def train_and_evaluate(hparams):
"""Run the training and evaluate using the high level API."""
train_input = lambda: model.input_fn(hparams.train_files,
num_epochs=hparams.num_epochs,
batch_size=hparams.train_batch_size)
# Don't shuffle evaluation data.
eval_input = lambda: model.input_fn(
hparams.eval_files, batch_size=hparams.eval_batch_size, shuffle=False)
train_spec = tf.estimator.TrainSpec(train_input,
max_steps=hparams.train_steps)
exporter = tf.estimator.FinalExporter(
'census', model.SERVING_FUNCTIONS[hparams.export_format])
eval_spec = tf.estimator.EvalSpec(eval_input,
steps=hparams.eval_steps,
exporters=[exporter],
name='census-eval')
model_fn = model.generate_model_fn(
embedding_size=hparams.embedding_size,
# Construct layers sizes with exponential decay.
hidden_units=[
max(2, int(hparams.first_layer_size * hparams.scale_factor**i))
for i in range(hparams.num_layers)
],
learning_rate=hparams.learning_rate)
estimator = tf.estimator.Estimator(model_fn=model_fn,
model_dir=hparams.job_dir)
tf.estimator.train_and_evaluate(estimator, train_spec, eval_spec)
def run_experiment(hparams):
"""Run the training and evaluate using the high level API"""
train_input = lambda: model.input_fn(hparams.train_files,
num_epochs=hparams.num_epochs,
batch_size=hparams.train_batch_size)
# Don't shuffle evaluation data
eval_input = lambda: model.input_fn(
hparams.eval_files, batch_size=hparams.eval_batch_size, shuffle=False)
# TODO: How should these train/eval spec's be further adjusted for using tf.estimator.train_and_evaluate()?
train_spec = tf.estimator.TrainSpec(train_input,
max_steps=hparams.train_steps)
exporter = tf.estimator.FinalExporter(
'census', model.SERVING_FUNCTIONS[hparams.export_format])
eval_spec = tf.estimator.EvalSpec(eval_input,
steps=hparams.eval_steps,
exporters=[exporter],
name='census-eval')
model_fn = model.generate_model_fn(
embedding_size=hparams.embedding_size,
# Construct layers sizes with exponetial decay
hidden_units=[
max(2, int(hparams.first_layer_size * hparams.scale_factor**i))
for i in range(hparams.num_layers)
],
learning_rate=hparams.learning_rate)
# TODO: unclear what config settings are needed for my model
config = tpu_config.RunConfig(
master=tpu_grpc_url,
evaluation_master=tpu_grpc_url,
model_dir=FLAGS.model_dir,
cluster=tpu_cluster_resolver,
tpu_config=tpu_config.TPUConfig(
iterations_per_loop=FLAGS.iterations_per_loop,
num_shards=FLAGS.num_cores))
estimator = tpu_estimator.Estimator(
use_tpu=True,
model_fn=model_fn,
model_dir=hparams.job_dir,
config=run_config,
# train_batch_size --> not being passed as I believe this should be handled in the train_spec...
)
tf.estimator.train_and_evaluate(estimator, train_spec, eval_spec)
def train_and_evaluate(args):
"""Run the training and evaluate using the high level API."""
def train_input():
"""Input function returning batches from the training
data set from training.
"""
return model.input_fn(args.train_files,
num_epochs=args.num_epochs,
batch_size=args.train_batch_size)
def eval_input():
"""Input function returning the entire validation data
set for evaluation. Shuffling is not required.
"""
return model.input_fn(args.eval_files,
batch_size=args.eval_batch_size,
shuffle=False)
train_spec = tf.estimator.TrainSpec(train_input,
max_steps=args.train_steps)
exporter = tf.estimator.FinalExporter(
'census', model.SERVING_FUNCTIONS[args.export_format])
eval_spec = tf.estimator.EvalSpec(eval_input,
steps=args.eval_steps,
exporters=[exporter],
name='census-eval')
model_fn = model.generate_model_fn(
embedding_size=args.embedding_size,
# Construct layers sizes with exponential decay.
hidden_units=[
max(2, int(args.first_layer_size * args.scale_factor**i))
for i in range(args.num_layers)
],
learning_rate=args.learning_rate)
estimator = tf.estimator.Estimator(model_fn=model_fn,
model_dir=args.job_dir)
tf.estimator.train_and_evaluate(estimator, train_spec, eval_spec)
def train_and_evaluate(args):
"""Run the training and evaluate using the high level API."""
train_input = lambda: model.input_fn(
args.train_files,
num_epochs=args.num_epochs,
batch_size=args.train_batch_size
)
# Don't shuffle evaluation data.
eval_input = lambda: model.input_fn(
args.eval_files,
batch_size=args.eval_batch_size,
shuffle=False
)
train_spec = tf.estimator.TrainSpec(
train_input, max_steps=args.train_steps)
exporter = tf.estimator.FinalExporter(
'census', model.SERVING_FUNCTIONS[args.export_format])
eval_spec = tf.estimator.EvalSpec(
eval_input,
steps=args.eval_steps,
exporters=[exporter],
name='census-eval')
model_fn = model.generate_model_fn(
embedding_size=args.embedding_size,
# Construct layers sizes with exponential decay.
hidden_units=[
max(2, int(args.first_layer_size * args.scale_factor**i))
for i in range(args.num_layers)
],
learning_rate=args.learning_rate)
estimator = tf.estimator.Estimator(
model_fn=model_fn, model_dir=args.job_dir)
tf.estimator.train_and_evaluate(estimator, train_spec, eval_spec)