[Feature] limit thinking tokens (hard limit) (#20859)
Signed-off-by: Sungjae Lee <33976427+llsj14@users.noreply.github.com> Signed-off-by: Sungjae Lee <sung-jae.lee@navercorp.com> Signed-off-by: Chauncey <chaunceyjiang@gmail.com> Co-authored-by: Chauncey <chaunceyjiang@gmail.com> Co-authored-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
This commit is contained in:
co-authored by
Chauncey
Harry Mellor
parent
8de5261e69
commit
4731884796
@@ -53,6 +53,7 @@ from vllm.config import (
|
||||
PoolerConfig,
|
||||
PrefetchOffloadConfig,
|
||||
ProfilerConfig,
|
||||
ReasoningConfig,
|
||||
SchedulerConfig,
|
||||
SpeculativeConfig,
|
||||
StructuredOutputsConfig,
|
||||
@@ -581,6 +582,7 @@ class EngineArgs:
|
||||
kv_events_config: KVEventsConfig | None = None
|
||||
|
||||
ec_transfer_config: ECTransferConfig | None = None
|
||||
reasoning_config: ReasoningConfig = get_field(VllmConfig, "reasoning_config")
|
||||
|
||||
generation_config: str = ModelConfig.generation_config
|
||||
enable_sleep_mode: bool = ModelConfig.enable_sleep_mode
|
||||
@@ -1297,6 +1299,7 @@ class EngineArgs:
|
||||
vllm_group.add_argument(
|
||||
"--attention-config", "-ac", **vllm_kwargs["attention_config"]
|
||||
)
|
||||
vllm_group.add_argument("--reasoning-config", **vllm_kwargs["reasoning_config"])
|
||||
vllm_group.add_argument("--kernel-config", **vllm_kwargs["kernel_config"])
|
||||
vllm_group.add_argument(
|
||||
"--additional-config", **vllm_kwargs["additional_config"]
|
||||
@@ -1958,6 +1961,7 @@ class EngineArgs:
|
||||
kv_transfer_config=self.kv_transfer_config,
|
||||
kv_events_config=self.kv_events_config,
|
||||
ec_transfer_config=self.ec_transfer_config,
|
||||
reasoning_config=self.reasoning_config,
|
||||
profiler_config=self.profiler_config,
|
||||
additional_config=self.additional_config,
|
||||
optimization_level=self.optimization_level,
|
||||
|
||||
Reference in New Issue
Block a user