idx, args.max_iters, batch_time=batch_time, reco_loss=reco_loss_value, iter_time=iter_time, loss=loss_value)) logger.info("remain_time: {}".format(remain_time)) # if loss < compare_loss: # print('save the lower loss iter, loss:',loss) # compare_loss = loss # torch.save(net.module.state_dict(), # '/data/CRAFT-pytorch/real_weights/lower_loss.pth' if index % args.eval_iter == 0 and index != 0: print('Saving state, index:', index) torch.save( net.module.state_dict(), './checkpoint/{}/synweights_'.format(args.exp_name) + repr(index) + '.pth') test('./checkpoint/{}/synweights_'.format(args.exp_name) + repr(index) + '.pth', args=args, result_folder='./checkpoint/{}/result/'.format( args.exp_name)) #test('/data/CRAFT-pytorch/craft_mlt_25k.pth') res_dict = getresult('./checkpoint/{}/result/'.format( args.exp_name)) logger.info(res_dict['method'])
loss = criterion(gh_label, gah_label, out1, out2, mask) loss.backward() optimizer.step() loss_value += loss.item() if index % 2 == 0 and index > 0: et = time.time() print( 'epoch {}:({}/{}) batch || training time for 2 batch {} || training loss {} ||' .format(epoch, index, len(train_loader), et - st, loss_value / 2)) loss_time = 0 loss_value = 0 st = time.time() # if loss < compare_loss: # print('save the lower loss iter, loss:',loss) # compare_loss = loss # torch.save(net.module.state_dict(), # '/data/CRAFT-pytorch/real_weights/lower_loss.pth' if index % 5000 == 0 and index != 0: print('Saving state, index:', index) torch.save( net.module.state_dict(), '/data/CRAFT-pytorch/synweights/synweights_' + repr(index) + '.pth') test('/data/CRAFT-pytorch/synweights/synweights_' + repr(index) + '.pth') #test('/data/CRAFT-pytorch/craft_mlt_25k.pth') getresult()
parser.add_argument('--cuda', default=True, type=str2bool, help='Use cuda to train model') parser.add_argument('--canvas_size', default=1920, type=int, help='image size for inference') parser.add_argument('--mag_ratio', default=2, type=float, help='image magnification ratio') parser.add_argument('--poly', default=False, action='store_true', help='enable polygon type') parser.add_argument('--show_time', default=False, action='store_true', help='show processing time') parser.add_argument('--test_folder', default='/data/', type=str, help='folder path to input images') args = parser.parse_args() test(args.trained_model, result_folder="./result/", args=args) res_dict = getresult('./result/') print(res_dict['method'])