femtoflow 0.1.7

Femtosense Model Optimization Toolkit

FemtoFlow

femtoflow is a Python package that enables pruning and quantization of TensorFlow models for optimized deployment to Femtosense's SPU (Sparse Processing Unit). The package provides a straightforward and flexible interface for reducing model size and complexity while maintaining high-performance inference on the SPU.

Features

• Pruning: Reduce model size by removing unimportant connections and neurons from your TensorFlow model.
• Quantization: Lower the precision of weights and activations to reduce memory requirements and computational costs while preserving model accuracy.

Getting Started

You can install femtoflow using pip:

pip install femtoflow

Documentation

Detailed documentation for femtoflow, including tutorials, API reference, and examples, can be found on the official website: https://femtoflow.femtosense.ai