Spaces:
Sleeping
Sleeping
File size: 3,300 Bytes
61a06c1 9b538d8 61a06c1 d592885 61a06c1 d592885 61a06c1 d592885 61a06c1 d592885 61a06c1 |
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 |
import streamlit as st
import pytube as pt
import os
import subprocess
import re
import whisper
from langchain.document_loaders import YoutubeLoader
from langchain.text_splitter import RecursiveCharacterTextSplitter
URL = 'URL'
TEXT = 'TEXT'
WHISPER = 'WHISPER'
PROCESSING = 'PROCESSING'
STATES = [URL, TEXT, WHISPER, PROCESSING]
AUDIO_FILE = "audio.mp3"
AUDIO_EXISTS = "AUDIO_EXISTS"
model = ''
st.title('Youtube Audio+Text')
def init_state():
if URL not in st.session_state:
st.session_state[URL] = ''
if TEXT not in st.session_state:
st.session_state[TEXT] = ''
if WHISPER not in st.session_state:
st.session_state[WHISPER] = ''
if AUDIO_EXISTS not in st.session_state:
st.session_state[AUDIO_EXISTS] = False
# if not st.session_state[URL]:
# clear_old_files()
def clear_old_files():
for file in os.listdir():
if file.endswith(".mp3") or file == 'transcript.txt':
os.remove(file)
print(f"Removed old files::{file}")
def extract_youtube_video_id(url):
regex = r"v=([^&]+)"
match = re.search(regex, url)
if match:
return match.group(1)
else:
return None
def load_whisper():
model = whisper.load_model("small")
print('Loaded Whisper Medium model')
if st.session_state[AUDIO_EXISTS]:
print('Transcribing with Whisper model')
result = model.transcribe("audio.mp3")
st.session_state[WHISPER] = result["text"]
write_file(result["text"], "transcript.txt")
def load_audio():
if os.path.exists(AUDIO_FILE):
st.session_state[AUDIO_EXISTS] = True
audio_file = open(AUDIO_FILE, 'rb')
audio_bytes = audio_file.read()
print(f"Audio file exists...{len(audio_bytes)}")
st.audio(audio_bytes, format="audio/mp3")
elif st.session_state[AUDIO_EXISTS]:
st.session_state[AUDIO_EXISTS] = False
def display():
container = st.container()
text_container = st.container()
load_audio()
#Download Button section
col1, col2 = st.columns(2)
with col1:
if st.session_state[AUDIO_EXISTS]:
st.download_button("Download Audio","file","audio.mp3","application/octet-stream")
with col2:
if os.path.exists("transcript.txt"):
st.download_button("Download Transcript",st.session_state[TEXT],"transcript.txt","text/plain")
with container:
with st.form(key='input_form'):
user_input = st.text_input("Youtube URL:", placeholder="http://www.youtube.com", key=URL)
input_submit_button = st.form_submit_button(label='Send')
if input_submit_button and user_input:
download()
load_whisper()
with text_container:
st.text_area(label="Youtube Transcript:",
height=200,
value=st.session_state[WHISPER])
def download():
id = extract_youtube_video_id(st.session_state[URL])
command = [f"yt-dlp --no-config -v --extract-audio --audio-format mp3 {st.session_state[URL]} -o audio.mp3"]
print(command)
out = subprocess.run(command, shell=True)
def write_file(text, filename):
with open(filename, "w") as f:
f.write(text)
def main():
init_state()
display()
if __name__ == "__main__":
main() |