search

prompt-security / ps-fuzz

Make your GenAI Apps Safe & Secure :rocket: Test & harden your system prompt
https://www.prompt.security/fuzzer
MIT License
404 stars 54 forks source link

Interactive mode: modifying target_model doesn't work #40

Closed vitaly-ps closed 7 months ago

vitaly-ps commented 7 months ago

In interactive mode, an attempt to modify the Target model name fails (no matter what you put there it reverts back to the default gpt-3.5-turbo).