Update app.py

app.py

@@ -6,7 +6,7 @@ import subprocess
 from faster_whisper import WhisperModel
 
 # ===============================
-# 1. Whisper Model (Lazy Load)
+# 1. Whisper Model
 # ===============================
 model = None
 
@@ -26,7 +26,7 @@ def get_ffmpeg_path():
     return path if path else "/usr/bin/ffmpeg"
 
 # ===============================
-# 3. Convert Video → Audio
+# 3. Extract Audio
 # ===============================
 def extract_audio(video_path):
     audio_path = "uploaded_audio.wav"
@@ -46,7 +46,7 @@ def extract_audio(video_path):
     return audio_path
 
 # ===============================
-# 4. Download Audio from ANY URL
+# 4. Download Audio
 # ===============================
 def download_audio_from_url(url):
     output = "url_audio.%(ext)s"
@@ -58,7 +58,6 @@ def download_audio_from_url(url):
         "postprocessors": [{
             "key": "FFmpegExtractAudio",
             "preferredcodec": "wav",
-            "preferredquality": "192",
         }],
         "quiet": True,
         "nocheckcertificate": True,
@@ -70,17 +69,16 @@ def download_audio_from_url(url):
     return "url_audio.wav"
 
 # ===============================
-# 5. Main Transcribe Logic
+# 5. Transcribe Function
 # ===============================
-def transcribe_media(url_input, file_input):
+def transcribe_media(url_input, file_input, language_choice):
 
     try:
         audio_path = None
 
-        # ---------- FILE UPLOAD ----------
+        # ---------- FILE ----------
         if file_input:
             ext = os.path.splitext(file_input)[1].lower()
-
             if ext in [".mp3", ".wav", ".m4a"]:
                 audio_path = file_input
             else:
@@ -91,21 +89,28 @@ def transcribe_media(url_input, file_input):
             audio_path = download_audio_from_url(url_input)
 
         else:
-            return "⚠️ Please paste a link or upload a file."
+            return "⚠️ Please provide URL or Upload file."
 
         if not os.path.exists(audio_path):
             return "❌ Audio processing failed."
 
         model = load_model()
 
-        segments, _ = model.transcribe(
+        # Language Handling
+        language = None if language_choice == "Auto Detect" else language_choice
+
+        segments, info = model.transcribe(
             audio_path,
             beam_size=1,
-            vad_filter=True
+            vad_filter=True,
+            language=language
         )
 
+        detected_lang = info.language
+
         text = " ".join(seg.text for seg in segments)
-        return text.strip() if text else "⚠️ No speech detected."
+
+        return f"🌍 Detected Language: {detected_lang}\n\n{text.strip()}"
 
     except Exception as e:
         return f"❌ Error: {str(e)}"
@@ -116,7 +121,7 @@ def transcribe_media(url_input, file_input):
 css = """
 .container {max-width: 900px; margin: auto;}
 .gr-button-primary {
-    background: linear-gradient(90deg,#ff416c,#ff4b2b);
+    background: linear-gradient(90deg,#667eea,#764ba2);
     border: none;
     color: white;
 }
@@ -124,18 +129,12 @@ css = """
 
 with gr.Blocks(theme=gr.themes.Soft(), css=css) as demo:
     with gr.Column(elem_classes="container"):
-        gr.Markdown("## 🚀 Universal Video Transcript Tool")
-        gr.Markdown(
-            "Supports **YouTube, TikTok, Instagram, Facebook, Twitter/X**\n\n"
-            "**OR** upload video/audio file."
-        )
+        gr.Markdown("## 🚀 Universal Transcript Tool")
+        gr.Markdown("Supports YouTube, TikTok, Instagram, Facebook, Twitter/X")
 
         with gr.Tabs():
             with gr.TabItem("🔗 Paste Link"):
-                url_in = gr.Textbox(
-                    label="Video URL",
-                    placeholder="https://youtube.com / tiktok.com / instagram.com"
-                )
+                url_in = gr.Textbox(label="Video URL")
                 btn_url = gr.Button("🎧 Transcribe Link", variant="primary")
 
             with gr.TabItem("📂 Upload File"):
@@ -145,9 +144,28 @@ with gr.Blocks(theme=gr.themes.Soft(), css=css) as demo:
                 )
                 btn_file = gr.Button("📂 Transcribe File", variant="primary")
 
-        output = gr.Code(label="Transcript Output", language="markdown", lines=15)
+        # 🌍 Language Selector
+        language_selector = gr.Dropdown(
+            choices=[
+                "Auto Detect",
+                "en",  # English
+                "hi",  # Hindi
+                "ur",  # Urdu
+                "ar",  # Arabic
+                "fr",  # French
+                "de",  # German
+                "es",  # Spanish
+                "ru",  # Russian
+                "ja",  # Japanese
+                "zh"   # Chinese
+            ],
+            value="Auto Detect",
+            label="🌍 Select Transcript Language"
+        )
+
+        output = gr.Code(label="Transcript Output", lines=15)
 
-    btn_url.click(transcribe_media, [url_in, gr.State(None)], output)
-    btn_file.click(transcribe_media, [gr.State(None), file_in], output)
+    btn_url.click(transcribe_media, [url_in, gr.State(None), language_selector], output)
+    btn_file.click(transcribe_media, [gr.State(None), file_in, language_selector], output)
 
 demo.launch()