FAQ from Terracotta
What is Terracotta?
Terracotta is a platform that allows users to experiment with Large Language Models (LLMs) quickly and intuitively. It enables users to manage, fine-tune, and evaluate multiple LLM models in one place. Users can securely store their data, fine-tune models for classification and text generation, and compare models qualitatively and quantitatively.
How to use Terracotta?
To use Terracotta, follow these steps:n1. Sign in with your Google or GitHub account.n2. Upload your data to securely store it for model fine-tuning.n3. Fine-tune models on your data for classification and text generation.n4. Create evaluations by comparing model outputs using prompts for qualitative evaluation or evaluating on datasets with various metrics for quantitative evaluation.
How do I use Terracotta?
To use Terracotta, sign in with your Google or GitHub account, upload your data for model fine-tuning, fine-tune models for classification and text generation, and create evaluations by comparing model outputs.
What are the core features of Terracotta?
Terracotta offers model management to handle multiple fine-tuned models, rapid iteration through qualitative and quantitative evaluation, and integration with multiple providers such as OpenAI and Cohere.
What are the use cases for Terracotta?
Terracotta can be used for purposes like optimizing natural language processing tasks, rapid iteration in research and development of models, and developing AI applications utilizing Large Language Models.