0.1.15

What's Changed

• Support for Anthropic Claude 3 and Phi-3 models by @Harsha-Nori
• [Feature] Metrics for Grammarless models by @riedgar-ms in #817
• Remove Instruct/Chat versions of models & introduce a new ChatTemplate API by @Harsha-Nori in #820
• Fix set_attribute by @slundberg in #835
• Fix for unstable tokenization schemes by @Harsha-Nori in #830
• Support null captures (Fix #826) by @slundberg in #834
• Allow capture of raw functions by @slundberg in #837
• [Build] Significant notebook testing updates by @riedgar-ms in #825
• [Docs] ReadTheDocs Now Works! by @riedgar-ms in #829
• add initial support for azure guidance server API by @paulbkoch in #827
• [Build] Adjust llama-cpp-python versions by @riedgar-ms in #832
• [Test] Fix for LlamaCpp tests by @paulbkoch in #839
• Remove default max token limit by @slundberg in #836
• [Build] General Maintenance by @riedgar-ms in #842
• Updated readme by @nking-1 in #810

Full Changelog: 0.1.14...0.1.15

0.1.14

This release brings a series of new features, bug fixes, and changes, including:

• Improved compatibility for loading many transformers models, including newly tested support for Phi-3
• New capabilities in the guidance.json function for generating json, including early support for pydantic schemas.
• Support for AzureAI models in new guidance.models.AzureAIStudio* classes
• Significant dependency reductions on the default guidance installation
• Small grammar optimizations
• Many improvements to the infrastructure, testing, and release pipelines :)

What's Changed

• replace catches of the ImportError exception with the more specific ModuleNotFoundError exception by @paulbkoch in #733
• fix for issue #660 where the call to llama_batch_free is failing during process shutdown by @paulbkoch in #734
• Fix #713 for mistral models in transformers by @slundberg in #740
• Update JSON API and example notebook by @riedgar-ms in #699
• remove redundant options from select() by @mmoskal in #712
• Add support for prefixItems, minItems, and maxItems for JSON array generation by @hudson-ai in #732
• [Docs] Expand documentation by @riedgar-ms in #750
• Generation with pydantic schemas by @hudson-ai in #724
• First stab at minimal mypy implementation by @ryanpeach in #738
• remove openai hard dependency by @slundberg in #763
• from guidance import guidance by @ryanpeach in #759
• Run black on codebase by @Harsha-Nori in #765
• add rust bindings by @paulbkoch in #736
• fix Transformers tokenizer issues when models have extended vocabulary and/or special tokens by @Harsha-Nori in #766
• [Install] Rework optional imports by @riedgar-ms in #767
• Update _llama_cpp.py based on new contract of llama_get_logits by @czoccolo in #769
• [Bug] Missing support for AzureAI by @riedgar-ms in #770
• [Doc] Add developer notes by @riedgar-ms in #774
• [Feature] Improved JSON Schema support by @riedgar-ms in #787
• [Bug] Fix Regex for identifying OpenAI models by @alexander-brady in #778
• Fix models like phi-3 which have a mismatched tokenizer definition and model output tensor size by @Harsha-Nori in #792
• [Feature] Add temperature to JSON generator by @riedgar-ms in #784
• [Feature] Support for Azure AI Studio by @riedgar-ms in #779
• Add guard for running build from setup.py, redirecting users to pip by @Harsha-Nori in #805
• Many testing and CI improvements by @riedgar-ms

New Contributors

• @mmoskal made their first contribution in #712
• @ryanpeach made their first contribution in #738
• @czoccolo made their first contribution in #769
• @alexander-brady made their first contribution in #778

Full Changelog: 0.1.13...0.1.14

0.1.13

What's Changed

• Fixed PyPi publishing
• Workflow for GPU Runner by @riedgar-ms in #694
• Add Llama to model test matrix by @riedgar-ms in #703

Full Changelog: 0.1.12...0.1.13

0.1.12

Lots of new features and bug fixes :)

What's Changed

• Adding Caching to OpenAI chat by @adamgordonbell in #603
• Minor Server Tweaks by @riedgar-ms in #612
• Dependency trimming by @riedgar-ms in #613
• Add Python 3.12 by @riedgar-ms in #623
• Grammar Builder for JSON Schema by @riedgar-ms in #619
• remove unused code by @paulbkoch in #629
• fix: missing tokenizer choice in AzureOpenAI subclasses by @imarquart in #641
• Pop first argument from guidance wrapper by @hudson-ai in #634
• Fix behavior of _json_schema_to_grammar for schemas with many properties by @hudson-ai in #645
• Exclude tests packages from distribution by @shawnz in #646
• Support multiple additions to ModelStream by @shawnz in #639
• Enable AzureOpenAI model testing by @riedgar-ms in #514
• CI Build Schedule by @riedgar-ms in #662
• Add extensible tests to transformers by @riedgar-ms in #663
• Update _llama_cpp.py to use new bindings by @Nking92 in #665
• Propagate errors in ModelStream by @shawnz in #640
• Fix test_transformers.py by @riedgar-ms in #673
• Together.ai Support by @minosvasilias in #654
• Move get_logits and EngineCallResponse out of the Engine.call function so that the remaining parts can be lowered to Rust or C++ in the future and for simplification of LLM servers that operate in batches by @paulbkoch in #647
• Make Nulls Terminal by @hudson-ai in #679
• [Test] Expand to other OSes by @riedgar-ms in #680
• JSON Schema Rewrite by @riedgar-ms in #651
• [Test] MacOS Issue with Server tests by @riedgar-ms in #684
• Support mutual recursion in GrammarFunctions by @hudson-ai in #683
• More JSON schema features by @hudson-ai in #685
• Improve negative test cases for JSON generation by @riedgar-ms in #693
• Separate out server tests by @riedgar-ms in #688
• Add fixture for an LLM by @riedgar-ms in #674
• Removed side effect on the stop argument from _gen by @VincentToups in #677
• Fix for test_gen by @riedgar-ms in #695
• Fix #681 and #682 by @slundberg in #704

New Contributors

• @adamgordonbell made their first contribution in #603
• @imarquart made their first contribution in #641
• @hudson-ai made their first contribution in #634
• @shawnz made their first contribution in #646
• @Nking92 made their first contribution in #665
• @minosvasilias made their first contribution in #654
• @VincentToups made their first contribution in #677

Full Changelog: 0.1.11...0.1.12

0.1.11

What's Changed

• Fix GPU usage by removing device from Transformers class wrapper to use the device/device_map directly exposed by HF Transformers in kwargs by @cpcdoy in #569
• Adding streaming support under guidance v0.1 by @hodlen in #568
• Fix typos in intro_to_guidance.ipynb by @trungsudo in #574
• Update "Basic Tutorial" link in README.md by @alexandremuzio in #580
• Prompt role blocks debug output + new prompt implementation example by @cpcdoy in #578
• WIP: fix: AzureOpenAI authentication to support API Key auth by @younes-io in #583
• Allow client/server guidance execution by @slundberg in #586

New Contributors

• @cpcdoy made their first contribution in #569
• @hodlen made their first contribution in #568
• @trungsudo made their first contribution in #574
• @alexandremuzio made their first contribution in #580
• @younes-io made their first contribution in #583

Full Changelog: 0.1.10...0.1.11

0.1.10

0.1.9 was missing the hugging face patch...so 0.1.10 is real 0.1.9 :)

Full Changelog: 0.1.9...0.1.10

0.1.9

What's Changed

• Bug patch for HuggingFace transformers tokenization issue.
• Fixed generation coloring for multiple sequence generator calls.
• Bug fix for the base AzureOpenAI class by @riedgar-ms in #543
• Google AI Studio's Gemini by @Anilturaga in #550
• add react notebook by @Sam1320 in #547

Full Changelog: 0.1.8...0.1.9

0.1.8

What's Changed

• Multi-model Gemini model support.
• Local model temperature fixes.
• Native Anthropic Claude support by @Anilturaga in #537

New Contributors

• @Anilturaga made their first contribution in #537

Full Changelog: 0.1.7...0.1.8

0.1.7

What's Changed

• Temperature controls now work for remote models.
• When optimistically running remote models grammar violations now have much better error messages/debugability.
• Huggingface transformers models now fully respect the training tokenization patterns.
• DOC Update notebooks by @riedgar-ms in #491
• Update tutorial for Azure OpenAI by @riedgar-ms in #468
• move llama.cpp batch object allocation into a separately allocated co… by @paulbkoch in #508
• Drop unused nest_asyncio by @evelkey in #510
• Fixed compile error when using Microsoft toolchain. by @wmiller in #511
• Using Model in a loop by @riedgar-ms in #512
• Set base_url default as None and remove redundant setting by @tshu-w in #519
• Adding outputs for Azure notebook by @riedgar-ms in #533
• Further notebook updates by @riedgar-ms in #522
• Update tutorial.ipynb by @mwieler in #534

New Contributors

• @evelkey made their first contribution in #510
• @wmiller made their first contribution in #511
• @tshu-w made their first contribution in #519
• @mwieler made their first contribution in #534

Full Changelog: 0.1.6...0.1.7

0.1.6

What's Changed

• preliminary support for log_probs using the compute_log_probs=True model constructor arg.
• A new guidance.cpp module that allows us to have high speed implementations of key objects.
• reduce memory requirements (and probably improve speed) for llama.cpp tests by @paulbkoch in #485
• Many bug fixes :)

Full Changelog: 0.1.5...0.1.6