Skip to content

Latest commit

 

History

History
124 lines (92 loc) · 3.79 KB

README.md

File metadata and controls

124 lines (92 loc) · 3.79 KB
Raw

Welcome to Dialog Flow SDK

This is an early alpha version of the DD-IDDE SDK. It is used in combination with the DeepPavlov's DD-IDDE available here.

Requirements

pip install -r requirements.txt
python -m spacy download en_core_web_sm

# install df_engine
pip install df_engine
# install dashboard for stats
pip install dff-node-stats[dashboard]

Environment

Item Requirements Comments
OS Debian-based distribution, e.g., Ubuntu or Windows 10 This version was tested on Ubuntu 18.04 under WSL2 on Windows 11 and Windows 10.
Python v3.9+ This version was tested on OS with Python 3.9.
Docker v20+ This version was tested with Docker v20.10.7 (64-bit).
Docker-Compose v1.29.2 This version was tested with Docker-Compose v1.29.2.

VS Code

Required Extensions

  • DD-IDDE
  • Python
  • Docker

Optional Extensions

  • Remote - WSL

Set WSL-based Terminal As Default One

If needed, set your WSL-based terminal app as the default one in your VS Code by following these instructions.

Python 3.9 - set as default (optional)

  1. Install the python3.9 package using apt-get

sudo apt-get install python3.9

Add Python3.6 & Python 3.9 to update-alternatives

sudo update-alternatives --install /usr/bin/python3 python3 /usr/bin/python3.6 1 sudo update-alternatives --install /usr/bin/python3 python3 /usr/bin/python3.9 2

Update Python 3 to point to Python 3.9:

sudo update-alternatives --config python3

Enter 2 for Python 3.9

Test the version of python:

python3 --version Python 3.9

Test the version of python used by pip3 command:

pip3 --version pip 21.3.1 from /home/danielko/.local/lib/python3.9/site-packages/pip (python 3.9)

Prerequisites

pip3 install lxml

Installation Process

Runtime

We use Dialog Flow Engine as the runtime for the open-domain/scenario-driven chatbots.

Follow these instructions to install Dialog Flow Engine:

# install df_engine
pip install df_engine
# install dashboard for stats
pip install dff-node-stats[dashboard]

Follow these requirements to prepare DD-IDDE SDK to run on your machine:

pip install -r requirements.txt

Discourse Moves Recommendation System

We use our Speech Functions Classifier & Predictor from our larger DeepPavlov Dream Multiskill AI Assistant Platform.

Follow these instructions to run the Discourse Moves Recommendation System:

docker-compose up -d --build

After that sf predictor is available on localhost:8107/annotation and sf classifier is availible on localhost:8108/annotation

Usage

DD-IDDE as Designer

Go to your local clone of this repo and run:

code .

This will ensure that your VS Code will run from this folder, and will (in case you use WSL) run through WSL.

Dashboard

Create ssh tunnel:

ssh -L 8501:localhost:8501 $HOST

Collect stats for food topic:

python examples/food.py 
Collect stats for artificial dialog:

python examples/stats_collection.py

After that run dashboard:

streamlit run examples/stats_dashboard.py


### Discourse Moves Recommendation System
TBD

## Generic Responses
TBD
## Entity Detection
TBD