关于模型部署的问题: paddle serving error

[fatsheep2020]

问题确认 Search before asking

• 我已经查询历史issue，没有发现相似的bug。I have searched the issues and found no similar bug report.

Bug组件 Bug Component

Deploy

Bug描述 Describe the Bug

按照 ~/deploy/README.md 中的指导导出paddleserving格式的模型：
python tools/export_model.py -c configs/yolov3/yolov3_mobilenet_v1_roadsign.yml -o weights=output/yolov3_mobilenet_v1_roadsign/best_model.pdparams --export_serving_model=True
会报错：
Traceback (most recent call last):
File "tools/export_model.py", line 118, in
main()
File "tools/export_model.py", line 114, in main
run(FLAGS, cfg)
File "tools/export_model.py", line 84, in run
from paddle_serving_client.io import inference_model_to_serving
File "/home/wade/.local/share/virtualenvs/3.paddle-l301jT_O/lib/python3.8/site-packages/paddle_serving_client/io/init.py", line 16, in
from paddle.fluid import Executor
ModuleNotFoundError: No module named 'paddle.fluid'

请问：现在云端部署还是用paddle-serving吗？还是用其他方法？
最近在做这方面的一个项目，打算用paddle做云端部署，还不知道改怎么弄，哪里有相关的文档？
比较着急，谢谢！

复现环境 Environment

ubuntu 20
paddle: 2.6.2
运行上述命令

Bug描述确认 Bug description confirmation

• 我确认已经提供了Bug复现步骤、代码改动说明、以及环境信息，确认问题是可以复现的。I confirm that the bug replication steps, code change instructions, and environment information have been provided, and the problem can be reproduced.

是否愿意提交PR？ Are you willing to submit a PR?

• 我愿意提交PR！I'd like to help by submitting a PR!

[liuhongen1234567]

您好，paddle的fluid 应该是比较老的paddle版本中采用的算子，可以尝试一些2.x的paddle版本。最新的部署方式建议您看一下 PaddleX文档的 3. 开发集成/部署部分。这个paddleX 中 paddle detection模型的文档 https://github.com/PaddlePaddle/PaddleX/blob/develop/docs/pipeline_usage/tutorials/cv_pipelines/object_detection.md

[fatsheep2020]

您好！ 我尝试回退paddle到 2.4.0下面， 再次运行上面的export_model.py 命令，发现还是报错：
xxx-OMEN:~/1.project/1.YOLO/3.paddle/PaddleDetection$ python export_model_test.py -c configs/yolov3/yolov3_darknet53_270e_coco.yml --export_serving_model True -o weights=https://paddledet.bj.bcebos.com/models/yolov3_darknet53_270e_coco.pdparams
/home/xxx/.local/share/virtualenvs/3.paddle-l301jT_O/lib/python3.8/site-packages/setuptools/command/easy_install.py:41: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html
import pkg_resources
/home/xxx/.local/share/virtualenvs/3.paddle-l301jT_O/lib/python3.8/site-packages/pkg_resources/init.py:3154: DeprecationWarning: Deprecated call to pkg_resources.declare_namespace('mpl_toolkits').
Implementing implicit namespace packages (as specified in PEP 420) is preferred to pkg_resources.declare_namespace. See https://setuptools.pypa.io/en/latest/references/keywords.html#keyword-namespace-packages
declare_namespace(pkg)
/home/xxx/.local/share/virtualenvs/3.paddle-l301jT_O/lib/python3.8/site-packages/pkg_resources/init.py:3154: DeprecationWarning: Deprecated call to pkg_resources.declare_namespace('google').
Implementing implicit namespace packages (as specified in PEP 420) is preferred to pkg_resources.declare_namespace. See https://setuptools.pypa.io/en/latest/references/keywords.html#keyword-namespace-packages
declare_namespace(pkg)
Traceback (most recent call last):
File "export_model_test.py", line 31, in
from ppdet.core.workspace import load_config, merge_config
File "/home/wade/1.project/1.YOLO/3.paddle/PaddleDetection/ppdet/init.py", line 15, in
from . import (core, data, engine, modeling, model_zoo, optimizer, metrics,
File "/home/wade/1.project/1.YOLO/3.paddle/PaddleDetection/ppdet/data/init.py", line 16, in
from . import transform
File "/home/wade/1.project/1.YOLO/3.paddle/PaddleDetection/ppdet/data/transform/init.py", line 15, in
from . import operators
File "/home/wade/1.project/1.YOLO/3.paddle/PaddleDetection/ppdet/data/transform/operators.py", line 57, in
from ppdet.modeling.keypoint_utils import get_affine_transform, affine_transform
File "/home/wade/1.project/1.YOLO/3.paddle/PaddleDetection/ppdet/modeling/init.py", line 23, in
from . import heads
File "/home/wade/1.project/1.YOLO/3.paddle/PaddleDetection/ppdet/modeling/heads/init.py", line 17, in
from . import yolo_head
File "/home/wade/1.project/1.YOLO/3.paddle/PaddleDetection/ppdet/modeling/heads/yolo_head.py", line 26, in
from ..losses import IouLoss
File "/home/wade/1.project/1.YOLO/3.paddle/PaddleDetection/ppdet/modeling/losses/init.py", line 34, in
from . import clrnet_loss
File "/home/wade/1.project/1.YOLO/3.paddle/PaddleDetection/ppdet/modeling/losses/clrnet_loss.py", line 6, in
from ppdet.modeling.assigners.clrnet_assigner import assign
File "/home/wade/1.project/1.YOLO/3.paddle/PaddleDetection/ppdet/modeling/assigners/init.py", line 18, in
from . import simota_assigner
File "/home/wade/1.project/1.YOLO/3.paddle/PaddleDetection/ppdet/modeling/assigners/simota_assigner.py", line 22, in
from ppdet.modeling.losses.varifocal_loss import varifocal_loss
File "/home/wade/1.project/1.YOLO/3.paddle/PaddleDetection/ppdet/modeling/losses/varifocal_loss.py", line 27, in
from paddle.base.framework import in_dygraph_mode
ModuleNotFoundError: No module named 'paddle.base'

[fatsheep2020]

上面的错误不知道是不是安装环境的问题？ 还是export_model.py脚本的问题？

Paddle X的部署我学习一下，然后再告诉你结果。

[fatsheep2020]

关于paddleX，我在飞桨的AI Studio 环境下安装好了paddle和paddlex但是运行paddlex还是会出错，是什么原因？

谢谢！
@liuhongen1234567
按照说明，启动paddlex服务，要用这个命令

[liuhongen1234567]

!pip install --user paddlex==3.0.0b2，然后AI studio 重启一下内核呢？

[fatsheep2020]

!pip install --user paddlex==3.0.0b2，然后AI studio 重启一下内核呢？

还是一样报错，重启了两次, 你那边可以试一下，这个命令不麻烦，安装也快. Paddle version 3.0.0-beta2

[fatsheep2020]

paddle 版本在2.4下面是不是paddle serving也报错？ paddle serving是不是不建议使用了？
@liuhonggen1234567

[fatsheep2020]

您好，paddle的fluid 应该是比较老的paddle版本中采用的算子，可以尝试一些2.x的paddle版本。最新的部署方式建议您看一下 PaddleX文档的 3. 开发集成/部署部分。这个paddleX 中 paddle detection模型的文档 https://github.com/PaddlePaddle/PaddleX/blob/develop/docs/pipeline_usage/tutorials/cv_pipelines/object_detection.md

关于paddleX上的部署，上面这个文档我看完了，似懂非懂，飞桨有没有一个典型例子能够实际演练一下，切身体会怎么做？
我没有明白的点是哪些是在服务端（云端）要做的事情，云端要不要提前申请资源？申请什么样的资源（先不说费用） 哪些是在客户端要做的，客户端和服务端的交互是怎样的？
@liuhongen1234567 这是我主要的问题，回到不了上面的问题，我们企业内部没法评估，谢谢！

[liuhongen1234567]

您好，paddlex 基本每1到2周都有直播课，包含了训练、推理、部署全流程，可以点击这个地方报名 参与打卡营，打卡营会免费发放算力来体验，也会有专门负责部署的老师答疑；
，

[liuhongen1234567]

这个是PaddleX主页链接 https://github.com/PaddlePaddle/PaddleX?tab=readme-ov-file

[fatsheep2020]

您好，paddlex 基本每1到2周都有直播课，包含了训练、推理、部署全流程，可以点击这个地方报名 参与打卡营，打卡营会免费发放算力来体验，也会有专门负责部署的老师答疑； ，

好的，谢谢提供信息，有没有比较详细的文档或者完整的例子？

[liuhongen1234567]

可以进群了解一下吧，参与一次打卡营应该就对流程比较清楚了，我不是搞部署方面的，

[fatsheep2020]

可以进群了解一下吧，参与一次打卡营应该就对流程比较清楚了，我不是搞部署方面的，

好的，谢谢！