def test_experiment_fit_gen(self, get_model, get_loss_metric,
get_custom_l, get_callback_fix):
new_session()
model, metrics, cust_objects = prepare_model(get_model(get_custom_l),
get_loss_metric,
get_custom_l)
model_name = model.__class__.__name__
_, data_val_use = make_data(train_samples, test_samples)
expe = Experiment(model)
for val in [1, data_val_use]:
gen, data, data_stream = make_gen(batch_size)
if val == 1:
val, data_2, data_stream_2 = make_gen(batch_size)
expe.fit_gen([gen], [val], nb_epoch=2,
model=model,
metrics=metrics,
custom_objects=cust_objects,
samples_per_epoch=64,
nb_val_samples=128,
verbose=2, overwrite=True,
callbacks=get_callback_fix)
close_gens(gen, data, data_stream)
if val == 1:
close_gens(val, data_2, data_stream_2)
if K.backend() == 'tensorflow':
K.clear_session()
print(self)
def test_experiment_fit_gen(self, get_model, get_loss_metric, get_custom_l,
get_callback_fix):
new_session()
model, metrics, cust_objects = prepare_model(get_model(get_custom_l),
get_loss_metric,
get_custom_l)
model_name = model.__class__.__name__
_, data_val_use = make_data(train_samples, test_samples)
expe = Experiment(model)
for val in [1, data_val_use]:
gen, data, data_stream = make_gen(batch_size)
if val == 1:
val, data_2, data_stream_2 = make_gen(batch_size)
expe.fit_gen([gen], [val],
nb_epoch=2,
model=model,
metrics=metrics,
custom_objects=cust_objects,
samples_per_epoch=64,
nb_val_samples=128,
verbose=2,
overwrite=True,
callbacks=get_callback_fix)
close_gens(gen, data, data_stream)
if val == 1:
close_gens(val, data_2, data_stream_2)
if K.backend() == 'tensorflow':
K.clear_session()
print(self)
def test_fit_gen_async(self):
gen, data, data_stream = make_gen(batch_size)
val, data_2, data_stream_2 = make_gen(batch_size)
experiments = make_experiments()
param_search = HParamsSearch(experiments, metric='loss', op=np.min)
param_search.fit_gen_async([gen], [val], nb_epoch=2,
verbose=2,
nb_val_samples=128,
samples_per_epoch=64,
overwrite=True)
param_search.summary(verbose=True, metrics={'val_loss': np.min})
close_gens(gen, data, data_stream)
close_gens(val, data_2, data_stream_2)
print(self)
def test_fit_gen_async(self):
gen, data, data_stream = make_gen(batch_size)
val, data_2, data_stream_2 = make_gen(batch_size)
experiments = make_experiments()
param_search = HParamsSearch(experiments, metric='loss', op=np.min)
param_search.fit_gen_async([gen], [val],
nb_epoch=2,
verbose=2,
nb_val_samples=128,
samples_per_epoch=64,
overwrite=True)
param_search.summary(verbose=True, metrics={'val_loss': np.min})
close_gens(gen, data, data_stream)
close_gens(val, data_2, data_stream_2)
print(self)
def test_utils():
assert get_function_name("bob") == "bob"
test_switch = switch_backend('sklearn')
assert test_switch is not None
gen, data, data_stream = make_gen(batch_size)
open_dataset_gen(data_stream)
gen.close()
data.close(None)
data_stream.close()
for i in range(1, 20):
utls.window(list(range(i * 2)), i)
def test_utils():
assert get_function_name("bob") == "bob"
test_switch = switch_backend('sklearn')
assert test_switch is not None
gen, data, data_stream = make_gen(batch_size)
open_dataset_gen(data_stream)
gen.close()
data.close(None)
data_stream.close()
for i in range(1, 20):
utls.window(list(range(i * 2)), i)
def test_experiment_fit_gen_async(self, get_model, get_loss_metric,
get_custom_l):
new_session()
model, metrics, cust_objects = prepare_model(get_model(get_custom_l),
get_loss_metric,
get_custom_l)
_, data_val_use = make_data(train_samples, test_samples)
expe = Experiment(model)
expected_value = 2
for val in [None, 1, data_val_use]:
gen, data, data_stream = make_gen(batch_size)
if val == 1:
val, data_2, data_stream_2 = make_gen(batch_size)
_, thread = expe.fit_gen_async([gen], [val],
nb_epoch=2,
model=model,
metrics=metrics,
custom_objects=cust_objects,
samples_per_epoch=64,
nb_val_samples=128,
verbose=2,
overwrite=True)
thread.join()
for k in expe.full_res['metrics']:
if 'iter' not in k:
assert len(expe.full_res['metrics'][k]) == expected_value
close_gens(gen, data, data_stream)
if val == 1:
close_gens(val, data_2, data_stream_2)
if K.backend() == 'tensorflow':
K.clear_session()
print(self)
def test_experiment_fit_gen_async(self, get_model, get_loss_metric,
get_custom_l):
new_session()
model, metrics, cust_objects = prepare_model(get_model(get_custom_l),
get_loss_metric,
get_custom_l)
_, data_val_use = make_data(train_samples, test_samples)
expe = Experiment(model)
expected_value = 2
for val in [None, 1, data_val_use]:
gen, data, data_stream = make_gen(batch_size)
if val == 1:
val, data_2, data_stream_2 = make_gen(batch_size)
_, thread = expe.fit_gen_async([gen], [val], nb_epoch=2,
model=model,
metrics=metrics,
custom_objects=cust_objects,
samples_per_epoch=64,
nb_val_samples=128,
verbose=2, overwrite=True)
thread.join()
for k in expe.full_res['metrics']:
if 'iter' not in k:
assert len(
expe.full_res['metrics'][k]) == expected_value
close_gens(gen, data, data_stream)
if val == 1:
close_gens(val, data_2, data_stream_2)
if K.backend() == 'tensorflow':
K.clear_session()
print(self)
def get_generators(request):
if request.param == 'one to many':
nb_train, nb_val = 4, 48
gen_t, data_t, d_stream_t = make_gen(batch_size, nb_train)
gen, data, d_stream = make_gen(batch_size, nb_val)
elif request.param == 'many to one':
nb_train, nb_val = 48, 4
gen_t, data_t, d_stream_t = make_gen(batch_size, nb_train)
gen, data, d_stream = make_gen(batch_size, nb_val)
elif request.param == 'val 1':
nb_train, nb_val = 4, 4
gen_t, data_t, d_stream_t = make_gen(batch_size, nb_train)
gen, data, d_stream = make_gen(batch_size, nb_val)
elif request.param == 'many to many':
nb_train, nb_val = 48, 48
gen_t, data_t, d_stream_t = make_gen(batch_size, nb_train)
gen, data, d_stream = make_gen(batch_size, nb_val)
return gen_t, data_t, d_stream_t, gen, data, d_stream, (nb_train, nb_val)
def get_generators(request):
if request.param == 'one to many':
nb_train, nb_val = 4, 48
gen_t, data_t, d_stream_t = make_gen(batch_size, nb_train)
gen, data, d_stream = make_gen(batch_size, nb_val)
elif request.param == 'many to one':
nb_train, nb_val = 48, 4
gen_t, data_t, d_stream_t = make_gen(batch_size, nb_train)
gen, data, d_stream = make_gen(batch_size, nb_val)
elif request.param == 'val 1':
nb_train, nb_val = 4, 4
gen_t, data_t, d_stream_t = make_gen(batch_size, nb_train)
gen, data, d_stream = make_gen(batch_size, nb_val)
elif request.param == 'many to many':
nb_train, nb_val = 48, 48
gen_t, data_t, d_stream_t = make_gen(batch_size, nb_train)
gen, data, d_stream = make_gen(batch_size, nb_val)
return gen_t, data_t, d_stream_t, gen, data, d_stream, (nb_train, nb_val)