Bug - large non-batch-splittable convolutions

[mgscox]

When pushing the length up (very) high. Seems like this could be avoided;

ComfyUI_windows_portable\python_embeded\Lib\site-packages\torch\nn\modules\conv.py:720: UserWarning: cuDNN cannot be used for large non-batch-splittable convolutions if the V8 API is not enabled or before cuDNN version 9.3+. Consider upgrading cuDNN and/or enabling the V8 API for better efficiency. (Triggered internally at C:\actions-runner_work\pytorch\pytorch\builder\windows\pytorch\aten\src\ATen\native\Convolution.cpp:432.)
return F.conv3d(
Warning: Ran out of memory when regular VAE decoding, retrying with tiled VAE decoding.