| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.997207114508305, | |
| "eval_steps": 1, | |
| "global_step": 850, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.023519035719535498, | |
| "grad_norm": 4.1415791511535645, | |
| "learning_rate": 3.060971262053483e-06, | |
| "logits/chosen": -4.5, | |
| "logits/rejected": -4.5, | |
| "logps/chosen": -624.0, | |
| "logps/rejected": -452.0, | |
| "loss": 0.6933, | |
| "rewards/accuracies": 0.08749999850988388, | |
| "rewards/chosen": -0.007080078125, | |
| "rewards/margins": -0.0089111328125, | |
| "rewards/rejected": 0.00183868408203125, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.047038071439070996, | |
| "grad_norm": 3.7295737266540527, | |
| "learning_rate": 3.9824154277970135e-06, | |
| "logits/chosen": -4.53125, | |
| "logits/rejected": -4.46875, | |
| "logps/chosen": -552.0, | |
| "logps/rejected": -422.0, | |
| "loss": 0.6766, | |
| "rewards/accuracies": 0.296875, | |
| "rewards/chosen": 0.0263671875, | |
| "rewards/margins": 0.044677734375, | |
| "rewards/rejected": -0.018310546875, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0705571071586065, | |
| "grad_norm": 3.65460205078125, | |
| "learning_rate": 4.521425711265269e-06, | |
| "logits/chosen": -4.46875, | |
| "logits/rejected": -4.40625, | |
| "logps/chosen": -592.0, | |
| "logps/rejected": -436.0, | |
| "loss": 0.6273, | |
| "rewards/accuracies": 0.5718749761581421, | |
| "rewards/chosen": 0.09423828125, | |
| "rewards/margins": 0.158203125, | |
| "rewards/rejected": -0.0634765625, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.09407614287814199, | |
| "grad_norm": 3.148322343826294, | |
| "learning_rate": 4.903859593540544e-06, | |
| "logits/chosen": -4.4375, | |
| "logits/rejected": -4.40625, | |
| "logps/chosen": -596.0, | |
| "logps/rejected": -456.0, | |
| "loss": 0.5544, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.185546875, | |
| "rewards/margins": 0.35546875, | |
| "rewards/rejected": -0.169921875, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1175951785976775, | |
| "grad_norm": 4.775664806365967, | |
| "learning_rate": 4.9628252788104095e-06, | |
| "logits/chosen": -4.4375, | |
| "logits/rejected": -4.4375, | |
| "logps/chosen": -576.0, | |
| "logps/rejected": -444.0, | |
| "loss": 0.4791, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": 0.2578125, | |
| "rewards/margins": 0.55859375, | |
| "rewards/rejected": -0.30078125, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.141114214317213, | |
| "grad_norm": 2.693018913269043, | |
| "learning_rate": 4.9008674101610905e-06, | |
| "logits/chosen": -4.5, | |
| "logits/rejected": -4.5, | |
| "logps/chosen": -584.0, | |
| "logps/rejected": -444.0, | |
| "loss": 0.4168, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": 0.39453125, | |
| "rewards/margins": 0.87890625, | |
| "rewards/rejected": -0.484375, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1646332500367485, | |
| "grad_norm": 2.0642476081848145, | |
| "learning_rate": 4.838909541511772e-06, | |
| "logits/chosen": -4.4375, | |
| "logits/rejected": -4.40625, | |
| "logps/chosen": -596.0, | |
| "logps/rejected": -454.0, | |
| "loss": 0.3891, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": 0.388671875, | |
| "rewards/margins": 1.0078125, | |
| "rewards/rejected": -0.625, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.18815228575628398, | |
| "grad_norm": 2.0007693767547607, | |
| "learning_rate": 4.776951672862453e-06, | |
| "logits/chosen": -4.53125, | |
| "logits/rejected": -4.5, | |
| "logps/chosen": -600.0, | |
| "logps/rejected": -492.0, | |
| "loss": 0.3605, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.49609375, | |
| "rewards/margins": 1.3515625, | |
| "rewards/rejected": -0.8515625, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2116713214758195, | |
| "grad_norm": 2.3289802074432373, | |
| "learning_rate": 4.714993804213135e-06, | |
| "logits/chosen": -4.4375, | |
| "logits/rejected": -4.4375, | |
| "logps/chosen": -584.0, | |
| "logps/rejected": -468.0, | |
| "loss": 0.3619, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": 0.396484375, | |
| "rewards/margins": 1.34375, | |
| "rewards/rejected": -0.9453125, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.235190357195355, | |
| "grad_norm": 2.2528398036956787, | |
| "learning_rate": 4.653035935563817e-06, | |
| "logits/chosen": -4.40625, | |
| "logits/rejected": -4.46875, | |
| "logps/chosen": -620.0, | |
| "logps/rejected": -468.0, | |
| "loss": 0.3334, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": 0.59765625, | |
| "rewards/margins": 1.609375, | |
| "rewards/rejected": -1.0078125, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2587093929148905, | |
| "grad_norm": 1.8650789260864258, | |
| "learning_rate": 4.591078066914498e-06, | |
| "logits/chosen": -4.40625, | |
| "logits/rejected": -4.375, | |
| "logps/chosen": -596.0, | |
| "logps/rejected": -416.0, | |
| "loss": 0.2843, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": 0.55859375, | |
| "rewards/margins": 1.8125, | |
| "rewards/rejected": -1.25, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.282228428634426, | |
| "grad_norm": 2.2233972549438477, | |
| "learning_rate": 4.52912019826518e-06, | |
| "logits/chosen": -4.40625, | |
| "logits/rejected": -4.4375, | |
| "logps/chosen": -596.0, | |
| "logps/rejected": -476.0, | |
| "loss": 0.2863, | |
| "rewards/accuracies": 0.8812500238418579, | |
| "rewards/chosen": 0.63671875, | |
| "rewards/margins": 1.8984375, | |
| "rewards/rejected": -1.2578125, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.3057474643539615, | |
| "grad_norm": 2.017451286315918, | |
| "learning_rate": 4.467162329615862e-06, | |
| "logits/chosen": -4.46875, | |
| "logits/rejected": -4.5, | |
| "logps/chosen": -612.0, | |
| "logps/rejected": -448.0, | |
| "loss": 0.2741, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": 0.64453125, | |
| "rewards/margins": 1.9453125, | |
| "rewards/rejected": -1.3046875, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.329266500073497, | |
| "grad_norm": 2.127619743347168, | |
| "learning_rate": 4.405204460966543e-06, | |
| "logits/chosen": -4.40625, | |
| "logits/rejected": -4.40625, | |
| "logps/chosen": -616.0, | |
| "logps/rejected": -454.0, | |
| "loss": 0.2705, | |
| "rewards/accuracies": 0.8843749761581421, | |
| "rewards/chosen": 0.8828125, | |
| "rewards/margins": 2.265625, | |
| "rewards/rejected": -1.375, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3527855357930325, | |
| "grad_norm": 1.8466318845748901, | |
| "learning_rate": 4.343246592317225e-06, | |
| "logits/chosen": -4.46875, | |
| "logits/rejected": -4.40625, | |
| "logps/chosen": -624.0, | |
| "logps/rejected": -492.0, | |
| "loss": 0.2757, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": 0.84375, | |
| "rewards/margins": 2.28125, | |
| "rewards/rejected": -1.4375, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.37630457151256796, | |
| "grad_norm": 1.8905202150344849, | |
| "learning_rate": 4.2812887236679065e-06, | |
| "logits/chosen": -4.40625, | |
| "logits/rejected": -4.375, | |
| "logps/chosen": -556.0, | |
| "logps/rejected": -450.0, | |
| "loss": 0.2506, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": 0.94140625, | |
| "rewards/margins": 2.328125, | |
| "rewards/rejected": -1.390625, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3998236072321035, | |
| "grad_norm": 1.59074866771698, | |
| "learning_rate": 4.219330855018588e-06, | |
| "logits/chosen": -4.53125, | |
| "logits/rejected": -4.4375, | |
| "logps/chosen": -576.0, | |
| "logps/rejected": -452.0, | |
| "loss": 0.2633, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": 0.9453125, | |
| "rewards/margins": 2.328125, | |
| "rewards/rejected": -1.390625, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.423342642951639, | |
| "grad_norm": 1.7788141965866089, | |
| "learning_rate": 4.157372986369269e-06, | |
| "logits/chosen": -4.4375, | |
| "logits/rejected": -4.4375, | |
| "logps/chosen": -600.0, | |
| "logps/rejected": -500.0, | |
| "loss": 0.2619, | |
| "rewards/accuracies": 0.859375, | |
| "rewards/chosen": 0.9765625, | |
| "rewards/margins": 2.375, | |
| "rewards/rejected": -1.3984375, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.4468616786711745, | |
| "grad_norm": 2.1362152099609375, | |
| "learning_rate": 4.095415117719951e-06, | |
| "logits/chosen": -4.4375, | |
| "logits/rejected": -4.46875, | |
| "logps/chosen": -592.0, | |
| "logps/rejected": -472.0, | |
| "loss": 0.2743, | |
| "rewards/accuracies": 0.918749988079071, | |
| "rewards/chosen": 1.09375, | |
| "rewards/margins": 2.578125, | |
| "rewards/rejected": -1.4921875, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.47038071439071, | |
| "grad_norm": 1.8583593368530273, | |
| "learning_rate": 4.033457249070632e-06, | |
| "logits/chosen": -4.46875, | |
| "logits/rejected": -4.40625, | |
| "logps/chosen": -588.0, | |
| "logps/rejected": -460.0, | |
| "loss": 0.2359, | |
| "rewards/accuracies": 0.9125000238418579, | |
| "rewards/chosen": 1.2265625, | |
| "rewards/margins": 2.640625, | |
| "rewards/rejected": -1.4140625, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4938997501102455, | |
| "grad_norm": 2.6427791118621826, | |
| "learning_rate": 3.971499380421314e-06, | |
| "logits/chosen": -4.46875, | |
| "logits/rejected": -4.5, | |
| "logps/chosen": -568.0, | |
| "logps/rejected": -456.0, | |
| "loss": 0.2418, | |
| "rewards/accuracies": 0.8968750238418579, | |
| "rewards/chosen": 1.2578125, | |
| "rewards/margins": 2.4375, | |
| "rewards/rejected": -1.1796875, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.517418785829781, | |
| "grad_norm": 1.6700748205184937, | |
| "learning_rate": 3.909541511771995e-06, | |
| "logits/chosen": -4.4375, | |
| "logits/rejected": -4.4375, | |
| "logps/chosen": -580.0, | |
| "logps/rejected": -442.0, | |
| "loss": 0.2332, | |
| "rewards/accuracies": 0.8968750238418579, | |
| "rewards/chosen": 1.140625, | |
| "rewards/margins": 2.8125, | |
| "rewards/rejected": -1.671875, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.5409378215493165, | |
| "grad_norm": 2.4413912296295166, | |
| "learning_rate": 3.847583643122677e-06, | |
| "logits/chosen": -4.5, | |
| "logits/rejected": -4.5, | |
| "logps/chosen": -612.0, | |
| "logps/rejected": -484.0, | |
| "loss": 0.2338, | |
| "rewards/accuracies": 0.9125000238418579, | |
| "rewards/chosen": 1.3125, | |
| "rewards/margins": 2.765625, | |
| "rewards/rejected": -1.453125, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.564456857268852, | |
| "grad_norm": 1.718529462814331, | |
| "learning_rate": 3.7856257744733583e-06, | |
| "logits/chosen": -4.46875, | |
| "logits/rejected": -4.46875, | |
| "logps/chosen": -592.0, | |
| "logps/rejected": -432.0, | |
| "loss": 0.2044, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 1.5546875, | |
| "rewards/margins": 3.09375, | |
| "rewards/rejected": -1.5390625, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5879758929883875, | |
| "grad_norm": 2.0037829875946045, | |
| "learning_rate": 3.7236679058240398e-06, | |
| "logits/chosen": -4.4375, | |
| "logits/rejected": -4.40625, | |
| "logps/chosen": -584.0, | |
| "logps/rejected": -458.0, | |
| "loss": 0.221, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": 1.375, | |
| "rewards/margins": 3.03125, | |
| "rewards/rejected": -1.6484375, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.611494928707923, | |
| "grad_norm": 1.673599362373352, | |
| "learning_rate": 3.6617100371747216e-06, | |
| "logits/chosen": -4.53125, | |
| "logits/rejected": -4.5, | |
| "logps/chosen": -564.0, | |
| "logps/rejected": -468.0, | |
| "loss": 0.2305, | |
| "rewards/accuracies": 0.9125000238418579, | |
| "rewards/chosen": 1.453125, | |
| "rewards/margins": 2.96875, | |
| "rewards/rejected": -1.515625, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.6350139644274585, | |
| "grad_norm": 1.8723180294036865, | |
| "learning_rate": 3.5997521685254026e-06, | |
| "logits/chosen": -4.46875, | |
| "logits/rejected": -4.46875, | |
| "logps/chosen": -564.0, | |
| "logps/rejected": -430.0, | |
| "loss": 0.2233, | |
| "rewards/accuracies": 0.921875, | |
| "rewards/chosen": 1.4296875, | |
| "rewards/margins": 3.0625, | |
| "rewards/rejected": -1.6328125, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.658533000146994, | |
| "grad_norm": 1.8895870447158813, | |
| "learning_rate": 3.5377942998760845e-06, | |
| "logits/chosen": -4.4375, | |
| "logits/rejected": -4.4375, | |
| "logps/chosen": -584.0, | |
| "logps/rejected": -468.0, | |
| "loss": 0.2191, | |
| "rewards/accuracies": 0.909375011920929, | |
| "rewards/chosen": 1.5234375, | |
| "rewards/margins": 3.21875, | |
| "rewards/rejected": -1.6953125, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6820520358665295, | |
| "grad_norm": 1.5127581357955933, | |
| "learning_rate": 3.4758364312267663e-06, | |
| "logits/chosen": -4.40625, | |
| "logits/rejected": -4.40625, | |
| "logps/chosen": -596.0, | |
| "logps/rejected": -450.0, | |
| "loss": 0.213, | |
| "rewards/accuracies": 0.9125000238418579, | |
| "rewards/chosen": 1.4609375, | |
| "rewards/margins": 2.9375, | |
| "rewards/rejected": -1.46875, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.705571071586065, | |
| "grad_norm": 2.202705144882202, | |
| "learning_rate": 3.4138785625774478e-06, | |
| "logits/chosen": -4.46875, | |
| "logits/rejected": -4.40625, | |
| "logps/chosen": -576.0, | |
| "logps/rejected": -468.0, | |
| "loss": 0.2195, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": 1.578125, | |
| "rewards/margins": 3.0, | |
| "rewards/rejected": -1.421875, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7290901073056004, | |
| "grad_norm": 1.8141247034072876, | |
| "learning_rate": 3.351920693928129e-06, | |
| "logits/chosen": -4.40625, | |
| "logits/rejected": -4.4375, | |
| "logps/chosen": -560.0, | |
| "logps/rejected": -436.0, | |
| "loss": 0.2075, | |
| "rewards/accuracies": 0.9156249761581421, | |
| "rewards/chosen": 1.578125, | |
| "rewards/margins": 3.1875, | |
| "rewards/rejected": -1.609375, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.7526091430251359, | |
| "grad_norm": 1.7854108810424805, | |
| "learning_rate": 3.2899628252788106e-06, | |
| "logits/chosen": -4.34375, | |
| "logits/rejected": -4.375, | |
| "logps/chosen": -616.0, | |
| "logps/rejected": -462.0, | |
| "loss": 0.2124, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": 1.5859375, | |
| "rewards/margins": 3.203125, | |
| "rewards/rejected": -1.6171875, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.7761281787446714, | |
| "grad_norm": 1.5797982215881348, | |
| "learning_rate": 3.2280049566294925e-06, | |
| "logits/chosen": -4.4375, | |
| "logits/rejected": -4.46875, | |
| "logps/chosen": -620.0, | |
| "logps/rejected": -496.0, | |
| "loss": 0.2276, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/chosen": 1.7734375, | |
| "rewards/margins": 3.515625, | |
| "rewards/rejected": -1.7421875, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.799647214464207, | |
| "grad_norm": 1.9662494659423828, | |
| "learning_rate": 3.1660470879801735e-06, | |
| "logits/chosen": -4.46875, | |
| "logits/rejected": -4.4375, | |
| "logps/chosen": -568.0, | |
| "logps/rejected": -456.0, | |
| "loss": 0.2161, | |
| "rewards/accuracies": 0.903124988079071, | |
| "rewards/chosen": 1.65625, | |
| "rewards/margins": 3.0625, | |
| "rewards/rejected": -1.3984375, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.8231662501837425, | |
| "grad_norm": 1.7154417037963867, | |
| "learning_rate": 3.1040892193308553e-06, | |
| "logits/chosen": -4.4375, | |
| "logits/rejected": -4.4375, | |
| "logps/chosen": -572.0, | |
| "logps/rejected": -462.0, | |
| "loss": 0.1925, | |
| "rewards/accuracies": 0.909375011920929, | |
| "rewards/chosen": 1.6640625, | |
| "rewards/margins": 3.15625, | |
| "rewards/rejected": -1.4921875, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.846685285903278, | |
| "grad_norm": 2.1363542079925537, | |
| "learning_rate": 3.042131350681537e-06, | |
| "logits/chosen": -4.40625, | |
| "logits/rejected": -4.375, | |
| "logps/chosen": -520.0, | |
| "logps/rejected": -410.0, | |
| "loss": 0.225, | |
| "rewards/accuracies": 0.9156249761581421, | |
| "rewards/chosen": 1.4609375, | |
| "rewards/margins": 3.171875, | |
| "rewards/rejected": -1.703125, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.8702043216228135, | |
| "grad_norm": 1.9305567741394043, | |
| "learning_rate": 2.980173482032218e-06, | |
| "logits/chosen": -4.5, | |
| "logits/rejected": -4.5, | |
| "logps/chosen": -572.0, | |
| "logps/rejected": -452.0, | |
| "loss": 0.1921, | |
| "rewards/accuracies": 0.918749988079071, | |
| "rewards/chosen": 1.671875, | |
| "rewards/margins": 3.453125, | |
| "rewards/rejected": -1.7890625, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.893723357342349, | |
| "grad_norm": 1.7296861410140991, | |
| "learning_rate": 2.9182156133829e-06, | |
| "logits/chosen": -4.46875, | |
| "logits/rejected": -4.5, | |
| "logps/chosen": -592.0, | |
| "logps/rejected": -450.0, | |
| "loss": 0.2142, | |
| "rewards/accuracies": 0.918749988079071, | |
| "rewards/chosen": 1.59375, | |
| "rewards/margins": 3.296875, | |
| "rewards/rejected": -1.703125, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.9172423930618845, | |
| "grad_norm": 1.9843934774398804, | |
| "learning_rate": 2.8562577447335815e-06, | |
| "logits/chosen": -4.4375, | |
| "logits/rejected": -4.46875, | |
| "logps/chosen": -576.0, | |
| "logps/rejected": -426.0, | |
| "loss": 0.213, | |
| "rewards/accuracies": 0.9281250238418579, | |
| "rewards/chosen": 1.625, | |
| "rewards/margins": 3.390625, | |
| "rewards/rejected": -1.7578125, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.94076142878142, | |
| "grad_norm": 2.4459681510925293, | |
| "learning_rate": 2.7942998760842625e-06, | |
| "logits/chosen": -4.375, | |
| "logits/rejected": -4.34375, | |
| "logps/chosen": -576.0, | |
| "logps/rejected": -496.0, | |
| "loss": 0.1966, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/chosen": 1.75, | |
| "rewards/margins": 3.515625, | |
| "rewards/rejected": -1.765625, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.9642804645009555, | |
| "grad_norm": 2.3723385334014893, | |
| "learning_rate": 2.7323420074349443e-06, | |
| "logits/chosen": -4.40625, | |
| "logits/rejected": -4.40625, | |
| "logps/chosen": -556.0, | |
| "logps/rejected": -452.0, | |
| "loss": 0.2096, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": 1.5703125, | |
| "rewards/margins": 3.234375, | |
| "rewards/rejected": -1.671875, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.987799500220491, | |
| "grad_norm": 2.0075254440307617, | |
| "learning_rate": 2.670384138785626e-06, | |
| "logits/chosen": -4.375, | |
| "logits/rejected": -4.40625, | |
| "logps/chosen": -588.0, | |
| "logps/rejected": -492.0, | |
| "loss": 0.219, | |
| "rewards/accuracies": 0.890625, | |
| "rewards/chosen": 1.7265625, | |
| "rewards/margins": 3.4375, | |
| "rewards/rejected": -1.7109375, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.0094076142878141, | |
| "grad_norm": 0.9782338738441467, | |
| "learning_rate": 2.6146220570012394e-06, | |
| "logits/chosen": -4.4375, | |
| "logits/rejected": -4.40625, | |
| "logps/chosen": -576.0, | |
| "logps/rejected": -480.0, | |
| "loss": 0.1502, | |
| "rewards/accuracies": 0.9319728016853333, | |
| "rewards/chosen": 1.78125, | |
| "rewards/margins": 3.625, | |
| "rewards/rejected": -1.84375, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.0329266500073497, | |
| "grad_norm": 0.8791906833648682, | |
| "learning_rate": 2.552664188351921e-06, | |
| "logits/chosen": -4.4375, | |
| "logits/rejected": -4.5, | |
| "logps/chosen": -568.0, | |
| "logps/rejected": -436.0, | |
| "loss": 0.1162, | |
| "rewards/accuracies": 0.971875011920929, | |
| "rewards/chosen": 1.890625, | |
| "rewards/margins": 4.125, | |
| "rewards/rejected": -2.25, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.0564456857268851, | |
| "grad_norm": 0.9551910161972046, | |
| "learning_rate": 2.4907063197026023e-06, | |
| "logits/chosen": -4.5, | |
| "logits/rejected": -4.5, | |
| "logps/chosen": -568.0, | |
| "logps/rejected": -434.0, | |
| "loss": 0.1221, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": 1.8359375, | |
| "rewards/margins": 3.84375, | |
| "rewards/rejected": -2.015625, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.0799647214464208, | |
| "grad_norm": 1.0174229145050049, | |
| "learning_rate": 2.4287484510532837e-06, | |
| "logits/chosen": -4.5, | |
| "logits/rejected": -4.53125, | |
| "logps/chosen": -576.0, | |
| "logps/rejected": -446.0, | |
| "loss": 0.122, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": 1.953125, | |
| "rewards/margins": 4.0, | |
| "rewards/rejected": -2.046875, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.1034837571659561, | |
| "grad_norm": 1.0029239654541016, | |
| "learning_rate": 2.3667905824039656e-06, | |
| "logits/chosen": -4.4375, | |
| "logits/rejected": -4.46875, | |
| "logps/chosen": -592.0, | |
| "logps/rejected": -458.0, | |
| "loss": 0.1126, | |
| "rewards/accuracies": 0.971875011920929, | |
| "rewards/chosen": 1.921875, | |
| "rewards/margins": 4.21875, | |
| "rewards/rejected": -2.296875, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.1270027928854918, | |
| "grad_norm": 1.4849553108215332, | |
| "learning_rate": 2.304832713754647e-06, | |
| "logits/chosen": -4.4375, | |
| "logits/rejected": -4.375, | |
| "logps/chosen": -600.0, | |
| "logps/rejected": -504.0, | |
| "loss": 0.1161, | |
| "rewards/accuracies": 0.9781249761581421, | |
| "rewards/chosen": 2.09375, | |
| "rewards/margins": 4.40625, | |
| "rewards/rejected": -2.3125, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.1505218286050272, | |
| "grad_norm": 0.9113921523094177, | |
| "learning_rate": 2.242874845105329e-06, | |
| "logits/chosen": -4.46875, | |
| "logits/rejected": -4.40625, | |
| "logps/chosen": -584.0, | |
| "logps/rejected": -474.0, | |
| "loss": 0.1067, | |
| "rewards/accuracies": 0.965624988079071, | |
| "rewards/chosen": 1.921875, | |
| "rewards/margins": 4.125, | |
| "rewards/rejected": -2.203125, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.1740408643245628, | |
| "grad_norm": 1.4038139581680298, | |
| "learning_rate": 2.1809169764560103e-06, | |
| "logits/chosen": -4.4375, | |
| "logits/rejected": -4.40625, | |
| "logps/chosen": -596.0, | |
| "logps/rejected": -450.0, | |
| "loss": 0.1208, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": 1.9375, | |
| "rewards/margins": 4.21875, | |
| "rewards/rejected": -2.265625, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.1975599000440982, | |
| "grad_norm": 0.9754643440246582, | |
| "learning_rate": 2.1189591078066917e-06, | |
| "logits/chosen": -4.40625, | |
| "logits/rejected": -4.375, | |
| "logps/chosen": -584.0, | |
| "logps/rejected": -456.0, | |
| "loss": 0.1145, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": 1.8515625, | |
| "rewards/margins": 4.0625, | |
| "rewards/rejected": -2.203125, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.2210789357636336, | |
| "grad_norm": 0.7279557585716248, | |
| "learning_rate": 2.057001239157373e-06, | |
| "logits/chosen": -4.40625, | |
| "logits/rejected": -4.375, | |
| "logps/chosen": -588.0, | |
| "logps/rejected": -460.0, | |
| "loss": 0.111, | |
| "rewards/accuracies": 0.984375, | |
| "rewards/chosen": 1.765625, | |
| "rewards/margins": 4.21875, | |
| "rewards/rejected": -2.46875, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.2445979714831692, | |
| "grad_norm": 1.1293213367462158, | |
| "learning_rate": 1.9950433705080545e-06, | |
| "logits/chosen": -4.4375, | |
| "logits/rejected": -4.46875, | |
| "logps/chosen": -572.0, | |
| "logps/rejected": -428.0, | |
| "loss": 0.1192, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": 1.9921875, | |
| "rewards/margins": 4.28125, | |
| "rewards/rejected": -2.296875, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.2681170072027048, | |
| "grad_norm": 0.7407346963882446, | |
| "learning_rate": 1.9330855018587364e-06, | |
| "logits/chosen": -4.46875, | |
| "logits/rejected": -4.40625, | |
| "logps/chosen": -596.0, | |
| "logps/rejected": -456.0, | |
| "loss": 0.1107, | |
| "rewards/accuracies": 0.965624988079071, | |
| "rewards/chosen": 2.015625, | |
| "rewards/margins": 4.25, | |
| "rewards/rejected": -2.25, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.2916360429222402, | |
| "grad_norm": 1.1689890623092651, | |
| "learning_rate": 1.8711276332094178e-06, | |
| "logits/chosen": -4.46875, | |
| "logits/rejected": -4.4375, | |
| "logps/chosen": -600.0, | |
| "logps/rejected": -472.0, | |
| "loss": 0.116, | |
| "rewards/accuracies": 0.9468749761581421, | |
| "rewards/chosen": 1.828125, | |
| "rewards/margins": 4.21875, | |
| "rewards/rejected": -2.375, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.3151550786417756, | |
| "grad_norm": 1.0410897731781006, | |
| "learning_rate": 1.8091697645600993e-06, | |
| "logits/chosen": -4.46875, | |
| "logits/rejected": -4.46875, | |
| "logps/chosen": -584.0, | |
| "logps/rejected": -436.0, | |
| "loss": 0.1166, | |
| "rewards/accuracies": 0.965624988079071, | |
| "rewards/chosen": 1.9375, | |
| "rewards/margins": 4.34375, | |
| "rewards/rejected": -2.40625, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.3386741143613112, | |
| "grad_norm": 0.8815547227859497, | |
| "learning_rate": 1.7472118959107809e-06, | |
| "logits/chosen": -4.4375, | |
| "logits/rejected": -4.40625, | |
| "logps/chosen": -580.0, | |
| "logps/rejected": -482.0, | |
| "loss": 0.1003, | |
| "rewards/accuracies": 0.953125, | |
| "rewards/chosen": 1.96875, | |
| "rewards/margins": 4.25, | |
| "rewards/rejected": -2.28125, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.3621931500808466, | |
| "grad_norm": 1.0066012144088745, | |
| "learning_rate": 1.6852540272614623e-06, | |
| "logits/chosen": -4.4375, | |
| "logits/rejected": -4.40625, | |
| "logps/chosen": -576.0, | |
| "logps/rejected": -516.0, | |
| "loss": 0.1079, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": 1.8671875, | |
| "rewards/margins": 4.03125, | |
| "rewards/rejected": -2.15625, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.3857121858003822, | |
| "grad_norm": 1.2409735918045044, | |
| "learning_rate": 1.623296158612144e-06, | |
| "logits/chosen": -4.4375, | |
| "logits/rejected": -4.46875, | |
| "logps/chosen": -600.0, | |
| "logps/rejected": -464.0, | |
| "loss": 0.1281, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": 1.890625, | |
| "rewards/margins": 4.71875, | |
| "rewards/rejected": -2.828125, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.4092312215199176, | |
| "grad_norm": 1.059398889541626, | |
| "learning_rate": 1.5613382899628254e-06, | |
| "logits/chosen": -4.5, | |
| "logits/rejected": -4.5, | |
| "logps/chosen": -568.0, | |
| "logps/rejected": -452.0, | |
| "loss": 0.1181, | |
| "rewards/accuracies": 0.953125, | |
| "rewards/chosen": 1.921875, | |
| "rewards/margins": 4.34375, | |
| "rewards/rejected": -2.421875, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.4327502572394533, | |
| "grad_norm": 0.8506196141242981, | |
| "learning_rate": 1.4993804213135068e-06, | |
| "logits/chosen": -4.4375, | |
| "logits/rejected": -4.4375, | |
| "logps/chosen": -576.0, | |
| "logps/rejected": -450.0, | |
| "loss": 0.1021, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": 1.921875, | |
| "rewards/margins": 4.40625, | |
| "rewards/rejected": -2.484375, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.4562692929589887, | |
| "grad_norm": 1.0724049806594849, | |
| "learning_rate": 1.4374225526641887e-06, | |
| "logits/chosen": -4.46875, | |
| "logits/rejected": -4.5, | |
| "logps/chosen": -576.0, | |
| "logps/rejected": -462.0, | |
| "loss": 0.1089, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": 1.796875, | |
| "rewards/margins": 4.21875, | |
| "rewards/rejected": -2.40625, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.479788328678524, | |
| "grad_norm": 0.8412142992019653, | |
| "learning_rate": 1.37546468401487e-06, | |
| "logits/chosen": -4.40625, | |
| "logits/rejected": -4.4375, | |
| "logps/chosen": -576.0, | |
| "logps/rejected": -440.0, | |
| "loss": 0.1178, | |
| "rewards/accuracies": 0.953125, | |
| "rewards/chosen": 1.9140625, | |
| "rewards/margins": 4.09375, | |
| "rewards/rejected": -2.1875, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.5033073643980597, | |
| "grad_norm": 1.080023169517517, | |
| "learning_rate": 1.3135068153655513e-06, | |
| "logits/chosen": -4.40625, | |
| "logits/rejected": -4.4375, | |
| "logps/chosen": -600.0, | |
| "logps/rejected": -452.0, | |
| "loss": 0.1097, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": 2.0, | |
| "rewards/margins": 4.5625, | |
| "rewards/rejected": -2.546875, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.5268264001175953, | |
| "grad_norm": 1.2255281209945679, | |
| "learning_rate": 1.2515489467162332e-06, | |
| "logits/chosen": -4.375, | |
| "logits/rejected": -4.40625, | |
| "logps/chosen": -572.0, | |
| "logps/rejected": -456.0, | |
| "loss": 0.1078, | |
| "rewards/accuracies": 0.971875011920929, | |
| "rewards/chosen": 1.96875, | |
| "rewards/margins": 4.40625, | |
| "rewards/rejected": -2.421875, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.5503454358371307, | |
| "grad_norm": 0.9179627299308777, | |
| "learning_rate": 1.1895910780669146e-06, | |
| "logits/chosen": -4.40625, | |
| "logits/rejected": -4.4375, | |
| "logps/chosen": -568.0, | |
| "logps/rejected": -444.0, | |
| "loss": 0.1105, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": 1.890625, | |
| "rewards/margins": 4.28125, | |
| "rewards/rejected": -2.40625, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.573864471556666, | |
| "grad_norm": 1.0199528932571411, | |
| "learning_rate": 1.127633209417596e-06, | |
| "logits/chosen": -4.46875, | |
| "logits/rejected": -4.46875, | |
| "logps/chosen": -548.0, | |
| "logps/rejected": -450.0, | |
| "loss": 0.1132, | |
| "rewards/accuracies": 0.965624988079071, | |
| "rewards/chosen": 1.8984375, | |
| "rewards/margins": 4.21875, | |
| "rewards/rejected": -2.3125, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.5973835072762017, | |
| "grad_norm": 1.190191626548767, | |
| "learning_rate": 1.0656753407682777e-06, | |
| "logits/chosen": -4.5, | |
| "logits/rejected": -4.53125, | |
| "logps/chosen": -596.0, | |
| "logps/rejected": -460.0, | |
| "loss": 0.1178, | |
| "rewards/accuracies": 0.965624988079071, | |
| "rewards/chosen": 1.9375, | |
| "rewards/margins": 4.3125, | |
| "rewards/rejected": -2.359375, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.6209025429957373, | |
| "grad_norm": 1.029616355895996, | |
| "learning_rate": 1.0037174721189593e-06, | |
| "logits/chosen": -4.5625, | |
| "logits/rejected": -4.59375, | |
| "logps/chosen": -552.0, | |
| "logps/rejected": -450.0, | |
| "loss": 0.1087, | |
| "rewards/accuracies": 0.9593750238418579, | |
| "rewards/chosen": 1.8515625, | |
| "rewards/margins": 4.375, | |
| "rewards/rejected": -2.515625, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.6444215787152727, | |
| "grad_norm": 1.2294584512710571, | |
| "learning_rate": 9.417596034696406e-07, | |
| "logits/chosen": -4.40625, | |
| "logits/rejected": -4.40625, | |
| "logps/chosen": -576.0, | |
| "logps/rejected": -444.0, | |
| "loss": 0.1049, | |
| "rewards/accuracies": 0.984375, | |
| "rewards/chosen": 1.8671875, | |
| "rewards/margins": 4.21875, | |
| "rewards/rejected": -2.34375, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.6679406144348081, | |
| "grad_norm": 1.0543959140777588, | |
| "learning_rate": 8.798017348203223e-07, | |
| "logits/chosen": -4.4375, | |
| "logits/rejected": -4.5, | |
| "logps/chosen": -580.0, | |
| "logps/rejected": -458.0, | |
| "loss": 0.1063, | |
| "rewards/accuracies": 0.965624988079071, | |
| "rewards/chosen": 1.9609375, | |
| "rewards/margins": 4.40625, | |
| "rewards/rejected": -2.46875, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.6914596501543437, | |
| "grad_norm": 1.0166019201278687, | |
| "learning_rate": 8.178438661710038e-07, | |
| "logits/chosen": -4.4375, | |
| "logits/rejected": -4.5, | |
| "logps/chosen": -560.0, | |
| "logps/rejected": -448.0, | |
| "loss": 0.1014, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": 1.8828125, | |
| "rewards/margins": 4.4375, | |
| "rewards/rejected": -2.5625, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.7149786858738791, | |
| "grad_norm": 1.0257729291915894, | |
| "learning_rate": 7.558859975216853e-07, | |
| "logits/chosen": -4.40625, | |
| "logits/rejected": -4.375, | |
| "logps/chosen": -604.0, | |
| "logps/rejected": -524.0, | |
| "loss": 0.1245, | |
| "rewards/accuracies": 0.9593750238418579, | |
| "rewards/chosen": 1.921875, | |
| "rewards/margins": 4.46875, | |
| "rewards/rejected": -2.546875, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.7384977215934145, | |
| "grad_norm": 1.058556318283081, | |
| "learning_rate": 6.939281288723669e-07, | |
| "logits/chosen": -4.40625, | |
| "logits/rejected": -4.40625, | |
| "logps/chosen": -588.0, | |
| "logps/rejected": -468.0, | |
| "loss": 0.1131, | |
| "rewards/accuracies": 0.953125, | |
| "rewards/chosen": 1.90625, | |
| "rewards/margins": 4.09375, | |
| "rewards/rejected": -2.1875, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.7620167573129502, | |
| "grad_norm": 1.1007747650146484, | |
| "learning_rate": 6.319702602230483e-07, | |
| "logits/chosen": -4.40625, | |
| "logits/rejected": -4.375, | |
| "logps/chosen": -620.0, | |
| "logps/rejected": -510.0, | |
| "loss": 0.0978, | |
| "rewards/accuracies": 0.971875011920929, | |
| "rewards/chosen": 1.9453125, | |
| "rewards/margins": 4.71875, | |
| "rewards/rejected": -2.765625, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.7855357930324858, | |
| "grad_norm": 1.0114929676055908, | |
| "learning_rate": 5.700123915737299e-07, | |
| "logits/chosen": -4.46875, | |
| "logits/rejected": -4.4375, | |
| "logps/chosen": -584.0, | |
| "logps/rejected": -472.0, | |
| "loss": 0.1013, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": 2.015625, | |
| "rewards/margins": 4.53125, | |
| "rewards/rejected": -2.53125, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.8090548287520212, | |
| "grad_norm": 0.9178210496902466, | |
| "learning_rate": 5.080545229244115e-07, | |
| "logits/chosen": -4.40625, | |
| "logits/rejected": -4.4375, | |
| "logps/chosen": -600.0, | |
| "logps/rejected": -470.0, | |
| "loss": 0.1082, | |
| "rewards/accuracies": 0.965624988079071, | |
| "rewards/chosen": 1.9375, | |
| "rewards/margins": 4.5, | |
| "rewards/rejected": -2.5625, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.8325738644715566, | |
| "grad_norm": 0.7771146893501282, | |
| "learning_rate": 4.4609665427509294e-07, | |
| "logits/chosen": -4.375, | |
| "logits/rejected": -4.4375, | |
| "logps/chosen": -600.0, | |
| "logps/rejected": -456.0, | |
| "loss": 0.1043, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": 1.765625, | |
| "rewards/margins": 4.5625, | |
| "rewards/rejected": -2.796875, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.8560929001910922, | |
| "grad_norm": 0.9903047680854797, | |
| "learning_rate": 3.8413878562577453e-07, | |
| "logits/chosen": -4.40625, | |
| "logits/rejected": -4.4375, | |
| "logps/chosen": -560.0, | |
| "logps/rejected": -462.0, | |
| "loss": 0.1029, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": 1.875, | |
| "rewards/margins": 4.4375, | |
| "rewards/rejected": -2.5625, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.8796119359106278, | |
| "grad_norm": 1.1646795272827148, | |
| "learning_rate": 3.22180916976456e-07, | |
| "logits/chosen": -4.5, | |
| "logits/rejected": -4.46875, | |
| "logps/chosen": -572.0, | |
| "logps/rejected": -444.0, | |
| "loss": 0.1103, | |
| "rewards/accuracies": 0.965624988079071, | |
| "rewards/chosen": 2.078125, | |
| "rewards/margins": 4.53125, | |
| "rewards/rejected": -2.4375, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.9031309716301632, | |
| "grad_norm": 0.7632185220718384, | |
| "learning_rate": 2.6022304832713755e-07, | |
| "logits/chosen": -4.4375, | |
| "logits/rejected": -4.40625, | |
| "logps/chosen": -584.0, | |
| "logps/rejected": -438.0, | |
| "loss": 0.0989, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": 1.9140625, | |
| "rewards/margins": 4.53125, | |
| "rewards/rejected": -2.625, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.9266500073496986, | |
| "grad_norm": 0.9221560955047607, | |
| "learning_rate": 1.982651796778191e-07, | |
| "logits/chosen": -4.53125, | |
| "logits/rejected": -4.53125, | |
| "logps/chosen": -592.0, | |
| "logps/rejected": -462.0, | |
| "loss": 0.1141, | |
| "rewards/accuracies": 0.9593750238418579, | |
| "rewards/chosen": 1.921875, | |
| "rewards/margins": 4.40625, | |
| "rewards/rejected": -2.484375, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.9501690430692342, | |
| "grad_norm": 1.339996576309204, | |
| "learning_rate": 1.3630731102850064e-07, | |
| "logits/chosen": -4.4375, | |
| "logits/rejected": -4.4375, | |
| "logps/chosen": -596.0, | |
| "logps/rejected": -454.0, | |
| "loss": 0.1099, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": 1.8203125, | |
| "rewards/margins": 4.46875, | |
| "rewards/rejected": -2.640625, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.9736880787887696, | |
| "grad_norm": 1.047699213027954, | |
| "learning_rate": 7.434944237918216e-08, | |
| "logits/chosen": -4.40625, | |
| "logits/rejected": -4.4375, | |
| "logps/chosen": -592.0, | |
| "logps/rejected": -446.0, | |
| "loss": 0.1054, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": 1.890625, | |
| "rewards/margins": 4.75, | |
| "rewards/rejected": -2.875, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.997207114508305, | |
| "grad_norm": 0.8556466698646545, | |
| "learning_rate": 1.2391573729863694e-08, | |
| "logits/chosen": -4.46875, | |
| "logits/rejected": -4.4375, | |
| "logps/chosen": -600.0, | |
| "logps/rejected": -516.0, | |
| "loss": 0.1072, | |
| "rewards/accuracies": 0.965624988079071, | |
| "rewards/chosen": 1.90625, | |
| "rewards/margins": 4.53125, | |
| "rewards/rejected": -2.625, | |
| "step": 850 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 850, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 250, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |