Skip to content

Navigation Menu

Sign in
Appearance settings

Search code, repositories, users, issues, pull requests...

Provide feedback

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly

Appearance settings
Discussion options

🚀 What's New?

  • Chronos-2 can now be deployed on AWS via SageMaker JumpStart. Please check this notebook for details.
  • Scaled Dot Product Attention (SDPA) is now used as the default attention implementation in Chronos-2. If you need to use the previous eager implementation, please load the model with Chronos2Pipeline.from_pretrained(..., attn_implementation="eager").
  • predict_df support has been added for older Chronos and Chronos-Bolt models. Now, all models (Chronos-2, Chronos-Bolt, Chronos) provide a unified pandas dataframe API. Note: Only Chronos-2 supports multivariate and covariate-informed forecasting.
  • Chronos2Pipeline.embed has been added, enabling users to extract embeddings from the last layer of the Chronos-2 encoder.

🐛 Bug Fixes

  • Fixed issues related to past-only covariates use during Chronos-2 fine-tuning. If you're fine-tuning Chronos-2 models, we strongly recommend upgrading to chronos-forecasting==2.1.0.
  • Fixed issue related to multiple workers on windows.

All Changes

New Contributors

Full Changelog: v2.0.1...v2.1.0


This discussion was created from the release 2.1.0.
You must be logged in to vote

Replies: 0 comments

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
1 participant
Morty Proxy This is a proxified and sanitized view of the page, visit original site.