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
Modified the project to support running of TensorFlow on GPU on Windows. (dotnet#4270)
* Modified the project to support running of TensorFlow on GPU on Windows.
Removed all dependencies of TensorFlow redist from the source projects,
and instead added the dependency to the Sample project.
Created separate sample project for GPU examples since gpu tensorflow requires cuda,
which may not be available on all machines, so it needs to be a separate
project.
Added documentation for setup as there is now some requirements.
* Added NuGet dependency to the test packages which need TensorFlow.
* Changed to use the Dependency variable for the TF version.
* removed erronious newlines.
Install [CUDA v10.0](https://developer.nvidia.com/cuda-10.0-download-archive) and [CUDNN v7.6.4](https://developer.nvidia.com/rdp/cudnn-download)
24
+
following [Nvidia's Install guide](https://docs.nvidia.com/cuda/cuda-quick-start-guide/index.html).
25
+
26
+
#### Usage
27
+
To use TensorFlow with GPU support take a NuGet dependency on the following package depending on your OS:
28
+
29
+
Windows -> SciSharp.TensorFlow.Redist-Windows-GPU
30
+
31
+
No code modification should be necessary to leverage the GPU for TensorFlow operations.
32
+
33
+
#### Troubleshooting
34
+
If you are not able to use your GPU after adding the GPU based TensorFlow NuGet,
35
+
make sure that there is only a dependency on the GPU based version. If you have
36
+
a dependency on both NuGets, the CPU based TensorFlow will run instead.
Copy file name to clipboardExpand all lines: src/Microsoft.ML.TensorFlow/TensorflowCatalog.cs
+6Lines changed: 6 additions & 0 deletions
Original file line number
Diff line number
Diff line change
@@ -14,6 +14,12 @@ public static class TensorflowCatalog
14
14
/// <summary>
15
15
/// Load TensorFlow model into memory. This is the convenience method that allows the model to be loaded once and subsequently use it for querying schema and creation of
16
16
/// <see cref="TensorFlowEstimator"/> using <see cref="TensorFlowModel.ScoreTensorFlowModel(string, string, bool)"/>.
17
+
/// usage of this API requires additional NuGet dependencies on TensorFlow redist, see linked document for more information.
0 commit comments