Stochastic gradient descent

As explained in Transformers in Action

Stochastic gradient descent is an iterative optimization method that updates parameters using gradients estimated from individual examples or randomly selected small batches.

Stochastic gradient descent (SGD) iteratively adjusts model weights and biases to reduce loss. Each update uses a gradient estimated from an individual example or a small batch, with the learning rate scaling the step. section 8.1.1

Training behavior depends on settings chosen before training:

• Learning rate sets the update size: a rate that is too high can overshoot the minimum, while one that is too low can slow training or get stuck in a local minimum. section 8.1.1
• Batch size sets how many data points contribute to each gradient, affecting convergence stability and speed. section 8.1.1
• More epochs can improve learning, but training too long can cause overfitting; early stopping can interrupt training when validation improvement stalls. section 8.1.1

Figure 8.1 Gradient descent with lr = 0.1, demonstrating faster convergence

Shows gradient descent with a learning rate of 0.1 and faster convergence. section 8.1.1

Figure 8.2 Gradient descent with lr = 0.01, illustrating stability in convergence

Shows gradient descent with a learning rate of 0.01 and stable convergence. section 8.1.1

How Transformers in Action covers it

How deep it goes

Explains parameter updates and learning-rate, batch-size, and epoch effects on training.

Examples

Compares gradient-descent figures using learning rates of 0.1 and 0.01.

What you can do after

Evaluate how learning rate, batch size, and epochs affect updates, convergence, and overfitting.

How it connects

Connects gradient updates to loss functions, learning rate, batch size, epochs, and early stopping.

liveBookmanning.com

Read Transformers in Action on liveBook Get the book at manning.com

Also explained in

• 

  liveBookmanning.com
  Grokking Machine LearningA gradient-descent variant that updates model parameters using one training data point at a time.
• 

  liveBookmanning.com
  Bayesian Optimization in ActionMini-batch gradient descent is an optimization technique that updates model parameters using loss and gradient estimates computed from small subsets of the training data.
• 

  liveBookmanning.com
  Deep Learning with Python, Third EditionA gradient-descent training method that computes updates from small batches of training examples rather than the full dataset at once.
• 

  liveBookmanning.com
  Machine Learning Algorithms in DepthStochastic gradient descent is an iterative optimization method that updates parameters using gradients estimated from random data subsets.
• 

  liveBookmanning.com
  Math and Architectures of Deep Learning
• 

  liveBookmanning.com
  Natural Language Processing in Action, Second EditionStochastic gradient descent is an iterative optimization technique that updates model parameters using randomized or sampled training examples rather than the entire dataset at once.
• 

  liveBookmanning.com
  Deep Learning with R, Third EditionMini-batch stochastic gradient descent is an optimization method that updates model parameters using gradients computed from randomly sampled small batches of training data.
• 

  liveBookmanning.com
  Machine Learning for Tabular Data: XGBoost, Deep Learning, and AIMini-batch gradient descent is an optimization algorithm that updates model parameters using gradients computed from small batches of examples.
• 

  liveBookmanning.com
  How Large Language Models WorkStochastic gradient descent is a gradient-descent technique that estimates parameter updates using a small random subset of training data.
• 

  liveBookmanning.com
  Deep Learning with PyTorch, Second EditionAn optimization method that updates parameters using stochastic gradient descent.
• 

  liveBookmanning.com
  Sutskever's List

Compare how 12 books explain stochastic gradient descent →

Related concepts

• gradient descentkind of · Gradient descent is an iterative optimization method that updates parameters opposite the gradient of an objective function to reduce its value.
• adamincludes · Adam is an adaptive stochastic-gradient optimization algorithm that updates model parameters using moving estimates of the first and second moments of gradients.
• batchinguses · Batching is the grouping of multiple data items, requests, or operations so they can be processed together as a unit.
• learning rateconfigured by · A learning rate is a step-size parameter that controls the magnitude of updates made by an iterative learning or optimization process.
• logistic regressionused by · Logistic regression is a statistical classification model that estimates categorical outcome probabilities by applying a logistic or softmax function to a linear combination of input features.

This explainer was generated by Manning's knowledge graph from the text of Transformers in Action. The section links point to the parts of the book it draws on.