| quant_stage: | |
| quant_modifiers: | |
| GPTQModifier: | |
| ignore: [lm_head, 're:.*gate$'] | |
| sequential_targets: [Qwen3MoeDecoderLayer] | |
| dampening_frac: 0.01 | |
| config_groups: | |
| group0: | |
| targets: [Linear] | |
| weights: {num_bits: 4, type: int, strategy: group, group_size: 128, symmetric: true, | |
| actorder: weight, observer: minmax} | |