search

fpadula / visualcollisionarm

End-to-End Visual Obstacle Avoidance for a Robotic Manipulator usingDeep Reinforcement Learning
10 stars 4 forks source link

Ask question #1

Closed Zhousucheng4811 closed 1 year ago

Zhousucheng4811 commented 1 year ago

Hello, after reviewing your project, I would like to ask if TD3VisualAugMixed3. py was used during training. But I didn't see using it in train.py

fpadula commented 1 year ago

Hi @Zhousucheng481, That architecture is used when architecture_type is set to TD3VisualAugMixed3 (as seen here)

The code responsible for evaluating the actor and critic networks (thus setting the model on the code) can be found, respectively, here and here.

I hope this answers your question.