train_utils.py 14.8 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
143
144
145
146
147
148
149
150
151
152
153
154
155
156
def deep_unsupervised_embedding(
    preprocessed_object: Tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray],
    batch_size: int = 256,
    encoding_size: int = 4,
    hparams: dict = None,
    kl_warmup: int = 0,
    loss: str = "ELBO",
    mmd_warmup: int = 0,
    montecarlo_kl: int = 10,
    n_components: int = 25,
    outpath: str = ".",
    phenotype_class: float = 0,
    predictor: float = 0,
    pretrained: str = False,
    save_checkpoints: bool = True,
    variational: bool = True,
):
157
158
    """Implementation function for deepof.data.coordinates.deep_unsupervised_embedding"""

159
    # Load data
160
161
162
163
164
    X_train, y_train, X_val, y_val = preprocessed_object

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

165
166
167
168
169
170
171
172
173
    # defines what to log on tensorboard (useful for trying out different models)
    logparam = {
        "encoding": encoding_size,
        "k": k,
        "loss": loss,
    }
    if phenotype_class:
        logparam["pheno_weight"] = phenotype_class

174
    # Load callbacks
175
176
177
178
179
    run_ID, tensorboard_callback, onecycle, cp_callback = get_callbacks(
        X_train=X_train,
        batch_size=batch_size,
        cp=save_checkpoints,
        variational=variational,
180
        phenotype_class=phenotype_class,
181
182
183
184
185
186
        predictor=predictor,
        loss=loss,
        logparam=logparam,
        outpath=output_path,
    )

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
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
    logparams = [
        hp.HParam(
            "encoding",
            hp.Discrete([2, 4, 6, 8, 12, 16]),
            display_name="encoding",
            description="encoding size dimensionality",
        ),
        hp.HParam(
            "k",
            hp.IntInterval(min_value=1, max_value=15),
            display_name="k",
            description="cluster_number",
        ),
        hp.HParam(
            "loss",
            hp.Discrete(["ELBO", "MMD", "ELBO+MMD"]),
            display_name="loss function",
            description="loss function",
        ),
        hp.HParam(
            "run",
            hp.Discrete([0, 1, 2]),
            display_name="trial run",
            description="trial run",
        ),
    ]

    rec = "reconstruction_" if phenotype_class else ""
    metrics = [
        hp.Metric("val_{}mae".format(rec), display_name="val_{}mae".format(rec)),
        hp.Metric("val_{}mse".format(rec), display_name="val_{}mse".format(rec)),
    ]
    logparam["run"] = run
    if phenotype_class:
        logparams.append(
            hp.HParam(
                "pheno_weight",
                hp.RealInterval(min_value=0.0, max_value=1000.0),
                display_name="pheno weight",
                description="weight applied to phenotypic classifier from the latent space",
            )
        )
        metrics += [
            hp.Metric(
                "phenotype_prediction_accuracy",
                display_name="phenotype_prediction_accuracy",
            ),
            hp.Metric(
                "phenotype_prediction_auc",
                display_name="phenotype_prediction_auc",
            ),
        ]

    with tf.summary.create_file_writer(
            os.path.join(output_path, "hparams", run_ID)
    ).as_default():
        hp.hparams_config(
            hparams=logparams,
            metrics=metrics,
        )

248
    # Build models
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
    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:
283
        # If pretrained models are specified, load weights and return
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
        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:

312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
            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)

331
332
333
334
335
336
337
            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:]]

338
            if phenotype_class > 0.0:
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
                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


358
def tune_search(
359
    data: List[np.array],
360
    encoding_size: int,
361
362
    hypertun_trials: int,
    hpt_type: str,
363
364
    hypermodel: str,
    k: int,
365
    kl_warmup_epochs: int,
366
    loss: str,
367
    mmd_warmup_epochs: int,
368
    overlap_loss: float,
369
    phenotype_class: float,
370
371
    predictor: float,
    project_name: str,
372
    callbacks: List,
373
    n_epochs: int = 30,
374
    n_replicas: int = 1,
375
) -> Union[bool, Tuple[Any, Any]]:
376
377
    """Define the search space using keras-tuner and bayesian optimization

378
379
380
381
382
383
384
385
386
387
388
    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
389
        - phenotype_class (float): adds an extra regularizing neural network to the model,
390
391
392
393
394
395
396
397
398
399
400
401
        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
402
403
404

    """

405
406
    X_train, y_train, X_val, y_val = data

407
408
409
410
    assert hpt_type in ["bayopt", "hyperband"], (
        "Invalid hyperparameter tuning framework. " "Select one of bayopt and hyperband"
    )

lucas_miranda's avatar
lucas_miranda committed
411
    if hypermodel == "S2SAE":  # pragma: no cover
412
        assert (
413
                predictor == 0.0 and phenotype_class == 0.0
414
        ), "Prediction branches are only available for variational models. See documentation for more details"
415
        hypermodel = deepof.hypermodels.SEQ_2_SEQ_AE(input_shape=X_train.shape)
416
417
418

    elif hypermodel == "S2SGMVAE":
        hypermodel = deepof.hypermodels.SEQ_2_SEQ_GMVAE(
419
            input_shape=X_train.shape,
420
            encoding=encoding_size,
421
            kl_warmup_epochs=kl_warmup_epochs,
422
            loss=loss,
423
            mmd_warmup_epochs=mmd_warmup_epochs,
424
            number_of_components=k,
425
            overlap_loss=overlap_loss,
426
            phenotype_predictor=phenotype_class,
427
            predictor=predictor,
428
        )
lucas_miranda's avatar
lucas_miranda committed
429

430
431
432
    else:
        return False

433
434
435
436
437
438
439
440
441
442
443
444
445
446
    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,
447
            hyperband_iterations=3,
lucas_miranda's avatar
lucas_miranda committed
448
            factor=2,
449
450
451
452
453
454
455
456
            **hpt_params
        )
    else:
        tuner = BayesianOptimization(
            directory="BayOpt_{}_{}".format(loss, str(date.today())),
            max_trials=hypertun_trials,
            **hpt_params
        )
457
458
459

    print(tuner.search_space_summary())

460
461
462
463
464
465
466
    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:]]

467
    if phenotype_class > 0.0:
468
469
470
        ys += [y_train]
        yvals += [y_val]

471
    tuner.search(
472
473
        Xs,
        ys,
474
        epochs=n_epochs,
475
        validation_data=(Xvals, yvals),
476
477
        verbose=1,
        batch_size=256,
lucas_miranda's avatar
lucas_miranda committed
478
        callbacks=callbacks,
479
480
481
482
483
    )

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

lucas_miranda's avatar
lucas_miranda committed
484
485
    print(tuner.results_summary())

486
    return best_hparams, best_run