Module pixart-diffusion-system not found

[gsls229]

Hello, I saw the issue above, but after following the solution above, I still encountered this error. Below is the specific issue after running inference.py: Traceback (most recent call last): File "/root/CraftsMan/craftsman/init.py", line 29, in find module = importlib.import_module(module_string, package=None) File "/root/miniconda3/envs/CraftsMan/lib/python3.10/importlib/init.py", line 126, in import_module return _bootstrap._gcd_import(name[level:], package, level) File "", line 1047, in _gcd_import File "", line 981, in _sanity_check ValueError: Empty module name

During handling of the above exception, another exception occurred:

Traceback (most recent call last): File "/root/CraftsMan/inference.py", line 10, in pipeline = CraftsManPipeline.from_pretrained("craftsman3d/craftsman-v1-5", device="cuda:0", torch_dtype=torch.float32) File "/root/CraftsMan/craftsman/pipeline.py", line 63, in from_pretrained system = craftsman.find(cfg.system_type)(cfg.system) File "/root/CraftsMan/craftsman/init.py", line 32, in find raise ValueError(f"Module {name} not found!") ValueError: Module pixart-diffusion-system not found!

[baizhanxu]

Hi, same issue here. It looks like pixart-diffusion.py is missing from ./craftsman/systems. Not sure how to resolve this — any help would be great!

[wyysf-98]

Since we have updated the version of doravae, if you want to load the v1.5 model in the current version, 1）you need to modify the "pixart-diffusion-system" in the config to "shape-diffusion-system," 2）or use the previous commit: https://github.com/wyysf-98/CraftsMan3D/tree/CraftsMan-v1.5.

I will work on making the current version support the weight loading of the v1.5 model as soon as possible. I sincerely apologize for the inconvenience.

[baizhanxu]

Thank you so much for the clarification and the detailed instructions!

I really appreciate the quick support and the effort you’re putting into making the current version compatible with v1.5 weights. no worries at all about the inconvenience, and thanks again for your great work!