Open zmhsmart opened 2 months ago
Hi,
Seems like I recently introduced this bug by setting aitviewer to use GPU by default, I need to fix that.
In the meanwhile, you can fix this by setting this parameter to device: "cpu"
https://github.com/MarilynKeller/aitviewer-skel/blob/cd4e851e05b00901e9ea9b3201d09f565b0dedf1/aitviewer/aitvconfig.yaml#L18
Or, if you want to use the GPU,
Move poses, betas and trans to the GPU poses = poses.to("cuda:0") ...
pbc@wcs-Pre:~/project/inference/SKEL$ python examples/skel_poses.py --gender male
Traceback (most recent call last):
File "examples/skel_poses.py", line 140, in
I first added pose, betas, and trans to the gpu, but the same error was still displayed. Then I added the model skel to the gpu, but the graphics card memory overflowed. My graphics card is 3090, and the video memory is 24g. The error shown in the picture appeared. How should I solve it? Thank you for your reply
Hi, then set the config to CPU as I suggested. The sequence is simply too long for being loaded as meshes on your GPU, which is fine.
Hello, thank you for your excellent work. I tried to reproduce your work and got the following error. Can you help me solve this problem? Thank you very much. $ python examples/skel_poses.py --gender male Error processing line 1 of /home/pbc/miniconda3/envs/zxh/lib/python3.8/site-packages/psbody-mesh-nspkg.pth:
Traceback (most recent call last): File "/home/pbc/miniconda3/envs/zxh/lib/python3.8/site.py", line 169, in addpackage exec(line) File "", line 1, in
File "", line 553, in module_from_spec
AttributeError: 'NoneType' object has no attribute 'loader'
Remainder of file ignored Traceback (most recent call last): File "examples/skel_poses.py", line 126, in
skel_seq = SKELSequence(skel_layer=skel_model, betas=betas, poses_body=poses, poses_type='skel',
File "/home/pbc/project/inference/SKEL/aitviewer-skel/aitviewer/renderables/skel.py", line 144, in init
skel_output = self.fk()
File "/home/pbc/project/inference/SKEL/aitviewer-skel/aitviewer/renderables/skel.py", line 441, in fk
skel_output = self.skel_layer(poses=poses_body, betas=betas, trans=trans, poses_type=self.poses_type)
File "/home/pbc/miniconda3/envs/zxh/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "/home/pbc/project/inference/SKEL/skel/skel_model.py", line 312, in forward
v_shaped = skin_v0 + torch.matmul(shapedirs, betas).view(B, Ns, 3)
RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cpu and cuda:0! (when checking argument for argument mat2 in method wrapper_CUDA_bmm)