Fine Tune with other languages

[Hilal-Urun]

Hi, I want to ask that when we are fine tune gpt2 english model with other languages such as Turkish, because of having historical data before fine tune, will there a conflict like when we are chatting the bot can it be use english words? If so, How can we fix it?

[Plkmoi]

Does the code work nowadays?

[Hilal-Urun]

Hi @Plkmoi Actually no, In part 5, when visualizing dataset I was getting stuck because for example I run the code in the morning and it did not gave me any results.
This part ;

lang = 'tr'
fname = f'all_texts_{lang}wiki.csv'
df = pd.read_csv(path_data/fname)
print(len(df))