| default_stage: | |
| default_modifiers: | |
| AWQModifier: | |
| config_groups: | |
| group_0: | |
| targets: [Linear] | |
| weights: | |
| num_bits: 4 | |
| type: int | |
| symmetric: true | |
| group_size: 32 | |
| strategy: group | |
| block_structure: null | |
| dynamic: false | |
| actorder: null | |
| observer: minmax | |
| observer_kwargs: {} | |
| input_activations: null | |
| output_activations: null | |
| format: null | |
| targets: [Linear] | |
| ignore: [model.embed_tokens, 're:.*block_sparse_moe[.]e_score_correction_bias$', 're:.*block_sparse_moe[.]gate$', | |
| 're:.*input_layernorm$', 're:.*post_attention_layernorm$', 're:.*k_norm$', 're:.*q_norm$', | |
| model.norm, lm_head] | |
| mappings: | |
| - smooth_layer: re:.*input_layernorm$ | |
| balance_layers: ['re:.*q_proj$', 're:.*k_proj$', 're:.*v_proj$'] | |
| - smooth_layer: re:.*v_proj$ | |
| balance_layers: ['re:.*o_proj$'] | |
| - smooth_layer: re:.*post_attention_layernorm$ | |
| balance_layers: ['re:.*w1$', 're:.*w3$'] | |
| - smooth_layer: re:.*w3$ | |
| balance_layers: ['re:.*w2$'] | |
| offload_device: !!python/object/apply:torch.device [cpu] | |
| duo_scaling: true | |