๐ฆ Goldener feature: Semantics aware sampling for better models
Goldener provides smart data sampling out of the box by combining 2 different GoldDoers (classes orchestrating data actions): 1๏ธโฃ GoldDescriptor: Unlock data semantics access via embeddings computed from foundation models. 2๏ธโฃ GoldSelector: Select samples automatically by digging into data semantics with coreset algorithms
Both the foundation model and coreset algorithm are fully customizable to achieve the selection goals from a few lines of Python code. The result? Goldener can replace the usual random selection and help release better models, faster!
๐ Introducing Goldener: The Python Data Orchestrator for more efficient ML
Machine Learning workflows often rely on randomness: selecting/splitting data for training, batching it variably, and monitoring real-world performance.
Nowadays, foundation models give access to the semantics of data. Goldener leverages this semantic to make the entire ML lifecycle more efficient!