Error loading "meta-llama/Meta-Llama-3.1-8B-Instruct" that was converted to gguf using llama.cpp convert_hf_to_gguf.py using q_8 quantization