koyeongmin / PINet

MIT License
207 stars 56 forks source link

monitoring training and cross-validation loss values #35

Open Mkarami3 opened 4 years ago

Mkarami3 commented 4 years ago

Hello, Many thanks for providing your code for free. I am training the model on my own dataset. I wonder how can I monitor the loss values for training and cross-validation? Mohammad

koyeongmin commented 4 years ago

Because TuSimple dataset has only about 3000 frames for training, the code evaluated the performance about test set every epoch.