Closed noahbald closed 9 months ago
Haven't got this working yet, but adding here for visibility
There was some crashing occurring when providing a model as a url but the tokeniser repo as one for hugging face in the vscode config. The fix for that has been pushed to https://github.com/huggingface/llm-ls/pull/40
The completion requests seem to be working after changing the casing for params to camel-case.
Adding adaptors here, as per https://github.com/huggingface/llm-ls/pull/40
This should close https://github.com/huggingface/llm-vscode/issues/60 and https://github.com/huggingface/llm-vscode/issues/108