def main():
config = Config('mycfg2.cfg')
# config.remove_peer('third')
# config.add_peer('third')
# config.create('192.168.8.1', 'eth0', 125625, 'ASDA5D5A2AS15D1AS61D61A6D')
# config.add_peer('shitpost', 'SSD5ASD25C25XV4FGGDC6', '192.168.8.9' )
config.remove_peer('shitpost')
def main():
"""Main app process. This controls every step of the process"""
# TODO: Allow users to supply alt configs?
try:
config = Config('./Ripmaster.ini')
# IOError will raise if iniFile is not found. ValueError will raise if
# iniFile is missing options.
except (IOError or ValueError), ex:
print ex
return
def save(self): # Update daemon config file
# Make a new Config object
fileConfig = Config(self.fileName, load=False)
# Copy values that could be changed to the new Config object and convert representation
ro = self.get('read-only', False)
fileConfig['read-only'] = '' if ro else None
fileConfig['overwrite'] = '' if self.get('overwrite',
False) and ro else None
fileConfig['startonstartofindicator'] = self.get(
'startonstartofindicator', True)
fileConfig['stoponexitfromindicator'] = self.get(
'stoponexitfromindicator', False)
exList = self.get('exclude-dirs', None)
fileConfig['exclude-dirs'] = (None if exList is None else
', '.join(v for v in CVal(exList)))
# Store changed values
fileConfig.save()
self.changed = False
def main():
if len(sys.argv) < 2:
print '\nUsage: ' + sys.argv[0] + ' fd.cfg\n'
sys.exit(-1)
config = Config.Config(sys.argv[1])
global_cfg = {
'global' : {
'server.socket_host' : str(config.http_ip),
'server.socket_port' : int(config.http_port),
'server.thread_pool' : int(config.thread_pool),
'server.request_queue_size' : int(config.request_queue_size),
'server.socket_timeout': int(config.timeout),
'request.show_tracebacks' : False,
'response.timeout': int(config.timeout),
'engine.autoreload_on' : False,
'log.screen': config.log_output,
'log.error_file': config.log_error_file,
'log.access_file': config.log_access_file,
'environment': config.environment,
'tools.gzip.on': config.gzip
}
}
current_dir = os.path.dirname(os.path.abspath(__file__))
cfg = {
'/css' : {
'tools.staticdir.on' : True,
'tools.staticdir.dir' : "css",
'tools.staticdir.root' : current_dir
}
}
cherrypy.config.update(global_cfg)
cherrypy.quickstart(FlightDaemon(config), '/', config = cfg)
import os
import os.path as osp
import torch
from tools import Config, setup_seed, initializeWeights
from BatchData import VD_REGLoader, get_iterator, get_data_file, load_vocab
from Trainer import REGTrainer
from REGer import REGModel
# choose GPU
torch.set_num_threads(3)
os.environ["CUDA_VISIBLE_DEVICES"] = '3'
setup_seed(1234)
# read config
cfg = Config('./configs/BaseModel.yml')
# set datapath
data_root = cfg.DATA_PATH
data_cfg = cfg.DATA
data_path = osp.join(data_root, data_cfg.DATA_SET, data_cfg.SPLIT + '_split')
# load vocab
vocab = load_vocab(data_path)
print('vocab_length:', len(vocab))
# load dataset
train_data = get_iterator(VD_REGLoader(vocab, cfg, split='train'),
cfg.TRAIN.T_BS)
eval_data = get_iterator(VD_REGLoader(vocab, cfg, split='val'), cfg.TRAIN.V_BS)
testA_data = get_iterator(VD_REGLoader(vocab, cfg, split='testA'),
cfg.TRAIN.V_BS)
config = Config(
# augmentation
sequence_change = 0.3,
zoom_range = (1.2, 1.2),
random_shear = 0.2,
random_rotation = 20,
# data config
rectangle_imgs = True,
path_to_data='./facs',
saved_paths='./paths.pkl',
# path_to_data='/mnt/course/datasets/facs/',
test_size = 0.1,
batch_size = 16,
img_height = 224,
img_width = 224,
n_emotions = 7,
n_action_units = 41, # target action units
landmark_size = 2*136, # number of features that provide dlib + delta coding
n_frames = 10, # number of images in sequence
# train config
path_to_summaries = './summaries',
path_to_log = './log.csv',
path_to_models = './models',
epochs = 100,
max_queue_size = 100,
workers = 1,
au_map = {64.0: 40, 1.0: 0, 2.0: 3, 43.0: 33, 4.0: 4, 5.0: 5, 6.0: 6,
1.5: 1, 9.0: 8, 10.0: 9, 11.0: 10, 12.0: 11, 13.0: 12, 14.0: 13,
15.0: 14, 16.0: 15, 17.0: 16, 18.0: 17, 1.7: 2, 21.0: 19, 22.0: 20,
23.0: 21, 24.0: 22, 25.0: 23, 26.0: 24, 27.0: 25, 28.0: 26,
29.0: 27, 30.0: 28, 31.0: 29, 34.0: 30, 38.0: 31, 39.0: 32,
7.0: 7, 44.0: 34, 45.0: 35, 54.0: 36, 20.0: 18, 61.0: 37,
62.0: 38, 63.0: 39},
emotion_map = {
0:'Anger',
1:'Contempt',
2:'Disgust',
3:'Fear',
4:'Happiness',
5:'Sadness',
6:'Surprise'
}
)
def initialize(self):
print("Initialize master:")
self.config = Config.Config("Config/pymosy.conf")
self.load_modules()
autostart, notifications, theme, fmextensions and daemons.
The dictionary 'config' stores the config settings for usage in code. Its values are saved to
config file on exit from the Menu.Preferences dialogue or when there is no configuration file
when application starts.
Note that daemon settings ('dir', 'read-only', 'overwrite' and 'exclude_dir') are stored
in ~/ .config/yandex-disk/config.cfg file. They are read in YDDaemon.__init__() method
(in dictionary YDDaemon.config). Their values are saved to daemon config file also
on exit from Menu.Preferences dialogue.
Additionally 'startonstartofindicator' and 'stoponexitfromindicator' values are added into daemon
configuration file to provide the functionality of obsolete 'startonstart' and 'stoponexit'
values for each daemon individually.
"""
APPCONF = Config(pathJoin(APPCONFPATH, APPNAME + '.conf'))
# Read some settings to variables, set default values and update some values
APPCONF['autostart'] = checkAutoStart(APPAUTOSTARTDST)
# Setup on-screen notification settings from config value
APPCONF.setdefault('notifications', True)
APPCONF.setdefault('theme', False)
APPCONF.setdefault('fmextensions', True)
APPCONF.setdefault('daemons', '~/.config/yandex-disk/config.cfg')
# Is it a first run?
if not APPCONF.readSuccess:
LOGGER.info('No config, probably it is a first run.')
# Create application config folders in ~/.config
try:
makeDirs(APPCONFPATH)
makeDirs(pathJoin(APPCONFPATH, 'icons/light'))
makeDirs(pathJoin(APPCONFPATH, 'icons/dark'))
from tools import Config
data_config = Config(path_to_train='./autoria/train',
path_to_test='./autoria/test',
valid_size=0.1,
batch_size=64,
img_height=224,
img_width=224)
train_config = Config(lr_min=1e-5,
lr_max=1e-3,
n_fozen_layers=[5, 15, 25],
path_to_summaries='./summaries',
path_to_log='./log.csv',
path_to_models='./models',
epochs=10,
max_queue_size=100,
workers=1)
config = Config(
scope='classifier',
data=data_config,
train=train_config,
)
logging.info('dial_id %s:' % (ref_id))
logging.info('Pred: %s' % (pred_sent))
logging.info('Dialog Pred: %s' % (dialog_list))
logging.info('Dialog: %s' % (dialog))
logging.info('REs: %s' % (sents))
logging.info('Entities: %s' % (entities))
return all_refer
eval_path = os.path.join(SAVE_PATH, 'eval')
os.makedirs(eval_path, exist_ok=True)
#'''
config_path = osp.join(SAVE_PATH, 'config', 'trainer_config.yml')
cfg = Config(config_path)
print('load model from ', SAVE_PATH)
check_point = torch.load(os.path.join(
SAVE_PATH, 'checkpoints', CHECKPOINT)) #,map_location = lambda _1,_2,:_1)
# load dataset
data_root = cfg.DATA_PATH
data_cfg = cfg.DATA
data_path = osp.join(data_root, data_cfg.DATA_SET, data_cfg.SPLIT + '_split')
vocab = load_vocab(data_path)
print('vocab_length:', len(vocab))
test_loader = VD_REGLoader(vocab, cfg, split=SPLIT)
model_cfg = cfg.MODEL
model = REGModel(vocab, cfg).cuda()
model.load_state_dict(check_point['state_dict'])
from tools import Config
data_config = Config(
path_to_data='/mnt/course/datasets/portraits/imgs',
path_to_masks='/mnt/course/datasets/portraits/masks',
test_size=0.1,
batch_size=16,
img_height=480, #after resize
img_width=400, # after resize
image_shape=(800, 600, 3) # original
)
train_config = Config(lr_min=1e-5,
lr_max=1e-3,
n_fozen_layers=[5, 15, 25],
path_to_summaries='./summaries',
path_to_log='./log.csv',
path_to_models='./models',
epochs=10,
max_queue_size=100,
workers=1)
config = Config(
scope='classifier',
data=data_config,
train=train_config,
)
from tools import Config
config = Config(
# data config
path_to_train_data='./facs',
test_size=0.1,
batch_size=2, ############################################
img_height=48, ############################################
img_width=48, ###############################################
n_emotions=7,
n_action_units=65, # target action units
landmark_size=136, # number of features that provide dlib
img_shape=(640, 490, 3), # original image shape
n_frames=10, # number of images in sequence
# train config
path_to_summaries='./summaries',
path_to_log='./log.csv',
path_to_models='./models',
epochs=100,
max_queue_size=100,
workers=1)
from tools import Config
config = Config(
path_to_train_imgs='/mnt/course/datasets/coco/train2017',
path_to_train_json='/mnt/course/datasets/coco/annotations/instances_train2017.json',
path_to_test_imgs='/mnt/course/datasets/coco/val2017',
path_to_test_json='/mnt/course/datasets/coco/annotations/instances_val2017.json',
test_size = 0.1,
batch_size = 2,
img_height = 240, #after resize
img_width = 320, # after resize
n_classes = 80, # 80 classes
n_obj = 10,
#train config
path_to_summaries = './summaries',
path_to_log = './log.csv',
path_to_models = './models',
epochs = 100,
max_queue_size = 100,
workers = 1
)
from tools import Config
config = Config(
path_to_texts='./OE',
test_size = 0.1,
batch_size = 1024,
max_len = 100,
#train config
path_to_summaries = './summaries',
path_to_log = './log.csv',
path_to_models = './models',
epochs = 200,
max_queue_size = 100,
workers = 1
)
