| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.996383363471971, | |
| "eval_steps": 500, | |
| "global_step": 966, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0020666494445879618, | |
| "grad_norm": 4.312192519671265, | |
| "learning_rate": 5.1546391752577325e-08, | |
| "loss": 3.0168, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01033324722293981, | |
| "grad_norm": 4.263469092495139, | |
| "learning_rate": 2.577319587628866e-07, | |
| "loss": 2.9984, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02066649444587962, | |
| "grad_norm": 4.090835140938876, | |
| "learning_rate": 5.154639175257732e-07, | |
| "loss": 2.9949, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.030999741668819428, | |
| "grad_norm": 3.0467308036273013, | |
| "learning_rate": 7.731958762886599e-07, | |
| "loss": 2.9694, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04133298889175924, | |
| "grad_norm": 2.673851155124283, | |
| "learning_rate": 1.0309278350515464e-06, | |
| "loss": 2.9361, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.051666236114699046, | |
| "grad_norm": 2.4196900473976966, | |
| "learning_rate": 1.288659793814433e-06, | |
| "loss": 2.8922, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.061999483337638855, | |
| "grad_norm": 2.194039815147728, | |
| "learning_rate": 1.5463917525773197e-06, | |
| "loss": 2.8477, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07233273056057866, | |
| "grad_norm": 1.816626011260457, | |
| "learning_rate": 1.8041237113402063e-06, | |
| "loss": 2.7875, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08266597778351847, | |
| "grad_norm": 1.8318006255995982, | |
| "learning_rate": 2.061855670103093e-06, | |
| "loss": 2.7757, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09299922500645828, | |
| "grad_norm": 2.114157643853681, | |
| "learning_rate": 2.3195876288659796e-06, | |
| "loss": 2.7489, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10333247222939809, | |
| "grad_norm": 2.023968470692679, | |
| "learning_rate": 2.577319587628866e-06, | |
| "loss": 2.7448, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1136657194523379, | |
| "grad_norm": 1.9532625510646646, | |
| "learning_rate": 2.8350515463917527e-06, | |
| "loss": 2.6986, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.12399896667527771, | |
| "grad_norm": 1.9398256152557316, | |
| "learning_rate": 3.0927835051546395e-06, | |
| "loss": 2.7067, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1343322138982175, | |
| "grad_norm": 2.0411858929451157, | |
| "learning_rate": 3.350515463917526e-06, | |
| "loss": 2.682, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14466546112115733, | |
| "grad_norm": 1.9182095651663165, | |
| "learning_rate": 3.6082474226804126e-06, | |
| "loss": 2.6537, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.15499870834409712, | |
| "grad_norm": 1.8939218115979524, | |
| "learning_rate": 3.865979381443299e-06, | |
| "loss": 2.689, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.16533195556703695, | |
| "grad_norm": 1.8800045882838237, | |
| "learning_rate": 4.123711340206186e-06, | |
| "loss": 2.6106, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.17566520278997674, | |
| "grad_norm": 1.903666598213381, | |
| "learning_rate": 4.381443298969073e-06, | |
| "loss": 2.6385, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.18599845001291657, | |
| "grad_norm": 1.9506327396922627, | |
| "learning_rate": 4.639175257731959e-06, | |
| "loss": 2.644, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.19633169723585636, | |
| "grad_norm": 1.8969467447381583, | |
| "learning_rate": 4.8969072164948455e-06, | |
| "loss": 2.6167, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.20666494445879618, | |
| "grad_norm": 2.063483326690761, | |
| "learning_rate": 4.9998529691533944e-06, | |
| "loss": 2.6128, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.21699819168173598, | |
| "grad_norm": 1.9617956961919498, | |
| "learning_rate": 4.998954509942391e-06, | |
| "loss": 2.5775, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.2273314389046758, | |
| "grad_norm": 1.6301435863496945, | |
| "learning_rate": 4.997239568521342e-06, | |
| "loss": 2.5848, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2376646861276156, | |
| "grad_norm": 2.0118752460927594, | |
| "learning_rate": 4.994708705212238e-06, | |
| "loss": 2.56, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.24799793335055542, | |
| "grad_norm": 1.792769014427482, | |
| "learning_rate": 4.991362746922835e-06, | |
| "loss": 2.583, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.25833118057349524, | |
| "grad_norm": 2.0389735927794557, | |
| "learning_rate": 4.987202786876483e-06, | |
| "loss": 2.5912, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.268664427796435, | |
| "grad_norm": 1.828850605237693, | |
| "learning_rate": 4.982230184254934e-06, | |
| "loss": 2.5457, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.27899767501937484, | |
| "grad_norm": 1.8015379458146892, | |
| "learning_rate": 4.976446563754262e-06, | |
| "loss": 2.5546, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.28933092224231466, | |
| "grad_norm": 1.7199552953613728, | |
| "learning_rate": 4.969853815054022e-06, | |
| "loss": 2.5483, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.2996641694652545, | |
| "grad_norm": 1.5772784625180778, | |
| "learning_rate": 4.962454092199839e-06, | |
| "loss": 2.5046, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.30999741668819425, | |
| "grad_norm": 1.5677450227507446, | |
| "learning_rate": 4.954249812899617e-06, | |
| "loss": 2.5291, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3203306639111341, | |
| "grad_norm": 1.456660409079126, | |
| "learning_rate": 4.9452436577336025e-06, | |
| "loss": 2.514, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.3306639111340739, | |
| "grad_norm": 1.5150592197514972, | |
| "learning_rate": 4.935438569278558e-06, | |
| "loss": 2.483, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3409971583570137, | |
| "grad_norm": 1.3377769703773317, | |
| "learning_rate": 4.924837751146339e-06, | |
| "loss": 2.4891, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3513304055799535, | |
| "grad_norm": 1.2629854028734164, | |
| "learning_rate": 4.913444666937174e-06, | |
| "loss": 2.4873, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.3616636528028933, | |
| "grad_norm": 1.2369841480907526, | |
| "learning_rate": 4.9012630391080105e-06, | |
| "loss": 2.4808, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.37199690002583313, | |
| "grad_norm": 1.2030471624360595, | |
| "learning_rate": 4.8882968477562754e-06, | |
| "loss": 2.452, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3823301472487729, | |
| "grad_norm": 1.2297488483872987, | |
| "learning_rate": 4.874550329319457e-06, | |
| "loss": 2.4817, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.3926633944717127, | |
| "grad_norm": 1.1865278901389154, | |
| "learning_rate": 4.8600279751909394e-06, | |
| "loss": 2.4779, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.40299664169465255, | |
| "grad_norm": 1.135041309414857, | |
| "learning_rate": 4.844734530252532e-06, | |
| "loss": 2.4839, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.41332988891759237, | |
| "grad_norm": 1.1971240713531093, | |
| "learning_rate": 4.8286749913241735e-06, | |
| "loss": 2.4727, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.42366313614053214, | |
| "grad_norm": 1.184700416763506, | |
| "learning_rate": 4.811854605531325e-06, | |
| "loss": 2.4893, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.43399638336347196, | |
| "grad_norm": 1.3114264862654226, | |
| "learning_rate": 4.79427886859058e-06, | |
| "loss": 2.4651, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.4443296305864118, | |
| "grad_norm": 1.2651626968475456, | |
| "learning_rate": 4.775953523014051e-06, | |
| "loss": 2.477, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.4546628778093516, | |
| "grad_norm": 1.1651412485565396, | |
| "learning_rate": 4.75688455623312e-06, | |
| "loss": 2.4875, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4649961250322914, | |
| "grad_norm": 1.183328978373714, | |
| "learning_rate": 4.7370781986421685e-06, | |
| "loss": 2.452, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.4753293722552312, | |
| "grad_norm": 1.2332127650589972, | |
| "learning_rate": 4.716540921562927e-06, | |
| "loss": 2.4488, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.485662619478171, | |
| "grad_norm": 1.1475089285046411, | |
| "learning_rate": 4.6952794351300964e-06, | |
| "loss": 2.463, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.49599586670111084, | |
| "grad_norm": 1.1862533875864938, | |
| "learning_rate": 4.673300686098957e-06, | |
| "loss": 2.4806, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5063291139240507, | |
| "grad_norm": 1.117191578266376, | |
| "learning_rate": 4.65061185557565e-06, | |
| "loss": 2.4535, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5166623611469905, | |
| "grad_norm": 1.1639210815368257, | |
| "learning_rate": 4.627220356670904e-06, | |
| "loss": 2.4536, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5269956083699302, | |
| "grad_norm": 1.1719032781427676, | |
| "learning_rate": 4.603133832077953e-06, | |
| "loss": 2.4366, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.53732885559287, | |
| "grad_norm": 1.1377513004294313, | |
| "learning_rate": 4.57836015157544e-06, | |
| "loss": 2.4524, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5476621028158098, | |
| "grad_norm": 1.1211864633428335, | |
| "learning_rate": 4.552907409456129e-06, | |
| "loss": 2.4202, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5579953500387497, | |
| "grad_norm": 1.1616507123396151, | |
| "learning_rate": 4.526783921882261e-06, | |
| "loss": 2.4273, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5683285972616895, | |
| "grad_norm": 1.1848691721187, | |
| "learning_rate": 4.499998224168417e-06, | |
| "loss": 2.4401, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5786618444846293, | |
| "grad_norm": 1.0582776568433816, | |
| "learning_rate": 4.47255906799278e-06, | |
| "loss": 2.4096, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5889950917075691, | |
| "grad_norm": 1.1638859190983761, | |
| "learning_rate": 4.444475418537707e-06, | |
| "loss": 2.4208, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.599328338930509, | |
| "grad_norm": 1.167542140637952, | |
| "learning_rate": 4.4157564515605315e-06, | |
| "loss": 2.4244, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6096615861534487, | |
| "grad_norm": 1.1838366384063863, | |
| "learning_rate": 4.386411550395576e-06, | |
| "loss": 2.435, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6199948333763885, | |
| "grad_norm": 1.081261572955448, | |
| "learning_rate": 4.356450302888343e-06, | |
| "loss": 2.4268, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6303280805993283, | |
| "grad_norm": 1.1052945392912799, | |
| "learning_rate": 4.32588249826288e-06, | |
| "loss": 2.388, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6406613278222681, | |
| "grad_norm": 1.08835668369336, | |
| "learning_rate": 4.294718123923349e-06, | |
| "loss": 2.3976, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.650994575045208, | |
| "grad_norm": 1.0478651783340451, | |
| "learning_rate": 4.262967362190851e-06, | |
| "loss": 2.4288, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6613278222681478, | |
| "grad_norm": 1.1501225106762079, | |
| "learning_rate": 4.23064058697656e-06, | |
| "loss": 2.4222, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6716610694910876, | |
| "grad_norm": 1.0831191994433091, | |
| "learning_rate": 4.197748360392256e-06, | |
| "loss": 2.4308, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.6819943167140274, | |
| "grad_norm": 1.059007370674017, | |
| "learning_rate": 4.164301429299374e-06, | |
| "loss": 2.3975, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.6923275639369671, | |
| "grad_norm": 1.092943166086201, | |
| "learning_rate": 4.130310721797691e-06, | |
| "loss": 2.4365, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.702660811159907, | |
| "grad_norm": 1.1348461717715252, | |
| "learning_rate": 4.095787343654787e-06, | |
| "loss": 2.4326, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7129940583828468, | |
| "grad_norm": 1.0225410363434102, | |
| "learning_rate": 4.060742574677467e-06, | |
| "loss": 2.4066, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7233273056057866, | |
| "grad_norm": 1.0792570056042612, | |
| "learning_rate": 4.025187865026311e-06, | |
| "loss": 2.4179, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7336605528287264, | |
| "grad_norm": 1.0656972631008976, | |
| "learning_rate": 3.989134831474575e-06, | |
| "loss": 2.3798, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7439938000516663, | |
| "grad_norm": 1.0796066558341972, | |
| "learning_rate": 3.952595253612639e-06, | |
| "loss": 2.4057, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7543270472746061, | |
| "grad_norm": 1.0659792080472803, | |
| "learning_rate": 3.915581069999279e-06, | |
| "loss": 2.4168, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7646602944975458, | |
| "grad_norm": 1.0221090674367013, | |
| "learning_rate": 3.878104374260974e-06, | |
| "loss": 2.4168, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7749935417204856, | |
| "grad_norm": 1.041831551248707, | |
| "learning_rate": 3.840177411140574e-06, | |
| "loss": 2.4064, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.7853267889434254, | |
| "grad_norm": 1.0939898104463206, | |
| "learning_rate": 3.8018125724965727e-06, | |
| "loss": 2.4241, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.7956600361663653, | |
| "grad_norm": 1.041903325778914, | |
| "learning_rate": 3.763022393254321e-06, | |
| "loss": 2.4209, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.8059932833893051, | |
| "grad_norm": 1.0113446521141738, | |
| "learning_rate": 3.723819547310504e-06, | |
| "loss": 2.3965, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8163265306122449, | |
| "grad_norm": 1.0264749571767495, | |
| "learning_rate": 3.6842168433921987e-06, | |
| "loss": 2.3796, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8266597778351847, | |
| "grad_norm": 1.1575314733187445, | |
| "learning_rate": 3.6442272208718873e-06, | |
| "loss": 2.4162, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8369930250581246, | |
| "grad_norm": 1.0237058117990445, | |
| "learning_rate": 3.6038637455397802e-06, | |
| "loss": 2.3887, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8473262722810643, | |
| "grad_norm": 1.0197966280482622, | |
| "learning_rate": 3.5631396053348387e-06, | |
| "loss": 2.3858, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8576595195040041, | |
| "grad_norm": 1.1566498283302307, | |
| "learning_rate": 3.5220681060358865e-06, | |
| "loss": 2.3637, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.8679927667269439, | |
| "grad_norm": 1.0095584831720887, | |
| "learning_rate": 3.4806626669142152e-06, | |
| "loss": 2.4122, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.8783260139498837, | |
| "grad_norm": 1.0666694477391483, | |
| "learning_rate": 3.4389368163491164e-06, | |
| "loss": 2.3899, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.8886592611728236, | |
| "grad_norm": 1.0506847225250229, | |
| "learning_rate": 3.396904187407759e-06, | |
| "loss": 2.3859, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.8989925083957634, | |
| "grad_norm": 1.1509707991212996, | |
| "learning_rate": 3.3545785133908637e-06, | |
| "loss": 2.3521, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9093257556187032, | |
| "grad_norm": 1.1044049179908932, | |
| "learning_rate": 3.3119736233456266e-06, | |
| "loss": 2.3635, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.919659002841643, | |
| "grad_norm": 1.046550942981578, | |
| "learning_rate": 3.269103437547367e-06, | |
| "loss": 2.3529, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.9299922500645827, | |
| "grad_norm": 1.1618469819333466, | |
| "learning_rate": 3.225981962951354e-06, | |
| "loss": 2.3838, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9403254972875226, | |
| "grad_norm": 1.040879291891829, | |
| "learning_rate": 3.182623288616328e-06, | |
| "loss": 2.3778, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9506587445104624, | |
| "grad_norm": 1.0092535371331217, | |
| "learning_rate": 3.139041581101187e-06, | |
| "loss": 2.4014, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9609919917334022, | |
| "grad_norm": 0.9581154923540901, | |
| "learning_rate": 3.0952510798363505e-06, | |
| "loss": 2.3724, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.971325238956342, | |
| "grad_norm": 1.019011267377941, | |
| "learning_rate": 3.0512660924713227e-06, | |
| "loss": 2.3851, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.9816584861792819, | |
| "grad_norm": 1.1025515727643884, | |
| "learning_rate": 3.0071009901999552e-06, | |
| "loss": 2.3533, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.9919917334022217, | |
| "grad_norm": 0.9955871899662452, | |
| "learning_rate": 2.9627702030649596e-06, | |
| "loss": 2.4072, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.9981916817359855, | |
| "eval_loss": 2.4291670322418213, | |
| "eval_runtime": 98.2505, | |
| "eval_samples_per_second": 106.32, | |
| "eval_steps_per_second": 4.438, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 1.0023249806251615, | |
| "grad_norm": 1.043119029342129, | |
| "learning_rate": 2.9182882152431813e-06, | |
| "loss": 2.3632, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.0126582278481013, | |
| "grad_norm": 0.9917731803771155, | |
| "learning_rate": 2.8736695603131953e-06, | |
| "loss": 2.3459, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.0229914750710412, | |
| "grad_norm": 1.059190202939381, | |
| "learning_rate": 2.8289288165067536e-06, | |
| "loss": 2.3504, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.033324722293981, | |
| "grad_norm": 1.047197875332299, | |
| "learning_rate": 2.784080601945651e-06, | |
| "loss": 2.3336, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0436579695169206, | |
| "grad_norm": 1.0722704531569385, | |
| "learning_rate": 2.739139569865547e-06, | |
| "loss": 2.3424, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.0539912167398604, | |
| "grad_norm": 1.0566564902419044, | |
| "learning_rate": 2.6941204038283285e-06, | |
| "loss": 2.3158, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.0643244639628002, | |
| "grad_norm": 0.968455085420159, | |
| "learning_rate": 2.64903781292455e-06, | |
| "loss": 2.3381, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.07465771118574, | |
| "grad_norm": 1.0149829246524174, | |
| "learning_rate": 2.6039065269675428e-06, | |
| "loss": 2.3765, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.0849909584086799, | |
| "grad_norm": 1.003506265130154, | |
| "learning_rate": 2.5587412916807508e-06, | |
| "loss": 2.3619, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.0953242056316197, | |
| "grad_norm": 1.0360443331960967, | |
| "learning_rate": 2.5135568638798647e-06, | |
| "loss": 2.3509, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.1056574528545595, | |
| "grad_norm": 1.03015456481267, | |
| "learning_rate": 2.4683680066513367e-06, | |
| "loss": 2.3532, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.1159907000774993, | |
| "grad_norm": 0.9947690934378423, | |
| "learning_rate": 2.423189484528844e-06, | |
| "loss": 2.3387, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.1263239473004392, | |
| "grad_norm": 0.9935757048613172, | |
| "learning_rate": 2.378036058669279e-06, | |
| "loss": 2.3449, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.136657194523379, | |
| "grad_norm": 1.0218876059810185, | |
| "learning_rate": 2.3329224820298454e-06, | |
| "loss": 2.3454, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.1469904417463188, | |
| "grad_norm": 1.013721944420629, | |
| "learning_rate": 2.287863494547828e-06, | |
| "loss": 2.3202, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.1573236889692586, | |
| "grad_norm": 0.99845936769635, | |
| "learning_rate": 2.242873818324625e-06, | |
| "loss": 2.3581, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.1676569361921985, | |
| "grad_norm": 1.024887085816387, | |
| "learning_rate": 2.1979681528155983e-06, | |
| "loss": 2.3335, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.1779901834151383, | |
| "grad_norm": 0.9636864445536957, | |
| "learning_rate": 2.15316117002733e-06, | |
| "loss": 2.3502, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.188323430638078, | |
| "grad_norm": 0.9607999961987078, | |
| "learning_rate": 2.1084675097238443e-06, | |
| "loss": 2.3295, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.1986566778610177, | |
| "grad_norm": 1.0062266520251049, | |
| "learning_rate": 2.0639017746433636e-06, | |
| "loss": 2.3619, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.2089899250839578, | |
| "grad_norm": 0.9874080934116608, | |
| "learning_rate": 2.0194785257271653e-06, | |
| "loss": 2.316, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.2193231723068974, | |
| "grad_norm": 0.9729847296931, | |
| "learning_rate": 1.9752122773620877e-06, | |
| "loss": 2.3287, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.2296564195298372, | |
| "grad_norm": 1.0507641768918863, | |
| "learning_rate": 1.931117492638257e-06, | |
| "loss": 2.3458, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.239989666752777, | |
| "grad_norm": 0.9964380928331144, | |
| "learning_rate": 1.8872085786235635e-06, | |
| "loss": 2.3299, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.2503229139757168, | |
| "grad_norm": 1.0014215477144741, | |
| "learning_rate": 1.8434998816564525e-06, | |
| "loss": 2.3473, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.2606561611986566, | |
| "grad_norm": 0.9875666662605547, | |
| "learning_rate": 1.800005682658547e-06, | |
| "loss": 2.3187, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.2709894084215965, | |
| "grad_norm": 0.9891384254217107, | |
| "learning_rate": 1.7567401924686512e-06, | |
| "loss": 2.3352, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.2813226556445363, | |
| "grad_norm": 0.9500870275985281, | |
| "learning_rate": 1.7137175471996525e-06, | |
| "loss": 2.3546, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.2916559028674761, | |
| "grad_norm": 0.9864694881362654, | |
| "learning_rate": 1.6709518036198307e-06, | |
| "loss": 2.3573, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.301989150090416, | |
| "grad_norm": 0.9724324214668948, | |
| "learning_rate": 1.628456934560102e-06, | |
| "loss": 2.3226, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.3123223973133558, | |
| "grad_norm": 1.0011381925537814, | |
| "learning_rate": 1.5862468243486783e-06, | |
| "loss": 2.358, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 1.3226556445362956, | |
| "grad_norm": 1.0283092775519347, | |
| "learning_rate": 1.5443352642746517e-06, | |
| "loss": 2.3421, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.3329888917592354, | |
| "grad_norm": 1.0189898038836651, | |
| "learning_rate": 1.5027359480819688e-06, | |
| "loss": 2.3148, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.3433221389821752, | |
| "grad_norm": 0.9802099195113333, | |
| "learning_rate": 1.4614624674952843e-06, | |
| "loss": 2.3327, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.3536553862051148, | |
| "grad_norm": 0.9516385215813331, | |
| "learning_rate": 1.4205283077791393e-06, | |
| "loss": 2.3454, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.3639886334280549, | |
| "grad_norm": 1.0000404439141914, | |
| "learning_rate": 1.3799468433319314e-06, | |
| "loss": 2.3542, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.3743218806509945, | |
| "grad_norm": 1.12899420231106, | |
| "learning_rate": 1.3397313333161007e-06, | |
| "loss": 2.3109, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.3846551278739345, | |
| "grad_norm": 0.9852088893899086, | |
| "learning_rate": 1.2998949173259712e-06, | |
| "loss": 2.364, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.3949883750968741, | |
| "grad_norm": 0.997466272202667, | |
| "learning_rate": 1.2604506110946599e-06, | |
| "loss": 2.3176, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.405321622319814, | |
| "grad_norm": 1.0065697210998936, | |
| "learning_rate": 1.2214113022414448e-06, | |
| "loss": 2.3405, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.4156548695427538, | |
| "grad_norm": 1.0113850437912417, | |
| "learning_rate": 1.1827897460610105e-06, | |
| "loss": 2.3524, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.4259881167656936, | |
| "grad_norm": 1.0142447327303221, | |
| "learning_rate": 1.1445985613559065e-06, | |
| "loss": 2.3289, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.4363213639886334, | |
| "grad_norm": 1.0377572762505685, | |
| "learning_rate": 1.1068502263136207e-06, | |
| "loss": 2.3502, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.4466546112115732, | |
| "grad_norm": 0.9765601945365431, | |
| "learning_rate": 1.0695570744295768e-06, | |
| "loss": 2.3203, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.456987858434513, | |
| "grad_norm": 1.003335350024317, | |
| "learning_rate": 1.032731290477429e-06, | |
| "loss": 2.3425, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.4673211056574529, | |
| "grad_norm": 1.1117543605228124, | |
| "learning_rate": 9.96384906527927e-07, | |
| "loss": 2.3282, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.4776543528803927, | |
| "grad_norm": 1.0082116359749547, | |
| "learning_rate": 9.605297980176905e-07, | |
| "loss": 2.3291, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.4879876001033325, | |
| "grad_norm": 1.0148823844175179, | |
| "learning_rate": 9.251776798691486e-07, | |
| "loss": 2.3322, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.4983208473262724, | |
| "grad_norm": 1.0009464672805306, | |
| "learning_rate": 8.903401026629283e-07, | |
| "loss": 2.3285, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.508654094549212, | |
| "grad_norm": 1.0232374800844972, | |
| "learning_rate": 8.560284488639448e-07, | |
| "loss": 2.33, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.518987341772152, | |
| "grad_norm": 1.00937657376153, | |
| "learning_rate": 8.222539291024079e-07, | |
| "loss": 2.3302, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.5293205889950916, | |
| "grad_norm": 0.9911043964041184, | |
| "learning_rate": 7.890275785109833e-07, | |
| "loss": 2.33, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.5396538362180316, | |
| "grad_norm": 0.9894490010184659, | |
| "learning_rate": 7.563602531192815e-07, | |
| "loss": 2.3083, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.5499870834409712, | |
| "grad_norm": 0.9765968142359513, | |
| "learning_rate": 7.242626263068825e-07, | |
| "loss": 2.3302, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.5603203306639113, | |
| "grad_norm": 0.9785368053784121, | |
| "learning_rate": 6.927451853160186e-07, | |
| "loss": 2.3488, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 1.570653577886851, | |
| "grad_norm": 0.9978746662987256, | |
| "learning_rate": 6.618182278250904e-07, | |
| "loss": 2.3268, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.5809868251097907, | |
| "grad_norm": 0.9671800637222838, | |
| "learning_rate": 6.314918585841026e-07, | |
| "loss": 2.3388, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 1.5913200723327305, | |
| "grad_norm": 1.0258803683412943, | |
| "learning_rate": 6.017759861131511e-07, | |
| "loss": 2.331, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.6016533195556704, | |
| "grad_norm": 1.003633352349549, | |
| "learning_rate": 5.72680319465016e-07, | |
| "loss": 2.348, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.6119865667786102, | |
| "grad_norm": 1.0073802630842454, | |
| "learning_rate": 5.442143650529233e-07, | |
| "loss": 2.33, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.62231981400155, | |
| "grad_norm": 0.982451589918844, | |
| "learning_rate": 5.163874235445248e-07, | |
| "loss": 2.337, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 1.6326530612244898, | |
| "grad_norm": 1.0215920984091988, | |
| "learning_rate": 4.892085868230881e-07, | |
| "loss": 2.3129, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.6429863084474297, | |
| "grad_norm": 0.9661027532299886, | |
| "learning_rate": 4.626867350169184e-07, | |
| "loss": 2.3305, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 1.6533195556703695, | |
| "grad_norm": 1.0107949913173304, | |
| "learning_rate": 4.368305335979514e-07, | |
| "loss": 2.3442, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.663652802893309, | |
| "grad_norm": 1.0338291893124987, | |
| "learning_rate": 4.1164843055049363e-07, | |
| "loss": 2.3306, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 1.6739860501162491, | |
| "grad_norm": 0.9842155309730392, | |
| "learning_rate": 3.871486536110089e-07, | |
| "loss": 2.3353, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.6843192973391887, | |
| "grad_norm": 1.0061784109162568, | |
| "learning_rate": 3.633392075798833e-07, | |
| "loss": 2.3475, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 1.6946525445621288, | |
| "grad_norm": 0.9944727072730667, | |
| "learning_rate": 3.402278717060134e-07, | |
| "loss": 2.3519, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.7049857917850684, | |
| "grad_norm": 1.0111005553437697, | |
| "learning_rate": 3.1782219714509844e-07, | |
| "loss": 2.3457, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.7153190390080084, | |
| "grad_norm": 0.9903500138490017, | |
| "learning_rate": 2.9612950449245727e-07, | |
| "loss": 2.3059, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.725652286230948, | |
| "grad_norm": 1.0201392949034442, | |
| "learning_rate": 2.75156881391167e-07, | |
| "loss": 2.3683, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 1.7359855334538878, | |
| "grad_norm": 1.050898965300813, | |
| "learning_rate": 2.549111802163298e-07, | |
| "loss": 2.3081, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.7463187806768277, | |
| "grad_norm": 1.0078158269898172, | |
| "learning_rate": 2.3539901583619186e-07, | |
| "loss": 2.3309, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 1.7566520278997675, | |
| "grad_norm": 0.9924247873358294, | |
| "learning_rate": 2.1662676345087591e-07, | |
| "loss": 2.3229, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.7669852751227073, | |
| "grad_norm": 0.9670839207903285, | |
| "learning_rate": 1.986005565094104e-07, | |
| "loss": 2.337, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 1.7773185223456471, | |
| "grad_norm": 0.9841412613184868, | |
| "learning_rate": 1.813262847057562e-07, | |
| "loss": 2.3188, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.787651769568587, | |
| "grad_norm": 0.9711799938673707, | |
| "learning_rate": 1.6480959205446483e-07, | |
| "loss": 2.3296, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 1.7979850167915268, | |
| "grad_norm": 0.9602656706710844, | |
| "learning_rate": 1.490558750466145e-07, | |
| "loss": 2.3302, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.8083182640144666, | |
| "grad_norm": 1.0216431924094667, | |
| "learning_rate": 1.3407028088661818e-07, | |
| "loss": 2.3393, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.8186515112374062, | |
| "grad_norm": 0.976140544478714, | |
| "learning_rate": 1.1985770581047662e-07, | |
| "loss": 2.3178, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.8289847584603462, | |
| "grad_norm": 0.9640709438478974, | |
| "learning_rate": 1.0642279348603762e-07, | |
| "loss": 2.332, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 1.8393180056832859, | |
| "grad_norm": 0.9965422182757674, | |
| "learning_rate": 9.376993349576896e-08, | |
| "loss": 2.3361, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.849651252906226, | |
| "grad_norm": 0.9956732940976856, | |
| "learning_rate": 8.421350989479299e-08, | |
| "loss": 2.3355, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 1.8599845001291655, | |
| "grad_norm": 0.9514205830424876, | |
| "learning_rate": 7.297858937089786e-08, | |
| "loss": 2.3176, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.8703177473521055, | |
| "grad_norm": 0.9798917287324781, | |
| "learning_rate": 6.253664838996393e-08, | |
| "loss": 2.3078, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 1.8806509945750451, | |
| "grad_norm": 0.9708517797491476, | |
| "learning_rate": 5.289109864236669e-08, | |
| "loss": 2.3019, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.890984241797985, | |
| "grad_norm": 0.9943948904413487, | |
| "learning_rate": 4.404509161395454e-08, | |
| "loss": 2.3227, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 1.9013174890209248, | |
| "grad_norm": 0.9560070616668404, | |
| "learning_rate": 3.600151755636744e-08, | |
| "loss": 2.3065, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.9116507362438646, | |
| "grad_norm": 0.9882401832339076, | |
| "learning_rate": 2.8763004542704255e-08, | |
| "loss": 2.349, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.9219839834668044, | |
| "grad_norm": 0.9652740208624518, | |
| "learning_rate": 2.233191760885406e-08, | |
| "loss": 2.3353, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.9323172306897443, | |
| "grad_norm": 0.9855071890028472, | |
| "learning_rate": 1.671035798077092e-08, | |
| "loss": 2.3222, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 1.942650477912684, | |
| "grad_norm": 0.9849452175341697, | |
| "learning_rate": 1.1900162387938896e-08, | |
| "loss": 2.3175, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.952983725135624, | |
| "grad_norm": 1.0077538293794988, | |
| "learning_rate": 7.90290246326042e-09, | |
| "loss": 2.3174, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 1.9633169723585637, | |
| "grad_norm": 0.9764187721877715, | |
| "learning_rate": 4.719884229555938e-09, | |
| "loss": 2.3102, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.9736502195815033, | |
| "grad_norm": 1.0135556832497523, | |
| "learning_rate": 2.3521476728480264e-09, | |
| "loss": 2.3504, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 1.9839834668044434, | |
| "grad_norm": 1.0016466106794726, | |
| "learning_rate": 8.004664025679232e-10, | |
| "loss": 2.3299, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.994316714027383, | |
| "grad_norm": 0.9845609766229912, | |
| "learning_rate": 6.534739879382468e-11, | |
| "loss": 2.3391, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 1.996383363471971, | |
| "eval_loss": 2.4173293113708496, | |
| "eval_runtime": 96.9594, | |
| "eval_samples_per_second": 107.736, | |
| "eval_steps_per_second": 4.497, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 1.996383363471971, | |
| "step": 966, | |
| "total_flos": 6353086637408256.0, | |
| "train_loss": 2.421678250374014, | |
| "train_runtime": 6320.1233, | |
| "train_samples_per_second": 29.398, | |
| "train_steps_per_second": 0.153 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 966, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 6353086637408256.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |