# Local LLM function calling
[![Documentation Status](https://readthedocs.org/projects/local-llm-function-calling/badge/?version=latest)](https://local-llm-function-calling.readthedocs.io/en/latest/?badge=latest) [![PyPI version](https://badge.fury.io/py/local-llm-function-calling.svg)](https://badge.fury.io/py/local-llm-function-calling)
## Overview
The `local-llm-function-calling` project is designed to constrain the generation of Hugging Face text generation models by enforcing a JSON schema and facilitating the formulation of prompts for function calls, similar to OpenAI's [function calling](https://openai.com/blog/function-calling-and-other-api-updates) feature, but actually enforcing the schema unlike OpenAI.
The project provides a `Generator` class that allows users to easily generate text while ensuring compliance with the provided prompt and JSON schema. By utilizing the `local-llm-function-calling` library, users can conveniently control the output of text generation models. It uses my own quickly sketched `json-schema-enforcer` project as the enforcer.
## Features
- Constrains the generation of Hugging Face text generation models to follow a JSON schema.
- Provides a mechanism for formulating prompts for function calls, enabling precise data extraction and formatting.
- Simplifies the text generation process through a user-friendly `Generator` class.
## Installation
To install the `local-llm-function-calling` library, use the following command:
```shell
pip install local-llm-function-calling
```
## Usage
Here's a simple example demonstrating how to use `local-llm-function-calling`:
```python
from local_llm_function_calling import Generator
# Define a function and models
functions = [
{
"name": "get_current_weather",
"description": "Get the current weather in a given location",
"parameters": {
"type": "object",
"properties": {
"location": {
"type": "string",
"description": "The city and state, e.g. San Francisco, CA",
"maxLength": 20,
},
"unit": {"type": "string", "enum": ["celsius", "fahrenheit"]},
},
"required": ["location"],
},
}
]
# Initialize the generator with the Hugging Face model and our functions
generator = Generator.hf(functions, "gpt2")
# Generate text using a prompt
function_call = generator.generate("What is the weather like today in Brooklyn?")
print(function_call)
```
## Custom constraints
You don't have to use my prompting methods; you can craft your own prompts and your own constraints, and still benefit from the constrained generation:
```python
from local_llm_function_calling import Constrainer
from local_llm_function_calling.model.huggingface import HuggingfaceModel
# Define your own constraint
# (you can also use local_llm_function_calling.JsonSchemaConstraint)
def lowercase_sentence_constraint(text: str):
# Has to return (is_valid, is_complete)
return [text.islower(), text.endswith(".")]
# Create the constrainer
constrainer = Constrainer(HuggingfaceModel("gpt2"))
# Generate your text
generated = constrainer.generate("Prefix.\n", lowercase_sentence_constraint, max_len=10)
```
## Extending and Customizing
To extend or customize the prompt structure, you can subclass the `TextPrompter` class. This allows you to modify the prompt generation process according to your specific requirements.
Raw data
{
"_id": null,
"home_page": "https://github.com/rizerphe/local-llm-function-calling",
"name": "local-llm-function-calling",
"maintainer": "",
"docs_url": null,
"requires_python": ">=3.11,<4.0",
"maintainer_email": "",
"keywords": "llm,jsonschema,huggingface,transformers,local,llama.cpp",
"author": "rizerphe",
"author_email": "44440399+rizerphe@users.noreply.github.com",
"download_url": "https://files.pythonhosted.org/packages/7c/f4/ee9a8c9cc1cbb85ac71c5dfffe933f8b23a0f925c325c5e156227e5d3eea/local_llm_function_calling-0.1.23.tar.gz",
"platform": null,
"description": "# Local LLM function calling\n\n[![Documentation Status](https://readthedocs.org/projects/local-llm-function-calling/badge/?version=latest)](https://local-llm-function-calling.readthedocs.io/en/latest/?badge=latest) [![PyPI version](https://badge.fury.io/py/local-llm-function-calling.svg)](https://badge.fury.io/py/local-llm-function-calling)\n\n## Overview\n\nThe `local-llm-function-calling` project is designed to constrain the generation of Hugging Face text generation models by enforcing a JSON schema and facilitating the formulation of prompts for function calls, similar to OpenAI's [function calling](https://openai.com/blog/function-calling-and-other-api-updates) feature, but actually enforcing the schema unlike OpenAI.\n\nThe project provides a `Generator` class that allows users to easily generate text while ensuring compliance with the provided prompt and JSON schema. By utilizing the `local-llm-function-calling` library, users can conveniently control the output of text generation models. It uses my own quickly sketched `json-schema-enforcer` project as the enforcer.\n\n## Features\n\n- Constrains the generation of Hugging Face text generation models to follow a JSON schema.\n- Provides a mechanism for formulating prompts for function calls, enabling precise data extraction and formatting.\n- Simplifies the text generation process through a user-friendly `Generator` class.\n\n## Installation\n\nTo install the `local-llm-function-calling` library, use the following command:\n\n```shell\npip install local-llm-function-calling\n```\n\n## Usage\n\nHere's a simple example demonstrating how to use `local-llm-function-calling`:\n\n```python\nfrom local_llm_function_calling import Generator\n\n# Define a function and models\nfunctions = [\n {\n \"name\": \"get_current_weather\",\n \"description\": \"Get the current weather in a given location\",\n \"parameters\": {\n \"type\": \"object\",\n \"properties\": {\n \"location\": {\n \"type\": \"string\",\n \"description\": \"The city and state, e.g. San Francisco, CA\",\n \"maxLength\": 20,\n },\n \"unit\": {\"type\": \"string\", \"enum\": [\"celsius\", \"fahrenheit\"]},\n },\n \"required\": [\"location\"],\n },\n }\n]\n\n# Initialize the generator with the Hugging Face model and our functions\ngenerator = Generator.hf(functions, \"gpt2\")\n\n# Generate text using a prompt\nfunction_call = generator.generate(\"What is the weather like today in Brooklyn?\")\nprint(function_call)\n```\n\n## Custom constraints\n\nYou don't have to use my prompting methods; you can craft your own prompts and your own constraints, and still benefit from the constrained generation:\n\n```python\nfrom local_llm_function_calling import Constrainer\nfrom local_llm_function_calling.model.huggingface import HuggingfaceModel\n\n# Define your own constraint\n# (you can also use local_llm_function_calling.JsonSchemaConstraint)\ndef lowercase_sentence_constraint(text: str):\n # Has to return (is_valid, is_complete)\n return [text.islower(), text.endswith(\".\")]\n\n# Create the constrainer\nconstrainer = Constrainer(HuggingfaceModel(\"gpt2\"))\n\n# Generate your text\ngenerated = constrainer.generate(\"Prefix.\\n\", lowercase_sentence_constraint, max_len=10)\n```\n\n## Extending and Customizing\n\nTo extend or customize the prompt structure, you can subclass the `TextPrompter` class. This allows you to modify the prompt generation process according to your specific requirements.\n",
"bugtrack_url": null,
"license": "MIT",
"summary": "A tool for generating function arguments and choosing what function to call with local LLMs",
"version": "0.1.23",
"project_urls": {
"Documentation": "https://local-llm-function-calling.readthedocs.io/",
"Homepage": "https://github.com/rizerphe/local-llm-function-calling"
},
"split_keywords": [
"llm",
"jsonschema",
"huggingface",
"transformers",
"local",
"llama.cpp"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "69b6e3f003e2d26735e0d37476487edb2c78931defe2a7358fed70f13ed2a15f",
"md5": "1f5f45ce1446fb369ccda5f0e262f169",
"sha256": "fdf7b982acfa016637e9a0aecb42efcfa60b21346b8ae5ae38fbb71e23dac3e2"
},
"downloads": -1,
"filename": "local_llm_function_calling-0.1.23-py3-none-any.whl",
"has_sig": false,
"md5_digest": "1f5f45ce1446fb369ccda5f0e262f169",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.11,<4.0",
"size": 18234,
"upload_time": "2024-03-12T12:29:41",
"upload_time_iso_8601": "2024-03-12T12:29:41.700601Z",
"url": "https://files.pythonhosted.org/packages/69/b6/e3f003e2d26735e0d37476487edb2c78931defe2a7358fed70f13ed2a15f/local_llm_function_calling-0.1.23-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "7cf4ee9a8c9cc1cbb85ac71c5dfffe933f8b23a0f925c325c5e156227e5d3eea",
"md5": "5c89f271a0ff4e33fb00471477d704cb",
"sha256": "c05fc5bad533fb671ee5c7c19052214428b6c2857dd95605d2751eaeaf235674"
},
"downloads": -1,
"filename": "local_llm_function_calling-0.1.23.tar.gz",
"has_sig": false,
"md5_digest": "5c89f271a0ff4e33fb00471477d704cb",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.11,<4.0",
"size": 14014,
"upload_time": "2024-03-12T12:29:43",
"upload_time_iso_8601": "2024-03-12T12:29:43.933078Z",
"url": "https://files.pythonhosted.org/packages/7c/f4/ee9a8c9cc1cbb85ac71c5dfffe933f8b23a0f925c325c5e156227e5d3eea/local_llm_function_calling-0.1.23.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2024-03-12 12:29:43",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "rizerphe",
"github_project": "local-llm-function-calling",
"travis_ci": false,
"coveralls": false,
"github_actions": false,
"lcname": "local-llm-function-calling"
}