engblogs

Building better GenAI through programmatic data development enhances performance of large language models for generative AI tasks by using weak supervision and automated curation techniques.

Using weak supervision to programmatically curate instruction tuning data for open-source LLMs to build a better GenAI.

Using weak supervision, we programmatically curated instruction tuning data for open-source LLMs to build a better GenAI.

Databricks has completed the acquisition of MosaicML, a platform for generative AI model creation and customization.

Integrating Amazon SageMaker Model Cards with the model registry for standardized model documentation and visibility into the model lifecycle.

Using a generative AI foundation model to harness your own dataset for text summarization and question answering

Introducing OpenHouse, an open source data lakehouse deployment for managing big data tables at LinkedIn.

Using weak supervision to programmatically curate instruction tuning data for open-source LLMs to improve GenAI.

Tracking and monitoring important metrics related to issues, pull requests, and discussions using the new Issue Metrics GitHub Action.

Discussing the use of control characters for prompt injection in ChatGPT and the importance of not overlooking natural language processing (NLP) techniques.

Learn how to create and customize Avatars for your Duolingo profile.

Exploring the impact of collaboration with the American Journalism Project on bolstering local news