Use more general heuristics to set the default value of --mem-fraction-static (#10975)

Co-authored-by: sglang-bot <sglangbot@gmail.com>
This commit is contained in:
Lianmin Zheng
2025-09-29 10:11:03 -07:00
committed by GitHub
parent 816b3a433a
commit a17e70f5cc
9 changed files with 167 additions and 151 deletions

View File

@@ -99,8 +99,6 @@ jobs:
needs: [check-changes, sgl-kernel-build-wheels]
if: needs.check-changes.outputs.sgl_kernel == 'true'
runs-on: 1-gpu-runner
env:
HF_TOKEN: ${{ secrets.HF_TOKEN }}
steps:
- uses: actions/checkout@v4
@@ -233,8 +231,6 @@ jobs:
if: always() && !failure() && !cancelled() &&
((needs.check-changes.outputs.main_package == 'true') || (needs.check-changes.outputs.sgl_kernel == 'true'))
runs-on: 1-gpu-runner
env:
HF_TOKEN: ${{ secrets.HF_TOKEN }}
strategy:
fail-fast: false
matrix:
@@ -266,8 +262,6 @@ jobs:
if: always() && !failure() && !cancelled() &&
((needs.check-changes.outputs.main_package == 'true') || (needs.check-changes.outputs.sgl_kernel == 'true'))
runs-on: 2-gpu-runner
env:
HF_TOKEN: ${{ secrets.HF_TOKEN }}
strategy:
fail-fast: false
matrix:
@@ -299,8 +293,6 @@ jobs:
if: always() && !failure() && !cancelled() &&
((needs.check-changes.outputs.main_package == 'true') || (needs.check-changes.outputs.sgl_kernel == 'true'))
runs-on: 4-gpu-runner
env:
HF_TOKEN: ${{ secrets.HF_TOKEN }}
strategy:
fail-fast: false
matrix:
@@ -332,8 +324,6 @@ jobs:
if: always() && !failure() && !cancelled() &&
((needs.check-changes.outputs.main_package == 'true') || (needs.check-changes.outputs.sgl_kernel == 'true'))
runs-on: 8-gpu-runner
env:
HF_TOKEN: ${{ secrets.HF_TOKEN }}
strategy:
fail-fast: false
matrix:
@@ -365,8 +355,6 @@ jobs:
if: always() && !failure() && !cancelled() &&
((needs.check-changes.outputs.main_package == 'true') || (needs.check-changes.outputs.sgl_kernel == 'true'))
runs-on: 1-gpu-runner
env:
HF_TOKEN: ${{ secrets.HF_TOKEN }}
steps:
- name: Checkout code
uses: actions/checkout@v4
@@ -426,8 +414,6 @@ jobs:
if: always() && !failure() && !cancelled() &&
((needs.check-changes.outputs.main_package == 'true') || (needs.check-changes.outputs.sgl_kernel == 'true'))
runs-on: 1-gpu-runner
env:
HF_TOKEN: ${{ secrets.HF_TOKEN }}
steps:
- name: Checkout code
uses: actions/checkout@v4
@@ -479,8 +465,6 @@ jobs:
if: always() && !failure() && !cancelled() &&
((needs.check-changes.outputs.main_package == 'true') || (needs.check-changes.outputs.sgl_kernel == 'true'))
runs-on: 2-gpu-runner
env:
HF_TOKEN: ${{ secrets.HF_TOKEN }}
steps:
- name: Checkout code
uses: actions/checkout@v4
@@ -538,8 +522,6 @@ jobs:
if: always() && !failure() && !cancelled() &&
((needs.check-changes.outputs.main_package == 'true') || (needs.check-changes.outputs.sgl_kernel == 'true'))
runs-on: 1-gpu-runner
env:
HF_TOKEN: ${{ secrets.HF_TOKEN }}
steps:
- name: Checkout code
uses: actions/checkout@v4
@@ -570,8 +552,6 @@ jobs:
if: always() && !failure() && !cancelled() &&
((needs.check-changes.outputs.main_package == 'true') || (needs.check-changes.outputs.sgl_kernel == 'true'))
runs-on: 2-gpu-runner
env:
HF_TOKEN: ${{ secrets.HF_TOKEN }}
steps:
- name: Checkout code
uses: actions/checkout@v4
@@ -602,8 +582,6 @@ jobs:
if: always() && !failure() && !cancelled() &&
((needs.check-changes.outputs.main_package == 'true') || (needs.check-changes.outputs.sgl_kernel == 'true'))
runs-on: 4-gpu-runner
env:
HF_TOKEN: ${{ secrets.HF_TOKEN }}
steps:
- name: Checkout code
uses: actions/checkout@v4
@@ -631,8 +609,6 @@ jobs:
if: always() && !failure() && !cancelled() &&
((needs.check-changes.outputs.main_package == 'true') || (needs.check-changes.outputs.sgl_kernel == 'true'))
runs-on: 8-gpu-runner
env:
HF_TOKEN: ${{ secrets.HF_TOKEN }}
steps:
- name: Checkout code
uses: actions/checkout@v4
@@ -660,8 +636,6 @@ jobs:
if: always() && !failure() && !cancelled() &&
((needs.check-changes.outputs.main_package == 'true') || (needs.check-changes.outputs.sgl_kernel == 'true'))
runs-on: 4-b200-runner
env:
HF_TOKEN: ${{ secrets.HF_TOKEN }}
strategy:
fail-fast: false
steps: