Got it, let's solve this once and for all—no more wasting time reinstalling OpenCV, Pandas, Numpy, or any other dependencies every time you fire up a Colab session. Here's a straightforward, repeatable workflow to create a virtual environment that lives in your Google Drive, so your dependencies stay intact between sessions:

Step 1: Mount Your Google Drive #

First, we need to link Colab to your Drive so we can save the environment there permanently. Run this code block and follow the authorization prompt to connect your account:

from google.colab import drive
drive.mount('/content/drive')

This will map your Drive to the /content/drive path in Colab—all our environment files will live here from now on.

Step 2: Create the Virtual Environment #

Colab comes with venv pre-installed, so we'll use that for simplicity. Run this command to create your environment in Drive (feel free to replace colab_env with a name that makes sense for your project):

!python -m venv /content/drive/MyDrive/colab_env

This creates a folder colab_env in your Drive's root directory—this is where all your environment's binaries and installed packages will be stored.

Step 3: Activate the Environment & Install Dependencies #

Now we need to activate the environment and install your required packages. In Colab, we'll chain the activation and pip install commands to ensure we're installing directly into our Drive-based environment (not Colab's default temporary one):

!source /content/drive/MyDrive/colab_env/bin/activate && pip install opencv-python pandas numpy

Add any other dependencies you need to the end of the pip install line—they'll all be saved in your Drive environment.

Optional: Save Your Dependencies List #

To keep track of exactly what's installed (and make it easy to restore if needed), save a requirements.txt file to your environment folder:

!source /content/drive/MyDrive/colab_env/bin/activate && pip freeze > /content/drive/MyDrive/colab_env/requirements.txt

Step 4: Reuse the Environment in Future Sessions #

Next time you open Colab, you don't need to recreate the environment—just follow these two quick steps:

1. Mount your Drive again (same as Step 1)
2. Activate the environment and start working:

!source /content/drive/MyDrive/colab_env/bin/activate

If you ever need to reinstall dependencies (e.g., if you accidentally modify the environment), you can use the saved requirements.txt:

!source /content/drive/MyDrive/colab_env/bin/activate && pip install -r /content/drive/MyDrive/colab_env/requirements.txt

Quick Tips to Avoid Headaches #

• Always activate before installing: If you skip activation, you'll install packages to Colab's default environment, which won't persist after the session ends.
• Verify your pip path: After activation, run !which pip to confirm it points to /content/drive/MyDrive/colab_env/bin/pip—this ensures you're using the environment's pip, not the global one.
• Watch Drive storage: Virtual environments can take up space, especially with large packages like OpenCV. Keep an eye on your Drive's storage limits if you're working with multiple environments.

内容的提问来源于stack exchange，提问作者Sahar Ben Mabrouk