Update README.md

README.md

@@ -1,21 +1,60 @@
-# Chat with PDF
+---
+license: mit
+language:
+- en
+base_model:
+- deepseek-ai/DeepSeek-R1
+- Qwen/Qwen2.5-VL-72B-Instruct
+---
 
-# Pre-requisites
-Install Ollama on your local machine from the [official website](https://ollama.com/). And then pull the Deepseek model:
+# Model Card for Model ID
 
-```bash
-ollama pull deepseek-r1:14b
-```
+AI that can answer user's queries based on the user's given pdf, url links and straight direct researchers
 
-Install the dependencies using pip:
+It has three main functions , PDF-ChatBot, Edu-Scraper, and Edu Researcher.
 
-```bash
-pip install -r requirements.txt
-```
 
-# Run
-Run the Streamlit app:
+## Installation 
 
-```bash
-streamlit run pdf_rag.py
-```
+1. ```cd Edu-Ai```
+
+2. ```pip install requirements.txt```
+
+3. ```pip install streamlit```
+
+4. ```streamlit run Hellopage.py```
+
+5. Please get your own hugging face token to add in secrets.toml so that you can use the models.
+
+### Model Description
+
+1. ```Hellopage.py``` : The starting page where you can choose the options between PDF-ChatBot, Edu-Scraper, Edu-Researcher.
+
+2. ```pdf_rag.py``` : Using RAG, user can input their queries and pdfs so that they can recieve their desired answers.
+
+3. ```Edu-Scraper.py``` : Using RAG and selenium, user can input their url and their queries so that they can recieve their desired answers.
+
+4. ```Edu-Researcher.py``` : Using DDGS and Stategraph, user can get straight to the point searches.
+
+
+
+- **Developed by:** [UIT batch 10 Section A]
+- **Model type:** [RAG]
+- **Language(s) (NLP):** [English]
+- **License:** [MIT]
+
+### Model Sources
+
+- **Repository:** [https://github.com/FightaHackathon/Edu-Ai]
+- **Demo:** [https://edu-ai.streamlit.app/]
+
+## Uses
+
+- Able to summarize pdfs and urls given by user's queries and straight to the point researches.
+
+
+### Important notes
+
+1. Users can choose between qwen and deepseek for their preference modes(Can add more models by just modifying the codes slightly
+
+2. Please get your own hugging face token from making an account in https://huggingface.co/welcome