ResearchTasksConstrained LLM DecodingFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedQwen2-14B INT8Pre³16.52Inference Time (ms)14Feb 26, 2026Llama-3-8BPre³11.77Inference Time (ms)14Feb 26, 2026Llama-2-70BPre³27.2Latency (ms)10Feb 26, 2026DeepSeek-V2-Lite-Chat 15.7BPre³49.91Inference Time (ms)10Feb 26, 2026