training set

The share of the available examples that a learning algorithm is allowed to learn from: whatever the model ends up encoding in its parametersparameter, n.: One of the adjustable numbers inside an AI model that are tuned during training. Large language models have billions of them, and together they store what the model has learned., whether a network's weightsweights, n.: The learned numbers inside a neural network that set how strongly each part influences the next. The finished set of weights is, in effect, the trained model itself. or a decision treedecision tree, n.: A model that makes a prediction by asking a series of yes-or-no questions about the input, like a flowchart, with each answer leading to the next question until it reaches a final answer.'s split points, is extracted from these examples. Conventionally it is one of three disjoint parts of a labeled dataset, alongside a validation setvalidation set, n.: A set of examples kept out of training and used during development to compare model versions and settings, like a practice exam taken before the real one. for model selectionmodel selection, n.: Choosing, from several candidate models or settings, the one most likely to work best on new data, rather than the one that merely fits the existing data best. and hyperparameter tuninghyperparameter tuning, n.: Trying out different settings for a model's training, such as learning speed or model size, and keeping the combination that performs best on data held back for checking. and a test settest set, n.: A set of examples kept completely apart from training and used at the end to measure how well a model works on data it has never seen, like a sealed final exam. reserved for the final, unbiased performance estimate; 60–80% commonly goes to training, though very large datasets may reserve only a small fraction for evaluation. Performance measured on the training set itself (training error) is optimistically biased, because a model with enough capacity can memorize it, and a large gap between training and validation performance signals overfittingoverfitting, n.: When a model learns its training examples too well — quirks and noise included — so it does great on data it has seen but poorly on anything new, like a student who memorizes answers without understanding.. The training set should resemble the data the model will meet in use; differences between the two are called distribution shiftdistribution shift, n.: When the data an AI system meets in real use differs from the data it was trained on — new users, places, times or habits — often making it perform worse than tests suggested.. For large language modelslarge language model, n.: A computer program trained on enormous amounts of text to predict the words that come next. At vast scale, that simple skill lets it chat, answer questions, translate, summarize and write code. the corresponding notion is the pretraining corpus or training datatraining data, n.: The examples an AI system learns from. Whatever patterns, skills, gaps and biases are in the training data tend to show up in the finished system., which is too vast to be split in the classical way.

We held out 20% of the records for validation and testing and used the remaining 80% as the training set.