huggingface/diffusers

Pull Requests Commits

DN6 committed 247 days ago

5c99566b

[modular] fallback to default_blocks_name when loading base block classes in ModularPipeline (#13193)

yiyixuxu committed 248 days ago

Verified 39188248

[tests] consistency tests for modular index (#13192)

sayakpaul committed 248 days ago

Verified 9b979324

[Modular] update the auto pipeline blocks doc (#13148)

yiyixuxu committed 249 days ago

Verified 680076fc

Fixing Kohya loras loading: Flux.1-dev loras with TE ("lora_te1_" prefix) (#13188)

christopher5106 committed 249 days ago

Verified 5910a1cc

Fix LTX-2 image-to-video generation failure in two stages generation (#13187)

Songrui625 committed 249 days ago

Verified 40e96454

Fix Flash Attention 3 interface for new FA3 return format (#13173)

veeceey committed 250 days ago

Verified 47455bd1

Fix wrong `do_classifier_free_guidance` threshold in ZImagePipeline (#13183)

kirillsst committed 251 days ago

Verified 97c2c6e3

Cosmos Transfer2.5 Auto-Regressive Inference Pipeline (#13114)

miguelmartin75 committed 251 days ago

Verified 212db7b9

[attention backends] use dedicated wrappers from fa3 for cp. (#13165)

sayakpaul committed 251 days ago

Verified 31058485

[docs] Fix torchrun command argument order in docs (#13181)

sayakpaul committed 252 days ago

Verified aac94bef

fix: graceful fallback when attention backends fail to import (#13060)

sym-bot committed 252 days ago

Verified 1f6ac1c3

migrate to `transformers` v5 (#12976)

sayakpaul committed 252 days ago

Verified 5e94d62e

Fix AutoModel `typing` Import Error (#13178)

dg845 committed 252 days ago

Verified 7ab20117

Allow Automodel to use `from_config` with custom code. (#13123)

DN6 committed 253 days ago

Verified 4890e9bf

Fix T5GemmaEncoder loading for transformers 5.x composite T5GemmaConfig (#13143)

DavidBert committed 253 days ago

Verified f1e59141

Support Flux Klein peft (fal) lora format (#13169)

asomoza committed 255 days ago

Verified a80b1921

[gguf][torch.compile time] Convert to plain tensor earlier in dequantize_gguf_tensor (#13166)

anijain2305 committed 256 days ago

Verified 01de02e8

[CI] Fix new LoRAHotswap tests (#13163)

DN6 committed 256 days ago

Verified db2d7e7b

remove deps related to test from ci (#13164)

sayakpaul committed 256 days ago

Verified f8d3db9c

[core] Enable CP for kernels-based attention backends (#12812)

sayakpaul committed 257 days ago

Verified 99daaa80

Fix `ftfy` import for PRX Pipeline (#13154)

dg845 committed 257 days ago

Verified fe78a7b7

[CI] Revert `setuptools` CI Fix as the Failing Pipelines are Deprecated (#13149)

dg845 committed 257 days ago

Verified 53e1d0e4

Flux2: Tensor tuples can cause issues for checkpointing (#12777)

dxqb committed 258 days ago

Verified a577ec36

[docs] add docs for qwenimagelayered (#13158)

stevhliu committed 258 days ago

Verified 6875490c

docs: improve docstring scheduling_flow_match_lcm.py (#13160)

delmalih committed 258 days ago

Verified 64734b21

[CI] Add ftfy as a test dependency (#13155)

DN6 committed 258 days ago

Verified f81e6531

[Bug] Fix QwenImageEditPlus Series on NPU (#13017)

zhangtao0408 committed 259 days ago

Verified bcbbded7

[core] support device type device_maps to work with offloading. (#12811)

sayakpaul committed 260 days ago

Verified 35086ac0

[tests] accept recompile_limit from the user in tests (#13150)

sayakpaul committed 260 days ago

Verified e390646f

Older