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, 09:45 UTC
[CI] Disable Failing Tests (#18165)
Robert Shaw committed
May 14, 2025 at 16:49 UTC
856865008e1a8ffce393901c0245df265b5dfc3f
2 files changed
+4
tests/spec_decode/e2e/test_eagle_correctness.py
+2
@@ -178,6 +178,8 @@ def test_eagle_e2e_greedy_correctness_cuda_graph(
178
batch_size, output_len, seed)
179
180
181
+# TRACKING: https://github.com/vllm-project/vllm/issues/18166
182
+@pytest.mark.skip(reason="RE-ENABLE: Failing on main.")
183
@pytest.mark.parametrize(
184
"common_llm_kwargs",
185
[{
tests/v1/engine/test_engine_core_client.py
+2
@@ -256,6 +256,8 @@ async def test_engine_core_client_asyncio(monkeypatch: pytest.MonkeyPatch):
256
client.shutdown()
257
258
259
+# TRACKING: https://github.com/vllm-project/vllm/issues/18167
260
+@pytest.mark.skip(reason="RE-ENABLE: this test is failing on main.")
261
@pytest.mark.parametrize(
262
"multiprocessing_mode,publisher_config",
263
[(True, "tcp"), (False, "inproc")],