File size: 3,030 Bytes
093e4b1 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 |
import tensorflow as tf
import numpy as np
# Function to tokenize input data
def tokenize_input(data):
tokenizer = tf.keras.preprocessing.text.Tokenizer()
tokenizer.fit_on_texts(data)
vocab_size = len(tokenizer.word_index) + 1
input_seq = tokenizer.texts_to_sequences(data)
max_len = max([len(x) for x in input_seq])
input_seq = tf.keras.preprocessing.sequence.pad_sequences(input_seq, maxlen=max_len, padding='post')
return tokenizer, input_seq, max_len, vocab_size
# Function to define the neural network model
def define_model(vocab_size, max_len):
model = tf.keras.Sequential([
tf.keras.layers.Embedding(vocab_size, 16),
tf.keras.layers.GlobalAveragePooling1D(),
tf.keras.layers.Dense(16, activation='relu'),
tf.keras.layers.Dense(max_len, activation='softmax')
])
model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])
return model
# Function to train the neural network
def train_model(model, input_seq, output, epochs=10):
model.fit(input_seq, output, epochs=epochs, verbose=1)
return model
# Function to test the model on new input
def test_model(model, test_data, tokenizer, max_len):
test_seq = tokenizer.texts_to_sequences(test_data)
test_seq = tf.keras.preprocessing.sequence.pad_sequences(test_seq, maxlen=max_len, padding='post')
predictions = model.predict(test_seq)
return predictions
# Define the input and output data
data = np.array(['create a python script', 'build a program', 'generate a code'])
output = np.array([['create', 'python', 'script'], ['build', 'program'], ['generate', 'code']])
# Tokenize the input data
tokenizer, input_seq, max_len, vocab_size = tokenize_input(data)
# Define the neural network model
model = define_model(vocab_size, max_len)
# Train the neural network model
model = train_model(model, input_seq, output)
# Define a function to generate a response based on the predicted output
def generate_response(predictions, tokenizer):
output = [tokenizer.index_word[i] for i in np.argmax(predictions, axis=1)]
if 'create' in output and 'python' in output and 'script' in output:
return 'Sure, I can help you with that! What kind of script would you like to create?'
elif 'build' in output and 'program' in output:
return 'Of course! What programming language would you like to use for your program?'
elif 'generate' in output and 'code' in output:
return 'Certainly! What kind of code do you need generated?'
else:
return "I'm sorry, I didn't understand your request. Can you please rephrase it?"
# Implement the chatbot
print("Hi, I'm BootLeggerAI, a chatbot that can help you with your programming needs! What can I assist you with today?")
while True:
user_input = input("Please enter your request: ")
test_data = np.array([user_input])
predictions = test_model(model, test_data, tokenizer, max_len)
response = generate_response(predictions, tokenizer)
print(response) |