# PyTorch tutorial codes for course EL-7143 Advanced Machine Learning, NYU, Spring 2018
# train.py: trainig convolutional or linear neural networks for MNIST classification
import time, datetime
from Pipeline.option import args
from Pipeline.run import train
start_time = datetime.datetime.now().replace(microsecond=0)
print('\n---Started training at---', (start_time))
for epoch in range(1, args.epochs + 1):
train(epoch)
current_time = datetime.datetime.now().replace(microsecond=0)
print('Time Interval:', current_time - start_time, '\n')
import time, datetime
import numpy as np
from Pipeline.option import args
from Pipeline.run import train, test
start_time = datetime.datetime.now().replace(microsecond=0)
print('\n---Started training at---', (start_time))
train_acc = np.zeros([args.epochs,2])
test_acc = np.zeros([args.epochs,2])
for epoch in range(1, args.epochs + 1):
tr_corr , tr_per = train(epoch)
ts_corr , ts_per = test()
train_acc[epoch-1,0] = tr_corr
train_acc[epoch-1,1] = tr_per
test_acc[epoch-1,0] = ts_corr
test_acc[epoch-1,1] = ts_per
current_time = datetime.datetime.now().replace(microsecond=0)
print('Time Interval:', current_time - start_time, '\n')
if args.aug == 0:
np.save('train_acc_'+args.model+'_bs'+str(args.batch_size)+'.npy',train_acc)
np.save('test_acc_'+args.model+'_bs'+str(args.batch_size)+'.npy',test_acc)
else:
np.save('train_acc_'+args.model+'_bs'+str(args.batch_size)+'_aug'+'.npy',train_acc)
np.save('test_acc_'+args.model+'_bs'+str(args.batch_size)+'_aug'+'.npy',test_acc)