AI and brand consistency

Member lesson

Training a model on your own product

Lesson 3 of 10 in AI and brand consistency.

How to Train a LoRA + Use It in Weavy (Step-by-Step) · Figma Weave
Lesson

A trained adapter is the difference between asking for a bottle like yours and asking for yours. Twenty to fifty photographs and a short run on a hosted trainer such as Replicate's, and the model can place your actual product in scenes you never shot. How the training itself works, from the LoRA to the trigger word and the zip file you upload, is in Training, fine-tuning and adapters. This lesson is about the photographs.

From above, a small bottle sits on a round turntable. Around it, twelve instant photos form a ring, each showing the bottle from a different angle and on a different surface, such as tiles, sand, linen, slate and grass.
Every angle, many grounds. Vary the backgrounds so the model learns the bottle, not the room.
Member

The rest of this is for members.

You have read the opening of this lesson. The rest of it comes with a membership.

One membership covers all four courses and Brands.

  • All lessons in the four Learn courses
  • All brand breakdowns in the Brands section
  • All quizzes, flashcards and crosswords
  • An AI teacher on every lesson, for up to 60 questions a day
  • All course material published during the membership

Price

Membership is from $12 a month on the Yearly plan, which is charged at $144 once a year. Lifetime access is $59, paid once.Membership is from ₹200 a month on the Yearly plan, which is charged at ₹2,400 + GST once a year. Lifetime access is ₹3,999 + GST, paid once.

Check what stuck

Members only

You can't take this check yet.

Members answer these 3 questions and keep the ticket for this lesson. Every lesson has one.

  1. Question 1, locked

    Which of these words means this? “The images you train on. Where nearly all the quality comes from.”

    • Quality gate
    • Regularisation
    • Dataset
    • Licensed typeface
  2. Question 2, locked

    Which of these words means this? “Every angle, every distance, every light. What stops a model learning one photograph.”

    • Coverage
    • Reference lock
    • Dataset
    • Measured, not judged
  3. Question 3, locked

    Which of these words means this? “Extra generic images that stop the model forgetting everything else.”

    • Rule with a number
    • Colour cast
    • Layout rules
    • Regularisation

Elsewhere in the library

Hear about new courses.

Leave an email and we will tell you when something new lands. We do not write often.