llama3 70B-instruct-q2_K Benchmark | LLM Performance Benchmarks