| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 754, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.002652519893899204, | |
| "grad_norm": 972.6011082224239, | |
| "learning_rate": 1.3157894736842107e-07, | |
| "loss": 12.3485, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.005305039787798408, | |
| "grad_norm": 1176.0761331054841, | |
| "learning_rate": 2.6315789473684213e-07, | |
| "loss": 12.3023, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.007957559681697613, | |
| "grad_norm": 971.133776912219, | |
| "learning_rate": 3.9473684210526315e-07, | |
| "loss": 12.2062, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.010610079575596816, | |
| "grad_norm": 1161.3124856806703, | |
| "learning_rate": 5.263157894736843e-07, | |
| "loss": 12.2629, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.013262599469496022, | |
| "grad_norm": 1332.84779500764, | |
| "learning_rate": 6.578947368421053e-07, | |
| "loss": 12.3509, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.015915119363395226, | |
| "grad_norm": 850.1180640186841, | |
| "learning_rate": 7.894736842105263e-07, | |
| "loss": 12.1591, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.01856763925729443, | |
| "grad_norm": 1123.22480648108, | |
| "learning_rate": 9.210526315789474e-07, | |
| "loss": 11.8747, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.021220159151193633, | |
| "grad_norm": 953.0673098616203, | |
| "learning_rate": 1.0526315789473685e-06, | |
| "loss": 11.6101, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.023872679045092837, | |
| "grad_norm": 1117.0646961596742, | |
| "learning_rate": 1.1842105263157894e-06, | |
| "loss": 11.4673, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.026525198938992044, | |
| "grad_norm": 1083.1435085272988, | |
| "learning_rate": 1.3157894736842106e-06, | |
| "loss": 10.2665, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.029177718832891247, | |
| "grad_norm": 925.8133851516698, | |
| "learning_rate": 1.4473684210526317e-06, | |
| "loss": 9.8587, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.03183023872679045, | |
| "grad_norm": 835.5900642092195, | |
| "learning_rate": 1.5789473684210526e-06, | |
| "loss": 9.5232, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.034482758620689655, | |
| "grad_norm": 737.5585272473943, | |
| "learning_rate": 1.710526315789474e-06, | |
| "loss": 8.6792, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.03713527851458886, | |
| "grad_norm": 973.6275891557045, | |
| "learning_rate": 1.8421052631578948e-06, | |
| "loss": 8.5067, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.03978779840848806, | |
| "grad_norm": 987.1822647467712, | |
| "learning_rate": 1.973684210526316e-06, | |
| "loss": 8.1794, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.042440318302387266, | |
| "grad_norm": 825.59398545438, | |
| "learning_rate": 2.105263157894737e-06, | |
| "loss": 7.7437, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.04509283819628647, | |
| "grad_norm": 664.0819933865747, | |
| "learning_rate": 2.236842105263158e-06, | |
| "loss": 6.4606, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.04774535809018567, | |
| "grad_norm": 399.57434205536475, | |
| "learning_rate": 2.368421052631579e-06, | |
| "loss": 5.6519, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.050397877984084884, | |
| "grad_norm": 504.5286526909389, | |
| "learning_rate": 2.5e-06, | |
| "loss": 5.3274, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.05305039787798409, | |
| "grad_norm": 388.29875810234256, | |
| "learning_rate": 2.631578947368421e-06, | |
| "loss": 5.0673, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05570291777188329, | |
| "grad_norm": 454.4658090119901, | |
| "learning_rate": 2.7631578947368424e-06, | |
| "loss": 4.7588, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.058355437665782495, | |
| "grad_norm": 366.33318334352083, | |
| "learning_rate": 2.8947368421052634e-06, | |
| "loss": 4.5846, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.0610079575596817, | |
| "grad_norm": 328.1174135802827, | |
| "learning_rate": 3.0263157894736843e-06, | |
| "loss": 4.4205, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.0636604774535809, | |
| "grad_norm": 466.9071960005728, | |
| "learning_rate": 3.157894736842105e-06, | |
| "loss": 4.2348, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.06631299734748011, | |
| "grad_norm": 331.2124348618216, | |
| "learning_rate": 3.289473684210527e-06, | |
| "loss": 3.9631, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06896551724137931, | |
| "grad_norm": 348.3850354268525, | |
| "learning_rate": 3.421052631578948e-06, | |
| "loss": 3.6177, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.07161803713527852, | |
| "grad_norm": 332.5916510702998, | |
| "learning_rate": 3.5526315789473687e-06, | |
| "loss": 3.5028, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.07427055702917772, | |
| "grad_norm": 456.9904876458593, | |
| "learning_rate": 3.6842105263157896e-06, | |
| "loss": 3.6164, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.07692307692307693, | |
| "grad_norm": 436.5446671330148, | |
| "learning_rate": 3.815789473684211e-06, | |
| "loss": 3.3163, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.07957559681697612, | |
| "grad_norm": 266.0326848946069, | |
| "learning_rate": 3.947368421052632e-06, | |
| "loss": 3.4249, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.08222811671087533, | |
| "grad_norm": 264.7876966039979, | |
| "learning_rate": 4.078947368421053e-06, | |
| "loss": 3.3034, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.08488063660477453, | |
| "grad_norm": 368.7949684184987, | |
| "learning_rate": 4.210526315789474e-06, | |
| "loss": 3.1039, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.08753315649867374, | |
| "grad_norm": 255.2314821527665, | |
| "learning_rate": 4.342105263157895e-06, | |
| "loss": 3.0517, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.09018567639257294, | |
| "grad_norm": 234.26369572444867, | |
| "learning_rate": 4.473684210526316e-06, | |
| "loss": 2.9529, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.09283819628647215, | |
| "grad_norm": 230.93299067113927, | |
| "learning_rate": 4.605263157894737e-06, | |
| "loss": 2.8815, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.09549071618037135, | |
| "grad_norm": 232.19712182995283, | |
| "learning_rate": 4.736842105263158e-06, | |
| "loss": 2.9011, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.09814323607427056, | |
| "grad_norm": 233.9410464893109, | |
| "learning_rate": 4.8684210526315795e-06, | |
| "loss": 2.7754, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.10079575596816977, | |
| "grad_norm": 240.55698339891316, | |
| "learning_rate": 5e-06, | |
| "loss": 2.7326, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.10344827586206896, | |
| "grad_norm": 204.71398942027966, | |
| "learning_rate": 5.131578947368422e-06, | |
| "loss": 2.7594, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.10610079575596817, | |
| "grad_norm": 191.82223594470645, | |
| "learning_rate": 5.263157894736842e-06, | |
| "loss": 2.665, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.10875331564986737, | |
| "grad_norm": 391.21195275764177, | |
| "learning_rate": 5.394736842105264e-06, | |
| "loss": 2.842, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.11140583554376658, | |
| "grad_norm": 199.44545824533185, | |
| "learning_rate": 5.526315789473685e-06, | |
| "loss": 2.7292, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.11405835543766578, | |
| "grad_norm": 214.39560903670093, | |
| "learning_rate": 5.657894736842106e-06, | |
| "loss": 2.7937, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.11671087533156499, | |
| "grad_norm": 135.53162028919942, | |
| "learning_rate": 5.789473684210527e-06, | |
| "loss": 2.6044, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.11936339522546419, | |
| "grad_norm": 184.52491951315736, | |
| "learning_rate": 5.921052631578948e-06, | |
| "loss": 2.6461, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1220159151193634, | |
| "grad_norm": 164.26425091990012, | |
| "learning_rate": 6.0526315789473685e-06, | |
| "loss": 2.6699, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.1246684350132626, | |
| "grad_norm": 156.31534034010923, | |
| "learning_rate": 6.18421052631579e-06, | |
| "loss": 2.6341, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.1273209549071618, | |
| "grad_norm": 133.568318629801, | |
| "learning_rate": 6.31578947368421e-06, | |
| "loss": 2.5945, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.129973474801061, | |
| "grad_norm": 129.47868792668547, | |
| "learning_rate": 6.447368421052632e-06, | |
| "loss": 2.4362, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.13262599469496023, | |
| "grad_norm": 148.49435032805644, | |
| "learning_rate": 6.578947368421054e-06, | |
| "loss": 2.6795, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.13527851458885942, | |
| "grad_norm": 140.9084313756403, | |
| "learning_rate": 6.710526315789474e-06, | |
| "loss": 2.5829, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.13793103448275862, | |
| "grad_norm": 152.83640328644313, | |
| "learning_rate": 6.842105263157896e-06, | |
| "loss": 2.9126, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.14058355437665782, | |
| "grad_norm": 123.55761772751443, | |
| "learning_rate": 6.973684210526316e-06, | |
| "loss": 2.6503, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.14323607427055704, | |
| "grad_norm": 100.61578026023533, | |
| "learning_rate": 7.1052631578947375e-06, | |
| "loss": 2.3635, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.14588859416445624, | |
| "grad_norm": 159.03053787861532, | |
| "learning_rate": 7.236842105263158e-06, | |
| "loss": 3.1021, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.14854111405835543, | |
| "grad_norm": 123.64692196540877, | |
| "learning_rate": 7.368421052631579e-06, | |
| "loss": 2.8577, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.15119363395225463, | |
| "grad_norm": 87.70627460451284, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 2.3376, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 86.58878221665253, | |
| "learning_rate": 7.631578947368423e-06, | |
| "loss": 2.4375, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.15649867374005305, | |
| "grad_norm": 134.90794157537533, | |
| "learning_rate": 7.763157894736843e-06, | |
| "loss": 2.5644, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.15915119363395225, | |
| "grad_norm": 115.31216244306997, | |
| "learning_rate": 7.894736842105265e-06, | |
| "loss": 2.4072, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.16180371352785147, | |
| "grad_norm": 120.1225075682861, | |
| "learning_rate": 8.026315789473685e-06, | |
| "loss": 2.5956, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.16445623342175067, | |
| "grad_norm": 136.94352276466043, | |
| "learning_rate": 8.157894736842106e-06, | |
| "loss": 2.4038, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.16710875331564987, | |
| "grad_norm": 111.78482749569982, | |
| "learning_rate": 8.289473684210526e-06, | |
| "loss": 2.3572, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.16976127320954906, | |
| "grad_norm": 100.8891133060052, | |
| "learning_rate": 8.421052631578948e-06, | |
| "loss": 2.257, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.1724137931034483, | |
| "grad_norm": 114.86478282978065, | |
| "learning_rate": 8.552631578947368e-06, | |
| "loss": 2.2236, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.17506631299734748, | |
| "grad_norm": 114.08371988056022, | |
| "learning_rate": 8.68421052631579e-06, | |
| "loss": 2.228, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.17771883289124668, | |
| "grad_norm": 113.07954964295763, | |
| "learning_rate": 8.81578947368421e-06, | |
| "loss": 2.2422, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.18037135278514588, | |
| "grad_norm": 109.31976151906129, | |
| "learning_rate": 8.947368421052632e-06, | |
| "loss": 2.1732, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.1830238726790451, | |
| "grad_norm": 90.99647003375713, | |
| "learning_rate": 9.078947368421054e-06, | |
| "loss": 2.1485, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.1856763925729443, | |
| "grad_norm": 120.51965849864614, | |
| "learning_rate": 9.210526315789474e-06, | |
| "loss": 2.2091, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.1883289124668435, | |
| "grad_norm": 96.16044170894298, | |
| "learning_rate": 9.342105263157895e-06, | |
| "loss": 2.0965, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.1909814323607427, | |
| "grad_norm": 113.04131258428713, | |
| "learning_rate": 9.473684210526315e-06, | |
| "loss": 2.232, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.19363395225464192, | |
| "grad_norm": 93.19087300109942, | |
| "learning_rate": 9.605263157894737e-06, | |
| "loss": 2.062, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.1962864721485411, | |
| "grad_norm": 122.34858082117903, | |
| "learning_rate": 9.736842105263159e-06, | |
| "loss": 2.2507, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.1989389920424403, | |
| "grad_norm": 123.75061956942726, | |
| "learning_rate": 9.868421052631579e-06, | |
| "loss": 2.0923, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.20159151193633953, | |
| "grad_norm": 95.31909307451943, | |
| "learning_rate": 1e-05, | |
| "loss": 2.0863, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.20424403183023873, | |
| "grad_norm": 111.39378682665635, | |
| "learning_rate": 9.999946324068588e-06, | |
| "loss": 2.0406, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.20689655172413793, | |
| "grad_norm": 80.94452155733912, | |
| "learning_rate": 9.999785297426788e-06, | |
| "loss": 2.0968, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.20954907161803712, | |
| "grad_norm": 84.78034162492699, | |
| "learning_rate": 9.999516923531906e-06, | |
| "loss": 1.9962, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.21220159151193635, | |
| "grad_norm": 78.02592069732914, | |
| "learning_rate": 9.999141208146029e-06, | |
| "loss": 1.9925, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.21485411140583555, | |
| "grad_norm": 73.73682362098175, | |
| "learning_rate": 9.998658159335903e-06, | |
| "loss": 1.9603, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.21750663129973474, | |
| "grad_norm": 90.67851878871835, | |
| "learning_rate": 9.998067787472772e-06, | |
| "loss": 1.8926, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.22015915119363394, | |
| "grad_norm": 62.82693612482471, | |
| "learning_rate": 9.997370105232134e-06, | |
| "loss": 1.8182, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.22281167108753316, | |
| "grad_norm": 75.05067588018298, | |
| "learning_rate": 9.99656512759349e-06, | |
| "loss": 1.8357, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.22546419098143236, | |
| "grad_norm": 49.91545803670374, | |
| "learning_rate": 9.995652871840006e-06, | |
| "loss": 1.7348, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.22811671087533156, | |
| "grad_norm": 91.12448174019961, | |
| "learning_rate": 9.994633357558158e-06, | |
| "loss": 1.8954, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.23076923076923078, | |
| "grad_norm": 52.94023906195075, | |
| "learning_rate": 9.993506606637297e-06, | |
| "loss": 1.8498, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.23342175066312998, | |
| "grad_norm": 58.195205266701194, | |
| "learning_rate": 9.992272643269181e-06, | |
| "loss": 1.7826, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.23607427055702918, | |
| "grad_norm": 66.84628315063024, | |
| "learning_rate": 9.990931493947467e-06, | |
| "loss": 1.7434, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.23872679045092837, | |
| "grad_norm": 73.69628201213949, | |
| "learning_rate": 9.989483187467128e-06, | |
| "loss": 1.8642, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.2413793103448276, | |
| "grad_norm": 52.58610935369465, | |
| "learning_rate": 9.987927754923844e-06, | |
| "loss": 1.836, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.2440318302387268, | |
| "grad_norm": 78.35414171000565, | |
| "learning_rate": 9.986265229713332e-06, | |
| "loss": 1.932, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.246684350132626, | |
| "grad_norm": 49.2141069227244, | |
| "learning_rate": 9.98449564753063e-06, | |
| "loss": 1.7129, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.2493368700265252, | |
| "grad_norm": 61.31351433697711, | |
| "learning_rate": 9.982619046369321e-06, | |
| "loss": 1.7202, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.2519893899204244, | |
| "grad_norm": 68.28056374656848, | |
| "learning_rate": 9.980635466520738e-06, | |
| "loss": 1.6906, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2546419098143236, | |
| "grad_norm": 46.951538432249045, | |
| "learning_rate": 9.978544950573075e-06, | |
| "loss": 1.632, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.2572944297082228, | |
| "grad_norm": 67.90324049138877, | |
| "learning_rate": 9.976347543410487e-06, | |
| "loss": 1.656, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.259946949602122, | |
| "grad_norm": 64.68477738459366, | |
| "learning_rate": 9.974043292212129e-06, | |
| "loss": 1.4681, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.2625994694960212, | |
| "grad_norm": 52.8724055352585, | |
| "learning_rate": 9.97163224645113e-06, | |
| "loss": 1.5811, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.26525198938992045, | |
| "grad_norm": 35.35995771432847, | |
| "learning_rate": 9.96911445789354e-06, | |
| "loss": 1.3361, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.26790450928381965, | |
| "grad_norm": 44.61978123978171, | |
| "learning_rate": 9.966489980597217e-06, | |
| "loss": 1.5174, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.27055702917771884, | |
| "grad_norm": 57.19406544164446, | |
| "learning_rate": 9.963758870910672e-06, | |
| "loss": 1.4109, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.27320954907161804, | |
| "grad_norm": 50.526490813756546, | |
| "learning_rate": 9.960921187471841e-06, | |
| "loss": 1.3191, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.27586206896551724, | |
| "grad_norm": 101.13725470627442, | |
| "learning_rate": 9.957976991206847e-06, | |
| "loss": 1.6531, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.27851458885941643, | |
| "grad_norm": 52.15413810029031, | |
| "learning_rate": 9.95492634532868e-06, | |
| "loss": 1.4429, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.28116710875331563, | |
| "grad_norm": 41.87999077534381, | |
| "learning_rate": 9.951769315335843e-06, | |
| "loss": 1.238, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.2838196286472148, | |
| "grad_norm": 37.40295368354187, | |
| "learning_rate": 9.94850596901095e-06, | |
| "loss": 1.1688, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.2864721485411141, | |
| "grad_norm": 68.04586288775573, | |
| "learning_rate": 9.94513637641926e-06, | |
| "loss": 1.461, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.2891246684350133, | |
| "grad_norm": 44.169402573464126, | |
| "learning_rate": 9.94166060990718e-06, | |
| "loss": 1.1791, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.2917771883289125, | |
| "grad_norm": 47.80762405569716, | |
| "learning_rate": 9.938078744100713e-06, | |
| "loss": 1.1185, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.29442970822281167, | |
| "grad_norm": 58.966613634353685, | |
| "learning_rate": 9.934390855903852e-06, | |
| "loss": 1.3944, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.29708222811671087, | |
| "grad_norm": 47.03397627383742, | |
| "learning_rate": 9.930597024496933e-06, | |
| "loss": 1.23, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.29973474801061006, | |
| "grad_norm": 40.39452844321203, | |
| "learning_rate": 9.926697331334924e-06, | |
| "loss": 1.0874, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.30238726790450926, | |
| "grad_norm": 40.58586118259995, | |
| "learning_rate": 9.922691860145696e-06, | |
| "loss": 1.1465, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.3050397877984085, | |
| "grad_norm": 81.40197786842415, | |
| "learning_rate": 9.918580696928206e-06, | |
| "loss": 1.5197, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 42.37482487824635, | |
| "learning_rate": 9.91436392995066e-06, | |
| "loss": 1.1805, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.3103448275862069, | |
| "grad_norm": 31.02071243624101, | |
| "learning_rate": 9.910041649748613e-06, | |
| "loss": 0.9539, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.3129973474801061, | |
| "grad_norm": 41.401274375888676, | |
| "learning_rate": 9.905613949123036e-06, | |
| "loss": 0.9762, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.3156498673740053, | |
| "grad_norm": 49.185437588379855, | |
| "learning_rate": 9.901080923138308e-06, | |
| "loss": 1.0671, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.3183023872679045, | |
| "grad_norm": 25.940285977239768, | |
| "learning_rate": 9.896442669120188e-06, | |
| "loss": 0.8426, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.3209549071618037, | |
| "grad_norm": 46.84002530549447, | |
| "learning_rate": 9.891699286653714e-06, | |
| "loss": 1.0983, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.32360742705570295, | |
| "grad_norm": 40.94275204536911, | |
| "learning_rate": 9.886850877581079e-06, | |
| "loss": 1.049, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.32625994694960214, | |
| "grad_norm": 28.321978107011155, | |
| "learning_rate": 9.88189754599943e-06, | |
| "loss": 0.8165, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.32891246684350134, | |
| "grad_norm": 30.5824466235061, | |
| "learning_rate": 9.87683939825864e-06, | |
| "loss": 0.7374, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.33156498673740054, | |
| "grad_norm": 30.73141505007011, | |
| "learning_rate": 9.87167654295903e-06, | |
| "loss": 0.7421, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.33421750663129973, | |
| "grad_norm": 38.284482417984805, | |
| "learning_rate": 9.866409090949023e-06, | |
| "loss": 1.0553, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.33687002652519893, | |
| "grad_norm": 51.217111013485166, | |
| "learning_rate": 9.861037155322777e-06, | |
| "loss": 0.907, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.3395225464190981, | |
| "grad_norm": 35.667190844261036, | |
| "learning_rate": 9.855560851417752e-06, | |
| "loss": 0.7419, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.3421750663129973, | |
| "grad_norm": 47.52472262750116, | |
| "learning_rate": 9.849980296812231e-06, | |
| "loss": 0.9425, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.3448275862068966, | |
| "grad_norm": 48.31204711246393, | |
| "learning_rate": 9.844295611322804e-06, | |
| "loss": 0.8446, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.34748010610079577, | |
| "grad_norm": 30.190906581474252, | |
| "learning_rate": 9.838506917001784e-06, | |
| "loss": 0.6881, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.35013262599469497, | |
| "grad_norm": 46.794947303199244, | |
| "learning_rate": 9.832614338134595e-06, | |
| "loss": 0.6701, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.35278514588859416, | |
| "grad_norm": 44.10169615851438, | |
| "learning_rate": 9.826618001237101e-06, | |
| "loss": 0.701, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.35543766578249336, | |
| "grad_norm": 40.06922950860214, | |
| "learning_rate": 9.82051803505289e-06, | |
| "loss": 0.7066, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.35809018567639256, | |
| "grad_norm": 51.10173470889434, | |
| "learning_rate": 9.814314570550506e-06, | |
| "loss": 0.7107, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.36074270557029176, | |
| "grad_norm": 45.42561699905777, | |
| "learning_rate": 9.808007740920647e-06, | |
| "loss": 0.7882, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.363395225464191, | |
| "grad_norm": 30.22023691550999, | |
| "learning_rate": 9.80159768157329e-06, | |
| "loss": 0.5774, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.3660477453580902, | |
| "grad_norm": 37.33347097999451, | |
| "learning_rate": 9.795084530134801e-06, | |
| "loss": 0.6183, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.3687002652519894, | |
| "grad_norm": 31.72442511930728, | |
| "learning_rate": 9.788468426444968e-06, | |
| "loss": 0.799, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.3713527851458886, | |
| "grad_norm": 44.85515304779163, | |
| "learning_rate": 9.781749512554e-06, | |
| "loss": 0.6424, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3740053050397878, | |
| "grad_norm": 39.0917840315655, | |
| "learning_rate": 9.774927932719484e-06, | |
| "loss": 0.6998, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.376657824933687, | |
| "grad_norm": 31.573363442317945, | |
| "learning_rate": 9.768003833403278e-06, | |
| "loss": 0.4961, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.3793103448275862, | |
| "grad_norm": 61.98565644042616, | |
| "learning_rate": 9.760977363268374e-06, | |
| "loss": 1.0158, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.3819628647214854, | |
| "grad_norm": 39.75425925803264, | |
| "learning_rate": 9.753848673175707e-06, | |
| "loss": 0.6281, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "grad_norm": 48.98336149254754, | |
| "learning_rate": 9.746617916180906e-06, | |
| "loss": 0.6136, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.38726790450928383, | |
| "grad_norm": 57.044641212074524, | |
| "learning_rate": 9.739285247531019e-06, | |
| "loss": 0.7083, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.38992042440318303, | |
| "grad_norm": 63.365556425165224, | |
| "learning_rate": 9.731850824661171e-06, | |
| "loss": 0.5737, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.3925729442970822, | |
| "grad_norm": 29.826608394688588, | |
| "learning_rate": 9.724314807191197e-06, | |
| "loss": 0.4653, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.3952254641909814, | |
| "grad_norm": 47.58176548578861, | |
| "learning_rate": 9.716677356922193e-06, | |
| "loss": 0.5239, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.3978779840848806, | |
| "grad_norm": 34.51576173886699, | |
| "learning_rate": 9.708938637833065e-06, | |
| "loss": 0.5865, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.4005305039787798, | |
| "grad_norm": 34.740787732790245, | |
| "learning_rate": 9.701098816076995e-06, | |
| "loss": 0.5071, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.40318302387267907, | |
| "grad_norm": 35.31607232875537, | |
| "learning_rate": 9.693158059977879e-06, | |
| "loss": 0.494, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.40583554376657827, | |
| "grad_norm": 51.83713992645971, | |
| "learning_rate": 9.685116540026703e-06, | |
| "loss": 0.6192, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.40848806366047746, | |
| "grad_norm": 37.44373095646274, | |
| "learning_rate": 9.6769744288779e-06, | |
| "loss": 0.5033, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.41114058355437666, | |
| "grad_norm": 109.25763221989088, | |
| "learning_rate": 9.668731901345632e-06, | |
| "loss": 0.8446, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.41379310344827586, | |
| "grad_norm": 72.46155458193397, | |
| "learning_rate": 9.660389134400034e-06, | |
| "loss": 0.7201, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.41644562334217505, | |
| "grad_norm": 92.07617516331763, | |
| "learning_rate": 9.651946307163417e-06, | |
| "loss": 0.9007, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.41909814323607425, | |
| "grad_norm": 28.589438406273647, | |
| "learning_rate": 9.643403600906433e-06, | |
| "loss": 0.3304, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.4217506631299735, | |
| "grad_norm": 33.709892226026334, | |
| "learning_rate": 9.634761199044165e-06, | |
| "loss": 0.5636, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.4244031830238727, | |
| "grad_norm": 158.4075248636022, | |
| "learning_rate": 9.626019287132202e-06, | |
| "loss": 0.4064, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.4270557029177719, | |
| "grad_norm": 38.574563627234824, | |
| "learning_rate": 9.617178052862648e-06, | |
| "loss": 0.4748, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.4297082228116711, | |
| "grad_norm": 45.60608001495772, | |
| "learning_rate": 9.608237686060099e-06, | |
| "loss": 0.6974, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.4323607427055703, | |
| "grad_norm": 47.48506948979431, | |
| "learning_rate": 9.599198378677559e-06, | |
| "loss": 0.4992, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.4350132625994695, | |
| "grad_norm": 31.370496495096628, | |
| "learning_rate": 9.590060324792328e-06, | |
| "loss": 0.4248, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.4376657824933687, | |
| "grad_norm": 30.604914876341343, | |
| "learning_rate": 9.580823720601824e-06, | |
| "loss": 0.3956, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.4403183023872679, | |
| "grad_norm": 35.21040129146704, | |
| "learning_rate": 9.571488764419381e-06, | |
| "loss": 0.4175, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.44297082228116713, | |
| "grad_norm": 35.4113416422159, | |
| "learning_rate": 9.562055656669988e-06, | |
| "loss": 0.5203, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.44562334217506633, | |
| "grad_norm": 39.91487718635988, | |
| "learning_rate": 9.552524599885982e-06, | |
| "loss": 0.378, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.4482758620689655, | |
| "grad_norm": 39.85559208094948, | |
| "learning_rate": 9.542895798702702e-06, | |
| "loss": 0.6038, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.4509283819628647, | |
| "grad_norm": 25.908557322404143, | |
| "learning_rate": 9.5331694598541e-06, | |
| "loss": 0.3786, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.4535809018567639, | |
| "grad_norm": 48.47896763638714, | |
| "learning_rate": 9.52334579216829e-06, | |
| "loss": 0.5274, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.4562334217506631, | |
| "grad_norm": 36.56071047636118, | |
| "learning_rate": 9.51342500656308e-06, | |
| "loss": 0.4753, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.4588859416445623, | |
| "grad_norm": 40.92239723888123, | |
| "learning_rate": 9.503407316041432e-06, | |
| "loss": 0.4009, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 24.061372567374256, | |
| "learning_rate": 9.493292935686896e-06, | |
| "loss": 0.3719, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.46419098143236076, | |
| "grad_norm": 32.5043140267615, | |
| "learning_rate": 9.483082082658984e-06, | |
| "loss": 0.348, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.46684350132625996, | |
| "grad_norm": 33.49855253180825, | |
| "learning_rate": 9.472774976188515e-06, | |
| "loss": 0.2976, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.46949602122015915, | |
| "grad_norm": 43.029659310923336, | |
| "learning_rate": 9.462371837572907e-06, | |
| "loss": 0.3728, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.47214854111405835, | |
| "grad_norm": 31.765940790477142, | |
| "learning_rate": 9.451872890171419e-06, | |
| "loss": 0.2834, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.47480106100795755, | |
| "grad_norm": 46.81294121573895, | |
| "learning_rate": 9.441278359400366e-06, | |
| "loss": 0.3757, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.47745358090185674, | |
| "grad_norm": 46.581740603921325, | |
| "learning_rate": 9.430588472728271e-06, | |
| "loss": 0.3292, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.48010610079575594, | |
| "grad_norm": 33.689497051794476, | |
| "learning_rate": 9.41980345967098e-06, | |
| "loss": 0.3834, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.4827586206896552, | |
| "grad_norm": 61.128747699132035, | |
| "learning_rate": 9.408923551786742e-06, | |
| "loss": 0.3354, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.4854111405835544, | |
| "grad_norm": 22.95346596441586, | |
| "learning_rate": 9.397948982671237e-06, | |
| "loss": 0.3053, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.4880636604774536, | |
| "grad_norm": 30.898534305601594, | |
| "learning_rate": 9.386879987952549e-06, | |
| "loss": 0.2642, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.4907161803713528, | |
| "grad_norm": 33.57899929124302, | |
| "learning_rate": 9.375716805286122e-06, | |
| "loss": 0.3217, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.493368700265252, | |
| "grad_norm": 62.90298927426579, | |
| "learning_rate": 9.364459674349642e-06, | |
| "loss": 0.3244, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.4960212201591512, | |
| "grad_norm": 43.421745744085165, | |
| "learning_rate": 9.353108836837907e-06, | |
| "loss": 0.4662, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.4986737400530504, | |
| "grad_norm": 84.88782280618557, | |
| "learning_rate": 9.341664536457626e-06, | |
| "loss": 0.4652, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.5013262599469496, | |
| "grad_norm": 41.48296017032388, | |
| "learning_rate": 9.330127018922195e-06, | |
| "loss": 0.4524, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.5039787798408488, | |
| "grad_norm": 54.307781009627426, | |
| "learning_rate": 9.318496531946411e-06, | |
| "loss": 0.4255, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.506631299734748, | |
| "grad_norm": 28.526745313141138, | |
| "learning_rate": 9.306773325241161e-06, | |
| "loss": 0.3938, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.5092838196286472, | |
| "grad_norm": 32.22292357278733, | |
| "learning_rate": 9.294957650508065e-06, | |
| "loss": 0.4339, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.5119363395225465, | |
| "grad_norm": 71.01658099069965, | |
| "learning_rate": 9.283049761434059e-06, | |
| "loss": 0.3714, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.5145888594164456, | |
| "grad_norm": 74.60447764546221, | |
| "learning_rate": 9.27104991368596e-06, | |
| "loss": 0.4595, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.5172413793103449, | |
| "grad_norm": 34.72647596903285, | |
| "learning_rate": 9.258958364904966e-06, | |
| "loss": 0.4545, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.519893899204244, | |
| "grad_norm": 26.331798545183236, | |
| "learning_rate": 9.246775374701139e-06, | |
| "loss": 0.3962, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.5225464190981433, | |
| "grad_norm": 24.700967086911067, | |
| "learning_rate": 9.234501204647814e-06, | |
| "loss": 0.3429, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.5251989389920424, | |
| "grad_norm": 54.171499277303575, | |
| "learning_rate": 9.222136118275996e-06, | |
| "loss": 0.4216, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.5278514588859416, | |
| "grad_norm": 27.43993544126369, | |
| "learning_rate": 9.209680381068698e-06, | |
| "loss": 0.6582, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.5305039787798409, | |
| "grad_norm": 22.598607279543927, | |
| "learning_rate": 9.197134260455233e-06, | |
| "loss": 0.4346, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.53315649867374, | |
| "grad_norm": 25.120430306898687, | |
| "learning_rate": 9.184498025805493e-06, | |
| "loss": 0.3572, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.5358090185676393, | |
| "grad_norm": 30.90653283841454, | |
| "learning_rate": 9.171771948424138e-06, | |
| "loss": 0.5035, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.5384615384615384, | |
| "grad_norm": 38.745294217181815, | |
| "learning_rate": 9.158956301544791e-06, | |
| "loss": 0.6001, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.5411140583554377, | |
| "grad_norm": 12.99132628044725, | |
| "learning_rate": 9.146051360324166e-06, | |
| "loss": 0.2217, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.5437665782493368, | |
| "grad_norm": 21.483605650056905, | |
| "learning_rate": 9.13305740183616e-06, | |
| "loss": 0.3407, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.5464190981432361, | |
| "grad_norm": 22.04202864170304, | |
| "learning_rate": 9.119974705065902e-06, | |
| "loss": 0.3155, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.5490716180371353, | |
| "grad_norm": 34.047701260544905, | |
| "learning_rate": 9.106803550903765e-06, | |
| "loss": 0.2755, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.5517241379310345, | |
| "grad_norm": 30.616049044407195, | |
| "learning_rate": 9.093544222139338e-06, | |
| "loss": 0.3207, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.5543766578249337, | |
| "grad_norm": 19.414012419163278, | |
| "learning_rate": 9.080197003455347e-06, | |
| "loss": 0.2854, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.5570291777188329, | |
| "grad_norm": 34.50661859771314, | |
| "learning_rate": 9.066762181421552e-06, | |
| "loss": 0.2571, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.5596816976127321, | |
| "grad_norm": 24.18802559390643, | |
| "learning_rate": 9.053240044488587e-06, | |
| "loss": 0.3284, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.5623342175066313, | |
| "grad_norm": 26.231041064560618, | |
| "learning_rate": 9.039630882981769e-06, | |
| "loss": 0.2911, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.5649867374005305, | |
| "grad_norm": 18.744950063838594, | |
| "learning_rate": 9.025934989094866e-06, | |
| "loss": 0.2078, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.5676392572944297, | |
| "grad_norm": 32.40690745848201, | |
| "learning_rate": 9.012152656883824e-06, | |
| "loss": 0.2154, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.5702917771883289, | |
| "grad_norm": 19.852273416933915, | |
| "learning_rate": 8.998284182260448e-06, | |
| "loss": 0.1813, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.5729442970822282, | |
| "grad_norm": 51.45945703972048, | |
| "learning_rate": 8.984329862986056e-06, | |
| "loss": 0.2579, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.5755968169761273, | |
| "grad_norm": 26.7842923687961, | |
| "learning_rate": 8.970289998665083e-06, | |
| "loss": 0.326, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.5782493368700266, | |
| "grad_norm": 25.798048935320374, | |
| "learning_rate": 8.956164890738643e-06, | |
| "loss": 0.2557, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.5809018567639257, | |
| "grad_norm": 31.147062160430725, | |
| "learning_rate": 8.941954842478071e-06, | |
| "loss": 0.3109, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.583554376657825, | |
| "grad_norm": 33.73392858792033, | |
| "learning_rate": 8.927660158978392e-06, | |
| "loss": 0.192, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.5862068965517241, | |
| "grad_norm": 52.12002946089713, | |
| "learning_rate": 8.913281147151793e-06, | |
| "loss": 0.3124, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.5888594164456233, | |
| "grad_norm": 31.949599003240756, | |
| "learning_rate": 8.898818115721009e-06, | |
| "loss": 0.3063, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.5915119363395226, | |
| "grad_norm": 46.81555399953881, | |
| "learning_rate": 8.884271375212714e-06, | |
| "loss": 0.3242, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.5941644562334217, | |
| "grad_norm": 35.262038994934144, | |
| "learning_rate": 8.86964123795085e-06, | |
| "loss": 0.2459, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.596816976127321, | |
| "grad_norm": 37.79974742141044, | |
| "learning_rate": 8.85492801804991e-06, | |
| "loss": 0.2955, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.5994694960212201, | |
| "grad_norm": 54.98818702954406, | |
| "learning_rate": 8.84013203140821e-06, | |
| "loss": 0.367, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.6021220159151194, | |
| "grad_norm": 40.40698875247452, | |
| "learning_rate": 8.825253595701097e-06, | |
| "loss": 0.2885, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.6047745358090185, | |
| "grad_norm": 52.621249803374496, | |
| "learning_rate": 8.810293030374126e-06, | |
| "loss": 0.2208, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.6074270557029178, | |
| "grad_norm": 63.43714192736091, | |
| "learning_rate": 8.795250656636207e-06, | |
| "loss": 0.4095, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.610079575596817, | |
| "grad_norm": 45.5316777547875, | |
| "learning_rate": 8.780126797452713e-06, | |
| "loss": 0.3347, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.6127320954907162, | |
| "grad_norm": 51.68778010061592, | |
| "learning_rate": 8.764921777538533e-06, | |
| "loss": 0.5593, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 187.21679917554656, | |
| "learning_rate": 8.749635923351108e-06, | |
| "loss": 0.2935, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.6180371352785146, | |
| "grad_norm": 34.59940203028243, | |
| "learning_rate": 8.734269563083424e-06, | |
| "loss": 0.2359, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.6206896551724138, | |
| "grad_norm": 38.825893758239005, | |
| "learning_rate": 8.71882302665696e-06, | |
| "loss": 0.1639, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.623342175066313, | |
| "grad_norm": 51.493519726499024, | |
| "learning_rate": 8.70329664571461e-06, | |
| "loss": 0.375, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.6259946949602122, | |
| "grad_norm": 47.202719237816765, | |
| "learning_rate": 8.687690753613554e-06, | |
| "loss": 0.2737, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.6286472148541115, | |
| "grad_norm": 47.06710432775105, | |
| "learning_rate": 8.672005685418115e-06, | |
| "loss": 0.329, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.6312997347480106, | |
| "grad_norm": 36.186199865765076, | |
| "learning_rate": 8.656241777892544e-06, | |
| "loss": 0.1865, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.6339522546419099, | |
| "grad_norm": 82.68094414992525, | |
| "learning_rate": 8.640399369493813e-06, | |
| "loss": 0.2824, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.636604774535809, | |
| "grad_norm": 37.23457078121522, | |
| "learning_rate": 8.624478800364332e-06, | |
| "loss": 0.2146, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.6392572944297082, | |
| "grad_norm": 49.777881838826744, | |
| "learning_rate": 8.608480412324652e-06, | |
| "loss": 0.444, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.6419098143236074, | |
| "grad_norm": 48.32237627009601, | |
| "learning_rate": 8.592404548866123e-06, | |
| "loss": 0.2079, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.6445623342175066, | |
| "grad_norm": 37.924814101167854, | |
| "learning_rate": 8.576251555143524e-06, | |
| "loss": 0.1501, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.6472148541114059, | |
| "grad_norm": 35.79857920052079, | |
| "learning_rate": 8.56002177796765e-06, | |
| "loss": 0.131, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.649867374005305, | |
| "grad_norm": 62.52788609216863, | |
| "learning_rate": 8.543715565797861e-06, | |
| "loss": 0.4995, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.6525198938992043, | |
| "grad_norm": 53.50944198275586, | |
| "learning_rate": 8.527333268734607e-06, | |
| "loss": 0.3744, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.6551724137931034, | |
| "grad_norm": 47.32416611175991, | |
| "learning_rate": 8.510875238511911e-06, | |
| "loss": 0.152, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.6578249336870027, | |
| "grad_norm": 56.236921384204315, | |
| "learning_rate": 8.494341828489812e-06, | |
| "loss": 0.3054, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.6604774535809018, | |
| "grad_norm": 57.63962006677401, | |
| "learning_rate": 8.477733393646787e-06, | |
| "loss": 0.2399, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.6631299734748011, | |
| "grad_norm": 42.888481005602536, | |
| "learning_rate": 8.461050290572114e-06, | |
| "loss": 0.2081, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.6657824933687002, | |
| "grad_norm": 20.99857739960341, | |
| "learning_rate": 8.444292877458238e-06, | |
| "loss": 0.1291, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.6684350132625995, | |
| "grad_norm": 61.31650373045394, | |
| "learning_rate": 8.427461514093056e-06, | |
| "loss": 0.3973, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.6710875331564987, | |
| "grad_norm": 62.39650389044618, | |
| "learning_rate": 8.410556561852212e-06, | |
| "loss": 0.3041, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.6737400530503979, | |
| "grad_norm": 24.6664128060848, | |
| "learning_rate": 8.39357838369133e-06, | |
| "loss": 0.1809, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.6763925729442971, | |
| "grad_norm": 67.09141289555055, | |
| "learning_rate": 8.376527344138222e-06, | |
| "loss": 0.1814, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.6790450928381963, | |
| "grad_norm": 42.36398464526459, | |
| "learning_rate": 8.359403809285054e-06, | |
| "loss": 0.1685, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.6816976127320955, | |
| "grad_norm": 36.868570222869444, | |
| "learning_rate": 8.342208146780504e-06, | |
| "loss": 0.2083, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.6843501326259946, | |
| "grad_norm": 187.91006984894173, | |
| "learning_rate": 8.324940725821853e-06, | |
| "loss": 0.2198, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.6870026525198939, | |
| "grad_norm": 94.1194320864834, | |
| "learning_rate": 8.30760191714706e-06, | |
| "loss": 0.1839, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.6896551724137931, | |
| "grad_norm": 90.43733640770276, | |
| "learning_rate": 8.290192093026805e-06, | |
| "loss": 0.2862, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.6923076923076923, | |
| "grad_norm": 33.625780709523845, | |
| "learning_rate": 8.272711627256501e-06, | |
| "loss": 0.1615, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.6949602122015915, | |
| "grad_norm": 36.64648281855228, | |
| "learning_rate": 8.255160895148263e-06, | |
| "loss": 0.2367, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.6976127320954907, | |
| "grad_norm": 50.56944146133072, | |
| "learning_rate": 8.237540273522844e-06, | |
| "loss": 0.1622, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.7002652519893899, | |
| "grad_norm": 182.53585445600774, | |
| "learning_rate": 8.219850140701557e-06, | |
| "loss": 0.1438, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.7029177718832891, | |
| "grad_norm": 69.8729850923303, | |
| "learning_rate": 8.202090876498144e-06, | |
| "loss": 0.2296, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.7055702917771883, | |
| "grad_norm": 31.529852166286762, | |
| "learning_rate": 8.184262862210624e-06, | |
| "loss": 0.1816, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.7082228116710876, | |
| "grad_norm": 29.112616136097394, | |
| "learning_rate": 8.166366480613107e-06, | |
| "loss": 0.1042, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.7108753315649867, | |
| "grad_norm": 21.42596725620029, | |
| "learning_rate": 8.14840211594757e-06, | |
| "loss": 0.1284, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.713527851458886, | |
| "grad_norm": 43.121642674406175, | |
| "learning_rate": 8.13037015391562e-06, | |
| "loss": 0.2312, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.7161803713527851, | |
| "grad_norm": 45.89885836530061, | |
| "learning_rate": 8.112270981670196e-06, | |
| "loss": 0.157, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.7188328912466844, | |
| "grad_norm": 46.12217694940257, | |
| "learning_rate": 8.09410498780727e-06, | |
| "loss": 0.1332, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.7214854111405835, | |
| "grad_norm": 37.25833887900994, | |
| "learning_rate": 8.075872562357502e-06, | |
| "loss": 0.129, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.7241379310344828, | |
| "grad_norm": 33.182430046980066, | |
| "learning_rate": 8.057574096777854e-06, | |
| "loss": 0.157, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.726790450928382, | |
| "grad_norm": 25.452121277422545, | |
| "learning_rate": 8.039209983943201e-06, | |
| "loss": 0.0775, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.7294429708222812, | |
| "grad_norm": 141.08365924116302, | |
| "learning_rate": 8.020780618137889e-06, | |
| "loss": 0.3046, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.7320954907161804, | |
| "grad_norm": 32.48697951888386, | |
| "learning_rate": 8.002286395047267e-06, | |
| "loss": 0.0674, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.7347480106100795, | |
| "grad_norm": 133.0319126654085, | |
| "learning_rate": 7.983727711749194e-06, | |
| "loss": 0.2422, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.7374005305039788, | |
| "grad_norm": 61.60012063683119, | |
| "learning_rate": 7.965104966705518e-06, | |
| "loss": 0.1673, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.7400530503978779, | |
| "grad_norm": 60.7112974795619, | |
| "learning_rate": 7.946418559753509e-06, | |
| "loss": 0.2377, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.7427055702917772, | |
| "grad_norm": 49.872485049267205, | |
| "learning_rate": 7.927668892097288e-06, | |
| "loss": 0.1291, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.7453580901856764, | |
| "grad_norm": 84.27985493725895, | |
| "learning_rate": 7.908856366299206e-06, | |
| "loss": 0.2068, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.7480106100795756, | |
| "grad_norm": 89.07028975843645, | |
| "learning_rate": 7.889981386271202e-06, | |
| "loss": 0.3291, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.7506631299734748, | |
| "grad_norm": 33.01333445164221, | |
| "learning_rate": 7.871044357266124e-06, | |
| "loss": 0.1875, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.753315649867374, | |
| "grad_norm": 61.66655390306763, | |
| "learning_rate": 7.852045685869046e-06, | |
| "loss": 0.233, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.7559681697612732, | |
| "grad_norm": 45.392626938408974, | |
| "learning_rate": 7.832985779988518e-06, | |
| "loss": 0.0932, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.7586206896551724, | |
| "grad_norm": 62.82412358551919, | |
| "learning_rate": 7.81386504884782e-06, | |
| "loss": 0.2112, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.7612732095490716, | |
| "grad_norm": 77.77135839682533, | |
| "learning_rate": 7.794683902976175e-06, | |
| "loss": 0.1791, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.7639257294429708, | |
| "grad_norm": 50.290161550536745, | |
| "learning_rate": 7.775442754199929e-06, | |
| "loss": 0.1481, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.76657824933687, | |
| "grad_norm": 67.88391445622501, | |
| "learning_rate": 7.75614201563372e-06, | |
| "loss": 0.2346, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 93.00871825203926, | |
| "learning_rate": 7.736782101671587e-06, | |
| "loss": 0.1737, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.7718832891246684, | |
| "grad_norm": 53.71872191270983, | |
| "learning_rate": 7.717363427978103e-06, | |
| "loss": 0.2112, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.7745358090185677, | |
| "grad_norm": 32.55623505216373, | |
| "learning_rate": 7.697886411479422e-06, | |
| "loss": 0.1045, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.7771883289124668, | |
| "grad_norm": 119.02364134570698, | |
| "learning_rate": 7.67835147035435e-06, | |
| "loss": 0.1283, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.7798408488063661, | |
| "grad_norm": 82.88296400008042, | |
| "learning_rate": 7.658759024025349e-06, | |
| "loss": 0.2001, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.7824933687002652, | |
| "grad_norm": 87.85759054670997, | |
| "learning_rate": 7.639109493149537e-06, | |
| "loss": 0.2218, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.7851458885941645, | |
| "grad_norm": 135.43767274810497, | |
| "learning_rate": 7.6194032996096685e-06, | |
| "loss": 0.1864, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.7877984084880637, | |
| "grad_norm": 35.39816057230434, | |
| "learning_rate": 7.599640866505058e-06, | |
| "loss": 0.1078, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.7904509283819628, | |
| "grad_norm": 86.97826726357368, | |
| "learning_rate": 7.579822618142505e-06, | |
| "loss": 0.1618, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.7931034482758621, | |
| "grad_norm": 63.73431809970673, | |
| "learning_rate": 7.559948980027189e-06, | |
| "loss": 0.2698, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.7957559681697612, | |
| "grad_norm": 60.183076177452044, | |
| "learning_rate": 7.540020378853523e-06, | |
| "loss": 0.1594, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7984084880636605, | |
| "grad_norm": 52.5406608746334, | |
| "learning_rate": 7.520037242496e-06, | |
| "loss": 0.1836, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.8010610079575596, | |
| "grad_norm": 58.3353288564282, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.2385, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.8037135278514589, | |
| "grad_norm": 95.12475164244071, | |
| "learning_rate": 7.479909081572587e-06, | |
| "loss": 0.149, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.8063660477453581, | |
| "grad_norm": 46.159718954423134, | |
| "learning_rate": 7.459764918573264e-06, | |
| "loss": 0.1241, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.8090185676392573, | |
| "grad_norm": 585.8543926676177, | |
| "learning_rate": 7.4395679435047175e-06, | |
| "loss": 0.2069, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.8116710875331565, | |
| "grad_norm": 43.62270741644705, | |
| "learning_rate": 7.419318590003524e-06, | |
| "loss": 0.2192, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.8143236074270557, | |
| "grad_norm": 80.80866203584026, | |
| "learning_rate": 7.399017292830848e-06, | |
| "loss": 0.1693, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.8169761273209549, | |
| "grad_norm": 140.06215238615462, | |
| "learning_rate": 7.3786644878631035e-06, | |
| "loss": 0.1179, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.8196286472148541, | |
| "grad_norm": 78.61750140083358, | |
| "learning_rate": 7.358260612082596e-06, | |
| "loss": 0.2927, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.8222811671087533, | |
| "grad_norm": 70.13628004985249, | |
| "learning_rate": 7.3378061035681415e-06, | |
| "loss": 0.1519, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.8249336870026526, | |
| "grad_norm": 57.22683921472029, | |
| "learning_rate": 7.317301401485657e-06, | |
| "loss": 0.1996, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.8275862068965517, | |
| "grad_norm": 46.56616511235378, | |
| "learning_rate": 7.296746946078737e-06, | |
| "loss": 0.1493, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.830238726790451, | |
| "grad_norm": 52.16793854587364, | |
| "learning_rate": 7.276143178659195e-06, | |
| "loss": 0.1366, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.8328912466843501, | |
| "grad_norm": 56.1327455558407, | |
| "learning_rate": 7.255490541597594e-06, | |
| "loss": 0.1955, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.8355437665782494, | |
| "grad_norm": 64.04061033319763, | |
| "learning_rate": 7.2347894783137485e-06, | |
| "loss": 0.1223, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.8381962864721485, | |
| "grad_norm": 41.30534249766932, | |
| "learning_rate": 7.2140404332671986e-06, | |
| "loss": 0.1542, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.8408488063660478, | |
| "grad_norm": 59.39598856442543, | |
| "learning_rate": 7.19324385194767e-06, | |
| "loss": 0.2146, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.843501326259947, | |
| "grad_norm": 37.41911932589385, | |
| "learning_rate": 7.172400180865514e-06, | |
| "loss": 0.0893, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.8461538461538461, | |
| "grad_norm": 73.2170655681548, | |
| "learning_rate": 7.1515098675421125e-06, | |
| "loss": 0.1732, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.8488063660477454, | |
| "grad_norm": 54.90046866257738, | |
| "learning_rate": 7.130573360500277e-06, | |
| "loss": 0.1439, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.8514588859416445, | |
| "grad_norm": 57.04949794426898, | |
| "learning_rate": 7.109591109254614e-06, | |
| "loss": 0.2042, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.8541114058355438, | |
| "grad_norm": 63.62163449990602, | |
| "learning_rate": 7.088563564301874e-06, | |
| "loss": 0.3106, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.8567639257294429, | |
| "grad_norm": 47.16915214970692, | |
| "learning_rate": 7.067491177111282e-06, | |
| "loss": 0.2004, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.8594164456233422, | |
| "grad_norm": 47.16540297885484, | |
| "learning_rate": 7.046374400114842e-06, | |
| "loss": 0.1318, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.8620689655172413, | |
| "grad_norm": 31.645829514913345, | |
| "learning_rate": 7.0252136866976205e-06, | |
| "loss": 0.1516, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.8647214854111406, | |
| "grad_norm": 24.012429851214378, | |
| "learning_rate": 7.004009491188023e-06, | |
| "loss": 0.0749, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.8673740053050398, | |
| "grad_norm": 32.73778677034213, | |
| "learning_rate": 6.982762268848024e-06, | |
| "loss": 0.125, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.870026525198939, | |
| "grad_norm": 82.59316907110794, | |
| "learning_rate": 6.961472475863406e-06, | |
| "loss": 0.2883, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.8726790450928382, | |
| "grad_norm": 62.849464785868946, | |
| "learning_rate": 6.940140569333953e-06, | |
| "loss": 0.0979, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.8753315649867374, | |
| "grad_norm": 43.47928012947127, | |
| "learning_rate": 6.918767007263646e-06, | |
| "loss": 0.1342, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.8779840848806366, | |
| "grad_norm": 34.882979257908005, | |
| "learning_rate": 6.897352248550828e-06, | |
| "loss": 0.1035, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.8806366047745358, | |
| "grad_norm": 29.19456412156272, | |
| "learning_rate": 6.875896752978345e-06, | |
| "loss": 0.0811, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.883289124668435, | |
| "grad_norm": 35.59516597305405, | |
| "learning_rate": 6.85440098120368e-06, | |
| "loss": 0.0848, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.8859416445623343, | |
| "grad_norm": 36.16086076919905, | |
| "learning_rate": 6.832865394749065e-06, | |
| "loss": 0.091, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.8885941644562334, | |
| "grad_norm": 177.05756609050815, | |
| "learning_rate": 6.811290455991561e-06, | |
| "loss": 0.3371, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.8912466843501327, | |
| "grad_norm": 92.92983470592152, | |
| "learning_rate": 6.7896766281531435e-06, | |
| "loss": 0.2157, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.8938992042440318, | |
| "grad_norm": 63.197994960078006, | |
| "learning_rate": 6.768024375290747e-06, | |
| "loss": 0.1932, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.896551724137931, | |
| "grad_norm": 48.34862843223494, | |
| "learning_rate": 6.7463341622863074e-06, | |
| "loss": 0.0846, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.8992042440318302, | |
| "grad_norm": 55.66528827842737, | |
| "learning_rate": 6.724606454836782e-06, | |
| "loss": 0.0901, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.9018567639257294, | |
| "grad_norm": 57.60148358043267, | |
| "learning_rate": 6.702841719444141e-06, | |
| "loss": 0.1768, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.9045092838196287, | |
| "grad_norm": 74.31896937236691, | |
| "learning_rate": 6.681040423405363e-06, | |
| "loss": 0.1168, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.9071618037135278, | |
| "grad_norm": 47.82878168837927, | |
| "learning_rate": 6.659203034802397e-06, | |
| "loss": 0.1596, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.9098143236074271, | |
| "grad_norm": 93.24911135040223, | |
| "learning_rate": 6.637330022492112e-06, | |
| "loss": 0.1042, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.9124668435013262, | |
| "grad_norm": 35.793207624941154, | |
| "learning_rate": 6.615421856096231e-06, | |
| "loss": 0.0924, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.9151193633952255, | |
| "grad_norm": 29.532406452075197, | |
| "learning_rate": 6.593479005991251e-06, | |
| "loss": 0.0592, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.9177718832891246, | |
| "grad_norm": 94.46539610098456, | |
| "learning_rate": 6.571501943298335e-06, | |
| "loss": 0.1761, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.9204244031830239, | |
| "grad_norm": 90.85615975016468, | |
| "learning_rate": 6.549491139873211e-06, | |
| "loss": 0.2461, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 61.149450630924925, | |
| "learning_rate": 6.527447068296026e-06, | |
| "loss": 0.132, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.9257294429708223, | |
| "grad_norm": 53.31973958738185, | |
| "learning_rate": 6.50537020186121e-06, | |
| "loss": 0.0903, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.9283819628647215, | |
| "grad_norm": 61.6331760199786, | |
| "learning_rate": 6.483261014567311e-06, | |
| "loss": 0.1165, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.9310344827586207, | |
| "grad_norm": 13.522720380850254, | |
| "learning_rate": 6.4611199811068196e-06, | |
| "loss": 0.0518, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.9336870026525199, | |
| "grad_norm": 56.84483914980899, | |
| "learning_rate": 6.4389475768559675e-06, | |
| "loss": 0.096, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.9363395225464191, | |
| "grad_norm": 43.113140594612666, | |
| "learning_rate": 6.416744277864541e-06, | |
| "loss": 0.0585, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.9389920424403183, | |
| "grad_norm": 30.744059362187798, | |
| "learning_rate": 6.394510560845637e-06, | |
| "loss": 0.0431, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.9416445623342176, | |
| "grad_norm": 89.21785052474625, | |
| "learning_rate": 6.372246903165445e-06, | |
| "loss": 0.1349, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.9442970822281167, | |
| "grad_norm": 36.9671448724562, | |
| "learning_rate": 6.349953782832991e-06, | |
| "loss": 0.0571, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.946949602122016, | |
| "grad_norm": 69.48840780248835, | |
| "learning_rate": 6.327631678489874e-06, | |
| "loss": 0.1355, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.9496021220159151, | |
| "grad_norm": 60.4321035877009, | |
| "learning_rate": 6.305281069399989e-06, | |
| "loss": 0.0963, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.9522546419098143, | |
| "grad_norm": 40.04735809339083, | |
| "learning_rate": 6.282902435439242e-06, | |
| "loss": 0.0928, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.9549071618037135, | |
| "grad_norm": 54.37683366544728, | |
| "learning_rate": 6.26049625708524e-06, | |
| "loss": 0.1326, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.9575596816976127, | |
| "grad_norm": 99.91926297776382, | |
| "learning_rate": 6.238063015406982e-06, | |
| "loss": 0.139, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.9602122015915119, | |
| "grad_norm": 101.75866107017366, | |
| "learning_rate": 6.215603192054523e-06, | |
| "loss": 0.1668, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.9628647214854111, | |
| "grad_norm": 98.85344488282533, | |
| "learning_rate": 6.1931172692486405e-06, | |
| "loss": 0.2849, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.9655172413793104, | |
| "grad_norm": 165.94536763450807, | |
| "learning_rate": 6.17060572977047e-06, | |
| "loss": 0.134, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.9681697612732095, | |
| "grad_norm": 54.285346936222666, | |
| "learning_rate": 6.1480690569511545e-06, | |
| "loss": 0.0964, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.9708222811671088, | |
| "grad_norm": 55.157535754858486, | |
| "learning_rate": 6.125507734661458e-06, | |
| "loss": 0.1046, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.9734748010610079, | |
| "grad_norm": 96.63952887374089, | |
| "learning_rate": 6.1029222473013705e-06, | |
| "loss": 0.159, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.9761273209549072, | |
| "grad_norm": 85.19730687411784, | |
| "learning_rate": 6.080313079789723e-06, | |
| "loss": 0.3654, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.9787798408488063, | |
| "grad_norm": 150.05384389958417, | |
| "learning_rate": 6.0576807175537654e-06, | |
| "loss": 0.3451, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.9814323607427056, | |
| "grad_norm": 58.19141403492245, | |
| "learning_rate": 6.035025646518747e-06, | |
| "loss": 0.2298, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.9840848806366048, | |
| "grad_norm": 736.1395772866435, | |
| "learning_rate": 6.012348353097484e-06, | |
| "loss": 0.1161, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.986737400530504, | |
| "grad_norm": 124.97267593433126, | |
| "learning_rate": 5.9896493241799115e-06, | |
| "loss": 0.1798, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.9893899204244032, | |
| "grad_norm": 39.63404839325758, | |
| "learning_rate": 5.966929047122641e-06, | |
| "loss": 0.0431, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.9920424403183024, | |
| "grad_norm": 42.1987745432697, | |
| "learning_rate": 5.944188009738483e-06, | |
| "loss": 0.131, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.9946949602122016, | |
| "grad_norm": 46.74565476536693, | |
| "learning_rate": 5.921426700285986e-06, | |
| "loss": 0.1121, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.9973474801061007, | |
| "grad_norm": 27.180360013119735, | |
| "learning_rate": 5.898645607458941e-06, | |
| "loss": 0.0807, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 52.54379466042234, | |
| "learning_rate": 5.8758452203758995e-06, | |
| "loss": 0.11, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.3463860750198364, | |
| "eval_runtime": 175.4859, | |
| "eval_samples_per_second": 12.052, | |
| "eval_steps_per_second": 1.51, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 1.0026525198938991, | |
| "grad_norm": 17.68142839472272, | |
| "learning_rate": 5.8530260285696674e-06, | |
| "loss": 0.0529, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 1.0053050397877985, | |
| "grad_norm": 51.89104013831488, | |
| "learning_rate": 5.830188521976794e-06, | |
| "loss": 0.1815, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 1.0079575596816976, | |
| "grad_norm": 50.26750376877717, | |
| "learning_rate": 5.807333190927054e-06, | |
| "loss": 0.1903, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.0106100795755968, | |
| "grad_norm": 26.284561902513307, | |
| "learning_rate": 5.784460526132918e-06, | |
| "loss": 0.0937, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 1.013262599469496, | |
| "grad_norm": 29.45075692230731, | |
| "learning_rate": 5.761571018679025e-06, | |
| "loss": 0.0689, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 1.0159151193633953, | |
| "grad_norm": 44.374685492191965, | |
| "learning_rate": 5.738665160011627e-06, | |
| "loss": 0.1024, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 1.0185676392572944, | |
| "grad_norm": 110.37886147075102, | |
| "learning_rate": 5.715743441928041e-06, | |
| "loss": 0.1542, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 1.0212201591511936, | |
| "grad_norm": 55.13375919414892, | |
| "learning_rate": 5.6928063565660955e-06, | |
| "loss": 0.1021, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.023872679045093, | |
| "grad_norm": 61.58755412552258, | |
| "learning_rate": 5.669854396393559e-06, | |
| "loss": 0.1664, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 1.026525198938992, | |
| "grad_norm": 16.119940601375582, | |
| "learning_rate": 5.646888054197568e-06, | |
| "loss": 0.0401, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 1.0291777188328912, | |
| "grad_norm": 34.5316730073702, | |
| "learning_rate": 5.623907823074044e-06, | |
| "loss": 0.0592, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 1.0318302387267904, | |
| "grad_norm": 17.85636087705779, | |
| "learning_rate": 5.600914196417112e-06, | |
| "loss": 0.0356, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 1.0344827586206897, | |
| "grad_norm": 82.08938241465589, | |
| "learning_rate": 5.577907667908505e-06, | |
| "loss": 0.318, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.0371352785145889, | |
| "grad_norm": 10.735136547662583, | |
| "learning_rate": 5.5548887315069575e-06, | |
| "loss": 0.0349, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 1.039787798408488, | |
| "grad_norm": 29.23400812860403, | |
| "learning_rate": 5.531857881437612e-06, | |
| "loss": 0.163, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 1.0424403183023874, | |
| "grad_norm": 9.144587638643738, | |
| "learning_rate": 5.508815612181401e-06, | |
| "loss": 0.0328, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 1.0450928381962865, | |
| "grad_norm": 91.3897853620049, | |
| "learning_rate": 5.48576241846443e-06, | |
| "loss": 0.1606, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 1.0477453580901857, | |
| "grad_norm": 64.31305092536942, | |
| "learning_rate": 5.462698795247357e-06, | |
| "loss": 0.2104, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.0503978779840848, | |
| "grad_norm": 20.209916148778973, | |
| "learning_rate": 5.4396252377147615e-06, | |
| "loss": 0.0633, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 1.0530503978779842, | |
| "grad_norm": 46.54402940059323, | |
| "learning_rate": 5.416542241264524e-06, | |
| "loss": 0.1654, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 1.0557029177718833, | |
| "grad_norm": 33.01089496895264, | |
| "learning_rate": 5.39345030149718e-06, | |
| "loss": 0.0621, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 1.0583554376657824, | |
| "grad_norm": 39.99501520680759, | |
| "learning_rate": 5.370349914205273e-06, | |
| "loss": 0.0498, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 1.0610079575596818, | |
| "grad_norm": 18.118774769065993, | |
| "learning_rate": 5.347241575362729e-06, | |
| "loss": 0.0299, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.063660477453581, | |
| "grad_norm": 62.17306334991743, | |
| "learning_rate": 5.324125781114193e-06, | |
| "loss": 0.1847, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 1.06631299734748, | |
| "grad_norm": 56.97150682105584, | |
| "learning_rate": 5.30100302776438e-06, | |
| "loss": 0.1063, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 1.0689655172413792, | |
| "grad_norm": 6.0117370575361475, | |
| "learning_rate": 5.277873811767415e-06, | |
| "loss": 0.0236, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 1.0716180371352786, | |
| "grad_norm": 60.033520469100175, | |
| "learning_rate": 5.254738629716186e-06, | |
| "loss": 0.2285, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 1.0742705570291777, | |
| "grad_norm": 63.69961951769621, | |
| "learning_rate": 5.231597978331669e-06, | |
| "loss": 0.0603, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.0769230769230769, | |
| "grad_norm": 143.5016156103116, | |
| "learning_rate": 5.208452354452275e-06, | |
| "loss": 0.2138, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 1.079575596816976, | |
| "grad_norm": 59.54079006680123, | |
| "learning_rate": 5.185302255023166e-06, | |
| "loss": 0.1195, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 1.0822281167108754, | |
| "grad_norm": 74.24827747223898, | |
| "learning_rate": 5.162148177085604e-06, | |
| "loss": 0.0795, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 1.0848806366047745, | |
| "grad_norm": 35.752176310460186, | |
| "learning_rate": 5.1389906177662705e-06, | |
| "loss": 0.1055, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 1.0875331564986737, | |
| "grad_norm": 44.45779507253205, | |
| "learning_rate": 5.115830074266592e-06, | |
| "loss": 0.0765, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.090185676392573, | |
| "grad_norm": 18.742694564415604, | |
| "learning_rate": 5.092667043852062e-06, | |
| "loss": 0.0398, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 1.0928381962864722, | |
| "grad_norm": 57.0358614235308, | |
| "learning_rate": 5.069502023841576e-06, | |
| "loss": 0.0715, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 1.0954907161803713, | |
| "grad_norm": 60.239344515637356, | |
| "learning_rate": 5.046335511596746e-06, | |
| "loss": 0.0848, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 1.0981432360742707, | |
| "grad_norm": 16.893018773823467, | |
| "learning_rate": 5.0231680045112174e-06, | |
| "loss": 0.0262, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 1.1007957559681698, | |
| "grad_norm": 22.200535604754553, | |
| "learning_rate": 5e-06, | |
| "loss": 0.0323, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.103448275862069, | |
| "grad_norm": 12.087786574422244, | |
| "learning_rate": 4.976831995488784e-06, | |
| "loss": 0.0322, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 1.106100795755968, | |
| "grad_norm": 3.4135212180684533, | |
| "learning_rate": 4.953664488403256e-06, | |
| "loss": 0.0224, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 1.1087533156498675, | |
| "grad_norm": 73.5459113588346, | |
| "learning_rate": 4.9304979761584256e-06, | |
| "loss": 0.1572, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 1.1114058355437666, | |
| "grad_norm": 119.11937925866647, | |
| "learning_rate": 4.90733295614794e-06, | |
| "loss": 0.074, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 1.1140583554376657, | |
| "grad_norm": 17.00988830972538, | |
| "learning_rate": 4.884169925733409e-06, | |
| "loss": 0.0396, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.1167108753315649, | |
| "grad_norm": 63.38848105296801, | |
| "learning_rate": 4.86100938223373e-06, | |
| "loss": 0.2442, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 1.1193633952254642, | |
| "grad_norm": 1.858090026430698, | |
| "learning_rate": 4.837851822914397e-06, | |
| "loss": 0.0173, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 1.1220159151193634, | |
| "grad_norm": 30.20171759259156, | |
| "learning_rate": 4.814697744976835e-06, | |
| "loss": 0.0414, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 1.1246684350132625, | |
| "grad_norm": 77.7716831140678, | |
| "learning_rate": 4.791547645547727e-06, | |
| "loss": 0.1056, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 1.1273209549071619, | |
| "grad_norm": 99.332034153341, | |
| "learning_rate": 4.768402021668332e-06, | |
| "loss": 0.1221, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.129973474801061, | |
| "grad_norm": 121.44507801917504, | |
| "learning_rate": 4.7452613702838166e-06, | |
| "loss": 0.1468, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 1.1326259946949602, | |
| "grad_norm": 58.54743022339587, | |
| "learning_rate": 4.722126188232586e-06, | |
| "loss": 0.0665, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 1.1352785145888595, | |
| "grad_norm": 32.52455340574876, | |
| "learning_rate": 4.698996972235622e-06, | |
| "loss": 0.0568, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 1.1379310344827587, | |
| "grad_norm": 5.838433990840112, | |
| "learning_rate": 4.6758742188858074e-06, | |
| "loss": 0.0192, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 1.1405835543766578, | |
| "grad_norm": 71.81629603584015, | |
| "learning_rate": 4.652758424637271e-06, | |
| "loss": 0.1831, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.143236074270557, | |
| "grad_norm": 28.144271443634484, | |
| "learning_rate": 4.629650085794728e-06, | |
| "loss": 0.0433, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 1.1458885941644563, | |
| "grad_norm": 10.923481408273847, | |
| "learning_rate": 4.606549698502824e-06, | |
| "loss": 0.0204, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 1.1485411140583555, | |
| "grad_norm": 35.06820640518518, | |
| "learning_rate": 4.583457758735477e-06, | |
| "loss": 0.031, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 1.1511936339522546, | |
| "grad_norm": 55.77005306913307, | |
| "learning_rate": 4.56037476228524e-06, | |
| "loss": 0.0803, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 1.1538461538461537, | |
| "grad_norm": 75.59003795948266, | |
| "learning_rate": 4.537301204752647e-06, | |
| "loss": 0.0309, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.156498673740053, | |
| "grad_norm": 1.3813703392066852, | |
| "learning_rate": 4.514237581535571e-06, | |
| "loss": 0.0151, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 1.1591511936339522, | |
| "grad_norm": 15.654231932623063, | |
| "learning_rate": 4.4911843878186e-06, | |
| "loss": 0.0232, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 1.1618037135278514, | |
| "grad_norm": 1.5271112385991394, | |
| "learning_rate": 4.468142118562389e-06, | |
| "loss": 0.0142, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 1.1644562334217508, | |
| "grad_norm": 41.808618529434774, | |
| "learning_rate": 4.4451112684930424e-06, | |
| "loss": 0.0365, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 1.16710875331565, | |
| "grad_norm": 8.875571871982734, | |
| "learning_rate": 4.422092332091497e-06, | |
| "loss": 0.0174, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.169761273209549, | |
| "grad_norm": 42.48178681496721, | |
| "learning_rate": 4.399085803582889e-06, | |
| "loss": 0.0619, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 1.1724137931034484, | |
| "grad_norm": 60.49769628624578, | |
| "learning_rate": 4.3760921769259585e-06, | |
| "loss": 0.0729, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 1.1750663129973475, | |
| "grad_norm": 26.574412523619465, | |
| "learning_rate": 4.353111945802433e-06, | |
| "loss": 0.0338, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 1.1777188328912467, | |
| "grad_norm": 3.355645813229191, | |
| "learning_rate": 4.3301456036064415e-06, | |
| "loss": 0.0146, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 1.1803713527851458, | |
| "grad_norm": 3.0080563293578133, | |
| "learning_rate": 4.307193643433907e-06, | |
| "loss": 0.0146, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.1830238726790452, | |
| "grad_norm": 63.60306018611037, | |
| "learning_rate": 4.28425655807196e-06, | |
| "loss": 0.0911, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 1.1856763925729443, | |
| "grad_norm": 89.9524066237151, | |
| "learning_rate": 4.261334839988375e-06, | |
| "loss": 0.0838, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 1.1883289124668435, | |
| "grad_norm": 33.0673426347805, | |
| "learning_rate": 4.2384289813209754e-06, | |
| "loss": 0.0254, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 1.1909814323607426, | |
| "grad_norm": 47.18288151519677, | |
| "learning_rate": 4.2155394738670814e-06, | |
| "loss": 0.0601, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 1.193633952254642, | |
| "grad_norm": 123.89391604420489, | |
| "learning_rate": 4.192666809072948e-06, | |
| "loss": 0.1175, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.1962864721485411, | |
| "grad_norm": 1.0021352013229665, | |
| "learning_rate": 4.1698114780232085e-06, | |
| "loss": 0.0121, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 1.1989389920424403, | |
| "grad_norm": 10.214692929672639, | |
| "learning_rate": 4.146973971430333e-06, | |
| "loss": 0.0181, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 1.2015915119363396, | |
| "grad_norm": 15.04816767414729, | |
| "learning_rate": 4.124154779624101e-06, | |
| "loss": 0.018, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 1.2042440318302388, | |
| "grad_norm": 132.58538119418554, | |
| "learning_rate": 4.101354392541061e-06, | |
| "loss": 0.0335, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 1.206896551724138, | |
| "grad_norm": 420.64703357856126, | |
| "learning_rate": 4.078573299714014e-06, | |
| "loss": 0.0501, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.209549071618037, | |
| "grad_norm": 280.33971081551715, | |
| "learning_rate": 4.055811990261518e-06, | |
| "loss": 0.2018, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 1.2122015915119364, | |
| "grad_norm": 88.97956230266567, | |
| "learning_rate": 4.033070952877362e-06, | |
| "loss": 0.0857, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 1.2148541114058355, | |
| "grad_norm": 163.09979246949874, | |
| "learning_rate": 4.010350675820091e-06, | |
| "loss": 0.0972, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 1.2175066312997347, | |
| "grad_norm": 42.15910822649088, | |
| "learning_rate": 3.987651646902518e-06, | |
| "loss": 0.0287, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 1.2201591511936338, | |
| "grad_norm": 64.1429638803959, | |
| "learning_rate": 3.964974353481254e-06, | |
| "loss": 0.1458, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.2228116710875332, | |
| "grad_norm": 30.51792590892679, | |
| "learning_rate": 3.942319282446236e-06, | |
| "loss": 0.024, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 1.2254641909814323, | |
| "grad_norm": 181.99918301562113, | |
| "learning_rate": 3.9196869202102775e-06, | |
| "loss": 0.098, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 1.2281167108753315, | |
| "grad_norm": 195.65169748487375, | |
| "learning_rate": 3.89707775269863e-06, | |
| "loss": 0.3297, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 1.2307692307692308, | |
| "grad_norm": 145.26469256015454, | |
| "learning_rate": 3.874492265338544e-06, | |
| "loss": 0.3517, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 1.23342175066313, | |
| "grad_norm": 35.48516271276079, | |
| "learning_rate": 3.851930943048845e-06, | |
| "loss": 0.0478, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.2360742705570291, | |
| "grad_norm": 65.85506728640877, | |
| "learning_rate": 3.829394270229531e-06, | |
| "loss": 0.0828, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 1.2387267904509285, | |
| "grad_norm": 0.6539317465235754, | |
| "learning_rate": 3.8068827307513624e-06, | |
| "loss": 0.0108, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 1.2413793103448276, | |
| "grad_norm": 3.9741081031290233, | |
| "learning_rate": 3.7843968079454773e-06, | |
| "loss": 0.0114, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 1.2440318302387268, | |
| "grad_norm": 53.32466974048667, | |
| "learning_rate": 3.7619369845930195e-06, | |
| "loss": 0.0723, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 1.246684350132626, | |
| "grad_norm": 90.17775858739552, | |
| "learning_rate": 3.7395037429147615e-06, | |
| "loss": 0.1513, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.2493368700265253, | |
| "grad_norm": 160.86802203362117, | |
| "learning_rate": 3.7170975645607587e-06, | |
| "loss": 0.0339, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 1.2519893899204244, | |
| "grad_norm": 5.200593648400127, | |
| "learning_rate": 3.694718930600012e-06, | |
| "loss": 0.0125, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 1.2546419098143236, | |
| "grad_norm": 117.79945671801583, | |
| "learning_rate": 3.672368321510128e-06, | |
| "loss": 0.2488, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 1.2572944297082227, | |
| "grad_norm": 165.35263196181958, | |
| "learning_rate": 3.6500462171670104e-06, | |
| "loss": 0.1481, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 1.259946949602122, | |
| "grad_norm": 129.4190841061859, | |
| "learning_rate": 3.6277530968345552e-06, | |
| "loss": 0.0589, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.2625994694960212, | |
| "grad_norm": 55.628019444318824, | |
| "learning_rate": 3.605489439154365e-06, | |
| "loss": 0.0454, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 1.2652519893899203, | |
| "grad_norm": 93.59588719854446, | |
| "learning_rate": 3.583255722135462e-06, | |
| "loss": 0.1275, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 1.2679045092838197, | |
| "grad_norm": 10.07453203891493, | |
| "learning_rate": 3.5610524231440324e-06, | |
| "loss": 0.0153, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 1.2705570291777188, | |
| "grad_norm": 26.44208504298303, | |
| "learning_rate": 3.5388800188931825e-06, | |
| "loss": 0.034, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 1.273209549071618, | |
| "grad_norm": 43.17905002513331, | |
| "learning_rate": 3.5167389854326907e-06, | |
| "loss": 0.0589, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.2758620689655173, | |
| "grad_norm": 86.23570826911956, | |
| "learning_rate": 3.4946297981387913e-06, | |
| "loss": 0.0662, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 1.2785145888594165, | |
| "grad_norm": 15.699132266261188, | |
| "learning_rate": 3.472552931703975e-06, | |
| "loss": 0.0244, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 1.2811671087533156, | |
| "grad_norm": 83.37866584966203, | |
| "learning_rate": 3.4505088601267913e-06, | |
| "loss": 0.0561, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 1.2838196286472148, | |
| "grad_norm": 7.371340772061292, | |
| "learning_rate": 3.428498056701665e-06, | |
| "loss": 0.0157, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 1.2864721485411141, | |
| "grad_norm": 72.03330352703713, | |
| "learning_rate": 3.4065209940087507e-06, | |
| "loss": 0.0528, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.2891246684350133, | |
| "grad_norm": 60.07439706689129, | |
| "learning_rate": 3.3845781439037695e-06, | |
| "loss": 0.1746, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 1.2917771883289124, | |
| "grad_norm": 15.830602220363893, | |
| "learning_rate": 3.3626699775078884e-06, | |
| "loss": 0.0183, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 1.2944297082228116, | |
| "grad_norm": 90.63531502713072, | |
| "learning_rate": 3.3407969651976045e-06, | |
| "loss": 0.047, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 1.297082228116711, | |
| "grad_norm": 74.83398542874417, | |
| "learning_rate": 3.3189595765946394e-06, | |
| "loss": 0.1173, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 1.29973474801061, | |
| "grad_norm": 9.886317265655553, | |
| "learning_rate": 3.2971582805558622e-06, | |
| "loss": 0.0155, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.3023872679045092, | |
| "grad_norm": 113.17084370887686, | |
| "learning_rate": 3.27539354516322e-06, | |
| "loss": 0.0786, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 1.3050397877984086, | |
| "grad_norm": 42.35804473126241, | |
| "learning_rate": 3.253665837713694e-06, | |
| "loss": 0.0489, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 1.3076923076923077, | |
| "grad_norm": 32.04559934238327, | |
| "learning_rate": 3.2319756247092552e-06, | |
| "loss": 0.0737, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 1.3103448275862069, | |
| "grad_norm": 22.395640501451613, | |
| "learning_rate": 3.2103233718468574e-06, | |
| "loss": 0.0298, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 1.3129973474801062, | |
| "grad_norm": 95.51322529061028, | |
| "learning_rate": 3.1887095440084402e-06, | |
| "loss": 0.0758, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.3156498673740054, | |
| "grad_norm": 28.65751993200879, | |
| "learning_rate": 3.167134605250938e-06, | |
| "loss": 0.0442, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 1.3183023872679045, | |
| "grad_norm": 78.32058560501815, | |
| "learning_rate": 3.14559901879632e-06, | |
| "loss": 0.1396, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 1.3209549071618036, | |
| "grad_norm": 55.49433188217796, | |
| "learning_rate": 3.1241032470216564e-06, | |
| "loss": 0.1229, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 1.323607427055703, | |
| "grad_norm": 23.12635143909747, | |
| "learning_rate": 3.102647751449174e-06, | |
| "loss": 0.0258, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 1.3262599469496021, | |
| "grad_norm": 28.230637007866875, | |
| "learning_rate": 3.081232992736355e-06, | |
| "loss": 0.037, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.3289124668435013, | |
| "grad_norm": 7.715649744709754, | |
| "learning_rate": 3.059859430666049e-06, | |
| "loss": 0.0146, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 1.3315649867374004, | |
| "grad_norm": 25.891769502451563, | |
| "learning_rate": 3.0385275241365965e-06, | |
| "loss": 0.02, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 1.3342175066312998, | |
| "grad_norm": 62.79349231746409, | |
| "learning_rate": 3.017237731151976e-06, | |
| "loss": 0.1742, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 1.336870026525199, | |
| "grad_norm": 30.499199180675188, | |
| "learning_rate": 2.9959905088119777e-06, | |
| "loss": 0.0302, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 1.339522546419098, | |
| "grad_norm": 73.98055141461457, | |
| "learning_rate": 2.9747863133023803e-06, | |
| "loss": 0.0736, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.3421750663129974, | |
| "grad_norm": 39.29927119144451, | |
| "learning_rate": 2.9536255998851615e-06, | |
| "loss": 0.0237, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 1.3448275862068966, | |
| "grad_norm": 98.33679695718767, | |
| "learning_rate": 2.93250882288872e-06, | |
| "loss": 0.0387, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 1.3474801061007957, | |
| "grad_norm": 35.11947851166149, | |
| "learning_rate": 2.9114364356981274e-06, | |
| "loss": 0.1034, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 1.350132625994695, | |
| "grad_norm": 88.45903761328185, | |
| "learning_rate": 2.8904088907453887e-06, | |
| "loss": 0.1778, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 1.3527851458885942, | |
| "grad_norm": 8.532979695014493, | |
| "learning_rate": 2.8694266394997238e-06, | |
| "loss": 0.0149, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.3554376657824934, | |
| "grad_norm": 6.249816237214853, | |
| "learning_rate": 2.8484901324578883e-06, | |
| "loss": 0.0133, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 1.3580901856763925, | |
| "grad_norm": 22.97585624735204, | |
| "learning_rate": 2.827599819134489e-06, | |
| "loss": 0.0341, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 1.3607427055702916, | |
| "grad_norm": 36.53367358492221, | |
| "learning_rate": 2.8067561480523315e-06, | |
| "loss": 0.0412, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 1.363395225464191, | |
| "grad_norm": 72.92777350676762, | |
| "learning_rate": 2.7859595667328027e-06, | |
| "loss": 0.0346, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 1.3660477453580901, | |
| "grad_norm": 4.019754908273797, | |
| "learning_rate": 2.7652105216862536e-06, | |
| "loss": 0.0115, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.3687002652519893, | |
| "grad_norm": 26.804417375354443, | |
| "learning_rate": 2.7445094584024067e-06, | |
| "loss": 0.0298, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 1.3713527851458887, | |
| "grad_norm": 45.990202357286485, | |
| "learning_rate": 2.723856821340806e-06, | |
| "loss": 0.0377, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 1.3740053050397878, | |
| "grad_norm": 37.03560383879199, | |
| "learning_rate": 2.703253053921266e-06, | |
| "loss": 0.0401, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 1.376657824933687, | |
| "grad_norm": 132.83236380385904, | |
| "learning_rate": 2.682698598514343e-06, | |
| "loss": 0.0593, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 1.3793103448275863, | |
| "grad_norm": 81.33023653955709, | |
| "learning_rate": 2.6621938964318593e-06, | |
| "loss": 0.0678, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.3819628647214854, | |
| "grad_norm": 19.815072857885436, | |
| "learning_rate": 2.6417393879174056e-06, | |
| "loss": 0.0162, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 1.3846153846153846, | |
| "grad_norm": 80.27696589823212, | |
| "learning_rate": 2.621335512136899e-06, | |
| "loss": 0.0956, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 1.387267904509284, | |
| "grad_norm": 75.6641316305186, | |
| "learning_rate": 2.600982707169154e-06, | |
| "loss": 0.0504, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 1.389920424403183, | |
| "grad_norm": 6.8828160616189855, | |
| "learning_rate": 2.580681409996477e-06, | |
| "loss": 0.0117, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 1.3925729442970822, | |
| "grad_norm": 79.52956610275369, | |
| "learning_rate": 2.5604320564952846e-06, | |
| "loss": 0.0378, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.3952254641909814, | |
| "grad_norm": 3.3412476079565105, | |
| "learning_rate": 2.5402350814267364e-06, | |
| "loss": 0.0116, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 1.3978779840848805, | |
| "grad_norm": 74.0836189917631, | |
| "learning_rate": 2.5200909184274125e-06, | |
| "loss": 0.1159, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 1.4005305039787799, | |
| "grad_norm": 52.65976926514463, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.0575, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 1.403183023872679, | |
| "grad_norm": 66.03843598072747, | |
| "learning_rate": 2.4799627575040014e-06, | |
| "loss": 0.0445, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 1.4058355437665782, | |
| "grad_norm": 2.2027472519213678, | |
| "learning_rate": 2.4599796211464772e-06, | |
| "loss": 0.0102, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.4084880636604775, | |
| "grad_norm": 44.9196517675172, | |
| "learning_rate": 2.4400510199728123e-06, | |
| "loss": 0.0195, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 1.4111405835543767, | |
| "grad_norm": 91.76177563325083, | |
| "learning_rate": 2.4201773818574956e-06, | |
| "loss": 0.0151, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 1.4137931034482758, | |
| "grad_norm": 23.09555777900529, | |
| "learning_rate": 2.400359133494944e-06, | |
| "loss": 0.0191, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 1.4164456233421752, | |
| "grad_norm": 118.53313959667297, | |
| "learning_rate": 2.3805967003903336e-06, | |
| "loss": 0.0666, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 1.4190981432360743, | |
| "grad_norm": 124.60220963221789, | |
| "learning_rate": 2.360890506850464e-06, | |
| "loss": 0.0947, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.4217506631299734, | |
| "grad_norm": 108.48487599136776, | |
| "learning_rate": 2.341240975974653e-06, | |
| "loss": 0.0334, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 1.4244031830238728, | |
| "grad_norm": 29.258330324047396, | |
| "learning_rate": 2.3216485296456514e-06, | |
| "loss": 0.0167, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 1.427055702917772, | |
| "grad_norm": 44.99491729254627, | |
| "learning_rate": 2.302113588520578e-06, | |
| "loss": 0.0196, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 1.429708222811671, | |
| "grad_norm": 194.73710142416067, | |
| "learning_rate": 2.2826365720218984e-06, | |
| "loss": 0.0788, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 1.4323607427055702, | |
| "grad_norm": 58.92853113805291, | |
| "learning_rate": 2.263217898328415e-06, | |
| "loss": 0.024, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.4350132625994694, | |
| "grad_norm": 131.34154138736488, | |
| "learning_rate": 2.243857984366284e-06, | |
| "loss": 0.0557, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 1.4376657824933687, | |
| "grad_norm": 22.727392713957276, | |
| "learning_rate": 2.2245572458000714e-06, | |
| "loss": 0.0102, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 1.4403183023872679, | |
| "grad_norm": 6.179370744669459, | |
| "learning_rate": 2.205316097023826e-06, | |
| "loss": 0.0123, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 1.442970822281167, | |
| "grad_norm": 5.1332994713955244, | |
| "learning_rate": 2.1861349511521817e-06, | |
| "loss": 0.0111, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 1.4456233421750664, | |
| "grad_norm": 39.86131150680043, | |
| "learning_rate": 2.1670142200114837e-06, | |
| "loss": 0.0329, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.4482758620689655, | |
| "grad_norm": 10.043074941303454, | |
| "learning_rate": 2.147954314130955e-06, | |
| "loss": 0.0127, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 1.4509283819628647, | |
| "grad_norm": 33.75209517992541, | |
| "learning_rate": 2.128955642733877e-06, | |
| "loss": 0.0108, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 1.453580901856764, | |
| "grad_norm": 4.281022797107646, | |
| "learning_rate": 2.1100186137288005e-06, | |
| "loss": 0.0094, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 1.4562334217506632, | |
| "grad_norm": 47.72341676502409, | |
| "learning_rate": 2.0911436337007935e-06, | |
| "loss": 0.0217, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 1.4588859416445623, | |
| "grad_norm": 58.00881611775122, | |
| "learning_rate": 2.072331107902713e-06, | |
| "loss": 0.048, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.4615384615384617, | |
| "grad_norm": 0.6359561449436009, | |
| "learning_rate": 2.0535814402464922e-06, | |
| "loss": 0.009, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 1.4641909814323608, | |
| "grad_norm": 3.323348183961715, | |
| "learning_rate": 2.034895033294483e-06, | |
| "loss": 0.0099, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 1.46684350132626, | |
| "grad_norm": 31.33520595629068, | |
| "learning_rate": 2.0162722882508072e-06, | |
| "loss": 0.0104, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 1.469496021220159, | |
| "grad_norm": 0.5675415392727176, | |
| "learning_rate": 1.9977136049527348e-06, | |
| "loss": 0.0088, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 1.4721485411140582, | |
| "grad_norm": 0.5205511687748441, | |
| "learning_rate": 1.9792193818621118e-06, | |
| "loss": 0.0085, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.4748010610079576, | |
| "grad_norm": 0.46492420932481743, | |
| "learning_rate": 1.960790016056801e-06, | |
| "loss": 0.0083, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 1.4774535809018567, | |
| "grad_norm": 0.6103064433138241, | |
| "learning_rate": 1.9424259032221482e-06, | |
| "loss": 0.0083, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 1.4801061007957559, | |
| "grad_norm": 101.84912553346997, | |
| "learning_rate": 1.9241274376425e-06, | |
| "loss": 0.0504, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 1.4827586206896552, | |
| "grad_norm": 0.5036290681501117, | |
| "learning_rate": 1.90589501219273e-06, | |
| "loss": 0.0082, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 1.4854111405835544, | |
| "grad_norm": 0.49630303098429807, | |
| "learning_rate": 1.8877290183298058e-06, | |
| "loss": 0.008, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.4880636604774535, | |
| "grad_norm": 1.7511111350237467, | |
| "learning_rate": 1.869629846084382e-06, | |
| "loss": 0.0084, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 1.490716180371353, | |
| "grad_norm": 145.93290092744815, | |
| "learning_rate": 1.8515978840524302e-06, | |
| "loss": 0.0746, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 1.493368700265252, | |
| "grad_norm": 2.600079267900106, | |
| "learning_rate": 1.8336335193868955e-06, | |
| "loss": 0.0086, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 1.4960212201591512, | |
| "grad_norm": 0.750605001982341, | |
| "learning_rate": 1.8157371377893769e-06, | |
| "loss": 0.008, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 1.4986737400530503, | |
| "grad_norm": 3.3186117605501178, | |
| "learning_rate": 1.7979091235018564e-06, | |
| "loss": 0.009, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.5013262599469495, | |
| "grad_norm": 0.43220741408065144, | |
| "learning_rate": 1.7801498592984445e-06, | |
| "loss": 0.0075, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 1.5039787798408488, | |
| "grad_norm": 0.498889987966915, | |
| "learning_rate": 1.762459726477157e-06, | |
| "loss": 0.0074, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 1.506631299734748, | |
| "grad_norm": 0.4530320806531492, | |
| "learning_rate": 1.7448391048517378e-06, | |
| "loss": 0.0074, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 1.509283819628647, | |
| "grad_norm": 0.4031770086436206, | |
| "learning_rate": 1.7272883727434996e-06, | |
| "loss": 0.0072, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 1.5119363395225465, | |
| "grad_norm": 0.42231287937693796, | |
| "learning_rate": 1.709807906973196e-06, | |
| "loss": 0.0072, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.5145888594164456, | |
| "grad_norm": 2.5024275258180335, | |
| "learning_rate": 1.6923980828529424e-06, | |
| "loss": 0.0075, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 1.5172413793103448, | |
| "grad_norm": 84.0292035953675, | |
| "learning_rate": 1.6750592741781496e-06, | |
| "loss": 0.0856, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 1.5198938992042441, | |
| "grad_norm": 0.5403791435068346, | |
| "learning_rate": 1.657791853219497e-06, | |
| "loss": 0.007, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 1.5225464190981433, | |
| "grad_norm": 0.4226395502128115, | |
| "learning_rate": 1.640596190714947e-06, | |
| "loss": 0.0069, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 1.5251989389920424, | |
| "grad_norm": 0.3768980080326811, | |
| "learning_rate": 1.623472655861782e-06, | |
| "loss": 0.0067, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.5278514588859418, | |
| "grad_norm": 0.3755495508872125, | |
| "learning_rate": 1.6064216163086716e-06, | |
| "loss": 0.0067, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 1.530503978779841, | |
| "grad_norm": 0.49910529427933337, | |
| "learning_rate": 1.589443438147789e-06, | |
| "loss": 0.0068, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 1.53315649867374, | |
| "grad_norm": 90.27478366870945, | |
| "learning_rate": 1.5725384859069454e-06, | |
| "loss": 0.0604, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 1.5358090185676394, | |
| "grad_norm": 0.3589735178642482, | |
| "learning_rate": 1.5557071225417648e-06, | |
| "loss": 0.0064, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 0.3576882586646404, | |
| "learning_rate": 1.5389497094278861e-06, | |
| "loss": 0.0063, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.5411140583554377, | |
| "grad_norm": 0.35892809435755246, | |
| "learning_rate": 1.5222666063532138e-06, | |
| "loss": 0.0063, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 1.5437665782493368, | |
| "grad_norm": 0.3976724392685113, | |
| "learning_rate": 1.5056581715101887e-06, | |
| "loss": 0.0063, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 1.546419098143236, | |
| "grad_norm": 0.5400185192537355, | |
| "learning_rate": 1.4891247614880904e-06, | |
| "loss": 0.0062, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 1.5490716180371353, | |
| "grad_norm": 2.049521701650788, | |
| "learning_rate": 1.472666731265394e-06, | |
| "loss": 0.0066, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 1.5517241379310345, | |
| "grad_norm": 0.3468430762292899, | |
| "learning_rate": 1.456284434202142e-06, | |
| "loss": 0.0061, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.5543766578249336, | |
| "grad_norm": 0.3467327054472945, | |
| "learning_rate": 1.4399782220323515e-06, | |
| "loss": 0.006, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 1.557029177718833, | |
| "grad_norm": 0.35228902598430756, | |
| "learning_rate": 1.4237484448564759e-06, | |
| "loss": 0.006, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 1.5596816976127321, | |
| "grad_norm": 2.6000097356754077, | |
| "learning_rate": 1.4075954511338784e-06, | |
| "loss": 0.0065, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 1.5623342175066313, | |
| "grad_norm": 1.237130912335302, | |
| "learning_rate": 1.3915195876753495e-06, | |
| "loss": 0.006, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 1.5649867374005306, | |
| "grad_norm": 3.584447291257898, | |
| "learning_rate": 1.3755211996356687e-06, | |
| "loss": 0.0058, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.5676392572944295, | |
| "grad_norm": 9.918392784619872, | |
| "learning_rate": 1.3596006305061888e-06, | |
| "loss": 0.0093, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 1.570291777188329, | |
| "grad_norm": 6.146405804061137, | |
| "learning_rate": 1.3437582221074574e-06, | |
| "loss": 0.0063, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 1.5729442970822283, | |
| "grad_norm": 0.3182841602870679, | |
| "learning_rate": 1.3279943145818874e-06, | |
| "loss": 0.0057, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 1.5755968169761272, | |
| "grad_norm": 0.5271389383491623, | |
| "learning_rate": 1.3123092463864456e-06, | |
| "loss": 0.0058, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 1.5782493368700266, | |
| "grad_norm": 0.31139484943580775, | |
| "learning_rate": 1.2967033542853918e-06, | |
| "loss": 0.0055, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.5809018567639257, | |
| "grad_norm": 0.31737735749701884, | |
| "learning_rate": 1.2811769733430406e-06, | |
| "loss": 0.0056, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 1.5835543766578248, | |
| "grad_norm": 122.25701299417396, | |
| "learning_rate": 1.2657304369165768e-06, | |
| "loss": 0.0109, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 1.5862068965517242, | |
| "grad_norm": 0.3525084942228702, | |
| "learning_rate": 1.250364076648894e-06, | |
| "loss": 0.0055, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 1.5888594164456233, | |
| "grad_norm": 0.3192413472102024, | |
| "learning_rate": 1.2350782224614689e-06, | |
| "loss": 0.0054, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 1.5915119363395225, | |
| "grad_norm": 0.29866301866270556, | |
| "learning_rate": 1.2198732025472876e-06, | |
| "loss": 0.0053, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.5941644562334218, | |
| "grad_norm": 106.17123075764732, | |
| "learning_rate": 1.2047493433637935e-06, | |
| "loss": 0.0891, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 1.596816976127321, | |
| "grad_norm": 0.6551855618573065, | |
| "learning_rate": 1.1897069696258756e-06, | |
| "loss": 0.0055, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 1.5994694960212201, | |
| "grad_norm": 0.3010392367594056, | |
| "learning_rate": 1.1747464042989037e-06, | |
| "loss": 0.0053, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 1.6021220159151195, | |
| "grad_norm": 0.2963399545869238, | |
| "learning_rate": 1.1598679685917901e-06, | |
| "loss": 0.0053, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 1.6047745358090184, | |
| "grad_norm": 7.673021209841488, | |
| "learning_rate": 1.1450719819500906e-06, | |
| "loss": 0.0059, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.6074270557029178, | |
| "grad_norm": 0.5894185742400115, | |
| "learning_rate": 1.1303587620491513e-06, | |
| "loss": 0.0053, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 1.6100795755968171, | |
| "grad_norm": 0.3022377778609233, | |
| "learning_rate": 1.1157286247872873e-06, | |
| "loss": 0.0053, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 1.612732095490716, | |
| "grad_norm": 0.998174163284724, | |
| "learning_rate": 1.1011818842789928e-06, | |
| "loss": 0.0053, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 1.6153846153846154, | |
| "grad_norm": 1.2605257767864775, | |
| "learning_rate": 1.0867188528482087e-06, | |
| "loss": 0.0053, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 1.6180371352785146, | |
| "grad_norm": 22.434991228580405, | |
| "learning_rate": 1.0723398410216085e-06, | |
| "loss": 0.0113, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.6206896551724137, | |
| "grad_norm": 0.30072458656709766, | |
| "learning_rate": 1.0580451575219304e-06, | |
| "loss": 0.0052, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 1.623342175066313, | |
| "grad_norm": 6.785389276141757, | |
| "learning_rate": 1.043835109261357e-06, | |
| "loss": 0.0067, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 1.6259946949602122, | |
| "grad_norm": 0.43895873989236295, | |
| "learning_rate": 1.0297100013349181e-06, | |
| "loss": 0.0052, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 1.6286472148541113, | |
| "grad_norm": 0.3095572825845269, | |
| "learning_rate": 1.0156701370139454e-06, | |
| "loss": 0.005, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 1.6312997347480107, | |
| "grad_norm": 0.2807082976360762, | |
| "learning_rate": 1.0017158177395531e-06, | |
| "loss": 0.0049, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.6339522546419099, | |
| "grad_norm": 0.2774911079059377, | |
| "learning_rate": 9.878473431161767e-07, | |
| "loss": 0.0049, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 1.636604774535809, | |
| "grad_norm": 0.39533257376305914, | |
| "learning_rate": 9.740650109051348e-07, | |
| "loss": 0.0049, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 1.6392572944297084, | |
| "grad_norm": 0.28296242314049025, | |
| "learning_rate": 9.603691170182316e-07, | |
| "loss": 0.005, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 1.6419098143236073, | |
| "grad_norm": 3.9041204188770458, | |
| "learning_rate": 9.467599555114137e-07, | |
| "loss": 0.0058, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 1.6445623342175066, | |
| "grad_norm": 0.2846533102236989, | |
| "learning_rate": 9.332378185784491e-07, | |
| "loss": 0.005, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.647214854111406, | |
| "grad_norm": 51.93209557405583, | |
| "learning_rate": 9.198029965446537e-07, | |
| "loss": 0.0101, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 1.649867374005305, | |
| "grad_norm": 0.2696954936621249, | |
| "learning_rate": 9.064557778606631e-07, | |
| "loss": 0.0048, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 1.6525198938992043, | |
| "grad_norm": 0.2695956972348163, | |
| "learning_rate": 8.931964490962364e-07, | |
| "loss": 0.0048, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 1.6551724137931034, | |
| "grad_norm": 94.98929934493466, | |
| "learning_rate": 8.800252949340998e-07, | |
| "loss": 0.0683, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 1.6578249336870026, | |
| "grad_norm": 1.2974916163318027, | |
| "learning_rate": 8.669425981638413e-07, | |
| "loss": 0.005, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.660477453580902, | |
| "grad_norm": 0.2710452417200414, | |
| "learning_rate": 8.539486396758357e-07, | |
| "loss": 0.0048, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 1.663129973474801, | |
| "grad_norm": 0.2666246456468241, | |
| "learning_rate": 8.410436984552112e-07, | |
| "loss": 0.0048, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 1.6657824933687002, | |
| "grad_norm": 1.6036131193583794, | |
| "learning_rate": 8.282280515758639e-07, | |
| "loss": 0.0049, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 1.6684350132625996, | |
| "grad_norm": 0.26275985762388926, | |
| "learning_rate": 8.15501974194508e-07, | |
| "loss": 0.0047, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 1.6710875331564987, | |
| "grad_norm": 0.2760173400456831, | |
| "learning_rate": 8.02865739544767e-07, | |
| "loss": 0.0047, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.6737400530503979, | |
| "grad_norm": 0.258375590884943, | |
| "learning_rate": 7.903196189313039e-07, | |
| "loss": 0.0046, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 1.6763925729442972, | |
| "grad_norm": 0.2622808578616796, | |
| "learning_rate": 7.778638817240042e-07, | |
| "loss": 0.0047, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 1.6790450928381961, | |
| "grad_norm": 0.25880678528941237, | |
| "learning_rate": 7.654987953521875e-07, | |
| "loss": 0.0046, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 1.6816976127320955, | |
| "grad_norm": 0.25847177799776694, | |
| "learning_rate": 7.532246252988617e-07, | |
| "loss": 0.0046, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 1.6843501326259946, | |
| "grad_norm": 9.543103543412947, | |
| "learning_rate": 7.410416350950333e-07, | |
| "loss": 0.006, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 1.6870026525198938, | |
| "grad_norm": 0.2583838978284882, | |
| "learning_rate": 7.289500863140414e-07, | |
| "loss": 0.0045, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 1.6896551724137931, | |
| "grad_norm": 0.25730496900642014, | |
| "learning_rate": 7.16950238565941e-07, | |
| "loss": 0.0046, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 1.6923076923076923, | |
| "grad_norm": 100.42652935388513, | |
| "learning_rate": 7.05042349491935e-07, | |
| "loss": 0.0583, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 1.6949602122015914, | |
| "grad_norm": 0.2532567668137234, | |
| "learning_rate": 6.932266747588395e-07, | |
| "loss": 0.0045, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 1.6976127320954908, | |
| "grad_norm": 0.3114113921947813, | |
| "learning_rate": 6.815034680535915e-07, | |
| "loss": 0.0045, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.70026525198939, | |
| "grad_norm": 0.3692338355058846, | |
| "learning_rate": 6.698729810778065e-07, | |
| "loss": 0.0046, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 1.702917771883289, | |
| "grad_norm": 6.091089006073549, | |
| "learning_rate": 6.583354635423755e-07, | |
| "loss": 0.0052, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 1.7055702917771884, | |
| "grad_norm": 26.336302510662076, | |
| "learning_rate": 6.46891163162095e-07, | |
| "loss": 0.0081, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 1.7082228116710876, | |
| "grad_norm": 0.29512973513581486, | |
| "learning_rate": 6.355403256503595e-07, | |
| "loss": 0.0045, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 1.7108753315649867, | |
| "grad_norm": 0.7223730285452392, | |
| "learning_rate": 6.242831947138806e-07, | |
| "loss": 0.0046, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.713527851458886, | |
| "grad_norm": 126.44512664163686, | |
| "learning_rate": 6.131200120474512e-07, | |
| "loss": 0.0318, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 1.716180371352785, | |
| "grad_norm": 0.40609491390564195, | |
| "learning_rate": 6.020510173287636e-07, | |
| "loss": 0.0046, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 1.7188328912466844, | |
| "grad_norm": 0.2507013793463241, | |
| "learning_rate": 5.910764482132575e-07, | |
| "loss": 0.0044, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 1.7214854111405835, | |
| "grad_norm": 0.24773838391502986, | |
| "learning_rate": 5.801965403290221e-07, | |
| "loss": 0.0044, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 1.7241379310344827, | |
| "grad_norm": 0.2486645179595358, | |
| "learning_rate": 5.694115272717326e-07, | |
| "loss": 0.0044, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.726790450928382, | |
| "grad_norm": 0.24730568044224677, | |
| "learning_rate": 5.587216405996343e-07, | |
| "loss": 0.0044, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 1.7294429708222812, | |
| "grad_norm": 0.2538856932859245, | |
| "learning_rate": 5.481271098285818e-07, | |
| "loss": 0.0043, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 1.7320954907161803, | |
| "grad_norm": 0.26593342998498926, | |
| "learning_rate": 5.376281624270946e-07, | |
| "loss": 0.0044, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 1.7347480106100797, | |
| "grad_norm": 0.248248124621863, | |
| "learning_rate": 5.272250238114857e-07, | |
| "loss": 0.0044, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 1.7374005305039788, | |
| "grad_norm": 0.2458121042836853, | |
| "learning_rate": 5.169179173410178e-07, | |
| "loss": 0.0044, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.740053050397878, | |
| "grad_norm": 0.24729370932843403, | |
| "learning_rate": 5.067070643131056e-07, | |
| "loss": 0.0044, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 1.7427055702917773, | |
| "grad_norm": 15.385259424248543, | |
| "learning_rate": 4.965926839585688e-07, | |
| "loss": 0.007, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 1.7453580901856764, | |
| "grad_norm": 0.2822390112635331, | |
| "learning_rate": 4.865749934369224e-07, | |
| "loss": 0.0043, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 1.7480106100795756, | |
| "grad_norm": 0.24839423194997115, | |
| "learning_rate": 4.766542078317121e-07, | |
| "loss": 0.0043, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 1.750663129973475, | |
| "grad_norm": 0.24111468553355503, | |
| "learning_rate": 4.668305401459022e-07, | |
| "loss": 0.0043, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.7533156498673739, | |
| "grad_norm": 0.2438447012071553, | |
| "learning_rate": 4.571042012972993e-07, | |
| "loss": 0.0043, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 1.7559681697612732, | |
| "grad_norm": 0.24361572051168376, | |
| "learning_rate": 4.4747540011401913e-07, | |
| "loss": 0.0043, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 1.7586206896551724, | |
| "grad_norm": 0.5275300344193206, | |
| "learning_rate": 4.379443433300129e-07, | |
| "loss": 0.0043, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 1.7612732095490715, | |
| "grad_norm": 0.23717312427507825, | |
| "learning_rate": 4.2851123558061927e-07, | |
| "loss": 0.0042, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 1.7639257294429709, | |
| "grad_norm": 0.2571916721136695, | |
| "learning_rate": 4.1917627939817793e-07, | |
| "loss": 0.0043, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.76657824933687, | |
| "grad_norm": 0.5553820884441111, | |
| "learning_rate": 4.0993967520767455e-07, | |
| "loss": 0.0044, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 1.7692307692307692, | |
| "grad_norm": 2.246712558405596, | |
| "learning_rate": 4.008016213224408e-07, | |
| "loss": 0.0045, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 1.7718832891246685, | |
| "grad_norm": 0.2358532445026429, | |
| "learning_rate": 3.9176231393990183e-07, | |
| "loss": 0.0042, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 1.7745358090185677, | |
| "grad_norm": 0.23255907341279805, | |
| "learning_rate": 3.8282194713735286e-07, | |
| "loss": 0.0041, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 1.7771883289124668, | |
| "grad_norm": 0.25068833373474997, | |
| "learning_rate": 3.739807128677986e-07, | |
| "loss": 0.0042, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.7798408488063662, | |
| "grad_norm": 0.28998360046758886, | |
| "learning_rate": 3.6523880095583554e-07, | |
| "loss": 0.0042, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 1.782493368700265, | |
| "grad_norm": 70.9543828204898, | |
| "learning_rate": 3.5659639909356725e-07, | |
| "loss": 0.0316, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 1.7851458885941645, | |
| "grad_norm": 11.517585657154022, | |
| "learning_rate": 3.480536928365824e-07, | |
| "loss": 0.0065, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 1.7877984084880638, | |
| "grad_norm": 0.24209238846563902, | |
| "learning_rate": 3.39610865599968e-07, | |
| "loss": 0.0042, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 1.7904509283819627, | |
| "grad_norm": 0.23781117806303836, | |
| "learning_rate": 3.3126809865436817e-07, | |
| "loss": 0.0041, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.793103448275862, | |
| "grad_norm": 0.2366381303709924, | |
| "learning_rate": 3.230255711220992e-07, | |
| "loss": 0.0042, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 1.7957559681697612, | |
| "grad_norm": 0.5459967361685063, | |
| "learning_rate": 3.1488345997329806e-07, | |
| "loss": 0.0042, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 1.7984084880636604, | |
| "grad_norm": 0.4047990607616113, | |
| "learning_rate": 3.0684194002212287e-07, | |
| "loss": 0.0041, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 1.8010610079575597, | |
| "grad_norm": 0.23141431270577473, | |
| "learning_rate": 2.9890118392300493e-07, | |
| "loss": 0.0041, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 1.8037135278514589, | |
| "grad_norm": 0.24515511136563692, | |
| "learning_rate": 2.910613621669356e-07, | |
| "loss": 0.0041, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.806366047745358, | |
| "grad_norm": 1.7050497926298187, | |
| "learning_rate": 2.83322643077808e-07, | |
| "loss": 0.0045, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 1.8090185676392574, | |
| "grad_norm": 0.23635479397726022, | |
| "learning_rate": 2.756851928088056e-07, | |
| "loss": 0.0042, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 1.8116710875331565, | |
| "grad_norm": 0.2342130054640727, | |
| "learning_rate": 2.681491753388282e-07, | |
| "loss": 0.0041, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 1.8143236074270557, | |
| "grad_norm": 0.23203537966106563, | |
| "learning_rate": 2.607147524689829e-07, | |
| "loss": 0.0041, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 1.816976127320955, | |
| "grad_norm": 0.2307593686090978, | |
| "learning_rate": 2.533820838190959e-07, | |
| "loss": 0.0041, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.819628647214854, | |
| "grad_norm": 0.2851962284663173, | |
| "learning_rate": 2.461513268242938e-07, | |
| "loss": 0.0042, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 1.8222811671087533, | |
| "grad_norm": 0.33440588668973653, | |
| "learning_rate": 2.390226367316262e-07, | |
| "loss": 0.0041, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 1.8249336870026527, | |
| "grad_norm": 0.23083456015449386, | |
| "learning_rate": 2.3199616659672352e-07, | |
| "loss": 0.0041, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 1.8275862068965516, | |
| "grad_norm": 0.8520328211146436, | |
| "learning_rate": 2.2507206728051732e-07, | |
| "loss": 0.0042, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 1.830238726790451, | |
| "grad_norm": 0.22849031893442262, | |
| "learning_rate": 2.1825048744600062e-07, | |
| "loss": 0.0041, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.83289124668435, | |
| "grad_norm": 143.7065462713553, | |
| "learning_rate": 2.1153157355503274e-07, | |
| "loss": 0.0296, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 1.8355437665782492, | |
| "grad_norm": 0.22881014891837861, | |
| "learning_rate": 2.0491546986519896e-07, | |
| "loss": 0.004, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 1.8381962864721486, | |
| "grad_norm": 0.2867682040530811, | |
| "learning_rate": 1.9840231842671087e-07, | |
| "loss": 0.0041, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 1.8408488063660478, | |
| "grad_norm": 0.23516372004144973, | |
| "learning_rate": 1.9199225907935492e-07, | |
| "loss": 0.004, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 1.843501326259947, | |
| "grad_norm": 0.23575605461699675, | |
| "learning_rate": 1.8568542944949474e-07, | |
| "loss": 0.0041, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "grad_norm": 0.23120379453374001, | |
| "learning_rate": 1.794819649471119e-07, | |
| "loss": 0.0041, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 1.8488063660477454, | |
| "grad_norm": 0.23494372122100343, | |
| "learning_rate": 1.7338199876289984e-07, | |
| "loss": 0.0041, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 1.8514588859416445, | |
| "grad_norm": 0.2267567245082868, | |
| "learning_rate": 1.6738566186540628e-07, | |
| "loss": 0.004, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 1.854111405835544, | |
| "grad_norm": 0.22946931691211364, | |
| "learning_rate": 1.6149308299821643e-07, | |
| "loss": 0.0041, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 1.8567639257294428, | |
| "grad_norm": 0.22698231845451255, | |
| "learning_rate": 1.5570438867719695e-07, | |
| "loss": 0.004, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.8594164456233422, | |
| "grad_norm": 0.22519946509907784, | |
| "learning_rate": 1.500197031877698e-07, | |
| "loss": 0.004, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 1.8620689655172413, | |
| "grad_norm": 0.2357152301031921, | |
| "learning_rate": 1.4443914858224938e-07, | |
| "loss": 0.0041, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 1.8647214854111405, | |
| "grad_norm": 0.23146666552008044, | |
| "learning_rate": 1.3896284467722398e-07, | |
| "loss": 0.0041, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 1.8673740053050398, | |
| "grad_norm": 0.22871909002642044, | |
| "learning_rate": 1.335909090509785e-07, | |
| "loss": 0.0041, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 1.870026525198939, | |
| "grad_norm": 0.2330036838970883, | |
| "learning_rate": 1.2832345704097082e-07, | |
| "loss": 0.0041, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.8726790450928381, | |
| "grad_norm": 0.22546368567502312, | |
| "learning_rate": 1.2316060174136e-07, | |
| "loss": 0.004, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 1.8753315649867375, | |
| "grad_norm": 0.2391827765733045, | |
| "learning_rate": 1.1810245400057152e-07, | |
| "loss": 0.0041, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 1.8779840848806366, | |
| "grad_norm": 0.23135321137439943, | |
| "learning_rate": 1.1314912241892184e-07, | |
| "loss": 0.004, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 1.8806366047745358, | |
| "grad_norm": 0.22616985941276838, | |
| "learning_rate": 1.0830071334628655e-07, | |
| "loss": 0.004, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 1.8832891246684351, | |
| "grad_norm": 0.22812730794748792, | |
| "learning_rate": 1.035573308798138e-07, | |
| "loss": 0.004, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.8859416445623343, | |
| "grad_norm": 0.2243059239586133, | |
| "learning_rate": 9.891907686169211e-08, | |
| "loss": 0.004, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 1.8885941644562334, | |
| "grad_norm": 0.4588092051271435, | |
| "learning_rate": 9.43860508769645e-08, | |
| "loss": 0.0041, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 1.8912466843501328, | |
| "grad_norm": 0.22536403743459937, | |
| "learning_rate": 8.995835025138677e-08, | |
| "loss": 0.004, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 1.8938992042440317, | |
| "grad_norm": 0.22565203184615454, | |
| "learning_rate": 8.563607004934193e-08, | |
| "loss": 0.004, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 1.896551724137931, | |
| "grad_norm": 0.22366314667273499, | |
| "learning_rate": 8.141930307179468e-08, | |
| "loss": 0.004, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.8992042440318302, | |
| "grad_norm": 0.2279783994579146, | |
| "learning_rate": 7.730813985430407e-08, | |
| "loss": 0.004, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 1.9018567639257293, | |
| "grad_norm": 0.24135436794835546, | |
| "learning_rate": 7.330266866507618e-08, | |
| "loss": 0.004, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 1.9045092838196287, | |
| "grad_norm": 0.23433143970955525, | |
| "learning_rate": 6.940297550306895e-08, | |
| "loss": 0.004, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 1.9071618037135278, | |
| "grad_norm": 0.22634949627182713, | |
| "learning_rate": 6.560914409614872e-08, | |
| "loss": 0.004, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 1.909814323607427, | |
| "grad_norm": 0.2387086723321628, | |
| "learning_rate": 6.192125589928821e-08, | |
| "loss": 0.004, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.9124668435013263, | |
| "grad_norm": 6.202647111806296, | |
| "learning_rate": 5.833939009282086e-08, | |
| "loss": 0.0053, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 1.9151193633952255, | |
| "grad_norm": 0.22863334490515452, | |
| "learning_rate": 5.486362358074093e-08, | |
| "loss": 0.0041, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 1.9177718832891246, | |
| "grad_norm": 0.22698135338110348, | |
| "learning_rate": 5.1494030989049926e-08, | |
| "loss": 0.004, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 1.920424403183024, | |
| "grad_norm": 0.22590216714405328, | |
| "learning_rate": 4.823068466415615e-08, | |
| "loss": 0.004, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "grad_norm": 0.22807141517652188, | |
| "learning_rate": 4.5073654671320965e-08, | |
| "loss": 0.0041, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.9257294429708223, | |
| "grad_norm": 1.201301894967121, | |
| "learning_rate": 4.202300879315446e-08, | |
| "loss": 0.0043, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 1.9283819628647216, | |
| "grad_norm": 0.23056138949041405, | |
| "learning_rate": 3.907881252816048e-08, | |
| "loss": 0.0041, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 1.9310344827586206, | |
| "grad_norm": 2.4312545444566362, | |
| "learning_rate": 3.6241129089329416e-08, | |
| "loss": 0.0043, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 1.93368700265252, | |
| "grad_norm": 0.22957360649656, | |
| "learning_rate": 3.351001940278209e-08, | |
| "loss": 0.004, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 1.936339522546419, | |
| "grad_norm": 0.22733977672942887, | |
| "learning_rate": 3.088554210646133e-08, | |
| "loss": 0.004, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.9389920424403182, | |
| "grad_norm": 0.22568780368466787, | |
| "learning_rate": 2.8367753548871335e-08, | |
| "loss": 0.004, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 1.9416445623342176, | |
| "grad_norm": 0.22705148098683284, | |
| "learning_rate": 2.595670778787196e-08, | |
| "loss": 0.004, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 1.9442970822281167, | |
| "grad_norm": 0.460666108715405, | |
| "learning_rate": 2.3652456589512983e-08, | |
| "loss": 0.0041, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 1.9469496021220158, | |
| "grad_norm": 0.22609962320812726, | |
| "learning_rate": 2.1455049426926666e-08, | |
| "loss": 0.004, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 1.9496021220159152, | |
| "grad_norm": 0.22507833764853438, | |
| "learning_rate": 1.9364533479263036e-08, | |
| "loss": 0.004, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.9522546419098143, | |
| "grad_norm": 0.23888217596144562, | |
| "learning_rate": 1.7380953630678488e-08, | |
| "loss": 0.004, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 1.9549071618037135, | |
| "grad_norm": 0.2330182908978619, | |
| "learning_rate": 1.5504352469371543e-08, | |
| "loss": 0.004, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 1.9575596816976129, | |
| "grad_norm": 0.22417303270400205, | |
| "learning_rate": 1.373477028666803e-08, | |
| "loss": 0.004, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 1.9602122015915118, | |
| "grad_norm": 0.22666476984553308, | |
| "learning_rate": 1.2072245076156786e-08, | |
| "loss": 0.004, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 1.9628647214854111, | |
| "grad_norm": 4.155083837928933, | |
| "learning_rate": 1.0516812532873622e-08, | |
| "loss": 0.005, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.9655172413793105, | |
| "grad_norm": 0.2266434772362444, | |
| "learning_rate": 9.068506052534732e-09, | |
| "loss": 0.004, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 1.9681697612732094, | |
| "grad_norm": 0.22759641996655722, | |
| "learning_rate": 7.727356730820035e-09, | |
| "loss": 0.0041, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 1.9708222811671088, | |
| "grad_norm": 0.22640459239936114, | |
| "learning_rate": 6.49339336270427e-09, | |
| "loss": 0.004, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 1.973474801061008, | |
| "grad_norm": 0.2269001817332957, | |
| "learning_rate": 5.366642441841374e-09, | |
| "loss": 0.004, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 1.976127320954907, | |
| "grad_norm": 0.22126857041830134, | |
| "learning_rate": 4.347128159993829e-09, | |
| "loss": 0.0039, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.9787798408488064, | |
| "grad_norm": 0.22613994671550425, | |
| "learning_rate": 3.4348724065119687e-09, | |
| "loss": 0.004, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 1.9814323607427056, | |
| "grad_norm": 0.22626971062577017, | |
| "learning_rate": 2.62989476786768e-09, | |
| "loss": 0.004, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 1.9840848806366047, | |
| "grad_norm": 0.22548279505487495, | |
| "learning_rate": 1.9322125272297488e-09, | |
| "loss": 0.004, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 1.986737400530504, | |
| "grad_norm": 0.2274724438060562, | |
| "learning_rate": 1.3418406640969272e-09, | |
| "loss": 0.004, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 1.9893899204244032, | |
| "grad_norm": 0.2259285197005954, | |
| "learning_rate": 8.587918539726403e-10, | |
| "loss": 0.004, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.9920424403183024, | |
| "grad_norm": 1.7457170404512745, | |
| "learning_rate": 4.830764680946453e-10, | |
| "loss": 0.0044, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 1.9946949602122017, | |
| "grad_norm": 0.22469337804785391, | |
| "learning_rate": 2.1470257321298815e-10, | |
| "loss": 0.004, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 1.9973474801061006, | |
| "grad_norm": 0.2279365937027892, | |
| "learning_rate": 5.3675931413477156e-11, | |
| "loss": 0.004, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.48703773792450233, | |
| "learning_rate": 0.0, | |
| "loss": 0.0042, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.590460479259491, | |
| "eval_runtime": 175.8842, | |
| "eval_samples_per_second": 12.025, | |
| "eval_steps_per_second": 1.507, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 754, | |
| "total_flos": 22077400891392.0, | |
| "train_loss": 0.6806143896792588, | |
| "train_runtime": 3320.7586, | |
| "train_samples_per_second": 3.632, | |
| "train_steps_per_second": 0.227 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 754, | |
| "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": 22077400891392.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |