| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 99, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.030303030303030304, | |
| "grad_norm": 1.5626156330108643, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 1.8564, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.06060606060606061, | |
| "grad_norm": 1.567307710647583, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 1.8964, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.09090909090909091, | |
| "grad_norm": 1.4846278429031372, | |
| "learning_rate": 2e-05, | |
| "loss": 1.7225, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.12121212121212122, | |
| "grad_norm": 1.5041078329086304, | |
| "learning_rate": 1.9994645874763657e-05, | |
| "loss": 1.7961, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.15151515151515152, | |
| "grad_norm": 1.3576281070709229, | |
| "learning_rate": 1.9978589232386036e-05, | |
| "loss": 1.5684, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 1.5841476917266846, | |
| "learning_rate": 1.995184726672197e-05, | |
| "loss": 1.6809, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.21212121212121213, | |
| "grad_norm": 1.451629638671875, | |
| "learning_rate": 1.9914448613738107e-05, | |
| "loss": 1.6862, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.24242424242424243, | |
| "grad_norm": 1.3648676872253418, | |
| "learning_rate": 1.9866433320848793e-05, | |
| "loss": 1.6725, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.2727272727272727, | |
| "grad_norm": 1.4545490741729736, | |
| "learning_rate": 1.9807852804032306e-05, | |
| "loss": 1.615, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.30303030303030304, | |
| "grad_norm": 1.4878568649291992, | |
| "learning_rate": 1.9738769792773338e-05, | |
| "loss": 1.567, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 1.9032509326934814, | |
| "learning_rate": 1.9659258262890683e-05, | |
| "loss": 1.6849, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 1.4631747007369995, | |
| "learning_rate": 1.956940335732209e-05, | |
| "loss": 1.7436, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.3939393939393939, | |
| "grad_norm": 1.3237619400024414, | |
| "learning_rate": 1.946930129495106e-05, | |
| "loss": 1.7103, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.42424242424242425, | |
| "grad_norm": 1.4598267078399658, | |
| "learning_rate": 1.935905926757326e-05, | |
| "loss": 1.813, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.45454545454545453, | |
| "grad_norm": 1.6514884233474731, | |
| "learning_rate": 1.9238795325112867e-05, | |
| "loss": 1.7952, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.48484848484848486, | |
| "grad_norm": 1.4365404844284058, | |
| "learning_rate": 1.910863824921176e-05, | |
| "loss": 1.7656, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.5151515151515151, | |
| "grad_norm": 1.253454566001892, | |
| "learning_rate": 1.8968727415326885e-05, | |
| "loss": 1.8244, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.5454545454545454, | |
| "grad_norm": 1.189319133758545, | |
| "learning_rate": 1.881921264348355e-05, | |
| "loss": 1.6908, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.5757575757575758, | |
| "grad_norm": 1.5841031074523926, | |
| "learning_rate": 1.866025403784439e-05, | |
| "loss": 1.7654, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.6060606060606061, | |
| "grad_norm": 1.2561309337615967, | |
| "learning_rate": 1.849202181526579e-05, | |
| "loss": 1.7271, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.6363636363636364, | |
| "grad_norm": 1.532762885093689, | |
| "learning_rate": 1.8314696123025456e-05, | |
| "loss": 1.7545, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 1.6211514472961426, | |
| "learning_rate": 1.8128466845916156e-05, | |
| "loss": 1.4955, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.696969696969697, | |
| "grad_norm": 1.5107672214508057, | |
| "learning_rate": 1.7933533402912354e-05, | |
| "loss": 1.5387, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 1.3635034561157227, | |
| "learning_rate": 1.773010453362737e-05, | |
| "loss": 1.5995, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.7575757575757576, | |
| "grad_norm": 1.3096054792404175, | |
| "learning_rate": 1.7518398074789776e-05, | |
| "loss": 1.7549, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.7878787878787878, | |
| "grad_norm": 1.2032136917114258, | |
| "learning_rate": 1.7298640726978357e-05, | |
| "loss": 1.634, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.8181818181818182, | |
| "grad_norm": 1.1950513124465942, | |
| "learning_rate": 1.7071067811865477e-05, | |
| "loss": 1.6768, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.8484848484848485, | |
| "grad_norm": 1.6178178787231445, | |
| "learning_rate": 1.6835923020228714e-05, | |
| "loss": 1.739, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.8787878787878788, | |
| "grad_norm": 1.3394091129302979, | |
| "learning_rate": 1.659345815100069e-05, | |
| "loss": 1.8401, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 1.3852413892745972, | |
| "learning_rate": 1.6343932841636455e-05, | |
| "loss": 1.6928, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.9393939393939394, | |
| "grad_norm": 1.2562795877456665, | |
| "learning_rate": 1.608761429008721e-05, | |
| "loss": 1.7373, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.9696969696969697, | |
| "grad_norm": 1.1925987005233765, | |
| "learning_rate": 1.5824776968678024e-05, | |
| "loss": 1.5096, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.7065600156784058, | |
| "learning_rate": 1.5555702330196024e-05, | |
| "loss": 1.3249, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.0303030303030303, | |
| "grad_norm": 1.791105031967163, | |
| "learning_rate": 1.528067850650368e-05, | |
| "loss": 1.2349, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.0606060606060606, | |
| "grad_norm": 1.6474976539611816, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 1.2779, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 1.5695178508758545, | |
| "learning_rate": 1.4713967368259981e-05, | |
| "loss": 1.2858, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.121212121212121, | |
| "grad_norm": 1.3161590099334717, | |
| "learning_rate": 1.4422886902190014e-05, | |
| "loss": 1.1867, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.1515151515151516, | |
| "grad_norm": 1.5779008865356445, | |
| "learning_rate": 1.4127070298043949e-05, | |
| "loss": 1.188, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.1818181818181819, | |
| "grad_norm": 1.5446536540985107, | |
| "learning_rate": 1.3826834323650899e-05, | |
| "loss": 1.1306, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.2121212121212122, | |
| "grad_norm": 1.9311134815216064, | |
| "learning_rate": 1.3522500479212337e-05, | |
| "loss": 1.2253, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.2424242424242424, | |
| "grad_norm": 2.0665199756622314, | |
| "learning_rate": 1.3214394653031616e-05, | |
| "loss": 1.2405, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.2727272727272727, | |
| "grad_norm": 1.340359091758728, | |
| "learning_rate": 1.2902846772544625e-05, | |
| "loss": 1.0486, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.303030303030303, | |
| "grad_norm": 1.5210260152816772, | |
| "learning_rate": 1.2588190451025209e-05, | |
| "loss": 1.1205, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 1.4104619026184082, | |
| "learning_rate": 1.2270762630343734e-05, | |
| "loss": 1.1148, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.3636363636363638, | |
| "grad_norm": 1.373587965965271, | |
| "learning_rate": 1.1950903220161286e-05, | |
| "loss": 1.1462, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.393939393939394, | |
| "grad_norm": 1.364762306213379, | |
| "learning_rate": 1.162895473394589e-05, | |
| "loss": 1.0527, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.4242424242424243, | |
| "grad_norm": 1.7049798965454102, | |
| "learning_rate": 1.130526192220052e-05, | |
| "loss": 1.1314, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 1.7813589572906494, | |
| "learning_rate": 1.098017140329561e-05, | |
| "loss": 1.1012, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.4848484848484849, | |
| "grad_norm": 1.595564603805542, | |
| "learning_rate": 1.0654031292301432e-05, | |
| "loss": 1.0513, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.5151515151515151, | |
| "grad_norm": 1.5489888191223145, | |
| "learning_rate": 1.0327190828217763e-05, | |
| "loss": 1.1117, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.5454545454545454, | |
| "grad_norm": 1.7180898189544678, | |
| "learning_rate": 1e-05, | |
| "loss": 1.1898, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.5757575757575757, | |
| "grad_norm": 1.3736799955368042, | |
| "learning_rate": 9.67280917178224e-06, | |
| "loss": 1.0677, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.606060606060606, | |
| "grad_norm": 1.3806053400039673, | |
| "learning_rate": 9.34596870769857e-06, | |
| "loss": 1.1602, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.6363636363636362, | |
| "grad_norm": 1.5797897577285767, | |
| "learning_rate": 9.019828596704394e-06, | |
| "loss": 1.0706, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 1.6175159215927124, | |
| "learning_rate": 8.694738077799487e-06, | |
| "loss": 1.147, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.696969696969697, | |
| "grad_norm": 1.6543636322021484, | |
| "learning_rate": 8.371045266054114e-06, | |
| "loss": 0.9402, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.7272727272727273, | |
| "grad_norm": 1.4852937459945679, | |
| "learning_rate": 8.04909677983872e-06, | |
| "loss": 1.0766, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.7575757575757576, | |
| "grad_norm": 1.6130305528640747, | |
| "learning_rate": 7.72923736965627e-06, | |
| "loss": 1.0769, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.7878787878787878, | |
| "grad_norm": 1.4735853672027588, | |
| "learning_rate": 7.411809548974792e-06, | |
| "loss": 0.9402, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 1.5684136152267456, | |
| "learning_rate": 7.097153227455379e-06, | |
| "loss": 1.1553, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.8484848484848486, | |
| "grad_norm": 1.4093765020370483, | |
| "learning_rate": 6.785605346968387e-06, | |
| "loss": 1.0747, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.878787878787879, | |
| "grad_norm": 1.2906310558319092, | |
| "learning_rate": 6.4774995207876654e-06, | |
| "loss": 1.0484, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.9090909090909092, | |
| "grad_norm": 1.5653777122497559, | |
| "learning_rate": 6.173165676349103e-06, | |
| "loss": 1.2066, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.9393939393939394, | |
| "grad_norm": 1.3291213512420654, | |
| "learning_rate": 5.872929701956054e-06, | |
| "loss": 1.0324, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.9696969696969697, | |
| "grad_norm": 1.45272958278656, | |
| "learning_rate": 5.5771130978099896e-06, | |
| "loss": 1.0972, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.6642159223556519, | |
| "learning_rate": 5.286032631740023e-06, | |
| "loss": 0.893, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.0303030303030303, | |
| "grad_norm": 1.6590158939361572, | |
| "learning_rate": 5.000000000000003e-06, | |
| "loss": 0.8978, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.0606060606060606, | |
| "grad_norm": 2.192969799041748, | |
| "learning_rate": 4.7193214934963204e-06, | |
| "loss": 0.891, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.090909090909091, | |
| "grad_norm": 1.7911643981933594, | |
| "learning_rate": 4.444297669803981e-06, | |
| "loss": 0.7529, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.121212121212121, | |
| "grad_norm": 1.4908910989761353, | |
| "learning_rate": 4.17522303132198e-06, | |
| "loss": 0.7556, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.1515151515151514, | |
| "grad_norm": 1.7466297149658203, | |
| "learning_rate": 3.912385709912794e-06, | |
| "loss": 0.8257, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "grad_norm": 1.5904890298843384, | |
| "learning_rate": 3.6560671583635467e-06, | |
| "loss": 0.7562, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.212121212121212, | |
| "grad_norm": 1.60899817943573, | |
| "learning_rate": 3.4065418489993118e-06, | |
| "loss": 0.7925, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.242424242424242, | |
| "grad_norm": 1.9628279209136963, | |
| "learning_rate": 3.1640769797712865e-06, | |
| "loss": 0.7809, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.2727272727272725, | |
| "grad_norm": 1.9455692768096924, | |
| "learning_rate": 2.9289321881345257e-06, | |
| "loss": 0.7837, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.303030303030303, | |
| "grad_norm": 1.9199553728103638, | |
| "learning_rate": 2.7013592730216464e-06, | |
| "loss": 0.9131, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 2.3333333333333335, | |
| "grad_norm": 2.134833335876465, | |
| "learning_rate": 2.4816019252102274e-06, | |
| "loss": 0.7476, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 2.3636363636363638, | |
| "grad_norm": 1.9508200883865356, | |
| "learning_rate": 2.26989546637263e-06, | |
| "loss": 0.7747, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 2.393939393939394, | |
| "grad_norm": 1.4582135677337646, | |
| "learning_rate": 2.0664665970876496e-06, | |
| "loss": 0.8055, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 2.4242424242424243, | |
| "grad_norm": 1.7351216077804565, | |
| "learning_rate": 1.8715331540838488e-06, | |
| "loss": 0.8209, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.4545454545454546, | |
| "grad_norm": 1.351690411567688, | |
| "learning_rate": 1.6853038769745466e-06, | |
| "loss": 0.7738, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 2.484848484848485, | |
| "grad_norm": 1.5985406637191772, | |
| "learning_rate": 1.5079781847342122e-06, | |
| "loss": 0.6174, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 2.515151515151515, | |
| "grad_norm": 1.7237467765808105, | |
| "learning_rate": 1.339745962155613e-06, | |
| "loss": 0.6738, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 2.5454545454545454, | |
| "grad_norm": 1.442694067955017, | |
| "learning_rate": 1.1807873565164507e-06, | |
| "loss": 0.913, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 2.5757575757575757, | |
| "grad_norm": 1.5762035846710205, | |
| "learning_rate": 1.0312725846731174e-06, | |
| "loss": 0.8292, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 2.606060606060606, | |
| "grad_norm": 1.6882027387619019, | |
| "learning_rate": 8.91361750788241e-07, | |
| "loss": 0.7793, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 2.6363636363636362, | |
| "grad_norm": 1.3225022554397583, | |
| "learning_rate": 7.612046748871327e-07, | |
| "loss": 0.7621, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 1.8009114265441895, | |
| "learning_rate": 6.409407324267448e-07, | |
| "loss": 0.7223, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 2.6969696969696972, | |
| "grad_norm": 1.5908373594284058, | |
| "learning_rate": 5.306987050489442e-07, | |
| "loss": 0.801, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 2.7272727272727275, | |
| "grad_norm": 1.5530376434326172, | |
| "learning_rate": 4.305966426779118e-07, | |
| "loss": 0.735, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 2.757575757575758, | |
| "grad_norm": 1.3646830320358276, | |
| "learning_rate": 3.4074173710931804e-07, | |
| "loss": 0.7068, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 2.787878787878788, | |
| "grad_norm": 1.6398669481277466, | |
| "learning_rate": 2.612302072266637e-07, | |
| "loss": 0.7385, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 2.8181818181818183, | |
| "grad_norm": 1.7059106826782227, | |
| "learning_rate": 1.921471959676957e-07, | |
| "loss": 0.6993, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 2.8484848484848486, | |
| "grad_norm": 1.6918838024139404, | |
| "learning_rate": 1.3356667915121025e-07, | |
| "loss": 0.8254, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 2.878787878787879, | |
| "grad_norm": 1.3561716079711914, | |
| "learning_rate": 8.555138626189619e-08, | |
| "loss": 0.668, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "grad_norm": 1.593574047088623, | |
| "learning_rate": 4.815273327803183e-08, | |
| "loss": 0.7856, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.9393939393939394, | |
| "grad_norm": 1.6517926454544067, | |
| "learning_rate": 2.1410767613965212e-08, | |
| "loss": 0.7489, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 2.9696969696969697, | |
| "grad_norm": 1.5703673362731934, | |
| "learning_rate": 5.354125236343155e-09, | |
| "loss": 0.7786, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.5764566659927368, | |
| "learning_rate": 0.0, | |
| "loss": 0.7356, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 99, | |
| "total_flos": 2.3635472472132813e+17, | |
| "train_loss": 1.1949122909343604, | |
| "train_runtime": 1081.5376, | |
| "train_samples_per_second": 2.854, | |
| "train_steps_per_second": 0.092 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 99, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 0, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.3635472472132813e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |