Release v1.13.0

What's Changed

• Dream distribution with SF-based DM by @nstsj in #572
• Feat/emotions multimodal (video) by @moon-strider in #564
• Add gigachat by @RafilGalimzyanov in #604

Full Changelog: v1.12.0...v1.13.0

Release v1.12.0

• update property extraction #576
• update OpenAI version to 1.2.3 #592
• Dream emotion dist #568
• Feat/update openai #593
• Dream_kg_prompted distribution #567
• feat: add mistral 7b 128k #589
• fix: user-km and skill tests for dream_kg #595
• feat: add system requirements #599
• Add configs for gpt4-turbo #598
• Add configs to envvars_to_send (08b5e03)
• Merge pull request #600 [from deeppavlov/fix/envvars-to-send]
• User km updates #597
• Feat/dff/voice skill #362
• component readme updates #536
• Dream kg prompted port fix #601
• Bot kg annotator #586
• fromage image skill fixed #603

Release v1.11.1

• updates to dream_kg distribution #581
• Feat/pytest tests #548
• Modify davinci config so it works better in russian #587
• fix: response selectors tags #588
• feat: dream russian persona based #590

Release v1.11.0

• update property extraction #482
• Feat/ocean #559
• Feat/mint #505
• Fix/midas cls #560
• user knowledge graph annotator #360
• KG skills #361
• Fromage #562
• Dist n2 for ac #569
• Fix/use gen configs in chatcompletion #573
• fix: werkzeug error #574
• Ru mtl #566
• Refactoring skills display name #578
• fixing path to chatgpt skill configurations #580

Release v1.10.1

• Feat/add external integration skill #504
• Fix/llms output postprocessing #556

Release v1.10.0

• Summarization models #393
• Models table upd #539
• fix: anthropic model params #547
• fix summarization annotator card #549
• Feat/complex tasks #533
• Feat/prompted robot #550
• fix: ranking and intent based response selector #552
• fix ru response selector component cards #553
• Feat/refactor dummy skill #530

Release v1.9.0

• fix: add AI to assistants
• fix: use half_precision for gpt-jt 6b
• fix: prompt-selector-ru cards
• fix #526
• Fix/llm response selector promp #524
• fix: refactor pipeline and llm #527
• Feat/terminusdb service https://github.com/deeppavlov/dream/pull/
• Fix/dreamtools cards #531
• feat: command selector distributions #314
• fix: sizes of llms
• feat: do not use sentence ranker url from env #535
• Feat/ruxglm prompted dist #528
• Feat/ru prompted dists #532
• update components.tsv #537
• Feat/rugpt 3.5 distribution #534
• feat: instruction how to add a new model
• fix ru prompt selector, remove unused component #538
• feat: replace oasst12b with gptjt #541
• Feat/utilize rugpt35 #540
• fix: ru_dists_names_and_prompts #543
• fix: cards for ru dists #544
• fix: rugpt35 config and envs #546

Release v1.8.2

• feat: use dream_faq in dream_openai_persona_prompted
• Feat/llm based resp selector dream #521
• fix: logs for resp selector
• Feat/doc readmes #522

Release v1.8.1

• Fix/component cards (#518)

Release v1.8.0

• Feat/add vicuna no dists #507
• Feat/edit fallback responses (#506)
• Move services to deeppavlov 1.2.0 (#495)
• fix pydantic under 2.0 (#509)
• Feat/new main dist (#508)
• fix: models.md indentation
• fix: models.md grammar
• Odqa with llm (#412)
• Add/reasoning and planning (#499)
• fix: assistants requirements
• New changes since you last viewed
• removed dff-google-api-skill from wait hosts in dream and dream_persona_openai_prompted (#513)
• fixed source in description_based_skill_selector (#512)
• fixed dockerfile path for reasoning skill (#514)
• Key components readmes - upd (#515)
• Feat/readme for prompted dists (#516)
• dream faq prompted skill (#517)