Sampler is failing with All input tensors need to be on the same GPU, but found some tensors to not be on a GPU:

[Morisgit148]

ERROR=
All input tensors need to be on the same GPU, but found some tensors to not be on a GPU:
[(torch.Size([1, 4718592]), device(type='cpu')), (torch.Size([147456]), device(type='cpu')), (torch.Size([3072, 3072]), device(type='cpu'))]

ran from GPU.bat
Failing for every sampler i have BITS and bytes installed updated Compfui tried everyting but getting same error i m able to run forge with the bnbnf4v2 and version1 failing on COMFYUI

running on gpu RTX 3060 = 6gb vram and corei9 16GB Ram

[teherauto]

Yeah, same problem. RTX3070 8GB VRAM and 32GB i7 notebook.

[KLL535]

same problem

[bluesky61]

Same error

[blitzk241]

same error

[PRANABWORKS]

same issue me too.

[Gmanskyi]

i got same trouble, 3080 10gb, pytorch version: 2.5.1+cu124, ComfyUI Version: v0.3.8, 49gb ram

KSampler
All input tensors need to be on the same GPU, but found some tensors to not be on a GPU:
[(torch.Size([4718592, 1]), device(type='cpu')), (torch.Size([1, 3072]), device(type='cuda', index=0)), (torch.Size([1, 3072]), device(type='cuda', index=0)), (torch.Size([147456]), device(type='cpu')), (torch.Size([16]), device(type='cpu'))]

[bluesky61]

This error seemed to be unfixable, so I moved to the GGUF version, which allowed us to run not only Flux, but also other models like CogVideo on GGUF.

[Sumit4647]

Yeah, please fixxxxxxxxxxxxxxxxxxx, helppppppppppppp someone