Dear authors,
Thanks for your open-sourced repo!
I have a question for gen_length when using TraDo-8B-Thinking. If gen_length is 2000, does it mean that the total number of generated tokens for thinking and final output after thinking (two kinds of tokens) is 2000?
Dear authors,
Thanks for your open-sourced repo!
I have a question for gen_length when using TraDo-8B-Thinking. If gen_length is 2000, does it mean that the total number of generated tokens for thinking and final output after thinking (two kinds of tokens) is 2000?