Skip to content

fix: improve numerical stability by conditionally using float32 in Anima with fp16 training#2303

Merged
kohya-ss merged 1 commit intomainfrom
sd3
Apr 2, 2026
Merged

fix: improve numerical stability by conditionally using float32 in Anima with fp16 training#2303
kohya-ss merged 1 commit intomainfrom
sd3

Conversation

@kohya-ss
Copy link
Copy Markdown
Owner

@kohya-ss kohya-ss commented Apr 2, 2026

Version 0.10.3

Closes #2293

…ima (#2302)

* fix: improve numerical stability by conditionally using float32 in block computations

* doc: update README for improvement stability for fp16 training on Anima in version 0.10.3
@kohya-ss kohya-ss merged commit 51435f1 into main Apr 2, 2026
5 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

NaN errors with fp16 training on Anima.

1 participant