TensorFlow Edge Documentation - Edge API

TensorFlow Edge provides a set of APIs and tools to run TensorFlow models on edge devices. The edge API allows you to run TensorFlow models on devices with limited computational resources.

Features

• Model Execution: Run TensorFlow models on edge devices.
• Model Conversion: Convert TensorFlow models to TensorFlow Lite format for edge deployment.
• Resource Management: Efficiently manage device resources like CPU, GPU, and memory.

Quick Start

To get started with TensorFlow Edge, follow these steps:

1. Download and Install TensorFlow Edge: Download TensorFlow Edge
2. Prepare Your Model: Convert your TensorFlow model to TensorFlow Lite format using the TensorFlow Lite Converter.
3. Deploy Your Model: Use the TensorFlow Edge runtime to deploy your model on an edge device.

Examples

Here are some examples of using the edge API:

• Load a Model: Load a TensorFlow Lite model on an edge device.

  import tensorflow as tf
  
  model = tf.lite.Interpreter(model_content=...")
  model.allocate_tensors()
  

• Run Inference: Run inference on the loaded model.

  input_details = model.get_input_details()
  output_details = model.get_output_details()
  
  input_data = ...
  model.set_tensor(input_details[0]['index'], input_data)
  model.invoke()
  output_data = model.get_tensor(output_details[0]['index'])
  

Learn More

For more detailed information, refer to the TensorFlow Edge documentation.

TensorFlow Edge Architecture