guiltylemon/gno-expansion-slim-retrieval-v1

GNO Slim Retrieval v1

Fine-tuned query expansion model for GNO hybrid retrieval.

Canonical promoted run: auto-entity-lock-default-mix-lr95 HF repo: guiltylemon/gno-expansion-slim-retrieval-v1 GGUF file: gno-expansion-auto-entity-lock-default-mix-lr95-f16.gguf

What This Model Is For

• retrieval-centric query expansion
• entity / negation preservation
• multilingual retrieval support
• JSON-structured expansion output for GNO

It is not optimized for standalone answer synthesis.

Promotion Decision

• confirmed incumbent artifact: confirmed-incumbent.json
• promotion target gate: promotion-target-check.json
• result: passed

Repeated Benchmark

• promoted median nDCG@10: 0.9250
• previous incumbent median nDCG@10: 0.9190
• promoted median ask Recall@5: 0.8750
• promoted median schema success: 100.0%
• promoted median p95: 4775.99ms

Shipped Slim Baseline Delta

• shipped slim nDCG@10: 0.8983
• promoted median delta vs shipped slim: 0.0267

Install In GNO

models:
  activePreset: slim-tuned
  presets:
    - id: slim-tuned
      name: GNO Slim Retrieval v1
      embed: "hf:gpustack/bge-m3-GGUF/bge-m3-Q4_K_M.gguf"
      rerank: "hf:ggml-org/Qwen3-Reranker-0.6B-Q8_0-GGUF/qwen3-reranker-0.6b-q8_0.gguf"
      gen: "hf:guiltylemon/gno-expansion-slim-retrieval-v1/gno-expansion-auto-entity-lock-default-mix-lr95-f16.gguf"

Published Artifact

• HF model URI: hf:guiltylemon/gno-expansion-slim-retrieval-v1/gno-expansion-auto-entity-lock-default-mix-lr95-f16.gguf
• best checkpoint: 0000500_adapters.safetensors

Source Artifacts

• benchmark summary: benchmark-summary.json
• repeat benchmark: repeat-benchmark.json
• promotion summary: promotion-summary.json
• release manifest: release-manifest.json