Skip to content

Commit 79fbb20

Browse files
authored
[ModelRunner] remove unused args (follow vllm changes) (#159)
### What this PR does / why we need it? The arg list of `Attention.forward()` is changed by vllm-project/vllm#13555. The unused args `kv_caches` and `attn_metadata` are removed. ### Does this PR introduce _any_ user-facing change? N/A ### How was this patch tested? CI passed with existing test. Signed-off-by: MengqingCao <[email protected]>
1 parent 51ae37b commit 79fbb20

File tree

1 file changed

+0
-2
lines changed

1 file changed

+0
-2
lines changed

vllm_ascend/model_runner.py

Lines changed: 0 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -1142,8 +1142,6 @@ def execute_model(
11421142
hidden_or_intermediate_states = model_executable(
11431143
input_ids=model_input.input_tokens,
11441144
positions=model_input.input_positions,
1145-
kv_caches=kv_caches,
1146-
attn_metadata=model_input.attn_metadata,
11471145
intermediate_tensors=intermediate_tensors,
11481146
**MultiModalKwargs.as_kwargs(multi_modal_kwargs,
11491147
device=self.device),

0 commit comments

Comments
 (0)