| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.9595687331536387, | |
| "eval_steps": 500, | |
| "global_step": 368, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05390835579514825, | |
| "grad_norm": 684.0, | |
| "learning_rate": 6.7567567567567575e-06, | |
| "log_odds_chosen": 5.920718193054199, | |
| "log_odds_ratio": -7.724557399749756, | |
| "logps/chosen": -24.315807342529297, | |
| "logps/rejected": -30.238388061523438, | |
| "loss": 193.2993, | |
| "nll_loss": 11.694974899291992, | |
| "rewards/accuracies": 0.484375, | |
| "rewards/chosen": -1.2157905101776123, | |
| "rewards/margins": 0.2961289882659912, | |
| "rewards/rejected": -1.511919379234314, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.1078167115902965, | |
| "grad_norm": 218.0, | |
| "learning_rate": 1.3513513513513515e-05, | |
| "log_odds_chosen": 4.100552082061768, | |
| "log_odds_ratio": -5.447903633117676, | |
| "logps/chosen": -20.245830535888672, | |
| "logps/rejected": -24.348098754882812, | |
| "loss": 161.9053, | |
| "nll_loss": 9.846681594848633, | |
| "rewards/accuracies": 0.4781250059604645, | |
| "rewards/chosen": -1.012291669845581, | |
| "rewards/margins": 0.2051134556531906, | |
| "rewards/rejected": -1.2174049615859985, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.16172506738544473, | |
| "grad_norm": 205.0, | |
| "learning_rate": 2.0270270270270273e-05, | |
| "log_odds_chosen": 3.8738200664520264, | |
| "log_odds_ratio": -6.855284690856934, | |
| "logps/chosen": -21.959657669067383, | |
| "logps/rejected": -25.834619522094727, | |
| "loss": 166.4944, | |
| "nll_loss": 10.06313419342041, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -1.0979830026626587, | |
| "rewards/margins": 0.19374798238277435, | |
| "rewards/rejected": -1.2917309999465942, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.215633423180593, | |
| "grad_norm": 648.0, | |
| "learning_rate": 2.702702702702703e-05, | |
| "log_odds_chosen": 0.7052766680717468, | |
| "log_odds_ratio": -7.368393898010254, | |
| "logps/chosen": -20.677827835083008, | |
| "logps/rejected": -21.383352279663086, | |
| "loss": 141.3241, | |
| "nll_loss": 8.464337348937988, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -1.0338914394378662, | |
| "rewards/margins": 0.03527634218335152, | |
| "rewards/rejected": -1.0691677331924438, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2695417789757412, | |
| "grad_norm": 182.0, | |
| "learning_rate": 3.3783783783783784e-05, | |
| "log_odds_chosen": -0.0883919969201088, | |
| "log_odds_ratio": -2.0752124786376953, | |
| "logps/chosen": -6.3797478675842285, | |
| "logps/rejected": -6.272956371307373, | |
| "loss": 62.7556, | |
| "nll_loss": 3.8184638023376465, | |
| "rewards/accuracies": 0.559374988079071, | |
| "rewards/chosen": -0.3189873993396759, | |
| "rewards/margins": -0.0053395600989460945, | |
| "rewards/rejected": -0.3136478364467621, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.32345013477088946, | |
| "grad_norm": 83.0, | |
| "learning_rate": 4.0540540540540545e-05, | |
| "log_odds_chosen": 0.24213147163391113, | |
| "log_odds_ratio": -0.7704355716705322, | |
| "logps/chosen": -1.9613044261932373, | |
| "logps/rejected": -2.175011157989502, | |
| "loss": 36.6421, | |
| "nll_loss": 2.2516086101531982, | |
| "rewards/accuracies": 0.565625011920929, | |
| "rewards/chosen": -0.09806521981954575, | |
| "rewards/margins": 0.010685333982110023, | |
| "rewards/rejected": -0.10875056684017181, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.37735849056603776, | |
| "grad_norm": 78.0, | |
| "learning_rate": 4.72972972972973e-05, | |
| "log_odds_chosen": 0.27958303689956665, | |
| "log_odds_ratio": -0.7305089831352234, | |
| "logps/chosen": -1.6892141103744507, | |
| "logps/rejected": -1.930132269859314, | |
| "loss": 30.9266, | |
| "nll_loss": 1.8963871002197266, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.08446070551872253, | |
| "rewards/margins": 0.01204589568078518, | |
| "rewards/rejected": -0.09650660306215286, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.431266846361186, | |
| "grad_norm": 62.5, | |
| "learning_rate": 4.998986632578596e-05, | |
| "log_odds_chosen": 0.2721244692802429, | |
| "log_odds_ratio": -0.7381750345230103, | |
| "logps/chosen": -1.7667934894561768, | |
| "logps/rejected": -1.9824241399765015, | |
| "loss": 31.6068, | |
| "nll_loss": 1.9385162591934204, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.0883396789431572, | |
| "rewards/margins": 0.010781526565551758, | |
| "rewards/rejected": -0.09912119805812836, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.48517520215633425, | |
| "grad_norm": 46.75, | |
| "learning_rate": 4.992796806510936e-05, | |
| "log_odds_chosen": 0.1983262300491333, | |
| "log_odds_ratio": -0.7065704464912415, | |
| "logps/chosen": -1.4196051359176636, | |
| "logps/rejected": -1.579620122909546, | |
| "loss": 28.8934, | |
| "nll_loss": 1.7705100774765015, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.0709802582859993, | |
| "rewards/margins": 0.008000737987458706, | |
| "rewards/rejected": -0.07898099720478058, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5390835579514824, | |
| "grad_norm": 33.0, | |
| "learning_rate": 4.980994056736854e-05, | |
| "log_odds_chosen": 0.1537928730249405, | |
| "log_odds_ratio": -0.7110344171524048, | |
| "logps/chosen": -1.3953869342803955, | |
| "logps/rejected": -1.5007522106170654, | |
| "loss": 27.7613, | |
| "nll_loss": 1.699528694152832, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.06976933777332306, | |
| "rewards/margins": 0.005268273409456015, | |
| "rewards/rejected": -0.07503761351108551, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5929919137466307, | |
| "grad_norm": 56.0, | |
| "learning_rate": 4.9636049590020475e-05, | |
| "log_odds_chosen": 0.2435292750597, | |
| "log_odds_ratio": -0.6781546473503113, | |
| "logps/chosen": -1.3146178722381592, | |
| "logps/rejected": -1.4984357357025146, | |
| "loss": 26.7408, | |
| "nll_loss": 1.6373910903930664, | |
| "rewards/accuracies": 0.6031249761581421, | |
| "rewards/chosen": -0.06573089212179184, | |
| "rewards/margins": 0.009190889075398445, | |
| "rewards/rejected": -0.07492177933454514, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.6469002695417789, | |
| "grad_norm": 65.0, | |
| "learning_rate": 4.940668667592439e-05, | |
| "log_odds_chosen": 0.13836432993412018, | |
| "log_odds_ratio": -0.7060701847076416, | |
| "logps/chosen": -1.2619632482528687, | |
| "logps/rejected": -1.3755035400390625, | |
| "loss": 25.7309, | |
| "nll_loss": 1.5728803873062134, | |
| "rewards/accuracies": 0.590624988079071, | |
| "rewards/chosen": -0.06309816986322403, | |
| "rewards/margins": 0.005677012261003256, | |
| "rewards/rejected": -0.06877517700195312, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7008086253369272, | |
| "grad_norm": 204.0, | |
| "learning_rate": 4.912236827172148e-05, | |
| "log_odds_chosen": 0.13251568377017975, | |
| "log_odds_ratio": -0.7122939229011536, | |
| "logps/chosen": -1.2452380657196045, | |
| "logps/rejected": -1.3495495319366455, | |
| "loss": 25.7811, | |
| "nll_loss": 1.5757039785385132, | |
| "rewards/accuracies": 0.5218750238418579, | |
| "rewards/chosen": -0.062261901795864105, | |
| "rewards/margins": 0.005215569399297237, | |
| "rewards/rejected": -0.06747747957706451, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.7547169811320755, | |
| "grad_norm": 95.0, | |
| "learning_rate": 4.878373456497416e-05, | |
| "log_odds_chosen": 0.16559086740016937, | |
| "log_odds_ratio": -0.6909226179122925, | |
| "logps/chosen": -1.2156684398651123, | |
| "logps/rejected": -1.3327122926712036, | |
| "loss": 25.469, | |
| "nll_loss": 1.5572690963745117, | |
| "rewards/accuracies": 0.546875, | |
| "rewards/chosen": -0.06078341603279114, | |
| "rewards/margins": 0.005852191708981991, | |
| "rewards/rejected": -0.0666356161236763, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.8086253369272237, | |
| "grad_norm": 130.0, | |
| "learning_rate": 4.839154804268313e-05, | |
| "log_odds_chosen": 0.24878835678100586, | |
| "log_odds_ratio": -0.6528117656707764, | |
| "logps/chosen": -1.171622633934021, | |
| "logps/rejected": -1.3498855829238892, | |
| "loss": 25.0146, | |
| "nll_loss": 1.530771255493164, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.05858113244175911, | |
| "rewards/margins": 0.00891314446926117, | |
| "rewards/rejected": -0.06749428063631058, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.862533692722372, | |
| "grad_norm": 119.5, | |
| "learning_rate": 4.7946691774428144e-05, | |
| "log_odds_chosen": 0.2302316427230835, | |
| "log_odds_ratio": -0.6437969207763672, | |
| "logps/chosen": -1.1122291088104248, | |
| "logps/rejected": -1.2697585821151733, | |
| "loss": 24.4222, | |
| "nll_loss": 1.4941973686218262, | |
| "rewards/accuracies": 0.6156250238418579, | |
| "rewards/chosen": -0.05561145395040512, | |
| "rewards/margins": 0.007876471616327763, | |
| "rewards/rejected": -0.0634879320859909, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.9164420485175202, | |
| "grad_norm": 48.0, | |
| "learning_rate": 4.745016742399804e-05, | |
| "log_odds_chosen": 0.18154019117355347, | |
| "log_odds_ratio": -0.6740174889564514, | |
| "logps/chosen": -1.0852258205413818, | |
| "logps/rejected": -1.2075302600860596, | |
| "loss": 23.767, | |
| "nll_loss": 1.451735496520996, | |
| "rewards/accuracies": 0.578125, | |
| "rewards/chosen": -0.05426129698753357, | |
| "rewards/margins": 0.006115221884101629, | |
| "rewards/rejected": -0.06037651747465134, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.9703504043126685, | |
| "grad_norm": 121.0, | |
| "learning_rate": 4.690309299398736e-05, | |
| "log_odds_chosen": 0.1481303721666336, | |
| "log_odds_ratio": -0.6858943700790405, | |
| "logps/chosen": -1.0602507591247559, | |
| "logps/rejected": -1.1643625497817993, | |
| "loss": 23.2092, | |
| "nll_loss": 1.416282296180725, | |
| "rewards/accuracies": 0.565625011920929, | |
| "rewards/chosen": -0.05301254242658615, | |
| "rewards/margins": 0.005205585155636072, | |
| "rewards/rejected": -0.05821812152862549, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.0215633423180592, | |
| "grad_norm": 65.0, | |
| "learning_rate": 4.630670030843785e-05, | |
| "log_odds_chosen": 0.2527507245540619, | |
| "log_odds_ratio": -0.6484514474868774, | |
| "logps/chosen": -0.9999241232872009, | |
| "logps/rejected": -1.1713911294937134, | |
| "loss": 21.3461, | |
| "nll_loss": 1.3719265460968018, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.04999620094895363, | |
| "rewards/margins": 0.00857335701584816, | |
| "rewards/rejected": -0.05856956169009209, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.0754716981132075, | |
| "grad_norm": 39.0, | |
| "learning_rate": 4.566233223919298e-05, | |
| "log_odds_chosen": 0.5452004671096802, | |
| "log_odds_ratio": -0.554630696773529, | |
| "logps/chosen": -0.8976788520812988, | |
| "logps/rejected": -1.2380362749099731, | |
| "loss": 20.9491, | |
| "nll_loss": 1.2815890312194824, | |
| "rewards/accuracies": 0.690625011920929, | |
| "rewards/chosen": -0.04488394409418106, | |
| "rewards/margins": 0.017017865553498268, | |
| "rewards/rejected": -0.06190181523561478, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.1293800539083558, | |
| "grad_norm": 36.25, | |
| "learning_rate": 4.4971439682211125e-05, | |
| "log_odds_chosen": 0.5778145790100098, | |
| "log_odds_ratio": -0.5227857828140259, | |
| "logps/chosen": -0.8327204585075378, | |
| "logps/rejected": -1.1775634288787842, | |
| "loss": 19.2746, | |
| "nll_loss": 1.1785202026367188, | |
| "rewards/accuracies": 0.7281249761581421, | |
| "rewards/chosen": -0.04163602367043495, | |
| "rewards/margins": 0.017242146655917168, | |
| "rewards/rejected": -0.05887817218899727, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.1832884097035041, | |
| "grad_norm": 51.0, | |
| "learning_rate": 4.4235578290645194e-05, | |
| "log_odds_chosen": 0.564378023147583, | |
| "log_odds_ratio": -0.5335876941680908, | |
| "logps/chosen": -0.8683027029037476, | |
| "logps/rejected": -1.225411057472229, | |
| "loss": 20.1128, | |
| "nll_loss": 1.2303677797317505, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.0434151366353035, | |
| "rewards/margins": 0.017855416983366013, | |
| "rewards/rejected": -0.06127055361866951, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.2371967654986522, | |
| "grad_norm": 34.5, | |
| "learning_rate": 4.3456404972045216e-05, | |
| "log_odds_chosen": 0.6416139602661133, | |
| "log_odds_ratio": -0.5017353296279907, | |
| "logps/chosen": -0.8975638151168823, | |
| "logps/rejected": -1.311712384223938, | |
| "loss": 19.926, | |
| "nll_loss": 1.220287799835205, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.044878192245960236, | |
| "rewards/margins": 0.02070743218064308, | |
| "rewards/rejected": -0.06558562815189362, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.2911051212938005, | |
| "grad_norm": 29.375, | |
| "learning_rate": 4.263567415757058e-05, | |
| "log_odds_chosen": 0.5536849498748779, | |
| "log_odds_ratio": -0.5178900957107544, | |
| "logps/chosen": -0.8433765172958374, | |
| "logps/rejected": -1.1795700788497925, | |
| "loss": 18.7183, | |
| "nll_loss": 1.1439992189407349, | |
| "rewards/accuracies": 0.7718750238418579, | |
| "rewards/chosen": -0.04216882586479187, | |
| "rewards/margins": 0.016809683293104172, | |
| "rewards/rejected": -0.05897850915789604, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.3450134770889488, | |
| "grad_norm": 40.25, | |
| "learning_rate": 4.177523385161264e-05, | |
| "log_odds_chosen": 0.5022929906845093, | |
| "log_odds_ratio": -0.5432526469230652, | |
| "logps/chosen": -0.830506443977356, | |
| "logps/rejected": -1.1401817798614502, | |
| "loss": 19.2524, | |
| "nll_loss": 1.176114559173584, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.041525326669216156, | |
| "rewards/margins": 0.015483764931559563, | |
| "rewards/rejected": -0.05700908973813057, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.398921832884097, | |
| "grad_norm": 45.5, | |
| "learning_rate": 4.087702147072241e-05, | |
| "log_odds_chosen": 0.7220330238342285, | |
| "log_odds_ratio": -0.4684603810310364, | |
| "logps/chosen": -0.8251385688781738, | |
| "logps/rejected": -1.28031325340271, | |
| "loss": 18.9627, | |
| "nll_loss": 1.1617438793182373, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -0.04125693440437317, | |
| "rewards/margins": 0.022758739069104195, | |
| "rewards/rejected": -0.06401567161083221, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.4528301886792452, | |
| "grad_norm": 42.75, | |
| "learning_rate": 3.9943059481212795e-05, | |
| "log_odds_chosen": 0.6586893796920776, | |
| "log_odds_ratio": -0.4884856641292572, | |
| "logps/chosen": -0.8284580111503601, | |
| "logps/rejected": -1.2332229614257812, | |
| "loss": 18.9402, | |
| "nll_loss": 1.1593413352966309, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -0.041422899812459946, | |
| "rewards/margins": 0.02023824118077755, | |
| "rewards/rejected": -0.061661142855882645, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.5067385444743935, | |
| "grad_norm": 40.0, | |
| "learning_rate": 3.89754508452579e-05, | |
| "log_odds_chosen": 0.5820509195327759, | |
| "log_odds_ratio": -0.507893443107605, | |
| "logps/chosen": -0.8388631939888, | |
| "logps/rejected": -1.1933777332305908, | |
| "loss": 18.6589, | |
| "nll_loss": 1.1407865285873413, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": -0.04194316640496254, | |
| "rewards/margins": 0.01772572472691536, | |
| "rewards/rejected": -0.0596688874065876, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.5606469002695418, | |
| "grad_norm": 29.625, | |
| "learning_rate": 3.797637428574326e-05, | |
| "log_odds_chosen": 0.5326761603355408, | |
| "log_odds_ratio": -0.5247709155082703, | |
| "logps/chosen": -0.8665750622749329, | |
| "logps/rejected": -1.1989167928695679, | |
| "loss": 20.4033, | |
| "nll_loss": 1.2489675283432007, | |
| "rewards/accuracies": 0.765625, | |
| "rewards/chosen": -0.043328750878572464, | |
| "rewards/margins": 0.01661708578467369, | |
| "rewards/rejected": -0.05994584038853645, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.61455525606469, | |
| "grad_norm": 46.0, | |
| "learning_rate": 3.694807938052887e-05, | |
| "log_odds_chosen": 0.5733472108840942, | |
| "log_odds_ratio": -0.5206255912780762, | |
| "logps/chosen": -0.8152993321418762, | |
| "logps/rejected": -1.168148159980774, | |
| "loss": 18.796, | |
| "nll_loss": 1.1487157344818115, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.04076496511697769, | |
| "rewards/margins": 0.017642442137002945, | |
| "rewards/rejected": -0.05840740725398064, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.6684636118598384, | |
| "grad_norm": 47.5, | |
| "learning_rate": 3.589288149717119e-05, | |
| "log_odds_chosen": 0.5549741387367249, | |
| "log_odds_ratio": -0.5415369868278503, | |
| "logps/chosen": -0.8091446161270142, | |
| "logps/rejected": -1.1404829025268555, | |
| "loss": 18.6901, | |
| "nll_loss": 1.1410565376281738, | |
| "rewards/accuracies": 0.746874988079071, | |
| "rewards/chosen": -0.04045723378658295, | |
| "rewards/margins": 0.016566917300224304, | |
| "rewards/rejected": -0.057024143636226654, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.7223719676549867, | |
| "grad_norm": 41.25, | |
| "learning_rate": 3.481315657950931e-05, | |
| "log_odds_chosen": 0.6500066518783569, | |
| "log_odds_ratio": -0.49195536971092224, | |
| "logps/chosen": -0.8281671404838562, | |
| "logps/rejected": -1.2156283855438232, | |
| "loss": 18.8162, | |
| "nll_loss": 1.1514135599136353, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -0.04140836000442505, | |
| "rewards/margins": 0.01937306486070156, | |
| "rewards/rejected": -0.06078142672777176, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.7762803234501348, | |
| "grad_norm": 22.5, | |
| "learning_rate": 3.3711335797853977e-05, | |
| "log_odds_chosen": 0.7445698976516724, | |
| "log_odds_ratio": -0.4790865480899811, | |
| "logps/chosen": -0.8083413243293762, | |
| "logps/rejected": -1.2825214862823486, | |
| "loss": 18.5217, | |
| "nll_loss": 1.1336489915847778, | |
| "rewards/accuracies": 0.809374988079071, | |
| "rewards/chosen": -0.04041706770658493, | |
| "rewards/margins": 0.0237090103328228, | |
| "rewards/rejected": -0.06412607431411743, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.830188679245283, | |
| "grad_norm": 39.0, | |
| "learning_rate": 3.2589900074825696e-05, | |
| "log_odds_chosen": 0.6931416988372803, | |
| "log_odds_ratio": -0.4947904944419861, | |
| "logps/chosen": -0.7949181795120239, | |
| "logps/rejected": -1.2068727016448975, | |
| "loss": 18.9459, | |
| "nll_loss": 1.1593761444091797, | |
| "rewards/accuracies": 0.778124988079071, | |
| "rewards/chosen": -0.03974590823054314, | |
| "rewards/margins": 0.02059771679341793, | |
| "rewards/rejected": -0.060343630611896515, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.8840970350404311, | |
| "grad_norm": 47.25, | |
| "learning_rate": 3.1451374499167586e-05, | |
| "log_odds_chosen": 0.5809050798416138, | |
| "log_odds_ratio": -0.5110514163970947, | |
| "logps/chosen": -0.8269885182380676, | |
| "logps/rejected": -1.1618260145187378, | |
| "loss": 18.2167, | |
| "nll_loss": 1.112992286682129, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -0.04134942218661308, | |
| "rewards/margins": 0.016741881147027016, | |
| "rewards/rejected": -0.05809130519628525, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.9380053908355794, | |
| "grad_norm": 33.75, | |
| "learning_rate": 3.029832264011133e-05, | |
| "log_odds_chosen": 0.5635145902633667, | |
| "log_odds_ratio": -0.5329084396362305, | |
| "logps/chosen": -0.8633018732070923, | |
| "logps/rejected": -1.2152156829833984, | |
| "loss": 19.8266, | |
| "nll_loss": 1.2125170230865479, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.043165091425180435, | |
| "rewards/margins": 0.017595689743757248, | |
| "rewards/rejected": -0.06076078489422798, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.9919137466307277, | |
| "grad_norm": 26.375, | |
| "learning_rate": 2.91333407750982e-05, | |
| "log_odds_chosen": 0.6361402869224548, | |
| "log_odds_ratio": -0.5029143691062927, | |
| "logps/chosen": -0.8135613203048706, | |
| "logps/rejected": -1.1941924095153809, | |
| "loss": 18.7586, | |
| "nll_loss": 1.1472654342651367, | |
| "rewards/accuracies": 0.784375011920929, | |
| "rewards/chosen": -0.04067806527018547, | |
| "rewards/margins": 0.019031552597880363, | |
| "rewards/rejected": -0.05970961973071098, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.0431266846361185, | |
| "grad_norm": 28.625, | |
| "learning_rate": 2.7959052043852464e-05, | |
| "log_odds_chosen": 1.1101665496826172, | |
| "log_odds_ratio": -0.3634699881076813, | |
| "logps/chosen": -0.6150734424591064, | |
| "logps/rejected": -1.237518548965454, | |
| "loss": 13.763, | |
| "nll_loss": 0.8872840404510498, | |
| "rewards/accuracies": 0.8980262875556946, | |
| "rewards/chosen": -0.030753672122955322, | |
| "rewards/margins": 0.031122254207730293, | |
| "rewards/rejected": -0.06187593191862106, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.0970350404312668, | |
| "grad_norm": 26.625, | |
| "learning_rate": 2.677810054197009e-05, | |
| "log_odds_chosen": 1.4561563730239868, | |
| "log_odds_ratio": -0.28991881012916565, | |
| "logps/chosen": -0.5946121215820312, | |
| "logps/rejected": -1.417449712753296, | |
| "loss": 14.7392, | |
| "nll_loss": 0.906706690788269, | |
| "rewards/accuracies": 0.9281250238418579, | |
| "rewards/chosen": -0.029730603098869324, | |
| "rewards/margins": 0.04114187881350517, | |
| "rewards/rejected": -0.0708724781870842, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.150943396226415, | |
| "grad_norm": 36.5, | |
| "learning_rate": 2.559314536732212e-05, | |
| "log_odds_chosen": 1.3712774515151978, | |
| "log_odds_ratio": -0.2994323670864105, | |
| "logps/chosen": -0.5892333984375, | |
| "logps/rejected": -1.355689287185669, | |
| "loss": 13.7443, | |
| "nll_loss": 0.8440448045730591, | |
| "rewards/accuracies": 0.934374988079071, | |
| "rewards/chosen": -0.02946167066693306, | |
| "rewards/margins": 0.03832279518246651, | |
| "rewards/rejected": -0.06778446584939957, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.2048517520215634, | |
| "grad_norm": 25.75, | |
| "learning_rate": 2.4406854632677883e-05, | |
| "log_odds_chosen": 1.447076439857483, | |
| "log_odds_ratio": -0.29277992248535156, | |
| "logps/chosen": -0.5634891390800476, | |
| "logps/rejected": -1.386910080909729, | |
| "loss": 13.4321, | |
| "nll_loss": 0.8248661756515503, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/chosen": -0.02817445993423462, | |
| "rewards/margins": 0.04117104411125183, | |
| "rewards/rejected": -0.06934549659490585, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.2587601078167117, | |
| "grad_norm": 27.75, | |
| "learning_rate": 2.3221899458029918e-05, | |
| "log_odds_chosen": 1.4350051879882812, | |
| "log_odds_ratio": -0.2998976409435272, | |
| "logps/chosen": -0.5992333292961121, | |
| "logps/rejected": -1.4122200012207031, | |
| "loss": 14.0534, | |
| "nll_loss": 0.8633407354354858, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/chosen": -0.02996166981756687, | |
| "rewards/margins": 0.04064933583140373, | |
| "rewards/rejected": -0.07061101496219635, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.31266846361186, | |
| "grad_norm": 28.25, | |
| "learning_rate": 2.204094795614755e-05, | |
| "log_odds_chosen": 1.4530997276306152, | |
| "log_odds_ratio": -0.2895173728466034, | |
| "logps/chosen": -0.5493010878562927, | |
| "logps/rejected": -1.3495293855667114, | |
| "loss": 13.5655, | |
| "nll_loss": 0.8333690762519836, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": -0.027465051040053368, | |
| "rewards/margins": 0.04001142084598541, | |
| "rewards/rejected": -0.06747647374868393, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.3665768194070083, | |
| "grad_norm": 26.125, | |
| "learning_rate": 2.0866659224901812e-05, | |
| "log_odds_chosen": 1.4535211324691772, | |
| "log_odds_ratio": -0.2938300371170044, | |
| "logps/chosen": -0.5619036555290222, | |
| "logps/rejected": -1.3884742259979248, | |
| "loss": 13.1624, | |
| "nll_loss": 0.8079586029052734, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/chosen": -0.02809518575668335, | |
| "rewards/margins": 0.04132852703332901, | |
| "rewards/rejected": -0.06942371279001236, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.420485175202156, | |
| "grad_norm": 44.75, | |
| "learning_rate": 1.970167735988867e-05, | |
| "log_odds_chosen": 1.4392869472503662, | |
| "log_odds_ratio": -0.291705459356308, | |
| "logps/chosen": -0.5864852666854858, | |
| "logps/rejected": -1.387615442276001, | |
| "loss": 13.6079, | |
| "nll_loss": 0.8359111547470093, | |
| "rewards/accuracies": 0.940625011920929, | |
| "rewards/chosen": -0.029324263334274292, | |
| "rewards/margins": 0.0400565080344677, | |
| "rewards/rejected": -0.06938077509403229, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.4743935309973044, | |
| "grad_norm": 25.75, | |
| "learning_rate": 1.854862550083241e-05, | |
| "log_odds_chosen": 1.425066590309143, | |
| "log_odds_ratio": -0.29249343276023865, | |
| "logps/chosen": -0.558022141456604, | |
| "logps/rejected": -1.3558303117752075, | |
| "loss": 13.6401, | |
| "nll_loss": 0.83788001537323, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/chosen": -0.02790110372006893, | |
| "rewards/margins": 0.03989041596651077, | |
| "rewards/rejected": -0.06779151409864426, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.5283018867924527, | |
| "grad_norm": 31.125, | |
| "learning_rate": 1.7410099925174307e-05, | |
| "log_odds_chosen": 1.3924018144607544, | |
| "log_odds_ratio": -0.3068452477455139, | |
| "logps/chosen": -0.5728594064712524, | |
| "logps/rejected": -1.3289473056793213, | |
| "loss": 13.4421, | |
| "nll_loss": 0.8247901797294617, | |
| "rewards/accuracies": 0.921875, | |
| "rewards/chosen": -0.028642971068620682, | |
| "rewards/margins": 0.03780439496040344, | |
| "rewards/rejected": -0.06644736230373383, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.582210242587601, | |
| "grad_norm": 25.375, | |
| "learning_rate": 1.628866420214603e-05, | |
| "log_odds_chosen": 1.424349308013916, | |
| "log_odds_ratio": -0.3004930913448334, | |
| "logps/chosen": -0.5925456285476685, | |
| "logps/rejected": -1.407995581626892, | |
| "loss": 14.0136, | |
| "nll_loss": 0.8608237504959106, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/chosen": -0.029627284035086632, | |
| "rewards/margins": 0.04077249392867088, | |
| "rewards/rejected": -0.07039977610111237, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.6361185983827493, | |
| "grad_norm": 28.375, | |
| "learning_rate": 1.5186843420490698e-05, | |
| "log_odds_chosen": 1.4117343425750732, | |
| "log_odds_ratio": -0.29593607783317566, | |
| "logps/chosen": -0.5560811758041382, | |
| "logps/rejected": -1.315725564956665, | |
| "loss": 14.3972, | |
| "nll_loss": 0.8850256204605103, | |
| "rewards/accuracies": 0.921875, | |
| "rewards/chosen": -0.02780405804514885, | |
| "rewards/margins": 0.03798222541809082, | |
| "rewards/rejected": -0.06578627973794937, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.6900269541778976, | |
| "grad_norm": 25.0, | |
| "learning_rate": 1.4107118502828814e-05, | |
| "log_odds_chosen": 1.5662610530853271, | |
| "log_odds_ratio": -0.27495214343070984, | |
| "logps/chosen": -0.5344637036323547, | |
| "logps/rejected": -1.4078654050827026, | |
| "loss": 13.1208, | |
| "nll_loss": 0.8063005208969116, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -0.026723185554146767, | |
| "rewards/margins": 0.043670084327459335, | |
| "rewards/rejected": -0.07039327919483185, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.743935309973046, | |
| "grad_norm": 25.625, | |
| "learning_rate": 1.3051920619471145e-05, | |
| "log_odds_chosen": 1.4818260669708252, | |
| "log_odds_ratio": -0.2981964945793152, | |
| "logps/chosen": -0.5319411754608154, | |
| "logps/rejected": -1.3410605192184448, | |
| "loss": 13.8473, | |
| "nll_loss": 0.8505468368530273, | |
| "rewards/accuracies": 0.9156249761581421, | |
| "rewards/chosen": -0.02659705840051174, | |
| "rewards/margins": 0.04045596718788147, | |
| "rewards/rejected": -0.06705302000045776, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.797843665768194, | |
| "grad_norm": 20.875, | |
| "learning_rate": 1.2023625714256743e-05, | |
| "log_odds_chosen": 1.3853600025177002, | |
| "log_odds_ratio": -0.3066624402999878, | |
| "logps/chosen": -0.5818651914596558, | |
| "logps/rejected": -1.3556396961212158, | |
| "loss": 13.8096, | |
| "nll_loss": 0.8477646112442017, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/chosen": -0.02909325622022152, | |
| "rewards/margins": 0.03868872672319412, | |
| "rewards/rejected": -0.06778198480606079, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.8517520215633425, | |
| "grad_norm": 25.875, | |
| "learning_rate": 1.1024549154742098e-05, | |
| "log_odds_chosen": 1.3135344982147217, | |
| "log_odds_ratio": -0.3150535225868225, | |
| "logps/chosen": -0.5827392339706421, | |
| "logps/rejected": -1.295180320739746, | |
| "loss": 13.5359, | |
| "nll_loss": 0.8302389979362488, | |
| "rewards/accuracies": 0.9281250238418579, | |
| "rewards/chosen": -0.029136961326003075, | |
| "rewards/margins": 0.03562205657362938, | |
| "rewards/rejected": -0.0647590309381485, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.9056603773584904, | |
| "grad_norm": 26.75, | |
| "learning_rate": 1.005694051878721e-05, | |
| "log_odds_chosen": 1.3994379043579102, | |
| "log_odds_ratio": -0.3157843351364136, | |
| "logps/chosen": -0.5678838491439819, | |
| "logps/rejected": -1.3345766067504883, | |
| "loss": 13.6712, | |
| "nll_loss": 0.8386627435684204, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/chosen": -0.028394188731908798, | |
| "rewards/margins": 0.03833463788032532, | |
| "rewards/rejected": -0.06672882288694382, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.9595687331536387, | |
| "grad_norm": 25.5, | |
| "learning_rate": 9.1229785292776e-06, | |
| "log_odds_chosen": 1.4966779947280884, | |
| "log_odds_ratio": -0.28297117352485657, | |
| "logps/chosen": -0.5280012488365173, | |
| "logps/rejected": -1.359018325805664, | |
| "loss": 13.0722, | |
| "nll_loss": 0.8028658628463745, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/chosen": -0.026400065049529076, | |
| "rewards/margins": 0.04155084863305092, | |
| "rewards/rejected": -0.06795090436935425, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.01078167115903, | |
| "grad_norm": 24.5, | |
| "learning_rate": 8.22476614838737e-06, | |
| "log_odds_chosen": 1.5248451232910156, | |
| "log_odds_ratio": -0.2791544198989868, | |
| "logps/chosen": -0.540464460849762, | |
| "logps/rejected": -1.3801342248916626, | |
| "loss": 12.3085, | |
| "nll_loss": 0.7958097457885742, | |
| "rewards/accuracies": 0.9407894611358643, | |
| "rewards/chosen": -0.027023224160075188, | |
| "rewards/margins": 0.04198349267244339, | |
| "rewards/rejected": -0.06900670379400253, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.0646900269541777, | |
| "grad_norm": 20.75, | |
| "learning_rate": 7.364325842429417e-06, | |
| "log_odds_chosen": 2.0129177570343018, | |
| "log_odds_ratio": -0.1976321041584015, | |
| "logps/chosen": -0.43267732858657837, | |
| "logps/rejected": -1.4722133874893188, | |
| "loss": 10.7779, | |
| "nll_loss": 0.6637340784072876, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.02163386344909668, | |
| "rewards/margins": 0.05197680741548538, | |
| "rewards/rejected": -0.07361067086458206, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.118598382749326, | |
| "grad_norm": 33.5, | |
| "learning_rate": 6.543595027954785e-06, | |
| "log_odds_chosen": 2.077929735183716, | |
| "log_odds_ratio": -0.2039026916027069, | |
| "logps/chosen": -0.4455755650997162, | |
| "logps/rejected": -1.5429691076278687, | |
| "loss": 10.7447, | |
| "nll_loss": 0.6613457798957825, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -0.02227878011763096, | |
| "rewards/margins": 0.05486968904733658, | |
| "rewards/rejected": -0.07714846730232239, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.1725067385444743, | |
| "grad_norm": 26.375, | |
| "learning_rate": 5.764421709354811e-06, | |
| "log_odds_chosen": 2.2563636302948, | |
| "log_odds_ratio": -0.17267544567584991, | |
| "logps/chosen": -0.4047982692718506, | |
| "logps/rejected": -1.6122785806655884, | |
| "loss": 10.4634, | |
| "nll_loss": 0.6453262567520142, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -0.02023991569876671, | |
| "rewards/margins": 0.06037401407957077, | |
| "rewards/rejected": -0.08061392605304718, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.2264150943396226, | |
| "grad_norm": 22.25, | |
| "learning_rate": 5.02856031778888e-06, | |
| "log_odds_chosen": 2.167142152786255, | |
| "log_odds_ratio": -0.19436317682266235, | |
| "logps/chosen": -0.43167462944984436, | |
| "logps/rejected": -1.5412400960922241, | |
| "loss": 10.7259, | |
| "nll_loss": 0.6606518030166626, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -0.021583732217550278, | |
| "rewards/margins": 0.055478282272815704, | |
| "rewards/rejected": -0.07706201076507568, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.280323450134771, | |
| "grad_norm": 22.75, | |
| "learning_rate": 4.337667760807018e-06, | |
| "log_odds_chosen": 2.055614471435547, | |
| "log_odds_ratio": -0.20666727423667908, | |
| "logps/chosen": -0.4421593248844147, | |
| "logps/rejected": -1.5431147813796997, | |
| "loss": 10.7846, | |
| "nll_loss": 0.6637014746665955, | |
| "rewards/accuracies": 0.971875011920929, | |
| "rewards/chosen": -0.022107966244220734, | |
| "rewards/margins": 0.05504777282476425, | |
| "rewards/rejected": -0.07715573906898499, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.334231805929919, | |
| "grad_norm": 27.875, | |
| "learning_rate": 3.69329969156216e-06, | |
| "log_odds_chosen": 2.183558940887451, | |
| "log_odds_ratio": -0.17826569080352783, | |
| "logps/chosen": -0.43023914098739624, | |
| "logps/rejected": -1.6062309741973877, | |
| "loss": 10.7638, | |
| "nll_loss": 0.663822591304779, | |
| "rewards/accuracies": 0.971875011920929, | |
| "rewards/chosen": -0.021511957049369812, | |
| "rewards/margins": 0.058799587190151215, | |
| "rewards/rejected": -0.08031155169010162, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.3881401617250675, | |
| "grad_norm": 27.875, | |
| "learning_rate": 3.096907006012639e-06, | |
| "log_odds_chosen": 2.1112418174743652, | |
| "log_odds_ratio": -0.19758489727973938, | |
| "logps/chosen": -0.4420015215873718, | |
| "logps/rejected": -1.5825992822647095, | |
| "loss": 10.8219, | |
| "nll_loss": 0.6664899587631226, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -0.02210007607936859, | |
| "rewards/margins": 0.05702989175915718, | |
| "rewards/rejected": -0.07912997156381607, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 3.442048517520216, | |
| "grad_norm": 23.875, | |
| "learning_rate": 2.549832576001965e-06, | |
| "log_odds_chosen": 2.2864553928375244, | |
| "log_odds_ratio": -0.1727539300918579, | |
| "logps/chosen": -0.42124342918395996, | |
| "logps/rejected": -1.643252968788147, | |
| "loss": 10.4624, | |
| "nll_loss": 0.6452642679214478, | |
| "rewards/accuracies": 0.9906250238418579, | |
| "rewards/chosen": -0.021062172949314117, | |
| "rewards/margins": 0.061100464314222336, | |
| "rewards/rejected": -0.08216263353824615, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 3.4959568733153636, | |
| "grad_norm": 24.75, | |
| "learning_rate": 2.0533082255718594e-06, | |
| "log_odds_chosen": 2.240377187728882, | |
| "log_odds_ratio": -0.18875715136528015, | |
| "logps/chosen": -0.39375439286231995, | |
| "logps/rejected": -1.5621250867843628, | |
| "loss": 9.993, | |
| "nll_loss": 0.6151254177093506, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -0.019687717780470848, | |
| "rewards/margins": 0.058418530970811844, | |
| "rewards/rejected": -0.07810626178979874, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 3.5498652291105124, | |
| "grad_norm": 28.5, | |
| "learning_rate": 1.6084519573168727e-06, | |
| "log_odds_chosen": 2.1224582195281982, | |
| "log_odds_ratio": -0.2051163613796234, | |
| "logps/chosen": -0.4335021376609802, | |
| "logps/rejected": -1.5549887418746948, | |
| "loss": 10.6481, | |
| "nll_loss": 0.6552503705024719, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -0.0216751080006361, | |
| "rewards/margins": 0.05607432872056961, | |
| "rewards/rejected": -0.07774944603443146, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 3.6037735849056602, | |
| "grad_norm": 21.625, | |
| "learning_rate": 1.2162654350258402e-06, | |
| "log_odds_chosen": 2.2741246223449707, | |
| "log_odds_ratio": -0.18007881939411163, | |
| "logps/chosen": -0.438800573348999, | |
| "logps/rejected": -1.675838828086853, | |
| "loss": 10.4112, | |
| "nll_loss": 0.6416950821876526, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.02194002829492092, | |
| "rewards/margins": 0.06185191869735718, | |
| "rewards/rejected": -0.08379194885492325, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 3.6576819407008085, | |
| "grad_norm": 27.5, | |
| "learning_rate": 8.776317282785163e-07, | |
| "log_odds_chosen": 2.210236072540283, | |
| "log_odds_ratio": -0.1864241361618042, | |
| "logps/chosen": -0.4287256598472595, | |
| "logps/rejected": -1.5987118482589722, | |
| "loss": 10.3796, | |
| "nll_loss": 0.6394068598747253, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.021436283364892006, | |
| "rewards/margins": 0.05849931389093399, | |
| "rewards/rejected": -0.07993558794260025, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 3.711590296495957, | |
| "grad_norm": 25.875, | |
| "learning_rate": 5.933133240756117e-07, | |
| "log_odds_chosen": 2.208949327468872, | |
| "log_odds_ratio": -0.18415012955665588, | |
| "logps/chosen": -0.44286638498306274, | |
| "logps/rejected": -1.6374422311782837, | |
| "loss": 10.6099, | |
| "nll_loss": 0.6539115905761719, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.022143317386507988, | |
| "rewards/margins": 0.05972879007458687, | |
| "rewards/rejected": -0.0818721130490303, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 3.765498652291105, | |
| "grad_norm": 21.375, | |
| "learning_rate": 3.6395040997952714e-07, | |
| "log_odds_chosen": 2.108987331390381, | |
| "log_odds_ratio": -0.19528530538082123, | |
| "logps/chosen": -0.42990946769714355, | |
| "logps/rejected": -1.5616236925125122, | |
| "loss": 10.8233, | |
| "nll_loss": 0.6666921377182007, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -0.021495472639799118, | |
| "rewards/margins": 0.05658571049571037, | |
| "rewards/rejected": -0.0780811756849289, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 3.8194070080862534, | |
| "grad_norm": 24.25, | |
| "learning_rate": 1.9005943263146198e-07, | |
| "log_odds_chosen": 2.360919237136841, | |
| "log_odds_ratio": -0.1664571613073349, | |
| "logps/chosen": -0.3908630311489105, | |
| "logps/rejected": -1.650352120399475, | |
| "loss": 10.2428, | |
| "nll_loss": 0.6318550109863281, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -0.019543152302503586, | |
| "rewards/margins": 0.0629744604229927, | |
| "rewards/rejected": -0.082517609000206, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 3.8733153638814017, | |
| "grad_norm": 29.25, | |
| "learning_rate": 7.203193489064009e-08, | |
| "log_odds_chosen": 2.1713168621063232, | |
| "log_odds_ratio": -0.1880660057067871, | |
| "logps/chosen": -0.41322654485702515, | |
| "logps/rejected": -1.57550048828125, | |
| "loss": 10.3482, | |
| "nll_loss": 0.6373571157455444, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.020661327987909317, | |
| "rewards/margins": 0.058113694190979004, | |
| "rewards/rejected": -0.07877503335475922, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 3.92722371967655, | |
| "grad_norm": 24.75, | |
| "learning_rate": 1.0133674214043987e-08, | |
| "log_odds_chosen": 2.3266632556915283, | |
| "log_odds_ratio": -0.16368618607521057, | |
| "logps/chosen": -0.3924325108528137, | |
| "logps/rejected": -1.6150646209716797, | |
| "loss": 9.8053, | |
| "nll_loss": 0.6046454310417175, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.019621625542640686, | |
| "rewards/margins": 0.06113160774111748, | |
| "rewards/rejected": -0.08075322955846786, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 3.9595687331536387, | |
| "step": 368, | |
| "total_flos": 0.0, | |
| "train_loss": 25.71915608903636, | |
| "train_runtime": 4429.3261, | |
| "train_samples_per_second": 5.361, | |
| "train_steps_per_second": 0.083 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 368, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |