A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 15, 11:00 UTC
[CI] Fix test_collective_rpc (#17858)
Signed-off-by: Russell Bryant <rbryant@redhat.com>
Russell Bryant committed
May 8, 2025 at 12:47 UTC
ec54d73c3105ccfdd9b036d1ca0feb62078e20f7
1 file changed
+2
-1
tests/entrypoints/llm/test_collective_rpc.py
+2
-1
@@ -10,7 +10,7 @@ from ...utils import create_new_process_for_each_test
10
@pytest.mark.parametrize("tp_size", [1, 2])
11
@pytest.mark.parametrize("backend", ["mp", "ray"])
12
@create_new_process_for_each_test()
13
-def test_collective_rpc(tp_size, backend):
13
+def test_collective_rpc(tp_size, backend, monkeypatch):
14
if tp_size == 1 and backend == "ray":
15
pytest.skip("Skip duplicate test case")
16
if tp_size == 1:
@@ -21,6 +21,7 @@ def test_collective_rpc(tp_size, backend):
21
def echo_rank(self):
22
return self.rank
23
24
+ monkeypatch.setenv("VLLM_ALLOW_INSECURE_SERIALIZATION", "1")
25
llm = LLM(model="meta-llama/Llama-3.2-1B-Instruct",
26
enforce_eager=True,
27
load_format="dummy",