Kristian MyramandClaude Sonnet 5.5 b7c78065db S2.61: cache the sine positions, pool without permutes ...

Sine position encodings depend only on the sizes: built once (150 ms of a frame).
do_pool takes the maximum over grouped cells instead of permute + max_pool2d +
permute (tested equal, odd sizes included). Encode 2.6 s -> 2.4 s per frame in
release on 16 threads (3.4 s at the start of the speed work); reference tests
unchanged. Chunked global attention was tried and did not help.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>

2026-10-30 08:43:31 +01:00

..

op_bench.rs

S2.61: cache the sine positions, pool without permutes

2026-10-30 08:43:31 +01:00

sam2_bench.rs

S2.61: cache the sine positions, pool without permutes

2026-10-30 08:43:31 +01:00