class Train(object):
def __init__(self):
self.vocab = Vocab(config.vocab_path, config.vocab_size)
self.batcher = Batcher(config.train_data_path, self.vocab, mode='train',
batch_size=config.batch_size, single_pass=False)
time.sleep(15)
train_dir = os.path.join(config.log_root, 'train_%d' % (int(time.time())))
if not os.path.exists(train_dir):
os.mkdir(train_dir)
self.model_dir = os.path.join(train_dir, 'model')
if not os.path.exists(self.model_dir):
os.mkdir(self.model_dir)
self.summary_writer = tf.summary.FileWriter(train_dir)
def save_model(self, running_avg_loss, iter):
state = {
'iter': iter,
'encoder_state_dict': self.model.encoder.state_dict(),
'section_encoder_state_dict': self.model.section_encoder.state_dict(),
'sentence_filterer_state_dict': self.model.sentence_filterer.state_dict(),
'decoder_state_dict': self.model.decoder.state_dict(),
'reduce_state_dict': self.model.reduce_state.state_dict(),
'section_reduce_state_dict': self.model.section_reduce_state.state_dict(),
'optimizer': self.optimizer.state_dict(),
'current_loss': running_avg_loss
}
model_save_path = os.path.join(self.model_dir, 'model_%d_%d' % (iter, int(time.time())))
torch.save(state, model_save_path)
def setup_train(self, model_file_path=None):
self.model = Model(model_file_path)
params = list(self.model.encoder.parameters()) + \
list(self.model.section_encoder.parameters()) + \
list(self.model.sentence_filterer.parameters()) + \
list(self.model.decoder.parameters()) + \
list(self.model.reduce_state.parameters()) + \
list(self.model.section_reduce_state.parameters())
initial_lr = config.lr_coverage if config.is_coverage else config.lr
self.optimizer = AdagradCustom(params, lr=initial_lr, initial_accumulator_value=config.adagrad_init_acc)
start_iter, start_loss = 0, 0
if model_file_path is not None:
state = torch.load(model_file_path, map_location= lambda storage, location: storage)
start_iter = state['iter']
start_loss = state['current_loss']
if not config.is_coverage and not config.is_sentence_filtering:
self.optimizer.load_state_dict(state['optimizer'])
if use_cuda:
for state in self.optimizer.state.values():
for k, v in state.items():
if torch.is_tensor(v):
state[k] = v.cuda()
return start_iter, start_loss
def train_one_batch(self, batch):
enc_batch, enc_padding_mask, enc_lens, enc_batch_extend_vocab, extra_zeros, c_t_1, coverage, sent_lens = \
get_input_from_batch(batch, use_cuda)
dec_batch, dec_padding_mask, max_dec_len, dec_lens_var, target_batch = \
get_output_from_batch(batch, use_cuda)
self.optimizer.zero_grad()
encoder_outputs, encoder_hidden, max_encoder_output = self.model.encoder(enc_batch, enc_lens)
s_t_1 = self.model.reduce_state(encoder_hidden)
if config.use_maxpool_init_ctx:
c_t_1 = max_encoder_output
gamma = None
if config.is_sentence_filtering:
gamma, sent_dists = self.model.sentence_filterer(encoder_outputs, sent_lens)
section_outputs, section_hidden = self.model.section_encoder(s_t_1)
s_t_1 = self.model.section_reduce_state(section_hidden)
step_losses = []
for di in range(min(max_dec_len, config.max_dec_steps)):
y_t_1 = dec_batch[:, di] # Teacher forcing
final_dist, s_t_1, c_t_1, attn_dist, p_gen, coverage = self.model.decoder(y_t_1, s_t_1,
encoder_outputs, section_outputs, enc_padding_mask,
c_t_1, extra_zeros, enc_batch_extend_vocab,
coverage, gamma)
target = target_batch[:, di]
gold_probs = torch.gather(final_dist, 1, target.unsqueeze(1)).squeeze()
step_loss = -torch.log(gold_probs + config.eps)
if config.is_coverage:
step_coverage_loss = torch.sum(torch.min(attn_dist, coverage.view(*attn_dist.shape)), 1)
step_loss = step_loss + config.cov_loss_wt * step_coverage_loss
step_mask = dec_padding_mask[:, di]
step_loss = step_loss * step_mask
step_losses.append(step_loss)
sum_losses = torch.sum(torch.stack(step_losses, 1), 1)
batch_avg_loss = sum_losses/dec_lens_var
loss = torch.mean(batch_avg_loss)
if config.is_sentence_filtering:
sim_scores = torch.FloatTensor(batch.sim_scores)
if use_cuda:
sim_scores = sim_scores.cuda()
sent_filter_loss = F.binary_cross_entropy(sent_dists, sim_scores)
loss += config.sent_loss_wt * sent_filter_loss
loss.backward()
clip_grad_norm_(self.model.encoder.parameters(), config.max_grad_norm)
clip_grad_norm_(self.model.section_encoder.parameters(), config.max_grad_norm)
clip_grad_norm_(self.model.sentence_filterer.parameters(), config.max_grad_norm)
clip_grad_norm_(self.model.decoder.parameters(), config.max_grad_norm)
clip_grad_norm_(self.model.reduce_state.parameters(), config.max_grad_norm)
clip_grad_norm_(self.model.section_reduce_state.parameters(), config.max_grad_norm)
self.optimizer.step()
return loss.item()
def trainIters(self, n_iters, model_file_path=None):
iter, running_avg_loss = self.setup_train(model_file_path)
start = time.time()
while iter < n_iters:
print('\rBatch %d' % iter, end="")
batch = self.batcher.next_batch()
loss = self.train_one_batch(batch)
running_avg_loss = calc_running_avg_loss(loss, running_avg_loss, self.summary_writer, iter)
iter += 1
if iter % 5000 == 0:
self.summary_writer.flush()
print_interval = 10
if iter % print_interval == 0:
print(' steps %d, seconds for %d batch: %.2f , loss: %f' % (iter, print_interval,
time.time() - start, loss))
start = time.time()
if iter % 1000 == 0:
self.save_model(running_avg_loss, iter)
class BeamSearch(object):
def __init__(self, model_file_path):
model_name = os.path.basename(model_file_path)
self._decode_dir = os.path.join(config.log_root, 'decode_%s' % (model_name))
self._rouge_ref_dir = os.path.join(self._decode_dir, 'rouge_ref')
self._rouge_dec_dir = os.path.join(self._decode_dir, 'rouge_dec_dir')
for p in [self._decode_dir, self._rouge_ref_dir, self._rouge_dec_dir]:
if not os.path.exists(p):
os.mkdir(p)
self.vocab = Vocab(config.vocab_path, config.vocab_size)
self.batcher = Batcher(config.decode_data_path, self.vocab, mode='decode',
batch_size=config.beam_size, single_pass=True)
time.sleep(15)
self.model = Model(model_file_path, is_eval=True)
def sort_beams(self, beams):
return sorted(beams, key=lambda h: h.avg_log_prob, reverse=True)
def decode(self):
start = time.time()
counter = 0
batch = self.batcher.next_batch()
while batch is not None:
# Run beam search to get best Hypothesis
with torch.no_grad():
best_summary = self.beam_search(batch)
# Extract the output ids from the hypothesis and convert back to words
output_ids = [int(t) for t in best_summary.tokens[1:]]
decoded_words = data.outputids2words(output_ids, self.vocab,
(batch.art_oovs[0] if config.pointer_gen else None))
# Remove the [STOP] token from decoded_words, if necessary
try:
fst_stop_idx = decoded_words.index(data.STOP_DECODING)
decoded_words = decoded_words[:fst_stop_idx]
except ValueError:
decoded_words = decoded_words
print("===============SUMMARY=============")
print(' '.join(decoded_words))
original_abstract_sents = batch.original_abstracts_sents[0]
write_for_rouge(original_abstract_sents, decoded_words, counter,
self._rouge_ref_dir, self._rouge_dec_dir)
counter += 1
if counter % 1000 == 0:
print('%d example in %d sec'%(counter, time.time() - start))
start = time.time()
batch = self.batcher.next_batch()
print("Decoder has finished reading dataset for single_pass.")
print("Now starting ROUGE eval...")
results_dict = rouge_eval(self._rouge_ref_dir, self._rouge_dec_dir)
rouge_log(results_dict, self._decode_dir)
def beam_search(self, batch):
#batch should have only one example
enc_batch, enc_padding_mask, enc_lens, enc_batch_extend_vocab, extra_zeros, c_t_0, coverage_t_0, sent_lens = \
get_input_from_batch(batch, use_cuda)
encoder_outputs, encoder_hidden, max_encoder_output = self.model.encoder(enc_batch, enc_lens)
s_t_0 = self.model.reduce_state(encoder_hidden)
if config.use_maxpool_init_ctx:
c_t_0 = max_encoder_output
gamma = None
if config.is_sentence_filtering:
gamma, sent_dists = self.model.sentence_filterer(encoder_outputs, sent_lens)
section_outputs, section_hidden = self.model.section_encoder(s_t_0)
s_t_0 = self.model.section_reduce_state(section_hidden)
dec_h, dec_c = s_t_0 # 1 x 2*hidden_size
dec_h = dec_h.squeeze()
dec_c = dec_c.squeeze()
#decoder batch preparation, it has beam_size example initially everything is repeated
beams = [Beam(tokens=[self.vocab.word2id(data.START_DECODING)],
log_probs=[0.0],
state=(dec_h[0], dec_c[0]),
context = c_t_0[0],
coverage=(coverage_t_0[0] if config.is_coverage else None))
for _ in range(config.beam_size)]
results = []
steps = 0
while steps < config.max_dec_steps and len(results) < config.beam_size:
latest_tokens = [h.latest_token for h in beams]
latest_tokens = [t if t < self.vocab.size() else self.vocab.word2id(data.UNKNOWN_TOKEN) \
for t in latest_tokens]
y_t_1 = Variable(torch.LongTensor(latest_tokens))
if use_cuda:
y_t_1 = y_t_1.cuda()
all_state_h =[]
all_state_c = []
all_context = []
for h in beams:
state_h, state_c = h.state
all_state_h.append(state_h)
all_state_c.append(state_c)
all_context.append(h.context)
s_t_1 = (torch.stack(all_state_h, 0).unsqueeze(0), torch.stack(all_state_c, 0).unsqueeze(0))
c_t_1 = torch.stack(all_context, 0)
coverage_t_1 = None
if config.is_coverage:
all_coverage = []
for h in beams:
all_coverage.append(h.coverage)
coverage_t_1 = torch.stack(all_coverage, 0)
final_dist, s_t, c_t, attn_dist, p_gen, coverage_t = self.model.decoder(y_t_1, s_t_1,
encoder_outputs, section_outputs, enc_padding_mask,
c_t_1, extra_zeros, enc_batch_extend_vocab, coverage_t_1, gamma)
topk_log_probs, topk_ids = torch.topk(final_dist, config.beam_size * 2)
dec_h, dec_c = s_t
dec_h = dec_h.squeeze()
dec_c = dec_c.squeeze()
all_beams = []
num_orig_beams = 1 if steps == 0 else len(beams)
for i in range(num_orig_beams):
h = beams[i]
state_i = (dec_h[i], dec_c[i])
context_i = c_t[i]
coverage_i = (coverage_t[i] if config.is_coverage else None)
for j in range(config.beam_size * 2): # for each of the top 2*beam_size hyps:
new_beam = h.extend(token=topk_ids[i, j].item(),
log_prob=topk_log_probs[i, j].item(),
state=state_i,
context=context_i,
coverage=coverage_i)
all_beams.append(new_beam)
beams = []
for h in self.sort_beams(all_beams):
if h.latest_token == self.vocab.word2id(data.STOP_DECODING):
if steps >= config.min_dec_steps:
results.append(h)
else:
beams.append(h)
if len(beams) == config.beam_size or len(results) == config.beam_size:
break
steps += 1
if len(results) == 0:
results = beams
beams_sorted = self.sort_beams(results)
return beams_sorted[0]
class Evaluate(object):
def __init__(self, model_file_path):
self.vocab = Vocab(config.vocab_path, config.vocab_size)
self.batcher = Batcher(config.eval_data_path,
self.vocab,
mode='eval',
batch_size=config.batch_size,
single_pass=True)
time.sleep(15)
model_name = os.path.basename(model_file_path)
eval_dir = os.path.join(config.log_root, 'eval_%s' % (model_name))
if not os.path.exists(eval_dir):
os.mkdir(eval_dir)
self.summary_writer = tf.summary.FileWriter(eval_dir)
self.model = Model(model_file_path, is_eval=True)
def eval_one_batch(self, batch):
enc_batch, enc_padding_mask, enc_lens, enc_batch_extend_vocab, extra_zeros, c_t_1, coverage = \
get_input_from_batch(batch, use_cuda)
dec_batch, dec_padding_mask, max_dec_len, dec_lens_var, target_batch = \
get_output_from_batch(batch, use_cuda)
encoder_outputs, encoder_hidden, max_encoder_output = self.model.encoder(
enc_batch, enc_lens)
s_t_1 = self.model.reduce_state(encoder_hidden)
if config.use_maxpool_init_ctx:
c_t_1 = max_encoder_output
step_losses = []
for di in range(min(max_dec_len, config.max_dec_steps)):
y_t_1 = dec_batch[:, di] # Teacher forcing
final_dist, s_t_1, c_t_1, attn_dist, p_gen, coverage = self.model.decoder(
y_t_1, s_t_1, encoder_outputs, enc_padding_mask, c_t_1,
extra_zeros, enc_batch_extend_vocab, coverage)
target = target_batch[:, di]
gold_probs = torch.gather(final_dist, 1,
target.unsqueeze(1)).squeeze()
step_loss = -torch.log(gold_probs + config.eps)
if config.is_coverage:
step_coverage_loss = torch.sum(torch.min(attn_dist, coverage),
1)
step_loss = step_loss + config.cov_loss_wt * step_coverage_loss
step_mask = dec_padding_mask[:, di]
step_loss = step_loss * step_mask
step_losses.append(step_loss)
sum_step_losses = torch.sum(torch.stack(step_losses, 1), 1)
batch_avg_loss = sum_step_losses / dec_lens_var
loss = torch.mean(batch_avg_loss)
return loss.data[0]
def run_eval(self):
running_avg_loss, iter = 0, 0
start = time.time()
batch = self.batcher.next_batch()
while batch is not None:
loss = self.eval_one_batch(batch)
running_avg_loss = calc_running_avg_loss(loss.item(),
running_avg_loss,
self.summary_writer, iter)
iter += 1
if iter % 100 == 0:
self.summary_writer.flush()
print_interval = 1
if iter % print_interval == 0:
print('steps %d, seconds for %d batch: %.2f , loss: %f' %
(iter, print_interval, time.time() - start,
running_avg_loss))
start = time.time()
batch = self.batcher.next_batch()