/Efficient-Inference

Predicts what class an arbitrary name belongs using transfer learning

Primary LanguageJupyter Notebook

NamePredict

This module predicts what class an arbitrary name belongs:

1:'Company',
2:'EducationalInstitution',
3:'Artist',
4:'Athlete',
5:'OfficeHolder',
6:'MeanOfTransportation',
7:'Building',
8:'NaturalPlace',
9:'Village',
10:'Animal',
11:'Plant',
12:'Album',
13:'Film',
14:'WrittenWork'

Install Dependencies

  • conda env create -f NP.yml --name NP
  • pandas should be 0.24.1 and scipy 1.1.0
  • conda install -c conda-forge pandas_ml
  • conda install -c conda-forge matplotlib
  • conda install -c anaconda seaborn

Train

  • Use NamePredict.ipynb

Run predict a name (examples)

  • python run_predict_name.py run --name Jhonson Brothers
  • python run_predict_name.py run --name Coca Cola
  • python run_predict_name.py run --name Hello