recipe.yaml

default_stage:
  default_modifiers:
    GPTQModifier:
      targets: [Linear]
      ignore: [lm_head, 're:.*mlp.gate$', 're:.*mlp.shared_expert_gate$']
      scheme: W4A16
      sequential_update: true
      block_size: 128
      dampening_frac: 0.01
      offload_hessians: false