An open-source machine learning framework developed by Google for building and deploying numerical computation and deep learning models across CPUs, GPUs and specialised accelerators, using a dataflow graph paradigm with automatic differentiation and a high-level Keras API.
Semantic Classification
Content
- TensorFlow represents computations as dataflow graphs of operations over multidimensional arrays called tensors. It provides automatic differentiation, distributed training and tools for serving models in production through TensorFlow Serving and TensorFlow Lite for mobile and embedded targets.
- The framework includes the high-level Keras API for model construction and supports deployment across servers, browsers and edge devices. It was among the first widely adopted industrial deep learning frameworks and remains in broad use, particularly in production engineering settings.