Longcat-Flash-Lite and "Ngram" Support

[TomLucidor]

ValueError: Model type longcat_flash_ngram not supported. for the model https://huggingface.co/mlx-community/LongCat-Flash-Lite-4bit when used with MLX-LM #433 (comment)

llama.cpp is WIP, vLLM has not made a move yet (and a lot of offloading discussion on how to best reduce memory usage) ggml-org/llama.cpp#19167

[M-I]

The Flash and flash lite are only similar in name. You need the latest git to run it. It has not yet been pushed to pypi.
uv pip install git+https://github.com/ml-explore/mlx-lm.git --upgrade or whatever tool you use to install packages.

[awni]

We will get a new release out soon. Closing this as resolved.