Loading and Preprocessing Data with TensorFlow
Chapter 13
First, let's import a few common modules, ensure MatplotLib plots figures inline and prepare a function to save the figures. We also check that Python 3.5 or later is installed (although Python 2.x may work, it is deprecated so we strongly recommend you use Python 3 instead), as well as Scikit-Learn ≥0.20 and TensorFlow ≥2.0.
# collapse-show
# Python ≥3.5 is required
import sys
assert sys.version_info >= (3, 5)
# Scikit-Learn ≥0.20 is required
import sklearn
assert sklearn.__version__ >= "0.20"
try:
# %tensorflow_version only exists in Colab.
%tensorflow_version 2.x
!pip install -q -U tfx==0.15.0rc0
print("You can safely ignore the package incompatibility errors.")
except Exception:
pass
# TensorFlow ≥2.0 is required
import tensorflow as tf
from tensorflow import keras
assert tf.__version__ >= "2.0"
# Common imports
import numpy as np
import os
# to make this notebook's output stable across runs
np.random.seed(42)
# To plot pretty figures
%matplotlib inline
import matplotlib as mpl
import matplotlib.pyplot as plt
mpl.rc('axes', labelsize=14)
mpl.rc('xtick', labelsize=12)
mpl.rc('ytick', labelsize=12)
# Where to save the figures
PROJECT_ROOT_DIR = "."
CHAPTER_ID = "data"
IMAGES_PATH = os.path.join(PROJECT_ROOT_DIR, "images", CHAPTER_ID)
os.makedirs(IMAGES_PATH, exist_ok=True)
def save_fig(fig_id, tight_layout=True, fig_extension="png", resolution=300):
path = os.path.join(IMAGES_PATH, fig_id + "." + fig_extension)
print("Saving figure", fig_id)
if tight_layout:
plt.tight_layout()
plt.savefig(path, format=fig_extension, dpi=resolution)
X = tf.range(10)
dataset = tf.data.Dataset.from_tensor_slices(X)
dataset
Equivalently:
dataset = tf.data.Dataset.range(10)
for item in dataset:
print(item)
dataset = dataset.repeat(3).batch(7)
for item in dataset:
print(item)
dataset = dataset.map(lambda x: x * 2)
for item in dataset:
print(item)
dataset = dataset.apply(tf.data.experimental.unbatch())
dataset = dataset.filter(lambda x: x < 10) # keep only items < 10
for item in dataset.take(3):
print(item)
dataset = tf.data.Dataset.range(10).repeat(3)
dataset = dataset.shuffle(buffer_size=3, seed=42).batch(7)
for item in dataset:
print(item)
Let's start by loading and preparing the California housing dataset. We first load it, then split it into a training set, a validation set and a test set, and finally we scale it:
from sklearn.datasets import fetch_california_housing
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler
housing = fetch_california_housing()
X_train_full, X_test, y_train_full, y_test = train_test_split(
housing.data, housing.target.reshape(-1, 1), random_state=42)
X_train, X_valid, y_train, y_valid = train_test_split(
X_train_full, y_train_full, random_state=42)
scaler = StandardScaler()
scaler.fit(X_train)
X_mean = scaler.mean_
X_std = scaler.scale_
For a very large dataset that does not fit in memory, you will typically want to split it into many files first, then have TensorFlow read these files in parallel. To demonstrate this, let's start by splitting the housing dataset and save it to 20 CSV files:
def save_to_multiple_csv_files(data, name_prefix, header=None, n_parts=10):
housing_dir = os.path.join("datasets", "housing")
os.makedirs(housing_dir, exist_ok=True)
path_format = os.path.join(housing_dir, "my_{}_{:02d}.csv")
filepaths = []
m = len(data)
for file_idx, row_indices in enumerate(np.array_split(np.arange(m), n_parts)):
part_csv = path_format.format(name_prefix, file_idx)
filepaths.append(part_csv)
with open(part_csv, "wt", encoding="utf-8") as f:
if header is not None:
f.write(header)
f.write("\n")
for row_idx in row_indices:
f.write(",".join([repr(col) for col in data[row_idx]]))
f.write("\n")
return filepaths
train_data = np.c_[X_train, y_train]
valid_data = np.c_[X_valid, y_valid]
test_data = np.c_[X_test, y_test]
header_cols = housing.feature_names + ["MedianHouseValue"]
header = ",".join(header_cols)
train_filepaths = save_to_multiple_csv_files(train_data, "train", header, n_parts=20)
valid_filepaths = save_to_multiple_csv_files(valid_data, "valid", header, n_parts=10)
test_filepaths = save_to_multiple_csv_files(test_data, "test", header, n_parts=10)
Okay, now let's take a peek at the first few lines of one of these CSV files:
import pandas as pd
pd.read_csv(train_filepaths[0]).head()
Or in text mode:
with open(train_filepaths[0]) as f:
for i in range(5):
print(f.readline(), end="")
train_filepaths
filepath_dataset = tf.data.Dataset.list_files(train_filepaths, seed=42)
for filepath in filepath_dataset:
print(filepath)
n_readers = 5
dataset = filepath_dataset.interleave(
lambda filepath: tf.data.TextLineDataset(filepath).skip(1),
cycle_length=n_readers)
for line in dataset.take(5):
print(line.numpy())
Notice that field 4 is interpreted as a string.
record_defaults=[0, np.nan, tf.constant(np.nan, dtype=tf.float64), "Hello", tf.constant([])]
parsed_fields = tf.io.decode_csv('1,2,3,4,5', record_defaults)
parsed_fields
Notice that all missing fields are replaced with their default value, when provided:
parsed_fields = tf.io.decode_csv(',,,,5', record_defaults)
parsed_fields
The 5th field is compulsory (since we provided tf.constant([]) as the "default value"), so we get an exception if we do not provide it:
try:
parsed_fields = tf.io.decode_csv(',,,,', record_defaults)
except tf.errors.InvalidArgumentError as ex:
print(ex)
The number of fields should match exactly the number of fields in the record_defaults:
try:
parsed_fields = tf.io.decode_csv('1,2,3,4,5,6,7', record_defaults)
except tf.errors.InvalidArgumentError as ex:
print(ex)
n_inputs = 8 # X_train.shape[-1]
@tf.function
def preprocess(line):
defs = [0.] * n_inputs + [tf.constant([], dtype=tf.float32)]
fields = tf.io.decode_csv(line, record_defaults=defs)
x = tf.stack(fields[:-1])
y = tf.stack(fields[-1:])
return (x - X_mean) / X_std, y
preprocess(b'4.2083,44.0,5.3232,0.9171,846.0,2.3370,37.47,-122.2,2.782')
def csv_reader_dataset(filepaths, repeat=1, n_readers=5,
n_read_threads=None, shuffle_buffer_size=10000,
n_parse_threads=5, batch_size=32):
dataset = tf.data.Dataset.list_files(filepaths).repeat(repeat)
dataset = dataset.interleave(
lambda filepath: tf.data.TextLineDataset(filepath).skip(1),
cycle_length=n_readers, num_parallel_calls=n_read_threads)
dataset = dataset.shuffle(shuffle_buffer_size)
dataset = dataset.map(preprocess, num_parallel_calls=n_parse_threads)
dataset = dataset.batch(batch_size)
return dataset.prefetch(1)
train_set = csv_reader_dataset(train_filepaths, batch_size=3)
for X_batch, y_batch in train_set.take(2):
print("X =", X_batch)
print("y =", y_batch)
print()
train_set = csv_reader_dataset(train_filepaths, repeat=None)
valid_set = csv_reader_dataset(valid_filepaths)
test_set = csv_reader_dataset(test_filepaths)
model = keras.models.Sequential([
keras.layers.Dense(30, activation="relu", input_shape=X_train.shape[1:]),
keras.layers.Dense(1),
])
model.compile(loss="mse", optimizer=keras.optimizers.SGD(lr=1e-3))
batch_size = 32
model.fit(train_set, steps_per_epoch=len(X_train) // batch_size, epochs=10,
validation_data=valid_set)
model.evaluate(test_set, steps=len(X_test) // batch_size)
new_set = test_set.map(lambda X, y: X) # we could instead just pass test_set, Keras would ignore the labels
X_new = X_test
model.predict(new_set, steps=len(X_new) // batch_size)
optimizer = keras.optimizers.Nadam(lr=0.01)
loss_fn = keras.losses.mean_squared_error
n_epochs = 5
batch_size = 32
n_steps_per_epoch = len(X_train) // batch_size
total_steps = n_epochs * n_steps_per_epoch
global_step = 0
for X_batch, y_batch in train_set.take(total_steps):
global_step += 1
print("\rGlobal step {}/{}".format(global_step, total_steps), end="")
with tf.GradientTape() as tape:
y_pred = model(X_batch)
main_loss = tf.reduce_mean(loss_fn(y_batch, y_pred))
loss = tf.add_n([main_loss] + model.losses)
gradients = tape.gradient(loss, model.trainable_variables)
optimizer.apply_gradients(zip(gradients, model.trainable_variables))
optimizer = keras.optimizers.Nadam(lr=0.01)
loss_fn = keras.losses.mean_squared_error
@tf.function
def train(model, n_epochs, batch_size=32,
n_readers=5, n_read_threads=5, shuffle_buffer_size=10000, n_parse_threads=5):
train_set = csv_reader_dataset(train_filepaths, repeat=n_epochs, n_readers=n_readers,
n_read_threads=n_read_threads, shuffle_buffer_size=shuffle_buffer_size,
n_parse_threads=n_parse_threads, batch_size=batch_size)
for X_batch, y_batch in train_set:
with tf.GradientTape() as tape:
y_pred = model(X_batch)
main_loss = tf.reduce_mean(loss_fn(y_batch, y_pred))
loss = tf.add_n([main_loss] + model.losses)
gradients = tape.gradient(loss, model.trainable_variables)
optimizer.apply_gradients(zip(gradients, model.trainable_variables))
train(model, 5)
optimizer = keras.optimizers.Nadam(lr=0.01)
loss_fn = keras.losses.mean_squared_error
@tf.function
def train(model, n_epochs, batch_size=32,
n_readers=5, n_read_threads=5, shuffle_buffer_size=10000, n_parse_threads=5):
train_set = csv_reader_dataset(train_filepaths, repeat=n_epochs, n_readers=n_readers,
n_read_threads=n_read_threads, shuffle_buffer_size=shuffle_buffer_size,
n_parse_threads=n_parse_threads, batch_size=batch_size)
n_steps_per_epoch = len(X_train) // batch_size
total_steps = n_epochs * n_steps_per_epoch
global_step = 0
for X_batch, y_batch in train_set.take(total_steps):
global_step += 1
if tf.equal(global_step % 100, 0):
tf.print("\rGlobal step", global_step, "/", total_steps)
with tf.GradientTape() as tape:
y_pred = model(X_batch)
main_loss = tf.reduce_mean(loss_fn(y_batch, y_pred))
loss = tf.add_n([main_loss] + model.losses)
gradients = tape.gradient(loss, model.trainable_variables)
optimizer.apply_gradients(zip(gradients, model.trainable_variables))
train(model, 5)
Here is a short description of each method in the Dataset class:
for m in dir(tf.data.Dataset):
if not (m.startswith("_") or m.endswith("_")):
func = getattr(tf.data.Dataset, m)
if hasattr(func, "__doc__"):
print("● {:21s}{}".format(m + "()", func.__doc__.split("\n")[0]))
A TFRecord file is just a list of binary records. You can create one using a tf.io.TFRecordWriter:
with tf.io.TFRecordWriter("my_data.tfrecord") as f:
f.write(b"This is the first record")
f.write(b"And this is the second record")
And you can read it using a tf.data.TFRecordDataset:
filepaths = ["my_data.tfrecord"]
dataset = tf.data.TFRecordDataset(filepaths)
for item in dataset:
print(item)
You can read multiple TFRecord files with just one TFRecordDataset. By default it will read them one at a time, but if you set num_parallel_reads=3, it will read 3 at a time in parallel and interleave their records:
filepaths = ["my_test_{}.tfrecord".format(i) for i in range(5)]
for i, filepath in enumerate(filepaths):
with tf.io.TFRecordWriter(filepath) as f:
for j in range(3):
f.write("File {} record {}".format(i, j).encode("utf-8"))
dataset = tf.data.TFRecordDataset(filepaths, num_parallel_reads=3)
for item in dataset:
print(item)
options = tf.io.TFRecordOptions(compression_type="GZIP")
with tf.io.TFRecordWriter("my_compressed.tfrecord", options) as f:
f.write(b"This is the first record")
f.write(b"And this is the second record")
dataset = tf.data.TFRecordDataset(["my_compressed.tfrecord"],
compression_type="GZIP")
for item in dataset:
print(item)
For this section you need to install protobuf. In general you will not have to do so when using TensorFlow, as it comes with functions to create and parse protocol buffers of type tf.train.Example, which are generally sufficient. However, in this section we will learn about protocol buffers by creating our own simple protobuf definition, so we need the protobuf compiler (protoc): we will use it to compile the protobuf definition to a Python module that we can then use in our code.
First let's write a simple protobuf definition:
%%writefile person.proto
syntax = "proto3";
message Person {
string name = 1;
int32 id = 2;
repeated string email = 3;
}
And let's compile it (the --descriptor_set_out and --include_imports options are only required for the tf.io.decode_proto() example below):
!protoc person.proto --python_out=. --descriptor_set_out=person.desc --include_imports
!ls person*
from person_pb2 import Person
person = Person(name="Al", id=123, email=["a@b.com"]) # create a Person
print(person) # display the Person
person.name # read a field
person.name = "Alice" # modify a field
person.email[0] # repeated fields can be accessed like arrays
person.email.append("c@d.com") # add an email address
s = person.SerializeToString() # serialize to a byte string
s
person2 = Person() # create a new Person
person2.ParseFromString(s) # parse the byte string (27 bytes)
person == person2 # now they are equal
In rare cases, you may want to parse a custom protobuf (like the one we just created) in TensorFlow. For this you can use the tf.io.decode_proto() function:
person_tf = tf.io.decode_proto(
bytes=s,
message_type="Person",
field_names=["name", "id", "email"],
output_types=[tf.string, tf.int32, tf.string],
descriptor_source="person.desc")
person_tf.values
For more details, see the tf.io.decode_proto() documentation.
Here is the definition of the tf.train.Example protobuf:
syntax = "proto3";
message BytesList { repeated bytes value = 1; }
message FloatList { repeated float value = 1 [packed = true]; }
message Int64List { repeated int64 value = 1 [packed = true]; }
message Feature {
oneof kind {
BytesList bytes_list = 1;
FloatList float_list = 2;
Int64List int64_list = 3;
}
};
message Features { map<string, Feature> feature = 1; };
message Example { Features features = 1; };
# WARNING: there's currently a bug preventing "from tensorflow.train import X"
# so we work around it by writing "X = tf.train.X"
#from tensorflow.train import BytesList, FloatList, Int64List
#from tensorflow.train import Feature, Features, Example
BytesList = tf.train.BytesList
FloatList = tf.train.FloatList
Int64List = tf.train.Int64List
Feature = tf.train.Feature
Features = tf.train.Features
Example = tf.train.Example
person_example = Example(
features=Features(
feature={
"name": Feature(bytes_list=BytesList(value=[b"Alice"])),
"id": Feature(int64_list=Int64List(value=[123])),
"emails": Feature(bytes_list=BytesList(value=[b"a@b.com", b"c@d.com"]))
}))
with tf.io.TFRecordWriter("my_contacts.tfrecord") as f:
f.write(person_example.SerializeToString())
feature_description = {
"name": tf.io.FixedLenFeature([], tf.string, default_value=""),
"id": tf.io.FixedLenFeature([], tf.int64, default_value=0),
"emails": tf.io.VarLenFeature(tf.string),
}
for serialized_example in tf.data.TFRecordDataset(["my_contacts.tfrecord"]):
parsed_example = tf.io.parse_single_example(serialized_example,
feature_description)
parsed_example
parsed_example
parsed_example["emails"].values[0]
tf.sparse.to_dense(parsed_example["emails"], default_value=b"")
parsed_example["emails"].values
from sklearn.datasets import load_sample_images
img = load_sample_images()["images"][0]
plt.imshow(img)
plt.axis("off")
plt.title("Original Image")
plt.show()
data = tf.io.encode_jpeg(img)
example_with_image = Example(features=Features(feature={
"image": Feature(bytes_list=BytesList(value=[data.numpy()]))}))
serialized_example = example_with_image.SerializeToString()
# then save to TFRecord
feature_description = { "image": tf.io.VarLenFeature(tf.string) }
example_with_image = tf.io.parse_single_example(serialized_example, feature_description)
decoded_img = tf.io.decode_jpeg(example_with_image["image"].values[0])
Or use decode_image() which supports BMP, GIF, JPEG and PNG formats:
decoded_img = tf.io.decode_image(example_with_image["image"].values[0])
plt.imshow(decoded_img)
plt.title("Decoded Image")
plt.axis("off")
plt.show()
Tensors can be serialized and parsed easily using tf.io.serialize_tensor() and tf.io.parse_tensor():
t = tf.constant([[0., 1.], [2., 3.], [4., 5.]])
s = tf.io.serialize_tensor(t)
s
tf.io.parse_tensor(s, out_type=tf.float32)
serialized_sparse = tf.io.serialize_sparse(parsed_example["emails"])
serialized_sparse
BytesList(value=serialized_sparse.numpy())
dataset = tf.data.TFRecordDataset(["my_contacts.tfrecord"]).batch(10)
for serialized_examples in dataset:
parsed_examples = tf.io.parse_example(serialized_examples,
feature_description)
parsed_examples
syntax = "proto3";
message FeatureList { repeated Feature feature = 1; };
message FeatureLists { map<string, FeatureList> feature_list = 1; };
message SequenceExample {
Features context = 1;
FeatureLists feature_lists = 2;
};
# WARNING: there's currently a bug preventing "from tensorflow.train import X"
# so we work around it by writing "X = tf.train.X"
#from tensorflow.train import FeatureList, FeatureLists, SequenceExample
FeatureList = tf.train.FeatureList
FeatureLists = tf.train.FeatureLists
SequenceExample = tf.train.SequenceExample
context = Features(feature={
"author_id": Feature(int64_list=Int64List(value=[123])),
"title": Feature(bytes_list=BytesList(value=[b"A", b"desert", b"place", b"."])),
"pub_date": Feature(int64_list=Int64List(value=[1623, 12, 25]))
})
content = [["When", "shall", "we", "three", "meet", "again", "?"],
["In", "thunder", ",", "lightning", ",", "or", "in", "rain", "?"]]
comments = [["When", "the", "hurlyburly", "'s", "done", "."],
["When", "the", "battle", "'s", "lost", "and", "won", "."]]
def words_to_feature(words):
return Feature(bytes_list=BytesList(value=[word.encode("utf-8")
for word in words]))
content_features = [words_to_feature(sentence) for sentence in content]
comments_features = [words_to_feature(comment) for comment in comments]
sequence_example = SequenceExample(
context=context,
feature_lists=FeatureLists(feature_list={
"content": FeatureList(feature=content_features),
"comments": FeatureList(feature=comments_features)
}))
sequence_example
serialized_sequence_example = sequence_example.SerializeToString()
context_feature_descriptions = {
"author_id": tf.io.FixedLenFeature([], tf.int64, default_value=0),
"title": tf.io.VarLenFeature(tf.string),
"pub_date": tf.io.FixedLenFeature([3], tf.int64, default_value=[0, 0, 0]),
}
sequence_feature_descriptions = {
"content": tf.io.VarLenFeature(tf.string),
"comments": tf.io.VarLenFeature(tf.string),
}
parsed_context, parsed_feature_lists = tf.io.parse_single_sequence_example(
serialized_sequence_example, context_feature_descriptions,
sequence_feature_descriptions)
parsed_context
parsed_context["title"].values
parsed_feature_lists
print(tf.RaggedTensor.from_sparse(parsed_feature_lists["content"]))
Let's use the variant of the California housing dataset that we used in Chapter 2, since it contains categorical features and missing values:
import os
import tarfile
import urllib
DOWNLOAD_ROOT = "https://raw.githubusercontent.com/ageron/handson-ml2/master/"
HOUSING_PATH = os.path.join("datasets", "housing")
HOUSING_URL = DOWNLOAD_ROOT + "datasets/housing/housing.tgz"
def fetch_housing_data(housing_url=HOUSING_URL, housing_path=HOUSING_PATH):
os.makedirs(housing_path, exist_ok=True)
tgz_path = os.path.join(housing_path, "housing.tgz")
urllib.request.urlretrieve(housing_url, tgz_path)
housing_tgz = tarfile.open(tgz_path)
housing_tgz.extractall(path=housing_path)
housing_tgz.close()
fetch_housing_data()
import pandas as pd
def load_housing_data(housing_path=HOUSING_PATH):
csv_path = os.path.join(housing_path, "housing.csv")
return pd.read_csv(csv_path)
housing = load_housing_data()
housing.head()
housing_median_age = tf.feature_column.numeric_column("housing_median_age")
age_mean, age_std = X_mean[1], X_std[1] # The median age is column in 1
housing_median_age = tf.feature_column.numeric_column(
"housing_median_age", normalizer_fn=lambda x: (x - age_mean) / age_std)
median_income = tf.feature_column.numeric_column("median_income")
bucketized_income = tf.feature_column.bucketized_column(
median_income, boundaries=[1.5, 3., 4.5, 6.])
bucketized_income
ocean_prox_vocab = ['<1H OCEAN', 'INLAND', 'ISLAND', 'NEAR BAY', 'NEAR OCEAN']
ocean_proximity = tf.feature_column.categorical_column_with_vocabulary_list(
"ocean_proximity", ocean_prox_vocab)
ocean_proximity
# Just an example, it's not used later on
city_hash = tf.feature_column.categorical_column_with_hash_bucket(
"city", hash_bucket_size=1000)
city_hash
bucketized_age = tf.feature_column.bucketized_column(
housing_median_age, boundaries=[-1., -0.5, 0., 0.5, 1.]) # age was scaled
age_and_ocean_proximity = tf.feature_column.crossed_column(
[bucketized_age, ocean_proximity], hash_bucket_size=100)
latitude = tf.feature_column.numeric_column("latitude")
longitude = tf.feature_column.numeric_column("longitude")
bucketized_latitude = tf.feature_column.bucketized_column(
latitude, boundaries=list(np.linspace(32., 42., 20 - 1)))
bucketized_longitude = tf.feature_column.bucketized_column(
longitude, boundaries=list(np.linspace(-125., -114., 20 - 1)))
location = tf.feature_column.crossed_column(
[bucketized_latitude, bucketized_longitude], hash_bucket_size=1000)
ocean_proximity_one_hot = tf.feature_column.indicator_column(ocean_proximity)
ocean_proximity_embed = tf.feature_column.embedding_column(ocean_proximity,
dimension=2)
median_house_value = tf.feature_column.numeric_column("median_house_value")
columns = [housing_median_age, median_house_value]
feature_descriptions = tf.feature_column.make_parse_example_spec(columns)
feature_descriptions
with tf.io.TFRecordWriter("my_data_with_features.tfrecords") as f:
for x, y in zip(X_train[:, 1:2], y_train):
example = Example(features=Features(feature={
"housing_median_age": Feature(float_list=FloatList(value=[x])),
"median_house_value": Feature(float_list=FloatList(value=[y]))
}))
f.write(example.SerializeToString())
def parse_examples(serialized_examples):
examples = tf.io.parse_example(serialized_examples, feature_descriptions)
targets = examples.pop("median_house_value") # separate the targets
return examples, targets
batch_size = 32
dataset = tf.data.TFRecordDataset(["my_data_with_features.tfrecords"])
dataset = dataset.repeat().shuffle(10000).batch(batch_size).map(parse_examples)
Warning: the DenseFeatures layer currently does not work with the Functional API, see TF issue #27416. Hopefully this will be resolved before the final release of TF 2.0.
columns_without_target = columns[:-1]
model = keras.models.Sequential([
keras.layers.DenseFeatures(feature_columns=columns_without_target),
keras.layers.Dense(1)
])
model.compile(loss="mse",
optimizer=keras.optimizers.SGD(lr=1e-3),
metrics=["accuracy"])
model.fit(dataset, steps_per_epoch=len(X_train) // batch_size, epochs=5)
some_columns = [ocean_proximity_embed, bucketized_income]
dense_features = keras.layers.DenseFeatures(some_columns)
dense_features({
"ocean_proximity": [["NEAR OCEAN"], ["INLAND"], ["INLAND"]],
"median_income": [[3.], [7.2], [1.]]
})
try:
import tensorflow_transform as tft
def preprocess(inputs): # inputs is a batch of input features
median_age = inputs["housing_median_age"]
ocean_proximity = inputs["ocean_proximity"]
standardized_age = tft.scale_to_z_score(median_age - tft.mean(median_age))
ocean_proximity_id = tft.compute_and_apply_vocabulary(ocean_proximity)
return {
"standardized_median_age": standardized_age,
"ocean_proximity_id": ocean_proximity_id
}
except ImportError:
print("TF Transform is not installed. Try running: pip3 install -U tensorflow-transform")
import tensorflow_datasets as tfds
datasets = tfds.load(name="mnist")
mnist_train, mnist_test = datasets["train"], datasets["test"]
print(tfds.list_builders())
plt.figure(figsize=(6,3))
mnist_train = mnist_train.repeat(5).batch(32).prefetch(1)
for item in mnist_train:
images = item["image"]
labels = item["label"]
for index in range(5):
plt.subplot(1, 5, index + 1)
image = images[index, ..., 0]
label = labels[index].numpy()
plt.imshow(image, cmap="binary")
plt.title(label)
plt.axis("off")
break # just showing part of the first batch
datasets = tfds.load(name="mnist")
mnist_train, mnist_test = datasets["train"], datasets["test"]
mnist_train = mnist_train.repeat(5).batch(32)
mnist_train = mnist_train.map(lambda items: (items["image"], items["label"]))
mnist_train = mnist_train.prefetch(1)
for images, labels in mnist_train.take(1):
print(images.shape)
print(labels.numpy())
datasets = tfds.load(name="mnist", batch_size=32, as_supervised=True)
mnist_train = datasets["train"].repeat().prefetch(1)
model = keras.models.Sequential([
keras.layers.Flatten(input_shape=[28, 28, 1]),
keras.layers.Lambda(lambda images: tf.cast(images, tf.float32)),
keras.layers.Dense(10, activation="softmax")])
model.compile(loss="sparse_categorical_crossentropy",
optimizer=keras.optimizers.SGD(lr=1e-3),
metrics=["accuracy"])
model.fit(mnist_train, steps_per_epoch=60000 // 32, epochs=5)
import tensorflow_hub as hub
hub_layer = hub.KerasLayer("https://tfhub.dev/google/tf2-preview/nnlm-en-dim50/1",
output_shape=[50], input_shape=[], dtype=tf.string)
model = keras.Sequential()
model.add(hub_layer)
model.add(keras.layers.Dense(16, activation='relu'))
model.add(keras.layers.Dense(1, activation='sigmoid'))
model.summary()
sentences = tf.constant(["It was a great movie", "The actors were amazing"])
embeddings = hub_layer(sentences)
embeddings