Skip to content

Instantly share code, notes, and snippets.

View stepinto's full-sized avatar

Chao Shi stepinto

View GitHub Profile
@stepinto
stepinto / a.sh
Created July 2, 2026 01:08
SWE-Bench PR 27010 Test
# Baseline: DS TP8 L3
sglang serve \
--model=deepseek-ai/DeepSeek-V4-Flash \
--tp-size 8 \
--moe-runner-backend flashinfer_mxfp4 \
--disable-cuda-graph \
--mem-fraction-static 0.85 \
--reasoning-parser deepseek-v4 \
--tool-call-parser deepseekv4 \
--host 0.0.0.0 \
eth0 ---- sw1a --- sw2a --- sw3a --- eth0
host1 X X host2
eth1 ---- sw1b --- sw2b --- sw3b --- eth1
eth0 ---- sw1a --- sw2a --- sw3a --- eth0
host1 X X host2
eth1 ---- sw1b --- sw2b --- eth1 --- eth1