Face announced today that its inference, batch processing, and cloud storage services—Inference Endpoints, Jobs, and Buckets—have been seamlessly integrated into the Papers with Code platform. This combination enables researchers and developers to quickly run preprocessed models, automatically store results, and conduct more precise searches across papers and source code.
With Inference Endpoints, users can expose their models as APIs and process inference requests in real time without the need for infrastructure setup. The Jobs service also runs heavy and batch processing tasks on a scheduled basis, while Buckets provide scalable storage space for results and intermediate data. These infrastructures have improved search response speed, increased the accuracy of matching code with papers, and reduced waiting times for Papers with Code users.