README.md

---
license: apache-2.0
---

# Model Card for Model ID

<!-- Provide a quick summary of what the model is/does. -->

**slim-sentiment-tool** is part of the SLIM ("Structured Language Instruction Model") model series, providing a set of small, specialized decoder-based LLMs, fine-tuned for function-calling.  

slim-sentiment-tool is a 4_K_M quantized GGUF version of slim-sentiment-tool, providing a fast, small inference implementation.

Load in your favorite GGUF inference engine, or try with llmware as follows:

    from llmware.models import ModelCatalog

    sentiment_tool = ModelCatalog().load_model("llmware/slim-sentiment-tool")
    response = sentiment_tool.function_call(text_sample, params=["sentiment"], function="classify")

Slim models can also be loaded even more simply as part of LLMfx calls:

    from llmware.agents import LLMfx

    llm_fx = LLMfx()
    llm_fx.load_tool("sentiment")
    response = llm_fx.sentiment(text)


### Model Description

<!-- Provide a longer summary of what this model is. -->

- **Developed by:** llmware
- **Model type:** GGUF 
- **Language(s) (NLP):** English
- **License:** Apache 2.0
- **Quantized from model:** llmware/slim-sentiment (finetuned tiny llama)

## Uses

<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->

The intended use of SLIM models is to re-imagine traditional 'hard-coded' classifiers through the use of function calls.  

Example:

    text = "The stock market declined yesterday as investors worried increasingly about the slowing economy."   

    model generation - {"sentiment": ["negative"]}

    keys = "sentiment"

All of the SLIM models use a novel prompt instruction structured as follows:

    "<human> " + text + "<classify> " + keys + "</classify>" + "/n<bot>: "


## Model Card Contact

Darren Oberst & llmware team