1.1.0

What's Changed

• Ingester & Retriever: Add support for Weaviate by @yassinsws in #64
• LLM: Remove BasicRequestHandler by @Hialus in #106
• Tutor Chat Pipeline with Lecture content. by @yassinsws in #104
• Pin Docker to 3.12.3 to fix Pydantic failure on production, by @yassinsws in #123
• LLM: Fix Ollama URL handling by @Hialus in #121
• CI: Add old deployment pipeline by @Hialus in #125
• Pipeline: Ingestion pipeline by @yassinsws in #96
• Feature/basic retrieval by @yassinsws in #122
• Pipeline: Interaction Suggestion Pipeline by @kaancayli in #115
• Fix merge conflicts of main by @yassinsws in #126
• Replace gpt-4-vision with gpt-o by @yassinsws in #127
• Exercise-Chat: Readd exercise chat suggestion by @kaancayli in #128
• Development: Add basic Sentry support by @Hialus in #134
• Development: Add Sentry reporting for the pipelines by @Hialus in #135
• Bugfix/index and parenthesis bug fixes by @yassinsws in #138
• Fix competency extraction feature by @MichaelOwenDyer in #145
• Enhance README.MD by @kaancayli in #152
• Make readme file more beginner friendly by @yassinsws in #153
• Add requirements list to competency extraction pipeline by @MaximilianAnzinger in #158
• Docker: Production setup without Nginx by @kaancayli in #159
• FastAPI: Support new settings system by @Hialus in #149
• Improve OpenAI error handling by @MichaelOwenDyer in #164
• Add Text Exercise Chat Pipeline by @MichaelOwenDyer in #161
• Track token usage of iris requests by @alexjoham in #165
• Exercise Chat: Implement native function calling agent by @kaancayli in #154
• Use one global DB client instead of one for each pipeline by @Hialus in #171
• Add links to citations by @yassinsws in #139
• Pipeline: Add lecture chat pipeline connection by @sebastianloose in #173
• Bugfix: Set result to none after sending the result by @kaancayli in #132
• Migrate the course-chat pipeline to native tool calling agent by @kaancayli in #179
• Ingestion status callback update by @yassinsws in #142
• Add ChatGPT wrapper pipeline by @wasnertobias in #196
• FAQ rewriting pipeline by @cremertim in #191
• Add inconsistency check pipeline by @FelixTJDietrich in #197
• Add problem statement rewriting to the rewriting pipeline by @FelixTJDietrich in #192
• Enable IRIS to use FAQs by @cremertim in #187
• Improved consistency check by @FelixTJDietrich in #203
• Exercise Chat: Add citation pipeline for FAQs by @cremertim in #201

New Contributors

• @MaximilianAnzinger made their first contribution in #158
• @alexjoham made their first contribution in #165
• @sebastianloose made their first contribution in #173
• @wasnertobias made their first contribution in #196
• @cremertim made their first contribution in #191
• @FelixTJDietrich made their first contribution in #197

Full Changelog: 1.0.1...1.1.0

1.0.1

What's Changed

• Development: Update dependencies by @Hialus in #91
• Pipeline: Use capabilities instead of hardcoded LLMs by @kaancayli in #95
• LLM: Add image recognition and generation support by @yassinsws & @Hialus in #89
• LLM: Add support for JSON mode by @Hialus in #102
• Development: Fix prod docker compose by @Hialus
• Development: Fix status stage handling by @Hialus

New Contributors

• @yassinsws made their first contribution in #89

Full Changelog: v1.0.0...1.0.1

1.0.0

What's Changed

• LLM: Add llm subsystem by @Hialus in #51
• LLM: Use pydantic for config parsing by @Hialus in #52
• LLM: Add Langchain LLM wrapper classes by @Hialus in #57
• LLM: Refactoring suggestions by @MichaelOwenDyer in #59
• Pipeline: Add new pipeline DTOs by @MichaelOwenDyer in #58
• Pipeline: Add pipeline subsystem by @kaancayli in #60
• Common: Improve langchain message converters by @Hialus in #65
• LLM: Add simple capability system by @Hialus in #69
• FastAPI: Add DTOs and pipeline run endpoint by @MichaelOwenDyer in #70
• Docker: Add initial Docker deployment setup by @Hialus in #75

New Contributors

• @kaancayli made their first contribution in #60

Full Changelog: https://github.com/ls1intum/Pyris/commits/v1.0.0