Metadata-Version: 2.4
Name: proofread-eng-scifi
Version: 0.1.0
Summary: An artisanal language model for proofreading prose in the style of classical English science fiction.
Author-email: Brandon Rohrer <brohrer@gmail.com>
License-Expression: MIT
License-File: LICENSE
Keywords: proofread,spell check,text processing
Classifier: Development Status :: 2 - Pre-Alpha
Classifier: Intended Audience :: Developers
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Classifier: Programming Language :: Python :: 3.12
Classifier: Programming Language :: Python :: 3.13
Classifier: Programming Language :: Python :: 3.14
Classifier: Programming Language :: Python :: 3.15
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
Classifier: Topic :: Software Development :: Documentation
Classifier: Topic :: Text Processing :: General
Requires-Python: >=3.12
Requires-Dist: numpy>=2.4.4
Requires-Dist: sentencepiece>=0.2.1
Description-Content-Type: text/markdown

# `proofread-eng-scify`, an ALM

An artisanal language model (ALM) for proofreading English prose in the
style of old-old-school science fiction, like Jules Verne, Lewis Carroll, and
L. Frank Baum. 

-----

## For someone who wants to use the proofreader

### Installation

### Using the proofreader

##

-----

## For someone who wants to modify the proofreader

### Testing

From the top level project directory, run

```
uv run pytest
```

### Running evaluations

`evals/`

### How the code is organized

**Model development code** sits in the top level `model_dev/` directory. Once
models have been created and tested and are ready to be used in the
proofreader, they are moved to the `src/proofread_eng_scify/models/`
directory. Versions are indicated with a two digit squential counter
appended to the model name, such as `tokenizer_13`.


**Tests** are scattered throughout the code, sitting close to the code they are
meant to test. They are primarily unit tests. For now, integration and
end-to-end testing gets covered by evaluations.

**Evaluations** are in `evals/`

