TensorFlow is an open source software library for high performance numerical computation. Its flexible architecture allows easy deployment of computation across a variety of platforms (CPUs, GPUs, TPUs), and from desktops to clusters of servers to mobile and edge devices.
TensorFlow was originally developed by researchers and engineers from the Google Brain team within Google’s AI organization.
|1||v2.7 for Python 2||v3.3+ for Python 3|
|2||NVIDIA CUDA 7.5||NVIDIA CUDA 8.0 required for Pascal GPUs|
|3||Hardware: NVIDIA® GPU card with CUDA® Compute Capability 3.5 or higher|
|4||Software: NVIDIA® GPU drivers —CUDA 9.0 requires 384.x or higher, CUDA® Toolkit —TensorFlow supports CUDA 9.0, CUPTI ships with the CUDA Toolkit, cuDNN SDK (>= 7.2)||Minimum Software plus: NCCL 2.2 for multiple GPU support, TensorRT 4.0 to improve latency and throughput for inference on some models|
Apache Mahout Apache MXNet (Incubating) Apache Spark Apache SystemML Eclipse Deeplearning4j MALLET Massive Online Analysis (MOA) mlpack OpenCV Orange PyTorch scikit-learn TensorFlow The Microsoft Cognitive Toolkit Torch Weka Yooreeka
Accord.NET AForge.NET Eclipse Deeplearning4j OpenCog TensorFlow The Microsoft Cognitive Toolkit
Apache MXNet (Incubating) Apache SystemML Caffe Eclipse Deeplearning4j OpenNN PyTorch TensorFlow The Microsoft Cognitive Toolkit Torch Weka