qdelta / lq-nets Goto Github PK
View Code? Open in Web Editor NEWLicense: MIT License
License: MIT License
โ Add sparsity-inducing optimization technique (e.g., soft thresholding) to the basis update step in the LQ-Net algorithm, achieving mixed-precision quantization schemes by inducing zeros in the basis vector.
โ Tune the strength of the sparsity-inducing optimization and observe the tradeoff between accuracy and average precision. How is it compared with the original LQ-Net performance?
Currently gradients of quantized activation are masked to 0 if the original activation is greater than maximum quantization level or less than minimum quantization level. This is consistent with the LQ-Nets paper.
But in the microsoft/LQ-Nets repo the implementation only masked the "greater than maximum" part but not the "less than minimum" part. Maybe we can try it.
ADD print info about latency and memory consumption to the training code.
It should be KB, not MB
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.