type=float,
default=0.05,
help='discretization minimum time interval')
parser.add_argument('--rnn_steps',
type=int,
default=10,
help='rnn readout steps')
args = parser.parse_args()
data_index = args.data
save_path = '/media/wang/DATASET/CARLA/town01/' + str(data_index) + '/'
log_path = '/home/cz/result/log/' + args.name + '/'
ckpt_path = '/home/cz/result/saved_models/%s' % args.name
logger = SummaryWriter(log_dir=log_path)
generator = Generator(input_dim=1 + 1 + args.vector_dim, output=2).to(device)
model = TD3(args=args,
buffer_size=1e5,
noise_decay_steps=3e3,
batch_size=64,
logger=logger,
policy_freq=5,
is_fix_policy_net=True) #48 85
# encoder = EncoderWithV(input_dim=6, out_dim=args.vector_dim).to(device)
try:
model.policy_net.load_state_dict(
torch.load(
'/home/cz/Downloads/learning-uncertainty-master/scripts/encoder_e2e.pth'
))
# model.policy_net.load_state_dict(torch.load('/home/cz/result/saved_models/rl-train-img-nav-04-train/115_policy_net.pkl'))
model.value_net1.load_state_dict(
if opt.test_mode: opt.batch_size = 1
description = 'dropout'
log_path = 'result/log/' + opt.dataset_name + '/'
os.makedirs('result/saved_models/%s' % opt.dataset_name, exist_ok=True)
os.makedirs('result/output/%s' % opt.dataset_name, exist_ok=True)
os.makedirs('result/output2/%s' % opt.dataset_name, exist_ok=True)
os.makedirs('result/output3/%s' % opt.dataset_name, exist_ok=True)
if not opt.test_mode:
logger = SummaryWriter(log_dir=log_path)
write_params(log_path, parser, description)
# generator = Generator(input_dim=128+32+1+1, output=2).to(device)
# discriminator = Discriminator(opt.points_num*2+32+1).to(device)
generator = Generator(input_dim=2 + 2 + 1 + 1, output=2).to(device)
discriminator = Discriminator(opt.points_num * 2 + 2 + 1).to(device)
# encoder = CNN(input_dim=1, out_dim=32).to(device)
encoder = CNNNorm(input_dim=1, out_dim=2).to(device)
encoder.load_state_dict(
torch.load('result/saved_models/il-uncertainty-02/encoder_119000.pth'))
# DO NOT TRAIN ENCODER
encoder.eval()
# discriminator.load_state_dict(torch.load('result/saved_models/train-gan-costmap-01/discriminator_120000.pth'))
generator.load_state_dict(
torch.load(
'result/saved_models/train-gan-costmap-01/generator_120000.pth'))
start_point_criterion = torch.nn.MSELoss()
criterion = torch.nn.BCELoss() #.to(device)
trajectory_criterion = torch.nn.MSELoss()
parser.add_argument('--max_dist', type=float, default=25., help='max distance')
parser.add_argument('--max_speed', type=float, default=10., help='max speed')
parser.add_argument('--max_t', type=float, default=3., help='max time')
opt = parser.parse_args()
if opt.test_mode: opt.batch_size = 1
description = 'wgan-gp mirror v0/opt.max_speed'
log_path = 'result/log/' + opt.dataset_name + '/'
os.makedirs('result/saved_models/%s' % opt.dataset_name, exist_ok=True)
os.makedirs('result/output/%s' % opt.dataset_name, exist_ok=True)
os.makedirs('result/output2/%s' % opt.dataset_name, exist_ok=True)
if not opt.test_mode:
logger = SummaryWriter(log_dir=log_path)
write_params(log_path, parser, description)
generator = Generator(opt.vector_dim + 2).to(device)
discriminator = Discriminator(opt.points_num * 2 + 1).to(device)
#discriminator.load_state_dict(torch.load('result/saved_models/wgan-gp-10/discriminator_40000.pth'))
#generator.load_state_dict(torch.load('result/saved_models/wgan-gp-10/generator_40000.pth'))
start_point_criterion = torch.nn.MSELoss()
criterion = torch.nn.BCELoss() #.to(device)
trajectory_criterion = torch.nn.MSELoss()
g_optimizer = torch.optim.RMSprop(generator.parameters(),
lr=opt.lr,
weight_decay=opt.weight_decay)
#g_optimizer = torch.optim.Adam(generator.parameters(), lr=opt.lr, weight_decay=opt.weight_decay)
d_optimizer = torch.optim.RMSprop(discriminator.parameters(),
lr=opt.lr,
weight_decay=opt.weight_decay)
#d_optimizer = torch.optim.Adam(discriminator.parameters(), lr=opt.lr, weight_decay=opt.weight_decay)
global_dict['max_steer_angle'] = 0.
global_dict['ipm_image'] = np.zeros((200, 400), dtype=np.uint8)
global_dict['ipm_image'].fill(255)
global_dict['trans_costmap_dict'] = {}
global_dict['state0'] = None
global_dict['start_control'] = False
random.seed(datetime.now())
torch.manual_seed(999)
torch.cuda.manual_seed(999)
device = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')
generator = GeneratorUNet()
generator = generator.to(device)
generator.load_state_dict(torch.load('../ckpt/g.pth'))
trajectory_model = Generator(4).to(device)
trajectory_model.load_state_dict(
torch.load('../result/saved_models/train-gan-02/generator_20000.pth'))
trajectory_model.eval()
generator.eval()
parser = argparse.ArgumentParser(description='Params')
parser.add_argument('--name',
type=str,
default="rl-train-04",
help='name of the script')
parser.add_argument('-s',
'--save',
type=bool,
default=False,
help='save result')
# 'callback':lidar_callback,
},
}
# param = Param()
param = cu.parse_yaml_file_unsafe('./param.yaml')
sensor = cu.PesudoSensor(sensor_dict['camera']['transform'], config['camera'])
sensor_master = CarlaSensorMaster(sensor,
sensor_dict['camera']['transform'],
binded=True)
# collect_perspective = CollectPerspectiveImage(param, sensor_master)
camera_param = cu.CameraParams(sensor)
# import pdb; pdb.set_trace()
pm = PerspectiveMapping(param, camera_param.K_augment, camera_param.T_img_imu)
generator = Generator(opt.vector_dim + 256 + 1 + 1).to(device)
discriminator = Discriminator(opt.points_num * 2 + 256 + 1).to(device)
encoder = CNN(input_dim=3, out_dim=256).to(device)
# discriminator.load_state_dict(torch.load('result/saved_models/train-cgan-12/discriminator_1000.pth'))
# generator.load_state_dict(torch.load('result/saved_models/train-cgan-12/generator_1000.pth'))
# encoder.load_state_dict(torch.load('result/saved_models/train-cgan-12/encoder_1000.pth'))
# discriminator.load_state_dict(torch.load('result/saved_models/train-cgan-01/discriminator_10000.pth'))
# generator.load_state_dict(torch.load('result/saved_models/train-cgan-01/generator_10000.pth'))
# encoder.load_state_dict(torch.load('result/saved_models/train-cgan-01/encoder_10000.pth'))
# discriminator.load_state_dict(torch.load('result/saved_models/train-cgan-09/discriminator_87000.pth'))
# generator.load_state_dict(torch.load('result/saved_models/train-cgan-09/generator_87000.pth'))
# encoder.load_state_dict(torch.load('result/saved_models/train-cgan-09/encoder_87000.pth'))
start_point_criterion = torch.nn.MSELoss()
trajectory_criterion = torch.nn.MSELoss()
g_optimizer = torch.optim.RMSprop(generator.parameters(),
parser.add_argument('--max_dist', type=float, default=25., help='max distance')
parser.add_argument('--max_speed', type=float, default=10., help='max speed')
parser.add_argument('--max_t', type=float, default=3., help='max time')
opt = parser.parse_args()
if opt.test_mode: opt.batch_size = 1
description = 'CNN'
log_path = 'result/log/' + opt.dataset_name + '/'
os.makedirs('result/saved_models/%s' % opt.dataset_name, exist_ok=True)
os.makedirs('result/output/%s' % opt.dataset_name, exist_ok=True)
os.makedirs('result/output2/%s' % opt.dataset_name, exist_ok=True)
if not opt.test_mode:
logger = SummaryWriter(log_dir=log_path)
write_params(log_path, parser, description)
generator = Generator(opt.vector_dim + 2, output=4).to(device)
# generator.load_state_dict(torch.load('./result/saved_models/il-uncertainty-02/generator_119000.pth'))
encoder = CNN(input_dim=1, out_dim=opt.vector_dim).to(device)
encoder.load_state_dict(
torch.load('./result/saved_models/il-uncertainty-02/encoder_119000.pth'))
generator.load_state_dict(
torch.load('./result/saved_models/il-uncertainty-02/generator_119000.pth'))
criterion = torch.nn.MSELoss()
e_optimizer = torch.optim.Adam(encoder.parameters(),
lr=opt.lr,
weight_decay=opt.weight_decay)
g_optimizer = torch.optim.Adam(generator.parameters(),
lr=opt.lr,