train_utils.py 7.77 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
    outpath: str = ".",
76
) -> List[Union[Any]]:
77
    """Generates callbacks for model training, including:
78
79
80
81
    - run_ID: run name, with coarse parameter details;
    - tensorboard_callback: for real-time visualization;
    - cp_callback: for checkpoint saving,
    - onecycle: for learning rate scheduling"""
82

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

96
    log_dir = os.path.abspath(os.path.join(outpath, "fit", run_ID))
97
    tensorboard_callback = tf.keras.callbacks.TensorBoard(
98
99
100
        log_dir=log_dir,
        histogram_freq=1,
        profile_batch=2,
101
102
103
    )

    onecycle = deepof.model_utils.one_cycle_scheduler(
104
105
        X_train.shape[0] // batch_size * 250,
        max_rate=0.005,
106
107
    )

108
109
110
111
    callbacks = [run_ID, tensorboard_callback, onecycle]

    if cp:
        cp_callback = tf.keras.callbacks.ModelCheckpoint(
112
            os.path.join(outpath,"checkpoints", run_ID + "/cp-{epoch:04d}.ckpt"),
113
114
115
116
117
118
119
120
            verbose=1,
            save_best_only=False,
            save_weights_only=True,
            save_freq="epoch",
        )
        callbacks.append(cp_callback)

    return callbacks
121
122
123


def tune_search(
124
    data: List[np.array],
125
    encoding_size: int,
126
127
    hypertun_trials: int,
    hpt_type: str,
128
129
    hypermodel: str,
    k: int,
130
    kl_warmup_epochs: int,
131
    loss: str,
132
    mmd_warmup_epochs: int,
133
    overlap_loss: float,
134
    pheno_class: float,
135
136
    predictor: float,
    project_name: str,
137
    callbacks: List,
138
    n_epochs: int = 30,
139
    n_replicas: int = 1,
140
) -> Union[bool, Tuple[Any, Any]]:
141
142
    """Define the search space using keras-tuner and bayesian optimization

143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
    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
167
168
169

    """

170
171
    X_train, y_train, X_val, y_val = data

172
173
174
175
    assert hpt_type in ["bayopt", "hyperband"], (
        "Invalid hyperparameter tuning framework. " "Select one of bayopt and hyperband"
    )

lucas_miranda's avatar
lucas_miranda committed
176
    if hypermodel == "S2SAE":  # pragma: no cover
177
178
179
        assert (
            predictor == 0.0 and pheno_class == 0.0
        ), "Prediction branches are only available for variational models. See documentation for more details"
180
        hypermodel = deepof.hypermodels.SEQ_2_SEQ_AE(input_shape=X_train.shape)
181
182
183

    elif hypermodel == "S2SGMVAE":
        hypermodel = deepof.hypermodels.SEQ_2_SEQ_GMVAE(
184
            input_shape=X_train.shape,
185
            encoding=encoding_size,
186
            kl_warmup_epochs=kl_warmup_epochs,
187
            loss=loss,
188
            mmd_warmup_epochs=mmd_warmup_epochs,
189
            number_of_components=k,
190
            overlap_loss=overlap_loss,
191
            phenotype_predictor=pheno_class,
192
            predictor=predictor,
193
        )
lucas_miranda's avatar
lucas_miranda committed
194

195
196
197
    else:
        return False

198
199
200
201
202
203
204
205
206
207
208
209
210
211
    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,
212
            hyperband_iterations=3,
lucas_miranda's avatar
lucas_miranda committed
213
            factor=2,
214
215
216
217
218
219
220
221
            **hpt_params
        )
    else:
        tuner = BayesianOptimization(
            directory="BayOpt_{}_{}".format(loss, str(date.today())),
            max_trials=hypertun_trials,
            **hpt_params
        )
222
223
224

    print(tuner.search_space_summary())

225
226
227
228
229
230
231
232
233
234
235
    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]

236
    tuner.search(
237
238
        Xs,
        ys,
239
        epochs=n_epochs,
240
        validation_data=(Xvals, yvals),
241
242
        verbose=1,
        batch_size=256,
lucas_miranda's avatar
lucas_miranda committed
243
        callbacks=callbacks,
244
245
246
247
248
    )

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

lucas_miranda's avatar
lucas_miranda committed
249
250
    print(tuner.results_summary())

251
    return best_hparams, best_run