Tomasz Hawro

It is also possible to send `tqdm` progress bar to log file via its `__str__` representation. My solution is to create separate logger for cmd logging and separate for file...

Here is the code I wrote to parse `png` mask to YOLO format: ``` import numpy as np import cv2 def find_polygons_in_binary_mask(binary_mask: np.ndarray) -> list[np.ndarray]: contours, _ = cv2.findContours(binary_mask, cv2.RETR_TREE,...

I had the same problem when saving the checkpoint with `ModelCheckpoint` initialized with `every_n_train_steps=1`. The problem disappeared when I used `every_n_epochs=1` instead of the `every_n_train_steps`. pytorch-lightning: 1.8.6