- Efficient tensor/matrix computation across multiple devices, including multiple CPUs, GPUs and distributed server nodes.
- Flexible symbolic manipulation to composite and construct state-of-the-art deep learning models.
For more details, see documentation below. Please also checkout the examples directory.
- Digit Recognition on MNIST
- Generating Random Sentence with LSTM RNN
- Installation Guide
- NDArray API
- Symbolic API
- Evaluation Metrics
- Data Providers
- Neural Network Factory
- Network Visualization