July 2022 Summaries

Activeloop Hub 2.7.1 introduces enhanced querying capabilities, allowing users to query large datasets in seconds using SQL-like queries. Users can now save query results or subsets of their data and optimize them for training. Additionally, the update allows Hub datasets to store references to original data using hub.link, making it easier to materialize data for training. Collaboration capabilities have also been improved, with up to 10 collaborators per organization on the Community plan. The Growth plan has been made more accessible at $495 per month. Additional features include deleting samples from datasets and accessing more information about dataset samples.

The recent release of Hub 2.7.1 by Activeloop introduces enhanced features aimed at improving data querying and collaboration capabilities. This update includes a highly performant query engine, allowing SQL-like queries on large datasets in seconds, and the ability to save and optimize query results for training. Users can now store references to original data using hub.link, enhancing data management without needing to convert data to Hub format. The collaboration aspect has been upgraded to allow up to 10 collaborators per organization under the Community plan, fostering teamwork in smaller machine learning teams. Additionally, the Growth plan has been made more affordable at $495 per month, and new functionalities such as deleting data samples and accessing data information as dictionaries have been added. These updates are complemented by performance optimizations and bug fixes, ensuring a more seamless experience for users.