Update app.py

app.py

@@ -7,8 +7,7 @@ model = AutoModelForCausalLM.from_pretrained(
     "microsoft/phi-3-mini-4k-instruct",
     device_map="cpu",
     torch_dtype="auto",
-    trust_remote_code=True,
-    attn_implementation='eager',
+    trust_remote_code=True
 )
 tokenizer = AutoTokenizer.from_pretrained("microsoft/phi-3-mini-4k-instruct")
 
@@ -33,8 +32,7 @@ pipe = pipeline(
     tokenizer="microsoft/phi-3-mini-4k-instruct",
     device_map="auto",  # Let Transformers automatically choose the best device
     torch_dtype="float16",  # Use half-precision for faster inference
-    trust_remote_code=True,
-    attn_implementation='eager',
+    trust_remote_code=True
 )
 
 # Predefined conversation responses