def train(self, conversation):
"""
Train the chat bot based on the provided list of
statements that represents a single conversation.
"""
previous_statement_text = None
previous_statement_search_text = ''
statements_to_create = []
for conversation_count, text in enumerate(conversation):
if self.show_training_progress:
utils.print_progress_bar('List Trainer',
conversation_count + 1,
len(conversation))
statement_search_text = self.chatbot.storage.tagger.get_text_index_string(
text)
statement = self.get_preprocessed_statement(
Statement(text=text,
search_text=statement_search_text,
in_response_to=previous_statement_text,
search_in_response_to=previous_statement_search_text,
conversation='training'))
previous_statement_text = statement.text
previous_statement_search_text = statement_search_text
statements_to_create.append(statement)
self.chatbot.storage.create_many(statements_to_create)
def train(self, conversation):
"""
Train the chat bot based on the provided list of
statements that represents a single conversation.
"""
previous_statement_text = None
for conversation_count, text in enumerate(conversation):
if self.show_training_progress:
utils.print_progress_bar('List Trainer',
conversation_count + 1,
len(conversation))
statement = self.get_preprocessed_statement(
Statement(text=text,
in_response_to=previous_statement_text,
conversation='training'))
previous_statement_text = statement.text
self.chatbot.storage.create(
text=statement.text,
in_response_to=statement.in_response_to,
conversation=statement.conversation,
tags=statement.tags)
def train(self, conversation):
"""
Train the chat bot based on the provided list of
statements that represents a single conversation.
"""
previous_statement_text = None
previous_statement_search_text = ''
statements_to_create = []
for conversation_count, text in enumerate(conversation):
if self.show_training_progress:
utils.print_progress_bar(
'List Trainer',
conversation_count + 1, len(conversation)
)
statement_search_text = self.stemmer.get_bigram_pair_string(text)
statement = self.get_preprocessed_statement(
Statement(
text=text,
search_text=statement_search_text,
in_response_to=previous_statement_text,
search_in_response_to=previous_statement_search_text,
conversation='training'
)
)
previous_statement_text = statement.text
previous_statement_search_text = statement_search_text
statements_to_create.append(statement)
self.chatbot.storage.create_many(statements_to_create)
def train(self, *corpus_paths):
#corpus_paths is a list of paths of corpus or one corpus path.
for corpus in corpus_paths:
print(corpus)
file = open(corpus)
l_line = []
l_line = file.readlines()
N = len(l_line)
for i in range(N):
l_line[i] = l_line[i].strip()
file.close()
previous_statement_text = None
for conversation_count, text in enumerate(l_line):
print_progress_bar("Subtitle Trainer", conversation_count + 1,
N)
statement = self.get_or_create(text)
if previous_statement_text:
statement.add_response(Response(previous_statement_text))
previous_statement_text = statement.text
self.storage.update(statement)
def train(self, NN=0, N=None, *corpus_paths):
#corpus_paths is a list of paths of corpus or one corpus path.
#read each corpus from line NN to N. By default the whole file is read.
print('N=', N)
for corpus in corpus_paths:
print(corpus)
file = open(corpus)
l_line = []
if N:
for i in range(NN, N):
line = file.readline().strip()
l_line.append(line)
else:
l_line = file.readlines()
N = len(l_line)
for i in range(N):
l_line[i] = l_line[i].strip()
file.close()
previous_statement_text = None
for conversation_count, text in enumerate(l_line):
print_progress_bar("Subtitle Trainer", conversation_count + 1,
N - NN)
statement = self.get_or_create(text)
if previous_statement_text:
statement.add_response(Response(previous_statement_text))
previous_statement_text = statement.text
self.storage.update(statement)
def train(self, *corpus_paths):
# Allow a list of corpora to be passed instead of arguments
if len(corpus_paths) == 1:
if isinstance(corpus_paths[0], list):
corpus_paths = corpus_paths[0]
# Train the chat bot with each statement and response pair
for corpus_path in corpus_paths:
corpora = self.corpus.load_corpus(corpus_path)
corpus_files = self.corpus.list_corpus_files(corpus_path)
for corpus_count, corpus in enumerate(corpora):
for conversation_count, conversation in enumerate(corpus):
if self.show_training_progress:
utils.print_progress_bar(
str(os.path.basename(corpus_files[corpus_count])) +
' Training', conversation_count + 1, len(corpus))
previous_statement_text = None
for text in conversation:
statement = self.get_or_create(text)
statement.add_tags(corpus.categories)
if previous_statement_text:
statement.add_response(
Response(previous_statement_text))
previous_statement_text = statement.text
self.chatbot.storage.update(statement)
def train(self, *corpus_paths):
from chatterbot.corpus import load_corpus, list_corpus_files
data_file_paths = []
# Get the paths to each file the bot will be trained with
for corpus_path in corpus_paths:
data_file_paths.extend(list_corpus_files(corpus_path))
for corpus, categories, file_path in load_corpus(*data_file_paths):
statements_to_create = []
# Train the chat bot with each statement and response pair
for conversation_count, conversation in enumerate(corpus):
if self.show_training_progress:
utils.print_progress_bar(
'Training ' + str(os.path.basename(file_path)),
conversation_count + 1, len(corpus))
previous_statement_text = None
previous_statement_search_text = ''
for text in conversation:
statement_search_text = self.stemmer.stem(text)
_statement = Statement(
text=text,
search_text=statement_search_text,
in_response_to=previous_statement_text,
search_in_response_to=previous_statement_search_text,
conversation='training')
_statement.add_tags(*categories)
statement = self.get_preprocessed_statement(_statement)
previous_statement_text = statement.text
previous_statement_search_text = statement_search_text
statements_to_create.append({
'text': statement.text,
'in_response_to': statement.in_response_to,
'conversation': statement.conversation,
'tags': statement.tags
})
self.chatbot.storage.create_many(statements_to_create)
def train(self, *corpus_paths):
from chatterbot.corpus import load_corpus, list_corpus_files
data_file_paths = []
# Get the paths to each file the bot will be trained with
for corpus_path in corpus_paths:
data_file_paths.extend(list_corpus_files(corpus_path))
for corpus, categories, file_path in load_corpus(*data_file_paths):
statements_to_create = []
# Train the chat bot with each statement and response pair
for conversation_count, conversation in enumerate(corpus):
if self.show_training_progress:
utils.print_progress_bar(
'Training ' + str(os.path.basename(file_path)),
conversation_count + 1,
len(corpus)
)
previous_statement_text = None
previous_statement_search_text = ''
for text in conversation:
statement_search_text = self.stemmer.get_bigram_pair_string(text)
statement = Statement(
text=text,
search_text=statement_search_text,
in_response_to=previous_statement_text,
search_in_response_to=previous_statement_search_text,
conversation='training'
)
statement.add_tags(*categories)
statement = self.get_preprocessed_statement(statement)
previous_statement_text = statement.text
previous_statement_search_text = statement_search_text
statements_to_create.append(statement)
self.chatbot.storage.create_many(statements_to_create)
def train(self, data, mimic_name):
"""
Train the chat bot based on the provided list of
statements that represents a single conversation.
"""
if not 'participants' in data:
return
m = self._merge_messages(data)
print('Total messages: ' + str(len(m)))
def clean(txt):
return txt.lower().replace('\'', '')
statements = []
for i in range(1, len(m)):
if self.show_training_progress:
utils.print_progress_bar('Fb Trainer', i + 1, len(m))
if m[i]['sender_name'] == mimic_name:
text = clean(m[i]['content'])[:75]
prev_text = clean(m[i - 1]['content'])[-75:]
if len(text) == 0 or len(prev_text) == 0:
continue
search_text = self.chatbot.storage.tagger.get_bigram_pair_string(
text)
prev_search_text = self.chatbot.storage.tagger.get_bigram_pair_string(
prev_text)
#print(prev_text + ' ==> ' + text)
statement = Statement(text=text,
search_text=search_text,
in_response_to=prev_text,
search_in_response_to=prev_search_text,
conversation='training')
#print("statement from %s with response from %s" % (m[i-1]['sender_name'], m[i]['sender_name']))
statements.append(statement)
if len(statements):
self.chatbot.storage.create_many(statements)
def train(self):
conversation = "hello"
previous_statement_text = None
results = self.data.get_all_generic_ingredient()
print(results)
for conversation_count, text in enumerate(conversation):
if self.show_training_progress:
utils.print_progress_bar('Recipe Trainer',
conversation_count + 1,
len(conversation))
statement = self.get_or_create(text)
if previous_statement_text:
statement.add_response(Response(previous_statement_text))
previous_statement_text = statement.text
self.chatbot.storage.update(statement)
def train(self, *corpus_paths):
# Allow a list of corpora to be passed instead of arguments
if len(corpus_paths) == 1:
if isinstance(corpus_paths[0], list):
corpus_paths = corpus_paths[0]
# Train the chat bot with each statement and response pair
for corpus_path in corpus_paths:
corpora = self.corpus.load_corpus(corpus_path)
corpus_files = self.corpus.list_corpus_files(corpus_path)
for corpus_count, corpus in enumerate(corpora):
for conversation_count, conversation in enumerate(corpus):
print_progress_bar(
str(os.path.basename(corpus_files[corpus_count])) +
" Training", conversation_count + 1, len(corpus))
previous_statement_line = []
statement_line = []
for line in conversation:
for text in line:
statement = self.get_or_create(text)
statement.add_tags(corpus.categories)
statement_line.append(statement.text)
if previous_statement_line != []:
for previous_statement_text in previous_statement_line:
statement.add_response(
Response(previous_statement_text))
self.storage.update(statement)
previous_statement_line = statement_line
statement_line = []
def train(self, *corpus_paths):
from chatterbot.corpus import load_corpus, list_corpus_files
data_file_paths = []
# Get the paths to each file the bot will be trained with
for corpus_path in corpus_paths:
data_file_paths.extend(list_corpus_files(corpus_path))
for corpus, categories, file_path in load_corpus(*data_file_paths):
statements_to_create = []
# Train the chat bot with each statement and response pair
for conversation_count, conversations in enumerate(corpus):
if self.show_training_progress:
utils.print_progress_bar(
'Training ' + str(os.path.basename(file_path)),
conversation_count + 1, len(corpus))
previous_statements_texts = [None]
previous_statements_search_texts = ['']
for conversation in conversations:
if isinstance(conversation, str):
conversation = [conversation]
statements_texts = []
statements_search_texts = []
for previous_statement_text, previous_statement_search_text in zip(
previous_statements_texts,
previous_statements_search_texts):
for text in conversation:
statement_search_text = self.chatbot.storage.tagger.get_bigram_pair_string(
text)
statement = Statement(
text=text,
search_text=statement_search_text,
in_response_to=previous_statement_text,
search_in_response_to=
previous_statement_search_text,
conversation='training')
statement.add_tags(*categories)
statement = self.get_preprocessed_statement(
statement)
statements_texts.append(statement.text)
statements_search_texts.append(
statement_search_text)
statements_to_create.append(statement)
previous_statements_texts = statements_texts
previous_statements_search_texts = statements_search_texts
if statements_to_create:
self.chatbot.storage.create_many(statements_to_create)
def train(self, *corpus_paths):
from chatterbot.corpus import load_corpus, list_corpus_files
data_file_paths = []
# Get the paths to each file the bot will be trained with
for corpus_path in corpus_paths:
data_file_paths.extend(list_corpus_files(corpus_path))
for corpus, categories, file_path in load_corpus(*data_file_paths):
statements_to_create = []
# Train the chat bot with each statement and response pair
for conversation_count, conversation in enumerate(corpus):
if self.show_training_progress:
utils.print_progress_bar(
'Training ' + str(os.path.basename(file_path)),
conversation_count + 1, len(corpus))
previous_statement_text = None
previous_statement_search_text = ''
for text in conversation:
suggestion_tags = []
if text.strip('.?!/;:\'\"') in constants.AFFIRMATIVES:
text = 'AFF'
elif text.strip('.?!/;:\'\"') in constants.NEGATIVES:
text = 'NEG'
elif text[0] is '^':
(suggestion, text) = text.split(maxsplit=1)
suggestion = suggestion[1:]
if not suggestion.find('/'):
suggestion_tags.append(suggestion)
else:
for suggestion in suggestion.split('/'):
suggestion_tags.append(suggestion)
statement_search_text = self.chatbot.storage.tagger.get_bigram_pair_string(
text)
statement = Statement(
text=text,
search_text=statement_search_text,
in_response_to=previous_statement_text,
search_in_response_to=previous_statement_search_text,
conversation='training')
# YesNoLogicAdapter deals with responses to AFF/NEG via statement tags.
# No need for statements in_response_to = AFF/NEG In fact, it was causing
# erroneous responses
if statement.in_response_to in ['AFF', 'NEG']:
statement.in_response_to = None
statement.search_in_response_to = None
statement.add_tags(*categories)
if suggestion_tags:
for suggestion in suggestion_tags:
statement.add_tags('SUGGESTION:' + suggestion)
if previous_statement_text:
if previous_statement_text == 'AFF':
statements_to_create[-2].add_tags('AFF:' +
statement.text)
elif previous_statement_text == 'NEG':
statements_to_create[-2].add_tags('NEG:' +
statement.text)
statement = self.get_preprocessed_statement(statement)
previous_statement_text = statement.text
previous_statement_search_text = statement_search_text
statements_to_create.append(statement)
# Using update() because create_many() makes duplicate statements. AFF/NEG tag data was lost on some.
for stmnts in statements_to_create:
self.chatbot.storage.update(stmnts)
