Skip to content
build

Implementation of tiled attention with bf16 and circular buffers which reduces memory requirements by 4x on longer context on gemma models. #6483

Sign in to view logs

Implementation of tiled attention with bf16 and circular buffers which reduces memory requirements by 4x on longer context on gemma models.

Implementation of tiled attention with bf16 and circular buffers which reduces memory requirements by 4x on longer context on gemma models. #6483

Triggered via pull request February 23, 2026 17:07
@copybara-servicecopybara-service[bot]
synchronize #839
test_864904207
Status Failure
Total duration 13m 47s
Artifacts 2

build.yml

on: pull_request
bazel
8m 6s
bazel
Matrix: build
Fit to window
Zoom out
Zoom in

Annotations

1 error and 1 warning
windows-latest (windows) Release
Process completed with exit code 1.
bazel
Failed to restore: Cache service responded with 400

Artifacts

Produced during runtime
Name Size Digest
gemma-macos-latest-make-Release
2.05 MB
sha256:c1e21eee6688984a4d8f5793744cff1eda90aac191b1cbdec776844b39f048ae
gemma-ubuntu-latest-make-Release
5.87 MB
sha256:336bf0c35479bd2c17ed1322cd48f76e34905ba2eb57fcba36fe451e3a1f65b9