README.md

---
license: other
license_name: lfm-open-license-v1.0
license_link: https://huggingface.co/LiquidAI/LFM2.5-Encoder-350M-Spellchecker/blob/main/LICENSE
library_name: minima-lfm
base_model: LiquidAI/LFM2.5-Encoder-350M-Spellchecker
pipeline_tag: token-classification
tags:
  - spellcheck
  - gec
  - ternary
  - encoder
---

# Minima Spellcheck

A strict packed W1.58A8 adaptation of
[LiquidAI/LFM2.5-Encoder-350M-Spellchecker](https://huggingface.co/LiquidAI/LFM2.5-Encoder-350M-Spellchecker),
built with [SSHDotCodes/minima](https://github.com/SSHDotCodes/minima).

- Logical matrix values: `{-1, 0, +1}` (1.585 bits)
- Physical artifact format: I2_S, four trits per byte
- Group size 128, **no recovery adapters**
- Full encoder context: 8,192 tokens
- Packed weight file: **90.7 MB** (95,128,472 bytes)
- CPU demo: [Minima Spellcheck](https://huggingface.co/spaces/ProCreations/minima-spellcheck)

## Use

```bash
pip install "minima-lfm @ git+https://github.com/SSHDotCodes/minima.git"
```

```python
from minima import MinimaModel

model = MinimaModel.from_pretrained("ProCreations/minima-spellcheck", device="cpu")
print(model.correct(
    ["I has went to the stor yesterday ."],
    max_iter=4,
    min_error_prob=0.0,
    rerank=False,
))
```

## Validation

Teacher-agreement diagnostics against the LiquidAI tagger with the optional
dense reranker disabled on both sides:

| Diagnostic | Result |
| --- | --- |
| Tag top-1 agreement within teacher candidates | 99.44% |
| Error-detection top-1 agreement | 99.49% |
| Exact correction agreement, held-out + reference examples | 38.2% (68 texts) |
| Exact agreement on the published-style examples | 25% |

| Input | Minima |
| --- | --- |
| `She go to school every day .` | `She go to school every day .` |
| `I has went to the stor yesterday .` | `I has went to the store yesterday .` |
| `Their are many reason to study hard .` | `Their are many reasons to study hard .` |
| `That 's a fair point , let 's discuss it tomorrow .` | `That 's a fair point , let 's discuss it tomorrow .` |

The upstream optional reranker is disabled because it contains a separate
1.42 GB dense encoder, which would invalidate this model's CPU memory profile.

## License

The weights remain subject to the LFM Open License v1.0 shipped in this
repository. The Minima runtime code is MIT licensed.