Commit Graph

  • f6b4bdc83f feat: block-wise fp8 quantization Kohya S 2025-09-18 21:20:54 +09:00
  • 2ce506e187 fix: fp8 casting not working Kohya S 2025-09-18 21:20:08 +09:00
  • f5b004009e fix: correct tensor indexing in HunyuanVAE2D class for blending and encoding functions Kohya S 2025-09-17 21:54:25 +09:00
  • cbe2a9da45 feat: add conversion script for LoRA models to ComfyUI format with reverse option Kohya S 2025-09-16 21:48:47 +09:00
  • f318ddaeea docs: update HunyuanImage-2.1 training guide with model download instructions and VRAM optimization settings (by Claude) kohya-ss 2025-09-16 21:18:01 +09:00
  • 39458ec0e3 fix: update default values for guidance_scale, image_size, infer_steps, and flow_shift in argument parser kohya-ss 2025-09-16 21:17:21 +09:00
  • 2732be0b29 Merge branch 'feat-hunyuan-image-2.1-inference' of https://github.com/kohya-ss/sd-scripts into feat-hunyuan-image-2.1-inference Kohya S 2025-09-14 20:49:24 +09:00
  • 1a73b5e8a5 feat: add script to convert LoRA format to ComfyUI format Kohya S 2025-09-14 20:49:20 +09:00
  • e04b9f0497 docs: add LoRA training guide for HunyuanImage-2.1 model (by Gemini CLI) kohya-ss 2025-09-13 22:06:10 +09:00
  • 29b0500e70 fix: restore files section in _typos.toml for exclusion configuration Kohya S 2025-09-13 21:18:50 +09:00
  • 4e2a80a6ca refactor: update imports to use safetensors_utils for memory-efficient operations Kohya S 2025-09-13 21:07:11 +09:00
  • d831c88832 fix: sample generation doesn't work with block swap Kohya S 2025-09-13 21:06:04 +09:00
  • bae7fa74eb Merge branch 'sd3' into feat-hunyuan-image-2.1-inference Kohya S 2025-09-13 20:13:58 +09:00
  • f5d44fd487 Merge pull request #2200 from kohya-ss/feat-faster-safetensors-load Kohya S. 2025-09-13 20:09:03 +09:00
  • 4568631b43 docs: update README to reflect improved loading speed of .safetensors files Kohya S 2025-09-13 20:05:39 +09:00
  • e1c666e97f Update library/safetensors_utils.py Kohya S. 2025-09-13 20:03:55 +09:00
  • 8783f8aed3 feat: faster safetensors load and split safetensor utils Kohya S 2025-09-13 19:51:38 +09:00
  • 9a61d61b22 feat: avoid unet type casting when fp8_scaled Kohya S 2025-09-12 22:18:29 +09:00
  • 7a651efd4d feat: add 'tak' to recognized words and update block swap method to support backward pass Kohya S 2025-09-12 22:00:41 +09:00
  • aa0af24d01 Merge branch 'sd3' into feat-hunyuan-image-2.1-inference Kohya S 2025-09-12 21:41:12 +09:00
  • 209c02dbb6 feat: HunyuanImage LoRA training Kohya S 2025-09-12 21:40:42 +09:00
  • 419a9c4af4 Merge pull request #2192 from kohya-ss/doc-update-for-latest-features Kohya S. 2025-09-12 20:28:42 +09:00
  • cbc9e1a3b1 feat: add byt5 to the list of recognized words in typos configuration Kohya S 2025-09-11 22:27:08 +09:00
  • a0f0afbb46 fix: revert constructor signature update Kohya S 2025-09-11 22:27:00 +09:00
  • 7f983c558d feat: block swap for inference and initial impl for HunyuanImage LoRA (not working) Kohya S 2025-09-11 22:15:22 +09:00
  • 5149be5a87 feat: initial commit for HunyuanImage-2.1 inference Kohya S 2025-09-11 12:54:12 +09:00
  • ee8e670765 Merge branch 'sd3' into doc-update-for-latest-features kohya-ss 2025-09-09 12:42:09 +09:00
  • f8337726cf Merge pull request #2196 from rockerBOO/validation-dataset-subset Kohya S. 2025-09-09 12:38:23 +09:00
  • fe4c18934c blocks_to_swap is supported for validation loss now rockerBOO 2025-09-08 14:28:55 -04:00
  • 78685b9c5f Move general settings to top to make more clear the validation bits rockerBOO 2025-09-08 14:18:50 -04:00
  • ef4397963b Fix validation dataset documentation to not use subsets rockerBOO 2025-09-08 14:16:33 -04:00
  • 0bb0d91615 doc: update introduction and clarify command line option priorities in config README kohya-ss 2025-09-06 19:52:54 +09:00
  • 952f9ce7be Update docs/train_textual_inversion.md Kohya S. 2025-09-04 19:46:04 +09:00
  • 884fc8c7f5 doc: remove SD3/FLUX.1 training guide kohya-ss 2025-09-04 18:40:21 +09:00
  • ddfb38e501 doc: add documentation for Textual Inversion training scripts kohya-ss 2025-09-04 18:39:52 +09:00
  • 6c82327dc8 doc: remove Japanese section on Gradual Latent options from gen_img README kohya-ss 2025-09-01 21:32:50 +09:00
  • 9984868154 doc: update README to include support for SDXL models and additional command-line options for gen_img.py kohya-ss 2025-09-01 21:32:24 +09:00
  • 142d0be180 doc: add comprehensive fine-tuning guide for various model architectures kohya-ss 2025-09-01 12:36:51 +09:00
  • c38b07d0da doc: add validation loss documentation for model training kohya-ss 2025-08-31 21:39:47 +09:00
  • 80710134d5 doc: add Sage Attention and sample batch size options to Lumina training guide kohya-ss 2025-08-31 21:19:28 +09:00
  • fe81d40202 doc: refactor structure for improved readability and maintainability kohya-ss 2025-08-31 21:14:45 +09:00
  • 989448afdd doc: enhance SD3/SDXL LoRA training guide kohya-ss 2025-08-31 19:19:10 +09:00
  • 884e07d73e Merge pull request #2191 from kohya-ss/fix-chroma-training-withtout-te-cache Kohya S. 2025-08-30 09:31:11 +09:00
  • e836b7f66d fix: chroma LoRA training without Text Encode caching fix-chroma-training-withtout-te-cache Kohya S 2025-08-30 09:30:24 +09:00
  • 8cadec68bb Merge pull request #2178 from kohya-ss/update-libraries Kohya S. 2025-08-28 08:42:53 +09:00
  • 5a5138d0ab doc: add PR reference for PyTorch and library versions update Kohya S 2025-08-28 08:38:14 +09:00
  • c52c45cd7a doc: update for PyTorch and libraries versions Kohya S 2025-08-28 08:36:09 +09:00
  • ac72cf88a7 feat: remove bitsandbytes version specification in requirements.txt Kohya S 2025-08-28 08:35:40 +09:00
  • f7acd2f7a3 fix: consolidate PyTorch versions in workflow matrix Kohya S 2025-08-24 17:16:17 +09:00
  • 4b12746d39 Merge branch 'dev' into sd3 Kohya S 2025-08-24 17:10:51 +09:00
  • 28cce2271b Merge branch 'main' into dev Kohya S 2025-08-24 17:09:32 +09:00
  • 1470cb8508 Merge pull request #2185 from kohya-ss/update-gitignore-for-ai-agent-and-testing Kohya S. 2025-08-24 17:09:05 +09:00
  • 69a85a0a11 gitignore: add CLAUDE.md, GEMINI.md, MagicMock and related files to ignore list Kohya S 2025-08-24 17:08:08 +09:00
  • 14ee64823f Merge pull request #2184 from kohya-ss/add-sponsor-logo Kohya S. 2025-08-24 17:05:51 +09:00
  • acba279b0b fix: update PyTorch version in workflow matrix Kohya S 2025-08-24 17:03:18 +09:00
  • 0e7f7808b0 doc: add sponsor logo and annoucements kohya-ss 2025-08-23 18:44:23 +09:00
  • f61c442f0b fix: use strategy for tokenizer and latent caching Kohya S 2025-08-16 22:03:52 +09:00
  • 6f24bce7cc fix: remove unnecessary super call in assert_extra_args method Kohya S 2025-08-16 22:03:31 +09:00
  • 6edbe00547 feat: update libraries, remove warnings Kohya S 2025-08-16 20:07:03 +09:00
  • 18e62515c4 Merge branch 'dev' into sd3 Kohya S 2025-08-15 19:05:07 +09:00
  • 1273af0cdc Merge branch 'main' into dev Kohya S 2025-08-15 19:04:23 +09:00
  • 2857f21abf Merge pull request #2175 from woct0rdho/resize-control-lora Kohya S. 2025-08-15 19:03:08 +09:00
  • 3ad71e1acf Refactor to avoid mutable global variable woctordho 2025-08-15 11:14:43 +08:00
  • c6fab554f4 Support resizing ControlLoRA woctordho 2025-08-15 03:00:40 +08:00
  • 3ce0c6e71f Fix. Symbiomatrix 2025-06-22 01:00:59 +03:00
  • 63ec59fc0b Support for multiple format loras. Symbiomatrix 2025-04-20 15:08:21 +03:00
  • f25c265836 Merge pull request #2174 from kohya-ss/fix-modelspec-sd3-finetune Kohya S. 2025-08-13 23:07:45 +09:00
  • 351bed965c fix model type handling in analyze_state_dict_state function for SD3 Kohya S 2025-08-13 21:38:51 +09:00
  • dcce057609 Merge pull request #2168 from rockerBOO/model-spec-1.0.1 Kohya S. 2025-08-13 21:31:21 +09:00
  • a125c10852 Merge pull request #2167 from rockerBOO/patch-6 Kohya S. 2025-08-12 19:30:15 +09:00
  • c149cf283b Add parser args for other trainers. rockerBOO 2025-08-03 00:58:25 -04:00
  • 9bb50c26c4 Set sai_model_spec to must rockerBOO 2025-08-03 00:43:09 -04:00
  • 10bfcb9ac5 Remove text model spec rockerBOO 2025-08-03 00:40:10 -04:00
  • bf0f86e797 Add sai_model_spec to train_network.py rockerBOO 2025-08-02 21:35:45 -04:00
  • 056472c2fc Add tests rockerBOO 2025-08-02 21:16:56 -04:00
  • d24d733892 Update model spec to 1.0.1. Refactor model spec rockerBOO 2025-08-02 21:14:27 -04:00
  • 0ad2cb854d Update flux_train_network.md Dave Lage 2025-08-02 17:27:55 -04:00
  • 24c605ee3b Update flux_train_network.md Dave Lage 2025-08-02 17:21:25 -04:00
  • b9c091eafc Fix validation documentation Dave Lage 2025-08-02 17:19:26 -04:00
  • 5249732a0f chore: update README to include --apply_t5_attn_mask requirement for Chroma training #2163 Kohya S 2025-08-01 23:38:02 +09:00
  • 75dd8c872b Merge pull request #2160 from kozistr/update/pytorch-optimizer Kohya S. 2025-08-01 21:55:41 +09:00
  • aebfea255f Merge pull request #2165 from kohya-ss/force_t5_attn_mask_for_chroma_training Kohya S. 2025-08-01 21:55:23 +09:00
  • bd6418a940 fix: add assertion for apply_t5_attn_mask requirement in Chroma Kohya S 2025-08-01 21:47:38 +09:00
  • 5dff02a65d Merge pull request #2157 from kohya-ss/feature-chroma-support Kohya S. 2025-07-30 22:13:37 +09:00
  • 250f0eb9b0 doc: update README and training guide with breaking changes for CFG scale and model download instructions Kohya S 2025-07-30 22:08:51 +09:00
  • 96feb61c0a feat: implement modulation vector extraction for Chroma and update related methods Kohya S 2025-07-30 21:34:49 +09:00
  • 450630c6bd fix: create network from weights not working Kohya S 2025-07-29 20:32:24 +09:00
  • 10de781806 build(deps): pytorch-optimizer to 3.7.0 kozistr 2025-07-28 23:40:38 +09:00
  • 6c8973c2da doc: add reference link for input vector gradient requirement in Chroma class Kohya S 2025-07-28 22:08:02 +09:00
  • af14eab6d7 doc: update section number for regex-based rank and learning rate configuration in FLUX.1 LoRA guide Kohya S 2025-07-26 19:37:15 +09:00
  • c28e7a47c3 feat: add regex-based rank and learning rate configuration for FLUX.1 LoRA Kohya S 2025-07-26 19:35:42 +09:00
  • 32f06012a7 doc: update flux train document and add about breaking changes in sample generation prompts kohya-ss 2025-07-21 21:48:06 +09:00
  • eef05504de Merge branch 'sd3' into feature-chroma-support Kohya S 2025-07-21 13:49:08 +09:00
  • 4987057701 Merge pull request #2042 from kohya-ss/update-docs Kohya S. 2025-07-21 13:48:34 +09:00
  • c84a163b32 docs: update README for documentation Kohya S 2025-07-21 13:40:03 +09:00
  • 7de68c1eb1 Merge branch 'sd3' into update-docs Kohya S 2025-07-21 13:32:43 +09:00
  • 9eda938876 Merge branch 'sd3' into feature-chroma-support Kohya S 2025-07-21 13:32:22 +09:00
  • d98400b06e Merge pull request #2138 from kohya-ss/feature-lumina-image Kohya S. 2025-07-21 13:21:26 +09:00
  • 518545bffb docs: add support information for Lumina-Image 2.0 in recent updates Kohya S 2025-07-21 13:16:42 +09:00
  • d300f19045 docs: update Lumina training guide to include inference script and options Kohya S 2025-07-21 13:15:09 +09:00