train_utils.py 7.72 KB
Newer Older
1
2
3
4
5
6
7
8
9
# @author lucasmiranda42
# encoding: utf-8
# module deepof

"""

Simple utility functions used in deepof example scripts. These are not part of the main package

"""
10
from datetime import date, datetime
11

12
from kerastuner import BayesianOptimization, Hyperband
13
from kerastuner import HyperParameters
14
from kerastuner_tensorboard_logger import TensorBoardLogger
15
from typing import Tuple, Union, Any, List
16
17
18
19
20
21
22
import deepof.hypermodels
import deepof.model_utils
import numpy as np
import os
import pickle
import tensorflow as tf

23
24
hp = HyperParameters()

25

26
def load_hparams(hparams):
27
28
29
30
31
32
33
34
    """Loads hyperparameters from a custom dictionary pickled on disc.
    Thought to be used with the output of hyperparameter_tuning.py"""

    if hparams is not None:
        with open(hparams, "rb") as handle:
            hparams = pickle.load(handle)
    else:
        hparams = {
35
36
37
38
39
            "bidirectional_merge": "ave",
            "clipvalue": 1.0,
            "dense_activation": "relu",
            "dense_layers_per_branch": 1,
            "dropout_rate": 1e-3,
40
            "learning_rate": 1e-3,
41
42
43
            "units_conv": 160,
            "units_dense2": 120,
            "units_lstm": 300,
44
45
46
47
48
49
50
51
52
53
54
        }
    return hparams


def load_treatments(train_path):
    """Loads a dictionary containing the treatments per individual,
    to be loaded as metadata in the coordinates class"""
    try:
        with open(
            os.path.join(
                train_path,
55
                [i for i in os.listdir(train_path) if i.endswith(".pkl")][0],
56
57
58
59
60
61
62
63
64
65
66
            ),
            "rb",
        ) as handle:
            treatment_dict = pickle.load(handle)
    except IndexError:
        treatment_dict = None

    return treatment_dict


def get_callbacks(
lucas_miranda's avatar
lucas_miranda committed
67
68
69
70
    X_train: np.array,
    batch_size: int,
    cp: bool,
    variational: bool,
71
    phenotype_class: float,
lucas_miranda's avatar
lucas_miranda committed
72
73
    predictor: float,
    loss: str,
74
    logparam: dict = None,
75
) -> List[Union[Any]]:
76
    """Generates callbacks for model training, including:
77
78
79
80
    - run_ID: run name, with coarse parameter details;
    - tensorboard_callback: for real-time visualization;
    - cp_callback: for checkpoint saving,
    - onecycle: for learning rate scheduling"""
81

82
    run_ID = "{}{}{}{}{}_{}".format(
83
84
        ("GMVAE" if variational else "AE"),
        ("P" if predictor > 0 and variational else ""),
85
        ("_Pheno" if phenotype_class > 0 else ""),
86
        ("_loss={}".format(loss) if variational else ""),
87
88
89
90
91
        (
            "_{}={}".format(list(logparam.keys())[0], list(logparam.values())[0])
            if logparam is not None
            else ""
        ),
92
93
94
95
        datetime.now().strftime("%Y%m%d-%H%M%S"),
    )

    log_dir = os.path.abspath("logs/fit/{}".format(run_ID))
96
    tensorboard_callback = tf.keras.callbacks.TensorBoard(
97
98
99
        log_dir=log_dir,
        histogram_freq=1,
        profile_batch=2,
100
101
102
    )

    onecycle = deepof.model_utils.one_cycle_scheduler(
103
104
        X_train.shape[0] // batch_size * 250,
        max_rate=0.005,
105
106
    )

107
108
109
110
111
112
113
114
115
116
117
118
119
    callbacks = [run_ID, tensorboard_callback, onecycle]

    if cp:
        cp_callback = tf.keras.callbacks.ModelCheckpoint(
            "./logs/checkpoints/" + run_ID + "/cp-{epoch:04d}.ckpt",
            verbose=1,
            save_best_only=False,
            save_weights_only=True,
            save_freq="epoch",
        )
        callbacks.append(cp_callback)

    return callbacks
120
121
122


def tune_search(
123
    data: List[np.array],
124
    encoding_size: int,
125
126
    hypertun_trials: int,
    hpt_type: str,
127
128
    hypermodel: str,
    k: int,
129
    kl_warmup_epochs: int,
130
    loss: str,
131
    mmd_warmup_epochs: int,
132
    overlap_loss: float,
133
    pheno_class: float,
134
135
    predictor: float,
    project_name: str,
136
    callbacks: List,
137
    n_epochs: int = 30,
138
    n_replicas: int = 1,
139
) -> Union[bool, Tuple[Any, Any]]:
140
141
    """Define the search space using keras-tuner and bayesian optimization

142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
    Parameters:
        - train (np.array): dataset to train the model on
        - test (np.array): dataset to validate the model on
        - hypertun_trials (int): number of Bayesian optimization iterations to run
        - hpt_type (str): specify one of Bayesian Optimization (bayopt) and Hyperband (hyperband)
        - hypermodel (str): hypermodel to load. Must be one of S2SAE (plain autoencoder)
        or S2SGMVAE (Gaussian Mixture Variational autoencoder).
        - k (int) number of components of the Gaussian Mixture
        - loss (str): one of [ELBO, MMD, ELBO+MMD]
        - overlap_loss (float): assigns as weight to an extra loss term which
        penalizes overlap between GM components
        - pheno_class (float): adds an extra regularizing neural network to the model,
        which tries to predict the phenotype of the animal from which the sequence comes
        - predictor (float): adds an extra regularizing neural network to the model,
        which tries to predict the next frame from the current one
        - project_name (str): ID of the current run
        - callbacks (list): list of callbacks for the training loop
        - n_epochs (int): optional. Number of epochs to train each run for
        - n_replicas (int): optional. Number of replicas per parameter set. Higher values
         will yield more robust results, but will affect performance severely

    Returns:
        - best_hparams (dict): dictionary with the best retrieved hyperparameters
        - best_run (tf.keras.Model): trained instance of the best model found
166
167
168

    """

169
170
    X_train, y_train, X_val, y_val = data

171
172
173
174
    assert hpt_type in ["bayopt", "hyperband"], (
        "Invalid hyperparameter tuning framework. " "Select one of bayopt and hyperband"
    )

lucas_miranda's avatar
lucas_miranda committed
175
    if hypermodel == "S2SAE":  # pragma: no cover
176
177
178
        assert (
            predictor == 0.0 and pheno_class == 0.0
        ), "Prediction branches are only available for variational models. See documentation for more details"
179
        hypermodel = deepof.hypermodels.SEQ_2_SEQ_AE(input_shape=X_train.shape)
180
181
182

    elif hypermodel == "S2SGMVAE":
        hypermodel = deepof.hypermodels.SEQ_2_SEQ_GMVAE(
183
            input_shape=X_train.shape,
184
            encoding=encoding_size,
185
            kl_warmup_epochs=kl_warmup_epochs,
186
            loss=loss,
187
            mmd_warmup_epochs=mmd_warmup_epochs,
188
            number_of_components=k,
189
            overlap_loss=overlap_loss,
190
            phenotype_predictor=pheno_class,
191
            predictor=predictor,
192
        )
lucas_miranda's avatar
lucas_miranda committed
193

194
195
196
    else:
        return False

197
198
199
200
201
202
203
204
205
206
207
208
209
210
    hpt_params = {
        "hypermodel": hypermodel,
        "executions_per_trial": n_replicas,
        "logger": TensorBoardLogger(metrics=["val_mae"], logdir="./logs/hparams"),
        "objective": "val_mae",
        "project_name": project_name,
        "seed": 42,
        "tune_new_entries": True,
    }

    if hpt_type == "hyperband":
        tuner = Hyperband(
            directory="HyperBandx_{}_{}".format(loss, str(date.today())),
            max_epochs=hypertun_trials,
211
            hyperband_iterations=3,
lucas_miranda's avatar
lucas_miranda committed
212
            factor=2,
213
214
215
216
217
218
219
220
            **hpt_params
        )
    else:
        tuner = BayesianOptimization(
            directory="BayOpt_{}_{}".format(loss, str(date.today())),
            max_trials=hypertun_trials,
            **hpt_params
        )
221
222
223

    print(tuner.search_space_summary())

224
225
226
227
228
229
230
231
232
233
234
    Xs, ys = [X_train], [X_train]
    Xvals, yvals = [X_val], [X_val]

    if predictor > 0.0:
        Xs, ys = X_train[:-1], [X_train[:-1], X_train[1:]]
        Xvals, yvals = X_val[:-1], [X_val[:-1], X_val[1:]]

    if pheno_class > 0.0:
        ys += [y_train]
        yvals += [y_val]

235
    tuner.search(
236
237
        Xs,
        ys,
238
        epochs=n_epochs,
239
        validation_data=(Xvals, yvals),
240
241
        verbose=1,
        batch_size=256,
lucas_miranda's avatar
lucas_miranda committed
242
        callbacks=callbacks,
243
244
245
246
247
    )

    best_hparams = tuner.get_best_hyperparameters(num_trials=1)[0]
    best_run = tuner.hypermodel.build(best_hparams)

lucas_miranda's avatar
lucas_miranda committed
248
249
    print(tuner.results_summary())

250
    return best_hparams, best_run