You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
* Ejhg osx arm64 lib fix (#21)
* Increase libtorch version to 2.4.0 and use libtorch for osx-arm64
- libtorch requires libomp.dylib, but it is only included with libtorch 2.4.0 and above
- compiling osx-arm64 against libtorch-macos-arm64 instead of conda py3.10 libraries source significantly improves performance (~6X on M2 Macbook Air)
* Clean up references to unused conda LibTorchArchiveSource conditions
* Update SHA hashes for libtorch 2.4.0 on Windows and Linux
* Update distribution filenames
* Set BuildLibTorchPackages to true to build artifacts in CI
* Pin mac-x64 target against latest libtorch 2.2.2
* Specify TargetOS to fix failing CI build
* Revert "Specify TargetOS to fix failing CI build"
This reverts commit bf723f5.
* Specify TargetOS to fix failing CI build for mac-x64
---------
Co-authored-by: Eugenio De Hoyos <[email protected]>
* Starting upgrade to 2.4.0
* Disable MacOS ARM64 builds
* Fixed CI/CD script
---------
Co-authored-by: Eugenio De Hoyos <[email protected]>
@@ -77,7 +77,7 @@ To change the TorchSharp package version update this [file](https://github.com/d
77
77
The TorchSharp package is pushed to nuget.org via Azure DevOps CI release pipeline. Assuming you're not building or updating the LibTorch packages
78
78
(`BuildLibTorchPackages` is `false` in [azure-pipelines.yml](azure-pipelines.yml)) this is pretty simple once you have the permissions:
79
79
80
-
1. Update the version number in [./build/BranchInfo.props](./build/BranchInfo.props) and in the [Release Notes](./RELEASENOTES.md) file and then submit a PR.
80
+
1. Update the version number in [./build/BranchInfo.props](./build/BranchInfo.props) and in the [Release Notes](./RELEASENOTES.md) file and then submit a PR.
81
81
82
82
Updating the major or minor version number should only be done after a discussion with repo admins. The patch number should be incremented by one each release and set to zero after a change to the major or minor version.
83
83
2. Integrate code to main and wait for CI to process
@@ -149,7 +149,7 @@ For this reason, we do the following
149
149
This project grabs LibTorch and makes a C API wrapper for it, then calls these from C#. When updating to a newer
150
150
version of PyTorch then quite a lot of careful work needs to be done.
151
151
152
-
0. Make sure you have plenty of disk space, e.g. 15GB.
152
+
0. Make sure you have plenty of disk space, e.g. 15GB.
153
153
154
154
1. Clean and reset to main
155
155
@@ -163,7 +163,7 @@ version of PyTorch then quite a lot of careful work needs to be done.
bytesthatisthelimit, **not**250*1024*1024. Inotherwords, itis250MB, not250MiB. NotethatWindowsExplorerwillshowfilesizesinKiB, notthousandsofbytes. Use'dir'fromaCMDwindowtogettheexactsizeinbytesfor each file. For example -- the file `libtorch_cpu.so` shows up as 511,872 KB in Windows Explorer, but 524,156,144 bytes in CMD. The 2.4% difference can be significant. Getting the partitioning right requires precision.
216
+
bytesthatisthelimit, **not**250*1024*1024. Inotherwords, itis250MB, not250MiB. NotethatWindowsExplorerwillshowfilesizesinKiB, notthousandsofbytes. Use'dir'fromaCMDwindowtogettheexactsizeinbytesfor each file. For example -- the file `libtorch_cpu.so` shows up as 511,872 KB in Windows Explorer, but 524,156,144 bytes in CMD. The 2.4% difference can be significant. Getting the partitioning right requires precision.
217
217
218
218
If the combined size of the files going into a part is smaller than 250MB, then everything is fine, and there is no need to split the part. It can be singular. If that is not the case, then the part should be fragmented into two or more parts that are linked together by their names.
219
219
@@ -229,12 +229,12 @@ On Linux / Mac:
229
229
They must all be called either 'primary,' which should be the first fragment, or 'fragmentN' where 'N' is the ordinal number of the fragment, starting with '1'. The current logic allows for as many as 10 non-primary fragments. If more are needed, the code in [FileRestitcher.cs](pkg/FileRestitcher/FileRestitcher/FileRestitcher.cs) and [RestitchPackage.targets](pkg/common/RestitchPackage.targets) needstobeupdated. Notethatthesizeofeachfragmentisexpressedinbytes, andthatfragmentstartmustbe
230
230
thesumofthesizeofallpreviousfragments. A'-1'shouldbeusedfor the last fragment (and only for the lastfragment):itmeansthatthefragmentsizewillbebasedonhowmuchthereisstillleftofthefile.
Copy file name to clipboardExpand all lines: README.md
+3-2Lines changed: 3 additions & 2 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -11,9 +11,10 @@
11
11
<br/>
12
12
Please check the [Release Notes](RELEASENOTES.md) file for news on what's been updated in each new release.
13
13
14
-
__TorchSharp is now in the .NET Foundation!__
15
14
16
-
If you are using TorchSharp from NuGet, you should be using a version >= 0.98.3 of TorchSharp, and >= 1.12.0 of the libtorch-xxx redistributable packages. We recommend using one of the 'bundled' packages: TorchSharp-cpu, TorchSharp-cuda-windows, or TorchSharp-cuda-linux. They will pull in the right LibTorch backends.
15
+
__TorchSharp no longer supports MacOS on Intel hardware.__
16
+
17
+
With libtorch release 2.4.0, Intel HW support was deprecated for libtorch. This means that the last version of TorchSharp to work on Intel Macintosh hardware is 0.102.8. Starting with 0.103.0, only Macs based on Apple Silicon are supported.
0 commit comments