Add support for MPT-7B.

[digisomni]

https://huggingface.co/mosaicml/mpt-7b-storywriter

The model appears to be a great alternative to LLaMA from surface observations, particularly for commercial use. I expect many groups to build upon this foundation going forward if it does hold up.

Can someone perhaps look into adding support for this model?

[merrymercy]

Contribution is welcome and you can see this PR (https://github.com/lm-sys/FastChat/pull/1019) for example

[EwoutH]

https://huggingface.co/mosaicml/mpt-7b-storywriter

To be clear: In https://github.com/lm-sys/FastChat/pull/1111 the mpt-7b-chat model was added, not the untuned mpt-7b or the mpt-7b-instruct or mpt-7b-storywriter tuned models.