Are you happy with your logging solution? Would you help us out by taking a 30-second survey? Click here


Star full 4f7b624809470f25b6493d5a7b30d9b9cb905931146e785d67c86ef0c205a402Star full 4f7b624809470f25b6493d5a7b30d9b9cb905931146e785d67c86ef0c205a402Star full 4f7b624809470f25b6493d5a7b30d9b9cb905931146e785d67c86ef0c205a402Star full 4f7b624809470f25b6493d5a7b30d9b9cb905931146e785d67c86ef0c205a402Star full 4f7b624809470f25b6493d5a7b30d9b9cb905931146e785d67c86ef0c205a402 (2 ratings)
Rated 5.0 out of 5
Subscribe to updates I use torch7

Statistics on torch7

Number of watchers on Github 7739
Number of open issues 231
Average time to close an issue 5 days
Main language C
Average time to merge a PR 2 days
Open pull requests 35+
Closed pull requests 29+
Last commit about 2 years ago
Repo Created about 6 years ago
Repo Last Updated over 1 year ago
Size 2.52 MB
Organization / Authortorch
Page Updated
Do you use torch7? Leave a review!
View open issues (231)
View torch7 activity
View on github
Fresh, new opensource launches 🚀🚀🚀
Trendy new open source projects in your inbox! View examples

Subscribe to our mailing list

Evaluating torch7 for your project? Score Explanation
Commits Score (?)
Issues & PR Score (?)

Join the chat at Build Status

Need help?

Torch Package Reference Manual

Torch is the main package in Torch7 where data structures for multi-dimensional tensors and mathematical operations over these are defined. Additionally, it provides many utilities for accessing files, serializing objects of arbitrary types and other useful utilities.

Torch Packages

  • Tensor Library
    • Tensor defines the all powerful tensor object that provides multi-dimensional numerical arrays with type templating.
    • Mathematical operations that are defined for the tensor object types.
    • Storage defines a simple storage interface that controls the underlying storage for any tensor object.
  • File I/O Interface Library
  • Useful Utilities
    • Timer provides functionality for measuring time.
    • Tester is a generic tester framework.
    • CmdLine is a command line argument parsing utility.
    • Random defines a random number generator package with various distributions.
    • Finally useful utility functions are provided for easy handling of torch tensor types and class inheritance.

Useful Links

torch7 open issues Ask a question     (View All Issues)
  • almost 3 years OpenBlas could not be compiled
  • almost 3 years torch.sum(x,x,2) not returning correct value
  • almost 3 years Error:expected arguments: CudaTensor~2D [CudaTensor2D] [float] CudaTensor2D CudaTensor2D
  • almost 3 years Convenience function to add a singleton dimension to a torch tensor.
  • almost 3 years not having neon does not imply x86
  • almost 3 years Calling 'random' method on CudaTensor fails.
  • almost 3 years TYPEStorage when given ptr acts weirds
  • almost 3 years Missing documents on several functions in torch class
  • about 3 years ARM64 NEON unknown register name 'r0' in asm
  • about 3 years Compilation on Raspberry Pi B+ armv6l
  • about 3 years can you recommend a IDE for me to debug lua?
  • about 3 years lots of strange issue on ubuntu16.04, never see it on ubuntu14.04
  • about 3 years VM Ware Player and UBUNTU / MINT installation error: CAN NOT COMPILE
  • about 3 years doesn't resize the result existing tensor
  • about 3 years How to run torch on 10 GPU system??
  • about 3 years MacOS 10.12 , qt fail
  • about 3 years [Feature request] Can we make index() accept 1xN or Nx1 LongTensors?
  • about 3 years Hi, what's the difference between the files?
  • about 3 years ./install-deps does not contemplate linux mint
  • about 3 years clearState() on gModule with moduleFromCriterion does not behave as expected
  • about 3 years NEON detection on aarch64 / TX1
  • about 3 years [Port] Portability problem introduced by THVector.c
  • about 3 years Handling Variable length sequences in training an RNN
  • about 3 years illegal memory access was encountered
  • about 3 years test stuck at GPU()
  • about 3 years [bug] torch.test() fails when the second time calling
  • about 3 years ../torch/File.lua:370 table index is nil
  • about 3 years on linux, could use linuxbrew instead of distribution-specific installs?
  • about 3 years THREAD FATAL ERROR: could not create lua state
  • about 3 years Architecture not supported
torch7 open pull requests (View All Pulls)
  • Making torch7 and his exports work with external lua; some bugfixes.
  • Add torch.load32 and torch.load64
  • torch.topk: use quickselect + quicksort
  • Added torch.equal function which performs a tensor equality check
  • not defining DLL_EXPORT by default
  • Added my opinions to ROADMAP
  • Define float/double math functions in generator
  • Update
  • Lua 5.3 integer support
  • Add {UShort, UInt, ULong}Tensor
  • External lua option: variables unification, propagation, export.
  • Remove junk values from right-singular vectors matrix in SVD
  • Make THInf type-specific
  • Add fmod(), remainder(), remove mod() and fix tensor operator % behavior
  • Alternate implementation of torch.chunk()
  • Fixed the Visual Studio CMake build.
  • Add function to get the amount of memory used by torch tensors
  • fix compilation issues under Windows
  • Document vec:addmv(scalar, mat, vec2) call
  • Added dyanamic dispatch for x86
  • add read/writePointer to be used in threads (de)serializePointer
  • Fix build on other platform than x86/arm with extension support.
  • Inroducing CPU "half" type for 16-bit float, copy and init only
  • Adding AVX Optimizations with Runtime Dispatch
  • half<->float conversion cleanup
  • New helper macro + fixes for old gcc
  • Disabling versioning for Android
  • allow torch to link with Intel MKL _ilp64 model (large integers)
  • Update
  • minor spelling tweaks
  • Update
  • Improved CmdLine help message formatting
  • for ubuntu 16.04
  • resolve torch/distro/issues/32
torch7 questions on Stackoverflow (View All Questions)
  • Torch7 alternative to MultiLabelMarginCriterion
  • Torch7 using weights with unbalanced training sets
  • Anaconda and Torch7 Install Issue Mac 10.10
  • Torch7 - not enough memory: you tried to allocate 0GB. Buy new RAM
  • Installing Torch7 with Luarocks on Windows with mingw build error
  • Simple Torch7 equivalent to numpy.roll
  • Sequence Classification i.e. assigning a single label to a whole sequence for LSTMs using Torch7
  • how to tackle HEX in torch7?
  • Torch7 Tensor Non-Contiguos Index (Similar to Numpy)
  • Installing Torch7. iPython installation error (mac)
  • How to install Torch7 on windows7?
  • Installing Torch7 on Kubuntu 14.04 LTS x64: "Error: Build error: Failed compiling object readline.o"
  • Call torch7 (Lua) function from python?
  • How to prepare data for torch7 deep learning convolutional neural network example?
  • Torch7 Lua, error loading module 'libpaths' (Linux)
  • Calculating Recall and precision for the Confusion matrix using torch7
  • Read images from local disk using torch7 while those images saved in different subfolders?
  • Installing Torch7 on Win7; cmake and PATH problems
  • Torch7 Access one element from a Tensor as a Tensor
  • plotting a 3D+colour scatter with gnuplot (on torch7)
  • Support Vector Machine in Torch7
  • Mean Centering a Tensor in torch7
  • 2-D PCA scatter on Neural-Netwok on Torch7
  • HTTP Server Using Lua/ Torch7
  • What is gradInput and gradOutput in Torch7's 'nn' package?
  • Fast way to initialize a tensor in torch7
  • torch7 : how to connect the neurons of the same layer?
  • torch7 : print matrix in text file with lines longer than 80 characters
  • torch7 neural network training error
  • interfacing Python and Torch7(Lua) via shared library
torch7 list of languages used
Other projects in C