biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[V1][PP] Optimization: continue scheduling prefill chunks (#17080)

Browse Source 
Signed-off-by: Rui Qiao <ruisearch42@gmail.com>
This commit is contained in:
Rui Qiao
2025-04-24 05:27:08 -07:00
committed by GitHub
parent a9138e85b1
commit c0dfd97519
5 changed files with 128 additions and 74 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
tests/v1/core/test_scheduler.py
View File

@@ -437,7 +437,6 @@ def test_stop_via_update_from_output():
req.num_computed_tokens = req.num_tokens
scheduler.requests[req.request_id] = req
scheduler.running.append(req)
scheduler.scheduled_req_ids.add(req.request_id)
scheduler_output = SchedulerOutput(scheduled_new_reqs=[],
scheduled_cached_reqs=[],
@@ -489,7 +488,6 @@ def test_stop_via_update_from_output():
req.num_computed_tokens = req.num_tokens
scheduler.requests[req.request_id] = req
scheduler.running.append(req)
scheduler.scheduled_req_ids.add(req.request_id)
scheduler_output = SchedulerOutput(scheduled_new_reqs=[],
scheduled_cached_reqs=[],
@@ -539,7 +537,6 @@ def test_stop_via_update_from_output():
req.num_computed_tokens = req.num_tokens
scheduler.requests[req.request_id] = req
scheduler.running.append(req)
scheduler.scheduled_req_ids.add(req.request_id)
scheduler_output = SchedulerOutput(scheduled_new_reqs=[],
scheduled_cached_reqs=[],
@@ -589,7 +586,6 @@ def test_stop_via_update_from_output():
requests[0].num_computed_tokens = requests[0].num_tokens
scheduler.requests[requests[0].request_id] = requests[0]
scheduler.running.append(requests[0])
scheduler.scheduled_req_ids.add(requests[0].request_id)
scheduler_output = SchedulerOutput(
scheduled_new_reqs=[],