[Core] Whisper enable FULL_DECODE_ONLY CudaGraph (#30072)

Signed-off-by: NickLucche <nlucches@redhat.com>
This commit is contained in:
Nicolò Lucchesi
2025-12-10 06:14:24 -08:00
committed by GitHub
parent d017bceb08
commit c756fb6781
3 changed files with 31 additions and 12 deletions
@@ -103,6 +103,8 @@ def run_test(
max_model_len=448,
tensor_parallel_size=tensor_parallel_size,
distributed_executor_backend=distributed_executor_backend,
# TODO (NickLucche) figure out output differences with non-eager and re-enable
enforce_eager=True,
) as vllm_model:
llm = vllm_model.llm