| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 0.951048951048951, |
| "eval_steps": 500, |
| "global_step": 17, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.055944055944055944, |
| "grad_norm": 112.59270477294922, |
| "learning_rate": 1.8823529411764705e-06, |
| "logits/chosen": 1.0625, |
| "logits/rejected": 1.125, |
| "logps/chosen": -17.75, |
| "logps/rejected": -15.75, |
| "loss": 0.691, |
| "rewards/accuracies": 0.03125, |
| "rewards/chosen": 0.000782012939453125, |
| "rewards/margins": 0.000782012939453125, |
| "rewards/rejected": 0.0, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.11188811188811189, |
| "grad_norm": 112.73638153076172, |
| "learning_rate": 1.764705882352941e-06, |
| "logits/chosen": 1.046875, |
| "logits/rejected": 1.09375, |
| "logps/chosen": -17.625, |
| "logps/rejected": -15.9375, |
| "loss": 0.691, |
| "rewards/accuracies": 0.46875, |
| "rewards/chosen": 0.0009765625, |
| "rewards/margins": 0.0033111572265625, |
| "rewards/rejected": -0.0023345947265625, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.16783216783216784, |
| "grad_norm": 108.02386474609375, |
| "learning_rate": 1.6470588235294116e-06, |
| "logits/chosen": 1.03125, |
| "logits/rejected": 1.1015625, |
| "logps/chosen": -17.5, |
| "logps/rejected": -16.0, |
| "loss": 0.6902, |
| "rewards/accuracies": 0.40625, |
| "rewards/chosen": 0.00921630859375, |
| "rewards/margins": 0.00567626953125, |
| "rewards/rejected": 0.0035247802734375, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.22377622377622378, |
| "grad_norm": 96.84840393066406, |
| "learning_rate": 1.5294117647058821e-06, |
| "logits/chosen": 1.078125, |
| "logits/rejected": 1.125, |
| "logps/chosen": -17.25, |
| "logps/rejected": -15.8125, |
| "loss": 0.6899, |
| "rewards/accuracies": 0.53125, |
| "rewards/chosen": 0.00921630859375, |
| "rewards/margins": 0.00567626953125, |
| "rewards/rejected": 0.0035247802734375, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.27972027972027974, |
| "grad_norm": 104.06653594970703, |
| "learning_rate": 1.411764705882353e-06, |
| "logits/chosen": 1.0859375, |
| "logits/rejected": 1.1328125, |
| "logps/chosen": -17.375, |
| "logps/rejected": -15.875, |
| "loss": 0.6913, |
| "rewards/accuracies": 0.4375, |
| "rewards/chosen": 0.0113525390625, |
| "rewards/margins": 0.0035247802734375, |
| "rewards/rejected": 0.0078125, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.3356643356643357, |
| "grad_norm": 103.72679901123047, |
| "learning_rate": 1.2941176470588235e-06, |
| "logits/chosen": 1.0546875, |
| "logits/rejected": 1.1015625, |
| "logps/chosen": -16.75, |
| "logps/rejected": -16.0, |
| "loss": 0.6918, |
| "rewards/accuracies": 0.4375, |
| "rewards/chosen": 0.015625, |
| "rewards/margins": 0.0025482177734375, |
| "rewards/rejected": 0.01312255859375, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.3916083916083916, |
| "grad_norm": 91.84857177734375, |
| "learning_rate": 1.176470588235294e-06, |
| "logits/chosen": 1.09375, |
| "logits/rejected": 1.109375, |
| "logps/chosen": -16.5, |
| "logps/rejected": -15.8125, |
| "loss": 0.6897, |
| "rewards/accuracies": 0.59375, |
| "rewards/chosen": 0.0208740234375, |
| "rewards/margins": 0.007049560546875, |
| "rewards/rejected": 0.01385498046875, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.44755244755244755, |
| "grad_norm": 95.65518188476562, |
| "learning_rate": 1.0588235294117646e-06, |
| "logits/chosen": 1.0234375, |
| "logits/rejected": 1.0859375, |
| "logps/chosen": -17.5, |
| "logps/rejected": -15.6875, |
| "loss": 0.689, |
| "rewards/accuracies": 0.53125, |
| "rewards/chosen": 0.0263671875, |
| "rewards/margins": 0.00836181640625, |
| "rewards/rejected": 0.0179443359375, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.5034965034965035, |
| "grad_norm": 96.71415710449219, |
| "learning_rate": 9.411764705882352e-07, |
| "logits/chosen": 1.078125, |
| "logits/rejected": 1.109375, |
| "logps/chosen": -16.75, |
| "logps/rejected": -15.9375, |
| "loss": 0.6836, |
| "rewards/accuracies": 0.59375, |
| "rewards/chosen": 0.03125, |
| "rewards/margins": 0.01953125, |
| "rewards/rejected": 0.01171875, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.5594405594405595, |
| "grad_norm": 101.46381378173828, |
| "learning_rate": 8.235294117647058e-07, |
| "logits/chosen": 1.046875, |
| "logits/rejected": 1.09375, |
| "logps/chosen": -16.875, |
| "logps/rejected": -15.625, |
| "loss": 0.6866, |
| "rewards/accuracies": 0.625, |
| "rewards/chosen": 0.032470703125, |
| "rewards/margins": 0.0133056640625, |
| "rewards/rejected": 0.0191650390625, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.6153846153846154, |
| "grad_norm": 102.90161895751953, |
| "learning_rate": 7.058823529411765e-07, |
| "logits/chosen": 1.0859375, |
| "logits/rejected": 1.1171875, |
| "logps/chosen": -16.875, |
| "logps/rejected": -15.8125, |
| "loss": 0.6827, |
| "rewards/accuracies": 0.65625, |
| "rewards/chosen": 0.0390625, |
| "rewards/margins": 0.0208740234375, |
| "rewards/rejected": 0.0181884765625, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.6713286713286714, |
| "grad_norm": 97.20207214355469, |
| "learning_rate": 5.88235294117647e-07, |
| "logits/chosen": 1.078125, |
| "logits/rejected": 1.09375, |
| "logps/chosen": -16.5, |
| "logps/rejected": -15.9375, |
| "loss": 0.691, |
| "rewards/accuracies": 0.53125, |
| "rewards/chosen": 0.0213623046875, |
| "rewards/margins": 0.004730224609375, |
| "rewards/rejected": 0.0166015625, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.7272727272727273, |
| "grad_norm": 88.96224975585938, |
| "learning_rate": 4.705882352941176e-07, |
| "logits/chosen": 1.0859375, |
| "logits/rejected": 1.125, |
| "logps/chosen": -16.875, |
| "logps/rejected": -15.625, |
| "loss": 0.6809, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": 0.03955078125, |
| "rewards/margins": 0.025146484375, |
| "rewards/rejected": 0.01446533203125, |
| "step": 13 |
| }, |
| { |
| "epoch": 0.7832167832167832, |
| "grad_norm": 85.17412567138672, |
| "learning_rate": 3.529411764705882e-07, |
| "logits/chosen": 1.0859375, |
| "logits/rejected": 1.1171875, |
| "logps/chosen": -16.625, |
| "logps/rejected": -15.8125, |
| "loss": 0.6852, |
| "rewards/accuracies": 0.5625, |
| "rewards/chosen": 0.0380859375, |
| "rewards/margins": 0.0164794921875, |
| "rewards/rejected": 0.021728515625, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.8391608391608392, |
| "grad_norm": 99.89457702636719, |
| "learning_rate": 2.352941176470588e-07, |
| "logits/chosen": 1.0703125, |
| "logits/rejected": 1.1171875, |
| "logps/chosen": -17.25, |
| "logps/rejected": -15.75, |
| "loss": 0.678, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": 0.048095703125, |
| "rewards/margins": 0.03125, |
| "rewards/rejected": 0.016845703125, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.8951048951048951, |
| "grad_norm": 100.7333984375, |
| "learning_rate": 1.176470588235294e-07, |
| "logits/chosen": 1.0703125, |
| "logits/rejected": 1.1171875, |
| "logps/chosen": -17.25, |
| "logps/rejected": -15.75, |
| "loss": 0.6777, |
| "rewards/accuracies": 0.59375, |
| "rewards/chosen": 0.060546875, |
| "rewards/margins": 0.0322265625, |
| "rewards/rejected": 0.0283203125, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.951048951048951, |
| "grad_norm": 92.44598388671875, |
| "learning_rate": 0.0, |
| "logits/chosen": 1.0546875, |
| "logits/rejected": 1.09375, |
| "logps/chosen": -16.875, |
| "logps/rejected": -16.0, |
| "loss": 0.683, |
| "rewards/accuracies": 0.5625, |
| "rewards/chosen": 0.043701171875, |
| "rewards/margins": 0.021240234375, |
| "rewards/rejected": 0.0224609375, |
| "step": 17 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 17, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 1, |
| "save_steps": 500, |
| "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": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
