diffusers

diffusers copied to clipboard

Add IP Adapter training script and update the docs with instructions

7

# What does this PR do? This PR adds the [original IP Adapter](https://github.com/tencent-ailab/IP-Adapter) training scripts, and also updates the documentation with instructions on how to use it. Fixes # (issue)...

AMohamedAakhil

Add SUPIR Upscaler

33

### Model/Pipeline/Scheduler description [SUPIR](https://supir.xpixel.group/) is a super-resolution model that looks like it produces excellent results Github Repo: https://github.com/Fanghua-Yu/SUPIR The model is quite memory intensive, so the optimisation features available in...

DN6

Flow Matching Euler Scheduler added

11

In PR I have implemented a scheduler for Flow Matching with Euler method. The popularity for Flow Matching models grows rapidly. Even novel models like SD3 (https://arxiv.org/pdf/2403.03206) use flow matching...

leffff

playground cuda error

10

https://github.com/huggingface/diffusers/blob/896fb6d8d7c10001eb2a92568be7b4bd3d5ddea3/src/diffusers/pipelines/stable_diffusion_xl/pipeline_stable_diffusion_xl_img2img.py#L728 When I initialize without using `to('cuda')`, the model exists in the cpu, so here self.device gets the handle on the cpu, refer to the code above, whether we change...

Honey-666

fix min-snr implementation

2

# What does this PR do? Fixes the implementation of min-snr training for v-prediction models Based on implementation seen here. https://github.com/kohya-ss/sd-scripts/blob/main/library/custom_train_functions.py#L66 These two graphs show the weight scheme based on...

ethansmith2000

Remove unnecessary lines

1

They are unnecessary lines, aren't they? Self-assigned vars have been found with [PLW0127](https://docs.astral.sh/ruff/rules/self-assigning-variable/). @yiyixuxu @DN6

tolgacangoz

[SD3] vae.config.shift_factor missing in dreambooth training examples

6

### Describe the bug shift_factor missing in traning code: https://github.com/huggingface/diffusers/blob/main/examples/dreambooth/train_dreambooth_lora_sd3.py#L1617, but used in inference code: https://github.com/huggingface/diffusers/blob/main/src/diffusers/pipelines/stable_diffusion_3/pipeline_stable_diffusion_3.py#L893 Is it resonable that when traning SD3, we do not need to norm latents...

bendanzzc

Rethinking the `encode_prompt()` method in pipelines

1

This thread is for discussing the possibility of making the most widely used `encode_prompt()` methods of our pipelines `classmethod`s. For historical context, I have made such attempts in the past...

sayakpaul

Fixes #8477 Similar to #3673 @dg845

tolgacangoz

Improve model card for `push_to_hub` trainers

1

# What does this PR do? Improves the model card of our trainers to better UX, hub integration, platform agnostic usage and better attribution (WIP)

apolinario