Stable Baselines3 Tutorial: Auto-Stop Training When Best Model Is Found | Demo on BipedalWalker-v3

Ғылым және технология

This tutorial walks thru the code that automatically stop training when the best model is found. We'll demonstrate by training the Gymnasium Bipedal Walker using Soft-Actor Critic.
Want more videos like this? Support me here: www.buymeacoffee.com/johnnycode
Get Started on SB3 part 1: • Get Started with Stabl...
Choosing SB3 Algorithm: • Stable Baselines3 Tuto...
Get the code: github.com/johnnycode8/gym_so...

Пікірлер: 3

  • @DQNLabsAI
    @DQNLabsAI2 ай бұрын

    You are a life saver! Amazing!

  • @IsakAslund
    @IsakAslundАй бұрын

    If min_evals=10k and eval_freq=10k it cannot stop via the "NoModelImp" criteria until 100 million time steps, not 1 million as you mention in the video (10k * 10k = 10M).

  • @johnnycode

    @johnnycode

    Ай бұрын

    You are right! I should have used a calculator😳 Correction on your formula: (10k * 10k = 100M)

Келесі