Jurassic-1 Instruct, our instruction-following language model, is now available in open beta! 🦖
Based on the Jurassic-1 Grande, Jurassic-1 Instruct was trained specifically to handle instructions-only prompts ("zero-shot") that do not require examples ("few-shot"). It is the most natural way to interact with large language models: simply enter a short prompt instructing the model what to do, and get a concise answer from the model.
Here’s an example of what Instruct can do - the World Cup. Even if you’re not a soccer fan, you can be part of the group. Ask Jurassic-1 Instruct to come up with a relevant soccer tweet:
Try it out and discover the power of zero-shot prompting!
Despite being in open beta, the Jurassic-1 Instruct displays phenomenal capabilities, and will continue to improve over time.
We hope you'll enjoy using Jurassic-1 Instruct.