QuantizedFluxTransformer2DModel save bug

[huangjun12]

Describe the bug

QuantizedFluxTransformer2DModel save bug

Reproduction

    class QuantizedFluxTransformer2DModel(QuantizedDiffusersModel):
        base_class = FluxTransformer2DModel

    transformer = FluxTransformer2DModel.from_pretrained(
        'black-forest-labs/FLUX.1-Fill-dev', subfolder="transformer", torch_dtype=torch.bfloat16, 
    ).to("cuda")

    qtransformer = QuantizedFluxTransformer2DModel.quantize(transformer, weights=qfloat8) 

    # for param in qtransformer.parameters(): param.data = param.data.contiguous() # useless

    qtransformer.save_pretrained('fluxfill_transformer_fp8')

Logs

ValueError: You are trying to save a non contiguous tensor: `time_text_embed.timestep_embedder.linear_1.weight._data` which is not allowed. It either means you are trying to save tensors which are reference of each other in which case it's recommended to save only the full tensors, and reslice at load time, or simply call `.contiguous()` on your tensor to pack it before saving.

System Info

python==3.12
torch==2.4.0 + cu121
transformers==4.47.0
optimum-quanto==0.2.6
diffusers main from 12.23

Who can help?

No response

[github-actions]

This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.

Please note that issues that do not follow the contributing guidelines are likely to be ignored.