Using-Fidle/Linux-installation-with-pip.md

+This installation procedure is for **Linux**  \
+...and is based on the official Tensorflow procedure, which is described here :
+
+<https://www.tensorflow.org/install/pip>
+
+**Note:** If you have a **GPU card**, things can be more complicated than this simplified procedure. We advise you to consult the documentation mentioned above.
+
+## First of all, you need Python !
+
+A version of python 3.6 to 3.9 is required. 
+
+If python is not available on your system, you can install the version proposed by your Linux distribution or (if you are a bit of a gambler) by compiling a version proposed on [python.org](https://www.python.org/downloads/source/)
+
+For example, under Debian :
+
+```
+apt-get install python3 python3-dev python3-pip python3-venv
+```
+
+## Create your pip virtual env
+
+From your working directory, where you have downloaded or cloned the fidle repository :
+
+```plaintext
+python -m venv --system-site-packages fidle-env
+```
+
+This will create a folder `fidle-env` which hosts your environment.
+
+## Install Tensorflow and additionnal packages
+
+Still from your working directory, activate your environment, by running :
+
+```plaintext
+source fidle-env/bin/activate
+```
+
+...and then, install Tensorflow, and fidle requested packages :
+
+```plaintext
+pip install --upgrade pip
+pip install --upgrade tensorflow
+pip install Scikit-image Scikit-learn Matplotlib Pandas Pandoc pyyaml Jupyterlab
+```
+
+It will take a little while, but should go very well :-)
+
+## Start Jupyter lab
+
+1. Go to your working directory
+2. Activate your environment
+3. Set your environment var
+4. Start Jupyter lab
+
+```plaintext
+cd <your working directory>
+source fidle-env/bin/activate
+export FIDLE_DATASETS_DIR=<path to the datasets dir>
+jupyter lab
+```
+
+[Back to the install procedure](Using%20Fidle/install%20fidle#linux-users)
\ No newline at end of file