| { | |
| "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": 9.303802312326823, | |
| "learning_rate": 9.000000000000001e-07, | |
| "log_odds_chosen": 1.083593726158142, | |
| "log_odds_ratio": -0.33476561307907104, | |
| "logits/chosen": -1.0812499523162842, | |
| "logits/rejected": 15.675000190734863, | |
| "logps/chosen": -1.661718726158142, | |
| "logps/rejected": -2.6156249046325684, | |
| "loss": 2.0143, | |
| "nll_loss": 1.99609375, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -0.083251953125, | |
| "rewards/margins": 0.04765624925494194, | |
| "rewards/rejected": -0.13066406548023224, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.025806451612903226, | |
| "grad_norm": 7.100174397164698, | |
| "learning_rate": 1.9000000000000002e-06, | |
| "log_odds_chosen": 1.1789062023162842, | |
| "log_odds_ratio": -0.31318360567092896, | |
| "logits/chosen": -1.015625, | |
| "logits/rejected": 16.8125, | |
| "logps/chosen": -1.6867187023162842, | |
| "logps/rejected": -2.729687452316284, | |
| "loss": 2.0014, | |
| "nll_loss": 1.9835937023162842, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.08432617038488388, | |
| "rewards/margins": 0.05207519605755806, | |
| "rewards/rejected": -0.13637694716453552, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.03870967741935484, | |
| "grad_norm": 7.008035670995982, | |
| "learning_rate": 2.9e-06, | |
| "log_odds_chosen": 0.9878906011581421, | |
| "log_odds_ratio": -0.35566407442092896, | |
| "logits/chosen": -1.28765869140625, | |
| "logits/rejected": 16.299999237060547, | |
| "logps/chosen": -1.661718726158142, | |
| "logps/rejected": -2.512500047683716, | |
| "loss": 1.9846, | |
| "nll_loss": 1.966406226158142, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -0.08310546725988388, | |
| "rewards/margins": 0.04267578199505806, | |
| "rewards/rejected": -0.12568359076976776, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.05161290322580645, | |
| "grad_norm": 5.261804897002812, | |
| "learning_rate": 3.900000000000001e-06, | |
| "log_odds_chosen": 0.8687499761581421, | |
| "log_odds_ratio": -0.3785156309604645, | |
| "logits/chosen": -0.40996092557907104, | |
| "logits/rejected": 16.706249237060547, | |
| "logps/chosen": -1.548437476158142, | |
| "logps/rejected": -2.284374952316284, | |
| "loss": 1.8731, | |
| "nll_loss": 1.8546874523162842, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -0.07744140923023224, | |
| "rewards/margins": 0.036865234375, | |
| "rewards/rejected": -0.11435546725988388, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.06451612903225806, | |
| "grad_norm": 3.9860274644043203, | |
| "learning_rate": 4.9000000000000005e-06, | |
| "log_odds_chosen": 0.748828113079071, | |
| "log_odds_ratio": -0.4212890565395355, | |
| "logits/chosen": 0.379638671875, | |
| "logits/rejected": 16.956249237060547, | |
| "logps/chosen": -1.467187523841858, | |
| "logps/rejected": -2.0882811546325684, | |
| "loss": 1.7385, | |
| "nll_loss": 1.7179687023162842, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -0.07333984225988388, | |
| "rewards/margins": 0.0311279296875, | |
| "rewards/rejected": -0.10444335639476776, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.07741935483870968, | |
| "grad_norm": 3.8803923662969386, | |
| "learning_rate": 5.9e-06, | |
| "log_odds_chosen": 0.49042969942092896, | |
| "log_odds_ratio": -0.5087890625, | |
| "logits/chosen": 0.905078113079071, | |
| "logits/rejected": 16.831249237060547, | |
| "logps/chosen": -1.4015624523162842, | |
| "logps/rejected": -1.791406273841858, | |
| "loss": 1.6458, | |
| "nll_loss": 1.62109375, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -0.070068359375, | |
| "rewards/margins": 0.019622802734375, | |
| "rewards/rejected": -0.089599609375, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.09032258064516129, | |
| "grad_norm": 2.9291160475725464, | |
| "learning_rate": 6.9e-06, | |
| "log_odds_chosen": 0.42890626192092896, | |
| "log_odds_ratio": -0.525585949420929, | |
| "logits/chosen": 1.0153319835662842, | |
| "logits/rejected": 17.274999618530273, | |
| "logps/chosen": -1.251562476158142, | |
| "logps/rejected": -1.5867187976837158, | |
| "loss": 1.4603, | |
| "nll_loss": 1.435156226158142, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -0.0626220703125, | |
| "rewards/margins": 0.01672973670065403, | |
| "rewards/rejected": -0.07929687201976776, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.1032258064516129, | |
| "grad_norm": 2.3895368722437746, | |
| "learning_rate": 7.9e-06, | |
| "log_odds_chosen": 0.20942382514476776, | |
| "log_odds_ratio": -0.6214843988418579, | |
| "logits/chosen": 1.1715819835662842, | |
| "logits/rejected": 17.331249237060547, | |
| "logps/chosen": -1.2156250476837158, | |
| "logps/rejected": -1.375, | |
| "loss": 1.3688, | |
| "nll_loss": 1.3390624523162842, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.06086425855755806, | |
| "rewards/margins": 0.0077911377884447575, | |
| "rewards/rejected": -0.06862793117761612, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.11612903225806452, | |
| "grad_norm": 2.1265814928397626, | |
| "learning_rate": 8.900000000000001e-06, | |
| "log_odds_chosen": 0.1634521484375, | |
| "log_odds_ratio": -0.649609386920929, | |
| "logits/chosen": 0.767163097858429, | |
| "logits/rejected": 17.131250381469727, | |
| "logps/chosen": -1.14453125, | |
| "logps/rejected": -1.26953125, | |
| "loss": 1.292, | |
| "nll_loss": 1.260156273841858, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.05727539211511612, | |
| "rewards/margins": 0.006190490908920765, | |
| "rewards/rejected": -0.06345214694738388, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.12903225806451613, | |
| "grad_norm": 2.459971487020149, | |
| "learning_rate": 9.9e-06, | |
| "log_odds_chosen": 0.14106445014476776, | |
| "log_odds_ratio": -0.6585937738418579, | |
| "logits/chosen": 0.06308593600988388, | |
| "logits/rejected": 16.306249618530273, | |
| "logps/chosen": -1.1515624523162842, | |
| "logps/rejected": -1.2664062976837158, | |
| "loss": 1.2808, | |
| "nll_loss": 1.247656226158142, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.05751953274011612, | |
| "rewards/margins": 0.005828857421875, | |
| "rewards/rejected": -0.06342773139476776, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.14193548387096774, | |
| "grad_norm": 2.4830433577540436, | |
| "learning_rate": 9.578262852211514e-06, | |
| "log_odds_chosen": 0.37322998046875, | |
| "log_odds_ratio": -0.560351550579071, | |
| "logits/chosen": 0.01640624925494194, | |
| "logits/rejected": 16.587499618530273, | |
| "logps/chosen": -1.126953125, | |
| "logps/rejected": -1.4070312976837158, | |
| "loss": 1.2312, | |
| "nll_loss": 1.2023437023162842, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.05644531175494194, | |
| "rewards/margins": 0.013935851864516735, | |
| "rewards/rejected": -0.07038573920726776, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.15483870967741936, | |
| "grad_norm": 8.467524109258711, | |
| "learning_rate": 9.166984970282113e-06, | |
| "log_odds_chosen": 0.886914074420929, | |
| "log_odds_ratio": -0.39765626192092896, | |
| "logits/chosen": -0.9283202886581421, | |
| "logits/rejected": 16.756250381469727, | |
| "logps/chosen": -1.103124976158142, | |
| "logps/rejected": -1.77734375, | |
| "loss": 1.1735, | |
| "nll_loss": 1.1531250476837158, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/chosen": -0.05517578125, | |
| "rewards/margins": 0.03373413160443306, | |
| "rewards/rejected": -0.08891601860523224, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.16774193548387098, | |
| "grad_norm": 2.492915426271658, | |
| "learning_rate": 8.804509063256239e-06, | |
| "log_odds_chosen": 1.274999976158142, | |
| "log_odds_ratio": -0.29912108182907104, | |
| "logits/chosen": -0.2676025331020355, | |
| "logits/rejected": 16.34375, | |
| "logps/chosen": -1.1281249523162842, | |
| "logps/rejected": -2.14453125, | |
| "loss": 1.1897, | |
| "nll_loss": 1.174218773841858, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -0.05637206882238388, | |
| "rewards/margins": 0.05068359524011612, | |
| "rewards/rejected": -0.10708007961511612, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.18064516129032257, | |
| "grad_norm": 2.259429852293854, | |
| "learning_rate": 8.481889296799709e-06, | |
| "log_odds_chosen": 1.3566405773162842, | |
| "log_odds_ratio": -0.2734375, | |
| "logits/chosen": -0.8609374761581421, | |
| "logits/rejected": 15.743749618530273, | |
| "logps/chosen": -1.103906273841858, | |
| "logps/rejected": -2.171875, | |
| "loss": 1.1491, | |
| "nll_loss": 1.1355469226837158, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.05512695387005806, | |
| "rewards/margins": 0.053466796875, | |
| "rewards/rejected": -0.10859374701976776, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.1935483870967742, | |
| "grad_norm": 2.3418100298894013, | |
| "learning_rate": 8.192319205190406e-06, | |
| "log_odds_chosen": 1.53125, | |
| "log_odds_ratio": -0.24775390326976776, | |
| "logits/chosen": -1.3630859851837158, | |
| "logits/rejected": 15.75, | |
| "logps/chosen": -1.092187523841858, | |
| "logps/rejected": -2.328125, | |
| "loss": 1.137, | |
| "nll_loss": 1.1242187023162842, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.05458984524011612, | |
| "rewards/margins": 0.06184082105755806, | |
| "rewards/rejected": -0.11635742336511612, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2064516129032258, | |
| "grad_norm": 3.440439256067863, | |
| "learning_rate": 7.930515857181441e-06, | |
| "log_odds_chosen": 1.54296875, | |
| "log_odds_ratio": -0.2630859315395355, | |
| "logits/chosen": -1.760156273841858, | |
| "logits/rejected": 15.199999809265137, | |
| "logps/chosen": -1.055078148841858, | |
| "logps/rejected": -2.293750047683716, | |
| "loss": 1.1039, | |
| "nll_loss": 1.0906250476837158, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -0.05275879055261612, | |
| "rewards/margins": 0.06201171875, | |
| "rewards/rejected": -0.11484374850988388, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.21935483870967742, | |
| "grad_norm": 2.237987495849613, | |
| "learning_rate": 7.692307692307692e-06, | |
| "log_odds_chosen": 1.7804687023162842, | |
| "log_odds_ratio": -0.20322266221046448, | |
| "logits/chosen": -1.688085913658142, | |
| "logits/rejected": 16.331249237060547, | |
| "logps/chosen": -1.01171875, | |
| "logps/rejected": -2.440624952316284, | |
| "loss": 1.0603, | |
| "nll_loss": 1.050390601158142, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.05051269382238388, | |
| "rewards/margins": 0.07155761867761612, | |
| "rewards/rejected": -0.12216796725988388, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.23225806451612904, | |
| "grad_norm": 3.0199229105723564, | |
| "learning_rate": 7.4743509275193595e-06, | |
| "log_odds_chosen": 1.7257812023162842, | |
| "log_odds_ratio": -0.20849609375, | |
| "logits/chosen": -1.4043457508087158, | |
| "logits/rejected": 15.712499618530273, | |
| "logps/chosen": -1.048828125, | |
| "logps/rejected": -2.4312500953674316, | |
| "loss": 1.075, | |
| "nll_loss": 1.0652344226837158, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.05239257961511612, | |
| "rewards/margins": 0.06918945163488388, | |
| "rewards/rejected": -0.12163086235523224, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.24516129032258063, | |
| "grad_norm": 6.554283287023061, | |
| "learning_rate": 7.27392967453308e-06, | |
| "log_odds_chosen": 1.7296874523162842, | |
| "log_odds_ratio": -0.21250000596046448, | |
| "logits/chosen": -0.91064453125, | |
| "logits/rejected": 16.212499618530273, | |
| "logps/chosen": -1.030859351158142, | |
| "logps/rejected": -2.418750047683716, | |
| "loss": 1.0727, | |
| "nll_loss": 1.0617187023162842, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -0.05158691480755806, | |
| "rewards/margins": 0.06936035305261612, | |
| "rewards/rejected": -0.12114258110523224, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.25806451612903225, | |
| "grad_norm": 2.3355013011756913, | |
| "learning_rate": 7.088812050083359e-06, | |
| "log_odds_chosen": 1.876562476158142, | |
| "log_odds_ratio": -0.18281249701976776, | |
| "logits/chosen": -1.1605956554412842, | |
| "logits/rejected": 16.456249237060547, | |
| "logps/chosen": -1.064453125, | |
| "logps/rejected": -2.590625047683716, | |
| "loss": 1.0956, | |
| "nll_loss": 1.0867187976837158, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.05319824069738388, | |
| "rewards/margins": 0.076416015625, | |
| "rewards/rejected": -0.12958984076976776, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.2709677419354839, | |
| "grad_norm": 2.663340541202873, | |
| "learning_rate": 6.917144638660746e-06, | |
| "log_odds_chosen": 1.8046875, | |
| "log_odds_ratio": -0.21416015923023224, | |
| "logits/chosen": -1.562109351158142, | |
| "logits/rejected": 15.65625, | |
| "logps/chosen": -1.050390601158142, | |
| "logps/rejected": -2.5093750953674316, | |
| "loss": 1.0719, | |
| "nll_loss": 1.060937523841858, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -0.05251464992761612, | |
| "rewards/margins": 0.07290039211511612, | |
| "rewards/rejected": -0.12548828125, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.2838709677419355, | |
| "grad_norm": 2.4359455832988774, | |
| "learning_rate": 6.757373783994861e-06, | |
| "log_odds_chosen": 1.8781249523162842, | |
| "log_odds_ratio": -0.21025390923023224, | |
| "logits/chosen": -0.8188232183456421, | |
| "logits/rejected": 16.125, | |
| "logps/chosen": -1.103906273841858, | |
| "logps/rejected": -2.6578125953674316, | |
| "loss": 1.13, | |
| "nll_loss": 1.119531273841858, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -0.05522460862994194, | |
| "rewards/margins": 0.07768554985523224, | |
| "rewards/rejected": -0.13300780951976776, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.2967741935483871, | |
| "grad_norm": 2.5381406849446044, | |
| "learning_rate": 6.6081860045508985e-06, | |
| "log_odds_chosen": 1.8273437023162842, | |
| "log_odds_ratio": -0.19570311903953552, | |
| "logits/chosen": -0.516894519329071, | |
| "logits/rejected": 15.912500381469727, | |
| "logps/chosen": -1.114843726158142, | |
| "logps/rejected": -2.606250047683716, | |
| "loss": 1.1259, | |
| "nll_loss": 1.1160156726837158, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.05568847805261612, | |
| "rewards/margins": 0.07460937649011612, | |
| "rewards/rejected": -0.13032226264476776, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.3096774193548387, | |
| "grad_norm": 2.249756481488481, | |
| "learning_rate": 6.46846227353151e-06, | |
| "log_odds_chosen": 1.7234375476837158, | |
| "log_odds_ratio": -0.22285155951976776, | |
| "logits/chosen": -1.727636694908142, | |
| "logits/rejected": 14.53125, | |
| "logps/chosen": -1.047265648841858, | |
| "logps/rejected": -2.4453125, | |
| "loss": 1.0836, | |
| "nll_loss": 1.0730469226837158, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.05241699144244194, | |
| "rewards/margins": 0.06984863430261612, | |
| "rewards/rejected": -0.12221679836511612, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.3225806451612903, | |
| "grad_norm": 2.4182796191238567, | |
| "learning_rate": 6.337242505244779e-06, | |
| "log_odds_chosen": 2.01171875, | |
| "log_odds_ratio": -0.17333984375, | |
| "logits/chosen": -2.0010743141174316, | |
| "logits/rejected": 15.53125, | |
| "logps/chosen": -1.0105469226837158, | |
| "logps/rejected": -2.6484375, | |
| "loss": 1.0448, | |
| "nll_loss": 1.03515625, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.05056152492761612, | |
| "rewards/margins": 0.08193359524011612, | |
| "rewards/rejected": -0.13247069716453552, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.33548387096774196, | |
| "grad_norm": 2.4058296328212965, | |
| "learning_rate": 6.213697660012001e-06, | |
| "log_odds_chosen": 1.8125, | |
| "log_odds_ratio": -0.2138671875, | |
| "logits/chosen": -1.9609375, | |
| "logits/rejected": 15.487500190734863, | |
| "logps/chosen": -1.032812476158142, | |
| "logps/rejected": -2.5, | |
| "loss": 1.0654, | |
| "nll_loss": 1.054296851158142, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.05168456956744194, | |
| "rewards/margins": 0.0733642578125, | |
| "rewards/rejected": -0.12495116889476776, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.34838709677419355, | |
| "grad_norm": 2.448833917691866, | |
| "learning_rate": 6.097107608496925e-06, | |
| "log_odds_chosen": 2.004687547683716, | |
| "log_odds_ratio": -0.18144531548023224, | |
| "logits/chosen": -1.5515625476837158, | |
| "logits/rejected": 15.40625, | |
| "logps/chosen": -1.0011718273162842, | |
| "logps/rejected": -2.6343750953674316, | |
| "loss": 1.0323, | |
| "nll_loss": 1.0226562023162842, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.050048828125, | |
| "rewards/margins": 0.08164062350988388, | |
| "rewards/rejected": -0.1318359375, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.36129032258064514, | |
| "grad_norm": 2.7061200750489762, | |
| "learning_rate": 5.986843400892497e-06, | |
| "log_odds_chosen": 2.0953125953674316, | |
| "log_odds_ratio": -0.16904297471046448, | |
| "logits/chosen": -1.35595703125, | |
| "logits/rejected": 15.625, | |
| "logps/chosen": -1.0671875476837158, | |
| "logps/rejected": -2.792187452316284, | |
| "loss": 1.0819, | |
| "nll_loss": 1.073828101158142, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.05336914211511612, | |
| "rewards/margins": 0.08642578125, | |
| "rewards/rejected": -0.13984374701976776, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.3741935483870968, | |
| "grad_norm": 2.4367758380161573, | |
| "learning_rate": 5.882352941176471e-06, | |
| "log_odds_chosen": 2.0132813453674316, | |
| "log_odds_ratio": -0.1748046875, | |
| "logits/chosen": -1.527734398841858, | |
| "logits/rejected": 15.881250381469727, | |
| "logps/chosen": -1.030859351158142, | |
| "logps/rejected": -2.6656250953674316, | |
| "loss": 1.0479, | |
| "nll_loss": 1.0390625, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.051513671875, | |
| "rewards/margins": 0.08173827826976776, | |
| "rewards/rejected": -0.13325195014476776, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.3870967741935484, | |
| "grad_norm": 3.3727479977644363, | |
| "learning_rate": 5.783149319662402e-06, | |
| "log_odds_chosen": 2.188281297683716, | |
| "log_odds_ratio": -0.182373046875, | |
| "logits/chosen": -1.51806640625, | |
| "logits/rejected": 15.743749618530273, | |
| "logps/chosen": -1.0539062023162842, | |
| "logps/rejected": -2.879687547683716, | |
| "loss": 1.0747, | |
| "nll_loss": 1.0656249523162842, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -0.05266113206744194, | |
| "rewards/margins": 0.09116210788488388, | |
| "rewards/rejected": -0.14374999701976776, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 2.812934888942385, | |
| "learning_rate": 5.688801239885744e-06, | |
| "log_odds_chosen": 2.254687547683716, | |
| "log_odds_ratio": -0.17075195908546448, | |
| "logits/chosen": -1.6027343273162842, | |
| "logits/rejected": 15.181249618530273, | |
| "logps/chosen": -1.0187499523162842, | |
| "logps/rejected": -2.8968749046325684, | |
| "loss": 1.0446, | |
| "nll_loss": 1.036718726158142, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -0.05092773586511612, | |
| "rewards/margins": 0.09372558444738388, | |
| "rewards/rejected": -0.14467772841453552, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.4129032258064516, | |
| "grad_norm": 2.751505791676006, | |
| "learning_rate": 5.598925109558544e-06, | |
| "log_odds_chosen": 2.2125000953674316, | |
| "log_odds_ratio": -0.169677734375, | |
| "logits/chosen": -1.304540991783142, | |
| "logits/rejected": 15.78125, | |
| "logps/chosen": -1.0320312976837158, | |
| "logps/rejected": -2.8687500953674316, | |
| "loss": 1.0443, | |
| "nll_loss": 1.036718726158142, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -0.05166015774011612, | |
| "rewards/margins": 0.09177245944738388, | |
| "rewards/rejected": -0.14340820908546448, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.4258064516129032, | |
| "grad_norm": 3.3773353240681008, | |
| "learning_rate": 5.513178464199713e-06, | |
| "log_odds_chosen": 2.3843750953674316, | |
| "log_odds_ratio": -0.14497070014476776, | |
| "logits/chosen": -1.611181616783142, | |
| "logits/rejected": 15.831250190734863, | |
| "logps/chosen": -1.0519530773162842, | |
| "logps/rejected": -3.057812452316284, | |
| "loss": 1.0883, | |
| "nll_loss": 1.08203125, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.0526123046875, | |
| "rewards/margins": 0.10014648735523224, | |
| "rewards/rejected": -0.15273436903953552, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.43870967741935485, | |
| "grad_norm": 2.9870987656093315, | |
| "learning_rate": 5.4312544659356846e-06, | |
| "log_odds_chosen": 2.4984374046325684, | |
| "log_odds_ratio": -0.13447265326976776, | |
| "logits/chosen": -1.1388671398162842, | |
| "logits/rejected": 15.675000190734863, | |
| "logps/chosen": -1.064062476158142, | |
| "logps/rejected": -3.1781249046325684, | |
| "loss": 1.0626, | |
| "nll_loss": 1.0574219226837158, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.05322265625, | |
| "rewards/margins": 0.10561523586511612, | |
| "rewards/rejected": -0.15878906846046448, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.45161290322580644, | |
| "grad_norm": 2.83867528933509, | |
| "learning_rate": 5.352877275721893e-06, | |
| "log_odds_chosen": 2.5406250953674316, | |
| "log_odds_ratio": -0.14262695610523224, | |
| "logits/chosen": -1.796777367591858, | |
| "logits/rejected": 15.524999618530273, | |
| "logps/chosen": -1.0109374523162842, | |
| "logps/rejected": -3.1484375, | |
| "loss": 1.0423, | |
| "nll_loss": 1.03515625, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.05056152492761612, | |
| "rewards/margins": 0.10683593899011612, | |
| "rewards/rejected": -0.15742187201976776, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.4645161290322581, | |
| "grad_norm": 2.9913815230527505, | |
| "learning_rate": 5.277798139692596e-06, | |
| "log_odds_chosen": 2.6484375, | |
| "log_odds_ratio": -0.13837890326976776, | |
| "logits/chosen": -1.791015625, | |
| "logits/rejected": 15.574999809265137, | |
| "logps/chosen": -1.0304687023162842, | |
| "logps/rejected": -3.284374952316284, | |
| "loss": 1.05, | |
| "nll_loss": 1.0437500476837158, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -0.051513671875, | |
| "rewards/margins": 0.11279296875, | |
| "rewards/rejected": -0.16445311903953552, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.4774193548387097, | |
| "grad_norm": 6.430960743025478, | |
| "learning_rate": 5.205792062953535e-06, | |
| "log_odds_chosen": 2.629687547683716, | |
| "log_odds_ratio": -0.14753417670726776, | |
| "logits/chosen": -1.828125, | |
| "logits/rejected": 15.9375, | |
| "logps/chosen": -1.038671851158142, | |
| "logps/rejected": -3.278125047683716, | |
| "loss": 1.0531, | |
| "nll_loss": 1.0457031726837158, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -0.05190429836511612, | |
| "rewards/margins": 0.112060546875, | |
| "rewards/rejected": -0.16396483778953552, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.49032258064516127, | |
| "grad_norm": 2.3749325025222587, | |
| "learning_rate": 5.136654969375142e-06, | |
| "log_odds_chosen": 2.176562547683716, | |
| "log_odds_ratio": -0.16562500596046448, | |
| "logits/chosen": -1.7742187976837158, | |
| "logits/rejected": 15.612500190734863, | |
| "logps/chosen": -1.0148437023162842, | |
| "logps/rejected": -2.7984375953674316, | |
| "loss": 1.0246, | |
| "nll_loss": 1.015234351158142, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.05070800706744194, | |
| "rewards/margins": 0.08916015923023224, | |
| "rewards/rejected": -0.139892578125, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.5032258064516129, | |
| "grad_norm": 2.9704201026587405, | |
| "learning_rate": 5.070201265633938e-06, | |
| "log_odds_chosen": 2.112499952316284, | |
| "log_odds_ratio": -0.1689453125, | |
| "logits/chosen": -1.837499976158142, | |
| "logits/rejected": 15.824999809265137, | |
| "logps/chosen": -1.0246093273162842, | |
| "logps/rejected": -2.7671875953674316, | |
| "loss": 1.0501, | |
| "nll_loss": 1.0421874523162842, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.05129394680261612, | |
| "rewards/margins": 0.08686523139476776, | |
| "rewards/rejected": -0.13833007216453552, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.5161290322580645, | |
| "grad_norm": 2.7572182399385303, | |
| "learning_rate": 5.006261743217589e-06, | |
| "log_odds_chosen": 2.305468797683716, | |
| "log_odds_ratio": -0.14760741591453552, | |
| "logits/chosen": -2.180468797683716, | |
| "logits/rejected": 15.449999809265137, | |
| "logps/chosen": -1.017968773841858, | |
| "logps/rejected": -2.9203124046325684, | |
| "loss": 1.0379, | |
| "nll_loss": 1.030859351158142, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.05087890475988388, | |
| "rewards/margins": 0.09526367485523224, | |
| "rewards/rejected": -0.14633789658546448, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5290322580645161, | |
| "grad_norm": 2.988129143495499, | |
| "learning_rate": 4.944681764341488e-06, | |
| "log_odds_chosen": 2.643749952316284, | |
| "log_odds_ratio": -0.12158203125, | |
| "logits/chosen": -1.9125487804412842, | |
| "logits/rejected": 15.543749809265137, | |
| "logps/chosen": -1.0125000476837158, | |
| "logps/rejected": -3.2484374046325684, | |
| "loss": 1.0289, | |
| "nll_loss": 1.023046851158142, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.0506591796875, | |
| "rewards/margins": 0.11171875149011612, | |
| "rewards/rejected": -0.16245117783546448, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.5419354838709678, | |
| "grad_norm": 3.3731155020181154, | |
| "learning_rate": 4.885319687460315e-06, | |
| "log_odds_chosen": 2.682812452316284, | |
| "log_odds_ratio": -0.14687499403953552, | |
| "logits/chosen": -1.4832031726837158, | |
| "logits/rejected": 15.387499809265137, | |
| "logps/chosen": -1.058203101158142, | |
| "logps/rejected": -3.375, | |
| "loss": 1.0678, | |
| "nll_loss": 1.060156226158142, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -0.05280761793255806, | |
| "rewards/margins": 0.115966796875, | |
| "rewards/rejected": -0.16875000298023224, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.5548387096774193, | |
| "grad_norm": 11.540099842491083, | |
| "learning_rate": 4.8280454958526765e-06, | |
| "log_odds_chosen": 2.6937499046325684, | |
| "log_odds_ratio": -0.13825683295726776, | |
| "logits/chosen": -0.983593761920929, | |
| "logits/rejected": 16.268749237060547, | |
| "logps/chosen": -1.044531226158142, | |
| "logps/rejected": -3.346874952316284, | |
| "loss": 1.0426, | |
| "nll_loss": 1.0363280773162842, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.05219726637005806, | |
| "rewards/margins": 0.11508788913488388, | |
| "rewards/rejected": -0.16728515923023224, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.567741935483871, | |
| "grad_norm": 3.6791814532749805, | |
| "learning_rate": 4.7727395990334775e-06, | |
| "log_odds_chosen": 2.5210938453674316, | |
| "log_odds_ratio": -0.14609375596046448, | |
| "logits/chosen": -1.860937476158142, | |
| "logits/rejected": 15.625, | |
| "logps/chosen": -1.04296875, | |
| "logps/rejected": -3.1859374046325684, | |
| "loss": 1.0679, | |
| "nll_loss": 1.060546875, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.05214843899011612, | |
| "rewards/margins": 0.10712890326976776, | |
| "rewards/rejected": -0.15937499701976776, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.5806451612903226, | |
| "grad_norm": 2.7953578580150538, | |
| "learning_rate": 4.719291781830088e-06, | |
| "log_odds_chosen": 2.6546874046325684, | |
| "log_odds_ratio": -0.14821776747703552, | |
| "logits/chosen": -1.664709448814392, | |
| "logits/rejected": 15.356249809265137, | |
| "logps/chosen": -1.074609398841858, | |
| "logps/rejected": -3.3515625, | |
| "loss": 1.0897, | |
| "nll_loss": 1.08203125, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.05378418043255806, | |
| "rewards/margins": 0.11376953125, | |
| "rewards/rejected": -0.16757813096046448, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.5935483870967742, | |
| "grad_norm": 3.005212830785336, | |
| "learning_rate": 4.667600280093366e-06, | |
| "log_odds_chosen": 3.120312452316284, | |
| "log_odds_ratio": -0.12114258110523224, | |
| "logits/chosen": -1.77734375, | |
| "logits/rejected": 15.712499618530273, | |
| "logps/chosen": -1.0207030773162842, | |
| "logps/rejected": -3.7265625, | |
| "loss": 1.0381, | |
| "nll_loss": 1.0324218273162842, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.05104980617761612, | |
| "rewards/margins": 0.13530273735523224, | |
| "rewards/rejected": -0.18632812798023224, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.6064516129032258, | |
| "grad_norm": 3.459726587570506, | |
| "learning_rate": 4.6175709653961015e-06, | |
| "log_odds_chosen": 2.854687452316284, | |
| "log_odds_ratio": -0.16249999403953552, | |
| "logits/chosen": -1.89453125, | |
| "logits/rejected": 15.081250190734863, | |
| "logps/chosen": -1.0656249523162842, | |
| "logps/rejected": -3.5640625953674316, | |
| "loss": 1.0723, | |
| "nll_loss": 1.0652344226837158, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/chosen": -0.05327148362994194, | |
| "rewards/margins": 0.124755859375, | |
| "rewards/rejected": -0.17812499403953552, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.6193548387096774, | |
| "grad_norm": 3.0057455144663665, | |
| "learning_rate": 4.569116623849539e-06, | |
| "log_odds_chosen": 3.1343750953674316, | |
| "log_odds_ratio": -0.11713866889476776, | |
| "logits/chosen": -1.96875, | |
| "logits/rejected": 15.893750190734863, | |
| "logps/chosen": -0.9652343988418579, | |
| "logps/rejected": -3.667187452316284, | |
| "loss": 0.9906, | |
| "nll_loss": 0.984375, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.04826660081744194, | |
| "rewards/margins": 0.13496093451976776, | |
| "rewards/rejected": -0.18330077826976776, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.632258064516129, | |
| "grad_norm": 3.351920982571353, | |
| "learning_rate": 4.522156316461347e-06, | |
| "log_odds_chosen": 2.721874952316284, | |
| "log_odds_ratio": -0.12812499701976776, | |
| "logits/chosen": -1.908203125, | |
| "logits/rejected": 16.024999618530273, | |
| "logps/chosen": -1.060937523841858, | |
| "logps/rejected": -3.395312547683716, | |
| "loss": 1.0772, | |
| "nll_loss": 1.070703148841858, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.05305175855755806, | |
| "rewards/margins": 0.11660156399011612, | |
| "rewards/rejected": -0.16972656548023224, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.6451612903225806, | |
| "grad_norm": 23.089712113718853, | |
| "learning_rate": 4.476614810358452e-06, | |
| "log_odds_chosen": 2.9468750953674316, | |
| "log_odds_ratio": -0.14716796576976776, | |
| "logits/chosen": -1.800390601158142, | |
| "logits/rejected": 15.518750190734863, | |
| "logps/chosen": -1.0910155773162842, | |
| "logps/rejected": -3.6624999046325684, | |
| "loss": 1.0891, | |
| "nll_loss": 1.082421898841858, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -0.05458984524011612, | |
| "rewards/margins": 0.12851563096046448, | |
| "rewards/rejected": -0.18310546875, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.6580645161290323, | |
| "grad_norm": 2.9945598420779853, | |
| "learning_rate": 4.432422071779363e-06, | |
| "log_odds_chosen": 3.317187547683716, | |
| "log_odds_ratio": -0.10064697265625, | |
| "logits/chosen": -2.010937452316284, | |
| "logits/rejected": 15.393750190734863, | |
| "logps/chosen": -0.9996093511581421, | |
| "logps/rejected": -3.8921875953674316, | |
| "loss": 1.0307, | |
| "nll_loss": 1.02734375, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.04997558519244194, | |
| "rewards/margins": 0.14467772841453552, | |
| "rewards/rejected": -0.19462890923023224, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.6709677419354839, | |
| "grad_norm": 7.310800315863408, | |
| "learning_rate": 4.389512813061471e-06, | |
| "log_odds_chosen": 3.265625, | |
| "log_odds_ratio": -0.10830078274011612, | |
| "logits/chosen": -2.186718702316284, | |
| "logits/rejected": 15.712499618530273, | |
| "logps/chosen": -1.0011718273162842, | |
| "logps/rejected": -3.840625047683716, | |
| "loss": 1.0184, | |
| "nll_loss": 1.014062523841858, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.05007324367761612, | |
| "rewards/margins": 0.14218750596046448, | |
| "rewards/rejected": -0.19218750298023224, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.6838709677419355, | |
| "grad_norm": 2.9539080812054275, | |
| "learning_rate": 4.347826086956522e-06, | |
| "log_odds_chosen": 3.0718750953674316, | |
| "log_odds_ratio": -0.12153320014476776, | |
| "logits/chosen": -0.94921875, | |
| "logits/rejected": 15.306249618530273, | |
| "logps/chosen": -0.98828125, | |
| "logps/rejected": -3.637500047683716, | |
| "loss": 1.0124, | |
| "nll_loss": 1.005859375, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.04941406100988388, | |
| "rewards/margins": 0.13232421875, | |
| "rewards/rejected": -0.18173828721046448, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.6967741935483871, | |
| "grad_norm": 2.8332863620261124, | |
| "learning_rate": 4.3073049225394805e-06, | |
| "log_odds_chosen": 3.135937452316284, | |
| "log_odds_ratio": -0.1195068359375, | |
| "logits/chosen": -1.21875, | |
| "logits/rejected": 16.206249237060547, | |
| "logps/chosen": -1.0390625, | |
| "logps/rejected": -3.745312452316284, | |
| "loss": 1.0254, | |
| "nll_loss": 1.019140601158142, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.05195312574505806, | |
| "rewards/margins": 0.13520507514476776, | |
| "rewards/rejected": -0.18710938096046448, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.7096774193548387, | |
| "grad_norm": 3.316803771710403, | |
| "learning_rate": 4.2678959977632e-06, | |
| "log_odds_chosen": 3.0453124046325684, | |
| "log_odds_ratio": -0.105224609375, | |
| "logits/chosen": -1.7107422351837158, | |
| "logits/rejected": 15.649999618530273, | |
| "logps/chosen": -1.0378906726837158, | |
| "logps/rejected": -3.674999952316284, | |
| "loss": 1.0426, | |
| "nll_loss": 1.037500023841858, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.05195312574505806, | |
| "rewards/margins": 0.13188476860523224, | |
| "rewards/rejected": -0.18369141221046448, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.7225806451612903, | |
| "grad_norm": 2.730881069069837, | |
| "learning_rate": 4.229549344378136e-06, | |
| "log_odds_chosen": 3.1968750953674316, | |
| "log_odds_ratio": -0.110595703125, | |
| "logits/chosen": -1.8035156726837158, | |
| "logits/rejected": 15.556249618530273, | |
| "logps/chosen": -1.029687523841858, | |
| "logps/rejected": -3.8203125, | |
| "loss": 1.0507, | |
| "nll_loss": 1.044921875, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.05146484449505806, | |
| "rewards/margins": 0.13959960639476776, | |
| "rewards/rejected": -0.19091796875, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.7354838709677419, | |
| "grad_norm": 5.293665189129163, | |
| "learning_rate": 4.192218081503185e-06, | |
| "log_odds_chosen": 3.5, | |
| "log_odds_ratio": -0.08615722507238388, | |
| "logits/chosen": -1.8792235851287842, | |
| "logits/rejected": 15.631250381469727, | |
| "logps/chosen": -0.999218761920929, | |
| "logps/rejected": -4.064062595367432, | |
| "loss": 1.017, | |
| "nll_loss": 1.0125000476837158, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.04990234225988388, | |
| "rewards/margins": 0.15317383408546448, | |
| "rewards/rejected": -0.203125, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.7483870967741936, | |
| "grad_norm": 2.7468345646696406, | |
| "learning_rate": 4.155858174616914e-06, | |
| "log_odds_chosen": 3.2906250953674316, | |
| "log_odds_ratio": -0.0904541015625, | |
| "logits/chosen": -2.1302733421325684, | |
| "logits/rejected": 15.012499809265137, | |
| "logps/chosen": -0.977343738079071, | |
| "logps/rejected": -3.8218750953674316, | |
| "loss": 1.0009, | |
| "nll_loss": 0.997265636920929, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.048828125, | |
| "rewards/margins": 0.14199218153953552, | |
| "rewards/rejected": -0.19082030653953552, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.7612903225806451, | |
| "grad_norm": 3.0853307261867635, | |
| "learning_rate": 4.120428217151647e-06, | |
| "log_odds_chosen": 3.4124999046325684, | |
| "log_odds_ratio": -0.09726562350988388, | |
| "logits/chosen": -2.3573241233825684, | |
| "logits/rejected": 15.337499618530273, | |
| "logps/chosen": -0.975390613079071, | |
| "logps/rejected": -3.9468750953674316, | |
| "loss": 1.0064, | |
| "nll_loss": 1.0011718273162842, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.0487060546875, | |
| "rewards/margins": 0.14853516221046448, | |
| "rewards/rejected": -0.197265625, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.7741935483870968, | |
| "grad_norm": 2.7361001609622426, | |
| "learning_rate": 4.0858892322271844e-06, | |
| "log_odds_chosen": 3.567187547683716, | |
| "log_odds_ratio": -0.08948974311351776, | |
| "logits/chosen": -1.8064453601837158, | |
| "logits/rejected": 15.625, | |
| "logps/chosen": -1.0207030773162842, | |
| "logps/rejected": -4.165625095367432, | |
| "loss": 1.0415, | |
| "nll_loss": 1.035546898841858, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.05104980617761612, | |
| "rewards/margins": 0.15751953423023224, | |
| "rewards/rejected": -0.20839843153953552, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.7870967741935484, | |
| "grad_norm": 3.4721246535677714, | |
| "learning_rate": 4.052204492365539e-06, | |
| "log_odds_chosen": 3.9296875, | |
| "log_odds_ratio": -0.07633056491613388, | |
| "logits/chosen": -2.706249952316284, | |
| "logits/rejected": 15.518750190734863, | |
| "logps/chosen": -1.033593773841858, | |
| "logps/rejected": -4.540625095367432, | |
| "loss": 1.0518, | |
| "nll_loss": 1.0476562976837158, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.0516357421875, | |
| "rewards/margins": 0.17529296875, | |
| "rewards/rejected": -0.22685547173023224, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 3.6600611060580515, | |
| "learning_rate": 4.019339355290704e-06, | |
| "log_odds_chosen": 3.6156249046325684, | |
| "log_odds_ratio": -0.1014404296875, | |
| "logits/chosen": -2.306640625, | |
| "logits/rejected": 15.862500190734863, | |
| "logps/chosen": -1.014062523841858, | |
| "logps/rejected": -4.1953125, | |
| "loss": 1.0323, | |
| "nll_loss": 1.027734398841858, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.05063476413488388, | |
| "rewards/margins": 0.15908202528953552, | |
| "rewards/rejected": -0.20966796576976776, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.8129032258064516, | |
| "grad_norm": 3.044997462446035, | |
| "learning_rate": 3.9872611141445005e-06, | |
| "log_odds_chosen": 3.4546875953674316, | |
| "log_odds_ratio": -0.09210205078125, | |
| "logits/chosen": -2.1363282203674316, | |
| "logits/rejected": 15.725000381469727, | |
| "logps/chosen": -0.983203113079071, | |
| "logps/rejected": -3.9921875, | |
| "loss": 0.9936, | |
| "nll_loss": 0.9898437261581421, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.04914550855755806, | |
| "rewards/margins": 0.15019531548023224, | |
| "rewards/rejected": -0.19951172173023224, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.8258064516129032, | |
| "grad_norm": 3.972247288109564, | |
| "learning_rate": 3.955938860646178e-06, | |
| "log_odds_chosen": 3.6546874046325684, | |
| "log_odds_ratio": -0.09271240234375, | |
| "logits/chosen": -1.830078125, | |
| "logits/rejected": 15.8125, | |
| "logps/chosen": -1.047265648841858, | |
| "logps/rejected": -4.262499809265137, | |
| "loss": 1.0426, | |
| "nll_loss": 1.0382812023162842, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.0523681640625, | |
| "rewards/margins": 0.16064453125, | |
| "rewards/rejected": -0.21318359673023224, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.8387096774193549, | |
| "grad_norm": 3.2312016718085466, | |
| "learning_rate": 3.925343359894298e-06, | |
| "log_odds_chosen": 4.076562404632568, | |
| "log_odds_ratio": -0.06367187201976776, | |
| "logits/chosen": -1.952734351158142, | |
| "logits/rejected": 16.381250381469727, | |
| "logps/chosen": -1.0363280773162842, | |
| "logps/rejected": -4.678124904632568, | |
| "loss": 1.0545, | |
| "nll_loss": 1.051171898841858, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.05180663987994194, | |
| "rewards/margins": 0.18183593451976776, | |
| "rewards/rejected": -0.23378905653953552, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.8516129032258064, | |
| "grad_norm": 3.1542224686668425, | |
| "learning_rate": 3.895446935658099e-06, | |
| "log_odds_chosen": 3.5687499046325684, | |
| "log_odds_ratio": -0.10651855170726776, | |
| "logits/chosen": -1.4596679210662842, | |
| "logits/rejected": 15.725000381469727, | |
| "logps/chosen": -1.042578101158142, | |
| "logps/rejected": -4.209374904632568, | |
| "loss": 1.047, | |
| "nll_loss": 1.041015625, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -0.05217285081744194, | |
| "rewards/margins": 0.15805664658546448, | |
| "rewards/rejected": -0.21044921875, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.864516129032258, | |
| "grad_norm": 2.76746682960287, | |
| "learning_rate": 3.866223365135295e-06, | |
| "log_odds_chosen": 3.481250047683716, | |
| "log_odds_ratio": -0.09975586086511612, | |
| "logits/chosen": -2.1744141578674316, | |
| "logits/rejected": 15.537500381469727, | |
| "logps/chosen": -1.017968773841858, | |
| "logps/rejected": -4.071875095367432, | |
| "loss": 1.0355, | |
| "nll_loss": 1.03125, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.0509033203125, | |
| "rewards/margins": 0.15253905951976776, | |
| "rewards/rejected": -0.20341797173023224, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.8774193548387097, | |
| "grad_norm": 3.339669169989708, | |
| "learning_rate": 3.837647782266681e-06, | |
| "log_odds_chosen": 3.8062500953674316, | |
| "log_odds_ratio": -0.08485107123851776, | |
| "logits/chosen": -1.7482421398162842, | |
| "logits/rejected": 15.668749809265137, | |
| "logps/chosen": -1.081640601158142, | |
| "logps/rejected": -4.474999904632568, | |
| "loss": 1.0689, | |
| "nll_loss": 1.064062476158142, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.05405273288488388, | |
| "rewards/margins": 0.16982421278953552, | |
| "rewards/rejected": -0.22402343153953552, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.8903225806451613, | |
| "grad_norm": 3.216293567417128, | |
| "learning_rate": 3.8096965887972973e-06, | |
| "log_odds_chosen": 3.6656250953674316, | |
| "log_odds_ratio": -0.1185302734375, | |
| "logits/chosen": -1.7449219226837158, | |
| "logits/rejected": 15.34375, | |
| "logps/chosen": -1.060937523841858, | |
| "logps/rejected": -4.292187690734863, | |
| "loss": 1.0248, | |
| "nll_loss": 1.019921898841858, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.05307617038488388, | |
| "rewards/margins": 0.16152343153953552, | |
| "rewards/rejected": -0.21445313096046448, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.9032258064516129, | |
| "grad_norm": 3.165994900484303, | |
| "learning_rate": 3.7823473723611694e-06, | |
| "log_odds_chosen": 3.528125047683716, | |
| "log_odds_ratio": -0.09687499701976776, | |
| "logits/chosen": -1.868749976158142, | |
| "logits/rejected": 15.487500190734863, | |
| "logps/chosen": -0.983593761920929, | |
| "logps/rejected": -4.068749904632568, | |
| "loss": 0.9975, | |
| "nll_loss": 0.993359386920929, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.0491943359375, | |
| "rewards/margins": 0.154296875, | |
| "rewards/rejected": -0.20351561903953552, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.9161290322580645, | |
| "grad_norm": 3.1804898751694877, | |
| "learning_rate": 3.7555788309433988e-06, | |
| "log_odds_chosen": 3.59375, | |
| "log_odds_ratio": -0.10922851413488388, | |
| "logits/chosen": -2.244140625, | |
| "logits/rejected": 15.425000190734863, | |
| "logps/chosen": -1.0078125, | |
| "logps/rejected": -4.1875, | |
| "loss": 1.0329, | |
| "nll_loss": 1.0265624523162842, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -0.0504150390625, | |
| "rewards/margins": 0.15878906846046448, | |
| "rewards/rejected": -0.20937499403953552, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.9290322580645162, | |
| "grad_norm": 2.8833066776928478, | |
| "learning_rate": 3.729370703141008e-06, | |
| "log_odds_chosen": 3.8125, | |
| "log_odds_ratio": -0.07578124850988388, | |
| "logits/chosen": -1.795996069908142, | |
| "logits/rejected": 15.637499809265137, | |
| "logps/chosen": -0.9906250238418579, | |
| "logps/rejected": -4.357812404632568, | |
| "loss": 1.0145, | |
| "nll_loss": 1.0105469226837158, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.04951171949505806, | |
| "rewards/margins": 0.16845703125, | |
| "rewards/rejected": -0.21787109971046448, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.9419354838709677, | |
| "grad_norm": 3.5368171320315573, | |
| "learning_rate": 3.7037037037037037e-06, | |
| "log_odds_chosen": 3.534374952316284, | |
| "log_odds_ratio": -0.09101562201976776, | |
| "logits/chosen": -2.108203172683716, | |
| "logits/rejected": 15.462499618530273, | |
| "logps/chosen": -1.011328101158142, | |
| "logps/rejected": -4.114062309265137, | |
| "loss": 1.0332, | |
| "nll_loss": 1.029296875, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.05058593675494194, | |
| "rewards/margins": 0.15507812798023224, | |
| "rewards/rejected": -0.20576171576976776, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.9548387096774194, | |
| "grad_norm": 2.88455714823336, | |
| "learning_rate": 3.6785594638885484e-06, | |
| "log_odds_chosen": 3.4359374046325684, | |
| "log_odds_ratio": -0.09562988579273224, | |
| "logits/chosen": -2.369335889816284, | |
| "logits/rejected": 15.0625, | |
| "logps/chosen": -0.989062488079071, | |
| "logps/rejected": -3.987499952316284, | |
| "loss": 0.9878, | |
| "nll_loss": 0.9839843511581421, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.0494384765625, | |
| "rewards/margins": 0.14960937201976776, | |
| "rewards/rejected": -0.19912108778953552, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.967741935483871, | |
| "grad_norm": 3.3000700003310293, | |
| "learning_rate": 3.6539204762093485e-06, | |
| "log_odds_chosen": 3.807812452316284, | |
| "log_odds_ratio": -0.07308349758386612, | |
| "logits/chosen": -2.13671875, | |
| "logits/rejected": 15.693750381469727, | |
| "logps/chosen": -1.0265624523162842, | |
| "logps/rejected": -4.396874904632568, | |
| "loss": 1.0483, | |
| "nll_loss": 1.044921875, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.05136718600988388, | |
| "rewards/margins": 0.16835936903953552, | |
| "rewards/rejected": -0.21962890028953552, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.9806451612903225, | |
| "grad_norm": 3.1320983844464814, | |
| "learning_rate": 3.6297700432031386e-06, | |
| "log_odds_chosen": 3.5453124046325684, | |
| "log_odds_ratio": -0.101806640625, | |
| "logits/chosen": -1.9267578125, | |
| "logits/rejected": 15.40625, | |
| "logps/chosen": -1.008203148841858, | |
| "logps/rejected": -4.128125190734863, | |
| "loss": 1.0326, | |
| "nll_loss": 1.0285155773162842, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.05043945461511612, | |
| "rewards/margins": 0.15605469048023224, | |
| "rewards/rejected": -0.20624999701976776, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.9935483870967742, | |
| "grad_norm": 5.872666548921313, | |
| "learning_rate": 3.606092229873094e-06, | |
| "log_odds_chosen": 3.6624999046325684, | |
| "log_odds_ratio": -0.09074707329273224, | |
| "logits/chosen": -2.439453125, | |
| "logits/rejected": 15.293749809265137, | |
| "logps/chosen": -1.0261719226837158, | |
| "logps/rejected": -4.246874809265137, | |
| "loss": 1.0374, | |
| "nll_loss": 1.0324218273162842, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.05126953125, | |
| "rewards/margins": 0.16127929091453552, | |
| "rewards/rejected": -0.21230468153953552, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 775, | |
| "total_flos": 0.0, | |
| "train_loss": 1.1359047379032259, | |
| "train_runtime": 1458.2503, | |
| "train_samples_per_second": 8.503, | |
| "train_steps_per_second": 0.531 | |
| } | |
| ], | |
| "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 | |
| } | |