biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Misc] Add max_seq_len to CommonAttentionMetadata (#23216)

Browse Source 
Signed-off-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
This commit is contained in:
Woosuk Kwon
2025-08-20 09:05:29 -07:00
committed by GitHub
parent 5efd6905bc
commit d6d13bd49e
12 changed files with 22 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
vllm/v1/attention/backends/rocm_aiter_fa.py
View File

@@ -270,7 +270,7 @@ class AiterFlashAttentionMetadataBuilder(
num_actual_tokens = common_attn_metadata.num_actual_tokens
max_query_len = common_attn_metadata.max_query_len
max_seq_len = int(common_attn_metadata.seq_lens_cpu.max())
max_seq_len = common_attn_metadata.max_seq_len
query_start_loc = common_attn_metadata.query_start_loc
seq_lens = common_attn_metadata.seq_lens
block_table_tensor = common_attn_metadata.block_table_tensor