[Serving] Simple serving YOLOv5 and PP-OCRv3 example, add uvicorn to fastdeploy tools (#986)

* ppocrv3 simple serving * add uvicorn to fd tools * update ppdet simple serving readme * yolov5 simple serving * not import simple serving by default * remove config from envs * update comment
2025-10-06 00:57:33 +08:00 · 2022-12-28 10:03:42 +08:00
parent 0ead9d27c2
commit aea454a856
18 changed files with 364 additions and 42 deletions
--- a/examples/vision/ocr/PP-OCRv3/python/serving/README.md
+++ b/examples/vision/ocr/PP-OCRv3/python/serving/README.md
@@ -0,0 +1 @@
+README_CN.md
--- a/examples/vision/ocr/PP-OCRv3/python/serving/README_CN.md
+++ b/examples/vision/ocr/PP-OCRv3/python/serving/README_CN.md
@@ -0,0 +1,44 @@
+简体中文 | [English](README_EN.md)
+
+# PP-OCRv3 Python轻量服务化部署示例
+
+在部署前，需确认以下两个步骤
+
+- 1. 软硬件环境满足要求，参考[FastDeploy环境要求](../../../../../../docs/cn/build_and_install/download_prebuilt_libraries.md)  
+- 2. FastDeploy Python whl包安装，参考[FastDeploy Python安装](../../../../../../docs/cn/build_and_install/download_prebuilt_libraries.md)
+
+服务端：
+```bash
+# 下载部署示例代码
+git clone https://github.com/PaddlePaddle/FastDeploy.git
+cd FastDeploy/examples/vision/ocr/PP-OCRv3/python/serving
+
+# 下载模型和字典文件
+wget https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_det_infer.tar
+tar xvf ch_PP-OCRv3_det_infer.tar
+
+wget https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_infer.tar
+tar -xvf ch_ppocr_mobile_v2.0_cls_infer.tar
+
+wget https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_infer.tar
+tar xvf ch_PP-OCRv3_rec_infer.tar
+
+wget https://gitee.com/paddlepaddle/PaddleOCR/raw/release/2.6/ppocr/utils/ppocr_keys_v1.txt
+
+# 启动服务，可修改server.py中的配置项来指定硬件、后端等
+# 可通过--host、--port指定IP和端口号
+fastdeploy simple_serving --app server:app
+```
+
+客户端：
+```bash
+# 下载部署示例代码
+git clone https://github.com/PaddlePaddle/FastDeploy.git
+cd FastDeploy/examples/vision/ocr/PP-OCRv3/python/serving
+
+# 下载测试图片
+wget https://gitee.com/paddlepaddle/PaddleOCR/raw/release/2.6/doc/imgs/12.jpg
+
+# 请求服务，获取推理结果（如有必要，请修改脚本中的IP和端口号）
+python client.py
+```
--- a/examples/vision/ocr/PP-OCRv3/python/serving/README_EN.md
+++ b/examples/vision/ocr/PP-OCRv3/python/serving/README_EN.md
@@ -0,0 +1,43 @@
+English | [简体中文](README_CN.md)
+
+# PP-OCRv3 Python Simple Serving Demo
+
+## Environment
+
+- 1. Prepare environment and install FastDeploy Python whl, refer to [download_prebuilt_libraries](../../../../../../docs/en/build_and_install/download_prebuilt_libraries.md)
+
+Server:
+```bash
+# Download demo code
+git clone https://github.com/PaddlePaddle/FastDeploy.git
+cd FastDeploy/examples/vision/ocr/PP-OCRv3/python/serving
+
+# Download models and labels
+wget https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_det_infer.tar
+tar xvf ch_PP-OCRv3_det_infer.tar
+
+wget https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_infer.tar
+tar -xvf ch_ppocr_mobile_v2.0_cls_infer.tar
+
+wget https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_infer.tar
+tar xvf ch_PP-OCRv3_rec_infer.tar
+
+wget https://gitee.com/paddlepaddle/PaddleOCR/raw/release/2.6/ppocr/utils/ppocr_keys_v1.txt
+
+# Launch server, change the configurations in server.py to select hardware, backend, etc.
+# and use --host, --port to specify IP and port
+fastdeploy simple_serving --app server:app
+```
+
+Client:
+```bash
+# Download demo code
+git clone https://github.com/PaddlePaddle/FastDeploy.git
+cd FastDeploy/examples/vision/ocr/PP-OCRv3/python/serving
+
+# Download test image
+wget https://gitee.com/paddlepaddle/PaddleOCR/raw/release/2.6/doc/imgs/12.jpg
+
+# Send request and get inference result (Please adapt the IP and port if necessary)
+python client.py
+```
--- a/examples/vision/ocr/PP-OCRv3/python/serving/client.py
+++ b/examples/vision/ocr/PP-OCRv3/python/serving/client.py
@@ -0,0 +1,24 @@
+import requests
+import json
+import cv2
+import fastdeploy as fd
+from fastdeploy.serving.utils import cv2_to_base64
+
+if __name__ == '__main__':
+    url = "http://127.0.0.1:8000/fd/ppocrv3"
+    headers = {"Content-Type": "application/json"}
+
+    im = cv2.imread("12.jpg")
+    data = {"data": {"image": cv2_to_base64(im)}, "parameters": {}}
+
+    resp = requests.post(url=url, headers=headers, data=json.dumps(data))
+    if resp.status_code == 200:
+        r_json = json.loads(resp.json()["result"])
+        print(r_json)
+        ocr_result = fd.vision.utils.json_to_ocr(r_json)
+        vis_im = fd.vision.vis_ppocr(im, ocr_result)
+        cv2.imwrite("visualized_result.jpg", vis_im)
+        print("Visualized result save in ./visualized_result.jpg")
+    else:
+        print("Error code:", resp.status_code)
+        print(resp.text)
--- a/examples/vision/ocr/PP-OCRv3/python/serving/server.py
+++ b/examples/vision/ocr/PP-OCRv3/python/serving/server.py
@@ -0,0 +1,80 @@
+import fastdeploy as fd
+from fastdeploy.serving.server import SimpleServer
+import os
+import logging
+
+logging.getLogger().setLevel(logging.INFO)
+
+# Configurations
+det_model_dir = 'ch_PP-OCRv3_det_infer'
+cls_model_dir = 'ch_ppocr_mobile_v2.0_cls_infer'
+rec_model_dir = 'ch_PP-OCRv3_rec_infer'
+rec_label_file = 'ppocr_keys_v1.txt'
+device = 'cpu'
+# backend: ['paddle', 'trt'], you can also use other backends, but need to modify
+# the runtime option below
+backend = 'paddle'
+
+# Prepare models
+# Detection model
+det_model_file = os.path.join(det_model_dir, "inference.pdmodel")
+det_params_file = os.path.join(det_model_dir, "inference.pdiparams")
+# Classification model
+cls_model_file = os.path.join(cls_model_dir, "inference.pdmodel")
+cls_params_file = os.path.join(cls_model_dir, "inference.pdiparams")
+# Recognition model
+rec_model_file = os.path.join(rec_model_dir, "inference.pdmodel")
+rec_params_file = os.path.join(rec_model_dir, "inference.pdiparams")
+
+# Setup runtime option to select hardware, backend, etc.
+option = fd.RuntimeOption()
+if device.lower() == 'gpu':
+    option.use_gpu()
+if backend == 'trt':
+    option.use_trt_backend()
+else:
+    option.use_paddle_infer_backend()
+
+det_option = option
+det_option.set_trt_input_shape("x", [1, 3, 64, 64], [1, 3, 640, 640],
+                               [1, 3, 960, 960])
+
+# det_option.set_trt_cache_file("det_trt_cache.trt")
+print(det_model_file, det_params_file)
+det_model = fd.vision.ocr.DBDetector(
+    det_model_file, det_params_file, runtime_option=det_option)
+
+cls_batch_size = 1
+rec_batch_size = 6
+
+cls_option = option
+cls_option.set_trt_input_shape("x", [1, 3, 48, 10],
+                               [cls_batch_size, 3, 48, 320],
+                               [cls_batch_size, 3, 48, 1024])
+
+# cls_option.set_trt_cache_file("cls_trt_cache.trt")
+cls_model = fd.vision.ocr.Classifier(
+    cls_model_file, cls_params_file, runtime_option=cls_option)
+
+rec_option = option
+rec_option.set_trt_input_shape("x", [1, 3, 48, 10],
+                               [rec_batch_size, 3, 48, 320],
+                               [rec_batch_size, 3, 48, 2304])
+
+# rec_option.set_trt_cache_file("rec_trt_cache.trt")
+rec_model = fd.vision.ocr.Recognizer(
+    rec_model_file, rec_params_file, rec_label_file, runtime_option=rec_option)
+
+# Create PPOCRv3 pipeline
+ppocr_v3 = fd.vision.ocr.PPOCRv3(
+    det_model=det_model, cls_model=cls_model, rec_model=rec_model)
+
+ppocr_v3.cls_batch_size = cls_batch_size
+ppocr_v3.rec_batch_size = rec_batch_size
+
+# Create server, setup REST API
+app = SimpleServer()
+app.register(
+    task_name="fd/ppocrv3",
+    model_handler=fd.serving.handler.VisionModelHandler,
+    predictor=ppocr_v3)