recipe.yaml

quant_stage:
  quant_modifiers:
    GPTQModifier:
      config_groups:
        group_0:
          targets: [Linear]
          weights:
            num_bits: 4
            type: int
            symmetric: true
            group_size: 128
            strategy: group
            block_structure: null
            dynamic: false
            actorder: !!python/object/apply:compressed_tensors.quantization.quant_args.ActivationOrdering [
              static]
            observer: minmax
            observer_kwargs: {}
          input_activations: null
          output_activations: null
          format: null
      targets: [Linear]
      ignore: [lm_head, '*block_sparse_moe.gate', 're:.*mlp.gate$', 're:.*mlp.shared_expert_gate$']
      sequential_targets: ['re:.*layers\.\d+$']
      block_size: 64
      dampening_frac: 0.001
      actorder: static
      offload_hessians: false