obscura/vllm
2
0
Fork 0
mirror of https://github.com/vllm-project/vllm.git synced 2026-08-22 21:50:15 +00:00
Code Issues Packages Projects Releases Wiki Activity

[CI][Bugfix] Reduce max_model_len in OOT embedding test to fix KV-cache OOM on small GPUs (#49351)

Browse Source 
Signed-off-by: sfeng33 <[email protected]>
This commit is contained in:
Flora Feng
2026-07-21 15:00:06 -04:00
committed by GitHub
parent 33178f9006
commit fca252d59e
1 changed files with 1 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files
tests/plugins_tests/test_oot_registration_offline.py
+1 -1
View File
@@ -50,7 +50,7 @@ def test_oot_registration_embedding(
m.setenv("VLLM_PLUGINS", "register_dummy_model")
prompts = ["Hello, my name is", "The text does not matter"]
llm = LLM(
model=dummy_gemma2_embedding_path, load_format="dummy", max_model_len=2048
model=dummy_gemma2_embedding_path, load_format="dummy", max_model_len=512
)
outputs = llm.embed(prompts)