Disty0
iGPUs are not officially supported. Every device officially supported by IPEX also supports BF16. There is test done by SDNext but IPEX seems to be computing nonsense instead of throwing...
> Also, it seems the SdNext default setting doesn't work, I have to switch to FP32 to make it use my iGPU. Default dtype for PyTorch (the setting in SDNext)...
Also noticed that your iGPU lists itself as fp64 capable according to the env info. Can you try exporting `IPEX_FORCE_ATTENTION_SLICE=1` This might be a 4GB allocation issue if the iGPU...
> if this was a 4G issue, I guess both fp16 and bf16 will get impacted? Yes.
Resizing the latents are not really possible. You need to use much higher denoise for it to work. SDXL latent space isn't compatible with resizing, rotating or similar image operations.
Closing as no response provided.
``` CUDA error: uncorrectable ECC error encountered ``` This is an hardware error. You should contact the compute provider because you've got a faulty node.
Try disabling non_blocking. It can have major issues with non-nvidia gpus.
> Non-blocking move operations, right? Yep, that is the setting i was talking about. > Torch tunable: enabled=True tuning=True Tunableop does use quite a lot of memory and takes quite...
RX 7600 isn't really supported by ROCm. Try overriding device architecture back to gfx1102 to see if it changes anything: ``` set HSA_OVERRIDE_GFX_VERSION=11.0.2 ```