Hugging Face announced that model pages will now include evaluation results from the Every Eval Ever (EEE) community project. The change brings a wide range of benchmark outcomes directly onto each model's page, rather than requiring users to search for results elsewhere.
According to the announcement, this integration is meant to make model comparison more straightforward. By surfacing community-contributed eval data in a standard location, Hugging Face aims to reduce friction for researchers and practitioners who need to assess performance across many models.
Since this is a single source, there are no conflicting accounts to note. The post frames the feature as a step toward more transparent and accessible evaluation on the platform.