Create app.py

app.py

@@ -0,0 +1,61 @@
+# 🗣️ Audio Translator | CPU-only HF Space
+
+import tempfile
+from deep_translator import GoogleTranslator
+from gtts import gTTS
+import gradio as gr
+from transformers import pipeline
+
+# 1. ASR pipeline (Whisper Tiny)
+asr = pipeline(
+    "automatic-speech-recognition",
+    model="openai/whisper-tiny",
+    device=-1,
+    chunk_length_s=30
+)
+
+def audio_translate(audio_path: str, target_lang: str):
+    if not audio_path:
+        return "", "", None
+
+    # 2. Transcribe
+    result = asr(audio_path)
+    transcript = result["text"].strip()
+
+    # 3. Translate
+    dest_code = "es" if target_lang == "Spanish" else "en"
+    translated = GoogleTranslator(source="auto", target=dest_code).translate(transcript)
+
+    # 4. Text-to-speech
+    tts = gTTS(translated, lang=dest_code)
+    tmp = tempfile.NamedTemporaryFile(suffix=".mp3", delete=False)
+    tts.write_to_fp(tmp)
+    tmp.flush()
+
+    return transcript, translated, tmp.name
+
+# 5. Gradio UI
+with gr.Blocks(title="🗣️ Audio Translator") as demo:
+    gr.Markdown(
+        "# 🗣️ Audio Translator\n"
+        "Upload an audio clip (any language), choose **English** or **Spanish**,\n"
+        "and get an immediate translated transcript and spoken audio."
+    )
+
+    with gr.Row():
+        audio_in   = gr.Audio(source="upload", type="filepath", label="Upload Audio")
+        lang_sel   = gr.Radio(["Spanish", "English"], value="Spanish", label="Translate to")
+    translate_btn = gr.Button("Translate 🔄", variant="primary")
+
+    orig_txt  = gr.Textbox(label="Transcription", interactive=False)
+    trans_txt = gr.Textbox(label="Translation", interactive=False)
+    audio_out = gr.Audio(label="Translated Speech", type="filepath")
+
+    translate_btn.click(
+        audio_translate,
+        inputs=[audio_in, lang_sel],
+        outputs=[orig_txt, trans_txt, audio_out]
+    )
+
+if __name__ == "__main__":
+    demo.launch(server_name="0.0.0.0")