Show HN: Terracotta – Platform for fine-tuning and evaluating LLMs

Terracotta is a free tool designed for fine-tuning and evaluating LLMs. We understand the challenges in experimenting with LLMs and have created a solution that streamlines the entire process, saving you valuable time and effort. Here are a few features that we have: 1. Training: Users can upload a dataset and fine-tune different LLMs provided by OpenAI within minutes using our training dashboard. 2. Qualitative evaluation: A playground for qualitatively comparing base models vs. fine-tuned models from OpenAI and Cohere. 3. Quantitative evaluation: An evaluation tool that lets you run inference on a dataset with any model with a few clicks, and compare several models across different metrics of your choice. We calculate relevant metrics depending on what kind of task you want to do.