What is VESSL Run?
VESSL Run is a unified YAML interface for training, fine-tuning, and scaling machine learning models. Run simplifies the complex compute backends and system details required to run open-source models into an intuitive YAML definition.
- Unified YAML interface: for training and scaling any models.
- Powerful VESSL CLI/SDK: Train the latest open-source models in seconds with
- Intuitive Web Console: Create a run on the web and keep track of the results on the dashboard.
- Complete Model Reproducibility: Run models without worrying about ML-specific peripherals like GPUs, CUDA configurations, and Python dependencies.
How it works
- Each YAML file defines a run spec for a model.
- In essence, with every
vessl run, you are launching a Kubernetes pod configured for ML.
- Our custom Docker Images provide run-proof, dockerized versions of popular models on GitHub.