Skip to content

v0.3.8: Add optional batch_size override to EQ_predict (#180)

Choose a tag to compare

View all tags
@github-actions github-actions released this 30 Apr 20:00
· 7 commits to main since this release
v0.3.8
8aeb217
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode. 
Inference is forward-only with no gradients, so larger batches than the
training default usually fit on the same GPU. Adds a `batch_size` CLI
field to `predict.yaml` (default null = inherit from training) and
applies it to `train_cfg.datamodule.batch_size` in memory before the
datamodule is instantiated, leaving the saved `resolved_config.yaml`
untouched.

Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>
Assets 6
Loading