Code to export results as txt for wider-face?

[ogvl]

I'm currently working on training yolox with the wider-face dataset. I would like to export the results as outlined in the competition here. Each text file should contain one row per detected bounding box in the format [left, top, width, height, score]. Basically, the export for each img file should look like this:

< image name i >
< number of faces detected in this image = im >
< face i1 >
< face i2 > 

Example:

0_Parade_marchingband_1_21 3
541 354 36 46 1.000 
31 254 32 45 0.980 
241 167 36 46 0.900

I know that I have to change the demo.py file to achieve this but I'm not quite sure which variables stand for the left, top, width, height, score information of each bounding box.