| { | |
| "best_metric": 0.6192739605903625, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-200", | |
| "epoch": 0.07923145488759037, | |
| "eval_steps": 50, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00039615727443795186, | |
| "grad_norm": 14.314478874206543, | |
| "learning_rate": 5e-06, | |
| "loss": 2.5987, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00039615727443795186, | |
| "eval_loss": 1.2088950872421265, | |
| "eval_runtime": 338.2836, | |
| "eval_samples_per_second": 12.569, | |
| "eval_steps_per_second": 6.285, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0007923145488759037, | |
| "grad_norm": 16.324403762817383, | |
| "learning_rate": 1e-05, | |
| "loss": 3.394, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0011884718233138556, | |
| "grad_norm": 18.80040168762207, | |
| "learning_rate": 1.5e-05, | |
| "loss": 3.0517, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0015846290977518075, | |
| "grad_norm": 17.43671989440918, | |
| "learning_rate": 2e-05, | |
| "loss": 3.2647, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0019807863721897595, | |
| "grad_norm": 14.420022010803223, | |
| "learning_rate": 2.5e-05, | |
| "loss": 2.9988, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0023769436466277113, | |
| "grad_norm": 13.038010597229004, | |
| "learning_rate": 3e-05, | |
| "loss": 2.9589, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.002773100921065663, | |
| "grad_norm": 11.008463859558105, | |
| "learning_rate": 3.5e-05, | |
| "loss": 2.6601, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.003169258195503615, | |
| "grad_norm": 11.67863941192627, | |
| "learning_rate": 4e-05, | |
| "loss": 2.6742, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0035654154699415667, | |
| "grad_norm": 12.343779563903809, | |
| "learning_rate": 4.5e-05, | |
| "loss": 2.9556, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.003961572744379519, | |
| "grad_norm": 10.546838760375977, | |
| "learning_rate": 5e-05, | |
| "loss": 2.6015, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.004357730018817471, | |
| "grad_norm": 11.106729507446289, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 2.5997, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.004753887293255423, | |
| "grad_norm": 10.504631996154785, | |
| "learning_rate": 6e-05, | |
| "loss": 3.0646, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.005150044567693374, | |
| "grad_norm": 9.306614875793457, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 2.5052, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.005546201842131326, | |
| "grad_norm": 9.76103401184082, | |
| "learning_rate": 7e-05, | |
| "loss": 2.272, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.005942359116569278, | |
| "grad_norm": 10.521080017089844, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 2.7484, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.00633851639100723, | |
| "grad_norm": 10.383809089660645, | |
| "learning_rate": 8e-05, | |
| "loss": 2.4941, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.006734673665445182, | |
| "grad_norm": 11.771020889282227, | |
| "learning_rate": 8.5e-05, | |
| "loss": 2.7839, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0071308309398831335, | |
| "grad_norm": 10.846237182617188, | |
| "learning_rate": 9e-05, | |
| "loss": 2.4915, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.007526988214321085, | |
| "grad_norm": 9.68586254119873, | |
| "learning_rate": 9.5e-05, | |
| "loss": 2.4326, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.007923145488759038, | |
| "grad_norm": 9.134086608886719, | |
| "learning_rate": 0.0001, | |
| "loss": 2.2999, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.008319302763196989, | |
| "grad_norm": 10.715767860412598, | |
| "learning_rate": 9.999238475781957e-05, | |
| "loss": 2.4463, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.008715460037634942, | |
| "grad_norm": 12.497233390808105, | |
| "learning_rate": 9.99695413509548e-05, | |
| "loss": 2.6965, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.009111617312072893, | |
| "grad_norm": 11.588178634643555, | |
| "learning_rate": 9.99314767377287e-05, | |
| "loss": 2.6387, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.009507774586510845, | |
| "grad_norm": 10.239289283752441, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 2.3716, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.009903931860948796, | |
| "grad_norm": 10.348134994506836, | |
| "learning_rate": 9.980973490458728e-05, | |
| "loss": 2.3706, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.010300089135386749, | |
| "grad_norm": 12.288321495056152, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 2.8137, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.0106962464098247, | |
| "grad_norm": 10.109082221984863, | |
| "learning_rate": 9.962730758206611e-05, | |
| "loss": 2.5793, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.011092403684262652, | |
| "grad_norm": 10.935514450073242, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 2.5139, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.011488560958700603, | |
| "grad_norm": 11.004712104797363, | |
| "learning_rate": 9.938441702975689e-05, | |
| "loss": 2.4287, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.011884718233138556, | |
| "grad_norm": 11.961386680603027, | |
| "learning_rate": 9.924038765061042e-05, | |
| "loss": 2.885, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.012280875507576509, | |
| "grad_norm": 10.97961139678955, | |
| "learning_rate": 9.908135917238321e-05, | |
| "loss": 2.7134, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.01267703278201446, | |
| "grad_norm": 13.087514877319336, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 3.0466, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.013073190056452412, | |
| "grad_norm": 9.89000129699707, | |
| "learning_rate": 9.871850323926177e-05, | |
| "loss": 2.0578, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.013469347330890363, | |
| "grad_norm": 13.183517456054688, | |
| "learning_rate": 9.851478631379982e-05, | |
| "loss": 2.8413, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.013865504605328316, | |
| "grad_norm": 10.432723999023438, | |
| "learning_rate": 9.829629131445342e-05, | |
| "loss": 2.3153, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.014261661879766267, | |
| "grad_norm": 10.83343505859375, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 2.3221, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.01465781915420422, | |
| "grad_norm": 11.4821138381958, | |
| "learning_rate": 9.781523779815179e-05, | |
| "loss": 2.7535, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.01505397642864217, | |
| "grad_norm": 13.579030990600586, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 3.1641, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.015450133703080123, | |
| "grad_norm": 12.277050018310547, | |
| "learning_rate": 9.727592877996585e-05, | |
| "loss": 2.5429, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.015846290977518076, | |
| "grad_norm": 11.074421882629395, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 2.3702, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.016242448251956025, | |
| "grad_norm": 14.457468032836914, | |
| "learning_rate": 9.667902132486009e-05, | |
| "loss": 3.2263, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.016638605526393978, | |
| "grad_norm": 12.959149360656738, | |
| "learning_rate": 9.635919272833938e-05, | |
| "loss": 2.4348, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.01703476280083193, | |
| "grad_norm": 17.107925415039062, | |
| "learning_rate": 9.602524267262203e-05, | |
| "loss": 3.3204, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.017430920075269883, | |
| "grad_norm": 14.488329887390137, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 2.845, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.017827077349707832, | |
| "grad_norm": 12.97142505645752, | |
| "learning_rate": 9.53153893518325e-05, | |
| "loss": 3.277, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.018223234624145785, | |
| "grad_norm": 16.733470916748047, | |
| "learning_rate": 9.493970231495835e-05, | |
| "loss": 3.6097, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.018619391898583738, | |
| "grad_norm": 13.195300102233887, | |
| "learning_rate": 9.45503262094184e-05, | |
| "loss": 2.9709, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.01901554917302169, | |
| "grad_norm": 13.836097717285156, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 2.7804, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.019411706447459643, | |
| "grad_norm": 12.61435317993164, | |
| "learning_rate": 9.373098535696979e-05, | |
| "loss": 2.2369, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.019807863721897592, | |
| "grad_norm": 17.996479034423828, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 3.2496, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.019807863721897592, | |
| "eval_loss": 0.7593479752540588, | |
| "eval_runtime": 340.1536, | |
| "eval_samples_per_second": 12.5, | |
| "eval_steps_per_second": 6.25, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.020204020996335545, | |
| "grad_norm": 14.763004302978516, | |
| "learning_rate": 9.285836503510562e-05, | |
| "loss": 3.4166, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.020600178270773498, | |
| "grad_norm": 8.999519348144531, | |
| "learning_rate": 9.24024048078213e-05, | |
| "loss": 2.2951, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.02099633554521145, | |
| "grad_norm": 7.654372215270996, | |
| "learning_rate": 9.193352839727121e-05, | |
| "loss": 2.9339, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.0213924928196494, | |
| "grad_norm": 6.517328262329102, | |
| "learning_rate": 9.145187862775209e-05, | |
| "loss": 2.4162, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.021788650094087352, | |
| "grad_norm": 7.003291130065918, | |
| "learning_rate": 9.09576022144496e-05, | |
| "loss": 2.2779, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.022184807368525305, | |
| "grad_norm": 8.182329177856445, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 2.9609, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.022580964642963258, | |
| "grad_norm": 7.554233074188232, | |
| "learning_rate": 8.993177550236464e-05, | |
| "loss": 2.6803, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.022977121917401207, | |
| "grad_norm": 7.749049186706543, | |
| "learning_rate": 8.940053768033609e-05, | |
| "loss": 2.5795, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.02337327919183916, | |
| "grad_norm": 7.0886149406433105, | |
| "learning_rate": 8.885729807284856e-05, | |
| "loss": 2.3511, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.023769436466277112, | |
| "grad_norm": 6.84679651260376, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 2.1122, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.024165593740715065, | |
| "grad_norm": 7.1356329917907715, | |
| "learning_rate": 8.773547901113862e-05, | |
| "loss": 2.3196, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.024561751015153017, | |
| "grad_norm": 7.6530046463012695, | |
| "learning_rate": 8.715724127386972e-05, | |
| "loss": 2.4788, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.024957908289590967, | |
| "grad_norm": 8.369495391845703, | |
| "learning_rate": 8.656768508095853e-05, | |
| "loss": 2.3375, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.02535406556402892, | |
| "grad_norm": 8.36403751373291, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 2.0959, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.025750222838466872, | |
| "grad_norm": 10.229287147521973, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 3.0627, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.026146380112904825, | |
| "grad_norm": 8.87305736541748, | |
| "learning_rate": 8.473291852294987e-05, | |
| "loss": 2.5706, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.026542537387342774, | |
| "grad_norm": 9.638481140136719, | |
| "learning_rate": 8.409991800312493e-05, | |
| "loss": 2.7759, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.026938694661780727, | |
| "grad_norm": 9.509964942932129, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 2.5595, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.02733485193621868, | |
| "grad_norm": 9.560708045959473, | |
| "learning_rate": 8.280295144952536e-05, | |
| "loss": 2.6391, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.027731009210656632, | |
| "grad_norm": 8.581311225891113, | |
| "learning_rate": 8.213938048432697e-05, | |
| "loss": 2.784, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.02812716648509458, | |
| "grad_norm": 7.63805627822876, | |
| "learning_rate": 8.146601955249188e-05, | |
| "loss": 2.2859, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.028523323759532534, | |
| "grad_norm": 8.626289367675781, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 2.7975, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.028919481033970486, | |
| "grad_norm": 11.270418167114258, | |
| "learning_rate": 8.009075115760243e-05, | |
| "loss": 2.5275, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.02931563830840844, | |
| "grad_norm": 9.097280502319336, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 2.696, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.02971179558284639, | |
| "grad_norm": 7.862690448760986, | |
| "learning_rate": 7.86788218175523e-05, | |
| "loss": 2.2284, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.03010795285728434, | |
| "grad_norm": 7.738744735717773, | |
| "learning_rate": 7.795964517353735e-05, | |
| "loss": 2.3424, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.030504110131722294, | |
| "grad_norm": 9.578076362609863, | |
| "learning_rate": 7.723195175075136e-05, | |
| "loss": 2.8572, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.030900267406160246, | |
| "grad_norm": 9.798004150390625, | |
| "learning_rate": 7.649596321166024e-05, | |
| "loss": 2.9643, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.0312964246805982, | |
| "grad_norm": 9.317380905151367, | |
| "learning_rate": 7.575190374550272e-05, | |
| "loss": 3.2559, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.03169258195503615, | |
| "grad_norm": 9.332989692687988, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 2.6451, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.032088739229474104, | |
| "grad_norm": 10.110783576965332, | |
| "learning_rate": 7.424048101231686e-05, | |
| "loss": 3.095, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.03248489650391205, | |
| "grad_norm": 9.339815139770508, | |
| "learning_rate": 7.347357813929454e-05, | |
| "loss": 2.3127, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.03288105377835, | |
| "grad_norm": 10.58879280090332, | |
| "learning_rate": 7.269952498697734e-05, | |
| "loss": 2.5449, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.033277211052787956, | |
| "grad_norm": 10.540449142456055, | |
| "learning_rate": 7.191855733945387e-05, | |
| "loss": 3.362, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.03367336832722591, | |
| "grad_norm": 12.278244972229004, | |
| "learning_rate": 7.113091308703498e-05, | |
| "loss": 2.5295, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.03406952560166386, | |
| "grad_norm": 9.675426483154297, | |
| "learning_rate": 7.033683215379002e-05, | |
| "loss": 2.4279, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.034465682876101814, | |
| "grad_norm": 8.322236061096191, | |
| "learning_rate": 6.953655642446368e-05, | |
| "loss": 2.1438, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.034861840150539766, | |
| "grad_norm": 15.984248161315918, | |
| "learning_rate": 6.873032967079561e-05, | |
| "loss": 2.4826, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.03525799742497772, | |
| "grad_norm": 11.099123001098633, | |
| "learning_rate": 6.7918397477265e-05, | |
| "loss": 2.8121, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.035654154699415665, | |
| "grad_norm": 10.237126350402832, | |
| "learning_rate": 6.710100716628344e-05, | |
| "loss": 2.6013, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.03605031197385362, | |
| "grad_norm": 12.89112663269043, | |
| "learning_rate": 6.627840772285784e-05, | |
| "loss": 2.3039, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.03644646924829157, | |
| "grad_norm": 8.39738655090332, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 2.0425, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.03684262652272952, | |
| "grad_norm": 10.789386749267578, | |
| "learning_rate": 6.461858523613684e-05, | |
| "loss": 2.7755, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.037238783797167475, | |
| "grad_norm": 11.95657730102539, | |
| "learning_rate": 6.378186779084995e-05, | |
| "loss": 2.5013, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.03763494107160543, | |
| "grad_norm": 9.726419448852539, | |
| "learning_rate": 6.294095225512603e-05, | |
| "loss": 2.6235, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.03803109834604338, | |
| "grad_norm": 12.768197059631348, | |
| "learning_rate": 6.209609477998338e-05, | |
| "loss": 2.7799, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.03842725562048133, | |
| "grad_norm": 11.749434471130371, | |
| "learning_rate": 6.124755271719325e-05, | |
| "loss": 2.9625, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.038823412894919286, | |
| "grad_norm": 14.08253002166748, | |
| "learning_rate": 6.0395584540887963e-05, | |
| "loss": 2.8971, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.03921957016935723, | |
| "grad_norm": 11.256560325622559, | |
| "learning_rate": 5.9540449768827246e-05, | |
| "loss": 2.3274, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.039615727443795185, | |
| "grad_norm": 12.625090599060059, | |
| "learning_rate": 5.868240888334653e-05, | |
| "loss": 2.6474, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.039615727443795185, | |
| "eval_loss": 0.6730939745903015, | |
| "eval_runtime": 340.0276, | |
| "eval_samples_per_second": 12.505, | |
| "eval_steps_per_second": 6.252, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.04001188471823314, | |
| "grad_norm": 6.478643894195557, | |
| "learning_rate": 5.782172325201155e-05, | |
| "loss": 2.6271, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.04040804199267109, | |
| "grad_norm": 6.2759222984313965, | |
| "learning_rate": 5.695865504800327e-05, | |
| "loss": 2.5561, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.04080419926710904, | |
| "grad_norm": 6.863723278045654, | |
| "learning_rate": 5.6093467170257374e-05, | |
| "loss": 2.5525, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.041200356541546995, | |
| "grad_norm": 5.635458946228027, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 2.254, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.04159651381598495, | |
| "grad_norm": 7.028830528259277, | |
| "learning_rate": 5.435778713738292e-05, | |
| "loss": 2.3026, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.0419926710904229, | |
| "grad_norm": 6.723957061767578, | |
| "learning_rate": 5.348782368720626e-05, | |
| "loss": 2.6904, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.042388828364860846, | |
| "grad_norm": 6.785745620727539, | |
| "learning_rate": 5.26167978121472e-05, | |
| "loss": 2.6896, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.0427849856392988, | |
| "grad_norm": 6.329504489898682, | |
| "learning_rate": 5.174497483512506e-05, | |
| "loss": 2.529, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.04318114291373675, | |
| "grad_norm": 6.532291412353516, | |
| "learning_rate": 5.0872620321864185e-05, | |
| "loss": 2.5266, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.043577300188174704, | |
| "grad_norm": 7.303381443023682, | |
| "learning_rate": 5e-05, | |
| "loss": 2.5342, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.04397345746261266, | |
| "grad_norm": 6.644681930541992, | |
| "learning_rate": 4.912737967813583e-05, | |
| "loss": 2.6088, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.04436961473705061, | |
| "grad_norm": 6.928483963012695, | |
| "learning_rate": 4.825502516487497e-05, | |
| "loss": 2.1421, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.04476577201148856, | |
| "grad_norm": 7.803322792053223, | |
| "learning_rate": 4.738320218785281e-05, | |
| "loss": 2.9061, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.045161929285926515, | |
| "grad_norm": 6.867422580718994, | |
| "learning_rate": 4.6512176312793736e-05, | |
| "loss": 2.3025, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.04555808656036447, | |
| "grad_norm": 7.223367214202881, | |
| "learning_rate": 4.564221286261709e-05, | |
| "loss": 2.4041, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.045954243834802413, | |
| "grad_norm": 19.110706329345703, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 1.9305, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.046350401109240366, | |
| "grad_norm": 6.7067999839782715, | |
| "learning_rate": 4.390653282974264e-05, | |
| "loss": 2.3248, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.04674655838367832, | |
| "grad_norm": 6.43788480758667, | |
| "learning_rate": 4.3041344951996746e-05, | |
| "loss": 1.9726, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.04714271565811627, | |
| "grad_norm": 7.696714401245117, | |
| "learning_rate": 4.2178276747988446e-05, | |
| "loss": 2.5576, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.047538872932554224, | |
| "grad_norm": 9.447258949279785, | |
| "learning_rate": 4.131759111665349e-05, | |
| "loss": 2.7564, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.04793503020699218, | |
| "grad_norm": 7.363338470458984, | |
| "learning_rate": 4.045955023117276e-05, | |
| "loss": 2.5536, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.04833118748143013, | |
| "grad_norm": 8.020831108093262, | |
| "learning_rate": 3.960441545911204e-05, | |
| "loss": 2.913, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.04872734475586808, | |
| "grad_norm": 8.864286422729492, | |
| "learning_rate": 3.875244728280676e-05, | |
| "loss": 2.8491, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.049123502030306035, | |
| "grad_norm": 8.777896881103516, | |
| "learning_rate": 3.790390522001662e-05, | |
| "loss": 2.9723, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.04951965930474398, | |
| "grad_norm": 7.069102764129639, | |
| "learning_rate": 3.705904774487396e-05, | |
| "loss": 2.0808, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.04991581657918193, | |
| "grad_norm": 8.090401649475098, | |
| "learning_rate": 3.6218132209150045e-05, | |
| "loss": 2.6407, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.050311973853619886, | |
| "grad_norm": 9.448684692382812, | |
| "learning_rate": 3.5381414763863166e-05, | |
| "loss": 2.8079, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.05070813112805784, | |
| "grad_norm": 7.581646919250488, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 2.2915, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.05110428840249579, | |
| "grad_norm": 8.184081077575684, | |
| "learning_rate": 3.372159227714218e-05, | |
| "loss": 2.2637, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.051500445676933744, | |
| "grad_norm": 6.465756893157959, | |
| "learning_rate": 3.289899283371657e-05, | |
| "loss": 2.2486, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.0518966029513717, | |
| "grad_norm": 9.846763610839844, | |
| "learning_rate": 3.2081602522734986e-05, | |
| "loss": 2.7469, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.05229276022580965, | |
| "grad_norm": 8.69058895111084, | |
| "learning_rate": 3.12696703292044e-05, | |
| "loss": 2.6517, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.052688917500247595, | |
| "grad_norm": 8.028298377990723, | |
| "learning_rate": 3.046344357553632e-05, | |
| "loss": 2.2827, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.05308507477468555, | |
| "grad_norm": 7.359259605407715, | |
| "learning_rate": 2.9663167846209998e-05, | |
| "loss": 2.0738, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.0534812320491235, | |
| "grad_norm": 10.206085205078125, | |
| "learning_rate": 2.886908691296504e-05, | |
| "loss": 3.022, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.05387738932356145, | |
| "grad_norm": 9.578158378601074, | |
| "learning_rate": 2.8081442660546125e-05, | |
| "loss": 2.643, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.054273546597999406, | |
| "grad_norm": 11.669113159179688, | |
| "learning_rate": 2.7300475013022663e-05, | |
| "loss": 2.6349, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.05466970387243736, | |
| "grad_norm": 10.829512596130371, | |
| "learning_rate": 2.6526421860705473e-05, | |
| "loss": 2.848, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.05506586114687531, | |
| "grad_norm": 10.7781400680542, | |
| "learning_rate": 2.575951898768315e-05, | |
| "loss": 2.3963, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.055462018421313264, | |
| "grad_norm": 9.932387351989746, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 2.9052, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.05585817569575122, | |
| "grad_norm": 9.059712409973145, | |
| "learning_rate": 2.4248096254497288e-05, | |
| "loss": 2.4298, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.05625433297018916, | |
| "grad_norm": 9.161842346191406, | |
| "learning_rate": 2.350403678833976e-05, | |
| "loss": 2.4514, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.056650490244627115, | |
| "grad_norm": 9.905587196350098, | |
| "learning_rate": 2.2768048249248648e-05, | |
| "loss": 2.4208, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.05704664751906507, | |
| "grad_norm": 9.493521690368652, | |
| "learning_rate": 2.2040354826462668e-05, | |
| "loss": 2.4486, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.05744280479350302, | |
| "grad_norm": 11.475030899047852, | |
| "learning_rate": 2.132117818244771e-05, | |
| "loss": 2.9187, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.05783896206794097, | |
| "grad_norm": 9.8174467086792, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 2.1594, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.058235119342378926, | |
| "grad_norm": 13.360136032104492, | |
| "learning_rate": 1.9909248842397584e-05, | |
| "loss": 2.7514, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.05863127661681688, | |
| "grad_norm": 9.616514205932617, | |
| "learning_rate": 1.9216926233717085e-05, | |
| "loss": 2.2589, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.05902743389125483, | |
| "grad_norm": 13.39029312133789, | |
| "learning_rate": 1.8533980447508137e-05, | |
| "loss": 3.0288, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.05942359116569278, | |
| "grad_norm": 14.151657104492188, | |
| "learning_rate": 1.7860619515673033e-05, | |
| "loss": 2.9837, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.05942359116569278, | |
| "eval_loss": 0.6290668845176697, | |
| "eval_runtime": 340.1758, | |
| "eval_samples_per_second": 12.499, | |
| "eval_steps_per_second": 6.25, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.05981974844013073, | |
| "grad_norm": 5.820888519287109, | |
| "learning_rate": 1.7197048550474643e-05, | |
| "loss": 2.1493, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.06021590571456868, | |
| "grad_norm": 5.287855625152588, | |
| "learning_rate": 1.6543469682057106e-05, | |
| "loss": 2.0881, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.060612062989006635, | |
| "grad_norm": 6.305063724517822, | |
| "learning_rate": 1.5900081996875083e-05, | |
| "loss": 2.6538, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.06100822026344459, | |
| "grad_norm": 6.2624101638793945, | |
| "learning_rate": 1.526708147705013e-05, | |
| "loss": 2.6299, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.06140437753788254, | |
| "grad_norm": 6.343729496002197, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 2.9052, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.06180053481232049, | |
| "grad_norm": 6.951181888580322, | |
| "learning_rate": 1.4033009983067452e-05, | |
| "loss": 2.5301, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.062196692086758446, | |
| "grad_norm": 5.954976558685303, | |
| "learning_rate": 1.3432314919041478e-05, | |
| "loss": 2.2918, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.0625928493611964, | |
| "grad_norm": 6.722479343414307, | |
| "learning_rate": 1.2842758726130283e-05, | |
| "loss": 2.3143, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.06298900663563435, | |
| "grad_norm": 6.913514137268066, | |
| "learning_rate": 1.22645209888614e-05, | |
| "loss": 2.6575, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.0633851639100723, | |
| "grad_norm": 7.978066444396973, | |
| "learning_rate": 1.1697777844051105e-05, | |
| "loss": 2.418, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.06378132118451026, | |
| "grad_norm": 8.855539321899414, | |
| "learning_rate": 1.1142701927151456e-05, | |
| "loss": 2.5445, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.06417747845894821, | |
| "grad_norm": 6.6415815353393555, | |
| "learning_rate": 1.0599462319663905e-05, | |
| "loss": 2.385, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.06457363573338615, | |
| "grad_norm": 5.9138360023498535, | |
| "learning_rate": 1.006822449763537e-05, | |
| "loss": 2.0081, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.0649697930078241, | |
| "grad_norm": 6.852816581726074, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 2.3153, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.06536595028226205, | |
| "grad_norm": 7.299200534820557, | |
| "learning_rate": 9.042397785550405e-06, | |
| "loss": 2.7809, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.0657621075567, | |
| "grad_norm": 8.31000804901123, | |
| "learning_rate": 8.548121372247918e-06, | |
| "loss": 2.7937, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.06615826483113796, | |
| "grad_norm": 7.229587554931641, | |
| "learning_rate": 8.066471602728803e-06, | |
| "loss": 2.2101, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.06655442210557591, | |
| "grad_norm": 6.925741672515869, | |
| "learning_rate": 7.597595192178702e-06, | |
| "loss": 2.2659, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.06695057938001386, | |
| "grad_norm": 8.263314247131348, | |
| "learning_rate": 7.1416349648943894e-06, | |
| "loss": 2.4601, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.06734673665445182, | |
| "grad_norm": 7.73466157913208, | |
| "learning_rate": 6.698729810778065e-06, | |
| "loss": 2.7857, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.06774289392888977, | |
| "grad_norm": 8.928947448730469, | |
| "learning_rate": 6.269014643030213e-06, | |
| "loss": 2.8635, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.06813905120332772, | |
| "grad_norm": 7.392125606536865, | |
| "learning_rate": 5.852620357053651e-06, | |
| "loss": 2.0613, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.06853520847776567, | |
| "grad_norm": 7.597949028015137, | |
| "learning_rate": 5.449673790581611e-06, | |
| "loss": 2.5247, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.06893136575220363, | |
| "grad_norm": 8.046714782714844, | |
| "learning_rate": 5.060297685041659e-06, | |
| "loss": 2.5918, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.06932752302664158, | |
| "grad_norm": 6.679835796356201, | |
| "learning_rate": 4.684610648167503e-06, | |
| "loss": 2.0561, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.06972368030107953, | |
| "grad_norm": 7.532476902008057, | |
| "learning_rate": 4.322727117869951e-06, | |
| "loss": 2.5625, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.07011983757551749, | |
| "grad_norm": 9.918126106262207, | |
| "learning_rate": 3.974757327377981e-06, | |
| "loss": 3.47, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.07051599484995544, | |
| "grad_norm": 8.339884757995605, | |
| "learning_rate": 3.6408072716606346e-06, | |
| "loss": 2.3804, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.07091215212439339, | |
| "grad_norm": 7.549655437469482, | |
| "learning_rate": 3.3209786751399187e-06, | |
| "loss": 2.2517, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.07130830939883133, | |
| "grad_norm": 8.312044143676758, | |
| "learning_rate": 3.0153689607045845e-06, | |
| "loss": 2.3566, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.07170446667326928, | |
| "grad_norm": 7.913268566131592, | |
| "learning_rate": 2.724071220034158e-06, | |
| "loss": 2.0139, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.07210062394770723, | |
| "grad_norm": 7.285165786743164, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 2.1933, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.07249678122214519, | |
| "grad_norm": 8.515385627746582, | |
| "learning_rate": 2.1847622018482283e-06, | |
| "loss": 2.2772, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.07289293849658314, | |
| "grad_norm": 10.071023941040039, | |
| "learning_rate": 1.9369152030840556e-06, | |
| "loss": 2.6814, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.07328909577102109, | |
| "grad_norm": 7.0660400390625, | |
| "learning_rate": 1.70370868554659e-06, | |
| "loss": 2.0909, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.07368525304545905, | |
| "grad_norm": 9.54623031616211, | |
| "learning_rate": 1.4852136862001764e-06, | |
| "loss": 2.8344, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.074081410319897, | |
| "grad_norm": 10.024079322814941, | |
| "learning_rate": 1.2814967607382432e-06, | |
| "loss": 2.2405, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.07447756759433495, | |
| "grad_norm": 9.234075546264648, | |
| "learning_rate": 1.0926199633097157e-06, | |
| "loss": 2.2105, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.0748737248687729, | |
| "grad_norm": 8.24449348449707, | |
| "learning_rate": 9.186408276168013e-07, | |
| "loss": 2.0602, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.07526988214321086, | |
| "grad_norm": 9.804146766662598, | |
| "learning_rate": 7.596123493895991e-07, | |
| "loss": 2.1116, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.07566603941764881, | |
| "grad_norm": 9.989800453186035, | |
| "learning_rate": 6.15582970243117e-07, | |
| "loss": 2.707, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.07606219669208676, | |
| "grad_norm": 11.14051628112793, | |
| "learning_rate": 4.865965629214819e-07, | |
| "loss": 2.9818, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.07645835396652471, | |
| "grad_norm": 9.280853271484375, | |
| "learning_rate": 3.7269241793390085e-07, | |
| "loss": 2.3858, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.07685451124096267, | |
| "grad_norm": 9.859801292419434, | |
| "learning_rate": 2.7390523158633554e-07, | |
| "loss": 2.5481, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.07725066851540062, | |
| "grad_norm": 11.552773475646973, | |
| "learning_rate": 1.9026509541272275e-07, | |
| "loss": 2.6877, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.07764682578983857, | |
| "grad_norm": 13.084033012390137, | |
| "learning_rate": 1.2179748700879012e-07, | |
| "loss": 3.1106, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.07804298306427651, | |
| "grad_norm": 12.237486839294434, | |
| "learning_rate": 6.852326227130834e-08, | |
| "loss": 3.2479, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.07843914033871446, | |
| "grad_norm": 12.215850830078125, | |
| "learning_rate": 3.04586490452119e-08, | |
| "loss": 3.1693, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.07883529761315242, | |
| "grad_norm": 11.5962553024292, | |
| "learning_rate": 7.615242180436522e-09, | |
| "loss": 2.6475, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.07923145488759037, | |
| "grad_norm": 12.424742698669434, | |
| "learning_rate": 0.0, | |
| "loss": 3.0626, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.07923145488759037, | |
| "eval_loss": 0.6192739605903625, | |
| "eval_runtime": 339.9709, | |
| "eval_samples_per_second": 12.507, | |
| "eval_steps_per_second": 6.253, | |
| "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": 2.861996017975296e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |