CodeProject.AI Version 2.0

cyberwolf_uk said:
I have a gt1030 running in my box and it works fine with GPU running YOLOv5.6.2 on 2.0.8, 2.1.6 & now 2.1.8. I must admit it did have all sorts of issues with detection etc on my test box running the same setup until I did a clean install of everything, I've mentioned it in a thread somewhere on here.

Here are the versions

View attachment 161952

View attachment 161953
Click to expand...


Having huge issues myself.

Removed an old 1.5x version of CPAI today and installed the 2.16 Beta from scratch as I had an old Nvidia GPU knocking about and I want to offload the AI to that.

I cannot get CUDA working for love nor money.


19:36:14:ObjectDetectionYolo: Installing CodeProject.AI Analysis Module
19:36:14:ObjectDetectionYolo: ========================================================================
19:36:14:ObjectDetectionYolo: CodeProject.AI Installer
19:36:14:ObjectDetectionYolo: ========================================================================
19:36:14:ObjectDetectionYolo: CUDA Present...True
19:36:14:ObjectDetectionYolo: Allowing GPU Support: Yes
19:36:15:ObjectDetectionYolo: Allowing CUDA Support: Yes
19:36:15:ObjectDetectionYolo: General CodeProject.AI setup
19:36:15:ObjectDetectionYolo: Creating Directories...Done


Server version: 2.1.6-Beta
Operating System: Windows (Microsoft Windows 11 version 10.0.22621)
CPUs: Intel(R) Core(TM) i7-8700K CPU @ 3.70GHz
1 CPU x 6 cores. 12 logical processors (x64)
GPU: NVIDIA GeForce RTX 2080 SUPER (8 GiB) (NVidia)
Driver: 531.79 CUDA: 12.1 Compute: 7.5
System RAM: 32 GiB
Target: Windows
BuildConfig: Release
Execution Env: Native
Runtime Env: Production
.NET framework: .NET 7.0.5
System GPU info:
GPU 3D Usage 3%
GPU RAM Usage 464 MiB
Video adapter info:
NVIDIA GeForce RTX 2080 SUPER:
Driver Version 31.0.15.3179
Video Processor NVIDIA GeForce RTX 2080 SUPER
Intel(R) UHD Graphics 630:
Driver Version 31.0.101.2111
Video Processor Intel(R) UHD Graphics Family
Global Environment variables:
CPAI_APPROOTPATH = C:\Program Files\CodeProject\AI
CPAI_PORT = 32168

I notice it's also picked up the CPU (in bold above) display adapter but everything seems to point at the GPU being enabled for processing but starting the YOLO always results in it using CPU.


19:39:46:Module 'Object Detection (YOLOv5 6.2)' (ID: ObjectDetectionYolo)
19:39:46:Module Path: C:\Program Files\CodeProject\AI\modules\ObjectDetectionYolo
19:39:46:AutoStart: True
19:39:46:Queue: objectdetection_queue
19:39:46:Platforms: all
19:39:46:GPU: Support enabled
19:39:46:Parallelism: 0
19:39:46:Accelerator:
19:39:46:Half Precis.: enable
19:39:46:Runtime: python37
19:39:46:Runtime Loc: Shared
19:39:46:FilePath: detect_adapter.py
19:39:46:Pre installed: False
19:39:46:Start pause: 1 sec
19:39:46:LogVerbosity:
19:39:46:Valid: True
19:39:46:Environment Variables
19:39:46:APPDIR = %CURRENT_MODULE_PATH%
19:39:46:CUSTOM_MODELS_DIR = %CURRENT_MODULE_PATH%/custom-models
19:39:46:MODELS_DIR = %CURRENT_MODULE_PATH%/assets
19:39:46:MODEL_SIZE = Medium
19:39:46:USE_CUDA = True

19:46:46:detect_adapter.py: Inference processing will occur on device 'CPU'
19:46:46:detect_adapter.py: Inference processing will occur on device 'CPU'
19:46:46:detect_adapter.py: Inference processing will occur on device 'CPU'

cpu.png


Anyone got any ideas?

Thanks,

Craig
 
Craig G said:
Having huge issues myself.

Removed an old 1.5x version of CPAI today and installed the 2.16 Beta from scratch as I had an old Nvidia GPU knocking about and I want to offload the AI to that.

I cannot get CUDA working for love nor money.


19:36:14:ObjectDetectionYolo: Installing CodeProject.AI Analysis Module
19:36:14:ObjectDetectionYolo: ========================================================================
19:36:14:ObjectDetectionYolo: CodeProject.AI Installer
19:36:14:ObjectDetectionYolo: ========================================================================
19:36:14:ObjectDetectionYolo: CUDA Present...True
19:36:14:ObjectDetectionYolo: Allowing GPU Support: Yes
19:36:15:ObjectDetectionYolo: Allowing CUDA Support: Yes
19:36:15:ObjectDetectionYolo: General CodeProject.AI setup
19:36:15:ObjectDetectionYolo: Creating Directories...Done


Server version: 2.1.6-Beta
Operating System: Windows (Microsoft Windows 11 version 10.0.22621)
CPUs: Intel(R) Core(TM) i7-8700K CPU @ 3.70GHz
1 CPU x 6 cores. 12 logical processors (x64)
GPU: NVIDIA GeForce RTX 2080 SUPER (8 GiB) (NVidia)
Driver: 531.79 CUDA: 12.1 Compute: 7.5
System RAM: 32 GiB
Target: Windows
BuildConfig: Release
Execution Env: Native
Runtime Env: Production
.NET framework: .NET 7.0.5
System GPU info:
GPU 3D Usage 3%
GPU RAM Usage 464 MiB
Video adapter info:
NVIDIA GeForce RTX 2080 SUPER:
Driver Version 31.0.15.3179
Video Processor NVIDIA GeForce RTX 2080 SUPER
Intel(R) UHD Graphics 630:
Driver Version 31.0.101.2111
Video Processor Intel(R) UHD Graphics Family
Global Environment variables:
CPAI_APPROOTPATH = C:\Program Files\CodeProject\AI
CPAI_PORT = 32168

I notice it's also picked up the CPU (in bold above) display adapter but everything seems to point at the GPU being enabled for processing but starting the YOLO always results in it using CPU.


19:39:46:Module 'Object Detection (YOLOv5 6.2)' (ID: ObjectDetectionYolo)
19:39:46:Module Path: C:\Program Files\CodeProject\AI\modules\ObjectDetectionYolo
19:39:46:AutoStart: True
19:39:46:Queue: objectdetection_queue
19:39:46:platforms: all
19:39:46:GPU: Support enabled
19:39:46:parallelism: 0
19:39:46:Accelerator:
19:39:46:Half Precis.: enable
19:39:46:Runtime: python37
19:39:46:Runtime Loc: Shared
19:39:46:FilePath: detect_adapter.py
19:39:46:pre installed: False
19:39:46:Start pause: 1 sec
19:39:46:LogVerbosity:
19:39:46:Valid: True
19:39:46:Environment Variables
19:39:46:APPDIR = %CURRENT_MODULE_PATH%
19:39:46:CUSTOM_MODELS_DIR = %CURRENT_MODULE_PATH%/custom-models
19:39:46:MODELS_DIR = %CURRENT_MODULE_PATH%/assets
19:39:46:MODEL_SIZE = Medium
19:39:46:USE_CUDA = True

19:46:46:detect_adapter.py: Inference processing will occur on device 'CPU'
19:46:46:detect_adapter.py: Inference processing will occur on device 'CPU'
19:46:46:detect_adapter.py: Inference processing will occur on device 'CPU'

View attachment 161976


Anyone got any ideas?

Thanks,

Craig
Click to expand...
Have your tried YOLOv5.NET
 
David L said:
Have your tried YOLOv5.NET
Click to expand...
Hi David,

I haven't since that specifically says in the description of the module 'MIT Provides Object Detection using YOLOv5 ONNX models with DirectML. This module is best for those on Windows and Linux without CUDA enabled GPUs '

I'll try anything at this point now.

I rolled back CUDA 12.1 to 11.7 and ran CUDnn - still nothing.

I'll give it a go!

Thanks
 
  • Like
Reactions: David L
Craig G said:
Hi David,

I haven't since that specifically says in the description of the module 'MIT Provides Object Detection using YOLOv5 ONNX models with DirectML. This module is best for those on Windows and Linux without CUDA enabled GPUs '

I'll try anything at this point now.

I rolled back CUDA 12.1 to 11.7 and ran CUDnn - still nothing.

I'll give it a go!

Thanks
Click to expand...
Nope, exactly the same.

Going to remove CPAI entirely and reinstall
 
Craig G said:
Hi David,

I haven't since that specifically says in the description of the module 'MIT Provides Object Detection using YOLOv5 ONNX models with DirectML. This module is best for those on Windows and Linux without CUDA enabled GPUs '

I'll try anything at this point now.

I rolled back CUDA 12.1 to 11.7 and ran CUDnn - still nothing.

I'll give it a go!

Thanks
Click to expand...
So I apologize ahead of time for anyone I mislead but I was told to try .NET and use ONNX for the custom models. Also be sure and turn off 1683315536482.png if doing so. But I am not on the latest Beta.
 
cyberwolf_uk said:
@Craig G did you use the cuDNN install script
Click to expand...
Hi Cyber, I did indeed.

I just removed CPAI, deleted folder, rebooted and re-installed.

Installed the .NET module this time but still stuck on CPU.

this is going to eat my bank holiday weekend :)

Although....

I removed all CUDA 12.1 components and installed 11.7 and ran the cuDNN bat file but CPAI still reports.....

GPU: NVIDIA GeForce RTX 2080 SUPER (8 GiB) (NVidia)
Driver: 531.79 CUDA: 12.1 Compute: 7.5
 
  • Like
Reactions: David L
Craig G said:
Hi Cyber, I did indeed.

I just removed CPAI, deleted folder, rebooted and re-installed.

Installed the .NET module this time but still stuck on CPU.

this is going to eat my bank holiday weekend :)

Although....

I removed all CUDA 12.1 components and installed 11.7 and ran the cuDNN bat file but CPAI still reports.....

GPU: NVIDIA GeForce RTX 2080 SUPER (8 GiB) (NVidia)
Driver: 531.79 CUDA: 12.1 Compute: 7.5
Click to expand...
You deleted both CodeProject folders in:

Program Files
ProgramData
 
David L said:
You deleted both CodeProject folders in:

Program Files
ProgramData
Click to expand...
Didn't, but have now.

I've removed cuda 11.7 and trying 12.x again as there's a cuDNN for that too since CPAI thought I was on that version.

So going for full CPAI reinstall 'after' CUDA 12.1 and cuDNN
 
  • Like
Reactions: David L
Craig G said:
Didn't, but have now.

I've removed cuda 11.7 and trying 12.x again as there's a cuDNN for that too since CPAI thought I was on that version.

So going for full CPAI reinstall 'after' CUDA 12.1 and cuDNN
Click to expand...
Damn, Nvidia want to sign up to their Dev network in order to download the latest cuDNN
 
  • Sad
Reactions: David L
tedrpi said:
Make sure you run the cuDNN install script, I had to run the script again when I upgraded to 2.1.6 despite my GPU working in previous versions.

File Download - CodeProject

www.codeproject.com www.codeproject.com
Click to expand...
Craig G said:
Hi Cyber, I did indeed.

I just removed CPAI, deleted folder, rebooted and re-installed.

Installed the .NET module this time but still stuck on CPU.

this is going to eat my bank holiday weekend :)

Although....

I removed all CUDA 12.1 components and installed 11.7 and ran the cuDNN bat file but CPAI still reports.....

GPU: NVIDIA GeForce RTX 2080 SUPER (8 GiB) (NVidia)
Driver: 531.79 CUDA: 12.1 Compute: 7.5
Click to expand...

I must admit on my test box before I reinstalled it from scratch, formatted the drive and installed everything from scratch I faced all sorts of issues with any version of CP when trying to get face detection working.

GPU wouldn't work, detection wouldn't work then only .net would work. After spending a day on and off trying to troubleshot it, I gave up and went for the fresh install which fixed all my issues. But for me it was simple as the test box is only used for that so the wipe meant nothing
 
  • Wow
Reactions: David L
So the latest cuDNN is just a zip file (after I signed up) with the libraries in it.

There's no BAT file to alter system variables. I suppose I could edit the scripts in the original BAT file but it's 9PM on Friday night, so.... NO

Going to install 11.7 again and cuDNN BAT before even installing CPAI (after deleting both folder locations)
 
Well guess I am glad I am moving to Windows Server for my BI box since you all have been having issues, so CP.AI will be a fresh install. Just hope I don't have any Video Driver issues with Windows Server 2016.

My homework will be to figure out how to properly backup BI.

1683317987053.png

Is this .reg file all I need? Does it have my Camera settings? Should I uninstall AI so when I Restore/Import this reg file I don't mess up 2.1.8-Beta install?
Haha, to many what ifs? :)
 
Craig G said:
Having huge issues myself.

Removed an old 1.5x version of CPAI today and installed the 2.16 Beta from scratch as I had an old Nvidia GPU knocking about and I want to offload the AI to that.

I cannot get CUDA working for love nor money.


19:36:14:ObjectDetectionYolo: Installing CodeProject.AI Analysis Module
19:36:14:ObjectDetectionYolo: ========================================================================
19:36:14:ObjectDetectionYolo: CodeProject.AI Installer
19:36:14:ObjectDetectionYolo: ========================================================================
19:36:14:ObjectDetectionYolo: CUDA Present...True
19:36:14:ObjectDetectionYolo: Allowing GPU Support: Yes
19:36:15:ObjectDetectionYolo: Allowing CUDA Support: Yes
19:36:15:ObjectDetectionYolo: General CodeProject.AI setup
19:36:15:ObjectDetectionYolo: Creating Directories...Done


Server version: 2.1.6-Beta
Operating System: Windows (Microsoft Windows 11 version 10.0.22621)
CPUs: Intel(R) Core(TM) i7-8700K CPU @ 3.70GHz
1 CPU x 6 cores. 12 logical processors (x64)
GPU: NVIDIA GeForce RTX 2080 SUPER (8 GiB) (NVidia)
Driver: 531.79 CUDA: 12.1 Compute: 7.5
System RAM: 32 GiB
Target: Windows
BuildConfig: Release
Execution Env: Native
Runtime Env: Production
.NET framework: .NET 7.0.5
System GPU info:
GPU 3D Usage 3%
GPU RAM Usage 464 MiB
Video adapter info:
NVIDIA GeForce RTX 2080 SUPER:
Driver Version 31.0.15.3179
Video Processor NVIDIA GeForce RTX 2080 SUPER
Intel(R) UHD Graphics 630:
Driver Version 31.0.101.2111
Video Processor Intel(R) UHD Graphics Family
Global Environment variables:
CPAI_APPROOTPATH = C:\Program Files\CodeProject\AI
CPAI_PORT = 32168

I notice it's also picked up the CPU (in bold above) display adapter but everything seems to point at the GPU being enabled for processing but starting the YOLO always results in it using CPU.


19:39:46:Module 'Object Detection (YOLOv5 6.2)' (ID: ObjectDetectionYolo)
19:39:46:Module Path: C:\Program Files\CodeProject\AI\modules\ObjectDetectionYolo
19:39:46:AutoStart: True
19:39:46:Queue: objectdetection_queue
19:39:46:platforms: all
19:39:46:GPU: Support enabled
19:39:46:parallelism: 0
19:39:46:Accelerator:
19:39:46:Half Precis.: enable
19:39:46:Runtime: python37
19:39:46:Runtime Loc: Shared
19:39:46:FilePath: detect_adapter.py
19:39:46:pre installed: False
19:39:46:Start pause: 1 sec
19:39:46:LogVerbosity:
19:39:46:Valid: True
19:39:46:Environment Variables
19:39:46:APPDIR = %CURRENT_MODULE_PATH%
19:39:46:CUSTOM_MODELS_DIR = %CURRENT_MODULE_PATH%/custom-models
19:39:46:MODELS_DIR = %CURRENT_MODULE_PATH%/assets
19:39:46:MODEL_SIZE = Medium
19:39:46:USE_CUDA = True

19:46:46:detect_adapter.py: Inference processing will occur on device 'CPU'
19:46:46:detect_adapter.py: Inference processing will occur on device 'CPU'
19:46:46:detect_adapter.py: Inference processing will occur on device 'CPU'

View attachment 161976


Anyone got any ideas?

Thanks,

Craig
Click to expand...
Why are you not installing 2.1.8?

I saw this in your post:
1683318328265.png

Just released yesterday:

CodeProject.AI Version 2.0

I shut down BI... and then I uninstalled v2.0.8. Downloaded v2.1.7 and installed it. Waited until all modules finished installing etc. which took about 10 minutes. I have an NVIDIA P400 GPU, should I stick with Object Detection YOLOv5 6.2 for best accuracy and speed for object detection. The...
ipcamtalk.com ipcamtalk.com
 
Craig G said:
So the latest cuDNN is just a zip file (after I signed up) with the libraries in it.

There's no BAT file to alter system variables. I suppose I could edit the scripts in the original BAT file but it's 9PM on Friday night, so.... NO

Going to install 11.7 again and cuDNN BAT before even installing CPAI (after deleting both folder locations)
Click to expand...

www.codeproject.com

CodeProject.AI Server: AI the easy way.

Version 2.6.5. Our fast, free, self-hosted Artificial Intelligence Server for any platform, any language
www.codeproject.com www.codeproject.com

The cudnn bat is from here
 
  • Like
Reactions: David L
I think I was using it because that's the one that's linked from the CPAI server UI. I assumed it was the latest but I guess the URL in the UI isn't up to date
 
  • Like
Reactions: David L
You must log in or register to reply here.
Top Bottom