| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.1807991321641656, | |
| "eval_steps": 500, | |
| "global_step": 500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0036159826432833123, | |
| "grad_norm": 0.3235728144645691, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.043861389160156, | |
| "logits/rejected": -18.043861389160156, | |
| "logps/chosen": -0.6726381778717041, | |
| "logps/rejected": -0.6726381778717041, | |
| "loss": 6.9675, | |
| "nll_loss": 1.6725690364837646, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.06726382672786713, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.06726382672786713, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0072319652865666245, | |
| "grad_norm": 0.3509086072444916, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.013948440551758, | |
| "logits/rejected": -18.013948440551758, | |
| "logps/chosen": -0.6865767240524292, | |
| "logps/rejected": -0.6865767240524292, | |
| "loss": 6.7227, | |
| "nll_loss": 1.6113628149032593, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.06865767389535904, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.06865767389535904, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.010847947929849937, | |
| "grad_norm": 0.38406211137771606, | |
| "learning_rate": 3e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.367061614990234, | |
| "logits/rejected": -18.367061614990234, | |
| "logps/chosen": -0.6837120056152344, | |
| "logps/rejected": -0.6837120056152344, | |
| "loss": 6.3986, | |
| "nll_loss": 1.5303384065628052, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.06837119162082672, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.06837119162082672, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.014463930573133249, | |
| "grad_norm": 0.4101921319961548, | |
| "learning_rate": 4.000000000000001e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -17.99128532409668, | |
| "logits/rejected": -17.99128532409668, | |
| "logps/chosen": -0.6463029980659485, | |
| "logps/rejected": -0.6463029980659485, | |
| "loss": 7.0291, | |
| "nll_loss": 1.6879545450210571, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.06463029980659485, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.06463029980659485, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.01807991321641656, | |
| "grad_norm": 0.444273978471756, | |
| "learning_rate": 5e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.08310890197754, | |
| "logits/rejected": -18.08310890197754, | |
| "logps/chosen": -0.6336182355880737, | |
| "logps/rejected": -0.6336182355880737, | |
| "loss": 6.5453, | |
| "nll_loss": 1.5670195817947388, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.06336182355880737, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.06336182355880737, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.021695895859699874, | |
| "grad_norm": 0.67641681432724, | |
| "learning_rate": 6e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.269500732421875, | |
| "logits/rejected": -18.269500732421875, | |
| "logps/chosen": -0.6496673822402954, | |
| "logps/rejected": -0.6496673822402954, | |
| "loss": 6.733, | |
| "nll_loss": 1.6139262914657593, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.06496674567461014, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.06496674567461014, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.025311878502983186, | |
| "grad_norm": 0.39636465907096863, | |
| "learning_rate": 7e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.115657806396484, | |
| "logits/rejected": -18.115657806396484, | |
| "logps/chosen": -0.6373459696769714, | |
| "logps/rejected": -0.6373459696769714, | |
| "loss": 6.8825, | |
| "nll_loss": 1.651307463645935, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.06373460590839386, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.06373460590839386, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.028927861146266498, | |
| "grad_norm": 0.5290302634239197, | |
| "learning_rate": 8.000000000000001e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.138675689697266, | |
| "logits/rejected": -18.138675689697266, | |
| "logps/chosen": -0.6702518463134766, | |
| "logps/rejected": -0.6702518463134766, | |
| "loss": 6.4063, | |
| "nll_loss": 1.5322693586349487, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.06702519208192825, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.06702519208192825, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.03254384378954981, | |
| "grad_norm": 0.2837754189968109, | |
| "learning_rate": 9e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.285091400146484, | |
| "logits/rejected": -18.285091400146484, | |
| "logps/chosen": -0.5695949196815491, | |
| "logps/rejected": -0.5695949196815491, | |
| "loss": 6.3599, | |
| "nll_loss": 1.520659327507019, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.056959498673677444, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.056959498673677444, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.03615982643283312, | |
| "grad_norm": 0.38816723227500916, | |
| "learning_rate": 1e-05, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.38321304321289, | |
| "logits/rejected": -18.38321304321289, | |
| "logps/chosen": -0.6220490336418152, | |
| "logps/rejected": -0.6220490336418152, | |
| "loss": 6.3579, | |
| "nll_loss": 1.5201700925827026, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.062204908579587936, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.062204908579587936, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.039775809076116435, | |
| "grad_norm": 0.485514760017395, | |
| "learning_rate": 9.88888888888889e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.24630355834961, | |
| "logits/rejected": -18.24630355834961, | |
| "logps/chosen": -0.58094322681427, | |
| "logps/rejected": -0.58094322681427, | |
| "loss": 6.8083, | |
| "nll_loss": 1.6327617168426514, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.058094322681427, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.058094322681427, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.04339179171939975, | |
| "grad_norm": 0.7200180292129517, | |
| "learning_rate": 9.777777777777779e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.222978591918945, | |
| "logits/rejected": -18.222978591918945, | |
| "logps/chosen": -0.6140703558921814, | |
| "logps/rejected": -0.6140703558921814, | |
| "loss": 7.0165, | |
| "nll_loss": 1.6848167181015015, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.06140704080462456, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.06140704080462456, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.04700777436268306, | |
| "grad_norm": 0.5130624175071716, | |
| "learning_rate": 9.666666666666667e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -17.98419189453125, | |
| "logits/rejected": -17.98419189453125, | |
| "logps/chosen": -0.6171376705169678, | |
| "logps/rejected": -0.6171376705169678, | |
| "loss": 6.9743, | |
| "nll_loss": 1.6742585897445679, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.06171376630663872, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.06171376630663872, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.05062375700596637, | |
| "grad_norm": 0.6743359565734863, | |
| "learning_rate": 9.555555555555556e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.43941879272461, | |
| "logits/rejected": -18.43941879272461, | |
| "logps/chosen": -0.5736243724822998, | |
| "logps/rejected": -0.5736243724822998, | |
| "loss": 6.4179, | |
| "nll_loss": 1.5351518392562866, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.057362429797649384, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.057362429797649384, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.054239739649249684, | |
| "grad_norm": 0.7007283568382263, | |
| "learning_rate": 9.444444444444445e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.523662567138672, | |
| "logits/rejected": -18.523662567138672, | |
| "logps/chosen": -0.5047087073326111, | |
| "logps/rejected": -0.5047087073326111, | |
| "loss": 6.4734, | |
| "nll_loss": 1.549041986465454, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.050470877438783646, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.050470877438783646, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.057855722292532996, | |
| "grad_norm": 0.8055678009986877, | |
| "learning_rate": 9.333333333333334e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.57369041442871, | |
| "logits/rejected": -18.57369041442871, | |
| "logps/chosen": -0.48697715997695923, | |
| "logps/rejected": -0.48697715997695923, | |
| "loss": 6.1125, | |
| "nll_loss": 1.4588209390640259, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.04869771748781204, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.04869771748781204, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.06147170493581631, | |
| "grad_norm": 0.5678962469100952, | |
| "learning_rate": 9.222222222222224e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.546266555786133, | |
| "logits/rejected": -18.546266555786133, | |
| "logps/chosen": -0.4713471531867981, | |
| "logps/rejected": -0.4713471531867981, | |
| "loss": 6.3675, | |
| "nll_loss": 1.5225670337677002, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.04713470861315727, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.04713470861315727, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.06508768757909962, | |
| "grad_norm": 0.9227916598320007, | |
| "learning_rate": 9.111111111111112e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.552295684814453, | |
| "logits/rejected": -18.552295684814453, | |
| "logps/chosen": -0.5135122537612915, | |
| "logps/rejected": -0.5135122537612915, | |
| "loss": 6.3925, | |
| "nll_loss": 1.5288182497024536, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.05135122686624527, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.05135122686624527, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.06870367022238293, | |
| "grad_norm": 1.3030140399932861, | |
| "learning_rate": 9e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.453067779541016, | |
| "logits/rejected": -18.453067779541016, | |
| "logps/chosen": -0.4881868362426758, | |
| "logps/rejected": -0.4881868362426758, | |
| "loss": 6.5908, | |
| "nll_loss": 1.5783830881118774, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.0488186851143837, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.0488186851143837, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.07231965286566625, | |
| "grad_norm": 1.0002509355545044, | |
| "learning_rate": 8.888888888888888e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.477296829223633, | |
| "logits/rejected": -18.477296829223633, | |
| "logps/chosen": -0.46165475249290466, | |
| "logps/rejected": -0.46165475249290466, | |
| "loss": 6.6221, | |
| "nll_loss": 1.5862023830413818, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.046165481209754944, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.046165481209754944, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.07593563550894955, | |
| "grad_norm": 0.7885683178901672, | |
| "learning_rate": 8.777777777777778e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.564895629882812, | |
| "logits/rejected": -18.564895629882812, | |
| "logps/chosen": -0.41595011949539185, | |
| "logps/rejected": -0.41595011949539185, | |
| "loss": 6.2789, | |
| "nll_loss": 1.500412940979004, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.04159501940011978, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.04159501940011978, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.07955161815223287, | |
| "grad_norm": 0.6379426717758179, | |
| "learning_rate": 8.666666666666668e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.577289581298828, | |
| "logits/rejected": -18.577289581298828, | |
| "logps/chosen": -0.43426617980003357, | |
| "logps/rejected": -0.43426617980003357, | |
| "loss": 6.4009, | |
| "nll_loss": 1.5309035778045654, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.043426621705293655, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.043426621705293655, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.08316760079551618, | |
| "grad_norm": 0.5890640020370483, | |
| "learning_rate": 8.555555555555556e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.817012786865234, | |
| "logits/rejected": -18.817012786865234, | |
| "logps/chosen": -0.38255172967910767, | |
| "logps/rejected": -0.38255172967910767, | |
| "loss": 6.021, | |
| "nll_loss": 1.4359278678894043, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.038255173712968826, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.038255173712968826, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.0867835834387995, | |
| "grad_norm": 0.5675965547561646, | |
| "learning_rate": 8.444444444444446e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.88204002380371, | |
| "logits/rejected": -18.88204002380371, | |
| "logps/chosen": -0.357022225856781, | |
| "logps/rejected": -0.357022225856781, | |
| "loss": 6.3796, | |
| "nll_loss": 1.5255934000015259, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.03570222482085228, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.03570222482085228, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.0903995660820828, | |
| "grad_norm": 0.284084677696228, | |
| "learning_rate": 8.344444444444445e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.659196853637695, | |
| "logits/rejected": -18.659196853637695, | |
| "logps/chosen": -0.35400137305259705, | |
| "logps/rejected": -0.35400137305259705, | |
| "loss": 6.1544, | |
| "nll_loss": 1.4692747592926025, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.03540014103055, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.03540014103055, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.09401554872536612, | |
| "grad_norm": 0.6424693465232849, | |
| "learning_rate": 8.233333333333335e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.762996673583984, | |
| "logits/rejected": -18.762996673583984, | |
| "logps/chosen": -0.40258026123046875, | |
| "logps/rejected": -0.40258026123046875, | |
| "loss": 6.3857, | |
| "nll_loss": 1.5271098613739014, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.040258027613162994, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.040258027613162994, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.09763153136864942, | |
| "grad_norm": 0.6682632565498352, | |
| "learning_rate": 8.122222222222223e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.719924926757812, | |
| "logits/rejected": -18.719924926757812, | |
| "logps/chosen": -0.3123939633369446, | |
| "logps/rejected": -0.3123939633369446, | |
| "loss": 5.7456, | |
| "nll_loss": 1.36708664894104, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.031239395961165428, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.031239395961165428, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.10124751401193274, | |
| "grad_norm": 0.9669603109359741, | |
| "learning_rate": 8.011111111111113e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.799579620361328, | |
| "logits/rejected": -18.799579620361328, | |
| "logps/chosen": -0.3817462623119354, | |
| "logps/rejected": -0.3817462623119354, | |
| "loss": 6.2838, | |
| "nll_loss": 1.5016276836395264, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.03817462921142578, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.03817462921142578, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.10486349665521605, | |
| "grad_norm": 0.7264003157615662, | |
| "learning_rate": 7.9e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -19.090185165405273, | |
| "logits/rejected": -19.090185165405273, | |
| "logps/chosen": -0.3294925093650818, | |
| "logps/rejected": -0.3294925093650818, | |
| "loss": 5.6699, | |
| "nll_loss": 1.3481695652008057, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.03294925019145012, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.03294925019145012, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.10847947929849937, | |
| "grad_norm": 0.564611554145813, | |
| "learning_rate": 7.788888888888889e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.767759323120117, | |
| "logits/rejected": -18.767759323120117, | |
| "logps/chosen": -0.26777949929237366, | |
| "logps/rejected": -0.26777949929237366, | |
| "loss": 6.5104, | |
| "nll_loss": 1.5582915544509888, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.026777952909469604, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.026777952909469604, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.11209546194178267, | |
| "grad_norm": 0.6952475905418396, | |
| "learning_rate": 7.677777777777778e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.896703720092773, | |
| "logits/rejected": -18.896703720092773, | |
| "logps/chosen": -0.24664482474327087, | |
| "logps/rejected": -0.24664482474327087, | |
| "loss": 5.9777, | |
| "nll_loss": 1.4251067638397217, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.02466448023915291, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.02466448023915291, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.11571144458506599, | |
| "grad_norm": 0.7094094753265381, | |
| "learning_rate": 7.566666666666667e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.993297576904297, | |
| "logits/rejected": -18.993297576904297, | |
| "logps/chosen": -0.27139216661453247, | |
| "logps/rejected": -0.27139216661453247, | |
| "loss": 5.848, | |
| "nll_loss": 1.3926928043365479, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.027139216661453247, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.027139216661453247, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.1193274272283493, | |
| "grad_norm": 0.8386672735214233, | |
| "learning_rate": 7.455555555555556e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.818639755249023, | |
| "logits/rejected": -18.818639755249023, | |
| "logps/chosen": -0.2643309533596039, | |
| "logps/rejected": -0.2643309533596039, | |
| "loss": 6.2247, | |
| "nll_loss": 1.4868563413619995, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.026433095335960388, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.026433095335960388, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.12294340987163262, | |
| "grad_norm": 0.6337667107582092, | |
| "learning_rate": 7.344444444444445e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -19.07388687133789, | |
| "logits/rejected": -19.07388687133789, | |
| "logps/chosen": -0.23179857432842255, | |
| "logps/rejected": -0.23179857432842255, | |
| "loss": 5.7871, | |
| "nll_loss": 1.3774592876434326, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.023179858922958374, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.023179858922958374, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.12655939251491594, | |
| "grad_norm": 1.056504726409912, | |
| "learning_rate": 7.233333333333334e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.8520450592041, | |
| "logits/rejected": -18.8520450592041, | |
| "logps/chosen": -0.24602051079273224, | |
| "logps/rejected": -0.24602051079273224, | |
| "loss": 5.8268, | |
| "nll_loss": 1.3873755931854248, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.024602051824331284, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.024602051824331284, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.13017537515819924, | |
| "grad_norm": 0.4462037682533264, | |
| "learning_rate": 7.122222222222222e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -19.04658317565918, | |
| "logits/rejected": -19.04658317565918, | |
| "logps/chosen": -0.2291136234998703, | |
| "logps/rejected": -0.2291136234998703, | |
| "loss": 5.8142, | |
| "nll_loss": 1.384232521057129, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.02291136048734188, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.02291136048734188, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.13379135780148255, | |
| "grad_norm": 0.5579663515090942, | |
| "learning_rate": 7.011111111111112e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -18.912765502929688, | |
| "logits/rejected": -18.912765502929688, | |
| "logps/chosen": -0.26528915762901306, | |
| "logps/rejected": -0.26528915762901306, | |
| "loss": 6.0316, | |
| "nll_loss": 1.4385900497436523, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.026528915390372276, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.026528915390372276, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.13740734044476585, | |
| "grad_norm": 1.2433960437774658, | |
| "learning_rate": 6.9e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -19.217700958251953, | |
| "logits/rejected": -19.217700958251953, | |
| "logps/chosen": -0.21606405079364777, | |
| "logps/rejected": -0.21606405079364777, | |
| "loss": 5.7944, | |
| "nll_loss": 1.3792803287506104, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.021606406196951866, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.021606406196951866, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.14102332308804919, | |
| "grad_norm": 0.7657280564308167, | |
| "learning_rate": 6.788888888888889e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -19.175457000732422, | |
| "logits/rejected": -19.175457000732422, | |
| "logps/chosen": -0.2413448989391327, | |
| "logps/rejected": -0.2413448989391327, | |
| "loss": 5.7839, | |
| "nll_loss": 1.3766554594039917, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.02413449063897133, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.02413449063897133, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.1446393057313325, | |
| "grad_norm": 0.7070457935333252, | |
| "learning_rate": 6.677777777777779e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -19.15035629272461, | |
| "logits/rejected": -19.15035629272461, | |
| "logps/chosen": -0.25096631050109863, | |
| "logps/rejected": -0.25096631050109863, | |
| "loss": 6.2307, | |
| "nll_loss": 1.4883568286895752, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.025096634402871132, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.025096634402871132, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.1482552883746158, | |
| "grad_norm": 0.43025216460227966, | |
| "learning_rate": 6.566666666666667e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -19.02223014831543, | |
| "logits/rejected": -19.02223014831543, | |
| "logps/chosen": -0.24732474982738495, | |
| "logps/rejected": -0.24732474982738495, | |
| "loss": 6.2865, | |
| "nll_loss": 1.5023012161254883, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.024732474237680435, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.024732474237680435, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.1518712710178991, | |
| "grad_norm": 1.3701528310775757, | |
| "learning_rate": 6.455555555555556e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -19.089460372924805, | |
| "logits/rejected": -19.089460372924805, | |
| "logps/chosen": -0.23749932646751404, | |
| "logps/rejected": -0.23749932646751404, | |
| "loss": 6.2238, | |
| "nll_loss": 1.4866377115249634, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.023749932646751404, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.023749932646751404, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.15548725366118243, | |
| "grad_norm": 0.9421939253807068, | |
| "learning_rate": 6.3444444444444454e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -19.231048583984375, | |
| "logits/rejected": -19.231048583984375, | |
| "logps/chosen": -0.19140982627868652, | |
| "logps/rejected": -0.19140982627868652, | |
| "loss": 6.0814, | |
| "nll_loss": 1.451047420501709, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.01914098486304283, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.01914098486304283, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.15910323630446574, | |
| "grad_norm": 0.9060840010643005, | |
| "learning_rate": 6.2333333333333335e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -19.107379913330078, | |
| "logits/rejected": -19.107379913330078, | |
| "logps/chosen": -0.2809773087501526, | |
| "logps/rejected": -0.2809773087501526, | |
| "loss": 6.3671, | |
| "nll_loss": 1.522457480430603, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.0280977301299572, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.0280977301299572, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.16271921894774904, | |
| "grad_norm": 0.5965518355369568, | |
| "learning_rate": 6.1222222222222224e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -19.34296417236328, | |
| "logits/rejected": -19.34296417236328, | |
| "logps/chosen": -0.19397786259651184, | |
| "logps/rejected": -0.19397786259651184, | |
| "loss": 5.7975, | |
| "nll_loss": 1.3800629377365112, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.019397784024477005, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.019397784024477005, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.16633520159103235, | |
| "grad_norm": 0.9895502328872681, | |
| "learning_rate": 6.011111111111112e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -19.35286521911621, | |
| "logits/rejected": -19.35286521911621, | |
| "logps/chosen": -0.20646443963050842, | |
| "logps/rejected": -0.20646443963050842, | |
| "loss": 5.8574, | |
| "nll_loss": 1.3950278759002686, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.02064644545316696, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.02064644545316696, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.16995118423431568, | |
| "grad_norm": 1.6251282691955566, | |
| "learning_rate": 5.9e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -19.29703140258789, | |
| "logits/rejected": -19.29703140258789, | |
| "logps/chosen": -0.23044386506080627, | |
| "logps/rejected": -0.23044386506080627, | |
| "loss": 6.0743, | |
| "nll_loss": 1.4492676258087158, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.023044386878609657, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.023044386878609657, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.173567166877599, | |
| "grad_norm": 0.7959076762199402, | |
| "learning_rate": 5.788888888888889e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -19.089004516601562, | |
| "logits/rejected": -19.089004516601562, | |
| "logps/chosen": -0.1713530272245407, | |
| "logps/rejected": -0.1713530272245407, | |
| "loss": 6.0034, | |
| "nll_loss": 1.4315412044525146, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.01713530346751213, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.01713530346751213, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.1771831495208823, | |
| "grad_norm": 0.6285834908485413, | |
| "learning_rate": 5.677777777777779e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -19.485876083374023, | |
| "logits/rejected": -19.485876083374023, | |
| "logps/chosen": -0.18361307680606842, | |
| "logps/rejected": -0.18361307680606842, | |
| "loss": 5.6858, | |
| "nll_loss": 1.3521363735198975, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.018361307680606842, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.018361307680606842, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.1807991321641656, | |
| "grad_norm": 0.9556539058685303, | |
| "learning_rate": 5.566666666666667e-06, | |
| "log_odds_chosen": 0.0, | |
| "log_odds_ratio": -0.6931472420692444, | |
| "logits/chosen": -19.300710678100586, | |
| "logits/rejected": -19.300710678100586, | |
| "logps/chosen": -0.17280462384223938, | |
| "logps/rejected": -0.17280462384223938, | |
| "loss": 6.0335, | |
| "nll_loss": 1.439051866531372, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": -0.01728046126663685, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": -0.01728046126663685, | |
| "step": 500 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1000, | |
| "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": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |