slim-sql-tool is a 4_K_M quantized GGUF version of slim-sql-1b-v0, providing a small, fast inference implementation, optimized for multi-model concurrent deployment.
slim-sql is part of the SLIM ("Structured Language Instruction Model") series, providing a set of small, specialized decoder-based LLMs, fine-tuned for function-calling.
Note: slim-sql is designed for small, fast, local prototyping and to be effective for 'one-table' lookups - it was not trained or optimized for complex joins and other sophisticated SQL queries.
To pull the model via API:
from huggingface_hub import snapshot_download
snapshot_download("llmware/slim-sql-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
# this one line will download the model and run a series of tests
# includes two sample table schema - go to llmware github repo for end-to-end example
ModelCatalog().tool_test_run("slim-sql-tool", verbose=True)
Slim models can also be orchestrated as part of multi-model, multi-step LLMfx calls:
