Skip to content

Use best model for decoding #23

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

Use best model for decoding #23

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
ff8649b
Use best model for decoding
pltrdy Jun 8, 2017
cfa8798
Best model loading less verbose
pltrdy Jun 8, 2017
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
2 changes: 1 addition & 1 deletion decode.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -51,7 +51,7 @@ def __init__(self, model, batcher, vocab):
self._sess = tf.Session(config=util.get_config())

# Load an initial checkpoint to use for decoding
ckpt_path = util.load_ckpt(self._saver, self._sess)
ckpt_path = util.load_ckpt(self._saver, self._sess, load_best=True)

if FLAGS.single_pass:
# Make a descriptive decode directory name
Expand Down
15 changes: 13 additions & 2 deletions util.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -27,12 +27,23 @@ def get_config():
config.gpu_options.allow_growth=True
return config

def load_ckpt(saver, sess):
def load_ckpt(saver, sess, load_best=False):
"""Load checkpoint from the train directory and restore it to saver and sess, waiting 10 secs in the case of failure. Also returns checkpoint name."""
while True:
try:
ckpt_state = None
if load_best:
eval_dir = os.path.join(FLAGS.log_root, "eval")
if os.path.exists(eval_dir):
try:
ckpt_state = tf.train.get_checkpoint_state(eval_dir, latest_filename="checkpoint_best")
except ValueError:
pass

train_dir = os.path.join(FLAGS.log_root, "train")
ckpt_state = tf.train.get_checkpoint_state(train_dir)
if ckpt_state is None:
ckpt_state = tf.train.get_checkpoint_state(train_dir)

tf.logging.info('Loading checkpoint %s', ckpt_state.model_checkpoint_path)
saver.restore(sess, ckpt_state.model_checkpoint_path)
return ckpt_state.model_checkpoint_path
Expand Down