Fix internlm after https://github.com/vllm-project/vllm/pull/2860 (#2861)

vllm/model_executor/models/llama.py

@@ -175,7 +175,8 @@ class LlamaDecoderLayer(nn.Module):
         self.self_attn = LlamaAttention(
             hidden_size=self.hidden_size,
             num_heads=config.num_attention_heads,
-            num_kv_heads=config.num_key_value_heads,
+            num_kv_heads=getattr(config, "num_key_value_heads",
+                                 config.num_attention_heads),
             rope_theta=rope_theta,
             rope_scaling=rope_scaling,
             max_position_embeddings=max_position_embeddings,