From 95d03147e7fac105a797dbf98f9cd2f64a408300 Mon Sep 17 00:00:00 2001
From: dylan-claude <dylan@project6.dev>
Date: Fri, 7 Aug 2026 07:16:28 +0000
Subject: [PATCH] =?UTF-8?q?fix(deploy):=20reduce=20max-model-len=20131072?=
 =?UTF-8?q?=E2=86=92100000,=20remove=20chunked-prefill,=20single-seq=20?=
 =?UTF-8?q?=E2=80=94=20prevent=20OOM=20crash?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

CCCL design reference: block_topk_air.cuh tile_items = threads * items must fit hardware.
max-model-len is vllm's tile size — 131072 overflows BI-V100 VRAM budget.
Submission 508 failed with 100% Connection refused = service never started.
---
 computility-run.yaml | 9 ++-------
 1 file changed, 2 insertions(+), 7 deletions(-)

diff --git a/computility-run.yaml b/computility-run.yaml
index f02fcda2..6c0d5cb1 100644
--- a/computility-run.yaml
+++ b/computility-run.yaml
@@ -8,21 +8,16 @@ command:
     - --served-model-name
     - llm
     - --max-model-len
-    - '131072'
+    - '100000'
     - --gpu-memory-utilization
     - '0.90'
     - --trust-remote-code
     - -tp
     - '4'
     - --max-num-seqs
-    - '2'
+    - '1'
     - --disable-log-requests
     - --disable-frontend-multiprocessing
-    - --max-num-batched-tokens
-    - '8192'
-    - --enable-chunked-prefill
-    - --max-seq-len-to-capture
-    - '32768'
     - --enforce-eager
     - --enable-auto-tool-choice
     - --tool-call-parser