guidance-ai / guidance

A guidance language for controlling large language models.

guidance-ai/guidance Issues

OpenAI models does not support multi modal?
Updated 7 days ago3
RuntimeError: Bad response to Guidance request (phi-3.5-mini-instruct serverless api)
Updated 12 days ago3
ollama support?
Updated 12 days ago6
reset() does not work as expected in connection with role blocks
Closed 16 days ago1
Prerelease 0.2.0rc1 no longer returns log probs with llama.cpp
Updated 16 days ago1
JSON and TypeAdapters produce unwanted values or empty list
Updated 16 days ago1
Support for Apple MLX backend
Updated a month ago2
Phi Model on Azure is Generating Random Results using the gen Grammar Function
Updated a month ago
Considering FastAPI Support for Enhanced Model Compatibility
Updated a month ago
json() function in guidance library generates content only in English, can't output in other languages
Closed a month ago10
How to cite text in guidance?
Closed a month ago2
A declarative framework on the top of guidance
Updated a month ago3
Round-trip encoding of tokens [!] failed, Warning: lexer error: too many states: 10406 >= 10000; stopping
Updated a month ago8
Default behavior of `json` generation is likely more verbose than users expect
Updated a month ago3
Do you support `json` when using `OpenAI` models?
Updated a month ago1
Bad link from Guidance README
Updated a month ago
Lots of HTML output
Updated a month ago8
[BUG] JSON: non-required properties that occur out-of-order are validated against `additionalProperties`
Closed a month ago1
[Query] True explanation behind the Guidance project?
Updated 2 months ago1
JSON: type-narrow `enum`, `oneOf`, and `anyOf`
Updated 2 months ago
support for minimum/maximum in JSON schemas
Closed 2 months ago
Transformers `past_key_values` deprecated
Updated 2 months ago4
attribute caching not thread-safe
Closed 2 months ago1
[Feature] `JSON` generation should allow `false` subschemas to blacklist keys
Updated 2 months ago1
Support for OpenAI structured output
Updated 3 months ago2
guidance.json() crash, TypeError: unhashable type: 'list'
Closed 3 months ago2
[Build] Notebook build failing after Rust merge
Updated 3 months ago
[Build] OpenAI test in CI build failing
Updated 3 months ago
Exceptions inside zero-arg stateless grammars only raise on first call
Closed 3 months ago
Telegram or Discord dedicated channels?
Closed 3 months ago2
[Phi3.5 Mini] Eval and text generation takes several times longer than with Gemma2 2B It
Closed 3 months ago2
Pypy support broken on MacOS/Linux
Updated 3 months ago
Can't use models that have bos_token null
Closed 3 months ago
Enhance docs
Closed 3 months ago4
Cannot instantiate `Transformers("microsoft/Phi-3-mini-4k-instruct")`
Closed 3 months ago15
Irrelevant link
Updated 3 months ago
CHAT_TEMPLATE_CACHE keys are not what I expect
Updated 3 months ago1
LlamaCpp always using gpt2 tokeniser
Updated 3 months ago1
Better error messages when failing to load models with sentencepiece based tokenizers
Closed 4 months ago3
Cannot append to a variable created with model.set
Closed 4 months ago2
[Bug] Latest Transformers disagrees with GPT2 on MacOS-ARM
Updated 4 months ago2
Stream() does not obey the roles function
Updated 4 months ago2
Add sample custom handler.py integration with HuggingFace
Updated 4 months ago3
dedent=True doesn't work in python 3.12
Closed 4 months ago5
LlamaCpp model crashes with multi-token characters
Updated 4 months ago5
Unable to connect to Hugging Face Dedicated Endpoints?
Updated 4 months ago3
support for transformer LoRA adapters
Updated 4 months ago5
Cannot import guidance w/ vertexai due to protobuf version mismatch
Closed 4 months ago5
`ModelStream` should be a drop-in replacement for `Model` or their API differences should be documented
Updated 5 months ago1
[Bug] `sequence()` not quite replacement for `zero_or_more()`
Updated 5 months ago