Hi All, here is a simple tool for testing your prompt and response quality. ️️🧪
I trained it across different quality matrices like hallucination, bias-ness, child-safety etc.
The result: Evaluation in tabular form with quantified matrices and providing reasoning behind the score.
https://chat.openai.com/g/g-WufHT9Sgj-orangepro
Basic instructions:
1. Type: “prompt: <……>”
2. Type: “response: <…..>”
3. See the evaluation in tabular form with quantified matrices and some reasoning behind it.
The context window is large, so a longer prompt or response is ok as well. You can also try just a prompt to see the prompt quality.
Let me know any feedback. This is a small feature of our testing platform leveraging recent GPTs fine tuning abilities.