Skip to content

Add optional max_length arg in batch_sentences #111

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
vliu15 wants to merge 2 commits into
base: master
Choose a base branch
from
Open

Add optional max_length arg in batch_sentences #111

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
5fa2636
max batch length argument
Oct 2, 2018
fbff27f
Create tqdm progress bar
vliu15 Oct 31, 2018
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
14 changes: 10 additions & 4 deletions bilm/data.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -205,14 +205,17 @@ def __init__(self, lm_vocab_file: str, max_token_length: int):
)
self._max_token_length = max_token_length

def batch_sentences(self, sentences: List[List[str]]):
def batch_sentences(self, sentences: List[List[str]], max_length=None):
'''
Batch the sentences as character ids
Each sentence is a list of tokens without <s> or </s>, e.g.
[['The', 'first', 'sentence', '.'], ['Second', '.']]
'''
n_sentences = len(sentences)
max_length = max(len(sentence) for sentence in sentences) + 2
if max_length == None:
max_length = max(len(sentence) for sentence in sentences) + 2
else:
max_length += 2

X_char_ids = np.zeros(
(n_sentences, max_length, self._max_token_length),
Expand Down Expand Up @@ -240,14 +243,17 @@ def __init__(self, lm_vocab_file: str):
'''
self._lm_vocab = Vocabulary(lm_vocab_file)

def batch_sentences(self, sentences: List[List[str]]):
def batch_sentences(self, sentences: List[List[str]], max_length=None):
'''
Batch the sentences as character ids
Each sentence is a list of tokens without <s> or </s>, e.g.
[['The', 'first', 'sentence', '.'], ['Second', '.']]
'''
n_sentences = len(sentences)
max_length = max(len(sentence) for sentence in sentences) + 2
if max_length == None:
max_length = max(len(sentence) for sentence in sentences) + 2
else:
max_length += 2

X_ids = np.zeros((n_sentences, max_length), dtype=np.int64)

Expand Down
4 changes: 4 additions & 0 deletions bilm/training.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -11,6 +11,7 @@
import numpy as np

from tensorflow.python.ops.init_ops import glorot_uniform_initializer
from tqdm import tqdm

from .data import Vocabulary, UnicodeCharsVocabulary, InvalidNumberOfCharacters

Expand Down Expand Up @@ -835,6 +836,7 @@ def train(options, data, n_gpus, tf_save_dir, tf_log_dir,

t1 = time.time()
data_gen = data.iter_batches(batch_size * n_gpus, unroll_steps)
pbar = tqdm(total=n_batches_total)
for batch_no, batch in enumerate(data_gen, start=1):

# slice the input in the batch for the feed_dict
Expand Down Expand Up @@ -893,7 +895,9 @@ def train(options, data, n_gpus, tf_save_dir, tf_log_dir,

if batch_no == n_batches_total:
# done training!
pbar.close()
break
pbar.update(1)


def clip_by_global_norm_summary(t_list, clip_norm, norm_name, variables):
Expand Down