a798e06dd2
Updated the support button to look more like a button
2026-05-25 07:48:04 -06:00
8042cbe9d2
Added virtulization for sample images to handle huge number of samples more efficientyly.
2026-05-25 07:22:24 -06:00
fbac1cb7f5
Dont force flash attention on hidream 01. Causes random issues and is slower.
2026-05-24 16:05:22 -06:00
307ff11bc5
Drastically improved the performance of the dataset viewer on large datasets by switching to virtualization. Fixed issue with images loading when they have two periods in a row ..
2026-05-24 15:13:55 -06:00
c6a7e81a70
Added a dataset image viewer. Fixed an issue where captions would show as not saved when they were saved.
2026-05-24 14:31:06 -06:00
12304e170f
Added some experimental loss targets
2026-05-24 14:13:23 -06:00
644a6f9246
Fix device casting for zimage in some instances
2026-05-23 10:37:33 -06:00
c6ecc03ccd
Fixed saving full model of z_image l2p for finetuning
2026-05-23 07:36:24 -06:00
6102370df9
Add support for ZImage L2P
2026-05-22 14:58:28 -06:00
6fc08a8928
Fix issue with saving images from new sample viewer
2026-05-22 11:58:32 -06:00
5579837c3f
Add hidream o1 to the readme
2026-05-21 07:56:54 -06:00
aecd554128
Add sapiens2 matting as a mask generator. Begin transition to model paths and model folders.
2026-05-20 08:56:16 -06:00
15d4fb89ff
Fixed overflow issue on modal
2026-05-19 10:12:43 -06:00
df851b3497
Made the UI mobile friendly, finally...
2026-05-19 09:27:36 -06:00
6ecaf679dc
Add ability to run small scripts from the ui and added a merge lora script
2026-05-18 14:33:47 -06:00
ec58dcde92
Add better gradient checkpointing to dfes
2026-05-18 09:25:23 -06:00
b42acb988f
Remove future steps from loss log if resuming from an earlier step
2026-05-18 09:23:55 -06:00
e03c6e4dc9
Fix potential inconsistency with different attention mentods in hidream01
2026-05-13 09:08:11 -06:00
4bfe944792
Scale dfe 7 with velocity_equiv_weight
2026-05-13 09:06:59 -06:00
fc4d6ebf39
Add support for fine-tuning Hidream O1 ( #831 )
...
* Initial support for hidream. Lora keys likely need work
* Fix saving for hidream-o1
* Remove dependence on flash attention for hidream o1
* Fix gradient checkpointing for hidream o1
* A lot of fixes for hidream. Handle loading and saving as comfy model.
* Omit layers not used in comfy. Fix issue with lora loading keys in comfy
* Version bumpo
2026-05-12 11:15:16 -06:00
f38de2a2fe
Add tipsv2 locally and fix gradient checkpointing for it
2026-05-10 14:47:44 -06:00
d144cb5ea6
Switch to uplot for loss graph and rework performance of graph. It is significantly more perfromant now.
2026-05-07 07:38:00 -06:00
a12ddd72a1
Change the velocity weight cap on dfe 9
2026-05-07 07:37:05 -06:00
6bb8acbffc
Add agreement_threshold default of 0.6 to automagic 2
2026-05-05 19:13:00 -06:00
963a9f42b2
Add decode latent to wan 2.1 models. Add gradinet checkpointing to wan vae.
2026-05-05 11:30:16 -06:00
4260a3c5b6
Add optimizer test suite and make minor speed adjustments to Automagicv2
2026-05-05 10:02:30 -06:00
aeca7fe404
Add Automagic v2 optimizer. It uses significantly less vram and is much more efficient.
2026-05-05 09:09:07 -06:00
0d91fcee9e
Allow edit of captioning job
2026-04-30 16:00:16 -06:00
eadc9a58af
Version bump
2026-04-30 06:00:58 -06:00
e9ab387dfd
Fixed issue with qwen image edit models when using multiple control images when not caching text embeddings.
2026-04-30 11:59:12 +00:00
deb409085a
Made it possible to use the flux 2 small decoder VAE when setting the vae_path manually for flux2 models
2026-04-30 05:42:31 -06:00
7ccec8ec2c
Add checkpointing and a proper decode for flux 2 VAEs so they can be used with DFE
2026-04-30 04:27:13 -06:00
b4f0efb025
Version Bump
2026-04-28 13:42:00 -06:00
af6458d1b5
Enable caching of ACE step latents.
2026-04-28 13:39:20 -06:00
77b8765939
Fix issue with hover styling on select components.
2026-04-28 10:55:20 -06:00
43989cc19e
Add advanced config section to the captioner
2026-04-28 10:49:46 -06:00
f972b750e6
Performance improvements to captioner. Add ability to set a default caption for ace captioner to avoid needing to caption the audio so we only transcribe
2026-04-28 10:15:11 -06:00
acc6a36214
Scale DFE 9 to a velocity equiv weight to match flow matching gradient strength. Probably need to rework all DFEs to do this as the math checks out.
2026-04-28 09:10:02 -06:00
1fc4ad3979
Add sapiens2 as a diffusion feature extractor
2026-04-27 15:59:03 -06:00
67d67f8c1d
Ignore hidden files when captioning
2026-04-25 17:38:31 -06:00
998a02f30e
Yet another pseudo_huber fix. Still drinking coffee.
2026-04-19 10:05:40 -06:00
fc85410c9a
Fix issue with precision on pseudo_huber loss
2026-04-19 10:02:27 -06:00
20a99258b8
Fix issue is is else on pseudo_huber loss
2026-04-19 09:59:18 -06:00
f4445cd78c
Added psuedo_huber loss
2026-04-19 09:51:46 -06:00
488878f354
Use hidden layers in the loss for DFE 7 and 8
2026-04-18 13:07:38 -06:00
beb40ae29b
Add DFE8 with partial step
2026-04-17 17:40:16 -06:00
7c4f18ce51
Fix ernie unpatchify
2026-04-17 12:03:39 -06:00
8cb9649382
Add decode latent to ernie pipe
2026-04-17 12:01:09 -06:00
67048df9f9
Merge branch 'main' of github.com:ostris/ai-toolkit
2026-04-17 06:05:40 -06:00
be54094704
Version bump
2026-04-17 06:05:26 -06:00
Jaret Burkett