FP8
Collection
FP8 compressed models
•
6 items
•
Updated
python3 -m vllm.entrypoints.openai.api_server \
--host 0.0.0.0 --port 8000 \
--model miike-ai/devstral-fp8 \
--tokenizer mistralai/Devstral-Small-2505 \
--tokenizer-mode mistral \
--trust-remote-code
Base model
mistralai/Devstral-Small-2505