def test_log_param(self):
submarine.log_param("name_1", "a")
# Validate params
with self.store.ManagedSessionMaker() as session:
params = session.query(SqlParam).options().filter(
SqlParam.id == JOB_ID).all()
assert params[0].key == "name_1"
assert params[0].value == "a"
assert params[0].id == JOB_ID
def log_param(self):
submarine.log_param("name_1", "a", "worker-1")
# Validate params
with self.store.ManagedSessionMaker() as session:
params = session \
.query(SqlParam) \
.options() \
.filter(SqlParam.job_name == JOB_NAME).all()
assert params[0].key == "name_1"
assert params[0].value == "a"
assert params[0].worker_index == "worker-1"
assert params[0].job_name == JOB_NAME
def test_log_param(tracking_uri_mock):
environ["SUBMARINE_JOB_NAME"] = JOB_NAME
submarine.log_param("name_1", "a", "worker-1")
tracking_uri = utils.get_tracking_uri()
store = utils.get_sqlalchemy_store(tracking_uri)
# Validate params
with store.ManagedSessionMaker() as session:
params = session \
.query(SqlParam) \
.options() \
.filter(SqlParam.job_name == JOB_NAME).all()
assert params[0].key == "name_1"
assert params[0].value == "a"
assert params[0].worker_index == "worker-1"
assert params[0].job_name == JOB_NAME
# contributor license agreements. See the NOTICE file distributed with
# this work for additional information regarding copyright ownership.
# The ASF licenses this file to You under the Apache License, Version 2.0
# (the "License"); you may not use this file except in compliance with
# the License. You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import numpy as np
from os import environ
from sklearn.linear_model import LogisticRegression
import submarine
if __name__ == "__main__":
# note: SUBMARINE_JOB_NAME should be set by submarine submitter
environ["SUBMARINE_JOB_NAME"] = "application_1234"
X = np.array([-2, -1, 0, 1, 2, 1]).reshape(-1, 1)
y = np.array([0, 0, 1, 1, 1, 0])
lr = LogisticRegression(solver='liblinear', max_iter=100)
submarine.log_param("max_iter", 100, "worker-1")
lr.fit(X, y)
score = lr.score(X, y)
print("Score: %s" % score)
submarine.log_metric("score", score, "worker-1")
"""
Licensed to the Apache Software Foundation (ASF) under one
or more contributor license agreements. See the NOTICE file
distributed with this work for additional information
regarding copyright ownership. The ASF licenses this file
to you under the Apache License, Version 2.0 (the
"License"); you may not use this file except in compliance
with the License. You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing,
software distributed under the License is distributed on an
"AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
KIND, either express or implied. See the License for the
specific language governing permissions and limitations
under the License.
"""
import random
import time
import submarine
if __name__ == "__main__":
submarine.log_param("learning_rate", random.random())
for i in range(100):
time.sleep(1)
submarine.log_metric("mse", random.random() * 100, i)
submarine.log_metric("acc", random.random(), i)
# Licensed to the Apache Software Foundation (ASF) under one or more
# contributor license agreements. See the NOTICE file distributed with
# this work for additional information regarding copyright ownership.
# The ASF licenses this file to You under the Apache License, Version 2.0
# (the "License"); you may not use this file except in compliance with
# the License. You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import numpy as np
from sklearn.linear_model import LogisticRegression
import submarine
if __name__ == "__main__":
X = np.array([-2, -1, 0, 1, 2, 1]).reshape(-1, 1)
y = np.array([0, 0, 1, 1, 1, 0])
lr = LogisticRegression(solver="liblinear", max_iter=100)
submarine.log_param("max_iter", 100)
lr.fit(X, y)
score = lr.score(X, y)
print("Score: %s" % score)
submarine.log_metric("score", score)
def on_epoch_end(self, epoch, logs=None):
print("\nLearning rate for epoch {} is {}".format(
epoch + 1, model.optimizer.lr.numpy()))
submarine.log_metric("lr", model.optimizer.lr.numpy())
submarine.save_model(model, "tensorflow", "mnist-tf")
# Put all the callbacks together.
callbacks = [
tf.keras.callbacks.TensorBoard(log_dir="./logs"),
tf.keras.callbacks.ModelCheckpoint(filepath=checkpoint_prefix,
save_weights_only=True),
tf.keras.callbacks.LearningRateScheduler(decay),
PrintLR(),
]
if __name__ == "__main__":
EPOCHS = 2
hist = model.fit(train_dataset, epochs=EPOCHS, callbacks=callbacks)
for i in range(EPOCHS):
submarine.log_metric("val_loss", hist.history["loss"][i], i)
submarine.log_metric("Val_accuracy", hist.history["accuracy"][i], i)
model.load_weights(tf.train.latest_checkpoint(checkpoint_dir))
eval_loss, eval_acc = model.evaluate(eval_dataset)
print("Eval loss: {}, Eval accuracy: {}".format(eval_loss, eval_acc))
submarine.log_param("loss", eval_loss)
submarine.log_param("acc", eval_acc)
"""Reference:
https://www.tensorflow.org/api_docs/python/tf/distribute/MirroredStrategy
"""
transform=transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.1307, ), (0.3081, ))
]),
),
batch_size=args.test_batch_size,
shuffle=False,
**kwargs)
model = Net().to(device)
if is_distributed():
Distributor = (nn.parallel.DistributedDataParallel
if use_cuda else nn.parallel.DistributedDataParallelCPU)
model = Distributor(model)
optimizer = optim.SGD(model.parameters(),
lr=args.lr,
momentum=args.momentum)
submarine.log_param("learning_rate", args.lr)
submarine.log_param("batch_size", args.batch_size)
for epoch in range(1, args.epochs + 1):
train(args, model, device, train_loader, optimizer, epoch, writer)
test(args, model, device, test_loader, writer, epoch)
if args.save_model:
torch.save(model.state_dict(), "mnist_cnn.pt")
"""
Reference:
https://github.com/kubeflow/pytorch-operator/blob/master/examples/mnist/mnist.py
"""
#experiment = Experiment(api_key="ej6XeyCVjqHM8uLDNj5VGrzjP",
# project_name="testing", workspace="pingsutw")
if __name__ == "__main__":
submarine.set_tracking_uri(
"mysql+pymysql://submarine:password@submarine-database/submarine")
print("TF_CONFIG", env.get_env("TF_CONFIG"))
print("JOB_NAME: ", env.get_env("JOB_NAME"))
print("TYPE: ", env.get_env("TPYE"))
print("TASK_INDEX: ", env.get_env("TASK_INDEX"))
print("CLUSTER_SPEC: ", env.get_env("CLUSTER_SPEC"))
print("RANK: ", env.get_env("RANK"))
submarine.log_param("max_iter", 100)
submarine.log_param("learning_rate", 0.0001)
submarine.log_param("alpha", 20)
submarine.log_param("batch_size", 256)
submarine.log_metric("score", 2)
submarine.log_metric("score", 5)
submarine.log_metric("score", 8)
submarine.log_metric("score", 5)
submarine.log_metric("score", 10)
submarine.log_metric("AUC", 0.62)
submarine.log_metric("AUC", 0.68)
submarine.log_metric("AUC", 0.75)
submarine.log_metric("AUC", 0.64)
submarine.log_metric("AUC", 0.79)