nn_course
MODELS HERE --> https://drive.google.com/drive/folders/1-8cUiut-AfxBJ02uHmxSznFxqGYV98wL?usp=sharing
model-res
dir including 2 models: first from finished trainig workflow named best.h5
, second in subdir best-checkpoint
is the best by val-accuracy from checkpoint during workflow.
Dataset and Model implemented
Train method:
#LBL1 shuffle training set and split it into train and validation sets (92%/8%)
#LBL2 added checkpoints - saving best by val_accuracy model
#LBL3 trainig process info
#LBL4 showing history