Commit Graph

  • 6290fdfda4 [Feat] TaylorSeer Cache (#12648) main Tran Thanh Luan 2025-12-06 07:09:54 +07:00
  • 256e010674 Improve docstrings and type hints in scheduling_deis_multistep.py (#12796) David El Malih 2025-12-05 17:48:01 +01:00
  • 8430ac2a2f [docs] minor fixes to kandinsky docs (#12797) Sayak Paul 2025-12-06 00:33:05 +08:00
  • 2532668363 up kandinsky-docs-update sayakpaul 2025-12-05 21:47:31 +07:00
  • bb9e713d02 move kandisnky docs. sayakpaul 2025-12-05 21:44:24 +07:00
  • c98c157a9e [Docs] Add Z-Image docs (#12775) Álvaro Somoza 2025-12-05 11:05:47 -03:00
  • 9de5739fb4 Merge branch 'main' into disable-mmap-pipeline disable-mmap-pipeline Dhruv Nair 2025-12-05 19:06:54 +05:30
  • 9a91821bd2 fix example usage. toilaluan-feat-taylorseer sayakpaul 2025-12-05 20:34:09 +07:00
  • a9c59b7de6 make fix-copies sayakpaul 2025-12-05 20:31:31 +07:00
  • 3bf9f9d281 Merge branch 'main' into feat-taylorseer Sayak Paul 2025-12-05 21:25:35 +08:00
  • f12d161d67 Fix broken group offloading with block_level for models with standalone layers (#12692) swappy 2025-12-05 18:54:05 +05:30
  • 5229769a94 update docs toilaluan 2025-12-05 07:31:57 +00:00
  • d009d451c2 Update src/diffusers/hooks/taylorseer_cache.py Tran Thanh Luan 2025-12-05 14:26:25 +07:00
  • 8d415a6f48 PRX Set downscale_freq_shift to 0 for consistency with internal implementation (#12791) David Bertoin 2025-12-04 21:57:14 +01:00
  • 7de51b826c [lora] support more ZImage LoRAs (#12790) Sayak Paul 2025-12-05 03:01:11 +08:00
  • cd00ba685b fix spatial compression ratio error for AutoEncoderKLWan doing tiled encode (#12753) Jiang 2025-12-05 02:57:13 +08:00
  • 2842c14c5f Improve docstrings and type hints in scheduling_unipc_multistep.py (#12767) David El Malih 2025-12-04 19:10:54 +01:00
  • 54ca26f44a Test hfh v1.2.0.rc0 ci-test-huggingface-hub-v1.2.0.rc0-release Hugging Face Bot (RC Testing) 2025-12-04 15:49:39 +00:00
  • c318686090 Update attention_backends.md to format kernels (#12757) Sayak Paul 2025-12-04 23:48:23 +08:00
  • 6028613226 Z-Image-Turbo from_single_file (#12756) hlky 2025-12-04 14:52:48 +00:00
  • f956ba0db1 resolve conflicts. migrate-lora-pytest sayakpaul 2025-12-04 20:07:15 +08:00
  • ca24569a2a Apply style fixes github-actions[bot] 2025-12-04 11:16:52 +00:00
  • 76494ca098 torch compile compatible toilaluan 2025-12-04 06:12:37 +00:00
  • 10d798a2d6 fix qwen-pipeline-mixin sayakpaul 2025-12-04 09:48:41 +08:00
  • 6e05c21f44 remove. sayakpaul 2025-12-04 09:45:17 +08:00
  • 3be4b01cda remove methods again sayakpaul 2025-12-04 09:43:17 +08:00
  • 16d67c4ae0 resolve conflicts for now, refactor again later in a bit. sayakpaul 2025-12-04 09:40:12 +08:00
  • a1f36ee3ef [Z-Image] various small changes, Z-Image transformer tests, etc. (#12741) Sayak Paul 2025-12-03 22:05:46 +08:00
  • d96cbacacd [tests] fix hunuyanvideo 1.5 offloading tests. (#12782) Sayak Paul 2025-12-03 20:37:59 +08:00
  • 5ab5946931 Fix: leaf_level offloading breaks after delete_adapters (#12639) Aditya Borate 2025-12-03 17:39:11 +05:30
  • d0c54e5563 Kandinsky 5.0 Video Pro and Image Lite (#12664) Lev Novitskiy 2025-12-03 13:46:37 +03:00
  • 4fb3f53b6c rename identifiers, use more expressive taylor predict loop toilaluan 2025-12-03 10:43:01 +00:00
  • 1908c47600 Deprecate upcast_vae in SDXL based pipelines (#12619) Dhruv Nair 2025-12-03 15:53:23 +05:30
  • f3593a8aa9 up sayakpaul 2025-12-03 17:42:36 +08:00
  • 1b6cdea043 Merge branch 'main' into migrate-lora-pytest Sayak Paul 2025-12-03 17:39:35 +08:00
  • 68eee98578 update sdxl-vae-fix DN6 2025-12-03 14:04:14 +05:30
  • 759ea58708 [core] reuse AttentionMixin for compatible classes (#12463) Sayak Paul 2025-12-03 16:28:33 +08:00
  • f48f9c250f [core] start varlen variants for attn backend kernels. (#12765) Sayak Paul 2025-12-03 16:04:52 +08:00
  • 3c05b9f71c Fixes #12673. record_stream in group offloading is not working properly (#12721) Kimbing Ng 2025-12-03 14:07:11 +08:00
  • 9379b2391b Fix TPU (torch_xla) compatibility Error about tensor repeat func along with empty dim. (#12770) Jerry Wu 2025-12-03 06:51:23 +08:00
  • 4f136f842c Add support for Ovis-Image (#12740) Guo-Hua Wang 2025-12-03 05:48:07 +08:00
  • e1ac8fd23e make fix-copies sayakpaul 2025-12-02 23:35:12 +08:00
  • e326b254d4 Merge branch 'main' into qwen-pipeline-mixin Sayak Paul 2025-12-02 21:01:31 +05:30
  • d287428460 Update tests/pipelines/qwenimage/test_qwenimage_edit_plus.py prompt-isolation-tests-qwen Sayak Paul 2025-12-02 19:08:05 +05:30
  • 93a470be0a Apply suggestions from code review Sayak Paul 2025-12-02 19:06:48 +05:30
  • 3eb1f0efe9 resolve conflicts custom-modular-tests sayakpaul 2025-12-02 21:29:28 +08:00
  • c541f9a49f Merge branch 'main' into prompt-isolation-tests-qwen Sayak Paul 2025-12-02 18:56:19 +05:30
  • edf36f5128 Add ZImage LoRA support and integrate into ZImagePipeline (#12750) CalamitousFelicitousness 2025-12-02 05:16:30 +00:00
  • 564079f295 [feat]: implement "local" caption upsampling for Flux.2 (#12718) Sayak Paul 2025-12-02 04:27:24 +05:30
  • 394a48d169 Update bria_fibo.md with minor fixes (#12731) Sayak Paul 2025-12-02 04:24:19 +05:30
  • 99784ae0d2 Rename BriaPipeline to BriaFiboPipeline in documentation (#12758) Gal Davidi 2025-12-01 21:34:47 +02:00
  • fffd964a0f fix FLUX.2 context parallel (#12737) DefTruth 2025-12-02 03:07:49 +08:00
  • 859b809031 Improve docstrings and type hints in scheduling_euler_ancestral_discrete.py (#12766) David El Malih 2025-12-01 19:38:01 +01:00
  • d769d8a13b Improve docstrings and type hints in scheduling_heun_discrete.py (#12726) David El Malih 2025-12-01 17:09:36 +01:00
  • c25582d509 [Docs] Update Imagen Video paper link in schedulers (#12724) David El Malih 2025-12-01 17:09:22 +01:00
  • 6156cf8f22 Hunyuanvideo15 (#12696) YiYi Xu 2025-11-30 20:27:59 -10:00
  • cfd857f3b3 up hunyuanvideo15-tests Sayak Paul 2025-12-01 05:55:00 +00:00
  • 921b959b9a update flux2-modular DN6 2025-12-01 10:42:35 +05:30
  • 475ec02d8c Remove TaylorSeerCacheTesterMixin from flux2 tests Tran Thanh Luan 2025-12-01 09:31:23 +07:00
  • 510cc66a2d Merge branch 'hunyuanvideo15' into hunyuanvideo15-tests Sayak Paul 2025-12-01 07:10:27 +05:30
  • 801f82c8a9 style. sayakpaul 2025-12-01 09:39:36 +08:00
  • 5efb5efbf2 up Sayak Paul 2025-12-01 01:39:05 +00:00
  • 404d3fa9a5 Update docs/source/en/api/pipelines/hunyuan_video15.md YiYi Xu 2025-11-30 15:24:18 -10:00
  • 5989014cfe Update docs/source/en/api/pipelines/hunyuan_video15.md YiYi Xu 2025-11-30 15:21:17 -10:00
  • 0dae8f956d Apply suggestions from code review YiYi Xu 2025-11-30 15:10:16 -10:00
  • c715470709 add a note on changing guidance_scale on doc yiyixuxu 2025-12-01 02:06:13 +01:00
  • 5f9a215178 up Sayak Paul 2025-12-01 00:51:38 +00:00
  • 2c018f8be6 Update docs/source/en/api/pipelines/hunyuan_video15.md YiYi Xu 2025-11-30 14:43:51 -10:00
  • 159270b067 Merge branch 'hunyuanvideo15' into hunyuanvideo15-tests Sayak Paul 2025-12-01 06:09:48 +05:30
  • bdfab30766 up yiyi@huggingface.co 2025-12-01 00:35:30 +00:00
  • d7f399d1b2 add a notes on the doc about attention backend yiyi@huggingface.co 2025-12-01 00:33:00 +00:00
  • 237d318e85 Apply suggestions from code review YiYi Xu 2025-11-30 13:22:43 -10:00
  • 7c100154f8 start tests. sayakpaul 2025-12-01 07:19:05 +08:00
  • 54f008e30b Update docs/source/en/api/pipelines/hunyuan_video15.md YiYi Xu 2025-11-30 13:17:31 -10:00
  • 289146e73e Apply style fixes github-actions[bot] 2025-11-30 21:40:03 +00:00
  • c3f45982b6 up yiyi@huggingface.co 2025-11-30 20:40:31 +00:00
  • 7aeab3f847 add to toctree yiyi@huggingface.co 2025-11-30 20:39:49 +00:00
  • 50abf504a1 add docs yiyi@huggingface.co 2025-11-30 20:38:31 +00:00
  • 8aa458ed46 copies yiyi@huggingface.co 2025-11-30 20:16:59 +00:00
  • 5029dbf763 style yiyi@huggingface.co 2025-11-30 20:14:36 +00:00
  • 3980f97d2f Merge branch 'main' into hunyuanvideo15 YiYi Xu 2025-11-30 09:54:37 -10:00
  • e1940341ff clean up a bit more pipelines yiyi@huggingface.co 2025-11-30 19:46:38 +00:00
  • e319d7207a simplify transformer yiyi@huggingface.co 2025-11-30 19:24:00 +00:00
  • f9cb82b64f a few small fix: proprocess, cpu_offloading, attention backend yiyi@huggingface.co 2025-11-30 18:38:33 +00:00
  • cd7e281b83 Update attention_backends.md to format kernels sayakpaul-patch-3 Sayak Paul 2025-11-30 20:13:15 +05:30
  • 152f7ca357 fix type-check for z-image transformer (#12739) DefTruth 2025-11-29 17:28:33 +08:00
  • e2dae7e432 add tests toilaluan 2025-11-29 07:21:01 +00:00
  • c22915d6c4 up up yiyi@huggingface.co 2025-11-29 05:36:33 +00:00
  • 0687a40768 remove use_meanflow yiyi@huggingface.co 2025-11-29 05:36:22 +00:00
  • 753d4075f9 add image to video pipeline yiyi@huggingface.co 2025-11-29 01:59:44 +00:00
  • e3301cbda4 add i2v pipeline yiyi@huggingface.co 2025-11-29 00:49:18 +00:00
  • 090ceb5d4f remove dtype from the _get_ encodeing methods yiyi@huggingface.co 2025-11-29 00:33:58 +00:00
  • 9391a5465d Merge branch 'main' into flux2-modular DN6 2025-11-28 22:23:38 +05:30
  • b010a8ce0c [Modular] Add single file support to Modular (#12383) Dhruv Nair 2025-11-28 22:23:04 +05:30
  • 15fb54de11 update sf-modular DN6 2025-11-28 21:34:32 +05:30
  • 6cf9c6ed3a Merge branch 'main' into qwen-pipeline-mixin Sayak Paul 2025-11-28 20:39:43 +05:30
  • 207847552d update DN6 2025-11-28 18:52:39 +05:30
  • 2cd7256ecb update DN6 2025-11-28 18:29:02 +05:30
  • 716dfe1468 Apply style fixes github-actions[bot] 2025-11-28 12:52:07 +00:00
  • ddc6164d18 Merge branch 'main' into feat-taylorseer Tran Thanh Luan 2025-11-28 19:29:04 +07:00