help="fold in [0,1,2,3,4]")
(options, args) = parser.parse_args()
fold = options.fold
batch_size = options.batch_size
mu = options.mu
epoch_num = 2000
sample_num_each_epoch = 500
patience = 25
best_epoch = 0
best_val_auc = 0.0
test_times = 20
if options.model_file == "":
minet = MINet()
minet = torch.nn.DataParallel(minet).cuda()
else:
print('not implemented')
learning_rate = options.learning_rate
optimizer = torch.optim.Adam([{
'params': minet.parameters()
}],
lr=learning_rate)
nn_loss_label = torch.nn.CrossEntropyLoss().cuda()
loss = 0
(options, args) = parser.parse_args()
fold = options.fold
batch_size = options.batch_size
mu = options.mu
embed_len = 10
epoch_num = 2000
inst_num = 5
sample_num_each_epoch = 500
patience = 25
best_epoch = 0
best_val_auc = 0.0
test_times = 10
if options.model_file == "":
minet = MINet(embed_len, inst_num)
minet = torch.nn.DataParallel(minet).cuda()
else:
print('not implemented')
learning_rate = options.learning_rate
optimizer = torch.optim.Adam([{
'params': minet.parameters()
}],
lr=learning_rate)
nn_loss_label = torch.nn.CrossEntropyLoss().cuda()
loss = 0