qwen3-1.7b-gptq-int4 / recipe.yaml
2imi9's picture
Upload GPTQ INT4 quantized Qwen3 1.7B
4ff774f verified
default_stage:
default_modifiers:
GPTQModifier:
config_groups:
group_0:
targets: [Linear]
weights:
num_bits: 4
type: int
symmetric: true
group_size: 16
strategy: group
block_structure: null
dynamic: false
actorder: null
observer: minmax
observer_kwargs: {}
input_activations: null
output_activations: null
format: null
targets: [Linear]
ignore: [lm_head]
sequential_update: true
block_size: 128
dampening_frac: 0.01
offload_hessians: false