Git Product home page Git Product logo

modelblocks-release's Introduction

README for ModelBlocks

This is the Modelblocks software package. It includes several resources for constructing and evaluating broad-coverage probabilistic models of cognitive processes, organized into projects centered around different tasks and data sets.

Documentation is available at: https://www.asc.ohio-state.edu/schuler.77/overview-mb.pdf.

Quickstart

To quickly get started using modelblocks, read through the quickstart guide in the neighboring QUICKSTART.md file. Below is more info about the structure of Modelblocks.

Use of Makefiles

Each project exists in a subdirectory of the main modelblocks directory. In order to ensure the reproducibility of experiments conducted using this resource, these project directories each contain a Makefile, which specifies how data sets, output files, and evaluation results are constructed. The repository contains several resource-XXX repositories that contain reusable libraries of code, project repositories that contain recipes for reproducing published experimental results, and a workspace for experimentation and development. Recipes will fail unless all dependencies to external resources (text corpora, experimental data, external code libraries, etc.) are satisfied. If you attempt to make a recipe that has a missing dependency, Make will exit with an error message about which dependency is missing and how you can access it.

For sandboxing and development, nearly all ModelBlocks recipes can be created from a single workspace. To initialize your workspace, simply type make at the repository root, then navigate to the workspace directory. To reproduce a published experiment, navigate to the relevant experiment directory (named by author/year) and type make. NOTE: We do not guarantee indefinite future support of all published results recipes. In some cases it may be necessary to revert the repository to some previous state in order to reproduce a result. If you are encountering errors as you try to reproduce a result, please contact the ModelBlocks development team.

Included Resources

Modelblocks makes use of several third-party data and software resources. Where licenses permit, these have been included directly in the modelblocks package, so as to avoid version compatibility issues and thereby ensure reproducibility. In some cases open-source software has been modified so as to produce a common data file format required by other software. All resources included in this package are distributed under the Gnu General Public License (see LICENSE file in this directory).

External Resources and USER-*.TXT Files

When licenses of resources used in Modelblocks projects do not permit redistribution, or when (usually data) resources are too large to be included, the Makefile will generate an appropriately-named user-*.txt configuration file in modelblocks-release/config/, in which a user may specify a path to an external copy of the resource. When Make is first invoked, ModelBlocks will create an incorrect pointer for each configuration file in the dependency chain, along with console output indicating which configuration files are needed to create the recipe. Before re-running Make, the needed third-party resources will need to be downloaded and the pointers updated in modelblocks-release/config. This use of user-*.txt files is intended to allow users to specify external resources or other user- specific data without having to modify the Makefile, which may be overwritten in subsequent updates to ModelBlocks.

External resources are described in RESOURCES.md. RESOURCES.md also specifies which user-*.txt files are associated with each external resource.

modelblocks-release's People

Contributors

coryshain avatar jaffe59 avatar byungdoh avatar vansky avatar christian-clark avatar dmhowcroft avatar manjuand avatar hongao0611 avatar nicalin avatar wschuler avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.