# Experiment builder
data = dataset.MiniImageNetDataSet(batch_size=batch_size,
classes_per_set=classes_per_set,
samples_per_class=samples_per_class,
shuffle_classes=True)
experiment = ExperimentBuilder(data)
one_shot_miniImagenet, losses, c_error_opt_op, init = experiment.build_experiment(
batch_size, classes_per_set, samples_per_class)
total_epochs = 120
total_train_batches = 1000
total_val_batches = int(250 * sp)
total_test_batches = int(250 * sp)
logs = "{}way{}shot learning problems, with {} tasks per task batch".format(
classes_per_set, samples_per_class, batch_size)
save_statistics(experiment_name, ["Experimental details: {}".format(logs)])
save_statistics(experiment_name, [
"epoch", "train_c_loss", "train_c_accuracy", "val_loss",
"val_accuracy", "test_c_loss", "test_c_accuracy", "learning_rate"
])
# Experiment initialization and running
config = tf.ConfigProto()
config.gpu_options.allow_growth = True
config.gpu_options.visible_device_list = "0"
with tf.Session(config=config) as sess:
sess.run(init)
saver = tf.train.Saver(max_to_keep=5)
if continue_from_epoch != -1: #load checkpoint if needed
print("Loading from checkpoint")
checkpoint = "saved_models/{}_{}.ckpt".format(
experiment = ExperimentBuilder(data)
one_shot_omniglot, losses, c_error_opt_op, init = experiment.build_experiment(
batch_size, classes_per_set, samples_per_class, fce)
total_train_batches = 18 #1000
total_val_batches = 3 #1000
total_test_batches = 3 #1000
#total_test_c_loss_mean, total_test_c_loss_std, total_test_accuracy_mean, total_test_accuracy_std \
#= -1, -1, -1, -1
saved_models_filepath, logs_filepath = build_experiment_folder(experiment_name)
save_statistics(logs_filepath, [
"epoch", "total_train_c_loss_mean", "total_train_c_loss_std",
"total_train_accuracy_mean", "total_train_accuracy_std",
"total_val_c_loss_mean", "total_val_c_loss_std", "total_val_accuracy_mean",
"total_val_accuracy_std", "total_test_c_loss_mean",
"total_test_c_loss_std", "total_test_accuracy_mean",
"total_test_accuracy_std"
],
create=True)
# Experiment initialization and running
with tf.Session() as sess:
sess.run(init)
train_saver = tf.train.Saver()
val_saver = tf.train.Saver()
if continue_from_epoch != -1: #load checkpoint if needed
checkpoint = "saved_models/{}_{}.ckpt".format(experiment_name,
continue_from_epoch)
variables_to_restore = []
for var in tf.get_collection(tf.GraphKeys.GLOBAL_VARIABLES):
else:
print("Unsupported dataset.")
assert False
experiment = ExperimentBuilder(data)
one_shot, losses, c_error_opt_op, init = experiment.build_experiment(batch_size,\
classes_train, classes_test, samples_per_class, queries_per_class, channels, image_size, fce, network_name)
#
total_epochs = 300
total_train_batches = 1000
total_val_batches = 100
total_test_batches = 250
save_statistics(experiment_name, [
"epoch", "train_c_loss", "train_c_accuracy", "val_loss", "val_accuracy",
"test_c_loss", "test_c_accuracy"
])
# Experiment initialization and running
with tf.Session() as sess:
sess.run(init)
saver = tf.train.Saver()
if continue_from_epoch != -1: #load checkpoint if needed
checkpoint = "{}_{}.ckpt".format(experiment_name, continue_from_epoch)
variables_to_restore = []
for var in tf.get_collection(tf.GraphKeys.VARIABLES):
print(var)
variables_to_restore.append(var)
tf.logging.info('Fine-tuning from %s' % checkpoint)