Sparse Adam cuda error

[ricshaw]

I'm getting a cuda error when using diff_gaussian_rasterization with sparse adam optimizer with cuda 12.
Without sparse adam optimizer it works just fine....

File "/home/ma-user/miniconda3/envs/gaussian_splatting/lib/python3.10/site-packages/torch/optim/optimizer.py", line 391, in wrapper
out = func(*args, **kwargs)
File "/home/ma-user/miniconda3/envs/gaussian_splatting/lib/python3.10/site-packages/torch/utils/_contextlib.py", line 115, in decorate_context
return func(*args, **kwargs)
File "/home/ma-user/miniconda3/envs/gaussian_splatting/lib/python3.10/site-packages/diff_gaussian_rasterization/init.py", line 263, in step
state['exp_avg'] = torch.zeros_like(param, memory_format=torch.preserve_format)
RuntimeError: CUDA error: invalid configuration argument
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1.
Compile with TORCH_USE_CUDA_DSA to enable device-side assertions.

[chenqi13814529300]

same problem

[BitCalSaul]

@ricshaw hi did you fix it?