recipe.yaml

quant_stage:
  quant_modifiers:
    GPTQModifier:
      sequential_update: false
      dampening_frac: 0.01
      block_size: 128
      ignore: [lm_head]
      config_groups:
        group_0:
          targets: [Linear]
          weights: {num_bits: 4, type: int, symmetric: true, strategy: group, group_size: 128}