Inspiration
By AI Dungeon
What it does
DawnDungeon is a D&D simulation, which will randomly generate attributes of the player such as their name, age, inventory and quests of the player. The player can perform certain actions in which the model will display any relevant output to fit the action.
How we built it
We used OpenAI and prompt it by giving it a role of a story master, additionally prompting it to return the response in a json format, in which our software can read to display any relevant information in the application
Challenges we ran into
Setting Up the API for the application to work with, initially we used FastAPI and Starlette to write the application, however we encountered some error in which we were unable to solved. Unfortunately, we amped to use Streamlit instead, since it did take quite a lot of effort to build the frontend application of Starlette only for us to not use it.
Accomplishments that we're proud of
We are very proud to have finish DawnDungeon, whilst it is not perfect, we plan to make it better and more accurate to what the players want, simulating battles, utilising weapons and handling in game events. It is a start that we hope to improve to the point that we will be satisfied with.
What we learned
Too much
What's next for DawnDungeon
Production
Built With
- javascript
- langchain
- mongodb
- openai
- pinecone
- python
- sqlalchemy
Log in or sign up for Devpost to join the conversation.