DeepLearning_Frameworks_Benchmark

Benchmark code and guidelines for article "Deep Learning with Theano, Torch, Caffe, TensorFlow, and Deeplearning4J: Which One Is the Best in Speed and Accuracy?"

This work was mainlyfunded by Altoros, a global provider of big data, cloud computing, and Platform-as-a-Service solutions. You can find more materials at Altoros research papers

DeepLearning Frameworks Benchmark (Main contents)

• 1. Installation instrutions on g2.2xlarge
• 2. Basic test: Logistic Regression
  • 2.1 Logistic Regression with Theano
    • 2.1.1 Theano: Blocks Framework
    • 2.1.2 Theano: Lasagne Framework
    • 2.1.3 Theano: Keras Framework
  • 2.2 Logistic Regression with Torch
  • 2.3 Logistic Regression with Caffe
  • 2.4 Logisitc Regression with Tensorflow
  • 2.5 Logistic Regression with Deeplearning4J
• 3. FCNN: Fully Connected Neural Network (on CPU)
  • 3.1 FCNN with Theano (Keras framework)
  • 3.2 FCNN with Torch
  • 3.3 FCNN with Caffe
  • 3.3 FCNN with TensorFlow
  • 3.4 FCNN with DeepLearning4J
  • 3.5 Conclusion
• 4. Distibuted TensorFlow: first steps
  • 4.1 Deep Learning Model for expriment with distributed TF
  • 4.2 Configuration of the virtual environment
  • 4.3 Preparing dataset
  • 4.4 Configure and Run distributed TF in VBox Env
  • 4.5 Check TF scalability on multi-GPU one-Node