Spaces:
Runtime error
Runtime error
File size: 6,927 Bytes
0d0cd1a |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 |
from yaspin import yaspin
from termcolor import colored
from keys import OPENAI_API_KEY
with yaspin(text="Waking agent...") as spinner:
import os
import time
import requests
import base64
import threading
import scipy.io.wavfile as wav
from queue import Queue
os.environ['PYGAME_HIDE_SUPPORT_PROMPT'] = "hide"
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'
from pygame import mixer
from modules.Whisper import transcribe
from modules.VoiceActivityDetection import VADDetector
import openai
from gtts import gTTS
from modules.command import CommandDetector
from modules.Yolo import Eyes
# from modules.google import GoogleManager
# from modules.github import create_repository
openai.api_key = OPENAI_API_KEY
mixer.init()
class GPTAssistant():
def __init__(self, startListening=False, startTexting=False, voice=False, local=False):
self.voice = voice
self.listening = startListening
self.texting = startTexting
self.vad = VADDetector(self.onSpeechStart, self.onSpeechEnd)
self.vad_data = Queue()
self.context = [
{"role": "system", "content": self.read_system_context("jvp.txt")}]
self.cdet = CommandDetector(model_path="./models/checkpoint-760")
# self.google = GoogleManager()
self.eyes = Eyes()
if startListening and not startTexting:
self.startListening()
t = threading.Thread(target=self.transcription_loop)
t.start()
else:
self.writingMessage()
def writingMessage(self):
text = ''
while True:
text = input(colored("[👨]: ", "magenta"))
self.build_context(role='user', content=text)
command = self.cdet.command_filter(text)
if (command != 'goodbye'):
if command == "vision":
vision = self.eyes.see()
self.build_context(
role='system', content=f'The vision module detected {vision}. Respond to the last user promt using this information.')
if command == "google":
self.google.get_query(text)
if (self.voice):
self.play_audio(response=self.google.notification,
exit=exit, response_name="google_notification.mp3")
search = self.google.search()
self.build_context(
role='system', content=f'The google module found {search}. Respond to the last user promt using this information.')
if command == "github":
repo = create_repository()
self.build_context(
role='system', content=f'The github module tried to create a repository and exited:\n {repo}. Tell the user what happened.')
self.send_to_GPT(messages=self.context)
else:
self.send_to_GPT(messages=self.context)
break
def startListening(self):
print(colored("Listening 👂", 'green'))
t = threading.Thread(target=self.vad.startListening)
t.start()
def toggleListening(self):
if not self.listening:
print()
print(colored("Listening 👂", 'green'))
while not self.vad_data.empty():
self.vad_data.get()
self.listening = not self.listening
def onSpeechStart(self):
pass
def onSpeechEnd(self, data):
if data.any():
self.vad_data.put(data)
def transcription_loop(self):
while True:
if not self.vad_data.empty():
data = self.vad_data.get()
if self.listening:
self.toggleListening()
text = transcribe(data)
if len(text) > 4 and text != "Thank you.":
print(colored(f'[👨]:{text}', 'magenta'))
self.build_context(role='user', content=text)
command = self.cdet.command_filter(text)
if (command != 'goodbye'):
if command == "vision":
vision = self.eyes.see()
self.build_context(
role='system', content=f'The vision module detected {vision}. Respond to the last user promt using this information.')
if command == "google":
self.google.get_query(text)
if (self.voice):
self.play_audio(
response=self.google.notification, exit=exit, response_name="google_notification.mp3")
search = self.google.search()
self.build_context(
role='system', content=f'The google module found {search}. Respond to the last user promt using this information.')
self.send_to_GPT(messages=self.context)
else:
self.send_to_GPT(messages=self.context, exit=True)
break
def read_system_context(self, file):
context = ''
with open(file) as f:
lines = f.readlines()
for line in lines:
context += line
return context
def build_context(self, role, content):
self.context.append({"role": role, "content": content})
def send_to_GPT(self, messages, exit=False):
completion = openai.ChatCompletion.create(
model='gpt-3.5-turbo',
messages=messages)
response = completion['choices'][0]['message']['content']
print(colored(f'[🤖]:{response}', 'green'))
self.build_context(role='assistant', content=response)
if (self.voice):
self.play_audio(response=response, exit=exit)
def play_audio(self, response, language="en", exit=False, response_name="GPT_response.mp3"):
speech = gTTS(text=response, lang=language, slow=False)
speech.save(response_name)
# play audio
mixer.music.load(response_name)
mixer.music.play()
# wait for audio to finish
duration = mixer.Sound(response_name).get_length()
time.sleep(duration + 1)
# unload and delete audio
mixer.music.unload()
os.remove(response_name)
# re-activate microphone
if (self.listening and not exit):
self.toggleListening()
if __name__ == '__main__':
assistant = GPTAssistant(
startListening=True, startTexting=True, voice=False, local=False)
context = [{"role": "user", "content": "Testing GPT"}]
assistant.send_to_GPT(messages=context)
|