Revert "Embedding parallel by attn_tp (#7623)" (#7880)

2025-07-08 22:03:09 -07:00
parent 0870232195
commit b6b6268ccf
1 changed files with 1 additions and 1 deletions
--- a/python/sglang/srt/models/deepseek_v2.py
+++ b/python/sglang/srt/models/deepseek_v2.py
@@ -1938,7 +1938,7 @@ class DeepseekV2Model(nn.Module):
        self.embed_tokens = VocabParallelEmbedding(
            config.vocab_size,
            config.hidden_size,
-            use_attn_tp_group=True,
+            enable_tp=not global_server_args_dict["enable_dp_attention"],
        )
        self.alt_stream = torch.cuda.Stream() if _is_cuda else None
        self.layers = nn.ModuleList(