TensorFlow Hub Pre-Trained Models

Explore and reuse existing models for image classification, text processing, and more.

What is TensorFlow Hub?

TensorFlow Hub is a repository for machine learning models, allowing you to reuse existing models for transfer learning or inference.

Pre-Trained Models

Access thousands of ready-to-use models for vision, NLP, and speech.

Transfer Learning

Modify or use fine-tuned models for your specific use case.

Model Export

Save, load, and share trained models in standard formats.

Featured Models

ResNet-50

State-of-the-art for image classification on large datasets.

View Model

BERT

Pre-trained model for natural language understanding tasks.

View Model

Wav2Vec2

State-of-the-art speech recognition model for audio processing.

View Model

Explore 4000+ Reusable Models

Search and reuse existing models for your machine learning projects. TensorFlow Hub gives you access to a broad range of pre-trained models for image, text, and audio tasks.

How to Use TensorFlow Hub

1. Find a Model

Search our repository by task, accuracy, or framework.

1000+ Image Models

2. Download

Get model files in SavedModel or Keras format.


                            import tensorflow_hub as hub

                            model = hub.load("https://tfhub.dev/some-model/1")

3. Deploy

Use for inference or customize for retraining.

curl -O https://tfhub.dev/my-model.tar.gz
tar -xzvf my-model.tar.gz