Meta-Llama-3.1-8B-Instruct-FP8 Benchmark | LLM Performance Benchmarks