shubham0204

import os
import yaml

dir_path = 'raw_data'
files_list = os.listdir(dir_path + os.sep)

questions = list()
answers = list()
for filepath in files_list:
    stream = open( dir_path + os.sep + filepath , 'rb')

shubham0204

import tensorflow as tf

encoder_inputs = tf.keras.layers.Input(shape=( None , ))
encoder_embedding = tf.keras.layers.Embedding( num_tokens, 200 , mask_zero=True) (encoder_inputs)
encoder_outputs , state_h , state_c = tf.keras.layers.LSTM( 200 , return_state=True )( encoder_embedding )
encoder_states = [ state_h , state_c ]

decoder_inputs = tf.keras.layers.Input(shape=( None ,  ))
decoder_embedding = tf.keras.layers.Embedding( num_tokens, 200 , mask_zero=True) (decoder_inputs)

shubham0204

def make_inference_models():
    
    encoder_model = tf.keras.models.Model(encoder_inputs, encoder_states)
    
    decoder_state_input_h = tf.keras.layers.Input(shape=( 200 ,))
    decoder_state_input_c = tf.keras.layers.Input(shape=( 200 ,))
    
    decoder_states_inputs = [decoder_state_input_h, decoder_state_input_c]
    

shubham0204

from tensorflow.keras import models, layers, activations, losses, optimizers
import tensorflow.keras.backend as K
import tensorflow as tf

DIMEN = 128 # dimension of the image

input_shape = ( (DIMEN**2) * 3 , )
convolution_shape = ( DIMEN , DIMEN , 3 )
kernel_size_1 = ( 4 , 4 )
kernel_size_2 = ( 3 , 3 )

shubham0204

custom_images = recognizer.prepare_images_from_dir( 'custom_images/'  )
class_1_images = recognizer.prepare_images_from_dir( 'images/p1/' )
class_2_images = recognizer.prepare_images_from_dir( 'images/p2/' )

scores = list()
labels = list()
for image in custom_images:
    label = list()
    score = list()
    for sample in class_1_images :

shubham0204

import tensorflow as tf
from tensorflow.keras import optimizers,losses,activations
from tensorflow.keras.layers import *

dropout_rate = 0.5
input_shape = ( maxlen ,  )
target_shape = ( maxlen , 1 )

# Note : activations.leaky_relu is a CUSTOM IMPLEMENTATION. It DOES NOT EXIST in the official TensorFlow build.

shubham0204

import json

with open( 'android/word_dict.json' , 'w' ) as file:
	json.dump( tokenizer.word_index , file )

shubham0204

import tensorflow as tf

converter = tf.lite.TFLiteConverter.from_keras_model_file( 'models/model.h5' )
converter.post_training_quantize = True
tflite_buffer = converter.convert()
open( 'android/model.tflite' , 'wb' ).write( tflite_buffer )

shubham0204

private fun loadJSONFromAsset(filename : String? ): String? {
    var json: String? = null
    try {
        val inputStream = context!!.assets.open(filename )
        val size = inputStream.available()
        val buffer = ByteArray(size)
        inputStream.read(buffer)
        inputStream.close()
        json = String(buffer)

shubham0204

fun tokenize ( message : String ): IntArray {
    val parts : List<String> = message.split(" " )
    val tokenizedMessage = ArrayList<Int>()
    for ( part in parts ) {
        if (part.trim() != ""){
            var index : Int? = 0
            if ( vocabData!![part] == null ) {
                index = 0
            }
            else{