Enable LWQ for onnxrt WOQ in 3.x API (#1625)

Signed-off-by: yuwenzho <yuwen.zhou@intel.com>

Author: yuwenzho

neural_compressor/onnxrt/algorithms/__init__.py

@@ -17,5 +17,6 @@
 from neural_compressor.onnxrt.algorithms.weight_only.rtn import apply_rtn_on_model
 from neural_compressor.onnxrt.algorithms.weight_only.gptq import apply_gptq_on_model
 from neural_compressor.onnxrt.algorithms.weight_only.awq import apply_awq_on_model
+from neural_compressor.onnxrt.algorithms.layer_wise import layer_wise_quant
 
-__all__ = ["Smoother", "apply_rtn_on_model", "apply_gptq_on_model", "apply_awq_on_model"]
+__all__ = ["Smoother", "apply_rtn_on_model", "apply_gptq_on_model", "apply_awq_on_model", "layer_wise_quant"]

neural_compressor/onnxrt/algorithms/layer_wise/__init__.py

@@ -0,0 +1,17 @@
+# Copyright (c) 2024 Intel Corporation
+#
+# 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.
+
+from neural_compressor.onnxrt.algorithms.layer_wise.core import layer_wise_quant
+
+__all__ = ["layer_wise_quant"]

neural_compressor/onnxrt/algorithms/layer_wise/core.py

@@ -0,0 +1,297 @@
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+#
+# Copyright (c) 2023 MIT HAN Lab
+# This source code is licensed under the MIT license
+#
+# Copyright (c) 2023 Intel Corporation
+#
+# 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.
+
+import os
+from copy import deepcopy
+from pathlib import Path
+from typing import Callable, List, Union
+
+import onnx
+import onnxruntime as ort
+import transformers
+from packaging.version import Version
+
+from neural_compressor.common import Logger
+from neural_compressor.onnxrt.quantization.calibrate import CalibrationDataReader
+from neural_compressor.onnxrt.utils.onnx_model import ONNXModel
+from neural_compressor.onnxrt.utils.utility import check_model_with_infer_shapes
+
+logger = Logger().get_logger()
+
+__all__ = [
+    "layer_wise_quant",
+]
+
+
+def layer_wise_quant(
+    model: Union[onnx.ModelProto, ONNXModel, Path, str],
+    quant_func: Callable,
+    weight_config: dict,
+    data_reader: CalibrationDataReader = None,
+    *args,
+    **kwargs
+) -> ONNXModel:
+    """Quantize model layer by layer to save memory.
+
+    Args:
+        model (Union[onnx.ModelProto, ONNXModel, Path, str]): onnx model.
+        quant_func (Callable): quantization algo function.
+        weight_config (dict): quantization config.
+        data_reader (CalibrationDataReader, optional): data_reader for calibration. Defaults to None.
+
+    Returns:
+        _type_: _description_
+    """
+    # TODO: remove the limitation for lwq
+    if Version(transformers.__version__) > Version("4.37.2"):
+        logger.warning(
+            "Model (such as llama-2) exported with transformers {} may fail in layer-wise quant. "
+            "we recommend downgrading transformers to 4.37.2 and try again.".format(transformers.__version__)
+        )
+
+    # check whether model shape is inferred
+    if not check_model_with_infer_shapes(model):
+        logger.error(
+            "Before applying layer-wise quantization, please make sure to "
+            "run symbolic shape inference on your model like follows:\n"
+            "import onnxruntime.tools.symbolic_shape_infer as symbolic_shape_infer\n"
+            "model = onnx.load(your_model_path)\n"
+            "out = symbolic_shape_infer.SymbolicShapeInference.infer_shapes(model, auto_merge=True)\n"
+            "onnx.save(out, infer_shape_model_path)\n"
+        )
+        raise ValueError("Fail to run layer-wise quantization.")
+
+    if not isinstance(model, ONNXModel):
+        model = ONNXModel(model, ignore_warning=True, load_external_data=False)
+
+    origin_model = deepcopy(model)
+
+    providers = kwargs.get("providers", ["CPUExecutionProvider"])
+
+    # get and check split nodes
+    split_nodes = origin_model.find_split_nodes()
+    if len(split_nodes) == 0:
+        logger.error(
+            "Can't find split nodes for layer-wise quantization. "
+            "We recommend applying graph optimization for your model like follows: \n"
+            "import onnxruntime as ort \n"
+            "sess_options = ort.SessionOptions() \n"
+            "sess_options.graph_optimization_level = ort.GraphOptimizationLevel.ORT_ENABLE_EXTENDED "
+            "# or ORT_ENABLE_BASIC \n"
+            "sess_options.optimized_model_filepath = 'optimized_model_path' \n"
+            "ort.InferenceSession(infer_shape_model_path, sess_options)"
+        )
+        raise ValueError("Fail to run layer-wise quantization.")
+    logger.info(
+        "Will split model into {} parts to do layer-wise quantization".format(
+            len([node.name for node in split_nodes]) + 1
+        )
+    )
+    logger.debug(
+        "Will split model with these nodes for layer-wise quantization: {}".format([node.name for node in split_nodes])
+    )
+
+    split_idx = 1
+    model_to_split = [origin_model]
+    quantized_model_merged = None
+
+    require_data_reader = data_reader is not None
+    if require_data_reader:
+        lwq_data_reader = [data_reader]
+
+    while len(model_to_split) != 0:
+        # prepare model, node and data_reader for current split
+        split_model = model_to_split.pop(0)
+        split_node = split_nodes.pop(0)
+        if require_data_reader:
+            current_data_reader = lwq_data_reader.pop(0)
+
+        # if no remaining split nodes, it means this is the last split, and the two split models will be saved.
+        save_both_split_models = True if len(split_nodes) == 0 else False
+
+        # split model with given split node
+        split_model_part_1, split_model_part_2 = split_model.split_model_with_node(
+            split_node.name, model.model_path, save_both_split_models
+        )
+        if not save_both_split_models:
+            # append split_model_part_2 to do next split
+            model_to_split.append(split_model_part_2)
+
+        logger.info("Quantize split model {}".format(split_idx))
+        if require_data_reader:
+            # process data_reader for current split and next split
+            current_data_reader = _filter_data_reader_for_current_split_model(
+                split_model_part_1.model, current_data_reader
+            )
+            next_data_reader = _prepare_data_reader_for_next_split_model(
+                split_model_part_1.model_path, current_data_reader, providers
+            )
+            lwq_data_reader.append(next_data_reader)
+
+            # perform quantization
+            split_model_part_1_quantized = quant_func(
+                split_model_part_1,
+                weight_config=weight_config,
+                data_reader=current_data_reader,
+                return_modelproto=False,
+                **kwargs
+            )
+        else:
+            # perform quantization
+            split_model_part_1_quantized = quant_func(
+                split_model_part_1, weight_config=weight_config, return_modelproto=False, **kwargs
+            )
+
+        # check split model is valid
+        try:
+            ort.InferenceSession(split_model_part_1_quantized.model.SerializeToString(), providers=providers)
+        except Exception as e:
+            logger.error(
+                "Layer-wise quantized model {} can't be inferred correctly. "
+                "Please check the raise exception".format(split_idx)
+            )
+            raise e
+
+        # merge split quantized model
+        if quantized_model_merged is None:
+            quantized_model_merged = split_model_part_1_quantized
+            quantized_model_merged.write_external_data_to_new_location(overwrite=True)
+        else:
+            quantized_model_merged.merge_split_models(split_model_part_1_quantized)
+
+        split_idx += 1
+        # if this is the last split, quantize the last split model
+        if save_both_split_models:
+            logger.info("Quantize split model {}".format(split_idx))
+
+            # quantize split model
+            if require_data_reader:
+                # process data_reader for current split
+                current_data_reader = lwq_data_reader.pop(0)
+                current_data_reader = _filter_data_reader_for_current_split_model(
+                    split_model_part_2.model, current_data_reader
+                )
+
+                # perform quantization
+                split_model_part_2_quantized = quant_func(
+                    split_model_part_2,
+                    weight_config=weight_config,
+                    data_reader=current_data_reader,
+                    return_modelproto=False,
+                    **kwargs
+                )
+            else:
+                # perform quantization
+                split_model_part_2_quantized = quant_func(
+                    split_model_part_2, weight_config=weight_config, return_modelproto=False, **kwargs
+                )
+
+            # check split model is valid
+            try:
+                ort.InferenceSession(split_model_part_2_quantized.model.SerializeToString(), providers=providers)
+            except Exception as e:
+                logger.error(
+                    "Layer-wise quantized model {} can't be inferred correctly. "
+                    "Please check the raise exception".format(split_idx)
+                )
+                raise e
+
+            # merge split quantized model
+            if quantized_model_merged is None:
+                quantized_model_merged = split_model_part_2_quantized
+                quantized_model_merged.write_external_data_to_new_location(overwrite=True)
+            else:
+                quantized_model_merged.merge_split_models(split_model_part_2_quantized)
+
+    # reload external data to prevent external data file path errors
+    from onnx.external_data_helper import load_external_data_for_model
+
+    load_external_data_for_model(quantized_model_merged.model, os.path.dirname(quantized_model_merged.model_path))
+
+    return quantized_model_merged
+
+
+class DataReader(CalibrationDataReader):
+    """Data reader for layer-wise quantization."""
+
+    def __init__(self, data_list):
+        self.data_list = data_list
+        self.iter_next = iter(self.data_list)
+
+    def get_next(self):
+        return next(self.iter_next, None)
+
+    def rewind(self):
+        self.iter_next = iter(self.data_list)
+
+
+def _filter_data_reader_for_current_split_model(model: onnx.ModelProto, data_reader: CalibrationDataReader):
+    """Filter data reader to remove data that is not in model input.
+
+    Args:
+        model (onnx.ModelProto): onnx model.
+        data_reader (CalibrationDataReader): data reader.
+
+    Returns:
+        CalibrationDataReader: filtered data reader.
+    """
+    filter_inputs = []
+    input_names = [input.name for input in model.graph.input]
+    while True:
+        inputs = data_reader.get_next()
+        if not inputs:
+            break
+        filter_input = {
+            input_name: input_tensor for input_name, input_tensor in inputs.items() if input_name in input_names
+        }
+        filter_inputs.append(filter_input)
+    return DataReader(filter_inputs)
+
+
+def _prepare_data_reader_for_next_split_model(
+    model_path: str,
+    data_reader: CalibrationDataReader,
+    providers: List[str] = ["CPUExecutionProvider"],
+):
+    """Prepare data reader for next split model.
+
+    Get data output of current split model and save for next split model.
+
+    Args:
+        model (str): path to onnx model.
+        data_reader (CalibrationDataReader): data reader
+        providers (List[str], optional): providers to use. Defaults to ["CPUExecutionProvider"].
+
+    Returns:
+        CalibrationDataReader: data reader for next split model.
+    """
+    data_reader = deepcopy(data_reader)
+
+    data_reader_for_next_split_model = []
+    session = ort.InferenceSession(model_path, providers=providers)
+    output_names = [output.name for output in session.get_outputs()]
+    while True:
+        inputs = data_reader.get_next()
+        if not inputs:
+            break
+        out = session.run(None, inputs)
+        inputs.update({name: value for name, value in zip(output_names, out)})
+        data_reader_for_next_split_model.append(inputs)
+    return DataReader(data_reader_for_next_split_model)

neural_compressor/onnxrt/algorithms/weight_only/awq.py

@@ -275,7 +275,7 @@ def _apply_awq_clip(model, weight_config, absorb_pairs, output_dicts, num_bits,
 
 def awq_quantize(
     model: Union[onnx.ModelProto, ONNXModel, Path, str],
-    dataloader: CalibrationDataReader,
+    data_reader: CalibrationDataReader,
     weight_config: dict = {},
     num_bits: int = 4,
     group_size: int = 32,
@@ -289,7 +289,7 @@ def awq_quantize(
 
     Args:
         model (Union[onnx.ModelProto, ONNXModel, Path, str]): onnx model.
-        dataloader (CalibrationDataReader): dataloader for calibration.
+        data_reader (CalibrationDataReader): data_reader for calibration.
         weight_config (dict, optional): quantization config
             For example,
             weight_config = {
@@ -323,8 +323,8 @@ def awq_quantize(
     full_ratio = {}
 
     if enable_mse_search:
-        inputs, so = prepare_inputs(model, dataloader, providers)
-        del dataloader
+        inputs, so = prepare_inputs(model, data_reader, providers)
+        del data_reader
 
         org_output = copy.deepcopy(model.model.graph.output)
         model.remove_tensors_from_outputs([i.name for i in org_output])
@@ -420,7 +420,7 @@ def apply_awq_on_model(
     Args:
         model (Union[onnx.ModelProto, ONNXModel, Path, str]): nnx model.
         quant_config (dict): quantization config.
-        calibration_data_reader (CalibrationDataReader): dataloader for calibration.
+        calibration_data_reader (CalibrationDataReader): data_reader for calibration.
 
     Returns:
         onnx.ModelProto: quantized onnx model.
@@ -434,4 +434,4 @@ def apply_awq_on_model(
         if isinstance(op_config, AWQConfig):
             quant_config[op_name_type] = op_config.to_dict()
 
-    return awq_quantize(model, dataloader=calibration_data_reader, weight_config=quant_config, **kwargs)
+    return awq_quantize(model, data_reader=calibration_data_reader, weight_config=quant_config, **kwargs)