Skip to main content
. 2021 Jan 25;7:e344. doi: 10.7717/peerj-cs.344

Figure 6. Training state performance.

Figure 6

Based on this figure, the best validation performance is achieved at epoch 26 where the gradient is 0.032875.