Update app.py

app.py

@@ -3,17 +3,18 @@ import torch
 from transformers import AutoProcessor, AutoModelForCausalLM
 from huggingface_hub import hf_hub_download
 from PIL import Image
+from datasets import load_dataset
 
 processor = AutoProcessor.from_pretrained("microsoft/git-base-vqav2")
-model = AutoModelForCausalLM.from_pretrained("microsoft/git-base-vqav2")
+model = AutoModelForCausalLM.from_pretrained("Multimodal-Fatima/OK-VQA_train")
 
-file_path = hf_hub_download(repo_id="Multimodal-Fatima/OK-VQA_train", filename="data", repo_type="dataset")
+dataset = load_dataset("question","answers","image")
+
+file_path = hf_hub_download(repo_id="Multimodal-Fatima/OK-VQA_train", repo_type="dataset")
 image = Image.open(file_path).convert("RGB")
 
 pixel_values = processor(images=image, return_tensors="pt").pixel_values
 
-question = "How many people are there?"
-
 input_ids = processor(text=question, add_special_tokens=False).input_ids
 input_ids = [processor.tokenizer.cls_token_id] + input_ids
 input_ids = torch.tensor(input_ids).unsqueeze(0)