LambdaFeedback/ EvaluationFunctionToolkit

A collection of utilities for creating Lambda Feedback evaluation functions for the Wolfram Language

Contributed by: Andreas Pfurtscheller

The EvaluationFunctionToolkit is designed to seamlessly integrate with the Lambda Feedback Platform, enabling efficient and reliable communication with the evaluation function shim. At its core, the Serve function launches a TCP server that allows the shim to interact with any custom evaluation function.

Installation Instructions

To install this paclet in your Wolfram Language environment, evaluate this code:
PacletInstall["LambdaFeedback/EvaluationFunctionToolkit"]


To load the code after installation, evaluate this code:
Needs["LambdaFeedback`EvaluationFunctionToolkit`"]

Details

Ensure the EVAL_RPC_TCP_ADDRESS environment variable is set correctly if you wish to specify a different address than the default.
The server runs indefinitely; use appropriate mechanisms to manage the lifecycle of the server in production environments.
Error handling and logging can be enhanced based on specific requirements.

Examples

Example

Create a simple evaluation function comparing "answer" and "response", and returning the evaluation result:

In[1]:=
evaluation[answer_, response_, params_] :=
 <|
  	"is_correct" -> TrueQ[answer == response]
  |>

Then, use "Serve" to serve the evaluation function:

In[2]:=
Serve[evaluation];

Publisher

Lambda Feedback

Disclosures

Compatibility

Wolfram Language Version 13.

Version History

  • 1.0.0 – 23 August 2024

License Information

MIT License

Paclet Source