Metadata-Version: 2.4
Name: sovos-families
Version: 0.1.0
Summary: Family cells (left/right × small/big) + GRPO reward functions for sovereign training
Author: Nicholas Templeman
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
Requires-Python: >=3.10
Description-Content-Type: text/markdown

# sovos-families

The 4-split cell structure inside each OWEM family, plus the GRPO
reward functions for sovereign training.

Absorbed 2026-08-11:

| File | Originally at | LOC |
|---|---|---|
| `src/sovos_families/family_cells.py`           | `family_cells.py` (top-level)           | 230 |
| `src/sovos_families/sov_reward_functions.py`   | `sov_reward_functions.py` (top-level)   | 423 |

## Use

```python
from sovos_families import family_cells, sov_reward_functions
```

Or directly:

```python
from sovos_families.family_cells import main as run_family_cells
from sovos_families.sov_reward_functions import build_reward_fns
```

The original scripts run standalone too:

```bash
python3 SOVOS/packages/sovos-families/src/sovos_families/family_cells.py --help
```

## What's in here (per brief)

- `family_cells.py` — each OWEM family holds a 4-cell brain
  (left/right × small/big). Cells hold the per-family lineage weights.
- `sov_reward_functions.py` — GRPO reward functions (DeepSeek-R1 / HF
  TRL pattern) for: sovereign knowledge, BFT, SIGIL, care floor,
  reasoning (ARC/GSM8K/logic), and calibration.

These are *operational substrate tooling* — not the chain math (that
lives in `sovos-chain` + `sovos-fisher-rao` + `sovos-jspace-hyperbolic`).
They back the training pipeline and the Hive (the absorbed
`sovos-hive` Rust kernel talks to these via the runtime).
