HAissa
/

M3_smoothquant098

Text Generation

text-generation-inference

8-bit precision

compressed-tensors

Model card Files Files and versions

M3_smoothquant098

File size: 171 Bytes

99e8a5c

default_stage:
  default_modifiers:
    SmoothQuantModifier: {smoothing_strength: 0.98}
    GPTQModifier:
      scheme: W4A8
      targets: Linear
      ignore: [lm_head]