Data labeling
Turning to human annotators is often the most effective and efficient way to collect labeled training data. But annotating sample data can not only be tedious, it also introduces a degree of human error that can impact model accuracy. This data labeling app showcases how you can collect several, detailed labels for a given image while streamlining the workflow for your labeling team. Labels are saved in a Baseten Postgres database so it can be used in model training.
Use cases
Data Labeling
Human-in-the-Loop
Resources
PostgreSQL

Explore more

Speech to text
Use OpenAI Whisper for incredibly accurate speech-to-text in dozens of languages.

Image generation
Use Stable Diffusion to generate original images from a text prompt.

Deploy XGBoost model
Serve an XGBoost model behind a REST API endpoint.

Deploy Keras model
Serve a Keras model behind a REST API endpoint.