File size: 191 Bytes
151ded2
 
 
 
 
 
 
1
2
3
4
5
6
7
8
default_stage:
  default_modifiers:
    QuantizationModifier:
      targets: [Linear]
      ignore: [lm_head, 're:.*mlp.gate[.].*']
      scheme: W8A8
      bypass_divisibility_checks: false