Update app.py

app.py

@@ -5,7 +5,7 @@ import spaces
 from transformers import pipeline
 
 BASE_MODEL_ID = "HuggingFaceTB/SmolVLM2-500M-Video-Instruct"
-FINE_TUNED_MODEL_ID = "ninjals/FoodExtract-Vision-SmolVLM2-500M-fine-tune-v1-VIDEO"
+FINE_TUNED_MODEL_ID = "ninjals/FoodExtract-Vision-SmolVLM2-500M-fine-tune-v1"
 OUTPUT_TOKENS = 256
 
 # Load original base model (no fine-tuning)
@@ -54,7 +54,7 @@ def extract_foods_from_image(input_image):
 demo_title = "🥑➡️📝 FoodExtract-Vision with a fine-tuned SmolVLM2-500M"
 demo_description = """* **Base model:** https://huggingface.co/HuggingFaceTB/SmolVLM-500M-Instruct
 * **Fine-tuning dataset:** https://huggingface.co/datasets/mrdbourke/FoodExtract-1k-Vision (1k food images and 500 not food images)
-* **Fine-tuned model:** https://huggingface.co/ninjals/FoodExtract-Vision-SmolVLM2-500M-fine-tune-v1-VIDEO
+* **Fine-tuned model:** https://huggingface.co/ninjals/FoodExtract-Vision-SmolVLM2-500M-fine-tune-v1
 ## Overview
 Extract food and drink items in a structured way from images.
 The original model outputs fail to capture the desired structure. But the fine-tuned model sticks to the output structure quite well.