Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

PaddleDetection的服务部署bug #7918

Open
3 tasks done
Coco0201 opened this issue Mar 12, 2023 · 7 comments
Open
3 tasks done

PaddleDetection的服务部署bug #7918

Coco0201 opened this issue Mar 12, 2023 · 7 comments
Assignees
Labels
status: to reply User has repied to this issue, and waiting R&D to confirm.

Comments

@Coco0201
Copy link

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

  1. PaddleDetection的服务部署的_--model_dir不生效_,必须得改config.yml,如下图的model_config与命令行参数不一致

python deploy/serving/python/web_service.py --model_dir inference_model/faster_rcnn_r50_1x_coco &
'''
image

'''
2. PaddleDetection的服务部署后,用图片测试只能输出一个box
python deploy/serving/python/pipeline_http_client.py --image_file demo/000000000139.jpg

'''
image

'''

复现环境 Environment

PaddlePaddle:2.4.2

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!
@Coco0201
Copy link
Author

@jerrywgz

@zhiboniu
Copy link
Collaborator

你好,
问题1,看起来应该是一样的吧,都是inference_model/faster_rcnn_r50_1x_coco?
问题2,感谢反馈,我们会检查一下代码实现

@zhiboniu
Copy link
Collaborator

image
看代码应该是有对多输出都做处理的,检查下是否是模型的问题呢

@Coco0201
Copy link
Author

服务部署时,命令行参数是
--model_dir inference_model/faster_rcnn_r50_1x_coco
但是执行后打印的不是以上模型,必须修改config.yml才行
image

@thinkthinking thinkthinking added the status: to reply User has repied to this issue, and waiting R&D to confirm. label Mar 13, 2023
@zhiboniu
Copy link
Collaborator

看起来好像确实有bug,我们修一下,感谢反馈

@jerrywgz jerrywgz added staus: replied R&D on duty has repied to this issue, and waiting users to confirm. and removed status: to reply User has repied to this issue, and waiting R&D to confirm. labels Mar 13, 2023
@zhiboniu
Copy link
Collaborator

梳理了代码,这边打印的时候还没有合并传参的model_dir,所以打印出来只是配置文件里的内容,后面实际是有合并,所以可以起作用的。已提pr更新打印显示为合并后的。

@Coco0201
Copy link
Author

Coco0201 commented Mar 14, 2023

image 看代码应该是有对多输出都做处理的,检查下是否是模型的问题呢

类Op的_run_process函数中1151行的input_offset_dict写死为[0,1]导致的错误
image
@thinkthinking

@thinkthinking thinkthinking added status: to reply User has repied to this issue, and waiting R&D to confirm. and removed staus: replied R&D on duty has repied to this issue, and waiting users to confirm. labels Mar 14, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
status: to reply User has repied to this issue, and waiting R&D to confirm.
Projects
None yet
Development

No branches or pull requests

4 participants