# ============================================================================
# LOAD CONFIGURATIONS
with open(args.cfg) as f:
config = yaml.load(f)
config = update(config, args)
test_mode = args.test
val_mode = args.val
if not test_mode and not val_mode:
opt = config['training_opt']
if not os.path.isdir(opt['log_dir']):
os.makedirs(opt['log_dir'])
pprint.pprint(config)
data = load_data(opt)
training_model = model(config, data, test=False)
training_model.train()
if val_mode:
opt = config['val_opt']
data = load_data(opt)
training_model = model(config, data, test=True)
# training_model.load_model(args.model_dir)
# pdb.set_trace()
training_model.eval('test', save=opt['log_dir'], window=opt['window'])
if test_mode:
opt = config['testting_opt']
data = load_data(opt)
training_model = model(config, data, test=True)
# load checkpoints
# training_model.load_model(args.model_dir)
training_model.eval('test', save=opt['log_dir'], window=opt['window'])
data['meta'] = dataloader.load_data(
data_root=data_root[dataset.rstrip('_LT')],
dataset=dataset,
phase='train' if 'CIFAR' in dataset else 'val',
batch_size=sampler_defs.get(
'meta_batch_size',
training_opt['batch_size'],
),
sampler_dic=cbs_sampler_dic,
num_workers=training_opt['num_workers'],
cifar_imb_ratio=training_opt['cifar_imb_ratio']
if 'cifar_imb_ratio' in training_opt else None,
meta=True)
training_model = model(config,
data,
test=False,
meta_sample=True,
learner=learner)
else:
training_model = model(config, data, test=False)
training_model.train()
else:
warnings.filterwarnings("ignore", "(Possibly )?corrupt EXIF data",
UserWarning)
print('Under testing phase, we load training data simply to calculate \
training data number for each class.')
if not os.path.isdir(training_opt['log_dir']):
os.makedirs(training_opt['log_dir'])
if not test_mode:
data = {
x: dataloader.load_data(data_root=config['data_root'],
dataset=dataset,
phase=x,
batch_size=args.batch_size,
use_sampler=parameters["sampler"],
num_workers=training_opt['num_workers'],
gamma=args.gamma)
for x in ['train', 'val', 'test']
}
training_model = model(config, data, parameters=parameters, test=False)
training_model.train()
else:
warnings.filterwarnings("ignore", "(Possibly )?corrupt EXIF data",
UserWarning)
print(
'Under testing phase, we load training data simply to calculate training data number for each class.'
)
data = {
x: dataloader.load_data(data_root=data_root[dataset.rstrip('_LT')],
dataset=dataset,
phase=x,
batch_size=args.batch_size,
num_workers=training_opt['num_workers'],
shuffle=False)
for x in ['train', 'test']
data = {
x: dataloader.load_data(data_root=data_root,
dataset=dataset,
phase=x,
batch_size=training_opt['batch_size'],
sampler_dic=sampler_dic,
num_workers=training_opt['num_workers'],
top_k_class=training_opt['top_k']
if 'top_k' in training_opt else None,
cifar_imb_ratio=training_opt['cifar_imb_ratio']
if 'cifar_imb_ratio' in training_opt else None,
reverse=args.train_reverse)
for x in splits
}
training_model = model(config, data, test=False)
training_model.train()
# ============================================================================
# TESTING
else:
warnings.filterwarnings("ignore", "(Possibly )?corrupt EXIF data",
UserWarning)
print(
'Under testing phase, we load training data simply to calculate training data number for each class.'
)
if 'iNaturalist' in dataset.rstrip('_LT'):
splits = ['train', 'val']
test_split = 'val'
else:
test_mode = args.test
test_open = args.test_open
if test_open:
test_mode = True
output_logits = args.output_logits
config = source_import(args.config).config
training_opt = config['training_opt']
# change
relatin_opt = config['memory']
dataset = training_opt['dataset']
data_loader = dataloader.load_data(data_root=data_root[dataset.rstrip('_LT')],
dataset=dataset,
phase='val',
batch_size=1,
num_workers=training_opt['num_workers'])
baseline_model = model(config, data_loader, test=False)
baseline_model.load_model()
for model in baseline_model.networks.values():
model.eval()
cls_weight = baseline_model.networks['classifier'].module.fc.weight
baseline_model.networks['classifier'].module.fc.bias
cls_weight = cls_weight.norm(dim=1).tolist()
df = pd.read_csv("./analysis/classifier_weight_norm.csv")
df[args.col] = cls_weight
df.to_csv("./analysis/classifier_weight_norm.csv", index=False)
def run_experiment(args):
args.config = './config/config.py'
config = source_import(args.config).config
training_opt = config['training_opt']
training_opt['dataset'] = args.data
training_opt['log_dir'] = './logs/'+args.data
dataset = training_opt['dataset']
training_opt['batch_size']=args.batch_size
training_opt['num_epochs']=args.num_epochs
if not os.path.isdir(training_opt['log_dir']):
os.makedirs(training_opt['log_dir'])
if args.preprocess:
process_data(dataset,training_opt['add_inverse_relation'])
return 0
print('Preparing knowledge graph base!')
kg = KnowledgeGraph(data_root[dataset],args,config)
kg.load_all_triples(data_root[dataset], add_reversed_edges=training_opt['add_inverse_relation'])
pprint.pprint(config)
if args.train:
print('Loading dataset from: %s' % data_root[dataset])
data = {x: dataloader.load_data_kg(data_root=data_root[dataset], dataset=dataset, phase=x,
batch_size=training_opt['batch_size'],kg_base=kg,
add_reversed_edges=training_opt['add_inverse_relation'])
for x in (['train', 'dev'])}
config['training_opt']['num_classes'] = len(data['train'].dataset.relation2id.keys())
training_model = model(config, data, test=False, args=args)
training_model.train()
elif args.test:
warnings.filterwarnings("ignore", "(Possibly )?corrupt EXIF data", UserWarning)
print('Under testing phase, we load training data simply to calculate training data number for each class.')
data = {x: dataloader.load_data_kg(data_root=data_root[dataset], dataset=dataset, phase=x,
batch_size=training_opt['batch_size'], kg_base=kg,
test_open=False,
shuffle=False,
add_reversed_edges=False)
for x in ['train', 'dev']}
config['training_opt']['num_classes'] = len(data['train'].dataset.relation2id.keys())
training_model = model(config, data, test=True,args=args)
training_model.load_model()
training_model.eval_batch(phase='dev', openset=False)
elif args.test_open:
warnings.filterwarnings("ignore", "(Possibly )?corrupt EXIF data", UserWarning)
print('Under testing phase, we load training data simply to calculate training data number for each class.')
data = {x: dataloader.load_data_kg(data_root=data_root[dataset], dataset=dataset, phase=x,
batch_size=training_opt['batch_size'], kg_base=kg,
test_open=True,
shuffle=False,
add_reversed_edges=False)
for x in ['train','dev']}
training_model = model(config, data, test=True,args=args,kg=kg)
training_model.load_model()
training_model.eval_batch(phase='dev', openset=args.test_open)
print('ALL COMPLETED.')
