| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9963459196102313, | |
| "eval_steps": 500, | |
| "global_step": 615, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.048721071863580996, | |
| "grad_norm": 556.0, | |
| "learning_rate": 5.000000000000001e-07, | |
| "log_odds_chosen": 0.11725157499313354, | |
| "log_odds_ratio": -0.7116702795028687, | |
| "logits/chosen": -2.241136312484741, | |
| "logits/rejected": -2.2574338912963867, | |
| "logps/chosen": -1.0354639291763306, | |
| "logps/rejected": -1.1154835224151611, | |
| "loss": 4.3799, | |
| "nll_loss": 4.344295501708984, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.05177319049835205, | |
| "rewards/margins": 0.004000981338322163, | |
| "rewards/rejected": -0.05577417463064194, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.09744214372716199, | |
| "grad_norm": 55.25, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "log_odds_chosen": 0.11950838565826416, | |
| "log_odds_ratio": -0.7092320322990417, | |
| "logits/chosen": -2.8477842807769775, | |
| "logits/rejected": -2.8416748046875, | |
| "logps/chosen": -0.9889892339706421, | |
| "logps/rejected": -1.076500654220581, | |
| "loss": 1.8135, | |
| "nll_loss": 1.7780277729034424, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.049449462443590164, | |
| "rewards/margins": 0.004375572316348553, | |
| "rewards/rejected": -0.053825028240680695, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.146163215590743, | |
| "grad_norm": 5.40625, | |
| "learning_rate": 1.5e-06, | |
| "log_odds_chosen": 0.15654945373535156, | |
| "log_odds_ratio": -0.690517783164978, | |
| "logits/chosen": -2.6702427864074707, | |
| "logits/rejected": -2.6656832695007324, | |
| "logps/chosen": -0.8676761388778687, | |
| "logps/rejected": -0.9682536125183105, | |
| "loss": 0.6655, | |
| "nll_loss": 0.6310166716575623, | |
| "rewards/accuracies": 0.590624988079071, | |
| "rewards/chosen": -0.04338381439447403, | |
| "rewards/margins": 0.005028870422393084, | |
| "rewards/rejected": -0.04841268062591553, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.19488428745432398, | |
| "grad_norm": 3.796875, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "log_odds_chosen": 0.17244641482830048, | |
| "log_odds_ratio": -0.6753009557723999, | |
| "logits/chosen": -2.296032190322876, | |
| "logits/rejected": -2.29901123046875, | |
| "logps/chosen": -0.8293353319168091, | |
| "logps/rejected": -0.9353267550468445, | |
| "loss": 0.5876, | |
| "nll_loss": 0.5538541078567505, | |
| "rewards/accuracies": 0.578125, | |
| "rewards/chosen": -0.041466765105724335, | |
| "rewards/margins": 0.005299570504575968, | |
| "rewards/rejected": -0.04676634073257446, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.243605359317905, | |
| "grad_norm": 3.96875, | |
| "learning_rate": 2.5e-06, | |
| "log_odds_chosen": 0.21714143455028534, | |
| "log_odds_ratio": -0.6588594913482666, | |
| "logits/chosen": -2.160338878631592, | |
| "logits/rejected": -2.1593730449676514, | |
| "logps/chosen": -0.7916803359985352, | |
| "logps/rejected": -0.9191333651542664, | |
| "loss": 0.5514, | |
| "nll_loss": 0.5184302926063538, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.03958401829004288, | |
| "rewards/margins": 0.0063726468943059444, | |
| "rewards/rejected": -0.04595666378736496, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.292326431181486, | |
| "grad_norm": 4.71875, | |
| "learning_rate": 3e-06, | |
| "log_odds_chosen": 0.25414514541625977, | |
| "log_odds_ratio": -0.6507801413536072, | |
| "logits/chosen": -2.1048073768615723, | |
| "logits/rejected": -2.1529033184051514, | |
| "logps/chosen": -0.8171817660331726, | |
| "logps/rejected": -0.9584506750106812, | |
| "loss": 0.591, | |
| "nll_loss": 0.5584203004837036, | |
| "rewards/accuracies": 0.621874988079071, | |
| "rewards/chosen": -0.04085909202694893, | |
| "rewards/margins": 0.007063449826091528, | |
| "rewards/rejected": -0.04792254418134689, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.341047503045067, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 3.5e-06, | |
| "log_odds_chosen": 0.2314440906047821, | |
| "log_odds_ratio": -0.6606293320655823, | |
| "logits/chosen": -2.1196396350860596, | |
| "logits/rejected": -2.0706403255462646, | |
| "logps/chosen": -0.8082675933837891, | |
| "logps/rejected": -0.9356250762939453, | |
| "loss": 0.5785, | |
| "nll_loss": 0.5454204082489014, | |
| "rewards/accuracies": 0.590624988079071, | |
| "rewards/chosen": -0.04041338339447975, | |
| "rewards/margins": 0.00636786874383688, | |
| "rewards/rejected": -0.046781253069639206, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.38976857490864797, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 4.000000000000001e-06, | |
| "log_odds_chosen": 0.2176472246646881, | |
| "log_odds_ratio": -0.6596580743789673, | |
| "logits/chosen": -2.1406569480895996, | |
| "logits/rejected": -2.1560683250427246, | |
| "logps/chosen": -0.7807229161262512, | |
| "logps/rejected": -0.8939612507820129, | |
| "loss": 0.5547, | |
| "nll_loss": 0.5217040181159973, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.03903614357113838, | |
| "rewards/margins": 0.005661919247359037, | |
| "rewards/rejected": -0.044698067009449005, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.438489646772229, | |
| "grad_norm": 4.5, | |
| "learning_rate": 4.5e-06, | |
| "log_odds_chosen": 0.1901816874742508, | |
| "log_odds_ratio": -0.6751978993415833, | |
| "logits/chosen": -2.1606240272521973, | |
| "logits/rejected": -2.130833387374878, | |
| "logps/chosen": -0.8014599084854126, | |
| "logps/rejected": -0.9177049398422241, | |
| "loss": 0.5793, | |
| "nll_loss": 0.5454980134963989, | |
| "rewards/accuracies": 0.5718749761581421, | |
| "rewards/chosen": -0.04007299616932869, | |
| "rewards/margins": 0.005812253803014755, | |
| "rewards/rejected": -0.045885246247053146, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.48721071863581, | |
| "grad_norm": 4.28125, | |
| "learning_rate": 5e-06, | |
| "log_odds_chosen": 0.24776022136211395, | |
| "log_odds_ratio": -0.6568383574485779, | |
| "logits/chosen": -2.105705499649048, | |
| "logits/rejected": -2.1442131996154785, | |
| "logps/chosen": -0.7829509973526001, | |
| "logps/rejected": -0.9177125692367554, | |
| "loss": 0.5837, | |
| "nll_loss": 0.5508524179458618, | |
| "rewards/accuracies": 0.590624988079071, | |
| "rewards/chosen": -0.039147552102804184, | |
| "rewards/margins": 0.006738076917827129, | |
| "rewards/rejected": -0.04588562995195389, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.535931790499391, | |
| "grad_norm": 4.5, | |
| "learning_rate": 4.767312946227961e-06, | |
| "log_odds_chosen": 0.16261616349220276, | |
| "log_odds_ratio": -0.6998116970062256, | |
| "logits/chosen": -2.0691206455230713, | |
| "logits/rejected": -2.097303867340088, | |
| "logps/chosen": -0.8038315773010254, | |
| "logps/rejected": -0.890066921710968, | |
| "loss": 0.5773, | |
| "nll_loss": 0.5423145890235901, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.04019157961010933, | |
| "rewards/margins": 0.004311761818826199, | |
| "rewards/rejected": -0.0445033423602581, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.584652862362972, | |
| "grad_norm": 3.8125, | |
| "learning_rate": 4.564354645876385e-06, | |
| "log_odds_chosen": 0.3128460645675659, | |
| "log_odds_ratio": -0.6486107110977173, | |
| "logits/chosen": -2.045989513397217, | |
| "logits/rejected": -2.06305193901062, | |
| "logps/chosen": -0.7878965139389038, | |
| "logps/rejected": -0.9589905738830566, | |
| "loss": 0.5752, | |
| "nll_loss": 0.5428130030632019, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.03939482569694519, | |
| "rewards/margins": 0.00855470634996891, | |
| "rewards/rejected": -0.04794953018426895, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.633373934226553, | |
| "grad_norm": 4.1875, | |
| "learning_rate": 4.385290096535147e-06, | |
| "log_odds_chosen": 0.26915115118026733, | |
| "log_odds_ratio": -0.6629001498222351, | |
| "logits/chosen": -2.160104751586914, | |
| "logits/rejected": -2.165109157562256, | |
| "logps/chosen": -0.7626894116401672, | |
| "logps/rejected": -0.8972024917602539, | |
| "loss": 0.5541, | |
| "nll_loss": 0.5209673643112183, | |
| "rewards/accuracies": 0.590624988079071, | |
| "rewards/chosen": -0.03813446685671806, | |
| "rewards/margins": 0.006725652609020472, | |
| "rewards/rejected": -0.044860124588012695, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.682095006090134, | |
| "grad_norm": 4.71875, | |
| "learning_rate": 4.2257712736425835e-06, | |
| "log_odds_chosen": 0.2980436682701111, | |
| "log_odds_ratio": -0.644582211971283, | |
| "logits/chosen": -2.170442581176758, | |
| "logits/rejected": -2.1650166511535645, | |
| "logps/chosen": -0.7542210817337036, | |
| "logps/rejected": -0.9323924779891968, | |
| "loss": 0.58, | |
| "nll_loss": 0.547797679901123, | |
| "rewards/accuracies": 0.621874988079071, | |
| "rewards/chosen": -0.03771105408668518, | |
| "rewards/margins": 0.008908570744097233, | |
| "rewards/rejected": -0.04661962762475014, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.730816077953715, | |
| "grad_norm": 3.375, | |
| "learning_rate": 4.082482904638631e-06, | |
| "log_odds_chosen": 0.30688202381134033, | |
| "log_odds_ratio": -0.6490530371665955, | |
| "logits/chosen": -2.1752607822418213, | |
| "logits/rejected": -2.20690655708313, | |
| "logps/chosen": -0.7646127939224243, | |
| "logps/rejected": -0.9283273816108704, | |
| "loss": 0.5539, | |
| "nll_loss": 0.5214100480079651, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.038230642676353455, | |
| "rewards/margins": 0.008185726590454578, | |
| "rewards/rejected": -0.04641636833548546, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.7795371498172959, | |
| "grad_norm": 4.25, | |
| "learning_rate": 3.952847075210474e-06, | |
| "log_odds_chosen": 0.29972273111343384, | |
| "log_odds_ratio": -0.6498032212257385, | |
| "logits/chosen": -2.222160816192627, | |
| "logits/rejected": -2.2057909965515137, | |
| "logps/chosen": -0.7570406198501587, | |
| "logps/rejected": -0.9201406240463257, | |
| "loss": 0.586, | |
| "nll_loss": 0.5535241365432739, | |
| "rewards/accuracies": 0.6156250238418579, | |
| "rewards/chosen": -0.03785203397274017, | |
| "rewards/margins": 0.008154998533427715, | |
| "rewards/rejected": -0.046007029712200165, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.8282582216808769, | |
| "grad_norm": 3.75, | |
| "learning_rate": 3.834824944236852e-06, | |
| "log_odds_chosen": 0.25258108973503113, | |
| "log_odds_ratio": -0.6628280878067017, | |
| "logits/chosen": -2.217949628829956, | |
| "logits/rejected": -2.2262425422668457, | |
| "logps/chosen": -0.7875797748565674, | |
| "logps/rejected": -0.9305909276008606, | |
| "loss": 0.5898, | |
| "nll_loss": 0.5566290616989136, | |
| "rewards/accuracies": 0.565625011920929, | |
| "rewards/chosen": -0.03937898576259613, | |
| "rewards/margins": 0.007150554563850164, | |
| "rewards/rejected": -0.04652954638004303, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.876979293544458, | |
| "grad_norm": 4.25, | |
| "learning_rate": 3.72677996249965e-06, | |
| "log_odds_chosen": 0.29803532361984253, | |
| "log_odds_ratio": -0.6474822759628296, | |
| "logits/chosen": -2.164008617401123, | |
| "logits/rejected": -2.1809074878692627, | |
| "logps/chosen": -0.7780044674873352, | |
| "logps/rejected": -0.935949981212616, | |
| "loss": 0.5573, | |
| "nll_loss": 0.5249563455581665, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.0389002226293087, | |
| "rewards/margins": 0.007897274568676949, | |
| "rewards/rejected": -0.0467974953353405, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.925700365408039, | |
| "grad_norm": 3.640625, | |
| "learning_rate": 3.6273812505500587e-06, | |
| "log_odds_chosen": 0.3355644941329956, | |
| "log_odds_ratio": -0.6257966756820679, | |
| "logits/chosen": -2.1745223999023438, | |
| "logits/rejected": -2.168818712234497, | |
| "logps/chosen": -0.7714961767196655, | |
| "logps/rejected": -0.9696172475814819, | |
| "loss": 0.5722, | |
| "nll_loss": 0.5409447550773621, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.038574814796447754, | |
| "rewards/margins": 0.009906048886477947, | |
| "rewards/rejected": -0.048480864614248276, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.97442143727162, | |
| "grad_norm": 3.734375, | |
| "learning_rate": 3.5355339059327378e-06, | |
| "log_odds_chosen": 0.24752295017242432, | |
| "log_odds_ratio": -0.6712836623191833, | |
| "logits/chosen": -2.1981072425842285, | |
| "logits/rejected": -2.2192599773406982, | |
| "logps/chosen": -0.7749738097190857, | |
| "logps/rejected": -0.9182379841804504, | |
| "loss": 0.6001, | |
| "nll_loss": 0.5665073394775391, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.038748692721128464, | |
| "rewards/margins": 0.007163210306316614, | |
| "rewards/rejected": -0.04591190069913864, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.0231425091352009, | |
| "grad_norm": 4.21875, | |
| "learning_rate": 3.450327796711771e-06, | |
| "log_odds_chosen": 0.5075238943099976, | |
| "log_odds_ratio": -0.5618590116500854, | |
| "logits/chosen": -2.1940836906433105, | |
| "logits/rejected": -2.2036900520324707, | |
| "logps/chosen": -0.683228611946106, | |
| "logps/rejected": -0.935465931892395, | |
| "loss": 0.5621, | |
| "nll_loss": 0.5340477228164673, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.03416142612695694, | |
| "rewards/margins": 0.012611865997314453, | |
| "rewards/rejected": -0.04677329212427139, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.071863580998782, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 3.3709993123162106e-06, | |
| "log_odds_chosen": 0.7354512214660645, | |
| "log_odds_ratio": -0.47807660698890686, | |
| "logits/chosen": -2.1200661659240723, | |
| "logits/rejected": -2.1521944999694824, | |
| "logps/chosen": -0.6185888051986694, | |
| "logps/rejected": -0.9742799997329712, | |
| "loss": 0.4596, | |
| "nll_loss": 0.4357299208641052, | |
| "rewards/accuracies": 0.765625, | |
| "rewards/chosen": -0.03092944249510765, | |
| "rewards/margins": 0.01778455823659897, | |
| "rewards/rejected": -0.04871400073170662, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.1205846528623629, | |
| "grad_norm": 3.9375, | |
| "learning_rate": 3.296902366978936e-06, | |
| "log_odds_chosen": 0.8707483410835266, | |
| "log_odds_ratio": -0.44475650787353516, | |
| "logits/chosen": -2.1387903690338135, | |
| "logits/rejected": -2.1741206645965576, | |
| "logps/chosen": -0.6047109365463257, | |
| "logps/rejected": -1.0183531045913696, | |
| "loss": 0.4634, | |
| "nll_loss": 0.44113415479660034, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -0.030235549435019493, | |
| "rewards/margins": 0.020682107657194138, | |
| "rewards/rejected": -0.05091765522956848, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.169305724725944, | |
| "grad_norm": 4.53125, | |
| "learning_rate": 3.2274861218395142e-06, | |
| "log_odds_chosen": 0.8473652005195618, | |
| "log_odds_ratio": -0.44497156143188477, | |
| "logits/chosen": -2.1265828609466553, | |
| "logits/rejected": -2.0939793586730957, | |
| "logps/chosen": -0.6094004511833191, | |
| "logps/rejected": -1.039342999458313, | |
| "loss": 0.4756, | |
| "nll_loss": 0.45334190130233765, | |
| "rewards/accuracies": 0.815625011920929, | |
| "rewards/chosen": -0.030470022931694984, | |
| "rewards/margins": 0.021497130393981934, | |
| "rewards/rejected": -0.05196715518832207, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.218026796589525, | |
| "grad_norm": 4.28125, | |
| "learning_rate": 3.1622776601683796e-06, | |
| "log_odds_chosen": 0.8328806161880493, | |
| "log_odds_ratio": -0.4545086920261383, | |
| "logits/chosen": -2.0806257724761963, | |
| "logits/rejected": -2.0717406272888184, | |
| "logps/chosen": -0.6114749312400818, | |
| "logps/rejected": -1.0371878147125244, | |
| "loss": 0.4832, | |
| "nll_loss": 0.4604315757751465, | |
| "rewards/accuracies": 0.796875, | |
| "rewards/chosen": -0.03057374618947506, | |
| "rewards/margins": 0.02128564938902855, | |
| "rewards/rejected": -0.051859401166439056, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.266747868453106, | |
| "grad_norm": 5.125, | |
| "learning_rate": 3.1008683647302113e-06, | |
| "log_odds_chosen": 0.8561245203018188, | |
| "log_odds_ratio": -0.4485153257846832, | |
| "logits/chosen": -2.0740833282470703, | |
| "logits/rejected": -2.1148364543914795, | |
| "logps/chosen": -0.603646457195282, | |
| "logps/rejected": -1.02623450756073, | |
| "loss": 0.482, | |
| "nll_loss": 0.4595448970794678, | |
| "rewards/accuracies": 0.8031250238418579, | |
| "rewards/chosen": -0.030182326212525368, | |
| "rewards/margins": 0.021129406988620758, | |
| "rewards/rejected": -0.05131173133850098, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.315468940316687, | |
| "grad_norm": 4.125, | |
| "learning_rate": 3.0429030972509227e-06, | |
| "log_odds_chosen": 0.9330509305000305, | |
| "log_odds_ratio": -0.43321356177330017, | |
| "logits/chosen": -2.0687127113342285, | |
| "logits/rejected": -2.073645830154419, | |
| "logps/chosen": -0.5700651407241821, | |
| "logps/rejected": -1.0073726177215576, | |
| "loss": 0.4551, | |
| "nll_loss": 0.43348783254623413, | |
| "rewards/accuracies": 0.815625011920929, | |
| "rewards/chosen": -0.028503259643912315, | |
| "rewards/margins": 0.021865375339984894, | |
| "rewards/rejected": -0.05036862939596176, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.364190012180268, | |
| "grad_norm": 3.78125, | |
| "learning_rate": 2.988071523335984e-06, | |
| "log_odds_chosen": 0.8365095257759094, | |
| "log_odds_ratio": -0.4509243369102478, | |
| "logits/chosen": -2.0936789512634277, | |
| "logits/rejected": -2.0444083213806152, | |
| "logps/chosen": -0.6157883405685425, | |
| "logps/rejected": -1.018973469734192, | |
| "loss": 0.4629, | |
| "nll_loss": 0.440369188785553, | |
| "rewards/accuracies": 0.809374988079071, | |
| "rewards/chosen": -0.030789416283369064, | |
| "rewards/margins": 0.02015925757586956, | |
| "rewards/rejected": -0.050948671996593475, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.412911084043849, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 2.9361010975735177e-06, | |
| "log_odds_chosen": 0.933612048625946, | |
| "log_odds_ratio": -0.42886632680892944, | |
| "logits/chosen": -2.07072114944458, | |
| "logits/rejected": -2.0590546131134033, | |
| "logps/chosen": -0.5958141684532166, | |
| "logps/rejected": -1.0696933269500732, | |
| "loss": 0.4313, | |
| "nll_loss": 0.4098600745201111, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -0.029790710657835007, | |
| "rewards/margins": 0.023693958297371864, | |
| "rewards/rejected": -0.05348466709256172, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.46163215590743, | |
| "grad_norm": 4.125, | |
| "learning_rate": 2.8867513459481293e-06, | |
| "log_odds_chosen": 0.8847057223320007, | |
| "log_odds_ratio": -0.44255512952804565, | |
| "logits/chosen": -2.094334840774536, | |
| "logits/rejected": -2.1165547370910645, | |
| "logps/chosen": -0.5963028073310852, | |
| "logps/rejected": -1.0416440963745117, | |
| "loss": 0.4379, | |
| "nll_loss": 0.41574278473854065, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -0.02981514297425747, | |
| "rewards/margins": 0.022267062216997147, | |
| "rewards/rejected": -0.05208220332860947, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.510353227771011, | |
| "grad_norm": 4.84375, | |
| "learning_rate": 2.839809171235324e-06, | |
| "log_odds_chosen": 0.9941097497940063, | |
| "log_odds_ratio": -0.4086780548095703, | |
| "logits/chosen": -2.080404758453369, | |
| "logits/rejected": -2.112900495529175, | |
| "logps/chosen": -0.5736903548240662, | |
| "logps/rejected": -1.081747055053711, | |
| "loss": 0.4603, | |
| "nll_loss": 0.4398314952850342, | |
| "rewards/accuracies": 0.840624988079071, | |
| "rewards/chosen": -0.028684521093964577, | |
| "rewards/margins": 0.02540283277630806, | |
| "rewards/rejected": -0.05408735200762749, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.559074299634592, | |
| "grad_norm": 4.3125, | |
| "learning_rate": 2.7950849718747376e-06, | |
| "log_odds_chosen": 0.883891224861145, | |
| "log_odds_ratio": -0.44051599502563477, | |
| "logits/chosen": -2.092296600341797, | |
| "logits/rejected": -2.066316604614258, | |
| "logps/chosen": -0.6078993678092957, | |
| "logps/rejected": -1.0367143154144287, | |
| "loss": 0.4823, | |
| "nll_loss": 0.46028652787208557, | |
| "rewards/accuracies": 0.8031250238418579, | |
| "rewards/chosen": -0.030394967645406723, | |
| "rewards/margins": 0.021440746262669563, | |
| "rewards/rejected": -0.051835715770721436, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.607795371498173, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 2.752409412815902e-06, | |
| "log_odds_chosen": 0.9730150103569031, | |
| "log_odds_ratio": -0.43135911226272583, | |
| "logits/chosen": -2.0477657318115234, | |
| "logits/rejected": -2.051565408706665, | |
| "logps/chosen": -0.5975598096847534, | |
| "logps/rejected": -1.0904568433761597, | |
| "loss": 0.4697, | |
| "nll_loss": 0.4481421113014221, | |
| "rewards/accuracies": 0.815625011920929, | |
| "rewards/chosen": -0.02987799420952797, | |
| "rewards/margins": 0.024644847959280014, | |
| "rewards/rejected": -0.054522834718227386, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.6565164433617539, | |
| "grad_norm": 4.09375, | |
| "learning_rate": 2.711630722733202e-06, | |
| "log_odds_chosen": 0.9206117391586304, | |
| "log_odds_ratio": -0.43162840604782104, | |
| "logits/chosen": -2.1549911499023438, | |
| "logits/rejected": -2.1604723930358887, | |
| "logps/chosen": -0.5994877815246582, | |
| "logps/rejected": -1.04535973072052, | |
| "loss": 0.4752, | |
| "nll_loss": 0.4536053538322449, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -0.02997438982129097, | |
| "rewards/margins": 0.02229359745979309, | |
| "rewards/rejected": -0.05226798728108406, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.705237515225335, | |
| "grad_norm": 4.6875, | |
| "learning_rate": 2.6726124191242444e-06, | |
| "log_odds_chosen": 0.9252131581306458, | |
| "log_odds_ratio": -0.44289445877075195, | |
| "logits/chosen": -2.0990488529205322, | |
| "logits/rejected": -2.1114163398742676, | |
| "logps/chosen": -0.601850688457489, | |
| "logps/rejected": -1.0636966228485107, | |
| "loss": 0.4599, | |
| "nll_loss": 0.43774834275245667, | |
| "rewards/accuracies": 0.815625011920929, | |
| "rewards/chosen": -0.03009253740310669, | |
| "rewards/margins": 0.023092303425073624, | |
| "rewards/rejected": -0.05318482965230942, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.753958587088916, | |
| "grad_norm": 3.8125, | |
| "learning_rate": 2.6352313834736496e-06, | |
| "log_odds_chosen": 0.9648879766464233, | |
| "log_odds_ratio": -0.4200161397457123, | |
| "logits/chosen": -2.1074578762054443, | |
| "logits/rejected": -2.1522512435913086, | |
| "logps/chosen": -0.5663856267929077, | |
| "logps/rejected": -1.0540645122528076, | |
| "loss": 0.4401, | |
| "nll_loss": 0.4191323220729828, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -0.028319280594587326, | |
| "rewards/margins": 0.02438395284116268, | |
| "rewards/rejected": -0.052703239023685455, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.802679658952497, | |
| "grad_norm": 4.90625, | |
| "learning_rate": 2.599376224550182e-06, | |
| "log_odds_chosen": 0.8812922239303589, | |
| "log_odds_ratio": -0.459873765707016, | |
| "logits/chosen": -2.059195041656494, | |
| "logits/rejected": -2.111880302429199, | |
| "logps/chosen": -0.6145040392875671, | |
| "logps/rejected": -1.0735037326812744, | |
| "loss": 0.4598, | |
| "nll_loss": 0.4368101954460144, | |
| "rewards/accuracies": 0.7906249761581421, | |
| "rewards/chosen": -0.030725205317139626, | |
| "rewards/margins": 0.022949980571866035, | |
| "rewards/rejected": -0.05367518588900566, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.8514007308160778, | |
| "grad_norm": 3.71875, | |
| "learning_rate": 2.564945880212886e-06, | |
| "log_odds_chosen": 0.8604068756103516, | |
| "log_odds_ratio": -0.4609376788139343, | |
| "logits/chosen": -2.0991945266723633, | |
| "logits/rejected": -2.093503952026367, | |
| "logps/chosen": -0.6222652196884155, | |
| "logps/rejected": -1.0542787313461304, | |
| "loss": 0.4708, | |
| "nll_loss": 0.4477114677429199, | |
| "rewards/accuracies": 0.778124988079071, | |
| "rewards/chosen": -0.031113261356949806, | |
| "rewards/margins": 0.02160067670047283, | |
| "rewards/rejected": -0.05271393805742264, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.900121802679659, | |
| "grad_norm": 5.09375, | |
| "learning_rate": 2.5318484177091667e-06, | |
| "log_odds_chosen": 0.9894090890884399, | |
| "log_odds_ratio": -0.4191601872444153, | |
| "logits/chosen": -2.1074106693267822, | |
| "logits/rejected": -2.1691222190856934, | |
| "logps/chosen": -0.5862478017807007, | |
| "logps/rejected": -1.0842490196228027, | |
| "loss": 0.4614, | |
| "nll_loss": 0.440448135137558, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -0.029312390834093094, | |
| "rewards/margins": 0.024900058284401894, | |
| "rewards/rejected": -0.05421245098114014, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.94884287454324, | |
| "grad_norm": 4.28125, | |
| "learning_rate": 2.5e-06, | |
| "log_odds_chosen": 0.8292287588119507, | |
| "log_odds_ratio": -0.4579356610774994, | |
| "logits/chosen": -2.0651729106903076, | |
| "logits/rejected": -2.0955402851104736, | |
| "logps/chosen": -0.6199088096618652, | |
| "logps/rejected": -1.0601669549942017, | |
| "loss": 0.4434, | |
| "nll_loss": 0.4204865097999573, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -0.030995439738035202, | |
| "rewards/margins": 0.02201290801167488, | |
| "rewards/rejected": -0.05300834774971008, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.997563946406821, | |
| "grad_norm": 3.8125, | |
| "learning_rate": 2.4693239916239746e-06, | |
| "log_odds_chosen": 1.0317885875701904, | |
| "log_odds_ratio": -0.4280542731285095, | |
| "logits/chosen": -2.066987991333008, | |
| "logits/rejected": -2.069795608520508, | |
| "logps/chosen": -0.6221650242805481, | |
| "logps/rejected": -1.1912611722946167, | |
| "loss": 0.451, | |
| "nll_loss": 0.4295925199985504, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": -0.031108248978853226, | |
| "rewards/margins": 0.02845480665564537, | |
| "rewards/rejected": -0.059563059359788895, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.0462850182704018, | |
| "grad_norm": 4.59375, | |
| "learning_rate": 2.4397501823713327e-06, | |
| "log_odds_chosen": 1.3798599243164062, | |
| "log_odds_ratio": -0.31996873021125793, | |
| "logits/chosen": -2.0742099285125732, | |
| "logits/rejected": -2.1023483276367188, | |
| "logps/chosen": -0.4915800094604492, | |
| "logps/rejected": -1.1646963357925415, | |
| "loss": 0.3809, | |
| "nll_loss": 0.3648887574672699, | |
| "rewards/accuracies": 0.909375011920929, | |
| "rewards/chosen": -0.02457900531589985, | |
| "rewards/margins": 0.033655811101198196, | |
| "rewards/rejected": -0.05823482200503349, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 2.095006090133983, | |
| "grad_norm": 4.0, | |
| "learning_rate": 2.411214110852061e-06, | |
| "log_odds_chosen": 1.586724042892456, | |
| "log_odds_ratio": -0.28053292632102966, | |
| "logits/chosen": -2.0577170848846436, | |
| "logits/rejected": -2.0551247596740723, | |
| "logps/chosen": -0.46747732162475586, | |
| "logps/rejected": -1.2207533121109009, | |
| "loss": 0.3688, | |
| "nll_loss": 0.35477322340011597, | |
| "rewards/accuracies": 0.921875, | |
| "rewards/chosen": -0.023373866453766823, | |
| "rewards/margins": 0.03766379505395889, | |
| "rewards/rejected": -0.06103766709566116, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 2.143727161997564, | |
| "grad_norm": 5.625, | |
| "learning_rate": 2.3836564731139807e-06, | |
| "log_odds_chosen": 1.6717441082000732, | |
| "log_odds_ratio": -0.26605507731437683, | |
| "logits/chosen": -2.0567383766174316, | |
| "logits/rejected": -2.067713737487793, | |
| "logps/chosen": -0.4729752540588379, | |
| "logps/rejected": -1.3001043796539307, | |
| "loss": 0.3854, | |
| "nll_loss": 0.3720557689666748, | |
| "rewards/accuracies": 0.940625011920929, | |
| "rewards/chosen": -0.023648761212825775, | |
| "rewards/margins": 0.04135645180940628, | |
| "rewards/rejected": -0.06500521302223206, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 2.192448233861145, | |
| "grad_norm": 4.96875, | |
| "learning_rate": 2.357022603955159e-06, | |
| "log_odds_chosen": 1.6444238424301147, | |
| "log_odds_ratio": -0.26991745829582214, | |
| "logits/chosen": -2.2070858478546143, | |
| "logits/rejected": -2.1682567596435547, | |
| "logps/chosen": -0.4466155469417572, | |
| "logps/rejected": -1.245265007019043, | |
| "loss": 0.4138, | |
| "nll_loss": 0.4003148674964905, | |
| "rewards/accuracies": 0.940625011920929, | |
| "rewards/chosen": -0.02233077771961689, | |
| "rewards/margins": 0.03993247076869011, | |
| "rewards/rejected": -0.06226325035095215, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.2411693057247257, | |
| "grad_norm": 5.71875, | |
| "learning_rate": 2.3312620206007847e-06, | |
| "log_odds_chosen": 1.5212812423706055, | |
| "log_odds_ratio": -0.2893092632293701, | |
| "logits/chosen": -2.103972911834717, | |
| "logits/rejected": -2.1150119304656982, | |
| "logps/chosen": -0.4900006353855133, | |
| "logps/rejected": -1.2329685688018799, | |
| "loss": 0.3907, | |
| "nll_loss": 0.37621983885765076, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -0.024500032886862755, | |
| "rewards/margins": 0.03714838624000549, | |
| "rewards/rejected": -0.0616484172642231, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 2.289890377588307, | |
| "grad_norm": 4.90625, | |
| "learning_rate": 2.3063280200722128e-06, | |
| "log_odds_chosen": 1.515355110168457, | |
| "log_odds_ratio": -0.2837259769439697, | |
| "logits/chosen": -2.0579495429992676, | |
| "logits/rejected": -2.06086802482605, | |
| "logps/chosen": -0.48503416776657104, | |
| "logps/rejected": -1.2265126705169678, | |
| "loss": 0.3629, | |
| "nll_loss": 0.3486883044242859, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -0.024251705035567284, | |
| "rewards/margins": 0.037073928862810135, | |
| "rewards/rejected": -0.06132563203573227, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 2.338611449451888, | |
| "grad_norm": 4.125, | |
| "learning_rate": 2.2821773229381924e-06, | |
| "log_odds_chosen": 1.5133790969848633, | |
| "log_odds_ratio": -0.2952267527580261, | |
| "logits/chosen": -2.070732831954956, | |
| "logits/rejected": -2.091557264328003, | |
| "logps/chosen": -0.4965738356113434, | |
| "logps/rejected": -1.2001986503601074, | |
| "loss": 0.3841, | |
| "nll_loss": 0.36938291788101196, | |
| "rewards/accuracies": 0.921875, | |
| "rewards/chosen": -0.02482869103550911, | |
| "rewards/margins": 0.03518123924732208, | |
| "rewards/rejected": -0.06000993400812149, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 2.387332521315469, | |
| "grad_norm": 3.828125, | |
| "learning_rate": 2.2587697572631284e-06, | |
| "log_odds_chosen": 1.6050021648406982, | |
| "log_odds_ratio": -0.26656216382980347, | |
| "logits/chosen": -2.079113006591797, | |
| "logits/rejected": -2.0700507164001465, | |
| "logps/chosen": -0.4672022759914398, | |
| "logps/rejected": -1.2427198886871338, | |
| "loss": 0.379, | |
| "nll_loss": 0.36565306782722473, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -0.02336011826992035, | |
| "rewards/margins": 0.03877587243914604, | |
| "rewards/rejected": -0.06213599443435669, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 2.43605359317905, | |
| "grad_norm": 5.375, | |
| "learning_rate": 2.23606797749979e-06, | |
| "log_odds_chosen": 1.6788904666900635, | |
| "log_odds_ratio": -0.27105051279067993, | |
| "logits/chosen": -2.0833709239959717, | |
| "logits/rejected": -2.119617462158203, | |
| "logps/chosen": -0.46906137466430664, | |
| "logps/rejected": -1.315861701965332, | |
| "loss": 0.3826, | |
| "nll_loss": 0.3690149486064911, | |
| "rewards/accuracies": 0.9593750238418579, | |
| "rewards/chosen": -0.023453067988157272, | |
| "rewards/margins": 0.042340025305747986, | |
| "rewards/rejected": -0.06579308956861496, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.484774665042631, | |
| "grad_norm": 4.90625, | |
| "learning_rate": 2.2140372138502386e-06, | |
| "log_odds_chosen": 1.6214439868927002, | |
| "log_odds_ratio": -0.2923833727836609, | |
| "logits/chosen": -2.117619037628174, | |
| "logits/rejected": -2.144683599472046, | |
| "logps/chosen": -0.483691543340683, | |
| "logps/rejected": -1.252079725265503, | |
| "loss": 0.4067, | |
| "nll_loss": 0.39209839701652527, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": -0.0241845790296793, | |
| "rewards/margins": 0.03841940313577652, | |
| "rewards/rejected": -0.06260398775339127, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 2.533495736906212, | |
| "grad_norm": 4.875, | |
| "learning_rate": 2.1926450482675734e-06, | |
| "log_odds_chosen": 1.5657522678375244, | |
| "log_odds_ratio": -0.2806842625141144, | |
| "logits/chosen": -2.1533446311950684, | |
| "logits/rejected": -2.1672301292419434, | |
| "logps/chosen": -0.4697963297367096, | |
| "logps/rejected": -1.2120527029037476, | |
| "loss": 0.3891, | |
| "nll_loss": 0.37509655952453613, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -0.02348981983959675, | |
| "rewards/margins": 0.03711282089352608, | |
| "rewards/rejected": -0.060602642595767975, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 2.582216808769793, | |
| "grad_norm": 4.625, | |
| "learning_rate": 2.1718612138153473e-06, | |
| "log_odds_chosen": 1.6662880182266235, | |
| "log_odds_ratio": -0.2821272611618042, | |
| "logits/chosen": -2.0334420204162598, | |
| "logits/rejected": -2.054309368133545, | |
| "logps/chosen": -0.4613516926765442, | |
| "logps/rejected": -1.2510273456573486, | |
| "loss": 0.3906, | |
| "nll_loss": 0.37650758028030396, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/chosen": -0.02306758239865303, | |
| "rewards/margins": 0.03948378562927246, | |
| "rewards/rejected": -0.06255136430263519, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 2.630937880633374, | |
| "grad_norm": 4.96875, | |
| "learning_rate": 2.151657414559676e-06, | |
| "log_odds_chosen": 1.5510852336883545, | |
| "log_odds_ratio": -0.2845192849636078, | |
| "logits/chosen": -2.105376958847046, | |
| "logits/rejected": -2.107239246368408, | |
| "logps/chosen": -0.481005996465683, | |
| "logps/rejected": -1.2433946132659912, | |
| "loss": 0.3943, | |
| "nll_loss": 0.3801063597202301, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": -0.02405030094087124, | |
| "rewards/margins": 0.03811942785978317, | |
| "rewards/rejected": -0.06216973066329956, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.679658952496955, | |
| "grad_norm": 4.4375, | |
| "learning_rate": 2.132007163556104e-06, | |
| "log_odds_chosen": 1.535161018371582, | |
| "log_odds_ratio": -0.28597292304039, | |
| "logits/chosen": -2.1408169269561768, | |
| "logits/rejected": -2.1645288467407227, | |
| "logps/chosen": -0.4741552472114563, | |
| "logps/rejected": -1.1751224994659424, | |
| "loss": 0.4205, | |
| "nll_loss": 0.4061834216117859, | |
| "rewards/accuracies": 0.9281250238418579, | |
| "rewards/chosen": -0.023707758635282516, | |
| "rewards/margins": 0.035048358142375946, | |
| "rewards/rejected": -0.05875612050294876, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.728380024360536, | |
| "grad_norm": 4.84375, | |
| "learning_rate": 2.1128856368212917e-06, | |
| "log_odds_chosen": 1.8219788074493408, | |
| "log_odds_ratio": -0.2337840050458908, | |
| "logits/chosen": -2.091169595718384, | |
| "logits/rejected": -2.1250526905059814, | |
| "logps/chosen": -0.4375804364681244, | |
| "logps/rejected": -1.326407551765442, | |
| "loss": 0.3892, | |
| "nll_loss": 0.3775223195552826, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.02187902107834816, | |
| "rewards/margins": 0.044441353529691696, | |
| "rewards/rejected": -0.06632037460803986, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.7771010962241167, | |
| "grad_norm": 4.0, | |
| "learning_rate": 2.0942695414584777e-06, | |
| "log_odds_chosen": 1.6477181911468506, | |
| "log_odds_ratio": -0.2635224461555481, | |
| "logits/chosen": -2.0645499229431152, | |
| "logits/rejected": -2.097146511077881, | |
| "logps/chosen": -0.477451890707016, | |
| "logps/rejected": -1.3132137060165405, | |
| "loss": 0.3748, | |
| "nll_loss": 0.361587256193161, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -0.02387259528040886, | |
| "rewards/margins": 0.041788093745708466, | |
| "rewards/rejected": -0.06566068530082703, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.825822168087698, | |
| "grad_norm": 4.625, | |
| "learning_rate": 2.0761369963434992e-06, | |
| "log_odds_chosen": 1.7657368183135986, | |
| "log_odds_ratio": -0.25923803448677063, | |
| "logits/chosen": -2.1280486583709717, | |
| "logits/rejected": -2.1294822692871094, | |
| "logps/chosen": -0.42457103729248047, | |
| "logps/rejected": -1.2813599109649658, | |
| "loss": 0.3643, | |
| "nll_loss": 0.35132378339767456, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -0.021228553727269173, | |
| "rewards/margins": 0.04283944517374039, | |
| "rewards/rejected": -0.06406799703836441, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.874543239951279, | |
| "grad_norm": 4.65625, | |
| "learning_rate": 2.058467423981546e-06, | |
| "log_odds_chosen": 1.6043407917022705, | |
| "log_odds_ratio": -0.28762462735176086, | |
| "logits/chosen": -2.1606733798980713, | |
| "logits/rejected": -2.171504259109497, | |
| "logps/chosen": -0.4778304100036621, | |
| "logps/rejected": -1.2797237634658813, | |
| "loss": 0.4117, | |
| "nll_loss": 0.3973469138145447, | |
| "rewards/accuracies": 0.909375011920929, | |
| "rewards/chosen": -0.023891519755125046, | |
| "rewards/margins": 0.04009466618299484, | |
| "rewards/rejected": -0.06398618966341019, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.92326431181486, | |
| "grad_norm": 3.96875, | |
| "learning_rate": 2.0412414523193154e-06, | |
| "log_odds_chosen": 1.6591618061065674, | |
| "log_odds_ratio": -0.28386634588241577, | |
| "logits/chosen": -2.105055809020996, | |
| "logits/rejected": -2.16955304145813, | |
| "logps/chosen": -0.4936911463737488, | |
| "logps/rejected": -1.315927267074585, | |
| "loss": 0.3883, | |
| "nll_loss": 0.37410837411880493, | |
| "rewards/accuracies": 0.909375011920929, | |
| "rewards/chosen": -0.02468455769121647, | |
| "rewards/margins": 0.04111180454492569, | |
| "rewards/rejected": -0.0657963678240776, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.971985383678441, | |
| "grad_norm": 5.09375, | |
| "learning_rate": 2.0244408254472904e-06, | |
| "log_odds_chosen": 1.6492611169815063, | |
| "log_odds_ratio": -0.2732219099998474, | |
| "logits/chosen": -2.124809741973877, | |
| "logits/rejected": -2.1507182121276855, | |
| "logps/chosen": -0.477315753698349, | |
| "logps/rejected": -1.2909691333770752, | |
| "loss": 0.3918, | |
| "nll_loss": 0.3781156837940216, | |
| "rewards/accuracies": 0.940625011920929, | |
| "rewards/chosen": -0.02386578544974327, | |
| "rewards/margins": 0.04068267345428467, | |
| "rewards/rejected": -0.06454845517873764, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.9963459196102313, | |
| "step": 615, | |
| "total_flos": 0.0, | |
| "train_loss": 0.5590515291787744, | |
| "train_runtime": 15994.549, | |
| "train_samples_per_second": 1.232, | |
| "train_steps_per_second": 0.038 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 615, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |