[Hi, currently i am doing some study base on your code. I do study on AutonomousDrivingCookbook
I tried to make the network so it can predict steering, throttle, and brake instead of just steering angle. I make some adjustment on data preparation so in can store image, steering, throttle, brake, and previous state.
I tried to train the network to predict steering, throttle and break. Training steering and throttle smoothly without error, but i got some error when training brake. I also tried using the original code only change steering to brake in training.ipynb and change data preparation to store image, brake, and prev state still got the same error. I dont know what cause that error since the number of train, eval, test, and batch size not change for
Here the screenshot of error i got
](url
)
I have some more question to ask
Is modifying the code so it can predict steering, throttle, and brake are possible task to do ?
If it was possible which one will you recomend me to do ? one is to make 1 network with 3 output or make 3 network with single output each and train separatedly ?
[Hi, currently i am doing some study base on your code. I do study on AutonomousDrivingCookbook I tried to make the network so it can predict steering, throttle, and brake instead of just steering angle. I make some adjustment on data preparation so in can store image, steering, throttle, brake, and previous state. I tried to train the network to predict steering, throttle and break. Training steering and throttle smoothly without error, but i got some error when training brake. I also tried using the original code only change steering to brake in training.ipynb and change data preparation to store image, brake, and prev state still got the same error. I dont know what cause that error since the number of train, eval, test, and batch size not change for
Here the screenshot of error i got ](url )
I have some more question to ask