joydeb28

class Prediction():
    def __init__(self):
        self.model = model_obj.model
        
    def predict_validation(self):
        valid_sentences = load_data_obj.validation_data_frame["query"].tolist()
        valid_labels = load_data_obj.validation_data_frame["category"].tolist()

        preprocess_bert_data_obj = PreprocessingBertData()
        val_x = preprocess_bert_data_obj.prepare_data_x(valid_sentences)

joydeb28

class Evaluation():
    def get_accuracy(self,actuals, predictions):
        acc = accuracy_score(actuals, predictions)
        return acc
      
eval_obj = Evaluation()
ytest,ypred = pred_obj.predict_validation()
#print(ytest,ypred)
acc = eval_obj.get_accuracy(ytest,ypred)
print("Auc: {:.2%}".format(acc))

joydeb28

class DesignModel():
    def __init__(self):
        self.model = None        
        self.train_data = [train_input_ids, train_input_masks, train_segment_ids]
        self.train_labels = train_labels
        
    def bert_model(self,max_seq_length): 
        in_id = Input(shape=(max_seq_length,), dtype=tf.int32, name="input_ids")
        in_mask = Input(shape=(max_seq_length,), dtype=tf.int32, name="input_masks")
        in_segment = Input(shape=(max_seq_length,), dtype=tf.int32, name="segment_ids")

joydeb28

class PreprocessingBertData():
    
    def prepare_data_x(self,train_sentences):
        x = bert_model_obj.get_input_array(train_sentences)
        return x
    
    def prepare_data_y(self,train_labels):
        y = list()
        for item in train_labels:
            label = item

joydeb28

class BertModel(object):
    
    def __init__(self):
        
        self.max_len = 128
        bert_path = "https://tfhub.dev/tensorflow/bert_en_uncased_L-12_H-768_A-12/1"
        FullTokenizer=bert.bert_tokenization.FullTokenizer
        
        self.bert_module = hub.KerasLayer(bert_path,trainable=True)

joydeb28

class LoadingData():
            
    def __init__(self):
        train_file_path = os.path.join("benchmarking_data","Train")
        validation_file_path = os.path.join("benchmarking_data","Validate")
        category_id = 0
        self.cat_to_intent = {}
        self.intent_to_cat = {}
        
        for dirname, _, filenames in os.walk(train_file_path):

joydeb28

100%|██████████| 1/1 [00:00<00:00, 1014.83it/s]
100%|██████████| 1/1 [00:00<00:00, 506.37it/s]
100%|██████████| 1/1 [00:00<00:00, 972.71it/s]
100%|██████████| 1/1 [00:00<00:00, 514.95it/s]
100%|██████████| 1/1 [00:00<00:00, 489.42it/s]
Predicted Intent: SearchCreativeWork	 Actual Intent: SearchScreeningEvent   Query: I want to see Medal for the General
Predicted Intent: BookRestaurant	 Actual Intent: BookRestaurant	       Query: Book a reservation for 5 people at the top-rated brasserie restaurant
Predicted Intent: AddToPlaylist	         Actual Intent: AddToPlaylist	       Query: Can I put this tune onto my sin estres playlist?
Predicted Intent: AddToPlaylist	         Actual Intent: AddToPlaylist	       Query: add the artist Pete Murray to my relaxing playlist
Predicted Intent: BookRestaurant	 Actual Intent: BookRestaurant	       Query: Book me a reservation for a party of 3 at a pub in Northern Mariana Islands

joydeb28

from numpy import array
import tensorflow as tf
from tensorflow.keras.preprocessing.text import Tokenizer
from tensorflow.keras.utils import to_categorical
from tensorflow.keras.preprocessing.sequence import pad_sequences
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense
from tensorflow.keras.layers import LSTM
from tensorflow.keras.layers import Dropout
from tensorflow.keras.layers import Embedding