| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 1000, | |
| "global_step": 229, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.004366812227074236, | |
| "grad_norm": 0.4543283495867887, | |
| "learning_rate": 2.173913043478261e-07, | |
| "logits/chosen": -0.9449982047080994, | |
| "logits/rejected": -0.915056049823761, | |
| "logps/chosen": -328.08038330078125, | |
| "logps/rejected": -274.1044921875, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.043668122270742356, | |
| "grad_norm": 0.47008608128274937, | |
| "learning_rate": 2.173913043478261e-06, | |
| "logits/chosen": -0.9651342034339905, | |
| "logits/rejected": -1.0432785749435425, | |
| "logps/chosen": -326.21295166015625, | |
| "logps/rejected": -277.12603759765625, | |
| "loss": 0.6928, | |
| "rewards/accuracies": 0.4583333432674408, | |
| "rewards/chosen": 0.000499483838211745, | |
| "rewards/margins": 0.00042585484334267676, | |
| "rewards/rejected": 7.362888572970405e-05, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.08733624454148471, | |
| "grad_norm": 0.5077455364034207, | |
| "learning_rate": 4.347826086956522e-06, | |
| "logits/chosen": -1.0618306398391724, | |
| "logits/rejected": -1.009238600730896, | |
| "logps/chosen": -334.39141845703125, | |
| "logps/rejected": -346.700927734375, | |
| "loss": 0.6882, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.0003057918802369386, | |
| "rewards/margins": 0.00985435489565134, | |
| "rewards/rejected": -0.010160146281123161, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.13100436681222707, | |
| "grad_norm": 0.5720215285315815, | |
| "learning_rate": 4.985768230048011e-06, | |
| "logits/chosen": -1.0507558584213257, | |
| "logits/rejected": -1.075430989265442, | |
| "logps/chosen": -308.62359619140625, | |
| "logps/rejected": -301.9061279296875, | |
| "loss": 0.6651, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": 0.0059205591678619385, | |
| "rewards/margins": 0.05983991548418999, | |
| "rewards/rejected": -0.053919363766908646, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.17467248908296942, | |
| "grad_norm": 0.8551226198033497, | |
| "learning_rate": 4.9164513914144005e-06, | |
| "logits/chosen": -1.1152777671813965, | |
| "logits/rejected": -1.1143274307250977, | |
| "logps/chosen": -288.18585205078125, | |
| "logps/rejected": -310.5125732421875, | |
| "loss": 0.5986, | |
| "rewards/accuracies": 0.9125000238418579, | |
| "rewards/chosen": -0.003953046165406704, | |
| "rewards/margins": 0.21825852990150452, | |
| "rewards/rejected": -0.2222115695476532, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2183406113537118, | |
| "grad_norm": 1.0672212168131259, | |
| "learning_rate": 4.791042480696179e-06, | |
| "logits/chosen": -1.1128385066986084, | |
| "logits/rejected": -1.0449755191802979, | |
| "logps/chosen": -317.44451904296875, | |
| "logps/rejected": -390.5851135253906, | |
| "loss": 0.4535, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": 0.03481646627187729, | |
| "rewards/margins": 0.6241441965103149, | |
| "rewards/rejected": -0.5893277525901794, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.26200873362445415, | |
| "grad_norm": 0.6925238532589472, | |
| "learning_rate": 4.612452562309975e-06, | |
| "logits/chosen": -1.0924533605575562, | |
| "logits/rejected": -1.0897819995880127, | |
| "logps/chosen": -312.33837890625, | |
| "logps/rejected": -432.6812438964844, | |
| "loss": 0.2332, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": 0.08799274265766144, | |
| "rewards/margins": 1.521544098854065, | |
| "rewards/rejected": -1.4335514307022095, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.3056768558951965, | |
| "grad_norm": 0.8913228040925837, | |
| "learning_rate": 4.384827169085993e-06, | |
| "logits/chosen": -1.0938880443572998, | |
| "logits/rejected": -1.1012872457504272, | |
| "logps/chosen": -322.24139404296875, | |
| "logps/rejected": -572.3203125, | |
| "loss": 0.114, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.0598033182322979, | |
| "rewards/margins": 2.8274989128112793, | |
| "rewards/rejected": -2.8873023986816406, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.34934497816593885, | |
| "grad_norm": 1.025805090974843, | |
| "learning_rate": 4.1134500737541026e-06, | |
| "logits/chosen": -1.0882785320281982, | |
| "logits/rejected": -0.9281471371650696, | |
| "logps/chosen": -279.4626770019531, | |
| "logps/rejected": -729.4642333984375, | |
| "loss": 0.0632, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.04784034937620163, | |
| "rewards/margins": 4.184046745300293, | |
| "rewards/rejected": -4.13620662689209, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.3930131004366812, | |
| "grad_norm": 0.7021281871664607, | |
| "learning_rate": 3.8046206389447916e-06, | |
| "logits/chosen": -0.9880226850509644, | |
| "logits/rejected": -0.9430360794067383, | |
| "logps/chosen": -297.018310546875, | |
| "logps/rejected": -715.1465454101562, | |
| "loss": 0.0491, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.10021056234836578, | |
| "rewards/margins": 4.483721733093262, | |
| "rewards/rejected": -4.383510589599609, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.4366812227074236, | |
| "grad_norm": 0.586209743573467, | |
| "learning_rate": 3.4655075927279576e-06, | |
| "logits/chosen": -0.9809297323226929, | |
| "logits/rejected": -0.8725347518920898, | |
| "logps/chosen": -311.753662109375, | |
| "logps/rejected": -830.88671875, | |
| "loss": 0.0353, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.36457541584968567, | |
| "rewards/margins": 4.913618564605713, | |
| "rewards/rejected": -5.278193950653076, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.48034934497816595, | |
| "grad_norm": 0.3979818436750899, | |
| "learning_rate": 3.1039826239365754e-06, | |
| "logits/chosen": -0.9446805119514465, | |
| "logits/rejected": -0.8940761685371399, | |
| "logps/chosen": -431.5677795410156, | |
| "logps/rejected": -925.68017578125, | |
| "loss": 0.0288, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -1.047907829284668, | |
| "rewards/margins": 5.1847357749938965, | |
| "rewards/rejected": -6.232643127441406, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.5240174672489083, | |
| "grad_norm": 0.44864039181837617, | |
| "learning_rate": 2.7284376599571776e-06, | |
| "logits/chosen": -0.9638189077377319, | |
| "logits/rejected": -0.8882731199264526, | |
| "logps/chosen": -392.0179748535156, | |
| "logps/rejected": -961.5611572265625, | |
| "loss": 0.0207, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.6613569259643555, | |
| "rewards/margins": 6.055957794189453, | |
| "rewards/rejected": -6.71731424331665, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.5676855895196506, | |
| "grad_norm": 0.15662373138721433, | |
| "learning_rate": 2.3475900684411027e-06, | |
| "logits/chosen": -0.97578364610672, | |
| "logits/rejected": -0.8818238973617554, | |
| "logps/chosen": -417.825439453125, | |
| "logps/rejected": -989.6531372070312, | |
| "loss": 0.0229, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.840344250202179, | |
| "rewards/margins": 5.920441627502441, | |
| "rewards/rejected": -6.760786533355713, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.611353711790393, | |
| "grad_norm": 1.0644930727731667, | |
| "learning_rate": 1.970280304707447e-06, | |
| "logits/chosen": -1.0020447969436646, | |
| "logits/rejected": -0.8326355814933777, | |
| "logps/chosen": -390.6119384765625, | |
| "logps/rejected": -1086.9949951171875, | |
| "loss": 0.012, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.8912075757980347, | |
| "rewards/margins": 6.5178542137146, | |
| "rewards/rejected": -7.40906286239624, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.6550218340611353, | |
| "grad_norm": 0.33838570745300933, | |
| "learning_rate": 1.6052667019636462e-06, | |
| "logits/chosen": -0.9777904748916626, | |
| "logits/rejected": -0.9008253216743469, | |
| "logps/chosen": -421.56658935546875, | |
| "logps/rejected": -1008.5904541015625, | |
| "loss": 0.0166, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.0421282052993774, | |
| "rewards/margins": 6.398242950439453, | |
| "rewards/rejected": -7.440371513366699, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.6986899563318777, | |
| "grad_norm": 0.1209081830310832, | |
| "learning_rate": 1.261022167792161e-06, | |
| "logits/chosen": -1.0033915042877197, | |
| "logits/rejected": -0.8852731585502625, | |
| "logps/chosen": -402.15692138671875, | |
| "logps/rejected": -1126.320556640625, | |
| "loss": 0.011, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.8826059103012085, | |
| "rewards/margins": 6.96124267578125, | |
| "rewards/rejected": -7.843849182128906, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.74235807860262, | |
| "grad_norm": 0.9669967772246254, | |
| "learning_rate": 9.455375061024319e-07, | |
| "logits/chosen": -0.9981080889701843, | |
| "logits/rejected": -0.914253830909729, | |
| "logps/chosen": -445.781982421875, | |
| "logps/rejected": -1072.5477294921875, | |
| "loss": 0.0161, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -1.2199878692626953, | |
| "rewards/margins": 6.577937126159668, | |
| "rewards/rejected": -7.797924995422363, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.7860262008733624, | |
| "grad_norm": 0.09826043440610506, | |
| "learning_rate": 6.661359299530626e-07, | |
| "logits/chosen": -0.9915900230407715, | |
| "logits/rejected": -0.8251992464065552, | |
| "logps/chosen": -402.2227478027344, | |
| "logps/rejected": -1161.5181884765625, | |
| "loss": 0.0089, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.1185193061828613, | |
| "rewards/margins": 7.358689308166504, | |
| "rewards/rejected": -8.477209091186523, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.8296943231441049, | |
| "grad_norm": 0.3084190328632994, | |
| "learning_rate": 4.293030708802834e-07, | |
| "logits/chosen": -1.0055569410324097, | |
| "logits/rejected": -0.8623663187026978, | |
| "logps/chosen": -418.1572265625, | |
| "logps/rejected": -1129.2655029296875, | |
| "loss": 0.0108, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.1186723709106445, | |
| "rewards/margins": 6.933996677398682, | |
| "rewards/rejected": -8.0526704788208, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.8733624454148472, | |
| "grad_norm": 0.10794356987220223, | |
| "learning_rate": 2.405364306547955e-07, | |
| "logits/chosen": -0.8744913339614868, | |
| "logits/rejected": -0.7203727960586548, | |
| "logps/chosen": -407.0746765136719, | |
| "logps/rejected": -1108.8079833984375, | |
| "loss": 0.0118, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.0695985555648804, | |
| "rewards/margins": 7.07146692276001, | |
| "rewards/rejected": -8.14106559753418, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.9170305676855895, | |
| "grad_norm": 1.7933618946506555, | |
| "learning_rate": 1.0421777008019663e-07, | |
| "logits/chosen": -1.047523856163025, | |
| "logits/rejected": -0.8576782941818237, | |
| "logps/chosen": -400.34527587890625, | |
| "logps/rejected": -1216.9423828125, | |
| "loss": 0.0123, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.9318978190422058, | |
| "rewards/margins": 7.647104740142822, | |
| "rewards/rejected": -8.579002380371094, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.9606986899563319, | |
| "grad_norm": 1.0019715180278645, | |
| "learning_rate": 2.351139701825267e-08, | |
| "logits/chosen": -0.9511087536811829, | |
| "logits/rejected": -0.8330629467964172, | |
| "logps/chosen": -415.8382873535156, | |
| "logps/rejected": -1095.948974609375, | |
| "loss": 0.0086, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.0705822706222534, | |
| "rewards/margins": 7.189767360687256, | |
| "rewards/rejected": -8.26034927368164, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 229, | |
| "total_flos": 0.0, | |
| "train_loss": 0.16486623261823405, | |
| "train_runtime": 2148.0634, | |
| "train_samples_per_second": 6.816, | |
| "train_steps_per_second": 0.107 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 229, | |
| "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": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |