From d9ee491f7083075f44c2ac750ee42639636f005b Mon Sep 17 00:00:00 2001 From: Angazenn <92204292+Angazenn@users.noreply.github.com> Date: Fri, 17 Oct 2025 11:19:41 +0800 Subject: [PATCH] [BugFix]Move to_list in foward_v1 with FIA earlier to build (#3185) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit ### What this PR does / why we need it? The current implementation of FIA will introduce an `to_list` operation for actual_seq_lengths_q and seq_lens，which comsumes extra time. These operation can be moved earlier into `build` operation of attention metadata. ### Does this PR introduce _any_ user-facing change? No. ### How was this patch tested? - vLLM version: v0.11.0rc3 - vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0 --------- Signed-off-by: Angazenn --- vllm_ascend/attention/attention_v1.py | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) diff --git a/vllm_ascend/attention/attention_v1.py b/vllm_ascend/attention/attention_v1.py index 27c289b..7dc309b 100644 --- a/vllm_ascend/attention/attention_v1.py +++ b/vllm_ascend/attention/attention_v1.py @@ -140,7 +140,13 @@ class AscendMetadata: # The sequence length per sequence. Sequence length means the computed # tokens + new tokens (is None if it is a decoding). # (batch_size,) + # TODO(Angazenn): The following parameters are quite redundant and + # contains similar information (such as seq_lens seq_lens_list). We + # should simplified these parameters once attention schema in vLLM-Ascend + # is unified. seq_lens: torch.Tensor = None + seq_lens_list: List[int] = None # type: ignore + actual_seq_lengths_q: List[int] = None # type: ignore query_start_loc: torch.Tensor = None query_lens: torch.Tensor = None @@ -229,7 +235,9 @@ class AscendAttentionMetadataBuilder: query_start_loc=query_start_loc, query_lens=query_lens, seq_lens=seq_lens, + seq_lens_list=seq_lens.tolist(), max_query_len=common_attn_metadata.max_query_len, + actual_seq_lengths_q=query_start_loc_cpu[1:].tolist(), slot_mapping=slot_mapping, attn_mask=attn_mask, attn_state=attn_state, @@ -528,8 +536,8 @@ class AscendAttentionBackendImpl(AttentionImpl): block_table=attn_metadata.block_tables, input_layout="TND", block_size=block_size, - actual_seq_lengths=attn_metadata.query_start_loc[1:], - actual_seq_lengths_kv=attn_metadata.seq_lens, + actual_seq_lengths=attn_metadata.actual_seq_lengths_q, + actual_seq_lengths_kv=attn_metadata.seq_lens_list, num_key_value_heads=self.num_kv_heads, num_heads=self.num_heads, scale=self.scale,