Reproducing ULMFiT classifier (hate speech detection)

I tried to open this repo for the hate speech detection task.
I followed the instructions in the Readme file and finally managed to customize the appropriate version.

I don’t quite understand everything here. I downloaded the data from poleval2019, where: the task6-1 folder contains test_set_clean_only_tags.txt and test_set_clean_only_text.txt while the test folder - task6-test contains folders task1 and task2, which contain files evaluate1.pl, evaluate2.pl and test_set_clean_only_tag_text .
Meanwhile…
In the evaluate.py file we have test_file = 'test1' and I don’t know what this test1 is

and

./test.sh test.txt - I also don’t know what the test.txt file is

Also, in the test.sh file there is cat poleval19-t6-s1.txt (?)

1 Like