GIFT-Eval
Summary
GIFT-Eval is a general-purpose time-series forecasting benchmark and leaderboard used by many time-series foundation-model papers, including Toto and Toto 2.0.
Official Artifacts
- Official Hugging Face: https://huggingface.co/datasets/Salesforce/GiftEval
- Official leaderboard: https://huggingface.co/spaces/Salesforce/GIFT-Eval
- Official code: https://github.com/SalesforceAIResearch/gift-eval
- Dataset metadata snapshot: gift-eval-2024
Role In The Wiki
GIFT-Eval should be used as a broad forecasting benchmark reference, not as direct evidence that a model solves observability-specific, HDTSF, or action-conditioned world-model tasks.
Relation To Foundation TSFM Agenda
Use the source-level agenda mapping in gift-eval-2024 rather than duplicating verdict rows here.
At the entity level, GIFT-Eval should be used as a broad forecasting benchmark reference, not as direct evidence that a model solves observability-specific, HDTSF, or action-conditioned world-model tasks. This page should stay as the object card; source pages carry slot-level verdicts, evidence, and missing pieces.