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

GLB to OBJ Conversion #18

Open
c0zaut opened this issue Dec 5, 2024 · 2 comments
Open

GLB to OBJ Conversion #18

c0zaut opened this issue Dec 5, 2024 · 2 comments

Comments

@c0zaut
Copy link

c0zaut commented Dec 5, 2024

Hi! First off, thank you for the amazing work that you all do!

I think I have a pretty close approximation of the Objaverse subset that you used (no animation, faces < 500, only glbs with descriptions,) but am having some issues converting the glb files into obj strings.

I can export the GLB into obj files, but would really appreciate some guidance on how you calculated the vertices so they did not need any texture or normal data, and as signed integers.

It also looks like maybe you capped vertex count, as well. If you did, could you share that limit, too?

Thank you, again!

@c0zaut c0zaut changed the title Feature Request - Training Data GLB to OBJ Conversion Dec 5, 2024
@thuwzy
Copy link
Collaborator

thuwzy commented Dec 7, 2024

Thank you for your interest in our work! We use the util functions in the repo of Pivotmesh for quantization of obj files.

@c0zaut
Copy link
Author

c0zaut commented Dec 7, 2024

THANK YOU!!!!! I will try out some tests later this weekend!

Can you also confirm the the max vertices, if any, of the objaverse dataset? That appears to be the only missing component. Once I have that, I can generate Parquet using pyarrow of the unified data. I plan on using it for static quantization with the Rockchip 3588 NPU SDK, RKLLM: https://github.com/airockchip/rknn-llm/blob/1a502afb52bddc49c2f4e9c0b025b7ab471467a9/rkllm-toolkit/examples/test.py#L28

My current implementation is here: https://github.com/c0zaut/RKLLM-Gradio

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