Update everything to use packages from ghcr.io

Author: Frankie Robertson

README.md

@@ -4,8 +4,8 @@
 <a href="https://gitlab.com/frankier/skelshop/-/commits/master">
   <img alt="pipeline status" src="https://gitlab.com/frankier/skelshop/badges/master/pipeline.svg" />
 </a>
-<a href="https://hub.docker.com/r/frankierr/skelshop/builds">
-  <img alt="DockerHub hosted images" src="https://img.shields.io/docker/pulls/frankierr/skelshop?style=flat" />
+<a href="https://github.com/frankier?tab=packages&repo_name=skelshop">
+  ghcr.io
 </a>
 <a href="https://frankier.github.io/skelshop/">
   <img alt="Documentation on GitHub pages" src="https://img.shields.io/badge/Docs-MkDocs-informational" />

docs/index.md

@@ -28,22 +28,22 @@ There are two Docker containers, one based on CUDA and one able to run
 CUDA 10 and [nvidia-docker](https://github.com/NVIDIA/nvidia-docker) on the
 host and then run:
 
-    $ docker run --nv frankierr/skelshop:focal_nvcaffe python -m skelshop --help
+    $ docker run --nv ghcr.io/frankier/skelshop_nvcaffe python -m skelshop --help
 
 For the CPU version:
 
-    $ docker run frankierr/skelshop:focal_cpu python -m skelshop --help
+    $ docker run ghcr.io/frankier/skelshop_cpu python -m skelshop --help
 
-For more information about the Docker containers see [their page on Docker
-Hub](https://hub.docker.com/r/frankierr/skelshop) and [the
+For more information about the Docker containers see [the GHCR
+packages](https://github.com/frankier?tab=packages&repo_name=skelshop) and [the
 openpose_containers repository where the bases are
 built](https://github.com/frankier/openpose_containers).
 
 ### Running the Docker container with Singularity
 
 You can also run the container with Singularity, which is convenient in HPC environments.
 
-    $ singularity pull skelshop.sif docker://frankierr/skelshop:focal_nvcaffe
+    $ singularity pull skelshop.sif docker://ghcr.io/frankier/skelshop_nvcaffe
     $ singularity run --nv skelshop.sif python -m skelshop --help
 
 ### Manual setup

docs/snakemake.md

@@ -40,7 +40,7 @@ First set up singslurm2:
 
 Now you can download the Docker image with Singularity:
 
-    $ singularity pull skelshop.sif docker://frankierr/skelshop:focal_nvcaffe
+    $ singularity pull skelshop.sif docker://ghcr.io/frankier/skelshop_nvcaffe
 
 Next, you need to create a JSON file specifying which type of nodes you would
 like to assign to different rules (steps in the workflow). [There is an example

docs/usage-examples.md

@@ -20,12 +20,12 @@ ourselves like so:
 We could also use the Singularity or Docker. OpenPose is installed in the image and
 everything is setup up for us so we can just run:
 
-    $ singularity pull skelshop.sif docker://frankierr/skelshop:focal_nvcaffe
+    $ singularity pull skelshop.sif docker://ghcr.io/frankier/skelshop_nvcaffe
     $ singularity exec --nv skelshop.sif python /opt/skelshop/skelshop dump video_in.mp4 pose_data.h5
 
 **OR**
 
-    $ docker run --nv frankierr/skelshop:focal_nvcaffe python /opt/skelshop/skelshop dump video_in.mp4 pose_data.h5
+    $ docker run --nv ghcr.io/frankier/skelshop_nvcaffe python /opt/skelshop/skelshop dump video_in.mp4 pose_data.h5
 
 You can track an existing dump using the `filter` command with the `--track`
 flag or apply tracking at the same time as dumping. Currently scene