my_num = my_num + 1
# if int(topi[0]) == 0:
# break
loss.backward()
encoder_optimizer1.step()
decoder_optimizer1.step()
return loss.data[0]
encoder = DenseNet121().cuda()
attn_decoder1 = AttnDecoderRNN(hidden_size, 112, dropout_p=0.1).cuda()
lr_rate = 0.00009
encoder_optimizer1 = torch.optim.Adam(encoder.parameters(), lr=lr_rate)
decoder_optimizer1 = torch.optim.Adam(attn_decoder1.parameters(), lr=lr_rate)
criterion = nn.CrossEntropyLoss()
loss_all_compare = 100
for epoch in range(1000):
# if using SGD optimizer
# if epoch%8 == 0:
# lr_rate = lr_rate/10
# encoder_optimizer1 = torch.optim.SGD(encoder.parameters(), lr=lr_rate,momentum=0.9)
# decoder_optimizer1 = torch.optim.SGD(attn_decoder1.parameters(), lr=lr_rate,momentum=0.9)
running_loss = 0
whole_loss = 0
torch.load(
'model_69/attn_decoder_lr0.00000_GN_te1_d05_SGD_bs6_mask_conv_bn_b_xavier.pkl'
))
decoder_input_init = torch.LongTensor([111] * batch_size).cuda()
decoder_hidden_init = torch.randn(batch_size, 1, hidden_size).cuda()
nn.init.xavier_uniform_(decoder_hidden_init)
# encoder_optimizer1 = torch.optim.Adam(encoder.parameters(), lr=lr_rate)
# decoder_optimizer1 = torch.optim.Adam(attn_decoder1.parameters(), lr=lr_rate)
for epoch in range(200):
encoder_optimizer1 = torch.optim.SGD(encoder.parameters(),
lr=lr_rate,
momentum=0.9)
decoder_optimizer1 = torch.optim.SGD(attn_decoder1.parameters(),
lr=lr_rate,
momentum=0.9)
# # if using SGD optimizer
# if epoch+1 == 50:
# lr_rate = lr_rate/10
# encoder_optimizer1 = torch.optim.SGD(encoder.parameters(), lr=lr_rate,momentum=0.9)
# decoder_optimizer1 = torch.optim.SGD(attn_decoder1.parameters(), lr=lr_rate,momentum=0.9)
# if epoch+1 == 75:
# lr_rate = lr_rate/10
# encoder_optimizer1 = torch.optim.SGD(encoder.parameters(), lr=lr_rate,momentum=0.9)
# decoder_optimizer1 = torch.optim.SGD(attn_decoder1.parameters(), lr=lr_rate,momentum=0.9)
running_loss = 0
whole_loss = 0
return numpy.array(pil_im.resize(sz))
criterion = nn.NLLLoss()
# encoder.load_state_dict(torch.load('model/encoder_lr0.00001_BN_te1_d05_SGD_bs8_mask_conv_bn_b.pkl'))
# attn_decoder1.load_state_dict(torch.load('model/attn_decoder_lr0.00001_BN_te1_d05_SGD_bs8_mask_conv_bn_b.pkl'))
decoder_input_init = torch.LongTensor([111]*batch_size).cuda()
decoder_hidden_init = torch.randn(batch_size, 1, hidden_size).cuda()
nn.init.xavier_uniform_(decoder_hidden_init)
# encoder_optimizer1 = torch.optim.Adam(encoder.parameters(), lr=lr_rate)
# decoder_optimizer1 = torch.optim.Adam(attn_decoder1.parameters(), lr=lr_rate)
for epoch in range(200):
encoder_optimizer1 = torch.optim.SGD(encoder.parameters(), lr=lr_rate,momentum=0.9)
decoder_optimizer1 = torch.optim.SGD(attn_decoder1.parameters(), lr=lr_rate,momentum=0.9)
# # if using SGD optimizer
# if epoch+1 == 50:
# lr_rate = lr_rate/10
# encoder_optimizer1 = torch.optim.SGD(encoder.parameters(), lr=lr_rate,momentum=0.9)
# decoder_optimizer1 = torch.optim.SGD(attn_decoder1.parameters(), lr=lr_rate,momentum=0.9)
# if epoch+1 == 75:
# lr_rate = lr_rate/10
# encoder_optimizer1 = torch.optim.SGD(encoder.parameters(), lr=lr_rate,momentum=0.9)
# decoder_optimizer1 = torch.optim.SGD(attn_decoder1.parameters(), lr=lr_rate,momentum=0.9)
running_loss=0
whole_loss = 0