Information Density: TensorFlow – Signal Evidence & AI Readability

Stay organized with collections

Save and categorize content based on your preferences.

An end-to-end platform for machine learning

Install TensorFlow

[H2]
Get started with TensorFlow

TensorFlow makes it easy to create ML models that can run in any environment.
Learn how to use the intuitive APIs through interactive code samples.

View tutorials

import tensorflow as tf
mnist = tf.keras.datasets.mnist
(x_train, y_train),(x_test, y_test) = mnist.load_data()
x_train, x_test = x_train / 255.0, x_test / 255.0
model = tf.keras.models.Sequential([
tf.keras.layers.Flatten(input_shape=(28, 28)),
tf.keras.layers.Dense(128, activation='relu'),
tf.keras.layers.Dropout(0.2),
tf.keras.layers.Dense(10, activation='softmax')
])
model.compile(optimizer='adam',
loss='sparse_categorical_crossentropy',
metrics=['accuracy'])
model.fit(x_train, y_train, epochs=5)
model.evaluate(x_test, y_test)

[H2]
Solve real-world problems with ML

Explore examples of how TensorFlow is used to advance research and build AI-powered applications.

TENSORFLOW.JS

Catch up on the latest from the Web AI Summit

Explore the latest advancements in running models client-side with speakers from Chrome, MediaPipe, Intel, Hugging Face, Microsoft, LangChain, and more.

Watch now

close

TensorFlow GNN

Analyze relational data using graph neural networks

GNNs can process complex relationships between objects, making them a powerful technique for traffic forecasting, medical discovery, and more.

Learn about TF GNN

TensorFlow Agents

Build recommendation systems with reinforcement learning

Learn how Spotify uses the TensorFlow ecosystem to design an extendable offline simulator and train RL Agents to generate playlists.

Read the blog

[H2]
What's new in TensorFlow

Read the latest announcements from the TensorFlow team and community.

[H2]
Explore the ecosystem

Discover production-tested tools to accelerate modeling, deployment, and other workflows.

Library

[H3]
TensorFlow.js
Train and run models directly in the browser using JavaScript or Node.js.

Library

[H3]
LiteRT
Deploy ML on mobile and edge devices such as Android, iOS, Raspberry Pi, and Edge TPU.

API

[H3]
tf.data
Preprocess data and create input pipelines for ML models.

Library

[H3]
TFX
Create production ML pipelines and implement MLOps best practices.

API

[H3]
tf.keras
Create ML models with TensorFlow's high-level API.

Resource

[H3]
Kaggle Models
Find pre-trained models ready for fine-tuning and deployment.

Resource

[H3]
TensorFlow Datasets
Browse the collection of standard datasets for initial training and validation.

Tool

[H3]
TensorBoard
Visualize and track development of ML models.

[H3]
ML models & datasets
Pretrained models and ready-to-use datasets for image, text, audio, and video use cases.

[H3]
Libraries & extensions
Packages for domain-specific applications and APIs for languages other than Python.

[H3]
Developer tools
Tools to evaluate models, optimize performance, and productionize ML workflows.

Join the community

Collaborate, find support, and share your projects by joining interest groups or attending developer events.

Get involved

Learn ML

New to machine learning? Begin with TensorFlow's curated curriculums or browse the resource library of books, online courses, and videos.

Explore resources

[H2]
Stay connected
Learn the latest in machine learning and TensorFlow by following our channels or signing up for the newsletter. View past newsletters in the archive.

[IMG: forum logo]

Forum

[IMG: x logo]

X

[IMG: youtube logo]

YouTube

[IMG: linkedin logo]

Linkedin

[IMG: forum logo]

Forum

[IMG: x logo]

X

[IMG: youtube logo]

YouTube

[IMG: linkedin logo]

Linkedin

[[["Easy to understand","easyToUnderstand","thumb-up"],["Solved my problem","solvedMyProblem","thumb-up"],["Other","otherUp","thumb-up"]],[["Missing the information I need","missingTheInformationINeed","thumb-down"],["Too complicated / too many steps","tooComplicatedTooManySteps","thumb-down"],["Out of date","outOfDate","thumb-down"],["Samples / code issue","samplesCodeIssue","thumb-down"],["Other","otherDown","thumb-down"]],[],[],[]]

Stay organized with collections

Save and categorize content based on your preferences.

[H1]
Models & datasets

Explore repositories and other resources to find available models, modules and datasets created by the TensorFlow community.

Kaggle Models

A comprehensive repository of trained models ready for fine-tuning and deployable anywhere.

Explore Kaggle Models

Model Garden

Machine learning models and examples built with TensorFlow's high-level APIs.

Explore GitHub

TensorFlow.js models

Pre-trained machine learning models ready-to-use in the web browser on the client side, or anywhere that JavaScript can run such as Node.js.

Explore GitHub

[H2]
Datasets

TensorFlow official datasets

A collection of datasets ready to use with TensorFlow.

Learn more

Google research datasets

Explore large-scale datasets released by Google research teams in a wide range of computer science disciplines.

Learn more

Additional dataset resources

Explore other datasets available to use with TensorFlow.

Dataset Search

Google Cloud public datasets

Kaggle datasets

[[["Easy to understand","easyToUnderstand","thumb-up"],["Solved my problem","solvedMyProblem","thumb-up"],["Other","otherUp","thumb-up"]],[["Missing the information I need","missingTheInformationINeed","thumb-down"],["Too complicated / too many steps","tooComplicatedTooManySteps","thumb-down"],["Out of date","outOfDate","thumb-down"],["Samples / code issue","samplesCodeIssue","thumb-down"],["Other","otherDown","thumb-down"]],[],[],[]]

Stay organized with collections

Save and categorize content based on your preferences.

The TensorFlow tutorials are written as Jupyter notebooks and run directly in Google Colab—a hosted notebook environment that requires no setup. At the top of each tutorial, you'll see a Run in Google Colab button. Click the button to open the notebook and run the code yourself.

[H2] For beginners
The best place to start is with the user-friendly Keras sequential API. Build models by plugging together building blocks. After these tutorials, read the Keras guide.

[H3] Beginner quickstart
This "Hello, World!" notebook shows the Keras Sequential API and model.fit.

[H3] Keras basics
This notebook collection demonstrates basic machine learning tasks using Keras.

[H3] Load data
These tutorials use tf.data to load various data formats and build input pipelines.

[H2] For experts
The Keras functional and subclassing APIs provide a define-by-run interface for customization and advanced research. Build your model, then write the forward and backward pass. Create custom layers, activations, and training loops.

[H3] Advanced quickstart
This "Hello, World!" notebook uses the Keras subclassing API and a custom training loop.

[H3] Customization
This notebook collection shows how to build custom layers and training loops in TensorFlow.

[H3] Distributed training
Distribute your model training across multiple GPUs, multiple machines or TPUs.

The Advanced section has many instructive notebooks examples, including Neural machine translation, Transformers, and CycleGAN.

[H2] Video tutorials
Check out these videos for an introduction to machine learning with TensorFlow:

[H3]

TensorFlow ML Zero to Hero

[H3]

Basic Computer Vision with ML

[H2] Libraries and extensions
Explore libraries to build advanced models or methods using TensorFlow, and access domain-specific application packages that extend TensorFlow. This is a sample of the tutorials available for these projects.

chevron_right

[H3] TensorBoard
Get started with TensorBoard Logging training metrics in Keras

chevron_right

[H3] TensorFlow Hub
Object detection Arbitrary style transfer

chevron_right

[H3] Model Optimization
Magnitude-based weight pruning with Keras Post-training quantization

chevron_right

[H3] TensorFlow Federated
Federated learning for image classification Federated learning for text generation

chevron_right

[H3] Neural Structured Learning
Natural graph regularization for document classification Synthetic graph regularization for sentiment classification

chevron_right

[H3] TensorFlow Graphics
Object pose alignment Mesh segmentation

chevron_right

[H3] SIG Addons
Image operations in TensorFlow Addons Normalization layers in TensorFlow Addons.

chevron_right

[H3] TFX
TFX developer tutorial Serve a model with TensorFlow Serving

chevron_right

[H3] Datasets
Using TensorFlow Datasets

chevron_right

[H3] Probability
TensorFlow distributions introduction Probabilistic regression

chevron_right

[H3] XLA
Classifying CIFAR-10 with XLA Use XLA with tf.function

chevron_right

[H3] Decision Forests
Train a decision forest model Use text and NN features with decision forests

chevron_right

[H3] TensorFlow Agents
Train a deep-Q network with TF Agents Reinforcement learning environments

chevron_right

[H3] TensorFlow Ranking
TF-Ranking Keras user guide TF Ranking for sparse features

chevron_right

[H3] Magenta
Generating Piano music with Transformer GANSynth

[H2] TensorFlow updates
Subscribe to the TensorFlow blog, YouTube channel, and Twitter for the latest updates.

Except as otherwise noted, the content of this page is licensed under the Creative Commons Attribution 4.0 License, and code samples are licensed under the Apache 2.0 License. For details, see the Google Developers Site Policies. Java is a registered trademark of Oracle and/or its affiliates.
Last updated 2023-09-19 UTC.

[[["Easy to understand","easyToUnderstand","thumb-up"],["Solved my problem","solvedMyProblem","thumb-up"],["Other","otherUp","thumb-up"]],[["Missing the information I need","missingTheInformationINeed","thumb-down"],["Too complicated / too many steps","tooComplicatedTooManySteps","thumb-down"],["Out of date","outOfDate","thumb-down"],["Samples / code issue","samplesCodeIssue","thumb-down"],["Other","otherDown","thumb-down"]],["Last updated 2023-09-19 UTC."],[],[]]

Stay organized with collections

Save and categorize content based on your preferences.

[H1]
Introduction to TensorFlow

TensorFlow makes it easy for beginners and experts to create machine learning models for desktop, mobile, web, and cloud. See the sections below to get started.

[H4]
TensorFlow

Learn the foundations of TensorFlow with tutorials for beginners and experts to help you create your next machine learning project.

Learn more

[H4]
For Web

Use TensorFlow.js to create new machine learning models and deploy existing models with JavaScript.

Learn more

[H4]
For Mobile & Edge

Run inference with LiteRT on mobile and embedded devices like Android, iOS, Edge TPU, and Raspberry Pi.

Learn more

[H4]
For Production

Deploy a production-ready ML pipeline for training and inference using TFX.

Learn more

[H2]
An end-to-end platform for machine learning

[H3]
Prepare and load data for successful ML outcomes

Data can be the most important factor in the success of your ML endeavors.
TensorFlow offers multiple data tools to help you consolidate, clean and preprocess data at scale:

Standard datasets for initial training and validation

Highly scalable data pipelines for loading data

Preprocessing layers for common input transformations

Tools to validate and transform large datasets

Additionally, responsible AI tools help you uncover and eliminate bias in your data to produce fair, ethical outcomes from your models.

[H4]
Try it in Colab

Load and preprocess an image dataset

Investigate and visualize datasets

[H3]
Build and fine-tune models with the TensorFlow ecosystem

Explore an entire ecosystem built on the Core framework that streamlines model construction, training, and export. TensorFlow supports distributed training, immediate model iteration and easy debugging with Keras, and much more. Tools like Model Analysis and TensorBoard help you track development and improvement through your model’s lifecycle. To help you get started, find collections of pre-trained models at TensorFlow Hub from Google and the community, or implementations of state-of-the art research models in the Model Garden. These libraries of high level components allow you to take powerful models, and fine-tune them on new data or customize them to perform new tasks.

[H4]
Try it in Colab

Train a neural network to classify images

Retrain an image classifier with transfer learning

[H3]
Deploy models on-device, in the browser, on-prem, or in the cloud

TensorFlow provides robust capabilities to deploy your models on any environment - servers, edge devices, browsers, mobile, microcontrollers, CPUs, GPUs, FPGAs. TensorFlow Serving can run ML models at production scale on the most advanced processors in the world, including Google's custom Tensor Processing Units (TPUs). If you need to analyze data close to its source to reduce latency and improve data privacy, the LiteRT framework lets you run models on mobile devices, edge computing devices, and even microcontrollers, and the TensorFlow.js framework lets you run machine learning with just a web browser.

[H4]
Try it in Colab

Serve a model with TensorFlow Serving

[H3]
Implement MLOps for production ML

The TensorFlow platform helps you implement best practices for data automation, model tracking, performance monitoring, and model retraining. Using production-level tools to automate and track model training over the lifetime of a product, service, or business process is critical to success. TFX provides software frameworks and tooling for full MLOps deployments, detecting issues as your data and models evolve over time.

[H4]
Try it in Colab

Create and run a simple TFX pipeline

Track lineage with ML Metadata

[H2]
Looking to expand your ML knowledge?

TensorFlow is easier to use with a basic understanding of machine learning principles and core concepts. Learn and apply fundamental machine learning practices to develop your skills.

Learn ML

Begin with curated curriculums to improve your skills in foundational ML areas.

Learn more

[[["Easy to understand","easyToUnderstand","thumb-up"],["Solved my problem","solvedMyProblem","thumb-up"],["Other","otherUp","thumb-up"]],[["Missing the information I need","missingTheInformationINeed","thumb-down"],["Too complicated / too many steps","tooComplicatedTooManySteps","thumb-down"],["Out of date","outOfDate","thumb-down"],["Samples / code issue","samplesCodeIssue","thumb-down"],["Other","otherDown","thumb-down"]],[],[],[]]