Filter dataset benchmark leaderboards by the number of model parameters. On any dataset leaderboard, pick a size range and the rankings refresh to that bucket. The top 3 models in each size category are marked with a 🏅.
Fetched June 1, 2026
A new Base only toggle on the Models page hides every finetune, adapter, merge, and quantization, leaving just the original base models. Th…
Hugging Face · Hugging Face Changelog
Trainer telemetry is now gated on an explicit class-name allowlist, restricting which trainer classes can send telemetry.
Hugging Face · Fine-tuningEvaluation models are now accessible to individual non-enterprise GitHub Copilot users and may be selected by Copilot auto model selection.…
GitHub · GitHub ChangelogEnterprise owners can now create targeted model rules to allow specific Copilot models for selected organizations instead of relying on ent…
GitHub · GitHub Changelog