| { | |
| "best_metric": 3.230548620223999, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-200", | |
| "epoch": 0.08714596949891068, | |
| "eval_steps": 50, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00043572984749455336, | |
| "grad_norm": 6.714541912078857, | |
| "learning_rate": 5e-06, | |
| "loss": 4.1538, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00043572984749455336, | |
| "eval_loss": 8.217473030090332, | |
| "eval_runtime": 340.3935, | |
| "eval_samples_per_second": 11.357, | |
| "eval_steps_per_second": 5.679, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0008714596949891067, | |
| "grad_norm": 10.93850040435791, | |
| "learning_rate": 1e-05, | |
| "loss": 6.0324, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.00130718954248366, | |
| "grad_norm": 10.759867668151855, | |
| "learning_rate": 1.5e-05, | |
| "loss": 6.2806, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0017429193899782135, | |
| "grad_norm": 8.95274543762207, | |
| "learning_rate": 2e-05, | |
| "loss": 6.2892, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.002178649237472767, | |
| "grad_norm": 8.26896858215332, | |
| "learning_rate": 2.5e-05, | |
| "loss": 6.3382, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.00261437908496732, | |
| "grad_norm": 7.07852840423584, | |
| "learning_rate": 3e-05, | |
| "loss": 6.1125, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0030501089324618735, | |
| "grad_norm": 5.926280975341797, | |
| "learning_rate": 3.5e-05, | |
| "loss": 6.0423, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.003485838779956427, | |
| "grad_norm": 5.2815117835998535, | |
| "learning_rate": 4e-05, | |
| "loss": 5.8223, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.00392156862745098, | |
| "grad_norm": 5.319513320922852, | |
| "learning_rate": 4.5e-05, | |
| "loss": 5.7601, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.004357298474945534, | |
| "grad_norm": 4.76923131942749, | |
| "learning_rate": 5e-05, | |
| "loss": 5.5735, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.004793028322440087, | |
| "grad_norm": 5.165170192718506, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 5.3123, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.00522875816993464, | |
| "grad_norm": 5.180248260498047, | |
| "learning_rate": 6e-05, | |
| "loss": 4.7714, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.005664488017429194, | |
| "grad_norm": 5.849486351013184, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 4.8962, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.006100217864923747, | |
| "grad_norm": 6.7091064453125, | |
| "learning_rate": 7e-05, | |
| "loss": 4.6161, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.006535947712418301, | |
| "grad_norm": 6.763986110687256, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 4.5908, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.006971677559912854, | |
| "grad_norm": 6.989901065826416, | |
| "learning_rate": 8e-05, | |
| "loss": 4.6449, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.007407407407407408, | |
| "grad_norm": 10.466228485107422, | |
| "learning_rate": 8.5e-05, | |
| "loss": 4.9358, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.00784313725490196, | |
| "grad_norm": 8.58240032196045, | |
| "learning_rate": 9e-05, | |
| "loss": 4.2366, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.008278867102396514, | |
| "grad_norm": 6.406922340393066, | |
| "learning_rate": 9.5e-05, | |
| "loss": 3.9878, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.008714596949891068, | |
| "grad_norm": 8.407312393188477, | |
| "learning_rate": 0.0001, | |
| "loss": 3.6146, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.009150326797385621, | |
| "grad_norm": 8.44864559173584, | |
| "learning_rate": 9.999238475781957e-05, | |
| "loss": 4.4908, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.009586056644880174, | |
| "grad_norm": 10.223307609558105, | |
| "learning_rate": 9.99695413509548e-05, | |
| "loss": 4.4297, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.010021786492374727, | |
| "grad_norm": 10.1259765625, | |
| "learning_rate": 9.99314767377287e-05, | |
| "loss": 4.2845, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.01045751633986928, | |
| "grad_norm": 9.56737995147705, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 4.4324, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.010893246187363835, | |
| "grad_norm": 10.609879493713379, | |
| "learning_rate": 9.980973490458728e-05, | |
| "loss": 4.434, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.011328976034858388, | |
| "grad_norm": 10.974640846252441, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 4.7324, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.011764705882352941, | |
| "grad_norm": 10.421369552612305, | |
| "learning_rate": 9.962730758206611e-05, | |
| "loss": 4.1938, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.012200435729847494, | |
| "grad_norm": 9.512422561645508, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 4.1048, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.012636165577342049, | |
| "grad_norm": 9.695697784423828, | |
| "learning_rate": 9.938441702975689e-05, | |
| "loss": 4.1034, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.013071895424836602, | |
| "grad_norm": 9.301131248474121, | |
| "learning_rate": 9.924038765061042e-05, | |
| "loss": 4.0044, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.013507625272331155, | |
| "grad_norm": 10.068182945251465, | |
| "learning_rate": 9.908135917238321e-05, | |
| "loss": 3.793, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.013943355119825708, | |
| "grad_norm": 11.07810115814209, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 3.7727, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.01437908496732026, | |
| "grad_norm": 9.415453910827637, | |
| "learning_rate": 9.871850323926177e-05, | |
| "loss": 3.7048, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.014814814814814815, | |
| "grad_norm": 7.458181381225586, | |
| "learning_rate": 9.851478631379982e-05, | |
| "loss": 3.383, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.015250544662309368, | |
| "grad_norm": 8.064324378967285, | |
| "learning_rate": 9.829629131445342e-05, | |
| "loss": 3.2161, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.01568627450980392, | |
| "grad_norm": 8.875123977661133, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 3.7818, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.016122004357298474, | |
| "grad_norm": 11.600370407104492, | |
| "learning_rate": 9.781523779815179e-05, | |
| "loss": 3.7597, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.016557734204793027, | |
| "grad_norm": 11.134902000427246, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 3.5337, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.01699346405228758, | |
| "grad_norm": 10.188698768615723, | |
| "learning_rate": 9.727592877996585e-05, | |
| "loss": 3.7508, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.017429193899782137, | |
| "grad_norm": 7.492838382720947, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 3.5272, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.01786492374727669, | |
| "grad_norm": 15.345321655273438, | |
| "learning_rate": 9.667902132486009e-05, | |
| "loss": 3.5264, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.018300653594771243, | |
| "grad_norm": 10.750115394592285, | |
| "learning_rate": 9.635919272833938e-05, | |
| "loss": 3.7775, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.018736383442265796, | |
| "grad_norm": 12.272331237792969, | |
| "learning_rate": 9.602524267262203e-05, | |
| "loss": 3.581, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.01917211328976035, | |
| "grad_norm": 10.027256965637207, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 3.4422, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.0196078431372549, | |
| "grad_norm": 12.036170959472656, | |
| "learning_rate": 9.53153893518325e-05, | |
| "loss": 3.3977, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.020043572984749455, | |
| "grad_norm": 13.455428123474121, | |
| "learning_rate": 9.493970231495835e-05, | |
| "loss": 3.8076, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.020479302832244008, | |
| "grad_norm": 21.514856338500977, | |
| "learning_rate": 9.45503262094184e-05, | |
| "loss": 3.4415, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.02091503267973856, | |
| "grad_norm": 16.04561424255371, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 3.4292, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.021350762527233117, | |
| "grad_norm": 26.326793670654297, | |
| "learning_rate": 9.373098535696979e-05, | |
| "loss": 4.6314, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.02178649237472767, | |
| "grad_norm": 35.12826156616211, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 5.3739, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.02178649237472767, | |
| "eval_loss": 6.510286808013916, | |
| "eval_runtime": 342.1833, | |
| "eval_samples_per_second": 11.298, | |
| "eval_steps_per_second": 5.649, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.022222222222222223, | |
| "grad_norm": 35.10013961791992, | |
| "learning_rate": 9.285836503510562e-05, | |
| "loss": 5.0736, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.022657952069716776, | |
| "grad_norm": 9.756641387939453, | |
| "learning_rate": 9.24024048078213e-05, | |
| "loss": 5.1805, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.02309368191721133, | |
| "grad_norm": 5.467214584350586, | |
| "learning_rate": 9.193352839727121e-05, | |
| "loss": 4.7327, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.023529411764705882, | |
| "grad_norm": 5.6564483642578125, | |
| "learning_rate": 9.145187862775209e-05, | |
| "loss": 4.4293, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.023965141612200435, | |
| "grad_norm": 3.8603763580322266, | |
| "learning_rate": 9.09576022144496e-05, | |
| "loss": 4.2583, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.024400871459694988, | |
| "grad_norm": 3.4143786430358887, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 4.0873, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.02483660130718954, | |
| "grad_norm": 3.591214179992676, | |
| "learning_rate": 8.993177550236464e-05, | |
| "loss": 4.1065, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.025272331154684097, | |
| "grad_norm": 3.990115165710449, | |
| "learning_rate": 8.940053768033609e-05, | |
| "loss": 3.9787, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.02570806100217865, | |
| "grad_norm": 3.732379674911499, | |
| "learning_rate": 8.885729807284856e-05, | |
| "loss": 3.9912, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.026143790849673203, | |
| "grad_norm": 4.2087016105651855, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 3.9105, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.026579520697167756, | |
| "grad_norm": 3.774808645248413, | |
| "learning_rate": 8.773547901113862e-05, | |
| "loss": 3.7629, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.02701525054466231, | |
| "grad_norm": 3.8952903747558594, | |
| "learning_rate": 8.715724127386972e-05, | |
| "loss": 3.9626, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.027450980392156862, | |
| "grad_norm": 4.1821160316467285, | |
| "learning_rate": 8.656768508095853e-05, | |
| "loss": 3.7911, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.027886710239651415, | |
| "grad_norm": 4.013959884643555, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 4.0033, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.02832244008714597, | |
| "grad_norm": 4.394720554351807, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 3.7035, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.02875816993464052, | |
| "grad_norm": 4.393320083618164, | |
| "learning_rate": 8.473291852294987e-05, | |
| "loss": 3.6583, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.029193899782135078, | |
| "grad_norm": 5.116196632385254, | |
| "learning_rate": 8.409991800312493e-05, | |
| "loss": 3.4642, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.02962962962962963, | |
| "grad_norm": 5.922344207763672, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 2.4069, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.030065359477124184, | |
| "grad_norm": 5.2199482917785645, | |
| "learning_rate": 8.280295144952536e-05, | |
| "loss": 2.8334, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.030501089324618737, | |
| "grad_norm": 5.466800212860107, | |
| "learning_rate": 8.213938048432697e-05, | |
| "loss": 2.7829, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.03093681917211329, | |
| "grad_norm": 5.415471076965332, | |
| "learning_rate": 8.146601955249188e-05, | |
| "loss": 2.926, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.03137254901960784, | |
| "grad_norm": 5.683098316192627, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 3.3565, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.0318082788671024, | |
| "grad_norm": 7.391642093658447, | |
| "learning_rate": 8.009075115760243e-05, | |
| "loss": 4.1612, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.03224400871459695, | |
| "grad_norm": 7.878474712371826, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 4.0754, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.032679738562091505, | |
| "grad_norm": 7.959073066711426, | |
| "learning_rate": 7.86788218175523e-05, | |
| "loss": 3.7957, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.033115468409586055, | |
| "grad_norm": 9.686983108520508, | |
| "learning_rate": 7.795964517353735e-05, | |
| "loss": 3.7651, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.03355119825708061, | |
| "grad_norm": 8.671168327331543, | |
| "learning_rate": 7.723195175075136e-05, | |
| "loss": 3.4634, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.03398692810457516, | |
| "grad_norm": 7.247128486633301, | |
| "learning_rate": 7.649596321166024e-05, | |
| "loss": 3.139, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.03442265795206972, | |
| "grad_norm": 8.371869087219238, | |
| "learning_rate": 7.575190374550272e-05, | |
| "loss": 3.1679, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.034858387799564274, | |
| "grad_norm": 7.804739952087402, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 3.082, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.03529411764705882, | |
| "grad_norm": 7.40488862991333, | |
| "learning_rate": 7.424048101231686e-05, | |
| "loss": 3.0061, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.03572984749455338, | |
| "grad_norm": 9.724720001220703, | |
| "learning_rate": 7.347357813929454e-05, | |
| "loss": 3.3592, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.03616557734204793, | |
| "grad_norm": 8.739630699157715, | |
| "learning_rate": 7.269952498697734e-05, | |
| "loss": 3.1958, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.036601307189542485, | |
| "grad_norm": 13.883004188537598, | |
| "learning_rate": 7.191855733945387e-05, | |
| "loss": 3.6667, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.037037037037037035, | |
| "grad_norm": 8.60599136352539, | |
| "learning_rate": 7.113091308703498e-05, | |
| "loss": 2.9883, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.03747276688453159, | |
| "grad_norm": 9.525105476379395, | |
| "learning_rate": 7.033683215379002e-05, | |
| "loss": 3.3996, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.03790849673202614, | |
| "grad_norm": 8.75131607055664, | |
| "learning_rate": 6.953655642446368e-05, | |
| "loss": 3.193, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.0383442265795207, | |
| "grad_norm": 7.756675720214844, | |
| "learning_rate": 6.873032967079561e-05, | |
| "loss": 3.2385, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.038779956427015254, | |
| "grad_norm": 8.15986442565918, | |
| "learning_rate": 6.7918397477265e-05, | |
| "loss": 3.0469, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.0392156862745098, | |
| "grad_norm": 9.440335273742676, | |
| "learning_rate": 6.710100716628344e-05, | |
| "loss": 3.2547, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.03965141612200436, | |
| "grad_norm": 7.82032585144043, | |
| "learning_rate": 6.627840772285784e-05, | |
| "loss": 3.0536, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.04008714596949891, | |
| "grad_norm": 8.337102890014648, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 3.3388, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.040522875816993466, | |
| "grad_norm": 9.669166564941406, | |
| "learning_rate": 6.461858523613684e-05, | |
| "loss": 3.6829, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.040958605664488015, | |
| "grad_norm": 8.827813148498535, | |
| "learning_rate": 6.378186779084995e-05, | |
| "loss": 3.2636, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.04139433551198257, | |
| "grad_norm": 7.568974018096924, | |
| "learning_rate": 6.294095225512603e-05, | |
| "loss": 3.064, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.04183006535947712, | |
| "grad_norm": 10.151366233825684, | |
| "learning_rate": 6.209609477998338e-05, | |
| "loss": 3.0729, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.04226579520697168, | |
| "grad_norm": 15.141189575195312, | |
| "learning_rate": 6.124755271719325e-05, | |
| "loss": 3.407, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.042701525054466234, | |
| "grad_norm": 12.019047737121582, | |
| "learning_rate": 6.0395584540887963e-05, | |
| "loss": 3.1992, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.043137254901960784, | |
| "grad_norm": 18.79649543762207, | |
| "learning_rate": 5.9540449768827246e-05, | |
| "loss": 3.5933, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.04357298474945534, | |
| "grad_norm": 13.088361740112305, | |
| "learning_rate": 5.868240888334653e-05, | |
| "loss": 4.619, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.04357298474945534, | |
| "eval_loss": 6.02884578704834, | |
| "eval_runtime": 342.2855, | |
| "eval_samples_per_second": 11.295, | |
| "eval_steps_per_second": 5.647, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.04400871459694989, | |
| "grad_norm": 42.128944396972656, | |
| "learning_rate": 5.782172325201155e-05, | |
| "loss": 5.0307, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.044444444444444446, | |
| "grad_norm": 19.087020874023438, | |
| "learning_rate": 5.695865504800327e-05, | |
| "loss": 4.9041, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.044880174291938996, | |
| "grad_norm": 8.862335205078125, | |
| "learning_rate": 5.6093467170257374e-05, | |
| "loss": 4.3268, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.04531590413943355, | |
| "grad_norm": 5.625029563903809, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 4.1794, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.0457516339869281, | |
| "grad_norm": 4.465734958648682, | |
| "learning_rate": 5.435778713738292e-05, | |
| "loss": 3.8786, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.04618736383442266, | |
| "grad_norm": 4.2806243896484375, | |
| "learning_rate": 5.348782368720626e-05, | |
| "loss": 3.5648, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.046623093681917215, | |
| "grad_norm": 4.626951217651367, | |
| "learning_rate": 5.26167978121472e-05, | |
| "loss": 3.8994, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.047058823529411764, | |
| "grad_norm": 4.95627498626709, | |
| "learning_rate": 5.174497483512506e-05, | |
| "loss": 3.6665, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.04749455337690632, | |
| "grad_norm": 3.974017858505249, | |
| "learning_rate": 5.0872620321864185e-05, | |
| "loss": 3.6386, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.04793028322440087, | |
| "grad_norm": 4.023645877838135, | |
| "learning_rate": 5e-05, | |
| "loss": 3.4637, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.048366013071895426, | |
| "grad_norm": 3.7657175064086914, | |
| "learning_rate": 4.912737967813583e-05, | |
| "loss": 3.6163, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.048801742919389976, | |
| "grad_norm": 4.199164867401123, | |
| "learning_rate": 4.825502516487497e-05, | |
| "loss": 3.4856, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.04923747276688453, | |
| "grad_norm": 4.223781108856201, | |
| "learning_rate": 4.738320218785281e-05, | |
| "loss": 3.4757, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.04967320261437908, | |
| "grad_norm": 4.435927867889404, | |
| "learning_rate": 4.6512176312793736e-05, | |
| "loss": 3.3169, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.05010893246187364, | |
| "grad_norm": 4.904012680053711, | |
| "learning_rate": 4.564221286261709e-05, | |
| "loss": 3.5495, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.050544662309368195, | |
| "grad_norm": 4.705389022827148, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 3.0275, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.050980392156862744, | |
| "grad_norm": 5.674432277679443, | |
| "learning_rate": 4.390653282974264e-05, | |
| "loss": 2.4125, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.0514161220043573, | |
| "grad_norm": 5.154523849487305, | |
| "learning_rate": 4.3041344951996746e-05, | |
| "loss": 2.4193, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.05185185185185185, | |
| "grad_norm": 5.284936904907227, | |
| "learning_rate": 4.2178276747988446e-05, | |
| "loss": 3.0715, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.05228758169934641, | |
| "grad_norm": 5.14494514465332, | |
| "learning_rate": 4.131759111665349e-05, | |
| "loss": 2.6556, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.052723311546840956, | |
| "grad_norm": 5.538849353790283, | |
| "learning_rate": 4.045955023117276e-05, | |
| "loss": 3.0354, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.05315904139433551, | |
| "grad_norm": 7.487570285797119, | |
| "learning_rate": 3.960441545911204e-05, | |
| "loss": 3.9713, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.05359477124183006, | |
| "grad_norm": 8.709990501403809, | |
| "learning_rate": 3.875244728280676e-05, | |
| "loss": 4.0759, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.05403050108932462, | |
| "grad_norm": 8.605923652648926, | |
| "learning_rate": 3.790390522001662e-05, | |
| "loss": 3.5372, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.054466230936819175, | |
| "grad_norm": 8.194428443908691, | |
| "learning_rate": 3.705904774487396e-05, | |
| "loss": 3.3761, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.054901960784313725, | |
| "grad_norm": 8.833706855773926, | |
| "learning_rate": 3.6218132209150045e-05, | |
| "loss": 3.4749, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.05533769063180828, | |
| "grad_norm": 8.356715202331543, | |
| "learning_rate": 3.5381414763863166e-05, | |
| "loss": 3.1606, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.05577342047930283, | |
| "grad_norm": 10.272514343261719, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 2.5981, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.05620915032679739, | |
| "grad_norm": 8.101070404052734, | |
| "learning_rate": 3.372159227714218e-05, | |
| "loss": 3.4551, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.05664488017429194, | |
| "grad_norm": 8.390080451965332, | |
| "learning_rate": 3.289899283371657e-05, | |
| "loss": 3.329, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.05708061002178649, | |
| "grad_norm": 7.999818325042725, | |
| "learning_rate": 3.2081602522734986e-05, | |
| "loss": 3.1951, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.05751633986928104, | |
| "grad_norm": 8.268589973449707, | |
| "learning_rate": 3.12696703292044e-05, | |
| "loss": 3.0559, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.0579520697167756, | |
| "grad_norm": 7.182888984680176, | |
| "learning_rate": 3.046344357553632e-05, | |
| "loss": 2.7636, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.058387799564270156, | |
| "grad_norm": 7.375523567199707, | |
| "learning_rate": 2.9663167846209998e-05, | |
| "loss": 2.9165, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.058823529411764705, | |
| "grad_norm": 7.143980026245117, | |
| "learning_rate": 2.886908691296504e-05, | |
| "loss": 2.835, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.05925925925925926, | |
| "grad_norm": 7.666155815124512, | |
| "learning_rate": 2.8081442660546125e-05, | |
| "loss": 2.9644, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.05969498910675381, | |
| "grad_norm": 8.490532875061035, | |
| "learning_rate": 2.7300475013022663e-05, | |
| "loss": 3.2043, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.06013071895424837, | |
| "grad_norm": 9.01056957244873, | |
| "learning_rate": 2.6526421860705473e-05, | |
| "loss": 3.2254, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.06056644880174292, | |
| "grad_norm": 9.418787956237793, | |
| "learning_rate": 2.575951898768315e-05, | |
| "loss": 3.1616, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.06100217864923747, | |
| "grad_norm": 10.980050086975098, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 2.756, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.06143790849673202, | |
| "grad_norm": 9.275108337402344, | |
| "learning_rate": 2.4248096254497288e-05, | |
| "loss": 3.0247, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.06187363834422658, | |
| "grad_norm": 9.312789916992188, | |
| "learning_rate": 2.350403678833976e-05, | |
| "loss": 3.2638, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.062309368191721136, | |
| "grad_norm": 8.673617362976074, | |
| "learning_rate": 2.2768048249248648e-05, | |
| "loss": 3.0619, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.06274509803921569, | |
| "grad_norm": 11.994880676269531, | |
| "learning_rate": 2.2040354826462668e-05, | |
| "loss": 3.6095, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.06318082788671024, | |
| "grad_norm": 12.651790618896484, | |
| "learning_rate": 2.132117818244771e-05, | |
| "loss": 2.9046, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.0636165577342048, | |
| "grad_norm": 14.072041511535645, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 3.1311, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.06405228758169934, | |
| "grad_norm": 11.299721717834473, | |
| "learning_rate": 1.9909248842397584e-05, | |
| "loss": 3.4079, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.0644880174291939, | |
| "grad_norm": 13.439027786254883, | |
| "learning_rate": 1.9216926233717085e-05, | |
| "loss": 3.3159, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.06492374727668845, | |
| "grad_norm": 21.56414222717285, | |
| "learning_rate": 1.8533980447508137e-05, | |
| "loss": 4.244, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.06535947712418301, | |
| "grad_norm": 19.723522186279297, | |
| "learning_rate": 1.7860619515673033e-05, | |
| "loss": 4.4341, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.06535947712418301, | |
| "eval_loss": 4.00246524810791, | |
| "eval_runtime": 342.244, | |
| "eval_samples_per_second": 11.296, | |
| "eval_steps_per_second": 5.648, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.06579520697167757, | |
| "grad_norm": 4.723827838897705, | |
| "learning_rate": 1.7197048550474643e-05, | |
| "loss": 3.2273, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.06623093681917211, | |
| "grad_norm": 4.662656784057617, | |
| "learning_rate": 1.6543469682057106e-05, | |
| "loss": 3.6659, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.06666666666666667, | |
| "grad_norm": 4.43446159362793, | |
| "learning_rate": 1.5900081996875083e-05, | |
| "loss": 3.5692, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.06710239651416122, | |
| "grad_norm": 4.775651931762695, | |
| "learning_rate": 1.526708147705013e-05, | |
| "loss": 3.5715, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.06753812636165578, | |
| "grad_norm": 3.828355073928833, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 3.5012, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.06797385620915032, | |
| "grad_norm": 3.4921183586120605, | |
| "learning_rate": 1.4033009983067452e-05, | |
| "loss": 3.2915, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.06840958605664488, | |
| "grad_norm": 3.7830638885498047, | |
| "learning_rate": 1.3432314919041478e-05, | |
| "loss": 3.0783, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.06884531590413943, | |
| "grad_norm": 3.9744088649749756, | |
| "learning_rate": 1.2842758726130283e-05, | |
| "loss": 3.4568, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.06928104575163399, | |
| "grad_norm": 4.1922407150268555, | |
| "learning_rate": 1.22645209888614e-05, | |
| "loss": 3.2042, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.06971677559912855, | |
| "grad_norm": 4.034817695617676, | |
| "learning_rate": 1.1697777844051105e-05, | |
| "loss": 3.2243, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.07015250544662309, | |
| "grad_norm": 4.016630172729492, | |
| "learning_rate": 1.1142701927151456e-05, | |
| "loss": 3.2557, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.07058823529411765, | |
| "grad_norm": 4.005847454071045, | |
| "learning_rate": 1.0599462319663905e-05, | |
| "loss": 3.3865, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.0710239651416122, | |
| "grad_norm": 3.8765485286712646, | |
| "learning_rate": 1.006822449763537e-05, | |
| "loss": 3.2679, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.07145969498910676, | |
| "grad_norm": 3.795775890350342, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 3.1198, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.0718954248366013, | |
| "grad_norm": 4.019495010375977, | |
| "learning_rate": 9.042397785550405e-06, | |
| "loss": 3.0646, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.07233115468409586, | |
| "grad_norm": 4.386630058288574, | |
| "learning_rate": 8.548121372247918e-06, | |
| "loss": 3.1476, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.07276688453159041, | |
| "grad_norm": 5.1531243324279785, | |
| "learning_rate": 8.066471602728803e-06, | |
| "loss": 3.281, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.07320261437908497, | |
| "grad_norm": 4.335279941558838, | |
| "learning_rate": 7.597595192178702e-06, | |
| "loss": 2.8631, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.07363834422657953, | |
| "grad_norm": 5.068178176879883, | |
| "learning_rate": 7.1416349648943894e-06, | |
| "loss": 2.6016, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.07407407407407407, | |
| "grad_norm": 5.061690807342529, | |
| "learning_rate": 6.698729810778065e-06, | |
| "loss": 2.7191, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.07450980392156863, | |
| "grad_norm": 5.518496990203857, | |
| "learning_rate": 6.269014643030213e-06, | |
| "loss": 3.0086, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.07494553376906318, | |
| "grad_norm": 6.2861762046813965, | |
| "learning_rate": 5.852620357053651e-06, | |
| "loss": 3.466, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.07538126361655774, | |
| "grad_norm": 11.34403133392334, | |
| "learning_rate": 5.449673790581611e-06, | |
| "loss": 3.7306, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.07581699346405228, | |
| "grad_norm": 12.873065948486328, | |
| "learning_rate": 5.060297685041659e-06, | |
| "loss": 3.6371, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.07625272331154684, | |
| "grad_norm": 11.45556926727295, | |
| "learning_rate": 4.684610648167503e-06, | |
| "loss": 3.4708, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.0766884531590414, | |
| "grad_norm": 13.728697776794434, | |
| "learning_rate": 4.322727117869951e-06, | |
| "loss": 3.6843, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.07712418300653595, | |
| "grad_norm": 12.337279319763184, | |
| "learning_rate": 3.974757327377981e-06, | |
| "loss": 3.1694, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.07755991285403051, | |
| "grad_norm": 11.851096153259277, | |
| "learning_rate": 3.6408072716606346e-06, | |
| "loss": 3.201, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.07799564270152505, | |
| "grad_norm": 13.997617721557617, | |
| "learning_rate": 3.3209786751399187e-06, | |
| "loss": 2.7749, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.0784313725490196, | |
| "grad_norm": 13.09975814819336, | |
| "learning_rate": 3.0153689607045845e-06, | |
| "loss": 3.0505, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.07886710239651416, | |
| "grad_norm": 10.723923683166504, | |
| "learning_rate": 2.724071220034158e-06, | |
| "loss": 2.8997, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.07930283224400872, | |
| "grad_norm": 10.451512336730957, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 2.8682, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.07973856209150326, | |
| "grad_norm": 9.712420463562012, | |
| "learning_rate": 2.1847622018482283e-06, | |
| "loss": 3.1159, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.08017429193899782, | |
| "grad_norm": 9.685064315795898, | |
| "learning_rate": 1.9369152030840556e-06, | |
| "loss": 2.9747, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.08061002178649238, | |
| "grad_norm": 9.208358764648438, | |
| "learning_rate": 1.70370868554659e-06, | |
| "loss": 3.1348, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.08104575163398693, | |
| "grad_norm": 10.657655715942383, | |
| "learning_rate": 1.4852136862001764e-06, | |
| "loss": 3.0165, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.08148148148148149, | |
| "grad_norm": 11.530529022216797, | |
| "learning_rate": 1.2814967607382432e-06, | |
| "loss": 3.1041, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.08191721132897603, | |
| "grad_norm": 10.34518814086914, | |
| "learning_rate": 1.0926199633097157e-06, | |
| "loss": 2.8182, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.08235294117647059, | |
| "grad_norm": 10.75024127960205, | |
| "learning_rate": 9.186408276168013e-07, | |
| "loss": 2.9237, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.08278867102396514, | |
| "grad_norm": 10.584077835083008, | |
| "learning_rate": 7.596123493895991e-07, | |
| "loss": 2.5154, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.0832244008714597, | |
| "grad_norm": 14.239422798156738, | |
| "learning_rate": 6.15582970243117e-07, | |
| "loss": 2.8776, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.08366013071895424, | |
| "grad_norm": 12.499034881591797, | |
| "learning_rate": 4.865965629214819e-07, | |
| "loss": 2.9535, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.0840958605664488, | |
| "grad_norm": 13.673118591308594, | |
| "learning_rate": 3.7269241793390085e-07, | |
| "loss": 3.1236, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.08453159041394336, | |
| "grad_norm": 13.16242790222168, | |
| "learning_rate": 2.7390523158633554e-07, | |
| "loss": 3.1829, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.08496732026143791, | |
| "grad_norm": 13.244391441345215, | |
| "learning_rate": 1.9026509541272275e-07, | |
| "loss": 3.0557, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.08540305010893247, | |
| "grad_norm": 15.455413818359375, | |
| "learning_rate": 1.2179748700879012e-07, | |
| "loss": 3.2272, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.08583877995642701, | |
| "grad_norm": 18.628986358642578, | |
| "learning_rate": 6.852326227130834e-08, | |
| "loss": 2.6964, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.08627450980392157, | |
| "grad_norm": 17.168880462646484, | |
| "learning_rate": 3.04586490452119e-08, | |
| "loss": 3.4646, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.08671023965141612, | |
| "grad_norm": 31.757919311523438, | |
| "learning_rate": 7.615242180436522e-09, | |
| "loss": 3.9619, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.08714596949891068, | |
| "grad_norm": 30.930551528930664, | |
| "learning_rate": 0.0, | |
| "loss": 4.0868, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.08714596949891068, | |
| "eval_loss": 3.230548620223999, | |
| "eval_runtime": 342.1999, | |
| "eval_samples_per_second": 11.297, | |
| "eval_steps_per_second": 5.649, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.9451989028962304e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |