RuntimeError: input tensor must fit into 32-bit index math

[OHLIA]

(StereoCrafter) root@7d0ec04d03ad:/StereoCrafter#  python depth_splatting_inference.py     --pre_trained_path ./weights/stable-video-diffusion-img2vid-xt-1-1     --unet_path ./weights/DepthCrafter     --input_video_path ./inputs/sample.mp4     --output_video_path ./outputs/sample_splatting_results.mp4
/opt/conda/envs/StereoCrafter/lib/python3.8/site-packages/Forward_Warp-0.0.1-py3.8.egg/Forward_Warp/forward_warp.py:19: SyntaxWarning: "is" with a literal. Did you mean "=="?
/opt/conda/envs/StereoCrafter/lib/python3.8/site-packages/Forward_Warp-0.0.1-py3.8.egg/Forward_Warp/forward_warp.py:57: SyntaxWarning: "is" with a literal. Did you mean "=="?
/opt/conda/envs/StereoCrafter/lib/python3.8/site-packages/Forward_Warp-0.0.1-py3.8.egg/Forward_Warp/forward_warp.py:58: SyntaxWarning: "is" with a literal. Did you mean "=="?
/opt/conda/envs/StereoCrafter/lib/python3.8/site-packages/Forward_Warp-0.0.1-py3.8.egg/Forward_Warp/forward_warp.py:19: SyntaxWarning: "is" with a literal. Did you mean "=="?
/opt/conda/envs/StereoCrafter/lib/python3.8/site-packages/Forward_Warp-0.0.1-py3.8.egg/Forward_Warp/forward_warp.py:57: SyntaxWarning: "is" with a literal. Did you mean "=="?
/opt/conda/envs/StereoCrafter/lib/python3.8/site-packages/Forward_Warp-0.0.1-py3.8.egg/Forward_Warp/forward_warp.py:58: SyntaxWarning: "is" with a literal. Did you mean "=="?
Loading pipeline components...: 100%|█████████████████████████████████████████████████████████████████████████████████| 5/5 [00:00<00:00, 16.90it/s]
==> processing video:  ./inputs/sample.mp4
==> original video shape:  (1416, 960, 1920, 3)
==> downsampled shape: (1416, 512, 1024, 3), with stride: 1
==> final processing shape: (1416, 512, 1024, 3)
Traceback (most recent call last):
  File "depth_splatting_inference.py", line 305, in <module>
    Fire(main)
  File "/opt/conda/envs/StereoCrafter/lib/python3.8/site-packages/fire/core.py", line 141, in Fire
    component_trace = _Fire(component, args, parsed_flag_args, context, name)
  File "/opt/conda/envs/StereoCrafter/lib/python3.8/site-packages/fire/core.py", line 475, in _Fire
    component, remaining_args = _CallAndUpdateTrace(
  File "/opt/conda/envs/StereoCrafter/lib/python3.8/site-packages/fire/core.py", line 691, in _CallAndUpdateTrace
    component = fn(*varargs, **kwargs)
  File "depth_splatting_inference.py", line 287, in main
    video_depth, depth_vis = depthcrafter_demo.infer(
  File "depth_splatting_inference.py", line 123, in infer
    res = self.pipe(
  File "/opt/conda/envs/StereoCrafter/lib/python3.8/site-packages/torch/utils/_contextlib.py", line 115, in decorate_context
    return func(*args, **kwargs)
  File "/StereoCrafter/dependency/DepthCrafter/depthcrafter/depth_crafter_ppl.py", line 162, in __call__
    video_embeddings = self.encode_video(
  File "/opt/conda/envs/StereoCrafter/lib/python3.8/site-packages/torch/utils/_contextlib.py", line 115, in decorate_context
    return func(*args, **kwargs)
  File "/StereoCrafter/dependency/DepthCrafter/depthcrafter/depth_crafter_ppl.py", line 32, in encode_video
    video_224 = _resize_with_antialiasing(video.float(), (224, 224))
  File "/opt/conda/envs/StereoCrafter/lib/python3.8/site-packages/diffusers/pipelines/stable_video_diffusion/pipeline_stable_video_diffusion.py", line 645, in _resize_with_antialiasing
    input = _gaussian_blur2d(input, ks, sigmas)
  File "/opt/conda/envs/StereoCrafter/lib/python3.8/site-packages/diffusers/pipelines/stable_video_diffusion/pipeline_stable_video_diffusion.py", line 723, in _gaussian_blur2d
    out_x = _filter2d(input, kernel_x[..., None, :])
  File "/opt/conda/envs/StereoCrafter/lib/python3.8/site-packages/diffusers/pipelines/stable_video_diffusion/pipeline_stable_video_diffusion.py", line 684, in _filter2d
    input = torch.nn.functional.pad(input, padding_shape, mode="reflect")
RuntimeError: input tensor must fit into 32-bit index math

原本视频样本分辨率为3840x1920，会报错。我改为1920x960之后，仍然同样的错误。请问我需要如何修改？

[a631807682]

同样的问题，有什么解决方法吗？