Start Simple: Your First h2oGPTe Chat
- Generative AI
- Enterprise h2oGPTe
This lesson walks through the five key steps for using h2oGPTe: Prompting, Collection, Configuration, RAG, and API, starting with the simplest one, regular chatting. It shows two example prompts typed into the chat field, first a basic identity question asking the AI who it is, then a more specific question about the factors behind drug desistance. It notes that even with a niche topic, the large language model draws on its training data to return a coherent, well reasoned answer without any uploaded documents. It sets up the following lessons, which build toward collections, configuration, and retrieval augmented generation.
Key takeaways
- What the five key steps for using h2oGPTe are: Prompting, Collection, Configuration, RAG, and API
- How a simple identity question shows h2oGPTe's basic prompting before any documents are involved
- Why a large language model can answer a niche question, like drug desistance factors, from training data alone
- The lesson sets up later steps that build toward collections, configuration, and RAG
Start owning your AI