Update app.py

app.py

@@ -99,14 +99,16 @@ tokenizer.pad_token = tokenizer.unk_token
 def predict_event(image):
     
 
-    img =  Image.open(image).convert("RGB")
+    # img =  Image.open(image).convert("RGB")
 
-    generated_caption = tokenizer.decode(model.generate(feature_extractor(img, return_tensors="pt").pixel_values.to("cuda"))[0])
+    generated_caption = tokenizer.decode(model.generate(feature_extractor(image, return_tensors="pt").pixel_values.to("cuda"))[0])
 
     # caption_vitgpt = generate_caption(model, image)
     #caption_vitgpt = generate_caption(feature_extractor, cap_model, image, tokenizer)
+    preds = [pred.strip() for pred in generated_caption]
+    return preds
 
-    return '\033[96m' +generated_caption[:85]+ '\033[0m'
+    #return '\033[96m' +generated_caption[:85]+ '\033[0m'
 
 
 
@@ -125,7 +127,7 @@ description = "final capstone"
 # inputs=gr.inputs.Image(type="pil")
 
 iface = gr.Interface(predict_event,
-                         inputs=["text"],
+                         inputs=["image"],
                          # gr.Image(type="pil"),
                          outputs=["text"] )