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 14, 05:30 UTC
[CI] Fix tensorizer test assertion (#24545)
Signed-off-by: Peter Schuurman <psch@google.com>
pwschuurman committed
Sep 10, 2025 at 06:57 UTC
fcc0a3130a9ffcb8d6ed12503372c2513f982d92
1 file changed
+4
-3
tests/tensorizer_loader/test_tensorizer.py
+4
-3
@@ -161,11 +161,11 @@ def test_load_without_tensorizer_load_format(vllm_runner, capfd, model_ref):
161
model = vllm_runner(
162
model_ref,
163
model_loader_extra_config=TensorizerConfig(tensorizer_uri="test"))
164
+ pytest.fail("Expected RuntimeError for extra config keys")
165
except RuntimeError:
166
out, err = capfd.readouterr()
167
combined_output = out + err
167
- assert ("ValueError: Model loader extra config "
168
- "is not supported for load "
168
+ assert ("ValueError: Unexpected extra config keys for load "
169
"format auto") in combined_output
170
finally:
171
del model
@@ -181,11 +181,12 @@ def test_raise_value_error_on_invalid_load_format(vllm_runner, capfd,
181
model_ref,
182
load_format="safetensors",
183
model_loader_extra_config=TensorizerConfig(tensorizer_uri="test"))
184
+ pytest.fail("Expected RuntimeError for extra config keys")
185
except RuntimeError:
186
out, err = capfd.readouterr()
187
188
combined_output = out + err
188
- assert ("ValueError: Model loader extra config is not supported "
189
+ assert ("ValueError: Unexpected extra config keys "
190
"for load format safetensors") in combined_output
191
finally:
192
del model