The new deadline for submission is July 24 due to community request.
This month’s community challenge is all about envisioning the future of AI-enabled apps. We would like to officially open this initiative to the community and challenge you to build a Dash app that integrates the ChatGPT API into your app.
The winning apps will be judged according to the following categories:
- Depth of integration of ChatGPT API into the Dash app
- App Design
- Data exploration and data analysis routines (e.g. numerical methods, machine learning, prediction, classification, optimization)
Integrate ChatGPT into an app in a safe manner is increasingly important. To keep your app safe, we encourage you to read all about Prompt Hacking.
The winners will be announced in the next Dash Club dispatch and will receive a reward of:
To submit your app, share it by replying to this Forum topic by midnight Monday,
July 10 (now July 24). Please include a link to the app, the code on GitHub, and a short description of the app.
For any questions, feel free to reply to the Forum topic or message me (Adam) directly.
Dash-ChatGPT minimal app example: GitHub repo with code.
Dash-ChatGPT app example with streaming text:
Great idea, let’s see what the community comes up with.
I have a Dash app that does entity extraction using ChatGPT.
You can paste an article in the textarea, and it will:
- Extract entities from the article
- Provide the Wikipedia URLs for each entity
- Provide the output in JSON-LD FORMAT
This was trained with a thousand examples (a fine-tuned model, instead of regular prompts).
Video in this tweet showing how it works.
An article with code explaining
how the app was built is being finalized how to finetune ChatGPT for entity extraction has been published. It also links to the GH repo, with full code, as well as the training data.
Looking forward to seeing other contributions.
Thanks for the submission, @eliasdabbas . Great way to add ChatGPT to the app
Thank you for sharing this information about the community challenge.
For this challenge, I have developed a Model Competition app that allows users to compare different LLM models in terms of speed and the quality of their responses. It provides an interactive platform where you can assess multiple models simultaneously and determine which one performs the best.
One can access the app through the following link:model_competition
GitHub : Link
One can check Medium Post:Link
Looking forward to seeing other contributions.
Thank you @abhinva. I’ve heard people ask which LLM is better. This app should help them make their own judgement
I also saw all the other apps you have on your portfolio. Really nice collection!
Thanks, Adam for your kind words.
AI Summarizer App using OpenAI GPT-3.5 API
I have developed an AI Summarizer app powered by OpenAI GPT-3.5 API, a cutting-edge language model. The app leverages the advanced capabilities of GPT-3.5 to generate accurate and coherent summaries of various types of texts, ranging from articles and blog posts to reports and documents.
- Advanced Language Model: The app utilizes OpenAI GPT-3.5, a state-of-the-art language model, to understand the context, extract key information, and generate concise and coherent summaries.
- Natural Language Understanding: GPT-3.5 comprehends the nuances of the input text, ensuring that the generated summaries capture the essential details and maintain the overall meaning and coherence.
- Customizable Summarization Options: Users have the flexibility to customize the length and style of the summaries, tailoring them to their specific requirements.
- User-Friendly Interface: The app features a user-friendly interface that allows users to easily input the text they want to summarize and obtain the summarized output within seconds.
You can access the source code and contribute to the development of the AI Summarizer app on my GitHub repository: [GitHub Repository Link]
Experience the power of AI-generated summaries by trying out the AI Summarizer app at the following link:
I am excited to have created this AI Summarizer app using OpenAI GPT-3.5 API, which enables users to obtain accurate and concise summaries effortlessly.
Cool! No API key required? Isn’t this app going to cost you a bunch of money?
I’m excited to check this app out. I copied two article links into the input field but I got this error:
An error occurred: Unexpected token O in JSON at position 0
Do you know where this is coming from?
My app was actually designed to fetch earthquake information from the records of the Kandilli Observatory and Earthquake Research Institute in Turkey. I retrieved the data by parsing the XML file in the desired format. Later, I visualized this information on various types of maps, including scatter, density, and hexagon layers. I also integrated a search bar with ChatGPT for brief interactions. Now, what I wanted to do was to introduce my dataframe to ChatGPT and have it provide me with the requested code snippets.GitHub - ceyhunsahin/earthquake, Afterwards, I planned to display the results using Python code. Although there are some CSS deficiencies, I shared a video on Twitter which you can review and provide feedback on. https://twitter.com/melkebir1/status/1671934468976787473?s=20
very nice @ceyhun
Were you able to deploy the app to the web?
I’ll try to deploy but there were 3 million row in my dataset so I need to get a sample. If you have a docker, you can open it directly from my GitHub explanation that will be more effective😂 but when you’re ask for something you need to insist with explained question😁
Hi @ceyhun . Yes, even if you just deploy with a small sample data set, that would be great.
Nice community challenge!
For the needs of a circle of friends and me, I’ve built the Dash app for:
- An easy communication with GPT-4 or 3.5 (first UI tab), which provides some AI personalities to chat with - or you can describe a custom one. It also supports markdown formatting and Latex for displaying equations
- A summarization (second UI tab) of arbitrarily long PDF documents and web pages. There are three summarization options (normal summary, bullet points, focus on a question) and it tries to provide a structure with headings and subheadings.
I can’t provide a link to the working app due to potential costs but I have published it on GitHub so anyone can use it with their API keys:
I hope it can be helpful for others, too.
Here’s my submission for the challenge!
Find the AI
In Find the AI, you will be presented with two quotes—one from a human and one generated by ChatGPT. Your objective is to identify which quote was written by the AI. You are presented with 10 quotes per day, and are given a score at the end of the game. The deployed app is available down below, and you can also run it locally by including your own OpenAI API key.
Some technical details:
- Calls to external API for random quote and ChatGPT for AI quote
- Responses from ChatGPT are cached to avoid unnecessary calls
- Clientside callbacks are used to update the state of the game
- New set of quotes are generated and saved to cache each day by scheduled task
Github: GitHub - ddl004/find-the-ai-dash: Real or AI quotes made with Dash
Very interesting app. Thanks for sharing, @LeRobert
Awesome app, @ddl004 . I got almost every puzzle wrong. I officially cannot tell the difference between humans and bots