| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 200, | |
| "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": 3.143817743419433, | |
| "learning_rate": 9.999956599329999e-06, | |
| "loss": 0.4703, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.005305039787798408, | |
| "grad_norm": 2.9153346228246715, | |
| "learning_rate": 9.999826398073438e-06, | |
| "loss": 0.4676, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.007957559681697613, | |
| "grad_norm": 2.1837013669916683, | |
| "learning_rate": 9.999609398490651e-06, | |
| "loss": 0.3752, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.010610079575596816, | |
| "grad_norm": 2.7009872655116145, | |
| "learning_rate": 9.999305604348804e-06, | |
| "loss": 0.494, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.013262599469496022, | |
| "grad_norm": 2.71509007103024, | |
| "learning_rate": 9.998915020921847e-06, | |
| "loss": 0.4101, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.015915119363395226, | |
| "grad_norm": 2.046773566123141, | |
| "learning_rate": 9.998437654990412e-06, | |
| "loss": 0.3118, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.01856763925729443, | |
| "grad_norm": 1.6805054009773157, | |
| "learning_rate": 9.997873514841703e-06, | |
| "loss": 0.2808, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.021220159151193633, | |
| "grad_norm": 1.9031301046910736, | |
| "learning_rate": 9.997222610269339e-06, | |
| "loss": 0.288, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.023872679045092837, | |
| "grad_norm": 2.2790419571248, | |
| "learning_rate": 9.996484952573203e-06, | |
| "loss": 0.3187, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.026525198938992044, | |
| "grad_norm": 2.005476307323896, | |
| "learning_rate": 9.995660554559225e-06, | |
| "loss": 0.2684, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.029177718832891247, | |
| "grad_norm": 2.1118955586623542, | |
| "learning_rate": 9.99474943053918e-06, | |
| "loss": 0.2861, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.03183023872679045, | |
| "grad_norm": 1.9084863447700302, | |
| "learning_rate": 9.993751596330421e-06, | |
| "loss": 0.2447, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.034482758620689655, | |
| "grad_norm": 2.443033048528926, | |
| "learning_rate": 9.99266706925562e-06, | |
| "loss": 0.295, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.03713527851458886, | |
| "grad_norm": 2.252694252488703, | |
| "learning_rate": 9.991495868142457e-06, | |
| "loss": 0.2895, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.03978779840848806, | |
| "grad_norm": 2.134403724246919, | |
| "learning_rate": 9.990238013323298e-06, | |
| "loss": 0.2704, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.042440318302387266, | |
| "grad_norm": 2.28401373031769, | |
| "learning_rate": 9.98889352663484e-06, | |
| "loss": 0.2648, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.04509283819628647, | |
| "grad_norm": 2.1064551964073286, | |
| "learning_rate": 9.987462431417732e-06, | |
| "loss": 0.2033, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.04774535809018567, | |
| "grad_norm": 2.1590960874250102, | |
| "learning_rate": 9.985944752516168e-06, | |
| "loss": 0.206, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.050397877984084884, | |
| "grad_norm": 1.787435615555535, | |
| "learning_rate": 9.984340516277464e-06, | |
| "loss": 0.2446, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.05305039787798409, | |
| "grad_norm": 2.154695176620596, | |
| "learning_rate": 9.982649750551589e-06, | |
| "loss": 0.2428, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05570291777188329, | |
| "grad_norm": 2.013092950603686, | |
| "learning_rate": 9.980872484690689e-06, | |
| "loss": 0.2095, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.058355437665782495, | |
| "grad_norm": 1.871221689867333, | |
| "learning_rate": 9.979008749548575e-06, | |
| "loss": 0.2119, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.0610079575596817, | |
| "grad_norm": 2.1428411294867975, | |
| "learning_rate": 9.977058577480192e-06, | |
| "loss": 0.2031, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.0636604774535809, | |
| "grad_norm": 2.029431401764423, | |
| "learning_rate": 9.975022002341045e-06, | |
| "loss": 0.2279, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.06631299734748011, | |
| "grad_norm": 1.9037807432106393, | |
| "learning_rate": 9.972899059486629e-06, | |
| "loss": 0.2236, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06896551724137931, | |
| "grad_norm": 1.5412274357921503, | |
| "learning_rate": 9.970689785771798e-06, | |
| "loss": 0.1816, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.07161803713527852, | |
| "grad_norm": 2.230329801114519, | |
| "learning_rate": 9.968394219550136e-06, | |
| "loss": 0.2428, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.07427055702917772, | |
| "grad_norm": 1.902794098881427, | |
| "learning_rate": 9.966012400673291e-06, | |
| "loss": 0.2154, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.07692307692307693, | |
| "grad_norm": 1.8780742288675212, | |
| "learning_rate": 9.96354437049027e-06, | |
| "loss": 0.1972, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.07957559681697612, | |
| "grad_norm": 1.8047024624739505, | |
| "learning_rate": 9.960990171846745e-06, | |
| "loss": 0.2003, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.08222811671087533, | |
| "grad_norm": 1.99420196004025, | |
| "learning_rate": 9.958349849084286e-06, | |
| "loss": 0.2229, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.08488063660477453, | |
| "grad_norm": 1.9295211213650523, | |
| "learning_rate": 9.955623448039605e-06, | |
| "loss": 0.2077, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.08753315649867374, | |
| "grad_norm": 1.9560816295756256, | |
| "learning_rate": 9.952811016043753e-06, | |
| "loss": 0.2035, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.09018567639257294, | |
| "grad_norm": 1.8994575592237317, | |
| "learning_rate": 9.949912601921306e-06, | |
| "loss": 0.2365, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.09283819628647215, | |
| "grad_norm": 1.7820353508663918, | |
| "learning_rate": 9.946928255989507e-06, | |
| "loss": 0.1889, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.09549071618037135, | |
| "grad_norm": 1.9972563802629155, | |
| "learning_rate": 9.943858030057404e-06, | |
| "loss": 0.1992, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.09814323607427056, | |
| "grad_norm": 1.9858847962279158, | |
| "learning_rate": 9.94070197742494e-06, | |
| "loss": 0.2126, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.10079575596816977, | |
| "grad_norm": 2.2468613321703375, | |
| "learning_rate": 9.937460152882035e-06, | |
| "loss": 0.1966, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.10344827586206896, | |
| "grad_norm": 1.9047848108957046, | |
| "learning_rate": 9.934132612707631e-06, | |
| "loss": 0.1929, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.10610079575596817, | |
| "grad_norm": 1.5198039037792352, | |
| "learning_rate": 9.93071941466872e-06, | |
| "loss": 0.1454, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.10875331564986737, | |
| "grad_norm": 1.942256631038327, | |
| "learning_rate": 9.927220618019331e-06, | |
| "loss": 0.1927, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.11140583554376658, | |
| "grad_norm": 1.9039836584761076, | |
| "learning_rate": 9.923636283499513e-06, | |
| "loss": 0.1683, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.11405835543766578, | |
| "grad_norm": 2.3882871093584015, | |
| "learning_rate": 9.919966473334278e-06, | |
| "loss": 0.1925, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.11671087533156499, | |
| "grad_norm": 1.9482729522397495, | |
| "learning_rate": 9.916211251232507e-06, | |
| "loss": 0.204, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.11936339522546419, | |
| "grad_norm": 1.9541832414635472, | |
| "learning_rate": 9.912370682385866e-06, | |
| "loss": 0.1718, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1220159151193634, | |
| "grad_norm": 2.2469271806449744, | |
| "learning_rate": 9.908444833467659e-06, | |
| "loss": 0.2099, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.1246684350132626, | |
| "grad_norm": 1.9273893068530785, | |
| "learning_rate": 9.904433772631674e-06, | |
| "loss": 0.1896, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.1273209549071618, | |
| "grad_norm": 2.020005570778388, | |
| "learning_rate": 9.900337569511003e-06, | |
| "loss": 0.2135, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.129973474801061, | |
| "grad_norm": 1.787168676896887, | |
| "learning_rate": 9.896156295216832e-06, | |
| "loss": 0.1895, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.13262599469496023, | |
| "grad_norm": 1.9492709359169609, | |
| "learning_rate": 9.891890022337201e-06, | |
| "loss": 0.1948, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.13527851458885942, | |
| "grad_norm": 1.4707392860636022, | |
| "learning_rate": 9.88753882493575e-06, | |
| "loss": 0.1494, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.13793103448275862, | |
| "grad_norm": 1.8405786005165676, | |
| "learning_rate": 9.883102778550434e-06, | |
| "loss": 0.1752, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.14058355437665782, | |
| "grad_norm": 1.8735255704087945, | |
| "learning_rate": 9.878581960192206e-06, | |
| "loss": 0.1779, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.14323607427055704, | |
| "grad_norm": 2.0141493172675746, | |
| "learning_rate": 9.873976448343685e-06, | |
| "loss": 0.2168, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.14588859416445624, | |
| "grad_norm": 1.7017677167638687, | |
| "learning_rate": 9.86928632295779e-06, | |
| "loss": 0.1589, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.14854111405835543, | |
| "grad_norm": 1.6283124591177214, | |
| "learning_rate": 9.864511665456355e-06, | |
| "loss": 0.1399, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.15119363395225463, | |
| "grad_norm": 2.2298913035183845, | |
| "learning_rate": 9.859652558728714e-06, | |
| "loss": 0.2046, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 2.12248040462289, | |
| "learning_rate": 9.854709087130261e-06, | |
| "loss": 0.1859, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.15649867374005305, | |
| "grad_norm": 2.4614522882556247, | |
| "learning_rate": 9.84968133648099e-06, | |
| "loss": 0.2284, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.15915119363395225, | |
| "grad_norm": 2.0728088827577578, | |
| "learning_rate": 9.844569394063997e-06, | |
| "loss": 0.1826, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.16180371352785147, | |
| "grad_norm": 1.626832715567298, | |
| "learning_rate": 9.839373348623976e-06, | |
| "loss": 0.144, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.16445623342175067, | |
| "grad_norm": 2.095258455090102, | |
| "learning_rate": 9.834093290365665e-06, | |
| "loss": 0.1762, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.16710875331564987, | |
| "grad_norm": 2.1180496536142495, | |
| "learning_rate": 9.828729310952292e-06, | |
| "loss": 0.1754, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.16976127320954906, | |
| "grad_norm": 2.523281048591522, | |
| "learning_rate": 9.823281503503976e-06, | |
| "loss": 0.2063, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.1724137931034483, | |
| "grad_norm": 1.7605117934687937, | |
| "learning_rate": 9.817749962596115e-06, | |
| "loss": 0.1513, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.17506631299734748, | |
| "grad_norm": 1.8068568519736308, | |
| "learning_rate": 9.812134784257743e-06, | |
| "loss": 0.1695, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.17771883289124668, | |
| "grad_norm": 2.194003391788825, | |
| "learning_rate": 9.80643606596986e-06, | |
| "loss": 0.1873, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.18037135278514588, | |
| "grad_norm": 1.8221102873639325, | |
| "learning_rate": 9.80065390666374e-06, | |
| "loss": 0.1709, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.1830238726790451, | |
| "grad_norm": 1.81681516720243, | |
| "learning_rate": 9.794788406719223e-06, | |
| "loss": 0.1611, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.1856763925729443, | |
| "grad_norm": 1.6215357337836858, | |
| "learning_rate": 9.788839667962956e-06, | |
| "loss": 0.1491, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.1883289124668435, | |
| "grad_norm": 1.9544453657582719, | |
| "learning_rate": 9.78280779366664e-06, | |
| "loss": 0.1666, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.1909814323607427, | |
| "grad_norm": 1.9891058976186013, | |
| "learning_rate": 9.77669288854523e-06, | |
| "loss": 0.1652, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.19363395225464192, | |
| "grad_norm": 1.7657716439562443, | |
| "learning_rate": 9.770495058755113e-06, | |
| "loss": 0.1665, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.1962864721485411, | |
| "grad_norm": 2.0702179893057657, | |
| "learning_rate": 9.764214411892283e-06, | |
| "loss": 0.1839, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.1989389920424403, | |
| "grad_norm": 1.94174040080256, | |
| "learning_rate": 9.757851056990446e-06, | |
| "loss": 0.1514, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.20159151193633953, | |
| "grad_norm": 2.0359509600829355, | |
| "learning_rate": 9.751405104519151e-06, | |
| "loss": 0.168, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.20424403183023873, | |
| "grad_norm": 1.8488897921580416, | |
| "learning_rate": 9.744876666381861e-06, | |
| "loss": 0.1642, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.20689655172413793, | |
| "grad_norm": 1.9493633096974006, | |
| "learning_rate": 9.738265855914014e-06, | |
| "loss": 0.1451, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.20954907161803712, | |
| "grad_norm": 1.969779554700488, | |
| "learning_rate": 9.731572787881045e-06, | |
| "loss": 0.1738, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.21220159151193635, | |
| "grad_norm": 1.6792825100474953, | |
| "learning_rate": 9.724797578476414e-06, | |
| "loss": 0.1243, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.21485411140583555, | |
| "grad_norm": 1.9140356369318925, | |
| "learning_rate": 9.71794034531957e-06, | |
| "loss": 0.1591, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.21750663129973474, | |
| "grad_norm": 1.7940577658445638, | |
| "learning_rate": 9.711001207453919e-06, | |
| "loss": 0.1411, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.22015915119363394, | |
| "grad_norm": 2.383028621766121, | |
| "learning_rate": 9.703980285344752e-06, | |
| "loss": 0.1752, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.22281167108753316, | |
| "grad_norm": 2.051261568492118, | |
| "learning_rate": 9.696877700877162e-06, | |
| "loss": 0.1974, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.22546419098143236, | |
| "grad_norm": 1.694394298477383, | |
| "learning_rate": 9.689693577353917e-06, | |
| "loss": 0.1395, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.22811671087533156, | |
| "grad_norm": 1.8051445315092405, | |
| "learning_rate": 9.682428039493325e-06, | |
| "loss": 0.1576, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.23076923076923078, | |
| "grad_norm": 1.988060978591615, | |
| "learning_rate": 9.675081213427076e-06, | |
| "loss": 0.152, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.23342175066312998, | |
| "grad_norm": 1.6413056627568852, | |
| "learning_rate": 9.667653226698033e-06, | |
| "loss": 0.157, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.23607427055702918, | |
| "grad_norm": 1.7068088087742874, | |
| "learning_rate": 9.660144208258039e-06, | |
| "loss": 0.1518, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.23872679045092837, | |
| "grad_norm": 2.0128746338247856, | |
| "learning_rate": 9.652554288465668e-06, | |
| "loss": 0.1792, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.2413793103448276, | |
| "grad_norm": 1.7874268938466609, | |
| "learning_rate": 9.644883599083959e-06, | |
| "loss": 0.1459, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.2440318302387268, | |
| "grad_norm": 2.15165919348902, | |
| "learning_rate": 9.637132273278135e-06, | |
| "loss": 0.1745, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.246684350132626, | |
| "grad_norm": 2.3979359764341415, | |
| "learning_rate": 9.629300445613294e-06, | |
| "loss": 0.2075, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.2493368700265252, | |
| "grad_norm": 2.250801983136737, | |
| "learning_rate": 9.62138825205206e-06, | |
| "loss": 0.1958, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.2519893899204244, | |
| "grad_norm": 2.021920489405669, | |
| "learning_rate": 9.613395829952233e-06, | |
| "loss": 0.1608, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2546419098143236, | |
| "grad_norm": 2.2291104046239076, | |
| "learning_rate": 9.605323318064403e-06, | |
| "loss": 0.196, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.2572944297082228, | |
| "grad_norm": 1.7681014524697132, | |
| "learning_rate": 9.59717085652954e-06, | |
| "loss": 0.1367, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.259946949602122, | |
| "grad_norm": 2.186660089739903, | |
| "learning_rate": 9.588938586876564e-06, | |
| "loss": 0.1485, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.2625994694960212, | |
| "grad_norm": 2.2507510523601044, | |
| "learning_rate": 9.580626652019878e-06, | |
| "loss": 0.1795, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.26525198938992045, | |
| "grad_norm": 1.9046066926852447, | |
| "learning_rate": 9.5722351962569e-06, | |
| "loss": 0.1676, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.26790450928381965, | |
| "grad_norm": 1.8505248856341252, | |
| "learning_rate": 9.563764365265553e-06, | |
| "loss": 0.1539, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.27055702917771884, | |
| "grad_norm": 2.0270020661057755, | |
| "learning_rate": 9.555214306101732e-06, | |
| "loss": 0.1702, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.27320954907161804, | |
| "grad_norm": 1.6795319874110077, | |
| "learning_rate": 9.546585167196755e-06, | |
| "loss": 0.1503, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.27586206896551724, | |
| "grad_norm": 2.341752387924688, | |
| "learning_rate": 9.537877098354787e-06, | |
| "loss": 0.1766, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.27851458885941643, | |
| "grad_norm": 2.5362314705747093, | |
| "learning_rate": 9.529090250750234e-06, | |
| "loss": 0.1854, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.28116710875331563, | |
| "grad_norm": 1.6421242469941448, | |
| "learning_rate": 9.52022477692513e-06, | |
| "loss": 0.12, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.2838196286472148, | |
| "grad_norm": 1.7705962516319114, | |
| "learning_rate": 9.511280830786471e-06, | |
| "loss": 0.15, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.2864721485411141, | |
| "grad_norm": 1.8018966800140481, | |
| "learning_rate": 9.502258567603563e-06, | |
| "loss": 0.1446, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.2891246684350133, | |
| "grad_norm": 2.154026465895243, | |
| "learning_rate": 9.493158144005314e-06, | |
| "loss": 0.175, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.2917771883289125, | |
| "grad_norm": 2.091201257414692, | |
| "learning_rate": 9.483979717977513e-06, | |
| "loss": 0.1533, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.29442970822281167, | |
| "grad_norm": 2.049226329317327, | |
| "learning_rate": 9.474723448860096e-06, | |
| "loss": 0.1582, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.29708222811671087, | |
| "grad_norm": 1.6937916325037254, | |
| "learning_rate": 9.46538949734438e-06, | |
| "loss": 0.1319, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.29973474801061006, | |
| "grad_norm": 2.2804043446780193, | |
| "learning_rate": 9.455978025470257e-06, | |
| "loss": 0.1744, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.30238726790450926, | |
| "grad_norm": 2.427110301571042, | |
| "learning_rate": 9.44648919662341e-06, | |
| "loss": 0.1972, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.3050397877984085, | |
| "grad_norm": 2.4719255725338307, | |
| "learning_rate": 9.436923175532442e-06, | |
| "loss": 0.1863, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 1.9198368860593722, | |
| "learning_rate": 9.427280128266049e-06, | |
| "loss": 0.1428, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.3103448275862069, | |
| "grad_norm": 1.9531949152663541, | |
| "learning_rate": 9.417560222230115e-06, | |
| "loss": 0.1723, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.3129973474801061, | |
| "grad_norm": 2.1363324241184456, | |
| "learning_rate": 9.407763626164812e-06, | |
| "loss": 0.1758, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.3156498673740053, | |
| "grad_norm": 1.8882375069693482, | |
| "learning_rate": 9.397890510141674e-06, | |
| "loss": 0.1248, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.3183023872679045, | |
| "grad_norm": 2.1194646896664584, | |
| "learning_rate": 9.387941045560641e-06, | |
| "loss": 0.1589, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.3209549071618037, | |
| "grad_norm": 1.5127624036717937, | |
| "learning_rate": 9.377915405147085e-06, | |
| "loss": 0.1446, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.32360742705570295, | |
| "grad_norm": 2.158233386717242, | |
| "learning_rate": 9.367813762948809e-06, | |
| "loss": 0.1584, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.32625994694960214, | |
| "grad_norm": 1.941105678166514, | |
| "learning_rate": 9.357636294333031e-06, | |
| "loss": 0.1272, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.32891246684350134, | |
| "grad_norm": 1.8205147183171333, | |
| "learning_rate": 9.347383175983333e-06, | |
| "loss": 0.1545, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.33156498673740054, | |
| "grad_norm": 1.6927779015644462, | |
| "learning_rate": 9.337054585896596e-06, | |
| "loss": 0.1202, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.33421750663129973, | |
| "grad_norm": 1.9276160632994672, | |
| "learning_rate": 9.326650703379913e-06, | |
| "loss": 0.1432, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.33687002652519893, | |
| "grad_norm": 1.9253639897291404, | |
| "learning_rate": 9.316171709047475e-06, | |
| "loss": 0.148, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.3395225464190981, | |
| "grad_norm": 1.8807065660047437, | |
| "learning_rate": 9.305617784817426e-06, | |
| "loss": 0.1401, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.3421750663129973, | |
| "grad_norm": 1.95250475420529, | |
| "learning_rate": 9.294989113908726e-06, | |
| "loss": 0.1548, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.3448275862068966, | |
| "grad_norm": 2.3715397805842255, | |
| "learning_rate": 9.284285880837947e-06, | |
| "loss": 0.1756, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.34748010610079577, | |
| "grad_norm": 1.93914417403847, | |
| "learning_rate": 9.273508271416082e-06, | |
| "loss": 0.1399, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.35013262599469497, | |
| "grad_norm": 1.9560254067556455, | |
| "learning_rate": 9.262656472745324e-06, | |
| "loss": 0.1389, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.35278514588859416, | |
| "grad_norm": 1.9150433290760882, | |
| "learning_rate": 9.251730673215802e-06, | |
| "loss": 0.1407, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.35543766578249336, | |
| "grad_norm": 1.8135844757186297, | |
| "learning_rate": 9.240731062502323e-06, | |
| "loss": 0.1306, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.35809018567639256, | |
| "grad_norm": 2.162238929625922, | |
| "learning_rate": 9.229657831561082e-06, | |
| "loss": 0.1496, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.36074270557029176, | |
| "grad_norm": 2.055713876885785, | |
| "learning_rate": 9.218511172626333e-06, | |
| "loss": 0.1753, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.363395225464191, | |
| "grad_norm": 1.6946475181680336, | |
| "learning_rate": 9.207291279207058e-06, | |
| "loss": 0.1335, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.3660477453580902, | |
| "grad_norm": 2.101107547619466, | |
| "learning_rate": 9.195998346083621e-06, | |
| "loss": 0.1493, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.3687002652519894, | |
| "grad_norm": 1.8055607972774566, | |
| "learning_rate": 9.184632569304365e-06, | |
| "loss": 0.1262, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.3713527851458886, | |
| "grad_norm": 1.690201779069974, | |
| "learning_rate": 9.173194146182219e-06, | |
| "loss": 0.1499, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3740053050397878, | |
| "grad_norm": 2.1397994585808955, | |
| "learning_rate": 9.161683275291275e-06, | |
| "loss": 0.157, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.376657824933687, | |
| "grad_norm": 1.9870860707292484, | |
| "learning_rate": 9.150100156463337e-06, | |
| "loss": 0.1529, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.3793103448275862, | |
| "grad_norm": 2.207156912182982, | |
| "learning_rate": 9.138444990784455e-06, | |
| "loss": 0.1859, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.3819628647214854, | |
| "grad_norm": 1.9713520675081648, | |
| "learning_rate": 9.126717980591422e-06, | |
| "loss": 0.1459, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "grad_norm": 1.7065232206005256, | |
| "learning_rate": 9.114919329468283e-06, | |
| "loss": 0.1461, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.38726790450928383, | |
| "grad_norm": 2.07601306564582, | |
| "learning_rate": 9.103049242242781e-06, | |
| "loss": 0.1785, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.38992042440318303, | |
| "grad_norm": 1.776835787519842, | |
| "learning_rate": 9.091107924982814e-06, | |
| "loss": 0.1525, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.3925729442970822, | |
| "grad_norm": 1.8663700533344434, | |
| "learning_rate": 9.079095584992848e-06, | |
| "loss": 0.1409, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.3952254641909814, | |
| "grad_norm": 2.2784042310709407, | |
| "learning_rate": 9.067012430810326e-06, | |
| "loss": 0.1785, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.3978779840848806, | |
| "grad_norm": 1.7721799125937119, | |
| "learning_rate": 9.05485867220204e-06, | |
| "loss": 0.1346, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.4005305039787798, | |
| "grad_norm": 2.13950597469927, | |
| "learning_rate": 9.0426345201605e-06, | |
| "loss": 0.1652, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.40318302387267907, | |
| "grad_norm": 1.8810097271179547, | |
| "learning_rate": 9.03034018690026e-06, | |
| "loss": 0.1449, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.40583554376657827, | |
| "grad_norm": 2.4159078987026863, | |
| "learning_rate": 9.01797588585424e-06, | |
| "loss": 0.1683, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.40848806366047746, | |
| "grad_norm": 2.0980888650490668, | |
| "learning_rate": 9.00554183167002e-06, | |
| "loss": 0.1424, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.41114058355437666, | |
| "grad_norm": 2.062288135177168, | |
| "learning_rate": 8.993038240206114e-06, | |
| "loss": 0.1392, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.41379310344827586, | |
| "grad_norm": 1.9204203170622036, | |
| "learning_rate": 8.98046532852822e-06, | |
| "loss": 0.1418, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.41644562334217505, | |
| "grad_norm": 2.08299140855318, | |
| "learning_rate": 8.967823314905452e-06, | |
| "loss": 0.1486, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.41909814323607425, | |
| "grad_norm": 1.844754808254566, | |
| "learning_rate": 8.95511241880656e-06, | |
| "loss": 0.1263, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.4217506631299735, | |
| "grad_norm": 1.902160167609469, | |
| "learning_rate": 8.942332860896102e-06, | |
| "loss": 0.1435, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.4244031830238727, | |
| "grad_norm": 1.4047023346474345, | |
| "learning_rate": 8.929484863030631e-06, | |
| "loss": 0.1204, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.4270557029177719, | |
| "grad_norm": 1.5383340677271746, | |
| "learning_rate": 8.91656864825483e-06, | |
| "loss": 0.1307, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.4297082228116711, | |
| "grad_norm": 1.8646076049063893, | |
| "learning_rate": 8.903584440797652e-06, | |
| "loss": 0.1403, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.4323607427055703, | |
| "grad_norm": 1.6149573488861944, | |
| "learning_rate": 8.890532466068417e-06, | |
| "loss": 0.1381, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.4350132625994695, | |
| "grad_norm": 2.1580924076315715, | |
| "learning_rate": 8.877412950652907e-06, | |
| "loss": 0.1577, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.4376657824933687, | |
| "grad_norm": 2.2128873070419375, | |
| "learning_rate": 8.864226122309423e-06, | |
| "loss": 0.1526, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.4403183023872679, | |
| "grad_norm": 2.174869482049831, | |
| "learning_rate": 8.850972209964837e-06, | |
| "loss": 0.1473, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.44297082228116713, | |
| "grad_norm": 1.6128773258978748, | |
| "learning_rate": 8.837651443710623e-06, | |
| "loss": 0.1217, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.44562334217506633, | |
| "grad_norm": 1.8324959002361592, | |
| "learning_rate": 8.824264054798852e-06, | |
| "loss": 0.1509, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.4482758620689655, | |
| "grad_norm": 1.708365841100358, | |
| "learning_rate": 8.810810275638183e-06, | |
| "loss": 0.124, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.4509283819628647, | |
| "grad_norm": 1.7502467085637508, | |
| "learning_rate": 8.797290339789827e-06, | |
| "loss": 0.1299, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.4535809018567639, | |
| "grad_norm": 2.060960880635761, | |
| "learning_rate": 8.783704481963498e-06, | |
| "loss": 0.1428, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.4562334217506631, | |
| "grad_norm": 1.8464905297693734, | |
| "learning_rate": 8.770052938013323e-06, | |
| "loss": 0.1502, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.4588859416445623, | |
| "grad_norm": 1.8087559826720645, | |
| "learning_rate": 8.756335944933768e-06, | |
| "loss": 0.1351, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 1.7162671809778718, | |
| "learning_rate": 8.742553740855507e-06, | |
| "loss": 0.1244, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.46419098143236076, | |
| "grad_norm": 2.25916831884875, | |
| "learning_rate": 8.728706565041296e-06, | |
| "loss": 0.176, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.46684350132625996, | |
| "grad_norm": 1.9927732799007924, | |
| "learning_rate": 8.714794657881818e-06, | |
| "loss": 0.1329, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.46949602122015915, | |
| "grad_norm": 1.9150506644030154, | |
| "learning_rate": 8.700818260891512e-06, | |
| "loss": 0.1455, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.47214854111405835, | |
| "grad_norm": 1.918021311795205, | |
| "learning_rate": 8.686777616704375e-06, | |
| "loss": 0.14, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.47480106100795755, | |
| "grad_norm": 2.0258194425535203, | |
| "learning_rate": 8.67267296906975e-06, | |
| "loss": 0.1676, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.47745358090185674, | |
| "grad_norm": 1.9171122906217821, | |
| "learning_rate": 8.658504562848104e-06, | |
| "loss": 0.1369, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.48010610079575594, | |
| "grad_norm": 2.218531055729975, | |
| "learning_rate": 8.644272644006764e-06, | |
| "loss": 0.152, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.4827586206896552, | |
| "grad_norm": 2.0038883017886375, | |
| "learning_rate": 8.629977459615655e-06, | |
| "loss": 0.1426, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.4854111405835544, | |
| "grad_norm": 1.7175445641951717, | |
| "learning_rate": 8.61561925784301e-06, | |
| "loss": 0.1249, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.4880636604774536, | |
| "grad_norm": 1.815071151822735, | |
| "learning_rate": 8.601198287951059e-06, | |
| "loss": 0.1301, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.4907161803713528, | |
| "grad_norm": 2.1561780015246463, | |
| "learning_rate": 8.586714800291704e-06, | |
| "loss": 0.1397, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.493368700265252, | |
| "grad_norm": 1.9294815320965064, | |
| "learning_rate": 8.572169046302174e-06, | |
| "loss": 0.141, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.4960212201591512, | |
| "grad_norm": 1.659333790717064, | |
| "learning_rate": 8.557561278500656e-06, | |
| "loss": 0.1214, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.4986737400530504, | |
| "grad_norm": 1.6753928986149604, | |
| "learning_rate": 8.542891750481913e-06, | |
| "loss": 0.1238, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.5013262599469496, | |
| "grad_norm": 1.909985436763465, | |
| "learning_rate": 8.528160716912882e-06, | |
| "loss": 0.1189, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.5039787798408488, | |
| "grad_norm": 1.956919058433968, | |
| "learning_rate": 8.513368433528255e-06, | |
| "loss": 0.1328, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.506631299734748, | |
| "grad_norm": 1.7825907922117281, | |
| "learning_rate": 8.498515157126038e-06, | |
| "loss": 0.1362, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.5092838196286472, | |
| "grad_norm": 2.1745948430153232, | |
| "learning_rate": 8.483601145563087e-06, | |
| "loss": 0.1403, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.5119363395225465, | |
| "grad_norm": 1.5767065409411627, | |
| "learning_rate": 8.46862665775064e-06, | |
| "loss": 0.121, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.5145888594164456, | |
| "grad_norm": 2.1206690113650737, | |
| "learning_rate": 8.45359195364982e-06, | |
| "loss": 0.1435, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.5172413793103449, | |
| "grad_norm": 2.0215180716784453, | |
| "learning_rate": 8.438497294267117e-06, | |
| "loss": 0.1362, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.519893899204244, | |
| "grad_norm": 2.2299825948556697, | |
| "learning_rate": 8.423342941649866e-06, | |
| "loss": 0.1337, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.5225464190981433, | |
| "grad_norm": 2.262534468463096, | |
| "learning_rate": 8.40812915888169e-06, | |
| "loss": 0.1538, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.5251989389920424, | |
| "grad_norm": 2.306393950737034, | |
| "learning_rate": 8.392856210077932e-06, | |
| "loss": 0.1696, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.5278514588859416, | |
| "grad_norm": 1.9803460856435193, | |
| "learning_rate": 8.37752436038108e-06, | |
| "loss": 0.1307, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.5305039787798409, | |
| "grad_norm": 1.7912221503357104, | |
| "learning_rate": 8.36213387595615e-06, | |
| "loss": 0.1292, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5305039787798409, | |
| "eval_loss": 0.14994381368160248, | |
| "eval_runtime": 1.3023, | |
| "eval_samples_per_second": 23.804, | |
| "eval_steps_per_second": 6.143, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.53315649867374, | |
| "grad_norm": 1.6059764600949016, | |
| "learning_rate": 8.34668502398608e-06, | |
| "loss": 0.106, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.5358090185676393, | |
| "grad_norm": 1.706895090532193, | |
| "learning_rate": 8.331178072667079e-06, | |
| "loss": 0.1261, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.5384615384615384, | |
| "grad_norm": 1.7309972720553148, | |
| "learning_rate": 8.315613291203977e-06, | |
| "loss": 0.1177, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.5411140583554377, | |
| "grad_norm": 2.119393668480105, | |
| "learning_rate": 8.299990949805551e-06, | |
| "loss": 0.1512, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.5437665782493368, | |
| "grad_norm": 2.1446534939795194, | |
| "learning_rate": 8.28431131967984e-06, | |
| "loss": 0.1451, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.5464190981432361, | |
| "grad_norm": 2.3040675508103057, | |
| "learning_rate": 8.268574673029415e-06, | |
| "loss": 0.1528, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.5490716180371353, | |
| "grad_norm": 1.828325753916526, | |
| "learning_rate": 8.252781283046688e-06, | |
| "loss": 0.1346, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.5517241379310345, | |
| "grad_norm": 1.8710705926707558, | |
| "learning_rate": 8.23693142390914e-06, | |
| "loss": 0.1463, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.5543766578249337, | |
| "grad_norm": 1.8384322195422618, | |
| "learning_rate": 8.22102537077457e-06, | |
| "loss": 0.1383, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.5570291777188329, | |
| "grad_norm": 1.9467404806792865, | |
| "learning_rate": 8.205063399776326e-06, | |
| "loss": 0.1415, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.5596816976127321, | |
| "grad_norm": 1.7702552779136205, | |
| "learning_rate": 8.189045788018502e-06, | |
| "loss": 0.141, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.5623342175066313, | |
| "grad_norm": 2.0515294131254525, | |
| "learning_rate": 8.172972813571132e-06, | |
| "loss": 0.1527, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.5649867374005305, | |
| "grad_norm": 2.0940031140217865, | |
| "learning_rate": 8.156844755465357e-06, | |
| "loss": 0.15, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.5676392572944297, | |
| "grad_norm": 1.9660531752973018, | |
| "learning_rate": 8.14066189368859e-06, | |
| "loss": 0.1408, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.5702917771883289, | |
| "grad_norm": 2.1910231506350186, | |
| "learning_rate": 8.124424509179648e-06, | |
| "loss": 0.1463, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.5729442970822282, | |
| "grad_norm": 1.5430710287754699, | |
| "learning_rate": 8.108132883823878e-06, | |
| "loss": 0.1097, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.5755968169761273, | |
| "grad_norm": 1.8265329192409332, | |
| "learning_rate": 8.091787300448264e-06, | |
| "loss": 0.1395, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.5782493368700266, | |
| "grad_norm": 1.8822040988497644, | |
| "learning_rate": 8.07538804281651e-06, | |
| "loss": 0.1298, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.5809018567639257, | |
| "grad_norm": 2.0002279443157445, | |
| "learning_rate": 8.058935395624128e-06, | |
| "loss": 0.1421, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.583554376657825, | |
| "grad_norm": 2.1521750298282964, | |
| "learning_rate": 8.042429644493479e-06, | |
| "loss": 0.1379, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.5862068965517241, | |
| "grad_norm": 1.5640533332285234, | |
| "learning_rate": 8.025871075968828e-06, | |
| "loss": 0.1228, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.5888594164456233, | |
| "grad_norm": 2.0211033096735864, | |
| "learning_rate": 8.00925997751136e-06, | |
| "loss": 0.1402, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.5915119363395226, | |
| "grad_norm": 1.7401492380650114, | |
| "learning_rate": 7.992596637494199e-06, | |
| "loss": 0.1223, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.5941644562334217, | |
| "grad_norm": 1.5482263543431414, | |
| "learning_rate": 7.975881345197394e-06, | |
| "loss": 0.1278, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.596816976127321, | |
| "grad_norm": 2.012786559192079, | |
| "learning_rate": 7.959114390802894e-06, | |
| "loss": 0.1322, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.5994694960212201, | |
| "grad_norm": 1.9101656055747547, | |
| "learning_rate": 7.942296065389528e-06, | |
| "loss": 0.1411, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.6021220159151194, | |
| "grad_norm": 1.8476462070875777, | |
| "learning_rate": 7.925426660927926e-06, | |
| "loss": 0.1325, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.6047745358090185, | |
| "grad_norm": 1.9254949919992999, | |
| "learning_rate": 7.908506470275474e-06, | |
| "loss": 0.1226, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.6074270557029178, | |
| "grad_norm": 2.0758812690239052, | |
| "learning_rate": 7.891535787171216e-06, | |
| "loss": 0.1263, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.610079575596817, | |
| "grad_norm": 1.9368560259395753, | |
| "learning_rate": 7.874514906230757e-06, | |
| "loss": 0.1308, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.6127320954907162, | |
| "grad_norm": 1.7832950240034287, | |
| "learning_rate": 7.857444122941155e-06, | |
| "loss": 0.1229, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 1.9284054045996681, | |
| "learning_rate": 7.84032373365578e-06, | |
| "loss": 0.1315, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.6180371352785146, | |
| "grad_norm": 1.8237229630031635, | |
| "learning_rate": 7.82315403558918e-06, | |
| "loss": 0.1263, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.6206896551724138, | |
| "grad_norm": 2.3980431877447552, | |
| "learning_rate": 7.805935326811913e-06, | |
| "loss": 0.1594, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.623342175066313, | |
| "grad_norm": 1.908328991517894, | |
| "learning_rate": 7.78866790624538e-06, | |
| "loss": 0.1221, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.6259946949602122, | |
| "grad_norm": 1.9070140557762951, | |
| "learning_rate": 7.771352073656628e-06, | |
| "loss": 0.1099, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.6286472148541115, | |
| "grad_norm": 1.9086850445705497, | |
| "learning_rate": 7.753988129653152e-06, | |
| "loss": 0.1217, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.6312997347480106, | |
| "grad_norm": 2.298658640053533, | |
| "learning_rate": 7.736576375677676e-06, | |
| "loss": 0.1534, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.6339522546419099, | |
| "grad_norm": 2.012272147602231, | |
| "learning_rate": 7.719117114002912e-06, | |
| "loss": 0.1367, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.636604774535809, | |
| "grad_norm": 2.054870475879218, | |
| "learning_rate": 7.701610647726323e-06, | |
| "loss": 0.1528, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.6392572944297082, | |
| "grad_norm": 1.931987531988978, | |
| "learning_rate": 7.684057280764855e-06, | |
| "loss": 0.1359, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.6419098143236074, | |
| "grad_norm": 1.6653702816540303, | |
| "learning_rate": 7.666457317849663e-06, | |
| "loss": 0.1271, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.6445623342175066, | |
| "grad_norm": 1.8059464643897476, | |
| "learning_rate": 7.648811064520821e-06, | |
| "loss": 0.1355, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.6472148541114059, | |
| "grad_norm": 1.8078224433971741, | |
| "learning_rate": 7.631118827122013e-06, | |
| "loss": 0.1202, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.649867374005305, | |
| "grad_norm": 1.948459636065191, | |
| "learning_rate": 7.613380912795225e-06, | |
| "loss": 0.1429, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.6525198938992043, | |
| "grad_norm": 2.2076773590525316, | |
| "learning_rate": 7.595597629475402e-06, | |
| "loss": 0.1516, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.6551724137931034, | |
| "grad_norm": 1.8887891109191988, | |
| "learning_rate": 7.57776928588511e-06, | |
| "loss": 0.1334, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.6578249336870027, | |
| "grad_norm": 2.3599876519640297, | |
| "learning_rate": 7.559896191529169e-06, | |
| "loss": 0.1523, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.6604774535809018, | |
| "grad_norm": 1.6073366184090068, | |
| "learning_rate": 7.54197865668929e-06, | |
| "loss": 0.1094, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.6631299734748011, | |
| "grad_norm": 1.6066208266739346, | |
| "learning_rate": 7.524016992418676e-06, | |
| "loss": 0.1133, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.6657824933687002, | |
| "grad_norm": 2.402024874117795, | |
| "learning_rate": 7.506011510536635e-06, | |
| "loss": 0.1683, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.6684350132625995, | |
| "grad_norm": 2.4880666218264955, | |
| "learning_rate": 7.487962523623159e-06, | |
| "loss": 0.1604, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.6710875331564987, | |
| "grad_norm": 2.1970208087255076, | |
| "learning_rate": 7.469870345013495e-06, | |
| "loss": 0.1261, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.6737400530503979, | |
| "grad_norm": 1.949833364046787, | |
| "learning_rate": 7.451735288792716e-06, | |
| "loss": 0.1283, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.6763925729442971, | |
| "grad_norm": 2.043764786717034, | |
| "learning_rate": 7.4335576697902546e-06, | |
| "loss": 0.1217, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.6790450928381963, | |
| "grad_norm": 1.7800715009896282, | |
| "learning_rate": 7.415337803574449e-06, | |
| "loss": 0.1134, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.6816976127320955, | |
| "grad_norm": 2.036030495974864, | |
| "learning_rate": 7.3970760064470634e-06, | |
| "loss": 0.1346, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.6843501326259946, | |
| "grad_norm": 1.7953348831678584, | |
| "learning_rate": 7.378772595437785e-06, | |
| "loss": 0.1068, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.6870026525198939, | |
| "grad_norm": 2.151685228258153, | |
| "learning_rate": 7.360427888298737e-06, | |
| "loss": 0.1374, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.6896551724137931, | |
| "grad_norm": 1.6830864009549993, | |
| "learning_rate": 7.342042203498952e-06, | |
| "loss": 0.0968, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.6923076923076923, | |
| "grad_norm": 2.0738297934683545, | |
| "learning_rate": 7.323615860218844e-06, | |
| "loss": 0.1274, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.6949602122015915, | |
| "grad_norm": 2.416943322684093, | |
| "learning_rate": 7.3051491783446705e-06, | |
| "loss": 0.1395, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.6976127320954907, | |
| "grad_norm": 1.8473933982642847, | |
| "learning_rate": 7.2866424784629806e-06, | |
| "loss": 0.1189, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.7002652519893899, | |
| "grad_norm": 1.990601440702459, | |
| "learning_rate": 7.26809608185504e-06, | |
| "loss": 0.1202, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.7029177718832891, | |
| "grad_norm": 1.904814403012209, | |
| "learning_rate": 7.249510310491268e-06, | |
| "loss": 0.1208, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.7055702917771883, | |
| "grad_norm": 1.8589441949655492, | |
| "learning_rate": 7.230885487025635e-06, | |
| "loss": 0.1181, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.7082228116710876, | |
| "grad_norm": 1.6795096831283485, | |
| "learning_rate": 7.212221934790067e-06, | |
| "loss": 0.1192, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.7108753315649867, | |
| "grad_norm": 1.723252840070737, | |
| "learning_rate": 7.193519977788834e-06, | |
| "loss": 0.1236, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.713527851458886, | |
| "grad_norm": 2.418801413723791, | |
| "learning_rate": 7.174779940692922e-06, | |
| "loss": 0.1544, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.7161803713527851, | |
| "grad_norm": 1.770338034548602, | |
| "learning_rate": 7.1560021488343956e-06, | |
| "loss": 0.1286, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.7188328912466844, | |
| "grad_norm": 1.6417602409918115, | |
| "learning_rate": 7.1371869282007545e-06, | |
| "loss": 0.1142, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.7214854111405835, | |
| "grad_norm": 2.2143362328692207, | |
| "learning_rate": 7.118334605429272e-06, | |
| "loss": 0.1542, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.7241379310344828, | |
| "grad_norm": 2.316061054037307, | |
| "learning_rate": 7.099445507801324e-06, | |
| "loss": 0.1376, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.726790450928382, | |
| "grad_norm": 2.184752996333078, | |
| "learning_rate": 7.080519963236706e-06, | |
| "loss": 0.1268, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.7294429708222812, | |
| "grad_norm": 1.840884510423046, | |
| "learning_rate": 7.0615583002879465e-06, | |
| "loss": 0.1346, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.7320954907161804, | |
| "grad_norm": 1.811306716138802, | |
| "learning_rate": 7.042560848134592e-06, | |
| "loss": 0.1149, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.7347480106100795, | |
| "grad_norm": 2.0428053121611476, | |
| "learning_rate": 7.023527936577507e-06, | |
| "loss": 0.1407, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.7374005305039788, | |
| "grad_norm": 1.6116834909252478, | |
| "learning_rate": 7.004459896033137e-06, | |
| "loss": 0.1193, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.7400530503978779, | |
| "grad_norm": 2.1213770604701976, | |
| "learning_rate": 6.985357057527774e-06, | |
| "loss": 0.1434, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.7427055702917772, | |
| "grad_norm": 1.91726573137005, | |
| "learning_rate": 6.966219752691814e-06, | |
| "loss": 0.125, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.7453580901856764, | |
| "grad_norm": 1.67695736254988, | |
| "learning_rate": 6.947048313753998e-06, | |
| "loss": 0.1338, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.7480106100795756, | |
| "grad_norm": 2.489354393443295, | |
| "learning_rate": 6.927843073535645e-06, | |
| "loss": 0.1473, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.7506631299734748, | |
| "grad_norm": 1.7295592511699343, | |
| "learning_rate": 6.9086043654448734e-06, | |
| "loss": 0.1256, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.753315649867374, | |
| "grad_norm": 1.9395856232269375, | |
| "learning_rate": 6.889332523470808e-06, | |
| "loss": 0.1195, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.7559681697612732, | |
| "grad_norm": 2.066706586377597, | |
| "learning_rate": 6.870027882177791e-06, | |
| "loss": 0.1358, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.7586206896551724, | |
| "grad_norm": 2.0770046146432395, | |
| "learning_rate": 6.850690776699574e-06, | |
| "loss": 0.1429, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.7612732095490716, | |
| "grad_norm": 1.881483786265112, | |
| "learning_rate": 6.831321542733482e-06, | |
| "loss": 0.1266, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.7639257294429708, | |
| "grad_norm": 1.695062472576317, | |
| "learning_rate": 6.811920516534616e-06, | |
| "loss": 0.1202, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.76657824933687, | |
| "grad_norm": 2.1054959851607506, | |
| "learning_rate": 6.7924880349099855e-06, | |
| "loss": 0.137, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 1.559652722034954, | |
| "learning_rate": 6.773024435212678e-06, | |
| "loss": 0.0985, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.7718832891246684, | |
| "grad_norm": 2.2208427457963342, | |
| "learning_rate": 6.753530055336006e-06, | |
| "loss": 0.1466, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.7745358090185677, | |
| "grad_norm": 2.1893216240483753, | |
| "learning_rate": 6.734005233707624e-06, | |
| "loss": 0.1537, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.7771883289124668, | |
| "grad_norm": 2.3583536185079703, | |
| "learning_rate": 6.714450309283671e-06, | |
| "loss": 0.1384, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.7798408488063661, | |
| "grad_norm": 2.116909607447533, | |
| "learning_rate": 6.694865621542873e-06, | |
| "loss": 0.1525, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.7824933687002652, | |
| "grad_norm": 1.553790399379545, | |
| "learning_rate": 6.675251510480662e-06, | |
| "loss": 0.1074, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.7851458885941645, | |
| "grad_norm": 1.702247544650182, | |
| "learning_rate": 6.655608316603257e-06, | |
| "loss": 0.1037, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.7877984084880637, | |
| "grad_norm": 1.4027176458517787, | |
| "learning_rate": 6.635936380921774e-06, | |
| "loss": 0.0842, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.7904509283819628, | |
| "grad_norm": 1.8615502107309188, | |
| "learning_rate": 6.616236044946283e-06, | |
| "loss": 0.1335, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.7931034482758621, | |
| "grad_norm": 1.6648222471112546, | |
| "learning_rate": 6.5965076506799e-06, | |
| "loss": 0.1181, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.7957559681697612, | |
| "grad_norm": 1.9881029740135223, | |
| "learning_rate": 6.576751540612835e-06, | |
| "loss": 0.1112, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7984084880636605, | |
| "grad_norm": 2.2013679784949667, | |
| "learning_rate": 6.556968057716457e-06, | |
| "loss": 0.1237, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.8010610079575596, | |
| "grad_norm": 2.1049495467849697, | |
| "learning_rate": 6.537157545437326e-06, | |
| "loss": 0.14, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.8037135278514589, | |
| "grad_norm": 1.9391684954075283, | |
| "learning_rate": 6.517320347691245e-06, | |
| "loss": 0.1432, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.8063660477453581, | |
| "grad_norm": 1.897903152535271, | |
| "learning_rate": 6.497456808857286e-06, | |
| "loss": 0.1242, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.8090185676392573, | |
| "grad_norm": 2.400907859514575, | |
| "learning_rate": 6.477567273771807e-06, | |
| "loss": 0.1363, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.8116710875331565, | |
| "grad_norm": 2.0519874786243095, | |
| "learning_rate": 6.4576520877224644e-06, | |
| "loss": 0.149, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.8143236074270557, | |
| "grad_norm": 1.613637672772973, | |
| "learning_rate": 6.437711596442228e-06, | |
| "loss": 0.1078, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.8169761273209549, | |
| "grad_norm": 2.0653398136129675, | |
| "learning_rate": 6.4177461461033675e-06, | |
| "loss": 0.1416, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.8196286472148541, | |
| "grad_norm": 2.2371819869165073, | |
| "learning_rate": 6.397756083311454e-06, | |
| "loss": 0.1395, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.8222811671087533, | |
| "grad_norm": 1.7682314840535782, | |
| "learning_rate": 6.377741755099334e-06, | |
| "loss": 0.1248, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.8249336870026526, | |
| "grad_norm": 1.7180875056368419, | |
| "learning_rate": 6.357703508921109e-06, | |
| "loss": 0.1146, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.8275862068965517, | |
| "grad_norm": 1.837539966421522, | |
| "learning_rate": 6.337641692646106e-06, | |
| "loss": 0.1063, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.830238726790451, | |
| "grad_norm": 1.9061111225956069, | |
| "learning_rate": 6.317556654552825e-06, | |
| "loss": 0.1261, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.8328912466843501, | |
| "grad_norm": 1.7785361235235786, | |
| "learning_rate": 6.297448743322918e-06, | |
| "loss": 0.1213, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.8355437665782494, | |
| "grad_norm": 2.285655601737043, | |
| "learning_rate": 6.277318308035109e-06, | |
| "loss": 0.136, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.8381962864721485, | |
| "grad_norm": 2.292822876847262, | |
| "learning_rate": 6.257165698159149e-06, | |
| "loss": 0.1437, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.8408488063660478, | |
| "grad_norm": 1.925635900709222, | |
| "learning_rate": 6.236991263549748e-06, | |
| "loss": 0.1244, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.843501326259947, | |
| "grad_norm": 1.8764634867307084, | |
| "learning_rate": 6.2167953544404955e-06, | |
| "loss": 0.1269, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.8461538461538461, | |
| "grad_norm": 1.9227813803757845, | |
| "learning_rate": 6.1965783214377895e-06, | |
| "loss": 0.1052, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.8488063660477454, | |
| "grad_norm": 1.6243389871471685, | |
| "learning_rate": 6.176340515514738e-06, | |
| "loss": 0.0946, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.8514588859416445, | |
| "grad_norm": 1.9721230457586518, | |
| "learning_rate": 6.156082288005078e-06, | |
| "loss": 0.1242, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.8541114058355438, | |
| "grad_norm": 1.939736016873224, | |
| "learning_rate": 6.135803990597066e-06, | |
| "loss": 0.1107, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.8567639257294429, | |
| "grad_norm": 1.910794512321765, | |
| "learning_rate": 6.115505975327382e-06, | |
| "loss": 0.1157, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.8594164456233422, | |
| "grad_norm": 2.3310014277848006, | |
| "learning_rate": 6.095188594575008e-06, | |
| "loss": 0.1473, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.8620689655172413, | |
| "grad_norm": 2.021424856899511, | |
| "learning_rate": 6.074852201055121e-06, | |
| "loss": 0.1412, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.8647214854111406, | |
| "grad_norm": 1.8755075770524758, | |
| "learning_rate": 6.054497147812962e-06, | |
| "loss": 0.1374, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.8673740053050398, | |
| "grad_norm": 2.0283809132037947, | |
| "learning_rate": 6.034123788217712e-06, | |
| "loss": 0.1225, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.870026525198939, | |
| "grad_norm": 1.77815086557222, | |
| "learning_rate": 6.013732475956352e-06, | |
| "loss": 0.1105, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.8726790450928382, | |
| "grad_norm": 1.698228853578335, | |
| "learning_rate": 5.993323565027528e-06, | |
| "loss": 0.0964, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.8753315649867374, | |
| "grad_norm": 1.8428502419687727, | |
| "learning_rate": 5.972897409735403e-06, | |
| "loss": 0.1227, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.8779840848806366, | |
| "grad_norm": 1.7797504117579037, | |
| "learning_rate": 5.952454364683507e-06, | |
| "loss": 0.1129, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.8806366047745358, | |
| "grad_norm": 2.5588750953515143, | |
| "learning_rate": 5.931994784768582e-06, | |
| "loss": 0.1512, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.883289124668435, | |
| "grad_norm": 2.4302985612966697, | |
| "learning_rate": 5.911519025174419e-06, | |
| "loss": 0.1419, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.8859416445623343, | |
| "grad_norm": 2.116715417993045, | |
| "learning_rate": 5.891027441365689e-06, | |
| "loss": 0.1248, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.8885941644562334, | |
| "grad_norm": 1.8450184290705178, | |
| "learning_rate": 5.870520389081782e-06, | |
| "loss": 0.1093, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.8912466843501327, | |
| "grad_norm": 1.8378061965275745, | |
| "learning_rate": 5.849998224330621e-06, | |
| "loss": 0.1103, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.8938992042440318, | |
| "grad_norm": 2.0712714287437963, | |
| "learning_rate": 5.829461303382484e-06, | |
| "loss": 0.1318, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.896551724137931, | |
| "grad_norm": 2.1172154065660034, | |
| "learning_rate": 5.808909982763825e-06, | |
| "loss": 0.1185, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.8992042440318302, | |
| "grad_norm": 2.600554195545317, | |
| "learning_rate": 5.788344619251076e-06, | |
| "loss": 0.1462, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.9018567639257294, | |
| "grad_norm": 2.08199632146431, | |
| "learning_rate": 5.767765569864459e-06, | |
| "loss": 0.1383, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.9045092838196287, | |
| "grad_norm": 1.806422329448522, | |
| "learning_rate": 5.747173191861788e-06, | |
| "loss": 0.0905, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.9071618037135278, | |
| "grad_norm": 1.8050690831990468, | |
| "learning_rate": 5.726567842732262e-06, | |
| "loss": 0.1156, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.9098143236074271, | |
| "grad_norm": 2.091128781060801, | |
| "learning_rate": 5.705949880190266e-06, | |
| "loss": 0.1195, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.9124668435013262, | |
| "grad_norm": 1.8670275726747676, | |
| "learning_rate": 5.685319662169157e-06, | |
| "loss": 0.1029, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.9151193633952255, | |
| "grad_norm": 2.039989303735285, | |
| "learning_rate": 5.664677546815043e-06, | |
| "loss": 0.1312, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.9177718832891246, | |
| "grad_norm": 1.859016569528762, | |
| "learning_rate": 5.644023892480583e-06, | |
| "loss": 0.1055, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.9204244031830239, | |
| "grad_norm": 1.5903154857571018, | |
| "learning_rate": 5.623359057718752e-06, | |
| "loss": 0.0973, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 1.9024912783079182, | |
| "learning_rate": 5.6026834012766155e-06, | |
| "loss": 0.1147, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.9257294429708223, | |
| "grad_norm": 2.4600571478007645, | |
| "learning_rate": 5.581997282089114e-06, | |
| "loss": 0.1274, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.9283819628647215, | |
| "grad_norm": 2.1252658277426493, | |
| "learning_rate": 5.561301059272821e-06, | |
| "loss": 0.1192, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.9310344827586207, | |
| "grad_norm": 2.2663754831446528, | |
| "learning_rate": 5.540595092119709e-06, | |
| "loss": 0.1368, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.9336870026525199, | |
| "grad_norm": 2.1656127015955637, | |
| "learning_rate": 5.519879740090918e-06, | |
| "loss": 0.1072, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.9363395225464191, | |
| "grad_norm": 2.1805472380714552, | |
| "learning_rate": 5.499155362810512e-06, | |
| "loss": 0.1475, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.9389920424403183, | |
| "grad_norm": 1.9313395207402062, | |
| "learning_rate": 5.478422320059231e-06, | |
| "loss": 0.1138, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.9416445623342176, | |
| "grad_norm": 1.9593835652421479, | |
| "learning_rate": 5.457680971768258e-06, | |
| "loss": 0.1236, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.9442970822281167, | |
| "grad_norm": 1.9523449371463568, | |
| "learning_rate": 5.436931678012956e-06, | |
| "loss": 0.1191, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.946949602122016, | |
| "grad_norm": 2.213437974106433, | |
| "learning_rate": 5.4161747990066235e-06, | |
| "loss": 0.157, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.9496021220159151, | |
| "grad_norm": 1.7830925038957812, | |
| "learning_rate": 5.395410695094246e-06, | |
| "loss": 0.1048, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.9522546419098143, | |
| "grad_norm": 1.9526782766464377, | |
| "learning_rate": 5.374639726746232e-06, | |
| "loss": 0.1298, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.9549071618037135, | |
| "grad_norm": 2.225209885017297, | |
| "learning_rate": 5.353862254552159e-06, | |
| "loss": 0.1114, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.9575596816976127, | |
| "grad_norm": 1.8717932182864292, | |
| "learning_rate": 5.333078639214511e-06, | |
| "loss": 0.114, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.9602122015915119, | |
| "grad_norm": 2.037766974038598, | |
| "learning_rate": 5.31228924154242e-06, | |
| "loss": 0.1061, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.9628647214854111, | |
| "grad_norm": 1.7551149566575617, | |
| "learning_rate": 5.2914944224454e-06, | |
| "loss": 0.1105, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.9655172413793104, | |
| "grad_norm": 1.6746920582972595, | |
| "learning_rate": 5.270694542927089e-06, | |
| "loss": 0.1129, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.9681697612732095, | |
| "grad_norm": 1.7421649434309925, | |
| "learning_rate": 5.249889964078965e-06, | |
| "loss": 0.1095, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.9708222811671088, | |
| "grad_norm": 1.705953446601723, | |
| "learning_rate": 5.2290810470740925e-06, | |
| "loss": 0.1009, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.9734748010610079, | |
| "grad_norm": 1.9937643181748503, | |
| "learning_rate": 5.2082681531608505e-06, | |
| "loss": 0.1374, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.9761273209549072, | |
| "grad_norm": 2.173137907211324, | |
| "learning_rate": 5.187451643656654e-06, | |
| "loss": 0.1279, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.9787798408488063, | |
| "grad_norm": 1.8358057318165615, | |
| "learning_rate": 5.166631879941686e-06, | |
| "loss": 0.1032, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.9814323607427056, | |
| "grad_norm": 2.097800711750244, | |
| "learning_rate": 5.145809223452625e-06, | |
| "loss": 0.119, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.9840848806366048, | |
| "grad_norm": 1.7975003014708764, | |
| "learning_rate": 5.124984035676366e-06, | |
| "loss": 0.0984, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.986737400530504, | |
| "grad_norm": 1.6452454858458356, | |
| "learning_rate": 5.1041566781437525e-06, | |
| "loss": 0.0995, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.9893899204244032, | |
| "grad_norm": 1.6148448625331262, | |
| "learning_rate": 5.083327512423294e-06, | |
| "loss": 0.087, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.9920424403183024, | |
| "grad_norm": 1.8548088873299826, | |
| "learning_rate": 5.062496900114887e-06, | |
| "loss": 0.1028, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.9946949602122016, | |
| "grad_norm": 2.286016586242463, | |
| "learning_rate": 5.041665202843543e-06, | |
| "loss": 0.1394, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.9973474801061007, | |
| "grad_norm": 1.9248524614602884, | |
| "learning_rate": 5.020832782253115e-06, | |
| "loss": 0.1206, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.7947903575090804, | |
| "learning_rate": 5e-06, | |
| "loss": 0.108, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 1.0026525198938991, | |
| "grad_norm": 1.1448965437517742, | |
| "learning_rate": 4.979167217746888e-06, | |
| "loss": 0.0474, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 1.0053050397877985, | |
| "grad_norm": 1.4782795600036578, | |
| "learning_rate": 4.9583347971564575e-06, | |
| "loss": 0.0546, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 1.0079575596816976, | |
| "grad_norm": 1.4977052145475058, | |
| "learning_rate": 4.937503099885115e-06, | |
| "loss": 0.0527, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.0106100795755968, | |
| "grad_norm": 1.3509330137945688, | |
| "learning_rate": 4.916672487576708e-06, | |
| "loss": 0.0581, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 1.013262599469496, | |
| "grad_norm": 1.3766430452266287, | |
| "learning_rate": 4.895843321856249e-06, | |
| "loss": 0.0543, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 1.0159151193633953, | |
| "grad_norm": 1.1838930001612804, | |
| "learning_rate": 4.875015964323635e-06, | |
| "loss": 0.0501, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 1.0185676392572944, | |
| "grad_norm": 1.1327638132836573, | |
| "learning_rate": 4.854190776547377e-06, | |
| "loss": 0.0488, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 1.0212201591511936, | |
| "grad_norm": 1.3242917574913802, | |
| "learning_rate": 4.833368120058317e-06, | |
| "loss": 0.0528, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.023872679045093, | |
| "grad_norm": 1.3957886304337215, | |
| "learning_rate": 4.812548356343347e-06, | |
| "loss": 0.0435, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 1.026525198938992, | |
| "grad_norm": 1.3976646283902743, | |
| "learning_rate": 4.79173184683915e-06, | |
| "loss": 0.0504, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 1.0291777188328912, | |
| "grad_norm": 1.594064921848588, | |
| "learning_rate": 4.770918952925908e-06, | |
| "loss": 0.0467, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 1.0318302387267904, | |
| "grad_norm": 1.8424058890187611, | |
| "learning_rate": 4.750110035921038e-06, | |
| "loss": 0.0592, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 1.0344827586206897, | |
| "grad_norm": 1.5198759678201303, | |
| "learning_rate": 4.729305457072913e-06, | |
| "loss": 0.0493, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.0371352785145889, | |
| "grad_norm": 1.63737016645813, | |
| "learning_rate": 4.708505577554601e-06, | |
| "loss": 0.0637, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 1.039787798408488, | |
| "grad_norm": 1.513812328618622, | |
| "learning_rate": 4.687710758457583e-06, | |
| "loss": 0.0452, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 1.0424403183023874, | |
| "grad_norm": 1.6651422093934103, | |
| "learning_rate": 4.6669213607854915e-06, | |
| "loss": 0.0498, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 1.0450928381962865, | |
| "grad_norm": 1.5964261962470236, | |
| "learning_rate": 4.646137745447843e-06, | |
| "loss": 0.0623, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 1.0477453580901857, | |
| "grad_norm": 1.7129301697378825, | |
| "learning_rate": 4.6253602732537685e-06, | |
| "loss": 0.0455, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.0503978779840848, | |
| "grad_norm": 1.5300658689026623, | |
| "learning_rate": 4.6045893049057544e-06, | |
| "loss": 0.0555, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 1.0530503978779842, | |
| "grad_norm": 1.3274779219248138, | |
| "learning_rate": 4.583825200993377e-06, | |
| "loss": 0.0461, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 1.0557029177718833, | |
| "grad_norm": 1.6225406314000637, | |
| "learning_rate": 4.563068321987047e-06, | |
| "loss": 0.0531, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 1.0583554376657824, | |
| "grad_norm": 1.5953486175793383, | |
| "learning_rate": 4.542319028231744e-06, | |
| "loss": 0.0477, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 1.0610079575596818, | |
| "grad_norm": 1.5805509793613008, | |
| "learning_rate": 4.521577679940769e-06, | |
| "loss": 0.0429, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.0610079575596818, | |
| "eval_loss": 0.1300889253616333, | |
| "eval_runtime": 1.2993, | |
| "eval_samples_per_second": 23.859, | |
| "eval_steps_per_second": 6.157, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.063660477453581, | |
| "grad_norm": 1.6980952582384556, | |
| "learning_rate": 4.50084463718949e-06, | |
| "loss": 0.06, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 1.06631299734748, | |
| "grad_norm": 1.5737277597483659, | |
| "learning_rate": 4.480120259909084e-06, | |
| "loss": 0.0488, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 1.0689655172413792, | |
| "grad_norm": 1.6771874133081281, | |
| "learning_rate": 4.459404907880293e-06, | |
| "loss": 0.0539, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 1.0716180371352786, | |
| "grad_norm": 1.6529014957836754, | |
| "learning_rate": 4.438698940727179e-06, | |
| "loss": 0.0572, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 1.0742705570291777, | |
| "grad_norm": 1.5970118458627822, | |
| "learning_rate": 4.418002717910887e-06, | |
| "loss": 0.06, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.0769230769230769, | |
| "grad_norm": 1.4471958707171573, | |
| "learning_rate": 4.397316598723385e-06, | |
| "loss": 0.0538, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 1.079575596816976, | |
| "grad_norm": 1.0348681849940542, | |
| "learning_rate": 4.37664094228125e-06, | |
| "loss": 0.035, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 1.0822281167108754, | |
| "grad_norm": 1.3544554476316786, | |
| "learning_rate": 4.3559761075194185e-06, | |
| "loss": 0.0417, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 1.0848806366047745, | |
| "grad_norm": 1.9097492769272106, | |
| "learning_rate": 4.335322453184959e-06, | |
| "loss": 0.0564, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 1.0875331564986737, | |
| "grad_norm": 1.337428576441879, | |
| "learning_rate": 4.314680337830847e-06, | |
| "loss": 0.0471, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.090185676392573, | |
| "grad_norm": 1.6610870125693307, | |
| "learning_rate": 4.294050119809735e-06, | |
| "loss": 0.0495, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 1.0928381962864722, | |
| "grad_norm": 1.648323552864328, | |
| "learning_rate": 4.273432157267739e-06, | |
| "loss": 0.0541, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 1.0954907161803713, | |
| "grad_norm": 1.5486084688969948, | |
| "learning_rate": 4.252826808138214e-06, | |
| "loss": 0.0559, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 1.0981432360742707, | |
| "grad_norm": 1.7853234719100994, | |
| "learning_rate": 4.232234430135542e-06, | |
| "loss": 0.0688, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 1.1007957559681698, | |
| "grad_norm": 1.5356971600614069, | |
| "learning_rate": 4.2116553807489255e-06, | |
| "loss": 0.0516, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.103448275862069, | |
| "grad_norm": 1.6247663576609628, | |
| "learning_rate": 4.191090017236177e-06, | |
| "loss": 0.0514, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 1.106100795755968, | |
| "grad_norm": 1.534189806530592, | |
| "learning_rate": 4.170538696617518e-06, | |
| "loss": 0.0524, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 1.1087533156498675, | |
| "grad_norm": 1.3158821951034483, | |
| "learning_rate": 4.15000177566938e-06, | |
| "loss": 0.0403, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 1.1114058355437666, | |
| "grad_norm": 1.5138948007814246, | |
| "learning_rate": 4.129479610918219e-06, | |
| "loss": 0.049, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 1.1140583554376657, | |
| "grad_norm": 1.2320891847094047, | |
| "learning_rate": 4.108972558634312e-06, | |
| "loss": 0.042, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.1167108753315649, | |
| "grad_norm": 1.6729860721769905, | |
| "learning_rate": 4.088480974825584e-06, | |
| "loss": 0.0527, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 1.1193633952254642, | |
| "grad_norm": 2.128191222629532, | |
| "learning_rate": 4.0680052152314185e-06, | |
| "loss": 0.064, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 1.1220159151193634, | |
| "grad_norm": 1.5415156129118721, | |
| "learning_rate": 4.047545635316494e-06, | |
| "loss": 0.0479, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 1.1246684350132625, | |
| "grad_norm": 1.4478769280496202, | |
| "learning_rate": 4.0271025902646e-06, | |
| "loss": 0.0461, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 1.1273209549071619, | |
| "grad_norm": 1.4168078919989058, | |
| "learning_rate": 4.006676434972474e-06, | |
| "loss": 0.0461, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.129973474801061, | |
| "grad_norm": 1.7163680394879697, | |
| "learning_rate": 3.98626752404365e-06, | |
| "loss": 0.0563, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 1.1326259946949602, | |
| "grad_norm": 1.4561458702273795, | |
| "learning_rate": 3.96587621178229e-06, | |
| "loss": 0.0451, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 1.1352785145888595, | |
| "grad_norm": 1.3054592699914147, | |
| "learning_rate": 3.94550285218704e-06, | |
| "loss": 0.0413, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 1.1379310344827587, | |
| "grad_norm": 1.410529232372516, | |
| "learning_rate": 3.92514779894488e-06, | |
| "loss": 0.0475, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 1.1405835543766578, | |
| "grad_norm": 1.1401984741886695, | |
| "learning_rate": 3.904811405424993e-06, | |
| "loss": 0.034, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.143236074270557, | |
| "grad_norm": 1.677403926891428, | |
| "learning_rate": 3.8844940246726206e-06, | |
| "loss": 0.0537, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 1.1458885941644563, | |
| "grad_norm": 2.0010687043565856, | |
| "learning_rate": 3.864196009402935e-06, | |
| "loss": 0.0571, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 1.1485411140583555, | |
| "grad_norm": 1.4276836740813328, | |
| "learning_rate": 3.843917711994923e-06, | |
| "loss": 0.039, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 1.1511936339522546, | |
| "grad_norm": 1.7186301264192605, | |
| "learning_rate": 3.823659484485264e-06, | |
| "loss": 0.0566, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 1.1538461538461537, | |
| "grad_norm": 1.6127857163652946, | |
| "learning_rate": 3.803421678562213e-06, | |
| "loss": 0.0392, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.156498673740053, | |
| "grad_norm": 1.6060058035757456, | |
| "learning_rate": 3.783204645559504e-06, | |
| "loss": 0.058, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 1.1591511936339522, | |
| "grad_norm": 1.8031413344908938, | |
| "learning_rate": 3.7630087364502545e-06, | |
| "loss": 0.0565, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 1.1618037135278514, | |
| "grad_norm": 1.3528132025697588, | |
| "learning_rate": 3.742834301840853e-06, | |
| "loss": 0.0488, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 1.1644562334217508, | |
| "grad_norm": 1.4452161790201725, | |
| "learning_rate": 3.722681691964892e-06, | |
| "loss": 0.052, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 1.16710875331565, | |
| "grad_norm": 1.506462400743829, | |
| "learning_rate": 3.702551256677083e-06, | |
| "loss": 0.0517, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.169761273209549, | |
| "grad_norm": 1.4482316284963157, | |
| "learning_rate": 3.6824433454471755e-06, | |
| "loss": 0.0486, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 1.1724137931034484, | |
| "grad_norm": 1.4651862565906717, | |
| "learning_rate": 3.662358307353897e-06, | |
| "loss": 0.0528, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 1.1750663129973475, | |
| "grad_norm": 1.7861455263109471, | |
| "learning_rate": 3.6422964910788917e-06, | |
| "loss": 0.068, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 1.1777188328912467, | |
| "grad_norm": 1.8641910946820022, | |
| "learning_rate": 3.6222582449006673e-06, | |
| "loss": 0.0557, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 1.1803713527851458, | |
| "grad_norm": 1.493182090383967, | |
| "learning_rate": 3.602243916688548e-06, | |
| "loss": 0.0468, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.1830238726790452, | |
| "grad_norm": 1.458973918132782, | |
| "learning_rate": 3.5822538538966333e-06, | |
| "loss": 0.0546, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 1.1856763925729443, | |
| "grad_norm": 1.1386325113824567, | |
| "learning_rate": 3.5622884035577743e-06, | |
| "loss": 0.044, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 1.1883289124668435, | |
| "grad_norm": 1.8301480285009966, | |
| "learning_rate": 3.542347912277537e-06, | |
| "loss": 0.0532, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 1.1909814323607426, | |
| "grad_norm": 1.4956585692430135, | |
| "learning_rate": 3.5224327262281956e-06, | |
| "loss": 0.0583, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 1.193633952254642, | |
| "grad_norm": 1.5435072195981274, | |
| "learning_rate": 3.502543191142713e-06, | |
| "loss": 0.0527, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.1962864721485411, | |
| "grad_norm": 1.2942065638898985, | |
| "learning_rate": 3.482679652308756e-06, | |
| "loss": 0.0458, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 1.1989389920424403, | |
| "grad_norm": 1.5335416817545946, | |
| "learning_rate": 3.462842454562677e-06, | |
| "loss": 0.0435, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 1.2015915119363396, | |
| "grad_norm": 1.4166517541759323, | |
| "learning_rate": 3.443031942283544e-06, | |
| "loss": 0.0423, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 1.2042440318302388, | |
| "grad_norm": 1.8664420121487215, | |
| "learning_rate": 3.423248459387165e-06, | |
| "loss": 0.0546, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 1.206896551724138, | |
| "grad_norm": 1.3815730200823022, | |
| "learning_rate": 3.403492349320101e-06, | |
| "loss": 0.043, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.209549071618037, | |
| "grad_norm": 1.581543542046171, | |
| "learning_rate": 3.3837639550537183e-06, | |
| "loss": 0.0563, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 1.2122015915119364, | |
| "grad_norm": 1.420673121312355, | |
| "learning_rate": 3.364063619078228e-06, | |
| "loss": 0.0364, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 1.2148541114058355, | |
| "grad_norm": 1.4971596799850846, | |
| "learning_rate": 3.344391683396744e-06, | |
| "loss": 0.0446, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 1.2175066312997347, | |
| "grad_norm": 1.701718210938796, | |
| "learning_rate": 3.3247484895193406e-06, | |
| "loss": 0.052, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 1.2201591511936338, | |
| "grad_norm": 1.4921935610407966, | |
| "learning_rate": 3.305134378457127e-06, | |
| "loss": 0.046, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.2228116710875332, | |
| "grad_norm": 1.6030117015462066, | |
| "learning_rate": 3.2855496907163296e-06, | |
| "loss": 0.0472, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 1.2254641909814323, | |
| "grad_norm": 1.2527893295170924, | |
| "learning_rate": 3.2659947662923767e-06, | |
| "loss": 0.0443, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 1.2281167108753315, | |
| "grad_norm": 1.9885749044696137, | |
| "learning_rate": 3.246469944663996e-06, | |
| "loss": 0.0619, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 1.2307692307692308, | |
| "grad_norm": 1.4623016319017155, | |
| "learning_rate": 3.226975564787322e-06, | |
| "loss": 0.042, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 1.23342175066313, | |
| "grad_norm": 1.372530389209216, | |
| "learning_rate": 3.2075119650900166e-06, | |
| "loss": 0.0406, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.2360742705570291, | |
| "grad_norm": 1.3245791205337774, | |
| "learning_rate": 3.1880794834653872e-06, | |
| "loss": 0.0471, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 1.2387267904509285, | |
| "grad_norm": 1.765042349753763, | |
| "learning_rate": 3.1686784572665176e-06, | |
| "loss": 0.0545, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 1.2413793103448276, | |
| "grad_norm": 1.6880726450034362, | |
| "learning_rate": 3.149309223300428e-06, | |
| "loss": 0.0417, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 1.2440318302387268, | |
| "grad_norm": 1.6867768845931341, | |
| "learning_rate": 3.12997211782221e-06, | |
| "loss": 0.064, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 1.246684350132626, | |
| "grad_norm": 1.6829202873944922, | |
| "learning_rate": 3.1106674765291943e-06, | |
| "loss": 0.0487, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.2493368700265253, | |
| "grad_norm": 1.482635437530892, | |
| "learning_rate": 3.0913956345551287e-06, | |
| "loss": 0.0437, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 1.2519893899204244, | |
| "grad_norm": 1.0243355165306798, | |
| "learning_rate": 3.072156926464356e-06, | |
| "loss": 0.0286, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 1.2546419098143236, | |
| "grad_norm": 1.7086250346584169, | |
| "learning_rate": 3.052951686246003e-06, | |
| "loss": 0.0524, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 1.2572944297082227, | |
| "grad_norm": 1.5669584106420342, | |
| "learning_rate": 3.033780247308187e-06, | |
| "loss": 0.0416, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 1.259946949602122, | |
| "grad_norm": 1.5314026636512095, | |
| "learning_rate": 3.0146429424722277e-06, | |
| "loss": 0.0459, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.2625994694960212, | |
| "grad_norm": 1.6510097722189716, | |
| "learning_rate": 2.9955401039668642e-06, | |
| "loss": 0.0451, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 1.2652519893899203, | |
| "grad_norm": 1.1998735054502896, | |
| "learning_rate": 2.976472063422493e-06, | |
| "loss": 0.0421, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 1.2679045092838197, | |
| "grad_norm": 1.8230740287638594, | |
| "learning_rate": 2.9574391518654077e-06, | |
| "loss": 0.0483, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 1.2705570291777188, | |
| "grad_norm": 1.8679306651778758, | |
| "learning_rate": 2.938441699712055e-06, | |
| "loss": 0.0541, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 1.273209549071618, | |
| "grad_norm": 1.5955196355028194, | |
| "learning_rate": 2.9194800367632946e-06, | |
| "loss": 0.0558, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.2758620689655173, | |
| "grad_norm": 1.8476586232402528, | |
| "learning_rate": 2.9005544921986774e-06, | |
| "loss": 0.0564, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 1.2785145888594165, | |
| "grad_norm": 1.4298742037812726, | |
| "learning_rate": 2.8816653945707286e-06, | |
| "loss": 0.0442, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 1.2811671087533156, | |
| "grad_norm": 1.287640370036359, | |
| "learning_rate": 2.8628130717992463e-06, | |
| "loss": 0.0373, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 1.2838196286472148, | |
| "grad_norm": 1.4067433199098056, | |
| "learning_rate": 2.8439978511656057e-06, | |
| "loss": 0.045, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 1.2864721485411141, | |
| "grad_norm": 1.4224319432151542, | |
| "learning_rate": 2.82522005930708e-06, | |
| "loss": 0.0385, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.2891246684350133, | |
| "grad_norm": 1.5527398960311367, | |
| "learning_rate": 2.8064800222111673e-06, | |
| "loss": 0.0413, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 1.2917771883289124, | |
| "grad_norm": 1.4424002142583947, | |
| "learning_rate": 2.787778065209934e-06, | |
| "loss": 0.0398, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 1.2944297082228116, | |
| "grad_norm": 1.3110803703822922, | |
| "learning_rate": 2.7691145129743645e-06, | |
| "loss": 0.048, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 1.297082228116711, | |
| "grad_norm": 1.5121856468389003, | |
| "learning_rate": 2.7504896895087317e-06, | |
| "loss": 0.0414, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 1.29973474801061, | |
| "grad_norm": 1.7902702451543397, | |
| "learning_rate": 2.7319039181449604e-06, | |
| "loss": 0.0616, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.3023872679045092, | |
| "grad_norm": 1.834255575531623, | |
| "learning_rate": 2.713357521537023e-06, | |
| "loss": 0.0479, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 1.3050397877984086, | |
| "grad_norm": 1.3600003850482532, | |
| "learning_rate": 2.6948508216553304e-06, | |
| "loss": 0.042, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 1.3076923076923077, | |
| "grad_norm": 1.9237485999672874, | |
| "learning_rate": 2.6763841397811576e-06, | |
| "loss": 0.0462, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 1.3103448275862069, | |
| "grad_norm": 1.2574491259586364, | |
| "learning_rate": 2.65795779650105e-06, | |
| "loss": 0.0382, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 1.3129973474801062, | |
| "grad_norm": 1.8546812652508733, | |
| "learning_rate": 2.6395721117012648e-06, | |
| "loss": 0.0582, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.3156498673740054, | |
| "grad_norm": 1.4303414853405272, | |
| "learning_rate": 2.6212274045622167e-06, | |
| "loss": 0.0437, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 1.3183023872679045, | |
| "grad_norm": 1.3475863489270175, | |
| "learning_rate": 2.6029239935529395e-06, | |
| "loss": 0.0374, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 1.3209549071618036, | |
| "grad_norm": 1.3885729140928034, | |
| "learning_rate": 2.5846621964255524e-06, | |
| "loss": 0.0378, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 1.323607427055703, | |
| "grad_norm": 1.6004829128070286, | |
| "learning_rate": 2.5664423302097462e-06, | |
| "loss": 0.0456, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 1.3262599469496021, | |
| "grad_norm": 1.4483756915235027, | |
| "learning_rate": 2.5482647112072857e-06, | |
| "loss": 0.0483, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.3289124668435013, | |
| "grad_norm": 1.5083194734193153, | |
| "learning_rate": 2.530129654986505e-06, | |
| "loss": 0.0403, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 1.3315649867374004, | |
| "grad_norm": 1.0241725585916064, | |
| "learning_rate": 2.5120374763768422e-06, | |
| "loss": 0.0309, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 1.3342175066312998, | |
| "grad_norm": 1.4775342162666805, | |
| "learning_rate": 2.493988489463366e-06, | |
| "loss": 0.0449, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 1.336870026525199, | |
| "grad_norm": 1.453802729887559, | |
| "learning_rate": 2.475983007581326e-06, | |
| "loss": 0.0552, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 1.339522546419098, | |
| "grad_norm": 1.4761120060128954, | |
| "learning_rate": 2.458021343310713e-06, | |
| "loss": 0.0538, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.3421750663129974, | |
| "grad_norm": 1.4602044460085675, | |
| "learning_rate": 2.4401038084708313e-06, | |
| "loss": 0.0367, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 1.3448275862068966, | |
| "grad_norm": 1.1064714894292946, | |
| "learning_rate": 2.422230714114891e-06, | |
| "loss": 0.0318, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 1.3474801061007957, | |
| "grad_norm": 1.5984242375021522, | |
| "learning_rate": 2.4044023705246e-06, | |
| "loss": 0.0451, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 1.350132625994695, | |
| "grad_norm": 1.8076623418841897, | |
| "learning_rate": 2.3866190872047775e-06, | |
| "loss": 0.055, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 1.3527851458885942, | |
| "grad_norm": 1.53849591572484, | |
| "learning_rate": 2.3688811728779875e-06, | |
| "loss": 0.0413, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.3554376657824934, | |
| "grad_norm": 1.3925465600995954, | |
| "learning_rate": 2.351188935479181e-06, | |
| "loss": 0.0405, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 1.3580901856763925, | |
| "grad_norm": 1.4430777345183743, | |
| "learning_rate": 2.333542682150339e-06, | |
| "loss": 0.0418, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 1.3607427055702916, | |
| "grad_norm": 1.7058397273368744, | |
| "learning_rate": 2.3159427192351467e-06, | |
| "loss": 0.0417, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 1.363395225464191, | |
| "grad_norm": 1.6101298515299973, | |
| "learning_rate": 2.2983893522736795e-06, | |
| "loss": 0.0436, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 1.3660477453580901, | |
| "grad_norm": 1.5663612607810036, | |
| "learning_rate": 2.2808828859970905e-06, | |
| "loss": 0.0415, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.3687002652519893, | |
| "grad_norm": 1.3127711478830895, | |
| "learning_rate": 2.263423624322326e-06, | |
| "loss": 0.0429, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 1.3713527851458887, | |
| "grad_norm": 1.4631985318610559, | |
| "learning_rate": 2.2460118703468475e-06, | |
| "loss": 0.0426, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 1.3740053050397878, | |
| "grad_norm": 2.056857172754129, | |
| "learning_rate": 2.228647926343373e-06, | |
| "loss": 0.0549, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 1.376657824933687, | |
| "grad_norm": 1.4596328383955115, | |
| "learning_rate": 2.211332093754622e-06, | |
| "loss": 0.0463, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 1.3793103448275863, | |
| "grad_norm": 1.4795909661114393, | |
| "learning_rate": 2.1940646731880887e-06, | |
| "loss": 0.0382, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.3819628647214854, | |
| "grad_norm": 1.79070132377592, | |
| "learning_rate": 2.1768459644108223e-06, | |
| "loss": 0.0493, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 1.3846153846153846, | |
| "grad_norm": 1.6598245589918261, | |
| "learning_rate": 2.159676266344222e-06, | |
| "loss": 0.0504, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 1.387267904509284, | |
| "grad_norm": 1.5697697079805137, | |
| "learning_rate": 2.142555877058847e-06, | |
| "loss": 0.056, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 1.389920424403183, | |
| "grad_norm": 1.5250399923513906, | |
| "learning_rate": 2.125485093769242e-06, | |
| "loss": 0.0451, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 1.3925729442970822, | |
| "grad_norm": 1.488259185423536, | |
| "learning_rate": 2.108464212828786e-06, | |
| "loss": 0.0403, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.3952254641909814, | |
| "grad_norm": 1.7336708194270882, | |
| "learning_rate": 2.091493529724528e-06, | |
| "loss": 0.0545, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 1.3978779840848805, | |
| "grad_norm": 1.457566134032916, | |
| "learning_rate": 2.0745733390720744e-06, | |
| "loss": 0.0401, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 1.4005305039787799, | |
| "grad_norm": 2.0446606459323173, | |
| "learning_rate": 2.057703934610474e-06, | |
| "loss": 0.0511, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 1.403183023872679, | |
| "grad_norm": 1.8179579055212491, | |
| "learning_rate": 2.0408856091971063e-06, | |
| "loss": 0.0415, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 1.4058355437665782, | |
| "grad_norm": 1.4476314828089778, | |
| "learning_rate": 2.024118654802608e-06, | |
| "loss": 0.043, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.4084880636604775, | |
| "grad_norm": 1.4148575078926533, | |
| "learning_rate": 2.007403362505802e-06, | |
| "loss": 0.0408, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 1.4111405835543767, | |
| "grad_norm": 1.6321368893473254, | |
| "learning_rate": 1.990740022488642e-06, | |
| "loss": 0.0381, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 1.4137931034482758, | |
| "grad_norm": 1.5202544820877246, | |
| "learning_rate": 1.9741289240311757e-06, | |
| "loss": 0.045, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 1.4164456233421752, | |
| "grad_norm": 1.5924780405260186, | |
| "learning_rate": 1.957570355506522e-06, | |
| "loss": 0.0453, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 1.4190981432360743, | |
| "grad_norm": 1.7962035777442131, | |
| "learning_rate": 1.9410646043758737e-06, | |
| "loss": 0.0514, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.4217506631299734, | |
| "grad_norm": 1.3406272487252804, | |
| "learning_rate": 1.9246119571834904e-06, | |
| "loss": 0.0333, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 1.4244031830238728, | |
| "grad_norm": 1.730688912537288, | |
| "learning_rate": 1.9082126995517376e-06, | |
| "loss": 0.0475, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 1.427055702917772, | |
| "grad_norm": 1.5376121127489066, | |
| "learning_rate": 1.8918671161761227e-06, | |
| "loss": 0.0343, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 1.429708222811671, | |
| "grad_norm": 1.4633947145114203, | |
| "learning_rate": 1.8755754908203528e-06, | |
| "loss": 0.0409, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 1.4323607427055702, | |
| "grad_norm": 1.6079257325417993, | |
| "learning_rate": 1.8593381063114113e-06, | |
| "loss": 0.0418, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.4350132625994694, | |
| "grad_norm": 1.36008605859255, | |
| "learning_rate": 1.8431552445346434e-06, | |
| "loss": 0.04, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 1.4376657824933687, | |
| "grad_norm": 1.7947332913494305, | |
| "learning_rate": 1.827027186428869e-06, | |
| "loss": 0.0601, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 1.4403183023872679, | |
| "grad_norm": 1.3927853248296074, | |
| "learning_rate": 1.8109542119815e-06, | |
| "loss": 0.0425, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 1.442970822281167, | |
| "grad_norm": 1.415485606603946, | |
| "learning_rate": 1.7949366002236762e-06, | |
| "loss": 0.0408, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 1.4456233421750664, | |
| "grad_norm": 1.3417207244235902, | |
| "learning_rate": 1.7789746292254313e-06, | |
| "loss": 0.0415, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.4482758620689655, | |
| "grad_norm": 1.5705396456844423, | |
| "learning_rate": 1.7630685760908623e-06, | |
| "loss": 0.0414, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 1.4509283819628647, | |
| "grad_norm": 1.4346574390804177, | |
| "learning_rate": 1.7472187169533128e-06, | |
| "loss": 0.0399, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 1.453580901856764, | |
| "grad_norm": 1.5831645649543884, | |
| "learning_rate": 1.7314253269705854e-06, | |
| "loss": 0.0456, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 1.4562334217506632, | |
| "grad_norm": 1.5188840072828362, | |
| "learning_rate": 1.7156886803201638e-06, | |
| "loss": 0.0508, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 1.4588859416445623, | |
| "grad_norm": 1.739652082715737, | |
| "learning_rate": 1.70000905019445e-06, | |
| "loss": 0.0495, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.4615384615384617, | |
| "grad_norm": 1.826264975830559, | |
| "learning_rate": 1.6843867087960252e-06, | |
| "loss": 0.0437, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 1.4641909814323608, | |
| "grad_norm": 1.30588237840256, | |
| "learning_rate": 1.6688219273329215e-06, | |
| "loss": 0.0383, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 1.46684350132626, | |
| "grad_norm": 1.4986591185883977, | |
| "learning_rate": 1.6533149760139206e-06, | |
| "loss": 0.0389, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 1.469496021220159, | |
| "grad_norm": 1.4584576788521406, | |
| "learning_rate": 1.6378661240438498e-06, | |
| "loss": 0.0398, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 1.4721485411140582, | |
| "grad_norm": 1.738259374230387, | |
| "learning_rate": 1.6224756396189216e-06, | |
| "loss": 0.0489, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.4748010610079576, | |
| "grad_norm": 1.4204392194550715, | |
| "learning_rate": 1.6071437899220688e-06, | |
| "loss": 0.0414, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 1.4774535809018567, | |
| "grad_norm": 1.5679568822862189, | |
| "learning_rate": 1.591870841118312e-06, | |
| "loss": 0.0527, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 1.4801061007957559, | |
| "grad_norm": 1.595033631355667, | |
| "learning_rate": 1.576657058350135e-06, | |
| "loss": 0.043, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 1.4827586206896552, | |
| "grad_norm": 1.738257177021954, | |
| "learning_rate": 1.561502705732883e-06, | |
| "loss": 0.0479, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 1.4854111405835544, | |
| "grad_norm": 1.365927651900338, | |
| "learning_rate": 1.546408046350183e-06, | |
| "loss": 0.0408, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.4880636604774535, | |
| "grad_norm": 1.5461936213104102, | |
| "learning_rate": 1.5313733422493626e-06, | |
| "loss": 0.0432, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 1.490716180371353, | |
| "grad_norm": 1.4928548013692324, | |
| "learning_rate": 1.516398854436914e-06, | |
| "loss": 0.04, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 1.493368700265252, | |
| "grad_norm": 1.640223366433278, | |
| "learning_rate": 1.501484842873963e-06, | |
| "loss": 0.0546, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 1.4960212201591512, | |
| "grad_norm": 1.7973098844829862, | |
| "learning_rate": 1.486631566471745e-06, | |
| "loss": 0.046, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 1.4986737400530503, | |
| "grad_norm": 1.6511015774830353, | |
| "learning_rate": 1.4718392830871192e-06, | |
| "loss": 0.0445, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.5013262599469495, | |
| "grad_norm": 2.1489351880145637, | |
| "learning_rate": 1.457108249518089e-06, | |
| "loss": 0.0626, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 1.5039787798408488, | |
| "grad_norm": 1.4283531671752674, | |
| "learning_rate": 1.4424387214993457e-06, | |
| "loss": 0.0388, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 1.506631299734748, | |
| "grad_norm": 1.9589195614651327, | |
| "learning_rate": 1.4278309536978275e-06, | |
| "loss": 0.0403, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 1.509283819628647, | |
| "grad_norm": 1.6259539938234473, | |
| "learning_rate": 1.4132851997082969e-06, | |
| "loss": 0.0453, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 1.5119363395225465, | |
| "grad_norm": 1.52635618170666, | |
| "learning_rate": 1.3988017120489417e-06, | |
| "loss": 0.0487, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.5145888594164456, | |
| "grad_norm": 1.166944034701237, | |
| "learning_rate": 1.384380742156991e-06, | |
| "loss": 0.0345, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 1.5172413793103448, | |
| "grad_norm": 1.4683349538414516, | |
| "learning_rate": 1.370022540384347e-06, | |
| "loss": 0.0456, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 1.5198938992042441, | |
| "grad_norm": 1.2722220574404348, | |
| "learning_rate": 1.3557273559932372e-06, | |
| "loss": 0.0321, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 1.5225464190981433, | |
| "grad_norm": 1.420343043420184, | |
| "learning_rate": 1.3414954371518968e-06, | |
| "loss": 0.0486, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 1.5251989389920424, | |
| "grad_norm": 1.613818956433735, | |
| "learning_rate": 1.32732703093025e-06, | |
| "loss": 0.0462, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.5278514588859418, | |
| "grad_norm": 1.5717020920513154, | |
| "learning_rate": 1.3132223832956265e-06, | |
| "loss": 0.0393, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 1.530503978779841, | |
| "grad_norm": 1.899833644607642, | |
| "learning_rate": 1.2991817391084887e-06, | |
| "loss": 0.0571, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 1.53315649867374, | |
| "grad_norm": 1.3467128729397928, | |
| "learning_rate": 1.2852053421181826e-06, | |
| "loss": 0.0436, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 1.5358090185676394, | |
| "grad_norm": 1.4684591070731103, | |
| "learning_rate": 1.2712934349587063e-06, | |
| "loss": 0.0409, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 1.4659121730561506, | |
| "learning_rate": 1.257446259144494e-06, | |
| "loss": 0.0343, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.5411140583554377, | |
| "grad_norm": 1.2296857151247282, | |
| "learning_rate": 1.2436640550662333e-06, | |
| "loss": 0.0315, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 1.5437665782493368, | |
| "grad_norm": 1.5658367180498545, | |
| "learning_rate": 1.2299470619866778e-06, | |
| "loss": 0.0468, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 1.546419098143236, | |
| "grad_norm": 1.5447401683597295, | |
| "learning_rate": 1.2162955180365033e-06, | |
| "loss": 0.0406, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 1.5490716180371353, | |
| "grad_norm": 1.563584941111247, | |
| "learning_rate": 1.2027096602101728e-06, | |
| "loss": 0.0385, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 1.5517241379310345, | |
| "grad_norm": 1.6434334673624074, | |
| "learning_rate": 1.1891897243618184e-06, | |
| "loss": 0.0452, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.5543766578249336, | |
| "grad_norm": 1.4760334609742514, | |
| "learning_rate": 1.1757359452011497e-06, | |
| "loss": 0.0366, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 1.557029177718833, | |
| "grad_norm": 1.232360571349415, | |
| "learning_rate": 1.1623485562893772e-06, | |
| "loss": 0.0373, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 1.5596816976127321, | |
| "grad_norm": 1.2603274281429182, | |
| "learning_rate": 1.1490277900351637e-06, | |
| "loss": 0.0336, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 1.5623342175066313, | |
| "grad_norm": 1.4672611395225155, | |
| "learning_rate": 1.1357738776905802e-06, | |
| "loss": 0.0499, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 1.5649867374005306, | |
| "grad_norm": 1.3938518014710124, | |
| "learning_rate": 1.1225870493470952e-06, | |
| "loss": 0.0407, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.5676392572944295, | |
| "grad_norm": 1.5266979684533877, | |
| "learning_rate": 1.1094675339315825e-06, | |
| "loss": 0.0389, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 1.570291777188329, | |
| "grad_norm": 1.8368858591777315, | |
| "learning_rate": 1.0964155592023483e-06, | |
| "loss": 0.0353, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 1.5729442970822283, | |
| "grad_norm": 1.6847631793792286, | |
| "learning_rate": 1.083431351745171e-06, | |
| "loss": 0.0492, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 1.5755968169761272, | |
| "grad_norm": 1.457143006494126, | |
| "learning_rate": 1.0705151369693712e-06, | |
| "loss": 0.0373, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 1.5782493368700266, | |
| "grad_norm": 1.4537669375386875, | |
| "learning_rate": 1.0576671391038996e-06, | |
| "loss": 0.0404, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.5809018567639257, | |
| "grad_norm": 1.570259411391911, | |
| "learning_rate": 1.0448875811934417e-06, | |
| "loss": 0.0374, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 1.5835543766578248, | |
| "grad_norm": 1.4551622624887812, | |
| "learning_rate": 1.0321766850945486e-06, | |
| "loss": 0.0349, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 1.5862068965517242, | |
| "grad_norm": 1.320645141600928, | |
| "learning_rate": 1.0195346714717813e-06, | |
| "loss": 0.0365, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 1.5888594164456233, | |
| "grad_norm": 1.7279741174166037, | |
| "learning_rate": 1.0069617597938869e-06, | |
| "loss": 0.048, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 1.5915119363395225, | |
| "grad_norm": 1.3238694043973904, | |
| "learning_rate": 9.944581683299804e-07, | |
| "loss": 0.0402, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.5915119363395225, | |
| "eval_loss": 0.1134517639875412, | |
| "eval_runtime": 1.2987, | |
| "eval_samples_per_second": 23.87, | |
| "eval_steps_per_second": 6.16, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.5941644562334218, | |
| "grad_norm": 1.3051093843669757, | |
| "learning_rate": 9.82024114145761e-07, | |
| "loss": 0.0382, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 1.596816976127321, | |
| "grad_norm": 1.162659102103195, | |
| "learning_rate": 9.696598130997415e-07, | |
| "loss": 0.036, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 1.5994694960212201, | |
| "grad_norm": 1.7546331982712047, | |
| "learning_rate": 9.57365479839501e-07, | |
| "loss": 0.0494, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 1.6021220159151195, | |
| "grad_norm": 1.3672793279747577, | |
| "learning_rate": 9.45141327797961e-07, | |
| "loss": 0.0322, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 1.6047745358090184, | |
| "grad_norm": 1.3661103077040575, | |
| "learning_rate": 9.32987569189675e-07, | |
| "loss": 0.0391, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.6074270557029178, | |
| "grad_norm": 1.5456527347090927, | |
| "learning_rate": 9.209044150071522e-07, | |
| "loss": 0.042, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 1.6100795755968171, | |
| "grad_norm": 1.6924497065293256, | |
| "learning_rate": 9.088920750171876e-07, | |
| "loss": 0.0415, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 1.612732095490716, | |
| "grad_norm": 1.74031693520605, | |
| "learning_rate": 8.969507577572189e-07, | |
| "loss": 0.0431, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 1.6153846153846154, | |
| "grad_norm": 1.551498689243845, | |
| "learning_rate": 8.850806705317183e-07, | |
| "loss": 0.0361, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 1.6180371352785146, | |
| "grad_norm": 1.542261921192749, | |
| "learning_rate": 8.732820194085794e-07, | |
| "loss": 0.0428, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.6206896551724137, | |
| "grad_norm": 1.78892779608987, | |
| "learning_rate": 8.615550092155478e-07, | |
| "loss": 0.0451, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 1.623342175066313, | |
| "grad_norm": 1.555805233678888, | |
| "learning_rate": 8.498998435366634e-07, | |
| "loss": 0.0365, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 1.6259946949602122, | |
| "grad_norm": 1.5060342455389297, | |
| "learning_rate": 8.383167247087259e-07, | |
| "loss": 0.04, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 1.6286472148541113, | |
| "grad_norm": 1.611093579148822, | |
| "learning_rate": 8.268058538177826e-07, | |
| "loss": 0.0375, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 1.6312997347480107, | |
| "grad_norm": 1.588047950817851, | |
| "learning_rate": 8.15367430695636e-07, | |
| "loss": 0.0456, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.6339522546419099, | |
| "grad_norm": 1.3816401812608983, | |
| "learning_rate": 8.040016539163792e-07, | |
| "loss": 0.0312, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 1.636604774535809, | |
| "grad_norm": 1.4853951274849326, | |
| "learning_rate": 7.927087207929418e-07, | |
| "loss": 0.0371, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 1.6392572944297084, | |
| "grad_norm": 1.4635552701452224, | |
| "learning_rate": 7.814888273736698e-07, | |
| "loss": 0.0398, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 1.6419098143236073, | |
| "grad_norm": 1.4395599081772887, | |
| "learning_rate": 7.70342168438919e-07, | |
| "loss": 0.0377, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 1.6445623342175066, | |
| "grad_norm": 1.4904004060637046, | |
| "learning_rate": 7.592689374976769e-07, | |
| "loss": 0.0422, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.647214854111406, | |
| "grad_norm": 1.3555033892449118, | |
| "learning_rate": 7.482693267842e-07, | |
| "loss": 0.0298, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 1.649867374005305, | |
| "grad_norm": 1.6737297709368, | |
| "learning_rate": 7.373435272546764e-07, | |
| "loss": 0.0426, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 1.6525198938992043, | |
| "grad_norm": 1.4564908618465116, | |
| "learning_rate": 7.264917285839168e-07, | |
| "loss": 0.0329, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 1.6551724137931034, | |
| "grad_norm": 1.2263290141969596, | |
| "learning_rate": 7.157141191620548e-07, | |
| "loss": 0.0405, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 1.6578249336870026, | |
| "grad_norm": 1.5061155245041684, | |
| "learning_rate": 7.050108860912752e-07, | |
| "loss": 0.0423, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.660477453580902, | |
| "grad_norm": 1.342531472718274, | |
| "learning_rate": 6.943822151825735e-07, | |
| "loss": 0.0381, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 1.663129973474801, | |
| "grad_norm": 1.529969425604032, | |
| "learning_rate": 6.838282909525268e-07, | |
| "loss": 0.0378, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 1.6657824933687002, | |
| "grad_norm": 1.4693891043473597, | |
| "learning_rate": 6.733492966200872e-07, | |
| "loss": 0.0409, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 1.6684350132625996, | |
| "grad_norm": 1.2856071654517978, | |
| "learning_rate": 6.629454141034053e-07, | |
| "loss": 0.0359, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 1.6710875331564987, | |
| "grad_norm": 1.261936660004472, | |
| "learning_rate": 6.526168240166686e-07, | |
| "loss": 0.0375, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.6737400530503979, | |
| "grad_norm": 1.9177656406918595, | |
| "learning_rate": 6.423637056669702e-07, | |
| "loss": 0.0528, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 1.6763925729442972, | |
| "grad_norm": 1.1739232814066218, | |
| "learning_rate": 6.321862370511922e-07, | |
| "loss": 0.0339, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 1.6790450928381961, | |
| "grad_norm": 1.5997908198741042, | |
| "learning_rate": 6.220845948529159e-07, | |
| "loss": 0.0365, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 1.6816976127320955, | |
| "grad_norm": 1.9896095612620952, | |
| "learning_rate": 6.120589544393596e-07, | |
| "loss": 0.0442, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 1.6843501326259946, | |
| "grad_norm": 1.4447133472783895, | |
| "learning_rate": 6.021094898583269e-07, | |
| "loss": 0.0417, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 1.6870026525198938, | |
| "grad_norm": 1.423453466191924, | |
| "learning_rate": 5.922363738351888e-07, | |
| "loss": 0.0416, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 1.6896551724137931, | |
| "grad_norm": 1.6412520906518056, | |
| "learning_rate": 5.824397777698859e-07, | |
| "loss": 0.0376, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 1.6923076923076923, | |
| "grad_norm": 1.3500786343856588, | |
| "learning_rate": 5.727198717339511e-07, | |
| "loss": 0.0365, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 1.6949602122015914, | |
| "grad_norm": 1.3281753235494516, | |
| "learning_rate": 5.630768244675583e-07, | |
| "loss": 0.0355, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 1.6976127320954908, | |
| "grad_norm": 1.5686317332318083, | |
| "learning_rate": 5.535108033765913e-07, | |
| "loss": 0.047, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.70026525198939, | |
| "grad_norm": 1.8508472340714566, | |
| "learning_rate": 5.440219745297432e-07, | |
| "loss": 0.049, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 1.702917771883289, | |
| "grad_norm": 1.6565736316956416, | |
| "learning_rate": 5.346105026556226e-07, | |
| "loss": 0.0397, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 1.7055702917771884, | |
| "grad_norm": 1.5711209136989794, | |
| "learning_rate": 5.252765511399044e-07, | |
| "loss": 0.0436, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 1.7082228116710876, | |
| "grad_norm": 1.5673409139078343, | |
| "learning_rate": 5.160202820224875e-07, | |
| "loss": 0.0441, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 1.7108753315649867, | |
| "grad_norm": 1.545089411507962, | |
| "learning_rate": 5.068418559946864e-07, | |
| "loss": 0.0348, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.713527851458886, | |
| "grad_norm": 1.432485291304248, | |
| "learning_rate": 4.977414323964364e-07, | |
| "loss": 0.0375, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 1.716180371352785, | |
| "grad_norm": 1.6399078665274962, | |
| "learning_rate": 4.88719169213529e-07, | |
| "loss": 0.039, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 1.7188328912466844, | |
| "grad_norm": 1.987599098491381, | |
| "learning_rate": 4.797752230748721e-07, | |
| "loss": 0.0515, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 1.7214854111405835, | |
| "grad_norm": 1.1831605370816431, | |
| "learning_rate": 4.7090974924976716e-07, | |
| "loss": 0.0313, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 1.7241379310344827, | |
| "grad_norm": 1.462425034427707, | |
| "learning_rate": 4.6212290164521554e-07, | |
| "loss": 0.0384, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.726790450928382, | |
| "grad_norm": 1.2550042615331936, | |
| "learning_rate": 4.534148328032456e-07, | |
| "loss": 0.0333, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 1.7294429708222812, | |
| "grad_norm": 2.006128939436255, | |
| "learning_rate": 4.4478569389826864e-07, | |
| "loss": 0.0362, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 1.7320954907161803, | |
| "grad_norm": 1.0400629902188836, | |
| "learning_rate": 4.3623563473444817e-07, | |
| "loss": 0.0389, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 1.7347480106100797, | |
| "grad_norm": 1.40192347825387, | |
| "learning_rate": 4.277648037430998e-07, | |
| "loss": 0.039, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 1.7374005305039788, | |
| "grad_norm": 1.183366553658279, | |
| "learning_rate": 4.193733479801232e-07, | |
| "loss": 0.0288, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.740053050397878, | |
| "grad_norm": 1.3787704383311228, | |
| "learning_rate": 4.110614131234375e-07, | |
| "loss": 0.0391, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 1.7427055702917773, | |
| "grad_norm": 1.5234821215536667, | |
| "learning_rate": 4.028291434704601e-07, | |
| "loss": 0.0402, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 1.7453580901856764, | |
| "grad_norm": 1.7063010359505684, | |
| "learning_rate": 3.946766819355985e-07, | |
| "loss": 0.0477, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 1.7480106100795756, | |
| "grad_norm": 1.6813393539550572, | |
| "learning_rate": 3.866041700477691e-07, | |
| "loss": 0.0451, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 1.750663129973475, | |
| "grad_norm": 1.7013026951961887, | |
| "learning_rate": 3.786117479479423e-07, | |
| "loss": 0.0375, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.7533156498673739, | |
| "grad_norm": 1.335932666043299, | |
| "learning_rate": 3.7069955438670704e-07, | |
| "loss": 0.0365, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 1.7559681697612732, | |
| "grad_norm": 1.3147540821216455, | |
| "learning_rate": 3.62867726721865e-07, | |
| "loss": 0.033, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 1.7586206896551724, | |
| "grad_norm": 1.6551386695113186, | |
| "learning_rate": 3.5511640091604293e-07, | |
| "loss": 0.0443, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 1.7612732095490715, | |
| "grad_norm": 1.2796382269681643, | |
| "learning_rate": 3.474457115343344e-07, | |
| "loss": 0.0341, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 1.7639257294429709, | |
| "grad_norm": 1.4215576260782057, | |
| "learning_rate": 3.398557917419626e-07, | |
| "loss": 0.0353, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.76657824933687, | |
| "grad_norm": 1.701821067856708, | |
| "learning_rate": 3.3234677330196865e-07, | |
| "loss": 0.0451, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 1.7692307692307692, | |
| "grad_norm": 1.182220074597278, | |
| "learning_rate": 3.2491878657292643e-07, | |
| "loss": 0.0366, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 1.7718832891246685, | |
| "grad_norm": 1.6450299221574771, | |
| "learning_rate": 3.175719605066746e-07, | |
| "loss": 0.0405, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 1.7745358090185677, | |
| "grad_norm": 1.9308466623075535, | |
| "learning_rate": 3.1030642264608393e-07, | |
| "loss": 0.0425, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 1.7771883289124668, | |
| "grad_norm": 2.0326661438505202, | |
| "learning_rate": 3.0312229912283884e-07, | |
| "loss": 0.048, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.7798408488063662, | |
| "grad_norm": 1.576985336913738, | |
| "learning_rate": 2.96019714655249e-07, | |
| "loss": 0.0385, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 1.782493368700265, | |
| "grad_norm": 1.4496388975425247, | |
| "learning_rate": 2.88998792546083e-07, | |
| "loss": 0.0366, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 1.7851458885941645, | |
| "grad_norm": 1.3474637091777282, | |
| "learning_rate": 2.820596546804316e-07, | |
| "loss": 0.0445, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 1.7877984084880638, | |
| "grad_norm": 1.6312145174464268, | |
| "learning_rate": 2.7520242152358767e-07, | |
| "loss": 0.047, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 1.7904509283819627, | |
| "grad_norm": 1.1102230652695808, | |
| "learning_rate": 2.6842721211895516e-07, | |
| "loss": 0.0309, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.793103448275862, | |
| "grad_norm": 1.6712723138079337, | |
| "learning_rate": 2.617341440859883e-07, | |
| "loss": 0.0391, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 1.7957559681697612, | |
| "grad_norm": 1.4567044509094966, | |
| "learning_rate": 2.551233336181386e-07, | |
| "loss": 0.0368, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 1.7984084880636604, | |
| "grad_norm": 1.4754171869798691, | |
| "learning_rate": 2.485948954808493e-07, | |
| "loss": 0.0386, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 1.8010610079575597, | |
| "grad_norm": 1.2965369200067902, | |
| "learning_rate": 2.421489430095547e-07, | |
| "loss": 0.0272, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 1.8037135278514589, | |
| "grad_norm": 1.4954201899471793, | |
| "learning_rate": 2.357855881077181e-07, | |
| "loss": 0.0436, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.806366047745358, | |
| "grad_norm": 1.081445215280861, | |
| "learning_rate": 2.2950494124488687e-07, | |
| "loss": 0.0355, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 1.8090185676392574, | |
| "grad_norm": 1.319457308590872, | |
| "learning_rate": 2.2330711145477247e-07, | |
| "loss": 0.0301, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 1.8116710875331565, | |
| "grad_norm": 1.5878396866106896, | |
| "learning_rate": 2.1719220633336147e-07, | |
| "loss": 0.0335, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 1.8143236074270557, | |
| "grad_norm": 1.4109191318255065, | |
| "learning_rate": 2.1116033203704534e-07, | |
| "loss": 0.0321, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 1.816976127320955, | |
| "grad_norm": 1.7498102959049076, | |
| "learning_rate": 2.0521159328077856e-07, | |
| "loss": 0.037, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.819628647214854, | |
| "grad_norm": 1.4473167506873907, | |
| "learning_rate": 1.993460933362601e-07, | |
| "loss": 0.0364, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 1.8222811671087533, | |
| "grad_norm": 1.2905934999941218, | |
| "learning_rate": 1.935639340301415e-07, | |
| "loss": 0.0304, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 1.8249336870026527, | |
| "grad_norm": 1.6396632866378924, | |
| "learning_rate": 1.8786521574225837e-07, | |
| "loss": 0.0492, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 1.8275862068965516, | |
| "grad_norm": 1.4628611231715085, | |
| "learning_rate": 1.8225003740388546e-07, | |
| "loss": 0.0384, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 1.830238726790451, | |
| "grad_norm": 1.4318855930579866, | |
| "learning_rate": 1.7671849649602502e-07, | |
| "loss": 0.0363, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.83289124668435, | |
| "grad_norm": 1.1555536281682646, | |
| "learning_rate": 1.7127068904770948e-07, | |
| "loss": 0.0316, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 1.8355437665782492, | |
| "grad_norm": 1.276197266535223, | |
| "learning_rate": 1.6590670963433642e-07, | |
| "loss": 0.0322, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 1.8381962864721486, | |
| "grad_norm": 1.540221628307776, | |
| "learning_rate": 1.6062665137602572e-07, | |
| "loss": 0.039, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 1.8408488063660478, | |
| "grad_norm": 1.488141776444817, | |
| "learning_rate": 1.5543060593600334e-07, | |
| "loss": 0.0497, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 1.843501326259947, | |
| "grad_norm": 1.4064075758646781, | |
| "learning_rate": 1.5031866351901182e-07, | |
| "loss": 0.0354, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "grad_norm": 1.7289030418499938, | |
| "learning_rate": 1.4529091286973994e-07, | |
| "loss": 0.047, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 1.8488063660477454, | |
| "grad_norm": 1.4042197094833593, | |
| "learning_rate": 1.403474412712874e-07, | |
| "loss": 0.0352, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 1.8514588859416445, | |
| "grad_norm": 1.0019472105701877, | |
| "learning_rate": 1.3548833454364641e-07, | |
| "loss": 0.024, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 1.854111405835544, | |
| "grad_norm": 1.773377229970334, | |
| "learning_rate": 1.3071367704221129e-07, | |
| "loss": 0.0444, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 1.8567639257294428, | |
| "grad_norm": 1.1177532519030482, | |
| "learning_rate": 1.260235516563163e-07, | |
| "loss": 0.0363, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.8594164456233422, | |
| "grad_norm": 1.2900623934609892, | |
| "learning_rate": 1.2141803980779464e-07, | |
| "loss": 0.0317, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 1.8620689655172413, | |
| "grad_norm": 1.8304735271621675, | |
| "learning_rate": 1.1689722144956672e-07, | |
| "loss": 0.0432, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 1.8647214854111405, | |
| "grad_norm": 1.4188940996465451, | |
| "learning_rate": 1.1246117506425014e-07, | |
| "loss": 0.0392, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 1.8673740053050398, | |
| "grad_norm": 1.4825923835504944, | |
| "learning_rate": 1.0810997766279974e-07, | |
| "loss": 0.0382, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 1.870026525198939, | |
| "grad_norm": 1.1014643278241296, | |
| "learning_rate": 1.0384370478316919e-07, | |
| "loss": 0.0311, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.8726790450928381, | |
| "grad_norm": 1.3678036067136727, | |
| "learning_rate": 9.966243048899704e-08, | |
| "loss": 0.0331, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 1.8753315649867375, | |
| "grad_norm": 1.673390021703709, | |
| "learning_rate": 9.556622736832665e-08, | |
| "loss": 0.0481, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 1.8779840848806366, | |
| "grad_norm": 1.5656233630923257, | |
| "learning_rate": 9.155516653234276e-08, | |
| "loss": 0.0442, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 1.8806366047745358, | |
| "grad_norm": 1.320780746379846, | |
| "learning_rate": 8.762931761413573e-08, | |
| "loss": 0.0382, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 1.8832891246684351, | |
| "grad_norm": 1.8359864667941603, | |
| "learning_rate": 8.378874876749433e-08, | |
| "loss": 0.0369, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.8859416445623343, | |
| "grad_norm": 1.6235154801266736, | |
| "learning_rate": 8.003352666572428e-08, | |
| "loss": 0.0404, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 1.8885941644562334, | |
| "grad_norm": 1.8357177886604772, | |
| "learning_rate": 7.636371650048658e-08, | |
| "loss": 0.0458, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 1.8912466843501328, | |
| "grad_norm": 1.4570512705105305, | |
| "learning_rate": 7.277938198066992e-08, | |
| "loss": 0.0418, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 1.8938992042440317, | |
| "grad_norm": 1.2604162586852787, | |
| "learning_rate": 6.928058533128112e-08, | |
| "loss": 0.0323, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 1.896551724137931, | |
| "grad_norm": 1.6009023727823846, | |
| "learning_rate": 6.58673872923693e-08, | |
| "loss": 0.0359, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.8992042440318302, | |
| "grad_norm": 1.2976436667625482, | |
| "learning_rate": 6.253984711796612e-08, | |
| "loss": 0.0371, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 1.9018567639257293, | |
| "grad_norm": 1.2870991478173943, | |
| "learning_rate": 5.929802257506112e-08, | |
| "loss": 0.045, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 1.9045092838196287, | |
| "grad_norm": 1.6902715191719975, | |
| "learning_rate": 5.6141969942596906e-08, | |
| "loss": 0.0389, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 1.9071618037135278, | |
| "grad_norm": 1.7050325681735479, | |
| "learning_rate": 5.307174401049275e-08, | |
| "loss": 0.0334, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 1.909814323607427, | |
| "grad_norm": 1.5808736002449604, | |
| "learning_rate": 5.0087398078694785e-08, | |
| "loss": 0.0417, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.9124668435013263, | |
| "grad_norm": 1.3430964938486658, | |
| "learning_rate": 4.718898395624671e-08, | |
| "loss": 0.0339, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 1.9151193633952255, | |
| "grad_norm": 1.489257458379661, | |
| "learning_rate": 4.437655196039559e-08, | |
| "loss": 0.0429, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 1.9177718832891246, | |
| "grad_norm": 1.4210307447912756, | |
| "learning_rate": 4.1650150915714674e-08, | |
| "loss": 0.0421, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 1.920424403183024, | |
| "grad_norm": 1.2959156290024842, | |
| "learning_rate": 3.900982815325582e-08, | |
| "loss": 0.0355, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "grad_norm": 1.4827766464308814, | |
| "learning_rate": 3.645562950973014e-08, | |
| "loss": 0.0369, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.9257294429708223, | |
| "grad_norm": 1.572959594152032, | |
| "learning_rate": 3.3987599326710806e-08, | |
| "loss": 0.0523, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 1.9283819628647216, | |
| "grad_norm": 1.519959930918152, | |
| "learning_rate": 3.160578044986373e-08, | |
| "loss": 0.0354, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 1.9310344827586206, | |
| "grad_norm": 1.2687605705293235, | |
| "learning_rate": 2.9310214228202016e-08, | |
| "loss": 0.0367, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 1.93368700265252, | |
| "grad_norm": 1.2589757041102903, | |
| "learning_rate": 2.7100940513370976e-08, | |
| "loss": 0.037, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 1.936339522546419, | |
| "grad_norm": 1.3802142564158149, | |
| "learning_rate": 2.4977997658954257e-08, | |
| "loss": 0.0384, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.9389920424403182, | |
| "grad_norm": 1.1728085570996578, | |
| "learning_rate": 2.29414225198088e-08, | |
| "loss": 0.0332, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 1.9416445623342176, | |
| "grad_norm": 1.4829028934918602, | |
| "learning_rate": 2.0991250451424806e-08, | |
| "loss": 0.0382, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 1.9442970822281167, | |
| "grad_norm": 1.0525766488908528, | |
| "learning_rate": 1.912751530931234e-08, | |
| "loss": 0.0307, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 1.9469496021220158, | |
| "grad_norm": 1.7747299163019683, | |
| "learning_rate": 1.735024944841235e-08, | |
| "loss": 0.0412, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 1.9496021220159152, | |
| "grad_norm": 1.3308814219146978, | |
| "learning_rate": 1.5659483722537117e-08, | |
| "loss": 0.0332, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.9522546419098143, | |
| "grad_norm": 1.6064157720427565, | |
| "learning_rate": 1.4055247483832356e-08, | |
| "loss": 0.0434, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 1.9549071618037135, | |
| "grad_norm": 1.334355441410993, | |
| "learning_rate": 1.2537568582269289e-08, | |
| "loss": 0.0407, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 1.9575596816976129, | |
| "grad_norm": 1.8576856489215392, | |
| "learning_rate": 1.110647336516002e-08, | |
| "loss": 0.041, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 1.9602122015915118, | |
| "grad_norm": 1.3915797639255931, | |
| "learning_rate": 9.761986676701251e-09, | |
| "loss": 0.0338, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 1.9628647214854111, | |
| "grad_norm": 1.696616311407856, | |
| "learning_rate": 8.504131857542952e-09, | |
| "loss": 0.0438, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.9655172413793105, | |
| "grad_norm": 1.2509363269884906, | |
| "learning_rate": 7.332930744380906e-09, | |
| "loss": 0.0359, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 1.9681697612732094, | |
| "grad_norm": 1.5867898086364787, | |
| "learning_rate": 6.24840366958035e-09, | |
| "loss": 0.0337, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 1.9708222811671088, | |
| "grad_norm": 1.0944506086655037, | |
| "learning_rate": 5.250569460822363e-09, | |
| "loss": 0.0304, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 1.973474801061008, | |
| "grad_norm": 1.6628593056793242, | |
| "learning_rate": 4.339445440776358e-09, | |
| "loss": 0.0377, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 1.976127320954907, | |
| "grad_norm": 1.7268640463935314, | |
| "learning_rate": 3.5150474267992007e-09, | |
| "loss": 0.0435, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.9787798408488064, | |
| "grad_norm": 1.469911347318018, | |
| "learning_rate": 2.7773897306615504e-09, | |
| "loss": 0.0377, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 1.9814323607427056, | |
| "grad_norm": 1.420365368396037, | |
| "learning_rate": 2.126485158298608e-09, | |
| "loss": 0.0384, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 1.9840848806366047, | |
| "grad_norm": 1.1547923580578376, | |
| "learning_rate": 1.5623450095880731e-09, | |
| "loss": 0.0289, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 1.986737400530504, | |
| "grad_norm": 1.7325651456934643, | |
| "learning_rate": 1.0849790781541913e-09, | |
| "loss": 0.0369, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 1.9893899204244032, | |
| "grad_norm": 1.6534729465614353, | |
| "learning_rate": 6.943956511973326e-10, | |
| "loss": 0.0513, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.9920424403183024, | |
| "grad_norm": 1.2257336512995223, | |
| "learning_rate": 3.9060150935077425e-10, | |
| "loss": 0.0341, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 1.9946949602122017, | |
| "grad_norm": 1.5013042404201287, | |
| "learning_rate": 1.7360192656246112e-10, | |
| "loss": 0.0428, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 1.9973474801061006, | |
| "grad_norm": 1.5234604261794584, | |
| "learning_rate": 4.340067000230264e-11, | |
| "loss": 0.0356, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.1204706677910448, | |
| "learning_rate": 0.0, | |
| "loss": 0.0257, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 754, | |
| "total_flos": 7513536872448.0, | |
| "train_loss": 0.09835794550769367, | |
| "train_runtime": 628.4293, | |
| "train_samples_per_second": 9.589, | |
| "train_steps_per_second": 1.2 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 754, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 5000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 7513536872448.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |