Upload MedVLThinker-7B-SFT_PMC model weights

README.md

@@ -19,6 +19,7 @@ pipeline_tag: image-text-to-text
 # MedVLThinker-7B-SFT_PMC
 
 Code: https://github.com/UCSC-VLAA/MedVLThinker
+Project Page: https://ucsc-vlaa.github.io/MedVLThinker/
 
 ## Model Description
 
@@ -34,13 +35,15 @@ This model has been trained using supervised fine-tuning on PMC-VQA dataset.
 
 ## Usage
 
+Check here for demo images: https://github.com/UCSC-VLAA/MedVLThinker?tab=readme-ov-file#demo
+
 ```python
-from transformers import Qwen2VLForConditionalGeneration, AutoTokenizer, AutoProcessor
+from transformers import Qwen2_5_VLForConditionalGeneration, AutoTokenizer, AutoProcessor
 from qwen_vl_utils import process_vision_info
 import torch
 
 # Load the model
-model = Qwen2VLForConditionalGeneration.from_pretrained(
+model = Qwen2_5_VLForConditionalGeneration.from_pretrained(
     "UCSC-VLAA/MedVLThinker-7B-SFT_PMC",
     torch_dtype=torch.bfloat16,
     device_map="auto"
@@ -49,6 +52,10 @@ processor = AutoProcessor.from_pretrained("UCSC-VLAA/MedVLThinker-7B-SFT_PMC")
 
 # Example usage
 messages = [
+    {
+        "role": "system",
+        "content": "You will solve a problem/request. You should provide your thoughts within <think> </think> tags before providing the answer. Write your final answer within <answer> </answer> tags.",
+    },
     {
         "role": "user",
         "content": [
@@ -76,7 +83,7 @@ inputs = processor(
 inputs = inputs.to("cuda")
 
 # Inference
-generated_ids = model.generate(**inputs, max_new_tokens=128)
+generated_ids = model.generate(**inputs, max_new_tokens=2048, temperature=0.6, top_p=0.95, do_sample=True)
 generated_ids_trimmed = [
     out_ids[len(in_ids) :] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)
 ]