Skip to content

Releases: LookforFPS/OJA

1.2.1

10 Dec 00:49
6bc8707

Choose a tag to compare

This release focuses on maintenance, dependency updates, and small polishing steps.

📝 Documentation

  • Fixed minor typos and improved Gradle examples (separated Groovy DSL and Kotlin DSL)

♻️ Refactorings

  • Refactored the Version class to use Number-based types
  • Slightly reordered and cleaned up the Gradle Kotlin DSL configuration

📦 Dependency & Build Management

  • Updated dependencies

1.2.0

27 Jun 15:08
7c35c99

Choose a tag to compare

This release primarily focuses on reorganization and improvements to the project’s maintainability, rather than delivering major new features. Below is a detailed summary of the updates:

✨ Features

  • Implemented a GitHub Workflow for automated validation.
  • Migrated the Gradle configuration to the Kotlin DSL.

📝 Documentation

  • Added a build status badge and improved the order of badges in the README.
  • Corrected typos and updated code examples for clarity.

♻️ Refactorings

  • Renamed all packages from me.lookforfps.oja to dev.lookforfps.oja to reflect the new project structure.
  • Renamed the groupId from me.lookforfps to dev.lookforfps.oja to reflect the new project structure.
  • Cleaned up the .gitignore.

🐛 Bugfixes

  • Fixed minor typographical errors in examples and workflow specifications.

1.1.0

15 Jan 18:13
1f9a581

Choose a tag to compare

In this release, I am excited to introduce new features, bug fixes, and several refactorings. Below is a detailed summary of the updates:

✨ Features

  • Added support for o1 models.
  • Implemented okhttp and reworked error handling.
  • Implemented moderation features and updated ModerationModel enum.

📝 Documentation

  • Added moderation example and reworked other examples.
  • Removed warning from the README.
  • Added embeddings example to the README.

♻️ Refactorings

  • Updated SystemMessage to support multiple content parts.
  • Added a shortcut for moderation image and changed the wording for moderation results.
  • Reworked logging system.
  • Refactored ChatCompletionResponse and Chunk model.
  • Shortened library names.
  • Updated setters to accept respective model enums.
  • Generalized InputNotSupportedException.
  • Refactored constructors and create methods; added error prevention for ChatCompletionService.
  • Updated assistant messages to support text, image, and refusal content parts.

🐛 Bugfixes

  • Fixed a bug where multi-modal input was not detected correctly.

🔧 Configuration

  • Added valid constructors for Function.
  • Added new fields (store, metadata, and maxCompletionTokens) to ChatCompletionConfiguration; marked maxTokens as deprecated.

🔥 Removals

  • Removed model enums for AudioModels and ImageGenerationModels.

⬆️ Dependencies

  • Upgraded dependencies.

1.0.2

19 Sep 18:44
8d76130

Choose a tag to compare

In this release, I’ve made adjustments to keep up with recent API changes related to Chat Completion and did some reworking.

Modifications due API changes:

  • added the completion_tokens_details field to Usage of Chat Completion.

Rework modifications:

  • separated the Usage model of Chat Completion from the Usage model of Embeddings
  • modified configuration of objectMapper to prevent mapping from breaking when new fields are added to the API

1.0.1

14 Aug 20:45
509c6eb

Choose a tag to compare

In this release, I’ve made adjustments to keep up with recent API changes related to Chat Completion. These updates include:

  • Adding the refusal field to AssistantMessage and Delta.
  • Implementing the new json_schema in ResponseFormat.

Additionally, I've updated a few dependencies.

1.0.0

01 Aug 18:19
30581ae

Choose a tag to compare

I'm excited to share that the OJA library has reached a point where it's ready for others to use.
Today marks the release of version 1.0.0.

Currently, the library only supports Chat Completion and Embeddings generation.
More features will be added in future updates.

This is my first public GitHub project, so please be a little lenient :)