Dataset Card for "prepared-yagpt"
Short Description
This dataset is aimed for training of chatbots on russian language.
It consists plenty of dialogues that allows you to train you model answer user prompts.
Notes
Special tokens
history, speaker1, speaker2 (history can be optionally removed, i.e. substituted on empty string)
