0.6.0

Release Highlights: Version 0.6.0 🌟

We're thrilled to announce the release of Version 0.6.0 for instructor. This version introduces a myriad of enhancements, new features, and critical fixes that significantly improve the developer experience. Here's what's new:

Features

• Instructor Hub Launch: Discover the newly introduced Instructor Hub, a central place for tutorials, examples, and a new CLI, making it easier for developers to get up and running.
• Enhanced Patching Methods: The introduction of new patching methods and modes brings flexibility and power to the customization capabilities of the library.
• Blog Addition: Dive into our latest blog post on llama-cpp-python and instructor library usage, offering insights and best practices.
• SQLModel Integration: A major leap forward with feat(Instructor): integrate with SQLModel, providing seamless integration, along with comprehensive documentation and examples.

Improvements and Fixes

• Workflow and Error Handling Enhancements: Modifications to GitHub Actions and parallel processing error handling improve reliability and developer workflow.
• Documentation Overhaul: Significant updates to documentation, including new reference links, formatting fixes, and synchronization of requirements, ensure developers have access to the latest and most accurate information.
• Type Annotations: The inclusion of types to various components enhances code readability and maintainability.

New Contributors

A special shoutout to our new contributors @ryanhalliday, @teome, @leobeeson, and @shiftbug for their valuable additions to the project.

We're excited for you to try out the new features and improvements in Version 0.6.0. Your feedback is invaluable to us, so please don't hesitate to share your thoughts and experiences as you explore this release. Happy coding!

What's Changed

• chore: Include types to instructor.distil and tests by @savarin in #396
• feat(Instructor): integrate with SQLModel, provide documentation and example by @shanktt in #418
• chore: Include types to instructor.dsl by @savarin in #419
• fix(github-actions): modify triggering conditions for workflows by @jxnl in #420
• fix(parallel): enhance error handling in get_types_array and add test case by @jxnl in #423
• docs(documentation): update reference link, add new file, modify mkdocs.yml by @jxnl in #424
• docs: small formatting fixes by @ryanhalliday in #427
• docs: Sync requirements-doc.txt with Poetry dev group till mkdocs works by @ryanhalliday in #428
• docs(patching): introduce new patching methods and modes, update documentation by @jxnl in #431
• fix(validation): increase max_retries from 1 to 2 by @jxnl in #433
• feat(blog): Add new post on llama-cpp-python and instructor library usage by @jxnl in #434
• fix: json modes - don't add json schema again in system message by @teome in #435
• chore: Include types to instructor.patch by @savarin in #422
• Revert "chore: Include types to instructor.patch" by @jxnl in #437
• feat(instructor/patch.py): change default mode from FUNCTIONS to TOOLS by @jxnl in #436
• Include new model names in _alias.py by @leobeeson in #426
• Update ollama.md by @shiftbug in #438
• Update pricing of models, and include 1106 and 0125 models. by @leobeeson in #425
• feat(Instructor): introduce Instructor Hub with tutorials, examples, and new CLI by @jxnl in #439

New Contributors

• @ryanhalliday made their first contribution in #427
• @teome made their first contribution in #435
• @leobeeson made their first contribution in #426
• @shiftbug made their first contribution in #438

Full Changelog: 0.5.2...0.6.0

0.5.2

What's Changed

• chore(pull_request_template): update to include conventional commits instructions by @jxnl in #411
• Update init.py to include handling models and remove unpatch in _… by @PrathamSoni in #409
• fix: Adding tests to make sure non-stream iterables work by @jxnl in #413

New Contributors

• @PrathamSoni made their first contribution in #409

Full Changelog: 0.5.1...0.5.2

0.5.1

What's Changed

• Include types to instructor.function_calls and tests by @savarin in #394
• lint documentation by @jxnl in #403
• Test all of our documentation. by @jxnl in #404
• CICD: Add set -e -o pipefail to ruff and mypy. by @gao-hongnan in #399
• fix: Fixed links in image to add copy by @lakshyaag in #367

Full Changelog: 0.5.0...0.5.1

0.5.0

What's Changed

• Be robust to null values while updating token counts by @indigoviolet in #359
• Together ai support, blog post by @jxnl in #366
• Fixes prompt typo by @skrawcz in #368
• Clean up streaming code by @jxnl in #377
• Types!!! by @jxnl in #372
• Implement Parallel Function Calls with List[Union[T]] by @jxnl in #378
• Attempt to implement new retires by @jxnl in #386
• Fix sp. errors and light copy editing. by @jhochenbaum in #388
• Remove unused imports by @savarin in #393
• Include types to most upstream modules by @savarin in #391
• Include types to instructor.cli.jobs by @savarin in #392
• Fix typos by @cdreetz in #395

New Contributors

• @indigoviolet made their first contribution in #359
• @skrawcz made their first contribution in #368
• @jhochenbaum made their first contribution in #388
• @savarin made their first contribution in #393
• @cdreetz made their first contribution in #395

Full Changelog: 0.4.8...0.5.0

0.4.8

What's Changed

• Adding support for caching async functions by @ivanleomk in #340
• fix a few spelling mistakes & typos by @petrus-jvrensburg in #349
• Update the readme and remove some wording duplication. by @thedayisntgray in #351
• Add ollama run.py with completion functionality by @Tedfulk in #352
• Ollama part2 by @jxnl in #356

New Contributors

• @petrus-jvrensburg made their first contribution in #349
• @thedayisntgray made their first contribution in #351

Full Changelog: 0.4.7...0.4.8

0.4.7

Improvements with field streaming and usage tokens

What's Changed

• Updated the Knowledge Graph article by @ivanleomk in #327
• test ollama and litellm by @jxnl in #326
• Revert "test ollama and litellm" by @jxnl in #328
• Update fields.md - fix typo, missing comma by @MrJarnould in #329
• Update function call mode and add deprecation warning by @Phodaie in #337
• Introduce total_usage variable to track cumulative token usage by @lazyhope in #343
• Field level streaming by @shanktt in #334

New Contributors

• @MrJarnould made their first contribution in #329
• @lazyhope made their first contribution in #343

Full Changelog: 0.4.6...0.4.7

0.4.6

What's Changed

• Update README.md by @lakshyaag in #291
• doc: fix typos and grammatical errors in Concepts by @lakshyaag in #292
• docs: example on Vision API (image to ad copy) by @lakshyaag in #293
• Improve tutorial quality by @jxnl in #298
• add ellipsis.yaml config file; by @hbrooks in #300
• Added example for using unions of models, useful for agents by @zby in #299
• More Tutorial Updates by @jxnl in #303
• [docs] Add example link on index page by @lakshyaag in #305
• Integrate and Enhance Linting with Ruff by @gao-hongnan in #295
• Update ellipsis.yaml by @jxnl in #312
• json scheme example change by @Phodaie in #316
• Small documentation updates by @CSRessel in #313
• init.py added for testing individual files in tests/openai by @zby in #318
• Add default openai_client in LLM Validator by @asimkhan73301 in #311
• Mc/toolmaxretry by @marcasty in #323
• Update Maybe typing by @dalberto in #325

New Contributors

• @lakshyaag made their first contribution in #291
• @hbrooks made their first contribution in #300
• @zby made their first contribution in #299
• @CSRessel made their first contribution in #313
• @asimkhan73301 made their first contribution in #311
• @marcasty made their first contribution in #323
• @dalberto made their first contribution in #325

Full Changelog: 0.4.5...0.4.6

0.4.5

What's Changed

• Update anyscale.md by @robertnishihara in #285
• example: Table extraction w/ Pydantic by @jxnl in #288
• Azure support in multitask.py by @arnavw in #290

New Contributors

• @robertnishihara made their first contribution in #285
• @arnavw made their first contribution in #290

Full Changelog: 0.4.3...0.4.5

0.4.4

What's Changed

• Add logging.error for Retry logic by @jxnl in #259
• Update contributing.md by @tadash10 in #241
• Adds ruff linting and formatting by @ggaabe in #262
• Json mode nested models by @shanktt in #263
• Fix async streaming by @Anmol6 in #261
• Add multitask tests by @jxnl in #265
• Parse availability of dates by @jxnl in #267
• Integrate and Enhance Type Checking with Mypy by @gao-hongnan in #264
• Add new Mode for Anyscale's json schema schema type by @jxnl in #273
• Revert "Add new Mode for Anyscale's json schema schema type " by @jxnl in #274
• Support JSON_OBJECT mode from Anyscale by @jxnl in #275
• Updates to Tutorials by @jxnl in #248
• added finish reason exception IncompleteOutputException by @ionflow in #279
• doc: small change to readablity. by @Tedfulk in #281
• feat: add debugging for retries by @jxnl in #283
• blog: Third party models by @Anmol6 in #284

New Contributors

• @tadash10 made their first contribution in #241
• @ggaabe made their first contribution in #262
• @shanktt made their first contribution in #263
• @gao-hongnan made their first contribution in #264
• @ionflow made their first contribution in #279
• @Tedfulk made their first contribution in #281

Full Changelog: 0.4.2...0.4.3

0.4.2

What's Changed

• Opensource examples - Runpod w/Text-Generation-WebUI API Endpoint by @PhiBrandon in #247
• Fixed the system prompt for JSON mode, enabling the use of Pydantic nested models by @aastroza in #249
• Fix "seperate" typo by @toolittlecakes in #251
• added missing angle bracket by @toolittlecakes in #252
• Small doc fix for FastAPI section by @waseemhnyc in #253

New Contributors

• @toolittlecakes made their first contribution in #251
• @waseemhnyc made their first contribution in #253

Full Changelog: 0.4.1...0.4.2