search

broadinstitute / AutoTrain

Using RL to solve overfitting in neural networks
0 stars 0 forks source link

Stochastic AutoTrain Environment? #3

Open ctrlnomad opened 3 years ago

ctrlnomad commented 3 years ago

Should we treat the actions as probability distributions or not, do we sample from them or treat as threshold values