File size: 253 Bytes
f2ff453 | 1 2 3 4 5 6 7 8 9 10 11 | default_stage:
default_modifiers:
GPTQModifier:
targets: [Linear]
ignore: [lm_head, mlp.gate, shared_expert_gate]
scheme: W4A16
block_size: 128
dampening_frac: 0.1
actorder: static
offload_hessians: false
|