regre

app.py

@@ -20,7 +20,7 @@ import time
 from PIL import Image
 import torch
 import subprocess
-# subprocess.run('pip install flash-attn --no-build-isolation', env={'FLASH_ATTENTION_SKIP_CUDA_BUILD': "TRUE"}, shell=True)
+subprocess.run('pip install flash-attn --no-build-isolation', env={'FLASH_ATTENTION_SKIP_CUDA_BUILD': "TRUE"}, shell=True)
 
 
 
@@ -45,7 +45,7 @@ def model_inference(
     # )
 
     #We recommend enabling flash_attention_2 for better acceleration and memory saving, especially in multi-image and video scenarios.
-    model = Qwen2VLForConditionalGeneration.from_pretrained("Qwen/Qwen2-VL-2B-Instruct", trust_remote_code=True, torch_dtype="auto").cuda().eval()
+    model = Qwen2VLForConditionalGeneration.from_pretrained("Qwen/Qwen2-VL-2B-Instruct",attn_implementation="flash_attention_2", trust_remote_code=True, torch_dtype="auto").cuda().eval()
 
     # default processer
     processor = AutoProcessor.from_pretrained("Qwen/Qwen2-VL-7B-Instruct")