File size: 1,977 Bytes
f89b70e 08b4a41 eecb5f8 08b4a41 10c2a9a 08b4a41 fc0df25 08b4a41 782ca10 dd4cb89 08b4a41 dd4cb89 fc0df25 1ce4968 08b4a41 dd4cb89 08b4a41 dd4cb89 10c2a9a fc0df25 08b4a41 eecb5f8 08b4a41 dd4cb89 08b4a41 dd4cb89 fc0df25 08b4a41 fc0df25 08b4a41 1183398 b82a2e5 08b4a41 dd4cb89 08b4a41 dd4cb89 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 | ---
license: apache-2.0
---
# SLIM-SQL-TOOL
<!-- Provide a quick summary of what the model is/does. -->
**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**](https://huggingface.co/llmware/slim-sql-1b-v0) 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.
*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:
from llmware.agents import LLMfx
llm_fx = LLMfx()
llm_fx.load_tool("sql")
response = llm_fx.sql(query, table_schema)
Note: please review [**config.json**](https://huggingface.co/llmware/slim-sql-tool/blob/main/config.json) in the repository for prompt wrapping information, details on the model, and full test set.
Note: two sample 'hello world' csv tables are included - this is fabricated data - any similarity with real people is coincidental.
## Model Card Contact
Darren Oberst & llmware team
[Any questions? Join us on Discord](https://discord.gg/MhZn5Nc39h)
|