Update app.py

app.py

@@ -6,52 +6,50 @@ from openai import OpenAI
 # 🔑 Load OpenAI API key from Hugging Face Space secrets
 client = OpenAI(api_key=os.environ.get("OPENAI_API_KEY"))
 
-# 🎧 Load Whisper model once (small = faster; use "base" for lighter)
+# 🎧 Load Whisper model once
 whisper_model = whisper.load_model("small")
 
-def transcribe_and_translate(audio_path):
+def audio_prompt_to_chatgpt(audio_path):
     if audio_path is None:
-        return "No audio detected.", "No translation."
+        return "No audio detected.", ""
 
-    # Step 1: Transcribe audio → Chinese text
+    # 1️⃣ Transcribe audio → Chinese text
     result = whisper_model.transcribe(audio_path, language="zh")
     chinese_text = result.get("text", "").strip()
 
     if not chinese_text:
-        return "⚠️ Could not transcribe audio.", "No translation."
-
-    # Step 2: Send transcription to ChatGPT for English translation
-    messages = [
-        {"role": "system", "content": "You are a professional translator. Translate the following Chinese into fluent, natural English."},
-        {"role": "user", "content": chinese_text}
-    ]
+        return "⚠️ Could not transcribe audio.", ""
 
+    # 2️⃣ Send transcription as a prompt to ChatGPT
     try:
         response = client.chat.completions.create(
-            model="gpt-4o-mini",  # can use "gpt-4o" for higher quality
-            messages=messages,
-            temperature=0
+            model="gpt-4o-mini",  # or gpt-4o for higher quality
+            messages=[
+                {"role": "system", "content": "You are a helpful assistant. Follow the user's instructions."},
+                {"role": "user", "content": chinese_text}
+            ],
+            temperature=0.7
         )
-        english_text = response.choices[0].message.content.strip()
+        chatgpt_reply = response.choices[0].message.content.strip()
     except Exception as e:
-        english_text = f"⚠️ Translation failed: {e}"
+        chatgpt_reply = f"⚠️ ChatGPT failed: {e}"
 
-    return chinese_text, english_text
+    return chinese_text, chatgpt_reply
 
 # 🚀 Gradio UI
 app = gr.Interface(
-    fn=transcribe_and_translate,
+    fn=audio_prompt_to_chatgpt,
     inputs=gr.Audio(
         sources=["microphone", "upload"],
         type="filepath",
         label="🎙️ Speak Chinese or Upload Audio"
     ),
     outputs=[
-        gr.Textbox(label="🈶 Chinese Transcription"),
-        gr.Textbox(label="🌍 English Translation (ChatGPT)")
+        gr.Textbox(label="🈶 Transcribed Chinese Prompt"),
+        gr.Textbox(label="🤖 ChatGPT Response")
     ],
-    title="🎤 Chinese Voice → English Translator",
-    description="Speak Chinese or upload audio. Whisper transcribes, then ChatGPT translates into English."
+    title="🎤 Chinese Voice → ChatGPT Prompt",
+    description="Speak Chinese or upload audio. Whisper transcribes, then ChatGPT executes your instructions."
 )
 
 if __name__ == "__main__":