SUITE OF DEVELOPMENT AND TRAINING TOOLS
GTI offers the tools to build and deploy Artificial Intelligence solutions on edge and cloud deployments.
SOFTWARE DEVELOPMENT KIT
GTI’s software development kit (SDK) provides a hardware-accelerated, Convolutional Neural Network (CNN) system and supporting software library – Implementing state-of-the-art algorithms on Lightspeeur® AI accelerator chips. The SDK includes:
- Drivers and prebuilt libraries for (ARM) Android & (ARM/x86) Linux environments.
- API for easy software integration.
- Pretrained CNN models.
- Sample use case source code.
MODEL DEVELOPMENT KITS
GTI’s model development kits (MDK) enable customers to build and train neural network models, that are compatible with GTI devices. Both Caffe and TensorFlow are fully supported frameworks.
These kits are used to port neural networks onto GTI devices with comparable performance to the original floating-point model. Train models from scratch or fine-tune from a pre-trained floating-point model. Working examples of VGG16, ResNet18 and MobileNet v1 are provided.
Plai™ Builder is a PyTorch based “full-stack” framework specifically designed to quickly train and build neural network models for GTI devices. Users have the option of a GUI interface for simple model creation, or command line access that offers more flexibility and custom options. Either way, users can easily run Plai™ Builder on a host PC, train and build quantized models optimized for GTI’s accelerator, and run inference.
Plai™ Plug (People Learning Artificial Intelligence) is designed for developers to create AI solutions on their Windows or Linux computer. Just plug in the accelerator to easily create apps and solutions that are 100% AI-enabled. Lightspeeur® 2801 and 2803 are now available.
Application source code is available to customers.