diffusers

diffusers copied to clipboard

SD3 ControlNet Script (and others?): dataset preprocessing cache depends on unrelated arguments

9

### Describe the bug When using the SD3 ControlNet training script, the training dataset embeddings are precomputed and the results are given a fingerprint based on the input script arguments,...

kentdan3msu

Bump torch from 2.2.0 to 2.7.1 in /examples/research_projects/realfill

1

Bumps [torch](https://github.com/pytorch/pytorch) from 2.2.0 to 2.7.1. Release notes Sourced from torch's releases. PyTorch 2.7.1 Release, bug fix release This release is meant to fix the following issues (regressions / silent...

dependabot[bot]

### Describe the bug If a Hidream lora is trained using OneTrainer, some keys of the checkpoint have the ".alpha" in them instead of all of them including 'lora'. So...

ali-afridi26

Implement TeaCache

14

# What does this PR do? ### What is TeaCache? [TeaCache](https://github.com/LiewFeng/TeaCache) (Timestep Embedding Aware Cache) is a training-free caching technique that speeds up diffusion model inference by **1.5x-2.6x** by reusing...

LawJarp-A

The Diffusers MVP 🚀

12

Hello folks 👋 ## What❓ We’re excited to bring a contributor-focused program to you! In this program, we want to work and collaborate with serious contributors to Diffusers and reward...

sayakpaul

Implement Selective Attention for Memory-Efficient Inference

3

# [Feature Request] Implement Selective Attention for Memory-Efficient Inference **Is your feature request related to a problem? Please describe.** Yes, it is hard to generate longer videos or process high-resolution...

cdutr

Add Unified Sequence Parallel attention

19

# What does this PR do? This is a draft implementation of the Unified SP attention approach. - Implements `_all_to_all_dim_exchange` with custom scatter and gather indices - Implements `TemplatedUnifiedAttention` Core...

Bissmella

Support for control-lora

88

This PR is a continuation of the following discussion https://github.com/huggingface/diffusers/issues/4679 https://github.com/huggingface/diffusers/pull/4899, and it addresses the following issues: 1. Loading SAI's control-lora files and enabling controlled image generation. 2. Building a...

lavinal712

Add support for LongCat-Image

5

# What does this PR do? This PR introduces the LongCat-Image Series into the diffusers library. The LongCat-Image Series comprises powerful and highly efficient 6B-parameter models, featuring both a state-of-the-art...

junqiangwu

`StateManager`, `Hook`, `State`: missing context setting (cond or uncond) in pipelines

2

In diffusion denoising model, we usually use `classifier free guidance` to control qualities. With [State Manager](https://github.com/huggingface/diffusers/blob/152f7ca357c066c4af3d1a58cdf17662ef5a2f87/src/diffusers/hooks/hooks.py#L42), we want each context hold different `state`, so we have to set which context...

toilaluan