From c704aeccae192d7f88e9138271682ddad1327a69 Mon Sep 17 00:00:00 2001 From: wuli666 <421774554@qq.com> Date: Tue, 8 Sep 2026 14:17:33 +0800 Subject: [PATCH] [Bugfix] Fix Step-Audio2 L4 CI KV cache allocation (#6839) Signed-off-by: wuli666 <421774554@qq.com> --- tests/config/test_config_factory.py | 2 +- vllm_omni/deploy/step_audio2_ci.yaml | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/tests/config/test_config_factory.py b/tests/config/test_config_factory.py index 54d04cb82..6cb328bc5 100644 --- a/tests/config/test_config_factory.py +++ b/tests/config/test_config_factory.py @@ -1597,7 +1597,7 @@ stages: assert stage.max_model_len == 1024 assert stage.max_num_batched_tokens == 1024 assert stage.max_num_seqs == 1 - assert stage.gpu_memory_utilization == 0.7 + assert stage.gpu_memory_utilization == 0.8 assert stage.skip_mm_profiling is True assert stage.enforce_eager is True assert stage.async_scheduling is False diff --git a/vllm_omni/deploy/step_audio2_ci.yaml b/vllm_omni/deploy/step_audio2_ci.yaml index 4373dfaa8..060ae6c54 100644 --- a/vllm_omni/deploy/step_audio2_ci.yaml +++ b/vllm_omni/deploy/step_audio2_ci.yaml @@ -15,7 +15,7 @@ stages: max_model_len: 1024 max_num_batched_tokens: 1024 max_num_seqs: 1 - gpu_memory_utilization: 0.7 + gpu_memory_utilization: 0.8 skip_mm_profiling: true enforce_eager: true async_scheduling: false