Impressive results from Meta's Llama adapting to various benchmarks. However, gaming performance seems lackluster compared to specialized models like Alpaca. It raises questions about the viability of large language models for complex, interactive tasks like gaming without more targeted fine-tuning. Exciting progress nonetheless!