ResearchDatasetsvLLM serving benchmarkvLLM serving benchmarkFollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsLarge Language Model ServingvLLM serving benchmark 128 prompts, 32 pre-fill tokens, 256 generation tokens76TTFT (ms)9
Large Language Model ServingvLLM serving benchmark 128 prompts, 32 pre-fill tokens, 256 generation tokens76TTFT (ms)9