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
[Bugfix] Fix silu_mul+quant fusion test (#24341)
Signed-off-by: elvischenv <219235043+elvischenv@users.noreply.github.com>
elvischenv committed
Sep 6, 2025 at 04:13 UTC
eedb2a2a102b47d527071b9acfb9edd541ef5de6
1 file changed
+2
-1
tests/compile/test_silu_mul_quant_fusion.py
+2
-1
@@ -118,7 +118,8 @@ def test_fusion_silu_and_mul_quant(num_tokens, hidden_size, model_class,
118
fusion_pass = ActivationQuantFusionPass(config)
119
120
backend = TestBackend(NoOpEliminationPass(config), fusion_pass)
121
- model = model_class(hidden_size, cuda_force_torch)
121
+ model = model_class(hidden_size=hidden_size,
122
+ cuda_force_torch=cuda_force_torch)
123
124
# First dimension dynamic
125
x = torch.rand(num_tokens, hidden_size * 2)