https://github.com/avik-pal/nnpack.jl

Julia wrapper for NNPACK

https://github.com/avik-pal/nnpack.jl

Science Score: 23.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • CITATION.cff file
  • codemeta.json file
    Found codemeta.json file
  • .zenodo.json file
  • DOI references
  • Academic publication links
  • Committers with academic emails
    1 of 1 committers (100.0%) from academic institutions
  • Institutional organization owner
  • JOSS paper metadata
  • Scientific vocabulary similarity
    Low similarity (14.1%) to scientific vocabulary

Keywords

convolutional-neural-networks deep-learning deep-neural-networks machine-learning multicore-cpu nnpack wrapper
Last synced: 5 months ago · JSON representation

Repository

Julia wrapper for NNPACK

Basic Info
  • Host: GitHub
  • Owner: avik-pal
  • License: mit
  • Language: Julia
  • Default Branch: master
  • Size: 40 KB
Statistics
  • Stars: 1
  • Watchers: 2
  • Forks: 0
  • Open Issues: 0
  • Releases: 1
Topics
convolutional-neural-networks deep-learning deep-neural-networks machine-learning multicore-cpu nnpack wrapper
Created over 7 years ago · Last pushed over 5 years ago
Metadata Files
Readme License

README.md

NOTE

This package is not officially maintained. It has been merged in NNlib.jl which offers deep learning primitives for a variety of backends. The source code for NNPACK.jl is now maintained here so please redirect any issue and pull requests to NNlib.jl.

NNPACK.jl

This is a wrapper for the low level deep learning acceleration package for multicore CPUs NNPACK.

Usage Instructions

NNPACK can be directly used with NNlib without any dependency on this package. The NNPACK wrapper resides inside the NNlib package. For integration with other deep learning libraries this wrapper may be used.

Installation

This package works on julia 1.0. So to install it simply do

julia ] add https://github.com/avik-pal/NNPACK.jl

This will work only on Linux. Follow the Installation for OSX to get it working on OSX

Exported Functions

  1. conv
  2. ∇conv_data
  3. ∇conv_filter
  4. softmax
  5. maxpool

To know more in detail about this functions refer to the src/interface.jl file. If you are familiar with the NNlib APIs then its pretty simple to understand this.

There are other NNPACK functions available but their API is not exposed. For using them refer to the src/libnnpack.jl file.

Environment Variables

  1. NNPACK_CPU_THREADS: Controls the number of threads NNPACK is allowed to use. Defaults to 4.
  2. NNPACK_FAST_OPS: Enables operations on Float64 Arrays. However, you might loose precision and the final output will also be in Float32. Defaults to 1.

Installation for OSX

```

Install PeachPy

$ git clone https://github.com/Maratyszcza/PeachPy.git $ cd PeachPy $ sudo pip install --upgrade -r requirements.txt $ python setup.py generate $ sudo pip install --upgrade .

Install Ninja Build System

$ sudo apt-get install ninja-build $ pip install ninja-syntax

Build NNPack shared library

$ cd ~ $ git clone --recursive https://github.com/Maratyszcza/NNPACK.git $ cd NNPACK $ python ./configure.py $ cmake . -G Ninja -DCMAKEPOSITIONINDEPENDENTCODE=ON -DNNPACKLIBRARY_TYPE="shared" ```

After this copy the libnnpack.so to ~/.julia/packages/NNPACK/deps/usr/lib/libnnpack.so.

Known Issues

  1. Windows is not supported by NNPACK. There is unofficial support for windows in nnpack-windows but the API is a bit different.
  2. OSX build will fail.
  3. Travis tests fail due to Unsupported Hardware error.

Owner

  • Name: Avik Pal
  • Login: avik-pal
  • Kind: user
  • Location: Cambridge, MA
  • Company: Massachusetts Institute of Technology

PhD Student @mit || Prev: BTech CSE IITK

GitHub Events

Total
Last Year

Committers

Last synced: over 1 year ago

All Time
  • Total Commits: 41
  • Total Committers: 1
  • Avg Commits per committer: 41.0
  • Development Distribution Score (DDS): 0.0
Past Year
  • Commits: 0
  • Committers: 0
  • Avg Commits per committer: 0.0
  • Development Distribution Score (DDS): 0.0
Top Committers
Name Email Commits
Avik Pal a****l@i****n 41
Committer Domains (Top 20 + Academic)

Issues and Pull Requests

Last synced: 11 months ago

All Time
  • Total issues: 0
  • Total pull requests: 0
  • Average time to close issues: N/A
  • Average time to close pull requests: N/A
  • Total issue authors: 0
  • Total pull request authors: 0
  • Average comments per issue: 0
  • Average comments per pull request: 0
  • Merged pull requests: 0
  • Bot issues: 0
  • Bot pull requests: 0
Past Year
  • Issues: 0
  • Pull requests: 0
  • Average time to close issues: N/A
  • Average time to close pull requests: N/A
  • Issue authors: 0
  • Pull request authors: 0
  • Average comments per issue: 0
  • Average comments per pull request: 0
  • Merged pull requests: 0
  • Bot issues: 0
  • Bot pull requests: 0
Top Authors
Issue Authors
Pull Request Authors
Top Labels
Issue Labels
Pull Request Labels