|
I constantly get this error caused by various errors:
01:28:20:objectdetection_coral_adapter.py: F driver/usb/usb_driver.cc:815] [0-0] bulk out for header failed. Abort. Unavailable: USB error -4 [AsyncBulkOutTransfer]
01:28:20:Module ObjectDetectionCoral has shutdown
02:31:33:objectdetection_coral_adapter.py: F driver/usb/usb_driver.cc:1148] HandleQueuedBulkIn transfer in failed. Not found: USB transfer error 5 [LibUsbDataInCallback]
02:31:34:Module ObjectDetectionCoral has shutdown
It starts working again if I restart the Docker container.
The container should detect this kind of thing and restart itself, instead of silently failing.
modified 20-Mar-24 11:12am.
|
|
|
|
|
This is on our TODO list
cheers
Chris Maunder
|
|
|
|
|
Blue Iris 5.8.8.8
Code Project AI 2.5.6
Windows 10 22H2
Unable to continue using the "Burn Label mark-up onto alert images" feature without causing a "framerate bottleneck" error on the camera preview sub stream that freezes the preview and stops processing further AI motion detection alerts until feed unfreezes in about 80 seconds.
Anyone familiar and if this is a Windows Defender exception needed that I am missing?
Not sure if this is more of a Blue Iris issue or Code Project AI issue.
|
|
|
|
|
This is definitely a Blue Iris issue. I've tried poking around some Blue Iris forums for solutions, but was not able to find any.
Thanks,
Sean Ewington
CodeProject
|
|
|
|
|
I have been using CodeProject AI 2.5.4 (I haven't update it yet) and it has been running YOLOv5 6.2 (1.9.1) and ALPR (3.0.2) on a 24/7 basis. The number of detections is over 100,000. However, every once in a while, I will get the CPAI error. I checked the log and it says:
24:19:21:Object Detection (YOLOv5 6.2): [RuntimeError] : Traceback (most recent call last):
File "C:\Program Files\CodeProject\AI\modules\ObjectDetectionYOLOv5-6.2\detect.py", line 141, in do_detection
det = detector(img, size=640)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\nn\modules\module.py", line 1190, in _call_impl
return forward_call(*input, **kwargs)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context
return func(*args, **kwargs)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\yolov5\models\common.py", line 705, in forward
y = self.model(x, augment=augment) # forward
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\nn\modules\module.py", line 1190, in _call_impl
return forward_call(*input, **kwargs)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\yolov5\models\common.py", line 515, in forward
y = self.model(im, augment=augment, visualize=visualize) if augment or visualize else self.model(im)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\nn\modules\module.py", line 1190, in _call_impl
return forward_call(*input, **kwargs)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\Lib\site-packages\yolov5\models\yolo.py", line 209, in forward
return self._forward_once(x, profile, visualize) # single-scale inference, train
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\Lib\site-packages\yolov5\models\yolo.py", line 121, in _forward_once
x = m(x) # run
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\lib\site-packages\torch\nn\modules\module.py", line 1190, in _call_impl
return forward_call(*input, **kwargs)
File "C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\Lib\site-packages\yolov5\models\yolo.py", line 74, in forward
xy = (xy * 2 + self.grid[i]) * self.stride[i] # xy
RuntimeError: The size of tensor a (60) must match the size of tensor b (32) at non-singleton dimension 2
I think this error is a bit deeper than something I can handle myself, so I wanted to see if anyone has any idea on how to solve this?
Update:
I just found out that YOLOv8 also supports ALPR and moved over to YOVOv8. We will see if YOLOv8 also gives me the same error or not.
modified 19-Mar-24 11:29am.
|
|
|
|
|
We've seen this error occasionally and haven't found the cause. How's YOLOv8?
cheers
Chris Maunder
|
|
|
|
|
I see. YOLOv8 is running very well, I have been running it 2 days in a row and did not get any error.
I also noticed it's performance is way better than YOLOv5! It can detect the plate that is on the other side of the street for some reason and most of the time is accurate! The give comparison, normally, the LPR detects one lane, occasionally 2 lanes I think (technology evolves so maybe 2 lanes are normal now). But with YOLOv8, it does 3 lanes, which I never seen it before!
|
|
|
|
|
I'm trying to get YOLOv5 6.2 to install on a clean install of CPAI 2.5.6.
After initial installation both Face Processing and Object Detection (YOLOv5 6.2) failed to start, giving the error "ModuleNotFoundError: No module named 'aiohttp' ".
I don't use Face processing, so I disabled it, but need to get Object Detection working.
Chris recommended uninstalling and reinstalling Object Detection in a post about the aiohttp module not found error. In another post it was mentioned to uninstall and then reinstall with download cache disabled.
I think I've now done 4 uninstall/reinstalls of ObjectDetectionYOLOv5-6.2 (which on my slow internet connection takes 20+ minutes) both with and without download cache enabled.
I even tried shutting down the service and deleting the C:\Program Files\CodeProject\AI\modules\ObjectDetectionYOLOv5-6.2 directory before trying the install with download cache disabled.
Each time the install gets part way through the process, throws the timeout error, but eventually comes back and says that "Module ObjectDetectionYOLOv5-6.2 installed successfully."
Unfortunately, when it tries to start Object Detection it still throws the ModuleNotFoundError, and detect_adapter.py exits.
Here is my log after the latest attempt:
13:53:22:System: Windows
13:53:22:Operating System: Windows (Microsoft Windows 11 version 10.0.22631)
13:53:22:CPUs: 13th Gen Intel(R) Core(TM) i7-13700K (Intel)
13:53:22: 1 CPU x 16 cores. 24 logical processors (x64)
13:53:22:GPU (Primary): NVIDIA GeForce RTX 4070 (12 GiB) (NVIDIA)
13:53:22: Driver: 546.12, CUDA: 12.3 (up to: 12.3), Compute: 8.9, cuDNN: 8.9
13:53:22:System RAM: 32 GiB
13:53:22:Platform: Windows
13:53:22:BuildConfig: Release
13:53:22:Execution Env: Native
13:53:22:Runtime Env: Production
13:53:22:Runtimes installed:
13:53:22: .NET runtime: 8.0.0
13:53:22: .NET SDK: 8.0.100
13:53:22: Default Python: Not found
13:53:22: Go: Not found
13:53:22: NodeJS: Not found
13:53:22:App DataDir: C:\ProgramData\CodeProject\AI
13:53:22:Video adapter info:
13:53:22: NVIDIA GeForce RTX 4070:
13:53:22: Driver Version 31.0.15.4612
13:53:22: Video Processor NVIDIA GeForce RTX 4070
13:53:22:STARTING CODEPROJECT.AI SERVER
13:53:22:RUNTIMES_PATH = C:\Program Files\CodeProject\AI\runtimes
13:53:22:PREINSTALLED_MODULES_PATH = C:\Program Files\CodeProject\AI\preinstalled-modules
13:53:22:MODULES_PATH = C:\Program Files\CodeProject\AI\modules
13:53:22:PYTHON_PATH = \bin\windows\%PYTHON_NAME%\venv\Scripts\python
13:53:22:Data Dir = C:\ProgramData\CodeProject\AI
13:53:22:Server version: 2.5.6
13:53:25:
13:53:25:Module 'License Plate Reader' 3.0.2 (ID: ALPR)
13:53:25:Valid: True
13:53:25:Module Path: <root>\modules\ALPR
13:53:25:AutoStart: True
13:53:25:Queue: alpr_queue
13:53:25:Runtime: python3.9
13:53:25:Runtime Loc: Local
13:53:25:FilePath: ALPR_adapter.py
13:53:25:Pre installed: False
13:53:25:Start pause: 3 sec
13:53:25:Parallelism: 0
13:53:25:LogVerbosity:
13:53:25:Platforms: all
13:53:25:GPU Libraries: installed if available
13:53:25:GPU Enabled: enabled
13:53:25:Accelerator:
13:53:25:Half Precis.: enable
13:53:25:Environment Variables
13:53:25:AUTO_PLATE_ROTATE = True
13:53:25:MIN_COMPUTE_CAPABILITY = 6
13:53:25:MIN_CUDNN_VERSION = 7
13:53:25:OCR_OPTIMAL_CHARACTER_HEIGHT = 60
13:53:25:OCR_OPTIMAL_CHARACTER_WIDTH = 30
13:53:25:OCR_OPTIMIZATION = True
13:53:25:PLATE_CONFIDENCE = 0.7
13:53:25:PLATE_RESCALE_FACTOR = 2
13:53:25:PLATE_ROTATE_DEG = 0
13:53:25:
13:53:25:Started License Plate Reader module
13:53:28:Server: This is the latest version
13:53:59:Preparing to install module 'ObjectDetectionYOLOv5-6.2'
13:53:59:Downloading module 'ObjectDetectionYOLOv5-6.2'
13:53:59:Installing module 'ObjectDetectionYOLOv5-6.2'
13:54:00:ObjectDetectionYOLOv5-6.2: Installing CodeProject.AI Analysis Module
13:54:00:ObjectDetectionYOLOv5-6.2: ======================================================================
13:54:00:ObjectDetectionYOLOv5-6.2: CodeProject.AI Installer
13:54:00:ObjectDetectionYOLOv5-6.2: ======================================================================
13:54:00:ObjectDetectionYOLOv5-6.2: 1582.6Gb of 1906Gb available on
13:54:00:ObjectDetectionYOLOv5-6.2: os, arch = windows x86_64
13:54:00:ObjectDetectionYOLOv5-6.2: systemName, platform = Windows, windows
13:54:00:ObjectDetectionYOLOv5-6.2: setupMode = InstallModule
13:54:00:ObjectDetectionYOLOv5-6.2: executionEnvironment = Production
13:54:00:ObjectDetectionYOLOv5-6.2: rootDirPath = C:\Program Files\CodeProject\AI
13:54:00:ObjectDetectionYOLOv5-6.2: appRootDirPath = C:\Program Files\CodeProject\AI\
13:54:00:ObjectDetectionYOLOv5-6.2: setupScriptDirPath = C:\Program Files\CodeProject\AI\
13:54:00:ObjectDetectionYOLOv5-6.2: sdkScriptsDirPath = C:\Program Files\CodeProject\AI\SDK\Scripts
13:54:00:ObjectDetectionYOLOv5-6.2: runtimesDirPath = C:\Program Files\CodeProject\AI\runtimes
13:54:00:ObjectDetectionYOLOv5-6.2: modulesDirPath = C:\Program Files\CodeProject\AI\modules
13:54:00:ObjectDetectionYOLOv5-6.2: downloadDirPath = C:\Program Files\CodeProject\AI\downloads
13:54:00:ObjectDetectionYOLOv5-6.2: General CodeProject.AI setup
13:54:00:ObjectDetectionYOLOv5-6.2: Creating Directories...Done
13:54:00:ObjectDetectionYOLOv5-6.2: GPU support
13:54:00:ObjectDetectionYOLOv5-6.2: CUDA Present...Yes (CUDA 12.3, cuDNN 8.9)
13:54:00:ObjectDetectionYOLOv5-6.2: ROCm Present...No
13:54:00:ObjectDetectionYOLOv5-6.2: Reading ObjectDetectionYOLOv5-6.2 settingsUsed modulesettings.json to get value Object Detection (YOLOv5 6.2)
13:54:01:ObjectDetectionYOLOv5-6.2: .Used modulesettings.json to get value 1.9.1
13:54:01:ObjectDetectionYOLOv5-6.2: .Used modulesettings.windows.json to get value python3.7
13:54:01:ObjectDetectionYOLOv5-6.2: .Used modulesettings.json to get value Shared
13:54:01:ObjectDetectionYOLOv5-6.2: .Used modulesettings.json to get value detect_adapter.py
13:54:01:ObjectDetectionYOLOv5-6.2: .Used modulesettings.json to get value true
13:54:01:ObjectDetectionYOLOv5-6.2: .Used modulesettings.json to get value ["all","raspberrypi","jetson"]
13:54:01:ObjectDetectionYOLOv5-6.2: .Done
13:54:01:ObjectDetectionYOLOv5-6.2: Installing module Object Detection (YOLOv5 6.2) 1.9.1
13:54:01:ObjectDetectionYOLOv5-6.2: Variable Dump
13:54:01:ObjectDetectionYOLOv5-6.2: moduleName = Object Detection (YOLOv5 6.2)
13:54:01:ObjectDetectionYOLOv5-6.2: moduleVersion = 1.9.1
13:54:02:ObjectDetectionYOLOv5-6.2: runtime = python3.7
13:54:02:ObjectDetectionYOLOv5-6.2: runtimeLocation = Shared
13:54:02:ObjectDetectionYOLOv5-6.2: installGPU = true
13:54:02:ObjectDetectionYOLOv5-6.2: pythonVersion = 3.7
13:54:02:ObjectDetectionYOLOv5-6.2: virtualEnvDirPath = C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv
13:54:02:ObjectDetectionYOLOv5-6.2: venvPythonCmdPath = C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\Scripts\python.exe
13:54:02:ObjectDetectionYOLOv5-6.2: packagesDirPath = C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\Lib\site-packages
13:54:02:ObjectDetectionYOLOv5-6.2: moduleStartFilePath = detect_adapter.py
13:54:02:ObjectDetectionYOLOv5-6.2: Installing Python 3.7
13:54:02:ObjectDetectionYOLOv5-6.2: Installing python37 in C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37
13:54:02:ObjectDetectionYOLOv5-6.2: Python 3.7 is already installed
13:54:02:ObjectDetectionYOLOv5-6.2: Creating Virtual Environment (Shared)...Virtual Environment already present
13:54:02:ObjectDetectionYOLOv5-6.2: Confirming we have Python 3.7 in our virtual environment...present
13:54:02:ObjectDetectionYOLOv5-6.2: Downloading models-yolo5-pt.zip to C:\Program Files\CodeProject\AI\downloads\ObjectDetectionYOLOv5-6.2\assets
13:54:02:ObjectDetectionYOLOv5-6.2: Downloading Standard YOLO models...Checking 'C:\Program Files\CodeProject\AI\downloads\ObjectDetectionYOLOv5-6.2\models-yolo5-pt.zip'
13:54:02:ObjectDetectionYOLOv5-6.2: already exists...Expanding...Heading to C:\Program Files\CodeProject\AI\downloads\ObjectDetectionYOLOv5-6.2
13:54:02:ObjectDetectionYOLOv5-6.2: Extracting models-yolo5-pt.zip
13:54:02:ObjectDetectionYOLOv5-6.2: Done.
13:54:03:ObjectDetectionYOLOv5-6.2: Copying contents of models-yolo5-pt.zip to assets...done
13:54:03:ObjectDetectionYOLOv5-6.2: Downloading custom-models-yolo5-pt.zip to C:\Program Files\CodeProject\AI\downloads\ObjectDetectionYOLOv5-6.2\custom-models
13:54:03:ObjectDetectionYOLOv5-6.2: Downloading Custom YOLO models...Checking 'C:\Program Files\CodeProject\AI\downloads\ObjectDetectionYOLOv5-6.2\custom-models-yolo5-pt.zip'
13:54:03:ObjectDetectionYOLOv5-6.2: already exists...Expanding...Heading to C:\Program Files\CodeProject\AI\downloads\ObjectDetectionYOLOv5-6.2
13:54:03:ObjectDetectionYOLOv5-6.2: Extracting custom-models-yolo5-pt.zip
13:54:04:ObjectDetectionYOLOv5-6.2: Done.
13:54:04:ObjectDetectionYOLOv5-6.2: Copying contents of custom-models-yolo5-pt.zip to custom-models...done
13:54:04:ObjectDetectionYOLOv5-6.2: Installing Python packages for Object Detection (YOLOv5 6.2)
13:54:04:ObjectDetectionYOLOv5-6.2: [0;Installing GPU-enabled libraries: If available
13:54:05:ObjectDetectionYOLOv5-6.2: Ensuring Python package manager (pip) is installed...Looking in links: c:\Windows\TEMP\tmpnayfpu86
13:54:05:ObjectDetectionYOLOv5-6.2: Requirement already satisfied: setuptools in c:\program files\codeproject\ai\runtimes\bin\windows\python37\venv\lib\site-packages (68.0.0)
13:54:05:ObjectDetectionYOLOv5-6.2: Requirement already satisfied: pip in c:\program files\codeproject\ai\runtimes\bin\windows\python37\venv\lib\site-packages (24.0)
13:54:05:ObjectDetectionYOLOv5-6.2: Done
13:54:07:ObjectDetectionYOLOv5-6.2: Ensuring Python package manager (pip) is up to date...Done
13:54:07:ObjectDetectionYOLOv5-6.2: Requirements Search Path is C:\Program Files\CodeProject\AI\modules\ObjectDetectionYOLOv5-6.2
13:54:07:ObjectDetectionYOLOv5-6.2: Searching for a suitable requirements.txts file in C:\Program Files\CodeProject\AI\modules\ObjectDetectionYOLOv5-6.2
13:54:07:ObjectDetectionYOLOv5-6.2: CUDA version is 12.3 (12_3 / 12)
13:54:07:ObjectDetectionYOLOv5-6.2: Python packages specified by requirements.windows.cuda.txt
13:54:07:ObjectDetectionYOLOv5-6.2: Choosing Python packages from C:\Program Files\CodeProject\AI\modules\ObjectDetectionYOLOv5-6.2\requirements.windows.cuda.txt
13:54:08:ObjectDetectionYOLOv5-6.2: - Installing Pandas, a data analysis / data manipulation tool...Already installed
13:54:09:ObjectDetectionYOLOv5-6.2: - Installing CoreMLTools, for working with .mlmodel format models...Already installed
13:54:09:ObjectDetectionYOLOv5-6.2: - Installing OpenCV, the Open source Computer Vision library...Already installed
13:54:10:ObjectDetectionYOLOv5-6.2: - Installing Pillow, a Python Image Library...Already installed
13:54:11:ObjectDetectionYOLOv5-6.2: - Installing SciPy, a library for mathematics, science, and engineering...Already installed
13:54:11:ObjectDetectionYOLOv5-6.2: - Installing PyYAML, a library for reading configuration files...Already installed
13:54:29:Response timeout. Try increasing the timeout value
14:14:00:Module ObjectDetectionYOLOv5-6.2 installed successfully.
14:14:00:
14:14:00:Module 'Object Detection (YOLOv5 6.2)' 1.9.1 (ID: ObjectDetectionYOLOv5-6.2)
14:14:00:Valid: True
14:14:00:Module Path: <root>\modules\ObjectDetectionYOLOv5-6.2
14:14:00:AutoStart: True
14:14:00:Queue: objectdetection_queue
14:14:00:Runtime: python3.7
14:14:00:Runtime Loc: Shared
14:14:00:FilePath: detect_adapter.py
14:14:00:Pre installed: False
14:14:00:Start pause: 1 sec
14:14:00:Parallelism: 0
14:14:00:LogVerbosity:
14:14:00:Platforms: all,!raspberrypi,!jetson
14:14:00:GPU Libraries: installed if available
14:14:00:GPU Enabled: enabled
14:14:00:ObjectDetectionYOLOv5-6.2: - Installing PyTorch, an open source machine learning framework...
14:14:00:Accelerator:
14:14:00:Half Precis.: enable
14:14:00:Environment Variables
14:14:00:APPDIR = <root>\modules\ObjectDetectionYOLOv5-6.2
14:14:00:CUSTOM_MODELS_DIR = <root>\modules\ObjectDetectionYOLOv5-6.2\custom-models
14:14:00:MODELS_DIR = <root>\modules\ObjectDetectionYOLOv5-6.2\assets
14:14:00:MODEL_SIZE = Medium
14:14:00:USE_CUDA = True
14:14:00:YOLOv5_AUTOINSTALL = false
14:14:00:YOLOv5_VERBOSE = false
14:14:00:
14:14:00:Started Object Detection (YOLOv5 6.2) module
14:14:00:Timed out attempting to install Module 'ObjectDetectionYOLOv5-6.2' (A task was canceled.)
14:14:00:detect_adapter.py: Traceback (most recent call last):
14:14:00:detect_adapter.py: File "C:\Program Files\CodeProject\AI\modules\ObjectDetectionYOLOv5-6.2\detect_adapter.py", line 13, in
14:14:00:detect_adapter.py: from module_runner import ModuleRunner
14:14:00:detect_adapter.py: File "../../SDK/Python\module_runner.py", line 30, in
14:14:00:detect_adapter.py: import aiohttp
14:14:00:detect_adapter.py: ModuleNotFoundError: No module named 'aiohttp'
14:14:00:Module ObjectDetectionYOLOv5-6.2 has shutdown
14:14:00:detect_adapter.py: has exited
14:14:01:Module ObjectDetectionYOLOv5-6.2 started successfully.
modified 18-Mar-24 11:17am.
|
|
|
|
|
What worked for me with python packages not downloading was to change this following setting in /AI/setup.bat from false to true
REM Whether or not to install all python packages in one step (-r requirements.txt)
REM or step by step. Doing this allows the PIP manager to handle incompatibilities
REM better.
REM ** WARNING ** There is a big tradeoff on keeping the users informed and speed/
REM reliability. Generally one-step shouldn't be needed. But it often is. And it
REM often doesn't actually solve problems either. Overall it's safer, but not a
REM panacea
set oneStepPIP=true
and then manually run ../../setup on that module.
YMMV
|
|
|
|
|
|
Hello,
is it possible to reduce the log spam?
Everytime when i open the frontend i need to wait until the whole log is parsed.
Maybe insert the function to change the logging level permanently?
modified 20-Mar-24 11:15am.
|
|
|
|
|
We've added that to our TODO. We should probably limit log retrieval to the last N items
cheers
Chris Maunder
|
|
|
|
|
is there a function to limit the log allready?
|
|
|
|
|
Using latest CPAI (2.5.6) with latest Coral module. Fresh install, every other module except Coral turned off. Blue Iris has face detection turned off.
I'm getting error 500 after almost every trigger from Blue Iris (latest stable).
Not using any custom models or anything. Literally a fresh install of CPAI from Docker on Linux, using Coral. (So default model: MobileNet SSD, default option: Multi-TPU).
modified 13-Mar-24 14:36pm.
|
|
|
|
|
|
Hi,
Tried updating yolov8 to v1.4.0 and got this error:
Unable to download module 'ObjectDetectionYOLOv8' from https:
That URL doesn't seem valid; any ideas? It uninstalled the previous version and I'm now stuck.
Thanks
modified 13-Mar-24 18:30pm.
|
|
|
|
|
I'm getting this error:
Error in Update ObjectDetectionYOLOv8: API server is offline (NetworkError when attempting to fetch resource.)
Also, no indication that an update was happening.
15:06:30:Server version: 2.5.6
15:06:35:Server: This is the latest version
15:06:51:Preparing to install module 'ObjectDetectionYOLOv8'
15:06:51:Downloading module 'ObjectDetectionYOLOv8'
15:06:51:Unable to download module 'ObjectDetectionYOLOv8' from https:
Now it's broken. No option to re-install
EDIT - Wait F5 to the rescue but have to re-install
EDIT 2 Nope, still broken
There also appears to be no way to revert to an older version. After trying to re-install YOLO V8 it just shows "Unknown"
modified 12-Mar-24 18:48pm.
|
|
|
|
|
Thanks very much for the report. We're looking into it.
Thanks,
Sean Ewington
CodeProject
|
|
|
|
|
Please try uninstalling and reinstalling the module, with "Do not use download cache" turned on.
Thanks,
Sean Ewington
CodeProject
modified 13-Mar-24 18:30pm.
|
|
|
|
|
It's working for me now, although I had to reboot to get is to start detecting.
|
|
|
|
|
Where did you find the update to 1.4.0 for YOLOV8 ?
|
|
|
|
|
We actually ended up reverting the module for 2.5.6. You want to be using no later than YOLOv8 1.3.1.
Thanks,
Sean Ewington
CodeProject
|
|
|
|
|
Nvidia 1050Ti (best I can do on my motherboard)
Hi,
In the hope it might fix my problem below, Is there a way to remove cuda 12.4.
I assume it was installed with Nvidia drivers. There is no mention in Windows Add/Remove. JUST 11.7 (& before 11.8), I had installed a variety in the past, removed 11.8 for the moment ..
ALPR GPU CUDA is started but does not detect anything, CPU mode does.
Now a few months after several days of trying to get CUDA enabled for ALPR, I gave up, CPU (only) worked but too slow.
New CPAI 2.5.6 thought have another go.
Only ever had CPU option before.
Object detection worked before & still OK in GPU, Cuda mode
ALPR CPU did work OK but was far too slow for what I wanted.
Back then tried dozens of combinations 11.7, 11.8, 12, CPAI versions, installs using all the procedures as described.
NOW, enabled but does not recognise any plates, No predictions returned using explorer & also in Blue Iris.
If I disable GPU mode & back to running in CPU it does find the plate OK.
08:21:32:System: Windows
08:21:32:Operating System: Windows (Microsoft Windows 10.0.19045)
08:21:32:CPUs: Intel(R) Core(TM) i7-4790 CPU @ 3.60GHz (Intel)
08:21:32: 1 CPU x 4 cores. 8 logical processors (x64)
08:21:32:GPU (Primary): NVIDIA GeForce GTX 1050 Ti (4 GiB) (NVIDIA)
08:21:32: Driver: 551.76, CUDA: 12.4 (up to: 12.4), Compute: 6.1, cuDNN: 8.9
08:21:32:System RAM: 16 GiB
08:21:32:Platform: Windows
08:21:32:BuildConfig: Release
08:21:32:Execution Env: Native
08:21:32:Runtime Env: Production
08:21:32:Runtimes installed:
08:21:32: .NET runtime: 7.0.10
08:21:32: .NET SDK: Not found
08:21:32: Default Python: 3.9.7
08:21:32: Go: Not found
08:21:32: NodeJS: Not found
08:21:32:App DataDir: C:\ProgramData\CodeProject\AI
08:21:32:Video adapter info:
08:21:32: NVIDIA GeForce GTX 1050 Ti:
08:21:32: Driver Version 31.0.15.5176
08:21:32: Video Processor NVIDIA GeForce GTX 1050 Ti
08:21:32: Intel(R) HD Graphics 4600:
08:21:32: Driver Version 20.19.15.5126
08:21:32: Video Processor Intel(R) HD Graphics Family
08:21:32:STARTING CODEPROJECT.AI SERVER
08:21:32:RUNTIMES_PATH = C:\Program Files\CodeProject\AI\runtimes
08:21:32:PREINSTALLED_MODULES_PATH = C:\Program Files\CodeProject\AI\preinstalled-modules
08:21:32:MODULES_PATH = C:\Program Files\CodeProject\AI\modules
08:21:32:PYTHON_PATH = \bin\windows\%PYTHON_NAME%\venv\Scripts\python
08:21:32:Data Dir = C:\ProgramData\CodeProject\AI
08:21:32:Server version: 2.5.6
08:21:36:
08:21:36:Attempting to start ALPR with C:\Program Files\CodeProject\AI\modules\ALPR\bin\windows\python39\venv\Scripts\python "C:\Program Files\CodeProject\AI\modules\ALPR\ALPR_adapter.py"
08:21:36:
08:21:36:Module 'License Plate Reader' 3.0.2 (ID: ALPR)
08:21:36:Valid: True
08:21:36:Module Path: <root>\modules\ALPR
08:21:36:AutoStart: True
08:21:36:Queue: alpr_queue
08:21:36:Runtime: python3.9
08:21:36:Runtime Loc: Local
08:21:36:FilePath: ALPR_adapter.py
08:21:36:Pre installed: False
08:21:36:Start pause: 3 sec
08:21:36:Parallelism: 0
08:21:36:LogVerbosity:
08:21:36:Platforms: all
08:21:36:GPU Libraries: installed if available
08:21:36:GPU Enabled: enabled
08:21:36:Accelerator:
08:21:36:Half Precis.: enable
08:21:36:Environment Variables
08:21:36:AUTO_PLATE_ROTATE = false
08:21:36:MIN_COMPUTE_CAPABILITY = 6
08:21:36:MIN_CUDNN_VERSION = 7
08:21:36:OCR_OPTIMAL_CHARACTER_HEIGHT = 60
08:21:36:OCR_OPTIMAL_CHARACTER_WIDTH = 30
08:21:36:OCR_OPTIMIZATION = True
08:21:36:PLATE_CONFIDENCE = 0.50
08:21:36:PLATE_RESCALE_FACTOR = 2
08:21:36:PLATE_ROTATE_DEG = 0
08:21:36:
08:21:36:Started License Plate Reader module
08:21:38:Current Version is 2.5.6
08:21:38:Server: This is the latest version
08:21:39:
08:21:39:Attempting to start ObjectDetectionYOLOv5-6.2 with C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\Scripts\python "C:\Program Files\CodeProject\AI\modules\ObjectDetectionYOLOv5-6.2\detect_adapter.py"
08:21:39:
08:21:39:Module 'Object Detection (YOLOv5 6.2)' 1.9.1 (ID: ObjectDetectionYOLOv5-6.2)
08:21:39:Valid: True
08:21:39:Module Path: <root>\modules\ObjectDetectionYOLOv5-6.2
08:21:39:AutoStart: True
08:21:39:Queue: objectdetection_queue
08:21:39:Runtime: python3.7
08:21:39:Runtime Loc: Shared
08:21:39:FilePath: detect_adapter.py
08:21:39:Pre installed: False
08:21:39:Start pause: 1 sec
08:21:39:Parallelism: 0
08:21:39:LogVerbosity:
08:21:39:Platforms: all,!raspberrypi,!jetson
08:21:39:GPU Libraries: installed if available
08:21:39:GPU Enabled: enabled
08:21:39:Accelerator:
08:21:39:Half Precis.: enable
08:21:39:Environment Variables
08:21:39:APPDIR = <root>\modules\ObjectDetectionYOLOv5-6.2
08:21:39:CPAI_MODULE_ENABLE_GPU = True
08:21:39:CUSTOM_MODELS_DIR = <root>\modules\ObjectDetectionYOLOv5-6.2\custom-models
08:21:39:MODELS_DIR = <root>\modules\ObjectDetectionYOLOv5-6.2\assets
08:21:39:MODEL_SIZE = Medium
08:21:39:USE_CUDA = True
08:21:39:YOLOv5_AUTOINSTALL = false
08:21:39:YOLOv5_VERBOSE = false
08:21:39:
08:21:39:Started Object Detection (YOLOv5 6.2) module
08:21:50:Sending shutdown request to python/ObjectDetectionYOLOv5-6.2
08:22:01:detect_adapter.py: APPDIR: C:\Program Files\CodeProject\AI\modules\ObjectDetectionYOLOv5-6.2
08:22:01:detect_adapter.py: MODEL_SIZE: medium
08:22:01:detect_adapter.py: MODELS_DIR: C:\Program Files\CodeProject\AI\modules\ObjectDetectionYOLOv5-6.2\assets
08:22:01:Object Detection (YOLOv5 6.2): Retrieved objectdetection_queue command 'Quit'
08:22:16:detect_adapter.py: Object Detection (YOLOv5 6.2) started.
08:22:17:Module ObjectDetectionYOLOv5-6.2 has shutdown
08:22:17:detect_adapter.py: has exited
08:22:23:ObjectDetectionYOLOv5-6.2 went quietly
08:22:23:
08:22:23:Attempting to start ObjectDetectionYOLOv5-6.2 with C:\Program Files\CodeProject\AI\runtimes\bin\windows\python37\venv\Scripts\python "C:\Program Files\CodeProject\AI\modules\ObjectDetectionYOLOv5-6.2\detect_adapter.py"
08:22:23:
08:22:23:Module 'Object Detection (YOLOv5 6.2)' 1.9.1 (ID: ObjectDetectionYOLOv5-6.2)
08:22:23:Valid: True
08:22:23:Module Path: <root>\modules\ObjectDetectionYOLOv5-6.2
08:22:23:AutoStart: True
08:22:23:Queue: objectdetection_queue
08:22:23:Runtime: python3.7
08:22:23:Runtime Loc: Shared
08:22:23:FilePath: detect_adapter.py
08:22:23:Pre installed: False
08:22:23:Start pause: 1 sec
08:22:23:Parallelism: 0
08:22:23:LogVerbosity:
08:22:23:Platforms: all,!raspberrypi,!jetson
08:22:23:GPU Libraries: installed if available
08:22:23:GPU Enabled: enabled
08:22:23:Accelerator:
08:22:23:Half Precis.: enable
08:22:23:Environment Variables
08:22:23:APPDIR = <root>\modules\ObjectDetectionYOLOv5-6.2
08:22:23:CPAI_MODULE_ENABLE_GPU = True
08:22:23:CUSTOM_MODELS_DIR = <root>\modules\ObjectDetectionYOLOv5-6.2\custom-models
08:22:23:MODELS_DIR = <root>\modules\ObjectDetectionYOLOv5-6.2\assets
08:22:23:MODEL_SIZE = Medium
08:22:23:USE_CUDA = True
08:22:23:YOLOv5_AUTOINSTALL = false
08:22:23:YOLOv5_VERBOSE = false
08:22:23:
08:22:23:Started Object Detection (YOLOv5 6.2) module
08:22:39:detect_adapter.py: APPDIR: C:\Program Files\CodeProject\AI\modules\ObjectDetectionYOLOv5-6.2
08:22:39:detect_adapter.py: MODEL_SIZE: medium
08:22:39:detect_adapter.py: MODELS_DIR: C:\Program Files\CodeProject\AI\modules\ObjectDetectionYOLOv5-6.2\assets
modified 13-Mar-24 14:37pm.
|
|
|
|
|
This user had the same issue as you and I think it may be worth going through their thread (though they're on a different system).
CodeProject.AI Server: AI the easy way.[^]
Thanks,
Sean Ewington
CodeProject
|
|
|
|
|
SOLVED. Windows 10
I cannot really advise what fixed this but just to advise of things I tried & there were dozens & dozens of variations, reboots etc etc. I had tried even more a while ago but was on an older CPAI. Now latest 2.5.6, not sure relevant or just finally hit the right pattern. Below might not be 100% as was rather bashing about in desperation. Once is did show Cuda enabled but did not read a plate, later crashed on other fiddling.
I have Licence plate working 1050Ti, Cuda 11.8 CuDNN to match.
11.8 is reported in Nvidia components (this was 12.4 before despite not being mentioned in Add/Remove)
I always had 11.8 with cmd "nvcc --version" despite Nvidia reporting 12.4
I uninstalled Licence Plate
I Add/Removed NVidia graphics driver, then from Device manager as seemed to remain.
I did not reinstall Nvidia. I used the BAT & it loaded 11.8 & this was confirmed in Nvidia/control panel/components.
(I Had tried below with LP installed & did not work, think an error when starting)
Lots of reboots & I think other bits & pieces but it eventually started working & reading the plates correctly!
I am sure had done these steps when I tried some months ago (I tried hundreds of combinations) so not sure what was the magic bullet this time.
modified 13-Mar-24 14:37pm.
|
|
|
|
|