File size: 2,872 Bytes
639fb99
 
 
 
 
 
 
 
 
d342e49
 
 
 
 
 
 
 
 
 
 
 
 
 
e9663ac
 
 
 
 
 
d342e49
 
 
 
 
 
 
 
 
 
 
e9663ac
 
d342e49
 
e9663ac
d342e49
 
 
 
e9663ac
d342e49
e9663ac
 
d342e49
 
e9663ac
 
 
d342e49
e9663ac
d342e49
 
e9663ac
 
d342e49
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
e9663ac
d342e49
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
---
license: mit
title: Audio Lecture to Notes Generator
sdk: gradio
emoji: πŸ“š
colorFrom: yellow
colorTo: yellow
short_description: An agent that converts lecture audio into clean Notes
---
# πŸŽ“ AI Lecture Notes Generator

Transform lecture audio into structured, comprehensive notes using AI.

## Features

- πŸŽ™οΈ Transcribes lecture audio using OpenAI's Whisper Large V3 model
- πŸ“ Generates well-structured markdown notes using Meta-Llama-3.1-8B
- 🧠 Identifies key concepts, points, and takeaways automatically
- πŸš€ Simple web interface built with Gradio
- ⚑ Optimized with 4-bit quantization for efficient inference

## Demo



<video controls autoplay src="https://cdn-uploads.huggingface.co/production/uploads/64ed91f11d42449431ee2253/0J7jWig2JLimC8n2NF0KO.mp4"></video>




## How It Works

1. **Audio Transcription**: Converts lecture audio to text using Whisper Large V3
2. **Note Generation**: Processes transcription through Llama 3.1 to create structured notes
3. **Format Cleaning**: Ensures consistent markdown formatting with proper headers

## Installation

```bash
# Clone the repository
git clone https://github.com/Abdulllah-Rizwan/AI_Lecture-s-Note-Generator.git
cd AI-Lecture-Notes-Generator

# Install dependencies
pip install requests torch bitsandbytes transformers sentencepiece accelerate gradio httpx==0.27.2
```

## Usage

### Option 1: Run the Jupyter Notebook

```bash
jupyter notebook "AI Lecture Notes Generator.ipynb"
```

Run all cells in the notebook to start the Gradio interface, then interact with the application through the displayed URL (typically http://127.0.0.1:7860/).

### Option 2: Use the functions directly in your code

You can import the functions from the notebook into your own Python code:

```python
# Import the necessary functions
from AI_Lecture_Notes_Generator import process_audio

# Generate notes from an audio file
notes = process_audio("path_to_lecture_audio.mp3")
print(notes)
```

## Requirements

- Python 3.8+
- PyTorch
- Transformers
- Gradio
- Hugging Face account with appropriate model access
- CUDA-compatible GPU (recommended)

## Output Format

The generated notes follow this structure:

```markdown
# Topic: [Lecture Topic]
## Instructor
[Instructor name if detected]
## Summary
[Concise summary of the lecture content]
## Key Concepts
- [Important concept 1]
- [Important concept 2]
...
## Key Points
- [Major point 1]
- [Major point 2]
...
## Takeaways
- [Takeaway 1]
- [Takeaway 2]
...
```

## Limitations

- Best suited for academic lectures with clear structure
- Audio quality significantly impacts transcription accuracy
- May require fine-tuning for specialized terminology

## License

MIT License.

## Acknowledgements

- OpenAI for the Whisper model
- Meta for the Llama 3.1 model
- Hugging Face for model hosting and libraries
- Gradio for the web interface framework