Ultralytics
yolov5/tutorials/pytorch_hub_model_loading/ #8255
giscus
bot
Replies: 8 comments 14 replies
-
|
python run.py Traceback (most recent call last): During handling of the above exception, another exception occurred: Traceback (most recent call last): The above exception was the direct cause of the following exception: Traceback (most recent call last): kakul@Teddy MINGW64 ~/OneDrive/Desktop/Collected_Image can someone help me to rectify this error |
Beta Was this translation helpful? Give feedback.
-
|
@LikhithaKakuluri it looks like you're encountering a To resolve this, you can try setting the model = torch.hub.load('ultralytics/yolov5', 'custom', path='C:/Users/kakul/OneDrive/Desktop/Collected_Image/best.pt', force_reload=True)If the issue persists, ensure that your file paths are correctly formatted for Windows (using backslashes For more detailed guidance, you might find the PyTorch Hub tutorial helpful. If you need further assistance, feel free to ask! 😊 |
Beta Was this translation helpful? Give feedback.
-
|
Hi @LikhithaKakuluri, I see you are using Windows. Why don't you follow this tutorial? There you have a nice step by step tutorial on how to prepare your environment and run yolov5 on Windows. |
Beta Was this translation helpful? Give feedback.
-
|
Nothing seems to work I have already done everything I could and I keep getting the same error over and over again can someone help me with this. Traceback (most recent call last): During handling of the above exception, another exception occurred: Traceback (most recent call last): The above exception was the direct cause of the following exception: Traceback (most recent call last): |
Beta Was this translation helpful? Give feedback.
-
|
@OFFT3ST hey there! 👋 It looks like you're running into a Instead of using raw string paths like from pathlib import Path
import torch
# Define model path using Path object
model_path = Path(r'C:\Users\anony\Desktop\cheats\best.pt')
# Load the model
model = torch.hub.load(str(Path(r'C:\Users\anony\Desktop\cheats\yolov5')), 'custom', path=model_path, source='local')Converting the paths to strings when passing them to Give this a try and let us know how it goes! If you run into any more issues, we're here to help. Happy coding! 🚀 |
Beta Was this translation helpful? Give feedback.
-
|
YOLOv5 v7.0-294-gdb125a20 Python-3.12.2 torch-2.2.1+cu121 CUDA:0 (NVIDIA
GeForce RTX 3070, 8192MiB)
Traceback (most recent call last):
File "C:\Users\anony\Desktop\cheats\yolov5\hubconf.py", line 50, in
_create
model = DetectMultiBackend(path, device=device, fuse=autoshape) #
detection model
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\anony\Desktop\cheats\yolov5\models\common.py", line 467,
in __init__
model = attempt_load(weights if isinstance(weights, list) else w,
device=device, inplace=True, fuse=fuse)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\anony\Desktop\cheats\yolov5\models\experimental.py", line
98, in attempt_load
ckpt = torch.load(attempt_download(w), map_location="cpu") # load
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File
"C:\Users\anony\Desktop\cheats\cheats-env\Lib\site-packages\torch\serialization.py",
line 1026, in load
return _load(opened_zipfile,
^^^^^^^^^^^^^^^^^^^^^
File
"C:\Users\anony\Desktop\cheats\cheats-env\Lib\site-packages\torch\serialization.py",
line 1438, in _load
result = unpickler.load()
^^^^^^^^^^^^^^^^
File
"C:\Users\anony\AppData\Local\Programs\Python\Python312\Lib\pathlib.py",
line 1422, in __new__
raise NotImplementedError(
NotImplementedError: cannot instantiate 'PosixPath' on your system
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "C:\Users\anony\Desktop\cheats\yolov5\hubconf.py", line 65, in
_create
model = attempt_load(path, device=device, fuse=False) # arbitrary model
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\anony\Desktop\cheats\yolov5\models\experimental.py", line
98, in attempt_load
ckpt = torch.load(attempt_download(w), map_location="cpu") # load
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File
"C:\Users\anony\Desktop\cheats\cheats-env\Lib\site-packages\torch\serialization.py",
line 1026, in load
return _load(opened_zipfile,
^^^^^^^^^^^^^^^^^^^^^
File
"C:\Users\anony\Desktop\cheats\cheats-env\Lib\site-packages\torch\serialization.py",
line 1438, in _load
result = unpickler.load()
^^^^^^^^^^^^^^^^
File
"C:\Users\anony\AppData\Local\Programs\Python\Python312\Lib\pathlib.py",
line 1422, in __new__
raise NotImplementedError(
NotImplementedError: cannot instantiate 'PosixPath' on your system
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "c:\Users\anony\Desktop\cheats\scripts\detections.py", line 16, in
<module>
model =
torch.hub.load(str(Path(r'C:\Users\anony\Desktop\cheats\yolov5')),
'custom', path=model_path, source='local')
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File
"C:\Users\anony\Desktop\cheats\cheats-env\Lib\site-packages\torch\hub.py",
line 566, in load
model = _load_local(repo_or_dir, model, *args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File
"C:\Users\anony\Desktop\cheats\cheats-env\Lib\site-packages\torch\hub.py",
line 595, in _load_local
model = entry(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\anony\Desktop\cheats\yolov5\hubconf.py", line 88, in custom
return _create(path, autoshape=autoshape, verbose=_verbose,
device=device)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\anony\Desktop\cheats\yolov5\hubconf.py", line 83, in
_create
raise Exception(s) from e
Exception: cannot instantiate 'PosixPath' on your system. Cache may be out
of date, try `force_reload=True` or see
https://docs.ultralytics.com/yolov5/tutorials/pytorch_hub_model_loading for
help.
I have done as you said and I still ran into the same error here is my code
from pathlib import Path
from mss import mss
import torch
import cv2
import numpy as np
MONITOR_WIDTH = 1920#game res
MONITOR_HEIGHT = 1080#game res
MONITOR_SCALE = 5#how much the screen shot is downsized by eg. 5 would be
one fifth of the monitor dimensions
region = (int(MONITOR_WIDTH/2-MONITOR_WIDTH/MONITOR_SCALE/2),int(
MONITOR_HEIGHT/2-MONITOR_HEIGHT/MONITOR_SCALE/2),int(MONITOR_WIDTH/2+
MONITOR_WIDTH/MONITOR_SCALE/2),int(MONITOR_HEIGHT/2+MONITOR_HEIGHT/
MONITOR_SCALE/2))
model_path = Path(r'C:\Users\anony\Desktop\cheats\best.pt')
model = torch.hub.load(str(Path(r'C:\Users\anony\Desktop\cheats\yolov5')),
'custom', path=model_path, source='local')
model.conf = 0.40
model.maxdet = 10
model.apm = True
with mss() as stc:
while True:
screenshot = np.array(stc.grab(region))
df = model(screenshot, size=736).pandas().xyxy[0]
for i in range(0,10):
try:
xmin = int(df.iloc[i,0])
ymin = int(df.iloc[i,1])
xmax = int(df.iloc[i,2])
ymax = int(df.iloc[i,3])
cv2.rectangle(screenshot,(xmin,ymin),(xmax,ymax), (255,0,0),
3)
except:
print("",end="")
cv2.imshow("frame",screenshot)
if(cv2.waitKey(1) == ord('l')):
cv2.destroyAllWindows()
break
…On Sat, Mar 16, 2024 at 4:33 AM Glenn Jocher ***@***.***> wrote:
@OFFT3ST <https://github.com/OFFT3ST> hey there! 👋 It looks like you're
running into a NotImplementedError related to PosixPath on Windows. This
usually happens because of the path handling difference between Unix-based
systems and Windows. Here's a quick fix you could try! 🛠️
Instead of using raw string paths like r'C:\path\to\your\model', you
might want to use Python's Path object from the pathlib library, which is
designed to handle system-dependent path formats gracefully. Here's how you
could update the script:
from pathlib import Pathimport torch
# Define model path using Path objectmodel_path = Path(r'C:\Users\anony\Desktop\cheats\best.pt')
# Load the modelmodel = torch.hub.load(str(Path(r'C:\Users\anony\Desktop\cheats\yolov5')), 'custom', path=model_path, source='local')
Converting the paths to strings when passing them to torch.hub.load might
help avoid the NotImplementedError related to PosixPath. The Path object
from pathlib should automatically handle the conversion to the correct
path string format for your operating system.
Give this a try and let us know how it goes! If you run into any more
issues, we're here to help. Happy coding! 🚀
—
Reply to this email directly, view it on GitHub
<#8255 (reply in thread)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/BFZPNKBMVY2WALBSERYCGULYYP7V3AVCNFSM6AAAAABDMQT4FWVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4DQMJQGIZDM>
.
You are receiving this because you were mentioned.Message ID:
***@***.***
com>
|
Beta Was this translation helpful? Give feedback.
-
|
from pathlib import Path MONITOR_WIDTH = 1920#game res MONITOR_SCALE = 5#how much the screen shot is downsized by eg. 5 would be one fifth of the monitor dimensions model_path = Path(r'C:\Users\anony\Desktop\cheats\best.pt') model = torch.hub.load(str(Path(r'C:\Users\anony\Desktop\cheats\yolov5')), 'custom', path=model_path, source='local') with mss() as stc: |
Beta Was this translation helpful? Give feedback.
-
|
Hey there! 😊 It looks like you're encountering a A quick workaround is to ensure you're passing strings (instead of Path objects) directly to Instead of using: model = torch.hub.load(str(Path(r'C:\Users\anony\Desktop\cheats\yolov5')), 'custom', path=model_path, source='local')You can directly pass the path as a string: model = torch.hub.load(r'C:\Users\anony\Desktop\cheats\yolov5', 'custom', path=str(model_path), source='local')Also, if you haven't already, consider following the PyTorch Hub guide here for more details on loading models. Make sure your environment is set up correctly with all necessary dependencies installed as outlined in the guide. If the issue persists, setting Hope this helps you get back on track! Let us know if there's anything else we can do to assist. Happy coding! |
Beta Was this translation helpful? Give feedback.
-
|
I encountered the error cannot instantiate 'WindowsPath' on your system, and none of the suggested solutions helped. For context, I was trying to run a custom YOLOv5 model on CVAT or simply run the model on my WSL, but each time I attempted it, I got the same error. I tried various solutions, but none worked. What I eventually realized was that I had trained my model on a Windows system using Windows CMD. So, when I tried to run the model on Linux, Ubuntu, or WSL on Windows, the error persisted. What I ended up doing was retraining the model, this time using an Ubuntu system. I used Google Colab for training, then tested the model on Ubuntu and CVAT, and it worked fine. I hope this helps anyone else encountering a similar problem, or maybe it could fix the 'PosixPath' error. |
Beta Was this translation helpful? Give feedback.
-
|
Hi there! 😊 Thanks for sharing your experience and solution for the For those encountering the For anyone reading this looking for a quick fix without retraining, consider converting WindowsPath to a string before using it in your scripts when running on Linux/WSL. Here’s a quick Python example on how to do it: from pathlib import Path
# Assuming 'windows_path' is your WindowsPath object
windows_path = Path("C:/your/model/path")
# Convert to string for universal compatibility
path_str = str(windows_path)
# Now 'path_str' can be used in your script without causing errorsThis approach ensures that your paths are interpreted correctly regardless of the underlying OS. Again, kudos to you for figuring out a workaround and sharing it with the community! If anyone has further questions or runs into similar issues, feel free to ask. Happy coding! 🚀 |
Beta Was this translation helpful? Give feedback.
-
|
Hi. I have trained a custom model of YOLOv5s and have the 'best.pt' file saved on my local machine. I am trying to load the model using torch.hub to use it for inference. I am encountering the following error. Using cache found in /home/kvothe/.cache/torch/hub/ultralytics_yolov5_master Traceback (most recent call last): During handling of the above exception, another exception occurred: Traceback (most recent call last): The above exception was the direct cause of the following exception: Traceback (most recent call last): Can anyone please help rectify this error. |
Beta Was this translation helpful? Give feedback.
-
|
Hey there! It looks like you're encountering errors while trying to load a custom-trained YOLOv5 model using The primary issue seems to be the path specification for loading the model. When loading a custom model with Here's how you can correct the code for loading your trained model: import torch
# Correct way to load a custom model
model = torch.hub.load('ultralytics/yolov5', 'custom', path='path/to/your/best.pt', source='local')Please make sure that the path to the If you've attempted this and are still facing issues, it might be related to version compatibility or cache problems. Considering forcing a cache reload with I hope this helps! If you have any more questions or run into other issues, feel free to ask. Happy coding! 😊 |
Beta Was this translation helpful? Give feedback.
-
|
Thank you Glenn! How do I check whether PyTorch and YOLOv5 versions are compatible? I was able to get the version of PyTorch. How do I check the version of YOLOv5? Thank you. |
Beta Was this translation helpful? Give feedback.
-
|
@edgeDancer256 hey! Thanks for updating me. It looks like you've already tried quite a few troubleshooting steps. To check the version of YOLOv5 that you are using, you can look for the version information in your YOLOv5 download, often mentioned in the initial comments or documentation strings of the code files. Alternatively, if you cloned the repository, you can check the git branch or tags to pinpoint the version. Regarding the compatibility check between PyTorch and YOLOv5, a good rule of thumb is to match the YOLOv5's requirements. The YOLOv5 GitHub page or the requirements.txt file you initially used should list the compatible PyTorch version. If you’ve installed PyTorch in a Python environment, ensure this matches the requirements specified for the version of YOLOv5 you have. Feel free to keep me posted if you encounter any more issues or have further questions. Happy coding! 🌟 |
Beta Was this translation helpful? Give feedback.
-
|
Thank you @glenn-jocher for your help. I was unable to get the model to load from the .pt files. I was instead able to do it by exporting the model to onnx (-dynamic) format. |
Beta Was this translation helpful? Give feedback.
-
|
@edgeDancer256 hey there! 👋 Great to hear that you managed to export your model to ONNX successfully. That's a solid workaround! If you need to use the |
Beta Was this translation helpful? Give feedback.
-
|
Hi, I am running a food detection project using pretrained yolov8s and yolov5. Yolov8s doesn't have issues, where it detected food accurately. However, yolov5 doesn't seem to work at all, I got the following error message. I loaded yolov5 using Pytorch. [2024-05-16 16:44:50,769] ERROR in app: Exception on /analyze [POST] It keeps showing this error, I guess it has something to do with theseus? |
Beta Was this translation helpful? Give feedback.
-
|
Hello! It looks like the
Try these adjustments, and it should help resolve the import error! If the issue persists, please share more details about your project’s structure. Happy coding! 😊 |
Beta Was this translation helpful? Give feedback.
-
|
Hi, for the latter solution, does it mean to include under if name == 'main':? I have tried it and it gives me the same error. |
Beta Was this translation helpful? Give feedback.
-
|
@TanWeeKang hello! It looks like you're trying to import modules from a specific directory structure. If you're encountering issues even after using Given your directory path
If you've tried these steps and are still facing issues, there might be something more specific at play. Feel free to share more details! 😊 |
Beta Was this translation helpful? Give feedback.
-
|
hi i keep getting this error again and again, and I tried the solutions here. this is my loading line: model = torch.hub.load("ultralytics/yolov5", 'custom', |
Beta Was this translation helpful? Give feedback.
-
|
Hello! It seems like there might be a problem with the file path specification or system compatibility on Windows. Here's a quick suggestion:
Try modifying your loading line to: model = torch.hub.load("ultralytics/yolov5", 'custom', path="models\\best.pt", force_reload=True)or model = torch.hub.load("ultralytics/yolov5", 'custom', path="models/best.pt", force_reload=True)
If you continue to face issues, it might be helpful to clear the cache manually and try running it again. You can clear the cache by deleting the contents of the |
Beta Was this translation helpful? Give feedback.
-
yolov5/tutorials/pytorch_hub_model_loading/
Detailed guide on loading YOLOv5 from PyTorch Hub. Includes examples & tips on inference settings, multi-GPU inference, training and more.
https://docs.ultralytics.com/yolov5/tutorials/pytorch_hub_model_loading/
Beta Was this translation helpful? Give feedback.
All reactions