Is deep-learning-models open source?

Yes — fchollet/deep-learning-models is open source, released under the MIT license.

What language is deep-learning-models written in?

fchollet/deep-learning-models is primarily written in Python.

How popular is deep-learning-models?

fchollet/deep-learning-models has 7.3k stars on GitHub.

Where can I find deep-learning-models?

fchollet/deep-learning-models is on GitHub at https://github.com/fchollet/deep-learning-models.

← all repositories

fchollet/deep-learning-models

Keras Image Models with a Tombstone README

A deprecated collection of ready-to-use Keras models—VGG, ResNet50, Inception v3, and a music-tagging CRNN—with pre-trained weights that download automatically on first use.

★7.3k stars Python ML Frameworks Image · Video · Audio

View on GitHub ↗

Not currently ranked — collecting fresh signals.

star history

What it does

This repository contains standalone Keras implementations of VGG16, VGG19, ResNet50, and Inception v3, plus a CRNN for music tagging. Instantiating a model can automatically pull pre-trained weights from the web and cache them in ~/.keras/models/. The code also negotiates TensorFlow versus Theano dimension ordering by reading your Keras config file.

The interesting bit

The README is essentially a tombstone: the project is deprecated, pull requests are ignored, and issues go unread. It survives as a read-only artifact now that these same architectures live in the keras.applications module.

Key highlights

Pre-trained weights available for ImageNet (image models) and the MSD dataset (music tagging).
Weight files are downloaded on demand and cached locally.
Handles both TensorFlow and Theano image_dim_ordering conventions.
Bundles imagenet_utils for input preprocessing and prediction decoding.
The music tagging CRNN is ported from an external project rather than trained here.

Caveats

Completely deprecated. The maintainers explicitly redirect all users to keras.applications and state that PRs and issues are not monitored.
Weight provenance is mixed: VGG and ResNet50 weights are ported from external academic releases, while the Inception v3 weights were trained in-house.

Verdict

A useful stop if you are debugging legacy Keras pipelines or studying how these classic architectures were first packaged for the framework. Otherwise, head straight to keras.applications.

Frequently asked

What is fchollet/deep-learning-models?: A deprecated collection of ready-to-use Keras models—VGG, ResNet50, Inception v3, and a music-tagging CRNN—with pre-trained weights that download automatically on first use.
Is deep-learning-models open source?: Yes — fchollet/deep-learning-models is open source, released under the MIT license.
What language is deep-learning-models written in?: fchollet/deep-learning-models is primarily written in Python.
How popular is deep-learning-models?: fchollet/deep-learning-models has 7.3k stars on GitHub.
Where can I find deep-learning-models?: fchollet/deep-learning-models is on GitHub at https://github.com/fchollet/deep-learning-models.