---

Introduction

Nexus is an AI-based Answer Sheet Evaluation System. Created to automate and accelerate the grading process using advanced Natural Language Processing (NLP) Techniques, offering fast and accurate results. Nexus stands for ${\color{blue}N}$ eural ${\color{blue}E}$ valuatory e ${\color{blue}X}$ pert ${\color{blue}U}$ nified ${\color{blue}S}$ ystem. More information about Nexus and how the system evaluate answers are described in this NexusAI pdf file.

Team Members

1. Umer Farooque
2. Prathyush N M
3. Sebinmon V R
4. Gautham Sugathan S

---

Note

You Must Have An HuggingFace and Streamlit Account For Running This Project.

---

Installation

1. Download the ZIP file, unzip it, and open the resulting folder in your preferred IDE, such as Visual Studio Code.

2. Install the libraries mentioned in requirements.txt In VsCode Terminal run the following command:

   pip install -r requirements.txt

---

Important

Since we are using models from huggingFace and calling it using Inference API (serverless) method you should have a huggingface account and api_key, This api key must be placed in apicall.py file.

3. Your api key will look like this : 'hf_XXXXXXXXXXXXXXXXXXXXX', copy and paste it in the mentioned area of apicall.py file.

4. Run the main.py python file.

   Streamlit run main.py

---

Tip

Since we are using Large NLP models, sometimes while running the main.py some errors like ${\color{red}TypeError: string indices must be integers}$ and ${\color{red}KeyError: 0}$ will occur so in streamlit tap ${\color{green}run}$ button ${\color{green}multiple times}$ to tackle this errors.