|
Do you see a log entry saying it restarted?
cheers
Chris Maunder
|
|
|
|
|
When i change the model size :
- it stops
- it restarts
- it crashes
Nothing more.
|
|
|
|
|
The stop and restart is normal when changing model size. If you are using Blue Iris the crashing might be caused by having Auto start/stop with Blue Iris enable. If you have it enable try disabling it.
|
|
|
|
|
I don't have that.
Nothing to do with BI.
|
|
|
|
|
Same problem on 2.4.7
Works fine on YOLOv5 6.2
|
|
|
|
|
Same problem on 2.5
|
|
|
|
|
Actually upgraded on 11/16/23. All is well. Great Job!!!!!
I was hoping to solve some Bi issues with Blueiris v 5.8.0.15 upgrade from v 5.7.7.18 However no joy.
Still suffer from Late alerts and Delayed Alert JPG or delayed indication in Alerts window. The Alert happens ( I get an alert trigger) but nothing shows up in BI to indicate we had an event it until much later.
If anyone else is seeing this please comment.
Late Alert Example:
BI Log shows a motion on Barn5 at 11/29/2023 6:45:55.094 PM However alert happens at 11/29/2023 7:20:43.216 PM
3 11/29/2023 6:45:55.094 PM Barn5 MOTION_A
3 11/29/2023 6:45:58.494 PM SunRoom3 MOTION_A
0 11/29/2023 6:46:00.777 PM SunRoom3 AI: [Objects] car:61% [356,535 668,776] 158ms
0 11/29/2023 6:46:00.777 PM SunRoom3 AI: car:61%
3 11/29/2023 6:46:11.472 PM FrontDrive MOTION_A
0 11/29/2023 6:46:13.418 PM FrontDrive AI: [Objects] car:59% [115,489 244,603] 208ms
0 11/29/2023 6:46:13.418 PM FrontDrive AI: car:59%
10 11/29/2023 6:48:50.300 PM CJ [::ffff:10.3.15.127]: Logout, 4:19
0 11/29/2023 6:50:27.948 PM New Move: 1 items 3.90G [6.34T/6.34T, 944.6G free]
0 11/29/2023 6:50:27.948 PM Stored Move: nothing to do [0B/5.00G, 948.5G free]
0 11/29/2023 6:50:27.941 PM Alerts Move: nothing to do [6.05G/940.0G, 948.5G free]
0 11/29/2023 6:50:27.954 PM Alerts Delete: nothing to do [6.05G/940.0G, 948.5G free]
0 11/29/2023 6:50:27.954 PM Aux 1 Delete: nothing to do [2.33T/7.27T, 4.94T free]
0 11/29/2023 6:56:05.191 PM New Move: nothing to do [6.34T/6.34T, 946.0G free]
0 11/29/2023 6:56:05.191 PM Stored Move: nothing to do [0B/5.00G, 946.0G free]
0 11/29/2023 6:56:05.193 PM Alerts Move: nothing to do [6.05G/940.0G, 946.0G free]
0 11/29/2023 6:56:05.211 PM Alerts Delete: nothing to do [6.05G/940.0G, 946.0G free]
0 11/29/2023 7:01:05.572 PM New Move: 1 items 3.90G [6.34T/6.34T, 943.7G free]
0 11/29/2023 7:01:05.572 PM Stored Move: nothing to do [0B/5.00G, 947.6G free]
0 11/29/2023 7:01:05.573 PM Alerts Move: nothing to do [6.05G/940.0G, 947.6G free]
0 11/29/2023 7:01:05.586 PM Alerts Delete: nothing to do [6.05G/940.0G, 947.6G free]
10 11/29/2023 7:05:13.116 PM Anonymous [::ffff:10.3.15.127]: Login
10 11/29/2023 7:05:13.955 PM CJ [::ffff:10.3.15.127]: Login
10 11/29/2023 7:06:15.690 PM Anonymous [::ffff:10.3.15.127]: Logout, 1:02
0 11/29/2023 7:06:40.350 PM New Move: 1 items 3.90G [6.34T/6.34T, 945.1G free]
0 11/29/2023 7:06:40.350 PM Stored Move: nothing to do [0B/5.00G, 949.0G free]
0 11/29/2023 7:06:40.351 PM Alerts Move: nothing to do [6.05G/940.0G, 949.0G free]
0 11/29/2023 7:06:40.366 PM Alerts Delete: nothing to do [6.05G/940.0G, 949.0G free]
0 11/29/2023 7:06:40.378 PM Aux 1 Delete: nothing to do [2.33T/7.27T, 4.93T free]
10 11/29/2023 7:08:47.876 PM CJ [::ffff:10.3.15.127]: Logout, 3:34
0 11/29/2023 7:12:15.246 PM New Move: nothing to do [6.34T/6.34T, 946.4G free]
0 11/29/2023 7:12:15.246 PM Stored Move: nothing to do [0B/5.00G, 946.4G free]
0 11/29/2023 7:12:15.248 PM Alerts Move: nothing to do [6.05G/940.0G, 946.4G free]
0 11/29/2023 7:12:15.254 PM Alerts Delete: nothing to do [6.05G/940.0G, 946.4G free]
0 11/29/2023 7:17:15.647 PM New Move: 3 items 1.93G [6.34T/6.34T, 944.2G free]
0 11/29/2023 7:17:15.647 PM Stored Move: nothing to do [0B/5.00G, 946.1G free]
0 11/29/2023 7:17:15.648 PM Alerts Move: nothing to do [6.05G/940.0G, 946.1G free]
0 11/29/2023 7:17:15.654 PM Alerts Delete: nothing to do [6.05G/940.0G, 946.1G free]
3 11/29/2023 7:20:42.401 PM Barn1 MOTION_A
0 11/29/2023 7:20:43.216 PM Barn5 AI: [Objects] car:64% [214,129 307,166] 267ms
0 11/29/2023 7:20:43.216 PM Barn5 AI: car:64%
Sound Event log - Log is from a custom .Net app I wrote to send Alerts to various places with different tones.
11/29/2023 7:20:43 PM - *******************************************************************
11/29/2023 7:20:43 PM - BI Sound Event Send v1.02 Aug 03, 2023 Started
11/29/2023 7:20:43 PM - *******************************************************************
11/29/2023 7:20:43 PM - FrontMotion - SoundType = 1 - AI = Barn5-West_car:64%
Then their is the "Delayed Alert".
This is an extreme Example as it's usually hours not days:
So at 12/14/2023 1:04:51 AM - We get an Alert that actually happened days earlier AND it's not even the correct alert. BI log shows no Alert in the log at 1:04 AM
BI log
3 11/23/2023 1:32:25.510 PM Garage4 MOTION_A
0 11/23/2023 1:32:26.233 PM Garage4 AI: [Objects] person:84% [987,1 1160,298] 194ms
0 11/23/2023 1:32:26.233 PM Garage4 AI: person:84%
Sound Event log:
12/14/2023 1:04:51 AM - *******************************************************************
12/14/2023 1:04:51 AM - BI Sound Event Send v1.02 Aug 03, 2023 Started
12/14/2023 1:04:51 AM - *******************************************************************
12/14/2023 1:04:51 AM - CAT - SoundType = 2 - AI = Garage4-BreezeWay_dog:72%__[{api:objects,found:{message:Found toilet, chair, dog...,count:4,predictions:[{confidence .5074509978294373,label:toilet,x_min:879,y_min:449,x_max:1145,y_max:723},{confidence .5580800175666809,label:chair,x_min:964,y_min:412,x_max:1498,y_max:952},{confidence .7169665694236755,label:dog,x_min:843,y_min:791,x_max:1026,y_max:1060},{confidence .7378189563751221,label:toilet,x_min:1027,y_min:896,x_max:1540,y_max:1070}],success:true,processMs:229,inferenceMs:228,code:200,command:detect,moduleId:ObjectDetectionYolo,executionProvider:CUDA,canUseGPU:true,analysisRoundTripMs:246}}]
Reviewing the .DAT file shows - person:84%
JPG Shows person:84%
BI Log around 12/14/2023 1:04AM
0 12/14/2023 12:59:41.914 AM New Move: nothing to do [6.34T/6.34T, 941.3G free]
0 12/14/2023 12:59:41.916 AM Stored Move: nothing to do [0B/5.00G, 941.3G free]
0 12/14/2023 12:59:41.918 AM Alerts Move: nothing to do [11.0G/940.0G, 941.3G free]
0 12/14/2023 12:59:41.972 AM Alerts Delete: nothing to do [11.0G/940.0G, 941.3G free]
0 12/14/2023 12:59:41.972 AM Aux 1 Delete: nothing to do [7.26T/7.27T, 8.83G free]
0 12/14/2023 1:04:42.392 AM New Move: 1 items 3.90G [6.34T/6.34T, 939.0G free]
0 12/14/2023 1:04:42.392 AM Stored Move: nothing to do [0B/5.00G, 942.9G free]
0 12/14/2023 1:04:42.393 AM Alerts Move: nothing to do [11.0G/940.0G, 942.9G free]
0 12/14/2023 1:04:42.448 AM Alerts Delete: nothing to do [11.0G/940.0G, 942.9G free]
0 12/14/2023 1:04:42.611 AM Aux 1 Delete: 2 items 2.69G [7.27T/7.27T, 4.92G free]
0 12/14/2023 1:10:16.726 AM New Move: nothing to do [6.34T/6.34T, 940.4G free]
0 12/14/2023 1:10:16.728 AM Stored Move: nothing to do [0B/5.00G, 940.4G free]
0 12/14/2023 1:10:16.721 AM Alerts Move: nothing to do [11.0G/940.0G, 940.4G free]
0 12/14/2023 1:10:16.776 AM Alerts Delete: nothing to do [11.0G/940.0G, 940.4G free]
0 12/14/2023 1:10:16.788 AM Aux 1 Delete: nothing to do [7.26T/7.27T, 7.62G free]
So the Fun Continues
=============================================
Installed 11/162023
Config:
CodeVersion 2.3.4-Beta
Windows installation
Using YOLOv5 6.2 for object detection only.
System:
Dell T3630 I9-9900K - 32GB
ZOTAC GeForce GTX 1660 SUPER 6GB GDDR5 (ZT-T16620F-10L)
GPU - Driver - 522.25-desktop-win10-win11-64bit-international-dch-whql.exe
(Ver 31.0.15.2225) - Reloaded after Win Update downgraded it to 31.0.15.1737 during CodeProject.Ai install (Now WUS is Disabled)
DISABLE Half-precision
NVIDIA Toolkit - cuda_11.7.1_516.94_windows.exe
Microsoft Windows 11 Pro
W11-22H2-22623.746.221004-1223
Modifications:
MSMG TookKit 12.7 - Strips out the Bloat
Changed: 'Adjust for best Performance of' = Background Services
Since it's ran offline - Disabled Windows update service including driver updates so drivers are left alone during Code Project install.
Added Open-Shell for us old guys.
Currently installed "classic" .NET Versions in the system:
2.0.50727.4927 Service Pack 2
3.0.30729.4926 Service Pack 2
3.5.30729.4926 Service Pack 1
4.0.0.0
4.8.09032
Blueiris v 5.8.0.15
16 Cams
Essaf & Cj
|
|
|
|
|
Which Object Detection are you using? Post a screenshot of Barn5 AI settings also CodeProject.AI log at the time you had the issue.
|
|
|
|
|
Dear Mike,
Thanks for the Help...
I am using YOLOv5 6.2 for object detection only.
The Code Project logs must get removed as the log starts today at 12am. I did however verify at the time that CP showed the same alert time as the BI log at 7:20pm. So BI Delayed that long before asking CP for an AI result.
I see this delay issue many time a week. Most of the time it is just a few minutes.
I also MUST restart BI ever few days else it crashes or just stops alerting ..... The Crash will result sometimes in a restart by BI but alerts will not work until I restart CP.
So I have a task that runs Sunday, Tuesday, Friday at 1:00pm that restarts both BI and CP services ( Rather shuts down BI then CP and then restarts CP with a delay to get up and running before BI comes up else BI gets cranky and will not alert).
Essaf & Cj
|
|
|
|
|
It could be that you are using all the models that YOLOv5 6.2 has because Custom models is blank. Post a screenshot of Blue Iris Main AI settings
|
|
|
|
|
Ok ... What should I change???
Doesn't the consideration of all models take place after BI make the call to CP?
|
|
|
|
|
Based on you main AI settings you are only using one model (Default object detection). One change you should make is change port 5000 to port 32168. The next time you see the delay post the log from CodeProject.AI so we can see what is happening.
|
|
|
|
|
|
Your not alone.
We started seeing the delays I think sometime after 5.3.9.15 with more than 16 4k cams in continuous record. The way we mitigated the problem was to overlap the coverage area so multiple cameras would trigger guaranteeing at least one event.
It's not CodeProject as the BlueI log shows the motion and then many minutes later the processed .DAT & .JPG file date/time show the same times and like yours and lineup with the notification times sent from BlueI.
You can disable AI and also see the delays between Motion detected and PROG Ran or SMTP send for example.
Good luck....
|
|
|
|
|
Thank you very much for your solution.
I have started repositioning cams where I can get the overlap and ordered more cams as needed.
Wow 5.3.9.15 brings back memories, the last version where the Traffic Light worked properly. After 5.3.9.15 we had to re-write our Ademco/EyesON Interface and add an Alert system to handle outbound e-mail-sms Alerts & Video clips outside of Bi. Now I will add code to look at the Bi Alert folder for overlap events and only send one alert event.....
Thank you again .... for responding....
Essaf & Cj
|
|
|
|
|
Hi,
I have asked earlier about LPR on CPU, won't use GPU (cuda) not even the option (cog) available.
Face & object Yolo is fine & both GPU (Cuda)
Reading so many with similar problems not expecting much luck so thought try more myself but cannot anywhere to download older modules.
I have since tried many uninstalls, difference cuda, now on 11.7, I think everything mentioned elsewhere.
I think is was the getcuda in utils.bat, changed the text.
Would be good to try perhaps older CodeprojectAI, are these available, am I missing something as cannot see any legacy.
Really want to get this working as object on cuda is so fast & a speedy LPR would be great, almost a deal breaker if not & may stick to my older plate recognizer setup.
thanks I/A
|
|
|
|
|
ALPR definitely uses GPU. I would try the latest pre-release (2.4.5) just to be sure you have the latest fixes
cheers
Chris Maunder
|
|
|
|
|
Thanks for your reply,
I did try 2.4.5 but actually would not let me install the LPR module at all!
just said failed, don't worry about that.
I had uninstalled previous CPAI but not the cuda etc. perhaps needed a really full clean install
Anyway back to previous and can at least install LPR & it works, just not cuda.
I am going to try & investigate ..
paddle install, seems installed OK but as face & object is GPU Cuda OK looking for things unique to LPR
The other thing & again as Object & Face OK not sure if likely to be relevant
Visual studio, not sure how to install separate with the extension necessary, so far just let the normal install handle that.
github lists as python code, visual studio not clear with ref. to CPAI
Is there something I can try ref.
1) paddle - for 11.8 - all seemed OK using below
python -m pip install paddlepaddle-gpu==2.5.2 -i https://mirror.baidu.com/pypi/simple
>>> import paddle
>>> paddle.utils.run_check()
Running verify PaddlePaddle program ...
I1219 08:09:34.013252 6960 interpretercore.cc:237] New Executor is Running.
W1219 08:09:34.016249 6960 gpu_resources.cc:96] The GPU architecture in your current machine is Pascal, which is not compatible with Paddle installation with arch: 70 75 80 86 , it is recommended to install the corresponding wheel package according to the installation information on the official Paddle website.
W1219 08:09:34.016249 6960 gpu_resources.cc:119] Please NOTE: device: 0, GPU Compute Capability: 6.1, Driver API Version: 11.8, Runtime API Version: 11.8
W1219 08:09:34.024252 6960 gpu_resources.cc:149] device: 0, cuDNN Version: 8.9.
I1219 08:09:36.909252 6960 interpreter_util.cc:518] Standalone Executor is Used.
PaddlePaddle works well on 1 GPU.
PaddlePaddle is installed successfully! Let's start deep learning with PaddlePaddle now.
2) Visual Studio,
Not sure where to start?
Any ideas?
|
|
|
|
|
|
Hi,
Thanks for your reply, uninstalled & done a fresh install
ANPR does work but just CPU, Object is super fast on the GPU (Cuda) & would be great to LPR
Did see this at the end,
CPAI_MODULE_REQUIRED_MB not found
<pre>Server version: 2.3.4-Beta
System: Windows
Operating System: Windows (Microsoft Windows 10.0.19045)
CPUs: Intel(R) Core(TM) i7-4790 CPU @ 3.60GHz (Intel)
1 CPU x 4 cores. 8 logical processors (x64)
GPU: NVIDIA GeForce GTX 1050 Ti (4 GiB) (NVIDIA)
Driver: 522.06 CUDA: 11.8.89 (max supported: 11.8) Compute: 6.1
System RAM: 16 GiB
Target: Windows
BuildConfig: Release
Execution Env: Native
Runtime Env: Production
.NET framework: .NET 7.0.10
Video adapter info:
NVIDIA GeForce GTX 1050 Ti:
Driver Version 31.0.15.2206
Video Processor NVIDIA GeForce GTX 1050 Ti
Intel(R) HD Graphics 4600:
Driver Version 20.19.15.5126
Video Processor Intel(R) HD Graphics Family
System GPU info:
GPU 3D Usage 1%
GPU RAM Usage 709 MiB
Global Environment variables:
CPAI_APPROOTPATH = <root>
CPAI_PORT = 32168
14:55:38:Preparing to install module 'ALPR'
14:55:38:Downloading module 'ALPR'
14:55:39:Installing module 'ALPR'
14:55:39:Installer script at 'C:\Program Files\CodeProject\AI\setup.bat'
14:55:39:ALPR: Installing CodeProject.AI Analysis Module
14:55:39:ALPR: ========================================================================
14:55:39:ALPR: CodeProject.AI Installer
14:55:39:ALPR: ========================================================================
14:55:39:ALPR: General CodeProject.AI setup
14:55:39:ALPR: Creating Directories...Done
14:55:39:ALPR: GPU support
14:55:40:ALPR: CUDA Present...Yes (version 11.8)
14:55:40:ALPR: ROCm Present...False
14:55:41:ALPR: Installing module ALPR 2.7.2
14:55:41:ALPR: Installing Python 3.7
14:55:41:ALPR: Python 3.7 is already installed
14:55:51:ALPR: Creating Virtual Environment (Local)...Done
14:55:51:ALPR: Confirming we have Python 3.7 in our virtual environment...present
14:55:51:ALPR: Downloading ALPR models...already exists...Expanding...Done.
14:55:51:ALPR: Copying contents of paddleocr-models.zip to paddleocr...done
14:55:51:ALPR: Cleaning up...done
14:55:51:ALPR: Installing Python packages for ALPR
14:55:51:ALPR: [0;Installing GPU-enabled libraries: If available
14:55:53:ALPR: Ensuring Python package manager (pip) is installed...Done
14:56:09:ALPR: Ensuring Python package manager (pip) is up to date...Done
14:56:09:ALPR: Python packages specified by requirements.windows.cuda11_8.txt
14:57:03:ALPR: - Installing PaddlePaddle, Parallel Distributed Deep Learning...(✔️ checked) Done
14:58:55:ALPR: - Installing PaddleOCR, the OCR toolkit based on PaddlePaddle...(✔️ checked) Done
14:58:59:ALPR: - Installing imutils, the image utilities library...(✔️ checked) Done
14:59:01:ALPR: - Installing Pillow, a Python Image Library...Already installed
14:59:02:ALPR: - Installing OpenCV, the Computer Vision library for Python...Already installed
14:59:03:ALPR: - Installing NumPy, a package for scientific computing...Already installed
14:59:04:ALPR: Installing Python packages for the CodeProject.AI Server SDK
14:59:05:ALPR: Ensuring Python package manager (pip) is installed...Done
14:59:09:ALPR: Ensuring Python package manager (pip) is up to date...Done
14:59:09:ALPR: Python packages specified by requirements.txt
14:59:10:ALPR: - Installing Pillow, a Python Image Library...Already installed
14:59:12:ALPR: - Installing Charset normalizer...Already installed
14:59:19:ALPR: - Installing aiohttp, the Async IO HTTP library...(✔️ checked) Done
14:59:23:ALPR: - Installing aiofiles, the Async IO Files library...(✔️ checked) Done
14:59:27:ALPR: - Installing py-cpuinfo to allow us to query CPU info...(✔️ checked) Done
14:59:29:ALPR: - Installing Requests, the HTTP library...Already installed
14:59:29:ALPR: Setup complete
14:59:29:Module ALPR installed successfully.
14:59:29:Command: C:\Program Files\CodeProject\AI\modules\ALPR\bin\windows\python37\venv\Scripts\python
14:59:29:
14:59:29:
14:59:29:Module 'License Plate Reader' 2.7.2 (ID: ALPR)
14:59:29:Module Path: C:\Program Files\CodeProject\AI\modules\ALPR
14:59:29:Starting C:\Program Files...ws\python37\venv\Scripts\python "C:\Program Files...\modules\ALPR\ALPR_adapter.py"
14:59:29:Attempting to start ALPR with C:\Program Files\CodeProject\AI\modules\ALPR\bin\windows\python37\venv\Scripts\python "C:\Program Files\CodeProject\AI\modules\ALPR\ALPR_adapter.py"
14:59:29:AutoStart: True
14:59:29:Queue: alpr_queue
14:59:29:Platforms: windows,linux,linux-arm64,macos,macos-arm64,!orangepi,!jetson
14:59:29:GPU Libraries: installed if available
14:59:29:GPU Enabled: enabled
14:59:29:Parallelism: 0
14:59:29:Accelerator:
14:59:29:Half Precis.: enable
14:59:29:Runtime: python3.7
14:59:29:Runtime Loc: Local
14:59:29:FilePath: ALPR_adapter.py
14:59:29:Pre installed: False
14:59:29:Start pause: 1 sec
14:59:29:LogVerbosity:
14:59:29:Valid: True
14:59:29:Environment Variables
14:59:29:AUTO_PLATE_ROTATE = True
14:59:29:MIN_COMPUTE_CAPABILITY = 7
14:59:29:MIN_CUDNN_VERSION = 7
14:59:29:OCR_OPTIMAL_CHARACTER_HEIGHT = 60
14:59:29:OCR_OPTIMAL_CHARACTER_WIDTH = 36
14:59:29:OCR_OPTIMIZATION = True
14:59:29:PLATE_CONFIDENCE = 0.7
14:59:29:PLATE_RESCALE_FACTOR = 2
14:59:29:PLATE_ROTATE_DEG = 0
14:59:29:
14:59:29:Started License Plate Reader module
14:59:29:Installer exited with code 0
14:59:29:Call to run Install on module ALPR has completed.
14:59:30:Module ALPR started successfully.
15:00:01:ALPR_adapter.py: CPAI_MODULE_REQUIRED_MB not found. Setting to default 0
15:00:01:ALPR_adapter.py: Timeout connecting to the server
15:00:01:ALPR_adapter.py: License Plate Reader started.License Plate Reader: License Plate Reader started.
15:00:01:License Plate Reader: License Plate Reader started.
|
|
|
|
|
Member 16163883 wrote: CPAI_MODULE_REQUIRED_MB not found. Setting to default 0
This is fine. It's just letting you know.
PaddlePaddle on CUDA 11.8 uses paddlepaddle-gpu==2.5.2 (-i https://mirror.baidu.com/pypi/simple) but 11.7 (which I have) uses paddlepaddle-gpu==2.5.2.post117 (https://www.paddlepaddle.org.cn/whl/linux/cudnnin/stable.html). I wonder if the 11.8 install just doesn't work as is advertised on the box.
If you were keen then you could try the following:
- Remove the Python libraries: Go to
C:\Program Files\CodeProject\AI\modules\ALPR and delete the bin directory. - Use the CUDA 11.7 setup for CUDA 11.8: Delete
requirements.linux.cuda11_7.txt and rename requirements.linux.cuda11_8.txt to requirements.linux.cuda11_7.txt - Redo the installation. In that same directory, open a Windows Command terminal and type
..\..\setup .
If this works please let me know and we'll make this the permanent solution.
cheers
Chris Maunder
|
|
|
|
|
Hi,
I have found this on the install, will investigate later but perhaps you can point me in the right direction.
11:14:25:ALPR: - Installing PaddlePaddle, Parallel Distributed Deep Learning...(✔️ checked) Done
11:16:10:ALPR: ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
11:16:10:ALPR: paddlepaddle-gpu 2.5.2 requires protobuf<=3.20.2,>=3.1.0; platform_system == "Windows", but you have protobuf 4.24.4 which is incompatible.
I did try briefly
Remove the Python libraries: Go to C:\Program Files\CodeProject\AI\modules\ALPR and delete the bin directory.
Use the CUDA 11.7 setup for CUDA 11.8: Delete requirements.linux.cuda11_7.txt and rename requirements.linux.cuda11_8.txt to requirements.linux.cuda11_7.txt
Redo the installation. In that same directory, open a Windows Command terminal and type ..\..\setup.
Did not help but little testing, will try again.
Mine is a windows install, I assume requirements.windows .. not linux as you advised above
I am assuming the ..\\..\setup does not download a fresh lot of all the requirements etc. overwriting any chages just made?
|
|
|
|
|
Hi,
sorry to persist, I have opened another thread, can someone can point me in the right direction of where these installing "Collected Packages" are called from.
I have searched for hours & just cannot see, suspect those in the know, would immediately know where the procedure/code/requirements, is located
15:09:26:ALPR: Using cached MarkupSafe-2.1.3-cp37-cp37m-win_amd64.whl (17 kB)
15:09:27:ALPR: Installing collected packages: pytz, pyclipper, lmdb, zipp, urllib3, typing-extensions, termcolor, soupsieve, six, rarfile, rapidfuzz, pyparsing, PyMuPDF, pycryptodome, psutil, protobuf, Pillow, packaging, numpy, networkx, MarkupSafe, lxml, itsdangerous, idna, future, fonttools, et-xmlfile, cython, cycler, cssselect, colorama, charset-normalizer, certifi, cachetools, Babel, Werkzeug, tqdm, tifffile, shapely, scipy, requests, PyWavelets, python-docx, python-dateutil, openpyxl, opencv-python, opencv-contrib-python, kiwisolver, Jinja2, importlib-metadata, imageio, fire, beautifulsoup4, bce-python-sdk, attrdict, scikit-image, pdf2docx, pandas, matplotlib, cssutils, click, premailer, imgaug, flask, Flask-Babel, visualdl, paddleocr
15:10:59:ALPR: Successfully installed Babel-2.14.0 Flask-Babel-3.1.0 Jinja2-3.1.2 MarkupSafe-2.1.3 Pillow-9.5.0 PyMuPDF-1.20.2 PyWavelets-1.3.0 Werkzeug-2.2.3 attrdict-2.0.1 bce-python-sdk-0.8.98 beautifulsoup4-4.12.2 cachetools-5.3.2 certifi-2023.11.17 charset-normalizer-3.3.2 click-8.1.7 colorama-0.4.6 cssselect-1.2.0 cssutils-2.7.1 cycler-0.11.0 cython-3.0.7 et-xmlfile-1.1.0 fire-0.5.0 flask-2.2.5 fonttools-4.38.0 future-0.18.3 idna-3.6 imageio-2.31.2 imgaug-0.4.0 importlib-metadata-6.7.0 itsdangerous-2.1.2 kiwisolver-1.4.5 lmdb-1.4.1 lxml-4.9.4 matplotlib-3.5.3 networkx-2.6.3 numpy-1.21.6 opencv-contrib-python-4.6.0.66 opencv-python-4.6.0.66 openpyxl-3.1.2 packaging-23.2 paddleocr-2.7.0.2 pandas-1.3.5 pdf2docx-0.5.6 premailer-3.10.0 protobuf-4.24.4 psutil-5.9.7 pyclipper-1.3.0.post5 pycryptodome-3.19.0 pyparsing-3.1.1 python-dateutil-2.8.2 python-docx-1.1.0 pytz-2023.3.post1 rapidfuzz-3.4.0 rarfile-4.1 requests-2.31.0 scikit-image-0.19.3 scipy-1.7.3 shapely-2.0.2 six-1.16.0 soupsieve-2.4.1 termcolor-2.3.0 tifffile-2021.11.2 tqdm-4.66.1 typing-extensions-4.7.1 urllib3-2.0.7 visualdl-2.5.3 zipp-3.15.0
15:10:59:ALPR: ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
15:10:59:ALPR: paddlepaddle-gpu 2.5.2 requires protobuf<=3.20.2,>=3.1.0; platform_system == "Windows", but you have protobuf 4.24.4 which is incompatible.
15:10:59:ALPR: WARNING: Target directory C:\Program Files\CodeProject\AI\modules\ALPR\bin\windows\python37\venv\Lib\site-packages\certifi already exists. Specify --upgrade to force replacement.
I want to stop if from installing protobuf 4.24.4, either 3.20.2 in its place or perhaps remove all together.
3.20.2 is installed earlier on in the log.
much earlier in the install/log is
15:08:35:ALPR: Installing collected packages: typing-extensions, sniffio, protobuf, Pillow, numpy, idna, exceptiongroup, decorator, certifi, astor, opt-einsum, h11, anyio, httpcore, httpx, paddlepaddle-gpu
15:09:09:ALPR: Successfully installed Pillow-9.5.0 anyio-3.7.1 astor-0.8.1 certifi-2023.11.17 decorator-5.1.1 exceptiongroup-1.2.0 h11-0.14.0 httpcore-0.17.3 httpx-0.24.1 idna-3.6 numpy-1.21.6 opt-einsum-3.3.0 paddlepaddle-gpu-2.5.2 protobuf-3.20.2 sniffio-1.3.0 typing-extensions-4.7.1
15:09:11:ALPR: (✔️ checked) Done
1
|
|
|
|
|
Tried uninstalling, removing directory reinstalling many times.
Cuda 11.8 as advised
did try 12, no GPU at all, 11.1 similar to 11.8, object enabled but CPU still just LPR
Confused as the other two are enabled & work fine. I don't use face but object so far is fast & reliable.
MSI Nvidia 1050 ti, latest drivers, used the script etc.
Face Processing 1.8.1
Stopped
GPU (CUDA)
0
License Plate Reader 2.7.2
Started
CPU
150
Object Detection (YOLOv5 .NET) 1.7.1
Stopped
CPU
0
Object Detection (YOLOv5 6.2) 1.7.1
Started
GPU (CUDA)
Server version: 2.3.4-Beta
System: Windows
Operating System: Windows (Microsoft Windows 10.0.19045)
CPUs: Intel(R) Core(TM) i7-4790 CPU @ 3.60GHz (Intel)
1 CPU x 4 cores. 8 logical processors (x64)
GPU: NVIDIA GeForce GTX 1050 Ti (4 GiB) (NVIDIA)
Driver: 522.06 CUDA: 11.8.89 (max supported: 11.8) Compute: 6.1
System RAM: 16 GiB
Target: Windows
BuildConfig: Release
Execution Env: Native
Runtime Env: Production
.NET framework: .NET 7.0.10
Video adapter info:
NVIDIA GeForce GTX 1050 Ti:
Driver Version 31.0.15.2206
Video Processor NVIDIA GeForce GTX 1050 Ti
Intel(R) HD Graphics 4600:
Driver Version 20.19.15.5126
Video Processor Intel(R) HD Graphics Family
System GPU info:
GPU 3D Usage 15%
GPU RAM Usage 828 MiB
Global Environment variables:
CPAI_APPROOTPATH = <root>
CPAI_PORT = 32168
|
|
|
|
|
|