EngineX/xc-llm-ascend
3
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

[main] rename device type (#5099)

Browse Source 
### What this PR does / why we need it?
Rename `_910B` to `A2`;
Rename `_910_93` to `A3`;
Rename `_910_95` to `A5`;

- vLLM version: v0.12.0
- vLLM main:
ad32e3e19c

Signed-off-by: zzzzwwjj <1183291235@qq.com>
This commit is contained in:
zzzzwwjj
2025-12-17 14:08:19 +08:00
committed by GitHub
parent 4144376e88
commit 06b82e7503
16 changed files with 47 additions and 48 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
vllm_ascend/worker/worker_v1.py
View File

@@ -90,7 +90,7 @@ class NPUWorker(WorkerBase):
# Register ops when worker init.
from vllm_ascend import ops
ops.register_dummy_fusion_op()
if get_ascend_device_type() != AscendDeviceType._910_95:
if get_ascend_device_type() != AscendDeviceType.A5:
_register_atb_extensions()
register_ascend_customop(vllm_config)
# init ascend config and soc version
@@ -360,7 +360,7 @@ class NPUWorker(WorkerBase):
self.model_runner.capture_model()
# Call ATB matmul to warm up; otherwise, the first operation (ReshapeAndCache)
# may cause performance degradation at runtime.
if get_ascend_device_type() != AscendDeviceType._910_95:
if get_ascend_device_type() != AscendDeviceType.A5:
self._warm_up_atb()
# Reset the seed to ensure that the random state is not affected by
# the model initialization and profiling.