Integrate StableLM (#254)

[glerzing]

Description

Enables to use Stable_LM, and adds a demo. The tuned version of Stable_LM is a bit special because, like ChatGPT, there are special tokens <|SYSTEM|>, <|USER|> and <|ASSISTANT|>. It is specially tuned with instruction fine-tuning for chat. These special tokens are also end tokens, so this needs a special treatment in the generate function, in order to stop when StableLM outputs <|USER|>.

Fixes #254

Type of change

Please delete options that are not relevant.

• New feature (non-breaking change which adds functionality)
• This change requires a documentation update

Checklist:

• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes
• I have not rewritten tests relating to key interfaces which would affect backward compatibility

[ArthurConmy]

Cool stuff! Two minor changes to pad_token_id and some deletions and let's merge this!