This is an example Streamlit application set up with a conda virtual environment created for educational purposes.
1- Run git init in empty project folder to set up git for deployment later.
2- Run conda create --name <ENV_NAME> python=<PY_VERSION> to create a virtual environment for easily sharing the used packages. Note: replace ENV_NAME with desired environment name e.g. myenv. Also replace PY_VERSION with the desired Python version e.g. 3.10.
3- Run conda activate <ENV_NAME> to activate the virtual environment you just created. This allows you to install the packages inside this package.
4- Create a Python file with any name you desire e.g. app.py.
5- Install Streamlit package via pip install streamlit.
6- Install any additional packages you need e.g. pip install langchain openai.
7- *Optional step* Create .env file that contain project secrets like OPENAI_API_KEY to your project folder.
8- *IMPORTANT* If you have created a .env file, make sure you also create a .gitignore file and write down all files you wish to ignore including .env.
9- Create a Streamlit app by following the official documentation at Streamlit docs.
10- Save the current state of your virtual environment using conda env export > environment.yml. This will create an environment.yml file that lists all the dependencies of your project.
11- To run your Streamlit application, use the command streamlit run app.py (replace app.py with the name of your Python file if different).
12- If you want to share your project, push your code to a Git repository and include the environment.yml file so others can recreate the same environment using conda env create -f environment.yml --name <ENV_NAME>.
13- For deployment, please use requirements.txt by executing pip freeze > requirements.txt as per Streamlit deployment guide or other hosting platforms like Heroku or AWS.
Happy coding!