| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 126, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.015873015873015872, | |
| "grad_norm": 17.046055731691766, | |
| "learning_rate": 9.998445910004082e-06, | |
| "loss": 1.0473, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.031746031746031744, | |
| "grad_norm": 6.368244201268006, | |
| "learning_rate": 9.993784606094612e-06, | |
| "loss": 0.6073, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.047619047619047616, | |
| "grad_norm": 3.937504515409542, | |
| "learning_rate": 9.986018985905901e-06, | |
| "loss": 0.4184, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.06349206349206349, | |
| "grad_norm": 3.1167770147122558, | |
| "learning_rate": 9.975153876827008e-06, | |
| "loss": 0.3656, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.07936507936507936, | |
| "grad_norm": 3.5451663774846858, | |
| "learning_rate": 9.961196033000862e-06, | |
| "loss": 0.3752, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.09523809523809523, | |
| "grad_norm": 1.8934015097083565, | |
| "learning_rate": 9.944154131125643e-06, | |
| "loss": 0.2923, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.1111111111111111, | |
| "grad_norm": 1.9773484964869892, | |
| "learning_rate": 9.924038765061042e-06, | |
| "loss": 0.2976, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.12698412698412698, | |
| "grad_norm": 1.8499922558582569, | |
| "learning_rate": 9.900862439242719e-06, | |
| "loss": 0.2858, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 2.1048532676599176, | |
| "learning_rate": 9.874639560909118e-06, | |
| "loss": 0.275, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.15873015873015872, | |
| "grad_norm": 1.5720262400150389, | |
| "learning_rate": 9.84538643114539e-06, | |
| "loss": 0.2756, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1746031746031746, | |
| "grad_norm": 1.684255407558622, | |
| "learning_rate": 9.81312123475006e-06, | |
| "loss": 0.2647, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.19047619047619047, | |
| "grad_norm": 1.635574387616813, | |
| "learning_rate": 9.777864028930705e-06, | |
| "loss": 0.2546, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.20634920634920634, | |
| "grad_norm": 1.422170916071452, | |
| "learning_rate": 9.73963673083566e-06, | |
| "loss": 0.2203, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 1.1575210545990935, | |
| "learning_rate": 9.698463103929542e-06, | |
| "loss": 0.197, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.23809523809523808, | |
| "grad_norm": 1.3304992288104052, | |
| "learning_rate": 9.654368743221022e-06, | |
| "loss": 0.206, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.25396825396825395, | |
| "grad_norm": 1.2029627869249027, | |
| "learning_rate": 9.60738105935204e-06, | |
| "loss": 0.2064, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2698412698412698, | |
| "grad_norm": 1.2168950922943917, | |
| "learning_rate": 9.557529261558367e-06, | |
| "loss": 0.2123, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 1.2061804273710595, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.2234, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.30158730158730157, | |
| "grad_norm": 1.3455088304193212, | |
| "learning_rate": 9.449359044057344e-06, | |
| "loss": 0.2256, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.31746031746031744, | |
| "grad_norm": 1.2713281285022595, | |
| "learning_rate": 9.391107866851143e-06, | |
| "loss": 0.2158, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 1.0727037071509673, | |
| "learning_rate": 9.330127018922195e-06, | |
| "loss": 0.1969, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.3492063492063492, | |
| "grad_norm": 1.3540864350754735, | |
| "learning_rate": 9.266454408160779e-06, | |
| "loss": 0.2598, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.36507936507936506, | |
| "grad_norm": 1.2847824839788686, | |
| "learning_rate": 9.200129615753858e-06, | |
| "loss": 0.2389, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.38095238095238093, | |
| "grad_norm": 1.122678154329176, | |
| "learning_rate": 9.131193871579975e-06, | |
| "loss": 0.2279, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.3968253968253968, | |
| "grad_norm": 1.1519429578533356, | |
| "learning_rate": 9.059690028579285e-06, | |
| "loss": 0.2076, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.4126984126984127, | |
| "grad_norm": 1.3613352350414323, | |
| "learning_rate": 8.985662536114614e-06, | |
| "loss": 0.22, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "grad_norm": 1.2339714490779135, | |
| "learning_rate": 8.90915741234015e-06, | |
| "loss": 0.2522, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 1.1236413741627491, | |
| "learning_rate": 8.83022221559489e-06, | |
| "loss": 0.1866, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.4603174603174603, | |
| "grad_norm": 1.3086093841693327, | |
| "learning_rate": 8.748906014838672e-06, | |
| "loss": 0.2333, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.47619047619047616, | |
| "grad_norm": 1.0158374499871152, | |
| "learning_rate": 8.665259359149132e-06, | |
| "loss": 0.2031, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.49206349206349204, | |
| "grad_norm": 0.9845474076221036, | |
| "learning_rate": 8.579334246298593e-06, | |
| "loss": 0.1895, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.5079365079365079, | |
| "grad_norm": 0.9272494831082883, | |
| "learning_rate": 8.491184090430365e-06, | |
| "loss": 0.1678, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.5238095238095238, | |
| "grad_norm": 1.1065938204746082, | |
| "learning_rate": 8.400863688854598e-06, | |
| "loss": 0.2131, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.5396825396825397, | |
| "grad_norm": 0.9645961460688046, | |
| "learning_rate": 8.308429187984298e-06, | |
| "loss": 0.1938, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.5555555555555556, | |
| "grad_norm": 1.0652512127256022, | |
| "learning_rate": 8.213938048432697e-06, | |
| "loss": 0.228, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 1.2041599554639741, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.227, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.5873015873015873, | |
| "grad_norm": 1.0747489173754983, | |
| "learning_rate": 8.019022051627387e-06, | |
| "loss": 0.2072, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.6031746031746031, | |
| "grad_norm": 1.2773460577215314, | |
| "learning_rate": 7.918718361173951e-06, | |
| "loss": 0.2173, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.6190476190476191, | |
| "grad_norm": 1.135886238064031, | |
| "learning_rate": 7.81660029031811e-06, | |
| "loss": 0.2142, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.6349206349206349, | |
| "grad_norm": 0.9260328676177088, | |
| "learning_rate": 7.712731319328798e-06, | |
| "loss": 0.1707, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.6507936507936508, | |
| "grad_norm": 1.3665126828704022, | |
| "learning_rate": 7.607176016897491e-06, | |
| "loss": 0.2181, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 1.3127794515877824, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.2439, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.6825396825396826, | |
| "grad_norm": 1.1360818137590631, | |
| "learning_rate": 7.391269893106592e-06, | |
| "loss": 0.2105, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.6984126984126984, | |
| "grad_norm": 1.1335666149334869, | |
| "learning_rate": 7.281053286765816e-06, | |
| "loss": 0.1929, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 1.1267539001855318, | |
| "learning_rate": 7.169418695587791e-06, | |
| "loss": 0.1984, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.7301587301587301, | |
| "grad_norm": 1.1936854411979159, | |
| "learning_rate": 7.056435515653059e-06, | |
| "loss": 0.2028, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.746031746031746, | |
| "grad_norm": 0.9015309462604437, | |
| "learning_rate": 6.942173981373474e-06, | |
| "loss": 0.1875, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.7619047619047619, | |
| "grad_norm": 0.8200713976990537, | |
| "learning_rate": 6.8267051218319766e-06, | |
| "loss": 0.163, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.7777777777777778, | |
| "grad_norm": 1.051998804086299, | |
| "learning_rate": 6.710100716628345e-06, | |
| "loss": 0.1953, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.7936507936507936, | |
| "grad_norm": 1.1445651214323342, | |
| "learning_rate": 6.592433251258423e-06, | |
| "loss": 0.2231, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.8095238095238095, | |
| "grad_norm": 1.1078932700031434, | |
| "learning_rate": 6.473775872054522e-06, | |
| "loss": 0.1944, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.8253968253968254, | |
| "grad_norm": 0.9313430929773627, | |
| "learning_rate": 6.354202340715027e-06, | |
| "loss": 0.196, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.8412698412698413, | |
| "grad_norm": 0.9270393870240721, | |
| "learning_rate": 6.233786988451468e-06, | |
| "loss": 0.1682, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 0.9164332807254034, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.1846, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.873015873015873, | |
| "grad_norm": 0.8740293066946664, | |
| "learning_rate": 5.990730715996989e-06, | |
| "loss": 0.1657, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 1.2633376067074007, | |
| "learning_rate": 5.8682408883346535e-06, | |
| "loss": 0.2077, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.9047619047619048, | |
| "grad_norm": 0.8792598852017888, | |
| "learning_rate": 5.745211330880872e-06, | |
| "loss": 0.1768, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.9206349206349206, | |
| "grad_norm": 1.0220047570734931, | |
| "learning_rate": 5.621718523237427e-06, | |
| "loss": 0.1881, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.9365079365079365, | |
| "grad_norm": 0.8703339387258027, | |
| "learning_rate": 5.497839232979084e-06, | |
| "loss": 0.1611, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.9523809523809523, | |
| "grad_norm": 1.0109865838643635, | |
| "learning_rate": 5.373650467932122e-06, | |
| "loss": 0.2008, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.9682539682539683, | |
| "grad_norm": 0.7417523893697854, | |
| "learning_rate": 5.249229428303486e-06, | |
| "loss": 0.1602, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.9841269841269841, | |
| "grad_norm": 0.9461546053888866, | |
| "learning_rate": 5.1246534586903655e-06, | |
| "loss": 0.1993, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.9660336029382152, | |
| "learning_rate": 5e-06, | |
| "loss": 0.191, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.0158730158730158, | |
| "grad_norm": 0.8446705493437469, | |
| "learning_rate": 4.875346541309637e-06, | |
| "loss": 0.1364, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.0317460317460316, | |
| "grad_norm": 0.9820307539758936, | |
| "learning_rate": 4.750770571696514e-06, | |
| "loss": 0.1587, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.0476190476190477, | |
| "grad_norm": 0.9058719502990771, | |
| "learning_rate": 4.626349532067879e-06, | |
| "loss": 0.1483, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.0634920634920635, | |
| "grad_norm": 0.9041764403564576, | |
| "learning_rate": 4.502160767020918e-06, | |
| "loss": 0.1414, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.0793650793650793, | |
| "grad_norm": 0.8001618713509455, | |
| "learning_rate": 4.3782814767625755e-06, | |
| "loss": 0.1338, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.0952380952380953, | |
| "grad_norm": 0.8319305596618536, | |
| "learning_rate": 4.254788669119127e-06, | |
| "loss": 0.1454, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.1111111111111112, | |
| "grad_norm": 0.8371055902592329, | |
| "learning_rate": 4.131759111665349e-06, | |
| "loss": 0.1402, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.126984126984127, | |
| "grad_norm": 0.7954052878494495, | |
| "learning_rate": 4.009269284003014e-06, | |
| "loss": 0.1228, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 0.7671843494889209, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.1221, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.1587301587301586, | |
| "grad_norm": 0.8711455715304539, | |
| "learning_rate": 3.7662130115485317e-06, | |
| "loss": 0.1501, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.1746031746031746, | |
| "grad_norm": 0.7847941425308853, | |
| "learning_rate": 3.6457976592849753e-06, | |
| "loss": 0.1122, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.1904761904761905, | |
| "grad_norm": 0.722628056722779, | |
| "learning_rate": 3.526224127945479e-06, | |
| "loss": 0.1111, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.2063492063492063, | |
| "grad_norm": 0.8245734562679857, | |
| "learning_rate": 3.4075667487415785e-06, | |
| "loss": 0.1212, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.2222222222222223, | |
| "grad_norm": 0.8905262611051837, | |
| "learning_rate": 3.289899283371657e-06, | |
| "loss": 0.1318, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.2380952380952381, | |
| "grad_norm": 0.8016171759163617, | |
| "learning_rate": 3.173294878168025e-06, | |
| "loss": 0.1194, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.253968253968254, | |
| "grad_norm": 0.8978165203445735, | |
| "learning_rate": 3.057826018626527e-06, | |
| "loss": 0.1167, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.2698412698412698, | |
| "grad_norm": 0.8629753908810158, | |
| "learning_rate": 2.9435644843469434e-06, | |
| "loss": 0.1309, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.2857142857142856, | |
| "grad_norm": 0.7772338824641468, | |
| "learning_rate": 2.83058130441221e-06, | |
| "loss": 0.1158, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.3015873015873016, | |
| "grad_norm": 0.8162122268225106, | |
| "learning_rate": 2.718946713234185e-06, | |
| "loss": 0.1205, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.3174603174603174, | |
| "grad_norm": 0.8161848415092973, | |
| "learning_rate": 2.608730106893411e-06, | |
| "loss": 0.1284, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.7777687958543732, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.1161, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.3492063492063493, | |
| "grad_norm": 0.9073998204298158, | |
| "learning_rate": 2.39282398310251e-06, | |
| "loss": 0.1192, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.3650793650793651, | |
| "grad_norm": 0.775388656338499, | |
| "learning_rate": 2.2872686806712037e-06, | |
| "loss": 0.1163, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.380952380952381, | |
| "grad_norm": 0.8224765782876229, | |
| "learning_rate": 2.1833997096818897e-06, | |
| "loss": 0.1172, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.3968253968253967, | |
| "grad_norm": 0.9413042727529922, | |
| "learning_rate": 2.081281638826052e-06, | |
| "loss": 0.1218, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.4126984126984126, | |
| "grad_norm": 0.8442720538674777, | |
| "learning_rate": 1.980977948372612e-06, | |
| "loss": 0.1231, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 0.7194237358207392, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.1045, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.4444444444444444, | |
| "grad_norm": 0.8127413072028286, | |
| "learning_rate": 1.7860619515673034e-06, | |
| "loss": 0.1233, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.4603174603174602, | |
| "grad_norm": 1.2439563617941711, | |
| "learning_rate": 1.6915708120157042e-06, | |
| "loss": 0.1275, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.4761904761904763, | |
| "grad_norm": 0.7770780245344372, | |
| "learning_rate": 1.5991363111454023e-06, | |
| "loss": 0.1181, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.492063492063492, | |
| "grad_norm": 0.7672775525058901, | |
| "learning_rate": 1.5088159095696365e-06, | |
| "loss": 0.1059, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.507936507936508, | |
| "grad_norm": 0.766143455209844, | |
| "learning_rate": 1.4206657537014078e-06, | |
| "loss": 0.1178, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.5238095238095237, | |
| "grad_norm": 0.724421182393947, | |
| "learning_rate": 1.3347406408508695e-06, | |
| "loss": 0.1149, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.5396825396825395, | |
| "grad_norm": 0.7658778028894299, | |
| "learning_rate": 1.2510939851613285e-06, | |
| "loss": 0.1225, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.5555555555555556, | |
| "grad_norm": 1.032494528371397, | |
| "learning_rate": 1.1697777844051105e-06, | |
| "loss": 0.143, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.5714285714285714, | |
| "grad_norm": 0.8317671879333141, | |
| "learning_rate": 1.0908425876598512e-06, | |
| "loss": 0.1226, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.5873015873015874, | |
| "grad_norm": 0.8645979946263267, | |
| "learning_rate": 1.0143374638853892e-06, | |
| "loss": 0.1251, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.6031746031746033, | |
| "grad_norm": 0.8455502085192623, | |
| "learning_rate": 9.403099714207175e-07, | |
| "loss": 0.1147, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.619047619047619, | |
| "grad_norm": 0.7822227098965502, | |
| "learning_rate": 8.688061284200266e-07, | |
| "loss": 0.1325, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.6349206349206349, | |
| "grad_norm": 0.7471909044045777, | |
| "learning_rate": 7.99870384246143e-07, | |
| "loss": 0.1056, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.6507936507936507, | |
| "grad_norm": 0.8069511438267241, | |
| "learning_rate": 7.33545591839222e-07, | |
| "loss": 0.1113, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 0.7983439950759835, | |
| "learning_rate": 6.698729810778065e-07, | |
| "loss": 0.1158, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.6825396825396826, | |
| "grad_norm": 0.8429996790936396, | |
| "learning_rate": 6.088921331488568e-07, | |
| "loss": 0.1106, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.6984126984126984, | |
| "grad_norm": 0.8241505435395763, | |
| "learning_rate": 5.506409559426573e-07, | |
| "loss": 0.1198, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "grad_norm": 0.7464229994435873, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.1133, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.7301587301587302, | |
| "grad_norm": 0.6974660696481206, | |
| "learning_rate": 4.4247073844163434e-07, | |
| "loss": 0.0932, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.746031746031746, | |
| "grad_norm": 0.7571169718329668, | |
| "learning_rate": 3.9261894064796136e-07, | |
| "loss": 0.1134, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.7619047619047619, | |
| "grad_norm": 0.8233122026769766, | |
| "learning_rate": 3.4563125677897936e-07, | |
| "loss": 0.1343, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 0.7642348427528916, | |
| "learning_rate": 3.015368960704584e-07, | |
| "loss": 0.1134, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.7936507936507935, | |
| "grad_norm": 0.7559524802350545, | |
| "learning_rate": 2.6036326916434153e-07, | |
| "loss": 0.1153, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.8095238095238095, | |
| "grad_norm": 0.7773214279041375, | |
| "learning_rate": 2.2213597106929608e-07, | |
| "loss": 0.1078, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.8253968253968254, | |
| "grad_norm": 0.7680760912308965, | |
| "learning_rate": 1.8687876524993987e-07, | |
| "loss": 0.1051, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.8412698412698414, | |
| "grad_norm": 0.7115438868122161, | |
| "learning_rate": 1.5461356885461077e-07, | |
| "loss": 0.113, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.8571428571428572, | |
| "grad_norm": 0.691478980586458, | |
| "learning_rate": 1.253604390908819e-07, | |
| "loss": 0.0988, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.873015873015873, | |
| "grad_norm": 0.687791401958618, | |
| "learning_rate": 9.913756075728088e-08, | |
| "loss": 0.096, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.8888888888888888, | |
| "grad_norm": 0.7262275320357141, | |
| "learning_rate": 7.59612349389599e-08, | |
| "loss": 0.1122, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.9047619047619047, | |
| "grad_norm": 0.7002743164354781, | |
| "learning_rate": 5.584586887435739e-08, | |
| "loss": 0.101, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.9206349206349205, | |
| "grad_norm": 0.6436376226113324, | |
| "learning_rate": 3.8803966999139686e-08, | |
| "loss": 0.0939, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.9365079365079365, | |
| "grad_norm": 0.8215986730670641, | |
| "learning_rate": 2.4846123172992953e-08, | |
| "loss": 0.113, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.9523809523809523, | |
| "grad_norm": 0.7252216992457547, | |
| "learning_rate": 1.3981014094099354e-08, | |
| "loss": 0.1029, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.9682539682539684, | |
| "grad_norm": 0.8386753220451895, | |
| "learning_rate": 6.215393905388278e-09, | |
| "loss": 0.1312, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.9841269841269842, | |
| "grad_norm": 0.6788368293698803, | |
| "learning_rate": 1.5540899959187727e-09, | |
| "loss": 0.0994, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.6890938486700772, | |
| "learning_rate": 0.0, | |
| "loss": 0.0922, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 126, | |
| "total_flos": 8257007517696.0, | |
| "train_loss": 0.1798131594818736, | |
| "train_runtime": 502.2354, | |
| "train_samples_per_second": 1.999, | |
| "train_steps_per_second": 0.251 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 126, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 70000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 8257007517696.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |