Performance Improvements - #775

Merged

Hananel-Hazan merged 15 commits into

Aug 4, 2026

Commits on Jul 24, 2026

Reduce CPU->GPU transfer every time step with learning.
C-Earl
committed
aad3361
Foldable pipelines
C-Earl
committed
1b6e4a2
More files for foldable pipelines
C-Earl
committed
2428476
Sparse compute
C-Earl
committed
677d196

Commits on Jul 26, 2026

Folding pipeline computation
C-Earl
committed
884010b
Fixed indentation error with learning rule class
C-Earl
committed
48c98c0
Test cases and minor bug fixes
C-Earl
committed
6f3c5a6
Benchmark tests
C-Earl
committed
a81c05d

Commits on Aug 4, 2026

black formater
Hananel-Hazan
committed
79de6b4
Correctness fixes for the folded pipeline

Hananel-Hazan
and
claude
committed
b0b055b
Performance: fold cache, rank-1 MSTDP, cached MSTDPET decays

Hananel-Hazan
and
claude
committed
e308123
Refactor docstring formatting in TorchvisionDatasetWrapper for improved readability
Hananel-Hazan
committed
951e2e0
Improve docstring formatting in TorchvisionDatasetWrapper for clarity
Hananel-Hazan
committed
ddd5163
fixing
Hananel-Hazan
committed
5215a46
Merge branch 'master' into performance
Hananel-Hazan
authored
f3cdad0