error with most recent vLLM

by elbiot - opened 5 days ago

5 days ago

I ran vllm serve models/Mistral-Small-3.1-24B-Instruct-2503-MAX-NEO-D_AU-Q4_K_S-imat.gguf --tokenizer DavidAU/Mistral-Small-3.1-24B-Instruct-2503-MAX-NEO-Imatrix-GGUF as per https://docs.vllm.ai/en/latest/features/quantization/gguf.html

The error I get is
ValueError: Unrecognized model in DavidAU/Mistral-Small-3.1-24B-Instruct-2503-MAX-NEO-Imatrix-GGUF. Should have a model_type key in its config.json

DavidAU

Owner 5 days ago

Hey;

--tokenizer

Have this point at the repo @mistral ; as the "GGUF" repo does not have tokenizer or other .json files.

elbiot

5 days ago

I can get past this point by using "mistralai/Mistral-Small-3.1-24B-Instruct-2503" as the tokenizer and hf_config_path. But then I get

RuntimeError: Unknown gguf model_type: pixtral

DavidAU

Owner 5 days ago

my bad; mine is text only.

Use:
mrfakename/mistral-small-3.1-24b-instruct-2503-hf

(source repo for the ggufs I made)

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

Your need to confirm your account before you can post a new comment.

· Sign up or log in to comment