RF-CLIP

1.1 verified Fri May 01 auth: no python

A library for performing inference with RF-CLIP (Risk Factor CLIP) models. Integrates CLIP-based vision-language understanding with risk factor analysis. Latest version: 1.1. Release cadence: infrequent.

pip install rf-clip

Common errors

error ModuleNotFoundError: No module named 'rfclip' ↓

cause Incorrect import name; package uses 'rf_clip' (underscore).

fix

Use 'from rf_clip import RFCLIP'

error ImportError: cannot import name 'RFCLIP' from 'rf_clip' ↓

cause Possibly outdated version or wrong class name.

fix

Check that you have the latest version (pip install --upgrade rf-clip) and verify the correct class name in the docs.

Warnings

deprecated The 'predict' method signature may change; check version notes for parameter deprecations. ↓

fix Always refer to the latest API docs for the correct parameters.

gotcha Import must use underscore 'rf_clip' not hyphen 'rf-clip'. ↓

fix Use 'from rf_clip import ...' even though the package is pip installed as rf-clip.

Imports

RFCLIP

wrong

from rfclip import RFCLIP

correct

from rf_clip import RFCLIP

Package uses underscore, not hyphen.

Quickstart

Basic usage: load the model and run inference on an image and text prompt.

from rf_clip import RFCLIP
model = RFCLIP()
result = model.predict(image_path='path/to/image.jpg', text='a risky scenario')
print(result)