Update log for RDBT Anima
Main timeline:
Now ~: rollback to cfg distillation.
v0.32 ~ v0.23: testing dmd2.
v0.23 ~ v0.1: cfg distilled.
Interesting old versions:
v0.29b: 12-step dmd2. Has best details and sharpest image. It was from RL that can "amplify" good features and can go beyond training data. However, this affect styles too much, not ideal for a base model.
(5/24/2026): b1 v0.37.1:
Improve high noise stability a little bit.
(5/18/2026): b1 v0.35:
No step distillation. Just guidance distilled.
I'm dropping step distillation. Anima official has their plan to do step distillation (aka, turbo, 4/8-step, or whatever). They have the money and recourse and full dataset. I don't. And my cheap step distillation is kind of sh*tty, tbh.
If you need higher stability or speed, you can stack the extracted cosmos dmd2 lora the anima-turbo, basically can achieve the same thing, probably even better. I prefer 0.2x cosmos dmd2 lora.
(5/12/2026): p3 v0.32.b:
Less step distillation (means higher diversity but less stability). 12 steps is still doable, 24 steps is recommended for complicated prompt.
Styles reinforcement learning. I did this in v0.29, but not in v0.32.
(5/10/2026): p3 v0.32:
No more green-ish, color shifting.
Trigger words have been reclassified to avoid model learning a unified style. See updated "Usage" section.
Old trigger words for backup (v0.29 and before):
"digital anime illustration": common 2d anime.
"digital art", 2d art but not anime, mostly digital art.
"anime sketch": simplified/unfinished anime drawing.
(4/29/2026) p3 v0.29.b: Further trained v0.29.
Known issues: v0.29 will add lots of details and make the images looks very sharp and clear. But this might affect other styles.
(4/27/2026): p3 v0.29: Distillation algorithm was almost completely rewritten.
Better details. This version can squeeze every single pixel out of the VAE.
(4/23/2026) p3 v0.27: Improved stability, details.
(4/18/2026) p3 v0.25: It's based on anima p3.
====== alpha testing ========
[base model version] [finetuned model version] [distillation method] [distillation version]
f = finetuned, cfg = cfg distilled, dmd2 = distribution matching distillation.
(4/14/2026) p3 v0.24f dmd2 b: Comparing with "v0.24f dmd2": Images should be sharper.
(4/11/2026) p3 v0.24f cfg b: Improved stability and diversity, comparing with previous 0.24f cfg.
(4/10/2026) p3 v0.24f dmd2: roughly 12~16 steps distillation. This is intentional. Low steps distillation = ai slop style without complex texture.
(4/8/2026) p3 v0.24f cfg: Rebased on preview3. Finetuned base model has -40% steps than v0.23. Less overfitted (?).
(4/4/2026) p2 v0.23f dmd2 b: Different distillation settings. Almost a 4 steps dmd2. Maximum stability.
(4/4/2026) p2 v0.23f dmd2: 8 steps dmd2. First dmd2 anima (?).
(3/28/2026) p2 v0.23f cfg: Rebased on preview2. Distillation: improved small details and stability (removed a regularization in distillation target and changed to second-order method).
Voting result: v0.20fd won. Thanks for the feedbacks.
(3/24/2026) preview1 v0.20f cfg b: Distillation: Different settings optimized for anime, high contrast and saturation.
(3/23/2026) p1 v0.20f cfg: Dataset: More furry. Finetuned base model: from v0.12 + more 100% steps. Distillation: Fixed noisy pixels this time, really.
(3/14/2026) preview v0.19f cfg b:
Updated dataset. Some private datasets have been dropped. You might notice the style changed.
Fixed high-freq artifacts in v0.12, now you should get a clear image without noisy pixels
b: Testing new distillation settings. Higher contrast. Aligned with common anime art.
(2/19/2026) preview v0.12f cfg:
Better stability and details, extended dataset.
(3/8/2026) preview v0.11f cfg 512px:
Prove of concept version for v0.12. Same dataset and settings as v0.12, except it was trained with 512px res.
Released by request, as it might be very useful. Running model in 512px and cfg1 is extremely fast (x10 faster, e.g. 30s -> 3s). If you don't have a beefy GPU. You can use this version to test your ideas/prompts in few seconds.
(2/12/2026) preview v0.6 cfg:
CFG distilled only. No finetuning. Cover images are using Animeyume v0.1.
(2/3/2026) preview v0.2f cfg:
Speedrun attempt, mainly for testing the training script. Limited training dataset. Only covered "1 person" images plus a little bit of "furry". But it works, and way better than what I expected.