4 February 2025, Vienna – Austrian synthetic data startup MOSTLY AI announces the release of the world’s first industry-grade open source toolkit for producing synthetic data from real customer data.

MOSTLY AI announces $100,000 global prize challenge to advance privacy-safe synthetic data for AI training. Entries close 3 ...

Databricks also announced updates to its Unity Catalog that expand support for Apache Iceberg — the open-source table format ...

In theory, synthetic data could create a near-infinite supply ... but fine-tunes existing open source models to add differential privacy and safety features, then packages those together to ...

Chinese social media giant’s cost-efficient AI breakthrough highlights Asia’s growing influence in enterprise AI development.

B-Preview, an open source AI coding model based on Deepseek-R1-Distilled-Qwen-14B. The model achieves a 60.6% pass rate on ...

From public and proprietary datasets to synthetic and behavioral ... academic institutions, and open-source projects. It includes census data, environmental reports, public financial filings ...

The open-source model was trained on TerraMesh ... TiM, as it’s known, involves self-generating additional synthetic training data from other modalities to enhance TerraMind’s performance ...

RedNote, also known as Xiaohongshu, has open-sourced its first large language model, “dots.llm1” amid global AI expansion.

For example, in Meta's flagship open-source model, Llama 3.1 405B, which the company introduced last week, the researchers made extensive use of synthetic data to "fine-tune" the model and to ...