def thread_way():
"""多线程方式"""
addresses = parse_args()
for i, address in enumerate(addresses):
t = Thread(target=get_poetry, args=(address, ))
t.start()
t.join()
def process_way():
"""多进程方式"""
addresses = parse_args()
p = Pool(4)
for i, address in enumerate(addresses):
p.apply_async(get_poetry, args=(address, ))
p.close()
p.join()
def sync_way():
"""同步方式"""
addresses = parse_args() # 8000 8001 8002
elapsed = datetime.timedelta()
for i, address in enumerate(addresses):
addr_fmt = format_address(address)
print('Task %d: get poetry from: %s' % (i + 1, addr_fmt))
start = datetime.datetime.now()
poem = nonblocking_way(address)
time = datetime.datetime.now() - start
msg = 'Task %d: got %d bytes of poetry from %s in %s'
print(msg % (i + 1, len(poem), addr_fmt, time))
elapsed += time
print('Got %d poems in %s' % (i + 1, elapsed))
import asyncio
import datetime
from parser_util import parse_args
addresses = list(parse_args())
async def fetch(url):
reader, writer = await asyncio.open_connection(host=url[0], port=url[1])
get = b'GET / HTTP/1.0\r\nHost: localhost\r\n\r\n'
writer.write(get)
response = await reader.read()
writer.close()
return response
def main():
loop = asyncio.get_event_loop() # 得到事件循环
tasks = [fetch(address) for address in addresses]
# loop.run_until_complete(asyncio.wait(tasks))
loop.run_until_complete(asyncio.gather(*tasks))
if __name__ == '__main__':
elapsed = datetime.timedelta()
start = datetime.datetime.now()
main()
time = datetime.datetime.now() - start
elapsed += time
print('done in %s' % elapsed)
import torch.utils.data.sampler
import os
import glob
import random
import time
import configs
from convnet import Convnet
from data.datamgr import SetDataManager
from parser_util import parse_args, get_best_file
from utils import set_device, euclidean_dist
if __name__ == '__main__':
params = parse_args('test')
device = set_device(params)
acc_all = []
image_size = 84
iter_num = 600
few_shot_params = dict(n_way=params.test_n_way, n_support=params.n_shot)
model = Convnet()
model = model.to(device)
checkpoint_dir = '%s/checkpoints/%s_%dway_%dshot' % (
configs.save_dir, params.dataset, params.train_n_way, params.n_shot)
if params.train_aug:
checkpoint_dir += '_aug'
outfile = os.path.join(params.checkpoint_dir, 'best_model.tar')
torch.save({'epoch': epoch, 'state': model.state_dict()}, outfile)
# torch.save(trlog, params.checkpoint_dir+'/trlog')
if (epoch % params.save_freq == 0) or (epoch == stop_epoch - 1):
outfile = os.path.join(params.checkpoint_dir,
'{:d}.tar'.format(epoch))
torch.save({'epoch': epoch, 'state': model.state_dict()}, outfile)
return model
if __name__ == '__main__':
params = parse_args('train')
###########################setting##############################
set_seed(params)
###########################dataset################################
base_file = configs.data_dir[params.dataset] + 'base.json'
val_file = configs.data_dir[params.dataset] + 'val.json'
if params.dataset == 'omniglot':
image_size = 28
else:
image_size = 84
if params.stop_epoch == -1:
if params.n_shot == 1: