From 3f1d40960fb79e6f1314755abf2d43d142e33363 Mon Sep 17 00:00:00 2001 From: AlexHuang Date: Sun, 26 Jul 2026 21:22:58 +0800 Subject: [PATCH] [KV Offload] Fix num_tokens_after_batch for different termination types (#49285) Signed-off-by: Alex Signed-off-by: Alex Signed-off-by: Alex Signed-off-by: Alex Co-authored-by: Claude Co-authored-by: Or Ozeri --- .../kv_transfer/kv_connector/v1/offloading/scheduler.py | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/vllm/distributed/kv_transfer/kv_connector/v1/offloading/scheduler.py b/vllm/distributed/kv_transfer/kv_connector/v1/offloading/scheduler.py index 6ccc820f25..f809dca5ba 100644 --- a/vllm/distributed/kv_transfer/kv_connector/v1/offloading/scheduler.py +++ b/vllm/distributed/kv_transfer/kv_connector/v1/offloading/scheduler.py @@ -50,7 +50,7 @@ from vllm.v1.kv_offload.base import ( make_offload_key, ) from vllm.v1.outputs import KVConnectorOutput -from vllm.v1.request import Request +from vllm.v1.request import Request, RequestStatus logger = init_logger(__name__) @@ -976,7 +976,9 @@ class OffloadingConnectorScheduler: continue req = req_status.req - if req.is_finished(): + if req.status is RequestStatus.FINISHED_ABORTED: + num_tokens_after_batch = req.num_computed_tokens + elif req.is_finished(): num_tokens_after_batch = req.num_tokens else: num_scheduled_tokens = scheduler_output.num_scheduled_tokens[req_id]