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, 22:15 UTC
[CI/Build] Fix V1 LoRA failure (#13767)
Jee Jee Li committed
Feb 25, 2025 at 18:01 UTC
37b6cb49852369945b4560efc8955a992f9b12cb
1 file changed
+2
tests/lora/test_gemma.py
+2
@@ -41,6 +41,8 @@ def v1(run_with_both_engines_lora):
41
pass
42
43
44
+# The V1 lora test for this model requires more than 24GB.
45
+@pytest.mark.skip_v1
46
@pytest.mark.xfail(current_platform.is_rocm(),
47
reason="There can be output mismatch on ROCm")
48
def test_gemma_lora(gemma_lora_files):