shuffling the dataset

Hi, thanks for your contribution, but I guess in [toxicity/train_dpo/pplm_dataset.py](https://github.com/ajyl/dpo_toxic/blob/0ab76996b2fa5e725ad00985de9e284a1121c31e/toxicity/train_dpo/pplm_dataset.py#L57), the code `random.shuffle(file_data)` should be `random.shuffle(data)` instead?