Hugging Face's logo

Spaces:
alisrbdni
/
Speech-Text-Generator-Speech
Runtime error

App Files Community
Speech-Text-Generator-Speech
File size: 8,695 Bytes 
6128691
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
228eb39
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
6128691
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
228eb39
 
6128691
 
 
 
 
 
 
 
 
 
 
 
 
 
228eb39
 
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
238
239
240
 
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("""
#    —
#    Sending audio to Whisper ...
#    —
#    """)
#    
#    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("———————————————————————————————————————————")  
#      
#    return transcription.text

def translate(audio):
    print("""
    —
    Sending audio to Whisper ...
    —
    """)
   
    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)