Today we’re making a big leap forward in enabling new applications and insights for every enterprise that uses visual data. We’re releasing the next version of Masterful, the platform that automates model development for computer vision. Masterful now includes a low-code interface that enables developers to build models faster than ever before. Using a simple command-line interface (CLI), any developer can build production-grade models without needing to write code in TensorFlow or PyTorch. Our low-code interface also opens up CV development to an even wider universe of developers!
Masterful already accelerates model development in two powerful ways: reducing the need for labeled data, and automatically tuning the training loop for maximum model accuracy and minimal training time. Now, with the CLI, developers just edit a simple config file, run one command to start training, and get back a production-grade model ready for deployment.
Masterful handles data pre-processing and model architecture setup automatically. First, just put your data into folders and Masterful does the rest to prepare it for training. Next, pick from a range of state-of-the-art model architecture in our model zoo. The final trained model can be saved in Open Neural Network Exchange (ONNX) format, which is compatible for use with other frameworks such as with PyTorch, or in TensorFlow format.
Power of the full platform
With the low-code interface, you'll still get the full suite of Masterful's powerful automation features:
- Semi-supervised learning to train with unlabeled data, which reduces labeling budgets while improving model accuracy.
- Comprehensive regularization (including data augmentation) that dynamically adjusts to each dataset, maximizing model robustness without any additional developer effort.
- Automatically setting hyperparameters like batch size and learning rate, saving developer time and reducing GPU bills.
- Visualization front-end, for full transparency into what Masterful is learning about your data and how it's maximizing the accuracy of your model.
The Masterful CLI Trainer is implemented entirely on top of the existing Masterful API. Developers who want finer-grained control or want to use specific Masterful features in their training pipeline can use the API directly.
We're excited about helping an even wider range of developers 10x their CV development and build better models faster than ever before!
Masterful is available for free for personal, academic, and trial use. Just run 'pip install masterful' in your Python terminal to get the latest release, then head over to the CLI Quickstart to start training your first model in minutes. Also join our Slack community - we're there to answer your questions and would love to hear your feedback and ideas on this latest release!