Tensor Flow
TensorFlow is an open-source machine learning framework developed by Google. It enables developers to build, train, and deploy AI and deep learning models. TensorFlow supports applications such as computer vision, NLP, and predictive analytics.
TensorFlow is a powerful open-source framework for machine learning and deep learning.
It provides tools and APIs for designing, training, evaluating, and deploying neural networks.
TensorFlow supports CPU, GPU, and distributed computing for scalable AI applications.
It is widely used for computer vision, natural language processing, recommendation systems, and time-series analysis.
With TensorFlow and Keras, developers can efficiently create and deploy production-ready AI models.
What You'll Learn
Tensors
What Is TensorFlow? Eager Mode and tf.function Explained
Learn what TensorFlow is and how tensors flow through operations. See what happens when one line of…
34 minTensorFlow Setup: Google Colab, Local Install and GPU Check
Set up TensorFlow 2.21 three ways: Google Colab, a local pip install in a virtual environment, or D…
29 minCreating Tensors in TensorFlow: tf.constant, tf.zeros and tf.random
Learn every common way to create a tensor: tf.constant from lists and NumPy, tf.zeros, tf.ones, tf.…
40 minTensor Shape, Rank, dtype and tf.reshape Explained
Understand the three facts every tensor has: shape, rank and dtype. Climb from rank 0 to rank 4 wit…
36 minTensor Indexing, Slicing and tf.gather in TensorFlow
Learn to pick parts of a tensor. Use indexing, negative indexes and start:stop:step slices on 1-D a…
37 minTensor Maths: Element-wise Ops, tf.matmul, Broadcasting and tf.reduce_*
Master the four kinds of maths behind every neural network: element-wise operations, matrix multipl…
46 minVariables and Automatic Gradients
tf.Variable vs tf.Tensor: Mutable Weights in TensorFlow
Learn why models store weights in tf.Variable and not tf.Tensor. Create variables, update them in p…
28 mintf.GradientTape Explained: Your First Gradient in TensorFlow
Understand gradients as slopes and compute your first one with tf.GradientTape in three lines. See …
35 minGradients for Many Variables with tf.GradientTape
Compute gradients for many variables in one backward pass. Pass lists or dicts of variables, match …
32 minPersistent GradientTape and Higher-Order Gradients
Learn why a normal GradientTape works only once and how persistent=True lets you call gradient many…
35 minGradient Descent by Hand in TensorFlow
Build gradient descent yourself with tf.GradientTape and assign_sub. Minimise a simple loss, then t…
38 minGraphs and Speed with tf.function
Eager vs Graph Mode in TensorFlow: How tf.function Speeds Up Code
Learn the difference between eager mode and graph mode in TensorFlow. See how tf.function turns Pyt…
29 min@tf.function and Tracing Explained: Concrete Functions and AutoGraph
Understand what happens when you call a tf.function: signature check, tracing with symbolic tensors…
28 minCommon tf.function Mistakes: Python Side Effects and Retracing
Avoid the five classic tf.function traps: Python side effects that run only once, frozen NumPy rand…
32 mininput_signature and tf.TensorSpec: Control Tracing in TensorFlow
Use tf.TensorSpec to describe the shape and dtype a tensor must have, with None for any size. Add i…
29 minSpeed Test: Eager vs Graph Mode Benchmark in TensorFlow
Measure the real speed-up of tf.function. Learn to benchmark fairly with warm-up, timeit and .numpy…
28 minNeural Network From Scratch
tf.Module Explained: The Base for Layers and Models in TensorFlow
Learn how tf.Module tracks variables and sub-modules for you. Build a module, nest layers inside a …
25 minWrite a Dense Layer From Scratch: Weights, Bias and Initialisation
Build a Dense layer with tf.Module: y = x @ W + b, its shapes and parameter count. See why zero wei…
32 minActivation Functions in TensorFlow: relu, sigmoid, tanh and softmax
Prove that stacked linear layers collapse into one, then meet the activations in tf.nn: relu, leaky…
30 minLoss Functions by Hand: MSE and Cross-Entropy in TensorFlow
Write loss functions yourself: MSE, MAE and Huber for regression, and binary and categorical cross-…
30 minWrite Optimizers From Scratch: SGD, Momentum and Adam
Write SGD, Momentum and Adam as tf.Module classes, with velocity and moment slots, bias correction …
30 minFull Training Loop From Scratch: Batches, Epochs and Accuracy
Put every piece together: a tf.data pipeline, an MLP of your own Dense layers, softmax cross-entrop…
41 minData Pipelines and Good Training
tf.data.Dataset Basics: Load Data From Tensors, CSV and Files
Learn what a tf.data.Dataset is and how to build one from NumPy arrays, dicts, Python generators, C…
31 minTFRecord Files in TensorFlow: Write, Read and Parse Examples
Learn the TFRecord format: pack features into tf.train.Example, write records with TFRecordWriter, …
29 minTrain/Validation Split and Early Stopping by Hand in TensorFlow
Split data into training and validation sets the right way, and avoid the shuffle leak that mixes t…
32 minRegularisation by Hand: L2, Dropout and Batch Norm in TensorFlow
Fight overfitting with three tools written by hand: an L2 weight penalty, tf.nn.dropout and a Batch…
37 mintf.summary and TensorBoard: Log Your Training in TensorFlow
Log training with tf.summary: create a file writer, record scalars, histograms, text and images, an…
31 mintf.train.Checkpoint: Save and Resume Training in TensorFlow
Save your model, optimizer and step counter with tf.train.Checkpoint, and keep the last few saves w…
28 min