| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.710508474576271, | |
| "eval_steps": 100, | |
| "global_step": 2000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.013559322033898305, | |
| "grad_norm": 154.68963322032346, | |
| "learning_rate": 4.504504504504504e-08, | |
| "logits/chosen": -1.6006476879119873, | |
| "logits/rejected": -1.7503880262374878, | |
| "logps/chosen": -136.20535278320312, | |
| "logps/rejected": -650.806396484375, | |
| "loss": 0.6902, | |
| "rewards/accuracies": 0.40000003576278687, | |
| "rewards/chosen": -0.002106652595102787, | |
| "rewards/margins": 0.007925467565655708, | |
| "rewards/rejected": -0.01003211922943592, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02711864406779661, | |
| "grad_norm": 143.60255538961036, | |
| "learning_rate": 9.009009009009008e-08, | |
| "logits/chosen": -1.4859544038772583, | |
| "logits/rejected": -1.6451891660690308, | |
| "logps/chosen": -125.66300964355469, | |
| "logps/rejected": -636.0386962890625, | |
| "loss": 0.688, | |
| "rewards/accuracies": 0.5374999642372131, | |
| "rewards/chosen": 0.003618550719693303, | |
| "rewards/margins": 0.012560315430164337, | |
| "rewards/rejected": -0.008941764943301678, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.04067796610169491, | |
| "grad_norm": 146.5406135926217, | |
| "learning_rate": 1.3513513513513515e-07, | |
| "logits/chosen": -1.6544740200042725, | |
| "logits/rejected": -1.831311583518982, | |
| "logps/chosen": -134.34368896484375, | |
| "logps/rejected": -654.5028686523438, | |
| "loss": 0.672, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.009306993335485458, | |
| "rewards/margins": 0.04563351720571518, | |
| "rewards/rejected": -0.03632652387022972, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.05423728813559322, | |
| "grad_norm": 134.93316662114188, | |
| "learning_rate": 1.8018018018018017e-07, | |
| "logits/chosen": -1.6334645748138428, | |
| "logits/rejected": -1.8082122802734375, | |
| "logps/chosen": -150.40078735351562, | |
| "logps/rejected": -714.775390625, | |
| "loss": 0.5965, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": 0.00473959231749177, | |
| "rewards/margins": 0.20695891976356506, | |
| "rewards/rejected": -0.20221932232379913, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.06779661016949153, | |
| "grad_norm": 85.1543469130844, | |
| "learning_rate": 2.2522522522522522e-07, | |
| "logits/chosen": -1.636574387550354, | |
| "logits/rejected": -1.8049229383468628, | |
| "logps/chosen": -142.1752471923828, | |
| "logps/rejected": -667.3391723632812, | |
| "loss": 0.4427, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": 0.007638626731932163, | |
| "rewards/margins": 0.6071017980575562, | |
| "rewards/rejected": -0.5994631052017212, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.08135593220338982, | |
| "grad_norm": 55.45578483376317, | |
| "learning_rate": 2.702702702702703e-07, | |
| "logits/chosen": -1.7117058038711548, | |
| "logits/rejected": -1.887060523033142, | |
| "logps/chosen": -117.88133239746094, | |
| "logps/rejected": -649.540283203125, | |
| "loss": 0.2845, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.0008934163488447666, | |
| "rewards/margins": 1.142379879951477, | |
| "rewards/rejected": -1.1414865255355835, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.09491525423728814, | |
| "grad_norm": 9.23309067720513, | |
| "learning_rate": 3.153153153153153e-07, | |
| "logits/chosen": -1.4379091262817383, | |
| "logits/rejected": -1.6131579875946045, | |
| "logps/chosen": -134.87896728515625, | |
| "logps/rejected": -696.23779296875, | |
| "loss": 0.0565, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.001966355135664344, | |
| "rewards/margins": 3.1120121479034424, | |
| "rewards/rejected": -3.113978624343872, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.10847457627118644, | |
| "grad_norm": 2.7786137414686887, | |
| "learning_rate": 3.6036036036036033e-07, | |
| "logits/chosen": -1.531922698020935, | |
| "logits/rejected": -1.6740307807922363, | |
| "logps/chosen": -130.48211669921875, | |
| "logps/rejected": -688.8399047851562, | |
| "loss": 0.0138, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.001919470028951764, | |
| "rewards/margins": 4.491945266723633, | |
| "rewards/rejected": -4.490025997161865, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.12203389830508475, | |
| "grad_norm": 0.011215121612896709, | |
| "learning_rate": 4.054054054054054e-07, | |
| "logits/chosen": -1.5453845262527466, | |
| "logits/rejected": -1.702711820602417, | |
| "logps/chosen": -137.26138305664062, | |
| "logps/rejected": -749.4444580078125, | |
| "loss": 0.001, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.00951399840414524, | |
| "rewards/margins": 8.38076400756836, | |
| "rewards/rejected": -8.37125015258789, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.13559322033898305, | |
| "grad_norm": 0.002748606315174295, | |
| "learning_rate": 4.5045045045045043e-07, | |
| "logits/chosen": -1.6784169673919678, | |
| "logits/rejected": -1.8867419958114624, | |
| "logps/chosen": -134.819091796875, | |
| "logps/rejected": -786.6911010742188, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.0101728904992342, | |
| "rewards/margins": 12.241233825683594, | |
| "rewards/rejected": -12.231060981750488, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.13559322033898305, | |
| "eval_logits/chosen": -1.5288071632385254, | |
| "eval_logits/rejected": -1.6835100650787354, | |
| "eval_logps/chosen": -129.43997192382812, | |
| "eval_logps/rejected": -748.4630126953125, | |
| "eval_loss": 3.7055913253425388e-06, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": 0.012690085917711258, | |
| "eval_rewards/margins": 13.008617401123047, | |
| "eval_rewards/rejected": -12.995927810668945, | |
| "eval_runtime": 23.2529, | |
| "eval_samples_per_second": 4.301, | |
| "eval_steps_per_second": 1.075, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.14915254237288136, | |
| "grad_norm": 0.0011093150463870978, | |
| "learning_rate": 4.954954954954955e-07, | |
| "logits/chosen": -1.57046377658844, | |
| "logits/rejected": -1.7331207990646362, | |
| "logps/chosen": -127.67959594726562, | |
| "logps/rejected": -752.2487182617188, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.004207385703921318, | |
| "rewards/margins": 13.458518981933594, | |
| "rewards/rejected": -13.462725639343262, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.16271186440677965, | |
| "grad_norm": 0.0005762634187664357, | |
| "learning_rate": 4.999773405362863e-07, | |
| "logits/chosen": -1.6726722717285156, | |
| "logits/rejected": -1.8734807968139648, | |
| "logps/chosen": -143.523193359375, | |
| "logps/rejected": -832.698974609375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.014902787283062935, | |
| "rewards/margins": 13.953229904174805, | |
| "rewards/rejected": -13.938325881958008, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.17627118644067796, | |
| "grad_norm": 0.0003332599393998685, | |
| "learning_rate": 4.998990167994546e-07, | |
| "logits/chosen": -1.6244195699691772, | |
| "logits/rejected": -1.790928602218628, | |
| "logps/chosen": -137.99049377441406, | |
| "logps/rejected": -785.167236328125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.014037198387086391, | |
| "rewards/margins": 13.896438598632812, | |
| "rewards/rejected": -13.882402420043945, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.18983050847457628, | |
| "grad_norm": 0.0003146655208168566, | |
| "learning_rate": 4.997647665674343e-07, | |
| "logits/chosen": -1.6478142738342285, | |
| "logits/rejected": -1.860294222831726, | |
| "logps/chosen": -147.6448516845703, | |
| "logps/rejected": -811.6314086914062, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.015513481572270393, | |
| "rewards/margins": 14.067631721496582, | |
| "rewards/rejected": -14.052118301391602, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.2033898305084746, | |
| "grad_norm": 0.0002489502956030974, | |
| "learning_rate": 4.995746198849412e-07, | |
| "logits/chosen": -1.5896422863006592, | |
| "logits/rejected": -1.7557127475738525, | |
| "logps/chosen": -140.1243896484375, | |
| "logps/rejected": -822.0531616210938, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.02557023987174034, | |
| "rewards/margins": 14.115985870361328, | |
| "rewards/rejected": -14.090415954589844, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.21694915254237288, | |
| "grad_norm": 0.00037978880201729675, | |
| "learning_rate": 4.993286193061145e-07, | |
| "logits/chosen": -1.6275484561920166, | |
| "logits/rejected": -1.8082895278930664, | |
| "logps/chosen": -145.8615264892578, | |
| "logps/rejected": -797.1973876953125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.005417819134891033, | |
| "rewards/margins": 14.054203033447266, | |
| "rewards/rejected": -14.048785209655762, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.2305084745762712, | |
| "grad_norm": 0.0011709678087963664, | |
| "learning_rate": 4.99026819884993e-07, | |
| "logits/chosen": -1.691068172454834, | |
| "logits/rejected": -1.8546725511550903, | |
| "logps/chosen": -140.33941650390625, | |
| "logps/rejected": -833.8118896484375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.005253595765680075, | |
| "rewards/margins": 14.147695541381836, | |
| "rewards/rejected": -14.142441749572754, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.2440677966101695, | |
| "grad_norm": 0.0005430831103481389, | |
| "learning_rate": 4.986692891631945e-07, | |
| "logits/chosen": -1.654329776763916, | |
| "logits/rejected": -1.8268946409225464, | |
| "logps/chosen": -142.9085235595703, | |
| "logps/rejected": -781.0303955078125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.004031305201351643, | |
| "rewards/margins": 13.904911041259766, | |
| "rewards/rejected": -13.900879859924316, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.2576271186440678, | |
| "grad_norm": 0.0002685256860972653, | |
| "learning_rate": 4.982561071548001e-07, | |
| "logits/chosen": -1.537014365196228, | |
| "logits/rejected": -1.6596903800964355, | |
| "logps/chosen": -123.82093811035156, | |
| "logps/rejected": -741.5814208984375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.018150925636291504, | |
| "rewards/margins": 14.061857223510742, | |
| "rewards/rejected": -14.043705940246582, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.2711864406779661, | |
| "grad_norm": 0.00047407562273882125, | |
| "learning_rate": 4.977873663284474e-07, | |
| "logits/chosen": -1.6226494312286377, | |
| "logits/rejected": -1.8021833896636963, | |
| "logps/chosen": -152.94691467285156, | |
| "logps/rejected": -808.9371337890625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.021439315751194954, | |
| "rewards/margins": 14.185613632202148, | |
| "rewards/rejected": -14.164173126220703, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.2711864406779661, | |
| "eval_logits/chosen": -1.5458643436431885, | |
| "eval_logits/rejected": -1.6861909627914429, | |
| "eval_logps/chosen": -129.42483520507812, | |
| "eval_logps/rejected": -757.507080078125, | |
| "eval_loss": 1.5668128980905749e-06, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": 0.014203321188688278, | |
| "eval_rewards/margins": 13.914541244506836, | |
| "eval_rewards/rejected": -13.900337219238281, | |
| "eval_runtime": 23.2677, | |
| "eval_samples_per_second": 4.298, | |
| "eval_steps_per_second": 1.074, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.2847457627118644, | |
| "grad_norm": 0.0005141024917889797, | |
| "learning_rate": 4.972631715866361e-07, | |
| "logits/chosen": -1.5850636959075928, | |
| "logits/rejected": -1.7246123552322388, | |
| "logps/chosen": -143.24696350097656, | |
| "logps/rejected": -778.4476318359375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.008326123468577862, | |
| "rewards/margins": 13.95406436920166, | |
| "rewards/rejected": -13.945737838745117, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.2983050847457627, | |
| "grad_norm": 0.0006130760590228257, | |
| "learning_rate": 4.966836402422515e-07, | |
| "logits/chosen": -1.6016970872879028, | |
| "logits/rejected": -1.7708076238632202, | |
| "logps/chosen": -132.4145050048828, | |
| "logps/rejected": -792.4443359375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.017194140702486038, | |
| "rewards/margins": 14.093986511230469, | |
| "rewards/rejected": -14.07679271697998, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.31186440677966104, | |
| "grad_norm": 0.000828819568409969, | |
| "learning_rate": 4.960489019923105e-07, | |
| "logits/chosen": -1.679842233657837, | |
| "logits/rejected": -1.8294236660003662, | |
| "logps/chosen": -136.24644470214844, | |
| "logps/rejected": -809.9606323242188, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.016033075749874115, | |
| "rewards/margins": 14.131494522094727, | |
| "rewards/rejected": -14.115462303161621, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.3254237288135593, | |
| "grad_norm": 0.0008292339546344823, | |
| "learning_rate": 4.95359098888935e-07, | |
| "logits/chosen": -1.6466869115829468, | |
| "logits/rejected": -1.8071887493133545, | |
| "logps/chosen": -144.3760986328125, | |
| "logps/rejected": -816.1300659179688, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.00023554242216050625, | |
| "rewards/margins": 14.113880157470703, | |
| "rewards/rejected": -14.113645553588867, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.3389830508474576, | |
| "grad_norm": 0.0007326232553653241, | |
| "learning_rate": 4.946143853075625e-07, | |
| "logits/chosen": -1.6482115983963013, | |
| "logits/rejected": -1.8068441152572632, | |
| "logps/chosen": -136.09292602539062, | |
| "logps/rejected": -786.30908203125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.0055378577671945095, | |
| "rewards/margins": 14.151810646057129, | |
| "rewards/rejected": -14.146272659301758, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.3525423728813559, | |
| "grad_norm": 0.0006201108834653548, | |
| "learning_rate": 4.938149279123959e-07, | |
| "logits/chosen": -1.614100694656372, | |
| "logits/rejected": -1.7720319032669067, | |
| "logps/chosen": -145.45883178710938, | |
| "logps/rejected": -772.9180908203125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.014799129217863083, | |
| "rewards/margins": 13.972376823425293, | |
| "rewards/rejected": -13.987175941467285, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.36610169491525424, | |
| "grad_norm": 0.0002824417146726921, | |
| "learning_rate": 4.929609056191057e-07, | |
| "logits/chosen": -1.6527501344680786, | |
| "logits/rejected": -1.8346607685089111, | |
| "logps/chosen": -147.9451904296875, | |
| "logps/rejected": -845.679931640625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 8.073175558820367e-05, | |
| "rewards/margins": 14.214624404907227, | |
| "rewards/rejected": -14.214543342590332, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.37966101694915255, | |
| "grad_norm": 0.0003459147020709031, | |
| "learning_rate": 4.920525095547895e-07, | |
| "logits/chosen": -1.700380802154541, | |
| "logits/rejected": -1.8940513134002686, | |
| "logps/chosen": -148.5547332763672, | |
| "logps/rejected": -806.921142578125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.0022345068864524364, | |
| "rewards/margins": 13.974288940429688, | |
| "rewards/rejected": -13.976522445678711, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.39322033898305087, | |
| "grad_norm": 0.0007686863218550116, | |
| "learning_rate": 4.910899430151973e-07, | |
| "logits/chosen": -1.690828800201416, | |
| "logits/rejected": -1.8689790964126587, | |
| "logps/chosen": -127.06858825683594, | |
| "logps/rejected": -781.9804077148438, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.014154733158648014, | |
| "rewards/margins": 13.95429801940918, | |
| "rewards/rejected": -13.940142631530762, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.4067796610169492, | |
| "grad_norm": 0.0007833273852497298, | |
| "learning_rate": 4.900734214192358e-07, | |
| "logits/chosen": -1.657442331314087, | |
| "logits/rejected": -1.8555173873901367, | |
| "logps/chosen": -142.8946075439453, | |
| "logps/rejected": -833.9373168945312, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.016108307987451553, | |
| "rewards/margins": 14.204878807067871, | |
| "rewards/rejected": -14.188769340515137, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.4067796610169492, | |
| "eval_logits/chosen": -1.5289416313171387, | |
| "eval_logits/rejected": -1.6836309432983398, | |
| "eval_logps/chosen": -129.3828582763672, | |
| "eval_logps/rejected": -758.2346801757812, | |
| "eval_loss": 1.4355653092934517e-06, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": 0.018402252346277237, | |
| "eval_rewards/margins": 13.991499900817871, | |
| "eval_rewards/rejected": -13.97309684753418, | |
| "eval_runtime": 23.0859, | |
| "eval_samples_per_second": 4.332, | |
| "eval_steps_per_second": 1.083, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.42033898305084744, | |
| "grad_norm": 0.00021728835196828888, | |
| "learning_rate": 4.890031722607576e-07, | |
| "logits/chosen": -1.5837361812591553, | |
| "logits/rejected": -1.7333931922912598, | |
| "logps/chosen": -133.32366943359375, | |
| "logps/rejected": -758.0372314453125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.0008433577604591846, | |
| "rewards/margins": 14.209864616394043, | |
| "rewards/rejected": -14.209020614624023, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.43389830508474575, | |
| "grad_norm": 0.00042770904659586716, | |
| "learning_rate": 4.878794350576498e-07, | |
| "logits/chosen": -1.577714443206787, | |
| "logits/rejected": -1.759061336517334, | |
| "logps/chosen": -128.18081665039062, | |
| "logps/rejected": -749.0996704101562, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.006202464923262596, | |
| "rewards/margins": 14.050898551940918, | |
| "rewards/rejected": -14.044694900512695, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.44745762711864406, | |
| "grad_norm": 0.0005431267045930932, | |
| "learning_rate": 4.867024612982295e-07, | |
| "logits/chosen": -1.5479421615600586, | |
| "logits/rejected": -1.6868603229522705, | |
| "logps/chosen": -145.11839294433594, | |
| "logps/rejected": -793.3886108398438, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.007111231796443462, | |
| "rewards/margins": 14.120939254760742, | |
| "rewards/rejected": -14.1138277053833, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.4610169491525424, | |
| "grad_norm": 0.0004452926543919869, | |
| "learning_rate": 4.854725143849631e-07, | |
| "logits/chosen": -1.5681132078170776, | |
| "logits/rejected": -1.7265939712524414, | |
| "logps/chosen": -135.96771240234375, | |
| "logps/rejected": -776.887939453125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.01606500893831253, | |
| "rewards/margins": 13.997052192687988, | |
| "rewards/rejected": -13.980987548828125, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.4745762711864407, | |
| "grad_norm": 0.0003403651283500103, | |
| "learning_rate": 4.841898695755167e-07, | |
| "logits/chosen": -1.6001578569412231, | |
| "logits/rejected": -1.7874952554702759, | |
| "logps/chosen": -128.35435485839844, | |
| "logps/rejected": -756.4078979492188, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.018694868311285973, | |
| "rewards/margins": 14.069332122802734, | |
| "rewards/rejected": -14.050636291503906, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.488135593220339, | |
| "grad_norm": 0.0005571850400383578, | |
| "learning_rate": 4.828548139211545e-07, | |
| "logits/chosen": -1.6838054656982422, | |
| "logits/rejected": -1.8665508031845093, | |
| "logps/chosen": -135.80624389648438, | |
| "logps/rejected": -794.292724609375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.02806948870420456, | |
| "rewards/margins": 14.219563484191895, | |
| "rewards/rejected": -14.19149398803711, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.5016949152542373, | |
| "grad_norm": 0.0007950002555106897, | |
| "learning_rate": 4.814676462024987e-07, | |
| "logits/chosen": -1.657123327255249, | |
| "logits/rejected": -1.8259265422821045, | |
| "logps/chosen": -151.13162231445312, | |
| "logps/rejected": -806.3003540039062, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.016831912100315094, | |
| "rewards/margins": 14.099932670593262, | |
| "rewards/rejected": -14.08310317993164, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.5152542372881356, | |
| "grad_norm": 0.0003242442253345165, | |
| "learning_rate": 4.800286768626621e-07, | |
| "logits/chosen": -1.657513976097107, | |
| "logits/rejected": -1.8597712516784668, | |
| "logps/chosen": -139.53028869628906, | |
| "logps/rejected": -826.2210083007812, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.014754395000636578, | |
| "rewards/margins": 14.141820907592773, | |
| "rewards/rejected": -14.127065658569336, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.5288135593220339, | |
| "grad_norm": 0.0005090620291333299, | |
| "learning_rate": 4.785382279377733e-07, | |
| "logits/chosen": -1.6210606098175049, | |
| "logits/rejected": -1.8241839408874512, | |
| "logps/chosen": -143.5242462158203, | |
| "logps/rejected": -836.617431640625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.02124781347811222, | |
| "rewards/margins": 14.220711708068848, | |
| "rewards/rejected": -14.199464797973633, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.5423728813559322, | |
| "grad_norm": 0.000825922097218788, | |
| "learning_rate": 4.769966329849054e-07, | |
| "logits/chosen": -1.6811935901641846, | |
| "logits/rejected": -1.8629109859466553, | |
| "logps/chosen": -150.13429260253906, | |
| "logps/rejected": -827.9759521484375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.0035764453932642937, | |
| "rewards/margins": 14.165657997131348, | |
| "rewards/rejected": -14.162080764770508, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5423728813559322, | |
| "eval_logits/chosen": -1.5275362730026245, | |
| "eval_logits/rejected": -1.6860315799713135, | |
| "eval_logps/chosen": -129.4043426513672, | |
| "eval_logps/rejected": -758.570556640625, | |
| "eval_loss": 1.3993409311297e-06, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": 0.016253653913736343, | |
| "eval_rewards/margins": 14.022944450378418, | |
| "eval_rewards/rejected": -14.006690979003906, | |
| "eval_runtime": 23.1703, | |
| "eval_samples_per_second": 4.316, | |
| "eval_steps_per_second": 1.079, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5559322033898305, | |
| "grad_norm": 0.0007397757326606043, | |
| "learning_rate": 4.7540423700742726e-07, | |
| "logits/chosen": -1.6916306018829346, | |
| "logits/rejected": -1.8490281105041504, | |
| "logps/chosen": -149.16802978515625, | |
| "logps/rejected": -804.88525390625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.008371539413928986, | |
| "rewards/margins": 14.190282821655273, | |
| "rewards/rejected": -14.18191146850586, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.5694915254237288, | |
| "grad_norm": 0.0005353994450256941, | |
| "learning_rate": 4.7376139637779354e-07, | |
| "logits/chosen": -1.6099956035614014, | |
| "logits/rejected": -1.7850123643875122, | |
| "logps/chosen": -122.01039123535156, | |
| "logps/rejected": -747.48974609375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.017607325688004494, | |
| "rewards/margins": 14.0084867477417, | |
| "rewards/rejected": -13.99087905883789, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.5830508474576271, | |
| "grad_norm": 0.00048794620873617576, | |
| "learning_rate": 4.7206847875778913e-07, | |
| "logits/chosen": -1.6391756534576416, | |
| "logits/rejected": -1.8201954364776611, | |
| "logps/chosen": -131.18875122070312, | |
| "logps/rejected": -759.473388671875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.01886049658060074, | |
| "rewards/margins": 14.103827476501465, | |
| "rewards/rejected": -14.084967613220215, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.5966101694915255, | |
| "grad_norm": 0.00044451168413258085, | |
| "learning_rate": 4.70325863016248e-07, | |
| "logits/chosen": -1.678541660308838, | |
| "logits/rejected": -1.8704532384872437, | |
| "logps/chosen": -144.02674865722656, | |
| "logps/rejected": -808.55859375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.015889765694737434, | |
| "rewards/margins": 14.10900592803955, | |
| "rewards/rejected": -14.093117713928223, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.6101694915254238, | |
| "grad_norm": 0.00038795072502636444, | |
| "learning_rate": 4.68533939144264e-07, | |
| "logits/chosen": -1.6224497556686401, | |
| "logits/rejected": -1.7962324619293213, | |
| "logps/chosen": -140.4901123046875, | |
| "logps/rejected": -825.280517578125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.005872346460819244, | |
| "rewards/margins": 14.365765571594238, | |
| "rewards/rejected": -14.359892845153809, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.6237288135593221, | |
| "grad_norm": 0.00038646259276520423, | |
| "learning_rate": 4.6669310816791184e-07, | |
| "logits/chosen": -1.7635605335235596, | |
| "logits/rejected": -1.951772928237915, | |
| "logps/chosen": -136.94752502441406, | |
| "logps/rejected": -807.515869140625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.0075184679590165615, | |
| "rewards/margins": 14.222395896911621, | |
| "rewards/rejected": -14.214877128601074, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.6372881355932203, | |
| "grad_norm": 0.0003666927496908433, | |
| "learning_rate": 4.6480378205849926e-07, | |
| "logits/chosen": -1.6419250965118408, | |
| "logits/rejected": -1.7922152280807495, | |
| "logps/chosen": -130.02239990234375, | |
| "logps/rejected": -759.2348022460938, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.014397606253623962, | |
| "rewards/margins": 14.155649185180664, | |
| "rewards/rejected": -14.141252517700195, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.6508474576271186, | |
| "grad_norm": 0.00046994952764087256, | |
| "learning_rate": 4.6286638364036905e-07, | |
| "logits/chosen": -1.6200618743896484, | |
| "logits/rejected": -1.797215461730957, | |
| "logps/chosen": -130.50233459472656, | |
| "logps/rejected": -829.3895263671875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.02378329634666443, | |
| "rewards/margins": 14.199002265930176, | |
| "rewards/rejected": -14.17521858215332, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.6644067796610169, | |
| "grad_norm": 0.00029735714895166485, | |
| "learning_rate": 4.6088134649627284e-07, | |
| "logits/chosen": -1.5747830867767334, | |
| "logits/rejected": -1.753875494003296, | |
| "logps/chosen": -138.4869842529297, | |
| "logps/rejected": -819.8696899414062, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.010798036120831966, | |
| "rewards/margins": 14.393420219421387, | |
| "rewards/rejected": -14.382623672485352, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.6779661016949152, | |
| "grad_norm": 0.00035937901840089274, | |
| "learning_rate": 4.5884911487033665e-07, | |
| "logits/chosen": -1.6192841529846191, | |
| "logits/rejected": -1.7925256490707397, | |
| "logps/chosen": -143.53257751464844, | |
| "logps/rejected": -787.250244140625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.016855549067258835, | |
| "rewards/margins": 14.227276802062988, | |
| "rewards/rejected": -14.210421562194824, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.6779661016949152, | |
| "eval_logits/chosen": -1.5297393798828125, | |
| "eval_logits/rejected": -1.6854040622711182, | |
| "eval_logps/chosen": -129.43572998046875, | |
| "eval_logps/rejected": -759.3566284179688, | |
| "eval_loss": 1.3430080798570998e-06, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": 0.01311517134308815, | |
| "eval_rewards/margins": 14.098403930664062, | |
| "eval_rewards/rejected": -14.085289001464844, | |
| "eval_runtime": 23.0065, | |
| "eval_samples_per_second": 4.347, | |
| "eval_steps_per_second": 1.087, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.6915254237288135, | |
| "grad_norm": 0.00020815218989513314, | |
| "learning_rate": 4.567701435686404e-07, | |
| "logits/chosen": -1.6033257246017456, | |
| "logits/rejected": -1.7678654193878174, | |
| "logps/chosen": -143.47254943847656, | |
| "logps/rejected": -833.3479614257812, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.018440861254930496, | |
| "rewards/margins": 14.311261177062988, | |
| "rewards/rejected": -14.292821884155273, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.7050847457627119, | |
| "grad_norm": 0.00044702973171401343, | |
| "learning_rate": 4.5464489785743454e-07, | |
| "logits/chosen": -1.588118076324463, | |
| "logits/rejected": -1.7491432428359985, | |
| "logps/chosen": -121.74967193603516, | |
| "logps/rejected": -763.8538818359375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.021985743194818497, | |
| "rewards/margins": 14.20384693145752, | |
| "rewards/rejected": -14.181861877441406, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.7186440677966102, | |
| "grad_norm": 0.00048331111256865464, | |
| "learning_rate": 4.5247385335901457e-07, | |
| "logits/chosen": -1.62116277217865, | |
| "logits/rejected": -1.798133134841919, | |
| "logps/chosen": -144.86624145507812, | |
| "logps/rejected": -807.5030517578125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.006796732544898987, | |
| "rewards/margins": 14.287379264831543, | |
| "rewards/rejected": -14.280582427978516, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.7322033898305085, | |
| "grad_norm": 0.0004276849725743657, | |
| "learning_rate": 4.5025749594527895e-07, | |
| "logits/chosen": -1.5937074422836304, | |
| "logits/rejected": -1.766377329826355, | |
| "logps/chosen": -130.42138671875, | |
| "logps/rejected": -770.9468994140625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.011606786400079727, | |
| "rewards/margins": 14.063494682312012, | |
| "rewards/rejected": -14.051887512207031, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.7457627118644068, | |
| "grad_norm": 0.00047701124764661757, | |
| "learning_rate": 4.4799632162899236e-07, | |
| "logits/chosen": -1.586263656616211, | |
| "logits/rejected": -1.7527462244033813, | |
| "logps/chosen": -138.99436950683594, | |
| "logps/rejected": -832.9317626953125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.01648543030023575, | |
| "rewards/margins": 14.375706672668457, | |
| "rewards/rejected": -14.359220504760742, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.7593220338983051, | |
| "grad_norm": 0.0005331903399639826, | |
| "learning_rate": 4.456908364527802e-07, | |
| "logits/chosen": -1.6967836618423462, | |
| "logits/rejected": -1.8608123064041138, | |
| "logps/chosen": -135.91598510742188, | |
| "logps/rejected": -800.591796875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.020324843004345894, | |
| "rewards/margins": 14.120574951171875, | |
| "rewards/rejected": -14.100250244140625, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.7728813559322034, | |
| "grad_norm": 0.000533186088613475, | |
| "learning_rate": 4.433415563758778e-07, | |
| "logits/chosen": -1.6483113765716553, | |
| "logits/rejected": -1.8176215887069702, | |
| "logps/chosen": -138.11244201660156, | |
| "logps/rejected": -791.1510620117188, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.01274613942950964, | |
| "rewards/margins": 14.025276184082031, | |
| "rewards/rejected": -14.012529373168945, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.7864406779661017, | |
| "grad_norm": 0.0003330593802523546, | |
| "learning_rate": 4.409490071586606e-07, | |
| "logits/chosen": -1.58778715133667, | |
| "logits/rejected": -1.7679705619812012, | |
| "logps/chosen": -143.45578002929688, | |
| "logps/rejected": -871.753173828125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.004776511341333389, | |
| "rewards/margins": 14.490696907043457, | |
| "rewards/rejected": -14.485920906066895, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.0004356549740596174, | |
| "learning_rate": 4.38513724244981e-07, | |
| "logits/chosen": -1.677300214767456, | |
| "logits/rejected": -1.823121428489685, | |
| "logps/chosen": -130.93118286132812, | |
| "logps/rejected": -767.2311401367188, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.01857886277139187, | |
| "rewards/margins": 14.194330215454102, | |
| "rewards/rejected": -14.175750732421875, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.8135593220338984, | |
| "grad_norm": 0.0006738152861953452, | |
| "learning_rate": 4.360362526423382e-07, | |
| "logits/chosen": -1.6017038822174072, | |
| "logits/rejected": -1.7450367212295532, | |
| "logps/chosen": -134.8054656982422, | |
| "logps/rejected": -829.2752075195312, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.023766428232192993, | |
| "rewards/margins": 14.333120346069336, | |
| "rewards/rejected": -14.309354782104492, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.8135593220338984, | |
| "eval_logits/chosen": -1.535141944885254, | |
| "eval_logits/rejected": -1.6854389905929565, | |
| "eval_logps/chosen": -129.3685302734375, | |
| "eval_logps/rejected": -759.9224853515625, | |
| "eval_loss": 1.2519085430540144e-06, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": 0.019834483042359352, | |
| "eval_rewards/margins": 14.161703109741211, | |
| "eval_rewards/rejected": -14.14186954498291, | |
| "eval_runtime": 23.2942, | |
| "eval_samples_per_second": 4.293, | |
| "eval_steps_per_second": 1.073, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.8271186440677966, | |
| "grad_norm": 0.00028376978865923564, | |
| "learning_rate": 4.3351714679990706e-07, | |
| "logits/chosen": -1.5073317289352417, | |
| "logits/rejected": -1.630771517753601, | |
| "logps/chosen": -108.68913269042969, | |
| "logps/rejected": -755.2454223632812, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.021552443504333496, | |
| "rewards/margins": 14.418371200561523, | |
| "rewards/rejected": -14.396819114685059, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.8406779661016949, | |
| "grad_norm": 0.00048164470828510026, | |
| "learning_rate": 4.3095697048445447e-07, | |
| "logits/chosen": -1.6341415643692017, | |
| "logits/rejected": -1.7738165855407715, | |
| "logps/chosen": -140.3278045654297, | |
| "logps/rejected": -794.1578369140625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.0028394628316164017, | |
| "rewards/margins": 14.288480758666992, | |
| "rewards/rejected": -14.28564167022705, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.8542372881355932, | |
| "grad_norm": 0.00046143931913359906, | |
| "learning_rate": 4.283562966541707e-07, | |
| "logits/chosen": -1.5982298851013184, | |
| "logits/rejected": -1.7638392448425293, | |
| "logps/chosen": -141.6815185546875, | |
| "logps/rejected": -788.85888671875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.018078766763210297, | |
| "rewards/margins": 14.164934158325195, | |
| "rewards/rejected": -14.146854400634766, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.8677966101694915, | |
| "grad_norm": 0.00032792057687413, | |
| "learning_rate": 4.25715707330443e-07, | |
| "logits/chosen": -1.5992499589920044, | |
| "logits/rejected": -1.7798690795898438, | |
| "logps/chosen": -128.59031677246094, | |
| "logps/rejected": -792.5185546875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.009904641658067703, | |
| "rewards/margins": 14.5419340133667, | |
| "rewards/rejected": -14.532029151916504, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.8813559322033898, | |
| "grad_norm": 0.00021017148877000384, | |
| "learning_rate": 4.2303579346760173e-07, | |
| "logits/chosen": -1.613771915435791, | |
| "logits/rejected": -1.801413655281067, | |
| "logps/chosen": -143.33572387695312, | |
| "logps/rejected": -811.8010864257812, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.015471378341317177, | |
| "rewards/margins": 14.446039199829102, | |
| "rewards/rejected": -14.430566787719727, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.8949152542372881, | |
| "grad_norm": 0.0003556697469879977, | |
| "learning_rate": 4.2031715482066655e-07, | |
| "logits/chosen": -1.585208535194397, | |
| "logits/rejected": -1.7628443241119385, | |
| "logps/chosen": -133.2225799560547, | |
| "logps/rejected": -783.7073974609375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.01283181644976139, | |
| "rewards/margins": 14.336665153503418, | |
| "rewards/rejected": -14.323833465576172, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.9084745762711864, | |
| "grad_norm": 0.00032469878909931687, | |
| "learning_rate": 4.1756039981112373e-07, | |
| "logits/chosen": -1.5636094808578491, | |
| "logits/rejected": -1.7128851413726807, | |
| "logps/chosen": -138.43899536132812, | |
| "logps/rejected": -804.6480712890625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.008499560877680779, | |
| "rewards/margins": 14.382369995117188, | |
| "rewards/rejected": -14.373868942260742, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.9220338983050848, | |
| "grad_norm": 0.0002716865446675816, | |
| "learning_rate": 4.147661453907635e-07, | |
| "logits/chosen": -1.6562086343765259, | |
| "logits/rejected": -1.8350610733032227, | |
| "logps/chosen": -153.1056365966797, | |
| "logps/rejected": -784.3397216796875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.011557278223335743, | |
| "rewards/margins": 14.29238510131836, | |
| "rewards/rejected": -14.280828475952148, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.9355932203389831, | |
| "grad_norm": 0.00021878390732588127, | |
| "learning_rate": 4.1193501690360834e-07, | |
| "logits/chosen": -1.582091212272644, | |
| "logits/rejected": -1.7254104614257812, | |
| "logps/chosen": -141.56887817382812, | |
| "logps/rejected": -816.9537353515625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.025769073516130447, | |
| "rewards/margins": 14.515641212463379, | |
| "rewards/rejected": -14.489871978759766, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.9491525423728814, | |
| "grad_norm": 0.0006094805270326929, | |
| "learning_rate": 4.0906764794596347e-07, | |
| "logits/chosen": -1.5609385967254639, | |
| "logits/rejected": -1.7095026969909668, | |
| "logps/chosen": -142.61056518554688, | |
| "logps/rejected": -779.0126953125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.011463040485978127, | |
| "rewards/margins": 14.310674667358398, | |
| "rewards/rejected": -14.299211502075195, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.9491525423728814, | |
| "eval_logits/chosen": -1.5444095134735107, | |
| "eval_logits/rejected": -1.686043620109558, | |
| "eval_logps/chosen": -129.370361328125, | |
| "eval_logps/rejected": -760.5565795898438, | |
| "eval_loss": 1.1656707101792563e-06, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": 0.019651925191283226, | |
| "eval_rewards/margins": 14.224937438964844, | |
| "eval_rewards/rejected": -14.205286026000977, | |
| "eval_runtime": 23.3494, | |
| "eval_samples_per_second": 4.283, | |
| "eval_steps_per_second": 1.071, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.9627118644067797, | |
| "grad_norm": 0.0006059188782453061, | |
| "learning_rate": 4.0616468022462013e-07, | |
| "logits/chosen": -1.6205476522445679, | |
| "logits/rejected": -1.785776138305664, | |
| "logps/chosen": -137.9543914794922, | |
| "logps/rejected": -812.373291015625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.015250766649842262, | |
| "rewards/margins": 14.447315216064453, | |
| "rewards/rejected": -14.432065963745117, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.976271186440678, | |
| "grad_norm": 0.00027120641837016974, | |
| "learning_rate": 4.0322676341324414e-07, | |
| "logits/chosen": -1.613734245300293, | |
| "logits/rejected": -1.7981483936309814, | |
| "logps/chosen": -133.99325561523438, | |
| "logps/rejected": -817.4915771484375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.011865440756082535, | |
| "rewards/margins": 14.308055877685547, | |
| "rewards/rejected": -14.296191215515137, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.9898305084745763, | |
| "grad_norm": 0.000231523046440184, | |
| "learning_rate": 4.002545550069808e-07, | |
| "logits/chosen": -1.6496268510818481, | |
| "logits/rejected": -1.841505527496338, | |
| "logps/chosen": -140.759765625, | |
| "logps/rejected": -784.6021118164062, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.014270871877670288, | |
| "rewards/margins": 14.330395698547363, | |
| "rewards/rejected": -14.31612491607666, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.0027118644067796, | |
| "grad_norm": 0.00021386925480897464, | |
| "learning_rate": 3.972487201753106e-07, | |
| "logits/chosen": -1.6074280738830566, | |
| "logits/rejected": -1.7795652151107788, | |
| "logps/chosen": -136.89691162109375, | |
| "logps/rejected": -791.8516845703125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.0169425867497921, | |
| "rewards/margins": 14.432229995727539, | |
| "rewards/rejected": -14.415288925170898, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.016271186440678, | |
| "grad_norm": 0.00040145408076981584, | |
| "learning_rate": 3.9420993161318615e-07, | |
| "logits/chosen": -1.6055034399032593, | |
| "logits/rejected": -1.77395761013031, | |
| "logps/chosen": -134.6471405029297, | |
| "logps/rejected": -791.5114135742188, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.014754224568605423, | |
| "rewards/margins": 14.438673973083496, | |
| "rewards/rejected": -14.423918724060059, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.0298305084745762, | |
| "grad_norm": 0.00025553044965608517, | |
| "learning_rate": 3.911388693904854e-07, | |
| "logits/chosen": -1.6558409929275513, | |
| "logits/rejected": -1.8468029499053955, | |
| "logps/chosen": -140.86526489257812, | |
| "logps/rejected": -796.6107788085938, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.010989008471369743, | |
| "rewards/margins": 14.3390531539917, | |
| "rewards/rejected": -14.32806396484375, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.0433898305084746, | |
| "grad_norm": 0.00032980079352337916, | |
| "learning_rate": 3.8803622079981496e-07, | |
| "logits/chosen": -1.578873872756958, | |
| "logits/rejected": -1.7349332571029663, | |
| "logps/chosen": -130.14547729492188, | |
| "logps/rejected": -764.939208984375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.0018688759300857782, | |
| "rewards/margins": 14.415413856506348, | |
| "rewards/rejected": -14.413544654846191, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.0569491525423729, | |
| "grad_norm": 0.00031172746262580516, | |
| "learning_rate": 3.8490268020269614e-07, | |
| "logits/chosen": -1.6029382944107056, | |
| "logits/rejected": -1.7419565916061401, | |
| "logps/chosen": -140.53709411621094, | |
| "logps/rejected": -811.80908203125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.016495605930685997, | |
| "rewards/margins": 14.483687400817871, | |
| "rewards/rejected": -14.467193603515625, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.0705084745762712, | |
| "grad_norm": 0.0003301712565175788, | |
| "learning_rate": 3.817389488741694e-07, | |
| "logits/chosen": -1.57521653175354, | |
| "logits/rejected": -1.7471016645431519, | |
| "logps/chosen": -133.0887451171875, | |
| "logps/rejected": -820.1575927734375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.027268720790743828, | |
| "rewards/margins": 14.49877643585205, | |
| "rewards/rejected": -14.471506118774414, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.0840677966101695, | |
| "grad_norm": 0.000600013552539676, | |
| "learning_rate": 3.785457348458516e-07, | |
| "logits/chosen": -1.6120831966400146, | |
| "logits/rejected": -1.783928394317627, | |
| "logps/chosen": -145.61685180664062, | |
| "logps/rejected": -781.661376953125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.0003941003233194351, | |
| "rewards/margins": 14.212892532348633, | |
| "rewards/rejected": -14.21249771118164, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.0840677966101695, | |
| "eval_logits/chosen": -1.5446038246154785, | |
| "eval_logits/rejected": -1.6855926513671875, | |
| "eval_logps/chosen": -129.36080932617188, | |
| "eval_logps/rejected": -761.1478881835938, | |
| "eval_loss": 1.1145154985570116e-06, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": 0.020607449114322662, | |
| "eval_rewards/margins": 14.285026550292969, | |
| "eval_rewards/rejected": -14.264419555664062, | |
| "eval_runtime": 23.5778, | |
| "eval_samples_per_second": 4.241, | |
| "eval_steps_per_second": 1.06, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.0976271186440678, | |
| "grad_norm": 0.00037026369248627336, | |
| "learning_rate": 3.753237527474812e-07, | |
| "logits/chosen": -1.5942147970199585, | |
| "logits/rejected": -1.7750327587127686, | |
| "logps/chosen": -137.2639923095703, | |
| "logps/rejected": -795.8204956054688, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.018559224903583527, | |
| "rewards/margins": 14.402730941772461, | |
| "rewards/rejected": -14.384172439575195, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.1111864406779661, | |
| "grad_norm": 0.0002646065707675486, | |
| "learning_rate": 3.7207372364698645e-07, | |
| "logits/chosen": -1.6505929231643677, | |
| "logits/rejected": -1.810257911682129, | |
| "logps/chosen": -137.0296173095703, | |
| "logps/rejected": -838.5280151367188, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.02975718304514885, | |
| "rewards/margins": 14.56403923034668, | |
| "rewards/rejected": -14.534282684326172, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.1247457627118644, | |
| "grad_norm": 0.0004516110377872378, | |
| "learning_rate": 3.687963748891131e-07, | |
| "logits/chosen": -1.5973923206329346, | |
| "logits/rejected": -1.7931503057479858, | |
| "logps/chosen": -148.9749755859375, | |
| "logps/rejected": -849.8316650390625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.007867500185966492, | |
| "rewards/margins": 14.508625030517578, | |
| "rewards/rejected": -14.50075626373291, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.1383050847457628, | |
| "grad_norm": 0.000400542368638696, | |
| "learning_rate": 3.6549243993264747e-07, | |
| "logits/chosen": -1.645015001296997, | |
| "logits/rejected": -1.8152064085006714, | |
| "logps/chosen": -134.02035522460938, | |
| "logps/rejected": -784.4261474609375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.00821724720299244, | |
| "rewards/margins": 14.368156433105469, | |
| "rewards/rejected": -14.359938621520996, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.151864406779661, | |
| "grad_norm": 0.0007766829343963702, | |
| "learning_rate": 3.6216265818627066e-07, | |
| "logits/chosen": -1.6092370748519897, | |
| "logits/rejected": -1.7626619338989258, | |
| "logps/chosen": -147.05447387695312, | |
| "logps/rejected": -797.9004516601562, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.005066628567874432, | |
| "rewards/margins": 14.369694709777832, | |
| "rewards/rejected": -14.36462688446045, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.1654237288135594, | |
| "grad_norm": 0.00034472860967714935, | |
| "learning_rate": 3.588077748430819e-07, | |
| "logits/chosen": -1.53743577003479, | |
| "logits/rejected": -1.6860629320144653, | |
| "logps/chosen": -130.87753295898438, | |
| "logps/rejected": -766.9425048828125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.017799021676182747, | |
| "rewards/margins": 14.47961139678955, | |
| "rewards/rejected": -14.461811065673828, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.1789830508474577, | |
| "grad_norm": 0.00026858839945804227, | |
| "learning_rate": 3.554285407138269e-07, | |
| "logits/chosen": -1.5462911128997803, | |
| "logits/rejected": -1.691042423248291, | |
| "logps/chosen": -129.9382781982422, | |
| "logps/rejected": -737.553955078125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.0059865121729671955, | |
| "rewards/margins": 14.388315200805664, | |
| "rewards/rejected": -14.382328987121582, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.192542372881356, | |
| "grad_norm": 0.00031207053329295663, | |
| "learning_rate": 3.5202571205886913e-07, | |
| "logits/chosen": -1.6693938970565796, | |
| "logits/rejected": -1.8746066093444824, | |
| "logps/chosen": -148.41986083984375, | |
| "logps/rejected": -816.6522827148438, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.023402290418744087, | |
| "rewards/margins": 14.48322582244873, | |
| "rewards/rejected": -14.459823608398438, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.2061016949152543, | |
| "grad_norm": 0.0004730088609444601, | |
| "learning_rate": 3.486000504189414e-07, | |
| "logits/chosen": -1.5571904182434082, | |
| "logits/rejected": -1.739476203918457, | |
| "logps/chosen": -131.6705322265625, | |
| "logps/rejected": -770.3902587890625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.011558154597878456, | |
| "rewards/margins": 14.456482887268066, | |
| "rewards/rejected": -14.444924354553223, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.2196610169491526, | |
| "grad_norm": 0.0004757504728478808, | |
| "learning_rate": 3.4515232244471606e-07, | |
| "logits/chosen": -1.4787291288375854, | |
| "logits/rejected": -1.661275029182434, | |
| "logps/chosen": -151.16482543945312, | |
| "logps/rejected": -888.5305786132812, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.014035153202712536, | |
| "rewards/margins": 14.75745964050293, | |
| "rewards/rejected": -14.743424415588379, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.2196610169491526, | |
| "eval_logits/chosen": -1.5419373512268066, | |
| "eval_logits/rejected": -1.685987949371338, | |
| "eval_logps/chosen": -129.385986328125, | |
| "eval_logps/rejected": -761.9303588867188, | |
| "eval_loss": 1.042011717800051e-06, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": 0.01808955892920494, | |
| "eval_rewards/margins": 14.360756874084473, | |
| "eval_rewards/rejected": -14.342667579650879, | |
| "eval_runtime": 23.2568, | |
| "eval_samples_per_second": 4.3, | |
| "eval_steps_per_second": 1.075, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.2332203389830507, | |
| "grad_norm": 0.0003716686876859065, | |
| "learning_rate": 3.41683299725231e-07, | |
| "logits/chosen": -1.6072585582733154, | |
| "logits/rejected": -1.7751410007476807, | |
| "logps/chosen": -121.9486312866211, | |
| "logps/rejected": -751.9805297851562, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.014703430235385895, | |
| "rewards/margins": 14.331403732299805, | |
| "rewards/rejected": -14.316699981689453, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.2467796610169493, | |
| "grad_norm": 0.00045583582930412936, | |
| "learning_rate": 3.3819375861521116e-07, | |
| "logits/chosen": -1.6996678113937378, | |
| "logits/rejected": -1.8710010051727295, | |
| "logps/chosen": -144.1748046875, | |
| "logps/rejected": -781.7122192382812, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.00705701345577836, | |
| "rewards/margins": 14.237908363342285, | |
| "rewards/rejected": -14.244966506958008, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.2603389830508474, | |
| "grad_norm": 0.00027042786850133124, | |
| "learning_rate": 3.346844800613229e-07, | |
| "logits/chosen": -1.635947823524475, | |
| "logits/rejected": -1.776626467704773, | |
| "logps/chosen": -122.72543334960938, | |
| "logps/rejected": -782.336669921875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.01217057928442955, | |
| "rewards/margins": 14.620203971862793, | |
| "rewards/rejected": -14.608033180236816, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.2738983050847459, | |
| "grad_norm": 0.00012613818860590913, | |
| "learning_rate": 3.311562494274009e-07, | |
| "logits/chosen": -1.5437029600143433, | |
| "logits/rejected": -1.7053155899047852, | |
| "logps/chosen": -146.37957763671875, | |
| "logps/rejected": -853.6106567382812, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.013366146944463253, | |
| "rewards/margins": 14.802824020385742, | |
| "rewards/rejected": -14.789458274841309, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.287457627118644, | |
| "grad_norm": 0.0009108583927914208, | |
| "learning_rate": 3.2760985631868716e-07, | |
| "logits/chosen": -1.6577401161193848, | |
| "logits/rejected": -1.8012009859085083, | |
| "logps/chosen": -128.98199462890625, | |
| "logps/rejected": -803.785400390625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.0071465689688920975, | |
| "rewards/margins": 14.534771919250488, | |
| "rewards/rejected": -14.527626037597656, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.3010169491525423, | |
| "grad_norm": 0.00018290415757236828, | |
| "learning_rate": 3.240460944051194e-07, | |
| "logits/chosen": -1.6779531240463257, | |
| "logits/rejected": -1.866011142730713, | |
| "logps/chosen": -139.95425415039062, | |
| "logps/rejected": -777.5191650390625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.005402210168540478, | |
| "rewards/margins": 14.424090385437012, | |
| "rewards/rejected": -14.41868782043457, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.3145762711864406, | |
| "grad_norm": 0.00032737612531831567, | |
| "learning_rate": 3.2046576124371106e-07, | |
| "logits/chosen": -1.6980135440826416, | |
| "logits/rejected": -1.8551394939422607, | |
| "logps/chosen": -135.28433227539062, | |
| "logps/rejected": -785.8731689453125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.01490587368607521, | |
| "rewards/margins": 14.301523208618164, | |
| "rewards/rejected": -14.286617279052734, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.328135593220339, | |
| "grad_norm": 0.00025153854347185144, | |
| "learning_rate": 3.1686965810006104e-07, | |
| "logits/chosen": -1.617310643196106, | |
| "logits/rejected": -1.804062008857727, | |
| "logps/chosen": -139.4668731689453, | |
| "logps/rejected": -800.4359130859375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.021481037139892578, | |
| "rewards/margins": 14.51413345336914, | |
| "rewards/rejected": -14.492652893066406, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.3416949152542372, | |
| "grad_norm": 0.000131467090970899, | |
| "learning_rate": 3.132585897690329e-07, | |
| "logits/chosen": -1.6868890523910522, | |
| "logits/rejected": -1.8882062435150146, | |
| "logps/chosen": -145.70948791503906, | |
| "logps/rejected": -857.4185180664062, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.008370953612029552, | |
| "rewards/margins": 14.516642570495605, | |
| "rewards/rejected": -14.508271217346191, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.3552542372881355, | |
| "grad_norm": 0.00031956784200775317, | |
| "learning_rate": 3.096333643946452e-07, | |
| "logits/chosen": -1.6276357173919678, | |
| "logits/rejected": -1.8146514892578125, | |
| "logps/chosen": -142.74488830566406, | |
| "logps/rejected": -807.506103515625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.012263872660696507, | |
| "rewards/margins": 14.435070037841797, | |
| "rewards/rejected": -14.422805786132812, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.3552542372881355, | |
| "eval_logits/chosen": -1.5454390048980713, | |
| "eval_logits/rejected": -1.6858749389648438, | |
| "eval_logps/chosen": -129.4228973388672, | |
| "eval_logps/rejected": -762.3302612304688, | |
| "eval_loss": 9.916642511598184e-07, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": 0.014398334547877312, | |
| "eval_rewards/margins": 14.39704704284668, | |
| "eval_rewards/rejected": -14.382649421691895, | |
| "eval_runtime": 23.2909, | |
| "eval_samples_per_second": 4.294, | |
| "eval_steps_per_second": 1.073, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.3688135593220339, | |
| "grad_norm": 0.0005110214664274667, | |
| "learning_rate": 3.059947932892113e-07, | |
| "logits/chosen": -1.581527590751648, | |
| "logits/rejected": -1.741445541381836, | |
| "logps/chosen": -130.16134643554688, | |
| "logps/rejected": -802.2377319335938, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.008667388930916786, | |
| "rewards/margins": 14.611343383789062, | |
| "rewards/rejected": -14.602675437927246, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.3823728813559322, | |
| "grad_norm": 0.000385109902014585, | |
| "learning_rate": 3.0234369075177105e-07, | |
| "logits/chosen": -1.6576886177062988, | |
| "logits/rejected": -1.8101606369018555, | |
| "logps/chosen": -139.96914672851562, | |
| "logps/rejected": -817.3280639648438, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.023594459518790245, | |
| "rewards/margins": 14.629423141479492, | |
| "rewards/rejected": -14.605830192565918, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.3959322033898305, | |
| "grad_norm": 0.00040091688342177467, | |
| "learning_rate": 2.9868087388585344e-07, | |
| "logits/chosen": -1.5842094421386719, | |
| "logits/rejected": -1.7406607866287231, | |
| "logps/chosen": -129.07803344726562, | |
| "logps/rejected": -770.8359985351562, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.013609381392598152, | |
| "rewards/margins": 14.5753812789917, | |
| "rewards/rejected": -14.561771392822266, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.4094915254237288, | |
| "grad_norm": 0.00026033035395961934, | |
| "learning_rate": 2.950071624166115e-07, | |
| "logits/chosen": -1.7157691717147827, | |
| "logits/rejected": -1.9252738952636719, | |
| "logps/chosen": -141.04685974121094, | |
| "logps/rejected": -809.7179565429688, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.015553833916783333, | |
| "rewards/margins": 14.591537475585938, | |
| "rewards/rejected": -14.575984954833984, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.423050847457627, | |
| "grad_norm": 0.00019886076594595082, | |
| "learning_rate": 2.9132337850737127e-07, | |
| "logits/chosen": -1.5866303443908691, | |
| "logits/rejected": -1.745476245880127, | |
| "logps/chosen": -136.6873779296875, | |
| "logps/rejected": -779.5513916015625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.01606130599975586, | |
| "rewards/margins": 14.531635284423828, | |
| "rewards/rejected": -14.51557445526123, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.4366101694915254, | |
| "grad_norm": 0.0003434669449456214, | |
| "learning_rate": 2.8763034657563425e-07, | |
| "logits/chosen": -1.594861626625061, | |
| "logits/rejected": -1.7643938064575195, | |
| "logps/chosen": -139.26364135742188, | |
| "logps/rejected": -812.1429443359375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.015910768881440163, | |
| "rewards/margins": 14.620904922485352, | |
| "rewards/rejected": -14.60499382019043, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.4501694915254237, | |
| "grad_norm": 0.00024295622286491588, | |
| "learning_rate": 2.839288931085761e-07, | |
| "logits/chosen": -1.6374831199645996, | |
| "logits/rejected": -1.8399394750595093, | |
| "logps/chosen": -146.24412536621094, | |
| "logps/rejected": -838.77587890625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.0162392258644104, | |
| "rewards/margins": 14.61961555480957, | |
| "rewards/rejected": -14.603377342224121, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.463728813559322, | |
| "grad_norm": 0.00021624383711316104, | |
| "learning_rate": 2.802198464780814e-07, | |
| "logits/chosen": -1.6567904949188232, | |
| "logits/rejected": -1.8536536693572998, | |
| "logps/chosen": -140.39337158203125, | |
| "logps/rejected": -793.81640625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.022600781172513962, | |
| "rewards/margins": 14.519584655761719, | |
| "rewards/rejected": -14.496984481811523, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.4772881355932204, | |
| "grad_norm": 0.00025925077468732225, | |
| "learning_rate": 2.765040367553572e-07, | |
| "logits/chosen": -1.6113574504852295, | |
| "logits/rejected": -1.8030925989151, | |
| "logps/chosen": -137.23849487304688, | |
| "logps/rejected": -808.6797485351562, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.005808805115520954, | |
| "rewards/margins": 14.623893737792969, | |
| "rewards/rejected": -14.618083953857422, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.4908474576271187, | |
| "grad_norm": 0.0003907081889108827, | |
| "learning_rate": 2.727822955251663e-07, | |
| "logits/chosen": -1.6336138248443604, | |
| "logits/rejected": -1.7939205169677734, | |
| "logps/chosen": -135.02374267578125, | |
| "logps/rejected": -803.9449462890625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.01596558652818203, | |
| "rewards/margins": 14.46599292755127, | |
| "rewards/rejected": -14.450026512145996, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.4908474576271187, | |
| "eval_logits/chosen": -1.5441298484802246, | |
| "eval_logits/rejected": -1.6875685453414917, | |
| "eval_logps/chosen": -129.44482421875, | |
| "eval_logps/rejected": -762.7908325195312, | |
| "eval_loss": 9.502831517238519e-07, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": 0.012204695492982864, | |
| "eval_rewards/margins": 14.44091510772705, | |
| "eval_rewards/rejected": -14.4287109375, | |
| "eval_runtime": 23.159, | |
| "eval_samples_per_second": 4.318, | |
| "eval_steps_per_second": 1.079, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.504406779661017, | |
| "grad_norm": 0.00022865739977937088, | |
| "learning_rate": 2.6905545569972124e-07, | |
| "logits/chosen": -1.5306434631347656, | |
| "logits/rejected": -1.7007068395614624, | |
| "logps/chosen": -141.87411499023438, | |
| "logps/rejected": -800.179931640625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.020658917725086212, | |
| "rewards/margins": 14.778863906860352, | |
| "rewards/rejected": -14.758204460144043, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.5179661016949153, | |
| "grad_norm": 0.00013390402806158302, | |
| "learning_rate": 2.6532435133228176e-07, | |
| "logits/chosen": -1.6342971324920654, | |
| "logits/rejected": -1.7988243103027344, | |
| "logps/chosen": -137.4864501953125, | |
| "logps/rejected": -778.6915283203125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.011233944445848465, | |
| "rewards/margins": 14.523719787597656, | |
| "rewards/rejected": -14.51248550415039, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.5315254237288136, | |
| "grad_norm": 0.0005026258076613325, | |
| "learning_rate": 2.615898174304967e-07, | |
| "logits/chosen": -1.6171948909759521, | |
| "logits/rejected": -1.7615530490875244, | |
| "logps/chosen": -157.30349731445312, | |
| "logps/rejected": -839.1676635742188, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.011025247164070606, | |
| "rewards/margins": 14.724686622619629, | |
| "rewards/rejected": -14.713661193847656, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.5450847457627117, | |
| "grad_norm": 0.00042097845856288896, | |
| "learning_rate": 2.5785268976953204e-07, | |
| "logits/chosen": -1.6210533380508423, | |
| "logits/rejected": -1.776179552078247, | |
| "logps/chosen": -138.84580993652344, | |
| "logps/rejected": -801.5419921875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.006160397548228502, | |
| "rewards/margins": 14.480558395385742, | |
| "rewards/rejected": -14.474397659301758, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.5586440677966102, | |
| "grad_norm": 0.00020478295097651352, | |
| "learning_rate": 2.541138047050281e-07, | |
| "logits/chosen": -1.5814204216003418, | |
| "logits/rejected": -1.7641853094100952, | |
| "logps/chosen": -137.55215454101562, | |
| "logps/rejected": -799.6993408203125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.021200550720095634, | |
| "rewards/margins": 14.847275733947754, | |
| "rewards/rejected": -14.826072692871094, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.5722033898305083, | |
| "grad_norm": 0.00023188287664193435, | |
| "learning_rate": 2.5037399898592537e-07, | |
| "logits/chosen": -1.5283310413360596, | |
| "logits/rejected": -1.697199821472168, | |
| "logps/chosen": -136.95700073242188, | |
| "logps/rejected": -806.5762329101562, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.025980135425925255, | |
| "rewards/margins": 14.58230209350586, | |
| "rewards/rejected": -14.55632209777832, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.5857627118644069, | |
| "grad_norm": 0.0002882430862738881, | |
| "learning_rate": 2.466341095672036e-07, | |
| "logits/chosen": -1.5664174556732178, | |
| "logits/rejected": -1.7281914949417114, | |
| "logps/chosen": -140.96392822265625, | |
| "logps/rejected": -860.5864868164062, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.015253797173500061, | |
| "rewards/margins": 14.810201644897461, | |
| "rewards/rejected": -14.794947624206543, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.599322033898305, | |
| "grad_norm": 0.00032442091374497006, | |
| "learning_rate": 2.428949734225744e-07, | |
| "logits/chosen": -1.5551780462265015, | |
| "logits/rejected": -1.7200207710266113, | |
| "logps/chosen": -116.5546646118164, | |
| "logps/rejected": -769.7291259765625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.020303260535001755, | |
| "rewards/margins": 14.525774002075195, | |
| "rewards/rejected": -14.505470275878906, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.6128813559322035, | |
| "grad_norm": 0.00026348293289965986, | |
| "learning_rate": 2.3915742735716914e-07, | |
| "logits/chosen": -1.5562556982040405, | |
| "logits/rejected": -1.721585988998413, | |
| "logps/chosen": -132.5054931640625, | |
| "logps/rejected": -780.1019287109375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.0179964117705822, | |
| "rewards/margins": 14.633602142333984, | |
| "rewards/rejected": -14.615605354309082, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.6264406779661016, | |
| "grad_norm": 0.0004614000733649941, | |
| "learning_rate": 2.3542230782026533e-07, | |
| "logits/chosen": -1.6813287734985352, | |
| "logits/rejected": -1.8313826322555542, | |
| "logps/chosen": -139.7794189453125, | |
| "logps/rejected": -794.0023803710938, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.018270863220095634, | |
| "rewards/margins": 14.51429271697998, | |
| "rewards/rejected": -14.496021270751953, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.6264406779661016, | |
| "eval_logits/chosen": -1.5334513187408447, | |
| "eval_logits/rejected": -1.6858574151992798, | |
| "eval_logps/chosen": -129.38221740722656, | |
| "eval_logps/rejected": -763.2522583007812, | |
| "eval_loss": 8.998525800052448e-07, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": 0.01846611313521862, | |
| "eval_rewards/margins": 14.493313789367676, | |
| "eval_rewards/rejected": -14.474847793579102, | |
| "eval_runtime": 23.3442, | |
| "eval_samples_per_second": 4.284, | |
| "eval_steps_per_second": 1.071, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.6400000000000001, | |
| "grad_norm": 0.0002485397852219414, | |
| "learning_rate": 2.3169045071809214e-07, | |
| "logits/chosen": -1.6468091011047363, | |
| "logits/rejected": -1.837449312210083, | |
| "logps/chosen": -132.42532348632812, | |
| "logps/rejected": -768.0850830078125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -8.118643017951399e-05, | |
| "rewards/margins": 14.453639030456543, | |
| "rewards/rejected": -14.453720092773438, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.6535593220338982, | |
| "grad_norm": 0.0002600309213234688, | |
| "learning_rate": 2.279626912267576e-07, | |
| "logits/chosen": -1.6016912460327148, | |
| "logits/rejected": -1.7917726039886475, | |
| "logps/chosen": -124.97222137451172, | |
| "logps/rejected": -763.31396484375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.0005855015479028225, | |
| "rewards/margins": 14.508626937866211, | |
| "rewards/rejected": -14.509212493896484, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.6671186440677968, | |
| "grad_norm": 0.0001866086825603447, | |
| "learning_rate": 2.2423986360533944e-07, | |
| "logits/chosen": -1.5999516248703003, | |
| "logits/rejected": -1.7527964115142822, | |
| "logps/chosen": -127.54315185546875, | |
| "logps/rejected": -779.42626953125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.03687942773103714, | |
| "rewards/margins": 14.615551948547363, | |
| "rewards/rejected": -14.578672409057617, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 1.6806779661016948, | |
| "grad_norm": 0.0001806454551986488, | |
| "learning_rate": 2.2052280100918053e-07, | |
| "logits/chosen": -1.6087480783462524, | |
| "logits/rejected": -1.733353614807129, | |
| "logps/chosen": -139.215576171875, | |
| "logps/rejected": -828.4327392578125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.012436101213097572, | |
| "rewards/margins": 14.784907341003418, | |
| "rewards/rejected": -14.77247142791748, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.6942372881355934, | |
| "grad_norm": 0.0003598095391952748, | |
| "learning_rate": 2.1681233530343226e-07, | |
| "logits/chosen": -1.6320774555206299, | |
| "logits/rejected": -1.770911455154419, | |
| "logps/chosen": -130.1085968017578, | |
| "logps/rejected": -781.075927734375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.008734717965126038, | |
| "rewards/margins": 14.645079612731934, | |
| "rewards/rejected": -14.636345863342285, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.7077966101694915, | |
| "grad_norm": 0.00021229397199613504, | |
| "learning_rate": 2.131092968768856e-07, | |
| "logits/chosen": -1.6914520263671875, | |
| "logits/rejected": -1.8534152507781982, | |
| "logps/chosen": -137.82667541503906, | |
| "logps/rejected": -795.1356811523438, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.01237824559211731, | |
| "rewards/margins": 14.534427642822266, | |
| "rewards/rejected": -14.522048950195312, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 1.7213559322033898, | |
| "grad_norm": 0.00034558435149114095, | |
| "learning_rate": 2.094145144561334e-07, | |
| "logits/chosen": -1.6149834394454956, | |
| "logits/rejected": -1.781679391860962, | |
| "logps/chosen": -135.41860961914062, | |
| "logps/rejected": -788.1875610351562, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.007942800410091877, | |
| "rewards/margins": 14.633131980895996, | |
| "rewards/rejected": -14.625189781188965, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 1.734915254237288, | |
| "grad_norm": 0.0003052738823642222, | |
| "learning_rate": 2.057288149201042e-07, | |
| "logits/chosen": -1.6356430053710938, | |
| "logits/rejected": -1.7994301319122314, | |
| "logps/chosen": -148.37130737304688, | |
| "logps/rejected": -835.1704711914062, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.00246133329346776, | |
| "rewards/margins": 14.854413032531738, | |
| "rewards/rejected": -14.851951599121094, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 1.7484745762711864, | |
| "grad_norm": 0.00021946391817271224, | |
| "learning_rate": 2.0205302311501e-07, | |
| "logits/chosen": -1.4806194305419922, | |
| "logits/rejected": -1.6502048969268799, | |
| "logps/chosen": -143.40792846679688, | |
| "logps/rejected": -826.89990234375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.022700820118188858, | |
| "rewards/margins": 14.935534477233887, | |
| "rewards/rejected": -14.912833213806152, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 1.7620338983050847, | |
| "grad_norm": 0.00018901278388815272, | |
| "learning_rate": 1.9838796166974835e-07, | |
| "logits/chosen": -1.6020697355270386, | |
| "logits/rejected": -1.7288063764572144, | |
| "logps/chosen": -118.82467651367188, | |
| "logps/rejected": -745.2700805664062, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.023728668689727783, | |
| "rewards/margins": 14.655136108398438, | |
| "rewards/rejected": -14.631406784057617, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.7620338983050847, | |
| "eval_logits/chosen": -1.541303038597107, | |
| "eval_logits/rejected": -1.6866235733032227, | |
| "eval_logps/chosen": -129.41481018066406, | |
| "eval_logps/rejected": -763.7774658203125, | |
| "eval_loss": 8.536229643141269e-07, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": 0.01520751602947712, | |
| "eval_rewards/margins": 14.542576789855957, | |
| "eval_rewards/rejected": -14.527369499206543, | |
| "eval_runtime": 23.2319, | |
| "eval_samples_per_second": 4.304, | |
| "eval_steps_per_second": 1.076, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.775593220338983, | |
| "grad_norm": 0.00024516636429012606, | |
| "learning_rate": 1.947344508118013e-07, | |
| "logits/chosen": -1.7006336450576782, | |
| "logits/rejected": -1.9313997030258179, | |
| "logps/chosen": -135.2695770263672, | |
| "logps/rejected": -803.58984375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.01268923282623291, | |
| "rewards/margins": 14.559757232666016, | |
| "rewards/rejected": -14.54706859588623, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 1.7891525423728813, | |
| "grad_norm": 0.0002825945001585042, | |
| "learning_rate": 1.9109330818367103e-07, | |
| "logits/chosen": -1.6026619672775269, | |
| "logits/rejected": -1.7476422786712646, | |
| "logps/chosen": -138.92730712890625, | |
| "logps/rejected": -774.2352294921875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.011833753436803818, | |
| "rewards/margins": 14.747849464416504, | |
| "rewards/rejected": -14.736017227172852, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 1.8027118644067797, | |
| "grad_norm": 0.00023343486470064288, | |
| "learning_rate": 1.8746534865989477e-07, | |
| "logits/chosen": -1.6764352321624756, | |
| "logits/rejected": -1.8599662780761719, | |
| "logps/chosen": -136.24755859375, | |
| "logps/rejected": -778.046630859375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.016610510647296906, | |
| "rewards/margins": 14.57569694519043, | |
| "rewards/rejected": -14.559085845947266, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 1.816271186440678, | |
| "grad_norm": 0.00020349358848448975, | |
| "learning_rate": 1.8385138416467886e-07, | |
| "logits/chosen": -1.619372010231018, | |
| "logits/rejected": -1.770355224609375, | |
| "logps/chosen": -137.48997497558594, | |
| "logps/rejected": -796.6224975585938, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.01347330305725336, | |
| "rewards/margins": 14.743675231933594, | |
| "rewards/rejected": -14.730201721191406, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 1.8298305084745763, | |
| "grad_norm": 0.000259228080505575, | |
| "learning_rate": 1.802522234901927e-07, | |
| "logits/chosen": -1.6564035415649414, | |
| "logits/rejected": -1.8519361019134521, | |
| "logps/chosen": -129.4320068359375, | |
| "logps/rejected": -815.1136474609375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.03748949244618416, | |
| "rewards/margins": 14.78825855255127, | |
| "rewards/rejected": -14.750767707824707, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.8433898305084746, | |
| "grad_norm": 0.0002871334654289991, | |
| "learning_rate": 1.7666867211556436e-07, | |
| "logits/chosen": -1.6476807594299316, | |
| "logits/rejected": -1.7843477725982666, | |
| "logps/chosen": -132.54351806640625, | |
| "logps/rejected": -779.9259643554688, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.004673462361097336, | |
| "rewards/margins": 14.616876602172852, | |
| "rewards/rejected": -14.612202644348145, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 1.856949152542373, | |
| "grad_norm": 0.00022022136706817435, | |
| "learning_rate": 1.7310153202661698e-07, | |
| "logits/chosen": -1.5748957395553589, | |
| "logits/rejected": -1.7366337776184082, | |
| "logps/chosen": -142.26052856445312, | |
| "logps/rejected": -840.873779296875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.01894519291818142, | |
| "rewards/margins": 14.885005950927734, | |
| "rewards/rejected": -14.866060256958008, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 1.8705084745762712, | |
| "grad_norm": 0.0002676827155917166, | |
| "learning_rate": 1.695516015363876e-07, | |
| "logits/chosen": -1.6750328540802002, | |
| "logits/rejected": -1.8339290618896484, | |
| "logps/chosen": -144.47813415527344, | |
| "logps/rejected": -832.0415649414062, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.023609986528754234, | |
| "rewards/margins": 14.777373313903809, | |
| "rewards/rejected": -14.753763198852539, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 1.8840677966101695, | |
| "grad_norm": 0.00024546650545054666, | |
| "learning_rate": 1.6601967510646718e-07, | |
| "logits/chosen": -1.725512981414795, | |
| "logits/rejected": -1.8677659034729004, | |
| "logps/chosen": -147.13035583496094, | |
| "logps/rejected": -842.0433349609375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.013028274290263653, | |
| "rewards/margins": 15.025127410888672, | |
| "rewards/rejected": -15.012099266052246, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 1.8976271186440679, | |
| "grad_norm": 0.00018426525023186526, | |
| "learning_rate": 1.6250654316920325e-07, | |
| "logits/chosen": -1.6960985660552979, | |
| "logits/rejected": -1.874474287033081, | |
| "logps/chosen": -138.58255004882812, | |
| "logps/rejected": -803.3692016601562, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.012537289410829544, | |
| "rewards/margins": 14.797500610351562, | |
| "rewards/rejected": -14.784963607788086, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.8976271186440679, | |
| "eval_logits/chosen": -1.5352431535720825, | |
| "eval_logits/rejected": -1.686131238937378, | |
| "eval_logps/chosen": -129.33590698242188, | |
| "eval_logps/rejected": -764.0711059570312, | |
| "eval_loss": 8.283525403385283e-07, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": 0.02309771627187729, | |
| "eval_rewards/margins": 14.579834938049316, | |
| "eval_rewards/rejected": -14.556737899780273, | |
| "eval_runtime": 23.3395, | |
| "eval_samples_per_second": 4.285, | |
| "eval_steps_per_second": 1.071, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.911186440677966, | |
| "grad_norm": 0.0002575697935659544, | |
| "learning_rate": 1.5901299195080392e-07, | |
| "logits/chosen": -1.6227922439575195, | |
| "logits/rejected": -1.7947490215301514, | |
| "logps/chosen": -130.98719787597656, | |
| "logps/rejected": -791.5205078125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.00749362725764513, | |
| "rewards/margins": 14.689105987548828, | |
| "rewards/rejected": -14.681612014770508, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 1.9247457627118645, | |
| "grad_norm": 0.00018017946205386775, | |
| "learning_rate": 1.5553980329538323e-07, | |
| "logits/chosen": -1.5999001264572144, | |
| "logits/rejected": -1.7650830745697021, | |
| "logps/chosen": -139.0697021484375, | |
| "logps/rejected": -810.0, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.010547073557972908, | |
| "rewards/margins": 14.652795791625977, | |
| "rewards/rejected": -14.642248153686523, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 1.9383050847457626, | |
| "grad_norm": 0.0001479458221901959, | |
| "learning_rate": 1.520877544899875e-07, | |
| "logits/chosen": -1.6732416152954102, | |
| "logits/rejected": -1.8719912767410278, | |
| "logps/chosen": -154.01966857910156, | |
| "logps/rejected": -817.1717529296875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.0030850698240101337, | |
| "rewards/margins": 14.651540756225586, | |
| "rewards/rejected": -14.654626846313477, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 1.951864406779661, | |
| "grad_norm": 0.0002763487512077753, | |
| "learning_rate": 1.4865761809064097e-07, | |
| "logits/chosen": -1.6813652515411377, | |
| "logits/rejected": -1.878997802734375, | |
| "logps/chosen": -125.57931518554688, | |
| "logps/rejected": -793.8370361328125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.015030990354716778, | |
| "rewards/margins": 14.654011726379395, | |
| "rewards/rejected": -14.638980865478516, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 1.9654237288135592, | |
| "grad_norm": 0.00016799758819048652, | |
| "learning_rate": 1.4525016174945103e-07, | |
| "logits/chosen": -1.7774598598480225, | |
| "logits/rejected": -1.9586834907531738, | |
| "logps/chosen": -128.68084716796875, | |
| "logps/rejected": -789.97021484375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.020244726911187172, | |
| "rewards/margins": 14.630430221557617, | |
| "rewards/rejected": -14.610186576843262, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.9789830508474577, | |
| "grad_norm": 0.00023868564973718838, | |
| "learning_rate": 1.4186614804280978e-07, | |
| "logits/chosen": -1.6796725988388062, | |
| "logits/rejected": -1.8473913669586182, | |
| "logps/chosen": -146.37466430664062, | |
| "logps/rejected": -850.5054931640625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.009393090382218361, | |
| "rewards/margins": 14.856435775756836, | |
| "rewards/rejected": -14.84704303741455, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 1.9925423728813558, | |
| "grad_norm": 0.000159343350698008, | |
| "learning_rate": 1.3850633430073286e-07, | |
| "logits/chosen": -1.5762964487075806, | |
| "logits/rejected": -1.7709100246429443, | |
| "logps/chosen": -151.92115783691406, | |
| "logps/rejected": -829.155029296875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.013924097642302513, | |
| "rewards/margins": 14.711325645446777, | |
| "rewards/rejected": -14.69740104675293, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 2.0054237288135592, | |
| "grad_norm": 0.00024916837620644227, | |
| "learning_rate": 1.3517147243737148e-07, | |
| "logits/chosen": -1.5804271697998047, | |
| "logits/rejected": -1.7189936637878418, | |
| "logps/chosen": -127.54490661621094, | |
| "logps/rejected": -792.7044067382812, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.021060097962617874, | |
| "rewards/margins": 14.793603897094727, | |
| "rewards/rejected": -14.772544860839844, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 2.0189830508474578, | |
| "grad_norm": 0.00028667386075358764, | |
| "learning_rate": 1.3186230878273653e-07, | |
| "logits/chosen": -1.6898704767227173, | |
| "logits/rejected": -1.8163692951202393, | |
| "logps/chosen": -133.3569793701172, | |
| "logps/rejected": -762.7427368164062, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.00874436367303133, | |
| "rewards/margins": 14.47949504852295, | |
| "rewards/rejected": -14.47075080871582, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 2.032542372881356, | |
| "grad_norm": 0.00030745885090509886, | |
| "learning_rate": 1.285795839156729e-07, | |
| "logits/chosen": -1.6830826997756958, | |
| "logits/rejected": -1.8209247589111328, | |
| "logps/chosen": -137.370361328125, | |
| "logps/rejected": -783.111083984375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.009560978040099144, | |
| "rewards/margins": 14.544479370117188, | |
| "rewards/rejected": -14.534917831420898, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.032542372881356, | |
| "eval_logits/chosen": -1.5413455963134766, | |
| "eval_logits/rejected": -1.6867271661758423, | |
| "eval_logps/chosen": -129.37197875976562, | |
| "eval_logps/rejected": -764.1895751953125, | |
| "eval_loss": 8.282594876618532e-07, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": 0.01949072629213333, | |
| "eval_rewards/margins": 14.588081359863281, | |
| "eval_rewards/rejected": -14.56859016418457, | |
| "eval_runtime": 23.3878, | |
| "eval_samples_per_second": 4.276, | |
| "eval_steps_per_second": 1.069, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.0461016949152544, | |
| "grad_norm": 0.00027440430133344047, | |
| "learning_rate": 1.2532403249812073e-07, | |
| "logits/chosen": -1.629027247428894, | |
| "logits/rejected": -1.7673338651657104, | |
| "logps/chosen": -143.42172241210938, | |
| "logps/rejected": -824.6468505859375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.004273869562894106, | |
| "rewards/margins": 14.795296669006348, | |
| "rewards/rejected": -14.791022300720215, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 2.0596610169491525, | |
| "grad_norm": 0.00023891678563468625, | |
| "learning_rate": 1.2209638311070024e-07, | |
| "logits/chosen": -1.6091238260269165, | |
| "logits/rejected": -1.7953072786331177, | |
| "logps/chosen": -127.51914978027344, | |
| "logps/rejected": -782.3043212890625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.0034638545475900173, | |
| "rewards/margins": 14.720000267028809, | |
| "rewards/rejected": -14.716536521911621, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 2.073220338983051, | |
| "grad_norm": 0.00024388492083099195, | |
| "learning_rate": 1.1889735808965853e-07, | |
| "logits/chosen": -1.6395263671875, | |
| "logits/rejected": -1.8181636333465576, | |
| "logps/chosen": -135.23330688476562, | |
| "logps/rejected": -783.9580078125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.009885641746222973, | |
| "rewards/margins": 14.711644172668457, | |
| "rewards/rejected": -14.701759338378906, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 2.086779661016949, | |
| "grad_norm": 5.583927886360231e-05, | |
| "learning_rate": 1.1572767336521322e-07, | |
| "logits/chosen": -1.7426564693450928, | |
| "logits/rejected": -1.9373841285705566, | |
| "logps/chosen": -163.0467529296875, | |
| "logps/rejected": -873.5623779296875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.01771201565861702, | |
| "rewards/margins": 14.745927810668945, | |
| "rewards/rejected": -14.728215217590332, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 2.1003389830508477, | |
| "grad_norm": 0.00017823331489187885, | |
| "learning_rate": 1.125880383013294e-07, | |
| "logits/chosen": -1.5471131801605225, | |
| "logits/rejected": -1.7288830280303955, | |
| "logps/chosen": -122.73018646240234, | |
| "logps/rejected": -782.6522216796875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.029936904087662697, | |
| "rewards/margins": 14.729475021362305, | |
| "rewards/rejected": -14.699538230895996, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.1138983050847457, | |
| "grad_norm": 0.0003091912161991017, | |
| "learning_rate": 1.0947915553696741e-07, | |
| "logits/chosen": -1.5830570459365845, | |
| "logits/rejected": -1.7391778230667114, | |
| "logps/chosen": -124.14356994628906, | |
| "logps/rejected": -795.6461791992188, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.01131663378328085, | |
| "rewards/margins": 14.700472831726074, | |
| "rewards/rejected": -14.689155578613281, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 2.127457627118644, | |
| "grad_norm": 0.00016494367259474433, | |
| "learning_rate": 1.0640172082883377e-07, | |
| "logits/chosen": -1.575786828994751, | |
| "logits/rejected": -1.7403693199157715, | |
| "logps/chosen": -143.2032470703125, | |
| "logps/rejected": -797.1260986328125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.01606675237417221, | |
| "rewards/margins": 14.676783561706543, | |
| "rewards/rejected": -14.660717010498047, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 2.1410169491525424, | |
| "grad_norm": 0.0001055398603476797, | |
| "learning_rate": 1.0335642289567453e-07, | |
| "logits/chosen": -1.5635545253753662, | |
| "logits/rejected": -1.6949539184570312, | |
| "logps/chosen": -137.00326538085938, | |
| "logps/rejected": -795.691650390625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.017065811902284622, | |
| "rewards/margins": 14.849153518676758, | |
| "rewards/rejected": -14.832088470458984, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 2.154576271186441, | |
| "grad_norm": 0.00015184404381463285, | |
| "learning_rate": 1.003439432641412e-07, | |
| "logits/chosen": -1.5640537738800049, | |
| "logits/rejected": -1.718758225440979, | |
| "logps/chosen": -145.822998046875, | |
| "logps/rejected": -822.9422607421875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.027230584993958473, | |
| "rewards/margins": 14.735852241516113, | |
| "rewards/rejected": -14.70862102508545, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 2.168135593220339, | |
| "grad_norm": 0.00017944006421282506, | |
| "learning_rate": 9.736495611626869e-08, | |
| "logits/chosen": -1.7426433563232422, | |
| "logits/rejected": -1.9045021533966064, | |
| "logps/chosen": -136.1781463623047, | |
| "logps/rejected": -826.3635864257812, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.009065323509275913, | |
| "rewards/margins": 14.765503883361816, | |
| "rewards/rejected": -14.774569511413574, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.168135593220339, | |
| "eval_logits/chosen": -1.5352487564086914, | |
| "eval_logits/rejected": -1.6869568824768066, | |
| "eval_logps/chosen": -129.40280151367188, | |
| "eval_logps/rejected": -764.318359375, | |
| "eval_loss": 8.185303954633127e-07, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": 0.016407504677772522, | |
| "eval_rewards/margins": 14.597875595092773, | |
| "eval_rewards/rejected": -14.58146858215332, | |
| "eval_runtime": 23.0522, | |
| "eval_samples_per_second": 4.338, | |
| "eval_steps_per_second": 1.084, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.181694915254237, | |
| "grad_norm": 0.0005351685788183639, | |
| "learning_rate": 9.442012813859495e-08, | |
| "logits/chosen": -1.7391018867492676, | |
| "logits/rejected": -1.9293949604034424, | |
| "logps/chosen": -147.13917541503906, | |
| "logps/rejected": -805.6615600585938, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.002429434796795249, | |
| "rewards/margins": 14.607394218444824, | |
| "rewards/rejected": -14.604965209960938, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 2.1952542372881356, | |
| "grad_norm": 0.0001727788888810501, | |
| "learning_rate": 9.151011837295967e-08, | |
| "logits/chosen": -1.5907230377197266, | |
| "logits/rejected": -1.779097318649292, | |
| "logps/chosen": -135.67388916015625, | |
| "logps/rejected": -828.772705078125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.02872363105416298, | |
| "rewards/margins": 14.719782829284668, | |
| "rewards/rejected": -14.691059112548828, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 2.2088135593220337, | |
| "grad_norm": 0.00025741808212769845, | |
| "learning_rate": 8.863557806901232e-08, | |
| "logits/chosen": -1.8002914190292358, | |
| "logits/rejected": -1.9408048391342163, | |
| "logps/chosen": -132.03172302246094, | |
| "logps/rejected": -784.93115234375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.011570243164896965, | |
| "rewards/margins": 14.813987731933594, | |
| "rewards/rejected": -14.80241584777832, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 2.2223728813559323, | |
| "grad_norm": 0.00021793936297483956, | |
| "learning_rate": 8.579715053846584e-08, | |
| "logits/chosen": -1.5967586040496826, | |
| "logits/rejected": -1.7385027408599854, | |
| "logps/chosen": -154.05177307128906, | |
| "logps/rejected": -800.445068359375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.005721386056393385, | |
| "rewards/margins": 14.734086036682129, | |
| "rewards/rejected": -14.728364944458008, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 2.2359322033898303, | |
| "grad_norm": 0.0002391850164899579, | |
| "learning_rate": 8.299547101112466e-08, | |
| "logits/chosen": -1.6040098667144775, | |
| "logits/rejected": -1.767350435256958, | |
| "logps/chosen": -144.35110473632812, | |
| "logps/rejected": -783.85107421875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.00950249470770359, | |
| "rewards/margins": 14.55084228515625, | |
| "rewards/rejected": -14.560342788696289, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.249491525423729, | |
| "grad_norm": 0.00031210724967463475, | |
| "learning_rate": 8.023116649272357e-08, | |
| "logits/chosen": -1.6417827606201172, | |
| "logits/rejected": -1.843612790107727, | |
| "logps/chosen": -140.8085479736328, | |
| "logps/rejected": -832.3277587890625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.01454589981585741, | |
| "rewards/margins": 14.897970199584961, | |
| "rewards/rejected": -14.883424758911133, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 2.263050847457627, | |
| "grad_norm": 0.0001537621460843329, | |
| "learning_rate": 7.750485562460529e-08, | |
| "logits/chosen": -1.637737512588501, | |
| "logits/rejected": -1.838837742805481, | |
| "logps/chosen": -151.76361083984375, | |
| "logps/rejected": -879.8406982421875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.028323961421847343, | |
| "rewards/margins": 15.002603530883789, | |
| "rewards/rejected": -14.974279403686523, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 2.2766101694915255, | |
| "grad_norm": 0.00014037312273128166, | |
| "learning_rate": 7.48171485452716e-08, | |
| "logits/chosen": -1.644343614578247, | |
| "logits/rejected": -1.8316127061843872, | |
| "logps/chosen": -148.27976989746094, | |
| "logps/rejected": -807.6490478515625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.002391491085290909, | |
| "rewards/margins": 14.774595260620117, | |
| "rewards/rejected": -14.776987075805664, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 2.2901694915254236, | |
| "grad_norm": 0.0003568019109756037, | |
| "learning_rate": 7.216864675383566e-08, | |
| "logits/chosen": -1.7114001512527466, | |
| "logits/rejected": -1.8946892023086548, | |
| "logps/chosen": -136.16168212890625, | |
| "logps/rejected": -822.518310546875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.0228084959089756, | |
| "rewards/margins": 14.955730438232422, | |
| "rewards/rejected": -14.93292236328125, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 2.303728813559322, | |
| "grad_norm": 0.00020642497846486078, | |
| "learning_rate": 6.955994297540946e-08, | |
| "logits/chosen": -1.582494854927063, | |
| "logits/rejected": -1.7430905103683472, | |
| "logps/chosen": -127.01213073730469, | |
| "logps/rejected": -749.1889038085938, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.004720491822808981, | |
| "rewards/margins": 14.59855842590332, | |
| "rewards/rejected": -14.593838691711426, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.303728813559322, | |
| "eval_logits/chosen": -1.5455721616744995, | |
| "eval_logits/rejected": -1.6867108345031738, | |
| "eval_logps/chosen": -129.34518432617188, | |
| "eval_logps/rejected": -764.3837890625, | |
| "eval_loss": 8.27629833111132e-07, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": 0.022169284522533417, | |
| "eval_rewards/margins": 14.610172271728516, | |
| "eval_rewards/rejected": -14.588001251220703, | |
| "eval_runtime": 23.2684, | |
| "eval_samples_per_second": 4.298, | |
| "eval_steps_per_second": 1.074, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.31728813559322, | |
| "grad_norm": 0.0003301800906268837, | |
| "learning_rate": 6.699162102845371e-08, | |
| "logits/chosen": -1.6968494653701782, | |
| "logits/rejected": -1.8610255718231201, | |
| "logps/chosen": -144.05172729492188, | |
| "logps/rejected": -792.3526000976562, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.0033090743236243725, | |
| "rewards/margins": 14.642684936523438, | |
| "rewards/rejected": -14.645994186401367, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 2.3308474576271188, | |
| "grad_norm": 0.00022300666262349114, | |
| "learning_rate": 6.446425569412145e-08, | |
| "logits/chosen": -1.5895376205444336, | |
| "logits/rejected": -1.7664618492126465, | |
| "logps/chosen": -147.58518981933594, | |
| "logps/rejected": -821.5425415039062, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.013544298708438873, | |
| "rewards/margins": 14.664928436279297, | |
| "rewards/rejected": -14.651383399963379, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 2.344406779661017, | |
| "grad_norm": 0.0002689016484107736, | |
| "learning_rate": 6.197841258762393e-08, | |
| "logits/chosen": -1.6732720136642456, | |
| "logits/rejected": -1.8248271942138672, | |
| "logps/chosen": -143.9190216064453, | |
| "logps/rejected": -798.0980224609375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.002302667824551463, | |
| "rewards/margins": 14.700139999389648, | |
| "rewards/rejected": -14.70244312286377, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 2.3579661016949154, | |
| "grad_norm": 0.00013823061030867806, | |
| "learning_rate": 5.95346480316484e-08, | |
| "logits/chosen": -1.6567462682724, | |
| "logits/rejected": -1.8002218008041382, | |
| "logps/chosen": -133.57330322265625, | |
| "logps/rejected": -790.8118896484375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.01070284005254507, | |
| "rewards/margins": 14.786917686462402, | |
| "rewards/rejected": -14.776214599609375, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 2.3715254237288135, | |
| "grad_norm": 0.00012708429822748938, | |
| "learning_rate": 5.713350893185462e-08, | |
| "logits/chosen": -1.540064811706543, | |
| "logits/rejected": -1.6900475025177002, | |
| "logps/chosen": -139.0050048828125, | |
| "logps/rejected": -813.723388671875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.011193746700882912, | |
| "rewards/margins": 14.780536651611328, | |
| "rewards/rejected": -14.769343376159668, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.385084745762712, | |
| "grad_norm": 0.00031972660791255883, | |
| "learning_rate": 5.4775532654479703e-08, | |
| "logits/chosen": -1.6927651166915894, | |
| "logits/rejected": -1.8659099340438843, | |
| "logps/chosen": -140.10357666015625, | |
| "logps/rejected": -804.2578125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.026886438950896263, | |
| "rewards/margins": 14.513021469116211, | |
| "rewards/rejected": -14.486133575439453, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 2.39864406779661, | |
| "grad_norm": 0.0002306535610201857, | |
| "learning_rate": 5.246124690607739e-08, | |
| "logits/chosen": -1.597672939300537, | |
| "logits/rejected": -1.7689580917358398, | |
| "logps/chosen": -136.6164093017578, | |
| "logps/rejected": -814.9773559570312, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.015394443646073341, | |
| "rewards/margins": 14.934272766113281, | |
| "rewards/rejected": -14.918877601623535, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 2.4122033898305086, | |
| "grad_norm": 0.00025827411726207306, | |
| "learning_rate": 5.019116961541928e-08, | |
| "logits/chosen": -1.6076412200927734, | |
| "logits/rejected": -1.787865400314331, | |
| "logps/chosen": -130.0896759033203, | |
| "logps/rejected": -771.5728149414062, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.018673747777938843, | |
| "rewards/margins": 14.67123794555664, | |
| "rewards/rejected": -14.652565002441406, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 2.4257627118644067, | |
| "grad_norm": 0.0001607149373365815, | |
| "learning_rate": 4.796580881758394e-08, | |
| "logits/chosen": -1.5243449211120605, | |
| "logits/rejected": -1.7028474807739258, | |
| "logps/chosen": -135.9856719970703, | |
| "logps/rejected": -788.3903198242188, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.013456443324685097, | |
| "rewards/margins": 14.943410873413086, | |
| "rewards/rejected": -14.92995548248291, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 2.4393220338983053, | |
| "grad_norm": 0.0008019666936544235, | |
| "learning_rate": 4.5785662540261035e-08, | |
| "logits/chosen": -1.5732166767120361, | |
| "logits/rejected": -1.7345067262649536, | |
| "logps/chosen": -123.94300842285156, | |
| "logps/rejected": -809.0182495117188, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.009817011654376984, | |
| "rewards/margins": 14.806741714477539, | |
| "rewards/rejected": -14.796926498413086, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.4393220338983053, | |
| "eval_logits/chosen": -1.5405449867248535, | |
| "eval_logits/rejected": -1.6869794130325317, | |
| "eval_logps/chosen": -129.44265747070312, | |
| "eval_logps/rejected": -764.5172729492188, | |
| "eval_loss": 8.009520229279588e-07, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": 0.012421955354511738, | |
| "eval_rewards/margins": 14.613780975341797, | |
| "eval_rewards/rejected": -14.601359367370605, | |
| "eval_runtime": 23.2676, | |
| "eval_samples_per_second": 4.298, | |
| "eval_steps_per_second": 1.074, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.4528813559322034, | |
| "grad_norm": 0.0005455920009654483, | |
| "learning_rate": 4.365121869229399e-08, | |
| "logits/chosen": -1.6582889556884766, | |
| "logits/rejected": -1.8295698165893555, | |
| "logps/chosen": -117.7739486694336, | |
| "logps/rejected": -741.6177368164062, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.011549338698387146, | |
| "rewards/margins": 14.508467674255371, | |
| "rewards/rejected": -14.496919631958008, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 2.4664406779661014, | |
| "grad_norm": 0.00023816862497805363, | |
| "learning_rate": 4.1562954954488194e-08, | |
| "logits/chosen": -1.6208901405334473, | |
| "logits/rejected": -1.81270432472229, | |
| "logps/chosen": -137.12130737304688, | |
| "logps/rejected": -798.8505249023438, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.011716622859239578, | |
| "rewards/margins": 14.657186508178711, | |
| "rewards/rejected": -14.64547061920166, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 0.0002564351546730618, | |
| "learning_rate": 3.952133867270749e-08, | |
| "logits/chosen": -1.6017396450042725, | |
| "logits/rejected": -1.760289192199707, | |
| "logps/chosen": -135.62570190429688, | |
| "logps/rejected": -787.9927978515625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.021429751068353653, | |
| "rewards/margins": 14.778738021850586, | |
| "rewards/rejected": -14.757308006286621, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 2.4935593220338985, | |
| "grad_norm": 0.00024431220204588973, | |
| "learning_rate": 3.7526826753284055e-08, | |
| "logits/chosen": -1.5733165740966797, | |
| "logits/rejected": -1.734438419342041, | |
| "logps/chosen": -135.39663696289062, | |
| "logps/rejected": -781.9830322265625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.021723030135035515, | |
| "rewards/margins": 14.634653091430664, | |
| "rewards/rejected": -14.612930297851562, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 2.5071186440677966, | |
| "grad_norm": 0.0002804770855324591, | |
| "learning_rate": 3.5579865560764086e-08, | |
| "logits/chosen": -1.6363115310668945, | |
| "logits/rejected": -1.8157660961151123, | |
| "logps/chosen": -143.70252990722656, | |
| "logps/rejected": -803.8099365234375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.008160289376974106, | |
| "rewards/margins": 14.804468154907227, | |
| "rewards/rejected": -14.796307563781738, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 2.5206779661016947, | |
| "grad_norm": 0.00042995034846870784, | |
| "learning_rate": 3.3680890818013506e-08, | |
| "logits/chosen": -1.7073619365692139, | |
| "logits/rejected": -1.8638834953308105, | |
| "logps/chosen": -136.8450469970703, | |
| "logps/rejected": -822.551025390625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.012727834284305573, | |
| "rewards/margins": 14.761212348937988, | |
| "rewards/rejected": -14.74848461151123, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 2.5342372881355932, | |
| "grad_norm": 0.0005052071520552922, | |
| "learning_rate": 3.183032750870443e-08, | |
| "logits/chosen": -1.7014129161834717, | |
| "logits/rejected": -1.890128493309021, | |
| "logps/chosen": -144.991943359375, | |
| "logps/rejected": -852.4500732421875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.019090861082077026, | |
| "rewards/margins": 14.904670715332031, | |
| "rewards/rejected": -14.885579109191895, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 2.5477966101694918, | |
| "grad_norm": 0.00029171939654467954, | |
| "learning_rate": 3.002858978220535e-08, | |
| "logits/chosen": -1.6215015649795532, | |
| "logits/rejected": -1.8319523334503174, | |
| "logps/chosen": -148.42018127441406, | |
| "logps/rejected": -826.58349609375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.017919525504112244, | |
| "rewards/margins": 14.893685340881348, | |
| "rewards/rejected": -14.875765800476074, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 2.56135593220339, | |
| "grad_norm": 0.00014374560387721727, | |
| "learning_rate": 2.8276080860896223e-08, | |
| "logits/chosen": -1.5488464832305908, | |
| "logits/rejected": -1.7413768768310547, | |
| "logps/chosen": -148.57704162597656, | |
| "logps/rejected": -838.784423828125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.015819568186998367, | |
| "rewards/margins": 15.020793914794922, | |
| "rewards/rejected": -15.004974365234375, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 2.574915254237288, | |
| "grad_norm": 0.00013430221339783704, | |
| "learning_rate": 2.65731929499286e-08, | |
| "logits/chosen": -1.6086156368255615, | |
| "logits/rejected": -1.7848008871078491, | |
| "logps/chosen": -144.15988159179688, | |
| "logps/rejected": -854.092529296875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.02485763281583786, | |
| "rewards/margins": 14.896612167358398, | |
| "rewards/rejected": -14.871755599975586, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.574915254237288, | |
| "eval_logits/chosen": -1.5360183715820312, | |
| "eval_logits/rejected": -1.686428189277649, | |
| "eval_logps/chosen": -129.38560485839844, | |
| "eval_logps/rejected": -764.64111328125, | |
| "eval_loss": 7.962561880958674e-07, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": 0.018127169460058212, | |
| "eval_rewards/margins": 14.63186264038086, | |
| "eval_rewards/rejected": -14.613737106323242, | |
| "eval_runtime": 23.1267, | |
| "eval_samples_per_second": 4.324, | |
| "eval_steps_per_second": 1.081, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.5884745762711865, | |
| "grad_norm": 0.0003822317334269131, | |
| "learning_rate": 2.492030714945162e-08, | |
| "logits/chosen": -1.5995548963546753, | |
| "logits/rejected": -1.76951003074646, | |
| "logps/chosen": -135.77032470703125, | |
| "logps/rejected": -790.2930908203125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.00021793553605675697, | |
| "rewards/margins": 14.569367408752441, | |
| "rewards/rejected": -14.569584846496582, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 2.6020338983050846, | |
| "grad_norm": 0.000792494216087496, | |
| "learning_rate": 2.3317793369322992e-08, | |
| "logits/chosen": -1.6509454250335693, | |
| "logits/rejected": -1.8071436882019043, | |
| "logps/chosen": -132.31130981445312, | |
| "logps/rejected": -799.173583984375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.030081693083047867, | |
| "rewards/margins": 14.740245819091797, | |
| "rewards/rejected": -14.710164070129395, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 2.615593220338983, | |
| "grad_norm": 0.0002708186052822774, | |
| "learning_rate": 2.1766010246324795e-08, | |
| "logits/chosen": -1.5743701457977295, | |
| "logits/rejected": -1.7702776193618774, | |
| "logps/chosen": -147.4547119140625, | |
| "logps/rejected": -827.8794555664062, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.006160350050777197, | |
| "rewards/margins": 14.79060173034668, | |
| "rewards/rejected": -14.784441947937012, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 2.629152542372881, | |
| "grad_norm": 0.00020442518354563292, | |
| "learning_rate": 2.026530506390156e-08, | |
| "logits/chosen": -1.5262815952301025, | |
| "logits/rejected": -1.6566275358200073, | |
| "logps/chosen": -147.53939819335938, | |
| "logps/rejected": -822.6124267578125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.014457417652010918, | |
| "rewards/margins": 14.995853424072266, | |
| "rewards/rejected": -14.981395721435547, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 2.6427118644067797, | |
| "grad_norm": 0.00014356861276548948, | |
| "learning_rate": 1.8816013674439885e-08, | |
| "logits/chosen": -1.6366736888885498, | |
| "logits/rejected": -1.816946268081665, | |
| "logps/chosen": -133.2000274658203, | |
| "logps/rejected": -794.9268798828125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.0276030246168375, | |
| "rewards/margins": 14.789546012878418, | |
| "rewards/rejected": -14.761942863464355, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 2.656271186440678, | |
| "grad_norm": 0.00018299695034949964, | |
| "learning_rate": 1.741846042410533e-08, | |
| "logits/chosen": -1.5427926778793335, | |
| "logits/rejected": -1.7075831890106201, | |
| "logps/chosen": -137.50941467285156, | |
| "logps/rejected": -807.8783569335938, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.010896590538322926, | |
| "rewards/margins": 14.576469421386719, | |
| "rewards/rejected": -14.565572738647461, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 2.6698305084745764, | |
| "grad_norm": 0.00016135810197605974, | |
| "learning_rate": 1.607295808025558e-08, | |
| "logits/chosen": -1.6068065166473389, | |
| "logits/rejected": -1.7660378217697144, | |
| "logps/chosen": -131.94476318359375, | |
| "logps/rejected": -811.60791015625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.019821301102638245, | |
| "rewards/margins": 14.834625244140625, | |
| "rewards/rejected": -14.814804077148438, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 2.6833898305084745, | |
| "grad_norm": 0.00023692180234191582, | |
| "learning_rate": 1.4779807761443635e-08, | |
| "logits/chosen": -1.6394813060760498, | |
| "logits/rejected": -1.8306336402893066, | |
| "logps/chosen": -143.0768585205078, | |
| "logps/rejected": -815.6318359375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.021681049838662148, | |
| "rewards/margins": 14.815585136413574, | |
| "rewards/rejected": -14.793903350830078, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 2.696949152542373, | |
| "grad_norm": 0.00026461222708204653, | |
| "learning_rate": 1.353929887002897e-08, | |
| "logits/chosen": -1.5212616920471191, | |
| "logits/rejected": -1.702871561050415, | |
| "logps/chosen": -137.1354522705078, | |
| "logps/rejected": -811.0231323242188, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.022010358050465584, | |
| "rewards/margins": 14.848760604858398, | |
| "rewards/rejected": -14.826749801635742, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 2.710508474576271, | |
| "grad_norm": 0.00034035092300092515, | |
| "learning_rate": 1.2351709027410145e-08, | |
| "logits/chosen": -1.5257236957550049, | |
| "logits/rejected": -1.6637029647827148, | |
| "logps/chosen": -125.89049530029297, | |
| "logps/rejected": -765.981201171875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.02222251147031784, | |
| "rewards/margins": 14.72728157043457, | |
| "rewards/rejected": -14.705059051513672, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.710508474576271, | |
| "eval_logits/chosen": -1.5505026578903198, | |
| "eval_logits/rejected": -1.686079740524292, | |
| "eval_logps/chosen": -129.3984832763672, | |
| "eval_logps/rejected": -764.6372680664062, | |
| "eval_loss": 7.91568538716092e-07, | |
| "eval_rewards/accuracies": 1.0, | |
| "eval_rewards/chosen": 0.016839873045682907, | |
| "eval_rewards/margins": 14.630191802978516, | |
| "eval_rewards/rejected": -14.613351821899414, | |
| "eval_runtime": 23.1883, | |
| "eval_samples_per_second": 4.313, | |
| "eval_steps_per_second": 1.078, | |
| "step": 2000 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 2211, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 400, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 145012404715520.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |