diffusers

diffusers copied to clipboard

[tests] add a test on torch compile for varied resolutions

3

# What does this PR do? After https://github.com/huggingface/diffusers/issues/11360 and https://github.com/huggingface/diffusers/pull/11297, we can notice how `torch.compile()` can also benefit scenarios that go through shape changes. This PR adds a test to...

sayakpaul

reset deterministic in tearDownClass

1

In #11738 , we enabled deterministic but forgot to reset it to False. Setting deterministic to False in tearDownClass will make the deterministic only have impacts for the test class...

jiqing-feng

fix "Expected all tensors to be on the same device, but found at least two devices" error

2

1. when run `pytest -rA tests/models/unets/test_models_unet_2d_condition.py::UNet2DConditionModelTests::test_load_sharded_checkpoint_device_map_from_hub_local` on 8 devices(CUDA, XPU), there will be a RuntimeError "RuntimeError: Expected all tensors to be on the same device, but found at least two...

yao-matrix

[single file] Cosmos

1

Possibly fixes #11798 We can run inference with the 7B Text-to-World model with the following code: ```python import torch from diffusers import CosmosTextToWorldPipeline, CosmosTransformer3DModel from diffusers.utils import export_to_video model_id =...

a-r-r-o-w

Support zero-3 for FLUX training

8

### Describe the bug Due to memory limitations, I am attempting to use Zero-3 for Flux training on 8 GPUs with 32GB each. I encountered a bug similar to the...

xiaoyewww

Run FLUX-controlnet zero3 training failed: 'weight' must be 2-D

14

### Describe the bug I am attempting to use Zero-3 for Flux Controlnet training on 8 GPUs following the guidance of [README](https://github.com/huggingface/diffusers/blob/main/examples/controlnet/README_flux.md#apply-deepspeed-zero3). The error below occured: ``` [rank0]: RuntimeError: 'weight'...

alien-0119

Single File and GGUF support of Cosmos-Predict2

1

As a Mac user I'm usually restricted in the memory reduction options I can use, so would like to ask for GGUF support for the new Comos models. I have...

Vargol

Add --lora_alpha and metadata handling to train_dreambooth_lora_sd3.py

2

# What does this PR do? Fixes #11730 ## Before submitting - [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if...

arnav-roy

[CI] Fix big GPU test marker

1

# What does this PR do? Big GPU test marker isn't registered in conftest, so tests aren't running. https://github.com/huggingface/diffusers/actions/runs/15838149705/job/44645818322 Fold it into `require_big_accelerator` so that we need to use fewer...

DN6

Add `--lora_alpha` and metadata handling for `train_dreambooth_lora_hidream`

1

# What does this PR do? Refs #11730 As described in the issue, I've updated the test and training scripts to making LoRA alpha also configurable and include the metadata...

ParagEkbote