diffusers

diffusers copied to clipboard

[research_projects] add shortened flux training script with quantization

4

adds shortened script to be referenced in this blogpost https://github.com/huggingface/blog/pull/2888

DerekLiu35

self.attn ValueError: not enough values to unpack (expected 2, got 1)

15

### Describe the bug I have encountered an error when trying to sample from SD3 transformer component, where line 208 in `attention.py` is asking for two outputs but only one...

tracyyy-zhu

Failed to load lora in int8 mode

1

Let me introduce the background. The fill model changes the picture background. Torchao quantization is used for inference, but it fails when loading Lora. If Lora is not quantized, it...

babyta

Optional 'prompt_2' parameter in encode_prompt method is not optional

2

### Describe the bug in pipeline_flux.py The docstring says prompt_2 is optional but it is not declared as Optional[] and has no default None value Minor issue, could be in...

japppie

[quant] add __repr__ for better printing of configs.

3

# What does this PR do? Adds repr for GGUF and Quanto config classes.

sayakpaul

# What does this PR do? See https://github.com/huggingface/diffusers/pull/11712

sayakpaul

# What does this PR do? For this: https://github.com/huggingface/diffusers/actions/runs/15646378986/job/44084543295#step:7:614

sayakpaul

# What does this PR do? Draft PR to address some issues with the attention modules. 1. Moves a number of transformer related blocks/objects in the attention module that would...

DN6

flux fill fp8 load failed

8

### Describe the bug cc. https://huggingface.co/AlekseyCalvin/FluxFillDev_fp8_Diffusers/discussions/1 I want to run flux fill with fp8 for faster inference but it failed ### Reproduction ``` from diffusers import FluxTransformer2DModel, FluxFillPipeline from transformers...

Suprhimp

[docs] quant_kwargs

3

Improves where you can find the `quant_kwargs` in the `PipelineQuantizationConfig` docs

stevhliu