Sinapsis Framework Converter

Templates for conversion between deep learning frameworks.

🐍 Installation • 🚀 Features 📙 Documentation • 🔍 License

The sinapsis-framework-converter module allows for the conversion between some of the most popular deep learning frameworks in the community:

• Keras -> Tensorflow
• Tensorflow -> ONNX
• Pytorch -> TensorRT
• Pytorch -> ONNX
• ONNX -> TensorRT

🐍 Installation

[!NOTE] CUDA-based templates in Sinapsis-framework-converter require NVIDIA driver version to be 550 or higher.

Install using your package manager of choice. We encourage the use of uv

Example with uv:

uv pip install sinapsis-framework-converter --extra-index-url https://pypi.sinapsis.tech

or with raw pip:

pip install sinapsis-framework-converter --extra-index-url https://pypi.sinapsis.tech

[!IMPORTANT] Templates in each package may require extra dependencies. For development, we recommend installing the package with all the optional dependencies:

Example with uv:

uv pip install sinapsis-framework-converter[all] --extra-index-url https://pypi.sinapsis.tech

or with raw pip:

pip install sinapsis-framework-converter[all] --extra-index-url https://pypi.sinapsis.tech

[!IMPORTANT] To enable tensorflow with cuda support please install tensorflow as follows:

uv pip install tensorflow[and-cuda]==2.18.0

or

pip install tensorflow[and-cuda]==2.18.0

🚀 Features

Templates Supported

The Sinapsis Framework Converter module provides multiple templates for deep learning framework conversion.

• KerasTensorFlowConverter: Converts Keras models to TensorFlow.
• ONNXTRTConverter: Converts ONNX models to TensorRT.
• TensorFlowONNXConverter: Converts TensorFlow models to ONNX.
• TorchONNXConverter: Converts PyTorch models to ONNX.
• TorchTRTConverter: Converts PyTorch models to TensorRT.

agent:
  name: conversion_agent

templates:
- template_name: InputTemplate
  class_name: InputTemplate
  attributes: {}

- template_name: TorchONNXConverter
  class_name: TorchONNXConverter
  template_input: InputTemplate
  attributes:
    model_name: resnet50
    save_model_path: true
    force_compilation: true
    opset_version: 12
    height: 224
    width: 224

📙 Documentation

Documentation is available on the sinapsis website

Tutorials for different projects within sinapsis are available at sinapsis tutorials page

🔍 License

This project is licensed under the AGPLv3 license, which encourages open collaboration and sharing. For more details, please refer to the LICENSE file.

For commercial use, please refer to our official Sinapsis website for information on obtaining a commercial license.