Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/Mahdi-s/llm.guts
https://github.com/Mahdi-s/llm.guts
Last synced: 3 days ago
JSON representation
- Host: GitHub
- URL: https://github.com/Mahdi-s/llm.guts
- Owner: Mahdi-s
- Created: 2024-10-17T02:06:20.000Z (3 months ago)
- Default Branch: main
- Last Pushed: 2024-10-24T05:59:16.000Z (3 months ago)
- Last Synced: 2024-10-24T22:28:10.040Z (3 months ago)
- Language: Python
- Size: 43 KB
- Stars: 2
- Watchers: 3
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
- awesome_ai_agents - Llm.Guts - A tool to visualize the internal computational graph of distilgpt2 model. (Building / Tools)
- awesome_ai_agents - Llm.Guts - A tool to visualize the internal computational graph of distilgpt2 model. (Building / Tools)
README
# 🔬 DistilGPT2 Visualization 🧠
📚 Educational Tool for Exploring Language Model Internals
Welcome to the DistilGPT2 Visualization app! This interactive tool allows you to peek inside the "brain" of a language model and understand its inner workings.
### 🌟 Features
- 🔍 Visualize various components of the DistilGPT2 model
- 💡 Explore embeddings, attention mechanisms, and more
- 📊 Interactive plots and heatmaps
- 🎓 Perfect for students, researchers, and AI enthusiasts
### 🚀 Getting Started
1. 📝 Enter your query in the text box
2. 🔢 Select a transformer block to visualize
3. 🖱️ Click "Submit" to generate visualizations
### 📊 Visualization Tabs
- 📌 Embeddings: See how words are represented numerically
- 🔁 Residual Streams: Observe information flow through the model
- 📏 LayerNorm1 & LayerNorm2: Understand normalization techniques
- 👀 Attention: Visualize how the model focuses on different parts of the input
- 🧮 MLP: Explore the feed-forward neural network components
### 🎯 Educational Goals
- 🧠 Demystify the inner workings of transformer-based language models
- 🔬 Provide hands-on experience with model internals
- 📈 Enhance understanding of NLP concepts through visual representations
### 🛠️ Technical Details
- Built with Python, Gradio, and PyTorch
- Utilizes the `transformer_lens` library for model introspection
- Generates interactive plots using Plotly
### 🏃♂️ Running Locally
To run this project on your local machine, follow these steps:
1. 📥 Clone the repository:
```
git clone https://github.com/Mahdi-s/llm.guts.git
cd llm.guts
```
2. 🐍 Create a virtual environment (optional but recommended):
```
python -m venv venv
source venv/bin/activate # On Windows, use `venv\Scripts\activate`
```
3. 📦 Install the required dependencies:
```
pip install -r requirements.txt
```
4. 🚀 Run the application:
```
python app.py
```
5. 🌐 Open your web browser and navigate to `http://localhost:7860` to access the application.
Note: Ensure you have Python 3.11+ installed on your system before running the application locally.
### 🐳 Running with Docker
Alternatively, you can run the application using Docker:
1. 🏗️ Build the Docker image:
```
docker build -t llm.guts .
```
2. 🐳 Run the Docker container:
```
docker run -p 7860:7860 llm.guts
```
3. 🌐 Access the application at `http://localhost:7860` in your web browser.
Dive in and start exploring the fascinating world of language models!
I'm working on adding support for additional models, let me know which models you'd like to see here. :)