Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

nvidia-smi显示的显存占用和torch.cuda.max_memory_reserved显示的不一致 #162

Open
isky-cd opened this issue Jan 3, 2025 · 0 comments

Comments

@isky-cd
Copy link

isky-cd commented Jan 3, 2025

镜像是你们提供的cuda12的镜像
MODEL_BASE=/mnt/jfs-hdd/share/models/hunyuan-ckpts/HunyuanVideo torchrun --nproc_per_node=${num_gpus} sample_video.py
--model-base /mnt/jfs-hdd/share/models/hunyuan-ckpts/HunyuanVideo
--dit-weight /mnt/jfs-hdd/share/models/hunyuan-ckpts/HunyuanVideo/hunyuan-video-t2v-720p/transformers/mp_rank_00_model_states.pt
--video-size 1280 720
--video-length 129
--num-videos 1
--infer-steps 20
--prompt "A cat walks on the grass, realistic style."
--flow-reverse
--seed 42
--ulysses-degree 8
--ring-degree 1
--save-path ./results
我的运行脚本如上:
八卡:ulysses-degree 8 ring-degree 1
torch.cuda.max_memory_reserved显示的显存占用:
image
nvidia-smi显示的最大显存占用:
image
四卡:ulysses-degree 2 ring-degree 2
torch.cuda.max_memory_reserved显示的显存占用:
image
nvidia-smi显示的最大显存占用:
image

麻烦问题下为什么会出现这种情况吗?
image
还有我想请问一下上面的60G和45G的运行的是什么配置呢?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant