sandy4321 / stand_ranking_en

Demo stand. Model: Ranking (English)

Installation and start

1. Clone the repo and cd to project root:

   git clone https://github.com/deepmipt/stand_ranking_en.git
   cd stand_ranking_en
   

2. Create a virtual environment with Python 3.6:

   virtualenv env -p python3.6
   

3. Activate the environment:

   source ./env/bin/activate
   

4. Install DeepPavlov:
   1. Clone the repo and cd to project root

      git clone https://github.com/deepmipt/DeepPavlov.git
      cd DeepPavlov
      

   2. Install the requirements:

      python setup.py develop
      

   3. Install spacy dependencies:

      pip install http://lnsigo.mipt.ru/export/en_core_web_sm-2.0.0.tar.gz
      python3.6 -m spacy link en_core_web_sm en --force
      

   4. Download model components:

      python3.6 -m deeppavlov.deep download deeppavlov/configs/ranking/insurance_config.json
      

   5. Download NLTK data:

   $ python3
   >>> import nltk
   >>> nltk.download('punkt')
   >>> exit()
   

5. Specify model endpoint host (host) and port (port) in common_defaults or corresponding model section of utils/server_utils/server_config.json
6. Return to the stand_ranking_en dir
7. Specify virtual environment path (if necessary) in run_en_ranking.sh
8. Run model:

   ./run_en_ranking.sh
   

Building and running with Docker:

1. If necessary, build base docker_cuda and docker_deeppavlov images from:

   https://github.com/deepmipt/stand_docker_base

2. Clone the repo and cd to project root:

   git clone https://github.com/deepmipt/stand_ranking_en.git
   cd stand_ranking_en
   

3. Build Docker image:

   sudo docker build -t stand/ranking_en .
   

4. Run Docker image:

   sudo docker run -p <host_port>:6009 -v /path/to/host/vol/map/dir:/logs stand/ranking_en