Caffe & Caffe2 on AWS

No 1 About

Caffe and Caffe2

Caffe, built by UC Berkeley, is an open source machine learning framework. It supports many different types of deep learning architectures geared towards image classification and image segmentation such as CNN, RCNN, LSTM and fully connected neural network designs. Caffe also supports GPU-based acceleration using CuDNN of Nvidia.

Built by NVidia and Facebook, Caffe2 builds upon Caffe and comes with native Python and C++ APIs that work interchangeably so you can prototype quickly now, and easily optimize later. Caffe2 fine tuned from the ground up to take full advantage of the latest NVIDIA Deep Learning SDK libraries, cuDNN , cuBLAS and NCCL , to deliver high-performance, multi-GPU acceleration for desktop, data centers, and embedded edge devices

Caffe & Caffe2 Facts:

  • Caffe was developed by UC Berkeley and formally released in 2017
  • Caffe2 was built as a partnership between Facebook and NVidia, it was also released in 2017
  • In addition to the neural networks supported in Caffe, Caffe2 also supports RNNs
  • Both Caffe and Caffe2 support Python and C++ APIs
Learn more about Caffe by click here

Learn more about Caffe2 by click here

No 2 Demo

Demo 1: Image file analysis

Built by UC Berkley, this demo showcases that Caffe is optimized for image-based AI/ML tasks. Upload a picture and see how quickly Caffe generates image tag for either a photo you upload or an example photo.