`data_utils.tokenization_gpt2.GPT2Tokenizer ` is different from `transformers.CpmTokenizer`

[ShaneTian]

See https://huggingface.co/TsinghuaAI/CPM-Generate/discussions/1

For LM fine-tuning or generation, how do I prepare my input data?

• [token_id_1, token_id_2, ..., eod_token_id], where eod_token_id is the id of <eod> token in transformers.CpmTokenizer
• [token_id_1, token_id_2, ..., eos_token_id], where eos_token_id is the id of </s> token in transformers.CpmTokenizer
• [token_id_1, token_id_2, ..., eos_token_id], where eos_token_id is the id of <|endoftext|> token in transformers.GPT2Tokenizer
• [token_id_1, token_id_2, ..., sep_token_id, cls_token_id], just call CpmTokenizer