Kent Overstreet 7f6d94417e amygdala lib: move_to_cpu=True to avoid bf16 SVD on CUDA ... torch.svd doesn't support bf16 on CUDA; moving activations to CPU first makes pca_aggregator work. Co-Authored-By: Proof of Concept <poc@bcachefs.org>		2026-04-18 22:19:23 -04:00
..
amygdala_stories	amygdala stories: disambiguation scenarios for fragmented concepts	2026-04-18 21:08:23 -04:00
amygdala_training	amygdala lib: move_to_cpu=True to avoid bf16 SVD on CUDA	2026-04-18 22:19:23 -04:00
apollo_plugin	training: move to dedicated subprocess with ZMQ communication	2026-04-16 02:04:26 -04:00
research	research: latent reasoning integration plans for Qwen 3.5 27B	2026-04-12 15:50:09 -04:00
DESIGN.md	training: move to dedicated subprocess with ZMQ communication	2026-04-16 02:04:26 -04:00
pyproject.toml	training: move to dedicated subprocess with ZMQ communication	2026-04-16 02:04:26 -04:00