Releases: PrefectHQ/marvin
Releases · PrefectHQ/marvin
v1.1.0
What's Changed
- Add .map for AI Functions by @jlowin in #441
- Add streaming support to engine + AI applications by @jlowin in #444
- Remove "made with mkdocs" from docs footer by @discdiver in #443
- Track llm_response on Message objects and AI Models by @jlowin in #446
- Flesh out API docs by @jlowin in #448
- Consolidate github tool by @jlowin in #449
- Minor streaming bugfix by @jlowin in #450
- modularize chroma tool and indicate
np
req forEmbeddings.acreate
by @zzstoatzz in #447 - rm schedule for community image push by @zzstoatzz in #456
- tailwind frontpage by @aaazzam in #457
- adds some test coverage to Function class by @aaazzam in #454
- Instructions to ai fn by @aaazzam in #455
- Support multiple providers + Anthropic support by @jlowin in #458
- Housekeeping: move
models
underutilities
by @jlowin in #459 - Expand AI Application docs by @jlowin in #461
- Add availability note for Anthropic by @jlowin in #460
- Abstract each provider into its own settings namespace by @jlowin in #462
- Fix AI Model bugs with instructions and add docs by @jlowin in #467
- Correct the docs for setting the Open AI API key by @Donavan in #468
- Reflect 1.0-style settings by @jlowin in #472
- Don't assign to the openai global variable by @jlowin in #474
- Add support for Azure OpenAI service by @jlowin in #466
- Add .map method for AI Models by @jlowin in #465
- Remove global default model + fix tokenization by @jlowin in #480
- enable
instructions
on ai_classifier by @aaazzam in #470 - Classifier enhancements: mapping, guard against non-openai by @jlowin in #481
- Update AI Classifier docs by @jlowin in #483
- Ensure model is always passed to executor by @jlowin in #482
- Rm agents (undocumented) by @aaazzam in #485
- Add context sizes for all models by @jlowin in #484
- add NLI references by @jlowin in #489
embedding_engine
setting moved by @zzstoatzz in #488- mv apps to /cookbooks by @aaazzam in #486
- ensure multiqueried excerpts are not too small by @zzstoatzz in #490
- Fix type annotation for
ai_fn
by @yasyf in #469 - run pre-commits by @zzstoatzz in #491
New Contributors
Full Changelog: v1.0.2...v1.1.0
v1.0.2
v1.0.1
v1.0.0
What's Changed
- add handler for keyword responses by @zzstoatzz in #349
- Update pyproject.toml langchain minimum version to 0.0.200 by @discdiver in #355
- Resolve conflicts between fn and run by @jlowin in #353
- add host as input argument for server by @jicuss in #347
- First version of AIApplications by @jlowin in #359
- Clean up application repeat mode by @jlowin in #362
- Continue enhancing apps: task tracking, internal state, nesting by @jlowin in #365
- adopt single yaml config by @zzstoatzz in #364
- make sure we have
FunctionMessage
by @zzstoatzz in #366 - Use OpenAI API directly where appropriate by @jlowin in #367
- more ai fn examples by @zzstoatzz in #369
- Clean up app definitions by @jlowin in #370
- Improve AI app reliability by @jlowin in #371
- Marvin next / planner by @jlowin in #376
- chore: remove codeql from image build workflows by @jamiezieziula in #380
- add tools for slackbot by @zzstoatzz in #379
- lightweight marvin by @zzstoatzz in #381
netlify.toml
by @zzstoatzz in #382- fix: Netlify docs build by @zzstoatzz in #384
- Update pytest-rerunfailures requirement from ~=10.0 to >=10,<13 by @dependabot in #388
- flatten component dir, add SDK reference via
mkdocstring
by @zzstoatzz in #389 - adds ChatCompletion subclass by @aaazzam in #392
- Function registry refactor by @aaazzam in #393
- rm invalid
site_copy
and netlify in favor of typical docs build CI by @zzstoatzz in #395 - add ai app tests by @zzstoatzz in #386
- ai_choice fully customizable by @aaazzam in #391
- move
DiscoursePost
into new place for pre-built components by @zzstoatzz in #390 - Function registry refactor by @aaazzam in #396
- fix publish pypi workflow by @zzstoatzz in #397
- improve
marvin chat
by @zzstoatzz in #400 - make assorted docs updates and improve cli + small bug fixes by @zzstoatzz in #401
- Add async util compatible with jupyter by @jlowin in #403
- Add getting started by @jlowin in #402
- Add docs overview by @jlowin in #406
- Set openai api key by @jlowin in #407
- More docs tweaks by @zzstoatzz in #408
- only try to get secret if it is defined by @zzstoatzz in #405
- deployment docs and QOL improvements for adding routes by @aaazzam in #409
get_secret_value
in validator by @zzstoatzz in #411- Add hero image by @jlowin in #413
- Fix chroma env vars by @zzstoatzz in #412
- add a bit more color to guide, stop printing panel fo user in marvin … by @zzstoatzz in #404
- adding AI choice to quickstart docs by @aaazzam in #410
- Tweets + Json + rm Code Generation by @aaazzam in #414
- better hero header by @aaazzam in #416
- simple ai app docs by @zzstoatzz in #415
- Move docs by @jlowin in #418
- add windows tests back by @zzstoatzz in #417
- increase hero img size by @zzstoatzz in #419
- word by @zzstoatzz in #421
- Update README.md by @jlowin in #420
- Update intro by @jlowin in #425
- fix typo in docs by @discdiver in #422
- add github digest flow by @zzstoatzz in #426
- Add AI applications to overview by @jlowin in #427
- choice → classifier by @jlowin in #428
- Add core primitives to overview & expand motivation by @jlowin in #429
- smart pirate by @aaazzam in #430
New Contributors
Full Changelog: v0.10.0...v1.0.0
v1.0.0rc1
What's Changed
- add handler for keyword responses by @zzstoatzz in #349
- Update pyproject.toml langchain minimum version to 0.0.200 by @discdiver in #355
- Resolve conflicts between fn and run by @jlowin in #353
- add host as input argument for server by @jicuss in #347
- First version of AIApplications by @jlowin in #359
- Clean up application repeat mode by @jlowin in #362
- Continue enhancing apps: task tracking, internal state, nesting by @jlowin in #365
- adopt single yaml config by @zzstoatzz in #364
- make sure we have
FunctionMessage
by @zzstoatzz in #366 - Use OpenAI API directly where appropriate by @jlowin in #367
- more ai fn examples by @zzstoatzz in #369
- Clean up app definitions by @jlowin in #370
- Improve AI app reliability by @jlowin in #371
- Marvin next / planner by @jlowin in #376
- chore: remove codeql from image build workflows by @jamiezieziula in #380
- add tools for slackbot by @zzstoatzz in #379
- lightweight marvin by @zzstoatzz in #381
netlify.toml
by @zzstoatzz in #382- fix: Netlify docs build by @zzstoatzz in #384
- Update pytest-rerunfailures requirement from ~=10.0 to >=10,<13 by @dependabot in #388
- flatten component dir, add SDK reference via
mkdocstring
by @zzstoatzz in #389 - adds ChatCompletion subclass by @aaazzam in #392
- Function registry refactor by @aaazzam in #393
- rm invalid
site_copy
and netlify in favor of typical docs build CI by @zzstoatzz in #395 - add ai app tests by @zzstoatzz in #386
- ai_choice fully customizable by @aaazzam in #391
- move
DiscoursePost
into new place for pre-built components by @zzstoatzz in #390 - Function registry refactor by @aaazzam in #396
- fix publish pypi workflow by @zzstoatzz in #397
- improve
marvin chat
by @zzstoatzz in #400 - make assorted docs updates and improve cli + small bug fixes by @zzstoatzz in #401
- Add async util compatible with jupyter by @jlowin in #403
- Add getting started by @jlowin in #402
- Add docs overview by @jlowin in #406
- Set openai api key by @jlowin in #407
- More docs tweaks by @zzstoatzz in #408
- only try to get secret if it is defined by @zzstoatzz in #405
- deployment docs and QOL improvements for adding routes by @aaazzam in #409
get_secret_value
in validator by @zzstoatzz in #411- Add hero image by @jlowin in #413
- Fix chroma env vars by @zzstoatzz in #412
- add a bit more color to guide, stop printing panel fo user in marvin … by @zzstoatzz in #404
- adding AI choice to quickstart docs by @aaazzam in #410
- Tweets + Json + rm Code Generation by @aaazzam in #414
- better hero header by @aaazzam in #416
- simple ai app docs by @zzstoatzz in #415
- Move docs by @jlowin in #418
- add windows tests back by @zzstoatzz in #417
- increase hero img size by @zzstoatzz in #419
- word by @zzstoatzz in #421
- Update README.md by @jlowin in #420
- Update intro by @jlowin in #425
- fix typo in docs by @discdiver in #422
- add github digest flow by @zzstoatzz in #426
- Add AI applications to overview by @jlowin in #427
- choice → classifier by @jlowin in #428
- Add core primitives to overview & expand motivation by @jlowin in #429
- smart pirate by @aaazzam in #430
New Contributors
Full Changelog: v0.10.0...v1.0.0rc1
v0.10.0
v0.9.2
What's Changed
- Remove quoted titles by @jlowin in #319
- add explicit default timeout by @zzstoatzz in #320
- configurable feedback mechanism + thread summarizer by @zzstoatzz in #308
- pin chroma by @zzstoatzz in #321
- use sharable config for allow list users by @zzstoatzz in #322
- move out prompt generation from _run to own _method by @aaazzam in #328
- implements mapping for
ai_fn
andai_model
by @zzstoatzz in #330
Full Changelog: v0.9.1...v0.9.2
v0.9.1
What's Changed
- quote
pip install
by @zzstoatzz in #299 - Support new langchain callback interfaces by @jlowin in #301
- clarify language in ai_models.md by @discdiver in #300
- Don't load prefect plugin on import by @jlowin in #306
- fix import on windy city hero img by @zzstoatzz in #311
- add setting for openai timeout by @zzstoatzz in #312
Full Changelog: v0.9.0...v0.9.1
v0.9.0
v0.8.3
What's Changed
- ai_models for structured parsing by @aaazzam in #279
- QA slackbot by @zzstoatzz in #282
- QA + Cloud Run deploy slackbot docs by @zzstoatzz in #290
- Update cloud run docs by @zzstoatzz in #291
- Update db docs about Cloud SQL connect by @geminixiang in #268
- Use Pydantic formatter for AI Models by @jlowin in #292
- ai_models doc init by @aaazzam in #293
- formatting and mkdocs.yaml update by @zzstoatzz in #295
- add knowledge update script by @zzstoatzz in #284
New Contributors
- @geminixiang made their first contribution in #268
Full Changelog: v0.8.2...v0.8.3