train_utils.py 12.2 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 tensorboard.plugins.hparams import api as hp
16
from typing import Tuple, Union, Any, List
17
18
19
20
21
22
23
import deepof.hypermodels
import deepof.model_utils
import numpy as np
import os
import pickle
import tensorflow as tf

24
25
hp = HyperParameters()

26

27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
class CustomStopper(tf.keras.callbacks.EarlyStopping):
    """ Custom callback for """

    def __init__(self, start_epoch, *args, **kwargs):
        super(CustomStopper, self).__init__(*args, **kwargs)
        self.start_epoch = start_epoch

    def get_config(self):  # pragma: no cover
        """Updates callback metadata"""

        config = super().get_config().copy()
        config.update({"start_epoch": self.start_epoch})
        return config

    def on_epoch_end(self, epoch, logs=None):
        if epoch > self.start_epoch:
            super().on_epoch_end(epoch, logs)


46
def load_hparams(hparams):
47
48
49
50
51
52
53
54
    """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 = {
55
56
57
58
59
            "bidirectional_merge": "ave",
            "clipvalue": 1.0,
            "dense_activation": "relu",
            "dense_layers_per_branch": 1,
            "dropout_rate": 1e-3,
60
            "learning_rate": 1e-3,
61
62
63
            "units_conv": 160,
            "units_dense2": 120,
            "units_lstm": 300,
64
65
66
67
68
69
70
71
72
73
74
        }
    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,
75
                [i for i in os.listdir(train_path) if i.endswith(".pkl")][0],
76
77
78
79
80
81
82
83
84
85
86
            ),
            "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
87
88
89
90
    X_train: np.array,
    batch_size: int,
    cp: bool,
    variational: bool,
91
    phenotype_class: float,
lucas_miranda's avatar
lucas_miranda committed
92
93
    predictor: float,
    loss: str,
94
    logparam: dict = None,
95
    outpath: str = ".",
96
) -> List[Union[Any]]:
97
    """Generates callbacks for model training, including:
98
99
100
101
    - run_ID: run name, with coarse parameter details;
    - tensorboard_callback: for real-time visualization;
    - cp_callback: for checkpoint saving,
    - onecycle: for learning rate scheduling"""
102

103
    run_ID = "{}{}{}{}{}{}_{}".format(
104
        ("GMVAE" if variational else "AE"),
105
106
        ("Pred={}".format(predictor) if predictor > 0 and variational else ""),
        ("_Pheno={}".format(phenotype_class) if phenotype_class > 0 else ""),
107
        ("_loss={}".format(loss) if variational else ""),
108
109
        ("_encoding={}".format(logparam["encoding"]) if logparam is not None else ""),
        ("_k={}".format(logparam["k"]) if logparam is not None else ""),
110
        (datetime.now().strftime("%Y%m%d-%H%M%S")),
111
112
    )

113
    log_dir = os.path.abspath(os.path.join(outpath, "fit", run_ID))
114
    tensorboard_callback = tf.keras.callbacks.TensorBoard(
115
116
117
        log_dir=log_dir,
        histogram_freq=1,
        profile_batch=2,
118
119
120
    )

    onecycle = deepof.model_utils.one_cycle_scheduler(
121
122
        X_train.shape[0] // batch_size * 250,
        max_rate=0.005,
123
124
    )

125
126
127
128
    callbacks = [run_ID, tensorboard_callback, onecycle]

    if cp:
        cp_callback = tf.keras.callbacks.ModelCheckpoint(
129
            os.path.join(outpath, "checkpoints", run_ID + "/cp-{epoch:04d}.ckpt"),
130
131
132
133
134
135
136
137
            verbose=1,
            save_best_only=False,
            save_weights_only=True,
            save_freq="epoch",
        )
        callbacks.append(cp_callback)

    return callbacks
138
139


140
141
142
def deep_unsupervised_embedding():
    """Implementation function for deepof.data.coordinates.deep_unsupervised_embedding"""

143
    # Load data
144
145
146
147
148
    X_train, y_train, X_val, y_val = preprocessed_object

    # To avoid stability issues
    tf.keras.backend.clear_session()

149
    # Load callbacks
150
151
152
153
154
155
156
157
158
159
160
161
    run_ID, tensorboard_callback, onecycle, cp_callback = get_callbacks(
        X_train=X_train,
        batch_size=batch_size,
        cp=save_checkpoints,
        variational=variational,
        phenotype_class=pheno_class,
        predictor=predictor,
        loss=loss,
        logparam=logparam,
        outpath=output_path,
    )

162
    # Build models
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
    if not variational:
        encoder, decoder, ae = deepof.models.SEQ_2_SEQ_AE(
            ({} if hparams is None else hparams)
        ).build(X_train.shape)
        return_list = (encoder, decoder, ae)

    else:
        (
            encoder,
            generator,
            grouper,
            ae,
            kl_warmup_callback,
            mmd_warmup_callback,
        ) = deepof.models.SEQ_2_SEQ_GMVAE(
            architecture_hparams=({} if hparams is None else hparams),
            batch_size=batch_size,
            compile_model=True,
            encoding=encoding_size,
            kl_warmup_epochs=kl_warmup,
            loss=loss,
            mmd_warmup_epochs=mmd_warmup,
            montecarlo_kl=montecarlo_kl,
            neuron_control=False,
            number_of_components=n_components,
            overlap_loss=False,
            phenotype_prediction=phenotype_class,
            predictor=predictor,
        ).build(
            X_train.shape
        )
        return_list = (encoder, generator, grouper, ae)

    if pretrained:
197
        # If pretrained models are specified, load weights and return
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
        ae.load_weights(pretrained)
        return return_list

    else:
        if not variational:

            ae.fit(
                x=X_train,
                y=X_train,
                epochs=35,
                batch_size=batch_size,
                verbose=1,
                validation_data=(X_val, X_val),
                callbacks=[
                    tensorboard_callback,
                    cp_callback,
                    onecycle,
                    CustomStopper(
                        monitor="val_loss",
                        patience=5,
                        restore_best_weights=True,
                        start_epoch=max(kl_wu, mmd_wu),
                    ),
                ],
            )

        else:

226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
            callbacks_ = [
                tensorboard_callback,
                # cp_callback,
                onecycle,
                CustomStopper(
                    monitor="val_loss",
                    patience=5,
                    restore_best_weights=True,
                    start_epoch=max(kl_wu, mmd_wu),
                ),
            ]

            if "ELBO" in loss and kl_wu > 0:
                # noinspection PyUnboundLocalVariable
                callbacks_.append(kl_warmup_callback)
            if "MMD" in loss and mmd_wu > 0:
                # noinspection PyUnboundLocalVariable
                callbacks_.append(mmd_warmup_callback)

245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
            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]

            gmvaep.fit(
                x=Xs,
                y=ys,
                epochs=35,
                batch_size=batch_size,
                verbose=1,
                validation_data=(
                    Xvals,
                    yvals,
                ),
                callbacks=callbacks_,
            )

    return return_list


272
def tune_search(
273
    data: List[np.array],
274
    encoding_size: int,
275
276
    hypertun_trials: int,
    hpt_type: str,
277
278
    hypermodel: str,
    k: int,
279
    kl_warmup_epochs: int,
280
    loss: str,
281
    mmd_warmup_epochs: int,
282
    overlap_loss: float,
283
    pheno_class: float,
284
285
    predictor: float,
    project_name: str,
286
    callbacks: List,
287
    n_epochs: int = 30,
288
    n_replicas: int = 1,
289
) -> Union[bool, Tuple[Any, Any]]:
290
291
    """Define the search space using keras-tuner and bayesian optimization

292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
    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
316
317
318

    """

319
320
    X_train, y_train, X_val, y_val = data

321
322
323
324
    assert hpt_type in ["bayopt", "hyperband"], (
        "Invalid hyperparameter tuning framework. " "Select one of bayopt and hyperband"
    )

lucas_miranda's avatar
lucas_miranda committed
325
    if hypermodel == "S2SAE":  # pragma: no cover
326
327
328
        assert (
            predictor == 0.0 and pheno_class == 0.0
        ), "Prediction branches are only available for variational models. See documentation for more details"
329
        hypermodel = deepof.hypermodels.SEQ_2_SEQ_AE(input_shape=X_train.shape)
330
331
332

    elif hypermodel == "S2SGMVAE":
        hypermodel = deepof.hypermodels.SEQ_2_SEQ_GMVAE(
333
            input_shape=X_train.shape,
334
            encoding=encoding_size,
335
            kl_warmup_epochs=kl_warmup_epochs,
336
            loss=loss,
337
            mmd_warmup_epochs=mmd_warmup_epochs,
338
            number_of_components=k,
339
            overlap_loss=overlap_loss,
340
            phenotype_predictor=pheno_class,
341
            predictor=predictor,
342
        )
lucas_miranda's avatar
lucas_miranda committed
343

344
345
346
    else:
        return False

347
348
349
350
351
352
353
354
355
356
357
358
359
360
    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,
361
            hyperband_iterations=3,
lucas_miranda's avatar
lucas_miranda committed
362
            factor=2,
363
364
365
366
367
368
369
370
            **hpt_params
        )
    else:
        tuner = BayesianOptimization(
            directory="BayOpt_{}_{}".format(loss, str(date.today())),
            max_trials=hypertun_trials,
            **hpt_params
        )
371
372
373

    print(tuner.search_space_summary())

374
375
376
377
378
379
380
381
382
383
384
    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]

385
    tuner.search(
386
387
        Xs,
        ys,
388
        epochs=n_epochs,
389
        validation_data=(Xvals, yvals),
390
391
        verbose=1,
        batch_size=256,
lucas_miranda's avatar
lucas_miranda committed
392
        callbacks=callbacks,
393
394
395
396
397
    )

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

lucas_miranda's avatar
lucas_miranda committed
398
399
    print(tuner.results_summary())

400
    return best_hparams, best_run