DeepEval

DeepEval provides a Pythonic way to run offline evaluations on your LLM pipelines so you can launch comfortably into production.

Why we wrote this library​

While the growth of LLMs, LangChain, LlamaIndex became prominent- we found that once these pipelines were built, it became really hard to continue iterating on these pipelines. Many engineers wanted to use LangChain as a quick start and then start adding guardrails, switch LLMs to Llama2.

Join our Discord​

We are continuing to evolve our evaluation platform and welcome discussion on our discord: https://discord.gg/a3K9c8GRGt