[Feature] support eager mode in model runner v2 (#5210)

### What this PR does / why we need it?
#5051 only implement a basic framework for model runner v2, but there
are still some bugs for e2e functionality, this PR aim to enable basic
functionality.
model runner v2 plans:
https://github.com/vllm-project/vllm-ascend/issues/5208

- vLLM version: release/v0.13.0
- vLLM main:
ad32e3e19c
---------
Signed-off-by: Ronald1995 <ronaldautomobile@163.com>
This commit is contained in:
Ronald
2025-12-29 15:28:34 +08:00
committed by GitHub
parent 4da46da9bf
commit e7e1a7dc05
19 changed files with 528 additions and 44 deletions

View File

@@ -233,8 +233,8 @@ class NPUWorker(WorkerBase):
self.device = self._init_device()
# Init ModelRunner here, so that we have access to self.device.
if self.use_v2_model_runner:
logger.error(
"npu model runner v2 is in developing, it can't work well for now."
logger.warning(
"npu model runner v2 is in developing, some features doesn't work for now."
)
from vllm_ascend.worker.v2.model_runner import \
NPUModelRunner as NPUModelRunnerV2