[K3] support recoverssm for K3 (#51855)

Signed-off-by: zjy0516 <riverclouds.zhu@qq.com>
Signed-off-by: Benjamin Chislett <bchislett@nvidia.com>
Co-authored-by: OpenAI Codex <codex@openai.com>
Co-authored-by: Benjamin Chislett <bchislett@nvidia.com>
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
This commit is contained in:
Jiangyun Zhu
2026-08-17 19:34:14 +08:00
committed by GitHub
co-authored by OpenAI Codex Benjamin Chislett mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
parent f27ae25473
commit 70afdedc10
19 changed files with 2235 additions and 75 deletions
+5 -3
View File
@@ -71,10 +71,12 @@ class MambaBase(AttentionLayerBase):
page_size_padded=page_size_padded,
mamba_type=self.mamba_type,
mamba_cache_mode=vllm_config.cache_config.mamba_cache_mode,
# RecoverSSM verifies the whole window off one checkpoint, so it
# never writes the baseline's per-draft-token state slots.
num_speculative_blocks=(
vllm_config.speculative_config.num_speculative_tokens
if vllm_config.speculative_config
else 0
0
if vllm_config.cache_config.use_kda_recoverssm
else vllm_config.num_speculative_tokens
),
)