Dioptra diagnoses your model to identify failure modes like bias and drift; and uses active learning techniques to curate the best data to fix it.

LLM Sandbox is an all-in-one environment to detect hallucinations, evaluate your prompts, select the right data to fine tune your models and track performance version after version.

Hey PH, This is Farah! With @parnoux and @jacques_arnoux, we just launched LLM Sandbox as part of our open source project: katiML. With LLM Sandbox, teams can detect hallucinations, improve their prompts and chain of models. They also find the best data for fine-tuning faster. Would you use this? Why / why not? Thanks for the feedback! Links: Free trial: https://dev.dioptra.ai/register Our github https://github.com/dioptra-ai/ka... Documentation https://dioptra.gitbook.io/diopt... Slack community channel https://join.slack.com/t/dioptra...

Congrats on the launch of LLM Sandbox, it sounds like a powerful tool for improving prompts and model chains. Looking forward to giving it a spin!

Congratulation on the launch LLM Sandbox!

Dioptra

The metrics engine for better data and better models

LLM Sandbox by Dioptra

Dioptra Launches

Do you use Dioptra?

You might also like

Dioptra Launches

You might also like