| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 882, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0011337868480725624, | |
| "grad_norm": 74.33029174804688, | |
| "learning_rate": 0.0, | |
| "loss": 2.7802, | |
| "num_tokens": 1145.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0022675736961451248, | |
| "grad_norm": 90.74027252197266, | |
| "learning_rate": 7.407407407407407e-07, | |
| "loss": 3.001, | |
| "num_tokens": 2223.0, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.003401360544217687, | |
| "grad_norm": 84.00408172607422, | |
| "learning_rate": 1.4814814814814815e-06, | |
| "loss": 3.0741, | |
| "num_tokens": 3315.0, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0045351473922902496, | |
| "grad_norm": 74.34835052490234, | |
| "learning_rate": 2.222222222222222e-06, | |
| "loss": 2.9966, | |
| "num_tokens": 4429.0, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.005668934240362812, | |
| "grad_norm": 74.42596435546875, | |
| "learning_rate": 2.962962962962963e-06, | |
| "loss": 2.8384, | |
| "num_tokens": 5594.0, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.006802721088435374, | |
| "grad_norm": 86.56283569335938, | |
| "learning_rate": 3.7037037037037037e-06, | |
| "loss": 2.9764, | |
| "num_tokens": 6751.0, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.007936507936507936, | |
| "grad_norm": 83.2591323852539, | |
| "learning_rate": 4.444444444444444e-06, | |
| "loss": 3.067, | |
| "num_tokens": 7768.0, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.009070294784580499, | |
| "grad_norm": 79.88240814208984, | |
| "learning_rate": 5.185185185185185e-06, | |
| "loss": 2.7835, | |
| "num_tokens": 8885.0, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.01020408163265306, | |
| "grad_norm": 62.56035232543945, | |
| "learning_rate": 5.925925925925926e-06, | |
| "loss": 2.8879, | |
| "num_tokens": 10034.0, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.011337868480725623, | |
| "grad_norm": 65.7270278930664, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 2.7762, | |
| "num_tokens": 11110.0, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.012471655328798186, | |
| "grad_norm": 66.46719360351562, | |
| "learning_rate": 7.4074074074074075e-06, | |
| "loss": 2.9042, | |
| "num_tokens": 12163.0, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.013605442176870748, | |
| "grad_norm": 54.481658935546875, | |
| "learning_rate": 8.148148148148148e-06, | |
| "loss": 2.8072, | |
| "num_tokens": 13214.0, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.01473922902494331, | |
| "grad_norm": 47.543304443359375, | |
| "learning_rate": 8.888888888888888e-06, | |
| "loss": 2.5354, | |
| "num_tokens": 14332.0, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.015873015873015872, | |
| "grad_norm": 43.516944885253906, | |
| "learning_rate": 9.62962962962963e-06, | |
| "loss": 2.6121, | |
| "num_tokens": 15486.0, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.017006802721088437, | |
| "grad_norm": 51.404109954833984, | |
| "learning_rate": 1.037037037037037e-05, | |
| "loss": 2.4756, | |
| "num_tokens": 16505.0, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.018140589569160998, | |
| "grad_norm": 49.24734878540039, | |
| "learning_rate": 1.1111111111111113e-05, | |
| "loss": 2.537, | |
| "num_tokens": 17529.0, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.01927437641723356, | |
| "grad_norm": 44.85908508300781, | |
| "learning_rate": 1.1851851851851852e-05, | |
| "loss": 2.5237, | |
| "num_tokens": 18625.0, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.02040816326530612, | |
| "grad_norm": 42.729736328125, | |
| "learning_rate": 1.2592592592592593e-05, | |
| "loss": 2.3821, | |
| "num_tokens": 19757.0, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.021541950113378686, | |
| "grad_norm": 42.92080307006836, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 2.3942, | |
| "num_tokens": 20818.0, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.022675736961451247, | |
| "grad_norm": 38.561309814453125, | |
| "learning_rate": 1.4074074074074075e-05, | |
| "loss": 2.4325, | |
| "num_tokens": 21939.0, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.023809523809523808, | |
| "grad_norm": 36.45574188232422, | |
| "learning_rate": 1.4814814814814815e-05, | |
| "loss": 2.1893, | |
| "num_tokens": 23014.0, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.024943310657596373, | |
| "grad_norm": 31.59442901611328, | |
| "learning_rate": 1.555555555555556e-05, | |
| "loss": 2.2218, | |
| "num_tokens": 24169.0, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.026077097505668934, | |
| "grad_norm": 32.13104248046875, | |
| "learning_rate": 1.6296296296296297e-05, | |
| "loss": 2.2421, | |
| "num_tokens": 25249.0, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.027210884353741496, | |
| "grad_norm": 30.995559692382812, | |
| "learning_rate": 1.7037037037037038e-05, | |
| "loss": 2.0935, | |
| "num_tokens": 26313.0, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.02834467120181406, | |
| "grad_norm": 33.38172912597656, | |
| "learning_rate": 1.7777777777777777e-05, | |
| "loss": 2.1663, | |
| "num_tokens": 27351.0, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.02947845804988662, | |
| "grad_norm": 30.938949584960938, | |
| "learning_rate": 1.851851851851852e-05, | |
| "loss": 1.9916, | |
| "num_tokens": 28367.0, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.030612244897959183, | |
| "grad_norm": 31.419017791748047, | |
| "learning_rate": 1.925925925925926e-05, | |
| "loss": 1.9991, | |
| "num_tokens": 29437.0, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.031746031746031744, | |
| "grad_norm": 28.316110610961914, | |
| "learning_rate": 2e-05, | |
| "loss": 1.9932, | |
| "num_tokens": 30576.0, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.032879818594104306, | |
| "grad_norm": 28.506546020507812, | |
| "learning_rate": 1.9976608187134504e-05, | |
| "loss": 1.9384, | |
| "num_tokens": 31772.0, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.034013605442176874, | |
| "grad_norm": 30.74509048461914, | |
| "learning_rate": 1.9953216374269007e-05, | |
| "loss": 1.897, | |
| "num_tokens": 32868.0, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.035147392290249435, | |
| "grad_norm": 29.190940856933594, | |
| "learning_rate": 1.992982456140351e-05, | |
| "loss": 2.0675, | |
| "num_tokens": 34025.0, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.036281179138321996, | |
| "grad_norm": 27.97025489807129, | |
| "learning_rate": 1.9906432748538015e-05, | |
| "loss": 1.9301, | |
| "num_tokens": 35175.0, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.03741496598639456, | |
| "grad_norm": 26.162158966064453, | |
| "learning_rate": 1.9883040935672515e-05, | |
| "loss": 1.8323, | |
| "num_tokens": 36242.0, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.03854875283446712, | |
| "grad_norm": 23.444116592407227, | |
| "learning_rate": 1.9859649122807017e-05, | |
| "loss": 1.9189, | |
| "num_tokens": 37352.0, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.03968253968253968, | |
| "grad_norm": 24.441686630249023, | |
| "learning_rate": 1.9836257309941523e-05, | |
| "loss": 1.8367, | |
| "num_tokens": 38425.0, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.04081632653061224, | |
| "grad_norm": 23.87506675720215, | |
| "learning_rate": 1.9812865497076026e-05, | |
| "loss": 1.7329, | |
| "num_tokens": 39617.0, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.04195011337868481, | |
| "grad_norm": 24.666210174560547, | |
| "learning_rate": 1.9789473684210528e-05, | |
| "loss": 1.8583, | |
| "num_tokens": 40651.0, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.04308390022675737, | |
| "grad_norm": 21.076745986938477, | |
| "learning_rate": 1.976608187134503e-05, | |
| "loss": 1.7904, | |
| "num_tokens": 41788.0, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.04421768707482993, | |
| "grad_norm": 18.543474197387695, | |
| "learning_rate": 1.9742690058479533e-05, | |
| "loss": 1.807, | |
| "num_tokens": 42935.0, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.045351473922902494, | |
| "grad_norm": 19.128664016723633, | |
| "learning_rate": 1.9719298245614036e-05, | |
| "loss": 1.6529, | |
| "num_tokens": 44074.0, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.046485260770975055, | |
| "grad_norm": 19.04387855529785, | |
| "learning_rate": 1.969590643274854e-05, | |
| "loss": 1.9339, | |
| "num_tokens": 45240.0, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.047619047619047616, | |
| "grad_norm": 18.83653450012207, | |
| "learning_rate": 1.9672514619883044e-05, | |
| "loss": 1.7952, | |
| "num_tokens": 46381.0, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.048752834467120185, | |
| "grad_norm": 20.276226043701172, | |
| "learning_rate": 1.9649122807017544e-05, | |
| "loss": 1.7274, | |
| "num_tokens": 47452.0, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.049886621315192746, | |
| "grad_norm": 18.112844467163086, | |
| "learning_rate": 1.962573099415205e-05, | |
| "loss": 1.7377, | |
| "num_tokens": 48609.0, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.05102040816326531, | |
| "grad_norm": 18.350265502929688, | |
| "learning_rate": 1.9602339181286552e-05, | |
| "loss": 1.4819, | |
| "num_tokens": 49744.0, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.05215419501133787, | |
| "grad_norm": 19.062835693359375, | |
| "learning_rate": 1.9578947368421055e-05, | |
| "loss": 1.8832, | |
| "num_tokens": 50850.0, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.05328798185941043, | |
| "grad_norm": 21.639110565185547, | |
| "learning_rate": 1.9555555555555557e-05, | |
| "loss": 1.7446, | |
| "num_tokens": 51963.0, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.05442176870748299, | |
| "grad_norm": 18.648723602294922, | |
| "learning_rate": 1.953216374269006e-05, | |
| "loss": 1.6512, | |
| "num_tokens": 53070.0, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.05555555555555555, | |
| "grad_norm": 21.16545867919922, | |
| "learning_rate": 1.9508771929824562e-05, | |
| "loss": 1.7065, | |
| "num_tokens": 54154.0, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.05668934240362812, | |
| "grad_norm": 18.10296630859375, | |
| "learning_rate": 1.9485380116959065e-05, | |
| "loss": 1.5725, | |
| "num_tokens": 55278.0, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05782312925170068, | |
| "grad_norm": 18.74391746520996, | |
| "learning_rate": 1.9461988304093568e-05, | |
| "loss": 1.8522, | |
| "num_tokens": 56381.0, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.05895691609977324, | |
| "grad_norm": 20.406055450439453, | |
| "learning_rate": 1.9438596491228074e-05, | |
| "loss": 1.6082, | |
| "num_tokens": 57413.0, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.060090702947845805, | |
| "grad_norm": 19.050594329833984, | |
| "learning_rate": 1.9415204678362573e-05, | |
| "loss": 1.7722, | |
| "num_tokens": 58555.0, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.061224489795918366, | |
| "grad_norm": 19.197467803955078, | |
| "learning_rate": 1.939181286549708e-05, | |
| "loss": 1.6229, | |
| "num_tokens": 59581.0, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.06235827664399093, | |
| "grad_norm": 20.112733840942383, | |
| "learning_rate": 1.936842105263158e-05, | |
| "loss": 1.71, | |
| "num_tokens": 60776.0, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.06349206349206349, | |
| "grad_norm": 19.704727172851562, | |
| "learning_rate": 1.9345029239766084e-05, | |
| "loss": 1.6764, | |
| "num_tokens": 61957.0, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.06462585034013606, | |
| "grad_norm": 19.901477813720703, | |
| "learning_rate": 1.9321637426900586e-05, | |
| "loss": 1.6584, | |
| "num_tokens": 63043.0, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.06575963718820861, | |
| "grad_norm": 19.40325927734375, | |
| "learning_rate": 1.929824561403509e-05, | |
| "loss": 1.7776, | |
| "num_tokens": 64129.0, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.06689342403628118, | |
| "grad_norm": 19.251150131225586, | |
| "learning_rate": 1.927485380116959e-05, | |
| "loss": 1.5785, | |
| "num_tokens": 65256.0, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.06802721088435375, | |
| "grad_norm": 18.591737747192383, | |
| "learning_rate": 1.9251461988304094e-05, | |
| "loss": 1.6411, | |
| "num_tokens": 66376.0, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.0691609977324263, | |
| "grad_norm": 18.51482582092285, | |
| "learning_rate": 1.9228070175438597e-05, | |
| "loss": 1.6408, | |
| "num_tokens": 67572.0, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.07029478458049887, | |
| "grad_norm": 20.782194137573242, | |
| "learning_rate": 1.9204678362573103e-05, | |
| "loss": 1.5661, | |
| "num_tokens": 68652.0, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.07142857142857142, | |
| "grad_norm": 20.952245712280273, | |
| "learning_rate": 1.9181286549707602e-05, | |
| "loss": 1.6445, | |
| "num_tokens": 69693.0, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.07256235827664399, | |
| "grad_norm": 19.55459213256836, | |
| "learning_rate": 1.9157894736842108e-05, | |
| "loss": 1.6929, | |
| "num_tokens": 70855.0, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.07369614512471655, | |
| "grad_norm": 19.258163452148438, | |
| "learning_rate": 1.913450292397661e-05, | |
| "loss": 1.5356, | |
| "num_tokens": 71899.0, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.07482993197278912, | |
| "grad_norm": 19.81553840637207, | |
| "learning_rate": 1.9111111111111113e-05, | |
| "loss": 1.5472, | |
| "num_tokens": 73092.0, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.07596371882086168, | |
| "grad_norm": 18.654260635375977, | |
| "learning_rate": 1.9087719298245616e-05, | |
| "loss": 1.6884, | |
| "num_tokens": 74232.0, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.07709750566893424, | |
| "grad_norm": 20.07318878173828, | |
| "learning_rate": 1.9064327485380118e-05, | |
| "loss": 1.7197, | |
| "num_tokens": 75347.0, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.0782312925170068, | |
| "grad_norm": 20.34659767150879, | |
| "learning_rate": 1.904093567251462e-05, | |
| "loss": 1.736, | |
| "num_tokens": 76455.0, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.07936507936507936, | |
| "grad_norm": 18.60862159729004, | |
| "learning_rate": 1.9017543859649123e-05, | |
| "loss": 1.7258, | |
| "num_tokens": 77648.0, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.08049886621315193, | |
| "grad_norm": 19.377056121826172, | |
| "learning_rate": 1.8994152046783626e-05, | |
| "loss": 1.7971, | |
| "num_tokens": 78688.0, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.08163265306122448, | |
| "grad_norm": 18.83735466003418, | |
| "learning_rate": 1.8970760233918132e-05, | |
| "loss": 1.4687, | |
| "num_tokens": 79760.0, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.08276643990929705, | |
| "grad_norm": 19.88797950744629, | |
| "learning_rate": 1.894736842105263e-05, | |
| "loss": 1.6935, | |
| "num_tokens": 80889.0, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.08390022675736962, | |
| "grad_norm": 20.648218154907227, | |
| "learning_rate": 1.8923976608187137e-05, | |
| "loss": 1.5982, | |
| "num_tokens": 81898.0, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.08503401360544217, | |
| "grad_norm": 20.697324752807617, | |
| "learning_rate": 1.890058479532164e-05, | |
| "loss": 1.7866, | |
| "num_tokens": 83019.0, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.08616780045351474, | |
| "grad_norm": 21.66934585571289, | |
| "learning_rate": 1.8877192982456142e-05, | |
| "loss": 1.7473, | |
| "num_tokens": 84034.0, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.0873015873015873, | |
| "grad_norm": 18.273752212524414, | |
| "learning_rate": 1.8853801169590645e-05, | |
| "loss": 1.6144, | |
| "num_tokens": 85221.0, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.08843537414965986, | |
| "grad_norm": 21.30136489868164, | |
| "learning_rate": 1.8830409356725147e-05, | |
| "loss": 1.6525, | |
| "num_tokens": 86291.0, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.08956916099773243, | |
| "grad_norm": 21.55545425415039, | |
| "learning_rate": 1.880701754385965e-05, | |
| "loss": 1.7933, | |
| "num_tokens": 87273.0, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.09070294784580499, | |
| "grad_norm": 20.109617233276367, | |
| "learning_rate": 1.8783625730994152e-05, | |
| "loss": 1.6552, | |
| "num_tokens": 88353.0, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.09183673469387756, | |
| "grad_norm": 20.001720428466797, | |
| "learning_rate": 1.8760233918128655e-05, | |
| "loss": 1.6845, | |
| "num_tokens": 89389.0, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.09297052154195011, | |
| "grad_norm": 20.805374145507812, | |
| "learning_rate": 1.873684210526316e-05, | |
| "loss": 1.597, | |
| "num_tokens": 90434.0, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.09410430839002268, | |
| "grad_norm": 18.642532348632812, | |
| "learning_rate": 1.871345029239766e-05, | |
| "loss": 1.5591, | |
| "num_tokens": 91566.0, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.09523809523809523, | |
| "grad_norm": 20.010791778564453, | |
| "learning_rate": 1.8690058479532166e-05, | |
| "loss": 1.6093, | |
| "num_tokens": 92699.0, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.0963718820861678, | |
| "grad_norm": 21.23207664489746, | |
| "learning_rate": 1.866666666666667e-05, | |
| "loss": 1.7362, | |
| "num_tokens": 93770.0, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.09750566893424037, | |
| "grad_norm": 22.148468017578125, | |
| "learning_rate": 1.864327485380117e-05, | |
| "loss": 1.6206, | |
| "num_tokens": 94927.0, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.09863945578231292, | |
| "grad_norm": 19.68358039855957, | |
| "learning_rate": 1.8619883040935674e-05, | |
| "loss": 1.7617, | |
| "num_tokens": 96064.0, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.09977324263038549, | |
| "grad_norm": 20.17274284362793, | |
| "learning_rate": 1.8596491228070176e-05, | |
| "loss": 1.6107, | |
| "num_tokens": 97169.0, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.10090702947845805, | |
| "grad_norm": 19.35859489440918, | |
| "learning_rate": 1.857309941520468e-05, | |
| "loss": 1.5928, | |
| "num_tokens": 98366.0, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.10204081632653061, | |
| "grad_norm": 21.101551055908203, | |
| "learning_rate": 1.854970760233918e-05, | |
| "loss": 1.6669, | |
| "num_tokens": 99457.0, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.10317460317460317, | |
| "grad_norm": 21.70389175415039, | |
| "learning_rate": 1.8526315789473684e-05, | |
| "loss": 1.7066, | |
| "num_tokens": 100599.0, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.10430839002267574, | |
| "grad_norm": 19.57090950012207, | |
| "learning_rate": 1.850292397660819e-05, | |
| "loss": 1.4318, | |
| "num_tokens": 101813.0, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.1054421768707483, | |
| "grad_norm": 20.19015884399414, | |
| "learning_rate": 1.847953216374269e-05, | |
| "loss": 1.5283, | |
| "num_tokens": 102920.0, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.10657596371882086, | |
| "grad_norm": 21.393760681152344, | |
| "learning_rate": 1.8456140350877195e-05, | |
| "loss": 1.574, | |
| "num_tokens": 104083.0, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.10770975056689343, | |
| "grad_norm": 19.35205078125, | |
| "learning_rate": 1.8432748538011698e-05, | |
| "loss": 1.4985, | |
| "num_tokens": 105252.0, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.10884353741496598, | |
| "grad_norm": 22.61254119873047, | |
| "learning_rate": 1.84093567251462e-05, | |
| "loss": 1.5807, | |
| "num_tokens": 106354.0, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.10997732426303855, | |
| "grad_norm": 21.425067901611328, | |
| "learning_rate": 1.8385964912280703e-05, | |
| "loss": 1.6194, | |
| "num_tokens": 107532.0, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.1111111111111111, | |
| "grad_norm": 20.971893310546875, | |
| "learning_rate": 1.8362573099415205e-05, | |
| "loss": 1.6884, | |
| "num_tokens": 108667.0, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.11224489795918367, | |
| "grad_norm": 22.08867073059082, | |
| "learning_rate": 1.833918128654971e-05, | |
| "loss": 1.6521, | |
| "num_tokens": 109714.0, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.11337868480725624, | |
| "grad_norm": 19.163150787353516, | |
| "learning_rate": 1.831578947368421e-05, | |
| "loss": 1.7947, | |
| "num_tokens": 110894.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.1145124716553288, | |
| "grad_norm": 20.42814826965332, | |
| "learning_rate": 1.8292397660818713e-05, | |
| "loss": 1.6677, | |
| "num_tokens": 111990.0, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.11564625850340136, | |
| "grad_norm": 20.525667190551758, | |
| "learning_rate": 1.826900584795322e-05, | |
| "loss": 1.5593, | |
| "num_tokens": 113085.0, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.11678004535147392, | |
| "grad_norm": 21.172090530395508, | |
| "learning_rate": 1.824561403508772e-05, | |
| "loss": 1.6892, | |
| "num_tokens": 114209.0, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.11791383219954649, | |
| "grad_norm": 19.004167556762695, | |
| "learning_rate": 1.8222222222222224e-05, | |
| "loss": 1.6172, | |
| "num_tokens": 115268.0, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.11904761904761904, | |
| "grad_norm": 21.883514404296875, | |
| "learning_rate": 1.8198830409356727e-05, | |
| "loss": 1.5572, | |
| "num_tokens": 116382.0, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.12018140589569161, | |
| "grad_norm": 20.1040096282959, | |
| "learning_rate": 1.817543859649123e-05, | |
| "loss": 1.6517, | |
| "num_tokens": 117504.0, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.12131519274376418, | |
| "grad_norm": 19.666257858276367, | |
| "learning_rate": 1.8152046783625732e-05, | |
| "loss": 1.6365, | |
| "num_tokens": 118698.0, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.12244897959183673, | |
| "grad_norm": 21.657516479492188, | |
| "learning_rate": 1.8128654970760235e-05, | |
| "loss": 1.7495, | |
| "num_tokens": 119791.0, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.1235827664399093, | |
| "grad_norm": 20.718935012817383, | |
| "learning_rate": 1.810526315789474e-05, | |
| "loss": 1.7333, | |
| "num_tokens": 120996.0, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.12471655328798185, | |
| "grad_norm": 20.30824851989746, | |
| "learning_rate": 1.808187134502924e-05, | |
| "loss": 1.5361, | |
| "num_tokens": 122066.0, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.12585034013605442, | |
| "grad_norm": 21.735916137695312, | |
| "learning_rate": 1.8058479532163746e-05, | |
| "loss": 1.5749, | |
| "num_tokens": 123250.0, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.12698412698412698, | |
| "grad_norm": 22.563383102416992, | |
| "learning_rate": 1.8035087719298248e-05, | |
| "loss": 1.562, | |
| "num_tokens": 124337.0, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.12811791383219956, | |
| "grad_norm": 22.8643856048584, | |
| "learning_rate": 1.8011695906432747e-05, | |
| "loss": 1.5494, | |
| "num_tokens": 125423.0, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.1292517006802721, | |
| "grad_norm": 22.234722137451172, | |
| "learning_rate": 1.7988304093567253e-05, | |
| "loss": 1.7195, | |
| "num_tokens": 126544.0, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.13038548752834467, | |
| "grad_norm": 21.315336227416992, | |
| "learning_rate": 1.7964912280701756e-05, | |
| "loss": 1.7304, | |
| "num_tokens": 127718.0, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.13151927437641722, | |
| "grad_norm": 22.4656925201416, | |
| "learning_rate": 1.794152046783626e-05, | |
| "loss": 1.6433, | |
| "num_tokens": 128897.0, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.1326530612244898, | |
| "grad_norm": 22.329336166381836, | |
| "learning_rate": 1.791812865497076e-05, | |
| "loss": 1.6484, | |
| "num_tokens": 130034.0, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.13378684807256236, | |
| "grad_norm": 20.504932403564453, | |
| "learning_rate": 1.7894736842105264e-05, | |
| "loss": 1.5951, | |
| "num_tokens": 131196.0, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.1349206349206349, | |
| "grad_norm": 20.69579315185547, | |
| "learning_rate": 1.787134502923977e-05, | |
| "loss": 1.589, | |
| "num_tokens": 132362.0, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.1360544217687075, | |
| "grad_norm": 21.02338218688965, | |
| "learning_rate": 1.784795321637427e-05, | |
| "loss": 1.6319, | |
| "num_tokens": 133454.0, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.13718820861678005, | |
| "grad_norm": 21.11552619934082, | |
| "learning_rate": 1.7824561403508775e-05, | |
| "loss": 1.6886, | |
| "num_tokens": 134552.0, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.1383219954648526, | |
| "grad_norm": 19.719133377075195, | |
| "learning_rate": 1.7801169590643277e-05, | |
| "loss": 1.6521, | |
| "num_tokens": 135670.0, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.13945578231292516, | |
| "grad_norm": 25.066865921020508, | |
| "learning_rate": 1.7777777777777777e-05, | |
| "loss": 1.743, | |
| "num_tokens": 136682.0, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.14058956916099774, | |
| "grad_norm": 21.543785095214844, | |
| "learning_rate": 1.7754385964912283e-05, | |
| "loss": 1.6218, | |
| "num_tokens": 137759.0, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.1417233560090703, | |
| "grad_norm": 21.717317581176758, | |
| "learning_rate": 1.7730994152046785e-05, | |
| "loss": 1.6657, | |
| "num_tokens": 138859.0, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 21.426856994628906, | |
| "learning_rate": 1.7707602339181288e-05, | |
| "loss": 1.4688, | |
| "num_tokens": 140019.0, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.14399092970521543, | |
| "grad_norm": 23.49262237548828, | |
| "learning_rate": 1.768421052631579e-05, | |
| "loss": 1.5918, | |
| "num_tokens": 141070.0, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.14512471655328799, | |
| "grad_norm": 23.134132385253906, | |
| "learning_rate": 1.7660818713450293e-05, | |
| "loss": 1.5323, | |
| "num_tokens": 142190.0, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.14625850340136054, | |
| "grad_norm": 21.138839721679688, | |
| "learning_rate": 1.76374269005848e-05, | |
| "loss": 1.4775, | |
| "num_tokens": 143354.0, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.1473922902494331, | |
| "grad_norm": 25.536008834838867, | |
| "learning_rate": 1.7614035087719298e-05, | |
| "loss": 1.5466, | |
| "num_tokens": 144426.0, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.14852607709750568, | |
| "grad_norm": 23.146469116210938, | |
| "learning_rate": 1.7590643274853804e-05, | |
| "loss": 1.5849, | |
| "num_tokens": 145548.0, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.14965986394557823, | |
| "grad_norm": 25.74847412109375, | |
| "learning_rate": 1.7567251461988307e-05, | |
| "loss": 1.5719, | |
| "num_tokens": 146672.0, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.15079365079365079, | |
| "grad_norm": 27.166261672973633, | |
| "learning_rate": 1.754385964912281e-05, | |
| "loss": 1.5535, | |
| "num_tokens": 147779.0, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.15192743764172337, | |
| "grad_norm": 36.15412139892578, | |
| "learning_rate": 1.752046783625731e-05, | |
| "loss": 1.5972, | |
| "num_tokens": 148689.0, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.15306122448979592, | |
| "grad_norm": 27.721406936645508, | |
| "learning_rate": 1.7497076023391814e-05, | |
| "loss": 1.7406, | |
| "num_tokens": 149900.0, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.15419501133786848, | |
| "grad_norm": 28.549758911132812, | |
| "learning_rate": 1.7473684210526317e-05, | |
| "loss": 1.5852, | |
| "num_tokens": 151079.0, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.15532879818594103, | |
| "grad_norm": 29.91474151611328, | |
| "learning_rate": 1.745029239766082e-05, | |
| "loss": 1.6233, | |
| "num_tokens": 152168.0, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.1564625850340136, | |
| "grad_norm": 29.763633728027344, | |
| "learning_rate": 1.7426900584795322e-05, | |
| "loss": 1.4513, | |
| "num_tokens": 153262.0, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.15759637188208617, | |
| "grad_norm": 30.80873680114746, | |
| "learning_rate": 1.7403508771929828e-05, | |
| "loss": 1.4502, | |
| "num_tokens": 154398.0, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.15873015873015872, | |
| "grad_norm": 32.4648323059082, | |
| "learning_rate": 1.7380116959064327e-05, | |
| "loss": 1.515, | |
| "num_tokens": 155438.0, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.1598639455782313, | |
| "grad_norm": 30.387754440307617, | |
| "learning_rate": 1.7356725146198833e-05, | |
| "loss": 1.5156, | |
| "num_tokens": 156535.0, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.16099773242630386, | |
| "grad_norm": 26.93579864501953, | |
| "learning_rate": 1.7333333333333336e-05, | |
| "loss": 1.5558, | |
| "num_tokens": 157676.0, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.1621315192743764, | |
| "grad_norm": 23.415470123291016, | |
| "learning_rate": 1.7309941520467838e-05, | |
| "loss": 1.5061, | |
| "num_tokens": 158849.0, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.16326530612244897, | |
| "grad_norm": 25.518234252929688, | |
| "learning_rate": 1.728654970760234e-05, | |
| "loss": 1.7354, | |
| "num_tokens": 159943.0, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.16439909297052155, | |
| "grad_norm": 20.67085838317871, | |
| "learning_rate": 1.7263157894736843e-05, | |
| "loss": 1.7097, | |
| "num_tokens": 161152.0, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.1655328798185941, | |
| "grad_norm": 20.089345932006836, | |
| "learning_rate": 1.7239766081871346e-05, | |
| "loss": 1.5232, | |
| "num_tokens": 162350.0, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.16666666666666666, | |
| "grad_norm": 22.614852905273438, | |
| "learning_rate": 1.721637426900585e-05, | |
| "loss": 1.3706, | |
| "num_tokens": 163374.0, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.16780045351473924, | |
| "grad_norm": 21.45526695251465, | |
| "learning_rate": 1.719298245614035e-05, | |
| "loss": 1.5264, | |
| "num_tokens": 164537.0, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.1689342403628118, | |
| "grad_norm": 23.220937728881836, | |
| "learning_rate": 1.7169590643274857e-05, | |
| "loss": 1.4362, | |
| "num_tokens": 165585.0, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.17006802721088435, | |
| "grad_norm": 24.269975662231445, | |
| "learning_rate": 1.7146198830409356e-05, | |
| "loss": 1.5262, | |
| "num_tokens": 166600.0, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.1712018140589569, | |
| "grad_norm": 23.387495040893555, | |
| "learning_rate": 1.7122807017543862e-05, | |
| "loss": 1.4802, | |
| "num_tokens": 167686.0, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.17233560090702948, | |
| "grad_norm": 23.54978370666504, | |
| "learning_rate": 1.7099415204678365e-05, | |
| "loss": 1.5707, | |
| "num_tokens": 168793.0, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.17346938775510204, | |
| "grad_norm": 25.489028930664062, | |
| "learning_rate": 1.7076023391812867e-05, | |
| "loss": 1.5879, | |
| "num_tokens": 169843.0, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.1746031746031746, | |
| "grad_norm": 23.096094131469727, | |
| "learning_rate": 1.705263157894737e-05, | |
| "loss": 1.4339, | |
| "num_tokens": 170851.0, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.17573696145124718, | |
| "grad_norm": 22.319326400756836, | |
| "learning_rate": 1.7029239766081872e-05, | |
| "loss": 1.6513, | |
| "num_tokens": 171960.0, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.17687074829931973, | |
| "grad_norm": 20.970462799072266, | |
| "learning_rate": 1.7005847953216375e-05, | |
| "loss": 1.5231, | |
| "num_tokens": 173095.0, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.17800453514739228, | |
| "grad_norm": 22.09793472290039, | |
| "learning_rate": 1.6982456140350878e-05, | |
| "loss": 1.5093, | |
| "num_tokens": 174221.0, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.17913832199546487, | |
| "grad_norm": 21.099365234375, | |
| "learning_rate": 1.695906432748538e-05, | |
| "loss": 1.5163, | |
| "num_tokens": 175405.0, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.18027210884353742, | |
| "grad_norm": 24.84507179260254, | |
| "learning_rate": 1.6935672514619886e-05, | |
| "loss": 1.4498, | |
| "num_tokens": 176445.0, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.18140589569160998, | |
| "grad_norm": 20.663394927978516, | |
| "learning_rate": 1.6912280701754385e-05, | |
| "loss": 1.4815, | |
| "num_tokens": 177648.0, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.18253968253968253, | |
| "grad_norm": 20.654090881347656, | |
| "learning_rate": 1.688888888888889e-05, | |
| "loss": 1.534, | |
| "num_tokens": 178831.0, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.1836734693877551, | |
| "grad_norm": 21.591787338256836, | |
| "learning_rate": 1.6865497076023394e-05, | |
| "loss": 1.4529, | |
| "num_tokens": 179986.0, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.18480725623582767, | |
| "grad_norm": 21.668750762939453, | |
| "learning_rate": 1.6842105263157896e-05, | |
| "loss": 1.5613, | |
| "num_tokens": 181201.0, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.18594104308390022, | |
| "grad_norm": 23.887989044189453, | |
| "learning_rate": 1.68187134502924e-05, | |
| "loss": 1.654, | |
| "num_tokens": 182310.0, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.1870748299319728, | |
| "grad_norm": 21.20265007019043, | |
| "learning_rate": 1.67953216374269e-05, | |
| "loss": 1.561, | |
| "num_tokens": 183432.0, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.18820861678004536, | |
| "grad_norm": 22.128807067871094, | |
| "learning_rate": 1.6771929824561408e-05, | |
| "loss": 1.6709, | |
| "num_tokens": 184597.0, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.1893424036281179, | |
| "grad_norm": 24.24042320251465, | |
| "learning_rate": 1.6748538011695907e-05, | |
| "loss": 1.5693, | |
| "num_tokens": 185668.0, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.19047619047619047, | |
| "grad_norm": 21.286191940307617, | |
| "learning_rate": 1.672514619883041e-05, | |
| "loss": 1.4641, | |
| "num_tokens": 186778.0, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.19160997732426305, | |
| "grad_norm": 21.220508575439453, | |
| "learning_rate": 1.6701754385964915e-05, | |
| "loss": 1.6059, | |
| "num_tokens": 187963.0, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.1927437641723356, | |
| "grad_norm": 20.513689041137695, | |
| "learning_rate": 1.6678362573099414e-05, | |
| "loss": 1.496, | |
| "num_tokens": 189061.0, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.19387755102040816, | |
| "grad_norm": 22.597923278808594, | |
| "learning_rate": 1.665497076023392e-05, | |
| "loss": 1.5342, | |
| "num_tokens": 190184.0, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.19501133786848074, | |
| "grad_norm": 22.039106369018555, | |
| "learning_rate": 1.6631578947368423e-05, | |
| "loss": 1.4233, | |
| "num_tokens": 191228.0, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.1961451247165533, | |
| "grad_norm": 21.834428787231445, | |
| "learning_rate": 1.6608187134502926e-05, | |
| "loss": 1.461, | |
| "num_tokens": 192388.0, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.19727891156462585, | |
| "grad_norm": 20.996660232543945, | |
| "learning_rate": 1.6584795321637428e-05, | |
| "loss": 1.5707, | |
| "num_tokens": 193528.0, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.1984126984126984, | |
| "grad_norm": 21.746488571166992, | |
| "learning_rate": 1.656140350877193e-05, | |
| "loss": 1.4698, | |
| "num_tokens": 194550.0, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.19954648526077098, | |
| "grad_norm": 22.856365203857422, | |
| "learning_rate": 1.6538011695906437e-05, | |
| "loss": 1.4731, | |
| "num_tokens": 195677.0, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.20068027210884354, | |
| "grad_norm": 21.368350982666016, | |
| "learning_rate": 1.6514619883040936e-05, | |
| "loss": 1.481, | |
| "num_tokens": 196912.0, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.2018140589569161, | |
| "grad_norm": 20.245771408081055, | |
| "learning_rate": 1.649122807017544e-05, | |
| "loss": 1.4977, | |
| "num_tokens": 198043.0, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.20294784580498867, | |
| "grad_norm": 25.99681282043457, | |
| "learning_rate": 1.6467836257309944e-05, | |
| "loss": 1.7099, | |
| "num_tokens": 199157.0, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.20408163265306123, | |
| "grad_norm": 22.558382034301758, | |
| "learning_rate": 1.6444444444444444e-05, | |
| "loss": 1.5767, | |
| "num_tokens": 200255.0, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.20521541950113378, | |
| "grad_norm": 22.796449661254883, | |
| "learning_rate": 1.642105263157895e-05, | |
| "loss": 1.4589, | |
| "num_tokens": 201297.0, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.20634920634920634, | |
| "grad_norm": 20.285388946533203, | |
| "learning_rate": 1.6397660818713452e-05, | |
| "loss": 1.4589, | |
| "num_tokens": 202492.0, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.20748299319727892, | |
| "grad_norm": 20.92437744140625, | |
| "learning_rate": 1.6374269005847955e-05, | |
| "loss": 1.5125, | |
| "num_tokens": 203650.0, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.20861678004535147, | |
| "grad_norm": 22.35742950439453, | |
| "learning_rate": 1.6350877192982457e-05, | |
| "loss": 1.442, | |
| "num_tokens": 204723.0, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.20975056689342403, | |
| "grad_norm": 24.2554931640625, | |
| "learning_rate": 1.632748538011696e-05, | |
| "loss": 1.4875, | |
| "num_tokens": 205792.0, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.2108843537414966, | |
| "grad_norm": 23.882658004760742, | |
| "learning_rate": 1.6304093567251466e-05, | |
| "loss": 1.6708, | |
| "num_tokens": 206878.0, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.21201814058956917, | |
| "grad_norm": 22.755496978759766, | |
| "learning_rate": 1.6280701754385965e-05, | |
| "loss": 1.5278, | |
| "num_tokens": 208071.0, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.21315192743764172, | |
| "grad_norm": 22.447935104370117, | |
| "learning_rate": 1.625730994152047e-05, | |
| "loss": 1.4872, | |
| "num_tokens": 209227.0, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.21428571428571427, | |
| "grad_norm": 22.08466339111328, | |
| "learning_rate": 1.6233918128654974e-05, | |
| "loss": 1.5934, | |
| "num_tokens": 210372.0, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.21541950113378686, | |
| "grad_norm": 21.285282135009766, | |
| "learning_rate": 1.6210526315789473e-05, | |
| "loss": 1.4961, | |
| "num_tokens": 211604.0, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.2165532879818594, | |
| "grad_norm": 21.292037963867188, | |
| "learning_rate": 1.618713450292398e-05, | |
| "loss": 1.503, | |
| "num_tokens": 212855.0, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.21768707482993196, | |
| "grad_norm": 21.15336036682129, | |
| "learning_rate": 1.616374269005848e-05, | |
| "loss": 1.4699, | |
| "num_tokens": 213956.0, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.21882086167800455, | |
| "grad_norm": 22.280893325805664, | |
| "learning_rate": 1.6140350877192984e-05, | |
| "loss": 1.6332, | |
| "num_tokens": 215120.0, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.2199546485260771, | |
| "grad_norm": 22.338592529296875, | |
| "learning_rate": 1.6116959064327486e-05, | |
| "loss": 1.4444, | |
| "num_tokens": 216206.0, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.22108843537414966, | |
| "grad_norm": 20.818641662597656, | |
| "learning_rate": 1.609356725146199e-05, | |
| "loss": 1.4245, | |
| "num_tokens": 217303.0, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 23.549814224243164, | |
| "learning_rate": 1.6070175438596495e-05, | |
| "loss": 1.5753, | |
| "num_tokens": 218327.0, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.2233560090702948, | |
| "grad_norm": 22.222686767578125, | |
| "learning_rate": 1.6046783625730994e-05, | |
| "loss": 1.3668, | |
| "num_tokens": 219485.0, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.22448979591836735, | |
| "grad_norm": 19.54814910888672, | |
| "learning_rate": 1.60233918128655e-05, | |
| "loss": 1.6546, | |
| "num_tokens": 220708.0, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.2256235827664399, | |
| "grad_norm": 20.543609619140625, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 1.5096, | |
| "num_tokens": 221781.0, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.22675736961451248, | |
| "grad_norm": 21.107006072998047, | |
| "learning_rate": 1.5976608187134505e-05, | |
| "loss": 1.5088, | |
| "num_tokens": 222829.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.22789115646258504, | |
| "grad_norm": 23.751380920410156, | |
| "learning_rate": 1.5953216374269008e-05, | |
| "loss": 1.4793, | |
| "num_tokens": 224071.0, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.2290249433106576, | |
| "grad_norm": 22.424638748168945, | |
| "learning_rate": 1.592982456140351e-05, | |
| "loss": 1.6096, | |
| "num_tokens": 225179.0, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.23015873015873015, | |
| "grad_norm": 22.971948623657227, | |
| "learning_rate": 1.5906432748538013e-05, | |
| "loss": 1.4364, | |
| "num_tokens": 226250.0, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.23129251700680273, | |
| "grad_norm": 21.397802352905273, | |
| "learning_rate": 1.5883040935672516e-05, | |
| "loss": 1.5396, | |
| "num_tokens": 227380.0, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.23242630385487528, | |
| "grad_norm": 46.37774658203125, | |
| "learning_rate": 1.5859649122807018e-05, | |
| "loss": 1.43, | |
| "num_tokens": 228565.0, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.23356009070294784, | |
| "grad_norm": 24.68637466430664, | |
| "learning_rate": 1.583625730994152e-05, | |
| "loss": 1.683, | |
| "num_tokens": 229639.0, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.23469387755102042, | |
| "grad_norm": 21.283252716064453, | |
| "learning_rate": 1.5812865497076023e-05, | |
| "loss": 1.5629, | |
| "num_tokens": 230753.0, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.23582766439909297, | |
| "grad_norm": 22.06300926208496, | |
| "learning_rate": 1.578947368421053e-05, | |
| "loss": 1.5683, | |
| "num_tokens": 231866.0, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.23696145124716553, | |
| "grad_norm": 21.9282283782959, | |
| "learning_rate": 1.5766081871345032e-05, | |
| "loss": 1.5273, | |
| "num_tokens": 233001.0, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.23809523809523808, | |
| "grad_norm": 21.07894515991211, | |
| "learning_rate": 1.5742690058479534e-05, | |
| "loss": 1.5438, | |
| "num_tokens": 234113.0, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.23922902494331066, | |
| "grad_norm": 19.786283493041992, | |
| "learning_rate": 1.5719298245614037e-05, | |
| "loss": 1.3995, | |
| "num_tokens": 235294.0, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.24036281179138322, | |
| "grad_norm": 23.547483444213867, | |
| "learning_rate": 1.569590643274854e-05, | |
| "loss": 1.4131, | |
| "num_tokens": 236392.0, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.24149659863945577, | |
| "grad_norm": 20.67225456237793, | |
| "learning_rate": 1.5672514619883042e-05, | |
| "loss": 1.4733, | |
| "num_tokens": 237522.0, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.24263038548752835, | |
| "grad_norm": 22.142709732055664, | |
| "learning_rate": 1.5649122807017545e-05, | |
| "loss": 1.5663, | |
| "num_tokens": 238635.0, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.2437641723356009, | |
| "grad_norm": 23.121814727783203, | |
| "learning_rate": 1.5625730994152047e-05, | |
| "loss": 1.3922, | |
| "num_tokens": 239690.0, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.24489795918367346, | |
| "grad_norm": 25.34332847595215, | |
| "learning_rate": 1.560233918128655e-05, | |
| "loss": 1.529, | |
| "num_tokens": 240712.0, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.24603174603174602, | |
| "grad_norm": 20.340042114257812, | |
| "learning_rate": 1.5578947368421052e-05, | |
| "loss": 1.4748, | |
| "num_tokens": 241899.0, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.2471655328798186, | |
| "grad_norm": 21.851383209228516, | |
| "learning_rate": 1.555555555555556e-05, | |
| "loss": 1.537, | |
| "num_tokens": 242975.0, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.24829931972789115, | |
| "grad_norm": 22.453187942504883, | |
| "learning_rate": 1.553216374269006e-05, | |
| "loss": 1.5451, | |
| "num_tokens": 244129.0, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.2494331065759637, | |
| "grad_norm": 22.221010208129883, | |
| "learning_rate": 1.5508771929824563e-05, | |
| "loss": 1.5423, | |
| "num_tokens": 245290.0, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.25056689342403626, | |
| "grad_norm": 22.524154663085938, | |
| "learning_rate": 1.5485380116959066e-05, | |
| "loss": 1.6421, | |
| "num_tokens": 246386.0, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.25170068027210885, | |
| "grad_norm": 21.442996978759766, | |
| "learning_rate": 1.546198830409357e-05, | |
| "loss": 1.4855, | |
| "num_tokens": 247440.0, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.2528344671201814, | |
| "grad_norm": 24.744312286376953, | |
| "learning_rate": 1.543859649122807e-05, | |
| "loss": 1.3662, | |
| "num_tokens": 248514.0, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.25396825396825395, | |
| "grad_norm": 23.523515701293945, | |
| "learning_rate": 1.5415204678362574e-05, | |
| "loss": 1.4826, | |
| "num_tokens": 249642.0, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.25510204081632654, | |
| "grad_norm": 23.982515335083008, | |
| "learning_rate": 1.5391812865497076e-05, | |
| "loss": 1.3806, | |
| "num_tokens": 250712.0, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.2562358276643991, | |
| "grad_norm": 22.8784236907959, | |
| "learning_rate": 1.536842105263158e-05, | |
| "loss": 1.4332, | |
| "num_tokens": 251897.0, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.25736961451247165, | |
| "grad_norm": 22.019994735717773, | |
| "learning_rate": 1.534502923976608e-05, | |
| "loss": 1.418, | |
| "num_tokens": 253031.0, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.2585034013605442, | |
| "grad_norm": 22.86067008972168, | |
| "learning_rate": 1.5321637426900587e-05, | |
| "loss": 1.6491, | |
| "num_tokens": 254172.0, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.25963718820861675, | |
| "grad_norm": 22.04460334777832, | |
| "learning_rate": 1.529824561403509e-05, | |
| "loss": 1.373, | |
| "num_tokens": 255350.0, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.26077097505668934, | |
| "grad_norm": 21.948341369628906, | |
| "learning_rate": 1.5274853801169593e-05, | |
| "loss": 1.441, | |
| "num_tokens": 256554.0, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.2619047619047619, | |
| "grad_norm": 22.99373435974121, | |
| "learning_rate": 1.5251461988304095e-05, | |
| "loss": 1.5293, | |
| "num_tokens": 257729.0, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.26303854875283444, | |
| "grad_norm": 25.595821380615234, | |
| "learning_rate": 1.5228070175438598e-05, | |
| "loss": 1.4453, | |
| "num_tokens": 258806.0, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.264172335600907, | |
| "grad_norm": 22.947647094726562, | |
| "learning_rate": 1.52046783625731e-05, | |
| "loss": 1.5168, | |
| "num_tokens": 259874.0, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.2653061224489796, | |
| "grad_norm": 25.86823081970215, | |
| "learning_rate": 1.5181286549707603e-05, | |
| "loss": 1.5426, | |
| "num_tokens": 260959.0, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.26643990929705214, | |
| "grad_norm": 20.40091896057129, | |
| "learning_rate": 1.5157894736842107e-05, | |
| "loss": 1.7117, | |
| "num_tokens": 262189.0, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.2675736961451247, | |
| "grad_norm": 21.5074462890625, | |
| "learning_rate": 1.5134502923976608e-05, | |
| "loss": 1.4967, | |
| "num_tokens": 263381.0, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.2687074829931973, | |
| "grad_norm": 20.92792320251465, | |
| "learning_rate": 1.5111111111111112e-05, | |
| "loss": 1.5214, | |
| "num_tokens": 264514.0, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.2698412698412698, | |
| "grad_norm": 22.8386287689209, | |
| "learning_rate": 1.5087719298245615e-05, | |
| "loss": 1.5155, | |
| "num_tokens": 265629.0, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.2709750566893424, | |
| "grad_norm": 21.064926147460938, | |
| "learning_rate": 1.5064327485380119e-05, | |
| "loss": 1.44, | |
| "num_tokens": 266750.0, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.272108843537415, | |
| "grad_norm": 22.04184913635254, | |
| "learning_rate": 1.504093567251462e-05, | |
| "loss": 1.5298, | |
| "num_tokens": 267781.0, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.2732426303854875, | |
| "grad_norm": 24.02085304260254, | |
| "learning_rate": 1.5017543859649124e-05, | |
| "loss": 1.425, | |
| "num_tokens": 268921.0, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.2743764172335601, | |
| "grad_norm": 22.363149642944336, | |
| "learning_rate": 1.4994152046783627e-05, | |
| "loss": 1.4996, | |
| "num_tokens": 269985.0, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.2755102040816326, | |
| "grad_norm": 23.14397430419922, | |
| "learning_rate": 1.497076023391813e-05, | |
| "loss": 1.5728, | |
| "num_tokens": 271085.0, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.2766439909297052, | |
| "grad_norm": 21.870580673217773, | |
| "learning_rate": 1.4947368421052632e-05, | |
| "loss": 1.5506, | |
| "num_tokens": 272207.0, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.2777777777777778, | |
| "grad_norm": 23.548410415649414, | |
| "learning_rate": 1.4923976608187136e-05, | |
| "loss": 1.4119, | |
| "num_tokens": 273195.0, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.2789115646258503, | |
| "grad_norm": 22.918581008911133, | |
| "learning_rate": 1.4900584795321637e-05, | |
| "loss": 1.593, | |
| "num_tokens": 274395.0, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.2800453514739229, | |
| "grad_norm": 24.52320671081543, | |
| "learning_rate": 1.4877192982456141e-05, | |
| "loss": 1.592, | |
| "num_tokens": 275413.0, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.2811791383219955, | |
| "grad_norm": 25.395458221435547, | |
| "learning_rate": 1.4853801169590644e-05, | |
| "loss": 1.3833, | |
| "num_tokens": 276425.0, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.282312925170068, | |
| "grad_norm": 20.522045135498047, | |
| "learning_rate": 1.4830409356725148e-05, | |
| "loss": 1.4719, | |
| "num_tokens": 277674.0, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.2834467120181406, | |
| "grad_norm": 22.4309024810791, | |
| "learning_rate": 1.4807017543859649e-05, | |
| "loss": 1.3931, | |
| "num_tokens": 278761.0, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.28458049886621317, | |
| "grad_norm": 23.054508209228516, | |
| "learning_rate": 1.4783625730994153e-05, | |
| "loss": 1.6357, | |
| "num_tokens": 279900.0, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 22.669353485107422, | |
| "learning_rate": 1.4760233918128658e-05, | |
| "loss": 1.5954, | |
| "num_tokens": 281152.0, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.2868480725623583, | |
| "grad_norm": 24.546659469604492, | |
| "learning_rate": 1.4736842105263159e-05, | |
| "loss": 1.5536, | |
| "num_tokens": 282236.0, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.28798185941043086, | |
| "grad_norm": 22.03496551513672, | |
| "learning_rate": 1.4713450292397661e-05, | |
| "loss": 1.4711, | |
| "num_tokens": 283294.0, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.2891156462585034, | |
| "grad_norm": 23.8269100189209, | |
| "learning_rate": 1.4690058479532165e-05, | |
| "loss": 1.5215, | |
| "num_tokens": 284345.0, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.29024943310657597, | |
| "grad_norm": 21.356388092041016, | |
| "learning_rate": 1.4666666666666666e-05, | |
| "loss": 1.6076, | |
| "num_tokens": 285419.0, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.29138321995464855, | |
| "grad_norm": 20.89931869506836, | |
| "learning_rate": 1.464327485380117e-05, | |
| "loss": 1.4129, | |
| "num_tokens": 286535.0, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.2925170068027211, | |
| "grad_norm": 21.1955623626709, | |
| "learning_rate": 1.4619883040935675e-05, | |
| "loss": 1.469, | |
| "num_tokens": 287631.0, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.29365079365079366, | |
| "grad_norm": 19.76694107055664, | |
| "learning_rate": 1.4596491228070177e-05, | |
| "loss": 1.6142, | |
| "num_tokens": 288754.0, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.2947845804988662, | |
| "grad_norm": 21.243698120117188, | |
| "learning_rate": 1.4573099415204678e-05, | |
| "loss": 1.5581, | |
| "num_tokens": 289958.0, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.29591836734693877, | |
| "grad_norm": 22.617021560668945, | |
| "learning_rate": 1.4549707602339183e-05, | |
| "loss": 1.4423, | |
| "num_tokens": 291137.0, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.29705215419501135, | |
| "grad_norm": 20.3275146484375, | |
| "learning_rate": 1.4526315789473687e-05, | |
| "loss": 1.5287, | |
| "num_tokens": 292312.0, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.2981859410430839, | |
| "grad_norm": 26.21241569519043, | |
| "learning_rate": 1.4502923976608188e-05, | |
| "loss": 1.4952, | |
| "num_tokens": 293355.0, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.29931972789115646, | |
| "grad_norm": 23.586990356445312, | |
| "learning_rate": 1.447953216374269e-05, | |
| "loss": 1.4412, | |
| "num_tokens": 294480.0, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.30045351473922904, | |
| "grad_norm": 23.049999237060547, | |
| "learning_rate": 1.4456140350877195e-05, | |
| "loss": 1.5597, | |
| "num_tokens": 295543.0, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.30158730158730157, | |
| "grad_norm": 22.225637435913086, | |
| "learning_rate": 1.4432748538011695e-05, | |
| "loss": 1.5182, | |
| "num_tokens": 296682.0, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.30272108843537415, | |
| "grad_norm": 22.19669532775879, | |
| "learning_rate": 1.44093567251462e-05, | |
| "loss": 1.5603, | |
| "num_tokens": 297762.0, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.30385487528344673, | |
| "grad_norm": 20.865692138671875, | |
| "learning_rate": 1.4385964912280704e-05, | |
| "loss": 1.4716, | |
| "num_tokens": 298916.0, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.30498866213151926, | |
| "grad_norm": 24.449691772460938, | |
| "learning_rate": 1.4362573099415207e-05, | |
| "loss": 1.3971, | |
| "num_tokens": 299936.0, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.30612244897959184, | |
| "grad_norm": 21.59602165222168, | |
| "learning_rate": 1.4339181286549707e-05, | |
| "loss": 1.5539, | |
| "num_tokens": 301074.0, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.3072562358276644, | |
| "grad_norm": 23.660263061523438, | |
| "learning_rate": 1.4315789473684212e-05, | |
| "loss": 1.4892, | |
| "num_tokens": 302125.0, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.30839002267573695, | |
| "grad_norm": 22.853479385375977, | |
| "learning_rate": 1.4292397660818716e-05, | |
| "loss": 1.4667, | |
| "num_tokens": 303263.0, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.30952380952380953, | |
| "grad_norm": 23.96630859375, | |
| "learning_rate": 1.4269005847953217e-05, | |
| "loss": 1.4877, | |
| "num_tokens": 304463.0, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.31065759637188206, | |
| "grad_norm": 21.596799850463867, | |
| "learning_rate": 1.4245614035087721e-05, | |
| "loss": 1.4767, | |
| "num_tokens": 305660.0, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.31179138321995464, | |
| "grad_norm": 23.900022506713867, | |
| "learning_rate": 1.4222222222222224e-05, | |
| "loss": 1.4882, | |
| "num_tokens": 306684.0, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.3129251700680272, | |
| "grad_norm": 21.063495635986328, | |
| "learning_rate": 1.4198830409356725e-05, | |
| "loss": 1.4495, | |
| "num_tokens": 307813.0, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.31405895691609975, | |
| "grad_norm": 23.05027198791504, | |
| "learning_rate": 1.4175438596491229e-05, | |
| "loss": 1.4677, | |
| "num_tokens": 308876.0, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.31519274376417233, | |
| "grad_norm": 20.81597900390625, | |
| "learning_rate": 1.4152046783625733e-05, | |
| "loss": 1.4085, | |
| "num_tokens": 310012.0, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.3163265306122449, | |
| "grad_norm": 23.965967178344727, | |
| "learning_rate": 1.4128654970760236e-05, | |
| "loss": 1.4615, | |
| "num_tokens": 311070.0, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.31746031746031744, | |
| "grad_norm": 22.640148162841797, | |
| "learning_rate": 1.4105263157894738e-05, | |
| "loss": 1.3822, | |
| "num_tokens": 312154.0, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.31859410430839, | |
| "grad_norm": 22.867446899414062, | |
| "learning_rate": 1.408187134502924e-05, | |
| "loss": 1.6723, | |
| "num_tokens": 313253.0, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.3197278911564626, | |
| "grad_norm": 22.407011032104492, | |
| "learning_rate": 1.4058479532163745e-05, | |
| "loss": 1.4687, | |
| "num_tokens": 314325.0, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.32086167800453513, | |
| "grad_norm": 21.54814338684082, | |
| "learning_rate": 1.4035087719298246e-05, | |
| "loss": 1.5139, | |
| "num_tokens": 315399.0, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.3219954648526077, | |
| "grad_norm": 21.578617095947266, | |
| "learning_rate": 1.401169590643275e-05, | |
| "loss": 1.3751, | |
| "num_tokens": 316543.0, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.3231292517006803, | |
| "grad_norm": 21.266693115234375, | |
| "learning_rate": 1.3988304093567253e-05, | |
| "loss": 1.4847, | |
| "num_tokens": 317658.0, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.3242630385487528, | |
| "grad_norm": 23.686180114746094, | |
| "learning_rate": 1.3964912280701755e-05, | |
| "loss": 1.491, | |
| "num_tokens": 318773.0, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.3253968253968254, | |
| "grad_norm": 22.65009117126465, | |
| "learning_rate": 1.3941520467836258e-05, | |
| "loss": 1.5428, | |
| "num_tokens": 319995.0, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.32653061224489793, | |
| "grad_norm": 20.74267578125, | |
| "learning_rate": 1.3918128654970762e-05, | |
| "loss": 1.3828, | |
| "num_tokens": 321078.0, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.3276643990929705, | |
| "grad_norm": 22.6479434967041, | |
| "learning_rate": 1.3894736842105265e-05, | |
| "loss": 1.4634, | |
| "num_tokens": 322140.0, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.3287981859410431, | |
| "grad_norm": 23.637678146362305, | |
| "learning_rate": 1.3871345029239767e-05, | |
| "loss": 1.4539, | |
| "num_tokens": 323326.0, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.3299319727891156, | |
| "grad_norm": 20.191329956054688, | |
| "learning_rate": 1.384795321637427e-05, | |
| "loss": 1.3562, | |
| "num_tokens": 324391.0, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.3310657596371882, | |
| "grad_norm": 24.18254852294922, | |
| "learning_rate": 1.3824561403508774e-05, | |
| "loss": 1.407, | |
| "num_tokens": 325423.0, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.3321995464852608, | |
| "grad_norm": 22.894956588745117, | |
| "learning_rate": 1.3801169590643275e-05, | |
| "loss": 1.4151, | |
| "num_tokens": 326471.0, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 25.60346221923828, | |
| "learning_rate": 1.377777777777778e-05, | |
| "loss": 1.5366, | |
| "num_tokens": 327570.0, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.3344671201814059, | |
| "grad_norm": 23.408321380615234, | |
| "learning_rate": 1.3754385964912282e-05, | |
| "loss": 1.5041, | |
| "num_tokens": 328672.0, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.3356009070294785, | |
| "grad_norm": 23.32335662841797, | |
| "learning_rate": 1.3730994152046784e-05, | |
| "loss": 1.5033, | |
| "num_tokens": 329696.0, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.336734693877551, | |
| "grad_norm": 20.227418899536133, | |
| "learning_rate": 1.3707602339181287e-05, | |
| "loss": 1.3922, | |
| "num_tokens": 330810.0, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.3378684807256236, | |
| "grad_norm": 20.30182647705078, | |
| "learning_rate": 1.3684210526315791e-05, | |
| "loss": 1.4642, | |
| "num_tokens": 331998.0, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.33900226757369617, | |
| "grad_norm": 22.34268569946289, | |
| "learning_rate": 1.3660818713450294e-05, | |
| "loss": 1.3948, | |
| "num_tokens": 333011.0, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.3401360544217687, | |
| "grad_norm": 21.91162872314453, | |
| "learning_rate": 1.3637426900584796e-05, | |
| "loss": 1.4822, | |
| "num_tokens": 334104.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.3412698412698413, | |
| "grad_norm": 22.55190658569336, | |
| "learning_rate": 1.3614035087719299e-05, | |
| "loss": 1.6097, | |
| "num_tokens": 335283.0, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.3424036281179138, | |
| "grad_norm": 21.861995697021484, | |
| "learning_rate": 1.3590643274853803e-05, | |
| "loss": 1.588, | |
| "num_tokens": 336363.0, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.3435374149659864, | |
| "grad_norm": 21.788677215576172, | |
| "learning_rate": 1.3567251461988304e-05, | |
| "loss": 1.582, | |
| "num_tokens": 337576.0, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.34467120181405897, | |
| "grad_norm": 20.740257263183594, | |
| "learning_rate": 1.3543859649122808e-05, | |
| "loss": 1.3907, | |
| "num_tokens": 338641.0, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.3458049886621315, | |
| "grad_norm": 21.97249412536621, | |
| "learning_rate": 1.3520467836257311e-05, | |
| "loss": 1.5256, | |
| "num_tokens": 339795.0, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.3469387755102041, | |
| "grad_norm": 21.80385971069336, | |
| "learning_rate": 1.3497076023391814e-05, | |
| "loss": 1.4808, | |
| "num_tokens": 340889.0, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.34807256235827666, | |
| "grad_norm": 22.424528121948242, | |
| "learning_rate": 1.3473684210526316e-05, | |
| "loss": 1.575, | |
| "num_tokens": 342007.0, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.3492063492063492, | |
| "grad_norm": 22.06599998474121, | |
| "learning_rate": 1.345029239766082e-05, | |
| "loss": 1.307, | |
| "num_tokens": 343151.0, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.35034013605442177, | |
| "grad_norm": 24.721797943115234, | |
| "learning_rate": 1.3426900584795323e-05, | |
| "loss": 1.5736, | |
| "num_tokens": 344208.0, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.35147392290249435, | |
| "grad_norm": 22.60153579711914, | |
| "learning_rate": 1.3403508771929826e-05, | |
| "loss": 1.5604, | |
| "num_tokens": 345374.0, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.3526077097505669, | |
| "grad_norm": 21.573318481445312, | |
| "learning_rate": 1.3380116959064328e-05, | |
| "loss": 1.524, | |
| "num_tokens": 346608.0, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.35374149659863946, | |
| "grad_norm": 25.30203628540039, | |
| "learning_rate": 1.3356725146198832e-05, | |
| "loss": 1.6416, | |
| "num_tokens": 347714.0, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.35487528344671204, | |
| "grad_norm": 22.905746459960938, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 1.5275, | |
| "num_tokens": 348930.0, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.35600907029478457, | |
| "grad_norm": 23.923063278198242, | |
| "learning_rate": 1.3309941520467838e-05, | |
| "loss": 1.5784, | |
| "num_tokens": 349999.0, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.35714285714285715, | |
| "grad_norm": 22.065505981445312, | |
| "learning_rate": 1.328654970760234e-05, | |
| "loss": 1.5701, | |
| "num_tokens": 351123.0, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.35827664399092973, | |
| "grad_norm": 22.438655853271484, | |
| "learning_rate": 1.3263157894736843e-05, | |
| "loss": 1.5035, | |
| "num_tokens": 352245.0, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.35941043083900226, | |
| "grad_norm": 21.77728271484375, | |
| "learning_rate": 1.3239766081871345e-05, | |
| "loss": 1.4594, | |
| "num_tokens": 353380.0, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.36054421768707484, | |
| "grad_norm": 26.78565788269043, | |
| "learning_rate": 1.321637426900585e-05, | |
| "loss": 1.6554, | |
| "num_tokens": 354375.0, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.36167800453514737, | |
| "grad_norm": 22.74375343322754, | |
| "learning_rate": 1.3192982456140354e-05, | |
| "loss": 1.4583, | |
| "num_tokens": 355502.0, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.36281179138321995, | |
| "grad_norm": 22.99059295654297, | |
| "learning_rate": 1.3169590643274855e-05, | |
| "loss": 1.4056, | |
| "num_tokens": 356615.0, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.36394557823129253, | |
| "grad_norm": 19.973115921020508, | |
| "learning_rate": 1.3146198830409357e-05, | |
| "loss": 1.4359, | |
| "num_tokens": 357688.0, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.36507936507936506, | |
| "grad_norm": 20.598600387573242, | |
| "learning_rate": 1.3122807017543862e-05, | |
| "loss": 1.3685, | |
| "num_tokens": 358832.0, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.36621315192743764, | |
| "grad_norm": 23.178176879882812, | |
| "learning_rate": 1.3099415204678362e-05, | |
| "loss": 1.4443, | |
| "num_tokens": 359904.0, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.3673469387755102, | |
| "grad_norm": 23.088056564331055, | |
| "learning_rate": 1.3076023391812867e-05, | |
| "loss": 1.3505, | |
| "num_tokens": 361014.0, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.36848072562358275, | |
| "grad_norm": 22.677352905273438, | |
| "learning_rate": 1.305263157894737e-05, | |
| "loss": 1.432, | |
| "num_tokens": 362163.0, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.36961451247165533, | |
| "grad_norm": 22.128244400024414, | |
| "learning_rate": 1.3029239766081872e-05, | |
| "loss": 1.4475, | |
| "num_tokens": 363222.0, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.3707482993197279, | |
| "grad_norm": 20.658218383789062, | |
| "learning_rate": 1.3005847953216374e-05, | |
| "loss": 1.4307, | |
| "num_tokens": 364332.0, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.37188208616780044, | |
| "grad_norm": 24.494672775268555, | |
| "learning_rate": 1.2982456140350879e-05, | |
| "loss": 1.5444, | |
| "num_tokens": 365433.0, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.373015873015873, | |
| "grad_norm": 24.10431671142578, | |
| "learning_rate": 1.2959064327485383e-05, | |
| "loss": 1.4431, | |
| "num_tokens": 366487.0, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.3741496598639456, | |
| "grad_norm": 20.076684951782227, | |
| "learning_rate": 1.2935672514619884e-05, | |
| "loss": 1.4626, | |
| "num_tokens": 367618.0, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.37528344671201813, | |
| "grad_norm": 22.610557556152344, | |
| "learning_rate": 1.2912280701754386e-05, | |
| "loss": 1.4713, | |
| "num_tokens": 368738.0, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.3764172335600907, | |
| "grad_norm": 25.21763801574707, | |
| "learning_rate": 1.288888888888889e-05, | |
| "loss": 1.5092, | |
| "num_tokens": 369720.0, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.37755102040816324, | |
| "grad_norm": 20.842025756835938, | |
| "learning_rate": 1.2865497076023392e-05, | |
| "loss": 1.4863, | |
| "num_tokens": 370885.0, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.3786848072562358, | |
| "grad_norm": 21.969791412353516, | |
| "learning_rate": 1.2842105263157896e-05, | |
| "loss": 1.3838, | |
| "num_tokens": 371958.0, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.3798185941043084, | |
| "grad_norm": 23.09484100341797, | |
| "learning_rate": 1.28187134502924e-05, | |
| "loss": 1.6291, | |
| "num_tokens": 373085.0, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.38095238095238093, | |
| "grad_norm": 20.638565063476562, | |
| "learning_rate": 1.2795321637426901e-05, | |
| "loss": 1.4701, | |
| "num_tokens": 374210.0, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.3820861678004535, | |
| "grad_norm": 21.330707550048828, | |
| "learning_rate": 1.2771929824561404e-05, | |
| "loss": 1.4856, | |
| "num_tokens": 375310.0, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.3832199546485261, | |
| "grad_norm": 22.242534637451172, | |
| "learning_rate": 1.2748538011695908e-05, | |
| "loss": 1.4001, | |
| "num_tokens": 376372.0, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.3843537414965986, | |
| "grad_norm": 22.377954483032227, | |
| "learning_rate": 1.2725146198830412e-05, | |
| "loss": 1.6322, | |
| "num_tokens": 377527.0, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.3854875283446712, | |
| "grad_norm": 22.19904327392578, | |
| "learning_rate": 1.2701754385964913e-05, | |
| "loss": 1.3738, | |
| "num_tokens": 378604.0, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.3866213151927438, | |
| "grad_norm": 21.403820037841797, | |
| "learning_rate": 1.2678362573099417e-05, | |
| "loss": 1.3584, | |
| "num_tokens": 379678.0, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.3877551020408163, | |
| "grad_norm": 21.75101089477539, | |
| "learning_rate": 1.265497076023392e-05, | |
| "loss": 1.4217, | |
| "num_tokens": 380729.0, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.3888888888888889, | |
| "grad_norm": 23.82843589782715, | |
| "learning_rate": 1.263157894736842e-05, | |
| "loss": 1.447, | |
| "num_tokens": 381876.0, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.3900226757369615, | |
| "grad_norm": 23.757530212402344, | |
| "learning_rate": 1.2608187134502925e-05, | |
| "loss": 1.4956, | |
| "num_tokens": 382962.0, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.391156462585034, | |
| "grad_norm": 23.748046875, | |
| "learning_rate": 1.258479532163743e-05, | |
| "loss": 1.5339, | |
| "num_tokens": 384096.0, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.3922902494331066, | |
| "grad_norm": 21.567293167114258, | |
| "learning_rate": 1.256140350877193e-05, | |
| "loss": 1.3744, | |
| "num_tokens": 385175.0, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.3934240362811791, | |
| "grad_norm": 22.490840911865234, | |
| "learning_rate": 1.2538011695906434e-05, | |
| "loss": 1.4251, | |
| "num_tokens": 386358.0, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.3945578231292517, | |
| "grad_norm": 21.747709274291992, | |
| "learning_rate": 1.2514619883040937e-05, | |
| "loss": 1.4746, | |
| "num_tokens": 387482.0, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.3956916099773243, | |
| "grad_norm": 19.943635940551758, | |
| "learning_rate": 1.2491228070175441e-05, | |
| "loss": 1.4278, | |
| "num_tokens": 388662.0, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.3968253968253968, | |
| "grad_norm": 22.539836883544922, | |
| "learning_rate": 1.2467836257309942e-05, | |
| "loss": 1.6124, | |
| "num_tokens": 389731.0, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.3979591836734694, | |
| "grad_norm": 21.41621208190918, | |
| "learning_rate": 1.2444444444444446e-05, | |
| "loss": 1.4527, | |
| "num_tokens": 390921.0, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.39909297052154197, | |
| "grad_norm": 21.026447296142578, | |
| "learning_rate": 1.2421052631578949e-05, | |
| "loss": 1.5012, | |
| "num_tokens": 392035.0, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.4002267573696145, | |
| "grad_norm": 25.59100914001465, | |
| "learning_rate": 1.239766081871345e-05, | |
| "loss": 1.5081, | |
| "num_tokens": 393093.0, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.4013605442176871, | |
| "grad_norm": 22.007505416870117, | |
| "learning_rate": 1.2374269005847954e-05, | |
| "loss": 1.5663, | |
| "num_tokens": 394239.0, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.40249433106575966, | |
| "grad_norm": 23.461524963378906, | |
| "learning_rate": 1.2350877192982458e-05, | |
| "loss": 1.4302, | |
| "num_tokens": 395291.0, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.4036281179138322, | |
| "grad_norm": 20.203760147094727, | |
| "learning_rate": 1.232748538011696e-05, | |
| "loss": 1.3541, | |
| "num_tokens": 396485.0, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.40476190476190477, | |
| "grad_norm": 19.759798049926758, | |
| "learning_rate": 1.2304093567251463e-05, | |
| "loss": 1.4233, | |
| "num_tokens": 397644.0, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.40589569160997735, | |
| "grad_norm": 22.704025268554688, | |
| "learning_rate": 1.2280701754385966e-05, | |
| "loss": 1.3426, | |
| "num_tokens": 398789.0, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.4070294784580499, | |
| "grad_norm": 21.620922088623047, | |
| "learning_rate": 1.225730994152047e-05, | |
| "loss": 1.5188, | |
| "num_tokens": 399878.0, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.40816326530612246, | |
| "grad_norm": 19.23693084716797, | |
| "learning_rate": 1.2233918128654971e-05, | |
| "loss": 1.3182, | |
| "num_tokens": 400987.0, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.409297052154195, | |
| "grad_norm": 21.466594696044922, | |
| "learning_rate": 1.2210526315789475e-05, | |
| "loss": 1.5335, | |
| "num_tokens": 402056.0, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.41043083900226757, | |
| "grad_norm": 21.63371467590332, | |
| "learning_rate": 1.2187134502923978e-05, | |
| "loss": 1.4244, | |
| "num_tokens": 403197.0, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.41156462585034015, | |
| "grad_norm": 20.83876609802246, | |
| "learning_rate": 1.216374269005848e-05, | |
| "loss": 1.6196, | |
| "num_tokens": 404342.0, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.4126984126984127, | |
| "grad_norm": 21.863889694213867, | |
| "learning_rate": 1.2140350877192983e-05, | |
| "loss": 1.5207, | |
| "num_tokens": 405485.0, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.41383219954648526, | |
| "grad_norm": 20.582901000976562, | |
| "learning_rate": 1.2116959064327487e-05, | |
| "loss": 1.4857, | |
| "num_tokens": 406682.0, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.41496598639455784, | |
| "grad_norm": 22.900217056274414, | |
| "learning_rate": 1.2093567251461988e-05, | |
| "loss": 1.453, | |
| "num_tokens": 407731.0, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.41609977324263037, | |
| "grad_norm": 24.468358993530273, | |
| "learning_rate": 1.2070175438596493e-05, | |
| "loss": 1.5036, | |
| "num_tokens": 408773.0, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.41723356009070295, | |
| "grad_norm": 22.844940185546875, | |
| "learning_rate": 1.2046783625730995e-05, | |
| "loss": 1.3458, | |
| "num_tokens": 409839.0, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.41836734693877553, | |
| "grad_norm": 23.0582332611084, | |
| "learning_rate": 1.20233918128655e-05, | |
| "loss": 1.4995, | |
| "num_tokens": 410856.0, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.41950113378684806, | |
| "grad_norm": 20.938154220581055, | |
| "learning_rate": 1.2e-05, | |
| "loss": 1.4394, | |
| "num_tokens": 411957.0, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.42063492063492064, | |
| "grad_norm": 21.037168502807617, | |
| "learning_rate": 1.1976608187134505e-05, | |
| "loss": 1.3752, | |
| "num_tokens": 413070.0, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.4217687074829932, | |
| "grad_norm": 23.35750961303711, | |
| "learning_rate": 1.1953216374269007e-05, | |
| "loss": 1.4844, | |
| "num_tokens": 414204.0, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.42290249433106575, | |
| "grad_norm": 20.994701385498047, | |
| "learning_rate": 1.192982456140351e-05, | |
| "loss": 1.4312, | |
| "num_tokens": 415467.0, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.42403628117913833, | |
| "grad_norm": 21.293333053588867, | |
| "learning_rate": 1.1906432748538012e-05, | |
| "loss": 1.4493, | |
| "num_tokens": 416638.0, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.42517006802721086, | |
| "grad_norm": 21.202180862426758, | |
| "learning_rate": 1.1883040935672517e-05, | |
| "loss": 1.4468, | |
| "num_tokens": 417771.0, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.42630385487528344, | |
| "grad_norm": 22.729934692382812, | |
| "learning_rate": 1.1859649122807017e-05, | |
| "loss": 1.5065, | |
| "num_tokens": 418899.0, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.427437641723356, | |
| "grad_norm": 20.654706954956055, | |
| "learning_rate": 1.1836257309941522e-05, | |
| "loss": 1.453, | |
| "num_tokens": 420087.0, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "grad_norm": 22.81949806213379, | |
| "learning_rate": 1.1812865497076024e-05, | |
| "loss": 1.4696, | |
| "num_tokens": 421175.0, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.42970521541950113, | |
| "grad_norm": 22.954544067382812, | |
| "learning_rate": 1.1789473684210527e-05, | |
| "loss": 1.4957, | |
| "num_tokens": 422301.0, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.4308390022675737, | |
| "grad_norm": 20.20991325378418, | |
| "learning_rate": 1.176608187134503e-05, | |
| "loss": 1.3547, | |
| "num_tokens": 423479.0, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.43197278911564624, | |
| "grad_norm": 21.98236846923828, | |
| "learning_rate": 1.1742690058479534e-05, | |
| "loss": 1.5336, | |
| "num_tokens": 424599.0, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.4331065759637188, | |
| "grad_norm": 22.076784133911133, | |
| "learning_rate": 1.1719298245614036e-05, | |
| "loss": 1.4168, | |
| "num_tokens": 425638.0, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.4342403628117914, | |
| "grad_norm": 20.801210403442383, | |
| "learning_rate": 1.1695906432748539e-05, | |
| "loss": 1.5112, | |
| "num_tokens": 426831.0, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.43537414965986393, | |
| "grad_norm": 21.711530685424805, | |
| "learning_rate": 1.1672514619883041e-05, | |
| "loss": 1.5072, | |
| "num_tokens": 428008.0, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.4365079365079365, | |
| "grad_norm": 20.649227142333984, | |
| "learning_rate": 1.1649122807017546e-05, | |
| "loss": 1.3347, | |
| "num_tokens": 429150.0, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.4376417233560091, | |
| "grad_norm": 24.207242965698242, | |
| "learning_rate": 1.1625730994152047e-05, | |
| "loss": 1.4249, | |
| "num_tokens": 430401.0, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.4387755102040816, | |
| "grad_norm": 22.359968185424805, | |
| "learning_rate": 1.160233918128655e-05, | |
| "loss": 1.4618, | |
| "num_tokens": 431572.0, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.4399092970521542, | |
| "grad_norm": 23.305208206176758, | |
| "learning_rate": 1.1578947368421053e-05, | |
| "loss": 1.6135, | |
| "num_tokens": 432748.0, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.4410430839002268, | |
| "grad_norm": 23.788869857788086, | |
| "learning_rate": 1.1555555555555556e-05, | |
| "loss": 1.4025, | |
| "num_tokens": 433673.0, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.4421768707482993, | |
| "grad_norm": 23.072362899780273, | |
| "learning_rate": 1.1532163742690059e-05, | |
| "loss": 1.4262, | |
| "num_tokens": 434803.0, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.4433106575963719, | |
| "grad_norm": 24.782913208007812, | |
| "learning_rate": 1.1508771929824563e-05, | |
| "loss": 1.5596, | |
| "num_tokens": 435895.0, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 22.748624801635742, | |
| "learning_rate": 1.1485380116959065e-05, | |
| "loss": 1.5855, | |
| "num_tokens": 436983.0, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.445578231292517, | |
| "grad_norm": 24.416845321655273, | |
| "learning_rate": 1.1461988304093568e-05, | |
| "loss": 1.5917, | |
| "num_tokens": 438073.0, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.4467120181405896, | |
| "grad_norm": 21.77821922302246, | |
| "learning_rate": 1.143859649122807e-05, | |
| "loss": 1.3617, | |
| "num_tokens": 439139.0, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.4478458049886621, | |
| "grad_norm": 22.38814926147461, | |
| "learning_rate": 1.1415204678362575e-05, | |
| "loss": 1.4778, | |
| "num_tokens": 440259.0, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.4489795918367347, | |
| "grad_norm": 21.56502342224121, | |
| "learning_rate": 1.1391812865497076e-05, | |
| "loss": 1.5352, | |
| "num_tokens": 441416.0, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.4501133786848073, | |
| "grad_norm": 20.565168380737305, | |
| "learning_rate": 1.136842105263158e-05, | |
| "loss": 1.5067, | |
| "num_tokens": 442549.0, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.4512471655328798, | |
| "grad_norm": 19.611265182495117, | |
| "learning_rate": 1.1345029239766083e-05, | |
| "loss": 1.4137, | |
| "num_tokens": 443711.0, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.4523809523809524, | |
| "grad_norm": 20.92315101623535, | |
| "learning_rate": 1.1321637426900585e-05, | |
| "loss": 1.4697, | |
| "num_tokens": 444840.0, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.45351473922902497, | |
| "grad_norm": 19.84340476989746, | |
| "learning_rate": 1.1298245614035088e-05, | |
| "loss": 1.5124, | |
| "num_tokens": 445920.0, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.4546485260770975, | |
| "grad_norm": 20.4117431640625, | |
| "learning_rate": 1.1274853801169592e-05, | |
| "loss": 1.5706, | |
| "num_tokens": 447140.0, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.4557823129251701, | |
| "grad_norm": 20.32288360595703, | |
| "learning_rate": 1.1251461988304096e-05, | |
| "loss": 1.3753, | |
| "num_tokens": 448247.0, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.45691609977324266, | |
| "grad_norm": 20.550880432128906, | |
| "learning_rate": 1.1228070175438597e-05, | |
| "loss": 1.3749, | |
| "num_tokens": 449442.0, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.4580498866213152, | |
| "grad_norm": 22.589542388916016, | |
| "learning_rate": 1.12046783625731e-05, | |
| "loss": 1.4055, | |
| "num_tokens": 450561.0, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.45918367346938777, | |
| "grad_norm": 20.559473037719727, | |
| "learning_rate": 1.1181286549707604e-05, | |
| "loss": 1.5184, | |
| "num_tokens": 451707.0, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.4603174603174603, | |
| "grad_norm": 19.94198226928711, | |
| "learning_rate": 1.1157894736842105e-05, | |
| "loss": 1.4678, | |
| "num_tokens": 452849.0, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.4614512471655329, | |
| "grad_norm": 21.226221084594727, | |
| "learning_rate": 1.1134502923976609e-05, | |
| "loss": 1.358, | |
| "num_tokens": 453908.0, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.46258503401360546, | |
| "grad_norm": 20.742549896240234, | |
| "learning_rate": 1.1111111111111113e-05, | |
| "loss": 1.3297, | |
| "num_tokens": 455072.0, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.463718820861678, | |
| "grad_norm": 20.73780059814453, | |
| "learning_rate": 1.1087719298245614e-05, | |
| "loss": 1.5569, | |
| "num_tokens": 456203.0, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.46485260770975056, | |
| "grad_norm": 21.602121353149414, | |
| "learning_rate": 1.1064327485380117e-05, | |
| "loss": 1.512, | |
| "num_tokens": 457306.0, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.46598639455782315, | |
| "grad_norm": 22.79952049255371, | |
| "learning_rate": 1.1040935672514621e-05, | |
| "loss": 1.5161, | |
| "num_tokens": 458435.0, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.4671201814058957, | |
| "grad_norm": 22.42172622680664, | |
| "learning_rate": 1.1017543859649125e-05, | |
| "loss": 1.2366, | |
| "num_tokens": 459465.0, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.46825396825396826, | |
| "grad_norm": 22.881622314453125, | |
| "learning_rate": 1.0994152046783626e-05, | |
| "loss": 1.5695, | |
| "num_tokens": 460630.0, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.46938775510204084, | |
| "grad_norm": 21.594409942626953, | |
| "learning_rate": 1.0970760233918129e-05, | |
| "loss": 1.446, | |
| "num_tokens": 461789.0, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.47052154195011336, | |
| "grad_norm": 23.080078125, | |
| "learning_rate": 1.0947368421052633e-05, | |
| "loss": 1.4408, | |
| "num_tokens": 462919.0, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.47165532879818595, | |
| "grad_norm": 22.40519142150879, | |
| "learning_rate": 1.0923976608187134e-05, | |
| "loss": 1.3175, | |
| "num_tokens": 463976.0, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.47278911564625853, | |
| "grad_norm": 23.810379028320312, | |
| "learning_rate": 1.0900584795321638e-05, | |
| "loss": 1.5058, | |
| "num_tokens": 465105.0, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.47392290249433106, | |
| "grad_norm": 21.764375686645508, | |
| "learning_rate": 1.0877192982456142e-05, | |
| "loss": 1.4464, | |
| "num_tokens": 466236.0, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.47505668934240364, | |
| "grad_norm": 21.064970016479492, | |
| "learning_rate": 1.0853801169590643e-05, | |
| "loss": 1.4898, | |
| "num_tokens": 467384.0, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.47619047619047616, | |
| "grad_norm": 21.655441284179688, | |
| "learning_rate": 1.0830409356725146e-05, | |
| "loss": 1.3508, | |
| "num_tokens": 468501.0, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.47732426303854875, | |
| "grad_norm": 23.890254974365234, | |
| "learning_rate": 1.080701754385965e-05, | |
| "loss": 1.597, | |
| "num_tokens": 469514.0, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.47845804988662133, | |
| "grad_norm": 22.981708526611328, | |
| "learning_rate": 1.0783625730994154e-05, | |
| "loss": 1.5223, | |
| "num_tokens": 470570.0, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.47959183673469385, | |
| "grad_norm": 23.430938720703125, | |
| "learning_rate": 1.0760233918128655e-05, | |
| "loss": 1.4387, | |
| "num_tokens": 471783.0, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.48072562358276644, | |
| "grad_norm": 22.37723159790039, | |
| "learning_rate": 1.073684210526316e-05, | |
| "loss": 1.4399, | |
| "num_tokens": 472934.0, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.481859410430839, | |
| "grad_norm": 22.111480712890625, | |
| "learning_rate": 1.0713450292397662e-05, | |
| "loss": 1.5395, | |
| "num_tokens": 474128.0, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.48299319727891155, | |
| "grad_norm": 22.287321090698242, | |
| "learning_rate": 1.0690058479532163e-05, | |
| "loss": 1.419, | |
| "num_tokens": 475240.0, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.48412698412698413, | |
| "grad_norm": 21.262020111083984, | |
| "learning_rate": 1.0666666666666667e-05, | |
| "loss": 1.5711, | |
| "num_tokens": 476354.0, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.4852607709750567, | |
| "grad_norm": 22.599716186523438, | |
| "learning_rate": 1.0643274853801172e-05, | |
| "loss": 1.5905, | |
| "num_tokens": 477408.0, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.48639455782312924, | |
| "grad_norm": 23.220651626586914, | |
| "learning_rate": 1.0619883040935672e-05, | |
| "loss": 1.5256, | |
| "num_tokens": 478458.0, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.4875283446712018, | |
| "grad_norm": 22.43808937072754, | |
| "learning_rate": 1.0596491228070177e-05, | |
| "loss": 1.4177, | |
| "num_tokens": 479577.0, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.4886621315192744, | |
| "grad_norm": 23.958772659301758, | |
| "learning_rate": 1.057309941520468e-05, | |
| "loss": 1.4556, | |
| "num_tokens": 480596.0, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.4897959183673469, | |
| "grad_norm": 23.579402923583984, | |
| "learning_rate": 1.0549707602339184e-05, | |
| "loss": 1.54, | |
| "num_tokens": 481636.0, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.4909297052154195, | |
| "grad_norm": 23.185216903686523, | |
| "learning_rate": 1.0526315789473684e-05, | |
| "loss": 1.5177, | |
| "num_tokens": 482749.0, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.49206349206349204, | |
| "grad_norm": 21.95652961730957, | |
| "learning_rate": 1.0502923976608189e-05, | |
| "loss": 1.4486, | |
| "num_tokens": 483894.0, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.4931972789115646, | |
| "grad_norm": 21.265581130981445, | |
| "learning_rate": 1.0479532163742691e-05, | |
| "loss": 1.5921, | |
| "num_tokens": 485099.0, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.4943310657596372, | |
| "grad_norm": 21.50389289855957, | |
| "learning_rate": 1.0456140350877194e-05, | |
| "loss": 1.5166, | |
| "num_tokens": 486238.0, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.4954648526077097, | |
| "grad_norm": 22.350412368774414, | |
| "learning_rate": 1.0432748538011696e-05, | |
| "loss": 1.4719, | |
| "num_tokens": 487264.0, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.4965986394557823, | |
| "grad_norm": 21.480653762817383, | |
| "learning_rate": 1.04093567251462e-05, | |
| "loss": 1.4648, | |
| "num_tokens": 488457.0, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.4977324263038549, | |
| "grad_norm": 23.51890754699707, | |
| "learning_rate": 1.0385964912280702e-05, | |
| "loss": 1.5191, | |
| "num_tokens": 489594.0, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.4988662131519274, | |
| "grad_norm": 21.588485717773438, | |
| "learning_rate": 1.0362573099415206e-05, | |
| "loss": 1.5039, | |
| "num_tokens": 490762.0, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 21.730770111083984, | |
| "learning_rate": 1.0339181286549708e-05, | |
| "loss": 1.4574, | |
| "num_tokens": 491869.0, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.5011337868480725, | |
| "grad_norm": 21.45746612548828, | |
| "learning_rate": 1.0315789473684213e-05, | |
| "loss": 1.5759, | |
| "num_tokens": 492996.0, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.5022675736961452, | |
| "grad_norm": 25.636009216308594, | |
| "learning_rate": 1.0292397660818714e-05, | |
| "loss": 1.5797, | |
| "num_tokens": 494026.0, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.5034013605442177, | |
| "grad_norm": 20.464859008789062, | |
| "learning_rate": 1.0269005847953218e-05, | |
| "loss": 1.469, | |
| "num_tokens": 495187.0, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.5045351473922902, | |
| "grad_norm": 21.59822654724121, | |
| "learning_rate": 1.024561403508772e-05, | |
| "loss": 1.4327, | |
| "num_tokens": 496377.0, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.5056689342403629, | |
| "grad_norm": 22.208587646484375, | |
| "learning_rate": 1.0222222222222223e-05, | |
| "loss": 1.5198, | |
| "num_tokens": 497515.0, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.5068027210884354, | |
| "grad_norm": 20.97501564025879, | |
| "learning_rate": 1.0198830409356726e-05, | |
| "loss": 1.3439, | |
| "num_tokens": 498679.0, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.5079365079365079, | |
| "grad_norm": 22.33463478088379, | |
| "learning_rate": 1.017543859649123e-05, | |
| "loss": 1.4406, | |
| "num_tokens": 499695.0, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.5090702947845805, | |
| "grad_norm": 23.04416275024414, | |
| "learning_rate": 1.015204678362573e-05, | |
| "loss": 1.397, | |
| "num_tokens": 500855.0, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.5102040816326531, | |
| "grad_norm": 21.69202995300293, | |
| "learning_rate": 1.0128654970760235e-05, | |
| "loss": 1.5077, | |
| "num_tokens": 501968.0, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.5113378684807256, | |
| "grad_norm": 24.2927303314209, | |
| "learning_rate": 1.0105263157894738e-05, | |
| "loss": 1.4229, | |
| "num_tokens": 502939.0, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.5124716553287982, | |
| "grad_norm": 22.449220657348633, | |
| "learning_rate": 1.0081871345029242e-05, | |
| "loss": 1.3924, | |
| "num_tokens": 503993.0, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.5136054421768708, | |
| "grad_norm": 23.610368728637695, | |
| "learning_rate": 1.0058479532163743e-05, | |
| "loss": 1.4355, | |
| "num_tokens": 505157.0, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.5147392290249433, | |
| "grad_norm": 21.155065536499023, | |
| "learning_rate": 1.0035087719298247e-05, | |
| "loss": 1.4111, | |
| "num_tokens": 506313.0, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.5158730158730159, | |
| "grad_norm": 23.355241775512695, | |
| "learning_rate": 1.001169590643275e-05, | |
| "loss": 1.4867, | |
| "num_tokens": 507464.0, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.5170068027210885, | |
| "grad_norm": 22.317737579345703, | |
| "learning_rate": 9.988304093567252e-06, | |
| "loss": 1.5623, | |
| "num_tokens": 508542.0, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.518140589569161, | |
| "grad_norm": 20.3128662109375, | |
| "learning_rate": 9.964912280701755e-06, | |
| "loss": 1.4996, | |
| "num_tokens": 509696.0, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.5192743764172335, | |
| "grad_norm": 21.98442268371582, | |
| "learning_rate": 9.941520467836257e-06, | |
| "loss": 1.5096, | |
| "num_tokens": 510844.0, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.5204081632653061, | |
| "grad_norm": 22.291845321655273, | |
| "learning_rate": 9.918128654970762e-06, | |
| "loss": 1.5076, | |
| "num_tokens": 511905.0, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.5215419501133787, | |
| "grad_norm": 21.11138153076172, | |
| "learning_rate": 9.894736842105264e-06, | |
| "loss": 1.4024, | |
| "num_tokens": 513025.0, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.5226757369614512, | |
| "grad_norm": 21.474123001098633, | |
| "learning_rate": 9.871345029239767e-06, | |
| "loss": 1.3739, | |
| "num_tokens": 514096.0, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.5238095238095238, | |
| "grad_norm": 22.77983283996582, | |
| "learning_rate": 9.84795321637427e-06, | |
| "loss": 1.537, | |
| "num_tokens": 515114.0, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.5249433106575964, | |
| "grad_norm": 22.172109603881836, | |
| "learning_rate": 9.824561403508772e-06, | |
| "loss": 1.4446, | |
| "num_tokens": 516218.0, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.5260770975056689, | |
| "grad_norm": 21.138439178466797, | |
| "learning_rate": 9.801169590643276e-06, | |
| "loss": 1.4733, | |
| "num_tokens": 517329.0, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.5272108843537415, | |
| "grad_norm": 22.17593765258789, | |
| "learning_rate": 9.777777777777779e-06, | |
| "loss": 1.5316, | |
| "num_tokens": 518435.0, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.528344671201814, | |
| "grad_norm": 22.072914123535156, | |
| "learning_rate": 9.754385964912281e-06, | |
| "loss": 1.3249, | |
| "num_tokens": 519509.0, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.5294784580498866, | |
| "grad_norm": 22.987707138061523, | |
| "learning_rate": 9.730994152046784e-06, | |
| "loss": 1.5979, | |
| "num_tokens": 520620.0, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.5306122448979592, | |
| "grad_norm": 20.49030876159668, | |
| "learning_rate": 9.707602339181286e-06, | |
| "loss": 1.3421, | |
| "num_tokens": 521806.0, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.5317460317460317, | |
| "grad_norm": 22.049985885620117, | |
| "learning_rate": 9.68421052631579e-06, | |
| "loss": 1.5731, | |
| "num_tokens": 522945.0, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.5328798185941043, | |
| "grad_norm": 21.49173927307129, | |
| "learning_rate": 9.660818713450293e-06, | |
| "loss": 1.4883, | |
| "num_tokens": 524157.0, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.5340136054421769, | |
| "grad_norm": 22.7564697265625, | |
| "learning_rate": 9.637426900584796e-06, | |
| "loss": 1.4914, | |
| "num_tokens": 525245.0, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.5351473922902494, | |
| "grad_norm": 21.756540298461914, | |
| "learning_rate": 9.614035087719298e-06, | |
| "loss": 1.4791, | |
| "num_tokens": 526398.0, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.536281179138322, | |
| "grad_norm": 20.813621520996094, | |
| "learning_rate": 9.590643274853801e-06, | |
| "loss": 1.5602, | |
| "num_tokens": 527591.0, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.5374149659863946, | |
| "grad_norm": 20.991981506347656, | |
| "learning_rate": 9.567251461988305e-06, | |
| "loss": 1.4139, | |
| "num_tokens": 528722.0, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.5385487528344671, | |
| "grad_norm": 21.36113166809082, | |
| "learning_rate": 9.543859649122808e-06, | |
| "loss": 1.3992, | |
| "num_tokens": 529926.0, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.5396825396825397, | |
| "grad_norm": 23.542648315429688, | |
| "learning_rate": 9.52046783625731e-06, | |
| "loss": 1.462, | |
| "num_tokens": 531002.0, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.5408163265306123, | |
| "grad_norm": 23.813064575195312, | |
| "learning_rate": 9.497076023391813e-06, | |
| "loss": 1.5113, | |
| "num_tokens": 532063.0, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.5419501133786848, | |
| "grad_norm": 23.0479736328125, | |
| "learning_rate": 9.473684210526315e-06, | |
| "loss": 1.3654, | |
| "num_tokens": 533122.0, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.5430839002267573, | |
| "grad_norm": 23.56155014038086, | |
| "learning_rate": 9.45029239766082e-06, | |
| "loss": 1.4703, | |
| "num_tokens": 534354.0, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.54421768707483, | |
| "grad_norm": 21.863500595092773, | |
| "learning_rate": 9.426900584795322e-06, | |
| "loss": 1.3612, | |
| "num_tokens": 535447.0, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.5453514739229025, | |
| "grad_norm": 20.762577056884766, | |
| "learning_rate": 9.403508771929825e-06, | |
| "loss": 1.5177, | |
| "num_tokens": 536573.0, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.546485260770975, | |
| "grad_norm": 21.422998428344727, | |
| "learning_rate": 9.380116959064327e-06, | |
| "loss": 1.435, | |
| "num_tokens": 537707.0, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.5476190476190477, | |
| "grad_norm": 21.787960052490234, | |
| "learning_rate": 9.35672514619883e-06, | |
| "loss": 1.5013, | |
| "num_tokens": 538818.0, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.5487528344671202, | |
| "grad_norm": 23.668500900268555, | |
| "learning_rate": 9.333333333333334e-06, | |
| "loss": 1.5828, | |
| "num_tokens": 539908.0, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.5498866213151927, | |
| "grad_norm": 22.385894775390625, | |
| "learning_rate": 9.309941520467837e-06, | |
| "loss": 1.4457, | |
| "num_tokens": 541045.0, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.5510204081632653, | |
| "grad_norm": 20.410974502563477, | |
| "learning_rate": 9.28654970760234e-06, | |
| "loss": 1.4135, | |
| "num_tokens": 542124.0, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.5521541950113379, | |
| "grad_norm": 22.673994064331055, | |
| "learning_rate": 9.263157894736842e-06, | |
| "loss": 1.4219, | |
| "num_tokens": 543220.0, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.5532879818594104, | |
| "grad_norm": 19.593944549560547, | |
| "learning_rate": 9.239766081871345e-06, | |
| "loss": 1.4269, | |
| "num_tokens": 544354.0, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.5544217687074829, | |
| "grad_norm": 23.19577407836914, | |
| "learning_rate": 9.216374269005849e-06, | |
| "loss": 1.5597, | |
| "num_tokens": 545379.0, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.5555555555555556, | |
| "grad_norm": 20.372406005859375, | |
| "learning_rate": 9.192982456140351e-06, | |
| "loss": 1.3499, | |
| "num_tokens": 546460.0, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.5566893424036281, | |
| "grad_norm": 18.678747177124023, | |
| "learning_rate": 9.169590643274856e-06, | |
| "loss": 1.3365, | |
| "num_tokens": 547604.0, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.5578231292517006, | |
| "grad_norm": 23.01070785522461, | |
| "learning_rate": 9.146198830409357e-06, | |
| "loss": 1.4461, | |
| "num_tokens": 548661.0, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.5589569160997733, | |
| "grad_norm": 23.752355575561523, | |
| "learning_rate": 9.12280701754386e-06, | |
| "loss": 1.5772, | |
| "num_tokens": 549763.0, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.5600907029478458, | |
| "grad_norm": 22.22798728942871, | |
| "learning_rate": 9.099415204678363e-06, | |
| "loss": 1.3951, | |
| "num_tokens": 550930.0, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.5612244897959183, | |
| "grad_norm": 21.435941696166992, | |
| "learning_rate": 9.076023391812866e-06, | |
| "loss": 1.5268, | |
| "num_tokens": 552031.0, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.562358276643991, | |
| "grad_norm": 22.89337730407715, | |
| "learning_rate": 9.05263157894737e-06, | |
| "loss": 1.4932, | |
| "num_tokens": 553188.0, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.5634920634920635, | |
| "grad_norm": 20.6674747467041, | |
| "learning_rate": 9.029239766081873e-06, | |
| "loss": 1.3984, | |
| "num_tokens": 554282.0, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.564625850340136, | |
| "grad_norm": 20.50768280029297, | |
| "learning_rate": 9.005847953216374e-06, | |
| "loss": 1.4644, | |
| "num_tokens": 555422.0, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.5657596371882087, | |
| "grad_norm": 20.704288482666016, | |
| "learning_rate": 8.982456140350878e-06, | |
| "loss": 1.6067, | |
| "num_tokens": 556575.0, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.5668934240362812, | |
| "grad_norm": 22.014612197875977, | |
| "learning_rate": 8.95906432748538e-06, | |
| "loss": 1.5485, | |
| "num_tokens": 557680.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.5680272108843537, | |
| "grad_norm": 23.033823013305664, | |
| "learning_rate": 8.935672514619885e-06, | |
| "loss": 1.4399, | |
| "num_tokens": 558791.0, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.5691609977324263, | |
| "grad_norm": 22.817472457885742, | |
| "learning_rate": 8.912280701754387e-06, | |
| "loss": 1.239, | |
| "num_tokens": 559835.0, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.5702947845804989, | |
| "grad_norm": 21.822277069091797, | |
| "learning_rate": 8.888888888888888e-06, | |
| "loss": 1.4371, | |
| "num_tokens": 560975.0, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 22.173824310302734, | |
| "learning_rate": 8.865497076023393e-06, | |
| "loss": 1.3944, | |
| "num_tokens": 562132.0, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.572562358276644, | |
| "grad_norm": 22.091264724731445, | |
| "learning_rate": 8.842105263157895e-06, | |
| "loss": 1.3903, | |
| "num_tokens": 563198.0, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.5736961451247166, | |
| "grad_norm": 20.748226165771484, | |
| "learning_rate": 8.8187134502924e-06, | |
| "loss": 1.3665, | |
| "num_tokens": 564323.0, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.5748299319727891, | |
| "grad_norm": 25.668231964111328, | |
| "learning_rate": 8.795321637426902e-06, | |
| "loss": 1.6003, | |
| "num_tokens": 565410.0, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.5759637188208617, | |
| "grad_norm": 21.842506408691406, | |
| "learning_rate": 8.771929824561405e-06, | |
| "loss": 1.4388, | |
| "num_tokens": 566583.0, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.5770975056689343, | |
| "grad_norm": 21.225013732910156, | |
| "learning_rate": 8.748538011695907e-06, | |
| "loss": 1.3724, | |
| "num_tokens": 567705.0, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.5782312925170068, | |
| "grad_norm": 22.67068862915039, | |
| "learning_rate": 8.72514619883041e-06, | |
| "loss": 1.3551, | |
| "num_tokens": 568800.0, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.5793650793650794, | |
| "grad_norm": 21.34926986694336, | |
| "learning_rate": 8.701754385964914e-06, | |
| "loss": 1.3908, | |
| "num_tokens": 569913.0, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.5804988662131519, | |
| "grad_norm": 22.283945083618164, | |
| "learning_rate": 8.678362573099417e-06, | |
| "loss": 1.4563, | |
| "num_tokens": 571017.0, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.5816326530612245, | |
| "grad_norm": 20.993438720703125, | |
| "learning_rate": 8.654970760233919e-06, | |
| "loss": 1.3512, | |
| "num_tokens": 572126.0, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.5827664399092971, | |
| "grad_norm": 22.38656234741211, | |
| "learning_rate": 8.631578947368422e-06, | |
| "loss": 1.5402, | |
| "num_tokens": 573264.0, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.5839002267573696, | |
| "grad_norm": 23.537073135375977, | |
| "learning_rate": 8.608187134502924e-06, | |
| "loss": 1.3946, | |
| "num_tokens": 574289.0, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.5850340136054422, | |
| "grad_norm": 22.79361915588379, | |
| "learning_rate": 8.584795321637429e-06, | |
| "loss": 1.5559, | |
| "num_tokens": 575350.0, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.5861678004535147, | |
| "grad_norm": 21.405969619750977, | |
| "learning_rate": 8.561403508771931e-06, | |
| "loss": 1.4803, | |
| "num_tokens": 576453.0, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.5873015873015873, | |
| "grad_norm": 23.284671783447266, | |
| "learning_rate": 8.538011695906434e-06, | |
| "loss": 1.5008, | |
| "num_tokens": 577497.0, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.5884353741496599, | |
| "grad_norm": 21.37405776977539, | |
| "learning_rate": 8.514619883040936e-06, | |
| "loss": 1.4739, | |
| "num_tokens": 578563.0, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.5895691609977324, | |
| "grad_norm": 24.321304321289062, | |
| "learning_rate": 8.491228070175439e-06, | |
| "loss": 1.5819, | |
| "num_tokens": 579721.0, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.590702947845805, | |
| "grad_norm": 21.786588668823242, | |
| "learning_rate": 8.467836257309943e-06, | |
| "loss": 1.4846, | |
| "num_tokens": 580835.0, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.5918367346938775, | |
| "grad_norm": 22.301950454711914, | |
| "learning_rate": 8.444444444444446e-06, | |
| "loss": 1.4472, | |
| "num_tokens": 582031.0, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.5929705215419501, | |
| "grad_norm": 22.732772827148438, | |
| "learning_rate": 8.421052631578948e-06, | |
| "loss": 1.4803, | |
| "num_tokens": 583095.0, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.5941043083900227, | |
| "grad_norm": 21.661388397216797, | |
| "learning_rate": 8.39766081871345e-06, | |
| "loss": 1.3937, | |
| "num_tokens": 584233.0, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.5952380952380952, | |
| "grad_norm": 20.343976974487305, | |
| "learning_rate": 8.374269005847953e-06, | |
| "loss": 1.353, | |
| "num_tokens": 585302.0, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.5963718820861678, | |
| "grad_norm": 22.84857940673828, | |
| "learning_rate": 8.350877192982458e-06, | |
| "loss": 1.4354, | |
| "num_tokens": 586371.0, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.5975056689342404, | |
| "grad_norm": 22.94053077697754, | |
| "learning_rate": 8.32748538011696e-06, | |
| "loss": 1.5107, | |
| "num_tokens": 587512.0, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.5986394557823129, | |
| "grad_norm": 22.8999080657959, | |
| "learning_rate": 8.304093567251463e-06, | |
| "loss": 1.6091, | |
| "num_tokens": 588596.0, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.5997732426303855, | |
| "grad_norm": 23.222049713134766, | |
| "learning_rate": 8.280701754385965e-06, | |
| "loss": 1.4075, | |
| "num_tokens": 589699.0, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.6009070294784581, | |
| "grad_norm": 20.573713302612305, | |
| "learning_rate": 8.257309941520468e-06, | |
| "loss": 1.5905, | |
| "num_tokens": 590891.0, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.6020408163265306, | |
| "grad_norm": 21.398988723754883, | |
| "learning_rate": 8.233918128654972e-06, | |
| "loss": 1.4535, | |
| "num_tokens": 592028.0, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.6031746031746031, | |
| "grad_norm": 22.95913314819336, | |
| "learning_rate": 8.210526315789475e-06, | |
| "loss": 1.371, | |
| "num_tokens": 593213.0, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.6043083900226758, | |
| "grad_norm": 21.470468521118164, | |
| "learning_rate": 8.187134502923977e-06, | |
| "loss": 1.6771, | |
| "num_tokens": 594442.0, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.6054421768707483, | |
| "grad_norm": 22.546823501586914, | |
| "learning_rate": 8.16374269005848e-06, | |
| "loss": 1.5234, | |
| "num_tokens": 595601.0, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.6065759637188208, | |
| "grad_norm": 25.772891998291016, | |
| "learning_rate": 8.140350877192983e-06, | |
| "loss": 1.4929, | |
| "num_tokens": 596802.0, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.6077097505668935, | |
| "grad_norm": 22.497135162353516, | |
| "learning_rate": 8.116959064327487e-06, | |
| "loss": 1.4492, | |
| "num_tokens": 597920.0, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.608843537414966, | |
| "grad_norm": 24.795047760009766, | |
| "learning_rate": 8.09356725146199e-06, | |
| "loss": 1.3766, | |
| "num_tokens": 598998.0, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.6099773242630385, | |
| "grad_norm": 20.771310806274414, | |
| "learning_rate": 8.070175438596492e-06, | |
| "loss": 1.6282, | |
| "num_tokens": 600212.0, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.6111111111111112, | |
| "grad_norm": 24.573957443237305, | |
| "learning_rate": 8.046783625730994e-06, | |
| "loss": 1.3345, | |
| "num_tokens": 601248.0, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.6122448979591837, | |
| "grad_norm": 20.914003372192383, | |
| "learning_rate": 8.023391812865497e-06, | |
| "loss": 1.4553, | |
| "num_tokens": 602367.0, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.6133786848072562, | |
| "grad_norm": 22.030210494995117, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.4759, | |
| "num_tokens": 603524.0, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.6145124716553289, | |
| "grad_norm": 22.657468795776367, | |
| "learning_rate": 7.976608187134504e-06, | |
| "loss": 1.4916, | |
| "num_tokens": 604618.0, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.6156462585034014, | |
| "grad_norm": 23.671598434448242, | |
| "learning_rate": 7.953216374269006e-06, | |
| "loss": 1.3016, | |
| "num_tokens": 605667.0, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.6167800453514739, | |
| "grad_norm": 22.971860885620117, | |
| "learning_rate": 7.929824561403509e-06, | |
| "loss": 1.5056, | |
| "num_tokens": 606768.0, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.6179138321995464, | |
| "grad_norm": 22.38102912902832, | |
| "learning_rate": 7.906432748538012e-06, | |
| "loss": 1.5457, | |
| "num_tokens": 607836.0, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.6190476190476191, | |
| "grad_norm": 21.6705379486084, | |
| "learning_rate": 7.883040935672516e-06, | |
| "loss": 1.4587, | |
| "num_tokens": 608915.0, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.6201814058956916, | |
| "grad_norm": 22.290454864501953, | |
| "learning_rate": 7.859649122807018e-06, | |
| "loss": 1.5118, | |
| "num_tokens": 609979.0, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.6213151927437641, | |
| "grad_norm": 22.28474998474121, | |
| "learning_rate": 7.836257309941521e-06, | |
| "loss": 1.5042, | |
| "num_tokens": 611160.0, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.6224489795918368, | |
| "grad_norm": 23.13262176513672, | |
| "learning_rate": 7.812865497076024e-06, | |
| "loss": 1.4265, | |
| "num_tokens": 612235.0, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.6235827664399093, | |
| "grad_norm": 25.128522872924805, | |
| "learning_rate": 7.789473684210526e-06, | |
| "loss": 1.6554, | |
| "num_tokens": 613310.0, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.6247165532879818, | |
| "grad_norm": 20.930503845214844, | |
| "learning_rate": 7.76608187134503e-06, | |
| "loss": 1.4841, | |
| "num_tokens": 614403.0, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.6258503401360545, | |
| "grad_norm": 21.933982849121094, | |
| "learning_rate": 7.742690058479533e-06, | |
| "loss": 1.4161, | |
| "num_tokens": 615531.0, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.626984126984127, | |
| "grad_norm": 20.2902889251709, | |
| "learning_rate": 7.719298245614036e-06, | |
| "loss": 1.4999, | |
| "num_tokens": 616780.0, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.6281179138321995, | |
| "grad_norm": 20.942724227905273, | |
| "learning_rate": 7.695906432748538e-06, | |
| "loss": 1.3684, | |
| "num_tokens": 617962.0, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.6292517006802721, | |
| "grad_norm": 21.612571716308594, | |
| "learning_rate": 7.67251461988304e-06, | |
| "loss": 1.5484, | |
| "num_tokens": 619151.0, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.6303854875283447, | |
| "grad_norm": 22.789594650268555, | |
| "learning_rate": 7.649122807017545e-06, | |
| "loss": 1.4786, | |
| "num_tokens": 620263.0, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.6315192743764172, | |
| "grad_norm": 20.90570640563965, | |
| "learning_rate": 7.625730994152048e-06, | |
| "loss": 1.6095, | |
| "num_tokens": 621484.0, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.6326530612244898, | |
| "grad_norm": 25.00489616394043, | |
| "learning_rate": 7.60233918128655e-06, | |
| "loss": 1.6409, | |
| "num_tokens": 622514.0, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.6337868480725624, | |
| "grad_norm": 20.243322372436523, | |
| "learning_rate": 7.578947368421054e-06, | |
| "loss": 1.5188, | |
| "num_tokens": 623693.0, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 0.6349206349206349, | |
| "grad_norm": 22.97846221923828, | |
| "learning_rate": 7.555555555555556e-06, | |
| "loss": 1.5657, | |
| "num_tokens": 624782.0, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.6360544217687075, | |
| "grad_norm": 24.399961471557617, | |
| "learning_rate": 7.5321637426900596e-06, | |
| "loss": 1.5053, | |
| "num_tokens": 625875.0, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 0.63718820861678, | |
| "grad_norm": 20.759233474731445, | |
| "learning_rate": 7.508771929824562e-06, | |
| "loss": 1.5293, | |
| "num_tokens": 627019.0, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 0.6383219954648526, | |
| "grad_norm": 21.16717529296875, | |
| "learning_rate": 7.485380116959065e-06, | |
| "loss": 1.3894, | |
| "num_tokens": 628139.0, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 0.6394557823129252, | |
| "grad_norm": 22.11092758178711, | |
| "learning_rate": 7.461988304093568e-06, | |
| "loss": 1.4995, | |
| "num_tokens": 629274.0, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.6405895691609977, | |
| "grad_norm": 27.72738265991211, | |
| "learning_rate": 7.438596491228071e-06, | |
| "loss": 1.3619, | |
| "num_tokens": 630243.0, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.6417233560090703, | |
| "grad_norm": 20.804826736450195, | |
| "learning_rate": 7.415204678362574e-06, | |
| "loss": 1.4518, | |
| "num_tokens": 631375.0, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 0.6428571428571429, | |
| "grad_norm": 21.256473541259766, | |
| "learning_rate": 7.391812865497077e-06, | |
| "loss": 1.4972, | |
| "num_tokens": 632486.0, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.6439909297052154, | |
| "grad_norm": 22.026762008666992, | |
| "learning_rate": 7.368421052631579e-06, | |
| "loss": 1.3551, | |
| "num_tokens": 633618.0, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.645124716553288, | |
| "grad_norm": 21.86498260498047, | |
| "learning_rate": 7.345029239766083e-06, | |
| "loss": 1.5152, | |
| "num_tokens": 634700.0, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 0.6462585034013606, | |
| "grad_norm": 22.414583206176758, | |
| "learning_rate": 7.321637426900585e-06, | |
| "loss": 1.3151, | |
| "num_tokens": 635812.0, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.6473922902494331, | |
| "grad_norm": 21.637290954589844, | |
| "learning_rate": 7.298245614035089e-06, | |
| "loss": 1.4313, | |
| "num_tokens": 636943.0, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 0.6485260770975056, | |
| "grad_norm": 22.154354095458984, | |
| "learning_rate": 7.274853801169591e-06, | |
| "loss": 1.3439, | |
| "num_tokens": 638032.0, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.6496598639455783, | |
| "grad_norm": 24.491914749145508, | |
| "learning_rate": 7.251461988304094e-06, | |
| "loss": 1.5135, | |
| "num_tokens": 639185.0, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 0.6507936507936508, | |
| "grad_norm": 23.253528594970703, | |
| "learning_rate": 7.228070175438597e-06, | |
| "loss": 1.4544, | |
| "num_tokens": 640432.0, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.6519274376417233, | |
| "grad_norm": 22.823341369628906, | |
| "learning_rate": 7.2046783625731e-06, | |
| "loss": 1.326, | |
| "num_tokens": 641575.0, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.6530612244897959, | |
| "grad_norm": 22.95077896118164, | |
| "learning_rate": 7.181286549707603e-06, | |
| "loss": 1.6407, | |
| "num_tokens": 642791.0, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.6541950113378685, | |
| "grad_norm": 22.988304138183594, | |
| "learning_rate": 7.157894736842106e-06, | |
| "loss": 1.4079, | |
| "num_tokens": 643851.0, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 0.655328798185941, | |
| "grad_norm": 22.63962745666504, | |
| "learning_rate": 7.134502923976608e-06, | |
| "loss": 1.3522, | |
| "num_tokens": 644917.0, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 0.6564625850340136, | |
| "grad_norm": 26.570436477661133, | |
| "learning_rate": 7.111111111111112e-06, | |
| "loss": 1.5682, | |
| "num_tokens": 645970.0, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 0.6575963718820862, | |
| "grad_norm": 22.55550765991211, | |
| "learning_rate": 7.087719298245614e-06, | |
| "loss": 1.4318, | |
| "num_tokens": 647052.0, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.6587301587301587, | |
| "grad_norm": 22.61783218383789, | |
| "learning_rate": 7.064327485380118e-06, | |
| "loss": 1.4444, | |
| "num_tokens": 648119.0, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.6598639455782312, | |
| "grad_norm": 22.553382873535156, | |
| "learning_rate": 7.04093567251462e-06, | |
| "loss": 1.359, | |
| "num_tokens": 649162.0, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.6609977324263039, | |
| "grad_norm": 21.678308486938477, | |
| "learning_rate": 7.017543859649123e-06, | |
| "loss": 1.3491, | |
| "num_tokens": 650283.0, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 0.6621315192743764, | |
| "grad_norm": 21.3242130279541, | |
| "learning_rate": 6.994152046783626e-06, | |
| "loss": 1.4176, | |
| "num_tokens": 651440.0, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.6632653061224489, | |
| "grad_norm": 24.34210777282715, | |
| "learning_rate": 6.970760233918129e-06, | |
| "loss": 1.5172, | |
| "num_tokens": 652549.0, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.6643990929705216, | |
| "grad_norm": 24.187849044799805, | |
| "learning_rate": 6.947368421052632e-06, | |
| "loss": 1.5697, | |
| "num_tokens": 653590.0, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 0.6655328798185941, | |
| "grad_norm": 22.247058868408203, | |
| "learning_rate": 6.923976608187135e-06, | |
| "loss": 1.3541, | |
| "num_tokens": 654714.0, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 21.859542846679688, | |
| "learning_rate": 6.9005847953216375e-06, | |
| "loss": 1.492, | |
| "num_tokens": 655964.0, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.6678004535147393, | |
| "grad_norm": 22.576213836669922, | |
| "learning_rate": 6.877192982456141e-06, | |
| "loss": 1.4299, | |
| "num_tokens": 657065.0, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 0.6689342403628118, | |
| "grad_norm": 22.9122257232666, | |
| "learning_rate": 6.8538011695906435e-06, | |
| "loss": 1.4698, | |
| "num_tokens": 658255.0, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.6700680272108843, | |
| "grad_norm": 23.04901123046875, | |
| "learning_rate": 6.830409356725147e-06, | |
| "loss": 1.5064, | |
| "num_tokens": 659325.0, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 0.671201814058957, | |
| "grad_norm": 23.730091094970703, | |
| "learning_rate": 6.8070175438596495e-06, | |
| "loss": 1.4925, | |
| "num_tokens": 660485.0, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.6723356009070295, | |
| "grad_norm": 22.81539535522461, | |
| "learning_rate": 6.783625730994152e-06, | |
| "loss": 1.3731, | |
| "num_tokens": 661513.0, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 0.673469387755102, | |
| "grad_norm": 23.25772476196289, | |
| "learning_rate": 6.7602339181286555e-06, | |
| "loss": 1.4893, | |
| "num_tokens": 662618.0, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.6746031746031746, | |
| "grad_norm": 23.883750915527344, | |
| "learning_rate": 6.736842105263158e-06, | |
| "loss": 1.2276, | |
| "num_tokens": 663737.0, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.6757369614512472, | |
| "grad_norm": 23.94465446472168, | |
| "learning_rate": 6.7134502923976615e-06, | |
| "loss": 1.4642, | |
| "num_tokens": 664857.0, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.6768707482993197, | |
| "grad_norm": 22.831388473510742, | |
| "learning_rate": 6.690058479532164e-06, | |
| "loss": 1.4457, | |
| "num_tokens": 665954.0, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 0.6780045351473923, | |
| "grad_norm": 22.41533851623535, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 1.5384, | |
| "num_tokens": 667045.0, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 0.6791383219954649, | |
| "grad_norm": 21.215373992919922, | |
| "learning_rate": 6.64327485380117e-06, | |
| "loss": 1.5598, | |
| "num_tokens": 668272.0, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 0.6802721088435374, | |
| "grad_norm": 21.893632888793945, | |
| "learning_rate": 6.619883040935673e-06, | |
| "loss": 1.5232, | |
| "num_tokens": 669382.0, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.68140589569161, | |
| "grad_norm": 22.526578903198242, | |
| "learning_rate": 6.596491228070177e-06, | |
| "loss": 1.5125, | |
| "num_tokens": 670453.0, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 0.6825396825396826, | |
| "grad_norm": 22.798067092895508, | |
| "learning_rate": 6.573099415204679e-06, | |
| "loss": 1.4695, | |
| "num_tokens": 671545.0, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.6836734693877551, | |
| "grad_norm": 21.24146842956543, | |
| "learning_rate": 6.549707602339181e-06, | |
| "loss": 1.46, | |
| "num_tokens": 672697.0, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.6848072562358276, | |
| "grad_norm": 24.16074562072754, | |
| "learning_rate": 6.526315789473685e-06, | |
| "loss": 1.4916, | |
| "num_tokens": 673798.0, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.6859410430839002, | |
| "grad_norm": 21.984033584594727, | |
| "learning_rate": 6.502923976608187e-06, | |
| "loss": 1.3227, | |
| "num_tokens": 674822.0, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.6870748299319728, | |
| "grad_norm": 23.14649772644043, | |
| "learning_rate": 6.4795321637426915e-06, | |
| "loss": 1.5123, | |
| "num_tokens": 675953.0, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.6882086167800453, | |
| "grad_norm": 21.985637664794922, | |
| "learning_rate": 6.456140350877193e-06, | |
| "loss": 1.5573, | |
| "num_tokens": 677069.0, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 0.6893424036281179, | |
| "grad_norm": 22.13504409790039, | |
| "learning_rate": 6.432748538011696e-06, | |
| "loss": 1.5341, | |
| "num_tokens": 678207.0, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.6904761904761905, | |
| "grad_norm": 22.876258850097656, | |
| "learning_rate": 6.4093567251462e-06, | |
| "loss": 1.5077, | |
| "num_tokens": 679333.0, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.691609977324263, | |
| "grad_norm": 22.648876190185547, | |
| "learning_rate": 6.385964912280702e-06, | |
| "loss": 1.5165, | |
| "num_tokens": 680472.0, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.6927437641723356, | |
| "grad_norm": 24.44489860534668, | |
| "learning_rate": 6.362573099415206e-06, | |
| "loss": 1.4595, | |
| "num_tokens": 681550.0, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 0.6938775510204082, | |
| "grad_norm": 23.062849044799805, | |
| "learning_rate": 6.339181286549709e-06, | |
| "loss": 1.5479, | |
| "num_tokens": 682665.0, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.6950113378684807, | |
| "grad_norm": 25.6104793548584, | |
| "learning_rate": 6.31578947368421e-06, | |
| "loss": 1.4372, | |
| "num_tokens": 683797.0, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 0.6961451247165533, | |
| "grad_norm": 22.21469497680664, | |
| "learning_rate": 6.292397660818715e-06, | |
| "loss": 1.4886, | |
| "num_tokens": 684941.0, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 0.6972789115646258, | |
| "grad_norm": 22.583446502685547, | |
| "learning_rate": 6.269005847953217e-06, | |
| "loss": 1.4647, | |
| "num_tokens": 686074.0, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.6984126984126984, | |
| "grad_norm": 23.686786651611328, | |
| "learning_rate": 6.245614035087721e-06, | |
| "loss": 1.3958, | |
| "num_tokens": 687271.0, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 0.699546485260771, | |
| "grad_norm": 23.05241584777832, | |
| "learning_rate": 6.222222222222223e-06, | |
| "loss": 1.4258, | |
| "num_tokens": 688341.0, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 0.7006802721088435, | |
| "grad_norm": 23.340354919433594, | |
| "learning_rate": 6.198830409356725e-06, | |
| "loss": 1.5196, | |
| "num_tokens": 689435.0, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 0.7018140589569161, | |
| "grad_norm": 22.070066452026367, | |
| "learning_rate": 6.175438596491229e-06, | |
| "loss": 1.3828, | |
| "num_tokens": 690569.0, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 0.7029478458049887, | |
| "grad_norm": 23.374649047851562, | |
| "learning_rate": 6.152046783625732e-06, | |
| "loss": 1.4309, | |
| "num_tokens": 691672.0, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.7040816326530612, | |
| "grad_norm": 23.68497085571289, | |
| "learning_rate": 6.128654970760235e-06, | |
| "loss": 1.4063, | |
| "num_tokens": 692947.0, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 0.7052154195011338, | |
| "grad_norm": 22.70351791381836, | |
| "learning_rate": 6.105263157894738e-06, | |
| "loss": 1.4569, | |
| "num_tokens": 694052.0, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 0.7063492063492064, | |
| "grad_norm": 21.42176055908203, | |
| "learning_rate": 6.08187134502924e-06, | |
| "loss": 1.4617, | |
| "num_tokens": 695248.0, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 0.7074829931972789, | |
| "grad_norm": 22.26094627380371, | |
| "learning_rate": 6.058479532163744e-06, | |
| "loss": 1.516, | |
| "num_tokens": 696356.0, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 0.7086167800453514, | |
| "grad_norm": 26.08213233947754, | |
| "learning_rate": 6.035087719298246e-06, | |
| "loss": 1.4949, | |
| "num_tokens": 697517.0, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.7097505668934241, | |
| "grad_norm": 22.348121643066406, | |
| "learning_rate": 6.01169590643275e-06, | |
| "loss": 1.4339, | |
| "num_tokens": 698582.0, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 0.7108843537414966, | |
| "grad_norm": 22.22833251953125, | |
| "learning_rate": 5.988304093567252e-06, | |
| "loss": 1.4808, | |
| "num_tokens": 699745.0, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 0.7120181405895691, | |
| "grad_norm": 23.78014373779297, | |
| "learning_rate": 5.964912280701755e-06, | |
| "loss": 1.4515, | |
| "num_tokens": 700778.0, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 0.7131519274376418, | |
| "grad_norm": 23.705251693725586, | |
| "learning_rate": 5.941520467836258e-06, | |
| "loss": 1.4206, | |
| "num_tokens": 701913.0, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 20.9232234954834, | |
| "learning_rate": 5.918128654970761e-06, | |
| "loss": 1.337, | |
| "num_tokens": 703034.0, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.7154195011337868, | |
| "grad_norm": 20.941648483276367, | |
| "learning_rate": 5.8947368421052634e-06, | |
| "loss": 1.3962, | |
| "num_tokens": 704183.0, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 0.7165532879818595, | |
| "grad_norm": 21.786916732788086, | |
| "learning_rate": 5.871345029239767e-06, | |
| "loss": 1.3946, | |
| "num_tokens": 705285.0, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 0.717687074829932, | |
| "grad_norm": 24.243087768554688, | |
| "learning_rate": 5.847953216374269e-06, | |
| "loss": 1.4448, | |
| "num_tokens": 706394.0, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 0.7188208616780045, | |
| "grad_norm": 24.048946380615234, | |
| "learning_rate": 5.824561403508773e-06, | |
| "loss": 1.5736, | |
| "num_tokens": 707507.0, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 0.719954648526077, | |
| "grad_norm": 20.490083694458008, | |
| "learning_rate": 5.801169590643275e-06, | |
| "loss": 1.531, | |
| "num_tokens": 708763.0, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.7210884353741497, | |
| "grad_norm": 23.876779556274414, | |
| "learning_rate": 5.777777777777778e-06, | |
| "loss": 1.4557, | |
| "num_tokens": 709792.0, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 0.7222222222222222, | |
| "grad_norm": 24.66008186340332, | |
| "learning_rate": 5.754385964912281e-06, | |
| "loss": 1.517, | |
| "num_tokens": 710874.0, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 0.7233560090702947, | |
| "grad_norm": 20.79994010925293, | |
| "learning_rate": 5.730994152046784e-06, | |
| "loss": 1.3886, | |
| "num_tokens": 712053.0, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 0.7244897959183674, | |
| "grad_norm": 23.969280242919922, | |
| "learning_rate": 5.707602339181287e-06, | |
| "loss": 1.3954, | |
| "num_tokens": 713114.0, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 0.7256235827664399, | |
| "grad_norm": 23.537185668945312, | |
| "learning_rate": 5.68421052631579e-06, | |
| "loss": 1.3893, | |
| "num_tokens": 714201.0, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.7267573696145124, | |
| "grad_norm": 22.46196937561035, | |
| "learning_rate": 5.6608187134502925e-06, | |
| "loss": 1.5846, | |
| "num_tokens": 715281.0, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 0.7278911564625851, | |
| "grad_norm": 20.890539169311523, | |
| "learning_rate": 5.637426900584796e-06, | |
| "loss": 1.4379, | |
| "num_tokens": 716404.0, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 0.7290249433106576, | |
| "grad_norm": 21.008331298828125, | |
| "learning_rate": 5.6140350877192985e-06, | |
| "loss": 1.4799, | |
| "num_tokens": 717488.0, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 0.7301587301587301, | |
| "grad_norm": 22.868465423583984, | |
| "learning_rate": 5.590643274853802e-06, | |
| "loss": 1.484, | |
| "num_tokens": 718571.0, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 0.7312925170068028, | |
| "grad_norm": 23.07869529724121, | |
| "learning_rate": 5.5672514619883045e-06, | |
| "loss": 1.484, | |
| "num_tokens": 719716.0, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.7324263038548753, | |
| "grad_norm": 24.570358276367188, | |
| "learning_rate": 5.543859649122807e-06, | |
| "loss": 1.5601, | |
| "num_tokens": 720905.0, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 0.7335600907029478, | |
| "grad_norm": 26.3187255859375, | |
| "learning_rate": 5.5204678362573105e-06, | |
| "loss": 1.5891, | |
| "num_tokens": 722076.0, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 0.7346938775510204, | |
| "grad_norm": 22.848508834838867, | |
| "learning_rate": 5.497076023391813e-06, | |
| "loss": 1.3089, | |
| "num_tokens": 723183.0, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 0.735827664399093, | |
| "grad_norm": 22.79510498046875, | |
| "learning_rate": 5.4736842105263165e-06, | |
| "loss": 1.5494, | |
| "num_tokens": 724329.0, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 0.7369614512471655, | |
| "grad_norm": 20.732099533081055, | |
| "learning_rate": 5.450292397660819e-06, | |
| "loss": 1.4599, | |
| "num_tokens": 725447.0, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.7380952380952381, | |
| "grad_norm": 22.604984283447266, | |
| "learning_rate": 5.426900584795322e-06, | |
| "loss": 1.3155, | |
| "num_tokens": 726477.0, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 0.7392290249433107, | |
| "grad_norm": 21.90055274963379, | |
| "learning_rate": 5.403508771929825e-06, | |
| "loss": 1.5524, | |
| "num_tokens": 727606.0, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 0.7403628117913832, | |
| "grad_norm": 21.966693878173828, | |
| "learning_rate": 5.380116959064328e-06, | |
| "loss": 1.5046, | |
| "num_tokens": 728718.0, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 0.7414965986394558, | |
| "grad_norm": 25.208890914916992, | |
| "learning_rate": 5.356725146198831e-06, | |
| "loss": 1.6362, | |
| "num_tokens": 729745.0, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 0.7426303854875284, | |
| "grad_norm": 24.090883255004883, | |
| "learning_rate": 5.333333333333334e-06, | |
| "loss": 1.3422, | |
| "num_tokens": 730924.0, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.7437641723356009, | |
| "grad_norm": 22.79339599609375, | |
| "learning_rate": 5.309941520467836e-06, | |
| "loss": 1.3421, | |
| "num_tokens": 731980.0, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 0.7448979591836735, | |
| "grad_norm": 23.833890914916992, | |
| "learning_rate": 5.28654970760234e-06, | |
| "loss": 1.5008, | |
| "num_tokens": 733073.0, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 0.746031746031746, | |
| "grad_norm": 19.78729820251465, | |
| "learning_rate": 5.263157894736842e-06, | |
| "loss": 1.3997, | |
| "num_tokens": 734355.0, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 0.7471655328798186, | |
| "grad_norm": 24.19782829284668, | |
| "learning_rate": 5.239766081871346e-06, | |
| "loss": 1.4371, | |
| "num_tokens": 735425.0, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 0.7482993197278912, | |
| "grad_norm": 23.320068359375, | |
| "learning_rate": 5.216374269005848e-06, | |
| "loss": 1.3869, | |
| "num_tokens": 736522.0, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.7494331065759637, | |
| "grad_norm": 22.0587158203125, | |
| "learning_rate": 5.192982456140351e-06, | |
| "loss": 1.4641, | |
| "num_tokens": 737633.0, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 0.7505668934240363, | |
| "grad_norm": 22.994808197021484, | |
| "learning_rate": 5.169590643274854e-06, | |
| "loss": 1.4685, | |
| "num_tokens": 738753.0, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 0.7517006802721088, | |
| "grad_norm": 22.301300048828125, | |
| "learning_rate": 5.146198830409357e-06, | |
| "loss": 1.4651, | |
| "num_tokens": 739872.0, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 0.7528344671201814, | |
| "grad_norm": 21.97808837890625, | |
| "learning_rate": 5.12280701754386e-06, | |
| "loss": 1.4382, | |
| "num_tokens": 741008.0, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 0.753968253968254, | |
| "grad_norm": 21.360986709594727, | |
| "learning_rate": 5.099415204678363e-06, | |
| "loss": 1.5411, | |
| "num_tokens": 742227.0, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.7551020408163265, | |
| "grad_norm": 22.520597457885742, | |
| "learning_rate": 5.076023391812865e-06, | |
| "loss": 1.4047, | |
| "num_tokens": 743294.0, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 0.7562358276643991, | |
| "grad_norm": 22.39497184753418, | |
| "learning_rate": 5.052631578947369e-06, | |
| "loss": 1.3934, | |
| "num_tokens": 744498.0, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 0.7573696145124716, | |
| "grad_norm": 21.658981323242188, | |
| "learning_rate": 5.029239766081871e-06, | |
| "loss": 1.4161, | |
| "num_tokens": 745753.0, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 0.7585034013605442, | |
| "grad_norm": 22.147428512573242, | |
| "learning_rate": 5.005847953216375e-06, | |
| "loss": 1.365, | |
| "num_tokens": 746899.0, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 0.7596371882086168, | |
| "grad_norm": 22.580663681030273, | |
| "learning_rate": 4.982456140350877e-06, | |
| "loss": 1.5026, | |
| "num_tokens": 747982.0, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.7607709750566893, | |
| "grad_norm": 22.9129581451416, | |
| "learning_rate": 4.959064327485381e-06, | |
| "loss": 1.3533, | |
| "num_tokens": 749043.0, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 0.7619047619047619, | |
| "grad_norm": 20.52309226989746, | |
| "learning_rate": 4.935672514619883e-06, | |
| "loss": 1.3866, | |
| "num_tokens": 750230.0, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 0.7630385487528345, | |
| "grad_norm": 25.36422348022461, | |
| "learning_rate": 4.912280701754386e-06, | |
| "loss": 1.5019, | |
| "num_tokens": 751311.0, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 0.764172335600907, | |
| "grad_norm": 20.621496200561523, | |
| "learning_rate": 4.888888888888889e-06, | |
| "loss": 1.4865, | |
| "num_tokens": 752533.0, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 0.7653061224489796, | |
| "grad_norm": 25.740957260131836, | |
| "learning_rate": 4.865497076023392e-06, | |
| "loss": 1.363, | |
| "num_tokens": 753621.0, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.7664399092970522, | |
| "grad_norm": 22.625244140625, | |
| "learning_rate": 4.842105263157895e-06, | |
| "loss": 1.4387, | |
| "num_tokens": 754720.0, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 0.7675736961451247, | |
| "grad_norm": 23.55186653137207, | |
| "learning_rate": 4.818713450292398e-06, | |
| "loss": 1.4493, | |
| "num_tokens": 755780.0, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 0.7687074829931972, | |
| "grad_norm": 22.29022789001465, | |
| "learning_rate": 4.7953216374269005e-06, | |
| "loss": 1.4572, | |
| "num_tokens": 756845.0, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 0.7698412698412699, | |
| "grad_norm": 23.95270538330078, | |
| "learning_rate": 4.771929824561404e-06, | |
| "loss": 1.4394, | |
| "num_tokens": 758014.0, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 0.7709750566893424, | |
| "grad_norm": 22.06515121459961, | |
| "learning_rate": 4.7485380116959065e-06, | |
| "loss": 1.5034, | |
| "num_tokens": 759116.0, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.7721088435374149, | |
| "grad_norm": 21.582685470581055, | |
| "learning_rate": 4.72514619883041e-06, | |
| "loss": 1.4249, | |
| "num_tokens": 760209.0, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 0.7732426303854876, | |
| "grad_norm": 23.25868797302246, | |
| "learning_rate": 4.7017543859649125e-06, | |
| "loss": 1.4951, | |
| "num_tokens": 761296.0, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 0.7743764172335601, | |
| "grad_norm": 22.286924362182617, | |
| "learning_rate": 4.678362573099415e-06, | |
| "loss": 1.473, | |
| "num_tokens": 762380.0, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 0.7755102040816326, | |
| "grad_norm": 22.370563507080078, | |
| "learning_rate": 4.6549707602339184e-06, | |
| "loss": 1.3724, | |
| "num_tokens": 763416.0, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 0.7766439909297053, | |
| "grad_norm": 22.624408721923828, | |
| "learning_rate": 4.631578947368421e-06, | |
| "loss": 1.3322, | |
| "num_tokens": 764519.0, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.7777777777777778, | |
| "grad_norm": 22.60655403137207, | |
| "learning_rate": 4.6081871345029244e-06, | |
| "loss": 1.4132, | |
| "num_tokens": 765685.0, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 0.7789115646258503, | |
| "grad_norm": 22.81845474243164, | |
| "learning_rate": 4.584795321637428e-06, | |
| "loss": 1.4029, | |
| "num_tokens": 766859.0, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 0.780045351473923, | |
| "grad_norm": 22.1308536529541, | |
| "learning_rate": 4.56140350877193e-06, | |
| "loss": 1.3881, | |
| "num_tokens": 768070.0, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 0.7811791383219955, | |
| "grad_norm": 23.998682022094727, | |
| "learning_rate": 4.538011695906433e-06, | |
| "loss": 1.5834, | |
| "num_tokens": 769163.0, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 0.782312925170068, | |
| "grad_norm": 23.69707489013672, | |
| "learning_rate": 4.5146198830409364e-06, | |
| "loss": 1.5529, | |
| "num_tokens": 770331.0, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.7834467120181405, | |
| "grad_norm": 24.159103393554688, | |
| "learning_rate": 4.491228070175439e-06, | |
| "loss": 1.3963, | |
| "num_tokens": 771348.0, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 0.7845804988662132, | |
| "grad_norm": 22.975770950317383, | |
| "learning_rate": 4.467836257309942e-06, | |
| "loss": 1.3646, | |
| "num_tokens": 772423.0, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 0.7857142857142857, | |
| "grad_norm": 24.724035263061523, | |
| "learning_rate": 4.444444444444444e-06, | |
| "loss": 1.3742, | |
| "num_tokens": 773491.0, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 0.7868480725623582, | |
| "grad_norm": 23.44858741760254, | |
| "learning_rate": 4.4210526315789476e-06, | |
| "loss": 1.5088, | |
| "num_tokens": 774638.0, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 0.7879818594104309, | |
| "grad_norm": 23.948352813720703, | |
| "learning_rate": 4.397660818713451e-06, | |
| "loss": 1.7207, | |
| "num_tokens": 775775.0, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.7891156462585034, | |
| "grad_norm": 21.847509384155273, | |
| "learning_rate": 4.3742690058479536e-06, | |
| "loss": 1.5141, | |
| "num_tokens": 776879.0, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 0.7902494331065759, | |
| "grad_norm": 23.483224868774414, | |
| "learning_rate": 4.350877192982457e-06, | |
| "loss": 1.5361, | |
| "num_tokens": 778025.0, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 0.7913832199546486, | |
| "grad_norm": 22.497262954711914, | |
| "learning_rate": 4.3274853801169596e-06, | |
| "loss": 1.5269, | |
| "num_tokens": 779182.0, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 0.7925170068027211, | |
| "grad_norm": 24.064851760864258, | |
| "learning_rate": 4.304093567251462e-06, | |
| "loss": 1.4504, | |
| "num_tokens": 780292.0, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 0.7936507936507936, | |
| "grad_norm": 25.360252380371094, | |
| "learning_rate": 4.2807017543859656e-06, | |
| "loss": 1.358, | |
| "num_tokens": 781293.0, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.7947845804988662, | |
| "grad_norm": 22.048860549926758, | |
| "learning_rate": 4.257309941520468e-06, | |
| "loss": 1.4928, | |
| "num_tokens": 782489.0, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 0.7959183673469388, | |
| "grad_norm": 25.04562759399414, | |
| "learning_rate": 4.2339181286549715e-06, | |
| "loss": 1.4883, | |
| "num_tokens": 783575.0, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 0.7970521541950113, | |
| "grad_norm": 23.770830154418945, | |
| "learning_rate": 4.210526315789474e-06, | |
| "loss": 1.512, | |
| "num_tokens": 784656.0, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 0.7981859410430839, | |
| "grad_norm": 22.704742431640625, | |
| "learning_rate": 4.187134502923977e-06, | |
| "loss": 1.3788, | |
| "num_tokens": 785780.0, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 0.7993197278911565, | |
| "grad_norm": 21.932666778564453, | |
| "learning_rate": 4.16374269005848e-06, | |
| "loss": 1.381, | |
| "num_tokens": 786993.0, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.800453514739229, | |
| "grad_norm": 22.353124618530273, | |
| "learning_rate": 4.140350877192983e-06, | |
| "loss": 1.3069, | |
| "num_tokens": 788042.0, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 0.8015873015873016, | |
| "grad_norm": 23.022003173828125, | |
| "learning_rate": 4.116959064327486e-06, | |
| "loss": 1.5089, | |
| "num_tokens": 789182.0, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 0.8027210884353742, | |
| "grad_norm": 22.555904388427734, | |
| "learning_rate": 4.093567251461989e-06, | |
| "loss": 1.4996, | |
| "num_tokens": 790326.0, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 0.8038548752834467, | |
| "grad_norm": 22.719186782836914, | |
| "learning_rate": 4.070175438596491e-06, | |
| "loss": 1.5449, | |
| "num_tokens": 791473.0, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 0.8049886621315193, | |
| "grad_norm": 21.924585342407227, | |
| "learning_rate": 4.046783625730995e-06, | |
| "loss": 1.3993, | |
| "num_tokens": 792646.0, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.8061224489795918, | |
| "grad_norm": 20.699281692504883, | |
| "learning_rate": 4.023391812865497e-06, | |
| "loss": 1.4973, | |
| "num_tokens": 793797.0, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 0.8072562358276644, | |
| "grad_norm": 22.318761825561523, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.4515, | |
| "num_tokens": 794861.0, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 0.808390022675737, | |
| "grad_norm": 21.21175765991211, | |
| "learning_rate": 3.976608187134503e-06, | |
| "loss": 1.3681, | |
| "num_tokens": 796010.0, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 0.8095238095238095, | |
| "grad_norm": 22.589601516723633, | |
| "learning_rate": 3.953216374269006e-06, | |
| "loss": 1.4962, | |
| "num_tokens": 797073.0, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 0.8106575963718821, | |
| "grad_norm": 21.697351455688477, | |
| "learning_rate": 3.929824561403509e-06, | |
| "loss": 1.4605, | |
| "num_tokens": 798194.0, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.8117913832199547, | |
| "grad_norm": 21.772241592407227, | |
| "learning_rate": 3.906432748538012e-06, | |
| "loss": 1.4382, | |
| "num_tokens": 799326.0, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 0.8129251700680272, | |
| "grad_norm": 21.598237991333008, | |
| "learning_rate": 3.883040935672515e-06, | |
| "loss": 1.4977, | |
| "num_tokens": 800462.0, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 0.8140589569160998, | |
| "grad_norm": 21.325088500976562, | |
| "learning_rate": 3.859649122807018e-06, | |
| "loss": 1.5207, | |
| "num_tokens": 801660.0, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 0.8151927437641724, | |
| "grad_norm": 24.253833770751953, | |
| "learning_rate": 3.83625730994152e-06, | |
| "loss": 1.4913, | |
| "num_tokens": 802681.0, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 0.8163265306122449, | |
| "grad_norm": 22.219438552856445, | |
| "learning_rate": 3.812865497076024e-06, | |
| "loss": 1.4802, | |
| "num_tokens": 803845.0, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.8174603174603174, | |
| "grad_norm": 22.79910659790039, | |
| "learning_rate": 3.789473684210527e-06, | |
| "loss": 1.4913, | |
| "num_tokens": 804919.0, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 0.81859410430839, | |
| "grad_norm": 22.189367294311523, | |
| "learning_rate": 3.7660818713450298e-06, | |
| "loss": 1.4382, | |
| "num_tokens": 806066.0, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 0.8197278911564626, | |
| "grad_norm": 21.064807891845703, | |
| "learning_rate": 3.7426900584795324e-06, | |
| "loss": 1.5483, | |
| "num_tokens": 807298.0, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 0.8208616780045351, | |
| "grad_norm": 22.164886474609375, | |
| "learning_rate": 3.7192982456140354e-06, | |
| "loss": 1.4852, | |
| "num_tokens": 808431.0, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 0.8219954648526077, | |
| "grad_norm": 22.567136764526367, | |
| "learning_rate": 3.6959064327485384e-06, | |
| "loss": 1.5195, | |
| "num_tokens": 809525.0, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.8231292517006803, | |
| "grad_norm": 22.86061668395996, | |
| "learning_rate": 3.6725146198830414e-06, | |
| "loss": 1.4455, | |
| "num_tokens": 810602.0, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 0.8242630385487528, | |
| "grad_norm": 24.17923927307129, | |
| "learning_rate": 3.6491228070175443e-06, | |
| "loss": 1.4668, | |
| "num_tokens": 811652.0, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 0.8253968253968254, | |
| "grad_norm": 23.038942337036133, | |
| "learning_rate": 3.625730994152047e-06, | |
| "loss": 1.5045, | |
| "num_tokens": 812716.0, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 0.826530612244898, | |
| "grad_norm": 21.556453704833984, | |
| "learning_rate": 3.60233918128655e-06, | |
| "loss": 1.3505, | |
| "num_tokens": 813814.0, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 0.8276643990929705, | |
| "grad_norm": 21.08284568786621, | |
| "learning_rate": 3.578947368421053e-06, | |
| "loss": 1.3864, | |
| "num_tokens": 814929.0, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.828798185941043, | |
| "grad_norm": 23.882568359375, | |
| "learning_rate": 3.555555555555556e-06, | |
| "loss": 1.3743, | |
| "num_tokens": 815979.0, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 0.8299319727891157, | |
| "grad_norm": 24.24713897705078, | |
| "learning_rate": 3.532163742690059e-06, | |
| "loss": 1.509, | |
| "num_tokens": 817290.0, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 0.8310657596371882, | |
| "grad_norm": 22.145536422729492, | |
| "learning_rate": 3.5087719298245615e-06, | |
| "loss": 1.501, | |
| "num_tokens": 818392.0, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 0.8321995464852607, | |
| "grad_norm": 21.552515029907227, | |
| "learning_rate": 3.4853801169590645e-06, | |
| "loss": 1.3933, | |
| "num_tokens": 819609.0, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 0.8333333333333334, | |
| "grad_norm": 25.56783676147461, | |
| "learning_rate": 3.4619883040935675e-06, | |
| "loss": 1.583, | |
| "num_tokens": 820656.0, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.8344671201814059, | |
| "grad_norm": 22.9251766204834, | |
| "learning_rate": 3.4385964912280705e-06, | |
| "loss": 1.3717, | |
| "num_tokens": 821762.0, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 0.8356009070294784, | |
| "grad_norm": 24.673242568969727, | |
| "learning_rate": 3.4152046783625735e-06, | |
| "loss": 1.5318, | |
| "num_tokens": 822936.0, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 0.8367346938775511, | |
| "grad_norm": 22.047882080078125, | |
| "learning_rate": 3.391812865497076e-06, | |
| "loss": 1.4858, | |
| "num_tokens": 824105.0, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 0.8378684807256236, | |
| "grad_norm": 26.981224060058594, | |
| "learning_rate": 3.368421052631579e-06, | |
| "loss": 1.6137, | |
| "num_tokens": 825116.0, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 0.8390022675736961, | |
| "grad_norm": 24.74795913696289, | |
| "learning_rate": 3.345029239766082e-06, | |
| "loss": 1.4405, | |
| "num_tokens": 826109.0, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.8401360544217688, | |
| "grad_norm": 23.003847122192383, | |
| "learning_rate": 3.321637426900585e-06, | |
| "loss": 1.4222, | |
| "num_tokens": 827151.0, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 0.8412698412698413, | |
| "grad_norm": 25.1625919342041, | |
| "learning_rate": 3.2982456140350885e-06, | |
| "loss": 1.538, | |
| "num_tokens": 828251.0, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 0.8424036281179138, | |
| "grad_norm": 23.558082580566406, | |
| "learning_rate": 3.2748538011695906e-06, | |
| "loss": 1.2967, | |
| "num_tokens": 829406.0, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 0.8435374149659864, | |
| "grad_norm": 22.346006393432617, | |
| "learning_rate": 3.2514619883040936e-06, | |
| "loss": 1.3927, | |
| "num_tokens": 830517.0, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 0.844671201814059, | |
| "grad_norm": 21.72648048400879, | |
| "learning_rate": 3.2280701754385966e-06, | |
| "loss": 1.5264, | |
| "num_tokens": 831751.0, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.8458049886621315, | |
| "grad_norm": 23.65290641784668, | |
| "learning_rate": 3.2046783625731e-06, | |
| "loss": 1.5314, | |
| "num_tokens": 832836.0, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 0.8469387755102041, | |
| "grad_norm": 22.38150978088379, | |
| "learning_rate": 3.181286549707603e-06, | |
| "loss": 1.6097, | |
| "num_tokens": 834005.0, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 0.8480725623582767, | |
| "grad_norm": 23.84773826599121, | |
| "learning_rate": 3.157894736842105e-06, | |
| "loss": 1.422, | |
| "num_tokens": 835035.0, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 0.8492063492063492, | |
| "grad_norm": 23.476564407348633, | |
| "learning_rate": 3.1345029239766086e-06, | |
| "loss": 1.4523, | |
| "num_tokens": 836174.0, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 0.8503401360544217, | |
| "grad_norm": 24.975629806518555, | |
| "learning_rate": 3.1111111111111116e-06, | |
| "loss": 1.4704, | |
| "num_tokens": 837345.0, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.8514739229024944, | |
| "grad_norm": 25.92818260192871, | |
| "learning_rate": 3.0877192982456146e-06, | |
| "loss": 1.354, | |
| "num_tokens": 838470.0, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 0.8526077097505669, | |
| "grad_norm": 21.76276969909668, | |
| "learning_rate": 3.0643274853801176e-06, | |
| "loss": 1.2758, | |
| "num_tokens": 839542.0, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 0.8537414965986394, | |
| "grad_norm": 23.923139572143555, | |
| "learning_rate": 3.04093567251462e-06, | |
| "loss": 1.5039, | |
| "num_tokens": 840615.0, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 0.854875283446712, | |
| "grad_norm": 20.571720123291016, | |
| "learning_rate": 3.017543859649123e-06, | |
| "loss": 1.5064, | |
| "num_tokens": 841800.0, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 0.8560090702947846, | |
| "grad_norm": 22.663013458251953, | |
| "learning_rate": 2.994152046783626e-06, | |
| "loss": 1.5114, | |
| "num_tokens": 842943.0, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 22.52815055847168, | |
| "learning_rate": 2.970760233918129e-06, | |
| "loss": 1.5805, | |
| "num_tokens": 844094.0, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 0.8582766439909297, | |
| "grad_norm": 22.185199737548828, | |
| "learning_rate": 2.9473684210526317e-06, | |
| "loss": 1.4734, | |
| "num_tokens": 845227.0, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 0.8594104308390023, | |
| "grad_norm": 21.498750686645508, | |
| "learning_rate": 2.9239766081871347e-06, | |
| "loss": 1.3279, | |
| "num_tokens": 846379.0, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 0.8605442176870748, | |
| "grad_norm": 22.456575393676758, | |
| "learning_rate": 2.9005847953216377e-06, | |
| "loss": 1.3714, | |
| "num_tokens": 847478.0, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 0.8616780045351474, | |
| "grad_norm": 25.112070083618164, | |
| "learning_rate": 2.8771929824561407e-06, | |
| "loss": 1.5006, | |
| "num_tokens": 848561.0, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.86281179138322, | |
| "grad_norm": 21.80204963684082, | |
| "learning_rate": 2.8538011695906437e-06, | |
| "loss": 1.4335, | |
| "num_tokens": 849697.0, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 0.8639455782312925, | |
| "grad_norm": 23.306922912597656, | |
| "learning_rate": 2.8304093567251463e-06, | |
| "loss": 1.4661, | |
| "num_tokens": 850827.0, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 0.8650793650793651, | |
| "grad_norm": 22.614166259765625, | |
| "learning_rate": 2.8070175438596493e-06, | |
| "loss": 1.4819, | |
| "num_tokens": 851967.0, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 0.8662131519274376, | |
| "grad_norm": 22.74590301513672, | |
| "learning_rate": 2.7836257309941523e-06, | |
| "loss": 1.4953, | |
| "num_tokens": 853111.0, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 0.8673469387755102, | |
| "grad_norm": 22.92290687561035, | |
| "learning_rate": 2.7602339181286553e-06, | |
| "loss": 1.4969, | |
| "num_tokens": 854225.0, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.8684807256235828, | |
| "grad_norm": 21.96657371520996, | |
| "learning_rate": 2.7368421052631583e-06, | |
| "loss": 1.5487, | |
| "num_tokens": 855400.0, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 0.8696145124716553, | |
| "grad_norm": 22.646835327148438, | |
| "learning_rate": 2.713450292397661e-06, | |
| "loss": 1.391, | |
| "num_tokens": 856571.0, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 0.8707482993197279, | |
| "grad_norm": 22.9639892578125, | |
| "learning_rate": 2.690058479532164e-06, | |
| "loss": 1.4171, | |
| "num_tokens": 857782.0, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 0.8718820861678005, | |
| "grad_norm": 23.083837509155273, | |
| "learning_rate": 2.666666666666667e-06, | |
| "loss": 1.5668, | |
| "num_tokens": 858959.0, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 0.873015873015873, | |
| "grad_norm": 22.602506637573242, | |
| "learning_rate": 2.64327485380117e-06, | |
| "loss": 1.4862, | |
| "num_tokens": 860124.0, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.8741496598639455, | |
| "grad_norm": 23.74089241027832, | |
| "learning_rate": 2.619883040935673e-06, | |
| "loss": 1.4123, | |
| "num_tokens": 861312.0, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 0.8752834467120182, | |
| "grad_norm": 22.073772430419922, | |
| "learning_rate": 2.5964912280701754e-06, | |
| "loss": 1.4083, | |
| "num_tokens": 862379.0, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 0.8764172335600907, | |
| "grad_norm": 23.299543380737305, | |
| "learning_rate": 2.5730994152046784e-06, | |
| "loss": 1.3877, | |
| "num_tokens": 863526.0, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 0.8775510204081632, | |
| "grad_norm": 23.912397384643555, | |
| "learning_rate": 2.5497076023391814e-06, | |
| "loss": 1.6234, | |
| "num_tokens": 864663.0, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 0.8786848072562359, | |
| "grad_norm": 24.13743019104004, | |
| "learning_rate": 2.5263157894736844e-06, | |
| "loss": 1.6174, | |
| "num_tokens": 865755.0, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.8798185941043084, | |
| "grad_norm": 23.58803939819336, | |
| "learning_rate": 2.5029239766081874e-06, | |
| "loss": 1.5354, | |
| "num_tokens": 866867.0, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 0.8809523809523809, | |
| "grad_norm": 22.72237205505371, | |
| "learning_rate": 2.4795321637426904e-06, | |
| "loss": 1.4695, | |
| "num_tokens": 867947.0, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 0.8820861678004536, | |
| "grad_norm": 23.858057022094727, | |
| "learning_rate": 2.456140350877193e-06, | |
| "loss": 1.4492, | |
| "num_tokens": 869047.0, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 0.8832199546485261, | |
| "grad_norm": 21.552154541015625, | |
| "learning_rate": 2.432748538011696e-06, | |
| "loss": 1.4438, | |
| "num_tokens": 870112.0, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 0.8843537414965986, | |
| "grad_norm": 24.152172088623047, | |
| "learning_rate": 2.409356725146199e-06, | |
| "loss": 1.4744, | |
| "num_tokens": 871206.0, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.8854875283446711, | |
| "grad_norm": 22.032583236694336, | |
| "learning_rate": 2.385964912280702e-06, | |
| "loss": 1.3994, | |
| "num_tokens": 872313.0, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 0.8866213151927438, | |
| "grad_norm": 22.46315574645996, | |
| "learning_rate": 2.362573099415205e-06, | |
| "loss": 1.4018, | |
| "num_tokens": 873407.0, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 0.8877551020408163, | |
| "grad_norm": 22.419082641601562, | |
| "learning_rate": 2.3391812865497075e-06, | |
| "loss": 1.3361, | |
| "num_tokens": 874507.0, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 22.973846435546875, | |
| "learning_rate": 2.3157894736842105e-06, | |
| "loss": 1.5251, | |
| "num_tokens": 875594.0, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 0.8900226757369615, | |
| "grad_norm": 21.73171043395996, | |
| "learning_rate": 2.292397660818714e-06, | |
| "loss": 1.4877, | |
| "num_tokens": 876759.0, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.891156462585034, | |
| "grad_norm": 22.867074966430664, | |
| "learning_rate": 2.2690058479532165e-06, | |
| "loss": 1.4999, | |
| "num_tokens": 877790.0, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 0.8922902494331065, | |
| "grad_norm": 23.208694458007812, | |
| "learning_rate": 2.2456140350877195e-06, | |
| "loss": 1.4502, | |
| "num_tokens": 878974.0, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 0.8934240362811792, | |
| "grad_norm": 23.5717830657959, | |
| "learning_rate": 2.222222222222222e-06, | |
| "loss": 1.3898, | |
| "num_tokens": 880052.0, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 0.8945578231292517, | |
| "grad_norm": 23.512678146362305, | |
| "learning_rate": 2.1988304093567255e-06, | |
| "loss": 1.4121, | |
| "num_tokens": 881239.0, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 0.8956916099773242, | |
| "grad_norm": 23.14679527282715, | |
| "learning_rate": 2.1754385964912285e-06, | |
| "loss": 1.4851, | |
| "num_tokens": 882281.0, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.8968253968253969, | |
| "grad_norm": 23.918848037719727, | |
| "learning_rate": 2.152046783625731e-06, | |
| "loss": 1.4963, | |
| "num_tokens": 883346.0, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 0.8979591836734694, | |
| "grad_norm": 26.56317710876465, | |
| "learning_rate": 2.128654970760234e-06, | |
| "loss": 1.5296, | |
| "num_tokens": 884352.0, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 0.8990929705215419, | |
| "grad_norm": 22.81806182861328, | |
| "learning_rate": 2.105263157894737e-06, | |
| "loss": 1.5657, | |
| "num_tokens": 885502.0, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 0.9002267573696145, | |
| "grad_norm": 23.83610725402832, | |
| "learning_rate": 2.08187134502924e-06, | |
| "loss": 1.5682, | |
| "num_tokens": 886626.0, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 0.9013605442176871, | |
| "grad_norm": 23.975093841552734, | |
| "learning_rate": 2.058479532163743e-06, | |
| "loss": 1.527, | |
| "num_tokens": 887757.0, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.9024943310657596, | |
| "grad_norm": 21.868783950805664, | |
| "learning_rate": 2.0350877192982456e-06, | |
| "loss": 1.4765, | |
| "num_tokens": 888939.0, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 0.9036281179138322, | |
| "grad_norm": 22.76094627380371, | |
| "learning_rate": 2.0116959064327486e-06, | |
| "loss": 1.3386, | |
| "num_tokens": 890157.0, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 0.9047619047619048, | |
| "grad_norm": 23.086637496948242, | |
| "learning_rate": 1.9883040935672516e-06, | |
| "loss": 1.6715, | |
| "num_tokens": 891398.0, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 0.9058956916099773, | |
| "grad_norm": 24.359127044677734, | |
| "learning_rate": 1.9649122807017546e-06, | |
| "loss": 1.3054, | |
| "num_tokens": 892408.0, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 0.9070294784580499, | |
| "grad_norm": 25.446590423583984, | |
| "learning_rate": 1.9415204678362576e-06, | |
| "loss": 1.4722, | |
| "num_tokens": 893592.0, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.9081632653061225, | |
| "grad_norm": 23.30593490600586, | |
| "learning_rate": 1.91812865497076e-06, | |
| "loss": 1.4158, | |
| "num_tokens": 894697.0, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 0.909297052154195, | |
| "grad_norm": 21.06711769104004, | |
| "learning_rate": 1.8947368421052634e-06, | |
| "loss": 1.592, | |
| "num_tokens": 895818.0, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 0.9104308390022676, | |
| "grad_norm": 23.522247314453125, | |
| "learning_rate": 1.8713450292397662e-06, | |
| "loss": 1.5568, | |
| "num_tokens": 896885.0, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 0.9115646258503401, | |
| "grad_norm": 22.43680763244629, | |
| "learning_rate": 1.8479532163742692e-06, | |
| "loss": 1.3984, | |
| "num_tokens": 898032.0, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 0.9126984126984127, | |
| "grad_norm": 22.576719284057617, | |
| "learning_rate": 1.8245614035087722e-06, | |
| "loss": 1.4604, | |
| "num_tokens": 899163.0, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.9138321995464853, | |
| "grad_norm": 22.583637237548828, | |
| "learning_rate": 1.801169590643275e-06, | |
| "loss": 1.4348, | |
| "num_tokens": 900304.0, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 0.9149659863945578, | |
| "grad_norm": 21.52495574951172, | |
| "learning_rate": 1.777777777777778e-06, | |
| "loss": 1.364, | |
| "num_tokens": 901464.0, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 0.9160997732426304, | |
| "grad_norm": 22.536762237548828, | |
| "learning_rate": 1.7543859649122807e-06, | |
| "loss": 1.5089, | |
| "num_tokens": 902652.0, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 0.9172335600907029, | |
| "grad_norm": 22.26384162902832, | |
| "learning_rate": 1.7309941520467837e-06, | |
| "loss": 1.4325, | |
| "num_tokens": 903725.0, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 0.9183673469387755, | |
| "grad_norm": 20.824234008789062, | |
| "learning_rate": 1.7076023391812867e-06, | |
| "loss": 1.396, | |
| "num_tokens": 904819.0, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.9195011337868481, | |
| "grad_norm": 21.5470027923584, | |
| "learning_rate": 1.6842105263157895e-06, | |
| "loss": 1.4437, | |
| "num_tokens": 905994.0, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 0.9206349206349206, | |
| "grad_norm": 22.454166412353516, | |
| "learning_rate": 1.6608187134502925e-06, | |
| "loss": 1.4207, | |
| "num_tokens": 907165.0, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 0.9217687074829932, | |
| "grad_norm": 22.63362693786621, | |
| "learning_rate": 1.6374269005847953e-06, | |
| "loss": 1.5246, | |
| "num_tokens": 908315.0, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 0.9229024943310657, | |
| "grad_norm": 25.17198371887207, | |
| "learning_rate": 1.6140350877192983e-06, | |
| "loss": 1.527, | |
| "num_tokens": 909504.0, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 0.9240362811791383, | |
| "grad_norm": 25.855079650878906, | |
| "learning_rate": 1.5906432748538015e-06, | |
| "loss": 1.6569, | |
| "num_tokens": 910579.0, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.9251700680272109, | |
| "grad_norm": 22.92259979248047, | |
| "learning_rate": 1.5672514619883043e-06, | |
| "loss": 1.4674, | |
| "num_tokens": 911639.0, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 0.9263038548752834, | |
| "grad_norm": 24.303924560546875, | |
| "learning_rate": 1.5438596491228073e-06, | |
| "loss": 1.5105, | |
| "num_tokens": 912701.0, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 0.927437641723356, | |
| "grad_norm": 21.863985061645508, | |
| "learning_rate": 1.52046783625731e-06, | |
| "loss": 1.6027, | |
| "num_tokens": 913766.0, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 0.9285714285714286, | |
| "grad_norm": 24.20486068725586, | |
| "learning_rate": 1.497076023391813e-06, | |
| "loss": 1.5517, | |
| "num_tokens": 914798.0, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 0.9297052154195011, | |
| "grad_norm": 24.408193588256836, | |
| "learning_rate": 1.4736842105263159e-06, | |
| "loss": 1.4336, | |
| "num_tokens": 915953.0, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.9308390022675737, | |
| "grad_norm": 22.91715431213379, | |
| "learning_rate": 1.4502923976608189e-06, | |
| "loss": 1.3284, | |
| "num_tokens": 917153.0, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 0.9319727891156463, | |
| "grad_norm": 23.430063247680664, | |
| "learning_rate": 1.4269005847953219e-06, | |
| "loss": 1.5508, | |
| "num_tokens": 918239.0, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 0.9331065759637188, | |
| "grad_norm": 24.82691764831543, | |
| "learning_rate": 1.4035087719298246e-06, | |
| "loss": 1.4022, | |
| "num_tokens": 919263.0, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 0.9342403628117913, | |
| "grad_norm": 21.78477668762207, | |
| "learning_rate": 1.3801169590643276e-06, | |
| "loss": 1.4777, | |
| "num_tokens": 920398.0, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 0.935374149659864, | |
| "grad_norm": 23.606664657592773, | |
| "learning_rate": 1.3567251461988304e-06, | |
| "loss": 1.479, | |
| "num_tokens": 921477.0, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.9365079365079365, | |
| "grad_norm": 22.557498931884766, | |
| "learning_rate": 1.3333333333333334e-06, | |
| "loss": 1.5736, | |
| "num_tokens": 922580.0, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 0.937641723356009, | |
| "grad_norm": 22.220712661743164, | |
| "learning_rate": 1.3099415204678364e-06, | |
| "loss": 1.4134, | |
| "num_tokens": 923731.0, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 0.9387755102040817, | |
| "grad_norm": 20.02456283569336, | |
| "learning_rate": 1.2865497076023392e-06, | |
| "loss": 1.4506, | |
| "num_tokens": 924867.0, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 0.9399092970521542, | |
| "grad_norm": 23.332258224487305, | |
| "learning_rate": 1.2631578947368422e-06, | |
| "loss": 1.3667, | |
| "num_tokens": 925946.0, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 0.9410430839002267, | |
| "grad_norm": 23.033876419067383, | |
| "learning_rate": 1.2397660818713452e-06, | |
| "loss": 1.4524, | |
| "num_tokens": 927211.0, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.9421768707482994, | |
| "grad_norm": 21.960102081298828, | |
| "learning_rate": 1.216374269005848e-06, | |
| "loss": 1.4102, | |
| "num_tokens": 928388.0, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 0.9433106575963719, | |
| "grad_norm": 22.837690353393555, | |
| "learning_rate": 1.192982456140351e-06, | |
| "loss": 1.4648, | |
| "num_tokens": 929505.0, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 0.9444444444444444, | |
| "grad_norm": 26.00808334350586, | |
| "learning_rate": 1.1695906432748538e-06, | |
| "loss": 1.3195, | |
| "num_tokens": 930491.0, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 0.9455782312925171, | |
| "grad_norm": 22.8687744140625, | |
| "learning_rate": 1.146198830409357e-06, | |
| "loss": 1.4152, | |
| "num_tokens": 931679.0, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 0.9467120181405896, | |
| "grad_norm": 23.170955657958984, | |
| "learning_rate": 1.1228070175438598e-06, | |
| "loss": 1.3588, | |
| "num_tokens": 932826.0, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.9478458049886621, | |
| "grad_norm": 21.579723358154297, | |
| "learning_rate": 1.0994152046783627e-06, | |
| "loss": 1.4296, | |
| "num_tokens": 933959.0, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 0.9489795918367347, | |
| "grad_norm": 22.338029861450195, | |
| "learning_rate": 1.0760233918128655e-06, | |
| "loss": 1.5036, | |
| "num_tokens": 935084.0, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 0.9501133786848073, | |
| "grad_norm": 24.644001007080078, | |
| "learning_rate": 1.0526315789473685e-06, | |
| "loss": 1.5511, | |
| "num_tokens": 936207.0, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 0.9512471655328798, | |
| "grad_norm": 20.820331573486328, | |
| "learning_rate": 1.0292397660818715e-06, | |
| "loss": 1.4014, | |
| "num_tokens": 937398.0, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 0.9523809523809523, | |
| "grad_norm": 22.744985580444336, | |
| "learning_rate": 1.0058479532163743e-06, | |
| "loss": 1.5348, | |
| "num_tokens": 938566.0, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.953514739229025, | |
| "grad_norm": 21.92036247253418, | |
| "learning_rate": 9.824561403508773e-07, | |
| "loss": 1.5325, | |
| "num_tokens": 939691.0, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 0.9546485260770975, | |
| "grad_norm": 21.75764274597168, | |
| "learning_rate": 9.5906432748538e-07, | |
| "loss": 1.468, | |
| "num_tokens": 940827.0, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 0.95578231292517, | |
| "grad_norm": 22.83157730102539, | |
| "learning_rate": 9.356725146198831e-07, | |
| "loss": 1.4777, | |
| "num_tokens": 941874.0, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 0.9569160997732427, | |
| "grad_norm": 24.487289428710938, | |
| "learning_rate": 9.122807017543861e-07, | |
| "loss": 1.2818, | |
| "num_tokens": 942976.0, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 0.9580498866213152, | |
| "grad_norm": 22.184720993041992, | |
| "learning_rate": 8.88888888888889e-07, | |
| "loss": 1.4944, | |
| "num_tokens": 944151.0, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.9591836734693877, | |
| "grad_norm": 24.632143020629883, | |
| "learning_rate": 8.654970760233919e-07, | |
| "loss": 1.4877, | |
| "num_tokens": 945188.0, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 0.9603174603174603, | |
| "grad_norm": 23.738849639892578, | |
| "learning_rate": 8.421052631578948e-07, | |
| "loss": 1.5023, | |
| "num_tokens": 946295.0, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 0.9614512471655329, | |
| "grad_norm": 23.87755012512207, | |
| "learning_rate": 8.187134502923977e-07, | |
| "loss": 1.5317, | |
| "num_tokens": 947340.0, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 0.9625850340136054, | |
| "grad_norm": 23.480918884277344, | |
| "learning_rate": 7.953216374269008e-07, | |
| "loss": 1.4858, | |
| "num_tokens": 948481.0, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 0.963718820861678, | |
| "grad_norm": 24.38697052001953, | |
| "learning_rate": 7.719298245614036e-07, | |
| "loss": 1.3701, | |
| "num_tokens": 949547.0, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.9648526077097506, | |
| "grad_norm": 23.775096893310547, | |
| "learning_rate": 7.485380116959065e-07, | |
| "loss": 1.4726, | |
| "num_tokens": 950588.0, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 0.9659863945578231, | |
| "grad_norm": 22.794851303100586, | |
| "learning_rate": 7.251461988304094e-07, | |
| "loss": 1.5002, | |
| "num_tokens": 951774.0, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 0.9671201814058957, | |
| "grad_norm": 21.191730499267578, | |
| "learning_rate": 7.017543859649123e-07, | |
| "loss": 1.4518, | |
| "num_tokens": 952871.0, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 0.9682539682539683, | |
| "grad_norm": 21.24921417236328, | |
| "learning_rate": 6.783625730994152e-07, | |
| "loss": 1.3679, | |
| "num_tokens": 954031.0, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 0.9693877551020408, | |
| "grad_norm": 23.635141372680664, | |
| "learning_rate": 6.549707602339182e-07, | |
| "loss": 1.3911, | |
| "num_tokens": 955054.0, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.9705215419501134, | |
| "grad_norm": 23.66080665588379, | |
| "learning_rate": 6.315789473684211e-07, | |
| "loss": 1.3363, | |
| "num_tokens": 956174.0, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 0.971655328798186, | |
| "grad_norm": 23.329038619995117, | |
| "learning_rate": 6.08187134502924e-07, | |
| "loss": 1.514, | |
| "num_tokens": 957337.0, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 0.9727891156462585, | |
| "grad_norm": 22.854223251342773, | |
| "learning_rate": 5.847953216374269e-07, | |
| "loss": 1.4398, | |
| "num_tokens": 958396.0, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 0.9739229024943311, | |
| "grad_norm": 22.29954719543457, | |
| "learning_rate": 5.614035087719299e-07, | |
| "loss": 1.5289, | |
| "num_tokens": 959510.0, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 0.9750566893424036, | |
| "grad_norm": 23.53312873840332, | |
| "learning_rate": 5.380116959064328e-07, | |
| "loss": 1.4513, | |
| "num_tokens": 960614.0, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.9761904761904762, | |
| "grad_norm": 21.8758602142334, | |
| "learning_rate": 5.146198830409358e-07, | |
| "loss": 1.422, | |
| "num_tokens": 961707.0, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 0.9773242630385488, | |
| "grad_norm": 23.302278518676758, | |
| "learning_rate": 4.912280701754387e-07, | |
| "loss": 1.4902, | |
| "num_tokens": 962890.0, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 0.9784580498866213, | |
| "grad_norm": 23.379724502563477, | |
| "learning_rate": 4.6783625730994155e-07, | |
| "loss": 1.4964, | |
| "num_tokens": 964107.0, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 0.9795918367346939, | |
| "grad_norm": 23.007959365844727, | |
| "learning_rate": 4.444444444444445e-07, | |
| "loss": 1.4419, | |
| "num_tokens": 965201.0, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 0.9807256235827665, | |
| "grad_norm": 24.184219360351562, | |
| "learning_rate": 4.210526315789474e-07, | |
| "loss": 1.3729, | |
| "num_tokens": 966250.0, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.981859410430839, | |
| "grad_norm": 22.903079986572266, | |
| "learning_rate": 3.976608187134504e-07, | |
| "loss": 1.4376, | |
| "num_tokens": 967412.0, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 0.9829931972789115, | |
| "grad_norm": 21.505783081054688, | |
| "learning_rate": 3.7426900584795327e-07, | |
| "loss": 1.4338, | |
| "num_tokens": 968670.0, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 0.9841269841269841, | |
| "grad_norm": 24.10959243774414, | |
| "learning_rate": 3.5087719298245616e-07, | |
| "loss": 1.3635, | |
| "num_tokens": 969688.0, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 0.9852607709750567, | |
| "grad_norm": 23.527223587036133, | |
| "learning_rate": 3.274853801169591e-07, | |
| "loss": 1.5074, | |
| "num_tokens": 970730.0, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 0.9863945578231292, | |
| "grad_norm": 22.13941764831543, | |
| "learning_rate": 3.04093567251462e-07, | |
| "loss": 1.4279, | |
| "num_tokens": 971851.0, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.9875283446712018, | |
| "grad_norm": 23.4210262298584, | |
| "learning_rate": 2.8070175438596494e-07, | |
| "loss": 1.4491, | |
| "num_tokens": 972955.0, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 0.9886621315192744, | |
| "grad_norm": 22.403488159179688, | |
| "learning_rate": 2.573099415204679e-07, | |
| "loss": 1.4925, | |
| "num_tokens": 974119.0, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 0.9897959183673469, | |
| "grad_norm": 22.045902252197266, | |
| "learning_rate": 2.3391812865497077e-07, | |
| "loss": 1.5381, | |
| "num_tokens": 975268.0, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 0.9909297052154195, | |
| "grad_norm": 22.418025970458984, | |
| "learning_rate": 2.105263157894737e-07, | |
| "loss": 1.3585, | |
| "num_tokens": 976366.0, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 0.9920634920634921, | |
| "grad_norm": 24.32988739013672, | |
| "learning_rate": 1.8713450292397663e-07, | |
| "loss": 1.3053, | |
| "num_tokens": 977432.0, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.9931972789115646, | |
| "grad_norm": 22.723827362060547, | |
| "learning_rate": 1.6374269005847955e-07, | |
| "loss": 1.461, | |
| "num_tokens": 978586.0, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 0.9943310657596371, | |
| "grad_norm": 22.09205436706543, | |
| "learning_rate": 1.4035087719298247e-07, | |
| "loss": 1.396, | |
| "num_tokens": 979670.0, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 0.9954648526077098, | |
| "grad_norm": 24.673280715942383, | |
| "learning_rate": 1.1695906432748539e-07, | |
| "loss": 1.3698, | |
| "num_tokens": 980667.0, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 0.9965986394557823, | |
| "grad_norm": 22.821020126342773, | |
| "learning_rate": 9.356725146198832e-08, | |
| "loss": 1.4921, | |
| "num_tokens": 981825.0, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 0.9977324263038548, | |
| "grad_norm": 23.99849510192871, | |
| "learning_rate": 7.017543859649123e-08, | |
| "loss": 1.5141, | |
| "num_tokens": 982883.0, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.9988662131519275, | |
| "grad_norm": 22.911663055419922, | |
| "learning_rate": 4.678362573099416e-08, | |
| "loss": 1.4862, | |
| "num_tokens": 983976.0, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 23.910690307617188, | |
| "learning_rate": 2.339181286549708e-08, | |
| "loss": 1.367, | |
| "num_tokens": 984517.0, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 882, | |
| "total_flos": 5752804923670528.0, | |
| "train_loss": 1.5323709486022827, | |
| "train_runtime": 132.5784, | |
| "train_samples_per_second": 53.169, | |
| "train_steps_per_second": 6.653 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 882, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5752804923670528.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |