Spaces:

JimmyK300
/

Qwen2.5-Math-1.5B-Instruct

Sleeping

JimmyK300 commited on Feb 18, 2025

Commit

5528772

verified ·

1 Parent(s): 4cf9177

Update app.py

Files changed (1) hide show

app.py CHANGED Viewed

@@ -12,7 +12,7 @@ from PIL import Image
 vl_model = Qwen2VLForConditionalGeneration.from_pretrained(
     "Qwen/Qwen2-VL-2B-Instruct",
     torch_dtype=torch.bfloat16,
-    attn_implementation="flash_attention_2",
     device_map="auto",
 )
 vl_processor = AutoProcessor.from_pretrained("Qwen/Qwen2-VL-2B-Instruct")

 vl_model = Qwen2VLForConditionalGeneration.from_pretrained(
     "Qwen/Qwen2-VL-2B-Instruct",
     torch_dtype=torch.bfloat16,
+    attn_implementation="flash_attention",
     device_map="auto",
 )
 vl_processor = AutoProcessor.from_pretrained("Qwen/Qwen2-VL-2B-Instruct")