Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

GouraudShader failing with AmbientLights #1043

Closed
samedii opened this issue Jan 16, 2022 · 11 comments
Closed

GouraudShader failing with AmbientLights #1043

samedii opened this issue Jan 16, 2022 · 11 comments
Assignees
@bottler

Comments

@samedii
Copy link

samedii commented Jan 16, 2022

🐛 Bugs / Unexpected behaviors

SoftGouraudShader and HardGouraudShader shader fail when SoftPhongShader and HardPhongShader work. I am using a TexturesVertex and I've tried with a few different meshes. I am just doing simple rendering.

This seems to be relevant:
pytorch/pytorch#55027

The detectron error is fixed on master/nightlies with #55292. 3rd party projects depending on torch are advised not to use thrust, and wrap cub in its own namespace to avoid these errors. Longer term, pytorch should stop using thrust and rely on namespaced cub only. #52663 is to track extension errors.

Do you know if there any way for me to avoid this issue by changing versions of packages etc?

Previous messages:

packages/pytorch3d/transforms/transform3d.py:780: UserWarning: R is not a valid rotation matrix
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [112,0,0], thread: [29,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
...
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [219,0,0], thread: [63,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.

Stacktrace:

  File "/home/aiwizo/Documents/generate-texture/.venv/lib/python3.8/site-packages/pytorch3d/renderer/mesh/shader.py", line 283, in forward
    pixel_colors = gouraud_shading(
  File "/home/aiwizo/Documents/generate-texture/.venv/lib/python3.8/site-packages/pytorch3d/renderer/mesh/shading.py", line 135, in gouraud_shading
    cameras = cameras.clone().gather_props(vert_to_mesh_idx)
  File "/home/aiwizo/Documents/generate-texture/.venv/lib/python3.8/site-packages/pytorch3d/renderer/cameras.py", line 346, in clone
    other = cam_type(device=self.device)
  File "/home/aiwizo/Documents/generate-texture/.venv/lib/python3.8/site-packages/pytorch3d/renderer/cameras.py", line 531, in __init__
    super().__init__(
  File "/home/aiwizo/Documents/generate-texture/.venv/lib/python3.8/site-packages/pytorch3d/renderer/utils.py", line 133, in __init__
    broadcasted_values = convert_to_tensors_and_broadcast(
  File "/home/aiwizo/Documents/generate-texture/.venv/lib/python3.8/site-packages/pytorch3d/renderer/utils.py", line 336, in convert_to_tensors_and_broadcast
    args_1d = [format_tensor(c, dtype, device) for c in args]
  File "/home/aiwizo/Documents/generate-texture/.venv/lib/python3.8/site-packages/pytorch3d/renderer/utils.py", line 336, in <listcomp>
    args_1d = [format_tensor(c, dtype, device) for c in args]
  File "/home/aiwizo/Documents/generate-texture/.venv/lib/python3.8/site-packages/pytorch3d/renderer/utils.py", line 298, in format_tensor
    input = torch.tensor(input, dtype=dtype, device=device_)
RuntimeError: CUDA error: device-side assert triggered
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.

Instructions To Reproduce the Issue:

Using poetry to install:

python = "3.8.8"
torch = "1.9.0"
torchvision = "0.10.0"
torchaudio = "0.9.0"
fvcore = "^0.1.5"
iopath = "^0.1.9"
ninja = "^1.10.2"
matplotlib = "^3.5.1"
tqdm = "^4.62.3"
Pillow = "^9.0.0"
kornia = "^0.6.2"
ftfy = "^6.0.3"
regex = "^2021.11.10"
requests = "^2.27.1"

CUDA 11.1

poetry run python -m pip install torch==1.9.0+cu111 torchvision==0.10.0+cu111 torchaudio==0.9.0 -f https://download.pytorch.org/whl/torch_stable.html

NVIDIA CUB requirement for CUDA 11.1

curl -LO https://github.com/NVIDIA/cub/archive/1.9.10-1.tar.gz
tar -xvzf 1.9.10-1.tar.gz
export CUB_HOME=$PWD/cub-1.9.10-1

pip install git+https://github.com/facebookresearch/pytorch3d.git@v0.6.1
@bottler
Copy link
Contributor

bottler commented Jan 16, 2022

Not sure what is happening here. Can you try again with the environment variable CUDA_LAUNCH_BLOCKING set to 1? It is possible that the stacktrace/error message will be more informative in that case. Do you call randperm in your code?

@bottler bottler self-assigned this Jan 16, 2022
@samedii
Copy link
Author

samedii commented Jan 18, 2022

I'm getting the same error message with CUDA_LAUNCH_BLOCKING=1 (see stack trace below). I am not using randperm.

I will install pytorch3d on another machine tomorrow and see if it works with CUDA 10.

➜ CUDA_LAUNCH_BLOCKING=1 poetry run python test_render.py
/home/aiwizo/Documents/generate-texture/.venv/lib/python3.8/site-packages/pytorch3d/transforms/transform3d.py:780: UserWarning: R is not a valid rotation matrix
  warnings.warn(msg)
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [32,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [33,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [34,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [35,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [36,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [37,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [38,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [39,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [40,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [41,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [42,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [43,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [44,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [45,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [46,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [47,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [48,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [49,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [50,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [51,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
Traceback (most recent call last):
  File "test_diffusion8.py", line 45, in <module>
    images = render(mesh.extend(BATCH_SIZE), azims, IMAGE_SIZE, DEVICE)
  File "/home/aiwizo/Documents/generate-texture/generate_texture/render.py", line 71, in render
    return renderer(azims, image_size, device)(mesh)
  File "/home/aiwizo/Documents/generate-texture/.venv/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl
    return forward_call(*input, **kwargs)
  File "/home/aiwizo/Documents/generate-texture/.venv/lib/python3.8/site-packages/pytorch3d/renderer/mesh/renderer.py", line 60, in forward
    images = self.shader(fragments, meshes_world, **kwargs)
  File "/home/aiwizo/Documents/generate-texture/.venv/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl
    return forward_call(*input, **kwargs)
  File "/home/aiwizo/Documents/generate-texture/.venv/lib/python3.8/site-packages/pytorch3d/renderer/mesh/shader.py", line 279, in forward
    pixel_colors = gouraud_shading(
  File "/home/aiwizo/Documents/generate-texture/.venv/lib/python3.8/site-packages/pytorch3d/renderer/mesh/shading.py", line 134, in gouraud_shading
    lights = lights.clone().gather_props(vert_to_mesh_idx)
  File "/home/aiwizo/Documents/generate-texture/.venv/lib/python3.8/site-packages/pytorch3d/renderer/utils.py", line 273, in gather_props
    v = v.gather(0, _batch_idx)
RuntimeError: CUDA error: device-side assert triggered

@bottler
Copy link
Contributor

bottler commented Jan 18, 2022

This is a more informative stack trace. Can you check the batch length of the lighting object matches the number of meshes in the batch?

@samedii
Copy link
Author

samedii commented Jan 18, 2022 •
edited
Loading

I just have ambient lights:

def renderer(azims, image_size, device):
    R, T = pytorch3d.renderer.look_at_view_transform(
        25.6, 15.0, azims, at=((0.0, 10.0, 0.0),)
    )
    cameras = pytorch3d.renderer.FoVPerspectiveCameras(
        device=device, R=R, T=T, znear=1, zfar=1000
    )
    return pytorch3d.renderer.MeshRenderer(
        rasterizer=pytorch3d.renderer.MeshRasterizer(
            cameras=cameras,
            raster_settings=pytorch3d.renderer.RasterizationSettings(
                image_size=image_size,
                blur_radius=np.log(1.0 / 1e-4 - 1.0) * 1e-5,
                faces_per_pixel=50,
            ),
        ),
        shader=pytorch3d.renderer.SoftGouraudShader(
            device=device,
            cameras=cameras,
            lights=pytorch3d.renderer.lighting.AmbientLights(device=device),
            blend_params=pytorch3d.renderer.BlendParams(
                background_color=(1.0, 1.0, 1.0),
            ),
        ),
    )

I tried changing the lights to (to have the same batch length) this but it gives the same type of error message (indexes have changed):

            lights=pytorch3d.renderer.lighting.AmbientLights(
                ambient_color=torch.ones([len(azims), 3]).to(azims), device=device
            ),

@bottler
Copy link
Contributor

bottler commented Jan 18, 2022

"same type of error message" - but this could have been a fix in the right direction. Can you set CUDA_LAUNCH_BLOCKING and look closely at the stack trace and then what is going on in the code?

Another way to simplify errors might be to try doing everything on the CPU.

@samedii
Copy link
Author

samedii commented Jan 18, 2022

This is the stacktrace:

CUDA_LAUNCH_BLOCKING=1 poetry run python test_diffusion8.py                                                                                                                                                      
/home/aiwizo/Documents/generate-texture/.venv/lib/python3.8/site-packages/pytorch3d/transforms/transform3d.py:780: UserWarning: R is not a valid rotation matrix                               
  warnings.warn(msg)                                                                                                                                                                                               
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [32,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [33,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [34,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [35,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [36,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [37,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [38,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [39,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [40,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [41,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [42,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [43,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [44,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [45,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [46,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [47,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [48,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [49,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [50,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [51,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [0,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed. 
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [1,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed. 
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [2,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed. 
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [3,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed. 
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [4,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed. 
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [5,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed. 
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [6,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.                     
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [7,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.                     
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [8,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.                     
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [9,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.                     
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [10,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.                    
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [11,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [12,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [13,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [14,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [15,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [16,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [17,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [18,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [19,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [20,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [21,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [22,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [23,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.                    
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [24,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [25,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.                    
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [26,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [27,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [28,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [29,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [30,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [31,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [39,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [40,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [41,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [42,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [43,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [44,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [45,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [46,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [47,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [48,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [49,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [50,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [51,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [52,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [53,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [54,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [55,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [56,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [57,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [58,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [59,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [60,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [61,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [62,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [63,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
Traceback (most recent call last):
  File "test_diffusion8.py", line 45, in <module>
    images = render(mesh.extend(BATCH_SIZE), azims, IMAGE_SIZE, DEVICE)
  File "/home/aiwizo/Documents/generate-texture/generate_texture/render.py", line 75, in render
    return renderer(azims, image_size, device)(mesh) 
  File "/home/aiwizo/Documents/generate-texture/.venv/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl
    return forward_call(*input, **kwargs)
  File "/home/aiwizo/Documents/generate-texture/.venv/lib/python3.8/site-packages/pytorch3d/renderer/mesh/renderer.py", line 60, in forward
    images = self.shader(fragments, meshes_world, **kwargs)
  File "/home/aiwizo/Documents/generate-texture/.venv/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl
    return forward_call(*input, **kwargs)
  File "/home/aiwizo/Documents/generate-texture/.venv/lib/python3.8/site-packages/pytorch3d/renderer/mesh/shader.py", line 279, in forward
    pixel_colors = gouraud_shading(
  File "/home/aiwizo/Documents/generate-texture/.venv/lib/python3.8/site-packages/pytorch3d/renderer/mesh/shading.py", line 134, in gouraud_shading
    lights = lights.clone().gather_props(vert_to_mesh_idx)
  File "/home/aiwizo/Documents/generate-texture/.venv/lib/python3.8/site-packages/pytorch3d/renderer/utils.py", line 273, in gather_props
    v = v.gather(0, _batch_idx)
RuntimeError: CUDA error: device-side assert triggered

I will just make a simple repository reproducing the error.

@samedii
Copy link
Author

samedii commented Jan 18, 2022

You seem to be onto something. Here is code that reproduces the error: https://github.com/samedii/pytorch3d-gouraud-failure/blob/master/test.py

When reproducing the error I found that BATCH_SIZE = 1 does not give an error.

BATCH_SIZE = 2 gives another error message:

➜ poetry run python test.py
/home/aiwizo/Documents/pytorch3d-gouraud-failure/.venv/lib/python3.8/site-packages/pytorch3d/transforms/transform3d.py:780: UserWarning: R is not a valid rotation matrix
  warnings.warn(msg)
Traceback (most recent call last):
  File "test.py", line 59, in <module>
    images = render(mesh.extend(BATCH_SIZE), azims, IMAGE_SIZE, DEVICE)
  File "test.py", line 42, in render
    return renderer(azims, image_size, device)(mesh)
  File "/home/aiwizo/Documents/pytorch3d-gouraud-failure/.venv/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl
    return forward_call(*input, **kwargs)
  File "/home/aiwizo/Documents/pytorch3d-gouraud-failure/.venv/lib/python3.8/site-packages/pytorch3d/renderer/mesh/renderer.py", line 60, in forward
    images = self.shader(fragments, meshes_world, **kwargs)
  File "/home/aiwizo/Documents/pytorch3d-gouraud-failure/.venv/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl
    return forward_call(*input, **kwargs)
  File "/home/aiwizo/Documents/pytorch3d-gouraud-failure/.venv/lib/python3.8/site-packages/pytorch3d/renderer/mesh/shader.py", line 279, in forward
    pixel_colors = gouraud_shading(
  File "/home/aiwizo/Documents/pytorch3d-gouraud-failure/.venv/lib/python3.8/site-packages/pytorch3d/renderer/mesh/shading.py", line 139, in gouraud_shading
    ambient, diffuse, specular = _apply_lighting(
  File "/home/aiwizo/Documents/pytorch3d-gouraud-failure/.venv/lib/python3.8/site-packages/pytorch3d/renderer/mesh/shading.py", line 39, in _apply_lighting
    ambient_color = materials.ambient_color * lights.ambient_color
RuntimeError: The size of tensor a (3) must match the size of tensor b (48922) at non-singleton dimension 1

BATCH_SIZE = 3 gives similar:

➜ poetry run python test.py
/home/aiwizo/Documents/pytorch3d-gouraud-failure/.venv/lib/python3.8/site-packages/pytorch3d/transforms/transform3d.py:780: UserWarning: R is not a valid rotation matrix
  warnings.warn(msg)
Traceback (most recent call last):
  File "test.py", line 59, in <module>
    images = render(mesh.extend(BATCH_SIZE), azims, IMAGE_SIZE, DEVICE)
  File "test.py", line 42, in render
    return renderer(azims, image_size, device)(mesh)
  File "/home/aiwizo/Documents/pytorch3d-gouraud-failure/.venv/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl
    return forward_call(*input, **kwargs)
  File "/home/aiwizo/Documents/pytorch3d-gouraud-failure/.venv/lib/python3.8/site-packages/pytorch3d/renderer/mesh/renderer.py", line 60, in forward
    images = self.shader(fragments, meshes_world, **kwargs)
  File "/home/aiwizo/Documents/pytorch3d-gouraud-failure/.venv/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl
    return forward_call(*input, **kwargs)
  File "/home/aiwizo/Documents/pytorch3d-gouraud-failure/.venv/lib/python3.8/site-packages/pytorch3d/renderer/mesh/shader.py", line 279, in forward
    pixel_colors = gouraud_shading(
  File "/home/aiwizo/Documents/pytorch3d-gouraud-failure/.venv/lib/python3.8/site-packages/pytorch3d/renderer/mesh/shading.py", line 139, in gouraud_shading
    ambient, diffuse, specular = _apply_lighting(
  File "/home/aiwizo/Documents/pytorch3d-gouraud-failure/.venv/lib/python3.8/site-packages/pytorch3d/renderer/mesh/shading.py", line 39, in _apply_lighting
    ambient_color = materials.ambient_color * lights.ambient_color
RuntimeError: The size of tensor a (3) must match the size of tensor b (73383) at non-singleton dimension 1

BATCH_SIZE = 4 gives the error from before:

Traceback (most recent call last):                                                                                                                                                                                 
  File "test.py", line 59, in <module>                                                                                                                                                                             
    images = render(mesh.extend(BATCH_SIZE), azims, IMAGE_SIZE, DEVICE)                                                                                                                                            
  File "test.py", line 42, in render                                                                                                                                                                               
    return renderer(azims, image_size, device)(mesh)                                                                                                                                                               
  File "/home/aiwizo/Documents/pytorch3d-gouraud-failure/.venv/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl                                                                   
    return forward_call(*input, **kwargs)                                                                                                                                                                          
  File "/home/aiwizo/Documents/pytorch3d-gouraud-failure/.venv/lib/python3.8/site-packages/pytorch3d/renderer/mesh/renderer.py", line 60, in forward                                           
    images = self.shader(fragments, meshes_world, **kwargs)                                                                                                                                                        
  File "/home/aiwizo/Documents/pytorch3d-gouraud-failure/.venv/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl                                               
    return forward_call(*input, **kwargs)                                                                                                                                                                          
  File "/home/aiwizo/Documents/pytorch3d-gouraud-failure/.venv/lib/python3.8/site-packages/pytorch3d/renderer/mesh/shader.py", line 279, in forward                                            
    pixel_colors = gouraud_shading(                                                                                                                                                                                
  File "/home/aiwizo/Documents/pytorch3d-gouraud-failure/.venv/lib/python3.8/site-packages/pytorch3d/renderer/mesh/shading.py", line 134, in gouraud_shading                                   
    lights = lights.clone().gather_props(vert_to_mesh_idx)                                                                                                                                                         
  File "/home/aiwizo/Documents/pytorch3d-gouraud-failure/.venv/lib/python3.8/site-packages/pytorch3d/renderer/utils.py", line 273, in gather_props                                             
    v = v.gather(0, _batch_idx)                                                                                                                                                                                    
RuntimeError: CUDA error: device-side assert triggered

Changing to CPU with BATCH_SIZE = 4 gives this error:

➜ CUDA_LAUNCH_BLOCKING=1 poetry run python test.py           
Traceback (most recent call last):
  File "test.py", line 59, in <module>
    images = render(mesh.extend(BATCH_SIZE), azims, IMAGE_SIZE, DEVICE)
  File "test.py", line 42, in render
    return renderer(azims, image_size, device)(mesh)
  File "/home/aiwizo/Documents/pytorch3d-gouraud-failure/.venv/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl
    return forward_call(*input, **kwargs)
  File "/home/aiwizo/Documents/pytorch3d-gouraud-failure/.venv/lib/python3.8/site-packages/pytorch3d/renderer/mesh/renderer.py", line 60, in forward
    images = self.shader(fragments, meshes_world, **kwargs)
  File "/home/aiwizo/Documents/pytorch3d-gouraud-failure/.venv/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl
    return forward_call(*input, **kwargs)
  File "/home/aiwizo/Documents/pytorch3d-gouraud-failure/.venv/lib/python3.8/site-packages/pytorch3d/renderer/mesh/shader.py", line 279, in forward
    pixel_colors = gouraud_shading(
  File "/home/aiwizo/Documents/pytorch3d-gouraud-failure/.venv/lib/python3.8/site-packages/pytorch3d/renderer/mesh/shading.py", line 134, in gouraud_shading
    lights = lights.clone().gather_props(vert_to_mesh_idx)
  File "/home/aiwizo/Documents/pytorch3d-gouraud-failure/.venv/lib/python3.8/site-packages/pytorch3d/renderer/utils.py", line 273, in gather_props
    v = v.gather(0, _batch_idx)
RuntimeError: index 3 is out of bounds for dimension 0 with size 3

@samedii
Copy link
Author

samedii commented Jan 18, 2022

Removing the light seems to fix the issue

pytorch3d.renderer.lighting.AmbientLights(device=device)

@bottler
Copy link
Contributor

bottler commented Jan 18, 2022

Yes. I think this is a specific problem with the AmbientLights class. You can use a PointLights with only an ambient component instead.

@samedii samedii changed the title GouraudShader failing for simple rendering GouraudShader failing with AmbientLights Jan 18, 2022
@spikeeeeeeeeee
Copy link

This is the stacktrace:

CUDA_LAUNCH_BLOCKING=1 poetry run python test_diffusion8.py                                                                                                                                                      
/home/aiwizo/Documents/generate-texture/.venv/lib/python3.8/site-packages/pytorch3d/transforms/transform3d.py:780: UserWarning: R is not a valid rotation matrix                               
  warnings.warn(msg)                                                                                                                                                                                               
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [32,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [33,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [34,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [35,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [36,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [37,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [38,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [39,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [40,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [41,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [42,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [43,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [44,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [45,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [46,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [47,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [48,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [49,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [50,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [382,0,0], thread: [51,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [0,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed. 
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [1,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed. 
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [2,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed. 
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [3,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed. 
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [4,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed. 
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [5,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed. 
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [6,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.                     
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [7,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.                     
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [8,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.                     
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [9,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.                     
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [10,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.                    
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [11,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [12,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [13,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [14,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [15,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [16,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [17,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [18,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [19,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [20,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [21,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [22,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [23,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.                    
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [24,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [25,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.                    
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [26,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [27,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [28,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [29,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [30,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [306,0,0], thread: [31,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [39,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [40,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [41,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [42,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [43,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [44,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [45,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [46,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [47,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [48,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [49,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [50,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [51,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [52,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [53,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [54,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [55,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [56,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [57,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [58,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [59,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [60,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [61,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [62,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
/pytorch/aten/src/ATen/native/cuda/ScatterGatherKernel.cu:115: operator(): block: [286,0,0], thread: [63,0,0] Assertion `idx_dim >= 0 && idx_dim < index_size && "index out of bounds"` failed.
Traceback (most recent call last):
  File "test_diffusion8.py", line 45, in <module>
    images = render(mesh.extend(BATCH_SIZE), azims, IMAGE_SIZE, DEVICE)
  File "/home/aiwizo/Documents/generate-texture/generate_texture/render.py", line 75, in render
    return renderer(azims, image_size, device)(mesh) 
  File "/home/aiwizo/Documents/generate-texture/.venv/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl
    return forward_call(*input, **kwargs)
  File "/home/aiwizo/Documents/generate-texture/.venv/lib/python3.8/site-packages/pytorch3d/renderer/mesh/renderer.py", line 60, in forward
    images = self.shader(fragments, meshes_world, **kwargs)
  File "/home/aiwizo/Documents/generate-texture/.venv/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl
    return forward_call(*input, **kwargs)
  File "/home/aiwizo/Documents/generate-texture/.venv/lib/python3.8/site-packages/pytorch3d/renderer/mesh/shader.py", line 279, in forward
    pixel_colors = gouraud_shading(
  File "/home/aiwizo/Documents/generate-texture/.venv/lib/python3.8/site-packages/pytorch3d/renderer/mesh/shading.py", line 134, in gouraud_shading
    lights = lights.clone().gather_props(vert_to_mesh_idx)
  File "/home/aiwizo/Documents/generate-texture/.venv/lib/python3.8/site-packages/pytorch3d/renderer/utils.py", line 273, in gather_props
    v = v.gather(0, _batch_idx)
RuntimeError: CUDA error: device-side assert triggered

I will just make a simple repository reproducing the error.

have you found the reason of this problem? it's redicilious.

@bottler
Copy link
Contributor

bottler commented Jun 21, 2024

Does the workaround of using a PointLights with only an ambient component work? Something probably has the wrong shape, and is (perhaps deliberately to avoid a slowdown) not being validated.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@bottler bottler

Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
3 participants
@bottler @samedii @spikeeeeeeeeee