File size: 2,693 Bytes
cfc4ac5
 
 
 
308f38a
cfc4ac5
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
308f38a
 
 
67ca6ac
cfc4ac5
 
 
 
 
5ed3033
cfc4ac5
67ca6ac
308f38a
67ca6ac
cfc4ac5
 
 
 
 
 
 
 
 
 
 
1cb05db
cfc4ac5
 
 
 
 
 
 
 
 
 
 
6dff63d
cfc4ac5
 
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
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
from elevenlabs import set_api_key
from elevenlabs import clone, generate, play
from transformers import AutoModelForSeq2SeqLM, AutoTokenizer
from pydub import AudioSegment
from huggingface_hub.hf_api import HfFolder
HfFolder.save_token('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')
access_token = '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'

import streamlit as st


#language translation function 
def translate(text): 
    tokenizer = AutoTokenizer.from_pretrained(".",use_auth_token=True)
    #model = AutoModelForSeq2SeqLM.from_pretrained("facebook/nllb-200-distilled-600M",use_auth_token=True)
    model = AutoModelForSeq2SeqLM.from_pretrained(".",use_auth_token=True)
    inputs = tokenizer(text, return_tensors="pt")
    translated_tokens = model.generate(**inputs, 
                                           forced_bos_token_id=tokenizer.lang_code_to_id["hin_Deva"], 
                                           max_length=100)
    result = tokenizer.batch_decode(translated_tokens, skip_special_tokens=True)[0]
    return result 


def save_uploaded_file_as_mp3(uploaded_file, output_file_path):
    audio = AudioSegment.from_file(uploaded_file)
    audio.export(output_file_path, format="mp3")

def ui(): 
    st.title('Multi Lingual Voice Cloner')
    st.markdown("Made by vageesh")

    #audio input box 
    audio_file = st.file_uploader("Upload an audio file that needs to be cloned", type=[ "wav,Mp4","Mp3"])
    if audio_file is not None:
        output_file_path = "./output_audio.mp3"
        save_uploaded_file_as_mp3(audio_file, output_file_path)
        st.success(f"Audio file saved as {output_file_path}")
        user_api_key = st.sidebar.text_input(
            label="#### Your Eleven Labs API key here 👇",
            placeholder="Paste your Eleven Labs API key API key, sk-",
            type="password")
    
        if user_api_key is not None and user_api_key.strip() != "":
            set_api_key(user_api_key)
            #making an voice 
            voice = clone(
                name="Alex",
                description="An middle aged American male voice with a slight hoarseness in his throat. Perfect for npodcast", # Optional
                files="./output_audio.mp3",
                model='eleven_multilingual_v1',
                language='Hindi'
            )

            in_text=st.text_input("Paste the text you want to hear from english to hindi")
            if in_text is not None and in_text.strip() != "":
                hin_text=translate(in_text)
                audio = generate(text=hin_text, voice=voice)

                st.audio(audio)

if __name__=="__main__": 
    ui()