| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 100, | |
| "global_step": 385, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 1.282051282051282e-07, | |
| "logits/chosen": -1.7278180122375488, | |
| "logits/rejected": -1.7377450466156006, | |
| "logps/chosen": -29.553977966308594, | |
| "logps/rejected": -42.813133239746094, | |
| "loss": 0.5, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.282051282051282e-06, | |
| "logits/chosen": -1.8668408393859863, | |
| "logits/rejected": -1.8711602687835693, | |
| "logps/chosen": -36.98978042602539, | |
| "logps/rejected": -33.66878890991211, | |
| "loss": 0.4962, | |
| "rewards/accuracies": 0.5416666865348816, | |
| "rewards/chosen": 0.006728413049131632, | |
| "rewards/margins": 0.022216208279132843, | |
| "rewards/rejected": -0.015487794764339924, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2.564102564102564e-06, | |
| "logits/chosen": -1.9970359802246094, | |
| "logits/rejected": -1.9996885061264038, | |
| "logps/chosen": -29.635208129882812, | |
| "logps/rejected": -29.063350677490234, | |
| "loss": 0.4994, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": 0.002794977743178606, | |
| "rewards/margins": 0.002757500857114792, | |
| "rewards/rejected": 3.747665323317051e-05, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 3.846153846153847e-06, | |
| "logits/chosen": -1.9208602905273438, | |
| "logits/rejected": -1.9181534051895142, | |
| "logps/chosen": -31.40317726135254, | |
| "logps/rejected": -33.23335647583008, | |
| "loss": 0.4982, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": 0.005168457515537739, | |
| "rewards/margins": 0.011126698926091194, | |
| "rewards/rejected": -0.005958239547908306, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.999896948438434e-06, | |
| "logits/chosen": -2.017474889755249, | |
| "logits/rejected": -2.008759021759033, | |
| "logps/chosen": -32.54490661621094, | |
| "logps/rejected": -32.49110794067383, | |
| "loss": 0.4994, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": 0.012759355828166008, | |
| "rewards/margins": 0.004133358132094145, | |
| "rewards/rejected": 0.008625999093055725, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.987541037542187e-06, | |
| "logits/chosen": -1.8631575107574463, | |
| "logits/rejected": -1.8523809909820557, | |
| "logps/chosen": -33.509056091308594, | |
| "logps/rejected": -35.39984893798828, | |
| "loss": 0.5001, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": 0.01936880685389042, | |
| "rewards/margins": -0.001629653968848288, | |
| "rewards/rejected": 0.020998459309339523, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 4.954691471941119e-06, | |
| "logits/chosen": -1.9419567584991455, | |
| "logits/rejected": -1.9438903331756592, | |
| "logps/chosen": -32.50743103027344, | |
| "logps/rejected": -33.188419342041016, | |
| "loss": 0.491, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 0.03710051625967026, | |
| "rewards/margins": 0.04391607269644737, | |
| "rewards/rejected": -0.006815555039793253, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.901618883413549e-06, | |
| "logits/chosen": -2.0737738609313965, | |
| "logits/rejected": -2.0787465572357178, | |
| "logps/chosen": -33.928836822509766, | |
| "logps/rejected": -36.540794372558594, | |
| "loss": 0.4971, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": 0.023859737440943718, | |
| "rewards/margins": 0.01262708194553852, | |
| "rewards/rejected": 0.01123266015201807, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.828760511501322e-06, | |
| "logits/chosen": -1.934011697769165, | |
| "logits/rejected": -1.9371341466903687, | |
| "logps/chosen": -34.23366928100586, | |
| "logps/rejected": -34.542274475097656, | |
| "loss": 0.4864, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 0.07765541225671768, | |
| "rewards/margins": 0.057602256536483765, | |
| "rewards/rejected": 0.020053153857588768, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.7367166013034295e-06, | |
| "logits/chosen": -1.9421571493148804, | |
| "logits/rejected": -1.9466804265975952, | |
| "logps/chosen": -32.304595947265625, | |
| "logps/rejected": -32.284873962402344, | |
| "loss": 0.4909, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.0651295855641365, | |
| "rewards/margins": 0.035516757518053055, | |
| "rewards/rejected": 0.02961282804608345, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.626245458345211e-06, | |
| "logits/chosen": -2.0406017303466797, | |
| "logits/rejected": -2.038613796234131, | |
| "logps/chosen": -32.027652740478516, | |
| "logps/rejected": -31.224151611328125, | |
| "loss": 0.4845, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.08527366816997528, | |
| "rewards/margins": 0.0684497207403183, | |
| "rewards/rejected": 0.01682395115494728, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "eval_logits/chosen": -2.234653949737549, | |
| "eval_logits/rejected": -2.229806423187256, | |
| "eval_logps/chosen": -33.91841125488281, | |
| "eval_logps/rejected": -37.42335891723633, | |
| "eval_loss": 0.4981803297996521, | |
| "eval_rewards/accuracies": 0.5510797500610352, | |
| "eval_rewards/chosen": 0.04645563289523125, | |
| "eval_rewards/margins": 0.009152057580649853, | |
| "eval_rewards/rejected": 0.03730357065796852, | |
| "eval_runtime": 145.8684, | |
| "eval_samples_per_second": 2.351, | |
| "eval_steps_per_second": 0.295, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.498257201263691e-06, | |
| "logits/chosen": -1.9949238300323486, | |
| "logits/rejected": -1.9925572872161865, | |
| "logps/chosen": -32.99268341064453, | |
| "logps/rejected": -33.90182113647461, | |
| "loss": 0.4869, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": 0.10045422613620758, | |
| "rewards/margins": 0.049433451145887375, | |
| "rewards/rejected": 0.051020748913288116, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 4.353806263777678e-06, | |
| "logits/chosen": -2.0055007934570312, | |
| "logits/rejected": -1.997157335281372, | |
| "logps/chosen": -32.192996978759766, | |
| "logps/rejected": -32.01173400878906, | |
| "loss": 0.491, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.10076460987329483, | |
| "rewards/margins": 0.04085635766386986, | |
| "rewards/rejected": 0.05990824103355408, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 4.1940827077152755e-06, | |
| "logits/chosen": -2.034147262573242, | |
| "logits/rejected": -2.026184558868408, | |
| "logps/chosen": -30.194936752319336, | |
| "logps/rejected": -31.905689239501953, | |
| "loss": 0.4876, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.11231068521738052, | |
| "rewards/margins": 0.05338172987103462, | |
| "rewards/rejected": 0.0589289590716362, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 4.0204024186666215e-06, | |
| "logits/chosen": -1.9643388986587524, | |
| "logits/rejected": -1.9745395183563232, | |
| "logps/chosen": -31.08599853515625, | |
| "logps/rejected": -32.422943115234375, | |
| "loss": 0.4795, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 0.13346607983112335, | |
| "rewards/margins": 0.08869143575429916, | |
| "rewards/rejected": 0.044774629175662994, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 3.834196265035119e-06, | |
| "logits/chosen": -1.8776963949203491, | |
| "logits/rejected": -1.8788686990737915, | |
| "logps/chosen": -33.690345764160156, | |
| "logps/rejected": -34.572776794433594, | |
| "loss": 0.4699, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.20093801617622375, | |
| "rewards/margins": 0.12714678049087524, | |
| "rewards/rejected": 0.07379122078418732, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.636998309800573e-06, | |
| "logits/chosen": -1.9286582469940186, | |
| "logits/rejected": -1.9253017902374268, | |
| "logps/chosen": -35.773475646972656, | |
| "logps/rejected": -32.47566604614258, | |
| "loss": 0.4851, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.15928010642528534, | |
| "rewards/margins": 0.06044477969408035, | |
| "rewards/rejected": 0.09883531928062439, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 3.4304331721118078e-06, | |
| "logits/chosen": -2.0298352241516113, | |
| "logits/rejected": -2.0225348472595215, | |
| "logps/chosen": -33.22509002685547, | |
| "logps/rejected": -31.19403648376465, | |
| "loss": 0.4603, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.22910937666893005, | |
| "rewards/margins": 0.165984109044075, | |
| "rewards/rejected": 0.06312531232833862, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 3.2162026428305436e-06, | |
| "logits/chosen": -2.036612033843994, | |
| "logits/rejected": -2.041813373565674, | |
| "logps/chosen": -31.960119247436523, | |
| "logps/rejected": -32.171165466308594, | |
| "loss": 0.4734, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.23386716842651367, | |
| "rewards/margins": 0.1082783117890358, | |
| "rewards/rejected": 0.12558886408805847, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 2.996071664294641e-06, | |
| "logits/chosen": -2.037198066711426, | |
| "logits/rejected": -2.0344595909118652, | |
| "logps/chosen": -31.026615142822266, | |
| "logps/rejected": -31.082998275756836, | |
| "loss": 0.4781, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.184935063123703, | |
| "rewards/margins": 0.09084881842136383, | |
| "rewards/rejected": 0.09408621490001678, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 2.7718537898066833e-06, | |
| "logits/chosen": -1.908818006515503, | |
| "logits/rejected": -1.913496732711792, | |
| "logps/chosen": -31.075199127197266, | |
| "logps/rejected": -32.616241455078125, | |
| "loss": 0.466, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 0.2124340832233429, | |
| "rewards/margins": 0.14255891740322113, | |
| "rewards/rejected": 0.06987515836954117, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "eval_logits/chosen": -2.23366379737854, | |
| "eval_logits/rejected": -2.2288384437561035, | |
| "eval_logps/chosen": -33.779903411865234, | |
| "eval_logps/rejected": -37.29892349243164, | |
| "eval_loss": 0.4965229630470276, | |
| "eval_rewards/accuracies": 0.545265793800354, | |
| "eval_rewards/chosen": 0.10185908526182175, | |
| "eval_rewards/margins": 0.014780867844820023, | |
| "eval_rewards/rejected": 0.08707821369171143, | |
| "eval_runtime": 145.8117, | |
| "eval_samples_per_second": 2.352, | |
| "eval_steps_per_second": 0.295, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 2.5453962426402006e-06, | |
| "logits/chosen": -2.0208163261413574, | |
| "logits/rejected": -2.0314111709594727, | |
| "logps/chosen": -31.514019012451172, | |
| "logps/rejected": -33.690643310546875, | |
| "loss": 0.4728, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 0.18291443586349487, | |
| "rewards/margins": 0.114091657102108, | |
| "rewards/rejected": 0.06882277131080627, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 2.3185646976551794e-06, | |
| "logits/chosen": -1.9141031503677368, | |
| "logits/rejected": -1.9288082122802734, | |
| "logps/chosen": -29.5712947845459, | |
| "logps/rejected": -31.429983139038086, | |
| "loss": 0.4615, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.2160976380109787, | |
| "rewards/margins": 0.16482076048851013, | |
| "rewards/rejected": 0.05127686262130737, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 2.0932279108998323e-06, | |
| "logits/chosen": -1.9712814092636108, | |
| "logits/rejected": -1.975285291671753, | |
| "logps/chosen": -32.82429885864258, | |
| "logps/rejected": -31.416866302490234, | |
| "loss": 0.4561, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 0.23726816475391388, | |
| "rewards/margins": 0.18429425358772278, | |
| "rewards/rejected": 0.052973903715610504, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 1.8712423238279358e-06, | |
| "logits/chosen": -1.969780683517456, | |
| "logits/rejected": -1.9480432271957397, | |
| "logps/chosen": -33.583518981933594, | |
| "logps/rejected": -34.8461799621582, | |
| "loss": 0.4567, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.22809162735939026, | |
| "rewards/margins": 0.18624703586101532, | |
| "rewards/rejected": 0.04184458404779434, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.6544367689701824e-06, | |
| "logits/chosen": -2.0109124183654785, | |
| "logits/rejected": -2.0076212882995605, | |
| "logps/chosen": -32.46331024169922, | |
| "logps/rejected": -35.97381591796875, | |
| "loss": 0.4799, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.18470348417758942, | |
| "rewards/margins": 0.08426074683666229, | |
| "rewards/rejected": 0.10044274479150772, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 1.4445974030621963e-06, | |
| "logits/chosen": -1.878321647644043, | |
| "logits/rejected": -1.8758872747421265, | |
| "logps/chosen": -33.721397399902344, | |
| "logps/rejected": -35.270362854003906, | |
| "loss": 0.4785, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 0.1868000328540802, | |
| "rewards/margins": 0.09309660643339157, | |
| "rewards/rejected": 0.09370341151952744, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.243452991757889e-06, | |
| "logits/chosen": -1.8631956577301025, | |
| "logits/rejected": -1.8606828451156616, | |
| "logps/chosen": -33.90094757080078, | |
| "logps/rejected": -31.57466697692871, | |
| "loss": 0.4722, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.198659747838974, | |
| "rewards/margins": 0.11999478191137314, | |
| "rewards/rejected": 0.07866497337818146, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 1.0526606671603523e-06, | |
| "logits/chosen": -1.9663877487182617, | |
| "logits/rejected": -1.9559704065322876, | |
| "logps/chosen": -34.72657775878906, | |
| "logps/rejected": -31.63601303100586, | |
| "loss": 0.4585, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.25202488899230957, | |
| "rewards/margins": 0.17003390192985535, | |
| "rewards/rejected": 0.08199100196361542, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 8.737922755071455e-07, | |
| "logits/chosen": -2.062107801437378, | |
| "logits/rejected": -2.0472733974456787, | |
| "logps/chosen": -30.40212059020996, | |
| "logps/rejected": -32.340721130371094, | |
| "loss": 0.4799, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.2088027447462082, | |
| "rewards/margins": 0.08435753732919693, | |
| "rewards/rejected": 0.12444518506526947, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 7.08321427484816e-07, | |
| "logits/chosen": -1.9330482482910156, | |
| "logits/rejected": -1.9305979013442993, | |
| "logps/chosen": -32.06965255737305, | |
| "logps/rejected": -30.65035629272461, | |
| "loss": 0.4349, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 0.3429831862449646, | |
| "rewards/margins": 0.27640262246131897, | |
| "rewards/rejected": 0.06658058613538742, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "eval_logits/chosen": -2.2318336963653564, | |
| "eval_logits/rejected": -2.2270052433013916, | |
| "eval_logps/chosen": -33.749427795410156, | |
| "eval_logps/rejected": -37.27041244506836, | |
| "eval_loss": 0.4961945414543152, | |
| "eval_rewards/accuracies": 0.5357142686843872, | |
| "eval_rewards/chosen": 0.11405016481876373, | |
| "eval_rewards/margins": 0.015567691065371037, | |
| "eval_rewards/rejected": 0.09848246723413467, | |
| "eval_runtime": 145.5479, | |
| "eval_samples_per_second": 2.357, | |
| "eval_steps_per_second": 0.295, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 5.576113578589035e-07, | |
| "logits/chosen": -1.9179494380950928, | |
| "logits/rejected": -1.9148216247558594, | |
| "logps/chosen": -31.037755966186523, | |
| "logps/rejected": -33.56406021118164, | |
| "loss": 0.4639, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": 0.22623713314533234, | |
| "rewards/margins": 0.15500028431415558, | |
| "rewards/rejected": 0.07123686373233795, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 4.229036944380913e-07, | |
| "logits/chosen": -1.9682966470718384, | |
| "logits/rejected": -1.9561439752578735, | |
| "logps/chosen": -34.023921966552734, | |
| "logps/rejected": -33.415985107421875, | |
| "loss": 0.4575, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.22886168956756592, | |
| "rewards/margins": 0.17971986532211304, | |
| "rewards/rejected": 0.04914180561900139, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 3.053082288996112e-07, | |
| "logits/chosen": -2.003788709640503, | |
| "logits/rejected": -2.002437114715576, | |
| "logps/chosen": -32.8862419128418, | |
| "logps/rejected": -32.21629333496094, | |
| "loss": 0.4679, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.23982055485248566, | |
| "rewards/margins": 0.13450448215007782, | |
| "rewards/rejected": 0.10531606525182724, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 2.0579377374915805e-07, | |
| "logits/chosen": -2.0905163288116455, | |
| "logits/rejected": -2.0749027729034424, | |
| "logps/chosen": -33.47309112548828, | |
| "logps/rejected": -32.81334686279297, | |
| "loss": 0.4618, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.2825208008289337, | |
| "rewards/margins": 0.16064420342445374, | |
| "rewards/rejected": 0.12187659740447998, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 1.2518018074041684e-07, | |
| "logits/chosen": -1.9627025127410889, | |
| "logits/rejected": -1.9618685245513916, | |
| "logps/chosen": -32.528175354003906, | |
| "logps/rejected": -32.22235107421875, | |
| "loss": 0.4522, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.30156710743904114, | |
| "rewards/margins": 0.20444798469543457, | |
| "rewards/rejected": 0.09711913019418716, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 6.41315865106129e-08, | |
| "logits/chosen": -1.918320655822754, | |
| "logits/rejected": -1.9285932779312134, | |
| "logps/chosen": -31.606945037841797, | |
| "logps/rejected": -34.98893356323242, | |
| "loss": 0.4718, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.2390974760055542, | |
| "rewards/margins": 0.11916828155517578, | |
| "rewards/rejected": 0.11992917954921722, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2.3150941078050325e-08, | |
| "logits/chosen": -2.0581445693969727, | |
| "logits/rejected": -2.051628828048706, | |
| "logps/chosen": -33.017967224121094, | |
| "logps/rejected": -28.989696502685547, | |
| "loss": 0.4612, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": 0.253601610660553, | |
| "rewards/margins": 0.15891048312187195, | |
| "rewards/rejected": 0.09469114243984222, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 2.575864278703266e-09, | |
| "logits/chosen": -1.9176830053329468, | |
| "logits/rejected": -1.919847846031189, | |
| "logps/chosen": -33.614540100097656, | |
| "logps/rejected": -30.760555267333984, | |
| "loss": 0.4519, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": 0.24980488419532776, | |
| "rewards/margins": 0.2049761563539505, | |
| "rewards/rejected": 0.04482869431376457, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 385, | |
| "total_flos": 0.0, | |
| "train_loss": 0.4751699732495593, | |
| "train_runtime": 3253.0203, | |
| "train_samples_per_second": 0.947, | |
| "train_steps_per_second": 0.118 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 385, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "total_flos": 0.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |