ResearchDatasetsMeta-LlamaMeta-LlamaFollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsLLM InferenceMeta-Llama-3-8B prompt/gen = 128 tokens5,463Memory Usage (MB)12Model provenance detectionMeta-Llama-3-8B Quantization83.3TPR5Model provenance detectionMeta-Llama-3-8B Post-training94.4TPR5Language Model Inference EfficiencyMeta-Llama-3-8B1,398Throughput (tokens/s)4