Release 2.3.0a

• Added Hydra.
• Small fixes.

Release 2.2.0

• added CORS
• added rating endpoints

2.1

Small improvements

• Pipeline configs can be chained
• New debug chat web interface
• New connector for creating dummy services
• Additional setup possibilities
• Working telegram client

2.0-alpha: Merge pull request #78 from deepmipt/dev

Briefly

• New configuration.
• New launch script.
• Current documentation is outdated.

Config refactoring

• Pipeline config in .json or .yml format.
• Explicit configuration of pipeline graph.
  • Explicit configuration of state manager method, used for service response saving.
  • Explicit configuration of dialog (format dialog for service) and response (format service response in order to save in dialog state).
• Database config in .json or .yml format.

Pipeline refactoring

• Pipeline graph is now fully customizable via a config file.
• It is possible to set a previous_services parameter in the service configuration part. These services should be skipped or responded with or without an error (at least one of them should return correct response) in order to send a request to the service.
  • Example: in a response selector service you can put all services that return hypotheses (potential bot response to user utterance).
• It is possible to set-up a required_previous_services parameter in the service configuration part. These services should return a correct response in order to send request to the service.
  • Example: a service is relied on another service response (like some model requires annotations, provided by another model). In that case required previous services.
• Check for cycles in pipeline graph.

Formatter refactoring

• Formatters are now divided into dialog (send to service) and response (receive from service) formatters.
• Dialog formatter should return a list of samples. Each of the samples represents a single task in a format that matches input structure of the service.

Agent refactoring

• Agent can create multiple tasks from one dialog. It can be useful in case you want to annotate all hypotheses before response selector, or other similar ones.
• Last chance service is introduced. It runs if an agent cannot continue processing an input utterance. For instance after crucial number of services were skipped or responded with error.
• Timeout service is introduced. It runs if deadline_timestamp is presented and processing time tooks more than that timestamp.

Release v1.0

Highload:

• asynchronous Agent
• independent input batchification for each connected service
• RabbitMQ connectors
• horizontal scaling support
• asynchronous Telegram connector

Other features:

• each Utterance has a corresponding Human or Bot object
• Response Selector can not only select a response but also rewrite it
• Skills are able to provide a list of response hypotheses at each human utterance
• output_formatters to change the Agent server output API
• arbitrary Agent server input
• logging of the Agent and each connected service time performance

Release v0.1

Main changes:

• AGENT_ENV_FILE variable for the path to the Agent environment variables
• ResponseSelector can run as a service
• multi-stage annotators
• empty annotators allowed
• save Mongo DB data even if the mongo container is killed
• DeepPavlov 0.6.0 compatibility

Useful utils:

• a script for Mongo DB data dump
• monitoring of current state (static) in a browser window
• a script for testing Agent HTTP API from ready dialogs
• a script for generating dialogs from phrases
• a script for testing the Agent in a batch mode

Release v0.1-beta

• Agent works as a web server in http mode
• Formatter functions support
• Added native support of DeepPavlov configs
• Full integration with the latest DeepPavlov

Release v0.1-alpha

• Running Agent services and state storage in separate Docker containers by a single command
• Conversation with the Agent via console or Telegram
• Out-of-the-box Agent with chitchat and odqa skills for the Russian language
• Out-of-the-box ner, obscenity and sentiment annotators for the Russian language
• GPU support in Docker containers