Build
Run your evaluations
Build
Run your evaluations
Simple and flexible way to run evaluations
Selene offers a simple and flexible option to run your evaluations using a python SDK.
You can run an evaluation using Atla default metrics or a custom metric you created using the Eval Copilot.
Selene not only outputs a score, but also a helpful Chain-of-Thought critique. You can use these to generate valuable insights across your LLM lifecycle.
Next steps:
Read about how you can use Selene’s scores and critiques to improve your LLM outputs here.