star-ga/mind-nerve-phase1

mind-nerve — Phase 1 (v1.1-oss)

Intent-classification preselector for agent runtimes.

A small, fast classifier that sits between a user request and the host runtime. It reads the request, decides which subset of available tools/skills/agents is relevant, and hands the host a short list — so the downstream LLM never sees the full library in its system prompt.

Result: library size decouples from token cost. Hosting 4,400 skills costs the same prompt budget as hosting 44, because only the top-K are ever loaded per turn.

Usage

pip install mind-nerve

from mind_nerve import route
result = route("git status", top_k=5)
for r in result.routes:
    print(r.score, r.name, r.kind)

The first call auto-downloads this checkpoint into ~/.local/share/mind-nerve/runtime/. To pre-seed or use a custom location, set MIND_NERVE_RUNTIME_DIR=/path/to/your/runtime/.

Model

• Base model: BAAI/bge-small-en-v1.5 (fine-tuned)
• Loss: MultipleNegativesRankingLoss
• Training: 3 epochs, batch 32, lr 2e-5, max_len 256, seed 1337
• Hardware: single CUDA GPU, 119.5 s wall-clock
• Training date: 2026-05-16T04:00:53Z

Catalog

• Version: v1.1-oss (public-clean — no STARGA-private content in the training corpus)
• Candidate pool: 11,922 routing candidates (skills + tools + agents)
• Corpus hash: 1cd130fa98255241b93aaa2fe6a8086bcbf6fc0627c904008cf48ba9f233536d
• Tokenizer hash: cc2a5502d0fa683c98d59da77af1e4ef3a3812e7e2f345c1d8d7a90bed99d817
• Model hash: 83d4d390469bc1bc6a6cac3b9ab8448dcfcd9ac2ba1ab9fce9348c64012681a6

Metrics (held-out eval, 1,193 pairs)

	Baseline (BGE off-the-shelf)	After Phase 1 fine-tune	Δ
Top-1	0.7527	0.8449	+0.092
Top-5	0.9296	0.9606	+0.031
Top-10	0.9489	0.9707	+0.022

What's in this repo

• checkpoint/ — sentence-transformers checkpoint (model.safetensors + tokenizer + config)
• manifest.json — full provenance (corpus_hash, model_hash, training config, metrics)
• route_table.npy — precomputed catalog embeddings (11,922 × 384, float32)
• route_table.jsonl — catalog metadata (one JSON object per row of route_table.npy)

Status

Phase 1, public alpha. Inference runs on PyTorch via the fine-tuned BGE encoder. Phase 2 (target Q3 2027) replaces the PyTorch path with a native MIND Q16.16 inference loop and adds:

• Cross-architecture bit-identity gate (x86 CPU vs CUDA)
• p95 ≤ 30 ms latency budget on 4-core CPU

Phase 2 is gated on mindc 0.2.6 (pub fn → C symbol export) and 0.3.0 (cdylib emit).

License

This model card and the weights it points at are released under Apache-2.0.

The PyPI wheel mind-nerve bundles a FORTRESS-protected libmindnerve.so whose source remains private (STARGA Commercial). The wheel is Apache-2.0; the bundled binary is the protected runtime layer that activates in Phase 2. The Phase 1 inference path published here does not depend on the protected binary.

For commercial deployments needing per-customer FORTRESS-locked builds of the runtime layer, contact license@star.ga.

Citation

@software{mind_nerve_2026,
  author  = {STARGA, Inc.},
  title   = {mind-nerve: Intent-classification preselector for agent runtimes},
  year    = {2026},
  url     = {https://github.com/star-ga/mind-nerve},
  version = {0.1.0-alpha.3},
}