| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 775, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.012903225806451613, | |
| "grad_norm": 35.01725185776426, | |
| "learning_rate": 4.5000000000000003e-07, | |
| "log_odds_chosen": 1.0187499523162842, | |
| "log_odds_ratio": -0.34882813692092896, | |
| "logits/chosen": -0.8787597417831421, | |
| "logits/rejected": 15.768750190734863, | |
| "logps/chosen": -1.615625023841858, | |
| "logps/rejected": -2.5015625953674316, | |
| "loss": 1.9536, | |
| "nll_loss": 1.9367187023162842, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -0.08076171576976776, | |
| "rewards/margins": 0.0443115234375, | |
| "rewards/rejected": -0.12509766221046448, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.025806451612903226, | |
| "grad_norm": 13.48550724489305, | |
| "learning_rate": 9.500000000000001e-07, | |
| "log_odds_chosen": 0.690234363079071, | |
| "log_odds_ratio": -0.4419921934604645, | |
| "logits/chosen": 0.51593017578125, | |
| "logits/rejected": 17.450000762939453, | |
| "logps/chosen": -1.4140625, | |
| "logps/rejected": -1.9765625, | |
| "loss": 1.6515, | |
| "nll_loss": 1.6281249523162842, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -0.07062987983226776, | |
| "rewards/margins": 0.02830810472369194, | |
| "rewards/rejected": -0.09902343899011612, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.03870967741935484, | |
| "grad_norm": 11.199724125486222, | |
| "learning_rate": 1.45e-06, | |
| "log_odds_chosen": 0.3243164122104645, | |
| "log_odds_ratio": -0.5767577886581421, | |
| "logits/chosen": 0.98828125, | |
| "logits/rejected": 17.287500381469727, | |
| "logps/chosen": -1.2507812976837158, | |
| "logps/rejected": -1.490625023841858, | |
| "loss": 1.4296, | |
| "nll_loss": 1.400781273841858, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.0625, | |
| "rewards/margins": 0.012020873837172985, | |
| "rewards/rejected": -0.07456054538488388, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.05161290322580645, | |
| "grad_norm": 9.337042800085895, | |
| "learning_rate": 1.9500000000000004e-06, | |
| "log_odds_chosen": 0.2557617127895355, | |
| "log_odds_ratio": -0.603515625, | |
| "logits/chosen": 0.38398438692092896, | |
| "logits/rejected": 17.081249237060547, | |
| "logps/chosen": -1.111718773841858, | |
| "logps/rejected": -1.2937500476837158, | |
| "loss": 1.2509, | |
| "nll_loss": 1.220312476158142, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.05561523512005806, | |
| "rewards/margins": 0.009097290225327015, | |
| "rewards/rejected": -0.06474609673023224, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.06451612903225806, | |
| "grad_norm": 8.845956747236016, | |
| "learning_rate": 2.4500000000000003e-06, | |
| "log_odds_chosen": 0.4603515565395355, | |
| "log_odds_ratio": -0.525390625, | |
| "logits/chosen": -0.30903321504592896, | |
| "logits/rejected": 16.443750381469727, | |
| "logps/chosen": -1.074609398841858, | |
| "logps/rejected": -1.40625, | |
| "loss": 1.1469, | |
| "nll_loss": 1.119531273841858, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": -0.05373535305261612, | |
| "rewards/margins": 0.01662902906537056, | |
| "rewards/rejected": -0.07036133110523224, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.07741935483870968, | |
| "grad_norm": 11.864656308978198, | |
| "learning_rate": 2.95e-06, | |
| "log_odds_chosen": 0.994140625, | |
| "log_odds_ratio": -0.3656249940395355, | |
| "logits/chosen": -0.944531261920929, | |
| "logits/rejected": 15.512499809265137, | |
| "logps/chosen": -1.09375, | |
| "logps/rejected": -1.865625023841858, | |
| "loss": 1.1414, | |
| "nll_loss": 1.1242187023162842, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": -0.0546875, | |
| "rewards/margins": 0.03852539137005806, | |
| "rewards/rejected": -0.09321288764476776, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.09032258064516129, | |
| "grad_norm": 8.085176699285602, | |
| "learning_rate": 3.45e-06, | |
| "log_odds_chosen": 1.61328125, | |
| "log_odds_ratio": -0.24765625596046448, | |
| "logits/chosen": -1.749609351158142, | |
| "logits/rejected": 15.681249618530273, | |
| "logps/chosen": -1.0242187976837158, | |
| "logps/rejected": -2.328906297683716, | |
| "loss": 1.0588, | |
| "nll_loss": 1.0457031726837158, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.05122070387005806, | |
| "rewards/margins": 0.065185546875, | |
| "rewards/rejected": -0.116455078125, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.1032258064516129, | |
| "grad_norm": 8.293601800252382, | |
| "learning_rate": 3.95e-06, | |
| "log_odds_chosen": 1.795312523841858, | |
| "log_odds_ratio": -0.2099609375, | |
| "logits/chosen": -1.5556640625, | |
| "logits/rejected": 15.699999809265137, | |
| "logps/chosen": -1.0593750476837158, | |
| "logps/rejected": -2.520312547683716, | |
| "loss": 1.0778, | |
| "nll_loss": 1.0675780773162842, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.052978515625, | |
| "rewards/margins": 0.07309570163488388, | |
| "rewards/rejected": -0.12607422471046448, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.11612903225806452, | |
| "grad_norm": 10.024940843527872, | |
| "learning_rate": 4.450000000000001e-06, | |
| "log_odds_chosen": 1.907812476158142, | |
| "log_odds_ratio": -0.21826171875, | |
| "logits/chosen": -1.687109351158142, | |
| "logits/rejected": 15.618749618530273, | |
| "logps/chosen": -1.0261719226837158, | |
| "logps/rejected": -2.5835938453674316, | |
| "loss": 1.0509, | |
| "nll_loss": 1.0398437976837158, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -0.05126953125, | |
| "rewards/margins": 0.078125, | |
| "rewards/rejected": -0.12929686903953552, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.12903225806451613, | |
| "grad_norm": 9.561260138128471, | |
| "learning_rate": 4.95e-06, | |
| "log_odds_chosen": 2.039843797683716, | |
| "log_odds_ratio": -0.19287109375, | |
| "logits/chosen": -1.9093749523162842, | |
| "logits/rejected": 15.118749618530273, | |
| "logps/chosen": -1.058984398841858, | |
| "logps/rejected": -2.7578125, | |
| "loss": 1.0856, | |
| "nll_loss": 1.075781226158142, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -0.05295410007238388, | |
| "rewards/margins": 0.08500976860523224, | |
| "rewards/rejected": -0.137939453125, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.14193548387096774, | |
| "grad_norm": 7.99980022257946, | |
| "learning_rate": 4.789131426105757e-06, | |
| "log_odds_chosen": 2.422656297683716, | |
| "log_odds_ratio": -0.15205077826976776, | |
| "logits/chosen": -1.204187035560608, | |
| "logits/rejected": 15.850000381469727, | |
| "logps/chosen": -1.05078125, | |
| "logps/rejected": -3.0999999046325684, | |
| "loss": 1.0735, | |
| "nll_loss": 1.064453125, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.052490234375, | |
| "rewards/margins": 0.10239257663488388, | |
| "rewards/rejected": -0.15498046576976776, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.15483870967741936, | |
| "grad_norm": 7.301318748826927, | |
| "learning_rate": 4.583492485141056e-06, | |
| "log_odds_chosen": 2.557812452316284, | |
| "log_odds_ratio": -0.16196289658546448, | |
| "logits/chosen": -2.0648436546325684, | |
| "logits/rejected": 16.181249618530273, | |
| "logps/chosen": -1.025390625, | |
| "logps/rejected": -3.1953125, | |
| "loss": 1.0461, | |
| "nll_loss": 1.0378906726837158, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -0.05124511569738388, | |
| "rewards/margins": 0.10859374701976776, | |
| "rewards/rejected": -0.15976563096046448, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.16774193548387098, | |
| "grad_norm": 7.9997556459437735, | |
| "learning_rate": 4.4022545316281195e-06, | |
| "log_odds_chosen": 2.9625000953674316, | |
| "log_odds_ratio": -0.11276855319738388, | |
| "logits/chosen": -1.30615234375, | |
| "logits/rejected": 15.875, | |
| "logps/chosen": -1.056640625, | |
| "logps/rejected": -3.6171875, | |
| "loss": 1.071, | |
| "nll_loss": 1.06640625, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.05278320237994194, | |
| "rewards/margins": 0.12812499701976776, | |
| "rewards/rejected": -0.18085937201976776, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.18064516129032257, | |
| "grad_norm": 8.787059338175244, | |
| "learning_rate": 4.240944648399854e-06, | |
| "log_odds_chosen": 2.53515625, | |
| "log_odds_ratio": -0.14907225966453552, | |
| "logits/chosen": -1.6693603992462158, | |
| "logits/rejected": 15.34375, | |
| "logps/chosen": -1.0421874523162842, | |
| "logps/rejected": -3.192187547683716, | |
| "loss": 1.0614, | |
| "nll_loss": 1.054296851158142, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.05207519605755806, | |
| "rewards/margins": 0.10761718451976776, | |
| "rewards/rejected": -0.15966796875, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.1935483870967742, | |
| "grad_norm": 7.326908745174855, | |
| "learning_rate": 4.096159602595203e-06, | |
| "log_odds_chosen": 2.682812452316284, | |
| "log_odds_ratio": -0.16826172173023224, | |
| "logits/chosen": -2.536914110183716, | |
| "logits/rejected": 15.175000190734863, | |
| "logps/chosen": -1.026953101158142, | |
| "logps/rejected": -3.34375, | |
| "loss": 1.0456, | |
| "nll_loss": 1.0378906726837158, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.05136718600988388, | |
| "rewards/margins": 0.11567382514476776, | |
| "rewards/rejected": -0.1669921875, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2064516129032258, | |
| "grad_norm": 9.22122535363528, | |
| "learning_rate": 3.965257928590721e-06, | |
| "log_odds_chosen": 2.785937547683716, | |
| "log_odds_ratio": -0.18808594346046448, | |
| "logits/chosen": -2.75390625, | |
| "logits/rejected": 14.600000381469727, | |
| "logps/chosen": -0.991015613079071, | |
| "logps/rejected": -3.4000000953674316, | |
| "loss": 1.0212, | |
| "nll_loss": 1.011328101158142, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.04953613132238388, | |
| "rewards/margins": 0.12050781399011612, | |
| "rewards/rejected": -0.17011718451976776, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.21935483870967742, | |
| "grad_norm": 7.10853327954206, | |
| "learning_rate": 3.846153846153846e-06, | |
| "log_odds_chosen": 3.2984375953674316, | |
| "log_odds_ratio": -0.1068115234375, | |
| "logits/chosen": -2.6039061546325684, | |
| "logits/rejected": 15.818750381469727, | |
| "logps/chosen": -0.961718738079071, | |
| "logps/rejected": -3.8218750953674316, | |
| "loss": 0.9891, | |
| "nll_loss": 0.9839843511581421, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.04804687574505806, | |
| "rewards/margins": 0.14321288466453552, | |
| "rewards/rejected": -0.19121094048023224, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.23225806451612904, | |
| "grad_norm": 7.942817891766315, | |
| "learning_rate": 3.7371754637596797e-06, | |
| "log_odds_chosen": 3.356250047683716, | |
| "log_odds_ratio": -0.09636230766773224, | |
| "logits/chosen": -2.181640625, | |
| "logits/rejected": 15.274999618530273, | |
| "logps/chosen": -0.979296863079071, | |
| "logps/rejected": -3.8890624046325684, | |
| "loss": 0.9928, | |
| "nll_loss": 0.98828125, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.0489501953125, | |
| "rewards/margins": 0.14541015028953552, | |
| "rewards/rejected": -0.19443359971046448, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.24516129032258063, | |
| "grad_norm": 15.366307517097038, | |
| "learning_rate": 3.63696483726654e-06, | |
| "log_odds_chosen": 3.3984375, | |
| "log_odds_ratio": -0.10626220703125, | |
| "logits/chosen": -1.9557616710662842, | |
| "logits/rejected": 15.681249618530273, | |
| "logps/chosen": -0.9742187261581421, | |
| "logps/rejected": -3.940624952316284, | |
| "loss": 1.0004, | |
| "nll_loss": 0.99609375, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -0.04877929762005806, | |
| "rewards/margins": 0.14833983778953552, | |
| "rewards/rejected": -0.19707031548023224, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.25806451612903225, | |
| "grad_norm": 6.891451760355594, | |
| "learning_rate": 3.5444060250416796e-06, | |
| "log_odds_chosen": 3.645312547683716, | |
| "log_odds_ratio": -0.0811767578125, | |
| "logits/chosen": -2.288281202316284, | |
| "logits/rejected": 15.893750190734863, | |
| "logps/chosen": -1.0066406726837158, | |
| "logps/rejected": -4.215624809265137, | |
| "loss": 1.0215, | |
| "nll_loss": 1.017187476158142, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.05036621168255806, | |
| "rewards/margins": 0.16035155951976776, | |
| "rewards/rejected": -0.21074219048023224, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.2709677419354839, | |
| "grad_norm": 8.135582655341244, | |
| "learning_rate": 3.458572319330373e-06, | |
| "log_odds_chosen": 3.387500047683716, | |
| "log_odds_ratio": -0.14912109076976776, | |
| "logits/chosen": -2.5718750953674316, | |
| "logits/rejected": 15.149999618530273, | |
| "logps/chosen": -0.981249988079071, | |
| "logps/rejected": -3.9453125, | |
| "loss": 0.998, | |
| "nll_loss": 0.991015613079071, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -0.04902343824505806, | |
| "rewards/margins": 0.148193359375, | |
| "rewards/rejected": -0.197265625, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.2838709677419355, | |
| "grad_norm": 7.518297474091935, | |
| "learning_rate": 3.3786868919974304e-06, | |
| "log_odds_chosen": 2.4242186546325684, | |
| "log_odds_ratio": -0.17851562798023224, | |
| "logits/chosen": -1.6906249523162842, | |
| "logits/rejected": 15.84375, | |
| "logps/chosen": -1.036718726158142, | |
| "logps/rejected": -3.0859375, | |
| "loss": 1.0572, | |
| "nll_loss": 1.0480468273162842, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -0.0518798828125, | |
| "rewards/margins": 0.10249023139476776, | |
| "rewards/rejected": -0.15439453721046448, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.2967741935483871, | |
| "grad_norm": 7.499198073553291, | |
| "learning_rate": 3.3040930022754493e-06, | |
| "log_odds_chosen": 2.407031297683716, | |
| "log_odds_ratio": -0.145263671875, | |
| "logits/chosen": -1.3737304210662842, | |
| "logits/rejected": 15.581250190734863, | |
| "logps/chosen": -1.0304687023162842, | |
| "logps/rejected": -3.043750047683716, | |
| "loss": 1.0455, | |
| "nll_loss": 1.0382812023162842, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.05156249925494194, | |
| "rewards/margins": 0.10073242336511612, | |
| "rewards/rejected": -0.15224608778953552, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.3096774193548387, | |
| "grad_norm": 7.726811415559842, | |
| "learning_rate": 3.234231136765755e-06, | |
| "log_odds_chosen": 2.612499952316284, | |
| "log_odds_ratio": -0.14296874403953552, | |
| "logits/chosen": -2.7847657203674316, | |
| "logits/rejected": 14.0625, | |
| "logps/chosen": -0.987500011920929, | |
| "logps/rejected": -3.1937499046325684, | |
| "loss": 1.0104, | |
| "nll_loss": 1.00390625, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.04938964918255806, | |
| "rewards/margins": 0.11020507663488388, | |
| "rewards/rejected": -0.15957030653953552, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.3225806451612903, | |
| "grad_norm": 7.365705719642815, | |
| "learning_rate": 3.1686212526223897e-06, | |
| "log_odds_chosen": 3.2109375, | |
| "log_odds_ratio": -0.09316406399011612, | |
| "logits/chosen": -3.19921875, | |
| "logits/rejected": 15.037500381469727, | |
| "logps/chosen": -0.948046863079071, | |
| "logps/rejected": -3.6968750953674316, | |
| "loss": 0.9679, | |
| "nll_loss": 0.964062511920929, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.04741210862994194, | |
| "rewards/margins": 0.13754883408546448, | |
| "rewards/rejected": -0.18476562201976776, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.33548387096774196, | |
| "grad_norm": 6.518152237715758, | |
| "learning_rate": 3.1068488300060003e-06, | |
| "log_odds_chosen": 2.875, | |
| "log_odds_ratio": -0.1068115234375, | |
| "logits/chosen": -2.6175780296325684, | |
| "logits/rejected": 15.149999618530273, | |
| "logps/chosen": -0.97265625, | |
| "logps/rejected": -3.4078125953674316, | |
| "loss": 0.9938, | |
| "nll_loss": 0.98828125, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.04868163913488388, | |
| "rewards/margins": 0.12158203125, | |
| "rewards/rejected": -0.17031249403953552, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.34838709677419355, | |
| "grad_norm": 7.237801132096072, | |
| "learning_rate": 3.0485538042484623e-06, | |
| "log_odds_chosen": 3.1875, | |
| "log_odds_ratio": -0.10446777194738388, | |
| "logits/chosen": -2.614062547683716, | |
| "logits/rejected": 14.987500190734863, | |
| "logps/chosen": -0.9515625238418579, | |
| "logps/rejected": -3.6890625953674316, | |
| "loss": 0.9738, | |
| "nll_loss": 0.9691406488418579, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.047607421875, | |
| "rewards/margins": 0.13681641221046448, | |
| "rewards/rejected": -0.1845703125, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.36129032258064514, | |
| "grad_norm": 7.797308798589523, | |
| "learning_rate": 2.9934217004462487e-06, | |
| "log_odds_chosen": 3.8890624046325684, | |
| "log_odds_ratio": -0.0784912109375, | |
| "logits/chosen": -2.3812499046325684, | |
| "logits/rejected": 15.181249618530273, | |
| "logps/chosen": -0.993359386920929, | |
| "logps/rejected": -4.421875, | |
| "loss": 0.9999, | |
| "nll_loss": 0.9976562261581421, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.04965820163488388, | |
| "rewards/margins": 0.17138671875, | |
| "rewards/rejected": -0.22119140625, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.3741935483870968, | |
| "grad_norm": 30.586016864966307, | |
| "learning_rate": 2.9411764705882355e-06, | |
| "log_odds_chosen": 4.060937404632568, | |
| "log_odds_ratio": -0.06879882514476776, | |
| "logits/chosen": -2.482421875, | |
| "logits/rejected": 15.481249809265137, | |
| "logps/chosen": -0.962109386920929, | |
| "logps/rejected": -4.540625095367432, | |
| "loss": 0.9723, | |
| "nll_loss": 0.9691406488418579, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.04816894605755806, | |
| "rewards/margins": 0.17919921875, | |
| "rewards/rejected": -0.22734375298023224, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.3870967741935484, | |
| "grad_norm": 8.595204860528385, | |
| "learning_rate": 2.891574659831201e-06, | |
| "log_odds_chosen": 4.564062595367432, | |
| "log_odds_ratio": -0.07656250149011612, | |
| "logits/chosen": -2.651562452316284, | |
| "logits/rejected": 15.306249618530273, | |
| "logps/chosen": -0.9839843511581421, | |
| "logps/rejected": -5.090624809265137, | |
| "loss": 0.9966, | |
| "nll_loss": 0.992968738079071, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.04916992038488388, | |
| "rewards/margins": 0.20488281548023224, | |
| "rewards/rejected": -0.25439453125, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 8.703268967809453, | |
| "learning_rate": 2.844400619942872e-06, | |
| "log_odds_chosen": 4.431250095367432, | |
| "log_odds_ratio": -0.07321777194738388, | |
| "logits/chosen": -2.5328125953674316, | |
| "logits/rejected": 14.806249618530273, | |
| "logps/chosen": -0.96875, | |
| "logps/rejected": -4.932812690734863, | |
| "loss": 0.9852, | |
| "nll_loss": 0.982421875, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.04848632961511612, | |
| "rewards/margins": 0.1982421875, | |
| "rewards/rejected": -0.24677734076976776, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.4129032258064516, | |
| "grad_norm": 7.421491749584791, | |
| "learning_rate": 2.799462554779272e-06, | |
| "log_odds_chosen": 4.453125, | |
| "log_odds_ratio": -0.07643432915210724, | |
| "logits/chosen": -2.30078125, | |
| "logits/rejected": 15.4375, | |
| "logps/chosen": -0.954296886920929, | |
| "logps/rejected": -4.931250095367432, | |
| "loss": 0.9638, | |
| "nll_loss": 0.959765613079071, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.04775390774011612, | |
| "rewards/margins": 0.19873046875, | |
| "rewards/rejected": -0.24628905951976776, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.4258064516129032, | |
| "grad_norm": 8.918312407814545, | |
| "learning_rate": 2.7565892320998566e-06, | |
| "log_odds_chosen": 4.756249904632568, | |
| "log_odds_ratio": -0.059814453125, | |
| "logits/chosen": -2.6878905296325684, | |
| "logits/rejected": 15.5, | |
| "logps/chosen": -0.982421875, | |
| "logps/rejected": -5.262499809265137, | |
| "loss": 1.011, | |
| "nll_loss": 1.0078125, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.04909668117761612, | |
| "rewards/margins": 0.21425780653953552, | |
| "rewards/rejected": -0.2630859315395355, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.43870967741935485, | |
| "grad_norm": 7.528569200319769, | |
| "learning_rate": 2.7156272329678423e-06, | |
| "log_odds_chosen": 4.699999809265137, | |
| "log_odds_ratio": -0.04852294921875, | |
| "logits/chosen": -2.184375047683716, | |
| "logits/rejected": 15.287500381469727, | |
| "logps/chosen": -0.987500011920929, | |
| "logps/rejected": -5.215624809265137, | |
| "loss": 0.9846, | |
| "nll_loss": 0.9820312261581421, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.04936523362994194, | |
| "rewards/margins": 0.21123047173023224, | |
| "rewards/rejected": -0.2607421875, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.45161290322580644, | |
| "grad_norm": 7.297861319360668, | |
| "learning_rate": 2.6764386378609465e-06, | |
| "log_odds_chosen": 4.709374904632568, | |
| "log_odds_ratio": -0.057373046875, | |
| "logits/chosen": -2.99609375, | |
| "logits/rejected": 15.09375, | |
| "logps/chosen": -0.9339843988418579, | |
| "logps/rejected": -5.134375095367432, | |
| "loss": 0.9631, | |
| "nll_loss": 0.9605468511581421, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.04672851413488388, | |
| "rewards/margins": 0.21025390923023224, | |
| "rewards/rejected": -0.2569335997104645, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.4645161290322581, | |
| "grad_norm": 8.07548290053001, | |
| "learning_rate": 2.638899069846298e-06, | |
| "log_odds_chosen": 4.639062404632568, | |
| "log_odds_ratio": -0.06218261644244194, | |
| "logits/chosen": -2.604687452316284, | |
| "logits/rejected": 15.231249809265137, | |
| "logps/chosen": -0.969531238079071, | |
| "logps/rejected": -5.131249904632568, | |
| "loss": 0.9815, | |
| "nll_loss": 0.979687511920929, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.04843749850988388, | |
| "rewards/margins": 0.2080078125, | |
| "rewards/rejected": -0.2564453184604645, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.4774193548387097, | |
| "grad_norm": 7.439409049540854, | |
| "learning_rate": 2.6028960314767675e-06, | |
| "log_odds_chosen": 4.8125, | |
| "log_odds_ratio": -0.06487731635570526, | |
| "logits/chosen": -3.2523436546325684, | |
| "logits/rejected": 15.425000190734863, | |
| "logps/chosen": -0.981249988079071, | |
| "logps/rejected": -5.306250095367432, | |
| "loss": 0.984, | |
| "nll_loss": 0.981640636920929, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.049072265625, | |
| "rewards/margins": 0.216796875, | |
| "rewards/rejected": -0.2657226622104645, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.49032258064516127, | |
| "grad_norm": 6.717751304060122, | |
| "learning_rate": 2.568327484687571e-06, | |
| "log_odds_chosen": 4.300000190734863, | |
| "log_odds_ratio": -0.09389648586511612, | |
| "logits/chosen": -2.848437547683716, | |
| "logits/rejected": 15.28125, | |
| "logps/chosen": -0.934765636920929, | |
| "logps/rejected": -4.760937690734863, | |
| "loss": 0.9459, | |
| "nll_loss": 0.94140625, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.0467529296875, | |
| "rewards/margins": 0.19101563096046448, | |
| "rewards/rejected": -0.23779296875, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.5032258064516129, | |
| "grad_norm": 7.456596504746751, | |
| "learning_rate": 2.535100632816969e-06, | |
| "log_odds_chosen": 4.435937404632568, | |
| "log_odds_ratio": -0.07844237983226776, | |
| "logits/chosen": -2.9749999046325684, | |
| "logits/rejected": 15.487500190734863, | |
| "logps/chosen": -0.935546875, | |
| "logps/rejected": -4.892187595367432, | |
| "loss": 0.9631, | |
| "nll_loss": 0.9593750238418579, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -0.0467529296875, | |
| "rewards/margins": 0.19755859673023224, | |
| "rewards/rejected": -0.24462890625, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.5161290322580645, | |
| "grad_norm": 6.423921141623107, | |
| "learning_rate": 2.5031308716087944e-06, | |
| "log_odds_chosen": 4.626562595367432, | |
| "log_odds_ratio": -0.05325927585363388, | |
| "logits/chosen": -3.141406297683716, | |
| "logits/rejected": 15.106249809265137, | |
| "logps/chosen": -0.951171875, | |
| "logps/rejected": -5.078125, | |
| "loss": 0.9649, | |
| "nll_loss": 0.962109386920929, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.04755859449505806, | |
| "rewards/margins": 0.20634765923023224, | |
| "rewards/rejected": -0.2538085877895355, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5290322580645161, | |
| "grad_norm": 8.037211110808377, | |
| "learning_rate": 2.472340882170744e-06, | |
| "log_odds_chosen": 4.784375190734863, | |
| "log_odds_ratio": -0.03513183444738388, | |
| "logits/chosen": -2.914843797683716, | |
| "logits/rejected": 15.162500381469727, | |
| "logps/chosen": -0.950390636920929, | |
| "logps/rejected": -5.237500190734863, | |
| "loss": 0.957, | |
| "nll_loss": 0.955078125, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.04758300632238388, | |
| "rewards/margins": 0.21406249701976776, | |
| "rewards/rejected": -0.26142579317092896, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.5419354838709678, | |
| "grad_norm": 7.975874669678937, | |
| "learning_rate": 2.4426598437301577e-06, | |
| "log_odds_chosen": 4.809374809265137, | |
| "log_odds_ratio": -0.06281737983226776, | |
| "logits/chosen": -2.657031297683716, | |
| "logits/rejected": 14.993749618530273, | |
| "logps/chosen": -0.962109386920929, | |
| "logps/rejected": -5.296875, | |
| "loss": 0.9726, | |
| "nll_loss": 0.969921886920929, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.048095703125, | |
| "rewards/margins": 0.21660156548023224, | |
| "rewards/rejected": -0.2645507752895355, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.5548387096774193, | |
| "grad_norm": 7.980706154694817, | |
| "learning_rate": 2.4140227479263382e-06, | |
| "log_odds_chosen": 4.993750095367432, | |
| "log_odds_ratio": -0.05314941331744194, | |
| "logits/chosen": -1.9533202648162842, | |
| "logits/rejected": 15.962499618530273, | |
| "logps/chosen": -0.9574218988418579, | |
| "logps/rejected": -5.465624809265137, | |
| "loss": 0.9602, | |
| "nll_loss": 0.957812488079071, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.04787597805261612, | |
| "rewards/margins": 0.22539062798023224, | |
| "rewards/rejected": -0.2728515565395355, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.567741935483871, | |
| "grad_norm": 9.156473187948562, | |
| "learning_rate": 2.3863697995167388e-06, | |
| "log_odds_chosen": 4.903124809265137, | |
| "log_odds_ratio": -0.04511108249425888, | |
| "logits/chosen": -3.05126953125, | |
| "logits/rejected": 15.243749618530273, | |
| "logps/chosen": -0.9671875238418579, | |
| "logps/rejected": -5.387499809265137, | |
| "loss": 0.9845, | |
| "nll_loss": 0.981640636920929, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.04831542819738388, | |
| "rewards/margins": 0.22089843451976776, | |
| "rewards/rejected": -0.26953125, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.5806451612903226, | |
| "grad_norm": 6.879386887391153, | |
| "learning_rate": 2.359645890915044e-06, | |
| "log_odds_chosen": 4.698437690734863, | |
| "log_odds_ratio": -0.07050170749425888, | |
| "logits/chosen": -2.684375047683716, | |
| "logits/rejected": 14.993749618530273, | |
| "logps/chosen": -0.983203113079071, | |
| "logps/rejected": -5.199999809265137, | |
| "loss": 1.0031, | |
| "nll_loss": 0.9996093511581421, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -0.0491943359375, | |
| "rewards/margins": 0.2109375, | |
| "rewards/rejected": -0.26005858182907104, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.5935483870967742, | |
| "grad_norm": 9.291098134467267, | |
| "learning_rate": 2.333800140046683e-06, | |
| "log_odds_chosen": 4.96875, | |
| "log_odds_ratio": -0.05540771409869194, | |
| "logits/chosen": -2.6537108421325684, | |
| "logits/rejected": 15.399999618530273, | |
| "logps/chosen": -0.9515625238418579, | |
| "logps/rejected": -5.418749809265137, | |
| "loss": 0.9651, | |
| "nll_loss": 0.962890625, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.04755859449505806, | |
| "rewards/margins": 0.22353515028953552, | |
| "rewards/rejected": -0.2708984315395355, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.6064516129032258, | |
| "grad_norm": 8.260430320302163, | |
| "learning_rate": 2.3087854826980507e-06, | |
| "log_odds_chosen": 4.625, | |
| "log_odds_ratio": -0.078857421875, | |
| "logits/chosen": -3.047656297683716, | |
| "logits/rejected": 14.675000190734863, | |
| "logps/chosen": -0.994921863079071, | |
| "logps/rejected": -5.173437595367432, | |
| "loss": 0.9971, | |
| "nll_loss": 0.9925781488418579, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.04975586012005806, | |
| "rewards/margins": 0.20878906548023224, | |
| "rewards/rejected": -0.25859373807907104, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.6193548387096774, | |
| "grad_norm": 7.97198152793078, | |
| "learning_rate": 2.2845583119247696e-06, | |
| "log_odds_chosen": 5.15625, | |
| "log_odds_ratio": -0.043426513671875, | |
| "logits/chosen": -2.921093702316284, | |
| "logits/rejected": 15.574999809265137, | |
| "logps/chosen": -0.9105468988418579, | |
| "logps/rejected": -5.540625095367432, | |
| "loss": 0.9281, | |
| "nll_loss": 0.926562488079071, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.04550781100988388, | |
| "rewards/margins": 0.23164062201976776, | |
| "rewards/rejected": -0.27714842557907104, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.632258064516129, | |
| "grad_norm": 8.119188050464615, | |
| "learning_rate": 2.2610781582306733e-06, | |
| "log_odds_chosen": 4.90625, | |
| "log_odds_ratio": -0.04886474460363388, | |
| "logits/chosen": -3.275390625, | |
| "logits/rejected": 15.618749618530273, | |
| "logps/chosen": -0.9800781011581421, | |
| "logps/rejected": -5.409375190734863, | |
| "loss": 0.9917, | |
| "nll_loss": 0.989453136920929, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.04899902269244194, | |
| "rewards/margins": 0.22128906846046448, | |
| "rewards/rejected": -0.27031248807907104, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.6451612903225806, | |
| "grad_norm": 7.067779726362276, | |
| "learning_rate": 2.238307405179226e-06, | |
| "log_odds_chosen": 4.7109375, | |
| "log_odds_ratio": -0.06611327826976776, | |
| "logits/chosen": -2.866406202316284, | |
| "logits/rejected": 15.162500381469727, | |
| "logps/chosen": -0.9976562261581421, | |
| "logps/rejected": -5.243750095367432, | |
| "loss": 1.0073, | |
| "nll_loss": 1.0046875476837158, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.04990234225988388, | |
| "rewards/margins": 0.21210937201976776, | |
| "rewards/rejected": -0.26201170682907104, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.6580645161290323, | |
| "grad_norm": 7.627869931724449, | |
| "learning_rate": 2.2162110358896814e-06, | |
| "log_odds_chosen": 5.0625, | |
| "log_odds_ratio": -0.04056396335363388, | |
| "logits/chosen": -3.075000047683716, | |
| "logits/rejected": 15.043749809265137, | |
| "logps/chosen": -0.9144531488418579, | |
| "logps/rejected": -5.449999809265137, | |
| "loss": 0.9441, | |
| "nll_loss": 0.943359375, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.04575195163488388, | |
| "rewards/margins": 0.22675780951976776, | |
| "rewards/rejected": -0.27275389432907104, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.6709677419354839, | |
| "grad_norm": 7.44284861722793, | |
| "learning_rate": 2.1947564065307354e-06, | |
| "log_odds_chosen": 4.876562595367432, | |
| "log_odds_ratio": -0.04053955152630806, | |
| "logits/chosen": -3.2132811546325684, | |
| "logits/rejected": 15.425000190734863, | |
| "logps/chosen": -0.913281261920929, | |
| "logps/rejected": -5.275000095367432, | |
| "loss": 0.9314, | |
| "nll_loss": 0.9292968511581421, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.045654296875, | |
| "rewards/margins": 0.21787109971046448, | |
| "rewards/rejected": -0.26347655057907104, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.6838709677419355, | |
| "grad_norm": 7.943793514721159, | |
| "learning_rate": 2.173913043478261e-06, | |
| "log_odds_chosen": 4.599999904632568, | |
| "log_odds_ratio": -0.06662597507238388, | |
| "logits/chosen": -2.092578172683716, | |
| "logits/rejected": 14.931249618530273, | |
| "logps/chosen": -0.911328136920929, | |
| "logps/rejected": -5.003125190734863, | |
| "loss": 0.9341, | |
| "nll_loss": 0.9312499761581421, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.04555664211511612, | |
| "rewards/margins": 0.20468750596046448, | |
| "rewards/rejected": -0.25029295682907104, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.6967741935483871, | |
| "grad_norm": 6.647015764786715, | |
| "learning_rate": 2.1536524612697402e-06, | |
| "log_odds_chosen": 4.793749809265137, | |
| "log_odds_ratio": -0.05656128004193306, | |
| "logits/chosen": -1.971093773841858, | |
| "logits/rejected": 16.006250381469727, | |
| "logps/chosen": -0.9496093988418579, | |
| "logps/rejected": -5.231249809265137, | |
| "loss": 0.9388, | |
| "nll_loss": 0.9359375238418579, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.0474853515625, | |
| "rewards/margins": 0.21416015923023224, | |
| "rewards/rejected": -0.26152342557907104, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.7096774193548387, | |
| "grad_norm": 8.078904372917599, | |
| "learning_rate": 2.1339479988816e-06, | |
| "log_odds_chosen": 4.9375, | |
| "log_odds_ratio": -0.03520507737994194, | |
| "logits/chosen": -2.5245604515075684, | |
| "logits/rejected": 15.412500381469727, | |
| "logps/chosen": -0.9507812261581421, | |
| "logps/rejected": -5.384375095367432, | |
| "loss": 0.9523, | |
| "nll_loss": 0.9515625238418579, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.04746093600988388, | |
| "rewards/margins": 0.22158202528953552, | |
| "rewards/rejected": -0.26923829317092896, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.7225806451612903, | |
| "grad_norm": 7.439562671796703, | |
| "learning_rate": 2.114774672189068e-06, | |
| "log_odds_chosen": 4.993750095367432, | |
| "log_odds_ratio": -0.044708251953125, | |
| "logits/chosen": -3.035937547683716, | |
| "logits/rejected": 15.199999809265137, | |
| "logps/chosen": -0.9527343511581421, | |
| "logps/rejected": -5.446875095367432, | |
| "loss": 0.9692, | |
| "nll_loss": 0.966796875, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.04768066480755806, | |
| "rewards/margins": 0.22470703721046448, | |
| "rewards/rejected": -0.27226561307907104, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.7354838709677419, | |
| "grad_norm": 6.850420581812698, | |
| "learning_rate": 2.0961090407515926e-06, | |
| "log_odds_chosen": 5.012499809265137, | |
| "log_odds_ratio": -0.03801269456744194, | |
| "logits/chosen": -2.707812547683716, | |
| "logits/rejected": 15.393750190734863, | |
| "logps/chosen": -0.926953136920929, | |
| "logps/rejected": -5.425000190734863, | |
| "loss": 0.9428, | |
| "nll_loss": 0.94140625, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.04636230319738388, | |
| "rewards/margins": 0.22451171278953552, | |
| "rewards/rejected": -0.2706054747104645, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.7483870967741936, | |
| "grad_norm": 7.650991563757063, | |
| "learning_rate": 2.077929087308457e-06, | |
| "log_odds_chosen": 4.650000095367432, | |
| "log_odds_ratio": -0.04947509616613388, | |
| "logits/chosen": -3.3359375, | |
| "logits/rejected": 14.625, | |
| "logps/chosen": -0.8941406011581421, | |
| "logps/rejected": -5.009375095367432, | |
| "loss": 0.9181, | |
| "nll_loss": 0.916015625, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.04475097730755806, | |
| "rewards/margins": 0.2060546875, | |
| "rewards/rejected": -0.25048828125, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.7612903225806451, | |
| "grad_norm": 7.115648653428975, | |
| "learning_rate": 2.0602141085758233e-06, | |
| "log_odds_chosen": 4.795312404632568, | |
| "log_odds_ratio": -0.04981689527630806, | |
| "logits/chosen": -3.2984375953674316, | |
| "logits/rejected": 15.037500381469727, | |
| "logps/chosen": -0.912890613079071, | |
| "logps/rejected": -5.192187309265137, | |
| "loss": 0.9367, | |
| "nll_loss": 0.9339843988418579, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.045654296875, | |
| "rewards/margins": 0.21416015923023224, | |
| "rewards/rejected": -0.25947266817092896, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.7741935483870968, | |
| "grad_norm": 6.8341435054086555, | |
| "learning_rate": 2.0429446161135922e-06, | |
| "log_odds_chosen": 4.859375, | |
| "log_odds_ratio": -0.04189453274011612, | |
| "logits/chosen": -2.752734422683716, | |
| "logits/rejected": 15.381250381469727, | |
| "logps/chosen": -0.94921875, | |
| "logps/rejected": -5.303124904632568, | |
| "loss": 0.9627, | |
| "nll_loss": 0.9609375, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.04743652418255806, | |
| "rewards/margins": 0.2177734375, | |
| "rewards/rejected": -0.26494139432907104, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.7870967741935484, | |
| "grad_norm": 7.6844935408533495, | |
| "learning_rate": 2.0261022461827697e-06, | |
| "log_odds_chosen": 5.337500095367432, | |
| "log_odds_ratio": -0.03539428859949112, | |
| "logits/chosen": -3.785937547683716, | |
| "logits/rejected": 15.237500190734863, | |
| "logps/chosen": -0.9476562738418579, | |
| "logps/rejected": -5.771874904632568, | |
| "loss": 0.9636, | |
| "nll_loss": 0.962109386920929, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.04733886569738388, | |
| "rewards/margins": 0.2412109375, | |
| "rewards/rejected": -0.28857421875, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 8.392646613992586, | |
| "learning_rate": 2.009669677645352e-06, | |
| "log_odds_chosen": 4.953125, | |
| "log_odds_ratio": -0.048187255859375, | |
| "logits/chosen": -3.5023436546325684, | |
| "logits/rejected": 15.518750190734863, | |
| "logps/chosen": -0.9253906011581421, | |
| "logps/rejected": -5.353125095367432, | |
| "loss": 0.9431, | |
| "nll_loss": 0.94140625, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.04624023288488388, | |
| "rewards/margins": 0.22138671576976776, | |
| "rewards/rejected": -0.26777344942092896, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.8129032258064516, | |
| "grad_norm": 7.4781250651760205, | |
| "learning_rate": 1.9936305570722503e-06, | |
| "log_odds_chosen": 4.815625190734863, | |
| "log_odds_ratio": -0.04139404371380806, | |
| "logits/chosen": -3.1523194313049316, | |
| "logits/rejected": 15.456250190734863, | |
| "logps/chosen": -0.9039062261581421, | |
| "logps/rejected": -5.181250095367432, | |
| "loss": 0.9141, | |
| "nll_loss": 0.912890613079071, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.04514160007238388, | |
| "rewards/margins": 0.2138671875, | |
| "rewards/rejected": -0.2591796815395355, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.8258064516129032, | |
| "grad_norm": 8.77906696889682, | |
| "learning_rate": 1.977969430323089e-06, | |
| "log_odds_chosen": 4.981249809265137, | |
| "log_odds_ratio": -0.035614013671875, | |
| "logits/chosen": -2.821337938308716, | |
| "logits/rejected": 15.518750190734863, | |
| "logps/chosen": -0.9605468511581421, | |
| "logps/rejected": -5.434374809265137, | |
| "loss": 0.973, | |
| "nll_loss": 0.97265625, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.04794921725988388, | |
| "rewards/margins": 0.22382812201976776, | |
| "rewards/rejected": -0.27177733182907104, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.8387096774193549, | |
| "grad_norm": 7.337971445424149, | |
| "learning_rate": 1.962671679947149e-06, | |
| "log_odds_chosen": 5.328125, | |
| "log_odds_ratio": -0.03910522535443306, | |
| "logits/chosen": -3.082812547683716, | |
| "logits/rejected": 16.075000762939453, | |
| "logps/chosen": -0.962109386920929, | |
| "logps/rejected": -5.787499904632568, | |
| "loss": 0.9758, | |
| "nll_loss": 0.974609375, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.04814453050494194, | |
| "rewards/margins": 0.2412109375, | |
| "rewards/rejected": -0.2894531190395355, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.8516129032258064, | |
| "grad_norm": 7.439098375728625, | |
| "learning_rate": 1.9477234678290497e-06, | |
| "log_odds_chosen": 4.9140625, | |
| "log_odds_ratio": -0.04149780422449112, | |
| "logits/chosen": -2.4859375953674316, | |
| "logits/rejected": 15.40625, | |
| "logps/chosen": -0.982421875, | |
| "logps/rejected": -5.409375190734863, | |
| "loss": 0.9789, | |
| "nll_loss": 0.977343738079071, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.049072265625, | |
| "rewards/margins": 0.22138671576976776, | |
| "rewards/rejected": -0.2705078125, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.864516129032258, | |
| "grad_norm": 7.0935668026210585, | |
| "learning_rate": 1.9331116825676476e-06, | |
| "log_odds_chosen": 4.962500095367432, | |
| "log_odds_ratio": -0.04017334058880806, | |
| "logits/chosen": -3.327343702316284, | |
| "logits/rejected": 15.162500381469727, | |
| "logps/chosen": -0.940625011920929, | |
| "logps/rejected": -5.387499809265137, | |
| "loss": 0.9549, | |
| "nll_loss": 0.953125, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.04702148586511612, | |
| "rewards/margins": 0.22285155951976776, | |
| "rewards/rejected": -0.2696289122104645, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.8774193548387097, | |
| "grad_norm": 8.300635787093205, | |
| "learning_rate": 1.9188238911333405e-06, | |
| "log_odds_chosen": 5.215624809265137, | |
| "log_odds_ratio": -0.02827148512005806, | |
| "logits/chosen": -2.7085938453674316, | |
| "logits/rejected": 15.381250381469727, | |
| "logps/chosen": -0.9664062261581421, | |
| "logps/rejected": -5.675000190734863, | |
| "loss": 0.9702, | |
| "nll_loss": 0.9691406488418579, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.04838867112994194, | |
| "rewards/margins": 0.23544922471046448, | |
| "rewards/rejected": -0.28369140625, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.8903225806451613, | |
| "grad_norm": 8.215237709287992, | |
| "learning_rate": 1.9048482943986487e-06, | |
| "log_odds_chosen": 4.873437404632568, | |
| "log_odds_ratio": -0.06497802585363388, | |
| "logits/chosen": -2.8578124046325684, | |
| "logits/rejected": 14.987500190734863, | |
| "logps/chosen": -0.9417968988418579, | |
| "logps/rejected": -5.318749904632568, | |
| "loss": 0.9407, | |
| "nll_loss": 0.938281238079071, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.04704589769244194, | |
| "rewards/margins": 0.21845702826976776, | |
| "rewards/rejected": -0.265625, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.9032258064516129, | |
| "grad_norm": 7.758338411265055, | |
| "learning_rate": 1.8911736861805847e-06, | |
| "log_odds_chosen": 4.796875, | |
| "log_odds_ratio": -0.04442138597369194, | |
| "logits/chosen": -2.7476563453674316, | |
| "logits/rejected": 15.206250190734863, | |
| "logps/chosen": -0.9046875238418579, | |
| "logps/rejected": -5.175000190734863, | |
| "loss": 0.918, | |
| "nll_loss": 0.9164062738418579, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.0452880859375, | |
| "rewards/margins": 0.21328124403953552, | |
| "rewards/rejected": -0.25859373807907104, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.9161290322580645, | |
| "grad_norm": 7.757899950179007, | |
| "learning_rate": 1.8777894154716994e-06, | |
| "log_odds_chosen": 5.040625095367432, | |
| "log_odds_ratio": -0.05682983249425888, | |
| "logits/chosen": -3.3187499046325684, | |
| "logits/rejected": 15.112500190734863, | |
| "logps/chosen": -0.9300781488418579, | |
| "logps/rejected": -5.46875, | |
| "loss": 0.9523, | |
| "nll_loss": 0.949999988079071, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.04653320461511612, | |
| "rewards/margins": 0.22666016221046448, | |
| "rewards/rejected": -0.2730468809604645, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.9290322580645162, | |
| "grad_norm": 6.925223393286465, | |
| "learning_rate": 1.864685351570504e-06, | |
| "log_odds_chosen": 5.334374904632568, | |
| "log_odds_ratio": -0.03135986253619194, | |
| "logits/chosen": -2.9644532203674316, | |
| "logits/rejected": 15.293749809265137, | |
| "logps/chosen": -0.904296875, | |
| "logps/rejected": -5.706250190734863, | |
| "loss": 0.9248, | |
| "nll_loss": 0.9242187738418579, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.04521484300494194, | |
| "rewards/margins": 0.24013671278953552, | |
| "rewards/rejected": -0.2855468690395355, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.9419354838709677, | |
| "grad_norm": 8.560013543994318, | |
| "learning_rate": 1.8518518518518519e-06, | |
| "log_odds_chosen": 4.846875190734863, | |
| "log_odds_ratio": -0.04491577297449112, | |
| "logits/chosen": -3.2242188453674316, | |
| "logits/rejected": 15.106249809265137, | |
| "logps/chosen": -0.938671886920929, | |
| "logps/rejected": -5.278124809265137, | |
| "loss": 0.9527, | |
| "nll_loss": 0.951953113079071, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.04694824293255806, | |
| "rewards/margins": 0.21738281846046448, | |
| "rewards/rejected": -0.26435548067092896, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.9548387096774194, | |
| "grad_norm": 7.44645859310936, | |
| "learning_rate": 1.8392797319442742e-06, | |
| "log_odds_chosen": 4.715624809265137, | |
| "log_odds_ratio": -0.05336914211511612, | |
| "logits/chosen": -3.6304688453674316, | |
| "logits/rejected": 14.675000190734863, | |
| "logps/chosen": -0.896484375, | |
| "logps/rejected": -5.090624809265137, | |
| "loss": 0.903, | |
| "nll_loss": 0.9007812738418579, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.0447998046875, | |
| "rewards/margins": 0.20957031846046448, | |
| "rewards/rejected": -0.25439453125, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.967741935483871, | |
| "grad_norm": 7.033644961041026, | |
| "learning_rate": 1.8269602381046742e-06, | |
| "log_odds_chosen": 5.084374904632568, | |
| "log_odds_ratio": -0.04151611402630806, | |
| "logits/chosen": -3.184375047683716, | |
| "logits/rejected": 15.412500381469727, | |
| "logps/chosen": -0.9359375238418579, | |
| "logps/rejected": -5.5, | |
| "loss": 0.9548, | |
| "nll_loss": 0.9535156488418579, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.04672851413488388, | |
| "rewards/margins": 0.22841796278953552, | |
| "rewards/rejected": -0.27509766817092896, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.9806451612903225, | |
| "grad_norm": 7.302442214130208, | |
| "learning_rate": 1.8148850216015693e-06, | |
| "log_odds_chosen": 4.909375190734863, | |
| "log_odds_ratio": -0.05724792554974556, | |
| "logits/chosen": -2.991406202316284, | |
| "logits/rejected": 15.149999618530273, | |
| "logps/chosen": -0.930859386920929, | |
| "logps/rejected": -5.34375, | |
| "loss": 0.9487, | |
| "nll_loss": 0.9453125, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.04653320461511612, | |
| "rewards/margins": 0.22050781548023224, | |
| "rewards/rejected": -0.2671875059604645, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.9935483870967742, | |
| "grad_norm": 25.02421316510306, | |
| "learning_rate": 1.803046114936547e-06, | |
| "log_odds_chosen": 5.053124904632568, | |
| "log_odds_ratio": -0.03767089918255806, | |
| "logits/chosen": -3.510937452316284, | |
| "logits/rejected": 15.0, | |
| "logps/chosen": -0.9378906488418579, | |
| "logps/rejected": -5.478125095367432, | |
| "loss": 0.9543, | |
| "nll_loss": 0.953906238079071, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.04685058444738388, | |
| "rewards/margins": 0.22675780951976776, | |
| "rewards/rejected": -0.27363282442092896, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 775, | |
| "total_flos": 0.0, | |
| "train_loss": 1.0152948588709678, | |
| "train_runtime": 1543.1546, | |
| "train_samples_per_second": 8.035, | |
| "train_steps_per_second": 0.502 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 775, | |
| "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": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |