Problem in training YOLOv3

[Midhilesh29]

Training problem
I was training yolov3 with batchsize=2. I got the following error:

RuntimeError                              Traceback (most recent call last)
/usr/local/lib/python3.6/dist-packages/torch/utils/data/dataloader.py in _try_get_data(self, timeout)
    760         try:
--> 761             data = self._data_queue.get(timeout=timeout)
    762             return (True, data)

11 frames
/usr/lib/python3.6/queue.py in get(self, block, timeout)
    172                         raise Empty
--> 173                     self.not_empty.wait(remaining)
    174             item = self._get()

/usr/lib/python3.6/threading.py in wait(self, timeout)
    298                 if timeout > 0:
--> 299                     gotit = waiter.acquire(True, timeout)
    300                 else:

/usr/local/lib/python3.6/dist-packages/torch/utils/data/_utils/signal_handling.py in handler(signum, frame)
     65         # Python can still get and update the process status successfully.
---> 66         _error_if_any_worker_fails()
     67         if previous_handler is not None:

RuntimeError: DataLoader worker (pid 2546) is killed by signal: Bus error. It is possible that dataloader's workers are out of shared memory. Please try to raise your shared memory limit.

During handling of the above exception, another exception occurred:

RuntimeError                              Traceback (most recent call last)
<ipython-input-37-5dcbfe173d75> in <module>()
----> 1 gtf.Train(num_epochs=1);

/content/Monk_Object_Detection/7_yolov3/lib/train_detector.py in Train(self, num_epochs)
    386                     # Train mutation

Inference problem
I tired to train yolov3-tiny instead for 1 epoch. With this trained yolov3-tiny, I tried to infer the model

img_path = "/content/drive/My Drive/Monkai/PRImA Layout Analysis Dataset/images/Train/00001317.tif";
gtf.Predict(img_path, conf_thres=0.3, iou_thres=0.5);

I got the following error

AssertionError                            Traceback (most recent call last)
<ipython-input-20-c510a274f973> in <module>()
      1 img_path = "/content/drive/My Drive/Monkai/PRImA Layout Analysis Dataset/images/Train/00001317.tif";
----> 2 gtf.Predict(img_path, conf_thres=0.3, iou_thres=0.5);
      3 #Image(filename='/content/drive/My Drive/Career/output.jpg')

1 frames
/content/Monk_Object_Detection/7_yolov3/lib/infer_detector.py in Predict(self, img_path, conf_thres, iou_thres)
    100         self.system_dict["local"]["dataset"] = LoadImages(source, 
    101                                                             img_size=self.system_dict["params"]["img_size"],
--> 102                                                             half=self.system_dict["params"]["half"]);
    103 
    104         # Get names and colors

/content/Monk_Object_Detection/7_yolov3/lib/utils/datasets.py in __init__(self, path, img_size, half)
     65         else:
     66             self.cap = None
---> 67         assert self.nF > 0, 'No images or videos found in ' + path
     68 
     69     def __iter__(self):

AssertionError: No images or videos found in tmp

Please help me with these problems.

[abhi-kumar]

Please check the path to dataset if it's proper or not

[abhi-kumar]

Let us know your code and directory structure if the error persists

[Shreyashwaghe]

I am also getting the ' No images or videos found in tmp' error,

[abhi-kumar]

Can you please share the code and folder directory details?

[Midhilesh29]

Hi
The link of my code: https://colab.research.google.com/drive/14A3siVGbRn3x-AbSwAJhThEu9iURlVFK?authuser=3#scrollTo=jHxFzDF5pABG
In this notebook, I have created a markdow as "error" the cell under that markdown has the error.
Also, I have solved this problem by copying the images to monkai folder the following command tells that

!cp "/content/drive/My Drive/Monkai/input.tif" "Monk_Object_Detection/example_notebooks/sample_dataset/ship/test/"

[abhi-kumar]

Ok. So it was more like a file not found error

[Midhilesh29]

yes, inside the yolov3 implementation there is a line

os.system("cp"  "<the path of image or video">  /tmp)

This particular line has some problem in execution. I also checked the /tmp/ folder !ls /tmp. It returned empty list. So what I think was the copy command which is being executed in os.system is working so files weren't copying to tmp folder and error has been thrown.

[THEFASHIONGEEK]

Try to change the .cfg file as mentioned in this link

[abhi-kumar]

We have made the necessary changes, please check if the error is now resolved for you?

[abhi-kumar]

Closing due to inactivity.

[gauravgund]

@abhi-kumar : I am getting an error while using pretrained yolov3 for Document Layout Analysis i.e.

AssertionError: No images or videos found in tmp

I am following this tutorial: https://pub.towardsai.net/object-detection-document-layout-analysis-using-monk-object-detection-toolkit-6c57200bde5

Also, are there any tutorials to fine tune the yolov3 models, that will be of great help! I have worked with darknet and ultralytics for yolov5. Din't see any information about fine-tuning on the Github page.

Thanks in advance!