InfrastructureHugging FaceOct 21How Specializing the Vocabulary of SLMs Can Cut Inference Cost by 77% in Real ApplicationsPrimary sourcecaught 12 h 25 min after publication
InfrastructureHugging FaceOct 9Impactful scheduling for GPU clustersImpactful scheduling for GPU clusters.Primary sourcecaught 14 h 40 min after publication