Open TaoDFang opened 2 years ago
Pytorch DataLoader tutorial: https://pytorch.org/tutorials/beginner/basics/data_tutorial.html
when build DNN model in Pytorch, built independent block atinit function and assembly them in forward function
pytorch tutorial: https://pytorch.org/tutorials/beginner/basics/quickstart_tutorial.html
Summary of One layer CNN: https://www.coursera.org/learn/convolutional-neural-networks/lecture/nsiuW/one-layer-of-a-convolutional-network
might need to increase histone bin size or decrease seq bin size?
not 30 minutes for each epoch, can reduce training size for debuging reason
test histone mark now but with very small bin size for example 1
change chromosome splitting as well
Here i would like to use similar model as mention in DeepHistone paper Input would be DNA sequence (one hot encode ) and histone marks
their code here: https://github.com/QijinYin/DeepHistone