OpenBLAS
rav1e
OpenBLAS | rav1e | |
---|---|---|
22 | 70 | |
5,983 | 3,586 | |
1.6% | 0.9% | |
9.8 | 9.1 | |
about 15 hours ago | 1 day ago | |
C | Assembly | |
BSD 3-clause "New" or "Revised" License | BSD 2-clause "Simplified" License |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
OpenBLAS
-
LLaMA Now Goes Faster on CPUs
The Fortran implementation is just a reference implementation. The goal of reference BLAS [0] is to provide relatively simple and easy to understand implementations which demonstrate the interface and are intended to give correct results to test against. Perhaps an exceptional Fortran compiler which doesn't yet exist could generate code which rivals hand (or automatically) tuned optimized BLAS libraries like OpenBLAS [1], MKL [2], ATLAS [3], and those based on BLIS [4], but in practice this is not observed.
Justine observed that the threading model for LLaMA makes it impractical to integrate one of these optimized BLAS libraries, so she wrote her own hand-tuned implementations following the same principles they use.
[0] https://en.wikipedia.org/wiki/Basic_Linear_Algebra_Subprogra...
[1] https://github.com/OpenMathLib/OpenBLAS
[2] https://www.intel.com/content/www/us/en/developer/tools/onea...
[3] https://en.wikipedia.org/wiki/Automatically_Tuned_Linear_Alg...
[4]https://en.wikipedia.org/wiki/BLIS_(software)
- Assume I'm an idiot - oogabooga LLaMa.cpp??!
-
Learn x86-64 assembly by writing a GUI from scratch
Yeah. I'm going to be helping to work on expanding CI for OpenBlas and have been diving into this stuff lately. See the discussion in this closed OpenBlas issue gh-1968 [0] for instance. OpenBlas's Skylake kernels do rely on intrinsics [1] for compilers that support them, but there's a wide range of architectures to support, and when hand-tuned assembly kernels work better, that's what are used. For example, [2].
[0] https://github.com/xianyi/OpenBLAS/issues/1968
[1] https://github.com/xianyi/OpenBLAS/blob/develop/kernel/x86_6...
[2] https://github.com/xianyi/OpenBLAS/blob/23693f09a26ffd8b60eb...
-
AI’s compute fragmentation: what matrix multiplication teaches us
We'll have to wait until part 2 to see what they are actually proposing, but they are trying to solve a real problem. To get a sense of things check out the handwritten assembly kernels in OpenBlas [0]. Note the level of granularity. There are micro-optimized implementations for specific chipsets.
If progress in ML will be aided by a proliferation of hyper-specialized hardware, then there really is a scalability issue around developing optimized matmul routines for each specialized chip. To be able to develop a custom ASIC for a particular application and then easily generate the necessary matrix libraries without having to write hand-crafted assembly for each specific case seems like it could be very powerful.
[0] https://github.com/xianyi/OpenBLAS/tree/develop/kernel
-
Trying downloading BCML
libraries mkl_rt not found in ['C:\python\lib', 'C:\', 'C:\python\libs'] ``` Install this and try again. Might need to reboot, never know with Windows https://www.openblas.net/
-
The Bitter Truth: Python 3.11 vs Cython vs C++ Performance for Simulations
There isn't any fortran code in the repo there itself but numpy itself can be linked with several numeric libraries. If you look through the wheels for numpy available on pypi, all the latest ones are packaged with OpenBLAS which uses Fortran quite a bit: https://github.com/xianyi/OpenBLAS
- Optimizing compilers reload vector constants needlessly
-
Just a quick question, can a programming language be as fast as C++ and efficient with as simple syntax like Python?
Sure - write functions in another language, export C bindings, and then call those functions from Python. An example is NumPy - a lot of its linear algebra functions are implemented in C and Fortran.
- OpenBLAS - optimized BLAS library based on GotoBLAS2 1.13 BSD version
-
How to include external libraries?
Read the official docs yet?
rav1e
-
Learn x86-64 assembly by writing a GUI from scratch
Sure. You'll see it very often in codec implementations. From rav1e, a fast AV1 encoder mostly written in Rust: https://github.com/xiph/rav1e/tree/master/src/x86
Large portions of the algorithm have been translated into assembly for ARM and x86. Shaving even a couple percent off something like motion compensation search will add up to meaningful gains.
Or the current reference implementation of JPEG: https://github.com/libjpeg-turbo/libjpeg-turbo/tree/main/sim...
-
SISVEL VP9/AV1 patent declared invalid in China
Again, if anything AOM would be the one restricting licenses to AV1 (if they chose to) except AOM has stated and also published AV1 in a way to allow license free access to development (which allows people to make forks of the official build like it's open source) and usage. (1)(2) I don't see why they would suddenly change this.
- Any new Opensource projects in (rust) looking for contributors. I want to start my journey as an OSS contributor.
- assembly from dav1d 1.1.0 now integrated into rav1e
-
A little script to parse large libraries to AV1, if you're interested
You can speed up the sampling process with --vmaf n_subsample=5, which in my experience works more accurately than either 2 or 4, possibly due to this bug/feature present in multiple encoders. You might also need to manually set the number of threads used for VMAF calculation with --vmaf n_threads=16, but YMMV.
-
rav1d: a Rust port of dav1d (currently experimental)
That remember me of https://github.com/xiph/rav1e which is an AV1 encoder
- A Safer High Performance AV1 Decoder
-
rav1e wrong mastering-display output?
I put in the request for ffmpeg passthrough mastering-display data a few years ago and haven't heard of any support yet.
-
HDR10, HDR10+, Dolby Vision with AV1?
It's getting there.. Initial steps for FFmepg: https://patchwork.ffmpeg.org/project/ffmpeg/list/?series=8444 rav1e: https://github.com/xiph/rav1e/pull/3000
-
Release Notes: Safari 16.4 Beta adds AV1 codec + hardware decode for WebRTC
It's entirely possible to re-use bits of other HW encoders for the first pass (motion estimation, etc).
What are some alternatives?
Eigen
SVT-AV1
GLM - OpenGL Mathematics (GLM)
dav1d - A read-only mirror of dav1d source code repository. The origin is at https://code.videolan.org/videolan/dav1d/
cblas - Netlib's C BLAS wrapper: http://www.netlib.org/blas/#_cblas
SVT-AV1 - Welcome to the GitHub repo for the SVT-AV1! This repo is set to read-only for archiving purposes. Please join us at https://gitlab.com/AOMediaCodec/SVT-AV1. We look forward to seeing you there
blaze
ffmpeg-build-script - The FFmpeg build script provides an easy way to build a static FFmpeg on OSX and Linux with non-free codecs included.
Boost.Multiprecision - Boost.Multiprecision
obs-amd-encoder - AMD Advanced Media Framework Encoder Plugin for Open Broadcaster Studio
ceres-solver - A large scale non-linear optimization library
libavif - libavif - Library for encoding and decoding .avif files