Cannot be demployed with SGLang
#2
by
zsytony
- opened
I cannot deploy this model with 32 H800 GPUs via SGLang, while it works for original deepseek-v3, could you give the detailed serving command either by vLLM or SGLang?
99 and 100 are missing
Thank you very much for pointing out the issue. We have already uploaded the two .safetensors that were accidentally missing (and verified that the ones on ModelScope are complete). If you have any other questions, feel free to ask!
ScienceOne-AI
changed discussion status to
closed