Update vision_llm.py

vision_llm.py

@@ -1,14 +1,14 @@
-# timeforge/vision_llm.py
+# vision_llm.py
 from transformers import AutoProcessor, AutoModelForVision2Seq
 import torch
 from PIL import Image
+
 class VisionLLM:
-    def __init__(self, device="cuda", model_id="microsoft/kosmos-2"):
+    def __init__(self, device="cuda", model_id="google/paligemma2-3b-pt-224"):  # Corrected model ID
         self.device = device
         self.processor = AutoProcessor.from_pretrained(model_id)
         self.model = AutoModelForVision2Seq.from_pretrained(model_id, torch_dtype=torch.float16).to(self.device)
 
-
     def describe_images(self, images, prompt="", max_length=128):
         if isinstance(images, list):
             images = [img.convert("RGB") for img in images]