def Run(args):
# create a Clairvoyante
logging.info("Loading model ...")
if args.v2 == True:
import utils_v2 as utils
utils.SetupEnv()
if args.slim == True:
import clairvoyante_v2_slim as cv
else:
import clairvoyante_v2 as cv
elif args.v3 == True:
import utils_v2 as utils # v3 network is using v2 utils
utils.SetupEnv()
if args.slim == True:
import clairvoyante_v3_slim as cv
else:
import clairvoyante_v3 as cv
if args.threads == None:
if args.tensor_fn == "PIPE":
param.NUM_THREADS = 4
else:
param.NUM_THREADS = args.threads
m = cv.Clairvoyante()
m.init()
m.restoreParameters(os.path.abspath(args.chkpnt_fn))
Test(args, m, utils)
def Run(args):
# create a Clairvoyante
logging.info("Loading model ...")
if args.v2 == True:
import utils_v2 as utils
utils.SetupEnv()
if args.slim == True:
import clairvoyante_v2_slim as cv
else:
import clairvoyante_v2 as cv
elif args.v3 == True:
import utils_v2 as utils # v3 network is using v2 utils
utils.SetupEnv()
if args.slim == True:
import clairvoyante_v3_slim as cv
else:
import clairvoyante_v3 as cv
if args.threads == None:
if args.tensor_fn == "PIPE":
param.NUM_THREADS = 4
else:
param.NUM_THREADS = args.threads
device = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')
m = cpt.Net()
if torch.cuda.device_count() > 0:
print("Let's use", torch.cuda.device_count(), "GPUs!")
if torch.cuda.device_count() > 0:
m.to(device)
m.restoreParameters(os.path.abspath(args.chkpnt_fn))
Test(args, m, utils)
def prepare_data(args):
import utils_v2 as utils # v3 network is using v2 utils
if args.slim == True:
import clairvoyante_v3_slim as cv
else:
import clairvoyante_v3 as cv
utils.SetupEnv()
m = cv.Clairvoyante()
m.init()
m.restoreParameters(args.chkpnt_fn)
if args.bin_fn != None:
with open(args.bin_fn, "rb") as fh:
total = pickle.load(fh)
XArrayCompressed = pickle.load(fh)
YArrayCompressed = pickle.load(fh)
posArrayCompressed = pickle.load(fh)
else:
total, XArrayCompressed, YArrayCompressed, posArrayCompressed = \
utils.GetTrainingArray(args.tensor_fn,
args.var_fn,
args.bed_fn)
return m, utils, total, XArrayCompressed, YArrayCompressed, posArrayCompressed
def Run(args):
# create a Clairvoyante
logging.info("Initializing model ...")
if args.v2 == True:
import utils_v2 as utils
if args.slim == True:
import clairvoyante_v2_slim as cv
else:
import clairvoyante_v2 as cv
elif args.v3 == True:
import utils_v2 as utils # v3 network is using v2 utils
if args.slim == True:
import clairvoyante_v3_slim as cv
else:
import clairvoyante_v3 as cv
utils.SetupEnv()
device = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')
m = cpt.Net()
if torch.cuda.device_count() > 0:
print("Let's use", torch.cuda.device_count(), "GPUs!")
if torch.cuda.device_count() > 0:
m.to(device)
if args.chkpnt_fn != None:
m.restoreParameters(os.path.abspath(args.chkpnt_fn))
TrainAll(args, m, utils)
def Run(args):
# create a Clairvoyante
logging.info("Initializing model ...")
utils.SetupEnv()
m = cv.Clairvoyante()
m.init()
TrainAll(args, m)
Test22(args, m)
def Run(args):
# create a Clairvoyante
import utils_v2 as utils
import clairvoyante_v3 as cv
utils.SetupEnv()
if args.threads == None:
if args.tensor_fn == "PIPE":
param.NUM_THREADS = 1
else:
param.NUM_THREADS = args.threads
m = cv.Clairvoyante()
m.init()
m.restoreParameters(os.path.abspath(args.chkpnt_fn))
Test(args, m, utils)
def Run(args):
# create a Clairvoyante
logging.info("Initializing model ...")
utils.SetupEnv()
device = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')
m = cpt.Net()
if torch.cuda.device_count() > 0:
print("Let's use", torch.cuda.device_count(), "GPUs!")
if torch.cuda.device_count() > 0:
m.to(device)
# m.init()
TrainAll(args, m)
Test22(args, m)
def Prepare(args):
import utils_v2 as utils # v3 network is using v2 utils
if args.slim == True:
import clairvoyante_v3_slim as cv
else:
import clairvoyante_v3 as cv
utils.SetupEnv()
m = cv.Clairvoyante()
m.init()
m.restoreParameters(args.chkpnt_fn)
total, XArrayCompressed, YArrayCompressed, posArrayCompressed = \
utils.GetTrainingArray(args.tensor_fn, args.var_fn, None)
return m, utils, total, XArrayCompressed, YArrayCompressed, posArrayCompressed
def Run(args):
# create a Clairvoyante
if args.v2 == True:
import utils_v2 as utils
if args.slim == True:
import clairvoyante_v2_slim as cv
else:
import clairvoyante_v2 as cv
elif args.v3 == True:
import utils_v2 as utils # v3 network is using v2 utils
if args.slim == True:
import clairvoyante_v3_slim as cv
else:
import clairvoyante_v3 as cv
utils.SetupEnv()
m = cv.Clairvoyante()
m.init()
CalcAll(args, m, utils)
def Run(args):
# create a Clairvoyante
logging.info("Loading model ...")
if args.v2 == True:
import utils_v2 as utils
if args.slim == True:
import clairvoyante_v2_slim as cv
else:
import clairvoyante_v2 as cv
elif args.v3 == True:
import utils_v2 as utils # v3 network is using v2 utils
if args.slim == True:
import clairvoyante_v3_slim as cv
else:
import clairvoyante_v3 as cv
utils.SetupEnv()
m = cv.Clairvoyante()
m.init()
m.restoreParameters(os.path.abspath(args.chkpnt_fn))
Test(args, m, utils)
def Run(args):
# create a Clairvoyante
if args.v2 == True:
import utils_v2 as utils
if args.slim == True:
import clairvoyante_v2_slim as cv
else:
import clairvoyante_v2 as cv
elif args.v3 == True:
import utils_v2 as utils # v3 network is using v2 utils
if args.slim == True:
import clairvoyante_v3_slim as cv
else:
import clairvoyante_v3 as cv
utils.SetupEnv()
m = cv.Clairvoyante()
m.init()
if args.bin_fn != None:
with open(args.bin_fn, "rb") as fh:
total = pickle.load(fh)
XArrayCompressed = pickle.load(fh)
YArrayCompressed = pickle.load(fh)
posArrayCompressed = pickle.load(fh)
else:
total, XArrayCompressed, YArrayCompressed, posArrayCompressed = \
utils.GetTrainingArray(args.tensor_fn,
args.var_fn,
args.bed_fn)
with open(args.chkpnt_list) as fh:
for row in fh:
row = row.rstrip()
logging.info("Working on model: %s" % (row))
m.restoreParameters(os.path.abspath(row))
Test(args, m, utils, total, XArrayCompressed, YArrayCompressed, posArrayCompressed)
def Run(args):
# create a Clairvoyante
logging.info("Initializing model ...")
if args.v2 == True:
import utils_v2 as utils
if args.slim == True:
import clairvoyante_v2_slim as cv
else:
import clairvoyante_v2 as cv
elif args.v3 == True:
import utils_v2 as utils # v3 network is using v2 utils
if args.slim == True:
import clairvoyante_v3_slim as cv
else:
import clairvoyante_v3 as cv
utils.SetupEnv()
m = cv.Clairvoyante()
m.init()
if args.ochk_prefix == None:
sys.exit("--chk_prefix must be defined in nonstop training mode")
if args.chkpnt_fn != None:
m.restoreParameters(os.path.abspath(args.chkpnt_fn))
TrainAll(args, m, utils)
def Run(args):
if args.v2 == True or args.v3 == True:
import utils_v2 as utils
utils.SetupEnv()
Convert(args, utils)
