Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

fixed legacy optimizer issue #42

Open
wants to merge 1 commit into
base: master
Choose a base branch
from
Open

fixed legacy optimizer issue #42

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
102 changes: 55 additions & 47 deletions neuralintents/main.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6,7 +6,7 @@
import numpy as np
import os

os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'
os.environ["TF_CPP_MIN_LOG_LEVEL"] = "3"

import nltk
from nltk.stem import WordNetLemmatizer
Expand All @@ -16,34 +16,33 @@
from tensorflow.keras.optimizers import SGD
from tensorflow.keras.models import load_model

nltk.download('punkt', quiet=True)
nltk.download('wordnet', quiet=True)
nltk.download("punkt", quiet=True)
nltk.download("wordnet", quiet=True)

class IAssistant(metaclass=ABCMeta):

class IAssistant(metaclass=ABCMeta):
@abstractmethod
def train_model(self):
""" Implemented in child class """
"""Implemented in child class"""

@abstractmethod
def request_tag(self, message):
""" Implemented in child class """
"""Implemented in child class"""

@abstractmethod
def get_tag_by_id(self, id):
""" Implemented in child class """
"""Implemented in child class"""

@abstractmethod
def request_method(self, message):
""" Implemented in child class """
"""Implemented in child class"""

@abstractmethod
def request(self, message):
""" Implemented in child class """
"""Implemented in child class"""


class GenericAssistant(IAssistant):

def __init__(self, intents, intent_methods={}, model_name="assistant_model"):
self.intents = intents
self.intent_methods = intent_methods
Expand All @@ -58,34 +57,37 @@ def load_json_intents(self, intents):
self.intents = json.loads(open(intents).read())

def train_model(self):

self.words = []
self.classes = []
documents = []
ignore_letters = ['!', '?', ',', '.']
ignore_letters = ["!", "?", ",", "."]

for intent in self.intents['intents']:
for pattern in intent['patterns']:
for intent in self.intents["intents"]:
for pattern in intent["patterns"]:
word = nltk.word_tokenize(pattern)
self.words.extend(word)
documents.append((word, intent['tag']))
if intent['tag'] not in self.classes:
self.classes.append(intent['tag'])

self.words = [self.lemmatizer.lemmatize(w.lower()) for w in self.words if w not in ignore_letters]
documents.append((word, intent["tag"]))
if intent["tag"] not in self.classes:
self.classes.append(intent["tag"])

self.words = [
self.lemmatizer.lemmatize(w.lower())
for w in self.words
if w not in ignore_letters
]
self.words = sorted(list(set(self.words)))

self.classes = sorted(list(set(self.classes)))



training = []
output_empty = [0] * len(self.classes)

for doc in documents:
bag = []
word_patterns = doc[0]
word_patterns = [self.lemmatizer.lemmatize(word.lower()) for word in word_patterns]
word_patterns = [
self.lemmatizer.lemmatize(word.lower()) for word in word_patterns
]
for word in self.words:
bag.append(1) if word in word_patterns else bag.append(0)

Expand All @@ -100,40 +102,46 @@ def train_model(self):
train_y = list(training[:, 1])

self.model = Sequential()
self.model.add(Dense(128, input_shape=(len(train_x[0]),), activation='relu'))
self.model.add(Dense(128, input_shape=(len(train_x[0]),), activation="relu"))
self.model.add(Dropout(0.5))
self.model.add(Dense(64, activation='relu'))
self.model.add(Dense(64, activation="relu"))
self.model.add(Dropout(0.5))
self.model.add(Dense(len(train_y[0]), activation='softmax'))
self.model.add(Dense(len(train_y[0]), activation="softmax"))

sgd = SGD(lr=0.01, decay=1e-6, momentum=0.9, nesterov=True)
self.model.compile(loss='categorical_crossentropy', optimizer=sgd, metrics=['accuracy'])
sgd = SGD(lr=0.01, momentum=0.9, nesterov=True)
self.model.compile(
loss="categorical_crossentropy", optimizer=sgd, metrics=["accuracy"]
)

self.hist = self.model.fit(np.array(train_x), np.array(train_y), epochs=200, batch_size=5, verbose=1)
self.hist = self.model.fit(
np.array(train_x), np.array(train_y), epochs=200, batch_size=5, verbose=1
)

def save_model(self, model_name=None):
if model_name is None:
self.model.save(f"{self.model_name}.h5", self.hist)
pickle.dump(self.words, open(f'{self.model_name}_words.pkl', 'wb'))
pickle.dump(self.classes, open(f'{self.model_name}_classes.pkl', 'wb'))
pickle.dump(self.words, open(f"{self.model_name}_words.pkl", "wb"))
pickle.dump(self.classes, open(f"{self.model_name}_classes.pkl", "wb"))
else:
self.model.save(f"{model_name}.h5", self.hist)
pickle.dump(self.words, open(f'{model_name}_words.pkl', 'wb'))
pickle.dump(self.classes, open(f'{model_name}_classes.pkl', 'wb'))
pickle.dump(self.words, open(f"{model_name}_words.pkl", "wb"))
pickle.dump(self.classes, open(f"{model_name}_classes.pkl", "wb"))

def load_model(self, model_name=None):
if model_name is None:
self.words = pickle.load(open(f'{self.model_name}_words.pkl', 'rb'))
self.classes = pickle.load(open(f'{self.model_name}_classes.pkl', 'rb'))
self.model = load_model(f'{self.model_name}.h5')
self.words = pickle.load(open(f"{self.model_name}_words.pkl", "rb"))
self.classes = pickle.load(open(f"{self.model_name}_classes.pkl", "rb"))
self.model = load_model(f"{self.model_name}.h5")
else:
self.words = pickle.load(open(f'{model_name}_words.pkl', 'rb'))
self.classes = pickle.load(open(f'{model_name}_classes.pkl', 'rb'))
self.model = load_model(f'{model_name}.h5')
self.words = pickle.load(open(f"{model_name}_words.pkl", "rb"))
self.classes = pickle.load(open(f"{model_name}_classes.pkl", "rb"))
self.model = load_model(f"{model_name}.h5")

def _clean_up_sentence(self, sentence):
sentence_words = nltk.word_tokenize(sentence)
sentence_words = [self.lemmatizer.lemmatize(word.lower()) for word in sentence_words]
sentence_words = [
self.lemmatizer.lemmatize(word.lower()) for word in sentence_words
]
return sentence_words

def _bag_of_words(self, sentence, words):
Expand All @@ -154,16 +162,16 @@ def _predict_class(self, sentence):
results.sort(key=lambda x: x[1], reverse=True)
return_list = []
for r in results:
return_list.append({'intent': self.classes[r[0]], 'probability': str(r[1])})
return_list.append({"intent": self.classes[r[0]], "probability": str(r[1])})
return return_list

def _get_response(self, ints, intents_json):
try:
tag = ints[0]['intent']
list_of_intents = intents_json['intents']
tag = ints[0]["intent"]
list_of_intents = intents_json["intents"]
for i in list_of_intents:
if i['tag'] == tag:
result = random.choice(i['responses'])
if i["tag"] == tag:
result = random.choice(i["responses"])
break
except IndexError:
result = "I don't understand!"
Expand All @@ -181,7 +189,7 @@ def request_method(self, message):
def request(self, message):
ints = self._predict_class(message)

if ints[0]['intent'] in self.intent_methods.keys():
self.intent_methods[ints[0]['intent']]()
if ints[0]["intent"] in self.intent_methods.keys():
self.intent_methods[ints[0]["intent"]]()
else:
return self._get_response(ints, self.intents)
return self._get_response(ints, self.intents)