Skip to content

Releases: cequence-io/pinecone-scala

Version 1.1.0

01 Sep 20:41
Compare
Choose a tag to compare
  • PineconeInferenceService - new function rerank with examples and test
  • PineconeInferenceFactory, PineconeAssistantFactory, and PineconeAssistantFileServiceFactory simplified/unified.

Version 1.0.0

25 Jul 11:50
Compare
Choose a tag to compare
  • New service PineconeInferenceService with createEmbeddings function
  • Two services for the assistant API introduced - PineconeAssistantService (listAssistants, createAssistant, etc), and PineconeAssistantFileService (chatWithAssistant, uploadFile, ...)
  • Demo projects moved as seprate modules: examples and openai-examples
  • New serverless regions added
  • Migrated to ws-client-play (io.cequence) as backend

Version 0.1.4

11 Jul 11:02
Compare
Choose a tag to compare

Migration to the newest ws-client (with Play engine/backend)

Version 0.1.3

02 Jul 11:56
Compare
Choose a tag to compare
  • Full support for pod-based and serverless index services: PineconePodBasedIndexService and PineconeServerlessIndexService
  • New PineconeInferenceService implementing embedData endpoint
  • Migration of infra to ws-client (io.cequence)

Version 0.1.2

06 May 06:28
Compare
Choose a tag to compare
  • QuerySettings - filter supporting Any value type
  • New endpoint listVectorIds with a response holder and JSON format
  • Proto support for serverless indexes (env optional)

Version 0.1.1

21 Dec 11:07
Compare
Choose a tag to compare
  • WS client close function exposed in PineconeIndexService and PineconeVectorService
  • Prep for Scala 3.3 support - Play WS libs added

Version 0.1.0

17 May 10:12
Compare
Choose a tag to compare
  • Refactoring EnumFormat such that it's Scala-2 and Scala-3 compatible
  • IndexDatabaseInfo -> pod_type made optional - undefined for indexes running on "free" envs
  • Handy new fun getMetadata(key: String) introduced for Match data holder

Version 0.0.1

27 Apr 14:01
Compare
Choose a tag to compare

Initial version already covering all the vector, index, and collection endpoints.