> ## Content Index
> Fetch the complete content index at: https://sudormrf.run/llms.txt
> Use this file to discover other available public pages before exploring further.

# Torch Wiki
- URL: https://sudormrf.run/2016/10/21/torch-wiki/
- Published: 2016-10-21T14:20:58.000Z
- Updated: 2022-09-18T06:45:53.000Z
- Author: Junho Cho
- Tags: Deep Learning, Torch, #Import 2026-10-24 03:36

개인적으로 Torch를 주력으로 사용하려합니다. 필요할 때마다 빠르게 검색하기 위해 모아놓은 것들을 wiki 처럼 공유합니다.

---

# [Official Homepage](torch.ch)

- Installation is easy.

# Tutorials

### Very useful

[15 min Lua](http://tylerneylon.com/a/learn-lua/?ref=sudormrf.run)

[Torch Video Tutorials](https://github.com/Atcold/torch-Video-Tutorials?ref=sudormrf.run)

[Bayarea DL summmer](https://github.com/alexbw/bayarea-dl-summerschool?ref=sudormrf.run)

[slide](http://mclab.eic.hust.edu.cn/valse2016/dl/%E7%9F%B3%E8%91%86%E5%85%89.pdf?ref=sudormrf.run)

### ref when needed

[demos](https://github.com/torch/demos?ref=sudormrf.run)

[DOCS](http://torch7.readthedocs.io/en/rtd/index.html?ref=sudormrf.run)

[tutorial](https://github.com/torch/tutorials?ref=sudormrf.run)

# Useful Libraries

[Torchnet](https://github.com/torchnet/torchnet?ref=sudormrf.run#tntengine)

- torchnet is a framework for torch which provides a set of abstractions aiming at encouraging code re-use as well as encouraging modular programming.  
At the moment, torchnet provides four set of important classes:  
Dataset: handling and pre-processing data in various ways.  
Engine: training/testing machine learning algorithm.  
Meter: meter performance or any other quantity.  
Log: output performance or any other string to file / disk in a consistent manner.  
For an overview of the torchnet framework, please also refer to this paper.

[Facebook : lua library](https://github.com/facebook/fblualib/?ref=sudormrf.run)

- Includes **debugger (very useful)**

[Deepmind : Cephes Mathematical Funcions Library](http://deepmind.github.io/torch-cephes/?ref=sudormrf.run)

- Provides and wraps the mathematical functions from the Cephes mathematical library, developed by Stephen L. Moshier. This C library provides a lot of mathematical functions. It is used, among many other places, at the heart of SciPy.

[Twitter Autograd](https://github.com/twitter/torch-autograd?ref=sudormrf.run)

- compute gradient automatically
- similar to tensorflow, theano auto differentiation.
- Useful to design very wierd architectures.
- ref: [Bayarea DL summmer](https://github.com/alexbw/bayarea-dl-summerschool?ref=sudormrf.run)

```lua
grad = require 'autograd'
grad.optimize(true) -- global
local df = grad(f, { optimize = true }) -- for this function only
local grads = df(params)

```

[conda](https://github.com/alexbw/conda-lua-recipes?ref=sudormrf.run)

[OptNet - reducing memory usage in torch neural networks](https://github.com/fmassa/optimize-net?ref=sudormrf.run)

- reduce memory

[Display](https://github.com/szym/display?ref=sudormrf.run)

- Web broswer
- Useful with SSH
- Almost only GUI with torch.
- socket based. Listening at certain port
- Developed from [gfx.js : a browser-based graphics server](https://github.com/clementfarabet/gfx.js/?ref=sudormrf.run)
- Depracated. Can use terminal

# From Torch.ch

- [torch](https://github.com/torch/torch7/blob/master/README.md?ref=sudormrf.run)
- [nn](https://github.com/torch/nn/blob/master/README.md?ref=sudormrf.run)
- [optim](https://github.com/torch/optim/blob/master/README.md?ref=sudormrf.run)
- [gnuplot](https://github.com/torch/gnuplot/blob/master/README.md?ref=sudormrf.run)
- [paths](https://github.com/torch/paths/blob/master/README.md?ref=sudormrf.run)
- [image](https://github.com/torch/image/blob/master/README.md?ref=sudormrf.run)
- [trepl](https://github.com/torch/trepl/blob/master/README.md?ref=sudormrf.run)
- [cwrap](https://github.com/torch/cwrap/blob/master/README.md?ref=sudormrf.run)
- [qtlua](https://github.com/torch/qtlua/blob/master/README.md?ref=sudormrf.run)
- [awsome-torch](https://github.com/carpedm20/awesome-torch?ref=sudormrf.run)

# Neural Network Package

This package provides an easy and modular way to build and train simple or complex neural networks using [Torch](https://github.com/torch/torch7/blob/master/README.md?ref=sudormrf.run):

- Modules are the bricks used to build neural networks. Each are themselves neural networks, but can be combined with other networks using containers to create complex neural networks:  
  - \[Module\]  
  ([https://github.com/torch/nn/blob/master/doc/module.md#nn.Module](https://github.com/torch/nn/blob/master/doc/module.md?ref=sudormrf.run#nn.Module)): abstract class inherited by all modules;
  - [Containers](https://github.com/torch/nn/blob/master/doc/containers.md?ref=sudormrf.run#nn.Containers): container classes like [Sequential](https://github.com/torch/nn/blob/master/doc/containers.md?ref=sudormrf.run#nn.Sequential), [Parallel](https://github.com/torch/nn/blob/master/doc/containers.md?ref=sudormrf.run#nn.Parallel) and [Concat](https://github.com/torch/nn/blob/master/doc/containers.md?ref=sudormrf.run#nn.Concat);
  - [Transfer functions](https://github.com/torch/nn/blob/master/doc/transfer.md?ref=sudormrf.run#nn.transfer.dok): non-linear functions like [Tanh](https://github.com/torch/nn/blob/master/doc/transfer.md?ref=sudormrf.run#nn.Tanh) and [Sigmoid](https://github.com/torch/nn/blob/master/doc/transfer.md?ref=sudormrf.run#nn.Sigmoid);
  - [Simple layers](https://github.com/torch/nn/blob/master/doc/simple.md?ref=sudormrf.run#nn.simplelayers.dok): like [Linear](https://github.com/torch/nn/blob/master/doc/simple.md?ref=sudormrf.run#nn.Linear), [Mean](https://github.com/torch/nn/blob/master/doc/simple.md?ref=sudormrf.run#nn.Mean), [Max](https://github.com/torch/nn/blob/master/doc/simple.md?ref=sudormrf.run#nn.Max) and [Reshape](https://github.com/torch/nn/blob/master/doc/simple.md?ref=sudormrf.run#nn.Reshape);
  - [Table layers](https://github.com/torch/nn/blob/master/doc/table.md?ref=sudormrf.run#nn.TableLayers): layers for manipulating `table`s like [SplitTable](https://github.com/torch/nn/blob/master/doc/table.md?ref=sudormrf.run#nn.SplitTable), [ConcatTable](https://github.com/torch/nn/blob/master/doc/table.md?ref=sudormrf.run#nn.ConcatTable) and [JoinTable](https://github.com/torch/nn/blob/master/doc/table.md?ref=sudormrf.run#nn.JoinTable);
  - [Convolution layers](https://github.com/torch/nn/blob/master/doc/convolution.md?ref=sudormrf.run#nn.convlayers.dok): [Temporal](https://github.com/torch/nn/blob/master/doc/convolution.md?ref=sudormrf.run#nn.TemporalModules), [Spatial](https://github.com/torch/nn/blob/master/doc/convolution.md?ref=sudormrf.run#nn.SpatialModules) and [Volumetric](https://github.com/torch/nn/blob/master/doc/convolution.md?ref=sudormrf.run#nn.VolumetricModules) convolutions;
- Criterions compute a gradient according to a given loss function given an input and a target:  
  - [Criterions](https://github.com/torch/nn/blob/master/doc/criterion.md?ref=sudormrf.run#nn.Criterions): a list of all criterions, including [Criterion](https://github.com/torch/nn/blob/master/doc/criterion.md?ref=sudormrf.run#nn.Criterion), the abstract class;
  - [MSECriterion](https://github.com/torch/nn/blob/master/doc/criterion.md?ref=sudormrf.run#nn.MSECriterion): the Mean Squared Error criterion used for regression;
  - [ClassNLLCriterion](https://github.com/torch/nn/blob/master/doc/criterion.md?ref=sudormrf.run#nn.ClassNLLCriterion): the Negative Log Likelihood criterion used for classification;
- Additional documentation:  
  - [Overview](https://github.com/torch/nn/blob/master/doc/overview.md?ref=sudormrf.run#nn.overview.dok) of the package essentials including modules, containers and training;
  - [Training](https://github.com/torch/nn/blob/master/doc/training.md?ref=sudormrf.run#nn.traningneuralnet.dok): how to train a neural network using [StochasticGradient](https://github.com/torch/nn/blob/master/doc/training.md?ref=sudormrf.run#nn.StochasticGradient);
  - [Testing](https://github.com/torch/nn/blob/master/doc/testing.md?ref=sudormrf.run): how to test your modules.
  - [Experimental Modules](https://github.com/clementfarabet/lua---nnx/blob/master/README.md?ref=sudormrf.run): a package containing experimental modules and criteria.