diffusers

diffusers copied to clipboard

implementing the ddpm paper in a simple lines of code ~ 100

Esmail-ibraheem

[docs] LCM

1

Refactors the LCM and LCM-LoRA docs together because their usage is so similar it is redundant/harder to navigate to have separate pages for them. todo: - [x] add LCM-LoRA

stevhliu

[docs] Distilled inference

1

This PR combines "Distilled Stable Diffusion inference" with the "Speed up inference" doc. It also: - updates the table with only "speed" results (adds tf32 and combined results) and not...

stevhliu

# What does this PR do? These tests download the `v1-5-pruned-emaonly.ckpt` for SD 1.5, however we've updated our `torch.load` to default to use the `weights_only=True` which is causing issues with...

DN6

FloatTensor type hint is incompatible with Tensor

3

### Describe the bug At present, the `torch.FloatTensor` type hint is commonly employed in the codebase. However, if one were to pass a `torch.Tensor`, Pyright would raise complaints about the...

tianze0926

Diffusion Transformers quantization

12

**Is your feature request related to a problem? Please describe.** Due to OpenAI's DALLE-3 and Sora's soaring success, many projects, such as StableDiffusion3, PixArt and Open-Sora, are trying to replicate...

kabachuha

Train text encoder in train_text_to_image_sdxl.py

7

I have used this script successfully on a A100 40GB to train a full SDXL model, I would like to try training the text_encoder as well and that is currently...

stpg06

bug in load lora weights when add align_device_hook to model

14

### Describe the bug i noticed that when i add ```align_device_hook``` to module in pipeline manually, then ```load_lora_weights``` function will enable the sequential cpu offload. so i dig deeper and...

zhangvia

# What does this PR do? Running the failing tests that have cache issues on our SSH runner which uses `/mnt/cache/.cache/huggingface:/mnt/cache/` and the tests are passing. Updating the cache here...

DN6

[Core] support saving and loading of sharded checkpoints

4

# What does this PR do? Follow-up of https://github.com/huggingface/diffusers/pull/6396. This PR adds support for saving a big model's state dict into multiple shards for efficient portability and loading. Adds support...

sayakpaul