Install GNO

← All features

Fine-Tuned Models

Measured query expansion with a tuned local model

GNO's slim-tuned preset uses a published fine-tuned model only for retrieval-time query expansion. It does not train on your documents and is separate from the model used for standalone answers.

Install GNORead the docs
Use cases
Users who want the shipped tuned expansion path
Researchers comparing query-expansion quality on a fixed fixture
Contributors developing and publishing retrieval models

What it gives you

  • Fine-tuned expansion role, separate from embeddings and answer generation
  • Published portable GGUF artifact on Hugging Face
  • nDCG@10 0.925 and Ask Recall@5 0.875 on the promoted retrieval fixture
  • Benchmark evidence and release manifest published with the model
  • No end-user training step or training on your indexed corpus

Try it yourself

Representative commands and entry points. Full reference lives in the documentation.

gno models use slim-tuned
gno models pull --expand
gno query 'topic' --thorough

Keep reading

Related features and docs.

Benchmarks

/features/benchmarks

Open

Local LLM answers

/features/local-llm

Open

Fine-tuned models reference

/docs/fine-tuned-models

Open

All features

Back to the feature index

Browse all