v0.8.0

0.8.0 (2025-01-29)

Changed

• ragbits-cli updated to version v0.8.0
• ragbits-conversations updated to version v0.8.0
• ragbits-document-search updated to version v0.8.0
  • DocumentSearch ingest accepts now a simple string format to determine sources; for example gcs://bucket/* (#264).
  • New CLI command to ingest documents (#305).
  • Add support for rerankers library (#284).
• ragbits-evaluate updated to version v0.8.0
• ragbits-guardrails updated to version v0.8.0
• ragbits-core updated to version v0.8.0
  • Add support for pgvector as VectorStore (#267).

v0.7.0

0.7.0 (2025-01-21)

Changed

• ragbits-cli updated to version v0.7.0
• ragbits-conversations updated to version v0.7.0
  • Added last message recontextualization (#271).
• ragbits-document-search updated to version v0.7.0
  • Add CLI command to perform search on DocumentSearch instance (#290).
  • New way to initialize DocumentSearch instances (#277).
  • ragbits-core updated to version v0.7.0
• ragbits-evaluate updated to version v0.7.0
  • Simplified interface to document-search evaluation (#258).
• ragbits-guardrails updated to version v0.7.0
• ragbits-core updated to version v0.7.0
  • Add nice-looking CLI logging for audit module (#273).
  • Add support for returning metadata from LLMs (#274).
  • Fix: limiting in qdrant vector store (#282).
  • Refactor: remove LLM client abstraction and lift it up to LLM (#270).

v0.6.0

0.6.0 (2024-12-27)

Changed

• ragbits-cli updated to version v0.6.0
  • Better error handling when dynamic importing fails in the CLI (#259).
  • Add option to choose what columns to display in the output (#257).
• ragbits-core updated to version v0.6.0
  • Add option to pass LiteLLM router to LLM instances (#262).
  • Add commands to browse vector stores (#244).
  • Implement generic Options class (#248).
  • Fix LiteLLM crash in python 3.13 (#245).
• ragbits-document-search updated to version v0.6.0
• ragbits-evaluate updated to version v0.6.0
• ragbits-guardrails updated to version v0.6.0

v0.5.1

0.5.1 (2024-12-09)

Changed

• ragbits-cli updated to version v0.5.1
• ragbits-document-search updated to version v0.5.1
• ragbits-evaluate updated to version v0.5.1
  • Document search evaluation now returns all Element types, rather than only TextElements (#241).
• ragbits-guardrails updated to version v0.5.1
• ragbits-core updated to version v0.5.1
  • Refactor: added standardized way to create ragbits objects from config (#233).

v0.5.0

0.5.0 (2024-12-05)

Changed

• ragbits-cli updated to version v0.5.0
  • Add global flag to specify output type: text or json (#232).
• ragbits-document-search updated to version v0.5.0
  • Distributed ingestion with usage of https://www.ray.io/ (#207)
  • Documents can be now replaced in existing VectorStore (#210)
  • Providers are now loaded dynamically (#219)
• ragbits-evaluate updated to version v0.5.0
• ragbits-guardrails updated to version v0.5.0
• ragbits-core updated to version v0.5.0
  • Default LLM factory when configuration is not provided (#209).
  • Add remove operation to VectorStore (#210).
  • Install litellm package by default (#236).

v0.4.0

0.4.0 (2024-11-27)

Changed

• ragbits-cli updated to version v0.4.0
• ragbits-document-search updated to version v0.4.0
  • Add support for batch ingestion (#185).
  • Ingesting images is now supported (#172).
• ragbits-evaluate updated to version v0.4.0
  • Introduced optimization with optuna (#177).
  • Add synthetic data generation pipeline (#165).
• ragbits-guardrails updated to version v0.4.0
• ragbits-core updated to version v0.4.0
  • Add support for Qdrant VectorStore (#163).
  • Add streaming interface to LLMs (#188).
  • Better images support in Prompt abstractions (#201).

v0.3.0

0.3.0 (2024-11-06)

Changlog

Added

• ragbits-guardrails v0.3.0:
  • Initial release of the package (#169).
  • First guardrail with OpenAI Moderation.

Changed

• ragbits-cli updated to version v0.3.0
• ragbits-document-search updated to version v0.3.0
  • Add location metadata to documents ingested into DocumentSearch (#122).
  • Add LiteLLM Reranker (#109).
  • ragbits-core updated to version v0.3.0
  • refactor: Add dynamic loading for modules that depend on optional dependencies (#148).
  • refactor: change the type in from_source method to Source (#156).
  • refactor: unified API for text representations of Element models (#171).
• ragbits-evaluate updated to version v0.3.0
• ragbits-core updated to version v0.3.0
  • Observability toolset, with initial support to export traces to OpenTelemetry (#168)
  • CLI commands to render / exec prompts (#146)
  • Support of images in Prompt abstractions (#149)
  • Support for different MetadataStores in VectorStore (#144)
  • Now LLMs can be configured separately for vision, text and structured. (#153)
  • refactor: Add dynamic loading for modules that depend on optional dependencies (#148).
  • refactor: Refactor vector store public API. (#151)

New Contributors

• @kdziedzic68 made their first contribution in #149
• @ds-ragbits-robot made their first contribution in #176

Full Changelog: v0.2.0...v0.3.0

v0.2.0

0.2.0 (2024-10-23)

Added

• ragbits-evaluate v0.2.0:
  • Initial release of the package (#91).
  • Evaluation pipeline framework with capability to define evaluators & metrics.
  • Evaluation pipeline for ragbits-document-search.

Changed

• ragbits-cli updated to version v0.2.0
  • Improved performance by lazy-loading the modules (#111 #113 #120)
• ragbits-document-search updated to version v0.2.0
  • Creation of DocumentSearch instances from config (#62).
  • Automatic detection of document type (#99).
  • LLM-based query rephrasing (#115).
  • Batch ingestion from sources (#112).
  • Support to image formats (#121).
  • HuggingFace sources (#106).
• ragbits-core updated to version v0.2.0
  • Project README.md (#103).
  • Listing entries API for VectorStores (#138).
  • Overrides for prompt discovery configurable in pyproject.toml file (#101).
  • Default LLM factory configurable in pyproject.toml file (#101).
  • Fixed bug in chromadb while returning multiple records (#117).
  • Fixed bug in prompt rendering for some pydantic models (#137).