Matt Deitke

Yes, see here: https://colab.research.google.com/drive/1NpR2MLs6ydXm4XWT_ovpM16CtzvGJVC7?usp=sharing Basically: ```python from objaverse.xl import get_annotations df = get_annotations() new_df = df[df["source"] != "sketchfab"] ```

yes - see the colab. it's released there :)

See https://huggingface.co/datasets/allenai/objaverse-xl/discussions/5 > Not sure on the storage size for the objects themselves. > For rendering, 12 renders for each object (plus the metadata and npy files) is about 1.75MB....

Ah, thanks will update! This comes from the old API before I refactored. Will update :)

Should be updated now :) https://github.com/allenai/objaverse-xl/commit/9a647bc844098776c86fca32270b826101194662

Might need to use a newer version of Python that supports f-strings. Maybe try with Python >3.8 :)

Thanks! It's not expected in the near term due to licensing being a bit harder.

I think the main thing that needs to be changed is the rendering engine should be changed from EEVEE to cycles. I believe this is a flag as part of...

If you’re rendering on a machine that doesn’t have a monitor, I believe you need to start an xserver to render on GPUs, which basically starts a virtual desktop. At...

Hi @whcpumpkin! The asset database is available [here](https://github.com/allenai/procthor/blob/main/procthor/databases/asset-database.json). I've also done a quick visualization of it [here](https://vigorous-lamarr-a0a9c0.netlify.app/). Note: the later website is not well maintained. It was just for my...