| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 3.0, |
| "eval_steps": 100, |
| "global_step": 2865, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.0010471204188481676, |
| "grad_norm": 0.006555078943058368, |
| "learning_rate": 1.7421602787456448e-07, |
| "logits/chosen": -9.6875, |
| "logits/rejected": -9.6875, |
| "logps/chosen": -980.0, |
| "logps/rejected": -980.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": 0.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": 0.0, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.010471204188481676, |
| "grad_norm": 0.005601721286211152, |
| "learning_rate": 1.7421602787456445e-06, |
| "logits/chosen": -9.875, |
| "logits/rejected": -9.875, |
| "logps/chosen": -1104.0, |
| "logps/rejected": -1104.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": 0.0022735595703125, |
| "rewards/margins": 0.0, |
| "rewards/rejected": 0.0022735595703125, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.020942408376963352, |
| "grad_norm": 0.0069638487839769276, |
| "learning_rate": 3.484320557491289e-06, |
| "logits/chosen": -10.5625, |
| "logits/rejected": -10.5625, |
| "logps/chosen": -944.0, |
| "logps/rejected": -944.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": 7.063150405883789e-06, |
| "rewards/margins": 0.0, |
| "rewards/rejected": 7.063150405883789e-06, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.031413612565445025, |
| "grad_norm": 0.006855279587352887, |
| "learning_rate": 5.226480836236934e-06, |
| "logits/chosen": -9.8125, |
| "logits/rejected": -9.8125, |
| "logps/chosen": -996.0, |
| "logps/rejected": -996.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -0.00421142578125, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -0.00421142578125, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.041884816753926704, |
| "grad_norm": 0.0063474904301538725, |
| "learning_rate": 6.968641114982578e-06, |
| "logits/chosen": -10.0, |
| "logits/rejected": -10.0, |
| "logps/chosen": -844.0, |
| "logps/rejected": -844.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -0.0079345703125, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -0.0079345703125, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.05235602094240838, |
| "grad_norm": 0.00620411474612823, |
| "learning_rate": 8.710801393728225e-06, |
| "logits/chosen": -10.25, |
| "logits/rejected": -10.25, |
| "logps/chosen": -848.0, |
| "logps/rejected": -848.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -0.0189208984375, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -0.0189208984375, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.06282722513089005, |
| "grad_norm": 0.006705380339555898, |
| "learning_rate": 1.0452961672473868e-05, |
| "logits/chosen": -10.125, |
| "logits/rejected": -10.125, |
| "logps/chosen": -968.0, |
| "logps/rejected": -968.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -0.037109375, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -0.037109375, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.07329842931937172, |
| "grad_norm": 0.00660900817995478, |
| "learning_rate": 1.2195121951219513e-05, |
| "logits/chosen": -10.3125, |
| "logits/rejected": -10.3125, |
| "logps/chosen": -908.0, |
| "logps/rejected": -908.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -0.05810546875, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -0.05810546875, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.08376963350785341, |
| "grad_norm": 0.006352759079716193, |
| "learning_rate": 1.3937282229965156e-05, |
| "logits/chosen": -9.9375, |
| "logits/rejected": -9.9375, |
| "logps/chosen": -912.0, |
| "logps/rejected": -912.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -0.0869140625, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -0.0869140625, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.09424083769633508, |
| "grad_norm": 0.007170138247269948, |
| "learning_rate": 1.56794425087108e-05, |
| "logits/chosen": -9.9375, |
| "logits/rejected": -9.9375, |
| "logps/chosen": -1064.0, |
| "logps/rejected": -1064.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -0.1279296875, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -0.1279296875, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.10471204188481675, |
| "grad_norm": 0.007343767311511533, |
| "learning_rate": 1.742160278745645e-05, |
| "logits/chosen": -9.625, |
| "logits/rejected": -9.625, |
| "logps/chosen": -872.0, |
| "logps/rejected": -872.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -0.2119140625, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -0.2119140625, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.10471204188481675, |
| "eval_logits/chosen": -10.25, |
| "eval_logits/rejected": -9.875, |
| "eval_logps/chosen": -340.0, |
| "eval_logps/rejected": -310.0, |
| "eval_loss": 0.6966249942779541, |
| "eval_rewards/accuracies": 0.45703125, |
| "eval_rewards/chosen": -0.22265625, |
| "eval_rewards/margins": -0.003387451171875, |
| "eval_rewards/rejected": -0.2197265625, |
| "eval_runtime": 48.7675, |
| "eval_samples_per_second": 41.011, |
| "eval_steps_per_second": 0.656, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.11518324607329843, |
| "grad_norm": 0.006984675454212801, |
| "learning_rate": 1.9163763066202093e-05, |
| "logits/chosen": -9.9375, |
| "logits/rejected": -9.9375, |
| "logps/chosen": -872.0, |
| "logps/rejected": -872.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -0.3046875, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -0.3046875, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.1256544502617801, |
| "grad_norm": 0.007272857723313964, |
| "learning_rate": 2.0905923344947736e-05, |
| "logits/chosen": -10.0, |
| "logits/rejected": -10.0, |
| "logps/chosen": -1032.0, |
| "logps/rejected": -1032.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -0.41796875, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -0.41796875, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.13612565445026178, |
| "grad_norm": 0.00836595394552543, |
| "learning_rate": 2.264808362369338e-05, |
| "logits/chosen": -10.0625, |
| "logits/rejected": -10.0625, |
| "logps/chosen": -940.0, |
| "logps/rejected": -940.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -0.53515625, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -0.53515625, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.14659685863874344, |
| "grad_norm": 0.00784807544822228, |
| "learning_rate": 2.4390243902439026e-05, |
| "logits/chosen": -10.25, |
| "logits/rejected": -10.25, |
| "logps/chosen": -1024.0, |
| "logps/rejected": -1024.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -0.76171875, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -0.76171875, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.15706806282722513, |
| "grad_norm": 0.009906290155426784, |
| "learning_rate": 2.6132404181184672e-05, |
| "logits/chosen": -11.375, |
| "logits/rejected": -11.375, |
| "logps/chosen": -1120.0, |
| "logps/rejected": -1120.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -1.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -1.0, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.16753926701570682, |
| "grad_norm": 0.012099590323176866, |
| "learning_rate": 2.7874564459930312e-05, |
| "logits/chosen": -12.9375, |
| "logits/rejected": -12.9375, |
| "logps/chosen": -1192.0, |
| "logps/rejected": -1192.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -1.6640625, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -1.6640625, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.17801047120418848, |
| "grad_norm": 0.023305629624208155, |
| "learning_rate": 2.9616724738675962e-05, |
| "logits/chosen": -13.125, |
| "logits/rejected": -13.125, |
| "logps/chosen": -1144.0, |
| "logps/rejected": -1144.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -2.9375, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -2.9375, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.18848167539267016, |
| "grad_norm": 0.04887998781295543, |
| "learning_rate": 3.13588850174216e-05, |
| "logits/chosen": -10.0, |
| "logits/rejected": -10.0, |
| "logps/chosen": -1392.0, |
| "logps/rejected": -1392.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -4.78125, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -4.78125, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.19895287958115182, |
| "grad_norm": 0.2068914203427902, |
| "learning_rate": 3.310104529616725e-05, |
| "logits/chosen": -9.4375, |
| "logits/rejected": -9.4375, |
| "logps/chosen": -1688.0, |
| "logps/rejected": -1688.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -6.6875, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -6.6875, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.2094240837696335, |
| "grad_norm": 0.19353065664497623, |
| "learning_rate": 3.48432055749129e-05, |
| "logits/chosen": -10.0625, |
| "logits/rejected": -10.0625, |
| "logps/chosen": -1528.0, |
| "logps/rejected": -1528.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -7.5, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -7.5, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.2094240837696335, |
| "eval_logits/chosen": -9.1875, |
| "eval_logits/rejected": -8.6875, |
| "eval_logps/chosen": -1440.0, |
| "eval_logps/rejected": -1328.0, |
| "eval_loss": 4.280875205993652, |
| "eval_rewards/accuracies": 0.431640625, |
| "eval_rewards/chosen": -11.1875, |
| "eval_rewards/margins": -0.85546875, |
| "eval_rewards/rejected": -10.375, |
| "eval_runtime": 46.5125, |
| "eval_samples_per_second": 42.999, |
| "eval_steps_per_second": 0.688, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.2198952879581152, |
| "grad_norm": 0.23477106176773008, |
| "learning_rate": 3.6585365853658535e-05, |
| "logits/chosen": -8.4375, |
| "logits/rejected": -8.4375, |
| "logps/chosen": -3568.0, |
| "logps/rejected": -3568.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -27.25, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -27.25, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.23036649214659685, |
| "grad_norm": 0.5852571533672618, |
| "learning_rate": 3.8327526132404185e-05, |
| "logits/chosen": -5.375, |
| "logits/rejected": -5.375, |
| "logps/chosen": -12032.0, |
| "logps/rejected": -12032.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -111.5, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -111.5, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.24083769633507854, |
| "grad_norm": 0.10713582241648234, |
| "learning_rate": 4.006968641114983e-05, |
| "logits/chosen": -2.09375, |
| "logits/rejected": -2.09375, |
| "logps/chosen": -26368.0, |
| "logps/rejected": -26368.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -254.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -254.0, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.2513089005235602, |
| "grad_norm": 0.054557900728832764, |
| "learning_rate": 4.181184668989547e-05, |
| "logits/chosen": -1.9140625, |
| "logits/rejected": -1.9140625, |
| "logps/chosen": -33792.0, |
| "logps/rejected": -33792.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -328.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -328.0, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.2617801047120419, |
| "grad_norm": 0.023634346316412404, |
| "learning_rate": 4.3554006968641115e-05, |
| "logits/chosen": -1.09375, |
| "logits/rejected": -1.09375, |
| "logps/chosen": -41216.0, |
| "logps/rejected": -41216.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -400.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -400.0, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.27225130890052357, |
| "grad_norm": 0.014667844099137744, |
| "learning_rate": 4.529616724738676e-05, |
| "logits/chosen": -0.94140625, |
| "logits/rejected": -0.94140625, |
| "logps/chosen": -39168.0, |
| "logps/rejected": -39168.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -384.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -384.0, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.28272251308900526, |
| "grad_norm": 0.012935492186638933, |
| "learning_rate": 4.703832752613241e-05, |
| "logits/chosen": -1.4921875, |
| "logits/rejected": -1.4921875, |
| "logps/chosen": -38656.0, |
| "logps/rejected": -38656.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -376.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -376.0, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.2931937172774869, |
| "grad_norm": 0.013092319913558765, |
| "learning_rate": 4.878048780487805e-05, |
| "logits/chosen": -2.046875, |
| "logits/rejected": -2.046875, |
| "logps/chosen": -41984.0, |
| "logps/rejected": -41984.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -412.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -412.0, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.3036649214659686, |
| "grad_norm": 0.009378391029981153, |
| "learning_rate": 4.999983293481418e-05, |
| "logits/chosen": -2.390625, |
| "logits/rejected": -2.390625, |
| "logps/chosen": -44032.0, |
| "logps/rejected": -44032.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -432.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -432.0, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.31413612565445026, |
| "grad_norm": 0.008713809636234436, |
| "learning_rate": 4.9996862949181595e-05, |
| "logits/chosen": -2.578125, |
| "logits/rejected": -2.578125, |
| "logps/chosen": -37888.0, |
| "logps/rejected": -37888.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -370.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -370.0, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.31413612565445026, |
| "eval_logits/chosen": -2.515625, |
| "eval_logits/rejected": -2.5, |
| "eval_logps/chosen": -32768.0, |
| "eval_logps/rejected": -28416.0, |
| "eval_loss": 95.71612548828125, |
| "eval_rewards/accuracies": 0.42578125, |
| "eval_rewards/chosen": -324.0, |
| "eval_rewards/margins": -44.0, |
| "eval_rewards/rejected": -280.0, |
| "eval_runtime": 46.5304, |
| "eval_samples_per_second": 42.983, |
| "eval_steps_per_second": 0.688, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.32460732984293195, |
| "grad_norm": 0.008303796218933282, |
| "learning_rate": 4.9990180911526896e-05, |
| "logits/chosen": -2.703125, |
| "logits/rejected": -2.703125, |
| "logps/chosen": -41728.0, |
| "logps/rejected": -41728.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -408.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -408.0, |
| "step": 310 |
| }, |
| { |
| "epoch": 0.33507853403141363, |
| "grad_norm": 0.007738817564694629, |
| "learning_rate": 4.997978781413745e-05, |
| "logits/chosen": -2.953125, |
| "logits/rejected": -2.953125, |
| "logps/chosen": -40704.0, |
| "logps/rejected": -40704.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -396.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -396.0, |
| "step": 320 |
| }, |
| { |
| "epoch": 0.34554973821989526, |
| "grad_norm": 0.007268383658364035, |
| "learning_rate": 4.996568520039565e-05, |
| "logits/chosen": -3.03125, |
| "logits/rejected": -3.03125, |
| "logps/chosen": -41984.0, |
| "logps/rejected": -41984.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -410.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -410.0, |
| "step": 330 |
| }, |
| { |
| "epoch": 0.35602094240837695, |
| "grad_norm": 0.011021065348790784, |
| "learning_rate": 4.994787516454973e-05, |
| "logits/chosen": -3.09375, |
| "logits/rejected": -3.09375, |
| "logps/chosen": -39424.0, |
| "logps/rejected": -39424.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -386.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -386.0, |
| "step": 340 |
| }, |
| { |
| "epoch": 0.36649214659685864, |
| "grad_norm": 0.008600185816883099, |
| "learning_rate": 4.992636035140273e-05, |
| "logits/chosen": -3.078125, |
| "logits/rejected": -3.078125, |
| "logps/chosen": -41216.0, |
| "logps/rejected": -41216.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -402.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -402.0, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.3769633507853403, |
| "grad_norm": 0.006923926754603383, |
| "learning_rate": 4.99011439559198e-05, |
| "logits/chosen": -3.171875, |
| "logits/rejected": -3.171875, |
| "logps/chosen": -40448.0, |
| "logps/rejected": -40448.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -396.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -396.0, |
| "step": 360 |
| }, |
| { |
| "epoch": 0.387434554973822, |
| "grad_norm": 0.007546587426387108, |
| "learning_rate": 4.987222972275369e-05, |
| "logits/chosen": -3.0625, |
| "logits/rejected": -3.0625, |
| "logps/chosen": -44032.0, |
| "logps/rejected": -44032.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -430.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -430.0, |
| "step": 370 |
| }, |
| { |
| "epoch": 0.39790575916230364, |
| "grad_norm": 0.007221581870219065, |
| "learning_rate": 4.983962194568872e-05, |
| "logits/chosen": -2.90625, |
| "logits/rejected": -2.90625, |
| "logps/chosen": -42496.0, |
| "logps/rejected": -42496.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -414.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -414.0, |
| "step": 380 |
| }, |
| { |
| "epoch": 0.4083769633507853, |
| "grad_norm": 0.0073727398185322165, |
| "learning_rate": 4.9803325467003084e-05, |
| "logits/chosen": -2.9375, |
| "logits/rejected": -2.9375, |
| "logps/chosen": -41216.0, |
| "logps/rejected": -41216.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -402.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -402.0, |
| "step": 390 |
| }, |
| { |
| "epoch": 0.418848167539267, |
| "grad_norm": 0.006526823547348211, |
| "learning_rate": 4.9763345676749816e-05, |
| "logits/chosen": -2.953125, |
| "logits/rejected": -2.953125, |
| "logps/chosen": -41984.0, |
| "logps/rejected": -41984.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -412.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -412.0, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.418848167539267, |
| "eval_logits/chosen": -2.640625, |
| "eval_logits/rejected": -2.625, |
| "eval_logps/chosen": -34048.0, |
| "eval_logps/rejected": -29440.0, |
| "eval_loss": 99.55337524414062, |
| "eval_rewards/accuracies": 0.427734375, |
| "eval_rewards/chosen": -338.0, |
| "eval_rewards/margins": -46.0, |
| "eval_rewards/rejected": -292.0, |
| "eval_runtime": 46.5201, |
| "eval_samples_per_second": 42.992, |
| "eval_steps_per_second": 0.688, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.4293193717277487, |
| "grad_norm": 0.006686742269692169, |
| "learning_rate": 4.9719688511956374e-05, |
| "logits/chosen": -2.921875, |
| "logits/rejected": -2.921875, |
| "logps/chosen": -38656.0, |
| "logps/rejected": -38656.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -376.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -376.0, |
| "step": 410 |
| }, |
| { |
| "epoch": 0.4397905759162304, |
| "grad_norm": 0.009588123381166306, |
| "learning_rate": 4.967236045574294e-05, |
| "logits/chosen": -2.875, |
| "logits/rejected": -2.875, |
| "logps/chosen": -43520.0, |
| "logps/rejected": -43520.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -426.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -426.0, |
| "step": 420 |
| }, |
| { |
| "epoch": 0.450261780104712, |
| "grad_norm": 0.007763726668584196, |
| "learning_rate": 4.96213685363597e-05, |
| "logits/chosen": -2.84375, |
| "logits/rejected": -2.84375, |
| "logps/chosen": -40448.0, |
| "logps/rejected": -40448.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -394.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -394.0, |
| "step": 430 |
| }, |
| { |
| "epoch": 0.4607329842931937, |
| "grad_norm": 0.007556216782699217, |
| "learning_rate": 4.956672032614315e-05, |
| "logits/chosen": -2.890625, |
| "logits/rejected": -2.890625, |
| "logps/chosen": -42752.0, |
| "logps/rejected": -42752.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -418.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -418.0, |
| "step": 440 |
| }, |
| { |
| "epoch": 0.4712041884816754, |
| "grad_norm": 0.00991493433328749, |
| "learning_rate": 4.950842394039157e-05, |
| "logits/chosen": -2.828125, |
| "logits/rejected": -2.828125, |
| "logps/chosen": -41216.0, |
| "logps/rejected": -41216.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -404.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -404.0, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.4816753926701571, |
| "grad_norm": 0.01050462220678678, |
| "learning_rate": 4.944648803615994e-05, |
| "logits/chosen": -2.71875, |
| "logits/rejected": -2.71875, |
| "logps/chosen": -41216.0, |
| "logps/rejected": -41216.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -404.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -404.0, |
| "step": 460 |
| }, |
| { |
| "epoch": 0.49214659685863876, |
| "grad_norm": 0.010846672297755051, |
| "learning_rate": 4.938092181097433e-05, |
| "logits/chosen": -2.765625, |
| "logits/rejected": -2.765625, |
| "logps/chosen": -44544.0, |
| "logps/rejected": -44544.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -436.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -436.0, |
| "step": 470 |
| }, |
| { |
| "epoch": 0.5026178010471204, |
| "grad_norm": 0.009130343389350109, |
| "learning_rate": 4.931173500146607e-05, |
| "logits/chosen": -2.828125, |
| "logits/rejected": -2.828125, |
| "logps/chosen": -41216.0, |
| "logps/rejected": -41216.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -402.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -402.0, |
| "step": 480 |
| }, |
| { |
| "epoch": 0.5130890052356021, |
| "grad_norm": 0.010601736365248969, |
| "learning_rate": 4.923893788192582e-05, |
| "logits/chosen": -2.6875, |
| "logits/rejected": -2.6875, |
| "logps/chosen": -45056.0, |
| "logps/rejected": -45056.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -440.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -440.0, |
| "step": 490 |
| }, |
| { |
| "epoch": 0.5235602094240838, |
| "grad_norm": 0.010216303180097145, |
| "learning_rate": 4.9162541262777906e-05, |
| "logits/chosen": -2.734375, |
| "logits/rejected": -2.734375, |
| "logps/chosen": -45312.0, |
| "logps/rejected": -45312.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -444.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -444.0, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.5235602094240838, |
| "eval_logits/chosen": -2.046875, |
| "eval_logits/rejected": -1.96875, |
| "eval_logps/chosen": -35328.0, |
| "eval_logps/rejected": -30592.0, |
| "eval_loss": 103.50824737548828, |
| "eval_rewards/accuracies": 0.427734375, |
| "eval_rewards/chosen": -352.0, |
| "eval_rewards/margins": -48.0, |
| "eval_rewards/rejected": -304.0, |
| "eval_runtime": 46.5236, |
| "eval_samples_per_second": 42.989, |
| "eval_steps_per_second": 0.688, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.5340314136125655, |
| "grad_norm": 0.008896621945470987, |
| "learning_rate": 4.90825564889749e-05, |
| "logits/chosen": -2.6875, |
| "logits/rejected": -2.6875, |
| "logps/chosen": -41728.0, |
| "logps/rejected": -41728.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -410.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -410.0, |
| "step": 510 |
| }, |
| { |
| "epoch": 0.5445026178010471, |
| "grad_norm": 0.011486140596960492, |
| "learning_rate": 4.899899543831288e-05, |
| "logits/chosen": -2.65625, |
| "logits/rejected": -2.65625, |
| "logps/chosen": -46080.0, |
| "logps/rejected": -46080.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -450.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -450.0, |
| "step": 520 |
| }, |
| { |
| "epoch": 0.5549738219895288, |
| "grad_norm": 0.011821024724913537, |
| "learning_rate": 4.891187051966765e-05, |
| "logits/chosen": -2.53125, |
| "logits/rejected": -2.53125, |
| "logps/chosen": -41728.0, |
| "logps/rejected": -41728.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -408.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -408.0, |
| "step": 530 |
| }, |
| { |
| "epoch": 0.5654450261780105, |
| "grad_norm": 0.01233621281173854, |
| "learning_rate": 4.882119467115191e-05, |
| "logits/chosen": -2.734375, |
| "logits/rejected": -2.734375, |
| "logps/chosen": -45568.0, |
| "logps/rejected": -45568.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -446.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -446.0, |
| "step": 540 |
| }, |
| { |
| "epoch": 0.5759162303664922, |
| "grad_norm": 0.00996381800251901, |
| "learning_rate": 4.872698135819402e-05, |
| "logits/chosen": -2.125, |
| "logits/rejected": -2.125, |
| "logps/chosen": -44032.0, |
| "logps/rejected": -44032.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -430.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -430.0, |
| "step": 550 |
| }, |
| { |
| "epoch": 0.5863874345549738, |
| "grad_norm": 0.01414442232829592, |
| "learning_rate": 4.862924457153831e-05, |
| "logits/chosen": -2.375, |
| "logits/rejected": -2.375, |
| "logps/chosen": -46336.0, |
| "logps/rejected": -46336.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -454.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -454.0, |
| "step": 560 |
| }, |
| { |
| "epoch": 0.5968586387434555, |
| "grad_norm": 0.010444564487859895, |
| "learning_rate": 4.8527998825167484e-05, |
| "logits/chosen": -2.1875, |
| "logits/rejected": -2.1875, |
| "logps/chosen": -44288.0, |
| "logps/rejected": -44288.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -432.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -432.0, |
| "step": 570 |
| }, |
| { |
| "epoch": 0.6073298429319371, |
| "grad_norm": 0.012240238178969813, |
| "learning_rate": 4.842325915414728e-05, |
| "logits/chosen": -2.1875, |
| "logits/rejected": -2.1875, |
| "logps/chosen": -44288.0, |
| "logps/rejected": -44288.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -432.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -432.0, |
| "step": 580 |
| }, |
| { |
| "epoch": 0.6178010471204188, |
| "grad_norm": 0.009697072668010585, |
| "learning_rate": 4.831504111239373e-05, |
| "logits/chosen": -2.296875, |
| "logits/rejected": -2.296875, |
| "logps/chosen": -46336.0, |
| "logps/rejected": -46336.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -454.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -454.0, |
| "step": 590 |
| }, |
| { |
| "epoch": 0.6282722513089005, |
| "grad_norm": 0.010340526372097427, |
| "learning_rate": 4.8203360770363425e-05, |
| "logits/chosen": -2.359375, |
| "logits/rejected": -2.359375, |
| "logps/chosen": -45312.0, |
| "logps/rejected": -45312.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -442.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -442.0, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.6282722513089005, |
| "eval_logits/chosen": -1.265625, |
| "eval_logits/rejected": -1.1328125, |
| "eval_logps/chosen": -36864.0, |
| "eval_logps/rejected": -31872.0, |
| "eval_loss": 107.68787384033203, |
| "eval_rewards/accuracies": 0.431640625, |
| "eval_rewards/chosen": -366.0, |
| "eval_rewards/margins": -50.0, |
| "eval_rewards/rejected": -316.0, |
| "eval_runtime": 46.5267, |
| "eval_samples_per_second": 42.986, |
| "eval_steps_per_second": 0.688, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.6387434554973822, |
| "grad_norm": 0.01103268438557989, |
| "learning_rate": 4.808823471266702e-05, |
| "logits/chosen": -2.203125, |
| "logits/rejected": -2.203125, |
| "logps/chosen": -44288.0, |
| "logps/rejected": -44288.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -434.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -434.0, |
| "step": 610 |
| }, |
| { |
| "epoch": 0.6492146596858639, |
| "grad_norm": 0.010407185584878428, |
| "learning_rate": 4.796968003560639e-05, |
| "logits/chosen": -2.234375, |
| "logits/rejected": -2.234375, |
| "logps/chosen": -48384.0, |
| "logps/rejected": -48384.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -472.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -472.0, |
| "step": 620 |
| }, |
| { |
| "epoch": 0.6596858638743456, |
| "grad_norm": 0.01267464370562806, |
| "learning_rate": 4.7847714344635864e-05, |
| "logits/chosen": -1.8359375, |
| "logits/rejected": -1.8359375, |
| "logps/chosen": -43776.0, |
| "logps/rejected": -43776.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -428.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -428.0, |
| "step": 630 |
| }, |
| { |
| "epoch": 0.6701570680628273, |
| "grad_norm": 0.012014336120441012, |
| "learning_rate": 4.772235575174776e-05, |
| "logits/chosen": -2.03125, |
| "logits/rejected": -2.03125, |
| "logps/chosen": -45824.0, |
| "logps/rejected": -45824.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -448.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -448.0, |
| "step": 640 |
| }, |
| { |
| "epoch": 0.680628272251309, |
| "grad_norm": 0.012385861590163928, |
| "learning_rate": 4.7593622872782754e-05, |
| "logits/chosen": -1.703125, |
| "logits/rejected": -1.703125, |
| "logps/chosen": -44544.0, |
| "logps/rejected": -44544.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -438.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -438.0, |
| "step": 650 |
| }, |
| { |
| "epoch": 0.6910994764397905, |
| "grad_norm": 0.011458855874050362, |
| "learning_rate": 4.746153482466542e-05, |
| "logits/chosen": -2.015625, |
| "logits/rejected": -2.015625, |
| "logps/chosen": -46592.0, |
| "logps/rejected": -46592.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -458.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -458.0, |
| "step": 660 |
| }, |
| { |
| "epoch": 0.7015706806282722, |
| "grad_norm": 0.011919153818594076, |
| "learning_rate": 4.732611122256537e-05, |
| "logits/chosen": -2.234375, |
| "logits/rejected": -2.234375, |
| "logps/chosen": -44800.0, |
| "logps/rejected": -44800.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -440.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -440.0, |
| "step": 670 |
| }, |
| { |
| "epoch": 0.7120418848167539, |
| "grad_norm": 0.010898422338145869, |
| "learning_rate": 4.718737217698435e-05, |
| "logits/chosen": -2.203125, |
| "logits/rejected": -2.203125, |
| "logps/chosen": -47872.0, |
| "logps/rejected": -47872.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -470.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -470.0, |
| "step": 680 |
| }, |
| { |
| "epoch": 0.7225130890052356, |
| "grad_norm": 0.013195063494328692, |
| "learning_rate": 4.704533829076982e-05, |
| "logits/chosen": -2.21875, |
| "logits/rejected": -2.21875, |
| "logps/chosen": -49152.0, |
| "logps/rejected": -49152.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -482.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -482.0, |
| "step": 690 |
| }, |
| { |
| "epoch": 0.7329842931937173, |
| "grad_norm": 0.01317048727887716, |
| "learning_rate": 4.6900030656055455e-05, |
| "logits/chosen": -2.5, |
| "logits/rejected": -2.5, |
| "logps/chosen": -52224.0, |
| "logps/rejected": -52224.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -512.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -512.0, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.7329842931937173, |
| "eval_logits/chosen": -0.703125, |
| "eval_logits/rejected": -0.51171875, |
| "eval_logps/chosen": -38400.0, |
| "eval_logps/rejected": -33024.0, |
| "eval_loss": 111.89299774169922, |
| "eval_rewards/accuracies": 0.4296875, |
| "eval_rewards/chosen": -380.0, |
| "eval_rewards/margins": -52.0, |
| "eval_rewards/rejected": -328.0, |
| "eval_runtime": 46.5281, |
| "eval_samples_per_second": 42.985, |
| "eval_steps_per_second": 0.688, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.743455497382199, |
| "grad_norm": 0.011649512564205263, |
| "learning_rate": 4.6751470851128906e-05, |
| "logits/chosen": -1.375, |
| "logits/rejected": -1.375, |
| "logps/chosen": -47360.0, |
| "logps/rejected": -47360.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -466.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -466.0, |
| "step": 710 |
| }, |
| { |
| "epoch": 0.7539267015706806, |
| "grad_norm": 0.011965694033072682, |
| "learning_rate": 4.6599680937227454e-05, |
| "logits/chosen": -1.4375, |
| "logits/rejected": -1.4375, |
| "logps/chosen": -43520.0, |
| "logps/rejected": -43520.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -428.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -428.0, |
| "step": 720 |
| }, |
| { |
| "epoch": 0.7643979057591623, |
| "grad_norm": 0.012813350426964918, |
| "learning_rate": 4.644468345526183e-05, |
| "logits/chosen": -1.7578125, |
| "logits/rejected": -1.7578125, |
| "logps/chosen": -50176.0, |
| "logps/rejected": -50176.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -492.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -492.0, |
| "step": 730 |
| }, |
| { |
| "epoch": 0.774869109947644, |
| "grad_norm": 0.014796812622249505, |
| "learning_rate": 4.628650142246894e-05, |
| "logits/chosen": -1.7265625, |
| "logits/rejected": -1.7265625, |
| "logps/chosen": -47872.0, |
| "logps/rejected": -47872.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -468.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -468.0, |
| "step": 740 |
| }, |
| { |
| "epoch": 0.7853403141361257, |
| "grad_norm": 0.013773988034363978, |
| "learning_rate": 4.6125158328993765e-05, |
| "logits/chosen": -1.609375, |
| "logits/rejected": -1.609375, |
| "logps/chosen": -51968.0, |
| "logps/rejected": -51968.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -510.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -510.0, |
| "step": 750 |
| }, |
| { |
| "epoch": 0.7958115183246073, |
| "grad_norm": 0.012553057898071013, |
| "learning_rate": 4.5960678134401064e-05, |
| "logits/chosen": -0.9765625, |
| "logits/rejected": -0.9765625, |
| "logps/chosen": -47872.0, |
| "logps/rejected": -47872.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -470.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -470.0, |
| "step": 760 |
| }, |
| { |
| "epoch": 0.806282722513089, |
| "grad_norm": 0.011553452725455915, |
| "learning_rate": 4.579308526411734e-05, |
| "logits/chosen": -0.703125, |
| "logits/rejected": -0.703125, |
| "logps/chosen": -45824.0, |
| "logps/rejected": -45824.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -450.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -450.0, |
| "step": 770 |
| }, |
| { |
| "epoch": 0.8167539267015707, |
| "grad_norm": 0.014357426173081807, |
| "learning_rate": 4.562240460580367e-05, |
| "logits/chosen": -0.76171875, |
| "logits/rejected": -0.76171875, |
| "logps/chosen": -46336.0, |
| "logps/rejected": -46336.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -454.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -454.0, |
| "step": 780 |
| }, |
| { |
| "epoch": 0.8272251308900523, |
| "grad_norm": 0.014858643680132504, |
| "learning_rate": 4.544866150565985e-05, |
| "logits/chosen": -0.63671875, |
| "logits/rejected": -0.63671875, |
| "logps/chosen": -51712.0, |
| "logps/rejected": -51712.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -510.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -510.0, |
| "step": 790 |
| }, |
| { |
| "epoch": 0.837696335078534, |
| "grad_norm": 0.01434040424070902, |
| "learning_rate": 4.527188176466051e-05, |
| "logits/chosen": -0.447265625, |
| "logits/rejected": -0.447265625, |
| "logps/chosen": -49408.0, |
| "logps/rejected": -49408.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -484.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -484.0, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.837696335078534, |
| "eval_logits/chosen": 1.078125, |
| "eval_logits/rejected": 1.390625, |
| "eval_logps/chosen": -39680.0, |
| "eval_logps/rejected": -34304.0, |
| "eval_loss": 116.29875183105469, |
| "eval_rewards/accuracies": 0.435546875, |
| "eval_rewards/chosen": -394.0, |
| "eval_rewards/margins": -54.0, |
| "eval_rewards/rejected": -340.0, |
| "eval_runtime": 46.5272, |
| "eval_samples_per_second": 42.986, |
| "eval_steps_per_second": 0.688, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.8481675392670157, |
| "grad_norm": 0.014079044240300665, |
| "learning_rate": 4.509209163472357e-05, |
| "logits/chosen": -1.0, |
| "logits/rejected": -1.0, |
| "logps/chosen": -51712.0, |
| "logps/rejected": -51712.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -508.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -508.0, |
| "step": 810 |
| }, |
| { |
| "epoch": 0.8586387434554974, |
| "grad_norm": 0.014646689770739395, |
| "learning_rate": 4.4909317814811896e-05, |
| "logits/chosen": -0.181640625, |
| "logits/rejected": -0.181640625, |
| "logps/chosen": -48128.0, |
| "logps/rejected": -48128.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -470.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -470.0, |
| "step": 820 |
| }, |
| { |
| "epoch": 0.8691099476439791, |
| "grad_norm": 0.012762084284474472, |
| "learning_rate": 4.472358744696845e-05, |
| "logits/chosen": -0.73828125, |
| "logits/rejected": -0.73828125, |
| "logps/chosen": -48128.0, |
| "logps/rejected": -48128.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -472.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -472.0, |
| "step": 830 |
| }, |
| { |
| "epoch": 0.8795811518324608, |
| "grad_norm": 0.014598321761178295, |
| "learning_rate": 4.4534928112285645e-05, |
| "logits/chosen": -0.859375, |
| "logits/rejected": -0.859375, |
| "logps/chosen": -52224.0, |
| "logps/rejected": -52224.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -512.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -512.0, |
| "step": 840 |
| }, |
| { |
| "epoch": 0.8900523560209425, |
| "grad_norm": 0.015155652246863886, |
| "learning_rate": 4.43433678268096e-05, |
| "logits/chosen": -0.640625, |
| "logits/rejected": -0.640625, |
| "logps/chosen": -52224.0, |
| "logps/rejected": -52224.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -512.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -512.0, |
| "step": 850 |
| }, |
| { |
| "epoch": 0.900523560209424, |
| "grad_norm": 0.014456743198814041, |
| "learning_rate": 4.414893503737968e-05, |
| "logits/chosen": -0.83203125, |
| "logits/rejected": -0.83203125, |
| "logps/chosen": -53504.0, |
| "logps/rejected": -53504.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -528.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -528.0, |
| "step": 860 |
| }, |
| { |
| "epoch": 0.9109947643979057, |
| "grad_norm": 0.013877388221773024, |
| "learning_rate": 4.395165861740414e-05, |
| "logits/chosen": -0.58203125, |
| "logits/rejected": -0.58203125, |
| "logps/chosen": -51456.0, |
| "logps/rejected": -51456.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -506.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -506.0, |
| "step": 870 |
| }, |
| { |
| "epoch": 0.9214659685863874, |
| "grad_norm": 0.015292529246672616, |
| "learning_rate": 4.375156786257241e-05, |
| "logits/chosen": -0.14453125, |
| "logits/rejected": -0.14453125, |
| "logps/chosen": -50944.0, |
| "logps/rejected": -50944.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -500.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -500.0, |
| "step": 880 |
| }, |
| { |
| "epoch": 0.9319371727748691, |
| "grad_norm": 0.014402166314922799, |
| "learning_rate": 4.354869248650466e-05, |
| "logits/chosen": -0.296875, |
| "logits/rejected": -0.296875, |
| "logps/chosen": -51968.0, |
| "logps/rejected": -51968.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -512.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -512.0, |
| "step": 890 |
| }, |
| { |
| "epoch": 0.9424083769633508, |
| "grad_norm": 0.015927841762181705, |
| "learning_rate": 4.334306261633931e-05, |
| "logits/chosen": -0.203125, |
| "logits/rejected": -0.203125, |
| "logps/chosen": -47616.0, |
| "logps/rejected": -47616.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -466.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -466.0, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.9424083769633508, |
| "eval_logits/chosen": 1.5390625, |
| "eval_logits/rejected": 1.90625, |
| "eval_logps/chosen": -41216.0, |
| "eval_logps/rejected": -35584.0, |
| "eval_loss": 120.7802505493164, |
| "eval_rewards/accuracies": 0.431640625, |
| "eval_rewards/chosen": -410.0, |
| "eval_rewards/margins": -55.75, |
| "eval_rewards/rejected": -354.0, |
| "eval_runtime": 46.5355, |
| "eval_samples_per_second": 42.978, |
| "eval_steps_per_second": 0.688, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.9528795811518325, |
| "grad_norm": 0.014593852011836841, |
| "learning_rate": 4.313470878825911e-05, |
| "logits/chosen": -0.5078125, |
| "logits/rejected": -0.5078125, |
| "logps/chosen": -48896.0, |
| "logps/rejected": -48896.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -480.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -480.0, |
| "step": 910 |
| }, |
| { |
| "epoch": 0.9633507853403142, |
| "grad_norm": 0.015161307568765613, |
| "learning_rate": 4.292366194295652e-05, |
| "logits/chosen": -0.5, |
| "logits/rejected": -0.5, |
| "logps/chosen": -51456.0, |
| "logps/rejected": -51456.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -504.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -504.0, |
| "step": 920 |
| }, |
| { |
| "epoch": 0.9738219895287958, |
| "grad_norm": 0.015949329901465116, |
| "learning_rate": 4.2709953421038964e-05, |
| "logits/chosen": -0.30859375, |
| "logits/rejected": -0.30859375, |
| "logps/chosen": -49920.0, |
| "logps/rejected": -49920.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -492.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -492.0, |
| "step": 930 |
| }, |
| { |
| "epoch": 0.9842931937172775, |
| "grad_norm": 0.013921757200701724, |
| "learning_rate": 4.249361495837478e-05, |
| "logits/chosen": 0.1982421875, |
| "logits/rejected": 0.1982421875, |
| "logps/chosen": -51968.0, |
| "logps/rejected": -51968.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -510.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -510.0, |
| "step": 940 |
| }, |
| { |
| "epoch": 0.9947643979057592, |
| "grad_norm": 0.020727223992253452, |
| "learning_rate": 4.227467868138035e-05, |
| "logits/chosen": -0.34765625, |
| "logits/rejected": -0.34765625, |
| "logps/chosen": -55552.0, |
| "logps/rejected": -55552.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -544.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -544.0, |
| "step": 950 |
| }, |
| { |
| "epoch": 1.0052356020942408, |
| "grad_norm": 0.021451480324750385, |
| "learning_rate": 4.205317710224938e-05, |
| "logits/chosen": 0.703125, |
| "logits/rejected": 0.703125, |
| "logps/chosen": -51968.0, |
| "logps/rejected": -51968.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -510.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -510.0, |
| "step": 960 |
| }, |
| { |
| "epoch": 1.0157068062827226, |
| "grad_norm": 0.016936544125958134, |
| "learning_rate": 4.1829143114124733e-05, |
| "logits/chosen": -0.12890625, |
| "logits/rejected": -0.12890625, |
| "logps/chosen": -54272.0, |
| "logps/rejected": -54272.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -532.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -532.0, |
| "step": 970 |
| }, |
| { |
| "epoch": 1.0261780104712042, |
| "grad_norm": 0.016165434028480866, |
| "learning_rate": 4.160260998621387e-05, |
| "logits/chosen": 0.267578125, |
| "logits/rejected": 0.267578125, |
| "logps/chosen": -53760.0, |
| "logps/rejected": -53760.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -528.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -528.0, |
| "step": 980 |
| }, |
| { |
| "epoch": 1.036649214659686, |
| "grad_norm": 0.014885625195633459, |
| "learning_rate": 4.137361135884824e-05, |
| "logits/chosen": 1.0703125, |
| "logits/rejected": 1.0703125, |
| "logps/chosen": -57344.0, |
| "logps/rejected": -57344.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -564.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -564.0, |
| "step": 990 |
| }, |
| { |
| "epoch": 1.0471204188481675, |
| "grad_norm": 0.013866038165299034, |
| "learning_rate": 4.114218123848777e-05, |
| "logits/chosen": 1.1875, |
| "logits/rejected": 1.1875, |
| "logps/chosen": -52224.0, |
| "logps/rejected": -52224.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -512.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -512.0, |
| "step": 1000 |
| }, |
| { |
| "epoch": 1.0471204188481675, |
| "eval_logits/chosen": 3.90625, |
| "eval_logits/rejected": 4.46875, |
| "eval_logps/chosen": -42752.0, |
| "eval_logps/rejected": -36864.0, |
| "eval_loss": 125.14350128173828, |
| "eval_rewards/accuracies": 0.435546875, |
| "eval_rewards/chosen": -424.0, |
| "eval_rewards/margins": -57.75, |
| "eval_rewards/rejected": -366.0, |
| "eval_runtime": 46.5438, |
| "eval_samples_per_second": 42.97, |
| "eval_steps_per_second": 0.688, |
| "step": 1000 |
| }, |
| { |
| "epoch": 1.057591623036649, |
| "grad_norm": 0.015437151774848444, |
| "learning_rate": 4.0908353992670824e-05, |
| "logits/chosen": 1.7265625, |
| "logits/rejected": 1.7265625, |
| "logps/chosen": -49664.0, |
| "logps/rejected": -49664.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -490.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -490.0, |
| "step": 1010 |
| }, |
| { |
| "epoch": 1.068062827225131, |
| "grad_norm": 0.014917753305771652, |
| "learning_rate": 4.067216434491059e-05, |
| "logits/chosen": 1.09375, |
| "logits/rejected": 1.09375, |
| "logps/chosen": -53760.0, |
| "logps/rejected": -53760.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -532.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -532.0, |
| "step": 1020 |
| }, |
| { |
| "epoch": 1.0785340314136125, |
| "grad_norm": 0.015254703189521537, |
| "learning_rate": 4.043364736953866e-05, |
| "logits/chosen": 0.9140625, |
| "logits/rejected": 0.9140625, |
| "logps/chosen": -52224.0, |
| "logps/rejected": -52224.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -516.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -516.0, |
| "step": 1030 |
| }, |
| { |
| "epoch": 1.0890052356020943, |
| "grad_norm": 0.01281163827033743, |
| "learning_rate": 4.019283848649643e-05, |
| "logits/chosen": 1.5234375, |
| "logits/rejected": 1.5234375, |
| "logps/chosen": -55040.0, |
| "logps/rejected": -55040.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -540.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -540.0, |
| "step": 1040 |
| }, |
| { |
| "epoch": 1.0994764397905759, |
| "grad_norm": 0.016356853991769983, |
| "learning_rate": 3.9949773456075215e-05, |
| "logits/chosen": 0.1337890625, |
| "logits/rejected": 0.1337890625, |
| "logps/chosen": -54272.0, |
| "logps/rejected": -54272.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -532.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -532.0, |
| "step": 1050 |
| }, |
| { |
| "epoch": 1.1099476439790577, |
| "grad_norm": 0.01592778301953466, |
| "learning_rate": 3.970448837360585e-05, |
| "logits/chosen": 1.40625, |
| "logits/rejected": 1.40625, |
| "logps/chosen": -53760.0, |
| "logps/rejected": -53760.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -528.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -528.0, |
| "step": 1060 |
| }, |
| { |
| "epoch": 1.1204188481675392, |
| "grad_norm": 0.014628017399864208, |
| "learning_rate": 3.945701966409846e-05, |
| "logits/chosen": 0.8125, |
| "logits/rejected": 0.8125, |
| "logps/chosen": -56064.0, |
| "logps/rejected": -56064.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -552.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -552.0, |
| "step": 1070 |
| }, |
| { |
| "epoch": 1.130890052356021, |
| "grad_norm": 0.013856549521764084, |
| "learning_rate": 3.9207404076833374e-05, |
| "logits/chosen": 1.84375, |
| "logits/rejected": 1.84375, |
| "logps/chosen": -48384.0, |
| "logps/rejected": -48384.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -476.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -476.0, |
| "step": 1080 |
| }, |
| { |
| "epoch": 1.1413612565445026, |
| "grad_norm": 0.01635887552342099, |
| "learning_rate": 3.895567867990379e-05, |
| "logits/chosen": 1.2734375, |
| "logits/rejected": 1.2734375, |
| "logps/chosen": -55040.0, |
| "logps/rejected": -55040.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -544.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -544.0, |
| "step": 1090 |
| }, |
| { |
| "epoch": 1.1518324607329844, |
| "grad_norm": 0.016150528406885033, |
| "learning_rate": 3.870188085471114e-05, |
| "logits/chosen": 1.65625, |
| "logits/rejected": 1.65625, |
| "logps/chosen": -54016.0, |
| "logps/rejected": -54016.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -532.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -532.0, |
| "step": 1100 |
| }, |
| { |
| "epoch": 1.1518324607329844, |
| "eval_logits/chosen": 3.515625, |
| "eval_logits/rejected": 4.03125, |
| "eval_logps/chosen": -44288.0, |
| "eval_logps/rejected": -38144.0, |
| "eval_loss": 129.38262939453125, |
| "eval_rewards/accuracies": 0.431640625, |
| "eval_rewards/chosen": -440.0, |
| "eval_rewards/margins": -59.75, |
| "eval_rewards/rejected": -380.0, |
| "eval_runtime": 46.546, |
| "eval_samples_per_second": 42.968, |
| "eval_steps_per_second": 0.687, |
| "step": 1100 |
| }, |
| { |
| "epoch": 1.162303664921466, |
| "grad_norm": 0.017998489212121434, |
| "learning_rate": 3.844604829041395e-05, |
| "logits/chosen": 1.21875, |
| "logits/rejected": 1.21875, |
| "logps/chosen": -59136.0, |
| "logps/rejected": -59136.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -584.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -584.0, |
| "step": 1110 |
| }, |
| { |
| "epoch": 1.1727748691099475, |
| "grad_norm": 0.019158510393387902, |
| "learning_rate": 3.818821897833095e-05, |
| "logits/chosen": 0.90234375, |
| "logits/rejected": 0.90234375, |
| "logps/chosen": -53248.0, |
| "logps/rejected": -53248.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -524.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -524.0, |
| "step": 1120 |
| }, |
| { |
| "epoch": 1.1832460732984293, |
| "grad_norm": 0.018058866561367026, |
| "learning_rate": 3.792843120629935e-05, |
| "logits/chosen": 1.6875, |
| "logits/rejected": 1.6875, |
| "logps/chosen": -47872.0, |
| "logps/rejected": -47872.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -470.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -470.0, |
| "step": 1130 |
| }, |
| { |
| "epoch": 1.193717277486911, |
| "grad_norm": 0.017687334559732766, |
| "learning_rate": 3.766672355298907e-05, |
| "logits/chosen": 1.5859375, |
| "logits/rejected": 1.5859375, |
| "logps/chosen": -54016.0, |
| "logps/rejected": -54016.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -532.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -532.0, |
| "step": 1140 |
| }, |
| { |
| "epoch": 1.2041884816753927, |
| "grad_norm": 0.021142007309678242, |
| "learning_rate": 3.740313488217373e-05, |
| "logits/chosen": 1.484375, |
| "logits/rejected": 1.484375, |
| "logps/chosen": -56576.0, |
| "logps/rejected": -56576.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -556.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -556.0, |
| "step": 1150 |
| }, |
| { |
| "epoch": 1.2146596858638743, |
| "grad_norm": 0.02065509696523931, |
| "learning_rate": 3.7137704336959464e-05, |
| "logits/chosen": 0.984375, |
| "logits/rejected": 0.984375, |
| "logps/chosen": -61440.0, |
| "logps/rejected": -61440.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -604.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -604.0, |
| "step": 1160 |
| }, |
| { |
| "epoch": 1.225130890052356, |
| "grad_norm": 0.023371613683585618, |
| "learning_rate": 3.687047133397201e-05, |
| "logits/chosen": 1.1328125, |
| "logits/rejected": 1.1328125, |
| "logps/chosen": -54272.0, |
| "logps/rejected": -54272.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -532.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -532.0, |
| "step": 1170 |
| }, |
| { |
| "epoch": 1.2356020942408377, |
| "grad_norm": 0.0405741898218784, |
| "learning_rate": 3.6601475557503415e-05, |
| "logits/chosen": 1.5546875, |
| "logits/rejected": 1.5546875, |
| "logps/chosen": -54784.0, |
| "logps/rejected": -54784.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -540.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -540.0, |
| "step": 1180 |
| }, |
| { |
| "epoch": 1.2460732984293195, |
| "grad_norm": 0.03691730960465931, |
| "learning_rate": 3.633075695361882e-05, |
| "logits/chosen": 0.59765625, |
| "logits/rejected": 0.59765625, |
| "logps/chosen": -57600.0, |
| "logps/rejected": -57600.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -568.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -568.0, |
| "step": 1190 |
| }, |
| { |
| "epoch": 1.256544502617801, |
| "grad_norm": 0.0400680254099473, |
| "learning_rate": 3.6058355724224477e-05, |
| "logits/chosen": 0.9453125, |
| "logits/rejected": 0.9453125, |
| "logps/chosen": -54016.0, |
| "logps/rejected": -54016.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -532.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -532.0, |
| "step": 1200 |
| }, |
| { |
| "epoch": 1.256544502617801, |
| "eval_logits/chosen": 3.21875, |
| "eval_logits/rejected": 3.84375, |
| "eval_logps/chosen": -45568.0, |
| "eval_logps/rejected": -39424.0, |
| "eval_loss": 133.65574645996094, |
| "eval_rewards/accuracies": 0.4296875, |
| "eval_rewards/chosen": -454.0, |
| "eval_rewards/margins": -62.0, |
| "eval_rewards/rejected": -392.0, |
| "eval_runtime": 46.5518, |
| "eval_samples_per_second": 42.963, |
| "eval_steps_per_second": 0.687, |
| "step": 1200 |
| }, |
| { |
| "epoch": 1.2670157068062826, |
| "grad_norm": 0.040783582420381936, |
| "learning_rate": 3.5784312321097744e-05, |
| "logits/chosen": 0.7265625, |
| "logits/rejected": 0.7265625, |
| "logps/chosen": -59648.0, |
| "logps/rejected": -59648.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -588.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -588.0, |
| "step": 1210 |
| }, |
| { |
| "epoch": 1.2774869109947644, |
| "grad_norm": 0.0447749159302651, |
| "learning_rate": 3.550866743987994e-05, |
| "logits/chosen": 0.90234375, |
| "logits/rejected": 0.90234375, |
| "logps/chosen": -59648.0, |
| "logps/rejected": -59648.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -588.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -588.0, |
| "step": 1220 |
| }, |
| { |
| "epoch": 1.2879581151832462, |
| "grad_norm": 0.0509025659083953, |
| "learning_rate": 3.5231462014033024e-05, |
| "logits/chosen": 0.169921875, |
| "logits/rejected": 0.169921875, |
| "logps/chosen": -54272.0, |
| "logps/rejected": -54272.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -536.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -536.0, |
| "step": 1230 |
| }, |
| { |
| "epoch": 1.2984293193717278, |
| "grad_norm": 0.04279423627870876, |
| "learning_rate": 3.495273720876095e-05, |
| "logits/chosen": -0.44140625, |
| "logits/rejected": -0.44140625, |
| "logps/chosen": -55552.0, |
| "logps/rejected": -55552.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -544.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -544.0, |
| "step": 1240 |
| }, |
| { |
| "epoch": 1.3089005235602094, |
| "grad_norm": 0.03706980219765163, |
| "learning_rate": 3.4672534414896616e-05, |
| "logits/chosen": -1.2890625, |
| "logits/rejected": -1.2890625, |
| "logps/chosen": -55808.0, |
| "logps/rejected": -55808.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -552.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -552.0, |
| "step": 1250 |
| }, |
| { |
| "epoch": 1.3193717277486912, |
| "grad_norm": 0.0420537974190497, |
| "learning_rate": 3.439089524275526e-05, |
| "logits/chosen": -0.451171875, |
| "logits/rejected": -0.451171875, |
| "logps/chosen": -58368.0, |
| "logps/rejected": -58368.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -572.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -572.0, |
| "step": 1260 |
| }, |
| { |
| "epoch": 1.3298429319371727, |
| "grad_norm": 0.036923355176261774, |
| "learning_rate": 3.410786151595533e-05, |
| "logits/chosen": 0.83984375, |
| "logits/rejected": 0.83984375, |
| "logps/chosen": -57856.0, |
| "logps/rejected": -57856.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -568.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -568.0, |
| "step": 1270 |
| }, |
| { |
| "epoch": 1.3403141361256545, |
| "grad_norm": 0.04890843728084653, |
| "learning_rate": 3.382347526520765e-05, |
| "logits/chosen": -0.4296875, |
| "logits/rejected": -0.4296875, |
| "logps/chosen": -60160.0, |
| "logps/rejected": -60160.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -592.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -592.0, |
| "step": 1280 |
| }, |
| { |
| "epoch": 1.350785340314136, |
| "grad_norm": 0.034260261289086394, |
| "learning_rate": 3.353777872207381e-05, |
| "logits/chosen": -0.3046875, |
| "logits/rejected": -0.3046875, |
| "logps/chosen": -62208.0, |
| "logps/rejected": -62208.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -612.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -612.0, |
| "step": 1290 |
| }, |
| { |
| "epoch": 1.3612565445026177, |
| "grad_norm": 0.03030933596622796, |
| "learning_rate": 3.325081431269475e-05, |
| "logits/chosen": -0.953125, |
| "logits/rejected": -0.953125, |
| "logps/chosen": -54016.0, |
| "logps/rejected": -54016.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -532.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -532.0, |
| "step": 1300 |
| }, |
| { |
| "epoch": 1.3612565445026177, |
| "eval_logits/chosen": 1.7890625, |
| "eval_logits/rejected": 2.109375, |
| "eval_logps/chosen": -46848.0, |
| "eval_logps/rejected": -40704.0, |
| "eval_loss": 137.50975036621094, |
| "eval_rewards/accuracies": 0.435546875, |
| "eval_rewards/chosen": -466.0, |
| "eval_rewards/margins": -63.5, |
| "eval_rewards/rejected": -404.0, |
| "eval_runtime": 46.5437, |
| "eval_samples_per_second": 42.97, |
| "eval_steps_per_second": 0.688, |
| "step": 1300 |
| }, |
| { |
| "epoch": 1.3717277486910995, |
| "grad_norm": 0.03227090536607956, |
| "learning_rate": 3.296262465149046e-05, |
| "logits/chosen": 0.51953125, |
| "logits/rejected": 0.51953125, |
| "logps/chosen": -51200.0, |
| "logps/rejected": -51200.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -504.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -504.0, |
| "step": 1310 |
| }, |
| { |
| "epoch": 1.3821989528795813, |
| "grad_norm": 0.024326298762496715, |
| "learning_rate": 3.2673252534831687e-05, |
| "logits/chosen": 0.5, |
| "logits/rejected": 0.5, |
| "logps/chosen": -60928.0, |
| "logps/rejected": -60928.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -600.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -600.0, |
| "step": 1320 |
| }, |
| { |
| "epoch": 1.3926701570680629, |
| "grad_norm": 0.02258510606016088, |
| "learning_rate": 3.2382740934684696e-05, |
| "logits/chosen": -0.291015625, |
| "logits/rejected": -0.291015625, |
| "logps/chosen": -58112.0, |
| "logps/rejected": -58112.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -572.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -572.0, |
| "step": 1330 |
| }, |
| { |
| "epoch": 1.4031413612565444, |
| "grad_norm": 0.023860546159110697, |
| "learning_rate": 3.209113299222982e-05, |
| "logits/chosen": -0.57421875, |
| "logits/rejected": -0.57421875, |
| "logps/chosen": -64000.0, |
| "logps/rejected": -64000.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -628.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -628.0, |
| "step": 1340 |
| }, |
| { |
| "epoch": 1.4136125654450262, |
| "grad_norm": 0.020840386441878106, |
| "learning_rate": 3.179847201145505e-05, |
| "logits/chosen": -0.028564453125, |
| "logits/rejected": -0.028564453125, |
| "logps/chosen": -57344.0, |
| "logps/rejected": -57344.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -568.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -568.0, |
| "step": 1350 |
| }, |
| { |
| "epoch": 1.4240837696335078, |
| "grad_norm": 0.021443057955016713, |
| "learning_rate": 3.150480145272528e-05, |
| "logits/chosen": 0.287109375, |
| "logits/rejected": 0.287109375, |
| "logps/chosen": -60928.0, |
| "logps/rejected": -60928.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -600.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -600.0, |
| "step": 1360 |
| }, |
| { |
| "epoch": 1.4345549738219896, |
| "grad_norm": 0.018891332406255266, |
| "learning_rate": 3.121016492632848e-05, |
| "logits/chosen": 0.328125, |
| "logits/rejected": 0.328125, |
| "logps/chosen": -62208.0, |
| "logps/rejected": -62208.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -612.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -612.0, |
| "step": 1370 |
| }, |
| { |
| "epoch": 1.4450261780104712, |
| "grad_norm": 0.015381554107551168, |
| "learning_rate": 3.091460618599951e-05, |
| "logits/chosen": 1.25, |
| "logits/rejected": 1.25, |
| "logps/chosen": -56064.0, |
| "logps/rejected": -56064.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -552.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -552.0, |
| "step": 1380 |
| }, |
| { |
| "epoch": 1.4554973821989527, |
| "grad_norm": 0.016181321714692026, |
| "learning_rate": 3.061816912242265e-05, |
| "logits/chosen": 0.63671875, |
| "logits/rejected": 0.63671875, |
| "logps/chosen": -59904.0, |
| "logps/rejected": -59904.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -588.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -588.0, |
| "step": 1390 |
| }, |
| { |
| "epoch": 1.4659685863874345, |
| "grad_norm": 0.02284512280333461, |
| "learning_rate": 3.0320897756713785e-05, |
| "logits/chosen": 0.57421875, |
| "logits/rejected": 0.57421875, |
| "logps/chosen": -64512.0, |
| "logps/rejected": -64512.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -636.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -636.0, |
| "step": 1400 |
| }, |
| { |
| "epoch": 1.4659685863874345, |
| "eval_logits/chosen": 2.765625, |
| "eval_logits/rejected": 3.234375, |
| "eval_logps/chosen": -48384.0, |
| "eval_logps/rejected": -41728.0, |
| "eval_loss": 141.6271209716797, |
| "eval_rewards/accuracies": 0.435546875, |
| "eval_rewards/chosen": -482.0, |
| "eval_rewards/margins": -65.5, |
| "eval_rewards/rejected": -416.0, |
| "eval_runtime": 46.5342, |
| "eval_samples_per_second": 42.979, |
| "eval_steps_per_second": 0.688, |
| "step": 1400 |
| }, |
| { |
| "epoch": 1.4764397905759163, |
| "grad_norm": 0.019967694319150568, |
| "learning_rate": 3.002283623388332e-05, |
| "logits/chosen": 0.90234375, |
| "logits/rejected": 0.90234375, |
| "logps/chosen": -57856.0, |
| "logps/rejected": -57856.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -568.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -568.0, |
| "step": 1410 |
| }, |
| { |
| "epoch": 1.486910994764398, |
| "grad_norm": 0.018360624862304626, |
| "learning_rate": 2.9724028816280507e-05, |
| "logits/chosen": 0.78515625, |
| "logits/rejected": 0.78515625, |
| "logps/chosen": -60672.0, |
| "logps/rejected": -60672.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -600.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -600.0, |
| "step": 1420 |
| }, |
| { |
| "epoch": 1.4973821989528795, |
| "grad_norm": 0.01790010533608198, |
| "learning_rate": 2.9424519877020524e-05, |
| "logits/chosen": 0.8984375, |
| "logits/rejected": 0.8984375, |
| "logps/chosen": -60160.0, |
| "logps/rejected": -60160.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -596.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -596.0, |
| "step": 1430 |
| }, |
| { |
| "epoch": 1.5078534031413613, |
| "grad_norm": 0.018427887906480415, |
| "learning_rate": 2.912435389339504e-05, |
| "logits/chosen": 0.98828125, |
| "logits/rejected": 0.98828125, |
| "logps/chosen": -59904.0, |
| "logps/rejected": -59904.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -588.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -588.0, |
| "step": 1440 |
| }, |
| { |
| "epoch": 1.518324607329843, |
| "grad_norm": 0.021735777025989843, |
| "learning_rate": 2.8823575440267226e-05, |
| "logits/chosen": 0.984375, |
| "logits/rejected": 0.984375, |
| "logps/chosen": -60672.0, |
| "logps/rejected": -60672.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -596.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -596.0, |
| "step": 1450 |
| }, |
| { |
| "epoch": 1.5287958115183247, |
| "grad_norm": 0.02488164983225801, |
| "learning_rate": 2.8522229183452465e-05, |
| "logits/chosen": 1.21875, |
| "logits/rejected": 1.21875, |
| "logps/chosen": -58880.0, |
| "logps/rejected": -58880.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -580.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -580.0, |
| "step": 1460 |
| }, |
| { |
| "epoch": 1.5392670157068062, |
| "grad_norm": 0.027814203660503543, |
| "learning_rate": 2.8220359873085373e-05, |
| "logits/chosen": 0.734375, |
| "logits/rejected": 0.734375, |
| "logps/chosen": -62720.0, |
| "logps/rejected": -62720.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -620.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -620.0, |
| "step": 1470 |
| }, |
| { |
| "epoch": 1.5497382198952878, |
| "grad_norm": 0.02678922835752938, |
| "learning_rate": 2.7918012336974388e-05, |
| "logits/chosen": 0.27734375, |
| "logits/rejected": 0.27734375, |
| "logps/chosen": -60928.0, |
| "logps/rejected": -60928.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -600.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -600.0, |
| "step": 1480 |
| }, |
| { |
| "epoch": 1.5602094240837696, |
| "grad_norm": 0.04002268051059004, |
| "learning_rate": 2.7615231473944815e-05, |
| "logits/chosen": 0.6484375, |
| "logits/rejected": 0.6484375, |
| "logps/chosen": -63488.0, |
| "logps/rejected": -63488.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -624.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -624.0, |
| "step": 1490 |
| }, |
| { |
| "epoch": 1.5706806282722514, |
| "grad_norm": 0.04956642035668694, |
| "learning_rate": 2.7312062247171327e-05, |
| "logits/chosen": 1.25, |
| "logits/rejected": 1.25, |
| "logps/chosen": -57600.0, |
| "logps/rejected": -57600.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -568.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -568.0, |
| "step": 1500 |
| }, |
| { |
| "epoch": 1.5706806282722514, |
| "eval_logits/chosen": 2.984375, |
| "eval_logits/rejected": 3.484375, |
| "eval_logps/chosen": -49664.0, |
| "eval_logps/rejected": -42752.0, |
| "eval_loss": 145.06924438476562, |
| "eval_rewards/accuracies": 0.43359375, |
| "eval_rewards/chosen": -492.0, |
| "eval_rewards/margins": -67.0, |
| "eval_rewards/rejected": -426.0, |
| "eval_runtime": 46.555, |
| "eval_samples_per_second": 42.96, |
| "eval_steps_per_second": 0.687, |
| "step": 1500 |
| }, |
| { |
| "epoch": 1.581151832460733, |
| "grad_norm": 0.029249030833769278, |
| "learning_rate": 2.7008549677500876e-05, |
| "logits/chosen": 1.1796875, |
| "logits/rejected": 1.1796875, |
| "logps/chosen": -61696.0, |
| "logps/rejected": -61696.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -604.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -604.0, |
| "step": 1510 |
| }, |
| { |
| "epoch": 1.5916230366492146, |
| "grad_norm": 0.03795737298660237, |
| "learning_rate": 2.670473883676709e-05, |
| "logits/chosen": 1.7265625, |
| "logits/rejected": 1.7265625, |
| "logps/chosen": -60928.0, |
| "logps/rejected": -60928.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -600.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -600.0, |
| "step": 1520 |
| }, |
| { |
| "epoch": 1.6020942408376964, |
| "grad_norm": 0.03528147973238564, |
| "learning_rate": 2.6400674841097074e-05, |
| "logits/chosen": 2.03125, |
| "logits/rejected": 2.03125, |
| "logps/chosen": -60416.0, |
| "logps/rejected": -60416.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -596.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -596.0, |
| "step": 1530 |
| }, |
| { |
| "epoch": 1.6125654450261782, |
| "grad_norm": 0.030932850850396734, |
| "learning_rate": 2.609640284421158e-05, |
| "logits/chosen": 2.171875, |
| "logits/rejected": 2.171875, |
| "logps/chosen": -57856.0, |
| "logps/rejected": -57856.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -572.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -572.0, |
| "step": 1540 |
| }, |
| { |
| "epoch": 1.6230366492146597, |
| "grad_norm": 0.020386201940892617, |
| "learning_rate": 2.5791968030719726e-05, |
| "logits/chosen": 1.125, |
| "logits/rejected": 1.125, |
| "logps/chosen": -58880.0, |
| "logps/rejected": -58880.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -580.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -580.0, |
| "step": 1550 |
| }, |
| { |
| "epoch": 1.6335078534031413, |
| "grad_norm": 0.02167021387525914, |
| "learning_rate": 2.548741560940902e-05, |
| "logits/chosen": 0.88671875, |
| "logits/rejected": 0.88671875, |
| "logps/chosen": -64768.0, |
| "logps/rejected": -64768.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -636.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -636.0, |
| "step": 1560 |
| }, |
| { |
| "epoch": 1.6439790575916229, |
| "grad_norm": 0.026240354628493872, |
| "learning_rate": 2.5182790806531787e-05, |
| "logits/chosen": 0.466796875, |
| "logits/rejected": 0.466796875, |
| "logps/chosen": -69120.0, |
| "logps/rejected": -69120.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -684.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -684.0, |
| "step": 1570 |
| }, |
| { |
| "epoch": 1.6544502617801047, |
| "grad_norm": 0.027697972817473052, |
| "learning_rate": 2.4878138859089074e-05, |
| "logits/chosen": 0.57421875, |
| "logits/rejected": 0.57421875, |
| "logps/chosen": -63232.0, |
| "logps/rejected": -63232.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -620.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -620.0, |
| "step": 1580 |
| }, |
| { |
| "epoch": 1.6649214659685865, |
| "grad_norm": 0.03104999859066827, |
| "learning_rate": 2.4573505008112924e-05, |
| "logits/chosen": 0.92578125, |
| "logits/rejected": 0.92578125, |
| "logps/chosen": -60672.0, |
| "logps/rejected": -60672.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -596.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -596.0, |
| "step": 1590 |
| }, |
| { |
| "epoch": 1.675392670157068, |
| "grad_norm": 0.03222787886163218, |
| "learning_rate": 2.4268934491948032e-05, |
| "logits/chosen": 0.66796875, |
| "logits/rejected": 0.66796875, |
| "logps/chosen": -61696.0, |
| "logps/rejected": -61696.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -608.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -608.0, |
| "step": 1600 |
| }, |
| { |
| "epoch": 1.675392670157068, |
| "eval_logits/chosen": 2.859375, |
| "eval_logits/rejected": 3.328125, |
| "eval_logps/chosen": -50688.0, |
| "eval_logps/rejected": -43776.0, |
| "eval_loss": 148.48387145996094, |
| "eval_rewards/accuracies": 0.4296875, |
| "eval_rewards/chosen": -504.0, |
| "eval_rewards/margins": -68.5, |
| "eval_rewards/rejected": -436.0, |
| "eval_runtime": 46.5434, |
| "eval_samples_per_second": 42.971, |
| "eval_steps_per_second": 0.688, |
| "step": 1600 |
| }, |
| { |
| "epoch": 1.6858638743455496, |
| "grad_norm": 0.027933418479643906, |
| "learning_rate": 2.396447253953385e-05, |
| "logits/chosen": 0.54296875, |
| "logits/rejected": 0.54296875, |
| "logps/chosen": -64512.0, |
| "logps/rejected": -64512.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -632.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -632.0, |
| "step": 1610 |
| }, |
| { |
| "epoch": 1.6963350785340314, |
| "grad_norm": 0.02858427756552721, |
| "learning_rate": 2.3660164363687998e-05, |
| "logits/chosen": 0.8046875, |
| "logits/rejected": 0.8046875, |
| "logps/chosen": -68096.0, |
| "logps/rejected": -68096.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -668.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -668.0, |
| "step": 1620 |
| }, |
| { |
| "epoch": 1.7068062827225132, |
| "grad_norm": 0.02087251231328732, |
| "learning_rate": 2.33560551543922e-05, |
| "logits/chosen": 1.4375, |
| "logits/rejected": 1.4375, |
| "logps/chosen": -62976.0, |
| "logps/rejected": -62976.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -620.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -620.0, |
| "step": 1630 |
| }, |
| { |
| "epoch": 1.7172774869109948, |
| "grad_norm": 0.023700722854211376, |
| "learning_rate": 2.3052190072081493e-05, |
| "logits/chosen": 1.15625, |
| "logits/rejected": 1.15625, |
| "logps/chosen": -63488.0, |
| "logps/rejected": -63488.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -624.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -624.0, |
| "step": 1640 |
| }, |
| { |
| "epoch": 1.7277486910994764, |
| "grad_norm": 0.02528105329254383, |
| "learning_rate": 2.2748614240937864e-05, |
| "logits/chosen": 1.484375, |
| "logits/rejected": 1.484375, |
| "logps/chosen": -62464.0, |
| "logps/rejected": -62464.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -616.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -616.0, |
| "step": 1650 |
| }, |
| { |
| "epoch": 1.738219895287958, |
| "grad_norm": 0.027398508316340463, |
| "learning_rate": 2.2445372742189334e-05, |
| "logits/chosen": 2.109375, |
| "logits/rejected": 2.109375, |
| "logps/chosen": -62208.0, |
| "logps/rejected": -62208.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -616.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -616.0, |
| "step": 1660 |
| }, |
| { |
| "epoch": 1.7486910994764397, |
| "grad_norm": 0.025862979301967695, |
| "learning_rate": 2.214251060741528e-05, |
| "logits/chosen": 1.28125, |
| "logits/rejected": 1.28125, |
| "logps/chosen": -61952.0, |
| "logps/rejected": -61952.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -612.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -612.0, |
| "step": 1670 |
| }, |
| { |
| "epoch": 1.7591623036649215, |
| "grad_norm": 0.0295944666348626, |
| "learning_rate": 2.1840072811859302e-05, |
| "logits/chosen": 1.1640625, |
| "logits/rejected": 1.1640625, |
| "logps/chosen": -67072.0, |
| "logps/rejected": -67072.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -664.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -664.0, |
| "step": 1680 |
| }, |
| { |
| "epoch": 1.7696335078534031, |
| "grad_norm": 0.02837634990956455, |
| "learning_rate": 2.1538104267750287e-05, |
| "logits/chosen": 1.6640625, |
| "logits/rejected": 1.6640625, |
| "logps/chosen": -64256.0, |
| "logps/rejected": -64256.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -632.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -632.0, |
| "step": 1690 |
| }, |
| { |
| "epoch": 1.7801047120418847, |
| "grad_norm": 0.037477698905215684, |
| "learning_rate": 2.1236649817632952e-05, |
| "logits/chosen": 1.3671875, |
| "logits/rejected": 1.3671875, |
| "logps/chosen": -63744.0, |
| "logps/rejected": -63744.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -628.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -628.0, |
| "step": 1700 |
| }, |
| { |
| "epoch": 1.7801047120418847, |
| "eval_logits/chosen": 3.21875, |
| "eval_logits/rejected": 3.703125, |
| "eval_logps/chosen": -51712.0, |
| "eval_logps/rejected": -44544.0, |
| "eval_loss": 151.19650268554688, |
| "eval_rewards/accuracies": 0.431640625, |
| "eval_rewards/chosen": -512.0, |
| "eval_rewards/margins": -69.5, |
| "eval_rewards/rejected": -444.0, |
| "eval_runtime": 46.5645, |
| "eval_samples_per_second": 42.951, |
| "eval_steps_per_second": 0.687, |
| "step": 1700 |
| }, |
| { |
| "epoch": 1.7905759162303665, |
| "grad_norm": 0.04122261373053859, |
| "learning_rate": 2.093575422770872e-05, |
| "logits/chosen": 1.1953125, |
| "logits/rejected": 1.1953125, |
| "logps/chosen": -61696.0, |
| "logps/rejected": -61696.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -608.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -608.0, |
| "step": 1710 |
| }, |
| { |
| "epoch": 1.8010471204188483, |
| "grad_norm": 0.028847606649896974, |
| "learning_rate": 2.0635462181187827e-05, |
| "logits/chosen": 0.89453125, |
| "logits/rejected": 0.89453125, |
| "logps/chosen": -62976.0, |
| "logps/rejected": -62976.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -620.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -620.0, |
| "step": 1720 |
| }, |
| { |
| "epoch": 1.8115183246073299, |
| "grad_norm": 0.02607133556562433, |
| "learning_rate": 2.0335818271653917e-05, |
| "logits/chosen": 0.8203125, |
| "logits/rejected": 0.8203125, |
| "logps/chosen": -66048.0, |
| "logps/rejected": -66048.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -652.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -652.0, |
| "step": 1730 |
| }, |
| { |
| "epoch": 1.8219895287958114, |
| "grad_norm": 0.032895608007411564, |
| "learning_rate": 2.0036866996441815e-05, |
| "logits/chosen": 1.546875, |
| "logits/rejected": 1.546875, |
| "logps/chosen": -63488.0, |
| "logps/rejected": -63488.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -624.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -624.0, |
| "step": 1740 |
| }, |
| { |
| "epoch": 1.8324607329842932, |
| "grad_norm": 0.025720950663877015, |
| "learning_rate": 1.9738652750029677e-05, |
| "logits/chosen": 1.203125, |
| "logits/rejected": 1.203125, |
| "logps/chosen": -64000.0, |
| "logps/rejected": -64000.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -632.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -632.0, |
| "step": 1750 |
| }, |
| { |
| "epoch": 1.8429319371727748, |
| "grad_norm": 0.03757848155792976, |
| "learning_rate": 1.944121981744631e-05, |
| "logits/chosen": 1.2734375, |
| "logits/rejected": 1.2734375, |
| "logps/chosen": -59392.0, |
| "logps/rejected": -59392.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -584.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -584.0, |
| "step": 1760 |
| }, |
| { |
| "epoch": 1.8534031413612566, |
| "grad_norm": 0.02334056857017759, |
| "learning_rate": 1.9144612367694887e-05, |
| "logits/chosen": 1.2265625, |
| "logits/rejected": 1.2265625, |
| "logps/chosen": -60672.0, |
| "logps/rejected": -60672.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -600.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -600.0, |
| "step": 1770 |
| }, |
| { |
| "epoch": 1.8638743455497382, |
| "grad_norm": 0.024611491427847163, |
| "learning_rate": 1.8848874447193803e-05, |
| "logits/chosen": 1.59375, |
| "logits/rejected": 1.59375, |
| "logps/chosen": -64256.0, |
| "logps/rejected": -64256.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -636.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -636.0, |
| "step": 1780 |
| }, |
| { |
| "epoch": 1.8743455497382198, |
| "grad_norm": 0.028047481025915307, |
| "learning_rate": 1.855404997323571e-05, |
| "logits/chosen": 1.6875, |
| "logits/rejected": 1.6875, |
| "logps/chosen": -65024.0, |
| "logps/rejected": -65024.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -640.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -640.0, |
| "step": 1790 |
| }, |
| { |
| "epoch": 1.8848167539267016, |
| "grad_norm": 0.02807485918220803, |
| "learning_rate": 1.82601827274658e-05, |
| "logits/chosen": 1.1015625, |
| "logits/rejected": 1.1015625, |
| "logps/chosen": -68096.0, |
| "logps/rejected": -68096.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -672.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -672.0, |
| "step": 1800 |
| }, |
| { |
| "epoch": 1.8848167539267016, |
| "eval_logits/chosen": 3.6875, |
| "eval_logits/rejected": 4.21875, |
| "eval_logps/chosen": -52736.0, |
| "eval_logps/rejected": -45568.0, |
| "eval_loss": 154.02149963378906, |
| "eval_rewards/accuracies": 0.43359375, |
| "eval_rewards/chosen": -524.0, |
| "eval_rewards/margins": -71.0, |
| "eval_rewards/rejected": -452.0, |
| "eval_runtime": 46.5307, |
| "eval_samples_per_second": 42.982, |
| "eval_steps_per_second": 0.688, |
| "step": 1800 |
| }, |
| { |
| "epoch": 1.8952879581151834, |
| "grad_norm": 0.02787448358261749, |
| "learning_rate": 1.7967316349380213e-05, |
| "logits/chosen": 1.21875, |
| "logits/rejected": 1.21875, |
| "logps/chosen": -69632.0, |
| "logps/rejected": -69632.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -688.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -688.0, |
| "step": 1810 |
| }, |
| { |
| "epoch": 1.905759162303665, |
| "grad_norm": 0.0264474553372622, |
| "learning_rate": 1.7675494329845516e-05, |
| "logits/chosen": 1.4140625, |
| "logits/rejected": 1.4140625, |
| "logps/chosen": -61952.0, |
| "logps/rejected": -61952.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -612.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -612.0, |
| "step": 1820 |
| }, |
| { |
| "epoch": 1.9162303664921465, |
| "grad_norm": 0.024951296381460725, |
| "learning_rate": 1.738476000464026e-05, |
| "logits/chosen": 1.4453125, |
| "logits/rejected": 1.4453125, |
| "logps/chosen": -61696.0, |
| "logps/rejected": -61696.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -608.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -608.0, |
| "step": 1830 |
| }, |
| { |
| "epoch": 1.9267015706806283, |
| "grad_norm": 0.02091953991407963, |
| "learning_rate": 1.7095156548019646e-05, |
| "logits/chosen": 1.390625, |
| "logits/rejected": 1.390625, |
| "logps/chosen": -67584.0, |
| "logps/rejected": -67584.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -664.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -664.0, |
| "step": 1840 |
| }, |
| { |
| "epoch": 1.93717277486911, |
| "grad_norm": 0.019682113536387894, |
| "learning_rate": 1.6806726966304064e-05, |
| "logits/chosen": 1.1953125, |
| "logits/rejected": 1.1953125, |
| "logps/chosen": -66560.0, |
| "logps/rejected": -66560.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -656.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -656.0, |
| "step": 1850 |
| }, |
| { |
| "epoch": 1.9476439790575917, |
| "grad_norm": 0.020216961927251108, |
| "learning_rate": 1.6519514091492625e-05, |
| "logits/chosen": 0.73828125, |
| "logits/rejected": 0.73828125, |
| "logps/chosen": -64000.0, |
| "logps/rejected": -64000.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -628.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -628.0, |
| "step": 1860 |
| }, |
| { |
| "epoch": 1.9581151832460733, |
| "grad_norm": 0.021870907704941635, |
| "learning_rate": 1.6233560574902627e-05, |
| "logits/chosen": 1.1796875, |
| "logits/rejected": 1.1796875, |
| "logps/chosen": -68096.0, |
| "logps/rejected": -68096.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -672.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -672.0, |
| "step": 1870 |
| }, |
| { |
| "epoch": 1.9685863874345548, |
| "grad_norm": 0.01697642758435497, |
| "learning_rate": 1.5948908880835753e-05, |
| "logits/chosen": 1.296875, |
| "logits/rejected": 1.296875, |
| "logps/chosen": -61184.0, |
| "logps/rejected": -61184.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -604.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -604.0, |
| "step": 1880 |
| }, |
| { |
| "epoch": 1.9790575916230366, |
| "grad_norm": 0.03571285404355297, |
| "learning_rate": 1.5665601280272124e-05, |
| "logits/chosen": 1.5703125, |
| "logits/rejected": 1.5703125, |
| "logps/chosen": -65536.0, |
| "logps/rejected": -65536.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -648.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -648.0, |
| "step": 1890 |
| }, |
| { |
| "epoch": 1.9895287958115184, |
| "grad_norm": 0.02817629986002779, |
| "learning_rate": 1.538367984459298e-05, |
| "logits/chosen": 1.3984375, |
| "logits/rejected": 1.3984375, |
| "logps/chosen": -63488.0, |
| "logps/rejected": -63488.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -624.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -624.0, |
| "step": 1900 |
| }, |
| { |
| "epoch": 1.9895287958115184, |
| "eval_logits/chosen": 2.875, |
| "eval_logits/rejected": 3.3125, |
| "eval_logps/chosen": -53504.0, |
| "eval_logps/rejected": -46080.0, |
| "eval_loss": 156.48974609375, |
| "eval_rewards/accuracies": 0.431640625, |
| "eval_rewards/chosen": -532.0, |
| "eval_rewards/margins": -72.5, |
| "eval_rewards/rejected": -460.0, |
| "eval_runtime": 46.5405, |
| "eval_samples_per_second": 42.973, |
| "eval_steps_per_second": 0.688, |
| "step": 1900 |
| }, |
| { |
| "epoch": 2.0, |
| "grad_norm": 0.021601933604289275, |
| "learning_rate": 1.5103186439333133e-05, |
| "logits/chosen": 1.234375, |
| "logits/rejected": 1.234375, |
| "logps/chosen": -66560.0, |
| "logps/rejected": -66560.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -656.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -656.0, |
| "step": 1910 |
| }, |
| { |
| "epoch": 2.0104712041884816, |
| "grad_norm": 0.022698513971145307, |
| "learning_rate": 1.4824162717963829e-05, |
| "logits/chosen": 0.58203125, |
| "logits/rejected": 0.58203125, |
| "logps/chosen": -70656.0, |
| "logps/rejected": -70656.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -700.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -700.0, |
| "step": 1920 |
| }, |
| { |
| "epoch": 2.020942408376963, |
| "grad_norm": 0.021365482780393418, |
| "learning_rate": 1.4546650115707192e-05, |
| "logits/chosen": 0.828125, |
| "logits/rejected": 0.828125, |
| "logps/chosen": -67584.0, |
| "logps/rejected": -67584.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -664.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -664.0, |
| "step": 1930 |
| }, |
| { |
| "epoch": 2.031413612565445, |
| "grad_norm": 0.05466916314665461, |
| "learning_rate": 1.4270689843383111e-05, |
| "logits/chosen": 0.68359375, |
| "logits/rejected": 0.68359375, |
| "logps/chosen": -64512.0, |
| "logps/rejected": -64512.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -636.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -636.0, |
| "step": 1940 |
| }, |
| { |
| "epoch": 2.0418848167539267, |
| "grad_norm": 0.029062423284783714, |
| "learning_rate": 1.3996322881289347e-05, |
| "logits/chosen": 0.423828125, |
| "logits/rejected": 0.423828125, |
| "logps/chosen": -69632.0, |
| "logps/rejected": -69632.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -684.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -684.0, |
| "step": 1950 |
| }, |
| { |
| "epoch": 2.0523560209424083, |
| "grad_norm": 0.029859130673154842, |
| "learning_rate": 1.3723589973115963e-05, |
| "logits/chosen": 1.2421875, |
| "logits/rejected": 1.2421875, |
| "logps/chosen": -61952.0, |
| "logps/rejected": -61952.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -612.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -612.0, |
| "step": 1960 |
| }, |
| { |
| "epoch": 2.06282722513089, |
| "grad_norm": 0.023398867638769735, |
| "learning_rate": 1.3452531619894837e-05, |
| "logits/chosen": 0.984375, |
| "logits/rejected": 0.984375, |
| "logps/chosen": -67072.0, |
| "logps/rejected": -67072.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -660.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -660.0, |
| "step": 1970 |
| }, |
| { |
| "epoch": 2.073298429319372, |
| "grad_norm": 0.023860638329438577, |
| "learning_rate": 1.3183188073985245e-05, |
| "logits/chosen": 1.25, |
| "logits/rejected": 1.25, |
| "logps/chosen": -62464.0, |
| "logps/rejected": -62464.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -616.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -616.0, |
| "step": 1980 |
| }, |
| { |
| "epoch": 2.0837696335078535, |
| "grad_norm": 0.025263943459597062, |
| "learning_rate": 1.2915599333096351e-05, |
| "logits/chosen": 0.55859375, |
| "logits/rejected": 0.55859375, |
| "logps/chosen": -70144.0, |
| "logps/rejected": -70144.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -692.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -692.0, |
| "step": 1990 |
| }, |
| { |
| "epoch": 2.094240837696335, |
| "grad_norm": 0.01673098039415528, |
| "learning_rate": 1.264980513434752e-05, |
| "logits/chosen": 0.80859375, |
| "logits/rejected": 0.80859375, |
| "logps/chosen": -60416.0, |
| "logps/rejected": -60416.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -596.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -596.0, |
| "step": 2000 |
| }, |
| { |
| "epoch": 2.094240837696335, |
| "eval_logits/chosen": 2.75, |
| "eval_logits/rejected": 3.1875, |
| "eval_logps/chosen": -54016.0, |
| "eval_logps/rejected": -46848.0, |
| "eval_loss": 158.3665008544922, |
| "eval_rewards/accuracies": 0.43359375, |
| "eval_rewards/chosen": -540.0, |
| "eval_rewards/margins": -73.0, |
| "eval_rewards/rejected": -466.0, |
| "eval_runtime": 46.5412, |
| "eval_samples_per_second": 42.973, |
| "eval_steps_per_second": 0.688, |
| "step": 2000 |
| }, |
| { |
| "epoch": 2.1047120418848166, |
| "grad_norm": 0.02138225248413285, |
| "learning_rate": 1.2385844948367324e-05, |
| "logits/chosen": 1.078125, |
| "logits/rejected": 1.078125, |
| "logps/chosen": -66560.0, |
| "logps/rejected": -66560.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -656.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -656.0, |
| "step": 2010 |
| }, |
| { |
| "epoch": 2.115183246073298, |
| "grad_norm": 0.023268156515249223, |
| "learning_rate": 1.2123757973432114e-05, |
| "logits/chosen": 0.78125, |
| "logits/rejected": 0.78125, |
| "logps/chosen": -67072.0, |
| "logps/rejected": -67072.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -660.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -660.0, |
| "step": 2020 |
| }, |
| { |
| "epoch": 2.1256544502617802, |
| "grad_norm": 0.027184544116301168, |
| "learning_rate": 1.1863583129645106e-05, |
| "logits/chosen": 0.953125, |
| "logits/rejected": 0.953125, |
| "logps/chosen": -66560.0, |
| "logps/rejected": -66560.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -656.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -656.0, |
| "step": 2030 |
| }, |
| { |
| "epoch": 2.136125654450262, |
| "grad_norm": 0.020420727955237146, |
| "learning_rate": 1.1605359053156607e-05, |
| "logits/chosen": 0.890625, |
| "logits/rejected": 0.890625, |
| "logps/chosen": -67584.0, |
| "logps/rejected": -67584.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -668.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -668.0, |
| "step": 2040 |
| }, |
| { |
| "epoch": 2.1465968586387434, |
| "grad_norm": 0.01891055575239768, |
| "learning_rate": 1.1349124090426644e-05, |
| "logits/chosen": 0.73046875, |
| "logits/rejected": 0.73046875, |
| "logps/chosen": -70144.0, |
| "logps/rejected": -70144.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -692.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -692.0, |
| "step": 2050 |
| }, |
| { |
| "epoch": 2.157068062827225, |
| "grad_norm": 0.020306967109288147, |
| "learning_rate": 1.1094916292530406e-05, |
| "logits/chosen": 0.84765625, |
| "logits/rejected": 0.84765625, |
| "logps/chosen": -66560.0, |
| "logps/rejected": -66560.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -656.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -656.0, |
| "step": 2060 |
| }, |
| { |
| "epoch": 2.167539267015707, |
| "grad_norm": 0.020174560894979297, |
| "learning_rate": 1.0842773409507623e-05, |
| "logits/chosen": 1.328125, |
| "logits/rejected": 1.328125, |
| "logps/chosen": -64768.0, |
| "logps/rejected": -64768.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -640.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -640.0, |
| "step": 2070 |
| }, |
| { |
| "epoch": 2.1780104712041886, |
| "grad_norm": 0.018672709132978943, |
| "learning_rate": 1.0592732884756754e-05, |
| "logits/chosen": 1.0234375, |
| "logits/rejected": 1.0234375, |
| "logps/chosen": -69120.0, |
| "logps/rejected": -69120.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -684.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -684.0, |
| "step": 2080 |
| }, |
| { |
| "epoch": 2.18848167539267, |
| "grad_norm": 0.019592883243389737, |
| "learning_rate": 1.0344831849474507e-05, |
| "logits/chosen": 1.3203125, |
| "logits/rejected": 1.3203125, |
| "logps/chosen": -65536.0, |
| "logps/rejected": -65536.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -644.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -644.0, |
| "step": 2090 |
| }, |
| { |
| "epoch": 2.1989528795811517, |
| "grad_norm": 0.017404526706102586, |
| "learning_rate": 1.0099107117141879e-05, |
| "logits/chosen": 0.79296875, |
| "logits/rejected": 0.79296875, |
| "logps/chosen": -74752.0, |
| "logps/rejected": -74752.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -736.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -736.0, |
| "step": 2100 |
| }, |
| { |
| "epoch": 2.1989528795811517, |
| "eval_logits/chosen": 2.890625, |
| "eval_logits/rejected": 3.34375, |
| "eval_logps/chosen": -54784.0, |
| "eval_logps/rejected": -47360.0, |
| "eval_loss": 160.32249450683594, |
| "eval_rewards/accuracies": 0.4296875, |
| "eval_rewards/chosen": -544.0, |
| "eval_rewards/margins": -74.0, |
| "eval_rewards/rejected": -470.0, |
| "eval_runtime": 46.5409, |
| "eval_samples_per_second": 42.973, |
| "eval_steps_per_second": 0.688, |
| "step": 2100 |
| }, |
| { |
| "epoch": 2.2094240837696333, |
| "grad_norm": 0.045387205858800586, |
| "learning_rate": 9.855595178057334e-06, |
| "logits/chosen": 0.98828125, |
| "logits/rejected": 0.98828125, |
| "logps/chosen": -70144.0, |
| "logps/rejected": -70144.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -692.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -692.0, |
| "step": 2110 |
| }, |
| { |
| "epoch": 2.2198952879581153, |
| "grad_norm": 0.019491867643683488, |
| "learning_rate": 9.614332193917949e-06, |
| "logits/chosen": 1.1171875, |
| "logits/rejected": 1.1171875, |
| "logps/chosen": -67072.0, |
| "logps/rejected": -67072.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -660.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -660.0, |
| "step": 2120 |
| }, |
| { |
| "epoch": 2.230366492146597, |
| "grad_norm": 0.018320401803212618, |
| "learning_rate": 9.375353992449385e-06, |
| "logits/chosen": 0.63671875, |
| "logits/rejected": 0.63671875, |
| "logps/chosen": -67072.0, |
| "logps/rejected": -67072.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -660.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -660.0, |
| "step": 2130 |
| }, |
| { |
| "epoch": 2.2408376963350785, |
| "grad_norm": 0.019214913692777353, |
| "learning_rate": 9.138696062085442e-06, |
| "logits/chosen": 0.59765625, |
| "logits/rejected": 0.59765625, |
| "logps/chosen": -66048.0, |
| "logps/rejected": -66048.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -648.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -648.0, |
| "step": 2140 |
| }, |
| { |
| "epoch": 2.25130890052356, |
| "grad_norm": 0.01733943235782046, |
| "learning_rate": 8.904393546698007e-06, |
| "logits/chosen": 0.90234375, |
| "logits/rejected": 0.90234375, |
| "logps/chosen": -69632.0, |
| "logps/rejected": -69632.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -684.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -684.0, |
| "step": 2150 |
| }, |
| { |
| "epoch": 2.261780104712042, |
| "grad_norm": 0.01896116142938523, |
| "learning_rate": 8.672481240378142e-06, |
| "logits/chosen": 0.7578125, |
| "logits/rejected": 0.7578125, |
| "logps/chosen": -70144.0, |
| "logps/rejected": -70144.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -692.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -692.0, |
| "step": 2160 |
| }, |
| { |
| "epoch": 2.2722513089005236, |
| "grad_norm": 0.024310842231388265, |
| "learning_rate": 8.44299358226919e-06, |
| "logits/chosen": 0.33984375, |
| "logits/rejected": 0.33984375, |
| "logps/chosen": -73728.0, |
| "logps/rejected": -73728.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -724.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -724.0, |
| "step": 2170 |
| }, |
| { |
| "epoch": 2.282722513089005, |
| "grad_norm": 0.022575534040367915, |
| "learning_rate": 8.215964651452457e-06, |
| "logits/chosen": 0.59375, |
| "logits/rejected": 0.59375, |
| "logps/chosen": -65536.0, |
| "logps/rejected": -65536.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -648.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -648.0, |
| "step": 2180 |
| }, |
| { |
| "epoch": 2.2931937172774868, |
| "grad_norm": 0.025076113573924773, |
| "learning_rate": 7.991428161886502e-06, |
| "logits/chosen": 0.46484375, |
| "logits/rejected": 0.46484375, |
| "logps/chosen": -74240.0, |
| "logps/rejected": -74240.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -732.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -732.0, |
| "step": 2190 |
| }, |
| { |
| "epoch": 2.303664921465969, |
| "grad_norm": 0.0282013130128055, |
| "learning_rate": 7.76941745740061e-06, |
| "logits/chosen": 0.6953125, |
| "logits/rejected": 0.6953125, |
| "logps/chosen": -67072.0, |
| "logps/rejected": -67072.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -664.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -664.0, |
| "step": 2200 |
| }, |
| { |
| "epoch": 2.303664921465969, |
| "eval_logits/chosen": 2.359375, |
| "eval_logits/rejected": 2.734375, |
| "eval_logps/chosen": -55296.0, |
| "eval_logps/rejected": -47616.0, |
| "eval_loss": 161.6043701171875, |
| "eval_rewards/accuracies": 0.431640625, |
| "eval_rewards/chosen": -548.0, |
| "eval_rewards/margins": -74.5, |
| "eval_rewards/rejected": -474.0, |
| "eval_runtime": 46.5397, |
| "eval_samples_per_second": 42.974, |
| "eval_steps_per_second": 0.688, |
| "step": 2200 |
| }, |
| { |
| "epoch": 2.3141361256544504, |
| "grad_norm": 0.02274527990316408, |
| "learning_rate": 7.549965506743123e-06, |
| "logits/chosen": 0.99609375, |
| "logits/rejected": 0.99609375, |
| "logps/chosen": -63488.0, |
| "logps/rejected": -63488.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -628.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -628.0, |
| "step": 2210 |
| }, |
| { |
| "epoch": 2.324607329842932, |
| "grad_norm": 0.016763842986592618, |
| "learning_rate": 7.333104898685631e-06, |
| "logits/chosen": 0.66796875, |
| "logits/rejected": 0.66796875, |
| "logps/chosen": -68608.0, |
| "logps/rejected": -68608.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -676.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -676.0, |
| "step": 2220 |
| }, |
| { |
| "epoch": 2.3350785340314135, |
| "grad_norm": 0.022082913141832444, |
| "learning_rate": 7.118867837183518e-06, |
| "logits/chosen": 1.0078125, |
| "logits/rejected": 1.0078125, |
| "logps/chosen": -65536.0, |
| "logps/rejected": -65536.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -644.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -644.0, |
| "step": 2230 |
| }, |
| { |
| "epoch": 2.345549738219895, |
| "grad_norm": 0.023822883351827116, |
| "learning_rate": 6.9072861365936045e-06, |
| "logits/chosen": 1.1015625, |
| "logits/rejected": 1.1015625, |
| "logps/chosen": -64256.0, |
| "logps/rejected": -64256.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -636.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -636.0, |
| "step": 2240 |
| }, |
| { |
| "epoch": 2.356020942408377, |
| "grad_norm": 0.02201829579670575, |
| "learning_rate": 6.6983912169497005e-06, |
| "logits/chosen": 0.2158203125, |
| "logits/rejected": 0.2158203125, |
| "logps/chosen": -70656.0, |
| "logps/rejected": -70656.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -696.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -696.0, |
| "step": 2250 |
| }, |
| { |
| "epoch": 2.3664921465968587, |
| "grad_norm": 0.025325491804689612, |
| "learning_rate": 6.49221409929677e-06, |
| "logits/chosen": 0.609375, |
| "logits/rejected": 0.609375, |
| "logps/chosen": -71168.0, |
| "logps/rejected": -71168.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -700.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -700.0, |
| "step": 2260 |
| }, |
| { |
| "epoch": 2.3769633507853403, |
| "grad_norm": 0.0240953293008413, |
| "learning_rate": 6.2887854010842075e-06, |
| "logits/chosen": 0.8359375, |
| "logits/rejected": 0.8359375, |
| "logps/chosen": -68096.0, |
| "logps/rejected": -68096.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -672.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -672.0, |
| "step": 2270 |
| }, |
| { |
| "epoch": 2.387434554973822, |
| "grad_norm": 0.024347997703830785, |
| "learning_rate": 6.0881353316191385e-06, |
| "logits/chosen": 0.43359375, |
| "logits/rejected": 0.43359375, |
| "logps/chosen": -68096.0, |
| "logps/rejected": -68096.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -672.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -672.0, |
| "step": 2280 |
| }, |
| { |
| "epoch": 2.3979057591623034, |
| "grad_norm": 0.025627642211693582, |
| "learning_rate": 5.890293687580381e-06, |
| "logits/chosen": 1.1796875, |
| "logits/rejected": 1.1796875, |
| "logps/chosen": -64000.0, |
| "logps/rejected": -64000.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -632.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -632.0, |
| "step": 2290 |
| }, |
| { |
| "epoch": 2.4083769633507854, |
| "grad_norm": 0.020077970537725823, |
| "learning_rate": 5.695289848593532e-06, |
| "logits/chosen": 0.6875, |
| "logits/rejected": 0.6875, |
| "logps/chosen": -65024.0, |
| "logps/rejected": -65024.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -640.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -640.0, |
| "step": 2300 |
| }, |
| { |
| "epoch": 2.4083769633507854, |
| "eval_logits/chosen": 2.40625, |
| "eval_logits/rejected": 2.828125, |
| "eval_logps/chosen": -55552.0, |
| "eval_logps/rejected": -48128.0, |
| "eval_loss": 162.53775024414062, |
| "eval_rewards/accuracies": 0.431640625, |
| "eval_rewards/chosen": -552.0, |
| "eval_rewards/margins": -75.0, |
| "eval_rewards/rejected": -478.0, |
| "eval_runtime": 46.5416, |
| "eval_samples_per_second": 42.972, |
| "eval_steps_per_second": 0.688, |
| "step": 2300 |
| }, |
| { |
| "epoch": 2.418848167539267, |
| "grad_norm": 0.019863870780562495, |
| "learning_rate": 5.50315277286812e-06, |
| "logits/chosen": 0.98046875, |
| "logits/rejected": 0.98046875, |
| "logps/chosen": -64256.0, |
| "logps/rejected": -64256.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -632.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -632.0, |
| "step": 2310 |
| }, |
| { |
| "epoch": 2.4293193717277486, |
| "grad_norm": 0.025340285858406048, |
| "learning_rate": 5.313910992897281e-06, |
| "logits/chosen": 0.90234375, |
| "logits/rejected": 0.90234375, |
| "logps/chosen": -68096.0, |
| "logps/rejected": -68096.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -668.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -668.0, |
| "step": 2320 |
| }, |
| { |
| "epoch": 2.4397905759162306, |
| "grad_norm": 0.023647170245640965, |
| "learning_rate": 5.127592611220658e-06, |
| "logits/chosen": 0.77734375, |
| "logits/rejected": 0.77734375, |
| "logps/chosen": -66560.0, |
| "logps/rejected": -66560.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -656.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -656.0, |
| "step": 2330 |
| }, |
| { |
| "epoch": 2.450261780104712, |
| "grad_norm": 0.028971862520499957, |
| "learning_rate": 4.944225296251159e-06, |
| "logits/chosen": 0.373046875, |
| "logits/rejected": 0.373046875, |
| "logps/chosen": -73216.0, |
| "logps/rejected": -73216.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -724.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -724.0, |
| "step": 2340 |
| }, |
| { |
| "epoch": 2.4607329842931938, |
| "grad_norm": 0.024237666342194673, |
| "learning_rate": 4.763836278166172e-06, |
| "logits/chosen": 0.75390625, |
| "logits/rejected": 0.75390625, |
| "logps/chosen": -65536.0, |
| "logps/rejected": -65536.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -648.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -648.0, |
| "step": 2350 |
| }, |
| { |
| "epoch": 2.4712041884816753, |
| "grad_norm": 0.02016821868498419, |
| "learning_rate": 4.586452344863876e-06, |
| "logits/chosen": 1.0234375, |
| "logits/rejected": 1.0234375, |
| "logps/chosen": -65024.0, |
| "logps/rejected": -65024.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -644.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -644.0, |
| "step": 2360 |
| }, |
| { |
| "epoch": 2.481675392670157, |
| "grad_norm": 0.02734977255352393, |
| "learning_rate": 4.412099837985192e-06, |
| "logits/chosen": 0.66015625, |
| "logits/rejected": 0.66015625, |
| "logps/chosen": -69632.0, |
| "logps/rejected": -69632.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -684.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -684.0, |
| "step": 2370 |
| }, |
| { |
| "epoch": 2.492146596858639, |
| "grad_norm": 0.028035619296303077, |
| "learning_rate": 4.240804649002089e-06, |
| "logits/chosen": 0.5546875, |
| "logits/rejected": 0.5546875, |
| "logps/chosen": -70144.0, |
| "logps/rejected": -70144.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -692.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -692.0, |
| "step": 2380 |
| }, |
| { |
| "epoch": 2.5026178010471205, |
| "grad_norm": 0.03298219557731263, |
| "learning_rate": 4.072592215372598e-06, |
| "logits/chosen": 0.9453125, |
| "logits/rejected": 0.9453125, |
| "logps/chosen": -71168.0, |
| "logps/rejected": -71168.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -700.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -700.0, |
| "step": 2390 |
| }, |
| { |
| "epoch": 2.513089005235602, |
| "grad_norm": 0.02818788109431593, |
| "learning_rate": 3.907487516763389e-06, |
| "logits/chosen": 0.77734375, |
| "logits/rejected": 0.77734375, |
| "logps/chosen": -71680.0, |
| "logps/rejected": -71680.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -708.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -708.0, |
| "step": 2400 |
| }, |
| { |
| "epoch": 2.513089005235602, |
| "eval_logits/chosen": 2.546875, |
| "eval_logits/rejected": 3.0, |
| "eval_logps/chosen": -55808.0, |
| "eval_logps/rejected": -48128.0, |
| "eval_loss": 163.31837463378906, |
| "eval_rewards/accuracies": 0.43359375, |
| "eval_rewards/chosen": -556.0, |
| "eval_rewards/margins": -75.5, |
| "eval_rewards/rejected": -480.0, |
| "eval_runtime": 46.5272, |
| "eval_samples_per_second": 42.986, |
| "eval_steps_per_second": 0.688, |
| "step": 2400 |
| }, |
| { |
| "epoch": 2.5235602094240837, |
| "grad_norm": 0.0204743474792107, |
| "learning_rate": 3.745515071340286e-06, |
| "logits/chosen": 0.5703125, |
| "logits/rejected": 0.5703125, |
| "logps/chosen": -69632.0, |
| "logps/rejected": -69632.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -688.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -688.0, |
| "step": 2410 |
| }, |
| { |
| "epoch": 2.5340314136125652, |
| "grad_norm": 0.02709018840295767, |
| "learning_rate": 3.586698932127236e-06, |
| "logits/chosen": 1.046875, |
| "logits/rejected": 1.046875, |
| "logps/chosen": -69120.0, |
| "logps/rejected": -69120.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -680.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -680.0, |
| "step": 2420 |
| }, |
| { |
| "epoch": 2.5445026178010473, |
| "grad_norm": 0.02128822310623819, |
| "learning_rate": 3.4310626834344744e-06, |
| "logits/chosen": 1.3359375, |
| "logits/rejected": 1.3359375, |
| "logps/chosen": -68608.0, |
| "logps/rejected": -68608.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -676.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -676.0, |
| "step": 2430 |
| }, |
| { |
| "epoch": 2.554973821989529, |
| "grad_norm": 0.02250789390550962, |
| "learning_rate": 3.2786294373562343e-06, |
| "logits/chosen": 1.5390625, |
| "logits/rejected": 1.5390625, |
| "logps/chosen": -66048.0, |
| "logps/rejected": -66048.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -652.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -652.0, |
| "step": 2440 |
| }, |
| { |
| "epoch": 2.5654450261780104, |
| "grad_norm": 0.020561374381945968, |
| "learning_rate": 3.1294218303385526e-06, |
| "logits/chosen": 1.0859375, |
| "logits/rejected": 1.0859375, |
| "logps/chosen": -70656.0, |
| "logps/rejected": -70656.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -696.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -696.0, |
| "step": 2450 |
| }, |
| { |
| "epoch": 2.5759162303664924, |
| "grad_norm": 0.02552327592015866, |
| "learning_rate": 2.9834620198177483e-06, |
| "logits/chosen": 0.82421875, |
| "logits/rejected": 0.82421875, |
| "logps/chosen": -66560.0, |
| "logps/rejected": -66560.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -656.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -656.0, |
| "step": 2460 |
| }, |
| { |
| "epoch": 2.5863874345549736, |
| "grad_norm": 0.023561188096015716, |
| "learning_rate": 2.840771680930068e-06, |
| "logits/chosen": 1.2890625, |
| "logits/rejected": 1.2890625, |
| "logps/chosen": -64000.0, |
| "logps/rejected": -64000.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -632.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -632.0, |
| "step": 2470 |
| }, |
| { |
| "epoch": 2.5968586387434556, |
| "grad_norm": 0.02846868283344398, |
| "learning_rate": 2.701372003292868e-06, |
| "logits/chosen": 0.87890625, |
| "logits/rejected": 0.87890625, |
| "logps/chosen": -75776.0, |
| "logps/rejected": -75776.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -744.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -744.0, |
| "step": 2480 |
| }, |
| { |
| "epoch": 2.607329842931937, |
| "grad_norm": 0.02652568001311377, |
| "learning_rate": 2.56528368785795e-06, |
| "logits/chosen": 1.3125, |
| "logits/rejected": 1.3125, |
| "logps/chosen": -65280.0, |
| "logps/rejected": -65280.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -644.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -644.0, |
| "step": 2490 |
| }, |
| { |
| "epoch": 2.6178010471204187, |
| "grad_norm": 0.02705623050186832, |
| "learning_rate": 2.432526943837471e-06, |
| "logits/chosen": 0.69921875, |
| "logits/rejected": 0.69921875, |
| "logps/chosen": -72192.0, |
| "logps/rejected": -72192.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -712.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -712.0, |
| "step": 2500 |
| }, |
| { |
| "epoch": 2.6178010471204187, |
| "eval_logits/chosen": 2.75, |
| "eval_logits/rejected": 3.1875, |
| "eval_logps/chosen": -56064.0, |
| "eval_logps/rejected": -48384.0, |
| "eval_loss": 163.9196319580078, |
| "eval_rewards/accuracies": 0.431640625, |
| "eval_rewards/chosen": -556.0, |
| "eval_rewards/margins": -75.5, |
| "eval_rewards/rejected": -482.0, |
| "eval_runtime": 46.537, |
| "eval_samples_per_second": 42.977, |
| "eval_steps_per_second": 0.688, |
| "step": 2500 |
| }, |
| { |
| "epoch": 2.6282722513089007, |
| "grad_norm": 0.02874819359097377, |
| "learning_rate": 2.3031214857028446e-06, |
| "logits/chosen": 1.203125, |
| "logits/rejected": 1.203125, |
| "logps/chosen": -67072.0, |
| "logps/rejected": -67072.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -660.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -660.0, |
| "step": 2510 |
| }, |
| { |
| "epoch": 2.6387434554973823, |
| "grad_norm": 0.027417391571089873, |
| "learning_rate": 2.177086530257139e-06, |
| "logits/chosen": 1.0234375, |
| "logits/rejected": 1.0234375, |
| "logps/chosen": -70144.0, |
| "logps/rejected": -70144.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -692.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -692.0, |
| "step": 2520 |
| }, |
| { |
| "epoch": 2.649214659685864, |
| "grad_norm": 0.02590818241825701, |
| "learning_rate": 2.054440793781354e-06, |
| "logits/chosen": 0.9765625, |
| "logits/rejected": 0.9765625, |
| "logps/chosen": -71168.0, |
| "logps/rejected": -71168.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -700.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -700.0, |
| "step": 2530 |
| }, |
| { |
| "epoch": 2.6596858638743455, |
| "grad_norm": 0.028079858446008157, |
| "learning_rate": 1.9352024892550587e-06, |
| "logits/chosen": 0.5859375, |
| "logits/rejected": 0.5859375, |
| "logps/chosen": -72704.0, |
| "logps/rejected": -72704.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -712.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -712.0, |
| "step": 2540 |
| }, |
| { |
| "epoch": 2.670157068062827, |
| "grad_norm": 0.02584256473218433, |
| "learning_rate": 1.8193893236517379e-06, |
| "logits/chosen": 1.265625, |
| "logits/rejected": 1.265625, |
| "logps/chosen": -69120.0, |
| "logps/rejected": -69120.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -680.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -680.0, |
| "step": 2550 |
| }, |
| { |
| "epoch": 2.680628272251309, |
| "grad_norm": 0.026138098555928222, |
| "learning_rate": 1.7070184953092956e-06, |
| "logits/chosen": 0.921875, |
| "logits/rejected": 0.921875, |
| "logps/chosen": -71168.0, |
| "logps/rejected": -71168.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -700.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -700.0, |
| "step": 2560 |
| }, |
| { |
| "epoch": 2.6910994764397906, |
| "grad_norm": 0.02077750861766315, |
| "learning_rate": 1.5981066913760918e-06, |
| "logits/chosen": 0.83203125, |
| "logits/rejected": 0.83203125, |
| "logps/chosen": -68096.0, |
| "logps/rejected": -68096.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -672.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -672.0, |
| "step": 2570 |
| }, |
| { |
| "epoch": 2.701570680628272, |
| "grad_norm": 0.023474194364334084, |
| "learning_rate": 1.492670085332884e-06, |
| "logits/chosen": 1.0546875, |
| "logits/rejected": 1.0546875, |
| "logps/chosen": -67584.0, |
| "logps/rejected": -67584.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -664.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -664.0, |
| "step": 2580 |
| }, |
| { |
| "epoch": 2.712041884816754, |
| "grad_norm": 0.023309298076190376, |
| "learning_rate": 1.3907243345910787e-06, |
| "logits/chosen": 1.3046875, |
| "logits/rejected": 1.3046875, |
| "logps/chosen": -64768.0, |
| "logps/rejected": -64768.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -640.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -640.0, |
| "step": 2590 |
| }, |
| { |
| "epoch": 2.7225130890052354, |
| "grad_norm": 0.02514569492305495, |
| "learning_rate": 1.292284578167538e-06, |
| "logits/chosen": 0.8046875, |
| "logits/rejected": 0.8046875, |
| "logps/chosen": -71168.0, |
| "logps/rejected": -71168.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -704.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -704.0, |
| "step": 2600 |
| }, |
| { |
| "epoch": 2.7225130890052354, |
| "eval_logits/chosen": 2.734375, |
| "eval_logits/rejected": 3.171875, |
| "eval_logps/chosen": -56064.0, |
| "eval_logps/rejected": -48640.0, |
| "eval_loss": 164.26974487304688, |
| "eval_rewards/accuracies": 0.4296875, |
| "eval_rewards/chosen": -556.0, |
| "eval_rewards/margins": -76.0, |
| "eval_rewards/rejected": -482.0, |
| "eval_runtime": 46.5533, |
| "eval_samples_per_second": 42.962, |
| "eval_steps_per_second": 0.687, |
| "step": 2600 |
| }, |
| { |
| "epoch": 2.7329842931937174, |
| "grad_norm": 0.02550138640729088, |
| "learning_rate": 1.1973654344364925e-06, |
| "logits/chosen": 1.3359375, |
| "logits/rejected": 1.3359375, |
| "logps/chosen": -65536.0, |
| "logps/rejected": -65536.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -648.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -648.0, |
| "step": 2610 |
| }, |
| { |
| "epoch": 2.743455497382199, |
| "grad_norm": 0.02268119160253597, |
| "learning_rate": 1.1059809989586583e-06, |
| "logits/chosen": 0.8828125, |
| "logits/rejected": 0.8828125, |
| "logps/chosen": -69120.0, |
| "logps/rejected": -69120.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -684.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -684.0, |
| "step": 2620 |
| }, |
| { |
| "epoch": 2.7539267015706805, |
| "grad_norm": 0.023205390999682877, |
| "learning_rate": 1.018144842388044e-06, |
| "logits/chosen": 0.78515625, |
| "logits/rejected": 0.78515625, |
| "logps/chosen": -66560.0, |
| "logps/rejected": -66560.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -656.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -656.0, |
| "step": 2630 |
| }, |
| { |
| "epoch": 2.7643979057591626, |
| "grad_norm": 0.0185911410326415, |
| "learning_rate": 9.338700084567109e-07, |
| "logits/chosen": 0.77734375, |
| "logits/rejected": 0.77734375, |
| "logps/chosen": -71168.0, |
| "logps/rejected": -71168.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -700.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -700.0, |
| "step": 2640 |
| }, |
| { |
| "epoch": 2.774869109947644, |
| "grad_norm": 0.024166869400554982, |
| "learning_rate": 8.531690120377606e-07, |
| "logits/chosen": 1.2265625, |
| "logits/rejected": 1.2265625, |
| "logps/chosen": -68096.0, |
| "logps/rejected": -68096.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -672.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -672.0, |
| "step": 2650 |
| }, |
| { |
| "epoch": 2.7853403141361257, |
| "grad_norm": 0.01889375914953375, |
| "learning_rate": 7.760538372868636e-07, |
| "logits/chosen": 1.3203125, |
| "logits/rejected": 1.3203125, |
| "logps/chosen": -62208.0, |
| "logps/rejected": -62208.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -616.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -616.0, |
| "step": 2660 |
| }, |
| { |
| "epoch": 2.7958115183246073, |
| "grad_norm": 0.024325408368237968, |
| "learning_rate": 7.025359358626166e-07, |
| "logits/chosen": 0.8828125, |
| "logits/rejected": 0.8828125, |
| "logps/chosen": -72704.0, |
| "logps/rejected": -72704.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -720.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -720.0, |
| "step": 2670 |
| }, |
| { |
| "epoch": 2.806282722513089, |
| "grad_norm": 0.057104076142959634, |
| "learning_rate": 6.326262252259463e-07, |
| "logits/chosen": 1.1484375, |
| "logits/rejected": 1.1484375, |
| "logps/chosen": -69632.0, |
| "logps/rejected": -69632.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -688.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -688.0, |
| "step": 2680 |
| }, |
| { |
| "epoch": 2.816753926701571, |
| "grad_norm": 0.025067239688892188, |
| "learning_rate": 5.66335087018871e-07, |
| "logits/chosen": 1.1484375, |
| "logits/rejected": 1.1484375, |
| "logps/chosen": -62976.0, |
| "logps/rejected": -62976.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -620.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -620.0, |
| "step": 2690 |
| }, |
| { |
| "epoch": 2.8272251308900525, |
| "grad_norm": 0.022099839828897808, |
| "learning_rate": 5.036723655228226e-07, |
| "logits/chosen": 1.0390625, |
| "logits/rejected": 1.0390625, |
| "logps/chosen": -68096.0, |
| "logps/rejected": -68096.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -672.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -672.0, |
| "step": 2700 |
| }, |
| { |
| "epoch": 2.8272251308900525, |
| "eval_logits/chosen": 2.71875, |
| "eval_logits/rejected": 3.15625, |
| "eval_logps/chosen": -56064.0, |
| "eval_logps/rejected": -48640.0, |
| "eval_loss": 164.35400390625, |
| "eval_rewards/accuracies": 0.4296875, |
| "eval_rewards/chosen": -560.0, |
| "eval_rewards/margins": -76.0, |
| "eval_rewards/rejected": -482.0, |
| "eval_runtime": 46.5423, |
| "eval_samples_per_second": 42.972, |
| "eval_steps_per_second": 0.688, |
| "step": 2700 |
| }, |
| { |
| "epoch": 2.837696335078534, |
| "grad_norm": 0.021342555273502758, |
| "learning_rate": 4.446473661967432e-07, |
| "logits/chosen": 0.890625, |
| "logits/rejected": 0.890625, |
| "logps/chosen": -68608.0, |
| "logps/rejected": -68608.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -672.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -672.0, |
| "step": 2710 |
| }, |
| { |
| "epoch": 2.8481675392670156, |
| "grad_norm": 0.026225163063370836, |
| "learning_rate": 3.892688542952411e-07, |
| "logits/chosen": 0.8828125, |
| "logits/rejected": 0.8828125, |
| "logps/chosen": -71168.0, |
| "logps/rejected": -71168.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -700.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -700.0, |
| "step": 2720 |
| }, |
| { |
| "epoch": 2.858638743455497, |
| "grad_norm": 0.025671354427707384, |
| "learning_rate": 3.3754505356694e-07, |
| "logits/chosen": 0.8359375, |
| "logits/rejected": 0.8359375, |
| "logps/chosen": -66560.0, |
| "logps/rejected": -66560.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -656.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -656.0, |
| "step": 2730 |
| }, |
| { |
| "epoch": 2.869109947643979, |
| "grad_norm": 0.02153044737028141, |
| "learning_rate": 2.8948364503322276e-07, |
| "logits/chosen": 0.55859375, |
| "logits/rejected": 0.55859375, |
| "logps/chosen": -78848.0, |
| "logps/rejected": -78848.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -776.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -776.0, |
| "step": 2740 |
| }, |
| { |
| "epoch": 2.8795811518324608, |
| "grad_norm": 0.0202096023215154, |
| "learning_rate": 2.4509176584762706e-07, |
| "logits/chosen": 1.25, |
| "logits/rejected": 1.25, |
| "logps/chosen": -66560.0, |
| "logps/rejected": -66560.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -656.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -656.0, |
| "step": 2750 |
| }, |
| { |
| "epoch": 2.8900523560209423, |
| "grad_norm": 0.025654819978643818, |
| "learning_rate": 2.0437600823595694e-07, |
| "logits/chosen": 1.4765625, |
| "logits/rejected": 1.4765625, |
| "logps/chosen": -66560.0, |
| "logps/rejected": -66560.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -656.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -656.0, |
| "step": 2760 |
| }, |
| { |
| "epoch": 2.900523560209424, |
| "grad_norm": 0.027063287797210487, |
| "learning_rate": 1.6734241851733e-07, |
| "logits/chosen": 0.5859375, |
| "logits/rejected": 0.5859375, |
| "logps/chosen": -75776.0, |
| "logps/rejected": -75776.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -748.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -748.0, |
| "step": 2770 |
| }, |
| { |
| "epoch": 2.9109947643979055, |
| "grad_norm": 0.028512586933205398, |
| "learning_rate": 1.339964962062984e-07, |
| "logits/chosen": 1.1015625, |
| "logits/rejected": 1.1015625, |
| "logps/chosen": -64512.0, |
| "logps/rejected": -64512.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -640.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -640.0, |
| "step": 2780 |
| }, |
| { |
| "epoch": 2.9214659685863875, |
| "grad_norm": 0.0251364871255187, |
| "learning_rate": 1.0434319319617136e-07, |
| "logits/chosen": 0.6328125, |
| "logits/rejected": 0.6328125, |
| "logps/chosen": -75264.0, |
| "logps/rejected": -75264.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -740.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -740.0, |
| "step": 2790 |
| }, |
| { |
| "epoch": 2.931937172774869, |
| "grad_norm": 0.024200605256505008, |
| "learning_rate": 7.838691302365363e-08, |
| "logits/chosen": 1.109375, |
| "logits/rejected": 1.109375, |
| "logps/chosen": -70144.0, |
| "logps/rejected": -70144.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -692.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -692.0, |
| "step": 2800 |
| }, |
| { |
| "epoch": 2.931937172774869, |
| "eval_logits/chosen": 2.703125, |
| "eval_logits/rejected": 3.15625, |
| "eval_logps/chosen": -56320.0, |
| "eval_logps/rejected": -48640.0, |
| "eval_loss": 164.41799926757812, |
| "eval_rewards/accuracies": 0.427734375, |
| "eval_rewards/chosen": -560.0, |
| "eval_rewards/margins": -76.0, |
| "eval_rewards/rejected": -482.0, |
| "eval_runtime": 46.5367, |
| "eval_samples_per_second": 42.977, |
| "eval_steps_per_second": 0.688, |
| "step": 2800 |
| }, |
| { |
| "epoch": 2.9424083769633507, |
| "grad_norm": 0.02336892708549287, |
| "learning_rate": 5.613151021490459e-08, |
| "logits/chosen": 1.203125, |
| "logits/rejected": 1.203125, |
| "logps/chosen": -72192.0, |
| "logps/rejected": -72192.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -712.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -712.0, |
| "step": 2810 |
| }, |
| { |
| "epoch": 2.9528795811518327, |
| "grad_norm": 0.02412563157346396, |
| "learning_rate": 3.7580289713159965e-08, |
| "logits/chosen": 1.03125, |
| "logits/rejected": 1.03125, |
| "logps/chosen": -71168.0, |
| "logps/rejected": -71168.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -704.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -704.0, |
| "step": 2820 |
| }, |
| { |
| "epoch": 2.9633507853403143, |
| "grad_norm": 0.023117545412084186, |
| "learning_rate": 2.2736006387927277e-08, |
| "logits/chosen": 0.8359375, |
| "logits/rejected": 0.8359375, |
| "logps/chosen": -69632.0, |
| "logps/rejected": -69632.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -692.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -692.0, |
| "step": 2830 |
| }, |
| { |
| "epoch": 2.973821989528796, |
| "grad_norm": 0.029753454582507805, |
| "learning_rate": 1.1600864625893604e-08, |
| "logits/chosen": 1.28125, |
| "logits/rejected": 1.28125, |
| "logps/chosen": -74240.0, |
| "logps/rejected": -74240.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -732.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -732.0, |
| "step": 2840 |
| }, |
| { |
| "epoch": 2.9842931937172774, |
| "grad_norm": 0.02190610918374442, |
| "learning_rate": 4.176518003573549e-09, |
| "logits/chosen": 0.6484375, |
| "logits/rejected": 0.6484375, |
| "logps/chosen": -67072.0, |
| "logps/rejected": -67072.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -660.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -660.0, |
| "step": 2850 |
| }, |
| { |
| "epoch": 2.994764397905759, |
| "grad_norm": 0.027010425293612455, |
| "learning_rate": 4.640690417528903e-10, |
| "logits/chosen": 0.91796875, |
| "logits/rejected": 0.91796875, |
| "logps/chosen": -69632.0, |
| "logps/rejected": -69632.0, |
| "loss": 0.6914, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": -688.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": -688.0, |
| "step": 2860 |
| }, |
| { |
| "epoch": 3.0, |
| "step": 2865, |
| "total_flos": 0.0, |
| "train_loss": 0.69140625, |
| "train_runtime": 12252.8173, |
| "train_samples_per_second": 14.964, |
| "train_steps_per_second": 0.234 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 2865, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 3, |
| "save_steps": 500, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": false, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 0.0, |
| "train_batch_size": 8, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
