| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 625, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 7.936507936507937e-08, | |
| "logits/chosen": -0.16161203384399414, | |
| "logits/rejected": -0.17215211689472198, | |
| "logps/chosen": -349.4459228515625, | |
| "logps/rejected": -405.1337890625, | |
| "loss": 2.2717, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 7.936507936507937e-07, | |
| "logits/chosen": -0.16214627027511597, | |
| "logits/rejected": -0.07371369749307632, | |
| "logps/chosen": -404.864990234375, | |
| "logps/rejected": -404.12548828125, | |
| "loss": 2.1676, | |
| "rewards/accuracies": 0.4166666567325592, | |
| "rewards/chosen": 0.00014017699868418276, | |
| "rewards/margins": -3.6880755942547694e-05, | |
| "rewards/rejected": 0.00017705775098875165, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.5873015873015873e-06, | |
| "logits/chosen": -0.14529487490653992, | |
| "logits/rejected": -0.10662194341421127, | |
| "logps/chosen": -417.0052185058594, | |
| "logps/rejected": -372.1432189941406, | |
| "loss": 2.2266, | |
| "rewards/accuracies": 0.41874998807907104, | |
| "rewards/chosen": -0.0004417522868607193, | |
| "rewards/margins": -3.551782720023766e-05, | |
| "rewards/rejected": -0.0004062344669364393, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2.380952380952381e-06, | |
| "logits/chosen": -0.18683600425720215, | |
| "logits/rejected": -0.12256692349910736, | |
| "logps/chosen": -439.61053466796875, | |
| "logps/rejected": -411.0472106933594, | |
| "loss": 2.2231, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.002564679365605116, | |
| "rewards/margins": 0.0009385704761371017, | |
| "rewards/rejected": -0.0035032504238188267, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 3.1746031746031746e-06, | |
| "logits/chosen": -0.15973249077796936, | |
| "logits/rejected": -0.06972485780715942, | |
| "logps/chosen": -432.1578674316406, | |
| "logps/rejected": -439.65692138671875, | |
| "loss": 2.1315, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.008030252531170845, | |
| "rewards/margins": 0.005707239266484976, | |
| "rewards/rejected": -0.013737490400671959, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 3.968253968253968e-06, | |
| "logits/chosen": -0.180991992354393, | |
| "logits/rejected": -0.12702423334121704, | |
| "logps/chosen": -445.7837829589844, | |
| "logps/rejected": -477.53204345703125, | |
| "loss": 2.1043, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.021977921947836876, | |
| "rewards/margins": 0.014845812693238258, | |
| "rewards/rejected": -0.036823734641075134, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.761904761904762e-06, | |
| "logits/chosen": -0.2151261866092682, | |
| "logits/rejected": -0.11040729284286499, | |
| "logps/chosen": -516.127685546875, | |
| "logps/rejected": -513.0391845703125, | |
| "loss": 2.0342, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.07830885052680969, | |
| "rewards/margins": 0.029339421540498734, | |
| "rewards/rejected": -0.10764826834201813, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 4.998086282661188e-06, | |
| "logits/chosen": -0.19878175854682922, | |
| "logits/rejected": -0.1446671187877655, | |
| "logps/chosen": -586.38037109375, | |
| "logps/rejected": -702.6896362304688, | |
| "loss": 1.9578, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.18542954325675964, | |
| "rewards/margins": 0.10293972492218018, | |
| "rewards/rejected": -0.2883692979812622, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.988720025682995e-06, | |
| "logits/chosen": -0.2566753923892975, | |
| "logits/rejected": -0.1441725641489029, | |
| "logps/chosen": -738.7784423828125, | |
| "logps/rejected": -817.1397705078125, | |
| "loss": 1.8572, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.3142145276069641, | |
| "rewards/margins": 0.1317647397518158, | |
| "rewards/rejected": -0.44597920775413513, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.9715789537359126e-06, | |
| "logits/chosen": -0.221205472946167, | |
| "logits/rejected": -0.1643596887588501, | |
| "logps/chosen": -786.4627685546875, | |
| "logps/rejected": -940.6861572265625, | |
| "loss": 1.7929, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.36182770133018494, | |
| "rewards/margins": 0.19319193065166473, | |
| "rewards/rejected": -0.5550196766853333, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 4.946716615897932e-06, | |
| "logits/chosen": -0.2160252034664154, | |
| "logits/rejected": -0.1826382726430893, | |
| "logps/chosen": -811.1300659179688, | |
| "logps/rejected": -964.46337890625, | |
| "loss": 1.8284, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.4078206419944763, | |
| "rewards/margins": 0.19059345126152039, | |
| "rewards/rejected": -0.5984141826629639, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.9142106826480114e-06, | |
| "logits/chosen": -0.22243019938468933, | |
| "logits/rejected": -0.1664743721485138, | |
| "logps/chosen": -861.86767578125, | |
| "logps/rejected": -994.6505737304688, | |
| "loss": 1.9512, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.44844961166381836, | |
| "rewards/margins": 0.15178686380386353, | |
| "rewards/rejected": -0.6002364158630371, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.874162703221823e-06, | |
| "logits/chosen": -0.2382921725511551, | |
| "logits/rejected": -0.1875036060810089, | |
| "logps/chosen": -862.8167114257812, | |
| "logps/rejected": -996.0676879882812, | |
| "loss": 1.8951, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.43303823471069336, | |
| "rewards/margins": 0.16045571863651276, | |
| "rewards/rejected": -0.5934939980506897, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.826697788369752e-06, | |
| "logits/chosen": -0.22683772444725037, | |
| "logits/rejected": -0.19122302532196045, | |
| "logps/chosen": -942.7552490234375, | |
| "logps/rejected": -1131.6470947265625, | |
| "loss": 1.7237, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.4950883984565735, | |
| "rewards/margins": 0.21038611233234406, | |
| "rewards/rejected": -0.7054744958877563, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.7719642195082224e-06, | |
| "logits/chosen": -0.1632569581270218, | |
| "logits/rejected": -0.17590998113155365, | |
| "logps/chosen": -867.0773315429688, | |
| "logps/rejected": -1070.54296875, | |
| "loss": 1.7874, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.47380104660987854, | |
| "rewards/margins": 0.20263484120368958, | |
| "rewards/rejected": -0.6764358878135681, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 4.710132985485355e-06, | |
| "logits/chosen": -0.17881760001182556, | |
| "logits/rejected": -0.148959681391716, | |
| "logps/chosen": -871.22216796875, | |
| "logps/rejected": -1014.7432861328125, | |
| "loss": 1.7998, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.43568116426467896, | |
| "rewards/margins": 0.19202783703804016, | |
| "rewards/rejected": -0.6277089715003967, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.641397248408122e-06, | |
| "logits/chosen": -0.16328132152557373, | |
| "logits/rejected": -0.1337490677833557, | |
| "logps/chosen": -840.59423828125, | |
| "logps/rejected": -991.12841796875, | |
| "loss": 1.7773, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.40007978677749634, | |
| "rewards/margins": 0.1872185319662094, | |
| "rewards/rejected": -0.5872983336448669, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.5659717401997655e-06, | |
| "logits/chosen": -0.1909216344356537, | |
| "logits/rejected": -0.14504285156726837, | |
| "logps/chosen": -804.4796142578125, | |
| "logps/rejected": -962.5818481445312, | |
| "loss": 1.9007, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.4276674687862396, | |
| "rewards/margins": 0.1687774360179901, | |
| "rewards/rejected": -0.5964449644088745, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.4840920917726425e-06, | |
| "logits/chosen": -0.18420439958572388, | |
| "logits/rejected": -0.15532034635543823, | |
| "logps/chosen": -728.2938232421875, | |
| "logps/rejected": -969.2843017578125, | |
| "loss": 1.8113, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.3586859405040741, | |
| "rewards/margins": 0.2266702651977539, | |
| "rewards/rejected": -0.5853561758995056, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 4.396014096912182e-06, | |
| "logits/chosen": -0.13352888822555542, | |
| "logits/rejected": -0.11805830150842667, | |
| "logps/chosen": -830.2320556640625, | |
| "logps/rejected": -1047.818359375, | |
| "loss": 1.7352, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.4081611633300781, | |
| "rewards/margins": 0.24260549247264862, | |
| "rewards/rejected": -0.6507667303085327, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 4.302012913171584e-06, | |
| "logits/chosen": -0.21690301597118378, | |
| "logits/rejected": -0.19134214520454407, | |
| "logps/chosen": -941.5948486328125, | |
| "logps/rejected": -1166.3424072265625, | |
| "loss": 1.8958, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.5229096412658691, | |
| "rewards/margins": 0.22313475608825684, | |
| "rewards/rejected": -0.746044397354126, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 4.202382202273702e-06, | |
| "logits/chosen": -0.2081947773694992, | |
| "logits/rejected": -0.14332859218120575, | |
| "logps/chosen": -888.3336791992188, | |
| "logps/rejected": -1034.1588134765625, | |
| "loss": 1.8719, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.45206698775291443, | |
| "rewards/margins": 0.20269028842449188, | |
| "rewards/rejected": -0.6547572612762451, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 4.097433212705492e-06, | |
| "logits/chosen": -0.15774083137512207, | |
| "logits/rejected": -0.12937268614768982, | |
| "logps/chosen": -831.2762451171875, | |
| "logps/rejected": -972.7151489257812, | |
| "loss": 1.7465, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.4013233780860901, | |
| "rewards/margins": 0.16350558400154114, | |
| "rewards/rejected": -0.5648289918899536, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 3.987493807371033e-06, | |
| "logits/chosen": -0.17701825499534607, | |
| "logits/rejected": -0.11814224720001221, | |
| "logps/chosen": -738.5556640625, | |
| "logps/rejected": -925.3514404296875, | |
| "loss": 1.7971, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.33476728200912476, | |
| "rewards/margins": 0.19879043102264404, | |
| "rewards/rejected": -0.5335577130317688, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.872907439340758e-06, | |
| "logits/chosen": -0.14647835493087769, | |
| "logits/rejected": -0.13141624629497528, | |
| "logps/chosen": -827.5966796875, | |
| "logps/rejected": -1106.738037109375, | |
| "loss": 1.6676, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.3911557197570801, | |
| "rewards/margins": 0.2823970317840576, | |
| "rewards/rejected": -0.6735527515411377, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 3.75403207889666e-06, | |
| "logits/chosen": -0.165244922041893, | |
| "logits/rejected": -0.10319925844669342, | |
| "logps/chosen": -897.0632934570312, | |
| "logps/rejected": -1235.7066650390625, | |
| "loss": 1.6872, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.44384971261024475, | |
| "rewards/margins": 0.32984811067581177, | |
| "rewards/rejected": -0.7736977934837341, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.631239095225417e-06, | |
| "logits/chosen": -0.18860653042793274, | |
| "logits/rejected": -0.20599384605884552, | |
| "logps/chosen": -895.7766723632812, | |
| "logps/rejected": -1160.929931640625, | |
| "loss": 1.6633, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.486502081155777, | |
| "rewards/margins": 0.2695665657520294, | |
| "rewards/rejected": -0.7560686469078064, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 3.5049120962530608e-06, | |
| "logits/chosen": -0.14848849177360535, | |
| "logits/rejected": -0.1331581175327301, | |
| "logps/chosen": -992.6770629882812, | |
| "logps/rejected": -1140.1298828125, | |
| "loss": 1.7584, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.510170578956604, | |
| "rewards/margins": 0.20152750611305237, | |
| "rewards/rejected": -0.7116979956626892, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 3.3754457302455464e-06, | |
| "logits/chosen": -0.13460226356983185, | |
| "logits/rejected": -0.08309068530797958, | |
| "logps/chosen": -913.01416015625, | |
| "logps/rejected": -1093.90234375, | |
| "loss": 1.7657, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.45719900727272034, | |
| "rewards/margins": 0.2582068145275116, | |
| "rewards/rejected": -0.7154058218002319, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 3.2432444529190714e-06, | |
| "logits/chosen": -0.16204790771007538, | |
| "logits/rejected": -0.15286168456077576, | |
| "logps/chosen": -848.2911987304688, | |
| "logps/rejected": -1192.6624755859375, | |
| "loss": 1.6853, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.4555201530456543, | |
| "rewards/margins": 0.324947327375412, | |
| "rewards/rejected": -0.7804675102233887, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 3.1087212639117057e-06, | |
| "logits/chosen": -0.145342618227005, | |
| "logits/rejected": -0.15378527343273163, | |
| "logps/chosen": -855.2142333984375, | |
| "logps/rejected": -1193.8294677734375, | |
| "loss": 1.7113, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.4761961102485657, | |
| "rewards/margins": 0.3075980544090271, | |
| "rewards/rejected": -0.7837941646575928, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.9722964165636263e-06, | |
| "logits/chosen": -0.12143807113170624, | |
| "logits/rejected": -0.16259871423244476, | |
| "logps/chosen": -939.7080078125, | |
| "logps/rejected": -1251.651123046875, | |
| "loss": 1.7493, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.4829345643520355, | |
| "rewards/margins": 0.3475037217140198, | |
| "rewards/rejected": -0.8304383158683777, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 2.8343961050366275e-06, | |
| "logits/chosen": -0.14929132163524628, | |
| "logits/rejected": -0.1324179470539093, | |
| "logps/chosen": -932.6903076171875, | |
| "logps/rejected": -1189.304443359375, | |
| "loss": 1.7696, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.5320444107055664, | |
| "rewards/margins": 0.24077489972114563, | |
| "rewards/rejected": -0.7728193998336792, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 2.695451132874385e-06, | |
| "logits/chosen": -0.18022935092449188, | |
| "logits/rejected": -0.11948423087596893, | |
| "logps/chosen": -945.2275390625, | |
| "logps/rejected": -1175.975341796875, | |
| "loss": 1.6686, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.5022936463356018, | |
| "rewards/margins": 0.2758629322052002, | |
| "rewards/rejected": -0.7781565189361572, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.5558955671628964e-06, | |
| "logits/chosen": -0.12632670998573303, | |
| "logits/rejected": -0.114730104804039, | |
| "logps/chosen": -963.2265625, | |
| "logps/rejected": -1251.21533203125, | |
| "loss": 1.6963, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.5192445516586304, | |
| "rewards/margins": 0.2935452461242676, | |
| "rewards/rejected": -0.812789797782898, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 2.4161653824955654e-06, | |
| "logits/chosen": -0.11149231344461441, | |
| "logits/rejected": -0.1491885930299759, | |
| "logps/chosen": -913.9876098632812, | |
| "logps/rejected": -1161.088623046875, | |
| "loss": 1.6863, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.4950644075870514, | |
| "rewards/margins": 0.2864512801170349, | |
| "rewards/rejected": -0.7815156579017639, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2.2766970989791697e-06, | |
| "logits/chosen": -0.09586126357316971, | |
| "logits/rejected": -0.06141890212893486, | |
| "logps/chosen": -859.8944091796875, | |
| "logps/rejected": -1119.4293212890625, | |
| "loss": 1.7026, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.4382789134979248, | |
| "rewards/margins": 0.2800602316856384, | |
| "rewards/rejected": -0.718339204788208, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 2.1379264185356545e-06, | |
| "logits/chosen": -0.0828716978430748, | |
| "logits/rejected": -0.0706620067358017, | |
| "logps/chosen": -909.8837890625, | |
| "logps/rejected": -1223.965576171875, | |
| "loss": 1.7231, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.47254592180252075, | |
| "rewards/margins": 0.340139776468277, | |
| "rewards/rejected": -0.8126856684684753, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 2.000286863759934e-06, | |
| "logits/chosen": -0.12816865742206573, | |
| "logits/rejected": -0.12074576318264008, | |
| "logps/chosen": -866.0748291015625, | |
| "logps/rejected": -1134.2685546875, | |
| "loss": 1.6912, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.45770812034606934, | |
| "rewards/margins": 0.29924601316452026, | |
| "rewards/rejected": -0.7569540739059448, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 1.8642084235859764e-06, | |
| "logits/chosen": -0.11092492192983627, | |
| "logits/rejected": -0.06964071094989777, | |
| "logps/chosen": -876.1008911132812, | |
| "logps/rejected": -1147.765625, | |
| "loss": 1.7318, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.40335574746131897, | |
| "rewards/margins": 0.2833205759525299, | |
| "rewards/rejected": -0.6866763234138489, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 1.7301162099921013e-06, | |
| "logits/chosen": -0.10882829129695892, | |
| "logits/rejected": -0.06481163948774338, | |
| "logps/chosen": -871.3518676757812, | |
| "logps/rejected": -1078.458740234375, | |
| "loss": 1.61, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.44269245862960815, | |
| "rewards/margins": 0.25641289353370667, | |
| "rewards/rejected": -0.6991053819656372, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 1.5984291299420117e-06, | |
| "logits/chosen": -0.08046077191829681, | |
| "logits/rejected": -0.05565166473388672, | |
| "logps/chosen": -783.994873046875, | |
| "logps/rejected": -1022.2307739257812, | |
| "loss": 1.8072, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.4169149398803711, | |
| "rewards/margins": 0.2162661999464035, | |
| "rewards/rejected": -0.6331812143325806, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.4695585767104092e-06, | |
| "logits/chosen": -0.11396467685699463, | |
| "logits/rejected": -0.10157410800457001, | |
| "logps/chosen": -899.9319458007812, | |
| "logps/rejected": -1116.689697265625, | |
| "loss": 1.7069, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.46262964606285095, | |
| "rewards/margins": 0.2319665402173996, | |
| "rewards/rejected": -0.6945961713790894, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.3439071446815452e-06, | |
| "logits/chosen": -0.11295236647129059, | |
| "logits/rejected": -0.08282504230737686, | |
| "logps/chosen": -887.32421875, | |
| "logps/rejected": -1047.8634033203125, | |
| "loss": 1.6503, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.43825840950012207, | |
| "rewards/margins": 0.21870870888233185, | |
| "rewards/rejected": -0.656967043876648, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.2218673716356919e-06, | |
| "logits/chosen": -0.07290169596672058, | |
| "logits/rejected": -0.06002408266067505, | |
| "logps/chosen": -929.90380859375, | |
| "logps/rejected": -1127.547119140625, | |
| "loss": 1.7147, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.4546731114387512, | |
| "rewards/margins": 0.27696481347084045, | |
| "rewards/rejected": -0.7316379547119141, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 1.103820512452661e-06, | |
| "logits/chosen": -0.10286526381969452, | |
| "logits/rejected": -0.071159228682518, | |
| "logps/chosen": -874.0986328125, | |
| "logps/rejected": -1238.300048828125, | |
| "loss": 1.6778, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.46896892786026, | |
| "rewards/margins": 0.3574833273887634, | |
| "rewards/rejected": -0.8264522552490234, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 9.901353480633468e-07, | |
| "logits/chosen": -0.05727295204997063, | |
| "logits/rejected": -0.08084852248430252, | |
| "logps/chosen": -864.6730346679688, | |
| "logps/rejected": -1181.377685546875, | |
| "loss": 1.6027, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.4679865837097168, | |
| "rewards/margins": 0.3022565245628357, | |
| "rewards/rejected": -0.7702431082725525, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 8.811670333701544e-07, | |
| "logits/chosen": -0.05785631015896797, | |
| "logits/rejected": -0.03484424576163292, | |
| "logps/chosen": -947.6015625, | |
| "logps/rejected": -1211.000732421875, | |
| "loss": 1.5945, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.4864828586578369, | |
| "rewards/margins": 0.31723761558532715, | |
| "rewards/rejected": -0.8037204742431641, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 7.772559877354341e-07, | |
| "logits/chosen": -0.08603869378566742, | |
| "logits/rejected": -0.0594349130988121, | |
| "logps/chosen": -976.7135009765625, | |
| "logps/rejected": -1181.75, | |
| "loss": 1.7634, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.5735828280448914, | |
| "rewards/margins": 0.23867933452129364, | |
| "rewards/rejected": -0.8122621774673462, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 6.787268315040604e-07, | |
| "logits/chosen": -0.1035851389169693, | |
| "logits/rejected": -0.10194051265716553, | |
| "logps/chosen": -891.7512817382812, | |
| "logps/rejected": -1107.476806640625, | |
| "loss": 1.6117, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.48722830414772034, | |
| "rewards/margins": 0.2562137544155121, | |
| "rewards/rejected": -0.7434421181678772, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 5.858873718824829e-07, | |
| "logits/chosen": -0.06155969947576523, | |
| "logits/rejected": -0.018813952803611755, | |
| "logps/chosen": -938.9776611328125, | |
| "logps/rejected": -1191.833740234375, | |
| "loss": 1.6467, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.5143369436264038, | |
| "rewards/margins": 0.2735580801963806, | |
| "rewards/rejected": -0.7878950834274292, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 4.990276413423817e-07, | |
| "logits/chosen": -0.06636036932468414, | |
| "logits/rejected": -0.10939071327447891, | |
| "logps/chosen": -859.03515625, | |
| "logps/rejected": -1074.6214599609375, | |
| "loss": 1.7273, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.480158269405365, | |
| "rewards/margins": 0.23960845172405243, | |
| "rewards/rejected": -0.7197667360305786, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 4.184189915529796e-07, | |
| "logits/chosen": -0.049408238381147385, | |
| "logits/rejected": -0.06695321202278137, | |
| "logps/chosen": -991.7838134765625, | |
| "logps/rejected": -1224.253173828125, | |
| "loss": 1.6447, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.529580295085907, | |
| "rewards/margins": 0.23929505050182343, | |
| "rewards/rejected": -0.7688754200935364, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 3.4431324567258176e-07, | |
| "logits/chosen": -0.0758766159415245, | |
| "logits/rejected": -0.06797490268945694, | |
| "logps/chosen": -894.0300903320312, | |
| "logps/rejected": -1302.058837890625, | |
| "loss": 1.654, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.4872868061065674, | |
| "rewards/margins": 0.3845263421535492, | |
| "rewards/rejected": -0.8718131184577942, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 2.769419116476052e-07, | |
| "logits/chosen": -0.12563803791999817, | |
| "logits/rejected": -0.09884506464004517, | |
| "logps/chosen": -938.1819458007812, | |
| "logps/rejected": -1275.197021484375, | |
| "loss": 1.6659, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.5264378786087036, | |
| "rewards/margins": 0.3382587432861328, | |
| "rewards/rejected": -0.8646965026855469, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 2.1651545897676512e-07, | |
| "logits/chosen": -0.11725203692913055, | |
| "logits/rejected": -0.11871466785669327, | |
| "logps/chosen": -898.6276245117188, | |
| "logps/rejected": -1201.9349365234375, | |
| "loss": 1.747, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.5061923861503601, | |
| "rewards/margins": 0.3278709053993225, | |
| "rewards/rejected": -0.8340633511543274, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.6322266119983222e-07, | |
| "logits/chosen": -0.04045483097434044, | |
| "logits/rejected": -0.05498255044221878, | |
| "logps/chosen": -983.4290771484375, | |
| "logps/rejected": -1337.211181640625, | |
| "loss": 1.5444, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.5375570058822632, | |
| "rewards/margins": 0.3453003168106079, | |
| "rewards/rejected": -0.8828572034835815, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 1.1723000616502167e-07, | |
| "logits/chosen": -0.09810743480920792, | |
| "logits/rejected": -0.09926000982522964, | |
| "logps/chosen": -835.6412353515625, | |
| "logps/rejected": -1187.4744873046875, | |
| "loss": 1.794, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.46745967864990234, | |
| "rewards/margins": 0.3435118794441223, | |
| "rewards/rejected": -0.8109715580940247, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 7.868117591737585e-08, | |
| "logits/chosen": -0.09739083796739578, | |
| "logits/rejected": -0.10297133773565292, | |
| "logps/chosen": -885.7447509765625, | |
| "logps/rejected": -1204.79443359375, | |
| "loss": 1.6582, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.5217578411102295, | |
| "rewards/margins": 0.28710317611694336, | |
| "rewards/rejected": -0.8088610768318176, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 4.769659783295383e-08, | |
| "logits/chosen": -0.11744657903909683, | |
| "logits/rejected": -0.07945755124092102, | |
| "logps/chosen": -882.5567626953125, | |
| "logps/rejected": -1246.40234375, | |
| "loss": 1.6835, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.487631618976593, | |
| "rewards/margins": 0.3489754796028137, | |
| "rewards/rejected": -0.8366070985794067, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2.4373068401120358e-08, | |
| "logits/chosen": -0.08680951595306396, | |
| "logits/rejected": -0.07239434868097305, | |
| "logps/chosen": -865.5779418945312, | |
| "logps/rejected": -1108.098388671875, | |
| "loss": 1.7686, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.48516979813575745, | |
| "rewards/margins": 0.2544567883014679, | |
| "rewards/rejected": -0.7396265864372253, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 8.78345083022425e-09, | |
| "logits/chosen": -0.09715618193149567, | |
| "logits/rejected": -0.08691917359828949, | |
| "logps/chosen": -904.9318237304688, | |
| "logps/rejected": -1211.5076904296875, | |
| "loss": 1.7077, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.5081444978713989, | |
| "rewards/margins": 0.2958123981952667, | |
| "rewards/rejected": -0.8039568662643433, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 9.764474213677654e-10, | |
| "logits/chosen": -0.05566658824682236, | |
| "logits/rejected": -0.07036083936691284, | |
| "logps/chosen": -910.61767578125, | |
| "logps/rejected": -1149.9638671875, | |
| "loss": 1.6468, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.5058663487434387, | |
| "rewards/margins": 0.25643160939216614, | |
| "rewards/rejected": -0.7622979879379272, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 625, | |
| "total_flos": 0.0, | |
| "train_loss": 1.7737385009765625, | |
| "train_runtime": 8419.2459, | |
| "train_samples_per_second": 3.563, | |
| "train_steps_per_second": 0.074 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 625, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "total_flos": 0.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |