| import gradio as gr | |
| from pydub import AudioSegment | |
| from faster_whisper import WhisperModel | |
| import os | |
| model = WhisperModel("openai/whisper-large-v3-turbo", compute_type="int8") | |
| def convert_to_wav(file_path): | |
| audio = AudioSegment.from_file(file_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 | |
| def detect_language(audio_file): | |
| wav = convert_to_wav(audio_file.name) | |
| segments, info = model.transcribe(wav) | |
| transcript = "\n".join([seg.text for seg in segments]) | |
| return f"🌐 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." | |
| ).launch() | |