[Bug] Fix DeepGEMM Attention Test (#26423)

Signed-off-by: yewentao256 <zhyanwentao@126.com>

pyproject.toml

@@ -184,6 +184,7 @@ ba = "ba"
 
 [tool.typos.type.py.extend-words]
 ba = "ba"
+nd = "nd"
 
 [tool.typos.type.cpp]
 extend-glob = ["*.cu"]

tests/kernels/attention/test_deepgemm_attention.py

@@ -82,8 +82,7 @@ def _ref_fp8_mqa_logits(
         torch.arange(0, seq_len_kv, device="cuda")[None, :] < cu_seqlen_ke[:, None]
     )
     mask = mask_lo & mask_hi
-
-    score = torch.einsum("mhd,and->hmn", q, k)
+    score = torch.einsum("mhd,nd->hmn", q, k)
     logits = (score.relu() * weights.unsqueeze(-1).transpose(0, 1)).sum(dim=0)
     logits = logits.masked_fill(~mask, float("-inf"))