{ "producer": { "name": "modelopt", "version": "0.0.0" }, "quantization": { "quant_algo": "MIXED_PRECISION", "kv_cache_quant_algo": "FP8", "quantized_layers": { "model.layers.0.mlp.gate_proj": { "quant_algo": "FP8" }, "model.layers.0.mlp.up_proj": { "quant_algo": "FP8" }, "model.layers.0.mlp.down_proj": { "quant_algo": "FP8" }, "model.layers.1.mlp.gate_proj": { "quant_algo": "FP8" }, "model.layers.1.mlp.up_proj": { "quant_algo": "FP8" }, "model.layers.1.mlp.down_proj": { "quant_algo": "FP8" }, "model.layers.2.mlp.gate_proj": { "quant_algo": "FP8" }, "model.layers.2.mlp.up_proj": { "quant_algo": "FP8" }, "model.layers.2.mlp.down_proj": { "quant_algo": "FP8" }, "model.layers.3.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.4.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.5.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.6.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.7.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.8.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.9.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.10.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.11.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.12.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.13.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.14.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.15.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.16.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.17.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.18.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.19.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.20.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.21.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.22.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.23.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.24.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.25.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.26.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.27.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.28.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.29.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.30.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.31.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.32.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.33.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.34.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.35.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.36.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.37.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.38.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.39.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.40.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.41.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.42.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.43.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.44.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.45.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.46.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.47.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.48.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.49.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.50.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.51.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.52.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.53.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.54.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.55.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.56.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.57.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.58.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.59.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.60.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.61.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.62.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.63.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.64.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.65.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.66.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.67.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.68.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.69.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.70.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.71.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.72.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.73.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.74.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.75.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.76.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.77.mlp.experts": { "quant_algo": "NVFP4", "group_size": 16 }, "model.layers.0.mlp.gate_up_proj": { "quant_algo": "FP8" }, "model.layers.1.mlp.gate_up_proj": { "quant_algo": "FP8" }, "model.layers.2.mlp.gate_up_proj": { "quant_algo": "FP8" } }, "exclude_modules": [ "model.layers.78*" ], "packed_modules_mapping": { "gate_up_proj": [ "gate_proj", "up_proj" ] } } }