On the Ray Dashboard, we’ve added new metrics for Ray Data in the Metrics tab.This release also features new Ray Data datasources for Databricks and BigQuery. This allows users to randomly shuffle input files for better model training accuracy. We’ve added support for PyTorch-compatible input files shuffling for Ray Data.In RLlib, we’ve moved 24 algorithms into rllib_contrib (still available within RLlib for Ray 2.8). More API changes in the release notes below. We’ve also added a new Java APIs that aligns with the Ray Serve 2.x APIs. The previously deprecated Ray Serve 1.x APIs have also been removed. Model composition will be supported through deployment handles providing more flexibility and stability. In Ray Serve, we are deprecating the previously experimental DAG API for deployment graphs.This release features stability improvements and API clean-ups across the Ray libraries.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |