[PD][LoRA] Gate decode admission on adapter slots (#39332)

This commit is contained in:
Yanbin Jiang
2026-09-15 13:11:04 +08:00
committed by GitHub
parent 3f871a246c
commit ebd37705e4
5 changed files with 184 additions and 2 deletions
@@ -127,6 +127,7 @@ class TestDecodeQueueCleanup(CustomTestCase):
scheduler.running_batch.reqs = []
scheduler.enable_priority_scheduling = False
scheduler.enable_hisparse = False
scheduler.enable_lora = False
scheduler.metrics_reporter.enable_metrics = False
scheduler.output_streamer = MagicMock()
queue.scheduler = scheduler
@@ -181,6 +182,7 @@ class TestDecodeQueueCleanup(CustomTestCase):
scheduler.running_batch.reqs = []
scheduler.enable_priority_scheduling = False
scheduler.enable_hisparse = False
scheduler.enable_lora = False
scheduler.output_streamer = MagicMock()
queue.scheduler = scheduler
@@ -241,6 +243,7 @@ class TestDecodeQueueCleanup(CustomTestCase):
scheduler.running_batch.reqs = []
scheduler.enable_priority_scheduling = False
scheduler.enable_hisparse = False
scheduler.enable_lora = False
scheduler.server_args.disaggregation_decode_enable_radix_cache = False
scheduler.output_streamer = MagicMock()
queue.scheduler = scheduler