diffusers

diffusers copied to clipboard

https://github.com/huggingface/diffusers/blob/7fc53b5d66867f9e59398f5a14dd3dd9fbc700dd/src/diffusers/loaders/lora_base.py#L839 I note that `transformer.__class__` is changed from `diffusers.models.transformers.transformer_flux.FluxTransformer2DModel` to `torch._dynamo.eval_frame.OptimizedModule` when I use `torch.compile`, so I can't use set_adapters.

chenxiao111222

prompt_embeds_scale in FluxPriorReduxPipeline seems to have no effect.

1

### Describe the bug When using the FluxPriorReduxPipeline the prompt_embeds_scale and pooled_prompt_embeds_scale seem to have no effect on the final generation. ### Reproduction ``` async def get_redux_embeds(image, prompt, strength): redux_repo...

Meatfucker

RepositoryNotFoundError on inpainting demo

1

### Describe the bug Diffusers page https://huggingface.co/tasks/image-to-image#useful-resources links to demo https://huggingface.co/spaces/diffusers/stable-diffusion-xl-inpainting which opens with this: ``` runtime error Exit code: 1. Reason: per( File "/home/user/.pyenv/versions/3.10.13/lib/python3.10/site-packages/huggingface_hub/file_download.py", line 402, in _request_wrapper response...

abitrolly

Avoid creating tensor in CosmosAttnProcessor2_0

1

https://github.com/huggingface/diffusers/blob/42077e6c734df2fc7bbed373abceab99635500ad/src/diffusers/models/transformers/transformer_cosmos.py#L188-L193 ``` # 4. Prepare for GQA query_idx = query.size(3) key_idx = key.size(3) value_idx = value.size(3) key = key.repeat_interleave(query_idx // key_idx, dim=3) value = value.repeat_interleave(query_idx // value_idx, dim=3) ``` Speedup...

chenxiao111222

### Describe the bug When using the index_for_timestep method from FlowMatchEulerDiscreteScheduler in scheduling_flow_match_euler_discrete.py, the self.timesteps variable after __init__ maintains floating point values for timesteps (by default, 1000 to 1). Some...

JStyborski

add T5 lora train and infer for FluxPipeline

1

train and infer code of transformer lora + clip text encoder lora + T5 text encoder lora with FluxPipeline and FluxFillPipeline

huangjun12

[GGUF] feat: support loading diffusers format gguf checkpoints.

4

# What does this PR do? Refer to https://github.com/ngxson/flux-to-gguf/pull/1 to know how to obtain the checkpoint. After the checkpoint is obtained, run the following code for inference: Expand ```py import...

sayakpaul

[wip] enable FastPersist to (potentially) speed up disk saving in group offloading

1

# What does this PR do? FastPersist and DeepNVMe: https://github.com/deepspeedai/DeepSpeedExamples/tree/master/deepnvme Blog post: https://pytorch.org/blog/deepnvme-affordable-i-o-scaling-for-deep-learning-applications/ ⛔️ Currently not ready for reviews.

sayakpaul

Add pruna integration for loading model through diffusers.from_pretrained / pipeline

5

**Is your feature request related to a problem? Please describe.** A clear and concise description of what the problem is. Ex. I'm always frustrated when [...]. [Pruna](https://github.com/PrunaAI/pruna) is an open-source...

davidberenstein1957

how to load lora weight with fp8 transfomer model?

22

Hi, I want to run fluxcontrolpipeline with transformer_fp8 reference the code : https://huggingface.co/docs/diffusers/api/pipelines/flux#quantization ``` import torch from diffusers import BitsAndBytesConfig as DiffusersBitsAndBytesConfig, FluxTransformer2DModel, FluxControlPipeline from transformers import BitsAndBytesConfig as BitsAndBytesConfig,...

Johnson-yue