Commit Graph

1 Commits

Author SHA1 Message Date
Krzysztof Rymski df162ead7c Implementation of tiled attention with bf16 and circular buffers which reduces memory requirements by 4x on longer context on gemma models.
It also supports better parallelism for small batch sizes / small models.
It also is able to utilize VDPBF16PS for nice 2x improvement on avx512

PiperOrigin-RevId: 874517319
2026-02-24 03:26:49 -08:00