Update app.py

app.py

@@ -3,33 +3,37 @@ import os
 import tempfile
 import yt_dlp
 import subprocess
+import requests
 from huggingface_hub import InferenceClient
 
-# ✅ Initialize clients
-whisper_client = InferenceClient("openai/whisper-large-v3")
-english_summarizer = InferenceClient("facebook/bart-large-cnn")
-urdu_summarizer = InferenceClient("openai/gpt-oss-120b")
+# ----------------------------
+# Initialize Hugging Face clients
+# ----------------------------
+WHISPER_MODEL = "openai/whisper-large-v3"
+EN_SUMMARY_MODEL = "facebook/bart-large-cnn"
+UR_SUMMARY_MODEL = "openai/gpt-oss-120b"
 
-# --- Helper Functions ---
+client = InferenceClient()
 
-def download_youtube_audio(url):
-    """Download YouTube audio using yt_dlp"""
+# ----------------------------
+# Helpers
+# ----------------------------
+def download_youtube_audio(url: str):
     try:
-        with tempfile.NamedTemporaryFile(delete=False, suffix=".mp3") as tmp:
-            ydl_opts = {
-                "format": "bestaudio/best",
-                "outtmpl": tmp.name,
-                "quiet": True,
-                "postprocessors": [{"key": "FFmpegExtractAudio", "preferredcodec": "mp3"}],
-            }
-            with yt_dlp.YoutubeDL(ydl_opts) as ydl:
-                ydl.download([url])
-            return tmp.name
+        tmp = tempfile.NamedTemporaryFile(delete=False, suffix=".mp3")
+        ydl_opts = {
+            "format": "bestaudio/best",
+            "outtmpl": tmp.name,
+            "quiet": True,
+            "postprocessors": [{"key": "FFmpegExtractAudio", "preferredcodec": "mp3"}],
+        }
+        with yt_dlp.YoutubeDL(ydl_opts) as ydl:
+            ydl.download([url])
+        return tmp.name
     except Exception as e:
-        raise RuntimeError(f"❌ Error downloading YouTube audio: {str(e)}")
+        raise RuntimeError(f"❌ Error downloading YouTube audio: {e}")
 
-def convert_to_wav(audio_path):
-    """Ensure audio is in .wav format"""
+def convert_to_wav(audio_path: str):
     wav_path = tempfile.mktemp(suffix=".wav")
     try:
         subprocess.run(
@@ -41,81 +45,97 @@ def convert_to_wav(audio_path):
     except subprocess.CalledProcessError as e:
         raise RuntimeError(f"❌ ffmpeg conversion failed: {e}")
 
+# ----------------------------
+# Transcription
+# ----------------------------
 def transcribe_audio(audio_file=None, youtube_url=None):
-    """Transcribe uploaded or YouTube audio"""
     try:
-        if youtube_url:
+        if youtube_url and youtube_url.strip():
             audio_path = download_youtube_audio(youtube_url)
         else:
             audio_path = audio_file
+        if not audio_path:
+            return "❌ Please upload or record audio or provide a YouTube link."
 
         wav_path = convert_to_wav(audio_path)
 
-        with open(wav_path, "rb") as f:
-            audio_data = f.read()
-
-        # ✅ Correct way to call Whisper for transcription
-        response = whisper_client.post(json=None, data=audio_data)
-        return response.get("text", "❌ No transcription returned.")
+        # --- Try API helper first ---
+        try:
+            result = client.audio_to_text(model=WHISPER_MODEL, audio=wav_path)
+            return result["text"] if isinstance(result, dict) else str(result)
+        except Exception:
+            # --- fallback to HTTP call ---
+            api_url = f"https://api-inference.huggingface.co/models/{WHISPER_MODEL}"
+            headers = {"Authorization": f"Bearer {os.environ.get('HUGGINGFACE_API_TOKEN','')}"}
+            with open(wav_path, "rb") as f:
+                resp = requests.post(api_url, headers=headers, data=f)
+            if resp.status_code != 200:
+                raise RuntimeError(f"HF API error: {resp.text}")
+            data = resp.json()
+            if isinstance(data, list) and len(data) and "text" in data[0]:
+                return data[0]["text"]
+            return str(data)
     except Exception as e:
         return f"❌ Error during transcription: {e}"
 
-def summarize_text(transcribed_text, language):
-    """Generate summary in English or Urdu"""
+# ----------------------------
+# Summarization
+# ----------------------------
+def summarize_text(text, language):
     try:
         if language == "English":
-            response = english_summarizer.text_generation(
-                prompt=f"Summarize the following text in detail:\n\n{transcribed_text}",
-                max_new_tokens=1024,
-            )
-            return response
+            result = client.summarization(model=EN_SUMMARY_MODEL, inputs=text, max_new_tokens=1024)
+            return result["summary_text"]
         else:
-            response = urdu_summarizer.text_generation(
-                prompt=f"مندرجہ ذیل انگریزی متن کا جامع اردو خلاصہ لکھیں:\n\n{transcribed_text}",
+            # Urdu summary via large model
+            resp = client.text_generation(
+                model=UR_SUMMARY_MODEL,
+                prompt=f"مندرجہ ذیل انگریزی متن کا جامع اردو خلاصہ لکھیں:\n\n{text}",
                 max_new_tokens=2048,
             )
-            return response
+            return resp
     except Exception as e:
         return f"❌ Summarization failed: {e}"
 
-def generate_tutorial(transcribed_text, summary, language):
-    """Craft a beginner-friendly tutorial in Urdu or English"""
+# ----------------------------
+# Tutorial Creator
+# ----------------------------
+def generate_tutorial(transcription, summary, language):
     try:
-        model = urdu_summarizer if language == "Urdu" else english_summarizer
         prompt = (
-            f"Write a simple, step-by-step tutorial in {language} for absolute beginners "
-            f"based on the following transcript and summary.\n\nTranscript:\n{transcribed_text}\n\nSummary:\n{summary}"
+            f"Write a detailed, beginner-friendly tutorial in {language} based on the following transcription and summary.\n\n"
+            f"Transcription:\n{transcription}\n\nSummary:\n{summary}"
         )
-        response = model.text_generation(prompt=prompt, max_new_tokens=2500)
-        return response
+        model = UR_SUMMARY_MODEL if language == "Urdu" else EN_SUMMARY_MODEL
+        result = client.text_generation(model=model, prompt=prompt, max_new_tokens=2200)
+        return result
     except Exception as e:
         return f"❌ Error generating tutorial: {e}"
 
-# --- Gradio UI ---
+# ----------------------------
+# Gradio Interface
+# ----------------------------
+with gr.Blocks(title="🎙️ Smart Transcriber & Tutorial Maker") as demo:
+    gr.Markdown("## 🎧 Smart Transcriber & Tutorial Generator")
 
-with gr.Blocks(title="🎙️ AI Audio Transcriber & Tutorial Maker") as demo:
-    gr.Markdown("## 🎧 AI Audio Transcriber, Summarizer & Tutorial Creator")
+    with gr.Row():
+        audio_in = gr.Audio(sources=["microphone", "upload"], type="filepath", label="🎙️ Record / Upload Audio")
+        yt_link = gr.Textbox(label="🎥 Or paste YouTube link")
 
-    with gr.Tab("🎤 Record / Upload Audio"):
-        with gr.Row():
-            audio_input = gr.Audio(sources=["microphone", "upload"], type="filepath", label="🎙️ Record or Upload Audio")
-            youtube_input = gr.Textbox(label="🎥 Or paste YouTube link")
-        transcribed_output = gr.Textbox(label="📝 Transcription", lines=8)
-        transcribe_btn = gr.Button("🚀 Transcribe Audio")
+    trans_btn = gr.Button("🚀 Transcribe")
+    transcription = gr.Textbox(label="📝 Transcription", lines=8)
 
-    with gr.Tab("🧠 Summarize"):
-        language_choice = gr.Radio(["English", "Urdu"], label="Choose Summary Language", value="English")
-        summary_output = gr.Textbox(label="📋 Detailed Summary", lines=10)
-        summarize_btn = gr.Button("✨ Generate Summary")
+    with gr.Row():
+        lang = gr.Radio(["English", "Urdu"], value="English", label="Select Summary Language")
+    sum_btn = gr.Button("🧠 Generate Detailed Summary")
+    summary_box = gr.Textbox(label="📋 Summary", lines=10)
 
-    with gr.Tab("📘 Tutorial Creator"):
-        tutorial_output = gr.Textbox(label="🎓 Beginner Tutorial", lines=12)
-        tutorial_btn = gr.Button("📚 Create Tutorial from Summary")
+    tut_btn = gr.Button("📘 Create Beginner Tutorial")
+    tutorial_box = gr.Textbox(label="🎓 Tutorial", lines=12)
 
-    # --- Button Actions ---
-    transcribe_btn.click(fn=transcribe_audio, inputs=[audio_input, youtube_input], outputs=transcribed_output)
-    summarize_btn.click(fn=summarize_text, inputs=[transcribed_output, language_choice], outputs=summary_output)
-    tutorial_btn.click(fn=generate_tutorial, inputs=[transcribed_output, summary_output, language_choice], outputs=tutorial_output)
+    # Actions
+    trans_btn.click(transcribe_audio, [audio_in, yt_link], transcription)
+    sum_btn.click(summarize_text, [transcription, lang], summary_box)
+    tut_btn.click(generate_tutorial, [transcription, summary_box, lang], tutorial_box)
 
-# Launch
 demo.launch()