Skip to content
Open

Main #11

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
772 commits
Select commit Hold shift + click to select a range
9e4a75b
[docs] Fix VAE scale factor calculation in distributed inference docs…
Men1scus Sep 2, 2025
006d092
[Flux LoRA] fix for prior preservation and mixed precision sampling, …
linoytsaban Sep 2, 2025
bcd4d77
[CI] Remove big accelerator requirements from Quanto Tests (#12266)
DN6 Sep 3, 2025
130fd8d
[core] use `kernels` to support `_flash_3_hub` attention backend (#12…
sayakpaul Sep 3, 2025
6549b04
[docs] AutoPipeline (#12160)
stevhliu Sep 3, 2025
4acbfbf
[Quantization] Add TRT-ModelOpt as a Backend (#11173)
ishan-modi Sep 3, 2025
ffc8c0c
[tests] feat: add AoT compilation tests (#12203)
sayakpaul Sep 3, 2025
6682956
Add AttentionMixin to WanVACETransformer3DModel (#12268)
J4BEZ Sep 3, 2025
764b624
fix some typos (#12265)
co63oc Sep 3, 2025
c2e5ece
[docs] Sharing pipelines/models (#12280)
stevhliu Sep 4, 2025
32798bf
[docs] Inference section cleanup (#12281)
stevhliu Sep 5, 2025
fc337d5
[docs] Models (#12248)
stevhliu Sep 5, 2025
f50b18e
[Modular] Qwen (#12220)
yiyixuxu Sep 8, 2025
4e36bb0
Support ControlNet-Inpainting for Qwen-Image (#12301)
haofanwang Sep 9, 2025
c222570
DeepSpeed adaption for flux-kontext (#12240)
leisuzz Sep 9, 2025
28106fc
gguf new quant type support (with demo) (#12076)
calcuis Sep 9, 2025
4067d6c
adjust criteria for marigold-intrinsics example on XPU (#12290)
kaixuanliu Sep 9, 2025
4345907
[core] feat: support group offloading at the pipeline level (#12283)
sayakpaul Sep 10, 2025
f7b7945
[modular] fix flux modular pipelines for t2i and i2i (#12272)
sayakpaul Sep 10, 2025
9e7ae56
[feat] cache allocator warmup for `from_single_model` (#12305)
sayakpaul Sep 10, 2025
e1b7f1f
fix for the qwen controlnet pipeline - wrong device can be used (#12309)
ttio2tech Sep 10, 2025
eb7ef26
[quant] allow `components_to_quantize` to be a non-list for single co…
sayakpaul Sep 10, 2025
55f0b3d
Fix AttributeError of `VisualClozeProcessor` (#12121)
Justin900429 Sep 10, 2025
5e181ed
Deprecate slicing and tiling methods from `DiffusionPipeline` (#12271)
sayakpaul Sep 11, 2025
f5c113e
Use SDP on BF16 in GPU/HPU migration (#12310)
dsocek Sep 12, 2025
b500140
Add Wan2.2 VACE - Fun (#12324)
linoytsaban Sep 15, 2025
751e250
fixed bug in defining embed dim for UNet1D (#12111)
SammyAgrawal Sep 16, 2025
8c72cd1
Added LucyEditPipeline (#12340)
sarihl Sep 16, 2025
d06750a
Fix autoencoder_kl_wan.py bugs for Wan2.2 VAE (#12335)
franciszzj Sep 16, 2025
efb7a29
Fix many type hint errors (#12289)
DefTruth Sep 17, 2025
eda9ff8
Add RequestScopedPipeline for safe concurrent inference, tokenizer lo…
FredyRivera-dev Sep 18, 2025
7e7e62c
Convert alphas for embedders for sd-scripts to ai toolkit conversion …
rockerBOO Sep 18, 2025
1eeddd0
Merge branch 'huggingface:main' into main
clementchadebec Sep 19, 2025
edd614e
[CI] Fix TRANSFORMERS_FLAX_WEIGHTS_NAME import issue (#12354)
DN6 Sep 20, 2025
df267ee
feat: Add QwenImageEditPlus to support future feature upgrades (#12357)
naykun Sep 21, 2025
c006a95
Fix example server install instructions (#12362)
jasonacox Sep 22, 2025
843355f
[tests] xfail some kandinsky tests. (#12364)
sayakpaul Sep 22, 2025
d83d35c
Fix bug with VAE slicing in autoencoder_dc.py (#12343)
mingyi456 Sep 22, 2025
78031c2
[Fix] enable_xformers_memory_efficient_attention() in Flux Pipeline (…
SahilCarterr Sep 22, 2025
d8310a8
[lora] factor out the overlaps in `save_lora_weights()`. (#12027)
sayakpaul Sep 22, 2025
5796735
add test and doc for QwenImageEdit Plus (#12363)
sayakpaul Sep 22, 2025
1448b03
[Fix] chroma docs (#12360)
SahilCarterr Sep 22, 2025
76810ec
[docs] Schedulers (#12246)
stevhliu Sep 23, 2025
80de641
Allow Automodel to support custom model code (#12353)
DN6 Sep 23, 2025
a72bc0c
[docs] Attention backends (#12320)
stevhliu Sep 23, 2025
09e777a
[tests] Single scheduler in lora tests (#12315)
sayakpaul Sep 24, 2025
7c54a7b
Fix Custom Code loading (#12378)
DN6 Sep 24, 2025
9ef1185
[tests] disable xformer tests for pipelines it isn't popular. (#12277)
sayakpaul Sep 24, 2025
7a58734
xpu enabling for 4 cases (#12345)
yao-matrix Sep 24, 2025
08c2902
fix marigold ut case fail on xpu (#12350)
yao-matrix Sep 24, 2025
043ab25
Fix WanVACEPipeline to allow prompt to be None and skip encoding step…
albchim Sep 24, 2025
dcb6dd9
Context Parallel w/ Ring & Ulysses & Unified Attention (#11941)
a-r-r-o-w Sep 24, 2025
310fdaf
Introduce cache-dit to community optimization (#12366)
DefTruth Sep 24, 2025
ec5449f
Support both huggingface_hub `v0.x` and `v1.x` (#12389)
Wauplin Sep 25, 2025
4588bbe
[CI] disable installing transformers from main in ci for now. (#12397)
sayakpaul Sep 26, 2025
9c09445
[docs] slight edits to the attention backends docs. (#12394)
sayakpaul Sep 26, 2025
041501a
[docs] remove docstrings from repeated methods in `lora_pipeline.py` …
sayakpaul Sep 26, 2025
19085ac
Don't skip Qwen model tests for group offloading with disk (#12382)
sayakpaul Sep 29, 2025
0a15111
Fix #12116: preserve boolean dtype for attention masks in ChromaPipe…
akshay-babbar Sep 29, 2025
64a5187
[quantization] feat: support aobaseconfig classes in `TorchAOConfig` …
sayakpaul Sep 29, 2025
ccedeca
[docs] Distributed inference (#12285)
stevhliu Sep 29, 2025
c07fcf7
[docs] Model formats (#12256)
stevhliu Sep 29, 2025
76d4e41
[modular]some small fix (#12307)
yiyixuxu Sep 29, 2025
20fd00b
[Tests] Add single file tester mixin for Models and remove unittest d…
DN6 Sep 30, 2025
0e12ba7
fix 3 xpu failures uts w/ latest pytorch (#12408)
yao-matrix Sep 30, 2025
b596545
Install latest prerelease from huggingface_hub when installing transf…
Wauplin Sep 30, 2025
d7a1a03
[docs] CP (#12331)
stevhliu Sep 30, 2025
cc5b31f
[docs] Migrate syntax (#12390)
stevhliu Sep 30, 2025
814d710
[tests] cache non lora pipeline outputs. (#12298)
sayakpaul Oct 1, 2025
9ae5b62
[ci] xfail failing tests in CI. (#12418)
sayakpaul Oct 2, 2025
b429796
[core] conditionally import torch distributed stuff. (#12420)
sayakpaul Oct 2, 2025
7242b5f
FIX Test to ignore warning for enable_lora_hotswap (#12421)
BenjaminBossan Oct 2, 2025
941ac9c
[training-scripts] Make more examples UV-compatible (follow up on #12…
linoytsaban Oct 3, 2025
2b7deff
fix scale_shift_factor being on cpu for wan and ltx (#12347)
vladmandic Oct 5, 2025
c3675d4
[core] support QwenImage Edit Plus in modular (#12416)
sayakpaul Oct 5, 2025
ce90f9b
[FIX] Text to image training peft version (#12434)
SahilCarterr Oct 6, 2025
7f3e9b8
make flux ready for mellon (#12419)
sayakpaul Oct 6, 2025
cf4b97b
[perf] Cache version checks (#12399)
cbensimon Oct 6, 2025
0974b4c
[i18n-KO] Fix typo and update translation in ethical_guidelines.md (#…
braintrue Oct 6, 2025
2d69bac
handle offload_state_dict when initing transformers models (#12438)
sayakpaul Oct 7, 2025
1066de8
[Qwen LoRA training] fix bug when offloading (#12440)
linoytsaban Oct 7, 2025
2dc3167
Align Flux modular more and more with Qwen modular (#12445)
sayakpaul Oct 8, 2025
35e538d
fix dockerfile definitions. (#12424)
sayakpaul Oct 8, 2025
345864e
fix more torch.distributed imports (#12425)
sayakpaul Oct 8, 2025
a519272
[ci] revisit the installations in CI. (#12450)
sayakpaul Oct 8, 2025
a9df12a
Update Dockerfile to include zip wget for doc-builder (#12451)
sayakpaul Oct 9, 2025
693d8a3
[modular] i2i and t2i support for kontext modular (#12454)
sayakpaul Oct 10, 2025
8abc7ae
[docs] Fix syntax (#12464)
stevhliu Oct 11, 2025
fa468c5
docs: api-pipelines-qwenimage typo fix (#12461)
Manith-Ratnayake Oct 13, 2025
a4bc845
Fix missing load_video documentation and load_video import in WanVide…
Meatfucker Oct 14, 2025
3eb4078
[docs] Prompting (#12312)
stevhliu Oct 14, 2025
b4e6dc3
[docs] Fix broken links (#12487)
stevhliu Oct 15, 2025
53a1051
remove unneeded checkpoint imports. (#12488)
sayakpaul Oct 15, 2025
aed636f
[tests] fix clapconfig for text backbone in audioldm2 (#12490)
sayakpaul Oct 15, 2025
f072c64
ltx0.9.8 (without IC lora, autoregressive sampling) (#12493)
yiyixuxu Oct 15, 2025
2647508
[docs] Attention checks (#12486)
stevhliu Oct 16, 2025
dbe4136
[CI] Check links (#12491)
stevhliu Oct 16, 2025
4715c5c
[ci] xfail more incorrect transformer imports. (#12455)
sayakpaul Oct 17, 2025
af76988
[tests] introduce `VAETesterMixin` to consolidate tests for slicing a…
sayakpaul Oct 17, 2025
1b456bd
docs: cleanup of runway model (#12503)
EazyAl Oct 17, 2025
23ebbb4
Kandinsky 5 is finally in Diffusers! (#12478)
leffff Oct 18, 2025
7853bfb
Remove Qwen Image Redundant RoPE Cache (#12452)
dg845 Oct 20, 2025
4830575
Raise warning instead of error when imports are missing for custom co…
DN6 Oct 20, 2025
decfa3c
Fix: Use incorrect temporary variable key when replacing adapter name…
FeiXie8 Oct 21, 2025
5b5fa49
[docs] Organize toctree by modality (#12514)
stevhliu Oct 21, 2025
b3e56e7
styling issues. (#12522)
sayakpaul Oct 21, 2025
cefc2cf
Add Photon model and pipeline support (#12456)
DavidBert Oct 21, 2025
57636ad
purge HF_HUB_ENABLE_HF_TRANSFER; promote Xet (#12497)
Vaibhavs10 Oct 21, 2025
dd07b19
Prx (#12525)
DavidBert Oct 22, 2025
a5a0ccf
[core] `AutoencoderMixin` to abstract common methods (#12473)
sayakpaul Oct 22, 2025
a0a51eb
Kandinsky5 No cfg fix (#12527)
asomoza Oct 22, 2025
bec2d8e
Fix: Add _skip_keys for AutoencoderKLWan (#12523)
yiyixuxu Oct 22, 2025
4c723d8
[CI] xfail the test_wuerstchen_prior test (#12530)
sayakpaul Oct 22, 2025
ccdd96c
[tests] Test attention backends (#12388)
sayakpaul Oct 23, 2025
85eb505
fix CI bug for kandinsky3_img2img case (#12474)
kaixuanliu Oct 23, 2025
74b5fed
Fix MPS compatibility in get_1d_sincos_pos_embed_from_grid #12432 (#1…
Aishwarya0811 Oct 23, 2025
9c3b58d
Handle deprecated transformer classes (#12517)
DN6 Oct 23, 2025
bc40398
fix constants.py to user `upper()` (#12479)
sayakpaul Oct 24, 2025
a138d71
HunyuanImage21 (#12333)
yiyixuxu Oct 24, 2025
7536f64
Loose the criteria tolerance appropriately for Intel XPU devices (#12…
kaixuanliu Oct 24, 2025
d34b18c
Deprecate Stable Cascade (#12537)
DN6 Oct 24, 2025
500b9cf
[chore] Move guiders experimental warning (#12543)
sayakpaul Oct 26, 2025
dc6bd15
Fix Chroma attention padding order and update docs to use `lodestones…
josephrocca Oct 27, 2025
250f5cb
Add AITER attention backend (#12549)
lauri9 Oct 27, 2025
6d1a648
Fix small inconsistency in output dimension of "_get_t5_prompt_embeds…
alirezafarashah Oct 27, 2025
5afbcce
Kandinsky 5 10 sec (NABLA suport) (#12520)
leffff Oct 28, 2025
303efd2
Improve pos embed for Flux.1 inference on Ascend NPU (#12534)
gameofdimension Oct 28, 2025
df0e2a4
support latest few-step wan LoRA. (#12541)
sayakpaul Oct 28, 2025
ecfbc8f
[Pipelines] Enable Wan VACE to run since single transformer (#12428)
DN6 Oct 28, 2025
dc622a9
fix crash if tiling mode is enabled (#12521)
sywangyi Oct 28, 2025
40528e9
Fix typos in kandinsky5 docs (#12552)
Meatfucker Oct 28, 2025
55d49d4
[ci] don't run sana layerwise casting tests in CI. (#12551)
sayakpaul Oct 28, 2025
84e1657
Bria fibo (#12545)
galbria Oct 28, 2025
9f3c0fd
Avoiding graph break by changing the way we infer dtype in vae.decode…
ppadjinTT Oct 30, 2025
df8dd77
[Modular] Fix for custom block kwargs (#12561)
DN6 Oct 30, 2025
d54622c
[Modular] Allow custom blocks to be saved to `local_dir` (#12381)
DN6 Oct 31, 2025
051c8a1
Fix Stable Diffusion 3.x pooled prompt embedding with multiple images…
friedrich Oct 31, 2025
5e8415a
Fix custom code loading in Automodel (#12571)
DN6 Nov 2, 2025
cdbf0ad
[modular] better warn message (#12573)
yiyixuxu Nov 2, 2025
8f80dda
[tests] add tests for flux modular (t2i, i2i, kontext) (#12566)
sayakpaul Nov 2, 2025
de6173c
[modular]pass hub_kwargs to load_config (#12577)
yiyixuxu Nov 3, 2025
1ec28a2
ulysses enabling in native attention path (#12563)
sywangyi Nov 3, 2025
325a950
Kandinsky 5.0 Docs fixes (#12582)
leffff Nov 4, 2025
ac5a1e2
[docs] sort doc (#12586)
sayakpaul Nov 4, 2025
dcfb18a
[LoRA] add support for more Qwen LoRAs (#12581)
linoytsaban Nov 4, 2025
6198f8a
[Modular] Allow ModularPipeline to load from revisions (#12592)
DN6 Nov 6, 2025
58f3771
Add optional precision-preserving preprocessing for examples/uncondit…
turian Nov 6, 2025
b3e9dfc
[SANA-Video] Adding 5s pre-trained 480p SANA-Video inference (#12584)
lawrence-cj Nov 6, 2025
e4393fa
Fix overflow and dtype handling in rgblike_to_depthmap (NumPy + PyTor…
MohammadSadeghSalehi Nov 6, 2025
8ac17cd
[Modular] Some clean up for Modular tests (#12579)
DN6 Nov 7, 2025
9f669e7
feat: enable attention dispatch for huanyuan video (#12591)
DefTruth Nov 7, 2025
a9cb08a
fix the crash in Wan-AI/Wan2.2-TI2V-5B-Diffusers if CP is enabled (#1…
sywangyi Nov 7, 2025
bc8fd86
[CI] Push test fix (#12617)
DN6 Nov 10, 2025
04f9d2b
add ChronoEdit (#12593)
zhangjiewu Nov 10, 2025
b455dc9
[modular] wan! (#12611)
yiyixuxu Nov 10, 2025
35d7033
[CI] Fix typo in uv install (#12618)
DN6 Nov 10, 2025
0fd58c7
fix: correct import path for load_model_dict_into_meta in conversion …
yashwantbezawada Nov 10, 2025
8d45f21
Fix Context Parallel validation checks (#12446)
DN6 Nov 10, 2025
8f6328c
[Modular] Clean up docs (#12604)
DN6 Nov 10, 2025
5a47442
Fix: update type hints for Tuple parameters across multiple files to …
cesaryuan Nov 10, 2025
66e6a02
[CI] Remove unittest dependency from `testing_utils.py` (#12621)
DN6 Nov 11, 2025
ff26394
Fix rotary positional embedding dimension mismatch in Wan and SkyReel…
charchit7 Nov 11, 2025
0c75892
fix copies (#12637)
yiyixuxu Nov 12, 2025
aecf0c5
Add MLU Support. (#12629)
a120092009 Nov 12, 2025
093cd3f
fix dispatch_attention_fn check (#12636)
yiyixuxu Nov 12, 2025
f5e5f34
[modular] add tests for qwen modular (#12585)
sayakpaul Nov 12, 2025
f3db38c
ArXiv -> HF Papers (#12583)
qgallouedec Nov 12, 2025
2f44d63
[docs] Update install instructions (#12626)
stevhliu Nov 12, 2025
d6c63bb
[modular] add a check (#12628)
yiyixuxu Nov 12, 2025
44c3101
Improve docstrings and type hints in scheduling_amused.py (#12623)
delmalih Nov 13, 2025
d8e4805
[WIP]Add Wan2.2 Animate Pipeline (Continuation of #12442 by tolgacang…
dg845 Nov 13, 2025
7a001c3
adjust unit tests for `test_save_load_float16` (#12500)
kaixuanliu Nov 13, 2025
cd3bbe2
skip autoencoderdl layerwise casting memory (#12647)
sayakpaul Nov 13, 2025
6a2309b
[utils] Update check_doc_toc (#12642)
stevhliu Nov 13, 2025
40de88a
[docs] AutoModel (#12644)
stevhliu Nov 13, 2025
6fe4a6f
Improve docstrings and type hints in scheduling_ddim.py (#12622)
delmalih Nov 13, 2025
3c1ca86
Improve docstrings and type hints in scheduling_ddpm.py (#12651)
delmalih Nov 13, 2025
eeae033
[Modular] Add Custom Blocks guide to doc (#12339)
DN6 Nov 14, 2025
63dd601
Improve docstrings and type hints in scheduling_euler_discrete.py (#1…
delmalih Nov 14, 2025
a9e4883
Update Wan Animate Docs (#12658)
dg845 Nov 15, 2025
01a5692
Rope in float32 for mps or npu compatibility (#12665)
DavidBert Nov 15, 2025
0c35b58
[PRX pipeline]: add 1024 resolution ratio bins (#12670)
DavidBert Nov 17, 2025
1afc218
SANA-Video Image to Video pipeline `SanaImageToVideoPipeline` support…
lawrence-cj Nov 17, 2025
3579fda
[CI] Make CI logs less verbose (#12674)
DN6 Nov 17, 2025
67dc65e
Revert `AutoencoderKLWan`'s `dim_mult` default value back to list (#1…
dg845 Nov 17, 2025
b7df4a5
[CI] Temporarily pin transformers (#12677)
DN6 Nov 18, 2025
ab71f3c
[core] Refactor hub attn kernels (#12475)
sayakpaul Nov 19, 2025
6d8973f
[CI] Fix indentation issue in workflow files (#12685)
DN6 Nov 19, 2025
a96b145
[CI] Fix failing Pipeline CPU tests (#12681)
DN6 Nov 19, 2025
15370f8
Improve docstrings and type hints in scheduling_pndm.py (#12676)
delmalih Nov 19, 2025
d5da453
Community Pipeline: FluxFillControlNetInpaintPipeline for FLUX Fill-B…
pratim4dasude Nov 19, 2025
6f1042e
Improve docstrings and type hints in scheduling_lms_discrete.py (#12678)
delmalih Nov 21, 2025
544ba67
Add FluxLoraLoaderMixin to Fibo pipeline (#12688)
SwayStar123 Nov 24, 2025
354d35a
bugfix: fix chrono-edit context parallel (#12660)
DefTruth Nov 24, 2025
d176f61
[core] support sage attention + FA2 through `kernels` (#12439)
sayakpaul Nov 24, 2025
fbcd3ba
[i8n-pt] Fix grammar and expand Portuguese documentation (#12598)
cdutr Nov 24, 2025
dde8754
Fix variable naming typos in community FluxControlNetFillInpaintPipel…
sqhuang Nov 24, 2025
d33d9f6
fix typo in docs (#12675)
lawrence-cj Nov 25, 2025
4088e8a
Add Support for Z-Image Series (#12703)
JerryWu-code Nov 25, 2025
5ffb73d
let's go Flux2 🚀 (#12711)
sayakpaul Nov 25, 2025
ac78646
Update script names in README for Flux2 training (#12713)
anvilarth Nov 26, 2025
b91e8c0
[lora]: Fix Flux2 LoRA NaN test (#12714)
sayakpaul Nov 26, 2025
94c9613
[docs] Correct flux2 links (#12716)
sayakpaul Nov 26, 2025
c8656ed
[docs] put autopipeline after overview and hunyuanimage in images (#1…
sayakpaul Nov 26, 2025
a88a7b4
Improve docstrings and type hints in scheduling_dpmsolver_multistep.p…
delmalih Nov 26, 2025
e6d4612
Support unittest for Z-image ⚡️ (#12715)
JerryWu-code Nov 26, 2025
6bf668c
[chore] remove torch.save from remnant code. (#12717)
sayakpaul Nov 27, 2025
01e3555
Enable regional compilation on z-image transformer model (#12736)
sayakpaul Nov 27, 2025
1b91856
Fix examples not loading LoRA adapter weights from checkpoint (#12690)
SurAyush Nov 28, 2025
b010a8c
[Modular] Add single file support to Modular (#12383)
DN6 Nov 28, 2025
152f7ca
fix type-check for z-image transformer (#12739)
DefTruth Nov 29, 2025
6156cf8
Hunyuanvideo15 (#12696)
yiyixuxu Dec 1, 2025
c25582d
[Docs] Update Imagen Video paper link in schedulers (#12724)
delmalih Dec 1, 2025
d769d8a
Improve docstrings and type hints in scheduling_heun_discrete.py (#12…
delmalih Dec 1, 2025
859b809
Improve docstrings and type hints in scheduling_euler_ancestral_discr…
delmalih Dec 1, 2025
fffd964
fix FLUX.2 context parallel (#12737)
DefTruth Dec 1, 2025
99784ae
Rename BriaPipeline to BriaFiboPipeline in documentation (#12758)
galbria Dec 1, 2025
394a48d
Update bria_fibo.md with minor fixes (#12731)
sayakpaul Dec 1, 2025
564079f
[feat]: implement "local" caption upsampling for Flux.2 (#12718)
sayakpaul Dec 1, 2025
edf36f5
Add ZImage LoRA support and integrate into ZImagePipeline (#12750)
CalamitousFelicitousness Dec 2, 2025
4f136f8
Add support for Ovis-Image (#12740)
DoctorKey Dec 2, 2025
9379b23
Fix TPU (torch_xla) compatibility Error about tensor repeat func alon…
JerryWu-code Dec 2, 2025
3c05b9f
Fixes #12673. `record_stream` in group offloading is not working prop…
KimbingNg Dec 3, 2025
f48f9c2
[core] start varlen variants for attn backend kernels. (#12765)
sayakpaul Dec 3, 2025
759ea58
[core] reuse `AttentionMixin` for compatible classes (#12463)
sayakpaul Dec 3, 2025
1908c47
Deprecate `upcast_vae` in SDXL based pipelines (#12619)
DN6 Dec 3, 2025
d0c54e5
Kandinsky 5.0 Video Pro and Image Lite (#12664)
leffff Dec 3, 2025
5ab5946
Fix: leaf_level offloading breaks after delete_adapters (#12639)
adi776borate Dec 3, 2025
d96cbac
[tests] fix hunuyanvideo 1.5 offloading tests. (#12782)
sayakpaul Dec 3, 2025
a1f36ee
[Z-Image] various small changes, Z-Image transformer tests, etc. (#12…
sayakpaul Dec 3, 2025
6028613
Z-Image-Turbo `from_single_file` (#12756)
hlky Dec 4, 2025
c318686
Update attention_backends.md to format kernels (#12757)
sayakpaul Dec 4, 2025
2842c14
Improve docstrings and type hints in scheduling_unipc_multistep.py (#…
delmalih Dec 4, 2025
cd00ba6
fix spatial compression ratio error for AutoEncoderKLWan doing tiled …
jerry2102 Dec 4, 2025
7de51b8
[lora] support more ZImage LoRAs (#12790)
sayakpaul Dec 4, 2025
8d415a6
PRX Set downscale_freq_shift to 0 for consistency with internal imple…
DavidBert Dec 4, 2025
f12d161
Fix broken group offloading with block_level for models with standalo…
rycerzes Dec 5, 2025
c98c157
[Docs] Add Z-Image docs (#12775)
asomoza Dec 5, 2025
bb9e713
move kandisnky docs.
sayakpaul Dec 5, 2025
8430ac2
[docs] minor fixes to kandinsky docs (#12797)
sayakpaul Dec 5, 2025
256e010
Improve docstrings and type hints in scheduling_deis_multistep.py (#1…
delmalih Dec 5, 2025
6290fdf
[Feat] TaylorSeer Cache (#12648)
toilaluan Dec 6, 2025
5a74319
Update the TensorRT-ModelOPT to Nvidia-ModelOPT (#12793)
jingyu-ml Dec 8, 2025
f67639b
add post init for safty checker (#12794)
jiqing-feng Dec 8, 2025
671149e
[HunyuanVideo1.5] support step-distilled (#12802)
yiyixuxu Dec 8, 2025
2246d2c
Add ZImageImg2ImgPipeline (#12751)
CalamitousFelicitousness Dec 8, 2025
3d02cd5
[PRX] Improve model compilation (#12787)
WaterKnight1998 Dec 8, 2025
54fa074
Improve docstrings and type hints in scheduling_dpmsolver_singlestep.…
delmalih Dec 8, 2025
c346366
Merge branch 'main' into main
clementchadebec Dec 9, 2025
e664396
Merge branch 'clipdrop-main' into main
clementchadebec Dec 9, 2025
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
The table of contents is too big for display.
Diff view
Diff view
  •  
  •  
  •  
50 changes: 35 additions & 15 deletions .github/workflows/benchmark.yml
Original file line number Diff line number Diff line change
Expand Up @@ -7,24 +7,25 @@ on:

env:
DIFFUSERS_IS_CI: yes
HF_HUB_ENABLE_HF_TRANSFER: 1
HF_XET_HIGH_PERFORMANCE: 1
HF_HOME: /mnt/cache
OMP_NUM_THREADS: 8
MKL_NUM_THREADS: 8
BASE_PATH: benchmark_outputs

jobs:
torch_pipelines_cuda_benchmark_tests:
torch_models_cuda_benchmark_tests:
env:
SLACK_WEBHOOK_URL: ${{ secrets.SLACK_WEBHOOK_URL_BENCHMARK }}
name: Torch Core Pipelines CUDA Benchmarking Tests
name: Torch Core Models CUDA Benchmarking Tests
strategy:
fail-fast: false
max-parallel: 1
runs-on:
group: aws-g6-4xlarge-plus
group: aws-g6e-4xlarge
container:
image: diffusers/diffusers-pytorch-compile-cuda
options: --shm-size "16gb" --ipc host --gpus 0
image: diffusers/diffusers-pytorch-cuda
options: --shm-size "16gb" --ipc host --gpus all
steps:
- name: Checkout diffusers
uses: actions/checkout@v3
Expand All @@ -35,27 +36,46 @@ jobs:
nvidia-smi
- name: Install dependencies
run: |
python -m venv /opt/venv && export PATH="/opt/venv/bin:$PATH"
python -m uv pip install -e [quality,test]
python -m uv pip install pandas peft
python -m uv pip uninstall transformers && python -m uv pip install transformers==4.48.0
apt update
apt install -y libpq-dev postgresql-client
uv pip install -e ".[quality]"
uv pip install -r benchmarks/requirements.txt
- name: Environment
run: |
python utils/print_env.py
- name: Diffusers Benchmarking
env:
HF_TOKEN: ${{ secrets.DIFFUSERS_BOT_TOKEN }}
BASE_PATH: benchmark_outputs
HF_TOKEN: ${{ secrets.DIFFUSERS_HF_HUB_READ_TOKEN }}
run: |
export TOTAL_GPU_MEMORY=$(python -c "import torch; print(torch.cuda.get_device_properties(0).total_memory / (1024**3))")
cd benchmarks && mkdir ${BASE_PATH} && python run_all.py && python push_results.py
cd benchmarks && python run_all.py

- name: Push results to the Hub
env:
HF_TOKEN: ${{ secrets.DIFFUSERS_BOT_TOKEN }}
run: |
cd benchmarks && python push_results.py
mkdir $BASE_PATH && cp *.csv $BASE_PATH

- name: Test suite reports artifacts
if: ${{ always() }}
uses: actions/upload-artifact@v4
with:
name: benchmark_test_reports
path: benchmarks/benchmark_outputs
path: benchmarks/${{ env.BASE_PATH }}

# TODO: enable this once the connection problem has been resolved.
- name: Update benchmarking results to DB
env:
PGDATABASE: metrics
PGHOST: ${{ secrets.DIFFUSERS_BENCHMARKS_PGHOST }}
PGUSER: transformers_benchmarks
PGPASSWORD: ${{ secrets.DIFFUSERS_BENCHMARKS_PGPASSWORD }}
BRANCH_NAME: ${{ github.head_ref || github.ref_name }}
run: |
git config --global --add safe.directory /__w/diffusers/diffusers
commit_id=$GITHUB_SHA
commit_msg=$(git show -s --format=%s "$commit_id" | cut -c1-70)
cd benchmarks && python populate_into_db.py "$BRANCH_NAME" "$commit_id" "$commit_msg"

- name: Report success status
if: ${{ success() }}
Expand Down
45 changes: 34 additions & 11 deletions .github/workflows/build_docker_images.yml
Original file line number Diff line number Diff line change
Expand Up @@ -38,15 +38,43 @@ jobs:
token: ${{ secrets.GITHUB_TOKEN }}

- name: Build Changed Docker Images
env:
CHANGED_FILES: ${{ steps.file_changes.outputs.all }}
run: |
CHANGED_FILES="${{ steps.file_changes.outputs.all }}"
echo "$CHANGED_FILES"
ALLOWED_IMAGES=(
diffusers-pytorch-cpu
diffusers-pytorch-cuda
diffusers-pytorch-xformers-cuda
diffusers-pytorch-minimum-cuda
diffusers-doc-builder
)

declare -A IMAGES_TO_BUILD=()

for FILE in $CHANGED_FILES; do
if [[ "$FILE" == docker/*Dockerfile ]]; then
DOCKER_PATH="${FILE%/Dockerfile}"
DOCKER_TAG=$(basename "$DOCKER_PATH")
echo "Building Docker image for $DOCKER_TAG"
docker build -t "$DOCKER_TAG" "$DOCKER_PATH"
# skip anything that isn't still on disk
if [[ ! -e "$FILE" ]]; then
echo "Skipping removed file $FILE"
continue
fi

for IMAGE in "${ALLOWED_IMAGES[@]}"; do
if [[ "$FILE" == docker/${IMAGE}/* ]]; then
IMAGES_TO_BUILD["$IMAGE"]=1
fi
done
done

if [[ ${#IMAGES_TO_BUILD[@]} -eq 0 ]]; then
echo "No relevant Docker changes detected."
exit 0
fi

for IMAGE in "${!IMAGES_TO_BUILD[@]}"; do
DOCKER_PATH="docker/${IMAGE}"
echo "Building Docker image for $IMAGE"
docker build -t "$IMAGE" "$DOCKER_PATH"
done
if: steps.file_changes.outputs.all != ''

Expand All @@ -65,13 +93,8 @@ jobs:
image-name:
- diffusers-pytorch-cpu
- diffusers-pytorch-cuda
- diffusers-pytorch-compile-cuda
- diffusers-pytorch-xformers-cuda
- diffusers-pytorch-minimum-cuda
- diffusers-flax-cpu
- diffusers-flax-tpu
- diffusers-onnxruntime-cpu
- diffusers-onnxruntime-cuda
- diffusers-doc-builder

steps:
Expand Down
26 changes: 26 additions & 0 deletions .github/workflows/build_pr_documentation.yml
Original file line number Diff line number Diff line change
Expand Up @@ -12,7 +12,33 @@ concurrency:
cancel-in-progress: true

jobs:
check-links:
runs-on: ubuntu-latest

steps:
- name: Checkout repository
uses: actions/checkout@v4

- name: Set up Python
uses: actions/setup-python@v5
with:
python-version: '3.10'

- name: Install uv
run: |
curl -LsSf https://astral.sh/uv/install.sh | sh
echo "$HOME/.cargo/bin" >> $GITHUB_PATH

- name: Install doc-builder
run: |
uv pip install --system git+https://github.com/huggingface/doc-builder.git@main

- name: Check documentation links
run: |
uv run doc-builder check-links docs/source/en

build:
needs: check-links
uses: huggingface/doc-builder/.github/workflows/build_pr_documentation.yml@main
with:
commit_sha: ${{ github.event.pull_request.head.sha }}
Expand Down
6 changes: 3 additions & 3 deletions .github/workflows/mirror_community_pipeline.yml
Original file line number Diff line number Diff line change
Expand Up @@ -74,19 +74,19 @@ jobs:
python-version: "3.10"
- name: Install dependencies
run: |
python -m pip install --upgrade pip
pip install --upgrade pip
pip install --upgrade huggingface_hub

# Check secret is set
- name: whoami
run: huggingface-cli whoami
run: hf auth whoami
env:
HF_TOKEN: ${{ secrets.HF_TOKEN_MIRROR_COMMUNITY_PIPELINES }}

# Push to HF! (under subfolder based on checkout ref)
# https://huggingface.co/datasets/diffusers/community-pipelines-mirror
- name: Mirror community pipeline to HF
run: huggingface-cli upload diffusers/community-pipelines-mirror ./examples/community ${PATH_IN_REPO} --repo-type dataset
run: hf upload diffusers/community-pipelines-mirror ./examples/community ${PATH_IN_REPO} --repo-type dataset
env:
PATH_IN_REPO: ${{ env.PATH_IN_REPO }}
HF_TOKEN: ${{ secrets.HF_TOKEN_MIRROR_COMMUNITY_PIPELINES }}
Expand Down
Loading
Loading