[FIX] Fix prefix test error on main (vllm-project#3286)

raaymax · Mar 9, 2024 · f48c679 · f48c679
1 parent c2c5e09
commit f48c679
Showing 1 changed file with 0 additions and 2 deletions.
diff --git a/vllm/model_executor/layers/attention/backends/flash_attn.py b/vllm/model_executor/layers/attention/backends/flash_attn.py
@@ -103,8 +103,6 @@ def forward(
                     key_cache,
                     value_cache,
                     input_metadata,
-                    self.num_heads,
-                    self.num_kv_heads,
                     self.alibi_slopes,
                 )
         else: