v1.0.2

genderfluid-tiny v1.0.2

Changes

• Lazy imports: --help and --version now 0.1s instead of 2.7s
• Multi-word names without quotes: genderfluid predict Elva Retta
• GitHub Pages landing page

Install

pip install genderfluid-tiny

v1.0.1

genderfluid-tiny v1.0.1

License change: Polyform Noncommercial License (free for personal/educational, paid for commercial).

Changes

• Polyform Noncommercial License 1.0.0 (was MIT)
• Added COMMERCIAL_LICENSE.md with licensing guidance
• Added CONTRIBUTING.md, CODE_OF_CONDUCT.md, SECURITY.md
• Added GitHub issue and PR templates
• Fixed pyproject.toml metadata (SPDX license, project URLs)
• Removed unused pyyaml dependency

Install

pip install genderfluid-tiny

License

Free for personal, educational, and noncommercial use.
Commercial use requires a license. See COMMERCIAL_LICENSE.md.

v1.0.0

genderfluid-tiny v1.0.0

Tiny offline classifier that estimates statistical associations between names and gendered naming conventions.

What it does

Given a name, estimates how strongly it aligns with feminine or masculine naming patterns in the training data. Outputs girl-associated, boy-associated, or uncertain.

Model

• Architecture: character n-gram features + logistic regression
• Size: 49 KB (0.05 MB)
• Trained on 102,927 real names (SSA 1880-2020 + Census 2020)
• Test accuracy: 68.9% | Macro F1: 0.629
• CPU only, no GPU, no internet required

Install

pip install genderfluid-tiny

Usage

from genderfluid import classify_name, is_girl_name

classify_name("Emma")    # "girl-associated"
is_girl_name("James")    # False

genderfluid predict "Elva Retta"
genderfluid predict --compare "Emma" "James" "Alex"
genderfluid stats

Files

• genderfluid_tiny-1.0.0-py3-none-any.whl (63 KB)
• genderfluid_tiny-1.0.0.tar.gz (68 KB)