| { | |
| "best_metric": 0.10305963456630707, | |
| "best_model_checkpoint": "saves/psy-course-doc-info/MentaLLaMA-chat-7B/train/fold7/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": 0.27619925141334534, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.4228, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.07901234567901234, | |
| "grad_norm": 0.24047105014324188, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.3852, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.11851851851851852, | |
| "grad_norm": 0.3186410367488861, | |
| "learning_rate": 2.307692307692308e-05, | |
| "loss": 0.4819, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.1580246913580247, | |
| "grad_norm": 0.28512120246887207, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.4286, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.19753086419753085, | |
| "grad_norm": 0.8059149980545044, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 0.9093, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.23703703703703705, | |
| "grad_norm": 1.308341383934021, | |
| "learning_rate": 4.615384615384616e-05, | |
| "loss": 1.7638, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.2765432098765432, | |
| "grad_norm": 0.2687425911426544, | |
| "learning_rate": 5.384615384615385e-05, | |
| "loss": 0.3847, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.3160493827160494, | |
| "grad_norm": 0.3323095738887787, | |
| "learning_rate": 6.153846153846155e-05, | |
| "loss": 0.424, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.35555555555555557, | |
| "grad_norm": 0.35097813606262207, | |
| "learning_rate": 6.923076923076924e-05, | |
| "loss": 0.4095, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.3950617283950617, | |
| "grad_norm": 0.2866114377975464, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 0.3696, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3950617283950617, | |
| "eval_loss": 0.6138448715209961, | |
| "eval_runtime": 38.481, | |
| "eval_samples_per_second": 1.169, | |
| "eval_steps_per_second": 1.169, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.4345679012345679, | |
| "grad_norm": 0.3299381136894226, | |
| "learning_rate": 8.461538461538461e-05, | |
| "loss": 0.4084, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.4740740740740741, | |
| "grad_norm": 1.5945664644241333, | |
| "learning_rate": 9.230769230769232e-05, | |
| "loss": 1.8608, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.5135802469135803, | |
| "grad_norm": 0.43234264850616455, | |
| "learning_rate": 0.0001, | |
| "loss": 0.4467, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.5530864197530864, | |
| "grad_norm": 0.23291274905204773, | |
| "learning_rate": 9.998033131915266e-05, | |
| "loss": 0.3466, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.5925925925925926, | |
| "grad_norm": 0.23256279528141022, | |
| "learning_rate": 9.992134075089084e-05, | |
| "loss": 0.2995, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.6320987654320988, | |
| "grad_norm": 0.2454572319984436, | |
| "learning_rate": 9.982307470588098e-05, | |
| "loss": 0.2912, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.671604938271605, | |
| "grad_norm": 0.19026127457618713, | |
| "learning_rate": 9.968561049466214e-05, | |
| "loss": 0.2579, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.7111111111111111, | |
| "grad_norm": 0.16097521781921387, | |
| "learning_rate": 9.950905626682228e-05, | |
| "loss": 0.2407, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.7506172839506173, | |
| "grad_norm": 0.18558672070503235, | |
| "learning_rate": 9.92935509259118e-05, | |
| "loss": 0.232, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.7901234567901234, | |
| "grad_norm": 0.20855776965618134, | |
| "learning_rate": 9.903926402016153e-05, | |
| "loss": 0.2069, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.7901234567901234, | |
| "eval_loss": 0.389056533575058, | |
| "eval_runtime": 38.7201, | |
| "eval_samples_per_second": 1.162, | |
| "eval_steps_per_second": 1.162, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.8296296296296296, | |
| "grad_norm": 0.21402069926261902, | |
| "learning_rate": 9.874639560909117e-05, | |
| "loss": 0.2248, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.8691358024691358, | |
| "grad_norm": 0.21493636071681976, | |
| "learning_rate": 9.841517610611309e-05, | |
| "loss": 0.2176, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.908641975308642, | |
| "grad_norm": 0.18079149723052979, | |
| "learning_rate": 9.804586609725499e-05, | |
| "loss": 0.2262, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.9481481481481482, | |
| "grad_norm": 0.19002017378807068, | |
| "learning_rate": 9.763875613614482e-05, | |
| "loss": 0.1979, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.9876543209876543, | |
| "grad_norm": 0.20085716247558594, | |
| "learning_rate": 9.719416651541839e-05, | |
| "loss": 0.2038, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.0271604938271606, | |
| "grad_norm": 0.29270461201667786, | |
| "learning_rate": 9.671244701472999e-05, | |
| "loss": 0.3211, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.0666666666666667, | |
| "grad_norm": 0.17304855585098267, | |
| "learning_rate": 9.619397662556435e-05, | |
| "loss": 0.2024, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 1.106172839506173, | |
| "grad_norm": 0.1423790603876114, | |
| "learning_rate": 9.563916325306594e-05, | |
| "loss": 0.1808, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.145679012345679, | |
| "grad_norm": 0.1679137498140335, | |
| "learning_rate": 9.504844339512095e-05, | |
| "loss": 0.1683, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "grad_norm": 0.1534942388534546, | |
| "learning_rate": 9.442228179894362e-05, | |
| "loss": 0.1724, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "eval_loss": 0.32005801796913147, | |
| "eval_runtime": 38.5278, | |
| "eval_samples_per_second": 1.168, | |
| "eval_steps_per_second": 1.168, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.2246913580246914, | |
| "grad_norm": 0.13248418271541595, | |
| "learning_rate": 9.376117109543769e-05, | |
| "loss": 0.1367, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.2641975308641975, | |
| "grad_norm": 2.301762580871582, | |
| "learning_rate": 9.306563141162046e-05, | |
| "loss": 1.1727, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.3037037037037038, | |
| "grad_norm": 0.7346077561378479, | |
| "learning_rate": 9.233620996141421e-05, | |
| "loss": 0.2848, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.34320987654321, | |
| "grad_norm": 0.1404714435338974, | |
| "learning_rate": 9.157348061512727e-05, | |
| "loss": 0.1423, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.382716049382716, | |
| "grad_norm": 0.1579016000032425, | |
| "learning_rate": 9.077804344796302e-05, | |
| "loss": 0.1581, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.4222222222222223, | |
| "grad_norm": 0.2410455346107483, | |
| "learning_rate": 8.995052426791247e-05, | |
| "loss": 0.1606, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.4617283950617284, | |
| "grad_norm": 0.15172675251960754, | |
| "learning_rate": 8.90915741234015e-05, | |
| "loss": 0.135, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.5012345679012347, | |
| "grad_norm": 0.1471657156944275, | |
| "learning_rate": 8.820186879108038e-05, | |
| "loss": 0.1414, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.5407407407407407, | |
| "grad_norm": 0.3015000522136688, | |
| "learning_rate": 8.728210824415827e-05, | |
| "loss": 0.223, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.5802469135802468, | |
| "grad_norm": 0.18128684163093567, | |
| "learning_rate": 8.633301610170135e-05, | |
| "loss": 0.1333, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.5802469135802468, | |
| "eval_loss": 0.25207382440567017, | |
| "eval_runtime": 38.7379, | |
| "eval_samples_per_second": 1.162, | |
| "eval_steps_per_second": 1.162, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.6197530864197531, | |
| "grad_norm": 0.17792671918869019, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 0.1314, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.6592592592592592, | |
| "grad_norm": 0.16997207701206207, | |
| "learning_rate": 8.434984630174509e-05, | |
| "loss": 0.1447, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.6987654320987655, | |
| "grad_norm": 0.18394654989242554, | |
| "learning_rate": 8.33173288976002e-05, | |
| "loss": 0.1425, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.7382716049382716, | |
| "grad_norm": 0.19959236681461334, | |
| "learning_rate": 8.225859917710439e-05, | |
| "loss": 0.1337, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 2.166421890258789, | |
| "learning_rate": 8.117449009293668e-05, | |
| "loss": 0.9819, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.817283950617284, | |
| "grad_norm": 0.19032002985477448, | |
| "learning_rate": 8.006585456492029e-05, | |
| "loss": 0.1197, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.8567901234567903, | |
| "grad_norm": 0.18904529511928558, | |
| "learning_rate": 7.89335648089903e-05, | |
| "loss": 0.1556, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.8962962962962964, | |
| "grad_norm": 0.21338751912117004, | |
| "learning_rate": 7.777851165098012e-05, | |
| "loss": 0.1276, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.9358024691358025, | |
| "grad_norm": 0.21389852464199066, | |
| "learning_rate": 7.660160382576683e-05, | |
| "loss": 0.1227, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.9753086419753085, | |
| "grad_norm": 0.7639189958572388, | |
| "learning_rate": 7.540376726232648e-05, | |
| "loss": 0.2458, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.9753086419753085, | |
| "eval_loss": 0.2010781615972519, | |
| "eval_runtime": 38.5178, | |
| "eval_samples_per_second": 1.168, | |
| "eval_steps_per_second": 1.168, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.0148148148148146, | |
| "grad_norm": 0.3677748143672943, | |
| "learning_rate": 7.4185944355262e-05, | |
| "loss": 0.1995, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 2.054320987654321, | |
| "grad_norm": 0.1963815540075302, | |
| "learning_rate": 7.294909322337689e-05, | |
| "loss": 0.1152, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 2.093827160493827, | |
| "grad_norm": 0.1908060759305954, | |
| "learning_rate": 7.169418695587791e-05, | |
| "loss": 0.0994, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 2.1333333333333333, | |
| "grad_norm": 0.21485035121440887, | |
| "learning_rate": 7.042221284679982e-05, | |
| "loss": 0.1413, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 2.1728395061728394, | |
| "grad_norm": 0.24795810878276825, | |
| "learning_rate": 6.91341716182545e-05, | |
| "loss": 0.1224, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.212345679012346, | |
| "grad_norm": 0.23560446500778198, | |
| "learning_rate": 6.783107663311565e-05, | |
| "loss": 0.1169, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 2.251851851851852, | |
| "grad_norm": 0.44824451208114624, | |
| "learning_rate": 6.651395309775837e-05, | |
| "loss": 0.2425, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.291358024691358, | |
| "grad_norm": 0.22434134781360626, | |
| "learning_rate": 6.518383725548074e-05, | |
| "loss": 0.108, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 2.330864197530864, | |
| "grad_norm": 0.20300117135047913, | |
| "learning_rate": 6.384177557124247e-05, | |
| "loss": 0.1073, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 2.3703703703703702, | |
| "grad_norm": 0.23574106395244598, | |
| "learning_rate": 6.248882390836135e-05, | |
| "loss": 0.0919, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.3703703703703702, | |
| "eval_loss": 0.15837843716144562, | |
| "eval_runtime": 38.4853, | |
| "eval_samples_per_second": 1.169, | |
| "eval_steps_per_second": 1.169, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.4098765432098768, | |
| "grad_norm": 0.24438481032848358, | |
| "learning_rate": 6.112604669781572e-05, | |
| "loss": 0.1235, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.449382716049383, | |
| "grad_norm": 0.1859017014503479, | |
| "learning_rate": 5.9754516100806423e-05, | |
| "loss": 0.1066, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 2.488888888888889, | |
| "grad_norm": 0.22820916771888733, | |
| "learning_rate": 5.837531116523682e-05, | |
| "loss": 0.0952, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 2.528395061728395, | |
| "grad_norm": 0.1828678846359253, | |
| "learning_rate": 5.698951697677498e-05, | |
| "loss": 0.094, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 2.567901234567901, | |
| "grad_norm": 1.702113389968872, | |
| "learning_rate": 5.559822380516539e-05, | |
| "loss": 0.5798, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.6074074074074076, | |
| "grad_norm": 0.24967199563980103, | |
| "learning_rate": 5.420252624646238e-05, | |
| "loss": 0.0863, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.6469135802469137, | |
| "grad_norm": 0.23501570522785187, | |
| "learning_rate": 5.2803522361859594e-05, | |
| "loss": 0.0985, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.68641975308642, | |
| "grad_norm": 0.21782447397708893, | |
| "learning_rate": 5.140231281379345e-05, | |
| "loss": 0.0859, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.725925925925926, | |
| "grad_norm": 0.2892228662967682, | |
| "learning_rate": 5e-05, | |
| "loss": 0.0813, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.765432098765432, | |
| "grad_norm": 0.22678381204605103, | |
| "learning_rate": 4.859768718620656e-05, | |
| "loss": 0.0889, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.765432098765432, | |
| "eval_loss": 0.1346510797739029, | |
| "eval_runtime": 38.6245, | |
| "eval_samples_per_second": 1.165, | |
| "eval_steps_per_second": 1.165, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.8049382716049385, | |
| "grad_norm": 0.19133037328720093, | |
| "learning_rate": 4.7196477638140404e-05, | |
| "loss": 0.0884, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.8444444444444446, | |
| "grad_norm": 0.1989874392747879, | |
| "learning_rate": 4.579747375353763e-05, | |
| "loss": 0.0987, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.8839506172839506, | |
| "grad_norm": 0.1637035757303238, | |
| "learning_rate": 4.4401776194834613e-05, | |
| "loss": 0.0731, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.9234567901234567, | |
| "grad_norm": 0.8744575381278992, | |
| "learning_rate": 4.3010483023225045e-05, | |
| "loss": 0.4245, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.962962962962963, | |
| "grad_norm": 0.19474665820598602, | |
| "learning_rate": 4.162468883476319e-05, | |
| "loss": 0.0793, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 3.0024691358024693, | |
| "grad_norm": 0.30547550320625305, | |
| "learning_rate": 4.0245483899193595e-05, | |
| "loss": 0.1227, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 3.0419753086419754, | |
| "grad_norm": 0.2047836184501648, | |
| "learning_rate": 3.887395330218429e-05, | |
| "loss": 0.0824, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 3.0814814814814815, | |
| "grad_norm": 0.6940678954124451, | |
| "learning_rate": 3.7511176091638653e-05, | |
| "loss": 0.4433, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 3.1209876543209876, | |
| "grad_norm": 0.1984521746635437, | |
| "learning_rate": 3.6158224428757535e-05, | |
| "loss": 0.0761, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 3.1604938271604937, | |
| "grad_norm": 0.2340918779373169, | |
| "learning_rate": 3.4816162744519263e-05, | |
| "loss": 0.0991, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.1604938271604937, | |
| "eval_loss": 0.11672868579626083, | |
| "eval_runtime": 38.6459, | |
| "eval_samples_per_second": 1.164, | |
| "eval_steps_per_second": 1.164, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.22067080438137054, | |
| "learning_rate": 3.3486046902241664e-05, | |
| "loss": 0.0817, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 3.2395061728395063, | |
| "grad_norm": 0.27402088046073914, | |
| "learning_rate": 3.216892336688435e-05, | |
| "loss": 0.1092, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 3.2790123456790123, | |
| "grad_norm": 0.208318293094635, | |
| "learning_rate": 3.086582838174551e-05, | |
| "loss": 0.0794, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 3.3185185185185184, | |
| "grad_norm": 0.1757229119539261, | |
| "learning_rate": 2.9577787153200197e-05, | |
| "loss": 0.0796, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 3.3580246913580245, | |
| "grad_norm": 0.21269826591014862, | |
| "learning_rate": 2.8305813044122097e-05, | |
| "loss": 0.072, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 3.397530864197531, | |
| "grad_norm": 0.23646800220012665, | |
| "learning_rate": 2.705090677662311e-05, | |
| "loss": 0.0813, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 3.437037037037037, | |
| "grad_norm": 0.2855721414089203, | |
| "learning_rate": 2.581405564473801e-05, | |
| "loss": 0.0774, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 3.476543209876543, | |
| "grad_norm": 0.2356238216161728, | |
| "learning_rate": 2.459623273767354e-05, | |
| "loss": 0.1151, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 3.5160493827160493, | |
| "grad_norm": 0.2751014232635498, | |
| "learning_rate": 2.3398396174233178e-05, | |
| "loss": 0.0831, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "grad_norm": 0.22889889776706696, | |
| "learning_rate": 2.2221488349019903e-05, | |
| "loss": 0.1154, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "eval_loss": 0.10971427708864212, | |
| "eval_runtime": 38.5026, | |
| "eval_samples_per_second": 1.169, | |
| "eval_steps_per_second": 1.169, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.595061728395062, | |
| "grad_norm": 0.24450689554214478, | |
| "learning_rate": 2.1066435191009715e-05, | |
| "loss": 0.0779, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 3.634567901234568, | |
| "grad_norm": 0.17681388556957245, | |
| "learning_rate": 1.9934145435079702e-05, | |
| "loss": 0.0707, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 3.674074074074074, | |
| "grad_norm": 0.2236029952764511, | |
| "learning_rate": 1.8825509907063327e-05, | |
| "loss": 0.0745, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 3.71358024691358, | |
| "grad_norm": 0.5732559561729431, | |
| "learning_rate": 1.774140082289563e-05, | |
| "loss": 0.2596, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 3.753086419753086, | |
| "grad_norm": 0.2240099310874939, | |
| "learning_rate": 1.6682671102399805e-05, | |
| "loss": 0.074, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 3.7925925925925927, | |
| "grad_norm": 0.2508302927017212, | |
| "learning_rate": 1.5650153698254916e-05, | |
| "loss": 0.0712, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 3.832098765432099, | |
| "grad_norm": 0.31908485293388367, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 0.066, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 3.871604938271605, | |
| "grad_norm": 0.20694778859615326, | |
| "learning_rate": 1.3666983898298657e-05, | |
| "loss": 0.0744, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 3.911111111111111, | |
| "grad_norm": 0.27899348735809326, | |
| "learning_rate": 1.2717891755841722e-05, | |
| "loss": 0.0808, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 3.950617283950617, | |
| "grad_norm": 0.1817609965801239, | |
| "learning_rate": 1.1798131208919627e-05, | |
| "loss": 0.07, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.950617283950617, | |
| "eval_loss": 0.10511435568332672, | |
| "eval_runtime": 38.5153, | |
| "eval_samples_per_second": 1.168, | |
| "eval_steps_per_second": 1.168, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.9901234567901236, | |
| "grad_norm": 0.3054462671279907, | |
| "learning_rate": 1.090842587659851e-05, | |
| "loss": 0.0906, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 4.029629629629629, | |
| "grad_norm": 0.3566385507583618, | |
| "learning_rate": 1.004947573208756e-05, | |
| "loss": 0.1023, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 4.069135802469136, | |
| "grad_norm": 0.1800849437713623, | |
| "learning_rate": 9.221956552036992e-06, | |
| "loss": 0.0726, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 4.108641975308642, | |
| "grad_norm": 0.25262001156806946, | |
| "learning_rate": 8.426519384872733e-06, | |
| "loss": 0.0691, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 4.148148148148148, | |
| "grad_norm": 0.1978868544101715, | |
| "learning_rate": 7.663790038585793e-06, | |
| "loss": 0.0657, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 4.187654320987654, | |
| "grad_norm": 0.21851716935634613, | |
| "learning_rate": 6.934368588379553e-06, | |
| "loss": 0.0732, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 4.22716049382716, | |
| "grad_norm": 0.1888604462146759, | |
| "learning_rate": 6.238828904562316e-06, | |
| "loss": 0.0894, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 4.266666666666667, | |
| "grad_norm": 0.17408150434494019, | |
| "learning_rate": 5.577718201056392e-06, | |
| "loss": 0.0767, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 4.306172839506173, | |
| "grad_norm": 0.18007762730121613, | |
| "learning_rate": 4.951556604879048e-06, | |
| "loss": 0.0832, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 4.345679012345679, | |
| "grad_norm": 0.18987688422203064, | |
| "learning_rate": 4.360836746934055e-06, | |
| "loss": 0.0668, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.345679012345679, | |
| "eval_loss": 0.10350405424833298, | |
| "eval_runtime": 38.7722, | |
| "eval_samples_per_second": 1.161, | |
| "eval_steps_per_second": 1.161, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.385185185185185, | |
| "grad_norm": 0.21131804585456848, | |
| "learning_rate": 3.8060233744356633e-06, | |
| "loss": 0.0591, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 4.424691358024692, | |
| "grad_norm": 0.20253707468509674, | |
| "learning_rate": 3.2875529852700147e-06, | |
| "loss": 0.0899, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 4.4641975308641975, | |
| "grad_norm": 0.18665483593940735, | |
| "learning_rate": 2.8058334845816213e-06, | |
| "loss": 0.0679, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 4.503703703703704, | |
| "grad_norm": 0.22282886505126953, | |
| "learning_rate": 2.361243863855184e-06, | |
| "loss": 0.0852, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 4.54320987654321, | |
| "grad_norm": 0.19947148859500885, | |
| "learning_rate": 1.9541339027450256e-06, | |
| "loss": 0.0706, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 4.582716049382716, | |
| "grad_norm": 0.17244380712509155, | |
| "learning_rate": 1.584823893886933e-06, | |
| "loss": 0.0758, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 4.622222222222222, | |
| "grad_norm": 0.16220329701900482, | |
| "learning_rate": 1.2536043909088191e-06, | |
| "loss": 0.0694, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 4.661728395061728, | |
| "grad_norm": 0.15722690522670746, | |
| "learning_rate": 9.607359798384785e-07, | |
| "loss": 0.0638, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 4.701234567901235, | |
| "grad_norm": 0.15548333525657654, | |
| "learning_rate": 7.064490740882057e-07, | |
| "loss": 0.0558, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 4.7407407407407405, | |
| "grad_norm": 0.38291072845458984, | |
| "learning_rate": 4.909437331777179e-07, | |
| "loss": 0.2276, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.7407407407407405, | |
| "eval_loss": 0.10305963456630707, | |
| "eval_runtime": 37.3529, | |
| "eval_samples_per_second": 1.205, | |
| "eval_steps_per_second": 1.205, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.780246913580247, | |
| "grad_norm": 0.23985083401203156, | |
| "learning_rate": 3.143895053378698e-07, | |
| "loss": 0.1301, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 4.8197530864197535, | |
| "grad_norm": 0.18469485640525818, | |
| "learning_rate": 1.7692529411904578e-07, | |
| "loss": 0.0707, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 4.859259259259259, | |
| "grad_norm": 0.2033776193857193, | |
| "learning_rate": 7.865924910916977e-08, | |
| "loss": 0.1056, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 4.898765432098766, | |
| "grad_norm": 0.1682732254266739, | |
| "learning_rate": 1.9668680847356735e-08, | |
| "loss": 0.0694, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 4.938271604938271, | |
| "grad_norm": 0.41569438576698303, | |
| "learning_rate": 0.0, | |
| "loss": 0.2328, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 4.938271604938271, | |
| "step": 125, | |
| "total_flos": 1.8139029154214707e+17, | |
| "train_loss": 0.20902034133672714, | |
| "train_runtime": 6019.0916, | |
| "train_samples_per_second": 0.336, | |
| "train_steps_per_second": 0.021 | |
| } | |
| ], | |
| "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": 1.8139029154214707e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |