[npugraph_ex]enable npugraph_ex by default (#6664)
### What this PR does / why we need it? This pull request enables the `npugraph_ex` backend by default to improve performance on Ascend NPUs, as proposed in the [RFC](https://github.com/vllm-project/vllm-ascend/issues/6214). ### Does this PR introduce _any_ user-facing change? Yes. `npugraph_ex` is now enabled by default. Users can disable it by setting `enable: false` in the `npugraph_ex_config` section of the `additional_config`. ### How was this patch tested? CI passed. The changes are covered by existing and new E2E tests (`test_aclgraph_accuracy.py`) and unit tests (`test_ascend_config.py`) that have been updated to reflect the new default behavior. The tests verify correctness and consistency with `npugraph_ex` enabled and disabled, as well as with the new static kernel option. Signed-off-by: huyuanquan1 <huyuanquan1@huawei.com> Co-authored-by: huyuanquan1 <huyuanquan1@huawei.com>
This commit is contained in:
@@ -260,7 +260,7 @@ class NpugraphExConfig:
|
||||
|
||||
def __init__(
|
||||
self,
|
||||
enable: bool = False,
|
||||
enable: bool = True,
|
||||
enable_static_kernel: bool = False,
|
||||
fuse_norm_quant: bool = True,
|
||||
fuse_qknorm_rope: bool = True,
|
||||
@@ -274,7 +274,7 @@ class NpugraphExConfig:
|
||||
enable (bool): Whether to enable npugraph_ex backend.
|
||||
When set to True, the Fx graph generated by Dymano will be
|
||||
optimized and compiled by the npugraph_ex backend.
|
||||
Default: False
|
||||
Default: True
|
||||
enable_static_kernel (bool): Whether to enable static kernel.
|
||||
Static kernel is suitable for scenarios with purely static shapes
|
||||
or minimal shape changes, and can improve network performance.
|
||||
|
||||
Reference in New Issue
Block a user