Update app.py

app.py

@@ -3,24 +3,26 @@ from pydub import AudioSegment
 from faster_whisper import WhisperModel
 import os
 
+# Load model from Hugging Face (it will download & cache automatically)
 model = WhisperModel("openai/whisper-large-v3-turbo", compute_type="int8")
 
-def convert_to_wav(file_path):
-    audio = AudioSegment.from_file(file_path)
+def convert_to_wav(input_path):
+    audio = AudioSegment.from_file(input_path)
     audio = audio.set_frame_rate(16000).set_channels(1).set_sample_width(2)
-    out_path = file_path.replace(".", "_ready.")
-    audio.export(out_path, format="wav")
-    return out_path
+    output_path = input_path.replace(".", "_ready.")
+    audio.export(output_path, format="wav")
+    return output_path
 
-def detect_language(audio_file):
-    wav = convert_to_wav(audio_file.name)
-    segments, info = model.transcribe(wav)
+def transcribe_and_detect_lang(file_path):
+    wav_path = convert_to_wav(file_path)
+    segments, info = model.transcribe(wav_path)
     transcript = "\n".join([seg.text for seg in segments])
-    return f"🌐 Language: {info.language}\n\n📝 Transcript:\n{transcript}"
+    return f"🌐 Detected Language: {info.language}\n\n📝 Transcript:\n{transcript}"
 
-gr.Interface(fn=detect_language,
-             inputs=gr.Audio(type="filepath"),
-             outputs="text",
-             title="🎧 Audio Language Detector",
-             description="Drop any MP3/WAV/FLAC to identify spoken language & get transcript."
+gr.Interface(
+    fn=transcribe_and_detect_lang,
+    inputs=gr.Audio(type="filepath", label="🎧 Upload Audio File"),
+    outputs=gr.Textbox(label="📋 Output"),
+    title="🌍 Whisper Language Identifier",
+    description="Upload an audio file in any language (Tamil, Hindi, English, etc.) and detect its language + get transcription."
 ).launch()