[Spec Decode][MRV2] Warm up top-k/top-p and greedy rejection kernels for speculative decoding (#59260)
Signed-off-by: zixi-qi <zixi@inferact.ai> Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
This commit is contained in:
co-authored by
Claude Opus 5.5
mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
parent
ae94e9c9e8
commit
c1407a7e86
@@ -458,9 +458,11 @@ class GPUModelRunner(LoRAModelRunnerMixin):
|
||||
self.vllm_config, self.req_states, self.is_pooling_model, config_processors
|
||||
)
|
||||
if self.is_last_pp_rank and not self.is_pooling_model:
|
||||
# V2 bypasses TopKTopPSampler, which registers native warmups.
|
||||
# ROCm only: on CUDA these warmups add ~2 min to every engine start.
|
||||
if current_platform.is_rocm():
|
||||
# Speculative verification uses native top-k/top-p masking even
|
||||
# when target-only sampling can use FlashInfer's fused sampler.
|
||||
if current_platform.is_rocm() or (
|
||||
current_platform.is_cuda() and self.speculative_config is not None
|
||||
):
|
||||
from vllm.v1.sample.ops.topk_topp_sampler import (
|
||||
register_top_k_top_p_warmups,
|
||||
)
|
||||
|
||||
Reference in New Issue
Block a user