README.md

Training is offline

1. for intent_classifier training python intent_classifier.py

   training is done using hindiR

2. for ner training python train.py

3. all data files are in their corresponding folders hindi - native hindi data hindiR - hindi WX data (converted using wxc converter)

Prediction is done using api

1. tornado server is used (will run on 8000 port by default)
2. python BotServer.py will start the server

there are total 4 service

1. ner (will do ner tagging) curl -X POST
   http://localhost:8000/ner/predict
   -H 'Cache-Control: no-cache'
   -H 'Postman-Token: 2b51822b-aa2e-b503-4ebf-e6a59e7c7f7a'
   -H 'content-type: application/json'
   -d '{ "queries":["मैं इंदिरानगर से यशवंतपुर रेलवे स्टेशन जाना चाहता हूं"] } '

result - { "data": { "output": [ { "Person": "मैं" }, { "Loc": "इंदिरानगर" }, { "Loc": "यशवंतपुर रेलवे स्टेशन" } ] } }

2. intent classification (smalltalk/cab_book) curl -X POST
   http://localhost:8000/intent/predict
   -H 'Cache-Control: no-cache'
   -H 'Postman-Token: 32549c36-614c-5303-16a6-fddd523b7441'
   -H 'content-type: application/json'
   -d '{ "queries":["मैं इंदिरानगर से यशवंतपुर रेलवे स्टेशन जाना चाहता हूं"] } ' result - { "data": { "output": "cab_book" } }

3. smalltalk response curl -X POST
   http://localhost:8000/smalltalk/predict
   -H 'Cache-Control: no-cache'
   -H 'Postman-Token: d86247fa-3ad2-5d85-4012-e9f7305dd827'
   -H 'content-type: application/json'
   -d '{ "queries":["आप कैसे हैं?"] } ' result - { "data": { "output": "मैं ठीक हूं, आपके क्या हाल हैं?" } }

4. all (it will run evrything in pipeline) curl -X POST
   http://localhost:8000/smalltalk/predict
   -H 'Cache-Control: no-cache'
   -H 'Postman-Token: d86247fa-3ad2-5d85-4012-e9f7305dd827'
   -H 'content-type: application/json'
   -d '{ "queries":["आप कैसे हैं?"] } ' result - { "data": { "output": "मैं ठीक हूं, आपके क्या हाल हैं?" } }