Update README.md

README.md

@@ -35,7 +35,7 @@ RadVLM is a compact, multitask vision-language model designed for conversational
   - Bounding Boxes (if applicable): Coordinates indicating the location of anatomical structures or abnormalities.
 
 ## Model Architecture
-- Backbone: LLaVA-OneVision-7B (https://huggingface.co/llava-hf/llava-onevision-qwen2-7b-si-hf/tree/main), a vision-language model adapted for medical tasks.
+- Backbone: LLaVA-OneVision-7B (https://huggingface.co/llava-hf/llava-onevision-qwen2-7b-si-hf), a vision-language model adapted for medical tasks.
 - Vision Encoder: SigLIP, used for image feature extraction.
 - Instruction Tuning: Fine-tuned with multi-task objectives, covering report generation, abnormality detection, and multi-turn Q&A.