Open tinudin opened 2 months ago
Hello, thanks for your kind words!
Which inference script are you running? It seems to be a issue related to huggfingface, could you try commands like "python -m lrm.inferrer --export_video --resume ./checkpoints/vfusion3dckpt" which is not related to hf?
python -m lrm.inferrer --export_mesh --resume ./checkpoints/vfusion3dckpt --source_path '.\assets\40_prompt_images\' This is what i tested.
ok i got it, hf might be used somewhere when creating the model.
maybe your connection to hf is not stable? Could you try typing "ping huggingface.co" command to see if you can establish a connection to hf.
that is not working. is it possible to run with only git?
this part used hf functions: https://github.com/facebookresearch/vfusion3d/blob/main/lrm/models/encoders/dino_wrapper2.py
so far I don't have any other versions that can skip hf.
Hi, thanks for you work. Tried to implement it, but unfortunatly when running the inference script i recieve the following error:
requests.exceptions.SSLError: (MaxRetryError("HTTPSConnectionPool(host='cdn-lfs.huggingface.co', port=443): Max retries exceeded with url:
The expected behaviour was: when downloading the model file from google Drive, this should be enough to run inference.