[WIP][Core] Enable async scheduling by default

Signed-off-by: Nick Hill <[email protected]>

Author: njhill

vllm/config/scheduler.py

@@ -128,11 +128,12 @@ class SchedulerConfig:
     like full attention and sliding window attention.
     """
 
-    async_scheduling: bool = False
-    """If set to True, perform async scheduling. This helps to avoid gaps in
-    GPU utilization, leading to better latency and throughput.
-    Async scheduling is currently not supported with some features such as
-    speculative decoding and pipeline parallelism.
+    async_scheduling: bool = Field(default=None)
+    """If set to False, disable async scheduling. Async scheduling helps to
+    avoid gaps in GPU utilization, leading to better latency and throughput.
+    It is currently not supported with some features such as
+    speculative decoding and pipeline parallelism, and will be automatically
+    disabled in those cases.
     """
 
     stream_interval: int = Field(default=1, ge=1)

vllm/config/vllm.py

@@ -381,7 +381,7 @@ def __post_init__(self):
                 if self.speculative_config.method not in get_args(EagleModelTypes):
                     raise ValueError(
                         "Currently, async scheduling is only supported "
-                        "with EAGLE/MTP kind of speculative decoding"
+                        "with EAGLE/MTP kind of speculative decoding."
                     )
                 if self.speculative_config.disable_padded_drafter_batch:
                     raise ValueError(
@@ -399,16 +399,27 @@ def __post_init__(self):
                 )
         elif self.scheduler_config.async_scheduling is None:
             # Enable async scheduling unless there is an incompatible option.
-            # NOTE: we won't reach here until async scheduling is enabled by default.
-            if (
-                self.parallel_config.pipeline_parallel_size > 1
-                or self.speculative_config is not None
-            ):
+            if self.parallel_config.pipeline_parallel_size > 1:
                 logger.warning(
-                    "Async scheduling is not yet supported with speculative decoding "
-                    " or pipeline_parallel_size > 1 and will be disabled."
+                    "Async scheduling is not yet supported with "
+                    "pipeline_parallel_size > 1 and will be disabled."
                 )
                 self.scheduler_config.async_scheduling = False
+            elif self.speculative_config is not None:
+                if self.speculative_config.method not in get_args(EagleModelTypes):
+                    logger.warning(
+                        "Async scheduling not supported with %s-based "
+                        "speculative decoding and will be disabled.",
+                        self.speculative_config.method,
+                    )
+                else:
+                    logger.warning(
+                        "Async scheduling will be disabled because some features do "
+                        "not currently work in conjunction with speculative decoding. "
+                        "To use async scheduling with spec decoding anyway, "
+                        "enable it explicitly via async_scheduling=True."
+                    )
+                self.scheduler_config.async_scheduling = False
             elif not executor_supports_async_sched:
                 logger.warning(
                     "Async scheduling will be disabled because it is not supported "
@@ -420,6 +431,11 @@ def __post_init__(self):
             else:
                 self.scheduler_config.async_scheduling = True
 
+        logger.info_once(
+            "Asynchronous scheduling is %s.",
+            "enabled" if self.scheduler_config.async_scheduling else "disabled",
+        )
+
         from vllm.platforms import current_platform
 
         if (