Need pad_token_id?

[ShiJiawenwen]

Thanks for your great work! I know openchat-3.5 use the padding-free training strategy. May I ask if it is necessary to add pad_token during the inference stage?

The example code is as follows:

tokenizer = AutoTokenizer.from_pretrained("openchat/openchat_3.5") tokenizer.pad_token="<|pad_0|>"

When I use this model for generation get the warning message: "The attention mask and the pad token id were not set. As a consequence, you may observe unexpected behavior. Please pass your input's attention_mask to obtain reliable results. Setting pad_token_id to eos_token_id:32000 for open-end generation."

[dependabot[bot]]

Looks like naive-ui is up-to-date now, so this is no longer needed.