ResearchDatasetsGeneral LLM EvaluationGeneral LLM EvaluationFollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsSpeculative DecodingGeneral LLM Evaluation (test)32Max Accepted Token Length8Text Generation AccelerationGeneral LLM Evaluation2.5Max Speedup Factor6