ResearchDatasetsDeepSeek-R1-Distill-LlamaDeepSeek-R1-Distill-LlamaFollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsMulti-step LLM InteractionDeepSeek-R1-Distill-Llama-8B Inference Performance (test)100High-Precision Ratio7End-to-end single-step decodingDeepSeek-R1-Distill-LLaMA-8B 32K Context23.4Latency (ms)4
Multi-step LLM InteractionDeepSeek-R1-Distill-Llama-8B Inference Performance (test)100High-Precision Ratio7