoch": 0.7685272784610533, + "grad_norm": 0.578125, + "learning_rate": 0.0002, + "loss": 2.0332, + "step": 2437 + }, + { + "epoch": 0.7688426363923052, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 2.0664, + "step": 2438 + }, + { + "epoch": 0.7691579943235572, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 2.3474, + "step": 2439 + }, + { + "epoch": 0.7694733522548092, + "grad_norm": 0.7109375, + "learning_rate": 0.0002, + "loss": 2.4222, + "step": 2440 + }, + { + "epoch": 0.7697887101860612, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 1.9639, + "step": 2441 + }, + { + "epoch": 0.7701040681173131, + "grad_norm": 0.7578125, + "learning_rate": 0.0002, + "loss": 2.307, + "step": 2442 + }, + { + "epoch": 0.7704194260485652, + "grad_norm": 0.7265625, + "learning_rate": 0.0002, + "loss": 2.3974, + "step": 2443 + }, + { + "epoch": 0.7707347839798171, + "grad_norm": 0.7890625, + "learning_rate": 0.0002, + "loss": 2.6291, + "step": 2444 + }, + { + "epoch": 0.771050141911069, + "grad_norm": 0.9609375, + "learning_rate": 0.0002, + "loss": 2.8769, + "step": 2445 + }, + { + "epoch": 0.7713654998423211, + "grad_norm": 0.73046875, + "learning_rate": 0.0002, + "loss": 2.5504, + "step": 2446 + }, + { + "epoch": 0.771680857773573, + "grad_norm": 1.140625, + "learning_rate": 0.0002, + "loss": 3.0071, + "step": 2447 + }, + { + "epoch": 0.771996215704825, + "grad_norm": 0.8828125, + "learning_rate": 0.0002, + "loss": 2.3819, + "step": 2448 + }, + { + "epoch": 0.7723115736360769, + "grad_norm": 0.7890625, + "learning_rate": 0.0002, + "loss": 2.3378, + "step": 2449 + }, + { + "epoch": 0.7726269315673289, + "grad_norm": 0.91015625, + "learning_rate": 0.0002, + "loss": 2.6399, + "step": 2450 + }, + { + "epoch": 0.7729422894985809, + "grad_norm": 0.140625, + "learning_rate": 0.0002, + "loss": 1.025, + "step": 2451 + }, + { + "epoch": 0.7732576474298328, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 0.976, + "step": 2452 + }, + { + "epoch": 0.7735730053610849, + "grad_norm": 0.154296875, + "learning_rate": 0.0002, + "loss": 1.003, + "step": 2453 + }, + { + "epoch": 0.7738883632923368, + "grad_norm": 0.1689453125, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 2454 + }, + { + "epoch": 0.7742037212235888, + "grad_norm": 0.1767578125, + "learning_rate": 0.0002, + "loss": 1.2349, + "step": 2455 + }, + { + "epoch": 0.7745190791548408, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.0683, + "step": 2456 + }, + { + "epoch": 0.7748344370860927, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.0986, + "step": 2457 + }, + { + "epoch": 0.7751497950173447, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.0735, + "step": 2458 + }, + { + "epoch": 0.7754651529485966, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.1457, + "step": 2459 + }, + { + "epoch": 0.7757805108798487, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.091, + "step": 2460 + }, + { + "epoch": 0.7760958688111006, + "grad_norm": 0.201171875, + "learning_rate": 0.0002, + "loss": 1.1252, + "step": 2461 + }, + { + "epoch": 0.7764112267423525, + "grad_norm": 0.1962890625, + "learning_rate": 0.0002, + "loss": 1.0686, + "step": 2462 + }, + { + "epoch": 0.7767265846736046, + "grad_norm": 0.197265625, + "learning_rate": 0.0002, + "loss": 1.1744, + "step": 2463 + }, + { + "epoch": 0.7770419426048565, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 2464 + }, + { + "epoch": 0.7773573005361085, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 1.2261, + "step": 2465 + }, + { + "epoch": 0.7776726584673604, + "grad_norm": 0.19140625, + "learning_rate": 0.0002, + "loss": 1.0922, + "step": 2466 + }, + { + "epoch": 0.7779880163986124, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 1.177, + "step": 2467 + }, + { + "epoch": 0.7783033743298644, + "grad_norm": 0.2392578125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 2468 + }, + { + "epoch": 0.7786187322611163, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 2469 + }, + { + "epoch": 0.7789340901923684, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 2470 + }, + { + "epoch": 0.7792494481236203, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 2471 + }, + { + "epoch": 0.7795648060548723, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 1.2969, + "step": 2472 + }, + { + "epoch": 0.7798801639861243, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 1.5209, + "step": 2473 + }, + { + "epoch": 0.7801955219173762, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 2474 + }, + { + "epoch": 0.7805108798486282, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 1.3984, + "step": 2475 + }, + { + "epoch": 0.7808262377798801, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.2485, + "step": 2476 + }, + { + "epoch": 0.7811415957111322, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.2799, + "step": 2477 + }, + { + "epoch": 0.7814569536423841, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 1.3789, + "step": 2478 + }, + { + "epoch": 0.781772311573636, + "grad_norm": 0.37109375, + "learning_rate": 0.0002, + "loss": 1.2747, + "step": 2479 + }, + { + "epoch": 0.7820876695048881, + "grad_norm": 0.400390625, + "learning_rate": 0.0002, + "loss": 1.6212, + "step": 2480 + }, + { + "epoch": 0.78240302743614, + "grad_norm": 0.48046875, + "learning_rate": 0.0002, + "loss": 1.7591, + "step": 2481 + }, + { + "epoch": 0.782718385367392, + "grad_norm": 0.45703125, + "learning_rate": 0.0002, + "loss": 1.8536, + "step": 2482 + }, + { + "epoch": 0.783033743298644, + "grad_norm": 0.46484375, + "learning_rate": 0.0002, + "loss": 1.6547, + "step": 2483 + }, + { + "epoch": 0.783349101229896, + "grad_norm": 0.50390625, + "learning_rate": 0.0002, + "loss": 1.6804, + "step": 2484 + }, + { + "epoch": 0.7836644591611479, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 1.9254, + "step": 2485 + }, + { + "epoch": 0.7839798170923998, + "grad_norm": 0.64453125, + "learning_rate": 0.0002, + "loss": 1.9247, + "step": 2486 + }, + { + "epoch": 0.7842951750236519, + "grad_norm": 0.66796875, + "learning_rate": 0.0002, + "loss": 2.1068, + "step": 2487 + }, + { + "epoch": 0.7846105329549038, + "grad_norm": 0.578125, + "learning_rate": 0.0002, + "loss": 2.311, + "step": 2488 + }, + { + "epoch": 0.7849258908861558, + "grad_norm": 0.61328125, + "learning_rate": 0.0002, + "loss": 2.2598, + "step": 2489 + }, + { + "epoch": 0.7852412488174078, + "grad_norm": 0.640625, + "learning_rate": 0.0002, + "loss": 2.113, + "step": 2490 + }, + { + "epoch": 0.7855566067486597, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 2.1202, + "step": 2491 + }, + { + "epoch": 0.7858719646799117, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 2.64, + "step": 2492 + }, + { + "epoch": 0.7861873226111636, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 2.9986, + "step": 2493 + }, + { + "epoch": 0.7865026805424157, + "grad_norm": 0.83203125, + "learning_rate": 0.0002, + "loss": 2.5397, + "step": 2494 + }, + { + "epoch": 0.7868180384736676, + "grad_norm": 0.84765625, + "learning_rate": 0.0002, + "loss": 2.1041, + "step": 2495 + }, + { + "epoch": 0.7871333964049196, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 2.7464, + "step": 2496 + }, + { + "epoch": 0.7874487543361716, + "grad_norm": 0.81640625, + "learning_rate": 0.0002, + "loss": 2.2531, + "step": 2497 + }, + { + "epoch": 0.7877641122674235, + "grad_norm": 0.72265625, + "learning_rate": 0.0002, + "loss": 2.3336, + "step": 2498 + }, + { + "epoch": 0.7880794701986755, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 2.6142, + "step": 2499 + }, + { + "epoch": 0.7883948281299274, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 2.593, + "step": 2500 + }, + { + "epoch": 0.7887101860611795, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.0761, + "step": 2501 + }, + { + "epoch": 0.7890255439924314, + "grad_norm": 0.1455078125, + "learning_rate": 0.0002, + "loss": 1.0343, + "step": 2502 + }, + { + "epoch": 0.7893409019236833, + "grad_norm": 0.1455078125, + "learning_rate": 0.0002, + "loss": 0.9651, + "step": 2503 + }, + { + "epoch": 0.7896562598549354, + "grad_norm": 0.166015625, + "learning_rate": 0.0002, + "loss": 1.1065, + "step": 2504 + }, + { + "epoch": 0.7899716177861873, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.2091, + "step": 2505 + }, + { + "epoch": 0.7902869757174393, + "grad_norm": 0.154296875, + "learning_rate": 0.0002, + "loss": 1.136, + "step": 2506 + }, + { + "epoch": 0.7906023336486913, + "grad_norm": 0.1591796875, + "learning_rate": 0.0002, + "loss": 1.0237, + "step": 2507 + }, + { + "epoch": 0.7909176915799432, + "grad_norm": 0.1630859375, + "learning_rate": 0.0002, + "loss": 1.1861, + "step": 2508 + }, + { + "epoch": 0.7912330495111952, + "grad_norm": 0.1689453125, + "learning_rate": 0.0002, + "loss": 1.0554, + "step": 2509 + }, + { + "epoch": 0.7915484074424471, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.2025, + "step": 2510 + }, + { + "epoch": 0.7918637653736992, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.1077, + "step": 2511 + }, + { + "epoch": 0.7921791233049511, + "grad_norm": 0.19140625, + "learning_rate": 0.0002, + "loss": 1.1101, + "step": 2512 + }, + { + "epoch": 0.7924944812362031, + "grad_norm": 0.193359375, + "learning_rate": 0.0002, + "loss": 1.1389, + "step": 2513 + }, + { + "epoch": 0.7928098391674551, + "grad_norm": 0.2451171875, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 2514 + }, + { + "epoch": 0.793125197098707, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.0499, + "step": 2515 + }, + { + "epoch": 0.793440555029959, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 1.2176, + "step": 2516 + }, + { + "epoch": 0.793755912961211, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.2636, + "step": 2517 + }, + { + "epoch": 0.794071270892463, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.1765, + "step": 2518 + }, + { + "epoch": 0.7943866288237149, + "grad_norm": 0.248046875, + "learning_rate": 0.0002, + "loss": 1.289, + "step": 2519 + }, + { + "epoch": 0.7947019867549668, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 1.2569, + "step": 2520 + }, + { + "epoch": 0.7950173446862189, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 2521 + }, + { + "epoch": 0.7953327026174708, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 1.2846, + "step": 2522 + }, + { + "epoch": 0.7956480605487228, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.4501, + "step": 2523 + }, + { + "epoch": 0.7959634184799748, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.2779, + "step": 2524 + }, + { + "epoch": 0.7962787764112268, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 1.378, + "step": 2525 + }, + { + "epoch": 0.7965941343424787, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 2526 + }, + { + "epoch": 0.7969094922737306, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 1.3808, + "step": 2527 + }, + { + "epoch": 0.7972248502049827, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 1.4038, + "step": 2528 + }, + { + "epoch": 0.7975402081362346, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 1.4244, + "step": 2529 + }, + { + "epoch": 0.7978555660674866, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 1.527, + "step": 2530 + }, + { + "epoch": 0.7981709239987386, + "grad_norm": 0.361328125, + "learning_rate": 0.0002, + "loss": 1.5969, + "step": 2531 + }, + { + "epoch": 0.7984862819299905, + "grad_norm": 0.412109375, + "learning_rate": 0.0002, + "loss": 1.6238, + "step": 2532 + }, + { + "epoch": 0.7988016398612425, + "grad_norm": 0.423828125, + "learning_rate": 0.0002, + "loss": 1.6179, + "step": 2533 + }, + { + "epoch": 0.7991169977924945, + "grad_norm": 0.55859375, + "learning_rate": 0.0002, + "loss": 2.0139, + "step": 2534 + }, + { + "epoch": 0.7994323557237465, + "grad_norm": 0.515625, + "learning_rate": 0.0002, + "loss": 1.9451, + "step": 2535 + }, + { + "epoch": 0.7997477136549984, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 2.1175, + "step": 2536 + }, + { + "epoch": 0.8000630715862503, + "grad_norm": 0.671875, + "learning_rate": 0.0002, + "loss": 2.1534, + "step": 2537 + }, + { + "epoch": 0.8003784295175024, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 2.0738, + "step": 2538 + }, + { + "epoch": 0.8006937874487543, + "grad_norm": 0.7265625, + "learning_rate": 0.0002, + "loss": 2.5575, + "step": 2539 + }, + { + "epoch": 0.8010091453800063, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 2.5412, + "step": 2540 + }, + { + "epoch": 0.8013245033112583, + "grad_norm": 0.87890625, + "learning_rate": 0.0002, + "loss": 2.8718, + "step": 2541 + }, + { + "epoch": 0.8016398612425103, + "grad_norm": 0.921875, + "learning_rate": 0.0002, + "loss": 2.8291, + "step": 2542 + }, + { + "epoch": 0.8019552191737622, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 2.5857, + "step": 2543 + }, + { + "epoch": 0.8022705771050141, + "grad_norm": 0.73828125, + "learning_rate": 0.0002, + "loss": 2.3453, + "step": 2544 + }, + { + "epoch": 0.8025859350362662, + "grad_norm": 0.765625, + "learning_rate": 0.0002, + "loss": 2.1989, + "step": 2545 + }, + { + "epoch": 0.8029012929675181, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 2.3728, + "step": 2546 + }, + { + "epoch": 0.8032166508987701, + "grad_norm": 0.87890625, + "learning_rate": 0.0002, + "loss": 2.7038, + "step": 2547 + }, + { + "epoch": 0.8035320088300221, + "grad_norm": 0.72265625, + "learning_rate": 0.0002, + "loss": 2.2536, + "step": 2548 + }, + { + "epoch": 0.803847366761274, + "grad_norm": 0.71875, + "learning_rate": 0.0002, + "loss": 2.2502, + "step": 2549 + }, + { + "epoch": 0.804162724692526, + "grad_norm": 0.94921875, + "learning_rate": 0.0002, + "loss": 2.1404, + "step": 2550 + }, + { + "epoch": 0.804478082623778, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.1654, + "step": 2551 + }, + { + "epoch": 0.80479344055503, + "grad_norm": 0.1630859375, + "learning_rate": 0.0002, + "loss": 1.0161, + "step": 2552 + }, + { + "epoch": 0.8051087984862819, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.1048, + "step": 2553 + }, + { + "epoch": 0.805424156417534, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.0498, + "step": 2554 + }, + { + "epoch": 0.8057395143487859, + "grad_norm": 0.166015625, + "learning_rate": 0.0002, + "loss": 1.1625, + "step": 2555 + }, + { + "epoch": 0.8060548722800378, + "grad_norm": 0.1728515625, + "learning_rate": 0.0002, + "loss": 1.0507, + "step": 2556 + }, + { + "epoch": 0.8063702302112898, + "grad_norm": 0.1728515625, + "learning_rate": 0.0002, + "loss": 1.1853, + "step": 2557 + }, + { + "epoch": 0.8066855881425418, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.1163, + "step": 2558 + }, + { + "epoch": 0.8070009460737938, + "grad_norm": 0.1787109375, + "learning_rate": 0.0002, + "loss": 1.0572, + "step": 2559 + }, + { + "epoch": 0.8073163040050457, + "grad_norm": 0.185546875, + "learning_rate": 0.0002, + "loss": 1.1052, + "step": 2560 + }, + { + "epoch": 0.8076316619362977, + "grad_norm": 0.1728515625, + "learning_rate": 0.0002, + "loss": 1.0688, + "step": 2561 + }, + { + "epoch": 0.8079470198675497, + "grad_norm": 0.17578125, + "learning_rate": 0.0002, + "loss": 1.0358, + "step": 2562 + }, + { + "epoch": 0.8082623777988016, + "grad_norm": 0.1865234375, + "learning_rate": 0.0002, + "loss": 1.1006, + "step": 2563 + }, + { + "epoch": 0.8085777357300536, + "grad_norm": 0.2177734375, + "learning_rate": 0.0002, + "loss": 1.2449, + "step": 2564 + }, + { + "epoch": 0.8088930936613056, + "grad_norm": 0.2060546875, + "learning_rate": 0.0002, + "loss": 1.1986, + "step": 2565 + }, + { + "epoch": 0.8092084515925575, + "grad_norm": 0.2041015625, + "learning_rate": 0.0002, + "loss": 1.2286, + "step": 2566 + }, + { + "epoch": 0.8095238095238095, + "grad_norm": 0.2265625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 2567 + }, + { + "epoch": 0.8098391674550615, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.1304, + "step": 2568 + }, + { + "epoch": 0.8101545253863135, + "grad_norm": 0.462890625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 2569 + }, + { + "epoch": 0.8104698833175654, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 2570 + }, + { + "epoch": 0.8107852412488175, + "grad_norm": 0.37109375, + "learning_rate": 0.0002, + "loss": 1.2463, + "step": 2571 + }, + { + "epoch": 0.8111005991800694, + "grad_norm": 0.244140625, + "learning_rate": 0.0002, + "loss": 1.3831, + "step": 2572 + }, + { + "epoch": 0.8114159571113213, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 2573 + }, + { + "epoch": 0.8117313150425733, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.4709, + "step": 2574 + }, + { + "epoch": 0.8120466729738253, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 1.4192, + "step": 2575 + }, + { + "epoch": 0.8123620309050773, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 1.4726, + "step": 2576 + }, + { + "epoch": 0.8126773888363292, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 2577 + }, + { + "epoch": 0.8129927467675812, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.3841, + "step": 2578 + }, + { + "epoch": 0.8133081046988332, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 2579 + }, + { + "epoch": 0.8136234626300851, + "grad_norm": 0.3828125, + "learning_rate": 0.0002, + "loss": 1.5953, + "step": 2580 + }, + { + "epoch": 0.8139388205613372, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 1.4726, + "step": 2581 + }, + { + "epoch": 0.8142541784925891, + "grad_norm": 0.431640625, + "learning_rate": 0.0002, + "loss": 1.9028, + "step": 2582 + }, + { + "epoch": 0.8145695364238411, + "grad_norm": 0.43359375, + "learning_rate": 0.0002, + "loss": 1.72, + "step": 2583 + }, + { + "epoch": 0.814884894355093, + "grad_norm": 0.416015625, + "learning_rate": 0.0002, + "loss": 1.7194, + "step": 2584 + }, + { + "epoch": 0.815200252286345, + "grad_norm": 0.4765625, + "learning_rate": 0.0002, + "loss": 1.8322, + "step": 2585 + }, + { + "epoch": 0.815515610217597, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 2.073, + "step": 2586 + }, + { + "epoch": 0.8158309681488489, + "grad_norm": 0.53515625, + "learning_rate": 0.0002, + "loss": 1.8685, + "step": 2587 + }, + { + "epoch": 0.816146326080101, + "grad_norm": 0.625, + "learning_rate": 0.0002, + "loss": 2.4779, + "step": 2588 + }, + { + "epoch": 0.8164616840113529, + "grad_norm": 0.53515625, + "learning_rate": 0.0002, + "loss": 1.8885, + "step": 2589 + }, + { + "epoch": 0.8167770419426048, + "grad_norm": 0.6171875, + "learning_rate": 0.0002, + "loss": 2.4094, + "step": 2590 + }, + { + "epoch": 0.8170923998738568, + "grad_norm": 0.73828125, + "learning_rate": 0.0002, + "loss": 2.4418, + "step": 2591 + }, + { + "epoch": 0.8174077578051088, + "grad_norm": 0.6953125, + "learning_rate": 0.0002, + "loss": 2.5129, + "step": 2592 + }, + { + "epoch": 0.8177231157363608, + "grad_norm": 0.52734375, + "learning_rate": 0.0002, + "loss": 1.7572, + "step": 2593 + }, + { + "epoch": 0.8180384736676127, + "grad_norm": 0.75, + "learning_rate": 0.0002, + "loss": 2.7357, + "step": 2594 + }, + { + "epoch": 0.8183538315988647, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 1.9594, + "step": 2595 + }, + { + "epoch": 0.8186691895301167, + "grad_norm": 0.7890625, + "learning_rate": 0.0002, + "loss": 2.4195, + "step": 2596 + }, + { + "epoch": 0.8189845474613686, + "grad_norm": 0.6796875, + "learning_rate": 0.0002, + "loss": 2.5404, + "step": 2597 + }, + { + "epoch": 0.8192999053926207, + "grad_norm": 0.8984375, + "learning_rate": 0.0002, + "loss": 2.6234, + "step": 2598 + }, + { + "epoch": 0.8196152633238726, + "grad_norm": 0.6953125, + "learning_rate": 0.0002, + "loss": 2.2414, + "step": 2599 + }, + { + "epoch": 0.8199306212551246, + "grad_norm": 0.96484375, + "learning_rate": 0.0002, + "loss": 2.7505, + "step": 2600 + }, + { + "epoch": 0.8202459791863765, + "grad_norm": 0.1513671875, + "learning_rate": 0.0002, + "loss": 1.0409, + "step": 2601 + }, + { + "epoch": 0.8205613371176285, + "grad_norm": 0.1640625, + "learning_rate": 0.0002, + "loss": 1.1216, + "step": 2602 + }, + { + "epoch": 0.8208766950488805, + "grad_norm": 0.150390625, + "learning_rate": 0.0002, + "loss": 1.0589, + "step": 2603 + }, + { + "epoch": 0.8211920529801324, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.1335, + "step": 2604 + }, + { + "epoch": 0.8215074109113845, + "grad_norm": 0.1630859375, + "learning_rate": 0.0002, + "loss": 1.0985, + "step": 2605 + }, + { + "epoch": 0.8218227688426364, + "grad_norm": 0.1513671875, + "learning_rate": 0.0002, + "loss": 1.1405, + "step": 2606 + }, + { + "epoch": 0.8221381267738883, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.0963, + "step": 2607 + }, + { + "epoch": 0.8224534847051403, + "grad_norm": 0.1796875, + "learning_rate": 0.0002, + "loss": 1.102, + "step": 2608 + }, + { + "epoch": 0.8227688426363923, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.0453, + "step": 2609 + }, + { + "epoch": 0.8230842005676443, + "grad_norm": 0.189453125, + "learning_rate": 0.0002, + "loss": 1.1377, + "step": 2610 + }, + { + "epoch": 0.8233995584988962, + "grad_norm": 0.17578125, + "learning_rate": 0.0002, + "loss": 1.115, + "step": 2611 + }, + { + "epoch": 0.8237149164301483, + "grad_norm": 0.197265625, + "learning_rate": 0.0002, + "loss": 1.1305, + "step": 2612 + }, + { + "epoch": 0.8240302743614002, + "grad_norm": 0.2001953125, + "learning_rate": 0.0002, + "loss": 1.1841, + "step": 2613 + }, + { + "epoch": 0.8243456322926521, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 1.2604, + "step": 2614 + }, + { + "epoch": 0.8246609902239042, + "grad_norm": 0.193359375, + "learning_rate": 0.0002, + "loss": 1.228, + "step": 2615 + }, + { + "epoch": 0.8249763481551561, + "grad_norm": 0.2119140625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 2616 + }, + { + "epoch": 0.8252917060864081, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 2617 + }, + { + "epoch": 0.82560706401766, + "grad_norm": 0.2138671875, + "learning_rate": 0.0002, + "loss": 1.1636, + "step": 2618 + }, + { + "epoch": 0.82560706401766, + "eval_loss": 1.6476565599441528, + "eval_runtime": 148.5236, + "eval_samples_per_second": 6.733, + "eval_steps_per_second": 6.733, + "step": 2618 + }, + { + "epoch": 0.82560706401766, + "mmlu_eval_accuracy": 0.6214867065853645, + "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182, + "mmlu_eval_accuracy_anatomy": 0.5714285714285714, + "mmlu_eval_accuracy_astronomy": 0.75, + "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, + "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034, + "mmlu_eval_accuracy_college_biology": 0.75, + "mmlu_eval_accuracy_college_chemistry": 0.625, + "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, + "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365, + "mmlu_eval_accuracy_college_medicine": 0.6818181818181818, + "mmlu_eval_accuracy_college_physics": 0.7272727272727273, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.5769230769230769, + "mmlu_eval_accuracy_econometrics": 0.4166666666666667, + "mmlu_eval_accuracy_electrical_engineering": 0.5, + "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634, + "mmlu_eval_accuracy_formal_logic": 0.5714285714285714, + "mmlu_eval_accuracy_global_facts": 0.4, + "mmlu_eval_accuracy_high_school_biology": 0.65625, + "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.8888888888888888, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8571428571428571, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442, + "mmlu_eval_accuracy_high_school_mathematics": 0.3448275862068966, + "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154, + "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354, + "mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667, + "mmlu_eval_accuracy_high_school_statistics": 0.4782608695652174, + "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727, + "mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693, + "mmlu_eval_accuracy_human_aging": 0.6086956521739131, + "mmlu_eval_accuracy_human_sexuality": 0.8333333333333334, + "mmlu_eval_accuracy_international_law": 0.8461538461538461, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.7777777777777778, + "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.88, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7558139534883721, + "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842, + "mmlu_eval_accuracy_moral_scenarios": 0.28, + "mmlu_eval_accuracy_nutrition": 0.6363636363636364, + "mmlu_eval_accuracy_philosophy": 0.7352941176470589, + "mmlu_eval_accuracy_prehistory": 0.5714285714285714, + "mmlu_eval_accuracy_professional_accounting": 0.22580645161290322, + "mmlu_eval_accuracy_professional_law": 0.40588235294117647, + "mmlu_eval_accuracy_professional_medicine": 0.7419354838709677, + "mmlu_eval_accuracy_professional_psychology": 0.6376811594202898, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7407407407407407, + "mmlu_eval_accuracy_sociology": 0.9090909090909091, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.5555555555555556, + "mmlu_eval_accuracy_world_religions": 0.8947368421052632, + "mmlu_loss": 6.15748606095666, + "step": 2618 + }, + { + "epoch": 0.825922421948912, + "grad_norm": 0.2490234375, + "learning_rate": 0.0002, + "loss": 1.4076, + "step": 2619 + }, + { + "epoch": 0.826237779880164, + "grad_norm": 0.2412109375, + "learning_rate": 0.0002, + "loss": 1.2286, + "step": 2620 + }, + { + "epoch": 0.8265531378114159, + "grad_norm": 0.240234375, + "learning_rate": 0.0002, + "loss": 1.1193, + "step": 2621 + }, + { + "epoch": 0.826868495742668, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.3, + "step": 2622 + }, + { + "epoch": 0.8271838536739199, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.3915, + "step": 2623 + }, + { + "epoch": 0.8274992116051719, + "grad_norm": 0.34765625, + "learning_rate": 0.0002, + "loss": 1.7888, + "step": 2624 + }, + { + "epoch": 0.8278145695364238, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 1.5007, + "step": 2625 + }, + { + "epoch": 0.8281299274676758, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 1.4139, + "step": 2626 + }, + { + "epoch": 0.8284452853989278, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 1.2953, + "step": 2627 + }, + { + "epoch": 0.8287606433301797, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 1.2675, + "step": 2628 + }, + { + "epoch": 0.8290760012614318, + "grad_norm": 0.353515625, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 2629 + }, + { + "epoch": 0.8293913591926837, + "grad_norm": 0.3671875, + "learning_rate": 0.0002, + "loss": 1.6895, + "step": 2630 + }, + { + "epoch": 0.8297067171239356, + "grad_norm": 0.482421875, + "learning_rate": 0.0002, + "loss": 1.8809, + "step": 2631 + }, + { + "epoch": 0.8300220750551877, + "grad_norm": 0.38671875, + "learning_rate": 0.0002, + "loss": 1.4017, + "step": 2632 + }, + { + "epoch": 0.8303374329864396, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 2.069, + "step": 2633 + }, + { + "epoch": 0.8306527909176916, + "grad_norm": 0.4609375, + "learning_rate": 0.0002, + "loss": 1.6091, + "step": 2634 + }, + { + "epoch": 0.8309681488489435, + "grad_norm": 0.64453125, + "learning_rate": 0.0002, + "loss": 1.8532, + "step": 2635 + }, + { + "epoch": 0.8312835067801955, + "grad_norm": 0.52734375, + "learning_rate": 0.0002, + "loss": 2.1941, + "step": 2636 + }, + { + "epoch": 0.8315988647114475, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 1.8197, + "step": 2637 + }, + { + "epoch": 0.8319142226426994, + "grad_norm": 0.81640625, + "learning_rate": 0.0002, + "loss": 2.1747, + "step": 2638 + }, + { + "epoch": 0.8322295805739515, + "grad_norm": 0.62109375, + "learning_rate": 0.0002, + "loss": 2.2379, + "step": 2639 + }, + { + "epoch": 0.8325449385052034, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 2.0997, + "step": 2640 + }, + { + "epoch": 0.8328602964364554, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 2.11, + "step": 2641 + }, + { + "epoch": 0.8331756543677074, + "grad_norm": 0.671875, + "learning_rate": 0.0002, + "loss": 2.3921, + "step": 2642 + }, + { + "epoch": 0.8334910122989593, + "grad_norm": 0.70703125, + "learning_rate": 0.0002, + "loss": 2.5424, + "step": 2643 + }, + { + "epoch": 0.8338063702302113, + "grad_norm": 0.640625, + "learning_rate": 0.0002, + "loss": 2.3225, + "step": 2644 + }, + { + "epoch": 0.8341217281614632, + "grad_norm": 0.7890625, + "learning_rate": 0.0002, + "loss": 2.4977, + "step": 2645 + }, + { + "epoch": 0.8344370860927153, + "grad_norm": 1.078125, + "learning_rate": 0.0002, + "loss": 2.969, + "step": 2646 + }, + { + "epoch": 0.8347524440239672, + "grad_norm": 0.66796875, + "learning_rate": 0.0002, + "loss": 2.1289, + "step": 2647 + }, + { + "epoch": 0.8350678019552191, + "grad_norm": 0.81640625, + "learning_rate": 0.0002, + "loss": 2.2049, + "step": 2648 + }, + { + "epoch": 0.8353831598864712, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 1.9446, + "step": 2649 + }, + { + "epoch": 0.8356985178177231, + "grad_norm": 0.79296875, + "learning_rate": 0.0002, + "loss": 2.03, + "step": 2650 + }, + { + "epoch": 0.8360138757489751, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.1252, + "step": 2651 + }, + { + "epoch": 0.836329233680227, + "grad_norm": 0.17578125, + "learning_rate": 0.0002, + "loss": 1.1802, + "step": 2652 + }, + { + "epoch": 0.8366445916114791, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 1.2507, + "step": 2653 + }, + { + "epoch": 0.836959949542731, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 0.9355, + "step": 2654 + }, + { + "epoch": 0.8372753074739829, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.1034, + "step": 2655 + }, + { + "epoch": 0.837590665405235, + "grad_norm": 0.18359375, + "learning_rate": 0.0002, + "loss": 1.1984, + "step": 2656 + }, + { + "epoch": 0.8379060233364869, + "grad_norm": 0.1796875, + "learning_rate": 0.0002, + "loss": 1.137, + "step": 2657 + }, + { + "epoch": 0.8382213812677389, + "grad_norm": 0.189453125, + "learning_rate": 0.0002, + "loss": 1.0283, + "step": 2658 + }, + { + "epoch": 0.8385367391989909, + "grad_norm": 0.18359375, + "learning_rate": 0.0002, + "loss": 1.0988, + "step": 2659 + }, + { + "epoch": 0.8388520971302428, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 1.1654, + "step": 2660 + }, + { + "epoch": 0.8391674550614948, + "grad_norm": 0.208984375, + "learning_rate": 0.0002, + "loss": 1.0617, + "step": 2661 + }, + { + "epoch": 0.8394828129927467, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.1755, + "step": 2662 + }, + { + "epoch": 0.8397981709239988, + "grad_norm": 0.201171875, + "learning_rate": 0.0002, + "loss": 1.0887, + "step": 2663 + }, + { + "epoch": 0.8401135288552507, + "grad_norm": 0.2060546875, + "learning_rate": 0.0002, + "loss": 1.2738, + "step": 2664 + }, + { + "epoch": 0.8404288867865026, + "grad_norm": 0.2001953125, + "learning_rate": 0.0002, + "loss": 1.1812, + "step": 2665 + }, + { + "epoch": 0.8407442447177547, + "grad_norm": 0.2001953125, + "learning_rate": 0.0002, + "loss": 1.1883, + "step": 2666 + }, + { + "epoch": 0.8410596026490066, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 1.2029, + "step": 2667 + }, + { + "epoch": 0.8413749605802586, + "grad_norm": 0.2275390625, + "learning_rate": 0.0002, + "loss": 1.197, + "step": 2668 + }, + { + "epoch": 0.8416903185115105, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 1.2503, + "step": 2669 + }, + { + "epoch": 0.8420056764427626, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.3572, + "step": 2670 + }, + { + "epoch": 0.8423210343740145, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 1.3702, + "step": 2671 + }, + { + "epoch": 0.8426363923052664, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 2672 + }, + { + "epoch": 0.8429517502365185, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 1.3754, + "step": 2673 + }, + { + "epoch": 0.8432671081677704, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.094, + "step": 2674 + }, + { + "epoch": 0.8435824660990224, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 2675 + }, + { + "epoch": 0.8438978240302744, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 1.4125, + "step": 2676 + }, + { + "epoch": 0.8442131819615263, + "grad_norm": 0.4609375, + "learning_rate": 0.0002, + "loss": 1.5151, + "step": 2677 + }, + { + "epoch": 0.8445285398927783, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 1.4808, + "step": 2678 + }, + { + "epoch": 0.8448438978240302, + "grad_norm": 0.392578125, + "learning_rate": 0.0002, + "loss": 1.6056, + "step": 2679 + }, + { + "epoch": 0.8451592557552823, + "grad_norm": 0.4375, + "learning_rate": 0.0002, + "loss": 1.7525, + "step": 2680 + }, + { + "epoch": 0.8454746136865342, + "grad_norm": 0.373046875, + "learning_rate": 0.0002, + "loss": 1.387, + "step": 2681 + }, + { + "epoch": 0.8457899716177862, + "grad_norm": 0.416015625, + "learning_rate": 0.0002, + "loss": 1.8009, + "step": 2682 + }, + { + "epoch": 0.8461053295490382, + "grad_norm": 0.50390625, + "learning_rate": 0.0002, + "loss": 1.793, + "step": 2683 + }, + { + "epoch": 0.8464206874802901, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 1.9718, + "step": 2684 + }, + { + "epoch": 0.8467360454115421, + "grad_norm": 0.4609375, + "learning_rate": 0.0002, + "loss": 1.6257, + "step": 2685 + }, + { + "epoch": 0.847051403342794, + "grad_norm": 0.42578125, + "learning_rate": 0.0002, + "loss": 1.6977, + "step": 2686 + }, + { + "epoch": 0.8473667612740461, + "grad_norm": 0.61328125, + "learning_rate": 0.0002, + "loss": 1.9708, + "step": 2687 + }, + { + "epoch": 0.847682119205298, + "grad_norm": 0.60546875, + "learning_rate": 0.0002, + "loss": 1.9837, + "step": 2688 + }, + { + "epoch": 0.8479974771365499, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 1.891, + "step": 2689 + }, + { + "epoch": 0.848312835067802, + "grad_norm": 0.8359375, + "learning_rate": 0.0002, + "loss": 2.7301, + "step": 2690 + }, + { + "epoch": 0.8486281929990539, + "grad_norm": 0.6640625, + "learning_rate": 0.0002, + "loss": 2.1532, + "step": 2691 + }, + { + "epoch": 0.8489435509303059, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 2.582, + "step": 2692 + }, + { + "epoch": 0.8492589088615579, + "grad_norm": 0.6796875, + "learning_rate": 0.0002, + "loss": 1.8989, + "step": 2693 + }, + { + "epoch": 0.8495742667928098, + "grad_norm": 0.7265625, + "learning_rate": 0.0002, + "loss": 2.5277, + "step": 2694 + }, + { + "epoch": 0.8498896247240618, + "grad_norm": 0.76953125, + "learning_rate": 0.0002, + "loss": 2.2218, + "step": 2695 + }, + { + "epoch": 0.8502049826553137, + "grad_norm": 0.76953125, + "learning_rate": 0.0002, + "loss": 2.3844, + "step": 2696 + }, + { + "epoch": 0.8505203405865658, + "grad_norm": 0.71484375, + "learning_rate": 0.0002, + "loss": 1.9499, + "step": 2697 + }, + { + "epoch": 0.8508356985178177, + "grad_norm": 0.6796875, + "learning_rate": 0.0002, + "loss": 1.8931, + "step": 2698 + }, + { + "epoch": 0.8511510564490697, + "grad_norm": 0.828125, + "learning_rate": 0.0002, + "loss": 2.1675, + "step": 2699 + }, + { + "epoch": 0.8514664143803217, + "grad_norm": 1.15625, + "learning_rate": 0.0002, + "loss": 2.9056, + "step": 2700 + }, + { + "epoch": 0.8517817723115736, + "grad_norm": 0.1455078125, + "learning_rate": 0.0002, + "loss": 1.1264, + "step": 2701 + }, + { + "epoch": 0.8520971302428256, + "grad_norm": 0.1455078125, + "learning_rate": 0.0002, + "loss": 1.1262, + "step": 2702 + }, + { + "epoch": 0.8524124881740776, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.1748, + "step": 2703 + }, + { + "epoch": 0.8527278461053296, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.05, + "step": 2704 + }, + { + "epoch": 0.8530432040365815, + "grad_norm": 0.1591796875, + "learning_rate": 0.0002, + "loss": 1.1157, + "step": 2705 + }, + { + "epoch": 0.8533585619678334, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.1096, + "step": 2706 + }, + { + "epoch": 0.8536739198990855, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.0779, + "step": 2707 + }, + { + "epoch": 0.8539892778303374, + "grad_norm": 0.1640625, + "learning_rate": 0.0002, + "loss": 1.1086, + "step": 2708 + }, + { + "epoch": 0.8543046357615894, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.1712, + "step": 2709 + }, + { + "epoch": 0.8546199936928414, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.008, + "step": 2710 + }, + { + "epoch": 0.8549353516240934, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 1.0295, + "step": 2711 + }, + { + "epoch": 0.8552507095553453, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 1.185, + "step": 2712 + }, + { + "epoch": 0.8555660674865972, + "grad_norm": 0.189453125, + "learning_rate": 0.0002, + "loss": 1.1652, + "step": 2713 + }, + { + "epoch": 0.8558814254178493, + "grad_norm": 0.189453125, + "learning_rate": 0.0002, + "loss": 1.1059, + "step": 2714 + }, + { + "epoch": 0.8561967833491012, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 1.1411, + "step": 2715 + }, + { + "epoch": 0.8565121412803532, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 1.2165, + "step": 2716 + }, + { + "epoch": 0.8568274992116052, + "grad_norm": 0.22265625, + "learning_rate": 0.0002, + "loss": 1.1951, + "step": 2717 + }, + { + "epoch": 0.8571428571428571, + "grad_norm": 0.37109375, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 2718 + }, + { + "epoch": 0.8574582150741091, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.199, + "step": 2719 + }, + { + "epoch": 0.8577735730053611, + "grad_norm": 0.2177734375, + "learning_rate": 0.0002, + "loss": 1.1458, + "step": 2720 + }, + { + "epoch": 0.8580889309366131, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 1.2002, + "step": 2721 + }, + { + "epoch": 0.858404288867865, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.2604, + "step": 2722 + }, + { + "epoch": 0.8587196467991169, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 1.5211, + "step": 2723 + }, + { + "epoch": 0.859035004730369, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 2724 + }, + { + "epoch": 0.8593503626616209, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 1.6514, + "step": 2725 + }, + { + "epoch": 0.8596657205928729, + "grad_norm": 0.330078125, + "learning_rate": 0.0002, + "loss": 1.5246, + "step": 2726 + }, + { + "epoch": 0.8599810785241249, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 1.2977, + "step": 2727 + }, + { + "epoch": 0.8602964364553769, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.2723, + "step": 2728 + }, + { + "epoch": 0.8606117943866288, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 1.642, + "step": 2729 + }, + { + "epoch": 0.8609271523178808, + "grad_norm": 0.34765625, + "learning_rate": 0.0002, + "loss": 1.5131, + "step": 2730 + }, + { + "epoch": 0.8612425102491328, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 1.3905, + "step": 2731 + }, + { + "epoch": 0.8615578681803847, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 1.5085, + "step": 2732 + }, + { + "epoch": 0.8618732261116367, + "grad_norm": 0.4453125, + "learning_rate": 0.0002, + "loss": 1.6843, + "step": 2733 + }, + { + "epoch": 0.8621885840428887, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 2.1759, + "step": 2734 + }, + { + "epoch": 0.8625039419741406, + "grad_norm": 0.53515625, + "learning_rate": 0.0002, + "loss": 1.7916, + "step": 2735 + }, + { + "epoch": 0.8628192999053926, + "grad_norm": 0.5703125, + "learning_rate": 0.0002, + "loss": 2.2492, + "step": 2736 + }, + { + "epoch": 0.8631346578366446, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 1.641, + "step": 2737 + }, + { + "epoch": 0.8634500157678966, + "grad_norm": 0.640625, + "learning_rate": 0.0002, + "loss": 2.1668, + "step": 2738 + }, + { + "epoch": 0.8637653736991485, + "grad_norm": 0.61328125, + "learning_rate": 0.0002, + "loss": 2.2757, + "step": 2739 + }, + { + "epoch": 0.8640807316304006, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 2.1881, + "step": 2740 + }, + { + "epoch": 0.8643960895616525, + "grad_norm": 1.15625, + "learning_rate": 0.0002, + "loss": 3.2461, + "step": 2741 + }, + { + "epoch": 0.8647114474929044, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 2.0547, + "step": 2742 + }, + { + "epoch": 0.8650268054241564, + "grad_norm": 0.84375, + "learning_rate": 0.0002, + "loss": 2.7016, + "step": 2743 + }, + { + "epoch": 0.8653421633554084, + "grad_norm": 0.7265625, + "learning_rate": 0.0002, + "loss": 2.5252, + "step": 2744 + }, + { + "epoch": 0.8656575212866604, + "grad_norm": 0.75, + "learning_rate": 0.0002, + "loss": 2.2324, + "step": 2745 + }, + { + "epoch": 0.8659728792179123, + "grad_norm": 0.76171875, + "learning_rate": 0.0002, + "loss": 2.3259, + "step": 2746 + }, + { + "epoch": 0.8662882371491643, + "grad_norm": 0.8828125, + "learning_rate": 0.0002, + "loss": 2.8784, + "step": 2747 + }, + { + "epoch": 0.8666035950804163, + "grad_norm": 0.63671875, + "learning_rate": 0.0002, + "loss": 2.001, + "step": 2748 + }, + { + "epoch": 0.8669189530116682, + "grad_norm": 0.80859375, + "learning_rate": 0.0002, + "loss": 2.3363, + "step": 2749 + }, + { + "epoch": 0.8672343109429203, + "grad_norm": 0.859375, + "learning_rate": 0.0002, + "loss": 2.2094, + "step": 2750 + }, + { + "epoch": 0.8675496688741722, + "grad_norm": 0.16015625, + "learning_rate": 0.0002, + "loss": 1.0767, + "step": 2751 + }, + { + "epoch": 0.8678650268054242, + "grad_norm": 0.16015625, + "learning_rate": 0.0002, + "loss": 1.0747, + "step": 2752 + }, + { + "epoch": 0.8681803847366761, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.0384, + "step": 2753 + }, + { + "epoch": 0.8684957426679281, + "grad_norm": 0.158203125, + "learning_rate": 0.0002, + "loss": 1.0726, + "step": 2754 + }, + { + "epoch": 0.8688111005991801, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.2577, + "step": 2755 + }, + { + "epoch": 0.869126458530432, + "grad_norm": 0.2041015625, + "learning_rate": 0.0002, + "loss": 1.1868, + "step": 2756 + }, + { + "epoch": 0.8694418164616841, + "grad_norm": 0.166015625, + "learning_rate": 0.0002, + "loss": 1.0894, + "step": 2757 + }, + { + "epoch": 0.869757174392936, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 1.2341, + "step": 2758 + }, + { + "epoch": 0.8700725323241879, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.1456, + "step": 2759 + }, + { + "epoch": 0.8703878902554399, + "grad_norm": 0.185546875, + "learning_rate": 0.0002, + "loss": 1.0651, + "step": 2760 + }, + { + "epoch": 0.8707032481866919, + "grad_norm": 0.193359375, + "learning_rate": 0.0002, + "loss": 1.2001, + "step": 2761 + }, + { + "epoch": 0.8710186061179439, + "grad_norm": 0.1884765625, + "learning_rate": 0.0002, + "loss": 1.2442, + "step": 2762 + }, + { + "epoch": 0.8713339640491958, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.2442, + "step": 2763 + }, + { + "epoch": 0.8716493219804478, + "grad_norm": 0.193359375, + "learning_rate": 0.0002, + "loss": 1.2184, + "step": 2764 + }, + { + "epoch": 0.8719646799116998, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 1.1915, + "step": 2765 + }, + { + "epoch": 0.8722800378429517, + "grad_norm": 0.208984375, + "learning_rate": 0.0002, + "loss": 1.1986, + "step": 2766 + }, + { + "epoch": 0.8725953957742038, + "grad_norm": 0.2138671875, + "learning_rate": 0.0002, + "loss": 1.2906, + "step": 2767 + }, + { + "epoch": 0.8729107537054557, + "grad_norm": 0.2177734375, + "learning_rate": 0.0002, + "loss": 1.2695, + "step": 2768 + }, + { + "epoch": 0.8732261116367077, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 2769 + }, + { + "epoch": 0.8735414695679596, + "grad_norm": 0.2294921875, + "learning_rate": 0.0002, + "loss": 1.1477, + "step": 2770 + }, + { + "epoch": 0.8738568274992116, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 1.2271, + "step": 2771 + }, + { + "epoch": 0.8741721854304636, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 2772 + }, + { + "epoch": 0.8744875433617155, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 1.2387, + "step": 2773 + }, + { + "epoch": 0.8748029012929676, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 1.4187, + "step": 2774 + }, + { + "epoch": 0.8751182592242195, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 1.3535, + "step": 2775 + }, + { + "epoch": 0.8754336171554714, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2776 + }, + { + "epoch": 0.8757489750867234, + "grad_norm": 0.330078125, + "learning_rate": 0.0002, + "loss": 1.4065, + "step": 2777 + }, + { + "epoch": 0.8760643330179754, + "grad_norm": 0.34765625, + "learning_rate": 0.0002, + "loss": 1.4056, + "step": 2778 + }, + { + "epoch": 0.8763796909492274, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 1.3006, + "step": 2779 + }, + { + "epoch": 0.8766950488804793, + "grad_norm": 0.40625, + "learning_rate": 0.0002, + "loss": 1.5757, + "step": 2780 + }, + { + "epoch": 0.8770104068117314, + "grad_norm": 0.408203125, + "learning_rate": 0.0002, + "loss": 1.6579, + "step": 2781 + }, + { + "epoch": 0.8773257647429833, + "grad_norm": 0.423828125, + "learning_rate": 0.0002, + "loss": 1.5053, + "step": 2782 + }, + { + "epoch": 0.8776411226742352, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 2.1916, + "step": 2783 + }, + { + "epoch": 0.8779564806054873, + "grad_norm": 0.60546875, + "learning_rate": 0.0002, + "loss": 1.6278, + "step": 2784 + }, + { + "epoch": 0.8782718385367392, + "grad_norm": 0.578125, + "learning_rate": 0.0002, + "loss": 1.9594, + "step": 2785 + }, + { + "epoch": 0.8785871964679912, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 2.0214, + "step": 2786 + }, + { + "epoch": 0.8789025543992431, + "grad_norm": 0.6171875, + "learning_rate": 0.0002, + "loss": 2.1394, + "step": 2787 + }, + { + "epoch": 0.8792179123304951, + "grad_norm": 0.74609375, + "learning_rate": 0.0002, + "loss": 2.3861, + "step": 2788 + }, + { + "epoch": 0.8795332702617471, + "grad_norm": 0.5546875, + "learning_rate": 0.0002, + "loss": 2.0088, + "step": 2789 + }, + { + "epoch": 0.879848628192999, + "grad_norm": 0.58984375, + "learning_rate": 0.0002, + "loss": 2.1111, + "step": 2790 + }, + { + "epoch": 0.8801639861242511, + "grad_norm": 0.9921875, + "learning_rate": 0.0002, + "loss": 2.0433, + "step": 2791 + }, + { + "epoch": 0.880479344055503, + "grad_norm": 0.74609375, + "learning_rate": 0.0002, + "loss": 2.3023, + "step": 2792 + }, + { + "epoch": 0.8807947019867549, + "grad_norm": 0.9765625, + "learning_rate": 0.0002, + "loss": 2.461, + "step": 2793 + }, + { + "epoch": 0.881110059918007, + "grad_norm": 0.984375, + "learning_rate": 0.0002, + "loss": 2.4896, + "step": 2794 + }, + { + "epoch": 0.8814254178492589, + "grad_norm": 0.71875, + "learning_rate": 0.0002, + "loss": 2.171, + "step": 2795 + }, + { + "epoch": 0.8817407757805109, + "grad_norm": 0.8828125, + "learning_rate": 0.0002, + "loss": 2.8272, + "step": 2796 + }, + { + "epoch": 0.8820561337117628, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 2.468, + "step": 2797 + }, + { + "epoch": 0.8823714916430149, + "grad_norm": 0.86328125, + "learning_rate": 0.0002, + "loss": 2.7454, + "step": 2798 + }, + { + "epoch": 0.8826868495742668, + "grad_norm": 0.73046875, + "learning_rate": 0.0002, + "loss": 1.8028, + "step": 2799 + }, + { + "epoch": 0.8830022075055187, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 2.6624, + "step": 2800 + }, + { + "epoch": 0.8833175654367708, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.0654, + "step": 2801 + }, + { + "epoch": 0.8836329233680227, + "grad_norm": 0.150390625, + "learning_rate": 0.0002, + "loss": 1.0404, + "step": 2802 + }, + { + "epoch": 0.8839482812992747, + "grad_norm": 0.158203125, + "learning_rate": 0.0002, + "loss": 1.1036, + "step": 2803 + }, + { + "epoch": 0.8842636392305266, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 1.0475, + "step": 2804 + }, + { + "epoch": 0.8845789971617786, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.1374, + "step": 2805 + }, + { + "epoch": 0.8845789971617786, + "eval_loss": 1.6614243984222412, + "eval_runtime": 149.1263, + "eval_samples_per_second": 6.706, + "eval_steps_per_second": 6.706, + "step": 2805 + }, + { + "epoch": 0.8845789971617786, + "mmlu_eval_accuracy": 0.6235704001434952, + "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, + "mmlu_eval_accuracy_anatomy": 0.5714285714285714, + "mmlu_eval_accuracy_astronomy": 0.75, + "mmlu_eval_accuracy_business_ethics": 0.45454545454545453, + "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034, + "mmlu_eval_accuracy_college_biology": 0.75, + "mmlu_eval_accuracy_college_chemistry": 0.625, + "mmlu_eval_accuracy_college_computer_science": 0.5454545454545454, + "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727, + "mmlu_eval_accuracy_college_medicine": 0.7272727272727273, + "mmlu_eval_accuracy_college_physics": 0.7272727272727273, + "mmlu_eval_accuracy_computer_security": 0.5454545454545454, + "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384, + "mmlu_eval_accuracy_econometrics": 0.5, + "mmlu_eval_accuracy_electrical_engineering": 0.5625, + "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244, + "mmlu_eval_accuracy_formal_logic": 0.42857142857142855, + "mmlu_eval_accuracy_global_facts": 0.4, + "mmlu_eval_accuracy_high_school_biology": 0.65625, + "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.8888888888888888, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8571428571428571, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209, + "mmlu_eval_accuracy_high_school_mathematics": 0.3103448275862069, + "mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923, + "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354, + "mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667, + "mmlu_eval_accuracy_high_school_statistics": 0.5217391304347826, + "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, + "mmlu_eval_accuracy_high_school_world_history": 0.8461538461538461, + "mmlu_eval_accuracy_human_aging": 0.6086956521739131, + "mmlu_eval_accuracy_human_sexuality": 0.75, + "mmlu_eval_accuracy_international_law": 0.9230769230769231, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.36363636363636365, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.84, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279, + "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842, + "mmlu_eval_accuracy_moral_scenarios": 0.33, + "mmlu_eval_accuracy_nutrition": 0.6666666666666666, + "mmlu_eval_accuracy_philosophy": 0.7352941176470589, + "mmlu_eval_accuracy_prehistory": 0.6, + "mmlu_eval_accuracy_professional_accounting": 0.1935483870967742, + "mmlu_eval_accuracy_professional_law": 0.3588235294117647, + "mmlu_eval_accuracy_professional_medicine": 0.6774193548387096, + "mmlu_eval_accuracy_professional_psychology": 0.6376811594202898, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7777777777777778, + "mmlu_eval_accuracy_sociology": 0.9545454545454546, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.5, + "mmlu_eval_accuracy_world_religions": 0.8947368421052632, + "mmlu_loss": 5.652842007144303, + "step": 2805 + }, + { + "epoch": 0.8848943550930306, + "grad_norm": 0.169921875, + "learning_rate": 0.0002, + "loss": 1.1368, + "step": 2806 + }, + { + "epoch": 0.8852097130242825, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.0709, + "step": 2807 + }, + { + "epoch": 0.8855250709555346, + "grad_norm": 0.18359375, + "learning_rate": 0.0002, + "loss": 1.2641, + "step": 2808 + }, + { + "epoch": 0.8858404288867865, + "grad_norm": 0.193359375, + "learning_rate": 0.0002, + "loss": 1.0416, + "step": 2809 + }, + { + "epoch": 0.8861557868180385, + "grad_norm": 0.205078125, + "learning_rate": 0.0002, + "loss": 1.1553, + "step": 2810 + }, + { + "epoch": 0.8864711447492905, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.0346, + "step": 2811 + }, + { + "epoch": 0.8867865026805424, + "grad_norm": 0.208984375, + "learning_rate": 0.0002, + "loss": 1.1348, + "step": 2812 + }, + { + "epoch": 0.8871018606117944, + "grad_norm": 0.2001953125, + "learning_rate": 0.0002, + "loss": 1.1157, + "step": 2813 + }, + { + "epoch": 0.8874172185430463, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 1.2594, + "step": 2814 + }, + { + "epoch": 0.8877325764742984, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 1.2249, + "step": 2815 + }, + { + "epoch": 0.8880479344055503, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 2816 + }, + { + "epoch": 0.8883632923368022, + "grad_norm": 0.2265625, + "learning_rate": 0.0002, + "loss": 1.2776, + "step": 2817 + }, + { + "epoch": 0.8886786502680543, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.1927, + "step": 2818 + }, + { + "epoch": 0.8889940081993062, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 1.3701, + "step": 2819 + }, + { + "epoch": 0.8893093661305582, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 1.2305, + "step": 2820 + }, + { + "epoch": 0.8896247240618101, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 2821 + }, + { + "epoch": 0.8899400819930621, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.4779, + "step": 2822 + }, + { + "epoch": 0.8902554399243141, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 1.2616, + "step": 2823 + }, + { + "epoch": 0.890570797855566, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 1.5007, + "step": 2824 + }, + { + "epoch": 0.8908861557868181, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 1.5053, + "step": 2825 + }, + { + "epoch": 0.89120151371807, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 2826 + }, + { + "epoch": 0.891516871649322, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 2827 + }, + { + "epoch": 0.891832229580574, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 1.5117, + "step": 2828 + }, + { + "epoch": 0.8921475875118259, + "grad_norm": 0.43359375, + "learning_rate": 0.0002, + "loss": 1.6907, + "step": 2829 + }, + { + "epoch": 0.8924629454430779, + "grad_norm": 0.39453125, + "learning_rate": 0.0002, + "loss": 1.7272, + "step": 2830 + }, + { + "epoch": 0.8927783033743298, + "grad_norm": 0.4375, + "learning_rate": 0.0002, + "loss": 1.7026, + "step": 2831 + }, + { + "epoch": 0.8930936613055819, + "grad_norm": 0.515625, + "learning_rate": 0.0002, + "loss": 1.7037, + "step": 2832 + }, + { + "epoch": 0.8934090192368338, + "grad_norm": 0.486328125, + "learning_rate": 0.0002, + "loss": 1.6669, + "step": 2833 + }, + { + "epoch": 0.8937243771680857, + "grad_norm": 0.52734375, + "learning_rate": 0.0002, + "loss": 1.9398, + "step": 2834 + }, + { + "epoch": 0.8940397350993378, + "grad_norm": 0.609375, + "learning_rate": 0.0002, + "loss": 1.8654, + "step": 2835 + }, + { + "epoch": 0.8943550930305897, + "grad_norm": 0.60546875, + "learning_rate": 0.0002, + "loss": 2.2186, + "step": 2836 + }, + { + "epoch": 0.8946704509618417, + "grad_norm": 0.66796875, + "learning_rate": 0.0002, + "loss": 2.4679, + "step": 2837 + }, + { + "epoch": 0.8949858088930936, + "grad_norm": 0.96875, + "learning_rate": 0.0002, + "loss": 2.238, + "step": 2838 + }, + { + "epoch": 0.8953011668243457, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 2.1007, + "step": 2839 + }, + { + "epoch": 0.8956165247555976, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 2.119, + "step": 2840 + }, + { + "epoch": 0.8959318826868495, + "grad_norm": 0.734375, + "learning_rate": 0.0002, + "loss": 2.8384, + "step": 2841 + }, + { + "epoch": 0.8962472406181016, + "grad_norm": 0.92578125, + "learning_rate": 0.0002, + "loss": 2.6056, + "step": 2842 + }, + { + "epoch": 0.8965625985493535, + "grad_norm": 0.74609375, + "learning_rate": 0.0002, + "loss": 2.3496, + "step": 2843 + }, + { + "epoch": 0.8968779564806055, + "grad_norm": 0.75, + "learning_rate": 0.0002, + "loss": 2.3906, + "step": 2844 + }, + { + "epoch": 0.8971933144118575, + "grad_norm": 0.6875, + "learning_rate": 0.0002, + "loss": 2.405, + "step": 2845 + }, + { + "epoch": 0.8975086723431094, + "grad_norm": 3.375, + "learning_rate": 0.0002, + "loss": 3.202, + "step": 2846 + }, + { + "epoch": 0.8978240302743614, + "grad_norm": 0.71875, + "learning_rate": 0.0002, + "loss": 2.0515, + "step": 2847 + }, + { + "epoch": 0.8981393882056133, + "grad_norm": 0.8046875, + "learning_rate": 0.0002, + "loss": 2.2276, + "step": 2848 + }, + { + "epoch": 0.8984547461368654, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 2.3253, + "step": 2849 + }, + { + "epoch": 0.8987701040681173, + "grad_norm": 0.984375, + "learning_rate": 0.0002, + "loss": 2.2014, + "step": 2850 + }, + { + "epoch": 0.8990854619993692, + "grad_norm": 0.142578125, + "learning_rate": 0.0002, + "loss": 1.1607, + "step": 2851 + }, + { + "epoch": 0.8994008199306213, + "grad_norm": 0.169921875, + "learning_rate": 0.0002, + "loss": 1.1963, + "step": 2852 + }, + { + "epoch": 0.8997161778618732, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 0.9569, + "step": 2853 + }, + { + "epoch": 0.9000315357931252, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.079, + "step": 2854 + }, + { + "epoch": 0.9003468937243772, + "grad_norm": 0.19140625, + "learning_rate": 0.0002, + "loss": 1.0962, + "step": 2855 + }, + { + "epoch": 0.9006622516556292, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.0605, + "step": 2856 + }, + { + "epoch": 0.9009776095868811, + "grad_norm": 0.1884765625, + "learning_rate": 0.0002, + "loss": 1.0927, + "step": 2857 + }, + { + "epoch": 0.901292967518133, + "grad_norm": 0.1689453125, + "learning_rate": 0.0002, + "loss": 1.1087, + "step": 2858 + }, + { + "epoch": 0.9016083254493851, + "grad_norm": 0.185546875, + "learning_rate": 0.0002, + "loss": 1.1041, + "step": 2859 + }, + { + "epoch": 0.901923683380637, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 1.123, + "step": 2860 + }, + { + "epoch": 0.902239041311889, + "grad_norm": 0.22265625, + "learning_rate": 0.0002, + "loss": 1.2895, + "step": 2861 + }, + { + "epoch": 0.902554399243141, + "grad_norm": 0.193359375, + "learning_rate": 0.0002, + "loss": 1.0479, + "step": 2862 + }, + { + "epoch": 0.9028697571743929, + "grad_norm": 0.193359375, + "learning_rate": 0.0002, + "loss": 1.0977, + "step": 2863 + }, + { + "epoch": 0.9031851151056449, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 0.9578, + "step": 2864 + }, + { + "epoch": 0.9035004730368968, + "grad_norm": 0.2060546875, + "learning_rate": 0.0002, + "loss": 1.1119, + "step": 2865 + }, + { + "epoch": 0.9038158309681489, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.12, + "step": 2866 + }, + { + "epoch": 0.9041311888994008, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.1158, + "step": 2867 + }, + { + "epoch": 0.9044465468306528, + "grad_norm": 0.2294921875, + "learning_rate": 0.0002, + "loss": 1.2085, + "step": 2868 + }, + { + "epoch": 0.9047619047619048, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.2224, + "step": 2869 + }, + { + "epoch": 0.9050772626931567, + "grad_norm": 0.2333984375, + "learning_rate": 0.0002, + "loss": 1.2043, + "step": 2870 + }, + { + "epoch": 0.9053926206244087, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 2871 + }, + { + "epoch": 0.9057079785556607, + "grad_norm": 0.2333984375, + "learning_rate": 0.0002, + "loss": 1.2488, + "step": 2872 + }, + { + "epoch": 0.9060233364869127, + "grad_norm": 0.255859375, + "learning_rate": 0.0002, + "loss": 1.2064, + "step": 2873 + }, + { + "epoch": 0.9063386944181646, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.4927, + "step": 2874 + }, + { + "epoch": 0.9066540523494165, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.2412, + "step": 2875 + }, + { + "epoch": 0.9069694102806686, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2876 + }, + { + "epoch": 0.9072847682119205, + "grad_norm": 0.40625, + "learning_rate": 0.0002, + "loss": 1.4592, + "step": 2877 + }, + { + "epoch": 0.9076001261431725, + "grad_norm": 0.349609375, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 2878 + }, + { + "epoch": 0.9079154840744245, + "grad_norm": 0.388671875, + "learning_rate": 0.0002, + "loss": 1.8721, + "step": 2879 + }, + { + "epoch": 0.9082308420056765, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 1.4328, + "step": 2880 + }, + { + "epoch": 0.9085461999369284, + "grad_norm": 0.384765625, + "learning_rate": 0.0002, + "loss": 1.4266, + "step": 2881 + }, + { + "epoch": 0.9088615578681803, + "grad_norm": 0.50390625, + "learning_rate": 0.0002, + "loss": 1.6709, + "step": 2882 + }, + { + "epoch": 0.9091769157994324, + "grad_norm": 0.56640625, + "learning_rate": 0.0002, + "loss": 2.0798, + "step": 2883 + }, + { + "epoch": 0.9094922737306843, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 1.9695, + "step": 2884 + }, + { + "epoch": 0.9098076316619363, + "grad_norm": 0.58984375, + "learning_rate": 0.0002, + "loss": 2.0552, + "step": 2885 + }, + { + "epoch": 0.9101229895931883, + "grad_norm": 0.75, + "learning_rate": 0.0002, + "loss": 2.0079, + "step": 2886 + }, + { + "epoch": 0.9104383475244402, + "grad_norm": 0.64453125, + "learning_rate": 0.0002, + "loss": 2.2637, + "step": 2887 + }, + { + "epoch": 0.9107537054556922, + "grad_norm": 0.61328125, + "learning_rate": 0.0002, + "loss": 2.1425, + "step": 2888 + }, + { + "epoch": 0.9110690633869442, + "grad_norm": 0.734375, + "learning_rate": 0.0002, + "loss": 2.5301, + "step": 2889 + }, + { + "epoch": 0.9113844213181962, + "grad_norm": 0.734375, + "learning_rate": 0.0002, + "loss": 2.2254, + "step": 2890 + }, + { + "epoch": 0.9116997792494481, + "grad_norm": 0.83984375, + "learning_rate": 0.0002, + "loss": 2.8518, + "step": 2891 + }, + { + "epoch": 0.9120151371807, + "grad_norm": 0.84375, + "learning_rate": 0.0002, + "loss": 2.3551, + "step": 2892 + }, + { + "epoch": 0.9123304951119521, + "grad_norm": 0.92578125, + "learning_rate": 0.0002, + "loss": 2.5807, + "step": 2893 + }, + { + "epoch": 0.912645853043204, + "grad_norm": 0.74609375, + "learning_rate": 0.0002, + "loss": 2.3539, + "step": 2894 + }, + { + "epoch": 0.912961210974456, + "grad_norm": 0.68359375, + "learning_rate": 0.0002, + "loss": 2.1911, + "step": 2895 + }, + { + "epoch": 0.913276568905708, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 2.2639, + "step": 2896 + }, + { + "epoch": 0.91359192683696, + "grad_norm": 0.83984375, + "learning_rate": 0.0002, + "loss": 2.5027, + "step": 2897 + }, + { + "epoch": 0.9139072847682119, + "grad_norm": 0.6875, + "learning_rate": 0.0002, + "loss": 2.2655, + "step": 2898 + }, + { + "epoch": 0.9142226426994638, + "grad_norm": 1.1796875, + "learning_rate": 0.0002, + "loss": 2.3187, + "step": 2899 + }, + { + "epoch": 0.9145380006307159, + "grad_norm": 1.4375, + "learning_rate": 0.0002, + "loss": 3.2871, + "step": 2900 + }, + { + "epoch": 0.9148533585619678, + "grad_norm": 0.1552734375, + "learning_rate": 0.0002, + "loss": 1.0588, + "step": 2901 + }, + { + "epoch": 0.9151687164932198, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.0324, + "step": 2902 + }, + { + "epoch": 0.9154840744244718, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.0369, + "step": 2903 + }, + { + "epoch": 0.9157994323557237, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.0786, + "step": 2904 + }, + { + "epoch": 0.9161147902869757, + "grad_norm": 0.158203125, + "learning_rate": 0.0002, + "loss": 1.0872, + "step": 2905 + }, + { + "epoch": 0.9164301482182277, + "grad_norm": 0.16015625, + "learning_rate": 0.0002, + "loss": 1.2096, + "step": 2906 + }, + { + "epoch": 0.9167455061494797, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.1087, + "step": 2907 + }, + { + "epoch": 0.9170608640807316, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.1072, + "step": 2908 + }, + { + "epoch": 0.9173762220119837, + "grad_norm": 0.1943359375, + "learning_rate": 0.0002, + "loss": 1.2005, + "step": 2909 + }, + { + "epoch": 0.9176915799432356, + "grad_norm": 0.17578125, + "learning_rate": 0.0002, + "loss": 1.1113, + "step": 2910 + }, + { + "epoch": 0.9180069378744875, + "grad_norm": 0.193359375, + "learning_rate": 0.0002, + "loss": 1.2651, + "step": 2911 + }, + { + "epoch": 0.9183222958057395, + "grad_norm": 0.1767578125, + "learning_rate": 0.0002, + "loss": 0.9908, + "step": 2912 + }, + { + "epoch": 0.9186376537369915, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 1.1871, + "step": 2913 + }, + { + "epoch": 0.9189530116682435, + "grad_norm": 0.19921875, + "learning_rate": 0.0002, + "loss": 1.2788, + "step": 2914 + }, + { + "epoch": 0.9192683695994954, + "grad_norm": 0.2001953125, + "learning_rate": 0.0002, + "loss": 1.1933, + "step": 2915 + }, + { + "epoch": 0.9195837275307474, + "grad_norm": 0.2177734375, + "learning_rate": 0.0002, + "loss": 1.29, + "step": 2916 + }, + { + "epoch": 0.9198990854619994, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.3938, + "step": 2917 + }, + { + "epoch": 0.9202144433932513, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.2255, + "step": 2918 + }, + { + "epoch": 0.9205298013245033, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 1.5373, + "step": 2919 + }, + { + "epoch": 0.9208451592557553, + "grad_norm": 0.2392578125, + "learning_rate": 0.0002, + "loss": 1.2894, + "step": 2920 + }, + { + "epoch": 0.9211605171870072, + "grad_norm": 0.2373046875, + "learning_rate": 0.0002, + "loss": 1.378, + "step": 2921 + }, + { + "epoch": 0.9214758751182592, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 1.2203, + "step": 2922 + }, + { + "epoch": 0.9217912330495112, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.4739, + "step": 2923 + }, + { + "epoch": 0.9221065909807632, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 1.4521, + "step": 2924 + }, + { + "epoch": 0.9224219489120151, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.2282, + "step": 2925 + }, + { + "epoch": 0.9227373068432672, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 1.571, + "step": 2926 + }, + { + "epoch": 0.9230526647745191, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 1.3932, + "step": 2927 + }, + { + "epoch": 0.923368022705771, + "grad_norm": 0.349609375, + "learning_rate": 0.0002, + "loss": 1.38, + "step": 2928 + }, + { + "epoch": 0.923683380637023, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 1.5119, + "step": 2929 + }, + { + "epoch": 0.923998738568275, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 1.5843, + "step": 2930 + }, + { + "epoch": 0.924314096499527, + "grad_norm": 0.369140625, + "learning_rate": 0.0002, + "loss": 1.5042, + "step": 2931 + }, + { + "epoch": 0.9246294544307789, + "grad_norm": 0.40234375, + "learning_rate": 0.0002, + "loss": 1.5426, + "step": 2932 + }, + { + "epoch": 0.9249448123620309, + "grad_norm": 0.4609375, + "learning_rate": 0.0002, + "loss": 1.803, + "step": 2933 + }, + { + "epoch": 0.9252601702932829, + "grad_norm": 0.45703125, + "learning_rate": 0.0002, + "loss": 1.8195, + "step": 2934 + }, + { + "epoch": 0.9255755282245348, + "grad_norm": 0.462890625, + "learning_rate": 0.0002, + "loss": 1.7978, + "step": 2935 + }, + { + "epoch": 0.9258908861557869, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 2.0293, + "step": 2936 + }, + { + "epoch": 0.9262062440870388, + "grad_norm": 3.9375, + "learning_rate": 0.0002, + "loss": 2.2722, + "step": 2937 + }, + { + "epoch": 0.9265216020182908, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 2.0289, + "step": 2938 + }, + { + "epoch": 0.9268369599495427, + "grad_norm": 0.8359375, + "learning_rate": 0.0002, + "loss": 2.8364, + "step": 2939 + }, + { + "epoch": 0.9271523178807947, + "grad_norm": 0.77734375, + "learning_rate": 0.0002, + "loss": 2.9, + "step": 2940 + }, + { + "epoch": 0.9274676758120467, + "grad_norm": 0.64453125, + "learning_rate": 0.0002, + "loss": 2.2801, + "step": 2941 + }, + { + "epoch": 0.9277830337432986, + "grad_norm": 0.76171875, + "learning_rate": 0.0002, + "loss": 2.4144, + "step": 2942 + }, + { + "epoch": 0.9280983916745507, + "grad_norm": 0.640625, + "learning_rate": 0.0002, + "loss": 2.2257, + "step": 2943 + }, + { + "epoch": 0.9284137496058026, + "grad_norm": 0.84375, + "learning_rate": 0.0002, + "loss": 2.8794, + "step": 2944 + }, + { + "epoch": 0.9287291075370545, + "grad_norm": 0.76171875, + "learning_rate": 0.0002, + "loss": 2.8294, + "step": 2945 + }, + { + "epoch": 0.9290444654683065, + "grad_norm": 0.7734375, + "learning_rate": 0.0002, + "loss": 2.4489, + "step": 2946 + }, + { + "epoch": 0.9293598233995585, + "grad_norm": 0.8984375, + "learning_rate": 0.0002, + "loss": 2.9276, + "step": 2947 + }, + { + "epoch": 0.9296751813308105, + "grad_norm": 0.74609375, + "learning_rate": 0.0002, + "loss": 2.4353, + "step": 2948 + }, + { + "epoch": 0.9299905392620624, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 2.1974, + "step": 2949 + }, + { + "epoch": 0.9303058971933144, + "grad_norm": 0.9765625, + "learning_rate": 0.0002, + "loss": 2.7593, + "step": 2950 + }, + { + "epoch": 0.9306212551245664, + "grad_norm": 0.255859375, + "learning_rate": 0.0002, + "loss": 1.1844, + "step": 2951 + }, + { + "epoch": 0.9309366130558183, + "grad_norm": 0.17578125, + "learning_rate": 0.0002, + "loss": 1.0736, + "step": 2952 + }, + { + "epoch": 0.9312519709870704, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.1248, + "step": 2953 + }, + { + "epoch": 0.9315673289183223, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 1.1194, + "step": 2954 + }, + { + "epoch": 0.9318826868495743, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.0035, + "step": 2955 + }, + { + "epoch": 0.9321980447808262, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.1176, + "step": 2956 + }, + { + "epoch": 0.9325134027120782, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.1537, + "step": 2957 + }, + { + "epoch": 0.9328287606433302, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.2007, + "step": 2958 + }, + { + "epoch": 0.9331441185745821, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 1.1401, + "step": 2959 + }, + { + "epoch": 0.9334594765058342, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.1555, + "step": 2960 + }, + { + "epoch": 0.9337748344370861, + "grad_norm": 0.1943359375, + "learning_rate": 0.0002, + "loss": 0.9713, + "step": 2961 + }, + { + "epoch": 0.934090192368338, + "grad_norm": 0.2041015625, + "learning_rate": 0.0002, + "loss": 1.0807, + "step": 2962 + }, + { + "epoch": 0.93440555029959, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 1.3015, + "step": 2963 + }, + { + "epoch": 0.934720908230842, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.1862, + "step": 2964 + }, + { + "epoch": 0.935036266162094, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.2548, + "step": 2965 + }, + { + "epoch": 0.9353516240933459, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 1.3539, + "step": 2966 + }, + { + "epoch": 0.935666982024598, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 1.1612, + "step": 2967 + }, + { + "epoch": 0.9359823399558499, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 1.3745, + "step": 2968 + }, + { + "epoch": 0.9362976978871018, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.5028, + "step": 2969 + }, + { + "epoch": 0.9366130558183539, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.3707, + "step": 2970 + }, + { + "epoch": 0.9369284137496058, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.2758, + "step": 2971 + }, + { + "epoch": 0.9372437716808578, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 1.4134, + "step": 2972 + }, + { + "epoch": 0.9375591296121097, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 1.3993, + "step": 2973 + }, + { + "epoch": 0.9378744875433617, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 1.3045, + "step": 2974 + }, + { + "epoch": 0.9381898454746137, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 1.4173, + "step": 2975 + }, + { + "epoch": 0.9385052034058656, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 1.6647, + "step": 2976 + }, + { + "epoch": 0.9388205613371177, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 1.6017, + "step": 2977 + }, + { + "epoch": 0.9391359192683696, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 1.5024, + "step": 2978 + }, + { + "epoch": 0.9394512771996215, + "grad_norm": 0.451171875, + "learning_rate": 0.0002, + "loss": 1.3916, + "step": 2979 + }, + { + "epoch": 0.9397666351308736, + "grad_norm": 0.39453125, + "learning_rate": 0.0002, + "loss": 1.5331, + "step": 2980 + }, + { + "epoch": 0.9400819930621255, + "grad_norm": 0.482421875, + "learning_rate": 0.0002, + "loss": 1.8516, + "step": 2981 + }, + { + "epoch": 0.9403973509933775, + "grad_norm": 0.427734375, + "learning_rate": 0.0002, + "loss": 1.6997, + "step": 2982 + }, + { + "epoch": 0.9407127089246294, + "grad_norm": 0.4609375, + "learning_rate": 0.0002, + "loss": 1.7164, + "step": 2983 + }, + { + "epoch": 0.9410280668558815, + "grad_norm": 0.482421875, + "learning_rate": 0.0002, + "loss": 1.6279, + "step": 2984 + }, + { + "epoch": 0.9413434247871334, + "grad_norm": 0.56640625, + "learning_rate": 0.0002, + "loss": 1.857, + "step": 2985 + }, + { + "epoch": 0.9416587827183853, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 2.2595, + "step": 2986 + }, + { + "epoch": 0.9419741406496374, + "grad_norm": 0.609375, + "learning_rate": 0.0002, + "loss": 2.4372, + "step": 2987 + }, + { + "epoch": 0.9422894985808893, + "grad_norm": 0.6640625, + "learning_rate": 0.0002, + "loss": 2.5082, + "step": 2988 + }, + { + "epoch": 0.9426048565121413, + "grad_norm": 0.80078125, + "learning_rate": 0.0002, + "loss": 2.2467, + "step": 2989 + }, + { + "epoch": 0.9429202144433932, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 2.197, + "step": 2990 + }, + { + "epoch": 0.9432355723746452, + "grad_norm": 0.671875, + "learning_rate": 0.0002, + "loss": 2.2568, + "step": 2991 + }, + { + "epoch": 0.9435509303058972, + "grad_norm": 0.90234375, + "learning_rate": 0.0002, + "loss": 2.2849, + "step": 2992 + }, + { + "epoch": 0.9435509303058972, + "eval_loss": 1.6418744325637817, + "eval_runtime": 149.0717, + "eval_samples_per_second": 6.708, + "eval_steps_per_second": 6.708, + "step": 2992 + }, + { + "epoch": 0.9435509303058972, + "mmlu_eval_accuracy": 0.615467452899633, + "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, + "mmlu_eval_accuracy_anatomy": 0.5, + "mmlu_eval_accuracy_astronomy": 0.6875, + "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, + "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034, + "mmlu_eval_accuracy_college_biology": 0.8125, + "mmlu_eval_accuracy_college_chemistry": 0.5, + "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, + "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365, + "mmlu_eval_accuracy_college_medicine": 0.6818181818181818, + "mmlu_eval_accuracy_college_physics": 0.5454545454545454, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.6538461538461539, + "mmlu_eval_accuracy_econometrics": 0.5, + "mmlu_eval_accuracy_electrical_engineering": 0.5, + "mmlu_eval_accuracy_elementary_mathematics": 0.4634146341463415, + "mmlu_eval_accuracy_formal_logic": 0.2857142857142857, + "mmlu_eval_accuracy_global_facts": 0.3, + "mmlu_eval_accuracy_high_school_biology": 0.65625, + "mmlu_eval_accuracy_high_school_chemistry": 0.5909090909090909, + "mmlu_eval_accuracy_high_school_computer_science": 0.7777777777777778, + "mmlu_eval_accuracy_high_school_european_history": 0.8888888888888888, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.4883720930232558, + "mmlu_eval_accuracy_high_school_mathematics": 0.3448275862068966, + "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, + "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, + "mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333, + "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, + "mmlu_eval_accuracy_high_school_us_history": 0.8636363636363636, + "mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693, + "mmlu_eval_accuracy_human_aging": 0.6521739130434783, + "mmlu_eval_accuracy_human_sexuality": 0.75, + "mmlu_eval_accuracy_international_law": 0.9230769230769231, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, + "mmlu_eval_accuracy_machine_learning": 0.2727272727272727, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.88, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7325581395348837, + "mmlu_eval_accuracy_moral_disputes": 0.5, + "mmlu_eval_accuracy_moral_scenarios": 0.24, + "mmlu_eval_accuracy_nutrition": 0.6666666666666666, + "mmlu_eval_accuracy_philosophy": 0.7352941176470589, + "mmlu_eval_accuracy_prehistory": 0.5428571428571428, + "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194, + "mmlu_eval_accuracy_professional_law": 0.36470588235294116, + "mmlu_eval_accuracy_professional_medicine": 0.7741935483870968, + "mmlu_eval_accuracy_professional_psychology": 0.6956521739130435, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7777777777777778, + "mmlu_eval_accuracy_sociology": 1.0, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.5, + "mmlu_eval_accuracy_world_religions": 0.9473684210526315, + "mmlu_loss": 5.921975422653009, + "step": 2992 + }, + { + "epoch": 0.9438662882371491, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 1.9797, + "step": 2993 + }, + { + "epoch": 0.9441816461684012, + "grad_norm": 0.90234375, + "learning_rate": 0.0002, + "loss": 2.5038, + "step": 2994 + }, + { + "epoch": 0.9444970040996531, + "grad_norm": 1.1640625, + "learning_rate": 0.0002, + "loss": 2.7903, + "step": 2995 + }, + { + "epoch": 0.9448123620309051, + "grad_norm": 0.828125, + "learning_rate": 0.0002, + "loss": 2.2498, + "step": 2996 + }, + { + "epoch": 0.945127719962157, + "grad_norm": 0.7890625, + "learning_rate": 0.0002, + "loss": 2.2077, + "step": 2997 + }, + { + "epoch": 0.945443077893409, + "grad_norm": 0.890625, + "learning_rate": 0.0002, + "loss": 2.334, + "step": 2998 + }, + { + "epoch": 0.945758435824661, + "grad_norm": 0.80078125, + "learning_rate": 0.0002, + "loss": 2.2372, + "step": 2999 + }, + { + "epoch": 0.9460737937559129, + "grad_norm": 1.0859375, + "learning_rate": 0.0002, + "loss": 2.6913, + "step": 3000 + }, + { + "epoch": 0.946389151687165, + "grad_norm": 0.142578125, + "learning_rate": 0.0002, + "loss": 1.1063, + "step": 3001 + }, + { + "epoch": 0.9467045096184169, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.1885, + "step": 3002 + }, + { + "epoch": 0.9470198675496688, + "grad_norm": 0.1630859375, + "learning_rate": 0.0002, + "loss": 1.116, + "step": 3003 + }, + { + "epoch": 0.9473352254809209, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.1427, + "step": 3004 + }, + { + "epoch": 0.9476505834121728, + "grad_norm": 0.1630859375, + "learning_rate": 0.0002, + "loss": 1.1703, + "step": 3005 + }, + { + "epoch": 0.9479659413434248, + "grad_norm": 0.158203125, + "learning_rate": 0.0002, + "loss": 1.0151, + "step": 3006 + }, + { + "epoch": 0.9482812992746767, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.085, + "step": 3007 + }, + { + "epoch": 0.9485966572059288, + "grad_norm": 0.1787109375, + "learning_rate": 0.0002, + "loss": 1.0716, + "step": 3008 + }, + { + "epoch": 0.9489120151371807, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 1.1744, + "step": 3009 + }, + { + "epoch": 0.9492273730684326, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.0199, + "step": 3010 + }, + { + "epoch": 0.9495427309996847, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.1445, + "step": 3011 + }, + { + "epoch": 0.9498580889309366, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.1637, + "step": 3012 + }, + { + "epoch": 0.9501734468621886, + "grad_norm": 0.23046875, + "learning_rate": 0.0002, + "loss": 1.0913, + "step": 3013 + }, + { + "epoch": 0.9504888047934406, + "grad_norm": 0.2119140625, + "learning_rate": 0.0002, + "loss": 1.1721, + "step": 3014 + }, + { + "epoch": 0.9508041627246925, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 1.1184, + "step": 3015 + }, + { + "epoch": 0.9511195206559445, + "grad_norm": 0.255859375, + "learning_rate": 0.0002, + "loss": 1.4525, + "step": 3016 + }, + { + "epoch": 0.9514348785871964, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 1.2342, + "step": 3017 + }, + { + "epoch": 0.9517502365184485, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.4143, + "step": 3018 + }, + { + "epoch": 0.9520655944497004, + "grad_norm": 0.2451171875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 3019 + }, + { + "epoch": 0.9523809523809523, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.3955, + "step": 3020 + }, + { + "epoch": 0.9526963103122044, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 1.378, + "step": 3021 + }, + { + "epoch": 0.9530116682434563, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.4967, + "step": 3022 + }, + { + "epoch": 0.9533270261747083, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 3023 + }, + { + "epoch": 0.9536423841059603, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 1.3905, + "step": 3024 + }, + { + "epoch": 0.9539577420372123, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 3025 + }, + { + "epoch": 0.9542730999684642, + "grad_norm": 0.349609375, + "learning_rate": 0.0002, + "loss": 1.5006, + "step": 3026 + }, + { + "epoch": 0.9545884578997161, + "grad_norm": 0.37109375, + "learning_rate": 0.0002, + "loss": 1.5254, + "step": 3027 + }, + { + "epoch": 0.9549038158309682, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.4727, + "step": 3028 + }, + { + "epoch": 0.9552191737622201, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 1.623, + "step": 3029 + }, + { + "epoch": 0.9555345316934721, + "grad_norm": 0.380859375, + "learning_rate": 0.0002, + "loss": 1.4998, + "step": 3030 + }, + { + "epoch": 0.9558498896247241, + "grad_norm": 0.37109375, + "learning_rate": 0.0002, + "loss": 1.4637, + "step": 3031 + }, + { + "epoch": 0.956165247555976, + "grad_norm": 0.41015625, + "learning_rate": 0.0002, + "loss": 1.9091, + "step": 3032 + }, + { + "epoch": 0.956480605487228, + "grad_norm": 0.49609375, + "learning_rate": 0.0002, + "loss": 1.5832, + "step": 3033 + }, + { + "epoch": 0.9567959634184799, + "grad_norm": 0.423828125, + "learning_rate": 0.0002, + "loss": 1.4172, + "step": 3034 + }, + { + "epoch": 0.957111321349732, + "grad_norm": 0.455078125, + "learning_rate": 0.0002, + "loss": 1.7496, + "step": 3035 + }, + { + "epoch": 0.9574266792809839, + "grad_norm": 0.546875, + "learning_rate": 0.0002, + "loss": 1.9299, + "step": 3036 + }, + { + "epoch": 0.9577420372122359, + "grad_norm": 0.6484375, + "learning_rate": 0.0002, + "loss": 1.8634, + "step": 3037 + }, + { + "epoch": 0.9580573951434879, + "grad_norm": 0.56640625, + "learning_rate": 0.0002, + "loss": 1.9918, + "step": 3038 + }, + { + "epoch": 0.9583727530747398, + "grad_norm": 0.671875, + "learning_rate": 0.0002, + "loss": 2.4615, + "step": 3039 + }, + { + "epoch": 0.9586881110059918, + "grad_norm": 0.7578125, + "learning_rate": 0.0002, + "loss": 2.3355, + "step": 3040 + }, + { + "epoch": 0.9590034689372438, + "grad_norm": 0.65625, + "learning_rate": 0.0002, + "loss": 1.9879, + "step": 3041 + }, + { + "epoch": 0.9593188268684958, + "grad_norm": 0.71875, + "learning_rate": 0.0002, + "loss": 2.5144, + "step": 3042 + }, + { + "epoch": 0.9596341847997477, + "grad_norm": 0.69921875, + "learning_rate": 0.0002, + "loss": 2.3263, + "step": 3043 + }, + { + "epoch": 0.9599495427309996, + "grad_norm": 0.83203125, + "learning_rate": 0.0002, + "loss": 2.6343, + "step": 3044 + }, + { + "epoch": 0.9602649006622517, + "grad_norm": 1.015625, + "learning_rate": 0.0002, + "loss": 2.1216, + "step": 3045 + }, + { + "epoch": 0.9605802585935036, + "grad_norm": 0.7421875, + "learning_rate": 0.0002, + "loss": 2.217, + "step": 3046 + }, + { + "epoch": 0.9608956165247556, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 2.2409, + "step": 3047 + }, + { + "epoch": 0.9612109744560076, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 3.0648, + "step": 3048 + }, + { + "epoch": 0.9615263323872595, + "grad_norm": 0.84765625, + "learning_rate": 0.0002, + "loss": 2.3585, + "step": 3049 + }, + { + "epoch": 0.9618416903185115, + "grad_norm": 0.875, + "learning_rate": 0.0002, + "loss": 2.6642, + "step": 3050 + }, + { + "epoch": 0.9621570482497634, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.2217, + "step": 3051 + }, + { + "epoch": 0.9624724061810155, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 0.9815, + "step": 3052 + }, + { + "epoch": 0.9627877641122674, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.0702, + "step": 3053 + }, + { + "epoch": 0.9631031220435194, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.0693, + "step": 3054 + }, + { + "epoch": 0.9634184799747714, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.1086, + "step": 3055 + }, + { + "epoch": 0.9637338379060233, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.1087, + "step": 3056 + }, + { + "epoch": 0.9640491958372753, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 0.9591, + "step": 3057 + }, + { + "epoch": 0.9643645537685273, + "grad_norm": 0.17578125, + "learning_rate": 0.0002, + "loss": 1.0538, + "step": 3058 + }, + { + "epoch": 0.9646799116997793, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 1.063, + "step": 3059 + }, + { + "epoch": 0.9649952696310312, + "grad_norm": 0.189453125, + "learning_rate": 0.0002, + "loss": 1.2636, + "step": 3060 + }, + { + "epoch": 0.9653106275622831, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.2582, + "step": 3061 + }, + { + "epoch": 0.9656259854935352, + "grad_norm": 0.201171875, + "learning_rate": 0.0002, + "loss": 1.2487, + "step": 3062 + }, + { + "epoch": 0.9659413434247871, + "grad_norm": 0.2001953125, + "learning_rate": 0.0002, + "loss": 1.252, + "step": 3063 + }, + { + "epoch": 0.9662567013560391, + "grad_norm": 0.201171875, + "learning_rate": 0.0002, + "loss": 1.2022, + "step": 3064 + }, + { + "epoch": 0.9665720592872911, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.3007, + "step": 3065 + }, + { + "epoch": 0.9668874172185431, + "grad_norm": 0.197265625, + "learning_rate": 0.0002, + "loss": 1.0546, + "step": 3066 + }, + { + "epoch": 0.967202775149795, + "grad_norm": 0.205078125, + "learning_rate": 0.0002, + "loss": 1.0161, + "step": 3067 + }, + { + "epoch": 0.967518133081047, + "grad_norm": 0.2275390625, + "learning_rate": 0.0002, + "loss": 1.2786, + "step": 3068 + }, + { + "epoch": 0.967833491012299, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 3069 + }, + { + "epoch": 0.9681488489435509, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 1.2835, + "step": 3070 + }, + { + "epoch": 0.968464206874803, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 1.2818, + "step": 3071 + }, + { + "epoch": 0.9687795648060549, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 3072 + }, + { + "epoch": 0.9690949227373068, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.3712, + "step": 3073 + }, + { + "epoch": 0.9694102806685588, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.4451, + "step": 3074 + }, + { + "epoch": 0.9697256385998108, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 1.4055, + "step": 3075 + }, + { + "epoch": 0.9700409965310628, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 1.2612, + "step": 3076 + }, + { + "epoch": 0.9703563544623147, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 3077 + }, + { + "epoch": 0.9706717123935666, + "grad_norm": 0.404296875, + "learning_rate": 0.0002, + "loss": 1.5172, + "step": 3078 + }, + { + "epoch": 0.9709870703248187, + "grad_norm": 0.412109375, + "learning_rate": 0.0002, + "loss": 1.4677, + "step": 3079 + }, + { + "epoch": 0.9713024282560706, + "grad_norm": 0.396484375, + "learning_rate": 0.0002, + "loss": 1.6139, + "step": 3080 + }, + { + "epoch": 0.9716177861873226, + "grad_norm": 0.365234375, + "learning_rate": 0.0002, + "loss": 1.5664, + "step": 3081 + }, + { + "epoch": 0.9719331441185746, + "grad_norm": 0.43359375, + "learning_rate": 0.0002, + "loss": 1.7854, + "step": 3082 + }, + { + "epoch": 0.9722485020498266, + "grad_norm": 0.490234375, + "learning_rate": 0.0002, + "loss": 1.8, + "step": 3083 + }, + { + "epoch": 0.9725638599810785, + "grad_norm": 0.453125, + "learning_rate": 0.0002, + "loss": 1.5997, + "step": 3084 + }, + { + "epoch": 0.9728792179123305, + "grad_norm": 0.484375, + "learning_rate": 0.0002, + "loss": 1.8671, + "step": 3085 + }, + { + "epoch": 0.9731945758435825, + "grad_norm": 0.447265625, + "learning_rate": 0.0002, + "loss": 1.5902, + "step": 3086 + }, + { + "epoch": 0.9735099337748344, + "grad_norm": 0.5703125, + "learning_rate": 0.0002, + "loss": 2.0488, + "step": 3087 + }, + { + "epoch": 0.9738252917060864, + "grad_norm": 0.56640625, + "learning_rate": 0.0002, + "loss": 2.1918, + "step": 3088 + }, + { + "epoch": 0.9741406496373384, + "grad_norm": 0.640625, + "learning_rate": 0.0002, + "loss": 2.1455, + "step": 3089 + }, + { + "epoch": 0.9744560075685903, + "grad_norm": 0.75, + "learning_rate": 0.0002, + "loss": 2.2726, + "step": 3090 + }, + { + "epoch": 0.9747713654998423, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 2.4869, + "step": 3091 + }, + { + "epoch": 0.9750867234310943, + "grad_norm": 0.69921875, + "learning_rate": 0.0002, + "loss": 2.2893, + "step": 3092 + }, + { + "epoch": 0.9754020813623463, + "grad_norm": 0.7109375, + "learning_rate": 0.0002, + "loss": 2.2712, + "step": 3093 + }, + { + "epoch": 0.9757174392935982, + "grad_norm": 0.67578125, + "learning_rate": 0.0002, + "loss": 1.9324, + "step": 3094 + }, + { + "epoch": 0.9760327972248503, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 2.5575, + "step": 3095 + }, + { + "epoch": 0.9763481551561022, + "grad_norm": 0.7109375, + "learning_rate": 0.0002, + "loss": 1.9556, + "step": 3096 + }, + { + "epoch": 0.9766635130873541, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 2.1783, + "step": 3097 + }, + { + "epoch": 0.9769788710186061, + "grad_norm": 0.984375, + "learning_rate": 0.0002, + "loss": 2.4076, + "step": 3098 + }, + { + "epoch": 0.9772942289498581, + "grad_norm": 0.765625, + "learning_rate": 0.0002, + "loss": 2.2945, + "step": 3099 + }, + { + "epoch": 0.9776095868811101, + "grad_norm": 1.015625, + "learning_rate": 0.0002, + "loss": 2.5899, + "step": 3100 + }, + { + "epoch": 0.977924944812362, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.0679, + "step": 3101 + }, + { + "epoch": 0.978240302743614, + "grad_norm": 0.1494140625, + "learning_rate": 0.0002, + "loss": 1.1129, + "step": 3102 + }, + { + "epoch": 0.978555660674866, + "grad_norm": 0.1640625, + "learning_rate": 0.0002, + "loss": 1.184, + "step": 3103 + }, + { + "epoch": 0.9788710186061179, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.0578, + "step": 3104 + }, + { + "epoch": 0.97918637653737, + "grad_norm": 0.16015625, + "learning_rate": 0.0002, + "loss": 1.1216, + "step": 3105 + }, + { + "epoch": 0.9795017344686219, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.0779, + "step": 3106 + }, + { + "epoch": 0.9798170923998738, + "grad_norm": 0.1884765625, + "learning_rate": 0.0002, + "loss": 1.2046, + "step": 3107 + }, + { + "epoch": 0.9801324503311258, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.0272, + "step": 3108 + }, + { + "epoch": 0.9804478082623778, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.0669, + "step": 3109 + }, + { + "epoch": 0.9807631661936298, + "grad_norm": 0.1767578125, + "learning_rate": 0.0002, + "loss": 1.0912, + "step": 3110 + }, + { + "epoch": 0.9810785241248817, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 1.1939, + "step": 3111 + }, + { + "epoch": 0.9813938820561338, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.1525, + "step": 3112 + }, + { + "epoch": 0.9817092399873857, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 1.1668, + "step": 3113 + }, + { + "epoch": 0.9820245979186376, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.3732, + "step": 3114 + }, + { + "epoch": 0.9823399558498896, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 1.2564, + "step": 3115 + }, + { + "epoch": 0.9826553137811416, + "grad_norm": 0.2275390625, + "learning_rate": 0.0002, + "loss": 1.136, + "step": 3116 + }, + { + "epoch": 0.9829706717123936, + "grad_norm": 0.220703125, + "learning_rate": 0.0002, + "loss": 1.1366, + "step": 3117 + }, + { + "epoch": 0.9832860296436455, + "grad_norm": 0.232421875, + "learning_rate": 0.0002, + "loss": 1.199, + "step": 3118 + }, + { + "epoch": 0.9836013875748975, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 3119 + }, + { + "epoch": 0.9839167455061495, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.4522, + "step": 3120 + }, + { + "epoch": 0.9842321034374014, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 1.1882, + "step": 3121 + }, + { + "epoch": 0.9845474613686535, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 1.0107, + "step": 3122 + }, + { + "epoch": 0.9848628192999054, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.2732, + "step": 3123 + }, + { + "epoch": 0.9851781772311574, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 1.2273, + "step": 3124 + }, + { + "epoch": 0.9854935351624093, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 3125 + }, + { + "epoch": 0.9858088930936613, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 1.4376, + "step": 3126 + }, + { + "epoch": 0.9861242510249133, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 1.6084, + "step": 3127 + }, + { + "epoch": 0.9864396089561652, + "grad_norm": 0.369140625, + "learning_rate": 0.0002, + "loss": 1.5739, + "step": 3128 + }, + { + "epoch": 0.9867549668874173, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 1.1431, + "step": 3129 + }, + { + "epoch": 0.9870703248186692, + "grad_norm": 0.361328125, + "learning_rate": 0.0002, + "loss": 1.4518, + "step": 3130 + }, + { + "epoch": 0.9873856827499211, + "grad_norm": 0.423828125, + "learning_rate": 0.0002, + "loss": 1.6746, + "step": 3131 + }, + { + "epoch": 0.9877010406811731, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 1.7388, + "step": 3132 + }, + { + "epoch": 0.9880163986124251, + "grad_norm": 0.484375, + "learning_rate": 0.0002, + "loss": 1.9546, + "step": 3133 + }, + { + "epoch": 0.9883317565436771, + "grad_norm": 0.5390625, + "learning_rate": 0.0002, + "loss": 1.9821, + "step": 3134 + }, + { + "epoch": 0.988647114474929, + "grad_norm": 0.56640625, + "learning_rate": 0.0002, + "loss": 2.0593, + "step": 3135 + }, + { + "epoch": 0.9889624724061811, + "grad_norm": 0.5390625, + "learning_rate": 0.0002, + "loss": 2.1689, + "step": 3136 + }, + { + "epoch": 0.989277830337433, + "grad_norm": 0.58203125, + "learning_rate": 0.0002, + "loss": 2.1914, + "step": 3137 + }, + { + "epoch": 0.9895931882686849, + "grad_norm": 0.734375, + "learning_rate": 0.0002, + "loss": 2.1328, + "step": 3138 + }, + { + "epoch": 0.989908546199937, + "grad_norm": 0.73046875, + "learning_rate": 0.0002, + "loss": 2.4486, + "step": 3139 + }, + { + "epoch": 0.9902239041311889, + "grad_norm": 0.6484375, + "learning_rate": 0.0002, + "loss": 2.4008, + "step": 3140 + }, + { + "epoch": 0.9905392620624409, + "grad_norm": 0.62890625, + "learning_rate": 0.0002, + "loss": 2.2022, + "step": 3141 + }, + { + "epoch": 0.9908546199936928, + "grad_norm": 0.73046875, + "learning_rate": 0.0002, + "loss": 2.2837, + "step": 3142 + }, + { + "epoch": 0.9911699779249448, + "grad_norm": 0.75, + "learning_rate": 0.0002, + "loss": 2.2455, + "step": 3143 + }, + { + "epoch": 0.9914853358561968, + "grad_norm": 0.70703125, + "learning_rate": 0.0002, + "loss": 2.2282, + "step": 3144 + }, + { + "epoch": 0.9918006937874487, + "grad_norm": 1.0, + "learning_rate": 0.0002, + "loss": 2.2936, + "step": 3145 + }, + { + "epoch": 0.9921160517187008, + "grad_norm": 0.74609375, + "learning_rate": 0.0002, + "loss": 1.825, + "step": 3146 + }, + { + "epoch": 0.9924314096499527, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 2.3309, + "step": 3147 + }, + { + "epoch": 0.9927467675812046, + "grad_norm": 0.8359375, + "learning_rate": 0.0002, + "loss": 2.5835, + "step": 3148 + }, + { + "epoch": 0.9930621255124567, + "grad_norm": 1.109375, + "learning_rate": 0.0002, + "loss": 2.8331, + "step": 3149 + }, + { + "epoch": 0.9933774834437086, + "grad_norm": 0.87109375, + "learning_rate": 0.0002, + "loss": 2.5929, + "step": 3150 + }, + { + "epoch": 0.9936928413749606, + "grad_norm": 0.1943359375, + "learning_rate": 0.0002, + "loss": 1.1385, + "step": 3151 + }, + { + "epoch": 0.9940081993062125, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 1.0368, + "step": 3152 + }, + { + "epoch": 0.9943235572374646, + "grad_norm": 0.1787109375, + "learning_rate": 0.0002, + "loss": 1.0114, + "step": 3153 + }, + { + "epoch": 0.9946389151687165, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.1246, + "step": 3154 + }, + { + "epoch": 0.9949542730999684, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.1485, + "step": 3155 + }, + { + "epoch": 0.9952696310312205, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.1004, + "step": 3156 + }, + { + "epoch": 0.9955849889624724, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.053, + "step": 3157 + }, + { + "epoch": 0.9959003468937244, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.2631, + "step": 3158 + }, + { + "epoch": 0.9962157048249763, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.1542, + "step": 3159 + }, + { + "epoch": 0.9965310627562283, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 1.3922, + "step": 3160 + }, + { + "epoch": 0.9968464206874803, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 1.2226, + "step": 3161 + }, + { + "epoch": 0.9971617786187322, + "grad_norm": 0.384765625, + "learning_rate": 0.0002, + "loss": 1.5588, + "step": 3162 + }, + { + "epoch": 0.9974771365499843, + "grad_norm": 0.46875, + "learning_rate": 0.0002, + "loss": 1.6681, + "step": 3163 + }, + { + "epoch": 0.9977924944812362, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 1.9082, + "step": 3164 + }, + { + "epoch": 0.9981078524124882, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 2.2168, + "step": 3165 + }, + { + "epoch": 0.9984232103437402, + "grad_norm": 0.734375, + "learning_rate": 0.0002, + "loss": 2.528, + "step": 3166 + }, + { + "epoch": 0.9987385682749921, + "grad_norm": 0.7265625, + "learning_rate": 0.0002, + "loss": 2.272, + "step": 3167 + }, + { + "epoch": 0.9990539262062441, + "grad_norm": 0.64453125, + "learning_rate": 0.0002, + "loss": 2.2394, + "step": 3168 + }, + { + "epoch": 0.999369284137496, + "grad_norm": 1.2109375, + "learning_rate": 0.0002, + "loss": 2.5989, + "step": 3169 + }, + { + "epoch": 0.9996846420687481, + "grad_norm": 0.9921875, + "learning_rate": 0.0002, + "loss": 2.0568, + "step": 3170 + }, + { + "epoch": 1.0, + "grad_norm": 0.734375, + "learning_rate": 0.0002, + "loss": 2.031, + "step": 3171 + }, + { + "epoch": 1.000315357931252, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 0.9765, + "step": 3172 + }, + { + "epoch": 1.0006307158625039, + "grad_norm": 0.177734375, + "learning_rate": 0.0002, + "loss": 1.0741, + "step": 3173 + }, + { + "epoch": 1.0009460737937559, + "grad_norm": 0.1640625, + "learning_rate": 0.0002, + "loss": 1.0415, + "step": 3174 + }, + { + "epoch": 1.001261431725008, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.034, + "step": 3175 + }, + { + "epoch": 1.00157678965626, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.0296, + "step": 3176 + }, + { + "epoch": 1.0018921475875118, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 0.9364, + "step": 3177 + }, + { + "epoch": 1.0022075055187638, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.015, + "step": 3178 + }, + { + "epoch": 1.0025228634500158, + "grad_norm": 0.169921875, + "learning_rate": 0.0002, + "loss": 1.0011, + "step": 3179 + }, + { + "epoch": 1.0025228634500158, + "eval_loss": 1.6409187316894531, + "eval_runtime": 149.1208, + "eval_samples_per_second": 6.706, + "eval_steps_per_second": 6.706, + "step": 3179 + }, + { + "epoch": 1.0025228634500158, + "mmlu_eval_accuracy": 0.6107586333415079, + "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, + "mmlu_eval_accuracy_anatomy": 0.5714285714285714, + "mmlu_eval_accuracy_astronomy": 0.6875, + "mmlu_eval_accuracy_business_ethics": 0.5454545454545454, + "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138, + "mmlu_eval_accuracy_college_biology": 0.875, + "mmlu_eval_accuracy_college_chemistry": 0.375, + "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, + "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365, + "mmlu_eval_accuracy_college_medicine": 0.6363636363636364, + "mmlu_eval_accuracy_college_physics": 0.6363636363636364, + "mmlu_eval_accuracy_computer_security": 0.5454545454545454, + "mmlu_eval_accuracy_conceptual_physics": 0.6538461538461539, + "mmlu_eval_accuracy_econometrics": 0.4166666666666667, + "mmlu_eval_accuracy_electrical_engineering": 0.5, + "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634, + "mmlu_eval_accuracy_formal_logic": 0.2857142857142857, + "mmlu_eval_accuracy_global_facts": 0.3, + "mmlu_eval_accuracy_high_school_biology": 0.65625, + "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365, + "mmlu_eval_accuracy_high_school_computer_science": 0.7777777777777778, + "mmlu_eval_accuracy_high_school_european_history": 0.8888888888888888, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325, + "mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724, + "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, + "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, + "mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333, + "mmlu_eval_accuracy_high_school_statistics": 0.4782608695652174, + "mmlu_eval_accuracy_high_school_us_history": 0.8636363636363636, + "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, + "mmlu_eval_accuracy_human_aging": 0.6521739130434783, + "mmlu_eval_accuracy_human_sexuality": 0.75, + "mmlu_eval_accuracy_international_law": 0.9230769230769231, + "mmlu_eval_accuracy_jurisprudence": 0.36363636363636365, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.36363636363636365, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.84, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7558139534883721, + "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842, + "mmlu_eval_accuracy_moral_scenarios": 0.32, + "mmlu_eval_accuracy_nutrition": 0.696969696969697, + "mmlu_eval_accuracy_philosophy": 0.7352941176470589, + "mmlu_eval_accuracy_prehistory": 0.5714285714285714, + "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194, + "mmlu_eval_accuracy_professional_law": 0.3941176470588235, + "mmlu_eval_accuracy_professional_medicine": 0.7096774193548387, + "mmlu_eval_accuracy_professional_psychology": 0.6666666666666666, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7407407407407407, + "mmlu_eval_accuracy_sociology": 1.0, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.5, + "mmlu_eval_accuracy_world_religions": 0.8421052631578947, + "mmlu_loss": 6.077162632671234, + "step": 3179 + }, + { + "epoch": 1.0028382213812677, + "grad_norm": 0.17578125, + "learning_rate": 0.0002, + "loss": 1.1036, + "step": 3180 + }, + { + "epoch": 1.0031535793125197, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 1.0446, + "step": 3181 + }, + { + "epoch": 1.0034689372437717, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 1.0825, + "step": 3182 + }, + { + "epoch": 1.0037842951750235, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 0.9996, + "step": 3183 + }, + { + "epoch": 1.0040996531062756, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 0.9619, + "step": 3184 + }, + { + "epoch": 1.0044150110375276, + "grad_norm": 0.2265625, + "learning_rate": 0.0002, + "loss": 1.1361, + "step": 3185 + }, + { + "epoch": 1.0047303689687797, + "grad_norm": 0.205078125, + "learning_rate": 0.0002, + "loss": 1.0186, + "step": 3186 + }, + { + "epoch": 1.0050457269000315, + "grad_norm": 0.2265625, + "learning_rate": 0.0002, + "loss": 1.2256, + "step": 3187 + }, + { + "epoch": 1.0053610848312835, + "grad_norm": 0.2216796875, + "learning_rate": 0.0002, + "loss": 1.1717, + "step": 3188 + }, + { + "epoch": 1.0056764427625355, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 1.0088, + "step": 3189 + }, + { + "epoch": 1.0059918006937874, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 0.9969, + "step": 3190 + }, + { + "epoch": 1.0063071586250394, + "grad_norm": 0.2431640625, + "learning_rate": 0.0002, + "loss": 1.1558, + "step": 3191 + }, + { + "epoch": 1.0066225165562914, + "grad_norm": 0.240234375, + "learning_rate": 0.0002, + "loss": 1.0926, + "step": 3192 + }, + { + "epoch": 1.0069378744875435, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.2016, + "step": 3193 + }, + { + "epoch": 1.0072532324187953, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.1969, + "step": 3194 + }, + { + "epoch": 1.0075685903500473, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 1.18, + "step": 3195 + }, + { + "epoch": 1.0078839482812993, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.0992, + "step": 3196 + }, + { + "epoch": 1.0081993062125512, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 1.0648, + "step": 3197 + }, + { + "epoch": 1.0085146641438032, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 3198 + }, + { + "epoch": 1.0088300220750552, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 1.2162, + "step": 3199 + }, + { + "epoch": 1.0091453800063073, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 1.2017, + "step": 3200 + }, + { + "epoch": 1.009460737937559, + "grad_norm": 0.41015625, + "learning_rate": 0.0002, + "loss": 1.1453, + "step": 3201 + }, + { + "epoch": 1.0097760958688111, + "grad_norm": 0.4140625, + "learning_rate": 0.0002, + "loss": 1.1606, + "step": 3202 + }, + { + "epoch": 1.0100914538000632, + "grad_norm": 0.49609375, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 3203 + }, + { + "epoch": 1.010406811731315, + "grad_norm": 0.625, + "learning_rate": 0.0002, + "loss": 1.6904, + "step": 3204 + }, + { + "epoch": 1.010722169662567, + "grad_norm": 0.52734375, + "learning_rate": 0.0002, + "loss": 1.6095, + "step": 3205 + }, + { + "epoch": 1.011037527593819, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 1.5699, + "step": 3206 + }, + { + "epoch": 1.0113528855250709, + "grad_norm": 0.5546875, + "learning_rate": 0.0002, + "loss": 1.5013, + "step": 3207 + }, + { + "epoch": 1.011668243456323, + "grad_norm": 0.73046875, + "learning_rate": 0.0002, + "loss": 2.0994, + "step": 3208 + }, + { + "epoch": 1.011983601387575, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 1.7507, + "step": 3209 + }, + { + "epoch": 1.012298959318827, + "grad_norm": 0.703125, + "learning_rate": 0.0002, + "loss": 1.8871, + "step": 3210 + }, + { + "epoch": 1.0126143172500788, + "grad_norm": 0.62890625, + "learning_rate": 0.0002, + "loss": 1.6135, + "step": 3211 + }, + { + "epoch": 1.0129296751813308, + "grad_norm": 0.84375, + "learning_rate": 0.0002, + "loss": 2.1584, + "step": 3212 + }, + { + "epoch": 1.0132450331125828, + "grad_norm": 0.9140625, + "learning_rate": 0.0002, + "loss": 2.271, + "step": 3213 + }, + { + "epoch": 1.0135603910438347, + "grad_norm": 0.91015625, + "learning_rate": 0.0002, + "loss": 2.2531, + "step": 3214 + }, + { + "epoch": 1.0138757489750867, + "grad_norm": 0.7578125, + "learning_rate": 0.0002, + "loss": 1.8806, + "step": 3215 + }, + { + "epoch": 1.0141911069063387, + "grad_norm": 1.171875, + "learning_rate": 0.0002, + "loss": 2.1231, + "step": 3216 + }, + { + "epoch": 1.0145064648375908, + "grad_norm": 0.703125, + "learning_rate": 0.0002, + "loss": 1.8405, + "step": 3217 + }, + { + "epoch": 1.0148218227688426, + "grad_norm": 0.97265625, + "learning_rate": 0.0002, + "loss": 2.6272, + "step": 3218 + }, + { + "epoch": 1.0151371807000946, + "grad_norm": 0.72265625, + "learning_rate": 0.0002, + "loss": 1.7084, + "step": 3219 + }, + { + "epoch": 1.0154525386313467, + "grad_norm": 0.92578125, + "learning_rate": 0.0002, + "loss": 1.9823, + "step": 3220 + }, + { + "epoch": 1.0157678965625985, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 2.3713, + "step": 3221 + }, + { + "epoch": 1.0160832544938505, + "grad_norm": 0.39453125, + "learning_rate": 0.0002, + "loss": 1.1318, + "step": 3222 + }, + { + "epoch": 1.0163986124251025, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.0224, + "step": 3223 + }, + { + "epoch": 1.0167139703563544, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 0.9838, + "step": 3224 + }, + { + "epoch": 1.0170293282876064, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 1.1018, + "step": 3225 + }, + { + "epoch": 1.0173446862188584, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 1.0557, + "step": 3226 + }, + { + "epoch": 1.0176600441501105, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.091, + "step": 3227 + }, + { + "epoch": 1.0179754020813623, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.0204, + "step": 3228 + }, + { + "epoch": 1.0182907600126143, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.0565, + "step": 3229 + }, + { + "epoch": 1.0186061179438664, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 0.9574, + "step": 3230 + }, + { + "epoch": 1.0189214758751182, + "grad_norm": 0.189453125, + "learning_rate": 0.0002, + "loss": 1.0932, + "step": 3231 + }, + { + "epoch": 1.0192368338063702, + "grad_norm": 0.1865234375, + "learning_rate": 0.0002, + "loss": 1.0356, + "step": 3232 + }, + { + "epoch": 1.0195521917376222, + "grad_norm": 0.19921875, + "learning_rate": 0.0002, + "loss": 1.0859, + "step": 3233 + }, + { + "epoch": 1.0198675496688743, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.0972, + "step": 3234 + }, + { + "epoch": 1.020182907600126, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 0.9797, + "step": 3235 + }, + { + "epoch": 1.0204982655313781, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 0.9034, + "step": 3236 + }, + { + "epoch": 1.0208136234626302, + "grad_norm": 0.2041015625, + "learning_rate": 0.0002, + "loss": 0.988, + "step": 3237 + }, + { + "epoch": 1.021128981393882, + "grad_norm": 0.22265625, + "learning_rate": 0.0002, + "loss": 1.118, + "step": 3238 + }, + { + "epoch": 1.021444339325134, + "grad_norm": 0.216796875, + "learning_rate": 0.0002, + "loss": 1.0407, + "step": 3239 + }, + { + "epoch": 1.021759697256386, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 3240 + }, + { + "epoch": 1.022075055187638, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 1.1088, + "step": 3241 + }, + { + "epoch": 1.02239041311889, + "grad_norm": 0.2392578125, + "learning_rate": 0.0002, + "loss": 0.9923, + "step": 3242 + }, + { + "epoch": 1.022705771050142, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.1666, + "step": 3243 + }, + { + "epoch": 1.023021128981394, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.0586, + "step": 3244 + }, + { + "epoch": 1.0233364869126458, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 1.202, + "step": 3245 + }, + { + "epoch": 1.0236518448438978, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 1.1118, + "step": 3246 + }, + { + "epoch": 1.0239672027751499, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 0.9841, + "step": 3247 + }, + { + "epoch": 1.0242825607064017, + "grad_norm": 0.330078125, + "learning_rate": 0.0002, + "loss": 1.1619, + "step": 3248 + }, + { + "epoch": 1.0245979186376537, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 1.0594, + "step": 3249 + }, + { + "epoch": 1.0249132765689057, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 1.2542, + "step": 3250 + }, + { + "epoch": 1.0252286345001578, + "grad_norm": 0.462890625, + "learning_rate": 0.0002, + "loss": 1.3932, + "step": 3251 + }, + { + "epoch": 1.0255439924314096, + "grad_norm": 0.416015625, + "learning_rate": 0.0002, + "loss": 1.1977, + "step": 3252 + }, + { + "epoch": 1.0258593503626616, + "grad_norm": 0.431640625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 3253 + }, + { + "epoch": 1.0261747082939137, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 1.6964, + "step": 3254 + }, + { + "epoch": 1.0264900662251655, + "grad_norm": 0.51953125, + "learning_rate": 0.0002, + "loss": 1.5015, + "step": 3255 + }, + { + "epoch": 1.0268054241564175, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 1.5653, + "step": 3256 + }, + { + "epoch": 1.0271207820876695, + "grad_norm": 0.60546875, + "learning_rate": 0.0002, + "loss": 1.6891, + "step": 3257 + }, + { + "epoch": 1.0274361400189216, + "grad_norm": 0.74609375, + "learning_rate": 0.0002, + "loss": 1.6393, + "step": 3258 + }, + { + "epoch": 1.0277514979501734, + "grad_norm": 0.6171875, + "learning_rate": 0.0002, + "loss": 1.7416, + "step": 3259 + }, + { + "epoch": 1.0280668558814254, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 1.8768, + "step": 3260 + }, + { + "epoch": 1.0283822138126775, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 2.0677, + "step": 3261 + }, + { + "epoch": 1.0286975717439293, + "grad_norm": 0.68359375, + "learning_rate": 0.0002, + "loss": 1.8406, + "step": 3262 + }, + { + "epoch": 1.0290129296751813, + "grad_norm": 0.75390625, + "learning_rate": 0.0002, + "loss": 2.0157, + "step": 3263 + }, + { + "epoch": 1.0293282876064334, + "grad_norm": 0.76953125, + "learning_rate": 0.0002, + "loss": 1.9768, + "step": 3264 + }, + { + "epoch": 1.0296436455376852, + "grad_norm": 0.9609375, + "learning_rate": 0.0002, + "loss": 2.3366, + "step": 3265 + }, + { + "epoch": 1.0299590034689372, + "grad_norm": 1.6171875, + "learning_rate": 0.0002, + "loss": 2.3527, + "step": 3266 + }, + { + "epoch": 1.0302743614001892, + "grad_norm": 0.73828125, + "learning_rate": 0.0002, + "loss": 1.7338, + "step": 3267 + }, + { + "epoch": 1.0305897193314413, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 2.472, + "step": 3268 + }, + { + "epoch": 1.030905077262693, + "grad_norm": 0.85546875, + "learning_rate": 0.0002, + "loss": 1.7746, + "step": 3269 + }, + { + "epoch": 1.0312204351939451, + "grad_norm": 0.9140625, + "learning_rate": 0.0002, + "loss": 2.0911, + "step": 3270 + }, + { + "epoch": 1.0315357931251972, + "grad_norm": 0.90234375, + "learning_rate": 0.0002, + "loss": 2.0933, + "step": 3271 + }, + { + "epoch": 1.031851151056449, + "grad_norm": 0.40625, + "learning_rate": 0.0002, + "loss": 1.1461, + "step": 3272 + }, + { + "epoch": 1.032166508987701, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.063, + "step": 3273 + }, + { + "epoch": 1.032481866918953, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 0.9726, + "step": 3274 + }, + { + "epoch": 1.032797224850205, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.0599, + "step": 3275 + }, + { + "epoch": 1.033112582781457, + "grad_norm": 0.1884765625, + "learning_rate": 0.0002, + "loss": 1.0833, + "step": 3276 + }, + { + "epoch": 1.033427940712709, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 1.0355, + "step": 3277 + }, + { + "epoch": 1.033743298643961, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.1631, + "step": 3278 + }, + { + "epoch": 1.0340586565752128, + "grad_norm": 0.2177734375, + "learning_rate": 0.0002, + "loss": 1.0762, + "step": 3279 + }, + { + "epoch": 1.0343740145064648, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 1.0874, + "step": 3280 + }, + { + "epoch": 1.0346893724377169, + "grad_norm": 0.2373046875, + "learning_rate": 0.0002, + "loss": 0.9974, + "step": 3281 + }, + { + "epoch": 1.0350047303689687, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 1.0405, + "step": 3282 + }, + { + "epoch": 1.0353200883002207, + "grad_norm": 0.2265625, + "learning_rate": 0.0002, + "loss": 1.148, + "step": 3283 + }, + { + "epoch": 1.0356354462314727, + "grad_norm": 0.2333984375, + "learning_rate": 0.0002, + "loss": 1.1007, + "step": 3284 + }, + { + "epoch": 1.0359508041627248, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 1.1115, + "step": 3285 + }, + { + "epoch": 1.0362661620939766, + "grad_norm": 0.2294921875, + "learning_rate": 0.0002, + "loss": 1.041, + "step": 3286 + }, + { + "epoch": 1.0365815200252286, + "grad_norm": 0.2294921875, + "learning_rate": 0.0002, + "loss": 0.9271, + "step": 3287 + }, + { + "epoch": 1.0368968779564807, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.1956, + "step": 3288 + }, + { + "epoch": 1.0372122358877325, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.0618, + "step": 3289 + }, + { + "epoch": 1.0375275938189845, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 1.0615, + "step": 3290 + }, + { + "epoch": 1.0378429517502366, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.0722, + "step": 3291 + }, + { + "epoch": 1.0381583096814886, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 1.1805, + "step": 3292 + }, + { + "epoch": 1.0384736676127404, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.2132, + "step": 3293 + }, + { + "epoch": 1.0387890255439924, + "grad_norm": 0.353515625, + "learning_rate": 0.0002, + "loss": 1.0142, + "step": 3294 + }, + { + "epoch": 1.0391043834752445, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 1.4408, + "step": 3295 + }, + { + "epoch": 1.0394197414064963, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 1.258, + "step": 3296 + }, + { + "epoch": 1.0397350993377483, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 1.1527, + "step": 3297 + }, + { + "epoch": 1.0400504572690004, + "grad_norm": 0.34765625, + "learning_rate": 0.0002, + "loss": 1.1707, + "step": 3298 + }, + { + "epoch": 1.0403658152002522, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 1.0546, + "step": 3299 + }, + { + "epoch": 1.0406811731315042, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 1.0674, + "step": 3300 + }, + { + "epoch": 1.0409965310627562, + "grad_norm": 0.373046875, + "learning_rate": 0.0002, + "loss": 1.0734, + "step": 3301 + }, + { + "epoch": 1.0413118889940083, + "grad_norm": 0.404296875, + "learning_rate": 0.0002, + "loss": 1.1772, + "step": 3302 + }, + { + "epoch": 1.04162724692526, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 1.4768, + "step": 3303 + }, + { + "epoch": 1.0419426048565121, + "grad_norm": 0.451171875, + "learning_rate": 0.0002, + "loss": 1.3591, + "step": 3304 + }, + { + "epoch": 1.0422579627877642, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 1.5377, + "step": 3305 + }, + { + "epoch": 1.042573320719016, + "grad_norm": 0.546875, + "learning_rate": 0.0002, + "loss": 1.3682, + "step": 3306 + }, + { + "epoch": 1.042888678650268, + "grad_norm": 0.58203125, + "learning_rate": 0.0002, + "loss": 1.6827, + "step": 3307 + }, + { + "epoch": 1.04320403658152, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 1.5953, + "step": 3308 + }, + { + "epoch": 1.043519394512772, + "grad_norm": 0.6796875, + "learning_rate": 0.0002, + "loss": 1.7257, + "step": 3309 + }, + { + "epoch": 1.043834752444024, + "grad_norm": 0.7734375, + "learning_rate": 0.0002, + "loss": 1.9081, + "step": 3310 + }, + { + "epoch": 1.044150110375276, + "grad_norm": 0.78125, + "learning_rate": 0.0002, + "loss": 2.089, + "step": 3311 + }, + { + "epoch": 1.044465468306528, + "grad_norm": 0.734375, + "learning_rate": 0.0002, + "loss": 1.9391, + "step": 3312 + }, + { + "epoch": 1.0447808262377798, + "grad_norm": 0.80859375, + "learning_rate": 0.0002, + "loss": 1.9228, + "step": 3313 + }, + { + "epoch": 1.0450961841690318, + "grad_norm": 0.81640625, + "learning_rate": 0.0002, + "loss": 1.855, + "step": 3314 + }, + { + "epoch": 1.0454115421002839, + "grad_norm": 0.71484375, + "learning_rate": 0.0002, + "loss": 1.8527, + "step": 3315 + }, + { + "epoch": 1.045726900031536, + "grad_norm": 0.890625, + "learning_rate": 0.0002, + "loss": 1.9459, + "step": 3316 + }, + { + "epoch": 1.0460422579627877, + "grad_norm": 0.921875, + "learning_rate": 0.0002, + "loss": 2.1118, + "step": 3317 + }, + { + "epoch": 1.0463576158940397, + "grad_norm": 0.85546875, + "learning_rate": 0.0002, + "loss": 2.1933, + "step": 3318 + }, + { + "epoch": 1.0466729738252918, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 1.4185, + "step": 3319 + }, + { + "epoch": 1.0469883317565436, + "grad_norm": 0.98046875, + "learning_rate": 0.0002, + "loss": 1.8693, + "step": 3320 + }, + { + "epoch": 1.0473036896877956, + "grad_norm": 1.1640625, + "learning_rate": 0.0002, + "loss": 2.1766, + "step": 3321 + }, + { + "epoch": 1.0476190476190477, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.0764, + "step": 3322 + }, + { + "epoch": 1.0479344055502995, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 1.0439, + "step": 3323 + }, + { + "epoch": 1.0482497634815515, + "grad_norm": 0.208984375, + "learning_rate": 0.0002, + "loss": 1.0525, + "step": 3324 + }, + { + "epoch": 1.0485651214128036, + "grad_norm": 0.2041015625, + "learning_rate": 0.0002, + "loss": 1.0798, + "step": 3325 + }, + { + "epoch": 1.0488804793440556, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 1.031, + "step": 3326 + }, + { + "epoch": 1.0491958372753074, + "grad_norm": 0.18359375, + "learning_rate": 0.0002, + "loss": 1.0484, + "step": 3327 + }, + { + "epoch": 1.0495111952065594, + "grad_norm": 0.1943359375, + "learning_rate": 0.0002, + "loss": 1.0821, + "step": 3328 + }, + { + "epoch": 1.0498265531378115, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.0097, + "step": 3329 + }, + { + "epoch": 1.0501419110690633, + "grad_norm": 0.232421875, + "learning_rate": 0.0002, + "loss": 1.1195, + "step": 3330 + }, + { + "epoch": 1.0504572690003153, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 1.147, + "step": 3331 + }, + { + "epoch": 1.0507726269315674, + "grad_norm": 0.193359375, + "learning_rate": 0.0002, + "loss": 1.0483, + "step": 3332 + }, + { + "epoch": 1.0510879848628194, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 0.9942, + "step": 3333 + }, + { + "epoch": 1.0514033427940712, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 0.992, + "step": 3334 + }, + { + "epoch": 1.0517187007253233, + "grad_norm": 0.2275390625, + "learning_rate": 0.0002, + "loss": 0.962, + "step": 3335 + }, + { + "epoch": 1.0520340586565753, + "grad_norm": 0.2216796875, + "learning_rate": 0.0002, + "loss": 1.0078, + "step": 3336 + }, + { + "epoch": 1.052349416587827, + "grad_norm": 0.224609375, + "learning_rate": 0.0002, + "loss": 1.0204, + "step": 3337 + }, + { + "epoch": 1.0526647745190791, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 1.1656, + "step": 3338 + }, + { + "epoch": 1.0529801324503312, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.127, + "step": 3339 + }, + { + "epoch": 1.0532954903815832, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.2622, + "step": 3340 + }, + { + "epoch": 1.053610848312835, + "grad_norm": 0.353515625, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 3341 + }, + { + "epoch": 1.053926206244087, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 1.1195, + "step": 3342 + }, + { + "epoch": 1.054241564175339, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 1.0641, + "step": 3343 + }, + { + "epoch": 1.054556922106591, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 0.9521, + "step": 3344 + }, + { + "epoch": 1.054872280037843, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 1.1428, + "step": 3345 + }, + { + "epoch": 1.055187637969095, + "grad_norm": 0.384765625, + "learning_rate": 0.0002, + "loss": 1.2688, + "step": 3346 + }, + { + "epoch": 1.0555029959003468, + "grad_norm": 0.396484375, + "learning_rate": 0.0002, + "loss": 1.1246, + "step": 3347 + }, + { + "epoch": 1.0558183538315988, + "grad_norm": 0.400390625, + "learning_rate": 0.0002, + "loss": 1.1884, + "step": 3348 + }, + { + "epoch": 1.0561337117628509, + "grad_norm": 0.41015625, + "learning_rate": 0.0002, + "loss": 1.3682, + "step": 3349 + }, + { + "epoch": 1.056449069694103, + "grad_norm": 0.447265625, + "learning_rate": 0.0002, + "loss": 1.2719, + "step": 3350 + }, + { + "epoch": 1.0567644276253547, + "grad_norm": 0.4765625, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 3351 + }, + { + "epoch": 1.0570797855566068, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 1.2916, + "step": 3352 + }, + { + "epoch": 1.0573951434878588, + "grad_norm": 0.58203125, + "learning_rate": 0.0002, + "loss": 1.4972, + "step": 3353 + }, + { + "epoch": 1.0577105014191106, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 1.4467, + "step": 3354 + }, + { + "epoch": 1.0580258593503626, + "grad_norm": 0.609375, + "learning_rate": 0.0002, + "loss": 1.6334, + "step": 3355 + }, + { + "epoch": 1.0583412172816147, + "grad_norm": 0.69921875, + "learning_rate": 0.0002, + "loss": 1.5164, + "step": 3356 + }, + { + "epoch": 1.0586565752128667, + "grad_norm": 0.76171875, + "learning_rate": 0.0002, + "loss": 2.007, + "step": 3357 + }, + { + "epoch": 1.0589719331441185, + "grad_norm": 0.81640625, + "learning_rate": 0.0002, + "loss": 1.9422, + "step": 3358 + }, + { + "epoch": 1.0592872910753706, + "grad_norm": 0.78125, + "learning_rate": 0.0002, + "loss": 2.1784, + "step": 3359 + }, + { + "epoch": 1.0596026490066226, + "grad_norm": 0.7109375, + "learning_rate": 0.0002, + "loss": 1.8977, + "step": 3360 + }, + { + "epoch": 1.0599180069378744, + "grad_norm": 0.98046875, + "learning_rate": 0.0002, + "loss": 2.0118, + "step": 3361 + }, + { + "epoch": 1.0602333648691264, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 1.804, + "step": 3362 + }, + { + "epoch": 1.0605487228003785, + "grad_norm": 0.9921875, + "learning_rate": 0.0002, + "loss": 1.9611, + "step": 3363 + }, + { + "epoch": 1.0608640807316303, + "grad_norm": 0.84375, + "learning_rate": 0.0002, + "loss": 2.2264, + "step": 3364 + }, + { + "epoch": 1.0611794386628823, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 2.016, + "step": 3365 + }, + { + "epoch": 1.0614947965941344, + "grad_norm": 1.25, + "learning_rate": 0.0002, + "loss": 2.5839, + "step": 3366 + }, + { + "epoch": 1.0614947965941344, + "eval_loss": 1.7018221616744995, + "eval_runtime": 152.4159, + "eval_samples_per_second": 6.561, + "eval_steps_per_second": 6.561, + "step": 3366 + }, + { + "epoch": 1.0614947965941344, + "mmlu_eval_accuracy": 0.6071666290506706, + "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, + "mmlu_eval_accuracy_anatomy": 0.5, + "mmlu_eval_accuracy_astronomy": 0.6875, + "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, + "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034, + "mmlu_eval_accuracy_college_biology": 0.875, + "mmlu_eval_accuracy_college_chemistry": 0.375, + "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, + "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454, + "mmlu_eval_accuracy_college_medicine": 0.5909090909090909, + "mmlu_eval_accuracy_college_physics": 0.5454545454545454, + "mmlu_eval_accuracy_computer_security": 0.5454545454545454, + "mmlu_eval_accuracy_conceptual_physics": 0.5769230769230769, + "mmlu_eval_accuracy_econometrics": 0.4166666666666667, + "mmlu_eval_accuracy_electrical_engineering": 0.5, + "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244, + "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, + "mmlu_eval_accuracy_global_facts": 0.2, + "mmlu_eval_accuracy_high_school_biology": 0.65625, + "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453, + "mmlu_eval_accuracy_high_school_computer_science": 0.7777777777777778, + "mmlu_eval_accuracy_high_school_european_history": 0.8888888888888888, + "mmlu_eval_accuracy_high_school_geography": 0.9545454545454546, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.4883720930232558, + "mmlu_eval_accuracy_high_school_mathematics": 0.3793103448275862, + "mmlu_eval_accuracy_high_school_microeconomics": 0.5, + "mmlu_eval_accuracy_high_school_physics": 0.058823529411764705, + "mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667, + "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173, + "mmlu_eval_accuracy_high_school_us_history": 0.8636363636363636, + "mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693, + "mmlu_eval_accuracy_human_aging": 0.6521739130434783, + "mmlu_eval_accuracy_human_sexuality": 0.6666666666666666, + "mmlu_eval_accuracy_international_law": 0.9230769230769231, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, + "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.8, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7093023255813954, + "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842, + "mmlu_eval_accuracy_moral_scenarios": 0.38, + "mmlu_eval_accuracy_nutrition": 0.7272727272727273, + "mmlu_eval_accuracy_philosophy": 0.7058823529411765, + "mmlu_eval_accuracy_prehistory": 0.6285714285714286, + "mmlu_eval_accuracy_professional_accounting": 0.41935483870967744, + "mmlu_eval_accuracy_professional_law": 0.36470588235294116, + "mmlu_eval_accuracy_professional_medicine": 0.7096774193548387, + "mmlu_eval_accuracy_professional_psychology": 0.6666666666666666, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7407407407407407, + "mmlu_eval_accuracy_sociology": 1.0, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.4444444444444444, + "mmlu_eval_accuracy_world_religions": 0.8947368421052632, + "mmlu_loss": 6.191603247283099, + "step": 3366 + }, + { + "epoch": 1.0618101545253864, + "grad_norm": 0.9609375, + "learning_rate": 0.0002, + "loss": 2.1898, + "step": 3367 + }, + { + "epoch": 1.0621255124566382, + "grad_norm": 0.90234375, + "learning_rate": 0.0002, + "loss": 1.8509, + "step": 3368 + }, + { + "epoch": 1.0624408703878903, + "grad_norm": 0.828125, + "learning_rate": 0.0002, + "loss": 1.71, + "step": 3369 + }, + { + "epoch": 1.0627562283191423, + "grad_norm": 0.9453125, + "learning_rate": 0.0002, + "loss": 1.758, + "step": 3370 + }, + { + "epoch": 1.063071586250394, + "grad_norm": 0.96484375, + "learning_rate": 0.0002, + "loss": 2.0631, + "step": 3371 + }, + { + "epoch": 1.0633869441816461, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.1022, + "step": 3372 + }, + { + "epoch": 1.0637023021128982, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 1.0294, + "step": 3373 + }, + { + "epoch": 1.0640176600441502, + "grad_norm": 0.220703125, + "learning_rate": 0.0002, + "loss": 1.0234, + "step": 3374 + }, + { + "epoch": 1.064333017975402, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.15, + "step": 3375 + }, + { + "epoch": 1.064648375906654, + "grad_norm": 0.2041015625, + "learning_rate": 0.0002, + "loss": 0.9549, + "step": 3376 + }, + { + "epoch": 1.064963733837906, + "grad_norm": 0.2041015625, + "learning_rate": 0.0002, + "loss": 1.2107, + "step": 3377 + }, + { + "epoch": 1.065279091769158, + "grad_norm": 0.212890625, + "learning_rate": 0.0002, + "loss": 0.996, + "step": 3378 + }, + { + "epoch": 1.06559444970041, + "grad_norm": 0.19140625, + "learning_rate": 0.0002, + "loss": 1.1096, + "step": 3379 + }, + { + "epoch": 1.065909807631662, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 1.0655, + "step": 3380 + }, + { + "epoch": 1.0662251655629138, + "grad_norm": 0.2001953125, + "learning_rate": 0.0002, + "loss": 0.9495, + "step": 3381 + }, + { + "epoch": 1.0665405234941658, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 1.0846, + "step": 3382 + }, + { + "epoch": 1.0668558814254179, + "grad_norm": 0.19921875, + "learning_rate": 0.0002, + "loss": 0.9858, + "step": 3383 + }, + { + "epoch": 1.06717123935667, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.0471, + "step": 3384 + }, + { + "epoch": 1.0674865972879217, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.1138, + "step": 3385 + }, + { + "epoch": 1.0678019552191738, + "grad_norm": 0.220703125, + "learning_rate": 0.0002, + "loss": 1.1094, + "step": 3386 + }, + { + "epoch": 1.0681173131504258, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.0618, + "step": 3387 + }, + { + "epoch": 1.0684326710816776, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 1.0364, + "step": 3388 + }, + { + "epoch": 1.0687480290129296, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.1614, + "step": 3389 + }, + { + "epoch": 1.0690633869441817, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.0864, + "step": 3390 + }, + { + "epoch": 1.0693787448754337, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.3653, + "step": 3391 + }, + { + "epoch": 1.0696941028066855, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 3392 + }, + { + "epoch": 1.0700094607379376, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 1.2158, + "step": 3393 + }, + { + "epoch": 1.0703248186691896, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 1.1373, + "step": 3394 + }, + { + "epoch": 1.0706401766004414, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 1.1338, + "step": 3395 + }, + { + "epoch": 1.0709555345316935, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 1.2667, + "step": 3396 + }, + { + "epoch": 1.0712708924629455, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 1.3523, + "step": 3397 + }, + { + "epoch": 1.0715862503941973, + "grad_norm": 0.34765625, + "learning_rate": 0.0002, + "loss": 1.0518, + "step": 3398 + }, + { + "epoch": 1.0719016083254493, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 1.0794, + "step": 3399 + }, + { + "epoch": 1.0722169662567014, + "grad_norm": 0.34765625, + "learning_rate": 0.0002, + "loss": 1.1878, + "step": 3400 + }, + { + "epoch": 1.0725323241879534, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 1.0949, + "step": 3401 + }, + { + "epoch": 1.0728476821192052, + "grad_norm": 0.44921875, + "learning_rate": 0.0002, + "loss": 1.2915, + "step": 3402 + }, + { + "epoch": 1.0731630400504573, + "grad_norm": 0.4453125, + "learning_rate": 0.0002, + "loss": 1.2027, + "step": 3403 + }, + { + "epoch": 1.0734783979817093, + "grad_norm": 0.498046875, + "learning_rate": 0.0002, + "loss": 1.4147, + "step": 3404 + }, + { + "epoch": 1.0737937559129611, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 1.4798, + "step": 3405 + }, + { + "epoch": 1.0741091138442131, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 1.5196, + "step": 3406 + }, + { + "epoch": 1.0744244717754652, + "grad_norm": 0.671875, + "learning_rate": 0.0002, + "loss": 1.4815, + "step": 3407 + }, + { + "epoch": 1.0747398297067172, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 1.5522, + "step": 3408 + }, + { + "epoch": 1.075055187637969, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 1.598, + "step": 3409 + }, + { + "epoch": 1.075370545569221, + "grad_norm": 0.84765625, + "learning_rate": 0.0002, + "loss": 2.2696, + "step": 3410 + }, + { + "epoch": 1.075685903500473, + "grad_norm": 0.71875, + "learning_rate": 0.0002, + "loss": 1.7457, + "step": 3411 + }, + { + "epoch": 1.076001261431725, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 2.3345, + "step": 3412 + }, + { + "epoch": 1.076316619362977, + "grad_norm": 0.796875, + "learning_rate": 0.0002, + "loss": 1.9934, + "step": 3413 + }, + { + "epoch": 1.076631977294229, + "grad_norm": 0.87109375, + "learning_rate": 0.0002, + "loss": 2.2436, + "step": 3414 + }, + { + "epoch": 1.0769473352254808, + "grad_norm": 0.87109375, + "learning_rate": 0.0002, + "loss": 2.0004, + "step": 3415 + }, + { + "epoch": 1.0772626931567328, + "grad_norm": 0.8515625, + "learning_rate": 0.0002, + "loss": 1.9103, + "step": 3416 + }, + { + "epoch": 1.0775780510879849, + "grad_norm": 1.2890625, + "learning_rate": 0.0002, + "loss": 1.6948, + "step": 3417 + }, + { + "epoch": 1.077893409019237, + "grad_norm": 0.984375, + "learning_rate": 0.0002, + "loss": 1.9969, + "step": 3418 + }, + { + "epoch": 1.0782087669504887, + "grad_norm": 1.59375, + "learning_rate": 0.0002, + "loss": 2.2548, + "step": 3419 + }, + { + "epoch": 1.0785241248817408, + "grad_norm": 0.94921875, + "learning_rate": 0.0002, + "loss": 1.8859, + "step": 3420 + }, + { + "epoch": 1.0788394828129928, + "grad_norm": 1.7265625, + "learning_rate": 0.0002, + "loss": 2.3352, + "step": 3421 + }, + { + "epoch": 1.0791548407442446, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 1.228, + "step": 3422 + }, + { + "epoch": 1.0794701986754967, + "grad_norm": 0.255859375, + "learning_rate": 0.0002, + "loss": 1.1252, + "step": 3423 + }, + { + "epoch": 1.0797855566067487, + "grad_norm": 0.23828125, + "learning_rate": 0.0002, + "loss": 1.0433, + "step": 3424 + }, + { + "epoch": 1.0801009145380007, + "grad_norm": 0.1943359375, + "learning_rate": 0.0002, + "loss": 0.9955, + "step": 3425 + }, + { + "epoch": 1.0804162724692525, + "grad_norm": 0.1884765625, + "learning_rate": 0.0002, + "loss": 1.0615, + "step": 3426 + }, + { + "epoch": 1.0807316304005046, + "grad_norm": 0.169921875, + "learning_rate": 0.0002, + "loss": 0.8747, + "step": 3427 + }, + { + "epoch": 1.0810469883317566, + "grad_norm": 0.19921875, + "learning_rate": 0.0002, + "loss": 1.0658, + "step": 3428 + }, + { + "epoch": 1.0813623462630084, + "grad_norm": 0.189453125, + "learning_rate": 0.0002, + "loss": 1.0488, + "step": 3429 + }, + { + "epoch": 1.0816777041942605, + "grad_norm": 0.1796875, + "learning_rate": 0.0002, + "loss": 1.0475, + "step": 3430 + }, + { + "epoch": 1.0819930621255125, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.0756, + "step": 3431 + }, + { + "epoch": 1.0823084200567645, + "grad_norm": 0.19140625, + "learning_rate": 0.0002, + "loss": 1.0041, + "step": 3432 + }, + { + "epoch": 1.0826237779880163, + "grad_norm": 0.205078125, + "learning_rate": 0.0002, + "loss": 1.0647, + "step": 3433 + }, + { + "epoch": 1.0829391359192684, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 1.1024, + "step": 3434 + }, + { + "epoch": 1.0832544938505204, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.0381, + "step": 3435 + }, + { + "epoch": 1.0835698517817722, + "grad_norm": 0.22265625, + "learning_rate": 0.0002, + "loss": 1.0121, + "step": 3436 + }, + { + "epoch": 1.0838852097130243, + "grad_norm": 0.208984375, + "learning_rate": 0.0002, + "loss": 1.0794, + "step": 3437 + }, + { + "epoch": 1.0842005676442763, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 0.9804, + "step": 3438 + }, + { + "epoch": 1.0845159255755283, + "grad_norm": 0.2373046875, + "learning_rate": 0.0002, + "loss": 1.1269, + "step": 3439 + }, + { + "epoch": 1.0848312835067802, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.193, + "step": 3440 + }, + { + "epoch": 1.0851466414380322, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 1.1776, + "step": 3441 + }, + { + "epoch": 1.0854619993692842, + "grad_norm": 0.2490234375, + "learning_rate": 0.0002, + "loss": 1.0061, + "step": 3442 + }, + { + "epoch": 1.085777357300536, + "grad_norm": 0.2451171875, + "learning_rate": 0.0002, + "loss": 1.1413, + "step": 3443 + }, + { + "epoch": 1.086092715231788, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 1.0726, + "step": 3444 + }, + { + "epoch": 1.0864080731630401, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.1652, + "step": 3445 + }, + { + "epoch": 1.086723431094292, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 1.0687, + "step": 3446 + }, + { + "epoch": 1.087038789025544, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 1.174, + "step": 3447 + }, + { + "epoch": 1.087354146956796, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 1.2713, + "step": 3448 + }, + { + "epoch": 1.087669504888048, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 1.0913, + "step": 3449 + }, + { + "epoch": 1.0879848628192998, + "grad_norm": 0.416015625, + "learning_rate": 0.0002, + "loss": 1.4997, + "step": 3450 + }, + { + "epoch": 1.0883002207505519, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 1.0955, + "step": 3451 + }, + { + "epoch": 1.088615578681804, + "grad_norm": 0.466796875, + "learning_rate": 0.0002, + "loss": 1.2778, + "step": 3452 + }, + { + "epoch": 1.0889309366130557, + "grad_norm": 0.431640625, + "learning_rate": 0.0002, + "loss": 1.373, + "step": 3453 + }, + { + "epoch": 1.0892462945443078, + "grad_norm": 0.419921875, + "learning_rate": 0.0002, + "loss": 1.3942, + "step": 3454 + }, + { + "epoch": 1.0895616524755598, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 1.4856, + "step": 3455 + }, + { + "epoch": 1.0898770104068118, + "grad_norm": 0.515625, + "learning_rate": 0.0002, + "loss": 1.3681, + "step": 3456 + }, + { + "epoch": 1.0901923683380637, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 3457 + }, + { + "epoch": 1.0905077262693157, + "grad_norm": 0.5390625, + "learning_rate": 0.0002, + "loss": 1.3904, + "step": 3458 + }, + { + "epoch": 1.0908230842005677, + "grad_norm": 0.60546875, + "learning_rate": 0.0002, + "loss": 1.6208, + "step": 3459 + }, + { + "epoch": 1.0911384421318195, + "grad_norm": 0.703125, + "learning_rate": 0.0002, + "loss": 1.6595, + "step": 3460 + }, + { + "epoch": 1.0914538000630716, + "grad_norm": 0.64453125, + "learning_rate": 0.0002, + "loss": 1.5849, + "step": 3461 + }, + { + "epoch": 1.0917691579943236, + "grad_norm": 0.71484375, + "learning_rate": 0.0002, + "loss": 1.5842, + "step": 3462 + }, + { + "epoch": 1.0920845159255754, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 2.1692, + "step": 3463 + }, + { + "epoch": 1.0923998738568275, + "grad_norm": 0.9609375, + "learning_rate": 0.0002, + "loss": 1.9786, + "step": 3464 + }, + { + "epoch": 1.0927152317880795, + "grad_norm": 0.84765625, + "learning_rate": 0.0002, + "loss": 1.8721, + "step": 3465 + }, + { + "epoch": 1.0930305897193315, + "grad_norm": 0.74609375, + "learning_rate": 0.0002, + "loss": 1.6533, + "step": 3466 + }, + { + "epoch": 1.0933459476505833, + "grad_norm": 0.921875, + "learning_rate": 0.0002, + "loss": 1.8377, + "step": 3467 + }, + { + "epoch": 1.0936613055818354, + "grad_norm": 0.89453125, + "learning_rate": 0.0002, + "loss": 2.1695, + "step": 3468 + }, + { + "epoch": 1.0939766635130874, + "grad_norm": 1.234375, + "learning_rate": 0.0002, + "loss": 2.2117, + "step": 3469 + }, + { + "epoch": 1.0942920214443392, + "grad_norm": 0.73046875, + "learning_rate": 0.0002, + "loss": 1.5814, + "step": 3470 + }, + { + "epoch": 1.0946073793755913, + "grad_norm": 1.1171875, + "learning_rate": 0.0002, + "loss": 2.1208, + "step": 3471 + }, + { + "epoch": 1.0949227373068433, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.0577, + "step": 3472 + }, + { + "epoch": 1.0952380952380953, + "grad_norm": 0.22265625, + "learning_rate": 0.0002, + "loss": 1.0196, + "step": 3473 + }, + { + "epoch": 1.0955534531693472, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.0333, + "step": 3474 + }, + { + "epoch": 1.0958688111005992, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 1.0437, + "step": 3475 + }, + { + "epoch": 1.0961841690318512, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 1.0821, + "step": 3476 + }, + { + "epoch": 1.096499526963103, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 1.0372, + "step": 3477 + }, + { + "epoch": 1.096814884894355, + "grad_norm": 0.1875, + "learning_rate": 0.0002, + "loss": 1.0016, + "step": 3478 + }, + { + "epoch": 1.0971302428256071, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 1.0842, + "step": 3479 + }, + { + "epoch": 1.097445600756859, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 1.0134, + "step": 3480 + }, + { + "epoch": 1.097760958688111, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.1215, + "step": 3481 + }, + { + "epoch": 1.098076316619363, + "grad_norm": 0.197265625, + "learning_rate": 0.0002, + "loss": 1.129, + "step": 3482 + }, + { + "epoch": 1.098391674550615, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 0.9961, + "step": 3483 + }, + { + "epoch": 1.0987070324818669, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 1.0414, + "step": 3484 + }, + { + "epoch": 1.0990223904131189, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.0858, + "step": 3485 + }, + { + "epoch": 1.099337748344371, + "grad_norm": 0.216796875, + "learning_rate": 0.0002, + "loss": 1.1193, + "step": 3486 + }, + { + "epoch": 1.0996531062756227, + "grad_norm": 0.244140625, + "learning_rate": 0.0002, + "loss": 1.0282, + "step": 3487 + }, + { + "epoch": 1.0999684642068748, + "grad_norm": 0.2470703125, + "learning_rate": 0.0002, + "loss": 1.1491, + "step": 3488 + }, + { + "epoch": 1.1002838221381268, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 1.1899, + "step": 3489 + }, + { + "epoch": 1.1005991800693788, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 0.9446, + "step": 3490 + }, + { + "epoch": 1.1009145380006307, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.1525, + "step": 3491 + }, + { + "epoch": 1.1012298959318827, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.1733, + "step": 3492 + }, + { + "epoch": 1.1015452538631347, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.0418, + "step": 3493 + }, + { + "epoch": 1.1018606117943865, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 1.2648, + "step": 3494 + }, + { + "epoch": 1.1021759697256386, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 1.2584, + "step": 3495 + }, + { + "epoch": 1.1024913276568906, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.9456, + "step": 3496 + }, + { + "epoch": 1.1028066855881424, + "grad_norm": 0.408203125, + "learning_rate": 0.0002, + "loss": 1.1921, + "step": 3497 + }, + { + "epoch": 1.1031220435193945, + "grad_norm": 0.361328125, + "learning_rate": 0.0002, + "loss": 1.1387, + "step": 3498 + }, + { + "epoch": 1.1034374014506465, + "grad_norm": 0.40625, + "learning_rate": 0.0002, + "loss": 1.2239, + "step": 3499 + }, + { + "epoch": 1.1037527593818985, + "grad_norm": 0.404296875, + "learning_rate": 0.0002, + "loss": 1.1509, + "step": 3500 + }, + { + "epoch": 1.1040681173131504, + "grad_norm": 0.390625, + "learning_rate": 0.0002, + "loss": 1.2274, + "step": 3501 + }, + { + "epoch": 1.1043834752444024, + "grad_norm": 0.46875, + "learning_rate": 0.0002, + "loss": 1.2212, + "step": 3502 + }, + { + "epoch": 1.1046988331756544, + "grad_norm": 0.5390625, + "learning_rate": 0.0002, + "loss": 1.456, + "step": 3503 + }, + { + "epoch": 1.1050141911069062, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 3504 + }, + { + "epoch": 1.1053295490381583, + "grad_norm": 0.5703125, + "learning_rate": 0.0002, + "loss": 1.5601, + "step": 3505 + }, + { + "epoch": 1.1056449069694103, + "grad_norm": 0.79296875, + "learning_rate": 0.0002, + "loss": 1.6284, + "step": 3506 + }, + { + "epoch": 1.1059602649006623, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 1.6014, + "step": 3507 + }, + { + "epoch": 1.1062756228319142, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 1.7176, + "step": 3508 + }, + { + "epoch": 1.1065909807631662, + "grad_norm": 0.68359375, + "learning_rate": 0.0002, + "loss": 1.8248, + "step": 3509 + }, + { + "epoch": 1.1069063386944182, + "grad_norm": 0.72265625, + "learning_rate": 0.0002, + "loss": 1.7601, + "step": 3510 + }, + { + "epoch": 1.10722169662567, + "grad_norm": 0.70703125, + "learning_rate": 0.0002, + "loss": 1.8524, + "step": 3511 + }, + { + "epoch": 1.107537054556922, + "grad_norm": 0.76171875, + "learning_rate": 0.0002, + "loss": 1.6578, + "step": 3512 + }, + { + "epoch": 1.1078524124881741, + "grad_norm": 0.703125, + "learning_rate": 0.0002, + "loss": 1.6774, + "step": 3513 + }, + { + "epoch": 1.108167770419426, + "grad_norm": 0.91796875, + "learning_rate": 0.0002, + "loss": 2.2785, + "step": 3514 + }, + { + "epoch": 1.108483128350678, + "grad_norm": 0.8046875, + "learning_rate": 0.0002, + "loss": 1.9546, + "step": 3515 + }, + { + "epoch": 1.10879848628193, + "grad_norm": 0.859375, + "learning_rate": 0.0002, + "loss": 2.0805, + "step": 3516 + }, + { + "epoch": 1.109113844213182, + "grad_norm": 0.83203125, + "learning_rate": 0.0002, + "loss": 2.0662, + "step": 3517 + }, + { + "epoch": 1.1094292021444339, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 1.9037, + "step": 3518 + }, + { + "epoch": 1.109744560075686, + "grad_norm": 0.99609375, + "learning_rate": 0.0002, + "loss": 1.8515, + "step": 3519 + }, + { + "epoch": 1.110059918006938, + "grad_norm": 0.953125, + "learning_rate": 0.0002, + "loss": 1.9306, + "step": 3520 + }, + { + "epoch": 1.1103752759381897, + "grad_norm": 1.4921875, + "learning_rate": 0.0002, + "loss": 2.4647, + "step": 3521 + }, + { + "epoch": 1.1106906338694418, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 1.1039, + "step": 3522 + }, + { + "epoch": 1.1110059918006938, + "grad_norm": 0.2373046875, + "learning_rate": 0.0002, + "loss": 1.0703, + "step": 3523 + }, + { + "epoch": 1.1113213497319459, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.0329, + "step": 3524 + }, + { + "epoch": 1.1116367076631977, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.0711, + "step": 3525 + }, + { + "epoch": 1.1119520655944497, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.0265, + "step": 3526 + }, + { + "epoch": 1.1122674235257017, + "grad_norm": 0.2119140625, + "learning_rate": 0.0002, + "loss": 0.9536, + "step": 3527 + }, + { + "epoch": 1.1125827814569536, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.0568, + "step": 3528 + }, + { + "epoch": 1.1128981393882056, + "grad_norm": 0.1884765625, + "learning_rate": 0.0002, + "loss": 0.9676, + "step": 3529 + }, + { + "epoch": 1.1132134973194576, + "grad_norm": 0.201171875, + "learning_rate": 0.0002, + "loss": 1.0826, + "step": 3530 + }, + { + "epoch": 1.1135288552507097, + "grad_norm": 0.2314453125, + "learning_rate": 0.0002, + "loss": 1.0603, + "step": 3531 + }, + { + "epoch": 1.1138442131819615, + "grad_norm": 0.19921875, + "learning_rate": 0.0002, + "loss": 0.9961, + "step": 3532 + }, + { + "epoch": 1.1141595711132135, + "grad_norm": 0.2177734375, + "learning_rate": 0.0002, + "loss": 0.9879, + "step": 3533 + }, + { + "epoch": 1.1144749290444655, + "grad_norm": 0.2392578125, + "learning_rate": 0.0002, + "loss": 1.0611, + "step": 3534 + }, + { + "epoch": 1.1147902869757174, + "grad_norm": 0.2275390625, + "learning_rate": 0.0002, + "loss": 1.1257, + "step": 3535 + }, + { + "epoch": 1.1151056449069694, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 0.9764, + "step": 3536 + }, + { + "epoch": 1.1154210028382214, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 3537 + }, + { + "epoch": 1.1157363607694735, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.0587, + "step": 3538 + }, + { + "epoch": 1.1160517187007253, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 1.1714, + "step": 3539 + }, + { + "epoch": 1.1163670766319773, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 0.9208, + "step": 3540 + }, + { + "epoch": 1.1166824345632294, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 1.0494, + "step": 3541 + }, + { + "epoch": 1.1169977924944812, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 1.208, + "step": 3542 + }, + { + "epoch": 1.1173131504257332, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 1.1427, + "step": 3543 + }, + { + "epoch": 1.1176285083569852, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 1.1009, + "step": 3544 + }, + { + "epoch": 1.117943866288237, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 1.0787, + "step": 3545 + }, + { + "epoch": 1.118259224219489, + "grad_norm": 0.462890625, + "learning_rate": 0.0002, + "loss": 1.1856, + "step": 3546 + }, + { + "epoch": 1.1185745821507411, + "grad_norm": 0.41796875, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 3547 + }, + { + "epoch": 1.1188899400819932, + "grad_norm": 0.400390625, + "learning_rate": 0.0002, + "loss": 1.1524, + "step": 3548 + }, + { + "epoch": 1.119205298013245, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 1.1796, + "step": 3549 + }, + { + "epoch": 1.119520655944497, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 1.1389, + "step": 3550 + }, + { + "epoch": 1.119836013875749, + "grad_norm": 0.40625, + "learning_rate": 0.0002, + "loss": 1.1711, + "step": 3551 + }, + { + "epoch": 1.1201513718070009, + "grad_norm": 0.400390625, + "learning_rate": 0.0002, + "loss": 1.1718, + "step": 3552 + }, + { + "epoch": 1.120466729738253, + "grad_norm": 0.49609375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 3553 + }, + { + "epoch": 1.120466729738253, + "eval_loss": 1.6859415769577026, + "eval_runtime": 148.9301, + "eval_samples_per_second": 6.715, + "eval_steps_per_second": 6.715, + "step": 3553 + }, + { + "epoch": 1.120466729738253, + "mmlu_eval_accuracy": 0.6147797729855283, + "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, + "mmlu_eval_accuracy_anatomy": 0.5714285714285714, + "mmlu_eval_accuracy_astronomy": 0.625, + "mmlu_eval_accuracy_business_ethics": 0.8181818181818182, + "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034, + "mmlu_eval_accuracy_college_biology": 0.8125, + "mmlu_eval_accuracy_college_chemistry": 0.375, + "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, + "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365, + "mmlu_eval_accuracy_college_medicine": 0.6818181818181818, + "mmlu_eval_accuracy_college_physics": 0.5454545454545454, + "mmlu_eval_accuracy_computer_security": 0.5454545454545454, + "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384, + "mmlu_eval_accuracy_econometrics": 0.5, + "mmlu_eval_accuracy_electrical_engineering": 0.4375, + "mmlu_eval_accuracy_elementary_mathematics": 0.5121951219512195, + "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, + "mmlu_eval_accuracy_global_facts": 0.2, + "mmlu_eval_accuracy_high_school_biology": 0.625, + "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453, + "mmlu_eval_accuracy_high_school_computer_science": 0.7777777777777778, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.9545454545454546, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209, + "mmlu_eval_accuracy_high_school_mathematics": 0.41379310344827586, + "mmlu_eval_accuracy_high_school_microeconomics": 0.5, + "mmlu_eval_accuracy_high_school_physics": 0.11764705882352941, + "mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667, + "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, + "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727, + "mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693, + "mmlu_eval_accuracy_human_aging": 0.6521739130434783, + "mmlu_eval_accuracy_human_sexuality": 0.6666666666666666, + "mmlu_eval_accuracy_international_law": 0.9230769230769231, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.8, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7209302325581395, + "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527, + "mmlu_eval_accuracy_moral_scenarios": 0.32, + "mmlu_eval_accuracy_nutrition": 0.7575757575757576, + "mmlu_eval_accuracy_philosophy": 0.7352941176470589, + "mmlu_eval_accuracy_prehistory": 0.5714285714285714, + "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644, + "mmlu_eval_accuracy_professional_law": 0.38823529411764707, + "mmlu_eval_accuracy_professional_medicine": 0.7419354838709677, + "mmlu_eval_accuracy_professional_psychology": 0.6811594202898551, + "mmlu_eval_accuracy_public_relations": 0.5833333333333334, + "mmlu_eval_accuracy_security_studies": 0.7037037037037037, + "mmlu_eval_accuracy_sociology": 0.9545454545454546, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.5, + "mmlu_eval_accuracy_world_religions": 0.8947368421052632, + "mmlu_loss": 6.176547575900167, + "step": 3553 + }, + { + "epoch": 1.120782087669505, + "grad_norm": 0.546875, + "learning_rate": 0.0002, + "loss": 1.4549, + "step": 3554 + }, + { + "epoch": 1.121097445600757, + "grad_norm": 0.55859375, + "learning_rate": 0.0002, + "loss": 1.4875, + "step": 3555 + }, + { + "epoch": 1.1214128035320088, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 1.5566, + "step": 3556 + }, + { + "epoch": 1.1217281614632608, + "grad_norm": 0.58984375, + "learning_rate": 0.0002, + "loss": 1.4477, + "step": 3557 + }, + { + "epoch": 1.1220435193945129, + "grad_norm": 0.6484375, + "learning_rate": 0.0002, + "loss": 1.644, + "step": 3558 + }, + { + "epoch": 1.1223588773257647, + "grad_norm": 0.6484375, + "learning_rate": 0.0002, + "loss": 1.6822, + "step": 3559 + }, + { + "epoch": 1.1226742352570167, + "grad_norm": 0.79296875, + "learning_rate": 0.0002, + "loss": 2.1476, + "step": 3560 + }, + { + "epoch": 1.1229895931882687, + "grad_norm": 0.671875, + "learning_rate": 0.0002, + "loss": 1.6542, + "step": 3561 + }, + { + "epoch": 1.1233049511195206, + "grad_norm": 0.71875, + "learning_rate": 0.0002, + "loss": 1.8541, + "step": 3562 + }, + { + "epoch": 1.1236203090507726, + "grad_norm": 0.97265625, + "learning_rate": 0.0002, + "loss": 2.3571, + "step": 3563 + }, + { + "epoch": 1.1239356669820246, + "grad_norm": 0.77734375, + "learning_rate": 0.0002, + "loss": 1.6506, + "step": 3564 + }, + { + "epoch": 1.1242510249132767, + "grad_norm": 0.83203125, + "learning_rate": 0.0002, + "loss": 1.9993, + "step": 3565 + }, + { + "epoch": 1.1245663828445285, + "grad_norm": 0.890625, + "learning_rate": 0.0002, + "loss": 2.0467, + "step": 3566 + }, + { + "epoch": 1.1248817407757805, + "grad_norm": 0.83203125, + "learning_rate": 0.0002, + "loss": 2.0859, + "step": 3567 + }, + { + "epoch": 1.1251970987070326, + "grad_norm": 0.93359375, + "learning_rate": 0.0002, + "loss": 2.1784, + "step": 3568 + }, + { + "epoch": 1.1255124566382844, + "grad_norm": 0.86328125, + "learning_rate": 0.0002, + "loss": 2.3036, + "step": 3569 + }, + { + "epoch": 1.1258278145695364, + "grad_norm": 0.796875, + "learning_rate": 0.0002, + "loss": 1.6357, + "step": 3570 + }, + { + "epoch": 1.1261431725007884, + "grad_norm": 0.8359375, + "learning_rate": 0.0002, + "loss": 1.6757, + "step": 3571 + }, + { + "epoch": 1.1264585304320405, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 1.1461, + "step": 3572 + }, + { + "epoch": 1.1267738883632923, + "grad_norm": 0.201171875, + "learning_rate": 0.0002, + "loss": 1.023, + "step": 3573 + }, + { + "epoch": 1.1270892462945443, + "grad_norm": 0.189453125, + "learning_rate": 0.0002, + "loss": 0.978, + "step": 3574 + }, + { + "epoch": 1.1274046042257964, + "grad_norm": 0.19921875, + "learning_rate": 0.0002, + "loss": 0.975, + "step": 3575 + }, + { + "epoch": 1.1277199621570482, + "grad_norm": 0.2060546875, + "learning_rate": 0.0002, + "loss": 1.0447, + "step": 3576 + }, + { + "epoch": 1.1280353200883002, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.1029, + "step": 3577 + }, + { + "epoch": 1.1283506780195522, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 1.0513, + "step": 3578 + }, + { + "epoch": 1.128666035950804, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.01, + "step": 3579 + }, + { + "epoch": 1.128981393882056, + "grad_norm": 0.2314453125, + "learning_rate": 0.0002, + "loss": 1.0792, + "step": 3580 + }, + { + "epoch": 1.1292967518133081, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 1.0484, + "step": 3581 + }, + { + "epoch": 1.1296121097445602, + "grad_norm": 0.2431640625, + "learning_rate": 0.0002, + "loss": 1.0499, + "step": 3582 + }, + { + "epoch": 1.129927467675812, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.1069, + "step": 3583 + }, + { + "epoch": 1.130242825607064, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 0.9118, + "step": 3584 + }, + { + "epoch": 1.130558183538316, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 1.1772, + "step": 3585 + }, + { + "epoch": 1.1308735414695679, + "grad_norm": 0.23828125, + "learning_rate": 0.0002, + "loss": 1.1095, + "step": 3586 + }, + { + "epoch": 1.13118889940082, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 1.1311, + "step": 3587 + }, + { + "epoch": 1.131504257332072, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 1.0762, + "step": 3588 + }, + { + "epoch": 1.131819615263324, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.1239, + "step": 3589 + }, + { + "epoch": 1.1321349731945758, + "grad_norm": 0.3671875, + "learning_rate": 0.0002, + "loss": 1.2068, + "step": 3590 + }, + { + "epoch": 1.1324503311258278, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 1.0939, + "step": 3591 + }, + { + "epoch": 1.1327656890570799, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.0757, + "step": 3592 + }, + { + "epoch": 1.1330810469883317, + "grad_norm": 0.34765625, + "learning_rate": 0.0002, + "loss": 1.2586, + "step": 3593 + }, + { + "epoch": 1.1333964049195837, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 1.1391, + "step": 3594 + }, + { + "epoch": 1.1337117628508357, + "grad_norm": 0.474609375, + "learning_rate": 0.0002, + "loss": 1.6515, + "step": 3595 + }, + { + "epoch": 1.1340271207820876, + "grad_norm": 0.365234375, + "learning_rate": 0.0002, + "loss": 1.2292, + "step": 3596 + }, + { + "epoch": 1.1343424787133396, + "grad_norm": 0.38671875, + "learning_rate": 0.0002, + "loss": 1.2719, + "step": 3597 + }, + { + "epoch": 1.1346578366445916, + "grad_norm": 0.42578125, + "learning_rate": 0.0002, + "loss": 1.2555, + "step": 3598 + }, + { + "epoch": 1.1349731945758437, + "grad_norm": 0.40234375, + "learning_rate": 0.0002, + "loss": 1.2174, + "step": 3599 + }, + { + "epoch": 1.1352885525070955, + "grad_norm": 0.4921875, + "learning_rate": 0.0002, + "loss": 1.1952, + "step": 3600 + }, + { + "epoch": 1.1356039104383475, + "grad_norm": 0.498046875, + "learning_rate": 0.0002, + "loss": 1.493, + "step": 3601 + }, + { + "epoch": 1.1359192683695996, + "grad_norm": 0.6484375, + "learning_rate": 0.0002, + "loss": 1.1842, + "step": 3602 + }, + { + "epoch": 1.1362346263008516, + "grad_norm": 0.49609375, + "learning_rate": 0.0002, + "loss": 1.2114, + "step": 3603 + }, + { + "epoch": 1.1365499842321034, + "grad_norm": 0.53515625, + "learning_rate": 0.0002, + "loss": 1.4364, + "step": 3604 + }, + { + "epoch": 1.1368653421633554, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 1.4227, + "step": 3605 + }, + { + "epoch": 1.1371807000946075, + "grad_norm": 0.78125, + "learning_rate": 0.0002, + "loss": 1.7187, + "step": 3606 + }, + { + "epoch": 1.1374960580258593, + "grad_norm": 0.7421875, + "learning_rate": 0.0002, + "loss": 1.8865, + "step": 3607 + }, + { + "epoch": 1.1378114159571113, + "grad_norm": 0.9140625, + "learning_rate": 0.0002, + "loss": 2.154, + "step": 3608 + }, + { + "epoch": 1.1381267738883634, + "grad_norm": 0.8984375, + "learning_rate": 0.0002, + "loss": 2.3594, + "step": 3609 + }, + { + "epoch": 1.1384421318196152, + "grad_norm": 0.76171875, + "learning_rate": 0.0002, + "loss": 1.9359, + "step": 3610 + }, + { + "epoch": 1.1387574897508672, + "grad_norm": 0.859375, + "learning_rate": 0.0002, + "loss": 2.0481, + "step": 3611 + }, + { + "epoch": 1.1390728476821192, + "grad_norm": 0.63671875, + "learning_rate": 0.0002, + "loss": 1.697, + "step": 3612 + }, + { + "epoch": 1.139388205613371, + "grad_norm": 0.921875, + "learning_rate": 0.0002, + "loss": 2.1183, + "step": 3613 + }, + { + "epoch": 1.139703563544623, + "grad_norm": 0.7734375, + "learning_rate": 0.0002, + "loss": 2.0349, + "step": 3614 + }, + { + "epoch": 1.1400189214758751, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 2.0597, + "step": 3615 + }, + { + "epoch": 1.1403342794071272, + "grad_norm": 0.96875, + "learning_rate": 0.0002, + "loss": 2.1503, + "step": 3616 + }, + { + "epoch": 1.140649637338379, + "grad_norm": 0.93359375, + "learning_rate": 0.0002, + "loss": 2.0752, + "step": 3617 + }, + { + "epoch": 1.140964995269631, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 2.2398, + "step": 3618 + }, + { + "epoch": 1.141280353200883, + "grad_norm": 0.92578125, + "learning_rate": 0.0002, + "loss": 1.8032, + "step": 3619 + }, + { + "epoch": 1.141595711132135, + "grad_norm": 1.0390625, + "learning_rate": 0.0002, + "loss": 1.977, + "step": 3620 + }, + { + "epoch": 1.141911069063387, + "grad_norm": 1.234375, + "learning_rate": 0.0002, + "loss": 2.4415, + "step": 3621 + }, + { + "epoch": 1.142226426994639, + "grad_norm": 0.212890625, + "learning_rate": 0.0002, + "loss": 1.1101, + "step": 3622 + }, + { + "epoch": 1.142541784925891, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 1.1197, + "step": 3623 + }, + { + "epoch": 1.1428571428571428, + "grad_norm": 0.2138671875, + "learning_rate": 0.0002, + "loss": 0.9494, + "step": 3624 + }, + { + "epoch": 1.1431725007883948, + "grad_norm": 0.19921875, + "learning_rate": 0.0002, + "loss": 0.9778, + "step": 3625 + }, + { + "epoch": 1.1434878587196469, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 1.0047, + "step": 3626 + }, + { + "epoch": 1.1438032166508987, + "grad_norm": 0.22265625, + "learning_rate": 0.0002, + "loss": 1.0502, + "step": 3627 + }, + { + "epoch": 1.1441185745821507, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 0.9605, + "step": 3628 + }, + { + "epoch": 1.1444339325134028, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 1.0431, + "step": 3629 + }, + { + "epoch": 1.1447492904446546, + "grad_norm": 0.2041015625, + "learning_rate": 0.0002, + "loss": 1.098, + "step": 3630 + }, + { + "epoch": 1.1450646483759066, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.0224, + "step": 3631 + }, + { + "epoch": 1.1453800063071586, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.0674, + "step": 3632 + }, + { + "epoch": 1.1456953642384107, + "grad_norm": 0.22265625, + "learning_rate": 0.0002, + "loss": 1.0605, + "step": 3633 + }, + { + "epoch": 1.1460107221696625, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.1406, + "step": 3634 + }, + { + "epoch": 1.1463260801009145, + "grad_norm": 0.22265625, + "learning_rate": 0.0002, + "loss": 0.9892, + "step": 3635 + }, + { + "epoch": 1.1466414380321666, + "grad_norm": 0.2265625, + "learning_rate": 0.0002, + "loss": 1.0082, + "step": 3636 + }, + { + "epoch": 1.1469567959634186, + "grad_norm": 0.2431640625, + "learning_rate": 0.0002, + "loss": 1.0641, + "step": 3637 + }, + { + "epoch": 1.1472721538946704, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.1713, + "step": 3638 + }, + { + "epoch": 1.1475875118259224, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 1.1088, + "step": 3639 + }, + { + "epoch": 1.1479028697571745, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.2692, + "step": 3640 + }, + { + "epoch": 1.1482182276884263, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 1.2193, + "step": 3641 + }, + { + "epoch": 1.1485335856196783, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 0.9935, + "step": 3642 + }, + { + "epoch": 1.1488489435509304, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 1.2118, + "step": 3643 + }, + { + "epoch": 1.1491643014821822, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 1.1554, + "step": 3644 + }, + { + "epoch": 1.1494796594134342, + "grad_norm": 0.40234375, + "learning_rate": 0.0002, + "loss": 1.1987, + "step": 3645 + }, + { + "epoch": 1.1497950173446863, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 1.2169, + "step": 3646 + }, + { + "epoch": 1.150110375275938, + "grad_norm": 0.380859375, + "learning_rate": 0.0002, + "loss": 1.2143, + "step": 3647 + }, + { + "epoch": 1.15042573320719, + "grad_norm": 0.484375, + "learning_rate": 0.0002, + "loss": 1.1924, + "step": 3648 + }, + { + "epoch": 1.1507410911384421, + "grad_norm": 0.44921875, + "learning_rate": 0.0002, + "loss": 1.3995, + "step": 3649 + }, + { + "epoch": 1.1510564490696942, + "grad_norm": 0.470703125, + "learning_rate": 0.0002, + "loss": 1.3999, + "step": 3650 + }, + { + "epoch": 1.151371807000946, + "grad_norm": 0.486328125, + "learning_rate": 0.0002, + "loss": 1.3007, + "step": 3651 + }, + { + "epoch": 1.151687164932198, + "grad_norm": 0.482421875, + "learning_rate": 0.0002, + "loss": 1.359, + "step": 3652 + }, + { + "epoch": 1.15200252286345, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 3653 + }, + { + "epoch": 1.152317880794702, + "grad_norm": 0.61328125, + "learning_rate": 0.0002, + "loss": 1.5314, + "step": 3654 + }, + { + "epoch": 1.152633238725954, + "grad_norm": 0.51953125, + "learning_rate": 0.0002, + "loss": 1.2039, + "step": 3655 + }, + { + "epoch": 1.152948596657206, + "grad_norm": 0.7578125, + "learning_rate": 0.0002, + "loss": 1.8937, + "step": 3656 + }, + { + "epoch": 1.153263954588458, + "grad_norm": 0.77734375, + "learning_rate": 0.0002, + "loss": 2.0666, + "step": 3657 + }, + { + "epoch": 1.1535793125197098, + "grad_norm": 0.77734375, + "learning_rate": 0.0002, + "loss": 1.9582, + "step": 3658 + }, + { + "epoch": 1.1538946704509618, + "grad_norm": 0.76953125, + "learning_rate": 0.0002, + "loss": 2.0212, + "step": 3659 + }, + { + "epoch": 1.1542100283822139, + "grad_norm": 1.03125, + "learning_rate": 0.0002, + "loss": 2.1688, + "step": 3660 + }, + { + "epoch": 1.1545253863134657, + "grad_norm": 0.87109375, + "learning_rate": 0.0002, + "loss": 1.8468, + "step": 3661 + }, + { + "epoch": 1.1548407442447177, + "grad_norm": 0.74609375, + "learning_rate": 0.0002, + "loss": 1.6897, + "step": 3662 + }, + { + "epoch": 1.1551561021759698, + "grad_norm": 1.125, + "learning_rate": 0.0002, + "loss": 2.2066, + "step": 3663 + }, + { + "epoch": 1.1554714601072218, + "grad_norm": 0.96484375, + "learning_rate": 0.0002, + "loss": 2.1667, + "step": 3664 + }, + { + "epoch": 1.1557868180384736, + "grad_norm": 0.7890625, + "learning_rate": 0.0002, + "loss": 1.8217, + "step": 3665 + }, + { + "epoch": 1.1561021759697256, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 2.0756, + "step": 3666 + }, + { + "epoch": 1.1564175339009777, + "grad_norm": 0.98046875, + "learning_rate": 0.0002, + "loss": 2.2185, + "step": 3667 + }, + { + "epoch": 1.1567328918322295, + "grad_norm": 0.98828125, + "learning_rate": 0.0002, + "loss": 1.7657, + "step": 3668 + }, + { + "epoch": 1.1570482497634815, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 1.5323, + "step": 3669 + }, + { + "epoch": 1.1573636076947336, + "grad_norm": 0.9765625, + "learning_rate": 0.0002, + "loss": 1.846, + "step": 3670 + }, + { + "epoch": 1.1576789656259856, + "grad_norm": 1.078125, + "learning_rate": 0.0002, + "loss": 2.1223, + "step": 3671 + }, + { + "epoch": 1.1579943235572374, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 0.9803, + "step": 3672 + }, + { + "epoch": 1.1583096814884895, + "grad_norm": 0.216796875, + "learning_rate": 0.0002, + "loss": 0.9973, + "step": 3673 + }, + { + "epoch": 1.1586250394197415, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 0.939, + "step": 3674 + }, + { + "epoch": 1.1589403973509933, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 1.1632, + "step": 3675 + }, + { + "epoch": 1.1592557552822453, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 1.0109, + "step": 3676 + }, + { + "epoch": 1.1595711132134974, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 0.9904, + "step": 3677 + }, + { + "epoch": 1.1598864711447492, + "grad_norm": 0.19921875, + "learning_rate": 0.0002, + "loss": 0.9995, + "step": 3678 + }, + { + "epoch": 1.1602018290760012, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 0.9314, + "step": 3679 + }, + { + "epoch": 1.1605171870072533, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 1.067, + "step": 3680 + }, + { + "epoch": 1.1608325449385053, + "grad_norm": 0.201171875, + "learning_rate": 0.0002, + "loss": 1.0402, + "step": 3681 + }, + { + "epoch": 1.161147902869757, + "grad_norm": 0.189453125, + "learning_rate": 0.0002, + "loss": 1.0843, + "step": 3682 + }, + { + "epoch": 1.1614632608010091, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.1085, + "step": 3683 + }, + { + "epoch": 1.1617786187322612, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.0203, + "step": 3684 + }, + { + "epoch": 1.162093976663513, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.1233, + "step": 3685 + }, + { + "epoch": 1.162409334594765, + "grad_norm": 0.2412109375, + "learning_rate": 0.0002, + "loss": 1.1893, + "step": 3686 + }, + { + "epoch": 1.162724692526017, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.0812, + "step": 3687 + }, + { + "epoch": 1.163040050457269, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 1.0411, + "step": 3688 + }, + { + "epoch": 1.163355408388521, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.0542, + "step": 3689 + }, + { + "epoch": 1.163670766319773, + "grad_norm": 0.2490234375, + "learning_rate": 0.0002, + "loss": 1.0491, + "step": 3690 + }, + { + "epoch": 1.163986124251025, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 1.0924, + "step": 3691 + }, + { + "epoch": 1.1643014821822768, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 1.1103, + "step": 3692 + }, + { + "epoch": 1.1646168401135288, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 0.8672, + "step": 3693 + }, + { + "epoch": 1.1649321980447809, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 0.9765, + "step": 3694 + }, + { + "epoch": 1.1652475559760327, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 1.0611, + "step": 3695 + }, + { + "epoch": 1.1655629139072847, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 1.1183, + "step": 3696 + }, + { + "epoch": 1.1658782718385368, + "grad_norm": 0.3671875, + "learning_rate": 0.0002, + "loss": 1.0915, + "step": 3697 + }, + { + "epoch": 1.1661936297697888, + "grad_norm": 0.439453125, + "learning_rate": 0.0002, + "loss": 1.4, + "step": 3698 + }, + { + "epoch": 1.1665089877010406, + "grad_norm": 0.400390625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 3699 + }, + { + "epoch": 1.1668243456322926, + "grad_norm": 0.3984375, + "learning_rate": 0.0002, + "loss": 1.1452, + "step": 3700 + }, + { + "epoch": 1.1671397035635447, + "grad_norm": 0.45703125, + "learning_rate": 0.0002, + "loss": 1.2758, + "step": 3701 + }, + { + "epoch": 1.1674550614947965, + "grad_norm": 0.74609375, + "learning_rate": 0.0002, + "loss": 1.4966, + "step": 3702 + }, + { + "epoch": 1.1677704194260485, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 1.6036, + "step": 3703 + }, + { + "epoch": 1.1680857773573006, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 1.5805, + "step": 3704 + }, + { + "epoch": 1.1684011352885526, + "grad_norm": 0.68359375, + "learning_rate": 0.0002, + "loss": 1.6867, + "step": 3705 + }, + { + "epoch": 1.1687164932198044, + "grad_norm": 0.60546875, + "learning_rate": 0.0002, + "loss": 1.4276, + "step": 3706 + }, + { + "epoch": 1.1690318511510565, + "grad_norm": 0.703125, + "learning_rate": 0.0002, + "loss": 1.7455, + "step": 3707 + }, + { + "epoch": 1.1693472090823085, + "grad_norm": 0.8203125, + "learning_rate": 0.0002, + "loss": 1.8387, + "step": 3708 + }, + { + "epoch": 1.1696625670135603, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 1.4123, + "step": 3709 + }, + { + "epoch": 1.1699779249448123, + "grad_norm": 0.7578125, + "learning_rate": 0.0002, + "loss": 1.6835, + "step": 3710 + }, + { + "epoch": 1.1702932828760644, + "grad_norm": 0.83203125, + "learning_rate": 0.0002, + "loss": 2.0572, + "step": 3711 + }, + { + "epoch": 1.1706086408073162, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 1.8621, + "step": 3712 + }, + { + "epoch": 1.1709239987385682, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 2.1537, + "step": 3713 + }, + { + "epoch": 1.1712393566698203, + "grad_norm": 0.9140625, + "learning_rate": 0.0002, + "loss": 2.253, + "step": 3714 + }, + { + "epoch": 1.1715547146010723, + "grad_norm": 0.94140625, + "learning_rate": 0.0002, + "loss": 2.1079, + "step": 3715 + }, + { + "epoch": 1.1718700725323241, + "grad_norm": 0.8515625, + "learning_rate": 0.0002, + "loss": 2.151, + "step": 3716 + }, + { + "epoch": 1.1721854304635762, + "grad_norm": 0.87890625, + "learning_rate": 0.0002, + "loss": 2.0435, + "step": 3717 + }, + { + "epoch": 1.1725007883948282, + "grad_norm": 1.2109375, + "learning_rate": 0.0002, + "loss": 1.898, + "step": 3718 + }, + { + "epoch": 1.1728161463260802, + "grad_norm": 0.70703125, + "learning_rate": 0.0002, + "loss": 1.6946, + "step": 3719 + }, + { + "epoch": 1.173131504257332, + "grad_norm": 0.765625, + "learning_rate": 0.0002, + "loss": 1.6488, + "step": 3720 + }, + { + "epoch": 1.173446862188584, + "grad_norm": 1.4296875, + "learning_rate": 0.0002, + "loss": 2.4379, + "step": 3721 + }, + { + "epoch": 1.173762220119836, + "grad_norm": 0.1884765625, + "learning_rate": 0.0002, + "loss": 1.067, + "step": 3722 + }, + { + "epoch": 1.174077578051088, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.0523, + "step": 3723 + }, + { + "epoch": 1.17439293598234, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 0.992, + "step": 3724 + }, + { + "epoch": 1.174708293913592, + "grad_norm": 0.1884765625, + "learning_rate": 0.0002, + "loss": 0.9778, + "step": 3725 + }, + { + "epoch": 1.1750236518448438, + "grad_norm": 0.1943359375, + "learning_rate": 0.0002, + "loss": 1.044, + "step": 3726 + }, + { + "epoch": 1.1753390097760958, + "grad_norm": 0.2001953125, + "learning_rate": 0.0002, + "loss": 1.0854, + "step": 3727 + }, + { + "epoch": 1.1756543677073479, + "grad_norm": 0.2041015625, + "learning_rate": 0.0002, + "loss": 1.0207, + "step": 3728 + }, + { + "epoch": 1.1759697256385997, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 1.0844, + "step": 3729 + }, + { + "epoch": 1.1762850835698517, + "grad_norm": 0.2216796875, + "learning_rate": 0.0002, + "loss": 1.054, + "step": 3730 + }, + { + "epoch": 1.1766004415011038, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.073, + "step": 3731 + }, + { + "epoch": 1.1769157994323558, + "grad_norm": 0.19921875, + "learning_rate": 0.0002, + "loss": 0.9874, + "step": 3732 + }, + { + "epoch": 1.1772311573636076, + "grad_norm": 0.23046875, + "learning_rate": 0.0002, + "loss": 1.0109, + "step": 3733 + }, + { + "epoch": 1.1775465152948597, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 1.1528, + "step": 3734 + }, + { + "epoch": 1.1778618732261117, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 1.0086, + "step": 3735 + }, + { + "epoch": 1.1781772311573637, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.1529, + "step": 3736 + }, + { + "epoch": 1.1784925890886155, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 0.9417, + "step": 3737 + }, + { + "epoch": 1.1788079470198676, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.0012, + "step": 3738 + }, + { + "epoch": 1.1791233049511196, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 1.0508, + "step": 3739 + }, + { + "epoch": 1.1794386628823714, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 1.1876, + "step": 3740 + }, + { + "epoch": 1.1794386628823714, + "eval_loss": 1.675485372543335, + "eval_runtime": 148.9816, + "eval_samples_per_second": 6.712, + "eval_steps_per_second": 6.712, + "step": 3740 + }, + { + "epoch": 1.1794386628823714, + "mmlu_eval_accuracy": 0.5984316704084249, + "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, + "mmlu_eval_accuracy_anatomy": 0.5714285714285714, + "mmlu_eval_accuracy_astronomy": 0.625, + "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, + "mmlu_eval_accuracy_clinical_knowledge": 0.6896551724137931, + "mmlu_eval_accuracy_college_biology": 0.8125, + "mmlu_eval_accuracy_college_chemistry": 0.375, + "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, + "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365, + "mmlu_eval_accuracy_college_medicine": 0.6818181818181818, + "mmlu_eval_accuracy_college_physics": 0.5454545454545454, + "mmlu_eval_accuracy_computer_security": 0.5454545454545454, + "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384, + "mmlu_eval_accuracy_econometrics": 0.4166666666666667, + "mmlu_eval_accuracy_electrical_engineering": 0.375, + "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244, + "mmlu_eval_accuracy_formal_logic": 0.2857142857142857, + "mmlu_eval_accuracy_global_facts": 0.2, + "mmlu_eval_accuracy_high_school_biology": 0.625, + "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453, + "mmlu_eval_accuracy_high_school_computer_science": 0.7777777777777778, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442, + "mmlu_eval_accuracy_high_school_mathematics": 0.3103448275862069, + "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154, + "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, + "mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333, + "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, + "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, + "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, + "mmlu_eval_accuracy_human_aging": 0.6521739130434783, + "mmlu_eval_accuracy_human_sexuality": 0.6666666666666666, + "mmlu_eval_accuracy_international_law": 0.8461538461538461, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.72, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7209302325581395, + "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527, + "mmlu_eval_accuracy_moral_scenarios": 0.28, + "mmlu_eval_accuracy_nutrition": 0.696969696969697, + "mmlu_eval_accuracy_philosophy": 0.7058823529411765, + "mmlu_eval_accuracy_prehistory": 0.6, + "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194, + "mmlu_eval_accuracy_professional_law": 0.3588235294117647, + "mmlu_eval_accuracy_professional_medicine": 0.6774193548387096, + "mmlu_eval_accuracy_professional_psychology": 0.6666666666666666, + "mmlu_eval_accuracy_public_relations": 0.4166666666666667, + "mmlu_eval_accuracy_security_studies": 0.7037037037037037, + "mmlu_eval_accuracy_sociology": 0.9545454545454546, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.5, + "mmlu_eval_accuracy_world_religions": 0.8421052631578947, + "mmlu_loss": 5.815515008025665, + "step": 3740 + }, + { + "epoch": 1.1797540208136235, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 0.9889, + "step": 3741 + }, + { + "epoch": 1.1800693787448755, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 1.1516, + "step": 3742 + }, + { + "epoch": 1.1803847366761273, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 1.1376, + "step": 3743 + }, + { + "epoch": 1.1807000946073793, + "grad_norm": 0.330078125, + "learning_rate": 0.0002, + "loss": 1.2196, + "step": 3744 + }, + { + "epoch": 1.1810154525386314, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 1.0521, + "step": 3745 + }, + { + "epoch": 1.1813308104698832, + "grad_norm": 0.3828125, + "learning_rate": 0.0002, + "loss": 1.2108, + "step": 3746 + }, + { + "epoch": 1.1816461684011352, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 1.2999, + "step": 3747 + }, + { + "epoch": 1.1819615263323873, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 1.0858, + "step": 3748 + }, + { + "epoch": 1.1822768842636393, + "grad_norm": 0.447265625, + "learning_rate": 0.0002, + "loss": 1.3862, + "step": 3749 + }, + { + "epoch": 1.1825922421948911, + "grad_norm": 0.353515625, + "learning_rate": 0.0002, + "loss": 1.1926, + "step": 3750 + }, + { + "epoch": 1.1829076001261432, + "grad_norm": 0.41796875, + "learning_rate": 0.0002, + "loss": 1.2472, + "step": 3751 + }, + { + "epoch": 1.1832229580573952, + "grad_norm": 0.470703125, + "learning_rate": 0.0002, + "loss": 1.4331, + "step": 3752 + }, + { + "epoch": 1.1835383159886472, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 3753 + }, + { + "epoch": 1.183853673919899, + "grad_norm": 0.53515625, + "learning_rate": 0.0002, + "loss": 1.4119, + "step": 3754 + }, + { + "epoch": 1.184169031851151, + "grad_norm": 0.5703125, + "learning_rate": 0.0002, + "loss": 1.444, + "step": 3755 + }, + { + "epoch": 1.1844843897824031, + "grad_norm": 0.609375, + "learning_rate": 0.0002, + "loss": 1.2427, + "step": 3756 + }, + { + "epoch": 1.184799747713655, + "grad_norm": 0.6171875, + "learning_rate": 0.0002, + "loss": 1.701, + "step": 3757 + }, + { + "epoch": 1.185115105644907, + "grad_norm": 0.71484375, + "learning_rate": 0.0002, + "loss": 1.6946, + "step": 3758 + }, + { + "epoch": 1.185430463576159, + "grad_norm": 0.765625, + "learning_rate": 0.0002, + "loss": 1.843, + "step": 3759 + }, + { + "epoch": 1.1857458215074108, + "grad_norm": 0.7265625, + "learning_rate": 0.0002, + "loss": 1.8329, + "step": 3760 + }, + { + "epoch": 1.1860611794386628, + "grad_norm": 1.0, + "learning_rate": 0.0002, + "loss": 1.9288, + "step": 3761 + }, + { + "epoch": 1.1863765373699149, + "grad_norm": 1.09375, + "learning_rate": 0.0002, + "loss": 2.2774, + "step": 3762 + }, + { + "epoch": 1.186691895301167, + "grad_norm": 1.03125, + "learning_rate": 0.0002, + "loss": 2.1545, + "step": 3763 + }, + { + "epoch": 1.1870072532324187, + "grad_norm": 0.80078125, + "learning_rate": 0.0002, + "loss": 1.846, + "step": 3764 + }, + { + "epoch": 1.1873226111636708, + "grad_norm": 1.2890625, + "learning_rate": 0.0002, + "loss": 2.0737, + "step": 3765 + }, + { + "epoch": 1.1876379690949228, + "grad_norm": 0.9765625, + "learning_rate": 0.0002, + "loss": 1.884, + "step": 3766 + }, + { + "epoch": 1.1879533270261746, + "grad_norm": 1.1875, + "learning_rate": 0.0002, + "loss": 2.3189, + "step": 3767 + }, + { + "epoch": 1.1882686849574267, + "grad_norm": 0.90234375, + "learning_rate": 0.0002, + "loss": 1.7346, + "step": 3768 + }, + { + "epoch": 1.1885840428886787, + "grad_norm": 0.91796875, + "learning_rate": 0.0002, + "loss": 1.579, + "step": 3769 + }, + { + "epoch": 1.1888994008199307, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 1.7031, + "step": 3770 + }, + { + "epoch": 1.1892147587511825, + "grad_norm": 1.125, + "learning_rate": 0.0002, + "loss": 2.1932, + "step": 3771 + }, + { + "epoch": 1.1895301166824346, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.0672, + "step": 3772 + }, + { + "epoch": 1.1898454746136866, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 1.1452, + "step": 3773 + }, + { + "epoch": 1.1901608325449384, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.0192, + "step": 3774 + }, + { + "epoch": 1.1904761904761905, + "grad_norm": 0.2060546875, + "learning_rate": 0.0002, + "loss": 1.0602, + "step": 3775 + }, + { + "epoch": 1.1907915484074425, + "grad_norm": 0.2392578125, + "learning_rate": 0.0002, + "loss": 0.9937, + "step": 3776 + }, + { + "epoch": 1.1911069063386943, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 1.0329, + "step": 3777 + }, + { + "epoch": 1.1914222642699464, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.0373, + "step": 3778 + }, + { + "epoch": 1.1917376222011984, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.1045, + "step": 3779 + }, + { + "epoch": 1.1920529801324504, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 1.0119, + "step": 3780 + }, + { + "epoch": 1.1923683380637022, + "grad_norm": 0.2177734375, + "learning_rate": 0.0002, + "loss": 1.0049, + "step": 3781 + }, + { + "epoch": 1.1926836959949543, + "grad_norm": 0.201171875, + "learning_rate": 0.0002, + "loss": 1.0365, + "step": 3782 + }, + { + "epoch": 1.1929990539262063, + "grad_norm": 0.224609375, + "learning_rate": 0.0002, + "loss": 1.1219, + "step": 3783 + }, + { + "epoch": 1.1933144118574581, + "grad_norm": 0.2451171875, + "learning_rate": 0.0002, + "loss": 1.1319, + "step": 3784 + }, + { + "epoch": 1.1936297697887102, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.0975, + "step": 3785 + }, + { + "epoch": 1.1939451277199622, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 0.9518, + "step": 3786 + }, + { + "epoch": 1.1942604856512142, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 1.1622, + "step": 3787 + }, + { + "epoch": 1.194575843582466, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 0.9851, + "step": 3788 + }, + { + "epoch": 1.194891201513718, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 3789 + }, + { + "epoch": 1.1952065594449701, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.1101, + "step": 3790 + }, + { + "epoch": 1.195521917376222, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 1.0248, + "step": 3791 + }, + { + "epoch": 1.195837275307474, + "grad_norm": 0.365234375, + "learning_rate": 0.0002, + "loss": 1.1755, + "step": 3792 + }, + { + "epoch": 1.196152633238726, + "grad_norm": 0.34765625, + "learning_rate": 0.0002, + "loss": 1.2263, + "step": 3793 + }, + { + "epoch": 1.1964679911699778, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 0.9941, + "step": 3794 + }, + { + "epoch": 1.1967833491012299, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 1.2028, + "step": 3795 + }, + { + "epoch": 1.197098707032482, + "grad_norm": 0.353515625, + "learning_rate": 0.0002, + "loss": 1.0515, + "step": 3796 + }, + { + "epoch": 1.197414064963734, + "grad_norm": 0.384765625, + "learning_rate": 0.0002, + "loss": 1.0216, + "step": 3797 + }, + { + "epoch": 1.1977294228949857, + "grad_norm": 0.45703125, + "learning_rate": 0.0002, + "loss": 1.2827, + "step": 3798 + }, + { + "epoch": 1.1980447808262378, + "grad_norm": 0.447265625, + "learning_rate": 0.0002, + "loss": 1.2698, + "step": 3799 + }, + { + "epoch": 1.1983601387574898, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 1.0376, + "step": 3800 + }, + { + "epoch": 1.1986754966887416, + "grad_norm": 0.515625, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 3801 + }, + { + "epoch": 1.1989908546199937, + "grad_norm": 0.4453125, + "learning_rate": 0.0002, + "loss": 1.2639, + "step": 3802 + }, + { + "epoch": 1.1993062125512457, + "grad_norm": 0.55859375, + "learning_rate": 0.0002, + "loss": 1.5286, + "step": 3803 + }, + { + "epoch": 1.1996215704824977, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 1.4875, + "step": 3804 + }, + { + "epoch": 1.1999369284137495, + "grad_norm": 0.60546875, + "learning_rate": 0.0002, + "loss": 1.4686, + "step": 3805 + }, + { + "epoch": 1.2002522863450016, + "grad_norm": 0.52734375, + "learning_rate": 0.0002, + "loss": 1.4374, + "step": 3806 + }, + { + "epoch": 1.2005676442762536, + "grad_norm": 0.76171875, + "learning_rate": 0.0002, + "loss": 1.6614, + "step": 3807 + }, + { + "epoch": 1.2008830022075054, + "grad_norm": 0.61328125, + "learning_rate": 0.0002, + "loss": 1.3737, + "step": 3808 + }, + { + "epoch": 1.2011983601387575, + "grad_norm": 0.76171875, + "learning_rate": 0.0002, + "loss": 1.5139, + "step": 3809 + }, + { + "epoch": 1.2015137180700095, + "grad_norm": 0.87890625, + "learning_rate": 0.0002, + "loss": 2.0056, + "step": 3810 + }, + { + "epoch": 1.2018290760012613, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 2.144, + "step": 3811 + }, + { + "epoch": 1.2021444339325134, + "grad_norm": 0.796875, + "learning_rate": 0.0002, + "loss": 2.1345, + "step": 3812 + }, + { + "epoch": 1.2024597918637654, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 1.9621, + "step": 3813 + }, + { + "epoch": 1.2027751497950174, + "grad_norm": 0.73828125, + "learning_rate": 0.0002, + "loss": 1.8271, + "step": 3814 + }, + { + "epoch": 1.2030905077262692, + "grad_norm": 0.8203125, + "learning_rate": 0.0002, + "loss": 1.9633, + "step": 3815 + }, + { + "epoch": 1.2034058656575213, + "grad_norm": 0.80859375, + "learning_rate": 0.0002, + "loss": 1.9335, + "step": 3816 + }, + { + "epoch": 1.2037212235887733, + "grad_norm": 0.80859375, + "learning_rate": 0.0002, + "loss": 1.846, + "step": 3817 + }, + { + "epoch": 1.2040365815200254, + "grad_norm": 0.7890625, + "learning_rate": 0.0002, + "loss": 1.7934, + "step": 3818 + }, + { + "epoch": 1.2043519394512772, + "grad_norm": 0.79296875, + "learning_rate": 0.0002, + "loss": 1.5013, + "step": 3819 + }, + { + "epoch": 1.2046672973825292, + "grad_norm": 1.03125, + "learning_rate": 0.0002, + "loss": 2.1411, + "step": 3820 + }, + { + "epoch": 1.2049826553137812, + "grad_norm": 0.99609375, + "learning_rate": 0.0002, + "loss": 2.1004, + "step": 3821 + }, + { + "epoch": 1.205298013245033, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 1.1607, + "step": 3822 + }, + { + "epoch": 1.205613371176285, + "grad_norm": 0.2275390625, + "learning_rate": 0.0002, + "loss": 1.1517, + "step": 3823 + }, + { + "epoch": 1.2059287291075371, + "grad_norm": 0.2041015625, + "learning_rate": 0.0002, + "loss": 1.0251, + "step": 3824 + }, + { + "epoch": 1.206244087038789, + "grad_norm": 0.201171875, + "learning_rate": 0.0002, + "loss": 1.1151, + "step": 3825 + }, + { + "epoch": 1.206559444970041, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 1.0488, + "step": 3826 + }, + { + "epoch": 1.206874802901293, + "grad_norm": 0.19921875, + "learning_rate": 0.0002, + "loss": 1.1138, + "step": 3827 + }, + { + "epoch": 1.2071901608325448, + "grad_norm": 0.208984375, + "learning_rate": 0.0002, + "loss": 1.1435, + "step": 3828 + }, + { + "epoch": 1.2075055187637969, + "grad_norm": 0.201171875, + "learning_rate": 0.0002, + "loss": 1.0309, + "step": 3829 + }, + { + "epoch": 1.207820876695049, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.0471, + "step": 3830 + }, + { + "epoch": 1.208136234626301, + "grad_norm": 0.2216796875, + "learning_rate": 0.0002, + "loss": 1.0661, + "step": 3831 + }, + { + "epoch": 1.2084515925575527, + "grad_norm": 0.2265625, + "learning_rate": 0.0002, + "loss": 1.0047, + "step": 3832 + }, + { + "epoch": 1.2087669504888048, + "grad_norm": 0.212890625, + "learning_rate": 0.0002, + "loss": 1.0562, + "step": 3833 + }, + { + "epoch": 1.2090823084200568, + "grad_norm": 0.2314453125, + "learning_rate": 0.0002, + "loss": 1.0705, + "step": 3834 + }, + { + "epoch": 1.2093976663513089, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.0648, + "step": 3835 + }, + { + "epoch": 1.2097130242825607, + "grad_norm": 0.22265625, + "learning_rate": 0.0002, + "loss": 0.9588, + "step": 3836 + }, + { + "epoch": 1.2100283822138127, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 1.0056, + "step": 3837 + }, + { + "epoch": 1.2103437401450647, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.0379, + "step": 3838 + }, + { + "epoch": 1.2106590980763166, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 1.1342, + "step": 3839 + }, + { + "epoch": 1.2109744560075686, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 1.1074, + "step": 3840 + }, + { + "epoch": 1.2112898139388206, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 1.0676, + "step": 3841 + }, + { + "epoch": 1.2116051718700724, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 1.0608, + "step": 3842 + }, + { + "epoch": 1.2119205298013245, + "grad_norm": 0.3828125, + "learning_rate": 0.0002, + "loss": 1.1897, + "step": 3843 + }, + { + "epoch": 1.2122358877325765, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 1.3657, + "step": 3844 + }, + { + "epoch": 1.2125512456638283, + "grad_norm": 0.3828125, + "learning_rate": 0.0002, + "loss": 1.192, + "step": 3845 + }, + { + "epoch": 1.2128666035950804, + "grad_norm": 0.384765625, + "learning_rate": 0.0002, + "loss": 1.0759, + "step": 3846 + }, + { + "epoch": 1.2131819615263324, + "grad_norm": 0.416015625, + "learning_rate": 0.0002, + "loss": 1.1305, + "step": 3847 + }, + { + "epoch": 1.2134973194575844, + "grad_norm": 0.47265625, + "learning_rate": 0.0002, + "loss": 1.3811, + "step": 3848 + }, + { + "epoch": 1.2138126773888362, + "grad_norm": 0.46484375, + "learning_rate": 0.0002, + "loss": 1.2404, + "step": 3849 + }, + { + "epoch": 1.2141280353200883, + "grad_norm": 0.515625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 3850 + }, + { + "epoch": 1.2144433932513403, + "grad_norm": 0.51953125, + "learning_rate": 0.0002, + "loss": 1.5785, + "step": 3851 + }, + { + "epoch": 1.2147587511825924, + "grad_norm": 0.5390625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 3852 + }, + { + "epoch": 1.2150741091138442, + "grad_norm": 0.6875, + "learning_rate": 0.0002, + "loss": 1.7573, + "step": 3853 + }, + { + "epoch": 1.2153894670450962, + "grad_norm": 0.56640625, + "learning_rate": 0.0002, + "loss": 1.4597, + "step": 3854 + }, + { + "epoch": 1.2157048249763482, + "grad_norm": 0.71484375, + "learning_rate": 0.0002, + "loss": 1.8658, + "step": 3855 + }, + { + "epoch": 1.2160201829076, + "grad_norm": 0.60546875, + "learning_rate": 0.0002, + "loss": 1.494, + "step": 3856 + }, + { + "epoch": 1.216335540838852, + "grad_norm": 0.65625, + "learning_rate": 0.0002, + "loss": 1.7624, + "step": 3857 + }, + { + "epoch": 1.2166508987701041, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 1.4557, + "step": 3858 + }, + { + "epoch": 1.216966256701356, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 2.2698, + "step": 3859 + }, + { + "epoch": 1.217281614632608, + "grad_norm": 0.74609375, + "learning_rate": 0.0002, + "loss": 1.8574, + "step": 3860 + }, + { + "epoch": 1.21759697256386, + "grad_norm": 0.81640625, + "learning_rate": 0.0002, + "loss": 1.9049, + "step": 3861 + }, + { + "epoch": 1.2179123304951118, + "grad_norm": 0.79296875, + "learning_rate": 0.0002, + "loss": 1.9179, + "step": 3862 + }, + { + "epoch": 1.2182276884263639, + "grad_norm": 0.95703125, + "learning_rate": 0.0002, + "loss": 2.1403, + "step": 3863 + }, + { + "epoch": 1.218543046357616, + "grad_norm": 1.109375, + "learning_rate": 0.0002, + "loss": 2.3441, + "step": 3864 + }, + { + "epoch": 1.218858404288868, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 1.9366, + "step": 3865 + }, + { + "epoch": 1.2191737622201197, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 1.9637, + "step": 3866 + }, + { + "epoch": 1.2194891201513718, + "grad_norm": 0.8046875, + "learning_rate": 0.0002, + "loss": 1.652, + "step": 3867 + }, + { + "epoch": 1.2198044780826238, + "grad_norm": 0.8359375, + "learning_rate": 0.0002, + "loss": 1.9583, + "step": 3868 + }, + { + "epoch": 1.2201198360138759, + "grad_norm": 0.86328125, + "learning_rate": 0.0002, + "loss": 1.574, + "step": 3869 + }, + { + "epoch": 1.2204351939451277, + "grad_norm": 1.890625, + "learning_rate": 0.0002, + "loss": 2.0062, + "step": 3870 + }, + { + "epoch": 1.2207505518763797, + "grad_norm": 0.84765625, + "learning_rate": 0.0002, + "loss": 1.7111, + "step": 3871 + }, + { + "epoch": 1.2210659098076317, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.0278, + "step": 3872 + }, + { + "epoch": 1.2213812677388836, + "grad_norm": 0.2041015625, + "learning_rate": 0.0002, + "loss": 1.041, + "step": 3873 + }, + { + "epoch": 1.2216966256701356, + "grad_norm": 0.22265625, + "learning_rate": 0.0002, + "loss": 1.0561, + "step": 3874 + }, + { + "epoch": 1.2220119836013876, + "grad_norm": 0.2060546875, + "learning_rate": 0.0002, + "loss": 0.9379, + "step": 3875 + }, + { + "epoch": 1.2223273415326394, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.0878, + "step": 3876 + }, + { + "epoch": 1.2226426994638915, + "grad_norm": 0.201171875, + "learning_rate": 0.0002, + "loss": 0.9713, + "step": 3877 + }, + { + "epoch": 1.2229580573951435, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.0015, + "step": 3878 + }, + { + "epoch": 1.2232734153263956, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 1.0114, + "step": 3879 + }, + { + "epoch": 1.2235887732576474, + "grad_norm": 0.1943359375, + "learning_rate": 0.0002, + "loss": 0.9542, + "step": 3880 + }, + { + "epoch": 1.2239041311888994, + "grad_norm": 0.208984375, + "learning_rate": 0.0002, + "loss": 1.0427, + "step": 3881 + }, + { + "epoch": 1.2242194891201514, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.0616, + "step": 3882 + }, + { + "epoch": 1.2245348470514033, + "grad_norm": 0.2314453125, + "learning_rate": 0.0002, + "loss": 1.0749, + "step": 3883 + }, + { + "epoch": 1.2248502049826553, + "grad_norm": 0.2412109375, + "learning_rate": 0.0002, + "loss": 1.1606, + "step": 3884 + }, + { + "epoch": 1.2251655629139073, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.1106, + "step": 3885 + }, + { + "epoch": 1.2254809208451594, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 0.9611, + "step": 3886 + }, + { + "epoch": 1.2257962787764112, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.0854, + "step": 3887 + }, + { + "epoch": 1.2261116367076632, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 1.1986, + "step": 3888 + }, + { + "epoch": 1.2264269946389152, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 1.2712, + "step": 3889 + }, + { + "epoch": 1.226742352570167, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 1.0883, + "step": 3890 + }, + { + "epoch": 1.227057710501419, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 1.152, + "step": 3891 + }, + { + "epoch": 1.2273730684326711, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 1.129, + "step": 3892 + }, + { + "epoch": 1.227688426363923, + "grad_norm": 0.349609375, + "learning_rate": 0.0002, + "loss": 1.2819, + "step": 3893 + }, + { + "epoch": 1.228003784295175, + "grad_norm": 0.4140625, + "learning_rate": 0.0002, + "loss": 1.4102, + "step": 3894 + }, + { + "epoch": 1.228319142226427, + "grad_norm": 0.34765625, + "learning_rate": 0.0002, + "loss": 1.1719, + "step": 3895 + }, + { + "epoch": 1.228634500157679, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 1.205, + "step": 3896 + }, + { + "epoch": 1.2289498580889309, + "grad_norm": 0.4375, + "learning_rate": 0.0002, + "loss": 1.3834, + "step": 3897 + }, + { + "epoch": 1.229265216020183, + "grad_norm": 0.408203125, + "learning_rate": 0.0002, + "loss": 1.2695, + "step": 3898 + }, + { + "epoch": 1.229580573951435, + "grad_norm": 0.41015625, + "learning_rate": 0.0002, + "loss": 1.2544, + "step": 3899 + }, + { + "epoch": 1.2298959318826868, + "grad_norm": 0.4296875, + "learning_rate": 0.0002, + "loss": 1.2191, + "step": 3900 + }, + { + "epoch": 1.2302112898139388, + "grad_norm": 0.44140625, + "learning_rate": 0.0002, + "loss": 1.1658, + "step": 3901 + }, + { + "epoch": 1.2305266477451908, + "grad_norm": 0.62890625, + "learning_rate": 0.0002, + "loss": 1.4976, + "step": 3902 + }, + { + "epoch": 1.2308420056764429, + "grad_norm": 0.4765625, + "learning_rate": 0.0002, + "loss": 1.2911, + "step": 3903 + }, + { + "epoch": 1.2311573636076947, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 1.547, + "step": 3904 + }, + { + "epoch": 1.2314727215389467, + "grad_norm": 0.75390625, + "learning_rate": 0.0002, + "loss": 1.6103, + "step": 3905 + }, + { + "epoch": 1.2317880794701987, + "grad_norm": 0.65625, + "learning_rate": 0.0002, + "loss": 1.661, + "step": 3906 + }, + { + "epoch": 1.2321034374014506, + "grad_norm": 0.73828125, + "learning_rate": 0.0002, + "loss": 1.7969, + "step": 3907 + }, + { + "epoch": 1.2324187953327026, + "grad_norm": 0.65625, + "learning_rate": 0.0002, + "loss": 1.5114, + "step": 3908 + }, + { + "epoch": 1.2327341532639546, + "grad_norm": 0.69921875, + "learning_rate": 0.0002, + "loss": 1.8201, + "step": 3909 + }, + { + "epoch": 1.2330495111952064, + "grad_norm": 0.7578125, + "learning_rate": 0.0002, + "loss": 1.7477, + "step": 3910 + }, + { + "epoch": 1.2333648691264585, + "grad_norm": 0.98046875, + "learning_rate": 0.0002, + "loss": 2.0979, + "step": 3911 + }, + { + "epoch": 1.2336802270577105, + "grad_norm": 0.83984375, + "learning_rate": 0.0002, + "loss": 2.0863, + "step": 3912 + }, + { + "epoch": 1.2339955849889626, + "grad_norm": 0.81640625, + "learning_rate": 0.0002, + "loss": 1.9357, + "step": 3913 + }, + { + "epoch": 1.2343109429202144, + "grad_norm": 1.078125, + "learning_rate": 0.0002, + "loss": 2.1604, + "step": 3914 + }, + { + "epoch": 1.2346263008514664, + "grad_norm": 0.94921875, + "learning_rate": 0.0002, + "loss": 2.0406, + "step": 3915 + }, + { + "epoch": 1.2349416587827184, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 1.9993, + "step": 3916 + }, + { + "epoch": 1.2352570167139705, + "grad_norm": 0.89453125, + "learning_rate": 0.0002, + "loss": 2.015, + "step": 3917 + }, + { + "epoch": 1.2355723746452223, + "grad_norm": 0.8984375, + "learning_rate": 0.0002, + "loss": 1.9357, + "step": 3918 + }, + { + "epoch": 1.2358877325764743, + "grad_norm": 0.95703125, + "learning_rate": 0.0002, + "loss": 1.9648, + "step": 3919 + }, + { + "epoch": 1.2362030905077264, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.8955, + "step": 3920 + }, + { + "epoch": 1.2365184484389782, + "grad_norm": 1.15625, + "learning_rate": 0.0002, + "loss": 2.0685, + "step": 3921 + }, + { + "epoch": 1.2368338063702302, + "grad_norm": 0.205078125, + "learning_rate": 0.0002, + "loss": 1.027, + "step": 3922 + }, + { + "epoch": 1.2371491643014823, + "grad_norm": 0.2451171875, + "learning_rate": 0.0002, + "loss": 1.1742, + "step": 3923 + }, + { + "epoch": 1.237464522232734, + "grad_norm": 0.2138671875, + "learning_rate": 0.0002, + "loss": 1.0795, + "step": 3924 + }, + { + "epoch": 1.237779880163986, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 1.1182, + "step": 3925 + }, + { + "epoch": 1.2380952380952381, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.015, + "step": 3926 + }, + { + "epoch": 1.23841059602649, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 0.9747, + "step": 3927 + }, + { + "epoch": 1.23841059602649, + "eval_loss": 1.688688039779663, + "eval_runtime": 148.9702, + "eval_samples_per_second": 6.713, + "eval_steps_per_second": 6.713, + "step": 3927 + }, + { + "epoch": 1.23841059602649, + "mmlu_eval_accuracy": 0.6129802117551629, + "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, + "mmlu_eval_accuracy_anatomy": 0.5714285714285714, + "mmlu_eval_accuracy_astronomy": 0.625, + "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, + "mmlu_eval_accuracy_clinical_knowledge": 0.6896551724137931, + "mmlu_eval_accuracy_college_biology": 0.8125, + "mmlu_eval_accuracy_college_chemistry": 0.375, + "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, + "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453, + "mmlu_eval_accuracy_college_medicine": 0.6818181818181818, + "mmlu_eval_accuracy_college_physics": 0.5454545454545454, + "mmlu_eval_accuracy_computer_security": 0.5454545454545454, + "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384, + "mmlu_eval_accuracy_econometrics": 0.4166666666666667, + "mmlu_eval_accuracy_electrical_engineering": 0.5, + "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244, + "mmlu_eval_accuracy_formal_logic": 0.2857142857142857, + "mmlu_eval_accuracy_global_facts": 0.2, + "mmlu_eval_accuracy_high_school_biology": 0.625, + "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453, + "mmlu_eval_accuracy_high_school_computer_science": 0.7777777777777778, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093, + "mmlu_eval_accuracy_high_school_mathematics": 0.3448275862068966, + "mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923, + "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, + "mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333, + "mmlu_eval_accuracy_high_school_statistics": 0.5217391304347826, + "mmlu_eval_accuracy_high_school_us_history": 0.8636363636363636, + "mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693, + "mmlu_eval_accuracy_human_aging": 0.6086956521739131, + "mmlu_eval_accuracy_human_sexuality": 0.6666666666666666, + "mmlu_eval_accuracy_international_law": 0.8461538461538461, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.36363636363636365, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.76, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7325581395348837, + "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527, + "mmlu_eval_accuracy_moral_scenarios": 0.26, + "mmlu_eval_accuracy_nutrition": 0.7878787878787878, + "mmlu_eval_accuracy_philosophy": 0.7058823529411765, + "mmlu_eval_accuracy_prehistory": 0.6285714285714286, + "mmlu_eval_accuracy_professional_accounting": 0.41935483870967744, + "mmlu_eval_accuracy_professional_law": 0.38823529411764707, + "mmlu_eval_accuracy_professional_medicine": 0.7096774193548387, + "mmlu_eval_accuracy_professional_psychology": 0.6521739130434783, + "mmlu_eval_accuracy_public_relations": 0.4166666666666667, + "mmlu_eval_accuracy_security_studies": 0.7777777777777778, + "mmlu_eval_accuracy_sociology": 0.9545454545454546, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.5, + "mmlu_eval_accuracy_world_religions": 0.8421052631578947, + "mmlu_loss": 5.5499920144258645, + "step": 3927 + }, + { + "epoch": 1.238725953957742, + "grad_norm": 0.216796875, + "learning_rate": 0.0002, + "loss": 1.0287, + "step": 3928 + }, + { + "epoch": 1.239041311888994, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.0764, + "step": 3929 + }, + { + "epoch": 1.239356669820246, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 0.9531, + "step": 3930 + }, + { + "epoch": 1.2396720277514979, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.0407, + "step": 3931 + }, + { + "epoch": 1.23998738568275, + "grad_norm": 0.2294921875, + "learning_rate": 0.0002, + "loss": 1.1047, + "step": 3932 + }, + { + "epoch": 1.240302743614002, + "grad_norm": 0.2177734375, + "learning_rate": 0.0002, + "loss": 0.9963, + "step": 3933 + }, + { + "epoch": 1.240618101545254, + "grad_norm": 0.23828125, + "learning_rate": 0.0002, + "loss": 1.1053, + "step": 3934 + }, + { + "epoch": 1.2409334594765058, + "grad_norm": 0.2392578125, + "learning_rate": 0.0002, + "loss": 1.1298, + "step": 3935 + }, + { + "epoch": 1.2412488174077578, + "grad_norm": 0.232421875, + "learning_rate": 0.0002, + "loss": 1.0339, + "step": 3936 + }, + { + "epoch": 1.2415641753390099, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 0.9549, + "step": 3937 + }, + { + "epoch": 1.2418795332702617, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 1.0805, + "step": 3938 + }, + { + "epoch": 1.2421948912015137, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.1497, + "step": 3939 + }, + { + "epoch": 1.2425102491327658, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 1.2048, + "step": 3940 + }, + { + "epoch": 1.2428256070640176, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 1.2852, + "step": 3941 + }, + { + "epoch": 1.2431409649952696, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 1.2751, + "step": 3942 + }, + { + "epoch": 1.2434563229265216, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 1.3635, + "step": 3943 + }, + { + "epoch": 1.2437716808577735, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 1.2514, + "step": 3944 + }, + { + "epoch": 1.2440870387890255, + "grad_norm": 0.369140625, + "learning_rate": 0.0002, + "loss": 1.1499, + "step": 3945 + }, + { + "epoch": 1.2444023967202775, + "grad_norm": 0.4453125, + "learning_rate": 0.0002, + "loss": 1.1681, + "step": 3946 + }, + { + "epoch": 1.2447177546515296, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 1.033, + "step": 3947 + }, + { + "epoch": 1.2450331125827814, + "grad_norm": 0.42578125, + "learning_rate": 0.0002, + "loss": 1.192, + "step": 3948 + }, + { + "epoch": 1.2453484705140334, + "grad_norm": 0.423828125, + "learning_rate": 0.0002, + "loss": 1.2068, + "step": 3949 + }, + { + "epoch": 1.2456638284452854, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 3950 + }, + { + "epoch": 1.2459791863765375, + "grad_norm": 0.458984375, + "learning_rate": 0.0002, + "loss": 1.1384, + "step": 3951 + }, + { + "epoch": 1.2462945443077893, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 1.4003, + "step": 3952 + }, + { + "epoch": 1.2466099022390413, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 3953 + }, + { + "epoch": 1.2469252601702934, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 1.4554, + "step": 3954 + }, + { + "epoch": 1.2472406181015452, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 1.5254, + "step": 3955 + }, + { + "epoch": 1.2475559760327972, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 1.39, + "step": 3956 + }, + { + "epoch": 1.2478713339640493, + "grad_norm": 0.64453125, + "learning_rate": 0.0002, + "loss": 1.7214, + "step": 3957 + }, + { + "epoch": 1.248186691895301, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 1.647, + "step": 3958 + }, + { + "epoch": 1.248502049826553, + "grad_norm": 0.81640625, + "learning_rate": 0.0002, + "loss": 1.8314, + "step": 3959 + }, + { + "epoch": 1.2488174077578051, + "grad_norm": 0.67578125, + "learning_rate": 0.0002, + "loss": 1.6004, + "step": 3960 + }, + { + "epoch": 1.249132765689057, + "grad_norm": 0.70703125, + "learning_rate": 0.0002, + "loss": 1.8374, + "step": 3961 + }, + { + "epoch": 1.249448123620309, + "grad_norm": 0.7265625, + "learning_rate": 0.0002, + "loss": 1.8613, + "step": 3962 + }, + { + "epoch": 1.249763481551561, + "grad_norm": 0.82421875, + "learning_rate": 0.0002, + "loss": 2.0487, + "step": 3963 + }, + { + "epoch": 1.250078839482813, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 2.1201, + "step": 3964 + }, + { + "epoch": 1.250394197414065, + "grad_norm": 0.8984375, + "learning_rate": 0.0002, + "loss": 1.8551, + "step": 3965 + }, + { + "epoch": 1.250709555345317, + "grad_norm": 0.7265625, + "learning_rate": 0.0002, + "loss": 1.3827, + "step": 3966 + }, + { + "epoch": 1.251024913276569, + "grad_norm": 1.0390625, + "learning_rate": 0.0002, + "loss": 2.2803, + "step": 3967 + }, + { + "epoch": 1.251340271207821, + "grad_norm": 0.828125, + "learning_rate": 0.0002, + "loss": 1.5871, + "step": 3968 + }, + { + "epoch": 1.2516556291390728, + "grad_norm": 0.8359375, + "learning_rate": 0.0002, + "loss": 1.7435, + "step": 3969 + }, + { + "epoch": 1.2519709870703248, + "grad_norm": 0.8359375, + "learning_rate": 0.0002, + "loss": 1.6138, + "step": 3970 + }, + { + "epoch": 1.2522863450015769, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 2.1425, + "step": 3971 + }, + { + "epoch": 1.2526017029328287, + "grad_norm": 0.2373046875, + "learning_rate": 0.0002, + "loss": 1.1404, + "step": 3972 + }, + { + "epoch": 1.2529170608640807, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.035, + "step": 3973 + }, + { + "epoch": 1.2532324187953328, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.1398, + "step": 3974 + }, + { + "epoch": 1.2535477767265846, + "grad_norm": 0.201171875, + "learning_rate": 0.0002, + "loss": 0.9641, + "step": 3975 + }, + { + "epoch": 1.2538631346578366, + "grad_norm": 0.19921875, + "learning_rate": 0.0002, + "loss": 1.1109, + "step": 3976 + }, + { + "epoch": 1.2541784925890886, + "grad_norm": 0.2041015625, + "learning_rate": 0.0002, + "loss": 1.0471, + "step": 3977 + }, + { + "epoch": 1.2544938505203405, + "grad_norm": 0.208984375, + "learning_rate": 0.0002, + "loss": 0.9847, + "step": 3978 + }, + { + "epoch": 1.2548092084515925, + "grad_norm": 0.2138671875, + "learning_rate": 0.0002, + "loss": 1.2021, + "step": 3979 + }, + { + "epoch": 1.2551245663828445, + "grad_norm": 0.201171875, + "learning_rate": 0.0002, + "loss": 1.0054, + "step": 3980 + }, + { + "epoch": 1.2554399243140966, + "grad_norm": 0.224609375, + "learning_rate": 0.0002, + "loss": 1.0921, + "step": 3981 + }, + { + "epoch": 1.2557552822453486, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.1162, + "step": 3982 + }, + { + "epoch": 1.2560706401766004, + "grad_norm": 0.2451171875, + "learning_rate": 0.0002, + "loss": 1.0804, + "step": 3983 + }, + { + "epoch": 1.2563859981078525, + "grad_norm": 0.240234375, + "learning_rate": 0.0002, + "loss": 1.125, + "step": 3984 + }, + { + "epoch": 1.2567013560391045, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.0193, + "step": 3985 + }, + { + "epoch": 1.2570167139703563, + "grad_norm": 0.2373046875, + "learning_rate": 0.0002, + "loss": 0.987, + "step": 3986 + }, + { + "epoch": 1.2573320719016083, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.0626, + "step": 3987 + }, + { + "epoch": 1.2576474298328604, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 1.0491, + "step": 3988 + }, + { + "epoch": 1.2579627877641122, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 1.2549, + "step": 3989 + }, + { + "epoch": 1.2582781456953642, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.1796, + "step": 3990 + }, + { + "epoch": 1.2585935036266163, + "grad_norm": 0.3515625, + "learning_rate": 0.0002, + "loss": 1.0875, + "step": 3991 + }, + { + "epoch": 1.258908861557868, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.0517, + "step": 3992 + }, + { + "epoch": 1.2592242194891201, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.9354, + "step": 3993 + }, + { + "epoch": 1.2595395774203721, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 1.0635, + "step": 3994 + }, + { + "epoch": 1.259854935351624, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 0.8966, + "step": 3995 + }, + { + "epoch": 1.260170293282876, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 1.0905, + "step": 3996 + }, + { + "epoch": 1.260485651214128, + "grad_norm": 0.380859375, + "learning_rate": 0.0002, + "loss": 1.1857, + "step": 3997 + }, + { + "epoch": 1.26080100914538, + "grad_norm": 0.39453125, + "learning_rate": 0.0002, + "loss": 1.0283, + "step": 3998 + }, + { + "epoch": 1.261116367076632, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 0.9048, + "step": 3999 + }, + { + "epoch": 1.261431725007884, + "grad_norm": 0.42578125, + "learning_rate": 0.0002, + "loss": 1.1652, + "step": 4000 + }, + { + "epoch": 1.261747082939136, + "grad_norm": 0.4921875, + "learning_rate": 0.0002, + "loss": 1.123, + "step": 4001 + }, + { + "epoch": 1.262062440870388, + "grad_norm": 0.515625, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 4002 + }, + { + "epoch": 1.2623777988016398, + "grad_norm": 0.6171875, + "learning_rate": 0.0002, + "loss": 1.5713, + "step": 4003 + }, + { + "epoch": 1.2626931567328918, + "grad_norm": 0.5546875, + "learning_rate": 0.0002, + "loss": 1.3782, + "step": 4004 + }, + { + "epoch": 1.2630085146641439, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 1.1014, + "step": 4005 + }, + { + "epoch": 1.2633238725953957, + "grad_norm": 0.66796875, + "learning_rate": 0.0002, + "loss": 1.5259, + "step": 4006 + }, + { + "epoch": 1.2636392305266477, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 1.5244, + "step": 4007 + }, + { + "epoch": 1.2639545884578998, + "grad_norm": 0.75, + "learning_rate": 0.0002, + "loss": 1.9931, + "step": 4008 + }, + { + "epoch": 1.2642699463891516, + "grad_norm": 0.8984375, + "learning_rate": 0.0002, + "loss": 1.7833, + "step": 4009 + }, + { + "epoch": 1.2645853043204036, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 2.0255, + "step": 4010 + }, + { + "epoch": 1.2649006622516556, + "grad_norm": 0.9453125, + "learning_rate": 0.0002, + "loss": 1.9262, + "step": 4011 + }, + { + "epoch": 1.2652160201829075, + "grad_norm": 0.89453125, + "learning_rate": 0.0002, + "loss": 2.0055, + "step": 4012 + }, + { + "epoch": 1.2655313781141595, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 1.6929, + "step": 4013 + }, + { + "epoch": 1.2658467360454115, + "grad_norm": 0.98828125, + "learning_rate": 0.0002, + "loss": 2.0968, + "step": 4014 + }, + { + "epoch": 1.2661620939766636, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 2.1452, + "step": 4015 + }, + { + "epoch": 1.2664774519079156, + "grad_norm": 0.90625, + "learning_rate": 0.0002, + "loss": 2.1603, + "step": 4016 + }, + { + "epoch": 1.2667928098391674, + "grad_norm": 1.328125, + "learning_rate": 0.0002, + "loss": 2.4564, + "step": 4017 + }, + { + "epoch": 1.2671081677704195, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 1.9036, + "step": 4018 + }, + { + "epoch": 1.2674235257016715, + "grad_norm": 0.95703125, + "learning_rate": 0.0002, + "loss": 1.9247, + "step": 4019 + }, + { + "epoch": 1.2677388836329233, + "grad_norm": 1.25, + "learning_rate": 0.0002, + "loss": 1.8985, + "step": 4020 + }, + { + "epoch": 1.2680542415641753, + "grad_norm": 1.2421875, + "learning_rate": 0.0002, + "loss": 2.0712, + "step": 4021 + }, + { + "epoch": 1.2683695994954274, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 0.9837, + "step": 4022 + }, + { + "epoch": 1.2686849574266792, + "grad_norm": 0.2265625, + "learning_rate": 0.0002, + "loss": 1.0938, + "step": 4023 + }, + { + "epoch": 1.2690003153579312, + "grad_norm": 0.2041015625, + "learning_rate": 0.0002, + "loss": 1.0386, + "step": 4024 + }, + { + "epoch": 1.2693156732891833, + "grad_norm": 0.2216796875, + "learning_rate": 0.0002, + "loss": 1.1878, + "step": 4025 + }, + { + "epoch": 1.269631031220435, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 1.1575, + "step": 4026 + }, + { + "epoch": 1.2699463891516871, + "grad_norm": 0.193359375, + "learning_rate": 0.0002, + "loss": 1.1025, + "step": 4027 + }, + { + "epoch": 1.2702617470829392, + "grad_norm": 0.220703125, + "learning_rate": 0.0002, + "loss": 1.0692, + "step": 4028 + }, + { + "epoch": 1.2705771050141912, + "grad_norm": 0.224609375, + "learning_rate": 0.0002, + "loss": 1.1521, + "step": 4029 + }, + { + "epoch": 1.270892462945443, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 0.9738, + "step": 4030 + }, + { + "epoch": 1.271207820876695, + "grad_norm": 0.201171875, + "learning_rate": 0.0002, + "loss": 1.1049, + "step": 4031 + }, + { + "epoch": 1.271523178807947, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.1981, + "step": 4032 + }, + { + "epoch": 1.271838536739199, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.0573, + "step": 4033 + }, + { + "epoch": 1.272153894670451, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 0.9194, + "step": 4034 + }, + { + "epoch": 1.272469252601703, + "grad_norm": 0.2412109375, + "learning_rate": 0.0002, + "loss": 1.0646, + "step": 4035 + }, + { + "epoch": 1.272784610532955, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 1.2335, + "step": 4036 + }, + { + "epoch": 1.2730999684642068, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.1828, + "step": 4037 + }, + { + "epoch": 1.2734153263954588, + "grad_norm": 0.2490234375, + "learning_rate": 0.0002, + "loss": 1.0081, + "step": 4038 + }, + { + "epoch": 1.2737306843267109, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.0502, + "step": 4039 + }, + { + "epoch": 1.2740460422579627, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 1.0364, + "step": 4040 + }, + { + "epoch": 1.2743614001892147, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 1.0828, + "step": 4041 + }, + { + "epoch": 1.2746767581204668, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.0198, + "step": 4042 + }, + { + "epoch": 1.2749921160517186, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 1.109, + "step": 4043 + }, + { + "epoch": 1.2753074739829706, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 1.0491, + "step": 4044 + }, + { + "epoch": 1.2756228319142227, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 1.0937, + "step": 4045 + }, + { + "epoch": 1.2759381898454747, + "grad_norm": 0.380859375, + "learning_rate": 0.0002, + "loss": 1.4069, + "step": 4046 + }, + { + "epoch": 1.2762535477767265, + "grad_norm": 0.671875, + "learning_rate": 0.0002, + "loss": 1.4274, + "step": 4047 + }, + { + "epoch": 1.2765689057079785, + "grad_norm": 0.330078125, + "learning_rate": 0.0002, + "loss": 0.9853, + "step": 4048 + }, + { + "epoch": 1.2768842636392306, + "grad_norm": 0.431640625, + "learning_rate": 0.0002, + "loss": 1.0784, + "step": 4049 + }, + { + "epoch": 1.2771996215704826, + "grad_norm": 0.421875, + "learning_rate": 0.0002, + "loss": 1.1957, + "step": 4050 + }, + { + "epoch": 1.2775149795017344, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 1.2831, + "step": 4051 + }, + { + "epoch": 1.2778303374329865, + "grad_norm": 0.494140625, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 4052 + }, + { + "epoch": 1.2781456953642385, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 1.4027, + "step": 4053 + }, + { + "epoch": 1.2784610532954903, + "grad_norm": 0.578125, + "learning_rate": 0.0002, + "loss": 1.4154, + "step": 4054 + }, + { + "epoch": 1.2787764112267423, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 1.4135, + "step": 4055 + }, + { + "epoch": 1.2790917691579944, + "grad_norm": 0.640625, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 4056 + }, + { + "epoch": 1.2794071270892462, + "grad_norm": 0.6953125, + "learning_rate": 0.0002, + "loss": 1.7331, + "step": 4057 + }, + { + "epoch": 1.2797224850204982, + "grad_norm": 0.66796875, + "learning_rate": 0.0002, + "loss": 1.7948, + "step": 4058 + }, + { + "epoch": 1.2800378429517503, + "grad_norm": 0.828125, + "learning_rate": 0.0002, + "loss": 1.9388, + "step": 4059 + }, + { + "epoch": 1.280353200883002, + "grad_norm": 0.80859375, + "learning_rate": 0.0002, + "loss": 1.8614, + "step": 4060 + }, + { + "epoch": 1.2806685588142541, + "grad_norm": 0.7578125, + "learning_rate": 0.0002, + "loss": 1.6803, + "step": 4061 + }, + { + "epoch": 1.2809839167455062, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 1.7406, + "step": 4062 + }, + { + "epoch": 1.2812992746767582, + "grad_norm": 0.84375, + "learning_rate": 0.0002, + "loss": 2.2313, + "step": 4063 + }, + { + "epoch": 1.2816146326080102, + "grad_norm": 0.953125, + "learning_rate": 0.0002, + "loss": 2.4612, + "step": 4064 + }, + { + "epoch": 1.281929990539262, + "grad_norm": 0.83984375, + "learning_rate": 0.0002, + "loss": 1.8681, + "step": 4065 + }, + { + "epoch": 1.282245348470514, + "grad_norm": 0.8359375, + "learning_rate": 0.0002, + "loss": 1.8177, + "step": 4066 + }, + { + "epoch": 1.2825607064017661, + "grad_norm": 0.94921875, + "learning_rate": 0.0002, + "loss": 2.0126, + "step": 4067 + }, + { + "epoch": 1.282876064333018, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 2.1254, + "step": 4068 + }, + { + "epoch": 1.28319142226427, + "grad_norm": 0.84375, + "learning_rate": 0.0002, + "loss": 1.5889, + "step": 4069 + }, + { + "epoch": 1.283506780195522, + "grad_norm": 0.9453125, + "learning_rate": 0.0002, + "loss": 1.4909, + "step": 4070 + }, + { + "epoch": 1.2838221381267738, + "grad_norm": 0.9296875, + "learning_rate": 0.0002, + "loss": 1.7395, + "step": 4071 + }, + { + "epoch": 1.2841374960580259, + "grad_norm": 0.19140625, + "learning_rate": 0.0002, + "loss": 1.0573, + "step": 4072 + }, + { + "epoch": 1.2844528539892779, + "grad_norm": 0.23046875, + "learning_rate": 0.0002, + "loss": 1.0859, + "step": 4073 + }, + { + "epoch": 1.2847682119205297, + "grad_norm": 0.2177734375, + "learning_rate": 0.0002, + "loss": 1.0969, + "step": 4074 + }, + { + "epoch": 1.2850835698517817, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 1.0754, + "step": 4075 + }, + { + "epoch": 1.2853989277830338, + "grad_norm": 0.197265625, + "learning_rate": 0.0002, + "loss": 0.9641, + "step": 4076 + }, + { + "epoch": 1.2857142857142856, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.0716, + "step": 4077 + }, + { + "epoch": 1.2860296436455376, + "grad_norm": 0.2001953125, + "learning_rate": 0.0002, + "loss": 1.13, + "step": 4078 + }, + { + "epoch": 1.2863450015767897, + "grad_norm": 0.205078125, + "learning_rate": 0.0002, + "loss": 1.1152, + "step": 4079 + }, + { + "epoch": 1.2866603595080417, + "grad_norm": 0.22265625, + "learning_rate": 0.0002, + "loss": 1.0455, + "step": 4080 + }, + { + "epoch": 1.2869757174392937, + "grad_norm": 0.224609375, + "learning_rate": 0.0002, + "loss": 1.1057, + "step": 4081 + }, + { + "epoch": 1.2872910753705455, + "grad_norm": 0.216796875, + "learning_rate": 0.0002, + "loss": 1.1292, + "step": 4082 + }, + { + "epoch": 1.2876064333017976, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.1252, + "step": 4083 + }, + { + "epoch": 1.2879217912330496, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 1.0059, + "step": 4084 + }, + { + "epoch": 1.2882371491643014, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 0.9977, + "step": 4085 + }, + { + "epoch": 1.2885525070955535, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 0.9901, + "step": 4086 + }, + { + "epoch": 1.2888678650268055, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 1.1663, + "step": 4087 + }, + { + "epoch": 1.2891832229580573, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.0515, + "step": 4088 + }, + { + "epoch": 1.2894985808893094, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.2022, + "step": 4089 + }, + { + "epoch": 1.2898139388205614, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.0798, + "step": 4090 + }, + { + "epoch": 1.2901292967518132, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 1.2097, + "step": 4091 + }, + { + "epoch": 1.2904446546830652, + "grad_norm": 0.380859375, + "learning_rate": 0.0002, + "loss": 1.1344, + "step": 4092 + }, + { + "epoch": 1.2907600126143173, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 1.2008, + "step": 4093 + }, + { + "epoch": 1.291075370545569, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 1.1492, + "step": 4094 + }, + { + "epoch": 1.2913907284768211, + "grad_norm": 0.353515625, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 4095 + }, + { + "epoch": 1.2917060864080732, + "grad_norm": 0.38671875, + "learning_rate": 0.0002, + "loss": 1.2088, + "step": 4096 + }, + { + "epoch": 1.2920214443393252, + "grad_norm": 0.5, + "learning_rate": 0.0002, + "loss": 1.2878, + "step": 4097 + }, + { + "epoch": 1.2923368022705772, + "grad_norm": 0.4375, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 4098 + }, + { + "epoch": 1.292652160201829, + "grad_norm": 0.4140625, + "learning_rate": 0.0002, + "loss": 1.2715, + "step": 4099 + }, + { + "epoch": 1.292967518133081, + "grad_norm": 0.392578125, + "learning_rate": 0.0002, + "loss": 1.0752, + "step": 4100 + }, + { + "epoch": 1.2932828760643331, + "grad_norm": 0.466796875, + "learning_rate": 0.0002, + "loss": 1.2344, + "step": 4101 + }, + { + "epoch": 1.293598233995585, + "grad_norm": 0.412109375, + "learning_rate": 0.0002, + "loss": 0.9928, + "step": 4102 + }, + { + "epoch": 1.293913591926837, + "grad_norm": 0.5703125, + "learning_rate": 0.0002, + "loss": 1.5043, + "step": 4103 + }, + { + "epoch": 1.294228949858089, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 1.5472, + "step": 4104 + }, + { + "epoch": 1.2945443077893408, + "grad_norm": 0.609375, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 4105 + }, + { + "epoch": 1.2948596657205929, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 1.5578, + "step": 4106 + }, + { + "epoch": 1.295175023651845, + "grad_norm": 0.70703125, + "learning_rate": 0.0002, + "loss": 1.8303, + "step": 4107 + }, + { + "epoch": 1.2954903815830967, + "grad_norm": 0.6875, + "learning_rate": 0.0002, + "loss": 1.5888, + "step": 4108 + }, + { + "epoch": 1.2958057395143487, + "grad_norm": 0.75390625, + "learning_rate": 0.0002, + "loss": 1.8287, + "step": 4109 + }, + { + "epoch": 1.2961210974456008, + "grad_norm": 0.7578125, + "learning_rate": 0.0002, + "loss": 1.9696, + "step": 4110 + }, + { + "epoch": 1.2964364553768526, + "grad_norm": 0.84375, + "learning_rate": 0.0002, + "loss": 1.8643, + "step": 4111 + }, + { + "epoch": 1.2967518133081046, + "grad_norm": 0.71875, + "learning_rate": 0.0002, + "loss": 1.6468, + "step": 4112 + }, + { + "epoch": 1.2970671712393567, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 1.8668, + "step": 4113 + }, + { + "epoch": 1.2973825291706087, + "grad_norm": 0.859375, + "learning_rate": 0.0002, + "loss": 2.0187, + "step": 4114 + }, + { + "epoch": 1.2973825291706087, + "eval_loss": 1.6921961307525635, + "eval_runtime": 149.0229, + "eval_samples_per_second": 6.71, + "eval_steps_per_second": 6.71, + "step": 4114 + }, + { + "epoch": 1.2973825291706087, + "mmlu_eval_accuracy": 0.621168438542835, + "mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453, + "mmlu_eval_accuracy_anatomy": 0.6428571428571429, + "mmlu_eval_accuracy_astronomy": 0.625, + "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, + "mmlu_eval_accuracy_clinical_knowledge": 0.6896551724137931, + "mmlu_eval_accuracy_college_biology": 0.8125, + "mmlu_eval_accuracy_college_chemistry": 0.375, + "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, + "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454, + "mmlu_eval_accuracy_college_medicine": 0.6363636363636364, + "mmlu_eval_accuracy_college_physics": 0.6363636363636364, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.5, + "mmlu_eval_accuracy_econometrics": 0.4166666666666667, + "mmlu_eval_accuracy_electrical_engineering": 0.375, + "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634, + "mmlu_eval_accuracy_formal_logic": 0.35714285714285715, + "mmlu_eval_accuracy_global_facts": 0.3, + "mmlu_eval_accuracy_high_school_biology": 0.625, + "mmlu_eval_accuracy_high_school_chemistry": 0.5, + "mmlu_eval_accuracy_high_school_computer_science": 0.8888888888888888, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.6046511627906976, + "mmlu_eval_accuracy_high_school_mathematics": 0.3448275862068966, + "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154, + "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, + "mmlu_eval_accuracy_high_school_psychology": 0.85, + "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, + "mmlu_eval_accuracy_high_school_us_history": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693, + "mmlu_eval_accuracy_human_aging": 0.5652173913043478, + "mmlu_eval_accuracy_human_sexuality": 0.6666666666666666, + "mmlu_eval_accuracy_international_law": 1.0, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.8, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7209302325581395, + "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527, + "mmlu_eval_accuracy_moral_scenarios": 0.31, + "mmlu_eval_accuracy_nutrition": 0.7575757575757576, + "mmlu_eval_accuracy_philosophy": 0.7058823529411765, + "mmlu_eval_accuracy_prehistory": 0.6, + "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644, + "mmlu_eval_accuracy_professional_law": 0.3941176470588235, + "mmlu_eval_accuracy_professional_medicine": 0.7741935483870968, + "mmlu_eval_accuracy_professional_psychology": 0.6811594202898551, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7037037037037037, + "mmlu_eval_accuracy_sociology": 0.9545454545454546, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.4444444444444444, + "mmlu_eval_accuracy_world_religions": 0.8421052631578947, + "mmlu_loss": 5.866786452230482, + "step": 4114 + }, + { + "epoch": 1.2976978871018607, + "grad_norm": 1.1171875, + "learning_rate": 0.0002, + "loss": 2.1802, + "step": 4115 + }, + { + "epoch": 1.2980132450331126, + "grad_norm": 0.8046875, + "learning_rate": 0.0002, + "loss": 1.8696, + "step": 4116 + }, + { + "epoch": 1.2983286029643646, + "grad_norm": 0.953125, + "learning_rate": 0.0002, + "loss": 2.0689, + "step": 4117 + }, + { + "epoch": 1.2986439608956166, + "grad_norm": 0.9765625, + "learning_rate": 0.0002, + "loss": 2.0097, + "step": 4118 + }, + { + "epoch": 1.2989593188268684, + "grad_norm": 0.96484375, + "learning_rate": 0.0002, + "loss": 2.0478, + "step": 4119 + }, + { + "epoch": 1.2992746767581205, + "grad_norm": 0.82421875, + "learning_rate": 0.0002, + "loss": 1.6629, + "step": 4120 + }, + { + "epoch": 1.2995900346893725, + "grad_norm": 1.296875, + "learning_rate": 0.0002, + "loss": 2.0682, + "step": 4121 + }, + { + "epoch": 1.2999053926206243, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.0099, + "step": 4122 + }, + { + "epoch": 1.3002207505518764, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.065, + "step": 4123 + }, + { + "epoch": 1.3005361084831284, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.0131, + "step": 4124 + }, + { + "epoch": 1.3008514664143802, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 0.9857, + "step": 4125 + }, + { + "epoch": 1.3011668243456322, + "grad_norm": 0.212890625, + "learning_rate": 0.0002, + "loss": 1.0993, + "step": 4126 + }, + { + "epoch": 1.3014821822768843, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.0207, + "step": 4127 + }, + { + "epoch": 1.3017975402081363, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.0153, + "step": 4128 + }, + { + "epoch": 1.3021128981393881, + "grad_norm": 0.216796875, + "learning_rate": 0.0002, + "loss": 1.0328, + "step": 4129 + }, + { + "epoch": 1.3024282560706402, + "grad_norm": 0.208984375, + "learning_rate": 0.0002, + "loss": 0.9596, + "step": 4130 + }, + { + "epoch": 1.3027436140018922, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 1.0278, + "step": 4131 + }, + { + "epoch": 1.3030589719331442, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.015, + "step": 4132 + }, + { + "epoch": 1.303374329864396, + "grad_norm": 0.240234375, + "learning_rate": 0.0002, + "loss": 1.0663, + "step": 4133 + }, + { + "epoch": 1.303689687795648, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 1.0296, + "step": 4134 + }, + { + "epoch": 1.3040050457269001, + "grad_norm": 0.2431640625, + "learning_rate": 0.0002, + "loss": 0.934, + "step": 4135 + }, + { + "epoch": 1.304320403658152, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 1.0487, + "step": 4136 + }, + { + "epoch": 1.304635761589404, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 1.0628, + "step": 4137 + }, + { + "epoch": 1.304951119520656, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 0.9939, + "step": 4138 + }, + { + "epoch": 1.3052664774519078, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.1354, + "step": 4139 + }, + { + "epoch": 1.3055818353831599, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 1.0062, + "step": 4140 + }, + { + "epoch": 1.305897193314412, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 0.9928, + "step": 4141 + }, + { + "epoch": 1.3062125512456637, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 1.1174, + "step": 4142 + }, + { + "epoch": 1.3065279091769157, + "grad_norm": 0.3515625, + "learning_rate": 0.0002, + "loss": 1.1963, + "step": 4143 + }, + { + "epoch": 1.3068432671081678, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 1.1018, + "step": 4144 + }, + { + "epoch": 1.3071586250394198, + "grad_norm": 0.365234375, + "learning_rate": 0.0002, + "loss": 1.2841, + "step": 4145 + }, + { + "epoch": 1.3074739829706716, + "grad_norm": 0.423828125, + "learning_rate": 0.0002, + "loss": 1.201, + "step": 4146 + }, + { + "epoch": 1.3077893409019237, + "grad_norm": 0.400390625, + "learning_rate": 0.0002, + "loss": 1.1975, + "step": 4147 + }, + { + "epoch": 1.3081046988331757, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 1.1957, + "step": 4148 + }, + { + "epoch": 1.3084200567644277, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 1.1377, + "step": 4149 + }, + { + "epoch": 1.3087354146956796, + "grad_norm": 0.447265625, + "learning_rate": 0.0002, + "loss": 1.2766, + "step": 4150 + }, + { + "epoch": 1.3090507726269316, + "grad_norm": 0.443359375, + "learning_rate": 0.0002, + "loss": 1.2756, + "step": 4151 + }, + { + "epoch": 1.3093661305581836, + "grad_norm": 0.4765625, + "learning_rate": 0.0002, + "loss": 1.3968, + "step": 4152 + }, + { + "epoch": 1.3096814884894354, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 1.4846, + "step": 4153 + }, + { + "epoch": 1.3099968464206875, + "grad_norm": 0.451171875, + "learning_rate": 0.0002, + "loss": 1.2386, + "step": 4154 + }, + { + "epoch": 1.3103122043519395, + "grad_norm": 0.62890625, + "learning_rate": 0.0002, + "loss": 1.5443, + "step": 4155 + }, + { + "epoch": 1.3106275622831913, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 1.4584, + "step": 4156 + }, + { + "epoch": 1.3109429202144434, + "grad_norm": 0.640625, + "learning_rate": 0.0002, + "loss": 1.6596, + "step": 4157 + }, + { + "epoch": 1.3112582781456954, + "grad_norm": 0.734375, + "learning_rate": 0.0002, + "loss": 1.8584, + "step": 4158 + }, + { + "epoch": 1.3115736360769472, + "grad_norm": 0.75, + "learning_rate": 0.0002, + "loss": 1.8434, + "step": 4159 + }, + { + "epoch": 1.3118889940081992, + "grad_norm": 0.85546875, + "learning_rate": 0.0002, + "loss": 1.8685, + "step": 4160 + }, + { + "epoch": 1.3122043519394513, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 1.7631, + "step": 4161 + }, + { + "epoch": 1.3125197098707033, + "grad_norm": 1.296875, + "learning_rate": 0.0002, + "loss": 2.1391, + "step": 4162 + }, + { + "epoch": 1.3128350678019554, + "grad_norm": 0.75, + "learning_rate": 0.0002, + "loss": 1.816, + "step": 4163 + }, + { + "epoch": 1.3131504257332072, + "grad_norm": 0.69921875, + "learning_rate": 0.0002, + "loss": 1.6398, + "step": 4164 + }, + { + "epoch": 1.3134657836644592, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 1.8274, + "step": 4165 + }, + { + "epoch": 1.3137811415957112, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 2.0279, + "step": 4166 + }, + { + "epoch": 1.314096499526963, + "grad_norm": 0.96484375, + "learning_rate": 0.0002, + "loss": 2.2145, + "step": 4167 + }, + { + "epoch": 1.314411857458215, + "grad_norm": 1.5234375, + "learning_rate": 0.0002, + "loss": 2.3529, + "step": 4168 + }, + { + "epoch": 1.3147272153894671, + "grad_norm": 0.83203125, + "learning_rate": 0.0002, + "loss": 1.8319, + "step": 4169 + }, + { + "epoch": 1.315042573320719, + "grad_norm": 0.91015625, + "learning_rate": 0.0002, + "loss": 1.7815, + "step": 4170 + }, + { + "epoch": 1.315357931251971, + "grad_norm": 1.265625, + "learning_rate": 0.0002, + "loss": 2.5793, + "step": 4171 + }, + { + "epoch": 1.315673289183223, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.0853, + "step": 4172 + }, + { + "epoch": 1.3159886471144748, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.058, + "step": 4173 + }, + { + "epoch": 1.3163040050457269, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 0.9576, + "step": 4174 + }, + { + "epoch": 1.316619362976979, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.0589, + "step": 4175 + }, + { + "epoch": 1.3169347209082307, + "grad_norm": 0.189453125, + "learning_rate": 0.0002, + "loss": 1.0005, + "step": 4176 + }, + { + "epoch": 1.3172500788394828, + "grad_norm": 0.2060546875, + "learning_rate": 0.0002, + "loss": 1.0322, + "step": 4177 + }, + { + "epoch": 1.3175654367707348, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 0.9762, + "step": 4178 + }, + { + "epoch": 1.3178807947019868, + "grad_norm": 0.2177734375, + "learning_rate": 0.0002, + "loss": 0.9257, + "step": 4179 + }, + { + "epoch": 1.3181961526332389, + "grad_norm": 0.2265625, + "learning_rate": 0.0002, + "loss": 1.1274, + "step": 4180 + }, + { + "epoch": 1.3185115105644907, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 1.0209, + "step": 4181 + }, + { + "epoch": 1.3188268684957427, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 0.8961, + "step": 4182 + }, + { + "epoch": 1.3191422264269947, + "grad_norm": 0.2314453125, + "learning_rate": 0.0002, + "loss": 1.0346, + "step": 4183 + }, + { + "epoch": 1.3194575843582466, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 1.2251, + "step": 4184 + }, + { + "epoch": 1.3197729422894986, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.0082, + "step": 4185 + }, + { + "epoch": 1.3200883002207506, + "grad_norm": 0.2373046875, + "learning_rate": 0.0002, + "loss": 1.0333, + "step": 4186 + }, + { + "epoch": 1.3204036581520024, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 1.1151, + "step": 4187 + }, + { + "epoch": 1.3207190160832545, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 1.11, + "step": 4188 + }, + { + "epoch": 1.3210343740145065, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 1.0389, + "step": 4189 + }, + { + "epoch": 1.3213497319457583, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 1.1328, + "step": 4190 + }, + { + "epoch": 1.3216650898770104, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 1.2676, + "step": 4191 + }, + { + "epoch": 1.3219804478082624, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 1.0919, + "step": 4192 + }, + { + "epoch": 1.3222958057395142, + "grad_norm": 0.34765625, + "learning_rate": 0.0002, + "loss": 1.2259, + "step": 4193 + }, + { + "epoch": 1.3226111636707663, + "grad_norm": 0.34765625, + "learning_rate": 0.0002, + "loss": 1.1393, + "step": 4194 + }, + { + "epoch": 1.3229265216020183, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 1.088, + "step": 4195 + }, + { + "epoch": 1.3232418795332703, + "grad_norm": 0.380859375, + "learning_rate": 0.0002, + "loss": 1.1647, + "step": 4196 + }, + { + "epoch": 1.3235572374645224, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 0.9122, + "step": 4197 + }, + { + "epoch": 1.3238725953957742, + "grad_norm": 0.435546875, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 4198 + }, + { + "epoch": 1.3241879533270262, + "grad_norm": 0.421875, + "learning_rate": 0.0002, + "loss": 1.2488, + "step": 4199 + }, + { + "epoch": 1.3245033112582782, + "grad_norm": 0.48046875, + "learning_rate": 0.0002, + "loss": 1.202, + "step": 4200 + }, + { + "epoch": 1.32481866918953, + "grad_norm": 0.46875, + "learning_rate": 0.0002, + "loss": 1.2767, + "step": 4201 + }, + { + "epoch": 1.325134027120782, + "grad_norm": 0.69921875, + "learning_rate": 0.0002, + "loss": 1.57, + "step": 4202 + }, + { + "epoch": 1.3254493850520341, + "grad_norm": 0.58203125, + "learning_rate": 0.0002, + "loss": 1.4014, + "step": 4203 + }, + { + "epoch": 1.325764742983286, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 1.4048, + "step": 4204 + }, + { + "epoch": 1.326080100914538, + "grad_norm": 0.73046875, + "learning_rate": 0.0002, + "loss": 1.4213, + "step": 4205 + }, + { + "epoch": 1.32639545884579, + "grad_norm": 0.62890625, + "learning_rate": 0.0002, + "loss": 1.553, + "step": 4206 + }, + { + "epoch": 1.3267108167770418, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 1.9328, + "step": 4207 + }, + { + "epoch": 1.3270261747082939, + "grad_norm": 0.6640625, + "learning_rate": 0.0002, + "loss": 1.6678, + "step": 4208 + }, + { + "epoch": 1.327341532639546, + "grad_norm": 0.76953125, + "learning_rate": 0.0002, + "loss": 1.7438, + "step": 4209 + }, + { + "epoch": 1.3276568905707977, + "grad_norm": 0.84765625, + "learning_rate": 0.0002, + "loss": 1.9366, + "step": 4210 + }, + { + "epoch": 1.3279722485020498, + "grad_norm": 0.7890625, + "learning_rate": 0.0002, + "loss": 1.8281, + "step": 4211 + }, + { + "epoch": 1.3282876064333018, + "grad_norm": 0.91796875, + "learning_rate": 0.0002, + "loss": 2.1019, + "step": 4212 + }, + { + "epoch": 1.3286029643645538, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 2.0374, + "step": 4213 + }, + { + "epoch": 1.3289183222958059, + "grad_norm": 0.6796875, + "learning_rate": 0.0002, + "loss": 1.663, + "step": 4214 + }, + { + "epoch": 1.3292336802270577, + "grad_norm": 1.0546875, + "learning_rate": 0.0002, + "loss": 2.3074, + "step": 4215 + }, + { + "epoch": 1.3295490381583097, + "grad_norm": 0.828125, + "learning_rate": 0.0002, + "loss": 1.5526, + "step": 4216 + }, + { + "epoch": 1.3298643960895618, + "grad_norm": 0.8046875, + "learning_rate": 0.0002, + "loss": 1.9091, + "step": 4217 + }, + { + "epoch": 1.3301797540208136, + "grad_norm": 0.96484375, + "learning_rate": 0.0002, + "loss": 1.9422, + "step": 4218 + }, + { + "epoch": 1.3304951119520656, + "grad_norm": 1.15625, + "learning_rate": 0.0002, + "loss": 2.2863, + "step": 4219 + }, + { + "epoch": 1.3308104698833176, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 1.9285, + "step": 4220 + }, + { + "epoch": 1.3311258278145695, + "grad_norm": 1.25, + "learning_rate": 0.0002, + "loss": 2.3399, + "step": 4221 + }, + { + "epoch": 1.3314411857458215, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.0805, + "step": 4222 + }, + { + "epoch": 1.3317565436770735, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.0208, + "step": 4223 + }, + { + "epoch": 1.3320719016083253, + "grad_norm": 0.224609375, + "learning_rate": 0.0002, + "loss": 0.9985, + "step": 4224 + }, + { + "epoch": 1.3323872595395774, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.1054, + "step": 4225 + }, + { + "epoch": 1.3327026174708294, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.0034, + "step": 4226 + }, + { + "epoch": 1.3330179754020814, + "grad_norm": 0.1962890625, + "learning_rate": 0.0002, + "loss": 0.954, + "step": 4227 + }, + { + "epoch": 1.3333333333333333, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.1639, + "step": 4228 + }, + { + "epoch": 1.3336486912645853, + "grad_norm": 0.2373046875, + "learning_rate": 0.0002, + "loss": 0.9768, + "step": 4229 + }, + { + "epoch": 1.3339640491958373, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 1.1513, + "step": 4230 + }, + { + "epoch": 1.3342794071270894, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 1.029, + "step": 4231 + }, + { + "epoch": 1.3345947650583412, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 1.0941, + "step": 4232 + }, + { + "epoch": 1.3349101229895932, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 0.9953, + "step": 4233 + }, + { + "epoch": 1.3352254809208453, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.1481, + "step": 4234 + }, + { + "epoch": 1.335540838852097, + "grad_norm": 0.248046875, + "learning_rate": 0.0002, + "loss": 1.0742, + "step": 4235 + }, + { + "epoch": 1.335856196783349, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 0.9848, + "step": 4236 + }, + { + "epoch": 1.3361715547146011, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.0248, + "step": 4237 + }, + { + "epoch": 1.336486912645853, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.1208, + "step": 4238 + }, + { + "epoch": 1.336802270577105, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 0.9839, + "step": 4239 + }, + { + "epoch": 1.337117628508357, + "grad_norm": 0.330078125, + "learning_rate": 0.0002, + "loss": 1.2012, + "step": 4240 + }, + { + "epoch": 1.3374329864396088, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 1.1135, + "step": 4241 + }, + { + "epoch": 1.3377483443708609, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.2338, + "step": 4242 + }, + { + "epoch": 1.338063702302113, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 1.1377, + "step": 4243 + }, + { + "epoch": 1.338379060233365, + "grad_norm": 0.390625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 4244 + }, + { + "epoch": 1.3386944181646168, + "grad_norm": 0.396484375, + "learning_rate": 0.0002, + "loss": 1.0545, + "step": 4245 + }, + { + "epoch": 1.3390097760958688, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 1.0523, + "step": 4246 + }, + { + "epoch": 1.3393251340271208, + "grad_norm": 0.416015625, + "learning_rate": 0.0002, + "loss": 1.3639, + "step": 4247 + }, + { + "epoch": 1.3396404919583729, + "grad_norm": 0.390625, + "learning_rate": 0.0002, + "loss": 1.0865, + "step": 4248 + }, + { + "epoch": 1.3399558498896247, + "grad_norm": 0.484375, + "learning_rate": 0.0002, + "loss": 1.2602, + "step": 4249 + }, + { + "epoch": 1.3402712078208767, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 1.3919, + "step": 4250 + }, + { + "epoch": 1.3405865657521288, + "grad_norm": 0.4921875, + "learning_rate": 0.0002, + "loss": 1.1809, + "step": 4251 + }, + { + "epoch": 1.3409019236833806, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 1.3738, + "step": 4252 + }, + { + "epoch": 1.3412172816146326, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 1.4123, + "step": 4253 + }, + { + "epoch": 1.3415326395458846, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 1.4621, + "step": 4254 + }, + { + "epoch": 1.3418479974771365, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 1.4558, + "step": 4255 + }, + { + "epoch": 1.3421633554083885, + "grad_norm": 0.66796875, + "learning_rate": 0.0002, + "loss": 1.6271, + "step": 4256 + }, + { + "epoch": 1.3424787133396405, + "grad_norm": 0.796875, + "learning_rate": 0.0002, + "loss": 2.0049, + "step": 4257 + }, + { + "epoch": 1.3427940712708923, + "grad_norm": 0.80078125, + "learning_rate": 0.0002, + "loss": 1.8345, + "step": 4258 + }, + { + "epoch": 1.3431094292021444, + "grad_norm": 0.76953125, + "learning_rate": 0.0002, + "loss": 1.6331, + "step": 4259 + }, + { + "epoch": 1.3434247871333964, + "grad_norm": 0.76171875, + "learning_rate": 0.0002, + "loss": 1.8873, + "step": 4260 + }, + { + "epoch": 1.3437401450646485, + "grad_norm": 1.0, + "learning_rate": 0.0002, + "loss": 2.1204, + "step": 4261 + }, + { + "epoch": 1.3440555029959003, + "grad_norm": 0.80078125, + "learning_rate": 0.0002, + "loss": 1.7963, + "step": 4262 + }, + { + "epoch": 1.3443708609271523, + "grad_norm": 0.79296875, + "learning_rate": 0.0002, + "loss": 2.0084, + "step": 4263 + }, + { + "epoch": 1.3446862188584043, + "grad_norm": 0.93359375, + "learning_rate": 0.0002, + "loss": 2.574, + "step": 4264 + }, + { + "epoch": 1.3450015767896564, + "grad_norm": 1.171875, + "learning_rate": 0.0002, + "loss": 2.367, + "step": 4265 + }, + { + "epoch": 1.3453169347209082, + "grad_norm": 0.88671875, + "learning_rate": 0.0002, + "loss": 1.9784, + "step": 4266 + }, + { + "epoch": 1.3456322926521602, + "grad_norm": 1.0, + "learning_rate": 0.0002, + "loss": 1.8395, + "step": 4267 + }, + { + "epoch": 1.3459476505834123, + "grad_norm": 0.88671875, + "learning_rate": 0.0002, + "loss": 2.1321, + "step": 4268 + }, + { + "epoch": 1.346263008514664, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 2.0085, + "step": 4269 + }, + { + "epoch": 1.346578366445916, + "grad_norm": 0.9296875, + "learning_rate": 0.0002, + "loss": 2.1479, + "step": 4270 + }, + { + "epoch": 1.3468937243771681, + "grad_norm": 0.91015625, + "learning_rate": 0.0002, + "loss": 1.8994, + "step": 4271 + }, + { + "epoch": 1.34720908230842, + "grad_norm": 0.2138671875, + "learning_rate": 0.0002, + "loss": 1.0292, + "step": 4272 + }, + { + "epoch": 1.347524440239672, + "grad_norm": 0.22265625, + "learning_rate": 0.0002, + "loss": 1.1486, + "step": 4273 + }, + { + "epoch": 1.347839798170924, + "grad_norm": 0.2216796875, + "learning_rate": 0.0002, + "loss": 1.108, + "step": 4274 + }, + { + "epoch": 1.3481551561021758, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 0.9517, + "step": 4275 + }, + { + "epoch": 1.3484705140334279, + "grad_norm": 0.216796875, + "learning_rate": 0.0002, + "loss": 0.9676, + "step": 4276 + }, + { + "epoch": 1.34878587196468, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.0013, + "step": 4277 + }, + { + "epoch": 1.349101229895932, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 1.0611, + "step": 4278 + }, + { + "epoch": 1.349416587827184, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 0.9913, + "step": 4279 + }, + { + "epoch": 1.3497319457584358, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.0466, + "step": 4280 + }, + { + "epoch": 1.3500473036896878, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.1596, + "step": 4281 + }, + { + "epoch": 1.3503626616209399, + "grad_norm": 0.23046875, + "learning_rate": 0.0002, + "loss": 1.0257, + "step": 4282 + }, + { + "epoch": 1.3506780195521917, + "grad_norm": 0.2314453125, + "learning_rate": 0.0002, + "loss": 1.0754, + "step": 4283 + }, + { + "epoch": 1.3509933774834437, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 1.1318, + "step": 4284 + }, + { + "epoch": 1.3513087354146958, + "grad_norm": 0.23046875, + "learning_rate": 0.0002, + "loss": 1.0461, + "step": 4285 + }, + { + "epoch": 1.3516240933459476, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 1.0573, + "step": 4286 + }, + { + "epoch": 1.3519394512771996, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.0892, + "step": 4287 + }, + { + "epoch": 1.3522548092084516, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.1208, + "step": 4288 + }, + { + "epoch": 1.3525701671397035, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 4289 + }, + { + "epoch": 1.3528855250709555, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.1365, + "step": 4290 + }, + { + "epoch": 1.3532008830022075, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 1.1055, + "step": 4291 + }, + { + "epoch": 1.3535162409334593, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 1.0923, + "step": 4292 + }, + { + "epoch": 1.3538315988647114, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 1.1121, + "step": 4293 + }, + { + "epoch": 1.3541469567959634, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 0.9603, + "step": 4294 + }, + { + "epoch": 1.3544623147272155, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 1.0689, + "step": 4295 + }, + { + "epoch": 1.3547776726584675, + "grad_norm": 0.353515625, + "learning_rate": 0.0002, + "loss": 1.1106, + "step": 4296 + }, + { + "epoch": 1.3550930305897193, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 0.9035, + "step": 4297 + }, + { + "epoch": 1.3554083885209713, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 0.9833, + "step": 4298 + }, + { + "epoch": 1.3557237464522234, + "grad_norm": 0.384765625, + "learning_rate": 0.0002, + "loss": 0.8974, + "step": 4299 + }, + { + "epoch": 1.3560391043834752, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 1.0399, + "step": 4300 + }, + { + "epoch": 1.3563544623147272, + "grad_norm": 0.38671875, + "learning_rate": 0.0002, + "loss": 1.0511, + "step": 4301 + }, + { + "epoch": 1.3563544623147272, + "eval_loss": 1.6826621294021606, + "eval_runtime": 148.5312, + "eval_samples_per_second": 6.733, + "eval_steps_per_second": 6.733, + "step": 4301 + }, + { + "epoch": 1.3563544623147272, + "mmlu_eval_accuracy": 0.6017731550510631, + "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, + "mmlu_eval_accuracy_anatomy": 0.5714285714285714, + "mmlu_eval_accuracy_astronomy": 0.625, + "mmlu_eval_accuracy_business_ethics": 0.8181818181818182, + "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034, + "mmlu_eval_accuracy_college_biology": 0.75, + "mmlu_eval_accuracy_college_chemistry": 0.25, + "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365, + "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454, + "mmlu_eval_accuracy_college_medicine": 0.5454545454545454, + "mmlu_eval_accuracy_college_physics": 0.5454545454545454, + "mmlu_eval_accuracy_computer_security": 0.36363636363636365, + "mmlu_eval_accuracy_conceptual_physics": 0.5, + "mmlu_eval_accuracy_econometrics": 0.4166666666666667, + "mmlu_eval_accuracy_electrical_engineering": 0.4375, + "mmlu_eval_accuracy_elementary_mathematics": 0.4634146341463415, + "mmlu_eval_accuracy_formal_logic": 0.35714285714285715, + "mmlu_eval_accuracy_global_facts": 0.2, + "mmlu_eval_accuracy_high_school_biology": 0.65625, + "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.9545454545454546, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093, + "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276, + "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154, + "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, + "mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333, + "mmlu_eval_accuracy_high_school_statistics": 0.5217391304347826, + "mmlu_eval_accuracy_high_school_us_history": 0.8636363636363636, + "mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693, + "mmlu_eval_accuracy_human_aging": 0.6086956521739131, + "mmlu_eval_accuracy_human_sexuality": 0.6666666666666666, + "mmlu_eval_accuracy_international_law": 0.9230769230769231, + "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, + "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, + "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.76, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7209302325581395, + "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527, + "mmlu_eval_accuracy_moral_scenarios": 0.33, + "mmlu_eval_accuracy_nutrition": 0.6666666666666666, + "mmlu_eval_accuracy_philosophy": 0.7058823529411765, + "mmlu_eval_accuracy_prehistory": 0.5428571428571428, + "mmlu_eval_accuracy_professional_accounting": 0.3870967741935484, + "mmlu_eval_accuracy_professional_law": 0.40588235294117647, + "mmlu_eval_accuracy_professional_medicine": 0.7419354838709677, + "mmlu_eval_accuracy_professional_psychology": 0.6666666666666666, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7037037037037037, + "mmlu_eval_accuracy_sociology": 0.9090909090909091, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.3888888888888889, + "mmlu_eval_accuracy_world_religions": 0.8421052631578947, + "mmlu_loss": 6.517086007250905, + "step": 4301 + }, + { + "epoch": 1.3566698202459793, + "grad_norm": 0.56640625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 4302 + }, + { + "epoch": 1.356985178177231, + "grad_norm": 0.474609375, + "learning_rate": 0.0002, + "loss": 1.1773, + "step": 4303 + }, + { + "epoch": 1.3573005361084831, + "grad_norm": 0.55859375, + "learning_rate": 0.0002, + "loss": 1.3962, + "step": 4304 + }, + { + "epoch": 1.3576158940397351, + "grad_norm": 0.56640625, + "learning_rate": 0.0002, + "loss": 1.3913, + "step": 4305 + }, + { + "epoch": 1.357931251970987, + "grad_norm": 0.609375, + "learning_rate": 0.0002, + "loss": 1.5258, + "step": 4306 + }, + { + "epoch": 1.358246609902239, + "grad_norm": 0.6015625, + "learning_rate": 0.0002, + "loss": 1.2741, + "step": 4307 + }, + { + "epoch": 1.358561967833491, + "grad_norm": 0.734375, + "learning_rate": 0.0002, + "loss": 1.6081, + "step": 4308 + }, + { + "epoch": 1.3588773257647428, + "grad_norm": 0.77734375, + "learning_rate": 0.0002, + "loss": 1.565, + "step": 4309 + }, + { + "epoch": 1.3591926836959949, + "grad_norm": 0.91796875, + "learning_rate": 0.0002, + "loss": 2.1412, + "step": 4310 + }, + { + "epoch": 1.359508041627247, + "grad_norm": 0.74609375, + "learning_rate": 0.0002, + "loss": 1.7302, + "step": 4311 + }, + { + "epoch": 1.359823399558499, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 1.9779, + "step": 4312 + }, + { + "epoch": 1.360138757489751, + "grad_norm": 0.859375, + "learning_rate": 0.0002, + "loss": 1.9883, + "step": 4313 + }, + { + "epoch": 1.3604541154210028, + "grad_norm": 1.03125, + "learning_rate": 0.0002, + "loss": 2.0645, + "step": 4314 + }, + { + "epoch": 1.3607694733522548, + "grad_norm": 1.15625, + "learning_rate": 0.0002, + "loss": 2.406, + "step": 4315 + }, + { + "epoch": 1.3610848312835069, + "grad_norm": 1.03125, + "learning_rate": 0.0002, + "loss": 2.2128, + "step": 4316 + }, + { + "epoch": 1.3614001892147587, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 2.2877, + "step": 4317 + }, + { + "epoch": 1.3617155471460107, + "grad_norm": 1.078125, + "learning_rate": 0.0002, + "loss": 2.3516, + "step": 4318 + }, + { + "epoch": 1.3620309050772628, + "grad_norm": 0.91015625, + "learning_rate": 0.0002, + "loss": 1.7486, + "step": 4319 + }, + { + "epoch": 1.3623462630085146, + "grad_norm": 1.109375, + "learning_rate": 0.0002, + "loss": 1.8358, + "step": 4320 + }, + { + "epoch": 1.3626616209397666, + "grad_norm": 1.03125, + "learning_rate": 0.0002, + "loss": 1.949, + "step": 4321 + }, + { + "epoch": 1.3629769788710187, + "grad_norm": 0.212890625, + "learning_rate": 0.0002, + "loss": 1.0198, + "step": 4322 + }, + { + "epoch": 1.3632923368022705, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 0.9872, + "step": 4323 + }, + { + "epoch": 1.3636076947335225, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 1.0014, + "step": 4324 + }, + { + "epoch": 1.3639230526647745, + "grad_norm": 0.22265625, + "learning_rate": 0.0002, + "loss": 1.0563, + "step": 4325 + }, + { + "epoch": 1.3642384105960264, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 1.0146, + "step": 4326 + }, + { + "epoch": 1.3645537685272784, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 1.1792, + "step": 4327 + }, + { + "epoch": 1.3648691264585304, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.0483, + "step": 4328 + }, + { + "epoch": 1.3651844843897825, + "grad_norm": 0.2392578125, + "learning_rate": 0.0002, + "loss": 0.8486, + "step": 4329 + }, + { + "epoch": 1.3654998423210345, + "grad_norm": 0.2333984375, + "learning_rate": 0.0002, + "loss": 0.9984, + "step": 4330 + }, + { + "epoch": 1.3658152002522863, + "grad_norm": 0.255859375, + "learning_rate": 0.0002, + "loss": 1.1192, + "step": 4331 + }, + { + "epoch": 1.3661305581835383, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 1.1022, + "step": 4332 + }, + { + "epoch": 1.3664459161147904, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 0.9202, + "step": 4333 + }, + { + "epoch": 1.3667612740460422, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.0103, + "step": 4334 + }, + { + "epoch": 1.3670766319772942, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 1.1009, + "step": 4335 + }, + { + "epoch": 1.3673919899085463, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.1342, + "step": 4336 + }, + { + "epoch": 1.367707347839798, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 1.0758, + "step": 4337 + }, + { + "epoch": 1.3680227057710501, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.0691, + "step": 4338 + }, + { + "epoch": 1.3683380637023022, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 1.0392, + "step": 4339 + }, + { + "epoch": 1.368653421633554, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.0919, + "step": 4340 + }, + { + "epoch": 1.368968779564806, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 1.2115, + "step": 4341 + }, + { + "epoch": 1.369284137496058, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 1.0749, + "step": 4342 + }, + { + "epoch": 1.36959949542731, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 1.0345, + "step": 4343 + }, + { + "epoch": 1.369914853358562, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 1.1956, + "step": 4344 + }, + { + "epoch": 1.370230211289814, + "grad_norm": 0.412109375, + "learning_rate": 0.0002, + "loss": 1.2002, + "step": 4345 + }, + { + "epoch": 1.370545569221066, + "grad_norm": 0.369140625, + "learning_rate": 0.0002, + "loss": 1.1375, + "step": 4346 + }, + { + "epoch": 1.370860927152318, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 1.0656, + "step": 4347 + }, + { + "epoch": 1.3711762850835698, + "grad_norm": 0.453125, + "learning_rate": 0.0002, + "loss": 1.209, + "step": 4348 + }, + { + "epoch": 1.3714916430148218, + "grad_norm": 0.41796875, + "learning_rate": 0.0002, + "loss": 1.163, + "step": 4349 + }, + { + "epoch": 1.3718070009460739, + "grad_norm": 0.41015625, + "learning_rate": 0.0002, + "loss": 1.1941, + "step": 4350 + }, + { + "epoch": 1.3721223588773257, + "grad_norm": 0.390625, + "learning_rate": 0.0002, + "loss": 1.0595, + "step": 4351 + }, + { + "epoch": 1.3724377168085777, + "grad_norm": 0.61328125, + "learning_rate": 0.0002, + "loss": 1.4779, + "step": 4352 + }, + { + "epoch": 1.3727530747398298, + "grad_norm": 0.5, + "learning_rate": 0.0002, + "loss": 1.3585, + "step": 4353 + }, + { + "epoch": 1.3730684326710816, + "grad_norm": 0.56640625, + "learning_rate": 0.0002, + "loss": 1.4967, + "step": 4354 + }, + { + "epoch": 1.3733837906023336, + "grad_norm": 0.66796875, + "learning_rate": 0.0002, + "loss": 1.5343, + "step": 4355 + }, + { + "epoch": 1.3736991485335857, + "grad_norm": 0.7265625, + "learning_rate": 0.0002, + "loss": 1.743, + "step": 4356 + }, + { + "epoch": 1.3740145064648375, + "grad_norm": 0.73828125, + "learning_rate": 0.0002, + "loss": 1.7426, + "step": 4357 + }, + { + "epoch": 1.3743298643960895, + "grad_norm": 0.671875, + "learning_rate": 0.0002, + "loss": 1.4103, + "step": 4358 + }, + { + "epoch": 1.3746452223273415, + "grad_norm": 0.80859375, + "learning_rate": 0.0002, + "loss": 1.9891, + "step": 4359 + }, + { + "epoch": 1.3749605802585936, + "grad_norm": 0.875, + "learning_rate": 0.0002, + "loss": 1.6945, + "step": 4360 + }, + { + "epoch": 1.3752759381898454, + "grad_norm": 0.85546875, + "learning_rate": 0.0002, + "loss": 1.8857, + "step": 4361 + }, + { + "epoch": 1.3755912961210974, + "grad_norm": 0.97265625, + "learning_rate": 0.0002, + "loss": 2.2108, + "step": 4362 + }, + { + "epoch": 1.3759066540523495, + "grad_norm": 1.0546875, + "learning_rate": 0.0002, + "loss": 2.0102, + "step": 4363 + }, + { + "epoch": 1.3762220119836015, + "grad_norm": 0.8984375, + "learning_rate": 0.0002, + "loss": 2.0344, + "step": 4364 + }, + { + "epoch": 1.3765373699148533, + "grad_norm": 0.94140625, + "learning_rate": 0.0002, + "loss": 2.2338, + "step": 4365 + }, + { + "epoch": 1.3768527278461054, + "grad_norm": 0.9765625, + "learning_rate": 0.0002, + "loss": 2.0332, + "step": 4366 + }, + { + "epoch": 1.3771680857773574, + "grad_norm": 0.82421875, + "learning_rate": 0.0002, + "loss": 1.6323, + "step": 4367 + }, + { + "epoch": 1.3774834437086092, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 2.0952, + "step": 4368 + }, + { + "epoch": 1.3777988016398612, + "grad_norm": 0.9375, + "learning_rate": 0.0002, + "loss": 1.783, + "step": 4369 + }, + { + "epoch": 1.3781141595711133, + "grad_norm": 0.76953125, + "learning_rate": 0.0002, + "loss": 1.4155, + "step": 4370 + }, + { + "epoch": 1.378429517502365, + "grad_norm": 1.25, + "learning_rate": 0.0002, + "loss": 2.3554, + "step": 4371 + }, + { + "epoch": 1.3787448754336171, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 1.0707, + "step": 4372 + }, + { + "epoch": 1.3790602333648692, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.0014, + "step": 4373 + }, + { + "epoch": 1.379375591296121, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 0.9387, + "step": 4374 + }, + { + "epoch": 1.379690949227373, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 0.9787, + "step": 4375 + }, + { + "epoch": 1.380006307158625, + "grad_norm": 0.2177734375, + "learning_rate": 0.0002, + "loss": 1.0554, + "step": 4376 + }, + { + "epoch": 1.380321665089877, + "grad_norm": 0.1875, + "learning_rate": 0.0002, + "loss": 1.0054, + "step": 4377 + }, + { + "epoch": 1.3806370230211291, + "grad_norm": 0.2041015625, + "learning_rate": 0.0002, + "loss": 1.0478, + "step": 4378 + }, + { + "epoch": 1.380952380952381, + "grad_norm": 0.22265625, + "learning_rate": 0.0002, + "loss": 1.0971, + "step": 4379 + }, + { + "epoch": 1.381267738883633, + "grad_norm": 0.2314453125, + "learning_rate": 0.0002, + "loss": 1.0866, + "step": 4380 + }, + { + "epoch": 1.381583096814885, + "grad_norm": 0.205078125, + "learning_rate": 0.0002, + "loss": 1.0592, + "step": 4381 + }, + { + "epoch": 1.3818984547461368, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.0623, + "step": 4382 + }, + { + "epoch": 1.3822138126773889, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.0465, + "step": 4383 + }, + { + "epoch": 1.382529170608641, + "grad_norm": 0.2392578125, + "learning_rate": 0.0002, + "loss": 1.0046, + "step": 4384 + }, + { + "epoch": 1.3828445285398927, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 0.9671, + "step": 4385 + }, + { + "epoch": 1.3831598864711447, + "grad_norm": 0.240234375, + "learning_rate": 0.0002, + "loss": 1.0571, + "step": 4386 + }, + { + "epoch": 1.3834752444023968, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.0799, + "step": 4387 + }, + { + "epoch": 1.3837906023336486, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 1.0017, + "step": 4388 + }, + { + "epoch": 1.3841059602649006, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 1.1886, + "step": 4389 + }, + { + "epoch": 1.3844213181961527, + "grad_norm": 0.349609375, + "learning_rate": 0.0002, + "loss": 1.145, + "step": 4390 + }, + { + "epoch": 1.3847366761274045, + "grad_norm": 0.431640625, + "learning_rate": 0.0002, + "loss": 1.3622, + "step": 4391 + }, + { + "epoch": 1.3850520340586565, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.0242, + "step": 4392 + }, + { + "epoch": 1.3853673919899085, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 1.1062, + "step": 4393 + }, + { + "epoch": 1.3856827499211606, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 1.0136, + "step": 4394 + }, + { + "epoch": 1.3859981078524126, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 0.9781, + "step": 4395 + }, + { + "epoch": 1.3863134657836644, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 1.0675, + "step": 4396 + }, + { + "epoch": 1.3866288237149165, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 1.0804, + "step": 4397 + }, + { + "epoch": 1.3869441816461685, + "grad_norm": 0.4609375, + "learning_rate": 0.0002, + "loss": 1.257, + "step": 4398 + }, + { + "epoch": 1.3872595395774203, + "grad_norm": 0.4140625, + "learning_rate": 0.0002, + "loss": 1.1737, + "step": 4399 + }, + { + "epoch": 1.3875748975086724, + "grad_norm": 0.408203125, + "learning_rate": 0.0002, + "loss": 1.2018, + "step": 4400 + }, + { + "epoch": 1.3878902554399244, + "grad_norm": 0.41015625, + "learning_rate": 0.0002, + "loss": 1.2447, + "step": 4401 + }, + { + "epoch": 1.3882056133711762, + "grad_norm": 0.470703125, + "learning_rate": 0.0002, + "loss": 1.2448, + "step": 4402 + }, + { + "epoch": 1.3885209713024282, + "grad_norm": 0.5, + "learning_rate": 0.0002, + "loss": 1.3834, + "step": 4403 + }, + { + "epoch": 1.3888363292336803, + "grad_norm": 0.5703125, + "learning_rate": 0.0002, + "loss": 1.5216, + "step": 4404 + }, + { + "epoch": 1.389151687164932, + "grad_norm": 0.625, + "learning_rate": 0.0002, + "loss": 1.4948, + "step": 4405 + }, + { + "epoch": 1.3894670450961841, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 1.651, + "step": 4406 + }, + { + "epoch": 1.3897824030274362, + "grad_norm": 0.625, + "learning_rate": 0.0002, + "loss": 1.556, + "step": 4407 + }, + { + "epoch": 1.390097760958688, + "grad_norm": 0.828125, + "learning_rate": 0.0002, + "loss": 1.9813, + "step": 4408 + }, + { + "epoch": 1.39041311888994, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 1.8958, + "step": 4409 + }, + { + "epoch": 1.390728476821192, + "grad_norm": 0.7578125, + "learning_rate": 0.0002, + "loss": 1.9137, + "step": 4410 + }, + { + "epoch": 1.391043834752444, + "grad_norm": 0.8203125, + "learning_rate": 0.0002, + "loss": 1.7872, + "step": 4411 + }, + { + "epoch": 1.3913591926836961, + "grad_norm": 0.88671875, + "learning_rate": 0.0002, + "loss": 2.045, + "step": 4412 + }, + { + "epoch": 1.391674550614948, + "grad_norm": 1.0, + "learning_rate": 0.0002, + "loss": 2.0314, + "step": 4413 + }, + { + "epoch": 1.3919899085462, + "grad_norm": 1.125, + "learning_rate": 0.0002, + "loss": 2.0538, + "step": 4414 + }, + { + "epoch": 1.392305266477452, + "grad_norm": 0.80859375, + "learning_rate": 0.0002, + "loss": 1.8975, + "step": 4415 + }, + { + "epoch": 1.3926206244087038, + "grad_norm": 1.0859375, + "learning_rate": 0.0002, + "loss": 2.2488, + "step": 4416 + }, + { + "epoch": 1.3929359823399559, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 2.3771, + "step": 4417 + }, + { + "epoch": 1.393251340271208, + "grad_norm": 0.87109375, + "learning_rate": 0.0002, + "loss": 2.1172, + "step": 4418 + }, + { + "epoch": 1.3935666982024597, + "grad_norm": 0.9453125, + "learning_rate": 0.0002, + "loss": 1.921, + "step": 4419 + }, + { + "epoch": 1.3938820561337117, + "grad_norm": 0.87890625, + "learning_rate": 0.0002, + "loss": 1.8631, + "step": 4420 + }, + { + "epoch": 1.3941974140649638, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 1.9014, + "step": 4421 + }, + { + "epoch": 1.3945127719962156, + "grad_norm": 0.185546875, + "learning_rate": 0.0002, + "loss": 0.9741, + "step": 4422 + }, + { + "epoch": 1.3948281299274676, + "grad_norm": 0.1865234375, + "learning_rate": 0.0002, + "loss": 0.9533, + "step": 4423 + }, + { + "epoch": 1.3951434878587197, + "grad_norm": 0.201171875, + "learning_rate": 0.0002, + "loss": 1.106, + "step": 4424 + }, + { + "epoch": 1.3954588457899715, + "grad_norm": 0.205078125, + "learning_rate": 0.0002, + "loss": 1.0932, + "step": 4425 + }, + { + "epoch": 1.3957742037212235, + "grad_norm": 0.478515625, + "learning_rate": 0.0002, + "loss": 1.0636, + "step": 4426 + }, + { + "epoch": 1.3960895616524756, + "grad_norm": 0.2275390625, + "learning_rate": 0.0002, + "loss": 1.0148, + "step": 4427 + }, + { + "epoch": 1.3964049195837276, + "grad_norm": 0.2275390625, + "learning_rate": 0.0002, + "loss": 1.1459, + "step": 4428 + }, + { + "epoch": 1.3967202775149796, + "grad_norm": 0.212890625, + "learning_rate": 0.0002, + "loss": 1.0428, + "step": 4429 + }, + { + "epoch": 1.3970356354462314, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 0.9585, + "step": 4430 + }, + { + "epoch": 1.3973509933774835, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 0.9783, + "step": 4431 + }, + { + "epoch": 1.3976663513087355, + "grad_norm": 0.2177734375, + "learning_rate": 0.0002, + "loss": 0.9915, + "step": 4432 + }, + { + "epoch": 1.3979817092399873, + "grad_norm": 0.2392578125, + "learning_rate": 0.0002, + "loss": 1.0596, + "step": 4433 + }, + { + "epoch": 1.3982970671712394, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.1438, + "step": 4434 + }, + { + "epoch": 1.3986124251024914, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 1.1849, + "step": 4435 + }, + { + "epoch": 1.3989277830337432, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.0347, + "step": 4436 + }, + { + "epoch": 1.3992431409649952, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 1.1057, + "step": 4437 + }, + { + "epoch": 1.3995584988962473, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 1.0619, + "step": 4438 + }, + { + "epoch": 1.399873856827499, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 1.199, + "step": 4439 + }, + { + "epoch": 1.4001892147587511, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 1.0727, + "step": 4440 + }, + { + "epoch": 1.4005045726900032, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 1.1441, + "step": 4441 + }, + { + "epoch": 1.4008199306212552, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 1.2541, + "step": 4442 + }, + { + "epoch": 1.401135288552507, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 1.1339, + "step": 4443 + }, + { + "epoch": 1.401450646483759, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 1.2213, + "step": 4444 + }, + { + "epoch": 1.401766004415011, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 1.0417, + "step": 4445 + }, + { + "epoch": 1.4020813623462631, + "grad_norm": 0.3515625, + "learning_rate": 0.0002, + "loss": 1.0985, + "step": 4446 + }, + { + "epoch": 1.402396720277515, + "grad_norm": 0.38671875, + "learning_rate": 0.0002, + "loss": 1.1636, + "step": 4447 + }, + { + "epoch": 1.402712078208767, + "grad_norm": 0.392578125, + "learning_rate": 0.0002, + "loss": 1.1159, + "step": 4448 + }, + { + "epoch": 1.403027436140019, + "grad_norm": 0.4609375, + "learning_rate": 0.0002, + "loss": 1.5031, + "step": 4449 + }, + { + "epoch": 1.4033427940712708, + "grad_norm": 0.451171875, + "learning_rate": 0.0002, + "loss": 1.2262, + "step": 4450 + }, + { + "epoch": 1.4036581520025229, + "grad_norm": 0.423828125, + "learning_rate": 0.0002, + "loss": 1.0854, + "step": 4451 + }, + { + "epoch": 1.403973509933775, + "grad_norm": 0.4453125, + "learning_rate": 0.0002, + "loss": 1.1403, + "step": 4452 + }, + { + "epoch": 1.4042888678650267, + "grad_norm": 0.490234375, + "learning_rate": 0.0002, + "loss": 1.2948, + "step": 4453 + }, + { + "epoch": 1.4046042257962787, + "grad_norm": 0.6796875, + "learning_rate": 0.0002, + "loss": 1.6132, + "step": 4454 + }, + { + "epoch": 1.4049195837275308, + "grad_norm": 0.6171875, + "learning_rate": 0.0002, + "loss": 1.594, + "step": 4455 + }, + { + "epoch": 1.4052349416587826, + "grad_norm": 0.6484375, + "learning_rate": 0.0002, + "loss": 1.5916, + "step": 4456 + }, + { + "epoch": 1.4055502995900346, + "grad_norm": 0.734375, + "learning_rate": 0.0002, + "loss": 1.6845, + "step": 4457 + }, + { + "epoch": 1.4058656575212867, + "grad_norm": 0.83203125, + "learning_rate": 0.0002, + "loss": 1.9111, + "step": 4458 + }, + { + "epoch": 1.4061810154525387, + "grad_norm": 0.8828125, + "learning_rate": 0.0002, + "loss": 1.8942, + "step": 4459 + }, + { + "epoch": 1.4064963733837905, + "grad_norm": 0.7734375, + "learning_rate": 0.0002, + "loss": 1.809, + "step": 4460 + }, + { + "epoch": 1.4068117313150426, + "grad_norm": 0.7890625, + "learning_rate": 0.0002, + "loss": 1.6869, + "step": 4461 + }, + { + "epoch": 1.4071270892462946, + "grad_norm": 0.8359375, + "learning_rate": 0.0002, + "loss": 1.952, + "step": 4462 + }, + { + "epoch": 1.4074424471775466, + "grad_norm": 0.78125, + "learning_rate": 0.0002, + "loss": 1.6923, + "step": 4463 + }, + { + "epoch": 1.4077578051087984, + "grad_norm": 1.09375, + "learning_rate": 0.0002, + "loss": 2.2388, + "step": 4464 + }, + { + "epoch": 1.4080731630400505, + "grad_norm": 0.8828125, + "learning_rate": 0.0002, + "loss": 1.9804, + "step": 4465 + }, + { + "epoch": 1.4083885209713025, + "grad_norm": 0.859375, + "learning_rate": 0.0002, + "loss": 1.8151, + "step": 4466 + }, + { + "epoch": 1.4087038789025543, + "grad_norm": 1.2421875, + "learning_rate": 0.0002, + "loss": 2.3868, + "step": 4467 + }, + { + "epoch": 1.4090192368338064, + "grad_norm": 0.875, + "learning_rate": 0.0002, + "loss": 1.9688, + "step": 4468 + }, + { + "epoch": 1.4093345947650584, + "grad_norm": 0.9921875, + "learning_rate": 0.0002, + "loss": 1.9134, + "step": 4469 + }, + { + "epoch": 1.4096499526963102, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 1.9368, + "step": 4470 + }, + { + "epoch": 1.4099653106275623, + "grad_norm": 2.28125, + "learning_rate": 0.0002, + "loss": 1.7365, + "step": 4471 + }, + { + "epoch": 1.4102806685588143, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 1.046, + "step": 4472 + }, + { + "epoch": 1.410596026490066, + "grad_norm": 0.2001953125, + "learning_rate": 0.0002, + "loss": 1.0097, + "step": 4473 + }, + { + "epoch": 1.4109113844213181, + "grad_norm": 0.19921875, + "learning_rate": 0.0002, + "loss": 0.8856, + "step": 4474 + }, + { + "epoch": 1.4112267423525702, + "grad_norm": 0.2138671875, + "learning_rate": 0.0002, + "loss": 1.0661, + "step": 4475 + }, + { + "epoch": 1.4115421002838222, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.0125, + "step": 4476 + }, + { + "epoch": 1.4118574582150742, + "grad_norm": 0.216796875, + "learning_rate": 0.0002, + "loss": 0.9954, + "step": 4477 + }, + { + "epoch": 1.412172816146326, + "grad_norm": 0.2041015625, + "learning_rate": 0.0002, + "loss": 1.1188, + "step": 4478 + }, + { + "epoch": 1.412488174077578, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.0394, + "step": 4479 + }, + { + "epoch": 1.4128035320088301, + "grad_norm": 0.2119140625, + "learning_rate": 0.0002, + "loss": 0.9903, + "step": 4480 + }, + { + "epoch": 1.413118889940082, + "grad_norm": 0.2216796875, + "learning_rate": 0.0002, + "loss": 0.956, + "step": 4481 + }, + { + "epoch": 1.413434247871334, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.0242, + "step": 4482 + }, + { + "epoch": 1.413749605802586, + "grad_norm": 0.392578125, + "learning_rate": 0.0002, + "loss": 1.0973, + "step": 4483 + }, + { + "epoch": 1.4140649637338378, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 1.1104, + "step": 4484 + }, + { + "epoch": 1.4143803216650899, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 1.0814, + "step": 4485 + }, + { + "epoch": 1.414695679596342, + "grad_norm": 0.255859375, + "learning_rate": 0.0002, + "loss": 1.1172, + "step": 4486 + }, + { + "epoch": 1.4150110375275937, + "grad_norm": 0.255859375, + "learning_rate": 0.0002, + "loss": 1.0349, + "step": 4487 + }, + { + "epoch": 1.4153263954588458, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.0223, + "step": 4488 + }, + { + "epoch": 1.4153263954588458, + "eval_loss": 1.6728718280792236, + "eval_runtime": 148.777, + "eval_samples_per_second": 6.721, + "eval_steps_per_second": 6.721, + "step": 4488 + }, + { + "epoch": 1.4153263954588458, + "mmlu_eval_accuracy": 0.6171812624334815, + "mmlu_eval_accuracy_abstract_algebra": 0.5454545454545454, + "mmlu_eval_accuracy_anatomy": 0.42857142857142855, + "mmlu_eval_accuracy_astronomy": 0.6875, + "mmlu_eval_accuracy_business_ethics": 0.7272727272727273, + "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034, + "mmlu_eval_accuracy_college_biology": 0.75, + "mmlu_eval_accuracy_college_chemistry": 0.375, + "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, + "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365, + "mmlu_eval_accuracy_college_medicine": 0.6363636363636364, + "mmlu_eval_accuracy_college_physics": 0.5454545454545454, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.5, + "mmlu_eval_accuracy_econometrics": 0.5, + "mmlu_eval_accuracy_electrical_engineering": 0.4375, + "mmlu_eval_accuracy_elementary_mathematics": 0.5609756097560976, + "mmlu_eval_accuracy_formal_logic": 0.35714285714285715, + "mmlu_eval_accuracy_global_facts": 0.3, + "mmlu_eval_accuracy_high_school_biology": 0.65625, + "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453, + "mmlu_eval_accuracy_high_school_computer_science": 0.7777777777777778, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.9545454545454546, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093, + "mmlu_eval_accuracy_high_school_mathematics": 0.3793103448275862, + "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154, + "mmlu_eval_accuracy_high_school_physics": 0.35294117647058826, + "mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333, + "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, + "mmlu_eval_accuracy_high_school_us_history": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693, + "mmlu_eval_accuracy_human_aging": 0.6521739130434783, + "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334, + "mmlu_eval_accuracy_international_law": 0.8461538461538461, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.36363636363636365, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.8, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7093023255813954, + "mmlu_eval_accuracy_moral_disputes": 0.631578947368421, + "mmlu_eval_accuracy_moral_scenarios": 0.31, + "mmlu_eval_accuracy_nutrition": 0.7272727272727273, + "mmlu_eval_accuracy_philosophy": 0.7058823529411765, + "mmlu_eval_accuracy_prehistory": 0.6, + "mmlu_eval_accuracy_professional_accounting": 0.3225806451612903, + "mmlu_eval_accuracy_professional_law": 0.4, + "mmlu_eval_accuracy_professional_medicine": 0.7741935483870968, + "mmlu_eval_accuracy_professional_psychology": 0.6666666666666666, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7407407407407407, + "mmlu_eval_accuracy_sociology": 0.9090909090909091, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.5, + "mmlu_eval_accuracy_world_religions": 0.8421052631578947, + "mmlu_loss": 5.858649192177182, + "step": 4488 + }, + { + "epoch": 1.4156417533900978, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 1.1861, + "step": 4489 + }, + { + "epoch": 1.4159571113213496, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 1.0885, + "step": 4490 + }, + { + "epoch": 1.4162724692526016, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 1.2323, + "step": 4491 + }, + { + "epoch": 1.4165878271838537, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 1.1165, + "step": 4492 + }, + { + "epoch": 1.4169031851151057, + "grad_norm": 0.39453125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 4493 + }, + { + "epoch": 1.4172185430463577, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 1.1119, + "step": 4494 + }, + { + "epoch": 1.4175339009776096, + "grad_norm": 0.369140625, + "learning_rate": 0.0002, + "loss": 1.1858, + "step": 4495 + }, + { + "epoch": 1.4178492589088616, + "grad_norm": 0.443359375, + "learning_rate": 0.0002, + "loss": 1.1901, + "step": 4496 + }, + { + "epoch": 1.4181646168401136, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 1.4955, + "step": 4497 + }, + { + "epoch": 1.4184799747713654, + "grad_norm": 0.466796875, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 4498 + }, + { + "epoch": 1.4187953327026175, + "grad_norm": 0.474609375, + "learning_rate": 0.0002, + "loss": 1.2239, + "step": 4499 + }, + { + "epoch": 1.4191106906338695, + "grad_norm": 0.47265625, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 4500 + }, + { + "epoch": 1.4194260485651213, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 1.6875, + "step": 4501 + }, + { + "epoch": 1.4197414064963734, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 1.7289, + "step": 4502 + }, + { + "epoch": 1.4200567644276254, + "grad_norm": 0.61328125, + "learning_rate": 0.0002, + "loss": 1.408, + "step": 4503 + }, + { + "epoch": 1.4203721223588772, + "grad_norm": 0.6953125, + "learning_rate": 0.0002, + "loss": 1.6072, + "step": 4504 + }, + { + "epoch": 1.4206874802901293, + "grad_norm": 0.6796875, + "learning_rate": 0.0002, + "loss": 1.701, + "step": 4505 + }, + { + "epoch": 1.4210028382213813, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 1.4259, + "step": 4506 + }, + { + "epoch": 1.421318196152633, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 1.9002, + "step": 4507 + }, + { + "epoch": 1.4216335540838851, + "grad_norm": 0.796875, + "learning_rate": 0.0002, + "loss": 1.9911, + "step": 4508 + }, + { + "epoch": 1.4219489120151372, + "grad_norm": 0.71484375, + "learning_rate": 0.0002, + "loss": 1.7728, + "step": 4509 + }, + { + "epoch": 1.4222642699463892, + "grad_norm": 0.875, + "learning_rate": 0.0002, + "loss": 1.7444, + "step": 4510 + }, + { + "epoch": 1.4225796278776413, + "grad_norm": 0.8203125, + "learning_rate": 0.0002, + "loss": 1.8744, + "step": 4511 + }, + { + "epoch": 1.422894985808893, + "grad_norm": 0.9375, + "learning_rate": 0.0002, + "loss": 1.9841, + "step": 4512 + }, + { + "epoch": 1.423210343740145, + "grad_norm": 0.8359375, + "learning_rate": 0.0002, + "loss": 2.021, + "step": 4513 + }, + { + "epoch": 1.4235257016713971, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 2.2217, + "step": 4514 + }, + { + "epoch": 1.423841059602649, + "grad_norm": 0.93359375, + "learning_rate": 0.0002, + "loss": 1.9487, + "step": 4515 + }, + { + "epoch": 1.424156417533901, + "grad_norm": 0.7578125, + "learning_rate": 0.0002, + "loss": 1.6396, + "step": 4516 + }, + { + "epoch": 1.424471775465153, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 2.0289, + "step": 4517 + }, + { + "epoch": 1.4247871333964048, + "grad_norm": 1.34375, + "learning_rate": 0.0002, + "loss": 2.587, + "step": 4518 + }, + { + "epoch": 1.4251024913276569, + "grad_norm": 0.68359375, + "learning_rate": 0.0002, + "loss": 1.5756, + "step": 4519 + }, + { + "epoch": 1.425417849258909, + "grad_norm": 0.79296875, + "learning_rate": 0.0002, + "loss": 1.5194, + "step": 4520 + }, + { + "epoch": 1.4257332071901607, + "grad_norm": 0.9765625, + "learning_rate": 0.0002, + "loss": 2.0389, + "step": 4521 + }, + { + "epoch": 1.4260485651214128, + "grad_norm": 0.185546875, + "learning_rate": 0.0002, + "loss": 1.0661, + "step": 4522 + }, + { + "epoch": 1.4263639230526648, + "grad_norm": 0.201171875, + "learning_rate": 0.0002, + "loss": 0.9554, + "step": 4523 + }, + { + "epoch": 1.4266792809839166, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 1.0512, + "step": 4524 + }, + { + "epoch": 1.4269946389151686, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 1.0179, + "step": 4525 + }, + { + "epoch": 1.4273099968464207, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.0896, + "step": 4526 + }, + { + "epoch": 1.4276253547776727, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 0.9629, + "step": 4527 + }, + { + "epoch": 1.4279407127089248, + "grad_norm": 0.2333984375, + "learning_rate": 0.0002, + "loss": 0.9853, + "step": 4528 + }, + { + "epoch": 1.4282560706401766, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.003, + "step": 4529 + }, + { + "epoch": 1.4285714285714286, + "grad_norm": 0.2177734375, + "learning_rate": 0.0002, + "loss": 1.0138, + "step": 4530 + }, + { + "epoch": 1.4288867865026806, + "grad_norm": 0.244140625, + "learning_rate": 0.0002, + "loss": 1.132, + "step": 4531 + }, + { + "epoch": 1.4292021444339325, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 0.9784, + "step": 4532 + }, + { + "epoch": 1.4295175023651845, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 1.0083, + "step": 4533 + }, + { + "epoch": 1.4298328602964365, + "grad_norm": 0.2421875, + "learning_rate": 0.0002, + "loss": 1.0724, + "step": 4534 + }, + { + "epoch": 1.4301482182276883, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 1.0271, + "step": 4535 + }, + { + "epoch": 1.4304635761589404, + "grad_norm": 0.244140625, + "learning_rate": 0.0002, + "loss": 1.1833, + "step": 4536 + }, + { + "epoch": 1.4307789340901924, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.0933, + "step": 4537 + }, + { + "epoch": 1.4310942920214442, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 1.2475, + "step": 4538 + }, + { + "epoch": 1.4314096499526963, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 1.065, + "step": 4539 + }, + { + "epoch": 1.4317250078839483, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 1.2217, + "step": 4540 + }, + { + "epoch": 1.4320403658152003, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 1.0596, + "step": 4541 + }, + { + "epoch": 1.4323557237464521, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 1.2098, + "step": 4542 + }, + { + "epoch": 1.4326710816777042, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 1.1505, + "step": 4543 + }, + { + "epoch": 1.4329864396089562, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 1.1062, + "step": 4544 + }, + { + "epoch": 1.4333017975402083, + "grad_norm": 0.404296875, + "learning_rate": 0.0002, + "loss": 1.2475, + "step": 4545 + }, + { + "epoch": 1.43361715547146, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 0.9741, + "step": 4546 + }, + { + "epoch": 1.433932513402712, + "grad_norm": 0.423828125, + "learning_rate": 0.0002, + "loss": 1.2191, + "step": 4547 + }, + { + "epoch": 1.4342478713339641, + "grad_norm": 0.404296875, + "learning_rate": 0.0002, + "loss": 1.2921, + "step": 4548 + }, + { + "epoch": 1.434563229265216, + "grad_norm": 0.39453125, + "learning_rate": 0.0002, + "loss": 1.152, + "step": 4549 + }, + { + "epoch": 1.434878587196468, + "grad_norm": 0.3828125, + "learning_rate": 0.0002, + "loss": 1.1114, + "step": 4550 + }, + { + "epoch": 1.43519394512772, + "grad_norm": 0.51953125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 4551 + }, + { + "epoch": 1.4355093030589718, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 1.1986, + "step": 4552 + }, + { + "epoch": 1.4358246609902239, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 1.6489, + "step": 4553 + }, + { + "epoch": 1.436140018921476, + "grad_norm": 0.625, + "learning_rate": 0.0002, + "loss": 1.2491, + "step": 4554 + }, + { + "epoch": 1.4364553768527277, + "grad_norm": 0.55859375, + "learning_rate": 0.0002, + "loss": 1.4072, + "step": 4555 + }, + { + "epoch": 1.4367707347839798, + "grad_norm": 0.73828125, + "learning_rate": 0.0002, + "loss": 1.6388, + "step": 4556 + }, + { + "epoch": 1.4370860927152318, + "grad_norm": 0.6640625, + "learning_rate": 0.0002, + "loss": 1.5704, + "step": 4557 + }, + { + "epoch": 1.4374014506464838, + "grad_norm": 0.7578125, + "learning_rate": 0.0002, + "loss": 1.6793, + "step": 4558 + }, + { + "epoch": 1.4377168085777356, + "grad_norm": 0.85546875, + "learning_rate": 0.0002, + "loss": 1.9264, + "step": 4559 + }, + { + "epoch": 1.4380321665089877, + "grad_norm": 0.75390625, + "learning_rate": 0.0002, + "loss": 1.6121, + "step": 4560 + }, + { + "epoch": 1.4383475244402397, + "grad_norm": 0.84765625, + "learning_rate": 0.0002, + "loss": 1.8738, + "step": 4561 + }, + { + "epoch": 1.4386628823714918, + "grad_norm": 0.796875, + "learning_rate": 0.0002, + "loss": 1.8856, + "step": 4562 + }, + { + "epoch": 1.4389782403027436, + "grad_norm": 1.3984375, + "learning_rate": 0.0002, + "loss": 2.1862, + "step": 4563 + }, + { + "epoch": 1.4392935982339956, + "grad_norm": 0.99609375, + "learning_rate": 0.0002, + "loss": 2.1264, + "step": 4564 + }, + { + "epoch": 1.4396089561652476, + "grad_norm": 1.109375, + "learning_rate": 0.0002, + "loss": 2.2477, + "step": 4565 + }, + { + "epoch": 1.4399243140964995, + "grad_norm": 0.99609375, + "learning_rate": 0.0002, + "loss": 1.5194, + "step": 4566 + }, + { + "epoch": 1.4402396720277515, + "grad_norm": 0.85546875, + "learning_rate": 0.0002, + "loss": 1.8753, + "step": 4567 + }, + { + "epoch": 1.4405550299590035, + "grad_norm": 1.2734375, + "learning_rate": 0.0002, + "loss": 2.5804, + "step": 4568 + }, + { + "epoch": 1.4408703878902553, + "grad_norm": 0.81640625, + "learning_rate": 0.0002, + "loss": 1.6239, + "step": 4569 + }, + { + "epoch": 1.4411857458215074, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.8593, + "step": 4570 + }, + { + "epoch": 1.4415011037527594, + "grad_norm": 0.984375, + "learning_rate": 0.0002, + "loss": 1.9956, + "step": 4571 + }, + { + "epoch": 1.4418164616840112, + "grad_norm": 0.2470703125, + "learning_rate": 0.0002, + "loss": 1.1102, + "step": 4572 + }, + { + "epoch": 1.4421318196152633, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 0.9789, + "step": 4573 + }, + { + "epoch": 1.4424471775465153, + "grad_norm": 0.2216796875, + "learning_rate": 0.0002, + "loss": 1.002, + "step": 4574 + }, + { + "epoch": 1.4427625354777673, + "grad_norm": 0.2373046875, + "learning_rate": 0.0002, + "loss": 1.0959, + "step": 4575 + }, + { + "epoch": 1.4430778934090194, + "grad_norm": 0.2275390625, + "learning_rate": 0.0002, + "loss": 1.0473, + "step": 4576 + }, + { + "epoch": 1.4433932513402712, + "grad_norm": 0.2138671875, + "learning_rate": 0.0002, + "loss": 0.9857, + "step": 4577 + }, + { + "epoch": 1.4437086092715232, + "grad_norm": 0.208984375, + "learning_rate": 0.0002, + "loss": 0.8748, + "step": 4578 + }, + { + "epoch": 1.4440239672027753, + "grad_norm": 0.2392578125, + "learning_rate": 0.0002, + "loss": 1.017, + "step": 4579 + }, + { + "epoch": 1.444339325134027, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.0922, + "step": 4580 + }, + { + "epoch": 1.444654683065279, + "grad_norm": 0.208984375, + "learning_rate": 0.0002, + "loss": 0.9391, + "step": 4581 + }, + { + "epoch": 1.4449700409965311, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.0713, + "step": 4582 + }, + { + "epoch": 1.445285398927783, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 1.1264, + "step": 4583 + }, + { + "epoch": 1.445600756859035, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.1331, + "step": 4584 + }, + { + "epoch": 1.445916114790287, + "grad_norm": 0.2412109375, + "learning_rate": 0.0002, + "loss": 1.0109, + "step": 4585 + }, + { + "epoch": 1.4462314727215388, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.1398, + "step": 4586 + }, + { + "epoch": 1.4465468306527909, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.0677, + "step": 4587 + }, + { + "epoch": 1.446862188584043, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.1134, + "step": 4588 + }, + { + "epoch": 1.4471775465152947, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 1.1102, + "step": 4589 + }, + { + "epoch": 1.4474929044465468, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 1.1458, + "step": 4590 + }, + { + "epoch": 1.4478082623777988, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 1.0561, + "step": 4591 + }, + { + "epoch": 1.4481236203090508, + "grad_norm": 0.330078125, + "learning_rate": 0.0002, + "loss": 1.1853, + "step": 4592 + }, + { + "epoch": 1.4484389782403029, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 1.0945, + "step": 4593 + }, + { + "epoch": 1.4487543361715547, + "grad_norm": 0.34765625, + "learning_rate": 0.0002, + "loss": 1.0901, + "step": 4594 + }, + { + "epoch": 1.4490696941028067, + "grad_norm": 0.40234375, + "learning_rate": 0.0002, + "loss": 1.3741, + "step": 4595 + }, + { + "epoch": 1.4493850520340588, + "grad_norm": 0.39453125, + "learning_rate": 0.0002, + "loss": 1.1888, + "step": 4596 + }, + { + "epoch": 1.4497004099653106, + "grad_norm": 0.37109375, + "learning_rate": 0.0002, + "loss": 1.22, + "step": 4597 + }, + { + "epoch": 1.4500157678965626, + "grad_norm": 0.408203125, + "learning_rate": 0.0002, + "loss": 1.2567, + "step": 4598 + }, + { + "epoch": 1.4503311258278146, + "grad_norm": 0.455078125, + "learning_rate": 0.0002, + "loss": 1.0554, + "step": 4599 + }, + { + "epoch": 1.4506464837590665, + "grad_norm": 0.484375, + "learning_rate": 0.0002, + "loss": 1.2349, + "step": 4600 + }, + { + "epoch": 1.4509618416903185, + "grad_norm": 0.431640625, + "learning_rate": 0.0002, + "loss": 1.1497, + "step": 4601 + }, + { + "epoch": 1.4512771996215705, + "grad_norm": 0.484375, + "learning_rate": 0.0002, + "loss": 1.2571, + "step": 4602 + }, + { + "epoch": 1.4515925575528223, + "grad_norm": 0.4921875, + "learning_rate": 0.0002, + "loss": 1.2632, + "step": 4603 + }, + { + "epoch": 1.4519079154840744, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 1.2711, + "step": 4604 + }, + { + "epoch": 1.4522232734153264, + "grad_norm": 0.58984375, + "learning_rate": 0.0002, + "loss": 1.4488, + "step": 4605 + }, + { + "epoch": 1.4525386313465782, + "grad_norm": 0.734375, + "learning_rate": 0.0002, + "loss": 1.8626, + "step": 4606 + }, + { + "epoch": 1.4528539892778303, + "grad_norm": 0.67578125, + "learning_rate": 0.0002, + "loss": 1.5825, + "step": 4607 + }, + { + "epoch": 1.4531693472090823, + "grad_norm": 0.7421875, + "learning_rate": 0.0002, + "loss": 1.695, + "step": 4608 + }, + { + "epoch": 1.4534847051403343, + "grad_norm": 0.78125, + "learning_rate": 0.0002, + "loss": 1.773, + "step": 4609 + }, + { + "epoch": 1.4538000630715864, + "grad_norm": 0.66796875, + "learning_rate": 0.0002, + "loss": 1.4413, + "step": 4610 + }, + { + "epoch": 1.4541154210028382, + "grad_norm": 0.8359375, + "learning_rate": 0.0002, + "loss": 2.0278, + "step": 4611 + }, + { + "epoch": 1.4544307789340902, + "grad_norm": 0.71875, + "learning_rate": 0.0002, + "loss": 1.6921, + "step": 4612 + }, + { + "epoch": 1.4547461368653423, + "grad_norm": 0.87890625, + "learning_rate": 0.0002, + "loss": 2.2232, + "step": 4613 + }, + { + "epoch": 1.455061494796594, + "grad_norm": 0.9296875, + "learning_rate": 0.0002, + "loss": 2.1536, + "step": 4614 + }, + { + "epoch": 1.4553768527278461, + "grad_norm": 0.84375, + "learning_rate": 0.0002, + "loss": 1.9236, + "step": 4615 + }, + { + "epoch": 1.4556922106590982, + "grad_norm": 1.0390625, + "learning_rate": 0.0002, + "loss": 2.3195, + "step": 4616 + }, + { + "epoch": 1.45600756859035, + "grad_norm": 0.9765625, + "learning_rate": 0.0002, + "loss": 2.0157, + "step": 4617 + }, + { + "epoch": 1.456322926521602, + "grad_norm": 0.99609375, + "learning_rate": 0.0002, + "loss": 2.1003, + "step": 4618 + }, + { + "epoch": 1.456638284452854, + "grad_norm": 0.94140625, + "learning_rate": 0.0002, + "loss": 1.8467, + "step": 4619 + }, + { + "epoch": 1.4569536423841059, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 2.1719, + "step": 4620 + }, + { + "epoch": 1.4572690003153579, + "grad_norm": 1.0546875, + "learning_rate": 0.0002, + "loss": 1.6824, + "step": 4621 + }, + { + "epoch": 1.45758435824661, + "grad_norm": 0.2216796875, + "learning_rate": 0.0002, + "loss": 1.1022, + "step": 4622 + }, + { + "epoch": 1.4578997161778617, + "grad_norm": 0.2138671875, + "learning_rate": 0.0002, + "loss": 1.0118, + "step": 4623 + }, + { + "epoch": 1.4582150741091138, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 0.9892, + "step": 4624 + }, + { + "epoch": 1.4585304320403658, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 1.1964, + "step": 4625 + }, + { + "epoch": 1.4588457899716178, + "grad_norm": 0.2314453125, + "learning_rate": 0.0002, + "loss": 1.1987, + "step": 4626 + }, + { + "epoch": 1.4591611479028699, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.0109, + "step": 4627 + }, + { + "epoch": 1.4594765058341217, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.0251, + "step": 4628 + }, + { + "epoch": 1.4597918637653737, + "grad_norm": 0.2265625, + "learning_rate": 0.0002, + "loss": 1.0297, + "step": 4629 + }, + { + "epoch": 1.4601072216966258, + "grad_norm": 0.2177734375, + "learning_rate": 0.0002, + "loss": 1.0569, + "step": 4630 + }, + { + "epoch": 1.4604225796278776, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.1076, + "step": 4631 + }, + { + "epoch": 1.4607379375591296, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 0.9665, + "step": 4632 + }, + { + "epoch": 1.4610532954903817, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.0555, + "step": 4633 + }, + { + "epoch": 1.4613686534216335, + "grad_norm": 0.2177734375, + "learning_rate": 0.0002, + "loss": 0.9634, + "step": 4634 + }, + { + "epoch": 1.4616840113528855, + "grad_norm": 0.248046875, + "learning_rate": 0.0002, + "loss": 0.9885, + "step": 4635 + }, + { + "epoch": 1.4619993692841375, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 1.0609, + "step": 4636 + }, + { + "epoch": 1.4623147272153894, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.2387, + "step": 4637 + }, + { + "epoch": 1.4626300851466414, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 1.2099, + "step": 4638 + }, + { + "epoch": 1.4629454430778934, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 1.1596, + "step": 4639 + }, + { + "epoch": 1.4632608010091455, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 0.9806, + "step": 4640 + }, + { + "epoch": 1.4635761589403973, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 1.1255, + "step": 4641 + }, + { + "epoch": 1.4638915168716493, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 1.1588, + "step": 4642 + }, + { + "epoch": 1.4642068748029013, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 1.2289, + "step": 4643 + }, + { + "epoch": 1.4645222327341534, + "grad_norm": 0.37109375, + "learning_rate": 0.0002, + "loss": 1.2659, + "step": 4644 + }, + { + "epoch": 1.4648375906654052, + "grad_norm": 0.39453125, + "learning_rate": 0.0002, + "loss": 1.1432, + "step": 4645 + }, + { + "epoch": 1.4651529485966572, + "grad_norm": 0.40625, + "learning_rate": 0.0002, + "loss": 1.1855, + "step": 4646 + }, + { + "epoch": 1.4654683065279093, + "grad_norm": 0.4765625, + "learning_rate": 0.0002, + "loss": 1.3032, + "step": 4647 + }, + { + "epoch": 1.465783664459161, + "grad_norm": 0.447265625, + "learning_rate": 0.0002, + "loss": 1.2275, + "step": 4648 + }, + { + "epoch": 1.4660990223904131, + "grad_norm": 0.443359375, + "learning_rate": 0.0002, + "loss": 1.169, + "step": 4649 + }, + { + "epoch": 1.4664143803216652, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 1.5818, + "step": 4650 + }, + { + "epoch": 1.466729738252917, + "grad_norm": 0.53515625, + "learning_rate": 0.0002, + "loss": 1.3673, + "step": 4651 + }, + { + "epoch": 1.467045096184169, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 1.3735, + "step": 4652 + }, + { + "epoch": 1.467360454115421, + "grad_norm": 0.6015625, + "learning_rate": 0.0002, + "loss": 1.5098, + "step": 4653 + }, + { + "epoch": 1.4676758120466729, + "grad_norm": 0.6484375, + "learning_rate": 0.0002, + "loss": 1.5073, + "step": 4654 + }, + { + "epoch": 1.467991169977925, + "grad_norm": 0.62890625, + "learning_rate": 0.0002, + "loss": 1.4514, + "step": 4655 + }, + { + "epoch": 1.468306527909177, + "grad_norm": 0.7734375, + "learning_rate": 0.0002, + "loss": 1.7178, + "step": 4656 + }, + { + "epoch": 1.468621885840429, + "grad_norm": 0.73046875, + "learning_rate": 0.0002, + "loss": 1.5924, + "step": 4657 + }, + { + "epoch": 1.4689372437716808, + "grad_norm": 0.7265625, + "learning_rate": 0.0002, + "loss": 1.7042, + "step": 4658 + }, + { + "epoch": 1.4692526017029328, + "grad_norm": 0.8359375, + "learning_rate": 0.0002, + "loss": 1.9477, + "step": 4659 + }, + { + "epoch": 1.4695679596341849, + "grad_norm": 1.0546875, + "learning_rate": 0.0002, + "loss": 1.8947, + "step": 4660 + }, + { + "epoch": 1.4698833175654369, + "grad_norm": 0.90234375, + "learning_rate": 0.0002, + "loss": 1.8546, + "step": 4661 + }, + { + "epoch": 1.4701986754966887, + "grad_norm": 0.89453125, + "learning_rate": 0.0002, + "loss": 2.2061, + "step": 4662 + }, + { + "epoch": 1.4705140334279407, + "grad_norm": 1.0390625, + "learning_rate": 0.0002, + "loss": 2.0766, + "step": 4663 + }, + { + "epoch": 1.4708293913591928, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 2.1213, + "step": 4664 + }, + { + "epoch": 1.4711447492904446, + "grad_norm": 0.9609375, + "learning_rate": 0.0002, + "loss": 2.2562, + "step": 4665 + }, + { + "epoch": 1.4714601072216966, + "grad_norm": 0.64453125, + "learning_rate": 0.0002, + "loss": 1.447, + "step": 4666 + }, + { + "epoch": 1.4717754651529487, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 1.9399, + "step": 4667 + }, + { + "epoch": 1.4720908230842005, + "grad_norm": 2.78125, + "learning_rate": 0.0002, + "loss": 2.5316, + "step": 4668 + }, + { + "epoch": 1.4724061810154525, + "grad_norm": 0.9765625, + "learning_rate": 0.0002, + "loss": 1.9574, + "step": 4669 + }, + { + "epoch": 1.4727215389467045, + "grad_norm": 0.9375, + "learning_rate": 0.0002, + "loss": 1.6584, + "step": 4670 + }, + { + "epoch": 1.4730368968779564, + "grad_norm": 1.109375, + "learning_rate": 0.0002, + "loss": 2.0895, + "step": 4671 + }, + { + "epoch": 1.4733522548092084, + "grad_norm": 0.1884765625, + "learning_rate": 0.0002, + "loss": 1.0289, + "step": 4672 + }, + { + "epoch": 1.4736676127404604, + "grad_norm": 0.2001953125, + "learning_rate": 0.0002, + "loss": 1.0681, + "step": 4673 + }, + { + "epoch": 1.4739829706717125, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 1.0518, + "step": 4674 + }, + { + "epoch": 1.4742983286029645, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.1295, + "step": 4675 + }, + { + "epoch": 1.4742983286029645, + "eval_loss": 1.6849108934402466, + "eval_runtime": 149.0406, + "eval_samples_per_second": 6.71, + "eval_steps_per_second": 6.71, + "step": 4675 + }, + { + "epoch": 1.4742983286029645, + "mmlu_eval_accuracy": 0.5877293553531255, + "mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453, + "mmlu_eval_accuracy_anatomy": 0.42857142857142855, + "mmlu_eval_accuracy_astronomy": 0.6875, + "mmlu_eval_accuracy_business_ethics": 0.7272727272727273, + "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034, + "mmlu_eval_accuracy_college_biology": 0.75, + "mmlu_eval_accuracy_college_chemistry": 0.25, + "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365, + "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453, + "mmlu_eval_accuracy_college_medicine": 0.5909090909090909, + "mmlu_eval_accuracy_college_physics": 0.5454545454545454, + "mmlu_eval_accuracy_computer_security": 0.5454545454545454, + "mmlu_eval_accuracy_conceptual_physics": 0.5, + "mmlu_eval_accuracy_econometrics": 0.4166666666666667, + "mmlu_eval_accuracy_electrical_engineering": 0.4375, + "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244, + "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, + "mmlu_eval_accuracy_global_facts": 0.3, + "mmlu_eval_accuracy_high_school_biology": 0.625, + "mmlu_eval_accuracy_high_school_chemistry": 0.5, + "mmlu_eval_accuracy_high_school_computer_science": 0.7777777777777778, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.4883720930232558, + "mmlu_eval_accuracy_high_school_mathematics": 0.3103448275862069, + "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384, + "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, + "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667, + "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, + "mmlu_eval_accuracy_high_school_us_history": 0.8636363636363636, + "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, + "mmlu_eval_accuracy_human_aging": 0.6521739130434783, + "mmlu_eval_accuracy_human_sexuality": 0.6666666666666666, + "mmlu_eval_accuracy_international_law": 0.9230769230769231, + "mmlu_eval_accuracy_jurisprudence": 0.36363636363636365, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.36363636363636365, + "mmlu_eval_accuracy_management": 0.8181818181818182, + "mmlu_eval_accuracy_marketing": 0.72, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7325581395348837, + "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527, + "mmlu_eval_accuracy_moral_scenarios": 0.31, + "mmlu_eval_accuracy_nutrition": 0.7272727272727273, + "mmlu_eval_accuracy_philosophy": 0.6470588235294118, + "mmlu_eval_accuracy_prehistory": 0.6, + "mmlu_eval_accuracy_professional_accounting": 0.3870967741935484, + "mmlu_eval_accuracy_professional_law": 0.3941176470588235, + "mmlu_eval_accuracy_professional_medicine": 0.7096774193548387, + "mmlu_eval_accuracy_professional_psychology": 0.6231884057971014, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.6666666666666666, + "mmlu_eval_accuracy_sociology": 0.8636363636363636, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.4444444444444444, + "mmlu_eval_accuracy_world_religions": 0.7368421052631579, + "mmlu_loss": 5.6143788559463275, + "step": 4675 + }, + { + "epoch": 1.4746136865342163, + "grad_norm": 0.216796875, + "learning_rate": 0.0002, + "loss": 1.0071, + "step": 4676 + }, + { + "epoch": 1.4749290444654684, + "grad_norm": 0.205078125, + "learning_rate": 0.0002, + "loss": 1.0518, + "step": 4677 + }, + { + "epoch": 1.4752444023967204, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.0869, + "step": 4678 + }, + { + "epoch": 1.4755597603279722, + "grad_norm": 0.2412109375, + "learning_rate": 0.0002, + "loss": 1.1922, + "step": 4679 + }, + { + "epoch": 1.4758751182592242, + "grad_norm": 0.2451171875, + "learning_rate": 0.0002, + "loss": 0.941, + "step": 4680 + }, + { + "epoch": 1.4761904761904763, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.0637, + "step": 4681 + }, + { + "epoch": 1.476505834121728, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 0.977, + "step": 4682 + }, + { + "epoch": 1.4768211920529801, + "grad_norm": 0.255859375, + "learning_rate": 0.0002, + "loss": 1.1656, + "step": 4683 + }, + { + "epoch": 1.4771365499842322, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 1.1593, + "step": 4684 + }, + { + "epoch": 1.477451907915484, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 1.0991, + "step": 4685 + }, + { + "epoch": 1.477767265846736, + "grad_norm": 0.416015625, + "learning_rate": 0.0002, + "loss": 1.0992, + "step": 4686 + }, + { + "epoch": 1.478082623777988, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 0.9923, + "step": 4687 + }, + { + "epoch": 1.4783979817092399, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.0232, + "step": 4688 + }, + { + "epoch": 1.478713339640492, + "grad_norm": 0.361328125, + "learning_rate": 0.0002, + "loss": 1.081, + "step": 4689 + }, + { + "epoch": 1.479028697571744, + "grad_norm": 0.3671875, + "learning_rate": 0.0002, + "loss": 1.1544, + "step": 4690 + }, + { + "epoch": 1.479344055502996, + "grad_norm": 0.349609375, + "learning_rate": 0.0002, + "loss": 1.1597, + "step": 4691 + }, + { + "epoch": 1.479659413434248, + "grad_norm": 0.361328125, + "learning_rate": 0.0002, + "loss": 1.2016, + "step": 4692 + }, + { + "epoch": 1.4799747713654998, + "grad_norm": 0.3984375, + "learning_rate": 0.0002, + "loss": 1.068, + "step": 4693 + }, + { + "epoch": 1.4802901292967519, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 1.0318, + "step": 4694 + }, + { + "epoch": 1.480605487228004, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 1.092, + "step": 4695 + }, + { + "epoch": 1.4809208451592557, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 1.1067, + "step": 4696 + }, + { + "epoch": 1.4812362030905077, + "grad_norm": 0.37109375, + "learning_rate": 0.0002, + "loss": 1.0743, + "step": 4697 + }, + { + "epoch": 1.4815515610217598, + "grad_norm": 0.48046875, + "learning_rate": 0.0002, + "loss": 1.2069, + "step": 4698 + }, + { + "epoch": 1.4818669189530116, + "grad_norm": 0.443359375, + "learning_rate": 0.0002, + "loss": 1.1157, + "step": 4699 + }, + { + "epoch": 1.4821822768842636, + "grad_norm": 0.47265625, + "learning_rate": 0.0002, + "loss": 1.1235, + "step": 4700 + }, + { + "epoch": 1.4824976348155157, + "grad_norm": 0.5390625, + "learning_rate": 0.0002, + "loss": 1.2728, + "step": 4701 + }, + { + "epoch": 1.4828129927467675, + "grad_norm": 0.625, + "learning_rate": 0.0002, + "loss": 1.5922, + "step": 4702 + }, + { + "epoch": 1.4831283506780195, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 1.457, + "step": 4703 + }, + { + "epoch": 1.4834437086092715, + "grad_norm": 0.984375, + "learning_rate": 0.0002, + "loss": 2.0945, + "step": 4704 + }, + { + "epoch": 1.4837590665405234, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 1.5922, + "step": 4705 + }, + { + "epoch": 1.4840744244717754, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 1.4625, + "step": 4706 + }, + { + "epoch": 1.4843897824030274, + "grad_norm": 0.921875, + "learning_rate": 0.0002, + "loss": 1.8197, + "step": 4707 + }, + { + "epoch": 1.4847051403342795, + "grad_norm": 0.68359375, + "learning_rate": 0.0002, + "loss": 1.7236, + "step": 4708 + }, + { + "epoch": 1.4850204982655315, + "grad_norm": 0.78125, + "learning_rate": 0.0002, + "loss": 1.6433, + "step": 4709 + }, + { + "epoch": 1.4853358561967833, + "grad_norm": 0.796875, + "learning_rate": 0.0002, + "loss": 1.7591, + "step": 4710 + }, + { + "epoch": 1.4856512141280354, + "grad_norm": 0.84765625, + "learning_rate": 0.0002, + "loss": 1.7765, + "step": 4711 + }, + { + "epoch": 1.4859665720592874, + "grad_norm": 0.85546875, + "learning_rate": 0.0002, + "loss": 1.7114, + "step": 4712 + }, + { + "epoch": 1.4862819299905392, + "grad_norm": 0.890625, + "learning_rate": 0.0002, + "loss": 1.9791, + "step": 4713 + }, + { + "epoch": 1.4865972879217912, + "grad_norm": 1.0, + "learning_rate": 0.0002, + "loss": 1.8683, + "step": 4714 + }, + { + "epoch": 1.4869126458530433, + "grad_norm": 0.81640625, + "learning_rate": 0.0002, + "loss": 1.7602, + "step": 4715 + }, + { + "epoch": 1.487228003784295, + "grad_norm": 0.75390625, + "learning_rate": 0.0002, + "loss": 1.6055, + "step": 4716 + }, + { + "epoch": 1.4875433617155471, + "grad_norm": 0.83203125, + "learning_rate": 0.0002, + "loss": 1.9912, + "step": 4717 + }, + { + "epoch": 1.4878587196467992, + "grad_norm": 0.94140625, + "learning_rate": 0.0002, + "loss": 1.8713, + "step": 4718 + }, + { + "epoch": 1.488174077578051, + "grad_norm": 0.88671875, + "learning_rate": 0.0002, + "loss": 1.8432, + "step": 4719 + }, + { + "epoch": 1.488489435509303, + "grad_norm": 0.98828125, + "learning_rate": 0.0002, + "loss": 1.9692, + "step": 4720 + }, + { + "epoch": 1.488804793440555, + "grad_norm": 1.0546875, + "learning_rate": 0.0002, + "loss": 2.1796, + "step": 4721 + }, + { + "epoch": 1.4891201513718069, + "grad_norm": 0.1865234375, + "learning_rate": 0.0002, + "loss": 0.9966, + "step": 4722 + }, + { + "epoch": 1.489435509303059, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.0908, + "step": 4723 + }, + { + "epoch": 1.489750867234311, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 1.0508, + "step": 4724 + }, + { + "epoch": 1.490066225165563, + "grad_norm": 0.2216796875, + "learning_rate": 0.0002, + "loss": 1.0414, + "step": 4725 + }, + { + "epoch": 1.490381583096815, + "grad_norm": 0.2177734375, + "learning_rate": 0.0002, + "loss": 1.078, + "step": 4726 + }, + { + "epoch": 1.4906969410280668, + "grad_norm": 0.216796875, + "learning_rate": 0.0002, + "loss": 1.0635, + "step": 4727 + }, + { + "epoch": 1.4910122989593189, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.0308, + "step": 4728 + }, + { + "epoch": 1.491327656890571, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.1804, + "step": 4729 + }, + { + "epoch": 1.4916430148218227, + "grad_norm": 0.224609375, + "learning_rate": 0.0002, + "loss": 1.0683, + "step": 4730 + }, + { + "epoch": 1.4919583727530747, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.1273, + "step": 4731 + }, + { + "epoch": 1.4922737306843268, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.0837, + "step": 4732 + }, + { + "epoch": 1.4925890886155786, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.047, + "step": 4733 + }, + { + "epoch": 1.4929044465468306, + "grad_norm": 0.23046875, + "learning_rate": 0.0002, + "loss": 1.0479, + "step": 4734 + }, + { + "epoch": 1.4932198044780827, + "grad_norm": 0.2490234375, + "learning_rate": 0.0002, + "loss": 1.1831, + "step": 4735 + }, + { + "epoch": 1.4935351624093345, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.136, + "step": 4736 + }, + { + "epoch": 1.4938505203405865, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 1.1801, + "step": 4737 + }, + { + "epoch": 1.4941658782718386, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 1.0866, + "step": 4738 + }, + { + "epoch": 1.4944812362030906, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.1006, + "step": 4739 + }, + { + "epoch": 1.4947965941343424, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 1.1678, + "step": 4740 + }, + { + "epoch": 1.4951119520655944, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 1.0677, + "step": 4741 + }, + { + "epoch": 1.4954273099968465, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 1.0616, + "step": 4742 + }, + { + "epoch": 1.4957426679280985, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 1.1091, + "step": 4743 + }, + { + "epoch": 1.4960580258593503, + "grad_norm": 0.41015625, + "learning_rate": 0.0002, + "loss": 1.0435, + "step": 4744 + }, + { + "epoch": 1.4963733837906024, + "grad_norm": 0.416015625, + "learning_rate": 0.0002, + "loss": 1.2355, + "step": 4745 + }, + { + "epoch": 1.4966887417218544, + "grad_norm": 0.443359375, + "learning_rate": 0.0002, + "loss": 1.202, + "step": 4746 + }, + { + "epoch": 1.4970040996531062, + "grad_norm": 0.34765625, + "learning_rate": 0.0002, + "loss": 1.0438, + "step": 4747 + }, + { + "epoch": 1.4973194575843582, + "grad_norm": 0.45703125, + "learning_rate": 0.0002, + "loss": 1.2132, + "step": 4748 + }, + { + "epoch": 1.4976348155156103, + "grad_norm": 0.4609375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 4749 + }, + { + "epoch": 1.497950173446862, + "grad_norm": 0.453125, + "learning_rate": 0.0002, + "loss": 1.2474, + "step": 4750 + }, + { + "epoch": 1.4982655313781141, + "grad_norm": 0.50390625, + "learning_rate": 0.0002, + "loss": 1.3954, + "step": 4751 + }, + { + "epoch": 1.4985808893093662, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 1.5925, + "step": 4752 + }, + { + "epoch": 1.498896247240618, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 1.435, + "step": 4753 + }, + { + "epoch": 1.49921160517187, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 1.3857, + "step": 4754 + }, + { + "epoch": 1.499526963103122, + "grad_norm": 0.578125, + "learning_rate": 0.0002, + "loss": 1.3948, + "step": 4755 + }, + { + "epoch": 1.499842321034374, + "grad_norm": 0.7109375, + "learning_rate": 0.0002, + "loss": 1.7528, + "step": 4756 + }, + { + "epoch": 1.5001576789656261, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 1.6624, + "step": 4757 + }, + { + "epoch": 1.500473036896878, + "grad_norm": 0.6640625, + "learning_rate": 0.0002, + "loss": 1.6025, + "step": 4758 + }, + { + "epoch": 1.50078839482813, + "grad_norm": 0.7109375, + "learning_rate": 0.0002, + "loss": 1.4021, + "step": 4759 + }, + { + "epoch": 1.501103752759382, + "grad_norm": 0.828125, + "learning_rate": 0.0002, + "loss": 1.7796, + "step": 4760 + }, + { + "epoch": 1.5014191106906338, + "grad_norm": 0.73828125, + "learning_rate": 0.0002, + "loss": 1.7697, + "step": 4761 + }, + { + "epoch": 1.5017344686218859, + "grad_norm": 0.97265625, + "learning_rate": 0.0002, + "loss": 2.1372, + "step": 4762 + }, + { + "epoch": 1.502049826553138, + "grad_norm": 0.96875, + "learning_rate": 0.0002, + "loss": 2.3764, + "step": 4763 + }, + { + "epoch": 1.5023651844843897, + "grad_norm": 1.09375, + "learning_rate": 0.0002, + "loss": 2.2457, + "step": 4764 + }, + { + "epoch": 1.5026805424156418, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 2.0756, + "step": 4765 + }, + { + "epoch": 1.5029959003468938, + "grad_norm": 0.94921875, + "learning_rate": 0.0002, + "loss": 2.1554, + "step": 4766 + }, + { + "epoch": 1.5033112582781456, + "grad_norm": 0.890625, + "learning_rate": 0.0002, + "loss": 1.914, + "step": 4767 + }, + { + "epoch": 1.5036266162093976, + "grad_norm": 0.8515625, + "learning_rate": 0.0002, + "loss": 2.0324, + "step": 4768 + }, + { + "epoch": 1.5039419741406497, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 2.0657, + "step": 4769 + }, + { + "epoch": 1.5042573320719015, + "grad_norm": 0.84765625, + "learning_rate": 0.0002, + "loss": 1.8619, + "step": 4770 + }, + { + "epoch": 1.5045726900031537, + "grad_norm": 1.1640625, + "learning_rate": 0.0002, + "loss": 2.3292, + "step": 4771 + }, + { + "epoch": 1.5048880479344056, + "grad_norm": 0.22265625, + "learning_rate": 0.0002, + "loss": 1.0046, + "step": 4772 + }, + { + "epoch": 1.5052034058656574, + "grad_norm": 0.2060546875, + "learning_rate": 0.0002, + "loss": 0.9236, + "step": 4773 + }, + { + "epoch": 1.5055187637969096, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.0705, + "step": 4774 + }, + { + "epoch": 1.5058341217281614, + "grad_norm": 0.2119140625, + "learning_rate": 0.0002, + "loss": 0.9702, + "step": 4775 + }, + { + "epoch": 1.5061494796594135, + "grad_norm": 0.2431640625, + "learning_rate": 0.0002, + "loss": 1.136, + "step": 4776 + }, + { + "epoch": 1.5064648375906655, + "grad_norm": 0.2060546875, + "learning_rate": 0.0002, + "loss": 1.0174, + "step": 4777 + }, + { + "epoch": 1.5067801955219173, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 1.0497, + "step": 4778 + }, + { + "epoch": 1.5070955534531694, + "grad_norm": 0.22265625, + "learning_rate": 0.0002, + "loss": 1.0521, + "step": 4779 + }, + { + "epoch": 1.5074109113844214, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 1.0045, + "step": 4780 + }, + { + "epoch": 1.5077262693156732, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.0437, + "step": 4781 + }, + { + "epoch": 1.5080416272469253, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 1.1237, + "step": 4782 + }, + { + "epoch": 1.5083569851781773, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 1.1433, + "step": 4783 + }, + { + "epoch": 1.508672343109429, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.0139, + "step": 4784 + }, + { + "epoch": 1.5089877010406811, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.0207, + "step": 4785 + }, + { + "epoch": 1.5093030589719332, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.0294, + "step": 4786 + }, + { + "epoch": 1.509618416903185, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 1.1076, + "step": 4787 + }, + { + "epoch": 1.5099337748344372, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.1149, + "step": 4788 + }, + { + "epoch": 1.510249132765689, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 4789 + }, + { + "epoch": 1.5105644906969409, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 1.0579, + "step": 4790 + }, + { + "epoch": 1.5108798486281931, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 1.1109, + "step": 4791 + }, + { + "epoch": 1.511195206559445, + "grad_norm": 0.353515625, + "learning_rate": 0.0002, + "loss": 1.0726, + "step": 4792 + }, + { + "epoch": 1.511510564490697, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 1.0576, + "step": 4793 + }, + { + "epoch": 1.511825922421949, + "grad_norm": 0.388671875, + "learning_rate": 0.0002, + "loss": 1.205, + "step": 4794 + }, + { + "epoch": 1.5121412803532008, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 1.2146, + "step": 4795 + }, + { + "epoch": 1.5124566382844529, + "grad_norm": 0.39453125, + "learning_rate": 0.0002, + "loss": 1.1745, + "step": 4796 + }, + { + "epoch": 1.512771996215705, + "grad_norm": 0.4296875, + "learning_rate": 0.0002, + "loss": 1.1861, + "step": 4797 + }, + { + "epoch": 1.5130873541469567, + "grad_norm": 0.443359375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 4798 + }, + { + "epoch": 1.5134027120782088, + "grad_norm": 0.42578125, + "learning_rate": 0.0002, + "loss": 1.1038, + "step": 4799 + }, + { + "epoch": 1.5137180700094608, + "grad_norm": 0.431640625, + "learning_rate": 0.0002, + "loss": 1.0751, + "step": 4800 + }, + { + "epoch": 1.5140334279407126, + "grad_norm": 0.447265625, + "learning_rate": 0.0002, + "loss": 1.2133, + "step": 4801 + }, + { + "epoch": 1.5143487858719646, + "grad_norm": 0.6015625, + "learning_rate": 0.0002, + "loss": 1.4114, + "step": 4802 + }, + { + "epoch": 1.5146641438032167, + "grad_norm": 0.55859375, + "learning_rate": 0.0002, + "loss": 1.494, + "step": 4803 + }, + { + "epoch": 1.5149795017344685, + "grad_norm": 0.61328125, + "learning_rate": 0.0002, + "loss": 1.7893, + "step": 4804 + }, + { + "epoch": 1.5152948596657208, + "grad_norm": 0.79296875, + "learning_rate": 0.0002, + "loss": 1.6546, + "step": 4805 + }, + { + "epoch": 1.5156102175969726, + "grad_norm": 0.58984375, + "learning_rate": 0.0002, + "loss": 1.4105, + "step": 4806 + }, + { + "epoch": 1.5159255755282244, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 1.599, + "step": 4807 + }, + { + "epoch": 1.5162409334594766, + "grad_norm": 0.71484375, + "learning_rate": 0.0002, + "loss": 1.65, + "step": 4808 + }, + { + "epoch": 1.5165562913907285, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 1.897, + "step": 4809 + }, + { + "epoch": 1.5168716493219805, + "grad_norm": 0.79296875, + "learning_rate": 0.0002, + "loss": 1.9424, + "step": 4810 + }, + { + "epoch": 1.5171870072532325, + "grad_norm": 0.66796875, + "learning_rate": 0.0002, + "loss": 1.4352, + "step": 4811 + }, + { + "epoch": 1.5175023651844843, + "grad_norm": 0.859375, + "learning_rate": 0.0002, + "loss": 1.8097, + "step": 4812 + }, + { + "epoch": 1.5178177231157364, + "grad_norm": 0.859375, + "learning_rate": 0.0002, + "loss": 1.6957, + "step": 4813 + }, + { + "epoch": 1.5181330810469884, + "grad_norm": 1.1171875, + "learning_rate": 0.0002, + "loss": 2.5774, + "step": 4814 + }, + { + "epoch": 1.5184484389782402, + "grad_norm": 1.03125, + "learning_rate": 0.0002, + "loss": 2.2339, + "step": 4815 + }, + { + "epoch": 1.5187637969094923, + "grad_norm": 1.3203125, + "learning_rate": 0.0002, + "loss": 2.4075, + "step": 4816 + }, + { + "epoch": 1.5190791548407443, + "grad_norm": 1.171875, + "learning_rate": 0.0002, + "loss": 2.2405, + "step": 4817 + }, + { + "epoch": 1.519394512771996, + "grad_norm": 0.90234375, + "learning_rate": 0.0002, + "loss": 1.6981, + "step": 4818 + }, + { + "epoch": 1.5197098707032481, + "grad_norm": 1.0390625, + "learning_rate": 0.0002, + "loss": 1.9532, + "step": 4819 + }, + { + "epoch": 1.5200252286345002, + "grad_norm": 0.890625, + "learning_rate": 0.0002, + "loss": 1.7248, + "step": 4820 + }, + { + "epoch": 1.520340586565752, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 2.0696, + "step": 4821 + }, + { + "epoch": 1.5206559444970043, + "grad_norm": 0.208984375, + "learning_rate": 0.0002, + "loss": 1.0353, + "step": 4822 + }, + { + "epoch": 1.520971302428256, + "grad_norm": 0.2490234375, + "learning_rate": 0.0002, + "loss": 1.1045, + "step": 4823 + }, + { + "epoch": 1.5212866603595079, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 1.1195, + "step": 4824 + }, + { + "epoch": 1.5216020182907601, + "grad_norm": 0.248046875, + "learning_rate": 0.0002, + "loss": 1.0741, + "step": 4825 + }, + { + "epoch": 1.521917376222012, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.0196, + "step": 4826 + }, + { + "epoch": 1.522232734153264, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 0.9773, + "step": 4827 + }, + { + "epoch": 1.522548092084516, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 1.1225, + "step": 4828 + }, + { + "epoch": 1.5228634500157678, + "grad_norm": 0.2275390625, + "learning_rate": 0.0002, + "loss": 1.0588, + "step": 4829 + }, + { + "epoch": 1.5231788079470199, + "grad_norm": 0.208984375, + "learning_rate": 0.0002, + "loss": 1.0616, + "step": 4830 + }, + { + "epoch": 1.523494165878272, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.0369, + "step": 4831 + }, + { + "epoch": 1.5238095238095237, + "grad_norm": 0.2216796875, + "learning_rate": 0.0002, + "loss": 1.0868, + "step": 4832 + }, + { + "epoch": 1.5241248817407758, + "grad_norm": 0.240234375, + "learning_rate": 0.0002, + "loss": 1.0353, + "step": 4833 + }, + { + "epoch": 1.5244402396720278, + "grad_norm": 0.23828125, + "learning_rate": 0.0002, + "loss": 1.0995, + "step": 4834 + }, + { + "epoch": 1.5247555976032796, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.0366, + "step": 4835 + }, + { + "epoch": 1.5250709555345316, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.1343, + "step": 4836 + }, + { + "epoch": 1.5253863134657837, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.0707, + "step": 4837 + }, + { + "epoch": 1.5257016713970355, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 0.9798, + "step": 4838 + }, + { + "epoch": 1.5260170293282878, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 1.1584, + "step": 4839 + }, + { + "epoch": 1.5263323872595396, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 1.221, + "step": 4840 + }, + { + "epoch": 1.5266477451907914, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.0623, + "step": 4841 + }, + { + "epoch": 1.5269631031220436, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 1.1851, + "step": 4842 + }, + { + "epoch": 1.5272784610532955, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 1.241, + "step": 4843 + }, + { + "epoch": 1.5275938189845475, + "grad_norm": 0.3515625, + "learning_rate": 0.0002, + "loss": 1.1376, + "step": 4844 + }, + { + "epoch": 1.5279091769157995, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 1.2099, + "step": 4845 + }, + { + "epoch": 1.5282245348470513, + "grad_norm": 0.416015625, + "learning_rate": 0.0002, + "loss": 1.1828, + "step": 4846 + }, + { + "epoch": 1.5285398927783034, + "grad_norm": 0.423828125, + "learning_rate": 0.0002, + "loss": 1.1378, + "step": 4847 + }, + { + "epoch": 1.5288552507095554, + "grad_norm": 0.45703125, + "learning_rate": 0.0002, + "loss": 1.2095, + "step": 4848 + }, + { + "epoch": 1.5291706086408072, + "grad_norm": 0.423828125, + "learning_rate": 0.0002, + "loss": 1.1802, + "step": 4849 + }, + { + "epoch": 1.5294859665720593, + "grad_norm": 0.4453125, + "learning_rate": 0.0002, + "loss": 1.3032, + "step": 4850 + }, + { + "epoch": 1.5298013245033113, + "grad_norm": 0.443359375, + "learning_rate": 0.0002, + "loss": 1.2626, + "step": 4851 + }, + { + "epoch": 1.5301166824345631, + "grad_norm": 0.4921875, + "learning_rate": 0.0002, + "loss": 1.2134, + "step": 4852 + }, + { + "epoch": 1.5304320403658151, + "grad_norm": 0.55859375, + "learning_rate": 0.0002, + "loss": 1.5389, + "step": 4853 + }, + { + "epoch": 1.5307473982970672, + "grad_norm": 0.65625, + "learning_rate": 0.0002, + "loss": 1.5241, + "step": 4854 + }, + { + "epoch": 1.531062756228319, + "grad_norm": 0.63671875, + "learning_rate": 0.0002, + "loss": 1.3554, + "step": 4855 + }, + { + "epoch": 1.5313781141595713, + "grad_norm": 0.625, + "learning_rate": 0.0002, + "loss": 1.4263, + "step": 4856 + }, + { + "epoch": 1.531693472090823, + "grad_norm": 0.6875, + "learning_rate": 0.0002, + "loss": 1.53, + "step": 4857 + }, + { + "epoch": 1.532008830022075, + "grad_norm": 0.6484375, + "learning_rate": 0.0002, + "loss": 1.6025, + "step": 4858 + }, + { + "epoch": 1.5323241879533271, + "grad_norm": 0.70703125, + "learning_rate": 0.0002, + "loss": 1.7299, + "step": 4859 + }, + { + "epoch": 1.532639545884579, + "grad_norm": 0.703125, + "learning_rate": 0.0002, + "loss": 1.5862, + "step": 4860 + }, + { + "epoch": 1.532954903815831, + "grad_norm": 0.875, + "learning_rate": 0.0002, + "loss": 2.2062, + "step": 4861 + }, + { + "epoch": 1.533270261747083, + "grad_norm": 0.83984375, + "learning_rate": 0.0002, + "loss": 2.0358, + "step": 4862 + }, + { + "epoch": 1.533270261747083, + "eval_loss": 1.6878618001937866, + "eval_runtime": 148.9941, + "eval_samples_per_second": 6.712, + "eval_steps_per_second": 6.712, + "step": 4862 + }, + { + "epoch": 1.533270261747083, + "mmlu_eval_accuracy": 0.5925609706938609, + "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, + "mmlu_eval_accuracy_anatomy": 0.5, + "mmlu_eval_accuracy_astronomy": 0.6875, + "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, + "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034, + "mmlu_eval_accuracy_college_biology": 0.8125, + "mmlu_eval_accuracy_college_chemistry": 0.25, + "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365, + "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454, + "mmlu_eval_accuracy_college_medicine": 0.6818181818181818, + "mmlu_eval_accuracy_college_physics": 0.5454545454545454, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231, + "mmlu_eval_accuracy_econometrics": 0.4166666666666667, + "mmlu_eval_accuracy_electrical_engineering": 0.4375, + "mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637, + "mmlu_eval_accuracy_formal_logic": 0.2857142857142857, + "mmlu_eval_accuracy_global_facts": 0.2, + "mmlu_eval_accuracy_high_school_biology": 0.65625, + "mmlu_eval_accuracy_high_school_chemistry": 0.5454545454545454, + "mmlu_eval_accuracy_high_school_computer_science": 0.7777777777777778, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209, + "mmlu_eval_accuracy_high_school_mathematics": 0.3448275862068966, + "mmlu_eval_accuracy_high_school_microeconomics": 0.5, + "mmlu_eval_accuracy_high_school_physics": 0.11764705882352941, + "mmlu_eval_accuracy_high_school_psychology": 0.85, + "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, + "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, + "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, + "mmlu_eval_accuracy_human_aging": 0.6521739130434783, + "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334, + "mmlu_eval_accuracy_international_law": 0.9230769230769231, + "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, + "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, + "mmlu_eval_accuracy_machine_learning": 0.36363636363636365, + "mmlu_eval_accuracy_management": 0.8181818181818182, + "mmlu_eval_accuracy_marketing": 0.8, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7093023255813954, + "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527, + "mmlu_eval_accuracy_moral_scenarios": 0.33, + "mmlu_eval_accuracy_nutrition": 0.7272727272727273, + "mmlu_eval_accuracy_philosophy": 0.6764705882352942, + "mmlu_eval_accuracy_prehistory": 0.5428571428571428, + "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194, + "mmlu_eval_accuracy_professional_law": 0.3941176470588235, + "mmlu_eval_accuracy_professional_medicine": 0.7096774193548387, + "mmlu_eval_accuracy_professional_psychology": 0.6376811594202898, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7037037037037037, + "mmlu_eval_accuracy_sociology": 0.9090909090909091, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.5555555555555556, + "mmlu_eval_accuracy_world_religions": 0.7894736842105263, + "mmlu_loss": 6.044349235930838, + "step": 4862 + }, + { + "epoch": 1.5335856196783348, + "grad_norm": 0.87109375, + "learning_rate": 0.0002, + "loss": 1.9358, + "step": 4863 + }, + { + "epoch": 1.5339009776095869, + "grad_norm": 1.296875, + "learning_rate": 0.0002, + "loss": 2.1545, + "step": 4864 + }, + { + "epoch": 1.534216335540839, + "grad_norm": 0.91796875, + "learning_rate": 0.0002, + "loss": 1.7614, + "step": 4865 + }, + { + "epoch": 1.5345316934720907, + "grad_norm": 0.7890625, + "learning_rate": 0.0002, + "loss": 1.7934, + "step": 4866 + }, + { + "epoch": 1.5348470514033428, + "grad_norm": 0.94140625, + "learning_rate": 0.0002, + "loss": 2.1164, + "step": 4867 + }, + { + "epoch": 1.5351624093345948, + "grad_norm": 0.98828125, + "learning_rate": 0.0002, + "loss": 2.1392, + "step": 4868 + }, + { + "epoch": 1.5354777672658466, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 2.1363, + "step": 4869 + }, + { + "epoch": 1.5357931251970987, + "grad_norm": 0.99609375, + "learning_rate": 0.0002, + "loss": 2.139, + "step": 4870 + }, + { + "epoch": 1.5361084831283507, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 2.3249, + "step": 4871 + }, + { + "epoch": 1.5364238410596025, + "grad_norm": 0.197265625, + "learning_rate": 0.0002, + "loss": 1.0297, + "step": 4872 + }, + { + "epoch": 1.5367391989908548, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 1.097, + "step": 4873 + }, + { + "epoch": 1.5370545569221066, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 1.1327, + "step": 4874 + }, + { + "epoch": 1.5373699148533586, + "grad_norm": 0.2265625, + "learning_rate": 0.0002, + "loss": 1.0189, + "step": 4875 + }, + { + "epoch": 1.5376852727846106, + "grad_norm": 0.2001953125, + "learning_rate": 0.0002, + "loss": 0.9875, + "step": 4876 + }, + { + "epoch": 1.5380006307158625, + "grad_norm": 0.208984375, + "learning_rate": 0.0002, + "loss": 0.9762, + "step": 4877 + }, + { + "epoch": 1.5383159886471145, + "grad_norm": 0.2265625, + "learning_rate": 0.0002, + "loss": 1.04, + "step": 4878 + }, + { + "epoch": 1.5386313465783665, + "grad_norm": 0.2294921875, + "learning_rate": 0.0002, + "loss": 1.0601, + "step": 4879 + }, + { + "epoch": 1.5389467045096183, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.0135, + "step": 4880 + }, + { + "epoch": 1.5392620624408704, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 0.9692, + "step": 4881 + }, + { + "epoch": 1.5395774203721224, + "grad_norm": 0.2490234375, + "learning_rate": 0.0002, + "loss": 1.0412, + "step": 4882 + }, + { + "epoch": 1.5398927783033742, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 0.9452, + "step": 4883 + }, + { + "epoch": 1.5402081362346263, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 0.9963, + "step": 4884 + }, + { + "epoch": 1.5405234941658783, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 1.0512, + "step": 4885 + }, + { + "epoch": 1.5408388520971301, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 1.0463, + "step": 4886 + }, + { + "epoch": 1.5411542100283824, + "grad_norm": 0.3515625, + "learning_rate": 0.0002, + "loss": 1.2858, + "step": 4887 + }, + { + "epoch": 1.5414695679596342, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 0.9683, + "step": 4888 + }, + { + "epoch": 1.541784925890886, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 1.1652, + "step": 4889 + }, + { + "epoch": 1.5421002838221383, + "grad_norm": 0.392578125, + "learning_rate": 0.0002, + "loss": 1.252, + "step": 4890 + }, + { + "epoch": 1.54241564175339, + "grad_norm": 0.40234375, + "learning_rate": 0.0002, + "loss": 1.2512, + "step": 4891 + }, + { + "epoch": 1.5427309996846421, + "grad_norm": 0.38671875, + "learning_rate": 0.0002, + "loss": 1.2463, + "step": 4892 + }, + { + "epoch": 1.5430463576158941, + "grad_norm": 0.373046875, + "learning_rate": 0.0002, + "loss": 1.1453, + "step": 4893 + }, + { + "epoch": 1.543361715547146, + "grad_norm": 0.404296875, + "learning_rate": 0.0002, + "loss": 1.3639, + "step": 4894 + }, + { + "epoch": 1.543677073478398, + "grad_norm": 0.412109375, + "learning_rate": 0.0002, + "loss": 1.1175, + "step": 4895 + }, + { + "epoch": 1.54399243140965, + "grad_norm": 0.408203125, + "learning_rate": 0.0002, + "loss": 1.1845, + "step": 4896 + }, + { + "epoch": 1.5443077893409018, + "grad_norm": 0.40234375, + "learning_rate": 0.0002, + "loss": 1.2194, + "step": 4897 + }, + { + "epoch": 1.5446231472721539, + "grad_norm": 0.494140625, + "learning_rate": 0.0002, + "loss": 1.2105, + "step": 4898 + }, + { + "epoch": 1.544938505203406, + "grad_norm": 0.451171875, + "learning_rate": 0.0002, + "loss": 1.178, + "step": 4899 + }, + { + "epoch": 1.5452538631346577, + "grad_norm": 0.4921875, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 4900 + }, + { + "epoch": 1.5455692210659098, + "grad_norm": 0.6171875, + "learning_rate": 0.0002, + "loss": 1.4415, + "step": 4901 + }, + { + "epoch": 1.5458845789971618, + "grad_norm": 0.609375, + "learning_rate": 0.0002, + "loss": 1.6896, + "step": 4902 + }, + { + "epoch": 1.5461999369284136, + "grad_norm": 0.58203125, + "learning_rate": 0.0002, + "loss": 1.3984, + "step": 4903 + }, + { + "epoch": 1.5465152948596659, + "grad_norm": 0.62109375, + "learning_rate": 0.0002, + "loss": 1.3727, + "step": 4904 + }, + { + "epoch": 1.5468306527909177, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 1.2883, + "step": 4905 + }, + { + "epoch": 1.5471460107221695, + "grad_norm": 0.58203125, + "learning_rate": 0.0002, + "loss": 1.3999, + "step": 4906 + }, + { + "epoch": 1.5474613686534218, + "grad_norm": 0.69921875, + "learning_rate": 0.0002, + "loss": 1.5646, + "step": 4907 + }, + { + "epoch": 1.5477767265846736, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 1.7522, + "step": 4908 + }, + { + "epoch": 1.5480920845159256, + "grad_norm": 0.8359375, + "learning_rate": 0.0002, + "loss": 1.8454, + "step": 4909 + }, + { + "epoch": 1.5484074424471777, + "grad_norm": 0.83203125, + "learning_rate": 0.0002, + "loss": 2.0066, + "step": 4910 + }, + { + "epoch": 1.5487228003784295, + "grad_norm": 0.84375, + "learning_rate": 0.0002, + "loss": 1.9689, + "step": 4911 + }, + { + "epoch": 1.5490381583096815, + "grad_norm": 0.94921875, + "learning_rate": 0.0002, + "loss": 2.1262, + "step": 4912 + }, + { + "epoch": 1.5493535162409335, + "grad_norm": 1.0390625, + "learning_rate": 0.0002, + "loss": 2.342, + "step": 4913 + }, + { + "epoch": 1.5496688741721854, + "grad_norm": 1.125, + "learning_rate": 0.0002, + "loss": 2.4227, + "step": 4914 + }, + { + "epoch": 1.5499842321034374, + "grad_norm": 1.2109375, + "learning_rate": 0.0002, + "loss": 2.2215, + "step": 4915 + }, + { + "epoch": 1.5502995900346894, + "grad_norm": 1.0, + "learning_rate": 0.0002, + "loss": 2.0459, + "step": 4916 + }, + { + "epoch": 1.5506149479659412, + "grad_norm": 1.015625, + "learning_rate": 0.0002, + "loss": 1.868, + "step": 4917 + }, + { + "epoch": 1.5509303058971933, + "grad_norm": 0.97265625, + "learning_rate": 0.0002, + "loss": 1.9965, + "step": 4918 + }, + { + "epoch": 1.5512456638284453, + "grad_norm": 1.1328125, + "learning_rate": 0.0002, + "loss": 2.1712, + "step": 4919 + }, + { + "epoch": 1.5515610217596971, + "grad_norm": 0.88671875, + "learning_rate": 0.0002, + "loss": 1.8041, + "step": 4920 + }, + { + "epoch": 1.5518763796909494, + "grad_norm": 1.4609375, + "learning_rate": 0.0002, + "loss": 2.4673, + "step": 4921 + }, + { + "epoch": 1.5521917376222012, + "grad_norm": 0.2001953125, + "learning_rate": 0.0002, + "loss": 0.9027, + "step": 4922 + }, + { + "epoch": 1.552507095553453, + "grad_norm": 0.208984375, + "learning_rate": 0.0002, + "loss": 0.9818, + "step": 4923 + }, + { + "epoch": 1.5528224534847053, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 0.9279, + "step": 4924 + }, + { + "epoch": 1.553137811415957, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 1.0784, + "step": 4925 + }, + { + "epoch": 1.5534531693472091, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 0.9576, + "step": 4926 + }, + { + "epoch": 1.5537685272784612, + "grad_norm": 0.220703125, + "learning_rate": 0.0002, + "loss": 1.0287, + "step": 4927 + }, + { + "epoch": 1.554083885209713, + "grad_norm": 0.216796875, + "learning_rate": 0.0002, + "loss": 0.9334, + "step": 4928 + }, + { + "epoch": 1.554399243140965, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 0.976, + "step": 4929 + }, + { + "epoch": 1.554714601072217, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.0761, + "step": 4930 + }, + { + "epoch": 1.5550299590034689, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 0.9334, + "step": 4931 + }, + { + "epoch": 1.555345316934721, + "grad_norm": 0.2421875, + "learning_rate": 0.0002, + "loss": 1.0357, + "step": 4932 + }, + { + "epoch": 1.555660674865973, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 1.0879, + "step": 4933 + }, + { + "epoch": 1.5559760327972247, + "grad_norm": 0.2470703125, + "learning_rate": 0.0002, + "loss": 1.1723, + "step": 4934 + }, + { + "epoch": 1.5562913907284768, + "grad_norm": 0.248046875, + "learning_rate": 0.0002, + "loss": 1.0019, + "step": 4935 + }, + { + "epoch": 1.5566067486597288, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 1.0628, + "step": 4936 + }, + { + "epoch": 1.5569221065909806, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.0439, + "step": 4937 + }, + { + "epoch": 1.5572374645222329, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 1.0, + "step": 4938 + }, + { + "epoch": 1.5575528224534847, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 1.1862, + "step": 4939 + }, + { + "epoch": 1.5578681803847365, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.0515, + "step": 4940 + }, + { + "epoch": 1.5581835383159888, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 1.0862, + "step": 4941 + }, + { + "epoch": 1.5584988962472406, + "grad_norm": 0.353515625, + "learning_rate": 0.0002, + "loss": 1.2925, + "step": 4942 + }, + { + "epoch": 1.5588142541784926, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 1.1128, + "step": 4943 + }, + { + "epoch": 1.5591296121097447, + "grad_norm": 0.37109375, + "learning_rate": 0.0002, + "loss": 1.1479, + "step": 4944 + }, + { + "epoch": 1.5594449700409965, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 1.185, + "step": 4945 + }, + { + "epoch": 1.5597603279722485, + "grad_norm": 0.416015625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 4946 + }, + { + "epoch": 1.5600756859035005, + "grad_norm": 0.388671875, + "learning_rate": 0.0002, + "loss": 1.2646, + "step": 4947 + }, + { + "epoch": 1.5603910438347524, + "grad_norm": 0.380859375, + "learning_rate": 0.0002, + "loss": 1.0935, + "step": 4948 + }, + { + "epoch": 1.5607064017660044, + "grad_norm": 0.447265625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 4949 + }, + { + "epoch": 1.5610217596972564, + "grad_norm": 0.498046875, + "learning_rate": 0.0002, + "loss": 1.2352, + "step": 4950 + }, + { + "epoch": 1.5613371176285082, + "grad_norm": 0.455078125, + "learning_rate": 0.0002, + "loss": 1.0508, + "step": 4951 + }, + { + "epoch": 1.5616524755597603, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 1.4475, + "step": 4952 + }, + { + "epoch": 1.5619678334910123, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 1.4335, + "step": 4953 + }, + { + "epoch": 1.5622831914222641, + "grad_norm": 0.80859375, + "learning_rate": 0.0002, + "loss": 1.6898, + "step": 4954 + }, + { + "epoch": 1.5625985493535164, + "grad_norm": 0.69921875, + "learning_rate": 0.0002, + "loss": 1.7439, + "step": 4955 + }, + { + "epoch": 1.5629139072847682, + "grad_norm": 0.71484375, + "learning_rate": 0.0002, + "loss": 1.7798, + "step": 4956 + }, + { + "epoch": 1.5632292652160202, + "grad_norm": 0.80078125, + "learning_rate": 0.0002, + "loss": 1.8537, + "step": 4957 + }, + { + "epoch": 1.5635446231472723, + "grad_norm": 0.70703125, + "learning_rate": 0.0002, + "loss": 1.5964, + "step": 4958 + }, + { + "epoch": 1.563859981078524, + "grad_norm": 0.71875, + "learning_rate": 0.0002, + "loss": 1.5285, + "step": 4959 + }, + { + "epoch": 1.5641753390097761, + "grad_norm": 0.953125, + "learning_rate": 0.0002, + "loss": 2.1087, + "step": 4960 + }, + { + "epoch": 1.5644906969410282, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 1.9085, + "step": 4961 + }, + { + "epoch": 1.56480605487228, + "grad_norm": 0.81640625, + "learning_rate": 0.0002, + "loss": 2.0214, + "step": 4962 + }, + { + "epoch": 1.565121412803532, + "grad_norm": 1.015625, + "learning_rate": 0.0002, + "loss": 2.4721, + "step": 4963 + }, + { + "epoch": 1.565436770734784, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 2.2663, + "step": 4964 + }, + { + "epoch": 1.5657521286660359, + "grad_norm": 0.7578125, + "learning_rate": 0.0002, + "loss": 1.7713, + "step": 4965 + }, + { + "epoch": 1.566067486597288, + "grad_norm": 1.171875, + "learning_rate": 0.0002, + "loss": 2.3474, + "step": 4966 + }, + { + "epoch": 1.56638284452854, + "grad_norm": 0.8359375, + "learning_rate": 0.0002, + "loss": 1.7166, + "step": 4967 + }, + { + "epoch": 1.5666982024597917, + "grad_norm": 1.0859375, + "learning_rate": 0.0002, + "loss": 2.0853, + "step": 4968 + }, + { + "epoch": 1.5670135603910438, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.9848, + "step": 4969 + }, + { + "epoch": 1.5673289183222958, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 1.7853, + "step": 4970 + }, + { + "epoch": 1.5676442762535476, + "grad_norm": 1.453125, + "learning_rate": 0.0002, + "loss": 2.8245, + "step": 4971 + }, + { + "epoch": 1.5679596341847999, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.0587, + "step": 4972 + }, + { + "epoch": 1.5682749921160517, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 1.0487, + "step": 4973 + }, + { + "epoch": 1.5685903500473037, + "grad_norm": 0.2060546875, + "learning_rate": 0.0002, + "loss": 0.8995, + "step": 4974 + }, + { + "epoch": 1.5689057079785558, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.1119, + "step": 4975 + }, + { + "epoch": 1.5692210659098076, + "grad_norm": 0.205078125, + "learning_rate": 0.0002, + "loss": 1.013, + "step": 4976 + }, + { + "epoch": 1.5695364238410596, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 0.9116, + "step": 4977 + }, + { + "epoch": 1.5698517817723117, + "grad_norm": 0.197265625, + "learning_rate": 0.0002, + "loss": 0.9965, + "step": 4978 + }, + { + "epoch": 1.5701671397035635, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 0.9665, + "step": 4979 + }, + { + "epoch": 1.5704824976348155, + "grad_norm": 0.2294921875, + "learning_rate": 0.0002, + "loss": 1.0742, + "step": 4980 + }, + { + "epoch": 1.5707978555660675, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 0.9743, + "step": 4981 + }, + { + "epoch": 1.5711132134973194, + "grad_norm": 0.240234375, + "learning_rate": 0.0002, + "loss": 1.1419, + "step": 4982 + }, + { + "epoch": 1.5714285714285714, + "grad_norm": 0.361328125, + "learning_rate": 0.0002, + "loss": 1.0653, + "step": 4983 + }, + { + "epoch": 1.5717439293598234, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.1301, + "step": 4984 + }, + { + "epoch": 1.5720592872910752, + "grad_norm": 0.2421875, + "learning_rate": 0.0002, + "loss": 1.0151, + "step": 4985 + }, + { + "epoch": 1.5723746452223275, + "grad_norm": 0.2470703125, + "learning_rate": 0.0002, + "loss": 0.9322, + "step": 4986 + }, + { + "epoch": 1.5726900031535793, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 1.1842, + "step": 4987 + }, + { + "epoch": 1.5730053610848311, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.1115, + "step": 4988 + }, + { + "epoch": 1.5733207190160834, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 1.0981, + "step": 4989 + }, + { + "epoch": 1.5736360769473352, + "grad_norm": 0.3671875, + "learning_rate": 0.0002, + "loss": 1.2851, + "step": 4990 + }, + { + "epoch": 1.5739514348785872, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 1.1574, + "step": 4991 + }, + { + "epoch": 1.5742667928098393, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 1.0815, + "step": 4992 + }, + { + "epoch": 1.574582150741091, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 1.0516, + "step": 4993 + }, + { + "epoch": 1.5748975086723431, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 1.4137, + "step": 4994 + }, + { + "epoch": 1.5752128666035952, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 1.0042, + "step": 4995 + }, + { + "epoch": 1.575528224534847, + "grad_norm": 0.37890625, + "learning_rate": 0.0002, + "loss": 1.274, + "step": 4996 + }, + { + "epoch": 1.575843582466099, + "grad_norm": 0.443359375, + "learning_rate": 0.0002, + "loss": 1.187, + "step": 4997 + }, + { + "epoch": 1.576158940397351, + "grad_norm": 0.396484375, + "learning_rate": 0.0002, + "loss": 1.2532, + "step": 4998 + }, + { + "epoch": 1.5764742983286029, + "grad_norm": 0.4375, + "learning_rate": 0.0002, + "loss": 1.1881, + "step": 4999 + }, + { + "epoch": 1.576789656259855, + "grad_norm": 0.4765625, + "learning_rate": 0.0002, + "loss": 1.2177, + "step": 5000 + }, + { + "epoch": 1.577105014191107, + "grad_norm": 0.439453125, + "learning_rate": 0.0002, + "loss": 1.238, + "step": 5001 + }, + { + "epoch": 1.5774203721223587, + "grad_norm": 0.515625, + "learning_rate": 0.0002, + "loss": 1.2735, + "step": 5002 + }, + { + "epoch": 1.577735730053611, + "grad_norm": 0.63671875, + "learning_rate": 0.0002, + "loss": 1.426, + "step": 5003 + }, + { + "epoch": 1.5780510879848628, + "grad_norm": 0.5, + "learning_rate": 0.0002, + "loss": 1.2688, + "step": 5004 + }, + { + "epoch": 1.5783664459161146, + "grad_norm": 0.76171875, + "learning_rate": 0.0002, + "loss": 1.7357, + "step": 5005 + }, + { + "epoch": 1.578681803847367, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 1.5136, + "step": 5006 + }, + { + "epoch": 1.5789971617786187, + "grad_norm": 0.73828125, + "learning_rate": 0.0002, + "loss": 1.4954, + "step": 5007 + }, + { + "epoch": 1.5793125197098707, + "grad_norm": 0.765625, + "learning_rate": 0.0002, + "loss": 1.8916, + "step": 5008 + }, + { + "epoch": 1.5796278776411228, + "grad_norm": 0.7734375, + "learning_rate": 0.0002, + "loss": 1.7652, + "step": 5009 + }, + { + "epoch": 1.5799432355723746, + "grad_norm": 0.76953125, + "learning_rate": 0.0002, + "loss": 1.8526, + "step": 5010 + }, + { + "epoch": 1.5802585935036266, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 1.7528, + "step": 5011 + }, + { + "epoch": 1.5805739514348787, + "grad_norm": 0.74609375, + "learning_rate": 0.0002, + "loss": 1.6637, + "step": 5012 + }, + { + "epoch": 1.5808893093661305, + "grad_norm": 1.0, + "learning_rate": 0.0002, + "loss": 2.0563, + "step": 5013 + }, + { + "epoch": 1.5812046672973825, + "grad_norm": 0.86328125, + "learning_rate": 0.0002, + "loss": 1.8735, + "step": 5014 + }, + { + "epoch": 1.5815200252286346, + "grad_norm": 0.9609375, + "learning_rate": 0.0002, + "loss": 2.1522, + "step": 5015 + }, + { + "epoch": 1.5818353831598864, + "grad_norm": 0.953125, + "learning_rate": 0.0002, + "loss": 1.9514, + "step": 5016 + }, + { + "epoch": 1.5821507410911384, + "grad_norm": 0.98828125, + "learning_rate": 0.0002, + "loss": 2.3104, + "step": 5017 + }, + { + "epoch": 1.5824660990223904, + "grad_norm": 1.1796875, + "learning_rate": 0.0002, + "loss": 2.6478, + "step": 5018 + }, + { + "epoch": 1.5827814569536423, + "grad_norm": 0.96875, + "learning_rate": 0.0002, + "loss": 1.8318, + "step": 5019 + }, + { + "epoch": 1.5830968148848945, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 1.6423, + "step": 5020 + }, + { + "epoch": 1.5834121728161463, + "grad_norm": 1.6015625, + "learning_rate": 0.0002, + "loss": 2.2413, + "step": 5021 + }, + { + "epoch": 1.5837275307473981, + "grad_norm": 0.1884765625, + "learning_rate": 0.0002, + "loss": 1.0051, + "step": 5022 + }, + { + "epoch": 1.5840428886786504, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 1.1618, + "step": 5023 + }, + { + "epoch": 1.5843582466099022, + "grad_norm": 0.2314453125, + "learning_rate": 0.0002, + "loss": 1.1493, + "step": 5024 + }, + { + "epoch": 1.5846736045411542, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 0.9738, + "step": 5025 + }, + { + "epoch": 1.5849889624724063, + "grad_norm": 0.2138671875, + "learning_rate": 0.0002, + "loss": 0.9752, + "step": 5026 + }, + { + "epoch": 1.585304320403658, + "grad_norm": 0.220703125, + "learning_rate": 0.0002, + "loss": 1.0694, + "step": 5027 + }, + { + "epoch": 1.5856196783349101, + "grad_norm": 0.2138671875, + "learning_rate": 0.0002, + "loss": 0.9741, + "step": 5028 + }, + { + "epoch": 1.5859350362661622, + "grad_norm": 0.212890625, + "learning_rate": 0.0002, + "loss": 1.0371, + "step": 5029 + }, + { + "epoch": 1.586250394197414, + "grad_norm": 0.212890625, + "learning_rate": 0.0002, + "loss": 1.1137, + "step": 5030 + }, + { + "epoch": 1.586565752128666, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 1.139, + "step": 5031 + }, + { + "epoch": 1.586881110059918, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.1146, + "step": 5032 + }, + { + "epoch": 1.5871964679911699, + "grad_norm": 0.23828125, + "learning_rate": 0.0002, + "loss": 1.0713, + "step": 5033 + }, + { + "epoch": 1.587511825922422, + "grad_norm": 0.2490234375, + "learning_rate": 0.0002, + "loss": 1.0846, + "step": 5034 + }, + { + "epoch": 1.587827183853674, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.1391, + "step": 5035 + }, + { + "epoch": 1.5881425417849258, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 1.2001, + "step": 5036 + }, + { + "epoch": 1.588457899716178, + "grad_norm": 0.244140625, + "learning_rate": 0.0002, + "loss": 1.024, + "step": 5037 + }, + { + "epoch": 1.5887732576474298, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 5038 + }, + { + "epoch": 1.5890886155786816, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.0499, + "step": 5039 + }, + { + "epoch": 1.589403973509934, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 1.1591, + "step": 5040 + }, + { + "epoch": 1.5897193314411857, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 1.1616, + "step": 5041 + }, + { + "epoch": 1.5900346893724377, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 1.0941, + "step": 5042 + }, + { + "epoch": 1.5903500473036898, + "grad_norm": 0.40234375, + "learning_rate": 0.0002, + "loss": 1.2934, + "step": 5043 + }, + { + "epoch": 1.5906654052349416, + "grad_norm": 0.34765625, + "learning_rate": 0.0002, + "loss": 1.0637, + "step": 5044 + }, + { + "epoch": 1.5909807631661936, + "grad_norm": 0.384765625, + "learning_rate": 0.0002, + "loss": 1.3843, + "step": 5045 + }, + { + "epoch": 1.5912961210974457, + "grad_norm": 0.40234375, + "learning_rate": 0.0002, + "loss": 1.1528, + "step": 5046 + }, + { + "epoch": 1.5916114790286975, + "grad_norm": 0.412109375, + "learning_rate": 0.0002, + "loss": 1.1908, + "step": 5047 + }, + { + "epoch": 1.5919268369599495, + "grad_norm": 0.42578125, + "learning_rate": 0.0002, + "loss": 1.1032, + "step": 5048 + }, + { + "epoch": 1.5922421948912016, + "grad_norm": 0.455078125, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 5049 + }, + { + "epoch": 1.5922421948912016, + "eval_loss": 1.6771923303604126, + "eval_runtime": 148.9853, + "eval_samples_per_second": 6.712, + "eval_steps_per_second": 6.712, + "step": 5049 + }, + { + "epoch": 1.5922421948912016, + "mmlu_eval_accuracy": 0.5763594830328238, + "mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453, + "mmlu_eval_accuracy_anatomy": 0.5, + "mmlu_eval_accuracy_astronomy": 0.6875, + "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, + "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034, + "mmlu_eval_accuracy_college_biology": 0.6875, + "mmlu_eval_accuracy_college_chemistry": 0.25, + "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365, + "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365, + "mmlu_eval_accuracy_college_medicine": 0.5909090909090909, + "mmlu_eval_accuracy_college_physics": 0.5454545454545454, + "mmlu_eval_accuracy_computer_security": 0.5454545454545454, + "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231, + "mmlu_eval_accuracy_econometrics": 0.4166666666666667, + "mmlu_eval_accuracy_electrical_engineering": 0.4375, + "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244, + "mmlu_eval_accuracy_formal_logic": 0.2857142857142857, + "mmlu_eval_accuracy_global_facts": 0.2, + "mmlu_eval_accuracy_high_school_biology": 0.65625, + "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365, + "mmlu_eval_accuracy_high_school_computer_science": 0.7777777777777778, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.9545454545454546, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093, + "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276, + "mmlu_eval_accuracy_high_school_microeconomics": 0.46153846153846156, + "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, + "mmlu_eval_accuracy_high_school_psychology": 0.8, + "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, + "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, + "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, + "mmlu_eval_accuracy_human_aging": 0.6086956521739131, + "mmlu_eval_accuracy_human_sexuality": 0.5, + "mmlu_eval_accuracy_international_law": 0.9230769230769231, + "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, + "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, + "mmlu_eval_accuracy_machine_learning": 0.36363636363636365, + "mmlu_eval_accuracy_management": 0.8181818181818182, + "mmlu_eval_accuracy_marketing": 0.72, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7209302325581395, + "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527, + "mmlu_eval_accuracy_moral_scenarios": 0.28, + "mmlu_eval_accuracy_nutrition": 0.7272727272727273, + "mmlu_eval_accuracy_philosophy": 0.6764705882352942, + "mmlu_eval_accuracy_prehistory": 0.5142857142857142, + "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194, + "mmlu_eval_accuracy_professional_law": 0.38235294117647056, + "mmlu_eval_accuracy_professional_medicine": 0.6774193548387096, + "mmlu_eval_accuracy_professional_psychology": 0.6521739130434783, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.5925925925925926, + "mmlu_eval_accuracy_sociology": 0.9090909090909091, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.4444444444444444, + "mmlu_eval_accuracy_world_religions": 0.7894736842105263, + "mmlu_loss": 5.924344757972085, + "step": 5049 + }, + { + "epoch": 1.5925575528224534, + "grad_norm": 0.421875, + "learning_rate": 0.0002, + "loss": 1.1523, + "step": 5050 + }, + { + "epoch": 1.5928729107537054, + "grad_norm": 0.478515625, + "learning_rate": 0.0002, + "loss": 1.1604, + "step": 5051 + }, + { + "epoch": 1.5931882686849574, + "grad_norm": 0.484375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 5052 + }, + { + "epoch": 1.5935036266162093, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 1.2807, + "step": 5053 + }, + { + "epoch": 1.5938189845474615, + "grad_norm": 0.62109375, + "learning_rate": 0.0002, + "loss": 1.5564, + "step": 5054 + }, + { + "epoch": 1.5941343424787133, + "grad_norm": 0.7578125, + "learning_rate": 0.0002, + "loss": 1.5671, + "step": 5055 + }, + { + "epoch": 1.5944497004099654, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 1.551, + "step": 5056 + }, + { + "epoch": 1.5947650583412174, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 1.7013, + "step": 5057 + }, + { + "epoch": 1.5950804162724692, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 1.9711, + "step": 5058 + }, + { + "epoch": 1.5953957742037213, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 2.1996, + "step": 5059 + }, + { + "epoch": 1.5957111321349733, + "grad_norm": 0.96484375, + "learning_rate": 0.0002, + "loss": 2.0482, + "step": 5060 + }, + { + "epoch": 1.596026490066225, + "grad_norm": 0.97265625, + "learning_rate": 0.0002, + "loss": 1.9273, + "step": 5061 + }, + { + "epoch": 1.5963418479974771, + "grad_norm": 0.859375, + "learning_rate": 0.0002, + "loss": 1.8249, + "step": 5062 + }, + { + "epoch": 1.5966572059287292, + "grad_norm": 1.3828125, + "learning_rate": 0.0002, + "loss": 2.195, + "step": 5063 + }, + { + "epoch": 1.596972563859981, + "grad_norm": 1.1171875, + "learning_rate": 0.0002, + "loss": 2.4511, + "step": 5064 + }, + { + "epoch": 1.597287921791233, + "grad_norm": 0.96484375, + "learning_rate": 0.0002, + "loss": 1.9753, + "step": 5065 + }, + { + "epoch": 1.597603279722485, + "grad_norm": 0.99609375, + "learning_rate": 0.0002, + "loss": 2.3296, + "step": 5066 + }, + { + "epoch": 1.5979186376537369, + "grad_norm": 1.234375, + "learning_rate": 0.0002, + "loss": 2.2667, + "step": 5067 + }, + { + "epoch": 1.598233995584989, + "grad_norm": 1.25, + "learning_rate": 0.0002, + "loss": 2.2576, + "step": 5068 + }, + { + "epoch": 1.598549353516241, + "grad_norm": 0.90234375, + "learning_rate": 0.0002, + "loss": 1.9613, + "step": 5069 + }, + { + "epoch": 1.5988647114474928, + "grad_norm": 0.97265625, + "learning_rate": 0.0002, + "loss": 2.0631, + "step": 5070 + }, + { + "epoch": 1.599180069378745, + "grad_norm": 1.078125, + "learning_rate": 0.0002, + "loss": 2.2722, + "step": 5071 + }, + { + "epoch": 1.5994954273099968, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 1.0521, + "step": 5072 + }, + { + "epoch": 1.5998107852412489, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.121, + "step": 5073 + }, + { + "epoch": 1.600126143172501, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.1034, + "step": 5074 + }, + { + "epoch": 1.6004415011037527, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 1.0795, + "step": 5075 + }, + { + "epoch": 1.6007568590350048, + "grad_norm": 0.208984375, + "learning_rate": 0.0002, + "loss": 1.1636, + "step": 5076 + }, + { + "epoch": 1.6010722169662568, + "grad_norm": 0.216796875, + "learning_rate": 0.0002, + "loss": 1.0503, + "step": 5077 + }, + { + "epoch": 1.6013875748975086, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.0865, + "step": 5078 + }, + { + "epoch": 1.6017029328287606, + "grad_norm": 0.232421875, + "learning_rate": 0.0002, + "loss": 1.0424, + "step": 5079 + }, + { + "epoch": 1.6020182907600127, + "grad_norm": 0.255859375, + "learning_rate": 0.0002, + "loss": 1.0543, + "step": 5080 + }, + { + "epoch": 1.6023336486912645, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 1.0584, + "step": 5081 + }, + { + "epoch": 1.6026490066225165, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 1.0839, + "step": 5082 + }, + { + "epoch": 1.6029643645537686, + "grad_norm": 0.23828125, + "learning_rate": 0.0002, + "loss": 1.0206, + "step": 5083 + }, + { + "epoch": 1.6032797224850204, + "grad_norm": 0.2490234375, + "learning_rate": 0.0002, + "loss": 1.1476, + "step": 5084 + }, + { + "epoch": 1.6035950804162726, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 1.2314, + "step": 5085 + }, + { + "epoch": 1.6039104383475244, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.0466, + "step": 5086 + }, + { + "epoch": 1.6042257962787763, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 1.0605, + "step": 5087 + }, + { + "epoch": 1.6045411542100285, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.2402, + "step": 5088 + }, + { + "epoch": 1.6048565121412803, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 1.1001, + "step": 5089 + }, + { + "epoch": 1.6051718700725324, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 1.0379, + "step": 5090 + }, + { + "epoch": 1.6054872280037844, + "grad_norm": 0.369140625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 5091 + }, + { + "epoch": 1.6058025859350362, + "grad_norm": 0.4375, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 5092 + }, + { + "epoch": 1.6061179438662883, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 1.0773, + "step": 5093 + }, + { + "epoch": 1.6064333017975403, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 1.0819, + "step": 5094 + }, + { + "epoch": 1.606748659728792, + "grad_norm": 0.408203125, + "learning_rate": 0.0002, + "loss": 1.111, + "step": 5095 + }, + { + "epoch": 1.6070640176600441, + "grad_norm": 0.369140625, + "learning_rate": 0.0002, + "loss": 1.1457, + "step": 5096 + }, + { + "epoch": 1.6073793755912962, + "grad_norm": 0.43359375, + "learning_rate": 0.0002, + "loss": 1.1709, + "step": 5097 + }, + { + "epoch": 1.607694733522548, + "grad_norm": 0.392578125, + "learning_rate": 0.0002, + "loss": 1.2424, + "step": 5098 + }, + { + "epoch": 1.6080100914538, + "grad_norm": 0.4921875, + "learning_rate": 0.0002, + "loss": 1.2218, + "step": 5099 + }, + { + "epoch": 1.608325449385052, + "grad_norm": 0.4609375, + "learning_rate": 0.0002, + "loss": 1.2444, + "step": 5100 + }, + { + "epoch": 1.6086408073163039, + "grad_norm": 0.400390625, + "learning_rate": 0.0002, + "loss": 0.9536, + "step": 5101 + }, + { + "epoch": 1.6089561652475561, + "grad_norm": 0.61328125, + "learning_rate": 0.0002, + "loss": 1.4142, + "step": 5102 + }, + { + "epoch": 1.609271523178808, + "grad_norm": 0.6015625, + "learning_rate": 0.0002, + "loss": 1.49, + "step": 5103 + }, + { + "epoch": 1.6095868811100598, + "grad_norm": 0.61328125, + "learning_rate": 0.0002, + "loss": 1.556, + "step": 5104 + }, + { + "epoch": 1.609902239041312, + "grad_norm": 0.51953125, + "learning_rate": 0.0002, + "loss": 1.1415, + "step": 5105 + }, + { + "epoch": 1.6102175969725638, + "grad_norm": 0.640625, + "learning_rate": 0.0002, + "loss": 1.4294, + "step": 5106 + }, + { + "epoch": 1.6105329549038159, + "grad_norm": 0.66796875, + "learning_rate": 0.0002, + "loss": 1.615, + "step": 5107 + }, + { + "epoch": 1.610848312835068, + "grad_norm": 0.86328125, + "learning_rate": 0.0002, + "loss": 1.6909, + "step": 5108 + }, + { + "epoch": 1.6111636707663197, + "grad_norm": 0.91796875, + "learning_rate": 0.0002, + "loss": 1.9264, + "step": 5109 + }, + { + "epoch": 1.6114790286975718, + "grad_norm": 0.796875, + "learning_rate": 0.0002, + "loss": 1.8801, + "step": 5110 + }, + { + "epoch": 1.6117943866288238, + "grad_norm": 0.8828125, + "learning_rate": 0.0002, + "loss": 1.9392, + "step": 5111 + }, + { + "epoch": 1.6121097445600756, + "grad_norm": 0.9921875, + "learning_rate": 0.0002, + "loss": 2.4256, + "step": 5112 + }, + { + "epoch": 1.6124251024913276, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 2.1807, + "step": 5113 + }, + { + "epoch": 1.6127404604225797, + "grad_norm": 0.8515625, + "learning_rate": 0.0002, + "loss": 1.7436, + "step": 5114 + }, + { + "epoch": 1.6130558183538315, + "grad_norm": 0.95703125, + "learning_rate": 0.0002, + "loss": 2.0113, + "step": 5115 + }, + { + "epoch": 1.6133711762850835, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 2.2849, + "step": 5116 + }, + { + "epoch": 1.6136865342163356, + "grad_norm": 0.8515625, + "learning_rate": 0.0002, + "loss": 1.7667, + "step": 5117 + }, + { + "epoch": 1.6140018921475874, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 2.2397, + "step": 5118 + }, + { + "epoch": 1.6143172500788396, + "grad_norm": 1.140625, + "learning_rate": 0.0002, + "loss": 2.2722, + "step": 5119 + }, + { + "epoch": 1.6146326080100915, + "grad_norm": 1.0390625, + "learning_rate": 0.0002, + "loss": 1.9065, + "step": 5120 + }, + { + "epoch": 1.6149479659413433, + "grad_norm": 1.25, + "learning_rate": 0.0002, + "loss": 2.2473, + "step": 5121 + }, + { + "epoch": 1.6152633238725955, + "grad_norm": 0.220703125, + "learning_rate": 0.0002, + "loss": 1.055, + "step": 5122 + }, + { + "epoch": 1.6155786818038473, + "grad_norm": 0.232421875, + "learning_rate": 0.0002, + "loss": 1.0749, + "step": 5123 + }, + { + "epoch": 1.6158940397350994, + "grad_norm": 0.208984375, + "learning_rate": 0.0002, + "loss": 0.9465, + "step": 5124 + }, + { + "epoch": 1.6162093976663514, + "grad_norm": 0.19921875, + "learning_rate": 0.0002, + "loss": 0.9945, + "step": 5125 + }, + { + "epoch": 1.6165247555976032, + "grad_norm": 0.2294921875, + "learning_rate": 0.0002, + "loss": 1.0243, + "step": 5126 + }, + { + "epoch": 1.6168401135288553, + "grad_norm": 0.220703125, + "learning_rate": 0.0002, + "loss": 1.0005, + "step": 5127 + }, + { + "epoch": 1.6171554714601073, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 0.9583, + "step": 5128 + }, + { + "epoch": 1.617470829391359, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 1.0297, + "step": 5129 + }, + { + "epoch": 1.6177861873226111, + "grad_norm": 0.23046875, + "learning_rate": 0.0002, + "loss": 1.0678, + "step": 5130 + }, + { + "epoch": 1.6181015452538632, + "grad_norm": 0.2314453125, + "learning_rate": 0.0002, + "loss": 1.0285, + "step": 5131 + }, + { + "epoch": 1.618416903185115, + "grad_norm": 0.2333984375, + "learning_rate": 0.0002, + "loss": 1.0582, + "step": 5132 + }, + { + "epoch": 1.618732261116367, + "grad_norm": 0.23046875, + "learning_rate": 0.0002, + "loss": 0.9439, + "step": 5133 + }, + { + "epoch": 1.619047619047619, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.06, + "step": 5134 + }, + { + "epoch": 1.6193629769788709, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 1.1269, + "step": 5135 + }, + { + "epoch": 1.6196783349101231, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 1.0278, + "step": 5136 + }, + { + "epoch": 1.619993692841375, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 1.2706, + "step": 5137 + }, + { + "epoch": 1.6203090507726268, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 1.0907, + "step": 5138 + }, + { + "epoch": 1.620624408703879, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 1.0111, + "step": 5139 + }, + { + "epoch": 1.6209397666351308, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 1.1081, + "step": 5140 + }, + { + "epoch": 1.6212551245663829, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 1.0663, + "step": 5141 + }, + { + "epoch": 1.621570482497635, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 1.1981, + "step": 5142 + }, + { + "epoch": 1.6218858404288867, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 5143 + }, + { + "epoch": 1.6222011983601388, + "grad_norm": 0.39453125, + "learning_rate": 0.0002, + "loss": 1.1534, + "step": 5144 + }, + { + "epoch": 1.6225165562913908, + "grad_norm": 0.3671875, + "learning_rate": 0.0002, + "loss": 1.1736, + "step": 5145 + }, + { + "epoch": 1.6228319142226426, + "grad_norm": 0.39453125, + "learning_rate": 0.0002, + "loss": 1.0691, + "step": 5146 + }, + { + "epoch": 1.6231472721538946, + "grad_norm": 0.3984375, + "learning_rate": 0.0002, + "loss": 1.1695, + "step": 5147 + }, + { + "epoch": 1.6234626300851467, + "grad_norm": 0.47265625, + "learning_rate": 0.0002, + "loss": 1.2859, + "step": 5148 + }, + { + "epoch": 1.6237779880163985, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 5149 + }, + { + "epoch": 1.6240933459476505, + "grad_norm": 0.458984375, + "learning_rate": 0.0002, + "loss": 1.1685, + "step": 5150 + }, + { + "epoch": 1.6244087038789026, + "grad_norm": 0.4765625, + "learning_rate": 0.0002, + "loss": 1.3035, + "step": 5151 + }, + { + "epoch": 1.6247240618101544, + "grad_norm": 0.515625, + "learning_rate": 0.0002, + "loss": 1.3722, + "step": 5152 + }, + { + "epoch": 1.6250394197414066, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 1.4437, + "step": 5153 + }, + { + "epoch": 1.6253547776726585, + "grad_norm": 0.609375, + "learning_rate": 0.0002, + "loss": 1.6459, + "step": 5154 + }, + { + "epoch": 1.6256701356039105, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 1.227, + "step": 5155 + }, + { + "epoch": 1.6259854935351625, + "grad_norm": 0.65625, + "learning_rate": 0.0002, + "loss": 1.6517, + "step": 5156 + }, + { + "epoch": 1.6263008514664143, + "grad_norm": 0.83203125, + "learning_rate": 0.0002, + "loss": 1.7303, + "step": 5157 + }, + { + "epoch": 1.6266162093976664, + "grad_norm": 0.7265625, + "learning_rate": 0.0002, + "loss": 1.6782, + "step": 5158 + }, + { + "epoch": 1.6269315673289184, + "grad_norm": 0.89453125, + "learning_rate": 0.0002, + "loss": 1.9444, + "step": 5159 + }, + { + "epoch": 1.6272469252601702, + "grad_norm": 0.85546875, + "learning_rate": 0.0002, + "loss": 1.988, + "step": 5160 + }, + { + "epoch": 1.6275622831914223, + "grad_norm": 0.875, + "learning_rate": 0.0002, + "loss": 2.3145, + "step": 5161 + }, + { + "epoch": 1.6278776411226743, + "grad_norm": 1.140625, + "learning_rate": 0.0002, + "loss": 1.9321, + "step": 5162 + }, + { + "epoch": 1.6281929990539261, + "grad_norm": 1.8125, + "learning_rate": 0.0002, + "loss": 2.055, + "step": 5163 + }, + { + "epoch": 1.6285083569851782, + "grad_norm": 0.98046875, + "learning_rate": 0.0002, + "loss": 2.4308, + "step": 5164 + }, + { + "epoch": 1.6288237149164302, + "grad_norm": 1.1328125, + "learning_rate": 0.0002, + "loss": 2.1361, + "step": 5165 + }, + { + "epoch": 1.629139072847682, + "grad_norm": 0.984375, + "learning_rate": 0.0002, + "loss": 2.1152, + "step": 5166 + }, + { + "epoch": 1.629454430778934, + "grad_norm": 0.84375, + "learning_rate": 0.0002, + "loss": 1.6886, + "step": 5167 + }, + { + "epoch": 1.629769788710186, + "grad_norm": 0.97265625, + "learning_rate": 0.0002, + "loss": 1.8483, + "step": 5168 + }, + { + "epoch": 1.6300851466414379, + "grad_norm": 0.90234375, + "learning_rate": 0.0002, + "loss": 1.6223, + "step": 5169 + }, + { + "epoch": 1.6304005045726901, + "grad_norm": 1.15625, + "learning_rate": 0.0002, + "loss": 1.9019, + "step": 5170 + }, + { + "epoch": 1.630715862503942, + "grad_norm": 1.328125, + "learning_rate": 0.0002, + "loss": 2.2396, + "step": 5171 + }, + { + "epoch": 1.631031220435194, + "grad_norm": 0.18359375, + "learning_rate": 0.0002, + "loss": 1.0325, + "step": 5172 + }, + { + "epoch": 1.631346578366446, + "grad_norm": 0.212890625, + "learning_rate": 0.0002, + "loss": 1.0291, + "step": 5173 + }, + { + "epoch": 1.6316619362976978, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.0474, + "step": 5174 + }, + { + "epoch": 1.6319772942289499, + "grad_norm": 0.232421875, + "learning_rate": 0.0002, + "loss": 1.0224, + "step": 5175 + }, + { + "epoch": 1.632292652160202, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 1.0311, + "step": 5176 + }, + { + "epoch": 1.6326080100914537, + "grad_norm": 0.2333984375, + "learning_rate": 0.0002, + "loss": 1.0984, + "step": 5177 + }, + { + "epoch": 1.6329233680227058, + "grad_norm": 0.255859375, + "learning_rate": 0.0002, + "loss": 1.0138, + "step": 5178 + }, + { + "epoch": 1.6332387259539578, + "grad_norm": 0.2412109375, + "learning_rate": 0.0002, + "loss": 1.1637, + "step": 5179 + }, + { + "epoch": 1.6335540838852096, + "grad_norm": 0.2314453125, + "learning_rate": 0.0002, + "loss": 0.9647, + "step": 5180 + }, + { + "epoch": 1.6338694418164617, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 1.0254, + "step": 5181 + }, + { + "epoch": 1.6341847997477137, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 1.0707, + "step": 5182 + }, + { + "epoch": 1.6345001576789655, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 1.081, + "step": 5183 + }, + { + "epoch": 1.6348155156102178, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.0847, + "step": 5184 + }, + { + "epoch": 1.6351308735414696, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 1.0418, + "step": 5185 + }, + { + "epoch": 1.6354462314727214, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 1.2573, + "step": 5186 + }, + { + "epoch": 1.6357615894039736, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 1.0944, + "step": 5187 + }, + { + "epoch": 1.6360769473352255, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 1.1143, + "step": 5188 + }, + { + "epoch": 1.6363923052664775, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 1.1658, + "step": 5189 + }, + { + "epoch": 1.6367076631977295, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 1.046, + "step": 5190 + }, + { + "epoch": 1.6370230211289813, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 1.0552, + "step": 5191 + }, + { + "epoch": 1.6373383790602334, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 1.0448, + "step": 5192 + }, + { + "epoch": 1.6376537369914854, + "grad_norm": 0.3671875, + "learning_rate": 0.0002, + "loss": 1.1187, + "step": 5193 + }, + { + "epoch": 1.6379690949227372, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 1.1107, + "step": 5194 + }, + { + "epoch": 1.6382844528539893, + "grad_norm": 0.408203125, + "learning_rate": 0.0002, + "loss": 1.2427, + "step": 5195 + }, + { + "epoch": 1.6385998107852413, + "grad_norm": 0.435546875, + "learning_rate": 0.0002, + "loss": 1.2581, + "step": 5196 + }, + { + "epoch": 1.6389151687164931, + "grad_norm": 0.4375, + "learning_rate": 0.0002, + "loss": 1.257, + "step": 5197 + }, + { + "epoch": 1.6392305266477452, + "grad_norm": 0.439453125, + "learning_rate": 0.0002, + "loss": 1.1059, + "step": 5198 + }, + { + "epoch": 1.6395458845789972, + "grad_norm": 0.46875, + "learning_rate": 0.0002, + "loss": 1.1255, + "step": 5199 + }, + { + "epoch": 1.639861242510249, + "grad_norm": 0.62890625, + "learning_rate": 0.0002, + "loss": 1.5186, + "step": 5200 + }, + { + "epoch": 1.6401766004415013, + "grad_norm": 0.58203125, + "learning_rate": 0.0002, + "loss": 1.3534, + "step": 5201 + }, + { + "epoch": 1.640491958372753, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 1.3796, + "step": 5202 + }, + { + "epoch": 1.640807316304005, + "grad_norm": 0.70703125, + "learning_rate": 0.0002, + "loss": 1.6939, + "step": 5203 + }, + { + "epoch": 1.6411226742352572, + "grad_norm": 0.75, + "learning_rate": 0.0002, + "loss": 1.6228, + "step": 5204 + }, + { + "epoch": 1.641438032166509, + "grad_norm": 0.95703125, + "learning_rate": 0.0002, + "loss": 2.283, + "step": 5205 + }, + { + "epoch": 1.641753390097761, + "grad_norm": 0.84765625, + "learning_rate": 0.0002, + "loss": 1.6913, + "step": 5206 + }, + { + "epoch": 1.642068748029013, + "grad_norm": 0.83984375, + "learning_rate": 0.0002, + "loss": 1.8686, + "step": 5207 + }, + { + "epoch": 1.6423841059602649, + "grad_norm": 0.86328125, + "learning_rate": 0.0002, + "loss": 1.8952, + "step": 5208 + }, + { + "epoch": 1.6426994638915169, + "grad_norm": 0.8984375, + "learning_rate": 0.0002, + "loss": 1.9856, + "step": 5209 + }, + { + "epoch": 1.643014821822769, + "grad_norm": 1.2109375, + "learning_rate": 0.0002, + "loss": 2.4723, + "step": 5210 + }, + { + "epoch": 1.6433301797540207, + "grad_norm": 0.890625, + "learning_rate": 0.0002, + "loss": 2.1353, + "step": 5211 + }, + { + "epoch": 1.6436455376852728, + "grad_norm": 0.92578125, + "learning_rate": 0.0002, + "loss": 2.0041, + "step": 5212 + }, + { + "epoch": 1.6439608956165248, + "grad_norm": 1.2109375, + "learning_rate": 0.0002, + "loss": 2.4213, + "step": 5213 + }, + { + "epoch": 1.6442762535477766, + "grad_norm": 0.875, + "learning_rate": 0.0002, + "loss": 1.7106, + "step": 5214 + }, + { + "epoch": 1.6445916114790287, + "grad_norm": 0.94140625, + "learning_rate": 0.0002, + "loss": 1.8976, + "step": 5215 + }, + { + "epoch": 1.6449069694102807, + "grad_norm": 1.25, + "learning_rate": 0.0002, + "loss": 2.4777, + "step": 5216 + }, + { + "epoch": 1.6452223273415325, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.9472, + "step": 5217 + }, + { + "epoch": 1.6455376852727848, + "grad_norm": 0.91796875, + "learning_rate": 0.0002, + "loss": 1.8547, + "step": 5218 + }, + { + "epoch": 1.6458530432040366, + "grad_norm": 0.87890625, + "learning_rate": 0.0002, + "loss": 1.7471, + "step": 5219 + }, + { + "epoch": 1.6461684011352884, + "grad_norm": 1.2734375, + "learning_rate": 0.0002, + "loss": 1.7565, + "step": 5220 + }, + { + "epoch": 1.6464837590665407, + "grad_norm": 1.296875, + "learning_rate": 0.0002, + "loss": 2.1827, + "step": 5221 + }, + { + "epoch": 1.6467991169977925, + "grad_norm": 0.197265625, + "learning_rate": 0.0002, + "loss": 1.0397, + "step": 5222 + }, + { + "epoch": 1.6471144749290445, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 0.9354, + "step": 5223 + }, + { + "epoch": 1.6474298328602965, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 1.0825, + "step": 5224 + }, + { + "epoch": 1.6477451907915484, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.0576, + "step": 5225 + }, + { + "epoch": 1.6480605487228004, + "grad_norm": 0.2119140625, + "learning_rate": 0.0002, + "loss": 1.0531, + "step": 5226 + }, + { + "epoch": 1.6483759066540524, + "grad_norm": 0.220703125, + "learning_rate": 0.0002, + "loss": 1.108, + "step": 5227 + }, + { + "epoch": 1.6486912645853042, + "grad_norm": 0.2275390625, + "learning_rate": 0.0002, + "loss": 1.0739, + "step": 5228 + }, + { + "epoch": 1.6490066225165563, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 1.0278, + "step": 5229 + }, + { + "epoch": 1.6493219804478083, + "grad_norm": 0.2216796875, + "learning_rate": 0.0002, + "loss": 1.0087, + "step": 5230 + }, + { + "epoch": 1.6496373383790601, + "grad_norm": 0.2470703125, + "learning_rate": 0.0002, + "loss": 1.0319, + "step": 5231 + }, + { + "epoch": 1.6499526963103122, + "grad_norm": 0.2265625, + "learning_rate": 0.0002, + "loss": 0.9943, + "step": 5232 + }, + { + "epoch": 1.6502680542415642, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.0676, + "step": 5233 + }, + { + "epoch": 1.650583412172816, + "grad_norm": 0.2451171875, + "learning_rate": 0.0002, + "loss": 1.0844, + "step": 5234 + }, + { + "epoch": 1.6508987701040683, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 1.0191, + "step": 5235 + }, + { + "epoch": 1.65121412803532, + "grad_norm": 0.23046875, + "learning_rate": 0.0002, + "loss": 1.0023, + "step": 5236 + }, + { + "epoch": 1.65121412803532, + "eval_loss": 1.677410364151001, + "eval_runtime": 148.3937, + "eval_samples_per_second": 6.739, + "eval_steps_per_second": 6.739, + "step": 5236 + }, + { + "epoch": 1.65121412803532, + "mmlu_eval_accuracy": 0.5977661524083827, + "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, + "mmlu_eval_accuracy_anatomy": 0.5, + "mmlu_eval_accuracy_astronomy": 0.625, + "mmlu_eval_accuracy_business_ethics": 0.7272727272727273, + "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138, + "mmlu_eval_accuracy_college_biology": 0.875, + "mmlu_eval_accuracy_college_chemistry": 0.375, + "mmlu_eval_accuracy_college_computer_science": 0.5454545454545454, + "mmlu_eval_accuracy_college_mathematics": 0.18181818181818182, + "mmlu_eval_accuracy_college_medicine": 0.5909090909090909, + "mmlu_eval_accuracy_college_physics": 0.5454545454545454, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231, + "mmlu_eval_accuracy_econometrics": 0.5, + "mmlu_eval_accuracy_electrical_engineering": 0.5, + "mmlu_eval_accuracy_elementary_mathematics": 0.4634146341463415, + "mmlu_eval_accuracy_formal_logic": 0.42857142857142855, + "mmlu_eval_accuracy_global_facts": 0.3, + "mmlu_eval_accuracy_high_school_biology": 0.65625, + "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091, + "mmlu_eval_accuracy_high_school_computer_science": 0.7777777777777778, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.9545454545454546, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325, + "mmlu_eval_accuracy_high_school_mathematics": 0.3103448275862069, + "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384, + "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, + "mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667, + "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, + "mmlu_eval_accuracy_high_school_us_history": 0.8636363636363636, + "mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693, + "mmlu_eval_accuracy_human_aging": 0.6086956521739131, + "mmlu_eval_accuracy_human_sexuality": 0.5, + "mmlu_eval_accuracy_international_law": 0.9230769230769231, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.36363636363636365, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.76, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7209302325581395, + "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, + "mmlu_eval_accuracy_moral_scenarios": 0.31, + "mmlu_eval_accuracy_nutrition": 0.7575757575757576, + "mmlu_eval_accuracy_philosophy": 0.6764705882352942, + "mmlu_eval_accuracy_prehistory": 0.6285714285714286, + "mmlu_eval_accuracy_professional_accounting": 0.2903225806451613, + "mmlu_eval_accuracy_professional_law": 0.4, + "mmlu_eval_accuracy_professional_medicine": 0.6774193548387096, + "mmlu_eval_accuracy_professional_psychology": 0.6811594202898551, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.6666666666666666, + "mmlu_eval_accuracy_sociology": 1.0, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.4444444444444444, + "mmlu_eval_accuracy_world_religions": 0.7368421052631579, + "mmlu_loss": 5.650790007109738, + "step": 5236 + }, + { + "epoch": 1.651529485966572, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 0.9535, + "step": 5237 + }, + { + "epoch": 1.6518448438978242, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 0.9762, + "step": 5238 + }, + { + "epoch": 1.652160201829076, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 1.1512, + "step": 5239 + }, + { + "epoch": 1.652475559760328, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 1.2372, + "step": 5240 + }, + { + "epoch": 1.65279091769158, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 1.1323, + "step": 5241 + }, + { + "epoch": 1.6531062756228319, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 1.3918, + "step": 5242 + }, + { + "epoch": 1.653421633554084, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 1.1014, + "step": 5243 + }, + { + "epoch": 1.653736991485336, + "grad_norm": 0.3828125, + "learning_rate": 0.0002, + "loss": 1.3713, + "step": 5244 + }, + { + "epoch": 1.6540523494165877, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 1.1076, + "step": 5245 + }, + { + "epoch": 1.6543677073478398, + "grad_norm": 0.421875, + "learning_rate": 0.0002, + "loss": 1.2154, + "step": 5246 + }, + { + "epoch": 1.6546830652790918, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 1.2664, + "step": 5247 + }, + { + "epoch": 1.6549984232103436, + "grad_norm": 0.40625, + "learning_rate": 0.0002, + "loss": 1.1866, + "step": 5248 + }, + { + "epoch": 1.6553137811415957, + "grad_norm": 0.3984375, + "learning_rate": 0.0002, + "loss": 1.1841, + "step": 5249 + }, + { + "epoch": 1.6556291390728477, + "grad_norm": 0.408203125, + "learning_rate": 0.0002, + "loss": 1.1904, + "step": 5250 + }, + { + "epoch": 1.6559444970040995, + "grad_norm": 0.486328125, + "learning_rate": 0.0002, + "loss": 1.1292, + "step": 5251 + }, + { + "epoch": 1.6562598549353518, + "grad_norm": 0.50390625, + "learning_rate": 0.0002, + "loss": 1.2643, + "step": 5252 + }, + { + "epoch": 1.6565752128666036, + "grad_norm": 0.46875, + "learning_rate": 0.0002, + "loss": 1.2785, + "step": 5253 + }, + { + "epoch": 1.6568905707978554, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 1.1895, + "step": 5254 + }, + { + "epoch": 1.6572059287291077, + "grad_norm": 0.66796875, + "learning_rate": 0.0002, + "loss": 1.6413, + "step": 5255 + }, + { + "epoch": 1.6575212866603595, + "grad_norm": 0.76953125, + "learning_rate": 0.0002, + "loss": 1.7528, + "step": 5256 + }, + { + "epoch": 1.6578366445916115, + "grad_norm": 0.87109375, + "learning_rate": 0.0002, + "loss": 1.7082, + "step": 5257 + }, + { + "epoch": 1.6581520025228635, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 5258 + }, + { + "epoch": 1.6584673604541154, + "grad_norm": 0.71484375, + "learning_rate": 0.0002, + "loss": 1.3889, + "step": 5259 + }, + { + "epoch": 1.6587827183853674, + "grad_norm": 0.76171875, + "learning_rate": 0.0002, + "loss": 1.7272, + "step": 5260 + }, + { + "epoch": 1.6590980763166194, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 1.709, + "step": 5261 + }, + { + "epoch": 1.6594134342478712, + "grad_norm": 1.1171875, + "learning_rate": 0.0002, + "loss": 1.7522, + "step": 5262 + }, + { + "epoch": 1.6597287921791233, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 2.033, + "step": 5263 + }, + { + "epoch": 1.6600441501103753, + "grad_norm": 0.9140625, + "learning_rate": 0.0002, + "loss": 2.0515, + "step": 5264 + }, + { + "epoch": 1.6603595080416271, + "grad_norm": 1.1171875, + "learning_rate": 0.0002, + "loss": 2.3653, + "step": 5265 + }, + { + "epoch": 1.6606748659728792, + "grad_norm": 1.2109375, + "learning_rate": 0.0002, + "loss": 2.0403, + "step": 5266 + }, + { + "epoch": 1.6609902239041312, + "grad_norm": 0.921875, + "learning_rate": 0.0002, + "loss": 1.9726, + "step": 5267 + }, + { + "epoch": 1.661305581835383, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 2.1711, + "step": 5268 + }, + { + "epoch": 1.6616209397666353, + "grad_norm": 1.0859375, + "learning_rate": 0.0002, + "loss": 2.132, + "step": 5269 + }, + { + "epoch": 1.661936297697887, + "grad_norm": 0.91796875, + "learning_rate": 0.0002, + "loss": 1.7789, + "step": 5270 + }, + { + "epoch": 1.6622516556291391, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 2.0748, + "step": 5271 + }, + { + "epoch": 1.6625670135603912, + "grad_norm": 0.19140625, + "learning_rate": 0.0002, + "loss": 0.9947, + "step": 5272 + }, + { + "epoch": 1.662882371491643, + "grad_norm": 0.1943359375, + "learning_rate": 0.0002, + "loss": 1.1443, + "step": 5273 + }, + { + "epoch": 1.663197729422895, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.1537, + "step": 5274 + }, + { + "epoch": 1.663513087354147, + "grad_norm": 0.2001953125, + "learning_rate": 0.0002, + "loss": 0.94, + "step": 5275 + }, + { + "epoch": 1.6638284452853989, + "grad_norm": 0.2041015625, + "learning_rate": 0.0002, + "loss": 1.0364, + "step": 5276 + }, + { + "epoch": 1.664143803216651, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 0.9712, + "step": 5277 + }, + { + "epoch": 1.664459161147903, + "grad_norm": 0.212890625, + "learning_rate": 0.0002, + "loss": 1.0381, + "step": 5278 + }, + { + "epoch": 1.6647745190791547, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.1168, + "step": 5279 + }, + { + "epoch": 1.6650898770104068, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.019, + "step": 5280 + }, + { + "epoch": 1.6654052349416588, + "grad_norm": 0.216796875, + "learning_rate": 0.0002, + "loss": 0.9255, + "step": 5281 + }, + { + "epoch": 1.6657205928729106, + "grad_norm": 0.240234375, + "learning_rate": 0.0002, + "loss": 1.1948, + "step": 5282 + }, + { + "epoch": 1.666035950804163, + "grad_norm": 0.2431640625, + "learning_rate": 0.0002, + "loss": 1.0241, + "step": 5283 + }, + { + "epoch": 1.6663513087354147, + "grad_norm": 0.2333984375, + "learning_rate": 0.0002, + "loss": 0.9733, + "step": 5284 + }, + { + "epoch": 1.6666666666666665, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 1.0249, + "step": 5285 + }, + { + "epoch": 1.6669820245979188, + "grad_norm": 0.255859375, + "learning_rate": 0.0002, + "loss": 0.8493, + "step": 5286 + }, + { + "epoch": 1.6672973825291706, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 1.1853, + "step": 5287 + }, + { + "epoch": 1.6676127404604226, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 1.1024, + "step": 5288 + }, + { + "epoch": 1.6679280983916747, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 5289 + }, + { + "epoch": 1.6682434563229265, + "grad_norm": 0.330078125, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 5290 + }, + { + "epoch": 1.6685588142541785, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 1.1073, + "step": 5291 + }, + { + "epoch": 1.6688741721854305, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 1.0734, + "step": 5292 + }, + { + "epoch": 1.6691895301166824, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 1.0194, + "step": 5293 + }, + { + "epoch": 1.6695048880479344, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 1.0898, + "step": 5294 + }, + { + "epoch": 1.6698202459791864, + "grad_norm": 0.43359375, + "learning_rate": 0.0002, + "loss": 1.2521, + "step": 5295 + }, + { + "epoch": 1.6701356039104382, + "grad_norm": 0.392578125, + "learning_rate": 0.0002, + "loss": 1.1242, + "step": 5296 + }, + { + "epoch": 1.6704509618416903, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 1.1322, + "step": 5297 + }, + { + "epoch": 1.6707663197729423, + "grad_norm": 0.3828125, + "learning_rate": 0.0002, + "loss": 1.043, + "step": 5298 + }, + { + "epoch": 1.6710816777041941, + "grad_norm": 0.44140625, + "learning_rate": 0.0002, + "loss": 1.2605, + "step": 5299 + }, + { + "epoch": 1.6713970356354464, + "grad_norm": 0.43359375, + "learning_rate": 0.0002, + "loss": 1.223, + "step": 5300 + }, + { + "epoch": 1.6717123935666982, + "grad_norm": 0.455078125, + "learning_rate": 0.0002, + "loss": 1.151, + "step": 5301 + }, + { + "epoch": 1.67202775149795, + "grad_norm": 0.66796875, + "learning_rate": 0.0002, + "loss": 1.3042, + "step": 5302 + }, + { + "epoch": 1.6723431094292023, + "grad_norm": 0.61328125, + "learning_rate": 0.0002, + "loss": 1.1996, + "step": 5303 + }, + { + "epoch": 1.672658467360454, + "grad_norm": 0.6796875, + "learning_rate": 0.0002, + "loss": 1.7224, + "step": 5304 + }, + { + "epoch": 1.6729738252917061, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 1.4452, + "step": 5305 + }, + { + "epoch": 1.6732891832229582, + "grad_norm": 0.796875, + "learning_rate": 0.0002, + "loss": 1.6726, + "step": 5306 + }, + { + "epoch": 1.67360454115421, + "grad_norm": 0.87890625, + "learning_rate": 0.0002, + "loss": 1.7403, + "step": 5307 + }, + { + "epoch": 1.673919899085462, + "grad_norm": 0.82421875, + "learning_rate": 0.0002, + "loss": 2.0744, + "step": 5308 + }, + { + "epoch": 1.674235257016714, + "grad_norm": 0.93359375, + "learning_rate": 0.0002, + "loss": 1.9351, + "step": 5309 + }, + { + "epoch": 1.6745506149479659, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 1.9561, + "step": 5310 + }, + { + "epoch": 1.674865972879218, + "grad_norm": 0.953125, + "learning_rate": 0.0002, + "loss": 2.2286, + "step": 5311 + }, + { + "epoch": 1.67518133081047, + "grad_norm": 0.85546875, + "learning_rate": 0.0002, + "loss": 1.9137, + "step": 5312 + }, + { + "epoch": 1.6754966887417218, + "grad_norm": 1.0, + "learning_rate": 0.0002, + "loss": 2.0911, + "step": 5313 + }, + { + "epoch": 1.6758120466729738, + "grad_norm": 0.9375, + "learning_rate": 0.0002, + "loss": 2.0646, + "step": 5314 + }, + { + "epoch": 1.6761274046042258, + "grad_norm": 1.5078125, + "learning_rate": 0.0002, + "loss": 2.5882, + "step": 5315 + }, + { + "epoch": 1.6764427625354776, + "grad_norm": 1.21875, + "learning_rate": 0.0002, + "loss": 2.5012, + "step": 5316 + }, + { + "epoch": 1.67675812046673, + "grad_norm": 1.1171875, + "learning_rate": 0.0002, + "loss": 2.4959, + "step": 5317 + }, + { + "epoch": 1.6770734783979817, + "grad_norm": 1.0390625, + "learning_rate": 0.0002, + "loss": 2.423, + "step": 5318 + }, + { + "epoch": 1.6773888363292335, + "grad_norm": 1.09375, + "learning_rate": 0.0002, + "loss": 2.2668, + "step": 5319 + }, + { + "epoch": 1.6777041942604858, + "grad_norm": 0.8984375, + "learning_rate": 0.0002, + "loss": 1.7261, + "step": 5320 + }, + { + "epoch": 1.6780195521917376, + "grad_norm": 1.46875, + "learning_rate": 0.0002, + "loss": 2.0442, + "step": 5321 + }, + { + "epoch": 1.6783349101229896, + "grad_norm": 0.23046875, + "learning_rate": 0.0002, + "loss": 1.0662, + "step": 5322 + }, + { + "epoch": 1.6786502680542417, + "grad_norm": 0.2177734375, + "learning_rate": 0.0002, + "loss": 0.9794, + "step": 5323 + }, + { + "epoch": 1.6789656259854935, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 1.1073, + "step": 5324 + }, + { + "epoch": 1.6792809839167455, + "grad_norm": 0.22265625, + "learning_rate": 0.0002, + "loss": 1.1259, + "step": 5325 + }, + { + "epoch": 1.6795963418479976, + "grad_norm": 0.224609375, + "learning_rate": 0.0002, + "loss": 1.1707, + "step": 5326 + }, + { + "epoch": 1.6799116997792494, + "grad_norm": 0.2294921875, + "learning_rate": 0.0002, + "loss": 1.0623, + "step": 5327 + }, + { + "epoch": 1.6802270577105014, + "grad_norm": 0.22265625, + "learning_rate": 0.0002, + "loss": 0.9843, + "step": 5328 + }, + { + "epoch": 1.6805424156417534, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 1.0302, + "step": 5329 + }, + { + "epoch": 1.6808577735730053, + "grad_norm": 0.23046875, + "learning_rate": 0.0002, + "loss": 1.0638, + "step": 5330 + }, + { + "epoch": 1.6811731315042573, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 1.0047, + "step": 5331 + }, + { + "epoch": 1.6814884894355093, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 1.1273, + "step": 5332 + }, + { + "epoch": 1.6818038473667611, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 1.2207, + "step": 5333 + }, + { + "epoch": 1.6821192052980134, + "grad_norm": 0.2490234375, + "learning_rate": 0.0002, + "loss": 1.0821, + "step": 5334 + }, + { + "epoch": 1.6824345632292652, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 0.9852, + "step": 5335 + }, + { + "epoch": 1.682749921160517, + "grad_norm": 0.244140625, + "learning_rate": 0.0002, + "loss": 0.9236, + "step": 5336 + }, + { + "epoch": 1.6830652790917693, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.0797, + "step": 5337 + }, + { + "epoch": 1.683380637023021, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 1.2524, + "step": 5338 + }, + { + "epoch": 1.6836959949542731, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 1.0977, + "step": 5339 + }, + { + "epoch": 1.6840113528855252, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 1.0824, + "step": 5340 + }, + { + "epoch": 1.684326710816777, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 1.2604, + "step": 5341 + }, + { + "epoch": 1.684642068748029, + "grad_norm": 0.34765625, + "learning_rate": 0.0002, + "loss": 1.084, + "step": 5342 + }, + { + "epoch": 1.684957426679281, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 1.007, + "step": 5343 + }, + { + "epoch": 1.6852727846105329, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 1.1202, + "step": 5344 + }, + { + "epoch": 1.685588142541785, + "grad_norm": 0.43359375, + "learning_rate": 0.0002, + "loss": 1.2282, + "step": 5345 + }, + { + "epoch": 1.685903500473037, + "grad_norm": 0.37109375, + "learning_rate": 0.0002, + "loss": 1.1095, + "step": 5346 + }, + { + "epoch": 1.6862188584042888, + "grad_norm": 0.41796875, + "learning_rate": 0.0002, + "loss": 1.2698, + "step": 5347 + }, + { + "epoch": 1.6865342163355408, + "grad_norm": 0.435546875, + "learning_rate": 0.0002, + "loss": 1.1777, + "step": 5348 + }, + { + "epoch": 1.6868495742667928, + "grad_norm": 0.4296875, + "learning_rate": 0.0002, + "loss": 1.0365, + "step": 5349 + }, + { + "epoch": 1.6871649321980446, + "grad_norm": 0.423828125, + "learning_rate": 0.0002, + "loss": 1.013, + "step": 5350 + }, + { + "epoch": 1.687480290129297, + "grad_norm": 0.55859375, + "learning_rate": 0.0002, + "loss": 1.3523, + "step": 5351 + }, + { + "epoch": 1.6877956480605487, + "grad_norm": 0.53515625, + "learning_rate": 0.0002, + "loss": 1.2795, + "step": 5352 + }, + { + "epoch": 1.6881110059918005, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 1.3831, + "step": 5353 + }, + { + "epoch": 1.6884263639230528, + "grad_norm": 1.5, + "learning_rate": 0.0002, + "loss": 1.5866, + "step": 5354 + }, + { + "epoch": 1.6887417218543046, + "grad_norm": 0.94921875, + "learning_rate": 0.0002, + "loss": 1.5665, + "step": 5355 + }, + { + "epoch": 1.6890570797855566, + "grad_norm": 0.7890625, + "learning_rate": 0.0002, + "loss": 1.6476, + "step": 5356 + }, + { + "epoch": 1.6893724377168087, + "grad_norm": 0.6796875, + "learning_rate": 0.0002, + "loss": 1.5802, + "step": 5357 + }, + { + "epoch": 1.6896877956480605, + "grad_norm": 0.75, + "learning_rate": 0.0002, + "loss": 1.7057, + "step": 5358 + }, + { + "epoch": 1.6900031535793125, + "grad_norm": 0.78125, + "learning_rate": 0.0002, + "loss": 1.8633, + "step": 5359 + }, + { + "epoch": 1.6903185115105646, + "grad_norm": 0.78125, + "learning_rate": 0.0002, + "loss": 1.8265, + "step": 5360 + }, + { + "epoch": 1.6906338694418164, + "grad_norm": 0.94921875, + "learning_rate": 0.0002, + "loss": 1.8953, + "step": 5361 + }, + { + "epoch": 1.6909492273730684, + "grad_norm": 0.8046875, + "learning_rate": 0.0002, + "loss": 1.7491, + "step": 5362 + }, + { + "epoch": 1.6912645853043204, + "grad_norm": 0.9765625, + "learning_rate": 0.0002, + "loss": 2.0843, + "step": 5363 + }, + { + "epoch": 1.6915799432355723, + "grad_norm": 1.125, + "learning_rate": 0.0002, + "loss": 2.5152, + "step": 5364 + }, + { + "epoch": 1.6918953011668243, + "grad_norm": 0.84765625, + "learning_rate": 0.0002, + "loss": 1.9461, + "step": 5365 + }, + { + "epoch": 1.6922106590980763, + "grad_norm": 1.15625, + "learning_rate": 0.0002, + "loss": 2.6891, + "step": 5366 + }, + { + "epoch": 1.6925260170293281, + "grad_norm": 0.84765625, + "learning_rate": 0.0002, + "loss": 1.8536, + "step": 5367 + }, + { + "epoch": 1.6928413749605804, + "grad_norm": 1.1640625, + "learning_rate": 0.0002, + "loss": 2.1295, + "step": 5368 + }, + { + "epoch": 1.6931567328918322, + "grad_norm": 0.82421875, + "learning_rate": 0.0002, + "loss": 1.7316, + "step": 5369 + }, + { + "epoch": 1.6934720908230843, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 1.9872, + "step": 5370 + }, + { + "epoch": 1.6937874487543363, + "grad_norm": 1.1640625, + "learning_rate": 0.0002, + "loss": 2.4541, + "step": 5371 + }, + { + "epoch": 1.694102806685588, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 1.0449, + "step": 5372 + }, + { + "epoch": 1.6944181646168401, + "grad_norm": 0.2041015625, + "learning_rate": 0.0002, + "loss": 1.0591, + "step": 5373 + }, + { + "epoch": 1.6947335225480922, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 0.9826, + "step": 5374 + }, + { + "epoch": 1.695048880479344, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 1.1642, + "step": 5375 + }, + { + "epoch": 1.695364238410596, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.018, + "step": 5376 + }, + { + "epoch": 1.695679596341848, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.139, + "step": 5377 + }, + { + "epoch": 1.6959949542730999, + "grad_norm": 0.212890625, + "learning_rate": 0.0002, + "loss": 0.9934, + "step": 5378 + }, + { + "epoch": 1.696310312204352, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.0302, + "step": 5379 + }, + { + "epoch": 1.696625670135604, + "grad_norm": 0.2392578125, + "learning_rate": 0.0002, + "loss": 0.9811, + "step": 5380 + }, + { + "epoch": 1.6969410280668558, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 0.9649, + "step": 5381 + }, + { + "epoch": 1.6972563859981078, + "grad_norm": 0.232421875, + "learning_rate": 0.0002, + "loss": 1.1059, + "step": 5382 + }, + { + "epoch": 1.6975717439293598, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 0.8933, + "step": 5383 + }, + { + "epoch": 1.6978871018606116, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 1.1247, + "step": 5384 + }, + { + "epoch": 1.698202459791864, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.1079, + "step": 5385 + }, + { + "epoch": 1.6985178177231157, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 1.1411, + "step": 5386 + }, + { + "epoch": 1.6988331756543678, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 5387 + }, + { + "epoch": 1.6991485335856198, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 1.2477, + "step": 5388 + }, + { + "epoch": 1.6994638915168716, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 1.0398, + "step": 5389 + }, + { + "epoch": 1.6997792494481236, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 1.1656, + "step": 5390 + }, + { + "epoch": 1.7000946073793757, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 1.2656, + "step": 5391 + }, + { + "epoch": 1.7004099653106275, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 0.9969, + "step": 5392 + }, + { + "epoch": 1.7007253232418795, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 1.1842, + "step": 5393 + }, + { + "epoch": 1.7010406811731316, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 1.1986, + "step": 5394 + }, + { + "epoch": 1.7013560391043834, + "grad_norm": 0.361328125, + "learning_rate": 0.0002, + "loss": 1.1337, + "step": 5395 + }, + { + "epoch": 1.7016713970356354, + "grad_norm": 0.349609375, + "learning_rate": 0.0002, + "loss": 1.1574, + "step": 5396 + }, + { + "epoch": 1.7019867549668874, + "grad_norm": 0.431640625, + "learning_rate": 0.0002, + "loss": 1.3805, + "step": 5397 + }, + { + "epoch": 1.7023021128981393, + "grad_norm": 0.484375, + "learning_rate": 0.0002, + "loss": 1.4126, + "step": 5398 + }, + { + "epoch": 1.7026174708293915, + "grad_norm": 0.462890625, + "learning_rate": 0.0002, + "loss": 1.4, + "step": 5399 + }, + { + "epoch": 1.7029328287606433, + "grad_norm": 0.41796875, + "learning_rate": 0.0002, + "loss": 1.1541, + "step": 5400 + }, + { + "epoch": 1.7032481866918951, + "grad_norm": 0.478515625, + "learning_rate": 0.0002, + "loss": 1.1806, + "step": 5401 + }, + { + "epoch": 1.7035635446231474, + "grad_norm": 0.58203125, + "learning_rate": 0.0002, + "loss": 1.4352, + "step": 5402 + }, + { + "epoch": 1.7038789025543992, + "grad_norm": 0.58984375, + "learning_rate": 0.0002, + "loss": 1.4763, + "step": 5403 + }, + { + "epoch": 1.7041942604856513, + "grad_norm": 0.62890625, + "learning_rate": 0.0002, + "loss": 1.4821, + "step": 5404 + }, + { + "epoch": 1.7045096184169033, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 1.403, + "step": 5405 + }, + { + "epoch": 1.704824976348155, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 1.4068, + "step": 5406 + }, + { + "epoch": 1.7051403342794071, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 1.8496, + "step": 5407 + }, + { + "epoch": 1.7054556922106592, + "grad_norm": 0.84375, + "learning_rate": 0.0002, + "loss": 1.7865, + "step": 5408 + }, + { + "epoch": 1.705771050141911, + "grad_norm": 0.875, + "learning_rate": 0.0002, + "loss": 2.0217, + "step": 5409 + }, + { + "epoch": 1.706086408073163, + "grad_norm": 0.7890625, + "learning_rate": 0.0002, + "loss": 1.9074, + "step": 5410 + }, + { + "epoch": 1.706401766004415, + "grad_norm": 0.80859375, + "learning_rate": 0.0002, + "loss": 1.8542, + "step": 5411 + }, + { + "epoch": 1.7067171239356669, + "grad_norm": 0.96484375, + "learning_rate": 0.0002, + "loss": 2.1501, + "step": 5412 + }, + { + "epoch": 1.707032481866919, + "grad_norm": 0.90234375, + "learning_rate": 0.0002, + "loss": 2.0355, + "step": 5413 + }, + { + "epoch": 1.707347839798171, + "grad_norm": 0.94921875, + "learning_rate": 0.0002, + "loss": 1.9515, + "step": 5414 + }, + { + "epoch": 1.7076631977294228, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 1.9763, + "step": 5415 + }, + { + "epoch": 1.707978555660675, + "grad_norm": 1.09375, + "learning_rate": 0.0002, + "loss": 1.9227, + "step": 5416 + }, + { + "epoch": 1.7082939135919268, + "grad_norm": 1.1171875, + "learning_rate": 0.0002, + "loss": 1.9985, + "step": 5417 + }, + { + "epoch": 1.7086092715231787, + "grad_norm": 1.03125, + "learning_rate": 0.0002, + "loss": 1.9394, + "step": 5418 + }, + { + "epoch": 1.708924629454431, + "grad_norm": 0.98046875, + "learning_rate": 0.0002, + "loss": 1.6718, + "step": 5419 + }, + { + "epoch": 1.7092399873856827, + "grad_norm": 0.984375, + "learning_rate": 0.0002, + "loss": 1.8309, + "step": 5420 + }, + { + "epoch": 1.7095553453169348, + "grad_norm": 0.84765625, + "learning_rate": 0.0002, + "loss": 1.6116, + "step": 5421 + }, + { + "epoch": 1.7098707032481868, + "grad_norm": 0.2001953125, + "learning_rate": 0.0002, + "loss": 1.0574, + "step": 5422 + }, + { + "epoch": 1.7101860611794386, + "grad_norm": 0.1943359375, + "learning_rate": 0.0002, + "loss": 1.0365, + "step": 5423 + }, + { + "epoch": 1.7101860611794386, + "eval_loss": 1.6905437707901, + "eval_runtime": 149.0528, + "eval_samples_per_second": 6.709, + "eval_steps_per_second": 6.709, + "step": 5423 + }, + { + "epoch": 1.7101860611794386, + "mmlu_eval_accuracy": 0.6056015971219283, + "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, + "mmlu_eval_accuracy_anatomy": 0.5714285714285714, + "mmlu_eval_accuracy_astronomy": 0.625, + "mmlu_eval_accuracy_business_ethics": 0.8181818181818182, + "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034, + "mmlu_eval_accuracy_college_biology": 0.8125, + "mmlu_eval_accuracy_college_chemistry": 0.125, + "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, + "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454, + "mmlu_eval_accuracy_college_medicine": 0.6363636363636364, + "mmlu_eval_accuracy_college_physics": 0.45454545454545453, + "mmlu_eval_accuracy_computer_security": 0.5454545454545454, + "mmlu_eval_accuracy_conceptual_physics": 0.5, + "mmlu_eval_accuracy_econometrics": 0.4166666666666667, + "mmlu_eval_accuracy_electrical_engineering": 0.5625, + "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244, + "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, + "mmlu_eval_accuracy_global_facts": 0.2, + "mmlu_eval_accuracy_high_school_biology": 0.625, + "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.9545454545454546, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8571428571428571, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.6046511627906976, + "mmlu_eval_accuracy_high_school_mathematics": 0.41379310344827586, + "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, + "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, + "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667, + "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, + "mmlu_eval_accuracy_high_school_us_history": 0.8636363636363636, + "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, + "mmlu_eval_accuracy_human_aging": 0.6086956521739131, + "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334, + "mmlu_eval_accuracy_international_law": 0.9230769230769231, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.72, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7325581395348837, + "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, + "mmlu_eval_accuracy_moral_scenarios": 0.28, + "mmlu_eval_accuracy_nutrition": 0.7575757575757576, + "mmlu_eval_accuracy_philosophy": 0.6764705882352942, + "mmlu_eval_accuracy_prehistory": 0.5714285714285714, + "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644, + "mmlu_eval_accuracy_professional_law": 0.3941176470588235, + "mmlu_eval_accuracy_professional_medicine": 0.7419354838709677, + "mmlu_eval_accuracy_professional_psychology": 0.6666666666666666, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7777777777777778, + "mmlu_eval_accuracy_sociology": 1.0, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.5, + "mmlu_eval_accuracy_world_religions": 0.7894736842105263, + "mmlu_loss": 5.898125943757599, + "step": 5423 + }, + { + "epoch": 1.7105014191106906, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.0502, + "step": 5424 + }, + { + "epoch": 1.7108167770419427, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.0441, + "step": 5425 + }, + { + "epoch": 1.7111321349731945, + "grad_norm": 0.224609375, + "learning_rate": 0.0002, + "loss": 1.0738, + "step": 5426 + }, + { + "epoch": 1.7114474929044465, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 0.9939, + "step": 5427 + }, + { + "epoch": 1.7117628508356986, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.0122, + "step": 5428 + }, + { + "epoch": 1.7120782087669504, + "grad_norm": 0.22265625, + "learning_rate": 0.0002, + "loss": 1.0449, + "step": 5429 + }, + { + "epoch": 1.7123935666982024, + "grad_norm": 0.2294921875, + "learning_rate": 0.0002, + "loss": 1.002, + "step": 5430 + }, + { + "epoch": 1.7127089246294545, + "grad_norm": 0.2431640625, + "learning_rate": 0.0002, + "loss": 0.9811, + "step": 5431 + }, + { + "epoch": 1.7130242825607063, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 1.1065, + "step": 5432 + }, + { + "epoch": 1.7133396404919585, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.0183, + "step": 5433 + }, + { + "epoch": 1.7136549984232103, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.0993, + "step": 5434 + }, + { + "epoch": 1.7139703563544622, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.089, + "step": 5435 + }, + { + "epoch": 1.7142857142857144, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.063, + "step": 5436 + }, + { + "epoch": 1.7146010722169662, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 1.1489, + "step": 5437 + }, + { + "epoch": 1.7149164301482183, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.1551, + "step": 5438 + }, + { + "epoch": 1.7152317880794703, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 1.1233, + "step": 5439 + }, + { + "epoch": 1.7155471460107221, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 1.2392, + "step": 5440 + }, + { + "epoch": 1.7158625039419741, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 1.1259, + "step": 5441 + }, + { + "epoch": 1.7161778618732262, + "grad_norm": 0.330078125, + "learning_rate": 0.0002, + "loss": 1.1283, + "step": 5442 + }, + { + "epoch": 1.716493219804478, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 1.253, + "step": 5443 + }, + { + "epoch": 1.71680857773573, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 1.1178, + "step": 5444 + }, + { + "epoch": 1.717123935666982, + "grad_norm": 0.412109375, + "learning_rate": 0.0002, + "loss": 1.2944, + "step": 5445 + }, + { + "epoch": 1.7174392935982339, + "grad_norm": 0.4375, + "learning_rate": 0.0002, + "loss": 1.2851, + "step": 5446 + }, + { + "epoch": 1.717754651529486, + "grad_norm": 0.494140625, + "learning_rate": 0.0002, + "loss": 1.1705, + "step": 5447 + }, + { + "epoch": 1.718070009460738, + "grad_norm": 0.419921875, + "learning_rate": 0.0002, + "loss": 1.1477, + "step": 5448 + }, + { + "epoch": 1.7183853673919898, + "grad_norm": 0.51953125, + "learning_rate": 0.0002, + "loss": 1.191, + "step": 5449 + }, + { + "epoch": 1.718700725323242, + "grad_norm": 0.50390625, + "learning_rate": 0.0002, + "loss": 1.4349, + "step": 5450 + }, + { + "epoch": 1.7190160832544938, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 1.4167, + "step": 5451 + }, + { + "epoch": 1.7193314411857457, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5452 + }, + { + "epoch": 1.719646799116998, + "grad_norm": 0.51953125, + "learning_rate": 0.0002, + "loss": 1.2725, + "step": 5453 + }, + { + "epoch": 1.7199621570482497, + "grad_norm": 0.5390625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 5454 + }, + { + "epoch": 1.7202775149795018, + "grad_norm": 0.93359375, + "learning_rate": 0.0002, + "loss": 1.5855, + "step": 5455 + }, + { + "epoch": 1.7205928729107538, + "grad_norm": 0.6953125, + "learning_rate": 0.0002, + "loss": 1.5193, + "step": 5456 + }, + { + "epoch": 1.7209082308420056, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 1.9073, + "step": 5457 + }, + { + "epoch": 1.7212235887732577, + "grad_norm": 0.80078125, + "learning_rate": 0.0002, + "loss": 1.7586, + "step": 5458 + }, + { + "epoch": 1.7215389467045097, + "grad_norm": 0.82421875, + "learning_rate": 0.0002, + "loss": 1.782, + "step": 5459 + }, + { + "epoch": 1.7218543046357615, + "grad_norm": 1.359375, + "learning_rate": 0.0002, + "loss": 2.7177, + "step": 5460 + }, + { + "epoch": 1.7221696625670135, + "grad_norm": 0.9765625, + "learning_rate": 0.0002, + "loss": 1.7343, + "step": 5461 + }, + { + "epoch": 1.7224850204982656, + "grad_norm": 0.9375, + "learning_rate": 0.0002, + "loss": 2.0733, + "step": 5462 + }, + { + "epoch": 1.7228003784295174, + "grad_norm": 0.97265625, + "learning_rate": 0.0002, + "loss": 2.0822, + "step": 5463 + }, + { + "epoch": 1.7231157363607694, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 2.1431, + "step": 5464 + }, + { + "epoch": 1.7234310942920215, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 2.0567, + "step": 5465 + }, + { + "epoch": 1.7237464522232733, + "grad_norm": 0.84375, + "learning_rate": 0.0002, + "loss": 1.9079, + "step": 5466 + }, + { + "epoch": 1.7240618101545255, + "grad_norm": 0.87890625, + "learning_rate": 0.0002, + "loss": 1.8339, + "step": 5467 + }, + { + "epoch": 1.7243771680857773, + "grad_norm": 0.94140625, + "learning_rate": 0.0002, + "loss": 2.1919, + "step": 5468 + }, + { + "epoch": 1.7246925260170294, + "grad_norm": 0.9375, + "learning_rate": 0.0002, + "loss": 2.3854, + "step": 5469 + }, + { + "epoch": 1.7250078839482814, + "grad_norm": 0.921875, + "learning_rate": 0.0002, + "loss": 1.6934, + "step": 5470 + }, + { + "epoch": 1.7253232418795332, + "grad_norm": 1.3828125, + "learning_rate": 0.0002, + "loss": 2.0414, + "step": 5471 + }, + { + "epoch": 1.7256385998107853, + "grad_norm": 0.1865234375, + "learning_rate": 0.0002, + "loss": 1.0798, + "step": 5472 + }, + { + "epoch": 1.7259539577420373, + "grad_norm": 0.19921875, + "learning_rate": 0.0002, + "loss": 1.0383, + "step": 5473 + }, + { + "epoch": 1.7262693156732891, + "grad_norm": 0.23046875, + "learning_rate": 0.0002, + "loss": 1.0638, + "step": 5474 + }, + { + "epoch": 1.7265846736045412, + "grad_norm": 0.205078125, + "learning_rate": 0.0002, + "loss": 0.9921, + "step": 5475 + }, + { + "epoch": 1.7269000315357932, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 1.0319, + "step": 5476 + }, + { + "epoch": 1.727215389467045, + "grad_norm": 0.220703125, + "learning_rate": 0.0002, + "loss": 0.9344, + "step": 5477 + }, + { + "epoch": 1.727530747398297, + "grad_norm": 0.2412109375, + "learning_rate": 0.0002, + "loss": 1.1506, + "step": 5478 + }, + { + "epoch": 1.727846105329549, + "grad_norm": 0.23828125, + "learning_rate": 0.0002, + "loss": 0.9418, + "step": 5479 + }, + { + "epoch": 1.728161463260801, + "grad_norm": 0.2451171875, + "learning_rate": 0.0002, + "loss": 1.0455, + "step": 5480 + }, + { + "epoch": 1.728476821192053, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.0412, + "step": 5481 + }, + { + "epoch": 1.728792179123305, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.0953, + "step": 5482 + }, + { + "epoch": 1.7291075370545568, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 1.1311, + "step": 5483 + }, + { + "epoch": 1.729422894985809, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.1224, + "step": 5484 + }, + { + "epoch": 1.7297382529170608, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.1871, + "step": 5485 + }, + { + "epoch": 1.7300536108483129, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.1803, + "step": 5486 + }, + { + "epoch": 1.730368968779565, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 1.2147, + "step": 5487 + }, + { + "epoch": 1.7306843267108167, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 1.0592, + "step": 5488 + }, + { + "epoch": 1.7309996846420688, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.0467, + "step": 5489 + }, + { + "epoch": 1.7313150425733208, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.0681, + "step": 5490 + }, + { + "epoch": 1.7316304005045726, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 1.0572, + "step": 5491 + }, + { + "epoch": 1.7319457584358247, + "grad_norm": 0.34765625, + "learning_rate": 0.0002, + "loss": 1.2116, + "step": 5492 + }, + { + "epoch": 1.7322611163670767, + "grad_norm": 0.423828125, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 5493 + }, + { + "epoch": 1.7325764742983285, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 1.1606, + "step": 5494 + }, + { + "epoch": 1.7328918322295805, + "grad_norm": 0.3671875, + "learning_rate": 0.0002, + "loss": 1.1846, + "step": 5495 + }, + { + "epoch": 1.7332071901608326, + "grad_norm": 0.373046875, + "learning_rate": 0.0002, + "loss": 1.0263, + "step": 5496 + }, + { + "epoch": 1.7335225480920844, + "grad_norm": 0.47265625, + "learning_rate": 0.0002, + "loss": 1.0224, + "step": 5497 + }, + { + "epoch": 1.7338379060233367, + "grad_norm": 0.396484375, + "learning_rate": 0.0002, + "loss": 1.1174, + "step": 5498 + }, + { + "epoch": 1.7341532639545885, + "grad_norm": 0.4609375, + "learning_rate": 0.0002, + "loss": 1.2784, + "step": 5499 + }, + { + "epoch": 1.7344686218858403, + "grad_norm": 0.49609375, + "learning_rate": 0.0002, + "loss": 1.1867, + "step": 5500 + }, + { + "epoch": 1.7347839798170925, + "grad_norm": 0.4609375, + "learning_rate": 0.0002, + "loss": 1.1089, + "step": 5501 + }, + { + "epoch": 1.7350993377483444, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 1.4201, + "step": 5502 + }, + { + "epoch": 1.7354146956795964, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 1.5657, + "step": 5503 + }, + { + "epoch": 1.7357300536108484, + "grad_norm": 0.62109375, + "learning_rate": 0.0002, + "loss": 1.4488, + "step": 5504 + }, + { + "epoch": 1.7360454115421002, + "grad_norm": 0.625, + "learning_rate": 0.0002, + "loss": 1.3668, + "step": 5505 + }, + { + "epoch": 1.7363607694733523, + "grad_norm": 0.65625, + "learning_rate": 0.0002, + "loss": 1.6544, + "step": 5506 + }, + { + "epoch": 1.7366761274046043, + "grad_norm": 0.69921875, + "learning_rate": 0.0002, + "loss": 1.559, + "step": 5507 + }, + { + "epoch": 1.7369914853358561, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 1.6882, + "step": 5508 + }, + { + "epoch": 1.7373068432671082, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 1.8012, + "step": 5509 + }, + { + "epoch": 1.7376222011983602, + "grad_norm": 0.765625, + "learning_rate": 0.0002, + "loss": 1.863, + "step": 5510 + }, + { + "epoch": 1.737937559129612, + "grad_norm": 0.94140625, + "learning_rate": 0.0002, + "loss": 2.1261, + "step": 5511 + }, + { + "epoch": 1.738252917060864, + "grad_norm": 0.9296875, + "learning_rate": 0.0002, + "loss": 1.9331, + "step": 5512 + }, + { + "epoch": 1.738568274992116, + "grad_norm": 1.2734375, + "learning_rate": 0.0002, + "loss": 2.3243, + "step": 5513 + }, + { + "epoch": 1.738883632923368, + "grad_norm": 0.96875, + "learning_rate": 0.0002, + "loss": 2.2611, + "step": 5514 + }, + { + "epoch": 1.7391989908546202, + "grad_norm": 1.0390625, + "learning_rate": 0.0002, + "loss": 2.1157, + "step": 5515 + }, + { + "epoch": 1.739514348785872, + "grad_norm": 0.8984375, + "learning_rate": 0.0002, + "loss": 1.9019, + "step": 5516 + }, + { + "epoch": 1.7398297067171238, + "grad_norm": 1.109375, + "learning_rate": 0.0002, + "loss": 2.0002, + "step": 5517 + }, + { + "epoch": 1.740145064648376, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 1.9032, + "step": 5518 + }, + { + "epoch": 1.7404604225796279, + "grad_norm": 1.125, + "learning_rate": 0.0002, + "loss": 1.9648, + "step": 5519 + }, + { + "epoch": 1.7407757805108799, + "grad_norm": 1.0390625, + "learning_rate": 0.0002, + "loss": 1.589, + "step": 5520 + }, + { + "epoch": 1.741091138442132, + "grad_norm": 1.265625, + "learning_rate": 0.0002, + "loss": 2.4533, + "step": 5521 + }, + { + "epoch": 1.7414064963733837, + "grad_norm": 0.1728515625, + "learning_rate": 0.0002, + "loss": 0.9368, + "step": 5522 + }, + { + "epoch": 1.7417218543046358, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.0237, + "step": 5523 + }, + { + "epoch": 1.7420372122358878, + "grad_norm": 0.212890625, + "learning_rate": 0.0002, + "loss": 1.1114, + "step": 5524 + }, + { + "epoch": 1.7423525701671396, + "grad_norm": 0.2177734375, + "learning_rate": 0.0002, + "loss": 1.1346, + "step": 5525 + }, + { + "epoch": 1.7426679280983917, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.0887, + "step": 5526 + }, + { + "epoch": 1.7429832860296437, + "grad_norm": 0.208984375, + "learning_rate": 0.0002, + "loss": 0.9844, + "step": 5527 + }, + { + "epoch": 1.7432986439608955, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.1466, + "step": 5528 + }, + { + "epoch": 1.7436140018921475, + "grad_norm": 0.2119140625, + "learning_rate": 0.0002, + "loss": 1.0388, + "step": 5529 + }, + { + "epoch": 1.7439293598233996, + "grad_norm": 0.22265625, + "learning_rate": 0.0002, + "loss": 1.0541, + "step": 5530 + }, + { + "epoch": 1.7442447177546514, + "grad_norm": 0.349609375, + "learning_rate": 0.0002, + "loss": 1.1185, + "step": 5531 + }, + { + "epoch": 1.7445600756859037, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 1.0652, + "step": 5532 + }, + { + "epoch": 1.7448754336171555, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 1.0123, + "step": 5533 + }, + { + "epoch": 1.7451907915484073, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.1216, + "step": 5534 + }, + { + "epoch": 1.7455061494796595, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 0.9668, + "step": 5535 + }, + { + "epoch": 1.7458215074109114, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 1.0477, + "step": 5536 + }, + { + "epoch": 1.7461368653421634, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.0598, + "step": 5537 + }, + { + "epoch": 1.7464522232734154, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.0918, + "step": 5538 + }, + { + "epoch": 1.7467675812046672, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 1.0227, + "step": 5539 + }, + { + "epoch": 1.7470829391359193, + "grad_norm": 0.353515625, + "learning_rate": 0.0002, + "loss": 1.1321, + "step": 5540 + }, + { + "epoch": 1.7473982970671713, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 1.1828, + "step": 5541 + }, + { + "epoch": 1.7477136549984231, + "grad_norm": 0.330078125, + "learning_rate": 0.0002, + "loss": 1.0675, + "step": 5542 + }, + { + "epoch": 1.7480290129296752, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 1.108, + "step": 5543 + }, + { + "epoch": 1.7483443708609272, + "grad_norm": 0.37109375, + "learning_rate": 0.0002, + "loss": 1.0571, + "step": 5544 + }, + { + "epoch": 1.748659728792179, + "grad_norm": 0.373046875, + "learning_rate": 0.0002, + "loss": 1.1798, + "step": 5545 + }, + { + "epoch": 1.748975086723431, + "grad_norm": 0.44921875, + "learning_rate": 0.0002, + "loss": 1.1903, + "step": 5546 + }, + { + "epoch": 1.749290444654683, + "grad_norm": 0.4296875, + "learning_rate": 0.0002, + "loss": 1.1732, + "step": 5547 + }, + { + "epoch": 1.749605802585935, + "grad_norm": 0.390625, + "learning_rate": 0.0002, + "loss": 1.0631, + "step": 5548 + }, + { + "epoch": 1.7499211605171872, + "grad_norm": 0.41015625, + "learning_rate": 0.0002, + "loss": 1.013, + "step": 5549 + }, + { + "epoch": 1.750236518448439, + "grad_norm": 0.44140625, + "learning_rate": 0.0002, + "loss": 1.2653, + "step": 5550 + }, + { + "epoch": 1.7505518763796908, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 1.3983, + "step": 5551 + }, + { + "epoch": 1.750867234310943, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 1.2, + "step": 5552 + }, + { + "epoch": 1.7511825922421949, + "grad_norm": 0.68359375, + "learning_rate": 0.0002, + "loss": 1.6079, + "step": 5553 + }, + { + "epoch": 1.751497950173447, + "grad_norm": 0.63671875, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 5554 + }, + { + "epoch": 1.751813308104699, + "grad_norm": 0.66796875, + "learning_rate": 0.0002, + "loss": 1.5163, + "step": 5555 + }, + { + "epoch": 1.7521286660359507, + "grad_norm": 0.76953125, + "learning_rate": 0.0002, + "loss": 1.834, + "step": 5556 + }, + { + "epoch": 1.7524440239672028, + "grad_norm": 0.77734375, + "learning_rate": 0.0002, + "loss": 1.795, + "step": 5557 + }, + { + "epoch": 1.7527593818984548, + "grad_norm": 0.70703125, + "learning_rate": 0.0002, + "loss": 1.6452, + "step": 5558 + }, + { + "epoch": 1.7530747398297066, + "grad_norm": 0.72265625, + "learning_rate": 0.0002, + "loss": 1.5421, + "step": 5559 + }, + { + "epoch": 1.7533900977609587, + "grad_norm": 0.6875, + "learning_rate": 0.0002, + "loss": 1.7337, + "step": 5560 + }, + { + "epoch": 1.7537054556922107, + "grad_norm": 0.859375, + "learning_rate": 0.0002, + "loss": 1.8799, + "step": 5561 + }, + { + "epoch": 1.7540208136234625, + "grad_norm": 0.84765625, + "learning_rate": 0.0002, + "loss": 2.0122, + "step": 5562 + }, + { + "epoch": 1.7543361715547146, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 2.0246, + "step": 5563 + }, + { + "epoch": 1.7546515294859666, + "grad_norm": 0.859375, + "learning_rate": 0.0002, + "loss": 1.8296, + "step": 5564 + }, + { + "epoch": 1.7549668874172184, + "grad_norm": 1.0390625, + "learning_rate": 0.0002, + "loss": 2.3264, + "step": 5565 + }, + { + "epoch": 1.7552822453484707, + "grad_norm": 0.98828125, + "learning_rate": 0.0002, + "loss": 1.9371, + "step": 5566 + }, + { + "epoch": 1.7555976032797225, + "grad_norm": 1.015625, + "learning_rate": 0.0002, + "loss": 1.9326, + "step": 5567 + }, + { + "epoch": 1.7559129612109745, + "grad_norm": 0.9765625, + "learning_rate": 0.0002, + "loss": 2.0955, + "step": 5568 + }, + { + "epoch": 1.7562283191422265, + "grad_norm": 1.140625, + "learning_rate": 0.0002, + "loss": 1.564, + "step": 5569 + }, + { + "epoch": 1.7565436770734784, + "grad_norm": 0.97265625, + "learning_rate": 0.0002, + "loss": 1.9556, + "step": 5570 + }, + { + "epoch": 1.7568590350047304, + "grad_norm": 1.140625, + "learning_rate": 0.0002, + "loss": 1.8699, + "step": 5571 + }, + { + "epoch": 1.7571743929359824, + "grad_norm": 0.212890625, + "learning_rate": 0.0002, + "loss": 1.1382, + "step": 5572 + }, + { + "epoch": 1.7574897508672342, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.0132, + "step": 5573 + }, + { + "epoch": 1.7578051087984863, + "grad_norm": 0.212890625, + "learning_rate": 0.0002, + "loss": 1.0463, + "step": 5574 + }, + { + "epoch": 1.7581204667297383, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.0458, + "step": 5575 + }, + { + "epoch": 1.7584358246609901, + "grad_norm": 0.205078125, + "learning_rate": 0.0002, + "loss": 1.0507, + "step": 5576 + }, + { + "epoch": 1.7587511825922422, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 1.0458, + "step": 5577 + }, + { + "epoch": 1.7590665405234942, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.0554, + "step": 5578 + }, + { + "epoch": 1.759381898454746, + "grad_norm": 0.2412109375, + "learning_rate": 0.0002, + "loss": 1.1494, + "step": 5579 + }, + { + "epoch": 1.759697256385998, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.0562, + "step": 5580 + }, + { + "epoch": 1.76001261431725, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 1.005, + "step": 5581 + }, + { + "epoch": 1.760327972248502, + "grad_norm": 0.2373046875, + "learning_rate": 0.0002, + "loss": 1.1213, + "step": 5582 + }, + { + "epoch": 1.7606433301797542, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 1.1646, + "step": 5583 + }, + { + "epoch": 1.760958688111006, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 1.1249, + "step": 5584 + }, + { + "epoch": 1.761274046042258, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.0319, + "step": 5585 + }, + { + "epoch": 1.76158940397351, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.0575, + "step": 5586 + }, + { + "epoch": 1.7619047619047619, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 1.1102, + "step": 5587 + }, + { + "epoch": 1.762220119836014, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.9959, + "step": 5588 + }, + { + "epoch": 1.762535477767266, + "grad_norm": 0.330078125, + "learning_rate": 0.0002, + "loss": 1.168, + "step": 5589 + }, + { + "epoch": 1.7628508356985177, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 1.1459, + "step": 5590 + }, + { + "epoch": 1.7631661936297698, + "grad_norm": 0.41015625, + "learning_rate": 0.0002, + "loss": 1.3031, + "step": 5591 + }, + { + "epoch": 1.7634815515610218, + "grad_norm": 0.361328125, + "learning_rate": 0.0002, + "loss": 1.1494, + "step": 5592 + }, + { + "epoch": 1.7637969094922736, + "grad_norm": 0.400390625, + "learning_rate": 0.0002, + "loss": 1.2318, + "step": 5593 + }, + { + "epoch": 1.7641122674235257, + "grad_norm": 0.384765625, + "learning_rate": 0.0002, + "loss": 1.2078, + "step": 5594 + }, + { + "epoch": 1.7644276253547777, + "grad_norm": 0.431640625, + "learning_rate": 0.0002, + "loss": 1.1867, + "step": 5595 + }, + { + "epoch": 1.7647429832860295, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 1.2838, + "step": 5596 + }, + { + "epoch": 1.7650583412172818, + "grad_norm": 0.451171875, + "learning_rate": 0.0002, + "loss": 1.0917, + "step": 5597 + }, + { + "epoch": 1.7653736991485336, + "grad_norm": 0.421875, + "learning_rate": 0.0002, + "loss": 1.0038, + "step": 5598 + }, + { + "epoch": 1.7656890570797854, + "grad_norm": 0.52734375, + "learning_rate": 0.0002, + "loss": 1.4083, + "step": 5599 + }, + { + "epoch": 1.7660044150110377, + "grad_norm": 0.498046875, + "learning_rate": 0.0002, + "loss": 1.1877, + "step": 5600 + }, + { + "epoch": 1.7663197729422895, + "grad_norm": 0.53515625, + "learning_rate": 0.0002, + "loss": 1.4296, + "step": 5601 + }, + { + "epoch": 1.7666351308735415, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 1.196, + "step": 5602 + }, + { + "epoch": 1.7669504888047936, + "grad_norm": 0.53515625, + "learning_rate": 0.0002, + "loss": 1.4602, + "step": 5603 + }, + { + "epoch": 1.7672658467360454, + "grad_norm": 0.6171875, + "learning_rate": 0.0002, + "loss": 1.2652, + "step": 5604 + }, + { + "epoch": 1.7675812046672974, + "grad_norm": 0.68359375, + "learning_rate": 0.0002, + "loss": 1.4458, + "step": 5605 + }, + { + "epoch": 1.7678965625985494, + "grad_norm": 0.8828125, + "learning_rate": 0.0002, + "loss": 1.8906, + "step": 5606 + }, + { + "epoch": 1.7682119205298013, + "grad_norm": 0.66796875, + "learning_rate": 0.0002, + "loss": 1.6153, + "step": 5607 + }, + { + "epoch": 1.7685272784610533, + "grad_norm": 0.83203125, + "learning_rate": 0.0002, + "loss": 2.0458, + "step": 5608 + }, + { + "epoch": 1.7688426363923053, + "grad_norm": 0.8515625, + "learning_rate": 0.0002, + "loss": 2.0332, + "step": 5609 + }, + { + "epoch": 1.7691579943235571, + "grad_norm": 0.92578125, + "learning_rate": 0.0002, + "loss": 1.8732, + "step": 5610 + }, + { + "epoch": 1.7691579943235571, + "eval_loss": 1.6834352016448975, + "eval_runtime": 149.0598, + "eval_samples_per_second": 6.709, + "eval_steps_per_second": 6.709, + "step": 5610 + }, + { + "epoch": 1.7691579943235571, + "mmlu_eval_accuracy": 0.6001710572530587, + "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, + "mmlu_eval_accuracy_anatomy": 0.5, + "mmlu_eval_accuracy_astronomy": 0.5625, + "mmlu_eval_accuracy_business_ethics": 0.8181818181818182, + "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138, + "mmlu_eval_accuracy_college_biology": 0.6875, + "mmlu_eval_accuracy_college_chemistry": 0.25, + "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, + "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454, + "mmlu_eval_accuracy_college_medicine": 0.7272727272727273, + "mmlu_eval_accuracy_college_physics": 0.45454545454545453, + "mmlu_eval_accuracy_computer_security": 0.6363636363636364, + "mmlu_eval_accuracy_conceptual_physics": 0.5, + "mmlu_eval_accuracy_econometrics": 0.4166666666666667, + "mmlu_eval_accuracy_electrical_engineering": 0.4375, + "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244, + "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, + "mmlu_eval_accuracy_global_facts": 0.2, + "mmlu_eval_accuracy_high_school_biology": 0.59375, + "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8571428571428571, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.627906976744186, + "mmlu_eval_accuracy_high_school_mathematics": 0.3103448275862069, + "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154, + "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354, + "mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, + "mmlu_eval_accuracy_high_school_us_history": 0.8636363636363636, + "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, + "mmlu_eval_accuracy_human_aging": 0.6521739130434783, + "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334, + "mmlu_eval_accuracy_international_law": 1.0, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.2727272727272727, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.76, + "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182, + "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279, + "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, + "mmlu_eval_accuracy_moral_scenarios": 0.29, + "mmlu_eval_accuracy_nutrition": 0.7878787878787878, + "mmlu_eval_accuracy_philosophy": 0.6764705882352942, + "mmlu_eval_accuracy_prehistory": 0.6, + "mmlu_eval_accuracy_professional_accounting": 0.5483870967741935, + "mmlu_eval_accuracy_professional_law": 0.4294117647058823, + "mmlu_eval_accuracy_professional_medicine": 0.7419354838709677, + "mmlu_eval_accuracy_professional_psychology": 0.6666666666666666, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.6666666666666666, + "mmlu_eval_accuracy_sociology": 0.9545454545454546, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.5, + "mmlu_eval_accuracy_world_religions": 0.7368421052631579, + "mmlu_loss": 5.960081269895229, + "step": 5610 + }, + { + "epoch": 1.7694733522548092, + "grad_norm": 0.859375, + "learning_rate": 0.0002, + "loss": 1.7698, + "step": 5611 + }, + { + "epoch": 1.7697887101860612, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 2.22, + "step": 5612 + }, + { + "epoch": 1.770104068117313, + "grad_norm": 0.8359375, + "learning_rate": 0.0002, + "loss": 1.7776, + "step": 5613 + }, + { + "epoch": 1.7704194260485653, + "grad_norm": 0.90625, + "learning_rate": 0.0002, + "loss": 1.9553, + "step": 5614 + }, + { + "epoch": 1.770734783979817, + "grad_norm": 1.359375, + "learning_rate": 0.0002, + "loss": 2.8596, + "step": 5615 + }, + { + "epoch": 1.771050141911069, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 2.3401, + "step": 5616 + }, + { + "epoch": 1.7713654998423212, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 2.4612, + "step": 5617 + }, + { + "epoch": 1.771680857773573, + "grad_norm": 0.9765625, + "learning_rate": 0.0002, + "loss": 2.0343, + "step": 5618 + }, + { + "epoch": 1.771996215704825, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 1.9207, + "step": 5619 + }, + { + "epoch": 1.772311573636077, + "grad_norm": 1.140625, + "learning_rate": 0.0002, + "loss": 2.0854, + "step": 5620 + }, + { + "epoch": 1.7726269315673289, + "grad_norm": 1.3359375, + "learning_rate": 0.0002, + "loss": 2.3614, + "step": 5621 + }, + { + "epoch": 1.772942289498581, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.0219, + "step": 5622 + }, + { + "epoch": 1.773257647429833, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 1.0633, + "step": 5623 + }, + { + "epoch": 1.7735730053610848, + "grad_norm": 0.208984375, + "learning_rate": 0.0002, + "loss": 1.0467, + "step": 5624 + }, + { + "epoch": 1.7738883632923368, + "grad_norm": 0.216796875, + "learning_rate": 0.0002, + "loss": 1.0263, + "step": 5625 + }, + { + "epoch": 1.7742037212235888, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.0192, + "step": 5626 + }, + { + "epoch": 1.7745190791548406, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 1.0189, + "step": 5627 + }, + { + "epoch": 1.7748344370860927, + "grad_norm": 0.255859375, + "learning_rate": 0.0002, + "loss": 1.124, + "step": 5628 + }, + { + "epoch": 1.7751497950173447, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.0282, + "step": 5629 + }, + { + "epoch": 1.7754651529485965, + "grad_norm": 0.23828125, + "learning_rate": 0.0002, + "loss": 1.0273, + "step": 5630 + }, + { + "epoch": 1.7757805108798488, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 0.8976, + "step": 5631 + }, + { + "epoch": 1.7760958688111006, + "grad_norm": 0.23046875, + "learning_rate": 0.0002, + "loss": 1.0241, + "step": 5632 + }, + { + "epoch": 1.7764112267423524, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.1079, + "step": 5633 + }, + { + "epoch": 1.7767265846736047, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 1.0137, + "step": 5634 + }, + { + "epoch": 1.7770419426048565, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 1.0551, + "step": 5635 + }, + { + "epoch": 1.7773573005361085, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 0.9313, + "step": 5636 + }, + { + "epoch": 1.7776726584673606, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 0.9816, + "step": 5637 + }, + { + "epoch": 1.7779880163986124, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.0702, + "step": 5638 + }, + { + "epoch": 1.7783033743298644, + "grad_norm": 0.330078125, + "learning_rate": 0.0002, + "loss": 1.164, + "step": 5639 + }, + { + "epoch": 1.7786187322611164, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 1.1518, + "step": 5640 + }, + { + "epoch": 1.7789340901923683, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 1.0106, + "step": 5641 + }, + { + "epoch": 1.7792494481236203, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 1.1034, + "step": 5642 + }, + { + "epoch": 1.7795648060548723, + "grad_norm": 0.349609375, + "learning_rate": 0.0002, + "loss": 1.2069, + "step": 5643 + }, + { + "epoch": 1.7798801639861241, + "grad_norm": 0.380859375, + "learning_rate": 0.0002, + "loss": 1.0819, + "step": 5644 + }, + { + "epoch": 1.7801955219173762, + "grad_norm": 0.380859375, + "learning_rate": 0.0002, + "loss": 1.2627, + "step": 5645 + }, + { + "epoch": 1.7805108798486282, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 1.0845, + "step": 5646 + }, + { + "epoch": 1.78082623777988, + "grad_norm": 0.427734375, + "learning_rate": 0.0002, + "loss": 1.2181, + "step": 5647 + }, + { + "epoch": 1.7811415957111323, + "grad_norm": 0.392578125, + "learning_rate": 0.0002, + "loss": 1.1614, + "step": 5648 + }, + { + "epoch": 1.781456953642384, + "grad_norm": 0.42578125, + "learning_rate": 0.0002, + "loss": 1.2024, + "step": 5649 + }, + { + "epoch": 1.781772311573636, + "grad_norm": 0.494140625, + "learning_rate": 0.0002, + "loss": 1.1403, + "step": 5650 + }, + { + "epoch": 1.7820876695048882, + "grad_norm": 0.46875, + "learning_rate": 0.0002, + "loss": 1.2027, + "step": 5651 + }, + { + "epoch": 1.78240302743614, + "grad_norm": 0.484375, + "learning_rate": 0.0002, + "loss": 1.1717, + "step": 5652 + }, + { + "epoch": 1.782718385367392, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 1.2959, + "step": 5653 + }, + { + "epoch": 1.783033743298644, + "grad_norm": 0.69921875, + "learning_rate": 0.0002, + "loss": 1.5141, + "step": 5654 + }, + { + "epoch": 1.7833491012298959, + "grad_norm": 0.6640625, + "learning_rate": 0.0002, + "loss": 1.4081, + "step": 5655 + }, + { + "epoch": 1.783664459161148, + "grad_norm": 0.7265625, + "learning_rate": 0.0002, + "loss": 1.7156, + "step": 5656 + }, + { + "epoch": 1.7839798170924, + "grad_norm": 0.79296875, + "learning_rate": 0.0002, + "loss": 1.7264, + "step": 5657 + }, + { + "epoch": 1.7842951750236518, + "grad_norm": 1.1875, + "learning_rate": 0.0002, + "loss": 1.8243, + "step": 5658 + }, + { + "epoch": 1.7846105329549038, + "grad_norm": 0.90234375, + "learning_rate": 0.0002, + "loss": 1.7706, + "step": 5659 + }, + { + "epoch": 1.7849258908861558, + "grad_norm": 0.765625, + "learning_rate": 0.0002, + "loss": 1.6783, + "step": 5660 + }, + { + "epoch": 1.7852412488174076, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 1.9225, + "step": 5661 + }, + { + "epoch": 1.7855566067486597, + "grad_norm": 0.94140625, + "learning_rate": 0.0002, + "loss": 1.9551, + "step": 5662 + }, + { + "epoch": 1.7858719646799117, + "grad_norm": 0.87109375, + "learning_rate": 0.0002, + "loss": 1.9248, + "step": 5663 + }, + { + "epoch": 1.7861873226111635, + "grad_norm": 1.109375, + "learning_rate": 0.0002, + "loss": 2.0365, + "step": 5664 + }, + { + "epoch": 1.7865026805424158, + "grad_norm": 0.83203125, + "learning_rate": 0.0002, + "loss": 1.914, + "step": 5665 + }, + { + "epoch": 1.7868180384736676, + "grad_norm": 0.92578125, + "learning_rate": 0.0002, + "loss": 1.8304, + "step": 5666 + }, + { + "epoch": 1.7871333964049196, + "grad_norm": 0.796875, + "learning_rate": 0.0002, + "loss": 1.8331, + "step": 5667 + }, + { + "epoch": 1.7874487543361717, + "grad_norm": 1.2421875, + "learning_rate": 0.0002, + "loss": 2.2471, + "step": 5668 + }, + { + "epoch": 1.7877641122674235, + "grad_norm": 1.171875, + "learning_rate": 0.0002, + "loss": 1.9496, + "step": 5669 + }, + { + "epoch": 1.7880794701986755, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 2.1075, + "step": 5670 + }, + { + "epoch": 1.7883948281299276, + "grad_norm": 1.234375, + "learning_rate": 0.0002, + "loss": 2.0441, + "step": 5671 + }, + { + "epoch": 1.7887101860611794, + "grad_norm": 0.2001953125, + "learning_rate": 0.0002, + "loss": 1.1063, + "step": 5672 + }, + { + "epoch": 1.7890255439924314, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.0005, + "step": 5673 + }, + { + "epoch": 1.7893409019236834, + "grad_norm": 0.2119140625, + "learning_rate": 0.0002, + "loss": 0.9687, + "step": 5674 + }, + { + "epoch": 1.7896562598549353, + "grad_norm": 0.220703125, + "learning_rate": 0.0002, + "loss": 0.9653, + "step": 5675 + }, + { + "epoch": 1.7899716177861873, + "grad_norm": 0.22265625, + "learning_rate": 0.0002, + "loss": 1.05, + "step": 5676 + }, + { + "epoch": 1.7902869757174393, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.0452, + "step": 5677 + }, + { + "epoch": 1.7906023336486911, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.0517, + "step": 5678 + }, + { + "epoch": 1.7909176915799432, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 1.0342, + "step": 5679 + }, + { + "epoch": 1.7912330495111952, + "grad_norm": 0.2265625, + "learning_rate": 0.0002, + "loss": 1.0422, + "step": 5680 + }, + { + "epoch": 1.791548407442447, + "grad_norm": 0.2294921875, + "learning_rate": 0.0002, + "loss": 1.1617, + "step": 5681 + }, + { + "epoch": 1.7918637653736993, + "grad_norm": 0.2421875, + "learning_rate": 0.0002, + "loss": 1.0553, + "step": 5682 + }, + { + "epoch": 1.792179123304951, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 1.0662, + "step": 5683 + }, + { + "epoch": 1.7924944812362031, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 1.1183, + "step": 5684 + }, + { + "epoch": 1.7928098391674552, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 1.1265, + "step": 5685 + }, + { + "epoch": 1.793125197098707, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.1764, + "step": 5686 + }, + { + "epoch": 1.793440555029959, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.0829, + "step": 5687 + }, + { + "epoch": 1.793755912961211, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.0498, + "step": 5688 + }, + { + "epoch": 1.7940712708924629, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 0.9749, + "step": 5689 + }, + { + "epoch": 1.794386628823715, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 1.1141, + "step": 5690 + }, + { + "epoch": 1.794701986754967, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 1.2816, + "step": 5691 + }, + { + "epoch": 1.7950173446862188, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 1.0424, + "step": 5692 + }, + { + "epoch": 1.7953327026174708, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 1.1429, + "step": 5693 + }, + { + "epoch": 1.7956480605487228, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 1.2252, + "step": 5694 + }, + { + "epoch": 1.7959634184799746, + "grad_norm": 0.38671875, + "learning_rate": 0.0002, + "loss": 1.1871, + "step": 5695 + }, + { + "epoch": 1.796278776411227, + "grad_norm": 0.37890625, + "learning_rate": 0.0002, + "loss": 1.0143, + "step": 5696 + }, + { + "epoch": 1.7965941343424787, + "grad_norm": 0.412109375, + "learning_rate": 0.0002, + "loss": 1.1906, + "step": 5697 + }, + { + "epoch": 1.7969094922737305, + "grad_norm": 0.46484375, + "learning_rate": 0.0002, + "loss": 1.2127, + "step": 5698 + }, + { + "epoch": 1.7972248502049828, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 1.2883, + "step": 5699 + }, + { + "epoch": 1.7975402081362346, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 1.4388, + "step": 5700 + }, + { + "epoch": 1.7978555660674866, + "grad_norm": 0.546875, + "learning_rate": 0.0002, + "loss": 1.3642, + "step": 5701 + }, + { + "epoch": 1.7981709239987387, + "grad_norm": 0.6171875, + "learning_rate": 0.0002, + "loss": 1.5699, + "step": 5702 + }, + { + "epoch": 1.7984862819299905, + "grad_norm": 0.58203125, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 5703 + }, + { + "epoch": 1.7988016398612425, + "grad_norm": 0.640625, + "learning_rate": 0.0002, + "loss": 1.5302, + "step": 5704 + }, + { + "epoch": 1.7991169977924946, + "grad_norm": 0.73046875, + "learning_rate": 0.0002, + "loss": 1.6818, + "step": 5705 + }, + { + "epoch": 1.7994323557237464, + "grad_norm": 0.73046875, + "learning_rate": 0.0002, + "loss": 1.759, + "step": 5706 + }, + { + "epoch": 1.7997477136549984, + "grad_norm": 0.703125, + "learning_rate": 0.0002, + "loss": 1.6969, + "step": 5707 + }, + { + "epoch": 1.8000630715862505, + "grad_norm": 0.92578125, + "learning_rate": 0.0002, + "loss": 1.8836, + "step": 5708 + }, + { + "epoch": 1.8003784295175023, + "grad_norm": 0.7734375, + "learning_rate": 0.0002, + "loss": 1.7577, + "step": 5709 + }, + { + "epoch": 1.8006937874487543, + "grad_norm": 0.87890625, + "learning_rate": 0.0002, + "loss": 1.9451, + "step": 5710 + }, + { + "epoch": 1.8010091453800063, + "grad_norm": 0.8203125, + "learning_rate": 0.0002, + "loss": 1.7634, + "step": 5711 + }, + { + "epoch": 1.8013245033112582, + "grad_norm": 0.77734375, + "learning_rate": 0.0002, + "loss": 1.6489, + "step": 5712 + }, + { + "epoch": 1.8016398612425104, + "grad_norm": 0.98828125, + "learning_rate": 0.0002, + "loss": 2.0926, + "step": 5713 + }, + { + "epoch": 1.8019552191737622, + "grad_norm": 1.1328125, + "learning_rate": 0.0002, + "loss": 2.5231, + "step": 5714 + }, + { + "epoch": 1.802270577105014, + "grad_norm": 1.171875, + "learning_rate": 0.0002, + "loss": 2.2136, + "step": 5715 + }, + { + "epoch": 1.8025859350362663, + "grad_norm": 0.94140625, + "learning_rate": 0.0002, + "loss": 2.3781, + "step": 5716 + }, + { + "epoch": 1.802901292967518, + "grad_norm": 0.81640625, + "learning_rate": 0.0002, + "loss": 1.6103, + "step": 5717 + }, + { + "epoch": 1.8032166508987701, + "grad_norm": 0.84765625, + "learning_rate": 0.0002, + "loss": 1.7832, + "step": 5718 + }, + { + "epoch": 1.8035320088300222, + "grad_norm": 1.140625, + "learning_rate": 0.0002, + "loss": 2.309, + "step": 5719 + }, + { + "epoch": 1.803847366761274, + "grad_norm": 0.921875, + "learning_rate": 0.0002, + "loss": 1.8368, + "step": 5720 + }, + { + "epoch": 1.804162724692526, + "grad_norm": 1.90625, + "learning_rate": 0.0002, + "loss": 2.3336, + "step": 5721 + }, + { + "epoch": 1.804478082623778, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.0174, + "step": 5722 + }, + { + "epoch": 1.8047934405550299, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 1.0554, + "step": 5723 + }, + { + "epoch": 1.805108798486282, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 1.0176, + "step": 5724 + }, + { + "epoch": 1.805424156417534, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 0.9723, + "step": 5725 + }, + { + "epoch": 1.8057395143487858, + "grad_norm": 0.2265625, + "learning_rate": 0.0002, + "loss": 1.1282, + "step": 5726 + }, + { + "epoch": 1.8060548722800378, + "grad_norm": 0.2216796875, + "learning_rate": 0.0002, + "loss": 1.0779, + "step": 5727 + }, + { + "epoch": 1.8063702302112898, + "grad_norm": 0.2294921875, + "learning_rate": 0.0002, + "loss": 1.027, + "step": 5728 + }, + { + "epoch": 1.8066855881425417, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 0.9959, + "step": 5729 + }, + { + "epoch": 1.807000946073794, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.0584, + "step": 5730 + }, + { + "epoch": 1.8073163040050457, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 1.0459, + "step": 5731 + }, + { + "epoch": 1.8076316619362975, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 1.0803, + "step": 5732 + }, + { + "epoch": 1.8079470198675498, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.2551, + "step": 5733 + }, + { + "epoch": 1.8082623777988016, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.1524, + "step": 5734 + }, + { + "epoch": 1.8085777357300536, + "grad_norm": 0.255859375, + "learning_rate": 0.0002, + "loss": 1.0335, + "step": 5735 + }, + { + "epoch": 1.8088930936613057, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 1.0355, + "step": 5736 + }, + { + "epoch": 1.8092084515925575, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.0685, + "step": 5737 + }, + { + "epoch": 1.8095238095238095, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.0867, + "step": 5738 + }, + { + "epoch": 1.8098391674550616, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 1.1604, + "step": 5739 + }, + { + "epoch": 1.8101545253863134, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 1.1689, + "step": 5740 + }, + { + "epoch": 1.8104698833175654, + "grad_norm": 0.421875, + "learning_rate": 0.0002, + "loss": 1.1594, + "step": 5741 + }, + { + "epoch": 1.8107852412488175, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 1.046, + "step": 5742 + }, + { + "epoch": 1.8111005991800693, + "grad_norm": 0.349609375, + "learning_rate": 0.0002, + "loss": 1.2321, + "step": 5743 + }, + { + "epoch": 1.8114159571113213, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 1.2708, + "step": 5744 + }, + { + "epoch": 1.8117313150425733, + "grad_norm": 0.43359375, + "learning_rate": 0.0002, + "loss": 1.2876, + "step": 5745 + }, + { + "epoch": 1.8120466729738252, + "grad_norm": 0.384765625, + "learning_rate": 0.0002, + "loss": 1.1662, + "step": 5746 + }, + { + "epoch": 1.8123620309050774, + "grad_norm": 0.365234375, + "learning_rate": 0.0002, + "loss": 1.1197, + "step": 5747 + }, + { + "epoch": 1.8126773888363292, + "grad_norm": 0.435546875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 5748 + }, + { + "epoch": 1.812992746767581, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 1.2371, + "step": 5749 + }, + { + "epoch": 1.8133081046988333, + "grad_norm": 0.50390625, + "learning_rate": 0.0002, + "loss": 1.2968, + "step": 5750 + }, + { + "epoch": 1.8136234626300851, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 1.4818, + "step": 5751 + }, + { + "epoch": 1.8139388205613372, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 5752 + }, + { + "epoch": 1.8142541784925892, + "grad_norm": 0.60546875, + "learning_rate": 0.0002, + "loss": 1.4775, + "step": 5753 + }, + { + "epoch": 1.814569536423841, + "grad_norm": 0.62109375, + "learning_rate": 0.0002, + "loss": 1.3595, + "step": 5754 + }, + { + "epoch": 1.814884894355093, + "grad_norm": 0.6796875, + "learning_rate": 0.0002, + "loss": 1.4131, + "step": 5755 + }, + { + "epoch": 1.815200252286345, + "grad_norm": 0.71484375, + "learning_rate": 0.0002, + "loss": 1.526, + "step": 5756 + }, + { + "epoch": 1.8155156102175969, + "grad_norm": 0.81640625, + "learning_rate": 0.0002, + "loss": 1.8825, + "step": 5757 + }, + { + "epoch": 1.815830968148849, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 1.8061, + "step": 5758 + }, + { + "epoch": 1.816146326080101, + "grad_norm": 0.85546875, + "learning_rate": 0.0002, + "loss": 1.7599, + "step": 5759 + }, + { + "epoch": 1.8164616840113528, + "grad_norm": 0.9140625, + "learning_rate": 0.0002, + "loss": 1.9384, + "step": 5760 + }, + { + "epoch": 1.8167770419426048, + "grad_norm": 0.8359375, + "learning_rate": 0.0002, + "loss": 1.7476, + "step": 5761 + }, + { + "epoch": 1.8170923998738568, + "grad_norm": 0.89453125, + "learning_rate": 0.0002, + "loss": 1.7291, + "step": 5762 + }, + { + "epoch": 1.8174077578051087, + "grad_norm": 0.87890625, + "learning_rate": 0.0002, + "loss": 1.752, + "step": 5763 + }, + { + "epoch": 1.817723115736361, + "grad_norm": 0.99609375, + "learning_rate": 0.0002, + "loss": 2.1727, + "step": 5764 + }, + { + "epoch": 1.8180384736676127, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 2.0937, + "step": 5765 + }, + { + "epoch": 1.8183538315988645, + "grad_norm": 1.015625, + "learning_rate": 0.0002, + "loss": 2.0368, + "step": 5766 + }, + { + "epoch": 1.8186691895301168, + "grad_norm": 0.94140625, + "learning_rate": 0.0002, + "loss": 1.6224, + "step": 5767 + }, + { + "epoch": 1.8189845474613686, + "grad_norm": 1.21875, + "learning_rate": 0.0002, + "loss": 2.2855, + "step": 5768 + }, + { + "epoch": 1.8192999053926207, + "grad_norm": 1.328125, + "learning_rate": 0.0002, + "loss": 2.3216, + "step": 5769 + }, + { + "epoch": 1.8196152633238727, + "grad_norm": 1.140625, + "learning_rate": 0.0002, + "loss": 2.0868, + "step": 5770 + }, + { + "epoch": 1.8199306212551245, + "grad_norm": 1.2109375, + "learning_rate": 0.0002, + "loss": 2.083, + "step": 5771 + }, + { + "epoch": 1.8202459791863765, + "grad_norm": 0.197265625, + "learning_rate": 0.0002, + "loss": 1.0369, + "step": 5772 + }, + { + "epoch": 1.8205613371176286, + "grad_norm": 0.1875, + "learning_rate": 0.0002, + "loss": 1.035, + "step": 5773 + }, + { + "epoch": 1.8208766950488804, + "grad_norm": 0.2119140625, + "learning_rate": 0.0002, + "loss": 1.0182, + "step": 5774 + }, + { + "epoch": 1.8211920529801324, + "grad_norm": 0.240234375, + "learning_rate": 0.0002, + "loss": 1.079, + "step": 5775 + }, + { + "epoch": 1.8215074109113845, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 1.1023, + "step": 5776 + }, + { + "epoch": 1.8218227688426363, + "grad_norm": 0.2373046875, + "learning_rate": 0.0002, + "loss": 0.9817, + "step": 5777 + }, + { + "epoch": 1.8221381267738883, + "grad_norm": 0.248046875, + "learning_rate": 0.0002, + "loss": 1.1328, + "step": 5778 + }, + { + "epoch": 1.8224534847051403, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.0421, + "step": 5779 + }, + { + "epoch": 1.8227688426363922, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.1028, + "step": 5780 + }, + { + "epoch": 1.8230842005676444, + "grad_norm": 0.248046875, + "learning_rate": 0.0002, + "loss": 1.1035, + "step": 5781 + }, + { + "epoch": 1.8233995584988962, + "grad_norm": 0.2412109375, + "learning_rate": 0.0002, + "loss": 1.0398, + "step": 5782 + }, + { + "epoch": 1.8237149164301483, + "grad_norm": 0.2431640625, + "learning_rate": 0.0002, + "loss": 1.1399, + "step": 5783 + }, + { + "epoch": 1.8240302743614003, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.0898, + "step": 5784 + }, + { + "epoch": 1.8243456322926521, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 1.0298, + "step": 5785 + }, + { + "epoch": 1.8246609902239042, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 1.0865, + "step": 5786 + }, + { + "epoch": 1.8249763481551562, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.996, + "step": 5787 + }, + { + "epoch": 1.825291706086408, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.1272, + "step": 5788 + }, + { + "epoch": 1.82560706401766, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 1.0164, + "step": 5789 + }, + { + "epoch": 1.825922421948912, + "grad_norm": 0.373046875, + "learning_rate": 0.0002, + "loss": 1.2347, + "step": 5790 + }, + { + "epoch": 1.826237779880164, + "grad_norm": 0.427734375, + "learning_rate": 0.0002, + "loss": 1.1762, + "step": 5791 + }, + { + "epoch": 1.826553137811416, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 1.0697, + "step": 5792 + }, + { + "epoch": 1.826868495742668, + "grad_norm": 0.369140625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 5793 + }, + { + "epoch": 1.8271838536739198, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 1.0812, + "step": 5794 + }, + { + "epoch": 1.827499211605172, + "grad_norm": 0.390625, + "learning_rate": 0.0002, + "loss": 1.0938, + "step": 5795 + }, + { + "epoch": 1.8278145695364238, + "grad_norm": 0.38671875, + "learning_rate": 0.0002, + "loss": 1.2296, + "step": 5796 + }, + { + "epoch": 1.8281299274676757, + "grad_norm": 0.37109375, + "learning_rate": 0.0002, + "loss": 1.0398, + "step": 5797 + }, + { + "epoch": 1.8281299274676757, + "eval_loss": 1.668998122215271, + "eval_runtime": 149.5383, + "eval_samples_per_second": 6.687, + "eval_steps_per_second": 6.687, + "step": 5797 + }, + { + "epoch": 1.8281299274676757, + "mmlu_eval_accuracy": 0.60451209043097, + "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, + "mmlu_eval_accuracy_anatomy": 0.42857142857142855, + "mmlu_eval_accuracy_astronomy": 0.6875, + "mmlu_eval_accuracy_business_ethics": 0.7272727272727273, + "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034, + "mmlu_eval_accuracy_college_biology": 0.75, + "mmlu_eval_accuracy_college_chemistry": 0.375, + "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, + "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453, + "mmlu_eval_accuracy_college_medicine": 0.5909090909090909, + "mmlu_eval_accuracy_college_physics": 0.6363636363636364, + "mmlu_eval_accuracy_computer_security": 0.5454545454545454, + "mmlu_eval_accuracy_conceptual_physics": 0.5, + "mmlu_eval_accuracy_econometrics": 0.5, + "mmlu_eval_accuracy_electrical_engineering": 0.4375, + "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244, + "mmlu_eval_accuracy_formal_logic": 0.2857142857142857, + "mmlu_eval_accuracy_global_facts": 0.2, + "mmlu_eval_accuracy_high_school_biology": 0.65625, + "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.9545454545454546, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8571428571428571, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.6046511627906976, + "mmlu_eval_accuracy_high_school_mathematics": 0.3103448275862069, + "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384, + "mmlu_eval_accuracy_high_school_physics": 0.35294117647058826, + "mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, + "mmlu_eval_accuracy_high_school_us_history": 0.8636363636363636, + "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, + "mmlu_eval_accuracy_human_aging": 0.6956521739130435, + "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334, + "mmlu_eval_accuracy_international_law": 1.0, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.2727272727272727, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.76, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7325581395348837, + "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842, + "mmlu_eval_accuracy_moral_scenarios": 0.31, + "mmlu_eval_accuracy_nutrition": 0.7575757575757576, + "mmlu_eval_accuracy_philosophy": 0.7058823529411765, + "mmlu_eval_accuracy_prehistory": 0.6, + "mmlu_eval_accuracy_professional_accounting": 0.3870967741935484, + "mmlu_eval_accuracy_professional_law": 0.4, + "mmlu_eval_accuracy_professional_medicine": 0.7419354838709677, + "mmlu_eval_accuracy_professional_psychology": 0.6811594202898551, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.6666666666666666, + "mmlu_eval_accuracy_sociology": 0.9545454545454546, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.5, + "mmlu_eval_accuracy_world_religions": 0.7894736842105263, + "mmlu_loss": 5.791989168725241, + "step": 5797 + }, + { + "epoch": 1.828445285398928, + "grad_norm": 0.421875, + "learning_rate": 0.0002, + "loss": 1.2152, + "step": 5798 + }, + { + "epoch": 1.8287606433301797, + "grad_norm": 0.453125, + "learning_rate": 0.0002, + "loss": 1.2706, + "step": 5799 + }, + { + "epoch": 1.8290760012614318, + "grad_norm": 0.4375, + "learning_rate": 0.0002, + "loss": 1.1407, + "step": 5800 + }, + { + "epoch": 1.8293913591926838, + "grad_norm": 0.5, + "learning_rate": 0.0002, + "loss": 1.13, + "step": 5801 + }, + { + "epoch": 1.8297067171239356, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 1.5162, + "step": 5802 + }, + { + "epoch": 1.8300220750551877, + "grad_norm": 0.73828125, + "learning_rate": 0.0002, + "loss": 1.5398, + "step": 5803 + }, + { + "epoch": 1.8303374329864397, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 1.4472, + "step": 5804 + }, + { + "epoch": 1.8306527909176915, + "grad_norm": 0.6640625, + "learning_rate": 0.0002, + "loss": 1.5029, + "step": 5805 + }, + { + "epoch": 1.8309681488489435, + "grad_norm": 0.8046875, + "learning_rate": 0.0002, + "loss": 1.7467, + "step": 5806 + }, + { + "epoch": 1.8312835067801956, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 1.5094, + "step": 5807 + }, + { + "epoch": 1.8315988647114474, + "grad_norm": 0.85546875, + "learning_rate": 0.0002, + "loss": 1.7707, + "step": 5808 + }, + { + "epoch": 1.8319142226426994, + "grad_norm": 0.69921875, + "learning_rate": 0.0002, + "loss": 1.5808, + "step": 5809 + }, + { + "epoch": 1.8322295805739515, + "grad_norm": 0.9140625, + "learning_rate": 0.0002, + "loss": 2.0162, + "step": 5810 + }, + { + "epoch": 1.8325449385052033, + "grad_norm": 0.96875, + "learning_rate": 0.0002, + "loss": 2.0748, + "step": 5811 + }, + { + "epoch": 1.8328602964364555, + "grad_norm": 0.87890625, + "learning_rate": 0.0002, + "loss": 1.7516, + "step": 5812 + }, + { + "epoch": 1.8331756543677074, + "grad_norm": 0.8984375, + "learning_rate": 0.0002, + "loss": 1.8669, + "step": 5813 + }, + { + "epoch": 1.8334910122989592, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 2.0286, + "step": 5814 + }, + { + "epoch": 1.8338063702302114, + "grad_norm": 0.9609375, + "learning_rate": 0.0002, + "loss": 2.3641, + "step": 5815 + }, + { + "epoch": 1.8341217281614632, + "grad_norm": 0.84375, + "learning_rate": 0.0002, + "loss": 1.9906, + "step": 5816 + }, + { + "epoch": 1.8344370860927153, + "grad_norm": 1.1328125, + "learning_rate": 0.0002, + "loss": 2.2269, + "step": 5817 + }, + { + "epoch": 1.8347524440239673, + "grad_norm": 0.92578125, + "learning_rate": 0.0002, + "loss": 1.7569, + "step": 5818 + }, + { + "epoch": 1.8350678019552191, + "grad_norm": 1.03125, + "learning_rate": 0.0002, + "loss": 2.046, + "step": 5819 + }, + { + "epoch": 1.8353831598864712, + "grad_norm": 1.0546875, + "learning_rate": 0.0002, + "loss": 1.9073, + "step": 5820 + }, + { + "epoch": 1.8356985178177232, + "grad_norm": 1.4296875, + "learning_rate": 0.0002, + "loss": 2.1473, + "step": 5821 + }, + { + "epoch": 1.836013875748975, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.187, + "step": 5822 + }, + { + "epoch": 1.836329233680227, + "grad_norm": 0.2119140625, + "learning_rate": 0.0002, + "loss": 1.034, + "step": 5823 + }, + { + "epoch": 1.836644591611479, + "grad_norm": 0.2333984375, + "learning_rate": 0.0002, + "loss": 0.9849, + "step": 5824 + }, + { + "epoch": 1.836959949542731, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.0474, + "step": 5825 + }, + { + "epoch": 1.837275307473983, + "grad_norm": 0.232421875, + "learning_rate": 0.0002, + "loss": 1.0463, + "step": 5826 + }, + { + "epoch": 1.837590665405235, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 1.1128, + "step": 5827 + }, + { + "epoch": 1.8379060233364868, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 1.0774, + "step": 5828 + }, + { + "epoch": 1.838221381267739, + "grad_norm": 0.2275390625, + "learning_rate": 0.0002, + "loss": 1.0602, + "step": 5829 + }, + { + "epoch": 1.8385367391989909, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.0752, + "step": 5830 + }, + { + "epoch": 1.8388520971302427, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 0.9737, + "step": 5831 + }, + { + "epoch": 1.839167455061495, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.0868, + "step": 5832 + }, + { + "epoch": 1.8394828129927467, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.0702, + "step": 5833 + }, + { + "epoch": 1.8397981709239988, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 1.0216, + "step": 5834 + }, + { + "epoch": 1.8401135288552508, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 1.2081, + "step": 5835 + }, + { + "epoch": 1.8404288867865026, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 1.1398, + "step": 5836 + }, + { + "epoch": 1.8407442447177547, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.0185, + "step": 5837 + }, + { + "epoch": 1.8410596026490067, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.2134, + "step": 5838 + }, + { + "epoch": 1.8413749605802585, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 1.2281, + "step": 5839 + }, + { + "epoch": 1.8416903185115105, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 1.1232, + "step": 5840 + }, + { + "epoch": 1.8420056764427626, + "grad_norm": 0.3671875, + "learning_rate": 0.0002, + "loss": 1.2084, + "step": 5841 + }, + { + "epoch": 1.8423210343740144, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 1.065, + "step": 5842 + }, + { + "epoch": 1.8426363923052664, + "grad_norm": 0.435546875, + "learning_rate": 0.0002, + "loss": 1.2833, + "step": 5843 + }, + { + "epoch": 1.8429517502365185, + "grad_norm": 0.3671875, + "learning_rate": 0.0002, + "loss": 1.1036, + "step": 5844 + }, + { + "epoch": 1.8432671081677703, + "grad_norm": 0.50390625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 5845 + }, + { + "epoch": 1.8435824660990225, + "grad_norm": 0.388671875, + "learning_rate": 0.0002, + "loss": 1.0062, + "step": 5846 + }, + { + "epoch": 1.8438978240302744, + "grad_norm": 0.388671875, + "learning_rate": 0.0002, + "loss": 1.1069, + "step": 5847 + }, + { + "epoch": 1.8442131819615262, + "grad_norm": 0.447265625, + "learning_rate": 0.0002, + "loss": 1.0332, + "step": 5848 + }, + { + "epoch": 1.8445285398927784, + "grad_norm": 0.498046875, + "learning_rate": 0.0002, + "loss": 1.5653, + "step": 5849 + }, + { + "epoch": 1.8448438978240302, + "grad_norm": 0.55859375, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 5850 + }, + { + "epoch": 1.8451592557552823, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 5851 + }, + { + "epoch": 1.8454746136865343, + "grad_norm": 0.5, + "learning_rate": 0.0002, + "loss": 1.2837, + "step": 5852 + }, + { + "epoch": 1.8457899716177861, + "grad_norm": 0.4609375, + "learning_rate": 0.0002, + "loss": 1.2614, + "step": 5853 + }, + { + "epoch": 1.8461053295490382, + "grad_norm": 0.63671875, + "learning_rate": 0.0002, + "loss": 1.3885, + "step": 5854 + }, + { + "epoch": 1.8464206874802902, + "grad_norm": 0.62109375, + "learning_rate": 0.0002, + "loss": 1.4146, + "step": 5855 + }, + { + "epoch": 1.846736045411542, + "grad_norm": 0.578125, + "learning_rate": 0.0002, + "loss": 1.2422, + "step": 5856 + }, + { + "epoch": 1.847051403342794, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 1.7041, + "step": 5857 + }, + { + "epoch": 1.847366761274046, + "grad_norm": 0.703125, + "learning_rate": 0.0002, + "loss": 1.6421, + "step": 5858 + }, + { + "epoch": 1.847682119205298, + "grad_norm": 0.984375, + "learning_rate": 0.0002, + "loss": 2.0207, + "step": 5859 + }, + { + "epoch": 1.84799747713655, + "grad_norm": 0.921875, + "learning_rate": 0.0002, + "loss": 1.7895, + "step": 5860 + }, + { + "epoch": 1.848312835067802, + "grad_norm": 0.96484375, + "learning_rate": 0.0002, + "loss": 2.2087, + "step": 5861 + }, + { + "epoch": 1.8486281929990538, + "grad_norm": 1.15625, + "learning_rate": 0.0002, + "loss": 2.3168, + "step": 5862 + }, + { + "epoch": 1.848943550930306, + "grad_norm": 0.94140625, + "learning_rate": 0.0002, + "loss": 2.0095, + "step": 5863 + }, + { + "epoch": 1.8492589088615579, + "grad_norm": 0.921875, + "learning_rate": 0.0002, + "loss": 1.9542, + "step": 5864 + }, + { + "epoch": 1.8495742667928097, + "grad_norm": 1.1328125, + "learning_rate": 0.0002, + "loss": 1.7065, + "step": 5865 + }, + { + "epoch": 1.849889624724062, + "grad_norm": 0.9140625, + "learning_rate": 0.0002, + "loss": 2.0667, + "step": 5866 + }, + { + "epoch": 1.8502049826553137, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 1.5504, + "step": 5867 + }, + { + "epoch": 1.8505203405865658, + "grad_norm": 1.015625, + "learning_rate": 0.0002, + "loss": 1.6372, + "step": 5868 + }, + { + "epoch": 1.8508356985178178, + "grad_norm": 1.15625, + "learning_rate": 0.0002, + "loss": 1.9548, + "step": 5869 + }, + { + "epoch": 1.8511510564490696, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 2.0861, + "step": 5870 + }, + { + "epoch": 1.8514664143803217, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 1.7755, + "step": 5871 + }, + { + "epoch": 1.8517817723115737, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.0626, + "step": 5872 + }, + { + "epoch": 1.8520971302428255, + "grad_norm": 0.2373046875, + "learning_rate": 0.0002, + "loss": 1.0394, + "step": 5873 + }, + { + "epoch": 1.8524124881740776, + "grad_norm": 0.212890625, + "learning_rate": 0.0002, + "loss": 0.9532, + "step": 5874 + }, + { + "epoch": 1.8527278461053296, + "grad_norm": 0.2119140625, + "learning_rate": 0.0002, + "loss": 1.0439, + "step": 5875 + }, + { + "epoch": 1.8530432040365814, + "grad_norm": 0.2421875, + "learning_rate": 0.0002, + "loss": 1.0695, + "step": 5876 + }, + { + "epoch": 1.8533585619678334, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 1.0704, + "step": 5877 + }, + { + "epoch": 1.8536739198990855, + "grad_norm": 0.23046875, + "learning_rate": 0.0002, + "loss": 1.1664, + "step": 5878 + }, + { + "epoch": 1.8539892778303373, + "grad_norm": 0.224609375, + "learning_rate": 0.0002, + "loss": 0.8932, + "step": 5879 + }, + { + "epoch": 1.8543046357615895, + "grad_norm": 0.248046875, + "learning_rate": 0.0002, + "loss": 1.1528, + "step": 5880 + }, + { + "epoch": 1.8546199936928414, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 0.9109, + "step": 5881 + }, + { + "epoch": 1.8549353516240934, + "grad_norm": 0.240234375, + "learning_rate": 0.0002, + "loss": 1.0283, + "step": 5882 + }, + { + "epoch": 1.8552507095553454, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 0.9848, + "step": 5883 + }, + { + "epoch": 1.8555660674865972, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.0496, + "step": 5884 + }, + { + "epoch": 1.8558814254178493, + "grad_norm": 0.255859375, + "learning_rate": 0.0002, + "loss": 0.9831, + "step": 5885 + }, + { + "epoch": 1.8561967833491013, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 1.2595, + "step": 5886 + }, + { + "epoch": 1.8565121412803531, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 1.0294, + "step": 5887 + }, + { + "epoch": 1.8568274992116052, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.0467, + "step": 5888 + }, + { + "epoch": 1.8571428571428572, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 0.9901, + "step": 5889 + }, + { + "epoch": 1.857458215074109, + "grad_norm": 0.34765625, + "learning_rate": 0.0002, + "loss": 1.1267, + "step": 5890 + }, + { + "epoch": 1.857773573005361, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 1.166, + "step": 5891 + }, + { + "epoch": 1.858088930936613, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 1.0887, + "step": 5892 + }, + { + "epoch": 1.858404288867865, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 1.1977, + "step": 5893 + }, + { + "epoch": 1.858719646799117, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 1.1831, + "step": 5894 + }, + { + "epoch": 1.859035004730369, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 1.1131, + "step": 5895 + }, + { + "epoch": 1.8593503626616208, + "grad_norm": 0.400390625, + "learning_rate": 0.0002, + "loss": 1.2741, + "step": 5896 + }, + { + "epoch": 1.859665720592873, + "grad_norm": 0.419921875, + "learning_rate": 0.0002, + "loss": 1.1033, + "step": 5897 + }, + { + "epoch": 1.8599810785241249, + "grad_norm": 0.423828125, + "learning_rate": 0.0002, + "loss": 1.0828, + "step": 5898 + }, + { + "epoch": 1.860296436455377, + "grad_norm": 0.470703125, + "learning_rate": 0.0002, + "loss": 1.2706, + "step": 5899 + }, + { + "epoch": 1.860611794386629, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 5900 + }, + { + "epoch": 1.8609271523178808, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 1.4556, + "step": 5901 + }, + { + "epoch": 1.8612425102491328, + "grad_norm": 0.6953125, + "learning_rate": 0.0002, + "loss": 1.5404, + "step": 5902 + }, + { + "epoch": 1.8615578681803848, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 1.523, + "step": 5903 + }, + { + "epoch": 1.8618732261116366, + "grad_norm": 0.65625, + "learning_rate": 0.0002, + "loss": 1.5736, + "step": 5904 + }, + { + "epoch": 1.8621885840428887, + "grad_norm": 0.828125, + "learning_rate": 0.0002, + "loss": 1.6405, + "step": 5905 + }, + { + "epoch": 1.8625039419741407, + "grad_norm": 0.671875, + "learning_rate": 0.0002, + "loss": 1.7437, + "step": 5906 + }, + { + "epoch": 1.8628192999053925, + "grad_norm": 0.86328125, + "learning_rate": 0.0002, + "loss": 1.8686, + "step": 5907 + }, + { + "epoch": 1.8631346578366446, + "grad_norm": 0.80078125, + "learning_rate": 0.0002, + "loss": 1.7135, + "step": 5908 + }, + { + "epoch": 1.8634500157678966, + "grad_norm": 1.015625, + "learning_rate": 0.0002, + "loss": 2.258, + "step": 5909 + }, + { + "epoch": 1.8637653736991484, + "grad_norm": 0.9296875, + "learning_rate": 0.0002, + "loss": 2.0685, + "step": 5910 + }, + { + "epoch": 1.8640807316304007, + "grad_norm": 0.9375, + "learning_rate": 0.0002, + "loss": 1.7957, + "step": 5911 + }, + { + "epoch": 1.8643960895616525, + "grad_norm": 0.88671875, + "learning_rate": 0.0002, + "loss": 1.9304, + "step": 5912 + }, + { + "epoch": 1.8647114474929043, + "grad_norm": 0.921875, + "learning_rate": 0.0002, + "loss": 2.1673, + "step": 5913 + }, + { + "epoch": 1.8650268054241566, + "grad_norm": 0.88671875, + "learning_rate": 0.0002, + "loss": 1.8262, + "step": 5914 + }, + { + "epoch": 1.8653421633554084, + "grad_norm": 0.9921875, + "learning_rate": 0.0002, + "loss": 2.2527, + "step": 5915 + }, + { + "epoch": 1.8656575212866604, + "grad_norm": 0.92578125, + "learning_rate": 0.0002, + "loss": 2.0216, + "step": 5916 + }, + { + "epoch": 1.8659728792179124, + "grad_norm": 1.2265625, + "learning_rate": 0.0002, + "loss": 2.3618, + "step": 5917 + }, + { + "epoch": 1.8662882371491643, + "grad_norm": 1.2734375, + "learning_rate": 0.0002, + "loss": 2.3895, + "step": 5918 + }, + { + "epoch": 1.8666035950804163, + "grad_norm": 1.3828125, + "learning_rate": 0.0002, + "loss": 2.7287, + "step": 5919 + }, + { + "epoch": 1.8669189530116683, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 1.9531, + "step": 5920 + }, + { + "epoch": 1.8672343109429201, + "grad_norm": 0.89453125, + "learning_rate": 0.0002, + "loss": 1.9326, + "step": 5921 + }, + { + "epoch": 1.8675496688741722, + "grad_norm": 0.1884765625, + "learning_rate": 0.0002, + "loss": 1.0459, + "step": 5922 + }, + { + "epoch": 1.8678650268054242, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.0475, + "step": 5923 + }, + { + "epoch": 1.868180384736676, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.0508, + "step": 5924 + }, + { + "epoch": 1.868495742667928, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 0.9877, + "step": 5925 + }, + { + "epoch": 1.86881110059918, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 0.9321, + "step": 5926 + }, + { + "epoch": 1.869126458530432, + "grad_norm": 0.2216796875, + "learning_rate": 0.0002, + "loss": 1.0022, + "step": 5927 + }, + { + "epoch": 1.8694418164616842, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 1.1331, + "step": 5928 + }, + { + "epoch": 1.869757174392936, + "grad_norm": 0.224609375, + "learning_rate": 0.0002, + "loss": 1.0246, + "step": 5929 + }, + { + "epoch": 1.8700725323241878, + "grad_norm": 0.2431640625, + "learning_rate": 0.0002, + "loss": 1.1624, + "step": 5930 + }, + { + "epoch": 1.87038789025544, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 1.1124, + "step": 5931 + }, + { + "epoch": 1.8707032481866919, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 0.9265, + "step": 5932 + }, + { + "epoch": 1.871018606117944, + "grad_norm": 0.248046875, + "learning_rate": 0.0002, + "loss": 1.0887, + "step": 5933 + }, + { + "epoch": 1.871333964049196, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.0708, + "step": 5934 + }, + { + "epoch": 1.8716493219804478, + "grad_norm": 0.3515625, + "learning_rate": 0.0002, + "loss": 1.2134, + "step": 5935 + }, + { + "epoch": 1.8719646799116998, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 1.0038, + "step": 5936 + }, + { + "epoch": 1.8722800378429518, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 0.9802, + "step": 5937 + }, + { + "epoch": 1.8725953957742036, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.2479, + "step": 5938 + }, + { + "epoch": 1.8729107537054557, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 1.2005, + "step": 5939 + }, + { + "epoch": 1.8732261116367077, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 1.1522, + "step": 5940 + }, + { + "epoch": 1.8735414695679595, + "grad_norm": 0.369140625, + "learning_rate": 0.0002, + "loss": 1.2382, + "step": 5941 + }, + { + "epoch": 1.8738568274992116, + "grad_norm": 0.353515625, + "learning_rate": 0.0002, + "loss": 1.102, + "step": 5942 + }, + { + "epoch": 1.8741721854304636, + "grad_norm": 0.373046875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 5943 + }, + { + "epoch": 1.8744875433617154, + "grad_norm": 0.380859375, + "learning_rate": 0.0002, + "loss": 1.0892, + "step": 5944 + }, + { + "epoch": 1.8748029012929677, + "grad_norm": 0.388671875, + "learning_rate": 0.0002, + "loss": 1.0877, + "step": 5945 + }, + { + "epoch": 1.8751182592242195, + "grad_norm": 0.400390625, + "learning_rate": 0.0002, + "loss": 1.2085, + "step": 5946 + }, + { + "epoch": 1.8754336171554713, + "grad_norm": 0.42578125, + "learning_rate": 0.0002, + "loss": 1.2814, + "step": 5947 + }, + { + "epoch": 1.8757489750867236, + "grad_norm": 0.3984375, + "learning_rate": 0.0002, + "loss": 1.16, + "step": 5948 + }, + { + "epoch": 1.8760643330179754, + "grad_norm": 0.412109375, + "learning_rate": 0.0002, + "loss": 1.1614, + "step": 5949 + }, + { + "epoch": 1.8763796909492274, + "grad_norm": 0.498046875, + "learning_rate": 0.0002, + "loss": 1.2267, + "step": 5950 + }, + { + "epoch": 1.8766950488804794, + "grad_norm": 0.6640625, + "learning_rate": 0.0002, + "loss": 1.1374, + "step": 5951 + }, + { + "epoch": 1.8770104068117313, + "grad_norm": 0.60546875, + "learning_rate": 0.0002, + "loss": 1.4612, + "step": 5952 + }, + { + "epoch": 1.8773257647429833, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 5953 + }, + { + "epoch": 1.8776411226742353, + "grad_norm": 0.6796875, + "learning_rate": 0.0002, + "loss": 1.4763, + "step": 5954 + }, + { + "epoch": 1.8779564806054871, + "grad_norm": 0.6015625, + "learning_rate": 0.0002, + "loss": 1.4555, + "step": 5955 + }, + { + "epoch": 1.8782718385367392, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 1.4262, + "step": 5956 + }, + { + "epoch": 1.8785871964679912, + "grad_norm": 0.6875, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 5957 + }, + { + "epoch": 1.878902554399243, + "grad_norm": 0.796875, + "learning_rate": 0.0002, + "loss": 1.6413, + "step": 5958 + }, + { + "epoch": 1.879217912330495, + "grad_norm": 0.91796875, + "learning_rate": 0.0002, + "loss": 1.5866, + "step": 5959 + }, + { + "epoch": 1.879533270261747, + "grad_norm": 0.82421875, + "learning_rate": 0.0002, + "loss": 1.8334, + "step": 5960 + }, + { + "epoch": 1.879848628192999, + "grad_norm": 0.9453125, + "learning_rate": 0.0002, + "loss": 2.0075, + "step": 5961 + }, + { + "epoch": 1.8801639861242512, + "grad_norm": 0.94140625, + "learning_rate": 0.0002, + "loss": 2.283, + "step": 5962 + }, + { + "epoch": 1.880479344055503, + "grad_norm": 0.92578125, + "learning_rate": 0.0002, + "loss": 1.8046, + "step": 5963 + }, + { + "epoch": 1.8807947019867548, + "grad_norm": 0.80078125, + "learning_rate": 0.0002, + "loss": 1.7835, + "step": 5964 + }, + { + "epoch": 1.881110059918007, + "grad_norm": 0.921875, + "learning_rate": 0.0002, + "loss": 1.8405, + "step": 5965 + }, + { + "epoch": 1.8814254178492589, + "grad_norm": 0.82421875, + "learning_rate": 0.0002, + "loss": 1.7978, + "step": 5966 + }, + { + "epoch": 1.881740775780511, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 2.0311, + "step": 5967 + }, + { + "epoch": 1.882056133711763, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 2.0749, + "step": 5968 + }, + { + "epoch": 1.8823714916430148, + "grad_norm": 0.9609375, + "learning_rate": 0.0002, + "loss": 1.761, + "step": 5969 + }, + { + "epoch": 1.8826868495742668, + "grad_norm": 1.3984375, + "learning_rate": 0.0002, + "loss": 2.1418, + "step": 5970 + }, + { + "epoch": 1.8830022075055188, + "grad_norm": 1.5546875, + "learning_rate": 0.0002, + "loss": 2.317, + "step": 5971 + }, + { + "epoch": 1.8833175654367706, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 1.0057, + "step": 5972 + }, + { + "epoch": 1.8836329233680227, + "grad_norm": 0.2138671875, + "learning_rate": 0.0002, + "loss": 1.1103, + "step": 5973 + }, + { + "epoch": 1.8839482812992747, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 0.9898, + "step": 5974 + }, + { + "epoch": 1.8842636392305265, + "grad_norm": 0.212890625, + "learning_rate": 0.0002, + "loss": 1.0873, + "step": 5975 + }, + { + "epoch": 1.8845789971617786, + "grad_norm": 0.232421875, + "learning_rate": 0.0002, + "loss": 1.1468, + "step": 5976 + }, + { + "epoch": 1.8848943550930306, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 1.0533, + "step": 5977 + }, + { + "epoch": 1.8852097130242824, + "grad_norm": 0.216796875, + "learning_rate": 0.0002, + "loss": 1.0764, + "step": 5978 + }, + { + "epoch": 1.8855250709555347, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 1.0939, + "step": 5979 + }, + { + "epoch": 1.8858404288867865, + "grad_norm": 0.2294921875, + "learning_rate": 0.0002, + "loss": 0.9756, + "step": 5980 + }, + { + "epoch": 1.8861557868180385, + "grad_norm": 0.2451171875, + "learning_rate": 0.0002, + "loss": 1.0658, + "step": 5981 + }, + { + "epoch": 1.8864711447492906, + "grad_norm": 0.2421875, + "learning_rate": 0.0002, + "loss": 1.2079, + "step": 5982 + }, + { + "epoch": 1.8867865026805424, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 0.9724, + "step": 5983 + }, + { + "epoch": 1.8871018606117944, + "grad_norm": 0.248046875, + "learning_rate": 0.0002, + "loss": 1.0103, + "step": 5984 + }, + { + "epoch": 1.8871018606117944, + "eval_loss": 1.666214108467102, + "eval_runtime": 148.8928, + "eval_samples_per_second": 6.716, + "eval_steps_per_second": 6.716, + "step": 5984 + }, + { + "epoch": 1.8871018606117944, + "mmlu_eval_accuracy": 0.607973638722505, + "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, + "mmlu_eval_accuracy_anatomy": 0.5, + "mmlu_eval_accuracy_astronomy": 0.625, + "mmlu_eval_accuracy_business_ethics": 0.7272727272727273, + "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034, + "mmlu_eval_accuracy_college_biology": 0.8125, + "mmlu_eval_accuracy_college_chemistry": 0.375, + "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, + "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454, + "mmlu_eval_accuracy_college_medicine": 0.5909090909090909, + "mmlu_eval_accuracy_college_physics": 0.5454545454545454, + "mmlu_eval_accuracy_computer_security": 0.5454545454545454, + "mmlu_eval_accuracy_conceptual_physics": 0.5, + "mmlu_eval_accuracy_econometrics": 0.4166666666666667, + "mmlu_eval_accuracy_electrical_engineering": 0.5625, + "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244, + "mmlu_eval_accuracy_formal_logic": 0.2857142857142857, + "mmlu_eval_accuracy_global_facts": 0.2, + "mmlu_eval_accuracy_high_school_biology": 0.65625, + "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8571428571428571, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209, + "mmlu_eval_accuracy_high_school_mathematics": 0.3793103448275862, + "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384, + "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354, + "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667, + "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, + "mmlu_eval_accuracy_high_school_us_history": 0.8636363636363636, + "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, + "mmlu_eval_accuracy_human_aging": 0.6521739130434783, + "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334, + "mmlu_eval_accuracy_international_law": 1.0, + "mmlu_eval_accuracy_jurisprudence": 0.2727272727272727, + "mmlu_eval_accuracy_logical_fallacies": 0.8333333333333334, + "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.76, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279, + "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842, + "mmlu_eval_accuracy_moral_scenarios": 0.29, + "mmlu_eval_accuracy_nutrition": 0.7272727272727273, + "mmlu_eval_accuracy_philosophy": 0.6470588235294118, + "mmlu_eval_accuracy_prehistory": 0.6571428571428571, + "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355, + "mmlu_eval_accuracy_professional_law": 0.40588235294117647, + "mmlu_eval_accuracy_professional_medicine": 0.7419354838709677, + "mmlu_eval_accuracy_professional_psychology": 0.6521739130434783, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.6666666666666666, + "mmlu_eval_accuracy_sociology": 0.9545454545454546, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.5, + "mmlu_eval_accuracy_world_religions": 0.7368421052631579, + "mmlu_loss": 5.803138671086397, + "step": 5984 + }, + { + "epoch": 1.8874172185430464, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 1.107, + "step": 5985 + }, + { + "epoch": 1.8877325764742983, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 0.9771, + "step": 5986 + }, + { + "epoch": 1.8880479344055503, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 1.1459, + "step": 5987 + }, + { + "epoch": 1.8883632923368023, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.0454, + "step": 5988 + }, + { + "epoch": 1.8886786502680541, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 1.2795, + "step": 5989 + }, + { + "epoch": 1.8889940081993062, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 1.1467, + "step": 5990 + }, + { + "epoch": 1.8893093661305582, + "grad_norm": 0.349609375, + "learning_rate": 0.0002, + "loss": 1.3691, + "step": 5991 + }, + { + "epoch": 1.88962472406181, + "grad_norm": 0.45703125, + "learning_rate": 0.0002, + "loss": 1.2713, + "step": 5992 + }, + { + "epoch": 1.889940081993062, + "grad_norm": 0.390625, + "learning_rate": 0.0002, + "loss": 1.171, + "step": 5993 + }, + { + "epoch": 1.890255439924314, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 1.271, + "step": 5994 + }, + { + "epoch": 1.890570797855566, + "grad_norm": 0.3984375, + "learning_rate": 0.0002, + "loss": 1.226, + "step": 5995 + }, + { + "epoch": 1.8908861557868182, + "grad_norm": 0.453125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 5996 + }, + { + "epoch": 1.89120151371807, + "grad_norm": 0.46875, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 5997 + }, + { + "epoch": 1.891516871649322, + "grad_norm": 0.46875, + "learning_rate": 0.0002, + "loss": 1.2424, + "step": 5998 + }, + { + "epoch": 1.891832229580574, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 1.4213, + "step": 5999 + }, + { + "epoch": 1.8921475875118259, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 1.3727, + "step": 6000 + }, + { + "epoch": 1.892462945443078, + "grad_norm": 0.53515625, + "learning_rate": 0.0002, + "loss": 1.1783, + "step": 6001 + }, + { + "epoch": 1.89277830337433, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 1.1996, + "step": 6002 + }, + { + "epoch": 1.8930936613055818, + "grad_norm": 0.66796875, + "learning_rate": 0.0002, + "loss": 1.6804, + "step": 6003 + }, + { + "epoch": 1.8934090192368338, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 1.4607, + "step": 6004 + }, + { + "epoch": 1.8937243771680858, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 1.4713, + "step": 6005 + }, + { + "epoch": 1.8940397350993377, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 1.416, + "step": 6006 + }, + { + "epoch": 1.8943550930305897, + "grad_norm": 0.6953125, + "learning_rate": 0.0002, + "loss": 1.5652, + "step": 6007 + }, + { + "epoch": 1.8946704509618417, + "grad_norm": 0.91015625, + "learning_rate": 0.0002, + "loss": 1.8565, + "step": 6008 + }, + { + "epoch": 1.8949858088930935, + "grad_norm": 0.94140625, + "learning_rate": 0.0002, + "loss": 1.8919, + "step": 6009 + }, + { + "epoch": 1.8953011668243458, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 1.8496, + "step": 6010 + }, + { + "epoch": 1.8956165247555976, + "grad_norm": 0.82421875, + "learning_rate": 0.0002, + "loss": 1.7242, + "step": 6011 + }, + { + "epoch": 1.8959318826868494, + "grad_norm": 0.87890625, + "learning_rate": 0.0002, + "loss": 1.9218, + "step": 6012 + }, + { + "epoch": 1.8962472406181017, + "grad_norm": 0.85546875, + "learning_rate": 0.0002, + "loss": 2.0602, + "step": 6013 + }, + { + "epoch": 1.8965625985493535, + "grad_norm": 0.90625, + "learning_rate": 0.0002, + "loss": 2.2177, + "step": 6014 + }, + { + "epoch": 1.8968779564806055, + "grad_norm": 1.2734375, + "learning_rate": 0.0002, + "loss": 1.7165, + "step": 6015 + }, + { + "epoch": 1.8971933144118576, + "grad_norm": 1.4296875, + "learning_rate": 0.0002, + "loss": 2.5045, + "step": 6016 + }, + { + "epoch": 1.8975086723431094, + "grad_norm": 1.1953125, + "learning_rate": 0.0002, + "loss": 2.8165, + "step": 6017 + }, + { + "epoch": 1.8978240302743614, + "grad_norm": 0.859375, + "learning_rate": 0.0002, + "loss": 1.7698, + "step": 6018 + }, + { + "epoch": 1.8981393882056135, + "grad_norm": 1.265625, + "learning_rate": 0.0002, + "loss": 2.3073, + "step": 6019 + }, + { + "epoch": 1.8984547461368653, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 1.9946, + "step": 6020 + }, + { + "epoch": 1.8987701040681173, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 2.1747, + "step": 6021 + }, + { + "epoch": 1.8990854619993693, + "grad_norm": 0.2275390625, + "learning_rate": 0.0002, + "loss": 1.0992, + "step": 6022 + }, + { + "epoch": 1.8994008199306212, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.0168, + "step": 6023 + }, + { + "epoch": 1.8997161778618732, + "grad_norm": 0.232421875, + "learning_rate": 0.0002, + "loss": 1.0675, + "step": 6024 + }, + { + "epoch": 1.9000315357931252, + "grad_norm": 0.2177734375, + "learning_rate": 0.0002, + "loss": 1.031, + "step": 6025 + }, + { + "epoch": 1.900346893724377, + "grad_norm": 0.2333984375, + "learning_rate": 0.0002, + "loss": 0.991, + "step": 6026 + }, + { + "epoch": 1.9006622516556293, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 1.0511, + "step": 6027 + }, + { + "epoch": 1.9009776095868811, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.0094, + "step": 6028 + }, + { + "epoch": 1.901292967518133, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 1.0053, + "step": 6029 + }, + { + "epoch": 1.9016083254493852, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 0.9405, + "step": 6030 + }, + { + "epoch": 1.901923683380637, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.0537, + "step": 6031 + }, + { + "epoch": 1.902239041311889, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.0281, + "step": 6032 + }, + { + "epoch": 1.902554399243141, + "grad_norm": 0.2333984375, + "learning_rate": 0.0002, + "loss": 1.0432, + "step": 6033 + }, + { + "epoch": 1.9028697571743929, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 1.152, + "step": 6034 + }, + { + "epoch": 1.903185115105645, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.1461, + "step": 6035 + }, + { + "epoch": 1.903500473036897, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.2302, + "step": 6036 + }, + { + "epoch": 1.9038158309681488, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 0.988, + "step": 6037 + }, + { + "epoch": 1.9041311888994008, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.0753, + "step": 6038 + }, + { + "epoch": 1.9044465468306528, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 1.1176, + "step": 6039 + }, + { + "epoch": 1.9047619047619047, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 1.2022, + "step": 6040 + }, + { + "epoch": 1.9050772626931567, + "grad_norm": 0.40234375, + "learning_rate": 0.0002, + "loss": 1.1219, + "step": 6041 + }, + { + "epoch": 1.9053926206244087, + "grad_norm": 0.40625, + "learning_rate": 0.0002, + "loss": 1.4635, + "step": 6042 + }, + { + "epoch": 1.9057079785556605, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 6043 + }, + { + "epoch": 1.9060233364869128, + "grad_norm": 0.3984375, + "learning_rate": 0.0002, + "loss": 1.1204, + "step": 6044 + }, + { + "epoch": 1.9063386944181646, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 0.9538, + "step": 6045 + }, + { + "epoch": 1.9066540523494164, + "grad_norm": 0.4375, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 6046 + }, + { + "epoch": 1.9069694102806687, + "grad_norm": 0.421875, + "learning_rate": 0.0002, + "loss": 1.0763, + "step": 6047 + }, + { + "epoch": 1.9072847682119205, + "grad_norm": 0.3828125, + "learning_rate": 0.0002, + "loss": 1.2457, + "step": 6048 + }, + { + "epoch": 1.9076001261431725, + "grad_norm": 0.455078125, + "learning_rate": 0.0002, + "loss": 1.2937, + "step": 6049 + }, + { + "epoch": 1.9079154840744246, + "grad_norm": 0.4921875, + "learning_rate": 0.0002, + "loss": 1.206, + "step": 6050 + }, + { + "epoch": 1.9082308420056764, + "grad_norm": 0.625, + "learning_rate": 0.0002, + "loss": 1.4715, + "step": 6051 + }, + { + "epoch": 1.9085461999369284, + "grad_norm": 0.50390625, + "learning_rate": 0.0002, + "loss": 1.2597, + "step": 6052 + }, + { + "epoch": 1.9088615578681805, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 1.2956, + "step": 6053 + }, + { + "epoch": 1.9091769157994323, + "grad_norm": 0.68359375, + "learning_rate": 0.0002, + "loss": 1.4144, + "step": 6054 + }, + { + "epoch": 1.9094922737306843, + "grad_norm": 0.69921875, + "learning_rate": 0.0002, + "loss": 1.4482, + "step": 6055 + }, + { + "epoch": 1.9098076316619363, + "grad_norm": 0.91015625, + "learning_rate": 0.0002, + "loss": 1.8544, + "step": 6056 + }, + { + "epoch": 1.9101229895931882, + "grad_norm": 0.7109375, + "learning_rate": 0.0002, + "loss": 1.5608, + "step": 6057 + }, + { + "epoch": 1.9104383475244402, + "grad_norm": 0.8046875, + "learning_rate": 0.0002, + "loss": 1.7738, + "step": 6058 + }, + { + "epoch": 1.9107537054556922, + "grad_norm": 0.91015625, + "learning_rate": 0.0002, + "loss": 1.8315, + "step": 6059 + }, + { + "epoch": 1.911069063386944, + "grad_norm": 0.8359375, + "learning_rate": 0.0002, + "loss": 2.0941, + "step": 6060 + }, + { + "epoch": 1.9113844213181963, + "grad_norm": 0.8359375, + "learning_rate": 0.0002, + "loss": 1.5968, + "step": 6061 + }, + { + "epoch": 1.9116997792494481, + "grad_norm": 0.96875, + "learning_rate": 0.0002, + "loss": 1.969, + "step": 6062 + }, + { + "epoch": 1.9120151371807, + "grad_norm": 0.91015625, + "learning_rate": 0.0002, + "loss": 1.9419, + "step": 6063 + }, + { + "epoch": 1.9123304951119522, + "grad_norm": 0.984375, + "learning_rate": 0.0002, + "loss": 1.933, + "step": 6064 + }, + { + "epoch": 1.912645853043204, + "grad_norm": 0.984375, + "learning_rate": 0.0002, + "loss": 2.2661, + "step": 6065 + }, + { + "epoch": 1.912961210974456, + "grad_norm": 1.515625, + "learning_rate": 0.0002, + "loss": 2.7135, + "step": 6066 + }, + { + "epoch": 1.913276568905708, + "grad_norm": 0.9765625, + "learning_rate": 0.0002, + "loss": 2.2919, + "step": 6067 + }, + { + "epoch": 1.9135919268369599, + "grad_norm": 0.79296875, + "learning_rate": 0.0002, + "loss": 1.7654, + "step": 6068 + }, + { + "epoch": 1.913907284768212, + "grad_norm": 0.8515625, + "learning_rate": 0.0002, + "loss": 1.6829, + "step": 6069 + }, + { + "epoch": 1.914222642699464, + "grad_norm": 1.03125, + "learning_rate": 0.0002, + "loss": 1.6107, + "step": 6070 + }, + { + "epoch": 1.9145380006307158, + "grad_norm": 1.078125, + "learning_rate": 0.0002, + "loss": 2.3335, + "step": 6071 + }, + { + "epoch": 1.9148533585619678, + "grad_norm": 0.2041015625, + "learning_rate": 0.0002, + "loss": 1.1398, + "step": 6072 + }, + { + "epoch": 1.9151687164932198, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 0.9455, + "step": 6073 + }, + { + "epoch": 1.9154840744244717, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 1.1648, + "step": 6074 + }, + { + "epoch": 1.9157994323557237, + "grad_norm": 0.2314453125, + "learning_rate": 0.0002, + "loss": 1.0546, + "step": 6075 + }, + { + "epoch": 1.9161147902869757, + "grad_norm": 0.205078125, + "learning_rate": 0.0002, + "loss": 1.0023, + "step": 6076 + }, + { + "epoch": 1.9164301482182275, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 1.2108, + "step": 6077 + }, + { + "epoch": 1.9167455061494798, + "grad_norm": 0.2060546875, + "learning_rate": 0.0002, + "loss": 1.0092, + "step": 6078 + }, + { + "epoch": 1.9170608640807316, + "grad_norm": 0.2392578125, + "learning_rate": 0.0002, + "loss": 1.0871, + "step": 6079 + }, + { + "epoch": 1.9173762220119837, + "grad_norm": 0.2275390625, + "learning_rate": 0.0002, + "loss": 0.9516, + "step": 6080 + }, + { + "epoch": 1.9176915799432357, + "grad_norm": 0.2451171875, + "learning_rate": 0.0002, + "loss": 0.9874, + "step": 6081 + }, + { + "epoch": 1.9180069378744875, + "grad_norm": 0.232421875, + "learning_rate": 0.0002, + "loss": 1.1217, + "step": 6082 + }, + { + "epoch": 1.9183222958057395, + "grad_norm": 0.2333984375, + "learning_rate": 0.0002, + "loss": 1.0182, + "step": 6083 + }, + { + "epoch": 1.9186376537369916, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 0.9918, + "step": 6084 + }, + { + "epoch": 1.9189530116682434, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 1.1672, + "step": 6085 + }, + { + "epoch": 1.9192683695994954, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 1.136, + "step": 6086 + }, + { + "epoch": 1.9195837275307475, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.1152, + "step": 6087 + }, + { + "epoch": 1.9198990854619993, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.0537, + "step": 6088 + }, + { + "epoch": 1.9202144433932513, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 1.1506, + "step": 6089 + }, + { + "epoch": 1.9205298013245033, + "grad_norm": 0.361328125, + "learning_rate": 0.0002, + "loss": 1.3766, + "step": 6090 + }, + { + "epoch": 1.9208451592557552, + "grad_norm": 0.369140625, + "learning_rate": 0.0002, + "loss": 1.1976, + "step": 6091 + }, + { + "epoch": 1.9211605171870072, + "grad_norm": 0.330078125, + "learning_rate": 0.0002, + "loss": 1.1004, + "step": 6092 + }, + { + "epoch": 1.9214758751182592, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 1.1825, + "step": 6093 + }, + { + "epoch": 1.921791233049511, + "grad_norm": 0.39453125, + "learning_rate": 0.0002, + "loss": 1.2818, + "step": 6094 + }, + { + "epoch": 1.9221065909807633, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 1.1654, + "step": 6095 + }, + { + "epoch": 1.9224219489120151, + "grad_norm": 0.3671875, + "learning_rate": 0.0002, + "loss": 1.1831, + "step": 6096 + }, + { + "epoch": 1.9227373068432672, + "grad_norm": 0.408203125, + "learning_rate": 0.0002, + "loss": 1.1764, + "step": 6097 + }, + { + "epoch": 1.9230526647745192, + "grad_norm": 0.50390625, + "learning_rate": 0.0002, + "loss": 1.2992, + "step": 6098 + }, + { + "epoch": 1.923368022705771, + "grad_norm": 0.4453125, + "learning_rate": 0.0002, + "loss": 1.0538, + "step": 6099 + }, + { + "epoch": 1.923683380637023, + "grad_norm": 0.52734375, + "learning_rate": 0.0002, + "loss": 1.5389, + "step": 6100 + }, + { + "epoch": 1.923998738568275, + "grad_norm": 0.4921875, + "learning_rate": 0.0002, + "loss": 1.1409, + "step": 6101 + }, + { + "epoch": 1.924314096499527, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 1.4007, + "step": 6102 + }, + { + "epoch": 1.924629454430779, + "grad_norm": 0.490234375, + "learning_rate": 0.0002, + "loss": 1.2636, + "step": 6103 + }, + { + "epoch": 1.924944812362031, + "grad_norm": 0.5390625, + "learning_rate": 0.0002, + "loss": 1.2258, + "step": 6104 + }, + { + "epoch": 1.9252601702932828, + "grad_norm": 0.578125, + "learning_rate": 0.0002, + "loss": 1.5219, + "step": 6105 + }, + { + "epoch": 1.9255755282245348, + "grad_norm": 1.1171875, + "learning_rate": 0.0002, + "loss": 2.0995, + "step": 6106 + }, + { + "epoch": 1.9258908861557869, + "grad_norm": 0.70703125, + "learning_rate": 0.0002, + "loss": 1.7002, + "step": 6107 + }, + { + "epoch": 1.9262062440870387, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 1.7653, + "step": 6108 + }, + { + "epoch": 1.926521602018291, + "grad_norm": 0.80078125, + "learning_rate": 0.0002, + "loss": 1.672, + "step": 6109 + }, + { + "epoch": 1.9268369599495427, + "grad_norm": 0.74609375, + "learning_rate": 0.0002, + "loss": 1.6693, + "step": 6110 + }, + { + "epoch": 1.9271523178807946, + "grad_norm": 0.890625, + "learning_rate": 0.0002, + "loss": 1.7838, + "step": 6111 + }, + { + "epoch": 1.9274676758120468, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 2.7141, + "step": 6112 + }, + { + "epoch": 1.9277830337432986, + "grad_norm": 0.94140625, + "learning_rate": 0.0002, + "loss": 2.0401, + "step": 6113 + }, + { + "epoch": 1.9280983916745507, + "grad_norm": 0.91015625, + "learning_rate": 0.0002, + "loss": 1.9483, + "step": 6114 + }, + { + "epoch": 1.9284137496058027, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 1.8893, + "step": 6115 + }, + { + "epoch": 1.9287291075370545, + "grad_norm": 1.0546875, + "learning_rate": 0.0002, + "loss": 2.5115, + "step": 6116 + }, + { + "epoch": 1.9290444654683065, + "grad_norm": 1.0390625, + "learning_rate": 0.0002, + "loss": 2.2338, + "step": 6117 + }, + { + "epoch": 1.9293598233995586, + "grad_norm": 1.1171875, + "learning_rate": 0.0002, + "loss": 2.4549, + "step": 6118 + }, + { + "epoch": 1.9296751813308104, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 2.0452, + "step": 6119 + }, + { + "epoch": 1.9299905392620624, + "grad_norm": 0.99609375, + "learning_rate": 0.0002, + "loss": 1.7953, + "step": 6120 + }, + { + "epoch": 1.9303058971933145, + "grad_norm": 1.0390625, + "learning_rate": 0.0002, + "loss": 1.8425, + "step": 6121 + }, + { + "epoch": 1.9306212551245663, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.0941, + "step": 6122 + }, + { + "epoch": 1.9309366130558183, + "grad_norm": 0.20703125, + "learning_rate": 0.0002, + "loss": 1.0144, + "step": 6123 + }, + { + "epoch": 1.9312519709870704, + "grad_norm": 0.2138671875, + "learning_rate": 0.0002, + "loss": 1.0521, + "step": 6124 + }, + { + "epoch": 1.9315673289183222, + "grad_norm": 0.2265625, + "learning_rate": 0.0002, + "loss": 1.1632, + "step": 6125 + }, + { + "epoch": 1.9318826868495744, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.09, + "step": 6126 + }, + { + "epoch": 1.9321980447808262, + "grad_norm": 0.2373046875, + "learning_rate": 0.0002, + "loss": 1.0342, + "step": 6127 + }, + { + "epoch": 1.932513402712078, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 0.9532, + "step": 6128 + }, + { + "epoch": 1.9328287606433303, + "grad_norm": 0.240234375, + "learning_rate": 0.0002, + "loss": 1.0816, + "step": 6129 + }, + { + "epoch": 1.9331441185745821, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 0.9774, + "step": 6130 + }, + { + "epoch": 1.9334594765058342, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 0.9706, + "step": 6131 + }, + { + "epoch": 1.9337748344370862, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 1.0685, + "step": 6132 + }, + { + "epoch": 1.934090192368338, + "grad_norm": 0.2470703125, + "learning_rate": 0.0002, + "loss": 1.0196, + "step": 6133 + }, + { + "epoch": 1.93440555029959, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.1079, + "step": 6134 + }, + { + "epoch": 1.934720908230842, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 1.1888, + "step": 6135 + }, + { + "epoch": 1.935036266162094, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.2029, + "step": 6136 + }, + { + "epoch": 1.935351624093346, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 1.1966, + "step": 6137 + }, + { + "epoch": 1.935666982024598, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 1.2734, + "step": 6138 + }, + { + "epoch": 1.9359823399558498, + "grad_norm": 0.330078125, + "learning_rate": 0.0002, + "loss": 1.1638, + "step": 6139 + }, + { + "epoch": 1.9362976978871018, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 1.0794, + "step": 6140 + }, + { + "epoch": 1.9366130558183539, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 1.0766, + "step": 6141 + }, + { + "epoch": 1.9369284137496057, + "grad_norm": 0.390625, + "learning_rate": 0.0002, + "loss": 1.385, + "step": 6142 + }, + { + "epoch": 1.937243771680858, + "grad_norm": 0.373046875, + "learning_rate": 0.0002, + "loss": 1.1157, + "step": 6143 + }, + { + "epoch": 1.9375591296121097, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 1.2257, + "step": 6144 + }, + { + "epoch": 1.9378744875433616, + "grad_norm": 0.42578125, + "learning_rate": 0.0002, + "loss": 1.2693, + "step": 6145 + }, + { + "epoch": 1.9381898454746138, + "grad_norm": 0.41796875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 6146 + }, + { + "epoch": 1.9385052034058656, + "grad_norm": 0.380859375, + "learning_rate": 0.0002, + "loss": 1.1885, + "step": 6147 + }, + { + "epoch": 1.9388205613371177, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 1.1342, + "step": 6148 + }, + { + "epoch": 1.9391359192683697, + "grad_norm": 0.45703125, + "learning_rate": 0.0002, + "loss": 1.2201, + "step": 6149 + }, + { + "epoch": 1.9394512771996215, + "grad_norm": 0.478515625, + "learning_rate": 0.0002, + "loss": 1.3533, + "step": 6150 + }, + { + "epoch": 1.9397666351308736, + "grad_norm": 0.466796875, + "learning_rate": 0.0002, + "loss": 1.2394, + "step": 6151 + }, + { + "epoch": 1.9400819930621256, + "grad_norm": 0.458984375, + "learning_rate": 0.0002, + "loss": 1.0797, + "step": 6152 + }, + { + "epoch": 1.9403973509933774, + "grad_norm": 0.5546875, + "learning_rate": 0.0002, + "loss": 1.2649, + "step": 6153 + }, + { + "epoch": 1.9407127089246294, + "grad_norm": 0.8046875, + "learning_rate": 0.0002, + "loss": 1.8821, + "step": 6154 + }, + { + "epoch": 1.9410280668558815, + "grad_norm": 0.640625, + "learning_rate": 0.0002, + "loss": 1.6848, + "step": 6155 + }, + { + "epoch": 1.9413434247871333, + "grad_norm": 0.75, + "learning_rate": 0.0002, + "loss": 1.5563, + "step": 6156 + }, + { + "epoch": 1.9416587827183853, + "grad_norm": 0.60546875, + "learning_rate": 0.0002, + "loss": 1.3894, + "step": 6157 + }, + { + "epoch": 1.9419741406496374, + "grad_norm": 0.75, + "learning_rate": 0.0002, + "loss": 1.7104, + "step": 6158 + }, + { + "epoch": 1.9422894985808892, + "grad_norm": 0.70703125, + "learning_rate": 0.0002, + "loss": 1.4135, + "step": 6159 + }, + { + "epoch": 1.9426048565121414, + "grad_norm": 1.015625, + "learning_rate": 0.0002, + "loss": 1.9167, + "step": 6160 + }, + { + "epoch": 1.9429202144433932, + "grad_norm": 0.7890625, + "learning_rate": 0.0002, + "loss": 2.0501, + "step": 6161 + }, + { + "epoch": 1.943235572374645, + "grad_norm": 0.8203125, + "learning_rate": 0.0002, + "loss": 1.9636, + "step": 6162 + }, + { + "epoch": 1.9435509303058973, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 2.1653, + "step": 6163 + }, + { + "epoch": 1.9438662882371491, + "grad_norm": 0.953125, + "learning_rate": 0.0002, + "loss": 1.9942, + "step": 6164 + }, + { + "epoch": 1.9441816461684012, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 2.5369, + "step": 6165 + }, + { + "epoch": 1.9444970040996532, + "grad_norm": 1.2265625, + "learning_rate": 0.0002, + "loss": 2.4261, + "step": 6166 + }, + { + "epoch": 1.944812362030905, + "grad_norm": 0.98828125, + "learning_rate": 0.0002, + "loss": 2.1261, + "step": 6167 + }, + { + "epoch": 1.945127719962157, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 2.2369, + "step": 6168 + }, + { + "epoch": 1.945443077893409, + "grad_norm": 0.9453125, + "learning_rate": 0.0002, + "loss": 1.7797, + "step": 6169 + }, + { + "epoch": 1.945758435824661, + "grad_norm": 0.953125, + "learning_rate": 0.0002, + "loss": 1.9609, + "step": 6170 + }, + { + "epoch": 1.946073793755913, + "grad_norm": 1.2109375, + "learning_rate": 0.0002, + "loss": 2.3888, + "step": 6171 + }, + { + "epoch": 1.946073793755913, + "eval_loss": 1.6738756895065308, + "eval_runtime": 148.4733, + "eval_samples_per_second": 6.735, + "eval_steps_per_second": 6.735, + "step": 6171 + }, + { + "epoch": 1.946073793755913, + "mmlu_eval_accuracy": 0.6013445476071988, + "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, + "mmlu_eval_accuracy_anatomy": 0.42857142857142855, + "mmlu_eval_accuracy_astronomy": 0.625, + "mmlu_eval_accuracy_business_ethics": 0.7272727272727273, + "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241, + "mmlu_eval_accuracy_college_biology": 0.75, + "mmlu_eval_accuracy_college_chemistry": 0.5, + "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, + "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453, + "mmlu_eval_accuracy_college_medicine": 0.5454545454545454, + "mmlu_eval_accuracy_college_physics": 0.5454545454545454, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384, + "mmlu_eval_accuracy_econometrics": 0.5, + "mmlu_eval_accuracy_electrical_engineering": 0.4375, + "mmlu_eval_accuracy_elementary_mathematics": 0.5121951219512195, + "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, + "mmlu_eval_accuracy_global_facts": 0.2, + "mmlu_eval_accuracy_high_school_biology": 0.625, + "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.9545454545454546, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8571428571428571, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.627906976744186, + "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276, + "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384, + "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, + "mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, + "mmlu_eval_accuracy_high_school_us_history": 0.8636363636363636, + "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, + "mmlu_eval_accuracy_human_aging": 0.6521739130434783, + "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334, + "mmlu_eval_accuracy_international_law": 0.9230769230769231, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.36363636363636365, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.76, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7558139534883721, + "mmlu_eval_accuracy_moral_disputes": 0.6578947368421053, + "mmlu_eval_accuracy_moral_scenarios": 0.32, + "mmlu_eval_accuracy_nutrition": 0.7575757575757576, + "mmlu_eval_accuracy_philosophy": 0.6470588235294118, + "mmlu_eval_accuracy_prehistory": 0.6571428571428571, + "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355, + "mmlu_eval_accuracy_professional_law": 0.37058823529411766, + "mmlu_eval_accuracy_professional_medicine": 0.8064516129032258, + "mmlu_eval_accuracy_professional_psychology": 0.6521739130434783, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7037037037037037, + "mmlu_eval_accuracy_sociology": 1.0, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.5, + "mmlu_eval_accuracy_world_religions": 0.7368421052631579, + "mmlu_loss": 5.373320624689741, + "step": 6171 + }, + { + "epoch": 1.946389151687165, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 1.0441, + "step": 6172 + }, + { + "epoch": 1.9467045096184168, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.0088, + "step": 6173 + }, + { + "epoch": 1.9470198675496688, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 0.993, + "step": 6174 + }, + { + "epoch": 1.9473352254809209, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.0155, + "step": 6175 + }, + { + "epoch": 1.9476505834121727, + "grad_norm": 0.2392578125, + "learning_rate": 0.0002, + "loss": 1.0949, + "step": 6176 + }, + { + "epoch": 1.947965941343425, + "grad_norm": 0.2314453125, + "learning_rate": 0.0002, + "loss": 1.0121, + "step": 6177 + }, + { + "epoch": 1.9482812992746767, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 1.001, + "step": 6178 + }, + { + "epoch": 1.9485966572059288, + "grad_norm": 0.240234375, + "learning_rate": 0.0002, + "loss": 0.8936, + "step": 6179 + }, + { + "epoch": 1.9489120151371808, + "grad_norm": 0.2392578125, + "learning_rate": 0.0002, + "loss": 0.982, + "step": 6180 + }, + { + "epoch": 1.9492273730684326, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 0.8514, + "step": 6181 + }, + { + "epoch": 1.9495427309996847, + "grad_norm": 0.2431640625, + "learning_rate": 0.0002, + "loss": 1.0755, + "step": 6182 + }, + { + "epoch": 1.9498580889309367, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 1.0948, + "step": 6183 + }, + { + "epoch": 1.9501734468621885, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 1.1692, + "step": 6184 + }, + { + "epoch": 1.9504888047934406, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 0.9606, + "step": 6185 + }, + { + "epoch": 1.9508041627246926, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 1.1203, + "step": 6186 + }, + { + "epoch": 1.9511195206559444, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 0.9865, + "step": 6187 + }, + { + "epoch": 1.9514348785871964, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 1.0446, + "step": 6188 + }, + { + "epoch": 1.9517502365184485, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 1.1423, + "step": 6189 + }, + { + "epoch": 1.9520655944497003, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 1.1558, + "step": 6190 + }, + { + "epoch": 1.9523809523809523, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 1.0654, + "step": 6191 + }, + { + "epoch": 1.9526963103122044, + "grad_norm": 0.380859375, + "learning_rate": 0.0002, + "loss": 1.0641, + "step": 6192 + }, + { + "epoch": 1.9530116682434562, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 1.0723, + "step": 6193 + }, + { + "epoch": 1.9533270261747084, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 0.9694, + "step": 6194 + }, + { + "epoch": 1.9536423841059603, + "grad_norm": 0.46875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 6195 + }, + { + "epoch": 1.9539577420372123, + "grad_norm": 0.421875, + "learning_rate": 0.0002, + "loss": 1.2374, + "step": 6196 + }, + { + "epoch": 1.9542730999684643, + "grad_norm": 0.3671875, + "learning_rate": 0.0002, + "loss": 1.0902, + "step": 6197 + }, + { + "epoch": 1.9545884578997161, + "grad_norm": 0.4609375, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 6198 + }, + { + "epoch": 1.9549038158309682, + "grad_norm": 0.486328125, + "learning_rate": 0.0002, + "loss": 1.2098, + "step": 6199 + }, + { + "epoch": 1.9552191737622202, + "grad_norm": 0.4609375, + "learning_rate": 0.0002, + "loss": 1.22, + "step": 6200 + }, + { + "epoch": 1.955534531693472, + "grad_norm": 0.4609375, + "learning_rate": 0.0002, + "loss": 1.2062, + "step": 6201 + }, + { + "epoch": 1.955849889624724, + "grad_norm": 0.4765625, + "learning_rate": 0.0002, + "loss": 1.0812, + "step": 6202 + }, + { + "epoch": 1.956165247555976, + "grad_norm": 0.515625, + "learning_rate": 0.0002, + "loss": 1.2682, + "step": 6203 + }, + { + "epoch": 1.956480605487228, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 1.2766, + "step": 6204 + }, + { + "epoch": 1.95679596341848, + "grad_norm": 0.63671875, + "learning_rate": 0.0002, + "loss": 1.6186, + "step": 6205 + }, + { + "epoch": 1.957111321349732, + "grad_norm": 0.71875, + "learning_rate": 0.0002, + "loss": 1.7181, + "step": 6206 + }, + { + "epoch": 1.9574266792809838, + "grad_norm": 0.6796875, + "learning_rate": 0.0002, + "loss": 1.6448, + "step": 6207 + }, + { + "epoch": 1.957742037212236, + "grad_norm": 0.75390625, + "learning_rate": 0.0002, + "loss": 1.6911, + "step": 6208 + }, + { + "epoch": 1.9580573951434879, + "grad_norm": 0.75, + "learning_rate": 0.0002, + "loss": 1.8729, + "step": 6209 + }, + { + "epoch": 1.9583727530747397, + "grad_norm": 0.87109375, + "learning_rate": 0.0002, + "loss": 1.8781, + "step": 6210 + }, + { + "epoch": 1.958688111005992, + "grad_norm": 0.83984375, + "learning_rate": 0.0002, + "loss": 1.9771, + "step": 6211 + }, + { + "epoch": 1.9590034689372438, + "grad_norm": 0.9296875, + "learning_rate": 0.0002, + "loss": 2.0067, + "step": 6212 + }, + { + "epoch": 1.9593188268684958, + "grad_norm": 0.9375, + "learning_rate": 0.0002, + "loss": 1.9497, + "step": 6213 + }, + { + "epoch": 1.9596341847997478, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 1.9869, + "step": 6214 + }, + { + "epoch": 1.9599495427309996, + "grad_norm": 1.15625, + "learning_rate": 0.0002, + "loss": 2.0369, + "step": 6215 + }, + { + "epoch": 1.9602649006622517, + "grad_norm": 1.1875, + "learning_rate": 0.0002, + "loss": 2.4176, + "step": 6216 + }, + { + "epoch": 1.9605802585935037, + "grad_norm": 0.82421875, + "learning_rate": 0.0002, + "loss": 1.747, + "step": 6217 + }, + { + "epoch": 1.9608956165247555, + "grad_norm": 0.92578125, + "learning_rate": 0.0002, + "loss": 2.055, + "step": 6218 + }, + { + "epoch": 1.9612109744560076, + "grad_norm": 1.4375, + "learning_rate": 0.0002, + "loss": 1.9239, + "step": 6219 + }, + { + "epoch": 1.9615263323872596, + "grad_norm": 1.2421875, + "learning_rate": 0.0002, + "loss": 1.9857, + "step": 6220 + }, + { + "epoch": 1.9618416903185114, + "grad_norm": 1.3125, + "learning_rate": 0.0002, + "loss": 2.563, + "step": 6221 + }, + { + "epoch": 1.9621570482497634, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 0.9285, + "step": 6222 + }, + { + "epoch": 1.9624724061810155, + "grad_norm": 0.224609375, + "learning_rate": 0.0002, + "loss": 1.019, + "step": 6223 + }, + { + "epoch": 1.9627877641122673, + "grad_norm": 0.244140625, + "learning_rate": 0.0002, + "loss": 0.9436, + "step": 6224 + }, + { + "epoch": 1.9631031220435196, + "grad_norm": 0.244140625, + "learning_rate": 0.0002, + "loss": 1.1363, + "step": 6225 + }, + { + "epoch": 1.9634184799747714, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.04, + "step": 6226 + }, + { + "epoch": 1.9637338379060232, + "grad_norm": 0.2431640625, + "learning_rate": 0.0002, + "loss": 1.1339, + "step": 6227 + }, + { + "epoch": 1.9640491958372754, + "grad_norm": 0.240234375, + "learning_rate": 0.0002, + "loss": 1.0424, + "step": 6228 + }, + { + "epoch": 1.9643645537685273, + "grad_norm": 0.240234375, + "learning_rate": 0.0002, + "loss": 1.1608, + "step": 6229 + }, + { + "epoch": 1.9646799116997793, + "grad_norm": 0.2421875, + "learning_rate": 0.0002, + "loss": 1.0209, + "step": 6230 + }, + { + "epoch": 1.9649952696310313, + "grad_norm": 0.2431640625, + "learning_rate": 0.0002, + "loss": 1.1156, + "step": 6231 + }, + { + "epoch": 1.9653106275622831, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.2398, + "step": 6232 + }, + { + "epoch": 1.9656259854935352, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 1.023, + "step": 6233 + }, + { + "epoch": 1.9659413434247872, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.0067, + "step": 6234 + }, + { + "epoch": 1.966256701356039, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 1.1252, + "step": 6235 + }, + { + "epoch": 1.966572059287291, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.2311, + "step": 6236 + }, + { + "epoch": 1.966887417218543, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 1.1348, + "step": 6237 + }, + { + "epoch": 1.967202775149795, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 1.1166, + "step": 6238 + }, + { + "epoch": 1.967518133081047, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 1.1656, + "step": 6239 + }, + { + "epoch": 1.967833491012299, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 1.0681, + "step": 6240 + }, + { + "epoch": 1.9681488489435508, + "grad_norm": 0.3515625, + "learning_rate": 0.0002, + "loss": 1.0895, + "step": 6241 + }, + { + "epoch": 1.968464206874803, + "grad_norm": 0.451171875, + "learning_rate": 0.0002, + "loss": 1.1312, + "step": 6242 + }, + { + "epoch": 1.9687795648060549, + "grad_norm": 0.384765625, + "learning_rate": 0.0002, + "loss": 1.1262, + "step": 6243 + }, + { + "epoch": 1.9690949227373067, + "grad_norm": 0.38671875, + "learning_rate": 0.0002, + "loss": 1.2046, + "step": 6244 + }, + { + "epoch": 1.969410280668559, + "grad_norm": 0.3828125, + "learning_rate": 0.0002, + "loss": 1.1661, + "step": 6245 + }, + { + "epoch": 1.9697256385998108, + "grad_norm": 0.39453125, + "learning_rate": 0.0002, + "loss": 1.1956, + "step": 6246 + }, + { + "epoch": 1.9700409965310628, + "grad_norm": 0.412109375, + "learning_rate": 0.0002, + "loss": 1.1437, + "step": 6247 + }, + { + "epoch": 1.9703563544623148, + "grad_norm": 0.451171875, + "learning_rate": 0.0002, + "loss": 1.2102, + "step": 6248 + }, + { + "epoch": 1.9706717123935666, + "grad_norm": 0.462890625, + "learning_rate": 0.0002, + "loss": 1.1443, + "step": 6249 + }, + { + "epoch": 1.9709870703248187, + "grad_norm": 0.47265625, + "learning_rate": 0.0002, + "loss": 1.1748, + "step": 6250 + }, + { + "epoch": 1.9713024282560707, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 6251 + }, + { + "epoch": 1.9716177861873225, + "grad_norm": 0.5703125, + "learning_rate": 0.0002, + "loss": 1.5608, + "step": 6252 + }, + { + "epoch": 1.9719331441185746, + "grad_norm": 0.640625, + "learning_rate": 0.0002, + "loss": 1.5696, + "step": 6253 + }, + { + "epoch": 1.9722485020498266, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 1.359, + "step": 6254 + }, + { + "epoch": 1.9725638599810784, + "grad_norm": 0.7578125, + "learning_rate": 0.0002, + "loss": 1.7245, + "step": 6255 + }, + { + "epoch": 1.9728792179123305, + "grad_norm": 0.7578125, + "learning_rate": 0.0002, + "loss": 1.8494, + "step": 6256 + }, + { + "epoch": 1.9731945758435825, + "grad_norm": 0.9296875, + "learning_rate": 0.0002, + "loss": 1.8159, + "step": 6257 + }, + { + "epoch": 1.9735099337748343, + "grad_norm": 0.921875, + "learning_rate": 0.0002, + "loss": 1.7569, + "step": 6258 + }, + { + "epoch": 1.9738252917060866, + "grad_norm": 0.97265625, + "learning_rate": 0.0002, + "loss": 2.1586, + "step": 6259 + }, + { + "epoch": 1.9741406496373384, + "grad_norm": 0.9765625, + "learning_rate": 0.0002, + "loss": 2.0865, + "step": 6260 + }, + { + "epoch": 1.9744560075685902, + "grad_norm": 0.80078125, + "learning_rate": 0.0002, + "loss": 1.9905, + "step": 6261 + }, + { + "epoch": 1.9747713654998424, + "grad_norm": 0.80078125, + "learning_rate": 0.0002, + "loss": 1.9823, + "step": 6262 + }, + { + "epoch": 1.9750867234310943, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 2.267, + "step": 6263 + }, + { + "epoch": 1.9754020813623463, + "grad_norm": 0.85546875, + "learning_rate": 0.0002, + "loss": 2.1153, + "step": 6264 + }, + { + "epoch": 1.9757174392935983, + "grad_norm": 0.984375, + "learning_rate": 0.0002, + "loss": 2.1151, + "step": 6265 + }, + { + "epoch": 1.9760327972248501, + "grad_norm": 1.6328125, + "learning_rate": 0.0002, + "loss": 2.5216, + "step": 6266 + }, + { + "epoch": 1.9763481551561022, + "grad_norm": 0.89453125, + "learning_rate": 0.0002, + "loss": 1.7367, + "step": 6267 + }, + { + "epoch": 1.9766635130873542, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 2.0851, + "step": 6268 + }, + { + "epoch": 1.976978871018606, + "grad_norm": 0.9921875, + "learning_rate": 0.0002, + "loss": 1.8833, + "step": 6269 + }, + { + "epoch": 1.977294228949858, + "grad_norm": 1.0, + "learning_rate": 0.0002, + "loss": 1.5694, + "step": 6270 + }, + { + "epoch": 1.97760958688111, + "grad_norm": 1.2734375, + "learning_rate": 0.0002, + "loss": 2.0359, + "step": 6271 + }, + { + "epoch": 1.977924944812362, + "grad_norm": 0.19921875, + "learning_rate": 0.0002, + "loss": 1.0426, + "step": 6272 + }, + { + "epoch": 1.978240302743614, + "grad_norm": 0.216796875, + "learning_rate": 0.0002, + "loss": 1.0864, + "step": 6273 + }, + { + "epoch": 1.978555660674866, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.1203, + "step": 6274 + }, + { + "epoch": 1.9788710186061178, + "grad_norm": 0.2275390625, + "learning_rate": 0.0002, + "loss": 1.1653, + "step": 6275 + }, + { + "epoch": 1.97918637653737, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 1.0879, + "step": 6276 + }, + { + "epoch": 1.9795017344686219, + "grad_norm": 0.224609375, + "learning_rate": 0.0002, + "loss": 1.0633, + "step": 6277 + }, + { + "epoch": 1.9798170923998737, + "grad_norm": 0.2333984375, + "learning_rate": 0.0002, + "loss": 1.1255, + "step": 6278 + }, + { + "epoch": 1.980132450331126, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 1.1729, + "step": 6279 + }, + { + "epoch": 1.9804478082623778, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 1.1171, + "step": 6280 + }, + { + "epoch": 1.9807631661936298, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 0.9225, + "step": 6281 + }, + { + "epoch": 1.9810785241248818, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 0.9674, + "step": 6282 + }, + { + "epoch": 1.9813938820561336, + "grad_norm": 0.2421875, + "learning_rate": 0.0002, + "loss": 1.0655, + "step": 6283 + }, + { + "epoch": 1.9817092399873857, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 0.9168, + "step": 6284 + }, + { + "epoch": 1.9820245979186377, + "grad_norm": 0.2373046875, + "learning_rate": 0.0002, + "loss": 0.9909, + "step": 6285 + }, + { + "epoch": 1.9823399558498895, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 1.2606, + "step": 6286 + }, + { + "epoch": 1.9826553137811416, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 1.0127, + "step": 6287 + }, + { + "epoch": 1.9829706717123936, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 1.1748, + "step": 6288 + }, + { + "epoch": 1.9832860296436454, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.1519, + "step": 6289 + }, + { + "epoch": 1.9836013875748975, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 1.0305, + "step": 6290 + }, + { + "epoch": 1.9839167455061495, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 1.1299, + "step": 6291 + }, + { + "epoch": 1.9842321034374013, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 0.9739, + "step": 6292 + }, + { + "epoch": 1.9845474613686536, + "grad_norm": 0.37109375, + "learning_rate": 0.0002, + "loss": 1.1076, + "step": 6293 + }, + { + "epoch": 1.9848628192999054, + "grad_norm": 0.427734375, + "learning_rate": 0.0002, + "loss": 1.1485, + "step": 6294 + }, + { + "epoch": 1.9851781772311574, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 1.1464, + "step": 6295 + }, + { + "epoch": 1.9854935351624095, + "grad_norm": 0.3828125, + "learning_rate": 0.0002, + "loss": 1.076, + "step": 6296 + }, + { + "epoch": 1.9858088930936613, + "grad_norm": 0.412109375, + "learning_rate": 0.0002, + "loss": 1.2134, + "step": 6297 + }, + { + "epoch": 1.9861242510249133, + "grad_norm": 0.435546875, + "learning_rate": 0.0002, + "loss": 1.1225, + "step": 6298 + }, + { + "epoch": 1.9864396089561653, + "grad_norm": 0.5, + "learning_rate": 0.0002, + "loss": 1.386, + "step": 6299 + }, + { + "epoch": 1.9867549668874172, + "grad_norm": 0.5, + "learning_rate": 0.0002, + "loss": 1.3598, + "step": 6300 + }, + { + "epoch": 1.9870703248186692, + "grad_norm": 0.47265625, + "learning_rate": 0.0002, + "loss": 1.362, + "step": 6301 + }, + { + "epoch": 1.9873856827499212, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 1.2954, + "step": 6302 + }, + { + "epoch": 1.987701040681173, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 1.416, + "step": 6303 + }, + { + "epoch": 1.988016398612425, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 1.2457, + "step": 6304 + }, + { + "epoch": 1.988331756543677, + "grad_norm": 0.70703125, + "learning_rate": 0.0002, + "loss": 1.567, + "step": 6305 + }, + { + "epoch": 1.988647114474929, + "grad_norm": 0.765625, + "learning_rate": 0.0002, + "loss": 1.6789, + "step": 6306 + }, + { + "epoch": 1.9889624724061812, + "grad_norm": 0.96875, + "learning_rate": 0.0002, + "loss": 1.8594, + "step": 6307 + }, + { + "epoch": 1.989277830337433, + "grad_norm": 0.76953125, + "learning_rate": 0.0002, + "loss": 1.6516, + "step": 6308 + }, + { + "epoch": 1.9895931882686848, + "grad_norm": 0.859375, + "learning_rate": 0.0002, + "loss": 1.7067, + "step": 6309 + }, + { + "epoch": 1.989908546199937, + "grad_norm": 0.83984375, + "learning_rate": 0.0002, + "loss": 1.8895, + "step": 6310 + }, + { + "epoch": 1.9902239041311889, + "grad_norm": 0.875, + "learning_rate": 0.0002, + "loss": 1.7202, + "step": 6311 + }, + { + "epoch": 1.990539262062441, + "grad_norm": 1.171875, + "learning_rate": 0.0002, + "loss": 2.0573, + "step": 6312 + }, + { + "epoch": 1.990854619993693, + "grad_norm": 0.875, + "learning_rate": 0.0002, + "loss": 1.8722, + "step": 6313 + }, + { + "epoch": 1.9911699779249448, + "grad_norm": 0.828125, + "learning_rate": 0.0002, + "loss": 1.7223, + "step": 6314 + }, + { + "epoch": 1.9914853358561968, + "grad_norm": 0.96484375, + "learning_rate": 0.0002, + "loss": 2.214, + "step": 6315 + }, + { + "epoch": 1.9918006937874488, + "grad_norm": 1.078125, + "learning_rate": 0.0002, + "loss": 2.2836, + "step": 6316 + }, + { + "epoch": 1.9921160517187007, + "grad_norm": 0.921875, + "learning_rate": 0.0002, + "loss": 1.965, + "step": 6317 + }, + { + "epoch": 1.9924314096499527, + "grad_norm": 1.265625, + "learning_rate": 0.0002, + "loss": 2.8641, + "step": 6318 + }, + { + "epoch": 1.9927467675812047, + "grad_norm": 1.859375, + "learning_rate": 0.0002, + "loss": 1.6513, + "step": 6319 + }, + { + "epoch": 1.9930621255124565, + "grad_norm": 1.1171875, + "learning_rate": 0.0002, + "loss": 2.2315, + "step": 6320 + }, + { + "epoch": 1.9933774834437086, + "grad_norm": 1.1328125, + "learning_rate": 0.0002, + "loss": 2.2914, + "step": 6321 + }, + { + "epoch": 1.9936928413749606, + "grad_norm": 0.2373046875, + "learning_rate": 0.0002, + "loss": 1.0916, + "step": 6322 + }, + { + "epoch": 1.9940081993062124, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.0814, + "step": 6323 + }, + { + "epoch": 1.9943235572374647, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 0.9841, + "step": 6324 + }, + { + "epoch": 1.9946389151687165, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 1.0647, + "step": 6325 + }, + { + "epoch": 1.9949542730999683, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.0997, + "step": 6326 + }, + { + "epoch": 1.9952696310312206, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 1.0395, + "step": 6327 + }, + { + "epoch": 1.9955849889624724, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.0033, + "step": 6328 + }, + { + "epoch": 1.9959003468937244, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 1.076, + "step": 6329 + }, + { + "epoch": 1.9962157048249765, + "grad_norm": 0.3515625, + "learning_rate": 0.0002, + "loss": 1.2326, + "step": 6330 + }, + { + "epoch": 1.9965310627562283, + "grad_norm": 0.3515625, + "learning_rate": 0.0002, + "loss": 1.1393, + "step": 6331 + }, + { + "epoch": 1.9968464206874803, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 0.9842, + "step": 6332 + }, + { + "epoch": 1.9971617786187323, + "grad_norm": 0.431640625, + "learning_rate": 0.0002, + "loss": 1.2578, + "step": 6333 + }, + { + "epoch": 1.9974771365499842, + "grad_norm": 0.498046875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 6334 + }, + { + "epoch": 1.9977924944812362, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 1.3885, + "step": 6335 + }, + { + "epoch": 1.9981078524124882, + "grad_norm": 0.63671875, + "learning_rate": 0.0002, + "loss": 1.6219, + "step": 6336 + }, + { + "epoch": 1.99842321034374, + "grad_norm": 0.734375, + "learning_rate": 0.0002, + "loss": 1.5857, + "step": 6337 + }, + { + "epoch": 1.998738568274992, + "grad_norm": 0.83984375, + "learning_rate": 0.0002, + "loss": 1.8052, + "step": 6338 + }, + { + "epoch": 1.9990539262062441, + "grad_norm": 0.7421875, + "learning_rate": 0.0002, + "loss": 1.7405, + "step": 6339 + }, + { + "epoch": 1.999369284137496, + "grad_norm": 1.1171875, + "learning_rate": 0.0002, + "loss": 2.4317, + "step": 6340 + }, + { + "epoch": 1.9996846420687482, + "grad_norm": 0.99609375, + "learning_rate": 0.0002, + "loss": 2.2286, + "step": 6341 + }, + { + "epoch": 2.0, + "grad_norm": 0.9453125, + "learning_rate": 0.0002, + "loss": 1.7481, + "step": 6342 + }, + { + "epoch": 2.000315357931252, + "grad_norm": 0.169921875, + "learning_rate": 0.0002, + "loss": 0.9776, + "step": 6343 + }, + { + "epoch": 2.000630715862504, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 0.8967, + "step": 6344 + }, + { + "epoch": 2.000946073793756, + "grad_norm": 0.18359375, + "learning_rate": 0.0002, + "loss": 0.9065, + "step": 6345 + }, + { + "epoch": 2.0012614317250077, + "grad_norm": 0.201171875, + "learning_rate": 0.0002, + "loss": 0.8953, + "step": 6346 + }, + { + "epoch": 2.00157678965626, + "grad_norm": 0.201171875, + "learning_rate": 0.0002, + "loss": 0.9768, + "step": 6347 + }, + { + "epoch": 2.0018921475875118, + "grad_norm": 0.2060546875, + "learning_rate": 0.0002, + "loss": 0.9099, + "step": 6348 + }, + { + "epoch": 2.0022075055187636, + "grad_norm": 0.2255859375, + "learning_rate": 0.0002, + "loss": 1.034, + "step": 6349 + }, + { + "epoch": 2.002522863450016, + "grad_norm": 0.224609375, + "learning_rate": 0.0002, + "loss": 0.8615, + "step": 6350 + }, + { + "epoch": 2.0028382213812677, + "grad_norm": 0.2490234375, + "learning_rate": 0.0002, + "loss": 0.8593, + "step": 6351 + }, + { + "epoch": 2.00315357931252, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 0.9498, + "step": 6352 + }, + { + "epoch": 2.0034689372437717, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 0.7871, + "step": 6353 + }, + { + "epoch": 2.0037842951750235, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 1.0166, + "step": 6354 + }, + { + "epoch": 2.004099653106276, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 0.9628, + "step": 6355 + }, + { + "epoch": 2.0044150110375276, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 0.8867, + "step": 6356 + }, + { + "epoch": 2.0047303689687794, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 0.853, + "step": 6357 + }, + { + "epoch": 2.0050457269000317, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 0.7728, + "step": 6358 + }, + { + "epoch": 2.0050457269000317, + "eval_loss": 1.6985900402069092, + "eval_runtime": 148.8807, + "eval_samples_per_second": 6.717, + "eval_steps_per_second": 6.717, + "step": 6358 + }, + { + "epoch": 2.0050457269000317, + "mmlu_eval_accuracy": 0.6021145704468607, + "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, + "mmlu_eval_accuracy_anatomy": 0.5, + "mmlu_eval_accuracy_astronomy": 0.5625, + "mmlu_eval_accuracy_business_ethics": 0.7272727272727273, + "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241, + "mmlu_eval_accuracy_college_biology": 0.8125, + "mmlu_eval_accuracy_college_chemistry": 0.375, + "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, + "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454, + "mmlu_eval_accuracy_college_medicine": 0.5909090909090909, + "mmlu_eval_accuracy_college_physics": 0.5454545454545454, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.5769230769230769, + "mmlu_eval_accuracy_econometrics": 0.5, + "mmlu_eval_accuracy_electrical_engineering": 0.5, + "mmlu_eval_accuracy_elementary_mathematics": 0.4878048780487805, + "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, + "mmlu_eval_accuracy_global_facts": 0.2, + "mmlu_eval_accuracy_high_school_biology": 0.625, + "mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 1.0, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8571428571428571, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.6046511627906976, + "mmlu_eval_accuracy_high_school_mathematics": 0.41379310344827586, + "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769, + "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, + "mmlu_eval_accuracy_high_school_psychology": 0.85, + "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, + "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, + "mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693, + "mmlu_eval_accuracy_human_aging": 0.6956521739130435, + "mmlu_eval_accuracy_human_sexuality": 0.5, + "mmlu_eval_accuracy_international_law": 1.0, + "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, + "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, + "mmlu_eval_accuracy_machine_learning": 0.36363636363636365, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.72, + "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182, + "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279, + "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, + "mmlu_eval_accuracy_moral_scenarios": 0.33, + "mmlu_eval_accuracy_nutrition": 0.696969696969697, + "mmlu_eval_accuracy_philosophy": 0.6764705882352942, + "mmlu_eval_accuracy_prehistory": 0.6857142857142857, + "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644, + "mmlu_eval_accuracy_professional_law": 0.3764705882352941, + "mmlu_eval_accuracy_professional_medicine": 0.7419354838709677, + "mmlu_eval_accuracy_professional_psychology": 0.6521739130434783, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7407407407407407, + "mmlu_eval_accuracy_sociology": 0.9090909090909091, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.5, + "mmlu_eval_accuracy_world_religions": 0.7368421052631579, + "mmlu_loss": 5.823521563307589, + "step": 6358 + }, + { + "epoch": 2.0053610848312835, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 0.8234, + "step": 6359 + }, + { + "epoch": 2.0056764427625353, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 0.773, + "step": 6360 + }, + { + "epoch": 2.0059918006937876, + "grad_norm": 0.404296875, + "learning_rate": 0.0002, + "loss": 0.9671, + "step": 6361 + }, + { + "epoch": 2.0063071586250394, + "grad_norm": 0.404296875, + "learning_rate": 0.0002, + "loss": 1.0573, + "step": 6362 + }, + { + "epoch": 2.006622516556291, + "grad_norm": 0.421875, + "learning_rate": 0.0002, + "loss": 1.0598, + "step": 6363 + }, + { + "epoch": 2.0069378744875435, + "grad_norm": 0.400390625, + "learning_rate": 0.0002, + "loss": 0.8304, + "step": 6364 + }, + { + "epoch": 2.0072532324187953, + "grad_norm": 0.455078125, + "learning_rate": 0.0002, + "loss": 0.8154, + "step": 6365 + }, + { + "epoch": 2.007568590350047, + "grad_norm": 0.46875, + "learning_rate": 0.0002, + "loss": 0.8355, + "step": 6366 + }, + { + "epoch": 2.0078839482812993, + "grad_norm": 0.51953125, + "learning_rate": 0.0002, + "loss": 0.9348, + "step": 6367 + }, + { + "epoch": 2.008199306212551, + "grad_norm": 0.5, + "learning_rate": 0.0002, + "loss": 0.9761, + "step": 6368 + }, + { + "epoch": 2.0085146641438034, + "grad_norm": 0.478515625, + "learning_rate": 0.0002, + "loss": 0.7922, + "step": 6369 + }, + { + "epoch": 2.0088300220750552, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 0.675, + "step": 6370 + }, + { + "epoch": 2.009145380006307, + "grad_norm": 0.64453125, + "learning_rate": 0.0002, + "loss": 0.8432, + "step": 6371 + }, + { + "epoch": 2.0094607379375593, + "grad_norm": 0.498046875, + "learning_rate": 0.0002, + "loss": 0.8609, + "step": 6372 + }, + { + "epoch": 2.009776095868811, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 0.8397, + "step": 6373 + }, + { + "epoch": 2.010091453800063, + "grad_norm": 0.609375, + "learning_rate": 0.0002, + "loss": 0.8863, + "step": 6374 + }, + { + "epoch": 2.010406811731315, + "grad_norm": 0.640625, + "learning_rate": 0.0002, + "loss": 0.9129, + "step": 6375 + }, + { + "epoch": 2.010722169662567, + "grad_norm": 0.6640625, + "learning_rate": 0.0002, + "loss": 0.9527, + "step": 6376 + }, + { + "epoch": 2.011037527593819, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 0.9601, + "step": 6377 + }, + { + "epoch": 2.011352885525071, + "grad_norm": 0.7578125, + "learning_rate": 0.0002, + "loss": 1.0903, + "step": 6378 + }, + { + "epoch": 2.011668243456323, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 1.2385, + "step": 6379 + }, + { + "epoch": 2.0119836013875747, + "grad_norm": 0.75, + "learning_rate": 0.0002, + "loss": 1.1461, + "step": 6380 + }, + { + "epoch": 2.012298959318827, + "grad_norm": 0.96875, + "learning_rate": 0.0002, + "loss": 1.2714, + "step": 6381 + }, + { + "epoch": 2.0126143172500788, + "grad_norm": 0.9765625, + "learning_rate": 0.0002, + "loss": 1.1492, + "step": 6382 + }, + { + "epoch": 2.0129296751813306, + "grad_norm": 0.984375, + "learning_rate": 0.0002, + "loss": 1.2872, + "step": 6383 + }, + { + "epoch": 2.013245033112583, + "grad_norm": 0.91796875, + "learning_rate": 0.0002, + "loss": 1.1215, + "step": 6384 + }, + { + "epoch": 2.0135603910438347, + "grad_norm": 1.4296875, + "learning_rate": 0.0002, + "loss": 1.6578, + "step": 6385 + }, + { + "epoch": 2.013875748975087, + "grad_norm": 1.1796875, + "learning_rate": 0.0002, + "loss": 1.559, + "step": 6386 + }, + { + "epoch": 2.0141911069063387, + "grad_norm": 1.3359375, + "learning_rate": 0.0002, + "loss": 1.6687, + "step": 6387 + }, + { + "epoch": 2.0145064648375905, + "grad_norm": 1.5, + "learning_rate": 0.0002, + "loss": 1.7101, + "step": 6388 + }, + { + "epoch": 2.014821822768843, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 1.646, + "step": 6389 + }, + { + "epoch": 2.0151371807000946, + "grad_norm": 0.96484375, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 6390 + }, + { + "epoch": 2.0154525386313464, + "grad_norm": 0.9921875, + "learning_rate": 0.0002, + "loss": 1.3809, + "step": 6391 + }, + { + "epoch": 2.0157678965625987, + "grad_norm": 1.3125, + "learning_rate": 0.0002, + "loss": 1.6275, + "step": 6392 + }, + { + "epoch": 2.0160832544938505, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 1.0309, + "step": 6393 + }, + { + "epoch": 2.0163986124251023, + "grad_norm": 0.384765625, + "learning_rate": 0.0002, + "loss": 0.9177, + "step": 6394 + }, + { + "epoch": 2.0167139703563546, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 0.8693, + "step": 6395 + }, + { + "epoch": 2.0170293282876064, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 0.9114, + "step": 6396 + }, + { + "epoch": 2.017344686218858, + "grad_norm": 0.2451171875, + "learning_rate": 0.0002, + "loss": 0.9719, + "step": 6397 + }, + { + "epoch": 2.0176600441501105, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 0.8671, + "step": 6398 + }, + { + "epoch": 2.0179754020813623, + "grad_norm": 0.3828125, + "learning_rate": 0.0002, + "loss": 0.9248, + "step": 6399 + }, + { + "epoch": 2.0182907600126145, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 0.9303, + "step": 6400 + }, + { + "epoch": 2.0186061179438664, + "grad_norm": 0.248046875, + "learning_rate": 0.0002, + "loss": 0.8384, + "step": 6401 + }, + { + "epoch": 2.018921475875118, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 0.8506, + "step": 6402 + }, + { + "epoch": 2.0192368338063704, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 0.8859, + "step": 6403 + }, + { + "epoch": 2.0195521917376222, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 0.8659, + "step": 6404 + }, + { + "epoch": 2.019867549668874, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 0.8153, + "step": 6405 + }, + { + "epoch": 2.0201829076001263, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 0.9104, + "step": 6406 + }, + { + "epoch": 2.020498265531378, + "grad_norm": 0.421875, + "learning_rate": 0.0002, + "loss": 0.8777, + "step": 6407 + }, + { + "epoch": 2.02081362346263, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 0.8876, + "step": 6408 + }, + { + "epoch": 2.021128981393882, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 0.8241, + "step": 6409 + }, + { + "epoch": 2.021444339325134, + "grad_norm": 0.373046875, + "learning_rate": 0.0002, + "loss": 0.8879, + "step": 6410 + }, + { + "epoch": 2.021759697256386, + "grad_norm": 0.330078125, + "learning_rate": 0.0002, + "loss": 0.8238, + "step": 6411 + }, + { + "epoch": 2.022075055187638, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 0.8821, + "step": 6412 + }, + { + "epoch": 2.02239041311889, + "grad_norm": 0.3828125, + "learning_rate": 0.0002, + "loss": 0.9695, + "step": 6413 + }, + { + "epoch": 2.0227057710501417, + "grad_norm": 0.515625, + "learning_rate": 0.0002, + "loss": 0.7807, + "step": 6414 + }, + { + "epoch": 2.023021128981394, + "grad_norm": 0.48828125, + "learning_rate": 0.0002, + "loss": 0.9297, + "step": 6415 + }, + { + "epoch": 2.023336486912646, + "grad_norm": 0.431640625, + "learning_rate": 0.0002, + "loss": 0.8089, + "step": 6416 + }, + { + "epoch": 2.023651844843898, + "grad_norm": 0.431640625, + "learning_rate": 0.0002, + "loss": 0.7951, + "step": 6417 + }, + { + "epoch": 2.02396720277515, + "grad_norm": 0.4375, + "learning_rate": 0.0002, + "loss": 0.8385, + "step": 6418 + }, + { + "epoch": 2.0242825607064017, + "grad_norm": 0.51953125, + "learning_rate": 0.0002, + "loss": 0.825, + "step": 6419 + }, + { + "epoch": 2.024597918637654, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 0.6679, + "step": 6420 + }, + { + "epoch": 2.0249132765689057, + "grad_norm": 0.6484375, + "learning_rate": 0.0002, + "loss": 0.8433, + "step": 6421 + }, + { + "epoch": 2.0252286345001576, + "grad_norm": 0.91796875, + "learning_rate": 0.0002, + "loss": 1.0427, + "step": 6422 + }, + { + "epoch": 2.02554399243141, + "grad_norm": 0.6171875, + "learning_rate": 0.0002, + "loss": 0.8756, + "step": 6423 + }, + { + "epoch": 2.0258593503626616, + "grad_norm": 0.734375, + "learning_rate": 0.0002, + "loss": 0.949, + "step": 6424 + }, + { + "epoch": 2.0261747082939134, + "grad_norm": 0.93359375, + "learning_rate": 0.0002, + "loss": 1.0162, + "step": 6425 + }, + { + "epoch": 2.0264900662251657, + "grad_norm": 0.82421875, + "learning_rate": 0.0002, + "loss": 1.0096, + "step": 6426 + }, + { + "epoch": 2.0268054241564175, + "grad_norm": 0.91796875, + "learning_rate": 0.0002, + "loss": 1.0464, + "step": 6427 + }, + { + "epoch": 2.0271207820876693, + "grad_norm": 0.8359375, + "learning_rate": 0.0002, + "loss": 1.0211, + "step": 6428 + }, + { + "epoch": 2.0274361400189216, + "grad_norm": 0.88671875, + "learning_rate": 0.0002, + "loss": 0.9558, + "step": 6429 + }, + { + "epoch": 2.0277514979501734, + "grad_norm": 0.9921875, + "learning_rate": 0.0002, + "loss": 1.2887, + "step": 6430 + }, + { + "epoch": 2.028066855881425, + "grad_norm": 0.8046875, + "learning_rate": 0.0002, + "loss": 1.0403, + "step": 6431 + }, + { + "epoch": 2.0283822138126775, + "grad_norm": 1.015625, + "learning_rate": 0.0002, + "loss": 1.1539, + "step": 6432 + }, + { + "epoch": 2.0286975717439293, + "grad_norm": 1.375, + "learning_rate": 0.0002, + "loss": 1.4023, + "step": 6433 + }, + { + "epoch": 2.0290129296751815, + "grad_norm": 1.265625, + "learning_rate": 0.0002, + "loss": 1.6736, + "step": 6434 + }, + { + "epoch": 2.0293282876064334, + "grad_norm": 1.3828125, + "learning_rate": 0.0002, + "loss": 1.7903, + "step": 6435 + }, + { + "epoch": 2.029643645537685, + "grad_norm": 1.140625, + "learning_rate": 0.0002, + "loss": 1.4265, + "step": 6436 + }, + { + "epoch": 2.0299590034689374, + "grad_norm": 1.1171875, + "learning_rate": 0.0002, + "loss": 1.4683, + "step": 6437 + }, + { + "epoch": 2.0302743614001892, + "grad_norm": 1.1875, + "learning_rate": 0.0002, + "loss": 1.4892, + "step": 6438 + }, + { + "epoch": 2.030589719331441, + "grad_norm": 0.9921875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 6439 + }, + { + "epoch": 2.0309050772626933, + "grad_norm": 0.9765625, + "learning_rate": 0.0002, + "loss": 1.4545, + "step": 6440 + }, + { + "epoch": 2.031220435193945, + "grad_norm": 0.93359375, + "learning_rate": 0.0002, + "loss": 1.3534, + "step": 6441 + }, + { + "epoch": 2.031535793125197, + "grad_norm": 1.5703125, + "learning_rate": 0.0002, + "loss": 1.869, + "step": 6442 + }, + { + "epoch": 2.031851151056449, + "grad_norm": 0.484375, + "learning_rate": 0.0002, + "loss": 0.8976, + "step": 6443 + }, + { + "epoch": 2.032166508987701, + "grad_norm": 0.466796875, + "learning_rate": 0.0002, + "loss": 1.0353, + "step": 6444 + }, + { + "epoch": 2.032481866918953, + "grad_norm": 0.388671875, + "learning_rate": 0.0002, + "loss": 1.0308, + "step": 6445 + }, + { + "epoch": 2.032797224850205, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 0.8112, + "step": 6446 + }, + { + "epoch": 2.033112582781457, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 0.9416, + "step": 6447 + }, + { + "epoch": 2.0334279407127087, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.0332, + "step": 6448 + }, + { + "epoch": 2.033743298643961, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 0.9228, + "step": 6449 + }, + { + "epoch": 2.034058656575213, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 0.8832, + "step": 6450 + }, + { + "epoch": 2.034374014506465, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 0.8815, + "step": 6451 + }, + { + "epoch": 2.034689372437717, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 1.0127, + "step": 6452 + }, + { + "epoch": 2.0350047303689687, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 0.9204, + "step": 6453 + }, + { + "epoch": 2.035320088300221, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.8444, + "step": 6454 + }, + { + "epoch": 2.0356354462314727, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 0.9197, + "step": 6455 + }, + { + "epoch": 2.0359508041627246, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 0.945, + "step": 6456 + }, + { + "epoch": 2.036266162093977, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 0.8449, + "step": 6457 + }, + { + "epoch": 2.0365815200252286, + "grad_norm": 0.3515625, + "learning_rate": 0.0002, + "loss": 0.8798, + "step": 6458 + }, + { + "epoch": 2.0368968779564804, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 0.9139, + "step": 6459 + }, + { + "epoch": 2.0372122358877327, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 0.7836, + "step": 6460 + }, + { + "epoch": 2.0375275938189845, + "grad_norm": 0.4140625, + "learning_rate": 0.0002, + "loss": 0.8101, + "step": 6461 + }, + { + "epoch": 2.0378429517502363, + "grad_norm": 0.4140625, + "learning_rate": 0.0002, + "loss": 0.8181, + "step": 6462 + }, + { + "epoch": 2.0381583096814886, + "grad_norm": 0.486328125, + "learning_rate": 0.0002, + "loss": 0.8841, + "step": 6463 + }, + { + "epoch": 2.0384736676127404, + "grad_norm": 0.419921875, + "learning_rate": 0.0002, + "loss": 0.9581, + "step": 6464 + }, + { + "epoch": 2.038789025543992, + "grad_norm": 0.3984375, + "learning_rate": 0.0002, + "loss": 0.6824, + "step": 6465 + }, + { + "epoch": 2.0391043834752445, + "grad_norm": 0.4453125, + "learning_rate": 0.0002, + "loss": 0.766, + "step": 6466 + }, + { + "epoch": 2.0394197414064963, + "grad_norm": 0.490234375, + "learning_rate": 0.0002, + "loss": 0.7583, + "step": 6467 + }, + { + "epoch": 2.0397350993377485, + "grad_norm": 0.4296875, + "learning_rate": 0.0002, + "loss": 0.7413, + "step": 6468 + }, + { + "epoch": 2.0400504572690004, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 0.9912, + "step": 6469 + }, + { + "epoch": 2.040365815200252, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 0.9059, + "step": 6470 + }, + { + "epoch": 2.0406811731315044, + "grad_norm": 0.68359375, + "learning_rate": 0.0002, + "loss": 0.7709, + "step": 6471 + }, + { + "epoch": 2.0409965310627562, + "grad_norm": 0.65625, + "learning_rate": 0.0002, + "loss": 0.7914, + "step": 6472 + }, + { + "epoch": 2.041311888994008, + "grad_norm": 0.671875, + "learning_rate": 0.0002, + "loss": 0.7792, + "step": 6473 + }, + { + "epoch": 2.0416272469252603, + "grad_norm": 0.6484375, + "learning_rate": 0.0002, + "loss": 0.8679, + "step": 6474 + }, + { + "epoch": 2.041942604856512, + "grad_norm": 0.77734375, + "learning_rate": 0.0002, + "loss": 1.0212, + "step": 6475 + }, + { + "epoch": 2.042257962787764, + "grad_norm": 0.640625, + "learning_rate": 0.0002, + "loss": 0.8398, + "step": 6476 + }, + { + "epoch": 2.042573320719016, + "grad_norm": 0.69921875, + "learning_rate": 0.0002, + "loss": 0.901, + "step": 6477 + }, + { + "epoch": 2.042888678650268, + "grad_norm": 0.73828125, + "learning_rate": 0.0002, + "loss": 1.149, + "step": 6478 + }, + { + "epoch": 2.04320403658152, + "grad_norm": 1.140625, + "learning_rate": 0.0002, + "loss": 1.0287, + "step": 6479 + }, + { + "epoch": 2.043519394512772, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 1.1539, + "step": 6480 + }, + { + "epoch": 2.043834752444024, + "grad_norm": 1.03125, + "learning_rate": 0.0002, + "loss": 1.1386, + "step": 6481 + }, + { + "epoch": 2.044150110375276, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 1.2774, + "step": 6482 + }, + { + "epoch": 2.044465468306528, + "grad_norm": 1.1328125, + "learning_rate": 0.0002, + "loss": 1.5076, + "step": 6483 + }, + { + "epoch": 2.04478082623778, + "grad_norm": 1.578125, + "learning_rate": 0.0002, + "loss": 1.4448, + "step": 6484 + }, + { + "epoch": 2.045096184169032, + "grad_norm": 1.5546875, + "learning_rate": 0.0002, + "loss": 1.5214, + "step": 6485 + }, + { + "epoch": 2.045411542100284, + "grad_norm": 1.671875, + "learning_rate": 0.0002, + "loss": 1.6178, + "step": 6486 + }, + { + "epoch": 2.0457269000315357, + "grad_norm": 1.078125, + "learning_rate": 0.0002, + "loss": 1.5895, + "step": 6487 + }, + { + "epoch": 2.046042257962788, + "grad_norm": 1.28125, + "learning_rate": 0.0002, + "loss": 1.3606, + "step": 6488 + }, + { + "epoch": 2.0463576158940397, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 1.443, + "step": 6489 + }, + { + "epoch": 2.0466729738252916, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 1.6067, + "step": 6490 + }, + { + "epoch": 2.046988331756544, + "grad_norm": 1.0, + "learning_rate": 0.0002, + "loss": 1.2672, + "step": 6491 + }, + { + "epoch": 2.0473036896877956, + "grad_norm": 0.95703125, + "learning_rate": 0.0002, + "loss": 1.487, + "step": 6492 + }, + { + "epoch": 2.0476190476190474, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.039, + "step": 6493 + }, + { + "epoch": 2.0479344055502997, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 0.9131, + "step": 6494 + }, + { + "epoch": 2.0482497634815515, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 0.8788, + "step": 6495 + }, + { + "epoch": 2.0485651214128033, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 0.8427, + "step": 6496 + }, + { + "epoch": 2.0488804793440556, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 0.8377, + "step": 6497 + }, + { + "epoch": 2.0491958372753074, + "grad_norm": 0.2490234375, + "learning_rate": 0.0002, + "loss": 0.8904, + "step": 6498 + }, + { + "epoch": 2.049511195206559, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 0.9885, + "step": 6499 + }, + { + "epoch": 2.0498265531378115, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 0.9463, + "step": 6500 + }, + { + "epoch": 2.0501419110690633, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 0.8414, + "step": 6501 + }, + { + "epoch": 2.0504572690003156, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 0.8191, + "step": 6502 + }, + { + "epoch": 2.0507726269315674, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 0.8541, + "step": 6503 + }, + { + "epoch": 2.051087984862819, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 0.7969, + "step": 6504 + }, + { + "epoch": 2.0514033427940714, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.0567, + "step": 6505 + }, + { + "epoch": 2.0517187007253233, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 0.8494, + "step": 6506 + }, + { + "epoch": 2.052034058656575, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 0.7732, + "step": 6507 + }, + { + "epoch": 2.0523494165878273, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.8602, + "step": 6508 + }, + { + "epoch": 2.052664774519079, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 0.8651, + "step": 6509 + }, + { + "epoch": 2.052980132450331, + "grad_norm": 0.380859375, + "learning_rate": 0.0002, + "loss": 0.8884, + "step": 6510 + }, + { + "epoch": 2.053295490381583, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 0.9221, + "step": 6511 + }, + { + "epoch": 2.053610848312835, + "grad_norm": 0.3671875, + "learning_rate": 0.0002, + "loss": 0.8655, + "step": 6512 + }, + { + "epoch": 2.053926206244087, + "grad_norm": 0.37890625, + "learning_rate": 0.0002, + "loss": 0.7997, + "step": 6513 + }, + { + "epoch": 2.054241564175339, + "grad_norm": 0.421875, + "learning_rate": 0.0002, + "loss": 0.929, + "step": 6514 + }, + { + "epoch": 2.054556922106591, + "grad_norm": 0.427734375, + "learning_rate": 0.0002, + "loss": 0.9507, + "step": 6515 + }, + { + "epoch": 2.054872280037843, + "grad_norm": 0.72265625, + "learning_rate": 0.0002, + "loss": 0.9764, + "step": 6516 + }, + { + "epoch": 2.055187637969095, + "grad_norm": 0.498046875, + "learning_rate": 0.0002, + "loss": 0.7683, + "step": 6517 + }, + { + "epoch": 2.055502995900347, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 0.9399, + "step": 6518 + }, + { + "epoch": 2.055818353831599, + "grad_norm": 0.46875, + "learning_rate": 0.0002, + "loss": 0.801, + "step": 6519 + }, + { + "epoch": 2.056133711762851, + "grad_norm": 0.62890625, + "learning_rate": 0.0002, + "loss": 0.8971, + "step": 6520 + }, + { + "epoch": 2.0564490696941027, + "grad_norm": 0.625, + "learning_rate": 0.0002, + "loss": 0.8238, + "step": 6521 + }, + { + "epoch": 2.056764427625355, + "grad_norm": 0.63671875, + "learning_rate": 0.0002, + "loss": 0.8786, + "step": 6522 + }, + { + "epoch": 2.0570797855566068, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 0.9909, + "step": 6523 + }, + { + "epoch": 2.0573951434878586, + "grad_norm": 0.83984375, + "learning_rate": 0.0002, + "loss": 1.1071, + "step": 6524 + }, + { + "epoch": 2.057710501419111, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 1.0933, + "step": 6525 + }, + { + "epoch": 2.0580258593503626, + "grad_norm": 0.71875, + "learning_rate": 0.0002, + "loss": 0.959, + "step": 6526 + }, + { + "epoch": 2.0583412172816145, + "grad_norm": 0.7109375, + "learning_rate": 0.0002, + "loss": 0.9804, + "step": 6527 + }, + { + "epoch": 2.0586565752128667, + "grad_norm": 0.89453125, + "learning_rate": 0.0002, + "loss": 1.0935, + "step": 6528 + }, + { + "epoch": 2.0589719331441185, + "grad_norm": 0.875, + "learning_rate": 0.0002, + "loss": 1.1069, + "step": 6529 + }, + { + "epoch": 2.0592872910753703, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 1.1968, + "step": 6530 + }, + { + "epoch": 2.0596026490066226, + "grad_norm": 1.171875, + "learning_rate": 0.0002, + "loss": 1.4076, + "step": 6531 + }, + { + "epoch": 2.0599180069378744, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 6532 + }, + { + "epoch": 2.0602333648691267, + "grad_norm": 1.078125, + "learning_rate": 0.0002, + "loss": 1.2462, + "step": 6533 + }, + { + "epoch": 2.0605487228003785, + "grad_norm": 1.015625, + "learning_rate": 0.0002, + "loss": 1.3529, + "step": 6534 + }, + { + "epoch": 2.0608640807316303, + "grad_norm": 1.109375, + "learning_rate": 0.0002, + "loss": 1.6657, + "step": 6535 + }, + { + "epoch": 2.0611794386628826, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 1.4573, + "step": 6536 + }, + { + "epoch": 2.0614947965941344, + "grad_norm": 1.140625, + "learning_rate": 0.0002, + "loss": 1.4683, + "step": 6537 + }, + { + "epoch": 2.061810154525386, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.6071, + "step": 6538 + }, + { + "epoch": 2.0621255124566384, + "grad_norm": 1.2421875, + "learning_rate": 0.0002, + "loss": 1.5303, + "step": 6539 + }, + { + "epoch": 2.0624408703878903, + "grad_norm": 1.5234375, + "learning_rate": 0.0002, + "loss": 1.1805, + "step": 6540 + }, + { + "epoch": 2.062756228319142, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 1.3838, + "step": 6541 + }, + { + "epoch": 2.0630715862503943, + "grad_norm": 1.296875, + "learning_rate": 0.0002, + "loss": 1.411, + "step": 6542 + }, + { + "epoch": 2.063386944181646, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 0.9522, + "step": 6543 + }, + { + "epoch": 2.063702302112898, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 0.9784, + "step": 6544 + }, + { + "epoch": 2.06401766004415, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 0.8759, + "step": 6545 + }, + { + "epoch": 2.06401766004415, + "eval_loss": 1.8574892282485962, + "eval_runtime": 148.9489, + "eval_samples_per_second": 6.714, + "eval_steps_per_second": 6.714, + "step": 6545 + }, + { + "epoch": 2.06401766004415, + "mmlu_eval_accuracy": 0.5972302303899769, + "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, + "mmlu_eval_accuracy_anatomy": 0.42857142857142855, + "mmlu_eval_accuracy_astronomy": 0.625, + "mmlu_eval_accuracy_business_ethics": 0.7272727272727273, + "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138, + "mmlu_eval_accuracy_college_biology": 0.875, + "mmlu_eval_accuracy_college_chemistry": 0.375, + "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365, + "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454, + "mmlu_eval_accuracy_college_medicine": 0.5909090909090909, + "mmlu_eval_accuracy_college_physics": 0.45454545454545453, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384, + "mmlu_eval_accuracy_econometrics": 0.5, + "mmlu_eval_accuracy_electrical_engineering": 0.375, + "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634, + "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, + "mmlu_eval_accuracy_global_facts": 0.3, + "mmlu_eval_accuracy_high_school_biology": 0.625, + "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778, + "mmlu_eval_accuracy_high_school_geography": 0.9545454545454546, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8571428571428571, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209, + "mmlu_eval_accuracy_high_school_mathematics": 0.3793103448275862, + "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, + "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413, + "mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, + "mmlu_eval_accuracy_high_school_us_history": 0.8636363636363636, + "mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693, + "mmlu_eval_accuracy_human_aging": 0.6086956521739131, + "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334, + "mmlu_eval_accuracy_international_law": 0.9230769230769231, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.36363636363636365, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.72, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279, + "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, + "mmlu_eval_accuracy_moral_scenarios": 0.32, + "mmlu_eval_accuracy_nutrition": 0.7878787878787878, + "mmlu_eval_accuracy_philosophy": 0.6470588235294118, + "mmlu_eval_accuracy_prehistory": 0.6285714285714286, + "mmlu_eval_accuracy_professional_accounting": 0.5161290322580645, + "mmlu_eval_accuracy_professional_law": 0.3941176470588235, + "mmlu_eval_accuracy_professional_medicine": 0.7096774193548387, + "mmlu_eval_accuracy_professional_psychology": 0.6231884057971014, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7037037037037037, + "mmlu_eval_accuracy_sociology": 0.9090909090909091, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.5555555555555556, + "mmlu_eval_accuracy_world_religions": 0.7368421052631579, + "mmlu_loss": 5.8482463436139165, + "step": 6545 + }, + { + "epoch": 2.064333017975402, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 0.9325, + "step": 6546 + }, + { + "epoch": 2.064648375906654, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 0.8478, + "step": 6547 + }, + { + "epoch": 2.064963733837906, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 1.0569, + "step": 6548 + }, + { + "epoch": 2.065279091769158, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 0.9769, + "step": 6549 + }, + { + "epoch": 2.06559444970041, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 0.8298, + "step": 6550 + }, + { + "epoch": 2.065909807631662, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 0.9474, + "step": 6551 + }, + { + "epoch": 2.066225165562914, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 0.7823, + "step": 6552 + }, + { + "epoch": 2.066540523494166, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 0.9405, + "step": 6553 + }, + { + "epoch": 2.066855881425418, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 0.8804, + "step": 6554 + }, + { + "epoch": 2.0671712393566697, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 0.857, + "step": 6555 + }, + { + "epoch": 2.067486597287922, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 0.8712, + "step": 6556 + }, + { + "epoch": 2.0678019552191738, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 0.8479, + "step": 6557 + }, + { + "epoch": 2.0681173131504256, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 0.7865, + "step": 6558 + }, + { + "epoch": 2.068432671081678, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 0.8953, + "step": 6559 + }, + { + "epoch": 2.0687480290129296, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 0.8487, + "step": 6560 + }, + { + "epoch": 2.0690633869441815, + "grad_norm": 0.404296875, + "learning_rate": 0.0002, + "loss": 0.8227, + "step": 6561 + }, + { + "epoch": 2.0693787448754337, + "grad_norm": 0.40234375, + "learning_rate": 0.0002, + "loss": 0.8179, + "step": 6562 + }, + { + "epoch": 2.0696941028066855, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 0.7743, + "step": 6563 + }, + { + "epoch": 2.0700094607379373, + "grad_norm": 0.390625, + "learning_rate": 0.0002, + "loss": 0.885, + "step": 6564 + }, + { + "epoch": 2.0703248186691896, + "grad_norm": 0.435546875, + "learning_rate": 0.0002, + "loss": 0.8802, + "step": 6565 + }, + { + "epoch": 2.0706401766004414, + "grad_norm": 0.42578125, + "learning_rate": 0.0002, + "loss": 0.7598, + "step": 6566 + }, + { + "epoch": 2.0709555345316937, + "grad_norm": 0.453125, + "learning_rate": 0.0002, + "loss": 0.748, + "step": 6567 + }, + { + "epoch": 2.0712708924629455, + "grad_norm": 0.5546875, + "learning_rate": 0.0002, + "loss": 0.9171, + "step": 6568 + }, + { + "epoch": 2.0715862503941973, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 0.8326, + "step": 6569 + }, + { + "epoch": 2.0719016083254496, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 0.8436, + "step": 6570 + }, + { + "epoch": 2.0722169662567014, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 0.8525, + "step": 6571 + }, + { + "epoch": 2.072532324187953, + "grad_norm": 0.6796875, + "learning_rate": 0.0002, + "loss": 0.8631, + "step": 6572 + }, + { + "epoch": 2.0728476821192054, + "grad_norm": 0.62890625, + "learning_rate": 0.0002, + "loss": 0.7634, + "step": 6573 + }, + { + "epoch": 2.0731630400504573, + "grad_norm": 0.69921875, + "learning_rate": 0.0002, + "loss": 0.9555, + "step": 6574 + }, + { + "epoch": 2.073478397981709, + "grad_norm": 0.86328125, + "learning_rate": 0.0002, + "loss": 1.0064, + "step": 6575 + }, + { + "epoch": 2.0737937559129613, + "grad_norm": 0.92578125, + "learning_rate": 0.0002, + "loss": 1.4846, + "step": 6576 + }, + { + "epoch": 2.074109113844213, + "grad_norm": 0.88671875, + "learning_rate": 0.0002, + "loss": 1.1653, + "step": 6577 + }, + { + "epoch": 2.074424471775465, + "grad_norm": 0.9296875, + "learning_rate": 0.0002, + "loss": 1.2657, + "step": 6578 + }, + { + "epoch": 2.074739829706717, + "grad_norm": 0.98828125, + "learning_rate": 0.0002, + "loss": 1.1027, + "step": 6579 + }, + { + "epoch": 2.075055187637969, + "grad_norm": 0.89453125, + "learning_rate": 0.0002, + "loss": 1.2456, + "step": 6580 + }, + { + "epoch": 2.0753705455692213, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 6581 + }, + { + "epoch": 2.075685903500473, + "grad_norm": 1.0859375, + "learning_rate": 0.0002, + "loss": 1.2409, + "step": 6582 + }, + { + "epoch": 2.076001261431725, + "grad_norm": 1.40625, + "learning_rate": 0.0002, + "loss": 1.6232, + "step": 6583 + }, + { + "epoch": 2.076316619362977, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 1.5345, + "step": 6584 + }, + { + "epoch": 2.076631977294229, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 1.3976, + "step": 6585 + }, + { + "epoch": 2.076947335225481, + "grad_norm": 1.078125, + "learning_rate": 0.0002, + "loss": 1.6009, + "step": 6586 + }, + { + "epoch": 2.077262693156733, + "grad_norm": 1.1953125, + "learning_rate": 0.0002, + "loss": 1.7146, + "step": 6587 + }, + { + "epoch": 2.077578051087985, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 1.5898, + "step": 6588 + }, + { + "epoch": 2.0778934090192367, + "grad_norm": 1.0, + "learning_rate": 0.0002, + "loss": 1.3509, + "step": 6589 + }, + { + "epoch": 2.078208766950489, + "grad_norm": 0.82421875, + "learning_rate": 0.0002, + "loss": 1.246, + "step": 6590 + }, + { + "epoch": 2.0785241248817408, + "grad_norm": 1.8125, + "learning_rate": 0.0002, + "loss": 1.8451, + "step": 6591 + }, + { + "epoch": 2.0788394828129926, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 1.7015, + "step": 6592 + }, + { + "epoch": 2.079154840744245, + "grad_norm": 0.2421875, + "learning_rate": 0.0002, + "loss": 0.8625, + "step": 6593 + }, + { + "epoch": 2.0794701986754967, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 1.0498, + "step": 6594 + }, + { + "epoch": 2.0797855566067485, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 1.0132, + "step": 6595 + }, + { + "epoch": 2.0801009145380007, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 0.956, + "step": 6596 + }, + { + "epoch": 2.0804162724692525, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 0.8351, + "step": 6597 + }, + { + "epoch": 2.0807316304005044, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 0.9819, + "step": 6598 + }, + { + "epoch": 2.0810469883317566, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 0.9556, + "step": 6599 + }, + { + "epoch": 2.0813623462630084, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 1.0227, + "step": 6600 + }, + { + "epoch": 2.0816777041942607, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 0.9166, + "step": 6601 + }, + { + "epoch": 2.0819930621255125, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 0.9335, + "step": 6602 + }, + { + "epoch": 2.0823084200567643, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 0.8824, + "step": 6603 + }, + { + "epoch": 2.0826237779880166, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 0.952, + "step": 6604 + }, + { + "epoch": 2.0829391359192684, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 0.9357, + "step": 6605 + }, + { + "epoch": 2.08325449385052, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 0.9357, + "step": 6606 + }, + { + "epoch": 2.0835698517817725, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 0.7782, + "step": 6607 + }, + { + "epoch": 2.0838852097130243, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.8366, + "step": 6608 + }, + { + "epoch": 2.084200567644276, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 0.843, + "step": 6609 + }, + { + "epoch": 2.0845159255755283, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 0.8525, + "step": 6610 + }, + { + "epoch": 2.08483128350678, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 0.8414, + "step": 6611 + }, + { + "epoch": 2.085146641438032, + "grad_norm": 0.388671875, + "learning_rate": 0.0002, + "loss": 0.9011, + "step": 6612 + }, + { + "epoch": 2.0854619993692842, + "grad_norm": 0.390625, + "learning_rate": 0.0002, + "loss": 0.8027, + "step": 6613 + }, + { + "epoch": 2.085777357300536, + "grad_norm": 0.4609375, + "learning_rate": 0.0002, + "loss": 0.8795, + "step": 6614 + }, + { + "epoch": 2.0860927152317883, + "grad_norm": 0.482421875, + "learning_rate": 0.0002, + "loss": 0.8477, + "step": 6615 + }, + { + "epoch": 2.08640807316304, + "grad_norm": 0.47265625, + "learning_rate": 0.0002, + "loss": 0.8492, + "step": 6616 + }, + { + "epoch": 2.086723431094292, + "grad_norm": 0.4921875, + "learning_rate": 0.0002, + "loss": 0.8522, + "step": 6617 + }, + { + "epoch": 2.087038789025544, + "grad_norm": 0.482421875, + "learning_rate": 0.0002, + "loss": 0.8907, + "step": 6618 + }, + { + "epoch": 2.087354146956796, + "grad_norm": 0.5703125, + "learning_rate": 0.0002, + "loss": 0.8347, + "step": 6619 + }, + { + "epoch": 2.087669504888048, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 0.8287, + "step": 6620 + }, + { + "epoch": 2.0879848628193, + "grad_norm": 0.5703125, + "learning_rate": 0.0002, + "loss": 0.8275, + "step": 6621 + }, + { + "epoch": 2.088300220750552, + "grad_norm": 0.625, + "learning_rate": 0.0002, + "loss": 0.8225, + "step": 6622 + }, + { + "epoch": 2.0886155786818037, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 0.9205, + "step": 6623 + }, + { + "epoch": 2.088930936613056, + "grad_norm": 0.83984375, + "learning_rate": 0.0002, + "loss": 1.0764, + "step": 6624 + }, + { + "epoch": 2.0892462945443078, + "grad_norm": 0.8203125, + "learning_rate": 0.0002, + "loss": 1.0865, + "step": 6625 + }, + { + "epoch": 2.0895616524755596, + "grad_norm": 0.9921875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 6626 + }, + { + "epoch": 2.089877010406812, + "grad_norm": 1.109375, + "learning_rate": 0.0002, + "loss": 1.0362, + "step": 6627 + }, + { + "epoch": 2.0901923683380637, + "grad_norm": 0.88671875, + "learning_rate": 0.0002, + "loss": 1.2938, + "step": 6628 + }, + { + "epoch": 2.0905077262693155, + "grad_norm": 0.8046875, + "learning_rate": 0.0002, + "loss": 1.2017, + "step": 6629 + }, + { + "epoch": 2.0908230842005677, + "grad_norm": 0.9765625, + "learning_rate": 0.0002, + "loss": 1.2974, + "step": 6630 + }, + { + "epoch": 2.0911384421318195, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 1.5406, + "step": 6631 + }, + { + "epoch": 2.091453800063072, + "grad_norm": 0.984375, + "learning_rate": 0.0002, + "loss": 1.4449, + "step": 6632 + }, + { + "epoch": 2.0917691579943236, + "grad_norm": 1.1171875, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 6633 + }, + { + "epoch": 2.0920845159255754, + "grad_norm": 1.375, + "learning_rate": 0.0002, + "loss": 1.4646, + "step": 6634 + }, + { + "epoch": 2.0923998738568277, + "grad_norm": 1.703125, + "learning_rate": 0.0002, + "loss": 1.9666, + "step": 6635 + }, + { + "epoch": 2.0927152317880795, + "grad_norm": 0.953125, + "learning_rate": 0.0002, + "loss": 1.5642, + "step": 6636 + }, + { + "epoch": 2.0930305897193313, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 1.5944, + "step": 6637 + }, + { + "epoch": 2.0933459476505836, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 1.5381, + "step": 6638 + }, + { + "epoch": 2.0936613055818354, + "grad_norm": 1.0390625, + "learning_rate": 0.0002, + "loss": 1.6648, + "step": 6639 + }, + { + "epoch": 2.093976663513087, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.6244, + "step": 6640 + }, + { + "epoch": 2.0942920214443395, + "grad_norm": 0.8515625, + "learning_rate": 0.0002, + "loss": 1.3855, + "step": 6641 + }, + { + "epoch": 2.0946073793755913, + "grad_norm": 1.2734375, + "learning_rate": 0.0002, + "loss": 1.533, + "step": 6642 + }, + { + "epoch": 2.094922737306843, + "grad_norm": 0.388671875, + "learning_rate": 0.0002, + "loss": 1.03, + "step": 6643 + }, + { + "epoch": 2.0952380952380953, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.0473, + "step": 6644 + }, + { + "epoch": 2.095553453169347, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 1.0732, + "step": 6645 + }, + { + "epoch": 2.095868811100599, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 0.8681, + "step": 6646 + }, + { + "epoch": 2.0961841690318512, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 0.9506, + "step": 6647 + }, + { + "epoch": 2.096499526963103, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 0.8312, + "step": 6648 + }, + { + "epoch": 2.0968148848943553, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 0.9742, + "step": 6649 + }, + { + "epoch": 2.097130242825607, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 0.9502, + "step": 6650 + }, + { + "epoch": 2.097445600756859, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 0.8247, + "step": 6651 + }, + { + "epoch": 2.097760958688111, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.8666, + "step": 6652 + }, + { + "epoch": 2.098076316619363, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 0.8013, + "step": 6653 + }, + { + "epoch": 2.098391674550615, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 0.8467, + "step": 6654 + }, + { + "epoch": 2.098707032481867, + "grad_norm": 0.361328125, + "learning_rate": 0.0002, + "loss": 0.8622, + "step": 6655 + }, + { + "epoch": 2.099022390413119, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 0.7412, + "step": 6656 + }, + { + "epoch": 2.0993377483443707, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 0.916, + "step": 6657 + }, + { + "epoch": 2.099653106275623, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 0.9271, + "step": 6658 + }, + { + "epoch": 2.0999684642068748, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.6721, + "step": 6659 + }, + { + "epoch": 2.1002838221381266, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 0.7966, + "step": 6660 + }, + { + "epoch": 2.100599180069379, + "grad_norm": 0.380859375, + "learning_rate": 0.0002, + "loss": 0.7564, + "step": 6661 + }, + { + "epoch": 2.1009145380006307, + "grad_norm": 0.37890625, + "learning_rate": 0.0002, + "loss": 0.9129, + "step": 6662 + }, + { + "epoch": 2.1012298959318825, + "grad_norm": 0.478515625, + "learning_rate": 0.0002, + "loss": 1.0456, + "step": 6663 + }, + { + "epoch": 2.1015452538631347, + "grad_norm": 0.48046875, + "learning_rate": 0.0002, + "loss": 0.8932, + "step": 6664 + }, + { + "epoch": 2.1018606117943865, + "grad_norm": 0.3984375, + "learning_rate": 0.0002, + "loss": 0.707, + "step": 6665 + }, + { + "epoch": 2.102175969725639, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 0.8679, + "step": 6666 + }, + { + "epoch": 2.1024913276568906, + "grad_norm": 0.5546875, + "learning_rate": 0.0002, + "loss": 0.9904, + "step": 6667 + }, + { + "epoch": 2.1028066855881424, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 0.7822, + "step": 6668 + }, + { + "epoch": 2.1031220435193947, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 0.8161, + "step": 6669 + }, + { + "epoch": 2.1034374014506465, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 0.842, + "step": 6670 + }, + { + "epoch": 2.1037527593818983, + "grad_norm": 0.71875, + "learning_rate": 0.0002, + "loss": 0.7779, + "step": 6671 + }, + { + "epoch": 2.1040681173131506, + "grad_norm": 0.7265625, + "learning_rate": 0.0002, + "loss": 0.935, + "step": 6672 + }, + { + "epoch": 2.1043834752444024, + "grad_norm": 0.80859375, + "learning_rate": 0.0002, + "loss": 1.0542, + "step": 6673 + }, + { + "epoch": 2.104698833175654, + "grad_norm": 0.7265625, + "learning_rate": 0.0002, + "loss": 0.9051, + "step": 6674 + }, + { + "epoch": 2.1050141911069065, + "grad_norm": 0.82421875, + "learning_rate": 0.0002, + "loss": 1.1569, + "step": 6675 + }, + { + "epoch": 2.1053295490381583, + "grad_norm": 0.8515625, + "learning_rate": 0.0002, + "loss": 0.8938, + "step": 6676 + }, + { + "epoch": 2.10564490696941, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 1.4579, + "step": 6677 + }, + { + "epoch": 2.1059602649006623, + "grad_norm": 0.92578125, + "learning_rate": 0.0002, + "loss": 1.1438, + "step": 6678 + }, + { + "epoch": 2.106275622831914, + "grad_norm": 0.93359375, + "learning_rate": 0.0002, + "loss": 1.2305, + "step": 6679 + }, + { + "epoch": 2.1065909807631664, + "grad_norm": 0.8828125, + "learning_rate": 0.0002, + "loss": 1.2609, + "step": 6680 + }, + { + "epoch": 2.1069063386944182, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 1.372, + "step": 6681 + }, + { + "epoch": 2.10722169662567, + "grad_norm": 1.5, + "learning_rate": 0.0002, + "loss": 1.4765, + "step": 6682 + }, + { + "epoch": 2.1075370545569223, + "grad_norm": 1.140625, + "learning_rate": 0.0002, + "loss": 1.5945, + "step": 6683 + }, + { + "epoch": 2.107852412488174, + "grad_norm": 0.984375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 6684 + }, + { + "epoch": 2.108167770419426, + "grad_norm": 1.28125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 6685 + }, + { + "epoch": 2.108483128350678, + "grad_norm": 1.078125, + "learning_rate": 0.0002, + "loss": 1.4746, + "step": 6686 + }, + { + "epoch": 2.10879848628193, + "grad_norm": 1.2734375, + "learning_rate": 0.0002, + "loss": 1.5918, + "step": 6687 + }, + { + "epoch": 2.109113844213182, + "grad_norm": 1.1328125, + "learning_rate": 0.0002, + "loss": 1.3612, + "step": 6688 + }, + { + "epoch": 2.109429202144434, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6689 + }, + { + "epoch": 2.109744560075686, + "grad_norm": 1.5390625, + "learning_rate": 0.0002, + "loss": 1.4182, + "step": 6690 + }, + { + "epoch": 2.1100599180069377, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.3552, + "step": 6691 + }, + { + "epoch": 2.11037527593819, + "grad_norm": 1.0390625, + "learning_rate": 0.0002, + "loss": 1.6094, + "step": 6692 + }, + { + "epoch": 2.110690633869442, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 1.002, + "step": 6693 + }, + { + "epoch": 2.1110059918006936, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 0.9687, + "step": 6694 + }, + { + "epoch": 2.111321349731946, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 0.9765, + "step": 6695 + }, + { + "epoch": 2.1116367076631977, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.001, + "step": 6696 + }, + { + "epoch": 2.1119520655944495, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.0844, + "step": 6697 + }, + { + "epoch": 2.1122674235257017, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 0.8777, + "step": 6698 + }, + { + "epoch": 2.1125827814569536, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 1.0233, + "step": 6699 + }, + { + "epoch": 2.112898139388206, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 0.8763, + "step": 6700 + }, + { + "epoch": 2.1132134973194576, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 0.8302, + "step": 6701 + }, + { + "epoch": 2.1135288552507094, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 0.8893, + "step": 6702 + }, + { + "epoch": 2.1138442131819617, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 0.8769, + "step": 6703 + }, + { + "epoch": 2.1141595711132135, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 0.8434, + "step": 6704 + }, + { + "epoch": 2.1144749290444653, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.9002, + "step": 6705 + }, + { + "epoch": 2.1147902869757176, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.9764, + "step": 6706 + }, + { + "epoch": 2.1151056449069694, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 1.0075, + "step": 6707 + }, + { + "epoch": 2.115421002838221, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 0.9324, + "step": 6708 + }, + { + "epoch": 2.1157363607694735, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 0.9909, + "step": 6709 + }, + { + "epoch": 2.1160517187007253, + "grad_norm": 0.40234375, + "learning_rate": 0.0002, + "loss": 0.9073, + "step": 6710 + }, + { + "epoch": 2.116367076631977, + "grad_norm": 0.439453125, + "learning_rate": 0.0002, + "loss": 0.8165, + "step": 6711 + }, + { + "epoch": 2.1166824345632294, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 0.8019, + "step": 6712 + }, + { + "epoch": 2.116997792494481, + "grad_norm": 0.37890625, + "learning_rate": 0.0002, + "loss": 0.7301, + "step": 6713 + }, + { + "epoch": 2.1173131504257334, + "grad_norm": 0.421875, + "learning_rate": 0.0002, + "loss": 0.946, + "step": 6714 + }, + { + "epoch": 2.1176285083569852, + "grad_norm": 0.48046875, + "learning_rate": 0.0002, + "loss": 0.8387, + "step": 6715 + }, + { + "epoch": 2.117943866288237, + "grad_norm": 0.451171875, + "learning_rate": 0.0002, + "loss": 0.8297, + "step": 6716 + }, + { + "epoch": 2.1182592242194893, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 0.8623, + "step": 6717 + }, + { + "epoch": 2.118574582150741, + "grad_norm": 0.52734375, + "learning_rate": 0.0002, + "loss": 0.8966, + "step": 6718 + }, + { + "epoch": 2.118889940081993, + "grad_norm": 0.70703125, + "learning_rate": 0.0002, + "loss": 0.8295, + "step": 6719 + }, + { + "epoch": 2.119205298013245, + "grad_norm": 0.53515625, + "learning_rate": 0.0002, + "loss": 0.8523, + "step": 6720 + }, + { + "epoch": 2.119520655944497, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 0.7975, + "step": 6721 + }, + { + "epoch": 2.119836013875749, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 0.8915, + "step": 6722 + }, + { + "epoch": 2.120151371807001, + "grad_norm": 0.9375, + "learning_rate": 0.0002, + "loss": 1.1323, + "step": 6723 + }, + { + "epoch": 2.120466729738253, + "grad_norm": 0.92578125, + "learning_rate": 0.0002, + "loss": 1.0133, + "step": 6724 + }, + { + "epoch": 2.1207820876695047, + "grad_norm": 0.8046875, + "learning_rate": 0.0002, + "loss": 1.0785, + "step": 6725 + }, + { + "epoch": 2.121097445600757, + "grad_norm": 0.85546875, + "learning_rate": 0.0002, + "loss": 1.1442, + "step": 6726 + }, + { + "epoch": 2.121412803532009, + "grad_norm": 0.91796875, + "learning_rate": 0.0002, + "loss": 1.4165, + "step": 6727 + }, + { + "epoch": 2.1217281614632606, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 1.2415, + "step": 6728 + }, + { + "epoch": 2.122043519394513, + "grad_norm": 0.8203125, + "learning_rate": 0.0002, + "loss": 1.0016, + "step": 6729 + }, + { + "epoch": 2.1223588773257647, + "grad_norm": 0.8203125, + "learning_rate": 0.0002, + "loss": 1.2382, + "step": 6730 + }, + { + "epoch": 2.122674235257017, + "grad_norm": 0.7578125, + "learning_rate": 0.0002, + "loss": 1.1498, + "step": 6731 + }, + { + "epoch": 2.1229895931882687, + "grad_norm": 1.1328125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 6732 + }, + { + "epoch": 2.1229895931882687, + "eval_loss": 1.8525458574295044, + "eval_runtime": 148.8182, + "eval_samples_per_second": 6.72, + "eval_steps_per_second": 6.72, + "step": 6732 + }, + { + "epoch": 2.1229895931882687, + "mmlu_eval_accuracy": 0.6017477347760478, + "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, + "mmlu_eval_accuracy_anatomy": 0.5714285714285714, + "mmlu_eval_accuracy_astronomy": 0.5625, + "mmlu_eval_accuracy_business_ethics": 0.7272727272727273, + "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241, + "mmlu_eval_accuracy_college_biology": 0.875, + "mmlu_eval_accuracy_college_chemistry": 0.25, + "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365, + "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454, + "mmlu_eval_accuracy_college_medicine": 0.5454545454545454, + "mmlu_eval_accuracy_college_physics": 0.45454545454545453, + "mmlu_eval_accuracy_computer_security": 0.5454545454545454, + "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384, + "mmlu_eval_accuracy_econometrics": 0.5833333333333334, + "mmlu_eval_accuracy_electrical_engineering": 0.375, + "mmlu_eval_accuracy_elementary_mathematics": 0.4878048780487805, + "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, + "mmlu_eval_accuracy_global_facts": 0.3, + "mmlu_eval_accuracy_high_school_biology": 0.65625, + "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209, + "mmlu_eval_accuracy_high_school_mathematics": 0.3793103448275862, + "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769, + "mmlu_eval_accuracy_high_school_physics": 0.11764705882352941, + "mmlu_eval_accuracy_high_school_psychology": 0.85, + "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, + "mmlu_eval_accuracy_high_school_us_history": 0.8636363636363636, + "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, + "mmlu_eval_accuracy_human_aging": 0.7391304347826086, + "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334, + "mmlu_eval_accuracy_international_law": 1.0, + "mmlu_eval_accuracy_jurisprudence": 0.36363636363636365, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.8, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7093023255813954, + "mmlu_eval_accuracy_moral_disputes": 0.631578947368421, + "mmlu_eval_accuracy_moral_scenarios": 0.34, + "mmlu_eval_accuracy_nutrition": 0.7575757575757576, + "mmlu_eval_accuracy_philosophy": 0.6764705882352942, + "mmlu_eval_accuracy_prehistory": 0.7142857142857143, + "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644, + "mmlu_eval_accuracy_professional_law": 0.38823529411764707, + "mmlu_eval_accuracy_professional_medicine": 0.7419354838709677, + "mmlu_eval_accuracy_professional_psychology": 0.6086956521739131, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7037037037037037, + "mmlu_eval_accuracy_sociology": 0.9090909090909091, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.5, + "mmlu_eval_accuracy_world_religions": 0.7368421052631579, + "mmlu_loss": 6.216100011597577, + "step": 6732 + }, + { + "epoch": 2.1233049511195206, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 1.2255, + "step": 6733 + }, + { + "epoch": 2.123620309050773, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 1.5895, + "step": 6734 + }, + { + "epoch": 2.1239356669820246, + "grad_norm": 1.125, + "learning_rate": 0.0002, + "loss": 1.6752, + "step": 6735 + }, + { + "epoch": 2.1242510249132764, + "grad_norm": 1.2578125, + "learning_rate": 0.0002, + "loss": 1.5121, + "step": 6736 + }, + { + "epoch": 2.1245663828445287, + "grad_norm": 1.3515625, + "learning_rate": 0.0002, + "loss": 2.0426, + "step": 6737 + }, + { + "epoch": 2.1248817407757805, + "grad_norm": 1.4921875, + "learning_rate": 0.0002, + "loss": 1.3712, + "step": 6738 + }, + { + "epoch": 2.1251970987070323, + "grad_norm": 0.98828125, + "learning_rate": 0.0002, + "loss": 1.2635, + "step": 6739 + }, + { + "epoch": 2.1255124566382846, + "grad_norm": 1.1953125, + "learning_rate": 0.0002, + "loss": 1.4258, + "step": 6740 + }, + { + "epoch": 2.1258278145695364, + "grad_norm": 3.296875, + "learning_rate": 0.0002, + "loss": 1.436, + "step": 6741 + }, + { + "epoch": 2.126143172500788, + "grad_norm": 3.296875, + "learning_rate": 0.0002, + "loss": 1.9302, + "step": 6742 + }, + { + "epoch": 2.1264585304320405, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 0.9721, + "step": 6743 + }, + { + "epoch": 2.1267738883632923, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 1.0201, + "step": 6744 + }, + { + "epoch": 2.127089246294544, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 0.8691, + "step": 6745 + }, + { + "epoch": 2.1274046042257964, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.9805, + "step": 6746 + }, + { + "epoch": 2.127719962157048, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 0.8913, + "step": 6747 + }, + { + "epoch": 2.1280353200883004, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 0.9101, + "step": 6748 + }, + { + "epoch": 2.1283506780195522, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.0827, + "step": 6749 + }, + { + "epoch": 2.128666035950804, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 0.9678, + "step": 6750 + }, + { + "epoch": 2.1289813938820563, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 0.9073, + "step": 6751 + }, + { + "epoch": 2.129296751813308, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 0.8667, + "step": 6752 + }, + { + "epoch": 2.12961210974456, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 0.8909, + "step": 6753 + }, + { + "epoch": 2.129927467675812, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.8875, + "step": 6754 + }, + { + "epoch": 2.130242825607064, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 0.8765, + "step": 6755 + }, + { + "epoch": 2.130558183538316, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 0.8235, + "step": 6756 + }, + { + "epoch": 2.130873541469568, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 0.8983, + "step": 6757 + }, + { + "epoch": 2.13118889940082, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 0.8858, + "step": 6758 + }, + { + "epoch": 2.1315042573320717, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 0.9038, + "step": 6759 + }, + { + "epoch": 2.131819615263324, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 0.8634, + "step": 6760 + }, + { + "epoch": 2.132134973194576, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 0.9045, + "step": 6761 + }, + { + "epoch": 2.1324503311258276, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 0.7182, + "step": 6762 + }, + { + "epoch": 2.13276568905708, + "grad_norm": 0.474609375, + "learning_rate": 0.0002, + "loss": 0.8451, + "step": 6763 + }, + { + "epoch": 2.1330810469883317, + "grad_norm": 0.7265625, + "learning_rate": 0.0002, + "loss": 0.968, + "step": 6764 + }, + { + "epoch": 2.133396404919584, + "grad_norm": 0.453125, + "learning_rate": 0.0002, + "loss": 0.947, + "step": 6765 + }, + { + "epoch": 2.1337117628508357, + "grad_norm": 0.4296875, + "learning_rate": 0.0002, + "loss": 0.7792, + "step": 6766 + }, + { + "epoch": 2.1340271207820876, + "grad_norm": 0.60546875, + "learning_rate": 0.0002, + "loss": 0.9491, + "step": 6767 + }, + { + "epoch": 2.13434247871334, + "grad_norm": 0.494140625, + "learning_rate": 0.0002, + "loss": 0.7736, + "step": 6768 + }, + { + "epoch": 2.1346578366445916, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 0.9332, + "step": 6769 + }, + { + "epoch": 2.1349731945758434, + "grad_norm": 0.58203125, + "learning_rate": 0.0002, + "loss": 0.8511, + "step": 6770 + }, + { + "epoch": 2.1352885525070957, + "grad_norm": 0.828125, + "learning_rate": 0.0002, + "loss": 0.9624, + "step": 6771 + }, + { + "epoch": 2.1356039104383475, + "grad_norm": 0.81640625, + "learning_rate": 0.0002, + "loss": 0.9889, + "step": 6772 + }, + { + "epoch": 2.1359192683695993, + "grad_norm": 0.80859375, + "learning_rate": 0.0002, + "loss": 1.0128, + "step": 6773 + }, + { + "epoch": 2.1362346263008516, + "grad_norm": 0.9765625, + "learning_rate": 0.0002, + "loss": 0.9617, + "step": 6774 + }, + { + "epoch": 2.1365499842321034, + "grad_norm": 0.953125, + "learning_rate": 0.0002, + "loss": 1.1108, + "step": 6775 + }, + { + "epoch": 2.136865342163355, + "grad_norm": 0.72265625, + "learning_rate": 0.0002, + "loss": 0.9289, + "step": 6776 + }, + { + "epoch": 2.1371807000946075, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 1.2519, + "step": 6777 + }, + { + "epoch": 2.1374960580258593, + "grad_norm": 0.9765625, + "learning_rate": 0.0002, + "loss": 1.2676, + "step": 6778 + }, + { + "epoch": 2.1378114159571115, + "grad_norm": 0.859375, + "learning_rate": 0.0002, + "loss": 1.1327, + "step": 6779 + }, + { + "epoch": 2.1381267738883634, + "grad_norm": 1.15625, + "learning_rate": 0.0002, + "loss": 1.3943, + "step": 6780 + }, + { + "epoch": 2.138442131819615, + "grad_norm": 0.9765625, + "learning_rate": 0.0002, + "loss": 1.474, + "step": 6781 + }, + { + "epoch": 2.1387574897508674, + "grad_norm": 1.1796875, + "learning_rate": 0.0002, + "loss": 1.1832, + "step": 6782 + }, + { + "epoch": 2.1390728476821192, + "grad_norm": 1.140625, + "learning_rate": 0.0002, + "loss": 1.5506, + "step": 6783 + }, + { + "epoch": 2.139388205613371, + "grad_norm": 1.3203125, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 6784 + }, + { + "epoch": 2.1397035635446233, + "grad_norm": 1.3046875, + "learning_rate": 0.0002, + "loss": 1.4783, + "step": 6785 + }, + { + "epoch": 2.140018921475875, + "grad_norm": 0.94140625, + "learning_rate": 0.0002, + "loss": 1.4069, + "step": 6786 + }, + { + "epoch": 2.140334279407127, + "grad_norm": 1.1953125, + "learning_rate": 0.0002, + "loss": 1.8722, + "step": 6787 + }, + { + "epoch": 2.140649637338379, + "grad_norm": 0.96484375, + "learning_rate": 0.0002, + "loss": 1.2043, + "step": 6788 + }, + { + "epoch": 2.140964995269631, + "grad_norm": 1.390625, + "learning_rate": 0.0002, + "loss": 1.5431, + "step": 6789 + }, + { + "epoch": 2.141280353200883, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 1.1879, + "step": 6790 + }, + { + "epoch": 2.141595711132135, + "grad_norm": 1.125, + "learning_rate": 0.0002, + "loss": 1.4227, + "step": 6791 + }, + { + "epoch": 2.141911069063387, + "grad_norm": 1.2578125, + "learning_rate": 0.0002, + "loss": 1.716, + "step": 6792 + }, + { + "epoch": 2.1422264269946387, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 0.9737, + "step": 6793 + }, + { + "epoch": 2.142541784925891, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 0.8982, + "step": 6794 + }, + { + "epoch": 2.142857142857143, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 0.9432, + "step": 6795 + }, + { + "epoch": 2.1431725007883946, + "grad_norm": 0.330078125, + "learning_rate": 0.0002, + "loss": 1.0265, + "step": 6796 + }, + { + "epoch": 2.143487858719647, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 0.8574, + "step": 6797 + }, + { + "epoch": 2.1438032166508987, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 0.9782, + "step": 6798 + }, + { + "epoch": 2.144118574582151, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 0.9817, + "step": 6799 + }, + { + "epoch": 2.1444339325134028, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 0.7468, + "step": 6800 + }, + { + "epoch": 2.1447492904446546, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 0.9319, + "step": 6801 + }, + { + "epoch": 2.145064648375907, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 0.9282, + "step": 6802 + }, + { + "epoch": 2.1453800063071586, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 0.9502, + "step": 6803 + }, + { + "epoch": 2.1456953642384105, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 0.888, + "step": 6804 + }, + { + "epoch": 2.1460107221696627, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.7909, + "step": 6805 + }, + { + "epoch": 2.1463260801009145, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 0.8783, + "step": 6806 + }, + { + "epoch": 2.1466414380321663, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 0.8778, + "step": 6807 + }, + { + "epoch": 2.1469567959634186, + "grad_norm": 0.462890625, + "learning_rate": 0.0002, + "loss": 0.8723, + "step": 6808 + }, + { + "epoch": 2.1472721538946704, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 0.8327, + "step": 6809 + }, + { + "epoch": 2.1475875118259222, + "grad_norm": 0.330078125, + "learning_rate": 0.0002, + "loss": 0.9163, + "step": 6810 + }, + { + "epoch": 2.1479028697571745, + "grad_norm": 0.404296875, + "learning_rate": 0.0002, + "loss": 0.9046, + "step": 6811 + }, + { + "epoch": 2.1482182276884263, + "grad_norm": 0.455078125, + "learning_rate": 0.0002, + "loss": 0.877, + "step": 6812 + }, + { + "epoch": 2.1485335856196786, + "grad_norm": 0.37109375, + "learning_rate": 0.0002, + "loss": 0.8659, + "step": 6813 + }, + { + "epoch": 2.1488489435509304, + "grad_norm": 0.40234375, + "learning_rate": 0.0002, + "loss": 0.7823, + "step": 6814 + }, + { + "epoch": 2.149164301482182, + "grad_norm": 0.416015625, + "learning_rate": 0.0002, + "loss": 0.826, + "step": 6815 + }, + { + "epoch": 2.1494796594134344, + "grad_norm": 0.474609375, + "learning_rate": 0.0002, + "loss": 0.8773, + "step": 6816 + }, + { + "epoch": 2.1497950173446863, + "grad_norm": 0.48828125, + "learning_rate": 0.0002, + "loss": 0.8247, + "step": 6817 + }, + { + "epoch": 2.150110375275938, + "grad_norm": 0.482421875, + "learning_rate": 0.0002, + "loss": 0.8195, + "step": 6818 + }, + { + "epoch": 2.1504257332071903, + "grad_norm": 0.63671875, + "learning_rate": 0.0002, + "loss": 1.0227, + "step": 6819 + }, + { + "epoch": 2.150741091138442, + "grad_norm": 0.51953125, + "learning_rate": 0.0002, + "loss": 0.7927, + "step": 6820 + }, + { + "epoch": 2.151056449069694, + "grad_norm": 0.578125, + "learning_rate": 0.0002, + "loss": 0.7765, + "step": 6821 + }, + { + "epoch": 2.151371807000946, + "grad_norm": 0.51953125, + "learning_rate": 0.0002, + "loss": 0.7539, + "step": 6822 + }, + { + "epoch": 2.151687164932198, + "grad_norm": 0.734375, + "learning_rate": 0.0002, + "loss": 0.8592, + "step": 6823 + }, + { + "epoch": 2.15200252286345, + "grad_norm": 0.76171875, + "learning_rate": 0.0002, + "loss": 1.0286, + "step": 6824 + }, + { + "epoch": 2.152317880794702, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 1.0353, + "step": 6825 + }, + { + "epoch": 2.152633238725954, + "grad_norm": 1.2890625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 6826 + }, + { + "epoch": 2.1529485966572057, + "grad_norm": 0.71484375, + "learning_rate": 0.0002, + "loss": 0.8533, + "step": 6827 + }, + { + "epoch": 2.153263954588458, + "grad_norm": 0.8203125, + "learning_rate": 0.0002, + "loss": 1.0355, + "step": 6828 + }, + { + "epoch": 2.15357931251971, + "grad_norm": 0.7890625, + "learning_rate": 0.0002, + "loss": 1.1446, + "step": 6829 + }, + { + "epoch": 2.1538946704509616, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 1.2475, + "step": 6830 + }, + { + "epoch": 2.154210028382214, + "grad_norm": 0.98046875, + "learning_rate": 0.0002, + "loss": 1.4264, + "step": 6831 + }, + { + "epoch": 2.1545253863134657, + "grad_norm": 0.92578125, + "learning_rate": 0.0002, + "loss": 1.5228, + "step": 6832 + }, + { + "epoch": 2.154840744244718, + "grad_norm": 1.4609375, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 6833 + }, + { + "epoch": 2.1551561021759698, + "grad_norm": 1.3203125, + "learning_rate": 0.0002, + "loss": 1.471, + "step": 6834 + }, + { + "epoch": 2.1554714601072216, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 1.4169, + "step": 6835 + }, + { + "epoch": 2.155786818038474, + "grad_norm": 1.078125, + "learning_rate": 0.0002, + "loss": 1.6204, + "step": 6836 + }, + { + "epoch": 2.1561021759697256, + "grad_norm": 1.125, + "learning_rate": 0.0002, + "loss": 1.5755, + "step": 6837 + }, + { + "epoch": 2.1564175339009775, + "grad_norm": 1.21875, + "learning_rate": 0.0002, + "loss": 1.7854, + "step": 6838 + }, + { + "epoch": 2.1567328918322297, + "grad_norm": 1.515625, + "learning_rate": 0.0002, + "loss": 1.8862, + "step": 6839 + }, + { + "epoch": 2.1570482497634815, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.6434, + "step": 6840 + }, + { + "epoch": 2.1573636076947333, + "grad_norm": 0.96484375, + "learning_rate": 0.0002, + "loss": 1.2812, + "step": 6841 + }, + { + "epoch": 2.1576789656259856, + "grad_norm": 2.0, + "learning_rate": 0.0002, + "loss": 1.8773, + "step": 6842 + }, + { + "epoch": 2.1579943235572374, + "grad_norm": 0.2412109375, + "learning_rate": 0.0002, + "loss": 0.9616, + "step": 6843 + }, + { + "epoch": 2.1583096814884892, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 0.8371, + "step": 6844 + }, + { + "epoch": 2.1586250394197415, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 0.9239, + "step": 6845 + }, + { + "epoch": 2.1589403973509933, + "grad_norm": 0.3515625, + "learning_rate": 0.0002, + "loss": 0.9735, + "step": 6846 + }, + { + "epoch": 2.1592557552822456, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 0.9055, + "step": 6847 + }, + { + "epoch": 2.1595711132134974, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 0.9053, + "step": 6848 + }, + { + "epoch": 2.159886471144749, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 1.0364, + "step": 6849 + }, + { + "epoch": 2.1602018290760014, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.8772, + "step": 6850 + }, + { + "epoch": 2.1605171870072533, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 0.9076, + "step": 6851 + }, + { + "epoch": 2.160832544938505, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 1.0285, + "step": 6852 + }, + { + "epoch": 2.1611479028697573, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 0.9368, + "step": 6853 + }, + { + "epoch": 2.161463260801009, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 0.9359, + "step": 6854 + }, + { + "epoch": 2.161778618732261, + "grad_norm": 0.3828125, + "learning_rate": 0.0002, + "loss": 0.854, + "step": 6855 + }, + { + "epoch": 2.162093976663513, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 0.975, + "step": 6856 + }, + { + "epoch": 2.162409334594765, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 0.9027, + "step": 6857 + }, + { + "epoch": 2.162724692526017, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 0.8713, + "step": 6858 + }, + { + "epoch": 2.163040050457269, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 0.8692, + "step": 6859 + }, + { + "epoch": 2.163355408388521, + "grad_norm": 0.46484375, + "learning_rate": 0.0002, + "loss": 1.0468, + "step": 6860 + }, + { + "epoch": 2.1636707663197727, + "grad_norm": 0.40625, + "learning_rate": 0.0002, + "loss": 0.8628, + "step": 6861 + }, + { + "epoch": 2.163986124251025, + "grad_norm": 0.416015625, + "learning_rate": 0.0002, + "loss": 0.8262, + "step": 6862 + }, + { + "epoch": 2.164301482182277, + "grad_norm": 0.455078125, + "learning_rate": 0.0002, + "loss": 0.7549, + "step": 6863 + }, + { + "epoch": 2.164616840113529, + "grad_norm": 0.423828125, + "learning_rate": 0.0002, + "loss": 1.0064, + "step": 6864 + }, + { + "epoch": 2.164932198044781, + "grad_norm": 0.4375, + "learning_rate": 0.0002, + "loss": 0.8122, + "step": 6865 + }, + { + "epoch": 2.1652475559760327, + "grad_norm": 0.4453125, + "learning_rate": 0.0002, + "loss": 0.6937, + "step": 6866 + }, + { + "epoch": 2.165562913907285, + "grad_norm": 0.515625, + "learning_rate": 0.0002, + "loss": 0.8166, + "step": 6867 + }, + { + "epoch": 2.1658782718385368, + "grad_norm": 0.56640625, + "learning_rate": 0.0002, + "loss": 0.8838, + "step": 6868 + }, + { + "epoch": 2.1661936297697886, + "grad_norm": 0.546875, + "learning_rate": 0.0002, + "loss": 0.8142, + "step": 6869 + }, + { + "epoch": 2.166508987701041, + "grad_norm": 0.68359375, + "learning_rate": 0.0002, + "loss": 0.9701, + "step": 6870 + }, + { + "epoch": 2.1668243456322926, + "grad_norm": 0.69921875, + "learning_rate": 0.0002, + "loss": 0.9403, + "step": 6871 + }, + { + "epoch": 2.1671397035635445, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 0.8408, + "step": 6872 + }, + { + "epoch": 2.1674550614947967, + "grad_norm": 0.77734375, + "learning_rate": 0.0002, + "loss": 0.9748, + "step": 6873 + }, + { + "epoch": 2.1677704194260485, + "grad_norm": 0.87109375, + "learning_rate": 0.0002, + "loss": 1.1465, + "step": 6874 + }, + { + "epoch": 2.1680857773573003, + "grad_norm": 0.93359375, + "learning_rate": 0.0002, + "loss": 1.1301, + "step": 6875 + }, + { + "epoch": 2.1684011352885526, + "grad_norm": 0.984375, + "learning_rate": 0.0002, + "loss": 1.1411, + "step": 6876 + }, + { + "epoch": 2.1687164932198044, + "grad_norm": 0.92578125, + "learning_rate": 0.0002, + "loss": 1.1926, + "step": 6877 + }, + { + "epoch": 2.1690318511510567, + "grad_norm": 0.9609375, + "learning_rate": 0.0002, + "loss": 1.1104, + "step": 6878 + }, + { + "epoch": 2.1693472090823085, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 1.1172, + "step": 6879 + }, + { + "epoch": 2.1696625670135603, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 1.4734, + "step": 6880 + }, + { + "epoch": 2.1699779249448126, + "grad_norm": 1.1328125, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 6881 + }, + { + "epoch": 2.1702932828760644, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 1.5857, + "step": 6882 + }, + { + "epoch": 2.170608640807316, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.6837, + "step": 6883 + }, + { + "epoch": 2.1709239987385685, + "grad_norm": 1.1875, + "learning_rate": 0.0002, + "loss": 1.3878, + "step": 6884 + }, + { + "epoch": 2.1712393566698203, + "grad_norm": 0.95703125, + "learning_rate": 0.0002, + "loss": 1.578, + "step": 6885 + }, + { + "epoch": 2.171554714601072, + "grad_norm": 1.03125, + "learning_rate": 0.0002, + "loss": 1.5877, + "step": 6886 + }, + { + "epoch": 2.1718700725323243, + "grad_norm": 1.3359375, + "learning_rate": 0.0002, + "loss": 1.5089, + "step": 6887 + }, + { + "epoch": 2.172185430463576, + "grad_norm": 1.8671875, + "learning_rate": 0.0002, + "loss": 2.2335, + "step": 6888 + }, + { + "epoch": 2.172500788394828, + "grad_norm": 1.21875, + "learning_rate": 0.0002, + "loss": 1.6846, + "step": 6889 + }, + { + "epoch": 2.1728161463260802, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 6890 + }, + { + "epoch": 2.173131504257332, + "grad_norm": 1.21875, + "learning_rate": 0.0002, + "loss": 1.458, + "step": 6891 + }, + { + "epoch": 2.173446862188584, + "grad_norm": 1.5859375, + "learning_rate": 0.0002, + "loss": 1.9691, + "step": 6892 + }, + { + "epoch": 2.173762220119836, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 0.8815, + "step": 6893 + }, + { + "epoch": 2.174077578051088, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 0.9246, + "step": 6894 + }, + { + "epoch": 2.1743929359823397, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 0.8797, + "step": 6895 + }, + { + "epoch": 2.174708293913592, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 0.9766, + "step": 6896 + }, + { + "epoch": 2.175023651844844, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 1.03, + "step": 6897 + }, + { + "epoch": 2.175339009776096, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 0.9645, + "step": 6898 + }, + { + "epoch": 2.175654367707348, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 0.8817, + "step": 6899 + }, + { + "epoch": 2.1759697256385997, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 0.8315, + "step": 6900 + }, + { + "epoch": 2.176285083569852, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 0.8531, + "step": 6901 + }, + { + "epoch": 2.1766004415011038, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 0.9091, + "step": 6902 + }, + { + "epoch": 2.1769157994323556, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 0.9021, + "step": 6903 + }, + { + "epoch": 2.177231157363608, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.0241, + "step": 6904 + }, + { + "epoch": 2.1775465152948597, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.8844, + "step": 6905 + }, + { + "epoch": 2.1778618732261115, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 0.8834, + "step": 6906 + }, + { + "epoch": 2.1781772311573637, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 0.8363, + "step": 6907 + }, + { + "epoch": 2.1784925890886155, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 0.8772, + "step": 6908 + }, + { + "epoch": 2.1788079470198674, + "grad_norm": 0.361328125, + "learning_rate": 0.0002, + "loss": 0.8734, + "step": 6909 + }, + { + "epoch": 2.1791233049511196, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 0.8024, + "step": 6910 + }, + { + "epoch": 2.1794386628823714, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 0.9208, + "step": 6911 + }, + { + "epoch": 2.1797540208136237, + "grad_norm": 0.373046875, + "learning_rate": 0.0002, + "loss": 0.9506, + "step": 6912 + }, + { + "epoch": 2.1800693787448755, + "grad_norm": 0.435546875, + "learning_rate": 0.0002, + "loss": 0.9807, + "step": 6913 + }, + { + "epoch": 2.1803847366761273, + "grad_norm": 0.3984375, + "learning_rate": 0.0002, + "loss": 0.8532, + "step": 6914 + }, + { + "epoch": 2.1807000946073796, + "grad_norm": 0.43359375, + "learning_rate": 0.0002, + "loss": 0.7925, + "step": 6915 + }, + { + "epoch": 2.1810154525386314, + "grad_norm": 0.396484375, + "learning_rate": 0.0002, + "loss": 0.7168, + "step": 6916 + }, + { + "epoch": 2.181330810469883, + "grad_norm": 0.486328125, + "learning_rate": 0.0002, + "loss": 0.8383, + "step": 6917 + }, + { + "epoch": 2.1816461684011355, + "grad_norm": 0.58203125, + "learning_rate": 0.0002, + "loss": 0.9802, + "step": 6918 + }, + { + "epoch": 2.1819615263323873, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 0.8286, + "step": 6919 + }, + { + "epoch": 2.1819615263323873, + "eval_loss": 1.7957621812820435, + "eval_runtime": 148.691, + "eval_samples_per_second": 6.725, + "eval_steps_per_second": 6.725, + "step": 6919 + }, + { + "epoch": 2.1819615263323873, + "mmlu_eval_accuracy": 0.5960640558859189, + "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, + "mmlu_eval_accuracy_anatomy": 0.5, + "mmlu_eval_accuracy_astronomy": 0.625, + "mmlu_eval_accuracy_business_ethics": 0.8181818181818182, + "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241, + "mmlu_eval_accuracy_college_biology": 0.8125, + "mmlu_eval_accuracy_college_chemistry": 0.25, + "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, + "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453, + "mmlu_eval_accuracy_college_medicine": 0.5, + "mmlu_eval_accuracy_college_physics": 0.45454545454545453, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.5769230769230769, + "mmlu_eval_accuracy_econometrics": 0.5, + "mmlu_eval_accuracy_electrical_engineering": 0.5625, + "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634, + "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, + "mmlu_eval_accuracy_global_facts": 0.2, + "mmlu_eval_accuracy_high_school_biology": 0.625, + "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8571428571428571, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209, + "mmlu_eval_accuracy_high_school_mathematics": 0.3103448275862069, + "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769, + "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, + "mmlu_eval_accuracy_high_school_psychology": 0.7666666666666667, + "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, + "mmlu_eval_accuracy_high_school_us_history": 0.8636363636363636, + "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923, + "mmlu_eval_accuracy_human_aging": 0.6956521739130435, + "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334, + "mmlu_eval_accuracy_international_law": 1.0, + "mmlu_eval_accuracy_jurisprudence": 0.36363636363636365, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.8, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7093023255813954, + "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, + "mmlu_eval_accuracy_moral_scenarios": 0.34, + "mmlu_eval_accuracy_nutrition": 0.7272727272727273, + "mmlu_eval_accuracy_philosophy": 0.6764705882352942, + "mmlu_eval_accuracy_prehistory": 0.6571428571428571, + "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355, + "mmlu_eval_accuracy_professional_law": 0.3941176470588235, + "mmlu_eval_accuracy_professional_medicine": 0.7096774193548387, + "mmlu_eval_accuracy_professional_psychology": 0.5652173913043478, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7037037037037037, + "mmlu_eval_accuracy_sociology": 0.9090909090909091, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.5, + "mmlu_eval_accuracy_world_religions": 0.7894736842105263, + "mmlu_loss": 5.572403286268009, + "step": 6919 + }, + { + "epoch": 2.182276884263639, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 0.8461, + "step": 6920 + }, + { + "epoch": 2.1825922421948913, + "grad_norm": 0.474609375, + "learning_rate": 0.0002, + "loss": 0.7308, + "step": 6921 + }, + { + "epoch": 2.182907600126143, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 0.8997, + "step": 6922 + }, + { + "epoch": 2.183222958057395, + "grad_norm": 0.60546875, + "learning_rate": 0.0002, + "loss": 0.8323, + "step": 6923 + }, + { + "epoch": 2.1835383159886472, + "grad_norm": 0.6875, + "learning_rate": 0.0002, + "loss": 1.0236, + "step": 6924 + }, + { + "epoch": 2.183853673919899, + "grad_norm": 0.703125, + "learning_rate": 0.0002, + "loss": 1.1037, + "step": 6925 + }, + { + "epoch": 2.184169031851151, + "grad_norm": 0.87109375, + "learning_rate": 0.0002, + "loss": 1.023, + "step": 6926 + }, + { + "epoch": 2.184484389782403, + "grad_norm": 0.87890625, + "learning_rate": 0.0002, + "loss": 1.0682, + "step": 6927 + }, + { + "epoch": 2.184799747713655, + "grad_norm": 1.015625, + "learning_rate": 0.0002, + "loss": 1.1459, + "step": 6928 + }, + { + "epoch": 2.1851151056449067, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 6929 + }, + { + "epoch": 2.185430463576159, + "grad_norm": 1.078125, + "learning_rate": 0.0002, + "loss": 1.3746, + "step": 6930 + }, + { + "epoch": 2.185745821507411, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 1.5047, + "step": 6931 + }, + { + "epoch": 2.186061179438663, + "grad_norm": 1.1171875, + "learning_rate": 0.0002, + "loss": 1.371, + "step": 6932 + }, + { + "epoch": 2.186376537369915, + "grad_norm": 1.03125, + "learning_rate": 0.0002, + "loss": 1.3927, + "step": 6933 + }, + { + "epoch": 2.1866918953011667, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 1.6032, + "step": 6934 + }, + { + "epoch": 2.187007253232419, + "grad_norm": 1.140625, + "learning_rate": 0.0002, + "loss": 1.5054, + "step": 6935 + }, + { + "epoch": 2.1873226111636708, + "grad_norm": 1.1875, + "learning_rate": 0.0002, + "loss": 1.5139, + "step": 6936 + }, + { + "epoch": 2.1876379690949226, + "grad_norm": 1.140625, + "learning_rate": 0.0002, + "loss": 1.3987, + "step": 6937 + }, + { + "epoch": 2.187953327026175, + "grad_norm": 0.95703125, + "learning_rate": 0.0002, + "loss": 1.5163, + "step": 6938 + }, + { + "epoch": 2.1882686849574267, + "grad_norm": 1.34375, + "learning_rate": 0.0002, + "loss": 2.3611, + "step": 6939 + }, + { + "epoch": 2.1885840428886785, + "grad_norm": 1.3125, + "learning_rate": 0.0002, + "loss": 1.5378, + "step": 6940 + }, + { + "epoch": 2.1888994008199307, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 6941 + }, + { + "epoch": 2.1892147587511825, + "grad_norm": 1.2578125, + "learning_rate": 0.0002, + "loss": 1.6343, + "step": 6942 + }, + { + "epoch": 2.1895301166824344, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 0.9611, + "step": 6943 + }, + { + "epoch": 2.1898454746136866, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 0.8671, + "step": 6944 + }, + { + "epoch": 2.1901608325449384, + "grad_norm": 0.369140625, + "learning_rate": 0.0002, + "loss": 0.9492, + "step": 6945 + }, + { + "epoch": 2.1904761904761907, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 0.9255, + "step": 6946 + }, + { + "epoch": 2.1907915484074425, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 1.0168, + "step": 6947 + }, + { + "epoch": 2.1911069063386943, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.0095, + "step": 6948 + }, + { + "epoch": 2.1914222642699466, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 0.8592, + "step": 6949 + }, + { + "epoch": 2.1917376222011984, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 0.9512, + "step": 6950 + }, + { + "epoch": 2.19205298013245, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.0145, + "step": 6951 + }, + { + "epoch": 2.1923683380637025, + "grad_norm": 0.353515625, + "learning_rate": 0.0002, + "loss": 0.9239, + "step": 6952 + }, + { + "epoch": 2.1926836959949543, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.8834, + "step": 6953 + }, + { + "epoch": 2.192999053926206, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.9607, + "step": 6954 + }, + { + "epoch": 2.1933144118574583, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 0.8871, + "step": 6955 + }, + { + "epoch": 2.19362976978871, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 0.9161, + "step": 6956 + }, + { + "epoch": 2.193945127719962, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 0.9642, + "step": 6957 + }, + { + "epoch": 2.1942604856512142, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 0.8556, + "step": 6958 + }, + { + "epoch": 2.194575843582466, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 0.7941, + "step": 6959 + }, + { + "epoch": 2.194891201513718, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 0.8805, + "step": 6960 + }, + { + "epoch": 2.19520655944497, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 0.8455, + "step": 6961 + }, + { + "epoch": 2.195521917376222, + "grad_norm": 0.46484375, + "learning_rate": 0.0002, + "loss": 0.8628, + "step": 6962 + }, + { + "epoch": 2.195837275307474, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 0.8778, + "step": 6963 + }, + { + "epoch": 2.196152633238726, + "grad_norm": 0.439453125, + "learning_rate": 0.0002, + "loss": 0.9961, + "step": 6964 + }, + { + "epoch": 2.196467991169978, + "grad_norm": 0.4453125, + "learning_rate": 0.0002, + "loss": 0.7859, + "step": 6965 + }, + { + "epoch": 2.19678334910123, + "grad_norm": 0.5, + "learning_rate": 0.0002, + "loss": 0.8907, + "step": 6966 + }, + { + "epoch": 2.197098707032482, + "grad_norm": 0.486328125, + "learning_rate": 0.0002, + "loss": 0.9611, + "step": 6967 + }, + { + "epoch": 2.1974140649637337, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 0.9055, + "step": 6968 + }, + { + "epoch": 2.197729422894986, + "grad_norm": 0.515625, + "learning_rate": 0.0002, + "loss": 0.7732, + "step": 6969 + }, + { + "epoch": 2.1980447808262378, + "grad_norm": 0.55859375, + "learning_rate": 0.0002, + "loss": 0.9237, + "step": 6970 + }, + { + "epoch": 2.1983601387574896, + "grad_norm": 0.625, + "learning_rate": 0.0002, + "loss": 0.829, + "step": 6971 + }, + { + "epoch": 2.198675496688742, + "grad_norm": 0.61328125, + "learning_rate": 0.0002, + "loss": 0.8692, + "step": 6972 + }, + { + "epoch": 2.1989908546199937, + "grad_norm": 0.6171875, + "learning_rate": 0.0002, + "loss": 0.8531, + "step": 6973 + }, + { + "epoch": 2.1993062125512455, + "grad_norm": 0.86328125, + "learning_rate": 0.0002, + "loss": 1.071, + "step": 6974 + }, + { + "epoch": 2.1996215704824977, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.0082, + "step": 6975 + }, + { + "epoch": 2.1999369284137495, + "grad_norm": 0.86328125, + "learning_rate": 0.0002, + "loss": 1.2352, + "step": 6976 + }, + { + "epoch": 2.200252286345002, + "grad_norm": 0.85546875, + "learning_rate": 0.0002, + "loss": 0.9552, + "step": 6977 + }, + { + "epoch": 2.2005676442762536, + "grad_norm": 0.91796875, + "learning_rate": 0.0002, + "loss": 1.0437, + "step": 6978 + }, + { + "epoch": 2.2008830022075054, + "grad_norm": 1.140625, + "learning_rate": 0.0002, + "loss": 1.1256, + "step": 6979 + }, + { + "epoch": 2.2011983601387577, + "grad_norm": 0.98046875, + "learning_rate": 0.0002, + "loss": 1.1437, + "step": 6980 + }, + { + "epoch": 2.2015137180700095, + "grad_norm": 1.171875, + "learning_rate": 0.0002, + "loss": 1.4362, + "step": 6981 + }, + { + "epoch": 2.2018290760012613, + "grad_norm": 1.34375, + "learning_rate": 0.0002, + "loss": 1.3958, + "step": 6982 + }, + { + "epoch": 2.2021444339325136, + "grad_norm": 1.0546875, + "learning_rate": 0.0002, + "loss": 1.4098, + "step": 6983 + }, + { + "epoch": 2.2024597918637654, + "grad_norm": 0.9765625, + "learning_rate": 0.0002, + "loss": 1.4985, + "step": 6984 + }, + { + "epoch": 2.202775149795017, + "grad_norm": 1.171875, + "learning_rate": 0.0002, + "loss": 1.5783, + "step": 6985 + }, + { + "epoch": 2.2030905077262695, + "grad_norm": 1.25, + "learning_rate": 0.0002, + "loss": 1.6364, + "step": 6986 + }, + { + "epoch": 2.2034058656575213, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 1.4589, + "step": 6987 + }, + { + "epoch": 2.203721223588773, + "grad_norm": 1.3046875, + "learning_rate": 0.0002, + "loss": 1.8718, + "step": 6988 + }, + { + "epoch": 2.2040365815200254, + "grad_norm": 0.98046875, + "learning_rate": 0.0002, + "loss": 1.5619, + "step": 6989 + }, + { + "epoch": 2.204351939451277, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.3553, + "step": 6990 + }, + { + "epoch": 2.204667297382529, + "grad_norm": 1.1796875, + "learning_rate": 0.0002, + "loss": 1.4364, + "step": 6991 + }, + { + "epoch": 2.2049826553137812, + "grad_norm": 1.3359375, + "learning_rate": 0.0002, + "loss": 1.8022, + "step": 6992 + }, + { + "epoch": 2.205298013245033, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 0.8711, + "step": 6993 + }, + { + "epoch": 2.205613371176285, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 0.9019, + "step": 6994 + }, + { + "epoch": 2.205928729107537, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 0.9906, + "step": 6995 + }, + { + "epoch": 2.206244087038789, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 0.9322, + "step": 6996 + }, + { + "epoch": 2.206559444970041, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 1.0008, + "step": 6997 + }, + { + "epoch": 2.206874802901293, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 0.8957, + "step": 6998 + }, + { + "epoch": 2.207190160832545, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 0.8706, + "step": 6999 + }, + { + "epoch": 2.207505518763797, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 1.0296, + "step": 7000 + }, + { + "epoch": 2.207820876695049, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 0.869, + "step": 7001 + }, + { + "epoch": 2.2081362346263007, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 0.8732, + "step": 7002 + }, + { + "epoch": 2.208451592557553, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 1.0064, + "step": 7003 + }, + { + "epoch": 2.208766950488805, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.7366, + "step": 7004 + }, + { + "epoch": 2.2090823084200566, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 0.9475, + "step": 7005 + }, + { + "epoch": 2.209397666351309, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 1.0214, + "step": 7006 + }, + { + "epoch": 2.2097130242825607, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 0.8709, + "step": 7007 + }, + { + "epoch": 2.2100283822138125, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 0.7272, + "step": 7008 + }, + { + "epoch": 2.2103437401450647, + "grad_norm": 0.3671875, + "learning_rate": 0.0002, + "loss": 0.8686, + "step": 7009 + }, + { + "epoch": 2.2106590980763166, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 0.737, + "step": 7010 + }, + { + "epoch": 2.210974456007569, + "grad_norm": 0.388671875, + "learning_rate": 0.0002, + "loss": 1.0156, + "step": 7011 + }, + { + "epoch": 2.2112898139388206, + "grad_norm": 0.349609375, + "learning_rate": 0.0002, + "loss": 0.8431, + "step": 7012 + }, + { + "epoch": 2.2116051718700724, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 0.8137, + "step": 7013 + }, + { + "epoch": 2.2119205298013247, + "grad_norm": 0.443359375, + "learning_rate": 0.0002, + "loss": 0.8708, + "step": 7014 + }, + { + "epoch": 2.2122358877325765, + "grad_norm": 0.453125, + "learning_rate": 0.0002, + "loss": 1.0639, + "step": 7015 + }, + { + "epoch": 2.2125512456638283, + "grad_norm": 0.416015625, + "learning_rate": 0.0002, + "loss": 0.8026, + "step": 7016 + }, + { + "epoch": 2.2128666035950806, + "grad_norm": 0.42578125, + "learning_rate": 0.0002, + "loss": 0.8183, + "step": 7017 + }, + { + "epoch": 2.2131819615263324, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 0.9182, + "step": 7018 + }, + { + "epoch": 2.213497319457584, + "grad_norm": 0.66796875, + "learning_rate": 0.0002, + "loss": 0.8701, + "step": 7019 + }, + { + "epoch": 2.2138126773888365, + "grad_norm": 0.58203125, + "learning_rate": 0.0002, + "loss": 0.8684, + "step": 7020 + }, + { + "epoch": 2.2141280353200883, + "grad_norm": 0.515625, + "learning_rate": 0.0002, + "loss": 0.7302, + "step": 7021 + }, + { + "epoch": 2.21444339325134, + "grad_norm": 0.78125, + "learning_rate": 0.0002, + "loss": 0.9651, + "step": 7022 + }, + { + "epoch": 2.2147587511825924, + "grad_norm": 0.671875, + "learning_rate": 0.0002, + "loss": 0.8427, + "step": 7023 + }, + { + "epoch": 2.215074109113844, + "grad_norm": 0.80859375, + "learning_rate": 0.0002, + "loss": 0.9541, + "step": 7024 + }, + { + "epoch": 2.215389467045096, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 0.9636, + "step": 7025 + }, + { + "epoch": 2.2157048249763482, + "grad_norm": 0.93359375, + "learning_rate": 0.0002, + "loss": 1.1261, + "step": 7026 + }, + { + "epoch": 2.2160201829076, + "grad_norm": 0.859375, + "learning_rate": 0.0002, + "loss": 0.9727, + "step": 7027 + }, + { + "epoch": 2.216335540838852, + "grad_norm": 1.3125, + "learning_rate": 0.0002, + "loss": 1.2481, + "step": 7028 + }, + { + "epoch": 2.216650898770104, + "grad_norm": 0.9921875, + "learning_rate": 0.0002, + "loss": 1.2809, + "step": 7029 + }, + { + "epoch": 2.216966256701356, + "grad_norm": 1.328125, + "learning_rate": 0.0002, + "loss": 1.2768, + "step": 7030 + }, + { + "epoch": 2.217281614632608, + "grad_norm": 1.34375, + "learning_rate": 0.0002, + "loss": 1.5957, + "step": 7031 + }, + { + "epoch": 2.21759697256386, + "grad_norm": 1.109375, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 7032 + }, + { + "epoch": 2.217912330495112, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 1.4642, + "step": 7033 + }, + { + "epoch": 2.218227688426364, + "grad_norm": 1.0546875, + "learning_rate": 0.0002, + "loss": 1.5338, + "step": 7034 + }, + { + "epoch": 2.218543046357616, + "grad_norm": 1.2265625, + "learning_rate": 0.0002, + "loss": 1.6111, + "step": 7035 + }, + { + "epoch": 2.2188584042888677, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 1.3986, + "step": 7036 + }, + { + "epoch": 2.21917376222012, + "grad_norm": 1.5625, + "learning_rate": 0.0002, + "loss": 1.7548, + "step": 7037 + }, + { + "epoch": 2.219489120151372, + "grad_norm": 1.0859375, + "learning_rate": 0.0002, + "loss": 1.4448, + "step": 7038 + }, + { + "epoch": 2.2198044780826236, + "grad_norm": 1.234375, + "learning_rate": 0.0002, + "loss": 1.6355, + "step": 7039 + }, + { + "epoch": 2.220119836013876, + "grad_norm": 1.4375, + "learning_rate": 0.0002, + "loss": 1.8009, + "step": 7040 + }, + { + "epoch": 2.2204351939451277, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 1.5448, + "step": 7041 + }, + { + "epoch": 2.2207505518763795, + "grad_norm": 1.3515625, + "learning_rate": 0.0002, + "loss": 2.0412, + "step": 7042 + }, + { + "epoch": 2.2210659098076317, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 0.9534, + "step": 7043 + }, + { + "epoch": 2.2213812677388836, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 0.9045, + "step": 7044 + }, + { + "epoch": 2.221696625670136, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 1.021, + "step": 7045 + }, + { + "epoch": 2.2220119836013876, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.007, + "step": 7046 + }, + { + "epoch": 2.2223273415326394, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 1.0215, + "step": 7047 + }, + { + "epoch": 2.2226426994638917, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.0984, + "step": 7048 + }, + { + "epoch": 2.2229580573951435, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 0.9836, + "step": 7049 + }, + { + "epoch": 2.2232734153263953, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 0.9041, + "step": 7050 + }, + { + "epoch": 2.2235887732576476, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 0.939, + "step": 7051 + }, + { + "epoch": 2.2239041311888994, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 0.9167, + "step": 7052 + }, + { + "epoch": 2.224219489120151, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 0.9383, + "step": 7053 + }, + { + "epoch": 2.2245348470514035, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 1.0348, + "step": 7054 + }, + { + "epoch": 2.2248502049826553, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 0.8409, + "step": 7055 + }, + { + "epoch": 2.225165562913907, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 0.8567, + "step": 7056 + }, + { + "epoch": 2.2254809208451594, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 0.8304, + "step": 7057 + }, + { + "epoch": 2.225796278776411, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 0.9581, + "step": 7058 + }, + { + "epoch": 2.226111636707663, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 0.7994, + "step": 7059 + }, + { + "epoch": 2.2264269946389152, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 0.9416, + "step": 7060 + }, + { + "epoch": 2.226742352570167, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 0.8577, + "step": 7061 + }, + { + "epoch": 2.2270577105014193, + "grad_norm": 0.365234375, + "learning_rate": 0.0002, + "loss": 0.8956, + "step": 7062 + }, + { + "epoch": 2.227373068432671, + "grad_norm": 0.39453125, + "learning_rate": 0.0002, + "loss": 0.8752, + "step": 7063 + }, + { + "epoch": 2.227688426363923, + "grad_norm": 0.435546875, + "learning_rate": 0.0002, + "loss": 0.9785, + "step": 7064 + }, + { + "epoch": 2.228003784295175, + "grad_norm": 0.3984375, + "learning_rate": 0.0002, + "loss": 0.8119, + "step": 7065 + }, + { + "epoch": 2.228319142226427, + "grad_norm": 0.423828125, + "learning_rate": 0.0002, + "loss": 0.8194, + "step": 7066 + }, + { + "epoch": 2.228634500157679, + "grad_norm": 0.68359375, + "learning_rate": 0.0002, + "loss": 0.9683, + "step": 7067 + }, + { + "epoch": 2.228949858088931, + "grad_norm": 0.48828125, + "learning_rate": 0.0002, + "loss": 0.8823, + "step": 7068 + }, + { + "epoch": 2.229265216020183, + "grad_norm": 0.52734375, + "learning_rate": 0.0002, + "loss": 0.8418, + "step": 7069 + }, + { + "epoch": 2.2295805739514347, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 0.8585, + "step": 7070 + }, + { + "epoch": 2.229895931882687, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 0.7859, + "step": 7071 + }, + { + "epoch": 2.230211289813939, + "grad_norm": 0.55859375, + "learning_rate": 0.0002, + "loss": 0.8974, + "step": 7072 + }, + { + "epoch": 2.2305266477451906, + "grad_norm": 0.6640625, + "learning_rate": 0.0002, + "loss": 1.0706, + "step": 7073 + }, + { + "epoch": 2.230842005676443, + "grad_norm": 0.6640625, + "learning_rate": 0.0002, + "loss": 0.8763, + "step": 7074 + }, + { + "epoch": 2.2311573636076947, + "grad_norm": 0.78125, + "learning_rate": 0.0002, + "loss": 0.9699, + "step": 7075 + }, + { + "epoch": 2.231472721538947, + "grad_norm": 0.71484375, + "learning_rate": 0.0002, + "loss": 1.0598, + "step": 7076 + }, + { + "epoch": 2.2317880794701987, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 1.1427, + "step": 7077 + }, + { + "epoch": 2.2321034374014506, + "grad_norm": 1.640625, + "learning_rate": 0.0002, + "loss": 1.121, + "step": 7078 + }, + { + "epoch": 2.232418795332703, + "grad_norm": 0.890625, + "learning_rate": 0.0002, + "loss": 1.2373, + "step": 7079 + }, + { + "epoch": 2.2327341532639546, + "grad_norm": 0.90625, + "learning_rate": 0.0002, + "loss": 1.2827, + "step": 7080 + }, + { + "epoch": 2.2330495111952064, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 7081 + }, + { + "epoch": 2.2333648691264587, + "grad_norm": 0.93359375, + "learning_rate": 0.0002, + "loss": 1.2245, + "step": 7082 + }, + { + "epoch": 2.2336802270577105, + "grad_norm": 1.1171875, + "learning_rate": 0.0002, + "loss": 1.3956, + "step": 7083 + }, + { + "epoch": 2.2339955849889623, + "grad_norm": 1.3359375, + "learning_rate": 0.0002, + "loss": 1.2615, + "step": 7084 + }, + { + "epoch": 2.2343109429202146, + "grad_norm": 1.40625, + "learning_rate": 0.0002, + "loss": 1.6562, + "step": 7085 + }, + { + "epoch": 2.2346263008514664, + "grad_norm": 1.03125, + "learning_rate": 0.0002, + "loss": 1.4308, + "step": 7086 + }, + { + "epoch": 2.234941658782718, + "grad_norm": 1.46875, + "learning_rate": 0.0002, + "loss": 1.6006, + "step": 7087 + }, + { + "epoch": 2.2352570167139705, + "grad_norm": 1.078125, + "learning_rate": 0.0002, + "loss": 1.7723, + "step": 7088 + }, + { + "epoch": 2.2355723746452223, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 1.6878, + "step": 7089 + }, + { + "epoch": 2.235887732576474, + "grad_norm": 1.1640625, + "learning_rate": 0.0002, + "loss": 1.4828, + "step": 7090 + }, + { + "epoch": 2.2362030905077264, + "grad_norm": 1.2109375, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 7091 + }, + { + "epoch": 2.236518448438978, + "grad_norm": 1.2421875, + "learning_rate": 0.0002, + "loss": 1.7639, + "step": 7092 + }, + { + "epoch": 2.23683380637023, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 0.9696, + "step": 7093 + }, + { + "epoch": 2.2371491643014823, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 0.9809, + "step": 7094 + }, + { + "epoch": 2.237464522232734, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.0291, + "step": 7095 + }, + { + "epoch": 2.2377798801639863, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 0.9684, + "step": 7096 + }, + { + "epoch": 2.238095238095238, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.8648, + "step": 7097 + }, + { + "epoch": 2.23841059602649, + "grad_norm": 0.4375, + "learning_rate": 0.0002, + "loss": 0.9585, + "step": 7098 + }, + { + "epoch": 2.238725953957742, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 0.8804, + "step": 7099 + }, + { + "epoch": 2.239041311888994, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 0.9422, + "step": 7100 + }, + { + "epoch": 2.239356669820246, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 0.9293, + "step": 7101 + }, + { + "epoch": 2.239672027751498, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 0.9019, + "step": 7102 + }, + { + "epoch": 2.23998738568275, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 0.8964, + "step": 7103 + }, + { + "epoch": 2.2403027436140017, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 1.0091, + "step": 7104 + }, + { + "epoch": 2.240618101545254, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 0.8661, + "step": 7105 + }, + { + "epoch": 2.240933459476506, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 0.9336, + "step": 7106 + }, + { + "epoch": 2.240933459476506, + "eval_loss": 1.7919546365737915, + "eval_runtime": 148.3392, + "eval_samples_per_second": 6.741, + "eval_steps_per_second": 6.741, + "step": 7106 + }, + { + "epoch": 2.240933459476506, + "mmlu_eval_accuracy": 0.5992283682534031, + "mmlu_eval_accuracy_abstract_algebra": 0.5454545454545454, + "mmlu_eval_accuracy_anatomy": 0.5714285714285714, + "mmlu_eval_accuracy_astronomy": 0.5625, + "mmlu_eval_accuracy_business_ethics": 0.8181818181818182, + "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034, + "mmlu_eval_accuracy_college_biology": 0.8125, + "mmlu_eval_accuracy_college_chemistry": 0.375, + "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, + "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453, + "mmlu_eval_accuracy_college_medicine": 0.5, + "mmlu_eval_accuracy_college_physics": 0.45454545454545453, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384, + "mmlu_eval_accuracy_econometrics": 0.4166666666666667, + "mmlu_eval_accuracy_electrical_engineering": 0.375, + "mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536, + "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, + "mmlu_eval_accuracy_global_facts": 0.2, + "mmlu_eval_accuracy_high_school_biology": 0.5625, + "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.6046511627906976, + "mmlu_eval_accuracy_high_school_mathematics": 0.3448275862068966, + "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769, + "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, + "mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, + "mmlu_eval_accuracy_high_school_us_history": 0.8636363636363636, + "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, + "mmlu_eval_accuracy_human_aging": 0.6521739130434783, + "mmlu_eval_accuracy_human_sexuality": 0.5, + "mmlu_eval_accuracy_international_law": 1.0, + "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, + "mmlu_eval_accuracy_management": 0.8181818181818182, + "mmlu_eval_accuracy_marketing": 0.72, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.686046511627907, + "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527, + "mmlu_eval_accuracy_moral_scenarios": 0.36, + "mmlu_eval_accuracy_nutrition": 0.7575757575757576, + "mmlu_eval_accuracy_philosophy": 0.6470588235294118, + "mmlu_eval_accuracy_prehistory": 0.6285714285714286, + "mmlu_eval_accuracy_professional_accounting": 0.5483870967741935, + "mmlu_eval_accuracy_professional_law": 0.4294117647058823, + "mmlu_eval_accuracy_professional_medicine": 0.7096774193548387, + "mmlu_eval_accuracy_professional_psychology": 0.5942028985507246, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7037037037037037, + "mmlu_eval_accuracy_sociology": 0.9090909090909091, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.5, + "mmlu_eval_accuracy_world_religions": 0.7894736842105263, + "mmlu_loss": 5.8061809511764, + "step": 7106 + }, + { + "epoch": 2.2412488174077576, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 0.974, + "step": 7107 + }, + { + "epoch": 2.24156417533901, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 0.8713, + "step": 7108 + }, + { + "epoch": 2.2418795332702617, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 0.833, + "step": 7109 + }, + { + "epoch": 2.242194891201514, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 0.7814, + "step": 7110 + }, + { + "epoch": 2.2425102491327658, + "grad_norm": 0.3828125, + "learning_rate": 0.0002, + "loss": 0.8854, + "step": 7111 + }, + { + "epoch": 2.2428256070640176, + "grad_norm": 0.3671875, + "learning_rate": 0.0002, + "loss": 0.8902, + "step": 7112 + }, + { + "epoch": 2.24314096499527, + "grad_norm": 0.4296875, + "learning_rate": 0.0002, + "loss": 0.8513, + "step": 7113 + }, + { + "epoch": 2.2434563229265216, + "grad_norm": 0.451171875, + "learning_rate": 0.0002, + "loss": 0.9236, + "step": 7114 + }, + { + "epoch": 2.2437716808577735, + "grad_norm": 0.439453125, + "learning_rate": 0.0002, + "loss": 0.9098, + "step": 7115 + }, + { + "epoch": 2.2440870387890257, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 0.9782, + "step": 7116 + }, + { + "epoch": 2.2444023967202775, + "grad_norm": 0.48828125, + "learning_rate": 0.0002, + "loss": 0.8362, + "step": 7117 + }, + { + "epoch": 2.2447177546515293, + "grad_norm": 0.5, + "learning_rate": 0.0002, + "loss": 0.8607, + "step": 7118 + }, + { + "epoch": 2.2450331125827816, + "grad_norm": 0.51953125, + "learning_rate": 0.0002, + "loss": 0.8761, + "step": 7119 + }, + { + "epoch": 2.2453484705140334, + "grad_norm": 0.609375, + "learning_rate": 0.0002, + "loss": 0.8221, + "step": 7120 + }, + { + "epoch": 2.2456638284452852, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 0.8029, + "step": 7121 + }, + { + "epoch": 2.2459791863765375, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 0.9011, + "step": 7122 + }, + { + "epoch": 2.2462945443077893, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 0.7938, + "step": 7123 + }, + { + "epoch": 2.246609902239041, + "grad_norm": 0.63671875, + "learning_rate": 0.0002, + "loss": 0.8822, + "step": 7124 + }, + { + "epoch": 2.2469252601702934, + "grad_norm": 0.72265625, + "learning_rate": 0.0002, + "loss": 1.0854, + "step": 7125 + }, + { + "epoch": 2.247240618101545, + "grad_norm": 0.81640625, + "learning_rate": 0.0002, + "loss": 1.1014, + "step": 7126 + }, + { + "epoch": 2.247555976032797, + "grad_norm": 0.76171875, + "learning_rate": 0.0002, + "loss": 1.0786, + "step": 7127 + }, + { + "epoch": 2.2478713339640493, + "grad_norm": 0.859375, + "learning_rate": 0.0002, + "loss": 1.0964, + "step": 7128 + }, + { + "epoch": 2.248186691895301, + "grad_norm": 0.91796875, + "learning_rate": 0.0002, + "loss": 1.1891, + "step": 7129 + }, + { + "epoch": 2.2485020498265533, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 1.0818, + "step": 7130 + }, + { + "epoch": 2.248817407757805, + "grad_norm": 0.99609375, + "learning_rate": 0.0002, + "loss": 1.2282, + "step": 7131 + }, + { + "epoch": 2.249132765689057, + "grad_norm": 1.140625, + "learning_rate": 0.0002, + "loss": 1.2293, + "step": 7132 + }, + { + "epoch": 2.249448123620309, + "grad_norm": 0.89453125, + "learning_rate": 0.0002, + "loss": 1.2774, + "step": 7133 + }, + { + "epoch": 2.249763481551561, + "grad_norm": 1.0546875, + "learning_rate": 0.0002, + "loss": 1.5866, + "step": 7134 + }, + { + "epoch": 2.250078839482813, + "grad_norm": 1.4140625, + "learning_rate": 0.0002, + "loss": 1.4369, + "step": 7135 + }, + { + "epoch": 2.250394197414065, + "grad_norm": 1.34375, + "learning_rate": 0.0002, + "loss": 1.7717, + "step": 7136 + }, + { + "epoch": 2.250709555345317, + "grad_norm": 1.265625, + "learning_rate": 0.0002, + "loss": 1.6198, + "step": 7137 + }, + { + "epoch": 2.2510249132765687, + "grad_norm": 0.97265625, + "learning_rate": 0.0002, + "loss": 1.472, + "step": 7138 + }, + { + "epoch": 2.251340271207821, + "grad_norm": 1.4609375, + "learning_rate": 0.0002, + "loss": 1.9111, + "step": 7139 + }, + { + "epoch": 2.251655629139073, + "grad_norm": 1.3046875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 7140 + }, + { + "epoch": 2.251970987070325, + "grad_norm": 1.1328125, + "learning_rate": 0.0002, + "loss": 1.4079, + "step": 7141 + }, + { + "epoch": 2.252286345001577, + "grad_norm": 1.5390625, + "learning_rate": 0.0002, + "loss": 1.8088, + "step": 7142 + }, + { + "epoch": 2.2526017029328287, + "grad_norm": 0.2470703125, + "learning_rate": 0.0002, + "loss": 0.996, + "step": 7143 + }, + { + "epoch": 2.252917060864081, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 0.9395, + "step": 7144 + }, + { + "epoch": 2.2532324187953328, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 0.9498, + "step": 7145 + }, + { + "epoch": 2.2535477767265846, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 1.015, + "step": 7146 + }, + { + "epoch": 2.253863134657837, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 0.9537, + "step": 7147 + }, + { + "epoch": 2.2541784925890886, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 0.8825, + "step": 7148 + }, + { + "epoch": 2.2544938505203405, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 0.8577, + "step": 7149 + }, + { + "epoch": 2.2548092084515927, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.9843, + "step": 7150 + }, + { + "epoch": 2.2551245663828445, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 0.8346, + "step": 7151 + }, + { + "epoch": 2.2554399243140963, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 0.9391, + "step": 7152 + }, + { + "epoch": 2.2557552822453486, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 0.9259, + "step": 7153 + }, + { + "epoch": 2.2560706401766004, + "grad_norm": 0.515625, + "learning_rate": 0.0002, + "loss": 0.9167, + "step": 7154 + }, + { + "epoch": 2.2563859981078522, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 0.9852, + "step": 7155 + }, + { + "epoch": 2.2567013560391045, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 0.8464, + "step": 7156 + }, + { + "epoch": 2.2570167139703563, + "grad_norm": 0.365234375, + "learning_rate": 0.0002, + "loss": 0.9255, + "step": 7157 + }, + { + "epoch": 2.257332071901608, + "grad_norm": 0.427734375, + "learning_rate": 0.0002, + "loss": 0.727, + "step": 7158 + }, + { + "epoch": 2.2576474298328604, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 0.8315, + "step": 7159 + }, + { + "epoch": 2.257962787764112, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 0.9278, + "step": 7160 + }, + { + "epoch": 2.258278145695364, + "grad_norm": 0.4140625, + "learning_rate": 0.0002, + "loss": 0.855, + "step": 7161 + }, + { + "epoch": 2.2585935036266163, + "grad_norm": 0.384765625, + "learning_rate": 0.0002, + "loss": 0.9005, + "step": 7162 + }, + { + "epoch": 2.258908861557868, + "grad_norm": 0.404296875, + "learning_rate": 0.0002, + "loss": 0.8053, + "step": 7163 + }, + { + "epoch": 2.2592242194891203, + "grad_norm": 0.392578125, + "learning_rate": 0.0002, + "loss": 0.8811, + "step": 7164 + }, + { + "epoch": 2.259539577420372, + "grad_norm": 0.412109375, + "learning_rate": 0.0002, + "loss": 0.8686, + "step": 7165 + }, + { + "epoch": 2.259854935351624, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 0.9387, + "step": 7166 + }, + { + "epoch": 2.260170293282876, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 0.9548, + "step": 7167 + }, + { + "epoch": 2.260485651214128, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 1.0229, + "step": 7168 + }, + { + "epoch": 2.26080100914538, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 0.8803, + "step": 7169 + }, + { + "epoch": 2.261116367076632, + "grad_norm": 0.69921875, + "learning_rate": 0.0002, + "loss": 0.8928, + "step": 7170 + }, + { + "epoch": 2.261431725007884, + "grad_norm": 0.703125, + "learning_rate": 0.0002, + "loss": 0.9545, + "step": 7171 + }, + { + "epoch": 2.2617470829391357, + "grad_norm": 0.828125, + "learning_rate": 0.0002, + "loss": 0.9782, + "step": 7172 + }, + { + "epoch": 2.262062440870388, + "grad_norm": 0.734375, + "learning_rate": 0.0002, + "loss": 1.1245, + "step": 7173 + }, + { + "epoch": 2.26237779880164, + "grad_norm": 0.78125, + "learning_rate": 0.0002, + "loss": 0.979, + "step": 7174 + }, + { + "epoch": 2.262693156732892, + "grad_norm": 0.90234375, + "learning_rate": 0.0002, + "loss": 1.1032, + "step": 7175 + }, + { + "epoch": 2.263008514664144, + "grad_norm": 0.98828125, + "learning_rate": 0.0002, + "loss": 1.2668, + "step": 7176 + }, + { + "epoch": 2.2633238725953957, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 1.2147, + "step": 7177 + }, + { + "epoch": 2.263639230526648, + "grad_norm": 0.84375, + "learning_rate": 0.0002, + "loss": 1.0713, + "step": 7178 + }, + { + "epoch": 2.2639545884578998, + "grad_norm": 0.9453125, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 7179 + }, + { + "epoch": 2.2642699463891516, + "grad_norm": 1.171875, + "learning_rate": 0.0002, + "loss": 1.3882, + "step": 7180 + }, + { + "epoch": 2.264585304320404, + "grad_norm": 1.3046875, + "learning_rate": 0.0002, + "loss": 1.7908, + "step": 7181 + }, + { + "epoch": 2.2649006622516556, + "grad_norm": 1.03125, + "learning_rate": 0.0002, + "loss": 1.3591, + "step": 7182 + }, + { + "epoch": 2.2652160201829075, + "grad_norm": 1.3984375, + "learning_rate": 0.0002, + "loss": 1.432, + "step": 7183 + }, + { + "epoch": 2.2655313781141597, + "grad_norm": 1.03125, + "learning_rate": 0.0002, + "loss": 1.5752, + "step": 7184 + }, + { + "epoch": 2.2658467360454115, + "grad_norm": 1.234375, + "learning_rate": 0.0002, + "loss": 1.555, + "step": 7185 + }, + { + "epoch": 2.2661620939766633, + "grad_norm": 1.6328125, + "learning_rate": 0.0002, + "loss": 1.9601, + "step": 7186 + }, + { + "epoch": 2.2664774519079156, + "grad_norm": 1.4921875, + "learning_rate": 0.0002, + "loss": 1.6169, + "step": 7187 + }, + { + "epoch": 2.2667928098391674, + "grad_norm": 1.515625, + "learning_rate": 0.0002, + "loss": 1.6977, + "step": 7188 + }, + { + "epoch": 2.2671081677704192, + "grad_norm": 1.15625, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 7189 + }, + { + "epoch": 2.2674235257016715, + "grad_norm": 2.125, + "learning_rate": 0.0002, + "loss": 2.1718, + "step": 7190 + }, + { + "epoch": 2.2677388836329233, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 1.197, + "step": 7191 + }, + { + "epoch": 2.268054241564175, + "grad_norm": 1.4609375, + "learning_rate": 0.0002, + "loss": 1.9017, + "step": 7192 + }, + { + "epoch": 2.2683695994954274, + "grad_norm": 0.23046875, + "learning_rate": 0.0002, + "loss": 1.0598, + "step": 7193 + }, + { + "epoch": 2.268684957426679, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 0.9321, + "step": 7194 + }, + { + "epoch": 2.269000315357931, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 0.8911, + "step": 7195 + }, + { + "epoch": 2.2693156732891833, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 0.9903, + "step": 7196 + }, + { + "epoch": 2.269631031220435, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 0.9065, + "step": 7197 + }, + { + "epoch": 2.2699463891516873, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.9536, + "step": 7198 + }, + { + "epoch": 2.270261747082939, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 0.8309, + "step": 7199 + }, + { + "epoch": 2.270577105014191, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 0.9329, + "step": 7200 + }, + { + "epoch": 2.2708924629454432, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 0.9428, + "step": 7201 + }, + { + "epoch": 2.271207820876695, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 0.8479, + "step": 7202 + }, + { + "epoch": 2.271523178807947, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 0.935, + "step": 7203 + }, + { + "epoch": 2.271838536739199, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.9151, + "step": 7204 + }, + { + "epoch": 2.272153894670451, + "grad_norm": 0.3515625, + "learning_rate": 0.0002, + "loss": 0.8029, + "step": 7205 + }, + { + "epoch": 2.272469252601703, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 1.0327, + "step": 7206 + }, + { + "epoch": 2.272784610532955, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 0.8652, + "step": 7207 + }, + { + "epoch": 2.273099968464207, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 1.0592, + "step": 7208 + }, + { + "epoch": 2.273415326395459, + "grad_norm": 0.3515625, + "learning_rate": 0.0002, + "loss": 0.9374, + "step": 7209 + }, + { + "epoch": 2.273730684326711, + "grad_norm": 0.40625, + "learning_rate": 0.0002, + "loss": 0.8516, + "step": 7210 + }, + { + "epoch": 2.2740460422579627, + "grad_norm": 0.39453125, + "learning_rate": 0.0002, + "loss": 0.8139, + "step": 7211 + }, + { + "epoch": 2.274361400189215, + "grad_norm": 0.408203125, + "learning_rate": 0.0002, + "loss": 0.8558, + "step": 7212 + }, + { + "epoch": 2.2746767581204668, + "grad_norm": 0.369140625, + "learning_rate": 0.0002, + "loss": 0.8378, + "step": 7213 + }, + { + "epoch": 2.2749921160517186, + "grad_norm": 0.447265625, + "learning_rate": 0.0002, + "loss": 0.8377, + "step": 7214 + }, + { + "epoch": 2.275307473982971, + "grad_norm": 0.431640625, + "learning_rate": 0.0002, + "loss": 0.7956, + "step": 7215 + }, + { + "epoch": 2.2756228319142227, + "grad_norm": 0.408203125, + "learning_rate": 0.0002, + "loss": 0.7346, + "step": 7216 + }, + { + "epoch": 2.2759381898454745, + "grad_norm": 0.52734375, + "learning_rate": 0.0002, + "loss": 0.9638, + "step": 7217 + }, + { + "epoch": 2.2762535477767267, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 0.9311, + "step": 7218 + }, + { + "epoch": 2.2765689057079785, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 0.8282, + "step": 7219 + }, + { + "epoch": 2.2768842636392304, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 0.9701, + "step": 7220 + }, + { + "epoch": 2.2771996215704826, + "grad_norm": 0.6171875, + "learning_rate": 0.0002, + "loss": 0.8307, + "step": 7221 + }, + { + "epoch": 2.2775149795017344, + "grad_norm": 0.69921875, + "learning_rate": 0.0002, + "loss": 0.953, + "step": 7222 + }, + { + "epoch": 2.2778303374329862, + "grad_norm": 0.69921875, + "learning_rate": 0.0002, + "loss": 0.7822, + "step": 7223 + }, + { + "epoch": 2.2781456953642385, + "grad_norm": 0.87890625, + "learning_rate": 0.0002, + "loss": 1.1003, + "step": 7224 + }, + { + "epoch": 2.2784610532954903, + "grad_norm": 0.9921875, + "learning_rate": 0.0002, + "loss": 1.1156, + "step": 7225 + }, + { + "epoch": 2.278776411226742, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 1.17, + "step": 7226 + }, + { + "epoch": 2.2790917691579944, + "grad_norm": 0.9609375, + "learning_rate": 0.0002, + "loss": 1.3931, + "step": 7227 + }, + { + "epoch": 2.279407127089246, + "grad_norm": 0.8359375, + "learning_rate": 0.0002, + "loss": 1.0892, + "step": 7228 + }, + { + "epoch": 2.2797224850204985, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.4012, + "step": 7229 + }, + { + "epoch": 2.2800378429517503, + "grad_norm": 1.328125, + "learning_rate": 0.0002, + "loss": 1.7998, + "step": 7230 + }, + { + "epoch": 2.280353200883002, + "grad_norm": 1.0390625, + "learning_rate": 0.0002, + "loss": 1.649, + "step": 7231 + }, + { + "epoch": 2.2806685588142543, + "grad_norm": 1.3125, + "learning_rate": 0.0002, + "loss": 1.4941, + "step": 7232 + }, + { + "epoch": 2.280983916745506, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 1.467, + "step": 7233 + }, + { + "epoch": 2.281299274676758, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 1.3659, + "step": 7234 + }, + { + "epoch": 2.2816146326080102, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.6784, + "step": 7235 + }, + { + "epoch": 2.281929990539262, + "grad_norm": 1.25, + "learning_rate": 0.0002, + "loss": 1.7605, + "step": 7236 + }, + { + "epoch": 2.282245348470514, + "grad_norm": 1.2265625, + "learning_rate": 0.0002, + "loss": 1.7113, + "step": 7237 + }, + { + "epoch": 2.282560706401766, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.5786, + "step": 7238 + }, + { + "epoch": 2.282876064333018, + "grad_norm": 1.0390625, + "learning_rate": 0.0002, + "loss": 2.0238, + "step": 7239 + }, + { + "epoch": 2.28319142226427, + "grad_norm": 1.171875, + "learning_rate": 0.0002, + "loss": 1.5917, + "step": 7240 + }, + { + "epoch": 2.283506780195522, + "grad_norm": 0.90234375, + "learning_rate": 0.0002, + "loss": 1.2026, + "step": 7241 + }, + { + "epoch": 2.283822138126774, + "grad_norm": 1.296875, + "learning_rate": 0.0002, + "loss": 1.5997, + "step": 7242 + }, + { + "epoch": 2.284137496058026, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 1.0279, + "step": 7243 + }, + { + "epoch": 2.284452853989278, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 0.957, + "step": 7244 + }, + { + "epoch": 2.2847682119205297, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 0.9128, + "step": 7245 + }, + { + "epoch": 2.285083569851782, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 0.952, + "step": 7246 + }, + { + "epoch": 2.2853989277830338, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 0.9262, + "step": 7247 + }, + { + "epoch": 2.2857142857142856, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 1.0006, + "step": 7248 + }, + { + "epoch": 2.286029643645538, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 0.9097, + "step": 7249 + }, + { + "epoch": 2.2863450015767897, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 0.8822, + "step": 7250 + }, + { + "epoch": 2.2866603595080415, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 1.0077, + "step": 7251 + }, + { + "epoch": 2.2869757174392937, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 0.7867, + "step": 7252 + }, + { + "epoch": 2.2872910753705455, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 0.9084, + "step": 7253 + }, + { + "epoch": 2.2876064333017974, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 0.7733, + "step": 7254 + }, + { + "epoch": 2.2879217912330496, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 1.0112, + "step": 7255 + }, + { + "epoch": 2.2882371491643014, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 0.8442, + "step": 7256 + }, + { + "epoch": 2.2885525070955532, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 0.9099, + "step": 7257 + }, + { + "epoch": 2.2888678650268055, + "grad_norm": 0.365234375, + "learning_rate": 0.0002, + "loss": 0.917, + "step": 7258 + }, + { + "epoch": 2.2891832229580573, + "grad_norm": 0.37890625, + "learning_rate": 0.0002, + "loss": 1.0538, + "step": 7259 + }, + { + "epoch": 2.289498580889309, + "grad_norm": 0.40234375, + "learning_rate": 0.0002, + "loss": 0.9764, + "step": 7260 + }, + { + "epoch": 2.2898139388205614, + "grad_norm": 0.369140625, + "learning_rate": 0.0002, + "loss": 0.8545, + "step": 7261 + }, + { + "epoch": 2.290129296751813, + "grad_norm": 0.365234375, + "learning_rate": 0.0002, + "loss": 0.8153, + "step": 7262 + }, + { + "epoch": 2.2904446546830655, + "grad_norm": 0.42578125, + "learning_rate": 0.0002, + "loss": 0.8939, + "step": 7263 + }, + { + "epoch": 2.2907600126143173, + "grad_norm": 0.455078125, + "learning_rate": 0.0002, + "loss": 0.8037, + "step": 7264 + }, + { + "epoch": 2.291075370545569, + "grad_norm": 0.41015625, + "learning_rate": 0.0002, + "loss": 0.7418, + "step": 7265 + }, + { + "epoch": 2.2913907284768213, + "grad_norm": 0.40625, + "learning_rate": 0.0002, + "loss": 0.7686, + "step": 7266 + }, + { + "epoch": 2.291706086408073, + "grad_norm": 0.56640625, + "learning_rate": 0.0002, + "loss": 1.0373, + "step": 7267 + }, + { + "epoch": 2.292021444339325, + "grad_norm": 0.609375, + "learning_rate": 0.0002, + "loss": 0.9273, + "step": 7268 + }, + { + "epoch": 2.2923368022705772, + "grad_norm": 0.6640625, + "learning_rate": 0.0002, + "loss": 1.0344, + "step": 7269 + }, + { + "epoch": 2.292652160201829, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 0.9535, + "step": 7270 + }, + { + "epoch": 2.292967518133081, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 0.8229, + "step": 7271 + }, + { + "epoch": 2.293282876064333, + "grad_norm": 0.6484375, + "learning_rate": 0.0002, + "loss": 0.8748, + "step": 7272 + }, + { + "epoch": 2.293598233995585, + "grad_norm": 0.8359375, + "learning_rate": 0.0002, + "loss": 1.0137, + "step": 7273 + }, + { + "epoch": 2.293913591926837, + "grad_norm": 0.71875, + "learning_rate": 0.0002, + "loss": 1.0846, + "step": 7274 + }, + { + "epoch": 2.294228949858089, + "grad_norm": 0.90234375, + "learning_rate": 0.0002, + "loss": 1.036, + "step": 7275 + }, + { + "epoch": 2.294544307789341, + "grad_norm": 0.93359375, + "learning_rate": 0.0002, + "loss": 1.2236, + "step": 7276 + }, + { + "epoch": 2.294859665720593, + "grad_norm": 0.9921875, + "learning_rate": 0.0002, + "loss": 1.3819, + "step": 7277 + }, + { + "epoch": 2.295175023651845, + "grad_norm": 0.96484375, + "learning_rate": 0.0002, + "loss": 1.2392, + "step": 7278 + }, + { + "epoch": 2.2954903815830967, + "grad_norm": 0.96875, + "learning_rate": 0.0002, + "loss": 1.2488, + "step": 7279 + }, + { + "epoch": 2.295805739514349, + "grad_norm": 1.1875, + "learning_rate": 0.0002, + "loss": 1.6564, + "step": 7280 + }, + { + "epoch": 2.296121097445601, + "grad_norm": 2.0625, + "learning_rate": 0.0002, + "loss": 1.5515, + "step": 7281 + }, + { + "epoch": 2.2964364553768526, + "grad_norm": 1.8203125, + "learning_rate": 0.0002, + "loss": 1.501, + "step": 7282 + }, + { + "epoch": 2.296751813308105, + "grad_norm": 1.125, + "learning_rate": 0.0002, + "loss": 1.5933, + "step": 7283 + }, + { + "epoch": 2.2970671712393567, + "grad_norm": 0.9609375, + "learning_rate": 0.0002, + "loss": 1.5532, + "step": 7284 + }, + { + "epoch": 2.2973825291706085, + "grad_norm": 1.171875, + "learning_rate": 0.0002, + "loss": 1.7417, + "step": 7285 + }, + { + "epoch": 2.2976978871018607, + "grad_norm": 1.1640625, + "learning_rate": 0.0002, + "loss": 1.9446, + "step": 7286 + }, + { + "epoch": 2.2980132450331126, + "grad_norm": 1.28125, + "learning_rate": 0.0002, + "loss": 1.459, + "step": 7287 + }, + { + "epoch": 2.2983286029643644, + "grad_norm": 1.1953125, + "learning_rate": 0.0002, + "loss": 1.5775, + "step": 7288 + }, + { + "epoch": 2.2986439608956166, + "grad_norm": 1.28125, + "learning_rate": 0.0002, + "loss": 1.5067, + "step": 7289 + }, + { + "epoch": 2.2989593188268684, + "grad_norm": 1.328125, + "learning_rate": 0.0002, + "loss": 1.4306, + "step": 7290 + }, + { + "epoch": 2.2992746767581203, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 1.4445, + "step": 7291 + }, + { + "epoch": 2.2995900346893725, + "grad_norm": 1.1171875, + "learning_rate": 0.0002, + "loss": 1.5591, + "step": 7292 + }, + { + "epoch": 2.2999053926206243, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.0528, + "step": 7293 + }, + { + "epoch": 2.2999053926206243, + "eval_loss": 1.9156546592712402, + "eval_runtime": 149.0404, + "eval_samples_per_second": 6.71, + "eval_steps_per_second": 6.71, + "step": 7293 + }, + { + "epoch": 2.2999053926206243, + "mmlu_eval_accuracy": 0.5942842763708238, + "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, + "mmlu_eval_accuracy_anatomy": 0.5714285714285714, + "mmlu_eval_accuracy_astronomy": 0.6875, + "mmlu_eval_accuracy_business_ethics": 0.7272727272727273, + "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034, + "mmlu_eval_accuracy_college_biology": 0.8125, + "mmlu_eval_accuracy_college_chemistry": 0.375, + "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365, + "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454, + "mmlu_eval_accuracy_college_medicine": 0.5, + "mmlu_eval_accuracy_college_physics": 0.45454545454545453, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384, + "mmlu_eval_accuracy_econometrics": 0.4166666666666667, + "mmlu_eval_accuracy_electrical_engineering": 0.4375, + "mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637, + "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, + "mmlu_eval_accuracy_global_facts": 0.2, + "mmlu_eval_accuracy_high_school_biology": 0.59375, + "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365, + "mmlu_eval_accuracy_high_school_computer_science": 0.7777777777777778, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.6046511627906976, + "mmlu_eval_accuracy_high_school_mathematics": 0.3793103448275862, + "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, + "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, + "mmlu_eval_accuracy_high_school_psychology": 0.7833333333333333, + "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, + "mmlu_eval_accuracy_high_school_us_history": 0.8636363636363636, + "mmlu_eval_accuracy_high_school_world_history": 0.8076923076923077, + "mmlu_eval_accuracy_human_aging": 0.6956521739130435, + "mmlu_eval_accuracy_human_sexuality": 0.5, + "mmlu_eval_accuracy_international_law": 1.0, + "mmlu_eval_accuracy_jurisprudence": 0.36363636363636365, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.84, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.6976744186046512, + "mmlu_eval_accuracy_moral_disputes": 0.6052631578947368, + "mmlu_eval_accuracy_moral_scenarios": 0.34, + "mmlu_eval_accuracy_nutrition": 0.7272727272727273, + "mmlu_eval_accuracy_philosophy": 0.6470588235294118, + "mmlu_eval_accuracy_prehistory": 0.5714285714285714, + "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355, + "mmlu_eval_accuracy_professional_law": 0.40588235294117647, + "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806, + "mmlu_eval_accuracy_professional_psychology": 0.6376811594202898, + "mmlu_eval_accuracy_public_relations": 0.4166666666666667, + "mmlu_eval_accuracy_security_studies": 0.7037037037037037, + "mmlu_eval_accuracy_sociology": 0.9090909090909091, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.4444444444444444, + "mmlu_eval_accuracy_world_religions": 0.7894736842105263, + "mmlu_loss": 5.7915960504368655, + "step": 7293 + }, + { + "epoch": 2.300220750551876, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 0.9746, + "step": 7294 + }, + { + "epoch": 2.3005361084831284, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 0.9569, + "step": 7295 + }, + { + "epoch": 2.30085146641438, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 0.9734, + "step": 7296 + }, + { + "epoch": 2.3011668243456325, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 0.935, + "step": 7297 + }, + { + "epoch": 2.3014821822768843, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 0.8677, + "step": 7298 + }, + { + "epoch": 2.301797540208136, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 0.9475, + "step": 7299 + }, + { + "epoch": 2.3021128981393884, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 0.9181, + "step": 7300 + }, + { + "epoch": 2.30242825607064, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 1.04, + "step": 7301 + }, + { + "epoch": 2.302743614001892, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 0.9005, + "step": 7302 + }, + { + "epoch": 2.3030589719331442, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 0.8954, + "step": 7303 + }, + { + "epoch": 2.303374329864396, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 0.9448, + "step": 7304 + }, + { + "epoch": 2.303689687795648, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 0.9528, + "step": 7305 + }, + { + "epoch": 2.3040050457269, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 0.8649, + "step": 7306 + }, + { + "epoch": 2.304320403658152, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 0.8715, + "step": 7307 + }, + { + "epoch": 2.304635761589404, + "grad_norm": 0.404296875, + "learning_rate": 0.0002, + "loss": 0.888, + "step": 7308 + }, + { + "epoch": 2.304951119520656, + "grad_norm": 0.43359375, + "learning_rate": 0.0002, + "loss": 0.9335, + "step": 7309 + }, + { + "epoch": 2.305266477451908, + "grad_norm": 0.79296875, + "learning_rate": 0.0002, + "loss": 0.9908, + "step": 7310 + }, + { + "epoch": 2.30558183538316, + "grad_norm": 0.41796875, + "learning_rate": 0.0002, + "loss": 0.848, + "step": 7311 + }, + { + "epoch": 2.305897193314412, + "grad_norm": 0.443359375, + "learning_rate": 0.0002, + "loss": 0.8758, + "step": 7312 + }, + { + "epoch": 2.3062125512456637, + "grad_norm": 0.396484375, + "learning_rate": 0.0002, + "loss": 0.954, + "step": 7313 + }, + { + "epoch": 2.306527909176916, + "grad_norm": 0.462890625, + "learning_rate": 0.0002, + "loss": 1.0427, + "step": 7314 + }, + { + "epoch": 2.306843267108168, + "grad_norm": 0.51953125, + "learning_rate": 0.0002, + "loss": 0.9863, + "step": 7315 + }, + { + "epoch": 2.3071586250394196, + "grad_norm": 0.578125, + "learning_rate": 0.0002, + "loss": 0.9447, + "step": 7316 + }, + { + "epoch": 2.307473982970672, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 0.9709, + "step": 7317 + }, + { + "epoch": 2.3077893409019237, + "grad_norm": 0.51953125, + "learning_rate": 0.0002, + "loss": 0.7493, + "step": 7318 + }, + { + "epoch": 2.3081046988331755, + "grad_norm": 0.64453125, + "learning_rate": 0.0002, + "loss": 0.9664, + "step": 7319 + }, + { + "epoch": 2.3084200567644277, + "grad_norm": 0.58203125, + "learning_rate": 0.0002, + "loss": 0.9541, + "step": 7320 + }, + { + "epoch": 2.3087354146956796, + "grad_norm": 0.6171875, + "learning_rate": 0.0002, + "loss": 0.9071, + "step": 7321 + }, + { + "epoch": 2.3090507726269314, + "grad_norm": 0.98046875, + "learning_rate": 0.0002, + "loss": 0.9828, + "step": 7322 + }, + { + "epoch": 2.3093661305581836, + "grad_norm": 0.87890625, + "learning_rate": 0.0002, + "loss": 1.1064, + "step": 7323 + }, + { + "epoch": 2.3096814884894354, + "grad_norm": 0.85546875, + "learning_rate": 0.0002, + "loss": 1.0307, + "step": 7324 + }, + { + "epoch": 2.3099968464206873, + "grad_norm": 1.0859375, + "learning_rate": 0.0002, + "loss": 1.2181, + "step": 7325 + }, + { + "epoch": 2.3103122043519395, + "grad_norm": 1.03125, + "learning_rate": 0.0002, + "loss": 1.0722, + "step": 7326 + }, + { + "epoch": 2.3106275622831913, + "grad_norm": 0.91796875, + "learning_rate": 0.0002, + "loss": 1.1439, + "step": 7327 + }, + { + "epoch": 2.3109429202144436, + "grad_norm": 1.1953125, + "learning_rate": 0.0002, + "loss": 1.37, + "step": 7328 + }, + { + "epoch": 2.3112582781456954, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 1.2854, + "step": 7329 + }, + { + "epoch": 2.311573636076947, + "grad_norm": 1.25, + "learning_rate": 0.0002, + "loss": 1.2151, + "step": 7330 + }, + { + "epoch": 2.3118889940081995, + "grad_norm": 1.296875, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 7331 + }, + { + "epoch": 2.3122043519394513, + "grad_norm": 1.140625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 7332 + }, + { + "epoch": 2.312519709870703, + "grad_norm": 1.1171875, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 7333 + }, + { + "epoch": 2.3128350678019554, + "grad_norm": 1.15625, + "learning_rate": 0.0002, + "loss": 1.5765, + "step": 7334 + }, + { + "epoch": 2.313150425733207, + "grad_norm": 1.21875, + "learning_rate": 0.0002, + "loss": 1.6546, + "step": 7335 + }, + { + "epoch": 2.313465783664459, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 1.7383, + "step": 7336 + }, + { + "epoch": 2.3137811415957112, + "grad_norm": 1.3984375, + "learning_rate": 0.0002, + "loss": 1.7689, + "step": 7337 + }, + { + "epoch": 2.314096499526963, + "grad_norm": 0.8984375, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 7338 + }, + { + "epoch": 2.3144118574582153, + "grad_norm": 1.359375, + "learning_rate": 0.0002, + "loss": 2.0625, + "step": 7339 + }, + { + "epoch": 2.314727215389467, + "grad_norm": 0.93359375, + "learning_rate": 0.0002, + "loss": 1.2778, + "step": 7340 + }, + { + "epoch": 2.315042573320719, + "grad_norm": 1.5078125, + "learning_rate": 0.0002, + "loss": 1.8479, + "step": 7341 + }, + { + "epoch": 2.315357931251971, + "grad_norm": 1.015625, + "learning_rate": 0.0002, + "loss": 1.4846, + "step": 7342 + }, + { + "epoch": 2.315673289183223, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.0135, + "step": 7343 + }, + { + "epoch": 2.315988647114475, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 0.9617, + "step": 7344 + }, + { + "epoch": 2.316304005045727, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 0.9487, + "step": 7345 + }, + { + "epoch": 2.316619362976979, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 1.0799, + "step": 7346 + }, + { + "epoch": 2.3169347209082307, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 1.0452, + "step": 7347 + }, + { + "epoch": 2.317250078839483, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 0.8681, + "step": 7348 + }, + { + "epoch": 2.317565436770735, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 0.9299, + "step": 7349 + }, + { + "epoch": 2.3178807947019866, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 0.8736, + "step": 7350 + }, + { + "epoch": 2.318196152633239, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 0.9287, + "step": 7351 + }, + { + "epoch": 2.3185115105644907, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 0.8639, + "step": 7352 + }, + { + "epoch": 2.3188268684957425, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.9322, + "step": 7353 + }, + { + "epoch": 2.3191422264269947, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 0.9632, + "step": 7354 + }, + { + "epoch": 2.3194575843582466, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 0.8439, + "step": 7355 + }, + { + "epoch": 2.3197729422894984, + "grad_norm": 0.3671875, + "learning_rate": 0.0002, + "loss": 0.8983, + "step": 7356 + }, + { + "epoch": 2.3200883002207506, + "grad_norm": 0.34765625, + "learning_rate": 0.0002, + "loss": 0.8936, + "step": 7357 + }, + { + "epoch": 2.3204036581520024, + "grad_norm": 0.392578125, + "learning_rate": 0.0002, + "loss": 0.9625, + "step": 7358 + }, + { + "epoch": 2.3207190160832543, + "grad_norm": 0.3984375, + "learning_rate": 0.0002, + "loss": 0.8556, + "step": 7359 + }, + { + "epoch": 2.3210343740145065, + "grad_norm": 0.384765625, + "learning_rate": 0.0002, + "loss": 0.9778, + "step": 7360 + }, + { + "epoch": 2.3213497319457583, + "grad_norm": 0.412109375, + "learning_rate": 0.0002, + "loss": 0.9371, + "step": 7361 + }, + { + "epoch": 2.3216650898770106, + "grad_norm": 0.427734375, + "learning_rate": 0.0002, + "loss": 0.8999, + "step": 7362 + }, + { + "epoch": 2.3219804478082624, + "grad_norm": 0.4296875, + "learning_rate": 0.0002, + "loss": 0.9473, + "step": 7363 + }, + { + "epoch": 2.322295805739514, + "grad_norm": 0.50390625, + "learning_rate": 0.0002, + "loss": 0.9235, + "step": 7364 + }, + { + "epoch": 2.3226111636707665, + "grad_norm": 0.48828125, + "learning_rate": 0.0002, + "loss": 0.8692, + "step": 7365 + }, + { + "epoch": 2.3229265216020183, + "grad_norm": 0.451171875, + "learning_rate": 0.0002, + "loss": 0.7636, + "step": 7366 + }, + { + "epoch": 2.32324187953327, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 0.8402, + "step": 7367 + }, + { + "epoch": 2.3235572374645224, + "grad_norm": 0.50390625, + "learning_rate": 0.0002, + "loss": 0.9369, + "step": 7368 + }, + { + "epoch": 2.323872595395774, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 0.9867, + "step": 7369 + }, + { + "epoch": 2.324187953327026, + "grad_norm": 0.546875, + "learning_rate": 0.0002, + "loss": 0.8111, + "step": 7370 + }, + { + "epoch": 2.3245033112582782, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 0.8215, + "step": 7371 + }, + { + "epoch": 2.32481866918953, + "grad_norm": 0.69921875, + "learning_rate": 0.0002, + "loss": 0.9631, + "step": 7372 + }, + { + "epoch": 2.3251340271207823, + "grad_norm": 0.72265625, + "learning_rate": 0.0002, + "loss": 1.0455, + "step": 7373 + }, + { + "epoch": 2.325449385052034, + "grad_norm": 0.73828125, + "learning_rate": 0.0002, + "loss": 1.0608, + "step": 7374 + }, + { + "epoch": 2.325764742983286, + "grad_norm": 0.80078125, + "learning_rate": 0.0002, + "loss": 1.0635, + "step": 7375 + }, + { + "epoch": 2.326080100914538, + "grad_norm": 0.93359375, + "learning_rate": 0.0002, + "loss": 1.1726, + "step": 7376 + }, + { + "epoch": 2.32639545884579, + "grad_norm": 0.953125, + "learning_rate": 0.0002, + "loss": 1.2149, + "step": 7377 + }, + { + "epoch": 2.326710816777042, + "grad_norm": 0.94140625, + "learning_rate": 0.0002, + "loss": 1.1446, + "step": 7378 + }, + { + "epoch": 2.327026174708294, + "grad_norm": 0.97265625, + "learning_rate": 0.0002, + "loss": 1.2534, + "step": 7379 + }, + { + "epoch": 2.327341532639546, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.2526, + "step": 7380 + }, + { + "epoch": 2.3276568905707977, + "grad_norm": 1.2265625, + "learning_rate": 0.0002, + "loss": 1.4683, + "step": 7381 + }, + { + "epoch": 2.32797224850205, + "grad_norm": 1.1640625, + "learning_rate": 0.0002, + "loss": 1.3846, + "step": 7382 + }, + { + "epoch": 2.328287606433302, + "grad_norm": 1.03125, + "learning_rate": 0.0002, + "loss": 1.2538, + "step": 7383 + }, + { + "epoch": 2.3286029643645536, + "grad_norm": 1.2109375, + "learning_rate": 0.0002, + "loss": 1.5881, + "step": 7384 + }, + { + "epoch": 2.328918322295806, + "grad_norm": 1.5234375, + "learning_rate": 0.0002, + "loss": 2.0365, + "step": 7385 + }, + { + "epoch": 2.3292336802270577, + "grad_norm": 1.140625, + "learning_rate": 0.0002, + "loss": 1.5248, + "step": 7386 + }, + { + "epoch": 2.3295490381583095, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.513, + "step": 7387 + }, + { + "epoch": 2.3298643960895618, + "grad_norm": 1.4921875, + "learning_rate": 0.0002, + "loss": 1.5131, + "step": 7388 + }, + { + "epoch": 2.3301797540208136, + "grad_norm": 1.3984375, + "learning_rate": 0.0002, + "loss": 1.6088, + "step": 7389 + }, + { + "epoch": 2.3304951119520654, + "grad_norm": 0.91015625, + "learning_rate": 0.0002, + "loss": 1.2278, + "step": 7390 + }, + { + "epoch": 2.3308104698833176, + "grad_norm": 1.1640625, + "learning_rate": 0.0002, + "loss": 1.4743, + "step": 7391 + }, + { + "epoch": 2.3311258278145695, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 1.7193, + "step": 7392 + }, + { + "epoch": 2.3314411857458213, + "grad_norm": 0.240234375, + "learning_rate": 0.0002, + "loss": 1.0086, + "step": 7393 + }, + { + "epoch": 2.3317565436770735, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.9233, + "step": 7394 + }, + { + "epoch": 2.3320719016083253, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 0.9651, + "step": 7395 + }, + { + "epoch": 2.3323872595395776, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 0.9995, + "step": 7396 + }, + { + "epoch": 2.3327026174708294, + "grad_norm": 0.46875, + "learning_rate": 0.0002, + "loss": 1.0495, + "step": 7397 + }, + { + "epoch": 2.3330179754020812, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 0.8882, + "step": 7398 + }, + { + "epoch": 2.3333333333333335, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 0.9746, + "step": 7399 + }, + { + "epoch": 2.3336486912645853, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.9546, + "step": 7400 + }, + { + "epoch": 2.333964049195837, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 0.8908, + "step": 7401 + }, + { + "epoch": 2.3342794071270894, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 0.7883, + "step": 7402 + }, + { + "epoch": 2.334594765058341, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 0.8583, + "step": 7403 + }, + { + "epoch": 2.334910122989593, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 0.9469, + "step": 7404 + }, + { + "epoch": 2.3352254809208453, + "grad_norm": 0.34765625, + "learning_rate": 0.0002, + "loss": 0.9726, + "step": 7405 + }, + { + "epoch": 2.335540838852097, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 0.8802, + "step": 7406 + }, + { + "epoch": 2.3358561967833493, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 0.8663, + "step": 7407 + }, + { + "epoch": 2.336171554714601, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 0.906, + "step": 7408 + }, + { + "epoch": 2.336486912645853, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 0.7412, + "step": 7409 + }, + { + "epoch": 2.336802270577105, + "grad_norm": 0.3984375, + "learning_rate": 0.0002, + "loss": 0.8156, + "step": 7410 + }, + { + "epoch": 2.337117628508357, + "grad_norm": 0.38671875, + "learning_rate": 0.0002, + "loss": 0.946, + "step": 7411 + }, + { + "epoch": 2.337432986439609, + "grad_norm": 0.453125, + "learning_rate": 0.0002, + "loss": 0.9848, + "step": 7412 + }, + { + "epoch": 2.337748344370861, + "grad_norm": 0.416015625, + "learning_rate": 0.0002, + "loss": 0.8285, + "step": 7413 + }, + { + "epoch": 2.338063702302113, + "grad_norm": 0.40625, + "learning_rate": 0.0002, + "loss": 0.9143, + "step": 7414 + }, + { + "epoch": 2.3383790602333647, + "grad_norm": 0.44140625, + "learning_rate": 0.0002, + "loss": 0.911, + "step": 7415 + }, + { + "epoch": 2.338694418164617, + "grad_norm": 0.455078125, + "learning_rate": 0.0002, + "loss": 0.9736, + "step": 7416 + }, + { + "epoch": 2.339009776095869, + "grad_norm": 0.419921875, + "learning_rate": 0.0002, + "loss": 0.8328, + "step": 7417 + }, + { + "epoch": 2.3393251340271206, + "grad_norm": 0.46875, + "learning_rate": 0.0002, + "loss": 0.9348, + "step": 7418 + }, + { + "epoch": 2.339640491958373, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 0.9046, + "step": 7419 + }, + { + "epoch": 2.3399558498896247, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 0.7828, + "step": 7420 + }, + { + "epoch": 2.3402712078208765, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 0.9381, + "step": 7421 + }, + { + "epoch": 2.3405865657521288, + "grad_norm": 0.64453125, + "learning_rate": 0.0002, + "loss": 0.8372, + "step": 7422 + }, + { + "epoch": 2.3409019236833806, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 0.9148, + "step": 7423 + }, + { + "epoch": 2.3412172816146324, + "grad_norm": 0.69921875, + "learning_rate": 0.0002, + "loss": 1.0593, + "step": 7424 + }, + { + "epoch": 2.3415326395458846, + "grad_norm": 0.7265625, + "learning_rate": 0.0002, + "loss": 0.9754, + "step": 7425 + }, + { + "epoch": 2.3418479974771365, + "grad_norm": 0.77734375, + "learning_rate": 0.0002, + "loss": 1.1018, + "step": 7426 + }, + { + "epoch": 2.3421633554083887, + "grad_norm": 0.80078125, + "learning_rate": 0.0002, + "loss": 1.0636, + "step": 7427 + }, + { + "epoch": 2.3424787133396405, + "grad_norm": 0.86328125, + "learning_rate": 0.0002, + "loss": 1.1547, + "step": 7428 + }, + { + "epoch": 2.3427940712708923, + "grad_norm": 0.8046875, + "learning_rate": 0.0002, + "loss": 1.1816, + "step": 7429 + }, + { + "epoch": 2.3431094292021446, + "grad_norm": 0.94921875, + "learning_rate": 0.0002, + "loss": 1.1818, + "step": 7430 + }, + { + "epoch": 2.3434247871333964, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.2771, + "step": 7431 + }, + { + "epoch": 2.3437401450646482, + "grad_norm": 1.1953125, + "learning_rate": 0.0002, + "loss": 1.6676, + "step": 7432 + }, + { + "epoch": 2.3440555029959005, + "grad_norm": 1.015625, + "learning_rate": 0.0002, + "loss": 1.4674, + "step": 7433 + }, + { + "epoch": 2.3443708609271523, + "grad_norm": 0.9765625, + "learning_rate": 0.0002, + "loss": 1.4259, + "step": 7434 + }, + { + "epoch": 2.344686218858404, + "grad_norm": 1.59375, + "learning_rate": 0.0002, + "loss": 1.809, + "step": 7435 + }, + { + "epoch": 2.3450015767896564, + "grad_norm": 1.34375, + "learning_rate": 0.0002, + "loss": 1.7856, + "step": 7436 + }, + { + "epoch": 2.345316934720908, + "grad_norm": 1.6171875, + "learning_rate": 0.0002, + "loss": 1.931, + "step": 7437 + }, + { + "epoch": 2.3456322926521604, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.6762, + "step": 7438 + }, + { + "epoch": 2.3459476505834123, + "grad_norm": 1.0390625, + "learning_rate": 0.0002, + "loss": 1.2442, + "step": 7439 + }, + { + "epoch": 2.346263008514664, + "grad_norm": 1.1953125, + "learning_rate": 0.0002, + "loss": 1.1587, + "step": 7440 + }, + { + "epoch": 2.3465783664459163, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 1.4514, + "step": 7441 + }, + { + "epoch": 2.346893724377168, + "grad_norm": 1.5, + "learning_rate": 0.0002, + "loss": 1.8374, + "step": 7442 + }, + { + "epoch": 2.34720908230842, + "grad_norm": 0.2333984375, + "learning_rate": 0.0002, + "loss": 0.8932, + "step": 7443 + }, + { + "epoch": 2.347524440239672, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 1.0404, + "step": 7444 + }, + { + "epoch": 2.347839798170924, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 0.916, + "step": 7445 + }, + { + "epoch": 2.348155156102176, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 0.9304, + "step": 7446 + }, + { + "epoch": 2.348470514033428, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 0.9212, + "step": 7447 + }, + { + "epoch": 2.34878587196468, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.0309, + "step": 7448 + }, + { + "epoch": 2.3491012298959317, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 0.9233, + "step": 7449 + }, + { + "epoch": 2.349416587827184, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 0.9509, + "step": 7450 + }, + { + "epoch": 2.349731945758436, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.8454, + "step": 7451 + }, + { + "epoch": 2.3500473036896876, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.8866, + "step": 7452 + }, + { + "epoch": 2.35036266162094, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 0.9414, + "step": 7453 + }, + { + "epoch": 2.3506780195521917, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 0.9431, + "step": 7454 + }, + { + "epoch": 2.3509933774834435, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 0.9244, + "step": 7455 + }, + { + "epoch": 2.3513087354146958, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 0.9181, + "step": 7456 + }, + { + "epoch": 2.3516240933459476, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 0.8632, + "step": 7457 + }, + { + "epoch": 2.3519394512771994, + "grad_norm": 0.3515625, + "learning_rate": 0.0002, + "loss": 0.8544, + "step": 7458 + }, + { + "epoch": 2.3522548092084516, + "grad_norm": 0.373046875, + "learning_rate": 0.0002, + "loss": 0.8388, + "step": 7459 + }, + { + "epoch": 2.3525701671397035, + "grad_norm": 0.365234375, + "learning_rate": 0.0002, + "loss": 0.907, + "step": 7460 + }, + { + "epoch": 2.3528855250709557, + "grad_norm": 0.404296875, + "learning_rate": 0.0002, + "loss": 1.0603, + "step": 7461 + }, + { + "epoch": 2.3532008830022075, + "grad_norm": 0.373046875, + "learning_rate": 0.0002, + "loss": 0.8822, + "step": 7462 + }, + { + "epoch": 2.3535162409334593, + "grad_norm": 0.478515625, + "learning_rate": 0.0002, + "loss": 0.9147, + "step": 7463 + }, + { + "epoch": 2.3538315988647116, + "grad_norm": 0.47265625, + "learning_rate": 0.0002, + "loss": 0.9975, + "step": 7464 + }, + { + "epoch": 2.3541469567959634, + "grad_norm": 0.4765625, + "learning_rate": 0.0002, + "loss": 0.9432, + "step": 7465 + }, + { + "epoch": 2.3544623147272152, + "grad_norm": 0.50390625, + "learning_rate": 0.0002, + "loss": 0.8792, + "step": 7466 + }, + { + "epoch": 2.3547776726584675, + "grad_norm": 0.416015625, + "learning_rate": 0.0002, + "loss": 0.8217, + "step": 7467 + }, + { + "epoch": 2.3550930305897193, + "grad_norm": 0.50390625, + "learning_rate": 0.0002, + "loss": 0.7161, + "step": 7468 + }, + { + "epoch": 2.355408388520971, + "grad_norm": 0.5546875, + "learning_rate": 0.0002, + "loss": 0.9358, + "step": 7469 + }, + { + "epoch": 2.3557237464522234, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 0.9326, + "step": 7470 + }, + { + "epoch": 2.356039104383475, + "grad_norm": 0.56640625, + "learning_rate": 0.0002, + "loss": 0.8513, + "step": 7471 + }, + { + "epoch": 2.3563544623147274, + "grad_norm": 0.73828125, + "learning_rate": 0.0002, + "loss": 0.7964, + "step": 7472 + }, + { + "epoch": 2.3566698202459793, + "grad_norm": 0.8359375, + "learning_rate": 0.0002, + "loss": 1.0189, + "step": 7473 + }, + { + "epoch": 2.356985178177231, + "grad_norm": 0.7265625, + "learning_rate": 0.0002, + "loss": 1.0048, + "step": 7474 + }, + { + "epoch": 2.3573005361084833, + "grad_norm": 0.89453125, + "learning_rate": 0.0002, + "loss": 1.1887, + "step": 7475 + }, + { + "epoch": 2.357615894039735, + "grad_norm": 0.953125, + "learning_rate": 0.0002, + "loss": 1.0386, + "step": 7476 + }, + { + "epoch": 2.357931251970987, + "grad_norm": 0.87890625, + "learning_rate": 0.0002, + "loss": 1.0143, + "step": 7477 + }, + { + "epoch": 2.358246609902239, + "grad_norm": 1.8515625, + "learning_rate": 0.0002, + "loss": 1.3767, + "step": 7478 + }, + { + "epoch": 2.358561967833491, + "grad_norm": 1.3046875, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 7479 + }, + { + "epoch": 2.358877325764743, + "grad_norm": 0.88671875, + "learning_rate": 0.0002, + "loss": 1.1672, + "step": 7480 + }, + { + "epoch": 2.358877325764743, + "eval_loss": 1.829485297203064, + "eval_runtime": 148.9469, + "eval_samples_per_second": 6.714, + "eval_steps_per_second": 6.714, + "step": 7480 + }, + { + "epoch": 2.358877325764743, + "mmlu_eval_accuracy": 0.603596431711575, + "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, + "mmlu_eval_accuracy_anatomy": 0.5714285714285714, + "mmlu_eval_accuracy_astronomy": 0.6875, + "mmlu_eval_accuracy_business_ethics": 0.8181818181818182, + "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138, + "mmlu_eval_accuracy_college_biology": 0.8125, + "mmlu_eval_accuracy_college_chemistry": 0.375, + "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453, + "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453, + "mmlu_eval_accuracy_college_medicine": 0.5, + "mmlu_eval_accuracy_college_physics": 0.45454545454545453, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.5769230769230769, + "mmlu_eval_accuracy_econometrics": 0.4166666666666667, + "mmlu_eval_accuracy_electrical_engineering": 0.5, + "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244, + "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, + "mmlu_eval_accuracy_global_facts": 0.3, + "mmlu_eval_accuracy_high_school_biology": 0.5625, + "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.6046511627906976, + "mmlu_eval_accuracy_high_school_mathematics": 0.3793103448275862, + "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769, + "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, + "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667, + "mmlu_eval_accuracy_high_school_statistics": 0.4782608695652174, + "mmlu_eval_accuracy_high_school_us_history": 0.8636363636363636, + "mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693, + "mmlu_eval_accuracy_human_aging": 0.6521739130434783, + "mmlu_eval_accuracy_human_sexuality": 0.5, + "mmlu_eval_accuracy_international_law": 0.9230769230769231, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.7777777777777778, + "mmlu_eval_accuracy_machine_learning": 0.6363636363636364, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.68, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7325581395348837, + "mmlu_eval_accuracy_moral_disputes": 0.631578947368421, + "mmlu_eval_accuracy_moral_scenarios": 0.31, + "mmlu_eval_accuracy_nutrition": 0.7272727272727273, + "mmlu_eval_accuracy_philosophy": 0.6176470588235294, + "mmlu_eval_accuracy_prehistory": 0.5714285714285714, + "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355, + "mmlu_eval_accuracy_professional_law": 0.4117647058823529, + "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806, + "mmlu_eval_accuracy_professional_psychology": 0.5942028985507246, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7037037037037037, + "mmlu_eval_accuracy_sociology": 0.9090909090909091, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.5, + "mmlu_eval_accuracy_world_religions": 0.7894736842105263, + "mmlu_loss": 6.061446502425171, + "step": 7480 + }, + { + "epoch": 2.359192683695995, + "grad_norm": 0.9296875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 7481 + }, + { + "epoch": 2.359508041627247, + "grad_norm": 2.5625, + "learning_rate": 0.0002, + "loss": 1.5587, + "step": 7482 + }, + { + "epoch": 2.3598233995584987, + "grad_norm": 1.640625, + "learning_rate": 0.0002, + "loss": 1.9194, + "step": 7483 + }, + { + "epoch": 2.360138757489751, + "grad_norm": 1.609375, + "learning_rate": 0.0002, + "loss": 1.5774, + "step": 7484 + }, + { + "epoch": 2.360454115421003, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 1.6176, + "step": 7485 + }, + { + "epoch": 2.3607694733522546, + "grad_norm": 1.1640625, + "learning_rate": 0.0002, + "loss": 1.5573, + "step": 7486 + }, + { + "epoch": 2.361084831283507, + "grad_norm": 1.328125, + "learning_rate": 0.0002, + "loss": 1.4922, + "step": 7487 + }, + { + "epoch": 2.3614001892147587, + "grad_norm": 1.140625, + "learning_rate": 0.0002, + "loss": 1.7699, + "step": 7488 + }, + { + "epoch": 2.3617155471460105, + "grad_norm": 1.2265625, + "learning_rate": 0.0002, + "loss": 1.3838, + "step": 7489 + }, + { + "epoch": 2.3620309050772628, + "grad_norm": 1.234375, + "learning_rate": 0.0002, + "loss": 1.4824, + "step": 7490 + }, + { + "epoch": 2.3623462630085146, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 1.5263, + "step": 7491 + }, + { + "epoch": 2.3626616209397664, + "grad_norm": 1.234375, + "learning_rate": 0.0002, + "loss": 1.76, + "step": 7492 + }, + { + "epoch": 2.3629769788710187, + "grad_norm": 0.23046875, + "learning_rate": 0.0002, + "loss": 0.9722, + "step": 7493 + }, + { + "epoch": 2.3632923368022705, + "grad_norm": 0.244140625, + "learning_rate": 0.0002, + "loss": 0.9461, + "step": 7494 + }, + { + "epoch": 2.3636076947335227, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 0.9116, + "step": 7495 + }, + { + "epoch": 2.3639230526647745, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 0.9295, + "step": 7496 + }, + { + "epoch": 2.3642384105960264, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.0047, + "step": 7497 + }, + { + "epoch": 2.3645537685272786, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 0.9939, + "step": 7498 + }, + { + "epoch": 2.3648691264585304, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 0.8427, + "step": 7499 + }, + { + "epoch": 2.3651844843897822, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 0.8443, + "step": 7500 + }, + { + "epoch": 2.3654998423210345, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 0.9873, + "step": 7501 + }, + { + "epoch": 2.3658152002522863, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.9403, + "step": 7502 + }, + { + "epoch": 2.366130558183538, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 0.9515, + "step": 7503 + }, + { + "epoch": 2.3664459161147904, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 0.9622, + "step": 7504 + }, + { + "epoch": 2.366761274046042, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 0.9992, + "step": 7505 + }, + { + "epoch": 2.3670766319772945, + "grad_norm": 0.361328125, + "learning_rate": 0.0002, + "loss": 1.0975, + "step": 7506 + }, + { + "epoch": 2.3673919899085463, + "grad_norm": 0.349609375, + "learning_rate": 0.0002, + "loss": 0.8923, + "step": 7507 + }, + { + "epoch": 2.367707347839798, + "grad_norm": 0.353515625, + "learning_rate": 0.0002, + "loss": 0.8978, + "step": 7508 + }, + { + "epoch": 2.3680227057710503, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 0.8182, + "step": 7509 + }, + { + "epoch": 2.368338063702302, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 0.8282, + "step": 7510 + }, + { + "epoch": 2.368653421633554, + "grad_norm": 0.400390625, + "learning_rate": 0.0002, + "loss": 0.9037, + "step": 7511 + }, + { + "epoch": 2.3689687795648062, + "grad_norm": 0.40234375, + "learning_rate": 0.0002, + "loss": 0.8868, + "step": 7512 + }, + { + "epoch": 2.369284137496058, + "grad_norm": 0.40234375, + "learning_rate": 0.0002, + "loss": 0.8995, + "step": 7513 + }, + { + "epoch": 2.36959949542731, + "grad_norm": 0.43359375, + "learning_rate": 0.0002, + "loss": 0.8574, + "step": 7514 + }, + { + "epoch": 2.369914853358562, + "grad_norm": 0.451171875, + "learning_rate": 0.0002, + "loss": 0.9757, + "step": 7515 + }, + { + "epoch": 2.370230211289814, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 0.9738, + "step": 7516 + }, + { + "epoch": 2.3705455692210657, + "grad_norm": 0.4921875, + "learning_rate": 0.0002, + "loss": 0.7741, + "step": 7517 + }, + { + "epoch": 2.370860927152318, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 0.9372, + "step": 7518 + }, + { + "epoch": 2.37117628508357, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 0.8928, + "step": 7519 + }, + { + "epoch": 2.3714916430148216, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 0.8874, + "step": 7520 + }, + { + "epoch": 2.371807000946074, + "grad_norm": 0.63671875, + "learning_rate": 0.0002, + "loss": 1.0196, + "step": 7521 + }, + { + "epoch": 2.3721223588773257, + "grad_norm": 0.61328125, + "learning_rate": 0.0002, + "loss": 0.9811, + "step": 7522 + }, + { + "epoch": 2.3724377168085775, + "grad_norm": 0.69921875, + "learning_rate": 0.0002, + "loss": 0.9286, + "step": 7523 + }, + { + "epoch": 2.3727530747398298, + "grad_norm": 0.6484375, + "learning_rate": 0.0002, + "loss": 0.935, + "step": 7524 + }, + { + "epoch": 2.3730684326710816, + "grad_norm": 0.984375, + "learning_rate": 0.0002, + "loss": 1.2853, + "step": 7525 + }, + { + "epoch": 2.373383790602334, + "grad_norm": 0.87109375, + "learning_rate": 0.0002, + "loss": 1.1182, + "step": 7526 + }, + { + "epoch": 2.3736991485335857, + "grad_norm": 0.9296875, + "learning_rate": 0.0002, + "loss": 1.1074, + "step": 7527 + }, + { + "epoch": 2.3740145064648375, + "grad_norm": 1.2421875, + "learning_rate": 0.0002, + "loss": 1.1748, + "step": 7528 + }, + { + "epoch": 2.3743298643960897, + "grad_norm": 0.91015625, + "learning_rate": 0.0002, + "loss": 1.2624, + "step": 7529 + }, + { + "epoch": 2.3746452223273415, + "grad_norm": 1.140625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 7530 + }, + { + "epoch": 2.3749605802585934, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 1.5364, + "step": 7531 + }, + { + "epoch": 2.3752759381898456, + "grad_norm": 1.3203125, + "learning_rate": 0.0002, + "loss": 1.614, + "step": 7532 + }, + { + "epoch": 2.3755912961210974, + "grad_norm": 1.0546875, + "learning_rate": 0.0002, + "loss": 1.49, + "step": 7533 + }, + { + "epoch": 2.3759066540523492, + "grad_norm": 0.984375, + "learning_rate": 0.0002, + "loss": 1.4885, + "step": 7534 + }, + { + "epoch": 2.3762220119836015, + "grad_norm": 1.4296875, + "learning_rate": 0.0002, + "loss": 1.4048, + "step": 7535 + }, + { + "epoch": 2.3765373699148533, + "grad_norm": 1.2265625, + "learning_rate": 0.0002, + "loss": 1.3825, + "step": 7536 + }, + { + "epoch": 2.3768527278461056, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 1.6041, + "step": 7537 + }, + { + "epoch": 2.3771680857773574, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.7361, + "step": 7538 + }, + { + "epoch": 2.377483443708609, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.2513, + "step": 7539 + }, + { + "epoch": 2.3777988016398615, + "grad_norm": 1.03125, + "learning_rate": 0.0002, + "loss": 1.1607, + "step": 7540 + }, + { + "epoch": 2.3781141595711133, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 1.407, + "step": 7541 + }, + { + "epoch": 2.378429517502365, + "grad_norm": 1.1171875, + "learning_rate": 0.0002, + "loss": 1.2736, + "step": 7542 + }, + { + "epoch": 2.3787448754336173, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 0.8494, + "step": 7543 + }, + { + "epoch": 2.379060233364869, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 0.9592, + "step": 7544 + }, + { + "epoch": 2.379375591296121, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 0.9749, + "step": 7545 + }, + { + "epoch": 2.3796909492273732, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 0.84, + "step": 7546 + }, + { + "epoch": 2.380006307158625, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 1.0191, + "step": 7547 + }, + { + "epoch": 2.380321665089877, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 0.9478, + "step": 7548 + }, + { + "epoch": 2.380637023021129, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 0.836, + "step": 7549 + }, + { + "epoch": 2.380952380952381, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 0.9762, + "step": 7550 + }, + { + "epoch": 2.3812677388836327, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 0.9092, + "step": 7551 + }, + { + "epoch": 2.381583096814885, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 0.9562, + "step": 7552 + }, + { + "epoch": 2.381898454746137, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 0.9038, + "step": 7553 + }, + { + "epoch": 2.3822138126773886, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.855, + "step": 7554 + }, + { + "epoch": 2.382529170608641, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 0.9334, + "step": 7555 + }, + { + "epoch": 2.3828445285398927, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 0.8563, + "step": 7556 + }, + { + "epoch": 2.3831598864711445, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 0.7908, + "step": 7557 + }, + { + "epoch": 2.3834752444023968, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 0.8986, + "step": 7558 + }, + { + "epoch": 2.3837906023336486, + "grad_norm": 0.45703125, + "learning_rate": 0.0002, + "loss": 0.9558, + "step": 7559 + }, + { + "epoch": 2.384105960264901, + "grad_norm": 0.470703125, + "learning_rate": 0.0002, + "loss": 1.0544, + "step": 7560 + }, + { + "epoch": 2.3844213181961527, + "grad_norm": 0.466796875, + "learning_rate": 0.0002, + "loss": 1.0485, + "step": 7561 + }, + { + "epoch": 2.3847366761274045, + "grad_norm": 0.44140625, + "learning_rate": 0.0002, + "loss": 0.8721, + "step": 7562 + }, + { + "epoch": 2.3850520340586567, + "grad_norm": 0.412109375, + "learning_rate": 0.0002, + "loss": 0.886, + "step": 7563 + }, + { + "epoch": 2.3853673919899085, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 1.0695, + "step": 7564 + }, + { + "epoch": 2.3856827499211604, + "grad_norm": 0.546875, + "learning_rate": 0.0002, + "loss": 0.7889, + "step": 7565 + }, + { + "epoch": 2.3859981078524126, + "grad_norm": 0.427734375, + "learning_rate": 0.0002, + "loss": 0.8256, + "step": 7566 + }, + { + "epoch": 2.3863134657836644, + "grad_norm": 0.5390625, + "learning_rate": 0.0002, + "loss": 0.9173, + "step": 7567 + }, + { + "epoch": 2.3866288237149162, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 0.8523, + "step": 7568 + }, + { + "epoch": 2.3869441816461685, + "grad_norm": 0.62109375, + "learning_rate": 0.0002, + "loss": 0.9724, + "step": 7569 + }, + { + "epoch": 2.3872595395774203, + "grad_norm": 0.625, + "learning_rate": 0.0002, + "loss": 0.9281, + "step": 7570 + }, + { + "epoch": 2.3875748975086726, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 0.9499, + "step": 7571 + }, + { + "epoch": 2.3878902554399244, + "grad_norm": 0.703125, + "learning_rate": 0.0002, + "loss": 0.9268, + "step": 7572 + }, + { + "epoch": 2.388205613371176, + "grad_norm": 0.953125, + "learning_rate": 0.0002, + "loss": 1.0733, + "step": 7573 + }, + { + "epoch": 2.3885209713024285, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 0.9426, + "step": 7574 + }, + { + "epoch": 2.3888363292336803, + "grad_norm": 0.69921875, + "learning_rate": 0.0002, + "loss": 0.8898, + "step": 7575 + }, + { + "epoch": 2.389151687164932, + "grad_norm": 1.0546875, + "learning_rate": 0.0002, + "loss": 1.1345, + "step": 7576 + }, + { + "epoch": 2.3894670450961843, + "grad_norm": 0.9375, + "learning_rate": 0.0002, + "loss": 1.0706, + "step": 7577 + }, + { + "epoch": 2.389782403027436, + "grad_norm": 0.73828125, + "learning_rate": 0.0002, + "loss": 1.0433, + "step": 7578 + }, + { + "epoch": 2.390097760958688, + "grad_norm": 0.8828125, + "learning_rate": 0.0002, + "loss": 1.1549, + "step": 7579 + }, + { + "epoch": 2.3904131188899402, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.2801, + "step": 7580 + }, + { + "epoch": 2.390728476821192, + "grad_norm": 1.125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 7581 + }, + { + "epoch": 2.391043834752444, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 1.5896, + "step": 7582 + }, + { + "epoch": 2.391359192683696, + "grad_norm": 0.98828125, + "learning_rate": 0.0002, + "loss": 1.2487, + "step": 7583 + }, + { + "epoch": 2.391674550614948, + "grad_norm": 1.0390625, + "learning_rate": 0.0002, + "loss": 1.3869, + "step": 7584 + }, + { + "epoch": 2.3919899085461998, + "grad_norm": 1.375, + "learning_rate": 0.0002, + "loss": 1.5337, + "step": 7585 + }, + { + "epoch": 2.392305266477452, + "grad_norm": 1.5234375, + "learning_rate": 0.0002, + "loss": 1.5766, + "step": 7586 + }, + { + "epoch": 2.392620624408704, + "grad_norm": 1.3203125, + "learning_rate": 0.0002, + "loss": 1.6183, + "step": 7587 + }, + { + "epoch": 2.3929359823399556, + "grad_norm": 2.1875, + "learning_rate": 0.0002, + "loss": 1.7167, + "step": 7588 + }, + { + "epoch": 2.393251340271208, + "grad_norm": 1.4609375, + "learning_rate": 0.0002, + "loss": 1.9134, + "step": 7589 + }, + { + "epoch": 2.3935666982024597, + "grad_norm": 1.5078125, + "learning_rate": 0.0002, + "loss": 1.9463, + "step": 7590 + }, + { + "epoch": 2.3938820561337115, + "grad_norm": 1.0859375, + "learning_rate": 0.0002, + "loss": 1.5954, + "step": 7591 + }, + { + "epoch": 2.394197414064964, + "grad_norm": 1.4921875, + "learning_rate": 0.0002, + "loss": 2.0755, + "step": 7592 + }, + { + "epoch": 2.3945127719962156, + "grad_norm": 0.2392578125, + "learning_rate": 0.0002, + "loss": 1.0047, + "step": 7593 + }, + { + "epoch": 2.394828129927468, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 0.9463, + "step": 7594 + }, + { + "epoch": 2.3951434878587197, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 0.8819, + "step": 7595 + }, + { + "epoch": 2.3954588457899715, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 0.9308, + "step": 7596 + }, + { + "epoch": 2.3957742037212237, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.0293, + "step": 7597 + }, + { + "epoch": 2.3960895616524756, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 0.9684, + "step": 7598 + }, + { + "epoch": 2.3964049195837274, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 1.0376, + "step": 7599 + }, + { + "epoch": 2.3967202775149796, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 0.8273, + "step": 7600 + }, + { + "epoch": 2.3970356354462314, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 0.8873, + "step": 7601 + }, + { + "epoch": 2.3973509933774833, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 0.9768, + "step": 7602 + }, + { + "epoch": 2.3976663513087355, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 1.0033, + "step": 7603 + }, + { + "epoch": 2.3979817092399873, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 0.816, + "step": 7604 + }, + { + "epoch": 2.3982970671712396, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 0.9828, + "step": 7605 + }, + { + "epoch": 2.3986124251024914, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 0.7396, + "step": 7606 + }, + { + "epoch": 2.398927783033743, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 0.9944, + "step": 7607 + }, + { + "epoch": 2.3992431409649955, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 0.894, + "step": 7608 + }, + { + "epoch": 2.3995584988962473, + "grad_norm": 0.44921875, + "learning_rate": 0.0002, + "loss": 0.8781, + "step": 7609 + }, + { + "epoch": 2.399873856827499, + "grad_norm": 0.390625, + "learning_rate": 0.0002, + "loss": 0.8223, + "step": 7610 + }, + { + "epoch": 2.4001892147587514, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 0.7262, + "step": 7611 + }, + { + "epoch": 2.400504572690003, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 0.8413, + "step": 7612 + }, + { + "epoch": 2.400819930621255, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 0.9357, + "step": 7613 + }, + { + "epoch": 2.4011352885525072, + "grad_norm": 0.478515625, + "learning_rate": 0.0002, + "loss": 0.7877, + "step": 7614 + }, + { + "epoch": 2.401450646483759, + "grad_norm": 0.4921875, + "learning_rate": 0.0002, + "loss": 0.8567, + "step": 7615 + }, + { + "epoch": 2.401766004415011, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 1.0791, + "step": 7616 + }, + { + "epoch": 2.402081362346263, + "grad_norm": 0.484375, + "learning_rate": 0.0002, + "loss": 0.8226, + "step": 7617 + }, + { + "epoch": 2.402396720277515, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 0.8887, + "step": 7618 + }, + { + "epoch": 2.4027120782087668, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 1.0028, + "step": 7619 + }, + { + "epoch": 2.403027436140019, + "grad_norm": 0.60546875, + "learning_rate": 0.0002, + "loss": 0.8304, + "step": 7620 + }, + { + "epoch": 2.403342794071271, + "grad_norm": 0.53515625, + "learning_rate": 0.0002, + "loss": 0.7783, + "step": 7621 + }, + { + "epoch": 2.4036581520025226, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 0.8789, + "step": 7622 + }, + { + "epoch": 2.403973509933775, + "grad_norm": 0.6953125, + "learning_rate": 0.0002, + "loss": 0.8799, + "step": 7623 + }, + { + "epoch": 2.4042888678650267, + "grad_norm": 1.0546875, + "learning_rate": 0.0002, + "loss": 1.2089, + "step": 7624 + }, + { + "epoch": 2.4046042257962785, + "grad_norm": 0.8046875, + "learning_rate": 0.0002, + "loss": 1.1189, + "step": 7625 + }, + { + "epoch": 2.404919583727531, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 0.9455, + "step": 7626 + }, + { + "epoch": 2.4052349416587826, + "grad_norm": 1.03125, + "learning_rate": 0.0002, + "loss": 1.1839, + "step": 7627 + }, + { + "epoch": 2.405550299590035, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 7628 + }, + { + "epoch": 2.4058656575212867, + "grad_norm": 0.8984375, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 7629 + }, + { + "epoch": 2.4061810154525385, + "grad_norm": 0.96484375, + "learning_rate": 0.0002, + "loss": 1.2733, + "step": 7630 + }, + { + "epoch": 2.4064963733837907, + "grad_norm": 1.140625, + "learning_rate": 0.0002, + "loss": 1.6562, + "step": 7631 + }, + { + "epoch": 2.4068117313150426, + "grad_norm": 1.109375, + "learning_rate": 0.0002, + "loss": 1.392, + "step": 7632 + }, + { + "epoch": 2.4071270892462944, + "grad_norm": 1.875, + "learning_rate": 0.0002, + "loss": 1.501, + "step": 7633 + }, + { + "epoch": 2.4074424471775466, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 1.4056, + "step": 7634 + }, + { + "epoch": 2.4077578051087984, + "grad_norm": 0.90625, + "learning_rate": 0.0002, + "loss": 1.2765, + "step": 7635 + }, + { + "epoch": 2.4080731630400507, + "grad_norm": 1.0859375, + "learning_rate": 0.0002, + "loss": 1.2687, + "step": 7636 + }, + { + "epoch": 2.4083885209713025, + "grad_norm": 1.296875, + "learning_rate": 0.0002, + "loss": 1.5276, + "step": 7637 + }, + { + "epoch": 2.4087038789025543, + "grad_norm": 1.1640625, + "learning_rate": 0.0002, + "loss": 1.4545, + "step": 7638 + }, + { + "epoch": 2.4090192368338066, + "grad_norm": 1.4453125, + "learning_rate": 0.0002, + "loss": 1.9291, + "step": 7639 + }, + { + "epoch": 2.4093345947650584, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 1.4292, + "step": 7640 + }, + { + "epoch": 2.40964995269631, + "grad_norm": 1.59375, + "learning_rate": 0.0002, + "loss": 1.5782, + "step": 7641 + }, + { + "epoch": 2.4099653106275625, + "grad_norm": 1.59375, + "learning_rate": 0.0002, + "loss": 1.9149, + "step": 7642 + }, + { + "epoch": 2.4102806685588143, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 0.9519, + "step": 7643 + }, + { + "epoch": 2.410596026490066, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 0.9499, + "step": 7644 + }, + { + "epoch": 2.4109113844213184, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 0.9603, + "step": 7645 + }, + { + "epoch": 2.41122674235257, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 1.0118, + "step": 7646 + }, + { + "epoch": 2.411542100283822, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 0.9459, + "step": 7647 + }, + { + "epoch": 2.4118574582150742, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 0.9022, + "step": 7648 + }, + { + "epoch": 2.412172816146326, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 0.8683, + "step": 7649 + }, + { + "epoch": 2.412488174077578, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 0.9138, + "step": 7650 + }, + { + "epoch": 2.41280353200883, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.9359, + "step": 7651 + }, + { + "epoch": 2.413118889940082, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 0.9562, + "step": 7652 + }, + { + "epoch": 2.4134342478713338, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 1.01, + "step": 7653 + }, + { + "epoch": 2.413749605802586, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 0.9327, + "step": 7654 + }, + { + "epoch": 2.414064963733838, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 0.77, + "step": 7655 + }, + { + "epoch": 2.4143803216650896, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 0.8559, + "step": 7656 + }, + { + "epoch": 2.414695679596342, + "grad_norm": 0.431640625, + "learning_rate": 0.0002, + "loss": 1.0243, + "step": 7657 + }, + { + "epoch": 2.4150110375275937, + "grad_norm": 0.41015625, + "learning_rate": 0.0002, + "loss": 0.9407, + "step": 7658 + }, + { + "epoch": 2.415326395458846, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 0.9582, + "step": 7659 + }, + { + "epoch": 2.415641753390098, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 0.865, + "step": 7660 + }, + { + "epoch": 2.4159571113213496, + "grad_norm": 0.396484375, + "learning_rate": 0.0002, + "loss": 0.7698, + "step": 7661 + }, + { + "epoch": 2.416272469252602, + "grad_norm": 0.50390625, + "learning_rate": 0.0002, + "loss": 0.9611, + "step": 7662 + }, + { + "epoch": 2.4165878271838537, + "grad_norm": 0.458984375, + "learning_rate": 0.0002, + "loss": 0.8427, + "step": 7663 + }, + { + "epoch": 2.4169031851151055, + "grad_norm": 0.4140625, + "learning_rate": 0.0002, + "loss": 0.7956, + "step": 7664 + }, + { + "epoch": 2.4172185430463577, + "grad_norm": 0.439453125, + "learning_rate": 0.0002, + "loss": 0.8373, + "step": 7665 + }, + { + "epoch": 2.4175339009776096, + "grad_norm": 0.4921875, + "learning_rate": 0.0002, + "loss": 0.9224, + "step": 7666 + }, + { + "epoch": 2.4178492589088614, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 0.9818, + "step": 7667 + }, + { + "epoch": 2.4178492589088614, + "eval_loss": 1.783203363418579, + "eval_runtime": 149.1463, + "eval_samples_per_second": 6.705, + "eval_steps_per_second": 6.705, + "step": 7667 + }, + { + "epoch": 2.4178492589088614, + "mmlu_eval_accuracy": 0.5855704251884402, + "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, + "mmlu_eval_accuracy_anatomy": 0.7142857142857143, + "mmlu_eval_accuracy_astronomy": 0.625, + "mmlu_eval_accuracy_business_ethics": 0.7272727272727273, + "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034, + "mmlu_eval_accuracy_college_biology": 0.8125, + "mmlu_eval_accuracy_college_chemistry": 0.375, + "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365, + "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365, + "mmlu_eval_accuracy_college_medicine": 0.5, + "mmlu_eval_accuracy_college_physics": 0.45454545454545453, + "mmlu_eval_accuracy_computer_security": 0.36363636363636365, + "mmlu_eval_accuracy_conceptual_physics": 0.5769230769230769, + "mmlu_eval_accuracy_econometrics": 0.5, + "mmlu_eval_accuracy_electrical_engineering": 0.625, + "mmlu_eval_accuracy_elementary_mathematics": 0.2926829268292683, + "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, + "mmlu_eval_accuracy_global_facts": 0.2, + "mmlu_eval_accuracy_high_school_biology": 0.5625, + "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093, + "mmlu_eval_accuracy_high_school_mathematics": 0.3793103448275862, + "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769, + "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354, + "mmlu_eval_accuracy_high_school_psychology": 0.7833333333333333, + "mmlu_eval_accuracy_high_school_statistics": 0.5217391304347826, + "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, + "mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693, + "mmlu_eval_accuracy_human_aging": 0.6521739130434783, + "mmlu_eval_accuracy_human_sexuality": 0.5, + "mmlu_eval_accuracy_international_law": 0.8461538461538461, + "mmlu_eval_accuracy_jurisprudence": 0.36363636363636365, + "mmlu_eval_accuracy_logical_fallacies": 0.7777777777777778, + "mmlu_eval_accuracy_machine_learning": 0.2727272727272727, + "mmlu_eval_accuracy_management": 0.8181818181818182, + "mmlu_eval_accuracy_marketing": 0.68, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7093023255813954, + "mmlu_eval_accuracy_moral_disputes": 0.6052631578947368, + "mmlu_eval_accuracy_moral_scenarios": 0.3, + "mmlu_eval_accuracy_nutrition": 0.696969696969697, + "mmlu_eval_accuracy_philosophy": 0.6470588235294118, + "mmlu_eval_accuracy_prehistory": 0.5714285714285714, + "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355, + "mmlu_eval_accuracy_professional_law": 0.4235294117647059, + "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806, + "mmlu_eval_accuracy_professional_psychology": 0.6231884057971014, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7037037037037037, + "mmlu_eval_accuracy_sociology": 0.8636363636363636, + "mmlu_eval_accuracy_us_foreign_policy": 1.0, + "mmlu_eval_accuracy_virology": 0.5, + "mmlu_eval_accuracy_world_religions": 0.7894736842105263, + "mmlu_loss": 5.95468822637467, + "step": 7667 + }, + { + "epoch": 2.4181646168401136, + "grad_norm": 0.56640625, + "learning_rate": 0.0002, + "loss": 0.9643, + "step": 7668 + }, + { + "epoch": 2.4184799747713654, + "grad_norm": 0.62109375, + "learning_rate": 0.0002, + "loss": 0.976, + "step": 7669 + }, + { + "epoch": 2.4187953327026177, + "grad_norm": 0.625, + "learning_rate": 0.0002, + "loss": 0.8839, + "step": 7670 + }, + { + "epoch": 2.4191106906338695, + "grad_norm": 0.84765625, + "learning_rate": 0.0002, + "loss": 1.039, + "step": 7671 + }, + { + "epoch": 2.4194260485651213, + "grad_norm": 0.56640625, + "learning_rate": 0.0002, + "loss": 0.8378, + "step": 7672 + }, + { + "epoch": 2.4197414064963736, + "grad_norm": 0.72265625, + "learning_rate": 0.0002, + "loss": 0.9248, + "step": 7673 + }, + { + "epoch": 2.4200567644276254, + "grad_norm": 0.796875, + "learning_rate": 0.0002, + "loss": 1.0087, + "step": 7674 + }, + { + "epoch": 2.420372122358877, + "grad_norm": 0.703125, + "learning_rate": 0.0002, + "loss": 0.9373, + "step": 7675 + }, + { + "epoch": 2.4206874802901295, + "grad_norm": 0.84765625, + "learning_rate": 0.0002, + "loss": 1.1936, + "step": 7676 + }, + { + "epoch": 2.4210028382213813, + "grad_norm": 0.87890625, + "learning_rate": 0.0002, + "loss": 1.203, + "step": 7677 + }, + { + "epoch": 2.421318196152633, + "grad_norm": 0.83203125, + "learning_rate": 0.0002, + "loss": 1.1347, + "step": 7678 + }, + { + "epoch": 2.4216335540838854, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 7679 + }, + { + "epoch": 2.421948912015137, + "grad_norm": 0.8046875, + "learning_rate": 0.0002, + "loss": 1.1064, + "step": 7680 + }, + { + "epoch": 2.422264269946389, + "grad_norm": 1.125, + "learning_rate": 0.0002, + "loss": 1.4136, + "step": 7681 + }, + { + "epoch": 2.4225796278776413, + "grad_norm": 1.1875, + "learning_rate": 0.0002, + "loss": 1.5922, + "step": 7682 + }, + { + "epoch": 2.422894985808893, + "grad_norm": 1.6484375, + "learning_rate": 0.0002, + "loss": 1.4737, + "step": 7683 + }, + { + "epoch": 2.423210343740145, + "grad_norm": 1.421875, + "learning_rate": 0.0002, + "loss": 1.7401, + "step": 7684 + }, + { + "epoch": 2.423525701671397, + "grad_norm": 1.015625, + "learning_rate": 0.0002, + "loss": 1.3927, + "step": 7685 + }, + { + "epoch": 2.423841059602649, + "grad_norm": 1.40625, + "learning_rate": 0.0002, + "loss": 1.8876, + "step": 7686 + }, + { + "epoch": 2.4241564175339008, + "grad_norm": 1.2265625, + "learning_rate": 0.0002, + "loss": 1.8026, + "step": 7687 + }, + { + "epoch": 2.424471775465153, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 1.6733, + "step": 7688 + }, + { + "epoch": 2.424787133396405, + "grad_norm": 1.5234375, + "learning_rate": 0.0002, + "loss": 1.6763, + "step": 7689 + }, + { + "epoch": 2.4251024913276567, + "grad_norm": 0.953125, + "learning_rate": 0.0002, + "loss": 1.5379, + "step": 7690 + }, + { + "epoch": 2.425417849258909, + "grad_norm": 1.328125, + "learning_rate": 0.0002, + "loss": 1.5872, + "step": 7691 + }, + { + "epoch": 2.4257332071901607, + "grad_norm": 1.6796875, + "learning_rate": 0.0002, + "loss": 1.7578, + "step": 7692 + }, + { + "epoch": 2.426048565121413, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 0.9793, + "step": 7693 + }, + { + "epoch": 2.426363923052665, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 0.9093, + "step": 7694 + }, + { + "epoch": 2.4266792809839166, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 0.9475, + "step": 7695 + }, + { + "epoch": 2.426994638915169, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 0.9787, + "step": 7696 + }, + { + "epoch": 2.4273099968464207, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 0.9519, + "step": 7697 + }, + { + "epoch": 2.4276253547776725, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 0.9081, + "step": 7698 + }, + { + "epoch": 2.4279407127089248, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 0.9044, + "step": 7699 + }, + { + "epoch": 2.4282560706401766, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 0.9011, + "step": 7700 + }, + { + "epoch": 2.4285714285714284, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 0.8671, + "step": 7701 + }, + { + "epoch": 2.4288867865026806, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 0.9018, + "step": 7702 + }, + { + "epoch": 2.4292021444339325, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 1.01, + "step": 7703 + }, + { + "epoch": 2.4295175023651847, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 0.8835, + "step": 7704 + }, + { + "epoch": 2.4298328602964365, + "grad_norm": 0.34765625, + "learning_rate": 0.0002, + "loss": 0.8636, + "step": 7705 + }, + { + "epoch": 2.4301482182276883, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 0.8463, + "step": 7706 + }, + { + "epoch": 2.4304635761589406, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 0.9092, + "step": 7707 + }, + { + "epoch": 2.4307789340901924, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 0.8474, + "step": 7708 + }, + { + "epoch": 2.4310942920214442, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 0.8233, + "step": 7709 + }, + { + "epoch": 2.4314096499526965, + "grad_norm": 0.392578125, + "learning_rate": 0.0002, + "loss": 0.9489, + "step": 7710 + }, + { + "epoch": 2.4317250078839483, + "grad_norm": 0.37109375, + "learning_rate": 0.0002, + "loss": 0.7379, + "step": 7711 + }, + { + "epoch": 2.4320403658152, + "grad_norm": 0.42578125, + "learning_rate": 0.0002, + "loss": 1.0252, + "step": 7712 + }, + { + "epoch": 2.4323557237464524, + "grad_norm": 0.478515625, + "learning_rate": 0.0002, + "loss": 0.8875, + "step": 7713 + }, + { + "epoch": 2.432671081677704, + "grad_norm": 0.470703125, + "learning_rate": 0.0002, + "loss": 0.9376, + "step": 7714 + }, + { + "epoch": 2.432986439608956, + "grad_norm": 0.5390625, + "learning_rate": 0.0002, + "loss": 0.9742, + "step": 7715 + }, + { + "epoch": 2.4333017975402083, + "grad_norm": 0.48046875, + "learning_rate": 0.0002, + "loss": 0.8406, + "step": 7716 + }, + { + "epoch": 2.43361715547146, + "grad_norm": 0.69921875, + "learning_rate": 0.0002, + "loss": 0.917, + "step": 7717 + }, + { + "epoch": 2.433932513402712, + "grad_norm": 0.6953125, + "learning_rate": 0.0002, + "loss": 0.9368, + "step": 7718 + }, + { + "epoch": 2.434247871333964, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 0.8143, + "step": 7719 + }, + { + "epoch": 2.434563229265216, + "grad_norm": 0.671875, + "learning_rate": 0.0002, + "loss": 0.9367, + "step": 7720 + }, + { + "epoch": 2.4348785871964678, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 0.9455, + "step": 7721 + }, + { + "epoch": 2.43519394512772, + "grad_norm": 0.63671875, + "learning_rate": 0.0002, + "loss": 0.9892, + "step": 7722 + }, + { + "epoch": 2.435509303058972, + "grad_norm": 0.73046875, + "learning_rate": 0.0002, + "loss": 0.9535, + "step": 7723 + }, + { + "epoch": 2.4358246609902237, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 1.1094, + "step": 7724 + }, + { + "epoch": 2.436140018921476, + "grad_norm": 0.87109375, + "learning_rate": 0.0002, + "loss": 0.9398, + "step": 7725 + }, + { + "epoch": 2.4364553768527277, + "grad_norm": 0.79296875, + "learning_rate": 0.0002, + "loss": 1.0819, + "step": 7726 + }, + { + "epoch": 2.43677073478398, + "grad_norm": 0.89453125, + "learning_rate": 0.0002, + "loss": 1.2351, + "step": 7727 + }, + { + "epoch": 2.437086092715232, + "grad_norm": 0.9453125, + "learning_rate": 0.0002, + "loss": 1.277, + "step": 7728 + }, + { + "epoch": 2.4374014506464836, + "grad_norm": 1.125, + "learning_rate": 0.0002, + "loss": 1.5914, + "step": 7729 + }, + { + "epoch": 2.437716808577736, + "grad_norm": 0.93359375, + "learning_rate": 0.0002, + "loss": 1.4466, + "step": 7730 + }, + { + "epoch": 2.4380321665089877, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 1.3803, + "step": 7731 + }, + { + "epoch": 2.4383475244402395, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 1.3864, + "step": 7732 + }, + { + "epoch": 2.4386628823714918, + "grad_norm": 1.125, + "learning_rate": 0.0002, + "loss": 1.6625, + "step": 7733 + }, + { + "epoch": 2.4389782403027436, + "grad_norm": 1.2734375, + "learning_rate": 0.0002, + "loss": 1.5798, + "step": 7734 + }, + { + "epoch": 2.439293598233996, + "grad_norm": 1.234375, + "learning_rate": 0.0002, + "loss": 1.3809, + "step": 7735 + }, + { + "epoch": 2.4396089561652476, + "grad_norm": 1.09375, + "learning_rate": 0.0002, + "loss": 1.5698, + "step": 7736 + }, + { + "epoch": 2.4399243140964995, + "grad_norm": 1.4609375, + "learning_rate": 0.0002, + "loss": 1.6379, + "step": 7737 + }, + { + "epoch": 2.4402396720277517, + "grad_norm": 1.34375, + "learning_rate": 0.0002, + "loss": 1.7346, + "step": 7738 + }, + { + "epoch": 2.4405550299590035, + "grad_norm": 1.1640625, + "learning_rate": 0.0002, + "loss": 1.6497, + "step": 7739 + }, + { + "epoch": 2.4408703878902553, + "grad_norm": 1.7890625, + "learning_rate": 0.0002, + "loss": 1.6913, + "step": 7740 + }, + { + "epoch": 2.4411857458215076, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 1.3749, + "step": 7741 + }, + { + "epoch": 2.4415011037527594, + "grad_norm": 1.625, + "learning_rate": 0.0002, + "loss": 1.9109, + "step": 7742 + }, + { + "epoch": 2.4418164616840112, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 1.0326, + "step": 7743 + }, + { + "epoch": 2.4421318196152635, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 0.9812, + "step": 7744 + }, + { + "epoch": 2.4424471775465153, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 0.9156, + "step": 7745 + }, + { + "epoch": 2.442762535477767, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 1.0231, + "step": 7746 + }, + { + "epoch": 2.4430778934090194, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 0.9984, + "step": 7747 + }, + { + "epoch": 2.443393251340271, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 0.8326, + "step": 7748 + }, + { + "epoch": 2.443708609271523, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 0.9745, + "step": 7749 + }, + { + "epoch": 2.4440239672027753, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 0.8224, + "step": 7750 + }, + { + "epoch": 2.444339325134027, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 0.9575, + "step": 7751 + }, + { + "epoch": 2.444654683065279, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 0.9189, + "step": 7752 + }, + { + "epoch": 2.444970040996531, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 0.7703, + "step": 7753 + }, + { + "epoch": 2.445285398927783, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 0.8904, + "step": 7754 + }, + { + "epoch": 2.4456007568590348, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 0.8609, + "step": 7755 + }, + { + "epoch": 2.445916114790287, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 0.8338, + "step": 7756 + }, + { + "epoch": 2.446231472721539, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 0.7211, + "step": 7757 + }, + { + "epoch": 2.446546830652791, + "grad_norm": 0.3515625, + "learning_rate": 0.0002, + "loss": 0.9167, + "step": 7758 + }, + { + "epoch": 2.446862188584043, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 0.9023, + "step": 7759 + }, + { + "epoch": 2.4471775465152947, + "grad_norm": 0.380859375, + "learning_rate": 0.0002, + "loss": 0.938, + "step": 7760 + }, + { + "epoch": 2.447492904446547, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 0.8255, + "step": 7761 + }, + { + "epoch": 2.447808262377799, + "grad_norm": 0.431640625, + "learning_rate": 0.0002, + "loss": 0.8974, + "step": 7762 + }, + { + "epoch": 2.4481236203090506, + "grad_norm": 0.39453125, + "learning_rate": 0.0002, + "loss": 1.0309, + "step": 7763 + }, + { + "epoch": 2.448438978240303, + "grad_norm": 0.421875, + "learning_rate": 0.0002, + "loss": 0.8893, + "step": 7764 + }, + { + "epoch": 2.4487543361715547, + "grad_norm": 0.443359375, + "learning_rate": 0.0002, + "loss": 0.9851, + "step": 7765 + }, + { + "epoch": 2.4490696941028065, + "grad_norm": 0.451171875, + "learning_rate": 0.0002, + "loss": 0.8267, + "step": 7766 + }, + { + "epoch": 2.4493850520340588, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 0.9261, + "step": 7767 + }, + { + "epoch": 2.4497004099653106, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 0.9021, + "step": 7768 + }, + { + "epoch": 2.450015767896563, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 0.9634, + "step": 7769 + }, + { + "epoch": 2.4503311258278146, + "grad_norm": 0.62890625, + "learning_rate": 0.0002, + "loss": 0.9349, + "step": 7770 + }, + { + "epoch": 2.4506464837590665, + "grad_norm": 0.62890625, + "learning_rate": 0.0002, + "loss": 0.9682, + "step": 7771 + }, + { + "epoch": 2.4509618416903187, + "grad_norm": 0.65625, + "learning_rate": 0.0002, + "loss": 0.9024, + "step": 7772 + }, + { + "epoch": 2.4512771996215705, + "grad_norm": 0.77734375, + "learning_rate": 0.0002, + "loss": 1.1598, + "step": 7773 + }, + { + "epoch": 2.4515925575528223, + "grad_norm": 0.67578125, + "learning_rate": 0.0002, + "loss": 0.9776, + "step": 7774 + }, + { + "epoch": 2.4519079154840746, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 1.2207, + "step": 7775 + }, + { + "epoch": 2.4522232734153264, + "grad_norm": 0.87890625, + "learning_rate": 0.0002, + "loss": 0.9946, + "step": 7776 + }, + { + "epoch": 2.4525386313465782, + "grad_norm": 0.8046875, + "learning_rate": 0.0002, + "loss": 0.9721, + "step": 7777 + }, + { + "epoch": 2.4528539892778305, + "grad_norm": 0.90625, + "learning_rate": 0.0002, + "loss": 1.1333, + "step": 7778 + }, + { + "epoch": 2.4531693472090823, + "grad_norm": 1.109375, + "learning_rate": 0.0002, + "loss": 1.2878, + "step": 7779 + }, + { + "epoch": 2.453484705140334, + "grad_norm": 1.015625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 7780 + }, + { + "epoch": 2.4538000630715864, + "grad_norm": 1.25, + "learning_rate": 0.0002, + "loss": 1.406, + "step": 7781 + }, + { + "epoch": 2.454115421002838, + "grad_norm": 1.484375, + "learning_rate": 0.0002, + "loss": 1.4138, + "step": 7782 + }, + { + "epoch": 2.45443077893409, + "grad_norm": 1.5234375, + "learning_rate": 0.0002, + "loss": 1.8527, + "step": 7783 + }, + { + "epoch": 2.4547461368653423, + "grad_norm": 1.2421875, + "learning_rate": 0.0002, + "loss": 1.6085, + "step": 7784 + }, + { + "epoch": 2.455061494796594, + "grad_norm": 1.0, + "learning_rate": 0.0002, + "loss": 1.3896, + "step": 7785 + }, + { + "epoch": 2.455376852727846, + "grad_norm": 1.03125, + "learning_rate": 0.0002, + "loss": 1.452, + "step": 7786 + }, + { + "epoch": 2.455692210659098, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 1.6402, + "step": 7787 + }, + { + "epoch": 2.45600756859035, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 1.7447, + "step": 7788 + }, + { + "epoch": 2.456322926521602, + "grad_norm": 0.94921875, + "learning_rate": 0.0002, + "loss": 1.5746, + "step": 7789 + }, + { + "epoch": 2.456638284452854, + "grad_norm": 1.2109375, + "learning_rate": 0.0002, + "loss": 1.5146, + "step": 7790 + }, + { + "epoch": 2.456953642384106, + "grad_norm": 1.15625, + "learning_rate": 0.0002, + "loss": 1.7345, + "step": 7791 + }, + { + "epoch": 2.457269000315358, + "grad_norm": 0.90625, + "learning_rate": 0.0002, + "loss": 1.6745, + "step": 7792 + }, + { + "epoch": 2.45758435824661, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 0.9811, + "step": 7793 + }, + { + "epoch": 2.4578997161778617, + "grad_norm": 0.2470703125, + "learning_rate": 0.0002, + "loss": 0.8887, + "step": 7794 + }, + { + "epoch": 2.458215074109114, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 0.8761, + "step": 7795 + }, + { + "epoch": 2.458530432040366, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 1.0111, + "step": 7796 + }, + { + "epoch": 2.4588457899716176, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 0.9406, + "step": 7797 + }, + { + "epoch": 2.45916114790287, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 0.8732, + "step": 7798 + }, + { + "epoch": 2.4594765058341217, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 0.9905, + "step": 7799 + }, + { + "epoch": 2.4597918637653735, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 0.9939, + "step": 7800 + }, + { + "epoch": 2.4601072216966258, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.9247, + "step": 7801 + }, + { + "epoch": 2.4604225796278776, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.022, + "step": 7802 + }, + { + "epoch": 2.46073793755913, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 0.8928, + "step": 7803 + }, + { + "epoch": 2.4610532954903817, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 0.8341, + "step": 7804 + }, + { + "epoch": 2.4613686534216335, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 0.8706, + "step": 7805 + }, + { + "epoch": 2.4616840113528857, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 1.0302, + "step": 7806 + }, + { + "epoch": 2.4619993692841375, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 0.9544, + "step": 7807 + }, + { + "epoch": 2.4623147272153894, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 0.8702, + "step": 7808 + }, + { + "epoch": 2.4626300851466416, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 0.9033, + "step": 7809 + }, + { + "epoch": 2.4629454430778934, + "grad_norm": 0.37109375, + "learning_rate": 0.0002, + "loss": 0.9557, + "step": 7810 + }, + { + "epoch": 2.4632608010091452, + "grad_norm": 0.390625, + "learning_rate": 0.0002, + "loss": 0.8916, + "step": 7811 + }, + { + "epoch": 2.4635761589403975, + "grad_norm": 0.400390625, + "learning_rate": 0.0002, + "loss": 0.9074, + "step": 7812 + }, + { + "epoch": 2.4638915168716493, + "grad_norm": 0.443359375, + "learning_rate": 0.0002, + "loss": 0.8555, + "step": 7813 + }, + { + "epoch": 2.464206874802901, + "grad_norm": 0.431640625, + "learning_rate": 0.0002, + "loss": 0.9838, + "step": 7814 + }, + { + "epoch": 2.4645222327341534, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 0.9641, + "step": 7815 + }, + { + "epoch": 2.464837590665405, + "grad_norm": 0.443359375, + "learning_rate": 0.0002, + "loss": 0.7914, + "step": 7816 + }, + { + "epoch": 2.465152948596657, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 0.9415, + "step": 7817 + }, + { + "epoch": 2.4654683065279093, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 0.7804, + "step": 7818 + }, + { + "epoch": 2.465783664459161, + "grad_norm": 0.52734375, + "learning_rate": 0.0002, + "loss": 0.8766, + "step": 7819 + }, + { + "epoch": 2.466099022390413, + "grad_norm": 0.64453125, + "learning_rate": 0.0002, + "loss": 0.9378, + "step": 7820 + }, + { + "epoch": 2.466414380321665, + "grad_norm": 0.65625, + "learning_rate": 0.0002, + "loss": 1.0745, + "step": 7821 + }, + { + "epoch": 2.466729738252917, + "grad_norm": 0.62890625, + "learning_rate": 0.0002, + "loss": 0.9967, + "step": 7822 + }, + { + "epoch": 2.467045096184169, + "grad_norm": 0.703125, + "learning_rate": 0.0002, + "loss": 0.9689, + "step": 7823 + }, + { + "epoch": 2.467360454115421, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 1.0476, + "step": 7824 + }, + { + "epoch": 2.467675812046673, + "grad_norm": 0.76171875, + "learning_rate": 0.0002, + "loss": 1.1364, + "step": 7825 + }, + { + "epoch": 2.467991169977925, + "grad_norm": 0.9140625, + "learning_rate": 0.0002, + "loss": 0.9222, + "step": 7826 + }, + { + "epoch": 2.468306527909177, + "grad_norm": 0.91796875, + "learning_rate": 0.0002, + "loss": 1.2229, + "step": 7827 + }, + { + "epoch": 2.4686218858404287, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 1.2624, + "step": 7828 + }, + { + "epoch": 2.468937243771681, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 1.3614, + "step": 7829 + }, + { + "epoch": 2.469252601702933, + "grad_norm": 1.0546875, + "learning_rate": 0.0002, + "loss": 1.274, + "step": 7830 + }, + { + "epoch": 2.4695679596341846, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 1.2903, + "step": 7831 + }, + { + "epoch": 2.469883317565437, + "grad_norm": 1.234375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 7832 + }, + { + "epoch": 2.4701986754966887, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 1.3788, + "step": 7833 + }, + { + "epoch": 2.470514033427941, + "grad_norm": 1.5546875, + "learning_rate": 0.0002, + "loss": 1.8632, + "step": 7834 + }, + { + "epoch": 2.4708293913591928, + "grad_norm": 1.453125, + "learning_rate": 0.0002, + "loss": 1.4677, + "step": 7835 + }, + { + "epoch": 2.4711447492904446, + "grad_norm": 1.4296875, + "learning_rate": 0.0002, + "loss": 1.8556, + "step": 7836 + }, + { + "epoch": 2.471460107221697, + "grad_norm": 9.25, + "learning_rate": 0.0002, + "loss": 2.0571, + "step": 7837 + }, + { + "epoch": 2.4717754651529487, + "grad_norm": 1.359375, + "learning_rate": 0.0002, + "loss": 1.5772, + "step": 7838 + }, + { + "epoch": 2.4720908230842005, + "grad_norm": 1.0, + "learning_rate": 0.0002, + "loss": 1.2094, + "step": 7839 + }, + { + "epoch": 2.4724061810154527, + "grad_norm": 0.94140625, + "learning_rate": 0.0002, + "loss": 1.3984, + "step": 7840 + }, + { + "epoch": 2.4727215389467045, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 1.6007, + "step": 7841 + }, + { + "epoch": 2.4730368968779564, + "grad_norm": 1.3359375, + "learning_rate": 0.0002, + "loss": 1.9602, + "step": 7842 + }, + { + "epoch": 2.4733522548092086, + "grad_norm": 0.248046875, + "learning_rate": 0.0002, + "loss": 0.9773, + "step": 7843 + }, + { + "epoch": 2.4736676127404604, + "grad_norm": 0.2412109375, + "learning_rate": 0.0002, + "loss": 0.9279, + "step": 7844 + }, + { + "epoch": 2.4739829706717122, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 0.9607, + "step": 7845 + }, + { + "epoch": 2.4742983286029645, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.0287, + "step": 7846 + }, + { + "epoch": 2.4746136865342163, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.9673, + "step": 7847 + }, + { + "epoch": 2.474929044465468, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 0.9053, + "step": 7848 + }, + { + "epoch": 2.4752444023967204, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 0.929, + "step": 7849 + }, + { + "epoch": 2.475559760327972, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 0.8858, + "step": 7850 + }, + { + "epoch": 2.475875118259224, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.9449, + "step": 7851 + }, + { + "epoch": 2.4761904761904763, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 0.8782, + "step": 7852 + }, + { + "epoch": 2.476505834121728, + "grad_norm": 0.34765625, + "learning_rate": 0.0002, + "loss": 1.0276, + "step": 7853 + }, + { + "epoch": 2.47682119205298, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 0.92, + "step": 7854 + }, + { + "epoch": 2.47682119205298, + "eval_loss": 1.789473533630371, + "eval_runtime": 148.7438, + "eval_samples_per_second": 6.723, + "eval_steps_per_second": 6.723, + "step": 7854 + }, + { + "epoch": 2.47682119205298, + "mmlu_eval_accuracy": 0.57962758324363, + "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, + "mmlu_eval_accuracy_anatomy": 0.5714285714285714, + "mmlu_eval_accuracy_astronomy": 0.625, + "mmlu_eval_accuracy_business_ethics": 0.6363636363636364, + "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034, + "mmlu_eval_accuracy_college_biology": 0.75, + "mmlu_eval_accuracy_college_chemistry": 0.25, + "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365, + "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454, + "mmlu_eval_accuracy_college_medicine": 0.5, + "mmlu_eval_accuracy_college_physics": 0.45454545454545453, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.5769230769230769, + "mmlu_eval_accuracy_econometrics": 0.5, + "mmlu_eval_accuracy_electrical_engineering": 0.5625, + "mmlu_eval_accuracy_elementary_mathematics": 0.3170731707317073, + "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, + "mmlu_eval_accuracy_global_facts": 0.2, + "mmlu_eval_accuracy_high_school_biology": 0.53125, + "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325, + "mmlu_eval_accuracy_high_school_mathematics": 0.41379310344827586, + "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539, + "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, + "mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_statistics": 0.5217391304347826, + "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, + "mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693, + "mmlu_eval_accuracy_human_aging": 0.6956521739130435, + "mmlu_eval_accuracy_human_sexuality": 0.5, + "mmlu_eval_accuracy_international_law": 0.9230769230769231, + "mmlu_eval_accuracy_jurisprudence": 0.2727272727272727, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.72, + "mmlu_eval_accuracy_medical_genetics": 1.0, + "mmlu_eval_accuracy_miscellaneous": 0.686046511627907, + "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842, + "mmlu_eval_accuracy_moral_scenarios": 0.3, + "mmlu_eval_accuracy_nutrition": 0.7272727272727273, + "mmlu_eval_accuracy_philosophy": 0.5882352941176471, + "mmlu_eval_accuracy_prehistory": 0.6285714285714286, + "mmlu_eval_accuracy_professional_accounting": 0.41935483870967744, + "mmlu_eval_accuracy_professional_law": 0.4294117647058823, + "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516, + "mmlu_eval_accuracy_professional_psychology": 0.5942028985507246, + "mmlu_eval_accuracy_public_relations": 0.3333333333333333, + "mmlu_eval_accuracy_security_studies": 0.6666666666666666, + "mmlu_eval_accuracy_sociology": 0.9090909090909091, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.4444444444444444, + "mmlu_eval_accuracy_world_religions": 0.8421052631578947, + "mmlu_loss": 5.7441266093824055, + "step": 7854 + }, + { + "epoch": 2.477136549984232, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 0.8908, + "step": 7855 + }, + { + "epoch": 2.477451907915484, + "grad_norm": 0.3515625, + "learning_rate": 0.0002, + "loss": 0.8912, + "step": 7856 + }, + { + "epoch": 2.4777672658467362, + "grad_norm": 0.34765625, + "learning_rate": 0.0002, + "loss": 0.9491, + "step": 7857 + }, + { + "epoch": 2.478082623777988, + "grad_norm": 0.353515625, + "learning_rate": 0.0002, + "loss": 1.0742, + "step": 7858 + }, + { + "epoch": 2.47839798170924, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 1.032, + "step": 7859 + }, + { + "epoch": 2.478713339640492, + "grad_norm": 0.373046875, + "learning_rate": 0.0002, + "loss": 0.831, + "step": 7860 + }, + { + "epoch": 2.479028697571744, + "grad_norm": 0.443359375, + "learning_rate": 0.0002, + "loss": 0.918, + "step": 7861 + }, + { + "epoch": 2.4793440555029957, + "grad_norm": 0.412109375, + "learning_rate": 0.0002, + "loss": 0.8258, + "step": 7862 + }, + { + "epoch": 2.479659413434248, + "grad_norm": 0.421875, + "learning_rate": 0.0002, + "loss": 0.8237, + "step": 7863 + }, + { + "epoch": 2.4799747713655, + "grad_norm": 0.4375, + "learning_rate": 0.0002, + "loss": 0.7866, + "step": 7864 + }, + { + "epoch": 2.4802901292967516, + "grad_norm": 0.46484375, + "learning_rate": 0.0002, + "loss": 0.9076, + "step": 7865 + }, + { + "epoch": 2.480605487228004, + "grad_norm": 0.447265625, + "learning_rate": 0.0002, + "loss": 0.8832, + "step": 7866 + }, + { + "epoch": 2.4809208451592557, + "grad_norm": 0.474609375, + "learning_rate": 0.0002, + "loss": 0.904, + "step": 7867 + }, + { + "epoch": 2.481236203090508, + "grad_norm": 0.5703125, + "learning_rate": 0.0002, + "loss": 0.9218, + "step": 7868 + }, + { + "epoch": 2.4815515610217598, + "grad_norm": 0.52734375, + "learning_rate": 0.0002, + "loss": 0.8271, + "step": 7869 + }, + { + "epoch": 2.4818669189530116, + "grad_norm": 0.75, + "learning_rate": 0.0002, + "loss": 0.9544, + "step": 7870 + }, + { + "epoch": 2.482182276884264, + "grad_norm": 0.6015625, + "learning_rate": 0.0002, + "loss": 0.909, + "step": 7871 + }, + { + "epoch": 2.4824976348155157, + "grad_norm": 0.6796875, + "learning_rate": 0.0002, + "loss": 0.8757, + "step": 7872 + }, + { + "epoch": 2.4828129927467675, + "grad_norm": 0.7734375, + "learning_rate": 0.0002, + "loss": 0.9604, + "step": 7873 + }, + { + "epoch": 2.4831283506780197, + "grad_norm": 0.74609375, + "learning_rate": 0.0002, + "loss": 1.1248, + "step": 7874 + }, + { + "epoch": 2.4834437086092715, + "grad_norm": 0.80078125, + "learning_rate": 0.0002, + "loss": 1.0242, + "step": 7875 + }, + { + "epoch": 2.4837590665405234, + "grad_norm": 1.078125, + "learning_rate": 0.0002, + "loss": 1.1504, + "step": 7876 + }, + { + "epoch": 2.4840744244717756, + "grad_norm": 1.0, + "learning_rate": 0.0002, + "loss": 1.0755, + "step": 7877 + }, + { + "epoch": 2.4843897824030274, + "grad_norm": 0.984375, + "learning_rate": 0.0002, + "loss": 1.1165, + "step": 7878 + }, + { + "epoch": 2.4847051403342792, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 1.2598, + "step": 7879 + }, + { + "epoch": 2.4850204982655315, + "grad_norm": 1.078125, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 7880 + }, + { + "epoch": 2.4853358561967833, + "grad_norm": 1.34375, + "learning_rate": 0.0002, + "loss": 1.6404, + "step": 7881 + }, + { + "epoch": 2.485651214128035, + "grad_norm": 0.9609375, + "learning_rate": 0.0002, + "loss": 1.2929, + "step": 7882 + }, + { + "epoch": 2.4859665720592874, + "grad_norm": 1.2890625, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 7883 + }, + { + "epoch": 2.486281929990539, + "grad_norm": 0.9921875, + "learning_rate": 0.0002, + "loss": 1.3896, + "step": 7884 + }, + { + "epoch": 2.486597287921791, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 1.4432, + "step": 7885 + }, + { + "epoch": 2.4869126458530433, + "grad_norm": 1.2265625, + "learning_rate": 0.0002, + "loss": 1.5605, + "step": 7886 + }, + { + "epoch": 2.487228003784295, + "grad_norm": 0.94140625, + "learning_rate": 0.0002, + "loss": 1.4261, + "step": 7887 + }, + { + "epoch": 2.487543361715547, + "grad_norm": 0.984375, + "learning_rate": 0.0002, + "loss": 1.6646, + "step": 7888 + }, + { + "epoch": 2.487858719646799, + "grad_norm": 1.125, + "learning_rate": 0.0002, + "loss": 1.6179, + "step": 7889 + }, + { + "epoch": 2.488174077578051, + "grad_norm": 0.95703125, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 7890 + }, + { + "epoch": 2.4884894355093032, + "grad_norm": 1.4296875, + "learning_rate": 0.0002, + "loss": 1.7171, + "step": 7891 + }, + { + "epoch": 2.488804793440555, + "grad_norm": 0.9140625, + "learning_rate": 0.0002, + "loss": 1.4005, + "step": 7892 + }, + { + "epoch": 2.489120151371807, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 0.9564, + "step": 7893 + }, + { + "epoch": 2.489435509303059, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 0.8884, + "step": 7894 + }, + { + "epoch": 2.489750867234311, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 0.8639, + "step": 7895 + }, + { + "epoch": 2.4900662251655628, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 0.9094, + "step": 7896 + }, + { + "epoch": 2.490381583096815, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 0.8879, + "step": 7897 + }, + { + "epoch": 2.490696941028067, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 0.9756, + "step": 7898 + }, + { + "epoch": 2.4910122989593186, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 0.9972, + "step": 7899 + }, + { + "epoch": 2.491327656890571, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 0.968, + "step": 7900 + }, + { + "epoch": 2.4916430148218227, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 0.9145, + "step": 7901 + }, + { + "epoch": 2.491958372753075, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 0.8573, + "step": 7902 + }, + { + "epoch": 2.492273730684327, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 1.0262, + "step": 7903 + }, + { + "epoch": 2.4925890886155786, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 0.8809, + "step": 7904 + }, + { + "epoch": 2.492904446546831, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 0.8741, + "step": 7905 + }, + { + "epoch": 2.4932198044780827, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 0.8983, + "step": 7906 + }, + { + "epoch": 2.4935351624093345, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 0.8085, + "step": 7907 + }, + { + "epoch": 2.4938505203405867, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 0.959, + "step": 7908 + }, + { + "epoch": 2.4941658782718386, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 0.9094, + "step": 7909 + }, + { + "epoch": 2.4944812362030904, + "grad_norm": 0.388671875, + "learning_rate": 0.0002, + "loss": 0.9378, + "step": 7910 + }, + { + "epoch": 2.4947965941343426, + "grad_norm": 0.3828125, + "learning_rate": 0.0002, + "loss": 0.9263, + "step": 7911 + }, + { + "epoch": 2.4951119520655944, + "grad_norm": 0.431640625, + "learning_rate": 0.0002, + "loss": 0.9257, + "step": 7912 + }, + { + "epoch": 2.4954273099968463, + "grad_norm": 0.3671875, + "learning_rate": 0.0002, + "loss": 0.7187, + "step": 7913 + }, + { + "epoch": 2.4957426679280985, + "grad_norm": 0.408203125, + "learning_rate": 0.0002, + "loss": 0.7735, + "step": 7914 + }, + { + "epoch": 2.4960580258593503, + "grad_norm": 0.482421875, + "learning_rate": 0.0002, + "loss": 0.8638, + "step": 7915 + }, + { + "epoch": 2.496373383790602, + "grad_norm": 0.46484375, + "learning_rate": 0.0002, + "loss": 0.831, + "step": 7916 + }, + { + "epoch": 2.4966887417218544, + "grad_norm": 0.494140625, + "learning_rate": 0.0002, + "loss": 0.9582, + "step": 7917 + }, + { + "epoch": 2.497004099653106, + "grad_norm": 0.515625, + "learning_rate": 0.0002, + "loss": 0.8631, + "step": 7918 + }, + { + "epoch": 2.497319457584358, + "grad_norm": 0.5390625, + "learning_rate": 0.0002, + "loss": 0.8059, + "step": 7919 + }, + { + "epoch": 2.4976348155156103, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 0.7739, + "step": 7920 + }, + { + "epoch": 2.497950173446862, + "grad_norm": 0.578125, + "learning_rate": 0.0002, + "loss": 0.9261, + "step": 7921 + }, + { + "epoch": 2.498265531378114, + "grad_norm": 0.6484375, + "learning_rate": 0.0002, + "loss": 0.945, + "step": 7922 + }, + { + "epoch": 2.498580889309366, + "grad_norm": 0.859375, + "learning_rate": 0.0002, + "loss": 1.1198, + "step": 7923 + }, + { + "epoch": 2.498896247240618, + "grad_norm": 0.70703125, + "learning_rate": 0.0002, + "loss": 0.8888, + "step": 7924 + }, + { + "epoch": 2.4992116051718702, + "grad_norm": 0.98046875, + "learning_rate": 0.0002, + "loss": 1.1008, + "step": 7925 + }, + { + "epoch": 2.499526963103122, + "grad_norm": 0.84375, + "learning_rate": 0.0002, + "loss": 1.0392, + "step": 7926 + }, + { + "epoch": 2.499842321034374, + "grad_norm": 0.8984375, + "learning_rate": 0.0002, + "loss": 1.1142, + "step": 7927 + }, + { + "epoch": 2.500157678965626, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 1.1005, + "step": 7928 + }, + { + "epoch": 2.500473036896878, + "grad_norm": 0.9453125, + "learning_rate": 0.0002, + "loss": 1.1397, + "step": 7929 + }, + { + "epoch": 2.50078839482813, + "grad_norm": 1.3046875, + "learning_rate": 0.0002, + "loss": 1.5638, + "step": 7930 + }, + { + "epoch": 2.501103752759382, + "grad_norm": 1.09375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 7931 + }, + { + "epoch": 2.501419110690634, + "grad_norm": 1.15625, + "learning_rate": 0.0002, + "loss": 1.5628, + "step": 7932 + }, + { + "epoch": 2.501734468621886, + "grad_norm": 1.421875, + "learning_rate": 0.0002, + "loss": 1.5438, + "step": 7933 + }, + { + "epoch": 2.502049826553138, + "grad_norm": 1.28125, + "learning_rate": 0.0002, + "loss": 1.4728, + "step": 7934 + }, + { + "epoch": 2.5023651844843897, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 1.539, + "step": 7935 + }, + { + "epoch": 2.502680542415642, + "grad_norm": 1.578125, + "learning_rate": 0.0002, + "loss": 1.9501, + "step": 7936 + }, + { + "epoch": 2.502995900346894, + "grad_norm": 1.3046875, + "learning_rate": 0.0002, + "loss": 1.6575, + "step": 7937 + }, + { + "epoch": 2.5033112582781456, + "grad_norm": 1.171875, + "learning_rate": 0.0002, + "loss": 1.4835, + "step": 7938 + }, + { + "epoch": 2.503626616209398, + "grad_norm": 1.0546875, + "learning_rate": 0.0002, + "loss": 1.6458, + "step": 7939 + }, + { + "epoch": 2.5039419741406497, + "grad_norm": 1.15625, + "learning_rate": 0.0002, + "loss": 1.7493, + "step": 7940 + }, + { + "epoch": 2.5042573320719015, + "grad_norm": 1.9453125, + "learning_rate": 0.0002, + "loss": 1.5446, + "step": 7941 + }, + { + "epoch": 2.5045726900031537, + "grad_norm": 0.98046875, + "learning_rate": 0.0002, + "loss": 1.5617, + "step": 7942 + }, + { + "epoch": 2.5048880479344056, + "grad_norm": 0.2373046875, + "learning_rate": 0.0002, + "loss": 1.0086, + "step": 7943 + }, + { + "epoch": 2.5052034058656574, + "grad_norm": 0.248046875, + "learning_rate": 0.0002, + "loss": 0.9482, + "step": 7944 + }, + { + "epoch": 2.5055187637969096, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 1.0015, + "step": 7945 + }, + { + "epoch": 2.5058341217281614, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 1.0, + "step": 7946 + }, + { + "epoch": 2.5061494796594133, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 1.0413, + "step": 7947 + }, + { + "epoch": 2.5064648375906655, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 1.0158, + "step": 7948 + }, + { + "epoch": 2.5067801955219173, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 0.924, + "step": 7949 + }, + { + "epoch": 2.507095553453169, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 0.8968, + "step": 7950 + }, + { + "epoch": 2.5074109113844214, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.8921, + "step": 7951 + }, + { + "epoch": 2.507726269315673, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 0.8466, + "step": 7952 + }, + { + "epoch": 2.508041627246925, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 0.958, + "step": 7953 + }, + { + "epoch": 2.5083569851781773, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 0.9355, + "step": 7954 + }, + { + "epoch": 2.508672343109429, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 0.8965, + "step": 7955 + }, + { + "epoch": 2.508987701040681, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 0.8837, + "step": 7956 + }, + { + "epoch": 2.509303058971933, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 0.9527, + "step": 7957 + }, + { + "epoch": 2.509618416903185, + "grad_norm": 0.380859375, + "learning_rate": 0.0002, + "loss": 0.9508, + "step": 7958 + }, + { + "epoch": 2.5099337748344372, + "grad_norm": 0.3671875, + "learning_rate": 0.0002, + "loss": 0.8579, + "step": 7959 + }, + { + "epoch": 2.510249132765689, + "grad_norm": 0.63671875, + "learning_rate": 0.0002, + "loss": 0.9101, + "step": 7960 + }, + { + "epoch": 2.510564490696941, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 0.8283, + "step": 7961 + }, + { + "epoch": 2.510879848628193, + "grad_norm": 0.427734375, + "learning_rate": 0.0002, + "loss": 0.8702, + "step": 7962 + }, + { + "epoch": 2.511195206559445, + "grad_norm": 0.4453125, + "learning_rate": 0.0002, + "loss": 0.984, + "step": 7963 + }, + { + "epoch": 2.511510564490697, + "grad_norm": 0.494140625, + "learning_rate": 0.0002, + "loss": 1.0321, + "step": 7964 + }, + { + "epoch": 2.511825922421949, + "grad_norm": 0.470703125, + "learning_rate": 0.0002, + "loss": 1.0057, + "step": 7965 + }, + { + "epoch": 2.512141280353201, + "grad_norm": 0.51953125, + "learning_rate": 0.0002, + "loss": 0.9354, + "step": 7966 + }, + { + "epoch": 2.512456638284453, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 1.0826, + "step": 7967 + }, + { + "epoch": 2.512771996215705, + "grad_norm": 0.478515625, + "learning_rate": 0.0002, + "loss": 0.8399, + "step": 7968 + }, + { + "epoch": 2.5130873541469567, + "grad_norm": 0.58984375, + "learning_rate": 0.0002, + "loss": 0.8791, + "step": 7969 + }, + { + "epoch": 2.513402712078209, + "grad_norm": 0.52734375, + "learning_rate": 0.0002, + "loss": 0.8662, + "step": 7970 + }, + { + "epoch": 2.513718070009461, + "grad_norm": 0.6640625, + "learning_rate": 0.0002, + "loss": 0.8954, + "step": 7971 + }, + { + "epoch": 2.5140334279407126, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 0.7725, + "step": 7972 + }, + { + "epoch": 2.514348785871965, + "grad_norm": 0.703125, + "learning_rate": 0.0002, + "loss": 0.9678, + "step": 7973 + }, + { + "epoch": 2.5146641438032167, + "grad_norm": 0.7109375, + "learning_rate": 0.0002, + "loss": 1.1902, + "step": 7974 + }, + { + "epoch": 2.5149795017344685, + "grad_norm": 0.765625, + "learning_rate": 0.0002, + "loss": 1.1479, + "step": 7975 + }, + { + "epoch": 2.5152948596657208, + "grad_norm": 0.8046875, + "learning_rate": 0.0002, + "loss": 1.1431, + "step": 7976 + }, + { + "epoch": 2.5156102175969726, + "grad_norm": 0.765625, + "learning_rate": 0.0002, + "loss": 0.982, + "step": 7977 + }, + { + "epoch": 2.5159255755282244, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 1.4204, + "step": 7978 + }, + { + "epoch": 2.5162409334594766, + "grad_norm": 1.078125, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 7979 + }, + { + "epoch": 2.5165562913907285, + "grad_norm": 0.984375, + "learning_rate": 0.0002, + "loss": 1.1948, + "step": 7980 + }, + { + "epoch": 2.5168716493219803, + "grad_norm": 1.734375, + "learning_rate": 0.0002, + "loss": 1.7344, + "step": 7981 + }, + { + "epoch": 2.5171870072532325, + "grad_norm": 1.2890625, + "learning_rate": 0.0002, + "loss": 1.3935, + "step": 7982 + }, + { + "epoch": 2.5175023651844843, + "grad_norm": 1.3046875, + "learning_rate": 0.0002, + "loss": 1.5443, + "step": 7983 + }, + { + "epoch": 2.517817723115736, + "grad_norm": 1.4375, + "learning_rate": 0.0002, + "loss": 1.5351, + "step": 7984 + }, + { + "epoch": 2.5181330810469884, + "grad_norm": 1.6015625, + "learning_rate": 0.0002, + "loss": 1.6094, + "step": 7985 + }, + { + "epoch": 2.5184484389782402, + "grad_norm": 1.46875, + "learning_rate": 0.0002, + "loss": 1.8896, + "step": 7986 + }, + { + "epoch": 2.518763796909492, + "grad_norm": 1.3828125, + "learning_rate": 0.0002, + "loss": 1.7159, + "step": 7987 + }, + { + "epoch": 2.5190791548407443, + "grad_norm": 1.7109375, + "learning_rate": 0.0002, + "loss": 1.7159, + "step": 7988 + }, + { + "epoch": 2.519394512771996, + "grad_norm": 1.21875, + "learning_rate": 0.0002, + "loss": 1.4852, + "step": 7989 + }, + { + "epoch": 2.519709870703248, + "grad_norm": 1.5859375, + "learning_rate": 0.0002, + "loss": 1.7149, + "step": 7990 + }, + { + "epoch": 2.5200252286345, + "grad_norm": 1.1875, + "learning_rate": 0.0002, + "loss": 1.4962, + "step": 7991 + }, + { + "epoch": 2.520340586565752, + "grad_norm": 1.875, + "learning_rate": 0.0002, + "loss": 1.9118, + "step": 7992 + }, + { + "epoch": 2.5206559444970043, + "grad_norm": 0.2294921875, + "learning_rate": 0.0002, + "loss": 1.0005, + "step": 7993 + }, + { + "epoch": 2.520971302428256, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.0166, + "step": 7994 + }, + { + "epoch": 2.521286660359508, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.0363, + "step": 7995 + }, + { + "epoch": 2.52160201829076, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 1.0384, + "step": 7996 + }, + { + "epoch": 2.521917376222012, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 0.982, + "step": 7997 + }, + { + "epoch": 2.522232734153264, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 1.0268, + "step": 7998 + }, + { + "epoch": 2.522548092084516, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 0.8524, + "step": 7999 + }, + { + "epoch": 2.522863450015768, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 0.8924, + "step": 8000 + }, + { + "epoch": 2.52317880794702, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 1.0317, + "step": 8001 + }, + { + "epoch": 2.523494165878272, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.8152, + "step": 8002 + }, + { + "epoch": 2.5238095238095237, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 0.8177, + "step": 8003 + }, + { + "epoch": 2.524124881740776, + "grad_norm": 0.353515625, + "learning_rate": 0.0002, + "loss": 0.9305, + "step": 8004 + }, + { + "epoch": 2.524440239672028, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 0.8104, + "step": 8005 + }, + { + "epoch": 2.5247555976032796, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 0.8905, + "step": 8006 + }, + { + "epoch": 2.525070955534532, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 0.8256, + "step": 8007 + }, + { + "epoch": 2.5253863134657837, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 0.8077, + "step": 8008 + }, + { + "epoch": 2.5257016713970355, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 0.9561, + "step": 8009 + }, + { + "epoch": 2.5260170293282878, + "grad_norm": 0.353515625, + "learning_rate": 0.0002, + "loss": 0.8691, + "step": 8010 + }, + { + "epoch": 2.5263323872595396, + "grad_norm": 0.43359375, + "learning_rate": 0.0002, + "loss": 0.9449, + "step": 8011 + }, + { + "epoch": 2.5266477451907914, + "grad_norm": 0.447265625, + "learning_rate": 0.0002, + "loss": 0.9661, + "step": 8012 + }, + { + "epoch": 2.5269631031220436, + "grad_norm": 0.46875, + "learning_rate": 0.0002, + "loss": 0.9586, + "step": 8013 + }, + { + "epoch": 2.5272784610532955, + "grad_norm": 0.48828125, + "learning_rate": 0.0002, + "loss": 0.8217, + "step": 8014 + }, + { + "epoch": 2.5275938189845473, + "grad_norm": 0.455078125, + "learning_rate": 0.0002, + "loss": 0.8303, + "step": 8015 + }, + { + "epoch": 2.5279091769157995, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 0.9311, + "step": 8016 + }, + { + "epoch": 2.5282245348470513, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 0.9015, + "step": 8017 + }, + { + "epoch": 2.528539892778303, + "grad_norm": 0.44921875, + "learning_rate": 0.0002, + "loss": 0.7227, + "step": 8018 + }, + { + "epoch": 2.5288552507095554, + "grad_norm": 0.625, + "learning_rate": 0.0002, + "loss": 0.925, + "step": 8019 + }, + { + "epoch": 2.5291706086408072, + "grad_norm": 0.58984375, + "learning_rate": 0.0002, + "loss": 0.9827, + "step": 8020 + }, + { + "epoch": 2.529485966572059, + "grad_norm": 0.625, + "learning_rate": 0.0002, + "loss": 0.967, + "step": 8021 + }, + { + "epoch": 2.5298013245033113, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 0.7522, + "step": 8022 + }, + { + "epoch": 2.530116682434563, + "grad_norm": 0.890625, + "learning_rate": 0.0002, + "loss": 1.1143, + "step": 8023 + }, + { + "epoch": 2.530432040365815, + "grad_norm": 0.8046875, + "learning_rate": 0.0002, + "loss": 1.1373, + "step": 8024 + }, + { + "epoch": 2.530747398297067, + "grad_norm": 0.83203125, + "learning_rate": 0.0002, + "loss": 1.0788, + "step": 8025 + }, + { + "epoch": 2.531062756228319, + "grad_norm": 1.015625, + "learning_rate": 0.0002, + "loss": 1.3973, + "step": 8026 + }, + { + "epoch": 2.5313781141595713, + "grad_norm": 0.85546875, + "learning_rate": 0.0002, + "loss": 1.0945, + "step": 8027 + }, + { + "epoch": 2.531693472090823, + "grad_norm": 1.15625, + "learning_rate": 0.0002, + "loss": 1.2907, + "step": 8028 + }, + { + "epoch": 2.5320088300220753, + "grad_norm": 0.921875, + "learning_rate": 0.0002, + "loss": 1.1861, + "step": 8029 + }, + { + "epoch": 2.532324187953327, + "grad_norm": 1.078125, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 8030 + }, + { + "epoch": 2.532639545884579, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 1.2926, + "step": 8031 + }, + { + "epoch": 2.532954903815831, + "grad_norm": 1.953125, + "learning_rate": 0.0002, + "loss": 1.7636, + "step": 8032 + }, + { + "epoch": 2.533270261747083, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.4394, + "step": 8033 + }, + { + "epoch": 2.533585619678335, + "grad_norm": 1.453125, + "learning_rate": 0.0002, + "loss": 1.7609, + "step": 8034 + }, + { + "epoch": 2.533900977609587, + "grad_norm": 1.65625, + "learning_rate": 0.0002, + "loss": 2.3289, + "step": 8035 + }, + { + "epoch": 2.534216335540839, + "grad_norm": 1.28125, + "learning_rate": 0.0002, + "loss": 1.4235, + "step": 8036 + }, + { + "epoch": 2.5345316934720907, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 1.502, + "step": 8037 + }, + { + "epoch": 2.534847051403343, + "grad_norm": 1.0, + "learning_rate": 0.0002, + "loss": 1.5229, + "step": 8038 + }, + { + "epoch": 2.535162409334595, + "grad_norm": 1.84375, + "learning_rate": 0.0002, + "loss": 1.6178, + "step": 8039 + }, + { + "epoch": 2.5354777672658466, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.3505, + "step": 8040 + }, + { + "epoch": 2.535793125197099, + "grad_norm": 1.0859375, + "learning_rate": 0.0002, + "loss": 1.1814, + "step": 8041 + }, + { + "epoch": 2.535793125197099, + "eval_loss": 1.8488632440567017, + "eval_runtime": 156.0812, + "eval_samples_per_second": 6.407, + "eval_steps_per_second": 6.407, + "step": 8041 + }, + { + "epoch": 2.535793125197099, + "mmlu_eval_accuracy": 0.5975481111934129, + "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182, + "mmlu_eval_accuracy_anatomy": 0.6428571428571429, + "mmlu_eval_accuracy_astronomy": 0.625, + "mmlu_eval_accuracy_business_ethics": 0.8181818181818182, + "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241, + "mmlu_eval_accuracy_college_biology": 0.8125, + "mmlu_eval_accuracy_college_chemistry": 0.375, + "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727, + "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453, + "mmlu_eval_accuracy_college_medicine": 0.5454545454545454, + "mmlu_eval_accuracy_college_physics": 0.36363636363636365, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384, + "mmlu_eval_accuracy_econometrics": 0.5833333333333334, + "mmlu_eval_accuracy_electrical_engineering": 0.625, + "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634, + "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, + "mmlu_eval_accuracy_global_facts": 0.3, + "mmlu_eval_accuracy_high_school_biology": 0.5625, + "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.8888888888888888, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209, + "mmlu_eval_accuracy_high_school_mathematics": 0.41379310344827586, + "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384, + "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, + "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667, + "mmlu_eval_accuracy_high_school_statistics": 0.5217391304347826, + "mmlu_eval_accuracy_high_school_us_history": 0.8636363636363636, + "mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693, + "mmlu_eval_accuracy_human_aging": 0.6086956521739131, + "mmlu_eval_accuracy_human_sexuality": 0.5, + "mmlu_eval_accuracy_international_law": 0.8461538461538461, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.72, + "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182, + "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279, + "mmlu_eval_accuracy_moral_disputes": 0.5, + "mmlu_eval_accuracy_moral_scenarios": 0.29, + "mmlu_eval_accuracy_nutrition": 0.7272727272727273, + "mmlu_eval_accuracy_philosophy": 0.6176470588235294, + "mmlu_eval_accuracy_prehistory": 0.6, + "mmlu_eval_accuracy_professional_accounting": 0.5161290322580645, + "mmlu_eval_accuracy_professional_law": 0.43529411764705883, + "mmlu_eval_accuracy_professional_medicine": 0.6774193548387096, + "mmlu_eval_accuracy_professional_psychology": 0.5942028985507246, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7037037037037037, + "mmlu_eval_accuracy_sociology": 0.9545454545454546, + "mmlu_eval_accuracy_us_foreign_policy": 1.0, + "mmlu_eval_accuracy_virology": 0.5555555555555556, + "mmlu_eval_accuracy_world_religions": 0.8421052631578947, + "mmlu_loss": 5.787406361079698, + "step": 8041 + }, + { + "epoch": 2.5361084831283507, + "grad_norm": 1.1328125, + "learning_rate": 0.0002, + "loss": 1.7876, + "step": 8042 + }, + { + "epoch": 2.5364238410596025, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 0.9116, + "step": 8043 + }, + { + "epoch": 2.5367391989908548, + "grad_norm": 0.255859375, + "learning_rate": 0.0002, + "loss": 0.8399, + "step": 8044 + }, + { + "epoch": 2.5370545569221066, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 0.936, + "step": 8045 + }, + { + "epoch": 2.5373699148533584, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 0.8672, + "step": 8046 + }, + { + "epoch": 2.5376852727846106, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 0.8371, + "step": 8047 + }, + { + "epoch": 2.5380006307158625, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 0.9004, + "step": 8048 + }, + { + "epoch": 2.5383159886471143, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 0.9237, + "step": 8049 + }, + { + "epoch": 2.5386313465783665, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.8143, + "step": 8050 + }, + { + "epoch": 2.5389467045096183, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 0.9218, + "step": 8051 + }, + { + "epoch": 2.53926206244087, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 0.9563, + "step": 8052 + }, + { + "epoch": 2.5395774203721224, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 0.8706, + "step": 8053 + }, + { + "epoch": 2.5398927783033742, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 0.9813, + "step": 8054 + }, + { + "epoch": 2.540208136234626, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 0.8784, + "step": 8055 + }, + { + "epoch": 2.5405234941658783, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 0.8739, + "step": 8056 + }, + { + "epoch": 2.54083885209713, + "grad_norm": 0.37109375, + "learning_rate": 0.0002, + "loss": 0.9888, + "step": 8057 + }, + { + "epoch": 2.5411542100283824, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 0.9138, + "step": 8058 + }, + { + "epoch": 2.541469567959634, + "grad_norm": 0.34765625, + "learning_rate": 0.0002, + "loss": 0.8076, + "step": 8059 + }, + { + "epoch": 2.541784925890886, + "grad_norm": 0.3984375, + "learning_rate": 0.0002, + "loss": 0.902, + "step": 8060 + }, + { + "epoch": 2.5421002838221383, + "grad_norm": 0.427734375, + "learning_rate": 0.0002, + "loss": 0.9558, + "step": 8061 + }, + { + "epoch": 2.54241564175339, + "grad_norm": 0.5390625, + "learning_rate": 0.0002, + "loss": 1.0392, + "step": 8062 + }, + { + "epoch": 2.5427309996846423, + "grad_norm": 0.484375, + "learning_rate": 0.0002, + "loss": 1.0403, + "step": 8063 + }, + { + "epoch": 2.543046357615894, + "grad_norm": 0.462890625, + "learning_rate": 0.0002, + "loss": 0.8284, + "step": 8064 + }, + { + "epoch": 2.543361715547146, + "grad_norm": 0.494140625, + "learning_rate": 0.0002, + "loss": 0.9115, + "step": 8065 + }, + { + "epoch": 2.543677073478398, + "grad_norm": 0.48828125, + "learning_rate": 0.0002, + "loss": 0.8954, + "step": 8066 + }, + { + "epoch": 2.54399243140965, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 0.9682, + "step": 8067 + }, + { + "epoch": 2.544307789340902, + "grad_norm": 0.52734375, + "learning_rate": 0.0002, + "loss": 0.8109, + "step": 8068 + }, + { + "epoch": 2.544623147272154, + "grad_norm": 0.62890625, + "learning_rate": 0.0002, + "loss": 0.9309, + "step": 8069 + }, + { + "epoch": 2.544938505203406, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 0.8756, + "step": 8070 + }, + { + "epoch": 2.5452538631346577, + "grad_norm": 0.61328125, + "learning_rate": 0.0002, + "loss": 0.8632, + "step": 8071 + }, + { + "epoch": 2.54556922106591, + "grad_norm": 0.7421875, + "learning_rate": 0.0002, + "loss": 1.1853, + "step": 8072 + }, + { + "epoch": 2.545884578997162, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 1.0085, + "step": 8073 + }, + { + "epoch": 2.5461999369284136, + "grad_norm": 0.87890625, + "learning_rate": 0.0002, + "loss": 1.1302, + "step": 8074 + }, + { + "epoch": 2.546515294859666, + "grad_norm": 0.8984375, + "learning_rate": 0.0002, + "loss": 1.1567, + "step": 8075 + }, + { + "epoch": 2.5468306527909177, + "grad_norm": 0.84765625, + "learning_rate": 0.0002, + "loss": 1.1616, + "step": 8076 + }, + { + "epoch": 2.5471460107221695, + "grad_norm": 0.89453125, + "learning_rate": 0.0002, + "loss": 1.0116, + "step": 8077 + }, + { + "epoch": 2.5474613686534218, + "grad_norm": 1.015625, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 8078 + }, + { + "epoch": 2.5477767265846736, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 8079 + }, + { + "epoch": 2.5480920845159254, + "grad_norm": 1.484375, + "learning_rate": 0.0002, + "loss": 1.5204, + "step": 8080 + }, + { + "epoch": 2.5484074424471777, + "grad_norm": 1.2578125, + "learning_rate": 0.0002, + "loss": 1.3773, + "step": 8081 + }, + { + "epoch": 2.5487228003784295, + "grad_norm": 1.1875, + "learning_rate": 0.0002, + "loss": 1.4833, + "step": 8082 + }, + { + "epoch": 2.5490381583096813, + "grad_norm": 1.5078125, + "learning_rate": 0.0002, + "loss": 1.6617, + "step": 8083 + }, + { + "epoch": 2.5493535162409335, + "grad_norm": 1.015625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 8084 + }, + { + "epoch": 2.5496688741721854, + "grad_norm": 1.0, + "learning_rate": 0.0002, + "loss": 1.492, + "step": 8085 + }, + { + "epoch": 2.549984232103437, + "grad_norm": 1.2578125, + "learning_rate": 0.0002, + "loss": 1.6013, + "step": 8086 + }, + { + "epoch": 2.5502995900346894, + "grad_norm": 2.078125, + "learning_rate": 0.0002, + "loss": 2.1712, + "step": 8087 + }, + { + "epoch": 2.5506149479659412, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 1.5238, + "step": 8088 + }, + { + "epoch": 2.550930305897193, + "grad_norm": 1.109375, + "learning_rate": 0.0002, + "loss": 1.5079, + "step": 8089 + }, + { + "epoch": 2.5512456638284453, + "grad_norm": 1.2109375, + "learning_rate": 0.0002, + "loss": 1.4108, + "step": 8090 + }, + { + "epoch": 2.551561021759697, + "grad_norm": 1.125, + "learning_rate": 0.0002, + "loss": 1.5853, + "step": 8091 + }, + { + "epoch": 2.5518763796909494, + "grad_norm": 2.8125, + "learning_rate": 0.0002, + "loss": 2.0119, + "step": 8092 + }, + { + "epoch": 2.552191737622201, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 0.9735, + "step": 8093 + }, + { + "epoch": 2.552507095553453, + "grad_norm": 0.2373046875, + "learning_rate": 0.0002, + "loss": 0.8634, + "step": 8094 + }, + { + "epoch": 2.5528224534847053, + "grad_norm": 0.2490234375, + "learning_rate": 0.0002, + "loss": 0.9954, + "step": 8095 + }, + { + "epoch": 2.553137811415957, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 1.0141, + "step": 8096 + }, + { + "epoch": 2.5534531693472093, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 0.9512, + "step": 8097 + }, + { + "epoch": 2.553768527278461, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 0.9317, + "step": 8098 + }, + { + "epoch": 2.554083885209713, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 0.933, + "step": 8099 + }, + { + "epoch": 2.5543992431409652, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 0.9745, + "step": 8100 + }, + { + "epoch": 2.554714601072217, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 0.9837, + "step": 8101 + }, + { + "epoch": 2.555029959003469, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.8884, + "step": 8102 + }, + { + "epoch": 2.555345316934721, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 0.9765, + "step": 8103 + }, + { + "epoch": 2.555660674865973, + "grad_norm": 0.3515625, + "learning_rate": 0.0002, + "loss": 0.9071, + "step": 8104 + }, + { + "epoch": 2.5559760327972247, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 0.8475, + "step": 8105 + }, + { + "epoch": 2.556291390728477, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 0.8244, + "step": 8106 + }, + { + "epoch": 2.556606748659729, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 0.9562, + "step": 8107 + }, + { + "epoch": 2.5569221065909806, + "grad_norm": 0.349609375, + "learning_rate": 0.0002, + "loss": 0.8613, + "step": 8108 + }, + { + "epoch": 2.557237464522233, + "grad_norm": 0.37109375, + "learning_rate": 0.0002, + "loss": 0.9081, + "step": 8109 + }, + { + "epoch": 2.5575528224534847, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 0.8743, + "step": 8110 + }, + { + "epoch": 2.5578681803847365, + "grad_norm": 0.3984375, + "learning_rate": 0.0002, + "loss": 0.8619, + "step": 8111 + }, + { + "epoch": 2.5581835383159888, + "grad_norm": 0.435546875, + "learning_rate": 0.0002, + "loss": 0.9258, + "step": 8112 + }, + { + "epoch": 2.5584988962472406, + "grad_norm": 0.486328125, + "learning_rate": 0.0002, + "loss": 0.9551, + "step": 8113 + }, + { + "epoch": 2.5588142541784924, + "grad_norm": 0.486328125, + "learning_rate": 0.0002, + "loss": 0.958, + "step": 8114 + }, + { + "epoch": 2.5591296121097447, + "grad_norm": 0.45703125, + "learning_rate": 0.0002, + "loss": 0.7597, + "step": 8115 + }, + { + "epoch": 2.5594449700409965, + "grad_norm": 0.46484375, + "learning_rate": 0.0002, + "loss": 0.8418, + "step": 8116 + }, + { + "epoch": 2.5597603279722483, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 0.8842, + "step": 8117 + }, + { + "epoch": 2.5600756859035005, + "grad_norm": 0.515625, + "learning_rate": 0.0002, + "loss": 0.895, + "step": 8118 + }, + { + "epoch": 2.5603910438347524, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 0.9372, + "step": 8119 + }, + { + "epoch": 2.560706401766004, + "grad_norm": 0.578125, + "learning_rate": 0.0002, + "loss": 0.9709, + "step": 8120 + }, + { + "epoch": 2.5610217596972564, + "grad_norm": 0.60546875, + "learning_rate": 0.0002, + "loss": 0.9647, + "step": 8121 + }, + { + "epoch": 2.5613371176285082, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 0.9067, + "step": 8122 + }, + { + "epoch": 2.56165247555976, + "grad_norm": 0.609375, + "learning_rate": 0.0002, + "loss": 0.8903, + "step": 8123 + }, + { + "epoch": 2.5619678334910123, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 1.0097, + "step": 8124 + }, + { + "epoch": 2.562283191422264, + "grad_norm": 0.71875, + "learning_rate": 0.0002, + "loss": 0.9252, + "step": 8125 + }, + { + "epoch": 2.5625985493535164, + "grad_norm": 0.7890625, + "learning_rate": 0.0002, + "loss": 0.9825, + "step": 8126 + }, + { + "epoch": 2.562913907284768, + "grad_norm": 0.94921875, + "learning_rate": 0.0002, + "loss": 1.089, + "step": 8127 + }, + { + "epoch": 2.5632292652160205, + "grad_norm": 0.87890625, + "learning_rate": 0.0002, + "loss": 1.2327, + "step": 8128 + }, + { + "epoch": 2.5635446231472723, + "grad_norm": 0.83984375, + "learning_rate": 0.0002, + "loss": 1.2508, + "step": 8129 + }, + { + "epoch": 2.563859981078524, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 1.4803, + "step": 8130 + }, + { + "epoch": 2.5641753390097763, + "grad_norm": 1.109375, + "learning_rate": 0.0002, + "loss": 1.5927, + "step": 8131 + }, + { + "epoch": 2.564490696941028, + "grad_norm": 0.90234375, + "learning_rate": 0.0002, + "loss": 1.2076, + "step": 8132 + }, + { + "epoch": 2.56480605487228, + "grad_norm": 1.171875, + "learning_rate": 0.0002, + "loss": 1.1966, + "step": 8133 + }, + { + "epoch": 2.5651214128035322, + "grad_norm": 1.1796875, + "learning_rate": 0.0002, + "loss": 1.5279, + "step": 8134 + }, + { + "epoch": 2.565436770734784, + "grad_norm": 1.3671875, + "learning_rate": 0.0002, + "loss": 1.5971, + "step": 8135 + }, + { + "epoch": 2.565752128666036, + "grad_norm": 1.421875, + "learning_rate": 0.0002, + "loss": 1.7763, + "step": 8136 + }, + { + "epoch": 2.566067486597288, + "grad_norm": 1.265625, + "learning_rate": 0.0002, + "loss": 1.4081, + "step": 8137 + }, + { + "epoch": 2.56638284452854, + "grad_norm": 1.34375, + "learning_rate": 0.0002, + "loss": 1.6554, + "step": 8138 + }, + { + "epoch": 2.5666982024597917, + "grad_norm": 1.1875, + "learning_rate": 0.0002, + "loss": 1.3949, + "step": 8139 + }, + { + "epoch": 2.567013560391044, + "grad_norm": 1.09375, + "learning_rate": 0.0002, + "loss": 1.402, + "step": 8140 + }, + { + "epoch": 2.567328918322296, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 1.1443, + "step": 8141 + }, + { + "epoch": 2.5676442762535476, + "grad_norm": 1.328125, + "learning_rate": 0.0002, + "loss": 1.8991, + "step": 8142 + }, + { + "epoch": 2.5679596341848, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.0126, + "step": 8143 + }, + { + "epoch": 2.5682749921160517, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 0.9594, + "step": 8144 + }, + { + "epoch": 2.5685903500473035, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 1.0717, + "step": 8145 + }, + { + "epoch": 2.5689057079785558, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.982, + "step": 8146 + }, + { + "epoch": 2.5692210659098076, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 0.9799, + "step": 8147 + }, + { + "epoch": 2.5695364238410594, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 0.9111, + "step": 8148 + }, + { + "epoch": 2.5698517817723117, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 0.9385, + "step": 8149 + }, + { + "epoch": 2.5701671397035635, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 0.9946, + "step": 8150 + }, + { + "epoch": 2.5704824976348153, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 0.8996, + "step": 8151 + }, + { + "epoch": 2.5707978555660675, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 0.9992, + "step": 8152 + }, + { + "epoch": 2.5711132134973194, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.9574, + "step": 8153 + }, + { + "epoch": 2.571428571428571, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 0.9977, + "step": 8154 + }, + { + "epoch": 2.5717439293598234, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 0.9474, + "step": 8155 + }, + { + "epoch": 2.5720592872910752, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 0.9477, + "step": 8156 + }, + { + "epoch": 2.5723746452223275, + "grad_norm": 0.392578125, + "learning_rate": 0.0002, + "loss": 0.9209, + "step": 8157 + }, + { + "epoch": 2.5726900031535793, + "grad_norm": 0.384765625, + "learning_rate": 0.0002, + "loss": 0.8344, + "step": 8158 + }, + { + "epoch": 2.573005361084831, + "grad_norm": 0.423828125, + "learning_rate": 0.0002, + "loss": 0.9805, + "step": 8159 + }, + { + "epoch": 2.5733207190160834, + "grad_norm": 0.3984375, + "learning_rate": 0.0002, + "loss": 0.8953, + "step": 8160 + }, + { + "epoch": 2.573636076947335, + "grad_norm": 0.384765625, + "learning_rate": 0.0002, + "loss": 0.898, + "step": 8161 + }, + { + "epoch": 2.5739514348785875, + "grad_norm": 0.451171875, + "learning_rate": 0.0002, + "loss": 0.8101, + "step": 8162 + }, + { + "epoch": 2.5742667928098393, + "grad_norm": 0.48046875, + "learning_rate": 0.0002, + "loss": 0.879, + "step": 8163 + }, + { + "epoch": 2.574582150741091, + "grad_norm": 0.47265625, + "learning_rate": 0.0002, + "loss": 0.9872, + "step": 8164 + }, + { + "epoch": 2.5748975086723433, + "grad_norm": 0.447265625, + "learning_rate": 0.0002, + "loss": 0.9662, + "step": 8165 + }, + { + "epoch": 2.575212866603595, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 0.9084, + "step": 8166 + }, + { + "epoch": 2.575528224534847, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 1.0015, + "step": 8167 + }, + { + "epoch": 2.5758435824660992, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 1.0664, + "step": 8168 + }, + { + "epoch": 2.576158940397351, + "grad_norm": 0.5703125, + "learning_rate": 0.0002, + "loss": 0.9359, + "step": 8169 + }, + { + "epoch": 2.576474298328603, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 0.8543, + "step": 8170 + }, + { + "epoch": 2.576789656259855, + "grad_norm": 0.625, + "learning_rate": 0.0002, + "loss": 0.8978, + "step": 8171 + }, + { + "epoch": 2.577105014191107, + "grad_norm": 0.65234375, + "learning_rate": 0.0002, + "loss": 0.8484, + "step": 8172 + }, + { + "epoch": 2.5774203721223587, + "grad_norm": 0.859375, + "learning_rate": 0.0002, + "loss": 1.1102, + "step": 8173 + }, + { + "epoch": 2.577735730053611, + "grad_norm": 0.87109375, + "learning_rate": 0.0002, + "loss": 1.2877, + "step": 8174 + }, + { + "epoch": 2.578051087984863, + "grad_norm": 0.859375, + "learning_rate": 0.0002, + "loss": 1.002, + "step": 8175 + }, + { + "epoch": 2.5783664459161146, + "grad_norm": 1.125, + "learning_rate": 0.0002, + "loss": 1.1986, + "step": 8176 + }, + { + "epoch": 2.578681803847367, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 8177 + }, + { + "epoch": 2.5789971617786187, + "grad_norm": 1.2421875, + "learning_rate": 0.0002, + "loss": 1.2304, + "step": 8178 + }, + { + "epoch": 2.5793125197098705, + "grad_norm": 1.546875, + "learning_rate": 0.0002, + "loss": 1.4181, + "step": 8179 + }, + { + "epoch": 2.579627877641123, + "grad_norm": 1.265625, + "learning_rate": 0.0002, + "loss": 1.5448, + "step": 8180 + }, + { + "epoch": 2.5799432355723746, + "grad_norm": 1.2109375, + "learning_rate": 0.0002, + "loss": 1.577, + "step": 8181 + }, + { + "epoch": 2.5802585935036264, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 1.6145, + "step": 8182 + }, + { + "epoch": 2.5805739514348787, + "grad_norm": 1.125, + "learning_rate": 0.0002, + "loss": 1.6414, + "step": 8183 + }, + { + "epoch": 2.5808893093661305, + "grad_norm": 1.125, + "learning_rate": 0.0002, + "loss": 1.4255, + "step": 8184 + }, + { + "epoch": 2.5812046672973823, + "grad_norm": 0.94140625, + "learning_rate": 0.0002, + "loss": 1.4128, + "step": 8185 + }, + { + "epoch": 2.5815200252286346, + "grad_norm": 1.1328125, + "learning_rate": 0.0002, + "loss": 1.4937, + "step": 8186 + }, + { + "epoch": 2.5818353831598864, + "grad_norm": 0.91015625, + "learning_rate": 0.0002, + "loss": 1.4695, + "step": 8187 + }, + { + "epoch": 2.582150741091138, + "grad_norm": 1.28125, + "learning_rate": 0.0002, + "loss": 1.7156, + "step": 8188 + }, + { + "epoch": 2.5824660990223904, + "grad_norm": 1.625, + "learning_rate": 0.0002, + "loss": 1.6202, + "step": 8189 + }, + { + "epoch": 2.5827814569536423, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.251, + "step": 8190 + }, + { + "epoch": 2.5830968148848945, + "grad_norm": 1.390625, + "learning_rate": 0.0002, + "loss": 1.55, + "step": 8191 + }, + { + "epoch": 2.5834121728161463, + "grad_norm": 1.0390625, + "learning_rate": 0.0002, + "loss": 1.7944, + "step": 8192 + }, + { + "epoch": 2.583727530747398, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 0.9796, + "step": 8193 + }, + { + "epoch": 2.5840428886786504, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.9593, + "step": 8194 + }, + { + "epoch": 2.584358246609902, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 1.0229, + "step": 8195 + }, + { + "epoch": 2.5846736045411545, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 0.8716, + "step": 8196 + }, + { + "epoch": 2.5849889624724063, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 0.8874, + "step": 8197 + }, + { + "epoch": 2.585304320403658, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 0.9051, + "step": 8198 + }, + { + "epoch": 2.5856196783349104, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 1.0108, + "step": 8199 + }, + { + "epoch": 2.585935036266162, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 0.8774, + "step": 8200 + }, + { + "epoch": 2.586250394197414, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 1.0042, + "step": 8201 + }, + { + "epoch": 2.5865657521286662, + "grad_norm": 0.349609375, + "learning_rate": 0.0002, + "loss": 0.9885, + "step": 8202 + }, + { + "epoch": 2.586881110059918, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 0.9125, + "step": 8203 + }, + { + "epoch": 2.58719646799117, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 1.0027, + "step": 8204 + }, + { + "epoch": 2.587511825922422, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 0.945, + "step": 8205 + }, + { + "epoch": 2.587827183853674, + "grad_norm": 0.3671875, + "learning_rate": 0.0002, + "loss": 0.8121, + "step": 8206 + }, + { + "epoch": 2.5881425417849258, + "grad_norm": 0.361328125, + "learning_rate": 0.0002, + "loss": 0.9001, + "step": 8207 + }, + { + "epoch": 2.588457899716178, + "grad_norm": 0.369140625, + "learning_rate": 0.0002, + "loss": 0.9146, + "step": 8208 + }, + { + "epoch": 2.58877325764743, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 0.82, + "step": 8209 + }, + { + "epoch": 2.5890886155786816, + "grad_norm": 0.390625, + "learning_rate": 0.0002, + "loss": 0.8192, + "step": 8210 + }, + { + "epoch": 2.589403973509934, + "grad_norm": 0.361328125, + "learning_rate": 0.0002, + "loss": 0.8077, + "step": 8211 + }, + { + "epoch": 2.5897193314411857, + "grad_norm": 0.515625, + "learning_rate": 0.0002, + "loss": 0.8385, + "step": 8212 + }, + { + "epoch": 2.5900346893724375, + "grad_norm": 0.41015625, + "learning_rate": 0.0002, + "loss": 0.8846, + "step": 8213 + }, + { + "epoch": 2.59035004730369, + "grad_norm": 0.44921875, + "learning_rate": 0.0002, + "loss": 0.98, + "step": 8214 + }, + { + "epoch": 2.5906654052349416, + "grad_norm": 0.49609375, + "learning_rate": 0.0002, + "loss": 0.9062, + "step": 8215 + }, + { + "epoch": 2.5909807631661934, + "grad_norm": 0.640625, + "learning_rate": 0.0002, + "loss": 0.8951, + "step": 8216 + }, + { + "epoch": 2.5912961210974457, + "grad_norm": 0.56640625, + "learning_rate": 0.0002, + "loss": 0.9997, + "step": 8217 + }, + { + "epoch": 2.5916114790286975, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 0.875, + "step": 8218 + }, + { + "epoch": 2.5919268369599493, + "grad_norm": 0.578125, + "learning_rate": 0.0002, + "loss": 1.0167, + "step": 8219 + }, + { + "epoch": 2.5922421948912016, + "grad_norm": 0.6015625, + "learning_rate": 0.0002, + "loss": 0.9366, + "step": 8220 + }, + { + "epoch": 2.5925575528224534, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 0.922, + "step": 8221 + }, + { + "epoch": 2.592872910753705, + "grad_norm": 0.6484375, + "learning_rate": 0.0002, + "loss": 0.8256, + "step": 8222 + }, + { + "epoch": 2.5931882686849574, + "grad_norm": 0.79296875, + "learning_rate": 0.0002, + "loss": 0.9383, + "step": 8223 + }, + { + "epoch": 2.5935036266162093, + "grad_norm": 0.84765625, + "learning_rate": 0.0002, + "loss": 0.9821, + "step": 8224 + }, + { + "epoch": 2.5938189845474615, + "grad_norm": 1.3671875, + "learning_rate": 0.0002, + "loss": 1.2563, + "step": 8225 + }, + { + "epoch": 2.5941343424787133, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 1.0725, + "step": 8226 + }, + { + "epoch": 2.5944497004099656, + "grad_norm": 0.828125, + "learning_rate": 0.0002, + "loss": 1.1832, + "step": 8227 + }, + { + "epoch": 2.5947650583412174, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 1.3869, + "step": 8228 + }, + { + "epoch": 2.5947650583412174, + "eval_loss": 1.8022663593292236, + "eval_runtime": 148.8518, + "eval_samples_per_second": 6.718, + "eval_steps_per_second": 6.718, + "step": 8228 + }, + { + "epoch": 2.5947650583412174, + "mmlu_eval_accuracy": 0.5931684002097192, + "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, + "mmlu_eval_accuracy_anatomy": 0.6428571428571429, + "mmlu_eval_accuracy_astronomy": 0.6875, + "mmlu_eval_accuracy_business_ethics": 0.8181818181818182, + "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138, + "mmlu_eval_accuracy_college_biology": 0.8125, + "mmlu_eval_accuracy_college_chemistry": 0.375, + "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727, + "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453, + "mmlu_eval_accuracy_college_medicine": 0.5454545454545454, + "mmlu_eval_accuracy_college_physics": 0.45454545454545453, + "mmlu_eval_accuracy_computer_security": 0.36363636363636365, + "mmlu_eval_accuracy_conceptual_physics": 0.5, + "mmlu_eval_accuracy_econometrics": 0.5, + "mmlu_eval_accuracy_electrical_engineering": 0.625, + "mmlu_eval_accuracy_elementary_mathematics": 0.2926829268292683, + "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, + "mmlu_eval_accuracy_global_facts": 0.2, + "mmlu_eval_accuracy_high_school_biology": 0.65625, + "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453, + "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556, + "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325, + "mmlu_eval_accuracy_high_school_mathematics": 0.3793103448275862, + "mmlu_eval_accuracy_high_school_microeconomics": 0.5, + "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354, + "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667, + "mmlu_eval_accuracy_high_school_statistics": 0.5217391304347826, + "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, + "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, + "mmlu_eval_accuracy_human_aging": 0.6086956521739131, + "mmlu_eval_accuracy_human_sexuality": 0.5, + "mmlu_eval_accuracy_international_law": 0.9230769230769231, + "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.68, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.6976744186046512, + "mmlu_eval_accuracy_moral_disputes": 0.631578947368421, + "mmlu_eval_accuracy_moral_scenarios": 0.28, + "mmlu_eval_accuracy_nutrition": 0.7575757575757576, + "mmlu_eval_accuracy_philosophy": 0.6470588235294118, + "mmlu_eval_accuracy_prehistory": 0.6857142857142857, + "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644, + "mmlu_eval_accuracy_professional_law": 0.4117647058823529, + "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806, + "mmlu_eval_accuracy_professional_psychology": 0.6086956521739131, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7037037037037037, + "mmlu_eval_accuracy_sociology": 1.0, + "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, + "mmlu_eval_accuracy_virology": 0.5, + "mmlu_eval_accuracy_world_religions": 0.7894736842105263, + "mmlu_loss": 5.807161321833116, + "step": 8228 + }, + { + "epoch": 2.595080416272469, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 1.1998, + "step": 8229 + }, + { + "epoch": 2.5953957742037215, + "grad_norm": 1.1328125, + "learning_rate": 0.0002, + "loss": 1.4068, + "step": 8230 + }, + { + "epoch": 2.5957111321349733, + "grad_norm": 1.0546875, + "learning_rate": 0.0002, + "loss": 1.6512, + "step": 8231 + }, + { + "epoch": 2.596026490066225, + "grad_norm": 0.99609375, + "learning_rate": 0.0002, + "loss": 1.4004, + "step": 8232 + }, + { + "epoch": 2.5963418479974774, + "grad_norm": 1.015625, + "learning_rate": 0.0002, + "loss": 1.271, + "step": 8233 + }, + { + "epoch": 2.596657205928729, + "grad_norm": 1.078125, + "learning_rate": 0.0002, + "loss": 1.3875, + "step": 8234 + }, + { + "epoch": 2.596972563859981, + "grad_norm": 1.5859375, + "learning_rate": 0.0002, + "loss": 1.6831, + "step": 8235 + }, + { + "epoch": 2.5972879217912332, + "grad_norm": 1.234375, + "learning_rate": 0.0002, + "loss": 1.4596, + "step": 8236 + }, + { + "epoch": 2.597603279722485, + "grad_norm": 1.4921875, + "learning_rate": 0.0002, + "loss": 1.6264, + "step": 8237 + }, + { + "epoch": 2.597918637653737, + "grad_norm": 1.5703125, + "learning_rate": 0.0002, + "loss": 1.6782, + "step": 8238 + }, + { + "epoch": 2.598233995584989, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 1.4148, + "step": 8239 + }, + { + "epoch": 2.598549353516241, + "grad_norm": 1.4375, + "learning_rate": 0.0002, + "loss": 1.7305, + "step": 8240 + }, + { + "epoch": 2.5988647114474928, + "grad_norm": 1.21875, + "learning_rate": 0.0002, + "loss": 1.7513, + "step": 8241 + }, + { + "epoch": 2.599180069378745, + "grad_norm": 1.109375, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 8242 + }, + { + "epoch": 2.599495427309997, + "grad_norm": 0.23828125, + "learning_rate": 0.0002, + "loss": 1.1504, + "step": 8243 + }, + { + "epoch": 2.5998107852412486, + "grad_norm": 0.2490234375, + "learning_rate": 0.0002, + "loss": 0.959, + "step": 8244 + }, + { + "epoch": 2.600126143172501, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 0.98, + "step": 8245 + }, + { + "epoch": 2.6004415011037527, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 0.9303, + "step": 8246 + }, + { + "epoch": 2.6007568590350045, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 1.021, + "step": 8247 + }, + { + "epoch": 2.601072216966257, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.8661, + "step": 8248 + }, + { + "epoch": 2.6013875748975086, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 0.9139, + "step": 8249 + }, + { + "epoch": 2.6017029328287604, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 0.9458, + "step": 8250 + }, + { + "epoch": 2.6020182907600127, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 0.925, + "step": 8251 + }, + { + "epoch": 2.6023336486912645, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.7844, + "step": 8252 + }, + { + "epoch": 2.6026490066225163, + "grad_norm": 0.369140625, + "learning_rate": 0.0002, + "loss": 1.0269, + "step": 8253 + }, + { + "epoch": 2.6029643645537686, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 0.8283, + "step": 8254 + }, + { + "epoch": 2.6032797224850204, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 1.0173, + "step": 8255 + }, + { + "epoch": 2.6035950804162726, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 0.9272, + "step": 8256 + }, + { + "epoch": 2.6039104383475244, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 0.9207, + "step": 8257 + }, + { + "epoch": 2.6042257962787763, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 0.8708, + "step": 8258 + }, + { + "epoch": 2.6045411542100285, + "grad_norm": 0.421875, + "learning_rate": 0.0002, + "loss": 1.0663, + "step": 8259 + }, + { + "epoch": 2.6048565121412803, + "grad_norm": 0.392578125, + "learning_rate": 0.0002, + "loss": 0.8391, + "step": 8260 + }, + { + "epoch": 2.6051718700725326, + "grad_norm": 0.427734375, + "learning_rate": 0.0002, + "loss": 1.0136, + "step": 8261 + }, + { + "epoch": 2.6054872280037844, + "grad_norm": 0.392578125, + "learning_rate": 0.0002, + "loss": 1.0452, + "step": 8262 + }, + { + "epoch": 2.605802585935036, + "grad_norm": 0.421875, + "learning_rate": 0.0002, + "loss": 0.867, + "step": 8263 + }, + { + "epoch": 2.6061179438662885, + "grad_norm": 0.486328125, + "learning_rate": 0.0002, + "loss": 0.9368, + "step": 8264 + }, + { + "epoch": 2.6064333017975403, + "grad_norm": 0.4921875, + "learning_rate": 0.0002, + "loss": 1.0356, + "step": 8265 + }, + { + "epoch": 2.606748659728792, + "grad_norm": 0.4609375, + "learning_rate": 0.0002, + "loss": 0.9427, + "step": 8266 + }, + { + "epoch": 2.6070640176600444, + "grad_norm": 0.609375, + "learning_rate": 0.0002, + "loss": 0.9194, + "step": 8267 + }, + { + "epoch": 2.607379375591296, + "grad_norm": 0.63671875, + "learning_rate": 0.0002, + "loss": 1.1008, + "step": 8268 + }, + { + "epoch": 2.607694733522548, + "grad_norm": 0.578125, + "learning_rate": 0.0002, + "loss": 0.7681, + "step": 8269 + }, + { + "epoch": 2.6080100914538002, + "grad_norm": 0.64453125, + "learning_rate": 0.0002, + "loss": 1.0306, + "step": 8270 + }, + { + "epoch": 2.608325449385052, + "grad_norm": 0.6875, + "learning_rate": 0.0002, + "loss": 0.9502, + "step": 8271 + }, + { + "epoch": 2.608640807316304, + "grad_norm": 0.72265625, + "learning_rate": 0.0002, + "loss": 0.8826, + "step": 8272 + }, + { + "epoch": 2.608956165247556, + "grad_norm": 0.98828125, + "learning_rate": 0.0002, + "loss": 1.2367, + "step": 8273 + }, + { + "epoch": 2.609271523178808, + "grad_norm": 0.81640625, + "learning_rate": 0.0002, + "loss": 1.1729, + "step": 8274 + }, + { + "epoch": 2.6095868811100598, + "grad_norm": 0.88671875, + "learning_rate": 0.0002, + "loss": 0.9989, + "step": 8275 + }, + { + "epoch": 2.609902239041312, + "grad_norm": 0.8515625, + "learning_rate": 0.0002, + "loss": 1.2143, + "step": 8276 + }, + { + "epoch": 2.610217596972564, + "grad_norm": 1.0546875, + "learning_rate": 0.0002, + "loss": 1.1996, + "step": 8277 + }, + { + "epoch": 2.6105329549038157, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.3961, + "step": 8278 + }, + { + "epoch": 2.610848312835068, + "grad_norm": 1.484375, + "learning_rate": 0.0002, + "loss": 1.544, + "step": 8279 + }, + { + "epoch": 2.6111636707663197, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 1.5317, + "step": 8280 + }, + { + "epoch": 2.6114790286975715, + "grad_norm": 0.96875, + "learning_rate": 0.0002, + "loss": 1.4831, + "step": 8281 + }, + { + "epoch": 2.611794386628824, + "grad_norm": 1.1953125, + "learning_rate": 0.0002, + "loss": 1.6597, + "step": 8282 + }, + { + "epoch": 2.6121097445600756, + "grad_norm": 0.9609375, + "learning_rate": 0.0002, + "loss": 1.4323, + "step": 8283 + }, + { + "epoch": 2.6124251024913274, + "grad_norm": 1.3671875, + "learning_rate": 0.0002, + "loss": 1.7265, + "step": 8284 + }, + { + "epoch": 2.6127404604225797, + "grad_norm": 1.71875, + "learning_rate": 0.0002, + "loss": 1.5453, + "step": 8285 + }, + { + "epoch": 2.6130558183538315, + "grad_norm": 1.4609375, + "learning_rate": 0.0002, + "loss": 1.6951, + "step": 8286 + }, + { + "epoch": 2.6133711762850833, + "grad_norm": 1.4765625, + "learning_rate": 0.0002, + "loss": 1.8578, + "step": 8287 + }, + { + "epoch": 2.6136865342163356, + "grad_norm": 1.34375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 8288 + }, + { + "epoch": 2.6140018921475874, + "grad_norm": 1.3203125, + "learning_rate": 0.0002, + "loss": 1.5772, + "step": 8289 + }, + { + "epoch": 2.6143172500788396, + "grad_norm": 1.21875, + "learning_rate": 0.0002, + "loss": 1.5467, + "step": 8290 + }, + { + "epoch": 2.6146326080100915, + "grad_norm": 1.2578125, + "learning_rate": 0.0002, + "loss": 1.53, + "step": 8291 + }, + { + "epoch": 2.6149479659413433, + "grad_norm": 1.6015625, + "learning_rate": 0.0002, + "loss": 1.7909, + "step": 8292 + }, + { + "epoch": 2.6152633238725955, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 1.0269, + "step": 8293 + }, + { + "epoch": 2.6155786818038473, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 0.9317, + "step": 8294 + }, + { + "epoch": 2.6158940397350996, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 0.9258, + "step": 8295 + }, + { + "epoch": 2.6162093976663514, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.9467, + "step": 8296 + }, + { + "epoch": 2.6165247555976032, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 1.0452, + "step": 8297 + }, + { + "epoch": 2.6168401135288555, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 0.9171, + "step": 8298 + }, + { + "epoch": 2.6171554714601073, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 0.9308, + "step": 8299 + }, + { + "epoch": 2.617470829391359, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 0.842, + "step": 8300 + }, + { + "epoch": 2.6177861873226114, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 0.9746, + "step": 8301 + }, + { + "epoch": 2.618101545253863, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.0559, + "step": 8302 + }, + { + "epoch": 2.618416903185115, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 0.9413, + "step": 8303 + }, + { + "epoch": 2.6187322611163673, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 0.863, + "step": 8304 + }, + { + "epoch": 2.619047619047619, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.8973, + "step": 8305 + }, + { + "epoch": 2.619362976978871, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 0.9183, + "step": 8306 + }, + { + "epoch": 2.619678334910123, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 1.1109, + "step": 8307 + }, + { + "epoch": 2.619993692841375, + "grad_norm": 0.361328125, + "learning_rate": 0.0002, + "loss": 0.8611, + "step": 8308 + }, + { + "epoch": 2.6203090507726268, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 0.9024, + "step": 8309 + }, + { + "epoch": 2.620624408703879, + "grad_norm": 0.37109375, + "learning_rate": 0.0002, + "loss": 0.9144, + "step": 8310 + }, + { + "epoch": 2.620939766635131, + "grad_norm": 0.400390625, + "learning_rate": 0.0002, + "loss": 0.9404, + "step": 8311 + }, + { + "epoch": 2.6212551245663827, + "grad_norm": 0.404296875, + "learning_rate": 0.0002, + "loss": 0.9479, + "step": 8312 + }, + { + "epoch": 2.621570482497635, + "grad_norm": 0.453125, + "learning_rate": 0.0002, + "loss": 1.0294, + "step": 8313 + }, + { + "epoch": 2.6218858404288867, + "grad_norm": 0.462890625, + "learning_rate": 0.0002, + "loss": 0.9669, + "step": 8314 + }, + { + "epoch": 2.6222011983601385, + "grad_norm": 0.462890625, + "learning_rate": 0.0002, + "loss": 0.899, + "step": 8315 + }, + { + "epoch": 2.622516556291391, + "grad_norm": 0.44140625, + "learning_rate": 0.0002, + "loss": 0.781, + "step": 8316 + }, + { + "epoch": 2.6228319142226426, + "grad_norm": 0.48046875, + "learning_rate": 0.0002, + "loss": 0.8528, + "step": 8317 + }, + { + "epoch": 2.6231472721538944, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 0.8636, + "step": 8318 + }, + { + "epoch": 2.6234626300851467, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 0.854, + "step": 8319 + }, + { + "epoch": 2.6237779880163985, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 0.9167, + "step": 8320 + }, + { + "epoch": 2.6240933459476503, + "grad_norm": 0.65625, + "learning_rate": 0.0002, + "loss": 0.9257, + "step": 8321 + }, + { + "epoch": 2.6244087038789026, + "grad_norm": 0.61328125, + "learning_rate": 0.0002, + "loss": 0.8945, + "step": 8322 + }, + { + "epoch": 2.6247240618101544, + "grad_norm": 0.6484375, + "learning_rate": 0.0002, + "loss": 0.9957, + "step": 8323 + }, + { + "epoch": 2.6250394197414066, + "grad_norm": 0.71484375, + "learning_rate": 0.0002, + "loss": 0.9734, + "step": 8324 + }, + { + "epoch": 2.6253547776726585, + "grad_norm": 0.97265625, + "learning_rate": 0.0002, + "loss": 1.2059, + "step": 8325 + }, + { + "epoch": 2.6256701356039107, + "grad_norm": 0.921875, + "learning_rate": 0.0002, + "loss": 1.1858, + "step": 8326 + }, + { + "epoch": 2.6259854935351625, + "grad_norm": 0.90625, + "learning_rate": 0.0002, + "loss": 1.0977, + "step": 8327 + }, + { + "epoch": 2.6263008514664143, + "grad_norm": 1.171875, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 8328 + }, + { + "epoch": 2.6266162093976666, + "grad_norm": 1.2265625, + "learning_rate": 0.0002, + "loss": 1.2199, + "step": 8329 + }, + { + "epoch": 2.6269315673289184, + "grad_norm": 1.03125, + "learning_rate": 0.0002, + "loss": 1.4074, + "step": 8330 + }, + { + "epoch": 2.6272469252601702, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 1.3872, + "step": 8331 + }, + { + "epoch": 2.6275622831914225, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 1.5265, + "step": 8332 + }, + { + "epoch": 2.6278776411226743, + "grad_norm": 1.328125, + "learning_rate": 0.0002, + "loss": 1.7081, + "step": 8333 + }, + { + "epoch": 2.628192999053926, + "grad_norm": 1.2578125, + "learning_rate": 0.0002, + "loss": 1.5806, + "step": 8334 + }, + { + "epoch": 2.6285083569851784, + "grad_norm": 1.5078125, + "learning_rate": 0.0002, + "loss": 1.762, + "step": 8335 + }, + { + "epoch": 2.62882371491643, + "grad_norm": 1.3671875, + "learning_rate": 0.0002, + "loss": 1.8312, + "step": 8336 + }, + { + "epoch": 2.629139072847682, + "grad_norm": 1.0546875, + "learning_rate": 0.0002, + "loss": 1.6145, + "step": 8337 + }, + { + "epoch": 2.6294544307789343, + "grad_norm": 1.3984375, + "learning_rate": 0.0002, + "loss": 2.0675, + "step": 8338 + }, + { + "epoch": 2.629769788710186, + "grad_norm": 1.4609375, + "learning_rate": 0.0002, + "loss": 1.6903, + "step": 8339 + }, + { + "epoch": 2.630085146641438, + "grad_norm": 1.765625, + "learning_rate": 0.0002, + "loss": 1.698, + "step": 8340 + }, + { + "epoch": 2.63040050457269, + "grad_norm": 0.94921875, + "learning_rate": 0.0002, + "loss": 1.2185, + "step": 8341 + }, + { + "epoch": 2.630715862503942, + "grad_norm": 1.734375, + "learning_rate": 0.0002, + "loss": 1.8262, + "step": 8342 + }, + { + "epoch": 2.6310312204351938, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 0.9256, + "step": 8343 + }, + { + "epoch": 2.631346578366446, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 0.8349, + "step": 8344 + }, + { + "epoch": 2.631661936297698, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 0.9716, + "step": 8345 + }, + { + "epoch": 2.6319772942289497, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 0.9538, + "step": 8346 + }, + { + "epoch": 2.632292652160202, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.9641, + "step": 8347 + }, + { + "epoch": 2.6326080100914537, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 0.9433, + "step": 8348 + }, + { + "epoch": 2.6329233680227055, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 0.8777, + "step": 8349 + }, + { + "epoch": 2.633238725953958, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 0.909, + "step": 8350 + }, + { + "epoch": 2.6335540838852096, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 0.9236, + "step": 8351 + }, + { + "epoch": 2.6338694418164614, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 0.9913, + "step": 8352 + }, + { + "epoch": 2.6341847997477137, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 0.9173, + "step": 8353 + }, + { + "epoch": 2.6345001576789655, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 0.9149, + "step": 8354 + }, + { + "epoch": 2.6348155156102178, + "grad_norm": 0.330078125, + "learning_rate": 0.0002, + "loss": 0.8632, + "step": 8355 + }, + { + "epoch": 2.6351308735414696, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 0.9797, + "step": 8356 + }, + { + "epoch": 2.6354462314727214, + "grad_norm": 0.353515625, + "learning_rate": 0.0002, + "loss": 0.9957, + "step": 8357 + }, + { + "epoch": 2.6357615894039736, + "grad_norm": 0.388671875, + "learning_rate": 0.0002, + "loss": 0.9713, + "step": 8358 + }, + { + "epoch": 2.6360769473352255, + "grad_norm": 0.392578125, + "learning_rate": 0.0002, + "loss": 0.7941, + "step": 8359 + }, + { + "epoch": 2.6363923052664777, + "grad_norm": 0.392578125, + "learning_rate": 0.0002, + "loss": 0.8083, + "step": 8360 + }, + { + "epoch": 2.6367076631977295, + "grad_norm": 0.42578125, + "learning_rate": 0.0002, + "loss": 0.9812, + "step": 8361 + }, + { + "epoch": 2.6370230211289813, + "grad_norm": 0.40625, + "learning_rate": 0.0002, + "loss": 0.9033, + "step": 8362 + }, + { + "epoch": 2.6373383790602336, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 1.0565, + "step": 8363 + }, + { + "epoch": 2.6376537369914854, + "grad_norm": 0.458984375, + "learning_rate": 0.0002, + "loss": 0.9257, + "step": 8364 + }, + { + "epoch": 2.6379690949227372, + "grad_norm": 0.46875, + "learning_rate": 0.0002, + "loss": 0.9317, + "step": 8365 + }, + { + "epoch": 2.6382844528539895, + "grad_norm": 0.466796875, + "learning_rate": 0.0002, + "loss": 0.8947, + "step": 8366 + }, + { + "epoch": 2.6385998107852413, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 0.8795, + "step": 8367 + }, + { + "epoch": 2.638915168716493, + "grad_norm": 0.7109375, + "learning_rate": 0.0002, + "loss": 0.8664, + "step": 8368 + }, + { + "epoch": 2.6392305266477454, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 0.9293, + "step": 8369 + }, + { + "epoch": 2.639545884578997, + "grad_norm": 0.56640625, + "learning_rate": 0.0002, + "loss": 0.8329, + "step": 8370 + }, + { + "epoch": 2.639861242510249, + "grad_norm": 0.6484375, + "learning_rate": 0.0002, + "loss": 0.8906, + "step": 8371 + }, + { + "epoch": 2.6401766004415013, + "grad_norm": 0.58984375, + "learning_rate": 0.0002, + "loss": 0.8953, + "step": 8372 + }, + { + "epoch": 2.640491958372753, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 0.8707, + "step": 8373 + }, + { + "epoch": 2.640807316304005, + "grad_norm": 0.8515625, + "learning_rate": 0.0002, + "loss": 1.0952, + "step": 8374 + }, + { + "epoch": 2.641122674235257, + "grad_norm": 0.875, + "learning_rate": 0.0002, + "loss": 1.0338, + "step": 8375 + }, + { + "epoch": 2.641438032166509, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 1.0022, + "step": 8376 + }, + { + "epoch": 2.641753390097761, + "grad_norm": 0.7421875, + "learning_rate": 0.0002, + "loss": 0.9621, + "step": 8377 + }, + { + "epoch": 2.642068748029013, + "grad_norm": 1.078125, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 8378 + }, + { + "epoch": 2.642384105960265, + "grad_norm": 0.93359375, + "learning_rate": 0.0002, + "loss": 1.3822, + "step": 8379 + }, + { + "epoch": 2.6426994638915167, + "grad_norm": 1.2421875, + "learning_rate": 0.0002, + "loss": 1.3891, + "step": 8380 + }, + { + "epoch": 2.643014821822769, + "grad_norm": 1.1640625, + "learning_rate": 0.0002, + "loss": 1.4226, + "step": 8381 + }, + { + "epoch": 2.6433301797540207, + "grad_norm": 1.2421875, + "learning_rate": 0.0002, + "loss": 1.581, + "step": 8382 + }, + { + "epoch": 2.6436455376852726, + "grad_norm": 1.28125, + "learning_rate": 0.0002, + "loss": 1.6177, + "step": 8383 + }, + { + "epoch": 2.643960895616525, + "grad_norm": 1.34375, + "learning_rate": 0.0002, + "loss": 1.6245, + "step": 8384 + }, + { + "epoch": 2.6442762535477766, + "grad_norm": 1.0546875, + "learning_rate": 0.0002, + "loss": 1.601, + "step": 8385 + }, + { + "epoch": 2.6445916114790284, + "grad_norm": 1.4765625, + "learning_rate": 0.0002, + "loss": 1.9155, + "step": 8386 + }, + { + "epoch": 2.6449069694102807, + "grad_norm": 1.1640625, + "learning_rate": 0.0002, + "loss": 1.8645, + "step": 8387 + }, + { + "epoch": 2.6452223273415325, + "grad_norm": 1.1640625, + "learning_rate": 0.0002, + "loss": 1.435, + "step": 8388 + }, + { + "epoch": 2.6455376852727848, + "grad_norm": 1.59375, + "learning_rate": 0.0002, + "loss": 1.6594, + "step": 8389 + }, + { + "epoch": 2.6458530432040366, + "grad_norm": 1.078125, + "learning_rate": 0.0002, + "loss": 1.4164, + "step": 8390 + }, + { + "epoch": 2.6461684011352884, + "grad_norm": 1.1640625, + "learning_rate": 0.0002, + "loss": 1.5453, + "step": 8391 + }, + { + "epoch": 2.6464837590665407, + "grad_norm": 1.765625, + "learning_rate": 0.0002, + "loss": 1.8041, + "step": 8392 + }, + { + "epoch": 2.6467991169977925, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.9193, + "step": 8393 + }, + { + "epoch": 2.6471144749290447, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.0166, + "step": 8394 + }, + { + "epoch": 2.6474298328602965, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 0.9998, + "step": 8395 + }, + { + "epoch": 2.6477451907915484, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 0.9931, + "step": 8396 + }, + { + "epoch": 2.6480605487228006, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.0153, + "step": 8397 + }, + { + "epoch": 2.6483759066540524, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 0.9009, + "step": 8398 + }, + { + "epoch": 2.6486912645853042, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.8218, + "step": 8399 + }, + { + "epoch": 2.6490066225165565, + "grad_norm": 0.330078125, + "learning_rate": 0.0002, + "loss": 0.9841, + "step": 8400 + }, + { + "epoch": 2.6493219804478083, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 0.8838, + "step": 8401 + }, + { + "epoch": 2.64963733837906, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 0.8797, + "step": 8402 + }, + { + "epoch": 2.6499526963103124, + "grad_norm": 0.330078125, + "learning_rate": 0.0002, + "loss": 0.966, + "step": 8403 + }, + { + "epoch": 2.650268054241564, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 0.8502, + "step": 8404 + }, + { + "epoch": 2.650583412172816, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 0.9347, + "step": 8405 + }, + { + "epoch": 2.6508987701040683, + "grad_norm": 0.349609375, + "learning_rate": 0.0002, + "loss": 1.006, + "step": 8406 + }, + { + "epoch": 2.65121412803532, + "grad_norm": 0.447265625, + "learning_rate": 0.0002, + "loss": 0.9867, + "step": 8407 + }, + { + "epoch": 2.651529485966572, + "grad_norm": 0.41015625, + "learning_rate": 0.0002, + "loss": 0.9442, + "step": 8408 + }, + { + "epoch": 2.651844843897824, + "grad_norm": 0.484375, + "learning_rate": 0.0002, + "loss": 1.007, + "step": 8409 + }, + { + "epoch": 2.652160201829076, + "grad_norm": 0.447265625, + "learning_rate": 0.0002, + "loss": 1.1441, + "step": 8410 + }, + { + "epoch": 2.652475559760328, + "grad_norm": 0.42578125, + "learning_rate": 0.0002, + "loss": 0.8795, + "step": 8411 + }, + { + "epoch": 2.65279091769158, + "grad_norm": 0.421875, + "learning_rate": 0.0002, + "loss": 0.8695, + "step": 8412 + }, + { + "epoch": 2.653106275622832, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 0.8928, + "step": 8413 + }, + { + "epoch": 2.6534216335540837, + "grad_norm": 0.5390625, + "learning_rate": 0.0002, + "loss": 1.0061, + "step": 8414 + }, + { + "epoch": 2.653736991485336, + "grad_norm": 0.48046875, + "learning_rate": 0.0002, + "loss": 0.8245, + "step": 8415 + }, + { + "epoch": 2.653736991485336, + "eval_loss": 1.7784637212753296, + "eval_runtime": 149.0243, + "eval_samples_per_second": 6.71, + "eval_steps_per_second": 6.71, + "step": 8415 + }, + { + "epoch": 2.653736991485336, + "mmlu_eval_accuracy": 0.6058561628413719, + "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365, + "mmlu_eval_accuracy_anatomy": 0.5714285714285714, + "mmlu_eval_accuracy_astronomy": 0.625, + "mmlu_eval_accuracy_business_ethics": 0.8181818181818182, + "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138, + "mmlu_eval_accuracy_college_biology": 0.875, + "mmlu_eval_accuracy_college_chemistry": 0.375, + "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365, + "mmlu_eval_accuracy_college_mathematics": 0.6363636363636364, + "mmlu_eval_accuracy_college_medicine": 0.45454545454545453, + "mmlu_eval_accuracy_college_physics": 0.45454545454545453, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.5769230769230769, + "mmlu_eval_accuracy_econometrics": 0.5833333333333334, + "mmlu_eval_accuracy_electrical_engineering": 0.5, + "mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536, + "mmlu_eval_accuracy_formal_logic": 0.14285714285714285, + "mmlu_eval_accuracy_global_facts": 0.2, + "mmlu_eval_accuracy_high_school_biology": 0.625, + "mmlu_eval_accuracy_high_school_chemistry": 0.5, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209, + "mmlu_eval_accuracy_high_school_mathematics": 0.41379310344827586, + "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384, + "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354, + "mmlu_eval_accuracy_high_school_psychology": 0.8, + "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, + "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, + "mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693, + "mmlu_eval_accuracy_human_aging": 0.6521739130434783, + "mmlu_eval_accuracy_human_sexuality": 0.5, + "mmlu_eval_accuracy_international_law": 0.9230769230769231, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.7777777777777778, + "mmlu_eval_accuracy_machine_learning": 0.5454545454545454, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.68, + "mmlu_eval_accuracy_medical_genetics": 1.0, + "mmlu_eval_accuracy_miscellaneous": 0.7093023255813954, + "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527, + "mmlu_eval_accuracy_moral_scenarios": 0.27, + "mmlu_eval_accuracy_nutrition": 0.7272727272727273, + "mmlu_eval_accuracy_philosophy": 0.6470588235294118, + "mmlu_eval_accuracy_prehistory": 0.6, + "mmlu_eval_accuracy_professional_accounting": 0.5161290322580645, + "mmlu_eval_accuracy_professional_law": 0.38235294117647056, + "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806, + "mmlu_eval_accuracy_professional_psychology": 0.6086956521739131, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7037037037037037, + "mmlu_eval_accuracy_sociology": 1.0, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.5, + "mmlu_eval_accuracy_world_religions": 0.8421052631578947, + "mmlu_loss": 6.248817355083064, + "step": 8415 + }, + { + "epoch": 2.6540523494165877, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 0.9636, + "step": 8416 + }, + { + "epoch": 2.6543677073478396, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 0.8851, + "step": 8417 + }, + { + "epoch": 2.654683065279092, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 0.8936, + "step": 8418 + }, + { + "epoch": 2.6549984232103436, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 0.8408, + "step": 8419 + }, + { + "epoch": 2.6553137811415954, + "grad_norm": 0.578125, + "learning_rate": 0.0002, + "loss": 0.801, + "step": 8420 + }, + { + "epoch": 2.6556291390728477, + "grad_norm": 0.6171875, + "learning_rate": 0.0002, + "loss": 0.9722, + "step": 8421 + }, + { + "epoch": 2.6559444970040995, + "grad_norm": 0.6875, + "learning_rate": 0.0002, + "loss": 1.1025, + "step": 8422 + }, + { + "epoch": 2.6562598549353518, + "grad_norm": 0.94140625, + "learning_rate": 0.0002, + "loss": 1.1493, + "step": 8423 + }, + { + "epoch": 2.6565752128666036, + "grad_norm": 0.84765625, + "learning_rate": 0.0002, + "loss": 1.1329, + "step": 8424 + }, + { + "epoch": 2.6568905707978554, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 1.3641, + "step": 8425 + }, + { + "epoch": 2.6572059287291077, + "grad_norm": 0.9765625, + "learning_rate": 0.0002, + "loss": 1.2831, + "step": 8426 + }, + { + "epoch": 2.6575212866603595, + "grad_norm": 0.94140625, + "learning_rate": 0.0002, + "loss": 1.3864, + "step": 8427 + }, + { + "epoch": 2.6578366445916117, + "grad_norm": 1.1328125, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 8428 + }, + { + "epoch": 2.6581520025228635, + "grad_norm": 1.4921875, + "learning_rate": 0.0002, + "loss": 1.398, + "step": 8429 + }, + { + "epoch": 2.6584673604541154, + "grad_norm": 1.015625, + "learning_rate": 0.0002, + "loss": 1.3721, + "step": 8430 + }, + { + "epoch": 2.6587827183853676, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 1.5505, + "step": 8431 + }, + { + "epoch": 2.6590980763166194, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 1.3684, + "step": 8432 + }, + { + "epoch": 2.6594134342478712, + "grad_norm": 1.328125, + "learning_rate": 0.0002, + "loss": 1.6557, + "step": 8433 + }, + { + "epoch": 2.6597287921791235, + "grad_norm": 1.2578125, + "learning_rate": 0.0002, + "loss": 1.5818, + "step": 8434 + }, + { + "epoch": 2.6600441501103753, + "grad_norm": 1.515625, + "learning_rate": 0.0002, + "loss": 1.7413, + "step": 8435 + }, + { + "epoch": 2.660359508041627, + "grad_norm": 1.265625, + "learning_rate": 0.0002, + "loss": 1.5215, + "step": 8436 + }, + { + "epoch": 2.6606748659728794, + "grad_norm": 1.2109375, + "learning_rate": 0.0002, + "loss": 1.7544, + "step": 8437 + }, + { + "epoch": 2.660990223904131, + "grad_norm": 1.296875, + "learning_rate": 0.0002, + "loss": 1.4633, + "step": 8438 + }, + { + "epoch": 2.661305581835383, + "grad_norm": 1.09375, + "learning_rate": 0.0002, + "loss": 1.1996, + "step": 8439 + }, + { + "epoch": 2.6616209397666353, + "grad_norm": 2.046875, + "learning_rate": 0.0002, + "loss": 1.7177, + "step": 8440 + }, + { + "epoch": 2.661936297697887, + "grad_norm": 1.234375, + "learning_rate": 0.0002, + "loss": 1.4886, + "step": 8441 + }, + { + "epoch": 2.662251655629139, + "grad_norm": 1.5, + "learning_rate": 0.0002, + "loss": 1.6401, + "step": 8442 + }, + { + "epoch": 2.662567013560391, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.134, + "step": 8443 + }, + { + "epoch": 2.662882371491643, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 1.0713, + "step": 8444 + }, + { + "epoch": 2.663197729422895, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 0.9351, + "step": 8445 + }, + { + "epoch": 2.663513087354147, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 0.8635, + "step": 8446 + }, + { + "epoch": 2.663828445285399, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 0.889, + "step": 8447 + }, + { + "epoch": 2.6641438032166507, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 0.974, + "step": 8448 + }, + { + "epoch": 2.664459161147903, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.0101, + "step": 8449 + }, + { + "epoch": 2.6647745190791547, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.0179, + "step": 8450 + }, + { + "epoch": 2.6650898770104066, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 0.9631, + "step": 8451 + }, + { + "epoch": 2.665405234941659, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 0.9014, + "step": 8452 + }, + { + "epoch": 2.6657205928729106, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 0.958, + "step": 8453 + }, + { + "epoch": 2.666035950804163, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 1.0076, + "step": 8454 + }, + { + "epoch": 2.6663513087354147, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 0.9114, + "step": 8455 + }, + { + "epoch": 2.6666666666666665, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 0.9491, + "step": 8456 + }, + { + "epoch": 2.6669820245979188, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 0.9706, + "step": 8457 + }, + { + "epoch": 2.6672973825291706, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 0.8619, + "step": 8458 + }, + { + "epoch": 2.667612740460423, + "grad_norm": 0.412109375, + "learning_rate": 0.0002, + "loss": 0.9969, + "step": 8459 + }, + { + "epoch": 2.6679280983916747, + "grad_norm": 0.42578125, + "learning_rate": 0.0002, + "loss": 0.9667, + "step": 8460 + }, + { + "epoch": 2.6682434563229265, + "grad_norm": 0.427734375, + "learning_rate": 0.0002, + "loss": 0.9044, + "step": 8461 + }, + { + "epoch": 2.6685588142541787, + "grad_norm": 0.400390625, + "learning_rate": 0.0002, + "loss": 0.866, + "step": 8462 + }, + { + "epoch": 2.6688741721854305, + "grad_norm": 0.4765625, + "learning_rate": 0.0002, + "loss": 0.8816, + "step": 8463 + }, + { + "epoch": 2.6691895301166824, + "grad_norm": 0.42578125, + "learning_rate": 0.0002, + "loss": 0.88, + "step": 8464 + }, + { + "epoch": 2.6695048880479346, + "grad_norm": 0.458984375, + "learning_rate": 0.0002, + "loss": 0.9269, + "step": 8465 + }, + { + "epoch": 2.6698202459791864, + "grad_norm": 0.494140625, + "learning_rate": 0.0002, + "loss": 0.9773, + "step": 8466 + }, + { + "epoch": 2.6701356039104382, + "grad_norm": 0.46875, + "learning_rate": 0.0002, + "loss": 0.8478, + "step": 8467 + }, + { + "epoch": 2.6704509618416905, + "grad_norm": 0.5390625, + "learning_rate": 0.0002, + "loss": 0.9933, + "step": 8468 + }, + { + "epoch": 2.6707663197729423, + "grad_norm": 0.63671875, + "learning_rate": 0.0002, + "loss": 0.8723, + "step": 8469 + }, + { + "epoch": 2.671081677704194, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 0.8921, + "step": 8470 + }, + { + "epoch": 2.6713970356354464, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 0.9355, + "step": 8471 + }, + { + "epoch": 2.671712393566698, + "grad_norm": 0.75, + "learning_rate": 0.0002, + "loss": 0.9279, + "step": 8472 + }, + { + "epoch": 2.67202775149795, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 0.8707, + "step": 8473 + }, + { + "epoch": 2.6723431094292023, + "grad_norm": 0.7578125, + "learning_rate": 0.0002, + "loss": 0.932, + "step": 8474 + }, + { + "epoch": 2.672658467360454, + "grad_norm": 0.765625, + "learning_rate": 0.0002, + "loss": 1.1987, + "step": 8475 + }, + { + "epoch": 2.672973825291706, + "grad_norm": 0.8046875, + "learning_rate": 0.0002, + "loss": 0.9958, + "step": 8476 + }, + { + "epoch": 2.673289183222958, + "grad_norm": 0.97265625, + "learning_rate": 0.0002, + "loss": 1.0953, + "step": 8477 + }, + { + "epoch": 2.67360454115421, + "grad_norm": 0.890625, + "learning_rate": 0.0002, + "loss": 1.0862, + "step": 8478 + }, + { + "epoch": 2.673919899085462, + "grad_norm": 1.0390625, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 8479 + }, + { + "epoch": 2.674235257016714, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 1.4424, + "step": 8480 + }, + { + "epoch": 2.674550614947966, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 1.3973, + "step": 8481 + }, + { + "epoch": 2.6748659728792177, + "grad_norm": 1.21875, + "learning_rate": 0.0002, + "loss": 1.7299, + "step": 8482 + }, + { + "epoch": 2.67518133081047, + "grad_norm": 1.109375, + "learning_rate": 0.0002, + "loss": 1.5917, + "step": 8483 + }, + { + "epoch": 2.6754966887417218, + "grad_norm": 1.09375, + "learning_rate": 0.0002, + "loss": 1.4928, + "step": 8484 + }, + { + "epoch": 2.6758120466729736, + "grad_norm": 1.078125, + "learning_rate": 0.0002, + "loss": 1.4942, + "step": 8485 + }, + { + "epoch": 2.676127404604226, + "grad_norm": 1.109375, + "learning_rate": 0.0002, + "loss": 1.8387, + "step": 8486 + }, + { + "epoch": 2.6764427625354776, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 1.4349, + "step": 8487 + }, + { + "epoch": 2.67675812046673, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 1.5294, + "step": 8488 + }, + { + "epoch": 2.6770734783979817, + "grad_norm": 1.4765625, + "learning_rate": 0.0002, + "loss": 1.6871, + "step": 8489 + }, + { + "epoch": 2.6773888363292335, + "grad_norm": 0.9765625, + "learning_rate": 0.0002, + "loss": 1.412, + "step": 8490 + }, + { + "epoch": 2.677704194260486, + "grad_norm": 0.94921875, + "learning_rate": 0.0002, + "loss": 1.4235, + "step": 8491 + }, + { + "epoch": 2.6780195521917376, + "grad_norm": 1.671875, + "learning_rate": 0.0002, + "loss": 1.9453, + "step": 8492 + }, + { + "epoch": 2.67833491012299, + "grad_norm": 0.244140625, + "learning_rate": 0.0002, + "loss": 0.905, + "step": 8493 + }, + { + "epoch": 2.6786502680542417, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 0.8523, + "step": 8494 + }, + { + "epoch": 2.6789656259854935, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 0.9937, + "step": 8495 + }, + { + "epoch": 2.6792809839167457, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 0.9909, + "step": 8496 + }, + { + "epoch": 2.6795963418479976, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 1.0547, + "step": 8497 + }, + { + "epoch": 2.6799116997792494, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 0.9141, + "step": 8498 + }, + { + "epoch": 2.6802270577105016, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 0.9216, + "step": 8499 + }, + { + "epoch": 2.6805424156417534, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 0.8996, + "step": 8500 + }, + { + "epoch": 2.6808577735730053, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 0.9013, + "step": 8501 + }, + { + "epoch": 2.6811731315042575, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 0.9566, + "step": 8502 + }, + { + "epoch": 2.6814884894355093, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 0.817, + "step": 8503 + }, + { + "epoch": 2.681803847366761, + "grad_norm": 0.408203125, + "learning_rate": 0.0002, + "loss": 1.0554, + "step": 8504 + }, + { + "epoch": 2.6821192052980134, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 0.9702, + "step": 8505 + }, + { + "epoch": 2.682434563229265, + "grad_norm": 0.380859375, + "learning_rate": 0.0002, + "loss": 0.9518, + "step": 8506 + }, + { + "epoch": 2.682749921160517, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 0.8241, + "step": 8507 + }, + { + "epoch": 2.6830652790917693, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 0.9721, + "step": 8508 + }, + { + "epoch": 2.683380637023021, + "grad_norm": 0.39453125, + "learning_rate": 0.0002, + "loss": 1.0466, + "step": 8509 + }, + { + "epoch": 2.683695994954273, + "grad_norm": 0.392578125, + "learning_rate": 0.0002, + "loss": 0.9418, + "step": 8510 + }, + { + "epoch": 2.684011352885525, + "grad_norm": 0.388671875, + "learning_rate": 0.0002, + "loss": 0.9628, + "step": 8511 + }, + { + "epoch": 2.684326710816777, + "grad_norm": 0.419921875, + "learning_rate": 0.0002, + "loss": 0.9359, + "step": 8512 + }, + { + "epoch": 2.684642068748029, + "grad_norm": 0.48046875, + "learning_rate": 0.0002, + "loss": 1.0206, + "step": 8513 + }, + { + "epoch": 2.684957426679281, + "grad_norm": 0.44921875, + "learning_rate": 0.0002, + "loss": 0.8374, + "step": 8514 + }, + { + "epoch": 2.685272784610533, + "grad_norm": 0.48828125, + "learning_rate": 0.0002, + "loss": 0.9136, + "step": 8515 + }, + { + "epoch": 2.6855881425417847, + "grad_norm": 0.486328125, + "learning_rate": 0.0002, + "loss": 0.9261, + "step": 8516 + }, + { + "epoch": 2.685903500473037, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 0.8003, + "step": 8517 + }, + { + "epoch": 2.6862188584042888, + "grad_norm": 0.6015625, + "learning_rate": 0.0002, + "loss": 1.0262, + "step": 8518 + }, + { + "epoch": 2.6865342163355406, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 0.9202, + "step": 8519 + }, + { + "epoch": 2.686849574266793, + "grad_norm": 0.5703125, + "learning_rate": 0.0002, + "loss": 0.8834, + "step": 8520 + }, + { + "epoch": 2.6871649321980446, + "grad_norm": 0.62890625, + "learning_rate": 0.0002, + "loss": 0.8742, + "step": 8521 + }, + { + "epoch": 2.687480290129297, + "grad_norm": 0.703125, + "learning_rate": 0.0002, + "loss": 0.9038, + "step": 8522 + }, + { + "epoch": 2.6877956480605487, + "grad_norm": 0.6640625, + "learning_rate": 0.0002, + "loss": 0.8713, + "step": 8523 + }, + { + "epoch": 2.6881110059918005, + "grad_norm": 0.83984375, + "learning_rate": 0.0002, + "loss": 1.0237, + "step": 8524 + }, + { + "epoch": 2.688426363923053, + "grad_norm": 0.84765625, + "learning_rate": 0.0002, + "loss": 0.9381, + "step": 8525 + }, + { + "epoch": 2.6887417218543046, + "grad_norm": 1.328125, + "learning_rate": 0.0002, + "loss": 1.1836, + "step": 8526 + }, + { + "epoch": 2.689057079785557, + "grad_norm": 1.515625, + "learning_rate": 0.0002, + "loss": 1.393, + "step": 8527 + }, + { + "epoch": 2.6893724377168087, + "grad_norm": 1.171875, + "learning_rate": 0.0002, + "loss": 1.2504, + "step": 8528 + }, + { + "epoch": 2.6896877956480605, + "grad_norm": 1.21875, + "learning_rate": 0.0002, + "loss": 1.4907, + "step": 8529 + }, + { + "epoch": 2.6900031535793127, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 1.4824, + "step": 8530 + }, + { + "epoch": 2.6903185115105646, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 1.2758, + "step": 8531 + }, + { + "epoch": 2.6906338694418164, + "grad_norm": 1.03125, + "learning_rate": 0.0002, + "loss": 1.3677, + "step": 8532 + }, + { + "epoch": 2.6909492273730686, + "grad_norm": 1.15625, + "learning_rate": 0.0002, + "loss": 1.5342, + "step": 8533 + }, + { + "epoch": 2.6912645853043204, + "grad_norm": 1.2109375, + "learning_rate": 0.0002, + "loss": 1.5117, + "step": 8534 + }, + { + "epoch": 2.6915799432355723, + "grad_norm": 1.125, + "learning_rate": 0.0002, + "loss": 1.7369, + "step": 8535 + }, + { + "epoch": 2.6918953011668245, + "grad_norm": 0.984375, + "learning_rate": 0.0002, + "loss": 1.4499, + "step": 8536 + }, + { + "epoch": 2.6922106590980763, + "grad_norm": 1.75, + "learning_rate": 0.0002, + "loss": 1.6979, + "step": 8537 + }, + { + "epoch": 2.692526017029328, + "grad_norm": 1.234375, + "learning_rate": 0.0002, + "loss": 1.6838, + "step": 8538 + }, + { + "epoch": 2.6928413749605804, + "grad_norm": 1.6328125, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 8539 + }, + { + "epoch": 2.693156732891832, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 1.7236, + "step": 8540 + }, + { + "epoch": 2.693472090823084, + "grad_norm": 1.6875, + "learning_rate": 0.0002, + "loss": 1.4675, + "step": 8541 + }, + { + "epoch": 2.6937874487543363, + "grad_norm": 1.1875, + "learning_rate": 0.0002, + "loss": 1.6234, + "step": 8542 + }, + { + "epoch": 2.694102806685588, + "grad_norm": 0.244140625, + "learning_rate": 0.0002, + "loss": 1.0407, + "step": 8543 + }, + { + "epoch": 2.69441816461684, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 1.0696, + "step": 8544 + }, + { + "epoch": 2.694733522548092, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 1.0691, + "step": 8545 + }, + { + "epoch": 2.695048880479344, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 0.9643, + "step": 8546 + }, + { + "epoch": 2.695364238410596, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.0169, + "step": 8547 + }, + { + "epoch": 2.695679596341848, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 0.9707, + "step": 8548 + }, + { + "epoch": 2.6959949542731, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 0.8954, + "step": 8549 + }, + { + "epoch": 2.6963103122043517, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 0.9504, + "step": 8550 + }, + { + "epoch": 2.696625670135604, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 1.0446, + "step": 8551 + }, + { + "epoch": 2.6969410280668558, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 1.0975, + "step": 8552 + }, + { + "epoch": 2.6972563859981076, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 0.8267, + "step": 8553 + }, + { + "epoch": 2.69757174392936, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 0.917, + "step": 8554 + }, + { + "epoch": 2.6978871018606116, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 0.9691, + "step": 8555 + }, + { + "epoch": 2.698202459791864, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 0.9339, + "step": 8556 + }, + { + "epoch": 2.6985178177231157, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 0.8607, + "step": 8557 + }, + { + "epoch": 2.698833175654368, + "grad_norm": 0.41796875, + "learning_rate": 0.0002, + "loss": 0.9507, + "step": 8558 + }, + { + "epoch": 2.69914853358562, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 0.9124, + "step": 8559 + }, + { + "epoch": 2.6994638915168716, + "grad_norm": 0.408203125, + "learning_rate": 0.0002, + "loss": 0.9933, + "step": 8560 + }, + { + "epoch": 2.699779249448124, + "grad_norm": 0.400390625, + "learning_rate": 0.0002, + "loss": 0.8861, + "step": 8561 + }, + { + "epoch": 2.7000946073793757, + "grad_norm": 0.4921875, + "learning_rate": 0.0002, + "loss": 0.9056, + "step": 8562 + }, + { + "epoch": 2.7004099653106275, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 1.0461, + "step": 8563 + }, + { + "epoch": 2.7007253232418797, + "grad_norm": 0.48828125, + "learning_rate": 0.0002, + "loss": 0.9703, + "step": 8564 + }, + { + "epoch": 2.7010406811731316, + "grad_norm": 0.55859375, + "learning_rate": 0.0002, + "loss": 1.0021, + "step": 8565 + }, + { + "epoch": 2.7013560391043834, + "grad_norm": 0.482421875, + "learning_rate": 0.0002, + "loss": 0.9137, + "step": 8566 + }, + { + "epoch": 2.7016713970356356, + "grad_norm": 0.5546875, + "learning_rate": 0.0002, + "loss": 0.904, + "step": 8567 + }, + { + "epoch": 2.7019867549668874, + "grad_norm": 0.484375, + "learning_rate": 0.0002, + "loss": 0.8743, + "step": 8568 + }, + { + "epoch": 2.7023021128981393, + "grad_norm": 0.60546875, + "learning_rate": 0.0002, + "loss": 0.8939, + "step": 8569 + }, + { + "epoch": 2.7026174708293915, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 0.8797, + "step": 8570 + }, + { + "epoch": 2.7029328287606433, + "grad_norm": 0.6640625, + "learning_rate": 0.0002, + "loss": 0.9881, + "step": 8571 + }, + { + "epoch": 2.703248186691895, + "grad_norm": 0.81640625, + "learning_rate": 0.0002, + "loss": 1.0076, + "step": 8572 + }, + { + "epoch": 2.7035635446231474, + "grad_norm": 0.90234375, + "learning_rate": 0.0002, + "loss": 1.1156, + "step": 8573 + }, + { + "epoch": 2.703878902554399, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 1.1386, + "step": 8574 + }, + { + "epoch": 2.704194260485651, + "grad_norm": 0.77734375, + "learning_rate": 0.0002, + "loss": 1.026, + "step": 8575 + }, + { + "epoch": 2.7045096184169033, + "grad_norm": 0.90234375, + "learning_rate": 0.0002, + "loss": 1.1362, + "step": 8576 + }, + { + "epoch": 2.704824976348155, + "grad_norm": 0.90625, + "learning_rate": 0.0002, + "loss": 1.0767, + "step": 8577 + }, + { + "epoch": 2.705140334279407, + "grad_norm": 0.83203125, + "learning_rate": 0.0002, + "loss": 1.2279, + "step": 8578 + }, + { + "epoch": 2.705455692210659, + "grad_norm": 1.1953125, + "learning_rate": 0.0002, + "loss": 1.4865, + "step": 8579 + }, + { + "epoch": 2.705771050141911, + "grad_norm": 1.1875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 8580 + }, + { + "epoch": 2.706086408073163, + "grad_norm": 0.98828125, + "learning_rate": 0.0002, + "loss": 1.2998, + "step": 8581 + }, + { + "epoch": 2.706401766004415, + "grad_norm": 1.0859375, + "learning_rate": 0.0002, + "loss": 1.7377, + "step": 8582 + }, + { + "epoch": 2.706717123935667, + "grad_norm": 1.2109375, + "learning_rate": 0.0002, + "loss": 1.6084, + "step": 8583 + }, + { + "epoch": 2.7070324818669187, + "grad_norm": 1.2265625, + "learning_rate": 0.0002, + "loss": 1.772, + "step": 8584 + }, + { + "epoch": 2.707347839798171, + "grad_norm": 1.3125, + "learning_rate": 0.0002, + "loss": 1.5837, + "step": 8585 + }, + { + "epoch": 2.7076631977294228, + "grad_norm": 2.046875, + "learning_rate": 0.0002, + "loss": 1.9561, + "step": 8586 + }, + { + "epoch": 2.707978555660675, + "grad_norm": 1.25, + "learning_rate": 0.0002, + "loss": 1.289, + "step": 8587 + }, + { + "epoch": 2.708293913591927, + "grad_norm": 1.2109375, + "learning_rate": 0.0002, + "loss": 1.6817, + "step": 8588 + }, + { + "epoch": 2.7086092715231787, + "grad_norm": 1.265625, + "learning_rate": 0.0002, + "loss": 1.6513, + "step": 8589 + }, + { + "epoch": 2.708924629454431, + "grad_norm": 1.3125, + "learning_rate": 0.0002, + "loss": 1.4062, + "step": 8590 + }, + { + "epoch": 2.7092399873856827, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 1.3922, + "step": 8591 + }, + { + "epoch": 2.709555345316935, + "grad_norm": 1.484375, + "learning_rate": 0.0002, + "loss": 1.7141, + "step": 8592 + }, + { + "epoch": 2.709870703248187, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 0.9034, + "step": 8593 + }, + { + "epoch": 2.7101860611794386, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 1.0435, + "step": 8594 + }, + { + "epoch": 2.710501419110691, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 0.9509, + "step": 8595 + }, + { + "epoch": 2.7108167770419427, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 0.9075, + "step": 8596 + }, + { + "epoch": 2.7111321349731945, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.8842, + "step": 8597 + }, + { + "epoch": 2.7114474929044468, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.9728, + "step": 8598 + }, + { + "epoch": 2.7117628508356986, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.0413, + "step": 8599 + }, + { + "epoch": 2.7120782087669504, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 0.8804, + "step": 8600 + }, + { + "epoch": 2.7123935666982026, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.9048, + "step": 8601 + }, + { + "epoch": 2.7127089246294545, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.8234, + "step": 8602 + }, + { + "epoch": 2.7127089246294545, + "eval_loss": 1.7827229499816895, + "eval_runtime": 148.672, + "eval_samples_per_second": 6.726, + "eval_steps_per_second": 6.726, + "step": 8602 + }, + { + "epoch": 2.7127089246294545, + "mmlu_eval_accuracy": 0.5968718355922845, + "mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453, + "mmlu_eval_accuracy_anatomy": 0.6428571428571429, + "mmlu_eval_accuracy_astronomy": 0.75, + "mmlu_eval_accuracy_business_ethics": 0.7272727272727273, + "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138, + "mmlu_eval_accuracy_college_biology": 0.75, + "mmlu_eval_accuracy_college_chemistry": 0.375, + "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727, + "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453, + "mmlu_eval_accuracy_college_medicine": 0.45454545454545453, + "mmlu_eval_accuracy_college_physics": 0.45454545454545453, + "mmlu_eval_accuracy_computer_security": 0.36363636363636365, + "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384, + "mmlu_eval_accuracy_econometrics": 0.5833333333333334, + "mmlu_eval_accuracy_electrical_engineering": 0.5625, + "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244, + "mmlu_eval_accuracy_formal_logic": 0.2857142857142857, + "mmlu_eval_accuracy_global_facts": 0.3, + "mmlu_eval_accuracy_high_school_biology": 0.53125, + "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093, + "mmlu_eval_accuracy_high_school_mathematics": 0.3793103448275862, + "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384, + "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354, + "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667, + "mmlu_eval_accuracy_high_school_statistics": 0.4782608695652174, + "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, + "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, + "mmlu_eval_accuracy_human_aging": 0.6521739130434783, + "mmlu_eval_accuracy_human_sexuality": 0.5, + "mmlu_eval_accuracy_international_law": 0.9230769230769231, + "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, + "mmlu_eval_accuracy_logical_fallacies": 0.7777777777777778, + "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.68, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7325581395348837, + "mmlu_eval_accuracy_moral_disputes": 0.631578947368421, + "mmlu_eval_accuracy_moral_scenarios": 0.22, + "mmlu_eval_accuracy_nutrition": 0.7575757575757576, + "mmlu_eval_accuracy_philosophy": 0.6176470588235294, + "mmlu_eval_accuracy_prehistory": 0.5714285714285714, + "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355, + "mmlu_eval_accuracy_professional_law": 0.40588235294117647, + "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806, + "mmlu_eval_accuracy_professional_psychology": 0.6231884057971014, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7407407407407407, + "mmlu_eval_accuracy_sociology": 0.9545454545454546, + "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, + "mmlu_eval_accuracy_virology": 0.5, + "mmlu_eval_accuracy_world_religions": 0.7368421052631579, + "mmlu_loss": 5.876555273690152, + "step": 8602 + }, + { + "epoch": 2.7130242825607063, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 0.979, + "step": 8603 + }, + { + "epoch": 2.7133396404919585, + "grad_norm": 0.34765625, + "learning_rate": 0.0002, + "loss": 0.8192, + "step": 8604 + }, + { + "epoch": 2.7136549984232103, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 0.8951, + "step": 8605 + }, + { + "epoch": 2.713970356354462, + "grad_norm": 0.37109375, + "learning_rate": 0.0002, + "loss": 0.997, + "step": 8606 + }, + { + "epoch": 2.7142857142857144, + "grad_norm": 0.41796875, + "learning_rate": 0.0002, + "loss": 0.8962, + "step": 8607 + }, + { + "epoch": 2.7146010722169662, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 0.9399, + "step": 8608 + }, + { + "epoch": 2.714916430148218, + "grad_norm": 0.4140625, + "learning_rate": 0.0002, + "loss": 0.8869, + "step": 8609 + }, + { + "epoch": 2.7152317880794703, + "grad_norm": 0.4375, + "learning_rate": 0.0002, + "loss": 0.9976, + "step": 8610 + }, + { + "epoch": 2.715547146010722, + "grad_norm": 0.421875, + "learning_rate": 0.0002, + "loss": 1.0515, + "step": 8611 + }, + { + "epoch": 2.715862503941974, + "grad_norm": 0.40234375, + "learning_rate": 0.0002, + "loss": 0.9828, + "step": 8612 + }, + { + "epoch": 2.716177861873226, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 0.9998, + "step": 8613 + }, + { + "epoch": 2.716493219804478, + "grad_norm": 0.50390625, + "learning_rate": 0.0002, + "loss": 1.0024, + "step": 8614 + }, + { + "epoch": 2.71680857773573, + "grad_norm": 0.44921875, + "learning_rate": 0.0002, + "loss": 0.9205, + "step": 8615 + }, + { + "epoch": 2.717123935666982, + "grad_norm": 0.486328125, + "learning_rate": 0.0002, + "loss": 0.8818, + "step": 8616 + }, + { + "epoch": 2.717439293598234, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 0.9112, + "step": 8617 + }, + { + "epoch": 2.7177546515294857, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 0.9105, + "step": 8618 + }, + { + "epoch": 2.718070009460738, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 0.8851, + "step": 8619 + }, + { + "epoch": 2.7183853673919898, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 0.9582, + "step": 8620 + }, + { + "epoch": 2.718700725323242, + "grad_norm": 0.7890625, + "learning_rate": 0.0002, + "loss": 1.1769, + "step": 8621 + }, + { + "epoch": 2.719016083254494, + "grad_norm": 0.5546875, + "learning_rate": 0.0002, + "loss": 0.911, + "step": 8622 + }, + { + "epoch": 2.7193314411857457, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 1.0185, + "step": 8623 + }, + { + "epoch": 2.719646799116998, + "grad_norm": 0.8515625, + "learning_rate": 0.0002, + "loss": 1.1763, + "step": 8624 + }, + { + "epoch": 2.7199621570482497, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 1.0549, + "step": 8625 + }, + { + "epoch": 2.720277514979502, + "grad_norm": 0.83984375, + "learning_rate": 0.0002, + "loss": 1.0238, + "step": 8626 + }, + { + "epoch": 2.720592872910754, + "grad_norm": 0.97265625, + "learning_rate": 0.0002, + "loss": 1.3687, + "step": 8627 + }, + { + "epoch": 2.7209082308420056, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 8628 + }, + { + "epoch": 2.721223588773258, + "grad_norm": 1.1796875, + "learning_rate": 0.0002, + "loss": 1.2306, + "step": 8629 + }, + { + "epoch": 2.7215389467045097, + "grad_norm": 1.2421875, + "learning_rate": 0.0002, + "loss": 1.4119, + "step": 8630 + }, + { + "epoch": 2.7218543046357615, + "grad_norm": 1.015625, + "learning_rate": 0.0002, + "loss": 1.3605, + "step": 8631 + }, + { + "epoch": 2.7221696625670138, + "grad_norm": 1.546875, + "learning_rate": 0.0002, + "loss": 1.6007, + "step": 8632 + }, + { + "epoch": 2.7224850204982656, + "grad_norm": 0.96484375, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 8633 + }, + { + "epoch": 2.7228003784295174, + "grad_norm": 0.99609375, + "learning_rate": 0.0002, + "loss": 1.3868, + "step": 8634 + }, + { + "epoch": 2.7231157363607696, + "grad_norm": 1.03125, + "learning_rate": 0.0002, + "loss": 1.5086, + "step": 8635 + }, + { + "epoch": 2.7234310942920215, + "grad_norm": 1.2109375, + "learning_rate": 0.0002, + "loss": 1.3647, + "step": 8636 + }, + { + "epoch": 2.7237464522232733, + "grad_norm": 1.140625, + "learning_rate": 0.0002, + "loss": 1.6972, + "step": 8637 + }, + { + "epoch": 2.7240618101545255, + "grad_norm": 1.109375, + "learning_rate": 0.0002, + "loss": 1.7756, + "step": 8638 + }, + { + "epoch": 2.7243771680857773, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 1.6014, + "step": 8639 + }, + { + "epoch": 2.724692526017029, + "grad_norm": 1.59375, + "learning_rate": 0.0002, + "loss": 1.8818, + "step": 8640 + }, + { + "epoch": 2.7250078839482814, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 1.4857, + "step": 8641 + }, + { + "epoch": 2.7253232418795332, + "grad_norm": 1.2265625, + "learning_rate": 0.0002, + "loss": 1.8357, + "step": 8642 + }, + { + "epoch": 2.725638599810785, + "grad_norm": 0.240234375, + "learning_rate": 0.0002, + "loss": 0.9693, + "step": 8643 + }, + { + "epoch": 2.7259539577420373, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 0.9802, + "step": 8644 + }, + { + "epoch": 2.726269315673289, + "grad_norm": 0.255859375, + "learning_rate": 0.0002, + "loss": 0.8928, + "step": 8645 + }, + { + "epoch": 2.726584673604541, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.0113, + "step": 8646 + }, + { + "epoch": 2.726900031535793, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 0.9075, + "step": 8647 + }, + { + "epoch": 2.727215389467045, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 1.0175, + "step": 8648 + }, + { + "epoch": 2.727530747398297, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.0166, + "step": 8649 + }, + { + "epoch": 2.727846105329549, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 0.9915, + "step": 8650 + }, + { + "epoch": 2.728161463260801, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 1.0107, + "step": 8651 + }, + { + "epoch": 2.7284768211920527, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 0.8958, + "step": 8652 + }, + { + "epoch": 2.728792179123305, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 0.9108, + "step": 8653 + }, + { + "epoch": 2.7291075370545568, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 0.9299, + "step": 8654 + }, + { + "epoch": 2.729422894985809, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 0.9751, + "step": 8655 + }, + { + "epoch": 2.729738252917061, + "grad_norm": 0.330078125, + "learning_rate": 0.0002, + "loss": 0.811, + "step": 8656 + }, + { + "epoch": 2.730053610848313, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 0.974, + "step": 8657 + }, + { + "epoch": 2.730368968779565, + "grad_norm": 0.40625, + "learning_rate": 0.0002, + "loss": 1.0373, + "step": 8658 + }, + { + "epoch": 2.7306843267108167, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 0.9956, + "step": 8659 + }, + { + "epoch": 2.730999684642069, + "grad_norm": 0.400390625, + "learning_rate": 0.0002, + "loss": 1.0711, + "step": 8660 + }, + { + "epoch": 2.731315042573321, + "grad_norm": 0.42578125, + "learning_rate": 0.0002, + "loss": 1.0385, + "step": 8661 + }, + { + "epoch": 2.7316304005045726, + "grad_norm": 0.43359375, + "learning_rate": 0.0002, + "loss": 0.9732, + "step": 8662 + }, + { + "epoch": 2.731945758435825, + "grad_norm": 0.46484375, + "learning_rate": 0.0002, + "loss": 1.0049, + "step": 8663 + }, + { + "epoch": 2.7322611163670767, + "grad_norm": 0.50390625, + "learning_rate": 0.0002, + "loss": 0.8765, + "step": 8664 + }, + { + "epoch": 2.7325764742983285, + "grad_norm": 0.625, + "learning_rate": 0.0002, + "loss": 0.8594, + "step": 8665 + }, + { + "epoch": 2.7328918322295808, + "grad_norm": 0.45703125, + "learning_rate": 0.0002, + "loss": 0.8661, + "step": 8666 + }, + { + "epoch": 2.7332071901608326, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 0.8482, + "step": 8667 + }, + { + "epoch": 2.7335225480920844, + "grad_norm": 0.61328125, + "learning_rate": 0.0002, + "loss": 0.8869, + "step": 8668 + }, + { + "epoch": 2.7338379060233367, + "grad_norm": 0.5703125, + "learning_rate": 0.0002, + "loss": 1.0652, + "step": 8669 + }, + { + "epoch": 2.7341532639545885, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 0.8867, + "step": 8670 + }, + { + "epoch": 2.7344686218858403, + "grad_norm": 0.6640625, + "learning_rate": 0.0002, + "loss": 1.0719, + "step": 8671 + }, + { + "epoch": 2.7347839798170925, + "grad_norm": 0.64453125, + "learning_rate": 0.0002, + "loss": 0.9785, + "step": 8672 + }, + { + "epoch": 2.7350993377483444, + "grad_norm": 0.64453125, + "learning_rate": 0.0002, + "loss": 0.9268, + "step": 8673 + }, + { + "epoch": 2.735414695679596, + "grad_norm": 0.70703125, + "learning_rate": 0.0002, + "loss": 0.9529, + "step": 8674 + }, + { + "epoch": 2.7357300536108484, + "grad_norm": 0.83984375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 8675 + }, + { + "epoch": 2.7360454115421002, + "grad_norm": 0.86328125, + "learning_rate": 0.0002, + "loss": 1.1507, + "step": 8676 + }, + { + "epoch": 2.736360769473352, + "grad_norm": 0.87890625, + "learning_rate": 0.0002, + "loss": 1.2389, + "step": 8677 + }, + { + "epoch": 2.7366761274046043, + "grad_norm": 1.1640625, + "learning_rate": 0.0002, + "loss": 1.382, + "step": 8678 + }, + { + "epoch": 2.736991485335856, + "grad_norm": 1.4921875, + "learning_rate": 0.0002, + "loss": 1.3843, + "step": 8679 + }, + { + "epoch": 2.737306843267108, + "grad_norm": 1.1875, + "learning_rate": 0.0002, + "loss": 1.5811, + "step": 8680 + }, + { + "epoch": 2.73762220119836, + "grad_norm": 1.1328125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 8681 + }, + { + "epoch": 2.737937559129612, + "grad_norm": 1.28125, + "learning_rate": 0.0002, + "loss": 1.8188, + "step": 8682 + }, + { + "epoch": 2.738252917060864, + "grad_norm": 1.296875, + "learning_rate": 0.0002, + "loss": 1.7543, + "step": 8683 + }, + { + "epoch": 2.738568274992116, + "grad_norm": 1.328125, + "learning_rate": 0.0002, + "loss": 1.4823, + "step": 8684 + }, + { + "epoch": 2.738883632923368, + "grad_norm": 2.796875, + "learning_rate": 0.0002, + "loss": 1.5844, + "step": 8685 + }, + { + "epoch": 2.73919899085462, + "grad_norm": 2.390625, + "learning_rate": 0.0002, + "loss": 1.9246, + "step": 8686 + }, + { + "epoch": 2.739514348785872, + "grad_norm": 1.09375, + "learning_rate": 0.0002, + "loss": 1.4598, + "step": 8687 + }, + { + "epoch": 2.739829706717124, + "grad_norm": 1.1640625, + "learning_rate": 0.0002, + "loss": 1.6225, + "step": 8688 + }, + { + "epoch": 2.740145064648376, + "grad_norm": 1.75, + "learning_rate": 0.0002, + "loss": 1.7871, + "step": 8689 + }, + { + "epoch": 2.740460422579628, + "grad_norm": 1.109375, + "learning_rate": 0.0002, + "loss": 1.3504, + "step": 8690 + }, + { + "epoch": 2.74077578051088, + "grad_norm": 1.1328125, + "learning_rate": 0.0002, + "loss": 1.523, + "step": 8691 + }, + { + "epoch": 2.741091138442132, + "grad_norm": 1.1953125, + "learning_rate": 0.0002, + "loss": 1.6453, + "step": 8692 + }, + { + "epoch": 2.7414064963733837, + "grad_norm": 0.2138671875, + "learning_rate": 0.0002, + "loss": 0.8973, + "step": 8693 + }, + { + "epoch": 2.741721854304636, + "grad_norm": 0.248046875, + "learning_rate": 0.0002, + "loss": 0.9473, + "step": 8694 + }, + { + "epoch": 2.742037212235888, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 0.9248, + "step": 8695 + }, + { + "epoch": 2.7423525701671396, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 0.973, + "step": 8696 + }, + { + "epoch": 2.742667928098392, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 0.9343, + "step": 8697 + }, + { + "epoch": 2.7429832860296437, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 0.9001, + "step": 8698 + }, + { + "epoch": 2.7432986439608955, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 0.9071, + "step": 8699 + }, + { + "epoch": 2.7436140018921478, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 1.0517, + "step": 8700 + }, + { + "epoch": 2.7439293598233996, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 1.1102, + "step": 8701 + }, + { + "epoch": 2.7442447177546514, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 0.9449, + "step": 8702 + }, + { + "epoch": 2.7445600756859037, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 1.0065, + "step": 8703 + }, + { + "epoch": 2.7448754336171555, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 0.8382, + "step": 8704 + }, + { + "epoch": 2.7451907915484073, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 1.0436, + "step": 8705 + }, + { + "epoch": 2.7455061494796595, + "grad_norm": 0.3515625, + "learning_rate": 0.0002, + "loss": 0.7826, + "step": 8706 + }, + { + "epoch": 2.7458215074109114, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 0.8978, + "step": 8707 + }, + { + "epoch": 2.746136865342163, + "grad_norm": 0.392578125, + "learning_rate": 0.0002, + "loss": 0.991, + "step": 8708 + }, + { + "epoch": 2.7464522232734154, + "grad_norm": 0.408203125, + "learning_rate": 0.0002, + "loss": 0.9495, + "step": 8709 + }, + { + "epoch": 2.7467675812046672, + "grad_norm": 0.41015625, + "learning_rate": 0.0002, + "loss": 0.8335, + "step": 8710 + }, + { + "epoch": 2.747082939135919, + "grad_norm": 0.412109375, + "learning_rate": 0.0002, + "loss": 0.9304, + "step": 8711 + }, + { + "epoch": 2.7473982970671713, + "grad_norm": 0.447265625, + "learning_rate": 0.0002, + "loss": 0.9322, + "step": 8712 + }, + { + "epoch": 2.747713654998423, + "grad_norm": 0.439453125, + "learning_rate": 0.0002, + "loss": 0.9498, + "step": 8713 + }, + { + "epoch": 2.748029012929675, + "grad_norm": 0.49609375, + "learning_rate": 0.0002, + "loss": 0.8936, + "step": 8714 + }, + { + "epoch": 2.748344370860927, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 1.0242, + "step": 8715 + }, + { + "epoch": 2.748659728792179, + "grad_norm": 0.52734375, + "learning_rate": 0.0002, + "loss": 0.8631, + "step": 8716 + }, + { + "epoch": 2.748975086723431, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 0.8661, + "step": 8717 + }, + { + "epoch": 2.749290444654683, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 0.8793, + "step": 8718 + }, + { + "epoch": 2.749605802585935, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 1.0127, + "step": 8719 + }, + { + "epoch": 2.749921160517187, + "grad_norm": 0.72265625, + "learning_rate": 0.0002, + "loss": 0.9742, + "step": 8720 + }, + { + "epoch": 2.750236518448439, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 1.1901, + "step": 8721 + }, + { + "epoch": 2.750551876379691, + "grad_norm": 0.75, + "learning_rate": 0.0002, + "loss": 1.1098, + "step": 8722 + }, + { + "epoch": 2.750867234310943, + "grad_norm": 0.90234375, + "learning_rate": 0.0002, + "loss": 1.0752, + "step": 8723 + }, + { + "epoch": 2.751182592242195, + "grad_norm": 0.94921875, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 8724 + }, + { + "epoch": 2.751497950173447, + "grad_norm": 0.87109375, + "learning_rate": 0.0002, + "loss": 1.293, + "step": 8725 + }, + { + "epoch": 2.751813308104699, + "grad_norm": 1.09375, + "learning_rate": 0.0002, + "loss": 1.4119, + "step": 8726 + }, + { + "epoch": 2.7521286660359507, + "grad_norm": 0.9609375, + "learning_rate": 0.0002, + "loss": 1.2217, + "step": 8727 + }, + { + "epoch": 2.752444023967203, + "grad_norm": 1.15625, + "learning_rate": 0.0002, + "loss": 1.2694, + "step": 8728 + }, + { + "epoch": 2.752759381898455, + "grad_norm": 0.94140625, + "learning_rate": 0.0002, + "loss": 1.2584, + "step": 8729 + }, + { + "epoch": 2.7530747398297066, + "grad_norm": 1.1796875, + "learning_rate": 0.0002, + "loss": 1.2808, + "step": 8730 + }, + { + "epoch": 2.753390097760959, + "grad_norm": 1.296875, + "learning_rate": 0.0002, + "loss": 1.5338, + "step": 8731 + }, + { + "epoch": 2.7537054556922107, + "grad_norm": 1.2890625, + "learning_rate": 0.0002, + "loss": 1.2695, + "step": 8732 + }, + { + "epoch": 2.7540208136234625, + "grad_norm": 1.484375, + "learning_rate": 0.0002, + "loss": 1.6996, + "step": 8733 + }, + { + "epoch": 2.7543361715547148, + "grad_norm": 1.296875, + "learning_rate": 0.0002, + "loss": 1.5579, + "step": 8734 + }, + { + "epoch": 2.7546515294859666, + "grad_norm": 1.515625, + "learning_rate": 0.0002, + "loss": 1.8452, + "step": 8735 + }, + { + "epoch": 2.7549668874172184, + "grad_norm": 1.21875, + "learning_rate": 0.0002, + "loss": 1.4171, + "step": 8736 + }, + { + "epoch": 2.7552822453484707, + "grad_norm": 2.546875, + "learning_rate": 0.0002, + "loss": 1.7525, + "step": 8737 + }, + { + "epoch": 2.7555976032797225, + "grad_norm": 1.21875, + "learning_rate": 0.0002, + "loss": 1.718, + "step": 8738 + }, + { + "epoch": 2.7559129612109743, + "grad_norm": 1.3046875, + "learning_rate": 0.0002, + "loss": 1.5285, + "step": 8739 + }, + { + "epoch": 2.7562283191422265, + "grad_norm": 0.98046875, + "learning_rate": 0.0002, + "loss": 1.2927, + "step": 8740 + }, + { + "epoch": 2.7565436770734784, + "grad_norm": 1.265625, + "learning_rate": 0.0002, + "loss": 1.7116, + "step": 8741 + }, + { + "epoch": 2.75685903500473, + "grad_norm": 1.390625, + "learning_rate": 0.0002, + "loss": 1.7959, + "step": 8742 + }, + { + "epoch": 2.7571743929359824, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 0.944, + "step": 8743 + }, + { + "epoch": 2.7574897508672342, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 0.9684, + "step": 8744 + }, + { + "epoch": 2.757805108798486, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 0.9603, + "step": 8745 + }, + { + "epoch": 2.7581204667297383, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 1.0109, + "step": 8746 + }, + { + "epoch": 2.75843582466099, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 0.9558, + "step": 8747 + }, + { + "epoch": 2.758751182592242, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.9348, + "step": 8748 + }, + { + "epoch": 2.759066540523494, + "grad_norm": 0.3828125, + "learning_rate": 0.0002, + "loss": 0.9527, + "step": 8749 + }, + { + "epoch": 2.759381898454746, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 0.9458, + "step": 8750 + }, + { + "epoch": 2.759697256385998, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 0.9037, + "step": 8751 + }, + { + "epoch": 2.76001261431725, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 0.8421, + "step": 8752 + }, + { + "epoch": 2.760327972248502, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 0.971, + "step": 8753 + }, + { + "epoch": 2.760643330179754, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 0.843, + "step": 8754 + }, + { + "epoch": 2.760958688111006, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 1.0411, + "step": 8755 + }, + { + "epoch": 2.7612740460422582, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 0.9402, + "step": 8756 + }, + { + "epoch": 2.76158940397351, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 1.0099, + "step": 8757 + }, + { + "epoch": 2.761904761904762, + "grad_norm": 0.4375, + "learning_rate": 0.0002, + "loss": 1.0038, + "step": 8758 + }, + { + "epoch": 2.762220119836014, + "grad_norm": 0.40234375, + "learning_rate": 0.0002, + "loss": 0.9638, + "step": 8759 + }, + { + "epoch": 2.762535477767266, + "grad_norm": 0.330078125, + "learning_rate": 0.0002, + "loss": 0.8553, + "step": 8760 + }, + { + "epoch": 2.7628508356985177, + "grad_norm": 0.40625, + "learning_rate": 0.0002, + "loss": 0.976, + "step": 8761 + }, + { + "epoch": 2.76316619362977, + "grad_norm": 0.4140625, + "learning_rate": 0.0002, + "loss": 0.8996, + "step": 8762 + }, + { + "epoch": 2.763481551561022, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 0.7729, + "step": 8763 + }, + { + "epoch": 2.7637969094922736, + "grad_norm": 0.43359375, + "learning_rate": 0.0002, + "loss": 0.8533, + "step": 8764 + }, + { + "epoch": 2.764112267423526, + "grad_norm": 0.42578125, + "learning_rate": 0.0002, + "loss": 0.8724, + "step": 8765 + }, + { + "epoch": 2.7644276253547777, + "grad_norm": 0.466796875, + "learning_rate": 0.0002, + "loss": 0.9347, + "step": 8766 + }, + { + "epoch": 2.7647429832860295, + "grad_norm": 0.55859375, + "learning_rate": 0.0002, + "loss": 0.9078, + "step": 8767 + }, + { + "epoch": 2.765058341217282, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 0.8748, + "step": 8768 + }, + { + "epoch": 2.7653736991485336, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 1.0211, + "step": 8769 + }, + { + "epoch": 2.7656890570797854, + "grad_norm": 0.640625, + "learning_rate": 0.0002, + "loss": 0.9573, + "step": 8770 + }, + { + "epoch": 2.7660044150110377, + "grad_norm": 0.58203125, + "learning_rate": 0.0002, + "loss": 0.9429, + "step": 8771 + }, + { + "epoch": 2.7663197729422895, + "grad_norm": 0.81640625, + "learning_rate": 0.0002, + "loss": 1.0252, + "step": 8772 + }, + { + "epoch": 2.7666351308735413, + "grad_norm": 0.80859375, + "learning_rate": 0.0002, + "loss": 1.1543, + "step": 8773 + }, + { + "epoch": 2.7669504888047936, + "grad_norm": 0.7109375, + "learning_rate": 0.0002, + "loss": 1.0539, + "step": 8774 + }, + { + "epoch": 2.7672658467360454, + "grad_norm": 0.83984375, + "learning_rate": 0.0002, + "loss": 1.1431, + "step": 8775 + }, + { + "epoch": 2.767581204667297, + "grad_norm": 0.75, + "learning_rate": 0.0002, + "loss": 0.9742, + "step": 8776 + }, + { + "epoch": 2.7678965625985494, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 1.2904, + "step": 8777 + }, + { + "epoch": 2.7682119205298013, + "grad_norm": 0.9453125, + "learning_rate": 0.0002, + "loss": 1.2969, + "step": 8778 + }, + { + "epoch": 2.768527278461053, + "grad_norm": 1.125, + "learning_rate": 0.0002, + "loss": 1.2958, + "step": 8779 + }, + { + "epoch": 2.7688426363923053, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.3575, + "step": 8780 + }, + { + "epoch": 2.769157994323557, + "grad_norm": 1.0, + "learning_rate": 0.0002, + "loss": 1.2178, + "step": 8781 + }, + { + "epoch": 2.769473352254809, + "grad_norm": 1.1328125, + "learning_rate": 0.0002, + "loss": 1.7033, + "step": 8782 + }, + { + "epoch": 2.769788710186061, + "grad_norm": 1.1953125, + "learning_rate": 0.0002, + "loss": 1.4154, + "step": 8783 + }, + { + "epoch": 2.770104068117313, + "grad_norm": 1.1875, + "learning_rate": 0.0002, + "loss": 1.669, + "step": 8784 + }, + { + "epoch": 2.7704194260485653, + "grad_norm": 0.9921875, + "learning_rate": 0.0002, + "loss": 1.4151, + "step": 8785 + }, + { + "epoch": 2.770734783979817, + "grad_norm": 1.6328125, + "learning_rate": 0.0002, + "loss": 1.5494, + "step": 8786 + }, + { + "epoch": 2.771050141911069, + "grad_norm": 1.28125, + "learning_rate": 0.0002, + "loss": 1.6466, + "step": 8787 + }, + { + "epoch": 2.771365499842321, + "grad_norm": 1.328125, + "learning_rate": 0.0002, + "loss": 1.675, + "step": 8788 + }, + { + "epoch": 2.771680857773573, + "grad_norm": 2.140625, + "learning_rate": 0.0002, + "loss": 1.6518, + "step": 8789 + }, + { + "epoch": 2.771680857773573, + "eval_loss": 1.8249502182006836, + "eval_runtime": 148.9389, + "eval_samples_per_second": 6.714, + "eval_steps_per_second": 6.714, + "step": 8789 + }, + { + "epoch": 2.771680857773573, + "mmlu_eval_accuracy": 0.5892045089801169, + "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, + "mmlu_eval_accuracy_anatomy": 0.6428571428571429, + "mmlu_eval_accuracy_astronomy": 0.625, + "mmlu_eval_accuracy_business_ethics": 0.8181818181818182, + "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034, + "mmlu_eval_accuracy_college_biology": 0.8125, + "mmlu_eval_accuracy_college_chemistry": 0.375, + "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727, + "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365, + "mmlu_eval_accuracy_college_medicine": 0.4090909090909091, + "mmlu_eval_accuracy_college_physics": 0.45454545454545453, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384, + "mmlu_eval_accuracy_econometrics": 0.5, + "mmlu_eval_accuracy_electrical_engineering": 0.5625, + "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244, + "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, + "mmlu_eval_accuracy_global_facts": 0.2, + "mmlu_eval_accuracy_high_school_biology": 0.59375, + "mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.627906976744186, + "mmlu_eval_accuracy_high_school_mathematics": 0.3793103448275862, + "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769, + "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354, + "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667, + "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087, + "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, + "mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693, + "mmlu_eval_accuracy_human_aging": 0.6521739130434783, + "mmlu_eval_accuracy_human_sexuality": 0.5, + "mmlu_eval_accuracy_international_law": 0.9230769230769231, + "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, + "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222, + "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.68, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7209302325581395, + "mmlu_eval_accuracy_moral_disputes": 0.6052631578947368, + "mmlu_eval_accuracy_moral_scenarios": 0.26, + "mmlu_eval_accuracy_nutrition": 0.7575757575757576, + "mmlu_eval_accuracy_philosophy": 0.5588235294117647, + "mmlu_eval_accuracy_prehistory": 0.5714285714285714, + "mmlu_eval_accuracy_professional_accounting": 0.5161290322580645, + "mmlu_eval_accuracy_professional_law": 0.4176470588235294, + "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516, + "mmlu_eval_accuracy_professional_psychology": 0.5942028985507246, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7037037037037037, + "mmlu_eval_accuracy_sociology": 0.9090909090909091, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.5, + "mmlu_eval_accuracy_world_religions": 0.8421052631578947, + "mmlu_loss": 6.025850145899338, + "step": 8789 + }, + { + "epoch": 2.7719962157048252, + "grad_norm": 1.265625, + "learning_rate": 0.0002, + "loss": 1.4231, + "step": 8790 + }, + { + "epoch": 2.772311573636077, + "grad_norm": 1.0859375, + "learning_rate": 0.0002, + "loss": 1.3826, + "step": 8791 + }, + { + "epoch": 2.772626931567329, + "grad_norm": 1.421875, + "learning_rate": 0.0002, + "loss": 2.0412, + "step": 8792 + }, + { + "epoch": 2.772942289498581, + "grad_norm": 0.23828125, + "learning_rate": 0.0002, + "loss": 0.9942, + "step": 8793 + }, + { + "epoch": 2.773257647429833, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.0898, + "step": 8794 + }, + { + "epoch": 2.7735730053610848, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.0221, + "step": 8795 + }, + { + "epoch": 2.773888363292337, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 0.982, + "step": 8796 + }, + { + "epoch": 2.774203721223589, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 0.8801, + "step": 8797 + }, + { + "epoch": 2.7745190791548406, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 0.9146, + "step": 8798 + }, + { + "epoch": 2.774834437086093, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 0.903, + "step": 8799 + }, + { + "epoch": 2.7751497950173447, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 1.0185, + "step": 8800 + }, + { + "epoch": 2.7754651529485965, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 0.8438, + "step": 8801 + }, + { + "epoch": 2.775780510879849, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 0.9793, + "step": 8802 + }, + { + "epoch": 2.7760958688111006, + "grad_norm": 0.353515625, + "learning_rate": 0.0002, + "loss": 0.871, + "step": 8803 + }, + { + "epoch": 2.7764112267423524, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 0.894, + "step": 8804 + }, + { + "epoch": 2.7767265846736047, + "grad_norm": 0.349609375, + "learning_rate": 0.0002, + "loss": 0.9639, + "step": 8805 + }, + { + "epoch": 2.7770419426048565, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 0.9022, + "step": 8806 + }, + { + "epoch": 2.7773573005361083, + "grad_norm": 0.34765625, + "learning_rate": 0.0002, + "loss": 0.9797, + "step": 8807 + }, + { + "epoch": 2.7776726584673606, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 0.9001, + "step": 8808 + }, + { + "epoch": 2.7779880163986124, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 0.8371, + "step": 8809 + }, + { + "epoch": 2.778303374329864, + "grad_norm": 0.408203125, + "learning_rate": 0.0002, + "loss": 0.9697, + "step": 8810 + }, + { + "epoch": 2.7786187322611164, + "grad_norm": 0.41796875, + "learning_rate": 0.0002, + "loss": 0.8763, + "step": 8811 + }, + { + "epoch": 2.7789340901923683, + "grad_norm": 0.43359375, + "learning_rate": 0.0002, + "loss": 0.9389, + "step": 8812 + }, + { + "epoch": 2.77924944812362, + "grad_norm": 0.515625, + "learning_rate": 0.0002, + "loss": 1.0082, + "step": 8813 + }, + { + "epoch": 2.7795648060548723, + "grad_norm": 0.498046875, + "learning_rate": 0.0002, + "loss": 0.8149, + "step": 8814 + }, + { + "epoch": 2.779880163986124, + "grad_norm": 0.50390625, + "learning_rate": 0.0002, + "loss": 0.8315, + "step": 8815 + }, + { + "epoch": 2.780195521917376, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 0.9257, + "step": 8816 + }, + { + "epoch": 2.780510879848628, + "grad_norm": 0.50390625, + "learning_rate": 0.0002, + "loss": 0.7886, + "step": 8817 + }, + { + "epoch": 2.78082623777988, + "grad_norm": 0.466796875, + "learning_rate": 0.0002, + "loss": 0.806, + "step": 8818 + }, + { + "epoch": 2.7811415957111323, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 0.8563, + "step": 8819 + }, + { + "epoch": 2.781456953642384, + "grad_norm": 0.71484375, + "learning_rate": 0.0002, + "loss": 0.9097, + "step": 8820 + }, + { + "epoch": 2.781772311573636, + "grad_norm": 0.77734375, + "learning_rate": 0.0002, + "loss": 0.8942, + "step": 8821 + }, + { + "epoch": 2.782087669504888, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 1.0195, + "step": 8822 + }, + { + "epoch": 2.78240302743614, + "grad_norm": 0.82421875, + "learning_rate": 0.0002, + "loss": 1.0928, + "step": 8823 + }, + { + "epoch": 2.7827183853673922, + "grad_norm": 0.76953125, + "learning_rate": 0.0002, + "loss": 1.148, + "step": 8824 + }, + { + "epoch": 2.783033743298644, + "grad_norm": 0.8046875, + "learning_rate": 0.0002, + "loss": 1.0947, + "step": 8825 + }, + { + "epoch": 2.783349101229896, + "grad_norm": 0.91015625, + "learning_rate": 0.0002, + "loss": 1.1035, + "step": 8826 + }, + { + "epoch": 2.783664459161148, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 1.1482, + "step": 8827 + }, + { + "epoch": 2.7839798170924, + "grad_norm": 1.3046875, + "learning_rate": 0.0002, + "loss": 1.2913, + "step": 8828 + }, + { + "epoch": 2.7842951750236518, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 1.3554, + "step": 8829 + }, + { + "epoch": 2.784610532954904, + "grad_norm": 1.234375, + "learning_rate": 0.0002, + "loss": 1.3804, + "step": 8830 + }, + { + "epoch": 2.784925890886156, + "grad_norm": 1.625, + "learning_rate": 0.0002, + "loss": 1.8033, + "step": 8831 + }, + { + "epoch": 2.7852412488174076, + "grad_norm": 1.3125, + "learning_rate": 0.0002, + "loss": 1.5417, + "step": 8832 + }, + { + "epoch": 2.78555660674866, + "grad_norm": 1.3359375, + "learning_rate": 0.0002, + "loss": 1.7324, + "step": 8833 + }, + { + "epoch": 2.7858719646799117, + "grad_norm": 1.3359375, + "learning_rate": 0.0002, + "loss": 1.9423, + "step": 8834 + }, + { + "epoch": 2.7861873226111635, + "grad_norm": 1.28125, + "learning_rate": 0.0002, + "loss": 1.6765, + "step": 8835 + }, + { + "epoch": 2.786502680542416, + "grad_norm": 1.3984375, + "learning_rate": 0.0002, + "loss": 1.5281, + "step": 8836 + }, + { + "epoch": 2.7868180384736676, + "grad_norm": 1.1796875, + "learning_rate": 0.0002, + "loss": 1.5967, + "step": 8837 + }, + { + "epoch": 2.7871333964049194, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 1.7768, + "step": 8838 + }, + { + "epoch": 2.7874487543361717, + "grad_norm": 0.95703125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 8839 + }, + { + "epoch": 2.7877641122674235, + "grad_norm": 1.4765625, + "learning_rate": 0.0002, + "loss": 1.9214, + "step": 8840 + }, + { + "epoch": 2.7880794701986753, + "grad_norm": 1.1875, + "learning_rate": 0.0002, + "loss": 1.7165, + "step": 8841 + }, + { + "epoch": 2.7883948281299276, + "grad_norm": 1.25, + "learning_rate": 0.0002, + "loss": 1.7883, + "step": 8842 + }, + { + "epoch": 2.7887101860611794, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 0.9813, + "step": 8843 + }, + { + "epoch": 2.789025543992431, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 1.0195, + "step": 8844 + }, + { + "epoch": 2.7893409019236834, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.003, + "step": 8845 + }, + { + "epoch": 2.7896562598549353, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.0125, + "step": 8846 + }, + { + "epoch": 2.789971617786187, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.0479, + "step": 8847 + }, + { + "epoch": 2.7902869757174393, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 0.934, + "step": 8848 + }, + { + "epoch": 2.790602333648691, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 1.0065, + "step": 8849 + }, + { + "epoch": 2.790917691579943, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 0.9795, + "step": 8850 + }, + { + "epoch": 2.791233049511195, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 0.9529, + "step": 8851 + }, + { + "epoch": 2.791548407442447, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 0.9138, + "step": 8852 + }, + { + "epoch": 2.7918637653736993, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 1.0316, + "step": 8853 + }, + { + "epoch": 2.792179123304951, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 0.9112, + "step": 8854 + }, + { + "epoch": 2.7924944812362034, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 0.9011, + "step": 8855 + }, + { + "epoch": 2.792809839167455, + "grad_norm": 0.37109375, + "learning_rate": 0.0002, + "loss": 0.8755, + "step": 8856 + }, + { + "epoch": 2.793125197098707, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 0.9168, + "step": 8857 + }, + { + "epoch": 2.7934405550299592, + "grad_norm": 0.404296875, + "learning_rate": 0.0002, + "loss": 0.9835, + "step": 8858 + }, + { + "epoch": 2.793755912961211, + "grad_norm": 0.38671875, + "learning_rate": 0.0002, + "loss": 0.8466, + "step": 8859 + }, + { + "epoch": 2.794071270892463, + "grad_norm": 0.41796875, + "learning_rate": 0.0002, + "loss": 0.8389, + "step": 8860 + }, + { + "epoch": 2.794386628823715, + "grad_norm": 0.515625, + "learning_rate": 0.0002, + "loss": 1.0245, + "step": 8861 + }, + { + "epoch": 2.794701986754967, + "grad_norm": 0.455078125, + "learning_rate": 0.0002, + "loss": 0.9438, + "step": 8862 + }, + { + "epoch": 2.7950173446862188, + "grad_norm": 0.40234375, + "learning_rate": 0.0002, + "loss": 0.9166, + "step": 8863 + }, + { + "epoch": 2.795332702617471, + "grad_norm": 0.55859375, + "learning_rate": 0.0002, + "loss": 0.9201, + "step": 8864 + }, + { + "epoch": 2.795648060548723, + "grad_norm": 0.546875, + "learning_rate": 0.0002, + "loss": 0.9023, + "step": 8865 + }, + { + "epoch": 2.7959634184799746, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 0.8653, + "step": 8866 + }, + { + "epoch": 2.796278776411227, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 1.0219, + "step": 8867 + }, + { + "epoch": 2.7965941343424787, + "grad_norm": 0.58984375, + "learning_rate": 0.0002, + "loss": 0.9558, + "step": 8868 + }, + { + "epoch": 2.7969094922737305, + "grad_norm": 0.640625, + "learning_rate": 0.0002, + "loss": 0.9165, + "step": 8869 + }, + { + "epoch": 2.797224850204983, + "grad_norm": 0.8203125, + "learning_rate": 0.0002, + "loss": 1.0774, + "step": 8870 + }, + { + "epoch": 2.7975402081362346, + "grad_norm": 0.83203125, + "learning_rate": 0.0002, + "loss": 1.0458, + "step": 8871 + }, + { + "epoch": 2.7978555660674864, + "grad_norm": 0.6953125, + "learning_rate": 0.0002, + "loss": 1.0387, + "step": 8872 + }, + { + "epoch": 2.7981709239987387, + "grad_norm": 0.99609375, + "learning_rate": 0.0002, + "loss": 1.0525, + "step": 8873 + }, + { + "epoch": 2.7984862819299905, + "grad_norm": 0.984375, + "learning_rate": 0.0002, + "loss": 1.2165, + "step": 8874 + }, + { + "epoch": 2.7988016398612423, + "grad_norm": 1.28125, + "learning_rate": 0.0002, + "loss": 1.229, + "step": 8875 + }, + { + "epoch": 2.7991169977924946, + "grad_norm": 0.9765625, + "learning_rate": 0.0002, + "loss": 1.4047, + "step": 8876 + }, + { + "epoch": 2.7994323557237464, + "grad_norm": 0.97265625, + "learning_rate": 0.0002, + "loss": 1.4698, + "step": 8877 + }, + { + "epoch": 2.799747713654998, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 1.3817, + "step": 8878 + }, + { + "epoch": 2.8000630715862505, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 8879 + }, + { + "epoch": 2.8003784295175023, + "grad_norm": 1.078125, + "learning_rate": 0.0002, + "loss": 1.3981, + "step": 8880 + }, + { + "epoch": 2.800693787448754, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 1.4737, + "step": 8881 + }, + { + "epoch": 2.8010091453800063, + "grad_norm": 1.4296875, + "learning_rate": 0.0002, + "loss": 1.5786, + "step": 8882 + }, + { + "epoch": 2.801324503311258, + "grad_norm": 1.1875, + "learning_rate": 0.0002, + "loss": 1.8231, + "step": 8883 + }, + { + "epoch": 2.8016398612425104, + "grad_norm": 1.296875, + "learning_rate": 0.0002, + "loss": 1.6189, + "step": 8884 + }, + { + "epoch": 2.8019552191737622, + "grad_norm": 1.484375, + "learning_rate": 0.0002, + "loss": 1.9084, + "step": 8885 + }, + { + "epoch": 2.802270577105014, + "grad_norm": 1.5078125, + "learning_rate": 0.0002, + "loss": 1.757, + "step": 8886 + }, + { + "epoch": 2.8025859350362663, + "grad_norm": 1.21875, + "learning_rate": 0.0002, + "loss": 1.5821, + "step": 8887 + }, + { + "epoch": 2.802901292967518, + "grad_norm": 1.2734375, + "learning_rate": 0.0002, + "loss": 1.9074, + "step": 8888 + }, + { + "epoch": 2.8032166508987704, + "grad_norm": 1.2421875, + "learning_rate": 0.0002, + "loss": 1.4676, + "step": 8889 + }, + { + "epoch": 2.803532008830022, + "grad_norm": 1.3828125, + "learning_rate": 0.0002, + "loss": 1.4174, + "step": 8890 + }, + { + "epoch": 2.803847366761274, + "grad_norm": 1.46875, + "learning_rate": 0.0002, + "loss": 1.3763, + "step": 8891 + }, + { + "epoch": 2.8041627246925263, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 1.8499, + "step": 8892 + }, + { + "epoch": 2.804478082623778, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 1.049, + "step": 8893 + }, + { + "epoch": 2.80479344055503, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 0.9218, + "step": 8894 + }, + { + "epoch": 2.805108798486282, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 0.8908, + "step": 8895 + }, + { + "epoch": 2.805424156417534, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 0.9456, + "step": 8896 + }, + { + "epoch": 2.8057395143487858, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.9057, + "step": 8897 + }, + { + "epoch": 2.806054872280038, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.009, + "step": 8898 + }, + { + "epoch": 2.80637023021129, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 0.8627, + "step": 8899 + }, + { + "epoch": 2.8066855881425417, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 0.8885, + "step": 8900 + }, + { + "epoch": 2.807000946073794, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 0.9481, + "step": 8901 + }, + { + "epoch": 2.8073163040050457, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 0.9683, + "step": 8902 + }, + { + "epoch": 2.8076316619362975, + "grad_norm": 0.353515625, + "learning_rate": 0.0002, + "loss": 0.9072, + "step": 8903 + }, + { + "epoch": 2.80794701986755, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 0.8369, + "step": 8904 + }, + { + "epoch": 2.8082623777988016, + "grad_norm": 0.3828125, + "learning_rate": 0.0002, + "loss": 1.0637, + "step": 8905 + }, + { + "epoch": 2.8085777357300534, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 0.9984, + "step": 8906 + }, + { + "epoch": 2.8088930936613057, + "grad_norm": 0.37109375, + "learning_rate": 0.0002, + "loss": 0.9434, + "step": 8907 + }, + { + "epoch": 2.8092084515925575, + "grad_norm": 0.3671875, + "learning_rate": 0.0002, + "loss": 0.9827, + "step": 8908 + }, + { + "epoch": 2.8095238095238093, + "grad_norm": 0.396484375, + "learning_rate": 0.0002, + "loss": 0.9374, + "step": 8909 + }, + { + "epoch": 2.8098391674550616, + "grad_norm": 0.384765625, + "learning_rate": 0.0002, + "loss": 0.9646, + "step": 8910 + }, + { + "epoch": 2.8101545253863134, + "grad_norm": 0.435546875, + "learning_rate": 0.0002, + "loss": 1.0249, + "step": 8911 + }, + { + "epoch": 2.810469883317565, + "grad_norm": 0.421875, + "learning_rate": 0.0002, + "loss": 1.0014, + "step": 8912 + }, + { + "epoch": 2.8107852412488175, + "grad_norm": 0.412109375, + "learning_rate": 0.0002, + "loss": 0.8639, + "step": 8913 + }, + { + "epoch": 2.8111005991800693, + "grad_norm": 0.44921875, + "learning_rate": 0.0002, + "loss": 1.0353, + "step": 8914 + }, + { + "epoch": 2.811415957111321, + "grad_norm": 0.49609375, + "learning_rate": 0.0002, + "loss": 0.977, + "step": 8915 + }, + { + "epoch": 2.8117313150425733, + "grad_norm": 0.48828125, + "learning_rate": 0.0002, + "loss": 0.9289, + "step": 8916 + }, + { + "epoch": 2.812046672973825, + "grad_norm": 0.49609375, + "learning_rate": 0.0002, + "loss": 0.9207, + "step": 8917 + }, + { + "epoch": 2.8123620309050774, + "grad_norm": 0.53515625, + "learning_rate": 0.0002, + "loss": 0.9475, + "step": 8918 + }, + { + "epoch": 2.8126773888363292, + "grad_norm": 0.5703125, + "learning_rate": 0.0002, + "loss": 0.8858, + "step": 8919 + }, + { + "epoch": 2.812992746767581, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 0.9333, + "step": 8920 + }, + { + "epoch": 2.8133081046988333, + "grad_norm": 0.6640625, + "learning_rate": 0.0002, + "loss": 0.9674, + "step": 8921 + }, + { + "epoch": 2.813623462630085, + "grad_norm": 0.65625, + "learning_rate": 0.0002, + "loss": 0.8812, + "step": 8922 + }, + { + "epoch": 2.8139388205613374, + "grad_norm": 0.71484375, + "learning_rate": 0.0002, + "loss": 0.9136, + "step": 8923 + }, + { + "epoch": 2.814254178492589, + "grad_norm": 0.74609375, + "learning_rate": 0.0002, + "loss": 1.129, + "step": 8924 + }, + { + "epoch": 2.814569536423841, + "grad_norm": 0.78125, + "learning_rate": 0.0002, + "loss": 1.2462, + "step": 8925 + }, + { + "epoch": 2.8148848943550933, + "grad_norm": 0.90234375, + "learning_rate": 0.0002, + "loss": 1.2528, + "step": 8926 + }, + { + "epoch": 2.815200252286345, + "grad_norm": 0.87890625, + "learning_rate": 0.0002, + "loss": 1.1841, + "step": 8927 + }, + { + "epoch": 2.815515610217597, + "grad_norm": 1.0, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 8928 + }, + { + "epoch": 2.815830968148849, + "grad_norm": 1.1875, + "learning_rate": 0.0002, + "loss": 1.3827, + "step": 8929 + }, + { + "epoch": 2.816146326080101, + "grad_norm": 1.1875, + "learning_rate": 0.0002, + "loss": 1.6567, + "step": 8930 + }, + { + "epoch": 2.8164616840113528, + "grad_norm": 1.1640625, + "learning_rate": 0.0002, + "loss": 1.4016, + "step": 8931 + }, + { + "epoch": 2.816777041942605, + "grad_norm": 1.2890625, + "learning_rate": 0.0002, + "loss": 1.5119, + "step": 8932 + }, + { + "epoch": 2.817092399873857, + "grad_norm": 1.3359375, + "learning_rate": 0.0002, + "loss": 1.5352, + "step": 8933 + }, + { + "epoch": 2.8174077578051087, + "grad_norm": 1.0546875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 8934 + }, + { + "epoch": 2.817723115736361, + "grad_norm": 1.71875, + "learning_rate": 0.0002, + "loss": 1.7965, + "step": 8935 + }, + { + "epoch": 2.8180384736676127, + "grad_norm": 1.3046875, + "learning_rate": 0.0002, + "loss": 1.9269, + "step": 8936 + }, + { + "epoch": 2.8183538315988645, + "grad_norm": 1.8046875, + "learning_rate": 0.0002, + "loss": 1.8032, + "step": 8937 + }, + { + "epoch": 2.818669189530117, + "grad_norm": 1.875, + "learning_rate": 0.0002, + "loss": 2.1296, + "step": 8938 + }, + { + "epoch": 2.8189845474613686, + "grad_norm": 1.328125, + "learning_rate": 0.0002, + "loss": 1.3494, + "step": 8939 + }, + { + "epoch": 2.8192999053926204, + "grad_norm": 1.2265625, + "learning_rate": 0.0002, + "loss": 1.2776, + "step": 8940 + }, + { + "epoch": 2.8196152633238727, + "grad_norm": 1.328125, + "learning_rate": 0.0002, + "loss": 1.4366, + "step": 8941 + }, + { + "epoch": 2.8199306212551245, + "grad_norm": 1.4765625, + "learning_rate": 0.0002, + "loss": 1.4985, + "step": 8942 + }, + { + "epoch": 2.8202459791863763, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 0.9794, + "step": 8943 + }, + { + "epoch": 2.8205613371176286, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 1.0461, + "step": 8944 + }, + { + "epoch": 2.8208766950488804, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 0.961, + "step": 8945 + }, + { + "epoch": 2.821192052980132, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 0.8348, + "step": 8946 + }, + { + "epoch": 2.8215074109113845, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 0.9051, + "step": 8947 + }, + { + "epoch": 2.8218227688426363, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 0.9972, + "step": 8948 + }, + { + "epoch": 2.822138126773888, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 1.0129, + "step": 8949 + }, + { + "epoch": 2.8224534847051403, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 0.903, + "step": 8950 + }, + { + "epoch": 2.822768842636392, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.943, + "step": 8951 + }, + { + "epoch": 2.8230842005676444, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 0.9683, + "step": 8952 + }, + { + "epoch": 2.8233995584988962, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 0.98, + "step": 8953 + }, + { + "epoch": 2.8237149164301485, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 0.8754, + "step": 8954 + }, + { + "epoch": 2.8240302743614003, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 0.8487, + "step": 8955 + }, + { + "epoch": 2.824345632292652, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 0.831, + "step": 8956 + }, + { + "epoch": 2.8246609902239044, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 0.8207, + "step": 8957 + }, + { + "epoch": 2.824976348155156, + "grad_norm": 0.4375, + "learning_rate": 0.0002, + "loss": 1.0379, + "step": 8958 + }, + { + "epoch": 2.825291706086408, + "grad_norm": 0.419921875, + "learning_rate": 0.0002, + "loss": 0.8918, + "step": 8959 + }, + { + "epoch": 2.8256070640176603, + "grad_norm": 0.46875, + "learning_rate": 0.0002, + "loss": 0.9999, + "step": 8960 + }, + { + "epoch": 2.825922421948912, + "grad_norm": 0.443359375, + "learning_rate": 0.0002, + "loss": 0.9985, + "step": 8961 + }, + { + "epoch": 2.826237779880164, + "grad_norm": 0.48828125, + "learning_rate": 0.0002, + "loss": 0.8997, + "step": 8962 + }, + { + "epoch": 2.826553137811416, + "grad_norm": 0.44921875, + "learning_rate": 0.0002, + "loss": 0.905, + "step": 8963 + }, + { + "epoch": 2.826868495742668, + "grad_norm": 0.494140625, + "learning_rate": 0.0002, + "loss": 0.7811, + "step": 8964 + }, + { + "epoch": 2.8271838536739198, + "grad_norm": 0.466796875, + "learning_rate": 0.0002, + "loss": 0.9283, + "step": 8965 + }, + { + "epoch": 2.827499211605172, + "grad_norm": 0.5546875, + "learning_rate": 0.0002, + "loss": 0.9192, + "step": 8966 + }, + { + "epoch": 2.827814569536424, + "grad_norm": 0.609375, + "learning_rate": 0.0002, + "loss": 1.0872, + "step": 8967 + }, + { + "epoch": 2.8281299274676757, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 0.7755, + "step": 8968 + }, + { + "epoch": 2.828445285398928, + "grad_norm": 0.609375, + "learning_rate": 0.0002, + "loss": 1.0021, + "step": 8969 + }, + { + "epoch": 2.8287606433301797, + "grad_norm": 0.6875, + "learning_rate": 0.0002, + "loss": 0.9479, + "step": 8970 + }, + { + "epoch": 2.8290760012614315, + "grad_norm": 0.7578125, + "learning_rate": 0.0002, + "loss": 1.0725, + "step": 8971 + }, + { + "epoch": 2.829391359192684, + "grad_norm": 0.81640625, + "learning_rate": 0.0002, + "loss": 0.9761, + "step": 8972 + }, + { + "epoch": 2.8297067171239356, + "grad_norm": 1.0, + "learning_rate": 0.0002, + "loss": 1.1108, + "step": 8973 + }, + { + "epoch": 2.8300220750551874, + "grad_norm": 0.7890625, + "learning_rate": 0.0002, + "loss": 1.0925, + "step": 8974 + }, + { + "epoch": 2.8303374329864397, + "grad_norm": 1.0390625, + "learning_rate": 0.0002, + "loss": 1.2325, + "step": 8975 + }, + { + "epoch": 2.8306527909176915, + "grad_norm": 1.109375, + "learning_rate": 0.0002, + "loss": 1.1769, + "step": 8976 + }, + { + "epoch": 2.8306527909176915, + "eval_loss": 1.8055412769317627, + "eval_runtime": 148.956, + "eval_samples_per_second": 6.713, + "eval_steps_per_second": 6.713, + "step": 8976 + }, + { + "epoch": 2.8306527909176915, + "mmlu_eval_accuracy": 0.5889018877964264, + "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, + "mmlu_eval_accuracy_anatomy": 0.5714285714285714, + "mmlu_eval_accuracy_astronomy": 0.6875, + "mmlu_eval_accuracy_business_ethics": 0.8181818181818182, + "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034, + "mmlu_eval_accuracy_college_biology": 0.9375, + "mmlu_eval_accuracy_college_chemistry": 0.25, + "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365, + "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365, + "mmlu_eval_accuracy_college_medicine": 0.5, + "mmlu_eval_accuracy_college_physics": 0.45454545454545453, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156, + "mmlu_eval_accuracy_econometrics": 0.4166666666666667, + "mmlu_eval_accuracy_electrical_engineering": 0.625, + "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244, + "mmlu_eval_accuracy_formal_logic": 0.2857142857142857, + "mmlu_eval_accuracy_global_facts": 0.2, + "mmlu_eval_accuracy_high_school_biology": 0.5625, + "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8571428571428571, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093, + "mmlu_eval_accuracy_high_school_mathematics": 0.3793103448275862, + "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154, + "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354, + "mmlu_eval_accuracy_high_school_psychology": 0.8, + "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, + "mmlu_eval_accuracy_high_school_us_history": 0.8636363636363636, + "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307, + "mmlu_eval_accuracy_human_aging": 0.6521739130434783, + "mmlu_eval_accuracy_human_sexuality": 0.5, + "mmlu_eval_accuracy_international_law": 0.9230769230769231, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, + "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.72, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279, + "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842, + "mmlu_eval_accuracy_moral_scenarios": 0.24, + "mmlu_eval_accuracy_nutrition": 0.7575757575757576, + "mmlu_eval_accuracy_philosophy": 0.6176470588235294, + "mmlu_eval_accuracy_prehistory": 0.5714285714285714, + "mmlu_eval_accuracy_professional_accounting": 0.5161290322580645, + "mmlu_eval_accuracy_professional_law": 0.38823529411764707, + "mmlu_eval_accuracy_professional_medicine": 0.6774193548387096, + "mmlu_eval_accuracy_professional_psychology": 0.6086956521739131, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.6666666666666666, + "mmlu_eval_accuracy_sociology": 0.9545454545454546, + "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, + "mmlu_eval_accuracy_virology": 0.5, + "mmlu_eval_accuracy_world_religions": 0.7368421052631579, + "mmlu_loss": 6.170890209178065, + "step": 8976 + }, + { + "epoch": 2.8309681488489433, + "grad_norm": 0.96484375, + "learning_rate": 0.0002, + "loss": 1.192, + "step": 8977 + }, + { + "epoch": 2.8312835067801956, + "grad_norm": 0.96484375, + "learning_rate": 0.0002, + "loss": 1.2481, + "step": 8978 + }, + { + "epoch": 2.8315988647114474, + "grad_norm": 0.9921875, + "learning_rate": 0.0002, + "loss": 1.2579, + "step": 8979 + }, + { + "epoch": 2.831914222642699, + "grad_norm": 1.1953125, + "learning_rate": 0.0002, + "loss": 1.4292, + "step": 8980 + }, + { + "epoch": 2.8322295805739515, + "grad_norm": 1.3046875, + "learning_rate": 0.0002, + "loss": 1.7413, + "step": 8981 + }, + { + "epoch": 2.8325449385052033, + "grad_norm": 1.1796875, + "learning_rate": 0.0002, + "loss": 1.589, + "step": 8982 + }, + { + "epoch": 2.8328602964364555, + "grad_norm": 1.0390625, + "learning_rate": 0.0002, + "loss": 1.7141, + "step": 8983 + }, + { + "epoch": 2.8331756543677074, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 1.9222, + "step": 8984 + }, + { + "epoch": 2.833491012298959, + "grad_norm": 1.0390625, + "learning_rate": 0.0002, + "loss": 1.6298, + "step": 8985 + }, + { + "epoch": 2.8338063702302114, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 1.6827, + "step": 8986 + }, + { + "epoch": 2.8341217281614632, + "grad_norm": 1.265625, + "learning_rate": 0.0002, + "loss": 1.1932, + "step": 8987 + }, + { + "epoch": 2.8344370860927155, + "grad_norm": 1.453125, + "learning_rate": 0.0002, + "loss": 1.8949, + "step": 8988 + }, + { + "epoch": 2.8347524440239673, + "grad_norm": 1.2109375, + "learning_rate": 0.0002, + "loss": 1.2414, + "step": 8989 + }, + { + "epoch": 2.835067801955219, + "grad_norm": 1.5703125, + "learning_rate": 0.0002, + "loss": 1.6804, + "step": 8990 + }, + { + "epoch": 2.8353831598864714, + "grad_norm": 1.5078125, + "learning_rate": 0.0002, + "loss": 1.9763, + "step": 8991 + }, + { + "epoch": 2.835698517817723, + "grad_norm": 2.078125, + "learning_rate": 0.0002, + "loss": 2.1071, + "step": 8992 + }, + { + "epoch": 2.836013875748975, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 0.9502, + "step": 8993 + }, + { + "epoch": 2.8363292336802273, + "grad_norm": 0.271484375, + "learning_rate": 0.0002, + "loss": 1.0338, + "step": 8994 + }, + { + "epoch": 2.836644591611479, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 0.9321, + "step": 8995 + }, + { + "epoch": 2.836959949542731, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.9598, + "step": 8996 + }, + { + "epoch": 2.837275307473983, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 0.9613, + "step": 8997 + }, + { + "epoch": 2.837590665405235, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.9878, + "step": 8998 + }, + { + "epoch": 2.837906023336487, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 0.9416, + "step": 8999 + }, + { + "epoch": 2.838221381267739, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.9374, + "step": 9000 + }, + { + "epoch": 2.838536739198991, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 0.8792, + "step": 9001 + }, + { + "epoch": 2.8388520971302427, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 1.0387, + "step": 9002 + }, + { + "epoch": 2.839167455061495, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 0.9348, + "step": 9003 + }, + { + "epoch": 2.8394828129927467, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 1.0171, + "step": 9004 + }, + { + "epoch": 2.8397981709239986, + "grad_norm": 0.3828125, + "learning_rate": 0.0002, + "loss": 0.9531, + "step": 9005 + }, + { + "epoch": 2.840113528855251, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 0.8877, + "step": 9006 + }, + { + "epoch": 2.8404288867865026, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 0.9478, + "step": 9007 + }, + { + "epoch": 2.8407442447177544, + "grad_norm": 0.38671875, + "learning_rate": 0.0002, + "loss": 0.8882, + "step": 9008 + }, + { + "epoch": 2.8410596026490067, + "grad_norm": 0.408203125, + "learning_rate": 0.0002, + "loss": 0.924, + "step": 9009 + }, + { + "epoch": 2.8413749605802585, + "grad_norm": 0.447265625, + "learning_rate": 0.0002, + "loss": 1.0386, + "step": 9010 + }, + { + "epoch": 2.8416903185115103, + "grad_norm": 0.451171875, + "learning_rate": 0.0002, + "loss": 0.9775, + "step": 9011 + }, + { + "epoch": 2.8420056764427626, + "grad_norm": 0.451171875, + "learning_rate": 0.0002, + "loss": 1.0121, + "step": 9012 + }, + { + "epoch": 2.8423210343740144, + "grad_norm": 0.419921875, + "learning_rate": 0.0002, + "loss": 0.7625, + "step": 9013 + }, + { + "epoch": 2.842636392305266, + "grad_norm": 0.46484375, + "learning_rate": 0.0002, + "loss": 0.9908, + "step": 9014 + }, + { + "epoch": 2.8429517502365185, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 0.8555, + "step": 9015 + }, + { + "epoch": 2.8432671081677703, + "grad_norm": 0.56640625, + "learning_rate": 0.0002, + "loss": 1.0102, + "step": 9016 + }, + { + "epoch": 2.8435824660990225, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 0.9352, + "step": 9017 + }, + { + "epoch": 2.8438978240302744, + "grad_norm": 0.578125, + "learning_rate": 0.0002, + "loss": 0.7829, + "step": 9018 + }, + { + "epoch": 2.844213181961526, + "grad_norm": 0.56640625, + "learning_rate": 0.0002, + "loss": 0.9584, + "step": 9019 + }, + { + "epoch": 2.8445285398927784, + "grad_norm": 0.578125, + "learning_rate": 0.0002, + "loss": 0.8962, + "step": 9020 + }, + { + "epoch": 2.8448438978240302, + "grad_norm": 0.7734375, + "learning_rate": 0.0002, + "loss": 0.9401, + "step": 9021 + }, + { + "epoch": 2.8451592557552825, + "grad_norm": 0.80859375, + "learning_rate": 0.0002, + "loss": 0.944, + "step": 9022 + }, + { + "epoch": 2.8454746136865343, + "grad_norm": 0.765625, + "learning_rate": 0.0002, + "loss": 1.1417, + "step": 9023 + }, + { + "epoch": 2.845789971617786, + "grad_norm": 1.0, + "learning_rate": 0.0002, + "loss": 0.9494, + "step": 9024 + }, + { + "epoch": 2.8461053295490384, + "grad_norm": 0.8359375, + "learning_rate": 0.0002, + "loss": 1.0249, + "step": 9025 + }, + { + "epoch": 2.84642068748029, + "grad_norm": 1.0546875, + "learning_rate": 0.0002, + "loss": 1.1367, + "step": 9026 + }, + { + "epoch": 2.846736045411542, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 1.2834, + "step": 9027 + }, + { + "epoch": 2.8470514033427943, + "grad_norm": 0.83984375, + "learning_rate": 0.0002, + "loss": 1.1066, + "step": 9028 + }, + { + "epoch": 2.847366761274046, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 1.2041, + "step": 9029 + }, + { + "epoch": 2.847682119205298, + "grad_norm": 1.03125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 9030 + }, + { + "epoch": 2.84799747713655, + "grad_norm": 1.125, + "learning_rate": 0.0002, + "loss": 1.449, + "step": 9031 + }, + { + "epoch": 2.848312835067802, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 1.7964, + "step": 9032 + }, + { + "epoch": 2.848628192999054, + "grad_norm": 1.2578125, + "learning_rate": 0.0002, + "loss": 1.5863, + "step": 9033 + }, + { + "epoch": 2.848943550930306, + "grad_norm": 1.375, + "learning_rate": 0.0002, + "loss": 1.679, + "step": 9034 + }, + { + "epoch": 2.849258908861558, + "grad_norm": 1.3046875, + "learning_rate": 0.0002, + "loss": 1.6231, + "step": 9035 + }, + { + "epoch": 2.8495742667928097, + "grad_norm": 1.390625, + "learning_rate": 0.0002, + "loss": 1.7126, + "step": 9036 + }, + { + "epoch": 2.849889624724062, + "grad_norm": 1.4296875, + "learning_rate": 0.0002, + "loss": 1.5515, + "step": 9037 + }, + { + "epoch": 2.8502049826553137, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 1.541, + "step": 9038 + }, + { + "epoch": 2.8505203405865656, + "grad_norm": 1.1953125, + "learning_rate": 0.0002, + "loss": 1.5253, + "step": 9039 + }, + { + "epoch": 2.850835698517818, + "grad_norm": 1.15625, + "learning_rate": 0.0002, + "loss": 1.4001, + "step": 9040 + }, + { + "epoch": 2.8511510564490696, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 9041 + }, + { + "epoch": 2.8514664143803214, + "grad_norm": 1.265625, + "learning_rate": 0.0002, + "loss": 1.7621, + "step": 9042 + }, + { + "epoch": 2.8517817723115737, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 0.9882, + "step": 9043 + }, + { + "epoch": 2.8520971302428255, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 0.8742, + "step": 9044 + }, + { + "epoch": 2.8524124881740773, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 0.9433, + "step": 9045 + }, + { + "epoch": 2.8527278461053296, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 0.9865, + "step": 9046 + }, + { + "epoch": 2.8530432040365814, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 0.9475, + "step": 9047 + }, + { + "epoch": 2.853358561967833, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.8969, + "step": 9048 + }, + { + "epoch": 2.8536739198990855, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 0.8434, + "step": 9049 + }, + { + "epoch": 2.8539892778303373, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 0.9128, + "step": 9050 + }, + { + "epoch": 2.8543046357615895, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.9541, + "step": 9051 + }, + { + "epoch": 2.8546199936928414, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 0.9439, + "step": 9052 + }, + { + "epoch": 2.8549353516240936, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 0.9496, + "step": 9053 + }, + { + "epoch": 2.8552507095553454, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 1.0241, + "step": 9054 + }, + { + "epoch": 2.8555660674865972, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 0.8955, + "step": 9055 + }, + { + "epoch": 2.8558814254178495, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 0.8801, + "step": 9056 + }, + { + "epoch": 2.8561967833491013, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 0.8589, + "step": 9057 + }, + { + "epoch": 2.856512141280353, + "grad_norm": 0.369140625, + "learning_rate": 0.0002, + "loss": 0.937, + "step": 9058 + }, + { + "epoch": 2.8568274992116054, + "grad_norm": 0.349609375, + "learning_rate": 0.0002, + "loss": 0.8966, + "step": 9059 + }, + { + "epoch": 2.857142857142857, + "grad_norm": 0.49609375, + "learning_rate": 0.0002, + "loss": 0.9214, + "step": 9060 + }, + { + "epoch": 2.857458215074109, + "grad_norm": 0.4296875, + "learning_rate": 0.0002, + "loss": 1.0518, + "step": 9061 + }, + { + "epoch": 2.8577735730053613, + "grad_norm": 0.40234375, + "learning_rate": 0.0002, + "loss": 0.8728, + "step": 9062 + }, + { + "epoch": 2.858088930936613, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 1.0353, + "step": 9063 + }, + { + "epoch": 2.858404288867865, + "grad_norm": 0.43359375, + "learning_rate": 0.0002, + "loss": 0.8742, + "step": 9064 + }, + { + "epoch": 2.858719646799117, + "grad_norm": 0.458984375, + "learning_rate": 0.0002, + "loss": 0.7565, + "step": 9065 + }, + { + "epoch": 2.859035004730369, + "grad_norm": 0.51953125, + "learning_rate": 0.0002, + "loss": 0.9136, + "step": 9066 + }, + { + "epoch": 2.859350362661621, + "grad_norm": 0.63671875, + "learning_rate": 0.0002, + "loss": 1.0123, + "step": 9067 + }, + { + "epoch": 2.859665720592873, + "grad_norm": 0.578125, + "learning_rate": 0.0002, + "loss": 0.8466, + "step": 9068 + }, + { + "epoch": 2.859981078524125, + "grad_norm": 0.66796875, + "learning_rate": 0.0002, + "loss": 0.9825, + "step": 9069 + }, + { + "epoch": 2.8602964364553767, + "grad_norm": 0.8359375, + "learning_rate": 0.0002, + "loss": 1.0654, + "step": 9070 + }, + { + "epoch": 2.860611794386629, + "grad_norm": 0.625, + "learning_rate": 0.0002, + "loss": 0.9707, + "step": 9071 + }, + { + "epoch": 2.8609271523178808, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 0.8612, + "step": 9072 + }, + { + "epoch": 2.8612425102491326, + "grad_norm": 0.8203125, + "learning_rate": 0.0002, + "loss": 1.1059, + "step": 9073 + }, + { + "epoch": 2.861557868180385, + "grad_norm": 1.0546875, + "learning_rate": 0.0002, + "loss": 1.1797, + "step": 9074 + }, + { + "epoch": 2.8618732261116366, + "grad_norm": 0.9140625, + "learning_rate": 0.0002, + "loss": 1.2514, + "step": 9075 + }, + { + "epoch": 2.8621885840428885, + "grad_norm": 0.94921875, + "learning_rate": 0.0002, + "loss": 1.1785, + "step": 9076 + }, + { + "epoch": 2.8625039419741407, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.3811, + "step": 9077 + }, + { + "epoch": 2.8628192999053925, + "grad_norm": 1.2109375, + "learning_rate": 0.0002, + "loss": 1.1887, + "step": 9078 + }, + { + "epoch": 2.8631346578366443, + "grad_norm": 1.15625, + "learning_rate": 0.0002, + "loss": 1.5655, + "step": 9079 + }, + { + "epoch": 2.8634500157678966, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.237, + "step": 9080 + }, + { + "epoch": 2.8637653736991484, + "grad_norm": 1.265625, + "learning_rate": 0.0002, + "loss": 1.3943, + "step": 9081 + }, + { + "epoch": 2.8640807316304007, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 1.4606, + "step": 9082 + }, + { + "epoch": 2.8643960895616525, + "grad_norm": 1.28125, + "learning_rate": 0.0002, + "loss": 1.5185, + "step": 9083 + }, + { + "epoch": 2.8647114474929043, + "grad_norm": 1.234375, + "learning_rate": 0.0002, + "loss": 1.599, + "step": 9084 + }, + { + "epoch": 2.8650268054241566, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 1.6656, + "step": 9085 + }, + { + "epoch": 2.8653421633554084, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 1.5984, + "step": 9086 + }, + { + "epoch": 2.8656575212866606, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 1.4598, + "step": 9087 + }, + { + "epoch": 2.8659728792179124, + "grad_norm": 1.3203125, + "learning_rate": 0.0002, + "loss": 2.0695, + "step": 9088 + }, + { + "epoch": 2.8662882371491643, + "grad_norm": 1.078125, + "learning_rate": 0.0002, + "loss": 1.3881, + "step": 9089 + }, + { + "epoch": 2.8666035950804165, + "grad_norm": 1.21875, + "learning_rate": 0.0002, + "loss": 1.5048, + "step": 9090 + }, + { + "epoch": 2.8669189530116683, + "grad_norm": 1.265625, + "learning_rate": 0.0002, + "loss": 1.4208, + "step": 9091 + }, + { + "epoch": 2.86723431094292, + "grad_norm": 0.94921875, + "learning_rate": 0.0002, + "loss": 1.592, + "step": 9092 + }, + { + "epoch": 2.8675496688741724, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 1.0889, + "step": 9093 + }, + { + "epoch": 2.867865026805424, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 1.0221, + "step": 9094 + }, + { + "epoch": 2.868180384736676, + "grad_norm": 0.255859375, + "learning_rate": 0.0002, + "loss": 0.9253, + "step": 9095 + }, + { + "epoch": 2.8684957426679283, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 0.9046, + "step": 9096 + }, + { + "epoch": 2.86881110059918, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 0.9602, + "step": 9097 + }, + { + "epoch": 2.869126458530432, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 0.9721, + "step": 9098 + }, + { + "epoch": 2.869441816461684, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 1.0476, + "step": 9099 + }, + { + "epoch": 2.869757174392936, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 0.9599, + "step": 9100 + }, + { + "epoch": 2.870072532324188, + "grad_norm": 0.369140625, + "learning_rate": 0.0002, + "loss": 0.9212, + "step": 9101 + }, + { + "epoch": 2.87038789025544, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 0.9403, + "step": 9102 + }, + { + "epoch": 2.870703248186692, + "grad_norm": 0.37890625, + "learning_rate": 0.0002, + "loss": 0.9766, + "step": 9103 + }, + { + "epoch": 2.8710186061179437, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 0.8448, + "step": 9104 + }, + { + "epoch": 2.871333964049196, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 0.9237, + "step": 9105 + }, + { + "epoch": 2.8716493219804478, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 0.953, + "step": 9106 + }, + { + "epoch": 2.8719646799116996, + "grad_norm": 0.431640625, + "learning_rate": 0.0002, + "loss": 0.9901, + "step": 9107 + }, + { + "epoch": 2.872280037842952, + "grad_norm": 0.3984375, + "learning_rate": 0.0002, + "loss": 1.0182, + "step": 9108 + }, + { + "epoch": 2.8725953957742036, + "grad_norm": 0.43359375, + "learning_rate": 0.0002, + "loss": 0.9215, + "step": 9109 + }, + { + "epoch": 2.8729107537054555, + "grad_norm": 0.48046875, + "learning_rate": 0.0002, + "loss": 0.8933, + "step": 9110 + }, + { + "epoch": 2.8732261116367077, + "grad_norm": 0.515625, + "learning_rate": 0.0002, + "loss": 0.9371, + "step": 9111 + }, + { + "epoch": 2.8735414695679595, + "grad_norm": 0.427734375, + "learning_rate": 0.0002, + "loss": 0.89, + "step": 9112 + }, + { + "epoch": 2.8738568274992113, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 0.8763, + "step": 9113 + }, + { + "epoch": 2.8741721854304636, + "grad_norm": 0.478515625, + "learning_rate": 0.0002, + "loss": 0.81, + "step": 9114 + }, + { + "epoch": 2.8744875433617154, + "grad_norm": 0.62109375, + "learning_rate": 0.0002, + "loss": 1.0891, + "step": 9115 + }, + { + "epoch": 2.8748029012929677, + "grad_norm": 0.5390625, + "learning_rate": 0.0002, + "loss": 1.122, + "step": 9116 + }, + { + "epoch": 2.8751182592242195, + "grad_norm": 0.49609375, + "learning_rate": 0.0002, + "loss": 0.8662, + "step": 9117 + }, + { + "epoch": 2.8754336171554713, + "grad_norm": 0.55859375, + "learning_rate": 0.0002, + "loss": 0.8953, + "step": 9118 + }, + { + "epoch": 2.8757489750867236, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 0.8628, + "step": 9119 + }, + { + "epoch": 2.8760643330179754, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 0.8141, + "step": 9120 + }, + { + "epoch": 2.8763796909492276, + "grad_norm": 0.703125, + "learning_rate": 0.0002, + "loss": 1.1121, + "step": 9121 + }, + { + "epoch": 2.8766950488804794, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 0.844, + "step": 9122 + }, + { + "epoch": 2.8770104068117313, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 0.9717, + "step": 9123 + }, + { + "epoch": 2.8773257647429835, + "grad_norm": 2.3125, + "learning_rate": 0.0002, + "loss": 1.207, + "step": 9124 + }, + { + "epoch": 2.8776411226742353, + "grad_norm": 0.80078125, + "learning_rate": 0.0002, + "loss": 1.0737, + "step": 9125 + }, + { + "epoch": 2.877956480605487, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 1.3813, + "step": 9126 + }, + { + "epoch": 2.8782718385367394, + "grad_norm": 1.390625, + "learning_rate": 0.0002, + "loss": 1.0059, + "step": 9127 + }, + { + "epoch": 2.878587196467991, + "grad_norm": 0.953125, + "learning_rate": 0.0002, + "loss": 1.1769, + "step": 9128 + }, + { + "epoch": 2.878902554399243, + "grad_norm": 1.03125, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 9129 + }, + { + "epoch": 2.8792179123304953, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 9130 + }, + { + "epoch": 2.879533270261747, + "grad_norm": 1.171875, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 9131 + }, + { + "epoch": 2.879848628192999, + "grad_norm": 1.09375, + "learning_rate": 0.0002, + "loss": 1.3944, + "step": 9132 + }, + { + "epoch": 2.880163986124251, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 1.6162, + "step": 9133 + }, + { + "epoch": 2.880479344055503, + "grad_norm": 1.28125, + "learning_rate": 0.0002, + "loss": 1.6316, + "step": 9134 + }, + { + "epoch": 2.880794701986755, + "grad_norm": 1.125, + "learning_rate": 0.0002, + "loss": 1.6732, + "step": 9135 + }, + { + "epoch": 2.881110059918007, + "grad_norm": 1.921875, + "learning_rate": 0.0002, + "loss": 1.5837, + "step": 9136 + }, + { + "epoch": 2.881425417849259, + "grad_norm": 1.453125, + "learning_rate": 0.0002, + "loss": 1.6252, + "step": 9137 + }, + { + "epoch": 2.8817407757805107, + "grad_norm": 1.390625, + "learning_rate": 0.0002, + "loss": 1.6695, + "step": 9138 + }, + { + "epoch": 2.882056133711763, + "grad_norm": 1.46875, + "learning_rate": 0.0002, + "loss": 1.5951, + "step": 9139 + }, + { + "epoch": 2.8823714916430148, + "grad_norm": 1.1796875, + "learning_rate": 0.0002, + "loss": 1.8997, + "step": 9140 + }, + { + "epoch": 2.8826868495742666, + "grad_norm": 1.5234375, + "learning_rate": 0.0002, + "loss": 1.6918, + "step": 9141 + }, + { + "epoch": 2.883002207505519, + "grad_norm": 1.0390625, + "learning_rate": 0.0002, + "loss": 1.8435, + "step": 9142 + }, + { + "epoch": 2.8833175654367706, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 0.9804, + "step": 9143 + }, + { + "epoch": 2.8836329233680225, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 0.918, + "step": 9144 + }, + { + "epoch": 2.8839482812992747, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 0.93, + "step": 9145 + }, + { + "epoch": 2.8842636392305265, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 0.938, + "step": 9146 + }, + { + "epoch": 2.8845789971617783, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.9425, + "step": 9147 + }, + { + "epoch": 2.8848943550930306, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 1.0491, + "step": 9148 + }, + { + "epoch": 2.8852097130242824, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.9534, + "step": 9149 + }, + { + "epoch": 2.8855250709555347, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 0.9029, + "step": 9150 + }, + { + "epoch": 2.8858404288867865, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 0.8812, + "step": 9151 + }, + { + "epoch": 2.8861557868180387, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 0.9079, + "step": 9152 + }, + { + "epoch": 2.8864711447492906, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 0.9387, + "step": 9153 + }, + { + "epoch": 2.8867865026805424, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 1.0157, + "step": 9154 + }, + { + "epoch": 2.8871018606117946, + "grad_norm": 0.3828125, + "learning_rate": 0.0002, + "loss": 0.9867, + "step": 9155 + }, + { + "epoch": 2.8874172185430464, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 0.9289, + "step": 9156 + }, + { + "epoch": 2.8877325764742983, + "grad_norm": 0.390625, + "learning_rate": 0.0002, + "loss": 0.8844, + "step": 9157 + }, + { + "epoch": 2.8880479344055505, + "grad_norm": 0.3984375, + "learning_rate": 0.0002, + "loss": 0.8964, + "step": 9158 + }, + { + "epoch": 2.8883632923368023, + "grad_norm": 0.39453125, + "learning_rate": 0.0002, + "loss": 0.8927, + "step": 9159 + }, + { + "epoch": 2.888678650268054, + "grad_norm": 0.40625, + "learning_rate": 0.0002, + "loss": 0.958, + "step": 9160 + }, + { + "epoch": 2.8889940081993064, + "grad_norm": 0.4375, + "learning_rate": 0.0002, + "loss": 0.7366, + "step": 9161 + }, + { + "epoch": 2.889309366130558, + "grad_norm": 0.498046875, + "learning_rate": 0.0002, + "loss": 0.9465, + "step": 9162 + }, + { + "epoch": 2.88962472406181, + "grad_norm": 0.69921875, + "learning_rate": 0.0002, + "loss": 0.881, + "step": 9163 + }, + { + "epoch": 2.88962472406181, + "eval_loss": 1.7740569114685059, + "eval_runtime": 149.0475, + "eval_samples_per_second": 6.709, + "eval_steps_per_second": 6.709, + "step": 9163 + }, + { + "epoch": 2.88962472406181, + "mmlu_eval_accuracy": 0.59246196665691, + "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, + "mmlu_eval_accuracy_anatomy": 0.5, + "mmlu_eval_accuracy_astronomy": 0.6875, + "mmlu_eval_accuracy_business_ethics": 0.8181818181818182, + "mmlu_eval_accuracy_clinical_knowledge": 0.6896551724137931, + "mmlu_eval_accuracy_college_biology": 0.8125, + "mmlu_eval_accuracy_college_chemistry": 0.375, + "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365, + "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453, + "mmlu_eval_accuracy_college_medicine": 0.5454545454545454, + "mmlu_eval_accuracy_college_physics": 0.45454545454545453, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.5, + "mmlu_eval_accuracy_econometrics": 0.5, + "mmlu_eval_accuracy_electrical_engineering": 0.5, + "mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536, + "mmlu_eval_accuracy_formal_logic": 0.35714285714285715, + "mmlu_eval_accuracy_global_facts": 0.3, + "mmlu_eval_accuracy_high_school_biology": 0.59375, + "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222, + "mmlu_eval_accuracy_high_school_geography": 0.8636363636363636, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8571428571428571, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.6046511627906976, + "mmlu_eval_accuracy_high_school_mathematics": 0.4482758620689655, + "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384, + "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354, + "mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, + "mmlu_eval_accuracy_high_school_us_history": 0.8636363636363636, + "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923, + "mmlu_eval_accuracy_human_aging": 0.6086956521739131, + "mmlu_eval_accuracy_human_sexuality": 0.5, + "mmlu_eval_accuracy_international_law": 0.9230769230769231, + "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, + "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, + "mmlu_eval_accuracy_machine_learning": 0.45454545454545453, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.72, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279, + "mmlu_eval_accuracy_moral_disputes": 0.5, + "mmlu_eval_accuracy_moral_scenarios": 0.24, + "mmlu_eval_accuracy_nutrition": 0.7575757575757576, + "mmlu_eval_accuracy_philosophy": 0.5882352941176471, + "mmlu_eval_accuracy_prehistory": 0.5428571428571428, + "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644, + "mmlu_eval_accuracy_professional_law": 0.38823529411764707, + "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806, + "mmlu_eval_accuracy_professional_psychology": 0.6521739130434783, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.7407407407407407, + "mmlu_eval_accuracy_sociology": 0.9545454545454546, + "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, + "mmlu_eval_accuracy_virology": 0.5, + "mmlu_eval_accuracy_world_religions": 0.7368421052631579, + "mmlu_loss": 6.164567319183549, + "step": 9163 + }, + { + "epoch": 2.8899400819930623, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 1.0874, + "step": 9164 + }, + { + "epoch": 2.890255439924314, + "grad_norm": 0.5390625, + "learning_rate": 0.0002, + "loss": 0.9896, + "step": 9165 + }, + { + "epoch": 2.890570797855566, + "grad_norm": 0.52734375, + "learning_rate": 0.0002, + "loss": 0.924, + "step": 9166 + }, + { + "epoch": 2.890886155786818, + "grad_norm": 0.6171875, + "learning_rate": 0.0002, + "loss": 0.9469, + "step": 9167 + }, + { + "epoch": 2.89120151371807, + "grad_norm": 0.67578125, + "learning_rate": 0.0002, + "loss": 1.0683, + "step": 9168 + }, + { + "epoch": 2.891516871649322, + "grad_norm": 0.6484375, + "learning_rate": 0.0002, + "loss": 0.9594, + "step": 9169 + }, + { + "epoch": 2.891832229580574, + "grad_norm": 0.640625, + "learning_rate": 0.0002, + "loss": 0.9733, + "step": 9170 + }, + { + "epoch": 2.892147587511826, + "grad_norm": 0.765625, + "learning_rate": 0.0002, + "loss": 1.019, + "step": 9171 + }, + { + "epoch": 2.8924629454430777, + "grad_norm": 0.6640625, + "learning_rate": 0.0002, + "loss": 1.0102, + "step": 9172 + }, + { + "epoch": 2.89277830337433, + "grad_norm": 1.125, + "learning_rate": 0.0002, + "loss": 1.4751, + "step": 9173 + }, + { + "epoch": 2.8930936613055818, + "grad_norm": 0.8828125, + "learning_rate": 0.0002, + "loss": 1.0672, + "step": 9174 + }, + { + "epoch": 2.8934090192368336, + "grad_norm": 1.09375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9175 + }, + { + "epoch": 2.893724377168086, + "grad_norm": 1.1171875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 9176 + }, + { + "epoch": 2.8940397350993377, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 1.2599, + "step": 9177 + }, + { + "epoch": 2.8943550930305895, + "grad_norm": 1.234375, + "learning_rate": 0.0002, + "loss": 1.6144, + "step": 9178 + }, + { + "epoch": 2.8946704509618417, + "grad_norm": 0.953125, + "learning_rate": 0.0002, + "loss": 1.1643, + "step": 9179 + }, + { + "epoch": 2.8949858088930935, + "grad_norm": 0.890625, + "learning_rate": 0.0002, + "loss": 1.2197, + "step": 9180 + }, + { + "epoch": 2.895301166824346, + "grad_norm": 1.2265625, + "learning_rate": 0.0002, + "loss": 1.494, + "step": 9181 + }, + { + "epoch": 2.8956165247555976, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 1.7919, + "step": 9182 + }, + { + "epoch": 2.8959318826868494, + "grad_norm": 1.3046875, + "learning_rate": 0.0002, + "loss": 1.4388, + "step": 9183 + }, + { + "epoch": 2.8962472406181017, + "grad_norm": 1.515625, + "learning_rate": 0.0002, + "loss": 2.0875, + "step": 9184 + }, + { + "epoch": 2.8965625985493535, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 1.8108, + "step": 9185 + }, + { + "epoch": 2.8968779564806058, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 1.7479, + "step": 9186 + }, + { + "epoch": 2.8971933144118576, + "grad_norm": 1.21875, + "learning_rate": 0.0002, + "loss": 1.6651, + "step": 9187 + }, + { + "epoch": 2.8975086723431094, + "grad_norm": 0.88671875, + "learning_rate": 0.0002, + "loss": 1.3928, + "step": 9188 + }, + { + "epoch": 2.8978240302743616, + "grad_norm": 1.171875, + "learning_rate": 0.0002, + "loss": 1.4801, + "step": 9189 + }, + { + "epoch": 2.8981393882056135, + "grad_norm": 1.0546875, + "learning_rate": 0.0002, + "loss": 1.4939, + "step": 9190 + }, + { + "epoch": 2.8984547461368653, + "grad_norm": 1.2890625, + "learning_rate": 0.0002, + "loss": 1.6091, + "step": 9191 + }, + { + "epoch": 2.8987701040681175, + "grad_norm": 1.171875, + "learning_rate": 0.0002, + "loss": 1.4234, + "step": 9192 + }, + { + "epoch": 2.8990854619993693, + "grad_norm": 0.255859375, + "learning_rate": 0.0002, + "loss": 1.0364, + "step": 9193 + }, + { + "epoch": 2.899400819930621, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 0.9084, + "step": 9194 + }, + { + "epoch": 2.8997161778618734, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 0.9167, + "step": 9195 + }, + { + "epoch": 2.9000315357931252, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 0.9351, + "step": 9196 + }, + { + "epoch": 2.900346893724377, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 0.9464, + "step": 9197 + }, + { + "epoch": 2.9006622516556293, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 0.863, + "step": 9198 + }, + { + "epoch": 2.900977609586881, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 1.0132, + "step": 9199 + }, + { + "epoch": 2.901292967518133, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 0.9459, + "step": 9200 + }, + { + "epoch": 2.901608325449385, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 0.9607, + "step": 9201 + }, + { + "epoch": 2.901923683380637, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 0.9272, + "step": 9202 + }, + { + "epoch": 2.902239041311889, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 0.9747, + "step": 9203 + }, + { + "epoch": 2.902554399243141, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 0.9192, + "step": 9204 + }, + { + "epoch": 2.902869757174393, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 0.9785, + "step": 9205 + }, + { + "epoch": 2.9031851151056447, + "grad_norm": 0.353515625, + "learning_rate": 0.0002, + "loss": 1.0312, + "step": 9206 + }, + { + "epoch": 2.903500473036897, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 0.9661, + "step": 9207 + }, + { + "epoch": 2.9038158309681488, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 0.9127, + "step": 9208 + }, + { + "epoch": 2.9041311888994006, + "grad_norm": 0.369140625, + "learning_rate": 0.0002, + "loss": 0.9552, + "step": 9209 + }, + { + "epoch": 2.904446546830653, + "grad_norm": 0.421875, + "learning_rate": 0.0002, + "loss": 0.9657, + "step": 9210 + }, + { + "epoch": 2.9047619047619047, + "grad_norm": 0.458984375, + "learning_rate": 0.0002, + "loss": 1.0528, + "step": 9211 + }, + { + "epoch": 2.9050772626931565, + "grad_norm": 0.41015625, + "learning_rate": 0.0002, + "loss": 1.0153, + "step": 9212 + }, + { + "epoch": 2.9053926206244087, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 0.9328, + "step": 9213 + }, + { + "epoch": 2.9057079785556605, + "grad_norm": 0.419921875, + "learning_rate": 0.0002, + "loss": 0.8887, + "step": 9214 + }, + { + "epoch": 2.906023336486913, + "grad_norm": 0.45703125, + "learning_rate": 0.0002, + "loss": 0.9299, + "step": 9215 + }, + { + "epoch": 2.9063386944181646, + "grad_norm": 0.470703125, + "learning_rate": 0.0002, + "loss": 1.0116, + "step": 9216 + }, + { + "epoch": 2.9066540523494164, + "grad_norm": 0.47265625, + "learning_rate": 0.0002, + "loss": 0.8092, + "step": 9217 + }, + { + "epoch": 2.9069694102806687, + "grad_norm": 0.61328125, + "learning_rate": 0.0002, + "loss": 1.0169, + "step": 9218 + }, + { + "epoch": 2.9072847682119205, + "grad_norm": 0.546875, + "learning_rate": 0.0002, + "loss": 0.7634, + "step": 9219 + }, + { + "epoch": 2.9076001261431728, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 0.9563, + "step": 9220 + }, + { + "epoch": 2.9079154840744246, + "grad_norm": 0.69921875, + "learning_rate": 0.0002, + "loss": 0.9444, + "step": 9221 + }, + { + "epoch": 2.9082308420056764, + "grad_norm": 0.58984375, + "learning_rate": 0.0002, + "loss": 0.894, + "step": 9222 + }, + { + "epoch": 2.9085461999369286, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 0.9738, + "step": 9223 + }, + { + "epoch": 2.9088615578681805, + "grad_norm": 0.78125, + "learning_rate": 0.0002, + "loss": 1.0867, + "step": 9224 + }, + { + "epoch": 2.9091769157994323, + "grad_norm": 0.890625, + "learning_rate": 0.0002, + "loss": 1.2521, + "step": 9225 + }, + { + "epoch": 2.9094922737306845, + "grad_norm": 0.81640625, + "learning_rate": 0.0002, + "loss": 1.0529, + "step": 9226 + }, + { + "epoch": 2.9098076316619363, + "grad_norm": 1.515625, + "learning_rate": 0.0002, + "loss": 1.2499, + "step": 9227 + }, + { + "epoch": 2.910122989593188, + "grad_norm": 0.90234375, + "learning_rate": 0.0002, + "loss": 1.0445, + "step": 9228 + }, + { + "epoch": 2.9104383475244404, + "grad_norm": 0.953125, + "learning_rate": 0.0002, + "loss": 1.1889, + "step": 9229 + }, + { + "epoch": 2.9107537054556922, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 1.1239, + "step": 9230 + }, + { + "epoch": 2.911069063386944, + "grad_norm": 1.5078125, + "learning_rate": 0.0002, + "loss": 1.3772, + "step": 9231 + }, + { + "epoch": 2.9113844213181963, + "grad_norm": 1.3203125, + "learning_rate": 0.0002, + "loss": 1.3686, + "step": 9232 + }, + { + "epoch": 2.911699779249448, + "grad_norm": 0.99609375, + "learning_rate": 0.0002, + "loss": 1.2162, + "step": 9233 + }, + { + "epoch": 2.9120151371807, + "grad_norm": 1.2578125, + "learning_rate": 0.0002, + "loss": 1.7522, + "step": 9234 + }, + { + "epoch": 2.912330495111952, + "grad_norm": 1.484375, + "learning_rate": 0.0002, + "loss": 1.6918, + "step": 9235 + }, + { + "epoch": 2.912645853043204, + "grad_norm": 1.15625, + "learning_rate": 0.0002, + "loss": 1.4655, + "step": 9236 + }, + { + "epoch": 2.912961210974456, + "grad_norm": 1.53125, + "learning_rate": 0.0002, + "loss": 1.6486, + "step": 9237 + }, + { + "epoch": 2.913276568905708, + "grad_norm": 1.234375, + "learning_rate": 0.0002, + "loss": 1.4942, + "step": 9238 + }, + { + "epoch": 2.91359192683696, + "grad_norm": 1.375, + "learning_rate": 0.0002, + "loss": 1.8481, + "step": 9239 + }, + { + "epoch": 2.9139072847682117, + "grad_norm": 1.328125, + "learning_rate": 0.0002, + "loss": 1.7406, + "step": 9240 + }, + { + "epoch": 2.914222642699464, + "grad_norm": 1.1328125, + "learning_rate": 0.0002, + "loss": 1.5874, + "step": 9241 + }, + { + "epoch": 2.9145380006307158, + "grad_norm": 1.3046875, + "learning_rate": 0.0002, + "loss": 1.6815, + "step": 9242 + }, + { + "epoch": 2.9148533585619676, + "grad_norm": 0.251953125, + "learning_rate": 0.0002, + "loss": 1.0118, + "step": 9243 + }, + { + "epoch": 2.91516871649322, + "grad_norm": 0.240234375, + "learning_rate": 0.0002, + "loss": 0.9145, + "step": 9244 + }, + { + "epoch": 2.9154840744244717, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 0.9035, + "step": 9245 + }, + { + "epoch": 2.9157994323557235, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 0.8951, + "step": 9246 + }, + { + "epoch": 2.9161147902869757, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 0.9083, + "step": 9247 + }, + { + "epoch": 2.9164301482182275, + "grad_norm": 0.294921875, + "learning_rate": 0.0002, + "loss": 0.9611, + "step": 9248 + }, + { + "epoch": 2.91674550614948, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 0.9879, + "step": 9249 + }, + { + "epoch": 2.9170608640807316, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 1.0168, + "step": 9250 + }, + { + "epoch": 2.917376222011984, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 0.9997, + "step": 9251 + }, + { + "epoch": 2.9176915799432357, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 0.913, + "step": 9252 + }, + { + "epoch": 2.9180069378744875, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 0.9985, + "step": 9253 + }, + { + "epoch": 2.9183222958057398, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 0.9503, + "step": 9254 + }, + { + "epoch": 2.9186376537369916, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 1.0742, + "step": 9255 + }, + { + "epoch": 2.9189530116682434, + "grad_norm": 0.384765625, + "learning_rate": 0.0002, + "loss": 0.9663, + "step": 9256 + }, + { + "epoch": 2.9192683695994956, + "grad_norm": 0.34765625, + "learning_rate": 0.0002, + "loss": 0.8125, + "step": 9257 + }, + { + "epoch": 2.9195837275307475, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 0.8229, + "step": 9258 + }, + { + "epoch": 2.9198990854619993, + "grad_norm": 0.40234375, + "learning_rate": 0.0002, + "loss": 1.0254, + "step": 9259 + }, + { + "epoch": 2.9202144433932515, + "grad_norm": 0.4765625, + "learning_rate": 0.0002, + "loss": 0.93, + "step": 9260 + }, + { + "epoch": 2.9205298013245033, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 1.0844, + "step": 9261 + }, + { + "epoch": 2.920845159255755, + "grad_norm": 0.451171875, + "learning_rate": 0.0002, + "loss": 0.9117, + "step": 9262 + }, + { + "epoch": 2.9211605171870074, + "grad_norm": 0.427734375, + "learning_rate": 0.0002, + "loss": 0.8144, + "step": 9263 + }, + { + "epoch": 2.9214758751182592, + "grad_norm": 0.58203125, + "learning_rate": 0.0002, + "loss": 1.0396, + "step": 9264 + }, + { + "epoch": 2.921791233049511, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 0.949, + "step": 9265 + }, + { + "epoch": 2.9221065909807633, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 0.8852, + "step": 9266 + }, + { + "epoch": 2.922421948912015, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 1.0405, + "step": 9267 + }, + { + "epoch": 2.922737306843267, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 0.8932, + "step": 9268 + }, + { + "epoch": 2.923052664774519, + "grad_norm": 0.58203125, + "learning_rate": 0.0002, + "loss": 1.0485, + "step": 9269 + }, + { + "epoch": 2.923368022705771, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 1.0364, + "step": 9270 + }, + { + "epoch": 2.923683380637023, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 0.8266, + "step": 9271 + }, + { + "epoch": 2.923998738568275, + "grad_norm": 0.67578125, + "learning_rate": 0.0002, + "loss": 0.924, + "step": 9272 + }, + { + "epoch": 2.924314096499527, + "grad_norm": 0.6875, + "learning_rate": 0.0002, + "loss": 0.8857, + "step": 9273 + }, + { + "epoch": 2.9246294544307787, + "grad_norm": 0.78125, + "learning_rate": 0.0002, + "loss": 1.0523, + "step": 9274 + }, + { + "epoch": 2.924944812362031, + "grad_norm": 0.79296875, + "learning_rate": 0.0002, + "loss": 1.1797, + "step": 9275 + }, + { + "epoch": 2.925260170293283, + "grad_norm": 0.88671875, + "learning_rate": 0.0002, + "loss": 1.141, + "step": 9276 + }, + { + "epoch": 2.9255755282245346, + "grad_norm": 0.95703125, + "learning_rate": 0.0002, + "loss": 1.2035, + "step": 9277 + }, + { + "epoch": 2.925890886155787, + "grad_norm": 1.09375, + "learning_rate": 0.0002, + "loss": 1.2603, + "step": 9278 + }, + { + "epoch": 2.9262062440870387, + "grad_norm": 1.3359375, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 9279 + }, + { + "epoch": 2.926521602018291, + "grad_norm": 1.078125, + "learning_rate": 0.0002, + "loss": 1.4533, + "step": 9280 + }, + { + "epoch": 2.9268369599495427, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.3676, + "step": 9281 + }, + { + "epoch": 2.9271523178807946, + "grad_norm": 1.2109375, + "learning_rate": 0.0002, + "loss": 1.676, + "step": 9282 + }, + { + "epoch": 2.927467675812047, + "grad_norm": 1.359375, + "learning_rate": 0.0002, + "loss": 1.6191, + "step": 9283 + }, + { + "epoch": 2.9277830337432986, + "grad_norm": 1.0703125, + "learning_rate": 0.0002, + "loss": 1.4751, + "step": 9284 + }, + { + "epoch": 2.928098391674551, + "grad_norm": 1.796875, + "learning_rate": 0.0002, + "loss": 1.8171, + "step": 9285 + }, + { + "epoch": 2.9284137496058027, + "grad_norm": 1.6953125, + "learning_rate": 0.0002, + "loss": 1.4691, + "step": 9286 + }, + { + "epoch": 2.9287291075370545, + "grad_norm": 1.3046875, + "learning_rate": 0.0002, + "loss": 1.5323, + "step": 9287 + }, + { + "epoch": 2.9290444654683068, + "grad_norm": 1.1640625, + "learning_rate": 0.0002, + "loss": 1.6528, + "step": 9288 + }, + { + "epoch": 2.9293598233995586, + "grad_norm": 1.234375, + "learning_rate": 0.0002, + "loss": 1.5795, + "step": 9289 + }, + { + "epoch": 2.9296751813308104, + "grad_norm": 1.3515625, + "learning_rate": 0.0002, + "loss": 1.706, + "step": 9290 + }, + { + "epoch": 2.9299905392620627, + "grad_norm": 1.640625, + "learning_rate": 0.0002, + "loss": 1.5317, + "step": 9291 + }, + { + "epoch": 2.9303058971933145, + "grad_norm": 1.5, + "learning_rate": 0.0002, + "loss": 1.6231, + "step": 9292 + }, + { + "epoch": 2.9306212551245663, + "grad_norm": 0.25, + "learning_rate": 0.0002, + "loss": 1.0414, + "step": 9293 + }, + { + "epoch": 2.9309366130558185, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 1.001, + "step": 9294 + }, + { + "epoch": 2.9312519709870704, + "grad_norm": 0.28515625, + "learning_rate": 0.0002, + "loss": 0.893, + "step": 9295 + }, + { + "epoch": 2.931567328918322, + "grad_norm": 0.287109375, + "learning_rate": 0.0002, + "loss": 0.9385, + "step": 9296 + }, + { + "epoch": 2.9318826868495744, + "grad_norm": 0.291015625, + "learning_rate": 0.0002, + "loss": 0.9436, + "step": 9297 + }, + { + "epoch": 2.9321980447808262, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 0.8855, + "step": 9298 + }, + { + "epoch": 2.932513402712078, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 0.9777, + "step": 9299 + }, + { + "epoch": 2.9328287606433303, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 0.9323, + "step": 9300 + }, + { + "epoch": 2.933144118574582, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 0.9104, + "step": 9301 + }, + { + "epoch": 2.933459476505834, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 0.8257, + "step": 9302 + }, + { + "epoch": 2.933774834437086, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 0.9413, + "step": 9303 + }, + { + "epoch": 2.934090192368338, + "grad_norm": 0.38671875, + "learning_rate": 0.0002, + "loss": 1.0291, + "step": 9304 + }, + { + "epoch": 2.93440555029959, + "grad_norm": 0.369140625, + "learning_rate": 0.0002, + "loss": 0.9101, + "step": 9305 + }, + { + "epoch": 2.934720908230842, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 1.0011, + "step": 9306 + }, + { + "epoch": 2.935036266162094, + "grad_norm": 0.40234375, + "learning_rate": 0.0002, + "loss": 1.0202, + "step": 9307 + }, + { + "epoch": 2.9353516240933457, + "grad_norm": 0.47265625, + "learning_rate": 0.0002, + "loss": 0.8333, + "step": 9308 + }, + { + "epoch": 2.935666982024598, + "grad_norm": 0.412109375, + "learning_rate": 0.0002, + "loss": 0.967, + "step": 9309 + }, + { + "epoch": 2.93598233995585, + "grad_norm": 0.44140625, + "learning_rate": 0.0002, + "loss": 0.9451, + "step": 9310 + }, + { + "epoch": 2.9362976978871016, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 1.0117, + "step": 9311 + }, + { + "epoch": 2.936613055818354, + "grad_norm": 0.498046875, + "learning_rate": 0.0002, + "loss": 1.0041, + "step": 9312 + }, + { + "epoch": 2.9369284137496057, + "grad_norm": 0.486328125, + "learning_rate": 0.0002, + "loss": 0.8895, + "step": 9313 + }, + { + "epoch": 2.937243771680858, + "grad_norm": 0.50390625, + "learning_rate": 0.0002, + "loss": 0.9342, + "step": 9314 + }, + { + "epoch": 2.9375591296121097, + "grad_norm": 0.5, + "learning_rate": 0.0002, + "loss": 0.8431, + "step": 9315 + }, + { + "epoch": 2.9378744875433616, + "grad_norm": 0.546875, + "learning_rate": 0.0002, + "loss": 0.8835, + "step": 9316 + }, + { + "epoch": 2.938189845474614, + "grad_norm": 0.52734375, + "learning_rate": 0.0002, + "loss": 0.8991, + "step": 9317 + }, + { + "epoch": 2.9385052034058656, + "grad_norm": 0.703125, + "learning_rate": 0.0002, + "loss": 0.979, + "step": 9318 + }, + { + "epoch": 2.938820561337118, + "grad_norm": 0.734375, + "learning_rate": 0.0002, + "loss": 0.8608, + "step": 9319 + }, + { + "epoch": 2.9391359192683697, + "grad_norm": 0.6875, + "learning_rate": 0.0002, + "loss": 0.9083, + "step": 9320 + }, + { + "epoch": 2.9394512771996215, + "grad_norm": 0.75390625, + "learning_rate": 0.0002, + "loss": 0.9914, + "step": 9321 + }, + { + "epoch": 2.9397666351308738, + "grad_norm": 0.8515625, + "learning_rate": 0.0002, + "loss": 1.1488, + "step": 9322 + }, + { + "epoch": 2.9400819930621256, + "grad_norm": 1.4765625, + "learning_rate": 0.0002, + "loss": 1.2908, + "step": 9323 + }, + { + "epoch": 2.9403973509933774, + "grad_norm": 0.81640625, + "learning_rate": 0.0002, + "loss": 1.1441, + "step": 9324 + }, + { + "epoch": 2.9407127089246297, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 1.0205, + "step": 9325 + }, + { + "epoch": 2.9410280668558815, + "grad_norm": 0.84765625, + "learning_rate": 0.0002, + "loss": 1.0605, + "step": 9326 + }, + { + "epoch": 2.9413434247871333, + "grad_norm": 1.1640625, + "learning_rate": 0.0002, + "loss": 1.2129, + "step": 9327 + }, + { + "epoch": 2.9416587827183855, + "grad_norm": 0.828125, + "learning_rate": 0.0002, + "loss": 0.9797, + "step": 9328 + }, + { + "epoch": 2.9419741406496374, + "grad_norm": 0.98828125, + "learning_rate": 0.0002, + "loss": 1.2913, + "step": 9329 + }, + { + "epoch": 2.942289498580889, + "grad_norm": 1.1796875, + "learning_rate": 0.0002, + "loss": 1.4266, + "step": 9330 + }, + { + "epoch": 2.9426048565121414, + "grad_norm": 1.0, + "learning_rate": 0.0002, + "loss": 1.2763, + "step": 9331 + }, + { + "epoch": 2.9429202144433932, + "grad_norm": 1.296875, + "learning_rate": 0.0002, + "loss": 1.4808, + "step": 9332 + }, + { + "epoch": 2.943235572374645, + "grad_norm": 1.28125, + "learning_rate": 0.0002, + "loss": 1.5636, + "step": 9333 + }, + { + "epoch": 2.9435509303058973, + "grad_norm": 1.2734375, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 9334 + }, + { + "epoch": 2.943866288237149, + "grad_norm": 1.375, + "learning_rate": 0.0002, + "loss": 1.4927, + "step": 9335 + }, + { + "epoch": 2.944181646168401, + "grad_norm": 1.3203125, + "learning_rate": 0.0002, + "loss": 1.672, + "step": 9336 + }, + { + "epoch": 2.944497004099653, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 1.8805, + "step": 9337 + }, + { + "epoch": 2.944812362030905, + "grad_norm": 1.1875, + "learning_rate": 0.0002, + "loss": 1.5614, + "step": 9338 + }, + { + "epoch": 2.945127719962157, + "grad_norm": 2.078125, + "learning_rate": 0.0002, + "loss": 1.4769, + "step": 9339 + }, + { + "epoch": 2.945443077893409, + "grad_norm": 1.1875, + "learning_rate": 0.0002, + "loss": 1.5824, + "step": 9340 + }, + { + "epoch": 2.945758435824661, + "grad_norm": 1.3828125, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 9341 + }, + { + "epoch": 2.9460737937559127, + "grad_norm": 1.3125, + "learning_rate": 0.0002, + "loss": 1.8198, + "step": 9342 + }, + { + "epoch": 2.946389151687165, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 1.0401, + "step": 9343 + }, + { + "epoch": 2.946704509618417, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 0.9061, + "step": 9344 + }, + { + "epoch": 2.9470198675496686, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 0.99, + "step": 9345 + }, + { + "epoch": 2.947335225480921, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.0261, + "step": 9346 + }, + { + "epoch": 2.9476505834121727, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 0.8755, + "step": 9347 + }, + { + "epoch": 2.947965941343425, + "grad_norm": 0.431640625, + "learning_rate": 0.0002, + "loss": 0.9962, + "step": 9348 + }, + { + "epoch": 2.9482812992746767, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 0.9763, + "step": 9349 + }, + { + "epoch": 2.948596657205929, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 0.8681, + "step": 9350 + }, + { + "epoch": 2.948596657205929, + "eval_loss": 1.7972837686538696, + "eval_runtime": 149.0413, + "eval_samples_per_second": 6.71, + "eval_steps_per_second": 6.71, + "step": 9350 + }, + { + "epoch": 2.948596657205929, + "mmlu_eval_accuracy": 0.5949288133748815, + "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, + "mmlu_eval_accuracy_anatomy": 0.5, + "mmlu_eval_accuracy_astronomy": 0.625, + "mmlu_eval_accuracy_business_ethics": 0.8181818181818182, + "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138, + "mmlu_eval_accuracy_college_biology": 0.8125, + "mmlu_eval_accuracy_college_chemistry": 0.25, + "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727, + "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454, + "mmlu_eval_accuracy_college_medicine": 0.5, + "mmlu_eval_accuracy_college_physics": 0.45454545454545453, + "mmlu_eval_accuracy_computer_security": 0.5454545454545454, + "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384, + "mmlu_eval_accuracy_econometrics": 0.4166666666666667, + "mmlu_eval_accuracy_electrical_engineering": 0.5625, + "mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637, + "mmlu_eval_accuracy_formal_logic": 0.35714285714285715, + "mmlu_eval_accuracy_global_facts": 0.3, + "mmlu_eval_accuracy_high_school_biology": 0.6875, + "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222, + "mmlu_eval_accuracy_high_school_geography": 0.8636363636363636, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8571428571428571, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209, + "mmlu_eval_accuracy_high_school_mathematics": 0.3448275862068966, + "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154, + "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354, + "mmlu_eval_accuracy_high_school_psychology": 0.85, + "mmlu_eval_accuracy_high_school_statistics": 0.4782608695652174, + "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, + "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923, + "mmlu_eval_accuracy_human_aging": 0.6521739130434783, + "mmlu_eval_accuracy_human_sexuality": 0.5, + "mmlu_eval_accuracy_international_law": 1.0, + "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, + "mmlu_eval_accuracy_logical_fallacies": 0.7777777777777778, + "mmlu_eval_accuracy_machine_learning": 0.2727272727272727, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.76, + "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182, + "mmlu_eval_accuracy_miscellaneous": 0.6744186046511628, + "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685, + "mmlu_eval_accuracy_moral_scenarios": 0.29, + "mmlu_eval_accuracy_nutrition": 0.7575757575757576, + "mmlu_eval_accuracy_philosophy": 0.6176470588235294, + "mmlu_eval_accuracy_prehistory": 0.6, + "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644, + "mmlu_eval_accuracy_professional_law": 0.40588235294117647, + "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806, + "mmlu_eval_accuracy_professional_psychology": 0.6376811594202898, + "mmlu_eval_accuracy_public_relations": 0.5833333333333334, + "mmlu_eval_accuracy_security_studies": 0.8518518518518519, + "mmlu_eval_accuracy_sociology": 0.9545454545454546, + "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, + "mmlu_eval_accuracy_virology": 0.5, + "mmlu_eval_accuracy_world_religions": 0.7894736842105263, + "mmlu_loss": 6.070316658231809, + "step": 9350 + }, + { + "epoch": 2.948912015137181, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 0.892, + "step": 9351 + }, + { + "epoch": 2.9492273730684326, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 0.9315, + "step": 9352 + }, + { + "epoch": 2.949542730999685, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 0.9683, + "step": 9353 + }, + { + "epoch": 2.9498580889309367, + "grad_norm": 0.349609375, + "learning_rate": 0.0002, + "loss": 0.9049, + "step": 9354 + }, + { + "epoch": 2.9501734468621885, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 0.9463, + "step": 9355 + }, + { + "epoch": 2.950488804793441, + "grad_norm": 0.330078125, + "learning_rate": 0.0002, + "loss": 0.9407, + "step": 9356 + }, + { + "epoch": 2.9508041627246926, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 0.9136, + "step": 9357 + }, + { + "epoch": 2.9511195206559444, + "grad_norm": 0.384765625, + "learning_rate": 0.0002, + "loss": 1.0027, + "step": 9358 + }, + { + "epoch": 2.9514348785871967, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 0.9762, + "step": 9359 + }, + { + "epoch": 2.9517502365184485, + "grad_norm": 0.380859375, + "learning_rate": 0.0002, + "loss": 0.7788, + "step": 9360 + }, + { + "epoch": 2.9520655944497003, + "grad_norm": 0.439453125, + "learning_rate": 0.0002, + "loss": 0.9355, + "step": 9361 + }, + { + "epoch": 2.9523809523809526, + "grad_norm": 0.400390625, + "learning_rate": 0.0002, + "loss": 0.9304, + "step": 9362 + }, + { + "epoch": 2.9526963103122044, + "grad_norm": 0.421875, + "learning_rate": 0.0002, + "loss": 0.8969, + "step": 9363 + }, + { + "epoch": 2.953011668243456, + "grad_norm": 0.474609375, + "learning_rate": 0.0002, + "loss": 0.9099, + "step": 9364 + }, + { + "epoch": 2.9533270261747084, + "grad_norm": 0.455078125, + "learning_rate": 0.0002, + "loss": 0.914, + "step": 9365 + }, + { + "epoch": 2.9536423841059603, + "grad_norm": 0.4765625, + "learning_rate": 0.0002, + "loss": 0.8358, + "step": 9366 + }, + { + "epoch": 2.953957742037212, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 0.9032, + "step": 9367 + }, + { + "epoch": 2.9542730999684643, + "grad_norm": 0.61328125, + "learning_rate": 0.0002, + "loss": 0.8768, + "step": 9368 + }, + { + "epoch": 2.954588457899716, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 0.7833, + "step": 9369 + }, + { + "epoch": 2.954903815830968, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 0.9484, + "step": 9370 + }, + { + "epoch": 2.95521917376222, + "grad_norm": 0.64453125, + "learning_rate": 0.0002, + "loss": 0.9892, + "step": 9371 + }, + { + "epoch": 2.955534531693472, + "grad_norm": 0.70703125, + "learning_rate": 0.0002, + "loss": 0.922, + "step": 9372 + }, + { + "epoch": 2.955849889624724, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 0.9177, + "step": 9373 + }, + { + "epoch": 2.956165247555976, + "grad_norm": 1.0, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 9374 + }, + { + "epoch": 2.956480605487228, + "grad_norm": 0.97265625, + "learning_rate": 0.0002, + "loss": 1.1663, + "step": 9375 + }, + { + "epoch": 2.9567959634184797, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 1.4151, + "step": 9376 + }, + { + "epoch": 2.957111321349732, + "grad_norm": 0.99609375, + "learning_rate": 0.0002, + "loss": 1.1749, + "step": 9377 + }, + { + "epoch": 2.957426679280984, + "grad_norm": 1.6484375, + "learning_rate": 0.0002, + "loss": 1.3788, + "step": 9378 + }, + { + "epoch": 2.957742037212236, + "grad_norm": 1.109375, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 9379 + }, + { + "epoch": 2.958057395143488, + "grad_norm": 1.3125, + "learning_rate": 0.0002, + "loss": 1.4747, + "step": 9380 + }, + { + "epoch": 2.9583727530747397, + "grad_norm": 1.3046875, + "learning_rate": 0.0002, + "loss": 1.7033, + "step": 9381 + }, + { + "epoch": 2.958688111005992, + "grad_norm": 1.453125, + "learning_rate": 0.0002, + "loss": 1.7084, + "step": 9382 + }, + { + "epoch": 2.9590034689372438, + "grad_norm": 1.4296875, + "learning_rate": 0.0002, + "loss": 1.6355, + "step": 9383 + }, + { + "epoch": 2.959318826868496, + "grad_norm": 1.296875, + "learning_rate": 0.0002, + "loss": 1.4644, + "step": 9384 + }, + { + "epoch": 2.959634184799748, + "grad_norm": 1.2734375, + "learning_rate": 0.0002, + "loss": 1.8401, + "step": 9385 + }, + { + "epoch": 2.9599495427309996, + "grad_norm": 1.25, + "learning_rate": 0.0002, + "loss": 1.3844, + "step": 9386 + }, + { + "epoch": 2.960264900662252, + "grad_norm": 1.40625, + "learning_rate": 0.0002, + "loss": 1.6307, + "step": 9387 + }, + { + "epoch": 2.9605802585935037, + "grad_norm": 1.21875, + "learning_rate": 0.0002, + "loss": 1.8325, + "step": 9388 + }, + { + "epoch": 2.9608956165247555, + "grad_norm": 1.4453125, + "learning_rate": 0.0002, + "loss": 1.8135, + "step": 9389 + }, + { + "epoch": 2.961210974456008, + "grad_norm": 1.3046875, + "learning_rate": 0.0002, + "loss": 1.8276, + "step": 9390 + }, + { + "epoch": 2.9615263323872596, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.5501, + "step": 9391 + }, + { + "epoch": 2.9618416903185114, + "grad_norm": 1.53125, + "learning_rate": 0.0002, + "loss": 1.8608, + "step": 9392 + }, + { + "epoch": 2.9621570482497637, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.0367, + "step": 9393 + }, + { + "epoch": 2.9624724061810155, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 0.9208, + "step": 9394 + }, + { + "epoch": 2.9627877641122673, + "grad_norm": 0.2734375, + "learning_rate": 0.0002, + "loss": 0.9349, + "step": 9395 + }, + { + "epoch": 2.9631031220435196, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 0.9004, + "step": 9396 + }, + { + "epoch": 2.9634184799747714, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 1.0183, + "step": 9397 + }, + { + "epoch": 2.963733837906023, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 0.9062, + "step": 9398 + }, + { + "epoch": 2.9640491958372754, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 0.9949, + "step": 9399 + }, + { + "epoch": 2.9643645537685273, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 0.998, + "step": 9400 + }, + { + "epoch": 2.964679911699779, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 0.9954, + "step": 9401 + }, + { + "epoch": 2.9649952696310313, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 0.8578, + "step": 9402 + }, + { + "epoch": 2.965310627562283, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 0.8917, + "step": 9403 + }, + { + "epoch": 2.965625985493535, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 0.8324, + "step": 9404 + }, + { + "epoch": 2.965941343424787, + "grad_norm": 0.349609375, + "learning_rate": 0.0002, + "loss": 0.8952, + "step": 9405 + }, + { + "epoch": 2.966256701356039, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 0.8728, + "step": 9406 + }, + { + "epoch": 2.966572059287291, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 0.9258, + "step": 9407 + }, + { + "epoch": 2.966887417218543, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 0.8735, + "step": 9408 + }, + { + "epoch": 2.967202775149795, + "grad_norm": 0.396484375, + "learning_rate": 0.0002, + "loss": 0.9931, + "step": 9409 + }, + { + "epoch": 2.9675181330810467, + "grad_norm": 0.431640625, + "learning_rate": 0.0002, + "loss": 0.9497, + "step": 9410 + }, + { + "epoch": 2.967833491012299, + "grad_norm": 0.400390625, + "learning_rate": 0.0002, + "loss": 0.9382, + "step": 9411 + }, + { + "epoch": 2.968148848943551, + "grad_norm": 0.416015625, + "learning_rate": 0.0002, + "loss": 0.9097, + "step": 9412 + }, + { + "epoch": 2.968464206874803, + "grad_norm": 0.443359375, + "learning_rate": 0.0002, + "loss": 0.8216, + "step": 9413 + }, + { + "epoch": 2.968779564806055, + "grad_norm": 0.435546875, + "learning_rate": 0.0002, + "loss": 0.8882, + "step": 9414 + }, + { + "epoch": 2.9690949227373067, + "grad_norm": 0.4765625, + "learning_rate": 0.0002, + "loss": 0.9522, + "step": 9415 + }, + { + "epoch": 2.969410280668559, + "grad_norm": 0.5, + "learning_rate": 0.0002, + "loss": 0.8577, + "step": 9416 + }, + { + "epoch": 2.9697256385998108, + "grad_norm": 0.5390625, + "learning_rate": 0.0002, + "loss": 0.9377, + "step": 9417 + }, + { + "epoch": 2.970040996531063, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 0.921, + "step": 9418 + }, + { + "epoch": 2.970356354462315, + "grad_norm": 0.625, + "learning_rate": 0.0002, + "loss": 1.0381, + "step": 9419 + }, + { + "epoch": 2.9706717123935666, + "grad_norm": 0.55859375, + "learning_rate": 0.0002, + "loss": 0.8954, + "step": 9420 + }, + { + "epoch": 2.970987070324819, + "grad_norm": 0.62890625, + "learning_rate": 0.0002, + "loss": 0.979, + "step": 9421 + }, + { + "epoch": 2.9713024282560707, + "grad_norm": 0.703125, + "learning_rate": 0.0002, + "loss": 0.995, + "step": 9422 + }, + { + "epoch": 2.9716177861873225, + "grad_norm": 0.6953125, + "learning_rate": 0.0002, + "loss": 0.9174, + "step": 9423 + }, + { + "epoch": 2.971933144118575, + "grad_norm": 0.87890625, + "learning_rate": 0.0002, + "loss": 1.0909, + "step": 9424 + }, + { + "epoch": 2.9722485020498266, + "grad_norm": 0.91796875, + "learning_rate": 0.0002, + "loss": 1.139, + "step": 9425 + }, + { + "epoch": 2.9725638599810784, + "grad_norm": 1.078125, + "learning_rate": 0.0002, + "loss": 1.2276, + "step": 9426 + }, + { + "epoch": 2.9728792179123307, + "grad_norm": 0.91796875, + "learning_rate": 0.0002, + "loss": 1.163, + "step": 9427 + }, + { + "epoch": 2.9731945758435825, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 1.0379, + "step": 9428 + }, + { + "epoch": 2.9735099337748343, + "grad_norm": 0.87890625, + "learning_rate": 0.0002, + "loss": 1.1338, + "step": 9429 + }, + { + "epoch": 2.9738252917060866, + "grad_norm": 1.015625, + "learning_rate": 0.0002, + "loss": 1.2574, + "step": 9430 + }, + { + "epoch": 2.9741406496373384, + "grad_norm": 1.03125, + "learning_rate": 0.0002, + "loss": 1.2248, + "step": 9431 + }, + { + "epoch": 2.97445600756859, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 9432 + }, + { + "epoch": 2.9747713654998424, + "grad_norm": 1.078125, + "learning_rate": 0.0002, + "loss": 1.5955, + "step": 9433 + }, + { + "epoch": 2.9750867234310943, + "grad_norm": 1.0546875, + "learning_rate": 0.0002, + "loss": 1.3785, + "step": 9434 + }, + { + "epoch": 2.975402081362346, + "grad_norm": 1.3125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9435 + }, + { + "epoch": 2.9757174392935983, + "grad_norm": 0.9921875, + "learning_rate": 0.0002, + "loss": 1.6206, + "step": 9436 + }, + { + "epoch": 2.97603279722485, + "grad_norm": 1.578125, + "learning_rate": 0.0002, + "loss": 1.8288, + "step": 9437 + }, + { + "epoch": 2.976348155156102, + "grad_norm": 1.28125, + "learning_rate": 0.0002, + "loss": 1.7441, + "step": 9438 + }, + { + "epoch": 2.976663513087354, + "grad_norm": 1.1953125, + "learning_rate": 0.0002, + "loss": 1.6316, + "step": 9439 + }, + { + "epoch": 2.976978871018606, + "grad_norm": 1.2109375, + "learning_rate": 0.0002, + "loss": 1.3642, + "step": 9440 + }, + { + "epoch": 2.977294228949858, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 1.4589, + "step": 9441 + }, + { + "epoch": 2.97760958688111, + "grad_norm": 0.953125, + "learning_rate": 0.0002, + "loss": 1.4455, + "step": 9442 + }, + { + "epoch": 2.977924944812362, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.0056, + "step": 9443 + }, + { + "epoch": 2.9782403027436137, + "grad_norm": 0.2490234375, + "learning_rate": 0.0002, + "loss": 0.8871, + "step": 9444 + }, + { + "epoch": 2.978555660674866, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 1.0214, + "step": 9445 + }, + { + "epoch": 2.978871018606118, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 0.8691, + "step": 9446 + }, + { + "epoch": 2.97918637653737, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 1.0863, + "step": 9447 + }, + { + "epoch": 2.979501734468622, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 0.9266, + "step": 9448 + }, + { + "epoch": 2.9798170923998737, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 0.935, + "step": 9449 + }, + { + "epoch": 2.980132450331126, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 0.9675, + "step": 9450 + }, + { + "epoch": 2.9804478082623778, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 0.9543, + "step": 9451 + }, + { + "epoch": 2.98076316619363, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 0.9309, + "step": 9452 + }, + { + "epoch": 2.981078524124882, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 0.9653, + "step": 9453 + }, + { + "epoch": 2.9813938820561336, + "grad_norm": 0.310546875, + "learning_rate": 0.0002, + "loss": 0.9005, + "step": 9454 + }, + { + "epoch": 2.981709239987386, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 0.9573, + "step": 9455 + }, + { + "epoch": 2.9820245979186377, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 0.9047, + "step": 9456 + }, + { + "epoch": 2.9823399558498895, + "grad_norm": 0.353515625, + "learning_rate": 0.0002, + "loss": 0.8914, + "step": 9457 + }, + { + "epoch": 2.982655313781142, + "grad_norm": 0.4375, + "learning_rate": 0.0002, + "loss": 0.9994, + "step": 9458 + }, + { + "epoch": 2.9829706717123936, + "grad_norm": 0.439453125, + "learning_rate": 0.0002, + "loss": 0.8799, + "step": 9459 + }, + { + "epoch": 2.9832860296436454, + "grad_norm": 0.44140625, + "learning_rate": 0.0002, + "loss": 1.1319, + "step": 9460 + }, + { + "epoch": 2.9836013875748977, + "grad_norm": 0.474609375, + "learning_rate": 0.0002, + "loss": 1.0024, + "step": 9461 + }, + { + "epoch": 2.9839167455061495, + "grad_norm": 0.443359375, + "learning_rate": 0.0002, + "loss": 0.9617, + "step": 9462 + }, + { + "epoch": 2.9842321034374013, + "grad_norm": 0.41796875, + "learning_rate": 0.0002, + "loss": 0.8228, + "step": 9463 + }, + { + "epoch": 2.9845474613686536, + "grad_norm": 0.486328125, + "learning_rate": 0.0002, + "loss": 0.9423, + "step": 9464 + }, + { + "epoch": 2.9848628192999054, + "grad_norm": 0.515625, + "learning_rate": 0.0002, + "loss": 0.7962, + "step": 9465 + }, + { + "epoch": 2.985178177231157, + "grad_norm": 0.55859375, + "learning_rate": 0.0002, + "loss": 1.1461, + "step": 9466 + }, + { + "epoch": 2.9854935351624095, + "grad_norm": 0.58984375, + "learning_rate": 0.0002, + "loss": 0.9428, + "step": 9467 + }, + { + "epoch": 2.9858088930936613, + "grad_norm": 0.578125, + "learning_rate": 0.0002, + "loss": 0.9771, + "step": 9468 + }, + { + "epoch": 2.986124251024913, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 0.8036, + "step": 9469 + }, + { + "epoch": 2.9864396089561653, + "grad_norm": 0.74609375, + "learning_rate": 0.0002, + "loss": 1.1844, + "step": 9470 + }, + { + "epoch": 2.986754966887417, + "grad_norm": 0.62890625, + "learning_rate": 0.0002, + "loss": 0.9559, + "step": 9471 + }, + { + "epoch": 2.987070324818669, + "grad_norm": 0.68359375, + "learning_rate": 0.0002, + "loss": 0.9953, + "step": 9472 + }, + { + "epoch": 2.9873856827499212, + "grad_norm": 0.6875, + "learning_rate": 0.0002, + "loss": 1.0774, + "step": 9473 + }, + { + "epoch": 2.987701040681173, + "grad_norm": 0.90625, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 9474 + }, + { + "epoch": 2.988016398612425, + "grad_norm": 0.80859375, + "learning_rate": 0.0002, + "loss": 1.0036, + "step": 9475 + }, + { + "epoch": 2.988331756543677, + "grad_norm": 0.7578125, + "learning_rate": 0.0002, + "loss": 1.0525, + "step": 9476 + }, + { + "epoch": 2.988647114474929, + "grad_norm": 1.2265625, + "learning_rate": 0.0002, + "loss": 1.5241, + "step": 9477 + }, + { + "epoch": 2.988962472406181, + "grad_norm": 1.2421875, + "learning_rate": 0.0002, + "loss": 1.2366, + "step": 9478 + }, + { + "epoch": 2.989277830337433, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 9479 + }, + { + "epoch": 2.989593188268685, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 1.4264, + "step": 9480 + }, + { + "epoch": 2.989908546199937, + "grad_norm": 1.625, + "learning_rate": 0.0002, + "loss": 1.5929, + "step": 9481 + }, + { + "epoch": 2.990223904131189, + "grad_norm": 1.2109375, + "learning_rate": 0.0002, + "loss": 1.4773, + "step": 9482 + }, + { + "epoch": 2.990539262062441, + "grad_norm": 1.1328125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 9483 + }, + { + "epoch": 2.990854619993693, + "grad_norm": 1.359375, + "learning_rate": 0.0002, + "loss": 1.7195, + "step": 9484 + }, + { + "epoch": 2.9911699779249448, + "grad_norm": 1.3671875, + "learning_rate": 0.0002, + "loss": 1.6849, + "step": 9485 + }, + { + "epoch": 2.991485335856197, + "grad_norm": 1.6328125, + "learning_rate": 0.0002, + "loss": 1.7355, + "step": 9486 + }, + { + "epoch": 2.991800693787449, + "grad_norm": 2.109375, + "learning_rate": 0.0002, + "loss": 1.9676, + "step": 9487 + }, + { + "epoch": 2.9921160517187007, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 1.8766, + "step": 9488 + }, + { + "epoch": 2.992431409649953, + "grad_norm": 1.3046875, + "learning_rate": 0.0002, + "loss": 1.8349, + "step": 9489 + }, + { + "epoch": 2.9927467675812047, + "grad_norm": 1.125, + "learning_rate": 0.0002, + "loss": 1.5915, + "step": 9490 + }, + { + "epoch": 2.9930621255124565, + "grad_norm": 1.1796875, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 9491 + }, + { + "epoch": 2.993377483443709, + "grad_norm": 1.46875, + "learning_rate": 0.0002, + "loss": 1.8276, + "step": 9492 + }, + { + "epoch": 2.9936928413749606, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 1.033, + "step": 9493 + }, + { + "epoch": 2.9940081993062124, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 1.0032, + "step": 9494 + }, + { + "epoch": 2.9943235572374647, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 1.1386, + "step": 9495 + }, + { + "epoch": 2.9946389151687165, + "grad_norm": 0.373046875, + "learning_rate": 0.0002, + "loss": 0.9866, + "step": 9496 + }, + { + "epoch": 2.9949542730999683, + "grad_norm": 0.353515625, + "learning_rate": 0.0002, + "loss": 0.8792, + "step": 9497 + }, + { + "epoch": 2.9952696310312206, + "grad_norm": 0.41796875, + "learning_rate": 0.0002, + "loss": 0.8932, + "step": 9498 + }, + { + "epoch": 2.9955849889624724, + "grad_norm": 0.5, + "learning_rate": 0.0002, + "loss": 0.9567, + "step": 9499 + }, + { + "epoch": 2.995900346893724, + "grad_norm": 0.58203125, + "learning_rate": 0.0002, + "loss": 0.9297, + "step": 9500 + }, + { + "epoch": 2.9962157048249765, + "grad_norm": 0.5390625, + "learning_rate": 0.0002, + "loss": 0.9026, + "step": 9501 + }, + { + "epoch": 2.9965310627562283, + "grad_norm": 0.79296875, + "learning_rate": 0.0002, + "loss": 1.0527, + "step": 9502 + }, + { + "epoch": 2.99684642068748, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 1.0265, + "step": 9503 + }, + { + "epoch": 2.9971617786187323, + "grad_norm": 0.66796875, + "learning_rate": 0.0002, + "loss": 0.9072, + "step": 9504 + }, + { + "epoch": 2.997477136549984, + "grad_norm": 0.94140625, + "learning_rate": 0.0002, + "loss": 1.0256, + "step": 9505 + }, + { + "epoch": 2.997792494481236, + "grad_norm": 0.87109375, + "learning_rate": 0.0002, + "loss": 1.031, + "step": 9506 + }, + { + "epoch": 2.9981078524124882, + "grad_norm": 0.90625, + "learning_rate": 0.0002, + "loss": 1.1539, + "step": 9507 + }, + { + "epoch": 2.99842321034374, + "grad_norm": 1.0390625, + "learning_rate": 0.0002, + "loss": 1.287, + "step": 9508 + }, + { + "epoch": 2.998738568274992, + "grad_norm": 1.375, + "learning_rate": 0.0002, + "loss": 1.614, + "step": 9509 + }, + { + "epoch": 2.999053926206244, + "grad_norm": 1.1953125, + "learning_rate": 0.0002, + "loss": 1.6854, + "step": 9510 + }, + { + "epoch": 2.999369284137496, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 1.4829, + "step": 9511 + }, + { + "epoch": 2.999684642068748, + "grad_norm": 1.0390625, + "learning_rate": 0.0002, + "loss": 1.2467, + "step": 9512 + }, + { + "epoch": 3.0, + "grad_norm": 1.5390625, + "learning_rate": 0.0002, + "loss": 1.6112, + "step": 9513 + }, + { + "epoch": 3.000315357931252, + "grad_norm": 0.2392578125, + "learning_rate": 0.0002, + "loss": 0.8448, + "step": 9514 + }, + { + "epoch": 3.000630715862504, + "grad_norm": 0.220703125, + "learning_rate": 0.0002, + "loss": 0.8296, + "step": 9515 + }, + { + "epoch": 3.000946073793756, + "grad_norm": 0.240234375, + "learning_rate": 0.0002, + "loss": 0.8067, + "step": 9516 + }, + { + "epoch": 3.0012614317250077, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 0.7907, + "step": 9517 + }, + { + "epoch": 3.00157678965626, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 0.7859, + "step": 9518 + }, + { + "epoch": 3.0018921475875118, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 0.7171, + "step": 9519 + }, + { + "epoch": 3.0022075055187636, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 0.8391, + "step": 9520 + }, + { + "epoch": 3.002522863450016, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 0.8305, + "step": 9521 + }, + { + "epoch": 3.0028382213812677, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 0.7674, + "step": 9522 + }, + { + "epoch": 3.00315357931252, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 0.6278, + "step": 9523 + }, + { + "epoch": 3.0034689372437717, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 0.714, + "step": 9524 + }, + { + "epoch": 3.0037842951750235, + "grad_norm": 0.373046875, + "learning_rate": 0.0002, + "loss": 0.6332, + "step": 9525 + }, + { + "epoch": 3.004099653106276, + "grad_norm": 0.419921875, + "learning_rate": 0.0002, + "loss": 0.665, + "step": 9526 + }, + { + "epoch": 3.0044150110375276, + "grad_norm": 0.4453125, + "learning_rate": 0.0002, + "loss": 0.6908, + "step": 9527 + }, + { + "epoch": 3.0047303689687794, + "grad_norm": 0.47265625, + "learning_rate": 0.0002, + "loss": 0.6075, + "step": 9528 + }, + { + "epoch": 3.0050457269000317, + "grad_norm": 0.453125, + "learning_rate": 0.0002, + "loss": 0.6481, + "step": 9529 + }, + { + "epoch": 3.0053610848312835, + "grad_norm": 0.4921875, + "learning_rate": 0.0002, + "loss": 0.6024, + "step": 9530 + }, + { + "epoch": 3.0056764427625353, + "grad_norm": 0.578125, + "learning_rate": 0.0002, + "loss": 0.5915, + "step": 9531 + }, + { + "epoch": 3.0059918006937876, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 0.6119, + "step": 9532 + }, + { + "epoch": 3.0063071586250394, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 0.7536, + "step": 9533 + }, + { + "epoch": 3.006622516556291, + "grad_norm": 0.578125, + "learning_rate": 0.0002, + "loss": 0.619, + "step": 9534 + }, + { + "epoch": 3.0069378744875435, + "grad_norm": 0.60546875, + "learning_rate": 0.0002, + "loss": 0.6347, + "step": 9535 + }, + { + "epoch": 3.0072532324187953, + "grad_norm": 0.8671875, + "learning_rate": 0.0002, + "loss": 0.6806, + "step": 9536 + }, + { + "epoch": 3.007568590350047, + "grad_norm": 0.49609375, + "learning_rate": 0.0002, + "loss": 0.5482, + "step": 9537 + }, + { + "epoch": 3.007568590350047, + "eval_loss": 1.9380700588226318, + "eval_runtime": 148.9636, + "eval_samples_per_second": 6.713, + "eval_steps_per_second": 6.713, + "step": 9537 + }, + { + "epoch": 3.007568590350047, + "mmlu_eval_accuracy": 0.5868460544634867, + "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, + "mmlu_eval_accuracy_anatomy": 0.5, + "mmlu_eval_accuracy_astronomy": 0.625, + "mmlu_eval_accuracy_business_ethics": 0.9090909090909091, + "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138, + "mmlu_eval_accuracy_college_biology": 0.8125, + "mmlu_eval_accuracy_college_chemistry": 0.375, + "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727, + "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454, + "mmlu_eval_accuracy_college_medicine": 0.5454545454545454, + "mmlu_eval_accuracy_college_physics": 0.5454545454545454, + "mmlu_eval_accuracy_computer_security": 0.5454545454545454, + "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231, + "mmlu_eval_accuracy_econometrics": 0.5, + "mmlu_eval_accuracy_electrical_engineering": 0.5, + "mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536, + "mmlu_eval_accuracy_formal_logic": 0.2857142857142857, + "mmlu_eval_accuracy_global_facts": 0.3, + "mmlu_eval_accuracy_high_school_biology": 0.65625, + "mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8571428571428571, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.627906976744186, + "mmlu_eval_accuracy_high_school_mathematics": 0.3793103448275862, + "mmlu_eval_accuracy_high_school_microeconomics": 0.46153846153846156, + "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, + "mmlu_eval_accuracy_high_school_psychology": 0.8, + "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, + "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, + "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923, + "mmlu_eval_accuracy_human_aging": 0.6086956521739131, + "mmlu_eval_accuracy_human_sexuality": 0.5, + "mmlu_eval_accuracy_international_law": 0.9230769230769231, + "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454, + "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, + "mmlu_eval_accuracy_machine_learning": 0.36363636363636365, + "mmlu_eval_accuracy_management": 0.8181818181818182, + "mmlu_eval_accuracy_marketing": 0.76, + "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182, + "mmlu_eval_accuracy_miscellaneous": 0.686046511627907, + "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842, + "mmlu_eval_accuracy_moral_scenarios": 0.27, + "mmlu_eval_accuracy_nutrition": 0.7575757575757576, + "mmlu_eval_accuracy_philosophy": 0.6176470588235294, + "mmlu_eval_accuracy_prehistory": 0.5714285714285714, + "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355, + "mmlu_eval_accuracy_professional_law": 0.4, + "mmlu_eval_accuracy_professional_medicine": 0.7096774193548387, + "mmlu_eval_accuracy_professional_psychology": 0.6086956521739131, + "mmlu_eval_accuracy_public_relations": 0.4166666666666667, + "mmlu_eval_accuracy_security_studies": 0.7777777777777778, + "mmlu_eval_accuracy_sociology": 0.9545454545454546, + "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091, + "mmlu_eval_accuracy_virology": 0.4444444444444444, + "mmlu_eval_accuracy_world_religions": 0.7894736842105263, + "mmlu_loss": 6.243690664202929, + "step": 9537 + }, + { + "epoch": 3.0078839482812993, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 0.5907, + "step": 9538 + }, + { + "epoch": 3.008199306212551, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 0.5451, + "step": 9539 + }, + { + "epoch": 3.0085146641438034, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 0.6096, + "step": 9540 + }, + { + "epoch": 3.0088300220750552, + "grad_norm": 0.7890625, + "learning_rate": 0.0002, + "loss": 0.7012, + "step": 9541 + }, + { + "epoch": 3.009145380006307, + "grad_norm": 0.91015625, + "learning_rate": 0.0002, + "loss": 0.6838, + "step": 9542 + }, + { + "epoch": 3.0094607379375593, + "grad_norm": 0.8984375, + "learning_rate": 0.0002, + "loss": 0.5947, + "step": 9543 + }, + { + "epoch": 3.009776095868811, + "grad_norm": 0.9375, + "learning_rate": 0.0002, + "loss": 0.6854, + "step": 9544 + }, + { + "epoch": 3.010091453800063, + "grad_norm": 0.95703125, + "learning_rate": 0.0002, + "loss": 0.6928, + "step": 9545 + }, + { + "epoch": 3.010406811731315, + "grad_norm": 1.140625, + "learning_rate": 0.0002, + "loss": 0.7343, + "step": 9546 + }, + { + "epoch": 3.010722169662567, + "grad_norm": 0.83984375, + "learning_rate": 0.0002, + "loss": 0.7939, + "step": 9547 + }, + { + "epoch": 3.011037527593819, + "grad_norm": 1.234375, + "learning_rate": 0.0002, + "loss": 0.7817, + "step": 9548 + }, + { + "epoch": 3.011352885525071, + "grad_norm": 1.40625, + "learning_rate": 0.0002, + "loss": 0.8452, + "step": 9549 + }, + { + "epoch": 3.011668243456323, + "grad_norm": 1.4609375, + "learning_rate": 0.0002, + "loss": 0.944, + "step": 9550 + }, + { + "epoch": 3.0119836013875747, + "grad_norm": 1.4609375, + "learning_rate": 0.0002, + "loss": 0.9711, + "step": 9551 + }, + { + "epoch": 3.012298959318827, + "grad_norm": 1.6328125, + "learning_rate": 0.0002, + "loss": 1.2912, + "step": 9552 + }, + { + "epoch": 3.0126143172500788, + "grad_norm": 3.0, + "learning_rate": 0.0002, + "loss": 1.2456, + "step": 9553 + }, + { + "epoch": 3.0129296751813306, + "grad_norm": 1.5625, + "learning_rate": 0.0002, + "loss": 1.3001, + "step": 9554 + }, + { + "epoch": 3.013245033112583, + "grad_norm": 1.796875, + "learning_rate": 0.0002, + "loss": 1.1407, + "step": 9555 + }, + { + "epoch": 3.0135603910438347, + "grad_norm": 1.8046875, + "learning_rate": 0.0002, + "loss": 1.134, + "step": 9556 + }, + { + "epoch": 3.013875748975087, + "grad_norm": 1.3359375, + "learning_rate": 0.0002, + "loss": 1.4343, + "step": 9557 + }, + { + "epoch": 3.0141911069063387, + "grad_norm": 1.1875, + "learning_rate": 0.0002, + "loss": 1.488, + "step": 9558 + }, + { + "epoch": 3.0145064648375905, + "grad_norm": 1.703125, + "learning_rate": 0.0002, + "loss": 1.2861, + "step": 9559 + }, + { + "epoch": 3.014821822768843, + "grad_norm": 1.140625, + "learning_rate": 0.0002, + "loss": 1.2967, + "step": 9560 + }, + { + "epoch": 3.0151371807000946, + "grad_norm": 0.82421875, + "learning_rate": 0.0002, + "loss": 1.0359, + "step": 9561 + }, + { + "epoch": 3.0154525386313464, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 1.074, + "step": 9562 + }, + { + "epoch": 3.0157678965625987, + "grad_norm": 1.2578125, + "learning_rate": 0.0002, + "loss": 1.1313, + "step": 9563 + }, + { + "epoch": 3.0160832544938505, + "grad_norm": 0.400390625, + "learning_rate": 0.0002, + "loss": 0.8915, + "step": 9564 + }, + { + "epoch": 3.0163986124251023, + "grad_norm": 0.376953125, + "learning_rate": 0.0002, + "loss": 0.824, + "step": 9565 + }, + { + "epoch": 3.0167139703563546, + "grad_norm": 0.38671875, + "learning_rate": 0.0002, + "loss": 0.7898, + "step": 9566 + }, + { + "epoch": 3.0170293282876064, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 0.7687, + "step": 9567 + }, + { + "epoch": 3.017344686218858, + "grad_norm": 0.349609375, + "learning_rate": 0.0002, + "loss": 0.8606, + "step": 9568 + }, + { + "epoch": 3.0176600441501105, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 0.7461, + "step": 9569 + }, + { + "epoch": 3.0179754020813623, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 0.623, + "step": 9570 + }, + { + "epoch": 3.0182907600126145, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 0.7688, + "step": 9571 + }, + { + "epoch": 3.0186061179438664, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 0.7569, + "step": 9572 + }, + { + "epoch": 3.018921475875118, + "grad_norm": 0.375, + "learning_rate": 0.0002, + "loss": 0.7149, + "step": 9573 + }, + { + "epoch": 3.0192368338063704, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 0.7207, + "step": 9574 + }, + { + "epoch": 3.0195521917376222, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 0.6762, + "step": 9575 + }, + { + "epoch": 3.019867549668874, + "grad_norm": 0.373046875, + "learning_rate": 0.0002, + "loss": 0.6169, + "step": 9576 + }, + { + "epoch": 3.0201829076001263, + "grad_norm": 0.3984375, + "learning_rate": 0.0002, + "loss": 0.6739, + "step": 9577 + }, + { + "epoch": 3.020498265531378, + "grad_norm": 0.48046875, + "learning_rate": 0.0002, + "loss": 0.6824, + "step": 9578 + }, + { + "epoch": 3.02081362346263, + "grad_norm": 0.39453125, + "learning_rate": 0.0002, + "loss": 0.6269, + "step": 9579 + }, + { + "epoch": 3.021128981393882, + "grad_norm": 0.451171875, + "learning_rate": 0.0002, + "loss": 0.6504, + "step": 9580 + }, + { + "epoch": 3.021444339325134, + "grad_norm": 0.484375, + "learning_rate": 0.0002, + "loss": 0.787, + "step": 9581 + }, + { + "epoch": 3.021759697256386, + "grad_norm": 0.4453125, + "learning_rate": 0.0002, + "loss": 0.7063, + "step": 9582 + }, + { + "epoch": 3.022075055187638, + "grad_norm": 0.52734375, + "learning_rate": 0.0002, + "loss": 0.606, + "step": 9583 + }, + { + "epoch": 3.02239041311889, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 0.6462, + "step": 9584 + }, + { + "epoch": 3.0227057710501417, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 0.6371, + "step": 9585 + }, + { + "epoch": 3.023021128981394, + "grad_norm": 0.5546875, + "learning_rate": 0.0002, + "loss": 0.5954, + "step": 9586 + }, + { + "epoch": 3.023336486912646, + "grad_norm": 0.53515625, + "learning_rate": 0.0002, + "loss": 0.5709, + "step": 9587 + }, + { + "epoch": 3.023651844843898, + "grad_norm": 0.58984375, + "learning_rate": 0.0002, + "loss": 0.7101, + "step": 9588 + }, + { + "epoch": 3.02396720277515, + "grad_norm": 0.6015625, + "learning_rate": 0.0002, + "loss": 0.597, + "step": 9589 + }, + { + "epoch": 3.0242825607064017, + "grad_norm": 0.8046875, + "learning_rate": 0.0002, + "loss": 0.7673, + "step": 9590 + }, + { + "epoch": 3.024597918637654, + "grad_norm": 0.81640625, + "learning_rate": 0.0002, + "loss": 0.6508, + "step": 9591 + }, + { + "epoch": 3.0249132765689057, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 0.6419, + "step": 9592 + }, + { + "epoch": 3.0252286345001576, + "grad_norm": 0.8359375, + "learning_rate": 0.0002, + "loss": 0.5689, + "step": 9593 + }, + { + "epoch": 3.02554399243141, + "grad_norm": 0.73046875, + "learning_rate": 0.0002, + "loss": 0.6618, + "step": 9594 + }, + { + "epoch": 3.0258593503626616, + "grad_norm": 1.0078125, + "learning_rate": 0.0002, + "loss": 0.8348, + "step": 9595 + }, + { + "epoch": 3.0261747082939134, + "grad_norm": 1.0546875, + "learning_rate": 0.0002, + "loss": 0.7489, + "step": 9596 + }, + { + "epoch": 3.0264900662251657, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 0.6195, + "step": 9597 + }, + { + "epoch": 3.0268054241564175, + "grad_norm": 0.62109375, + "learning_rate": 0.0002, + "loss": 0.6364, + "step": 9598 + }, + { + "epoch": 3.0271207820876693, + "grad_norm": 1.03125, + "learning_rate": 0.0002, + "loss": 0.6616, + "step": 9599 + }, + { + "epoch": 3.0274361400189216, + "grad_norm": 0.921875, + "learning_rate": 0.0002, + "loss": 0.8875, + "step": 9600 + }, + { + "epoch": 3.0277514979501734, + "grad_norm": 1.3203125, + "learning_rate": 0.0002, + "loss": 0.9012, + "step": 9601 + }, + { + "epoch": 3.028066855881425, + "grad_norm": 2.15625, + "learning_rate": 0.0002, + "loss": 0.9506, + "step": 9602 + }, + { + "epoch": 3.0283822138126775, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 0.9822, + "step": 9603 + }, + { + "epoch": 3.0286975717439293, + "grad_norm": 0.9609375, + "learning_rate": 0.0002, + "loss": 0.9078, + "step": 9604 + }, + { + "epoch": 3.0290129296751815, + "grad_norm": 1.4921875, + "learning_rate": 0.0002, + "loss": 1.228, + "step": 9605 + }, + { + "epoch": 3.0293282876064334, + "grad_norm": 1.9765625, + "learning_rate": 0.0002, + "loss": 1.1882, + "step": 9606 + }, + { + "epoch": 3.029643645537685, + "grad_norm": 0.93359375, + "learning_rate": 0.0002, + "loss": 1.0426, + "step": 9607 + }, + { + "epoch": 3.0299590034689374, + "grad_norm": 1.21875, + "learning_rate": 0.0002, + "loss": 1.6427, + "step": 9608 + }, + { + "epoch": 3.0302743614001892, + "grad_norm": 1.0546875, + "learning_rate": 0.0002, + "loss": 1.2852, + "step": 9609 + }, + { + "epoch": 3.030589719331441, + "grad_norm": 3.625, + "learning_rate": 0.0002, + "loss": 1.4931, + "step": 9610 + }, + { + "epoch": 3.0309050772626933, + "grad_norm": 1.3515625, + "learning_rate": 0.0002, + "loss": 1.1129, + "step": 9611 + }, + { + "epoch": 3.031220435193945, + "grad_norm": 1.109375, + "learning_rate": 0.0002, + "loss": 1.0054, + "step": 9612 + }, + { + "epoch": 3.031535793125197, + "grad_norm": 2.359375, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 9613 + }, + { + "epoch": 3.031851151056449, + "grad_norm": 0.2890625, + "learning_rate": 0.0002, + "loss": 0.9268, + "step": 9614 + }, + { + "epoch": 3.032166508987701, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 0.8836, + "step": 9615 + }, + { + "epoch": 3.032481866918953, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 0.773, + "step": 9616 + }, + { + "epoch": 3.032797224850205, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 0.8497, + "step": 9617 + }, + { + "epoch": 3.033112582781457, + "grad_norm": 0.314453125, + "learning_rate": 0.0002, + "loss": 0.7906, + "step": 9618 + }, + { + "epoch": 3.0334279407127087, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 0.7845, + "step": 9619 + }, + { + "epoch": 3.033743298643961, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 0.7187, + "step": 9620 + }, + { + "epoch": 3.034058656575213, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 0.7386, + "step": 9621 + }, + { + "epoch": 3.034374014506465, + "grad_norm": 0.3515625, + "learning_rate": 0.0002, + "loss": 0.6743, + "step": 9622 + }, + { + "epoch": 3.034689372437717, + "grad_norm": 0.31640625, + "learning_rate": 0.0002, + "loss": 0.6572, + "step": 9623 + }, + { + "epoch": 3.0350047303689687, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 0.749, + "step": 9624 + }, + { + "epoch": 3.035320088300221, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 0.6521, + "step": 9625 + }, + { + "epoch": 3.0356354462314727, + "grad_norm": 0.3828125, + "learning_rate": 0.0002, + "loss": 0.7467, + "step": 9626 + }, + { + "epoch": 3.0359508041627246, + "grad_norm": 0.349609375, + "learning_rate": 0.0002, + "loss": 0.6213, + "step": 9627 + }, + { + "epoch": 3.036266162093977, + "grad_norm": 0.470703125, + "learning_rate": 0.0002, + "loss": 0.7644, + "step": 9628 + }, + { + "epoch": 3.0365815200252286, + "grad_norm": 0.46875, + "learning_rate": 0.0002, + "loss": 0.6426, + "step": 9629 + }, + { + "epoch": 3.0368968779564804, + "grad_norm": 0.49609375, + "learning_rate": 0.0002, + "loss": 0.7209, + "step": 9630 + }, + { + "epoch": 3.0372122358877327, + "grad_norm": 0.470703125, + "learning_rate": 0.0002, + "loss": 0.6474, + "step": 9631 + }, + { + "epoch": 3.0375275938189845, + "grad_norm": 0.46875, + "learning_rate": 0.0002, + "loss": 0.6988, + "step": 9632 + }, + { + "epoch": 3.0378429517502363, + "grad_norm": 0.462890625, + "learning_rate": 0.0002, + "loss": 0.5912, + "step": 9633 + }, + { + "epoch": 3.0381583096814886, + "grad_norm": 0.7578125, + "learning_rate": 0.0002, + "loss": 0.6387, + "step": 9634 + }, + { + "epoch": 3.0384736676127404, + "grad_norm": 0.546875, + "learning_rate": 0.0002, + "loss": 0.5576, + "step": 9635 + }, + { + "epoch": 3.038789025543992, + "grad_norm": 0.5390625, + "learning_rate": 0.0002, + "loss": 0.5747, + "step": 9636 + }, + { + "epoch": 3.0391043834752445, + "grad_norm": 0.65625, + "learning_rate": 0.0002, + "loss": 0.6539, + "step": 9637 + }, + { + "epoch": 3.0394197414064963, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 0.6182, + "step": 9638 + }, + { + "epoch": 3.0397350993377485, + "grad_norm": 0.65625, + "learning_rate": 0.0002, + "loss": 0.6306, + "step": 9639 + }, + { + "epoch": 3.0400504572690004, + "grad_norm": 0.578125, + "learning_rate": 0.0002, + "loss": 0.6341, + "step": 9640 + }, + { + "epoch": 3.040365815200252, + "grad_norm": 0.6875, + "learning_rate": 0.0002, + "loss": 0.7109, + "step": 9641 + }, + { + "epoch": 3.0406811731315044, + "grad_norm": 0.91796875, + "learning_rate": 0.0002, + "loss": 0.7064, + "step": 9642 + }, + { + "epoch": 3.0409965310627562, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 0.5317, + "step": 9643 + }, + { + "epoch": 3.041311888994008, + "grad_norm": 0.7109375, + "learning_rate": 0.0002, + "loss": 0.5612, + "step": 9644 + }, + { + "epoch": 3.0416272469252603, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 0.7099, + "step": 9645 + }, + { + "epoch": 3.041942604856512, + "grad_norm": 0.98046875, + "learning_rate": 0.0002, + "loss": 0.788, + "step": 9646 + }, + { + "epoch": 3.042257962787764, + "grad_norm": 1.1328125, + "learning_rate": 0.0002, + "loss": 0.9258, + "step": 9647 + }, + { + "epoch": 3.042573320719016, + "grad_norm": 1.1875, + "learning_rate": 0.0002, + "loss": 0.7219, + "step": 9648 + }, + { + "epoch": 3.042888678650268, + "grad_norm": 1.3984375, + "learning_rate": 0.0002, + "loss": 0.8218, + "step": 9649 + }, + { + "epoch": 3.04320403658152, + "grad_norm": 1.0546875, + "learning_rate": 0.0002, + "loss": 0.8561, + "step": 9650 + }, + { + "epoch": 3.043519394512772, + "grad_norm": 1.296875, + "learning_rate": 0.0002, + "loss": 0.9527, + "step": 9651 + }, + { + "epoch": 3.043834752444024, + "grad_norm": 1.015625, + "learning_rate": 0.0002, + "loss": 1.2758, + "step": 9652 + }, + { + "epoch": 3.044150110375276, + "grad_norm": 1.3203125, + "learning_rate": 0.0002, + "loss": 1.169, + "step": 9653 + }, + { + "epoch": 3.044465468306528, + "grad_norm": 1.09375, + "learning_rate": 0.0002, + "loss": 1.1941, + "step": 9654 + }, + { + "epoch": 3.04478082623778, + "grad_norm": 1.125, + "learning_rate": 0.0002, + "loss": 1.0929, + "step": 9655 + }, + { + "epoch": 3.045096184169032, + "grad_norm": 1.1171875, + "learning_rate": 0.0002, + "loss": 1.2858, + "step": 9656 + }, + { + "epoch": 3.045411542100284, + "grad_norm": 1.5703125, + "learning_rate": 0.0002, + "loss": 1.1435, + "step": 9657 + }, + { + "epoch": 3.0457269000315357, + "grad_norm": 1.625, + "learning_rate": 0.0002, + "loss": 1.4967, + "step": 9658 + }, + { + "epoch": 3.046042257962788, + "grad_norm": 1.4609375, + "learning_rate": 0.0002, + "loss": 1.2679, + "step": 9659 + }, + { + "epoch": 3.0463576158940397, + "grad_norm": 1.1640625, + "learning_rate": 0.0002, + "loss": 1.132, + "step": 9660 + }, + { + "epoch": 3.0466729738252916, + "grad_norm": 1.6875, + "learning_rate": 0.0002, + "loss": 1.2314, + "step": 9661 + }, + { + "epoch": 3.046988331756544, + "grad_norm": 0.94921875, + "learning_rate": 0.0002, + "loss": 0.9875, + "step": 9662 + }, + { + "epoch": 3.0473036896877956, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 1.2083, + "step": 9663 + }, + { + "epoch": 3.0476190476190474, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 0.9508, + "step": 9664 + }, + { + "epoch": 3.0479344055502997, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 0.8139, + "step": 9665 + }, + { + "epoch": 3.0482497634815515, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 0.8378, + "step": 9666 + }, + { + "epoch": 3.0485651214128033, + "grad_norm": 0.400390625, + "learning_rate": 0.0002, + "loss": 0.7973, + "step": 9667 + }, + { + "epoch": 3.0488804793440556, + "grad_norm": 0.353515625, + "learning_rate": 0.0002, + "loss": 0.7855, + "step": 9668 + }, + { + "epoch": 3.0491958372753074, + "grad_norm": 0.3515625, + "learning_rate": 0.0002, + "loss": 0.7866, + "step": 9669 + }, + { + "epoch": 3.049511195206559, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 0.8585, + "step": 9670 + }, + { + "epoch": 3.0498265531378115, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 0.7943, + "step": 9671 + }, + { + "epoch": 3.0501419110690633, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 0.637, + "step": 9672 + }, + { + "epoch": 3.0504572690003156, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 0.7632, + "step": 9673 + }, + { + "epoch": 3.0507726269315674, + "grad_norm": 0.3671875, + "learning_rate": 0.0002, + "loss": 0.6841, + "step": 9674 + }, + { + "epoch": 3.051087984862819, + "grad_norm": 0.365234375, + "learning_rate": 0.0002, + "loss": 0.7594, + "step": 9675 + }, + { + "epoch": 3.0514033427940714, + "grad_norm": 0.4453125, + "learning_rate": 0.0002, + "loss": 0.8018, + "step": 9676 + }, + { + "epoch": 3.0517187007253233, + "grad_norm": 0.482421875, + "learning_rate": 0.0002, + "loss": 0.6065, + "step": 9677 + }, + { + "epoch": 3.052034058656575, + "grad_norm": 0.38671875, + "learning_rate": 0.0002, + "loss": 0.6595, + "step": 9678 + }, + { + "epoch": 3.0523494165878273, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 0.8247, + "step": 9679 + }, + { + "epoch": 3.052664774519079, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 0.6867, + "step": 9680 + }, + { + "epoch": 3.052980132450331, + "grad_norm": 0.458984375, + "learning_rate": 0.0002, + "loss": 0.6432, + "step": 9681 + }, + { + "epoch": 3.053295490381583, + "grad_norm": 0.49609375, + "learning_rate": 0.0002, + "loss": 0.6181, + "step": 9682 + }, + { + "epoch": 3.053610848312835, + "grad_norm": 0.6953125, + "learning_rate": 0.0002, + "loss": 0.6369, + "step": 9683 + }, + { + "epoch": 3.053926206244087, + "grad_norm": 0.490234375, + "learning_rate": 0.0002, + "loss": 0.614, + "step": 9684 + }, + { + "epoch": 3.054241564175339, + "grad_norm": 0.4921875, + "learning_rate": 0.0002, + "loss": 0.6286, + "step": 9685 + }, + { + "epoch": 3.054556922106591, + "grad_norm": 0.515625, + "learning_rate": 0.0002, + "loss": 0.6677, + "step": 9686 + }, + { + "epoch": 3.054872280037843, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 0.6994, + "step": 9687 + }, + { + "epoch": 3.055187637969095, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 0.6021, + "step": 9688 + }, + { + "epoch": 3.055502995900347, + "grad_norm": 0.6171875, + "learning_rate": 0.0002, + "loss": 0.7105, + "step": 9689 + }, + { + "epoch": 3.055818353831599, + "grad_norm": 0.68359375, + "learning_rate": 0.0002, + "loss": 0.6836, + "step": 9690 + }, + { + "epoch": 3.056133711762851, + "grad_norm": 0.59375, + "learning_rate": 0.0002, + "loss": 0.5107, + "step": 9691 + }, + { + "epoch": 3.0564490696941027, + "grad_norm": 0.7421875, + "learning_rate": 0.0002, + "loss": 0.6423, + "step": 9692 + }, + { + "epoch": 3.056764427625355, + "grad_norm": 0.73046875, + "learning_rate": 0.0002, + "loss": 0.5345, + "step": 9693 + }, + { + "epoch": 3.0570797855566068, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 0.5594, + "step": 9694 + }, + { + "epoch": 3.0573951434878586, + "grad_norm": 0.96875, + "learning_rate": 0.0002, + "loss": 0.7349, + "step": 9695 + }, + { + "epoch": 3.057710501419111, + "grad_norm": 0.8125, + "learning_rate": 0.0002, + "loss": 0.6797, + "step": 9696 + }, + { + "epoch": 3.0580258593503626, + "grad_norm": 0.83203125, + "learning_rate": 0.0002, + "loss": 0.7709, + "step": 9697 + }, + { + "epoch": 3.0583412172816145, + "grad_norm": 0.9921875, + "learning_rate": 0.0002, + "loss": 0.7901, + "step": 9698 + }, + { + "epoch": 3.0586565752128667, + "grad_norm": 0.953125, + "learning_rate": 0.0002, + "loss": 0.8321, + "step": 9699 + }, + { + "epoch": 3.0589719331441185, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 0.6784, + "step": 9700 + }, + { + "epoch": 3.0592872910753703, + "grad_norm": 0.90625, + "learning_rate": 0.0002, + "loss": 0.935, + "step": 9701 + }, + { + "epoch": 3.0596026490066226, + "grad_norm": 1.234375, + "learning_rate": 0.0002, + "loss": 1.2005, + "step": 9702 + }, + { + "epoch": 3.0599180069378744, + "grad_norm": 1.171875, + "learning_rate": 0.0002, + "loss": 1.2039, + "step": 9703 + }, + { + "epoch": 3.0602333648691267, + "grad_norm": 1.3203125, + "learning_rate": 0.0002, + "loss": 1.0325, + "step": 9704 + }, + { + "epoch": 3.0605487228003785, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 9705 + }, + { + "epoch": 3.0608640807316303, + "grad_norm": 1.40625, + "learning_rate": 0.0002, + "loss": 1.0815, + "step": 9706 + }, + { + "epoch": 3.0611794386628826, + "grad_norm": 1.2109375, + "learning_rate": 0.0002, + "loss": 1.4965, + "step": 9707 + }, + { + "epoch": 3.0614947965941344, + "grad_norm": 1.34375, + "learning_rate": 0.0002, + "loss": 1.5561, + "step": 9708 + }, + { + "epoch": 3.061810154525386, + "grad_norm": 1.28125, + "learning_rate": 0.0002, + "loss": 1.0898, + "step": 9709 + }, + { + "epoch": 3.0621255124566384, + "grad_norm": 1.1328125, + "learning_rate": 0.0002, + "loss": 1.2331, + "step": 9710 + }, + { + "epoch": 3.0624408703878903, + "grad_norm": 1.921875, + "learning_rate": 0.0002, + "loss": 1.1502, + "step": 9711 + }, + { + "epoch": 3.062756228319142, + "grad_norm": 1.28125, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 9712 + }, + { + "epoch": 3.0630715862503943, + "grad_norm": 0.953125, + "learning_rate": 0.0002, + "loss": 1.3868, + "step": 9713 + }, + { + "epoch": 3.063386944181646, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 0.8527, + "step": 9714 + }, + { + "epoch": 3.063702302112898, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 0.7363, + "step": 9715 + }, + { + "epoch": 3.06401766004415, + "grad_norm": 0.349609375, + "learning_rate": 0.0002, + "loss": 0.8556, + "step": 9716 + }, + { + "epoch": 3.064333017975402, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 0.756, + "step": 9717 + }, + { + "epoch": 3.064648375906654, + "grad_norm": 0.3203125, + "learning_rate": 0.0002, + "loss": 0.7247, + "step": 9718 + }, + { + "epoch": 3.064963733837906, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 0.8308, + "step": 9719 + }, + { + "epoch": 3.065279091769158, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 0.7607, + "step": 9720 + }, + { + "epoch": 3.06559444970041, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 0.7187, + "step": 9721 + }, + { + "epoch": 3.065909807631662, + "grad_norm": 0.349609375, + "learning_rate": 0.0002, + "loss": 0.7534, + "step": 9722 + }, + { + "epoch": 3.066225165562914, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 0.6926, + "step": 9723 + }, + { + "epoch": 3.066540523494166, + "grad_norm": 0.380859375, + "learning_rate": 0.0002, + "loss": 0.6616, + "step": 9724 + }, + { + "epoch": 3.066540523494166, + "eval_loss": 1.9542231559753418, + "eval_runtime": 149.103, + "eval_samples_per_second": 6.707, + "eval_steps_per_second": 6.707, + "step": 9724 + }, + { + "epoch": 3.066540523494166, + "mmlu_eval_accuracy": 0.5900854222334173, + "mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453, + "mmlu_eval_accuracy_anatomy": 0.5714285714285714, + "mmlu_eval_accuracy_astronomy": 0.6875, + "mmlu_eval_accuracy_business_ethics": 0.9090909090909091, + "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138, + "mmlu_eval_accuracy_college_biology": 0.875, + "mmlu_eval_accuracy_college_chemistry": 0.375, + "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727, + "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454, + "mmlu_eval_accuracy_college_medicine": 0.5, + "mmlu_eval_accuracy_college_physics": 0.36363636363636365, + "mmlu_eval_accuracy_computer_security": 0.36363636363636365, + "mmlu_eval_accuracy_conceptual_physics": 0.5, + "mmlu_eval_accuracy_econometrics": 0.5, + "mmlu_eval_accuracy_electrical_engineering": 0.4375, + "mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536, + "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, + "mmlu_eval_accuracy_global_facts": 0.3, + "mmlu_eval_accuracy_high_school_biology": 0.6875, + "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334, + "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.627906976744186, + "mmlu_eval_accuracy_high_school_mathematics": 0.41379310344827586, + "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769, + "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, + "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667, + "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, + "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182, + "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539, + "mmlu_eval_accuracy_human_aging": 0.5652173913043478, + "mmlu_eval_accuracy_human_sexuality": 0.5, + "mmlu_eval_accuracy_international_law": 0.8461538461538461, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, + "mmlu_eval_accuracy_machine_learning": 0.2727272727272727, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.76, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.7093023255813954, + "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527, + "mmlu_eval_accuracy_moral_scenarios": 0.25, + "mmlu_eval_accuracy_nutrition": 0.7272727272727273, + "mmlu_eval_accuracy_philosophy": 0.6176470588235294, + "mmlu_eval_accuracy_prehistory": 0.6571428571428571, + "mmlu_eval_accuracy_professional_accounting": 0.5806451612903226, + "mmlu_eval_accuracy_professional_law": 0.3588235294117647, + "mmlu_eval_accuracy_professional_medicine": 0.7096774193548387, + "mmlu_eval_accuracy_professional_psychology": 0.5942028985507246, + "mmlu_eval_accuracy_public_relations": 0.5, + "mmlu_eval_accuracy_security_studies": 0.6666666666666666, + "mmlu_eval_accuracy_sociology": 0.9545454545454546, + "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, + "mmlu_eval_accuracy_virology": 0.4444444444444444, + "mmlu_eval_accuracy_world_religions": 0.7894736842105263, + "mmlu_loss": 5.988420398931423, + "step": 9724 + }, + { + "epoch": 3.066855881425418, + "grad_norm": 0.361328125, + "learning_rate": 0.0002, + "loss": 0.6493, + "step": 9725 + }, + { + "epoch": 3.0671712393566697, + "grad_norm": 0.388671875, + "learning_rate": 0.0002, + "loss": 0.6378, + "step": 9726 + }, + { + "epoch": 3.067486597287922, + "grad_norm": 0.392578125, + "learning_rate": 0.0002, + "loss": 0.6599, + "step": 9727 + }, + { + "epoch": 3.0678019552191738, + "grad_norm": 0.412109375, + "learning_rate": 0.0002, + "loss": 0.634, + "step": 9728 + }, + { + "epoch": 3.0681173131504256, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 0.6605, + "step": 9729 + }, + { + "epoch": 3.068432671081678, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 0.689, + "step": 9730 + }, + { + "epoch": 3.0687480290129296, + "grad_norm": 0.51953125, + "learning_rate": 0.0002, + "loss": 0.6328, + "step": 9731 + }, + { + "epoch": 3.0690633869441815, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 0.5226, + "step": 9732 + }, + { + "epoch": 3.0693787448754337, + "grad_norm": 0.515625, + "learning_rate": 0.0002, + "loss": 0.6069, + "step": 9733 + }, + { + "epoch": 3.0696941028066855, + "grad_norm": 0.5390625, + "learning_rate": 0.0002, + "loss": 0.6587, + "step": 9734 + }, + { + "epoch": 3.0700094607379373, + "grad_norm": 0.66015625, + "learning_rate": 0.0002, + "loss": 0.614, + "step": 9735 + }, + { + "epoch": 3.0703248186691896, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 0.6328, + "step": 9736 + }, + { + "epoch": 3.0706401766004414, + "grad_norm": 0.62890625, + "learning_rate": 0.0002, + "loss": 0.6561, + "step": 9737 + }, + { + "epoch": 3.0709555345316937, + "grad_norm": 0.60546875, + "learning_rate": 0.0002, + "loss": 0.7507, + "step": 9738 + }, + { + "epoch": 3.0712708924629455, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 0.5708, + "step": 9739 + }, + { + "epoch": 3.0715862503941973, + "grad_norm": 0.64453125, + "learning_rate": 0.0002, + "loss": 0.603, + "step": 9740 + }, + { + "epoch": 3.0719016083254496, + "grad_norm": 0.73828125, + "learning_rate": 0.0002, + "loss": 0.5526, + "step": 9741 + }, + { + "epoch": 3.0722169662567014, + "grad_norm": 0.85546875, + "learning_rate": 0.0002, + "loss": 0.618, + "step": 9742 + }, + { + "epoch": 3.072532324187953, + "grad_norm": 0.67578125, + "learning_rate": 0.0002, + "loss": 0.606, + "step": 9743 + }, + { + "epoch": 3.0728476821192054, + "grad_norm": 0.76171875, + "learning_rate": 0.0002, + "loss": 0.6644, + "step": 9744 + }, + { + "epoch": 3.0731630400504573, + "grad_norm": 0.83984375, + "learning_rate": 0.0002, + "loss": 0.6574, + "step": 9745 + }, + { + "epoch": 3.073478397981709, + "grad_norm": 0.984375, + "learning_rate": 0.0002, + "loss": 0.6981, + "step": 9746 + }, + { + "epoch": 3.0737937559129613, + "grad_norm": 0.93359375, + "learning_rate": 0.0002, + "loss": 0.8201, + "step": 9747 + }, + { + "epoch": 3.074109113844213, + "grad_norm": 0.8203125, + "learning_rate": 0.0002, + "loss": 0.8174, + "step": 9748 + }, + { + "epoch": 3.074424471775465, + "grad_norm": 0.9140625, + "learning_rate": 0.0002, + "loss": 0.6713, + "step": 9749 + }, + { + "epoch": 3.074739829706717, + "grad_norm": 1.1171875, + "learning_rate": 0.0002, + "loss": 0.9157, + "step": 9750 + }, + { + "epoch": 3.075055187637969, + "grad_norm": 0.92578125, + "learning_rate": 0.0002, + "loss": 0.7511, + "step": 9751 + }, + { + "epoch": 3.0753705455692213, + "grad_norm": 0.9765625, + "learning_rate": 0.0002, + "loss": 0.884, + "step": 9752 + }, + { + "epoch": 3.075685903500473, + "grad_norm": 1.15625, + "learning_rate": 0.0002, + "loss": 1.0204, + "step": 9753 + }, + { + "epoch": 3.076001261431725, + "grad_norm": 1.6953125, + "learning_rate": 0.0002, + "loss": 1.1332, + "step": 9754 + }, + { + "epoch": 3.076316619362977, + "grad_norm": 1.296875, + "learning_rate": 0.0002, + "loss": 1.067, + "step": 9755 + }, + { + "epoch": 3.076631977294229, + "grad_norm": 2.84375, + "learning_rate": 0.0002, + "loss": 1.1203, + "step": 9756 + }, + { + "epoch": 3.076947335225481, + "grad_norm": 1.5625, + "learning_rate": 0.0002, + "loss": 1.4864, + "step": 9757 + }, + { + "epoch": 3.077262693156733, + "grad_norm": 1.0, + "learning_rate": 0.0002, + "loss": 1.2943, + "step": 9758 + }, + { + "epoch": 3.077578051087985, + "grad_norm": 1.2109375, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 9759 + }, + { + "epoch": 3.0778934090192367, + "grad_norm": 1.2265625, + "learning_rate": 0.0002, + "loss": 1.164, + "step": 9760 + }, + { + "epoch": 3.078208766950489, + "grad_norm": 1.7890625, + "learning_rate": 0.0002, + "loss": 1.1606, + "step": 9761 + }, + { + "epoch": 3.0785241248817408, + "grad_norm": 0.91796875, + "learning_rate": 0.0002, + "loss": 1.0192, + "step": 9762 + }, + { + "epoch": 3.0788394828129926, + "grad_norm": 1.1953125, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 9763 + }, + { + "epoch": 3.079154840744245, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 0.8393, + "step": 9764 + }, + { + "epoch": 3.0794701986754967, + "grad_norm": 0.3984375, + "learning_rate": 0.0002, + "loss": 0.9468, + "step": 9765 + }, + { + "epoch": 3.0797855566067485, + "grad_norm": 0.3515625, + "learning_rate": 0.0002, + "loss": 0.8253, + "step": 9766 + }, + { + "epoch": 3.0801009145380007, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 0.8413, + "step": 9767 + }, + { + "epoch": 3.0804162724692525, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 0.7947, + "step": 9768 + }, + { + "epoch": 3.0807316304005044, + "grad_norm": 0.330078125, + "learning_rate": 0.0002, + "loss": 0.717, + "step": 9769 + }, + { + "epoch": 3.0810469883317566, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 0.7487, + "step": 9770 + }, + { + "epoch": 3.0813623462630084, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 0.8105, + "step": 9771 + }, + { + "epoch": 3.0816777041942607, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 0.7353, + "step": 9772 + }, + { + "epoch": 3.0819930621255125, + "grad_norm": 0.3515625, + "learning_rate": 0.0002, + "loss": 0.791, + "step": 9773 + }, + { + "epoch": 3.0823084200567643, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 0.7328, + "step": 9774 + }, + { + "epoch": 3.0826237779880166, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 0.6659, + "step": 9775 + }, + { + "epoch": 3.0829391359192684, + "grad_norm": 0.4375, + "learning_rate": 0.0002, + "loss": 0.8017, + "step": 9776 + }, + { + "epoch": 3.08325449385052, + "grad_norm": 0.419921875, + "learning_rate": 0.0002, + "loss": 0.6766, + "step": 9777 + }, + { + "epoch": 3.0835698517817725, + "grad_norm": 0.400390625, + "learning_rate": 0.0002, + "loss": 0.7223, + "step": 9778 + }, + { + "epoch": 3.0838852097130243, + "grad_norm": 0.443359375, + "learning_rate": 0.0002, + "loss": 0.6933, + "step": 9779 + }, + { + "epoch": 3.084200567644276, + "grad_norm": 0.494140625, + "learning_rate": 0.0002, + "loss": 0.6793, + "step": 9780 + }, + { + "epoch": 3.0845159255755283, + "grad_norm": 0.46875, + "learning_rate": 0.0002, + "loss": 0.6931, + "step": 9781 + }, + { + "epoch": 3.08483128350678, + "grad_norm": 0.462890625, + "learning_rate": 0.0002, + "loss": 0.5551, + "step": 9782 + }, + { + "epoch": 3.085146641438032, + "grad_norm": 0.4375, + "learning_rate": 0.0002, + "loss": 0.6003, + "step": 9783 + }, + { + "epoch": 3.0854619993692842, + "grad_norm": 0.8984375, + "learning_rate": 0.0002, + "loss": 0.7331, + "step": 9784 + }, + { + "epoch": 3.085777357300536, + "grad_norm": 0.58984375, + "learning_rate": 0.0002, + "loss": 0.6959, + "step": 9785 + }, + { + "epoch": 3.0860927152317883, + "grad_norm": 0.55078125, + "learning_rate": 0.0002, + "loss": 0.6265, + "step": 9786 + }, + { + "epoch": 3.08640807316304, + "grad_norm": 0.5546875, + "learning_rate": 0.0002, + "loss": 0.5831, + "step": 9787 + }, + { + "epoch": 3.086723431094292, + "grad_norm": 0.55859375, + "learning_rate": 0.0002, + "loss": 0.5694, + "step": 9788 + }, + { + "epoch": 3.087038789025544, + "grad_norm": 0.7421875, + "learning_rate": 0.0002, + "loss": 0.7514, + "step": 9789 + }, + { + "epoch": 3.087354146956796, + "grad_norm": 0.76171875, + "learning_rate": 0.0002, + "loss": 0.5609, + "step": 9790 + }, + { + "epoch": 3.087669504888048, + "grad_norm": 0.71875, + "learning_rate": 0.0002, + "loss": 0.7239, + "step": 9791 + }, + { + "epoch": 3.0879848628193, + "grad_norm": 0.76953125, + "learning_rate": 0.0002, + "loss": 0.6872, + "step": 9792 + }, + { + "epoch": 3.088300220750552, + "grad_norm": 0.8046875, + "learning_rate": 0.0002, + "loss": 0.5697, + "step": 9793 + }, + { + "epoch": 3.0886155786818037, + "grad_norm": 0.8515625, + "learning_rate": 0.0002, + "loss": 0.6614, + "step": 9794 + }, + { + "epoch": 3.088930936613056, + "grad_norm": 0.89453125, + "learning_rate": 0.0002, + "loss": 0.7819, + "step": 9795 + }, + { + "epoch": 3.0892462945443078, + "grad_norm": 0.734375, + "learning_rate": 0.0002, + "loss": 0.7848, + "step": 9796 + }, + { + "epoch": 3.0895616524755596, + "grad_norm": 0.9453125, + "learning_rate": 0.0002, + "loss": 0.8611, + "step": 9797 + }, + { + "epoch": 3.089877010406812, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 0.9354, + "step": 9798 + }, + { + "epoch": 3.0901923683380637, + "grad_norm": 1.375, + "learning_rate": 0.0002, + "loss": 0.9084, + "step": 9799 + }, + { + "epoch": 3.0905077262693155, + "grad_norm": 1.515625, + "learning_rate": 0.0002, + "loss": 1.1268, + "step": 9800 + }, + { + "epoch": 3.0908230842005677, + "grad_norm": 1.6328125, + "learning_rate": 0.0002, + "loss": 1.0787, + "step": 9801 + }, + { + "epoch": 3.0911384421318195, + "grad_norm": 1.3359375, + "learning_rate": 0.0002, + "loss": 1.0188, + "step": 9802 + }, + { + "epoch": 3.091453800063072, + "grad_norm": 1.3984375, + "learning_rate": 0.0002, + "loss": 1.2487, + "step": 9803 + }, + { + "epoch": 3.0917691579943236, + "grad_norm": 1.2109375, + "learning_rate": 0.0002, + "loss": 1.051, + "step": 9804 + }, + { + "epoch": 3.0920845159255754, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 1.2125, + "step": 9805 + }, + { + "epoch": 3.0923998738568277, + "grad_norm": 1.859375, + "learning_rate": 0.0002, + "loss": 1.5213, + "step": 9806 + }, + { + "epoch": 3.0927152317880795, + "grad_norm": 1.34375, + "learning_rate": 0.0002, + "loss": 1.6901, + "step": 9807 + }, + { + "epoch": 3.0930305897193313, + "grad_norm": 1.2578125, + "learning_rate": 0.0002, + "loss": 1.1607, + "step": 9808 + }, + { + "epoch": 3.0933459476505836, + "grad_norm": 0.9140625, + "learning_rate": 0.0002, + "loss": 1.1863, + "step": 9809 + }, + { + "epoch": 3.0936613055818354, + "grad_norm": 0.9609375, + "learning_rate": 0.0002, + "loss": 1.1153, + "step": 9810 + }, + { + "epoch": 3.093976663513087, + "grad_norm": 0.90234375, + "learning_rate": 0.0002, + "loss": 1.2714, + "step": 9811 + }, + { + "epoch": 3.0942920214443395, + "grad_norm": 1.3125, + "learning_rate": 0.0002, + "loss": 1.093, + "step": 9812 + }, + { + "epoch": 3.0946073793755913, + "grad_norm": 0.98046875, + "learning_rate": 0.0002, + "loss": 1.7525, + "step": 9813 + }, + { + "epoch": 3.094922737306843, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 0.9086, + "step": 9814 + }, + { + "epoch": 3.0952380952380953, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 0.7795, + "step": 9815 + }, + { + "epoch": 3.095553453169347, + "grad_norm": 0.30078125, + "learning_rate": 0.0002, + "loss": 0.6875, + "step": 9816 + }, + { + "epoch": 3.095868811100599, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 0.8693, + "step": 9817 + }, + { + "epoch": 3.0961841690318512, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 0.7979, + "step": 9818 + }, + { + "epoch": 3.096499526963103, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 0.6691, + "step": 9819 + }, + { + "epoch": 3.0968148848943553, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 0.8037, + "step": 9820 + }, + { + "epoch": 3.097130242825607, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 0.7151, + "step": 9821 + }, + { + "epoch": 3.097445600756859, + "grad_norm": 0.33203125, + "learning_rate": 0.0002, + "loss": 0.6967, + "step": 9822 + }, + { + "epoch": 3.097760958688111, + "grad_norm": 0.369140625, + "learning_rate": 0.0002, + "loss": 0.7014, + "step": 9823 + }, + { + "epoch": 3.098076316619363, + "grad_norm": 0.3671875, + "learning_rate": 0.0002, + "loss": 0.7291, + "step": 9824 + }, + { + "epoch": 3.098391674550615, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 0.6688, + "step": 9825 + }, + { + "epoch": 3.098707032481867, + "grad_norm": 0.369140625, + "learning_rate": 0.0002, + "loss": 0.7784, + "step": 9826 + }, + { + "epoch": 3.099022390413119, + "grad_norm": 0.38671875, + "learning_rate": 0.0002, + "loss": 0.5853, + "step": 9827 + }, + { + "epoch": 3.0993377483443707, + "grad_norm": 0.42578125, + "learning_rate": 0.0002, + "loss": 0.6985, + "step": 9828 + }, + { + "epoch": 3.099653106275623, + "grad_norm": 0.46875, + "learning_rate": 0.0002, + "loss": 0.5999, + "step": 9829 + }, + { + "epoch": 3.0999684642068748, + "grad_norm": 0.44921875, + "learning_rate": 0.0002, + "loss": 0.7391, + "step": 9830 + }, + { + "epoch": 3.1002838221381266, + "grad_norm": 0.46875, + "learning_rate": 0.0002, + "loss": 0.6601, + "step": 9831 + }, + { + "epoch": 3.100599180069379, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 0.6311, + "step": 9832 + }, + { + "epoch": 3.1009145380006307, + "grad_norm": 0.515625, + "learning_rate": 0.0002, + "loss": 0.6252, + "step": 9833 + }, + { + "epoch": 3.1012298959318825, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 0.6005, + "step": 9834 + }, + { + "epoch": 3.1015452538631347, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 0.5963, + "step": 9835 + }, + { + "epoch": 3.1018606117943865, + "grad_norm": 0.58203125, + "learning_rate": 0.0002, + "loss": 0.7811, + "step": 9836 + }, + { + "epoch": 3.102175969725639, + "grad_norm": 0.68359375, + "learning_rate": 0.0002, + "loss": 0.7284, + "step": 9837 + }, + { + "epoch": 3.1024913276568906, + "grad_norm": 0.796875, + "learning_rate": 0.0002, + "loss": 0.6537, + "step": 9838 + }, + { + "epoch": 3.1028066855881424, + "grad_norm": 0.5390625, + "learning_rate": 0.0002, + "loss": 0.5044, + "step": 9839 + }, + { + "epoch": 3.1031220435193947, + "grad_norm": 0.9140625, + "learning_rate": 0.0002, + "loss": 0.6262, + "step": 9840 + }, + { + "epoch": 3.1034374014506465, + "grad_norm": 0.78125, + "learning_rate": 0.0002, + "loss": 0.6331, + "step": 9841 + }, + { + "epoch": 3.1037527593818983, + "grad_norm": 0.984375, + "learning_rate": 0.0002, + "loss": 0.5808, + "step": 9842 + }, + { + "epoch": 3.1040681173131506, + "grad_norm": 0.90234375, + "learning_rate": 0.0002, + "loss": 0.7453, + "step": 9843 + }, + { + "epoch": 3.1043834752444024, + "grad_norm": 0.97265625, + "learning_rate": 0.0002, + "loss": 0.7159, + "step": 9844 + }, + { + "epoch": 3.104698833175654, + "grad_norm": 0.87109375, + "learning_rate": 0.0002, + "loss": 0.7329, + "step": 9845 + }, + { + "epoch": 3.1050141911069065, + "grad_norm": 1.0234375, + "learning_rate": 0.0002, + "loss": 0.7484, + "step": 9846 + }, + { + "epoch": 3.1053295490381583, + "grad_norm": 1.15625, + "learning_rate": 0.0002, + "loss": 0.9697, + "step": 9847 + }, + { + "epoch": 3.10564490696941, + "grad_norm": 1.0625, + "learning_rate": 0.0002, + "loss": 0.8294, + "step": 9848 + }, + { + "epoch": 3.1059602649006623, + "grad_norm": 1.15625, + "learning_rate": 0.0002, + "loss": 0.8429, + "step": 9849 + }, + { + "epoch": 3.106275622831914, + "grad_norm": 1.1875, + "learning_rate": 0.0002, + "loss": 1.0871, + "step": 9850 + }, + { + "epoch": 3.1065909807631664, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 1.2559, + "step": 9851 + }, + { + "epoch": 3.1069063386944182, + "grad_norm": 1.3125, + "learning_rate": 0.0002, + "loss": 1.0318, + "step": 9852 + }, + { + "epoch": 3.10722169662567, + "grad_norm": 1.90625, + "learning_rate": 0.0002, + "loss": 0.9389, + "step": 9853 + }, + { + "epoch": 3.1075370545569223, + "grad_norm": 1.1171875, + "learning_rate": 0.0002, + "loss": 1.2723, + "step": 9854 + }, + { + "epoch": 3.107852412488174, + "grad_norm": 1.1171875, + "learning_rate": 0.0002, + "loss": 1.3645, + "step": 9855 + }, + { + "epoch": 3.108167770419426, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 1.2183, + "step": 9856 + }, + { + "epoch": 3.108483128350678, + "grad_norm": 1.8828125, + "learning_rate": 0.0002, + "loss": 1.3688, + "step": 9857 + }, + { + "epoch": 3.10879848628193, + "grad_norm": 1.1328125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9858 + }, + { + "epoch": 3.109113844213182, + "grad_norm": 1.2109375, + "learning_rate": 0.0002, + "loss": 1.1301, + "step": 9859 + }, + { + "epoch": 3.109429202144434, + "grad_norm": 2.21875, + "learning_rate": 0.0002, + "loss": 1.5874, + "step": 9860 + }, + { + "epoch": 3.109744560075686, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 1.4653, + "step": 9861 + }, + { + "epoch": 3.1100599180069377, + "grad_norm": 2.046875, + "learning_rate": 0.0002, + "loss": 1.1567, + "step": 9862 + }, + { + "epoch": 3.11037527593819, + "grad_norm": 1.359375, + "learning_rate": 0.0002, + "loss": 1.457, + "step": 9863 + }, + { + "epoch": 3.110690633869442, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 0.8867, + "step": 9864 + }, + { + "epoch": 3.1110059918006936, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 0.7683, + "step": 9865 + }, + { + "epoch": 3.111321349731946, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 0.7695, + "step": 9866 + }, + { + "epoch": 3.1116367076631977, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 0.7808, + "step": 9867 + }, + { + "epoch": 3.1119520655944495, + "grad_norm": 0.30859375, + "learning_rate": 0.0002, + "loss": 0.7406, + "step": 9868 + }, + { + "epoch": 3.1122674235257017, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 0.7167, + "step": 9869 + }, + { + "epoch": 3.1125827814569536, + "grad_norm": 0.388671875, + "learning_rate": 0.0002, + "loss": 0.7693, + "step": 9870 + }, + { + "epoch": 3.112898139388206, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 0.7389, + "step": 9871 + }, + { + "epoch": 3.1132134973194576, + "grad_norm": 0.3359375, + "learning_rate": 0.0002, + "loss": 0.7136, + "step": 9872 + }, + { + "epoch": 3.1135288552507094, + "grad_norm": 0.373046875, + "learning_rate": 0.0002, + "loss": 0.6948, + "step": 9873 + }, + { + "epoch": 3.1138442131819617, + "grad_norm": 0.4296875, + "learning_rate": 0.0002, + "loss": 0.755, + "step": 9874 + }, + { + "epoch": 3.1141595711132135, + "grad_norm": 0.396484375, + "learning_rate": 0.0002, + "loss": 0.7913, + "step": 9875 + }, + { + "epoch": 3.1144749290444653, + "grad_norm": 0.373046875, + "learning_rate": 0.0002, + "loss": 0.7299, + "step": 9876 + }, + { + "epoch": 3.1147902869757176, + "grad_norm": 0.40625, + "learning_rate": 0.0002, + "loss": 0.6224, + "step": 9877 + }, + { + "epoch": 3.1151056449069694, + "grad_norm": 0.48046875, + "learning_rate": 0.0002, + "loss": 0.7433, + "step": 9878 + }, + { + "epoch": 3.115421002838221, + "grad_norm": 0.439453125, + "learning_rate": 0.0002, + "loss": 0.6561, + "step": 9879 + }, + { + "epoch": 3.1157363607694735, + "grad_norm": 0.482421875, + "learning_rate": 0.0002, + "loss": 0.6703, + "step": 9880 + }, + { + "epoch": 3.1160517187007253, + "grad_norm": 0.48046875, + "learning_rate": 0.0002, + "loss": 0.7482, + "step": 9881 + }, + { + "epoch": 3.116367076631977, + "grad_norm": 0.470703125, + "learning_rate": 0.0002, + "loss": 0.6028, + "step": 9882 + }, + { + "epoch": 3.1166824345632294, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 0.6573, + "step": 9883 + }, + { + "epoch": 3.116997792494481, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 0.7135, + "step": 9884 + }, + { + "epoch": 3.1173131504257334, + "grad_norm": 0.89453125, + "learning_rate": 0.0002, + "loss": 0.6962, + "step": 9885 + }, + { + "epoch": 3.1176285083569852, + "grad_norm": 0.58203125, + "learning_rate": 0.0002, + "loss": 0.5804, + "step": 9886 + }, + { + "epoch": 3.117943866288237, + "grad_norm": 0.60546875, + "learning_rate": 0.0002, + "loss": 0.6361, + "step": 9887 + }, + { + "epoch": 3.1182592242194893, + "grad_norm": 0.59765625, + "learning_rate": 0.0002, + "loss": 0.6692, + "step": 9888 + }, + { + "epoch": 3.118574582150741, + "grad_norm": 0.640625, + "learning_rate": 0.0002, + "loss": 0.5918, + "step": 9889 + }, + { + "epoch": 3.118889940081993, + "grad_norm": 0.73828125, + "learning_rate": 0.0002, + "loss": 0.6562, + "step": 9890 + }, + { + "epoch": 3.119205298013245, + "grad_norm": 0.68359375, + "learning_rate": 0.0002, + "loss": 0.6131, + "step": 9891 + }, + { + "epoch": 3.119520655944497, + "grad_norm": 0.98046875, + "learning_rate": 0.0002, + "loss": 0.675, + "step": 9892 + }, + { + "epoch": 3.119836013875749, + "grad_norm": 0.640625, + "learning_rate": 0.0002, + "loss": 0.5631, + "step": 9893 + }, + { + "epoch": 3.120151371807001, + "grad_norm": 0.8828125, + "learning_rate": 0.0002, + "loss": 0.6476, + "step": 9894 + }, + { + "epoch": 3.120466729738253, + "grad_norm": 1.2578125, + "learning_rate": 0.0002, + "loss": 0.7292, + "step": 9895 + }, + { + "epoch": 3.1207820876695047, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 0.7775, + "step": 9896 + }, + { + "epoch": 3.121097445600757, + "grad_norm": 0.87890625, + "learning_rate": 0.0002, + "loss": 0.6019, + "step": 9897 + }, + { + "epoch": 3.121412803532009, + "grad_norm": 0.9296875, + "learning_rate": 0.0002, + "loss": 0.7849, + "step": 9898 + }, + { + "epoch": 3.1217281614632606, + "grad_norm": 0.84765625, + "learning_rate": 0.0002, + "loss": 0.6187, + "step": 9899 + }, + { + "epoch": 3.122043519394513, + "grad_norm": 1.03125, + "learning_rate": 0.0002, + "loss": 0.7959, + "step": 9900 + }, + { + "epoch": 3.1223588773257647, + "grad_norm": 1.921875, + "learning_rate": 0.0002, + "loss": 1.0467, + "step": 9901 + }, + { + "epoch": 3.122674235257017, + "grad_norm": 1.3125, + "learning_rate": 0.0002, + "loss": 1.0689, + "step": 9902 + }, + { + "epoch": 3.1229895931882687, + "grad_norm": 1.140625, + "learning_rate": 0.0002, + "loss": 1.0334, + "step": 9903 + }, + { + "epoch": 3.1233049511195206, + "grad_norm": 1.6328125, + "learning_rate": 0.0002, + "loss": 1.2073, + "step": 9904 + }, + { + "epoch": 3.123620309050773, + "grad_norm": 1.0859375, + "learning_rate": 0.0002, + "loss": 0.9743, + "step": 9905 + }, + { + "epoch": 3.1239356669820246, + "grad_norm": 1.3515625, + "learning_rate": 0.0002, + "loss": 1.3744, + "step": 9906 + }, + { + "epoch": 3.1242510249132764, + "grad_norm": 1.5859375, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 9907 + }, + { + "epoch": 3.1245663828445287, + "grad_norm": 1.1640625, + "learning_rate": 0.0002, + "loss": 1.286, + "step": 9908 + }, + { + "epoch": 3.1248817407757805, + "grad_norm": 1.375, + "learning_rate": 0.0002, + "loss": 1.1599, + "step": 9909 + }, + { + "epoch": 3.1251970987070323, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 1.1013, + "step": 9910 + }, + { + "epoch": 3.1255124566382846, + "grad_norm": 1.5, + "learning_rate": 0.0002, + "loss": 1.4274, + "step": 9911 + }, + { + "epoch": 3.1255124566382846, + "eval_loss": 2.024993896484375, + "eval_runtime": 148.9102, + "eval_samples_per_second": 6.715, + "eval_steps_per_second": 6.715, + "step": 9911 + }, + { + "epoch": 3.1255124566382846, + "mmlu_eval_accuracy": 0.5857903615316072, + "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727, + "mmlu_eval_accuracy_anatomy": 0.5714285714285714, + "mmlu_eval_accuracy_astronomy": 0.625, + "mmlu_eval_accuracy_business_ethics": 0.9090909090909091, + "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138, + "mmlu_eval_accuracy_college_biology": 0.8125, + "mmlu_eval_accuracy_college_chemistry": 0.375, + "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727, + "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453, + "mmlu_eval_accuracy_college_medicine": 0.5454545454545454, + "mmlu_eval_accuracy_college_physics": 0.45454545454545453, + "mmlu_eval_accuracy_computer_security": 0.45454545454545453, + "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156, + "mmlu_eval_accuracy_econometrics": 0.4166666666666667, + "mmlu_eval_accuracy_electrical_engineering": 0.5625, + "mmlu_eval_accuracy_elementary_mathematics": 0.3170731707317073, + "mmlu_eval_accuracy_formal_logic": 0.21428571428571427, + "mmlu_eval_accuracy_global_facts": 0.3, + "mmlu_eval_accuracy_high_school_biology": 0.59375, + "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091, + "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666, + "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222, + "mmlu_eval_accuracy_high_school_geography": 0.9545454545454546, + "mmlu_eval_accuracy_high_school_government_and_politics": 0.9047619047619048, + "mmlu_eval_accuracy_high_school_macroeconomics": 0.6744186046511628, + "mmlu_eval_accuracy_high_school_mathematics": 0.3793103448275862, + "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384, + "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882, + "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667, + "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216, + "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727, + "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923, + "mmlu_eval_accuracy_human_aging": 0.6086956521739131, + "mmlu_eval_accuracy_human_sexuality": 0.5, + "mmlu_eval_accuracy_international_law": 1.0, + "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453, + "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666, + "mmlu_eval_accuracy_machine_learning": 0.36363636363636365, + "mmlu_eval_accuracy_management": 0.9090909090909091, + "mmlu_eval_accuracy_marketing": 0.72, + "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091, + "mmlu_eval_accuracy_miscellaneous": 0.686046511627907, + "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842, + "mmlu_eval_accuracy_moral_scenarios": 0.27, + "mmlu_eval_accuracy_nutrition": 0.7272727272727273, + "mmlu_eval_accuracy_philosophy": 0.5882352941176471, + "mmlu_eval_accuracy_prehistory": 0.6285714285714286, + "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644, + "mmlu_eval_accuracy_professional_law": 0.38823529411764707, + "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516, + "mmlu_eval_accuracy_professional_psychology": 0.6231884057971014, + "mmlu_eval_accuracy_public_relations": 0.5833333333333334, + "mmlu_eval_accuracy_security_studies": 0.6296296296296297, + "mmlu_eval_accuracy_sociology": 1.0, + "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182, + "mmlu_eval_accuracy_virology": 0.5, + "mmlu_eval_accuracy_world_religions": 0.7894736842105263, + "mmlu_loss": 6.094484817273161, + "step": 9911 + }, + { + "epoch": 3.1258278145695364, + "grad_norm": 1.25, + "learning_rate": 0.0002, + "loss": 1.1503, + "step": 9912 + }, + { + "epoch": 3.126143172500788, + "grad_norm": 1.140625, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 9913 + }, + { + "epoch": 3.1264585304320405, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 0.8948, + "step": 9914 + }, + { + "epoch": 3.1267738883632923, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 0.8834, + "step": 9915 + }, + { + "epoch": 3.127089246294544, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 0.818, + "step": 9916 + }, + { + "epoch": 3.1274046042257964, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 0.7567, + "step": 9917 + }, + { + "epoch": 3.127719962157048, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 0.7605, + "step": 9918 + }, + { + "epoch": 3.1280353200883004, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 0.6732, + "step": 9919 + }, + { + "epoch": 3.1283506780195522, + "grad_norm": 0.330078125, + "learning_rate": 0.0002, + "loss": 0.7421, + "step": 9920 + }, + { + "epoch": 3.128666035950804, + "grad_norm": 0.361328125, + "learning_rate": 0.0002, + "loss": 0.7421, + "step": 9921 + }, + { + "epoch": 3.1289813938820563, + "grad_norm": 0.353515625, + "learning_rate": 0.0002, + "loss": 0.6806, + "step": 9922 + }, + { + "epoch": 3.129296751813308, + "grad_norm": 0.4140625, + "learning_rate": 0.0002, + "loss": 0.8404, + "step": 9923 + }, + { + "epoch": 3.12961210974456, + "grad_norm": 0.361328125, + "learning_rate": 0.0002, + "loss": 0.7295, + "step": 9924 + }, + { + "epoch": 3.129927467675812, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 0.6858, + "step": 9925 + }, + { + "epoch": 3.130242825607064, + "grad_norm": 0.388671875, + "learning_rate": 0.0002, + "loss": 0.6325, + "step": 9926 + }, + { + "epoch": 3.130558183538316, + "grad_norm": 0.462890625, + "learning_rate": 0.0002, + "loss": 0.6999, + "step": 9927 + }, + { + "epoch": 3.130873541469568, + "grad_norm": 0.435546875, + "learning_rate": 0.0002, + "loss": 0.6017, + "step": 9928 + }, + { + "epoch": 3.13118889940082, + "grad_norm": 0.419921875, + "learning_rate": 0.0002, + "loss": 0.5765, + "step": 9929 + }, + { + "epoch": 3.1315042573320717, + "grad_norm": 0.56640625, + "learning_rate": 0.0002, + "loss": 0.834, + "step": 9930 + }, + { + "epoch": 3.131819615263324, + "grad_norm": 0.64453125, + "learning_rate": 0.0002, + "loss": 0.6319, + "step": 9931 + }, + { + "epoch": 3.132134973194576, + "grad_norm": 0.5546875, + "learning_rate": 0.0002, + "loss": 0.6042, + "step": 9932 + }, + { + "epoch": 3.1324503311258276, + "grad_norm": 0.53125, + "learning_rate": 0.0002, + "loss": 0.6103, + "step": 9933 + }, + { + "epoch": 3.13276568905708, + "grad_norm": 0.62109375, + "learning_rate": 0.0002, + "loss": 0.5482, + "step": 9934 + }, + { + "epoch": 3.1330810469883317, + "grad_norm": 0.63671875, + "learning_rate": 0.0002, + "loss": 0.6609, + "step": 9935 + }, + { + "epoch": 3.133396404919584, + "grad_norm": 0.66796875, + "learning_rate": 0.0002, + "loss": 0.5118, + "step": 9936 + }, + { + "epoch": 3.1337117628508357, + "grad_norm": 0.78125, + "learning_rate": 0.0002, + "loss": 0.6242, + "step": 9937 + }, + { + "epoch": 3.1340271207820876, + "grad_norm": 0.703125, + "learning_rate": 0.0002, + "loss": 0.6517, + "step": 9938 + }, + { + "epoch": 3.13434247871334, + "grad_norm": 0.78515625, + "learning_rate": 0.0002, + "loss": 0.6681, + "step": 9939 + }, + { + "epoch": 3.1346578366445916, + "grad_norm": 0.765625, + "learning_rate": 0.0002, + "loss": 0.6692, + "step": 9940 + }, + { + "epoch": 3.1349731945758434, + "grad_norm": 0.80078125, + "learning_rate": 0.0002, + "loss": 0.6131, + "step": 9941 + }, + { + "epoch": 3.1352885525070957, + "grad_norm": 1.0, + "learning_rate": 0.0002, + "loss": 0.7451, + "step": 9942 + }, + { + "epoch": 3.1356039104383475, + "grad_norm": 0.7890625, + "learning_rate": 0.0002, + "loss": 0.7483, + "step": 9943 + }, + { + "epoch": 3.1359192683695993, + "grad_norm": 0.8046875, + "learning_rate": 0.0002, + "loss": 1.0, + "step": 9944 + }, + { + "epoch": 3.1362346263008516, + "grad_norm": 0.9375, + "learning_rate": 0.0002, + "loss": 0.8213, + "step": 9945 + }, + { + "epoch": 3.1365499842321034, + "grad_norm": 0.671875, + "learning_rate": 0.0002, + "loss": 0.8172, + "step": 9946 + }, + { + "epoch": 3.136865342163355, + "grad_norm": 1.015625, + "learning_rate": 0.0002, + "loss": 0.8236, + "step": 9947 + }, + { + "epoch": 3.1371807000946075, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 0.8943, + "step": 9948 + }, + { + "epoch": 3.1374960580258593, + "grad_norm": 1.359375, + "learning_rate": 0.0002, + "loss": 1.1422, + "step": 9949 + }, + { + "epoch": 3.1378114159571115, + "grad_norm": 1.3046875, + "learning_rate": 0.0002, + "loss": 0.9237, + "step": 9950 + }, + { + "epoch": 3.1381267738883634, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 0.907, + "step": 9951 + }, + { + "epoch": 3.138442131819615, + "grad_norm": 1.2421875, + "learning_rate": 0.0002, + "loss": 1.0166, + "step": 9952 + }, + { + "epoch": 3.1387574897508674, + "grad_norm": 0.96875, + "learning_rate": 0.0002, + "loss": 1.0034, + "step": 9953 + }, + { + "epoch": 3.1390728476821192, + "grad_norm": 1.25, + "learning_rate": 0.0002, + "loss": 1.278, + "step": 9954 + }, + { + "epoch": 3.139388205613371, + "grad_norm": 0.8984375, + "learning_rate": 0.0002, + "loss": 1.0871, + "step": 9955 + }, + { + "epoch": 3.1397035635446233, + "grad_norm": 2.171875, + "learning_rate": 0.0002, + "loss": 1.1751, + "step": 9956 + }, + { + "epoch": 3.140018921475875, + "grad_norm": 0.96875, + "learning_rate": 0.0002, + "loss": 1.1824, + "step": 9957 + }, + { + "epoch": 3.140334279407127, + "grad_norm": 0.9375, + "learning_rate": 0.0002, + "loss": 1.0635, + "step": 9958 + }, + { + "epoch": 3.140649637338379, + "grad_norm": 1.171875, + "learning_rate": 0.0002, + "loss": 1.1221, + "step": 9959 + }, + { + "epoch": 3.140964995269631, + "grad_norm": 1.1484375, + "learning_rate": 0.0002, + "loss": 1.1046, + "step": 9960 + }, + { + "epoch": 3.141280353200883, + "grad_norm": 1.078125, + "learning_rate": 0.0002, + "loss": 1.0763, + "step": 9961 + }, + { + "epoch": 3.141595711132135, + "grad_norm": 1.140625, + "learning_rate": 0.0002, + "loss": 1.0586, + "step": 9962 + }, + { + "epoch": 3.141911069063387, + "grad_norm": 1.734375, + "learning_rate": 0.0002, + "loss": 1.5955, + "step": 9963 + }, + { + "epoch": 3.1422264269946387, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 0.8511, + "step": 9964 + }, + { + "epoch": 3.142541784925891, + "grad_norm": 0.330078125, + "learning_rate": 0.0002, + "loss": 0.8113, + "step": 9965 + }, + { + "epoch": 3.142857142857143, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 0.7415, + "step": 9966 + }, + { + "epoch": 3.1431725007883946, + "grad_norm": 0.36328125, + "learning_rate": 0.0002, + "loss": 0.7113, + "step": 9967 + }, + { + "epoch": 3.143487858719647, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 0.7799, + "step": 9968 + }, + { + "epoch": 3.1438032166508987, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 0.7077, + "step": 9969 + }, + { + "epoch": 3.144118574582151, + "grad_norm": 0.349609375, + "learning_rate": 0.0002, + "loss": 0.737, + "step": 9970 + }, + { + "epoch": 3.1444339325134028, + "grad_norm": 0.3984375, + "learning_rate": 0.0002, + "loss": 0.814, + "step": 9971 + }, + { + "epoch": 3.1447492904446546, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 0.7808, + "step": 9972 + }, + { + "epoch": 3.145064648375907, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 0.7805, + "step": 9973 + }, + { + "epoch": 3.1453800063071586, + "grad_norm": 0.361328125, + "learning_rate": 0.0002, + "loss": 0.7091, + "step": 9974 + }, + { + "epoch": 3.1456953642384105, + "grad_norm": 0.396484375, + "learning_rate": 0.0002, + "loss": 0.7859, + "step": 9975 + }, + { + "epoch": 3.1460107221696627, + "grad_norm": 0.396484375, + "learning_rate": 0.0002, + "loss": 0.769, + "step": 9976 + }, + { + "epoch": 3.1463260801009145, + "grad_norm": 0.3984375, + "learning_rate": 0.0002, + "loss": 0.6831, + "step": 9977 + }, + { + "epoch": 3.1466414380321663, + "grad_norm": 0.404296875, + "learning_rate": 0.0002, + "loss": 0.5702, + "step": 9978 + }, + { + "epoch": 3.1469567959634186, + "grad_norm": 0.447265625, + "learning_rate": 0.0002, + "loss": 0.6944, + "step": 9979 + }, + { + "epoch": 3.1472721538946704, + "grad_norm": 0.443359375, + "learning_rate": 0.0002, + "loss": 0.6896, + "step": 9980 + }, + { + "epoch": 3.1475875118259222, + "grad_norm": 0.46484375, + "learning_rate": 0.0002, + "loss": 0.6684, + "step": 9981 + }, + { + "epoch": 3.1479028697571745, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 0.6548, + "step": 9982 + }, + { + "epoch": 3.1482182276884263, + "grad_norm": 0.52734375, + "learning_rate": 0.0002, + "loss": 0.6118, + "step": 9983 + }, + { + "epoch": 3.1485335856196786, + "grad_norm": 0.47265625, + "learning_rate": 0.0002, + "loss": 0.6063, + "step": 9984 + }, + { + "epoch": 3.1488489435509304, + "grad_norm": 0.5625, + "learning_rate": 0.0002, + "loss": 0.6446, + "step": 9985 + }, + { + "epoch": 3.149164301482182, + "grad_norm": 0.5703125, + "learning_rate": 0.0002, + "loss": 0.676, + "step": 9986 + }, + { + "epoch": 3.1494796594134344, + "grad_norm": 0.7265625, + "learning_rate": 0.0002, + "loss": 0.6466, + "step": 9987 + }, + { + "epoch": 3.1497950173446863, + "grad_norm": 0.578125, + "learning_rate": 0.0002, + "loss": 0.6153, + "step": 9988 + }, + { + "epoch": 3.150110375275938, + "grad_norm": 0.62890625, + "learning_rate": 0.0002, + "loss": 0.4852, + "step": 9989 + }, + { + "epoch": 3.1504257332071903, + "grad_norm": 0.7734375, + "learning_rate": 0.0002, + "loss": 0.5952, + "step": 9990 + }, + { + "epoch": 3.150741091138442, + "grad_norm": 0.68359375, + "learning_rate": 0.0002, + "loss": 0.6719, + "step": 9991 + }, + { + "epoch": 3.151056449069694, + "grad_norm": 0.84765625, + "learning_rate": 0.0002, + "loss": 0.6491, + "step": 9992 + }, + { + "epoch": 3.151371807000946, + "grad_norm": 0.8046875, + "learning_rate": 0.0002, + "loss": 0.6762, + "step": 9993 + }, + { + "epoch": 3.151687164932198, + "grad_norm": 1.109375, + "learning_rate": 0.0002, + "loss": 0.6445, + "step": 9994 + }, + { + "epoch": 3.15200252286345, + "grad_norm": 1.046875, + "learning_rate": 0.0002, + "loss": 0.7611, + "step": 9995 + }, + { + "epoch": 3.152317880794702, + "grad_norm": 0.890625, + "learning_rate": 0.0002, + "loss": 0.9491, + "step": 9996 + }, + { + "epoch": 3.152633238725954, + "grad_norm": 1.125, + "learning_rate": 0.0002, + "loss": 0.7131, + "step": 9997 + }, + { + "epoch": 3.1529485966572057, + "grad_norm": 1.328125, + "learning_rate": 0.0002, + "loss": 0.9073, + "step": 9998 + }, + { + "epoch": 3.153263954588458, + "grad_norm": 1.28125, + "learning_rate": 0.0002, + "loss": 0.9119, + "step": 9999 + }, + { + "epoch": 3.15357931251971, + "grad_norm": 1.203125, + "learning_rate": 0.0002, + "loss": 0.8966, + "step": 10000 + }, + { + "epoch": 3.15357931251971, + "step": 10000, + "total_flos": 1.1687984226206024e+18, + "train_loss": 1.3553110431969166, + "train_runtime": 191744.1538, + "train_samples_per_second": 0.834, + "train_steps_per_second": 0.052 + } + ], + "logging_steps": 1, + "max_steps": 10000, + "num_input_tokens_seen": 0, + "num_train_epochs": 4, + "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": 1.1687984226206024e+18, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +}