import matplotlib
matplotlib.use('Agg')
import os
import sys
import json
import argparse
import time
import numpy as np
import tensorflow as tf
from tensorflow.python import debug as tf_debug
from args import argument_parser, prepare_args, model_kwards, learn_kwards
parser = argument_parser()
args = parser.parse_args()
args = prepare_args(args)
if args.dataset_name == 'gpsamples':
from data.gpsampler import GPSampler
data = np.load("gpsamples_var05.npz")
train_data = {"xs": data['xs'][:50000], "ys": data['ys'][:50000]}
val_data = {"xs": data['xs'][50000:60000], "ys": data['ys'][50000:60000]}
train_set = GPSampler(input_range=[-2., 2.],
var_range=[0.5, 0.5],
max_num_samples=200,
data=train_data)
val_set = GPSampler(input_range=[-2., 2.],
var_range=[0.5, 0.5],
max_num_samples=200,
data=val_data)
elif args.dataset_name == 'sinusoid':
from data.sinusoid import Sinusoid
import omni_torch.visualize.basic as vb
from omni_torch.networks.optimizer import *
import time
import torch
from torch.autograd import Variable
import torch.nn as nn
import torch.optim as optim
import torch.backends.cudnn as cudnn
import torch.nn.init as init
import torch.utils.data as data
import numpy as np
from args import prepare_args
import mmdet.ops.dcn as dcn
from layers.visualization import *
args = prepare_args(VOC_ROOT)
TMPJPG = os.path.expanduser("~/Pictures/tmp.jpg")
torch.set_default_tensor_type('torch.cuda.FloatTensor')
dt = datetime.datetime.now().strftime("%Y-%m-%d_%H:%M")
def avg(list):
return sum(list) / len(list)
def old_fit(args, cfg, net, train_set, optimizer, criterion):
step_index = 0
train_loader = data.DataLoader(train_set,
args.batch_size,
num_workers=args.num_workers,