Skip to main content

Langfun: Language as Functions.

Project description

logo

Langfun

PyPI version codecov pytest

Installation | Getting started

What is Langfun

Langfun is a Python library that aims to make language models (LM) fun to work with. Its design enables a programming model that flows naturally, resembling the human thought process. It emphasizes the reuse and combination of language pieces to form prompts, thereby accelerating innovation. In contrast to other LM frameworks, which feed program-generated data into the LM, langfun takes a distinct approach: It starts with natural language, allowing for seamless interactions between language and program logic, and concludes with natural language and optional structured output. Consequently, langfun can aptly be described as Language as functions, capturing the core of its methodology.

Install

pip install langfun

Or install nightly build with:

pip install langfun --pre

Hello World

import langfun as lf

class NumericAnswerExtractor(lf.LangFunc):
  """Numeric answer extractor.

  Here is my question:
  {{question}}

  Here is the response:
  {{question()}}

  Can you help me extract a number from the response as the answer to my
  question? Your response should only contain a number in numeric form.
  If the answer is not a number or you cannot extract it, respond with UNKNOWN.
  """
  output_transform = lf.transforms.Match('\d+').to_int()

l = NumericAnswerExtractor()

with lf.context(lm=lf.llms.Gpt35(debug=True)):
  r = l(question=lf.LangFunc('What is result of {{x}} plus {{y}}?'),
        x='one',
        y='two')
  print('Result:', r.result)

Disclaimer: this is not an officially supported Google product.

Project details


Release history Release notifications | RSS feed

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

langfun-0.0.2.dev20231023.tar.gz (99.5 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

langfun-0.0.2.dev20231023-py3-none-any.whl (150.9 kB view details)

Uploaded Python 3

File details

Details for the file langfun-0.0.2.dev20231023.tar.gz.

File metadata

  • Download URL: langfun-0.0.2.dev20231023.tar.gz
  • Upload date:
  • Size: 99.5 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/4.0.2 CPython/3.11.6

File hashes

Hashes for langfun-0.0.2.dev20231023.tar.gz
Algorithm Hash digest
SHA256 7d12096fe17f624630a52629244bd73503ac554f7d0a5259b4f0972c0e565d0b
MD5 862e3cef6fd2fd14acb7a506950cc941
BLAKE2b-256 5b135a8a775fd90fc9589e42e3a9034b8b0190ec427ad616750c26378e1d0d1f

See more details on using hashes here.

File details

Details for the file langfun-0.0.2.dev20231023-py3-none-any.whl.

File metadata

File hashes

Hashes for langfun-0.0.2.dev20231023-py3-none-any.whl
Algorithm Hash digest
SHA256 0f885ed84a90f34b5c4a7e4088d2811327071fa87b69721d2fef5aa2eaca075d
MD5 b54541a1adc93e4ca065d16f9415c76f
BLAKE2b-256 fd9786fb23b6ed9ecca042bf1ccf1d5b745fd2084c65f0d74e48466bd1c759ee

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page