1
- [ ![ Build Status] ( https://migueldeicaza.visualstudio.com/TorchSharp/_apis/build/status/TorchSharp-CI )] ( https://migueldeicaza.visualstudio.com/TorchSharp/_build/latest?definitionId=5 )
2
-
3
1
TorchSharp
4
2
==========
5
3
6
4
TorchSharp is a .NET library that provides access to the library that powers
7
5
PyTorch. It is a work in progress, but already provides a .NET API that can
8
- be used to perform various operations on Tensors.
6
+ be used to perform (1) various operations on ATen Tensors; (2) scoring of
7
+ TorchScript models; (3) Training of simple neural networks.
9
8
10
- Our current focus is to bind the entire C API surfaced by libtorch.
9
+ Our current focus is to bind the entire API surfaced by libtorch.
11
10
12
11
Things that you can try:
13
12
14
13
``` csharp
15
- using TorchSharp ;
14
+ using AtenSharp ;
16
15
17
16
var x = new FloatTensor (100 ); // 1D-tensor with 100 elements
18
17
FloatTensor result = new FloatTensor (100 );
@@ -29,27 +28,20 @@ We have a chat room here:
29
28
30
29
https://gitter.im/xamarin/TorchSharp
31
30
32
- Status
33
- ======
34
-
35
- Currently I am binding the Tensor APIs, opportunities to bind other APIs
36
- are wide open.
37
-
38
- [ API Documentation] ( https://xamarin.github.io/TorchSharp/api/TorchSharp.html )
39
-
40
31
Running this
41
32
============
42
- To run this, you will need libtorch and its dependencies (on Mac that
33
+ To run this, you will need libTorchSharp, libTorch and its dependencies (on Mac that
43
34
includes libc10) installed in a location that your dynamic linker can
44
35
get to.
45
36
46
37
47
- Running On Linux (on Windows)
38
+ Running On Windows
48
39
-----------------------------
49
40
50
- The following was tested on Ubuntu 18.04 on the Windows Subsytem for Linux .
41
+ The following was tested on Windows.
51
42
52
- 1 . Install the [ .NET Core SDK] ( https://www.microsoft.com/net/download ) .
53
- 2 . Install [ libtorch] ( https://pytorch.org/ ) , and make it available to the
43
+ 1 . Install [ libTorch] ( https://pytorch.org/ ) , and make it available to the
44
+ dynamic linker.
45
+ 2 . Download and compile [ libTorchSharp] ( https://github.com/interesaaat/LibTorchSharp/ ) , and make it available to the
54
46
dynamic linker.
55
- 3 . Run ` dotnet run ` in the ` Tester ` subfolder.
47
+ 3 . Run ` dotnet run ` in the ` Test ` subfolder.
0 commit comments