From 9e74d9d003d546c17dca472c3f4b48be651f1d7c Mon Sep 17 00:00:00 2001
From: Charles Riggins <liqianchen123@foxmail.com>
Date: Tue, 18 Jun 2024 00:05:33 +0800
Subject: [PATCH] Correct alignment in the seq_len diagram. (#5592)

Co-authored-by: Liqian Chen <liqian.chen@deeplang.ai>
---
 vllm/attention/backends/flash_attn.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/vllm/attention/backends/flash_attn.py b/vllm/attention/backends/flash_attn.py
index 300bab72..1c48e2a0 100644
--- a/vllm/attention/backends/flash_attn.py
+++ b/vllm/attention/backends/flash_attn.py
@@ -83,7 +83,7 @@ class FlashAttentionMetadata(AttentionMetadata):
     # |---------------- N iteration ---------------------|
     # |- tokenA -|......................|-- newTokens ---|
     # |---------- context_len ----------|
-    # |-------------------- seq_len ----------------------|
+    # |-------------------- seq_len ---------------------|
     #                                   |-- query_len ---|
 
     # Maximum query length in the batch. None for decoding.