def main(args): # pylint:disable=redefined-outer-name
if args.cuda and not torch.cuda.is_available(): # cuda is not available
args.cuda = False
# Set the random seed for the program
torch.manual_seed(args.random_seed)
np.random.seed(args.random_seed)
np.set_printoptions(precision=8)
if args.cuda:
torch.cuda.manual_seed(args.random_seed)
utils.makedirs(args.dataset)
search_space, action_list, submodel_manager = \
build_search_space(args)
if args.optimizer == 'EA':
model_selector = EA_Selector(args, search_space, action_list,
submodel_manager)
elif args.optimizer == 'RL':
model_selector = RL_Selector(args, search_space, action_list,
submodel_manager)
elif args.optimizer == 'GS':
model_selector = GridSearch_Selector(args, search_space, action_list,
submodel_manager)
elif args.optimizer == 'RS':
model_selector = RandomSearch_Selector(args, search_space, action_list,
submodel_manager)
else:
raise Exception("[!] Optimizer not found: ", args.optimizer)
exit(1)
print('on selector main: ', args)
model_selector.select()
def main(args): # pylint:disable=redefined-outer-name
# os.environ['CUDA_VISIBLE_DEVICE'] = args.device
if args.cuda and not torch.cuda.is_available(): # cuda is not available
args.cuda = False
print("CUDA NOT AVAILABLE!")
# args.max_epoch = 1
# args.controller_max_step = 1
# args.derive_num_sample = 1
torch.manual_seed(args.random_seed)
if args.cuda:
# print("Device Num", torch.cuda.current_device())
torch.cuda.manual_seed(args.random_seed)
utils.makedirs(args.dataset)
trnr = trainer.Trainer(args)
if args.mode == 'train':
print(args)
trnr.train()
elif args.mode == 'derive':
trnr.derive()
else:
raise Exception(f"[!] Mode not found: {args.mode}")
def main(args): # pylint:disable=redefined-outer-name
if args.cuda and not torch.cuda.is_available(): # cuda is not available
args.cuda = False
# args.max_epoch = 10
# args.controller_max_step = 1
# args.derive_num_sample = 1
torch.manual_seed(args.random_seed)
if args.cuda:
torch.cuda.manual_seed(args.random_seed)
# 创建包含数据集Citeseer名称目录
utils.makedirs(args.dataset)
"""
def makedirs(path):
if not os.path.exists(path):
logger.info("[*] Make directories : {}".format(path))
os.makedirs(path)
"""
#训练入口,获取最终GNN入口
trnr = trainer.Trainer(args)
if args.mode == 'train':
print(args)
trnr.train()
elif args.mode == 'derive':
trnr.derive()
else:
raise Exception(f"[!] Mode not found: {args.mode}")
def main(args): # pylint:disable=redefined-outer-name
if args.cuda and not torch.cuda.is_available(): # cuda is not available
args.cuda = False
torch.manual_seed(args.random_seed)
np.random.seed(args.random_seed)
np.set_printoptions(precision=8)
if args.cuda:
torch.cuda.manual_seed(args.random_seed)
utils.makedirs(args.dataset)
if args.optimizer == 'EA':
trainer = Evolution_Trainer(args)
elif args.optimizer == 'RL':
trainer = RL_Trainer(args)
elif args.optimizer == 'GS':
trainer = GridSearch_Trainer(args)
elif args.optimizer == 'RS':
trainer = RandomSearch_Trainer(args)
else:
raise Exception("[!] Optimizer not found: ", args.optimizer)
if args.mode == 'train':
print(args)
trainer.train()
elif args.mode == 'derive':
trainer.derive()
else:
raise Exception("[!] Mode not found: ", args.mode)
def __init__(self, args):
if args.cuda and not torch.cuda.is_available():
args.cuda = False
torch.manual_seed(args.random_seed)
np.random.seed(args.random_seed)
np.set_printoptions(precision=8)
if args.cuda:
torch.cuda.manual_seed(args.random_seed)
utils.makedirs(args.dataset)
trainer = Evolution_Trainer(args)
trainer.train()
def main(args): # pylint:disable=redefined-outer-name
if args.cuda and not torch.cuda.is_available(): # cuda is not available
args.cuda = False
torch.manual_seed(args.random_seed)
np.random.seed(args.random_seed)
np.set_printoptions(precision=8)
if args.cuda:
torch.cuda.manual_seed(args.random_seed)
utils.makedirs(args.dataset)
trainer = Evolution_Trainer(args)
trainer.train()
def main(args): # pylint:disable=redefined-outer-name
if args.cuda and not torch.cuda.is_available(): # cuda is not available
args.cuda = False
# args.max_epoch = 1
# args.controller_max_step = 1
# args.derive_num_sample = 1
torch.manual_seed(args.random_seed)
if args.cuda:
torch.cuda.manual_seed(args.random_seed)
utils.makedirs(args.dataset)
trnr = trainer.Trainer(args)
if args.mode == 'train':
print(args)
trnr.train()
elif args.mode == 'derive':
trnr.derive()
else:
raise Exception(f"[!] Mode not found: {args.mode}")
def fitness_func(solution, solution_idx):
args = build_args()
if args.cuda and not torch.cuda.is_available(): # cuda is not available
args.cuda = False
torch.manual_seed(args.random_seed)
if args.cuda:
torch.cuda.manual_seed(args.random_seed)
utils.makedirs(args.dataset)
global trnr
search_space = MacroSearchSpace().search_space
gnn = []
gnn.append(search_space['attention_type'][int(solution[0])])
gnn.append(search_space['aggregator_type'][int(solution[1])])
gnn.append(search_space['activate_function'][int(solution[2])])
gnn.append(int(solution[3]))
gnn.append(int(solution[4]))
gnn.append(search_space['attention_type'][int(solution[5])])
gnn.append(search_space['aggregator_type'][int(solution[6])])
gnn.append(search_space['activate_function'][int(solution[7])])
gnn.append(int(solution[8]))
gnn.append(6)
fitness = trnr.genetic_get_reward(gnn)[1]
return fitness
args1 = Parameter(method1)
args2 = Parameter(method2)
args3 = Parameter(method3)
if args1.cuda and not torch.cuda.is_available():
args1.cuda = False
torch.manual_seed(args1.random_seed)
np.random.seed(args1.random_seed)
np.set_printoptions(precision=8)
if args1.cuda:
torch.cuda.manual_seed(args1.random_seed)
utils.makedirs(args1.dataset)
trainer1 = Evolution_Trainer.remote(args1)
trainer2 = Evolution_Trainer.remote(args2)
trainer3 = Evolution_Trainer.remote(args3)
initialize_pop1, initialize_acc1 = trainer1.initialize_population_Random.remote(
)
initialize_pop2, initialize_acc2 = trainer2.initialize_population_Random.remote(
)
initialize_pop3, initialize_acc3 = trainer3.initialize_population_Random.remote(
)
task1 = [initialize_pop1, initialize_acc1]
task2 = [initialize_pop2, initialize_acc2]
task3 = [initialize_pop3, initialize_acc3]