Skip to content
Z

zeroshot

Efficient and simple image classification for developers

Free

#2263in Developer Tools
0In stores

Description

Zeroshot is an open source tool that lets developers build image classifiers from text descriptions instead of labelled training data. You d…

Highlights

  • Create image classifiers from text descriptions
  • No data labelling required
  • Python package via pip or conda
  • Runs locally on CPU or GPU, offline
  • Web interface for training custom models

Ratings & Reviews

No ratings yet

Be the first to rate zeroshot.

People also visit

L

Lobe

Train models easily, create custom models in a few steps

Developer Tools
Launch
C

classifier.dev docs

Zero-shot text classification over plain HTTP — no API key, no account

AI Skills
Get
T

tao-finetune-clip

CLIP vision-language model for image-text retrieval, zero-shot classification

AI Skills
Get
M

MachineBox

Easily create, deploy, and analyze ML models for image recognition

Developer Tools
Launch
L

Label Sleuth

Open source no-code system for text annotation and building of text classifiers

Developer Tools
Launch
T

tao-train-image-classification

PyTorch-based TAO image classification

AI Skills
Get

zeroshot FAQ

How do I create a classifier?

Describe your classes in the Zeroshot web interface to train a custom model in a few minutes; no training data is needed.

How do I use it in code?

Install with pip or conda, then create zeroshot.Classifier with your model string or path, build a preprocessing function with zeroshot.create_preprocess_fn(), and call classifier.predict on the preprocessed image.

Does it need an internet connection?

No. Models run locally, so you can use and deploy Zeroshot offline.

Is it free?

Yes. Zeroshot is open source and free to use.

What are the best zeroshot alternatives?

Popular alternatives to zeroshot include Lobe, MachineBox and Label Sleuth. See all zeroshot alternatives

Developer Stuff

Badges