recipe.yaml

quant_stage:
  quant_modifiers:
    GPTQModifier:
      ignore: ['re:.*lm_head.*', 're:.*embed_tokens.*', 're:vision_tower.*', 're:multi_modal_projector.*']
      sequential_targets: [Gemma3DecoderLayer]
      dampening_frac: 0.05
      config_groups:
        group0:
          targets: [Linear]
          weights: {num_bits: 4, type: int, strategy: group, group_size: 128, symmetric: true,
            actorder: weight, observer: mse}