| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.8859229638692638, | |
| "eval_steps": 500, | |
| "global_step": 1900000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.004662752441417178, | |
| "grad_norm": 1.3671598434448242, | |
| "learning_rate": 4.976690900545356e-05, | |
| "loss": 4.8243, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.009325504882834356, | |
| "grad_norm": 1.3568960428237915, | |
| "learning_rate": 4.953384132466932e-05, | |
| "loss": 3.7236, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.013988257324251536, | |
| "grad_norm": 1.4119852781295776, | |
| "learning_rate": 4.930079695764729e-05, | |
| "loss": 3.4332, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 0.018651009765668712, | |
| "grad_norm": 1.1216883659362793, | |
| "learning_rate": 4.906772927686305e-05, | |
| "loss": 3.2825, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 0.023313762207085892, | |
| "grad_norm": 1.2323102951049805, | |
| "learning_rate": 4.883468490984102e-05, | |
| "loss": 3.1866, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 0.02797651464850307, | |
| "grad_norm": 0.9565121531486511, | |
| "learning_rate": 4.860161722905678e-05, | |
| "loss": 3.1126, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 0.03263926708992025, | |
| "grad_norm": 1.1173287630081177, | |
| "learning_rate": 4.836861948955917e-05, | |
| "loss": 3.0577, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 0.037302019531337424, | |
| "grad_norm": 1.4626446962356567, | |
| "learning_rate": 4.813555180877493e-05, | |
| "loss": 3.0145, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 0.041964771972754604, | |
| "grad_norm": 1.0016125440597534, | |
| "learning_rate": 4.790246081422848e-05, | |
| "loss": 2.973, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 0.046627524414171784, | |
| "grad_norm": 1.3417049646377563, | |
| "learning_rate": 4.766941644720645e-05, | |
| "loss": 2.9423, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 0.051290276855588963, | |
| "grad_norm": 1.2270045280456543, | |
| "learning_rate": 4.7436372080184424e-05, | |
| "loss": 2.9127, | |
| "step": 110000 | |
| }, | |
| { | |
| "epoch": 0.05595302929700614, | |
| "grad_norm": 1.0300427675247192, | |
| "learning_rate": 4.7203327713162395e-05, | |
| "loss": 2.8823, | |
| "step": 120000 | |
| }, | |
| { | |
| "epoch": 0.060615781738423316, | |
| "grad_norm": 0.8841068148612976, | |
| "learning_rate": 4.6970283346140366e-05, | |
| "loss": 2.8588, | |
| "step": 130000 | |
| }, | |
| { | |
| "epoch": 0.0652785341798405, | |
| "grad_norm": 1.0524730682373047, | |
| "learning_rate": 4.673721566535613e-05, | |
| "loss": 2.8425, | |
| "step": 140000 | |
| }, | |
| { | |
| "epoch": 0.06994128662125768, | |
| "grad_norm": 0.9874018430709839, | |
| "learning_rate": 4.650417129833409e-05, | |
| "loss": 2.8257, | |
| "step": 150000 | |
| }, | |
| { | |
| "epoch": 0.07460403906267485, | |
| "grad_norm": 0.9634119272232056, | |
| "learning_rate": 4.6271126931312064e-05, | |
| "loss": 2.8114, | |
| "step": 160000 | |
| }, | |
| { | |
| "epoch": 0.07926679150409204, | |
| "grad_norm": 0.885671854019165, | |
| "learning_rate": 4.603808256429003e-05, | |
| "loss": 2.7911, | |
| "step": 170000 | |
| }, | |
| { | |
| "epoch": 0.08392954394550921, | |
| "grad_norm": 1.0135940313339233, | |
| "learning_rate": 4.58050148835058e-05, | |
| "loss": 2.7786, | |
| "step": 180000 | |
| }, | |
| { | |
| "epoch": 0.0885922963869264, | |
| "grad_norm": 1.011932611465454, | |
| "learning_rate": 4.557194720272156e-05, | |
| "loss": 2.7636, | |
| "step": 190000 | |
| }, | |
| { | |
| "epoch": 0.09325504882834357, | |
| "grad_norm": 0.7796096205711365, | |
| "learning_rate": 4.533892614946173e-05, | |
| "loss": 2.753, | |
| "step": 200000 | |
| }, | |
| { | |
| "epoch": 0.09791780126976074, | |
| "grad_norm": 1.1194034814834595, | |
| "learning_rate": 4.5105858468677495e-05, | |
| "loss": 2.7371, | |
| "step": 210000 | |
| }, | |
| { | |
| "epoch": 0.10258055371117793, | |
| "grad_norm": 1.1135520935058594, | |
| "learning_rate": 4.4872814101655467e-05, | |
| "loss": 2.7285, | |
| "step": 220000 | |
| }, | |
| { | |
| "epoch": 0.1072433061525951, | |
| "grad_norm": 0.7772097587585449, | |
| "learning_rate": 4.463976973463344e-05, | |
| "loss": 2.7173, | |
| "step": 230000 | |
| }, | |
| { | |
| "epoch": 0.11190605859401229, | |
| "grad_norm": 1.096358299255371, | |
| "learning_rate": 4.44067020538492e-05, | |
| "loss": 2.7141, | |
| "step": 240000 | |
| }, | |
| { | |
| "epoch": 0.11656881103542946, | |
| "grad_norm": 0.8112640380859375, | |
| "learning_rate": 4.417363437306496e-05, | |
| "loss": 2.7073, | |
| "step": 250000 | |
| }, | |
| { | |
| "epoch": 0.12123156347684663, | |
| "grad_norm": 1.0931545495986938, | |
| "learning_rate": 4.394059000604293e-05, | |
| "loss": 2.6931, | |
| "step": 260000 | |
| }, | |
| { | |
| "epoch": 0.1258943159182638, | |
| "grad_norm": 1.1369918584823608, | |
| "learning_rate": 4.37075456390209e-05, | |
| "loss": 2.6824, | |
| "step": 270000 | |
| }, | |
| { | |
| "epoch": 0.130557068359681, | |
| "grad_norm": 1.258300542831421, | |
| "learning_rate": 4.347450127199887e-05, | |
| "loss": 2.6741, | |
| "step": 280000 | |
| }, | |
| { | |
| "epoch": 0.13521982080109818, | |
| "grad_norm": 0.9752686023712158, | |
| "learning_rate": 4.324143359121463e-05, | |
| "loss": 2.6645, | |
| "step": 290000 | |
| }, | |
| { | |
| "epoch": 0.13988257324251535, | |
| "grad_norm": 1.0001367330551147, | |
| "learning_rate": 4.3008412537954805e-05, | |
| "loss": 2.6592, | |
| "step": 300000 | |
| }, | |
| { | |
| "epoch": 0.14454532568393252, | |
| "grad_norm": 1.0314422845840454, | |
| "learning_rate": 4.277534485717057e-05, | |
| "loss": 2.6523, | |
| "step": 310000 | |
| }, | |
| { | |
| "epoch": 0.1492080781253497, | |
| "grad_norm": 0.9287506937980652, | |
| "learning_rate": 4.254230049014854e-05, | |
| "loss": 2.6503, | |
| "step": 320000 | |
| }, | |
| { | |
| "epoch": 0.1538708305667669, | |
| "grad_norm": 0.8209073543548584, | |
| "learning_rate": 4.23092328093643e-05, | |
| "loss": 2.6379, | |
| "step": 330000 | |
| }, | |
| { | |
| "epoch": 0.15853358300818407, | |
| "grad_norm": 0.8727386593818665, | |
| "learning_rate": 4.207618844234227e-05, | |
| "loss": 2.6332, | |
| "step": 340000 | |
| }, | |
| { | |
| "epoch": 0.16319633544960124, | |
| "grad_norm": 0.9841961860656738, | |
| "learning_rate": 4.184314407532024e-05, | |
| "loss": 2.6279, | |
| "step": 350000 | |
| }, | |
| { | |
| "epoch": 0.16785908789101842, | |
| "grad_norm": 0.7831237316131592, | |
| "learning_rate": 4.1610099708298214e-05, | |
| "loss": 2.6237, | |
| "step": 360000 | |
| }, | |
| { | |
| "epoch": 0.1725218403324356, | |
| "grad_norm": 0.9184048175811768, | |
| "learning_rate": 4.137707865503839e-05, | |
| "loss": 2.6167, | |
| "step": 370000 | |
| }, | |
| { | |
| "epoch": 0.1771845927738528, | |
| "grad_norm": 0.9598727822303772, | |
| "learning_rate": 4.114401097425415e-05, | |
| "loss": 2.6082, | |
| "step": 380000 | |
| }, | |
| { | |
| "epoch": 0.18184734521526996, | |
| "grad_norm": 0.8814136981964111, | |
| "learning_rate": 4.0910966607232115e-05, | |
| "loss": 2.6027, | |
| "step": 390000 | |
| }, | |
| { | |
| "epoch": 0.18651009765668713, | |
| "grad_norm": 0.9080318212509155, | |
| "learning_rate": 4.067789892644788e-05, | |
| "loss": 2.6014, | |
| "step": 400000 | |
| }, | |
| { | |
| "epoch": 0.1911728500981043, | |
| "grad_norm": 0.8321977257728577, | |
| "learning_rate": 4.044485455942585e-05, | |
| "loss": 2.5964, | |
| "step": 410000 | |
| }, | |
| { | |
| "epoch": 0.19583560253952148, | |
| "grad_norm": 1.3375693559646606, | |
| "learning_rate": 4.021178687864161e-05, | |
| "loss": 2.5906, | |
| "step": 420000 | |
| }, | |
| { | |
| "epoch": 0.20049835498093868, | |
| "grad_norm": 0.8211286067962646, | |
| "learning_rate": 3.997874251161958e-05, | |
| "loss": 2.5866, | |
| "step": 430000 | |
| }, | |
| { | |
| "epoch": 0.20516110742235585, | |
| "grad_norm": 0.7890422940254211, | |
| "learning_rate": 3.9745698144597546e-05, | |
| "loss": 2.5817, | |
| "step": 440000 | |
| }, | |
| { | |
| "epoch": 0.20982385986377303, | |
| "grad_norm": 1.0580294132232666, | |
| "learning_rate": 3.951263046381331e-05, | |
| "loss": 2.5776, | |
| "step": 450000 | |
| }, | |
| { | |
| "epoch": 0.2144866123051902, | |
| "grad_norm": 1.0666168928146362, | |
| "learning_rate": 3.927956278302907e-05, | |
| "loss": 2.5729, | |
| "step": 460000 | |
| }, | |
| { | |
| "epoch": 0.21914936474660737, | |
| "grad_norm": 1.0440067052841187, | |
| "learning_rate": 3.904651841600704e-05, | |
| "loss": 2.5748, | |
| "step": 470000 | |
| }, | |
| { | |
| "epoch": 0.22381211718802457, | |
| "grad_norm": 0.8746099472045898, | |
| "learning_rate": 3.88134507352228e-05, | |
| "loss": 2.5704, | |
| "step": 480000 | |
| }, | |
| { | |
| "epoch": 0.22847486962944175, | |
| "grad_norm": 0.882897675037384, | |
| "learning_rate": 3.858040636820078e-05, | |
| "loss": 2.5623, | |
| "step": 490000 | |
| }, | |
| { | |
| "epoch": 0.23313762207085892, | |
| "grad_norm": 0.8458369970321655, | |
| "learning_rate": 3.834740862870316e-05, | |
| "loss": 2.5612, | |
| "step": 500000 | |
| }, | |
| { | |
| "epoch": 0.2378003745122761, | |
| "grad_norm": 0.9579658508300781, | |
| "learning_rate": 3.811434094791892e-05, | |
| "loss": 2.5551, | |
| "step": 510000 | |
| }, | |
| { | |
| "epoch": 0.24246312695369326, | |
| "grad_norm": 1.0498722791671753, | |
| "learning_rate": 3.78813198946591e-05, | |
| "loss": 2.5502, | |
| "step": 520000 | |
| }, | |
| { | |
| "epoch": 0.24712587939511046, | |
| "grad_norm": 1.032334804534912, | |
| "learning_rate": 3.764825221387486e-05, | |
| "loss": 2.5534, | |
| "step": 530000 | |
| }, | |
| { | |
| "epoch": 0.2517886318365276, | |
| "grad_norm": 0.9145790934562683, | |
| "learning_rate": 3.7415184533090624e-05, | |
| "loss": 2.547, | |
| "step": 540000 | |
| }, | |
| { | |
| "epoch": 0.2564513842779448, | |
| "grad_norm": 1.0633904933929443, | |
| "learning_rate": 3.71821634798308e-05, | |
| "loss": 2.543, | |
| "step": 550000 | |
| }, | |
| { | |
| "epoch": 0.261114136719362, | |
| "grad_norm": 0.9828123450279236, | |
| "learning_rate": 3.694911911280877e-05, | |
| "loss": 2.5398, | |
| "step": 560000 | |
| }, | |
| { | |
| "epoch": 0.2657768891607792, | |
| "grad_norm": 0.8735861778259277, | |
| "learning_rate": 3.671607474578674e-05, | |
| "loss": 2.5345, | |
| "step": 570000 | |
| }, | |
| { | |
| "epoch": 0.27043964160219636, | |
| "grad_norm": 1.1347264051437378, | |
| "learning_rate": 3.6483053692526915e-05, | |
| "loss": 2.5312, | |
| "step": 580000 | |
| }, | |
| { | |
| "epoch": 0.27510239404361353, | |
| "grad_norm": 0.8275557160377502, | |
| "learning_rate": 3.625000932550489e-05, | |
| "loss": 2.5299, | |
| "step": 590000 | |
| }, | |
| { | |
| "epoch": 0.2797651464850307, | |
| "grad_norm": 0.8891148567199707, | |
| "learning_rate": 3.601696495848285e-05, | |
| "loss": 2.5301, | |
| "step": 600000 | |
| }, | |
| { | |
| "epoch": 0.2844278989264479, | |
| "grad_norm": 0.9649547338485718, | |
| "learning_rate": 3.578389727769861e-05, | |
| "loss": 2.5304, | |
| "step": 610000 | |
| }, | |
| { | |
| "epoch": 0.28909065136786505, | |
| "grad_norm": 1.0385549068450928, | |
| "learning_rate": 3.5550876224438794e-05, | |
| "loss": 2.5179, | |
| "step": 620000 | |
| }, | |
| { | |
| "epoch": 0.2937534038092822, | |
| "grad_norm": 1.1167703866958618, | |
| "learning_rate": 3.5317785229892346e-05, | |
| "loss": 2.5168, | |
| "step": 630000 | |
| }, | |
| { | |
| "epoch": 0.2984161562506994, | |
| "grad_norm": 0.8988041281700134, | |
| "learning_rate": 3.508476417663252e-05, | |
| "loss": 2.519, | |
| "step": 640000 | |
| }, | |
| { | |
| "epoch": 0.30307890869211657, | |
| "grad_norm": 1.2292715311050415, | |
| "learning_rate": 3.485169649584828e-05, | |
| "loss": 2.5102, | |
| "step": 650000 | |
| }, | |
| { | |
| "epoch": 0.3077416611335338, | |
| "grad_norm": 0.9778872728347778, | |
| "learning_rate": 3.4618652128826254e-05, | |
| "loss": 2.5115, | |
| "step": 660000 | |
| }, | |
| { | |
| "epoch": 0.31240441357495097, | |
| "grad_norm": 0.9290140271186829, | |
| "learning_rate": 3.4385584448042016e-05, | |
| "loss": 2.5068, | |
| "step": 670000 | |
| }, | |
| { | |
| "epoch": 0.31706716601636814, | |
| "grad_norm": 0.8549484610557556, | |
| "learning_rate": 3.415254008101999e-05, | |
| "loss": 2.5045, | |
| "step": 680000 | |
| }, | |
| { | |
| "epoch": 0.3217299184577853, | |
| "grad_norm": 0.9482207894325256, | |
| "learning_rate": 3.391949571399796e-05, | |
| "loss": 2.5002, | |
| "step": 690000 | |
| }, | |
| { | |
| "epoch": 0.3263926708992025, | |
| "grad_norm": 0.9213985204696655, | |
| "learning_rate": 3.368642803321372e-05, | |
| "loss": 2.499, | |
| "step": 700000 | |
| }, | |
| { | |
| "epoch": 0.33105542334061966, | |
| "grad_norm": 0.9165827631950378, | |
| "learning_rate": 3.345338366619169e-05, | |
| "loss": 2.4978, | |
| "step": 710000 | |
| }, | |
| { | |
| "epoch": 0.33571817578203683, | |
| "grad_norm": 0.9712676405906677, | |
| "learning_rate": 3.3220315985407453e-05, | |
| "loss": 2.4923, | |
| "step": 720000 | |
| }, | |
| { | |
| "epoch": 0.340380928223454, | |
| "grad_norm": 0.9476392865180969, | |
| "learning_rate": 3.2987271618385425e-05, | |
| "loss": 2.4921, | |
| "step": 730000 | |
| }, | |
| { | |
| "epoch": 0.3450436806648712, | |
| "grad_norm": 0.8675413727760315, | |
| "learning_rate": 3.275422725136339e-05, | |
| "loss": 2.4913, | |
| "step": 740000 | |
| }, | |
| { | |
| "epoch": 0.34970643310628835, | |
| "grad_norm": 1.0163840055465698, | |
| "learning_rate": 3.252115957057915e-05, | |
| "loss": 2.4876, | |
| "step": 750000 | |
| }, | |
| { | |
| "epoch": 0.3543691855477056, | |
| "grad_norm": 0.8246685862541199, | |
| "learning_rate": 3.228809188979491e-05, | |
| "loss": 2.4848, | |
| "step": 760000 | |
| }, | |
| { | |
| "epoch": 0.35903193798912275, | |
| "grad_norm": 1.1230520009994507, | |
| "learning_rate": 3.205507083653509e-05, | |
| "loss": 2.4856, | |
| "step": 770000 | |
| }, | |
| { | |
| "epoch": 0.3636946904305399, | |
| "grad_norm": 1.0105488300323486, | |
| "learning_rate": 3.182200315575085e-05, | |
| "loss": 2.4815, | |
| "step": 780000 | |
| }, | |
| { | |
| "epoch": 0.3683574428719571, | |
| "grad_norm": 1.0298407077789307, | |
| "learning_rate": 3.158893547496662e-05, | |
| "loss": 2.4801, | |
| "step": 790000 | |
| }, | |
| { | |
| "epoch": 0.37302019531337427, | |
| "grad_norm": 1.1969521045684814, | |
| "learning_rate": 3.135589110794458e-05, | |
| "loss": 2.4776, | |
| "step": 800000 | |
| }, | |
| { | |
| "epoch": 0.37768294775479144, | |
| "grad_norm": 0.9831893444061279, | |
| "learning_rate": 3.112287005468476e-05, | |
| "loss": 2.4787, | |
| "step": 810000 | |
| }, | |
| { | |
| "epoch": 0.3823457001962086, | |
| "grad_norm": 1.2115819454193115, | |
| "learning_rate": 3.0889825687662735e-05, | |
| "loss": 2.4721, | |
| "step": 820000 | |
| }, | |
| { | |
| "epoch": 0.3870084526376258, | |
| "grad_norm": 1.0241882801055908, | |
| "learning_rate": 3.06567813206407e-05, | |
| "loss": 2.4714, | |
| "step": 830000 | |
| }, | |
| { | |
| "epoch": 0.39167120507904296, | |
| "grad_norm": 1.0166583061218262, | |
| "learning_rate": 3.042373695361867e-05, | |
| "loss": 2.4736, | |
| "step": 840000 | |
| }, | |
| { | |
| "epoch": 0.39633395752046013, | |
| "grad_norm": 0.9006229639053345, | |
| "learning_rate": 3.0190645959072226e-05, | |
| "loss": 2.4678, | |
| "step": 850000 | |
| }, | |
| { | |
| "epoch": 0.40099670996187736, | |
| "grad_norm": 1.0126405954360962, | |
| "learning_rate": 2.9957624905812404e-05, | |
| "loss": 2.4651, | |
| "step": 860000 | |
| }, | |
| { | |
| "epoch": 0.40565946240329454, | |
| "grad_norm": 1.3557987213134766, | |
| "learning_rate": 2.972458053879037e-05, | |
| "loss": 2.466, | |
| "step": 870000 | |
| }, | |
| { | |
| "epoch": 0.4103222148447117, | |
| "grad_norm": 1.241245150566101, | |
| "learning_rate": 2.9491559485530546e-05, | |
| "loss": 2.4611, | |
| "step": 880000 | |
| }, | |
| { | |
| "epoch": 0.4149849672861289, | |
| "grad_norm": 1.0998499393463135, | |
| "learning_rate": 2.9258491804746315e-05, | |
| "loss": 2.4618, | |
| "step": 890000 | |
| }, | |
| { | |
| "epoch": 0.41964771972754605, | |
| "grad_norm": 0.9139440655708313, | |
| "learning_rate": 2.902544743772428e-05, | |
| "loss": 2.461, | |
| "step": 900000 | |
| }, | |
| { | |
| "epoch": 0.4243104721689632, | |
| "grad_norm": 0.8467987179756165, | |
| "learning_rate": 2.8792426384464453e-05, | |
| "loss": 2.4584, | |
| "step": 910000 | |
| }, | |
| { | |
| "epoch": 0.4289732246103804, | |
| "grad_norm": 1.090574860572815, | |
| "learning_rate": 2.8559335389918012e-05, | |
| "loss": 2.4567, | |
| "step": 920000 | |
| }, | |
| { | |
| "epoch": 0.43363597705179757, | |
| "grad_norm": 1.063254714012146, | |
| "learning_rate": 2.832629102289598e-05, | |
| "loss": 2.4572, | |
| "step": 930000 | |
| }, | |
| { | |
| "epoch": 0.43829872949321474, | |
| "grad_norm": 0.962209165096283, | |
| "learning_rate": 2.8093223342111742e-05, | |
| "loss": 2.4525, | |
| "step": 940000 | |
| }, | |
| { | |
| "epoch": 0.4429614819346319, | |
| "grad_norm": 0.9544676542282104, | |
| "learning_rate": 2.7860202288851923e-05, | |
| "loss": 2.4518, | |
| "step": 950000 | |
| }, | |
| { | |
| "epoch": 0.44762423437604915, | |
| "grad_norm": 1.0543245077133179, | |
| "learning_rate": 2.7627134608067685e-05, | |
| "loss": 2.4423, | |
| "step": 960000 | |
| }, | |
| { | |
| "epoch": 0.4522869868174663, | |
| "grad_norm": 1.1137259006500244, | |
| "learning_rate": 2.7394066927283447e-05, | |
| "loss": 2.445, | |
| "step": 970000 | |
| }, | |
| { | |
| "epoch": 0.4569497392588835, | |
| "grad_norm": 0.8332359790802002, | |
| "learning_rate": 2.716102256026141e-05, | |
| "loss": 2.4483, | |
| "step": 980000 | |
| }, | |
| { | |
| "epoch": 0.46161249170030066, | |
| "grad_norm": 0.9382948875427246, | |
| "learning_rate": 2.6927954879477173e-05, | |
| "loss": 2.4451, | |
| "step": 990000 | |
| }, | |
| { | |
| "epoch": 0.46627524414171784, | |
| "grad_norm": 0.8949111700057983, | |
| "learning_rate": 2.6694910512455145e-05, | |
| "loss": 2.4431, | |
| "step": 1000000 | |
| }, | |
| { | |
| "epoch": 0.470937996583135, | |
| "grad_norm": 0.8998690843582153, | |
| "learning_rate": 2.6461866145433116e-05, | |
| "loss": 2.4419, | |
| "step": 1010000 | |
| }, | |
| { | |
| "epoch": 0.4756007490245522, | |
| "grad_norm": 0.9611771702766418, | |
| "learning_rate": 2.6228821778411084e-05, | |
| "loss": 2.4425, | |
| "step": 1020000 | |
| }, | |
| { | |
| "epoch": 0.48026350146596936, | |
| "grad_norm": 1.0314167737960815, | |
| "learning_rate": 2.5995754097626846e-05, | |
| "loss": 2.4393, | |
| "step": 1030000 | |
| }, | |
| { | |
| "epoch": 0.48492625390738653, | |
| "grad_norm": 1.0637314319610596, | |
| "learning_rate": 2.5762709730604817e-05, | |
| "loss": 2.4373, | |
| "step": 1040000 | |
| }, | |
| { | |
| "epoch": 0.4895890063488037, | |
| "grad_norm": 1.1488375663757324, | |
| "learning_rate": 2.552964204982058e-05, | |
| "loss": 2.4348, | |
| "step": 1050000 | |
| }, | |
| { | |
| "epoch": 0.49425175879022093, | |
| "grad_norm": 1.1408883333206177, | |
| "learning_rate": 2.529657436903634e-05, | |
| "loss": 2.4352, | |
| "step": 1060000 | |
| }, | |
| { | |
| "epoch": 0.4989145112316381, | |
| "grad_norm": 0.9368888735771179, | |
| "learning_rate": 2.5063530002014312e-05, | |
| "loss": 2.4321, | |
| "step": 1070000 | |
| }, | |
| { | |
| "epoch": 0.5035772636730552, | |
| "grad_norm": 0.9766519069671631, | |
| "learning_rate": 2.483046232123007e-05, | |
| "loss": 2.4319, | |
| "step": 1080000 | |
| }, | |
| { | |
| "epoch": 0.5082400161144724, | |
| "grad_norm": 0.9681417942047119, | |
| "learning_rate": 2.4597417954208042e-05, | |
| "loss": 2.4329, | |
| "step": 1090000 | |
| }, | |
| { | |
| "epoch": 0.5129027685558896, | |
| "grad_norm": 1.0913608074188232, | |
| "learning_rate": 2.4364373587186014e-05, | |
| "loss": 2.428, | |
| "step": 1100000 | |
| }, | |
| { | |
| "epoch": 0.5175655209973068, | |
| "grad_norm": 0.9350466132164001, | |
| "learning_rate": 2.413132922016398e-05, | |
| "loss": 2.4241, | |
| "step": 1110000 | |
| }, | |
| { | |
| "epoch": 0.522228273438724, | |
| "grad_norm": 0.8676067590713501, | |
| "learning_rate": 2.3898261539379743e-05, | |
| "loss": 2.4227, | |
| "step": 1120000 | |
| }, | |
| { | |
| "epoch": 0.5268910258801411, | |
| "grad_norm": 0.8786779046058655, | |
| "learning_rate": 2.366521717235771e-05, | |
| "loss": 2.4208, | |
| "step": 1130000 | |
| }, | |
| { | |
| "epoch": 0.5315537783215584, | |
| "grad_norm": 0.9234575629234314, | |
| "learning_rate": 2.3432126177811267e-05, | |
| "loss": 2.4249, | |
| "step": 1140000 | |
| }, | |
| { | |
| "epoch": 0.5362165307629755, | |
| "grad_norm": 1.3854731321334839, | |
| "learning_rate": 2.3199105124551445e-05, | |
| "loss": 2.4245, | |
| "step": 1150000 | |
| }, | |
| { | |
| "epoch": 0.5408792832043927, | |
| "grad_norm": 1.0635942220687866, | |
| "learning_rate": 2.2966060757529416e-05, | |
| "loss": 2.4202, | |
| "step": 1160000 | |
| }, | |
| { | |
| "epoch": 0.5455420356458098, | |
| "grad_norm": 0.8858787417411804, | |
| "learning_rate": 2.2732993076745178e-05, | |
| "loss": 2.418, | |
| "step": 1170000 | |
| }, | |
| { | |
| "epoch": 0.5502047880872271, | |
| "grad_norm": 1.309348702430725, | |
| "learning_rate": 2.2499948709723146e-05, | |
| "loss": 2.4148, | |
| "step": 1180000 | |
| }, | |
| { | |
| "epoch": 0.5548675405286442, | |
| "grad_norm": 0.9109322428703308, | |
| "learning_rate": 2.2266904342701114e-05, | |
| "loss": 2.4151, | |
| "step": 1190000 | |
| }, | |
| { | |
| "epoch": 0.5595302929700614, | |
| "grad_norm": 1.0102558135986328, | |
| "learning_rate": 2.2033836661916876e-05, | |
| "loss": 2.4134, | |
| "step": 1200000 | |
| }, | |
| { | |
| "epoch": 0.5641930454114785, | |
| "grad_norm": 1.0807286500930786, | |
| "learning_rate": 2.1800815608657053e-05, | |
| "loss": 2.4135, | |
| "step": 1210000 | |
| }, | |
| { | |
| "epoch": 0.5688557978528958, | |
| "grad_norm": 0.9256259799003601, | |
| "learning_rate": 2.156774792787282e-05, | |
| "loss": 2.4106, | |
| "step": 1220000 | |
| }, | |
| { | |
| "epoch": 0.573518550294313, | |
| "grad_norm": 0.9988642930984497, | |
| "learning_rate": 2.1334703560850786e-05, | |
| "loss": 2.4127, | |
| "step": 1230000 | |
| }, | |
| { | |
| "epoch": 0.5781813027357301, | |
| "grad_norm": 1.1274610757827759, | |
| "learning_rate": 2.1101659193828754e-05, | |
| "loss": 2.4096, | |
| "step": 1240000 | |
| }, | |
| { | |
| "epoch": 0.5828440551771473, | |
| "grad_norm": 1.0102494955062866, | |
| "learning_rate": 2.0868591513044516e-05, | |
| "loss": 2.4066, | |
| "step": 1250000 | |
| }, | |
| { | |
| "epoch": 0.5875068076185644, | |
| "grad_norm": 1.011261224746704, | |
| "learning_rate": 2.0635547146022484e-05, | |
| "loss": 2.4046, | |
| "step": 1260000 | |
| }, | |
| { | |
| "epoch": 0.5921695600599817, | |
| "grad_norm": 1.1067317724227905, | |
| "learning_rate": 2.0402502779000456e-05, | |
| "loss": 2.4054, | |
| "step": 1270000 | |
| }, | |
| { | |
| "epoch": 0.5968323125013988, | |
| "grad_norm": 1.117375135421753, | |
| "learning_rate": 2.0169435098216217e-05, | |
| "loss": 2.4057, | |
| "step": 1280000 | |
| }, | |
| { | |
| "epoch": 0.601495064942816, | |
| "grad_norm": 1.067470669746399, | |
| "learning_rate": 1.993636741743198e-05, | |
| "loss": 2.4066, | |
| "step": 1290000 | |
| }, | |
| { | |
| "epoch": 0.6061578173842331, | |
| "grad_norm": 1.0158133506774902, | |
| "learning_rate": 1.9703346364172157e-05, | |
| "loss": 2.4029, | |
| "step": 1300000 | |
| }, | |
| { | |
| "epoch": 0.6108205698256504, | |
| "grad_norm": 1.1179207563400269, | |
| "learning_rate": 1.947027868338792e-05, | |
| "loss": 2.4006, | |
| "step": 1310000 | |
| }, | |
| { | |
| "epoch": 0.6154833222670676, | |
| "grad_norm": 0.8885159492492676, | |
| "learning_rate": 1.923721100260368e-05, | |
| "loss": 2.4008, | |
| "step": 1320000 | |
| }, | |
| { | |
| "epoch": 0.6201460747084847, | |
| "grad_norm": 0.9562169313430786, | |
| "learning_rate": 1.9004143321819446e-05, | |
| "loss": 2.4014, | |
| "step": 1330000 | |
| }, | |
| { | |
| "epoch": 0.6248088271499019, | |
| "grad_norm": 1.0893275737762451, | |
| "learning_rate": 1.8771098954797414e-05, | |
| "loss": 2.3992, | |
| "step": 1340000 | |
| }, | |
| { | |
| "epoch": 0.629471579591319, | |
| "grad_norm": 1.1396783590316772, | |
| "learning_rate": 1.853807790153759e-05, | |
| "loss": 2.3961, | |
| "step": 1350000 | |
| }, | |
| { | |
| "epoch": 0.6341343320327363, | |
| "grad_norm": 0.894639790058136, | |
| "learning_rate": 1.830503353451556e-05, | |
| "loss": 2.3949, | |
| "step": 1360000 | |
| }, | |
| { | |
| "epoch": 0.6387970844741534, | |
| "grad_norm": 1.0523122549057007, | |
| "learning_rate": 1.807196585373132e-05, | |
| "loss": 2.3924, | |
| "step": 1370000 | |
| }, | |
| { | |
| "epoch": 0.6434598369155706, | |
| "grad_norm": 1.4329748153686523, | |
| "learning_rate": 1.7838898172947086e-05, | |
| "loss": 2.3965, | |
| "step": 1380000 | |
| }, | |
| { | |
| "epoch": 0.6481225893569877, | |
| "grad_norm": 0.9407207369804382, | |
| "learning_rate": 1.7605853805925054e-05, | |
| "loss": 2.3944, | |
| "step": 1390000 | |
| }, | |
| { | |
| "epoch": 0.652785341798405, | |
| "grad_norm": 1.1153851747512817, | |
| "learning_rate": 1.7372809438903022e-05, | |
| "loss": 2.3951, | |
| "step": 1400000 | |
| }, | |
| { | |
| "epoch": 0.6574480942398221, | |
| "grad_norm": 1.4270461797714233, | |
| "learning_rate": 1.7139741758118784e-05, | |
| "loss": 2.3903, | |
| "step": 1410000 | |
| }, | |
| { | |
| "epoch": 0.6621108466812393, | |
| "grad_norm": 0.9156707525253296, | |
| "learning_rate": 1.6906697391096756e-05, | |
| "loss": 2.387, | |
| "step": 1420000 | |
| }, | |
| { | |
| "epoch": 0.6667735991226565, | |
| "grad_norm": 1.0517213344573975, | |
| "learning_rate": 1.6673653024074724e-05, | |
| "loss": 2.3908, | |
| "step": 1430000 | |
| }, | |
| { | |
| "epoch": 0.6714363515640737, | |
| "grad_norm": 1.1789027452468872, | |
| "learning_rate": 1.64406319708149e-05, | |
| "loss": 2.3857, | |
| "step": 1440000 | |
| }, | |
| { | |
| "epoch": 0.6760991040054909, | |
| "grad_norm": 0.9410611391067505, | |
| "learning_rate": 1.6207564290030663e-05, | |
| "loss": 2.3851, | |
| "step": 1450000 | |
| }, | |
| { | |
| "epoch": 0.680761856446908, | |
| "grad_norm": 1.2597123384475708, | |
| "learning_rate": 1.597451992300863e-05, | |
| "loss": 2.3853, | |
| "step": 1460000 | |
| }, | |
| { | |
| "epoch": 0.6854246088883252, | |
| "grad_norm": 1.111659288406372, | |
| "learning_rate": 1.5741452242224393e-05, | |
| "loss": 2.3849, | |
| "step": 1470000 | |
| }, | |
| { | |
| "epoch": 0.6900873613297424, | |
| "grad_norm": 1.114686131477356, | |
| "learning_rate": 1.5508384561440158e-05, | |
| "loss": 2.3844, | |
| "step": 1480000 | |
| }, | |
| { | |
| "epoch": 0.6947501137711596, | |
| "grad_norm": 1.3087519407272339, | |
| "learning_rate": 1.527531688065592e-05, | |
| "loss": 2.3811, | |
| "step": 1490000 | |
| }, | |
| { | |
| "epoch": 0.6994128662125767, | |
| "grad_norm": 1.2704778909683228, | |
| "learning_rate": 1.5042319141158304e-05, | |
| "loss": 2.3793, | |
| "step": 1500000 | |
| }, | |
| { | |
| "epoch": 0.7040756186539939, | |
| "grad_norm": 1.0817821025848389, | |
| "learning_rate": 1.4809251460374065e-05, | |
| "loss": 2.3793, | |
| "step": 1510000 | |
| }, | |
| { | |
| "epoch": 0.7087383710954112, | |
| "grad_norm": 1.1640921831130981, | |
| "learning_rate": 1.4576230407114241e-05, | |
| "loss": 2.3826, | |
| "step": 1520000 | |
| }, | |
| { | |
| "epoch": 0.7134011235368283, | |
| "grad_norm": 1.5091464519500732, | |
| "learning_rate": 1.4343162726330003e-05, | |
| "loss": 2.379, | |
| "step": 1530000 | |
| }, | |
| { | |
| "epoch": 0.7180638759782455, | |
| "grad_norm": 1.3562886714935303, | |
| "learning_rate": 1.4110118359307974e-05, | |
| "loss": 2.3748, | |
| "step": 1540000 | |
| }, | |
| { | |
| "epoch": 0.7227266284196626, | |
| "grad_norm": 0.9998787641525269, | |
| "learning_rate": 1.3877073992285944e-05, | |
| "loss": 2.375, | |
| "step": 1550000 | |
| }, | |
| { | |
| "epoch": 0.7273893808610798, | |
| "grad_norm": 1.163294792175293, | |
| "learning_rate": 1.3644006311501706e-05, | |
| "loss": 2.3776, | |
| "step": 1560000 | |
| }, | |
| { | |
| "epoch": 0.732052133302497, | |
| "grad_norm": 1.0799118280410767, | |
| "learning_rate": 1.3410985258241882e-05, | |
| "loss": 2.3732, | |
| "step": 1570000 | |
| }, | |
| { | |
| "epoch": 0.7367148857439142, | |
| "grad_norm": 0.9467183351516724, | |
| "learning_rate": 1.317794089121985e-05, | |
| "loss": 2.3705, | |
| "step": 1580000 | |
| }, | |
| { | |
| "epoch": 0.7413776381853313, | |
| "grad_norm": 1.2810046672821045, | |
| "learning_rate": 1.2944873210435612e-05, | |
| "loss": 2.3721, | |
| "step": 1590000 | |
| }, | |
| { | |
| "epoch": 0.7460403906267485, | |
| "grad_norm": 1.2798866033554077, | |
| "learning_rate": 1.2711828843413585e-05, | |
| "loss": 2.3738, | |
| "step": 1600000 | |
| }, | |
| { | |
| "epoch": 0.7507031430681657, | |
| "grad_norm": 1.221845030784607, | |
| "learning_rate": 1.2478784476391553e-05, | |
| "loss": 2.3683, | |
| "step": 1610000 | |
| }, | |
| { | |
| "epoch": 0.7553658955095829, | |
| "grad_norm": 1.2743821144104004, | |
| "learning_rate": 1.2245740109369522e-05, | |
| "loss": 2.3724, | |
| "step": 1620000 | |
| }, | |
| { | |
| "epoch": 0.7600286479510001, | |
| "grad_norm": 1.1069179773330688, | |
| "learning_rate": 1.201269574234749e-05, | |
| "loss": 2.3662, | |
| "step": 1630000 | |
| }, | |
| { | |
| "epoch": 0.7646914003924172, | |
| "grad_norm": 1.4689267873764038, | |
| "learning_rate": 1.177965137532546e-05, | |
| "loss": 2.3713, | |
| "step": 1640000 | |
| }, | |
| { | |
| "epoch": 0.7693541528338345, | |
| "grad_norm": 1.0129334926605225, | |
| "learning_rate": 1.154660700830343e-05, | |
| "loss": 2.3689, | |
| "step": 1650000 | |
| }, | |
| { | |
| "epoch": 0.7740169052752516, | |
| "grad_norm": 0.9776953458786011, | |
| "learning_rate": 1.1313539327519193e-05, | |
| "loss": 2.363, | |
| "step": 1660000 | |
| }, | |
| { | |
| "epoch": 0.7786796577166688, | |
| "grad_norm": 1.1849191188812256, | |
| "learning_rate": 1.1080494960497161e-05, | |
| "loss": 2.3671, | |
| "step": 1670000 | |
| }, | |
| { | |
| "epoch": 0.7833424101580859, | |
| "grad_norm": 1.0659184455871582, | |
| "learning_rate": 1.0847427279712923e-05, | |
| "loss": 2.363, | |
| "step": 1680000 | |
| }, | |
| { | |
| "epoch": 0.7880051625995032, | |
| "grad_norm": 1.0228557586669922, | |
| "learning_rate": 1.0614382912690895e-05, | |
| "loss": 2.362, | |
| "step": 1690000 | |
| }, | |
| { | |
| "epoch": 0.7926679150409203, | |
| "grad_norm": 0.9540805816650391, | |
| "learning_rate": 1.0381315231906656e-05, | |
| "loss": 2.366, | |
| "step": 1700000 | |
| }, | |
| { | |
| "epoch": 0.7973306674823375, | |
| "grad_norm": 1.1381940841674805, | |
| "learning_rate": 1.0148270864884624e-05, | |
| "loss": 2.3592, | |
| "step": 1710000 | |
| }, | |
| { | |
| "epoch": 0.8019934199237547, | |
| "grad_norm": 1.1460505723953247, | |
| "learning_rate": 9.915226497862596e-06, | |
| "loss": 2.3591, | |
| "step": 1720000 | |
| }, | |
| { | |
| "epoch": 0.8066561723651718, | |
| "grad_norm": 1.0586894750595093, | |
| "learning_rate": 9.682182130840564e-06, | |
| "loss": 2.3592, | |
| "step": 1730000 | |
| }, | |
| { | |
| "epoch": 0.8113189248065891, | |
| "grad_norm": 1.3877402544021606, | |
| "learning_rate": 9.449114450056326e-06, | |
| "loss": 2.3635, | |
| "step": 1740000 | |
| }, | |
| { | |
| "epoch": 0.8159816772480062, | |
| "grad_norm": 1.2622848749160767, | |
| "learning_rate": 9.216046769272089e-06, | |
| "loss": 2.3577, | |
| "step": 1750000 | |
| }, | |
| { | |
| "epoch": 0.8206444296894234, | |
| "grad_norm": 1.1290611028671265, | |
| "learning_rate": 8.983002402250059e-06, | |
| "loss": 2.3587, | |
| "step": 1760000 | |
| }, | |
| { | |
| "epoch": 0.8253071821308405, | |
| "grad_norm": 1.0407214164733887, | |
| "learning_rate": 8.74993472146582e-06, | |
| "loss": 2.3562, | |
| "step": 1770000 | |
| }, | |
| { | |
| "epoch": 0.8299699345722578, | |
| "grad_norm": 1.1062073707580566, | |
| "learning_rate": 8.51689035444379e-06, | |
| "loss": 2.358, | |
| "step": 1780000 | |
| }, | |
| { | |
| "epoch": 0.8346326870136749, | |
| "grad_norm": 1.04072904586792, | |
| "learning_rate": 8.28384598742176e-06, | |
| "loss": 2.3518, | |
| "step": 1790000 | |
| }, | |
| { | |
| "epoch": 0.8392954394550921, | |
| "grad_norm": 1.0454237461090088, | |
| "learning_rate": 8.050801620399728e-06, | |
| "loss": 2.3587, | |
| "step": 1800000 | |
| }, | |
| { | |
| "epoch": 0.8439581918965092, | |
| "grad_norm": 1.2492414712905884, | |
| "learning_rate": 7.817780567139906e-06, | |
| "loss": 2.3552, | |
| "step": 1810000 | |
| }, | |
| { | |
| "epoch": 0.8486209443379265, | |
| "grad_norm": 1.2101612091064453, | |
| "learning_rate": 7.584712886355667e-06, | |
| "loss": 2.358, | |
| "step": 1820000 | |
| }, | |
| { | |
| "epoch": 0.8532836967793437, | |
| "grad_norm": 1.0315169095993042, | |
| "learning_rate": 7.351668519333638e-06, | |
| "loss": 2.3515, | |
| "step": 1830000 | |
| }, | |
| { | |
| "epoch": 0.8579464492207608, | |
| "grad_norm": 1.130194902420044, | |
| "learning_rate": 7.118624152311607e-06, | |
| "loss": 2.3535, | |
| "step": 1840000 | |
| }, | |
| { | |
| "epoch": 0.862609201662178, | |
| "grad_norm": 1.1591068506240845, | |
| "learning_rate": 6.885579785289576e-06, | |
| "loss": 2.3484, | |
| "step": 1850000 | |
| }, | |
| { | |
| "epoch": 0.8672719541035951, | |
| "grad_norm": 1.1694544553756714, | |
| "learning_rate": 6.652535418267545e-06, | |
| "loss": 2.3473, | |
| "step": 1860000 | |
| }, | |
| { | |
| "epoch": 0.8719347065450124, | |
| "grad_norm": 1.2773854732513428, | |
| "learning_rate": 6.419491051245514e-06, | |
| "loss": 2.3464, | |
| "step": 1870000 | |
| }, | |
| { | |
| "epoch": 0.8765974589864295, | |
| "grad_norm": 1.0938977003097534, | |
| "learning_rate": 6.186423370461277e-06, | |
| "loss": 2.3468, | |
| "step": 1880000 | |
| }, | |
| { | |
| "epoch": 0.8812602114278467, | |
| "grad_norm": 1.178916573524475, | |
| "learning_rate": 5.9534023172014535e-06, | |
| "loss": 2.3455, | |
| "step": 1890000 | |
| }, | |
| { | |
| "epoch": 0.8859229638692638, | |
| "grad_norm": 1.2058972120285034, | |
| "learning_rate": 5.720334636417215e-06, | |
| "loss": 2.3433, | |
| "step": 1900000 | |
| } | |
| ], | |
| "logging_steps": 10000, | |
| "max_steps": 2144656, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9223372036854775807, | |
| "save_steps": 100000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.684660451731086e+19, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |