diff --git a/vllm/v1/core/sched/scheduler.py b/vllm/v1/core/sched/scheduler.py
index 61640e856ac1c..4323141c435b7 100644
--- a/vllm/v1/core/sched/scheduler.py
+++ b/vllm/v1/core/sched/scheduler.py
@@ -472,9 +472,9 @@ class Scheduler(SchedulerInterface):
                     num_computed_tokens = (
                         num_new_local_computed_tokens + num_external_computed_tokens
                     )
-                # KVTransfer: WAITING reqs have num_computed_tokens > 0
-                # after async KV recvs are completed.
                 else:
+                    # KVTransfer: WAITING reqs have num_computed_tokens > 0
+                    # after async KV recvs are completed.
                     new_computed_blocks = self.kv_cache_manager.empty_kv_cache_blocks
                     num_new_local_computed_tokens = 0
                     num_computed_tokens = request.num_computed_tokens
@@ -483,12 +483,12 @@ class Scheduler(SchedulerInterface):
                 external_load_encoder_input = []
                 new_encoder_compute_budget = encoder_compute_budget
 
-                # KVTransfer: loading remote KV, do not allocate for new work.
                 if load_kv_async:
+                    # KVTransfer: loading remote KV, do not allocate for new work.
                     assert num_external_computed_tokens > 0
                     num_new_tokens = 0
-                # Number of tokens to be scheduled.
                 else:
+                    # Number of tokens to be scheduled.
                     # We use `request.num_tokens` instead of
                     # `request.num_prompt_tokens` to consider the resumed
                     # requests, which have output tokens.