ResearchBenchmarksEnd-to-end dialog modeling on CamRest676Follow98.5Match ScoreESAinsTOD92.46894.03495.697.166Mar 10, 2026Evaluation ResultsMethodMethodLinksMatch ScoreSuccess F1 ScoreBLEU ScoreCombined ScoreESAinsTODBackbone=Llama 2 7BBackbone=Llama 2 7B2026.0398.588.4526.92120.39SPACEFine-tuned individuall...Fine-tuned individually=true2026.0397.7488.2423.68116.67LABES2026.0396.48325.5115.2FSDM2026.0393.586.225.8115.65SEDST2026.0392.775.423.6107.65TSCP2026.0392.785.425.3114.35