update documentation cpu vs gpu

Author: mathrip

README.md

@@ -12,14 +12,15 @@ This package is a pipeline to segment FCD-lesions from MRI scans.
 
 ## UPDATE
 
-**<span style="color: red;">SIGN UP TO THE MELD GRAPH MAILING LIST</span>**:
-We request that all MELD Graph users sign up to the mailing list. If you are using MELD Graph, please send an email to `meld.study@gmail.com` with the subject 'Request to be added to the MELD Graph mailing list' and provide use with your name and institute. This will ensure that we can update you about bug fixs and new releases. 
+**<span style="color: red;">REGISTER TO MELD GRAPH TO GET THE MELD LICENSE</span>**:
+We request that all MELD Graph users are registered by filling the [MELD form](https://docs.google.com/forms/d/e/1FAIpQLSdocMWtxbmh9T7Sv8NT4f0Kpev-tmRI-kngDhUeBF9VcZXcfg/viewform?usp=header). Following registration you will received a license file. This file will be needed for use of all future MELD Graph versions v2.2.4 and above. Your email adress will be added to the MELD Graph mailing list. This will ensure that we can update you about bug fixs and new releases. 
 
-**<span style="color: red;">IF YOU ARE STILL RUNNING WITH V2.2.1 - PLEASE UPDATE TO VERSION V2.2.2</span>**:
-We have released MELD Graph V2.2.2 which fixes a couple of issues found by users. For more information about the release please see [MELD Graph V2.2.2](https://github.com/MELDProject/meld_graph/releases/tag/v2.2.2). To update your code please follow the guidelines [Updating MELD Graph to V2.2.2](https://meld-graph.readthedocs.io/en/latest/FAQs.html#Updating-MELD-Graph-to-V2.2.2) from our FAQ.
+**<span style="color: red;">PLEASE UPDATE TO V2.2.4</span>**: 
+We have released MELD Graph v2.2.4 and v2.2.4_gpu, the new stable versions of MELD Graph with updated licensing and small bug fix.
+If you have GPU ressources with at least 20GB of VRAM and would like to use GPU for Fastsurfer segmentation and accelerated prediction, please install v2.2.4_gpu. 
 
-**<span style="color: red;">NEW RELEASE V2.2.3 IN TEST - FOR GPU USERS</span>**: 
-We have released MELD Graph v2.2.3 which enable to use the Docker with GPU. If you already have MELD Graph V2.2.2 and do not need to use the GPU ressources, we recommand to keep your current version. If you want to use the GPU please update your code to get the latest MELD Graph v2.2.3 docker. 
+**<span style="color: red;">IF YOU ARE STILL RUNNING WITH V2.2.1 - PLEASE UPDATE TO THE LATEST VERSION</span>**:
+We have released MELD Graph V2.2.2 and V2.2.4 which fixes a couple of issues found by users. For more information about the release please see [MELD Graph V2.2.2](https://github.com/MELDProject/meld_graph/releases/tag/v2.2.2) and [MELD Graph V2.2.4](https://github.com/MELDProject/meld_graph/releases/tag/v2.2.4). To update your code please follow the guidelines [Updating MELD Graph to V2.2.2 and above](https://meld-graph.readthedocs.io/en/latest/FAQs.html#Updating-MELD-Graph-to-V2.2.2) from our FAQ.
 
 ![overview](https://raw.githubusercontent.com//MELDProject/meld_graph/main/docs/images/Fig1_pipeline.jpg)
 

docs/install_docker.md

@@ -7,7 +7,7 @@ The Docker container has all the prerequisites embedded on it which makes it eas
 Notes: 
 - Currently only tested on **Linux**. HPC users should use the [Singularity version](https://meld-graph.readthedocs.io/en/latest/install_singularity.html). Mac M chip computers have to do a [install_native](https://meld-graph.readthedocs.io/en/latest/install_native.html)
 - You will need **~13GB of space** to install the container
-- The docker image contains Miniconda 3, Freesurfer V7.2, Fastsurfer V1.1.2 and torch 1.10.0+cu113. The whole image is 20 GB.
+- The docker image contains Miniconda 3, Freesurfer V7.2, Fastsurfer V1.1.2 and torch 1.10.0. The whole image is ~13 GB.
 - The predictions stage can use over **20GB of RAM/VRAM**, therefore we recommend using a computer of **at least 24GB of RAM (for CPU) or VRAM (for GPU)**.
 
 Here is the video tutorial detailing how to install the Docker - [Docker Installation](https://youtu.be/oduOe6NDXLA).
@@ -28,9 +28,9 @@ On windows, Docker should be [using WSL2](https://docs.docker.com/desktop/wsl/).
 :::
 
 
-## Enable GPUs
+## Enable GPUs (compatible with MELD Graph GPU version only )
 
-Enabling your computer's GPUs for running the pipeline accelerates the brain segmentation when using Fastsurfer and the predictions. Follow instructions for your operating system to install.
+Enabling your computer's GPUs for running the pipeline accelerates the brain segmentation when using Fastsurfer and the predictions. Ensure you have installed a MELD Graph version compatible with GPU (see [release versions](https://docs.google.com/forms/d/e/1FAIpQLSdocMWtxbmh9T7Sv8NT4f0Kpev-tmRI-kngDhUeBF9VcZXcfg/viewform?usp=header)). Follow instructions for your operating system to install.
 
 ::::{tab-set}
 

docs/install_native.md

@@ -107,14 +107,18 @@ Then activate your environment by running the following:
 conda env create -f environment.yml
 # activate the environment
 conda activate meld_graph
-# add the torch packages 
+# add the torch CPU packages (see below for GPU)
 pip install --no-cache-dir torch==1.10.0 torchvision==0.11.1 && pip install -e . && pip install torch-scatter -f https://data.pyg.org/whl/torch-1.10.0.html && pip install torch-geometric==2.4.0 && pip install captum==0.6.0
+# for use of GPU , install the Torch packages below instead
+# pip install --no-cache-dir torch==1.10.0+cu113  torchvision==0.11.1+cu113  && pip install -e . && pip install torch-scatter -f https://data.pyg.org/whl/torch-1.10.0+cu111.html && pip install torch-geometric==2.4.0 && pip install captum==0.6.0
 # install meld_graph with pip (with `-e`, the development mode, to allow changes in the code to be immediately visible in the installation)
 pip install -e .
 ```
 :::
 ::::
 
+
+
 ## MELD license
 In order to run MELD Graph you need to have a `meld_license.txt` in the meld graph folder. To get this file, please fill out the [MELD registration form](https://docs.google.com/forms/d/e/1FAIpQLSdocMWtxbmh9T7Sv8NT4f0Kpev-tmRI-kngDhUeBF9VcZXcfg/viewform?usp=header). Once submitted, your application will be automatically reviewed and the meld_license.txt file will be send to your email. 
 

docs/install_singularity.md

@@ -7,7 +7,7 @@ The Singularity container has been created to be used on HPC supporting Linux as
 Notes: 
 - The Singularity image is built from the Docker container. 
 - You will need **~13GB of space** to install the container
-- The image contains Miniconda 3, Freesurfer V7.2, Fastsurfer V1.1.2 and torch 1.10.0+cu113. The whole image is 20 GB.  
+- The image contains Miniconda 3, Freesurfer V7.2, Fastsurfer V1.1.2 and torch 1.10.0. The whole image is 13 GB.  
 
 ## Prerequisites
 
@@ -29,7 +29,7 @@ In order to run MELD Graph you need to have a `meld_license.txt` in the meld gra
 ## Configuration
 In order to run the singularity image, you'll need to build the singularity image from the meld_graph docker image. This will create a singularity image called meld_graph.sif where you ran the command. 
 
-Make sure you have 20GB of storage space available for the docker
+Make sure you have 13GB of storage space available for the docker
 
 ```bash
 singularity build meld_graph.sif docker://meldproject/meld_graph:latest