24 open source projects by ctuning

Collective Knowledge components for TensorFlow (code, data sets, models, packages, workflows):

Collective Knowledge workflow for the MILEPOST GCC (machine learning based compiler). See how it is used in the collaborative project with the Raspberry Pi foundation to support collaborative research for multi-objective autotuning and machine learning techniques, and prototype reproducible papers with portable workflows:

Shared artifacts in the Collective Knowledge Format as a proof-of-concept to reproduce our recent Collective Mind- and Collective Knowledge-related papers

Collective Knowledge workflow for ARM's workload automation tool: an open framework for gathering and sharing knowledge about system design and optimization using real-world workloads.

Collective Knowledge framework (CK) helps to organize black-box research software as a database of reusable components and micro-services with common APIs, automation actions and extensible meta descriptions. See real-world use cases from Arm, General Motors, ACM, Raspberry Pi foundation and others:

Public data sets and their properties in the Collective Knowledge Format with JSON API and JSON meta information to be easily pluggable to customizable and reproducible CK experimental workflows (such as collaborative program analysis and optimization):

Portable and customizable Collective Knowledge workflows for MXNet:

CK repository with components and automation actions to enable portable workflows across diverse platforms including Linux, Windows, MacOS and Android. It includes software detection plugins and meta packages (code, data sets, models, scripts, etc) with the possibility of multiple versions to co-exist in a user or system environment:

Android application to participate in experiment crowdsourcing (such as workload crowd-benchmarking and crowd-tuning) using Collective Knowledge Framework and open repositories of knowledge:

Collective Knowledge web extension to browse CK repositories, visualize interactive graphs and articles, render CK-based websites, implement simple web services with JSON API (for example to crowdsource experiments or unify access to DNN). Demos of interactive articles, graphs and crowdsourced experiments:

CK dissemination modules

Collective Knowledge extension with unified and customizable benchmarks (with extensible JSON meta information) to be easily integrated with customizable and portable Collective Knowledge workflows. You can easily compile and run these benchmarks using different compilers, environments, hardware and OS (Linux, MacOS, Windows, Android). More info:

Collective Knowledge packages for various mathematical libs to be plugged into portable and customizable CK research workflows:

CK workflow, portable packages and other artifacts for the ReQuEST-ASPLOS'18 submission:

Integration of Caffe2 to Collective Knowledge workflow framework to provide unified CK JSON API for AI (customized builds across diverse libraries and hardware, unified AI API, collaborative experiments, performance optimization and model/data set tuning):

Demonstration of compiler autotuning, crowd-tuning and machine learning on RPi3 via customizable Collective Knowledge workflow framework with a portable package manager. This technology supports Pareto-efficient software/hardware co-design tournaments of deep learning in terms of speed, accuracy, energy, costs:

Shared artifacts and workflows from the EU FP7 Carp project in the customizable, portable and reusable Collective Knowledge format:

Public scenarios to crowdsource experiments (such as DNN crowd-benchmarking and crowd-tuning) using Collective Knowledge Framework across diverse mobile devices provided by volunteers. Results are continuously aggregated at the open repository of knowledge:

Collective Knowledge extension to crowdsource bug detection in OpenCL compilers using CLSmith tool from Imperial College London

Shared artifacts and workflows from the EPSRC Pamela project in the customizable, portable and reusable Collective Knowledge format:

CK-NNTest: collaboratively validating, benchmarking and optimizing neural net operators across platforms, frameworks and datasets

This repository is outdated! Join the open MLPerf workgroup to participate in the development of the next generation of automation workflows for MLPerf benchmarks:

Portable and customizable Collective Knowledge workflows for TVM and VTA:

Collective Knowledge repository with actions to unify the access to different predictive analytics engines (scipy, R, DNN) from software, command line and web-services via CK JSON API: