You can store your data or connect stream resources and build your data assets into authorized repos. Powered by industry standard big data platforms, you can scale up very easily. Through this unified repo, you can store, authorize and classify data assets for multiple users and projects centrally and ensure proper backup and high availability strategies.
Put your assets into projects and prepare your data for training process, which includes cleaning and sanitizing your data from noise and anomalies. You can put filters on your data to remove unwanted fragments temparing with your data model. Add as much tags as you want and get your data ready to feed into your machine learning model training process.
You can build you own data training pipelines on popular platforms such as Tensorflow, Torch, Theano, MXNet and Caffee. You can use same assets for different projects trained on different platforms. Lion Gate Arena utilizes your GPU farms and stores and versions your outputs so that you can use them in your models.