v0.3.0

What's Changed

Breaking Changes

• fix!: raise ValueError for unsupported layer_index in vLLM backend by @j341nono in #20

Other Changes

• feat: add input validation and improved type hints to core.py by @j341nono in #20
• test: add comprehensive error handling tests with mocked backends by @j341nono in #20
• chore: bump version to 0.3.0 by @j341nono in #21

Full Changelog: v0.2.2...v0.3.0

v0.2.2

What's Changed

• feat!: refactor pooling strategy into orthogonal parameters (pooling, layer, template) by @j341nono
• feat: implement smart defaults for pooling method and layer index by @j341nono
• build!: make bitsandbytes a core dependency (GPU-first default) by @j341nono
• feat: add comprehensive GPU verification script (transformers+vllm) by @j341nono
• style: fix Ruff E501 line too long errors by @j341nono
• chore!: set version to 0.2.2 (breaking changes included) by @j341nono

Full Changelog: v0.2.1...v0.2.2

v0.2.1

What's Changed

• chore: bump version to v0.2.1 to fix pypi upload collision by @j341nono

Full Changelog: v0.2.0...v0.2.1

v0.2.0

What's Changed

• feat: implement batch processing in Encoder.encode by @j341nono in #14
• feat: supports vLLM Backend and improve compatibility by @rice8y in #15
• fix(transformers): improve robustness and remove redundant GC in TransformersBackend by @rice8y in #12
• ci: add GitHub Actions workflow to deploy MkDocs docs via uv by @j341nono in #11

Full Changelog: v0.1.0...v0.2.0

v0.1.0

What's Changed

• feat: implement transformers backend and vllm backend by @j341nono in #2
• feat: flexible transformers config by @j341nono in #3
• fix: vllm embedding implementation and verification script by @j341nono in #4
• fix: resolve duplicate kwarg error in vllm backend by @j341nono in #5
• chore: rename package to llm-embed and bump version by @j341nono in #7
• fix: change lib name by @j341nono in #9