huggingface/diffusers

Pull Requests Commits

Merge branch 'main' into refactor/lora-aware-qkv-fusion

dg845 committed 18 days ago

Verified 0b0e3e77

fix(bria_fibo): fix guidance_embeds, prompt_embeds, tensor-image and multi-image crashes (#13981)

akshan-main committed 18 days ago

Verified d8d3f906

fix claude code review fix in PRs. (#14058)

sayakpaul committed 18 days ago

Verified eb0a900c

[.ai] document single-file model layout and "don't reimplement Diffus… (#14048)

yiyixuxu committed 18 days ago

Verified 03040e51

Fix model cuda tests (#13975)

sayakpaul committed 20 days ago

Verified 6a71b6e3

Krea 2 LoRA DreamBooth trainer (#14046)

apolinario committed 20 days ago

Verified 3993de59

fix(flux): tighten check_inputs validation (#13955)

akshan-main committed 20 days ago

Verified afd77654

Enable BitsAndBytes quantization in MPS (#13915)

LucasSte committed 20 days ago

Verified e97a1ad2

[.ai doc] Refine .ai attention-mask and component-mutation guidance (#13982)

yiyixuxu committed 20 days ago

Verified aed5c641

Use inplace=False in ChronoEdit for consistency

dg845 committed 20 days ago

799b8cf7

Add inplace argument to model-overriden methods for compatibility (Wan, Helios, ChronoEdit)

dg845 committed 20 days ago

e41dee65

Merge branch 'main' into refactor/lora-aware-qkv-fusion

dg845 committed 20 days ago

a7512dbd

make style and make quality

dg845 committed 20 days ago

b761f4a3

Add Krea 2 (K2) text-to-image pipeline and transformer (#14045)

yiyixuxu committed 20 days ago

Verified 7104cb43

Fix Claude review comments (typo, get_added_qkv docstring)

dg845 committed 22 days ago

37c0f70c

Add tests for refactored QKV fusion API in AttentionMixin/AttentionModuleMixin

dg845 committed 22 days ago

658f0719

Refactor QKV fusion: add inplace option, model-level helper to restore to original fusion state, module-level helpers to get Q,K,V in both fused and split cases

dg845 committed 22 days ago

4e188842

[Fix] Fix three final_layer LoRA conversion bugs in _convert_sd_scripts_to_ai_toolkit (#14001)

lcheng321 committed 23 days ago

Verified 2d0110f8

Fix `Ideogram4MRoPE` collapsing under `torch.autocast` (compute rotary in float32) (#13922)

HaozheZhang6 committed 23 days ago

Verified 6e8c8c4c

Fix LTX2 connector token/register layout (regression from #13564) (#13931)

Boffee committed 24 days ago

Verified 7bf00006

ci: use hosted runners (#13987)

tarekziade committed 25 days ago

Verified b596c83a

ci: don't remind on prs from admins, etc. (#13965)

sayakpaul committed 25 days ago

Verified 9a72cd3e

[Neuron] Enable `torch.compile` compatibility with Neuron device (#13485)

JingyaHuang committed 25 days ago

Verified 236e5dd9

Enable LoRA loading on `Ideogram4ModularPipeline` (#13980)

SamuelTallet committed 25 days ago

Verified 6d9331ea

Add Ideogram4LoraLoaderMixin (LoRA loading for Ideogram4) (#13921)

linoytsaban committed 26 days ago

Verified 3467efa6

[tests] port final set of model tests and others (#13974)

sayakpaul committed 26 days ago

Verified 5e7540fc

Add PRXPixelPipeline: pixel-space PRX text-to-image pipeline (#13928)

DavidBert committed 26 days ago

Verified a487e2f1

Add Sound Encoder to Cosmos3 (#13911)

MaciejBalaNV committed 26 days ago

Verified 212a98d4

sayakpaul committed 27 days ago

4d85a3ae

sayakpaul committed 27 days ago

26ae69b1

Older