Learner

Basic Neural Network as a service, written in Crystal.

Installation

Requires Crystal 0.27.0.

git clone [email protected]:suruja/learner.git
cd learner
shards install

Usage

Start

crystal run src/learner.cr

API

Every value must be castable to Float64. Either input, output or category must be castable to a vector, aka Array(Float64). Pass your query parameters in JSON format.

CRUD your engine

To create, replace or append your engine training data, you can either upload a CSV file (using ; as separator) file containing one row per data item, of INPUT_SIZE columns, resulting output in the last OUTPUT_SIZE columns, or pass your JSON-encoded data in data query parameter.

Create your engine

POST /:engine_id?input_size=INPUT_SIZE&output_size=OUTPUT_SIZE

Once your data (CSV file or data query parameter) is successfully processed, you will get a token you must keep.

Append your engine training data

PATCH /:engine_id?input_size=INPUT_SIZE&output_size=OUTPUT_SIZE&token=TOKEN

Uploading CSV file or pass data query parameter. You must provide the creation TOKEN as query parameter.

Replace your engine training data

PUT /:engine_id?input_size=INPUT_SIZE&output_size=OUTPUT_SIZE&token=TOKEN

Uploading CSV file or pass data query parameter. You must provide the creation TOKEN as query parameter.

Destroy your engine

DELETE /:engine_id?token=TOKEN

You must provide the creation TOKEN as query parameter.

Consume your engine

Run

GET /:engine_id/run?value=VALUE

Run your engine with your VALUE.

Classify

GET /:engine_id/classify?value=VALUE

Classify your VALUE among the categories found in your data.

Specs

KEMAL_ENV=test crystal spec

Contributing

1. Fork it (https://github.com/suruja/learner/fork)
2. Create your feature branch (git checkout -b my-new-feature)
3. Commit your changes (git commit -am 'Add some feature')
4. Push to the branch (git push origin my-new-feature)
5. Create a new Pull Request

Contributors

• suruja - creator and maintainer