Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/shaltielshmid/torchsharp.pybridge
A library enabling easy transfer and handling of PyTorch models between .NET and Python environments
https://github.com/shaltielshmid/torchsharp.pybridge
libtorch pybridge pytorch torchsharp
Last synced: 1 day ago
JSON representation
A library enabling easy transfer and handling of PyTorch models between .NET and Python environments
- Host: GitHub
- URL: https://github.com/shaltielshmid/torchsharp.pybridge
- Owner: shaltielshmid
- License: mit
- Created: 2023-11-19T14:30:13.000Z (about 1 year ago)
- Default Branch: main
- Last Pushed: 2024-12-02T17:07:13.000Z (about 1 month ago)
- Last Synced: 2024-12-02T18:23:06.273Z (about 1 month ago)
- Topics: libtorch, pybridge, pytorch, torchsharp
- Language: C#
- Homepage:
- Size: 214 KB
- Stars: 17
- Watchers: 2
- Forks: 4
- Open Issues: 2
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# TorchSharp.PyBridge
[](https://www.nuget.org/packages/TorchSharp.PyBridge/)
TorchSharp.PyBridge is an extension library for [TorchSharp](https://github.com/dotnet/TorchSharp), providing seamless interoperability between .NET and Python for model serialization. It simplifies the process of saving and loading PyTorch models in a .NET environment, enabling developers to easily develop models in both .NET and Python and transfer models easily.
## Features
- `module.load_py(...)`, `optim.load_py(...)`: Extension method for modules and optimizers for easily loading PyTorch models saved in the standard Python format (using `torch.save`) directly into TorchSharp.
> This only works for when the `state_dict` was saved and not the whole model, see example below.
- `module.save_py(...)`, `optim.save_py(...)`: Extension method for modules and optimizers for easily saving TorchSharp models in a format that can be directly loaded in PyTorch (using `torch.load`), offering cross-platform model compatibility.
- `module.load_safetensors(...)`, `module.save_safetensors(...)`: Extension methods for modules for easily saving and loading model weights using the [safetensors](https://github.com/huggingface/safetensors) format.
- `module.load_checkpoint(...)`: Extension method for loading in a checkpoint (both safetensors and regular pytorch, including sharded models) from a directory saved using HuggingFace's `PreTrainedModel.save_pretrained()` method.
## Getting Started
### Installation
TorchSharp.PyBridge is available on NuGet. You can install it using the following command:
#### .NET CLI
```bash
dotnet add package TorchSharp.PyBridge
```
#### NuGet Package Manager
```powershell
Install-Package TorchSharp.PyBridge
```
### Prerequisites
- .NET SDK
- TorchSharp library
## Usage
### Loading a PyTorch Model in .NET
Saving the model in Python:
```python
import torch
model = ...
torch.save(model.state_dict(), 'path_to_your_model.pth')
```
Loading it in C#:
```csharp
using TorchSharp.PyBridge;
var model = ...;
model.load_py("path_to_your_model.pth");
```
### Saving a TorchSharp Model for PyTorch
To save a model in a format compatible with PyTorch:
```csharp
using TorchSharp.PyBridge;
var model = ...;
model.save_py("path_to_save_model.pth");
```
And loading it in in Python:
```python
import torch
model = ...
model.load_state_dict(torch.load('path_to_save_model.pth'))
```
## Contributing
Contributions to TorchSharp.PyBridge are welcome.
## Acknowledgments
This project makes use of the `pickle` library, a Java and .NET implementation of Python's pickle serialization protocol, developed by Irmen de Jong. The `pickle` library plays a vital role in enabling the serialization features within TorchSharp.PyBridge. We extend our thanks to the developer for their significant contributions to the open-source community. For more details about the `pickle` library, please visit their [GitHub repository](https://github.com/irmen/pickle).
## Support and Contact
For support, questions, or feedback, please open an issue in the [GitHub repository](https://github.com/shaltielshmid/TorchSharp.PyBridge).