-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathindex.html
183 lines (170 loc) · 8.66 KB
/
index.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
<!DOCTYPE html>
<!--[if IE]><![endif]-->
<html>
<head>
<meta charset="utf-8">
<meta http-equiv="X-UA-Compatible" content="IE=edge,chrome=1">
<title>Tensor for F# | Tensor for F# </title>
<meta name="viewport" content="width=device-width">
<meta name="title" content="Tensor for F# | Tensor for F# ">
<meta name="generator" content="docfx 2.35.0.0">
<link rel="shortcut icon" href="images/tensor.ico">
<link rel="stylesheet" href="styles/docfx.vendor.css">
<link rel="stylesheet" href="styles/docfx.css">
<link rel="stylesheet" href="styles/main.css">
<meta property="docfx:navrel" content="toc.html">
<meta property="docfx:tocrel" content="toc.html">
<meta property="docfx:rel" content="">
</head>
<body data-spy="scroll" data-target="#affix">
<div id="wrapper">
<header>
<nav id="autocollapse" class="navbar navbar-inverse ng-scope" role="navigation">
<div class="container">
<div class="navbar-header">
<button type="button" class="navbar-toggle" data-toggle="collapse" data-target="#navbar">
<span class="sr-only">Toggle navigation</span>
<span class="icon-bar"></span>
<span class="icon-bar"></span>
<span class="icon-bar"></span>
</button>
<a class="navbar-brand" href="index.html">
<img id="logo" class="svg" src="images/tensor.svg" alt="">
</a>
</div>
<div class="collapse navbar-collapse" id="navbar">
<form class="navbar-form navbar-right" role="search" id="search">
<div class="form-group">
<input type="text" class="form-control" id="search-query" placeholder="Search" autocomplete="off">
</div>
</form>
</div>
</div>
</nav>
<div class="subnav navbar navbar-default">
<div class="container hide-when-search" id="breadcrumb">
<ul class="breadcrumb">
<li></li>
</ul>
</div>
</div>
</header>
<div class="container body-content">
<div id="search-results">
<div class="search-list"></div>
<div class="sr-items">
<p><i class="glyphicon glyphicon-refresh index-loading"></i></p>
</div>
<ul id="pagination"></ul>
</div>
</div>
<div role="main" class="container body-content hide-when-search">
<div class="article row grid">
<div class="col-md-10">
<article class="content wrap" id="_content" data-uid="">
<h1 id="tensor-for-f">Tensor for F#</h1>
<p><a href="https://www.travis-ci.org/DeepMLNet/DeepNet"><img src="https://www.travis-ci.org/DeepMLNet/DeepNet.svg?branch=master" alt="Build Status"></a>
<a href="https://ci.appveyor.com/project/surban/deepnet/branch/master"><img src="https://ci.appveyor.com/api/projects/status/7qrfufbj0mvb6llv/branch/master?svg=true" alt="Build status"></a></p>
<p>A <em>tensor</em> is an n-dimensional array of an arbitrary data type (for example <code>single</code> or <code>double</code>).
Tensors of data type <code>'T</code> are implemented by the <a class="xref" href="api/Tensor.Tensor-1.html">Tensor<'T></a> type.</p>
<p>A tensor can be either stored in host memory or in the memory of a GPU computing device.
Currently only nVidia cards implementing the <a href="https://developer.nvidia.com/cuda-zone">CUDA API</a> are supported.
The API for host and GPU stored tensors is mostly equal, thus a program can make use of GPU accelerated operations without porting effort.</p>
<p>The tensor library provides functionality similar to <a href="http://docs.scipy.org/doc/numpy-1.10.0/reference/arrays.html">Numpy's Ndarray</a> and <a href="http://www.mathworks.com/help/matlab/matrices-and-arrays.html">MATLAB arrays</a>, including vector-wise operations, reshaping, slicing, broadcasting, masked assignment, reduction operations and BLAS operations.</p>
<p>This open source library is written in <a href="http://fsharp.org/">F#</a> and targets the <a href="https://github.com/dotnet/standard/blob/master/docs/versions/netstandard2.0.md">.NET Standard 2.0 platform</a> with Linux, MacOS and Microsoft Windows as supported operating systems.</p>
<h3 id="features-provided-by-the-core-tensor-library">Features provided by the core Tensor library</h3>
<ul>
<li>Core features<ul>
<li>n-dimensional arrays (tensors) in host memory or on CUDA GPUs</li>
<li>element-wise operations (addition, multiplication, absolute value, etc.)</li>
<li>basic linear algebra operations (dot product, SVD decomposition, matrix inverse, etc.)</li>
<li>reduction operations (sum, product, average, maximum, arg max, etc.)</li>
<li>logic operations (comparison, and, or, etc.)</li>
<li>views, slicing, reshaping, broadcasting (similar to NumPy)</li>
<li>scatter and gather by indices</li>
<li>standard functional operations (map, fold, etc.)</li>
</ul>
</li>
<li>Data exchange<ul>
<li>read/write support for HDF5 (.h5)</li>
<li>interop with standard F# types (Seq, List, Array, Array2D, Array3D, etc.)</li>
</ul>
</li>
<li>Performance<ul>
<li>host: SIMD and Intel MKL accelerated operations</li>
<li>CUDA GPU: all operations performed locally on GPU and cuBLAS used for matrix operations</li>
</ul>
</li>
</ul>
<h3 id="additional-features-provided-by-tensoralgorithm">Additional features provided by Tensor.Algorithm</h3>
<ul>
<li>Matrix algebra (integer, rational)<ul>
<li>Row echelon form</li>
<li>Smith normal form</li>
<li>Kernel, co-kernel and (pseudo-)inverse</li>
</ul>
</li>
<li>Matrix decomposition (floating point)<ul>
<li>Principal component analysis (PCA)</li>
<li>ZCA whitening</li>
</ul>
</li>
<li>Misc<ul>
<li>Bezout's identity</li>
<li>Loading of NumPy's .npy and .npz files.</li>
</ul>
</li>
</ul>
<h2 id="news">News</h2>
<ul>
<li><p><strong>2018/05/07</strong>. Packages are now available on public NuGet. <a href="https://www.nuget.org/packages/Tensor">Tensor NuGet package</a> and <a href="https://www.nuget.org/packages/Tensor.Algorithm">Tensor.Algorithm NuGet package</a>.</p>
</li>
<li><p><strong>2018/05/05</strong>. <a href="articles/ReleaseNotes.html">The CUDA SDK is no longer required for using GPU acceleration</a>.
An up-to-date nVidia GPU driver must be installed, however.</p>
</li>
<li><p><strong>2018/05/02</strong>. <a href="articles/Benchmarks.html">Initial benchmarks are available.</a></p>
</li>
</ul>
<h2 id="documentation">Documentation</h2>
<p>To get an overview of available functions by category, see <a href="articles/Tensor.html">Tensor at a glance</a>.
We also provide <a class="xref" href="api/Tensor.html">full reference documentation</a>.</p>
<p>To start using the library, follow the <a href="articles/Guide-Installation.html">installation guide</a> and the <a href="articles/Guide-Creation.html">getting started guide</a>.</p>
<h2 id="example">Example</h2>
<p>Check out our example project that demonstrates essential features of the library at <a href="https://github.com/DeepMLNet/Tensor.Sample">https://github.com/DeepMLNet/Tensor.Sample</a>.</p>
<h2 id="authors">Authors</h2>
<p><a href="https://github.com/surban/">Sebastian Urban</a><br>Marcus Basalla<br>Johannes Langer<br>Wiebke Köpp </p>
</article>
</div>
<div class="hidden-sm col-md-2" role="complementary">
<div class="sideaffix">
<div class="contribution">
<ul class="nav">
<li>
<a href="https://github.com/DeepMLNet/DeepNet/blob/6b91e28ab3e30de89fe76f63e2f97544a6bbd478/Tensor/Tensor.Docs/index.md/#L1" class="contribution-link">Improve this Doc</a>
</li>
</ul>
</div>
<nav class="bs-docs-sidebar hidden-print hidden-xs hidden-sm affix" id="affix">
<!-- <p><a class="back-to-top" href="#top">Back to top</a><p> -->
</nav>
</div>
</div>
</div>
</div>
<footer>
<div class="grad-bottom"></div>
<div class="footer">
<div class="container">
<span class="pull-right">
<a href="#top">Back to top</a>
</span>
<span>Generated by <strong>DocFX</strong></span>
</div>
</div>
</footer>
</div>
<script type="text/javascript" src="styles/docfx.vendor.js"></script>
<script type="text/javascript" src="styles/docfx.js"></script>
<script type="text/javascript" src="styles/main.js"></script>
</body>
</html>