From 97e06f34235fcce34f3ba508dc166ada75b41839 Mon Sep 17 00:00:00 2001
From: Wing Lian <wing.lian@gmail.com>
Date: Tue, 23 Jan 2024 10:06:22 -0500
Subject: [PATCH] make sure flash attention is enabled for the test

---
 tests/e2e/patched/test_phi_multipack.py | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/tests/e2e/patched/test_phi_multipack.py b/tests/e2e/patched/test_phi_multipack.py
index eefd912e71..e45842b20b 100644
--- a/tests/e2e/patched/test_phi_multipack.py
+++ b/tests/e2e/patched/test_phi_multipack.py
@@ -35,6 +35,7 @@ def test_ft_packed(self, temp_dir):
                 "tokenizer_type": "AutoTokenizer",
                 "sequence_len": 2048,
                 "sample_packing": True,
+                "flash_attention": True,
                 "pad_to_sequence_len": True,
                 "load_in_8bit": False,
                 "adapter": None,
@@ -82,6 +83,7 @@ def test_qlora_packed(self, temp_dir):
                 "tokenizer_type": "AutoTokenizer",
                 "sequence_len": 2048,
                 "sample_packing": True,
+                "flash_attention": True,
                 "pad_to_sequence_len": True,
                 "load_in_8bit": False,
                 "adapter": "qlora",