ResearchBenchmarksOpen-domain Question Answering on TriviaQA (Mixed LLM Suite Metrics)Follow77LAMA ScoreREMIX71.38472.84274.375.758Nov 11, 2024Evaluation ResultsMethodMethodLinksLAMA ScoreEntityQA ScoreWebQA ScoreFactoid Avg AccuracyGSM8K AccuracyMATH AccuracyEvolCode AccuracyAPPS AccuracyUltraChat ScoreNon-Factoid Avg AccuracyREMIXModel=Mistral-7B-v0.3,...Model=Mistral-7B-v0.3, Mixing Strategy=Random / K-Pile2024.117781.583.180.51.691.195.397.790.775.3No MixingModel=Mistral-7B-v0.3,...Model=Mistral-7B-v0.3, Mixing Strategy=None2024.1171.686.491.583.24.89995.979.99775.3