Create app.py

app.py

@@ -0,0 +1,100 @@
+import whisper
+import gradio as gr
+import tempfile
+import os
+from pathlib import Path
+
+MODEL_SIZE = "small"  # Change to "base", "small", "medium", or "large"
+
+class GradioSRTGenerator:
+    def __init__(self):
+        print(f"Loading Whisper model: {MODEL_SIZE}")
+        self.model = whisper.load_model(MODEL_SIZE)
+        print("Model loaded successfully")
+
+    @staticmethod
+    def format_time(seconds: float) -> str:
+        """Convert seconds to SRT time format (HH:MM:SS,mmm)"""
+        hours = int(seconds // 3600)
+        minutes = int((seconds % 3600) // 60)
+        secs = int(seconds % 60)
+        millis = int((seconds - int(seconds)) * 1000)
+        return f"{hours:02d}:{minutes:02d}:{secs:02d},{millis:03d}"
+
+    def build_srt_file(self, segments, srt_path: str):
+        """Write Whisper segments into an SRT file at srt_path."""
+        with open(srt_path, "w", encoding="utf-8") as srt_file:
+            for i, seg in enumerate(segments, start=1):
+                start_time = self.format_time(seg["start"])
+                end_time = self.format_time(seg["end"])
+                text = seg["text"].strip()
+                srt_file.write(f"{i}\n")
+                srt_file.write(f"{start_time} --> {end_time}\n")
+                srt_file.write(f"{text}\n\n")
+
+    def generate_srt(self, video_path, progress=gr.Progress()):
+        """
+        Transcribe the uploaded video and produce an SRT file.
+        Returns: (path_to_srt, status_message)
+        """
+        if not video_path:
+            return None, "⚠️ Please upload a video first."
+
+        try:
+            progress(0.1, desc="Transcribing audio with Whisper...")
+            result = self.model.transcribe(video_path, task="translate")
+            progress(0.6, desc="Building SRT file...")
+
+            # Create a temporary file to hold the SRT
+            with tempfile.NamedTemporaryFile(delete=False, suffix=".srt", mode="w", encoding="utf-8") as tmp:
+                srt_path = tmp.name
+
+            # Write segments into that SRT
+            self.build_srt_file(result["segments"], srt_path)
+            progress(1.0, desc="✅ SRT ready!")
+
+            return srt_path, "✅ Transcription complete. Download your SRT below."
+
+        except Exception as e:
+            return None, f"❌ Error during transcription: {str(e)}"
+
+
+def create_ui():
+    generator = GradioSRTGenerator()
+
+    with gr.Blocks(theme=gr.themes.Base(primary_hue="blue", secondary_hue="indigo")) as app:
+        gr.Markdown(
+            """
+            # 📝 Video → SRT Generator (Queued)
+
+            Only one transcription job will run at a time; additional users will be placed in a queue.
+            Upload any supported video file (MP4, MOV, AVI, MKV, etc.) and click **Generate SRT**. 
+            Whisper will transcribe and produce an SRT subtitle file you can download immediately.
+            """
+        )
+
+        with gr.Row():
+            with gr.Column():
+                input_video = gr.Video(label="Upload Video", source="upload", interactive=True)
+                generate_btn = gr.Button("🎯 Generate SRT", variant="primary")
+                status_text = gr.Textbox(label="Status", interactive=False, show_copy_button=True)
+
+            with gr.Column():
+                srt_download = gr.File(label="Download SRT", interactive=False, visible=False)
+
+        # Bind the button and wrap the handler in .queue()
+        generate_btn.click(
+            fn=generator.generate_srt,
+            inputs=[input_video],
+            outputs=[srt_download, status_text],
+            api_name="generate_srt"
+        ).queue()  # <-- ensures this function is queued if another request is running
+
+    # Add a queue to the entire Blocks app with exactly 1 worker
+    app.queue(concurrency_count=1, max_size=10)
+    return app
+
+app = create_ui()
+
+if __name__ == "__main__":
+    app.launch()