def fit(self, train_x, train_y, validation_data_fit, epochs, **kwargs):
val_x, val_y = validation_data_fit
patience = 2
callbacks = [
keras.callbacks.EarlyStopping(monitor='val_loss',
patience=patience)
]
if self._is_multilabel:
train_y = train_y
val_y = val_y
else:
train_y = ohe2cat(train_y)
val_y = ohe2cat(val_y)
self._model.fit(
train_x,
train_y,
epochs=epochs,
callbacks=callbacks,
# validation_data=(val_x, val_y),
verbose=1, # Logs once per epoch.
batch_size=32,
shuffle=True,
# use_multiprocessing=True
)
def fit(self, train_x, train_y, validation_data_fit, params, epochs,
**kwargs):
patience = 2
callbacks = [
keras.callbacks.EarlyStopping(monitor='val_loss',
patience=patience)
]
val_x, val_y = validation_data_fit
if self._class_num == 2:
train_y = ohe2cat(train_y)
val_y = ohe2cat(val_y)
batch_size = params["batch_size"]
steps_per_epoch = int(len(train_x) // batch_size)
train_data_generator = ModelSequenceDataGenerator(
train_x, train_y, **params)
history = self._model.fit_generator(train_data_generator,
steps_per_epoch=steps_per_epoch,
validation_data=(val_x, val_y),
epochs=epochs,
max_queue_size=10,
callbacks=callbacks,
use_multiprocessing=False,
workers=1,
verbose=VERBOSE)
return history
def fit(self,
train_x,
train_y,
validation_data_fit,
epochs,
batch_size=32,
**kwargs):
val_x, val_y = validation_data_fit
callbacks = []
if epochs > 1:
early_stop = keras.callbacks.EarlyStopping(monitor='val_loss',
patience=3)
callbacks.append(early_stop)
self._model.fit(
train_x,
ohe2cat(train_y),
epochs=epochs,
callbacks=callbacks,
validation_data=(val_x, ohe2cat(val_y)),
verbose=1, # Logs once per epoch.
batch_size=batch_size,
shuffle=True)
else:
self._model.fit(
train_x,
ohe2cat(train_y),
epochs=1,
callbacks=callbacks,
verbose=1, # Logs once per epoch.
batch_size=batch_size,
shuffle=True)
def fit(self, train_x, train_y, validation_data_fit, epochs, **kwargs):
val_x, val_y = validation_data_fit
callbacks = [keras.callbacks.EarlyStopping(
monitor='val_loss', patience=3)]
self._model.fit(train_x, ohe2cat(train_y),
epochs=epochs,
callbacks=callbacks,
validation_data=(val_x, ohe2cat(val_y)),
verbose=1, # Logs once per epoch.
batch_size=32,
shuffle=True)
def fit(self,
train_x,
train_y,
validation_data_fit,
epochs,
cur_model_run_loop,
batch_size=32,
**kwargs):
val_x, val_y = validation_data_fit
callbacks = []
if self._use_step_decay:
lrate = LearningRateScheduler(self.step_decay)
callbacks.append(lrate)
if self._is_multilabel:
train_y = train_y
val_y = val_y
else:
train_y = ohe2cat(train_y)
val_y = ohe2cat(val_y)
if epochs > 1:
early_stop = keras.callbacks.EarlyStopping(monitor='val_loss',
patience=3)
callbacks.append(early_stop)
self._model.fit(
train_x,
train_y,
epochs=epochs,
callbacks=callbacks,
validation_data=(val_x, val_y),
verbose=1, # Logs once per epoch.
batch_size=batch_size,
shuffle=True)
else:
self._model.fit(
train_x,
train_y,
epochs=1,
callbacks=callbacks,
verbose=1, # Logs once per epoch.
batch_size=batch_size,
shuffle=True)
def fit(self,
train_x,
train_y,
validation_data_fit,
epochs,
cur_model_run_loop,
batch_size=64,
**kwargs):
index = [i for i in range(len(train_y))]
np.random.shuffle(index)
X = train_x
train_y = ohe2cat(train_y)
Y_b = np.eye(self.num_classes)[train_y]
X = X[index]
Y_b = Y_b[index]
if batch_size >= len(X):
batch_size = int(len(X) / 2)
print("TF model---------- batch_size:%d, X length: %d" %
(batch_size, len(X)))
train_step = tf.train.AdamOptimizer(0.0025).minimize(self.losses)
init_global = tf.global_variables_initializer()
init_local = tf.local_variables_initializer()
saver = tf.train.Saver()
with tf.Session() as sess:
sess.run([init_global, init_local])
rounds = min(int(len(X) / batch_size), self.rr)
print("TF model ---------- Rounds:%d" % rounds)
for i in range(rounds):
start = i * batch_size
end = (i + 1) * batch_size
_ = sess.run(train_step,
feed_dict={
self.input_x: X[start:end],
self.input_y: Y_b[start:end]
})
self.rr += self.add_rr
saver.save(sess, './ft.ckpt')
return 0
def fit(self, x_train, y_train, *args, **kwargs):
# sscaler = StandardScaler()
# x_train = sscaler.fit_transform(x_train[:, :])
self._existed_classes = set(ohe2cat(y_train))
print("=== svm class {}".format(len(self._existed_classes)))
self._model.fit(x_train, ohe2cat(y_train))
def fit(self, x_train, y_train, *args, **kwargs):
print("=== lr fit {}".format(y_train.shape))
self._existed_classes = set(ohe2cat(y_train))
print("=== lr class {}".format(len(self._existed_classes)))
self._model.fit(x_train, ohe2cat(y_train))
def fit(self, x_train, y_train, *args, **kwargs):
if not self._is_multilabel:
self._model.fit(x_train, ohe2cat(y_train))
else:
self._model.fit(x_train, y_train)
