Update README.md
Browse files
README.md
CHANGED
|
@@ -12,53 +12,68 @@ pinned: true
|
|
| 12 |
|
| 13 |
|
| 14 |
|
| 15 |
-
|
| 16 |
-
ExplainAnything.AI is an interactive multimodal science explainer built with Gradio. It allows users to:
|
| 17 |
|
| 18 |
-
|
| 19 |
|
| 20 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 21 |
|
| 22 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 23 |
|
| 24 |
-
|
| 25 |
|
| 26 |
-
|
| 27 |
|
| 28 |
-
|
| 29 |
-
Ask a Question
|
| 30 |
-
Type a science question (e.g., How do volcanoes erupt?)
|
| 31 |
-
→ Choose level: Kid, Beginner, or Advanced
|
| 32 |
|
| 33 |
-
|
| 34 |
-
You can upload a PDF, image, or diagram to get contextual explanations
|
| 35 |
|
| 36 |
-
|
| 37 |
-
The app generates a clear explanation and a visual illustration using generative models
|
| 38 |
|
| 39 |
-
|
| 40 |
-
Answer auto-generated multiple choice questions to reinforce learning
|
| 41 |
|
| 42 |
-
|
| 43 |
-
Export everything to a PDF or Markdown file for later review
|
| 44 |
|
| 45 |
-
|
| 46 |
-
Gradio (UI)
|
| 47 |
|
| 48 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 49 |
|
| 50 |
-
|
| 51 |
|
| 52 |
-
|
| 53 |
|
| 54 |
-
|
|
|
|
|
|
|
| 55 |
|
| 56 |
-
|
| 57 |
|
| 58 |
-
|
| 59 |
-
Make sure to add the following API keys as secrets in your Space:
|
| 60 |
|
|
|
|
|
|
|
|
|
|
| 61 |
|
| 62 |
-
GEMINI_API_KEY
|
| 63 |
-
OPENROUTER_API_KEY
|
| 64 |
-
HF_TOKEN
|
|
|
|
| 12 |
|
| 13 |
|
| 14 |
|
| 15 |
+
# ExplainAnything.AI
|
|
|
|
| 16 |
|
| 17 |
+
**Track:** agent-demo-track
|
| 18 |
|
| 19 |
+
**ExplainAnything.AI** is a multimodal agent that helps users understand any science topic — either by **asking a question** or **uploading an image or PDF**. It then builds an interactive explainer package with:
|
| 20 |
+
- 🧠 Easy-to-understand explanation (via Mistral)
|
| 21 |
+
- 🖼️ Auto-generated visual diagram (via Flux)
|
| 22 |
+
- ❓ Quiz questions to test understanding (via Mistral)
|
| 23 |
+
- 📄 Downloadable report summarizing everything
|
| 24 |
|
| 25 |
+
---
|
| 26 |
+
|
| 27 |
+
## 🚀 How It Works
|
| 28 |
+
|
| 29 |
+
You have two options to get started:
|
| 30 |
+
1. **Ask a question**, e.g. *"How do solar panels work?"*
|
| 31 |
+
2. **Upload an image or PDF**, like a diagram or worksheet.
|
| 32 |
+
|
| 33 |
+
The agent then:
|
| 34 |
+
- Uses **Gemini Vision** (for image/PDF) or **your question** as input
|
| 35 |
+
- Generates an explanation using **Mistral**
|
| 36 |
+
- Creates a visual diagram with **Flux**
|
| 37 |
+
- Generates quiz questions using **Mistral**
|
| 38 |
+
- Compiles everything into a downloadable **learning report**
|
| 39 |
+
|
| 40 |
+
---
|
| 41 |
|
| 42 |
+
## 🚧 Build Status
|
| 43 |
|
| 44 |
+
This Space is currently **building**, but all logic, tools, and functionality were submitted **before the deadline**.
|
| 45 |
|
| 46 |
+
---
|
|
|
|
|
|
|
|
|
|
| 47 |
|
| 48 |
+
## 🎥 Video Overview
|
|
|
|
| 49 |
|
| 50 |
+
**Video Overview:** [Coming Soon – will be added post-deadline]
|
|
|
|
| 51 |
|
| 52 |
+
*A short walkthrough of the app’s flow and learning experience will be uploaded here.*
|
|
|
|
| 53 |
|
| 54 |
+
---
|
|
|
|
| 55 |
|
| 56 |
+
## 🛠️ Tech Stack
|
|
|
|
| 57 |
|
| 58 |
+
- **Mistral** – for science explanations + quiz generation
|
| 59 |
+
- **Flux** – for diagram/image generation
|
| 60 |
+
- **Gemini Vision** – for reading image and PDF content
|
| 61 |
+
- **Gradio** – chat + upload interface
|
| 62 |
+
- Manual orchestration (no MCP yet)
|
| 63 |
|
| 64 |
+
---
|
| 65 |
|
| 66 |
+
## 📘 Use Cases
|
| 67 |
|
| 68 |
+
- 🧑🎓 Students learning STEM with visual + interactive help
|
| 69 |
+
- 👩🏫 Teachers turning textbook pages into visual lessons
|
| 70 |
+
- 🧠 Self-learners asking "why/how" questions and getting full reports
|
| 71 |
|
| 72 |
+
---
|
| 73 |
|
| 74 |
+
## 🧠 Future Plans
|
|
|
|
| 75 |
|
| 76 |
+
- Integrate Hugging Face MCP for agent orchestration
|
| 77 |
+
- Add TTS narration for accessibility
|
| 78 |
+
- Generate downloadable PDF learning packs
|
| 79 |
|
|
|
|
|
|
|
|
|