File size: 3,038 Bytes
079312c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
aa9c213
079312c
aa9c213
 
4d96d8c
aa9c213
 
 
 
 
 
 
 
079312c
 
 
 
 
 
 
 
 
 
 
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
import streamlit as st
import ssl
ssl._create_default_https_context = ssl._create_unverified_context
import glob
import os

def vid_to_audio(url=None):
    # importing packages 
    from pytube import YouTube 
    import os 

    # url input from user 
    yt = YouTube(url) 

    # extract only audio 
    video = yt.streams.filter(only_audio=True).first() 

    # check for destination to save file 
    destination = '.'

    # download the file 
    out_file = video.download(output_path=destination) 

    # save the file 
    base, ext = os.path.splitext(out_file) 
    new_file = base + '.mp3'
    os.rename(out_file, new_file) 

    # result of success 
    print(yt.title + " has been successfully downloaded.")

    return "OK"

#vid_to_text(url='https://youtu.be/FE5tva_o7ew?si=ztkKeO7qwcpC36AS')

def audio_to_text():
    import torch
    
    
    from transformers import AutoModelForSpeechSeq2Seq, AutoProcessor, pipeline
    


    device = "cuda:0" if torch.cuda.is_available() else "cpu"
    torch_dtype = torch.float16 if torch.cuda.is_available() else torch.float32

    model_id = "openai/whisper-tiny"

    model = AutoModelForSpeechSeq2Seq.from_pretrained(
        model_id, torch_dtype=torch_dtype, low_cpu_mem_usage=True, use_safetensors=True
    )
    model.to(device)

    processor = AutoProcessor.from_pretrained(model_id)
    #
    pipe = pipeline(
    "automatic-speech-recognition",
    model=model,
    tokenizer=processor.tokenizer,
    feature_extractor=processor.feature_extractor,
    max_new_tokens=128,
    chunk_length_s=30,
    batch_size=16,

    torch_dtype=torch_dtype,
    device=device,
    )
    
    #files = glob.glob('*.mp3')[0]
    files = os.listdir()
    # Get a list of all files in the current directory
    files = os.listdir()
    #st.write(files)

    # Create an empty list to collect results
    results = []
    
    # Iterate through the files
    for i in files:
        if ".mp3" in i:
            # Build the full path to the MP3 file
            file_path = os.path.join(os.getcwd(), i)
    
            # Display information (optional)
            st.write("Current Directory:", os.getcwd())
            st.write("File Path:", file_path)
            
            
            result = pipe(file_path)
            #print(result)
            return result['text']
    


def summarize():
    transcript = audio_to_text()
    import requests
    
    API_URL = "https://api-inference.huggingface.co/models/Azma-AI/bart-large-text-summarizer"
    headers = {"Authorization": f"Bearer {API_TOKEN}"}
    
    def query(payload):
    	response = requests.post(API_URL, headers=headers, json=payload)
    	return response.json()
    	
    output = query({
    	"inputs": transcript["text"],
    })
    return output

yt_link = st.text_input("Enter the YouTube URL: ")

if st.button("Start Summarization"):
    
    with st.status("Downloading the video..."):
        vid_to_audio(url=yt_link)
    with st.status("Summarizing..."):
        s = summarize()
        st.write(s)