From 05c09a2801aa122a81a6e7785b7a349ee8ad748e Mon Sep 17 00:00:00 2001
From: project6-dev <dev@project6.ai>
Date: Wed, 12 Aug 2026 08:35:42 +0000
Subject: [PATCH] =?UTF-8?q?fix:=20revert=20max-num-seqs=20to=201=20?=
 =?UTF-8?q?=E2=80=94=20segfault=20during=20profile=5Frun=20with=20n=3D2?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Profile run with 2 sequences triggers segfault in torch.distributed.gather
during determine_num_available_blocks. t2_n_2 is worth 1 point, not worth
the crash risk. max_completion_tokens fix is the priority.
---
 computility-run.yaml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/computility-run.yaml b/computility-run.yaml
index ac5b8ec2..41351f46 100644
--- a/computility-run.yaml
+++ b/computility-run.yaml
@@ -15,7 +15,7 @@ command:
     - -tp
     - '4'
     - --max-num-seqs
-    - '2'
+    - '1'
     - --disable-log-requests
     - --disable-frontend-multiprocessing
     - --max-num-batched-tokens