lmms-lab

/

llava-next-72b

like 13

Follow

LMMs-Lab 24

Qwen1.5-72B need more memory for kv cache, it's not a friendly-to-deploy model. May Qwen1.5-32B is suitable for most situation for deployment.