Finally, the data set is marked, and the training starts, and then the following error repeatedly occurs:UnicodeDecodeError: 'gbk' codec can't decode byte 0x8b in position 233: illegal multibyte sequenceHowever, the strange thing is that the position of each error is not the same! After all, there was a problem reading the training set and test set images.The path of the configuration file was changed AGAIN and AGAIN, the default parameters were changed AGAIN and AGAIN, the whole yoloV5 model was installed on GitHub AGAIN and consulted by many seniors, the final result was — AGAIN!!Since the problem is the import of the dataset, take a look at my.yaml fileIn the final modification, when I removed the blank lines of line 4, line 7, and line 10, I was surprised to find that it was running! So, as a final note, this may not be a one-size-fits-all solution: If you get the same error, check to see if your yamL file is correct. If possible, delete the space between lines, especially the last line.