kcculhwch / prayer2vec

Word2vec and Paragraph2vec Workspace for payer processing.

prayer2vec

Directories

• corpus
  • Repository of texts
• models
  • Storage for models
• normalized
  • Storage for normalized texts
• build
  • Code for the project

Installing and Building

Build Instructions Here

Running the tools

From the root of the project run p2v [queue] OPTIONS

queue is used to submit the job for running on cloud resources

Options

word2vec ARGS to run word2vec command directly

pre corpus/x/x to normalize the text

raw COMMANDS to run raw commands in the env

sync_up to sync local data to cloud archives for processing

sync_down to sync changes on the cloud back to the local file system

TODO

Will add more options for particular functions as we need them.

• train /corpus/x/x/ /model/x/x/
• etc

Corpus

Corpus will be bundled in /corpus in the container. It is also mounted over the top of the container so you will not need to rebuild if yout make changes to it.

Models

Models will be bundled in /models in the container. It is also mounted over the top of the container so you will n ot need to rebuild if yout make changes to it.

Normalized

Normalized text filed are bundled in /normalized in the container. It is also mounted over the top of the container so you will not need to rebuild if yout make changes to it.