diff --git a/deploy/serving/python/config.yml b/deploy/serving/python/config.yml
new file mode 100644
index 0000000000000000000000000000000000000000..5ec4285257d618f6c5a7ed02aab5c34dae9a96e1
--- /dev/null
+++ b/deploy/serving/python/config.yml
@@ -0,0 +1,31 @@
+#worker_num, 最大并发数。当build_dag_each_worker=True时, 框架会创建worker_num个进程，每个进程内构建grpcSever和DAG
+##当build_dag_each_worker=False时，框架会设置主线程grpc线程池的max_workers=worker_num
+worker_num: 20
+
+#http端口, rpc_port和http_port不允许同时为空。当rpc_port可用且http_port为空时，不自动生成http_port
+http_port: 18093
+rpc_port: 9993
+
+dag:
+    #op资源类型, True, 为线程模型；False，为进程模型
+    is_thread_op: False
+op:
+    #op名称，与web_service中的TIPCExampleService初始化name参数一致
+    ppdet:
+        #并发数，is_thread_op=True时，为线程并发；否则为进程并发
+        concurrency: 1
+
+        #当op配置没有server_endpoints时，从local_service_conf读取本地服务配置
+        local_service_conf:
+
+            #uci模型路径
+            model_config: "./serving_server"
+
+            #计算硬件类型: 空缺时由devices决定(CPU/GPU)，0=cpu, 1=gpu, 2=tensorRT, 3=arm cpu, 4=kunlun xpu
+            device_type:
+
+            #计算硬件ID，当devices为""或不写时为CPU预测；当devices为"0", "0,1,2"时为GPU预测，表示使用的GPU卡
+            devices: "0" # "0,1"
+
+            #client类型，包括brpc, grpc和local_predictor.local_predictor不启动Serving服务，进程内预测
+            client_type: local_predictor
diff --git a/deploy/serving/python/pipeline_http_client.py b/deploy/serving/python/pipeline_http_client.py
new file mode 100644
index 0000000000000000000000000000000000000000..33eb70dcf3b4dc1e9f56aea399d9c40e20091078
--- /dev/null
+++ b/deploy/serving/python/pipeline_http_client.py
@@ -0,0 +1,86 @@
+# 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 glob
+import requests
+import json
+import base64
+import os
+import argparse
+
+parser = argparse.ArgumentParser(description="args for paddleserving")
+parser.add_argument("--image_dir", type=str)
+parser.add_argument("--image_file", type=str)
+args = parser.parse_args()
+
+
+def get_test_images(infer_dir, infer_img):
+    """
+    Get image path list in TEST mode
+    """
+    assert infer_img is not None or infer_dir is not None, \
+        "--image_file or --image_dir should be set"
+    assert infer_img is None or os.path.isfile(infer_img), \
+            "{} is not a file".format(infer_img)
+    assert infer_dir is None or os.path.isdir(infer_dir), \
+            "{} is not a directory".format(infer_dir)
+
+    # infer_img has a higher priority
+    if infer_img and os.path.isfile(infer_img):
+        return [infer_img]
+
+    images = set()
+    infer_dir = os.path.abspath(infer_dir)
+    assert os.path.isdir(infer_dir), \
+        "infer_dir {} is not a directory".format(infer_dir)
+    exts = ['jpg', 'jpeg', 'png', 'bmp']
+    exts += [ext.upper() for ext in exts]
+    for ext in exts:
+        images.update(glob.glob('{}/*.{}'.format(infer_dir, ext)))
+    images = list(images)
+
+    assert len(images) > 0, "no image found in {}".format(infer_dir)
+    print("Found {} inference images in total.".format(len(images)))
+
+    return images
+
+
+def cv2_to_base64(image):
+    """cv2_to_base64
+
+    Convert an numpy array to a base64 object.
+
+    Args:
+        image: Input array.
+
+    Returns: Base64 output of the input.
+    """
+    return base64.b64encode(image).decode('utf8')
+
+
+if __name__ == "__main__":
+    url = "http://127.0.0.1:18093/ppdet/prediction"
+    logid = 10000
+    img_list = get_test_images(args.image_dir, args.image_file)
+
+    for img_file in img_list:
+        with open(img_file, 'rb') as file:
+            image_data1 = file.read()
+
+        image = cv2_to_base64(image_data1)
+
+        data = {"key": ["image_0"], "value": [image], "logid": logid}
+        # send requests
+        r = requests.post(url=url, data=json.dumps(data))
+        print(r.json())
diff --git a/deploy/serving/python/preprocess_ops.py b/deploy/serving/python/preprocess_ops.py
new file mode 100644
index 0000000000000000000000000000000000000000..eaf013744e938164f6e42a99c5517c35e0beeb83
--- /dev/null
+++ b/deploy/serving/python/preprocess_ops.py
@@ -0,0 +1,414 @@
+import numpy as np
+import cv2
+import copy
+
+
+def decode_image(im, img_info):
+    im = np.array(im)
+    img_info['im_shape'] = np.array(im.shape[:2], dtype=np.float32)
+    img_info['scale_factor'] = np.array([1., 1.], dtype=np.float32)
+    return im, img_info
+
+
+class Resize(object):
+    """resize image by target_size and max_size
+    Args:
+        target_size (int): the target size of image
+        keep_ratio (bool): whether keep_ratio or not, default true
+        interp (int): method of resize
+    """
+
+    def __init__(self, target_size, keep_ratio=True, interp=cv2.INTER_LINEAR):
+        if isinstance(target_size, int):
+            target_size = [target_size, target_size]
+        self.target_size = target_size
+        self.keep_ratio = keep_ratio
+        self.interp = interp
+
+    def __call__(self, im, im_info):
+        """
+        Args:
+            im (np.ndarray): image (np.ndarray)
+            im_info (dict): info of image
+        Returns:
+            im (np.ndarray):  processed image (np.ndarray)
+            im_info (dict): info of processed image
+        """
+        assert len(self.target_size) == 2
+        assert self.target_size[0] > 0 and self.target_size[1] > 0
+        im_channel = im.shape[2]
+        im_scale_y, im_scale_x = self.generate_scale(im)
+        im = cv2.resize(
+            im,
+            None,
+            None,
+            fx=im_scale_x,
+            fy=im_scale_y,
+            interpolation=self.interp)
+        im_info['im_shape'] = np.array(im.shape[:2]).astype('float32')
+        im_info['scale_factor'] = np.array(
+            [im_scale_y, im_scale_x]).astype('float32')
+        return im, im_info
+
+    def generate_scale(self, im):
+        """
+        Args:
+            im (np.ndarray): image (np.ndarray)
+        Returns:
+            im_scale_x: the resize ratio of X
+            im_scale_y: the resize ratio of Y
+        """
+        origin_shape = im.shape[:2]
+        im_c = im.shape[2]
+        if self.keep_ratio:
+            im_size_min = np.min(origin_shape)
+            im_size_max = np.max(origin_shape)
+            target_size_min = np.min(self.target_size)
+            target_size_max = np.max(self.target_size)
+            im_scale = float(target_size_min) / float(im_size_min)
+            if np.round(im_scale * im_size_max) > target_size_max:
+                im_scale = float(target_size_max) / float(im_size_max)
+            im_scale_x = im_scale
+            im_scale_y = im_scale
+        else:
+            resize_h, resize_w = self.target_size
+            im_scale_y = resize_h / float(origin_shape[0])
+            im_scale_x = resize_w / float(origin_shape[1])
+        return im_scale_y, im_scale_x
+
+
+class NormalizeImage(object):
+    """normalize image
+    Args:
+        mean (list): im - mean
+        std (list): im / std
+        is_scale (bool): whether need im / 255
+        is_channel_first (bool): if True: image shape is CHW, else: HWC
+    """
+
+    def __init__(self, mean, std, is_scale=True):
+        self.mean = mean
+        self.std = std
+        self.is_scale = is_scale
+
+    def __call__(self, im, im_info):
+        """
+        Args:
+            im (np.ndarray): image (np.ndarray)
+            im_info (dict): info of image
+        Returns:
+            im (np.ndarray):  processed image (np.ndarray)
+            im_info (dict): info of processed image
+        """
+        im = im.astype(np.float32, copy=False)
+        mean = np.array(self.mean)[np.newaxis, np.newaxis, :]
+        std = np.array(self.std)[np.newaxis, np.newaxis, :]
+
+        if self.is_scale:
+            im = im / 255.0
+        im -= mean
+        im /= std
+        return im, im_info
+
+
+class Permute(object):
+    """permute image
+    Args:
+        to_bgr (bool): whether convert RGB to BGR
+        channel_first (bool): whether convert HWC to CHW
+    """
+
+    def __init__(self, ):
+        super(Permute, self).__init__()
+
+    def __call__(self, im, im_info):
+        """
+        Args:
+            im (np.ndarray): image (np.ndarray)
+            im_info (dict): info of image
+        Returns:
+            im (np.ndarray):  processed image (np.ndarray)
+            im_info (dict): info of processed image
+        """
+        im = im.transpose((2, 0, 1)).copy()
+        return im, im_info
+
+
+class PadStride(object):
+    """ padding image for model with FPN, instead PadBatch(pad_to_stride) in original config
+    Args:
+        stride (bool): model with FPN need image shape % stride == 0
+    """
+
+    def __init__(self, stride=0):
+        self.coarsest_stride = stride
+
+    def __call__(self, im, im_info):
+        """
+        Args:
+            im (np.ndarray): image (np.ndarray)
+            im_info (dict): info of image
+        Returns:
+            im (np.ndarray):  processed image (np.ndarray)
+            im_info (dict): info of processed image
+        """
+        coarsest_stride = self.coarsest_stride
+        if coarsest_stride <= 0:
+            return im, im_info
+        im_c, im_h, im_w = im.shape
+        pad_h = int(np.ceil(float(im_h) / coarsest_stride) * coarsest_stride)
+        pad_w = int(np.ceil(float(im_w) / coarsest_stride) * coarsest_stride)
+        padding_im = np.zeros((im_c, pad_h, pad_w), dtype=np.float32)
+        padding_im[:, :im_h, :im_w] = im
+        return padding_im, im_info
+
+
+class LetterBoxResize(object):
+    def __init__(self, target_size):
+        """
+        Resize image to target size, convert normalized xywh to pixel xyxy
+        format ([x_center, y_center, width, height] -> [x0, y0, x1, y1]).
+        Args:
+            target_size (int|list): image target size.
+        """
+        super(LetterBoxResize, self).__init__()
+        if isinstance(target_size, int):
+            target_size = [target_size, target_size]
+        self.target_size = target_size
+
+    def letterbox(self, img, height, width, color=(127.5, 127.5, 127.5)):
+        # letterbox: resize a rectangular image to a padded rectangular
+        shape = img.shape[:2]  # [height, width]
+        ratio_h = float(height) / shape[0]
+        ratio_w = float(width) / shape[1]
+        ratio = min(ratio_h, ratio_w)
+        new_shape = (round(shape[1] * ratio),
+                     round(shape[0] * ratio))  # [width, height]
+        padw = (width - new_shape[0]) / 2
+        padh = (height - new_shape[1]) / 2
+        top, bottom = round(padh - 0.1), round(padh + 0.1)
+        left, right = round(padw - 0.1), round(padw + 0.1)
+
+        img = cv2.resize(
+            img, new_shape, interpolation=cv2.INTER_AREA)  # resized, no border
+        img = cv2.copyMakeBorder(
+            img, top, bottom, left, right, cv2.BORDER_CONSTANT,
+            value=color)  # padded rectangular
+        return img, ratio, padw, padh
+
+    def __call__(self, im, im_info):
+        """
+        Args:
+            im (np.ndarray): image (np.ndarray)
+            im_info (dict): info of image
+        Returns:
+            im (np.ndarray):  processed image (np.ndarray)
+            im_info (dict): info of processed image
+        """
+        assert len(self.target_size) == 2
+        assert self.target_size[0] > 0 and self.target_size[1] > 0
+        height, width = self.target_size
+        h, w = im.shape[:2]
+        im, ratio, padw, padh = self.letterbox(im, height=height, width=width)
+
+        new_shape = [round(h * ratio), round(w * ratio)]
+        im_info['im_shape'] = np.array(new_shape, dtype=np.float32)
+        im_info['scale_factor'] = np.array([ratio, ratio], dtype=np.float32)
+        return im, im_info
+
+
+class Pad(object):
+    def __init__(self, size, fill_value=[114.0, 114.0, 114.0]):
+        """
+        Pad image to a specified size.
+        Args:
+            size (list[int]): image target size
+            fill_value (list[float]): rgb value of pad area, default (114.0, 114.0, 114.0)
+        """
+        super(Pad, self).__init__()
+        if isinstance(size, int):
+            size = [size, size]
+        self.size = size
+        self.fill_value = fill_value
+
+    def __call__(self, im, im_info):
+        im_h, im_w = im.shape[:2]
+        h, w = self.size
+        if h == im_h and w == im_w:
+            im = im.astype(np.float32)
+            return im, im_info
+
+        canvas = np.ones((h, w, 3), dtype=np.float32)
+        canvas *= np.array(self.fill_value, dtype=np.float32)
+        canvas[0:im_h, 0:im_w, :] = im.astype(np.float32)
+        im = canvas
+        return im, im_info
+
+
+def rotate_point(pt, angle_rad):
+    """Rotate a point by an angle.
+
+    Args:
+        pt (list[float]): 2 dimensional point to be rotated
+        angle_rad (float): rotation angle by radian
+
+    Returns:
+        list[float]: Rotated point.
+    """
+    assert len(pt) == 2
+    sn, cs = np.sin(angle_rad), np.cos(angle_rad)
+    new_x = pt[0] * cs - pt[1] * sn
+    new_y = pt[0] * sn + pt[1] * cs
+    rotated_pt = [new_x, new_y]
+
+    return rotated_pt
+
+
+def _get_3rd_point(a, b):
+    """To calculate the affine matrix, three pairs of points are required. This
+    function is used to get the 3rd point, given 2D points a & b.
+
+    The 3rd point is defined by rotating vector `a - b` by 90 degrees
+    anticlockwise, using b as the rotation center.
+
+    Args:
+        a (np.ndarray): point(x,y)
+        b (np.ndarray): point(x,y)
+
+    Returns:
+        np.ndarray: The 3rd point.
+    """
+    assert len(a) == 2
+    assert len(b) == 2
+    direction = a - b
+    third_pt = b + np.array([-direction[1], direction[0]], dtype=np.float32)
+
+    return third_pt
+
+
+def get_affine_transform(center,
+                         input_size,
+                         rot,
+                         output_size,
+                         shift=(0., 0.),
+                         inv=False):
+    """Get the affine transform matrix, given the center/scale/rot/output_size.
+
+    Args:
+        center (np.ndarray[2, ]): Center of the bounding box (x, y).
+        scale (np.ndarray[2, ]): Scale of the bounding box
+            wrt [width, height].
+        rot (float): Rotation angle (degree).
+        output_size (np.ndarray[2, ]): Size of the destination heatmaps.
+        shift (0-100%): Shift translation ratio wrt the width/height.
+            Default (0., 0.).
+        inv (bool): Option to inverse the affine transform direction.
+            (inv=False: src->dst or inv=True: dst->src)
+
+    Returns:
+        np.ndarray: The transform matrix.
+    """
+    assert len(center) == 2
+    assert len(output_size) == 2
+    assert len(shift) == 2
+    if not isinstance(input_size, (np.ndarray, list)):
+        input_size = np.array([input_size, input_size], dtype=np.float32)
+    scale_tmp = input_size
+
+    shift = np.array(shift)
+    src_w = scale_tmp[0]
+    dst_w = output_size[0]
+    dst_h = output_size[1]
+
+    rot_rad = np.pi * rot / 180
+    src_dir = rotate_point([0., src_w * -0.5], rot_rad)
+    dst_dir = np.array([0., dst_w * -0.5])
+
+    src = np.zeros((3, 2), dtype=np.float32)
+    src[0, :] = center + scale_tmp * shift
+    src[1, :] = center + src_dir + scale_tmp * shift
+    src[2, :] = _get_3rd_point(src[0, :], src[1, :])
+
+    dst = np.zeros((3, 2), dtype=np.float32)
+    dst[0, :] = [dst_w * 0.5, dst_h * 0.5]
+    dst[1, :] = np.array([dst_w * 0.5, dst_h * 0.5]) + dst_dir
+    dst[2, :] = _get_3rd_point(dst[0, :], dst[1, :])
+
+    if inv:
+        trans = cv2.getAffineTransform(np.float32(dst), np.float32(src))
+    else:
+        trans = cv2.getAffineTransform(np.float32(src), np.float32(dst))
+
+    return trans
+
+
+class WarpAffine(object):
+    """Warp affine the image
+    """
+
+    def __init__(self,
+                 keep_res=False,
+                 pad=31,
+                 input_h=512,
+                 input_w=512,
+                 scale=0.4,
+                 shift=0.1):
+        self.keep_res = keep_res
+        self.pad = pad
+        self.input_h = input_h
+        self.input_w = input_w
+        self.scale = scale
+        self.shift = shift
+
+    def __call__(self, im, im_info):
+        """
+        Args:
+            im (np.ndarray): image (np.ndarray)
+            im_info (dict): info of image
+        Returns:
+            im (np.ndarray):  processed image (np.ndarray)
+            im_info (dict): info of processed image
+        """
+        img = cv2.cvtColor(im, cv2.COLOR_RGB2BGR)
+
+        h, w = img.shape[:2]
+
+        if self.keep_res:
+            input_h = (h | self.pad) + 1
+            input_w = (w | self.pad) + 1
+            s = np.array([input_w, input_h], dtype=np.float32)
+            c = np.array([w // 2, h // 2], dtype=np.float32)
+
+        else:
+            s = max(h, w) * 1.0
+            input_h, input_w = self.input_h, self.input_w
+            c = np.array([w / 2., h / 2.], dtype=np.float32)
+
+        trans_input = get_affine_transform(c, s, 0, [input_w, input_h])
+        img = cv2.resize(img, (w, h))
+        inp = cv2.warpAffine(
+            img, trans_input, (input_w, input_h), flags=cv2.INTER_LINEAR)
+        return inp, im_info
+
+
+class Compose:
+    def __init__(self, transforms):
+        self.transforms = []
+        for op_info in transforms:
+            new_op_info = op_info.copy()
+            op_type = new_op_info.pop('type')
+            self.transforms.append(eval(op_type)(**new_op_info))
+
+        self.im_info = {
+            'scale_factor': np.array(
+                [1., 1.], dtype=np.float32),
+            'im_shape': None
+        }
+
+    def __call__(self, img):
+        img, self.im_info = decode_image(img, self.im_info)
+        for t in self.transforms:
+            img, self.im_info = t(img, self.im_info)
+        inputs = copy.deepcopy(self.im_info)
+        inputs['image'] = img
+        return inputs
diff --git a/deploy/serving/python/web_service.py b/deploy/serving/python/web_service.py
new file mode 100644
index 0000000000000000000000000000000000000000..5a7018f98d670a2f9dc3715888d9b0191091d088
--- /dev/null
+++ b/deploy/serving/python/web_service.py
@@ -0,0 +1,247 @@
+# 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 copy
+
+from paddle_serving_server.web_service import WebService, Op
+from paddle_serving_server.proto import general_model_config_pb2 as m_config
+import google.protobuf.text_format
+
+import os
+import numpy as np
+import base64
+from PIL import Image
+import io
+from preprocess_ops import Compose
+
+from argparse import ArgumentParser, RawDescriptionHelpFormatter
+import yaml
+
+# Global dictionary
+SUPPORT_MODELS = {
+    'YOLO',
+    'RCNN',
+    'SSD',
+    'Face',
+    'FCOS',
+    'SOLOv2',
+    'TTFNet',
+    'S2ANet',
+    'JDE',
+    'FairMOT',
+    'DeepSORT',
+    'GFL',
+    'PicoDet',
+    'CenterNet',
+    'TOOD',
+    'RetinaNet',
+    'StrongBaseline',
+    'STGCN',
+    'YOLOX',
+}
+
+GLOBAL_VAR = {}
+
+
+class ArgsParser(ArgumentParser):
+    def __init__(self):
+        super(ArgsParser, self).__init__(
+            formatter_class=RawDescriptionHelpFormatter)
+        self.add_argument(
+            "-c",
+            "--config",
+            default="deploy/serving/python/config.yml",
+            help="configuration file to use")
+        self.add_argument(
+            "--model_dir",
+            type=str,
+            default=None,
+            help=("Directory include:'model.pdiparams', 'model.pdmodel', "
+                  "'infer_cfg.yml', created by tools/export_model.py."),
+            required=True)
+        self.add_argument(
+            "-o", "--opt", nargs='+', help="set configuration options")
+
+    def parse_args(self, argv=None):
+        args = super(ArgsParser, self).parse_args(argv)
+        assert args.config is not None, \
+            "Please specify --config=configure_file_path."
+        args.service_config = self._parse_opt(args.opt, args.config)
+        print("args config:", args.service_config)
+        args.model_config = PredictConfig(args.model_dir)
+        return args
+
+    def _parse_helper(self, v):
+        if v.isnumeric():
+            if "." in v:
+                v = float(v)
+            else:
+                v = int(v)
+        elif v == "True" or v == "False":
+            v = (v == "True")
+        return v
+
+    def _parse_opt(self, opts, conf_path):
+        f = open(conf_path)
+        config = yaml.load(f, Loader=yaml.Loader)
+        if not opts:
+            return config
+        for s in opts:
+            s = s.strip()
+            k, v = s.split('=')
+            v = self._parse_helper(v)
+            if "devices" in k:
+                v = str(v)
+            print(k, v, type(v))
+            cur = config
+            parent = cur
+            for kk in k.split("."):
+                if kk not in cur:
+                    cur[kk] = {}
+                    parent = cur
+                    cur = cur[kk]
+                else:
+                    parent = cur
+                    cur = cur[kk]
+            parent[k.split(".")[-1]] = v
+        return config
+
+
+class PredictConfig(object):
+    """set config of preprocess, postprocess and visualize
+    Args:
+        model_dir (str): root path of infer_cfg.yml
+    """
+
+    def __init__(self, model_dir):
+        # parsing Yaml config for Preprocess
+        deploy_file = os.path.join(model_dir, 'infer_cfg.yml')
+        with open(deploy_file) as f:
+            yml_conf = yaml.safe_load(f)
+        self.check_model(yml_conf)
+        self.arch = yml_conf['arch']
+        self.preprocess_infos = yml_conf['Preprocess']
+        self.min_subgraph_size = yml_conf['min_subgraph_size']
+        self.labels = yml_conf['label_list']
+        self.use_dynamic_shape = yml_conf['use_dynamic_shape']
+        self.draw_threshold = yml_conf.get("draw_threshold", 0.5)
+        self.mask = yml_conf.get("mask", False)
+        self.tracker = yml_conf.get("tracker", None)
+        self.nms = yml_conf.get("NMS", None)
+        self.fpn_stride = yml_conf.get("fpn_stride", None)
+        if self.arch == 'RCNN' and yml_conf.get('export_onnx', False):
+            print(
+                'The RCNN export model is used for ONNX and it only supports batch_size = 1'
+            )
+        self.print_config()
+
+    def check_model(self, yml_conf):
+        """
+        Raises:
+            ValueError: loaded model not in supported model type
+        """
+        for support_model in SUPPORT_MODELS:
+            if support_model in yml_conf['arch']:
+                return True
+        raise ValueError("Unsupported arch: {}, expect {}".format(yml_conf[
+            'arch'], SUPPORT_MODELS))
+
+    def print_config(self):
+        print('-----------  Model Configuration -----------')
+        print('%s: %s' % ('Model Arch', self.arch))
+        print('%s: ' % ('Transform Order'))
+        for op_info in self.preprocess_infos:
+            print('--%s: %s' % ('transform op', op_info['type']))
+        print('--------------------------------------------')
+
+
+class DetectorOp(Op):
+    def init_op(self):
+        self.preprocess_pipeline = Compose(GLOBAL_VAR['preprocess_ops'])
+
+    def preprocess(self, input_dicts, data_id, log_id):
+        (_, input_dict), = input_dicts.items()
+        inputs = []
+        for key, data in input_dict.items():
+            data = base64.b64decode(data.encode('utf8'))
+            byte_stream = io.BytesIO(data)
+            img = Image.open(byte_stream).convert("RGB")
+            inputs.append(self.preprocess_pipeline(img))
+        inputs = self.collate_inputs(inputs)
+        return inputs, False, None, ""
+
+    def postprocess(self, input_dicts, fetch_dict, data_id, log_id):
+        (_, input_dict), = input_dicts.items()
+        bboxes = fetch_dict["multiclass_nms3_0.tmp_0"]
+        bboxes_num = fetch_dict["multiclass_nms3_0.tmp_2"]
+        draw_threshold = GLOBAL_VAR['model_config'].draw_threshold
+        idx = 0
+        result = {}
+        for k, num in zip(input_dict.keys(), bboxes_num):
+            bbox = bboxes[idx:idx + num]
+            result[k] = self.parse_det_result(bbox, draw_threshold,
+                                              GLOBAL_VAR['model_config'].labels)
+        return result, None, ""
+
+    def collate_inputs(self, inputs):
+        collate_inputs = {k: [] for k in inputs[0].keys()}
+        for info in inputs:
+            for k in collate_inputs.keys():
+                collate_inputs[k].append(info[k])
+        return {
+            k: np.stack(v)
+            for k, v in collate_inputs.items() if k in GLOBAL_VAR['feed_vars']
+        }
+
+    def parse_det_result(self, bbox, draw_threshold, label_list):
+        result = []
+        for line in bbox:
+            if line[1] > draw_threshold:
+                result.append(f"{label_list[int(line[0])]} {line[1]} "
+                              f"{line[2]} {line[3]} {line[4]} {line[5]}")
+        return result
+
+
+class DetectorService(WebService):
+    def get_pipeline_response(self, read_op):
+        return DetectorOp(name="ppdet", input_ops=[read_op])
+
+
+def get_model_vars(model_dir, service_config):
+    serving_server_dir = os.path.join(model_dir, "serving_server")
+    # rewrite model_config
+    service_config['op']['ppdet']['local_service_conf'][
+        'model_config'] = serving_server_dir
+    f = open(
+        os.path.join(serving_server_dir, "serving_server_conf.prototxt"), 'r')
+    model_var = google.protobuf.text_format.Merge(
+        str(f.read()), m_config.GeneralModelConfig())
+    feed_vars = [var.name for var in model_var.feed_var]
+    fetch_vars = [var.name for var in model_var.fetch_var]
+    return feed_vars, fetch_vars
+
+
+if __name__ == '__main__':
+    # load config and prepare the service
+    FLAGS = ArgsParser().parse_args()
+    feed_vars, fetch_vars = get_model_vars(FLAGS.model_dir,
+                                           FLAGS.service_config)
+    GLOBAL_VAR['feed_vars'] = feed_vars
+    GLOBAL_VAR['fetch_vars'] = fetch_vars
+    GLOBAL_VAR['preprocess_ops'] = FLAGS.model_config.preprocess_infos
+    GLOBAL_VAR['model_config'] = FLAGS.model_config
+    # define the service
+    uci_service = DetectorService(name="ppdet")
+    uci_service.prepare_pipeline_config(yml_dict=FLAGS.service_config)
+    # start the service
+    uci_service.run_service()
diff --git a/test_tipc/configs/cascade_rcnn/cascade_mask_rcnn_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/cascade_rcnn/cascade_mask_rcnn_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index e9bc4e640da9e4576ffae9b3751f8164eb91d279..5a38b046f8ffba07fcc78c94432986bea4451687 100644
--- a/test_tipc/configs/cascade_rcnn/cascade_mask_rcnn_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/cascade_rcnn/cascade_mask_rcnn_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/cascade_mask_rcnn_r50_fpn_1x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/cascade_rcnn/cascade_mask_rcnn_r50_fpn_1x_coco.yml -o
diff --git a/test_tipc/configs/cascade_rcnn/cascade_mask_rcnn_r50_vd_fpn_ssld_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/cascade_rcnn/cascade_mask_rcnn_r50_vd_fpn_ssld_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 446bcd0b1dcf56e156a7ae3f980476f3bc96e7a4..5d49ea4bd86f8d98b7a2c732566334a69a48e33a 100644
--- a/test_tipc/configs/cascade_rcnn/cascade_mask_rcnn_r50_vd_fpn_ssld_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/cascade_rcnn/cascade_mask_rcnn_r50_vd_fpn_ssld_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/cascade_mask_rcnn_r50_vd_fpn_ssld_1x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/cascade_rcnn/cascade_mask_rcnn_r50_vd_fpn_ssld_1x_coco.yml -o
diff --git a/test_tipc/configs/cascade_rcnn/cascade_rcnn_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/cascade_rcnn/cascade_rcnn_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 4ce15810e0099195815525fba7eb2d5805500161..f1c3e3f378109459e1846374944c825936a0d774 100644
--- a/test_tipc/configs/cascade_rcnn/cascade_rcnn_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/cascade_rcnn/cascade_rcnn_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/cascade_rcnn_r50_fpn_1x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/cascade_rcnn/cascade_rcnn_r50_fpn_1x_coco.yml -o
diff --git a/test_tipc/configs/cascade_rcnn/cascade_rcnn_r50_vd_fpn_ssld_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/cascade_rcnn/cascade_rcnn_r50_vd_fpn_ssld_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index c9c013025d7312303e94a57209efecace3c65087..277437f295d86106b4d1a913bc1f3271328a5750 100644
--- a/test_tipc/configs/cascade_rcnn/cascade_rcnn_r50_vd_fpn_ssld_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/cascade_rcnn/cascade_rcnn_r50_vd_fpn_ssld_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/cascade_rcnn_r50_vd_fpn_ssld_1x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/cascade_rcnn/cascade_rcnn_r50_vd_fpn_ssld_1x_coco.yml -o
diff --git a/test_tipc/configs/deformable_detr/deformable_detr_r50_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/deformable_detr/deformable_detr_r50_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 7ac82f945d5b39a8ebbb6e201fd21a8233fa4b94..49821ac5320b46b627cb193dd9e587bf3cd0fa31 100644
--- a/test_tipc/configs/deformable_detr/deformable_detr_r50_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/deformable_detr/deformable_detr_r50_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/deformable_detr_r50_1x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/deformable_detr/deformable_detr_r50_1x_coco.yml -o
diff --git a/test_tipc/configs/detr/detr_r50_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/detr/detr_r50_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 4b49de9da047810de28bcf3c64374ee0f2ea480b..b6073456f289fc52b8feb02c09943efafbacaa87 100644
--- a/test_tipc/configs/detr/detr_r50_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/detr/detr_r50_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/detr_r50_1x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/detr/detr_r50_1x_coco.yml -o
diff --git a/test_tipc/configs/dota/s2anet_1x_spine_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/dota/s2anet_1x_spine_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 287c53afa96d0de164331705c5044e29be17c0d7..08f29765f364639e33d82767909b7d7362bce637 100644
--- a/test_tipc/configs/dota/s2anet_1x_spine_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/dota/s2anet_1x_spine_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/s2anet_1x_spine.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/dota/s2anet_1x_spine.yml -o
diff --git a/test_tipc/configs/dota/s2anet_alignconv_2x_spine_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/dota/s2anet_alignconv_2x_spine_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 00475ef3e8867cdb95f0423ec72f562542af3a9d..04b2082c53c7ec3a5aab35cb13506912a3e002f5 100644
--- a/test_tipc/configs/dota/s2anet_alignconv_2x_spine_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/dota/s2anet_alignconv_2x_spine_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=1|lite_train_whole_infer=1|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/s2anet_alignconv_2x_dota.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/spine_coco/test/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c test_tipc/configs/dota/s2anet_alignconv_2x_spine.yml -o
diff --git a/test_tipc/configs/dota/s2anet_conv_2x_spine_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/dota/s2anet_conv_2x_spine_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 8f2d0992d39e2ecbd8a948bdc3dffe0786a20da4..f88d2241a15a26786e4d917ccf6634b71178b503 100644
--- a/test_tipc/configs/dota/s2anet_conv_2x_spine_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/dota/s2anet_conv_2x_spine_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=1|lite_train_whole_infer=1|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/s2anet_conv_2x_dota.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/spine_coco/test/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c test_tipc/configs/dota/s2anet_conv_2x_spine.yml -o
diff --git a/test_tipc/configs/face_detection/blazeface_1000e_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/face_detection/blazeface_1000e_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index a4d45d52e7933f7d0ccbda06872ac5d9d9b0af7c..1300f81d8660c56d8c257d802559268232baf0fb 100644
--- a/test_tipc/configs/face_detection/blazeface_1000e_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/face_detection/blazeface_1000e_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/blazeface_1000e.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/wider_face/WIDER_val/images/0--Parade/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/face_detection/blazeface_1000e.yml -o
diff --git a/test_tipc/configs/face_detection/blazeface_fpn_ssh_1000e_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/face_detection/blazeface_fpn_ssh_1000e_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index c654b6de5254dd0c10c9ab658971364a442c8790..b80b650d9332fe9b8157122934c95077824b0261 100644
--- a/test_tipc/configs/face_detection/blazeface_fpn_ssh_1000e_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/face_detection/blazeface_fpn_ssh_1000e_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/blazeface_fpn_ssh_1000e.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/wider_face/WIDER_val/images/0--Parade/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/face_detection/blazeface_fpn_ssh_1000e.yml -o
diff --git a/test_tipc/configs/faster_rcnn/faster_rcnn_r101_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/faster_rcnn/faster_rcnn_r101_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 4d824f157ff34268da1858aed4413790f3d88e71..e2c96858a453f12e02654381ef48b47c0a4f0832 100644
--- a/test_tipc/configs/faster_rcnn/faster_rcnn_r101_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/faster_rcnn/faster_rcnn_r101_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/faster_rcnn_r101_1x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/faster_rcnn/faster_rcnn_r101_1x_coco.yml -o
diff --git a/test_tipc/configs/faster_rcnn/faster_rcnn_r101_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/faster_rcnn/faster_rcnn_r101_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 56406cc53a4dc60e84386dc48345da85fcc11c36..e6ef4154e5f4d55d89fc474e73a6ac2b8d5c2760 100644
--- a/test_tipc/configs/faster_rcnn/faster_rcnn_r101_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/faster_rcnn/faster_rcnn_r101_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/faster_rcnn_r101_fpn_1x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/faster_rcnn/faster_rcnn_r101_fpn_1x_coco.yml -o
diff --git a/test_tipc/configs/faster_rcnn/faster_rcnn_r101_vd_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/faster_rcnn/faster_rcnn_r101_vd_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 6423f3afc7f8b1d4685c46931460ab785e433526..0eb12969a30cf590234a435adb923421f1b2f95c 100644
--- a/test_tipc/configs/faster_rcnn/faster_rcnn_r101_vd_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/faster_rcnn/faster_rcnn_r101_vd_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/faster_rcnn_r101_vd_fpn_1x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/faster_rcnn/faster_rcnn_r101_vd_fpn_1x_coco.yml -o
diff --git a/test_tipc/configs/faster_rcnn/faster_rcnn_r34_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/faster_rcnn/faster_rcnn_r34_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index fcc33579b31fc36d48f1e286a41f291f8ce6f66e..a7ea74d437ebf0ac25b2f47a2a5ed05bcd448f66 100644
--- a/test_tipc/configs/faster_rcnn/faster_rcnn_r34_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/faster_rcnn/faster_rcnn_r34_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/faster_rcnn_r34_fpn_1x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/faster_rcnn/faster_rcnn_r34_fpn_1x_coco.yml -o
diff --git a/test_tipc/configs/faster_rcnn/faster_rcnn_r34_vd_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/faster_rcnn/faster_rcnn_r34_vd_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 57a10cb8647cd6e8452a68cec6ddf00fcc7ac511..9c45cd7723e0970b016f66b3bae6843a6d3f57b0 100644
--- a/test_tipc/configs/faster_rcnn/faster_rcnn_r34_vd_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/faster_rcnn/faster_rcnn_r34_vd_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/faster_rcnn_r34_vd_fpn_1x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/faster_rcnn/faster_rcnn_r34_vd_fpn_1x_coco.yml -o
diff --git a/test_tipc/configs/faster_rcnn/faster_rcnn_r50_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/faster_rcnn/faster_rcnn_r50_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 5ce75c465bc5f05c15c661949c466d21fdde99a0..a1584b90e148fce097fe74350fa2a59f47235375 100644
--- a/test_tipc/configs/faster_rcnn/faster_rcnn_r50_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/faster_rcnn/faster_rcnn_r50_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/faster_rcnn_r50_1x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/faster_rcnn/faster_rcnn_r50_1x_coco.yml -o
diff --git a/test_tipc/configs/faster_rcnn/faster_rcnn_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/faster_rcnn/faster_rcnn_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index d886e94af321cf383689f59bbbfd000b5117e66f..da2470ac26a357395c5565ea7fd44fd308ec6427 100644
--- a/test_tipc/configs/faster_rcnn/faster_rcnn_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/faster_rcnn/faster_rcnn_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/faster_rcnn_r50_fpn_1x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/faster_rcnn/faster_rcnn_r50_fpn_1x_coco.yml -o
diff --git a/test_tipc/configs/faster_rcnn/faster_rcnn_r50_vd_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/faster_rcnn/faster_rcnn_r50_vd_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 4c6cc960c28d77213869e7ca0b980a4d9a87aabc..c12efa6cd545c340f9372d2ec9e05b80ba7b0737 100644
--- a/test_tipc/configs/faster_rcnn/faster_rcnn_r50_vd_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/faster_rcnn/faster_rcnn_r50_vd_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/faster_rcnn_r50_vd_1x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/faster_rcnn/faster_rcnn_r50_vd_1x_coco.yml -o
diff --git a/test_tipc/configs/faster_rcnn/faster_rcnn_r50_vd_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/faster_rcnn/faster_rcnn_r50_vd_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index a8ca7c6103f5da650123e589288001ae587888cf..a852290aae86d8e2074076f6b9018a8f7b7e7960 100644
--- a/test_tipc/configs/faster_rcnn/faster_rcnn_r50_vd_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/faster_rcnn/faster_rcnn_r50_vd_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/faster_rcnn_r50_vd_fpn_1x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/faster_rcnn/faster_rcnn_r50_vd_fpn_1x_coco.yml -o
diff --git a/test_tipc/configs/faster_rcnn/faster_rcnn_swin_tiny_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/faster_rcnn/faster_rcnn_swin_tiny_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index ef1a06f00b7e8a819d77115ee1f16001380ed907..6fc1c6edcfb4976bcac3527919aaeaeea5ed6399 100644
--- a/test_tipc/configs/faster_rcnn/faster_rcnn_swin_tiny_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/faster_rcnn/faster_rcnn_swin_tiny_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/faster_rcnn_swin_tiny_fpn_1x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/faster_rcnn/faster_rcnn_swin_tiny_fpn_1x_coco.yml -o
diff --git a/test_tipc/configs/faster_rcnn/faster_rcnn_x101_vd_64x4d_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/faster_rcnn/faster_rcnn_x101_vd_64x4d_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 7cfcac208416a6ec729ebb2caa19446400e91394..9dda811bb7d258e44c98e230925092541f98a9d4 100644
--- a/test_tipc/configs/faster_rcnn/faster_rcnn_x101_vd_64x4d_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/faster_rcnn/faster_rcnn_x101_vd_64x4d_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/faster_rcnn_x101_vd_64x4d_fpn_1x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/faster_rcnn/faster_rcnn_x101_vd_64x4d_fpn_1x_coco.yml -o
diff --git a/test_tipc/configs/fcos/fcos_dcn_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/fcos/fcos_dcn_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 2320b41f824bf975bef50e79024f64161632af79..88bc7aa0818f693b8268e8f0bad362d92eebb1c7 100644
--- a/test_tipc/configs/fcos/fcos_dcn_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/fcos/fcos_dcn_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/fcos_dcn_r50_fpn_1x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/fcos/fcos_dcn_r50_fpn_1x_coco.yml -o
diff --git a/test_tipc/configs/fcos/fcos_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/fcos/fcos_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 420417f02e13ba1934f81608edddde815fd9ab20..6475ac8863fdda3732bab9dcece7c9a08f5e6e58 100644
--- a/test_tipc/configs/fcos/fcos_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/fcos/fcos_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/fcos_r50_fpn_1x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/fcos/fcos_r50_fpn_1x_coco.yml -o
diff --git a/test_tipc/configs/gfl/gfl_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/gfl/gfl_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index d491ec0ca0f85746641fd142197a2eed7304cb93..22169af14ebe878050187d0a8daa8fec28f082dc 100644
--- a/test_tipc/configs/gfl/gfl_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/gfl/gfl_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/gfl_r50_fpn_1x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/gfl/gfl_r50_fpn_1x_coco.yml -o
diff --git a/test_tipc/configs/keypoint/dark_hrnet_w32_256x192_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/keypoint/dark_hrnet_w32_256x192_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 8b334240c1883d9b38184bfeb44a3056fd2d8e64..7875054063fef898ff3bce85e3c661900358cfc9 100644
--- a/test_tipc/configs/keypoint/dark_hrnet_w32_256x192_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/keypoint/dark_hrnet_w32_256x192_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/keypoint/dark_hrnet_w32_256x192.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/keypoint/hrnet/dark_hrnet_w32_256x192.yml -o
diff --git a/test_tipc/configs/keypoint/dark_hrnet_w48_256x192_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/keypoint/dark_hrnet_w48_256x192_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index b50aa8f9c832ca13b8c7302547d1beb106822c6f..8a6c6f17a1d93e50cc39044ac52e581d909348ca 100644
--- a/test_tipc/configs/keypoint/dark_hrnet_w48_256x192_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/keypoint/dark_hrnet_w48_256x192_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/keypoint/dark_hrnet_w48_256x192.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/keypoint/hrnet/dark_hrnet_w48_256x192.yml -o
diff --git a/test_tipc/configs/keypoint/higherhrnet_hrnet_w32_512_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/keypoint/higherhrnet_hrnet_w32_512_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index c3496b41d3bb54b122d2a056ef4d54927cf32844..9b5e0503edb7f8a30a26f7eb53e515faaafe93f5 100644
--- a/test_tipc/configs/keypoint/higherhrnet_hrnet_w32_512_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/keypoint/higherhrnet_hrnet_w32_512_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/keypoint/higherhrnet_hrnet_w32_512.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/keypoint/higherhrnet/higherhrnet_hrnet_w32_512.yml -o
diff --git a/test_tipc/configs/keypoint/hrnet_w32_256x192_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/keypoint/hrnet_w32_256x192_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 9f8aaf54fae2a77381747623c1fe03a4e41914e8..26c6a3df0ebb235724237ae027473817bbe8fe79 100644
--- a/test_tipc/configs/keypoint/hrnet_w32_256x192_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/keypoint/hrnet_w32_256x192_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/keypoint/hrnet_w32_256x192.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/keypoint/hrnet/hrnet_w32_256x192.yml -o
diff --git a/test_tipc/configs/keypoint/tinypose_128x96_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/keypoint/tinypose_128x96_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index e0198b289aed988be5a2cf4ccb9e6a3f435811f0..98111895219632c6390bdbfeb79d772dfc8c0ebc 100644
--- a/test_tipc/configs/keypoint/tinypose_128x96_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/keypoint/tinypose_128x96_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/keypoint/tinypose_128x96.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c test_tipc/configs/keypoint/tinypose_128x96.yml -o
diff --git a/test_tipc/configs/mask_rcnn/mask_rcnn_r101_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/mask_rcnn/mask_rcnn_r101_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index cebdb83fbedb8348684716ad57247cc6acfbddd5..2be777af8468d68b8ddeb4d60d7156a34bb9f277 100644
--- a/test_tipc/configs/mask_rcnn/mask_rcnn_r101_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/mask_rcnn/mask_rcnn_r101_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/mask_rcnn_r101_fpn_1x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/mask_rcnn/mask_rcnn_r101_fpn_1x_coco.yml -o
diff --git a/test_tipc/configs/mask_rcnn/mask_rcnn_r101_vd_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/mask_rcnn/mask_rcnn_r101_vd_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 0d7f36ab68adc683504d77e08e7603d8407803f7..8e26f344365cdedd08fef6f0c4b252f958ef0c89 100644
--- a/test_tipc/configs/mask_rcnn/mask_rcnn_r101_vd_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/mask_rcnn/mask_rcnn_r101_vd_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/mask_rcnn_r101_vd_fpn_1x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/mask_rcnn/mask_rcnn_r101_vd_fpn_1x_coco.yml -o
diff --git a/test_tipc/configs/mask_rcnn/mask_rcnn_r50_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/mask_rcnn/mask_rcnn_r50_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index ede384570137429eb9e8568d99cd595e51966fa1..843ac40e9cb4722f264cb6b56774c00aef616a53 100644
--- a/test_tipc/configs/mask_rcnn/mask_rcnn_r50_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/mask_rcnn/mask_rcnn_r50_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/mask_rcnn_r50_1x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/mask_rcnn/mask_rcnn_r50_1x_coco.yml -o
diff --git a/test_tipc/configs/mask_rcnn/mask_rcnn_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/mask_rcnn/mask_rcnn_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 46447daad963c125f206e938656a8fed56867772..af061e0559103607d4ad774be1c500ac82b5c030 100644
--- a/test_tipc/configs/mask_rcnn/mask_rcnn_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/mask_rcnn/mask_rcnn_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/mask_rcnn_r50_fpn_1x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/mask_rcnn/mask_rcnn_r50_fpn_1x_coco.yml -o
diff --git a/test_tipc/configs/mask_rcnn/mask_rcnn_r50_vd_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/mask_rcnn/mask_rcnn_r50_vd_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 56d3e26121831ab068101c2c191c405a3cd4db17..c2cc1429efcde5502d8c32fc713a348167797a10 100644
--- a/test_tipc/configs/mask_rcnn/mask_rcnn_r50_vd_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/mask_rcnn/mask_rcnn_r50_vd_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/mask_rcnn_r50_vd_fpn_1x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/mask_rcnn/mask_rcnn_r50_vd_fpn_1x_coco.yml -o
diff --git a/test_tipc/configs/mask_rcnn/mask_rcnn_x101_vd_64x4d_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/mask_rcnn/mask_rcnn_x101_vd_64x4d_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 705f0b3ed50e40efff9a0b57f634369fb14f7ce2..d2c226a1faa6ae7dbb3c4d0018354616de6a092e 100644
--- a/test_tipc/configs/mask_rcnn/mask_rcnn_x101_vd_64x4d_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/mask_rcnn/mask_rcnn_x101_vd_64x4d_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/mask_rcnn_x101_vd_64x4d_fpn_1x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/mask_rcnn/mask_rcnn_x101_vd_64x4d_fpn_1x_coco.yml -o
diff --git a/test_tipc/configs/mot/fairmot_dla34_30e_1088x608_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/mot/fairmot_dla34_30e_1088x608_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index db624ce64c15e26fc748ccb1846e0df657e04613..914b74175545a32c9d7da9e6b87e190c4e6d6311 100644
--- a/test_tipc/configs/mot/fairmot_dla34_30e_1088x608_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/mot/fairmot_dla34_30e_1088x608_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/mot/fairmot_dla34_30e_1088x608.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/mot/test.mp4
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/mot/fairmot/fairmot_dla34_30e_1088x608.yml -o
diff --git a/test_tipc/configs/mot/fairmot_dla34_30e_576x320_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/mot/fairmot_dla34_30e_576x320_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 0de7ed6ce4dabf0e23bdb35f938b8f4c645a2b7d..2cb87aaa28933b7ee6ed1eb2753cb93115f85794 100644
--- a/test_tipc/configs/mot/fairmot_dla34_30e_576x320_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/mot/fairmot_dla34_30e_576x320_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/mot/fairmot_dla34_30e_576x320.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/mot/test.mp4
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/mot/fairmot/fairmot_dla34_30e_576x320.yml -o
diff --git a/test_tipc/configs/mot/fairmot_hrnetv2_w18_dlafpn_30e_576x320_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/mot/fairmot_hrnetv2_w18_dlafpn_30e_576x320_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index bc4e37e6841d942918fd5d13a1fd0be5ec1b4176..f4407bd75f577f3ed40d3fecd56f446a7aa8dd45 100644
--- a/test_tipc/configs/mot/fairmot_hrnetv2_w18_dlafpn_30e_576x320_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/mot/fairmot_hrnetv2_w18_dlafpn_30e_576x320_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/mot/fairmot_hrnetv2_w18_dlafpn_30e_576x320.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/mot/test.mp4
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c test_tipc/configs/mot/fairmot_hrnetv2_w18_dlafpn_30e_576x320.yml -o
diff --git a/test_tipc/configs/mot/jde_darknet53_30e_1088x608_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/mot/jde_darknet53_30e_1088x608_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 7decaedc0d9dd077596cdb4759e0e7767339f066..45b8d17c5a3aecf8abb4eaa3450e6afd705ff696 100644
--- a/test_tipc/configs/mot/jde_darknet53_30e_1088x608_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/mot/jde_darknet53_30e_1088x608_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/mot/jde_darknet53_30e_1088x608.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/mot/test.mp4
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/mot/jde/jde_darknet53_30e_1088x608.yml -o
diff --git a/test_tipc/configs/mot/jde_darknet53_30e_576x320_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/mot/jde_darknet53_30e_576x320_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 18a0549f5c4493505ee785dc2f0e4e7520f06565..40d98490feb9e9c53df1e4404759a0305550d136 100644
--- a/test_tipc/configs/mot/jde_darknet53_30e_576x320_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/mot/jde_darknet53_30e_576x320_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/mot/jde_darknet53_30e_576x320.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/mot/test.mp4
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/mot/jde/jde_darknet53_30e_576x320.yml -o
diff --git a/test_tipc/configs/picodet/picodet_l_640_coco_lcnet_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/picodet/picodet_l_640_coco_lcnet_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index a2aa277cf69de061157b89cf89fc283e123605b7..4fb5661f5ecd3e4e992db6fe8306418b4bbae14b 100644
--- a/test_tipc/configs/picodet/picodet_l_640_coco_lcnet_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/picodet/picodet_l_640_coco_lcnet_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/picodet_l_640_coco_lcnet.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/picodet/picodet_l_640_coco_lcnet.yml -o
diff --git a/test_tipc/configs/picodet/picodet_l_640_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/picodet/picodet_l_640_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index c8ba16da1dff8928ea5e38edb8af6161d14b8b07..6b54d654369b46e2c5583d39ace3f2e1a476a390 100644
--- a/test_tipc/configs/picodet/picodet_l_640_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/picodet/picodet_l_640_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/picodet_l_640_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/picodet/legacy_model/picodet_l_640_coco.yml -o
diff --git a/test_tipc/configs/picodet/picodet_lcnet_1_5x_416_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/picodet/picodet_lcnet_1_5x_416_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index ba0d31ccf34c06558b9de4538dac2d0aacdd69ce..28272ad589218350c25b58d9750de335518e4d24 100644
--- a/test_tipc/configs/picodet/picodet_lcnet_1_5x_416_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/picodet/picodet_lcnet_1_5x_416_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/picodet_lcnet_1_5x_416_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/picodet/legacy_model/more_config/picodet_lcnet_1_5x_416_coco.yml -o
diff --git a/test_tipc/configs/picodet/picodet_m_416_coco_lcnet_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/picodet/picodet_m_416_coco_lcnet_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 200c0e448303a30b56100fc55ff39fed3c51258f..b347c375ff6c187614b9a428b4881aa6214711f8 100644
--- a/test_tipc/configs/picodet/picodet_m_416_coco_lcnet_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/picodet/picodet_m_416_coco_lcnet_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/picodet_m_416_coco_lcnet.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/picodet/picodet_m_416_coco_lcnet.yml -o
diff --git a/test_tipc/configs/picodet/picodet_m_416_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/picodet/picodet_m_416_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 5e1507633853bd843f2a584719244e9659a601ec..8d99e83d695ce056c95d723b5ce90fc814f953b4 100644
--- a/test_tipc/configs/picodet/picodet_m_416_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/picodet/picodet_m_416_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/picodet_m_416_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/picodet/legacy_model/picodet_m_416_coco.yml -o
diff --git a/test_tipc/configs/picodet/picodet_mobilenetv3_large_1x_416_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/picodet/picodet_mobilenetv3_large_1x_416_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index e9fe5361c049b0225f2d8ab6c0de738990944b67..3db11e54f39e00bdc0c17b9537c241aa0dac1b3c 100644
--- a/test_tipc/configs/picodet/picodet_mobilenetv3_large_1x_416_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/picodet/picodet_mobilenetv3_large_1x_416_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/picodet_mobilenetv3_large_1x_416_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/picodet/legacy_model/more_config/picodet_mobilenetv3_large_1x_416_coco.yml -o
diff --git a/test_tipc/configs/picodet/picodet_r18_640_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/picodet/picodet_r18_640_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 87997a3a51c430e2f64a7fd26f3d9d3f27fd1e23..8a5e487937ea348c6af435b2f6fa1d0d7aad5e9d 100644
--- a/test_tipc/configs/picodet/picodet_r18_640_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/picodet/picodet_r18_640_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/picodet_r18_640_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/picodet/legacy_model/more_config/picodet_r18_640_coco.yml -o
diff --git a/test_tipc/configs/picodet/picodet_s_320_coco_lcnet_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/picodet/picodet_s_320_coco_lcnet_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 776d02ed438e8bba45a57c533bbdc4b910cb05cf..cac5a70bba188654ad61483f44d4c61510c04646 100644
--- a/test_tipc/configs/picodet/picodet_s_320_coco_lcnet_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/picodet/picodet_s_320_coco_lcnet_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/picodet_s_320_coco_lcnet.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/picodet/picodet_s_320_coco_lcnet.yml -o
diff --git a/test_tipc/configs/picodet/picodet_s_320_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/picodet/picodet_s_320_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 5765a14582f7f60b8cf23576a268546e1e6ebd38..b1a87cdd7bf5f5219d91d2c660c5183837419d99 100644
--- a/test_tipc/configs/picodet/picodet_s_320_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/picodet/picodet_s_320_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/picodet_s_320_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/picodet/legacy_model/picodet_s_320_coco.yml -o
diff --git a/test_tipc/configs/picodet/picodet_shufflenetv2_1x_416_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/picodet/picodet_shufflenetv2_1x_416_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 72cc4dc0dbe989e71df628becb8839a70d7fd90b..065abae6679b832014227cf2262d830e08f907b0 100644
--- a/test_tipc/configs/picodet/picodet_shufflenetv2_1x_416_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/picodet/picodet_shufflenetv2_1x_416_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/picodet_shufflenetv2_1x_416_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/picodet/legacy_model/more_config/picodet_shufflenetv2_1x_416_coco.yml -o
diff --git a/test_tipc/configs/picodet/picodet_xs_320_coco_lcnet_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/picodet/picodet_xs_320_coco_lcnet_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 89934723b78cc5807f13b1f61320f95045289f44..c84d14c384d4980d3696b70dd3ae71a610777022 100644
--- a/test_tipc/configs/picodet/picodet_xs_320_coco_lcnet_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/picodet/picodet_xs_320_coco_lcnet_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/picodet_xs_320_coco_lcnet.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/picodet/picodet_xs_320_coco_lcnet.yml -o
diff --git a/test_tipc/configs/ppyolo/ppyolo_mbv3_large_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ppyolo/ppyolo_mbv3_large_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index ea56d0c6e433840f6b6b77b8e500b5e8d57f2078..6ee39190da3fc44687386dc5c17246cd5f6ad702 100644
--- a/test_tipc/configs/ppyolo/ppyolo_mbv3_large_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ppyolo/ppyolo_mbv3_large_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/ppyolo_mbv3_large_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/ppyolo/ppyolo_mbv3_large_coco.yml -o
diff --git a/test_tipc/configs/ppyolo/ppyolo_mbv3_small_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ppyolo/ppyolo_mbv3_small_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 75a1903d5d5bd087394a5bcaee5802571c98f5af..ae3322eeec948bf8ccaab6090c6a193e0c1d4056 100644
--- a/test_tipc/configs/ppyolo/ppyolo_mbv3_small_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ppyolo/ppyolo_mbv3_small_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/ppyolo_mbv3_small_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/ppyolo/ppyolo_mbv3_small_coco.yml -o
diff --git a/test_tipc/configs/ppyolo/ppyolo_r18vd_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ppyolo/ppyolo_r18vd_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 1adeaf4b490be0ecd0a76bad3c2a834101ecd46b..fd2ef0dcaefdc7e58234a760a8746d00d5d95588 100644
--- a/test_tipc/configs/ppyolo/ppyolo_r18vd_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ppyolo/ppyolo_r18vd_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/ppyolo_r18vd_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/ppyolo/ppyolo_r18vd_coco.yml -o
diff --git a/test_tipc/configs/ppyolo/ppyolo_r50vd_dcn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ppyolo/ppyolo_r50vd_dcn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 3208f48fffa4c505838da1f8fedc5e702e4d9076..6a4235a0bb89c33093955c7786058f934c343ff6 100644
--- a/test_tipc/configs/ppyolo/ppyolo_r50vd_dcn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ppyolo/ppyolo_r50vd_dcn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/ppyolo_r50vd_dcn_1x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/ppyolo/ppyolo_r50vd_dcn_1x_coco.yml -o
diff --git a/test_tipc/configs/ppyolo/ppyolo_tiny_650e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ppyolo/ppyolo_tiny_650e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index caabd60b3a6a6032bd1c13c62cb745d4e87ecc53..cb0659a7bdb8398ae4f29072b9acbbed7191e110 100644
--- a/test_tipc/configs/ppyolo/ppyolo_tiny_650e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ppyolo/ppyolo_tiny_650e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/ppyolo_tiny_650e_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/ppyolo/ppyolo_tiny_650e_coco.yml -o
diff --git a/test_tipc/configs/ppyolo/ppyolov2_r101vd_dcn_365e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ppyolo/ppyolov2_r101vd_dcn_365e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index d42bdbaaa59b4a1306359e6d0bc54e2a6acb65ca..fc2ca676860301d2fb9bfea261d0330bdde5a56a 100644
--- a/test_tipc/configs/ppyolo/ppyolov2_r101vd_dcn_365e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ppyolo/ppyolov2_r101vd_dcn_365e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/ppyolov2_r101vd_dcn_365e_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/ppyolo/ppyolov2_r101vd_dcn_365e_coco.yml -o
diff --git a/test_tipc/configs/ppyolo/ppyolov2_r50vd_dcn_365e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ppyolo/ppyolov2_r50vd_dcn_365e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 3c520c9e0d027aef12a9cfaea038b250f0cd079d..721db09e03f235ad0a1064e2f2f6853ab9fab5aa 100644
--- a/test_tipc/configs/ppyolo/ppyolov2_r50vd_dcn_365e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ppyolo/ppyolov2_r50vd_dcn_365e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/ppyolov2_r50vd_dcn_365e_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/ppyolo/ppyolov2_r50vd_dcn_365e_coco.yml -o
diff --git a/test_tipc/configs/ppyoloe/ppyoloe_crn_l_300e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ppyoloe/ppyoloe_crn_l_300e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index b6f75f7bd29ebc8f0fd4b9b1b60d46d2b8485695..3df9e020b206e202603b4a64f52d828e4d02fee1 100644
--- a/test_tipc/configs/ppyoloe/ppyoloe_crn_l_300e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ppyoloe/ppyoloe_crn_l_300e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/ppyoloe_crn_l_300e_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/ppyoloe/ppyoloe_crn_l_300e_coco.yml -o
diff --git a/test_tipc/configs/ppyoloe/ppyoloe_crn_m_300e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ppyoloe/ppyoloe_crn_m_300e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index aafe6a093993f060755d321f8e51dd41c4012b50..607113b9d58d7de9f3af9e522ab4c3b91bc9d436 100644
--- a/test_tipc/configs/ppyoloe/ppyoloe_crn_m_300e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ppyoloe/ppyoloe_crn_m_300e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/ppyoloe_crn_m_300e_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/ppyoloe/ppyoloe_crn_m_300e_coco.yml -o
diff --git a/test_tipc/configs/ppyoloe/ppyoloe_crn_s_300e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ppyoloe/ppyoloe_crn_s_300e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 5ea846972dfc412f1a67afe9bacf6c8c48c9887c..815a2393742957b371bf59fd1a9efb0c92745796 100644
--- a/test_tipc/configs/ppyoloe/ppyoloe_crn_s_300e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ppyoloe/ppyoloe_crn_s_300e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/ppyoloe_crn_s_300e_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/ppyoloe/ppyoloe_crn_s_300e_coco.yml -o
diff --git a/test_tipc/configs/ppyoloe/ppyoloe_crn_x_300e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ppyoloe/ppyoloe_crn_x_300e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 7ed12a7a82a5d1dc9ebad79a7987f33f266eaaef..38f2fcc8fb9f6d00f6691c1e9e92aefe374eda51 100644
--- a/test_tipc/configs/ppyoloe/ppyoloe_crn_x_300e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ppyoloe/ppyoloe_crn_x_300e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/ppyoloe_crn_x_300e_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/ppyoloe/ppyoloe_crn_x_300e_coco.yml -o
diff --git a/test_tipc/configs/retinanet/retinanet_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/retinanet/retinanet_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 76a6b9e57c9234ab93f64e0a13575059eafc60de..585c99f38e95587349a2fbc8a1807ddb2819974f 100644
--- a/test_tipc/configs/retinanet/retinanet_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/retinanet/retinanet_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/retinanet_r50_fpn_1x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/retinanet/retinanet_r50_fpn_1x_coco.yml -o
diff --git a/test_tipc/configs/solov2/solov2_r101_vd_fpn_3x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/solov2/solov2_r101_vd_fpn_3x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 142d0354333077a41798f9f8ac516f3b94785a81..df50fd3c40774077fdac52fde7034ff0abb35b7f 100644
--- a/test_tipc/configs/solov2/solov2_r101_vd_fpn_3x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/solov2/solov2_r101_vd_fpn_3x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/solov2_r101_vd_fpn_3x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/solov2/solov2_r101_vd_fpn_3x_coco.yml -o
diff --git a/test_tipc/configs/solov2/solov2_r50_enhance_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/solov2/solov2_r50_enhance_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 4195c11a48cf58374239d07088105642745fadb4..e2b43e45fa3af3a323385de126611854252c643f 100644
--- a/test_tipc/configs/solov2/solov2_r50_enhance_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/solov2/solov2_r50_enhance_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/solov2_r50_enhance_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/solov2/solov2_r50_enhance_coco.yml -o
diff --git a/test_tipc/configs/solov2/solov2_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/solov2/solov2_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 41ed5ef46338387caf56a16d3542ec19b47ca81f..3d00504021d868613cccebd4ca039d5d3f8b0dd1 100644
--- a/test_tipc/configs/solov2/solov2_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/solov2/solov2_r50_fpn_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/solov2_r50_fpn_1x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/solov2/solov2_r50_fpn_1x_coco.yml -o
diff --git a/test_tipc/configs/sparse_rcnn/sparse_rcnn_r50_fpn_3x_pro100_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/sparse_rcnn/sparse_rcnn_r50_fpn_3x_pro100_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index f4dbf76f5ddbbcbd37d58a90b2cedcf7328525a0..a461713b3441ea786a05ba383f31607e86f75f55 100644
--- a/test_tipc/configs/sparse_rcnn/sparse_rcnn_r50_fpn_3x_pro100_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/sparse_rcnn/sparse_rcnn_r50_fpn_3x_pro100_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/sparse_rcnn_r50_fpn_3x_pro100_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/sparse_rcnn/sparse_rcnn_r50_fpn_3x_pro100_coco.yml -o
diff --git a/test_tipc/configs/ssd/ssdlite_mobilenet_v1_300_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ssd/ssdlite_mobilenet_v1_300_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index ec50a05a88a9e2e2c6ac87acf786fa671c4986b4..8d2036f55fdff4bfc92e091851668a0a433a8473 100644
--- a/test_tipc/configs/ssd/ssdlite_mobilenet_v1_300_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ssd/ssdlite_mobilenet_v1_300_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/ssdlite_mobilenet_v1_300_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/ssd/ssdlite_mobilenet_v1_300_coco.yml -o
diff --git a/test_tipc/configs/ttfnet/ttfnet_darknet53_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/ttfnet/ttfnet_darknet53_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 1d33fe8124a2a98d563a346a53f1f489816db44d..a167ce93d8f874a16c2a97f6d2441f9cbc8ba226 100644
--- a/test_tipc/configs/ttfnet/ttfnet_darknet53_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/ttfnet/ttfnet_darknet53_1x_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/ttfnet_darknet53_1x_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/ttfnet/ttfnet_darknet53_1x_coco.yml -o
diff --git a/test_tipc/configs/yolov3/yolov3_darknet53_270e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/yolov3/yolov3_darknet53_270e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index cfe8534bb1036a866cffe2768de595868698f0d8..b3142b6eac8b046c55c11ccd674612c743ea5d45 100644
--- a/test_tipc/configs/yolov3/yolov3_darknet53_270e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/yolov3/yolov3_darknet53_270e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/yolov3_darknet53_270e_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/yolov3/yolov3_darknet53_270e_coco.yml -o
diff --git a/test_tipc/configs/yolov3/yolov3_mobilenet_v1_270e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/yolov3/yolov3_mobilenet_v1_270e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 6f9cb142400e4ad66b0c502a26a28fb357a1506e..5e9d35f2a11ff89b8d8400590682fb110f763c22 100644
--- a/test_tipc/configs/yolov3/yolov3_mobilenet_v1_270e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/yolov3/yolov3_mobilenet_v1_270e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/yolov3_mobilenet_v1_270e_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/yolov3/yolov3_mobilenet_v1_270e_coco.yml -o
diff --git a/test_tipc/configs/yolov3/yolov3_mobilenet_v3_large_270e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/yolov3/yolov3_mobilenet_v3_large_270e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index be099dd9d24f54604ae4f6524b1e12a4f719cd1b..49668ed6cc0c6f2440d7b59cc55d51fb7f1e93e8 100644
--- a/test_tipc/configs/yolov3/yolov3_mobilenet_v3_large_270e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/yolov3/yolov3_mobilenet_v3_large_270e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/yolov3_mobilenet_v3_large_270e_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/yolov3/yolov3_mobilenet_v3_large_270e_coco.yml -o
diff --git a/test_tipc/configs/yolov3/yolov3_r34_270e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/yolov3/yolov3_r34_270e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index 667b6c86543881c66a156fe30ab5a289a48ab670..6b8720cdc1aa7647b8ba498ddcd2c9742607d818 100644
--- a/test_tipc/configs/yolov3/yolov3_r34_270e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/yolov3/yolov3_r34_270e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/yolov3_r34_270e_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/yolov3/yolov3_r34_270e_coco.yml -o
diff --git a/test_tipc/configs/yolov3/yolov3_r50vd_dcn_270e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt b/test_tipc/configs/yolov3/yolov3_r50vd_dcn_270e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
index a1d20bab82d5b969c69d452c9c04c3cccfbba8a2..4282bb38c302acf8a1d2300c67060d6e924139ac 100644
--- a/test_tipc/configs/yolov3/yolov3_r50vd_dcn_270e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+++ b/test_tipc/configs/yolov3/yolov3_r50vd_dcn_270e_coco_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
@@ -10,7 +10,7 @@ TrainReader.batch_size:lite_train_lite_infer=2|lite_train_whole_infer=2|whole_tr
 pretrain_weights:https://paddledet.bj.bcebos.com/models/yolov3_r50vd_dcn_270e_coco.pdparams
 trained_model_name:model_final.pdparams
 train_infer_img_dir:./dataset/coco/test2017/
-filename:null
+amp_level:O2
 ##
 trainer:norm_train
 norm_train:tools/train.py -c configs/yolov3/yolov3_r50vd_dcn_270e_coco.yml -o