Update app.py
Browse files
app.py
CHANGED
|
@@ -4,7 +4,7 @@ import torch
|
|
| 4 |
from transformers import WhisperProcessor, WhisperForConditionalGeneration
|
| 5 |
import requests
|
| 6 |
from bs4 import BeautifulSoup
|
| 7 |
-
import tempfile
|
| 8 |
import os
|
| 9 |
import soundfile as sf
|
| 10 |
from spellchecker import SpellChecker
|
|
@@ -12,9 +12,12 @@ from pydub import AudioSegment
|
|
| 12 |
import librosa
|
| 13 |
import numpy as np
|
| 14 |
from pyannote.audio import Pipeline
|
| 15 |
-
|
| 16 |
-
from
|
| 17 |
-
|
|
|
|
|
|
|
|
|
|
| 18 |
|
| 19 |
# Initialize the speaker diarization pipeline
|
| 20 |
try:
|
|
@@ -171,18 +174,57 @@ def transcribe_video(url):
|
|
| 171 |
print(error_message)
|
| 172 |
return error_message
|
| 173 |
|
| 174 |
-
|
| 175 |
-
|
| 176 |
-
|
| 177 |
-
|
| 178 |
-
|
| 179 |
-
|
| 180 |
-
|
| 181 |
-
|
| 182 |
-
|
| 183 |
-
|
| 184 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 185 |
|
| 186 |
if __name__ == '__main__':
|
| 187 |
-
|
| 188 |
-
start_server(video_transcription, port=7860, debug=True)
|
|
|
|
| 4 |
from transformers import WhisperProcessor, WhisperForConditionalGeneration
|
| 5 |
import requests
|
| 6 |
from bs4 import BeautifulSoup
|
| 7 |
+
import tempfile
|
| 8 |
import os
|
| 9 |
import soundfile as sf
|
| 10 |
from spellchecker import SpellChecker
|
|
|
|
| 12 |
import librosa
|
| 13 |
import numpy as np
|
| 14 |
from pyannote.audio import Pipeline
|
| 15 |
+
import dash
|
| 16 |
+
from dash import dcc, html, Input, Output, State
|
| 17 |
+
import dash_bootstrap_components as dbc
|
| 18 |
+
from dash.exceptions import PreventUpdate
|
| 19 |
+
import base64
|
| 20 |
+
import threading
|
| 21 |
|
| 22 |
# Initialize the speaker diarization pipeline
|
| 23 |
try:
|
|
|
|
| 174 |
print(error_message)
|
| 175 |
return error_message
|
| 176 |
|
| 177 |
+
app = dash.Dash(__name__, external_stylesheets=[dbc.themes.BOOTSTRAP])
|
| 178 |
+
|
| 179 |
+
app.layout = dbc.Container([
|
| 180 |
+
dbc.Row([
|
| 181 |
+
dbc.Col([
|
| 182 |
+
html.H1("Video Transcription", className="text-center mb-4"),
|
| 183 |
+
dbc.Card([
|
| 184 |
+
dbc.CardBody([
|
| 185 |
+
dbc.Input(id="video-url", type="text", placeholder="Enter video URL"),
|
| 186 |
+
dbc.Button("Transcribe", id="transcribe-button", color="primary", className="mt-3"),
|
| 187 |
+
dbc.Spinner(html.Div(id="transcription-output", className="mt-3")),
|
| 188 |
+
dcc.Download(id="download-transcript")
|
| 189 |
+
])
|
| 190 |
+
])
|
| 191 |
+
], width=12)
|
| 192 |
+
])
|
| 193 |
+
], fluid=True)
|
| 194 |
+
|
| 195 |
+
@app.callback(
|
| 196 |
+
Output("transcription-output", "children"),
|
| 197 |
+
Output("download-transcript", "data"),
|
| 198 |
+
Input("transcribe-button", "n_clicks"),
|
| 199 |
+
State("video-url", "value"),
|
| 200 |
+
prevent_initial_call=True
|
| 201 |
+
)
|
| 202 |
+
def update_transcription(n_clicks, url):
|
| 203 |
+
if not url:
|
| 204 |
+
raise PreventUpdate
|
| 205 |
+
|
| 206 |
+
def transcribe():
|
| 207 |
+
transcript = transcribe_video(url)
|
| 208 |
+
return transcript
|
| 209 |
+
|
| 210 |
+
# Run transcription in a separate thread
|
| 211 |
+
thread = threading.Thread(target=transcribe)
|
| 212 |
+
thread.start()
|
| 213 |
+
thread.join()
|
| 214 |
+
|
| 215 |
+
transcript = transcribe()
|
| 216 |
+
|
| 217 |
+
if transcript:
|
| 218 |
+
download_data = dict(content=transcript, filename="transcript.txt")
|
| 219 |
+
return dbc.Card([
|
| 220 |
+
dbc.CardBody([
|
| 221 |
+
html.H5("Transcription Result"),
|
| 222 |
+
html.Pre(transcript),
|
| 223 |
+
dbc.Button("Download Transcript", id="btn-download", color="secondary", className="mt-3")
|
| 224 |
+
])
|
| 225 |
+
]), download_data
|
| 226 |
+
else:
|
| 227 |
+
return "Failed to transcribe video.", None
|
| 228 |
|
| 229 |
if __name__ == '__main__':
|
| 230 |
+
app.run(debug=True, host='0.0.0.0', port=7860)
|
|
|