jtmer
jtmer
属于是知识点整理了
I got the same error. I only have 1 A800 gpu card. When I just run the benchmark command: ``` python -m sglang.bench_latency --model-path /path_to_my_local_model/llama3_8B_chat_self_play_wenxin --batch 32 --input-len 256 --output-len...
In addition, `pip install -r requirements/requirements.txt` will download pytorch for cuda12, so I uninstalled it and installed pytorch for cuda11.8 after running this command
> Hi, What is your nvcc version? 我把镜像改了，问题解决了。但是解决以后遇到了新问题,,我放弃了