FilesExpand file tree

main

/

mkdocs.yml

Copy path

Blame

More file actions

Blame

More file actions

Latest commit

 

History

History

History

313 lines (309 loc) · 13.5 KB

main

/

mkdocs.yml

Copy path

Top

File metadata and controls

• Code
• Blame

313 lines (309 loc) · 13.5 KB

Raw

Copy raw file

Download raw file

Open symbols panel

Edit and raw actions

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

64

65

66

67

68

69

70

71

72

73

74

75

76

77

78

79

80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

96

97

98

99

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

186

187

188

189

190

191

192

193

194

195

196

197

198

199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222

223

224

225

226

227

228

229

230

231

232

233

234

235

236

237

238

239

240

241

242

243

244

245

246

247

248

249

250

251

252

253

254

255

256

257

258

259

260

261

262

263

264

265

266

267

268

269

270

271

272

273

274

275

276

277

278

279

280

281

282

283

284

285

286

287

288

289

290

291

292

293

294

295

296

297

298

299

300

301

302

303

304

305

306

307

308

309

310

311

312

313

# Project information

site_name: GPUStack

site_url: https://docs.gpustack.ai

site_author: GPUStack.ai

site_description: >-

GPUStack is an open-source GPU cluster manager designed for efficient AI model deployment.

It lets you run models efficiently on your own GPU hardware by choosing the best inference engines,

scheduling GPU resources, analyzing model architectures, and automatically configuring deployment parameters.

# Hooks

hooks:

- hack/mkdocs_i18n.py

- hack/mkdocs_multilingual.py

# Repository

repo_name: gpustack/gpustack

repo_url: https://github.com/gpustack/gpustack

# Copyright

copyright: Copyright © 2026 GPUStack.ai

theme:

name: material

logo: assets/logo-white.png

favicon: assets/logo.png

custom_dir: docs/overrides

palette:

- media: "(prefers-color-scheme)"

# This entry is what Material writes into the served HTML, before its

# JS picks the light/dark entry below. Leaving scheme/primary unset

# here means the page ships as Material's default indigo and repaints

# to white only once the script runs — a visible purple header flash

# on a slow connection. Stating the light values makes the initial

# paint already correct for the common case.

scheme: default

primary: white

accent: indigo

toggle:

icon: material/brightness-auto

name: Switch to light mode

- media: "(prefers-color-scheme: light)"

scheme: default

# `white` / `black` keep the header out of a saturated colour block;

# the brand hue lives in links and accents instead (see extra.css).

primary: white

accent: indigo

toggle:

icon: material/brightness-7

name: Switch to dark mode

- media: "(prefers-color-scheme: dark)"

scheme: slate

primary: black

accent: indigo

toggle:

icon: material/brightness-4

name: Switch to system preference

features:

- search.suggest

- search.highlight

- content.tabs.link

- navigation.indexes

- content.tooltips

- navigation.path

- navigation.tabs

- content.code.annotate

- content.code.copy

- content.code.select

- content.action.view

- content.action.edit

- navigation.top

- navigation.footer

- navigation.tracking

- search.share

- toc.follow

icon:

repo: fontawesome/brands/github

# `false` stops Material from emitting the fonts.googleapis.com link.

# Inter and JetBrains Mono — the pair gpustack.ai uses — are served from

# docs/assets/fonts/ instead; see the @font-face block at the top of

# stylesheets/extra.css, which also sets --md-text-font/--md-code-font

# so Material builds its stacks around them.

font: false

language: en

# A mapping rather than a list, because mkdocs.offline.yml adds a plugin to this

# set through INHERIT. A child config replaces a list wholesale but merges a

# mapping, so the list form would force that file to restate every entry below —

# and to silently lose whichever one a later edit here forgot to copy over.

plugins:

glightbox: {}

search: {}

redirects:

redirect_maps:

'installation/nvidia/installation.md': 'installation/requirements.md'

'installation/amd/installation.md': 'installation/requirements.md'

'installation/ascend/installation.md': 'installation/requirements.md'

'installation/hygon/installation.md': 'installation/requirements.md'

'installation/mthreads/installation.md': 'installation/requirements.md'

'installation/iluvatar/installation.md': 'installation/requirements.md'

'installation/cambricon/installation.md': 'installation/requirements.md'

'installation/metax/installation.md': 'installation/requirements.md'

minify:

minify_html: true

minify_css: true

minify_js: true

cache_safe: true

htmlmin_opts:

remove_comments: true

extra:

generator: false

version:

provider: mike

social:

- icon: fontawesome/brands/github

link: https://github.com/gpustack

- icon: fontawesome/brands/docker

link: https://hub.docker.com/u/gpustack

- icon: fontawesome/brands/python

link: https://pypi.org/project/gpustack/

- icon: fontawesome/brands/discord

link: https://discord.gg/VXYJzuaqwD

extra_javascript:

- image-selector/assets/extra.js

- assets/katex/katex.min.js

- assets/katex/contrib/auto-render.min.js

- javascripts/katex.js

- https://unpkg.com/tablesort@5.3.0/dist/tablesort.min.js

- javascripts/tablesort.js

extra_css:

- stylesheets/extra.css

- assets/katex/katex.min.css

markdown_extensions:

- abbr

- def_list

- attr_list

- md_in_html

- pymdownx.highlight:

linenums: true

line_spans: __span

anchor_linenums: true

pygments_lang_class: true

- pymdownx.inlinehilite

- pymdownx.superfences:

custom_fences:

- name: mermaid

class: mermaid

format: !!python/name:pymdownx.superfences.fence_code_format

- pymdownx.tabbed:

alternate_style: true

- admonition

- sane_lists

- pymdownx.details

- pymdownx.tilde

- pymdownx.betterem

- pymdownx.caret

- pymdownx.tasklist:

custom_checkbox: true

- pymdownx.arithmatex:

generic: true

- pymdownx.emoji:

emoji_generator: !!python/name:material.extensions.emoji.to_svg

emoji_index: !!python/name:material.extensions.emoji.twemoji

- pymdownx.critic

- pymdownx.keys

- pymdownx.snippets:

auto_append:

- docs/includes/abbreviations.md

edit_uri: edit/main/docs/

nav:

- Home:

- Overview: overview.md

- Quickstart: quickstart.md

- Installation:

- Requirements: installation/requirements.md

- Installation via Docker: installation/installation.md

- Installation via Helm: installation/helm.md

- Air-Gapped Installation: installation/air-gapped.md

- Uninstallation: installation/uninstallation.md

- Upgrade:

- Upgrade via Docker: upgrade/docker.md

- Upgrade via Helm: upgrade/helm.md

- Upgrade a Cluster Deployment: upgrade/cluster.md

- Migration: migration.md

- User Guide:

- Playground:

- Overview: user-guide/playground/index.md

- Chat: user-guide/playground/chat.md

- Image: user-guide/playground/image.md

- Audio: user-guide/playground/audio.md

- Embedding: user-guide/playground/embedding.md

- Rerank: user-guide/playground/rerank.md

- Decision: user-guide/playground/decision.md

- Model Catalog: user-guide/model-catalog.md

- Model Deployment Management: user-guide/model-deployment-management.md

- PD Disaggregation: user-guide/pd-disaggregation.md

- Model Route Management: user-guide/model-route-management.md

- Model Provider Management: user-guide/model-provider-management.md

- Inference Backend Management: user-guide/inference-backend-management.md

- Built-in Inference Backends: user-guide/built-in-inference-backends.md

- Cache Service Management: user-guide/cache-service-management.md

- OTA Content Updates: user-guide/ota-content-updates.md

- Compatibility Check: user-guide/compatibility-check.md

- Model File management: user-guide/model-file-management.md

- Cluster Management: user-guide/cluster-management.md

- GPU Service:

- GPU Instances: user-guide/gpuservice-instances.md

- Instance Types: user-guide/gpuservice-instance-types.md

- Instance Templates: user-guide/gpuservice-instance-templates.md

- Storage: user-guide/gpuservice-storage.md

- Storage Types: user-guide/gpuservice-storage-types.md

- SSH Public Keys: user-guide/gpuservice-ssh-public-keys.md

- Cloud Credential Management: user-guide/cloud-credential-management.md

- API Key Management: user-guide/api-key-management.md

- User Management: user-guide/user-management.md

- Usage: user-guide/usage.md

- Single Sign-On (SSO) Authentication: user-guide/sso.md

- Observability: user-guide/observability.md

- Benchmarking: user-guide/benchmarking.md

- Using Models:

- Using Large Language Models: using-models/using-large-language-models.md

- Using Vision Language Models: using-models/using-vision-language-models.md

- Using Embedding Models: using-models/using-embedding-models.md

- Using Reranker Models: using-models/using-reranker-models.md

- Using Decision Models: using-models/using-decision-models.md

- Using Image Generation Models: using-models/using-image-generation-models.md

- Using Audio Models: using-models/using-audio-models.md

- Tutorials:

- Running DeepSeek R1 671B with Distributed vLLM: tutorials/running-deepseek-r1-671b-with-distributed-vllm.md

- Running Distributed vLLM with the MultiProcessing Backend: tutorials/running-distributed-vllm-with-multiprocessing.md

- Running DeepSeek R1 671B with Distributed Ascend Mindie: tutorials/running-deepseek-r1-671b-with-distributed-ascend-mindie.md

- Inference On CPUs: tutorials/inference-on-cpus.md

- Inference with Tool Calling: tutorials/inference-with-tool-calling.md

- Using Custom Inference Backend: tutorials/using-custom-backends.md

- Adding a GPU Cluster Using DigitalOcean: tutorials/adding-gpucluster-using-digitalocean.md

- Adding a GPU Cluster Using Shuihua: tutorials/adding-gpucluster-using-shuihua.md

- Adding a GPU Cluster Using Kubernetes: tutorials/adding-gpucluster-using-kubernetes.md

- Managing Model Routes: tutorials/managing-model-routes.md

- Routing with LB Policies: tutorials/routing-with-lb-policies.md

- Serving Models with LoRA Adapters: tutorials/serving-with-lora-adapters.md

- Integrations:

- Inference APIs: integrations/inference-apis.md

- Integrate with Claude Code: integrations/integrate-with-claude-code.md

- Integrate with Dify: integrations/integrate-with-dify.md

- Integrate with RAGFlow: integrations/integrate-with-ragflow.md

- Integrate with Cherry Studio: integrations/integrate-with-cherrystudio.md

- Integrate with OpenClaw: integrations/integrate-with-openclaw.md

- Integrate with n8n: integrations/integrate-with-n8n.md

- Integrate with MaxKB: integrations/integrate-with-maxkb.md

- Integrate with NewAPI: integrations/integrate-with-newapi.md

- Integrate with LiteLLM: integrations/integrate-with-litellm.md

- Architecture: architecture.md

- Scheduler: scheduler.md

- Troubleshooting: troubleshooting.md

- FAQ: faq.md

- API Reference: api-reference.md

- CLI Reference:

- Start: cli-reference/start.md

- Download Tools: cli-reference/download-tools.md

- Reload Config: cli-reference/reload-config.md

- List Images: cli-reference/list-images.md

- Save Images: cli-reference/save-images.md

- Load Images: cli-reference/load-images.md

- Copy Images: cli-reference/copy-images.md

- Environment Variables: environment-variables.md

- Inference Performance Lab:

- Overview: performance-lab/overview.md

- Optimizing Throughput:

- Qwen3.5-35B-A3B:

- H200: performance-lab/qwen3.5-35b-a3b/h200.md

- Qwen3.5-9B:

- H100: performance-lab/qwen3.5-9b/h100.md

- GLM-4.5-Air:

- A100: performance-lab/glm-4.5-air/a100.md

- H100: performance-lab/glm-4.5-air/h100.md

- GLM-4.x:

- A100: performance-lab/glm-4.x/a100.md

- H100: performance-lab/glm-4.x/h100.md

- H200: performance-lab/glm-4.x/h200.md

- GPT-OSS-20B:

- A100: performance-lab/gpt-oss-20b/a100.md

- H100: performance-lab/gpt-oss-20b/h100.md

- GPT-OSS-120B:

- A100: performance-lab/gpt-oss-120b/a100.md

- H100: performance-lab/gpt-oss-120b/h100.md

- DeepSeek-R1:

- H200: performance-lab/deepseek-r1/h200.md

- DeepSeek-V3.2:

- H200: performance-lab/deepseek-v3.2/h200.md

- GLM-5.2:

- H200: performance-lab/glm-5.2/h200.md

- Qwen3-8B:

- 910B: performance-lab/qwen3-8b/910b.md

- Qwen3-14B:

- A100: performance-lab/qwen3-14b/a100.md

- H100: performance-lab/qwen3-14b/h100.md

- Qwen3-32B:

- A100: performance-lab/qwen3-32b/a100.md

- H100: performance-lab/qwen3-32b/h100.md

- Qwen3-30B-A3B:

- 910B: performance-lab/qwen3-30b-a3b/910b.md

- Qwen3-235B-A22B:

- A100: performance-lab/qwen3-235b-a22b/a100.md

- H100: performance-lab/qwen3-235b-a22b/h100.md

- Optimizing Latency:

- Qwen3.5-35B-A3B:

- H200: performance-lab/qwen3.5-35b-a3b/h200-latency.md

- Qwen3.5-9B:

- H100: performance-lab/qwen3.5-9b/h100-latency.md

- Qwen3-8B:

- H100: performance-lab/qwen3-8b/h100-latency.md

- References:

- The Impact of Quantization on vLLM Inference Performance: performance-lab/references/the-impact-of-quantization-on-vllm-inference-performance.md

- Evaluating LMCache Prefill Acceleration in vLLM: performance-lab/references/evaluating-lmcache-prefill-acceleration-in-vllm.md

- Container Image Selector: image-selector/index.html