stanford_alpaca
stanford_alpaca copied to clipboard
calculate max_tokens based on prompt tokens
The idea is to maximize the new tokens, and a first step to adjust to users that change the prompt.txt. #47 I dont know whether the code is correct, please have a look.