Skip to content

juldpnt/flower-classification-transfer-learning

Repository files navigation

This repository follows the tutorial of this Gitlab made by Gaspard Dussert & Vincent Miele (CNRS/LBBE).

It is recommended to use a virtual environment to avoid conflicts with other projects, using venv or conda. Venv won't be explained here, but you can find more information here.

If you don't have anaconda, you can either install the lightweigth version miniconda or install the full version anaconda.

You can create a virtual environment with the following command with an anaconda/miniconda prompt:

conda create -n myenv pip

Then, you can activate your virtual environment with the following command:

conda activate myenv

You can move to the directory of the project with the following command:

cd path/to/directory

And finally, you can install the required packages with the following command while in the right directory:

pip install -r requirements.txt 

If you want to run the notebook, write the following command while being in the right directory, it will open a jupyter notebook in your browser:

jupyter lab

About

Classification of few flowers based on transfer learning

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published