Skip to content
master
Go to file
Code
This branch is 7 commits ahead of gewimmer-neuro:master.

Latest commit

 

Git stats

Files

Permalink
Failed to load latest commit information.
Type
Name
Latest commit message
Commit time
css
 
 
 
 
js
 
 
 
 
 
 
 
 
 
 

README.md

Experiment Factory Experiment

Hi Friend! This is an Experiment (adopted from gewimmer-neuro/reward_learning_js) that is friendly for use in the Experiment Factory. You can run it locally by putting these files in a web server, or use the Experiment Factory to generate a reproducible container. Check out the documentation above for more information, or post an issue if you have any questions.

https://expfactory.github.io/expfactory/img/expfactoryticketyellow.png

I (@vsoch) did not create the experiment, but it looks like images (and their corresponding id lookup) are stored under images and then listed in js/list.js

Build an Experiment Container

These instructions are also available here.

First, create a working directory

mkdir -p /tmp/reward-learning
cd /tmp/reward-learning

Then see experiments available

docker run quay.io/vanessa/expfactory-builder list

Generate a container with the reward-learning-task

docker run -v $PWD:/data quay.io/vanessa/expfactory-builder build reward-learning-task

The message will tell you the next step - to build your container! And actually, you would be best off (if you want to share or reproduce this) to add the Dockerfile to a GitHub repository and then have an automated build.

Expfactory Version: 3.16
LOG Recipe written to /data/Dockerfile

To build, cd to directory with Dockerfile and:
              docker build --no-cache -t expfactory/experiments .

The /data folder in the container is where you just bound the present working directory, so our Dockerfile and entrypoint script are actually right here!

$ ls
Dockerfile  startscript.sh

We could build that as follows:

docker build -t reward-learning .

And then run it on port 80:

mkdir -p /tmp/reward-learning/data
docker run -v /tmp/reward-learning/data/:/scif/data -p 80:80 reward-learning start 

And then open your browser to 127.0.0.1 to see the interface!

images/experiment.png

By default, the data is saved to the filesystem where you mounted the local data folder.

And of course see the documentation pages for how to customize the database, and other configuration. If you need to customize the experiment repository cloned from (e.g., if you want to make changes) you can edit this in the Dockerfile:

LABEL EXPERIMENT_reward-learning-task /scif/apps/reward-learning-task
WORKDIR /scif/apps
RUN expfactory install https://www.github.com/<username>/reward-learning-task

About

A reward learning task for the Experiment factory.

Resources

License

Releases

No releases published

Packages

No packages published
You can’t perform that action at this time.