| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.4994054696789536, | |
| "eval_steps": 105, | |
| "global_step": 210, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0023781212841854932, | |
| "grad_norm": 32.382706174573094, | |
| "learning_rate": 5.000000000000001e-07, | |
| "loss": 2.9481, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0023781212841854932, | |
| "eval_loss": 3.3739373683929443, | |
| "eval_runtime": 152.3599, | |
| "eval_samples_per_second": 4.194, | |
| "eval_steps_per_second": 0.525, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0047562425683709865, | |
| "grad_norm": 36.523706488132554, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 2.8267, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.007134363852556481, | |
| "grad_norm": 30.823955989385283, | |
| "learning_rate": 1.5e-06, | |
| "loss": 2.9936, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.009512485136741973, | |
| "grad_norm": 27.113313789497223, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 2.8945, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.011890606420927468, | |
| "grad_norm": 13.505949871248614, | |
| "learning_rate": 2.5e-06, | |
| "loss": 2.7689, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.014268727705112961, | |
| "grad_norm": 15.0338641814596, | |
| "learning_rate": 3e-06, | |
| "loss": 2.6611, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.016646848989298454, | |
| "grad_norm": 14.319329036066911, | |
| "learning_rate": 3.5e-06, | |
| "loss": 2.8985, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.019024970273483946, | |
| "grad_norm": 14.23176344511718, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 2.834, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.02140309155766944, | |
| "grad_norm": 13.977227656063016, | |
| "learning_rate": 4.5e-06, | |
| "loss": 2.7566, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.023781212841854936, | |
| "grad_norm": 11.952136262918579, | |
| "learning_rate": 5e-06, | |
| "loss": 2.5198, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.026159334126040427, | |
| "grad_norm": 8.093250140015975, | |
| "learning_rate": 5.500000000000001e-06, | |
| "loss": 2.6486, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.028537455410225922, | |
| "grad_norm": 4.914080620863233, | |
| "learning_rate": 6e-06, | |
| "loss": 2.4014, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.030915576694411414, | |
| "grad_norm": 4.175963213321, | |
| "learning_rate": 6.5000000000000004e-06, | |
| "loss": 2.3925, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.03329369797859691, | |
| "grad_norm": 4.098569808536842, | |
| "learning_rate": 7e-06, | |
| "loss": 2.5433, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0356718192627824, | |
| "grad_norm": 5.46644902034287, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 2.5083, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.03804994054696789, | |
| "grad_norm": 5.787961659158823, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 2.6687, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.04042806183115339, | |
| "grad_norm": 3.034638828158752, | |
| "learning_rate": 8.5e-06, | |
| "loss": 2.4193, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.04280618311533888, | |
| "grad_norm": 3.6371355055713352, | |
| "learning_rate": 9e-06, | |
| "loss": 2.3729, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.04518430439952437, | |
| "grad_norm": 2.1215238627086714, | |
| "learning_rate": 9.5e-06, | |
| "loss": 2.4799, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.04756242568370987, | |
| "grad_norm": 3.987132196757572, | |
| "learning_rate": 1e-05, | |
| "loss": 2.3981, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.04994054696789536, | |
| "grad_norm": 1.8643899752237216, | |
| "learning_rate": 1.0500000000000001e-05, | |
| "loss": 2.6625, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.052318668252080855, | |
| "grad_norm": 1.9762840454371524, | |
| "learning_rate": 1.1000000000000001e-05, | |
| "loss": 2.4424, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.054696789536266346, | |
| "grad_norm": 1.6430481169991258, | |
| "learning_rate": 1.15e-05, | |
| "loss": 2.3021, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.057074910820451845, | |
| "grad_norm": 2.539281280765958, | |
| "learning_rate": 1.2e-05, | |
| "loss": 2.5471, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.059453032104637336, | |
| "grad_norm": 1.457528590578011, | |
| "learning_rate": 1.25e-05, | |
| "loss": 2.2526, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06183115338882283, | |
| "grad_norm": 2.1155531281068884, | |
| "learning_rate": 1.3000000000000001e-05, | |
| "loss": 2.8667, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.06420927467300833, | |
| "grad_norm": 1.73342938568033, | |
| "learning_rate": 1.3500000000000001e-05, | |
| "loss": 2.4915, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.06658739595719382, | |
| "grad_norm": 1.7317610327219517, | |
| "learning_rate": 1.4e-05, | |
| "loss": 2.498, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.06896551724137931, | |
| "grad_norm": 1.8363122567034704, | |
| "learning_rate": 1.45e-05, | |
| "loss": 2.4243, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.0713436385255648, | |
| "grad_norm": 1.6309580804999364, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 2.1101, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07372175980975029, | |
| "grad_norm": 1.334654998985815, | |
| "learning_rate": 1.55e-05, | |
| "loss": 2.1046, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.07609988109393578, | |
| "grad_norm": 1.2376006409087055, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 2.4625, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.07847800237812129, | |
| "grad_norm": 1.8515124675338408, | |
| "learning_rate": 1.65e-05, | |
| "loss": 2.3255, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.08085612366230678, | |
| "grad_norm": 2.4630792631286575, | |
| "learning_rate": 1.7e-05, | |
| "loss": 2.3863, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.08323424494649227, | |
| "grad_norm": 1.4437945302508726, | |
| "learning_rate": 1.7500000000000002e-05, | |
| "loss": 2.6266, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08561236623067776, | |
| "grad_norm": 1.6987676732458694, | |
| "learning_rate": 1.8e-05, | |
| "loss": 2.2871, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.08799048751486326, | |
| "grad_norm": 1.8155955386416776, | |
| "learning_rate": 1.8500000000000002e-05, | |
| "loss": 2.2534, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.09036860879904875, | |
| "grad_norm": 1.1727394458786922, | |
| "learning_rate": 1.9e-05, | |
| "loss": 2.3458, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.09274673008323424, | |
| "grad_norm": 1.4758138327709573, | |
| "learning_rate": 1.95e-05, | |
| "loss": 2.4411, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.09512485136741974, | |
| "grad_norm": 2.683368502920404, | |
| "learning_rate": 2e-05, | |
| "loss": 2.5046, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09750297265160524, | |
| "grad_norm": 1.6044285552596205, | |
| "learning_rate": 1.9999981652287733e-05, | |
| "loss": 2.413, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.09988109393579073, | |
| "grad_norm": 1.2802177698200177, | |
| "learning_rate": 1.999992660921826e-05, | |
| "loss": 2.2159, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.10225921521997622, | |
| "grad_norm": 1.8164033444979393, | |
| "learning_rate": 1.999983487099356e-05, | |
| "loss": 2.5287, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.10463733650416171, | |
| "grad_norm": 1.8420609096797744, | |
| "learning_rate": 1.999970643795027e-05, | |
| "loss": 2.3484, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.1070154577883472, | |
| "grad_norm": 1.4183887604989038, | |
| "learning_rate": 1.9999541310559686e-05, | |
| "loss": 2.4876, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10939357907253269, | |
| "grad_norm": 1.419110897512839, | |
| "learning_rate": 1.9999339489427746e-05, | |
| "loss": 2.3251, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.1117717003567182, | |
| "grad_norm": 1.6548987738436802, | |
| "learning_rate": 1.9999100975295046e-05, | |
| "loss": 2.358, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.11414982164090369, | |
| "grad_norm": 1.7822053646902, | |
| "learning_rate": 1.999882576903682e-05, | |
| "loss": 2.5853, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.11652794292508918, | |
| "grad_norm": 1.1885208188962406, | |
| "learning_rate": 1.9998513871662945e-05, | |
| "loss": 2.3794, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.11890606420927467, | |
| "grad_norm": 1.7227839639402496, | |
| "learning_rate": 1.9998165284317944e-05, | |
| "loss": 2.1933, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.12128418549346016, | |
| "grad_norm": 1.3654158039316031, | |
| "learning_rate": 1.999778000828098e-05, | |
| "loss": 2.0518, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.12366230677764566, | |
| "grad_norm": 1.3281085763249942, | |
| "learning_rate": 1.9997358044965833e-05, | |
| "loss": 2.4352, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.12604042806183116, | |
| "grad_norm": 1.1185472935459022, | |
| "learning_rate": 1.9996899395920915e-05, | |
| "loss": 2.4415, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.12841854934601665, | |
| "grad_norm": 1.343325378951958, | |
| "learning_rate": 1.999640406282926e-05, | |
| "loss": 2.2661, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.13079667063020214, | |
| "grad_norm": 1.371349840462257, | |
| "learning_rate": 1.9995872047508516e-05, | |
| "loss": 2.2813, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.13317479191438764, | |
| "grad_norm": 1.2560078627168356, | |
| "learning_rate": 1.9995303351910934e-05, | |
| "loss": 2.494, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.13555291319857313, | |
| "grad_norm": 1.3801141775649166, | |
| "learning_rate": 1.9994697978123363e-05, | |
| "loss": 2.1525, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.13793103448275862, | |
| "grad_norm": 2.8895107871445167, | |
| "learning_rate": 1.9994055928367256e-05, | |
| "loss": 2.6727, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.1403091557669441, | |
| "grad_norm": 1.231028014153545, | |
| "learning_rate": 1.999337720499863e-05, | |
| "loss": 2.3854, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.1426872770511296, | |
| "grad_norm": 1.5588938179669447, | |
| "learning_rate": 1.99926618105081e-05, | |
| "loss": 2.033, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1450653983353151, | |
| "grad_norm": 1.1690094119529322, | |
| "learning_rate": 1.9991909747520835e-05, | |
| "loss": 2.1, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.14744351961950058, | |
| "grad_norm": 1.5407871708756533, | |
| "learning_rate": 1.999112101879656e-05, | |
| "loss": 2.3949, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.14982164090368608, | |
| "grad_norm": 1.0249789290696483, | |
| "learning_rate": 1.9990295627229544e-05, | |
| "loss": 2.463, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.15219976218787157, | |
| "grad_norm": 1.5221618005694364, | |
| "learning_rate": 1.99894335758486e-05, | |
| "loss": 2.0575, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.1545778834720571, | |
| "grad_norm": 1.0758079202744963, | |
| "learning_rate": 1.9988534867817065e-05, | |
| "loss": 2.0227, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.15695600475624258, | |
| "grad_norm": 1.3430324490071632, | |
| "learning_rate": 1.9987599506432785e-05, | |
| "loss": 2.467, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.15933412604042807, | |
| "grad_norm": 1.0714364507553935, | |
| "learning_rate": 1.9986627495128105e-05, | |
| "loss": 2.3812, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.16171224732461356, | |
| "grad_norm": 1.296279438065536, | |
| "learning_rate": 1.9985618837469864e-05, | |
| "loss": 2.4045, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.16409036860879905, | |
| "grad_norm": 1.1423020273292732, | |
| "learning_rate": 1.998457353715938e-05, | |
| "loss": 2.386, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.16646848989298454, | |
| "grad_norm": 1.5165435155836928, | |
| "learning_rate": 1.998349159803241e-05, | |
| "loss": 2.3063, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.16884661117717004, | |
| "grad_norm": 2.758568140385349, | |
| "learning_rate": 1.9982373024059195e-05, | |
| "loss": 2.4007, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.17122473246135553, | |
| "grad_norm": 1.268753870102164, | |
| "learning_rate": 1.998121781934438e-05, | |
| "loss": 2.1938, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.17360285374554102, | |
| "grad_norm": 1.1116353498200209, | |
| "learning_rate": 1.9980025988127037e-05, | |
| "loss": 2.1202, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.1759809750297265, | |
| "grad_norm": 1.266229091733924, | |
| "learning_rate": 1.9978797534780646e-05, | |
| "loss": 2.4397, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.178359096313912, | |
| "grad_norm": 1.2442419764935988, | |
| "learning_rate": 1.9977532463813064e-05, | |
| "loss": 2.4345, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.1807372175980975, | |
| "grad_norm": 1.0991664177678122, | |
| "learning_rate": 1.9976230779866527e-05, | |
| "loss": 2.3604, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.18311533888228299, | |
| "grad_norm": 1.4648399555465317, | |
| "learning_rate": 1.9974892487717613e-05, | |
| "loss": 2.4796, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.18549346016646848, | |
| "grad_norm": 1.4075214014031838, | |
| "learning_rate": 1.997351759227725e-05, | |
| "loss": 2.4604, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.187871581450654, | |
| "grad_norm": 1.1658962187610111, | |
| "learning_rate": 1.9972106098590665e-05, | |
| "loss": 2.3639, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.1902497027348395, | |
| "grad_norm": 1.0920203834125308, | |
| "learning_rate": 1.9970658011837404e-05, | |
| "loss": 2.1962, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.19262782401902498, | |
| "grad_norm": 1.0547465850317588, | |
| "learning_rate": 1.9969173337331283e-05, | |
| "loss": 2.2381, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.19500594530321047, | |
| "grad_norm": 0.9700896578179515, | |
| "learning_rate": 1.996765208052037e-05, | |
| "loss": 1.9818, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.19738406658739596, | |
| "grad_norm": 1.3721631541122385, | |
| "learning_rate": 1.9966094246986983e-05, | |
| "loss": 2.1842, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.19976218787158145, | |
| "grad_norm": 1.1909292893353944, | |
| "learning_rate": 1.9964499842447665e-05, | |
| "loss": 2.5704, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.20214030915576695, | |
| "grad_norm": 1.0845849173658786, | |
| "learning_rate": 1.9962868872753144e-05, | |
| "loss": 2.1158, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.20451843043995244, | |
| "grad_norm": 1.1512213319968665, | |
| "learning_rate": 1.996120134388834e-05, | |
| "loss": 2.3564, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.20689655172413793, | |
| "grad_norm": 1.1653706380489515, | |
| "learning_rate": 1.995949726197231e-05, | |
| "loss": 2.4664, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.20927467300832342, | |
| "grad_norm": 1.2375083589347724, | |
| "learning_rate": 1.9957756633258264e-05, | |
| "loss": 2.2763, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.2116527942925089, | |
| "grad_norm": 1.180995352120129, | |
| "learning_rate": 1.9955979464133515e-05, | |
| "loss": 2.28, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.2140309155766944, | |
| "grad_norm": 1.0060015116786936, | |
| "learning_rate": 1.995416576111945e-05, | |
| "loss": 2.0939, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.2164090368608799, | |
| "grad_norm": 1.123230557994905, | |
| "learning_rate": 1.9952315530871537e-05, | |
| "loss": 2.2575, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.21878715814506539, | |
| "grad_norm": 1.033250223812201, | |
| "learning_rate": 1.9950428780179274e-05, | |
| "loss": 2.2192, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.2211652794292509, | |
| "grad_norm": 1.01496288741104, | |
| "learning_rate": 1.994850551596617e-05, | |
| "loss": 2.3692, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.2235434007134364, | |
| "grad_norm": 1.2509512883238079, | |
| "learning_rate": 1.9946545745289727e-05, | |
| "loss": 2.5349, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.2259215219976219, | |
| "grad_norm": 1.1631889518067213, | |
| "learning_rate": 1.9944549475341404e-05, | |
| "loss": 2.2335, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.22829964328180738, | |
| "grad_norm": 1.0176487393302203, | |
| "learning_rate": 1.99425167134466e-05, | |
| "loss": 2.325, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.23067776456599287, | |
| "grad_norm": 1.0766159235170416, | |
| "learning_rate": 1.9940447467064624e-05, | |
| "loss": 2.4656, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.23305588585017836, | |
| "grad_norm": 1.1250639831138038, | |
| "learning_rate": 1.9938341743788658e-05, | |
| "loss": 2.1741, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.23543400713436385, | |
| "grad_norm": 1.1024674310720775, | |
| "learning_rate": 1.9936199551345744e-05, | |
| "loss": 2.1336, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.23781212841854935, | |
| "grad_norm": 1.2527734635640946, | |
| "learning_rate": 1.9934020897596752e-05, | |
| "loss": 2.2741, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.24019024970273484, | |
| "grad_norm": 0.9699919156060421, | |
| "learning_rate": 1.9931805790536342e-05, | |
| "loss": 2.2369, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.24256837098692033, | |
| "grad_norm": 1.1834325873202396, | |
| "learning_rate": 1.9929554238292944e-05, | |
| "loss": 2.0419, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.24494649227110582, | |
| "grad_norm": 1.1078024399344104, | |
| "learning_rate": 1.992726624912872e-05, | |
| "loss": 2.4991, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.2473246135552913, | |
| "grad_norm": 1.1465140647519878, | |
| "learning_rate": 1.992494183143955e-05, | |
| "loss": 2.6218, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.2497027348394768, | |
| "grad_norm": 1.143333452765713, | |
| "learning_rate": 1.9922580993754985e-05, | |
| "loss": 2.3428, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.2497027348394768, | |
| "eval_loss": 2.499577760696411, | |
| "eval_runtime": 151.1827, | |
| "eval_samples_per_second": 4.227, | |
| "eval_steps_per_second": 0.529, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.2520808561236623, | |
| "grad_norm": 1.0672145334944887, | |
| "learning_rate": 1.9920183744738208e-05, | |
| "loss": 2.3485, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.2544589774078478, | |
| "grad_norm": 1.0044582600402383, | |
| "learning_rate": 1.9917750093186036e-05, | |
| "loss": 2.1678, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.2568370986920333, | |
| "grad_norm": 1.1029447003396373, | |
| "learning_rate": 1.9915280048028853e-05, | |
| "loss": 2.3967, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.25921521997621877, | |
| "grad_norm": 1.260079077740416, | |
| "learning_rate": 1.9912773618330595e-05, | |
| "loss": 2.4385, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.2615933412604043, | |
| "grad_norm": 1.0846663572847435, | |
| "learning_rate": 1.9910230813288713e-05, | |
| "loss": 2.1431, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.26397146254458975, | |
| "grad_norm": 0.991674367348856, | |
| "learning_rate": 1.9907651642234138e-05, | |
| "loss": 1.9523, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.26634958382877527, | |
| "grad_norm": 1.105394178981242, | |
| "learning_rate": 1.9905036114631247e-05, | |
| "loss": 2.3063, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.26872770511296074, | |
| "grad_norm": 0.9979450972746343, | |
| "learning_rate": 1.990238424007783e-05, | |
| "loss": 2.2218, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.27110582639714625, | |
| "grad_norm": 1.2796796522419402, | |
| "learning_rate": 1.989969602830505e-05, | |
| "loss": 2.2271, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.2734839476813318, | |
| "grad_norm": 1.1720433923322253, | |
| "learning_rate": 1.9896971489177417e-05, | |
| "loss": 2.2736, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.27586206896551724, | |
| "grad_norm": 1.0744949490065667, | |
| "learning_rate": 1.9894210632692745e-05, | |
| "loss": 2.1864, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.27824019024970276, | |
| "grad_norm": 1.9078078540875143, | |
| "learning_rate": 1.9891413468982112e-05, | |
| "loss": 2.2678, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.2806183115338882, | |
| "grad_norm": 1.0427118571864202, | |
| "learning_rate": 1.988858000830983e-05, | |
| "loss": 2.1757, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.28299643281807374, | |
| "grad_norm": 1.8597307152991636, | |
| "learning_rate": 1.9885710261073402e-05, | |
| "loss": 2.2913, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.2853745541022592, | |
| "grad_norm": 1.0401590654407316, | |
| "learning_rate": 1.9882804237803487e-05, | |
| "loss": 2.0629, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2877526753864447, | |
| "grad_norm": 1.5079024059776849, | |
| "learning_rate": 1.9879861949163863e-05, | |
| "loss": 2.1827, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.2901307966706302, | |
| "grad_norm": 1.1499157717530648, | |
| "learning_rate": 1.9876883405951378e-05, | |
| "loss": 2.2959, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.2925089179548157, | |
| "grad_norm": 1.030399166377001, | |
| "learning_rate": 1.987386861909593e-05, | |
| "loss": 2.279, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.29488703923900117, | |
| "grad_norm": 1.062349469745721, | |
| "learning_rate": 1.98708175996604e-05, | |
| "loss": 2.2865, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.2972651605231867, | |
| "grad_norm": 1.0853089306877393, | |
| "learning_rate": 1.986773035884064e-05, | |
| "loss": 2.332, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.29964328180737215, | |
| "grad_norm": 1.0066223498076698, | |
| "learning_rate": 1.9864606907965407e-05, | |
| "loss": 2.3971, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.30202140309155767, | |
| "grad_norm": 1.101756427229776, | |
| "learning_rate": 1.986144725849634e-05, | |
| "loss": 2.2862, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.30439952437574314, | |
| "grad_norm": 1.1264625682995106, | |
| "learning_rate": 1.9858251422027903e-05, | |
| "loss": 2.0974, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.30677764565992865, | |
| "grad_norm": 1.0012622487091893, | |
| "learning_rate": 1.9855019410287355e-05, | |
| "loss": 2.2648, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.3091557669441142, | |
| "grad_norm": 1.0030908712085922, | |
| "learning_rate": 1.98517512351347e-05, | |
| "loss": 2.2626, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.31153388822829964, | |
| "grad_norm": 1.1376149636856583, | |
| "learning_rate": 1.9848446908562647e-05, | |
| "loss": 2.2315, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.31391200951248516, | |
| "grad_norm": 0.9616294646184662, | |
| "learning_rate": 1.9845106442696563e-05, | |
| "loss": 2.4033, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.3162901307966706, | |
| "grad_norm": 1.3402719458865533, | |
| "learning_rate": 1.9841729849794427e-05, | |
| "loss": 2.4429, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.31866825208085614, | |
| "grad_norm": 1.3575140959134089, | |
| "learning_rate": 1.983831714224679e-05, | |
| "loss": 2.291, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.3210463733650416, | |
| "grad_norm": 0.9521047622766085, | |
| "learning_rate": 1.9834868332576727e-05, | |
| "loss": 2.2754, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.3234244946492271, | |
| "grad_norm": 1.0875381930222732, | |
| "learning_rate": 1.9831383433439798e-05, | |
| "loss": 2.1469, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.3258026159334126, | |
| "grad_norm": 1.0257090212605473, | |
| "learning_rate": 1.982786245762398e-05, | |
| "loss": 2.0848, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.3281807372175981, | |
| "grad_norm": 1.0372156134974286, | |
| "learning_rate": 1.9824305418049645e-05, | |
| "loss": 2.4043, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.33055885850178357, | |
| "grad_norm": 0.9832029689246473, | |
| "learning_rate": 1.9820712327769503e-05, | |
| "loss": 2.177, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.3329369797859691, | |
| "grad_norm": 1.0472107045649877, | |
| "learning_rate": 1.9817083199968552e-05, | |
| "loss": 2.3309, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.33531510107015455, | |
| "grad_norm": 1.0277879008926316, | |
| "learning_rate": 1.9813418047964025e-05, | |
| "loss": 2.1389, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.3376932223543401, | |
| "grad_norm": 1.15382327979194, | |
| "learning_rate": 1.9809716885205363e-05, | |
| "loss": 2.3254, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.3400713436385256, | |
| "grad_norm": 0.9933827586398313, | |
| "learning_rate": 1.980597972527413e-05, | |
| "loss": 2.2454, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.34244946492271106, | |
| "grad_norm": 0.9637942573486198, | |
| "learning_rate": 1.9802206581883992e-05, | |
| "loss": 2.1945, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.3448275862068966, | |
| "grad_norm": 0.918837452187297, | |
| "learning_rate": 1.979839746888067e-05, | |
| "loss": 2.1599, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.34720570749108204, | |
| "grad_norm": 0.962272148586432, | |
| "learning_rate": 1.979455240024186e-05, | |
| "loss": 2.1504, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.34958382877526756, | |
| "grad_norm": 1.0902104459187203, | |
| "learning_rate": 1.97906713900772e-05, | |
| "loss": 2.1671, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.351961950059453, | |
| "grad_norm": 0.9528336584077055, | |
| "learning_rate": 1.9786754452628226e-05, | |
| "loss": 2.113, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.35434007134363854, | |
| "grad_norm": 0.9216300894356183, | |
| "learning_rate": 1.9782801602268306e-05, | |
| "loss": 1.9271, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.356718192627824, | |
| "grad_norm": 0.9148117236258577, | |
| "learning_rate": 1.9778812853502592e-05, | |
| "loss": 2.0221, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3590963139120095, | |
| "grad_norm": 0.9737793407143137, | |
| "learning_rate": 1.9774788220967968e-05, | |
| "loss": 2.2007, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.361474435196195, | |
| "grad_norm": 0.8979472123958888, | |
| "learning_rate": 1.9770727719432994e-05, | |
| "loss": 2.1295, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.3638525564803805, | |
| "grad_norm": 1.0391719811646754, | |
| "learning_rate": 1.9766631363797852e-05, | |
| "loss": 2.2383, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.36623067776456597, | |
| "grad_norm": 1.0898893562871452, | |
| "learning_rate": 1.9762499169094288e-05, | |
| "loss": 2.085, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.3686087990487515, | |
| "grad_norm": 2.9046586074281686, | |
| "learning_rate": 1.9758331150485576e-05, | |
| "loss": 2.2903, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.37098692033293695, | |
| "grad_norm": 1.0438135264691892, | |
| "learning_rate": 1.9754127323266426e-05, | |
| "loss": 2.349, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.3733650416171225, | |
| "grad_norm": 1.2082790526216014, | |
| "learning_rate": 1.9749887702862972e-05, | |
| "loss": 2.2182, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.375743162901308, | |
| "grad_norm": 1.2415740026323197, | |
| "learning_rate": 1.9745612304832672e-05, | |
| "loss": 2.4834, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.37812128418549346, | |
| "grad_norm": 0.9901876124346225, | |
| "learning_rate": 1.9741301144864284e-05, | |
| "loss": 2.2873, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.380499405469679, | |
| "grad_norm": 1.1185971951047096, | |
| "learning_rate": 1.9736954238777793e-05, | |
| "loss": 2.2114, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.38287752675386444, | |
| "grad_norm": 1.0186645068648283, | |
| "learning_rate": 1.9732571602524353e-05, | |
| "loss": 2.3323, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.38525564803804996, | |
| "grad_norm": 0.9856339888297305, | |
| "learning_rate": 1.972815325218624e-05, | |
| "loss": 2.2638, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.3876337693222354, | |
| "grad_norm": 1.287711819624049, | |
| "learning_rate": 1.9723699203976768e-05, | |
| "loss": 2.3897, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.39001189060642094, | |
| "grad_norm": 0.9474533284935532, | |
| "learning_rate": 1.9719209474240263e-05, | |
| "loss": 1.8287, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.3923900118906064, | |
| "grad_norm": 1.0505224096035144, | |
| "learning_rate": 1.971468407945198e-05, | |
| "loss": 2.3906, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3947681331747919, | |
| "grad_norm": 0.9322039774829307, | |
| "learning_rate": 1.9710123036218044e-05, | |
| "loss": 2.0246, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.3971462544589774, | |
| "grad_norm": 1.1428006052468438, | |
| "learning_rate": 1.97055263612754e-05, | |
| "loss": 2.0085, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.3995243757431629, | |
| "grad_norm": 0.9233456322532203, | |
| "learning_rate": 1.9700894071491736e-05, | |
| "loss": 2.0657, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.40190249702734837, | |
| "grad_norm": 1.1387607148614496, | |
| "learning_rate": 1.9696226183865436e-05, | |
| "loss": 2.2507, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.4042806183115339, | |
| "grad_norm": 1.0240739510681864, | |
| "learning_rate": 1.969152271552552e-05, | |
| "loss": 2.1685, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.40665873959571935, | |
| "grad_norm": 1.2665975670284688, | |
| "learning_rate": 1.9686783683731557e-05, | |
| "loss": 2.3869, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.4090368608799049, | |
| "grad_norm": 1.0148421037850517, | |
| "learning_rate": 1.9682009105873633e-05, | |
| "loss": 2.1379, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.4114149821640904, | |
| "grad_norm": 1.0117482642225601, | |
| "learning_rate": 1.9677198999472257e-05, | |
| "loss": 2.1104, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.41379310344827586, | |
| "grad_norm": 0.9720066057353862, | |
| "learning_rate": 1.967235338217832e-05, | |
| "loss": 2.2884, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.4161712247324614, | |
| "grad_norm": 1.2836956527083296, | |
| "learning_rate": 1.9667472271773026e-05, | |
| "loss": 2.281, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.41854934601664684, | |
| "grad_norm": 0.9738075739171279, | |
| "learning_rate": 1.9662555686167808e-05, | |
| "loss": 2.2039, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.42092746730083236, | |
| "grad_norm": 1.0382703606377657, | |
| "learning_rate": 1.965760364340429e-05, | |
| "loss": 2.1142, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.4233055885850178, | |
| "grad_norm": 0.9318167386351257, | |
| "learning_rate": 1.9652616161654204e-05, | |
| "loss": 2.2409, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.42568370986920334, | |
| "grad_norm": 1.3023888510009893, | |
| "learning_rate": 1.9647593259219328e-05, | |
| "loss": 2.1972, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.4280618311533888, | |
| "grad_norm": 2.3590946551757583, | |
| "learning_rate": 1.964253495453141e-05, | |
| "loss": 2.1922, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.4304399524375743, | |
| "grad_norm": 1.277353180787431, | |
| "learning_rate": 1.963744126615212e-05, | |
| "loss": 2.384, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.4328180737217598, | |
| "grad_norm": 0.9912373962254282, | |
| "learning_rate": 1.9632312212772956e-05, | |
| "loss": 2.3005, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.4351961950059453, | |
| "grad_norm": 1.3088391002508755, | |
| "learning_rate": 1.9627147813215207e-05, | |
| "loss": 2.2924, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.43757431629013077, | |
| "grad_norm": 1.0523417119399674, | |
| "learning_rate": 1.9621948086429847e-05, | |
| "loss": 2.2255, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.4399524375743163, | |
| "grad_norm": 1.386504728655603, | |
| "learning_rate": 1.9616713051497496e-05, | |
| "loss": 2.1963, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.4423305588585018, | |
| "grad_norm": 0.9623064139518317, | |
| "learning_rate": 1.9611442727628344e-05, | |
| "loss": 2.3175, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.4447086801426873, | |
| "grad_norm": 1.7130236244329065, | |
| "learning_rate": 1.960613713416206e-05, | |
| "loss": 2.2246, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.4470868014268728, | |
| "grad_norm": 1.2183030146426241, | |
| "learning_rate": 1.9600796290567747e-05, | |
| "loss": 2.2345, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.44946492271105826, | |
| "grad_norm": 2.1564007093614497, | |
| "learning_rate": 1.9595420216443864e-05, | |
| "loss": 2.269, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.4518430439952438, | |
| "grad_norm": 1.1685306329495788, | |
| "learning_rate": 1.9590008931518133e-05, | |
| "loss": 2.4817, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.45422116527942924, | |
| "grad_norm": 1.9314760188762214, | |
| "learning_rate": 1.9584562455647494e-05, | |
| "loss": 2.2482, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.45659928656361476, | |
| "grad_norm": 1.2361103482091276, | |
| "learning_rate": 1.9579080808818035e-05, | |
| "loss": 2.2229, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.4589774078478002, | |
| "grad_norm": 1.7230964883799738, | |
| "learning_rate": 1.9573564011144873e-05, | |
| "loss": 2.142, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.46135552913198574, | |
| "grad_norm": 1.2344882256405865, | |
| "learning_rate": 1.9568012082872148e-05, | |
| "loss": 2.0984, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.4637336504161712, | |
| "grad_norm": 1.563594408096457, | |
| "learning_rate": 1.9562425044372884e-05, | |
| "loss": 1.9245, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.4661117717003567, | |
| "grad_norm": 1.942680386259649, | |
| "learning_rate": 1.9556802916148963e-05, | |
| "loss": 2.2633, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.4684898929845422, | |
| "grad_norm": 1.3339191105911707, | |
| "learning_rate": 1.955114571883102e-05, | |
| "loss": 2.1356, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.4708680142687277, | |
| "grad_norm": 2.204421390949867, | |
| "learning_rate": 1.9545453473178384e-05, | |
| "loss": 2.2506, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.47324613555291317, | |
| "grad_norm": 1.106826809772669, | |
| "learning_rate": 1.9539726200078987e-05, | |
| "loss": 2.0526, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.4756242568370987, | |
| "grad_norm": 2.7022504927488136, | |
| "learning_rate": 1.9533963920549307e-05, | |
| "loss": 2.3654, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4780023781212842, | |
| "grad_norm": 2.20401718688023, | |
| "learning_rate": 1.9528166655734267e-05, | |
| "loss": 2.3523, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.4803804994054697, | |
| "grad_norm": 1.7098249314197853, | |
| "learning_rate": 1.9522334426907185e-05, | |
| "loss": 2.0789, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.4827586206896552, | |
| "grad_norm": 2.893610442346491, | |
| "learning_rate": 1.951646725546966e-05, | |
| "loss": 2.2456, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.48513674197384066, | |
| "grad_norm": 2.2821472751414937, | |
| "learning_rate": 1.9510565162951538e-05, | |
| "loss": 2.3351, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.4875148632580262, | |
| "grad_norm": 1.300863641854685, | |
| "learning_rate": 1.950462817101079e-05, | |
| "loss": 2.3617, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.48989298454221164, | |
| "grad_norm": 1.3899023575579752, | |
| "learning_rate": 1.9498656301433466e-05, | |
| "loss": 2.0628, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.49227110582639716, | |
| "grad_norm": 1.018395556873715, | |
| "learning_rate": 1.9492649576133594e-05, | |
| "loss": 2.2362, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.4946492271105826, | |
| "grad_norm": 1.4600451736607643, | |
| "learning_rate": 1.94866080171531e-05, | |
| "loss": 2.2194, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.49702734839476814, | |
| "grad_norm": 4.244045726741938, | |
| "learning_rate": 1.9480531646661753e-05, | |
| "loss": 2.434, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.4994054696789536, | |
| "grad_norm": 1.2125371460079588, | |
| "learning_rate": 1.9474420486957045e-05, | |
| "loss": 2.228, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.4994054696789536, | |
| "eval_loss": 2.4567697048187256, | |
| "eval_runtime": 151.5129, | |
| "eval_samples_per_second": 4.217, | |
| "eval_steps_per_second": 0.528, | |
| "step": 210 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1680, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 210, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.3027053762772992e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |