transformer/quantization_config.json

{
  "bits": 4,
  "data_type": "int",
  "group_size": 128,
  "sym": true,
  "batch_size": 1,
  "iters": 100,
  "lr": 0.005,
  "low_gpu_mem_usage": true,
  "minmax_lr": 0.005,
  "nsamples": 32,
  "autoround_version": "0.13.0",
  "block_name_to_quantize": "blocks",
  "quant_method": "auto-round",
  "packing_format": "auto_round:auto_gptq"
}