Merge pull request #418 from hellohaptik/removed_unused_crf_code

Removed unused crf code

.gitignore

@@ -5,9 +5,8 @@ __pycache__/
 
 # haptik ner config file
 config
-model_config
 
-# C extensions
+# C Extensions
 *.so
 
 # Distribution / packaging

chatbot_ner/config.py

@@ -3,13 +3,10 @@
 import logging.handlers
 import os
 
-import dotenv
 from elasticsearch import RequestsHttpConnection
 from requests_aws4auth import AWS4Auth
 
 BASE_DIR = os.path.dirname(os.path.dirname(__file__))
-MODEL_CONFIG_PATH = os.path.join(BASE_DIR, 'model_config')
-
 LOG_PATH = os.path.join(BASE_DIR, 'logs')
 
 # TODO: Set this up via Django LOGGING
@@ -116,34 +113,7 @@
     ner_logger.warning('`ES_AWS_SERVICE` and `ES_AWS_REGION` are not set. '
                        'This is not a problem if you are using self hosted ES')
 
-# TODO: Remove non functional crf code and cleanup
-# Model Vars
-# Crf Model Specific (Mandatory to use CRF Model)
-CRF_MODELS_PATH = os.environ.get('MODELS_PATH')
-CRF_EMBEDDINGS_PATH_VOCAB = os.environ.get('EMBEDDINGS_PATH_VOCAB')
-CRF_EMBEDDINGS_PATH_VECTORS = os.environ.get('EMBEDDINGS_PATH_VECTORS')
-
-if os.path.exists(MODEL_CONFIG_PATH):
-    dotenv.read_dotenv(MODEL_CONFIG_PATH)
-else:
-    ner_logger.warning('Warning: no file named "model_config" found at %s. This is not a problem if you '
-                       'dont want to run NER with ML models', MODEL_CONFIG_PATH)
-
-CITY_MODEL_TYPE = os.environ.get('CITY_MODEL_TYPE')
-CITY_MODEL_PATH = os.environ.get('CITY_MODEL_PATH')
-DATE_MODEL_TYPE = os.environ.get('DATE_MODEL_TYPE')
-DATE_MODEL_PATH = os.environ.get('DATE_MODEL_PATH')
-if not CITY_MODEL_PATH:
-    CITY_MODEL_PATH = os.path.join(BASE_DIR, 'data', 'models', 'crf', 'city', 'model_13062017.crf')
-if not DATE_MODEL_PATH:
-    DATE_MODEL_PATH = os.path.join(BASE_DIR, 'data', 'models', 'crf', 'date', 'model_date.crf')
-
-# Crf Model Specific with additional AWS storage (optional)
-CRF_MODEL_S3_BUCKET_NAME = os.environ.get('CRF_MODEL_S3_BUCKET_NAME')
-CRF_MODEL_S3_BUCKET_REGION = os.environ.get('CRF_MODEL_S3_BUCKET_REGION')
-WORD_EMBEDDING_REMOTE_URL = os.environ.get('WORD_EMBEDDING_REMOTE_URL')
 GOOGLE_TRANSLATE_API_KEY = os.environ.get('GOOGLE_TRANSLATE_API_KEY')
-
 if not GOOGLE_TRANSLATE_API_KEY:
     ner_logger.warning('Google Translate API key is null or not set')
     GOOGLE_TRANSLATE_API_KEY = ''

chatbot_ner/settings.py

@@ -48,7 +48,6 @@
     'datastore',
     'ner_v1',
     'ner_v2',
-    'models',
     'django_nose'
 ]
 
@@ -137,7 +136,7 @@ def __getitem__(self, item):
     '--ignore-files=constants.py',
     '--ignore-files=run_postman_tests.py',
     '--cover-erase',
-    '--cover-package=datastore,external_api,language_utilities,lib,models,ner_v1,ner_v2',
+    '--cover-package=datastore,external_api,language_utilities,lib,ner_v1,ner_v2',
     '--cover-inclusive',
 ]
 

chatbot_ner/urls.py

@@ -54,9 +54,6 @@
     url(r'^entities/get_crf_training_data', external_api.get_crf_training_data),
     url(r'^entities/update_crf_training_data', external_api.update_crf_training_data),
 
-    # Deprecated train crf model
-    url(r'^entities/train_crf_model', external_api.train_crf_model),
-
     url(r'^entities/languages/v1/(?P<entity_name>.+)$', external_api.entity_language_view),
     url(r'^entities/data/v1/(?P<entity_name>.+)$', external_api.entity_data_view),
 

config.example

@@ -53,14 +53,3 @@ DESTINATION_PORT=
 # In order to enable entity detection for multiple languages, we use google translate. Please enter the key(optional)
 GOOGLE_TRANSLATE_API_KEY=
 
-# Deprecated CRF models configuration
-MODELS_PATH=
-WORD_EMBEDDING_REMOTE_URL=
-EMBEDDINGS_PATH_VECTORS=
-EMBEDDINGS_PATH_VOCAB=
-CITY_MODEL_PATH=
-CITY_MODEL_TYPE=crf
-CRF_MODEL_S3_BUCKET_NAME=
-CRF_MODEL_S3_BUCKET_REGION=
-DATE_MODEL_PATH=
-DATE_MODEL_TYPE=

docker/Dockerfile-python3

@@ -1,5 +1,3 @@
-# This is to automated chatbot_ner installation
-
 FROM python:3.6.10
 
 RUN apt-get update && apt-get install -y wget build-essential curl nginx supervisor
@@ -15,11 +13,9 @@ COPY docker/install.sh nltk_setup.py datastore_setup.py /app/
 COPY docker/supervisord.conf /etc/supervisor/conf.d/supervisord.conf
 
 # cython is installed because pandas build fails otherwise
-RUN mkdir -p ~/model_lib && \
-    mkdir -p /root/models && \
-    /app/install.sh && \
+RUN /app/install.sh && \
     touch /app/config && \
-    touch /app/model_config && \
+    pip install -U pip && \
     pip install --no-cache-dir -I uwsgi && \
     pip install cython
 

docker/install.sh

@@ -1,21 +1,5 @@
 #!/bin/bash
-mkdir -p ~/model_lib
-cd /tmp/
-wget ftp://ftp.netbsd.org/pub/pkgsrc/distfiles/CRF++-0.58.tar.gz
-tar -xzf CRF++-0.58.tar.gz -C ~/model_lib/
-cd ~/model_lib/CRF++-0.58/
-./configure
-make
-make install
-
-echo "export LD_LIBRARY_PATH=/usr/local/lib" >> ~/.bashrc
-
-cd python
-python setup.py build
-python setup.py install
-
 # Get sample Nginx file for routing and GUI
-
 cd /tmp
 wget "https://s3-us-west-2.amazonaws.com/chatbotner/chatbot_ner_nginx/default"
 bash -c "cat /tmp/default > /etc/nginx/sites-available/default"

external_api/api.py

@@ -11,11 +11,9 @@
     FetchIndexForAliasException, DeleteIndexFromAliasException
 from chatbot_ner.config import ner_logger
 from external_api.constants import ENTITY_DATA, ENTITY_NAME, LANGUAGE_SCRIPT, ENTITY_LIST, \
-    EXTERNAL_API_DATA, SENTENCE_LIST, READ_MODEL_FROM_S3, ES_CONFIG, READ_EMBEDDINGS_FROM_REMOTE_URL, \
-    LIVE_CRF_MODEL_PATH, SENTENCES, LANGUAGES
+    EXTERNAL_API_DATA, SENTENCES, LANGUAGES
 
 from django.views.decorators.csrf import csrf_exempt
-from models.crf_v2.crf_train import CrfTrain
 
 from external_api.lib import dictionary_utils
 from external_api.response_utils import external_api_response_wrapper
@@ -210,60 +208,6 @@ def update_crf_training_data(request):
     return HttpResponse(json.dumps(response), content_type='application/json', status=200)
 
 
-@csrf_exempt
-def train_crf_model(request):
-    """
-    This method is used to train crf model.
-    Args:
-        request (HttpResponse): HTTP response from url
-    Returns:
-        HttpResponse : HttpResponse with appropriate status and error message.
-    Post Request Body:
-    key: "external_api_data"
-    value: {
-    "entity_name": "crf_test",
-    "read_model_from_s3": true,
-    "es_config": true,
-    "read_embeddings_from_remote_url": true
-    }
-    """
-    response = {"success": False, "error": "", "result": {}}
-    try:
-        external_api_data = json.loads(request.POST.get(EXTERNAL_API_DATA))
-        entity_name = external_api_data.get(ENTITY_NAME)
-        read_model_from_s3 = external_api_data.get(READ_MODEL_FROM_S3)
-        es_config = external_api_data.get(ES_CONFIG)
-        read_embeddings_from_remote_url = external_api_data.get(READ_EMBEDDINGS_FROM_REMOTE_URL)
-        crf_model = CrfTrain(entity_name=entity_name,
-                             read_model_from_s3=read_model_from_s3,
-                             read_embeddings_from_remote_url=read_embeddings_from_remote_url)
-
-        if es_config:
-            model_path = crf_model.train_model_from_es_data()
-        else:
-            sentence_list = external_api_data.get(SENTENCE_LIST)
-            entity_list = external_api_data.get(ENTITY_LIST)
-            model_path = crf_model.train_crf_model_from_list(sentence_list=sentence_list, entity_list=entity_list)
-
-        response['result'] = {LIVE_CRF_MODEL_PATH: model_path}
-        response['success'] = True
-
-    except (IndexNotFoundException, InvalidESURLException,
-            SourceDestinationSimilarException, InternalBackupException, AliasNotFoundException,
-            PointIndexToAliasException, FetchIndexForAliasException, DeleteIndexFromAliasException,
-            AliasForTransferException, IndexForTransferException, NonESEngineTransferException) as error_message:
-        response['error'] = str(error_message)
-        ner_logger.exception('Error: %s' % error_message)
-        return HttpResponse(json.dumps(response), content_type='application/json', status=500)
-
-    except Exception as e:
-        response['error'] = str(e)
-        ner_logger.exception('Error: %s' % e)
-        return HttpResponse(json.dumps(response), content_type='application/json', status=500)
-
-    return HttpResponse(json.dumps(response), content_type='application/json', status=200)
-
-
 @csrf_exempt
 @external_api_response_wrapper
 def entity_language_view(request, entity_name):

external_api/constants.py

@@ -10,9 +10,4 @@
 LANGUAGE_SCRIPT = 'language_script'
 ENTITY_LIST = 'entity_list'
 SENTENCE_LIST = 'sentence_list'
-
-
-READ_MODEL_FROM_S3 = 'read_model_from_s3'
 ES_CONFIG = 'es_config'
-READ_EMBEDDINGS_FROM_REMOTE_URL = 'read_embeddings_from_remote_url'
-LIVE_CRF_MODEL_PATH = 'live_crf_model_path'