A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 16, 03:15 UTC
fix: [doc] fix typo (#11751)
Co-authored-by: Lancer <maruixiang6688@gmail.com>
Lancer committed
Jan 6, 2025 at 00:48 UTC
eba17173d34548a39989eae2530dce53496a1f3d
1 file changed
+1
-1
vllm/core/block/block_table.py
+1
-1
@@ -23,7 +23,7 @@ class BlockTable:
23
blocks to initialize the BlockTable with. If not provided, an empty
24
BlockTable is created.
25
max_block_sliding_window (Optional[int], optional): The number of
26
- blocks to keep around for each sequance. If None, all blocks
26
+ blocks to keep around for each sequence. If None, all blocks
27
are kept (eg., when sliding window is not used).
28
It should at least fit the sliding window size of the model.
29