| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.9975757575757576, | |
| "eval_steps": 123, | |
| "global_step": 1236, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0016161616161616162, | |
| "grad_norm": 280.0, | |
| "learning_rate": 8.064516129032259e-08, | |
| "loss": 4.116, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0032323232323232323, | |
| "grad_norm": 268.0, | |
| "learning_rate": 1.6129032258064518e-07, | |
| "loss": 4.16, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0048484848484848485, | |
| "grad_norm": 316.0, | |
| "learning_rate": 2.4193548387096775e-07, | |
| "loss": 4.0967, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.006464646464646465, | |
| "grad_norm": 280.0, | |
| "learning_rate": 3.2258064516129035e-07, | |
| "loss": 4.1925, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.00808080808080808, | |
| "grad_norm": 300.0, | |
| "learning_rate": 4.032258064516129e-07, | |
| "loss": 4.0364, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.009696969696969697, | |
| "grad_norm": 290.0, | |
| "learning_rate": 4.838709677419355e-07, | |
| "loss": 3.9787, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.011313131313131313, | |
| "grad_norm": 272.0, | |
| "learning_rate": 5.645161290322581e-07, | |
| "loss": 3.9483, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.01292929292929293, | |
| "grad_norm": 212.0, | |
| "learning_rate": 6.451612903225807e-07, | |
| "loss": 3.8711, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.014545454545454545, | |
| "grad_norm": 195.0, | |
| "learning_rate": 7.258064516129033e-07, | |
| "loss": 3.6116, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.01616161616161616, | |
| "grad_norm": 183.0, | |
| "learning_rate": 8.064516129032258e-07, | |
| "loss": 3.6654, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.017777777777777778, | |
| "grad_norm": 151.0, | |
| "learning_rate": 8.870967741935485e-07, | |
| "loss": 3.3067, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.019393939393939394, | |
| "grad_norm": 130.0, | |
| "learning_rate": 9.67741935483871e-07, | |
| "loss": 3.2813, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.02101010101010101, | |
| "grad_norm": 126.0, | |
| "learning_rate": 1.0483870967741936e-06, | |
| "loss": 3.1032, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.022626262626262626, | |
| "grad_norm": 94.5, | |
| "learning_rate": 1.1290322580645162e-06, | |
| "loss": 3.2868, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.024242424242424242, | |
| "grad_norm": 98.5, | |
| "learning_rate": 1.2096774193548388e-06, | |
| "loss": 2.9182, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.02585858585858586, | |
| "grad_norm": 82.0, | |
| "learning_rate": 1.2903225806451614e-06, | |
| "loss": 2.9145, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.027474747474747475, | |
| "grad_norm": 98.5, | |
| "learning_rate": 1.3709677419354838e-06, | |
| "loss": 2.7479, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.02909090909090909, | |
| "grad_norm": 94.5, | |
| "learning_rate": 1.4516129032258066e-06, | |
| "loss": 2.5529, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.030707070707070707, | |
| "grad_norm": 71.0, | |
| "learning_rate": 1.5322580645161292e-06, | |
| "loss": 2.6872, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.03232323232323232, | |
| "grad_norm": 63.25, | |
| "learning_rate": 1.6129032258064516e-06, | |
| "loss": 2.3578, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.03393939393939394, | |
| "grad_norm": 76.0, | |
| "learning_rate": 1.6935483870967742e-06, | |
| "loss": 2.2456, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.035555555555555556, | |
| "grad_norm": 111.5, | |
| "learning_rate": 1.774193548387097e-06, | |
| "loss": 2.0884, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.037171717171717175, | |
| "grad_norm": 49.5, | |
| "learning_rate": 1.8548387096774196e-06, | |
| "loss": 2.1233, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.03878787878787879, | |
| "grad_norm": 47.75, | |
| "learning_rate": 1.935483870967742e-06, | |
| "loss": 1.8103, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.04040404040404041, | |
| "grad_norm": 49.5, | |
| "learning_rate": 2.0161290322580646e-06, | |
| "loss": 1.6413, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.04202020202020202, | |
| "grad_norm": 45.75, | |
| "learning_rate": 2.096774193548387e-06, | |
| "loss": 2.0003, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.04363636363636364, | |
| "grad_norm": 43.5, | |
| "learning_rate": 2.17741935483871e-06, | |
| "loss": 1.7371, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.04525252525252525, | |
| "grad_norm": 40.0, | |
| "learning_rate": 2.2580645161290324e-06, | |
| "loss": 1.6431, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.04686868686868687, | |
| "grad_norm": 33.75, | |
| "learning_rate": 2.338709677419355e-06, | |
| "loss": 1.7026, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.048484848484848485, | |
| "grad_norm": 38.5, | |
| "learning_rate": 2.4193548387096776e-06, | |
| "loss": 1.6614, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.050101010101010104, | |
| "grad_norm": 33.25, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.6833, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.05171717171717172, | |
| "grad_norm": 34.75, | |
| "learning_rate": 2.580645161290323e-06, | |
| "loss": 1.3788, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.05333333333333334, | |
| "grad_norm": 30.5, | |
| "learning_rate": 2.6612903225806454e-06, | |
| "loss": 1.4505, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.05494949494949495, | |
| "grad_norm": 34.5, | |
| "learning_rate": 2.7419354838709676e-06, | |
| "loss": 1.3686, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.05656565656565657, | |
| "grad_norm": 29.125, | |
| "learning_rate": 2.822580645161291e-06, | |
| "loss": 1.5094, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.05818181818181818, | |
| "grad_norm": 27.625, | |
| "learning_rate": 2.903225806451613e-06, | |
| "loss": 1.1449, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.0597979797979798, | |
| "grad_norm": 26.0, | |
| "learning_rate": 2.983870967741936e-06, | |
| "loss": 1.2051, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.061414141414141414, | |
| "grad_norm": 30.5, | |
| "learning_rate": 3.0645161290322584e-06, | |
| "loss": 1.4705, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.06303030303030303, | |
| "grad_norm": 27.125, | |
| "learning_rate": 3.145161290322581e-06, | |
| "loss": 1.4405, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.06464646464646465, | |
| "grad_norm": 26.5, | |
| "learning_rate": 3.225806451612903e-06, | |
| "loss": 1.2536, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.06626262626262626, | |
| "grad_norm": 28.375, | |
| "learning_rate": 3.306451612903226e-06, | |
| "loss": 1.3682, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.06787878787878789, | |
| "grad_norm": 28.375, | |
| "learning_rate": 3.3870967741935484e-06, | |
| "loss": 1.2269, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.0694949494949495, | |
| "grad_norm": 25.75, | |
| "learning_rate": 3.4677419354838714e-06, | |
| "loss": 1.155, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.07111111111111111, | |
| "grad_norm": 23.75, | |
| "learning_rate": 3.548387096774194e-06, | |
| "loss": 1.3323, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.07272727272727272, | |
| "grad_norm": 27.125, | |
| "learning_rate": 3.6290322580645166e-06, | |
| "loss": 1.4642, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.07434343434343435, | |
| "grad_norm": 25.875, | |
| "learning_rate": 3.7096774193548392e-06, | |
| "loss": 1.1611, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.07595959595959596, | |
| "grad_norm": 25.375, | |
| "learning_rate": 3.7903225806451614e-06, | |
| "loss": 1.36, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.07757575757575758, | |
| "grad_norm": 25.5, | |
| "learning_rate": 3.870967741935484e-06, | |
| "loss": 1.463, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.07919191919191919, | |
| "grad_norm": 28.25, | |
| "learning_rate": 3.951612903225807e-06, | |
| "loss": 1.3702, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.08080808080808081, | |
| "grad_norm": 25.625, | |
| "learning_rate": 4.032258064516129e-06, | |
| "loss": 1.3247, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.08242424242424243, | |
| "grad_norm": 19.625, | |
| "learning_rate": 4.112903225806452e-06, | |
| "loss": 1.1308, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.08404040404040404, | |
| "grad_norm": 22.0, | |
| "learning_rate": 4.193548387096774e-06, | |
| "loss": 1.1602, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.08565656565656565, | |
| "grad_norm": 22.0, | |
| "learning_rate": 4.274193548387097e-06, | |
| "loss": 1.2107, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.08727272727272728, | |
| "grad_norm": 22.0, | |
| "learning_rate": 4.35483870967742e-06, | |
| "loss": 1.1575, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.08888888888888889, | |
| "grad_norm": 23.5, | |
| "learning_rate": 4.435483870967742e-06, | |
| "loss": 1.3044, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.0905050505050505, | |
| "grad_norm": 20.625, | |
| "learning_rate": 4.516129032258065e-06, | |
| "loss": 0.9626, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.09212121212121212, | |
| "grad_norm": 23.0, | |
| "learning_rate": 4.596774193548387e-06, | |
| "loss": 1.5338, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.09373737373737374, | |
| "grad_norm": 21.5, | |
| "learning_rate": 4.67741935483871e-06, | |
| "loss": 1.3822, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.09535353535353536, | |
| "grad_norm": 18.0, | |
| "learning_rate": 4.758064516129033e-06, | |
| "loss": 0.8616, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.09696969696969697, | |
| "grad_norm": 20.75, | |
| "learning_rate": 4.838709677419355e-06, | |
| "loss": 1.133, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.09858585858585858, | |
| "grad_norm": 21.125, | |
| "learning_rate": 4.919354838709678e-06, | |
| "loss": 1.2883, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.10020202020202021, | |
| "grad_norm": 16.875, | |
| "learning_rate": 5e-06, | |
| "loss": 1.0195, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.10181818181818182, | |
| "grad_norm": 16.25, | |
| "learning_rate": 5.080645161290323e-06, | |
| "loss": 1.1219, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.10343434343434343, | |
| "grad_norm": 16.75, | |
| "learning_rate": 5.161290322580646e-06, | |
| "loss": 1.183, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.10505050505050505, | |
| "grad_norm": 20.375, | |
| "learning_rate": 5.241935483870968e-06, | |
| "loss": 1.1628, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.10666666666666667, | |
| "grad_norm": 18.75, | |
| "learning_rate": 5.322580645161291e-06, | |
| "loss": 0.9526, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.10828282828282829, | |
| "grad_norm": 18.25, | |
| "learning_rate": 5.4032258064516126e-06, | |
| "loss": 0.9687, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.1098989898989899, | |
| "grad_norm": 17.625, | |
| "learning_rate": 5.483870967741935e-06, | |
| "loss": 0.9895, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.11151515151515151, | |
| "grad_norm": 18.75, | |
| "learning_rate": 5.564516129032258e-06, | |
| "loss": 0.9116, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.11313131313131314, | |
| "grad_norm": 19.625, | |
| "learning_rate": 5.645161290322582e-06, | |
| "loss": 1.3456, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.11474747474747475, | |
| "grad_norm": 19.375, | |
| "learning_rate": 5.725806451612904e-06, | |
| "loss": 0.9474, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.11636363636363636, | |
| "grad_norm": 15.75, | |
| "learning_rate": 5.806451612903226e-06, | |
| "loss": 1.0595, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.11797979797979798, | |
| "grad_norm": 18.0, | |
| "learning_rate": 5.887096774193549e-06, | |
| "loss": 1.0755, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.1195959595959596, | |
| "grad_norm": 17.5, | |
| "learning_rate": 5.967741935483872e-06, | |
| "loss": 0.9928, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.12121212121212122, | |
| "grad_norm": 25.125, | |
| "learning_rate": 6.048387096774194e-06, | |
| "loss": 0.8585, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.12282828282828283, | |
| "grad_norm": 17.0, | |
| "learning_rate": 6.129032258064517e-06, | |
| "loss": 1.0376, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.12444444444444444, | |
| "grad_norm": 18.875, | |
| "learning_rate": 6.209677419354839e-06, | |
| "loss": 0.8824, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.12606060606060607, | |
| "grad_norm": 19.875, | |
| "learning_rate": 6.290322580645162e-06, | |
| "loss": 1.0713, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.12767676767676767, | |
| "grad_norm": 16.875, | |
| "learning_rate": 6.370967741935485e-06, | |
| "loss": 1.1012, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.1292929292929293, | |
| "grad_norm": 20.5, | |
| "learning_rate": 6.451612903225806e-06, | |
| "loss": 1.2101, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.13090909090909092, | |
| "grad_norm": 16.375, | |
| "learning_rate": 6.532258064516129e-06, | |
| "loss": 1.4869, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.13252525252525252, | |
| "grad_norm": 18.375, | |
| "learning_rate": 6.612903225806452e-06, | |
| "loss": 1.1957, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.13414141414141414, | |
| "grad_norm": 17.0, | |
| "learning_rate": 6.693548387096774e-06, | |
| "loss": 0.9294, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.13575757575757577, | |
| "grad_norm": 14.8125, | |
| "learning_rate": 6.774193548387097e-06, | |
| "loss": 0.7969, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.13737373737373737, | |
| "grad_norm": 16.625, | |
| "learning_rate": 6.854838709677419e-06, | |
| "loss": 0.9239, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.138989898989899, | |
| "grad_norm": 13.625, | |
| "learning_rate": 6.935483870967743e-06, | |
| "loss": 0.9914, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.1406060606060606, | |
| "grad_norm": 13.6875, | |
| "learning_rate": 7.0161290322580654e-06, | |
| "loss": 0.643, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.14222222222222222, | |
| "grad_norm": 15.0625, | |
| "learning_rate": 7.096774193548388e-06, | |
| "loss": 1.0152, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.14383838383838385, | |
| "grad_norm": 14.3125, | |
| "learning_rate": 7.177419354838711e-06, | |
| "loss": 0.7406, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.14545454545454545, | |
| "grad_norm": 16.875, | |
| "learning_rate": 7.258064516129033e-06, | |
| "loss": 1.0667, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.14707070707070707, | |
| "grad_norm": 14.8125, | |
| "learning_rate": 7.338709677419356e-06, | |
| "loss": 0.958, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.1486868686868687, | |
| "grad_norm": 15.4375, | |
| "learning_rate": 7.4193548387096784e-06, | |
| "loss": 0.9992, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.1503030303030303, | |
| "grad_norm": 16.25, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.7043, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.15191919191919193, | |
| "grad_norm": 16.875, | |
| "learning_rate": 7.580645161290323e-06, | |
| "loss": 1.2198, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.15353535353535352, | |
| "grad_norm": 19.5, | |
| "learning_rate": 7.661290322580646e-06, | |
| "loss": 1.2321, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.15515151515151515, | |
| "grad_norm": 15.4375, | |
| "learning_rate": 7.741935483870968e-06, | |
| "loss": 0.9691, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.15676767676767678, | |
| "grad_norm": 15.625, | |
| "learning_rate": 7.822580645161291e-06, | |
| "loss": 0.7549, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.15838383838383838, | |
| "grad_norm": 14.6875, | |
| "learning_rate": 7.903225806451613e-06, | |
| "loss": 0.8696, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 14.0, | |
| "learning_rate": 7.983870967741935e-06, | |
| "loss": 0.9764, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.16161616161616163, | |
| "grad_norm": 14.5, | |
| "learning_rate": 8.064516129032258e-06, | |
| "loss": 0.8763, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.16323232323232323, | |
| "grad_norm": 13.5625, | |
| "learning_rate": 8.145161290322582e-06, | |
| "loss": 0.9208, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.16484848484848486, | |
| "grad_norm": 16.0, | |
| "learning_rate": 8.225806451612904e-06, | |
| "loss": 1.1447, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.16646464646464645, | |
| "grad_norm": 16.25, | |
| "learning_rate": 8.306451612903227e-06, | |
| "loss": 0.9742, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.16808080808080808, | |
| "grad_norm": 17.125, | |
| "learning_rate": 8.387096774193549e-06, | |
| "loss": 0.9766, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.1696969696969697, | |
| "grad_norm": 15.4375, | |
| "learning_rate": 8.467741935483872e-06, | |
| "loss": 1.0782, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.1713131313131313, | |
| "grad_norm": 13.75, | |
| "learning_rate": 8.548387096774194e-06, | |
| "loss": 0.9839, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.17292929292929293, | |
| "grad_norm": 13.75, | |
| "learning_rate": 8.629032258064517e-06, | |
| "loss": 0.8056, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.17454545454545456, | |
| "grad_norm": 13.875, | |
| "learning_rate": 8.70967741935484e-06, | |
| "loss": 0.8916, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.17616161616161616, | |
| "grad_norm": 14.6875, | |
| "learning_rate": 8.790322580645163e-06, | |
| "loss": 0.8902, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.17777777777777778, | |
| "grad_norm": 18.125, | |
| "learning_rate": 8.870967741935484e-06, | |
| "loss": 1.027, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.17939393939393938, | |
| "grad_norm": 13.3125, | |
| "learning_rate": 8.951612903225806e-06, | |
| "loss": 0.9191, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.181010101010101, | |
| "grad_norm": 14.0625, | |
| "learning_rate": 9.03225806451613e-06, | |
| "loss": 0.9115, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.18262626262626264, | |
| "grad_norm": 14.8125, | |
| "learning_rate": 9.112903225806451e-06, | |
| "loss": 1.4586, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.18424242424242424, | |
| "grad_norm": 14.3125, | |
| "learning_rate": 9.193548387096775e-06, | |
| "loss": 0.6925, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.18585858585858586, | |
| "grad_norm": 13.875, | |
| "learning_rate": 9.274193548387097e-06, | |
| "loss": 1.0962, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.1874747474747475, | |
| "grad_norm": 14.25, | |
| "learning_rate": 9.35483870967742e-06, | |
| "loss": 0.8336, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.1890909090909091, | |
| "grad_norm": 24.375, | |
| "learning_rate": 9.435483870967743e-06, | |
| "loss": 1.1903, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.1907070707070707, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 9.516129032258065e-06, | |
| "loss": 0.6861, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.1923232323232323, | |
| "grad_norm": 19.25, | |
| "learning_rate": 9.596774193548389e-06, | |
| "loss": 1.1889, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.19393939393939394, | |
| "grad_norm": 14.4375, | |
| "learning_rate": 9.67741935483871e-06, | |
| "loss": 1.0123, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.19555555555555557, | |
| "grad_norm": 15.625, | |
| "learning_rate": 9.758064516129034e-06, | |
| "loss": 0.9163, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.19717171717171716, | |
| "grad_norm": 14.25, | |
| "learning_rate": 9.838709677419356e-06, | |
| "loss": 0.9699, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.1987878787878788, | |
| "grad_norm": 13.25, | |
| "learning_rate": 9.919354838709679e-06, | |
| "loss": 0.9863, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.1987878787878788, | |
| "eval_loss": 0.8790417313575745, | |
| "eval_runtime": 35.1712, | |
| "eval_samples_per_second": 31.276, | |
| "eval_steps_per_second": 3.924, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.20040404040404042, | |
| "grad_norm": 14.1875, | |
| "learning_rate": 1e-05, | |
| "loss": 1.1048, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.20202020202020202, | |
| "grad_norm": 13.375, | |
| "learning_rate": 9.999980046003546e-06, | |
| "loss": 0.7441, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.20363636363636364, | |
| "grad_norm": 12.9375, | |
| "learning_rate": 9.999920184173449e-06, | |
| "loss": 0.9276, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.20525252525252524, | |
| "grad_norm": 13.125, | |
| "learning_rate": 9.999820414987502e-06, | |
| "loss": 0.9395, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.20686868686868687, | |
| "grad_norm": 14.625, | |
| "learning_rate": 9.999680739242022e-06, | |
| "loss": 1.0927, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.2084848484848485, | |
| "grad_norm": 13.875, | |
| "learning_rate": 9.999501158051846e-06, | |
| "loss": 0.8983, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.2101010101010101, | |
| "grad_norm": 12.9375, | |
| "learning_rate": 9.999281672850317e-06, | |
| "loss": 0.8245, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.21171717171717172, | |
| "grad_norm": 15.5625, | |
| "learning_rate": 9.99902228538928e-06, | |
| "loss": 0.8986, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.21333333333333335, | |
| "grad_norm": 15.625, | |
| "learning_rate": 9.99872299773906e-06, | |
| "loss": 1.0636, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.21494949494949495, | |
| "grad_norm": 13.125, | |
| "learning_rate": 9.998383812288451e-06, | |
| "loss": 0.8439, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.21656565656565657, | |
| "grad_norm": 12.25, | |
| "learning_rate": 9.998004731744696e-06, | |
| "loss": 0.8364, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.21818181818181817, | |
| "grad_norm": 12.0, | |
| "learning_rate": 9.997585759133463e-06, | |
| "loss": 0.8303, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.2197979797979798, | |
| "grad_norm": 11.5625, | |
| "learning_rate": 9.997126897798826e-06, | |
| "loss": 0.8335, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.22141414141414142, | |
| "grad_norm": 13.625, | |
| "learning_rate": 9.996628151403226e-06, | |
| "loss": 0.9116, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.22303030303030302, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 9.996089523927461e-06, | |
| "loss": 0.7399, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.22464646464646465, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 9.995511019670639e-06, | |
| "loss": 1.1413, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.22626262626262628, | |
| "grad_norm": 11.625, | |
| "learning_rate": 9.994892643250147e-06, | |
| "loss": 1.0564, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.22787878787878788, | |
| "grad_norm": 11.6875, | |
| "learning_rate": 9.99423439960162e-06, | |
| "loss": 1.013, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.2294949494949495, | |
| "grad_norm": 11.5, | |
| "learning_rate": 9.993536293978892e-06, | |
| "loss": 1.154, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.2311111111111111, | |
| "grad_norm": 11.4375, | |
| "learning_rate": 9.992798331953962e-06, | |
| "loss": 0.7173, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.23272727272727273, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 9.99202051941695e-06, | |
| "loss": 0.6389, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.23434343434343435, | |
| "grad_norm": 11.5, | |
| "learning_rate": 9.991202862576036e-06, | |
| "loss": 1.1159, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.23595959595959595, | |
| "grad_norm": 12.625, | |
| "learning_rate": 9.99034536795744e-06, | |
| "loss": 0.9197, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.23757575757575758, | |
| "grad_norm": 13.0, | |
| "learning_rate": 9.989448042405328e-06, | |
| "loss": 0.9929, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.2391919191919192, | |
| "grad_norm": 11.875, | |
| "learning_rate": 9.9885108930818e-06, | |
| "loss": 0.7332, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.2408080808080808, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 9.987533927466804e-06, | |
| "loss": 0.8305, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.24242424242424243, | |
| "grad_norm": 12.375, | |
| "learning_rate": 9.986517153358086e-06, | |
| "loss": 0.7982, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.24404040404040403, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 9.98546057887113e-06, | |
| "loss": 1.0044, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.24565656565656566, | |
| "grad_norm": 12.75, | |
| "learning_rate": 9.984364212439089e-06, | |
| "loss": 0.9286, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.24727272727272728, | |
| "grad_norm": 11.5, | |
| "learning_rate": 9.98322806281272e-06, | |
| "loss": 1.0573, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.24888888888888888, | |
| "grad_norm": 11.0, | |
| "learning_rate": 9.982052139060312e-06, | |
| "loss": 0.8163, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.2505050505050505, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 9.980836450567619e-06, | |
| "loss": 0.8431, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.25212121212121213, | |
| "grad_norm": 12.5, | |
| "learning_rate": 9.979581007037776e-06, | |
| "loss": 0.6704, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.25373737373737376, | |
| "grad_norm": 10.875, | |
| "learning_rate": 9.978285818491232e-06, | |
| "loss": 0.6866, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.25535353535353533, | |
| "grad_norm": 14.375, | |
| "learning_rate": 9.976950895265657e-06, | |
| "loss": 1.0205, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.25696969696969696, | |
| "grad_norm": 9.75, | |
| "learning_rate": 9.975576248015878e-06, | |
| "loss": 0.9211, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.2585858585858586, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 9.974161887713775e-06, | |
| "loss": 1.1265, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.2602020202020202, | |
| "grad_norm": 9.0, | |
| "learning_rate": 9.972707825648204e-06, | |
| "loss": 0.8746, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.26181818181818184, | |
| "grad_norm": 12.8125, | |
| "learning_rate": 9.971214073424906e-06, | |
| "loss": 0.846, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.2634343434343434, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 9.969680642966409e-06, | |
| "loss": 0.6882, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.26505050505050504, | |
| "grad_norm": 13.5625, | |
| "learning_rate": 9.968107546511942e-06, | |
| "loss": 0.9636, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.26666666666666666, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 9.966494796617328e-06, | |
| "loss": 0.6918, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.2682828282828283, | |
| "grad_norm": 9.375, | |
| "learning_rate": 9.964842406154892e-06, | |
| "loss": 1.0313, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.2698989898989899, | |
| "grad_norm": 10.625, | |
| "learning_rate": 9.963150388313347e-06, | |
| "loss": 0.9354, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.27151515151515154, | |
| "grad_norm": 12.3125, | |
| "learning_rate": 9.961418756597703e-06, | |
| "loss": 0.9895, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.2731313131313131, | |
| "grad_norm": 12.3125, | |
| "learning_rate": 9.959647524829148e-06, | |
| "loss": 0.9479, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.27474747474747474, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 9.957836707144943e-06, | |
| "loss": 1.0128, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.27636363636363637, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 9.95598631799831e-06, | |
| "loss": 0.7783, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.277979797979798, | |
| "grad_norm": 11.8125, | |
| "learning_rate": 9.95409637215831e-06, | |
| "loss": 0.9492, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.2795959595959596, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 9.952166884709735e-06, | |
| "loss": 0.8773, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.2812121212121212, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 9.950197871052974e-06, | |
| "loss": 1.0219, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.2828282828282828, | |
| "grad_norm": 10.25, | |
| "learning_rate": 9.94818934690391e-06, | |
| "loss": 0.6188, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.28444444444444444, | |
| "grad_norm": 12.1875, | |
| "learning_rate": 9.94614132829377e-06, | |
| "loss": 0.7879, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.28606060606060607, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 9.94405383156902e-06, | |
| "loss": 0.7926, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.2876767676767677, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 9.941926873391223e-06, | |
| "loss": 0.7642, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.28929292929292927, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 9.939760470736902e-06, | |
| "loss": 0.8427, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.2909090909090909, | |
| "grad_norm": 12.5, | |
| "learning_rate": 9.937554640897414e-06, | |
| "loss": 0.7917, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.2925252525252525, | |
| "grad_norm": 11.875, | |
| "learning_rate": 9.935309401478808e-06, | |
| "loss": 0.8325, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.29414141414141415, | |
| "grad_norm": 10.0, | |
| "learning_rate": 9.933024770401682e-06, | |
| "loss": 0.7708, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.2957575757575758, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 9.930700765901046e-06, | |
| "loss": 1.1573, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.2973737373737374, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 9.928337406526172e-06, | |
| "loss": 1.0751, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.298989898989899, | |
| "grad_norm": 8.25, | |
| "learning_rate": 9.925934711140444e-06, | |
| "loss": 0.6648, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.3006060606060606, | |
| "grad_norm": 9.625, | |
| "learning_rate": 9.923492698921214e-06, | |
| "loss": 1.0982, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.3022222222222222, | |
| "grad_norm": 9.5, | |
| "learning_rate": 9.921011389359638e-06, | |
| "loss": 0.9648, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.30383838383838385, | |
| "grad_norm": 13.4375, | |
| "learning_rate": 9.918490802260538e-06, | |
| "loss": 0.7589, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.3054545454545455, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 9.915930957742228e-06, | |
| "loss": 0.6699, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.30707070707070705, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 9.913331876236358e-06, | |
| "loss": 0.9228, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.3086868686868687, | |
| "grad_norm": 9.75, | |
| "learning_rate": 9.910693578487755e-06, | |
| "loss": 0.6985, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.3103030303030303, | |
| "grad_norm": 9.0625, | |
| "learning_rate": 9.90801608555425e-06, | |
| "loss": 0.8004, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.31191919191919193, | |
| "grad_norm": 9.25, | |
| "learning_rate": 9.905299418806517e-06, | |
| "loss": 0.8232, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.31353535353535356, | |
| "grad_norm": 10.625, | |
| "learning_rate": 9.902543599927903e-06, | |
| "loss": 1.3909, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.3151515151515151, | |
| "grad_norm": 9.25, | |
| "learning_rate": 9.899748650914245e-06, | |
| "loss": 0.9673, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.31676767676767675, | |
| "grad_norm": 12.0, | |
| "learning_rate": 9.896914594073703e-06, | |
| "loss": 0.9208, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.3183838383838384, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 9.894041452026584e-06, | |
| "loss": 0.8361, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 9.891129247705153e-06, | |
| "loss": 1.0054, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.32161616161616163, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 9.888178004353456e-06, | |
| "loss": 0.8643, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.32323232323232326, | |
| "grad_norm": 8.75, | |
| "learning_rate": 9.885187745527132e-06, | |
| "loss": 0.738, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.32484848484848483, | |
| "grad_norm": 10.0, | |
| "learning_rate": 9.88215849509323e-06, | |
| "loss": 0.9793, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.32646464646464646, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 9.879090277230005e-06, | |
| "loss": 0.7254, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.3280808080808081, | |
| "grad_norm": 11.5, | |
| "learning_rate": 9.875983116426745e-06, | |
| "loss": 0.8806, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.3296969696969697, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 9.87283703748356e-06, | |
| "loss": 0.9895, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.33131313131313134, | |
| "grad_norm": 8.4375, | |
| "learning_rate": 9.869652065511188e-06, | |
| "loss": 0.7317, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.3329292929292929, | |
| "grad_norm": 9.5, | |
| "learning_rate": 9.866428225930798e-06, | |
| "loss": 0.6969, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.33454545454545453, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 9.863165544473781e-06, | |
| "loss": 0.8589, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.33616161616161616, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 9.859864047181551e-06, | |
| "loss": 0.6515, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.3377777777777778, | |
| "grad_norm": 7.84375, | |
| "learning_rate": 9.856523760405339e-06, | |
| "loss": 0.6014, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.3393939393939394, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 9.853144710805966e-06, | |
| "loss": 0.7315, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.341010101010101, | |
| "grad_norm": 11.25, | |
| "learning_rate": 9.849726925353655e-06, | |
| "loss": 0.721, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.3426262626262626, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 9.846270431327793e-06, | |
| "loss": 0.8893, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.34424242424242424, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 9.842775256316732e-06, | |
| "loss": 0.9498, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.34585858585858587, | |
| "grad_norm": 11.125, | |
| "learning_rate": 9.839241428217553e-06, | |
| "loss": 0.8841, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.3474747474747475, | |
| "grad_norm": 10.125, | |
| "learning_rate": 9.835668975235857e-06, | |
| "loss": 0.9036, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.3490909090909091, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 9.832057925885526e-06, | |
| "loss": 0.8493, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.3507070707070707, | |
| "grad_norm": 9.375, | |
| "learning_rate": 9.828408308988506e-06, | |
| "loss": 0.8034, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.3523232323232323, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 9.824720153674578e-06, | |
| "loss": 0.7762, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.35393939393939394, | |
| "grad_norm": 11.6875, | |
| "learning_rate": 9.820993489381114e-06, | |
| "loss": 0.986, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.35555555555555557, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 9.817228345852853e-06, | |
| "loss": 0.9599, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3571717171717172, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 9.813424753141658e-06, | |
| "loss": 0.7152, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.35878787878787877, | |
| "grad_norm": 7.75, | |
| "learning_rate": 9.809582741606283e-06, | |
| "loss": 0.828, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.3604040404040404, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 9.805702341912117e-06, | |
| "loss": 0.9487, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.362020202020202, | |
| "grad_norm": 9.125, | |
| "learning_rate": 9.801783585030959e-06, | |
| "loss": 0.8378, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 9.797826502240746e-06, | |
| "loss": 1.0327, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.3652525252525253, | |
| "grad_norm": 38.75, | |
| "learning_rate": 9.79383112512533e-06, | |
| "loss": 0.877, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.36686868686868684, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 9.789797485574204e-06, | |
| "loss": 1.0547, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.36848484848484847, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 9.785725615782262e-06, | |
| "loss": 0.6674, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.3701010101010101, | |
| "grad_norm": 14.75, | |
| "learning_rate": 9.781615548249533e-06, | |
| "loss": 0.8929, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.3717171717171717, | |
| "grad_norm": 8.5, | |
| "learning_rate": 9.777467315780926e-06, | |
| "loss": 0.6194, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.37333333333333335, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 9.773280951485967e-06, | |
| "loss": 0.7332, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.374949494949495, | |
| "grad_norm": 7.5625, | |
| "learning_rate": 9.769056488778538e-06, | |
| "loss": 0.6245, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.37656565656565655, | |
| "grad_norm": 8.25, | |
| "learning_rate": 9.764793961376602e-06, | |
| "loss": 0.8382, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.3781818181818182, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 9.760493403301941e-06, | |
| "loss": 0.6372, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.3797979797979798, | |
| "grad_norm": 11.875, | |
| "learning_rate": 9.756154848879885e-06, | |
| "loss": 0.9138, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.3814141414141414, | |
| "grad_norm": 8.25, | |
| "learning_rate": 9.751778332739033e-06, | |
| "loss": 0.9099, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.38303030303030305, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 9.74736388981098e-06, | |
| "loss": 0.864, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.3846464646464646, | |
| "grad_norm": 8.4375, | |
| "learning_rate": 9.74291155533004e-06, | |
| "loss": 0.6037, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.38626262626262625, | |
| "grad_norm": 8.875, | |
| "learning_rate": 9.738421364832956e-06, | |
| "loss": 0.6612, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.3878787878787879, | |
| "grad_norm": 8.25, | |
| "learning_rate": 9.733893354158628e-06, | |
| "loss": 0.772, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.3894949494949495, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 9.729327559447817e-06, | |
| "loss": 0.7761, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.39111111111111113, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 9.724724017142869e-06, | |
| "loss": 0.8807, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.3927272727272727, | |
| "grad_norm": 8.5, | |
| "learning_rate": 9.720082763987404e-06, | |
| "loss": 0.8429, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.39434343434343433, | |
| "grad_norm": 7.5625, | |
| "learning_rate": 9.715403837026046e-06, | |
| "loss": 0.4624, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.39595959595959596, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 9.710687273604111e-06, | |
| "loss": 0.865, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.3975757575757576, | |
| "grad_norm": 6.875, | |
| "learning_rate": 9.705933111367314e-06, | |
| "loss": 0.7918, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.3975757575757576, | |
| "eval_loss": 0.8323823809623718, | |
| "eval_runtime": 35.1362, | |
| "eval_samples_per_second": 31.307, | |
| "eval_steps_per_second": 3.928, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.3991919191919192, | |
| "grad_norm": 8.375, | |
| "learning_rate": 9.701141388261471e-06, | |
| "loss": 0.7728, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.40080808080808084, | |
| "grad_norm": 9.5, | |
| "learning_rate": 9.69631214253219e-06, | |
| "loss": 0.8228, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.4024242424242424, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 9.691445412724576e-06, | |
| "loss": 0.8661, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.40404040404040403, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 9.68654123768291e-06, | |
| "loss": 0.8111, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.40565656565656566, | |
| "grad_norm": 8.5, | |
| "learning_rate": 9.681599656550346e-06, | |
| "loss": 0.7355, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.4072727272727273, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 9.676620708768608e-06, | |
| "loss": 0.8194, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.4088888888888889, | |
| "grad_norm": 10.25, | |
| "learning_rate": 9.67160443407765e-06, | |
| "loss": 0.8502, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.4105050505050505, | |
| "grad_norm": 16.625, | |
| "learning_rate": 9.666550872515367e-06, | |
| "loss": 0.841, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.4121212121212121, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 9.661460064417263e-06, | |
| "loss": 1.1208, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.41373737373737374, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 9.656332050416118e-06, | |
| "loss": 0.9173, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.41535353535353536, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 9.651166871441684e-06, | |
| "loss": 1.1702, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.416969696969697, | |
| "grad_norm": 9.875, | |
| "learning_rate": 9.645964568720345e-06, | |
| "loss": 0.8802, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.41858585858585856, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 9.640725183774797e-06, | |
| "loss": 0.9126, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.4202020202020202, | |
| "grad_norm": 8.25, | |
| "learning_rate": 9.635448758423703e-06, | |
| "loss": 0.6081, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.4218181818181818, | |
| "grad_norm": 7.6875, | |
| "learning_rate": 9.630135334781373e-06, | |
| "loss": 0.7091, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.42343434343434344, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 9.624784955257423e-06, | |
| "loss": 0.7537, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.42505050505050507, | |
| "grad_norm": 8.375, | |
| "learning_rate": 9.619397662556434e-06, | |
| "loss": 0.7439, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.4266666666666667, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 9.613973499677613e-06, | |
| "loss": 0.9897, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.42828282828282827, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 9.608512509914453e-06, | |
| "loss": 0.7775, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.4298989898989899, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 9.60301473685438e-06, | |
| "loss": 0.8957, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.4315151515151515, | |
| "grad_norm": 11.0, | |
| "learning_rate": 9.597480224378412e-06, | |
| "loss": 1.0668, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.43313131313131314, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 9.591909016660806e-06, | |
| "loss": 0.8081, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.43474747474747477, | |
| "grad_norm": 9.25, | |
| "learning_rate": 9.586301158168706e-06, | |
| "loss": 0.8119, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.43636363636363634, | |
| "grad_norm": 11.4375, | |
| "learning_rate": 9.580656693661787e-06, | |
| "loss": 0.8829, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.43797979797979797, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 9.574975668191899e-06, | |
| "loss": 0.7646, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.4395959595959596, | |
| "grad_norm": 10.25, | |
| "learning_rate": 9.569258127102708e-06, | |
| "loss": 0.8685, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.4412121212121212, | |
| "grad_norm": 10.0, | |
| "learning_rate": 9.56350411602933e-06, | |
| "loss": 1.093, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.44282828282828285, | |
| "grad_norm": 51.5, | |
| "learning_rate": 9.55771368089797e-06, | |
| "loss": 0.8206, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 9.551886867925565e-06, | |
| "loss": 1.0301, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.44606060606060605, | |
| "grad_norm": 10.125, | |
| "learning_rate": 9.546023723619387e-06, | |
| "loss": 1.0908, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.4476767676767677, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 9.540124294776704e-06, | |
| "loss": 0.7014, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.4492929292929293, | |
| "grad_norm": 7.84375, | |
| "learning_rate": 9.534188628484391e-06, | |
| "loss": 0.682, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.4509090909090909, | |
| "grad_norm": 7.59375, | |
| "learning_rate": 9.528216772118553e-06, | |
| "loss": 0.7439, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.45252525252525255, | |
| "grad_norm": 8.125, | |
| "learning_rate": 9.522208773344147e-06, | |
| "loss": 0.7983, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.4541414141414141, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 9.516164680114612e-06, | |
| "loss": 0.4831, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.45575757575757575, | |
| "grad_norm": 10.375, | |
| "learning_rate": 9.510084540671471e-06, | |
| "loss": 0.7558, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.4573737373737374, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 9.50396840354396e-06, | |
| "loss": 0.5978, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.458989898989899, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 9.497816317548625e-06, | |
| "loss": 0.9693, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.46060606060606063, | |
| "grad_norm": 7.65625, | |
| "learning_rate": 9.491628331788949e-06, | |
| "loss": 0.6546, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.4622222222222222, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 9.485404495654953e-06, | |
| "loss": 0.9247, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.4638383838383838, | |
| "grad_norm": 8.0, | |
| "learning_rate": 9.479144858822795e-06, | |
| "loss": 0.6692, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.46545454545454545, | |
| "grad_norm": 10.125, | |
| "learning_rate": 9.472849471254386e-06, | |
| "loss": 1.1385, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.4670707070707071, | |
| "grad_norm": 8.125, | |
| "learning_rate": 9.466518383196981e-06, | |
| "loss": 0.6915, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.4686868686868687, | |
| "grad_norm": 6.625, | |
| "learning_rate": 9.460151645182784e-06, | |
| "loss": 0.6693, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.4703030303030303, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 9.453749308028542e-06, | |
| "loss": 0.641, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.4719191919191919, | |
| "grad_norm": 8.75, | |
| "learning_rate": 9.447311422835141e-06, | |
| "loss": 0.8019, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.47353535353535353, | |
| "grad_norm": 9.0, | |
| "learning_rate": 9.440838040987194e-06, | |
| "loss": 0.8253, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.47515151515151516, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 9.43432921415264e-06, | |
| "loss": 1.1923, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.4767676767676768, | |
| "grad_norm": 7.90625, | |
| "learning_rate": 9.427784994282318e-06, | |
| "loss": 0.7972, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.4783838383838384, | |
| "grad_norm": 8.0, | |
| "learning_rate": 9.421205433609568e-06, | |
| "loss": 0.7207, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 9.414590584649798e-06, | |
| "loss": 0.861, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.4816161616161616, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 9.407940500200082e-06, | |
| "loss": 0.8226, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.48323232323232324, | |
| "grad_norm": 7.75, | |
| "learning_rate": 9.40125523333872e-06, | |
| "loss": 0.9907, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.48484848484848486, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 9.39453483742483e-06, | |
| "loss": 0.7025, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.4864646464646465, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 9.387779366097915e-06, | |
| "loss": 0.5553, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.48808080808080806, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 9.380988873277436e-06, | |
| "loss": 0.6758, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.4896969696969697, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 9.37416341316238e-06, | |
| "loss": 0.5991, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.4913131313131313, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 9.367303040230828e-06, | |
| "loss": 0.9227, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.49292929292929294, | |
| "grad_norm": 23.75, | |
| "learning_rate": 9.360407809239527e-06, | |
| "loss": 0.6559, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.49454545454545457, | |
| "grad_norm": 9.625, | |
| "learning_rate": 9.35347777522344e-06, | |
| "loss": 0.6758, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.49616161616161614, | |
| "grad_norm": 8.875, | |
| "learning_rate": 9.346512993495315e-06, | |
| "loss": 0.8509, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.49777777777777776, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 9.33951351964525e-06, | |
| "loss": 0.6474, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.4993939393939394, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 9.33247940954023e-06, | |
| "loss": 0.7196, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.501010101010101, | |
| "grad_norm": 8.125, | |
| "learning_rate": 9.3254107193237e-06, | |
| "loss": 0.7781, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.5026262626262626, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 9.318307505415108e-06, | |
| "loss": 1.1145, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.5042424242424243, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 9.311169824509454e-06, | |
| "loss": 0.979, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.5058585858585859, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 9.303997733576846e-06, | |
| "loss": 1.0086, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.5074747474747475, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 9.296791289862031e-06, | |
| "loss": 0.5292, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.509090909090909, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 9.289550550883952e-06, | |
| "loss": 0.7044, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.5107070707070707, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 9.28227557443528e-06, | |
| "loss": 0.9252, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.5123232323232323, | |
| "grad_norm": 7.125, | |
| "learning_rate": 9.274966418581958e-06, | |
| "loss": 0.5309, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.5139393939393939, | |
| "grad_norm": 8.0, | |
| "learning_rate": 9.267623141662734e-06, | |
| "loss": 0.6308, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.5155555555555555, | |
| "grad_norm": 8.875, | |
| "learning_rate": 9.260245802288694e-06, | |
| "loss": 0.7667, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.5171717171717172, | |
| "grad_norm": 12.9375, | |
| "learning_rate": 9.252834459342801e-06, | |
| "loss": 0.8895, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.5187878787878788, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 9.24538917197942e-06, | |
| "loss": 0.3923, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.5204040404040404, | |
| "grad_norm": 11.125, | |
| "learning_rate": 9.237909999623847e-06, | |
| "loss": 0.9289, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.522020202020202, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 9.230397001971829e-06, | |
| "loss": 0.8841, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.5236363636363637, | |
| "grad_norm": 10.0, | |
| "learning_rate": 9.222850238989104e-06, | |
| "loss": 1.0293, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.5252525252525253, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 9.215269770910897e-06, | |
| "loss": 0.9055, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.5268686868686868, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 9.207655658241469e-06, | |
| "loss": 0.4496, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.5284848484848484, | |
| "grad_norm": 9.25, | |
| "learning_rate": 9.200007961753605e-06, | |
| "loss": 1.0731, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.5301010101010101, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 9.192326742488153e-06, | |
| "loss": 0.672, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.5317171717171717, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 9.184612061753518e-06, | |
| "loss": 1.0366, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.5333333333333333, | |
| "grad_norm": 8.625, | |
| "learning_rate": 9.176863981125185e-06, | |
| "loss": 0.9249, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.534949494949495, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 9.169082562445228e-06, | |
| "loss": 0.6611, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.5365656565656566, | |
| "grad_norm": 7.625, | |
| "learning_rate": 9.161267867821802e-06, | |
| "loss": 0.7343, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.5381818181818182, | |
| "grad_norm": 8.75, | |
| "learning_rate": 9.153419959628666e-06, | |
| "loss": 0.9306, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.5397979797979798, | |
| "grad_norm": 6.75, | |
| "learning_rate": 9.14553890050467e-06, | |
| "loss": 0.7274, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.5414141414141415, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 9.137624753353268e-06, | |
| "loss": 0.5954, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.5430303030303031, | |
| "grad_norm": 7.125, | |
| "learning_rate": 9.129677581342e-06, | |
| "loss": 1.0952, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.5446464646464646, | |
| "grad_norm": 8.375, | |
| "learning_rate": 9.121697447902006e-06, | |
| "loss": 0.6458, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.5462626262626262, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 9.113684416727511e-06, | |
| "loss": 0.6578, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.5478787878787879, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 9.10563855177531e-06, | |
| "loss": 0.6899, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.5494949494949495, | |
| "grad_norm": 10.375, | |
| "learning_rate": 9.097559917264268e-06, | |
| "loss": 0.7607, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.5511111111111111, | |
| "grad_norm": 9.875, | |
| "learning_rate": 9.089448577674799e-06, | |
| "loss": 0.6255, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.5527272727272727, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 9.081304597748366e-06, | |
| "loss": 0.9227, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.5543434343434344, | |
| "grad_norm": 6.71875, | |
| "learning_rate": 9.073128042486945e-06, | |
| "loss": 0.6744, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.555959595959596, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 9.064918977152517e-06, | |
| "loss": 0.4926, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.5575757575757576, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 9.056677467266543e-06, | |
| "loss": 0.5658, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.5591919191919192, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 9.048403578609454e-06, | |
| "loss": 0.9654, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.5608080808080808, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 9.040097377220103e-06, | |
| "loss": 0.8399, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.5624242424242424, | |
| "grad_norm": 8.875, | |
| "learning_rate": 9.031758929395259e-06, | |
| "loss": 0.7202, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.564040404040404, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 9.02338830168906e-06, | |
| "loss": 0.5778, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.5656565656565656, | |
| "grad_norm": 6.375, | |
| "learning_rate": 9.014985560912499e-06, | |
| "loss": 0.4703, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5672727272727273, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 9.00655077413288e-06, | |
| "loss": 0.8051, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.5688888888888889, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 8.998084008673284e-06, | |
| "loss": 0.5963, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.5705050505050505, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 8.989585332112039e-06, | |
| "loss": 0.9701, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.5721212121212121, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 8.981054812282162e-06, | |
| "loss": 1.0809, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.5737373737373738, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 8.972492517270845e-06, | |
| "loss": 0.8909, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.5753535353535354, | |
| "grad_norm": 7.71875, | |
| "learning_rate": 8.963898515418885e-06, | |
| "loss": 0.5967, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.576969696969697, | |
| "grad_norm": 7.34375, | |
| "learning_rate": 8.955272875320159e-06, | |
| "loss": 0.74, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.5785858585858585, | |
| "grad_norm": 7.71875, | |
| "learning_rate": 8.946615665821059e-06, | |
| "loss": 0.6306, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.5802020202020202, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 8.937926956019957e-06, | |
| "loss": 0.5952, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.5818181818181818, | |
| "grad_norm": 9.0, | |
| "learning_rate": 8.929206815266653e-06, | |
| "loss": 0.5917, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.5834343434343434, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 8.920455313161801e-06, | |
| "loss": 0.6248, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.585050505050505, | |
| "grad_norm": 8.375, | |
| "learning_rate": 8.911672519556386e-06, | |
| "loss": 0.7684, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.5866666666666667, | |
| "grad_norm": 7.125, | |
| "learning_rate": 8.902858504551136e-06, | |
| "loss": 0.6592, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.5882828282828283, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 8.89401333849598e-06, | |
| "loss": 0.6805, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.5898989898989899, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 8.885137091989488e-06, | |
| "loss": 0.8623, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.5915151515151515, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 8.87622983587829e-06, | |
| "loss": 1.0672, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.5931313131313132, | |
| "grad_norm": 7.8125, | |
| "learning_rate": 8.867291641256534e-06, | |
| "loss": 1.091, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.5947474747474748, | |
| "grad_norm": 12.1875, | |
| "learning_rate": 8.8583225794653e-06, | |
| "loss": 0.7798, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.5963636363636363, | |
| "grad_norm": 6.53125, | |
| "learning_rate": 8.849322722092037e-06, | |
| "loss": 0.5133, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.5963636363636363, | |
| "eval_loss": 0.7914893627166748, | |
| "eval_runtime": 35.1895, | |
| "eval_samples_per_second": 31.259, | |
| "eval_steps_per_second": 3.922, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.597979797979798, | |
| "grad_norm": 7.65625, | |
| "learning_rate": 8.840292140969995e-06, | |
| "loss": 0.9505, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.5995959595959596, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 8.83123090817765e-06, | |
| "loss": 0.6006, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.6012121212121212, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 8.82213909603812e-06, | |
| "loss": 0.7318, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.6028282828282828, | |
| "grad_norm": 8.125, | |
| "learning_rate": 8.813016777118604e-06, | |
| "loss": 0.725, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.6044444444444445, | |
| "grad_norm": 8.0, | |
| "learning_rate": 8.803864024229786e-06, | |
| "loss": 0.8758, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.6060606060606061, | |
| "grad_norm": 7.6875, | |
| "learning_rate": 8.79468091042527e-06, | |
| "loss": 0.8636, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.6076767676767677, | |
| "grad_norm": 9.75, | |
| "learning_rate": 8.78546750900098e-06, | |
| "loss": 0.7711, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.6092929292929293, | |
| "grad_norm": 7.21875, | |
| "learning_rate": 8.77622389349459e-06, | |
| "loss": 0.5907, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.610909090909091, | |
| "grad_norm": 6.15625, | |
| "learning_rate": 8.766950137684929e-06, | |
| "loss": 0.6627, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.6125252525252525, | |
| "grad_norm": 8.125, | |
| "learning_rate": 8.757646315591391e-06, | |
| "loss": 0.9963, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.6141414141414141, | |
| "grad_norm": 6.25, | |
| "learning_rate": 8.748312501473351e-06, | |
| "loss": 0.8233, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.6157575757575757, | |
| "grad_norm": 7.5, | |
| "learning_rate": 8.738948769829565e-06, | |
| "loss": 0.5933, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.6173737373737374, | |
| "grad_norm": 6.8125, | |
| "learning_rate": 8.72955519539758e-06, | |
| "loss": 0.4457, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.618989898989899, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 8.72013185315314e-06, | |
| "loss": 0.6291, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.6206060606060606, | |
| "grad_norm": 4.6875, | |
| "learning_rate": 8.710678818309576e-06, | |
| "loss": 0.5969, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.6222222222222222, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 8.701196166317221e-06, | |
| "loss": 0.4308, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.6238383838383839, | |
| "grad_norm": 9.5, | |
| "learning_rate": 8.691683972862792e-06, | |
| "loss": 0.8222, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.6254545454545455, | |
| "grad_norm": 8.125, | |
| "learning_rate": 8.682142313868802e-06, | |
| "loss": 0.8346, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.6270707070707071, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 8.672571265492944e-06, | |
| "loss": 0.7311, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.6286868686868687, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 8.66297090412748e-06, | |
| "loss": 0.7654, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.6303030303030303, | |
| "grad_norm": 9.25, | |
| "learning_rate": 8.653341306398645e-06, | |
| "loss": 0.8764, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.6319191919191919, | |
| "grad_norm": 7.75, | |
| "learning_rate": 8.643682549166021e-06, | |
| "loss": 0.8139, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.6335353535353535, | |
| "grad_norm": 7.4375, | |
| "learning_rate": 8.63399470952193e-06, | |
| "loss": 0.8704, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.6351515151515151, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 8.62427786479082e-06, | |
| "loss": 0.5697, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.6367676767676768, | |
| "grad_norm": 7.46875, | |
| "learning_rate": 8.614532092528645e-06, | |
| "loss": 0.6327, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.6383838383838384, | |
| "grad_norm": 7.75, | |
| "learning_rate": 8.604757470522247e-06, | |
| "loss": 0.7221, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 7.28125, | |
| "learning_rate": 8.594954076788736e-06, | |
| "loss": 0.7422, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.6416161616161616, | |
| "grad_norm": 14.1875, | |
| "learning_rate": 8.585121989574865e-06, | |
| "loss": 0.6289, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.6432323232323233, | |
| "grad_norm": 7.65625, | |
| "learning_rate": 8.575261287356407e-06, | |
| "loss": 0.8722, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.6448484848484849, | |
| "grad_norm": 7.125, | |
| "learning_rate": 8.565372048837529e-06, | |
| "loss": 0.7178, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.6464646464646465, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 8.555454352950161e-06, | |
| "loss": 0.6246, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.648080808080808, | |
| "grad_norm": 8.4375, | |
| "learning_rate": 8.545508278853373e-06, | |
| "loss": 0.8206, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.6496969696969697, | |
| "grad_norm": 10.125, | |
| "learning_rate": 8.535533905932739e-06, | |
| "loss": 0.9123, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.6513131313131313, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 8.525531313799693e-06, | |
| "loss": 0.9784, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.6529292929292929, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 8.515500582290914e-06, | |
| "loss": 0.9171, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.6545454545454545, | |
| "grad_norm": 5.84375, | |
| "learning_rate": 8.505441791467674e-06, | |
| "loss": 0.5419, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.6561616161616162, | |
| "grad_norm": 6.625, | |
| "learning_rate": 8.495355021615204e-06, | |
| "loss": 0.6661, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.6577777777777778, | |
| "grad_norm": 7.625, | |
| "learning_rate": 8.48524035324205e-06, | |
| "loss": 0.8925, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.6593939393939394, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 8.475097867079437e-06, | |
| "loss": 0.5441, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.661010101010101, | |
| "grad_norm": 6.21875, | |
| "learning_rate": 8.464927644080617e-06, | |
| "loss": 0.5148, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.6626262626262627, | |
| "grad_norm": 8.125, | |
| "learning_rate": 8.454729765420228e-06, | |
| "loss": 0.7649, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.6642424242424242, | |
| "grad_norm": 7.65625, | |
| "learning_rate": 8.444504312493642e-06, | |
| "loss": 0.6308, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.6658585858585858, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 8.434251366916323e-06, | |
| "loss": 0.6534, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.6674747474747474, | |
| "grad_norm": 7.46875, | |
| "learning_rate": 8.423971010523165e-06, | |
| "loss": 0.8665, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.6690909090909091, | |
| "grad_norm": 6.78125, | |
| "learning_rate": 8.413663325367845e-06, | |
| "loss": 0.5408, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.6707070707070707, | |
| "grad_norm": 6.59375, | |
| "learning_rate": 8.403328393722169e-06, | |
| "loss": 0.6455, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.6723232323232323, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 8.392966298075413e-06, | |
| "loss": 0.6351, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.673939393939394, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 8.382577121133664e-06, | |
| "loss": 0.6295, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.6755555555555556, | |
| "grad_norm": 7.625, | |
| "learning_rate": 8.372160945819164e-06, | |
| "loss": 1.2937, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.6771717171717172, | |
| "grad_norm": 9.25, | |
| "learning_rate": 8.361717855269643e-06, | |
| "loss": 0.731, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.6787878787878788, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 8.351247932837655e-06, | |
| "loss": 0.5192, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.6804040404040405, | |
| "grad_norm": 8.125, | |
| "learning_rate": 8.34075126208992e-06, | |
| "loss": 0.8637, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.682020202020202, | |
| "grad_norm": 7.78125, | |
| "learning_rate": 8.330227926806652e-06, | |
| "loss": 0.6211, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.6836363636363636, | |
| "grad_norm": 7.71875, | |
| "learning_rate": 8.319678010980883e-06, | |
| "loss": 0.7818, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.6852525252525252, | |
| "grad_norm": 5.5625, | |
| "learning_rate": 8.309101598817812e-06, | |
| "loss": 0.4556, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.6868686868686869, | |
| "grad_norm": 5.875, | |
| "learning_rate": 8.298498774734114e-06, | |
| "loss": 0.6321, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.6884848484848485, | |
| "grad_norm": 7.375, | |
| "learning_rate": 8.287869623357275e-06, | |
| "loss": 0.8331, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.6901010101010101, | |
| "grad_norm": 8.625, | |
| "learning_rate": 8.277214229524913e-06, | |
| "loss": 0.7505, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.6917171717171717, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 8.266532678284103e-06, | |
| "loss": 0.6113, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.6933333333333334, | |
| "grad_norm": 8.0, | |
| "learning_rate": 8.255825054890705e-06, | |
| "loss": 0.5578, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.694949494949495, | |
| "grad_norm": 9.0, | |
| "learning_rate": 8.245091444808663e-06, | |
| "loss": 1.0137, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.6965656565656566, | |
| "grad_norm": 7.34375, | |
| "learning_rate": 8.234331933709353e-06, | |
| "loss": 0.5748, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.6981818181818182, | |
| "grad_norm": 7.15625, | |
| "learning_rate": 8.223546607470863e-06, | |
| "loss": 0.6973, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.6997979797979798, | |
| "grad_norm": 8.375, | |
| "learning_rate": 8.212735552177347e-06, | |
| "loss": 0.6268, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.7014141414141414, | |
| "grad_norm": 6.875, | |
| "learning_rate": 8.201898854118301e-06, | |
| "loss": 0.6871, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.703030303030303, | |
| "grad_norm": 9.5, | |
| "learning_rate": 8.191036599787908e-06, | |
| "loss": 0.8489, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.7046464646464646, | |
| "grad_norm": 6.875, | |
| "learning_rate": 8.18014887588431e-06, | |
| "loss": 0.6502, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.7062626262626263, | |
| "grad_norm": 6.21875, | |
| "learning_rate": 8.169235769308959e-06, | |
| "loss": 0.5236, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.7078787878787879, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 8.158297367165885e-06, | |
| "loss": 0.8233, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.7094949494949495, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 8.147333756761027e-06, | |
| "loss": 0.7471, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.7111111111111111, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 8.13634502560152e-06, | |
| "loss": 0.627, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.7127272727272728, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 8.125331261395004e-06, | |
| "loss": 0.6162, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.7143434343434344, | |
| "grad_norm": 9.25, | |
| "learning_rate": 8.114292552048925e-06, | |
| "loss": 0.7379, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.7159595959595959, | |
| "grad_norm": 6.125, | |
| "learning_rate": 8.10322898566983e-06, | |
| "loss": 0.6932, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.7175757575757575, | |
| "grad_norm": 6.0, | |
| "learning_rate": 8.092140650562665e-06, | |
| "loss": 0.5414, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.7191919191919192, | |
| "grad_norm": 6.8125, | |
| "learning_rate": 8.08102763523007e-06, | |
| "loss": 0.6522, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.7208080808080808, | |
| "grad_norm": 6.15625, | |
| "learning_rate": 8.069890028371672e-06, | |
| "loss": 0.5511, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.7224242424242424, | |
| "grad_norm": 8.5, | |
| "learning_rate": 8.058727918883376e-06, | |
| "loss": 0.9194, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.724040404040404, | |
| "grad_norm": 6.15625, | |
| "learning_rate": 8.047541395856661e-06, | |
| "loss": 0.5101, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.7256565656565657, | |
| "grad_norm": 6.78125, | |
| "learning_rate": 8.036330548577866e-06, | |
| "loss": 0.664, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 7.21875, | |
| "learning_rate": 8.025095466527468e-06, | |
| "loss": 0.6514, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.7288888888888889, | |
| "grad_norm": 8.25, | |
| "learning_rate": 8.013836239379388e-06, | |
| "loss": 0.8009, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.7305050505050505, | |
| "grad_norm": 6.59375, | |
| "learning_rate": 8.002552957000254e-06, | |
| "loss": 0.5222, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.7321212121212122, | |
| "grad_norm": 6.71875, | |
| "learning_rate": 7.991245709448697e-06, | |
| "loss": 0.6088, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.7337373737373737, | |
| "grad_norm": 7.65625, | |
| "learning_rate": 7.979914586974628e-06, | |
| "loss": 0.8188, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.7353535353535353, | |
| "grad_norm": 6.96875, | |
| "learning_rate": 7.968559680018518e-06, | |
| "loss": 0.696, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.7369696969696969, | |
| "grad_norm": 7.90625, | |
| "learning_rate": 7.957181079210676e-06, | |
| "loss": 0.7024, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.7385858585858586, | |
| "grad_norm": 9.25, | |
| "learning_rate": 7.945778875370527e-06, | |
| "loss": 0.9176, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.7402020202020202, | |
| "grad_norm": 7.46875, | |
| "learning_rate": 7.934353159505885e-06, | |
| "loss": 0.9633, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.7418181818181818, | |
| "grad_norm": 7.28125, | |
| "learning_rate": 7.922904022812228e-06, | |
| "loss": 0.6307, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.7434343434343434, | |
| "grad_norm": 6.84375, | |
| "learning_rate": 7.911431556671967e-06, | |
| "loss": 0.6905, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.7450505050505051, | |
| "grad_norm": 6.875, | |
| "learning_rate": 7.899935852653723e-06, | |
| "loss": 0.4554, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.7466666666666667, | |
| "grad_norm": 6.21875, | |
| "learning_rate": 7.888417002511592e-06, | |
| "loss": 0.5635, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.7482828282828283, | |
| "grad_norm": 8.875, | |
| "learning_rate": 7.876875098184411e-06, | |
| "loss": 1.0392, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.74989898989899, | |
| "grad_norm": 5.875, | |
| "learning_rate": 7.865310231795026e-06, | |
| "loss": 0.5778, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.7515151515151515, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 7.853722495649558e-06, | |
| "loss": 0.6214, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.7531313131313131, | |
| "grad_norm": 7.1875, | |
| "learning_rate": 7.842111982236669e-06, | |
| "loss": 0.8281, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.7547474747474747, | |
| "grad_norm": 7.9375, | |
| "learning_rate": 7.83047878422681e-06, | |
| "loss": 0.7688, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.7563636363636363, | |
| "grad_norm": 9.0625, | |
| "learning_rate": 7.818822994471504e-06, | |
| "loss": 0.6517, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.757979797979798, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 7.807144706002582e-06, | |
| "loss": 0.7734, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.7595959595959596, | |
| "grad_norm": 7.4375, | |
| "learning_rate": 7.79544401203146e-06, | |
| "loss": 0.7312, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.7612121212121212, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 7.783721005948374e-06, | |
| "loss": 0.8063, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.7628282828282829, | |
| "grad_norm": 7.96875, | |
| "learning_rate": 7.771975781321655e-06, | |
| "loss": 0.6811, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.7644444444444445, | |
| "grad_norm": 6.4375, | |
| "learning_rate": 7.760208431896971e-06, | |
| "loss": 0.866, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.7660606060606061, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 7.748419051596586e-06, | |
| "loss": 0.8808, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.7676767676767676, | |
| "grad_norm": 7.9375, | |
| "learning_rate": 7.736607734518593e-06, | |
| "loss": 0.9093, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.7692929292929293, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 7.72477457493619e-06, | |
| "loss": 0.7615, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.7709090909090909, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 7.712919667296902e-06, | |
| "loss": 0.5422, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.7725252525252525, | |
| "grad_norm": 5.8125, | |
| "learning_rate": 7.701043106221847e-06, | |
| "loss": 0.5013, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.7741414141414141, | |
| "grad_norm": 9.25, | |
| "learning_rate": 7.689144986504966e-06, | |
| "loss": 0.8879, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.7757575757575758, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 7.677225403112277e-06, | |
| "loss": 0.5958, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.7773737373737374, | |
| "grad_norm": 6.59375, | |
| "learning_rate": 7.665284451181106e-06, | |
| "loss": 0.5568, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.778989898989899, | |
| "grad_norm": 8.25, | |
| "learning_rate": 7.653322226019341e-06, | |
| "loss": 0.8891, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.7806060606060606, | |
| "grad_norm": 6.59375, | |
| "learning_rate": 7.64133882310466e-06, | |
| "loss": 0.6576, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.7822222222222223, | |
| "grad_norm": 7.5, | |
| "learning_rate": 7.629334338083774e-06, | |
| "loss": 0.7541, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.7838383838383839, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 7.617308866771667e-06, | |
| "loss": 0.9125, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.7854545454545454, | |
| "grad_norm": 7.25, | |
| "learning_rate": 7.605262505150819e-06, | |
| "loss": 0.7351, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.787070707070707, | |
| "grad_norm": 6.625, | |
| "learning_rate": 7.5931953493704545e-06, | |
| "loss": 0.5468, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.7886868686868687, | |
| "grad_norm": 7.21875, | |
| "learning_rate": 7.58110749574577e-06, | |
| "loss": 0.5995, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.7903030303030303, | |
| "grad_norm": 7.78125, | |
| "learning_rate": 7.568999040757157e-06, | |
| "loss": 0.6156, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.7919191919191919, | |
| "grad_norm": 6.46875, | |
| "learning_rate": 7.556870081049444e-06, | |
| "loss": 0.7599, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.7935353535353535, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 7.5447207134311195e-06, | |
| "loss": 0.7025, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.7951515151515152, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 7.532551034873558e-06, | |
| "loss": 0.5702, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.7951515151515152, | |
| "eval_loss": 0.7591387629508972, | |
| "eval_runtime": 40.0722, | |
| "eval_samples_per_second": 27.45, | |
| "eval_steps_per_second": 3.444, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.7967676767676768, | |
| "grad_norm": 7.28125, | |
| "learning_rate": 7.520361142510248e-06, | |
| "loss": 0.6548, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.7983838383838384, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 7.50815113363602e-06, | |
| "loss": 0.8229, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 7.59375, | |
| "learning_rate": 7.49592110570626e-06, | |
| "loss": 1.0282, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.8016161616161617, | |
| "grad_norm": 9.375, | |
| "learning_rate": 7.483671156336142e-06, | |
| "loss": 0.9156, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.8032323232323232, | |
| "grad_norm": 7.25, | |
| "learning_rate": 7.471401383299847e-06, | |
| "loss": 0.6175, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.8048484848484848, | |
| "grad_norm": 7.78125, | |
| "learning_rate": 7.459111884529775e-06, | |
| "loss": 0.7255, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.8064646464646464, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 7.446802758115775e-06, | |
| "loss": 0.8077, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.8080808080808081, | |
| "grad_norm": 5.34375, | |
| "learning_rate": 7.43447410230435e-06, | |
| "loss": 0.7292, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.8096969696969697, | |
| "grad_norm": 7.71875, | |
| "learning_rate": 7.422126015497884e-06, | |
| "loss": 0.7707, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.8113131313131313, | |
| "grad_norm": 5.53125, | |
| "learning_rate": 7.409758596253849e-06, | |
| "loss": 0.5303, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.812929292929293, | |
| "grad_norm": 6.59375, | |
| "learning_rate": 7.397371943284017e-06, | |
| "loss": 0.4893, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.8145454545454546, | |
| "grad_norm": 7.625, | |
| "learning_rate": 7.384966155453686e-06, | |
| "loss": 0.6416, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.8161616161616162, | |
| "grad_norm": 7.8125, | |
| "learning_rate": 7.372541331780871e-06, | |
| "loss": 0.8562, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.8177777777777778, | |
| "grad_norm": 6.71875, | |
| "learning_rate": 7.360097571435527e-06, | |
| "loss": 0.6703, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.8193939393939393, | |
| "grad_norm": 7.6875, | |
| "learning_rate": 7.347634973738753e-06, | |
| "loss": 1.2725, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.821010101010101, | |
| "grad_norm": 6.125, | |
| "learning_rate": 7.335153638162005e-06, | |
| "loss": 0.6887, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.8226262626262626, | |
| "grad_norm": 6.84375, | |
| "learning_rate": 7.322653664326289e-06, | |
| "loss": 0.7664, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.8242424242424242, | |
| "grad_norm": 8.25, | |
| "learning_rate": 7.310135152001381e-06, | |
| "loss": 0.9012, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.8258585858585858, | |
| "grad_norm": 6.96875, | |
| "learning_rate": 7.29759820110502e-06, | |
| "loss": 0.6616, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.8274747474747475, | |
| "grad_norm": 7.5625, | |
| "learning_rate": 7.285042911702116e-06, | |
| "loss": 0.6404, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.8290909090909091, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 7.27246938400395e-06, | |
| "loss": 0.437, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.8307070707070707, | |
| "grad_norm": 8.75, | |
| "learning_rate": 7.259877718367372e-06, | |
| "loss": 0.7375, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.8323232323232324, | |
| "grad_norm": 9.0, | |
| "learning_rate": 7.2472680152940015e-06, | |
| "loss": 0.5881, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.833939393939394, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 7.234640375429427e-06, | |
| "loss": 0.628, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.8355555555555556, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 7.2219948995624035e-06, | |
| "loss": 0.7878, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.8371717171717171, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 7.209331688624039e-06, | |
| "loss": 0.5814, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.8387878787878787, | |
| "grad_norm": 7.71875, | |
| "learning_rate": 7.1966508436870044e-06, | |
| "loss": 1.1205, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.8404040404040404, | |
| "grad_norm": 7.34375, | |
| "learning_rate": 7.183952465964711e-06, | |
| "loss": 0.8673, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.842020202020202, | |
| "grad_norm": 6.375, | |
| "learning_rate": 7.171236656810513e-06, | |
| "loss": 0.4973, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.8436363636363636, | |
| "grad_norm": 7.59375, | |
| "learning_rate": 7.158503517716894e-06, | |
| "loss": 0.6818, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.8452525252525253, | |
| "grad_norm": 8.375, | |
| "learning_rate": 7.145753150314661e-06, | |
| "loss": 0.763, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.8468686868686869, | |
| "grad_norm": 8.4375, | |
| "learning_rate": 7.132985656372126e-06, | |
| "loss": 0.7879, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.8484848484848485, | |
| "grad_norm": 6.25, | |
| "learning_rate": 7.120201137794301e-06, | |
| "loss": 0.3823, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.8501010101010101, | |
| "grad_norm": 5.84375, | |
| "learning_rate": 7.1073996966220835e-06, | |
| "loss": 0.4849, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.8517171717171718, | |
| "grad_norm": 6.625, | |
| "learning_rate": 7.0945814350314346e-06, | |
| "loss": 0.5307, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.8533333333333334, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 7.0817464553325764e-06, | |
| "loss": 0.8415, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.8549494949494949, | |
| "grad_norm": 7.5, | |
| "learning_rate": 7.068894859969162e-06, | |
| "loss": 0.9093, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.8565656565656565, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 7.0560267515174685e-06, | |
| "loss": 0.6398, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.8581818181818182, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 7.043142232685572e-06, | |
| "loss": 0.7314, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.8597979797979798, | |
| "grad_norm": 7.0, | |
| "learning_rate": 7.030241406312528e-06, | |
| "loss": 0.6262, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.8614141414141414, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 7.0173243753675544e-06, | |
| "loss": 0.7267, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.863030303030303, | |
| "grad_norm": 6.28125, | |
| "learning_rate": 7.004391242949209e-06, | |
| "loss": 0.5482, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.8646464646464647, | |
| "grad_norm": 7.46875, | |
| "learning_rate": 6.991442112284561e-06, | |
| "loss": 0.8512, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.8662626262626263, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 6.978477086728375e-06, | |
| "loss": 0.4953, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.8678787878787879, | |
| "grad_norm": 7.71875, | |
| "learning_rate": 6.96549626976228e-06, | |
| "loss": 0.4961, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.8694949494949495, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 6.952499764993945e-06, | |
| "loss": 0.9078, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.8711111111111111, | |
| "grad_norm": 7.59375, | |
| "learning_rate": 6.9394876761562555e-06, | |
| "loss": 0.631, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.8727272727272727, | |
| "grad_norm": 7.5625, | |
| "learning_rate": 6.926460107106483e-06, | |
| "loss": 0.6733, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.8743434343434343, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 6.913417161825449e-06, | |
| "loss": 0.8478, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.8759595959595959, | |
| "grad_norm": 5.71875, | |
| "learning_rate": 6.900358944416712e-06, | |
| "loss": 0.503, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.8775757575757576, | |
| "grad_norm": 5.84375, | |
| "learning_rate": 6.887285559105721e-06, | |
| "loss": 0.8665, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.8791919191919192, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 6.874197110238986e-06, | |
| "loss": 0.5614, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.8808080808080808, | |
| "grad_norm": 6.53125, | |
| "learning_rate": 6.861093702283254e-06, | |
| "loss": 0.7636, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.8824242424242424, | |
| "grad_norm": 11.6875, | |
| "learning_rate": 6.847975439824669e-06, | |
| "loss": 0.74, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.8840404040404041, | |
| "grad_norm": 7.46875, | |
| "learning_rate": 6.834842427567934e-06, | |
| "loss": 0.7544, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.8856565656565657, | |
| "grad_norm": 6.4375, | |
| "learning_rate": 6.8216947703354815e-06, | |
| "loss": 0.8037, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.8872727272727273, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 6.808532573066635e-06, | |
| "loss": 0.7922, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 6.78125, | |
| "learning_rate": 6.795355940816768e-06, | |
| "loss": 0.5742, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.8905050505050505, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 6.7821649787564706e-06, | |
| "loss": 0.6695, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.8921212121212121, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 6.7689597921707065e-06, | |
| "loss": 0.8931, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.8937373737373737, | |
| "grad_norm": 7.6875, | |
| "learning_rate": 6.755740486457973e-06, | |
| "loss": 0.8255, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.8953535353535353, | |
| "grad_norm": 7.65625, | |
| "learning_rate": 6.742507167129465e-06, | |
| "loss": 0.5539, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.896969696969697, | |
| "grad_norm": 7.9375, | |
| "learning_rate": 6.729259939808223e-06, | |
| "loss": 0.6744, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.8985858585858586, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 6.715998910228296e-06, | |
| "loss": 0.5843, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.9002020202020202, | |
| "grad_norm": 6.4375, | |
| "learning_rate": 6.702724184233904e-06, | |
| "loss": 0.6715, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.9018181818181819, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 6.689435867778578e-06, | |
| "loss": 0.5671, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.9034343434343435, | |
| "grad_norm": 11.875, | |
| "learning_rate": 6.676134066924325e-06, | |
| "loss": 0.8063, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 0.9050505050505051, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 6.6628188878407806e-06, | |
| "loss": 0.4545, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.9066666666666666, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 6.64949043680436e-06, | |
| "loss": 0.7163, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 0.9082828282828282, | |
| "grad_norm": 8.4375, | |
| "learning_rate": 6.636148820197409e-06, | |
| "loss": 0.8048, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 0.9098989898989899, | |
| "grad_norm": 7.65625, | |
| "learning_rate": 6.622794144507356e-06, | |
| "loss": 0.7068, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 0.9115151515151515, | |
| "grad_norm": 8.75, | |
| "learning_rate": 6.609426516325859e-06, | |
| "loss": 0.5749, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.9131313131313131, | |
| "grad_norm": 7.375, | |
| "learning_rate": 6.596046042347964e-06, | |
| "loss": 0.5476, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.9147474747474748, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 6.58265282937124e-06, | |
| "loss": 0.9369, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 0.9163636363636364, | |
| "grad_norm": 9.75, | |
| "learning_rate": 6.569246984294938e-06, | |
| "loss": 0.9137, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.917979797979798, | |
| "grad_norm": 7.75, | |
| "learning_rate": 6.555828614119132e-06, | |
| "loss": 0.6025, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.9195959595959596, | |
| "grad_norm": 7.28125, | |
| "learning_rate": 6.542397825943867e-06, | |
| "loss": 0.6785, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 0.9212121212121213, | |
| "grad_norm": 8.5, | |
| "learning_rate": 6.528954726968302e-06, | |
| "loss": 0.7209, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.9228282828282828, | |
| "grad_norm": 6.8125, | |
| "learning_rate": 6.515499424489857e-06, | |
| "loss": 0.5772, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 0.9244444444444444, | |
| "grad_norm": 7.9375, | |
| "learning_rate": 6.502032025903356e-06, | |
| "loss": 0.7312, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.926060606060606, | |
| "grad_norm": 7.75, | |
| "learning_rate": 6.4885526387001654e-06, | |
| "loss": 0.8718, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 0.9276767676767677, | |
| "grad_norm": 6.25, | |
| "learning_rate": 6.475061370467346e-06, | |
| "loss": 0.7196, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.9292929292929293, | |
| "grad_norm": 7.84375, | |
| "learning_rate": 6.461558328886786e-06, | |
| "loss": 0.6602, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.9309090909090909, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 6.4480436217343366e-06, | |
| "loss": 0.6234, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.9325252525252525, | |
| "grad_norm": 7.25, | |
| "learning_rate": 6.434517356878974e-06, | |
| "loss": 0.5093, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 0.9341414141414142, | |
| "grad_norm": 8.875, | |
| "learning_rate": 6.420979642281909e-06, | |
| "loss": 0.7573, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 0.9357575757575758, | |
| "grad_norm": 7.625, | |
| "learning_rate": 6.4074305859957475e-06, | |
| "loss": 1.0258, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 0.9373737373737374, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 6.393870296163616e-06, | |
| "loss": 0.7432, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.938989898989899, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 6.380298881018307e-06, | |
| "loss": 1.0014, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.9406060606060606, | |
| "grad_norm": 7.4375, | |
| "learning_rate": 6.366716448881407e-06, | |
| "loss": 0.7915, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.9422222222222222, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 6.353123108162435e-06, | |
| "loss": 0.5849, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 0.9438383838383838, | |
| "grad_norm": 5.65625, | |
| "learning_rate": 6.339518967357985e-06, | |
| "loss": 0.785, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.9454545454545454, | |
| "grad_norm": 7.125, | |
| "learning_rate": 6.325904135050844e-06, | |
| "loss": 0.5567, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.9470707070707071, | |
| "grad_norm": 7.125, | |
| "learning_rate": 6.312278719909138e-06, | |
| "loss": 0.8426, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 0.9486868686868687, | |
| "grad_norm": 6.625, | |
| "learning_rate": 6.298642830685464e-06, | |
| "loss": 0.6369, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 0.9503030303030303, | |
| "grad_norm": 7.34375, | |
| "learning_rate": 6.284996576216014e-06, | |
| "loss": 0.6441, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.9519191919191919, | |
| "grad_norm": 7.0, | |
| "learning_rate": 6.271340065419715e-06, | |
| "loss": 0.7032, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 0.9535353535353536, | |
| "grad_norm": 6.71875, | |
| "learning_rate": 6.257673407297352e-06, | |
| "loss": 0.5248, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.9551515151515152, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 6.243996710930705e-06, | |
| "loss": 0.5664, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 0.9567676767676768, | |
| "grad_norm": 7.3125, | |
| "learning_rate": 6.230310085481677e-06, | |
| "loss": 0.4842, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.9583838383838383, | |
| "grad_norm": 6.28125, | |
| "learning_rate": 6.216613640191414e-06, | |
| "loss": 0.5453, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 8.0, | |
| "learning_rate": 6.2029074843794445e-06, | |
| "loss": 0.6635, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.9616161616161616, | |
| "grad_norm": 6.71875, | |
| "learning_rate": 6.189191727442807e-06, | |
| "loss": 0.6053, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.9632323232323232, | |
| "grad_norm": 6.15625, | |
| "learning_rate": 6.175466478855161e-06, | |
| "loss": 0.5201, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.9648484848484848, | |
| "grad_norm": 6.53125, | |
| "learning_rate": 6.1617318481659364e-06, | |
| "loss": 0.6634, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 0.9664646464646465, | |
| "grad_norm": 7.90625, | |
| "learning_rate": 6.147987944999441e-06, | |
| "loss": 0.7641, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 0.9680808080808081, | |
| "grad_norm": 7.3125, | |
| "learning_rate": 6.13423487905399e-06, | |
| "loss": 0.7179, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 0.9696969696969697, | |
| "grad_norm": 6.53125, | |
| "learning_rate": 6.1204727601010396e-06, | |
| "loss": 0.7291, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.9713131313131314, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 6.106701697984296e-06, | |
| "loss": 0.7674, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 0.972929292929293, | |
| "grad_norm": 7.1875, | |
| "learning_rate": 6.092921802618849e-06, | |
| "loss": 0.7901, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.9745454545454545, | |
| "grad_norm": 5.75, | |
| "learning_rate": 6.079133183990292e-06, | |
| "loss": 0.6376, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.9761616161616161, | |
| "grad_norm": 7.125, | |
| "learning_rate": 6.065335952153846e-06, | |
| "loss": 0.7688, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.9777777777777777, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 6.051530217233473e-06, | |
| "loss": 0.7441, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.9793939393939394, | |
| "grad_norm": 7.21875, | |
| "learning_rate": 6.037716089421011e-06, | |
| "loss": 0.5193, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.981010101010101, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 6.023893678975281e-06, | |
| "loss": 0.6732, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 0.9826262626262626, | |
| "grad_norm": 6.625, | |
| "learning_rate": 6.010063096221215e-06, | |
| "loss": 1.2178, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.9842424242424243, | |
| "grad_norm": 6.0, | |
| "learning_rate": 5.996224451548974e-06, | |
| "loss": 0.432, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.9858585858585859, | |
| "grad_norm": 6.875, | |
| "learning_rate": 5.982377855413063e-06, | |
| "loss": 0.6985, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.9874747474747475, | |
| "grad_norm": 6.15625, | |
| "learning_rate": 5.968523418331453e-06, | |
| "loss": 0.4962, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 0.9890909090909091, | |
| "grad_norm": 6.0, | |
| "learning_rate": 5.954661250884704e-06, | |
| "loss": 0.5834, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.9907070707070708, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 5.940791463715068e-06, | |
| "loss": 0.7957, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 0.9923232323232323, | |
| "grad_norm": 8.25, | |
| "learning_rate": 5.926914167525618e-06, | |
| "loss": 0.6498, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 0.9939393939393939, | |
| "grad_norm": 6.96875, | |
| "learning_rate": 5.913029473079365e-06, | |
| "loss": 0.7897, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.9939393939393939, | |
| "eval_loss": 0.6975554823875427, | |
| "eval_runtime": 35.4933, | |
| "eval_samples_per_second": 30.992, | |
| "eval_steps_per_second": 3.888, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.9955555555555555, | |
| "grad_norm": 6.5, | |
| "learning_rate": 5.899137491198364e-06, | |
| "loss": 0.9353, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 0.9971717171717172, | |
| "grad_norm": 5.40625, | |
| "learning_rate": 5.885238332762837e-06, | |
| "loss": 0.6782, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 0.9987878787878788, | |
| "grad_norm": 6.75, | |
| "learning_rate": 5.871332108710292e-06, | |
| "loss": 0.6207, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 1.0004040404040404, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 5.8574189300346215e-06, | |
| "loss": 0.4077, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 1.002020202020202, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 5.843498907785236e-06, | |
| "loss": 0.8129, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.0036363636363637, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 5.829572153066166e-06, | |
| "loss": 0.3507, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 1.0052525252525253, | |
| "grad_norm": 5.8125, | |
| "learning_rate": 5.815638777035175e-06, | |
| "loss": 0.4918, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 1.006868686868687, | |
| "grad_norm": 7.375, | |
| "learning_rate": 5.801698890902877e-06, | |
| "loss": 0.451, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 1.0084848484848485, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 5.78775260593185e-06, | |
| "loss": 0.5601, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 1.0101010101010102, | |
| "grad_norm": 12.6875, | |
| "learning_rate": 5.77380003343574e-06, | |
| "loss": 0.8753, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.0117171717171718, | |
| "grad_norm": 6.28125, | |
| "learning_rate": 5.759841284778379e-06, | |
| "loss": 0.5141, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 1.0133333333333334, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 5.7458764713729e-06, | |
| "loss": 0.5174, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 1.014949494949495, | |
| "grad_norm": 8.875, | |
| "learning_rate": 5.731905704680834e-06, | |
| "loss": 0.3319, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 1.0165656565656567, | |
| "grad_norm": 9.875, | |
| "learning_rate": 5.717929096211232e-06, | |
| "loss": 0.4503, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 1.018181818181818, | |
| "grad_norm": 7.28125, | |
| "learning_rate": 5.703946757519777e-06, | |
| "loss": 0.3714, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.0197979797979797, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 5.689958800207879e-06, | |
| "loss": 0.42, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 1.0214141414141413, | |
| "grad_norm": 10.625, | |
| "learning_rate": 5.6759653359218e-06, | |
| "loss": 0.6338, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 1.023030303030303, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 5.661966476351756e-06, | |
| "loss": 0.4457, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 1.0246464646464646, | |
| "grad_norm": 6.28125, | |
| "learning_rate": 5.647962333231021e-06, | |
| "loss": 0.4263, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 1.0262626262626262, | |
| "grad_norm": 8.5, | |
| "learning_rate": 5.633953018335047e-06, | |
| "loss": 0.6676, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 1.0278787878787878, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 5.6199386434805615e-06, | |
| "loss": 0.4648, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 1.0294949494949495, | |
| "grad_norm": 7.1875, | |
| "learning_rate": 5.605919320524677e-06, | |
| "loss": 0.4956, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 1.031111111111111, | |
| "grad_norm": 6.8125, | |
| "learning_rate": 5.591895161364006e-06, | |
| "loss": 0.4562, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 1.0327272727272727, | |
| "grad_norm": 7.0, | |
| "learning_rate": 5.577866277933752e-06, | |
| "loss": 0.3596, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 1.0343434343434343, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 5.563832782206835e-06, | |
| "loss": 0.5325, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.035959595959596, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 5.549794786192984e-06, | |
| "loss": 0.3813, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 1.0375757575757576, | |
| "grad_norm": 7.15625, | |
| "learning_rate": 5.535752401937846e-06, | |
| "loss": 0.5279, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 1.0391919191919192, | |
| "grad_norm": 6.625, | |
| "learning_rate": 5.521705741522096e-06, | |
| "loss": 0.4679, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 1.0408080808080808, | |
| "grad_norm": 6.5, | |
| "learning_rate": 5.507654917060541e-06, | |
| "loss": 0.396, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 1.0424242424242425, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 5.49360004070122e-06, | |
| "loss": 0.5639, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.044040404040404, | |
| "grad_norm": 5.8125, | |
| "learning_rate": 5.4795412246245126e-06, | |
| "loss": 0.5076, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 1.0456565656565657, | |
| "grad_norm": 6.15625, | |
| "learning_rate": 5.465478581042248e-06, | |
| "loss": 0.4221, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 1.0472727272727274, | |
| "grad_norm": 6.0625, | |
| "learning_rate": 5.451412222196801e-06, | |
| "loss": 0.5823, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 1.048888888888889, | |
| "grad_norm": 7.28125, | |
| "learning_rate": 5.4373422603602025e-06, | |
| "loss": 0.5174, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 1.0505050505050506, | |
| "grad_norm": 7.3125, | |
| "learning_rate": 5.4232688078332384e-06, | |
| "loss": 0.4169, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.0521212121212122, | |
| "grad_norm": 7.625, | |
| "learning_rate": 5.409191976944559e-06, | |
| "loss": 0.8318, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 1.0537373737373736, | |
| "grad_norm": 5.28125, | |
| "learning_rate": 5.395111880049775e-06, | |
| "loss": 0.2674, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 1.0553535353535353, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 5.3810286295305715e-06, | |
| "loss": 0.5487, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 1.056969696969697, | |
| "grad_norm": 7.125, | |
| "learning_rate": 5.366942337793798e-06, | |
| "loss": 0.4007, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 1.0585858585858585, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 5.352853117270581e-06, | |
| "loss": 0.7315, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.0602020202020201, | |
| "grad_norm": 6.5, | |
| "learning_rate": 5.338761080415425e-06, | |
| "loss": 0.502, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 1.0618181818181818, | |
| "grad_norm": 7.3125, | |
| "learning_rate": 5.324666339705307e-06, | |
| "loss": 0.4223, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 1.0634343434343434, | |
| "grad_norm": 7.25, | |
| "learning_rate": 5.310569007638795e-06, | |
| "loss": 0.5229, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 1.065050505050505, | |
| "grad_norm": 7.15625, | |
| "learning_rate": 5.296469196735134e-06, | |
| "loss": 0.7768, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 1.0666666666666667, | |
| "grad_norm": 5.6875, | |
| "learning_rate": 5.28236701953335e-06, | |
| "loss": 0.4436, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.0682828282828283, | |
| "grad_norm": 6.0, | |
| "learning_rate": 5.268262588591364e-06, | |
| "loss": 0.5257, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 1.06989898989899, | |
| "grad_norm": 7.96875, | |
| "learning_rate": 5.254156016485082e-06, | |
| "loss": 0.571, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 1.0715151515151515, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 5.240047415807499e-06, | |
| "loss": 0.5387, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 1.0731313131313132, | |
| "grad_norm": 7.1875, | |
| "learning_rate": 5.225936899167803e-06, | |
| "loss": 0.8149, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 1.0747474747474748, | |
| "grad_norm": 5.6875, | |
| "learning_rate": 5.211824579190473e-06, | |
| "loss": 0.5172, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.0763636363636364, | |
| "grad_norm": 5.03125, | |
| "learning_rate": 5.197710568514381e-06, | |
| "loss": 0.2866, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 1.077979797979798, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 5.183594979791896e-06, | |
| "loss": 0.3989, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 1.0795959595959597, | |
| "grad_norm": 5.09375, | |
| "learning_rate": 5.169477925687981e-06, | |
| "loss": 0.4623, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 1.0812121212121213, | |
| "grad_norm": 6.28125, | |
| "learning_rate": 5.155359518879295e-06, | |
| "loss": 0.535, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 1.082828282828283, | |
| "grad_norm": 6.375, | |
| "learning_rate": 5.141239872053294e-06, | |
| "loss": 0.4243, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.0844444444444445, | |
| "grad_norm": 5.96875, | |
| "learning_rate": 5.1271190979073296e-06, | |
| "loss": 0.4805, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 1.086060606060606, | |
| "grad_norm": 5.375, | |
| "learning_rate": 5.112997309147753e-06, | |
| "loss": 0.3017, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 1.0876767676767676, | |
| "grad_norm": 7.28125, | |
| "learning_rate": 5.098874618489017e-06, | |
| "loss": 0.3953, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 1.0892929292929292, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 5.084751138652764e-06, | |
| "loss": 0.4452, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 5.070626982366943e-06, | |
| "loss": 0.6184, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.0925252525252525, | |
| "grad_norm": 7.3125, | |
| "learning_rate": 5.0565022623649e-06, | |
| "loss": 0.5576, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 1.094141414141414, | |
| "grad_norm": 6.75, | |
| "learning_rate": 5.0423770913844795e-06, | |
| "loss": 0.4086, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 1.0957575757575757, | |
| "grad_norm": 9.625, | |
| "learning_rate": 5.028251582167126e-06, | |
| "loss": 0.6037, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 1.0973737373737373, | |
| "grad_norm": 7.84375, | |
| "learning_rate": 5.0141258474569855e-06, | |
| "loss": 0.6863, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 1.098989898989899, | |
| "grad_norm": 7.4375, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4384, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.1006060606060606, | |
| "grad_norm": 5.96875, | |
| "learning_rate": 4.985874152543015e-06, | |
| "loss": 0.5086, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 1.1022222222222222, | |
| "grad_norm": 7.75, | |
| "learning_rate": 4.971748417832876e-06, | |
| "loss": 0.4695, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 1.1038383838383838, | |
| "grad_norm": 7.78125, | |
| "learning_rate": 4.957622908615522e-06, | |
| "loss": 0.466, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 1.1054545454545455, | |
| "grad_norm": 5.78125, | |
| "learning_rate": 4.943497737635103e-06, | |
| "loss": 0.5558, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 1.107070707070707, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 4.92937301763306e-06, | |
| "loss": 0.3521, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.1086868686868687, | |
| "grad_norm": 7.21875, | |
| "learning_rate": 4.915248861347239e-06, | |
| "loss": 0.5391, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 1.1103030303030303, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 4.901125381510986e-06, | |
| "loss": 0.4364, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 1.111919191919192, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 4.887002690852249e-06, | |
| "loss": 0.5012, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 1.1135353535353536, | |
| "grad_norm": 6.59375, | |
| "learning_rate": 4.872880902092673e-06, | |
| "loss": 0.4982, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 1.1151515151515152, | |
| "grad_norm": 6.84375, | |
| "learning_rate": 4.858760127946707e-06, | |
| "loss": 0.4903, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.1167676767676769, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 4.8446404811207055e-06, | |
| "loss": 0.6549, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 1.1183838383838385, | |
| "grad_norm": 7.0, | |
| "learning_rate": 4.830522074312019e-06, | |
| "loss": 0.6238, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 6.125, | |
| "learning_rate": 4.816405020208105e-06, | |
| "loss": 0.2768, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 1.1216161616161617, | |
| "grad_norm": 5.875, | |
| "learning_rate": 4.80228943148562e-06, | |
| "loss": 0.3653, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 1.1232323232323231, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 4.788175420809528e-06, | |
| "loss": 0.4203, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.1248484848484848, | |
| "grad_norm": 6.8125, | |
| "learning_rate": 4.774063100832199e-06, | |
| "loss": 0.3433, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 1.1264646464646464, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 4.759952584192502e-06, | |
| "loss": 0.3582, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 1.128080808080808, | |
| "grad_norm": 5.84375, | |
| "learning_rate": 4.745843983514919e-06, | |
| "loss": 0.3239, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 1.1296969696969696, | |
| "grad_norm": 7.78125, | |
| "learning_rate": 4.731737411408637e-06, | |
| "loss": 0.6284, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 1.1313131313131313, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 4.717632980466652e-06, | |
| "loss": 0.4641, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.132929292929293, | |
| "grad_norm": 7.96875, | |
| "learning_rate": 4.703530803264868e-06, | |
| "loss": 0.4764, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 1.1345454545454545, | |
| "grad_norm": 6.375, | |
| "learning_rate": 4.6894309923612055e-06, | |
| "loss": 0.3617, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 1.1361616161616162, | |
| "grad_norm": 6.28125, | |
| "learning_rate": 4.675333660294693e-06, | |
| "loss": 0.3734, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 1.1377777777777778, | |
| "grad_norm": 7.5, | |
| "learning_rate": 4.661238919584578e-06, | |
| "loss": 0.789, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 1.1393939393939394, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 4.647146882729422e-06, | |
| "loss": 0.8066, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.141010101010101, | |
| "grad_norm": 7.78125, | |
| "learning_rate": 4.633057662206205e-06, | |
| "loss": 0.5443, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 1.1426262626262627, | |
| "grad_norm": 6.84375, | |
| "learning_rate": 4.618971370469431e-06, | |
| "loss": 0.5255, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 1.1442424242424243, | |
| "grad_norm": 7.65625, | |
| "learning_rate": 4.6048881199502265e-06, | |
| "loss": 0.4648, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 1.145858585858586, | |
| "grad_norm": 7.21875, | |
| "learning_rate": 4.590808023055444e-06, | |
| "loss": 0.4431, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 1.1474747474747475, | |
| "grad_norm": 6.59375, | |
| "learning_rate": 4.576731192166762e-06, | |
| "loss": 0.4793, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.1490909090909092, | |
| "grad_norm": 6.46875, | |
| "learning_rate": 4.562657739639798e-06, | |
| "loss": 0.467, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 1.1507070707070708, | |
| "grad_norm": 5.96875, | |
| "learning_rate": 4.548587777803198e-06, | |
| "loss": 0.6655, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 1.1523232323232324, | |
| "grad_norm": 7.125, | |
| "learning_rate": 4.534521418957753e-06, | |
| "loss": 0.4943, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 1.1539393939393938, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 4.520458775375488e-06, | |
| "loss": 0.4532, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 1.1555555555555554, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 4.506399959298781e-06, | |
| "loss": 0.7173, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.157171717171717, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 4.49234508293946e-06, | |
| "loss": 0.4683, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 1.1587878787878787, | |
| "grad_norm": 5.84375, | |
| "learning_rate": 4.4782942584779046e-06, | |
| "loss": 0.3148, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 1.1604040404040403, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 4.464247598062156e-06, | |
| "loss": 0.4546, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 1.162020202020202, | |
| "grad_norm": 7.0, | |
| "learning_rate": 4.450205213807018e-06, | |
| "loss": 0.4984, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 1.1636363636363636, | |
| "grad_norm": 5.46875, | |
| "learning_rate": 4.436167217793167e-06, | |
| "loss": 0.445, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.1652525252525252, | |
| "grad_norm": 6.53125, | |
| "learning_rate": 4.4221337220662495e-06, | |
| "loss": 0.5306, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 1.1668686868686868, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 4.408104838635996e-06, | |
| "loss": 0.974, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 1.1684848484848485, | |
| "grad_norm": 5.5625, | |
| "learning_rate": 4.394080679475324e-06, | |
| "loss": 0.3139, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 1.17010101010101, | |
| "grad_norm": 6.0625, | |
| "learning_rate": 4.38006135651944e-06, | |
| "loss": 0.5718, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 1.1717171717171717, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 4.366046981664954e-06, | |
| "loss": 0.6332, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.1733333333333333, | |
| "grad_norm": 7.15625, | |
| "learning_rate": 4.3520376667689815e-06, | |
| "loss": 0.7909, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 1.174949494949495, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 4.338033523648247e-06, | |
| "loss": 0.4901, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 1.1765656565656566, | |
| "grad_norm": 6.375, | |
| "learning_rate": 4.3240346640782014e-06, | |
| "loss": 0.4123, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 1.1781818181818182, | |
| "grad_norm": 7.625, | |
| "learning_rate": 4.3100411997921235e-06, | |
| "loss": 0.6332, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 1.1797979797979798, | |
| "grad_norm": 5.71875, | |
| "learning_rate": 4.2960532424802235e-06, | |
| "loss": 0.4957, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.1814141414141415, | |
| "grad_norm": 7.59375, | |
| "learning_rate": 4.282070903788767e-06, | |
| "loss": 0.5822, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 1.183030303030303, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 4.268094295319167e-06, | |
| "loss": 0.549, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 1.1846464646464647, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 4.2541235286271e-06, | |
| "loss": 0.5147, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 1.1862626262626264, | |
| "grad_norm": 7.46875, | |
| "learning_rate": 4.240158715221621e-06, | |
| "loss": 0.6269, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 1.187878787878788, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 4.2261999665642615e-06, | |
| "loss": 0.3747, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.1894949494949496, | |
| "grad_norm": 6.25, | |
| "learning_rate": 4.212247394068151e-06, | |
| "loss": 0.5491, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 1.1911111111111112, | |
| "grad_norm": 5.3125, | |
| "learning_rate": 4.198301109097124e-06, | |
| "loss": 0.4715, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 1.1927272727272726, | |
| "grad_norm": 14.9375, | |
| "learning_rate": 4.184361222964827e-06, | |
| "loss": 0.5872, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 1.1927272727272726, | |
| "eval_loss": 0.676762580871582, | |
| "eval_runtime": 35.4724, | |
| "eval_samples_per_second": 31.01, | |
| "eval_steps_per_second": 3.89, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 1.1943434343434343, | |
| "grad_norm": 6.03125, | |
| "learning_rate": 4.170427846933835e-06, | |
| "loss": 0.4474, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 1.195959595959596, | |
| "grad_norm": 5.78125, | |
| "learning_rate": 4.1565010922147644e-06, | |
| "loss": 0.2681, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.1975757575757575, | |
| "grad_norm": 7.5, | |
| "learning_rate": 4.142581069965379e-06, | |
| "loss": 0.4707, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 1.1991919191919191, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 4.1286678912897095e-06, | |
| "loss": 0.5952, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 1.2008080808080808, | |
| "grad_norm": 6.53125, | |
| "learning_rate": 4.114761667237164e-06, | |
| "loss": 0.4401, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 1.2024242424242424, | |
| "grad_norm": 6.875, | |
| "learning_rate": 4.100862508801639e-06, | |
| "loss": 0.4933, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 1.204040404040404, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 4.086970526920638e-06, | |
| "loss": 0.4491, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.2056565656565656, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 4.0730858324743845e-06, | |
| "loss": 0.2304, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 1.2072727272727273, | |
| "grad_norm": 6.96875, | |
| "learning_rate": 4.059208536284936e-06, | |
| "loss": 0.492, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 1.208888888888889, | |
| "grad_norm": 7.125, | |
| "learning_rate": 4.045338749115299e-06, | |
| "loss": 0.4473, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 1.2105050505050505, | |
| "grad_norm": 7.25, | |
| "learning_rate": 4.031476581668549e-06, | |
| "loss": 0.5541, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 1.2121212121212122, | |
| "grad_norm": 7.3125, | |
| "learning_rate": 4.017622144586938e-06, | |
| "loss": 0.5086, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.2137373737373738, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 4.003775548451028e-06, | |
| "loss": 0.4265, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 1.2153535353535354, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 3.989936903778785e-06, | |
| "loss": 0.5193, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 1.216969696969697, | |
| "grad_norm": 6.84375, | |
| "learning_rate": 3.97610632102472e-06, | |
| "loss": 0.4394, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 1.2185858585858587, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 3.962283910578991e-06, | |
| "loss": 0.7359, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 1.2202020202020203, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 3.948469782766528e-06, | |
| "loss": 0.4886, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 1.221818181818182, | |
| "grad_norm": 7.1875, | |
| "learning_rate": 3.934664047846157e-06, | |
| "loss": 0.4353, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 1.2234343434343433, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 3.9208668160097095e-06, | |
| "loss": 0.5272, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 1.225050505050505, | |
| "grad_norm": 6.03125, | |
| "learning_rate": 3.907078197381153e-06, | |
| "loss": 0.4523, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 1.2266666666666666, | |
| "grad_norm": 10.625, | |
| "learning_rate": 3.893298302015705e-06, | |
| "loss": 0.6854, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 1.2282828282828282, | |
| "grad_norm": 6.75, | |
| "learning_rate": 3.879527239898962e-06, | |
| "loss": 0.3579, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.2298989898989898, | |
| "grad_norm": 5.3125, | |
| "learning_rate": 3.8657651209460104e-06, | |
| "loss": 0.2616, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 1.2315151515151515, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 3.85201205500056e-06, | |
| "loss": 0.5633, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 1.233131313131313, | |
| "grad_norm": 6.03125, | |
| "learning_rate": 3.838268151834065e-06, | |
| "loss": 0.3226, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 1.2347474747474747, | |
| "grad_norm": 13.3125, | |
| "learning_rate": 3.8245335211448404e-06, | |
| "loss": 0.4658, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 1.2363636363636363, | |
| "grad_norm": 6.125, | |
| "learning_rate": 3.810808272557196e-06, | |
| "loss": 0.493, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 1.237979797979798, | |
| "grad_norm": 6.46875, | |
| "learning_rate": 3.797092515620557e-06, | |
| "loss": 0.4253, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 1.2395959595959596, | |
| "grad_norm": 7.28125, | |
| "learning_rate": 3.7833863598085897e-06, | |
| "loss": 0.4679, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 1.2412121212121212, | |
| "grad_norm": 8.5, | |
| "learning_rate": 3.769689914518326e-06, | |
| "loss": 0.7057, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 1.2428282828282828, | |
| "grad_norm": 8.25, | |
| "learning_rate": 3.7560032890692967e-06, | |
| "loss": 0.4716, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 1.2444444444444445, | |
| "grad_norm": 7.9375, | |
| "learning_rate": 3.7423265927026473e-06, | |
| "loss": 0.9959, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.246060606060606, | |
| "grad_norm": 5.65625, | |
| "learning_rate": 3.728659934580286e-06, | |
| "loss": 0.3721, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 1.2476767676767677, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 3.715003423783986e-06, | |
| "loss": 0.454, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 1.2492929292929293, | |
| "grad_norm": 11.25, | |
| "learning_rate": 3.701357169314536e-06, | |
| "loss": 0.7889, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 1.250909090909091, | |
| "grad_norm": 6.5, | |
| "learning_rate": 3.6877212800908625e-06, | |
| "loss": 0.8363, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 1.2525252525252526, | |
| "grad_norm": 7.59375, | |
| "learning_rate": 3.6740958649491575e-06, | |
| "loss": 0.8208, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.2541414141414142, | |
| "grad_norm": 6.4375, | |
| "learning_rate": 3.660481032642016e-06, | |
| "loss": 0.5369, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 1.2557575757575759, | |
| "grad_norm": 6.71875, | |
| "learning_rate": 3.646876891837566e-06, | |
| "loss": 0.3812, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 1.2573737373737375, | |
| "grad_norm": 5.96875, | |
| "learning_rate": 3.633283551118595e-06, | |
| "loss": 0.401, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 1.258989898989899, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 3.6197011189816945e-06, | |
| "loss": 0.5256, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 1.2606060606060607, | |
| "grad_norm": 5.5, | |
| "learning_rate": 3.6061297038363853e-06, | |
| "loss": 0.3921, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.2622222222222224, | |
| "grad_norm": 7.28125, | |
| "learning_rate": 3.5925694140042546e-06, | |
| "loss": 0.5519, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 1.2638383838383838, | |
| "grad_norm": 6.03125, | |
| "learning_rate": 3.579020357718092e-06, | |
| "loss": 0.6655, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 1.2654545454545454, | |
| "grad_norm": 6.0625, | |
| "learning_rate": 3.5654826431210273e-06, | |
| "loss": 0.7744, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 1.267070707070707, | |
| "grad_norm": 5.5625, | |
| "learning_rate": 3.5519563782656642e-06, | |
| "loss": 0.4487, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 1.2686868686868686, | |
| "grad_norm": 5.625, | |
| "learning_rate": 3.5384416711132173e-06, | |
| "loss": 0.449, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 1.2703030303030303, | |
| "grad_norm": 7.5, | |
| "learning_rate": 3.524938629532656e-06, | |
| "loss": 0.5011, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 1.271919191919192, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 3.5114473612998367e-06, | |
| "loss": 0.5925, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 1.2735353535353535, | |
| "grad_norm": 7.46875, | |
| "learning_rate": 3.497967974096647e-06, | |
| "loss": 0.4433, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 1.2751515151515151, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 3.4845005755101446e-06, | |
| "loss": 0.4608, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 1.2767676767676768, | |
| "grad_norm": 7.28125, | |
| "learning_rate": 3.4710452730316978e-06, | |
| "loss": 0.5457, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.2783838383838384, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 3.457602174056134e-06, | |
| "loss": 0.5484, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 3.4441713858808684e-06, | |
| "loss": 0.4372, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 1.2816161616161617, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 3.4307530157050624e-06, | |
| "loss": 0.4388, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 1.2832323232323233, | |
| "grad_norm": 6.53125, | |
| "learning_rate": 3.4173471706287607e-06, | |
| "loss": 0.7221, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 1.284848484848485, | |
| "grad_norm": 6.0625, | |
| "learning_rate": 3.4039539576520376e-06, | |
| "loss": 0.4252, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 1.2864646464646465, | |
| "grad_norm": 6.125, | |
| "learning_rate": 3.3905734836741415e-06, | |
| "loss": 0.5235, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 1.2880808080808082, | |
| "grad_norm": 7.0, | |
| "learning_rate": 3.3772058554926456e-06, | |
| "loss": 0.5727, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 1.2896969696969696, | |
| "grad_norm": 5.0, | |
| "learning_rate": 3.3638511798025927e-06, | |
| "loss": 0.4155, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 1.2913131313131312, | |
| "grad_norm": 7.8125, | |
| "learning_rate": 3.350509563195641e-06, | |
| "loss": 0.4508, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 1.2929292929292928, | |
| "grad_norm": 7.15625, | |
| "learning_rate": 3.3371811121592203e-06, | |
| "loss": 0.5727, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.2945454545454544, | |
| "grad_norm": 5.75, | |
| "learning_rate": 3.3238659330756774e-06, | |
| "loss": 0.4035, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 1.296161616161616, | |
| "grad_norm": 8.0, | |
| "learning_rate": 3.3105641322214238e-06, | |
| "loss": 0.7284, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 1.2977777777777777, | |
| "grad_norm": 5.75, | |
| "learning_rate": 3.2972758157660966e-06, | |
| "loss": 0.3584, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 1.2993939393939393, | |
| "grad_norm": 5.125, | |
| "learning_rate": 3.2840010897717045e-06, | |
| "loss": 0.4011, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 1.301010101010101, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 3.270740060191779e-06, | |
| "loss": 0.4707, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 1.3026262626262626, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 3.257492832870537e-06, | |
| "loss": 0.5322, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 1.3042424242424242, | |
| "grad_norm": 7.375, | |
| "learning_rate": 3.2442595135420287e-06, | |
| "loss": 0.5998, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 1.3058585858585858, | |
| "grad_norm": 6.03125, | |
| "learning_rate": 3.2310402078292956e-06, | |
| "loss": 0.4243, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 1.3074747474747475, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 3.2178350212435315e-06, | |
| "loss": 0.3165, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 1.309090909090909, | |
| "grad_norm": 5.71875, | |
| "learning_rate": 3.2046440591832323e-06, | |
| "loss": 0.5172, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.3107070707070707, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 3.191467426933365e-06, | |
| "loss": 0.5222, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 1.3123232323232323, | |
| "grad_norm": 6.8125, | |
| "learning_rate": 3.178305229664519e-06, | |
| "loss": 0.4126, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 1.313939393939394, | |
| "grad_norm": 5.0625, | |
| "learning_rate": 3.1651575724320665e-06, | |
| "loss": 0.2886, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 1.3155555555555556, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 3.152024560175332e-06, | |
| "loss": 0.3833, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 1.3171717171717172, | |
| "grad_norm": 6.59375, | |
| "learning_rate": 3.1389062977167474e-06, | |
| "loss": 0.3823, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 1.3187878787878788, | |
| "grad_norm": 7.5, | |
| "learning_rate": 3.125802889761016e-06, | |
| "loss": 0.5198, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 1.3204040404040405, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 3.112714440894281e-06, | |
| "loss": 0.3382, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 1.322020202020202, | |
| "grad_norm": 8.125, | |
| "learning_rate": 3.099641055583289e-06, | |
| "loss": 0.6866, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 1.3236363636363637, | |
| "grad_norm": 6.5, | |
| "learning_rate": 3.0865828381745515e-06, | |
| "loss": 0.3631, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 1.3252525252525253, | |
| "grad_norm": 9.375, | |
| "learning_rate": 3.073539892893519e-06, | |
| "loss": 0.5295, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.326868686868687, | |
| "grad_norm": 7.34375, | |
| "learning_rate": 3.0605123238437457e-06, | |
| "loss": 0.4532, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 1.3284848484848486, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 3.047500235006056e-06, | |
| "loss": 0.4369, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 1.3301010101010102, | |
| "grad_norm": 6.625, | |
| "learning_rate": 3.0345037302377216e-06, | |
| "loss": 0.3557, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 1.3317171717171719, | |
| "grad_norm": 7.84375, | |
| "learning_rate": 3.021522913271627e-06, | |
| "loss": 0.7774, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 6.375, | |
| "learning_rate": 3.008557887715441e-06, | |
| "loss": 0.5469, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.3349494949494949, | |
| "grad_norm": 6.84375, | |
| "learning_rate": 2.995608757050793e-06, | |
| "loss": 0.5189, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 1.3365656565656565, | |
| "grad_norm": 6.84375, | |
| "learning_rate": 2.9826756246324477e-06, | |
| "loss": 0.4486, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 1.3381818181818181, | |
| "grad_norm": 7.1875, | |
| "learning_rate": 2.969758593687475e-06, | |
| "loss": 0.9107, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 1.3397979797979798, | |
| "grad_norm": 11.375, | |
| "learning_rate": 2.95685776731443e-06, | |
| "loss": 0.9758, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 1.3414141414141414, | |
| "grad_norm": 8.625, | |
| "learning_rate": 2.9439732484825323e-06, | |
| "loss": 0.3861, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.343030303030303, | |
| "grad_norm": 7.4375, | |
| "learning_rate": 2.931105140030838e-06, | |
| "loss": 0.3662, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 1.3446464646464646, | |
| "grad_norm": 6.75, | |
| "learning_rate": 2.9182535446674244e-06, | |
| "loss": 0.4409, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 1.3462626262626263, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 2.905418564968565e-06, | |
| "loss": 0.4852, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 1.347878787878788, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 2.8926003033779194e-06, | |
| "loss": 0.6186, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 1.3494949494949495, | |
| "grad_norm": 13.4375, | |
| "learning_rate": 2.8797988622057005e-06, | |
| "loss": 0.3766, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 1.3511111111111112, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 2.8670143436278757e-06, | |
| "loss": 0.5832, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 1.3527272727272728, | |
| "grad_norm": 13.625, | |
| "learning_rate": 2.8542468496853404e-06, | |
| "loss": 0.3591, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 1.3543434343434344, | |
| "grad_norm": 8.625, | |
| "learning_rate": 2.8414964822831063e-06, | |
| "loss": 0.7896, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 1.355959595959596, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 2.8287633431894877e-06, | |
| "loss": 0.8662, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 1.3575757575757577, | |
| "grad_norm": 8.75, | |
| "learning_rate": 2.8160475340352913e-06, | |
| "loss": 0.4789, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.359191919191919, | |
| "grad_norm": 6.78125, | |
| "learning_rate": 2.8033491563129977e-06, | |
| "loss": 0.5332, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 1.3608080808080807, | |
| "grad_norm": 6.78125, | |
| "learning_rate": 2.790668311375962e-06, | |
| "loss": 0.5736, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 1.3624242424242423, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 2.7780051004375986e-06, | |
| "loss": 0.3348, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 1.364040404040404, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 2.765359624570574e-06, | |
| "loss": 0.5082, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 1.3656565656565656, | |
| "grad_norm": 11.125, | |
| "learning_rate": 2.752731984706e-06, | |
| "loss": 0.4075, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 1.3672727272727272, | |
| "grad_norm": 8.5, | |
| "learning_rate": 2.7401222816326316e-06, | |
| "loss": 0.5683, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 1.3688888888888888, | |
| "grad_norm": 5.34375, | |
| "learning_rate": 2.7275306159960522e-06, | |
| "loss": 0.2826, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 1.3705050505050504, | |
| "grad_norm": 6.28125, | |
| "learning_rate": 2.714957088297886e-06, | |
| "loss": 0.3661, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 1.372121212121212, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 2.7024017988949802e-06, | |
| "loss": 0.9216, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 1.3737373737373737, | |
| "grad_norm": 7.5, | |
| "learning_rate": 2.6898648479986187e-06, | |
| "loss": 0.5641, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.3753535353535353, | |
| "grad_norm": 7.15625, | |
| "learning_rate": 2.677346335673713e-06, | |
| "loss": 0.4403, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 1.376969696969697, | |
| "grad_norm": 5.59375, | |
| "learning_rate": 2.664846361837997e-06, | |
| "loss": 0.3652, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 1.3785858585858586, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 2.6523650262612476e-06, | |
| "loss": 0.4473, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 1.3802020202020202, | |
| "grad_norm": 8.25, | |
| "learning_rate": 2.639902428564475e-06, | |
| "loss": 0.6482, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 1.3818181818181818, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 2.6274586682191306e-06, | |
| "loss": 0.3526, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 1.3834343434343435, | |
| "grad_norm": 6.59375, | |
| "learning_rate": 2.6150338445463146e-06, | |
| "loss": 0.4567, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 1.385050505050505, | |
| "grad_norm": 5.5625, | |
| "learning_rate": 2.6026280567159847e-06, | |
| "loss": 0.4546, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 1.3866666666666667, | |
| "grad_norm": 6.375, | |
| "learning_rate": 2.5902414037461544e-06, | |
| "loss": 0.355, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 1.3882828282828283, | |
| "grad_norm": 9.0, | |
| "learning_rate": 2.5778739845021177e-06, | |
| "loss": 0.6155, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 1.38989898989899, | |
| "grad_norm": 6.71875, | |
| "learning_rate": 2.565525897695651e-06, | |
| "loss": 0.5631, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.3915151515151516, | |
| "grad_norm": 7.0, | |
| "learning_rate": 2.5531972418842266e-06, | |
| "loss": 0.4242, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 1.3915151515151516, | |
| "eval_loss": 0.6649448275566101, | |
| "eval_runtime": 35.1088, | |
| "eval_samples_per_second": 31.331, | |
| "eval_steps_per_second": 3.931, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 1.3931313131313132, | |
| "grad_norm": 7.125, | |
| "learning_rate": 2.540888115470226e-06, | |
| "loss": 0.6219, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 1.3947474747474748, | |
| "grad_norm": 9.5, | |
| "learning_rate": 2.5285986167001565e-06, | |
| "loss": 0.3452, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 1.3963636363636365, | |
| "grad_norm": 7.78125, | |
| "learning_rate": 2.51632884366386e-06, | |
| "loss": 0.5206, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 1.397979797979798, | |
| "grad_norm": 5.40625, | |
| "learning_rate": 2.5040788942937423e-06, | |
| "loss": 0.3927, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 1.3995959595959597, | |
| "grad_norm": 5.8125, | |
| "learning_rate": 2.4918488663639824e-06, | |
| "loss": 0.3102, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 1.4012121212121211, | |
| "grad_norm": 7.21875, | |
| "learning_rate": 2.479638857489753e-06, | |
| "loss": 0.4863, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 1.4028282828282828, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 2.4674489651264433e-06, | |
| "loss": 0.415, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 1.4044444444444444, | |
| "grad_norm": 7.5, | |
| "learning_rate": 2.4552792865688814e-06, | |
| "loss": 0.4191, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 1.406060606060606, | |
| "grad_norm": 6.21875, | |
| "learning_rate": 2.4431299189505563e-06, | |
| "loss": 0.5285, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.4076767676767676, | |
| "grad_norm": 6.875, | |
| "learning_rate": 2.4310009592428442e-06, | |
| "loss": 0.569, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 1.4092929292929293, | |
| "grad_norm": 6.4375, | |
| "learning_rate": 2.418892504254231e-06, | |
| "loss": 0.5699, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 1.410909090909091, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 2.4068046506295455e-06, | |
| "loss": 0.5794, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 1.4125252525252525, | |
| "grad_norm": 7.65625, | |
| "learning_rate": 2.394737494849184e-06, | |
| "loss": 0.4674, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 1.4141414141414141, | |
| "grad_norm": 9.375, | |
| "learning_rate": 2.3826911332283364e-06, | |
| "loss": 0.619, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.4157575757575758, | |
| "grad_norm": 10.25, | |
| "learning_rate": 2.3706656619162278e-06, | |
| "loss": 0.7549, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 1.4173737373737374, | |
| "grad_norm": 5.5625, | |
| "learning_rate": 2.358661176895342e-06, | |
| "loss": 0.3443, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 1.418989898989899, | |
| "grad_norm": 6.96875, | |
| "learning_rate": 2.34667777398066e-06, | |
| "loss": 0.6654, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 1.4206060606060606, | |
| "grad_norm": 6.875, | |
| "learning_rate": 2.3347155488188944e-06, | |
| "loss": 0.6335, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 1.4222222222222223, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 2.322774596887726e-06, | |
| "loss": 0.4496, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.423838383838384, | |
| "grad_norm": 5.3125, | |
| "learning_rate": 2.3108550134950352e-06, | |
| "loss": 0.3563, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 1.4254545454545455, | |
| "grad_norm": 5.59375, | |
| "learning_rate": 2.298956893778154e-06, | |
| "loss": 0.4681, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 1.427070707070707, | |
| "grad_norm": 8.0, | |
| "learning_rate": 2.2870803327030995e-06, | |
| "loss": 0.6678, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 1.4286868686868686, | |
| "grad_norm": 5.34375, | |
| "learning_rate": 2.275225425063813e-06, | |
| "loss": 0.3222, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 1.4303030303030302, | |
| "grad_norm": 7.90625, | |
| "learning_rate": 2.2633922654814082e-06, | |
| "loss": 0.5111, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 1.4319191919191918, | |
| "grad_norm": 5.8125, | |
| "learning_rate": 2.251580948403416e-06, | |
| "loss": 0.2793, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 1.4335353535353534, | |
| "grad_norm": 6.5, | |
| "learning_rate": 2.23979156810303e-06, | |
| "loss": 0.3486, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 1.435151515151515, | |
| "grad_norm": 5.6875, | |
| "learning_rate": 2.2280242186783473e-06, | |
| "loss": 0.4422, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 1.4367676767676767, | |
| "grad_norm": 7.0, | |
| "learning_rate": 2.2162789940516268e-06, | |
| "loss": 0.4664, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 1.4383838383838383, | |
| "grad_norm": 7.8125, | |
| "learning_rate": 2.204555987968541e-06, | |
| "loss": 0.408, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 2.192855293997418e-06, | |
| "loss": 0.3908, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 1.4416161616161616, | |
| "grad_norm": 6.71875, | |
| "learning_rate": 2.1811770055284968e-06, | |
| "loss": 0.6109, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 1.4432323232323232, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 2.1695212157731904e-06, | |
| "loss": 0.3194, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 1.4448484848484848, | |
| "grad_norm": 7.3125, | |
| "learning_rate": 2.157888017763333e-06, | |
| "loss": 0.5721, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 1.4464646464646465, | |
| "grad_norm": 6.4375, | |
| "learning_rate": 2.1462775043504423e-06, | |
| "loss": 0.4683, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 1.448080808080808, | |
| "grad_norm": 7.90625, | |
| "learning_rate": 2.134689768204975e-06, | |
| "loss": 0.3507, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 1.4496969696969697, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 2.12312490181559e-06, | |
| "loss": 0.5323, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 1.4513131313131313, | |
| "grad_norm": 6.15625, | |
| "learning_rate": 2.1115829974884097e-06, | |
| "loss": 0.4394, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 1.452929292929293, | |
| "grad_norm": 6.0, | |
| "learning_rate": 2.1000641473462773e-06, | |
| "loss": 0.4603, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 2.0885684433280336e-06, | |
| "loss": 0.4744, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.4561616161616162, | |
| "grad_norm": 7.375, | |
| "learning_rate": 2.077095977187773e-06, | |
| "loss": 0.4406, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 1.4577777777777778, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 2.065646840494115e-06, | |
| "loss": 0.3625, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 1.4593939393939395, | |
| "grad_norm": 5.46875, | |
| "learning_rate": 2.054221124629473e-06, | |
| "loss": 0.3098, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 1.461010101010101, | |
| "grad_norm": 5.4375, | |
| "learning_rate": 2.042818920789326e-06, | |
| "loss": 0.3083, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 1.4626262626262627, | |
| "grad_norm": 6.4375, | |
| "learning_rate": 2.0314403199814846e-06, | |
| "loss": 0.614, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 1.4642424242424243, | |
| "grad_norm": 5.65625, | |
| "learning_rate": 2.020085413025375e-06, | |
| "loss": 0.3322, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 1.465858585858586, | |
| "grad_norm": 6.21875, | |
| "learning_rate": 2.008754290551306e-06, | |
| "loss": 0.4278, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 1.4674747474747476, | |
| "grad_norm": 7.25, | |
| "learning_rate": 1.9974470429997482e-06, | |
| "loss": 0.6041, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 1.4690909090909092, | |
| "grad_norm": 6.375, | |
| "learning_rate": 1.9861637606206134e-06, | |
| "loss": 0.593, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 1.4707070707070706, | |
| "grad_norm": 8.25, | |
| "learning_rate": 1.974904533472532e-06, | |
| "loss": 0.4545, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.4723232323232323, | |
| "grad_norm": 11.125, | |
| "learning_rate": 1.9636694514221356e-06, | |
| "loss": 0.6286, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 1.4739393939393939, | |
| "grad_norm": 5.8125, | |
| "learning_rate": 1.9524586041433393e-06, | |
| "loss": 0.3233, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 1.4755555555555555, | |
| "grad_norm": 5.1875, | |
| "learning_rate": 1.9412720811166248e-06, | |
| "loss": 0.263, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 1.4771717171717171, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 1.9301099716283293e-06, | |
| "loss": 0.504, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 1.4787878787878788, | |
| "grad_norm": 6.46875, | |
| "learning_rate": 1.918972364769932e-06, | |
| "loss": 0.4073, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 1.4804040404040404, | |
| "grad_norm": 6.8125, | |
| "learning_rate": 1.907859349437336e-06, | |
| "loss": 0.3346, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 1.482020202020202, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 1.896771014330171e-06, | |
| "loss": 0.4134, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 1.4836363636363636, | |
| "grad_norm": 6.28125, | |
| "learning_rate": 1.8857074479510761e-06, | |
| "loss": 0.3783, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 1.4852525252525253, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 1.8746687386049972e-06, | |
| "loss": 0.3182, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 1.486868686868687, | |
| "grad_norm": 7.0, | |
| "learning_rate": 1.8636549743984815e-06, | |
| "loss": 0.431, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.4884848484848485, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 1.8526662432389752e-06, | |
| "loss": 0.3474, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 1.4901010101010101, | |
| "grad_norm": 7.875, | |
| "learning_rate": 1.8417026328341158e-06, | |
| "loss": 0.511, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 1.4917171717171718, | |
| "grad_norm": 7.84375, | |
| "learning_rate": 1.8307642306910423e-06, | |
| "loss": 0.3839, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 1.4933333333333334, | |
| "grad_norm": 7.25, | |
| "learning_rate": 1.8198511241156902e-06, | |
| "loss": 0.3581, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 1.494949494949495, | |
| "grad_norm": 8.375, | |
| "learning_rate": 1.8089634002120953e-06, | |
| "loss": 0.4388, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.4965656565656564, | |
| "grad_norm": 7.375, | |
| "learning_rate": 1.7981011458816988e-06, | |
| "loss": 0.6741, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 1.498181818181818, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 1.7872644478226563e-06, | |
| "loss": 0.3082, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 1.4997979797979797, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 1.776453392529139e-06, | |
| "loss": 0.6083, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 1.5014141414141413, | |
| "grad_norm": 8.25, | |
| "learning_rate": 1.7656680662906495e-06, | |
| "loss": 0.5088, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 1.503030303030303, | |
| "grad_norm": 7.65625, | |
| "learning_rate": 1.7549085551913358e-06, | |
| "loss": 0.5867, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.5046464646464646, | |
| "grad_norm": 7.15625, | |
| "learning_rate": 1.744174945109296e-06, | |
| "loss": 0.4078, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 1.5062626262626262, | |
| "grad_norm": 6.59375, | |
| "learning_rate": 1.7334673217158976e-06, | |
| "loss": 0.7761, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 1.5078787878787878, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 1.7227857704750888e-06, | |
| "loss": 0.3331, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 1.5094949494949494, | |
| "grad_norm": 6.84375, | |
| "learning_rate": 1.7121303766427266e-06, | |
| "loss": 0.3509, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 1.511111111111111, | |
| "grad_norm": 5.25, | |
| "learning_rate": 1.7015012252658864e-06, | |
| "loss": 0.2569, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 1.5127272727272727, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 1.6908984011821883e-06, | |
| "loss": 0.3199, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 1.5143434343434343, | |
| "grad_norm": 28.5, | |
| "learning_rate": 1.6803219890191174e-06, | |
| "loss": 0.8267, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 1.515959595959596, | |
| "grad_norm": 9.25, | |
| "learning_rate": 1.669772073193352e-06, | |
| "loss": 0.5876, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 1.5175757575757576, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 1.6592487379100819e-06, | |
| "loss": 0.4238, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 1.5191919191919192, | |
| "grad_norm": 8.125, | |
| "learning_rate": 1.6487520671623469e-06, | |
| "loss": 0.349, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.5208080808080808, | |
| "grad_norm": 7.75, | |
| "learning_rate": 1.6382821447303593e-06, | |
| "loss": 0.4342, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 1.5224242424242425, | |
| "grad_norm": 6.625, | |
| "learning_rate": 1.6278390541808364e-06, | |
| "loss": 0.4918, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 1.524040404040404, | |
| "grad_norm": 6.84375, | |
| "learning_rate": 1.6174228788663366e-06, | |
| "loss": 0.4434, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 1.5256565656565657, | |
| "grad_norm": 8.25, | |
| "learning_rate": 1.6070337019245896e-06, | |
| "loss": 0.381, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 1.5272727272727273, | |
| "grad_norm": 7.125, | |
| "learning_rate": 1.5966716062778336e-06, | |
| "loss": 0.6952, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 1.528888888888889, | |
| "grad_norm": 6.28125, | |
| "learning_rate": 1.5863366746321578e-06, | |
| "loss": 0.5478, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 1.5305050505050506, | |
| "grad_norm": 6.125, | |
| "learning_rate": 1.5760289894768377e-06, | |
| "loss": 0.329, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 1.5321212121212122, | |
| "grad_norm": 8.0, | |
| "learning_rate": 1.5657486330836786e-06, | |
| "loss": 0.5797, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 1.5337373737373738, | |
| "grad_norm": 8.75, | |
| "learning_rate": 1.5554956875063587e-06, | |
| "loss": 0.5311, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 1.5353535353535355, | |
| "grad_norm": 7.25, | |
| "learning_rate": 1.5452702345797738e-06, | |
| "loss": 0.5798, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.536969696969697, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 1.5350723559193837e-06, | |
| "loss": 0.428, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 1.5385858585858587, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 1.5249021329205638e-06, | |
| "loss": 0.4676, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 1.5402020202020203, | |
| "grad_norm": 6.875, | |
| "learning_rate": 1.51475964675795e-06, | |
| "loss": 0.5698, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 1.541818181818182, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 1.5046449783847965e-06, | |
| "loss": 0.5788, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 1.5434343434343434, | |
| "grad_norm": 6.875, | |
| "learning_rate": 1.4945582085323275e-06, | |
| "loss": 0.5295, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 1.545050505050505, | |
| "grad_norm": 5.96875, | |
| "learning_rate": 1.4844994177090871e-06, | |
| "loss": 0.5326, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 1.5466666666666666, | |
| "grad_norm": 6.0, | |
| "learning_rate": 1.4744686862003082e-06, | |
| "loss": 0.4787, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 1.5482828282828283, | |
| "grad_norm": 6.59375, | |
| "learning_rate": 1.4644660940672628e-06, | |
| "loss": 0.561, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 1.5498989898989899, | |
| "grad_norm": 9.0, | |
| "learning_rate": 1.4544917211466264e-06, | |
| "loss": 0.6131, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 1.5515151515151515, | |
| "grad_norm": 6.5, | |
| "learning_rate": 1.4445456470498392e-06, | |
| "loss": 0.4817, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.5531313131313131, | |
| "grad_norm": 5.1875, | |
| "learning_rate": 1.4346279511624728e-06, | |
| "loss": 0.4069, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 1.5547474747474748, | |
| "grad_norm": 6.21875, | |
| "learning_rate": 1.4247387126435957e-06, | |
| "loss": 0.5329, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 1.5563636363636364, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 1.414878010425137e-06, | |
| "loss": 0.4998, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 1.557979797979798, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 1.4050459232112652e-06, | |
| "loss": 0.4898, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 1.5595959595959596, | |
| "grad_norm": 6.71875, | |
| "learning_rate": 1.3952425294777533e-06, | |
| "loss": 0.5784, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 1.561212121212121, | |
| "grad_norm": 7.375, | |
| "learning_rate": 1.3854679074713557e-06, | |
| "loss": 0.546, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 1.5628282828282827, | |
| "grad_norm": 7.5, | |
| "learning_rate": 1.3757221352091805e-06, | |
| "loss": 0.5018, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 1.5644444444444443, | |
| "grad_norm": 7.46875, | |
| "learning_rate": 1.3660052904780707e-06, | |
| "loss": 0.3646, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 1.566060606060606, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 1.3563174508339793e-06, | |
| "loss": 0.6095, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 1.5676767676767676, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 1.3466586936013548e-06, | |
| "loss": 0.4514, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.5692929292929292, | |
| "grad_norm": 6.71875, | |
| "learning_rate": 1.337029095872519e-06, | |
| "loss": 0.6468, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 1.5709090909090908, | |
| "grad_norm": 5.625, | |
| "learning_rate": 1.3274287345070564e-06, | |
| "loss": 0.4297, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 1.5725252525252524, | |
| "grad_norm": 9.5, | |
| "learning_rate": 1.3178576861311988e-06, | |
| "loss": 0.3784, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 1.574141414141414, | |
| "grad_norm": 9.875, | |
| "learning_rate": 1.3083160271372092e-06, | |
| "loss": 0.5492, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 1.5757575757575757, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 1.2988038336827814e-06, | |
| "loss": 0.4873, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 1.5773737373737373, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 1.2893211816904243e-06, | |
| "loss": 0.3761, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 1.578989898989899, | |
| "grad_norm": 7.4375, | |
| "learning_rate": 1.2798681468468605e-06, | |
| "loss": 0.5108, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 1.5806060606060606, | |
| "grad_norm": 5.75, | |
| "learning_rate": 1.2704448046024192e-06, | |
| "loss": 0.2867, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 1.5822222222222222, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 1.261051230170437e-06, | |
| "loss": 0.3189, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 1.5838383838383838, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 1.2516874985266508e-06, | |
| "loss": 0.3282, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.5854545454545454, | |
| "grad_norm": 6.15625, | |
| "learning_rate": 1.2423536844086105e-06, | |
| "loss": 0.6539, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 1.587070707070707, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 1.2330498623150722e-06, | |
| "loss": 0.3337, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 1.5886868686868687, | |
| "grad_norm": 7.25, | |
| "learning_rate": 1.2237761065054106e-06, | |
| "loss": 0.3437, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 1.5903030303030303, | |
| "grad_norm": 6.84375, | |
| "learning_rate": 1.2145324909990202e-06, | |
| "loss": 0.5222, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 1.5903030303030303, | |
| "eval_loss": 0.6608893275260925, | |
| "eval_runtime": 35.0868, | |
| "eval_samples_per_second": 31.351, | |
| "eval_steps_per_second": 3.933, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 1.591919191919192, | |
| "grad_norm": 5.25, | |
| "learning_rate": 1.2053190895747324e-06, | |
| "loss": 0.3712, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 1.5935353535353536, | |
| "grad_norm": 6.78125, | |
| "learning_rate": 1.1961359757702151e-06, | |
| "loss": 0.54, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 1.5951515151515152, | |
| "grad_norm": 8.375, | |
| "learning_rate": 1.1869832228813987e-06, | |
| "loss": 0.5581, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 1.5967676767676768, | |
| "grad_norm": 5.78125, | |
| "learning_rate": 1.1778609039618804e-06, | |
| "loss": 0.4538, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 1.5983838383838385, | |
| "grad_norm": 6.0, | |
| "learning_rate": 1.168769091822351e-06, | |
| "loss": 0.4087, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 6.84375, | |
| "learning_rate": 1.1597078590300054e-06, | |
| "loss": 0.3286, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.6016161616161617, | |
| "grad_norm": 7.75, | |
| "learning_rate": 1.150677277907964e-06, | |
| "loss": 0.3279, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 1.6032323232323233, | |
| "grad_norm": 5.625, | |
| "learning_rate": 1.1416774205347015e-06, | |
| "loss": 0.6002, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 1.604848484848485, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 1.1327083587434662e-06, | |
| "loss": 0.3384, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 1.6064646464646466, | |
| "grad_norm": 6.53125, | |
| "learning_rate": 1.1237701641217097e-06, | |
| "loss": 0.3252, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 1.6080808080808082, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 1.1148629080105128e-06, | |
| "loss": 0.7288, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 1.6096969696969698, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 1.1059866615040205e-06, | |
| "loss": 0.731, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 1.6113131313131315, | |
| "grad_norm": 6.46875, | |
| "learning_rate": 1.0971414954488657e-06, | |
| "loss": 0.4641, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 1.6129292929292929, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 1.0883274804436155e-06, | |
| "loss": 0.5988, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 1.6145454545454545, | |
| "grad_norm": 6.125, | |
| "learning_rate": 1.0795446868381987e-06, | |
| "loss": 0.3712, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 1.6161616161616161, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 1.0707931847333487e-06, | |
| "loss": 0.495, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.6177777777777778, | |
| "grad_norm": 6.375, | |
| "learning_rate": 1.0620730439800424e-06, | |
| "loss": 0.5169, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 1.6193939393939394, | |
| "grad_norm": 6.0, | |
| "learning_rate": 1.053384334178944e-06, | |
| "loss": 0.6019, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 1.621010101010101, | |
| "grad_norm": 6.5, | |
| "learning_rate": 1.0447271246798446e-06, | |
| "loss": 0.5812, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 1.6226262626262626, | |
| "grad_norm": 7.34375, | |
| "learning_rate": 1.036101484581117e-06, | |
| "loss": 0.4106, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 1.6242424242424243, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 1.027507482729157e-06, | |
| "loss": 0.8248, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 1.625858585858586, | |
| "grad_norm": 7.21875, | |
| "learning_rate": 1.0189451877178386e-06, | |
| "loss": 0.4905, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 1.6274747474747475, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 1.0104146678879633e-06, | |
| "loss": 0.4978, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 1.6290909090909091, | |
| "grad_norm": 9.25, | |
| "learning_rate": 1.0019159913267156e-06, | |
| "loss": 0.5907, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 1.6307070707070705, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 9.934492258671214e-07, | |
| "loss": 0.384, | |
| "step": 1009 | |
| }, | |
| { | |
| "epoch": 1.6323232323232322, | |
| "grad_norm": 5.625, | |
| "learning_rate": 9.850144390875022e-07, | |
| "loss": 0.2712, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.6339393939393938, | |
| "grad_norm": 6.59375, | |
| "learning_rate": 9.76611698310942e-07, | |
| "loss": 0.5483, | |
| "step": 1011 | |
| }, | |
| { | |
| "epoch": 1.6355555555555554, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 9.68241070604743e-07, | |
| "loss": 0.5543, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 1.637171717171717, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 9.59902622779898e-07, | |
| "loss": 0.6036, | |
| "step": 1013 | |
| }, | |
| { | |
| "epoch": 1.6387878787878787, | |
| "grad_norm": 7.125, | |
| "learning_rate": 9.51596421390547e-07, | |
| "loss": 0.4629, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 1.6404040404040403, | |
| "grad_norm": 6.71875, | |
| "learning_rate": 9.433225327334567e-07, | |
| "loss": 0.4255, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 1.642020202020202, | |
| "grad_norm": 6.78125, | |
| "learning_rate": 9.350810228474855e-07, | |
| "loss": 0.6725, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 1.6436363636363636, | |
| "grad_norm": 8.875, | |
| "learning_rate": 9.268719575130563e-07, | |
| "loss": 0.4409, | |
| "step": 1017 | |
| }, | |
| { | |
| "epoch": 1.6452525252525252, | |
| "grad_norm": 5.5, | |
| "learning_rate": 9.186954022516343e-07, | |
| "loss": 0.4973, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 1.6468686868686868, | |
| "grad_norm": 7.5, | |
| "learning_rate": 9.105514223252016e-07, | |
| "loss": 0.4426, | |
| "step": 1019 | |
| }, | |
| { | |
| "epoch": 1.6484848484848484, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 9.024400827357344e-07, | |
| "loss": 0.5614, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.65010101010101, | |
| "grad_norm": 8.875, | |
| "learning_rate": 8.943614482246915e-07, | |
| "loss": 0.6925, | |
| "step": 1021 | |
| }, | |
| { | |
| "epoch": 1.6517171717171717, | |
| "grad_norm": 7.65625, | |
| "learning_rate": 8.863155832724895e-07, | |
| "loss": 0.5336, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 1.6533333333333333, | |
| "grad_norm": 7.3125, | |
| "learning_rate": 8.783025520979932e-07, | |
| "loss": 0.383, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 1.654949494949495, | |
| "grad_norm": 7.59375, | |
| "learning_rate": 8.703224186580012e-07, | |
| "loss": 0.6345, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 1.6565656565656566, | |
| "grad_norm": 5.75, | |
| "learning_rate": 8.623752466467344e-07, | |
| "loss": 0.4726, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 1.6581818181818182, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 8.544610994953317e-07, | |
| "loss": 0.4826, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 1.6597979797979798, | |
| "grad_norm": 11.875, | |
| "learning_rate": 8.465800403713365e-07, | |
| "loss": 0.4592, | |
| "step": 1027 | |
| }, | |
| { | |
| "epoch": 1.6614141414141415, | |
| "grad_norm": 5.78125, | |
| "learning_rate": 8.387321321781977e-07, | |
| "loss": 0.4643, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 1.663030303030303, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 8.309174375547724e-07, | |
| "loss": 0.5006, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 1.6646464646464647, | |
| "grad_norm": 10.25, | |
| "learning_rate": 8.23136018874815e-07, | |
| "loss": 0.5035, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.6662626262626263, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 8.153879382464836e-07, | |
| "loss": 0.3378, | |
| "step": 1031 | |
| }, | |
| { | |
| "epoch": 1.667878787878788, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 8.07673257511849e-07, | |
| "loss": 0.5713, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 1.6694949494949496, | |
| "grad_norm": 7.3125, | |
| "learning_rate": 7.999920382463955e-07, | |
| "loss": 0.3865, | |
| "step": 1033 | |
| }, | |
| { | |
| "epoch": 1.6711111111111112, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 7.923443417585324e-07, | |
| "loss": 0.3851, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 1.6727272727272728, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 7.847302290891029e-07, | |
| "loss": 0.5374, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 1.6743434343434345, | |
| "grad_norm": 7.84375, | |
| "learning_rate": 7.771497610108981e-07, | |
| "loss": 0.8689, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 1.675959595959596, | |
| "grad_norm": 7.46875, | |
| "learning_rate": 7.696029980281722e-07, | |
| "loss": 0.557, | |
| "step": 1037 | |
| }, | |
| { | |
| "epoch": 1.6775757575757577, | |
| "grad_norm": 7.8125, | |
| "learning_rate": 7.620900003761561e-07, | |
| "loss": 0.4862, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 1.6791919191919193, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 7.546108280205816e-07, | |
| "loss": 0.658, | |
| "step": 1039 | |
| }, | |
| { | |
| "epoch": 1.680808080808081, | |
| "grad_norm": 9.0625, | |
| "learning_rate": 7.471655406572003e-07, | |
| "loss": 0.6738, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.6824242424242424, | |
| "grad_norm": 5.875, | |
| "learning_rate": 7.397541977113076e-07, | |
| "loss": 0.4773, | |
| "step": 1041 | |
| }, | |
| { | |
| "epoch": 1.684040404040404, | |
| "grad_norm": 7.625, | |
| "learning_rate": 7.323768583372675e-07, | |
| "loss": 0.4712, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 1.6856565656565656, | |
| "grad_norm": 7.21875, | |
| "learning_rate": 7.250335814180432e-07, | |
| "loss": 0.7264, | |
| "step": 1043 | |
| }, | |
| { | |
| "epoch": 1.6872727272727273, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 7.177244255647209e-07, | |
| "loss": 0.6175, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 1.6888888888888889, | |
| "grad_norm": 6.21875, | |
| "learning_rate": 7.10449449116049e-07, | |
| "loss": 0.3118, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 1.6905050505050505, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 7.032087101379703e-07, | |
| "loss": 0.6699, | |
| "step": 1046 | |
| }, | |
| { | |
| "epoch": 1.6921212121212121, | |
| "grad_norm": 6.875, | |
| "learning_rate": 6.960022664231558e-07, | |
| "loss": 0.3863, | |
| "step": 1047 | |
| }, | |
| { | |
| "epoch": 1.6937373737373738, | |
| "grad_norm": 6.84375, | |
| "learning_rate": 6.888301754905469e-07, | |
| "loss": 0.51, | |
| "step": 1048 | |
| }, | |
| { | |
| "epoch": 1.6953535353535354, | |
| "grad_norm": 8.0, | |
| "learning_rate": 6.816924945848941e-07, | |
| "loss": 0.5381, | |
| "step": 1049 | |
| }, | |
| { | |
| "epoch": 1.696969696969697, | |
| "grad_norm": 8.5, | |
| "learning_rate": 6.745892806763015e-07, | |
| "loss": 0.6998, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.6985858585858584, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 6.675205904597715e-07, | |
| "loss": 0.4438, | |
| "step": 1051 | |
| }, | |
| { | |
| "epoch": 1.70020202020202, | |
| "grad_norm": 6.96875, | |
| "learning_rate": 6.604864803547511e-07, | |
| "loss": 0.615, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 1.7018181818181817, | |
| "grad_norm": 5.75, | |
| "learning_rate": 6.534870065046839e-07, | |
| "loss": 0.2991, | |
| "step": 1053 | |
| }, | |
| { | |
| "epoch": 1.7034343434343433, | |
| "grad_norm": 7.21875, | |
| "learning_rate": 6.465222247765618e-07, | |
| "loss": 0.4271, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 1.705050505050505, | |
| "grad_norm": 5.8125, | |
| "learning_rate": 6.395921907604747e-07, | |
| "loss": 0.4636, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 1.7066666666666666, | |
| "grad_norm": 7.125, | |
| "learning_rate": 6.326969597691724e-07, | |
| "loss": 0.4994, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 1.7082828282828282, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 6.258365868376215e-07, | |
| "loss": 0.5785, | |
| "step": 1057 | |
| }, | |
| { | |
| "epoch": 1.7098989898989898, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 6.190111267225651e-07, | |
| "loss": 0.8024, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 1.7115151515151514, | |
| "grad_norm": 7.75, | |
| "learning_rate": 6.122206339020854e-07, | |
| "loss": 0.3282, | |
| "step": 1059 | |
| }, | |
| { | |
| "epoch": 1.713131313131313, | |
| "grad_norm": 7.46875, | |
| "learning_rate": 6.054651625751717e-07, | |
| "loss": 0.4669, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.7147474747474747, | |
| "grad_norm": 7.25, | |
| "learning_rate": 5.987447666612823e-07, | |
| "loss": 0.584, | |
| "step": 1061 | |
| }, | |
| { | |
| "epoch": 1.7163636363636363, | |
| "grad_norm": 6.875, | |
| "learning_rate": 5.920594997999202e-07, | |
| "loss": 0.5449, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 1.717979797979798, | |
| "grad_norm": 10.125, | |
| "learning_rate": 5.854094153502022e-07, | |
| "loss": 0.4517, | |
| "step": 1063 | |
| }, | |
| { | |
| "epoch": 1.7195959595959596, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 5.787945663904332e-07, | |
| "loss": 0.6394, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 1.7212121212121212, | |
| "grad_norm": 5.34375, | |
| "learning_rate": 5.722150057176823e-07, | |
| "loss": 0.395, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 1.7228282828282828, | |
| "grad_norm": 8.375, | |
| "learning_rate": 5.65670785847362e-07, | |
| "loss": 0.4517, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 1.7244444444444444, | |
| "grad_norm": 9.375, | |
| "learning_rate": 5.591619590128067e-07, | |
| "loss": 0.4588, | |
| "step": 1067 | |
| }, | |
| { | |
| "epoch": 1.726060606060606, | |
| "grad_norm": 6.96875, | |
| "learning_rate": 5.526885771648599e-07, | |
| "loss": 0.2908, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 1.7276767676767677, | |
| "grad_norm": 6.875, | |
| "learning_rate": 5.462506919714583e-07, | |
| "loss": 0.3348, | |
| "step": 1069 | |
| }, | |
| { | |
| "epoch": 1.7292929292929293, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 5.398483548172162e-07, | |
| "loss": 0.3347, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.730909090909091, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 5.334816168030204e-07, | |
| "loss": 0.6141, | |
| "step": 1071 | |
| }, | |
| { | |
| "epoch": 1.7325252525252526, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 5.271505287456153e-07, | |
| "loss": 0.437, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 1.7341414141414142, | |
| "grad_norm": 6.53125, | |
| "learning_rate": 5.208551411772056e-07, | |
| "loss": 0.4954, | |
| "step": 1073 | |
| }, | |
| { | |
| "epoch": 1.7357575757575758, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 5.145955043450484e-07, | |
| "loss": 0.4811, | |
| "step": 1074 | |
| }, | |
| { | |
| "epoch": 1.7373737373737375, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 5.083716682110512e-07, | |
| "loss": 0.6364, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 1.738989898989899, | |
| "grad_norm": 5.46875, | |
| "learning_rate": 5.021836824513759e-07, | |
| "loss": 0.2266, | |
| "step": 1076 | |
| }, | |
| { | |
| "epoch": 1.7406060606060607, | |
| "grad_norm": 6.53125, | |
| "learning_rate": 4.960315964560425e-07, | |
| "loss": 0.3873, | |
| "step": 1077 | |
| }, | |
| { | |
| "epoch": 1.7422222222222223, | |
| "grad_norm": 5.375, | |
| "learning_rate": 4.899154593285294e-07, | |
| "loss": 0.3454, | |
| "step": 1078 | |
| }, | |
| { | |
| "epoch": 1.743838383838384, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 4.838353198853884e-07, | |
| "loss": 0.2861, | |
| "step": 1079 | |
| }, | |
| { | |
| "epoch": 1.7454545454545456, | |
| "grad_norm": 5.75, | |
| "learning_rate": 4.777912266558532e-07, | |
| "loss": 0.4245, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.7470707070707072, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 4.717832278814488e-07, | |
| "loss": 0.735, | |
| "step": 1081 | |
| }, | |
| { | |
| "epoch": 1.7486868686868688, | |
| "grad_norm": 7.21875, | |
| "learning_rate": 4.658113715156098e-07, | |
| "loss": 0.7658, | |
| "step": 1082 | |
| }, | |
| { | |
| "epoch": 1.7503030303030302, | |
| "grad_norm": 6.78125, | |
| "learning_rate": 4.5987570522329773e-07, | |
| "loss": 0.5362, | |
| "step": 1083 | |
| }, | |
| { | |
| "epoch": 1.7519191919191919, | |
| "grad_norm": 5.96875, | |
| "learning_rate": 4.5397627638061604e-07, | |
| "loss": 0.3993, | |
| "step": 1084 | |
| }, | |
| { | |
| "epoch": 1.7535353535353535, | |
| "grad_norm": 4.96875, | |
| "learning_rate": 4.481131320744386e-07, | |
| "loss": 0.4245, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 1.7551515151515151, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 4.422863191020299e-07, | |
| "loss": 0.4338, | |
| "step": 1086 | |
| }, | |
| { | |
| "epoch": 1.7567676767676768, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 4.3649588397067177e-07, | |
| "loss": 0.4729, | |
| "step": 1087 | |
| }, | |
| { | |
| "epoch": 1.7583838383838384, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 4.307418728972934e-07, | |
| "loss": 0.7243, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 4.2502433180810153e-07, | |
| "loss": 0.3767, | |
| "step": 1089 | |
| }, | |
| { | |
| "epoch": 1.7616161616161616, | |
| "grad_norm": 7.5, | |
| "learning_rate": 4.193433063382135e-07, | |
| "loss": 0.3824, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.7632323232323233, | |
| "grad_norm": 7.28125, | |
| "learning_rate": 4.1369884183129485e-07, | |
| "loss": 0.7881, | |
| "step": 1091 | |
| }, | |
| { | |
| "epoch": 1.7648484848484849, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 4.080909833391944e-07, | |
| "loss": 0.8348, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 1.7664646464646465, | |
| "grad_norm": 6.5, | |
| "learning_rate": 4.0251977562158873e-07, | |
| "loss": 0.4176, | |
| "step": 1093 | |
| }, | |
| { | |
| "epoch": 1.768080808080808, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 3.9698526314562114e-07, | |
| "loss": 1.021, | |
| "step": 1094 | |
| }, | |
| { | |
| "epoch": 1.7696969696969695, | |
| "grad_norm": 7.9375, | |
| "learning_rate": 3.9148749008554807e-07, | |
| "loss": 0.6396, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 1.7713131313131312, | |
| "grad_norm": 7.6875, | |
| "learning_rate": 3.8602650032238675e-07, | |
| "loss": 0.5306, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 1.7729292929292928, | |
| "grad_norm": 7.8125, | |
| "learning_rate": 3.8060233744356634e-07, | |
| "loss": 0.5574, | |
| "step": 1097 | |
| }, | |
| { | |
| "epoch": 1.7745454545454544, | |
| "grad_norm": 7.4375, | |
| "learning_rate": 3.752150447425773e-07, | |
| "loss": 0.6494, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 1.776161616161616, | |
| "grad_norm": 6.125, | |
| "learning_rate": 3.698646652186272e-07, | |
| "loss": 0.5226, | |
| "step": 1099 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 8.625, | |
| "learning_rate": 3.6455124157629805e-07, | |
| "loss": 0.7729, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.7793939393939393, | |
| "grad_norm": 6.96875, | |
| "learning_rate": 3.5927481622520463e-07, | |
| "loss": 0.5724, | |
| "step": 1101 | |
| }, | |
| { | |
| "epoch": 1.781010101010101, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 3.5403543127965514e-07, | |
| "loss": 0.5896, | |
| "step": 1102 | |
| }, | |
| { | |
| "epoch": 1.7826262626262626, | |
| "grad_norm": 7.625, | |
| "learning_rate": 3.488331285583174e-07, | |
| "loss": 0.5196, | |
| "step": 1103 | |
| }, | |
| { | |
| "epoch": 1.7842424242424242, | |
| "grad_norm": 5.65625, | |
| "learning_rate": 3.436679495838835e-07, | |
| "loss": 0.4324, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 1.7858585858585858, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 3.38539935582739e-07, | |
| "loss": 0.2973, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 1.7874747474747474, | |
| "grad_norm": 5.8125, | |
| "learning_rate": 3.334491274846324e-07, | |
| "loss": 0.3711, | |
| "step": 1106 | |
| }, | |
| { | |
| "epoch": 1.789090909090909, | |
| "grad_norm": 4.71875, | |
| "learning_rate": 3.283955659223509e-07, | |
| "loss": 0.2609, | |
| "step": 1107 | |
| }, | |
| { | |
| "epoch": 1.789090909090909, | |
| "eval_loss": 0.6599124670028687, | |
| "eval_runtime": 35.4231, | |
| "eval_samples_per_second": 31.053, | |
| "eval_steps_per_second": 3.896, | |
| "step": 1107 | |
| }, | |
| { | |
| "epoch": 1.7907070707070707, | |
| "grad_norm": 5.65625, | |
| "learning_rate": 3.233792912313943e-07, | |
| "loss": 0.4951, | |
| "step": 1108 | |
| }, | |
| { | |
| "epoch": 1.7923232323232323, | |
| "grad_norm": 7.90625, | |
| "learning_rate": 3.1840034344965355e-07, | |
| "loss": 0.5709, | |
| "step": 1109 | |
| }, | |
| { | |
| "epoch": 1.793939393939394, | |
| "grad_norm": 7.25, | |
| "learning_rate": 3.134587623170909e-07, | |
| "loss": 0.4571, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.7955555555555556, | |
| "grad_norm": 5.65625, | |
| "learning_rate": 3.085545872754242e-07, | |
| "loss": 0.5489, | |
| "step": 1111 | |
| }, | |
| { | |
| "epoch": 1.7971717171717172, | |
| "grad_norm": 6.875, | |
| "learning_rate": 3.0368785746780925e-07, | |
| "loss": 0.3417, | |
| "step": 1112 | |
| }, | |
| { | |
| "epoch": 1.7987878787878788, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 2.9885861173852946e-07, | |
| "loss": 0.5436, | |
| "step": 1113 | |
| }, | |
| { | |
| "epoch": 1.8004040404040405, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 2.940668886326864e-07, | |
| "loss": 0.407, | |
| "step": 1114 | |
| }, | |
| { | |
| "epoch": 1.802020202020202, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 2.8931272639589005e-07, | |
| "loss": 0.5136, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 1.8036363636363637, | |
| "grad_norm": 6.21875, | |
| "learning_rate": 2.8459616297395464e-07, | |
| "loss": 0.3449, | |
| "step": 1116 | |
| }, | |
| { | |
| "epoch": 1.8052525252525253, | |
| "grad_norm": 7.0, | |
| "learning_rate": 2.7991723601259677e-07, | |
| "loss": 0.3335, | |
| "step": 1117 | |
| }, | |
| { | |
| "epoch": 1.806868686868687, | |
| "grad_norm": 6.5, | |
| "learning_rate": 2.7527598285713387e-07, | |
| "loss": 0.7282, | |
| "step": 1118 | |
| }, | |
| { | |
| "epoch": 1.8084848484848486, | |
| "grad_norm": 6.15625, | |
| "learning_rate": 2.706724405521838e-07, | |
| "loss": 0.3397, | |
| "step": 1119 | |
| }, | |
| { | |
| "epoch": 1.8101010101010102, | |
| "grad_norm": 5.78125, | |
| "learning_rate": 2.6610664584137413e-07, | |
| "loss": 0.3164, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.8117171717171718, | |
| "grad_norm": 7.5625, | |
| "learning_rate": 2.615786351670452e-07, | |
| "loss": 0.548, | |
| "step": 1121 | |
| }, | |
| { | |
| "epoch": 1.8133333333333335, | |
| "grad_norm": 12.9375, | |
| "learning_rate": 2.570884446699612e-07, | |
| "loss": 0.5915, | |
| "step": 1122 | |
| }, | |
| { | |
| "epoch": 1.814949494949495, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 2.5263611018901957e-07, | |
| "loss": 0.371, | |
| "step": 1123 | |
| }, | |
| { | |
| "epoch": 1.8165656565656567, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 2.482216672609677e-07, | |
| "loss": 0.4942, | |
| "step": 1124 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 7.375, | |
| "learning_rate": 2.438451511201162e-07, | |
| "loss": 0.4319, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 1.8197979797979797, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 2.3950659669806033e-07, | |
| "loss": 0.5259, | |
| "step": 1126 | |
| }, | |
| { | |
| "epoch": 1.8214141414141414, | |
| "grad_norm": 7.8125, | |
| "learning_rate": 2.3520603862339952e-07, | |
| "loss": 0.4634, | |
| "step": 1127 | |
| }, | |
| { | |
| "epoch": 1.823030303030303, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 2.3094351122146307e-07, | |
| "loss": 0.446, | |
| "step": 1128 | |
| }, | |
| { | |
| "epoch": 1.8246464646464646, | |
| "grad_norm": 5.96875, | |
| "learning_rate": 2.2671904851403303e-07, | |
| "loss": 0.6715, | |
| "step": 1129 | |
| }, | |
| { | |
| "epoch": 1.8262626262626263, | |
| "grad_norm": 6.71875, | |
| "learning_rate": 2.2253268421907503e-07, | |
| "loss": 0.4167, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.8278787878787879, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 2.1838445175046796e-07, | |
| "loss": 0.6699, | |
| "step": 1131 | |
| }, | |
| { | |
| "epoch": 1.8294949494949495, | |
| "grad_norm": 6.8125, | |
| "learning_rate": 2.142743842177386e-07, | |
| "loss": 0.4689, | |
| "step": 1132 | |
| }, | |
| { | |
| "epoch": 1.8311111111111111, | |
| "grad_norm": 6.71875, | |
| "learning_rate": 2.1020251442579632e-07, | |
| "loss": 0.6284, | |
| "step": 1133 | |
| }, | |
| { | |
| "epoch": 1.8327272727272728, | |
| "grad_norm": 6.15625, | |
| "learning_rate": 2.0616887487467096e-07, | |
| "loss": 0.4624, | |
| "step": 1134 | |
| }, | |
| { | |
| "epoch": 1.8343434343434344, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 2.0217349775925487e-07, | |
| "loss": 0.4118, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 1.835959595959596, | |
| "grad_norm": 11.125, | |
| "learning_rate": 1.98216414969043e-07, | |
| "loss": 0.478, | |
| "step": 1136 | |
| }, | |
| { | |
| "epoch": 1.8375757575757574, | |
| "grad_norm": 5.34375, | |
| "learning_rate": 1.942976580878825e-07, | |
| "loss": 0.334, | |
| "step": 1137 | |
| }, | |
| { | |
| "epoch": 1.839191919191919, | |
| "grad_norm": 7.5625, | |
| "learning_rate": 1.9041725839371805e-07, | |
| "loss": 0.4451, | |
| "step": 1138 | |
| }, | |
| { | |
| "epoch": 1.8408080808080807, | |
| "grad_norm": 7.71875, | |
| "learning_rate": 1.8657524685834206e-07, | |
| "loss": 0.3302, | |
| "step": 1139 | |
| }, | |
| { | |
| "epoch": 1.8424242424242423, | |
| "grad_norm": 9.5, | |
| "learning_rate": 1.8277165414714858e-07, | |
| "loss": 0.5024, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.844040404040404, | |
| "grad_norm": 5.125, | |
| "learning_rate": 1.7900651061888762e-07, | |
| "loss": 0.4093, | |
| "step": 1141 | |
| }, | |
| { | |
| "epoch": 1.8456565656565656, | |
| "grad_norm": 7.78125, | |
| "learning_rate": 1.752798463254235e-07, | |
| "loss": 0.9122, | |
| "step": 1142 | |
| }, | |
| { | |
| "epoch": 1.8472727272727272, | |
| "grad_norm": 5.625, | |
| "learning_rate": 1.7159169101149398e-07, | |
| "loss": 0.493, | |
| "step": 1143 | |
| }, | |
| { | |
| "epoch": 1.8488888888888888, | |
| "grad_norm": 6.21875, | |
| "learning_rate": 1.6794207411447548e-07, | |
| "loss": 0.6134, | |
| "step": 1144 | |
| }, | |
| { | |
| "epoch": 1.8505050505050504, | |
| "grad_norm": 7.4375, | |
| "learning_rate": 1.6433102476414386e-07, | |
| "loss": 0.612, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 1.852121212121212, | |
| "grad_norm": 7.6875, | |
| "learning_rate": 1.6075857178244613e-07, | |
| "loss": 0.5233, | |
| "step": 1146 | |
| }, | |
| { | |
| "epoch": 1.8537373737373737, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 1.5722474368326802e-07, | |
| "loss": 0.7897, | |
| "step": 1147 | |
| }, | |
| { | |
| "epoch": 1.8553535353535353, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 1.5372956867220678e-07, | |
| "loss": 0.3268, | |
| "step": 1148 | |
| }, | |
| { | |
| "epoch": 1.856969696969697, | |
| "grad_norm": 6.8125, | |
| "learning_rate": 1.5027307464634656e-07, | |
| "loss": 0.8276, | |
| "step": 1149 | |
| }, | |
| { | |
| "epoch": 1.8585858585858586, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 1.4685528919403446e-07, | |
| "loss": 0.5105, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.8602020202020202, | |
| "grad_norm": 6.46875, | |
| "learning_rate": 1.434762395946626e-07, | |
| "loss": 0.5541, | |
| "step": 1151 | |
| }, | |
| { | |
| "epoch": 1.8618181818181818, | |
| "grad_norm": 5.96875, | |
| "learning_rate": 1.4013595281844872e-07, | |
| "loss": 0.3773, | |
| "step": 1152 | |
| }, | |
| { | |
| "epoch": 1.8634343434343434, | |
| "grad_norm": 7.8125, | |
| "learning_rate": 1.368344555262202e-07, | |
| "loss": 0.3672, | |
| "step": 1153 | |
| }, | |
| { | |
| "epoch": 1.865050505050505, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 1.335717740692033e-07, | |
| "loss": 0.47, | |
| "step": 1154 | |
| }, | |
| { | |
| "epoch": 1.8666666666666667, | |
| "grad_norm": 6.53125, | |
| "learning_rate": 1.30347934488812e-07, | |
| "loss": 0.458, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 1.8682828282828283, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 1.2716296251644e-07, | |
| "loss": 0.3902, | |
| "step": 1156 | |
| }, | |
| { | |
| "epoch": 1.86989898989899, | |
| "grad_norm": 7.84375, | |
| "learning_rate": 1.2401688357325415e-07, | |
| "loss": 0.2978, | |
| "step": 1157 | |
| }, | |
| { | |
| "epoch": 1.8715151515151516, | |
| "grad_norm": 6.8125, | |
| "learning_rate": 1.2090972276999513e-07, | |
| "loss": 0.4911, | |
| "step": 1158 | |
| }, | |
| { | |
| "epoch": 1.8731313131313132, | |
| "grad_norm": 7.90625, | |
| "learning_rate": 1.1784150490677215e-07, | |
| "loss": 0.501, | |
| "step": 1159 | |
| }, | |
| { | |
| "epoch": 1.8747474747474748, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 1.1481225447286803e-07, | |
| "loss": 0.3906, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.8763636363636365, | |
| "grad_norm": 6.125, | |
| "learning_rate": 1.1182199564654495e-07, | |
| "loss": 0.5567, | |
| "step": 1161 | |
| }, | |
| { | |
| "epoch": 1.877979797979798, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 1.0887075229484789e-07, | |
| "loss": 0.4789, | |
| "step": 1162 | |
| }, | |
| { | |
| "epoch": 1.8795959595959597, | |
| "grad_norm": 7.1875, | |
| "learning_rate": 1.0595854797341709e-07, | |
| "loss": 0.3514, | |
| "step": 1163 | |
| }, | |
| { | |
| "epoch": 1.8812121212121213, | |
| "grad_norm": 7.125, | |
| "learning_rate": 1.0308540592629756e-07, | |
| "loss": 0.5602, | |
| "step": 1164 | |
| }, | |
| { | |
| "epoch": 1.882828282828283, | |
| "grad_norm": 6.03125, | |
| "learning_rate": 1.0025134908575707e-07, | |
| "loss": 0.2898, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 1.8844444444444446, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 9.745640007209844e-08, | |
| "loss": 0.2949, | |
| "step": 1166 | |
| }, | |
| { | |
| "epoch": 1.8860606060606062, | |
| "grad_norm": 6.625, | |
| "learning_rate": 9.470058119348257e-08, | |
| "loss": 0.5841, | |
| "step": 1167 | |
| }, | |
| { | |
| "epoch": 1.8876767676767678, | |
| "grad_norm": 8.25, | |
| "learning_rate": 9.198391444575072e-08, | |
| "loss": 0.4759, | |
| "step": 1168 | |
| }, | |
| { | |
| "epoch": 1.8892929292929292, | |
| "grad_norm": 8.25, | |
| "learning_rate": 8.930642151224634e-08, | |
| "loss": 0.6854, | |
| "step": 1169 | |
| }, | |
| { | |
| "epoch": 1.8909090909090909, | |
| "grad_norm": 6.71875, | |
| "learning_rate": 8.666812376364187e-08, | |
| "loss": 0.3352, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.8925252525252525, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 8.406904225777223e-08, | |
| "loss": 0.3283, | |
| "step": 1171 | |
| }, | |
| { | |
| "epoch": 1.8941414141414141, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 8.150919773946165e-08, | |
| "loss": 0.4876, | |
| "step": 1172 | |
| }, | |
| { | |
| "epoch": 1.8957575757575758, | |
| "grad_norm": 5.6875, | |
| "learning_rate": 7.898861064036257e-08, | |
| "loss": 0.6124, | |
| "step": 1173 | |
| }, | |
| { | |
| "epoch": 1.8973737373737374, | |
| "grad_norm": 7.6875, | |
| "learning_rate": 7.650730107878812e-08, | |
| "loss": 0.5362, | |
| "step": 1174 | |
| }, | |
| { | |
| "epoch": 1.898989898989899, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 7.406528885955665e-08, | |
| "loss": 0.4961, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 1.9006060606060606, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 7.166259347382854e-08, | |
| "loss": 0.3595, | |
| "step": 1176 | |
| }, | |
| { | |
| "epoch": 1.9022222222222223, | |
| "grad_norm": 5.8125, | |
| "learning_rate": 6.929923409895345e-08, | |
| "loss": 0.3119, | |
| "step": 1177 | |
| }, | |
| { | |
| "epoch": 1.9038383838383839, | |
| "grad_norm": 6.0625, | |
| "learning_rate": 6.697522959831837e-08, | |
| "loss": 0.3637, | |
| "step": 1178 | |
| }, | |
| { | |
| "epoch": 1.9054545454545453, | |
| "grad_norm": 8.375, | |
| "learning_rate": 6.46905985211932e-08, | |
| "loss": 0.3966, | |
| "step": 1179 | |
| }, | |
| { | |
| "epoch": 1.907070707070707, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 6.244535910258697e-08, | |
| "loss": 0.6424, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.9086868686868685, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 6.023952926309917e-08, | |
| "loss": 0.5532, | |
| "step": 1181 | |
| }, | |
| { | |
| "epoch": 1.9103030303030302, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 5.8073126608778064e-08, | |
| "loss": 0.5825, | |
| "step": 1182 | |
| }, | |
| { | |
| "epoch": 1.9119191919191918, | |
| "grad_norm": 6.84375, | |
| "learning_rate": 5.594616843097978e-08, | |
| "loss": 0.3147, | |
| "step": 1183 | |
| }, | |
| { | |
| "epoch": 1.9135353535353534, | |
| "grad_norm": 8.125, | |
| "learning_rate": 5.3858671706230605e-08, | |
| "loss": 0.309, | |
| "step": 1184 | |
| }, | |
| { | |
| "epoch": 1.915151515151515, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 5.181065309609212e-08, | |
| "loss": 0.3509, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 1.9167676767676767, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 4.98021289470263e-08, | |
| "loss": 0.3494, | |
| "step": 1186 | |
| }, | |
| { | |
| "epoch": 1.9183838383838383, | |
| "grad_norm": 5.375, | |
| "learning_rate": 4.7833115290266686e-08, | |
| "loss": 0.2529, | |
| "step": 1187 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 4.590362784169022e-08, | |
| "loss": 0.6008, | |
| "step": 1188 | |
| }, | |
| { | |
| "epoch": 1.9216161616161616, | |
| "grad_norm": 6.53125, | |
| "learning_rate": 4.401368200169009e-08, | |
| "loss": 0.5284, | |
| "step": 1189 | |
| }, | |
| { | |
| "epoch": 1.9232323232323232, | |
| "grad_norm": 6.375, | |
| "learning_rate": 4.2163292855056936e-08, | |
| "loss": 0.5237, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.9248484848484848, | |
| "grad_norm": 7.34375, | |
| "learning_rate": 4.035247517085283e-08, | |
| "loss": 0.5049, | |
| "step": 1191 | |
| }, | |
| { | |
| "epoch": 1.9264646464646464, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 3.858124340229863e-08, | |
| "loss": 0.5204, | |
| "step": 1192 | |
| }, | |
| { | |
| "epoch": 1.928080808080808, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 3.684961168665457e-08, | |
| "loss": 0.4971, | |
| "step": 1193 | |
| }, | |
| { | |
| "epoch": 1.9296969696969697, | |
| "grad_norm": 11.0, | |
| "learning_rate": 3.515759384510986e-08, | |
| "loss": 0.4429, | |
| "step": 1194 | |
| }, | |
| { | |
| "epoch": 1.9313131313131313, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 3.350520338267216e-08, | |
| "loss": 0.5476, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 1.932929292929293, | |
| "grad_norm": 7.1875, | |
| "learning_rate": 3.1892453488058803e-08, | |
| "loss": 0.454, | |
| "step": 1196 | |
| }, | |
| { | |
| "epoch": 1.9345454545454546, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 3.0319357033591904e-08, | |
| "loss": 0.5787, | |
| "step": 1197 | |
| }, | |
| { | |
| "epoch": 1.9361616161616162, | |
| "grad_norm": 8.125, | |
| "learning_rate": 2.87859265750956e-08, | |
| "loss": 0.6485, | |
| "step": 1198 | |
| }, | |
| { | |
| "epoch": 1.9377777777777778, | |
| "grad_norm": 7.0, | |
| "learning_rate": 2.7292174351797317e-08, | |
| "loss": 0.3967, | |
| "step": 1199 | |
| }, | |
| { | |
| "epoch": 1.9393939393939394, | |
| "grad_norm": 7.46875, | |
| "learning_rate": 2.5838112286226123e-08, | |
| "loss": 0.7807, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.941010101010101, | |
| "grad_norm": 7.625, | |
| "learning_rate": 2.4423751984122835e-08, | |
| "loss": 0.834, | |
| "step": 1201 | |
| }, | |
| { | |
| "epoch": 1.9426262626262627, | |
| "grad_norm": 8.375, | |
| "learning_rate": 2.304910473434341e-08, | |
| "loss": 0.4306, | |
| "step": 1202 | |
| }, | |
| { | |
| "epoch": 1.9442424242424243, | |
| "grad_norm": 6.8125, | |
| "learning_rate": 2.1714181508770138e-08, | |
| "loss": 0.4254, | |
| "step": 1203 | |
| }, | |
| { | |
| "epoch": 1.945858585858586, | |
| "grad_norm": 6.125, | |
| "learning_rate": 2.0418992962224495e-08, | |
| "loss": 0.5607, | |
| "step": 1204 | |
| }, | |
| { | |
| "epoch": 1.9474747474747476, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 1.916354943238219e-08, | |
| "loss": 0.6082, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 1.9490909090909092, | |
| "grad_norm": 7.28125, | |
| "learning_rate": 1.7947860939688255e-08, | |
| "loss": 0.5441, | |
| "step": 1206 | |
| }, | |
| { | |
| "epoch": 1.9507070707070708, | |
| "grad_norm": 5.6875, | |
| "learning_rate": 1.6771937187280986e-08, | |
| "loss": 0.4209, | |
| "step": 1207 | |
| }, | |
| { | |
| "epoch": 1.9523232323232325, | |
| "grad_norm": 9.0625, | |
| "learning_rate": 1.563578756091144e-08, | |
| "loss": 0.5322, | |
| "step": 1208 | |
| }, | |
| { | |
| "epoch": 1.953939393939394, | |
| "grad_norm": 8.625, | |
| "learning_rate": 1.4539421128870745e-08, | |
| "loss": 0.455, | |
| "step": 1209 | |
| }, | |
| { | |
| "epoch": 1.9555555555555557, | |
| "grad_norm": 7.90625, | |
| "learning_rate": 1.3482846641914572e-08, | |
| "loss": 0.6719, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.9571717171717171, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 1.2466072533197093e-08, | |
| "loss": 0.4821, | |
| "step": 1211 | |
| }, | |
| { | |
| "epoch": 1.9587878787878787, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 1.1489106918200487e-08, | |
| "loss": 0.5824, | |
| "step": 1212 | |
| }, | |
| { | |
| "epoch": 1.9604040404040404, | |
| "grad_norm": 6.53125, | |
| "learning_rate": 1.0551957594672201e-08, | |
| "loss": 0.5726, | |
| "step": 1213 | |
| }, | |
| { | |
| "epoch": 1.962020202020202, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 9.654632042562229e-09, | |
| "loss": 0.3726, | |
| "step": 1214 | |
| }, | |
| { | |
| "epoch": 1.9636363636363636, | |
| "grad_norm": 5.65625, | |
| "learning_rate": 8.797137423963153e-09, | |
| "loss": 0.382, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 1.9652525252525253, | |
| "grad_norm": 6.5, | |
| "learning_rate": 7.979480583052423e-09, | |
| "loss": 0.5233, | |
| "step": 1216 | |
| }, | |
| { | |
| "epoch": 1.9668686868686869, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 7.201668046038501e-09, | |
| "loss": 0.501, | |
| "step": 1217 | |
| }, | |
| { | |
| "epoch": 1.9684848484848485, | |
| "grad_norm": 7.75, | |
| "learning_rate": 6.4637060211092395e-09, | |
| "loss": 0.5653, | |
| "step": 1218 | |
| }, | |
| { | |
| "epoch": 1.9701010101010101, | |
| "grad_norm": 5.40625, | |
| "learning_rate": 5.765600398381366e-09, | |
| "loss": 0.3169, | |
| "step": 1219 | |
| }, | |
| { | |
| "epoch": 1.9717171717171718, | |
| "grad_norm": 7.125, | |
| "learning_rate": 5.107356749853298e-09, | |
| "loss": 0.4751, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.9733333333333334, | |
| "grad_norm": 7.1875, | |
| "learning_rate": 4.488980329361847e-09, | |
| "loss": 0.3681, | |
| "step": 1221 | |
| }, | |
| { | |
| "epoch": 1.9749494949494948, | |
| "grad_norm": 7.9375, | |
| "learning_rate": 3.910476072539471e-09, | |
| "loss": 0.3618, | |
| "step": 1222 | |
| }, | |
| { | |
| "epoch": 1.9765656565656564, | |
| "grad_norm": 6.375, | |
| "learning_rate": 3.3718485967748626e-09, | |
| "loss": 0.3531, | |
| "step": 1223 | |
| }, | |
| { | |
| "epoch": 1.978181818181818, | |
| "grad_norm": 7.46875, | |
| "learning_rate": 2.8731022011757593e-09, | |
| "loss": 0.5474, | |
| "step": 1224 | |
| }, | |
| { | |
| "epoch": 1.9797979797979797, | |
| "grad_norm": 7.1875, | |
| "learning_rate": 2.4142408665367433e-09, | |
| "loss": 0.4938, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 1.9814141414141413, | |
| "grad_norm": 5.375, | |
| "learning_rate": 1.9952682553042722e-09, | |
| "loss": 0.3083, | |
| "step": 1226 | |
| }, | |
| { | |
| "epoch": 1.983030303030303, | |
| "grad_norm": 7.3125, | |
| "learning_rate": 1.6161877115494773e-09, | |
| "loss": 0.3612, | |
| "step": 1227 | |
| }, | |
| { | |
| "epoch": 1.9846464646464645, | |
| "grad_norm": 8.0, | |
| "learning_rate": 1.2770022609409628e-09, | |
| "loss": 0.393, | |
| "step": 1228 | |
| }, | |
| { | |
| "epoch": 1.9862626262626262, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 9.777146107209367e-10, | |
| "loss": 0.5605, | |
| "step": 1229 | |
| }, | |
| { | |
| "epoch": 1.9878787878787878, | |
| "grad_norm": 7.5, | |
| "learning_rate": 7.18327149683562e-10, | |
| "loss": 0.4834, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 1.9878787878787878, | |
| "eval_loss": 0.6601172089576721, | |
| "eval_runtime": 35.1117, | |
| "eval_samples_per_second": 31.329, | |
| "eval_steps_per_second": 3.93, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 1.9894949494949494, | |
| "grad_norm": 6.875, | |
| "learning_rate": 4.988419481549711e-10, | |
| "loss": 0.5459, | |
| "step": 1231 | |
| }, | |
| { | |
| "epoch": 1.991111111111111, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 3.1926075797827914e-10, | |
| "loss": 0.4832, | |
| "step": 1232 | |
| }, | |
| { | |
| "epoch": 1.9927272727272727, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 1.7958501249859538e-10, | |
| "loss": 0.4882, | |
| "step": 1233 | |
| }, | |
| { | |
| "epoch": 1.9943434343434343, | |
| "grad_norm": 8.375, | |
| "learning_rate": 7.9815826551366e-11, | |
| "loss": 0.6767, | |
| "step": 1234 | |
| }, | |
| { | |
| "epoch": 1.995959595959596, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 1.9953996454047386e-11, | |
| "loss": 0.4317, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 1.9975757575757576, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 0.0, | |
| "loss": 0.5997, | |
| "step": 1236 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1236, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "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": 2.926082991007826e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |