Inspiration
the youtube series deathbattle
What it does
uses tensor flow to find the best fight for a villain using a framework we built and then used gbt api to create story around it
How we built it
used tensor flow to create a linear representation of the data after parsing it and then used it to find the most similar output to create a fight. the fight charaters were then used in a promt that was passed into the api which then returned the story and descriptions of who they are and how they fight.
Challenges we ran into
first time using tensor flow pandas and numpy so it was a steep learning curve of learning how to parse data, lots of index errors, implantations of functions into the front end, and tried to create it without chunking which resulted in large data files that created crashing situations
Accomplishments that we're proud of
that the tensor flow works well and can accurately find similar matches also we learned a new front end library in python
What we learned
how to use steam lit, numpy , tensorflow , pandas, and chatgbt api
What's next for Power Match
add online debates character displays random generation
Log in or sign up for Devpost to join the conversation.