NumPy is one of the most used library of Python in Data Science field. TensorFlow NumPy implements a subset of the full NumPy spec. While more symbols will be added over time, there are systematic features that will not be supported in the near future.

NOTE: If you don’t know about NumPy library you should visit this article “ NumPy: Everything A Data Scientist Should Know”.

NumPy API on TensorFlow :

All the benefits of TensorFlow Ecosystem for NumPy:

  • Accelerate Numpy code on CPU/TPU/GPU.
  • Auto-Differentiate through NumPy code.
  • Optimize execution using compilation and auto-vectorization.
  • Run distributed using tf.distribute .
  • Combine seamlessly with TensorFlow APIs (tf.linalg , tf.signal , tf.data , tf.keras ).
  • SavedModel Serialization.

Contents

  • Getting started
  • Performance Comparison
  • TensorFlow NumPy support and flexibility
  • Interoperability with NumPy and TensorFlow
  • Adding new NumPy operations
  • Case studies

#machine-learning #tensorflow #deep-learning #python

NumPy using TensorFlow : A Faster Way To Do NumPy Operations
31.15 GEEK