AI Studio by Metric Coders: A No-Code Software to train, download and deploy Large Language Models.
MIT License
AI Studio is a GitHub repository created by Metric Coders. It is a no-code software to create and train various Large Language Models by configuring different hyperparameters through the UI.
Due to lack of processing power, we are currently focussing only on providing a robust platform for creating ML models using scikit-learn.
The platform is designed to simplify the process of building and experimenting with LLMs, including ML models, making it accessible for users with varying levels of expertise in machine learning. This application is built using Streamlit to provide an interactive and user-friendly experience.
The platform offers a one-click deployment zip file that contains everything you need to directly deploy the model as an API into your development and production environments.
Here is how the webpage looks when pre-trained datasets are loaded
Here is how the homepage looks when a custom dataset is loaded
Visualization your data effortlessly in different charts.
Visualization your data effortlessly in different tables.
git clone https://github.com/MetricCoders/AI-Studio.git
cd AI-Studio
pip install -r requirements.txt
To start the Streamlit application, use the following command:
streamlit run Home.py
This will launch the AI Studio web UI. Open your web browser and navigate to the URL provided in the terminal to start using the platform.
After training, AI Studio provides performance metrics and visualizations to help you evaluate the effectiveness of your model. Use these insights to refine your model by adjusting hyperparameters or selecting different algorithms.
The "Download Model" button allows you to instantly download the model to your local directory.
The "Download Deployment Zip" button allows you to instantly download a zip file that contains everything you need to run the model, including the model itself. It contains:
pip install -r requirements.txt
In the menu on the left, Large Language Models section allows you to train and fine-tune large models. You can add as many layers as you need. Due to Streamlit cloud constraints for the demo purpose, the values are limited. In your local, you can change it to whatever you need depending on your system.
The number of epochs have been limited to 10 and the batch_size is 16.
The datasets should have headers and the column by the name 'target' should contain the output/result. A sample is given below:
feature1,feature2,feature3,target
1,2,3,0
2,3,4,0
3,4,5,1
4,5,6,1
5,6,7,0
6,7,8,1
7,8,9,0
8,9,10,1
9,10,11,1
10,11,12,0
This project is licensed under the MIT License. See the LICENSE
file for details.
For questions, feedback, or support, please contact me via email at [email protected].
AI Studio by Metric Coders aims to democratize machine learning by providing an intuitive platform for model creation and experimentation. With upcoming features and continuous improvements, it strives to be a comprehensive tool for both novice and experienced data scientists.