Braintrust Weekly Update
We’ve been heads down at Braintrust working on an exciting new feature. We’ll announce it soon! In the meantime, we’ve shipped some user experience improvements and fixes this week:
We wanted to try out open source models like Mistral, Codellama, Llama2, etc. and compare them to OpenAI's new models. So, we added them to the playground and also used Perplexity's API for the OS models. They are so fast!
- Improved selectors for diffing and comparison modes on experiment view
- Added support for new OpenAI models (GPT4 preview, 3.5turbo-1106) in playground
- Added support for open source models (Mistral, Codellama, Llama2,etc.) in the playground using Perplexity's APIs (opens in a new tab)
- OpenAI's reproducible outputs makes evals and testing 10x better (opens in a new tab)
- Have fun building AI apps with your team (opens in a new tab)
Braintrust is the enterprise-grade stack for building AI products. From evaluations, to prompt playground, to data management, we take uncertainty and tedium out of incorporating AI into your business.