Update app.py

app.py

@@ -13,8 +13,14 @@ from pptx import Presentation
 import subprocess
 import shlex
 
-# Load the Whisper model (smaller model for faster transcription)
-model = whisper.load_model("tiny")
+# Define available Whisper models
+whisper_models = {
+    "Tiny (Fast, Less Accurate)": "tiny",
+    "Base (Faster, Moderate Accuracy)": "base",
+    "Small (Moderate Speed, Good Accuracy)": "small",
+    "Medium (Slower, High Accuracy)": "medium",
+    "Large (Slow, Very High Accuracy)": "large",
+}
 
 # Load M2M100 translation model for different languages
 def load_translation_model(target_language):
@@ -178,7 +184,10 @@ def write_ppt(transcription, output_file, tokenizer=None, translation_model=None
     ppt.save(output_file)
 
 # Transcribing video and generating output
-def transcribe_video(video_file, language, target_language, output_format):
+def transcribe_video(video_file, language, target_language, output_format, model_name):
+    actual_model_name = whisper_models[model_name]  # Map user selection to model name
+    model = whisper.load_model(actual_model_name)  # Load the selected model
+
     if video_file is not None:  # Ensure the video_file is not None
         video_file_path = video_file.name
     else:
@@ -218,15 +227,17 @@ def transcribe_video(video_file, language, target_language, output_format):
         ppt_file = f"{video_name}.pptx"
         write_ppt(result, ppt_file, tokenizer, translation_model)
         return ppt_file
+    else:
+        raise ValueError("Invalid output format selected.")
 
-
-# Gradio interface without YouTube URL
+# Gradio interface
 iface = gr.Interface(
     fn=transcribe_video,
     inputs=[
-        gr.File(label="Upload Video File"),  
+        gr.File(label="Upload Video File"),
         gr.Dropdown(label="Select Original Video Language", choices=["en", "es", "fr", "de", "it", "pt"], value="en"),
         gr.Dropdown(label="Select Subtitle Translation Language", choices=["en", "fa", "es", "de", "fr", "it", "pt"], value="fa"),
+        gr.Dropdown(label="Select Whisper Model", choices=list(whisper_models.keys()), value="Tiny (Fast, Less Accurate)"),
         gr.Radio(label="Choose Output Format", choices=["SRT", "Video with Hardsub", "Word", "PDF", "PowerPoint"], value="Video with Hardsub")
     ],
     outputs=gr.File(label="Download File"),