Llama-3-8B-Instruct-Coder-AWQ/quant_config.json

{
  "zero_point": true,
  "q_group_size": 128,
  "w_bit": 4,
  "version": "GEMM"
}