CodeProject.AI Version 2.0

[Skinny1]

Codeproject kept crashing so I deleted it and all files in the codeproject directory and reinstalled Codeproject Ai 2.4.6-Beta. Now I am getting this error, and the ALPR isn't working.

The below is the error I am getting.
15:51:22:ALPR_adapter.py: Traceback (most recent call last):
15:51:22:ALPR_adapter.py: File "C:\Program Files\CodeProject\AI\modules\ALPR\ALPR_adapter.py", line 7, in
15:51:22:ALPR_adapter.py: from request_data import RequestData
15:51:22:ALPR_adapter.py: File "../../SDK/Python\request_data.py", line 7, in
15:51:22:ALPR_adapter.py: from PIL import Image
15:51:22:ALPR_adapter.py: ModuleNotFoundError: No module named 'PIL'

I have no modules named PIL nor have I had any in previous versions of Codeproject Ai.

Any ideas or suggestions??

Hey Mike, (Disregard) I just got it working. I uninstalled the LPR module twice and reinstalled it before I got it to work.)
)
I just downloaded Codeproject Ai 2.4.7-Beta, deleted 2.4.6-Beta and installed 2.4.7-Beta and I am now getting this error (not the same error message above that you helped me fix) -

12:42:07:Started License Plate Reader module
12:42:07:ALPR_adapter.py: Traceback (most recent call last):
12:42:07:ALPR_adapter.py: File "C:\Program Files\CodeProject\AI\modules\ALPR\ALPR_adapter.py", line 8, in
12:42:07:ALPR_adapter.py: from module_runner import ModuleRunner
12:42:07:ALPR_adapter.py: File "../../SDK/Python\module_runner.py", line 37, in
12:42:07:ALPR_adapter.py: import aiohttp
12:42:07:ALPR_adapter.py: ModuleNotFoundError: No module named 'aiohttp'
12:42:07:Module ALPR has shutdown

Any ideas on this new error message with 2.4.7-Beta?

Thank you,

Rick

 

[prsmith777]

Hey Mike,

I just downloaded Codeproject Ai 2.4.7-Beta, deleted 2.4.6-Beta and installed 2.4.7-Beta and I am now getting this error (not the same error message above that you helped me fix) -

12:42:07:Started License Plate Reader module
12:42:07:ALPR_adapter.py: Traceback (most recent call last):
12:42:07:ALPR_adapter.py: File "C:\Program Files\CodeProject\AI\modules\ALPR\ALPR_adapter.py", line 8, in
12:42:07:ALPR_adapter.py: from module_runner import ModuleRunner
12:42:07:ALPR_adapter.py: File "../../SDK/Python\module_runner.py", line 37, in
12:42:07:ALPR_adapter.py: import aiohttp
12:42:07:ALPR_adapter.py: ModuleNotFoundError: No module named 'aiohttp'
12:42:07:Module ALPR has shutdown

Any ideas on this new error message with 2.4.7-Beta?

Thank you,

Rick

I had a similar error. Seems like something's broken in 2.4.7. I reverted back to 2.3.4

 

[CCTVCam]

Maybe it would be worth recruiting a group of private Beta testers rather than just calling something Beta and releasing it to the wider community. I'm sure there are some out there who are IT based and can run virtual machines or system cloning software etc to make testing really easy in so far as rolling back is one click when bugs are indentified plus many of these people could set up testing rigs on spare machines. The issue with having beta software out to the whole community is many download the latest version only to experience isues with bugs, and in my experience with BI / CPAI combination, rolling back hasn't worked. I've literally had to reinstall windows from scratch every single time I've encountered a CPAI issue as neither rolling back, using Windows Rollback, DISM commands or any other means of restoring windows has helped. The only other way to do this safely is for the rollback facilities of both CPAI and BI to improved to where they're 100% reliable in rolliong back every aspect of the programmes changes including registry entries and changes to all windows records / sub systems / features. A clean up tool might also help although the issue with removing programes is yuou lose settings as well and in a programme like BI, they can be quite complex and time consuming to re-enter.

 

[weigle2]

I'm surprised CPAI does not have a clean-up tool to completely remove not only the software, but also the Registry entries created. I tried to upgrade CPAI from 2.0.8 to 2.3.4 and ran into numerous problems, including the Python issue that prevented 2.3.4 from even running. I gave up trying to figure out how to fix it, and just did a clean install of Win10Pro and starting from scratch again.

Maybe it would be worth recruiting a group of private Beta testers rather than just calling something Beta and releasing it to the wider community. I'm sure there are some out there who are IT based and can run virtual machines or system cloning software etc to make testing really easy in so far as rolling back is one click when bugs are indentified plus many of these people could set up testing rigs on spare machines. The issue with having beta software out to the whole community is many download the latest version only to experience isues with bugs, and in my experience with BI / CPAI combination, rolling back hasn't worked. I've literally had to reinstall windows from scratch every single time I've encountered a CPAI issue as neither rolling back, using Windows Rollback, DISM commands or any other means of restoring windows has helped. The only other way to do this safely is for the rollback facilities of both CPAI and BI to improved to where they're 100% reliable in rolliong back every aspect of the programmes changes including registry entries and changes to all windows records / sub systems / features. A clean up tool might also help although the issue with removing programes is yuou lose settings as well and in a programme like BI, they can be quite complex and time consuming to re-enter.

 

[fenderman]

Maybe it would be worth recruiting a group of private Beta testers rather than just calling something Beta and releasing it to the wider community. I'm sure there are some out there who are IT based and can run virtual machines or system cloning software etc to make testing really easy in so far as rolling back is one click when bugs are indentified plus many of these people could set up testing rigs on spare machines. The issue with having beta software out to the whole community is many download the latest version only to experience isues with bugs, and in my experience with BI / CPAI combination, rolling back hasn't worked. I've literally had to reinstall windows from scratch every single time I've encountered a CPAI issue as neither rolling back, using Windows Rollback, DISM commands or any other means of restoring windows has helped. The only other way to do this safely is for the rollback facilities of both CPAI and BI to improved to where they're 100% reliable in rolliong back every aspect of the programmes changes including registry entries and changes to all windows records / sub systems / features. A clean up tool might also help although the issue with removing programes is yuou lose settings as well and in a programme like BI, they can be quite complex and time consuming to re-enter.

This is silly. No reason to hold back releases because someone can't control themselves and updates...

 

[jeremypsl]

What is your advise to avoid false alert ? There is some shadows recognize as a person with more than 80% of confidence. I used the Static Object Analysis but it seem to not be very helpfull...

 

[wpiman]

I'm surprised CPAI does not have a clean-up tool to completely remove not only the software, but also the Registry entries created. I tried to upgrade CPAI from 2.0.8 to 2.3.4 and ran into numerous problems, including the Python issue that prevented 2.3.4 from even running. I gave up trying to figure out how to fix it, and just did a clean install of Win10Pro and starting from scratch again.

How does the software run on Docker for Windows? I have my Coral AI TPUs on my windows machine... Code Project runs like a dog on my windows machine and maybe this is the reason.

I assume GPU and TPU passthrough is possible on docker for windows?

 

[weigle2]

How does the software run on Docker for Windows? I have my Coral AI TPUs on my windows machine... Code Project runs like a dog on my windows machine and maybe this is the reason.

I assume GPU and TPU passthrough is possible on docker for windows?

I can't answer any questions about Docker for Windows, as I am not using it. Running it on Win10Pro with a SFF i8700 and just using CPU. No hardware acceleration of GPU, and I consistently get sub 100ms to 200 ms detection times running the .NET detection engine. This is on 2.0.8 CPAI.

 

[CCTVCam]

I'm surprised CPAI does not have a clean-up tool to completely remove not only the software, but also the Registry entries created. I tried to upgrade CPAI from 2.0.8 to 2.3.4 and ran into numerous problems, including the Python issue that prevented 2.3.4 from even running. I gave up trying to figure out how to fix it, and just did a clean install of Win10Pro and starting from scratch again.

That's the issue I've had. You never know if the next update is going to break or make it. When it breaks it, nothing less than a clean Windows Installation seems to fix it. I have no issue with Apps being beta if a fix is as simple as rolling back the App. However, when a new clean installtion of Windows is required, it makes any kind of upgrade Russian Roulette.

 

[fenderman]

That's the issue I've had. You never know if the next update is going to break or make it. When it breaks it, nothing less than a clean Windows Installation seems to fix it. I have no issue with Apps being beta if a fix is as simple as rolling back the App. However, when a new clean installtion of Windows is required, it makes any kind of upgrade Russian Roulette.

This is not true I've been able to revert back to older versions no problem... Simply uninstall cpai and manually delete the two code project folders...

 

[kaltertod]

This is not true I've been able to revert back to older versions no problem... Simply uninstall cpai and manually delete the two code project folders...

100% never have needed to clean the registry or anything else beyond deleting the CPAI folders after an uninstall... Hell I have never even had to reboot the computer to make the AI server work.

 

[elesjuan]

Hey all, it's me again. I've got about 50 hours of troubleshooting invested in trying to get CPAI to work on my ridiculously overcomplicated setup, and I'm very happy to report that I've reached the 95% mark! After completely rebuilding my linux container that hosts docker 3 times, and installing easily 20 different versions of CPAI, finally found a couple Nvidia/docker related things that were missing. CPAI runs in docker on my proxmox node using GPU sharing with CUDA! MOSTLY, which is why I'm back with one last issue:

Face Processing and YOLO v5 6.2 both work with CUDA GPU.
License Plate Reader 2.6, refuses to.

I've uninstalled and reinstalled the license plate reader module multiple times watching the logs, forced it to download without using local cache, still can't get it working with CUDA, the info tab says "Support Disabled" for GPU. This machine has 3 GPUs in it, two CUDA enabled. I'm manually assigning a GPU not accessible by any other device/process to weed out interference or out of memory issues.

Host is Proxmox PVE6.4 Bare metal
Docker running in a privlidged LXC container with full access to the GPUs (ubuntu 20.04)
CPAI: 2.2.4-Beta
Run command: docker run --name CodeProject.AI-Server3 -d -p 32168:32168 --gpus device=GPU-eb6187de-dff9-83e5-ce33-140d9e466b12 codeproject/ai-server:gpu
CUDA 11.7, 515.105.01 Driver
GPU: NVidia Tesla P4 8gb




Most recent install log for License Plate Reader:

Installation Log
2024-01-06 00:59:06: Hi Docker! We will disable shared python installs for downloaded modules
2024-01-06 00:59:06: No schemas installed
2024-01-06 00:59:06: (No schemas means: we can't detect if you're in light or dark mode)
2024-01-06 00:59:06: sh: 1: lsmod: not found
2024-01-06 00:59:06: Installing CodeProject.AI Analysis Module
2024-01-06 00:59:06: ======================================================================
2024-01-06 00:59:06: CodeProject.AI Installer
2024-01-06 00:59:06: ======================================================================
2024-01-06 00:59:06: 253.02 GiB available
2024-01-06 00:59:06: Checking GPU support
2024-01-06 00:59:06: CUDA Present...Yes
2024-01-06 00:59:06: Allowing GPU Support: Yes
2024-01-06 00:59:06: Allowing CUDA Support: Yes
2024-01-06 00:59:06: General CodeProject.AI setup
2024-01-06 00:59:06: Setting permissions...Done
2024-01-06 00:59:06: Processing SDK
2024-01-06 00:59:12: Adding toolchain repo... Done
2024-01-06 00:59:15: Checking for installed dependencies (from gcc-4.9 apt-utils)...Done
2024-01-06 00:59:20: Installing missing dependencies: gcc-4.9...Done
2024-01-06 00:59:22: Upgrading libstdc... Done
2024-01-06 00:59:34: Checking for installed dependencies (from ca-certificates gnupg libc6-dev libfontconfig1 libgdiplus libjpeg-dev zlib1g-dev)...Done
2024-01-06 00:59:34: All dependencies already installed.
2024-01-06 00:59:48: Checking for installed dependencies (from ffmpeg libsm6 libxext6 mesa-utils curl rsync unzip wget)...Done
2024-01-06 00:59:48: All dependencies already installed.
2024-01-06 00:59:48: The command could not be loaded, possibly because:
2024-01-06 00:59:48: * You intended to execute a .NET application:
2024-01-06 00:59:48: The application '--version' does not exist.
2024-01-06 00:59:48: * You intended to execute a .NET SDK command:
2024-01-06 00:59:48: No .NET SDKs were found.
2024-01-06 00:59:48: Download a .NET SDK:
2024-01-06 00:59:48: .NET Downloads (Linux, macOS, and Windows)
2024-01-06 00:59:48: Learn about SDK resolution:
2024-01-06 00:59:48: Select which .NET version to use - .NET
2024-01-06 00:59:48: Checking for .NET >= 7.0...All good. Current .NET is , requested was 7.0
2024-01-06 00:59:48: Installing module ALPR 2.6
2024-01-06 00:59:48: Python 3.8 is already installed
2024-01-06 00:59:50: Installing Virtual Environment tools for Linux... Done
2024-01-06 00:59:55: Creating Virtual Environment (Local)... Done
2024-01-06 00:59:55: Checking for Python 3.8...Found Python 3.8.18. present
2024-01-06 00:59:55: Checking for CUDA...CUDA Present
2024-01-06 00:59:55: CUDA version is:
2024-01-06 00:59:57: Installing PIP... Done
2024-01-06 01:00:00: Ensuring PIP compatibility... Done
2024-01-06 01:00:04: Updating PIP... Done
2024-01-06 01:00:04: Installing setuptools... Done
2024-01-06 01:00:04: Choosing packages from requirements.linux.txt
2024-01-06 01:01:20: Installing Packages into Virtual Environment... Success
2024-01-06 01:01:21: Downloading OCR models... already exists...Expanding... Done.
2024-01-06 01:01:21: Applying PaddleOCR patch
2024-01-06 01:01:21: Installing Server SDK support:
2024-01-06 01:01:21: Checking for CUDA...CUDA Present
2024-01-06 01:01:21: CUDA version is:
2024-01-06 01:01:23: Installing PIP... Done
2024-01-06 01:01:26: Ensuring PIP compatibility... Done
2024-01-06 01:01:30: Updating PIP... Done
2024-01-06 01:01:31: Installing setuptools... Done
2024-01-06 01:01:31: Choosing packages from requirements.txt
2024-01-06 01:01:39: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/requests already exists. Specify --upgrade to force replacement.
2024-01-06 01:01:39: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/Pillow.libs already exists. Specify --upgrade to force replacement.
2024-01-06 01:01:39: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/idna already exists. Specify --upgrade to force replacement.
2024-01-06 01:01:39: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/certifi-2023.11.17.dist-info already exists. Specify --upgrade to force replacement.
2024-01-06 01:01:39: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/urllib3-2.1.0.dist-info already exists. Specify --upgrade to force replacement.
2024-01-06 01:01:39: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/PIL already exists. Specify --upgrade to force replacement.
2024-01-06 01:01:39: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/urllib3 already exists. Specify --upgrade to force replacement.
2024-01-06 01:01:39: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/requests-2.31.0.dist-info already exists. Specify --upgrade to force replacement.
2024-01-06 01:01:39: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/certifi already exists. Specify --upgrade to force replacement.
2024-01-06 01:01:39: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/Pillow-9.5.0.dist-info already exists. Specify --upgrade to force replacement.
2024-01-06 01:01:39: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/idna-3.6.dist-info already exists. Specify --upgrade to force replacement.
2024-01-06 01:01:39: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/charset_normalizer already exists. Specify --upgrade to force replacement.
2024-01-06 01:01:39: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/bin already exists. Specify --upgrade to force replacement.
2024-01-06 01:01:39: Installing Packages into Virtual Environment... Success
2024-01-06 01:01:39: Setup complete
Installer exited with code 0

System Info Tab:

 

[kaltertod]

try this

windows key + s, Type cmd right click on it and run as administrator, then cd c:\Program Files\Code*\AI\modules\ (not sure of the ALPR directory) then run ..\..\setup in the command line and let the program run.

See if this gets ALPR working

Source

 

[elesjuan]

try this

windows key + s, Type cmd right click on it and run as administrator, then cd c:\Program Files\Code*\AI\modules\ (not sure of the ALPR directory) then run ..\..\setup in the command line and let the program run.

See if this gets ALPR working

Source

Appreciate the input, but there is not a Windows machine involved in any of this, sans a headless desktop PC running Windows Server 2019 that hosts Blue Iris. Everything on the CPAI side is Debian based Linux running Docker containers.

 

[MikeLud1]

Hey all, it's me again. I've got about 50 hours of troubleshooting invested in trying to get CPAI to work on my ridiculously overcomplicated setup, and I'm very happy to report that I've reached the 95% mark! After completely rebuilding my linux container that hosts docker 3 times, and installing easily 20 different versions of CPAI, finally found a couple Nvidia/docker related things that were missing. CPAI runs in docker on my proxmox node using GPU sharing with CUDA! MOSTLY, which is why I'm back with one last issue:

Face Processing and YOLO v5 6.2 both work with CUDA GPU.
License Plate Reader 2.6, refuses to.

I've uninstalled and reinstalled the license plate reader module multiple times watching the logs, forced it to download without using local cache, still can't get it working with CUDA, the info tab says "Support Disabled" for GPU. This machine has 3 GPUs in it, two CUDA enabled. I'm manually assigning a GPU not accessible by any other device/process to weed out interference or out of memory issues.

Host is Proxmox PVE6.4 Bare metal
Docker running in a privlidged LXC container with full access to the GPUs (ubuntu 20.04)
CPAI: 2.2.4-Beta
Run command: docker run --name CodeProject.AI-Server3 -d -p 32168:32168 --gpus device=GPU-eb6187de-dff9-83e5-ce33-140d9e466b12 codeproject/ai-server:gpu
CUDA 11.7, 515.105.01 Driver
GPU: NVidia Tesla P4 8gb

View attachment 181894
View attachment 181895

Most recent install log for License Plate Reader:

Installation Log
2024-01-06 00:59:06: Hi Docker! We will disable shared python installs for downloaded modules
2024-01-06 00:59:06: No schemas installed
2024-01-06 00:59:06: (No schemas means: we can't detect if you're in light or dark mode)
2024-01-06 00:59:06: sh: 1: lsmod: not found
2024-01-06 00:59:06: Installing CodeProject.AI Analysis Module
2024-01-06 00:59:06: ======================================================================
2024-01-06 00:59:06: CodeProject.AI Installer
2024-01-06 00:59:06: ======================================================================
2024-01-06 00:59:06: 253.02 GiB available
2024-01-06 00:59:06: Checking GPU support
2024-01-06 00:59:06: CUDA Present...Yes
2024-01-06 00:59:06: Allowing GPU Support: Yes
2024-01-06 00:59:06: Allowing CUDA Support: Yes
2024-01-06 00:59:06: General CodeProject.AI setup
2024-01-06 00:59:06: Setting permissions...Done
2024-01-06 00:59:06: Processing SDK
2024-01-06 00:59:12: Adding toolchain repo... Done
2024-01-06 00:59:15: Checking for installed dependencies (from gcc-4.9 apt-utils)...Done
2024-01-06 00:59:20: Installing missing dependencies: gcc-4.9...Done
2024-01-06 00:59:22: Upgrading libstdc... Done
2024-01-06 00:59:34: Checking for installed dependencies (from ca-certificates gnupg libc6-dev libfontconfig1 libgdiplus libjpeg-dev zlib1g-dev)...Done
2024-01-06 00:59:34: All dependencies already installed.
2024-01-06 00:59:48: Checking for installed dependencies (from ffmpeg libsm6 libxext6 mesa-utils curl rsync unzip wget)...Done
2024-01-06 00:59:48: All dependencies already installed.
2024-01-06 00:59:48: The command could not be loaded, possibly because:
2024-01-06 00:59:48: * You intended to execute a .NET application:
2024-01-06 00:59:48: The application '--version' does not exist.
2024-01-06 00:59:48: * You intended to execute a .NET SDK command:
2024-01-06 00:59:48: No .NET SDKs were found.
2024-01-06 00:59:48: Download a .NET SDK:
2024-01-06 00:59:48: .NET Downloads (Linux, macOS, and Windows)
2024-01-06 00:59:48: Learn about SDK resolution:
2024-01-06 00:59:48: Select which .NET version to use - .NET
2024-01-06 00:59:48: Checking for .NET >= 7.0...All good. Current .NET is , requested was 7.0
2024-01-06 00:59:48: Installing module ALPR 2.6
2024-01-06 00:59:48: Python 3.8 is already installed
2024-01-06 00:59:50: Installing Virtual Environment tools for Linux... Done
2024-01-06 00:59:55: Creating Virtual Environment (Local)... Done
2024-01-06 00:59:55: Checking for Python 3.8...Found Python 3.8.18. present
2024-01-06 00:59:55: Checking for CUDA...CUDA Present
2024-01-06 00:59:55: CUDA version is:
2024-01-06 00:59:57: Installing PIP... Done
2024-01-06 01:00:00: Ensuring PIP compatibility... Done
2024-01-06 01:00:04: Updating PIP... Done
2024-01-06 01:00:04: Installing setuptools... Done
2024-01-06 01:00:04: Choosing packages from requirements.linux.txt
2024-01-06 01:01:20: Installing Packages into Virtual Environment... Success
2024-01-06 01:01:21: Downloading OCR models... already exists...Expanding... Done.
2024-01-06 01:01:21: Applying PaddleOCR patch
2024-01-06 01:01:21: Installing Server SDK support:
2024-01-06 01:01:21: Checking for CUDA...CUDA Present
2024-01-06 01:01:21: CUDA version is:
2024-01-06 01:01:23: Installing PIP... Done
2024-01-06 01:01:26: Ensuring PIP compatibility... Done
2024-01-06 01:01:30: Updating PIP... Done
2024-01-06 01:01:31: Installing setuptools... Done
2024-01-06 01:01:31: Choosing packages from requirements.txt
2024-01-06 01:01:39: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/requests already exists. Specify --upgrade to force replacement.
2024-01-06 01:01:39: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/Pillow.libs already exists. Specify --upgrade to force replacement.
2024-01-06 01:01:39: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/idna already exists. Specify --upgrade to force replacement.
2024-01-06 01:01:39: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/certifi-2023.11.17.dist-info already exists. Specify --upgrade to force replacement.
2024-01-06 01:01:39: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/urllib3-2.1.0.dist-info already exists. Specify --upgrade to force replacement.
2024-01-06 01:01:39: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/PIL already exists. Specify --upgrade to force replacement.
2024-01-06 01:01:39: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/urllib3 already exists. Specify --upgrade to force replacement.
2024-01-06 01:01:39: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/requests-2.31.0.dist-info already exists. Specify --upgrade to force replacement.
2024-01-06 01:01:39: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/certifi already exists. Specify --upgrade to force replacement.
2024-01-06 01:01:39: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/Pillow-9.5.0.dist-info already exists. Specify --upgrade to force replacement.
2024-01-06 01:01:39: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/idna-3.6.dist-info already exists. Specify --upgrade to force replacement.
2024-01-06 01:01:39: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/charset_normalizer already exists. Specify --upgrade to force replacement.
2024-01-06 01:01:39: WARNING: Target directory /app/modules/ALPR/bin/linux/python38/venv/lib/python3.8/site-packages/bin already exists. Specify --upgrade to force replacement.
2024-01-06 01:01:39: Installing Packages into Virtual Environment... Success
2024-01-06 01:01:39: Setup complete
Installer exited with code 0

System Info Tab:
View attachment 181896

View attachment 181897

The ALPR module at this time only supports GPU Compute >= 7.0 your GPU is 6.1

 

[elesjuan]

The ALPR module at this time only supports GPU Compute >= 7.0 your GPU is 6.1

View attachment 181907

Dammit. Hopefully this is where you're going to tell me that a future version of ALPR will support compute 6.1? I certainly hope...


Either way, I GREATLY appreciate your input. I can stop banging my head on the wall, for now at least!

 

[kaltertod]

Appreciate the input, but there is not a Windows machine involved in any of this, sans a headless desktop PC running Windows Server 2019 that hosts Blue Iris. Everything on the CPAI side is Debian based Linux running Docker containers.

Duly Noted, I did not note the OS info in the screen shots, that is my fault. I just assumed it was windows, we deal with more windows users here than the linux variants.

 

[prsmith777]

The ALPR module at this time only supports GPU Compute >= 7.0 your GPU is 6.1

This is news to me. Explains why I could never get my 1060ti with compute 6.1 to work. Upgraded to 3060 and works fine

 

[MikeLud1]

This is news to me. Explains why I could never get my 1060ti with compute 6.1 to work. Upgraded to 3060 and works fine

When I was writing the code I only had 3060, 3060ti, 3090, and 4090 GPUs to test with. Chris from CP.AI added in the restriction into the code because older GPU was not working.

 

[elesjuan]

This is news to me. Explains why I could never get my 1060ti with compute 6.1 to work. Upgraded to 3060 and works fine

I find it moderately frustrating that despite hours of reading about CUDA with CPAI, I never saw this specific information anywhere. My selection of the Tesla P4 was based on cost, performance, and more specifically power requirements as they're installed in a commercial server chassis with redundant PSUs and no provision for GPU aux power. Everything I read lead me to believe the P4 should be fully compatible with CPAI, which to be fair, YOLO 6.2 is smokin' fast with CUDA. My NVR runs on a SFF desktop i7-8700 with 10 1080p cameras and 2 8MP cameras, all cameras are decoding with Intel+VPP so the CPU usage is almost nothing, until I started up CPAI. On low percentage lows, you could expect 2000ms results, but my plate cams were 10000ms on a good day, IF it didn't time out before resolving. Still running the LPR via CPAI on the desktop, but moved all the other object identification off to the docker container and the highest time I've seen in the last few hours was around 400ms with averages below 150ms.

Definitely happy that I at least got YOLO working, which took a HUGE load off the NVRs poor CPU.