Skip to content

readme.md

Latest commit

 

History

History
103 lines (93 loc) · 21.5 KB

readme.md

File metadata and controls

103 lines (93 loc) · 21.5 KB

NLU Server

This package contains the Botpress Standalone NLU server.

Documentation

 Available Routes
 
 /**
  * Gets the current version of the NLU engine being used. Usefull to test if your installation is working.
  * @returns info: version, health and supported languages.
 */
 GET http://HOST:PORT/v1/info
 
 /**
  * Starts a training.
  * @header x-app-id Application ID to make sure there's no collision between models of different applications.
  * @body_parameter language Language to use for training.
  * @body_parameter intents Intents definitions.
  * @body_parameter contexts All available contexts.
  * @body_parameter entities Entities definitions.
  * @body_parameter seed Number to seed random number generators used during training (beta feature). ** Optionnal **
  * @returns modelId A model id for futur API calls
  */
 POST http://HOST:PORT/v1/train
 
 /**
  * Gets a training progress status.
  * @header x-app-id Application ID to make sure there's no collision between models of different applications.
  * @path_parameter modelId The model id for which you seek the training progress.
  * @returns session A training state data structure with information on desired model.
  */
 GET http://HOST:PORT/v1/train/:modelId
 
 /**
  * List all models for a given app Id and secret.
  * @header x-app-id Application ID to make sure there's no collision between models of different applications.
  * @path_parameter modelId The model id for which you seek the training progress.
  * @returns models Array of strings model ids available for prediction.
  */
 GET http://HOST:PORT/v1/models/:modelId
 
 /**
  * Cancels a training.
  * @header x-app-id Application ID to make sure there's no collision between models of different applications.
  * @path_parameter modelId The model id for which you want to cancel the training.
  * @returns models Array of strings model ids that where pruned.
  */
 POST http://HOST:PORT/v1/models/prune
 
 /**
  * Cancels a training.
  * @header x-app-id Application ID to make sure there's no collision between models of different applications.
  * @path_parameter modelId The model id for which you want to cancel the training.
  */
 POST http://HOST:PORT/v1/train/:modelId/cancel
 
 /**
  * Perform prediction for a text input.
  * @header x-app-id Application ID to make sure there's no collision between models of different applications.
  * @path_parameter modelId The model id you want to use for prediction.
  * @body_parameter utterances Array of text for which you want a prediction.
  * @returns predictions Array of predictions; Each prediction is a data structure reprensenting the understanding of the text.
  */
 POST http://HOST:PORT/v1/predict/:modelId
 
 /**
  * Perform prediction for a text input.
  * @header x-app-id Application ID to make sure there's no collision between models of different applications.
  * @path_parameter modelId The model id you want to use for prediction.
  * @body_parameter utterances Array of text for which you want a prediction.
  * @body_parameter models Array of strings model ids you want to use to detect language. ** Optionnal **
  * @returns detectedLanguages Array of string language codes.
  */
 POST http://HOST:PORT/v1/detect-lang
 
 For more detailed information on typings, see
 https://github.com/botpress/nlu/blob/master/packages/nlu-client/src/typings/sdk.d.ts.
 
 For a complete example on training input, see
 https://github.com/botpress/nlu/blob/master/packages/nlu-server/examples/train-example.json.
 
 If you plan on querying from a nodejs environment, see
 https://www.npmjs.com/package/@botpress/nlu-client.

To display this documentation, launch the nlu server with argument --doc.

⚠️⚠️ Disclaimer ⚠️⚠️

The NLU Server does not enforce authentication in any way. This means it is completely exposed to many attacks. If you plan on using the nlu-server in your local Botpress setup, makes sure it is not publicly exposed. If you plan on exposing the NLU server, make sure it his hidden behind a reverse proxy which ensures a proper authentication. This reverse proxy should:

  • Ensure each appId (X-App-Id header) is unique.
  • Ensure a user with appId user1 can't call the nlu server with header X-App-Id set to anything other than user1.
  • Ensure only calls with a registered appId can call the nlu server except maybe for the GET /info route.

The NLU Server's only purpose is to do NLU.

Licensing

This software is protected by the same license as the main Botpress repository. You can find the license file here.