Conversational AI with Rasa: Creating a New Assistant

Поделиться
HTML-код
  • Опубликовано: 11 дек 2024

Комментарии • 16

  • @aswinb9838
    @aswinb9838 2 года назад +4

    I installed rasa but how to activate conda environment in vs code pls guide me in this. I can activate conda environment in anaconda prompt but i cant in vs code.

    • @rimabenrejeb5152
      @rimabenrejeb5152 2 года назад

      in vscode:
      shift + cmd + P
      Search Select Interpreter
      Select it and it will show you the list of your virtual environment created via conda and other python versions
      select the environment and you are ready to go.

  • @RobertoRuiz-bw8ln
    @RobertoRuiz-bw8ln 2 года назад

    From which link can i download the project?

  • @syedshahab8471
    @syedshahab8471 3 года назад +2

    Awesome!

  • @gamez447
    @gamez447 2 года назад

    How to reduce rasa voice playback speed

  • @rimabenrejeb5152
    @rimabenrejeb5152 2 года назад

    is Rasa open source 3. x not yet compatible with rasa x ??

    • @RasaHQ
      @RasaHQ  2 года назад

      Not the current version of Rasa X, but it will be with Rasa X 1.1, which we're hoping to get out soon.

  • @aldoleal7660
    @aldoleal7660 2 года назад

    Thank you for these videos. Should I have created a moodbot project before this step in VS?

    • @RasaHQ
      @RasaHQ  2 года назад

      nope, `rasa init` will do that for you - Rachael

  • @abir95571
    @abir95571 3 года назад

    Can we have a custom pipeline video where we can use our own models to do let's say intent classification.

    • @RasaHQ
      @RasaHQ  3 года назад

      (Vincent here)
      I wouldn't recommend putting your own models in the pipeline. We certainly allow it, but our Research Team spends a lot of time researching algorithms so you don't have to. Especially when you're starting out your primary concern should be getting a meaningful dataset first.
      That said, I maintain a project with custom models that could be used as a starting point, here; github.com/RasaHQ/rasa-nlu-examples

    • @abir95571
      @abir95571 3 года назад

      @@RasaHQ actually I’ve used RASA during its beta phase around 2018 for a banking client using the embedding pipeline and I once used to be an active community member . Our current client has this weird requirement of making the Intent recogniser and Entity classifier as standalone without RASA but for Stories they want RASA . Being a Data Scientist I’ll be able prepare the model but i didn’t find any example on how one can integrate an external model with rasa stories.
      i shall check your given link

  • @youness127
    @youness127 3 года назад

    can we use the default pipeline for language:fr???

    • @RasaHQ
      @RasaHQ  3 года назад +1

      (Vincent here)
      Since French is whitespace tokenizable (at least to a very large extent) I'd argue it's fine to use the default pipeline as a starting point. You can always add embeddings later, but especially when starting out I recommend just using the default pipeline.

    • @youness127
      @youness127 3 года назад

      @@RasaHQ thank you very much

  • @inhthimyngoc7249
    @inhthimyngoc7249 3 года назад

    Rasa voice ..