Update app.py

app.py

@@ -11,6 +11,13 @@ from transformers import (
     TextIteratorStreamer,
     LlamaTokenizer,
 )
+import subprocess
+
+subprocess.run(
+    "pip install flash-attn --no-build-isolation",
+    env={"FLASH_ATTENTION_SKIP_CUDA_BUILD": "TRUE"},
+    shell=True,
+)
 
 MAX_MAX_NEW_TOKENS = 1024
 DEFAULT_MAX_NEW_TOKENS = 50
@@ -28,8 +35,7 @@ model = AutoModelForCausalLM.from_pretrained(
     "microsoft/Phi-3-mini-4k-instruct", 
     device_map="cuda", 
     torch_dtype="auto", 
-    trust_remote_code=True,
-    attn_implementation='eager',
+    trust_remote_code=True
 )
 tokenizer = AutoTokenizer.from_pretrained("microsoft/Phi-3-mini-4k-instruct")