Tiny add command line args for prefill delayer and unify names (#16830)
This commit is contained in:
@@ -489,6 +489,18 @@ _warn_deprecated_env_to_cli_flag(
|
||||
"SGLANG_SUPPORT_CUTLASS_BLOCK_FP8",
|
||||
"It will be completely removed in 0.5.7. Please use '--fp8-gemm-backend=cutlass' instead.",
|
||||
)
|
||||
_warn_deprecated_env_to_cli_flag(
|
||||
"SGLANG_SCHEDULER_DECREASE_PREFILL_IDLE",
|
||||
"Please use '--enable-prefill-delayer' instead.",
|
||||
)
|
||||
_warn_deprecated_env_to_cli_flag(
|
||||
"SGLANG_PREFILL_DELAYER_MAX_DELAY_PASSES",
|
||||
"Please use '--prefill-delayer-max-delay-passes' instead.",
|
||||
)
|
||||
_warn_deprecated_env_to_cli_flag(
|
||||
"SGLANG_PREFILL_DELAYER_TOKEN_USAGE_LOW_WATERMARK",
|
||||
"Please use '--prefill-delayer-token-usage-low-watermark' instead.",
|
||||
)
|
||||
|
||||
|
||||
def example_with_exit_stack():
|
||||
|
||||
Reference in New Issue
Block a user