File size: 6,493 Bytes
969ab0b
 
 
 
 
 
 
 
 
 
 
 
cee36ef
969ab0b
 
 
 
911caa6
8cc4893
cee36ef
969ab0b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
%%writefile app.py
from IPython.display import Javascript
from IPython import display
from google.colab import output
from base64 import b64decode
import datetime
import whisper
import openai
import os
import base64
from Crypto.Cipher import AES
from streamlit_bokeh_events import streamlit_bokeh_events
import streamlit as st
from bokeh.models.widgets import Button
from bokeh.models.widgets.buttons import Button
from bokeh.models import CustomJS
from streamlit_bokeh_events import streamlit_bokeh_events



RECORD = """
const sleep  = time => new Promise(resolve => setTimeout(resolve, time))
const b2text = blob => new Promise(resolve => {
  const reader = new FileReader()
  reader.onloadend = e => resolve(e.srcElement.result)
  reader.readAsDataURL(blob)
})
var record = time => new Promise(async resolve => {
  stream = await navigator.mediaDevices.getUserMedia({ audio: true })
  recorder = new MediaRecorder(stream)
  chunks = []
  recorder.ondataavailable = e => chunks.push(e.data)
  recorder.start()
  await sleep(time)
  recorder.onstop = async ()=>{
    blob = new Blob(chunks)
    text = await b2text(blob)
    resolve(text)
  }
  recorder.stop()
})
"""

openai.api_key = os.environ["API_KEY"]

with open("encrypt.txt", "r") as encfile:
  encoder_txt = encfile.read()

with open("decrypt.txt", "r") as decfile:
  decoder_txt = decfile.read()

def openai_fun(myprompt):
  response_encoded = openai.Completion.create(
                engine="text-davinci-003",
                prompt = myprompt,
                max_tokens=1024,
                n=1,
                stop=None,
                temperature=0.5,
            )
  return response_encoded

def record(sec=5):
  display.display(Javascript(RECORD))
  s = output.eval_js('record(%d)' % (sec*1000))
  b = b64decode(s.split(',')[1])
  ts = datetime.datetime.now()
  filename = ts.strftime("%Y_%m_%d_%H_%M_%S") 
  with open(f'{filename}.wav','wb') as f:
    f.write(b)
  return f'{filename}.wav'  # or webm ?

model = whisper.load_model("base")
transcribed = []

while True:
  user_choice = st.text_input("Do you want to record a new audio for transcription?[y/n]")
  
  if user_choice == 'y':
    st.write('Recording! (5 seconds)')
    record(5)
    folder_path = "/content"
    audio_files = [f for f in os.listdir(folder_path) if f.endswith(".wav")]

    audio_files.sort(key=lambda x: os.path.getmtime(os.path.join(folder_path, x)), reverse=True)
    last_audio_file_path = os.path.join(folder_path, audio_files[0])
    st.write('Transcribing audio file: ',last_audio_file_path)
    # COMMENT IF NOT NEEDED:
    if os.path.exists(last_audio_file_path) and not last_audio_file_path in transcribed:
      audio = whisper.load_audio(last_audio_file_path)
      audio = whisper.pad_or_trim(audio)
      mel = whisper.log_mel_spectrogram(audio).to(model.device)
      options = whisper.DecodingOptions(language= 'en', fp16=False)

      result = whisper.decode(model, mel, options)

      if result.no_speech_prob < 0.5:
          mymsg = result.text
          st.write("Actual Message: ",mymsg)

          enc_prompt = encoder_txt + mymsg

          openai_fun(enc_prompt)

          if openai_fun(enc_prompt)['choices'][0]['text'] != "":
              # print(response_encoded['choices'][0]['text'])
              exec(openai_fun(enc_prompt)['choices'][0]['text'])
              encoded_msg = enc(mymsg)
              print("The encoded message: ", encoded_msg)

              decode_ = st.text_input("Do you wish to decode the message?[y/n]")
              if decode_ == "y":
                    dec_prompt = decoder_txt + str(encoded_msg)
                    response_decoded = openai.Completion.create(
                      engine="text-davinci-003",
                      prompt = dec_prompt,
                      max_tokens=500,
                      n=1,
                      stop=None,
                      temperature=0.5,
                    )

                    if response_decoded['choices'][0]['text'] != "":
                        print(response_decoded['choices'][0]['text'])
                        exec(response_decoded['choices'][0]['text'])
                        decoded_msg = dec(encoded_msg, key)
                        print("The decoded message: ", decoded_msg)


          else:
              st.write('Retry! The message could')


    break # exit the loop
  
  elif user_choice == 'n':
    uc1 = input('Do you want to transcribe an existing audio?[y/n]')
    
    if uc1 == 'y':
      folder_path = "/content"
      audio_files = [f for f in os.listdir(folder_path) if f.endswith(".wav")]
      print('Audio files present: ',audio_files)

      audio_files.sort(key=lambda x: os.path.getmtime(os.path.join(folder_path, x)), reverse=True)
      last_audio_file_path = os.path.join(folder_path, audio_files[0])
      print('Transcribing last audio file: ',last_audio_file_path)

      # COMMENT IF NOT NEEDED:
      if os.path.exists(last_audio_file_path) and not last_audio_file_path in transcribed:
        audio = whisper.load_audio(last_audio_file_path)
        audio = whisper.pad_or_trim(audio)
        mel = whisper.log_mel_spectrogram(audio).to(model.device)
        options = whisper.DecodingOptions(language= 'en', fp16=False)

        result = whisper.decode(model, mel, options)

        if result.no_speech_prob < 0.5:
            mymsg = result.text
            print("Actual Message: ",mymsg)
            enc_prompt = encoder_txt + result.text
            response_encoded = openai.Completion.create(
                engine="text-davinci-003",
                prompt = enc_prompt,
                max_tokens=1024,
                n=1,
                stop=None,
                temperature=0.5,
            )
            
            if response_encoded['choices'][0]['text'] != "":
              # print(response_encoded['choices'][0]['text'])
              exec(response_encoded['choices'][0]['text'])
              encoded_msg = enc(mymsg)
              st.write("The encoded message: ", encoded_msg)

            else:
                st.write('Retry! The message could')
            
            # DELETE audio

      break # exit the loop

    elif uc1 == 'n':
      continue # continue the loop, prompting for input again

    else:
      st.write('Invalid input, please enter y or n')
      continue # continue the loop, prompting for input again

  else:
    st.write('Invalid input, please enter y or n')
    continue # continue the loop, prompting for input again