tag | 7732d3f18a10301ec4ffca2e643087eda16463ba | |
---|---|---|
tagger | The Android Open Source Project <initial-contribution@android.com> | Wed Apr 13 23:29:54 2022 -0700 |
object | d62f84754c0d964d447c548b4d7137d059a74d51 |
Android Mainline 12.0.0 Release 63 (7993649,com.google.android.networkstack)
commit | d62f84754c0d964d447c548b4d7137d059a74d51 | [log] [tgz] |
---|---|---|
author | Android Build Coastguard Worker <android-build-coastguard-worker@google.com> | Mon Nov 15 06:22:57 2021 +0000 |
committer | Android Build Coastguard Worker <android-build-coastguard-worker@google.com> | Mon Nov 15 06:22:57 2021 +0000 |
tree | 420dd314b49b6a7f5772ae3cb6e5f738520572b8 | |
parent | ce6dad96d41226dff1c94feefaee8a30f8c10a48 [diff] | |
parent | 0cdefb826c9c26f407aa89f380987a0eda162ff4 [diff] |
Snap for 7914678 from 0cdefb826c9c26f407aa89f380987a0eda162ff4 to mainline-networking-release Change-Id: Ie2ce85c93fae2927290783c491e44427945b60dc
This is not an officially supported Google product.
ruy is a matrix multiplication library. Its focus is to cover the matrix multiplication needs of neural network inference engines. Its initial user has been TensorFlow Lite, where it is used by default on the ARM CPU architecture.
ruy supports both floating-point and 8bit-integer-quantized matrices.
ruy is designed to achieve high performance not just on very large sizes, as is the focus of many established libraries, but on whatever are the actual sizes and shapes of matrices most critical in current TensorFlow Lite applications. This often means quite small sizes, e.g. 100x100 or even 50x50, and all sorts of rectangular shapes. It's not as fast as completely specialized code for each shape, but it aims to offer a good compromise of speed across all shapes and a small binary size.
Some documentation will eventually be available in the doc/ directory, see doc/README.md.