6551AI

this project takes two embeddings of books one by Nietzsche and Pride and prejudice. It takes a user's input query and embeds this query and performs a nearest neighbour search to generate a new embeddings that may be used to fine tune the original model or to create a new model. This new embedding is minted as an erc-6551 NFT to be actionable, transferrable.

Python with flask. Embeddings were created with open ai, most of it done on a colab notebook. NFT.storage for ipfs storage. started with an open ai gpt demo source. it's a simple script that takes embeddings in the file system on cvs files. Manipulates the embeddings, calls openai to generate an embedding with the user query prompted from the front end of the app. The embedding is turned into JSON. the JSON is rendered on the front end. There is a button for the user to click on to mint an nft with the JSON as metadata to the NFT. The JSON metadata is stored on ipfs with NFT.storage