CUDA error: operation not supported

[Pancat009]

CUDA error: operation not supported 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.

my env: CUDA:12.4 Python version: 3.10.11 pytorch version: 2.4.0+cu124 xformers version: 0.0.27.post2

[Pancat009]

CUDA: 12.6 reports the same error

[Pancat009]

if you are using 秋叶版aki-ui, try to switch the CUDA memory allocation change it to native

[KannManMachen]

Same problem / error.

Happened in the ltxv-13b-i2v-base workflow when reaching the LTXV Tiled Sampler node.

As suggested in the yellow note beneath the LTXV Tiled Sampler node: changing the mentioned values (tiles, SplitSigmas steps) did help to solve this problem (although I do not know which "noise levels" I should / could change as well...?).