Update app.py

app.py

@@ -19,13 +19,13 @@ import subprocess
 subprocess.run('pip install flash-attn --no-build-isolation', env={'FLASH_ATTENTION_SKIP_CUDA_BUILD': "TRUE"}, shell=True)
 
 model = Qwen2_5_VLForConditionalGeneration.from_pretrained(
-    "lingshu-medical-mllm/Lingshu-32B",
+    "lingshu-medical-mllm/Lingshu-7B",
     torch_dtype=torch.bfloat16,
     attn_implementation="flash_attention_2",
     device_map="auto",
 )
 
-processor = AutoProcessor.from_pretrained("lingshu-medical-mllm/Lingshu-32B")
+processor = AutoProcessor.from_pretrained("lingshu-medical-mllm/Lingshu-7B")
 
 
 MAX_NUM_IMAGES = int(os.getenv("MAX_NUM_IMAGES", "5"))
@@ -216,7 +216,7 @@ def run(message: dict, history: list[dict], system_prompt: str = "", max_new_tok
 
 
 DESCRIPTION = """\
-This is a demo of Lingshu 32B, a multimodal model trained for performance on medical text and image comprehension.
+This is a demo of Lingshu 7B, a multimodal model trained for performance on medical text and image comprehension.
 Lingshu supports more than 12 medical imaging modalities, including X-Ray, CT Scan, MRI, Microscopy, Ultrasound, Histopathology, Dermoscopy, Fundus, OCT, Digital Photography, Endoscopy, and PET.
 """
 
@@ -231,7 +231,7 @@ demo = gr.ChatInterface(
         gr.Slider(label="Max New Tokens", minimum=100, maximum=8192, step=10, value=2048),
     ],
     stop_btn=False,
-    title="Lingshu 32B",
+    title="Lingshu 7B",
     description=DESCRIPTION,
     run_examples_on_click=False,
     cache_examples=False,