Terracotta is an easy-to-use platform that allows users to experiment with large language models (LLMs) quickly and efficiently. With Terracotta, users can manage multiple fine-tuned models in one place, enabling them to iterate and improve their models quickly.
The tool offers both qualitative and quantitative evaluation features, allowing users to compare model outputs and evaluate performance using various metrics such as accuracy, BLEU, and confusion matrices. One of Terracotta's key functionalities is the ability to securely upload data, which is crucial for the fine-tuning process.
Users can store their data on the platform and then use it to fine-tune their models for both classification and text generation tasks. The fine-tuning process is streamlined, requiring just a few clicks to get started. Terracotta also seamlessly integrates with popular LLM providers like OpenAI and Cohere, giving users the flexibility to connect and work with different providers within the platform.
Terracotta was created by Beri Kohen and Lucas Pauker, both Stanford AI graduates with a passion for LLMs. As a tool designed by experts in the field, Terracotta promises to streamline the LLM development workflow and make it more accessible for users. Sign up for their email list to stay updated on their latest developments.
