| { |
| "config_groups": { |
| "group_mxfp8_mtp_routed_experts": { |
| "input_activations": { |
| "dynamic": true, |
| "group_size": 32, |
| "num_bits": 8, |
| "type": "float" |
| }, |
| "targets": [ |
| "model.language_model.layers.45.mlp.experts" |
| ], |
| "weights": { |
| "dynamic": false, |
| "group_size": 32, |
| "num_bits": 8, |
| "type": "float", |
| "weight_block_size": [ |
| 1, |
| 32 |
| ] |
| } |
| }, |
| "group_nvfp4_routed_experts": { |
| "input_activations": { |
| "dynamic": false, |
| "group_size": 16, |
| "num_bits": 4, |
| "type": "float" |
| }, |
| "targets": [ |
| "model.language_model.layers.3.mlp.experts", |
| "model.language_model.layers.4.mlp.experts", |
| "model.language_model.layers.5.mlp.experts", |
| "model.language_model.layers.6.mlp.experts", |
| "model.language_model.layers.7.mlp.experts", |
| "model.language_model.layers.8.mlp.experts", |
| "model.language_model.layers.9.mlp.experts", |
| "model.language_model.layers.10.mlp.experts", |
| "model.language_model.layers.11.mlp.experts", |
| "model.language_model.layers.12.mlp.experts", |
| "model.language_model.layers.13.mlp.experts", |
| "model.language_model.layers.14.mlp.experts", |
| "model.language_model.layers.15.mlp.experts", |
| "model.language_model.layers.16.mlp.experts", |
| "model.language_model.layers.17.mlp.experts", |
| "model.language_model.layers.18.mlp.experts", |
| "model.language_model.layers.19.mlp.experts", |
| "model.language_model.layers.20.mlp.experts", |
| "model.language_model.layers.21.mlp.experts", |
| "model.language_model.layers.22.mlp.experts", |
| "model.language_model.layers.23.mlp.experts", |
| "model.language_model.layers.24.mlp.experts", |
| "model.language_model.layers.25.mlp.experts", |
| "model.language_model.layers.26.mlp.experts", |
| "model.language_model.layers.27.mlp.experts", |
| "model.language_model.layers.28.mlp.experts", |
| "model.language_model.layers.29.mlp.experts", |
| "model.language_model.layers.30.mlp.experts", |
| "model.language_model.layers.31.mlp.experts", |
| "model.language_model.layers.32.mlp.experts", |
| "model.language_model.layers.33.mlp.experts", |
| "model.language_model.layers.34.mlp.experts", |
| "model.language_model.layers.35.mlp.experts", |
| "model.language_model.layers.36.mlp.experts", |
| "model.language_model.layers.37.mlp.experts", |
| "model.language_model.layers.38.mlp.experts", |
| "model.language_model.layers.39.mlp.experts", |
| "model.language_model.layers.40.mlp.experts", |
| "model.language_model.layers.41.mlp.experts", |
| "model.language_model.layers.42.mlp.experts", |
| "model.language_model.layers.43.mlp.experts", |
| "model.language_model.layers.44.mlp.experts" |
| ], |
| "weights": { |
| "dynamic": false, |
| "group_size": 16, |
| "num_bits": 4, |
| "type": "float" |
| } |
| } |
| }, |
| "ignore": [], |
| "producer": { |
| "name": "modelopt", |
| "version": "0.39.0.dev290+gf9d9a71de.d20260407" |
| }, |
| "quant_algo": "MIXED_PRECISION", |
| "quant_method": "modelopt", |
| "quantized_layers": { |
| "model.language_model.layers.10.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.11.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.12.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.13.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.14.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.15.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.16.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.17.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.18.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.19.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.20.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.21.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.22.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.23.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.24.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.25.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.26.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.27.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.28.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.29.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.3.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.30.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.31.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.32.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.33.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.34.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.35.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.36.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.37.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.38.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.39.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.4.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.40.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.41.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.42.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.43.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.44.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.45.mlp.experts": { |
| "group_size": 32, |
| "quant_algo": "MXFP8" |
| }, |
| "model.language_model.layers.5.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.6.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.7.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.8.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| }, |
| "model.language_model.layers.9.mlp.experts": { |
| "group_size": 16, |
| "quant_algo": "NVFP4" |
| } |
| } |
| } |
|
|