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

fix some typos (#24616)

Browse Source 
Signed-off-by: co63oc <co63oc@users.noreply.github.com>
This commit is contained in:
co63oc
2025-09-12 01:48:46 +08:00
committed by GitHub
parent c1eda615ba
commit e26fef8397
5 changed files with 6 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
tests/kernels/mamba/test_mamba_ssm_ssd.py
View File

@@ -345,7 +345,7 @@ def test_mamba_chunk_scan_cont_batch_prefill_chunking(chunk_size, seqlens):
# in the mamba2 ssd kernels, by comparing concatenation (in the sequence
# dimension) of chunked results with the full sequence result.
# It is different from test_mamba_chunk_scan_cont_batch by:
# 1. Not using the naive torch implementaion (ssd_minimal_discrete) to get
# 1. Not using the naive torch implementation (ssd_minimal_discrete) to get
# reference outputs. Instead, it compares chunked kernel outputs to full
# sequence kernel outputs. This is the most straightforward way to
# assert chunked prefill correctness.