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

可以使用nuscenes mini-v1.0 进行训练吗 #45

Open
lzl-mt opened this issue Jan 8, 2025 · 3 comments
Open

可以使用nuscenes mini-v1.0 进行训练吗 #45

lzl-mt opened this issue Jan 8, 2025 · 3 comments

Comments

@lzl-mt
Copy link

lzl-mt commented Jan 8, 2025

训练报错 FileNotFoundError: NuScenesDataset: LoadBEVSegmentationS: [Errno 2] No such file or directory: '/cpfs01/shared/public/ADLab/datasets/nuscenes/maps/expansion/singapore-onenorth.json'

@jokester-zzz
Copy link
Collaborator

需要修改以下的数据集路径:

dataset_root_nuscenes: /cpfs01/shared/public/ADLab/datasets/nuscenes

mini数据集的量可能太少了,训练效果不会太好。

@lzl-mt
Copy link
Author

lzl-mt commented Jan 10, 2025

请问如果需要修改模型的结构应该在哪里修改,我看现在默认的head dimension是512,这个会不会有点大

@lzl-mt
Copy link
Author

lzl-mt commented Jan 13, 2025

third_party/diffusers/src/diffusers/models/vae.py 这里面encoder和decoder的attn_num_head_channels可以从默认的None改成128吗

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

2 participants