github/ai-toolkit
1
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity
1,041 Commits 9 Branches 0 Tags
9ca58e9aa2accf8fa3cd7209fd5266d6775bae5b
Commit Graph

1041 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Jaret Burkett
9ca58e9aa2 Fixed offload and quantize order of ltx 2.3 text encoder. 2026-04-07 15:11:50 -06:00
Jaret Burkett
0dcbabf6af Fix merge nertwork ref 2026-04-01 10:38:31 -06:00
M. Hofer
f213e3b1e5 Fix FLUX2 Klein load-time VRAM spikes on low-memory GPUs. (#726) 
Keep the transformer and Qwen text encoder off CUDA during initial load/quantization in low-VRAM mode so model startup avoids full-model OOM before offloading and quantization can take effect.

Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Jaret Burkett <jaretburkett@gmail.com>
 2026-04-01 09:36:55 -06:00
Jaret Burkett
da2a79590f Add a merge network on save strength 2026-04-01 09:21:08 -06:00
Jaret Burkett
853ffaf207 Add light mode support. 2026-03-31 16:54:55 -06:00
Jaret Burkett
ad474e3d06 Update and reformat the readme 2026-03-31 12:31:05 -06:00
Jaret Burkett
4a3251640a More work on compiling models 2026-03-31 12:11:56 -06:00
Jaret Burkett
358d684f6f Move compiiling the model after accelerate manipulation 2026-03-31 09:52:27 -06:00
Jaret Burkett
0045260af7 Fix issue where compile true did not actually compile the model 2026-03-31 09:27:54 -06:00
Jaret Burkett
e22039e4aa Add more optimizers to the ui 2026-03-31 09:20:20 -06:00
Jaret Burkett
bf56217c37 Fixed issue where job would fail if DB is locked. 2026-03-31 09:10:33 -06:00
Jaret Burkett
dcb7f465ec Version Bump 2026-03-30 15:57:40 -06:00
Jaret Burkett
626d9674ea Add info about automated coding agent pull requests. So sick of them. 2026-03-30 11:02:12 -06:00
Jaret Burkett
b43ea6c2d3 Abort caption requests when they are not in view to tax the server less. 2026-03-30 10:46:25 -06:00
Jaret Burkett
a484e55d66 Rework dataset model and file dragging. Use single model for dragging, uploading, and selecting images. 2026-03-30 10:33:10 -06:00
Jaret Burkett
ac82ebd852 Dont try to list hidden files in datasets 2026-03-30 10:03:10 -06:00
Jaret Burkett
171535833a Add Mac OS support for Apple Silicon (#770) 
* Made an install script and auto updates env for mac

* GPU sensors and initial training working for MAC. Still WIP.

* Switch dataloader to single threaded until I can work around some mac pickeling issues.

* Get quantization working on mac

* Fix mac exclusive imports so they don't break other builds.

* Add mac instructions to the UI
 2026-03-30 09:37:47 -06:00
Jaret Burkett
bc47fd6755 Make a requirements base file to make it easier to maintain requirements across platforms. 2026-03-29 14:04:46 -06:00
Jaret Burkett
fbda10d088 Add a duplicate dataset function to the ui 2026-03-29 13:51:18 -06:00
Jaret Burkett
86dcf39eee Allow user to set a training seed via env vars for repeat result testing 2026-03-29 13:34:46 -06:00
Jaret Burkett
45e99664b9 Add icons to the top bar on the job page 2026-03-29 12:38:47 -06:00
Jaret Burkett
540659709d Improved the load time of dataset and sample images and videos by switching to streaming 2026-03-29 10:38:34 -06:00
Jaret Burkett
e030f4f2e0 Show the control images in the image viewer when clicked so they can be easily previewed for reference. 2026-03-29 10:00:54 -06:00
Jaret Burkett
affa411edc Fixed an issue where Flux.2 model VAE can be left offloaded to CPU when encoding control images while caching latents 2026-03-29 09:49:10 -06:00
Jaret Burkett
6a1fc54779 Add t0 loss target 2026-03-28 13:35:21 -06:00
Jaret Burkett
8302b21f8f Version Bump 2026-03-28 13:23:52 -06:00
willhsmit
20929b93df Fix onChange path for EMA Decay input (#695) 
Changes to the EMA Decay input don't get preserved when switching back and forth between Advanced and Simple view. I believe the onChange is not writing it correctly here.
 2026-03-28 13:02:32 -06:00
abionda-sc
4ef5cbe5bc Fixing bug where width and height are inverted for control image resizing (#707) 2026-03-28 13:00:32 -06:00
Rob Ballantyne
700c4b53d0 Pin timm==1.0.22 (#633) 
* Pin timm==1.0.22

* Added timm version pinn to dgx

---------

Co-authored-by: Jaret Burkett <jaretburkett@gmail.com>
 2026-03-28 12:52:41 -06:00
Rayane
ca72eb1515 Add 1328 native resolution for Qwen Image training (#749) 
* Add 1328 native resolution for Qwen Image training

Qwen-Image and Qwen-Image-2512 have a native 1:1 resolution of 1328x1328
as documented in the official model card's aspect ratio table. Adding it
to the resolution buckets and UI allows training at the model's native
resolution for improved quality.

* Revert example config change (24GB OOM at 1328)
 2026-03-28 12:09:15 -06:00
Jaret Burkett
5ce87fa48b Version bump 2026-03-27 20:26:31 -06:00
Jaret Burkett
740657e25e Improve dataset uploader. Upload the files one at a time instead of one huge chunk. Show progress for each file. 2026-03-27 09:26:22 -06:00
Jaret Burkett
f85bf065bf Use pooler embeddings for DFE v6 with dino v3 2026-03-27 07:02:07 -06:00
Jaret Burkett
a802014ec5 Update the torch versions in the README 2026-03-26 12:15:32 -06:00
Jaret Burkett
2782df02c3 Allow HF_HUB_ENABLE_HF_TRANSFER to be set via env variable 2026-03-26 10:45:49 -06:00
Jaret Burkett
2c8d2acdcb On jobs table, sort idle jobs by last updated so recent active ones are at the top 2026-03-26 10:33:17 -06:00
Jaret Burkett
9a77389653 On a new training job, or when editing one, load everything before allowing editing 2026-03-26 10:23:42 -06:00
Jaret Burkett
a7bb4ddb2c Work on loss graph. Add smoothed overlay. Allow user to hilite a secton of the graph to zoom into. 2026-03-26 10:05:09 -06:00
Jaret Burkett
401f7df425 Merge branch 'main' of github.com:ostris/ai-toolkit 2026-03-26 09:11:50 -06:00
Jaret Burkett
4df3b0463f Save job pid to the database and sing sigint to kill it when stopping so it stops immediatly. 2026-03-26 09:10:37 -06:00
科林 KELIN
489b194231 Fix CPU/CUDA device mismatch in Klein edit control image encoding (#742) 
When training Klein models with a `control_path` (edit/kontext-style
paired datasets), `encode_image_refs()` returns tensors that reside on
the VAE's device (CPU, since the VAE weights are loaded via
`load_file(..., device="cpu")` and are never explicitly moved to the
training device).  Concatenating those CPU tensors with the training
latents (`packed_latents`) that live on CUDA raises:

    RuntimeError: Expected all tensors to be on the same device

Fix: move `img_cond_seq` and `img_cond_seq_ids` to the same device
(and dtype) as `img_input` / `img_input_ids` before concatenation.

Co-authored-by: HuangYuChuh <HuangYuChuh@users.noreply.github.com>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
 2026-03-25 11:45:38 -06:00
Jaret Burkett
89d2090962 Fixed race condition that would occasionally set the dataset path to the first one when editing a job 2026-03-25 11:22:42 -06:00
Jaret Burkett
3f7a3d8d87 Shorten stal action to 3 months 2026-03-25 10:42:18 -06:00
Jaret Burkett
45647c15d3 Added github actions to close stale issues automatically. Hopefully it doesnt break things 2026-03-25 10:25:56 -06:00
Jaret Burkett
899ee528f9 Update git ignore 2026-03-25 10:16:36 -06:00
Jaret Burkett
5d5a8ef9da Fixed issue with deleting datasets and jobs with newer version of node.js. Bumped minimum version of node js to 20 2026-03-25 10:04:36 -06:00
Jaret Burkett
dfde30f231 Fix issue with ltx2 custom te repo path 2026-03-25 09:50:18 -06:00
Jaret Burkett
b8000dbcbc Bump version 2026-03-25 08:18:42 -06:00
Rodrigo Reis
54f4732c9b Fix the bug in temporal_compression data loader (#754) 2026-03-25 08:16:44 -06:00
Jaret Burkett
7f3309b291 Add support for audo frame count so datasets can have varrying length videos. Varous ltx 2.3 VAE optimizations such as removing tiling articacts, and doing frame split encoding to reduce vram on encoding/decoding. 2026-03-24 12:20:09 -06:00