-
Notifications
You must be signed in to change notification settings - Fork 995
/
model.py
221 lines (172 loc) · 6.23 KB
/
model.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
"""Data prep, train and evaluate DNN model."""
import logging
import os
import numpy as np
import tensorflow as tf
from tensorflow.keras import callbacks, models
from tensorflow.keras.layers import (
Concatenate,
Dense,
Discretization,
Embedding,
Flatten,
Input,
Lambda,
)
from tensorflow.keras.layers.experimental.preprocessing import HashedCrossing
logging.info(tf.version.VERSION)
CSV_COLUMNS = [
"fare_amount",
"pickup_datetime",
"pickup_longitude",
"pickup_latitude",
"dropoff_longitude",
"dropoff_latitude",
"passenger_count",
"key",
]
LABEL_COLUMN = "fare_amount"
DEFAULTS = [[0.0], ["na"], [0.0], [0.0], [0.0], [0.0], [0.0], ["na"]]
UNWANTED_COLS = ["pickup_datetime", "key"]
INPUT_COLS = [
c for c in CSV_COLUMNS if c != LABEL_COLUMN and c not in UNWANTED_COLS
]
def features_and_labels(row_data):
for unwanted_col in UNWANTED_COLS:
row_data.pop(unwanted_col)
label = row_data.pop(LABEL_COLUMN)
return row_data, label
def load_dataset(pattern, batch_size, num_repeat):
dataset = tf.data.experimental.make_csv_dataset(
file_pattern=pattern,
batch_size=batch_size,
column_names=CSV_COLUMNS,
column_defaults=DEFAULTS,
num_epochs=num_repeat,
shuffle_buffer_size=1000000,
)
return dataset.map(features_and_labels)
def create_train_dataset(pattern, batch_size):
dataset = load_dataset(pattern, batch_size, num_repeat=None)
return dataset.prefetch(1)
def create_eval_dataset(pattern, batch_size):
dataset = load_dataset(pattern, batch_size, num_repeat=1)
return dataset.prefetch(1)
def euclidean(params):
lon1, lat1, lon2, lat2 = params
londiff = lon2 - lon1
latdiff = lat2 - lat1
return tf.sqrt(londiff * londiff + latdiff * latdiff)
def scale_longitude(lon_column):
return (lon_column + 78) / 8.0
def scale_latitude(lat_column):
return (lat_column - 37) / 8.0
def transform(inputs, nbuckets):
transformed = {}
# Scaling longitude from range [-70, -78] to [0, 1]
transformed["scaled_plon"] = Lambda(scale_longitude, name="scale_plon")(
inputs["pickup_longitude"]
)
transformed["scaled_dlon"] = Lambda(scale_longitude, name="scale_dlon")(
inputs["dropoff_longitude"]
)
# Scaling latitude from range [37, 45] to [0, 1]
transformed["scaled_plat"] = Lambda(scale_latitude, name="scale_plat")(
inputs["pickup_latitude"]
)
transformed["scaled_dlat"] = Lambda(scale_latitude, name="scale_dlat")(
inputs["dropoff_latitude"]
)
# Apply euclidean function
transformed["euclidean_distance"] = Lambda(euclidean, name="euclidean")(
[
inputs["pickup_longitude"],
inputs["pickup_latitude"],
inputs["dropoff_longitude"],
inputs["dropoff_latitude"],
]
)
latbuckets = np.linspace(start=0.0, stop=1.0, num=nbuckets).tolist()
lonbuckets = np.linspace(start=0.0, stop=1.0, num=nbuckets).tolist()
# Bucketization with Discretization layer
plon = Discretization(lonbuckets, name="plon_bkt")(
transformed["scaled_plon"]
)
plat = Discretization(latbuckets, name="plat_bkt")(
transformed["scaled_plat"]
)
dlon = Discretization(lonbuckets, name="dlon_bkt")(
transformed["scaled_dlon"]
)
dlat = Discretization(latbuckets, name="dlat_bkt")(
transformed["scaled_dlat"]
)
# Feature Cross with HashedCrossing layer
p_fc = HashedCrossing(num_bins=nbuckets * nbuckets, name="p_fc")(
(plon, plat)
)
d_fc = HashedCrossing(num_bins=nbuckets * nbuckets, name="d_fc")(
(dlon, dlat)
)
pd_fc = HashedCrossing(num_bins=nbuckets**4, name="pd_fc")((p_fc, d_fc))
# Embedding with Embedding layer
transformed["pd_embed"] = Flatten()(
Embedding(input_dim=nbuckets**4, output_dim=10, name="pd_embed")(pd_fc)
)
transformed["passenger_count"] = inputs["passenger_count"]
return transformed
def rmse(y_true, y_pred):
return tf.sqrt(tf.reduce_mean(tf.square(y_pred - y_true)))
def build_dnn_model(nbuckets, nnsize, lr):
inputs = {
colname: Input(name=colname, shape=(1,), dtype="float32")
for colname in INPUT_COLS
}
# transforms
transformed = transform(inputs, nbuckets)
dnn_inputs = Concatenate()(transformed.values())
x = dnn_inputs
for layer, nodes in enumerate(nnsize):
x = Dense(nodes, activation="relu", name=f"h{layer}")(x)
output = Dense(1, name="fare")(x)
model = models.Model(inputs, output)
# TODO 1a
lr_optimizer = tf.keras.optimizers.Adam(learning_rate=lr)
model.compile(optimizer=lr_optimizer, loss="mse", metrics=[rmse, "mse"])
return model
def train_and_evaluate(hparams):
# TODO 1b
batch_size = hparams["batch_size"]
nbuckets = hparams["nbuckets"]
lr = hparams["lr"]
nnsize = [int(s) for s in hparams["nnsize"].split()]
eval_data_path = hparams["eval_data_path"]
num_evals = hparams["num_evals"]
num_examples_to_train_on = hparams["num_examples_to_train_on"]
output_dir = hparams["output_dir"]
train_data_path = hparams["train_data_path"]
model_export_path = os.path.join(output_dir, "savedmodel")
checkpoint_path = os.path.join(output_dir, "checkpoints")
tensorboard_path = os.path.join(output_dir, "tensorboard")
if tf.io.gfile.exists(output_dir):
tf.io.gfile.rmtree(output_dir)
model = build_dnn_model(nbuckets, nnsize, lr)
logging.info(model.summary())
trainds = create_train_dataset(train_data_path, batch_size)
evalds = create_eval_dataset(eval_data_path, batch_size)
steps_per_epoch = num_examples_to_train_on // (batch_size * num_evals)
checkpoint_cb = callbacks.ModelCheckpoint(
checkpoint_path, save_weights_only=True, verbose=1
)
tensorboard_cb = callbacks.TensorBoard(tensorboard_path, histogram_freq=1)
history = model.fit(
trainds,
validation_data=evalds,
epochs=num_evals,
steps_per_epoch=max(1, steps_per_epoch),
verbose=2, # 0=silent, 1=progress bar, 2=one line per epoch
callbacks=[checkpoint_cb, tensorboard_cb],
)
# Exporting the model with default serving function.
model.save(model_export_path)
return history