[Bugfix][Kimi-K3] Declare max_tp_shards on the DSpark MLA KV cache spec (#59733)

Signed-off-by: hyukjlee <hyukjlee@amd.com>
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
Co-authored-by: Summer Yang <girasoleyang@gmail.com>
This commit is contained in:
Hyukjoon Lee
2026-10-08 00:33:32 -04:00
committed by GitHub
co-authored by Claude Opus 5.5 mergify[bot] <37929162+mergify[bot]@users.noreply.github.com> Summer Yang
parent 5dd4a628dd
commit 1fbbfa1657
2 changed files with 34 additions and 0 deletions
+1
View File
@@ -422,6 +422,7 @@ class MultiHeadLatentAttention(nn.Module, AttentionLayerBase):
return MLAAttentionSpec( # type: ignore[call-arg]
block_size=vllm_config.cache_config.block_size,
num_kv_heads=1,
max_tp_shards=1,
head_size=self.head_size,
dtype=kv_cache_dtype,
cache_dtype_str=self.kv_cache_dtype,