Oops something went wrong, there might be issues with this model (YOLOV11)

Community Help
1

Hi, I managed to upload my weights from Google Colab to Roboflow. However, it seems that it is unable to visualize the model

issue
issue1440×700 152 KB

I managed to upload the weights, but it seems that there are problems with the inference in Google Colab too

Any ways to solve this? This is not the first time where I have encountered this problem. Initially the deployment of model was also not reflected on Roboflow, it took more than 15 minutes and multiple times of re-running the “deploy” line on Colab for it to currently display the metrics, but fails to visualize it

Project Type: Instance Segmentation
Operating System & Browser: Microsoft Edge, Windows 11
Project Universe Link or Workspace/Project ID: testna

2

Issue of Inference in Google Colab

image
image1266×714 42.4 KB

3

I tried rerunning the code in Google Colab and this was the error when installing inference, it says that my package rich version is not incompatible even though i specified it to be the exact same. I was initially prompted to restart the session which I did, but now the issue still persists

Hence, I am unable to do inference in the subsequent code with the error below

RoboflowAPIUnsuccessfulRequestError: Unsuccessful request to Roboflow API with response code: 500

Error in installing inference:

image
image1441×412 38.7 KB

4

Hi, I am also experiencing the same error in my end. Have you found any solutions yet? Here is my training notebook for reference.

5

Hey I have not, waiting for a reply from a Roboflow representative on this matter. I am still unable to visualize my model in Roboflow even though the weights have been successfully uploaded.

6

Hello can anybody help with this error? I have retrained another model and there seems to be problems with Roboflow Inference

7

Hello @Looi_Kai_Xuan,

Thank you for reporting this issue.

Am I correct to assume you uploaded your weights file (*.pt) into Roboflow through roboflow cli?

8

Dear @Grzegorz, my bad I have fixed the issue. I realized I was uploading the instance segmentation model weights to an Object Detection Project. I have changed the project type and the model visualizer in Roboflow works now :slight_smile:

However, I do still face issues with Inference API at the end of the Google Colab Instance Segmentation Training on YOLOv11. May I know if its an API error or if there is something wrong in my code?

image
image874×813 24.3 KB

image
image1401×616 76 KB

9

Thank you for confirming the original issue is solved!

With regards to API error, I noticed your model is public, so I tried to load it myself:

import cv2 as cv

from inference import get_model

model_id = "testing-awz3x/2"
model = get_model(model_id=model_id, api_key=<secret>)
img_path = "/path/to/my/file.jpeg"
results = model.infer(
    image=img_path,
    confidence=0.1,
)

I received no error. Can you try to run your model in simplified environment like this and confirm if error is still coming up?

Thanks,
Grzegorz

10

Hey @Grzegorz, I tried running it and seems like there are no issues about it. But, I do get the API error from time to time previously when I was training on Google Colab.

Thank you very much for your support! Cheers

Regards,
Kai Xuan

1 Like