print(' train with model: %s'%params.model)
if 'Conv' in params.model:
image_size = 84
else:
image_size = 224
n_query = max(1, int(16* params.test_n_way/params.train_n_way))
train_few_shot_params = dict(n_way = params.train_n_way, n_support = params.n_shot)
base_datamgr = SetDataManager(image_size, n_query = n_query, n_eposide = 100, **train_few_shot_params)
test_few_shot_params = dict(n_way = params.test_n_way, n_support = params.n_shot)
val_datamgr = SetDataManager(image_size, n_query = n_query, n_eposide = 100, **test_few_shot_params)
val_loader = val_datamgr.get_data_loader( val_file, aug = False)
val_loader_nd = val_datamgr.get_data_loader( val_file, aug = False)
model = MAML(params, tf_path=params.tf_dir)
model.cuda()
# resume training
start_epoch = params.start_epoch
stop_epoch = params.stop_epoch
if params.resume != '':
resume_file = get_resume_file('%s/checkpoints/%s'%(params.save_dir, params.resume), params.resume_epoch)
if resume_file is not None:
start_epoch = model.resume(resume_file)
print(' resume the training with at {} epoch (model file {})'.format(start_epoch, params.resume))
else:
raise ValueError('No resume file')
# load pre-trained feature encoder
else:
if params.warmup == 'scratch':
pass
# prepare model
print('--- prepare model {} (backbone {}) ---'.format(
params.method, params.model))
if params.method in ['maml', 'maml_approx']:
backbone.ConvBlock.maml = True
backbone.SimpleBlock.maml = True
backbone.BottleneckBlock.maml = True
backbone.ResNet.maml = True
model = MAML(model_dict[params.model],
approx=(params.method == 'maml_approx'),
tf_path=params.tf_dir,
**train_few_shot_params)
else:
raise ValueError('Unknown method')
model = model.cuda()
max_acc = 0
total_it = 0
start_epoch = params.start_epoch
stop_epoch = params.stop_epoch
if params.method not in ['baseline', 'baseline++']:
stop_epoch = params.stop_epoch * model.batch_size
# resume/warmup or not
if params.resume != '':
resume_file = get_resume_file('%s/checkpoints/%s' %
(params.save_dir, params.resume))
if resume_file is not None:
print(' load resume file: {}'.format(params.resume))
tmp = torch.load(resume_file)