Crowd-benchmarking and crowd-tuning deep learning

This is a prototype of our open-source client based on the open-source Collective Knowledge framework (https://github.com/ctuning/ck) to help the community co-design Pareto-efficient software and hardware stack for deep learning and other emerging workloads in terms of speed, accuracy, energy and costs across diverse models, data sets, libraries, frameworks and devices provided provided by volunteers similar to SETI@HOME:
* http://cKnowledge.org/dnn-crowd-benchmarking-results
* http://cKnowledge.org/request
* http://cKnowledge.org/ai
* https://github.com/ctuning/ck

License: permissive 3-clause BSD

Source code: https://github.com/dividiti/crowdsource-video-experiments-on-android

Maintainers: non-profit cTuning foundation (cTuning.org) and dividiti (dividiti.com)
Read more
4.5
30 total
5
4
3
2
1
Loading...

What's New

Added ArmCL 18.05 OpenCL library with 3 MobileNets v1 scenario; updated scoreboard at http://cKnowledge.org/dnn-crowd-benchmarking-results (sorting by throughput and latency)
Read more

Additional Information

Updated
July 6, 2018
Size
1.9M
Installs
1,000+
Current Version
2.12
Requires Android
4.2 and up
Content Rating
Everyone
Interactive Elements
Users Interact
Permissions
Offered By
dividiti Ltd
©2018 GoogleSite Terms of ServicePrivacyDevelopersArtistsAbout Google|Location: United StatesLanguage: English
By purchasing this item, you are transacting with Google Payments and agreeing to the Google Payments Terms of Service and Privacy Notice.