| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9997038791827065, | |
| "eval_steps": 500, | |
| "global_step": 1688, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.005922416345869114, | |
| "grad_norm": 25.375, | |
| "learning_rate": 2.9585798816568044e-08, | |
| "log_odds_chosen": -0.4997142255306244, | |
| "log_odds_ratio": -1.0621646642684937, | |
| "logits/chosen": -2.2295050621032715, | |
| "logits/rejected": -2.215860366821289, | |
| "logps/chosen": -0.7159513235092163, | |
| "logps/rejected": -0.47170203924179077, | |
| "loss": 1.2686, | |
| "nll_loss": 1.285839319229126, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.011844832691738229, | |
| "grad_norm": 25.25, | |
| "learning_rate": 5.917159763313609e-08, | |
| "log_odds_chosen": -0.6078722476959229, | |
| "log_odds_ratio": -1.1548207998275757, | |
| "logits/chosen": -2.1872293949127197, | |
| "logits/rejected": -2.1639022827148438, | |
| "logps/chosen": -0.8250460624694824, | |
| "logps/rejected": -0.4715689718723297, | |
| "loss": 1.2301, | |
| "nll_loss": 1.2283066511154175, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.017767249037607343, | |
| "grad_norm": 27.125, | |
| "learning_rate": 8.875739644970414e-08, | |
| "log_odds_chosen": -0.5964034199714661, | |
| "log_odds_ratio": -1.1720728874206543, | |
| "logits/chosen": -2.155057191848755, | |
| "logits/rejected": -2.146630048751831, | |
| "logps/chosen": -0.8543933033943176, | |
| "logps/rejected": -0.4923427104949951, | |
| "loss": 1.2398, | |
| "nll_loss": 1.313323736190796, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.023689665383476458, | |
| "grad_norm": 28.25, | |
| "learning_rate": 1.1834319526627217e-07, | |
| "log_odds_chosen": -0.5351605415344238, | |
| "log_odds_ratio": -1.0927046537399292, | |
| "logits/chosen": -2.2190463542938232, | |
| "logits/rejected": -2.206223964691162, | |
| "logps/chosen": -0.7575310468673706, | |
| "logps/rejected": -0.4635254740715027, | |
| "loss": 1.2735, | |
| "nll_loss": 1.2356092929840088, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.029612081729345572, | |
| "grad_norm": 26.0, | |
| "learning_rate": 1.4792899408284022e-07, | |
| "log_odds_chosen": -0.4550475478172302, | |
| "log_odds_ratio": -1.0262255668640137, | |
| "logits/chosen": -2.163825750350952, | |
| "logits/rejected": -2.148223400115967, | |
| "logps/chosen": -0.7005314826965332, | |
| "logps/rejected": -0.47106480598449707, | |
| "loss": 1.2103, | |
| "nll_loss": 1.2403192520141602, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.035534498075214686, | |
| "grad_norm": 21.375, | |
| "learning_rate": 1.7751479289940827e-07, | |
| "log_odds_chosen": -0.6598173975944519, | |
| "log_odds_ratio": -1.2315865755081177, | |
| "logits/chosen": -2.2192461490631104, | |
| "logits/rejected": -2.1879701614379883, | |
| "logps/chosen": -0.8897407650947571, | |
| "logps/rejected": -0.4609861969947815, | |
| "loss": 1.2416, | |
| "nll_loss": 1.2300336360931396, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.041456914421083804, | |
| "grad_norm": 27.375, | |
| "learning_rate": 2.0710059171597633e-07, | |
| "log_odds_chosen": -0.5370969772338867, | |
| "log_odds_ratio": -1.1101651191711426, | |
| "logits/chosen": -2.233755588531494, | |
| "logits/rejected": -2.201343297958374, | |
| "logps/chosen": -0.7967244386672974, | |
| "logps/rejected": -0.4630069136619568, | |
| "loss": 1.2546, | |
| "nll_loss": 1.2548679113388062, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.047379330766952915, | |
| "grad_norm": 26.625, | |
| "learning_rate": 2.3668639053254435e-07, | |
| "log_odds_chosen": -0.5750253796577454, | |
| "log_odds_ratio": -1.1556330919265747, | |
| "logits/chosen": -2.19846773147583, | |
| "logits/rejected": -2.187711715698242, | |
| "logps/chosen": -0.7946293950080872, | |
| "logps/rejected": -0.4594718813896179, | |
| "loss": 1.2238, | |
| "nll_loss": 1.2166999578475952, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.05330174711282203, | |
| "grad_norm": 22.375, | |
| "learning_rate": 2.662721893491124e-07, | |
| "log_odds_chosen": -0.502492368221283, | |
| "log_odds_ratio": -1.0737704038619995, | |
| "logits/chosen": -2.18656063079834, | |
| "logits/rejected": -2.1636054515838623, | |
| "logps/chosen": -0.7198958992958069, | |
| "logps/rejected": -0.4653542935848236, | |
| "loss": 1.1987, | |
| "nll_loss": 1.2203375101089478, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.059224163458691144, | |
| "grad_norm": 22.375, | |
| "learning_rate": 2.9585798816568045e-07, | |
| "log_odds_chosen": -0.41660839319229126, | |
| "log_odds_ratio": -0.9962056279182434, | |
| "logits/chosen": -2.247572422027588, | |
| "logits/rejected": -2.2023332118988037, | |
| "logps/chosen": -0.6892199516296387, | |
| "logps/rejected": -0.4913715422153473, | |
| "loss": 1.1737, | |
| "nll_loss": 1.2142550945281982, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.06514657980456026, | |
| "grad_norm": 17.25, | |
| "learning_rate": 3.254437869822485e-07, | |
| "log_odds_chosen": -0.4817970395088196, | |
| "log_odds_ratio": -1.0484726428985596, | |
| "logits/chosen": -2.1959776878356934, | |
| "logits/rejected": -2.172440767288208, | |
| "logps/chosen": -0.7387205958366394, | |
| "logps/rejected": -0.47952842712402344, | |
| "loss": 1.1196, | |
| "nll_loss": 1.1109485626220703, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.07106899615042937, | |
| "grad_norm": 17.75, | |
| "learning_rate": 3.5502958579881655e-07, | |
| "log_odds_chosen": -0.5072614550590515, | |
| "log_odds_ratio": -1.073188066482544, | |
| "logits/chosen": -2.2234084606170654, | |
| "logits/rejected": -2.212110996246338, | |
| "logps/chosen": -0.7518635988235474, | |
| "logps/rejected": -0.4725222587585449, | |
| "loss": 1.1538, | |
| "nll_loss": 1.1456319093704224, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.07699141249629848, | |
| "grad_norm": 22.875, | |
| "learning_rate": 3.8461538461538463e-07, | |
| "log_odds_chosen": -0.6446342468261719, | |
| "log_odds_ratio": -1.2178680896759033, | |
| "logits/chosen": -2.2080233097076416, | |
| "logits/rejected": -2.1998672485351562, | |
| "logps/chosen": -0.8730036020278931, | |
| "logps/rejected": -0.4481457769870758, | |
| "loss": 1.1644, | |
| "nll_loss": 1.1509124040603638, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.08291382884216761, | |
| "grad_norm": 16.25, | |
| "learning_rate": 4.1420118343195265e-07, | |
| "log_odds_chosen": -0.45662721991539, | |
| "log_odds_ratio": -1.0147430896759033, | |
| "logits/chosen": -2.237990617752075, | |
| "logits/rejected": -2.2128589153289795, | |
| "logps/chosen": -0.6646671295166016, | |
| "logps/rejected": -0.44373393058776855, | |
| "loss": 1.0906, | |
| "nll_loss": 1.0673267841339111, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.08883624518803672, | |
| "grad_norm": 26.5, | |
| "learning_rate": 4.437869822485207e-07, | |
| "log_odds_chosen": -0.46678367257118225, | |
| "log_odds_ratio": -1.0147194862365723, | |
| "logits/chosen": -2.167670488357544, | |
| "logits/rejected": -2.1592793464660645, | |
| "logps/chosen": -0.7025789618492126, | |
| "logps/rejected": -0.47203493118286133, | |
| "loss": 1.1008, | |
| "nll_loss": 1.1650502681732178, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.09475866153390583, | |
| "grad_norm": 12.0625, | |
| "learning_rate": 4.733727810650887e-07, | |
| "log_odds_chosen": -0.3221941888332367, | |
| "log_odds_ratio": -0.9352226257324219, | |
| "logits/chosen": -2.247824192047119, | |
| "logits/rejected": -2.2287344932556152, | |
| "logps/chosen": -0.6016725301742554, | |
| "logps/rejected": -0.4531864523887634, | |
| "loss": 1.0947, | |
| "nll_loss": 1.0781590938568115, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.10068107787977496, | |
| "grad_norm": 11.5625, | |
| "learning_rate": 4.999994653198566e-07, | |
| "log_odds_chosen": -0.4564700722694397, | |
| "log_odds_ratio": -1.0602452754974365, | |
| "logits/chosen": -2.2789835929870605, | |
| "logits/rejected": -2.2523741722106934, | |
| "logps/chosen": -0.7461049556732178, | |
| "logps/rejected": -0.48730534315109253, | |
| "loss": 1.1159, | |
| "nll_loss": 1.0658115148544312, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.10660349422564407, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 4.999353064699471e-07, | |
| "log_odds_chosen": -0.5452951192855835, | |
| "log_odds_ratio": -1.1454532146453857, | |
| "logits/chosen": -2.237121820449829, | |
| "logits/rejected": -2.202718496322632, | |
| "logps/chosen": -0.827674388885498, | |
| "logps/rejected": -0.49005183577537537, | |
| "loss": 0.9748, | |
| "nll_loss": 1.0014435052871704, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.11252591057151318, | |
| "grad_norm": 10.375, | |
| "learning_rate": 4.99764243036258e-07, | |
| "log_odds_chosen": -0.4207037091255188, | |
| "log_odds_ratio": -0.9974331855773926, | |
| "logits/chosen": -2.27175235748291, | |
| "logits/rejected": -2.242116689682007, | |
| "logps/chosen": -0.6407202482223511, | |
| "logps/rejected": -0.4429788589477539, | |
| "loss": 1.0095, | |
| "nll_loss": 1.017865777015686, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.11844832691738229, | |
| "grad_norm": 9.5, | |
| "learning_rate": 4.994863481875841e-07, | |
| "log_odds_chosen": -0.4031923711299896, | |
| "log_odds_ratio": -0.973800003528595, | |
| "logits/chosen": -2.221717119216919, | |
| "logits/rejected": -2.18719482421875, | |
| "logps/chosen": -0.6306296586990356, | |
| "logps/rejected": -0.43233147263526917, | |
| "loss": 1.0045, | |
| "nll_loss": 0.9697571992874146, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.12437074326325141, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 4.991017407876165e-07, | |
| "log_odds_chosen": -0.4411424994468689, | |
| "log_odds_ratio": -1.0120642185211182, | |
| "logits/chosen": -2.238583564758301, | |
| "logits/rejected": -2.1919620037078857, | |
| "logps/chosen": -0.7006498575210571, | |
| "logps/rejected": -0.4852658808231354, | |
| "loss": 0.9832, | |
| "nll_loss": 1.0057976245880127, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.13029315960912052, | |
| "grad_norm": 9.0, | |
| "learning_rate": 4.98610585344102e-07, | |
| "log_odds_chosen": -0.25588923692703247, | |
| "log_odds_ratio": -0.9158498048782349, | |
| "logits/chosen": -2.258283853530884, | |
| "logits/rejected": -2.2223126888275146, | |
| "logps/chosen": -0.5977104306221008, | |
| "logps/rejected": -0.4761990010738373, | |
| "loss": 1.02, | |
| "nll_loss": 1.0466753244400024, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.13621557595498965, | |
| "grad_norm": 8.4375, | |
| "learning_rate": 4.980130919384768e-07, | |
| "log_odds_chosen": -0.5824810266494751, | |
| "log_odds_ratio": -1.1220190525054932, | |
| "logits/chosen": -2.2531580924987793, | |
| "logits/rejected": -2.2409615516662598, | |
| "logps/chosen": -0.7504315972328186, | |
| "logps/rejected": -0.43684881925582886, | |
| "loss": 1.0183, | |
| "nll_loss": 1.0061722993850708, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.14213799230085875, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 4.973095161360105e-07, | |
| "log_odds_chosen": -0.44555410742759705, | |
| "log_odds_ratio": -1.0208032131195068, | |
| "logits/chosen": -2.2470836639404297, | |
| "logits/rejected": -2.214434862136841, | |
| "logps/chosen": -0.6731461882591248, | |
| "logps/rejected": -0.4670758843421936, | |
| "loss": 1.0354, | |
| "nll_loss": 1.0512316226959229, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.14806040864672787, | |
| "grad_norm": 8.875, | |
| "learning_rate": 4.965001588764913e-07, | |
| "log_odds_chosen": -0.4621347486972809, | |
| "log_odds_ratio": -1.0333962440490723, | |
| "logits/chosen": -2.274649143218994, | |
| "logits/rejected": -2.241596221923828, | |
| "logps/chosen": -0.6809024214744568, | |
| "logps/rejected": -0.4347941279411316, | |
| "loss": 1.0076, | |
| "nll_loss": 1.007010817527771, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.15398282499259697, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 4.955853663455072e-07, | |
| "log_odds_chosen": -0.3350891172885895, | |
| "log_odds_ratio": -0.9613872766494751, | |
| "logits/chosen": -2.260413885116577, | |
| "logits/rejected": -2.2278614044189453, | |
| "logps/chosen": -0.6426165699958801, | |
| "logps/rejected": -0.45530933141708374, | |
| "loss": 0.9607, | |
| "nll_loss": 0.9523956179618835, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.1599052413384661, | |
| "grad_norm": 9.125, | |
| "learning_rate": 4.945655298263713e-07, | |
| "log_odds_chosen": -0.4467865824699402, | |
| "log_odds_ratio": -1.0078147649765015, | |
| "logits/chosen": -2.2099037170410156, | |
| "logits/rejected": -2.183701992034912, | |
| "logps/chosen": -0.6576748490333557, | |
| "logps/rejected": -0.4429934620857239, | |
| "loss": 1.0429, | |
| "nll_loss": 1.075627326965332, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.16582765768433522, | |
| "grad_norm": 7.96875, | |
| "learning_rate": 4.934410855327585e-07, | |
| "log_odds_chosen": -0.38402479887008667, | |
| "log_odds_ratio": -0.9679163098335266, | |
| "logits/chosen": -2.292367458343506, | |
| "logits/rejected": -2.2721431255340576, | |
| "logps/chosen": -0.6379308104515076, | |
| "logps/rejected": -0.442401647567749, | |
| "loss": 0.9621, | |
| "nll_loss": 1.016234278678894, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.1717500740302043, | |
| "grad_norm": 9.0, | |
| "learning_rate": 4.922125144221252e-07, | |
| "log_odds_chosen": -0.4171718955039978, | |
| "log_odds_ratio": -0.9991844296455383, | |
| "logits/chosen": -2.259284257888794, | |
| "logits/rejected": -2.205514430999756, | |
| "logps/chosen": -0.6329622268676758, | |
| "logps/rejected": -0.43889325857162476, | |
| "loss": 1.014, | |
| "nll_loss": 1.0359452962875366, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.17767249037607344, | |
| "grad_norm": 8.125, | |
| "learning_rate": 4.90880341989989e-07, | |
| "log_odds_chosen": -0.33935636281967163, | |
| "log_odds_ratio": -0.9420417547225952, | |
| "logits/chosen": -2.2610156536102295, | |
| "logits/rejected": -2.2359061241149902, | |
| "logps/chosen": -0.6213563084602356, | |
| "logps/rejected": -0.44430437684059143, | |
| "loss": 0.9833, | |
| "nll_loss": 0.9867600202560425, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.18359490672194256, | |
| "grad_norm": 8.375, | |
| "learning_rate": 4.894451380451589e-07, | |
| "log_odds_chosen": -0.5468162298202515, | |
| "log_odds_ratio": -1.0870132446289062, | |
| "logits/chosen": -2.241508722305298, | |
| "logits/rejected": -2.22690749168396, | |
| "logps/chosen": -0.7115592360496521, | |
| "logps/rejected": -0.43017569184303284, | |
| "loss": 1.0006, | |
| "nll_loss": 0.994620680809021, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.18951732306781166, | |
| "grad_norm": 9.25, | |
| "learning_rate": 4.879075164660124e-07, | |
| "log_odds_chosen": -0.3401740491390228, | |
| "log_odds_ratio": -0.9383065104484558, | |
| "logits/chosen": -2.2438132762908936, | |
| "logits/rejected": -2.209188938140869, | |
| "logps/chosen": -0.5985551476478577, | |
| "logps/rejected": -0.43559733033180237, | |
| "loss": 0.94, | |
| "nll_loss": 0.9133344888687134, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.19543973941368079, | |
| "grad_norm": 10.5, | |
| "learning_rate": 4.862681349379212e-07, | |
| "log_odds_chosen": -0.3794914484024048, | |
| "log_odds_ratio": -0.9728193283081055, | |
| "logits/chosen": -2.2533066272735596, | |
| "logits/rejected": -2.1980607509613037, | |
| "logps/chosen": -0.6138342022895813, | |
| "logps/rejected": -0.44097796082496643, | |
| "loss": 1.0041, | |
| "nll_loss": 1.0256803035736084, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.2013621557595499, | |
| "grad_norm": 7.6875, | |
| "learning_rate": 4.8452769467194e-07, | |
| "log_odds_chosen": -0.40433868765830994, | |
| "log_odds_ratio": -0.9825445413589478, | |
| "logits/chosen": -2.2585511207580566, | |
| "logits/rejected": -2.233630657196045, | |
| "logps/chosen": -0.6160660982131958, | |
| "logps/rejected": -0.4248103201389313, | |
| "loss": 0.9778, | |
| "nll_loss": 0.9514611959457397, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.207284572105419, | |
| "grad_norm": 7.625, | |
| "learning_rate": 4.82686940104879e-07, | |
| "log_odds_chosen": -0.4215853214263916, | |
| "log_odds_ratio": -1.01924729347229, | |
| "logits/chosen": -2.30430269241333, | |
| "logits/rejected": -2.272357702255249, | |
| "logps/chosen": -0.645369291305542, | |
| "logps/rejected": -0.4133967459201813, | |
| "loss": 0.9287, | |
| "nll_loss": 0.9160087704658508, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.21320698845128813, | |
| "grad_norm": 8.125, | |
| "learning_rate": 4.807466585808856e-07, | |
| "log_odds_chosen": -0.3686332702636719, | |
| "log_odds_ratio": -0.9627587199211121, | |
| "logits/chosen": -2.282811403274536, | |
| "logits/rejected": -2.2714035511016846, | |
| "logps/chosen": -0.5806415677070618, | |
| "logps/rejected": -0.4163896143436432, | |
| "loss": 0.987, | |
| "nll_loss": 0.9767228960990906, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.21912940479715723, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 4.787076800146752e-07, | |
| "log_odds_chosen": -0.34714585542678833, | |
| "log_odds_ratio": -0.9853572845458984, | |
| "logits/chosen": -2.2601521015167236, | |
| "logits/rejected": -2.2084286212921143, | |
| "logps/chosen": -0.6458638906478882, | |
| "logps/rejected": -0.4349249005317688, | |
| "loss": 0.908, | |
| "nll_loss": 0.8895160555839539, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.22505182114302635, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 4.765708765365526e-07, | |
| "log_odds_chosen": -0.30534738302230835, | |
| "log_odds_ratio": -0.9373781085014343, | |
| "logits/chosen": -2.2653586864471436, | |
| "logits/rejected": -2.254210948944092, | |
| "logps/chosen": -0.576322615146637, | |
| "logps/rejected": -0.4294815957546234, | |
| "loss": 1.0005, | |
| "nll_loss": 0.958886981010437, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.23097423748889548, | |
| "grad_norm": 9.875, | |
| "learning_rate": 4.7433716211937587e-07, | |
| "log_odds_chosen": -0.5105515122413635, | |
| "log_odds_ratio": -1.0566070079803467, | |
| "logits/chosen": -2.328101396560669, | |
| "logits/rejected": -2.302281141281128, | |
| "logps/chosen": -0.6501199007034302, | |
| "logps/rejected": -0.404310941696167, | |
| "loss": 0.9396, | |
| "nll_loss": 0.9967532157897949, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.23689665383476458, | |
| "grad_norm": 7.8125, | |
| "learning_rate": 4.720074921876245e-07, | |
| "log_odds_chosen": -0.45067232847213745, | |
| "log_odds_ratio": -1.0197547674179077, | |
| "logits/chosen": -2.340407133102417, | |
| "logits/rejected": -2.293402910232544, | |
| "logps/chosen": -0.6130900382995605, | |
| "logps/rejected": -0.41540417075157166, | |
| "loss": 0.9442, | |
| "nll_loss": 0.9423254132270813, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.2428190701806337, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 4.6958286320873593e-07, | |
| "log_odds_chosen": -0.43627676367759705, | |
| "log_odds_ratio": -0.9863921403884888, | |
| "logits/chosen": -2.2813560962677, | |
| "logits/rejected": -2.275886058807373, | |
| "logps/chosen": -0.6022886633872986, | |
| "logps/rejected": -0.4021386504173279, | |
| "loss": 0.9658, | |
| "nll_loss": 0.9948114156723022, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.24874148652650283, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 4.6706431226688804e-07, | |
| "log_odds_chosen": -0.3637348413467407, | |
| "log_odds_ratio": -0.9635465741157532, | |
| "logits/chosen": -2.2663254737854004, | |
| "logits/rejected": -2.2325570583343506, | |
| "logps/chosen": -0.6079740524291992, | |
| "logps/rejected": -0.42877498269081116, | |
| "loss": 0.971, | |
| "nll_loss": 0.9684462547302246, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.25466390287237195, | |
| "grad_norm": 7.71875, | |
| "learning_rate": 4.6445291661940777e-07, | |
| "log_odds_chosen": -0.29998743534088135, | |
| "log_odds_ratio": -0.9151178598403931, | |
| "logits/chosen": -2.288652181625366, | |
| "logits/rejected": -2.28438138961792, | |
| "logps/chosen": -0.5727067589759827, | |
| "logps/rejected": -0.43537649512290955, | |
| "loss": 0.9344, | |
| "nll_loss": 0.8895971179008484, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.26058631921824105, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 4.6174979323599715e-07, | |
| "log_odds_chosen": -0.5159381031990051, | |
| "log_odds_ratio": -1.0749253034591675, | |
| "logits/chosen": -2.2701315879821777, | |
| "logits/rejected": -2.2190845012664795, | |
| "logps/chosen": -0.7043232321739197, | |
| "logps/rejected": -0.4317665696144104, | |
| "loss": 0.9929, | |
| "nll_loss": 1.0871878862380981, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.26650873556411014, | |
| "grad_norm": 7.9375, | |
| "learning_rate": 4.5895609832097277e-07, | |
| "log_odds_chosen": -0.38775309920310974, | |
| "log_odds_ratio": -1.0040466785430908, | |
| "logits/chosen": -2.2794992923736572, | |
| "logits/rejected": -2.2638792991638184, | |
| "logps/chosen": -0.6565039157867432, | |
| "logps/rejected": -0.43878334760665894, | |
| "loss": 0.9716, | |
| "nll_loss": 0.9555328488349915, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.2724311519099793, | |
| "grad_norm": 8.25, | |
| "learning_rate": 4.560730268187236e-07, | |
| "log_odds_chosen": -0.3349025249481201, | |
| "log_odds_ratio": -0.9378219842910767, | |
| "logits/chosen": -2.282761812210083, | |
| "logits/rejected": -2.244011878967285, | |
| "logps/chosen": -0.5650533437728882, | |
| "logps/rejected": -0.4169080853462219, | |
| "loss": 0.9547, | |
| "nll_loss": 0.9367356300354004, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.2783535682558484, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 4.531018119025989e-07, | |
| "log_odds_chosen": -0.24693968892097473, | |
| "log_odds_ratio": -0.9230139851570129, | |
| "logits/chosen": -2.338200807571411, | |
| "logits/rejected": -2.3114407062530518, | |
| "logps/chosen": -0.5866008996963501, | |
| "logps/rejected": -0.498542845249176, | |
| "loss": 0.9863, | |
| "nll_loss": 1.0312178134918213, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.2842759846017175, | |
| "grad_norm": 7.59375, | |
| "learning_rate": 4.5004372444744376e-07, | |
| "log_odds_chosen": -0.259705126285553, | |
| "log_odds_ratio": -0.9033578634262085, | |
| "logits/chosen": -2.281229257583618, | |
| "logits/rejected": -2.259384870529175, | |
| "logps/chosen": -0.6026913523674011, | |
| "logps/rejected": -0.46954187750816345, | |
| "loss": 0.9559, | |
| "nll_loss": 0.9717810750007629, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.2901984009475866, | |
| "grad_norm": 9.0, | |
| "learning_rate": 4.4690007248600967e-07, | |
| "log_odds_chosen": -0.3773840069770813, | |
| "log_odds_ratio": -0.9825248718261719, | |
| "logits/chosen": -2.2721426486968994, | |
| "logits/rejected": -2.2558834552764893, | |
| "logps/chosen": -0.629915714263916, | |
| "logps/rejected": -0.43304410576820374, | |
| "loss": 0.954, | |
| "nll_loss": 0.9644275903701782, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.29612081729345574, | |
| "grad_norm": 7.71875, | |
| "learning_rate": 4.436722006494701e-07, | |
| "log_odds_chosen": -0.5259193778038025, | |
| "log_odds_ratio": -1.1190059185028076, | |
| "logits/chosen": -2.266916275024414, | |
| "logits/rejected": -2.243081569671631, | |
| "logps/chosen": -0.7579408884048462, | |
| "logps/rejected": -0.4302619397640228, | |
| "loss": 0.9695, | |
| "nll_loss": 0.9956067204475403, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.30204323363932484, | |
| "grad_norm": 8.5, | |
| "learning_rate": 4.4036148959228356e-07, | |
| "log_odds_chosen": -0.4430968165397644, | |
| "log_odds_ratio": -1.0375418663024902, | |
| "logits/chosen": -2.300400733947754, | |
| "logits/rejected": -2.2604432106018066, | |
| "logps/chosen": -0.6584800481796265, | |
| "logps/rejected": -0.4138873517513275, | |
| "loss": 0.9756, | |
| "nll_loss": 0.9368442296981812, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.30796564998519393, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 4.3696935540164705e-07, | |
| "log_odds_chosen": -0.3859555423259735, | |
| "log_odds_ratio": -0.9752845764160156, | |
| "logits/chosen": -2.2633957862854004, | |
| "logits/rejected": -2.2417874336242676, | |
| "logps/chosen": -0.6037057638168335, | |
| "logps/rejected": -0.41955527663230896, | |
| "loss": 0.9235, | |
| "nll_loss": 0.9441665410995483, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.3138880663310631, | |
| "grad_norm": 7.71875, | |
| "learning_rate": 4.334972489917947e-07, | |
| "log_odds_chosen": -0.29654431343078613, | |
| "log_odds_ratio": -0.9258224368095398, | |
| "logits/chosen": -2.3264002799987793, | |
| "logits/rejected": -2.269259214401245, | |
| "logps/chosen": -0.5935055017471313, | |
| "logps/rejected": -0.4429333806037903, | |
| "loss": 0.9497, | |
| "nll_loss": 0.9263819456100464, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.3198104826769322, | |
| "grad_norm": 9.0, | |
| "learning_rate": 4.299466554833997e-07, | |
| "log_odds_chosen": -0.400839239358902, | |
| "log_odds_ratio": -0.9843107461929321, | |
| "logits/chosen": -2.30580472946167, | |
| "logits/rejected": -2.256434440612793, | |
| "logps/chosen": -0.5819273591041565, | |
| "logps/rejected": -0.408183753490448, | |
| "loss": 0.9515, | |
| "nll_loss": 0.92247474193573, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.3257328990228013, | |
| "grad_norm": 7.46875, | |
| "learning_rate": 4.263190935683449e-07, | |
| "log_odds_chosen": -0.32894009351730347, | |
| "log_odds_ratio": -0.9418984651565552, | |
| "logits/chosen": -2.282500743865967, | |
| "logits/rejected": -2.24668025970459, | |
| "logps/chosen": -0.5584912896156311, | |
| "logps/rejected": -0.4048989713191986, | |
| "loss": 0.8853, | |
| "nll_loss": 0.8602296113967896, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.33165531536867043, | |
| "grad_norm": 9.0, | |
| "learning_rate": 4.2261611486013437e-07, | |
| "log_odds_chosen": -0.39398467540740967, | |
| "log_odds_ratio": -0.9864169955253601, | |
| "logits/chosen": -2.3277463912963867, | |
| "logits/rejected": -2.2908778190612793, | |
| "logps/chosen": -0.618613064289093, | |
| "logps/rejected": -0.4362561106681824, | |
| "loss": 0.961, | |
| "nll_loss": 0.9670404195785522, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.33757773171453953, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 4.188393032302233e-07, | |
| "log_odds_chosen": -0.2161109894514084, | |
| "log_odds_ratio": -0.8888469934463501, | |
| "logits/chosen": -2.266890048980713, | |
| "logits/rejected": -2.2078969478607178, | |
| "logps/chosen": -0.5593982934951782, | |
| "logps/rejected": -0.48008909821510315, | |
| "loss": 0.9285, | |
| "nll_loss": 0.9204473495483398, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.3435001480604086, | |
| "grad_norm": 10.75, | |
| "learning_rate": 4.1499027413055e-07, | |
| "log_odds_chosen": -0.41526442766189575, | |
| "log_odds_ratio": -0.9975423812866211, | |
| "logits/chosen": -2.2734172344207764, | |
| "logits/rejected": -2.2457797527313232, | |
| "logps/chosen": -0.6249933838844299, | |
| "logps/rejected": -0.42492228746414185, | |
| "loss": 0.9404, | |
| "nll_loss": 0.9193958044052124, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.3494225644062778, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 4.1107067390256056e-07, | |
| "log_odds_chosen": -0.45963993668556213, | |
| "log_odds_ratio": -1.0648995637893677, | |
| "logits/chosen": -2.3240678310394287, | |
| "logits/rejected": -2.2981557846069336, | |
| "logps/chosen": -0.723495364189148, | |
| "logps/rejected": -0.454792320728302, | |
| "loss": 0.9656, | |
| "nll_loss": 1.0240063667297363, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.3553449807521469, | |
| "grad_norm": 9.875, | |
| "learning_rate": 4.0708217907302047e-07, | |
| "log_odds_chosen": -0.4009949564933777, | |
| "log_odds_ratio": -0.9855114221572876, | |
| "logits/chosen": -2.2710177898406982, | |
| "logits/rejected": -2.237403392791748, | |
| "logps/chosen": -0.6120108366012573, | |
| "logps/rejected": -0.43240681290626526, | |
| "loss": 0.9575, | |
| "nll_loss": 0.9712766408920288, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.361267397098016, | |
| "grad_norm": 12.0, | |
| "learning_rate": 4.030264956369157e-07, | |
| "log_odds_chosen": -0.39438915252685547, | |
| "log_odds_ratio": -0.9749253988265991, | |
| "logits/chosen": -2.31217098236084, | |
| "logits/rejected": -2.273338794708252, | |
| "logps/chosen": -0.5822636485099792, | |
| "logps/rejected": -0.4064372181892395, | |
| "loss": 0.9477, | |
| "nll_loss": 0.9778239130973816, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.3671898134438851, | |
| "grad_norm": 9.375, | |
| "learning_rate": 3.989053583277492e-07, | |
| "log_odds_chosen": -0.5915114879608154, | |
| "log_odds_ratio": -1.1409562826156616, | |
| "logits/chosen": -2.3212368488311768, | |
| "logits/rejected": -2.3024001121520996, | |
| "logps/chosen": -0.7573744654655457, | |
| "logps/rejected": -0.4205297827720642, | |
| "loss": 0.9491, | |
| "nll_loss": 0.9616823196411133, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.3731122297897542, | |
| "grad_norm": 8.125, | |
| "learning_rate": 3.947205298755447e-07, | |
| "log_odds_chosen": -0.32023632526397705, | |
| "log_odds_ratio": -0.9460951685905457, | |
| "logits/chosen": -2.2791507244110107, | |
| "logits/rejected": -2.2480525970458984, | |
| "logps/chosen": -0.6137298345565796, | |
| "logps/rejected": -0.4523869454860687, | |
| "loss": 0.9577, | |
| "nll_loss": 0.9420009851455688, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.3790346461356233, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 3.9047380025287634e-07, | |
| "log_odds_chosen": -0.31926944851875305, | |
| "log_odds_ratio": -0.9340398907661438, | |
| "logits/chosen": -2.288464069366455, | |
| "logits/rejected": -2.257875442504883, | |
| "logps/chosen": -0.5796951055526733, | |
| "logps/rejected": -0.43441399931907654, | |
| "loss": 0.9511, | |
| "nll_loss": 0.935884952545166, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.3849570624814925, | |
| "grad_norm": 10.125, | |
| "learning_rate": 3.8616698590924523e-07, | |
| "log_odds_chosen": -0.3541373610496521, | |
| "log_odds_ratio": -0.9547072649002075, | |
| "logits/chosen": -2.3075475692749023, | |
| "logits/rejected": -2.261488437652588, | |
| "logps/chosen": -0.6222845315933228, | |
| "logps/rejected": -0.4457763135433197, | |
| "loss": 0.9292, | |
| "nll_loss": 0.941017746925354, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.39087947882736157, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 3.8180192899413123e-07, | |
| "log_odds_chosen": -0.373871773481369, | |
| "log_odds_ratio": -0.963890552520752, | |
| "logits/chosen": -2.3060686588287354, | |
| "logits/rejected": -2.2961385250091553, | |
| "logps/chosen": -0.5887154936790466, | |
| "logps/rejected": -0.41268324851989746, | |
| "loss": 0.9644, | |
| "nll_loss": 0.9328317642211914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.39680189517323067, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 3.7738049656905225e-07, | |
| "log_odds_chosen": -0.3005954623222351, | |
| "log_odds_ratio": -0.9146180152893066, | |
| "logits/chosen": -2.241210460662842, | |
| "logits/rejected": -2.197197437286377, | |
| "logps/chosen": -0.5695523023605347, | |
| "logps/rejected": -0.4374919533729553, | |
| "loss": 0.9635, | |
| "nll_loss": 0.9454113841056824, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.4027243115190998, | |
| "grad_norm": 8.25, | |
| "learning_rate": 3.7290457980896787e-07, | |
| "log_odds_chosen": -0.2508184611797333, | |
| "log_odds_ratio": -0.8950401544570923, | |
| "logits/chosen": -2.310917377471924, | |
| "logits/rejected": -2.2810654640197754, | |
| "logps/chosen": -0.5575405955314636, | |
| "logps/rejected": -0.44296175241470337, | |
| "loss": 0.9245, | |
| "nll_loss": 0.9060578346252441, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.4086467278649689, | |
| "grad_norm": 8.375, | |
| "learning_rate": 3.68376093193369e-07, | |
| "log_odds_chosen": -0.35061341524124146, | |
| "log_odds_ratio": -0.9449998140335083, | |
| "logits/chosen": -2.3210480213165283, | |
| "logits/rejected": -2.281230926513672, | |
| "logps/chosen": -0.5540003776550293, | |
| "logps/rejected": -0.4034114480018616, | |
| "loss": 0.9207, | |
| "nll_loss": 0.9037810564041138, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.414569144210838, | |
| "grad_norm": 8.4375, | |
| "learning_rate": 3.637969736873992e-07, | |
| "log_odds_chosen": -0.29555535316467285, | |
| "log_odds_ratio": -0.9367197155952454, | |
| "logits/chosen": -2.2944698333740234, | |
| "logits/rejected": -2.2611544132232666, | |
| "logps/chosen": -0.5592113733291626, | |
| "logps/rejected": -0.42597031593322754, | |
| "loss": 0.9637, | |
| "nll_loss": 0.9748933911323547, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.4204915605567071, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 3.591691799133587e-07, | |
| "log_odds_chosen": -0.27811819314956665, | |
| "log_odds_ratio": -0.8972823023796082, | |
| "logits/chosen": -2.3404221534729004, | |
| "logits/rejected": -2.3104233741760254, | |
| "logps/chosen": -0.5548882484436035, | |
| "logps/rejected": -0.42241740226745605, | |
| "loss": 0.9489, | |
| "nll_loss": 0.9462203979492188, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.42641397690257626, | |
| "grad_norm": 7.8125, | |
| "learning_rate": 3.5449469131294476e-07, | |
| "log_odds_chosen": -0.282146155834198, | |
| "log_odds_ratio": -0.9153865575790405, | |
| "logits/chosen": -2.3050596714019775, | |
| "logits/rejected": -2.2582859992980957, | |
| "logps/chosen": -0.5491407513618469, | |
| "logps/rejected": -0.4175952970981598, | |
| "loss": 0.9258, | |
| "nll_loss": 0.9185633659362793, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.43233639324844536, | |
| "grad_norm": 7.78125, | |
| "learning_rate": 3.497755073005868e-07, | |
| "log_odds_chosen": -0.17704807221889496, | |
| "log_odds_ratio": -0.8550702333450317, | |
| "logits/chosen": -2.304471254348755, | |
| "logits/rejected": -2.2704811096191406, | |
| "logps/chosen": -0.5407411456108093, | |
| "logps/rejected": -0.43452388048171997, | |
| "loss": 0.9148, | |
| "nll_loss": 0.8776341676712036, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.43825880959431446, | |
| "grad_norm": 14.0, | |
| "learning_rate": 3.4501364640823926e-07, | |
| "log_odds_chosen": -0.4160383343696594, | |
| "log_odds_ratio": -0.9982725381851196, | |
| "logits/chosen": -2.3177871704101562, | |
| "logits/rejected": -2.291195869445801, | |
| "logps/chosen": -0.6620553135871887, | |
| "logps/rejected": -0.4509620666503906, | |
| "loss": 0.9449, | |
| "nll_loss": 0.9611420631408691, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.4441812259401836, | |
| "grad_norm": 8.4375, | |
| "learning_rate": 3.402111454219966e-07, | |
| "log_odds_chosen": -0.2541792690753937, | |
| "log_odds_ratio": -0.8975493311882019, | |
| "logits/chosen": -2.3212180137634277, | |
| "logits/rejected": -2.2709405422210693, | |
| "logps/chosen": -0.5664907693862915, | |
| "logps/rejected": -0.4353105127811432, | |
| "loss": 0.9301, | |
| "nll_loss": 0.9432824850082397, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.4501036422860527, | |
| "grad_norm": 7.28125, | |
| "learning_rate": 3.353700585109005e-07, | |
| "log_odds_chosen": -0.2790587842464447, | |
| "log_odds_ratio": -0.9118951559066772, | |
| "logits/chosen": -2.3148138523101807, | |
| "logits/rejected": -2.2849326133728027, | |
| "logps/chosen": -0.5668213963508606, | |
| "logps/rejected": -0.4337525963783264, | |
| "loss": 0.9239, | |
| "nll_loss": 0.9522818326950073, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.4560260586319218, | |
| "grad_norm": 7.90625, | |
| "learning_rate": 3.304924563483129e-07, | |
| "log_odds_chosen": -0.31332454085350037, | |
| "log_odds_ratio": -0.9554667472839355, | |
| "logits/chosen": -2.329709053039551, | |
| "logits/rejected": -2.3159825801849365, | |
| "logps/chosen": -0.6328242421150208, | |
| "logps/rejected": -0.4525510370731354, | |
| "loss": 0.9725, | |
| "nll_loss": 0.9982641935348511, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.46194847497779096, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 3.255804252262283e-07, | |
| "log_odds_chosen": -0.26954448223114014, | |
| "log_odds_ratio": -0.9041155576705933, | |
| "logits/chosen": -2.26902437210083, | |
| "logits/rejected": -2.2395756244659424, | |
| "logps/chosen": -0.5438047647476196, | |
| "logps/rejected": -0.4186398386955261, | |
| "loss": 0.9454, | |
| "nll_loss": 0.9862927198410034, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.46787089132366005, | |
| "grad_norm": 6.78125, | |
| "learning_rate": 3.2063606616290626e-07, | |
| "log_odds_chosen": -0.40437692403793335, | |
| "log_odds_ratio": -0.991305947303772, | |
| "logits/chosen": -2.2474241256713867, | |
| "logits/rejected": -2.2076640129089355, | |
| "logps/chosen": -0.5990616083145142, | |
| "logps/rejected": -0.40946364402770996, | |
| "loss": 0.8641, | |
| "nll_loss": 0.827691376209259, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.47379330766952915, | |
| "grad_norm": 12.25, | |
| "learning_rate": 3.1566149400420523e-07, | |
| "log_odds_chosen": -0.3424193859100342, | |
| "log_odds_ratio": -0.9442498087882996, | |
| "logits/chosen": -2.300968885421753, | |
| "logits/rejected": -2.289825201034546, | |
| "logps/chosen": -0.6068278551101685, | |
| "logps/rejected": -0.4422214925289154, | |
| "loss": 0.9572, | |
| "nll_loss": 0.9411390423774719, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.4797157240153983, | |
| "grad_norm": 8.75, | |
| "learning_rate": 3.1065883651900087e-07, | |
| "log_odds_chosen": -0.3020106852054596, | |
| "log_odds_ratio": -0.9359525442123413, | |
| "logits/chosen": -2.288480281829834, | |
| "logits/rejected": -2.246896982192993, | |
| "logps/chosen": -0.5823680758476257, | |
| "logps/rejected": -0.44429031014442444, | |
| "loss": 0.9677, | |
| "nll_loss": 0.9093478918075562, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.4856381403612674, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 3.056302334890786e-07, | |
| "log_odds_chosen": -0.38523969054222107, | |
| "log_odds_ratio": -0.9780759811401367, | |
| "logits/chosen": -2.294841766357422, | |
| "logits/rejected": -2.2723891735076904, | |
| "logps/chosen": -0.6043334603309631, | |
| "logps/rejected": -0.4185991883277893, | |
| "loss": 0.9121, | |
| "nll_loss": 0.9030720591545105, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.4915605567071365, | |
| "grad_norm": 8.75, | |
| "learning_rate": 3.0057783579388586e-07, | |
| "log_odds_chosen": -0.24561011791229248, | |
| "log_odds_ratio": -0.8836873173713684, | |
| "logits/chosen": -2.2996482849121094, | |
| "logits/rejected": -2.258457660675049, | |
| "logps/chosen": -0.5478182435035706, | |
| "logps/rejected": -0.440875768661499, | |
| "loss": 0.928, | |
| "nll_loss": 0.9274915456771851, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.49748297305300565, | |
| "grad_norm": 7.8125, | |
| "learning_rate": 2.9550380449053907e-07, | |
| "log_odds_chosen": -0.26652732491493225, | |
| "log_odds_ratio": -0.9008363485336304, | |
| "logits/chosen": -2.2529563903808594, | |
| "logits/rejected": -2.2309823036193848, | |
| "logps/chosen": -0.5522275567054749, | |
| "logps/rejected": -0.42188987135887146, | |
| "loss": 0.914, | |
| "nll_loss": 0.8120133280754089, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.5034053893988747, | |
| "grad_norm": 6.4375, | |
| "learning_rate": 2.904103098894767e-07, | |
| "log_odds_chosen": -0.3553586006164551, | |
| "log_odds_ratio": -0.9903032183647156, | |
| "logits/chosen": -2.291224956512451, | |
| "logits/rejected": -2.2469000816345215, | |
| "logps/chosen": -0.6393681764602661, | |
| "logps/rejected": -0.42175260186195374, | |
| "loss": 0.9129, | |
| "nll_loss": 0.9170019030570984, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.5093278057447439, | |
| "grad_norm": 9.75, | |
| "learning_rate": 2.852995306261545e-07, | |
| "log_odds_chosen": -0.2889431416988373, | |
| "log_odds_ratio": -0.9156063795089722, | |
| "logits/chosen": -2.318115472793579, | |
| "logits/rejected": -2.2808139324188232, | |
| "logps/chosen": -0.5667640566825867, | |
| "logps/rejected": -0.44533196091651917, | |
| "loss": 0.9772, | |
| "nll_loss": 0.9936600923538208, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.515250222090613, | |
| "grad_norm": 8.75, | |
| "learning_rate": 2.801736527291797e-07, | |
| "log_odds_chosen": -0.3678986728191376, | |
| "log_odds_ratio": -0.9755579233169556, | |
| "logits/chosen": -2.2834322452545166, | |
| "logits/rejected": -2.2398197650909424, | |
| "logps/chosen": -0.6312032341957092, | |
| "logps/rejected": -0.4318135380744934, | |
| "loss": 0.9391, | |
| "nll_loss": 0.8974191546440125, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.5211726384364821, | |
| "grad_norm": 7.84375, | |
| "learning_rate": 2.750348686852836e-07, | |
| "log_odds_chosen": -0.40664905309677124, | |
| "log_odds_ratio": -0.9781969785690308, | |
| "logits/chosen": -2.3389241695404053, | |
| "logits/rejected": -2.2737958431243896, | |
| "logps/chosen": -0.6092024445533752, | |
| "logps/rejected": -0.4250633120536804, | |
| "loss": 0.971, | |
| "nll_loss": 0.9957748651504517, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.5270950547823512, | |
| "grad_norm": 9.5, | |
| "learning_rate": 2.69885376501531e-07, | |
| "log_odds_chosen": -0.31569716334342957, | |
| "log_odds_ratio": -0.9389151334762573, | |
| "logits/chosen": -2.2705588340759277, | |
| "logits/rejected": -2.2545580863952637, | |
| "logps/chosen": -0.6091697812080383, | |
| "logps/rejected": -0.45014920830726624, | |
| "loss": 0.9568, | |
| "nll_loss": 0.9439749717712402, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.5330174711282203, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 2.647273787651687e-07, | |
| "log_odds_chosen": -0.27334731817245483, | |
| "log_odds_ratio": -0.8902351260185242, | |
| "logits/chosen": -2.3029747009277344, | |
| "logits/rejected": -2.2809951305389404, | |
| "logps/chosen": -0.5580970644950867, | |
| "logps/rejected": -0.43120306730270386, | |
| "loss": 0.9333, | |
| "nll_loss": 0.9487207531929016, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.5389398874740894, | |
| "grad_norm": 11.5, | |
| "learning_rate": 2.5956308170151526e-07, | |
| "log_odds_chosen": -0.5443618893623352, | |
| "log_odds_ratio": -1.1218284368515015, | |
| "logits/chosen": -2.275094985961914, | |
| "logits/rejected": -2.2452805042266846, | |
| "logps/chosen": -0.7426999807357788, | |
| "logps/rejected": -0.4159156382083893, | |
| "loss": 1.0105, | |
| "nll_loss": 0.9728318452835083, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.5448623038199586, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 2.543946942302944e-07, | |
| "log_odds_chosen": -0.30027318000793457, | |
| "log_odds_ratio": -0.9199014902114868, | |
| "logits/chosen": -2.265780210494995, | |
| "logits/rejected": -2.2284467220306396, | |
| "logps/chosen": -0.5695372819900513, | |
| "logps/rejected": -0.4176994264125824, | |
| "loss": 0.9145, | |
| "nll_loss": 0.9305332899093628, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.5507847201658277, | |
| "grad_norm": 9.375, | |
| "learning_rate": 2.492244270208158e-07, | |
| "log_odds_chosen": -0.2501292824745178, | |
| "log_odds_ratio": -0.8865777850151062, | |
| "logits/chosen": -2.273998260498047, | |
| "logits/rejected": -2.247119665145874, | |
| "logps/chosen": -0.562545895576477, | |
| "logps/rejected": -0.44244521856307983, | |
| "loss": 0.9004, | |
| "nll_loss": 0.9349013566970825, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.5567071365116968, | |
| "grad_norm": 8.25, | |
| "learning_rate": 2.440544915464078e-07, | |
| "log_odds_chosen": -0.29280886054039, | |
| "log_odds_ratio": -0.9178652763366699, | |
| "logits/chosen": -2.304103374481201, | |
| "logits/rejected": -2.2641754150390625, | |
| "logps/chosen": -0.5466963052749634, | |
| "logps/rejected": -0.41387224197387695, | |
| "loss": 0.9221, | |
| "nll_loss": 0.9058642387390137, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.5626295528575659, | |
| "grad_norm": 7.875, | |
| "learning_rate": 2.3888709913850593e-07, | |
| "log_odds_chosen": -0.30006080865859985, | |
| "log_odds_ratio": -0.9208847880363464, | |
| "logits/chosen": -2.350160837173462, | |
| "logits/rejected": -2.3101677894592285, | |
| "logps/chosen": -0.5762113928794861, | |
| "logps/rejected": -0.43552321195602417, | |
| "loss": 0.9856, | |
| "nll_loss": 0.938804030418396, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.568551969203435, | |
| "grad_norm": 7.8125, | |
| "learning_rate": 2.337244600408025e-07, | |
| "log_odds_chosen": -0.39082369208335876, | |
| "log_odds_ratio": -0.9911519289016724, | |
| "logits/chosen": -2.3191308975219727, | |
| "logits/rejected": -2.2875494956970215, | |
| "logps/chosen": -0.6351069211959839, | |
| "logps/rejected": -0.4366016387939453, | |
| "loss": 0.9507, | |
| "nll_loss": 0.9739691615104675, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.5744743855493041, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 2.2856878246386085e-07, | |
| "log_odds_chosen": -0.2834867537021637, | |
| "log_odds_ratio": -0.9153987765312195, | |
| "logits/chosen": -2.3153960704803467, | |
| "logits/rejected": -2.2916574478149414, | |
| "logps/chosen": -0.5750494003295898, | |
| "logps/rejected": -0.4384193420410156, | |
| "loss": 0.9944, | |
| "nll_loss": 1.0159144401550293, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.5803968018951732, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 2.2342227164060035e-07, | |
| "log_odds_chosen": -0.37823957204818726, | |
| "log_odds_ratio": -0.9799555540084839, | |
| "logits/chosen": -2.2767786979675293, | |
| "logits/rejected": -2.2212002277374268, | |
| "logps/chosen": -0.6210430860519409, | |
| "logps/rejected": -0.4381546378135681, | |
| "loss": 0.9342, | |
| "nll_loss": 0.9048612713813782, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.5863192182410424, | |
| "grad_norm": 7.625, | |
| "learning_rate": 2.182871288830533e-07, | |
| "log_odds_chosen": -0.3980916738510132, | |
| "log_odds_ratio": -0.9920517206192017, | |
| "logits/chosen": -2.3002982139587402, | |
| "logits/rejected": -2.237112522125244, | |
| "logps/chosen": -0.6255283951759338, | |
| "logps/rejected": -0.4390384256839752, | |
| "loss": 0.9667, | |
| "nll_loss": 0.9580439329147339, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.5922416345869115, | |
| "grad_norm": 7.8125, | |
| "learning_rate": 2.131655506408007e-07, | |
| "log_odds_chosen": -0.3284297287464142, | |
| "log_odds_ratio": -0.9404581785202026, | |
| "logits/chosen": -2.3054287433624268, | |
| "logits/rejected": -2.263627767562866, | |
| "logps/chosen": -0.5948997139930725, | |
| "logps/rejected": -0.4438301622867584, | |
| "loss": 0.9222, | |
| "nll_loss": 0.9223626852035522, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.5981640509327806, | |
| "grad_norm": 7.65625, | |
| "learning_rate": 2.0805972756148643e-07, | |
| "log_odds_chosen": -0.48507261276245117, | |
| "log_odds_ratio": -1.0847915410995483, | |
| "logits/chosen": -2.2977192401885986, | |
| "logits/rejected": -2.2829782962799072, | |
| "logps/chosen": -0.748909592628479, | |
| "logps/rejected": -0.43740910291671753, | |
| "loss": 0.9675, | |
| "nll_loss": 0.9915729761123657, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.6040864672786497, | |
| "grad_norm": 7.59375, | |
| "learning_rate": 2.0297184355381432e-07, | |
| "log_odds_chosen": -0.3442012667655945, | |
| "log_odds_ratio": -0.9447819590568542, | |
| "logits/chosen": -2.3124001026153564, | |
| "logits/rejected": -2.2729249000549316, | |
| "logps/chosen": -0.5671228170394897, | |
| "logps/rejected": -0.42450952529907227, | |
| "loss": 0.9345, | |
| "nll_loss": 0.9476312398910522, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.6100088836245188, | |
| "grad_norm": 7.71875, | |
| "learning_rate": 1.9790407485342638e-07, | |
| "log_odds_chosen": -0.5493720769882202, | |
| "log_odds_ratio": -1.1290843486785889, | |
| "logits/chosen": -2.3353946208953857, | |
| "logits/rejected": -2.2935006618499756, | |
| "logps/chosen": -0.7500286102294922, | |
| "logps/rejected": -0.4041396975517273, | |
| "loss": 0.9066, | |
| "nll_loss": 0.930087685585022, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.6159312999703879, | |
| "grad_norm": 9.125, | |
| "learning_rate": 1.928585890920641e-07, | |
| "log_odds_chosen": -0.2760196626186371, | |
| "log_odds_ratio": -0.9122495651245117, | |
| "logits/chosen": -2.303187608718872, | |
| "logits/rejected": -2.267937183380127, | |
| "logps/chosen": -0.5642871856689453, | |
| "logps/rejected": -0.42797571420669556, | |
| "loss": 0.9504, | |
| "nll_loss": 0.9056134223937988, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.6218537163162571, | |
| "grad_norm": 8.25, | |
| "learning_rate": 1.8783754437040902e-07, | |
| "log_odds_chosen": -0.34286069869995117, | |
| "log_odds_ratio": -0.9555438756942749, | |
| "logits/chosen": -2.282454013824463, | |
| "logits/rejected": -2.247560739517212, | |
| "logps/chosen": -0.5602587461471558, | |
| "logps/rejected": -0.41198721528053284, | |
| "loss": 0.9096, | |
| "nll_loss": 0.8938838243484497, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.6277761326621262, | |
| "grad_norm": 7.53125, | |
| "learning_rate": 1.8284308833500118e-07, | |
| "log_odds_chosen": -0.30499863624572754, | |
| "log_odds_ratio": -0.9316195249557495, | |
| "logits/chosen": -2.2879326343536377, | |
| "logits/rejected": -2.2628307342529297, | |
| "logps/chosen": -0.5727280378341675, | |
| "logps/rejected": -0.4329405725002289, | |
| "loss": 0.9415, | |
| "nll_loss": 0.9282618761062622, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.6336985490079953, | |
| "grad_norm": 9.0, | |
| "learning_rate": 1.7787735725962756e-07, | |
| "log_odds_chosen": -0.371854692697525, | |
| "log_odds_ratio": -0.9633452296257019, | |
| "logits/chosen": -2.295264482498169, | |
| "logits/rejected": -2.2589457035064697, | |
| "logps/chosen": -0.608909010887146, | |
| "logps/rejected": -0.43556636571884155, | |
| "loss": 0.9874, | |
| "nll_loss": 0.9842734336853027, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.6396209653538644, | |
| "grad_norm": 7.4375, | |
| "learning_rate": 1.7294247513157616e-07, | |
| "log_odds_chosen": -0.30945563316345215, | |
| "log_odds_ratio": -0.9170244932174683, | |
| "logits/chosen": -2.32027530670166, | |
| "logits/rejected": -2.2682743072509766, | |
| "logps/chosen": -0.5618667006492615, | |
| "logps/rejected": -0.42478686571121216, | |
| "loss": 0.9273, | |
| "nll_loss": 0.9494869112968445, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.6455433816997335, | |
| "grad_norm": 8.25, | |
| "learning_rate": 1.6804055274314494e-07, | |
| "log_odds_chosen": -0.28144484758377075, | |
| "log_odds_ratio": -0.9044340252876282, | |
| "logits/chosen": -2.2794411182403564, | |
| "logits/rejected": -2.256417989730835, | |
| "logps/chosen": -0.5532391667366028, | |
| "logps/rejected": -0.43312329053878784, | |
| "loss": 0.9267, | |
| "nll_loss": 0.8960529565811157, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.6514657980456026, | |
| "grad_norm": 7.8125, | |
| "learning_rate": 1.6317368678879496e-07, | |
| "log_odds_chosen": -0.28822919726371765, | |
| "log_odds_ratio": -0.9030183553695679, | |
| "logits/chosen": -2.3176181316375732, | |
| "logits/rejected": -2.2776379585266113, | |
| "logps/chosen": -0.5691734552383423, | |
| "logps/rejected": -0.4426758885383606, | |
| "loss": 0.9774, | |
| "nll_loss": 0.9514939188957214, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.6573882143914718, | |
| "grad_norm": 10.0, | |
| "learning_rate": 1.5834395896833281e-07, | |
| "log_odds_chosen": -0.3918454051017761, | |
| "log_odds_ratio": -0.9793996810913086, | |
| "logits/chosen": -2.3301963806152344, | |
| "logits/rejected": -2.274294376373291, | |
| "logps/chosen": -0.6070703864097595, | |
| "logps/rejected": -0.42063984274864197, | |
| "loss": 0.9521, | |
| "nll_loss": 0.956030547618866, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.6633106307373409, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 1.535534350965075e-07, | |
| "log_odds_chosen": -0.3459760546684265, | |
| "log_odds_ratio": -0.9434119462966919, | |
| "logits/chosen": -2.3291070461273193, | |
| "logits/rejected": -2.3154056072235107, | |
| "logps/chosen": -0.5571088790893555, | |
| "logps/rejected": -0.3965280055999756, | |
| "loss": 0.9099, | |
| "nll_loss": 0.9126838445663452, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.66923304708321, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 1.4880416421940154e-07, | |
| "log_odds_chosen": -0.35412847995758057, | |
| "log_odds_ratio": -0.9552983045578003, | |
| "logits/chosen": -2.2702512741088867, | |
| "logits/rejected": -2.246307849884033, | |
| "logps/chosen": -0.6233401298522949, | |
| "logps/rejected": -0.4397760033607483, | |
| "loss": 1.0082, | |
| "nll_loss": 1.0226290225982666, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.6751554634290791, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 1.4409817773799459e-07, | |
| "log_odds_chosen": -0.30409640073776245, | |
| "log_odds_ratio": -0.933831512928009, | |
| "logits/chosen": -2.2992305755615234, | |
| "logits/rejected": -2.2570438385009766, | |
| "logps/chosen": -0.6016424894332886, | |
| "logps/rejected": -0.44892677664756775, | |
| "loss": 0.9551, | |
| "nll_loss": 0.9244022369384766, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.6810778797749482, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 1.3943748853927385e-07, | |
| "log_odds_chosen": -0.41090458631515503, | |
| "log_odds_ratio": -1.001075029373169, | |
| "logits/chosen": -2.299729824066162, | |
| "logits/rejected": -2.2916903495788574, | |
| "logps/chosen": -0.6525920033454895, | |
| "logps/rejected": -0.4317931532859802, | |
| "loss": 0.9303, | |
| "nll_loss": 0.9261299967765808, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.6870002961208173, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 1.3482409013526436e-07, | |
| "log_odds_chosen": -0.42632365226745605, | |
| "log_odds_ratio": -1.0060193538665771, | |
| "logits/chosen": -2.2864601612091064, | |
| "logits/rejected": -2.2797439098358154, | |
| "logps/chosen": -0.615047037601471, | |
| "logps/rejected": -0.4228528141975403, | |
| "loss": 0.9621, | |
| "nll_loss": 0.9737777709960938, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.6929227124666865, | |
| "grad_norm": 7.65625, | |
| "learning_rate": 1.302599558103456e-07, | |
| "log_odds_chosen": -0.3416286110877991, | |
| "log_odds_ratio": -0.971416175365448, | |
| "logits/chosen": -2.3440544605255127, | |
| "logits/rejected": -2.3086702823638916, | |
| "logps/chosen": -0.6329351663589478, | |
| "logps/rejected": -0.44783586263656616, | |
| "loss": 0.9418, | |
| "nll_loss": 0.9567440152168274, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.6988451288125556, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 1.257470377772214e-07, | |
| "log_odds_chosen": -0.37471523880958557, | |
| "log_odds_ratio": -0.9742682576179504, | |
| "logits/chosen": -2.320568323135376, | |
| "logits/rejected": -2.2932517528533936, | |
| "logps/chosen": -0.5956822633743286, | |
| "logps/rejected": -0.410876601934433, | |
| "loss": 0.9619, | |
| "nll_loss": 0.9405835270881653, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.7047675451584247, | |
| "grad_norm": 7.96875, | |
| "learning_rate": 1.2128726634190046e-07, | |
| "log_odds_chosen": -0.3462384343147278, | |
| "log_odds_ratio": -0.9395328760147095, | |
| "logits/chosen": -2.3269436359405518, | |
| "logits/rejected": -2.2818374633789062, | |
| "logps/chosen": -0.5803397297859192, | |
| "logps/rejected": -0.4099668860435486, | |
| "loss": 0.9172, | |
| "nll_loss": 0.9042008519172668, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.7106899615042938, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 1.1688254907804992e-07, | |
| "log_odds_chosen": -0.3338465392589569, | |
| "log_odds_ratio": -0.9492910504341125, | |
| "logits/chosen": -2.282212495803833, | |
| "logits/rejected": -2.2418830394744873, | |
| "logps/chosen": -0.6125479340553284, | |
| "logps/rejected": -0.45005935430526733, | |
| "loss": 0.9674, | |
| "nll_loss": 0.9735835194587708, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.7166123778501629, | |
| "grad_norm": 7.5, | |
| "learning_rate": 1.1253477001106956e-07, | |
| "log_odds_chosen": -0.26436474919319153, | |
| "log_odds_ratio": -0.8985050916671753, | |
| "logits/chosen": -2.259978771209717, | |
| "logits/rejected": -2.223177433013916, | |
| "logps/chosen": -0.5662964582443237, | |
| "logps/rejected": -0.4440518915653229, | |
| "loss": 0.9408, | |
| "nll_loss": 0.90367591381073, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.722534794196032, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.0824578881224065e-07, | |
| "log_odds_chosen": -0.24436886608600616, | |
| "log_odds_ratio": -0.8822824358940125, | |
| "logits/chosen": -2.332968235015869, | |
| "logits/rejected": -2.3182759284973145, | |
| "logps/chosen": -0.539296567440033, | |
| "logps/rejected": -0.42076578736305237, | |
| "loss": 0.8899, | |
| "nll_loss": 0.8598573803901672, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.728457210541901, | |
| "grad_norm": 8.125, | |
| "learning_rate": 1.0401744000328918e-07, | |
| "log_odds_chosen": -0.28977444767951965, | |
| "log_odds_ratio": -0.9303587675094604, | |
| "logits/chosen": -2.2798142433166504, | |
| "logits/rejected": -2.2761147022247314, | |
| "logps/chosen": -0.5908230543136597, | |
| "logps/rejected": -0.449887216091156, | |
| "loss": 0.9195, | |
| "nll_loss": 0.9145529866218567, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.7343796268877703, | |
| "grad_norm": 9.375, | |
| "learning_rate": 9.985153217170902e-08, | |
| "log_odds_chosen": -0.357065886259079, | |
| "log_odds_ratio": -0.9588850140571594, | |
| "logits/chosen": -2.3385989665985107, | |
| "logits/rejected": -2.323024034500122, | |
| "logps/chosen": -0.6093414425849915, | |
| "logps/rejected": -0.4347008168697357, | |
| "loss": 1.008, | |
| "nll_loss": 1.0087924003601074, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.7403020432336394, | |
| "grad_norm": 10.5, | |
| "learning_rate": 9.574984719717553e-08, | |
| "log_odds_chosen": -0.3260490894317627, | |
| "log_odds_ratio": -0.9441172480583191, | |
| "logits/chosen": -2.321216344833374, | |
| "logits/rejected": -2.302063226699829, | |
| "logps/chosen": -0.5826759338378906, | |
| "logps/rejected": -0.4284025728702545, | |
| "loss": 0.9308, | |
| "nll_loss": 0.9714264869689941, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.7462244595795084, | |
| "grad_norm": 8.25, | |
| "learning_rate": 9.171413948938459e-08, | |
| "log_odds_chosen": -0.3101581037044525, | |
| "log_odds_ratio": -0.929049015045166, | |
| "logits/chosen": -2.310981273651123, | |
| "logits/rejected": -2.2564587593078613, | |
| "logps/chosen": -0.6019686460494995, | |
| "logps/rejected": -0.45062392950057983, | |
| "loss": 0.9576, | |
| "nll_loss": 0.9787800908088684, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.7521468759253775, | |
| "grad_norm": 7.53125, | |
| "learning_rate": 8.774613523764049e-08, | |
| "log_odds_chosen": -0.369983971118927, | |
| "log_odds_ratio": -0.9563344120979309, | |
| "logits/chosen": -2.2968955039978027, | |
| "logits/rejected": -2.248944044113159, | |
| "logps/chosen": -0.5875356197357178, | |
| "logps/rejected": -0.4163094162940979, | |
| "loss": 0.9305, | |
| "nll_loss": 0.900018572807312, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.7580692922712466, | |
| "grad_norm": 9.125, | |
| "learning_rate": 8.384753167251412e-08, | |
| "log_odds_chosen": -0.32507094740867615, | |
| "log_odds_ratio": -0.9397505521774292, | |
| "logits/chosen": -2.2516260147094727, | |
| "logits/rejected": -2.226477861404419, | |
| "logps/chosen": -0.5629323124885559, | |
| "logps/rejected": -0.41151052713394165, | |
| "loss": 0.8916, | |
| "nll_loss": 0.8657590746879578, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.7639917086171157, | |
| "grad_norm": 8.375, | |
| "learning_rate": 8.001999633988942e-08, | |
| "log_odds_chosen": -0.3611569106578827, | |
| "log_odds_ratio": -0.957983672618866, | |
| "logits/chosen": -2.3322999477386475, | |
| "logits/rejected": -2.283409833908081, | |
| "logps/chosen": -0.5784574747085571, | |
| "logps/rejected": -0.4157342314720154, | |
| "loss": 0.8973, | |
| "nll_loss": 0.8929991722106934, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.769914124962985, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 7.62651663877042e-08, | |
| "log_odds_chosen": -0.26533371210098267, | |
| "log_odds_ratio": -0.9089031219482422, | |
| "logits/chosen": -2.2688136100769043, | |
| "logits/rejected": -2.2409274578094482, | |
| "logps/chosen": -0.5685082077980042, | |
| "logps/rejected": -0.4421761631965637, | |
| "loss": 0.9965, | |
| "nll_loss": 0.9676351547241211, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.775836541308854, | |
| "grad_norm": 7.1875, | |
| "learning_rate": 7.258464786569549e-08, | |
| "log_odds_chosen": -0.28731244802474976, | |
| "log_odds_ratio": -0.9186748266220093, | |
| "logits/chosen": -2.3306045532226562, | |
| "logits/rejected": -2.2782888412475586, | |
| "logps/chosen": -0.5666372776031494, | |
| "logps/rejected": -0.43749627470970154, | |
| "loss": 0.9659, | |
| "nll_loss": 0.9544159770011902, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.7817589576547231, | |
| "grad_norm": 9.75, | |
| "learning_rate": 6.898001503844483e-08, | |
| "log_odds_chosen": -0.5405977964401245, | |
| "log_odds_ratio": -1.1180508136749268, | |
| "logits/chosen": -2.3619232177734375, | |
| "logits/rejected": -2.3188281059265137, | |
| "logps/chosen": -0.7727476358413696, | |
| "logps/rejected": -0.4373859465122223, | |
| "loss": 0.9429, | |
| "nll_loss": 0.9921876192092896, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.7876813740005922, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 6.545280971202014e-08, | |
| "log_odds_chosen": -0.2534041702747345, | |
| "log_odds_ratio": -0.8981307744979858, | |
| "logits/chosen": -2.320126533508301, | |
| "logits/rejected": -2.289376974105835, | |
| "logps/chosen": -0.5534666180610657, | |
| "logps/rejected": -0.4292474687099457, | |
| "loss": 0.9168, | |
| "nll_loss": 0.9440558552742004, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.7936037903464613, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 6.200454057450022e-08, | |
| "log_odds_chosen": -0.36177825927734375, | |
| "log_odds_ratio": -0.9495649337768555, | |
| "logits/chosen": -2.2736241817474365, | |
| "logits/rejected": -2.226933479309082, | |
| "logps/chosen": -0.59937584400177, | |
| "logps/rejected": -0.42542099952697754, | |
| "loss": 0.9704, | |
| "nll_loss": 0.9059191942214966, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.7995262066923304, | |
| "grad_norm": 9.0, | |
| "learning_rate": 5.863668255066492e-08, | |
| "log_odds_chosen": -0.313324511051178, | |
| "log_odds_ratio": -0.919102668762207, | |
| "logits/chosen": -2.270073890686035, | |
| "logits/rejected": -2.2372827529907227, | |
| "logps/chosen": -0.5768779516220093, | |
| "logps/rejected": -0.43518179655075073, | |
| "loss": 0.9124, | |
| "nll_loss": 0.9345908164978027, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.8054486230381996, | |
| "grad_norm": 8.375, | |
| "learning_rate": 5.53506761711274e-08, | |
| "log_odds_chosen": -0.2887657880783081, | |
| "log_odds_ratio": -0.912114143371582, | |
| "logits/chosen": -2.305987596511841, | |
| "logits/rejected": -2.2752127647399902, | |
| "logps/chosen": -0.587549090385437, | |
| "logps/rejected": -0.4469973146915436, | |
| "loss": 0.9423, | |
| "nll_loss": 0.9892560243606567, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.8113710393840687, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 5.2147926956177174e-08, | |
| "log_odds_chosen": -0.45511436462402344, | |
| "log_odds_ratio": -1.0450800657272339, | |
| "logits/chosen": -2.294468641281128, | |
| "logits/rejected": -2.283860445022583, | |
| "logps/chosen": -0.6607165336608887, | |
| "logps/rejected": -0.4220455586910248, | |
| "loss": 0.9395, | |
| "nll_loss": 0.9603630304336548, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.8172934557299378, | |
| "grad_norm": 15.3125, | |
| "learning_rate": 4.902980481459834e-08, | |
| "log_odds_chosen": -0.26191025972366333, | |
| "log_odds_ratio": -0.9043244123458862, | |
| "logits/chosen": -2.277843475341797, | |
| "logits/rejected": -2.248347520828247, | |
| "logps/chosen": -0.5755423307418823, | |
| "logps/rejected": -0.4490273594856262, | |
| "loss": 0.9003, | |
| "nll_loss": 0.9170975685119629, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.8232158720758069, | |
| "grad_norm": 9.25, | |
| "learning_rate": 4.5997643457719646e-08, | |
| "log_odds_chosen": -0.35434719920158386, | |
| "log_odds_ratio": -0.9506848454475403, | |
| "logits/chosen": -2.295780658721924, | |
| "logits/rejected": -2.2894127368927, | |
| "logps/chosen": -0.5851874351501465, | |
| "logps/rejected": -0.4212135672569275, | |
| "loss": 0.8958, | |
| "nll_loss": 0.9056984186172485, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.829138288421676, | |
| "grad_norm": 8.75, | |
| "learning_rate": 4.305273982894772e-08, | |
| "log_odds_chosen": -0.33616143465042114, | |
| "log_odds_ratio": -0.9503694772720337, | |
| "logits/chosen": -2.3287465572357178, | |
| "logits/rejected": -2.286414623260498, | |
| "logps/chosen": -0.6181541681289673, | |
| "logps/rejected": -0.4454525113105774, | |
| "loss": 0.9425, | |
| "nll_loss": 0.9352006912231445, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.8350607047675451, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 4.0196353549026786e-08, | |
| "log_odds_chosen": -0.30044835805892944, | |
| "log_odds_ratio": -0.9167086482048035, | |
| "logits/chosen": -2.298393487930298, | |
| "logits/rejected": -2.26066255569458, | |
| "logps/chosen": -0.5805574655532837, | |
| "logps/rejected": -0.43902960419654846, | |
| "loss": 0.9712, | |
| "nll_loss": 1.0039526224136353, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.8409831211134142, | |
| "grad_norm": 7.75, | |
| "learning_rate": 3.742970637726181e-08, | |
| "log_odds_chosen": -0.179987370967865, | |
| "log_odds_ratio": -0.8582404255867004, | |
| "logits/chosen": -2.3168177604675293, | |
| "logits/rejected": -2.269207000732422, | |
| "logps/chosen": -0.5285545587539673, | |
| "logps/rejected": -0.44142407178878784, | |
| "loss": 0.9201, | |
| "nll_loss": 0.9034355878829956, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.8469055374592834, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 3.4753981688937284e-08, | |
| "log_odds_chosen": -0.3474799394607544, | |
| "log_odds_ratio": -0.9500767588615417, | |
| "logits/chosen": -2.2948362827301025, | |
| "logits/rejected": -2.2666220664978027, | |
| "logps/chosen": -0.5843050479888916, | |
| "logps/rejected": -0.4246344566345215, | |
| "loss": 0.9549, | |
| "nll_loss": 0.9555429220199585, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.8528279538051525, | |
| "grad_norm": 8.25, | |
| "learning_rate": 3.217032396915265e-08, | |
| "log_odds_chosen": -0.40568438172340393, | |
| "log_odds_ratio": -1.0082272291183472, | |
| "logits/chosen": -2.3024380207061768, | |
| "logits/rejected": -2.268986701965332, | |
| "logps/chosen": -0.6786967515945435, | |
| "logps/rejected": -0.44073349237442017, | |
| "loss": 0.9636, | |
| "nll_loss": 0.9871211051940918, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.8587503701510216, | |
| "grad_norm": 9.75, | |
| "learning_rate": 2.9679838323293404e-08, | |
| "log_odds_chosen": -0.4226885437965393, | |
| "log_odds_ratio": -1.0269486904144287, | |
| "logits/chosen": -2.2947676181793213, | |
| "logits/rejected": -2.2655680179595947, | |
| "logps/chosen": -0.6772679090499878, | |
| "logps/rejected": -0.44576793909072876, | |
| "loss": 0.9064, | |
| "nll_loss": 0.9240104556083679, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.8646727864968907, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 2.728359000434488e-08, | |
| "log_odds_chosen": -0.34337377548217773, | |
| "log_odds_ratio": -0.9470311403274536, | |
| "logits/chosen": -2.3264529705047607, | |
| "logits/rejected": -2.290132761001587, | |
| "logps/chosen": -0.5465956926345825, | |
| "logps/rejected": -0.41503897309303284, | |
| "loss": 0.9481, | |
| "nll_loss": 0.8961936831474304, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.8705952028427598, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 2.498260395725302e-08, | |
| "log_odds_chosen": -0.3448273539543152, | |
| "log_odds_ratio": -0.9480770230293274, | |
| "logits/chosen": -2.293290376663208, | |
| "logits/rejected": -2.278653144836426, | |
| "logps/chosen": -0.5983850955963135, | |
| "logps/rejected": -0.44902753829956055, | |
| "loss": 0.9406, | |
| "nll_loss": 0.93559330701828, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.8765176191886289, | |
| "grad_norm": 8.5, | |
| "learning_rate": 2.2777864380525426e-08, | |
| "log_odds_chosen": -0.29847949743270874, | |
| "log_odds_ratio": -0.9293072819709778, | |
| "logits/chosen": -2.2964632511138916, | |
| "logits/rejected": -2.2744767665863037, | |
| "logps/chosen": -0.59266597032547, | |
| "logps/rejected": -0.4337979853153229, | |
| "loss": 0.9143, | |
| "nll_loss": 0.8765565752983093, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.8824400355344981, | |
| "grad_norm": 8.4375, | |
| "learning_rate": 2.0670314305261423e-08, | |
| "log_odds_chosen": -0.2956581115722656, | |
| "log_odds_ratio": -0.914827823638916, | |
| "logits/chosen": -2.312617540359497, | |
| "logits/rejected": -2.2852249145507812, | |
| "logps/chosen": -0.5551884770393372, | |
| "logps/rejected": -0.42470401525497437, | |
| "loss": 0.8906, | |
| "nll_loss": 0.9051159620285034, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.8883624518803672, | |
| "grad_norm": 8.5, | |
| "learning_rate": 1.866085519178995e-08, | |
| "log_odds_chosen": -0.3115543723106384, | |
| "log_odds_ratio": -0.9432921409606934, | |
| "logits/chosen": -2.294912815093994, | |
| "logits/rejected": -2.2731730937957764, | |
| "logps/chosen": -0.6179423332214355, | |
| "logps/rejected": -0.47528520226478577, | |
| "loss": 0.9807, | |
| "nll_loss": 0.9959957003593445, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.8942848682262363, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 1.675034654408894e-08, | |
| "log_odds_chosen": -0.3954925239086151, | |
| "log_odds_ratio": -0.966839611530304, | |
| "logits/chosen": -2.331923007965088, | |
| "logits/rejected": -2.3019633293151855, | |
| "logps/chosen": -0.5650435090065002, | |
| "logps/rejected": -0.40297931432724, | |
| "loss": 0.9211, | |
| "nll_loss": 0.9447514414787292, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.9002072845721054, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 1.4939605542150595e-08, | |
| "log_odds_chosen": -0.2961687445640564, | |
| "log_odds_ratio": -0.9432598352432251, | |
| "logits/chosen": -2.31669545173645, | |
| "logits/rejected": -2.2790303230285645, | |
| "logps/chosen": -0.6368409395217896, | |
| "logps/rejected": -0.466596782207489, | |
| "loss": 0.9972, | |
| "nll_loss": 0.9845758676528931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.9061297009179745, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 1.3229406692449791e-08, | |
| "log_odds_chosen": -0.22676777839660645, | |
| "log_odds_ratio": -0.8955879211425781, | |
| "logits/chosen": -2.2553787231445312, | |
| "logits/rejected": -2.228121280670166, | |
| "logps/chosen": -0.5664690136909485, | |
| "logps/rejected": -0.45375269651412964, | |
| "loss": 0.9651, | |
| "nll_loss": 0.9422292709350586, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.9120521172638436, | |
| "grad_norm": 7.6875, | |
| "learning_rate": 1.162048149666503e-08, | |
| "log_odds_chosen": -0.28153032064437866, | |
| "log_odds_ratio": -0.9496873617172241, | |
| "logits/chosen": -2.3055193424224854, | |
| "logits/rejected": -2.2621009349823, | |
| "logps/chosen": -0.6041845083236694, | |
| "logps/rejected": -0.4735984802246094, | |
| "loss": 0.9464, | |
| "nll_loss": 0.9381273984909058, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.9179745336097128, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 1.0113518138794047e-08, | |
| "log_odds_chosen": -0.3274211287498474, | |
| "log_odds_ratio": -0.9445363283157349, | |
| "logits/chosen": -2.2565391063690186, | |
| "logits/rejected": -2.233027935028076, | |
| "logps/chosen": -0.5867388844490051, | |
| "logps/rejected": -0.4352657198905945, | |
| "loss": 0.9594, | |
| "nll_loss": 0.9287152290344238, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.9238969499555819, | |
| "grad_norm": 8.625, | |
| "learning_rate": 8.709161190797565e-09, | |
| "log_odds_chosen": -0.23092766106128693, | |
| "log_odds_ratio": -0.8911072611808777, | |
| "logits/chosen": -2.3257815837860107, | |
| "logits/rejected": -2.29530668258667, | |
| "logps/chosen": -0.5526595115661621, | |
| "logps/rejected": -0.4391084611415863, | |
| "loss": 0.9337, | |
| "nll_loss": 0.9090098142623901, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.929819366301451, | |
| "grad_norm": 9.0, | |
| "learning_rate": 7.408011336897141e-09, | |
| "log_odds_chosen": -0.4632336497306824, | |
| "log_odds_ratio": -1.089444875717163, | |
| "logits/chosen": -2.346909284591675, | |
| "logits/rejected": -2.334372043609619, | |
| "logps/chosen": -0.7571093440055847, | |
| "logps/rejected": -0.4586968421936035, | |
| "loss": 0.9736, | |
| "nll_loss": 1.0062029361724854, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.9357417826473201, | |
| "grad_norm": 7.46875, | |
| "learning_rate": 6.210625116645135e-09, | |
| "log_odds_chosen": -0.41088682413101196, | |
| "log_odds_ratio": -0.9913327097892761, | |
| "logits/chosen": -2.3505208492279053, | |
| "logits/rejected": -2.3092150688171387, | |
| "logps/chosen": -0.6176980137825012, | |
| "logps/rejected": -0.42517074942588806, | |
| "loss": 0.9028, | |
| "nll_loss": 0.8675041198730469, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.9416641989931892, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 5.117514686876378e-09, | |
| "log_odds_chosen": -0.29983749985694885, | |
| "log_odds_ratio": -0.9341946840286255, | |
| "logits/chosen": -2.3163905143737793, | |
| "logits/rejected": -2.281881809234619, | |
| "logps/chosen": -0.569345235824585, | |
| "logps/rejected": -0.4335504174232483, | |
| "loss": 0.946, | |
| "nll_loss": 0.9458128809928894, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.9475866153390583, | |
| "grad_norm": 8.25, | |
| "learning_rate": 4.1291476026441565e-09, | |
| "log_odds_chosen": -0.21767720580101013, | |
| "log_odds_ratio": -0.8741191029548645, | |
| "logits/chosen": -2.2778666019439697, | |
| "logits/rejected": -2.2554242610931396, | |
| "logps/chosen": -0.560379147529602, | |
| "logps/rejected": -0.44631558656692505, | |
| "loss": 0.897, | |
| "nll_loss": 0.8650028109550476, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.9535090316849274, | |
| "grad_norm": 7.96875, | |
| "learning_rate": 3.2459466172331253e-09, | |
| "log_odds_chosen": -0.35443753004074097, | |
| "log_odds_ratio": -0.9958807229995728, | |
| "logits/chosen": -2.2850985527038574, | |
| "logits/rejected": -2.264432430267334, | |
| "logps/chosen": -0.6543992757797241, | |
| "logps/rejected": -0.436093807220459, | |
| "loss": 0.9838, | |
| "nll_loss": 0.9750429391860962, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.9594314480307966, | |
| "grad_norm": 10.25, | |
| "learning_rate": 2.4682895013354854e-09, | |
| "log_odds_chosen": -0.3622002899646759, | |
| "log_odds_ratio": -0.9864064455032349, | |
| "logits/chosen": -2.287553548812866, | |
| "logits/rejected": -2.2655410766601562, | |
| "logps/chosen": -0.6380153894424438, | |
| "logps/rejected": -0.4244503378868103, | |
| "loss": 0.9213, | |
| "nll_loss": 0.9489747881889343, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.9653538643766657, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 1.7965088814675677e-09, | |
| "log_odds_chosen": -0.4782753586769104, | |
| "log_odds_ratio": -1.0607492923736572, | |
| "logits/chosen": -2.290717840194702, | |
| "logits/rejected": -2.272459030151367, | |
| "logps/chosen": -0.6781035661697388, | |
| "logps/rejected": -0.4276870787143707, | |
| "loss": 0.9333, | |
| "nll_loss": 0.9644565582275391, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.9712762807225348, | |
| "grad_norm": 9.125, | |
| "learning_rate": 1.2308920976958348e-09, | |
| "log_odds_chosen": -0.29858607053756714, | |
| "log_odds_ratio": -0.9460258483886719, | |
| "logits/chosen": -2.269747734069824, | |
| "logits/rejected": -2.247730016708374, | |
| "logps/chosen": -0.6245580911636353, | |
| "logps/rejected": -0.45077449083328247, | |
| "loss": 0.906, | |
| "nll_loss": 0.9039252996444702, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.9771986970684039, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 7.716810807330276e-10, | |
| "log_odds_chosen": -0.4411376416683197, | |
| "log_odds_ratio": -1.0094521045684814, | |
| "logits/chosen": -2.2869138717651367, | |
| "logits/rejected": -2.24787974357605, | |
| "logps/chosen": -0.6290577054023743, | |
| "logps/rejected": -0.4051317572593689, | |
| "loss": 0.9304, | |
| "nll_loss": 0.9296571016311646, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.983121113414273, | |
| "grad_norm": 10.125, | |
| "learning_rate": 4.190722484575804e-10, | |
| "log_odds_chosen": -0.3509382903575897, | |
| "log_odds_ratio": -0.9882933497428894, | |
| "logits/chosen": -2.2927916049957275, | |
| "logits/rejected": -2.262193202972412, | |
| "logps/chosen": -0.6660831570625305, | |
| "logps/rejected": -0.4522073268890381, | |
| "loss": 0.9544, | |
| "nll_loss": 0.9784467816352844, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.9890435297601421, | |
| "grad_norm": 7.71875, | |
| "learning_rate": 1.732164218998522e-10, | |
| "log_odds_chosen": -0.35314035415649414, | |
| "log_odds_ratio": -0.9515780210494995, | |
| "logits/chosen": -2.265188694000244, | |
| "logits/rejected": -2.2218968868255615, | |
| "logps/chosen": -0.5966526865959167, | |
| "logps/rejected": -0.4354891777038574, | |
| "loss": 0.9111, | |
| "nll_loss": 0.9058610796928406, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.9949659461060113, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 3.4218760731730136e-11, | |
| "log_odds_chosen": -0.3034502863883972, | |
| "log_odds_ratio": -0.9285211563110352, | |
| "logits/chosen": -2.339616298675537, | |
| "logits/rejected": -2.2971951961517334, | |
| "logps/chosen": -0.5794862508773804, | |
| "logps/rejected": -0.43610063195228577, | |
| "loss": 0.9681, | |
| "nll_loss": 0.9744264483451843, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.9997038791827065, | |
| "step": 1688, | |
| "total_flos": 0.0, | |
| "train_loss": 0.9736523162132191, | |
| "train_runtime": 25409.1611, | |
| "train_samples_per_second": 2.126, | |
| "train_steps_per_second": 0.066 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1688, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "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": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |