xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2026-01-07 18:43:04 +08:00
Code Issues Packages Projects Releases Wiki Activity

[Perf][Async Scheduling] Remove CPU->GPU sync in dummy_run (#27455)

Browse Source 
Signed-off-by: Lehua Ding <lehuading@tencent.com>
This commit is contained in:
Lehua Ding 2025-10-25 04:45:36 +08:00 committed by GitHub
parent 17af6aa0da
commit 0402428200
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 4 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
vllm/v1/worker/gpu_model_runner.py
View File

@ -3492,7 +3492,10 @@ class GPUModelRunner(LoRAModelRunnerMixin, KVConnectorModelRunnerMixin):
self.eplb_step(is_dummy=True, is_profile=is_profile)
logit_indices = np.cumsum(num_scheduled_tokens) - 1
return hidden_states, hidden_states[logit_indices]
logit_indices_device = torch.from_numpy(logit_indices).to(
self.device, non_blocking=True
)
return hidden_states, hidden_states[logit_indices_device]
@torch.inference_mode()
def _dummy_sampler_run(