update code

deploy/auto_compression/configs/yolov5_reader.yml

+metric: COCO
+num_classes: 80
+
+# Datset configuration
+TrainDataset:
+  !COCODataSet
+    image_dir: train2017
+    anno_path: annotations/instances_train2017.json
+    dataset_dir: dataset/coco/
+
+EvalDataset:
+  !COCODataSet
+    image_dir: val2017
+    anno_path: annotations/instances_val2017.json
+    dataset_dir: dataset/coco/
+
+worker_num: 0
+
+# preprocess reader in test
+TestReader:
+  sample_transforms:
+    - Decode: {}
+    - Resize: {target_size: [640, 640], keep_ratio: True, interp: 1}
+    - Pad: {size: [640, 640], fill_value: [114., 114., 114.]}
+    - Permute: {}
+  batch_size: 1

deploy/auto_compression/configs/yolov5_s_qat_dis.yml

+
+Global:
+  reader_config: configs/yolov5_reader.yml
+  include_nms: True
+  Evaluation: True
+  model_dir: ./yolov5_s_300e_coco
+  model_filename: model.pdmodel
+  params_filename: model.pdiparams
+
+Distillation:
+  alpha: 1.0
+  loss: soft_label
+
+QuantAware:
+  use_pact: true
+  activation_quantize_type: 'moving_average_abs_max'
+  quantize_op_types:
+  - conv2d
+  - depthwise_conv2d
+
+TrainConfig:
+  train_iter: 3000
+  eval_iter: 1000
+  learning_rate: 0.00001
+  optimizer_builder:
+    optimizer:
+      type: SGD
+    weight_decay: 4.0e-05
+  target_metric: 0.365

deploy/auto_compression/configs/yolov6mt_s_qat_dis.yaml

+
+Global:
+  reader_config: configs/yolov5_reader.yml
+  include_nms: True
+  Evaluation: True
+  model_dir: ./yolov6mt_s_400e_coco
+  model_filename: model.pdmodel
+  params_filename: model.pdiparams
+
+Distillation:
+  alpha: 1.0
+  loss: soft_label
+
+QuantAware:
+  activation_quantize_type: 'moving_average_abs_max'
+  quantize_op_types:
+  - conv2d
+  - depthwise_conv2d
+
+TrainConfig:
+  train_iter: 8000
+  eval_iter: 1000
+  learning_rate: 
+    type: CosineAnnealingDecay 
+    learning_rate: 0.00003
+    T_max: 8000
+  optimizer_builder:
+    optimizer:
+      type: SGD
+    weight_decay: 0.00004

deploy/auto_compression/configs/yolov7_l_qat_dis.yaml

+
+Global:
+  reader_config: configs/yolov5_reader.yml
+  include_nms: True
+  Evaluation: True
+  model_dir: ./yolov7_l_300e_coco
+  model_filename: model.pdmodel
+  params_filename: model.pdiparams
+
+Distillation:
+  alpha: 1.0
+  loss: soft_label
+
+QuantAware:
+  activation_quantize_type: 'moving_average_abs_max'
+  quantize_op_types:
+  - conv2d
+  - depthwise_conv2d
+
+TrainConfig:
+  train_iter: 8000
+  eval_iter: 1000
+  learning_rate: 
+    type: CosineAnnealingDecay 
+    learning_rate: 0.00003
+    T_max: 8000
+  optimizer_builder:
+    optimizer:
+      type: SGD
+    weight_decay: 0.00004
\ No newline at end of file

deploy/auto_compression/configs/yolov8_reader.yml

+metric: COCO
+num_classes: 80
+
+# Dataset configuration
+TrainDataset:
+  !COCODataSet
+    image_dir: train2017
+    anno_path: annotations/instances_train2017.json
+    dataset_dir: dataset/coco/
+
+EvalDataset:
+  !COCODataSet
+    image_dir: val2017
+    anno_path: annotations/instances_val2017.json
+    dataset_dir: dataset/coco/
+
+worker_num: 0
+
+# preprocess reader in test
+EvalReader:
+  sample_transforms:
+    - Decode: {}
+    - Resize: {target_size: [640, 640], keep_ratio: True, interp: 1}
+    - Pad: {size: [640, 640], fill_value: [114., 114., 114.]}
+    - NormalizeImage: {mean: [0., 0., 0.], std: [1., 1., 1.], norm_type: none}
+    - Permute: {}
+  batch_size: 4

deploy/auto_compression/configs/yolov8_s_qat_dis.yaml

+
+Global:
+  reader_config: configs/yolov8_reader.yml
+  include_nms: True
+  Evaluation: True
+  model_dir: ./yolov8_s_500e_coco_trt_nms/
+  model_filename: model.pdmodel
+  params_filename: model.pdiparams
+
+Distillation:
+  alpha: 1.0
+  loss: soft_label
+
+QuantAware:
+  onnx_format: true
+  activation_quantize_type: 'moving_average_abs_max'
+  quantize_op_types:
+  - conv2d
+  - depthwise_conv2d
+
+TrainConfig:
+  train_iter: 8000
+  eval_iter: 1000
+  learning_rate:  
+    type: CosineAnnealingDecay
+    learning_rate: 0.00003
+    T_max: 10000
+  optimizer_builder:
+    optimizer: 
+      type: SGD
+    weight_decay: 4.0e-05
+

deploy/auto_compression/eval.py

+# Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
+#
+# 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
+import sys
+import numpy as np
+import argparse
+import paddle
+from ppdet.core.workspace import load_config, merge_config
+from ppdet.core.workspace import create
+from ppdet.metrics import COCOMetric, VOCMetric, KeyPointTopDownCOCOEval
+from paddleslim.auto_compression.config_helpers import load_config as load_slim_config
+from post_process import PPYOLOEPostProcess
+
+
+def argsparser():
+    parser = argparse.ArgumentParser(description=__doc__)
+    parser.add_argument(
+        '--config_path',
+        type=str,
+        default=None,
+        help="path of compression strategy config.",
+        required=True)
+    parser.add_argument(
+        '--devices',
+        type=str,
+        default='gpu',
+        help="which device used to compress.")
+
+    return parser
+
+
+def reader_wrapper(reader, input_list):
+    def gen():
+        for data in reader:
+            in_dict = {}
+            if isinstance(input_list, list):
+                for input_name in input_list:
+                    in_dict[input_name] = data[input_name]
+            elif isinstance(input_list, dict):
+                for input_name in input_list.keys():
+                    in_dict[input_list[input_name]] = data[input_name]
+            yield in_dict
+
+    return gen
+
+
+def convert_numpy_data(data, metric):
+    data_all = {}
+    data_all = {k: np.array(v) for k, v in data.items()}
+    if isinstance(metric, VOCMetric):
+        for k, v in data_all.items():
+            if not isinstance(v[0], np.ndarray):
+                tmp_list = []
+                for t in v:
+                    tmp_list.append(np.array(t))
+                data_all[k] = np.array(tmp_list)
+    else:
+        data_all = {k: np.array(v) for k, v in data.items()}
+    return data_all
+
+
+def eval():
+
+    place = paddle.CUDAPlace(0) if FLAGS.devices == 'gpu' else paddle.CPUPlace()
+    exe = paddle.static.Executor(place)
+
+    val_program, feed_target_names, fetch_targets = paddle.static.load_inference_model(
+        global_config["model_dir"].rstrip('/'),
+        exe,
+        model_filename=global_config["model_filename"],
+        params_filename=global_config["params_filename"])
+    print('Loaded model from: {}'.format(global_config["model_dir"]))
+
+    metric = global_config['metric']
+    for batch_id, data in enumerate(val_loader):
+        data_all = convert_numpy_data(data, metric)
+        data_input = {}
+        for k, v in data.items():
+            if isinstance(global_config['input_list'], list):
+                if k in global_config['input_list']:
+                    data_input[k] = np.array(v)
+            elif isinstance(global_config['input_list'], dict):
+                if k in global_config['input_list'].keys():
+                    data_input[global_config['input_list'][k]] = np.array(v)
+
+        outs = exe.run(val_program,
+                       feed=data_input,
+                       fetch_list=fetch_targets,
+                       return_numpy=False)
+        res = {}
+        if 'arch' in global_config and global_config['arch'] == 'PPYOLOE':
+            postprocess = PPYOLOEPostProcess(
+                score_threshold=0.01, nms_threshold=0.6)
+            res = postprocess(np.array(outs[0]), data_all['scale_factor'])
+        else:
+            for out in outs:
+                v = np.array(out)
+                if len(v.shape) > 1:
+                    res['bbox'] = v
+                else:
+                    res['bbox_num'] = v
+        metric.update(data_all, res)
+        if batch_id % 100 == 0:
+            print('Eval iter:', batch_id)
+    metric.accumulate()
+    metric.log()
+    metric.reset()
+
+
+def main():
+    global global_config
+    all_config = load_slim_config(FLAGS.config_path)
+    assert "Global" in all_config, "Key 'Global' not found in config file."
+    global_config = all_config["Global"]
+    reader_cfg = load_config(global_config['reader_config'])
+
+    dataset = reader_cfg['EvalDataset']
+    global val_loader
+    val_loader = create('EvalReader')(reader_cfg['EvalDataset'],
+                                      reader_cfg['worker_num'],
+                                      return_list=True)
+    metric = None
+    if reader_cfg['metric'] == 'COCO':
+        clsid2catid = {v: k for k, v in dataset.catid2clsid.items()}
+        anno_file = dataset.get_anno()
+        metric = COCOMetric(
+            anno_file=anno_file, clsid2catid=clsid2catid, IouType='bbox')
+    elif reader_cfg['metric'] == 'VOC':
+        metric = VOCMetric(
+            label_list=dataset.get_label_list(),
+            class_num=reader_cfg['num_classes'],
+            map_type=reader_cfg['map_type'])
+    elif reader_cfg['metric'] == 'KeyPointTopDownCOCOEval':
+        anno_file = dataset.get_anno()
+        metric = KeyPointTopDownCOCOEval(anno_file,
+                                         len(dataset), 17, 'output_eval')
+    else:
+        raise ValueError("metric currently only supports COCO and VOC.")
+    global_config['metric'] = metric
+
+    eval()
+
+
+if __name__ == '__main__':
+    paddle.enable_static()
+    parser = argsparser()
+    FLAGS = parser.parse_args()
+    assert FLAGS.devices in ['cpu', 'gpu', 'xpu', 'npu']
+    paddle.set_device(FLAGS.devices)
+
+    main()

deploy/auto_compression/post_process.py

+# Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
+#
+# 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 numpy as np
+import cv2
+
+
+def hard_nms(box_scores, iou_threshold, top_k=-1, candidate_size=200):
+    """
+    Args:
+        box_scores (N, 5): boxes in corner-form and probabilities.
+        iou_threshold: intersection over union threshold.
+        top_k: keep top_k results. If k <= 0, keep all the results.
+        candidate_size: only consider the candidates with the highest scores.
+    Returns:
+         picked: a list of indexes of the kept boxes
+    """
+    scores = box_scores[:, -1]
+    boxes = box_scores[:, :-1]
+    picked = []
+    indexes = np.argsort(scores)
+    indexes = indexes[-candidate_size:]
+    while len(indexes) > 0:
+        current = indexes[-1]
+        picked.append(current)
+        if 0 < top_k == len(picked) or len(indexes) == 1:
+            break
+        current_box = boxes[current, :]
+        indexes = indexes[:-1]
+        rest_boxes = boxes[indexes, :]
+        iou = iou_of(
+            rest_boxes,
+            np.expand_dims(
+                current_box, axis=0), )
+        indexes = indexes[iou <= iou_threshold]
+
+    return box_scores[picked, :]
+
+
+def iou_of(boxes0, boxes1, eps=1e-5):
+    """Return intersection-over-union (Jaccard index) of boxes.
+    Args:
+        boxes0 (N, 4): ground truth boxes.
+        boxes1 (N or 1, 4): predicted boxes.
+        eps: a small number to avoid 0 as denominator.
+    Returns:
+        iou (N): IoU values.
+    """
+    overlap_left_top = np.maximum(boxes0[..., :2], boxes1[..., :2])
+    overlap_right_bottom = np.minimum(boxes0[..., 2:], boxes1[..., 2:])
+
+    overlap_area = area_of(overlap_left_top, overlap_right_bottom)
+    area0 = area_of(boxes0[..., :2], boxes0[..., 2:])
+    area1 = area_of(boxes1[..., :2], boxes1[..., 2:])
+    return overlap_area / (area0 + area1 - overlap_area + eps)
+
+
+def area_of(left_top, right_bottom):
+    """Compute the areas of rectangles given two corners.
+    Args:
+        left_top (N, 2): left top corner.
+        right_bottom (N, 2): right bottom corner.
+    Returns:
+        area (N): return the area.
+    """
+    hw = np.clip(right_bottom - left_top, 0.0, None)
+    return hw[..., 0] * hw[..., 1]
+
+
+class PPYOLOEPostProcess(object):
+    """
+    Args:
+        input_shape (int): network input image size
+        scale_factor (float): scale factor of ori image
+    """
+
+    def __init__(self,
+                 score_threshold=0.4,
+                 nms_threshold=0.5,
+                 nms_top_k=10000,
+                 keep_top_k=300):
+        self.score_threshold = score_threshold
+        self.nms_threshold = nms_threshold
+        self.nms_top_k = nms_top_k
+        self.keep_top_k = keep_top_k
+
+    def _non_max_suppression(self, prediction, scale_factor):
+        batch_size = prediction.shape[0]
+        out_boxes_list = []
+        box_num_list = []
+        for batch_id in range(batch_size):
+            bboxes, confidences = prediction[batch_id][..., :4], prediction[
+                batch_id][..., 4:]
+            # nms
+            picked_box_probs = []
+            picked_labels = []
+            for class_index in range(0, confidences.shape[1]):
+                probs = confidences[:, class_index]
+                mask = probs > self.score_threshold
+                probs = probs[mask]
+                if probs.shape[0] == 0:
+                    continue
+                subset_boxes = bboxes[mask, :]
+                box_probs = np.concatenate(
+                    [subset_boxes, probs.reshape(-1, 1)], axis=1)
+                box_probs = hard_nms(
+                    box_probs,
+                    iou_threshold=self.nms_threshold,
+                    top_k=self.nms_top_k)
+                picked_box_probs.append(box_probs)
+                picked_labels.extend([class_index] * box_probs.shape[0])
+
+            if len(picked_box_probs) == 0:
+                out_boxes_list.append(np.empty((0, 4)))
+
+            else:
+                picked_box_probs = np.concatenate(picked_box_probs)
+                # resize output boxes
+                picked_box_probs[:, 0] /= scale_factor[batch_id][1]
+                picked_box_probs[:, 2] /= scale_factor[batch_id][1]
+                picked_box_probs[:, 1] /= scale_factor[batch_id][0]
+                picked_box_probs[:, 3] /= scale_factor[batch_id][0]
+
+                # clas score box
+                out_box = np.concatenate(
+                    [
+                        np.expand_dims(
+                            np.array(picked_labels), axis=-1), np.expand_dims(
+                                picked_box_probs[:, 4], axis=-1),
+                        picked_box_probs[:, :4]
+                    ],
+                    axis=1)
+                if out_box.shape[0] > self.keep_top_k:
+                    out_box = out_box[out_box[:, 1].argsort()[::-1]
+                                      [:self.keep_top_k]]
+                out_boxes_list.append(out_box)
+                box_num_list.append(out_box.shape[0])
+
+        out_boxes_list = np.concatenate(out_boxes_list, axis=0)
+        box_num_list = np.array(box_num_list)
+        return out_boxes_list, box_num_list
+
+    def __call__(self, outs, scale_factor):
+        out_boxes_list, box_num_list = self._non_max_suppression(outs,
+                                                                 scale_factor)
+        return {'bbox': out_boxes_list, 'bbox_num': box_num_list}

deploy/benchmark/benchmark.sh

+# All rights `PaddleDetection` reserved
+#!/bin/bash
+model_dir=$1
+model_name=$2
+
+export img_dir="demo"
+export log_path="output_pipeline"
+
+
+echo "model_dir : ${model_dir}"
+echo "img_dir: ${img_dir}"
+
+# TODO: support batch size>1
+for use_mkldnn in "True" "False"; do
+    for threads in "1" "6"; do
+            echo "${model_name}  ${model_dir}, use_mkldnn: ${use_mkldnn}   threads: ${threads}"
+            python deploy/python/infer.py \
+		 --model_dir=${model_dir} \
+		 --run_benchmark True \
+		 --enable_mkldnn=${use_mkldnn} \
+		 --device=CPU \
+		 --cpu_threads=${threads} \
+		 --image_dir=${img_dir}  2>&1 | tee ${log_path}/${model_name}_cpu_usemkldnn_${use_mkldnn}_cputhreads_${threads}_bs1_infer.log
+    done
+done
+
+for run_mode in "fluid" "trt_fp32" "trt_fp16"; do
+    echo "${model_name}  ${model_dir}, run_mode: ${run_mode}"
+    python deploy/python/infer.py \
+	--model_dir=${model_dir} \
+	--run_benchmark=True \
+	--device=GPU \
+	--run_mode=${run_mode} \
+	--image_dir=${img_dir}  2>&1 | tee ${log_path}/${model_name}_gpu_runmode_${run_mode}_bs1_infer.log
+done
+

deploy/benchmark/benchmark_quant.sh

+# All rights `PaddleDetection` reserved
+#!/bin/bash
+model_dir=$1
+model_name=$2
+
+export img_dir="demo"
+export log_path="output_pipeline"
+
+
+echo "model_dir : ${model_dir}"
+echo "img_dir: ${img_dir}"
+
+# TODO: support batch size>1
+for run_mode in "trt_int8"; do
+    echo "${model_name}  ${model_dir}, run_mode: ${run_mode}"
+    python deploy/python/infer.py \
+	--model_dir=${model_dir} \
+	--run_benchmark=True \
+	--device=GPU \
+	--run_mode=${run_mode} \
+	--image_dir=${img_dir}  2>&1 | tee ${log_path}/${model_name}_gpu_runmode_${run_mode}_bs1_infer.log
+done
+

deploy/cpp/cmake/yaml-cpp.cmake

+
+find_package(Git REQUIRED)
+
+include(ExternalProject)
+
+message("${CMAKE_BUILD_TYPE}")
+
+ExternalProject_Add(
+        ext-yaml-cpp
+        URL https://bj.bcebos.com/paddlex/deploy/deps/yaml-cpp.zip
+        URL_MD5 9542d6de397d1fbd649ed468cb5850e6
+        CMAKE_ARGS
+        -DYAML_CPP_BUILD_TESTS=OFF
+		-DYAML_CPP_BUILD_TOOLS=OFF
+        -DYAML_CPP_INSTALL=OFF
+        -DYAML_CPP_BUILD_CONTRIB=OFF
+		-DMSVC_SHARED_RT=OFF
+		-DBUILD_SHARED_LIBS=OFF
+        -DCMAKE_BUILD_TYPE=${CMAKE_BUILD_TYPE}
+        -DCMAKE_CXX_FLAGS=${CMAKE_CXX_FLAGS}
+        -DCMAKE_CXX_FLAGS_DEBUG=${CMAKE_CXX_FLAGS_DEBUG}
+        -DCMAKE_CXX_FLAGS_RELEASE=${CMAKE_CXX_FLAGS_RELEASE}
+        -DCMAKE_LIBRARY_OUTPUT_DIRECTORY=${CMAKE_BINARY_DIR}/ext/yaml-cpp/lib
+        -DCMAKE_ARCHIVE_OUTPUT_DIRECTORY=${CMAKE_BINARY_DIR}/ext/yaml-cpp/lib
+        PREFIX "${CMAKE_BINARY_DIR}/ext/yaml-cpp"
+        # Disable install step
+        INSTALL_COMMAND ""
+	    LOG_DOWNLOAD ON
+        LOG_BUILD 1
+)