weaviate

Helm chart

Last changed

Request a free trial

Contact our team to test out this Helm chart and related images for free. Please also indicate any other images you would like to evaluate.

Tag:

image:

# registry where weaviate image is stored

registry: cgr.dev

# Tag of weaviate image to deploy

# Note: We strongly recommend you overwrite this value in your own values.yaml.

# Otherwise a mere upgrade of the chart could lead to an unexpected upgrade

# of weaviate. In accordance with Infra-as-code, you should pin this value

# down and only change it if you explicitly want to upgrade the Weaviate

# version.

tag: 1.39.4-r0@sha256:a5bc025dcda857cb479e8be181ba00fe0a7942f4be5a6e2a53c481a8a44a7fc3

repo: scratch-images/test-tmp/weaviate

# Image pull policy: https://kubernetes.io/docs/concepts/containers/images/#image-pull-policy

pullPolicy: IfNotPresent

pullSecrets: []

# overwrite command and args if you want to run specific startup scripts, for

# example setting the nofile limit

command: ["/bin/weaviate"]

args:

- '--host'

- '0.0.0.0'

- '--port'

- '8080'

- '--scheme'

- 'http'

- '--config-file'

- '/weaviate-config/conf.yaml'

- --read-timeout=60s

- --write-timeout=60s

# below is an example that can be used to set an arbitrary nofile limit at

# startup:

# command:

# - "/bin/sh"

# args:

# - "-c"

# - "ulimit -n 65535 && /bin/weaviate --host 0.0.0.0 --port 8080 --scheme http --config-file /weaviate-config/conf.yaml"

# it is possible to change the sysctl's 'vm.max_map_count' using initContainer for Weaviate,

# the init Container runs before Weaviate Container and sets the value for the WHOLE node

# to the one provided below.

# it is possible to run additional initContainer before Weaviate is up and running. You can specify the

# containers as a list in `extraInitContainers`, exactly how they are defined in a kubernetes manifest:

# https://kubernetes.io/docs/concepts/workloads/pods/init-containers/

initContainers:

sysctlInitContainer:

enabled: true

sysctlVmMaxMapCount: 524288

image:

registry: cgr.dev

repo: scratch-images/test-tmp/busybox

tag: glibc-1.38.0-r2@sha256:2a9dc729fa177c41c3fb59a2d6c4a8b0d5dc51f2f999854a3fa2b46171b72041

pullPolicy: IfNotPresent

ensureFileOwnershipContainer:

# This init container sets the file ownerships of /var/lib/weaviate directory to the ones set in

# containerSecurityContext.runAsUser and containerSecurityContext.fsGroup settings to ensure that Weaviate is able

# to start in unprivileged configuration.

# Enable this init container only if Weaviate was configured previously without security context

# and now containerSecurityContext is provided to run Weaviate container with non-root user.

# Please be sure to set at least containerSecurityContext.runAsUser and containerSecurityContext.fsGroup.

enabled: false

extraInitContainers: {}

# - image: some-image

# name: some-name

# Scale replicas of Weaviate. Note that as of v1.8.0 dynamic scaling is limited

# to cases where no data is imported yet. Scaling down after importing data may

# break usability. Full dynamic scalability will be added in a future release.

replicas: 1

# Define how pods will be created. Possible values: OrderedReady | Parallel

# OrderedReady - pods will be created one after another

# Parallel - all pods will be created at once

podManagementPolicy: Parallel

updateStrategy:

type: RollingUpdate

# This setting is only available in K8s v1.24 and higher.

# Setting maxUnavailable to 100% results in removing all of the pods

# and re-creating them in parallel all at once.

# rollingUpdate:

# maxUnavailable: 100%

resources: {}

# requests:

# cpu: '500m'

# memory: '300Mi'

# limits:

# cpu: '1000m'

# memory: '1Gi'

# security Context for the Weaviate Pods. The configurations are the same as setting them

# as described here: https://kubernetes.io/docs/tasks/configure-pod-container/security-context/

securityContext: {}

# Security context for the Weaviate container. Override overlapping settings made at the Pod level.

containerSecurityContext: {}

# runAsUser: 1000

# runAsGroup: 1000

# fsGroup: 1000

# fsGroupChangePolicy: "OnRootMismatch"

# runAsNonRoot: true

# allowPrivilegeEscalation: false

# privileged: false

# readOnlyRootFilesystem: true

100

101

# Add a service account to the Weaviate pods if you need Weaviate to have permissions to

102

# access kubernetes resources or cloud provider resources. For example for it to have

103

# access to a backup up bucket, or if you want to restrict Weaviate pod in any way.

104

# By default, use the default ServiceAccount

105

serviceAccountName:

106

# Kubernetes Cluster domain name, used for resolving intra-cluster requests, i.e

107

# between instances of weaviate.

108

# Note: The final '.' on the end of the hostname makes it a FQDN, and is required for

109

# DNS to resolve in all kubernetes environments.

110

# See https://github.com/weaviate/weaviate-helm/issues/175 for details.

111

clusterDomain: cluster.local.

112

# The Persistent Volume Claim settings for Weaviate. The given name will be used as the

113

# prefix for the volumes (e.g. first replica will use {{ .Values.storage.name }}-weaviate-0)

114

storage:

115

116

size: 32Gi

117

storageClassName: ""

118

# The service controls how weaviate is exposed to the outside world. If you

119

# don't want a public load balancer, you can also choose 'ClusterIP' to make

120

# weaviate only accessible within your cluster.

121

service:

122

123

ports:

124

- name: http

125

protocol: TCP

126

port: 80

127

# Target port is going to be the same for every port

128

type: LoadBalancer

129

loadBalancerSourceRanges: []

130

# optionally set cluster IP if you want to set a static IP

131

clusterIP:

132

annotations: {}

133

# The service controls how weaviate gRPC endpoint is exposed to the outside world.

134

# If you don't want a public load balancer, you can also choose 'ClusterIP' or `NodePort`

135

# to make weaviate gRPC port be only accessible within your cluster.

136

# This service is by default enabled but if you don't want it to be deployed in your

137

# environment then it can be disabled by setting enabled: false option.

138

grpcService:

139

enabled: true

140

141

ports:

142

- name: grpc

143

protocol: TCP

144

port: 50051

145

# Target port is going to be the same for every port

146

type: LoadBalancer

147

loadBalancerSourceRanges: []

148

# optionally set cluster IP if you want to set a static IP

149

clusterIP:

150

annotations: {}

151

# MCP (Model Context Protocol) allows LLM clients to interact with Weaviate.

152

# MCP is served on the main REST port at /v1/mcp, no separate port is needed.

153

# This is by default disabled. Set enabled to true to enable the MCP server.

154

mcp:

155

enabled: false

156

# Write access allows MCP clients to create/update objects.

157

# Disabled by default for safety.

158

writeAccessEnabled: false

159

# Custom MCP server configuration for tool descriptions.

160

# When set, a ConfigMap is created and mounted automatically.

161

# Example:

162

# config:

163

# tools:

164

# weaviate-collections-get-config:

165

# description: "Retrieves the schema configuration for one or all collections in the Weaviate database."

166

# weaviate-tenants-list:

167

# description: "Lists all tenants for a specific multi-tenant collection."

168

# weaviate-query-hybrid:

169

# description: "Performs a hybrid search combining vector similarity and keyword matching (BM25) on a collection."

170

# weaviate-objects-upsert:

171

# description: "Upserts (inserts or updates) one or more objects into a collection in batch."

172

config:

173

# The service monitor defines prometheus monitoring for a set of services

174

# https://github.com/prometheus-operator/prometheus-operator/blob/main/Documentation/api.md#monitoring.coreos.com/v1.ServiceMonitor

175

# Make sure to set the following prometheus values if deploying observability with the kube-prometheus-stack helm chart:

176

# - prometheus.prometheusSpec.serviceMonitorSelectorNilUsesHelmValues: false

177

serviceMonitor:

178

enabled: false

179

interval: 30s

180

scrapeTimeout: 10s

181

# Adjust liveness, readiness and startup probes configuration

182

# below is an example that can be used to switch the probeType to exec command

183

# readinessProbe: # (Compatible with liveness, readiness and startup probe configurations)

184

# probeType: exec

185

# probe:

186

# exec:

187

# command: ["/bin/sh", "-c", "wget --spider --server-response --tries=1 --timeout=30 -o /dev/null localhost:8080/v1/.well-known/ready"]

188

startupProbe:

189

# For kubernetes versions prior to 1.18 startupProbe is not supported thus can be disabled.

190

enabled: false

191

probeType: httpGet

192

probe:

193

httpGet:

194

path: /v1/.well-known/ready

195

port: 8080

196

initialDelaySeconds: 300

197

periodSeconds: 60

198

failureThreshold: 50

199

successThreshold: 1

200

timeoutSeconds: 3

201

livenessProbe:

202

probeType: httpGet

203

probe:

204

httpGet:

205

path: /v1/.well-known/live

206

port: 8080

207

initialDelaySeconds: 900

208

periodSeconds: 10

209

failureThreshold: 30

210

successThreshold: 1

211

timeoutSeconds: 3

212

readinessProbe:

213

probeType: httpGet

214

probe:

215

httpGet:

216

path: /v1/.well-known/ready

217

port: 8080

218

initialDelaySeconds: 3

219

periodSeconds: 10

220

failureThreshold: 3

221

successThreshold: 1

222

timeoutSeconds: 3

223

terminationGracePeriodSeconds: 600

224

# Weaviate Config

225

226

# The following settings allow you to customize Weaviate to your needs, for

227

# example set authentication and authorization options. See weaviate docs

228

# (https://www.weaviate.io/developers/weaviate/) for all

229

# configuration.

230

authentication:

231

anonymous_access:

232

enabled: true

233

# This configuration allows to add API keys to Weaviate. This configuration allows only

234

# plain text API Keys, if you want to store the API Keys in a Kubernetes secret you can

235

# configure the same configuration with ENV Vars. Read the `env` section below on what

236

# needs to be configured. If using ENV Vars over this make sure to comment out the whole

237

# `apikey` section (as it is by default). ENV Vars has priority over this config.

238

# apikey:

239

# enabled: false

240

# # Any number of allowed API Keys as plain text

241

# allowed_keys:

242

# - readOnly-plainText-API-Key

243

# - admin-plainText-API-Key

244

# # You can either set a single user for all the listed Allowed API keys OR

245

# # one user per API Key, i.e. length(apikey.allowed_keys) == length(apikey.users) OR

246

# # length(apikey.users) == 1

247

# # NOTE: Make sure the lister Users are added to the Authorization as well.

248

# users:

249

# - api-key-user-readOnly

250

# - api-key-user-admin

251

oidc:

252

enabled: false

253

# issuer: ''

254

# username_claim: ''

255

# groups_claim: ''

256

# client_id: ''

257

# skip_client_id_check: false

258

# namespace_claim: ''

259

# global_principal_claim: ''

260

authorization:

261

rbac:

262

enabled: false

263

# root_users:

264

# - admin_user1

265

# - admin_user2

266

admin_list:

267

enabled: false

268

# users:

269

# - admin_user1

270

# - admin_user2

271

# - api-key-user-admin

272

# read_only_users:

273

# - readonly_user1

274

# - readonly_user2

275

# - api-key-user-readOnly

276

query_defaults:

277

limit: 100

278

debug: false

279

# Enable namespaces support in the weaviate cluster. Once enabled, the cluster can't be migrated to a non-namespaced cluster and vice versa.

280

namespaces:

281

enabled: false

282

# Insert any custom environment variables or envSecrets by putting the exact name

283

# and desired value into the settings below. Any env name passed will be automatically

284

# set for the statefulSet.

285

env:

286

CLUSTER_GOSSIP_BIND_PORT: 7000

287

CLUSTER_DATA_BIND_PORT: 7001

288

# Set RAFT cluster expected number of voter nodes at bootstrap.

289

# By default helm automatically sets this value based on the cluster size.

290

# RAFT_BOOTSTRAP_EXPECT: 1

291

292

# Set RAFT cluster bootstrap timeout (in seconds), default is 600 (seconds)

293

# which should be sufficient for most of the deployments.

294

RAFT_BOOTSTRAP_TIMEOUT: 600

295

# Set manually RAFT voter nodes.

296

# RAFT_JOIN value is automatically generated by "raft_configuration"

297

# template, but if someone wants to set this value manually then it can be done

298

# by setting RAFT_JOIN environment variable, example: RAFT_JOIN: "weaviate-0,weaviate-1"

299

# Please notice that in this case RAFT_BOOTSTRAP_EXPECT setting needs to be also adjusted manually

300

# to match the number of RAFT voters, so if there are 2 nodes set using RAFT_JOIN variable

301

# then RAFT_BOOTSTRAP_EXPECT needs to be equal 2 also.

302

# RAFT_JOIN: "weaviate-0"

303

304

# Set to true if voters nodes should handle only schema. With this setting enabled

305

# voter nodes will not accept any data, one needs to resize the cluster using replicas

306

# setting so that replicas > voters.

307

# RAFT_METADATA_ONLY_VOTERS: false

308

309

# RAFT_ENABLE_FQDN_RESOLVER setting changes the node name to node ip resolution to use DNS lookups

310

# instead of memberlist lookup. That means that when weaviate raft component wants to contact `weaviate-0`

311

# it's going to lookup the dns name `weaviate-0` instead of looking for the node-id in memberlist.

312

# This is particularly useful if running in an environment where you're using services (for example k8s)

313

# where the IP of the services is different from the actual node IP, but it proxies the connection to the node.

314

# RAFT_ENABLE_FQDN_RESOLVER: false

315

316

# RAFT_FQDN_RESOLVER_TLD setting acts in combination with RAFT_ENABLE_FQDN_RESOLVER and is appended

317

# in the format "[node-id].[tld]" when resolving a node-id to an ip.

318

# RAFT_FQDN_RESOLVER_TLD: "weaviate-0."

319

320

# The aggressiveness of the Go Garbage Collector. 100 is the default value.

321

GOGC: 100

322

# Expose metrics on port 2112 for Prometheus to scrape

323

PROMETHEUS_MONITORING_ENABLED: false

324

PROMETHEUS_MONITORING_GROUP: false

325

# Set a MEM limit for the Weaviate Pod so it can help you both increase GC-related

326

# performance as well as avoid GC-related out-of-memory (“OOM”) situations

327

# GOMEMLIMIT: 6GiB

328

329

# Maximum results Weaviate can query with/without pagination

330

# NOTE: Affects performance, do NOT set to a very high value.

331

# The default is 100K

332

QUERY_MAXIMUM_RESULTS: 100000

333

# whether to enable vector dimensions tracking metric

334

TRACK_VECTOR_DIMENSIONS: false

335

# whether to re-index/-compute the vector dimensions metric (needed if upgrading from weaviate < v1.16.0)

336

REINDEX_VECTOR_DIMENSIONS_AT_STARTUP: false

337

##########################

338

# API Keys with ENV Vars #

339

##########################

340

# If using ENV Vars to set up API Keys make sure to have `authentication.apikey` block commented out

341

# to avoid any future changes. ENV Vars has priority over the config above `authentication.apikey`.

342

# If using `authentication.apikey `the below ENV Vars will be used because they have priority,

343

# so comment them out to avoid any future changes. The same applies for the RBAC configuration

344

# under the authorization block.

345

# Enables API key authentication. If it is set to 'false' the AUTHENTICATION_APIKEY_ALLOWED_KEYS

346

# and AUTHENTICATION_APIKEY_USERS will not have any effect.

347

# AUTHENTICATION_APIKEY_ENABLED: 'true'

348

349

# List one or more keys, separated by commas. Each key corresponds to a specific user identity below.

350

# If you want to use a kubernetes secret for the API Keys comment out this Variable and use the one in `envSecrets` below

351

# AUTHENTICATION_APIKEY_ALLOWED_KEYS: 'jane-secret-key,ian-secret-key' (plain text)

352

353

# List one or more user identities, separated by commas. You can have only one User for all the keys or one user per key.

354

# The User/s can be a simple name or an email, no matter if it exists or not.

355

# NOTE: Make sure to add the users to the authorization above overwise they will not be allowed to interact with Weaviate.

356

# AUTHENTICATION_APIKEY_USERS: 'jane@doe.com,ian-smith'

357

358

# Enabling RBAC authorization. It is mutually exclusive with the AUTHORIZATION_ADMIN_LISTS variable. Either RBAC or the

359

# admin lists mechanism can be used.

360

# AUTHORIZATION_ENABLE_RBAC: "true"

361

362

# Users with admin's RBAC role. List one or more user identities, separated by commas, which will

363

# have the admin role assigned to. This role provides all permissions to the user, but it's required at least

364

# in one of the user for managing the cluster.

365

# AUTHORIZATION_ADMIN_USERS: "admin-user"

366

367

# Users with viewer's RBAC role. List one or more user identities, separated by commas, which will

368

# have the viewer role assigned to. This role allows read permissions in all different areas. Once assigned via

369

# config, it can't be revoked via API AuthZ calls.

370

# AUTHORIZATION_VIEWER_USERS: "viewer-user"

371

envSecrets:

372

# create a Kubernetes secret with AUTHENTICATION_APIKEY_ALLOWED_KEYS key and its respective value

373

# AUTHENTICATION_APIKEY_ALLOWED_KEYS: name-of-the-k8s-secret-containing-the-comma-separated-api-keys

374

375

# Configure offload providers

376

offload:

377

s3:

378

enabled: false

379

# If one is using AWS EKS and has already configured K8s Service Account

380

# that holds the AWS credentials one can pass a name of that service account

381

# here using this setting.

382

# NOTE: the root `serviceAccountName` config has priority over this one, and

383

# if the root one is set this one will NOT overwrite it. This one is here for

384

# backwards compatibility.

385

serviceAccountName:

386

envconfig:

387

# Configure bucket where data should be saved, this setting is mandatory

388

OFFLOAD_S3_BUCKET: weaviate-offload

389

# Optional setting. Defaults to AWS S3 (s3.amazonaws.com).

390

# Set this option if you have a MinIO storage configured in your environment

391

# and want to use it instead of the AWS S3.

392

# OFFLOAD_S3_ENDPOINT: custom.minio.endpoint.address

393

394

# Optional setting. Defaults to true.

395

# Set this option if you don't want to use SSL.

396

# OFFLOAD_S3_USE_SSL: true

397

398

# Optional setting. Defaults to false.

399

# Set this option if you wan't Weaviate to create

400

# the bucket used for offloading tenants. Otherwise,

401

# if set to false Weaviate expects the bucket to be

402

# already created with the OFFLOAD_S3_BUCKET name

403

# OFFLOAD_S3_BUCKET_AUTO_CREATE: true

404

405

# You can pass environment AWS settings here:

406

# Define the region

407

# AWS_REGION: eu-west-1

408

# For Weaviate to be able to create bucket objects it needs a user credentials to authenticate to AWS.

409

# The User must have permissions to read/create/delete bucket objects.

410

# You can pass the User credentials (access-key id and access-secret-key) in 2 ways:

411

# 1. by setting the AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY plain values in the `secrets` section below

412

# this chart will create a kubernetes secret for you with these key-values pairs

413

# 2. create Kubernetes secret/s with AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY keys and their respective values

414

# Set the Key and the secret where it is set in `envSecrets` section below

415

secrets: {}

416

# AWS_ACCESS_KEY_ID: access-key-id (plain text)

417

# AWS_SECRET_ACCESS_KEY: secret-access-key (plain text)

418

419

# If one has already defined secrets with AWS credentials one can pass them using

420

# this setting:

421

envSecrets: {}

422

# AWS_ACCESS_KEY_ID: name-of-the-k8s-secret-containing-the-key-id

423

# AWS_SECRET_ACCESS_KEY: name-of-the-k8s-secret-containing-the-key

424

# Configure usage metrics providers

425

usage:

426

gcs:

427

enabled: false

428

envconfig:

429

# Configure bucket where usage metrics should be saved, this setting is mandatory

430

USAGE_GCS_BUCKET: weaviate-usage

431

# Optional setting. Defaults to empty string.

432

# Set this option if you want to save usage metrics to a given location

433

# inside the bucket

434

# USAGE_GCS_PREFIX: path/inside/bucket

435

436

# Optional setting. Defaults to 1h.

437

# Set this option if you want collection interval to be different than 1 hour

438

# USAGE_SCRAPE_INTERVAL: 1h

439

s3:

440

enabled: false

441

envconfig:

442

# Configure bucket where usage metrics should be saved, this setting is mandatory

443

USAGE_S3_BUCKET: weaviate-usage

444

# Optional setting. Defaults to empty string.

445

# Set this option if you want to save usage metrics to a given location

446

# inside the bucket

447

# USAGE_S3_PREFIX: path/inside/bucket

448

449

# Optional setting. Defaults to 1h.

450

# Set this option if you want collection interval to be different than 1 hour

451

# USAGE_SCRAPE_INTERVAL: 1h

452

# Configure backup providers

453

backups:

454

# The backup-filesystem module enables creation of the DB backups in

455

# the local filesystem

456

filesystem:

457

enabled: false

458

envconfig:

459

# Configure folder where backups should be saved

460

BACKUP_FILESYSTEM_PATH: /tmp/backups

461

s3:

462

enabled: false

463

# If one is using AWS EKS and has already configured K8s Service Account

464

# that holds the AWS credentials one can pass a name of that service account

465

# here using this setting.

466

# NOTE: the root `serviceAccountName` config has priority over this one, and

467

# if the root one is set this one will NOT overwrite it. This one is here for

468

# backwards compatibility.

469

serviceAccountName:

470

envconfig:

471

# Configure bucket where backups should be saved, this setting is mandatory

472

BACKUP_S3_BUCKET: weaviate-backups

473

# Optional setting. Defaults to empty string.

474

# Set this option if you want to save backups to a given location

475

# inside the bucket

476

# BACKUP_S3_PATH: path/inside/bucket

477

478

# Optional setting. Defaults to AWS S3 (s3.amazonaws.com).

479

# Set this option if you have a MinIO storage configured in your environment

480

# and want to use it instead of the AWS S3.

481

# BACKUP_S3_ENDPOINT: custom.minio.endpoint.address

482

483

# Optional setting. Defaults to true.

484

# Set this option if you don't want to use SSL.

485

# BACKUP_S3_USE_SSL: true

486

487

# You can pass environment AWS settings here:

488

# Define the region

489

# AWS_REGION: eu-west-1

490

# For Weaviate to be able to create bucket objects it needs a user credentials to authenticate to AWS.

491

# The User must have permissions to read/create/delete bucket objects.

492

# You can pass the User credentials (access-key id and access-secret-key) in 2 ways:

493

# 1. by setting the AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY plain values in the `secrets` section below

494

# this chart will create a kubernetes secret for you with these key-values pairs

495

# 2. create Kubernetes secret/s with AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY keys and their respective values

496

# Set the Key and the secret where it is set in `envSecrets` section below

497

secrets: {}

498

# AWS_ACCESS_KEY_ID: access-key-id (plain text)

499

# AWS_SECRET_ACCESS_KEY: secret-access-key (plain text)

500

501

# If one has already defined secrets with AWS credentials one can pass them using

502

# this setting:

503

envSecrets: {}

504

# AWS_ACCESS_KEY_ID: name-of-the-k8s-secret-containing-the-key-id

505

# AWS_SECRET_ACCESS_KEY: name-of-the-k8s-secret-containing-the-key

506

gcs:

507

enabled: false

508

envconfig:

509

# Configure bucket where backups should be saved, this setting is mandatory

510

BACKUP_GCS_BUCKET: weaviate-backups

511

# Optional setting. Defaults to empty string.

512

# Set this option if you want to save backups to a given location

513

# inside the bucket

514

# BACKUP_GCS_PATH: path/inside/bucket

515

516

# You can pass environment Google settings here:

517

# Define the project

518

# GOOGLE_CLOUD_PROJECT: project-id

519

# For Weaviate to be able to create bucket objects it needs a ServiceAccount credentials to authenticate to GCP.

520

# The ServiceAccount must have permissions to read/create/delete bucket objects.

521

# You can pass the ServiceAccount credentials (as JSON) in 2 ways:

522

# 1. by setting the GOOGLE_APPLICATION_CREDENTIALS json as plain text in the `secrets` section below

523

# this chart will create a kubernetes secret for you with this key-values pairs

524

# 2. create a Kubernetes secret with GOOGLE_APPLICATION_CREDENTIALS key and its respective value

525

# Set the Key and the secret where it is set in `envSecrets` section below

526

secrets: {}

527

# GOOGLE_APPLICATION_CREDENTIALS: credentials-json-string (plain text)

528

529

# If one has already defined a secret with GOOGLE_APPLICATION_CREDENTIALS one can pass them using

530

# this setting:

531

envSecrets: {}

532

# GOOGLE_APPLICATION_CREDENTIALS: name-of-the-k8s-secret-containing-the-key

533

azure:

534

enabled: false

535

envconfig:

536

# Configure container where backups should be saved, this setting is mandatory

537

BACKUP_AZURE_CONTAINER: weaviate-backups

538

# Optional setting. Defaults to empty string.

539

# Set this option if you want to save backups to a given location

540

# inside the container

541

# BACKUP_AZURE_PATH: path/inside/container

542

# For Weaviate to be able to create container objects it needs a user credentials to authenticate to Azure Storage.

543

# The User must have permissions to read/create/delete container objects.

544

# You can pass the User credentials (account-name id and account-key or connection-string) in 2 ways:

545

# 1. by setting the AZURE_STORAGE_ACCOUNT and AZURE_STORAGE_KEY

546

# or AZURE_STORAGE_CONNECTION_STRING plain values in the `secrets` section below

547

# this chart will create a kubernetes secret for you with these key-values pairs

548

# 2. create Kubernetes secret/s with AZURE_STORAGE_ACCOUNT and AZURE_STORAGE_KEY

549

# or AZURE_STORAGE_CONNECTION_STRING and their respective values

550

# Set the Key and the secret where it is set in `envSecrets` section below

551

secrets: {}

552

# AZURE_STORAGE_ACCOUNT: account-name (plain text)

553

# AZURE_STORAGE_KEY: account-key (plain text)

554

# AZURE_STORAGE_CONNECTION_STRING: connection-string (plain text)

555

556

# If one has already defined secrets with Azure Storage credentials one can pass them using

557

# this setting:

558

envSecrets: {}

559

# AZURE_STORAGE_ACCOUNT: name-of-the-k8s-secret-containing-the-account-name

560

# AZURE_STORAGE_KEY: name-of-the-k8s-secret-containing-account-key

561

# AZURE_STORAGE_CONNECTION_STRING: name-of-the-k8s-secret-containing-connection-string

562

# Configure collection export

563

collectionExport:

564

enabled: false

565

envconfig:

566

# Configure bucket where exports should be saved, this setting is mandatory.

567

# The bucket must exist before enabling collection export, otherwise exports will fail.

568

EXPORT_DEFAULT_BUCKET: weaviate-export

569

# Required setting. Bucket path in which to save exports. Defaults to empty string.

570

# Set this option if you want to save exports to a given path inside the bucket. Must be a valid bucket path.

571

EXPORT_DEFAULT_PATH: ""

572

# Optional setting. Defaults to 0 (GOMAXPROCS at runtime).

573

# Set this option to control the number of concurrent scan workers per export.

574

# EXPORT_PARALLELISM: 0

575

# modules are extensions to Weaviate, they can be used to support various

576

# ML-models, but also other features unrelated to model inference.

577

# An inference/vectorizer module is not required, you can also run without any

578

# modules and import your own vectors.

579

modules:

580

# The text2vec-contextionary module uses a fastText-based vector-space to

581

# derive vector embeddings for your objects. It is very efficient on CPUs,

582

# but in some situations it cannot reach the same level of accuracy as

583

# transformers-based models.

584

text2vec-contextionary:

585

# Enable deployment of this module

586

enabled: false

587

# The configuration below is ignored if enabled==false

588

fullnameOverride: contextionary

589

tag: en0.16.0-v1.2.1

590

repo: semitechnologies/contextionary

591

registry: cr.weaviate.io

592

replicas: 1

593

strategy:

594

type: RollingUpdate

595

imagePullPolicy: IfNotPresent

596

imagePullSecrets: []

597

priorityClassName: ""

598

livenessProbe:

599

initialDelaySeconds: 120

600

periodSeconds: 3

601

timeoutSeconds: 3

602

readinessProbe:

603

initialDelaySeconds: 120

604

periodSeconds: 3

605

envconfig:

606

occurrence_weight_linear_factor: 0.75

607

neighbor_occurrence_ignore_percentile: 5

608

enable_compound_splitting: false

609

extensions_storage_mode: weaviate

610

resources: {}

611

# requests:

612

# cpu: '500m'

613

# memory: '500Mi'

614

# limits:

615

# cpu: '1000m'

616

# memory: '5000Mi'

617

618

# security Context for the Contextionary Pods. The configurations are the same as setting them

619

# as described here: https://kubernetes.io/docs/tasks/configure-pod-container/security-context/

620

securityContext: {}

621

# It is possible to add a ServiceAccount to this module's Pods, it can be

622

# used in cases where the module is in a private registry and you want to

623

# give access to the registry only to this pod.

624

# NOTE: if not set the root `serviceAccountName` config will be used.

625

serviceAccountName:

626

# You can guide where the pods are scheduled on a per-module basis,

627

# as well as for Weaviate overall. Each module accepts nodeSelector,

628

# tolerations, and affinity configuration. If it is set on a per-

629

# module basis, this configuration overrides the global config.

630

nodeSelector:

631

tolerations:

632

affinity:

633

# The text2vec-transformers modules uses neural networks, such as BERT,

634

# DistilBERT, etc. to dynamically compute vector embeddings based on the

635

# sentence's context. It is very slow on CPUs and should run with

636

# CUDA-enabled GPUs for optimal performance.

637

text2vec-transformers:

638

# Enable deployment of this module

639

enabled: false

640

# You can set directly an inference URL of this module without deploying it with this release.

641

# You can do so by setting a value for the `inferenceUrl` here AND by setting the `enable` to `false`

642

inferenceUrl: {}

643

# The configuration below is ignored if enabled==false

644

645

# replace with model of choice, see

646

# https://weaviate.io/developers/weaviate/modules/retriever-vectorizer-modules/text2vec-transformers

647

# for all supported models or build your own container.

648

tag: sentence-transformers-paraphrase-multilingual-MiniLM-L12-v2

649

repo: semitechnologies/transformers-inference

650

registry: cr.weaviate.io

651

replicas: 1

652

strategy:

653

type: RollingUpdate

654

imagePullPolicy: IfNotPresent

655

imagePullSecrets: []

656

priorityClassName: ""

657

fullnameOverride: transformers-inference

658

# Deprecated setting use initialDelaySeconds instead in each probe instead

659

# probeInitialDelaySeconds: 120

660

livenessProbe:

661

initialDelaySeconds: 120

662

periodSeconds: 3

663

timeoutSeconds: 3

664

readinessProbe:

665

initialDelaySeconds: 120

666

periodSeconds: 3

667

envconfig:

668

# enable for CUDA support. Your K8s cluster needs to be configured

669

# accordingly and you need to explicitly set GPU requests & limits below

670

enable_cuda: false

671

# only used when CUDA is enabled

672

nvidia_visible_devices: all

673

nvidia_driver_capabilities: compute,utility

674

# only used when CUDA is enabled

675

ld_library_path: /usr/local/nvidia/lib64

676

# enable this setting to use all avaialble cuda cores

677

use_sentence_transformers_multi_process: false

678

# set it to true to enable simple in memory request caching

679

enable_cache: false

680

resources: {}

681

# requests:

682

# cpu: '1000m'

683

# memory: '3000Mi'

684

# enable if running with CUDA support

685

# nvidia.com/gpu: 1

686

# limits:

687

# cpu: '1000m'

688

# memory: '5000Mi'

689

# enable if running with CUDA support

690

# nvidia.com/gpu: 1

691

692

# security Context for the Transformers Pods. The configurations are the same as setting them

693

# as described here: https://kubernetes.io/docs/tasks/configure-pod-container/security-context/

694

# applies to passageQueryService below

695

securityContext: {}

696

# It is possible to add a ServiceAccount to this module's Pods, it can be

697

# used in cases where the module is in a private registry and you want to

698

# give access to the registry only to this pod.

699

# NOTE: if not set the root `serviceAccountName` config will be used.

700

serviceAccountName:

701

# You can guide where the pods are scheduled on a per-module basis,

702

# as well as for Weaviate overall. Each module accepts nodeSelector,

703

# tolerations, and affinity configuration. If it is set on a per-

704

# module basis, this configuration overrides the global config.

705

nodeSelector:

706

tolerations:

707

affinity:

708

passageQueryServices:

709

passage:

710

enabled: false

711

# You can set directly an inference URL of this module without deploying it with this release.

712

# You can do so by setting a value for the `inferenceUrl` here AND by setting the `enable` to `false`

713

inferenceUrl: {}

714

# The configuration below is ignored if enabled==false

715

tag: facebook-dpr-ctx_encoder-single-nq-base

716

repo: semitechnologies/transformers-inference

717

registry: cr.weaviate.io

718

imagePullPolicy: IfNotPresent

719

imagePullSecrets: []

720

priorityClassName: ""

721

replicas: 1

722

strategy:

723

type: RollingUpdate

724

fullnameOverride: transformers-inference-passage

725

livenessProbe:

726

initialDelaySeconds: 120

727

periodSeconds: 3

728

timeoutSeconds: 3

729

readinessProbe:

730

initialDelaySeconds: 120

731

periodSeconds: 3

732

envconfig:

733

# enable for CUDA support. Your K8s cluster needs to be configured

734

# accordingly and you need to explicitly set GPU requests & limits below

735

enable_cuda: false

736

# only used when CUDA is enabled

737

nvidia_visible_devices: all

738

nvidia_driver_capabilities: compute,utility

739

# only used when CUDA is enabled

740

ld_library_path: /usr/local/nvidia/lib64

741

# enable this setting to use all avaialble cuda cores

742

use_sentence_transformers_multi_process: false

743

# set it to true to enable simple in memory request caching

744

enable_cache: false

745

resources: {}

746

# requests:

747

# cpu: '1000m'

748

# memory: '3000Mi'

749

# enable if running with CUDA support

750

# nvidia.com/gpu: 1

751

# limits:

752

# cpu: '1000m'

753

# memory: '5000Mi'

754

# enable if running with CUDA support

755

# nvidia.com/gpu: 1

756

757

# security Context for the Transformers Pods. The configurations are the same as setting them

758

# as described here: https://kubernetes.io/docs/tasks/configure-pod-container/security-context/

759

# applies to passageQueryService below

760

securityContext: {}

761

# It is possible to add a ServiceAccount to this module's Pods, it can be

762

# used in cases where the module is in a private registry and you want to

763

# give access to the registry only to this pod.

764

# NOTE: if not set the root `serviceAccountName` config will be used.

765

serviceAccountName:

766

# You can guide where the pods are scheduled on a per-module basis,

767

# as well as for Weaviate overall. Each module accepts nodeSelector,

768

# tolerations, and affinity configuration. If it is set on a per-

769

# module basis, this configuration overrides the global config.

770

nodeSelector:

771

tolerations:

772

affinity:

773

query:

774

enabled: false

775

# You can set directly an inference URL of this module without deploying it with this release.

776

# You can do so by setting a value for the `inferenceUrl` here AND by setting the `enable` to `false`

777

inferenceUrl: {}

778

# The configuration below is ignored if enabled==false

779

tag: facebook-dpr-question_encoder-single-nq-base

780

repo: semitechnologies/transformers-inference

781

registry: cr.weaviate.io

782

imagePullPolicy: IfNotPresent

783

imagePullSecrets: []

784

priorityClassName: ""

785

replicas: 1

786

strategy:

787

type: RollingUpdate

788

fullnameOverride: transformers-inference-query

789

livenessProbe:

790

initialDelaySeconds: 120

791

periodSeconds: 3

792

timeoutSeconds: 3

793

readinessProbe:

794

initialDelaySeconds: 120

795

periodSeconds: 3

796

envconfig:

797

# enable for CUDA support. Your K8s cluster needs to be configured

798

# accordingly and you need to explicitly set GPU requests & limits below

799

enable_cuda: false

800

# only used when CUDA is enabled

801

nvidia_visible_devices: all

802

nvidia_driver_capabilities: compute,utility

803

# only used when CUDA is enabled

804

ld_library_path: /usr/local/nvidia/lib64

805

# enable this setting to use all avaialble cuda cores

806

use_sentence_transformers_multi_process: false

807

# set it to true to enable simple in memory request caching

808

enable_cache: false

809

resources: {}

810

# requests:

811

# cpu: '1000m'

812

# memory: '3000Mi'

813

# enable if running with CUDA support

814

# nvidia.com/gpu: 1

815

816

# limits:

817

# cpu: '1000m'

818

# memory: '5000Mi'

819

# enable if running with CUDA support

820

# nvidia.com/gpu: 1

821

822

# security Context for the Transformers Pods. The configurations are the same as setting them

823

# as described here: https://kubernetes.io/docs/tasks/configure-pod-container/security-context/

824

# applies to passageQueryService below

825

securityContext: {}

826

# It is possible to add a ServiceAccount to this module's Pods, it can be

827

# used in cases where the module is in a private registry and you want to

828

# give access to the registry only to this pod.

829

# NOTE: if not set the root `serviceAccountName` config will be used.

830

serviceAccountName:

831

# You can guide where the pods are scheduled on a per-module basis,

832

# as well as for Weaviate overall. Each module accepts nodeSelector,

833

# tolerations, and affinity configuration. If it is set on a per-

834

# module basis, this configuration overrides the global config.

835

nodeSelector:

836

tolerations:

837

affinity:

838

# The text2vec-gpt4all is a vectorizer module that allows to run ML models from

839

# nomic-ai/gpt4all: https://docs.gpt4all.io/gpt4all_python_embedding.html

840

# These models run only on CPU only and on x86_64 arch

841

# The ML model is containerized in a Weaviate compatible way.

842

# If you want to run a different model that published ones you can follow the

843

# tutorial from here on how to create such a container: https://github.com/weaviate/t2v-gpt4all-models

844

text2vec-gpt4all:

845

# Enable deployment of this module

846

enabled: false

847

# You can set directly an inference URL of this module without deploying it with this release.

848

# You can do so by setting a value for the `inferenceUrl` here AND by setting the `enable` to `false`

849

inferenceUrl: {}

850

# The configuration below is ignored if enabled==false

851

tag: all-MiniLM-L6-v2

852

repo: semitechnologies/gpt4all-inference

853

registry: cr.weaviate.io

854

replicas: 1

855

strategy:

856

type: RollingUpdate

857

imagePullPolicy: IfNotPresent

858

imagePullSecrets: []

859

priorityClassName: ""

860

fullnameOverride: gpt4all-inference

861

livenessProbe:

862

initialDelaySeconds: 120

863

periodSeconds: 3

864

timeoutSeconds: 3

865

readinessProbe:

866

initialDelaySeconds: 120

867

periodSeconds: 3

868

resources: {}

869

# requests:

870

# cpu: '1000m'

871

# memory: '3000Mi'

872

873

# limits:

874

# cpu: '1000m'

875

# memory: '5000Mi'

876

877

# security Context for the Contextionary Pods. The configurations are the same as setting them

878

# as described here: https://kubernetes.io/docs/tasks/configure-pod-container/security-context/

879

securityContext: {}

880

# It is possible to add a ServiceAccount to this module's Pods, it can be

881

# used in cases where the module is in a private registry and you want to

882

# give access to the registry only to this pod.

883

# NOTE: if not set the root `serviceAccountName` config will be used.

884

serviceAccountName:

885

# You can guide where the pods are scheduled on a per-module basis,

886

# as well as for Weaviate overall. Each module accepts nodeSelector,

887

# tolerations, and affinity configuration. If it is set on a per-

888

# module basis, this configuration overrides the global config.

889

nodeSelector:

890

tolerations:

891

affinity:

892

# The text2vec-model2vec is a vectorizer module that allows to run static embeddings models

893

# More information about the project can be found here:

894

# https://github.com/weaviate/t2v-model2vec-models

895

text2vec-model2vec:

896

# Enable deployment of this module

897

enabled: false

898

# You can set directly an inference URL of this module without deploying it with this release.

899

# You can do so by setting a value for the `inferenceUrl` here AND by setting the `enable` to `false`

900

inferenceUrl: {}

901

# The configuration below is ignored if enabled==false

902

tag: minishlab-potion-retrieval-32M

903

repo: semitechnologies/model2vec-inference

904

registry: cr.weaviate.io

905

replicas: 1

906

strategy:

907

type: RollingUpdate

908

imagePullPolicy: IfNotPresent

909

imagePullSecrets: []

910

priorityClassName: ""

911

fullnameOverride: model2vec-inference

912

livenessProbe:

913

initialDelaySeconds: 120

914

periodSeconds: 3

915

timeoutSeconds: 3

916

readinessProbe:

917

initialDelaySeconds: 120

918

periodSeconds: 3

919

resources: {}

920

# requests:

921

# cpu: '1000m'

922

# memory: '3000Mi'

923

924

# limits:

925

# cpu: '1000m'

926

# memory: '5000Mi'

927

928

# security Context for the Model2Vec Pods. The configurations are the same as setting them

929

# as described here: https://kubernetes.io/docs/tasks/configure-pod-container/security-context/

930

securityContext: {}

931

# It is possible to add a ServiceAccount to this module's Pods, it can be

932

# used in cases where the module is in a private registry and you want to

933

# give access to the registry only to this pod.

934

# NOTE: if not set the root `serviceAccountName` config will be used.

935

serviceAccountName:

936

# You can guide where the pods are scheduled on a per-module basis,

937

# as well as for Weaviate overall. Each module accepts nodeSelector,

938

# tolerations, and affinity configuration. If it is set on a per-

939

# module basis, this configuration overrides the global config.

940

nodeSelector:

941

tolerations:

942

affinity:

943

# The text2vec-openai module uses OpenAI Embeddings API

944

# to dynamically compute vector embeddings based on the

945

# sentence's context.

946

# More information about OpenAI Embeddings API can be found here:

947

# https://beta.openai.com/docs/guides/embeddings/what-are-embeddings

948

text2vec-openai:

949

# enable if you want to use OpenAI module

950

enabled: false

951

# Set your OpenAI API Key to be passed to Weaviate pod as

952

# an environment variable. You can set either the OpenAI's Service API key

953

# OR the Azure OpenAI's Service Key. Note that you cannot set both of the at the same time.

954

apiKey: ''

955

azureApiKey: ''

956

# The text2vec-huggingface module uses HuggingFace API

957

# to dynamically compute vector embeddings based on the

958

# sentence's context.

959

# More information about HuggingFace API can be found here:

960

# https://huggingface.co/docs/api-inference/detailed_parameters#feature-extraction-task

961

text2vec-huggingface:

962

# enable if you want to use HuggingFace module

963

enabled: false

964

# Set your HuggingFace API Key to be passed to Weaviate pod as

965

# an environment variable

966

apiKey: ''

967

# The text2vec-cohere module uses Cohere API

968

# to dynamically compute vector embeddings based on the

969

# sentence's context.

970

# More information about Cohere API can be found here: https://docs.cohere.ai/

971

text2vec-cohere:

972

# enable if you want to use Cohere module

973

enabled: false

974

# Set your Cohere API Key to be passed to Weaviate pod as

975

# an environment variable

976

apiKey: ''

977

# The text2vec-google module uses Google Embeddings API

978

# to dynamically compute vector embeddings based on the

979

# sentence's context.

980

# More information about Google Embeddings API can be found here:

981

# https://developers.generativeai.google/

982

text2vec-google:

983

# enable if you want to use Google module

984

enabled: false

985

# Set your Google API Key to be passed to Weaviate pod as

986

# an environment variable

987

apiKey: ''

988

# Deprecated setting, please use text2vec-google instead

989

990

# This setting will be removed with charts version 20.

991

text2vec-palm:

992

# enable if you want to use Google PaLM module

993

enabled: false

994

# Set your Google PaLM API Key to be passed to Weaviate pod as

995

# an environment variable

996

apiKey: ''

997

# The text2vec-jinaai module uses JinaAI Embeddings API

998

# to dynamically compute vector embeddings based on the

999

# sentence's context.

1000

# More information about JinaAI Embeddings API can be found here:

1001

# https://jina.ai/embeddings

1002

text2vec-jinaai:

1003

# enable if you want to use JinaAI module

1004

enabled: false

1005

# Set your JinaAI API Key to be passed to Weaviate pod as

1006

# an environment variable

1007

apiKey: ''

1008

# Deprecated setting, please use text2multivec-jinaai instead

1009

1010

# This setting will be removed with charts version 20.

1011

text2colbert-jinaai:

1012

# enable if you want to use JinaAI module

1013

enabled: false

1014

# Set your JinaAI API Key to be passed to Weaviate pod as

1015

# an environment variable

1016

apiKey: ''

1017

# The text2multivec-jinaai module uses JinaAI Embeddings API

1018

# to dynamically compute ColBERT embeddings based on the

1019

# sentence's context.

1020

# More information about JinaAI Embeddings API can be found here:

1021

# https://jina.ai/embeddings

1022

text2multivec-jinaai:

1023

# enable if you want to use JinaAI module

1024

enabled: false

1025

# Set your JinaAI API Key to be passed to Weaviate pod as

1026

# an environment variable

1027

apiKey: ''

1028

# The multi2multivec-jinaai module uses JinaAI Embeddings API

1029

# to dynamically compute ColBERT embeddings.

1030

# More information about JinaAI Embeddings API can be found here:

1031

# https://jina.ai/embeddings

1032

multi2multivec-jinaai:

1033

# enable if you want to use JinaAI module

1034

enabled: false

1035

# Set your JinaAI API Key to be passed to Weaviate pod as

1036

# an environment variable

1037

apiKey: ''

1038

# The ref2vec-centroid module

1039

ref2vec-centroid:

1040

# enable if you want to use Centroid module

1041

enabled: false

1042

# The text2vec-aws module uses AWS Bedrock service

1043

# to dynamically compute vector embeddings based on the

1044

# sentence's context.

1045

# More information about AWS Bedrock service can be found here: https://docs.aws.amazon.com/bedrock/

1046

text2vec-aws:

1047

# enable if you want to use AWS module

1048

enabled: false

1049

# For Weaviate to be able to use AWS Bedrock service it needs a user credentials to authenticate to AWS.

1050

# The User must have permissions to use AWS Bedrock service AI.

1051

# You can pass the User credentials (access-key id and access-secret-key) in 2 ways:

1052

# 1. by setting the AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY plain values in the `secrets` section below

1053

# this chart will create a kubernetes secret for you with these key-values pairs

1054

# 2. create Kubernetes secret/s with AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY keys and their respective values

1055

# Set the Key and the secret where it is set in `envSecrets` section below

1056

secrets: {}

1057

# AWS_ACCESS_KEY_ID: access-key-id (plain text)

1058

# AWS_SECRET_ACCESS_KEY: secret-access-key (plain text)

1059

1060

# If one has already defined secrets with AWS credentials one can pass them using

1061

# this setting:

1062

envSecrets: {}

1063

# AWS_ACCESS_KEY_ID: name-of-the-k8s-secret-containing-the-key-id

1064

# AWS_SECRET_ACCESS_KEY: name-of-the-k8s-secret-containing-the-key

1065

# The text2vec-voyageai module uses VoaygeAI Embeddings API

1066

# to dynamically compute vector embeddings based on the

1067

# sentence's context.

1068

# More information about VoaygeAI Embeddings API can be found here:

1069

# https://www.voyageai.com/

1070

text2vec-voyageai:

1071

# enable if you want to use VoaygeAI module

1072

enabled: false

1073

# Set your VoaygeAI API Key to be passed to Weaviate pod as

1074

# an environment variable

1075

apiKey: ''

1076

# The text2vec-ollama module uses Ollama Embeddings API

1077

# to dynamically compute vector embeddings based on the

1078

# sentence's context.

1079

# More information about Ollama Embeddings API can be found here:

1080

# https://github.com/ollama/ollama/blob/main/docs/api.md#generate-embeddings

1081

# Please note that you have to deploy Ollama container by yourself

1082

# and then properly point the Ollama endpoint using module settings.

1083

# More about this configuration can be found here:

1084

# https://weaviate.io/developers/weaviate/modules/retriever-vectorizer-modules/text2vec-ollama#ollama-endpoint

1085

text2vec-ollama:

1086

# enable if you want to use Ollama module

1087

enabled: false

1088

# The text2vec-octoai module uses OctoAI Embeddings API

1089

# to dynamically compute vector embeddings based on the

1090

# sentence's context.

1091

# More information about OctoAI Embeddings API can be found here:

1092

# https://octo.ai/docs/text-gen-solution/getting-started

1093

text2vec-octoai:

1094

# enable if you want to use OctoAI module

1095

enabled: false

1096

# Set your OctoAI API Key to be passed to Weaviate pod as

1097

# an environment variable

1098

apiKey: ''

1099

# The text2vec-mistral module uses MISTRAL AI Embeddings API

1100

# More information about MISTRAL AI's Embeddings API can be found here:

1101

# https://docs.mistral.ai/api/#operation/createEmbedding

1102

text2vec-mistral:

1103

# enable if you want to use MISTRAL AI module

1104

enabled: false

1105

# Set your MISTRAL API Key to be passed to Weaviate pod as

1106

# an environment variable

1107

apiKey: ''

1108

# The text2vec-databricks module uses Databricks Embeddings API

1109

# to dynamically compute vector embeddings based on the

1110

# sentence's context.

1111

# More information about Databricks Embeddings API can be found here:

1112

# https://docs.databricks.com/en/machine-learning/foundation-models/api-reference.html#embedding-task

1113

text2vec-databricks:

1114

# enable if you want to use Databricks module

1115

enabled: false

1116

# Set your Databricks Token to be passed to Weaviate pod as

1117

# an environment variable

1118

token: ''

1119

# The text2vec-nvidia module uses NVIDIA API

1120

# to dynamically compute vector embeddings based on the

1121

# sentence's context.

1122

# More information about NVIDIA's API can be found here:

1123

# https://docs.api.nvidia.com/nim/reference/retrieval-apis

1124

text2vec-nvidia:

1125

# enable if you want to use NVIDIA text2vec module

1126

enabled: false

1127

# Set your NVIDIA API Key to be passed to Weaviate pod as

1128

# an environment variable

1129

apiKey: ''

1130

# The text2vec-morph module uses Morph Embeddings API

1131

# to dynamically compute vector embeddings.

1132

# More information about morph Embeddings API can be found here:

1133

# https://docs.morphllm.com/quickstart

1134

text2vec-morph:

1135

# enable if you want to use morph module

1136

enabled: false

1137

# Set your Morph API Key to be passed to Weaviate pod as

1138

# an environment variable

1139

apiKey: ''

1140

# The text2vec-digitalocean module uses DigitalOcean Embeddings API

1141

# to dynamically compute vector embeddings based on the

1142

# sentence's context.

1143

# More information about DigitalOcean Embeddings API can be found here:

1144

# https://docs.digitalocean.com/products/gradient-ai-platform/

1145

text2vec-digitalocean:

1146

# enable if you want to use DigitalOcean module

1147

enabled: false

1148

# Set your DigitalOcean API Key to be passed to Weaviate pod as

1149

# an environment variable

1150

apiKey: ''

1151

# The multi2vec-clip modules uses CLIP transformers to vectorize both images

1152

# and text in the same vector space. It is typically slow(er) on CPUs and should

1153

# run with CUDA-enabled GPUs for optimal performance.

1154

multi2vec-clip:

1155

# Enable deployment of this module

1156

enabled: false

1157

# You can set directly an inference URL of this module without deploying it with this release.

1158

# You can do so by setting a value for the `inferenceUrl` here AND by setting the `enable` to `false`

1159

inferenceUrl: {}

1160

# The configuration below is ignored if enabled==false

1161

1162

# replace with model of choice, see

1163

# https://weaviate.io/developers/weaviate/modules/retriever-vectorizer-modules/multi2vec-clip

1164

# for all supported models or build your own container.

1165

tag: sentence-transformers-clip-ViT-B-32-multilingual-v1

1166

repo: semitechnologies/multi2vec-clip

1167

registry: cr.weaviate.io

1168

replicas: 1

1169

strategy:

1170

type: RollingUpdate

1171

imagePullPolicy: IfNotPresent

1172

imagePullSecrets: []

1173

priorityClassName: ""

1174

fullnameOverride: clip-inference

1175

livenessProbe:

1176

initialDelaySeconds: 120

1177

periodSeconds: 3

1178

timeoutSeconds: 3

1179

readinessProbe:

1180

initialDelaySeconds: 120

1181

periodSeconds: 3

1182

envconfig:

1183

# enable for CUDA support. Your K8s cluster needs to be configured

1184

# accordingly and you need to explicitly set GPU requests & limits below

1185

enable_cuda: false

1186

# only used when CUDA is enabled

1187

nvidia_visible_devices: all

1188

nvidia_driver_capabilities: compute,utility

1189

# only used when CUDA is enabled

1190

ld_library_path: /usr/local/nvidia/lib64

1191

resources: {}

1192

# requests:

1193

# cpu: '1000m'

1194

# memory: '3000Mi'

1195

# enable if running with CUDA support

1196

# nvidia.com/gpu: 1

1197

# limits:

1198

# cpu: '1000m'

1199

# memory: '5000Mi'

1200

# enable if running with CUDA support

1201

# nvidia.com/gpu: 1

1202

1203

# security Context for the Contextionary Pods. The configurations are the same as setting them

1204

# as described here: https://kubernetes.io/docs/tasks/configure-pod-container/security-context/

1205

securityContext: {}

1206

# It is possible to add a ServiceAccount to this module's Pods, it can be

1207

# used in cases where the module is in a private registry and you want to

1208

# give access to the registry only to this pod.

1209

# NOTE: if not set the root `serviceAccountName` config will be used.

1210

serviceAccountName:

1211

annotations:

1212

nodeSelector:

1213

tolerations:

1214

# The multi2vec-bind is a multi modal inference container it uses Meta's open

1215

# source ImageBind implementation as base for this module:

1216

# https://github.com/facebookresearch/ImageBind

1217

# It is possible to create your own version of this module by following the instructions here:

1218

# https://github.com/weaviate/multi2vec-bind-inference

1219

multi2vec-bind:

1220

# Enable deployment of this module

1221

enabled: false

1222

# You can set directly an inference URL of this module without deploying it with this release.

1223

# You can do so by setting a value for the `inferenceUrl` here AND by setting the `enable` to `false`

1224

inferenceUrl: {}

1225

# The configuration below is ignored if enabled==false

1226

tag: imagebind

1227

repo: semitechnologies/multi2vec-bind

1228

registry: cr.weaviate.io

1229

replicas: 1

1230

strategy:

1231

type: RollingUpdate

1232

imagePullPolicy: IfNotPresent

1233

imagePullSecrets: []

1234

priorityClassName: ""

1235

fullnameOverride: bind-inference

1236

livenessProbe:

1237

initialDelaySeconds: 120

1238

periodSeconds: 3

1239

timeoutSeconds: 3

1240

readinessProbe:

1241

initialDelaySeconds: 120

1242

periodSeconds: 3

1243

envconfig:

1244

# enable for CUDA support. Your K8s cluster needs to be configured

1245

# accordingly and you need to explicitly set GPU requests & limits below

1246

enable_cuda: false

1247

# only used when CUDA is enabled

1248

nvidia_visible_devices: all

1249

nvidia_driver_capabilities: compute,utility

1250

# only used when CUDA is enabled

1251

ld_library_path: /usr/local/nvidia/lib64

1252

resources: {}

1253

# requests:

1254

# cpu: '1000m'

1255

# memory: '7000Mi'

1256

# enable if running with CUDA support

1257

# nvidia.com/gpu: 1

1258

# limits:

1259

# cpu: '1000m'

1260

# memory: '7000Mi'

1261

# enable if running with CUDA support

1262

# nvidia.com/gpu: 1

1263

1264

# security Context for the Contextionary Pods. The configurations are the same as setting them

1265

# as described here: https://kubernetes.io/docs/tasks/configure-pod-container/security-context/

1266

securityContext: {}

1267

# It is possible to add a ServiceAccount to this module's Pods, it can be

1268

# used in cases where the module is in a private registry and you want to

1269

# give access to the registry only to this pod.

1270

# NOTE: if not set the root `serviceAccountName` config will be used.

1271

serviceAccountName:

1272

annotations:

1273

nodeSelector:

1274

tolerations:

1275

# The multi2vec-google module uses Google Embeddings API

1276

# to dynamically compute vector embeddings.

1277

# More information about Google Embeddings API can be found here:

1278

# https://cloud.google.com/vertex-ai/generative-ai/docs/embeddings/get-multimodal-embeddings

1279

multi2vec-google:

1280

# enable if you want to use Google Multimodal module

1281

enabled: false

1282

# Set your Google API Key to be passed to Weaviate pod as

1283

# an environment variable

1284

apiKey: ''

1285

# Deprecated setting, please use multi2vec-google instead

1286

1287

# This setting will be removed with charts version 20.

1288

multi2vec-palm:

1289

# enable if you want to use Google Multimodal PaLM module

1290

enabled: false

1291

# Set your Google PaLM API Key to be passed to Weaviate pod as

1292

# an environment variable

1293

apiKey: ''

1294

# The multi2vec-cohere module uses Cohere API

1295

# to dynamically compute vector embeddings.

1296

# More information about Cohere API can be found here: https://docs.cohere.ai/

1297

multi2vec-cohere:

1298

# enable if you want to use Cohere module

1299

enabled: false

1300

# Set your Cohere API Key to be passed to Weaviate pod as

1301

# an environment variable

1302

apiKey: ''

1303

# The multi2vec-jinaai module uses JinaAI Embeddings API

1304

# to dynamically compute vector embeddings.

1305

# More information about JinaAI Embeddings API can be found here:

1306

# https://jina.ai/embeddings

1307

multi2vec-jinaai:

1308

# enable if you want to use JinaAI module

1309

enabled: false

1310

# Set your JinaAI API Key to be passed to Weaviate pod as

1311

# an environment variable

1312

apiKey: ''

1313

# The multi2vec-voyageai module uses VoaygeAI Multi Modal Embeddings API

1314

# to dynamically compute vector embeddings.

1315

# More information about VoaygeAI Embeddings API can be found here:

1316

# https://www.voyageai.com/

1317

multi2vec-voyageai:

1318

# enable if you want to use VoaygeAI module

1319

enabled: false

1320

# Set your VoaygeAI API Key to be passed to Weaviate pod as

1321

# an environment variable

1322

apiKey: ''

1323

# The multi2vec-nvidia module uses NVIDIA API

1324

# to dynamically compute vector embeddings.

1325

# More information about NVIDIA's API can be found here:

1326

# https://docs.api.nvidia.com/nim/reference/retrieval-apis

1327

multi2vec-nvidia:

1328

# enable if you want to use NVIDIA multi2vec module

1329

enabled: false

1330

# Set your NVIDIA API Key to be passed to Weaviate pod as

1331

# an environment variable

1332

apiKey: ''

1333

# The multi2vec-aws module uses AWS Bedrock service

1334

# to dynamically compute vector embeddings.

1335

# More information about AWS Bedrock service can be found here: https://docs.aws.amazon.com/bedrock/

1336

multi2vec-aws:

1337

# enable if you want to use AWS module

1338

enabled: false

1339

# For Weaviate to be able to use AWS Bedrock service it needs a user credentials to authenticate to AWS.

1340

# The User must have permissions to use AWS Bedrock service AI.

1341

# You can pass the User credentials (access-key id and access-secret-key) in 2 ways:

1342

# 1. by setting the AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY plain values in the `secrets` section below

1343

# this chart will create a kubernetes secret for you with these key-values pairs

1344

# 2. create Kubernetes secret/s with AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY keys and their respective values

1345

# Set the Key and the secret where it is set in `envSecrets` section below

1346

secrets: {}

1347

# AWS_ACCESS_KEY_ID: access-key-id (plain text)

1348

# AWS_SECRET_ACCESS_KEY: secret-access-key (plain text)

1349

1350

# If one has already defined secrets with AWS credentials one can pass them using

1351

# this setting:

1352

envSecrets: {}

1353

# AWS_ACCESS_KEY_ID: name-of-the-k8s-secret-containing-the-key-id

1354

# AWS_SECRET_ACCESS_KEY: name-of-the-k8s-secret-containing-the-key

1355

# The qna-transformers module uses neural networks, such as BERT,

1356

# DistilBERT, to find an answer in text to a given question

1357

qna-transformers:

1358

# Enable deployment of this module

1359

enabled: false

1360

# You can set directly an inference URL of this module without deploying it with this release.

1361

# You can do so by setting a value for the `inferenceUrl` here AND by setting the `enable` to `false`

1362

inferenceUrl: {}

1363

# The configuration below is ignored if enabled==false

1364

tag: bert-large-uncased-whole-word-masking-finetuned-squad

1365

repo: semitechnologies/qna-transformers

1366

registry: cr.weaviate.io

1367

replicas: 1

1368

strategy:

1369

type: RollingUpdate

1370

imagePullPolicy: IfNotPresent

1371

imagePullSecrets: []

1372

priorityClassName: ""

1373

fullnameOverride: qna-transformers

1374

livenessProbe:

1375

initialDelaySeconds: 120

1376

periodSeconds: 3

1377

timeoutSeconds: 3

1378

readinessProbe:

1379

initialDelaySeconds: 120

1380

periodSeconds: 3

1381

envconfig:

1382

# enable for CUDA support. Your K8s cluster needs to be configured

1383

# accordingly and you need to explicitly set GPU requests & limits below

1384

enable_cuda: false

1385

# only used when CUDA is enabled

1386

nvidia_visible_devices: all

1387

nvidia_driver_capabilities: compute,utility

1388

# only used when CUDA is enabled

1389

ld_library_path: /usr/local/nvidia/lib64

1390

resources: {}

1391

# requests:

1392

# cpu: '1000m'

1393

# memory: '3000Mi'

1394

# enable if running with CUDA support

1395

# nvidia.com/gpu: 1

1396

# limits:

1397

# cpu: '1000m'

1398

# memory: '5000Mi'

1399

# enable if running with CUDA support

1400

# nvidia.com/gpu: 1

1401

1402

# security Context for the Contextionary Pods. The configurations are the same as setting them

1403

# as described here: https://kubernetes.io/docs/tasks/configure-pod-container/security-context/

1404

securityContext: {}

1405

# It is possible to add a ServiceAccount to this module's Pods, it can be

1406

# used in cases where the module is in a private registry and you want to

1407

# give access to the registry only to this pod.

1408

# NOTE: if not set the root `serviceAccountName` config will be used.

1409

serviceAccountName:

1410

# You can guide where the pods are scheduled on a per-module basis,

1411

# as well as for Weaviate overall. Each module accepts nodeSelector,

1412

# tolerations, and affinity configuration. If it is set on a per-

1413

# module basis, this configuration overrides the global config.

1414

nodeSelector:

1415

tolerations:

1416

affinity:

1417

# The qna-openai module uses OpenAI Completions API

1418

# to dynamically answer given questions.

1419

# More information about OpenAI Completions API can be found here:

1420

# https://beta.openai.com/docs/api-reference/completions

1421

qna-openai:

1422

# enable if you want to use OpenAI module

1423

enabled: false

1424

# Set your OpenAI API Key to be passed to Weaviate pod as

1425

# an environment variable. You can set either the OpenAI's Service API key

1426

# OR the Azure OpenAI's Service Key. Note that you cannot set both of the at the same time.

1427

apiKey: ''

1428

azureApiKey: ''

1429

# The generative-openai module uses OpenAI Completions API

1430

# along with text-davinci-003 model to behave as ChatGPT.

1431

# More information about OpenAI Completions API can be found here:

1432

# https://beta.openai.com/docs/api-reference/completions

1433

generative-openai:

1434

# enable if you want to use OpenAI module

1435

enabled: false

1436

# Set your OpenAI API Key to be passed to Weaviate pod as

1437

# an environment variable. You can set either the OpenAI's Service API key

1438

# OR the Azure OpenAI's Service Key. Note that you cannot set both of the at the same time.

1439

apiKey: ''

1440

azureApiKey: ''

1441

# The generative-cohere module uses Cohere Generate API

1442

# More information about Cohere's Generate API can be found here:

1443

# https://docs.cohere.com/reference/generate

1444

generative-cohere:

1445

# enable if you want to use Cohere generative module

1446

enabled: false

1447

# Set your Cohere API Key to be passed to Weaviate pod as

1448

# an environment variable

1449

apiKey: ''

1450

# The generative-google module uses Google API.

1451

# More information about Google API can be found here:

1452

# https://developers.generativeai.google/

1453

generative-google:

1454

# enable if you want to use Google module

1455

enabled: false

1456

# Set your Google API Key to be passed to Weaviate pod as

1457

# an environment variable

1458

apiKey: ''

1459

# Deprecated setting, please use generative-google instead

1460

1461

# This setting will be removed with charts version 20.

1462

generative-palm:

1463

# enable if you want to use Google PaLM module

1464

enabled: false

1465

# Set your Google PaLM API Key to be passed to Weaviate pod as

1466

# an environment variable

1467

apiKey: ''

1468

# The generative-aws module uses AWS Bedrock service.

1469

# More information about AWS Bedrock service can be found here:

1470

# https://docs.aws.amazon.com/bedrock/

1471

generative-aws:

1472

# enable if you want to use AWS module

1473

enabled: false

1474

# For Weaviate to be able to use AWS Bedrock service it needs a user credentials to authenticate to AWS.

1475

# The User must have permissions to use AWS Bedrock service AI.

1476

# You can pass the User credentials (access-key id and access-secret-key) in 2 ways:

1477

# 1. by setting the AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY plain values in the `secrets` section below

1478

# this chart will create a kubernetes secret for you with these key-values pairs

1479

# 2. create Kubernetes secret/s with AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY keys and their respective values

1480

# Set the Key and the secret where it is set in `envSecrets` section below

1481

secrets: {}

1482

# AWS_ACCESS_KEY_ID: access-key-id (plain text)

1483

# AWS_SECRET_ACCESS_KEY: secret-access-key (plain text)

1484

1485

# If one has already defined secrets with AWS credentials one can pass them using

1486

# this setting:

1487

envSecrets: {}

1488

# AWS_ACCESS_KEY_ID: name-of-the-k8s-secret-containing-the-key-id

1489

# AWS_SECRET_ACCESS_KEY: name-of-the-k8s-secret-containing-the-key

1490

# The generative-anyscale module uses Anyscale API

1491

# More information about Anyscale API can be found here:

1492

# https://docs.anyscale.com/endpoints/overview

1493

generative-anyscale:

1494

# enable if you want to use Anyscale module

1495

enabled: false

1496

# Set your Anyscale API Key to be passed to Weaviate pod as

1497

# an environment variable.

1498

apiKey: ''

1499

# The generative-mistral module uses MISTRAL AI Generative API

1500

# More information about MISTRAL AI's Generative API can be found here:

1501

# https://docs.mistral.ai/

1502

generative-mistral:

1503

# enable if you want to use MISTRAL AI generative module

1504

enabled: false

1505

# Set your MISTRAL API Key to be passed to Weaviate pod as

1506

# an environment variable

1507

apiKey: ''

1508

# The generative-ollama module uses Ollama Generate API

1509

# More information about Ollama's Generate API can be found here:

1510

# https://github.com/ollama/ollama/blob/main/docs/api.md#generate-a-completion

1511

# Please note that you have to deploy Ollama container by yourself

1512

# and then properly point the Ollama endpoint using module settings.

1513

# More about this configuration can be found here:

1514

# https://weaviate.io/developers/weaviate/modules/reader-generator-modules/generative-ollama#ollama-endpoint

1515

generative-ollama:

1516

# enable if you want to use Ollama module

1517

enabled: false

1518

# The generative-octoai module uses OctoAI API.

1519

# More information about OctoAI API can be found here:

1520

# https://octo.ai/docs/text-gen-solution/getting-started

1521

generative-octoai:

1522

# enable if you want to use OctoAI module

1523

enabled: false

1524

# Set your OctoAI API Key to be passed to Weaviate pod as

1525

# an environment variable

1526

apiKey: ''

1527

# The generative-anthropic module uses Anthropic API.

1528

# More information about Anthropic API can be found here:

1529

# https://docs.anthropic.com/en/api/getting-started

1530

generative-anthropic:

1531

# enable if you want to use Anthropic module

1532

enabled: false

1533

# Set your Anthropic API Key to be passed to Weaviate pod as

1534

# an environment variable

1535

apiKey: ''

1536

# The generative-friendliai module uses FriendliAI API.

1537

# More information about FriendliAI API can be found here:

1538

# https://docs.friendli.ai/openapi/create-chat-completions

1539

generative-friendliai:

1540

# enable if you want to use FriendliAI module

1541

enabled: false

1542

# Set your FriendliAI Token to be passed to Weaviate pod as

1543

# an environment variable

1544

token: ''

1545

# The generative-databricks module uses Databricks API.

1546

# More information about Databricks API can be found here:

1547

# https://docs.databricks.com/en/machine-learning/foundation-models/api-reference.html#completion-task

1548

generative-databricks:

1549

# enable if you want to use Databricks module

1550

enabled: false

1551

# Set your Databricks Token to be passed to Weaviate pod as

1552

# an environment variable

1553

token: ''

1554

# The generative-nvidia module uses NVIDIA API

1555

# More information about NVIDIA's API can be found here:

1556

# https://docs.api.nvidia.com/nim/reference/llm-apis

1557

generative-nvidia:

1558

# enable if you want to use NVIDIA generative module

1559

enabled: false

1560

# Set your NVIDIA API Key to be passed to Weaviate pod as

1561

# an environment variable

1562

apiKey: ''

1563

# The generative-xai module uses xAI API

1564

# More information about xAI's API can be found here:

1565

# https://docs.x.ai/docs/overview

1566

generative-xai:

1567

# enable if you want to use xAI generative module

1568

enabled: false

1569

# Set your xAI API Key to be passed to Weaviate pod as

1570

# an environment variable

1571

apiKey: ''

1572

# The generative-contextualai module uses Contextual AI API

1573

# More information about Contextual AI's API can be found here:

1574

# https://docs.contextual.ai/api-reference/generate/generate

1575

generative-contextualai:

1576

# enable if you want to use contextualai generative module

1577

enabled: false

1578

# Set your Contextual AI API Key to be passed to Weaviate pod as

1579

# an environment variable

1580

apiKey: ''

1581

# The img2vec-neural module uses neural networks, to generate

1582

# a vector representation of the image

1583

img2vec-neural:

1584

# Enable deployment of this module

1585

enabled: false

1586

# You can set directly an inference URL of this module without deploying it with this release.

1587

# You can do so by setting a value for the `inferenceUrl` here AND by setting the `enable` to `false`

1588

inferenceUrl: {}

1589

tag: resnet50

1590

repo: semitechnologies/img2vec-pytorch

1591

registry: cr.weaviate.io

1592

replicas: 1

1593

strategy:

1594

type: RollingUpdate

1595

imagePullPolicy: IfNotPresent

1596

imagePullSecrets: []

1597

priorityClassName: ""

1598

fullnameOverride: img2vec-neural

1599

livenessProbe:

1600

initialDelaySeconds: 120

1601

periodSeconds: 3

1602

timeoutSeconds: 3

1603

readinessProbe:

1604

initialDelaySeconds: 120

1605

periodSeconds: 3

1606

envconfig:

1607

# enable for CUDA support. Your K8s cluster needs to be configured

1608

# accordingly and you need to explicitly set GPU requests & limits below

1609

enable_cuda: false

1610

# only used when CUDA is enabled

1611

nvidia_visible_devices: all

1612

nvidia_driver_capabilities: compute,utility

1613

# only used when CUDA is enabled

1614

ld_library_path: /usr/local/nvidia/lib64

1615

resources: {}

1616

# requests:

1617

# cpu: '1000m'

1618

# memory: '3000Mi'

1619

# enable if running with CUDA support

1620

# nvidia.com/gpu: 1

1621

# limits:

1622

# cpu: '1000m'

1623

# memory: '5000Mi'

1624

# enable if running with CUDA support

1625

# nvidia.com/gpu: 1

1626

1627

# security Context for the Contextionary Pods. The configurations are the same as setting them

1628

# as described here: https://kubernetes.io/docs/tasks/configure-pod-container/security-context/

1629

securityContext: {}

1630

# It is possible to add a ServiceAccount to this module's Pods, it can be

1631

# used in cases where the module is in a private registry and you want to

1632

# give access to the registry only to this pod.

1633

# NOTE: if not set the root `serviceAccountName` config will be used.

1634

serviceAccountName:

1635

# You can guide where the pods are scheduled on a per-module basis,

1636

# as well as for Weaviate overall. Each module accepts nodeSelector,

1637

# tolerations, and affinity configuration. If it is set on a per-

1638

# module basis, this configuration overrides the global config.

1639

nodeSelector:

1640

tolerations:

1641

affinity:

1642

# The reranker-cohere module uses Cohere API

1643

# to dynamically compute a score for the relevance

1644

# of the query with each of the initial search results.

1645

# More information about Cohere API can be found here: https://docs.cohere.com/docs/rerank-guide

1646

reranker-cohere:

1647

# enable if you want to use Cohere rerank module

1648

enabled: false

1649

# Set your Cohere API Key to be passed to Weaviate pod as

1650

# an environment variable

1651

apiKey: ''

1652

# The reranker-voyageai module uses VoaygeAI API

1653

# to dynamically compute a score for the relevance

1654

# of the query with each of the initial search results.

1655

# More information about Cohere API can be found here: https://www.voyageai.com/

1656

reranker-voyageai:

1657

# enable if you want to use VoaygeAI module

1658

enabled: false

1659

# Set your VoaygeAI API Key to be passed to Weaviate pod as

1660

# an environment variable

1661

apiKey: ''

1662

# The reranker-jinaai module uses JinaAI API

1663

# to dynamically compute a score for the relevance

1664

# of the query with each of the initial search results.

1665

# More information about JinaAI API can be found here: https://jina.ai/reranker

1666

reranker-jinaai:

1667

# enable if you want to use JinaAI module

1668

enabled: false

1669

# Set your JinaAI API Key to be passed to Weaviate pod as

1670

# an environment variable

1671

apiKey: ''

1672

# The reranker-nvidia module uses NVIDIA API

1673

# to dynamically compute a score for the relevance

1674

# of the query with each of the initial search results.

1675

# More information about NVIDIA's API can be found here:

1676

# https://docs.api.nvidia.com/nim/reference/retrieval-apis

1677

reranker-nvidia:

1678

# enable if you want to use NVIDIA reranker module

1679

enabled: false

1680

# Set your NVIDIA API Key to be passed to Weaviate pod as

1681

# an environment variable

1682

apiKey: ''

1683

# The reranker-contextualai module uses Contextual AI API

1684

# to dynamically compute a score for the relevance

1685

# of the query with each of the initial search results.

1686

# More information about Contextual AI's API can be found here:

1687

# https://docs.api.nvidia.com/nim/reference/retrieval-apis

1688

reranker-contextualai:

1689

# enable if you want to use Contextual AI reranker module

1690

enabled: false

1691

# Set your Contextual AI API Key to be passed to Weaviate pod as

1692

# an environment variable

1693

apiKey: ''

1694

# The reranker-transformers module uses Cross-Encoders for

1695

# sentence pair scoring and sentence pair classification tasks.

1696

# More information about Cross-Encoders can be found here:

1697

# https://www.sbert.net/examples/applications/cross-encoder/README.html

1698

reranker-transformers:

1699

# Enable deployment of this module

1700

enabled: false

1701

# You can set directly an inference URL of this module without deploying it with this release.

1702

# You can do so by setting a value for the `inferenceUrl` here AND by setting the `enable` to `false`

1703

inferenceUrl: {}

1704

# The configuration below is ignored if enabled==false

1705

tag: cross-encoder-ms-marco-MiniLM-L-6-v2

1706

repo: semitechnologies/reranker-transformers

1707

registry: cr.weaviate.io

1708

replicas: 1

1709

strategy:

1710

type: RollingUpdate

1711

imagePullPolicy: IfNotPresent

1712

imagePullSecrets: []

1713

priorityClassName: ""

1714

fullnameOverride: reranker-transformers

1715

livenessProbe:

1716

initialDelaySeconds: 120

1717

periodSeconds: 3

1718

timeoutSeconds: 3

1719

readinessProbe:

1720

initialDelaySeconds: 120

1721

periodSeconds: 3

1722

envconfig:

1723

# enable for CUDA support. Your K8s cluster needs to be configured

1724

# accordingly and you need to explicitly set GPU requests & limits below

1725

enable_cuda: false

1726

# only used when CUDA is enabled

1727

nvidia_visible_devices: all

1728

nvidia_driver_capabilities: compute,utility

1729

# only used when CUDA is enabled

1730

ld_library_path: /usr/local/nvidia/lib64

1731

resources: {}

1732

# requests:

1733

# cpu: '1000m'

1734

# memory: '3000Mi'

1735

# enable if running with CUDA support

1736

# nvidia.com/gpu: 1

1737

# limits:

1738

# cpu: '1000m'

1739

# memory: '5000Mi'

1740

# enable if running with CUDA support

1741

# nvidia.com/gpu: 1

1742

1743

# security Context for the Contextionary Pods. The configurations are the same as setting them

1744

# as described here: https://kubernetes.io/docs/tasks/configure-pod-container/security-context/

1745

securityContext: {}

1746

# It is possible to add a ServiceAccount to this module's Pods, it can be

1747

# used in cases where the module is in a private registry and you want to

1748

# give access to the registry only to this pod.

1749

# NOTE: if not set the root `serviceAccountName` config will be used.

1750

serviceAccountName:

1751

# You can guide where the pods are scheduled on a per-module basis,

1752

# as well as for Weaviate overall. Each module accepts nodeSelector,

1753

# tolerations, and affinity configuration. If it is set on a per-

1754

# module basis, this configuration overrides the global config.

1755

nodeSelector:

1756

tolerations:

1757

affinity:

1758

# The text-spellcheck module uses spellchecker library to check

1759

# misspellings in a given text

1760

text-spellcheck:

1761

# Enable deployment of this module

1762

enabled: false

1763

# You can set directly an inference URL of this module without deploying it with this release.

1764

# You can do so by setting a value for the `inferenceUrl` here AND by setting the `enable` to `false`

1765

inferenceUrl: {}

1766

# The configuration below is ignored if enabled==false

1767

tag: pyspellchecker-en

1768

repo: semitechnologies/text-spellcheck-model

1769

registry: cr.weaviate.io

1770

replicas: 1

1771

strategy:

1772

type: RollingUpdate

1773

imagePullPolicy: IfNotPresent

1774

imagePullSecrets: []

1775

priorityClassName: ""

1776

fullnameOverride: text-spellcheck

1777

livenessProbe:

1778

initialDelaySeconds: 120

1779

periodSeconds: 3

1780

timeoutSeconds: 3

1781

readinessProbe:

1782

initialDelaySeconds: 120

1783

periodSeconds: 3

1784

resources: {}

1785

# requests:

1786

# cpu: '400m'

1787

# memory: '400Mi'

1788

# limits:

1789

# cpu: '500m'

1790

# memory: '500Mi'

1791

1792

# security Context for the Contextionary Pods. The configurations are the same as setting them

1793

# as described here: https://kubernetes.io/docs/tasks/configure-pod-container/security-context/

1794

securityContext: {}

1795

# It is possible to add a ServiceAccount to this module's Pods, it can be

1796

# used in cases where the module is in a private registry and you want to

1797

# give access to the registry only to this pod.

1798

# NOTE: if not set the root `serviceAccountName` config will be used.

1799

serviceAccountName:

1800

# You can guide where the pods are scheduled on a per-module basis,

1801

# as well as for Weaviate overall. Each module accepts nodeSelector,

1802

# tolerations, and affinity configuration. If it is set on a per-

1803

# module basis, this configuration overrides the global config.

1804

nodeSelector:

1805

tolerations:

1806

affinity:

1807

# The ner-transformers module uses spellchecker library to check

1808

# misspellings in a given text

1809

ner-transformers:

1810

# Enable deployment of this module

1811

enabled: false

1812

# You can set directly an inference URL of this module without deploying it with this release.

1813

# You can do so by setting a value for the `inferenceUrl` here AND by setting the `enable` to `false`

1814

inferenceUrl: {}

1815

# The configuration below is ignored if enabled==false

1816

tag: dbmdz-bert-large-cased-finetuned-conll03-english

1817

repo: semitechnologies/ner-transformers

1818

registry: cr.weaviate.io

1819

replicas: 1

1820

strategy:

1821

type: RollingUpdate

1822

imagePullPolicy: IfNotPresent

1823

imagePullSecrets: []

1824

priorityClassName: ""

1825

fullnameOverride: ner-transformers

1826

livenessProbe:

1827

initialDelaySeconds: 120

1828

periodSeconds: 3

1829

timeoutSeconds: 3

1830

readinessProbe:

1831

initialDelaySeconds: 120

1832

periodSeconds: 3

1833

envconfig:

1834

# enable for CUDA support. Your K8s cluster needs to be configured

1835

# accordingly and you need to explicitly set GPU requests & limits below

1836

enable_cuda: false

1837

# only used when CUDA is enabled

1838

nvidia_visible_devices: all

1839

nvidia_driver_capabilities: compute,utility

1840

# only used when CUDA is enabled

1841

ld_library_path: /usr/local/nvidia/lib64

1842

resources: {}

1843

# requests:

1844

# cpu: '1000m'

1845

# memory: '3000Mi'

1846

1847

# enable if running with CUDA support

1848

# nvidia.com/gpu: 1

1849

# limits:

1850

# cpu: '1000m'

1851

# memory: '5000Mi'

1852

1853

# enable if running with CUDA support

1854

# nvidia.com/gpu: 1

1855

1856

# security Context for the Contextionary Pods. The configurations are the same as setting them

1857

# as described here: https://kubernetes.io/docs/tasks/configure-pod-container/security-context/

1858

securityContext: {}

1859

# It is possible to add a ServiceAccount to this module's Pods, it can be

1860

# used in cases where the module is in a private registry and you want to

1861

# give access to the registry only to this pod.

1862

# NOTE: if not set the root `serviceAccountName` config will be used.

1863

serviceAccountName:

1864

# You can guide where the pods are scheduled on a per-module basis,

1865

# as well as for Weaviate overall. Each module accepts nodeSelector,

1866

# tolerations, and affinity configuration. If it is set on a per-

1867

# module basis, this configuration overrides the global config.

1868

nodeSelector:

1869

tolerations:

1870

affinity:

1871

# The sum-transformers module makes result texts summarizations

1872

sum-transformers:

1873

# Enable deployment of this module

1874

enabled: false

1875

# You can set directly an inference URL of this module without deploying it with this release.

1876

# You can do so by setting a value for the `inferenceUrl` here AND by setting the `enable` to `false`

1877

inferenceUrl: {}

1878

# The configuration below is ignored if enabled==false

1879

tag: facebook-bart-large-cnn

1880

repo: semitechnologies/sum-transformers

1881

registry: cr.weaviate.io

1882

replicas: 1

1883

strategy:

1884

type: RollingUpdate

1885

imagePullPolicy: IfNotPresent

1886

imagePullSecrets: []

1887

priorityClassName: ""

1888

fullnameOverride: sum-transformers

1889

livenessProbe:

1890

initialDelaySeconds: 120

1891

periodSeconds: 3

1892

timeoutSeconds: 3

1893

readinessProbe:

1894

initialDelaySeconds: 120

1895

periodSeconds: 3

1896

envconfig:

1897

# enable for CUDA support. Your K8s cluster needs to be configured

1898

# accordingly and you need to explicitly set GPU requests & limits below

1899

enable_cuda: false

1900

# only used when CUDA is enabled

1901

nvidia_visible_devices: all

1902

nvidia_driver_capabilities: compute,utility

1903

# only used when CUDA is enabled

1904

ld_library_path: /usr/local/nvidia/lib64

1905

resources: {}

1906

# requests:

1907

# cpu: '1000m'

1908

# memory: '3000Mi'

1909

# enable if running with CUDA support

1910

# nvidia.com/gpu: 1

1911

# limits:

1912

# cpu: '1000m'

1913

# memory: '5000Mi'

1914

# enable if running with CUDA support

1915

# nvidia.com/gpu: 1

1916

1917

# security Context for the Contextionary Pods. The configurations are the same as setting them

1918

# as described here: https://kubernetes.io/docs/tasks/configure-pod-container/security-context/

1919

securityContext: {}

1920

# It is possible to add a ServiceAccount to this module's Pods, it can be

1921

# used in cases where the module is in a private registry and you want to

1922

# give access to the registry only to this pod.

1923

# NOTE: if not set the root `serviceAccountName` config will be used.

1924

serviceAccountName:

1925

# You can guide where the pods are scheduled on a per-module basis,

1926

# as well as for Weaviate overall. Each module accepts nodeSelector,

1927

# tolerations, and affinity configuration. If it is set on a per-

1928

# module basis, this configuration overrides the global config.

1929

nodeSelector:

1930

tolerations:

1931

affinity:

1932

# by choosing the default vectorizer module, you can tell Weaviate to always

1933

# use this module as the vectorizer if nothing else is specified. Can be

1934

# overwritten on a per-class basis.

1935

# set to text2vec-transformers if running with transformers instead

1936

default_vectorizer_module: none

1937

# It is also possible to configure authentication and authorization through a

1938

# custom configmap The authorization and authentication values defined in

1939

# values.yaml will be ignored when defining a custom config map.

1940

custom_config_map:

1941

enabled: false

1942

1943

# runtime_overrides is a way to override some weaviate configs during runtime without needing pod restarts.

1944

runtime_overrides:

1945

enabled: false

1946

path: "/config/overrides.yaml"

1947

load_interval: "2m"

1948

# Uncomment following lines to add custom runtime overrides

1949

values: {}

1950

# maximum_allowed_collections_count: 100

1951

# auto_schema_enabled: false

1952

# async_replication_disabled: false

1953

# Pass any annotations to Weaviate pods

1954

annotations:

1955

extraVolumeMounts:

1956

extraVolumes:

1957

nodeSelector:

1958

tolerations:

1959

hostAliases:

1960

affinity:

1961

podAntiAffinity:

1962

preferredDuringSchedulingIgnoredDuringExecution:

1963

- weight: 1

1964

podAffinityTerm:

1965

topologyKey: "kubernetes.io/hostname"

1966

labelSelector:

1967

matchExpressions:

1968

- key: "app"

1969

operator: In

1970

values:

1971

- weaviate

1972

## Optionally specify priorityClass name for the pod

1973

## ref: https://kubernetes.io/docs/concepts/scheduling-eviction/pod-priority-preemption/#pod-priority

1974

1975

priorityClassName: ""

1976

globalPriorityClassName: ""

1977

# Optionally enable a pod disruption budget for a given service.

1978

# One should define either minAvailable or maxUnavailable, but not both, as they are mutually exclusive.

1979

# We recommend to set Weaviate's minAvailable value equal to the number of nodes which provide quorum.

1980

# Please note that pod disruption budgets are only applied when the corresponding module is enabled.

1981

# For instance, if the text2vec-transformers module is disabled, its associated pod disruption budget will not be deployed.

1982

# A pod disruption budget is only activated when its respective module is enabled, allowing you to configure it as needed.

1983

# ref: https://kubernetes.io/docs/tasks/run-application/configure-pdb/

1984

podDisruptionBudgets:

1985

weaviate:

1986

enabled: false

1987

spec:

1988

minAvailable: 50%

1989

# maxUnavailable: 0

1990

text2vec-transformers:

1991

enabled: false

1992

spec:

1993

minAvailable: 1

1994

# maxUnavailable: 0

1995

text2vec-contextionary:

1996

enabled: false

1997

spec:

1998

minAvailable: 1

1999

# maxUnavailable: 0

2000

multi2vec-clip:

2001

enabled: false

2002

spec:

2003

minAvailable: 1

2004

# maxUnavailable: 0

2005

reranker-transformers:

2006

enabled: false

2007

spec:

2008

minAvailable: 1

2009

# maxUnavailable: 0

2010

img2vec-neural:

2011

enabled: false

2012

spec:

2013

minAvailable: 1

2014

# maxUnavailable: 0

2015

multi2vec-bind:

2016

enabled: false

2017

spec:

2018

minAvailable: 1

2019

# maxUnavailable: 0

2020

ner-transformers:

2021

enabled: false

2022

spec:

2023

minAvailable: 1

2024

# maxUnavailable: 0

2025

qna-transformers:

2026

enabled: false

2027

spec:

2028

minAvailable: 1

2029

# maxUnavailable: 0

2030

sum-transformers:

2031

enabled: false

2032

spec:

2033

minAvailable: 1

2034

# maxUnavailable: 0

2035

text2vec-gpt4all:

2036

enabled: false

2037

spec:

2038

minAvailable: 1

2039

# maxUnavailable: 0

2040

text-spellcheck:

2041

enabled: false

2042

spec:

2043

minAvailable: 1

2044

# maxUnavailable: 0

2045

text2vec-model2vec:

2046

enabled: false

2047

spec:

2048

minAvailable: 1

2049

# maxUnavailable: 0

2050

The trusted source for open source

Talk to an expert

Privacy

Terms

© 2026 Chainguard, Inc. All Rights Reserved.
Chainguard® and the Chainguard logo are registered trademarks of Chainguard, Inc. in the United States and/or other countries.
The other respective trademarks mentioned on this page are owned by the respective companies and use of them does not imply any affiliation or endorsement.

weaviate

The trusted source for open source

Product

Solutions

Customers

Resources

Company