fixed example caching

app.py

@@ -14,7 +14,7 @@ def process_image_and_generate_output(image, model_selection):
         return "Please select an image", None
     # (Trained only for 15 epochs without any hyperparameter tuning, utilizing inception v3)'
     #                            (SOTA model for Image captioning)
-    if model_selection == ('Basic Model'):
+    if model_selection == 'Basic Model':
         result = inference_script.evaluate(image)
         pred_caption = ' '.join(result).rsplit(' ', 1)[0]
         pred_caption = pred_caption.replace('<unk>', '')
@@ -39,22 +39,13 @@ def process_image_and_generate_output(image, model_selection):
     return pred_caption, audio_content
 
 
-# sample_images = [
-#     [os.path.join(os.path.dirname(__file__), "sample_images/1.jpg"), 'Basic Model'],
-#     [os.path.join(os.path.dirname(__file__), "sample_images/2.jpg"), 'Basic Model'],
-#     [os.path.join(os.path.dirname(__file__), "sample_images/3.jpg"), 'Basic Model'],
-#     [os.path.join(os.path.dirname(__file__), "sample_images/4.jpg"), "ViT-GPT2"],
-#     [os.path.join(os.path.dirname(__file__), "sample_images/5.jpg"), "ViT-GPT2"],
-#     [os.path.join(os.path.dirname(__file__), "sample_images/6.jpg"), "ViT-GPT2"]
-# ]
-
 sample_images = [
-    ["sample_images/1.jpg"],
-    ["sample_images/2.jpg"],
-    ["sample_images/3.jpg"],
-    ["sample_images/4.jpg"],
-    ["sample_images/5.jpg"],
-    ["sample_images/6.jpg"]
+    [os.path.join(os.path.dirname(__file__), "sample_images/1.jpg"), "ViT-GPT2"],
+    [os.path.join(os.path.dirname(__file__), "sample_images/1.jpg"), 'Basic Model'],
+    [os.path.join(os.path.dirname(__file__), "sample_images/2.jpg"), "ViT-GPT2"],
+    [os.path.join(os.path.dirname(__file__), "sample_images/2.jpg"), 'Basic Model'],
+    [os.path.join(os.path.dirname(__file__), "sample_images/3.jpg"), "ViT-GPT2"],
+    [os.path.join(os.path.dirname(__file__), "sample_images/3.jpg"), 'Basic Model']
 ]
 
 
@@ -70,7 +61,7 @@ iface = gr.Interface(fn=process_image_and_generate_output,
                      inputs=[image_input, model_selection_input],
                      outputs=["text", "audio"],
                      examples=sample_images,
-                     cache_examples=False,
+                     cache_examples=True,
                      allow_flagging='never',
                      title="Eye For Blind | Image Captioning & TTS",
                      description="To be added")