Move W4A16_NVFP4_CFG to YAML

main added W4A16_NVFP4_CFG (weight-only NVFP4 for all linears) after the
YAML-conversion branch was forked, so it landed in choices but had no
YAML preset and no load_config call after rebase. Add the YAML preset
and the load_config call to match the rest of the constants.

Signed-off-by: Shengliang Xu <shengliangx@nvidia.com>

Author: shengliangxu

modelopt/torch/quantization/config.py

@@ -1385,6 +1385,10 @@ def _load_quantizer_cfg_dict_list(config_path: str) -> list[dict[str, Any]]:
     "configs/ptq/presets/model/w4a8_nvfp4_fp8",
     schema_type=QuantizeConfig,
 )
+W4A16_NVFP4_CFG: QuantizeConfig = load_config(
+    "configs/ptq/presets/model/w4a16_nvfp4",
+    schema_type=QuantizeConfig,
+)
 MXFP4_MLP_WEIGHT_ONLY_CFG: QuantizeConfig = load_config(
     "configs/ptq/presets/model/mxfp4_mlp_weight_only",
     schema_type=QuantizeConfig,

modelopt_recipes/configs/ptq/presets/model/w4a16_nvfp4.yaml

@@ -0,0 +1,30 @@
+# SPDX-FileCopyrightText: Copyright (c) 2024 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-License-Identifier: Apache-2.0
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+# QuantizeConfig preset for NVFP4 weight-only quantization on all layers (W4A16).
+
+# modelopt-schema: modelopt.torch.quantization.config.QuantizeConfig
+imports:
+  base_disable_all: configs/ptq/units/base_disable_all
+  default_disabled_quantizers: configs/ptq/units/default_disabled_quantizers
+  nvfp4: configs/numerics/nvfp4
+
+algorithm: max
+quant_cfg:
+  - $import: base_disable_all
+  - quantizer_name: '*weight_quantizer'
+    cfg:
+      $import: nvfp4
+  - $import: default_disabled_quantizers