llmware / slim-intent-tool

SLIM-INTENT-TOOL

slim-intent-tool is a 4_K_M quantized GGUF version of slim-intent, providing a small, fast inference implementation, optimized for multi-model concurrent deployment.

slim-intent is part of the SLIM (" S tructured L anguage I nstruction M odel") series, providing a set of small, specialized decoder-based LLMs, fine-tuned for function-calling.

To pull the model via API:

from huggingface_hub import snapshot_download           
snapshot_download("llmware/slim-intent-tool", local_dir="/path/on/your/machine/", local_dir_use_symlinks=False)  

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

from llmware.models import ModelCatalog  

# to load the model and make a basic inference
model = ModelCatalog().load_model("slim-intent-tool")
response = model.function_call(text_sample)  

# this one line will download the model and run a series of tests
ModelCatalog().tool_test_run("slim-intent-tool", verbose=True)  

Slim models can also orchestrated as part of a multi-model, multi-step LLMfx calls:

from llmware.agents import LLMfx

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

Note: please review config.json in the repository for prompt wrapping information, details on the model, and full test set.

Model Card Contact

Darren Oberst & llmware team

Any questions? Join us on Discord