How models learn
← The series

Chapter 10 · Not written yet

The loop

Everything, end to end.

Predict, measure, assign the blame, adjust, and start again. Watching the error curve fall is the first thing anyone does, and reading it properly is a skill.

  • epochs
  • training curve
  • diagnosis

How this chapter will go

Name every stage of the cycle, and read a training curve.

The image it is built onA day in the workshop. Produce, inspect, adjust the machines, start again tomorrow.

  1. Four moves, on one figure

    Predict, measure, hand out the blame, adjust. Each station on the cycle links back to the chapter that built it.

    predictmeasurehand out blameadjustepoch 0step 00.000.251.004.0016.000102030epochscostwhat it predictsfalls, and settles
    this problem stops converging above 0.2082
    at
    predict
    epoch
    0
    steps
    0
    cost
    12.365
    step size
    0.0833

    Four presses of One station is one adjustment. Press Run to let it go on its own.

    The whole of training. Four presses of One station is one adjustment, and the curve beside the ring is the run those adjustments produce. The three presets are three shapes worth knowing.
  2. Watch one run

    Start a small network and let the cost curve fall in real time. This is the first thing anyone does, and it is worth doing slowly once.

  3. Three shapes worth knowing

    A curve that flattens too early, a curve that oscillates, a curve that climbs. Each has a cause, and each is available as a preset so the shape can be produced on purpose.

  4. What an epoch counts

    One epoch is one pass through the whole training set. The word is defined by what it counts, not by translating it.

The misreading to head off

A falling cost is not proof of anything good. It only says the model is fitting what it was shown, which is the subject of the next chapter.

There is no date, no list to join and nothing to subscribe to. The chapter simply appears here once the drawing works.