srbdni
Update app.py
228eb39
import tempfile
import gradio as gr
from neon_tts_plugin_coqui import CoquiTTS
LANGUAGES = list(CoquiTTS.langs.keys())
default_lang = "en"
import telnetlib
#import whisper
#whisper_model = whisper.load_model("small")
#whisper = gr.Interface.load(name="spaces/sanchit-gandhi/whisper-large-v2")
chatgpt = gr.Blocks.load(name="spaces/fffiloni/whisper-to-chatGPT")
import os
import json
session_token = os.environ.get('SessionToken')
bypass_node = "https://gpt.pawan.krd"
#api_endpoint = os.environ.get('API_EndPoint')
# ChatGPT
#from revChatGPT.ChatGPT import Chatbot
#chatbot = Chatbot({"session_token": session_token}) # You can start a custom conversation
import asyncio
from pygpt import PyGPT
import argparse
import sys
import asyncio
from ChatGPT_lite.ChatGPT import Chatbot
import os
os.system("pip install numpy==1.23.1")
os.system("pip install gradio==2.7.5.2")
os.system("python -m pip install paddlepaddle-gpu==2.2.1.post112 -f https://www.paddlepaddle.org.cn/whl/linux/mkl/avx/stable.html")
os.system("hub install wav2lip==1.0.0")
import gradio as gr
import paddlehub as hub
module = hub.Module(name="wav2lip")
def inference(image,audio):
module.wav2lip_transfer(face=image, audio=audio, output_dir='.', use_gpu=False)
return "result.mp4"
title = ""
description = ""
article = ""
examples=[['monatest.jpeg',"game.wav"]]
title = "Speech to ChatGPT to Speech"
#info = "more info at [Neon Coqui TTS Plugin](https://github.com/NeonGeckoCom/neon-tts-plugin-coqui), [Coqui TTS](https://github.com/coqui-ai/TTS)"
#badge = "https://visitor-badge-reloaded.herokuapp.com/badge?page_id=neongeckocom.neon-tts-plugin-coqui"
coquiTTS = CoquiTTS()
chat_id = {'conversation_id': None, 'parent_id': None}
headers = {'Authorization': 'yusin'}
##############################################################
async def async_main(prompt):
if session_token is None:
print("Please provide a session token")
print(session_token, bypass_node)
chat = Chatbot(session_token, bypass_node)
await asyncio.gather(chat.wait_for_ready())
while True:
response = await chat.ask(prompt)
print(f"\nBot: {response['answer']}\n")
# Close sockets
chat.close()
# exit
#sys.exit(0)
print(response)
return response['answer']
def sync_main(prompt):
chat = Chatbot(session_token, bypass_node)
# Create loop
loop = asyncio.new_event_loop()
# Set
asyncio.set_event_loop(loop)
# Run
loop.run_until_complete(chat.wait_for_ready())
while True:
response = loop.run_until_complete(chat.ask(prompt))
print(f"\nBot: {response['answer']}\n")
# Close sockets
chat.close()
# stop asyncio event loop
loop.stop()
# exit
#sys.exit(0)
print(response)
return response['answer']
###########################################
parser = argparse.ArgumentParser()
parser.add_argument('--session_token', type=str, default=None)
parser.add_argument('--bypass_node', type=str, default="https://gpt.pawan.krd")
parser.add_argument('--async_mode', action='store_true')
args = parser.parse_args()
args.session_token = session_token
##########################################
async def chat_gpt_ask(prompt):
print(session_token)
chat_gpt = PyGPT(session_token)
await chat_gpt.connect()
await chat_gpt.wait_for_ready()
print(prompt)
answer = await chat_gpt.ask(prompt)
print(answer)
await chat_gpt.disconnect()
# ChatGPT
def chat_hf(audio, custom_token, language):
output = chatgpt(audio, "transcribe", fn_index=0)
whisper_text, gpt_response = output[0], output[1]
#whisper_text = translate(audio)
#gpt_response = asyncio.run(async_main(whisper_text))
'''
try:
whisper_text = translate(audio)
if whisper_text == "ERROR: You have to either use the microphone or upload an audio file":
gpt_response = "MISSING AUDIO: Record your voice by clicking the microphone button, do not forget to stop recording before sending your message ;)"
else:
#gpt_response = chatbot.ask(whisper_text, conversation_id=conversation_id, parent_id=None)
#gpt_response = asyncio.run(chat_gpt_ask(whisper_text))
gpt_response = asyncio.run(async_main(whisper_text))
#gpt_response = async_main(whisper_text)
#if chat_id['conversation_id'] != None:
# data = {"content": whisper_text, "conversation_id": chat_id['conversation_id'], "parent_id": chat_id['parent_id']}
#else:
# data = {"content": whisper_text}
#print(data)
#res = requests.get('http://myip.ipip.net', timeout=5).text
#print(res)
#response = requests.post('api_endpoint', headers=headers, json=data, verify=False, timeout=5)
#print('this is my answear', response.text)
#chat_id['parent_id'] = response.json()["response_id"]
#chat_id['conversation_id'] = response.json()["conversation_id"]
#gpt_response = response.json()["content"]
#response = requests.get('https://api.pawan.krd/chat/gpt?text=' + whisper_text + '&cache=false', verify=False, timeout=5)
#print(response.text)
#whisper_text = translate(audio)
#api = ChatGPT(session_token)
#resp = api.send_message(whisper_text)
#api.refresh_auth() # refresh the authorization token
#api.reset_conversation() # reset the conversation
#gpt_response = resp['message']
except:
whisper_text = translate(audio)
gpt_response = """Sorry, I'm quite busy right now, but please try again later :)"""
#whisper_text = translate(audio)
#api = ChatGPT(custom_token)
#resp = api.send_message(whisper_text)
#api.refresh_auth() # refresh the authorization token
#api.reset_conversation() # reset the conversation
#gpt_response = resp['message']
'''
# to voice
with tempfile.NamedTemporaryFile(suffix=".wav", delete=False) as fp:
coquiTTS.get_tts(gpt_response, fp, speaker = {"language" : language})
return whisper_text, gpt_response, fp.name
# whisper
#def translate(audio):
# print("""
# β€”
# Sending audio to Whisper ...
# β€”
# """)
#
# audio = whisper.load_audio(audio)
# audio = whisper.pad_or_trim(audio)
#
# mel = whisper.log_mel_spectrogram(audio).to(whisper_model.device)
#
# _, probs = whisper_model.detect_language(mel)
#
# transcript_options = whisper.DecodingOptions(task="transcribe", fp16 = False)
#
# transcription = whisper.decode(whisper_model, mel, transcript_options)
#
# print("language spoken: " + transcription.language)
# print("transcript: " + transcription.text)
# print("β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”β€”")
#
# return transcription.text
def translate(audio):
print("""
β€”
Sending audio to Whisper ...
β€”
""")
text_result = whisper(audio, None, "transcribe", fn_index=0)
print(text_result)
return text_result
with gr.Blocks() as blocks:
gr.Markdown("<h1 style='text-align: center; margin-bottom: 1rem'>"
+ title
+ "</h1>")
#gr.Markdown(description)
radio = gr.Radio(label="Language",choices=LANGUAGES,value=default_lang)
with gr.Row(equal_height=True):# equal_height=False
with gr.Column():# variant="panel"
audio_file = gr.Audio(source="microphone",type="filepath")
custom_token = gr.Textbox(label='If it fails, use your own session token', placeholder="your own session token")
with gr.Row():# mobile_collapse=False
submit = gr.Button("Submit", variant="primary")
with gr.Column():
text1 = gr.Textbox(label="Speech to Text")
text2 = gr.Textbox(label="ChatGPT Response")
audio = gr.Audio(label="Output", interactive=False)
iface = gr.Interface(inference, [gr.inputs.Image(type="filepath"),audio],
outputs=gr.outputs.Video(label="Output Video"),article=article,description=description)
#gr.Markdown(info)
#gr.Markdown("<center>"
# +f'<img src={badge} alt="visitors badge"/>'
# +"</center>")
# actions
submit.click(
chat_hf,
[audio_file, custom_token, radio],
[text1, text2, audio],
)
radio.change(lambda lang: CoquiTTS.langs[lang]["sentence"], radio, text2)
blocks.launch(debug=True,cache_examples=True,enable_queue=True)