1
0
mirror of https://github.com/huggingface/diffusers.git synced 2026-01-29 07:22:12 +03:00

Commit Graph

  • 27c1158b23 add a to-do for guider cconfig mixin yiyixuxu 2025-05-18 18:50:03 +02:00
  • 3c5a4857e0 Revert error to warning when loading LoRA from repo with multiple weights revert-error-to-warning apolinário 2025-05-17 05:58:25 +02:00
  • 8ceed7d3ae Initial commit: Chroma as a FLUX.1 variant. Hameer Abbasi 2025-05-17 05:20:26 +05:00
  • b8317da20f remove central registry based on review Aryan 2025-05-16 12:53:36 +02:00
  • a5fe2bd4fd remove unused function Aryan 2025-05-16 12:37:01 +02:00
  • 153cf0c393 fix Aryan 2025-05-16 12:33:52 +02:00
  • 495fddb8ae support mochi Aryan 2025-05-15 21:42:19 +02:00
  • 367fdef96d support cogvideox Aryan 2025-05-15 21:41:45 +02:00
  • 82fa9df1c7 Merge branch 'main' into integrations/first-block-cache-2 Aryan 2025-05-15 21:38:57 +02:00
  • 9836f0e000 [docs] Regional compilation docs (#11556) Sayak Paul 2025-05-15 19:11:24 +05:30
  • 20379d9d13 [tests] add tests for combining layerwise upcasting and groupoffloading. (#11558) Sayak Paul 2025-05-15 17:16:44 +05:30
  • 3a6caba8e4 [gguf] Refactor __torch_function__ to avoid unnecessary computation (#11551) Animesh Jain 2025-05-15 02:08:18 -07:00
  • 4267d8f4eb [Single File] GGUF/Single File Support for HiDream (#11550) Dhruv Nair 2025-05-15 08:55:18 +02:00
  • 8a12103891 Merge branch 'main' into training-group-offloading-tests training-group-offloading-tests Sayak Paul 2025-05-15 10:08:59 +05:30
  • 96ce6744fe after_denoise -> decoders yiyixuxu 2025-05-15 00:45:45 +02:00
  • 4bac6acce6 Merge branch 'main' into hidream-torch-compile Sayak Paul 2025-05-14 20:26:44 +05:30
  • fb229b54bb metadata registration with decorators instead of centralized Aryan 2025-05-14 14:19:48 +02:00
  • 0a44380a36 address review comments Aryan 2025-05-14 12:14:24 +02:00
  • 2ed59c178d Merge branch 'main' into integrations/first-block-cache-2 Aryan 2025-05-14 11:48:12 +02:00
  • f4fa3beee7 [tests] Add torch.compile test for UNet2DConditionModel (#11537) Seokhyeon Jeong 2025-05-14 14:56:12 +09:00
  • 8ad14a52cb make generator intermediates (it is mutable) yiyixuxu 2025-05-13 23:25:56 +02:00
  • a7fb2d2a22 remove the output step yiyixuxu 2025-05-13 22:15:54 +02:00
  • a0deefb606 fix more yiyixuxu 2025-05-13 20:51:21 +02:00
  • e2491af650 fix import yiyixuxu 2025-05-13 20:42:57 +02:00
  • 7e3353196c Fix deprecation warnings in test_ltx_image2video.py (#11538) Anwesha Chowdhury 2025-05-13 23:35:52 +05:30
  • 8c249d1401 Update pipeline_flux_img2img.py to add missing vae_slicing and vae_tiling calls. (#11545) Meatfucker 2025-05-13 12:31:45 -04:00
  • f9662ed8ce resolce conflicts., sayakpaul 2025-05-13 21:03:59 +05:30
  • b555a03723 [tests] Enable testing for HiDream transformer (#11478) Sayak Paul 2025-05-13 21:01:20 +05:30
  • 4e2c346992 update hidream-single-file Dhruv Nair 2025-05-13 16:08:01 +02:00
  • 5de8cb7ddb update Dhruv Nair 2025-05-13 15:51:09 +02:00
  • 406a656d9a update Dhruv Nair 2025-05-13 14:10:56 +02:00
  • 971794454f update DN6 2025-05-13 15:44:14 +05:30
  • 06fee551e9 LTX Video 0.9.7 (#11516) Aryan 2025-05-13 14:57:03 +05:30
  • 4436f54a80 update DN6 2025-05-13 14:56:12 +05:30
  • 2a19302e65 update DN6 2025-05-13 14:50:13 +05:30
  • 721375bb81 update DN6 2025-05-13 14:48:20 +05:30
  • 8b99f7e157 [LoRA] small change to support Hunyuan LoRA Loading for FramePack (#11546) Linoy Tsaban 2025-05-13 10:15:06 +03:00
  • 506a8ea09c fix imports yiyixuxu 2025-05-13 04:36:06 +02:00
  • 07dd6f8c0e [train_dreambooth.py] Fix the LR Schedulers when num_train_epochs is passed in a distributed training env (#11239) Kenneth Gerald Hamilton 2025-05-13 05:04:01 +03:00
  • f8d4a1e283 fix: remove torch_dtype="auto" option from docstrings (#11513) johannaSommer 2025-05-13 03:42:35 +02:00
  • ddd0cfb497 Fix typo in train_diffusion_orpo_sdxl_lora_wds.py (#11541) Abdellah Oumida 2025-05-13 03:28:29 +02:00
  • 58358c2d00 decode block, if skip decoding do not need to update latent yiyixuxu 2025-05-13 01:57:47 +02:00
  • 5cde77f915 make inputs truly immutable, remove the output logic in sequential pipeline, and update so that intermediates_outputs are only new variables yiyixuxu 2025-05-13 01:52:51 +02:00
  • 4f438de35a Add VisualCloze (#11377) Zhong-Yu Li 2025-05-13 05:16:51 +08:00
  • c8a7617536 update modular-refactor-custom-code DN6 2025-05-12 19:37:28 +05:30
  • 98cc6d05e4 [test_models_transformer_ltx.py] help us test torch.compile() for impactful models (#11512) Evan Han 2025-05-12 22:56:15 +09:00
  • c3726153fd enable several pipeline integration tests on XPU (#11526) Yao Matrix 2025-05-12 19:21:37 +08:00
  • e48f6aeeb4 Hunyuan Video Framepack F1 (#11534) Aryan 2025-05-12 16:11:10 +05:30
  • d84dbabba6 Merge branch 'main' into training-group-offloading-tests Sayak Paul 2025-05-12 12:58:40 +05:30
  • 522e827625 move block mappings to its own file yiyixuxu 2025-05-12 01:17:45 +02:00
  • 144eae4e0b add block state will also make sure modifed intermediates_inputs will be updated yiyixuxu 2025-05-12 01:16:42 +02:00
  • 796453cad1 add notes yiyixuxu 2025-05-12 01:14:43 +02:00
  • 01abfc8736 [tests] add tests for framepack transformer model. (#11520) Sayak Paul 2025-05-11 09:50:06 +05:30
  • 153ae34ff6 update __init__ yiyixuxu 2025-05-10 03:50:47 +02:00
  • 0acb5e1460 made a modular_pipelines folder! yiyixuxu 2025-05-10 03:50:31 +02:00
  • 462429b687 remove modular reelated change from pipelines folder yiyixuxu 2025-05-10 03:50:10 +02:00
  • cf01aaeb49 update imports on guiders yiyixuxu 2025-05-10 03:49:30 +02:00
  • 92fe689f06 Change Framepack transformer layer initialization order (#11535) Aryan 2025-05-09 23:09:50 +05:30
  • 0ba1f76d4d enable print_env on xpu (#11507) Yao Matrix 2025-05-09 19:00:51 +08:00
  • d6bf268a4a enable dit integration cases on xpu (#11523) Yao Matrix 2025-05-09 18:36:50 +08:00
  • 3c0a0129fe [LTXPipeline] Update latents dtype to match VAE dtype (#11533) James Xu 2025-05-09 03:35:21 -07:00
  • 2d380895e5 enable 7 cases on XPU (#11503) Yao Matrix 2025-05-09 18:22:08 +08:00
  • 0c47c954f3 [LoRA] support non-diffusers hidream loras (#11532) Sayak Paul 2025-05-09 14:42:39 +05:30
  • cc0335154f Merge branch 'main' into hidream-non-diffusers-lora hidream-non-diffusers-lora Sayak Paul 2025-05-09 13:55:40 +05:30
  • 2017ae5624 fix auto denoise so all tests pass yiyixuxu 2025-05-09 08:19:24 +02:00
  • 2b361a2413 fix get_execusion blocks with loopsequential yiyixuxu 2025-05-09 08:17:10 +02:00
  • c677d528e4 change warning to debug yiyixuxu 2025-05-09 08:16:24 +02:00
  • e1968390c1 make fix-copies sayakpaul 2025-05-09 11:30:08 +05:30
  • 7acf8345f6 [Tests] Enable more general testing for torch.compile() with LoRA hotswapping (#11322) Sayak Paul 2025-05-09 11:29:06 +05:30
  • 3c84a3c108 support non-diffusers hidream loras sayakpaul 2025-05-09 11:05:06 +05:30
  • 599c887164 feat: pipeline-level quantization config (#11130) Sayak Paul 2025-05-09 10:04:44 +05:30
  • 895428be9c Merge branch 'main' into repr-quant-config Sayak Paul 2025-05-09 09:35:22 +05:30
  • 90f934df8e Merge branch 'main' into training-group-offloading-tests Sayak Paul 2025-05-09 09:34:59 +05:30
  • 393aefcdc7 [tests] fix audioldm2 for transformers main. (#11522) Sayak Paul 2025-05-08 21:13:42 +05:30
  • 6674a5157f Conditionally import torchvision in Cosmos transformer (#11524) Aryan 2025-05-08 19:37:47 +05:30
  • 784db0eaab Add cross attention type for Sana-Sprint training in diffusers. (#11514) scxue 2025-05-08 21:25:29 +08:00
  • 50507ea7c4 Merge branch 'main' into training-group-offloading-tests Sayak Paul 2025-05-08 18:42:05 +05:30
  • 0f0618ff2b refactor the denoiseestep using LoopSequential! also add a new file for denoise step yiyixuxu 2025-05-08 11:28:52 +02:00
  • d89631fc50 update input formating, consider kwarggs_type inputs with no name, e/g *_controlnet_kwargs yiyixuxu 2025-05-08 11:27:17 +02:00
  • 16b6583fa8 allow input_fields as input & update message yiyixuxu 2025-05-08 11:25:31 +02:00
  • 66e50d4e24 [LoRA] make lora alpha and dropout configurable (#11467) Linoy Tsaban 2025-05-08 11:54:50 +03:00
  • 3997436a07 fix audioldm2 for transformers main. audioldm2-fix sayakpaul 2025-05-08 11:31:32 +05:30
  • c5c34a4591 Revert "fix audioldm" sayakpaul 2025-05-08 11:30:29 +05:30
  • 87e508f11f fix audioldm sayakpaul 2025-05-08 11:30:11 +05:30
  • fa493e376b Merge branch 'main' into training-group-offloading-tests Sayak Paul 2025-05-08 09:12:29 +05:30
  • 53bd367b03 clean up the __Init__ for stable_diffusion (#11500) YiYi Xu 2025-05-07 07:01:17 -10:00
  • 7b904941bc Cosmos (#10660) Aryan 2025-05-07 20:59:09 +05:30
  • 131ed8ed16 add test for group_offloading with training. sayakpaul 2025-05-07 14:37:35 +05:30
  • 5761c84190 Test hfh v0.31.0.rc0 ci-test-huggingface-hub-v0.31.0.rc0-release Hugging Face Bot (RC Testing) 2025-05-06 20:47:51 +00:00
  • fb29132b98 [docs] minor updates to bitsandbytes docs. (#11509) Sayak Paul 2025-05-06 18:52:18 +05:30
  • 79371661d1 [lora_conversion] Enhance key handling for OneTrainer components in LORA conversion utility (#11441) (#11487) Valeriy Selitskiy 2025-05-06 15:14:58 +02:00
  • 8c661ea586 enable lora cases on XPU (#11506) Yao Matrix 2025-05-06 17:29:50 +08:00
  • d7ffe60166 Hunyuan Video Framepack (#11428) Aryan 2025-05-06 14:59:38 +05:30
  • f552773572 remove controlnet union denoise step, refactor & reuse controlnet denoisee step to accept aditional contrlnet kwargs yiyixuxu 2025-05-06 10:00:14 +02:00
  • dc4dbfe107 reefactor pipeline/block states so that it can dynamically accept kwargs yiyixuxu 2025-05-06 09:58:44 +02:00
  • 10bee525e7 [LoRA] use removeprefix to preserve sanity. (#11493) Sayak Paul 2025-05-06 12:17:57 +05:30
  • d88ae1f52a update dep table. (#11504) Sayak Paul 2025-05-06 11:14:07 +05:30
  • 9862c93d6d fix dep-table-update sayakpaul 2025-05-06 10:59:19 +05:30
  • 71ae84da4a update dep table. sayakpaul 2025-05-06 10:57:38 +05:30
  • 53f1043cbb Update setup.py to pin min version of peft (#11502) Sayak Paul 2025-05-06 10:23:16 +05:30