def main():
# parse the options
opts = parse_args()
# create the dataloaders
dataloader = {'train': create_dataloader('train_valid' if opts.no_validation else 'train', opts),
'valid': create_dataloader('valid', opts)}
# create the model
model = Prover(opts)
model.to(opts.device)
# crete the optimizer
optimizer = torch.optim.RMSprop(model.parameters(), lr=opts.learning_rate,
momentum=opts.momentum,
weight_decay=opts.l2)
if opts.no_validation:
scheduler = StepLR(optimizer, step_size=opts.lr_reduce_steps, gamma=0.1)
else:
scheduler = ReduceLROnPlateau(optimizer, patience=opts.lr_reduce_patience, verbose=True)
# load the checkpoint
start_epoch = 0
if opts.resume != None:
log('loading model checkpoint from %s..' % opts.resume)
if opts.device.type == 'cpu':
checkpoint = torch.load(opts.resume, map_location='cpu')
else:
checkpoint = torch.load(opts.resume)
model.load_state_dict(checkpoint['state_dict'])
optimizer.load_state_dict(checkpoint['optimizer'])
start_epoch = checkpoint['n_epoch'] + 1
model.to(opts.device)
agent = Agent(model, optimizer, dataloader, opts)
best_acc = -1.
for n_epoch in range(start_epoch, start_epoch + opts.num_epochs):
log('EPOCH #%d' % n_epoch)
# training
loss_train = agent.train(n_epoch)
# save the model checkpoint
if n_epoch % opts.save_model_epochs == 0:
agent.save(n_epoch, opts.checkpoint_dir)
# validation
if not opts.no_validation:
loss_valid = agent.valid(n_epoch)
# reduce the learning rate
if opts.no_validation:
scheduler.step()
else:
scheduler.step(loss_valid)
torch.manual_seed(opts.seed)
torch.backends.cudnn.deterministic = True
torch.backends.cudnn.benchmark = False
np.random.seed(opts.seed)
random.seed(opts.seed)
if "ours" in opts.method:
model = Prover(opts)
log("loading model checkpoint from %s.." % opts.path)
if opts.device.type == "cpu":
checkpoint = torch.load(opts.path, map_location="cpu")
else:
checkpoint = torch.load(opts.path)
model.load_state_dict(checkpoint["state_dict"])
model.to(opts.device)
else:
model = None
agent = Agent(model, None, None, opts)
if opts.file:
files = [opts.file]
else:
files = []
projs = json.load(open(opts.projs_split))["projs_" + opts.split]
for proj in projs:
files.extend(
glob(os.path.join(opts.datapath, "%s/**/*.json" % proj),
recursive=True))