| { | |
| "best_metric": 0.13593938946723938, | |
| "best_model_checkpoint": "saves/psy-course-info/MentaLLaMA-chat-7B/train/fold10/checkpoint-120", | |
| "epoch": 4.938271604938271, | |
| "eval_steps": 10, | |
| "global_step": 125, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03950617283950617, | |
| "grad_norm": 0.8390358686447144, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 1.0419, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.07901234567901234, | |
| "grad_norm": 0.8197033405303955, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.9963, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.11851851851851852, | |
| "grad_norm": 0.9033790230751038, | |
| "learning_rate": 2.307692307692308e-05, | |
| "loss": 0.9637, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.1580246913580247, | |
| "grad_norm": 0.8509941697120667, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 1.0582, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.19753086419753085, | |
| "grad_norm": 0.8880956172943115, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 1.04, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.23703703703703705, | |
| "grad_norm": 0.9652888178825378, | |
| "learning_rate": 4.615384615384616e-05, | |
| "loss": 1.0651, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.2765432098765432, | |
| "grad_norm": 1.030725121498108, | |
| "learning_rate": 5.384615384615385e-05, | |
| "loss": 0.9158, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.3160493827160494, | |
| "grad_norm": 1.2006186246871948, | |
| "learning_rate": 6.153846153846155e-05, | |
| "loss": 0.9412, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.35555555555555557, | |
| "grad_norm": 1.4533060789108276, | |
| "learning_rate": 6.923076923076924e-05, | |
| "loss": 0.9712, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.3950617283950617, | |
| "grad_norm": 1.1821893453598022, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 0.7237, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3950617283950617, | |
| "eval_loss": 0.6588842868804932, | |
| "eval_runtime": 23.1164, | |
| "eval_samples_per_second": 1.947, | |
| "eval_steps_per_second": 1.947, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.4345679012345679, | |
| "grad_norm": 1.2356246709823608, | |
| "learning_rate": 8.461538461538461e-05, | |
| "loss": 0.6476, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.4740740740740741, | |
| "grad_norm": 0.8601561188697815, | |
| "learning_rate": 9.230769230769232e-05, | |
| "loss": 0.5812, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.5135802469135803, | |
| "grad_norm": 0.8983665108680725, | |
| "learning_rate": 0.0001, | |
| "loss": 0.4996, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.5530864197530864, | |
| "grad_norm": 0.628352701663971, | |
| "learning_rate": 9.998033131915266e-05, | |
| "loss": 0.3582, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.5925925925925926, | |
| "grad_norm": 0.49270370602607727, | |
| "learning_rate": 9.992134075089084e-05, | |
| "loss": 0.3912, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.6320987654320988, | |
| "grad_norm": 0.5510178208351135, | |
| "learning_rate": 9.982307470588098e-05, | |
| "loss": 0.2828, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.671604938271605, | |
| "grad_norm": 0.5726563334465027, | |
| "learning_rate": 9.968561049466214e-05, | |
| "loss": 0.3039, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.7111111111111111, | |
| "grad_norm": 0.5539241433143616, | |
| "learning_rate": 9.950905626682228e-05, | |
| "loss": 0.2842, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.7506172839506173, | |
| "grad_norm": 0.44315189123153687, | |
| "learning_rate": 9.92935509259118e-05, | |
| "loss": 0.2978, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.7901234567901234, | |
| "grad_norm": 0.39687928557395935, | |
| "learning_rate": 9.903926402016153e-05, | |
| "loss": 0.2412, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.7901234567901234, | |
| "eval_loss": 0.27791768312454224, | |
| "eval_runtime": 23.1717, | |
| "eval_samples_per_second": 1.942, | |
| "eval_steps_per_second": 1.942, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.8296296296296296, | |
| "grad_norm": 0.5034745335578918, | |
| "learning_rate": 9.874639560909117e-05, | |
| "loss": 0.3352, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.8691358024691358, | |
| "grad_norm": 0.5247141122817993, | |
| "learning_rate": 9.841517610611309e-05, | |
| "loss": 0.2029, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.908641975308642, | |
| "grad_norm": 0.47732657194137573, | |
| "learning_rate": 9.804586609725499e-05, | |
| "loss": 0.2955, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.9481481481481482, | |
| "grad_norm": 0.44496268033981323, | |
| "learning_rate": 9.763875613614482e-05, | |
| "loss": 0.2153, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.9876543209876543, | |
| "grad_norm": 0.40034201741218567, | |
| "learning_rate": 9.719416651541839e-05, | |
| "loss": 0.2228, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.0271604938271606, | |
| "grad_norm": 0.8534076809883118, | |
| "learning_rate": 9.671244701472999e-05, | |
| "loss": 0.3916, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.0666666666666667, | |
| "grad_norm": 0.39993032813072205, | |
| "learning_rate": 9.619397662556435e-05, | |
| "loss": 0.2572, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 1.106172839506173, | |
| "grad_norm": 0.2911875545978546, | |
| "learning_rate": 9.563916325306594e-05, | |
| "loss": 0.187, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.145679012345679, | |
| "grad_norm": 0.37117239832878113, | |
| "learning_rate": 9.504844339512095e-05, | |
| "loss": 0.2074, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "grad_norm": 0.31329572200775146, | |
| "learning_rate": 9.442228179894362e-05, | |
| "loss": 0.1983, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "eval_loss": 0.21022391319274902, | |
| "eval_runtime": 23.1557, | |
| "eval_samples_per_second": 1.943, | |
| "eval_steps_per_second": 1.943, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.2246913580246914, | |
| "grad_norm": 0.3860533535480499, | |
| "learning_rate": 9.376117109543769e-05, | |
| "loss": 0.1528, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.2641975308641975, | |
| "grad_norm": 0.37998268008232117, | |
| "learning_rate": 9.306563141162046e-05, | |
| "loss": 0.198, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.3037037037037038, | |
| "grad_norm": 0.3685644865036011, | |
| "learning_rate": 9.233620996141421e-05, | |
| "loss": 0.2229, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.34320987654321, | |
| "grad_norm": 0.3412479758262634, | |
| "learning_rate": 9.157348061512727e-05, | |
| "loss": 0.2049, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.382716049382716, | |
| "grad_norm": 0.4064359664916992, | |
| "learning_rate": 9.077804344796302e-05, | |
| "loss": 0.1855, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.4222222222222223, | |
| "grad_norm": 0.3913111984729767, | |
| "learning_rate": 8.995052426791247e-05, | |
| "loss": 0.1711, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.4617283950617284, | |
| "grad_norm": 0.40914851427078247, | |
| "learning_rate": 8.90915741234015e-05, | |
| "loss": 0.1906, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.5012345679012347, | |
| "grad_norm": 0.312428742647171, | |
| "learning_rate": 8.820186879108038e-05, | |
| "loss": 0.1446, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.5407407407407407, | |
| "grad_norm": 0.32728713750839233, | |
| "learning_rate": 8.728210824415827e-05, | |
| "loss": 0.1503, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.5802469135802468, | |
| "grad_norm": 0.4225092828273773, | |
| "learning_rate": 8.633301610170135e-05, | |
| "loss": 0.2083, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.5802469135802468, | |
| "eval_loss": 0.1740482896566391, | |
| "eval_runtime": 23.1603, | |
| "eval_samples_per_second": 1.943, | |
| "eval_steps_per_second": 1.943, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.6197530864197531, | |
| "grad_norm": 0.3866317868232727, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 0.1906, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.6592592592592592, | |
| "grad_norm": 0.2778302729129791, | |
| "learning_rate": 8.434984630174509e-05, | |
| "loss": 0.1146, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.6987654320987655, | |
| "grad_norm": 0.31596770882606506, | |
| "learning_rate": 8.33173288976002e-05, | |
| "loss": 0.14, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.7382716049382716, | |
| "grad_norm": 0.41009825468063354, | |
| "learning_rate": 8.225859917710439e-05, | |
| "loss": 0.1683, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 0.427985280752182, | |
| "learning_rate": 8.117449009293668e-05, | |
| "loss": 0.1483, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.817283950617284, | |
| "grad_norm": 0.3274437189102173, | |
| "learning_rate": 8.006585456492029e-05, | |
| "loss": 0.1846, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.8567901234567903, | |
| "grad_norm": 0.2898247539997101, | |
| "learning_rate": 7.89335648089903e-05, | |
| "loss": 0.134, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.8962962962962964, | |
| "grad_norm": 0.36800283193588257, | |
| "learning_rate": 7.777851165098012e-05, | |
| "loss": 0.1905, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.9358024691358025, | |
| "grad_norm": 0.22297610342502594, | |
| "learning_rate": 7.660160382576683e-05, | |
| "loss": 0.1043, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.9753086419753085, | |
| "grad_norm": 0.3482544422149658, | |
| "learning_rate": 7.540376726232648e-05, | |
| "loss": 0.1367, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.9753086419753085, | |
| "eval_loss": 0.15801812708377838, | |
| "eval_runtime": 23.1355, | |
| "eval_samples_per_second": 1.945, | |
| "eval_steps_per_second": 1.945, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.0148148148148146, | |
| "grad_norm": 0.6955103278160095, | |
| "learning_rate": 7.4185944355262e-05, | |
| "loss": 0.2412, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 2.054320987654321, | |
| "grad_norm": 0.3053165078163147, | |
| "learning_rate": 7.294909322337689e-05, | |
| "loss": 0.1151, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 2.093827160493827, | |
| "grad_norm": 0.3078591823577881, | |
| "learning_rate": 7.169418695587791e-05, | |
| "loss": 0.1326, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 2.1333333333333333, | |
| "grad_norm": 0.3687678575515747, | |
| "learning_rate": 7.042221284679982e-05, | |
| "loss": 0.1802, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 2.1728395061728394, | |
| "grad_norm": 0.2860524654388428, | |
| "learning_rate": 6.91341716182545e-05, | |
| "loss": 0.1332, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.212345679012346, | |
| "grad_norm": 0.39456039667129517, | |
| "learning_rate": 6.783107663311565e-05, | |
| "loss": 0.1674, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 2.251851851851852, | |
| "grad_norm": 0.39261484146118164, | |
| "learning_rate": 6.651395309775837e-05, | |
| "loss": 0.1821, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.291358024691358, | |
| "grad_norm": 0.3944259285926819, | |
| "learning_rate": 6.518383725548074e-05, | |
| "loss": 0.1483, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 2.330864197530864, | |
| "grad_norm": 0.30228763818740845, | |
| "learning_rate": 6.384177557124247e-05, | |
| "loss": 0.1152, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 2.3703703703703702, | |
| "grad_norm": 0.39069655537605286, | |
| "learning_rate": 6.248882390836135e-05, | |
| "loss": 0.1382, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.3703703703703702, | |
| "eval_loss": 0.14894644916057587, | |
| "eval_runtime": 23.1056, | |
| "eval_samples_per_second": 1.948, | |
| "eval_steps_per_second": 1.948, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.4098765432098768, | |
| "grad_norm": 0.3270443379878998, | |
| "learning_rate": 6.112604669781572e-05, | |
| "loss": 0.1228, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.449382716049383, | |
| "grad_norm": 0.24159280955791473, | |
| "learning_rate": 5.9754516100806423e-05, | |
| "loss": 0.1038, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 2.488888888888889, | |
| "grad_norm": 0.28394487500190735, | |
| "learning_rate": 5.837531116523682e-05, | |
| "loss": 0.1023, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 2.528395061728395, | |
| "grad_norm": 0.39842841029167175, | |
| "learning_rate": 5.698951697677498e-05, | |
| "loss": 0.1386, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 2.567901234567901, | |
| "grad_norm": 0.3707628548145294, | |
| "learning_rate": 5.559822380516539e-05, | |
| "loss": 0.1089, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.6074074074074076, | |
| "grad_norm": 0.28499382734298706, | |
| "learning_rate": 5.420252624646238e-05, | |
| "loss": 0.0947, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.6469135802469137, | |
| "grad_norm": 0.3631799519062042, | |
| "learning_rate": 5.2803522361859594e-05, | |
| "loss": 0.1038, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.68641975308642, | |
| "grad_norm": 0.3704788088798523, | |
| "learning_rate": 5.140231281379345e-05, | |
| "loss": 0.1339, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.725925925925926, | |
| "grad_norm": 0.2988777458667755, | |
| "learning_rate": 5e-05, | |
| "loss": 0.1194, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.765432098765432, | |
| "grad_norm": 0.38090386986732483, | |
| "learning_rate": 4.859768718620656e-05, | |
| "loss": 0.1333, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.765432098765432, | |
| "eval_loss": 0.14429892599582672, | |
| "eval_runtime": 23.1836, | |
| "eval_samples_per_second": 1.941, | |
| "eval_steps_per_second": 1.941, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.8049382716049385, | |
| "grad_norm": 0.34871533513069153, | |
| "learning_rate": 4.7196477638140404e-05, | |
| "loss": 0.1027, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.8444444444444446, | |
| "grad_norm": 0.46063342690467834, | |
| "learning_rate": 4.579747375353763e-05, | |
| "loss": 0.1424, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.8839506172839506, | |
| "grad_norm": 0.35923025012016296, | |
| "learning_rate": 4.4401776194834613e-05, | |
| "loss": 0.1311, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.9234567901234567, | |
| "grad_norm": 0.3711896240711212, | |
| "learning_rate": 4.3010483023225045e-05, | |
| "loss": 0.1223, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.962962962962963, | |
| "grad_norm": 0.38326412439346313, | |
| "learning_rate": 4.162468883476319e-05, | |
| "loss": 0.1131, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 3.0024691358024693, | |
| "grad_norm": 0.7552617788314819, | |
| "learning_rate": 4.0245483899193595e-05, | |
| "loss": 0.1576, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 3.0419753086419754, | |
| "grad_norm": 0.3829554319381714, | |
| "learning_rate": 3.887395330218429e-05, | |
| "loss": 0.115, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 3.0814814814814815, | |
| "grad_norm": 0.2932409346103668, | |
| "learning_rate": 3.7511176091638653e-05, | |
| "loss": 0.1045, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 3.1209876543209876, | |
| "grad_norm": 0.2691555321216583, | |
| "learning_rate": 3.6158224428757535e-05, | |
| "loss": 0.0897, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 3.1604938271604937, | |
| "grad_norm": 0.3032447397708893, | |
| "learning_rate": 3.4816162744519263e-05, | |
| "loss": 0.0986, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.1604938271604937, | |
| "eval_loss": 0.14191414415836334, | |
| "eval_runtime": 23.164, | |
| "eval_samples_per_second": 1.943, | |
| "eval_steps_per_second": 1.943, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.3649228513240814, | |
| "learning_rate": 3.3486046902241664e-05, | |
| "loss": 0.0928, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 3.2395061728395063, | |
| "grad_norm": 0.366616815328598, | |
| "learning_rate": 3.216892336688435e-05, | |
| "loss": 0.1232, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 3.2790123456790123, | |
| "grad_norm": 0.4049418270587921, | |
| "learning_rate": 3.086582838174551e-05, | |
| "loss": 0.1014, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 3.3185185185185184, | |
| "grad_norm": 0.37535908818244934, | |
| "learning_rate": 2.9577787153200197e-05, | |
| "loss": 0.1063, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 3.3580246913580245, | |
| "grad_norm": 0.30204907059669495, | |
| "learning_rate": 2.8305813044122097e-05, | |
| "loss": 0.0966, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 3.397530864197531, | |
| "grad_norm": 0.3174695074558258, | |
| "learning_rate": 2.705090677662311e-05, | |
| "loss": 0.1036, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 3.437037037037037, | |
| "grad_norm": 0.3384718596935272, | |
| "learning_rate": 2.581405564473801e-05, | |
| "loss": 0.1499, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 3.476543209876543, | |
| "grad_norm": 0.3292217254638672, | |
| "learning_rate": 2.459623273767354e-05, | |
| "loss": 0.088, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 3.5160493827160493, | |
| "grad_norm": 0.28059694170951843, | |
| "learning_rate": 2.3398396174233178e-05, | |
| "loss": 0.08, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "grad_norm": 0.3763943910598755, | |
| "learning_rate": 2.2221488349019903e-05, | |
| "loss": 0.0952, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "eval_loss": 0.13908639550209045, | |
| "eval_runtime": 23.1373, | |
| "eval_samples_per_second": 1.945, | |
| "eval_steps_per_second": 1.945, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.595061728395062, | |
| "grad_norm": 0.42026203870773315, | |
| "learning_rate": 2.1066435191009715e-05, | |
| "loss": 0.1162, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 3.634567901234568, | |
| "grad_norm": 0.2964298129081726, | |
| "learning_rate": 1.9934145435079702e-05, | |
| "loss": 0.0783, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 3.674074074074074, | |
| "grad_norm": 0.291548490524292, | |
| "learning_rate": 1.8825509907063327e-05, | |
| "loss": 0.0995, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 3.71358024691358, | |
| "grad_norm": 0.37142956256866455, | |
| "learning_rate": 1.774140082289563e-05, | |
| "loss": 0.0789, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 3.753086419753086, | |
| "grad_norm": 0.38268908858299255, | |
| "learning_rate": 1.6682671102399805e-05, | |
| "loss": 0.1274, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 3.7925925925925927, | |
| "grad_norm": 0.3901118040084839, | |
| "learning_rate": 1.5650153698254916e-05, | |
| "loss": 0.1262, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 3.832098765432099, | |
| "grad_norm": 0.30913418531417847, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 0.1148, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 3.871604938271605, | |
| "grad_norm": 0.43375906348228455, | |
| "learning_rate": 1.3666983898298657e-05, | |
| "loss": 0.1245, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 3.911111111111111, | |
| "grad_norm": 0.310648649930954, | |
| "learning_rate": 1.2717891755841722e-05, | |
| "loss": 0.1043, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 3.950617283950617, | |
| "grad_norm": 0.3259522020816803, | |
| "learning_rate": 1.1798131208919627e-05, | |
| "loss": 0.0878, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.950617283950617, | |
| "eval_loss": 0.1373450607061386, | |
| "eval_runtime": 23.1306, | |
| "eval_samples_per_second": 1.945, | |
| "eval_steps_per_second": 1.945, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.9901234567901236, | |
| "grad_norm": 0.41351118683815, | |
| "learning_rate": 1.090842587659851e-05, | |
| "loss": 0.1418, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 4.029629629629629, | |
| "grad_norm": 0.5473781824111938, | |
| "learning_rate": 1.004947573208756e-05, | |
| "loss": 0.1311, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 4.069135802469136, | |
| "grad_norm": 0.3283960819244385, | |
| "learning_rate": 9.221956552036992e-06, | |
| "loss": 0.085, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 4.108641975308642, | |
| "grad_norm": 0.28756096959114075, | |
| "learning_rate": 8.426519384872733e-06, | |
| "loss": 0.1015, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 4.148148148148148, | |
| "grad_norm": 0.31263357400894165, | |
| "learning_rate": 7.663790038585793e-06, | |
| "loss": 0.0641, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 4.187654320987654, | |
| "grad_norm": 0.37136438488960266, | |
| "learning_rate": 6.934368588379553e-06, | |
| "loss": 0.0895, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 4.22716049382716, | |
| "grad_norm": 0.37439683079719543, | |
| "learning_rate": 6.238828904562316e-06, | |
| "loss": 0.0847, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 4.266666666666667, | |
| "grad_norm": 0.29002976417541504, | |
| "learning_rate": 5.577718201056392e-06, | |
| "loss": 0.0768, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 4.306172839506173, | |
| "grad_norm": 0.31060555577278137, | |
| "learning_rate": 4.951556604879048e-06, | |
| "loss": 0.0935, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 4.345679012345679, | |
| "grad_norm": 0.3601924180984497, | |
| "learning_rate": 4.360836746934055e-06, | |
| "loss": 0.0892, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.345679012345679, | |
| "eval_loss": 0.1362769901752472, | |
| "eval_runtime": 23.1714, | |
| "eval_samples_per_second": 1.942, | |
| "eval_steps_per_second": 1.942, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.385185185185185, | |
| "grad_norm": 0.2731637954711914, | |
| "learning_rate": 3.8060233744356633e-06, | |
| "loss": 0.0741, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 4.424691358024692, | |
| "grad_norm": 0.369502991437912, | |
| "learning_rate": 3.2875529852700147e-06, | |
| "loss": 0.1249, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 4.4641975308641975, | |
| "grad_norm": 0.33320462703704834, | |
| "learning_rate": 2.8058334845816213e-06, | |
| "loss": 0.1029, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 4.503703703703704, | |
| "grad_norm": 0.391130656003952, | |
| "learning_rate": 2.361243863855184e-06, | |
| "loss": 0.109, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 4.54320987654321, | |
| "grad_norm": 0.35531750321388245, | |
| "learning_rate": 1.9541339027450256e-06, | |
| "loss": 0.0946, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 4.582716049382716, | |
| "grad_norm": 0.35947561264038086, | |
| "learning_rate": 1.584823893886933e-06, | |
| "loss": 0.1033, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 4.622222222222222, | |
| "grad_norm": 0.39352288842201233, | |
| "learning_rate": 1.2536043909088191e-06, | |
| "loss": 0.1055, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 4.661728395061728, | |
| "grad_norm": 0.3489411175251007, | |
| "learning_rate": 9.607359798384785e-07, | |
| "loss": 0.1007, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 4.701234567901235, | |
| "grad_norm": 0.31264597177505493, | |
| "learning_rate": 7.064490740882057e-07, | |
| "loss": 0.075, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 4.7407407407407405, | |
| "grad_norm": 0.3497534394264221, | |
| "learning_rate": 4.909437331777179e-07, | |
| "loss": 0.1224, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.7407407407407405, | |
| "eval_loss": 0.13593938946723938, | |
| "eval_runtime": 23.1707, | |
| "eval_samples_per_second": 1.942, | |
| "eval_steps_per_second": 1.942, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.780246913580247, | |
| "grad_norm": 0.37891966104507446, | |
| "learning_rate": 3.143895053378698e-07, | |
| "loss": 0.077, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 4.8197530864197535, | |
| "grad_norm": 0.3828718066215515, | |
| "learning_rate": 1.7692529411904578e-07, | |
| "loss": 0.1337, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 4.859259259259259, | |
| "grad_norm": 0.3418748378753662, | |
| "learning_rate": 7.865924910916977e-08, | |
| "loss": 0.0884, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 4.898765432098766, | |
| "grad_norm": 0.32816141843795776, | |
| "learning_rate": 1.9668680847356735e-08, | |
| "loss": 0.092, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 4.938271604938271, | |
| "grad_norm": 0.27710503339767456, | |
| "learning_rate": 0.0, | |
| "loss": 0.0774, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 4.938271604938271, | |
| "step": 125, | |
| "total_flos": 1.3796864787303629e+17, | |
| "train_loss": 0.22307482421398162, | |
| "train_runtime": 4046.1096, | |
| "train_samples_per_second": 0.5, | |
| "train_steps_per_second": 0.031 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 125, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 10, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.3796864787303629e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |