| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 15.886524822695035, | |
| "eval_steps": 80, | |
| "global_step": 840, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.6052009456264775, | |
| "grad_norm": 9.684629440307617, | |
| "learning_rate": 1.9047619047619045e-07, | |
| "log_odds_chosen": 0.08431098610162735, | |
| "log_odds_ratio": -0.7315660715103149, | |
| "logits/chosen": -2.4399943351745605, | |
| "logits/rejected": -2.418248414993286, | |
| "logps/chosen": -1.316224455833435, | |
| "logps/rejected": -1.3803966045379639, | |
| "loss": 1.6469, | |
| "nll_loss": 1.5446076393127441, | |
| "rewards/accuracies": 0.54296875, | |
| "rewards/chosen": -0.1974336802959442, | |
| "rewards/margins": 0.009625822305679321, | |
| "rewards/rejected": -0.20705950260162354, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.210401891252955, | |
| "grad_norm": 6.682727813720703, | |
| "learning_rate": 3.809523809523809e-07, | |
| "log_odds_chosen": 0.1358582228422165, | |
| "log_odds_ratio": -0.6844438314437866, | |
| "logits/chosen": -2.4866111278533936, | |
| "logits/rejected": -2.473512649536133, | |
| "logps/chosen": -1.2416539192199707, | |
| "logps/rejected": -1.338365912437439, | |
| "loss": 1.5513, | |
| "nll_loss": 1.43682861328125, | |
| "rewards/accuracies": 0.55859375, | |
| "rewards/chosen": -0.1862480789422989, | |
| "rewards/margins": 0.01450679823756218, | |
| "rewards/rejected": -0.20075488090515137, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.5130023640661938, | |
| "eval_log_odds_chosen": 1.1365413665771484, | |
| "eval_log_odds_ratio": -0.2909667193889618, | |
| "eval_logits/chosen": -2.8555617332458496, | |
| "eval_logits/rejected": -2.7511401176452637, | |
| "eval_logps/chosen": -1.1086950302124023, | |
| "eval_logps/rejected": -1.990875244140625, | |
| "eval_loss": 1.1764631271362305, | |
| "eval_nll_loss": 1.1883823871612549, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": -0.16630426049232483, | |
| "eval_rewards/margins": 0.13232707977294922, | |
| "eval_rewards/rejected": -0.29863131046295166, | |
| "eval_runtime": 0.7889, | |
| "eval_samples_per_second": 173.657, | |
| "eval_steps_per_second": 6.338, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.8156028368794326, | |
| "grad_norm": 5.326328754425049, | |
| "learning_rate": 4.996892303047305e-07, | |
| "log_odds_chosen": 0.18785640597343445, | |
| "log_odds_ratio": -0.6766043901443481, | |
| "logits/chosen": -2.4521989822387695, | |
| "logits/rejected": -2.457139492034912, | |
| "logps/chosen": -1.1742055416107178, | |
| "logps/rejected": -1.3268922567367554, | |
| "loss": 1.4307, | |
| "nll_loss": 1.3236442804336548, | |
| "rewards/accuracies": 0.5390625, | |
| "rewards/chosen": -0.17613083124160767, | |
| "rewards/margins": 0.022903023287653923, | |
| "rewards/rejected": -0.19903387129306793, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.42080378250591, | |
| "grad_norm": 6.345442295074463, | |
| "learning_rate": 4.958326378681848e-07, | |
| "log_odds_chosen": 0.20125526189804077, | |
| "log_odds_ratio": -0.6606975793838501, | |
| "logits/chosen": -2.4338855743408203, | |
| "logits/rejected": -2.4032998085021973, | |
| "logps/chosen": -1.2052891254425049, | |
| "logps/rejected": -1.3293428421020508, | |
| "loss": 1.3822, | |
| "nll_loss": 1.3110582828521729, | |
| "rewards/accuracies": 0.609375, | |
| "rewards/chosen": -0.18079334497451782, | |
| "rewards/margins": 0.018608052283525467, | |
| "rewards/rejected": -0.19940140843391418, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 3.0260047281323876, | |
| "grad_norm": 6.9922590255737305, | |
| "learning_rate": 4.876353872369572e-07, | |
| "log_odds_chosen": 0.2849215567111969, | |
| "log_odds_ratio": -0.6226438283920288, | |
| "logits/chosen": -2.28694748878479, | |
| "logits/rejected": -2.2813074588775635, | |
| "logps/chosen": -1.177689552307129, | |
| "logps/rejected": -1.3626967668533325, | |
| "loss": 1.3457, | |
| "nll_loss": 1.2851402759552002, | |
| "rewards/accuracies": 0.68359375, | |
| "rewards/chosen": -0.1766534298658371, | |
| "rewards/margins": 0.027751106768846512, | |
| "rewards/rejected": -0.20440451800823212, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 3.0260047281323876, | |
| "eval_log_odds_chosen": 1.289251685142517, | |
| "eval_log_odds_ratio": -0.2533319592475891, | |
| "eval_logits/chosen": -2.6994073390960693, | |
| "eval_logits/rejected": -2.5944201946258545, | |
| "eval_logps/chosen": -1.1015231609344482, | |
| "eval_logps/rejected": -2.1139886379241943, | |
| "eval_loss": 1.1738542318344116, | |
| "eval_nll_loss": 1.1927688121795654, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": -0.1652284860610962, | |
| "eval_rewards/margins": 0.15186984837055206, | |
| "eval_rewards/rejected": -0.31709831953048706, | |
| "eval_runtime": 0.7972, | |
| "eval_samples_per_second": 171.841, | |
| "eval_steps_per_second": 6.272, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 3.631205673758865, | |
| "grad_norm": 6.0752034187316895, | |
| "learning_rate": 4.752422169756047e-07, | |
| "log_odds_chosen": 0.29861366748809814, | |
| "log_odds_ratio": -0.6187925338745117, | |
| "logits/chosen": -2.278367280960083, | |
| "logits/rejected": -2.197380781173706, | |
| "logps/chosen": -1.1402652263641357, | |
| "logps/rejected": -1.323896884918213, | |
| "loss": 1.3117, | |
| "nll_loss": 1.1778795719146729, | |
| "rewards/accuracies": 0.69140625, | |
| "rewards/chosen": -0.17103978991508484, | |
| "rewards/margins": 0.0275447778403759, | |
| "rewards/rejected": -0.19858455657958984, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 4.236406619385343, | |
| "grad_norm": 6.758876800537109, | |
| "learning_rate": 4.588719528532341e-07, | |
| "log_odds_chosen": 0.32968199253082275, | |
| "log_odds_ratio": -0.6064258813858032, | |
| "logits/chosen": -2.2086422443389893, | |
| "logits/rejected": -2.213918447494507, | |
| "logps/chosen": -1.2367851734161377, | |
| "logps/rejected": -1.4456892013549805, | |
| "loss": 1.2914, | |
| "nll_loss": 1.220529317855835, | |
| "rewards/accuracies": 0.703125, | |
| "rewards/chosen": -0.1855177879333496, | |
| "rewards/margins": 0.03133557736873627, | |
| "rewards/rejected": -0.21685336530208588, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 4.539007092198582, | |
| "eval_log_odds_chosen": 1.4739799499511719, | |
| "eval_log_odds_ratio": -0.21903792023658752, | |
| "eval_logits/chosen": -2.6473140716552734, | |
| "eval_logits/rejected": -2.5277247428894043, | |
| "eval_logps/chosen": -1.2001134157180786, | |
| "eval_logps/rejected": -2.410470485687256, | |
| "eval_loss": 1.1219241619110107, | |
| "eval_nll_loss": 1.137636423110962, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": -0.18001702427864075, | |
| "eval_rewards/margins": 0.18155357241630554, | |
| "eval_rewards/rejected": -0.3615706264972687, | |
| "eval_runtime": 0.7853, | |
| "eval_samples_per_second": 174.46, | |
| "eval_steps_per_second": 6.367, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 4.84160756501182, | |
| "grad_norm": 6.2288618087768555, | |
| "learning_rate": 4.3881364404463375e-07, | |
| "log_odds_chosen": 0.46286657452583313, | |
| "log_odds_ratio": -0.5630860328674316, | |
| "logits/chosen": -2.216555595397949, | |
| "logits/rejected": -2.1122565269470215, | |
| "logps/chosen": -1.1853437423706055, | |
| "logps/rejected": -1.4803636074066162, | |
| "loss": 1.2727, | |
| "nll_loss": 1.1509523391723633, | |
| "rewards/accuracies": 0.77734375, | |
| "rewards/chosen": -0.17780157923698425, | |
| "rewards/margins": 0.04425298422574997, | |
| "rewards/rejected": -0.222054585814476, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 5.446808510638298, | |
| "grad_norm": 8.837624549865723, | |
| "learning_rate": 4.154214593992149e-07, | |
| "log_odds_chosen": 0.571550726890564, | |
| "log_odds_ratio": -0.5269472002983093, | |
| "logits/chosen": -2.189985990524292, | |
| "logits/rejected": -2.085646152496338, | |
| "logps/chosen": -1.2091223001480103, | |
| "logps/rejected": -1.5827255249023438, | |
| "loss": 1.2582, | |
| "nll_loss": 1.1602920293807983, | |
| "rewards/accuracies": 0.78515625, | |
| "rewards/chosen": -0.18136833608150482, | |
| "rewards/margins": 0.05604049190878868, | |
| "rewards/rejected": -0.237408846616745, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 6.052009456264775, | |
| "grad_norm": 10.696549415588379, | |
| "learning_rate": 3.891084338941603e-07, | |
| "log_odds_chosen": 0.5793906450271606, | |
| "log_odds_ratio": -0.5278146862983704, | |
| "logits/chosen": -2.0601658821105957, | |
| "logits/rejected": -2.04327130317688, | |
| "logps/chosen": -1.2197258472442627, | |
| "logps/rejected": -1.6007359027862549, | |
| "loss": 1.261, | |
| "nll_loss": 1.1714200973510742, | |
| "rewards/accuracies": 0.76953125, | |
| "rewards/chosen": -0.18295888602733612, | |
| "rewards/margins": 0.057151518762111664, | |
| "rewards/rejected": -0.2401103973388672, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 6.052009456264775, | |
| "eval_log_odds_chosen": 1.5137661695480347, | |
| "eval_log_odds_ratio": -0.21324041485786438, | |
| "eval_logits/chosen": -2.5153615474700928, | |
| "eval_logits/rejected": -2.389270067214966, | |
| "eval_logps/chosen": -1.2208881378173828, | |
| "eval_logps/rejected": -2.475594997406006, | |
| "eval_loss": 1.0919252634048462, | |
| "eval_nll_loss": 1.1019929647445679, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": -0.18313322961330414, | |
| "eval_rewards/margins": 0.1882060021162033, | |
| "eval_rewards/rejected": -0.37133923172950745, | |
| "eval_runtime": 0.8032, | |
| "eval_samples_per_second": 170.562, | |
| "eval_steps_per_second": 6.225, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 6.657210401891253, | |
| "grad_norm": 14.943062782287598, | |
| "learning_rate": 3.6033917569043597e-07, | |
| "log_odds_chosen": 0.6752089858055115, | |
| "log_odds_ratio": -0.5140572190284729, | |
| "logits/chosen": -2.0157763957977295, | |
| "logits/rejected": -2.004314422607422, | |
| "logps/chosen": -1.2377139329910278, | |
| "logps/rejected": -1.692950963973999, | |
| "loss": 1.252, | |
| "nll_loss": 1.1812970638275146, | |
| "rewards/accuracies": 0.75390625, | |
| "rewards/chosen": -0.1856570839881897, | |
| "rewards/margins": 0.06828554719686508, | |
| "rewards/rejected": -0.25394266843795776, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 7.26241134751773, | |
| "grad_norm": 24.735410690307617, | |
| "learning_rate": 3.296216625629211e-07, | |
| "log_odds_chosen": 0.8964105844497681, | |
| "log_odds_ratio": -0.46805524826049805, | |
| "logits/chosen": -1.9912079572677612, | |
| "logits/rejected": -1.955162525177002, | |
| "logps/chosen": -1.2576524019241333, | |
| "logps/rejected": -1.9145023822784424, | |
| "loss": 1.2436, | |
| "nll_loss": 1.170907974243164, | |
| "rewards/accuracies": 0.7578125, | |
| "rewards/chosen": -0.18864786624908447, | |
| "rewards/margins": 0.09852751344442368, | |
| "rewards/rejected": -0.28717538714408875, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 7.5650118203309695, | |
| "eval_log_odds_chosen": 1.5138293504714966, | |
| "eval_log_odds_ratio": -0.21318714320659637, | |
| "eval_logits/chosen": -2.423281669616699, | |
| "eval_logits/rejected": -2.29191255569458, | |
| "eval_logps/chosen": -1.25368332862854, | |
| "eval_logps/rejected": -2.5206007957458496, | |
| "eval_loss": 1.0984269380569458, | |
| "eval_nll_loss": 1.1060694456100464, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": -0.18805250525474548, | |
| "eval_rewards/margins": 0.19003766775131226, | |
| "eval_rewards/rejected": -0.37809017300605774, | |
| "eval_runtime": 0.7836, | |
| "eval_samples_per_second": 174.83, | |
| "eval_steps_per_second": 6.381, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 7.867612293144208, | |
| "grad_norm": 17.922605514526367, | |
| "learning_rate": 2.974982725547975e-07, | |
| "log_odds_chosen": 1.3920270204544067, | |
| "log_odds_ratio": -0.4031583368778229, | |
| "logits/chosen": -1.9781438112258911, | |
| "logits/rejected": -1.9409700632095337, | |
| "logps/chosen": -1.2178527116775513, | |
| "logps/rejected": -2.3021371364593506, | |
| "loss": 1.2447, | |
| "nll_loss": 1.1541370153427124, | |
| "rewards/accuracies": 0.7734375, | |
| "rewards/chosen": -0.18267790973186493, | |
| "rewards/margins": 0.16264265775680542, | |
| "rewards/rejected": -0.34532058238983154, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 8.472813238770685, | |
| "grad_norm": 18.74088478088379, | |
| "learning_rate": 2.6453620722761895e-07, | |
| "log_odds_chosen": 1.4127886295318604, | |
| "log_odds_ratio": -0.45162278413772583, | |
| "logits/chosen": -1.9333115816116333, | |
| "logits/rejected": -1.9409185647964478, | |
| "logps/chosen": -1.3663212060928345, | |
| "logps/rejected": -2.5033111572265625, | |
| "loss": 1.2511, | |
| "nll_loss": 1.2117295265197754, | |
| "rewards/accuracies": 0.76953125, | |
| "rewards/chosen": -0.20494820177555084, | |
| "rewards/margins": 0.17054852843284607, | |
| "rewards/rejected": -0.37549668550491333, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 9.078014184397164, | |
| "grad_norm": 22.932889938354492, | |
| "learning_rate": 2.3131747660339394e-07, | |
| "log_odds_chosen": 1.6455353498458862, | |
| "log_odds_ratio": -0.44734472036361694, | |
| "logits/chosen": -1.923959493637085, | |
| "logits/rejected": -1.9963738918304443, | |
| "logps/chosen": -1.3357137441635132, | |
| "logps/rejected": -2.676542043685913, | |
| "loss": 1.2336, | |
| "nll_loss": 1.197737455368042, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.2003570795059204, | |
| "rewards/margins": 0.2011241912841797, | |
| "rewards/rejected": -0.4014812707901001, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 9.078014184397164, | |
| "eval_log_odds_chosen": 1.4782460927963257, | |
| "eval_log_odds_ratio": -0.21918949484825134, | |
| "eval_logits/chosen": -2.3784029483795166, | |
| "eval_logits/rejected": -2.244081497192383, | |
| "eval_logps/chosen": -1.296876072883606, | |
| "eval_logps/rejected": -2.5429465770721436, | |
| "eval_loss": 1.1173924207687378, | |
| "eval_nll_loss": 1.114406704902649, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": -0.1945313960313797, | |
| "eval_rewards/margins": 0.18691061437129974, | |
| "eval_rewards/rejected": -0.38144201040267944, | |
| "eval_runtime": 0.8027, | |
| "eval_samples_per_second": 170.681, | |
| "eval_steps_per_second": 6.229, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 9.68321513002364, | |
| "grad_norm": 27.342180252075195, | |
| "learning_rate": 1.984286226342056e-07, | |
| "log_odds_chosen": 2.1517550945281982, | |
| "log_odds_ratio": -0.3539085388183594, | |
| "logits/chosen": -2.0003914833068848, | |
| "logits/rejected": -1.958829641342163, | |
| "logps/chosen": -1.2932095527648926, | |
| "logps/rejected": -3.1245839595794678, | |
| "loss": 1.2519, | |
| "nll_loss": 1.177643060684204, | |
| "rewards/accuracies": 0.796875, | |
| "rewards/chosen": -0.19398145377635956, | |
| "rewards/margins": 0.2747061550617218, | |
| "rewards/rejected": -0.46868759393692017, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 10.288416075650119, | |
| "grad_norm": 21.41707992553711, | |
| "learning_rate": 1.6645036265170313e-07, | |
| "log_odds_chosen": 2.206300973892212, | |
| "log_odds_ratio": -0.41246891021728516, | |
| "logits/chosen": -1.9226995706558228, | |
| "logits/rejected": -1.9452672004699707, | |
| "logps/chosen": -1.3459149599075317, | |
| "logps/rejected": -3.243032455444336, | |
| "loss": 1.2377, | |
| "nll_loss": 1.1743229627609253, | |
| "rewards/accuracies": 0.73828125, | |
| "rewards/chosen": -0.20188726484775543, | |
| "rewards/margins": 0.28456762433052063, | |
| "rewards/rejected": -0.48645487427711487, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 10.591016548463356, | |
| "eval_log_odds_chosen": 1.4928518533706665, | |
| "eval_log_odds_ratio": -0.21719364821910858, | |
| "eval_logits/chosen": -2.406038999557495, | |
| "eval_logits/rejected": -2.2726240158081055, | |
| "eval_logps/chosen": -1.2588163614273071, | |
| "eval_logps/rejected": -2.5091373920440674, | |
| "eval_loss": 1.0936493873596191, | |
| "eval_nll_loss": 1.0926154851913452, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": -0.18882247805595398, | |
| "eval_rewards/margins": 0.18754813075065613, | |
| "eval_rewards/rejected": -0.3763706088066101, | |
| "eval_runtime": 0.7924, | |
| "eval_samples_per_second": 172.882, | |
| "eval_steps_per_second": 6.31, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 10.893617021276595, | |
| "grad_norm": 28.251604080200195, | |
| "learning_rate": 1.3594733566170925e-07, | |
| "log_odds_chosen": 1.9746395349502563, | |
| "log_odds_ratio": -0.39856040477752686, | |
| "logits/chosen": -1.9557344913482666, | |
| "logits/rejected": -1.9873769283294678, | |
| "logps/chosen": -1.3181676864624023, | |
| "logps/rejected": -2.9867465496063232, | |
| "loss": 1.2325, | |
| "nll_loss": 1.2073771953582764, | |
| "rewards/accuracies": 0.796875, | |
| "rewards/chosen": -0.19772517681121826, | |
| "rewards/margins": 0.2502868175506592, | |
| "rewards/rejected": -0.44801196455955505, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 11.498817966903074, | |
| "grad_norm": 30.6031551361084, | |
| "learning_rate": 1.0745813253325956e-07, | |
| "log_odds_chosen": 2.6722493171691895, | |
| "log_odds_ratio": -0.35422736406326294, | |
| "logits/chosen": -1.9136030673980713, | |
| "logits/rejected": -1.8901042938232422, | |
| "logps/chosen": -1.233724594116211, | |
| "logps/rejected": -3.5643980503082275, | |
| "loss": 1.2295, | |
| "nll_loss": 1.1232733726501465, | |
| "rewards/accuracies": 0.79296875, | |
| "rewards/chosen": -0.18505869805812836, | |
| "rewards/margins": 0.3496010899543762, | |
| "rewards/rejected": -0.534659743309021, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 12.10401891252955, | |
| "grad_norm": 24.5542049407959, | |
| "learning_rate": 8.148578611867113e-08, | |
| "log_odds_chosen": 2.4431307315826416, | |
| "log_odds_ratio": -0.3909703195095062, | |
| "logits/chosen": -1.8700110912322998, | |
| "logits/rejected": -1.9457833766937256, | |
| "logps/chosen": -1.2895874977111816, | |
| "logps/rejected": -3.4220337867736816, | |
| "loss": 1.2212, | |
| "nll_loss": 1.1831854581832886, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.19343814253807068, | |
| "rewards/margins": 0.31986698508262634, | |
| "rewards/rejected": -0.5133051872253418, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 12.10401891252955, | |
| "eval_log_odds_chosen": 1.515697956085205, | |
| "eval_log_odds_ratio": -0.213613823056221, | |
| "eval_logits/chosen": -2.373903274536133, | |
| "eval_logits/rejected": -2.2387218475341797, | |
| "eval_logps/chosen": -1.2586114406585693, | |
| "eval_logps/rejected": -2.530747175216675, | |
| "eval_loss": 1.0882254838943481, | |
| "eval_nll_loss": 1.085294485092163, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": -0.18879172205924988, | |
| "eval_rewards/margins": 0.19082039594650269, | |
| "eval_rewards/rejected": -0.37961211800575256, | |
| "eval_runtime": 0.8037, | |
| "eval_samples_per_second": 170.464, | |
| "eval_steps_per_second": 6.221, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 12.709219858156029, | |
| "grad_norm": 19.59035301208496, | |
| "learning_rate": 5.848888922025552e-08, | |
| "log_odds_chosen": 2.298971176147461, | |
| "log_odds_ratio": -0.4026568830013275, | |
| "logits/chosen": -1.9471426010131836, | |
| "logits/rejected": -1.9182159900665283, | |
| "logps/chosen": -1.3048053979873657, | |
| "logps/rejected": -3.2884950637817383, | |
| "loss": 1.2173, | |
| "nll_loss": 1.1782861948013306, | |
| "rewards/accuracies": 0.76171875, | |
| "rewards/chosen": -0.19572080671787262, | |
| "rewards/margins": 0.2975533902645111, | |
| "rewards/rejected": -0.4932742416858673, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 13.314420803782506, | |
| "grad_norm": 29.107627868652344, | |
| "learning_rate": 3.887349723342303e-08, | |
| "log_odds_chosen": 2.654096841812134, | |
| "log_odds_ratio": -0.38686317205429077, | |
| "logits/chosen": -1.9159326553344727, | |
| "logits/rejected": -1.9410839080810547, | |
| "logps/chosen": -1.2585792541503906, | |
| "logps/rejected": -3.5659923553466797, | |
| "loss": 1.2261, | |
| "nll_loss": 1.1586594581604004, | |
| "rewards/accuracies": 0.76953125, | |
| "rewards/chosen": -0.18878689408302307, | |
| "rewards/margins": 0.34611204266548157, | |
| "rewards/rejected": -0.5348988771438599, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 13.617021276595745, | |
| "eval_log_odds_chosen": 1.5199410915374756, | |
| "eval_log_odds_ratio": -0.21233825385570526, | |
| "eval_logits/chosen": -2.366751194000244, | |
| "eval_logits/rejected": -2.2316524982452393, | |
| "eval_logps/chosen": -1.2421499490737915, | |
| "eval_logps/rejected": -2.5126953125, | |
| "eval_loss": 1.0794531106948853, | |
| "eval_nll_loss": 1.0783252716064453, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": -0.18632249534130096, | |
| "eval_rewards/margins": 0.190581813454628, | |
| "eval_rewards/rejected": -0.37690430879592896, | |
| "eval_runtime": 0.7829, | |
| "eval_samples_per_second": 175.001, | |
| "eval_steps_per_second": 6.387, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 13.919621749408984, | |
| "grad_norm": 22.09730339050293, | |
| "learning_rate": 2.298595844092377e-08, | |
| "log_odds_chosen": 2.671109199523926, | |
| "log_odds_ratio": -0.3479268252849579, | |
| "logits/chosen": -2.0034313201904297, | |
| "logits/rejected": -1.870398998260498, | |
| "logps/chosen": -1.2474991083145142, | |
| "logps/rejected": -3.5732622146606445, | |
| "loss": 1.2194, | |
| "nll_loss": 1.1217308044433594, | |
| "rewards/accuracies": 0.80859375, | |
| "rewards/chosen": -0.18712489306926727, | |
| "rewards/margins": 0.3488644063472748, | |
| "rewards/rejected": -0.5359892845153809, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 14.52482269503546, | |
| "grad_norm": 24.665891647338867, | |
| "learning_rate": 1.1106798553464802e-08, | |
| "log_odds_chosen": 2.7288260459899902, | |
| "log_odds_ratio": -0.34282395243644714, | |
| "logits/chosen": -1.9847919940948486, | |
| "logits/rejected": -1.9164719581604004, | |
| "logps/chosen": -1.250791311264038, | |
| "logps/rejected": -3.6267054080963135, | |
| "loss": 1.2016, | |
| "nll_loss": 1.1445385217666626, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -0.18761873245239258, | |
| "rewards/margins": 0.3563871383666992, | |
| "rewards/rejected": -0.5440058708190918, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 15.130023640661939, | |
| "grad_norm": 47.552574157714844, | |
| "learning_rate": 3.4457674771554422e-09, | |
| "log_odds_chosen": 2.26895809173584, | |
| "log_odds_ratio": -0.3884022831916809, | |
| "logits/chosen": -1.9943946599960327, | |
| "logits/rejected": -1.961279034614563, | |
| "logps/chosen": -1.3015713691711426, | |
| "logps/rejected": -3.266200304031372, | |
| "loss": 1.2176, | |
| "nll_loss": 1.1944975852966309, | |
| "rewards/accuracies": 0.76953125, | |
| "rewards/chosen": -0.1952357143163681, | |
| "rewards/margins": 0.29469433426856995, | |
| "rewards/rejected": -0.48993009328842163, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 15.130023640661939, | |
| "eval_log_odds_chosen": 1.5204813480377197, | |
| "eval_log_odds_ratio": -0.21171817183494568, | |
| "eval_logits/chosen": -2.370277166366577, | |
| "eval_logits/rejected": -2.233962059020996, | |
| "eval_logps/chosen": -1.2457667589187622, | |
| "eval_logps/rejected": -2.517129898071289, | |
| "eval_loss": 1.081107497215271, | |
| "eval_nll_loss": 1.0765924453735352, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": -0.18686501681804657, | |
| "eval_rewards/margins": 0.19070449471473694, | |
| "eval_rewards/rejected": -0.3775694966316223, | |
| "eval_runtime": 0.8025, | |
| "eval_samples_per_second": 170.724, | |
| "eval_steps_per_second": 6.231, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 15.735224586288416, | |
| "grad_norm": 17.138763427734375, | |
| "learning_rate": 1.3813576683111006e-10, | |
| "log_odds_chosen": 2.620955228805542, | |
| "log_odds_ratio": -0.34706899523735046, | |
| "logits/chosen": -1.9596831798553467, | |
| "logits/rejected": -1.9431588649749756, | |
| "logps/chosen": -1.2250535488128662, | |
| "logps/rejected": -3.4995248317718506, | |
| "loss": 1.2191, | |
| "nll_loss": 1.129616379737854, | |
| "rewards/accuracies": 0.78515625, | |
| "rewards/chosen": -0.18375803530216217, | |
| "rewards/margins": 0.34117066860198975, | |
| "rewards/rejected": -0.5249287486076355, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 15.886524822695035, | |
| "grad_norm": 26.069650650024414, | |
| "learning_rate": 0.0, | |
| "log_odds_chosen": 2.719402551651001, | |
| "log_odds_ratio": -0.3425367772579193, | |
| "logits/chosen": -2.0392448902130127, | |
| "logits/rejected": -1.9386732578277588, | |
| "logps/chosen": -1.2458713054656982, | |
| "logps/rejected": -3.6082699298858643, | |
| "loss": 1.2157, | |
| "nll_loss": 1.1475220918655396, | |
| "rewards/accuracies": 0.828125, | |
| "rewards/chosen": -0.1868807077407837, | |
| "rewards/margins": 0.35435977578163147, | |
| "rewards/rejected": -0.5412405133247375, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 15.886524822695035, | |
| "eval_log_odds_chosen": 1.5112136602401733, | |
| "eval_log_odds_ratio": -0.2135576754808426, | |
| "eval_logits/chosen": -2.3702893257141113, | |
| "eval_logits/rejected": -2.2342278957366943, | |
| "eval_logps/chosen": -1.250806450843811, | |
| "eval_logps/rejected": -2.515653133392334, | |
| "eval_loss": 1.0810010433197021, | |
| "eval_nll_loss": 1.0763533115386963, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": -0.18762096762657166, | |
| "eval_rewards/margins": 0.18972699344158173, | |
| "eval_rewards/rejected": -0.37734800577163696, | |
| "eval_runtime": 0.7924, | |
| "eval_samples_per_second": 172.894, | |
| "eval_steps_per_second": 6.31, | |
| "step": 840 | |
| } | |
| ], | |
| "logging_steps": 32, | |
| "max_steps": 840, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 17, | |
| "save_steps": 80, | |
| "total_flos": 0.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |