3 ms·Did anyone compare Alpaca responses between LLaMA fp16 vs 4-bit quantized?by code51 4y agoDid anyone compare Alpaca responses between LLaMA fp16 vs 4-bit quantized?