| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 50, | |
| "global_step": 476, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02100840336134454, | |
| "grad_norm": 16.28906943780765, | |
| "learning_rate": 1.0416666666666667e-07, | |
| "logits/chosen": -1.109375, | |
| "logits/rejected": -1.09375, | |
| "logps/chosen": -3552.0, | |
| "logps/rejected": -3360.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.04201680672268908, | |
| "grad_norm": 17.42578535458299, | |
| "learning_rate": 2.0833333333333333e-07, | |
| "logits/chosen": -1.171875, | |
| "logits/rejected": -1.109375, | |
| "logps/chosen": -4256.0, | |
| "logps/rejected": -3408.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06302521008403361, | |
| "grad_norm": 19.8731189533398, | |
| "learning_rate": 3.1249999999999997e-07, | |
| "logits/chosen": -1.3203125, | |
| "logits/rejected": -1.2890625, | |
| "logps/chosen": -4288.0, | |
| "logps/rejected": -3888.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.08403361344537816, | |
| "grad_norm": 14.405736256208325, | |
| "learning_rate": 4.1666666666666667e-07, | |
| "logits/chosen": -1.7265625, | |
| "logits/rejected": -1.6796875, | |
| "logps/chosen": -4192.0, | |
| "logps/rejected": -3312.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.10504201680672269, | |
| "grad_norm": 9.363286794597972, | |
| "learning_rate": 4.999730614378151e-07, | |
| "logits/chosen": -2.375, | |
| "logits/rejected": -2.34375, | |
| "logps/chosen": -3728.0, | |
| "logps/rejected": -3392.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.10504201680672269, | |
| "eval_logits/chosen": -2.5, | |
| "eval_logits/rejected": -2.484375, | |
| "eval_logps/chosen": -3440.0, | |
| "eval_logps/rejected": -3024.0, | |
| "eval_loss": 0.69140625, | |
| "eval_rewards/accuracies": 0.0, | |
| "eval_rewards/chosen": 0.0, | |
| "eval_rewards/margins": 0.0, | |
| "eval_rewards/rejected": 0.0, | |
| "eval_runtime": 102.4147, | |
| "eval_samples_per_second": 19.392, | |
| "eval_steps_per_second": 0.312, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.12605042016806722, | |
| "grad_norm": 11.711357348387072, | |
| "learning_rate": 4.99030821197584e-07, | |
| "logits/chosen": -2.5, | |
| "logits/rejected": -2.5, | |
| "logps/chosen": -3216.0, | |
| "logps/rejected": -2848.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.14705882352941177, | |
| "grad_norm": 10.992333167872541, | |
| "learning_rate": 4.967474527158459e-07, | |
| "logits/chosen": -2.53125, | |
| "logits/rejected": -2.515625, | |
| "logps/chosen": -3584.0, | |
| "logps/rejected": -3344.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.16806722689075632, | |
| "grad_norm": 10.760162307426238, | |
| "learning_rate": 4.931352528237397e-07, | |
| "logits/chosen": -2.421875, | |
| "logits/rejected": -2.421875, | |
| "logps/chosen": -3312.0, | |
| "logps/rejected": -3024.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.18907563025210083, | |
| "grad_norm": 12.926141811678383, | |
| "learning_rate": 4.882136746285794e-07, | |
| "logits/chosen": -2.28125, | |
| "logits/rejected": -2.28125, | |
| "logps/chosen": -3120.0, | |
| "logps/rejected": -2784.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.21008403361344538, | |
| "grad_norm": 7.053588811204541, | |
| "learning_rate": 4.820092227512735e-07, | |
| "logits/chosen": -2.171875, | |
| "logits/rejected": -2.15625, | |
| "logps/chosen": -3392.0, | |
| "logps/rejected": -2928.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.21008403361344538, | |
| "eval_logits/chosen": -2.046875, | |
| "eval_logits/rejected": -2.03125, | |
| "eval_logps/chosen": -2880.0, | |
| "eval_logps/rejected": -2544.0, | |
| "eval_loss": 0.69140625, | |
| "eval_rewards/accuracies": 0.0, | |
| "eval_rewards/chosen": 0.0, | |
| "eval_rewards/margins": 0.0, | |
| "eval_rewards/rejected": 0.0, | |
| "eval_runtime": 102.9839, | |
| "eval_samples_per_second": 19.285, | |
| "eval_steps_per_second": 0.311, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.23109243697478993, | |
| "grad_norm": 9.09261170699122, | |
| "learning_rate": 4.745553105885882e-07, | |
| "logits/chosen": -1.953125, | |
| "logits/rejected": -1.9453125, | |
| "logps/chosen": -2832.0, | |
| "logps/rejected": -2432.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.25210084033613445, | |
| "grad_norm": 12.748620762742686, | |
| "learning_rate": 4.658920803689553e-07, | |
| "logits/chosen": -1.8671875, | |
| "logits/rejected": -1.859375, | |
| "logps/chosen": -3056.0, | |
| "logps/rejected": -2544.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.27310924369747897, | |
| "grad_norm": 7.022734911341551, | |
| "learning_rate": 4.560661869708929e-07, | |
| "logits/chosen": -1.8984375, | |
| "logits/rejected": -1.890625, | |
| "logps/chosen": -2928.0, | |
| "logps/rejected": -2768.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.29411764705882354, | |
| "grad_norm": 9.578363785870927, | |
| "learning_rate": 4.4513054666826144e-07, | |
| "logits/chosen": -1.7734375, | |
| "logits/rejected": -1.765625, | |
| "logps/chosen": -3136.0, | |
| "logps/rejected": -2704.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.31512605042016806, | |
| "grad_norm": 8.139451643424685, | |
| "learning_rate": 4.3314405215546036e-07, | |
| "logits/chosen": -1.6875, | |
| "logits/rejected": -1.6953125, | |
| "logps/chosen": -2608.0, | |
| "logps/rejected": -2528.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.31512605042016806, | |
| "eval_logits/chosen": -1.5859375, | |
| "eval_logits/rejected": -1.59375, | |
| "eval_logps/chosen": -2736.0, | |
| "eval_logps/rejected": -2432.0, | |
| "eval_loss": 0.69140625, | |
| "eval_rewards/accuracies": 0.0, | |
| "eval_rewards/chosen": 0.0, | |
| "eval_rewards/margins": 0.0, | |
| "eval_rewards/rejected": 0.0, | |
| "eval_runtime": 103.2169, | |
| "eval_samples_per_second": 19.241, | |
| "eval_steps_per_second": 0.31, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.33613445378151263, | |
| "grad_norm": 17.892723758314286, | |
| "learning_rate": 4.201712553872657e-07, | |
| "logits/chosen": -1.6875, | |
| "logits/rejected": -1.6953125, | |
| "logps/chosen": -2688.0, | |
| "logps/rejected": -2544.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.35714285714285715, | |
| "grad_norm": 22.27823672845172, | |
| "learning_rate": 4.062820199413401e-07, | |
| "logits/chosen": -1.90625, | |
| "logits/rejected": -1.8984375, | |
| "logps/chosen": -2816.0, | |
| "logps/rejected": -2336.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.37815126050420167, | |
| "grad_norm": 14.06298479728037, | |
| "learning_rate": 3.9155114477557926e-07, | |
| "logits/chosen": -1.734375, | |
| "logits/rejected": -1.7421875, | |
| "logps/chosen": -2944.0, | |
| "logps/rejected": -2512.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.39915966386554624, | |
| "grad_norm": 7.672499104925795, | |
| "learning_rate": 3.760579614065076e-07, | |
| "logits/chosen": -1.5625, | |
| "logits/rejected": -1.5625, | |
| "logps/chosen": -3040.0, | |
| "logps/rejected": -2512.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.42016806722689076, | |
| "grad_norm": 6.6620635447942975, | |
| "learning_rate": 3.598859066780754e-07, | |
| "logits/chosen": -1.6796875, | |
| "logits/rejected": -1.6796875, | |
| "logps/chosen": -2832.0, | |
| "logps/rejected": -2432.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.42016806722689076, | |
| "eval_logits/chosen": -1.6875, | |
| "eval_logits/rejected": -1.6875, | |
| "eval_logps/chosen": -2688.0, | |
| "eval_logps/rejected": -2384.0, | |
| "eval_loss": 0.69140625, | |
| "eval_rewards/accuracies": 0.0, | |
| "eval_rewards/chosen": 0.0, | |
| "eval_rewards/margins": 0.0, | |
| "eval_rewards/rejected": 0.0, | |
| "eval_runtime": 102.6512, | |
| "eval_samples_per_second": 19.347, | |
| "eval_steps_per_second": 0.312, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4411764705882353, | |
| "grad_norm": 13.458763562501465, | |
| "learning_rate": 3.4312207342166483e-07, | |
| "logits/chosen": -1.7109375, | |
| "logits/rejected": -1.71875, | |
| "logps/chosen": -2464.0, | |
| "logps/rejected": -2192.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.46218487394957986, | |
| "grad_norm": 11.168375013873582, | |
| "learning_rate": 3.2585674142717477e-07, | |
| "logits/chosen": -1.7265625, | |
| "logits/rejected": -1.7265625, | |
| "logps/chosen": -3024.0, | |
| "logps/rejected": -2480.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4831932773109244, | |
| "grad_norm": 14.465492326709429, | |
| "learning_rate": 3.081828912510901e-07, | |
| "logits/chosen": -1.6640625, | |
| "logits/rejected": -1.671875, | |
| "logps/chosen": -2880.0, | |
| "logps/rejected": -2448.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5042016806722689, | |
| "grad_norm": 9.34309193516938, | |
| "learning_rate": 2.9019570347986706e-07, | |
| "logits/chosen": -1.6953125, | |
| "logits/rejected": -1.6953125, | |
| "logps/chosen": -2944.0, | |
| "logps/rejected": -2416.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5252100840336135, | |
| "grad_norm": 8.777510895642173, | |
| "learning_rate": 2.719920461452991e-07, | |
| "logits/chosen": -1.6484375, | |
| "logits/rejected": -1.6484375, | |
| "logps/chosen": -2624.0, | |
| "logps/rejected": -2256.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5252100840336135, | |
| "eval_logits/chosen": -1.703125, | |
| "eval_logits/rejected": -1.703125, | |
| "eval_logps/chosen": -2640.0, | |
| "eval_logps/rejected": -2352.0, | |
| "eval_loss": 0.69140625, | |
| "eval_rewards/accuracies": 0.0, | |
| "eval_rewards/chosen": 0.0, | |
| "eval_rewards/margins": 0.0, | |
| "eval_rewards/rejected": 0.0, | |
| "eval_runtime": 102.4791, | |
| "eval_samples_per_second": 19.38, | |
| "eval_steps_per_second": 0.312, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5462184873949579, | |
| "grad_norm": 10.81202800911272, | |
| "learning_rate": 2.536699530523291e-07, | |
| "logits/chosen": -1.71875, | |
| "logits/rejected": -1.71875, | |
| "logps/chosen": -3072.0, | |
| "logps/rejected": -2480.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5672268907563025, | |
| "grad_norm": 7.1452276225234295, | |
| "learning_rate": 2.353280958287195e-07, | |
| "logits/chosen": -1.6953125, | |
| "logits/rejected": -1.703125, | |
| "logps/chosen": -2832.0, | |
| "logps/rejected": -2592.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5882352941176471, | |
| "grad_norm": 16.243346282591897, | |
| "learning_rate": 2.1706525253979534e-07, | |
| "logits/chosen": -1.5859375, | |
| "logits/rejected": -1.59375, | |
| "logps/chosen": -2880.0, | |
| "logps/rejected": -2464.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6092436974789915, | |
| "grad_norm": 12.28591647805525, | |
| "learning_rate": 1.9897977572998152e-07, | |
| "logits/chosen": -1.5234375, | |
| "logits/rejected": -1.5234375, | |
| "logps/chosen": -2960.0, | |
| "logps/rejected": -2624.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6302521008403361, | |
| "grad_norm": 10.712221070111289, | |
| "learning_rate": 1.8116906275593507e-07, | |
| "logits/chosen": -1.65625, | |
| "logits/rejected": -1.6484375, | |
| "logps/chosen": -2976.0, | |
| "logps/rejected": -2480.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6302521008403361, | |
| "eval_logits/chosen": -1.671875, | |
| "eval_logits/rejected": -1.6796875, | |
| "eval_logps/chosen": -2624.0, | |
| "eval_logps/rejected": -2336.0, | |
| "eval_loss": 0.69140625, | |
| "eval_rewards/accuracies": 0.0, | |
| "eval_rewards/chosen": 0.0, | |
| "eval_rewards/margins": 0.0, | |
| "eval_rewards/rejected": 0.0, | |
| "eval_runtime": 103.0042, | |
| "eval_samples_per_second": 19.281, | |
| "eval_steps_per_second": 0.311, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6512605042016807, | |
| "grad_norm": 11.873878441097949, | |
| "learning_rate": 1.6372903126373684e-07, | |
| "logits/chosen": -1.6953125, | |
| "logits/rejected": -1.6875, | |
| "logps/chosen": -2864.0, | |
| "logps/rejected": -2784.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6722689075630253, | |
| "grad_norm": 16.36161277353074, | |
| "learning_rate": 1.4675360263490295e-07, | |
| "logits/chosen": -1.65625, | |
| "logits/rejected": -1.65625, | |
| "logps/chosen": -2608.0, | |
| "logps/rejected": -2560.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6932773109243697, | |
| "grad_norm": 16.718448992720084, | |
| "learning_rate": 1.3033419618306018e-07, | |
| "logits/chosen": -1.578125, | |
| "logits/rejected": -1.5703125, | |
| "logps/chosen": -2768.0, | |
| "logps/rejected": -2224.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 17.049047455790024, | |
| "learning_rate": 1.1455923682523475e-07, | |
| "logits/chosen": -1.59375, | |
| "logits/rejected": -1.59375, | |
| "logps/chosen": -2752.0, | |
| "logps/rejected": -2256.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7352941176470589, | |
| "grad_norm": 8.628681230007127, | |
| "learning_rate": 9.951367887913573e-08, | |
| "logits/chosen": -1.6015625, | |
| "logits/rejected": -1.609375, | |
| "logps/chosen": -3328.0, | |
| "logps/rejected": -2880.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7352941176470589, | |
| "eval_logits/chosen": -1.6015625, | |
| "eval_logits/rejected": -1.6015625, | |
| "eval_logps/chosen": -2640.0, | |
| "eval_logps/rejected": -2352.0, | |
| "eval_loss": 0.69140625, | |
| "eval_rewards/accuracies": 0.0, | |
| "eval_rewards/chosen": 0.0, | |
| "eval_rewards/margins": 0.0, | |
| "eval_rewards/rejected": 0.0, | |
| "eval_runtime": 102.7271, | |
| "eval_samples_per_second": 19.333, | |
| "eval_steps_per_second": 0.312, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7563025210084033, | |
| "grad_norm": 11.363934563365467, | |
| "learning_rate": 8.527854855097224e-08, | |
| "logits/chosen": -1.6171875, | |
| "logits/rejected": -1.6015625, | |
| "logps/chosen": -2688.0, | |
| "logps/rejected": -2352.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7773109243697479, | |
| "grad_norm": 14.871357784427335, | |
| "learning_rate": 7.193050757768431e-08, | |
| "logits/chosen": -1.578125, | |
| "logits/rejected": -1.5703125, | |
| "logps/chosen": -2720.0, | |
| "logps/rejected": -2208.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7983193277310925, | |
| "grad_norm": 12.038701774247873, | |
| "learning_rate": 5.9541440373546445e-08, | |
| "logits/chosen": -1.625, | |
| "logits/rejected": -1.625, | |
| "logps/chosen": -2832.0, | |
| "logps/rejected": -2320.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.819327731092437, | |
| "grad_norm": 14.978538565892313, | |
| "learning_rate": 4.8178066904518894e-08, | |
| "logits/chosen": -1.6015625, | |
| "logits/rejected": -1.6015625, | |
| "logps/chosen": -2592.0, | |
| "logps/rejected": -2288.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8403361344537815, | |
| "grad_norm": 8.293782679846284, | |
| "learning_rate": 3.790158337517127e-08, | |
| "logits/chosen": -1.578125, | |
| "logits/rejected": -1.578125, | |
| "logps/chosen": -2944.0, | |
| "logps/rejected": -2736.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8403361344537815, | |
| "eval_logits/chosen": -1.578125, | |
| "eval_logits/rejected": -1.578125, | |
| "eval_logps/chosen": -2640.0, | |
| "eval_logps/rejected": -2368.0, | |
| "eval_loss": 0.69140625, | |
| "eval_rewards/accuracies": 0.0, | |
| "eval_rewards/chosen": 0.0, | |
| "eval_rewards/margins": 0.0, | |
| "eval_rewards/rejected": 0.0, | |
| "eval_runtime": 102.8311, | |
| "eval_samples_per_second": 19.313, | |
| "eval_steps_per_second": 0.311, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8613445378151261, | |
| "grad_norm": 17.01214709788288, | |
| "learning_rate": 2.876733266321765e-08, | |
| "logits/chosen": -1.5625, | |
| "logits/rejected": -1.5546875, | |
| "logps/chosen": -3056.0, | |
| "logps/rejected": -2560.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8823529411764706, | |
| "grad_norm": 13.078800555411027, | |
| "learning_rate": 2.0824506276503894e-08, | |
| "logits/chosen": -1.5703125, | |
| "logits/rejected": -1.5625, | |
| "logps/chosen": -2912.0, | |
| "logps/rejected": -2592.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9033613445378151, | |
| "grad_norm": 9.301856974075692, | |
| "learning_rate": 1.4115879437524043e-08, | |
| "logits/chosen": -1.578125, | |
| "logits/rejected": -1.578125, | |
| "logps/chosen": -2736.0, | |
| "logps/rejected": -2432.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9243697478991597, | |
| "grad_norm": 9.51276981883413, | |
| "learning_rate": 8.677580722139671e-09, | |
| "logits/chosen": -1.6015625, | |
| "logits/rejected": -1.6015625, | |
| "logps/chosen": -2752.0, | |
| "logps/rejected": -2432.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9453781512605042, | |
| "grad_norm": 15.942237520125405, | |
| "learning_rate": 4.538897493087112e-09, | |
| "logits/chosen": -1.5625, | |
| "logits/rejected": -1.5625, | |
| "logps/chosen": -2816.0, | |
| "logps/rejected": -2336.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9453781512605042, | |
| "eval_logits/chosen": -1.578125, | |
| "eval_logits/rejected": -1.578125, | |
| "eval_logps/chosen": -2624.0, | |
| "eval_logps/rejected": -2352.0, | |
| "eval_loss": 0.69140625, | |
| "eval_rewards/accuracies": 0.0, | |
| "eval_rewards/chosen": 0.0, | |
| "eval_rewards/margins": 0.0, | |
| "eval_rewards/rejected": 0.0, | |
| "eval_runtime": 101.63, | |
| "eval_samples_per_second": 19.541, | |
| "eval_steps_per_second": 0.315, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9663865546218487, | |
| "grad_norm": 13.30372792679492, | |
| "learning_rate": 1.722118176089915e-09, | |
| "logits/chosen": -1.59375, | |
| "logits/rejected": -1.5859375, | |
| "logps/chosen": -2720.0, | |
| "logps/rejected": -2416.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9873949579831933, | |
| "grad_norm": 8.647302168816173, | |
| "learning_rate": 2.424122279805485e-10, | |
| "logits/chosen": -1.5703125, | |
| "logits/rejected": -1.5625, | |
| "logps/chosen": -2672.0, | |
| "logps/rejected": -2320.0, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 476, | |
| "total_flos": 0.0, | |
| "train_loss": 0.69140625, | |
| "train_runtime": 12796.3799, | |
| "train_samples_per_second": 4.76, | |
| "train_steps_per_second": 0.037 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 476, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |