Commit Graph

  • 93202c7a2b Training working for Qwen Image Jaret Burkett 2025-08-04 21:14:30 +00:00
  • 9da8b5408e Initial but untested support for qwen_image Jaret Burkett 2025-08-04 13:29:37 -06:00
  • 9dfb614755 Initial work for training wan first and last frame Jaret Burkett 2025-08-04 11:37:26 -06:00
  • ef1d60ba34 Update wan 2.2 5b timestep distribution to weighted. Jaret Burkett 2025-07-30 10:13:22 -06:00
  • 75f688766d Version bump Jaret Burkett 2025-07-29 09:30:54 -06:00
  • a558d5b68f Move transformer back to device on aggresive wan 2.2 pipeline after generation. Jaret Burkett 2025-07-29 09:13:47 -06:00
  • 1d1199b15b Fix bug that prevented training wan 2.2 with batch size greater than 1 Jaret Burkett 2025-07-29 09:06:25 -06:00
  • f453e28ea3 Fixed deprecation of lumina pipeline error Jaret Burkett 2025-07-29 08:26:51 -06:00
  • ca7c5c950b Add support for Wan2.2 5B Jaret Burkett 2025-07-29 05:31:54 -06:00
  • e55116d8c9 Added hidream low vram options Jaret Burkett 2025-07-27 18:29:46 -06:00
  • 99705ec8be Add support in UI for Hidream E1 Jaret Burkett 2025-07-27 18:13:36 -06:00
  • ed8d14225f Add ability to set the quantization type for text encoders and transformer in the ui Jaret Burkett 2025-07-27 18:00:53 -06:00
  • b717586ee2 Version bump Jaret Burkett 2025-07-27 15:13:28 -06:00
  • cefa2ca5fe Added initial support for Hidream E1 training Jaret Burkett 2025-07-27 15:12:56 -06:00
  • 3f518d9951 Add sharpening before losses with a split loss on vae training Jaret Burkett 2025-07-27 15:11:56 -06:00
  • 77dc38a574 Some work on caching text embeddings Jaret Burkett 2025-07-26 09:22:04 -06:00
  • 0d89c44624 Bug fixes on vae trainer. Allow to target params for vae training. Jaret Burkett 2025-07-26 09:20:22 -06:00
  • 3e14a674ac Fix upload progress for datasets in the ui Jaret Burkett 2025-07-26 09:07:30 -06:00
  • 523c159579 Add vram flag to some models in the ui Jaret Burkett 2025-07-24 07:02:46 -06:00
  • c5eb763342 Improvements to VAE trainer. Allow CLIP loss. Jaret Burkett 2025-07-24 06:50:56 -06:00
  • ca5cf827a1 Version bump Jaret Burkett 2025-07-20 12:20:46 -06:00
  • b1bff66d52 Merge pull request #343 from davertor/fix_kontext_bs Jaret Burkett 2025-07-20 12:00:55 -06:00
  • a77ba5a089 fix: Guidance incorrect shape Daniel Verdu 2025-07-18 12:49:18 +02:00
  • 8610c6ed7f Made it easy to add control images to the samples in the UI Jaret Burkett 2025-07-17 12:00:48 -06:00
  • e25d2feddf Use scale shift in vae latent space for vae trainer Jaret Burkett 2025-07-17 08:14:07 -06:00
  • f500b9f240 Add ability to do more advanced sample prompt objects to prepart for a UI rework on control images and other things. Jaret Burkett 2025-07-17 07:13:35 -06:00
  • 3916e67455 Scale target vae latent before targeting it Jaret Burkett 2025-07-17 07:12:21 -06:00
  • e5ed450dc7 Allow finetuning tiny autoencoder in vae trainer Jaret Burkett 2025-07-16 07:13:30 -06:00
  • 1930c3edea Fix naming with wan i2v new keys in lora Jaret Burkett 2025-07-14 07:34:01 -06:00
  • ef5149180c Switch i2v ui defaults to weighted Jaret Burkett 2025-07-12 21:30:04 -06:00
  • 998e8b6537 Bump version Jaret Burkett 2025-07-12 16:57:09 -06:00
  • 755f0e207c Fix issue with wan i2v scaling. Adjust aggressive loader to be compatable with updated diffusers. Jaret Burkett 2025-07-12 16:56:27 -06:00
  • 2e84b3d5b1 Update VAE trainer to handle fixed latent target. Also minor bug fixes and improvements Jaret Burkett 2025-07-12 16:55:15 -06:00
  • 7ab44ae0cd Fix issue with getting captions on runpod Jaret Burkett 2025-07-11 19:16:50 +00:00
  • 47002b067f Add path to image for datasets on the image card Jaret Burkett 2025-07-11 11:44:48 -06:00
  • 8537a8557f Add simple ui settings to train Wan i2v models. Jaret Burkett 2025-07-11 11:28:40 -06:00
  • 6e2beef8dd Version Bump Jaret Burkett 2025-07-09 13:55:33 -06:00
  • 611969ec1f Allow control image for omnigen training and sampling Jaret Burkett 2025-07-09 13:54:55 -06:00
  • bbb57de6ec Speed up omnigen TE loading Jaret Burkett 2025-07-05 09:32:00 -06:00
  • 5906a76666 Fixed issue with flux kontext forcing generation image sizes Jaret Burkett 2025-06-29 05:38:20 -06:00
  • 57a81bc0db Update base model version for kontext meta Jaret Burkett 2025-06-28 14:48:36 -06:00
  • 843be31138 Update readme changelog Jaret Burkett 2025-06-28 12:55:23 -06:00
  • 8fb01e96e4 Update sponsors Jaret Burkett 2025-06-28 10:05:17 -06:00
  • 01a3c8a9b1 Fix device issue Jaret Burkett 2025-06-26 19:14:25 -06:00
  • 4f91cb7148 Fix issue with gradient checkpointing and flux kontext Jaret Burkett 2025-06-26 19:03:12 -06:00
  • 446b0b6989 Remove revision for kontext Jaret Burkett 2025-06-26 16:46:58 -06:00
  • 60ef2f1df7 Added support for FLUX.1-Kontext-dev Jaret Burkett 2025-06-26 15:24:37 -06:00
  • 8d9c47316a Work on mean flow. Minor bug fixes. Omnigen improvements Jaret Burkett 2025-06-26 13:46:20 -06:00
  • 84c6edca7e Merge branch 'main' into dev Jaret Burkett 2025-06-25 14:10:25 -06:00
  • 24cd94929e Fix bug that can happen with fast processing dataset Jaret Burkett 2025-06-25 14:01:08 -06:00
  • 19ea8ecc38 Added support for finetuning OmniGen2. Jaret Burkett 2025-06-25 13:58:16 -06:00
  • 18513ec866 Merged in from main Jaret Burkett 2025-06-24 10:56:54 -06:00
  • 5e733764aa Update version Jaret Burkett 2025-06-24 10:37:13 -06:00
  • 03bc431279 Fixed an issue training lumina 2 Jaret Burkett 2025-06-24 10:29:47 -06:00
  • f3eb1dff42 Add a config flag to trigger fast image size db builder. Add config flag to set unconditional prompt for guidance loss Jaret Burkett 2025-06-24 08:51:29 -06:00
  • ba1274d99e Added a guidance burning loss. Modified DFE to work with new model. Bug fixes Jaret Burkett 2025-06-23 08:38:27 -06:00
  • 8602470952 Updated diffusion feature extractor Jaret Burkett 2025-06-19 15:36:10 -06:00
  • 4586eb5392 Added social links to sidebar Jaret Burkett 2025-06-17 13:25:24 -06:00
  • 989ebfaa11 Added a basic torch profiler that can be used in config during development to find some obvious issues. Jaret Burkett 2025-06-17 13:03:39 -06:00
  • ff617fdaea Started doing info bubble docs on the simple ui Jaret Burkett 2025-06-17 11:00:24 -06:00
  • 595a6f1735 Initial setup for a cron working on the ui for various tasks Jaret Burkett 2025-06-17 07:43:34 -06:00
  • 1cc663a664 Performance optimizations for pre processing the batch Jaret Burkett 2025-06-17 07:37:41 -06:00
  • 11f2eee53a Hide control images from ui image viewer Jaret Burkett 2025-06-16 07:18:43 -06:00
  • 1c2b7298dd More work on mean flow loss. Moved it to an adapter. Still not functioning properly though. Jaret Burkett 2025-06-16 07:17:35 -06:00
  • c0314ba325 Fixed some issues with training mean flow algo. Still testing WIP Jaret Burkett 2025-06-14 12:24:00 -06:00
  • cbf04b8d53 Fixed some issues with training mean flow algo. Still testing WIP Jaret Burkett 2025-06-14 12:24:00 -06:00
  • 0946a66576 Merge branch 'main' into dev Jaret Burkett 2025-06-12 08:11:19 -06:00
  • 3f0ae99d48 Version bump Jaret Burkett 2025-06-12 08:01:26 -06:00
  • fc83eb7691 WIP on mean flow loss. Still a WIP. Jaret Burkett 2025-06-12 08:00:51 -06:00
  • cf11f128b9 Merge pull request #304 from hameerabbasi/fix-caption-loads Jaret Burkett 2025-06-12 07:44:12 -06:00
  • 5e86139e0a Fix NameError. Hameer Abbasi 2025-06-11 15:07:20 +02:00
  • c5d6b74fea Fix caption loading. Hameer Abbasi 2025-06-11 14:47:26 +02:00
  • ba5196dd4a Merge branch 'main' into dev Jaret Burkett 2025-06-10 10:26:11 -06:00
  • ffb5fe0667 Version bump Jaret Burkett 2025-06-10 10:04:32 -06:00
  • f8fb3b9c45 Added support for sdxl and sd1.5 to the ui. Jaret Burkett 2025-06-10 10:03:54 -06:00
  • d5c547da43 Fixed DOP typo Jaret Burkett 2025-06-10 08:44:47 -06:00
  • f19f7f9486 Fixed issue with wan2.1 training in ui. Name had a typo Jaret Burkett 2025-06-10 08:42:18 -06:00
  • 7317ed58af Adjust the ui of the sidebar Jaret Burkett 2025-06-10 08:40:51 -06:00
  • 517bc294fa Update support link Jaret Burkett 2025-06-10 08:28:30 -06:00
  • 97e101522c Increase ema feedback amount. Normalize the dfe 4 image embeds Jaret Burkett 2025-06-10 08:01:13 -06:00
  • eefa93f16e Various code to support experiments. Jaret Burkett 2025-06-09 11:19:21 -06:00
  • 22cdfadab6 Added new timestep weighing strategy Jaret Burkett 2025-06-04 01:16:02 -06:00
  • adc31ec77d Small updates and bug fixes for various things Jaret Burkett 2025-06-03 20:08:35 -06:00
  • 82b90b902e Double tap torch install to force blackwell compatability Jaret Burkett 2025-06-02 20:22:22 -06:00
  • 85f4b47e79 Fix issue with setup tools requirements Jaret Burkett 2025-06-02 06:54:35 -06:00
  • e20a869dc1 Update docker install for blacwell Jaret Burkett 2025-06-01 19:06:53 -06:00
  • 12fa109910 Updated cuda arch list on docker build Jaret Burkett 2025-06-01 13:37:11 -06:00
  • 7d76165dcf Merge branch 'main' into dev Jaret Burkett 2025-06-01 13:33:40 -06:00
  • b6d25fcd10 Improvements to vae trainer. Adjust denoise prediction of DFE v3 Jaret Burkett 2025-05-30 12:06:47 -06:00
  • ffaf2f154a Fix issue with the way chroma handled gradient checkpointing. Jaret Burkett 2025-05-28 08:41:47 -06:00
  • 34f4c14cd6 Work on vae trainer Jaret Burkett 2025-05-28 07:42:48 -06:00
  • 79bb9be92b Fix issue with saving chroma full finetune. Jaret Burkett 2025-05-28 07:42:30 -06:00
  • 79499fa795 Allow fine tuning pruned versions of chroma. Allow flash attention 2 for chroma if it is installed. Jaret Burkett 2025-05-21 07:02:50 -06:00
  • 48e11cf843 Fallback unwrapping logic if fails Jaret Burkett 2025-05-21 03:10:33 -06:00
  • 7045a01375 Fixed issue saving optimizer in some instances. Jaret Burkett 2025-05-21 02:27:55 -06:00
  • fca7fd6c38 Merge branch 'main' of github.com:ostris/ai-toolkit Jaret Burkett 2025-05-21 02:20:06 -06:00
  • e5181d23cd Added some experimental training techniques. Ignore for now. Still in testing. Jaret Burkett 2025-05-21 02:19:54 -06:00
  • 4f896c0d8a Fixed issue where sampling fails if doing a full finetune for some models Jaret Burkett 2025-05-17 19:37:55 +00:00
  • 01101be196 version bump Jaret Burkett 2025-05-17 05:50:12 -06:00
  • 6174ba474e Fixed issue with chroma sampling Jaret Burkett 2025-05-10 18:30:23 +00:00