experiments
/
STT
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Add trial pruning to lm_optimizer.py

This commit is contained in:
Caleb Moses 2020-03-18 10:14:47 +13:00
parent 29a2ac37f0
commit c9e6cbc958
2 changed files with 16 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
DeepSpeech.py
View File

@ -75,7 +75,7 @@ def create_overlapping_windows(batch_x):
def dense(name, x, units, dropout_rate=None, relu=True):
with tfv1.variable_scope(name):
with tfv1.variable_scope(name, reuse=tf.AUTO_REUSE):
bias = variable_on_cpu('bias', [units], tfv1.zeros_initializer())
weights = variable_on_cpu('weights', [x.shape[-1], units], tfv1.keras.initializers.VarianceScaling(scale=1.0, mode="fan_avg", distribution="uniform"))
@ -91,7 +91,7 @@ def dense(name, x, units, dropout_rate=None, relu=True):
def rnn_impl_lstmblockfusedcell(x, seq_length, previous_state, reuse):
with tfv1.variable_scope('cudnn_lstm/rnn/multi_rnn_cell/cell_0'):
with tfv1.variable_scope('cudnn_lstm/rnn/multi_rnn_cell/cell_0', reuse=tf.AUTO_REUSE):
fw_cell = tf.contrib.rnn.LSTMBlockFusedCell(Config.n_cell_dim,
forget_bias=0,
reuse=reuse,
@ -133,7 +133,7 @@ rnn_impl_cudnn_rnn.cell = None
def rnn_impl_static_rnn(x, seq_length, previous_state, reuse):
with tfv1.variable_scope('cudnn_lstm/rnn/multi_rnn_cell'):
with tfv1.variable_scope('cudnn_lstm/rnn/multi_rnn_cell', reuse=tf.AUTO_REUSE):
# Forward direction cell:
fw_cell = tfv1.nn.rnn_cell.LSTMCell(Config.n_cell_dim,
forget_bias=0,

14
lm_optimizer.py
View File

@ -29,10 +29,22 @@ def objective(trial):
FLAGS.lm_beta = trial.suggest_uniform('lm_beta', 0, FLAGS.lm_beta_max)
tfv1.reset_default_graph()
samples = evaluate(FLAGS.test_files.split(','), create_model)
is_character_based = trial.study.user_attrs['is_character_based']
samples = []
for step, test_file in enumerate(FLAGS.test_files.split(',')):
current_samples = evaluate([test_file], create_model, try_loading)
samples += current_samples
# Report intermediate objective value.
wer, cer = wer_cer_batch(current_samples)
trial.report(cer if is_character_based else wer, step)
# Handle pruning based on the intermediate value.
if trial.should_prune():
raise optuna.exceptions.TrialPruned()
wer, cer = wer_cer_batch(samples)
return cer if is_character_based else wer