ostris/ai-toolkit
1
0
Fork 0
mirror of https://github.com/ostris/ai-toolkit.git synced 2026-01-26 16:39:47 +00:00
Code Issues Packages Projects Releases Wiki Activity
698 Commits 10 Branches 0 Tags
22cdfadab6e59bf27bd22bf28a650a7da23ca966
Commit Graph

159 Commits

Author SHA1 Message Date
Jaret Burkett
22cdfadab6 Added new timestep weighing strategy 2025-06-04 01:16:02 -06:00
Jaret Burkett
adc31ec77d Small updates and bug fixes for various things 2025-06-03 20:08:35 -06:00
Jaret Burkett
e5181d23cd Added some experimental training techniques. Ignore for now. Still in testing. 2025-05-21 02:19:54 -06:00
Jaret Burkett
5890e67a46 Various bug fixes 2025-04-29 09:30:33 -06:00
Jaret Burkett
6fb44db6a0 Finished up first frame for i2v adapter 2025-04-12 17:13:04 -06:00
Jaret Burkett
cd37ccfc2e Use gradient checkpointing on DFE models if set 2025-04-11 10:45:39 -06:00
Jaret Burkett
059155174a Added mask diffirential mask dialation for flex2. Handle video for the i2v adapter 2025-04-10 11:50:01 -06:00
Jaret Burkett
a8680c75eb Added initial support for finetuning wan i2v WIP 2025-04-07 20:34:38 -06:00
Jaret Burkett
a42c5a1de5 Adjust buckets for flex2 2025-04-02 06:47:41 -06:00
Jaret Burkett
c083a0e5ea Allow DFE to not have a VAE 2025-03-30 09:23:01 -06:00
Jaret Burkett
5365200da1 Added ability to add models to finetune as plugins. Also added flux2 new arch via that method. 2025-03-27 16:07:00 -06:00
Jaret Burkett
ce4c5291a0 Added experimental wavelet loss 2025-03-26 18:11:23 -06:00
Jaret Burkett
45be82d5d6 Handle inpainting training for control_lora adapter 2025-03-24 13:17:47 -06:00
Jaret Burkett
f10937e6da Handle multi control inputs for control lora training 2025-03-23 07:37:08 -06:00
Jaret Burkett
3812957bc9 Added ability to train control loras. Other important bug fixes thrown in 2025-03-14 18:03:00 -06:00
Jaret Burkett
242c04a0b8 Fix error with training video models with batch greater than 1 2025-03-08 18:47:27 -07:00
Jaret Burkett
6f6fb90812 Added cogview4. Loss still needs work. 2025-03-04 18:43:52 -07:00
Jaret Burkett
60539c0b0f Allow using prior loss with a custom adapter 2025-03-01 08:01:14 -07:00
Jaret Burkett
d360e76661 fixed issue with dop prompt replacement 2025-02-26 13:35:18 -07:00
Jaret Burkett
6ec23ed226 Fixed issue when doing inverted masked prior with flowmatching algos 2025-02-26 12:12:32 -07:00
Jaret Burkett
f6e16e582a Added Differential Output Preservation Loss to trainer and ui 2025-02-25 20:12:36 -07:00
Jaret Burkett
7679105d52 Added llm text encoder adapter 2025-02-13 08:28:32 -07:00
Jaret Burkett
0b8a32def7 merged in lumina2 branch 2025-02-12 09:33:03 -07:00
Jaret Burkett
787bb37e76 Small fixed for DFE, polar guidance, and other things 2025-02-12 09:27:44 -07:00
Jaret Burkett
d138f07365 Imitial lumina3 support 2025-02-08 10:59:53 -07:00
Jaret Burkett
ff3d54bb5b Make the mean of the mask multiplier be 1.0 for a more balanced loss. 2025-02-06 10:57:06 +00:00
Jaret Burkett
15a57bc89f Add new version of DFE. Kitchen sink 2025-01-31 11:42:27 -07:00
Jaret Burkett
5e663746b8 Working multi gpu training. Still need a lot of tweaks and testing. 2025-01-25 16:46:20 -07:00
Jaret Burkett
bbfba0c188 Added v2 of dfp 2025-01-22 16:32:13 -07:00
Jaret Burkett
04abe57c76 Added weighing to DFE 2025-01-22 08:50:57 -07:00
Jaret Burkett
29122b1a54 Added code to handle diffusion feature extraction loss 2025-01-21 14:21:34 -07:00
Jaret Burkett
8ef07a9c36 Added training for an experimental decoratgor embedding. Allow for turning off guidance embedding on flux (for unreleased model). Various bug fixes and modifications 2024-12-15 08:59:27 -07:00
Jaret Burkett
473d455f44 Process empty clip image if there is not one for reg images when training a custom adapter 2024-10-15 08:28:04 -06:00
Jaret Burkett
ab22674980 Allow for a default caption file in the folder. Minor bug fixes. 2024-10-10 07:31:33 -06:00
Jaret Burkett
2776221497 Added option to cache empty prompt or trigger and unload text encoders while training 2024-09-21 20:54:09 -06:00
Plat
79b4e04b80 Feat: Wandb logging (#95) 
* wandb logging

* fix: start logging before train loop

* chore: add wandb dir to gitignore

* fix: wrap wandb functions

* fix: forget to send last samples

* chore: use valid type

* chore: use None when not type-checking

* chore: resolved complicated logic

* fix: follow log_every

---------

Co-authored-by: Plat <github@p1at.dev>
Co-authored-by: Jaret Burkett <jaretburkett@gmail.com>
 2024-09-19 20:01:01 -06:00
Jaret Burkett
3a1f464132 Added support for training vision direct weight adapters 2024-09-05 10:11:44 -06:00
Jaret Burkett
5c8fcc8a4e Fix bug with zeroing out gradients when accumulating 2024-09-03 08:29:15 -06:00
Jaret Burkett
121a760c19 Added proper grad accumulation 2024-09-03 07:24:18 -06:00
Jaret Burkett
d44d4eb61a Added a new experimental linear weighing technique 2024-09-02 09:22:13 -06:00
Jaret Burkett
60232def91 Made peleminary arch for flux ip adapter training 2024-08-28 08:55:39 -06:00
Jaret Burkett
3843e0d148 Added support for vision direct adapter for flux 2024-08-26 16:27:28 -06:00
Jaret Burkett
e07a98a50c Bugfixes for full finetuning at bf16 2024-08-22 05:15:33 -06:00
Jaret Burkett
165510ace2 Dumb typo 2024-08-15 12:59:32 -06:00
Jaret Burkett
b99d36dfdb fixed issue with batch sizes larget than 1 2024-08-15 12:21:38 -06:00
Jaret Burkett
9ee1ef2a0a Added experimental modified sigma sqrt weight mapping for linear timestep scheduling for flowmatching 2024-08-12 17:03:09 -06:00
Jaret Burkett
653fe60f16 Updates to flow matching algo 2024-08-07 15:04:17 +00:00
Jaret Burkett
c2424087d6 8 bit training working on flux 2024-08-06 11:53:27 -06:00
Jaret Burkett
edb7e827ee Adjusted flow matching so target noise multiplier works properly with it. 2024-08-05 11:40:05 -06:00
Jaret Burkett
f321de7bdb Setup to retrain guidance embedding for flux. Use defualt timestep distribution for flux 2024-08-04 10:37:23 -06:00