def train(self, train_db, val_db, test_db):
##################################################################
## LOG
##################################################################
logz.configure_output_dir(self.cfg.model_dir)
logz.save_config(self.cfg)
##################################################################
## Main loop
##################################################################
start = time()
min_val_loss = 100000000
for epoch in range(self.epoch, self.cfg.n_epochs):
##################################################################
## Training
##################################################################
torch.cuda.empty_cache()
train_loss, train_accu = self.train_epoch(train_db, epoch)
##################################################################
## Validation
##################################################################
torch.cuda.empty_cache()
val_loss, val_accu = self.validate_epoch(val_db, epoch)
##################################################################
## Logging
##################################################################
# update optim scheduler
current_val_loss = np.mean(val_loss[:,0])
# self.optimizer.update(current_val_loss, epoch)
logz.log_tabular("Time", time() - start)
logz.log_tabular("Iteration", epoch)
logz.log_tabular("AverageLoss", np.mean(train_loss[:, 0]))
logz.log_tabular("AveragePredLoss", np.mean(train_loss[:, 1]))
logz.log_tabular("AverageEmbedLoss", np.mean(train_loss[:, 2]))
logz.log_tabular("AverageAttnLoss", np.mean(train_loss[:, 3]))
logz.log_tabular("AverageObjAccu", np.mean(train_accu[:, 0]))
logz.log_tabular("AverageCoordAccu", np.mean(train_accu[:, 1]))
logz.log_tabular("AverageScaleAccu", np.mean(train_accu[:, 2]))
logz.log_tabular("AverageRatioAccu", np.mean(train_accu[:, 3]))
logz.log_tabular("ValAverageLoss", np.mean(val_loss[:, 0]))
logz.log_tabular("ValAveragePredLoss", np.mean(val_loss[:, 1]))
logz.log_tabular("ValAverageEmbedLoss", np.mean(val_loss[:, 2]))
logz.log_tabular("ValAverageAttnLoss", np.mean(val_loss[:, 3]))
logz.log_tabular("ValAverageObjAccu", np.mean(val_accu[:, 0]))
logz.log_tabular("ValAverageCoordAccu", np.mean(val_accu[:, 1]))
logz.log_tabular("ValAverageScaleAccu", np.mean(val_accu[:, 2]))
logz.log_tabular("ValAverageRatioAccu", np.mean(val_accu[:, 3]))
logz.dump_tabular()
##################################################################
## Checkpoint
##################################################################
self.save_checkpoint(epoch)
def train_PG(opt):
start = time.time()
# Configure output directory for logging
logz.configure_output_dir(opt.logdir)
# Log experimental parameters
logz.save_config(opt)
# Set random seeds
np.random.seed(opt.seed)
torch.manual_seed(opt.seed)
if opt.cuda:
torch.cuda.manual_seed_all(opt.seed)
# Make the gym environment
env = gym.make(opt.env_name)
# Is this env continuous, or discrete?
discrete = isinstance(env.action_space, gym.spaces.Discrete)
# Maximum length for episodes
max_path_length = opt.max_path_length or env.spec.max_episode_steps
# Observation and action sizes
ob_dim = env.observation_space.shape[0]
ac_dim = env.action_space.n if discrete else env.action_space.shape[0]
# Policy net, the underlying model is a mlp
policy_net = PolicyNet(ob_dim, ac_dim, opt.n_layers, opt.size, discrete)
optimizer = optim.Adam(policy_net.parameters(), lr=opt.learning_rate)
# Neural network baseline (the mean reward to reduce the variance of the gradient)
if opt.nn_baseline:
baseline_net = PolicyNet(ob_dim, 1, opt.n_layers, opt.size, False)
baseline_optimizer = optim.Adam(baseline_net.parameters(),
lr=opt.learning_rate)
total_timesteps = 0
for itr in range(opt.n_iter):
print("********** Iteration %i ************" % itr)
# Collect paths until we have enough timesteps
# policy net turns to evaluation mode
policy_net.eval()
# the batch size of this iteration
# yes, the batch size for each iteration varies
timesteps_this_batch = 0
paths = []
while True:
ob = env.reset()
obs, acs, rewards = [], [], []
animate_this_episode = (len(paths) == 0 and (itr % 10 == 0)
and opt.render)
steps = 0
while True:
if animate_this_episode:
env.render()
time.sleep(0.05)
# collect observations
obs.append(ob)
# run the policy net to collect the actions
obs_th = torch.from_numpy(ob[None, :]).float()
if opt.cuda:
obs_th = obs_th.cuda()
if discrete:
probs, _ = policy_net(Variable(obs_th))
# multinomial sampling
m = torch.distributions.Categorical(probs)
ac = m.sample().data.numpy()
else:
mean, sigma = policy_net(Variable(obs_th))
ac = sigma * (mean + Variable(torch.randn(mean.size())))
ac = ac.data.numpy()
ac = ac[0]
acs.append(ac)
# collect rewards
ob, rew, done, _ = env.step(ac)
rewards.append(rew)
steps += 1
if done or steps > max_path_length:
break
path = {
"observation": np.array(obs),
"reward": np.array(rewards),
"action": np.array(acs)
}
paths.append(path)
timesteps_this_batch += pathlength(path)
if timesteps_this_batch > opt.batch_size:
break
total_timesteps += timesteps_this_batch
# Build arrays for observation, action for the policy gradient update by concatenating
# across paths
ob_no = np.concatenate([path["observation"] for path in paths])
ac_na = np.concatenate([path["action"] for path in paths])
# discount for reward computation
gamma = opt.discount
# trajectory rewards
q_n = []
if opt.reward_to_go:
for path in paths:
n_samples = len(path["reward"])
tr = 0.0
pr = []
for k in range(n_samples - 1, -1, -1):
cr = path["reward"][k]
tr = gamma * tr + cr
pr.append(tr)
q_n.extend(pr[::-1])
else:
for path in paths:
n_samples = len(path["reward"])
tr = 0.0
for k in range(n_samples - 1, -1, -1):
cr = path["reward"][k]
tr = gamma * tr + cr
pr = np.ones((n_samples, )) * tr
q_n.extend(pr.tolist())
q_n = np.array(q_n)
# If the neural network baseline is used
# The predicted mean rewards should be subtracted from
# the Q values
if opt.nn_baseline:
baseline_net.eval()
obs_th = torch.from_numpy(ob_no).float()
if opt.cuda:
obs_th = obs_th.cuda()
b_n, _ = baseline_net(Variable(obs_th))
b_n = torch.squeeze(b_n, dim=-1)
b_n = b_n.data.numpy()
adv_n = q_n - b_n
else:
adv_n = q_n.copy()
# Normalize the advantages
# an empirical way to reduce the variance of the gradient
# may or may not help, just an option to try
if not (opt.dont_normalize_advantages):
mu = np.mean(adv_n)
sigma = np.std(adv_n)
adv_n = (adv_n - mu) / sigma
# Pytorch tensors
batch_size = ob_no.shape[0]
ob_no_th = torch.from_numpy(ob_no).float()
ac_na_th = torch.from_numpy(ac_na).float()
adv_n_th = torch.from_numpy(adv_n).float()
if opt.cuda:
ob_no_th = ob_no_th.cuda()
ac_na_th = ac_na_th.cuda()
adv_n_th = adv_n_th.cuda()
if opt.nn_baseline:
# train the baseline network
q_mu = np.mean(q_n)
q_sigma = np.std(q_n)
q_n_th = torch.from_numpy((q_n - q_mu) / q_sigma).float()
if opt.cuda:
q_n_th = q_n_th.cuda()
baseline_net.train()
baseline_criterion = nn.MSELoss()
baseline_net.zero_grad()
pred, _ = baseline_net(Variable(ob_no_th))
pred = torch.squeeze(pred, dim=-1)
baseline_loss = baseline_criterion(pred, Variable(q_n_th))
baseline_loss.backward()
baseline_optimizer.step()
baseline_error = baseline_loss.data[0]
policy_net.train()
policy_net.zero_grad()
if discrete:
probs, _ = policy_net(Variable(ob_no_th))
# multinomial sampling, a biased exploration
m = torch.distributions.Categorical(probs)
loss = -m.log_prob(Variable(ac_na_th)) * Variable(adv_n_th)
loss = loss.mean()
else:
means, sigma = policy_net(Variable(ob_no_th))
diff = (means - Variable(ac_na_th, requires_grad=False) /
(sigma + 1e-8))**2
loss = torch.mean(torch.sum(diff, -1) * Variable(adv_n_th))
loss.backward()
optimizer.step()
error = loss.data[0]
# Log diagnostics
returns = [path["reward"].sum() for path in paths]
ep_lengths = [pathlength(path) for path in paths]
logz.log_tabular("Time", time.time() - start)
logz.log_tabular("Iteration", itr)
logz.log_tabular("AverageReturn", np.mean(returns))
logz.log_tabular("StdReturn", np.std(returns))
logz.log_tabular("MaxReturn", np.max(returns))
logz.log_tabular("MinReturn", np.min(returns))
logz.log_tabular("EpLenMean", np.mean(ep_lengths))
logz.log_tabular("EpLenStd", np.std(ep_lengths))
logz.log_tabular("TimestepsThisBatch", timesteps_this_batch)
logz.log_tabular("TimestepsSoFar", total_timesteps)
logz.log_tabular("Loss", error)
if opt.nn_baseline:
logz.log_tabular("BaselineLoss", baseline_error)
logz.dump_tabular()
def train(self, train_db, val_db, test_db):
##################################################################
## Optimizer
##################################################################
image_encoder_trainable_paras = \
filter(lambda p: p.requires_grad, self.net.image_encoder.parameters())
raw_optimizer = optim.Adam([
{'params': self.net.text_encoder.embedding.parameters(), 'lr': self.cfg.finetune_lr},
{'params': image_encoder_trainable_paras, 'lr': self.cfg.finetune_lr},
{'params': self.net.text_encoder.rnn.parameters()},
{'params': self.net.what_decoder.parameters()},
{'params': self.net.where_decoder.parameters()}
], lr=self.cfg.lr)
optimizer = Optimizer(raw_optimizer, max_grad_norm=self.cfg.grad_norm_clipping)
# scheduler = optim.lr_scheduler.ReduceLROnPlateau(optimizer.optimizer, factor=0.8, patience=3)
scheduler = optim.lr_scheduler.StepLR(optimizer.optimizer, step_size=3, gamma=0.8)
optimizer.set_scheduler(scheduler)
##################################################################
## LOG
##################################################################
logz.configure_output_dir(self.cfg.model_dir)
logz.save_config(self.cfg)
##################################################################
## Main loop
##################################################################
start = time()
for epoch in range(self.cfg.n_epochs):
##################################################################
## Training
##################################################################
torch.cuda.empty_cache()
train_pred_loss, train_attn_loss, train_eos_loss, train_accu = \
self.train_epoch(train_db, optimizer, epoch)
##################################################################
## Validation
##################################################################
torch.cuda.empty_cache()
val_loss, val_accu, val_infos = self.validate_epoch(val_db)
##################################################################
## Sample
##################################################################
torch.cuda.empty_cache()
self.sample(epoch, test_db, self.cfg.n_samples)
torch.cuda.empty_cache()
##################################################################
## Logging
##################################################################
# update optim scheduler
optimizer.update(np.mean(val_loss), epoch)
logz.log_tabular("Time", time() - start)
logz.log_tabular("Iteration", epoch)
logz.log_tabular("TrainAverageError", np.mean(train_pred_loss))
logz.log_tabular("TrainStdError", np.std(train_pred_loss))
logz.log_tabular("TrainMaxError", np.max(train_pred_loss))
logz.log_tabular("TrainMinError", np.min(train_pred_loss))
logz.log_tabular("TrainAverageAccu", np.mean(train_accu))
logz.log_tabular("TrainStdAccu", np.std(train_accu))
logz.log_tabular("TrainMaxAccu", np.max(train_accu))
logz.log_tabular("TrainMinAccu", np.min(train_accu))
logz.log_tabular("ValAverageError", np.mean(val_loss))
logz.log_tabular("ValStdError", np.std(val_loss))
logz.log_tabular("ValMaxError", np.max(val_loss))
logz.log_tabular("ValMinError", np.min(val_loss))
logz.log_tabular("ValAverageAccu", np.mean(val_accu))
logz.log_tabular("ValStdAccu", np.std(val_accu))
logz.log_tabular("ValMaxAccu", np.max(val_accu))
logz.log_tabular("ValMinAccu", np.min(val_accu))
logz.log_tabular("ValAverageObjAccu", np.mean(val_accu[:, 0]))
logz.log_tabular("ValStdObjAccu", np.std(val_accu[:, 0]))
logz.log_tabular("ValMaxObjAccu", np.max(val_accu[:, 0]))
logz.log_tabular("ValMinObjAccu", np.min(val_accu[:, 0]))
logz.log_tabular("ValAveragePoseAccu", np.mean(val_accu[:, 1]))
logz.log_tabular("ValStdPoseAccu", np.std(val_accu[:, 1]))
logz.log_tabular("ValMaxPoseAccu", np.max(val_accu[:, 1]))
logz.log_tabular("ValMinPoseAccu", np.min(val_accu[:, 1]))
logz.log_tabular("ValAverageExprAccu", np.mean(val_accu[:, 2]))
logz.log_tabular("ValStdExprAccu", np.std(val_accu[:, 2]))
logz.log_tabular("ValMaxExprAccu", np.max(val_accu[:, 2]))
logz.log_tabular("ValMinExprAccu", np.min(val_accu[:, 2]))
logz.log_tabular("ValAverageCoordAccu", np.mean(val_accu[:, 3]))
logz.log_tabular("ValStdCoordAccu", np.std(val_accu[:, 3]))
logz.log_tabular("ValMaxCoordAccu", np.max(val_accu[:, 3]))
logz.log_tabular("ValMinCoordAccu", np.min(val_accu[:, 3]))
logz.log_tabular("ValAverageScaleAccu", np.mean(val_accu[:, 4]))
logz.log_tabular("ValStdScaleAccu", np.std(val_accu[:, 4]))
logz.log_tabular("ValMaxScaleAccu", np.max(val_accu[:, 4]))
logz.log_tabular("ValMinScaleAccu", np.min(val_accu[:, 4]))
logz.log_tabular("ValAverageFlipAccu", np.mean(val_accu[:, 5]))
logz.log_tabular("ValStdFlipAccu", np.std(val_accu[:, 5]))
logz.log_tabular("ValMaxFlipAccu", np.max(val_accu[:, 5]))
logz.log_tabular("ValMinFlipAccu", np.min(val_accu[:, 5]))
logz.log_tabular("ValUnigramF3", np.mean(val_infos.unigram_F3()))
logz.log_tabular("ValBigramF3", np.mean(val_infos.bigram_F3()))
logz.log_tabular("ValUnigramP", np.mean(val_infos.unigram_P()))
logz.log_tabular("ValUnigramR", np.mean(val_infos.unigram_R()))
logz.log_tabular("ValBigramP", val_infos.mean_bigram_P())
logz.log_tabular("ValBigramR", val_infos.mean_bigram_R())
logz.log_tabular("ValUnigramPose", np.mean(val_infos.pose()))
logz.log_tabular("ValUnigramExpr", np.mean(val_infos.expr()))
logz.log_tabular("ValUnigramScale", np.mean(val_infos.scale()))
logz.log_tabular("ValUnigramFlip", np.mean(val_infos.flip()))
logz.log_tabular("ValUnigramSim", np.mean(val_infos.unigram_coord()))
logz.log_tabular("ValBigramSim", val_infos.mean_bigram_coord())
logz.dump_tabular()
##################################################################
## Checkpoint
##################################################################
log_info = [np.mean(val_loss), np.mean(val_accu)]
self.save_checkpoint(epoch, log_info)
torch.cuda.empty_cache()
def test(self, test_db):
##################################################################
## LOG
##################################################################
logz.configure_output_dir(self.cfg.model_dir)
logz.save_config(self.cfg)
start = time()
test_loaddb = region_loader(test_db)
test_loader = DataLoader(test_loaddb,
batch_size=self.cfg.batch_size,
shuffle=False,
num_workers=self.cfg.num_workers,
collate_fn=region_collate_fn)
sample_mode = 0 if self.cfg.rl_finetune > 0 else self.cfg.explore_mode
all_txt_feats, all_img_feats, all_img_masks, losses = [], [], [], []
self.net.eval()
for cnt, batched in enumerate(test_loader):
##################################################################
## Batched data
##################################################################
scene_inds, sent_inds, sent_msks, region_feats, region_masks, region_clses = self.batch_data(
batched)
##################################################################
## Inference one step
##################################################################
with torch.no_grad():
img_feats, masked_feats, txt_feats, subspace_masks, sample_logits, sample_indices = \
self.net(scene_inds, sent_inds, sent_msks, None, None, None, region_feats, region_clses, region_masks, sample_mode=sample_mode)
txt_masks = txt_feats.new_ones(txt_feats.size(0),
txt_feats.size(1))
batch_losses = self.net.final_loss(img_feats, masked_feats,
region_masks, txt_feats,
txt_masks, sample_logits,
sample_indices)
loss = torch.sum(torch.mean(batch_losses, -1))
losses.append(loss.cpu().data.item())
all_txt_feats.append(txt_feats)
all_img_masks.append(region_masks)
if self.cfg.subspace_alignment_mode > 0:
all_img_feats.append(masked_feats)
else:
all_img_feats.append(img_feats)
##################################################################
## Print info
##################################################################
if cnt % self.cfg.log_per_steps == 0:
print('Iter %07d:' % (cnt))
tmp_losses = np.stack(losses, 0)
print('mean loss: ', np.mean(tmp_losses))
print('-------------------------')
torch.cuda.empty_cache()
losses = np.array(losses)
all_img_feats = torch.cat(all_img_feats, 0)
all_img_masks = torch.cat(all_img_masks, 0)
all_txt_feats = torch.cat(all_txt_feats, 0)
all_txt_masks = all_txt_feats.new_ones(all_txt_feats.size(0),
all_txt_feats.size(1))
# print('all_img_feats', all_img_feats.size())
all_img_feats_np = all_img_feats.cpu().data.numpy()
all_img_masks_np = all_img_masks.cpu().data.numpy()
with open(
osp.join(self.cfg.model_dir,
'img_features_%d.pkl' % self.cfg.n_feature_dim),
'wb') as fid:
pickle.dump({
'feats': all_img_feats_np,
'masks': all_img_masks_np
}, fid, pickle.HIGHEST_PROTOCOL)
##################################################################
## Evaluation
##################################################################
print('Evaluating the per-turn performance, may take a while.')
metrics, caches_results = self.net.evaluate(all_img_feats,
all_img_masks,
all_txt_feats)
with open(osp.join(self.cfg.model_dir, 'test_metrics.json'),
'w') as fp:
json.dump(metrics, fp, indent=4, sort_keys=True)
with open(osp.join(self.cfg.model_dir, 'test_caches.pkl'),
'wb') as fid:
pickle.dump(caches_results, fid, pickle.HIGHEST_PROTOCOL)
visualize(self.cfg.exp_name, metrics,
osp.join(self.cfg.model_dir, 'evaluation.png'))
return losses, metrics, caches_results
def train(self, train_db, val_db, test_db):
##################################################################
## LOG
##################################################################
logz.configure_output_dir(self.cfg.model_dir)
logz.save_config(self.cfg)
##################################################################
## Main loop
##################################################################
start = time()
min_val_loss = 1000.0
max_val_recall = -1.0
train_loaddb = region_loader(train_db)
val_loaddb = region_loader(val_db)
#TODO
train_loader = DataLoader(train_loaddb,
batch_size=self.cfg.batch_size,
shuffle=True,
num_workers=self.cfg.num_workers,
collate_fn=region_collate_fn)
val_loader = DataLoader(val_loaddb,
batch_size=self.cfg.batch_size,
shuffle=False,
num_workers=self.cfg.num_workers,
collate_fn=region_collate_fn)
for epoch in range(self.epoch, self.cfg.n_epochs):
##################################################################
## Training
##################################################################
if self.cfg.coco_mode >= 0:
self.cfg.coco_mode = np.random.randint(0, self.cfg.max_turns)
torch.cuda.empty_cache()
train_losses = self.train_epoch(train_loaddb, train_loader, epoch)
##################################################################
## Validation
##################################################################
if self.cfg.coco_mode >= 0:
self.cfg.coco_mode = 0
torch.cuda.empty_cache()
val_losses, val_metrics, caches_results = self.validate_epoch(
val_loaddb, val_loader, epoch)
#################################################################
# Logging
#################################################################
# update optim scheduler
current_val_loss = np.mean(val_losses)
self.optimizer.update(current_val_loss, epoch)
logz.log_tabular("Time", time() - start)
logz.log_tabular("Iteration", epoch)
logz.log_tabular("TrainAverageLoss", np.mean(train_losses))
logz.log_tabular("ValAverageLoss", current_val_loss)
mmm = np.zeros((5, ), dtype=np.float64)
for k, v in val_metrics.items():
mmm = mmm + np.array(v)
mmm /= len(val_metrics)
logz.log_tabular("t2i_R1", mmm[0])
logz.log_tabular("t2i_R5", mmm[1])
logz.log_tabular("t2i_R10", mmm[2])
logz.log_tabular("t2i_medr", mmm[3])
logz.log_tabular("t2i_meanr", mmm[4])
logz.dump_tabular()
current_val_recall = np.mean(mmm[:3])
##################################################################
## Checkpoint
##################################################################
if self.cfg.rl_finetune == 0 and self.cfg.coco_mode < 0:
if min_val_loss > current_val_loss:
min_val_loss = current_val_loss
self.save_checkpoint(epoch)
with open(
osp.join(self.cfg.model_dir,
'val_metrics_%d.json' % epoch),
'w') as fp:
json.dump(val_metrics, fp, indent=4, sort_keys=True)
with open(
osp.join(self.cfg.model_dir,
'val_top5_inds_%d.pkl' % epoch),
'wb') as fid:
pickle.dump(caches_results, fid,
pickle.HIGHEST_PROTOCOL)
else:
if max_val_recall < current_val_recall:
max_val_recall = current_val_recall
self.save_checkpoint(epoch)
with open(
osp.join(self.cfg.model_dir,
'val_metrics_%d.json' % epoch),
'w') as fp:
json.dump(val_metrics, fp, indent=4, sort_keys=True)
with open(
osp.join(self.cfg.model_dir,
'val_top5_inds_%d.pkl' % epoch),
'wb') as fid:
pickle.dump(caches_results, fid,
pickle.HIGHEST_PROTOCOL)
def test(self, test_db):
##################################################################
## LOG
##################################################################
logz.configure_output_dir(self.cfg.model_dir)
logz.save_config(self.cfg)
start = time()
test_loaddb = caption_loader(test_db)
test_loader = DataLoader(test_loaddb,
batch_size=self.cfg.batch_size,
shuffle=False,
num_workers=self.cfg.num_workers,
collate_fn=caption_collate_fn)
all_txt_feats, all_img_feats, losses = [], [], []
self.net.eval()
for cnt, batched in enumerate(test_loader):
##################################################################
## Batched data
##################################################################
images, sent_inds, sent_msks = self.batch_data(batched)
##################################################################
## Inference one step
##################################################################
with torch.no_grad():
img_feats, txt_feats = self.net(sent_inds, sent_msks, None,
images)
loss = self.net.forward_loss(img_feats, txt_feats)
losses.append(loss.cpu().data.item())
all_txt_feats.append(txt_feats)
all_img_feats.append(img_feats)
##################################################################
## Print info
##################################################################
if cnt % self.cfg.log_per_steps == 0:
print('Iter %07d:' % (cnt))
tmp_losses = np.stack(losses, 0)
print('mean loss: ', np.mean(tmp_losses))
print('-------------------------')
torch.cuda.empty_cache()
losses = np.array(losses)
all_img_feats = torch.cat(all_img_feats, 0)
all_txt_feats = torch.cat(all_txt_feats, 0)
# print('all_img_feats', all_img_feats.size())
all_img_feats_np = all_img_feats.cpu().data.numpy()
with open(
osp.join(self.cfg.model_dir,
'img_features_%d.pkl' % self.cfg.n_feature_dim),
'wb') as fid:
pickle.dump(all_img_feats_np, fid, pickle.HIGHEST_PROTOCOL)
##################################################################
## Evaluation
##################################################################
print('Evaluating the per-turn performance, may take a while.')
metrics, caches_results = self.net.evaluate(all_img_feats,
all_txt_feats)
with open(osp.join(self.cfg.model_dir, 'test_metrics.json'),
'w') as fp:
json.dump(metrics, fp, indent=4, sort_keys=True)
with open(osp.join(self.cfg.model_dir, 'test_caches.pkl'),
'wb') as fid:
pickle.dump(caches_results, fid, pickle.HIGHEST_PROTOCOL)
visualize(self.cfg.exp_name, metrics,
osp.join(self.cfg.model_dir, 'evaluation.png'))
return losses, metrics, caches_results
def train(self, train_db, val_db, test_db):
##################################################################
## Optimizer
##################################################################
if self.cfg.cuda and self.cfg.parallel:
net = self.net.module
else:
net = self.net
image_encoder_trainable_paras = \
filter(lambda p: p.requires_grad, net.image_encoder.parameters())
raw_optimizer = optim.Adam([
{
'params': image_encoder_trainable_paras
},
{
'params': net.text_encoder.embedding.parameters(),
'lr': self.cfg.finetune_lr
},
{
'params': net.text_encoder.rnn.parameters()
},
{
'params': net.what_decoder.parameters()
},
{
'params': net.where_decoder.parameters()
},
{
'params': net.shape_encoder.parameters()
},
],
lr=self.cfg.lr)
optimizer = Optimizer(raw_optimizer,
max_grad_norm=self.cfg.grad_norm_clipping)
# scheduler = optim.lr_scheduler.ReduceLROnPlateau(optimizer.optimizer, factor=0.8, patience=3)
# scheduler = optim.lr_scheduler.StepLR(optimizer.optimizer, step_size=3, gamma=0.8)
# optimizer.set_scheduler(scheduler)
##################################################################
## LOG
##################################################################
logz.configure_output_dir(self.cfg.model_dir)
logz.save_config(self.cfg)
##################################################################
## Main loop
##################################################################
start = time()
min_val_loss = 100000000
for epoch in range(self.cfg.n_epochs):
##################################################################
## Training
##################################################################
torch.cuda.empty_cache()
train_loss = self.train_epoch(train_db, optimizer, epoch)
##################################################################
## Validation
##################################################################
torch.cuda.empty_cache()
val_loss = self.validate_epoch(val_db, epoch)
##################################################################
## Logging
##################################################################
# update optim scheduler
current_val_loss = np.mean(val_loss[:, 0])
# optimizer.update(current_val_loss, epoch)
logz.log_tabular("Time", time() - start)
logz.log_tabular("Iteration", epoch)
logz.log_tabular("AverageLoss", np.mean(train_loss[:, 0]))
logz.log_tabular("AverageEmbedLoss", np.mean(train_loss[:, 1]))
logz.log_tabular("AverageAttnLoss", np.mean(train_loss[:, 2]))
logz.log_tabular("ValAverageLoss", np.mean(val_loss[:, 0]))
logz.log_tabular("ValAverageEmbedLoss", np.mean(val_loss[:, 1]))
logz.log_tabular("ValAverageAttnLoss", np.mean(val_loss[:, 2]))
logz.dump_tabular()
##################################################################
## Checkpoint
##################################################################
if min_val_loss > current_val_loss:
min_val_loss = current_val_loss
# log_info = [np.mean(val_loss), np.mean(val_accu)]
# self.save_checkpoint(epoch, log_info)
self.save_best_checkpoint()
torch.cuda.empty_cache()
def train(self, train_db, val_db, test_db):
##################################################################
## Optimizer
##################################################################
if self.cfg.cuda and self.cfg.parallel:
net = self.net.module
else:
net = self.net
image_encoder_trainable_paras = \
filter(lambda p: p.requires_grad, net.image_encoder.parameters())
# raw_optimizer = optim.Adam([
# {'params': net.text_encoder.parameters(), 'lr': self.cfg.finetune_lr},
# {'params': image_encoder_trainable_paras},
# {'params': net.what_decoder.parameters()},
# {'params': net.where_decoder.parameters()}
# ], lr=self.cfg.lr)
raw_optimizer = optim.Adam([{
'params': image_encoder_trainable_paras,
'initial_lr': self.cfg.lr
}, {
'params': net.what_decoder.parameters(),
'initial_lr': self.cfg.lr
}, {
'params': net.where_decoder.parameters(),
'initial_lr': self.cfg.lr
}],
lr=self.cfg.lr)
self.optimizer = Optimizer(raw_optimizer,
max_grad_norm=self.cfg.grad_norm_clipping)
# scheduler = optim.lr_scheduler.ReduceLROnPlateau(optimizer.optimizer, factor=0.8, patience=3)
scheduler = optim.lr_scheduler.StepLR(self.optimizer.optimizer,
step_size=3,
gamma=0.8,
last_epoch=self.start_epoch - 1)
self.optimizer.set_scheduler(scheduler)
num_train_steps = int(
len(train_db) / self.cfg.accumulation_steps * self.cfg.n_epochs)
num_warmup_steps = int(num_train_steps * self.cfg.warmup)
self.bert_optimizer = AdamW([{
'params': net.text_encoder.parameters(),
'initial_lr': self.cfg.finetune_lr
}],
lr=self.cfg.finetune_lr)
self.bert_scheduler = get_linear_schedule_with_warmup(
self.bert_optimizer,
num_warmup_steps=num_warmup_steps,
num_training_steps=num_train_steps,
last_epoch=self.start_epoch - 1)
bucket_boundaries = [4, 8, 12, 16, 22] # [4,8,12,16,22]
print('preparing training bucket sampler')
self.train_bucket_sampler = BucketSampler(
train_db, bucket_boundaries, batch_size=self.cfg.batch_size)
print('preparing validation bucket sampler')
self.val_bucket_sampler = BucketSampler(val_db,
bucket_boundaries,
batch_size=4)
##################################################################
## LOG
##################################################################
logz.configure_output_dir(self.cfg.model_dir)
logz.save_config(self.cfg)
##################################################################
## Main loop
##################################################################
start = time()
for epoch in range(self.start_epoch, self.cfg.n_epochs):
##################################################################
## Training
##################################################################
print('Training...')
torch.cuda.empty_cache()
train_pred_loss, train_attn_loss, train_eos_loss, train_accu, train_mse = \
self.train_epoch(train_db, self.optimizer, epoch)
##################################################################
## Validation
##################################################################
print('Validation...')
val_loss, val_accu, val_mse, val_infos = self.validate_epoch(
val_db)
##################################################################
## Sample
##################################################################
if self.cfg.if_sample:
print('Sample...')
torch.cuda.empty_cache()
self.sample(epoch, test_db, self.cfg.n_samples)
torch.cuda.empty_cache()
##################################################################
## Logging
##################################################################
# update optim scheduler
print('Loging...')
self.optimizer.update(np.mean(val_loss), epoch)
logz.log_tabular("Time", time() - start)
logz.log_tabular("Iteration", epoch)
logz.log_tabular("TrainAverageError", np.mean(train_pred_loss))
logz.log_tabular("TrainAverageAccu", np.mean(train_accu))
logz.log_tabular("TrainAverageMse", np.mean(train_mse))
logz.log_tabular("ValAverageError", np.mean(val_loss))
logz.log_tabular("ValAverageAccu", np.mean(val_accu))
logz.log_tabular("ValAverageObjAccu", np.mean(val_accu[:, 0]))
logz.log_tabular("ValAverageCoordAccu", np.mean(val_accu[:, 1]))
logz.log_tabular("ValAverageScaleAccu", np.mean(val_accu[:, 2]))
logz.log_tabular("ValAverageRatioAccu", np.mean(val_accu[:, 3]))
logz.log_tabular("ValAverageMse", np.mean(val_mse))
logz.log_tabular("ValAverageXMse", np.mean(val_mse[:, 0]))
logz.log_tabular("ValAverageYMse", np.mean(val_mse[:, 1]))
logz.log_tabular("ValAverageWMse", np.mean(val_mse[:, 2]))
logz.log_tabular("ValAverageHMse", np.mean(val_mse[:, 3]))
logz.log_tabular("ValUnigramF3", np.mean(val_infos.unigram_F3()))
logz.log_tabular("ValBigramF3", np.mean(val_infos.bigram_F3()))
logz.log_tabular("ValUnigramP", np.mean(val_infos.unigram_P()))
logz.log_tabular("ValUnigramR", np.mean(val_infos.unigram_R()))
logz.log_tabular("ValBigramP", val_infos.mean_bigram_P())
logz.log_tabular("ValBigramR", val_infos.mean_bigram_R())
logz.log_tabular("ValUnigramScale", np.mean(val_infos.scale()))
logz.log_tabular("ValUnigramRatio", np.mean(val_infos.ratio()))
logz.log_tabular("ValUnigramSim",
np.mean(val_infos.unigram_coord()))
logz.log_tabular("ValBigramSim", val_infos.mean_bigram_coord())
logz.dump_tabular()
##################################################################
## Checkpoint
##################################################################
print('Saving checkpoint...')
log_info = [np.mean(val_loss), np.mean(val_accu)]
self.save_checkpoint(epoch, log_info)
torch.cuda.empty_cache()
def train(self, train_db, val_db):
# if self.cfg.cuda and self.cfg.parallel:
# net = self.net.module
# else:
# net = self.net
# net = self.net
# print(net)
if self.cfg.tensorboard_logdir is not None:
summary_writer = SummaryWriter(self.cfg.tensorboard_logdir)
else:
summary_writer = SummaryWriter(
osp.join(self.cfg.log_dir, self.cfg.task, 'tensorboard',
self.cfg.model_name))
# log_per_steps = self.cfg.accumulation_steps * self.cfg.log_per_steps
log_dir = osp.join(self.cfg.log_dir, self.cfg.task,
self.cfg.model_name)
if not osp.exists(log_dir):
os.makedirs(log_dir)
code_dir = osp.join(log_dir, 'code')
if not osp.exists(code_dir):
os.makedirs(code_dir)
shutil.copy('./train.py', osp.join(code_dir, 'train.py'))
shutil.copy('./commonsense_dataset.py',
osp.join(code_dir, 'commonsense_dataset.py'))
logz.configure_output_dir(log_dir)
logz.save_config(self.cfg)
train_loader = DataLoader(train_db,
batch_size=self.cfg.batch_size,
shuffle=True,
num_workers=self.cfg.num_workers)
# self.optimizer = BertAdam(net.parameters(), lr=cfg.lr, warmup=cfg.warmup)
# self.scheduler = optim.lr_self.scheduler.StepLR(self.optimizer, step_size=3, gamma=0.8)
num_train_steps = int(
len(train_loader) / self.cfg.accumulation_steps * self.cfg.epochs)
num_warmup_steps = int(num_train_steps * self.cfg.warmup)
no_decay = ['bias', 'LayerNorm.weight']
not_optim = []
optimizer_grouped_parameters = [{
'params': [
p for n, p in self.net.named_parameters()
if (not any(nd in n for nd in no_decay)) and (not any(
nd in n for nd in not_optim))
],
'weight_decay':
self.cfg.weight_decay
}, {
'params': [
p for n, p in self.net.named_parameters()
if (any(nd in n
for nd in no_decay)) and (not any(nd in n
for nd in not_optim))
],
'weight_decay':
0.0
}]
if self.cfg.fix_emb:
for p in self.net.embedding.embeddings.parameters():
p.requires_grad = False
if self.cfg.ft_last_layer:
for p in self.net.embedding.embeddings.parameters():
p.requires_grad = False
for i in range(10):
for p in self.net.embedding.encoder.layer[i]:
p.requires_grad = False
self.optimizer = AdamW(optimizer_grouped_parameters,
lr=self.cfg.lr,
eps=self.cfg.adam_eps,
betas=eval(self.cfg.adam_betas))
# self.optimizer = AdamW(self.net.parameters(), lr=self.cfg.lr, eps=1e-8)
self.scheduler = WarmupLinearSchedule(self.optimizer,
warmup_steps=num_warmup_steps,
t_total=num_train_steps)
loss_func = nn.CrossEntropyLoss()
if self.cfg.fp16:
try:
from apex import amp
except ImportError:
raise ImportError(
"Please install apex from https://www.github.com/nvidia/apex to use fp16 training."
)
self.net, self.optimizer = amp.initialize(
self.net, self.optimizer, opt_level=self.cfg.fp16_opt_level)
# self.scheduler = get_linear_schedule_with_warmup(optimizer, num_warmup_steps=num_warmup_steps, num_training_steps=num_train_steps)
# self.optimizer.set_self.scheduler(self.scheduler)
torch.cuda.synchronize()
self.start = time.time()
self.net.zero_grad()
self.batch_loss, self.batch_acc = [], []
self.global_step = 0
for epoch in range(self.start_epoch, self.cfg.epochs):
print('Training...')
torch.cuda.empty_cache()
self.batch_loss, self.batch_acc = [], []
for cnt, batch in tqdm(enumerate(train_loader)):
self.net.train()
input_ids, input_mask, segment_ids, features, fea_mask, labels, input_indexs = self.batch_data(
batch)
batch_input = (input_ids, input_mask, segment_ids, features,
fea_mask)
# self.net.zero_grad()
logits, probabilities, preds = self.net(
input_ids, input_mask, segment_ids, features, fea_mask)
loss = loss_func(logits, labels).mean()
# print(probabilities)
# one_hot_labels = nn.functional.one_hot(labels, num_classes = Number_class[self.cfg.task.lower()]).float()
# per_example_loss = -torch.sum(one_hot_labels * log_probs, dim=-1)
# loss = torch.mean(per_example_loss)
if self.cfg.accumulation_steps > 1:
loss = loss / self.cfg.accumulation_steps
if self.cfg.fp16:
with amp.scale_loss(loss, self.optimizer) as scaled_loss:
scaled_loss.backward()
if self.cfg.max_grad_norm > 0.0:
torch.nn.utils.clip_grad_norm_(
amp.master_params(self.optimizer),
self.cfg.max_grad_norm)
else:
loss.backward()
if self.cfg.max_grad_norm > 0.0:
nn.utils.clip_grad_norm_(self.net.parameters(),
self.cfg.max_grad_norm)
acc, _, _, _ = self.evaluate(preds, labels, input_indexs)
self.batch_loss.append(loss.cpu().data.item() / len(input_ids))
self.batch_acc.append(acc)
if self.global_step == 0 and cnt == 0:
_ = self.update_log(summary_writer, epoch, val_db)
if ((cnt + 1) % self.cfg.accumulation_steps) == 0:
# print(nn.utils.clip_grad_norm_(self.net.parameters(), max_norm=1e5))
self.optimizer.step()
self.scheduler.step()
self.net.zero_grad()
self.global_step += 1
if self.global_step % self.cfg.log_per_steps == 0:
val_acc = self.update_log(summary_writer, epoch,
val_db)
self.batch_loss, self.batch_acc = [], []
if self.cfg.save_ckpt:
if epoch >= (self.cfg.epochs / 4):
if self.best < val_acc:
print('Saving checkpoint...')
self.save_checkpoint(epoch, acc=val_acc)
self.best = val_acc
##################################################################
## Checkpoint
##################################################################
if len(self.batch_loss) > 0:
val_acc = self.update_log(summary_writer, epoch, val_db)
self.best = max(self.best, val_acc)
self.batch_loss, self.batch_acc = [], []
# val_wrong_qa = []
# for q, a in zip(val_wrong, val_wrong_answer):
# val_wrong_qa.append([val_db.index2qid[q], trainer.index2label[a]])
# epoch_wrong = {epoch: val_wrong_qa}
if self.cfg.save_ckpt:
if epoch >= (self.cfg.epochs / 4):
print('Saving checkpoint...')
self.save_checkpoint(epoch, True, acc=val_acc)
torch.cuda.empty_cache()
summary_writer.close()
def train(self, train_db, val_db, test_db):
##################################################################
## LOG
##################################################################
logz.configure_output_dir(self.cfg.model_dir)
logz.save_config(self.cfg)
##################################################################
## NN table
##################################################################
if self.cfg.use_hard_mining:
self.train_tables = AllCategoriesTables(train_db)
self.val_tables = AllCategoriesTables(val_db)
self.train_tables.build_nntables_for_all_categories(True)
self.val_tables.build_nntables_for_all_categories(True)
##################################################################
## Main loop
##################################################################
start = time()
min_val_loss = 100000000
for epoch in range(self.epoch, self.cfg.n_epochs):
##################################################################
## Training
##################################################################
torch.cuda.empty_cache()
train_loss, train_accu = self.train_epoch(train_db, epoch)
##################################################################
## Validation
##################################################################
torch.cuda.empty_cache()
val_loss, val_accu = self.validate_epoch(val_db, epoch)
##################################################################
## Logging
##################################################################
# update optim scheduler
current_val_loss = np.mean(val_loss[:, 0])
# self.optimizer.update(current_val_loss, epoch)
logz.log_tabular("Time", time() - start)
logz.log_tabular("Iteration", epoch)
logz.log_tabular("AverageLoss", np.mean(train_loss[:, 0]))
logz.log_tabular("AveragePredLoss", np.mean(train_loss[:, 1]))
logz.log_tabular("AverageEmbedLoss", np.mean(train_loss[:, 2]))
logz.log_tabular("AverageAttnLoss", np.mean(train_loss[:, 3]))
logz.log_tabular("AverageObjAccu", np.mean(train_accu[:, 0]))
logz.log_tabular("AverageCoordAccu", np.mean(train_accu[:, 1]))
logz.log_tabular("AverageScaleAccu", np.mean(train_accu[:, 2]))
logz.log_tabular("AverageRatioAccu", np.mean(train_accu[:, 3]))
logz.log_tabular("ValAverageLoss", np.mean(val_loss[:, 0]))
logz.log_tabular("ValAveragePredLoss", np.mean(val_loss[:, 1]))
logz.log_tabular("ValAverageEmbedLoss", np.mean(val_loss[:, 2]))
logz.log_tabular("ValAverageAttnLoss", np.mean(val_loss[:, 3]))
logz.log_tabular("ValAverageObjAccu", np.mean(val_accu[:, 0]))
logz.log_tabular("ValAverageCoordAccu", np.mean(val_accu[:, 1]))
logz.log_tabular("ValAverageScaleAccu", np.mean(val_accu[:, 2]))
logz.log_tabular("ValAverageRatioAccu", np.mean(val_accu[:, 3]))
logz.dump_tabular()
##################################################################
## Checkpoint
##################################################################
if self.cfg.use_hard_mining:
if (epoch + 1) % 3 == 0:
torch.cuda.empty_cache()
t0 = time()
self.dump_shape_vectors(train_db)
torch.cuda.empty_cache()
self.dump_shape_vectors(val_db)
print("Dump shape vectors completes (time %.2fs)" %
(time() - t0))
torch.cuda.empty_cache()
t0 = time()
self.train_tables.build_nntables_for_all_categories(False)
self.val_tables.build_nntables_for_all_categories(False)
print("NN completes (time %.2fs)" % (time() - t0))
self.save_checkpoint(epoch)
def train(self, train_db, val_db, test_db):
##################################################################
## LOG
##################################################################
logz.configure_output_dir(self.cfg.model_dir)
logz.save_config(self.cfg)
##################################################################
## Main loop
##################################################################
start = time()
min_val_loss = 100000000
for epoch in range(self.epoch, self.cfg.n_epochs):
##################################################################
## Training
##################################################################
torch.cuda.empty_cache()
train_loss = self.train_epoch(train_db, epoch)
##################################################################
## Validation
##################################################################
torch.cuda.empty_cache()
val_loss = self.validate_epoch(val_db, epoch)
# val_loss = train_loss
##################################################################
## Sample
##################################################################
torch.cuda.empty_cache()
self.sample_for_vis(epoch, test_db, self.cfg.n_samples)
torch.cuda.empty_cache()
##################################################################
## Logging
##################################################################
# update optim scheduler
current_val_loss = np.mean(val_loss)
logz.log_tabular("Time", time() - start)
logz.log_tabular("Iteration", epoch)
logz.log_tabular("AverageTotalError", np.mean(train_loss[:, 0]))
logz.log_tabular("AveragePredError", np.mean(train_loss[:, 1]))
logz.log_tabular("AverageImageError", np.mean(train_loss[:, 2]))
logz.log_tabular("AverageFeat0Error", np.mean(train_loss[:, 3]))
logz.log_tabular("AverageFeat1Error", np.mean(train_loss[:, 4]))
logz.log_tabular("AverageFeat2Error", np.mean(train_loss[:, 5]))
logz.log_tabular("AverageFeat3Error", np.mean(train_loss[:, 6]))
logz.log_tabular("AverageFeat4Error", np.mean(train_loss[:, 7]))
logz.log_tabular("ValAverageTotalError", np.mean(val_loss[:, 0]))
logz.log_tabular("ValAveragePredError", np.mean(val_loss[:, 1]))
logz.log_tabular("ValAverageImageError", np.mean(val_loss[:, 2]))
logz.log_tabular("ValAverageFeat0Error", np.mean(val_loss[:, 3]))
logz.log_tabular("ValAverageFeat1Error", np.mean(val_loss[:, 4]))
logz.log_tabular("ValAverageFeat2Error", np.mean(val_loss[:, 5]))
logz.log_tabular("ValAverageFeat3Error", np.mean(val_loss[:, 6]))
logz.log_tabular("ValAverageFeat4Error", np.mean(val_loss[:, 7]))
logz.dump_tabular()
##################################################################
## Checkpoint
##################################################################
if min_val_loss > current_val_loss:
min_val_loss = current_val_loss
self.save_checkpoint(epoch)
torch.cuda.empty_cache()
def train(self, train_db, val_db):
##################################################################
## Optimizer
##################################################################
if self.cfg.cuda and self.cfg.parallel:
net = self.net.module
else:
net = self.net
optimizer = optim.Adam([{
'params': net.encoder.parameters()
}, {
'params': net.decoder.parameters()
}],
lr=self.cfg.lr)
##################################################################
## LOG
##################################################################
logz.configure_output_dir(self.cfg.model_dir)
logz.save_config(self.cfg)
##################################################################
## Main loop
##################################################################
start = time()
min_val_loss = 100000000
for epoch in range(self.cfg.n_epochs):
##################################################################
## Training
##################################################################
torch.cuda.empty_cache()
train_loss = self.train_epoch(train_db, optimizer, epoch)
##################################################################
## Validation
##################################################################
torch.cuda.empty_cache()
val_loss = self.validate_epoch(val_db, epoch)
# val_loss = train_loss
##################################################################
## Sample
##################################################################
torch.cuda.empty_cache()
self.sample(epoch, val_db, self.cfg.n_samples)
torch.cuda.empty_cache()
##################################################################
## Logging
##################################################################
# update optim scheduler
current_val_loss = np.mean(val_loss)
logz.log_tabular("Time", time() - start)
logz.log_tabular("Iteration", epoch)
logz.log_tabular("AverageTotalError", np.mean(train_loss[:, 0]))
logz.log_tabular("AveragePredError", np.mean(train_loss[:, 1]))
logz.log_tabular("AverageImageError", np.mean(train_loss[:, 2]))
logz.log_tabular("AverageFeat0Error", np.mean(train_loss[:, 3]))
logz.log_tabular("AverageFeat1Error", np.mean(train_loss[:, 4]))
logz.log_tabular("AverageFeat2Error", np.mean(train_loss[:, 5]))
logz.log_tabular("AverageFeat3Error", np.mean(train_loss[:, 6]))
logz.log_tabular("AverageFeat4Error", np.mean(train_loss[:, 7]))
logz.log_tabular("ValAverageTotalError", np.mean(val_loss[:, 0]))
logz.log_tabular("ValAveragePredError", np.mean(val_loss[:, 1]))
logz.log_tabular("ValAverageImageError", np.mean(val_loss[:, 2]))
logz.log_tabular("ValAverageFeat0Error", np.mean(val_loss[:, 3]))
logz.log_tabular("ValAverageFeat1Error", np.mean(val_loss[:, 4]))
logz.log_tabular("ValAverageFeat2Error", np.mean(val_loss[:, 5]))
logz.log_tabular("ValAverageFeat3Error", np.mean(val_loss[:, 6]))
logz.log_tabular("ValAverageFeat4Error", np.mean(val_loss[:, 7]))
logz.dump_tabular()
##################################################################
## Checkpoint
##################################################################
if min_val_loss > current_val_loss:
min_val_loss = current_val_loss
self.save_checkpoint()
torch.cuda.empty_cache()