llm
llm copied to clipboard
Support for `GPTBigCode`
What we need is local cpp for llm "HuggingFaceH4/starchat-alpha A". It is much better than any other open source chat models at coding. And it does very well at chatting too. This is better than Vicuna.
https://huggingface.co/docs/transformers/model_doc/gpt_bigcode
https://huggingface.co/HuggingFaceH4/starchat-alpha/blob/main/config.json
Looks to be a GPT-2 variant.
This should be solved with the last GPT-2 changes.