This is a submission for the Cloudflare AI Challenge.
What I Built
UI may appear broken, but I couldn't have done better.
Technologies used in project:
- FastAPi
- React
- Tailwind
⚠️Initially my idea was to create a simple image generator inspired by the impressive visuals of some NFTs. I wanted to throw myself deep down into the possibility of using a model to generate NFTs based on user input.
✨Additionally, I included a feature to help users improve their commands by upgrading and adding extra details into given command.
Also another feature allows users to request a quote along with an image of its author, adding a touch of authenticity.
🎈Moreover, Ask-me supports content in multiple languages by automatically translating it into English. Users can also upload audio files to receive a transcript containing all the words spoken, along with a word count.
Demo
My Code
itz-Amethyst
/
ask-me
Your go-to for personalized NFT creation , quotes with it author image translate any context into English, can Identify the text inside an audio file (Cloudflare Hackathon)
Ask-me is a service designed to assist with various tasks, such as NFT creation, generating quotes, and translation.
Technologies used in project:
- FastAPi
- React
- Tailwind
UI may appear broken, but I couldn't have done better.
✨Additionally, I included a feature to help users improve their commands by providing feedback and suggestions. Another feature allows users to request a quote along with an image of its author, adding a touch of authenticity.
🎈Moreover, Ask-me supports content in multiple languages by automatically translating it into English. Users can also upload audio files to receive a transcript containing all the words spoken, along with a word count.
NFT_Model = "stable-diffusion-xl-lightning"
Chat_Model…Journey
Given that this was my first touch with workers and AI models, the circumstances were undeniably challenging. In particular, the pressure to complete the task within just 1/5 day it did cost sacrificing sleep
If I had more time, there are a few things I'd like to add or improve:
- Enhance the user experience on the frontend, especially in terms of design. Making it more visually appealing can make a big difference.
- Implement a voice recording feature for whispers. It could add a personal touch and make interactions more engaging.
- Enable a feature to save pictures to the D1 database, associating them with user IP and browser information as identifiers.
- Find a solution to deploy Python workers along with their requirements on Cloudflare. This could streamline the deployment process and make scaling easier.
Multiple Models and/or Triple Task Types
NFT_Model = "stable-diffusion-xl-lightning"
Chat_Model = "llama-2-7b-chat-fp16"
Improve_NFT_Command_Model = "hermes-2-pro-mistral-7b"
Quote_Model = "mistral-7b-instruct-v0.2"
Translate_Model = "m2m100-1.2b"
Audio_Model = "whisper"
Top comments (0)