def main(_):
console.start('RNN task')
# Configurations
th = NlsHub(as_global=True)
th.memory_depth = 10
th.num_blocks = 1
th.multiplier = 8
th.hidden_dim = th.memory_depth * th.multiplier
th.num_steps = 32
th.epoch = 100000
th.batch_size = 32
th.learning_rate = 1e-4
th.validation_per_round = 20
th.print_cycle = 0
# th.train = False
th.smart_train = True
th.max_bad_apples = 4
th.lr_decay = 0.6
th.early_stop = True
th.idle_tol = 20
th.save_mode = SaveMode.ON_RECORD
th.warm_up_thres = 1
th.at_most_save_once_per_round = True
th.overwrite = True # Default: False
th.export_note = True
th.summary = True
th.monitor_preact = False
th.save_model = True
th.allow_growth = False
th.gpu_memory_fraction = 0.4
description = '0'
th.mark = 'rnn-{}x({}x{})-{}steps-{}'.format(
th.num_blocks, th.memory_depth, th.multiplier, th.num_steps, description)
# Get model
model = model_lib.rnn0(th)
# Load data
train_set, val_set, test_set = load_wiener_hammerstein(
th.data_dir, depth=th.memory_depth, validation_size=2000)
assert isinstance(train_set, DataSet)
assert isinstance(val_set, DataSet)
assert isinstance(test_set, DataSet)
# Train or evaluate
if th.train:
model.nn.train(train_set, validation_set=val_set, trainer_hub=th)
else:
console.show_status('Evaluating ...')
model.evaluate(train_set, start_at=th.memory_depth)
model.evaluate(val_set, start_at=th.memory_depth)
model.evaluate(test_set, start_at=th.memory_depth)
# End
console.end()
def main(_):
console.start('LSTM task')
# Configurations
th = core.th
th.model = models.lstm_test
# th.model = models.lstm
th.num_blocks = 1
th.memory_depth = 3
th.hidden_dim = 100
th.epoch = 50000
th.learning_rate = 1e-4
th.batch_size = 512
th.num_steps = 100
th.val_preheat = 500
th.validation_per_round = 0
th.validate_cycle = 0
th.print_cycle = 2
th.train = True
th.smart_train = False
th.max_bad_apples = 4
th.lr_decay = 0.5
th.save_model = True
th.overwrite = True
th.export_note = True
th.summary = False
th.monitor = False
description = ''
th.mark = '{}x{}{}'.format(th.num_blocks, th.memory_depth, description)
core.activate()
def main(_):
# Configuration
# FLAGS.train = False
# FLAGS.smart_train = True
FLAGS.overwrite = True
FLAGS.summary = True
FLAGS.save_model = False
FLAGS.snapshot = False
MEMORY_DEPTH = 1
EPOCH = 2
# Start
console.start('rnn_task')
# Initiate model
model = rnn_models.vanilla_RNN('rnn00')
# Load data
train_set, val_set, test_set = load_wiener_hammerstein(
r'../data/wiener_hammerstein/whb.tfd', depth=MEMORY_DEPTH)
assert isinstance(train_set, DataSet)
assert isinstance(val_set, DataSet)
assert isinstance(test_set, DataSet)
# Train or evaluate
if FLAGS.train:
pass
else:
console.show_status('Evaluating ...')
# End
console.end()
def main(_):
console.start('GPAT Classification task (MLP)')
# Configurations
th = core.th
th.model = models.mlp
th.num_blocks = 2
th.hidden_dim = 500
th.actype1 = 'relu'
th.idle_tol = 30
th.epoch = 500
th.learning_rate = 1e-3
th.batch_size = 64
th.validation_per_round = 1
th.print_cycle = 1
th.shuffle = True
# th.train = False
th.smart_train = False
th.max_bad_apples = 4
th.lr_decay = 0.6
th.save_model = True
th.overwrite = True
th.export_note = True
th.summary = True
th.monitor = False
description = 'demo'
th.mark = 'mlp_{}x{}{}'.format(th.hidden_dim, th.num_blocks, description)
core.activate()
def main(_):
console.start('mlp task')
# Configurations
th = NlsHub(as_global=True)
th.memory_depth = 6
th.num_blocks = 2
th.multiplier = 2
th.hidden_dim = th.memory_depth * th.multiplier
# th.actype1 = 'lrelu' # Default: relu
th.epoch = 10
th.batch_size = 32
th.learning_rate = 1e-4
th.validation_per_round = 5
th.print_cycle = 100
th.train = True
# th.smart_train = True
# th.max_bad_apples = 4
# th.lr_decay = 0.6
th.early_stop = True
th.idle_tol = 20
th.save_mode = SaveMode.NAIVE
# th.warm_up_thres = 1
# th.at_most_save_once_per_round = True
th.overwrite = True
th.export_note = True
th.summary = False
# th.monitor = True
th.save_model = True
th.allow_growth = False
th.gpu_memory_fraction = 0.40
description = '0'
th.mark = 'mlp-{}x({}x{})-{}'.format(th.num_blocks, th.memory_depth,
th.multiplier, description)
# Get model
model = nlsf_model_lib.mlp_00(th)
# Load data
train_set, val_set, test_set = load_data(th.data_dir,
depth=th.memory_depth)
assert isinstance(train_set, DataSet)
assert isinstance(val_set, DataSet)
assert isinstance(test_set, DataSet)
# Train or evaluate
if th.train:
model.nn.train(train_set, validation_set=val_set, trainer_hub=th)
else:
console.show_status('Evaluating ...')
model.evaluate(train_set, start_at=th.memory_depth)
model.evaluate(val_set, start_at=th.memory_depth)
model.evaluate(test_set, start_at=th.memory_depth, plot=True)
# End
console.end()
def main(_):
console.suppress_logging()
FLAGS.train = True
FLAGS.overwrite = True
# Start
console.start('MNIST VANILLA VAE')
# Get model
model = models.vanilla('vanilla_00')
if FLAGS.train:
mnist = load_mnist('../../data/MNIST',
flatten=True,
validation_size=0,
one_hot=True)
model.train(training_set=mnist[pedia.training],
epoch=1000,
batch_size=128,
print_cycle=50,
snapshot_cycle=200)
else:
samples = model.generate(sample_num=16)
console.show_status('{} samples generated'.format(samples.shape[0]))
imtool.gan_grid_plot(samples, show=True)
# End
console.end()
def main(_):
console.start('WHBM task (MLP model)')
# Configurations
th = core.th
th.model = models.mlp
th.num_blocks = 2
th.memory_depth = 80
multiplier = 4
th.hidden_dim = th.memory_depth * multiplier
th.actype1 = 'relu'
th.epoch = 50000
th.learning_rate = 1e-4
th.batch_size = 64
th.validation_per_round = 5
th.print_cycle = 50
th.shuffle = True
# th.train = False
th.smart_train = True
th.max_bad_apples = 4
th.lr_decay = 0.5
th.save_model = True
th.overwrite = True
th.export_note = True
th.summary = True
th.monitor = False
description = ''
th.mark = '{}x[{}x{}]{}'.format(th.num_blocks, th.memory_depth, multiplier,
description)
core.activate()
def main(_):
console.suppress_logging()
FLAGS.train = True
FLAGS.overwrite = False
# Start
console.start("MNIST DCGAN DEMO")
# Get model
model = models.dcgan('dcgan_002')
# model = models.dcgan_h3_rs_nbn()
# Train or test
if FLAGS.train:
mnist = load_mnist('../../data/MNIST', flatten=False, validation_size=0,
one_hot=True)
model.train(training_set=mnist[pedia.training], epoch=10, batch_size=128,
print_cycle=20, snapshot_cycle=200, D_times=1, G_times=1)
else:
samples = model.generate(sample_num=16)
console.show_status('{} samples generated'.format(samples.shape[0]))
imtool.gan_grid_plot(samples, show=True)
# End
console.end()
def main(_):
console.start('CNN task ')
# Configurations
th = core.th
# th.model = models.conv_test
th.model = models.conv_2d_test
th.actype1 = 'relu'
th.patience = 100
th.epoch = 5000
th.learning_rate = 1e-3
th.batch_size = 64
th.validation_per_round = 1
th.print_cycle = 10
th.shuffle = False
# th.train = False
th.smart_train = True
th.max_bad_apples = 4
th.lr_decay = 0.6
th.save_model = True
th.overwrite = True
th.export_note = True
th.summary = True
th.monitor = False
th.allow_growth = True
# th.gpu_memory_fraction =
description = 'conv_2d_add_noise'
th.mark = 'cnn_{}x{}{}'.format(th.hidden_dim, th.num_blocks, description)
core.activate()
def main(_):
console.suppress_logging()
FLAGS.overwrite = True
FLAGS.train = True
# Start
console.start()
# Get or define model
model = models.vanilla('vanilla_nov9_02_h2_c', bn=False)
# model = models.dcgan('dcgan_c00')
# model = models.vanilla_h3_rs_nbn('vanilla_nov9_01_h3_nbn_opdef')
# return
# Train or test
if FLAGS.train:
mnist = load_mnist('../../data/MNIST',
flatten=True,
validation_size=0,
one_hot=True)
model.train(training_set=mnist[pedia.training],
epoch=1000,
batch_size=128,
print_cycle=20,
snapshot_cycle=150,
sample_num=25)
else:
samples = model.generate(sample_num=16)
console.show_status('{} samples generated'.format(samples.shape[0]))
imtool.gan_grid_plot(samples, show=True)
# End
console.end()
def main(_):
console.suppress_logging()
# Setting
FLAGS.train = False
FLAGS.overwrite = True
# FLAGS.shuffle = True
show_false_pred = True
# Start
console.start('CIFAR-10 CONV DEMO')
# Get model
# model = models.deep_conv('dper_do0p5_reg0p2')
model = models.deep_conv('001_pre_bn')
# Train or test
cifar10 = load_cifar10('../../data/CIFAR-10',
flatten=False,
validation_size=5000,
one_hot=True)
if FLAGS.train:
model.train(training_set=cifar10[pedia.training],
validation_set=cifar10[pedia.validation],
epoch=120,
batch_size=64,
print_cycle=100)
else:
model.evaluate_model(cifar10[pedia.test], with_false=show_false_pred)
# End
console.end()
def main(_):
console.suppress_logging()
FLAGS.train = TRAIN
FLAGS.overwrite = OVERWRITE
console.start('EXP OB 01')
# Define system
system = define_system()
# Generate data
training_set, validation_set, test_set = generate_data(system)
if len(SYS_LOCK_ORDERS) == 1:
homogeneous_check(system, SYS_LOCK_ORDERS[0], training_set.signls[0],
training_set.responses[0])
# Identification
# .. wiener
wiener = Wiener(degree=WN_DEGREE, memory_depth=WN_MEN_DEPTH)
if WIENER_ON: wiener.identify(training_set, validation_set)
# .. vn
homo_strs = NN_HOMO_STRS
vns = collections.OrderedDict()
for homo_str in homo_strs:
console.show_status('Volterra Net h**o-strength = {:.2f}'.format(homo_str))
vn = init_vn('vn_{:.2f}{}'.format(homo_str, POSTFIX), homo_str=homo_str)
vns[homo_str] = vn
if FLAGS.train:
vn.identify(training_set, validation_set,
batch_size=50, print_cycle=100, epoch=EPOCH)
# Verification
verify(vns, wiener, system, test_set)
# End
console.end()
def main(_):
console.start('MLP task')
# Configurations
th = core.th
th.model = models.mlp
th.fc_dims = [800, 500]
th.actype1 = 'relu'
th.epoch = 50
th.learning_rate = 1e-5
th.batch_size = 64
th.validation_per_round = 2
th.print_cycle = 20
th.shuffle = True
# th.train = False
th.smart_train = False
th.max_bad_apples = 4
th.lr_decay = 0.6
th.save_model = True
th.overwrite = True
th.export_note = True
th.summary = True
th.monitor = False
description = ''
th.mark = 'mlp_{}{}'.format(ms(th.fc_dims), description)
export_false = True
core.activate(export_false=export_false)
def main(_):
console.suppress_logging()
FLAGS.train = True
FLAGS.overwrite = False
# Start
console.start('CIFAR-10 DCGAN')
# Get model
model = models.dcgan('dcgan_00')
if FLAGS.train:
cifar10 = load_cifar10('../../data/CIFAR-10',
validation_size=0,
one_hot=True)
model.train(training_set=cifar10[pedia.training],
epoch=20000,
batch_size=128,
print_cycle=20,
snapshot_cycle=2000)
else:
samples = model.generate(sample_num=16)
console.show_status('{} samples generated'.format(samples.shape[0]))
imtool.gan_grid_plot(samples, show=True)
# End
console.end()
def main(_):
console.suppress_logging()
FLAGS.train = False
FLAGS.overwrite = True
show_false = True
flatten = False
# Start
console.start('MNIST DEMO')
# model = models.vanilla('003_post')
model = models.deep_conv('dc_000')
mnist = load_mnist('../../data/MNIST',
flatten=flatten,
validation_size=5000,
one_hot=True)
# Train or test
if FLAGS.train:
model.train(training_set=mnist[pedia.training],
validation_set=mnist[pedia.validation],
epoch=30,
batch_size=100,
print_cycle=50)
else:
model.evaluate_model(mnist[pedia.test], with_false=show_false)
# End
console.end()
def main(_):
FLAGS.overwrite = False
FLAGS.train = True
play = True
console.suppress_logging()
console.start('TD Gomoku - vanilla')
with tf.Graph().as_default():
model = models.mlp00('mlp00_00')
with tf.Graph().as_default():
opponent = models.mlp00('mlp00_00')
game = Game()
if FLAGS.train:
model.train(game,
episodes=500000,
print_cycle=20,
snapshot_cycle=300,
match_cycle=2000,
rounds=5,
rate_thresh=1.0,
shadow=opponent,
save_cycle=200,
snapshot_function=game.snapshot)
else:
if play:
TkBoard(player=model).show()
else:
model.compete(game, rounds=100, opponent=opponent)
console.end()
def main(_):
console.start('Task CNN (MEMBRANE)')
th = core.th
th.job_dir = './records_unet_alpha'
th.model = models.unet
th.suffix = '01'
th.batch_size = 2
th.learning_rate = 1e-4
th.epoch = 3
th.early_stop = True
th.patience = 5
th.print_cycle = 1
th.validation_per_round = 4
th.val_batch_size = 10
th.validate_train_set = True
th.export_tensors_upon_validation = True
# th.probe_cycle = 1
th.warm_up = False
th.save_model = True
th.overwrite = True
th.gather_note = True
th.summary = False
th.warm_up_thres = 0
#
th.train = False
th.mark = 'unet_{}'.format('x')
core.activate()
def main(_):
console.start('{} on TIMIT task'.format(model_name.upper()))
th = core.th
# ---------------------------------------------------------------------------
# 1. folder/file names and device
# ---------------------------------------------------------------------------
th.job_dir += '/{:02d}_{}'.format(id, model_name)
th.prefix = '{}_'.format(date_string())
summ_name = model_name
th.suffix = ''
th.visible_gpu_id = 0
# ---------------------------------------------------------------------------
# 2. model setup
# ---------------------------------------------------------------------------
th.model = model
"""
Layers state_size #params
L1
2x44: 10189
3x30: 10225
4x23: 10444
L2
2x29: 10146
3x20: 9945
L3
2x23: 9938
3x16: 9785
"""
th.num_layers = 1
th.unit_size = 3
th.num_units = 30
# Setting truncate_grad to False works better
th.truncate_grad = False
# ---------------------------------------------------------------------------
# 3. trainer setup
# ---------------------------------------------------------------------------
th.epoch = 1000
th.batch_size = 1
th.optimizer = 'adam'
th.learning_rate = 0.0008
th.validation_per_round = 4
# ---------------------------------------------------------------------------
# 4. summary and note setup
# ---------------------------------------------------------------------------
th.train = True
th.overwrite = True
# ---------------------------------------------------------------------------
# 5. other stuff and activate
# ---------------------------------------------------------------------------
th.mark = '{}({}x{})'.format(model_name, th.unit_size, th.num_units)
th.gather_summ_name = th.prefix + summ_name + th.suffix + '.sum'
core.activate()
def main(_):
console.start('{} on CIFAR-10 task'.format(model_name.upper()))
th = core.th
# ---------------------------------------------------------------------------
# 0. date set setup
# ---------------------------------------------------------------------------
# ---------------------------------------------------------------------------
# 1. folder/file names and device
# ---------------------------------------------------------------------------
th.job_dir += '/{:02d}_{}'.format(id, model_name)
summ_name = model_name
th.prefix = '{}_'.format(date_string())
th.suffix = '_t00'
th.visible_gpu_id = 0
# ---------------------------------------------------------------------------
# 2. model setup
# ---------------------------------------------------------------------------
th.model = model
th.centralize_data = True
th.num_layers = 50
th.layer_width = 100
th.spatial_activation = 'tanh'
th.bias_initializer = -5.
# ---------------------------------------------------------------------------
# 3. trainer setup
# ---------------------------------------------------------------------------
th.epoch = 200
th.batch_size = 128
th.validation_per_round = 1
th.optimizer = tf.train.AdamOptimizer
th.learning_rate = 0.0004
th.patience = 5
th.early_stop = False
th.validate_train_set = True
th.val_decimals = 6
# ---------------------------------------------------------------------------
# 4. summary and note setup
th.export_tensors_upon_validation = True
# th.export_gates = True
th.train = True
th.save_model = True
th.overwrite = True
# ---------------------------------------------------------------------------
# 5. other stuff and activate
# ---------------------------------------------------------------------------
th.mark = '{}({}x{}-{})'.format(
model_name, th.layer_width, th.num_layers, th.spatial_activation)
th.gather_summ_name = th.prefix + summ_name + th.suffix + '.sum'
core.activate()
def main(_):
console.start('{} on TIMIT task'.format(model_name.upper()))
th = core.th
# ---------------------------------------------------------------------------
# 1. folder/file names and device
# ---------------------------------------------------------------------------
th.job_dir += '/{:02d}_{}'.format(id, model_name)
th.prefix = '{}_'.format(date_string())
summ_name = model_name
th.suffix = ''
th.visible_gpu_id = 0
# ---------------------------------------------------------------------------
# 2. model setup
# ---------------------------------------------------------------------------
th.model = model
"""For SxN GDU, params # = 2x(14+SxN)x(SxN)+25xSxN+25 = (53+2xSxN)xSxN+25
Denote SxN as x, # = 2*x^2 + 53*x + 25
x = (sqrt(2609+8*#)-53)/4
5000: 38; 10000: 58.6
2 layers: 37; 3 layers: 29;
"""
th.num_layers = 1
layer2config = {
1: '15x2+5x5+3x1',
2: '15x2+5x1+2x1',
3: '15x1+12x1+2x1',
}
th.gdu_string = '15x2+7x4'
th.gdu_string = layer2config[th.num_layers]
th.state_size = sum([
np.prod([int(x) for x in g.split('x')])
for g in th.gdu_string.split('+')
])
# ---------------------------------------------------------------------------
# 3. trainer setup
# ---------------------------------------------------------------------------
th.epoch = 1000
th.batch_size = 1
th.optimizer = 'adam'
th.learning_rate = 0.003
th.validation_per_round = 4
# ---------------------------------------------------------------------------
# 4. summary and note setup
# ---------------------------------------------------------------------------
th.train = True
th.overwrite = True
# ---------------------------------------------------------------------------
# 5. other stuff and activate
# ---------------------------------------------------------------------------
th.mark = '{}({})'.format(model_name, th.gdu_string)
th.gather_summ_name = th.prefix + summ_name + th.suffix + '.sum'
core.activate()
def main(_):
console.start('BResNet task')
description = '0'
# Configurations
th = NlsHub(as_global=True)
th.memory_depth = 80
th.num_blocks = 3
th.multiplier = 1
th.hidden_dim = th.memory_depth * th.multiplier
th.mark = 'bres-{}x({}x{})-{}'.format(th.num_blocks, th.memory_depth,
th.multiplier, description)
th.epoch = 50000
th.batch_size = 64
th.learning_rate = 0.0001
th.start_at = 0
th.reg_strength = 0.000
th.validation_per_round = 30
th.train = True
th.smart_train = True
th.idle_tol = 30
th.max_bad_apples = 5
th.lr_decay = 0.6
th.early_stop = True
th.save_mode = SaveMode.ON_RECORD
th.warm_up_rounds = 50
th.overwrite = True
th.export_note = True
th.summary = False
th.save_model = False
# Smoothen
th.overwrite = th.overwrite and th.start_at == 0
# Get model
model = model_lib.bres_net_res0(th)
# Load data
train_set, val_set, test_set = load_wiener_hammerstein(
th.data_dir, depth=th.memory_depth)
assert isinstance(train_set, DataSet)
assert isinstance(val_set, DataSet)
assert isinstance(test_set, DataSet)
# Train or evaluate
if th.train:
model.nn.train(train_set,
validation_set=val_set,
trainer_hub=th,
start_at=th.start_at)
else:
model.evaluate(train_set, start_at=th.memory_depth)
model.evaluate(val_set, start_at=th.memory_depth)
model.evaluate(test_set, start_at=th.memory_depth)
# End
console.end()
def main(_):
console.start('trainer.task')
EPOCH = 1000
# FLAGS.train = False
FLAGS.overwrite = True
# FLAGS.save_best = True
FLAGS.smart_train = True
# Hyper parameters
LEARNING_RATE = 0.001
LAYER_NUM = 4
BATCH_SIZE = 32
MEMORY_DEPTH = 80
LAYER_DIM = MEMORY_DEPTH * 2
ACTIVATION = 'relu'
# Set default flags
FLAGS.progress_bar = True
FLAGS.save_model = True
FLAGS.summary = False
FLAGS.snapshot = False
PRINT_CYCLE = 100
WH_PATH = os.path.join(nls_root, 'data/wiener_hammerstein/whb.tfd')
MARK = 'mlp00'
# Get model
model = model_lib.mlp_00(MARK,
MEMORY_DEPTH,
LAYER_DIM,
LAYER_NUM,
LEARNING_RATE,
activation=ACTIVATION)
# Load data set
train_set, val_set, test_set = load_wiener_hammerstein(WH_PATH,
depth=MEMORY_DEPTH)
assert isinstance(train_set, DataSet)
assert isinstance(val_set, DataSet)
assert isinstance(test_set, DataSet)
# Train or evaluate
if FLAGS.train:
model.identify(train_set,
val_set,
batch_size=BATCH_SIZE,
print_cycle=PRINT_CYCLE,
epoch=EPOCH)
else:
model.evaluate(train_set, start_at=MEMORY_DEPTH, plot=False)
model.evaluate(val_set, start_at=MEMORY_DEPTH, plot=False)
model.evaluate(test_set, start_at=MEMORY_DEPTH, plot=False)
console.end()
def main(_):
console.start('{} on sCIFAR-10 task'.format(model_name.upper()))
th = core.th
# ---------------------------------------------------------------------------
# 0. date set setup
# ---------------------------------------------------------------------------
# ---------------------------------------------------------------------------
# 1. folder/file names and device
# ---------------------------------------------------------------------------
th.job_dir += '/{:02d}_{}'.format(id, model_name)
summ_name = model_name
th.prefix = '{}_'.format(date_string())
th.suffix = ''
th.visible_gpu_id = 0
# ---------------------------------------------------------------------------
# 2. model setup
# ---------------------------------------------------------------------------
th.model = model
th.gdu_string = '5x60'
th.use_reset_gate = True
th.sog_version = 1 # sog_v1 is much faster
th.dropout = 0.1
th.output_dropout = 0.2
# ---------------------------------------------------------------------------
# 3. trainer setup
# ---------------------------------------------------------------------------
th.epoch = 10000
th.batch_size = 128
th.validation_per_round = 10
th.optimizer = tf.train.AdamOptimizer
th.learning_rate = 0.001
th.clip_threshold = 1.0
th.clip_method = 'value'
# ---------------------------------------------------------------------------
# 4. summary and note setup
# ---------------------------------------------------------------------------
th.train = True
th.save_model = True
th.overwrite = False
# ---------------------------------------------------------------------------
# 5. other stuff and activate
# ---------------------------------------------------------------------------
th.mark = GDU.mark()
th.mark += '_rdp{}odp{}gc{}'.format(th.dropout, th.output_dropout,
th.clip_threshold)
th.gather_summ_name = th.prefix + summ_name + th.suffix + '.sum'
core.activate()
def main(_):
console.start('basis task')
# Configurations
th = core.th
id = 11
th.job_dir = 'basis_task'
th.model = models.multinput_ver_only
th.actype1 = 'relu'
th.epoch = 5000
th.learning_rate = 1e-3
th.batch_size = 32
th.validation_per_round = 1
th.val_batch_size = th.batch_size
th.print_cycle = 20
th.patience = 100
th.shuffle = True
# th.train = False
th.smart_train = False
th.max_bad_apples = 4
th.lr_decay = 0.6
th.rand_over_classes = False
th.save_model = True
th.overwrite = True
th.export_note = True
th.summary = True
th.monitor = False
th.allow_growth = False
th.gpu_memory_fraction = 0.3
th.raw_keep_prob = 0.9
th.mfcc_keep_prob = 0.7
th.concat_keep_prob = 0.9
th.fold = 1
# th.shuffle = False
th.rand_pos = True
th.test_all = True
th.val_on_train_set = False
th.visible_gpu_id = '1'
# description = 'raw_data_mfcc_dropout_{}_random_{}_fold_{}'.format(
# th.mfcc_keep_prob, th.concat_keep_prob, th.fold)
# description = 'raw_data_mfcc_dropout_{}_{}'.format(th.mfcc_keep_prob,
# th.concat_keep_prob)
description = 'raw_data_mfcc_model_{}'.format(id)
# description = 'raw_data_mfcc_simlified_dropout_0.7_reg_0.2_sap_all'
th.mark = 'cnn_{}'.format(description)
export_false = True
core.activate()
def main(_):
console.start('mlp task')
description = 'm'
# Configurations
th = NlsHub(as_global=True)
th.memory_depth = 40
th.num_blocks = 2
th.multiplier = 2
th.hidden_dim = th.memory_depth * th.multiplier
th.mark = 'mlp-{}x({}x{})-{}'.format(th.num_blocks, th.memory_depth,
th.multiplier, description)
th.epoch = 50000
th.batch_size = 64
th.learning_rate = 0.001
th.validation_per_round = 20
th.train = True
th.smart_train = False
th.idle_tol = 20
th.max_bad_apples = 4
th.lr_decay = 0.6
th.early_stop = True
th.save_mode = SaveMode.ON_RECORD
th.warm_up_rounds = 50
th.overwrite = True
th.export_note = True
th.summary = True
th.monitor = True
th.save_model = False
th.allow_growth = False
th.gpu_memory_fraction = 0.4
# Get model
model = model_lib.mlp_00(th)
# Load data
train_set, val_set, test_set = load_wiener_hammerstein(
th.data_dir, depth=th.memory_depth)
assert isinstance(train_set, DataSet)
assert isinstance(val_set, DataSet)
assert isinstance(test_set, DataSet)
# Train or evaluate
if th.train:
model.nn.train(train_set, validation_set=val_set, trainer_hub=th)
else:
console.show_status('Evaluating ...')
model.evaluate(train_set, start_at=th.memory_depth)
model.evaluate(val_set, start_at=th.memory_depth)
model.evaluate(test_set, start_at=th.memory_depth, plot=True)
# End
console.end()
def main(_):
console.start('trainer.task')
# Set default flags
FLAGS.train = True
if FLAGS.use_default:
FLAGS.overwrite = True
FLAGS.smart_train = False
FLAGS.save_best = False
FLAGS.smart_train = True
FLAGS.save_best = False
WH_PATH = FLAGS.data_dir
MARK = 'lottery02'
MEMORY_DEPTH = 80
PRINT_CYCLE = 50
EPOCH = 1000
LR = 0.000088
LAYER_DIM = MEMORY_DEPTH * FLAGS.coe
# ACTIVATION = FLAGS.activation
ACTIVATION = 'relu'
# BRANCHES = FLAGS.branches
BRANCHES = 6
LR_LIST = [FLAGS.lr1] * (BRANCHES + 1)
FLAGS.smart_train = True
# Get model
model = model_lib.mlp02(MARK,
MEMORY_DEPTH,
BRANCHES,
LAYER_DIM,
LR,
ACTIVATION,
identity_init=True)
# Load data set
train_set, val_set, test_set = load_wiener_hammerstein(WH_PATH,
depth=MEMORY_DEPTH)
assert isinstance(train_set, DataSet)
assert isinstance(val_set, DataSet)
assert isinstance(test_set, DataSet)
# Train
if FLAGS.train:
model.identify(train_set,
val_set,
batch_size=64,
print_cycle=PRINT_CYCLE,
epoch=EPOCH,
lr_list=LR_LIST)
console.end()
def main(_):
console.start('{} on TO task'.format(model_name.upper()))
th = core.th
# ---------------------------------------------------------------------------
# 0. date set setup
# ---------------------------------------------------------------------------
th.sequence_length = 100
th.bits = 3
# ---------------------------------------------------------------------------
# 1. folder/file names and device
# ---------------------------------------------------------------------------
th.job_dir += '/{:02d}_{}'.format(id, model_name)
summ_name = model_name
th.visible_gpu_id = 0
th.prefix = '{}_'.format(date_string())
th.suffix = ''
# ---------------------------------------------------------------------------
# 2. model setup
# ---------------------------------------------------------------------------
th.model = model
th.gam_config = '6x10'
th.head_size = 10
th.hyper_kernel = 'gru'
th.state_size = 60
th.num_layers = 1
# ---------------------------------------------------------------------------
# 3. trainer setup
# ---------------------------------------------------------------------------
th.max_iterations = 50000
th.optimizer = tf.train.AdamOptimizer
th.learning_rate = 0.001
# ---------------------------------------------------------------------------
# 4. summary and note setup
# ---------------------------------------------------------------------------
th.export_tensors_upon_validation = True
# ---------------------------------------------------------------------------
# 5. other stuff and activate
# ---------------------------------------------------------------------------
tail = '_{}bits_L{}'.format(th.bits, th.sequence_length)
th.mark = GamRHN.mark() + tail
th.gather_summ_name = th.prefix + summ_name + tail + th.suffix + '.sum'
core.activate()
def main(_):
console.start('Multinput task')
# Configurations
th = core.th
th.job_dir = 'res_task'
th.model = models.res_00
th.actype1 = 'relu'
th.epoch = 5000
th.learning_rate = 1e-3
th.batch_size = 32
th.validation_per_round = 1
th.val_batch_size = th.batch_size
th.print_cycle = 20
th.patience = 100
th.shuffle = True
# th.train = False
th.smart_train = False
th.max_bad_apples = 4
th.lr_decay = 0.6
th.rand_over_classes = False
th.save_model = True
th.overwrite = True
th.export_note = True
th.summary = True
th.monitor = False
th.allow_growth = False
th.gpu_memory_fraction = 0.4
th.raw_keep_prob = 0.9
th.mfcc_keep_prob = 0.7
th.concat_keep_prob = 0.9
th.fold = 0
# th.shuffle = False
th.rand_pos = True
th.visible_gpu_id = '0'
# description = 'cnn_raw_data_mfcc_random_rand'
description = 'raw_data_mfcc_dropout_{}_{}'.format(th.mfcc_keep_prob,
th.concat_keep_prob)
th.mark = 'cnn_{}'.format(description)
export_false = True
core.activate()
def main(_):
console.start('{} on TIMIT task'.format(model_name.upper()))
th = core.th
# ---------------------------------------------------------------------------
# 1. folder/file names and device
# ---------------------------------------------------------------------------
th.job_dir += '/{:02d}_{}'.format(id, model_name)
th.prefix = '{}_'.format(date_string())
summ_name = model_name
th.suffix = ''
th.visible_gpu_id = 0
# ---------------------------------------------------------------------------
# 2. model setup
# ---------------------------------------------------------------------------
th.model = model
"""
Layers state_size #params
1 33 10156
2 21 10378
3 17 10752
"""
th.num_layers = 1
th.state_size = 33
layer2size = {1: 33, 2: 21, 3: 17, 4: 14}
th.state_size = layer2size[th.num_layers]
# ---------------------------------------------------------------------------
# 3. trainer setup
# ---------------------------------------------------------------------------
th.epoch = 1000
th.batch_size = 1
th.optimizer = 'adam'
th.learning_rate = 0.003
th.validation_per_round = 4
# ---------------------------------------------------------------------------
# 4. summary and note setup
# ---------------------------------------------------------------------------
th.train = True
th.overwrite = True
# ---------------------------------------------------------------------------
# 5. other stuff and activate
# ---------------------------------------------------------------------------
th.mark = '{}({})'.format(model_name, th.state_size)
th.gather_summ_name = th.prefix + summ_name + th.suffix + '.sum'
core.activate()
def main(_):
console.start('{} on TO task'.format(model_name.upper()))
th = core.th
# ---------------------------------------------------------------------------
# 0. date set setup
# ---------------------------------------------------------------------------
th.sequence_length = 100
th.bits = 3
# ---------------------------------------------------------------------------
# 1. folder/file names and device
# ---------------------------------------------------------------------------
th.job_dir += '/{:02d}_{}'.format(id, model_name)
summ_name = model_name
th.visible_gpu_id = 0
th.prefix = '{}_'.format(date_string())
th.suffix = ''
# ---------------------------------------------------------------------------
# 2. model setup
# ---------------------------------------------------------------------------
th.model = model
th.fast_layers = 2
table = {2: 41, 3: 45, 4: 47, 5: 49, 6: 50}
th.fast_size = table[th.fast_layers]
th.slow_size = th.fast_size
th.hyper_kernel = 'lstm'
th.forget_bias_initializer = 2.0
# ---------------------------------------------------------------------------
# 3. trainer setup
# ---------------------------------------------------------------------------
th.max_iterations = 50000
th.optimizer = tf.train.AdamOptimizer
th.learning_rate = 0.001
# ---------------------------------------------------------------------------
# 4. summary and note setup
# ---------------------------------------------------------------------------
th.export_tensors_upon_validation = True
# ---------------------------------------------------------------------------
# 5. other stuff and activate
# ---------------------------------------------------------------------------
tail = '_{}bits_L{}'.format(th.bits, th.sequence_length)
th.mark = FastSlow.mark() + tail
th.gather_summ_name = th.prefix + summ_name + tail + th.suffix + '.sum'
core.activate()