llama3: 8b-instruct Benchmark | LLM Performance Benchmarks