textcurator-llm-py

A Python package for processing user-provided texts, such as headlines or short descriptions, and generating structured summaries or categorizations using language models. Designed to facilitate content curation, news aggregation, and event highlighting by producing consistent, formatted outputs with key information extracted automatically.

Features

• Accepts various text inputs and outputs structured summaries.
• Uses the ChatLLM7 model from the langchain_llm7 package by default.
• Supports custom language model instances for increased flexibility.
• Implements regex-based pattern matching to extract data reliably.
• Suitable for applications like newsletters, databases, and alert systems.

Installation

Install the package via pip:

pip install textcurator_llm_py

Usage

Import the main function and invoke it with your input text. You can specify your preferred LLM instance or rely on the default ChatLLM7. If not provided, the package will use the API key from the environment variable LLM7_API_KEY.

from textcurator_llm_py import textcurator_llm_py

results = textcurator_llm_py(
    user_input="City's Best Winter Show Is in Its Pitch-Dark Skies",
    api_key="your_api_key_here"  # optional, if not set in environment
)
print(results)

Using a custom language model

You can pass your own LLM instance, such as ChatOpenAI, ChatAnthropic, or others, to tailor the processing:

from langchain_openai import ChatOpenAI
from textcurator_llm_py import textcurator_llm_py

llm = ChatOpenAI()
results = textcurator_llm_py(
    user_input="Example headline about an upcoming event.",
    llm=llm
)
print(results)

Similarly, support exists for other LLMs:

from langchain_anthropic import ChatAnthropic

llm = ChatAnthropic()
results = textcurator_llm_py(
    user_input="News about recent developments.",
    llm=llm
)
print(results)

from langchain_google_genai import ChatGoogleGenerativeAI

llm = ChatGoogleGenerativeAI()
results = textcurator_llm_py(
    user_input="Short description of an event.",
    llm=llm
)
print(results)

Configuration

The function optionally accepts an api_key parameter for the ChatLLM7 model. If not provided, it defaults to the environment variable LLM7_API_KEY. For higher rate limits, obtain an API key at https://token.llm7.io/ and set it accordingly.

Limitations

• The package relies on regex patterns defined within the source code (pattern). Ensure these patterns are suitable for your input data.
• The default ChatLLM7 model is suitable for most use cases; however, users can provide custom LLMs for broader compatibility.

Issues and Support

For issues, please visit the GitHub repository: https://github.com/chigwell/textcurator-llm-py

Author

Eugene Evstafev
Email: hi@euegne.plus
GitHub: chigwell