| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 0, | |
| "global_step": 528, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.001893939393939394, | |
| "grad_norm": 0.451171875, | |
| "learning_rate": 9.981060606060606e-06, | |
| "loss": 1.7963, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.003787878787878788, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 9.962121212121212e-06, | |
| "loss": 1.7754, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.005681818181818182, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 9.943181818181819e-06, | |
| "loss": 1.7396, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.007575757575757576, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 9.924242424242425e-06, | |
| "loss": 1.7643, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.00946969696969697, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 9.90530303030303e-06, | |
| "loss": 1.7028, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.011363636363636364, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 9.886363636363637e-06, | |
| "loss": 1.7429, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.013257575757575758, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 9.867424242424243e-06, | |
| "loss": 1.751, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.015151515151515152, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 9.84848484848485e-06, | |
| "loss": 1.7441, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.017045454545454544, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 9.829545454545455e-06, | |
| "loss": 1.7713, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.01893939393939394, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 9.810606060606061e-06, | |
| "loss": 1.6467, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.020833333333333332, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 9.791666666666666e-06, | |
| "loss": 1.7119, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.022727272727272728, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 9.772727272727273e-06, | |
| "loss": 1.6096, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.02462121212121212, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 9.75378787878788e-06, | |
| "loss": 1.6245, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.026515151515151516, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 9.734848484848486e-06, | |
| "loss": 1.6297, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.028409090909090908, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 9.715909090909091e-06, | |
| "loss": 1.6394, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.030303030303030304, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 9.696969696969698e-06, | |
| "loss": 1.5655, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.032196969696969696, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 9.678030303030304e-06, | |
| "loss": 1.593, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.03409090909090909, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 9.65909090909091e-06, | |
| "loss": 1.6071, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.03598484848484849, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 9.640151515151516e-06, | |
| "loss": 1.4755, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.03787878787878788, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 9.621212121212122e-06, | |
| "loss": 1.562, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.03977272727272727, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 9.602272727272727e-06, | |
| "loss": 1.558, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.041666666666666664, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 9.583333333333335e-06, | |
| "loss": 1.5019, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.043560606060606064, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 9.56439393939394e-06, | |
| "loss": 1.59, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.045454545454545456, | |
| "grad_norm": 0.2392578125, | |
| "learning_rate": 9.545454545454547e-06, | |
| "loss": 1.47, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.04734848484848485, | |
| "grad_norm": 0.23046875, | |
| "learning_rate": 9.526515151515152e-06, | |
| "loss": 1.5116, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.04924242424242424, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 9.507575757575758e-06, | |
| "loss": 1.463, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.05113636363636364, | |
| "grad_norm": 0.2255859375, | |
| "learning_rate": 9.488636363636365e-06, | |
| "loss": 1.4856, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.05303030303030303, | |
| "grad_norm": 0.2119140625, | |
| "learning_rate": 9.469696969696971e-06, | |
| "loss": 1.4342, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.054924242424242424, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 9.450757575757576e-06, | |
| "loss": 1.4546, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.056818181818181816, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 9.431818181818183e-06, | |
| "loss": 1.4768, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.058712121212121215, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 9.41287878787879e-06, | |
| "loss": 1.3967, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.06060606060606061, | |
| "grad_norm": 0.2080078125, | |
| "learning_rate": 9.393939393939396e-06, | |
| "loss": 1.4417, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.0625, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 9.375000000000001e-06, | |
| "loss": 1.3645, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.06439393939393939, | |
| "grad_norm": 0.2080078125, | |
| "learning_rate": 9.356060606060606e-06, | |
| "loss": 1.4291, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.06628787878787878, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 9.337121212121212e-06, | |
| "loss": 1.3972, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.06818181818181818, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 9.318181818181819e-06, | |
| "loss": 1.3673, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.07007575757575757, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 9.299242424242425e-06, | |
| "loss": 1.3402, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.07196969696969698, | |
| "grad_norm": 0.201171875, | |
| "learning_rate": 9.28030303030303e-06, | |
| "loss": 1.4012, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.07386363636363637, | |
| "grad_norm": 0.2333984375, | |
| "learning_rate": 9.261363636363637e-06, | |
| "loss": 1.2972, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.07575757575757576, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 9.242424242424244e-06, | |
| "loss": 1.409, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07765151515151515, | |
| "grad_norm": 0.2021484375, | |
| "learning_rate": 9.22348484848485e-06, | |
| "loss": 1.3533, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.07954545454545454, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 9.204545454545455e-06, | |
| "loss": 1.4493, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.08143939393939394, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 9.185606060606062e-06, | |
| "loss": 1.3153, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.08333333333333333, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 9.166666666666666e-06, | |
| "loss": 1.3608, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.08522727272727272, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 9.147727272727273e-06, | |
| "loss": 1.3115, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.08712121212121213, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 9.12878787878788e-06, | |
| "loss": 1.324, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.08901515151515152, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 9.109848484848486e-06, | |
| "loss": 1.3352, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.09090909090909091, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 9.090909090909091e-06, | |
| "loss": 1.3193, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.0928030303030303, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 9.071969696969698e-06, | |
| "loss": 1.2729, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.0946969696969697, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 9.053030303030304e-06, | |
| "loss": 1.3045, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.09659090909090909, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 9.03409090909091e-06, | |
| "loss": 1.2968, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.09848484848484848, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 9.015151515151516e-06, | |
| "loss": 1.3272, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.10037878787878787, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 8.996212121212122e-06, | |
| "loss": 1.2825, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.10227272727272728, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 8.977272727272727e-06, | |
| "loss": 1.2588, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.10416666666666667, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 8.958333333333334e-06, | |
| "loss": 1.2485, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.10606060606060606, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 8.93939393939394e-06, | |
| "loss": 1.2276, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.10795454545454546, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 8.920454545454547e-06, | |
| "loss": 1.3342, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.10984848484848485, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 8.901515151515152e-06, | |
| "loss": 1.2678, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.11174242424242424, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 8.882575757575758e-06, | |
| "loss": 1.2523, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.11363636363636363, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 8.863636363636365e-06, | |
| "loss": 1.2582, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.11553030303030302, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 8.844696969696971e-06, | |
| "loss": 1.2021, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.11742424242424243, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 8.825757575757576e-06, | |
| "loss": 1.2091, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.11931818181818182, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 8.806818181818183e-06, | |
| "loss": 1.3111, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.12121212121212122, | |
| "grad_norm": 0.193359375, | |
| "learning_rate": 8.787878787878788e-06, | |
| "loss": 1.2265, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.12310606060606061, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 8.768939393939394e-06, | |
| "loss": 1.243, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.125, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 1.2032, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.1268939393939394, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 8.731060606060606e-06, | |
| "loss": 1.2213, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.12878787878787878, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 8.712121212121212e-06, | |
| "loss": 1.1738, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.13068181818181818, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 8.693181818181819e-06, | |
| "loss": 1.2173, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.13257575757575757, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 8.674242424242426e-06, | |
| "loss": 1.2439, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.13446969696969696, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 8.65530303030303e-06, | |
| "loss": 1.1949, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.13636363636363635, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 8.636363636363637e-06, | |
| "loss": 1.2421, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.13825757575757575, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 8.617424242424242e-06, | |
| "loss": 1.149, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.14015151515151514, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 8.59848484848485e-06, | |
| "loss": 1.1985, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.14204545454545456, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 8.579545454545455e-06, | |
| "loss": 1.2139, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.14393939393939395, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 8.560606060606062e-06, | |
| "loss": 1.2225, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.14583333333333334, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 8.541666666666666e-06, | |
| "loss": 1.2365, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.14772727272727273, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 8.522727272727273e-06, | |
| "loss": 1.2121, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.14962121212121213, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 8.50378787878788e-06, | |
| "loss": 1.1981, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.15151515151515152, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 8.484848484848486e-06, | |
| "loss": 1.1861, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.1534090909090909, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 8.465909090909091e-06, | |
| "loss": 1.2362, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.1553030303030303, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 8.446969696969698e-06, | |
| "loss": 1.2322, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.1571969696969697, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 8.428030303030304e-06, | |
| "loss": 1.1466, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.1590909090909091, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 8.40909090909091e-06, | |
| "loss": 1.1789, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.16098484848484848, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 8.390151515151516e-06, | |
| "loss": 1.2105, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.16287878787878787, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 8.371212121212122e-06, | |
| "loss": 1.1877, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.16477272727272727, | |
| "grad_norm": 0.2255859375, | |
| "learning_rate": 8.352272727272727e-06, | |
| "loss": 1.1701, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.16666666666666666, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 1.2096, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.16856060606060605, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 8.31439393939394e-06, | |
| "loss": 1.1431, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.17045454545454544, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 8.295454545454547e-06, | |
| "loss": 1.2117, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.17234848484848486, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 8.276515151515152e-06, | |
| "loss": 1.1393, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.17424242424242425, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 8.257575757575758e-06, | |
| "loss": 1.171, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.17613636363636365, | |
| "grad_norm": 0.8984375, | |
| "learning_rate": 8.238636363636365e-06, | |
| "loss": 1.1707, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.17803030303030304, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 8.219696969696971e-06, | |
| "loss": 1.1527, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.17992424242424243, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 8.200757575757576e-06, | |
| "loss": 1.1672, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 8.181818181818183e-06, | |
| "loss": 1.1511, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.18371212121212122, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 8.162878787878788e-06, | |
| "loss": 1.1525, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.1856060606060606, | |
| "grad_norm": 0.228515625, | |
| "learning_rate": 8.143939393939394e-06, | |
| "loss": 1.1724, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.1875, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 8.125000000000001e-06, | |
| "loss": 1.1699, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.1893939393939394, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 8.106060606060606e-06, | |
| "loss": 1.1535, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.19128787878787878, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 8.087121212121212e-06, | |
| "loss": 1.1673, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.19318181818181818, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 8.068181818181819e-06, | |
| "loss": 1.1941, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.19507575757575757, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 8.049242424242426e-06, | |
| "loss": 1.198, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.19696969696969696, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 8.03030303030303e-06, | |
| "loss": 1.1431, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.19886363636363635, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 8.011363636363637e-06, | |
| "loss": 1.1525, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.20075757575757575, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 7.992424242424242e-06, | |
| "loss": 1.1667, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.20265151515151514, | |
| "grad_norm": 0.2431640625, | |
| "learning_rate": 7.973484848484849e-06, | |
| "loss": 1.1767, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.20454545454545456, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 7.954545454545455e-06, | |
| "loss": 1.1665, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.20643939393939395, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 7.935606060606062e-06, | |
| "loss": 1.1504, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.20833333333333334, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 7.916666666666667e-06, | |
| "loss": 1.1342, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.21022727272727273, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 7.897727272727273e-06, | |
| "loss": 1.1203, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.21212121212121213, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 7.87878787878788e-06, | |
| "loss": 1.1428, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.21401515151515152, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 7.859848484848486e-06, | |
| "loss": 1.1109, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.2159090909090909, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 7.840909090909091e-06, | |
| "loss": 1.1376, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.2178030303030303, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 7.821969696969698e-06, | |
| "loss": 1.1574, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2196969696969697, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 7.803030303030303e-06, | |
| "loss": 1.1442, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.2215909090909091, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 7.784090909090911e-06, | |
| "loss": 1.1367, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.22348484848484848, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 7.765151515151516e-06, | |
| "loss": 1.154, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.22537878787878787, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 7.746212121212122e-06, | |
| "loss": 1.1311, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.22727272727272727, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 7.727272727272727e-06, | |
| "loss": 1.1048, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.22916666666666666, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 7.708333333333334e-06, | |
| "loss": 1.0927, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.23106060606060605, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 7.68939393939394e-06, | |
| "loss": 1.1566, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.23295454545454544, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 7.670454545454547e-06, | |
| "loss": 1.1359, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.23484848484848486, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 7.651515151515152e-06, | |
| "loss": 1.196, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.23674242424242425, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 7.632575757575758e-06, | |
| "loss": 1.0799, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.23863636363636365, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 7.613636363636364e-06, | |
| "loss": 1.1465, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.24053030303030304, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 7.594696969696971e-06, | |
| "loss": 1.1252, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.24242424242424243, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 7.5757575757575764e-06, | |
| "loss": 1.0775, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.24431818181818182, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 7.556818181818183e-06, | |
| "loss": 1.1184, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.24621212121212122, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 7.537878787878789e-06, | |
| "loss": 1.1603, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2481060606060606, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 7.518939393939395e-06, | |
| "loss": 1.1033, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.23046875, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 1.1378, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.2518939393939394, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 7.481060606060606e-06, | |
| "loss": 1.1165, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.2537878787878788, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 7.4621212121212125e-06, | |
| "loss": 1.1678, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.2556818181818182, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 7.443181818181818e-06, | |
| "loss": 1.1559, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.25757575757575757, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 7.424242424242425e-06, | |
| "loss": 1.0984, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.25946969696969696, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 7.4053030303030305e-06, | |
| "loss": 1.1306, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.26136363636363635, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 7.386363636363637e-06, | |
| "loss": 1.1568, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.26325757575757575, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 7.367424242424243e-06, | |
| "loss": 1.1173, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.26515151515151514, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 7.348484848484849e-06, | |
| "loss": 1.0802, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.26704545454545453, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 7.329545454545455e-06, | |
| "loss": 1.1028, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.2689393939393939, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 7.310606060606062e-06, | |
| "loss": 1.0935, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.2708333333333333, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 7.291666666666667e-06, | |
| "loss": 1.1048, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.2727272727272727, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 7.272727272727273e-06, | |
| "loss": 1.1398, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.2746212121212121, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 7.253787878787879e-06, | |
| "loss": 1.122, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.2765151515151515, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 7.2348484848484855e-06, | |
| "loss": 1.0773, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.2784090909090909, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 7.215909090909091e-06, | |
| "loss": 1.066, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.2803030303030303, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 7.196969696969698e-06, | |
| "loss": 1.1131, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.2821969696969697, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 7.1780303030303035e-06, | |
| "loss": 1.1247, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.2840909090909091, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 7.15909090909091e-06, | |
| "loss": 1.1012, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2859848484848485, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 7.140151515151516e-06, | |
| "loss": 1.107, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.2878787878787879, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 7.121212121212122e-06, | |
| "loss": 1.0854, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.2897727272727273, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 7.102272727272727e-06, | |
| "loss": 1.0889, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.2916666666666667, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 7.083333333333335e-06, | |
| "loss": 1.1018, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.2935606060606061, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 7.0643939393939396e-06, | |
| "loss": 1.133, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.29545454545454547, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 7.045454545454546e-06, | |
| "loss": 1.0746, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.29734848484848486, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 7.026515151515152e-06, | |
| "loss": 1.0956, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.29924242424242425, | |
| "grad_norm": 0.1826171875, | |
| "learning_rate": 7.0075757575757585e-06, | |
| "loss": 1.0624, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.30113636363636365, | |
| "grad_norm": 0.21875, | |
| "learning_rate": 6.988636363636364e-06, | |
| "loss": 1.0742, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.30303030303030304, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 6.969696969696971e-06, | |
| "loss": 1.1139, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.30492424242424243, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 6.9507575757575765e-06, | |
| "loss": 1.1004, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.3068181818181818, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 6.931818181818183e-06, | |
| "loss": 1.0942, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.3087121212121212, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 6.912878787878789e-06, | |
| "loss": 1.1033, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.3106060606060606, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 6.893939393939395e-06, | |
| "loss": 1.0978, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.3125, | |
| "grad_norm": 0.1826171875, | |
| "learning_rate": 6.875e-06, | |
| "loss": 1.1675, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3143939393939394, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 6.856060606060606e-06, | |
| "loss": 1.0792, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.3162878787878788, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 6.8371212121212125e-06, | |
| "loss": 1.097, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.3181818181818182, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 6.818181818181818e-06, | |
| "loss": 1.0423, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.32007575757575757, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 6.799242424242425e-06, | |
| "loss": 1.1157, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.32196969696969696, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 6.780303030303031e-06, | |
| "loss": 1.0916, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.32386363636363635, | |
| "grad_norm": 0.1826171875, | |
| "learning_rate": 6.761363636363637e-06, | |
| "loss": 1.1438, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.32575757575757575, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 6.742424242424243e-06, | |
| "loss": 1.1348, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.32765151515151514, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 6.7234848484848495e-06, | |
| "loss": 1.0739, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.32954545454545453, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 6.704545454545454e-06, | |
| "loss": 1.0961, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.3314393939393939, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 6.685606060606061e-06, | |
| "loss": 1.0229, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 1.1111, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.3352272727272727, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 6.647727272727273e-06, | |
| "loss": 1.0451, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.3371212121212121, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 6.628787878787879e-06, | |
| "loss": 1.0819, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.3390151515151515, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 6.6098484848484855e-06, | |
| "loss": 1.0276, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.3409090909090909, | |
| "grad_norm": 0.220703125, | |
| "learning_rate": 6.590909090909091e-06, | |
| "loss": 1.1011, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3428030303030303, | |
| "grad_norm": 0.2021484375, | |
| "learning_rate": 6.571969696969698e-06, | |
| "loss": 1.0703, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.3446969696969697, | |
| "grad_norm": 0.232421875, | |
| "learning_rate": 6.5530303030303036e-06, | |
| "loss": 1.0276, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.3465909090909091, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 6.53409090909091e-06, | |
| "loss": 1.123, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.3484848484848485, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 6.515151515151516e-06, | |
| "loss": 1.0628, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.3503787878787879, | |
| "grad_norm": 0.203125, | |
| "learning_rate": 6.4962121212121224e-06, | |
| "loss": 1.1604, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.3522727272727273, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 6.477272727272727e-06, | |
| "loss": 1.0925, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.3541666666666667, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 6.458333333333334e-06, | |
| "loss": 1.0547, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.3560606060606061, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 6.43939393939394e-06, | |
| "loss": 1.0688, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.35795454545454547, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 6.420454545454546e-06, | |
| "loss": 1.031, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.35984848484848486, | |
| "grad_norm": 0.193359375, | |
| "learning_rate": 6.401515151515152e-06, | |
| "loss": 1.09, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.36174242424242425, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 6.3825757575757585e-06, | |
| "loss": 1.096, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 6.363636363636364e-06, | |
| "loss": 1.037, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.36553030303030304, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 6.344696969696971e-06, | |
| "loss": 1.0587, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.36742424242424243, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 6.3257575757575765e-06, | |
| "loss": 1.138, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.3693181818181818, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 6.306818181818183e-06, | |
| "loss": 1.0685, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.3712121212121212, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 6.287878787878788e-06, | |
| "loss": 1.0301, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.3731060606060606, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 6.2689393939393946e-06, | |
| "loss": 1.0451, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.375, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 6.25e-06, | |
| "loss": 1.0362, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.3768939393939394, | |
| "grad_norm": 0.201171875, | |
| "learning_rate": 6.231060606060606e-06, | |
| "loss": 1.0863, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.3787878787878788, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 6.212121212121213e-06, | |
| "loss": 1.1231, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.3806818181818182, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 6.193181818181818e-06, | |
| "loss": 1.0892, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.38257575757575757, | |
| "grad_norm": 0.2021484375, | |
| "learning_rate": 6.174242424242425e-06, | |
| "loss": 1.1051, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.38446969696969696, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 6.155303030303031e-06, | |
| "loss": 1.0526, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.38636363636363635, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 6.136363636363637e-06, | |
| "loss": 1.0809, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.38825757575757575, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 6.117424242424242e-06, | |
| "loss": 1.0893, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.39015151515151514, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 6.0984848484848495e-06, | |
| "loss": 1.0341, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.39204545454545453, | |
| "grad_norm": 0.1826171875, | |
| "learning_rate": 6.079545454545454e-06, | |
| "loss": 1.0463, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.3939393939393939, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 6.060606060606061e-06, | |
| "loss": 1.039, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.3958333333333333, | |
| "grad_norm": 0.2138671875, | |
| "learning_rate": 6.041666666666667e-06, | |
| "loss": 1.1095, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.3977272727272727, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 6.022727272727273e-06, | |
| "loss": 1.0593, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.3996212121212121, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 6.003787878787879e-06, | |
| "loss": 1.1322, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.4015151515151515, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 5.984848484848486e-06, | |
| "loss": 1.1224, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.4034090909090909, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 5.965909090909091e-06, | |
| "loss": 1.059, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.4053030303030303, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 5.946969696969698e-06, | |
| "loss": 1.1349, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.4071969696969697, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 5.928030303030304e-06, | |
| "loss": 1.0584, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.4090909090909091, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 5.90909090909091e-06, | |
| "loss": 1.0697, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.4109848484848485, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 5.890151515151515e-06, | |
| "loss": 1.0758, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.4128787878787879, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 5.871212121212122e-06, | |
| "loss": 1.038, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.4147727272727273, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 5.852272727272727e-06, | |
| "loss": 1.0454, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.4166666666666667, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 5.833333333333334e-06, | |
| "loss": 1.0517, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4185606060606061, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 5.81439393939394e-06, | |
| "loss": 1.0624, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.42045454545454547, | |
| "grad_norm": 0.23046875, | |
| "learning_rate": 5.795454545454546e-06, | |
| "loss": 1.0628, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.42234848484848486, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 5.776515151515152e-06, | |
| "loss": 1.0505, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.42424242424242425, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 5.7575757575757586e-06, | |
| "loss": 1.0563, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.42613636363636365, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 5.738636363636364e-06, | |
| "loss": 1.0539, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.42803030303030304, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 5.719696969696971e-06, | |
| "loss": 1.0244, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.42992424242424243, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 5.700757575757576e-06, | |
| "loss": 1.055, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.4318181818181818, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 5.681818181818183e-06, | |
| "loss": 1.077, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.4337121212121212, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 5.662878787878788e-06, | |
| "loss": 1.0469, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.4356060606060606, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 5.643939393939395e-06, | |
| "loss": 1.0753, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.4375, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 5.625e-06, | |
| "loss": 1.1051, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.4393939393939394, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 5.606060606060606e-06, | |
| "loss": 1.0168, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.4412878787878788, | |
| "grad_norm": 0.2119140625, | |
| "learning_rate": 5.587121212121213e-06, | |
| "loss": 1.0703, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.4431818181818182, | |
| "grad_norm": 0.193359375, | |
| "learning_rate": 5.568181818181818e-06, | |
| "loss": 1.0704, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.44507575757575757, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 5.549242424242425e-06, | |
| "loss": 1.024, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.44696969696969696, | |
| "grad_norm": 0.2109375, | |
| "learning_rate": 5.530303030303031e-06, | |
| "loss": 1.0036, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.44886363636363635, | |
| "grad_norm": 0.2294921875, | |
| "learning_rate": 5.511363636363637e-06, | |
| "loss": 1.0425, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.45075757575757575, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 5.492424242424242e-06, | |
| "loss": 1.0471, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.45265151515151514, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 5.473484848484849e-06, | |
| "loss": 1.0404, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.45454545454545453, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 5.4545454545454545e-06, | |
| "loss": 1.0796, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.4564393939393939, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 5.435606060606061e-06, | |
| "loss": 1.0376, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.4583333333333333, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 5.416666666666667e-06, | |
| "loss": 1.0419, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.4602272727272727, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 5.397727272727273e-06, | |
| "loss": 1.0938, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.4621212121212121, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 5.378787878787879e-06, | |
| "loss": 1.068, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.4640151515151515, | |
| "grad_norm": 0.2353515625, | |
| "learning_rate": 5.359848484848486e-06, | |
| "loss": 1.0211, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.4659090909090909, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 5.340909090909091e-06, | |
| "loss": 1.0627, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.4678030303030303, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 5.321969696969698e-06, | |
| "loss": 1.0802, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.4696969696969697, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 5.303030303030303e-06, | |
| "loss": 1.0905, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.4715909090909091, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 5.28409090909091e-06, | |
| "loss": 1.0209, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.4734848484848485, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 5.265151515151515e-06, | |
| "loss": 1.0466, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.4753787878787879, | |
| "grad_norm": 0.216796875, | |
| "learning_rate": 5.246212121212122e-06, | |
| "loss": 1.106, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.4772727272727273, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 5.2272727272727274e-06, | |
| "loss": 1.105, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.4791666666666667, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 5.208333333333334e-06, | |
| "loss": 1.0542, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.4810606060606061, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 5.18939393939394e-06, | |
| "loss": 1.0407, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.48295454545454547, | |
| "grad_norm": 0.21875, | |
| "learning_rate": 5.170454545454546e-06, | |
| "loss": 1.0475, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.48484848484848486, | |
| "grad_norm": 0.21484375, | |
| "learning_rate": 5.151515151515152e-06, | |
| "loss": 1.0253, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.48674242424242425, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 5.132575757575759e-06, | |
| "loss": 1.0138, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.48863636363636365, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 5.113636363636364e-06, | |
| "loss": 1.0474, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.49053030303030304, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 5.094696969696971e-06, | |
| "loss": 1.1214, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.49242424242424243, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 5.075757575757576e-06, | |
| "loss": 1.0277, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.4943181818181818, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 5.056818181818182e-06, | |
| "loss": 1.065, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.4962121212121212, | |
| "grad_norm": 0.2255859375, | |
| "learning_rate": 5.037878787878788e-06, | |
| "loss": 1.0555, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.4981060606060606, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 5.018939393939395e-06, | |
| "loss": 1.0351, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 5e-06, | |
| "loss": 1.0455, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.5018939393939394, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 4.981060606060606e-06, | |
| "loss": 1.0422, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5037878787878788, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 4.962121212121213e-06, | |
| "loss": 1.0157, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.5056818181818182, | |
| "grad_norm": 0.2236328125, | |
| "learning_rate": 4.9431818181818184e-06, | |
| "loss": 1.0829, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.5075757575757576, | |
| "grad_norm": 0.2158203125, | |
| "learning_rate": 4.924242424242425e-06, | |
| "loss": 1.0898, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.509469696969697, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 4.905303030303031e-06, | |
| "loss": 1.0327, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.5113636363636364, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 4.8863636363636365e-06, | |
| "loss": 1.0568, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5132575757575758, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 4.867424242424243e-06, | |
| "loss": 1.0147, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.5151515151515151, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 4.848484848484849e-06, | |
| "loss": 1.0506, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.5170454545454546, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 4.829545454545455e-06, | |
| "loss": 1.0745, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.5189393939393939, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 4.810606060606061e-06, | |
| "loss": 1.049, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.5208333333333334, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 4.791666666666668e-06, | |
| "loss": 1.0643, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5227272727272727, | |
| "grad_norm": 0.439453125, | |
| "learning_rate": 4.772727272727273e-06, | |
| "loss": 1.0551, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.5246212121212122, | |
| "grad_norm": 0.2158203125, | |
| "learning_rate": 4.753787878787879e-06, | |
| "loss": 1.0604, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.5265151515151515, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 4.734848484848486e-06, | |
| "loss": 1.0536, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.5284090909090909, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 4.715909090909091e-06, | |
| "loss": 1.0339, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.5303030303030303, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 4.696969696969698e-06, | |
| "loss": 0.996, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5321969696969697, | |
| "grad_norm": 0.193359375, | |
| "learning_rate": 4.678030303030303e-06, | |
| "loss": 1.0682, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.5340909090909091, | |
| "grad_norm": 0.2080078125, | |
| "learning_rate": 4.6590909090909095e-06, | |
| "loss": 1.0002, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.5359848484848485, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 4.640151515151515e-06, | |
| "loss": 1.043, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.5378787878787878, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 4.621212121212122e-06, | |
| "loss": 1.0011, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.5397727272727273, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 4.6022727272727275e-06, | |
| "loss": 0.9569, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.5416666666666666, | |
| "grad_norm": 0.220703125, | |
| "learning_rate": 4.583333333333333e-06, | |
| "loss": 1.0664, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.5435606060606061, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 4.56439393939394e-06, | |
| "loss": 1.0537, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.5454545454545454, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 4.5454545454545455e-06, | |
| "loss": 1.004, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.5473484848484849, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 4.526515151515152e-06, | |
| "loss": 1.0854, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.5492424242424242, | |
| "grad_norm": 0.2158203125, | |
| "learning_rate": 4.507575757575758e-06, | |
| "loss": 1.0354, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.5511363636363636, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 4.4886363636363636e-06, | |
| "loss": 1.0331, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.553030303030303, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 4.46969696969697e-06, | |
| "loss": 1.0081, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.5549242424242424, | |
| "grad_norm": 0.228515625, | |
| "learning_rate": 4.450757575757576e-06, | |
| "loss": 1.0208, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.5568181818181818, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 4.4318181818181824e-06, | |
| "loss": 1.0202, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.5587121212121212, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 4.412878787878788e-06, | |
| "loss": 1.0211, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.5606060606060606, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 4.393939393939394e-06, | |
| "loss": 1.0362, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.5625, | |
| "grad_norm": 0.193359375, | |
| "learning_rate": 4.3750000000000005e-06, | |
| "loss": 1.0356, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.5643939393939394, | |
| "grad_norm": 0.21484375, | |
| "learning_rate": 4.356060606060606e-06, | |
| "loss": 1.0634, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.5662878787878788, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 4.337121212121213e-06, | |
| "loss": 1.0213, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.5681818181818182, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 4.3181818181818185e-06, | |
| "loss": 1.0089, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.5700757575757576, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 4.299242424242425e-06, | |
| "loss": 1.031, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.571969696969697, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 4.280303030303031e-06, | |
| "loss": 1.1084, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.5738636363636364, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 4.2613636363636365e-06, | |
| "loss": 1.0512, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.5757575757575758, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 4.242424242424243e-06, | |
| "loss": 1.0471, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.5776515151515151, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 4.223484848484849e-06, | |
| "loss": 0.985, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.5795454545454546, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 4.204545454545455e-06, | |
| "loss": 1.0386, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.5814393939393939, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 4.185606060606061e-06, | |
| "loss": 0.9627, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.5833333333333334, | |
| "grad_norm": 0.203125, | |
| "learning_rate": 4.166666666666667e-06, | |
| "loss": 1.0047, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.5852272727272727, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 4.1477272727272734e-06, | |
| "loss": 1.0111, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.5871212121212122, | |
| "grad_norm": 0.193359375, | |
| "learning_rate": 4.128787878787879e-06, | |
| "loss": 1.0222, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.5890151515151515, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 4.109848484848486e-06, | |
| "loss": 1.0487, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.5909090909090909, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 4.0909090909090915e-06, | |
| "loss": 1.0101, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.5928030303030303, | |
| "grad_norm": 0.193359375, | |
| "learning_rate": 4.071969696969697e-06, | |
| "loss": 1.0388, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.5946969696969697, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 4.053030303030303e-06, | |
| "loss": 1.0022, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.5965909090909091, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 4.0340909090909095e-06, | |
| "loss": 1.0172, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.5984848484848485, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 4.015151515151515e-06, | |
| "loss": 1.0265, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.6003787878787878, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 3.996212121212121e-06, | |
| "loss": 1.0261, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.6022727272727273, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 3.9772727272727275e-06, | |
| "loss": 1.0162, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.6041666666666666, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 3.958333333333333e-06, | |
| "loss": 1.0726, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.6060606060606061, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 3.93939393939394e-06, | |
| "loss": 1.0004, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6079545454545454, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 3.9204545454545456e-06, | |
| "loss": 1.0371, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.6098484848484849, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 3.901515151515151e-06, | |
| "loss": 1.0491, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.6117424242424242, | |
| "grad_norm": 0.220703125, | |
| "learning_rate": 3.882575757575758e-06, | |
| "loss": 1.0305, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.6136363636363636, | |
| "grad_norm": 0.236328125, | |
| "learning_rate": 3.863636363636364e-06, | |
| "loss": 1.0276, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.615530303030303, | |
| "grad_norm": 0.2392578125, | |
| "learning_rate": 3.84469696969697e-06, | |
| "loss": 1.0749, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.6174242424242424, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 3.825757575757576e-06, | |
| "loss": 1.0202, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.6193181818181818, | |
| "grad_norm": 0.228515625, | |
| "learning_rate": 3.806818181818182e-06, | |
| "loss": 1.0087, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.6212121212121212, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 3.7878787878787882e-06, | |
| "loss": 1.0445, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.6231060606060606, | |
| "grad_norm": 0.224609375, | |
| "learning_rate": 3.7689393939393944e-06, | |
| "loss": 1.0177, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.625, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.9965, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.6268939393939394, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 3.7310606060606062e-06, | |
| "loss": 1.069, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.6287878787878788, | |
| "grad_norm": 0.2138671875, | |
| "learning_rate": 3.7121212121212124e-06, | |
| "loss": 1.0465, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.6306818181818182, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 3.6931818181818186e-06, | |
| "loss": 1.05, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.6325757575757576, | |
| "grad_norm": 0.2080078125, | |
| "learning_rate": 3.6742424242424247e-06, | |
| "loss": 1.034, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.634469696969697, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 3.655303030303031e-06, | |
| "loss": 1.0072, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.6363636363636364, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 3.6363636363636366e-06, | |
| "loss": 1.0752, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.6382575757575758, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 3.6174242424242427e-06, | |
| "loss": 1.0422, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.6401515151515151, | |
| "grad_norm": 0.2421875, | |
| "learning_rate": 3.598484848484849e-06, | |
| "loss": 0.9971, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.6420454545454546, | |
| "grad_norm": 0.212890625, | |
| "learning_rate": 3.579545454545455e-06, | |
| "loss": 0.971, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.6439393939393939, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 3.560606060606061e-06, | |
| "loss": 1.025, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.6458333333333334, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 3.5416666666666673e-06, | |
| "loss": 1.024, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.6477272727272727, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 3.522727272727273e-06, | |
| "loss": 1.0118, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.6496212121212122, | |
| "grad_norm": 0.2158203125, | |
| "learning_rate": 3.5037878787878792e-06, | |
| "loss": 0.9955, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.6515151515151515, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 3.4848484848484854e-06, | |
| "loss": 1.0528, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.6534090909090909, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 3.4659090909090915e-06, | |
| "loss": 1.0043, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.6553030303030303, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 3.4469696969696977e-06, | |
| "loss": 1.0287, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.6571969696969697, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 3.428030303030303e-06, | |
| "loss": 1.0232, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.6590909090909091, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 3.409090909090909e-06, | |
| "loss": 1.0388, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.6609848484848485, | |
| "grad_norm": 0.2099609375, | |
| "learning_rate": 3.3901515151515153e-06, | |
| "loss": 1.0505, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.6628787878787878, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 3.3712121212121214e-06, | |
| "loss": 1.0098, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.6647727272727273, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 3.352272727272727e-06, | |
| "loss": 1.024, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 1.059, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.6685606060606061, | |
| "grad_norm": 0.2216796875, | |
| "learning_rate": 3.3143939393939395e-06, | |
| "loss": 1.0339, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.6704545454545454, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 3.2954545454545456e-06, | |
| "loss": 1.021, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.6723484848484849, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 3.2765151515151518e-06, | |
| "loss": 1.0581, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.6742424242424242, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 3.257575757575758e-06, | |
| "loss": 1.0264, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.6761363636363636, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 3.2386363636363637e-06, | |
| "loss": 1.0273, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.678030303030303, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 3.21969696969697e-06, | |
| "loss": 0.9971, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.6799242424242424, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 3.200757575757576e-06, | |
| "loss": 1.0068, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.6818181818181818, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 3.181818181818182e-06, | |
| "loss": 1.009, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.6837121212121212, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 3.1628787878787883e-06, | |
| "loss": 1.0302, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.6856060606060606, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 3.143939393939394e-06, | |
| "loss": 1.0213, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.6875, | |
| "grad_norm": 0.2021484375, | |
| "learning_rate": 3.125e-06, | |
| "loss": 1.0763, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.6893939393939394, | |
| "grad_norm": 0.220703125, | |
| "learning_rate": 3.1060606060606063e-06, | |
| "loss": 1.0518, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.6912878787878788, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 3.0871212121212125e-06, | |
| "loss": 1.0182, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.6931818181818182, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 3.0681818181818186e-06, | |
| "loss": 1.01, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.6950757575757576, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 3.0492424242424248e-06, | |
| "loss": 1.0204, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.696969696969697, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 3.0303030303030305e-06, | |
| "loss": 0.9715, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.6988636363636364, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 3.0113636363636366e-06, | |
| "loss": 1.0144, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.7007575757575758, | |
| "grad_norm": 0.2080078125, | |
| "learning_rate": 2.992424242424243e-06, | |
| "loss": 1.0633, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7026515151515151, | |
| "grad_norm": 0.2099609375, | |
| "learning_rate": 2.973484848484849e-06, | |
| "loss": 1.0299, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.7045454545454546, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 2.954545454545455e-06, | |
| "loss": 1.0252, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.7064393939393939, | |
| "grad_norm": 0.220703125, | |
| "learning_rate": 2.935606060606061e-06, | |
| "loss": 0.9969, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.7083333333333334, | |
| "grad_norm": 0.21484375, | |
| "learning_rate": 2.916666666666667e-06, | |
| "loss": 1.1488, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.7102272727272727, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 2.897727272727273e-06, | |
| "loss": 0.9887, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.7121212121212122, | |
| "grad_norm": 0.201171875, | |
| "learning_rate": 2.8787878787878793e-06, | |
| "loss": 1.0506, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.7140151515151515, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 2.8598484848484854e-06, | |
| "loss": 1.0164, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.7159090909090909, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 2.8409090909090916e-06, | |
| "loss": 1.0289, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.7178030303030303, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 2.8219696969696973e-06, | |
| "loss": 1.0498, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.7196969696969697, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 2.803030303030303e-06, | |
| "loss": 1.0029, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.7215909090909091, | |
| "grad_norm": 0.193359375, | |
| "learning_rate": 2.784090909090909e-06, | |
| "loss": 1.0073, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.7234848484848485, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 2.7651515151515153e-06, | |
| "loss": 0.965, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.7253787878787878, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 2.746212121212121e-06, | |
| "loss": 1.0399, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 2.7272727272727272e-06, | |
| "loss": 1.0859, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.7291666666666666, | |
| "grad_norm": 0.2099609375, | |
| "learning_rate": 2.7083333333333334e-06, | |
| "loss": 1.0315, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.7310606060606061, | |
| "grad_norm": 0.201171875, | |
| "learning_rate": 2.6893939393939395e-06, | |
| "loss": 1.0363, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.7329545454545454, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 2.6704545454545457e-06, | |
| "loss": 1.0575, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.7348484848484849, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 2.6515151515151514e-06, | |
| "loss": 0.9732, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.7367424242424242, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 2.6325757575757576e-06, | |
| "loss": 1.0255, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.7386363636363636, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 2.6136363636363637e-06, | |
| "loss": 1.0481, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.740530303030303, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 2.59469696969697e-06, | |
| "loss": 1.0179, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.7424242424242424, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 2.575757575757576e-06, | |
| "loss": 1.0121, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.7443181818181818, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 2.556818181818182e-06, | |
| "loss": 1.0497, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.7462121212121212, | |
| "grad_norm": 0.2021484375, | |
| "learning_rate": 2.537878787878788e-06, | |
| "loss": 1.0452, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.7481060606060606, | |
| "grad_norm": 0.2333984375, | |
| "learning_rate": 2.518939393939394e-06, | |
| "loss": 0.9853, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.0029, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.7518939393939394, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 2.4810606060606064e-06, | |
| "loss": 1.0078, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.7537878787878788, | |
| "grad_norm": 0.216796875, | |
| "learning_rate": 2.4621212121212125e-06, | |
| "loss": 1.0086, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.7556818181818182, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 2.4431818181818182e-06, | |
| "loss": 1.0296, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.7575757575757576, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 2.4242424242424244e-06, | |
| "loss": 1.0285, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.759469696969697, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 2.4053030303030305e-06, | |
| "loss": 1.0278, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.7613636363636364, | |
| "grad_norm": 0.2021484375, | |
| "learning_rate": 2.3863636363636367e-06, | |
| "loss": 1.003, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.7632575757575758, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 2.367424242424243e-06, | |
| "loss": 1.0172, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.7651515151515151, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 2.348484848484849e-06, | |
| "loss": 1.0342, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.7670454545454546, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 2.3295454545454547e-06, | |
| "loss": 1.0667, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.7689393939393939, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 2.310606060606061e-06, | |
| "loss": 0.9957, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.7708333333333334, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 2.2916666666666666e-06, | |
| "loss": 1.0137, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.7727272727272727, | |
| "grad_norm": 0.193359375, | |
| "learning_rate": 2.2727272727272728e-06, | |
| "loss": 1.0105, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.7746212121212122, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 2.253787878787879e-06, | |
| "loss": 1.0261, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.7765151515151515, | |
| "grad_norm": 0.22265625, | |
| "learning_rate": 2.234848484848485e-06, | |
| "loss": 1.0223, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.7784090909090909, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 2.2159090909090912e-06, | |
| "loss": 1.0132, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.7803030303030303, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 2.196969696969697e-06, | |
| "loss": 1.0187, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.7821969696969697, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 2.178030303030303e-06, | |
| "loss": 1.0506, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.7840909090909091, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 2.1590909090909092e-06, | |
| "loss": 1.0259, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.7859848484848485, | |
| "grad_norm": 0.2080078125, | |
| "learning_rate": 2.1401515151515154e-06, | |
| "loss": 1.0628, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.7878787878787878, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 2.1212121212121216e-06, | |
| "loss": 1.0166, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.7897727272727273, | |
| "grad_norm": 0.2353515625, | |
| "learning_rate": 2.1022727272727277e-06, | |
| "loss": 1.0129, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.7916666666666666, | |
| "grad_norm": 0.2109375, | |
| "learning_rate": 2.0833333333333334e-06, | |
| "loss": 1.0543, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.7935606060606061, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 2.0643939393939396e-06, | |
| "loss": 0.9744, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.7954545454545454, | |
| "grad_norm": 0.2109375, | |
| "learning_rate": 2.0454545454545457e-06, | |
| "loss": 1.0075, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.7973484848484849, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 2.0265151515151515e-06, | |
| "loss": 1.0368, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.7992424242424242, | |
| "grad_norm": 0.193359375, | |
| "learning_rate": 2.0075757575757576e-06, | |
| "loss": 1.0257, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.8011363636363636, | |
| "grad_norm": 0.193359375, | |
| "learning_rate": 1.9886363636363638e-06, | |
| "loss": 1.0123, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.803030303030303, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 1.96969696969697e-06, | |
| "loss": 1.0215, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.8049242424242424, | |
| "grad_norm": 0.2109375, | |
| "learning_rate": 1.9507575757575757e-06, | |
| "loss": 1.0081, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.8068181818181818, | |
| "grad_norm": 0.201171875, | |
| "learning_rate": 1.931818181818182e-06, | |
| "loss": 1.0203, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.8087121212121212, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 1.912878787878788e-06, | |
| "loss": 0.9781, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.8106060606060606, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 1.8939393939393941e-06, | |
| "loss": 1.0437, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.8125, | |
| "grad_norm": 0.220703125, | |
| "learning_rate": 1.8750000000000003e-06, | |
| "loss": 1.0094, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.8143939393939394, | |
| "grad_norm": 0.25, | |
| "learning_rate": 1.8560606060606062e-06, | |
| "loss": 1.0487, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.8162878787878788, | |
| "grad_norm": 0.224609375, | |
| "learning_rate": 1.8371212121212124e-06, | |
| "loss": 1.0197, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.8181818181818182, | |
| "grad_norm": 0.201171875, | |
| "learning_rate": 1.8181818181818183e-06, | |
| "loss": 0.9761, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.8200757575757576, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 1.7992424242424244e-06, | |
| "loss": 1.0172, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.821969696969697, | |
| "grad_norm": 0.2138671875, | |
| "learning_rate": 1.7803030303030306e-06, | |
| "loss": 1.0417, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.8238636363636364, | |
| "grad_norm": 0.2216796875, | |
| "learning_rate": 1.7613636363636365e-06, | |
| "loss": 1.055, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.8257575757575758, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 1.7424242424242427e-06, | |
| "loss": 1.0217, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.8276515151515151, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 1.7234848484848488e-06, | |
| "loss": 1.0212, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.8295454545454546, | |
| "grad_norm": 0.2177734375, | |
| "learning_rate": 1.7045454545454546e-06, | |
| "loss": 1.0046, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.8314393939393939, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 1.6856060606060607e-06, | |
| "loss": 1.0271, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.8333333333333334, | |
| "grad_norm": 0.2060546875, | |
| "learning_rate": 1.6666666666666667e-06, | |
| "loss": 0.9767, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.8352272727272727, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 1.6477272727272728e-06, | |
| "loss": 1.0363, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.8371212121212122, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 1.628787878787879e-06, | |
| "loss": 1.0145, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.8390151515151515, | |
| "grad_norm": 0.203125, | |
| "learning_rate": 1.609848484848485e-06, | |
| "loss": 1.0217, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.8409090909090909, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 1.590909090909091e-06, | |
| "loss": 0.9888, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.8428030303030303, | |
| "grad_norm": 0.8203125, | |
| "learning_rate": 1.571969696969697e-06, | |
| "loss": 1.0196, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.8446969696969697, | |
| "grad_norm": 0.2177734375, | |
| "learning_rate": 1.5530303030303032e-06, | |
| "loss": 1.0016, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.8465909090909091, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 1.5340909090909093e-06, | |
| "loss": 1.0335, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.8484848484848485, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 1.5151515151515152e-06, | |
| "loss": 1.0176, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.8503787878787878, | |
| "grad_norm": 0.212890625, | |
| "learning_rate": 1.4962121212121214e-06, | |
| "loss": 1.0166, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.8522727272727273, | |
| "grad_norm": 0.2236328125, | |
| "learning_rate": 1.4772727272727275e-06, | |
| "loss": 1.0212, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.8541666666666666, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 1.4583333333333335e-06, | |
| "loss": 1.029, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.8560606060606061, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 1.4393939393939396e-06, | |
| "loss": 0.9911, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.8579545454545454, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 1.4204545454545458e-06, | |
| "loss": 1.026, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.8598484848484849, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 1.4015151515151515e-06, | |
| "loss": 0.9957, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.8617424242424242, | |
| "grad_norm": 0.23046875, | |
| "learning_rate": 1.3825757575757577e-06, | |
| "loss": 1.0497, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.8636363636363636, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 1.3636363636363636e-06, | |
| "loss": 1.0472, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.865530303030303, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 1.3446969696969698e-06, | |
| "loss": 1.0525, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.8674242424242424, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 1.3257575757575757e-06, | |
| "loss": 1.0084, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.8693181818181818, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 1.3068181818181819e-06, | |
| "loss": 1.008, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.8712121212121212, | |
| "grad_norm": 0.2099609375, | |
| "learning_rate": 1.287878787878788e-06, | |
| "loss": 0.9981, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.8731060606060606, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 1.268939393939394e-06, | |
| "loss": 1.0033, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.875, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 1.25e-06, | |
| "loss": 1.0505, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.8768939393939394, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 1.2310606060606063e-06, | |
| "loss": 1.0223, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.8787878787878788, | |
| "grad_norm": 0.2099609375, | |
| "learning_rate": 1.2121212121212122e-06, | |
| "loss": 0.9805, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.8806818181818182, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 1.1931818181818183e-06, | |
| "loss": 1.0563, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.8825757575757576, | |
| "grad_norm": 0.2158203125, | |
| "learning_rate": 1.1742424242424245e-06, | |
| "loss": 1.0264, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.884469696969697, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 1.1553030303030304e-06, | |
| "loss": 1.0075, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.8863636363636364, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 1.1363636363636364e-06, | |
| "loss": 1.0187, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.8882575757575758, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 1.1174242424242425e-06, | |
| "loss": 1.0015, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.8901515151515151, | |
| "grad_norm": 0.2294921875, | |
| "learning_rate": 1.0984848484848485e-06, | |
| "loss": 1.0642, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.8920454545454546, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 1.0795454545454546e-06, | |
| "loss": 1.0108, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.8939393939393939, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 1.0606060606060608e-06, | |
| "loss": 1.0096, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.8958333333333334, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 1.0416666666666667e-06, | |
| "loss": 1.0537, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.8977272727272727, | |
| "grad_norm": 0.2158203125, | |
| "learning_rate": 1.0227272727272729e-06, | |
| "loss": 1.0312, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.8996212121212122, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 1.0037878787878788e-06, | |
| "loss": 1.0218, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.9015151515151515, | |
| "grad_norm": 0.216796875, | |
| "learning_rate": 9.84848484848485e-07, | |
| "loss": 1.06, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.9034090909090909, | |
| "grad_norm": 0.2255859375, | |
| "learning_rate": 9.65909090909091e-07, | |
| "loss": 0.9917, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.9053030303030303, | |
| "grad_norm": 0.2021484375, | |
| "learning_rate": 9.469696969696971e-07, | |
| "loss": 1.0538, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.9071969696969697, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 9.280303030303031e-07, | |
| "loss": 1.0648, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 9.090909090909091e-07, | |
| "loss": 1.0167, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.9109848484848485, | |
| "grad_norm": 0.2138671875, | |
| "learning_rate": 8.901515151515153e-07, | |
| "loss": 1.0157, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.9128787878787878, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 8.712121212121213e-07, | |
| "loss": 0.983, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.9147727272727273, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 8.522727272727273e-07, | |
| "loss": 0.9711, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.9166666666666666, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 8.333333333333333e-07, | |
| "loss": 1.0326, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.9185606060606061, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 8.143939393939395e-07, | |
| "loss": 1.0072, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.9204545454545454, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 7.954545454545455e-07, | |
| "loss": 0.9935, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.9223484848484849, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 7.765151515151516e-07, | |
| "loss": 0.972, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.9242424242424242, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 7.575757575757576e-07, | |
| "loss": 0.9696, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.9261363636363636, | |
| "grad_norm": 0.203125, | |
| "learning_rate": 7.386363636363638e-07, | |
| "loss": 1.0179, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.928030303030303, | |
| "grad_norm": 0.21875, | |
| "learning_rate": 7.196969696969698e-07, | |
| "loss": 1.0141, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.9299242424242424, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 7.007575757575758e-07, | |
| "loss": 0.9915, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.9318181818181818, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 6.818181818181818e-07, | |
| "loss": 0.9909, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.9337121212121212, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 6.628787878787879e-07, | |
| "loss": 1.0208, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.9356060606060606, | |
| "grad_norm": 0.203125, | |
| "learning_rate": 6.43939393939394e-07, | |
| "loss": 1.043, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.9375, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 6.25e-07, | |
| "loss": 0.9941, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.9393939393939394, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 6.060606060606061e-07, | |
| "loss": 1.033, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.9412878787878788, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 5.871212121212122e-07, | |
| "loss": 0.9724, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.9431818181818182, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 5.681818181818182e-07, | |
| "loss": 1.0303, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.9450757575757576, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 5.492424242424242e-07, | |
| "loss": 1.0489, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.946969696969697, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 5.303030303030304e-07, | |
| "loss": 1.0209, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.9488636363636364, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 5.113636363636364e-07, | |
| "loss": 1.0542, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.9507575757575758, | |
| "grad_norm": 0.2109375, | |
| "learning_rate": 4.924242424242425e-07, | |
| "loss": 1.0406, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.9526515151515151, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 4.7348484848484853e-07, | |
| "loss": 0.991, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.9545454545454546, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 4.5454545454545457e-07, | |
| "loss": 1.045, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.9564393939393939, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 4.3560606060606067e-07, | |
| "loss": 1.0152, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.9583333333333334, | |
| "grad_norm": 0.193359375, | |
| "learning_rate": 4.1666666666666667e-07, | |
| "loss": 1.0243, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.9602272727272727, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 3.9772727272727276e-07, | |
| "loss": 1.0035, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.9621212121212122, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 3.787878787878788e-07, | |
| "loss": 0.991, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.9640151515151515, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 3.598484848484849e-07, | |
| "loss": 1.0024, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.9659090909090909, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 3.409090909090909e-07, | |
| "loss": 1.0288, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.9678030303030303, | |
| "grad_norm": 0.2060546875, | |
| "learning_rate": 3.21969696969697e-07, | |
| "loss": 1.0204, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.9696969696969697, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 3.0303030303030305e-07, | |
| "loss": 1.0213, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.9715909090909091, | |
| "grad_norm": 0.25, | |
| "learning_rate": 2.840909090909091e-07, | |
| "loss": 1.0228, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.9734848484848485, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 2.651515151515152e-07, | |
| "loss": 1.0101, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.9753787878787878, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 2.4621212121212124e-07, | |
| "loss": 1.068, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.9772727272727273, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 2.2727272727272729e-07, | |
| "loss": 1.0313, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.9791666666666666, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 2.0833333333333333e-07, | |
| "loss": 0.9904, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.9810606060606061, | |
| "grad_norm": 0.2060546875, | |
| "learning_rate": 1.893939393939394e-07, | |
| "loss": 1.0292, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.9829545454545454, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 1.7045454545454545e-07, | |
| "loss": 1.0392, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.9848484848484849, | |
| "grad_norm": 0.2109375, | |
| "learning_rate": 1.5151515151515152e-07, | |
| "loss": 1.0461, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.9867424242424242, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 1.325757575757576e-07, | |
| "loss": 1.0209, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.9886363636363636, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 1.1363636363636364e-07, | |
| "loss": 1.0252, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.990530303030303, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 9.46969696969697e-08, | |
| "loss": 0.9835, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.9924242424242424, | |
| "grad_norm": 0.203125, | |
| "learning_rate": 7.575757575757576e-08, | |
| "loss": 1.0126, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.9943181818181818, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 5.681818181818182e-08, | |
| "loss": 0.9871, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.9962121212121212, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 3.787878787878788e-08, | |
| "loss": 1.0243, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.9981060606060606, | |
| "grad_norm": 0.21484375, | |
| "learning_rate": 1.893939393939394e-08, | |
| "loss": 1.0246, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.2431640625, | |
| "learning_rate": 0.0, | |
| "loss": 0.9495, | |
| "step": 528 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 528, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 0, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.695542908054143e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |