Image Prediction - Search Space and Hyperparameter Optimization (HPO)

While the Image Prediction - Quick Start introduced basic usage of AutoGluon fit, evaluate, predict with default configurations, this tutorial dives into the various options that you can specify for more advanced control over the fitting process.

These options include: - Defining the search space of various hyperparameter values for the training of neural networks - Specifying how to search through your choosen hyperparameter space - Specifying how to schedule jobs to train a network under a particular hyperparameter configuration.

The advanced functionalities of AutoGluon enable you to use your external knowledge about your particular prediction problem and computing resources to guide the training process. If properly used, you may be able to achieve superior performance within less training time.

Tip: If you are new to AutoGluon, review Image Prediction - Quick Start to learn the basics of the AutoGluon API.

Since our task is to classify images, we will use AutoGluon to produce an ImagePredictor:

import autogluon.core as ag
from autogluon.vision import ImagePredictor, ImageDataset
/var/lib/jenkins/workspace/workspace/autogluon-tutorial-image-classification-v3/venv/lib/python3.7/site-packages/gluoncv/__init__.py:40: UserWarning: Both mxnet==1.7.0 and torch==1.7.1+cu101 are installed. You might encounter increased GPU memory footprint if both framework are used at the same time.
  warnings.warn(f'Both mxnet=={mx.__version__} and torch=={torch.__version__} are installed. '

Create AutoGluon Dataset

Let’s first create the dataset using the same subset of the Shopee-IET dataset as the Image Prediction - Quick Start tutorial. Recall that there’s no validation split in original data, a 90/10 train/validation split is automatically performed when fit with train_data.

train_data, _, test_data = ImageDataset.from_folders('https://autogluon.s3.amazonaws.com/datasets/shopee-iet.zip')
data/
├── test/
└── train/

Specify which Networks to Try

We start with specifying the pretrained neural network candidates. Given such a list, AutoGluon tries to train different networks from this list to identify the best-performing candidate. This is an example of a autogluon.core.space.Categorical search space, in which there are a limited number of values to choose from.

model = ag.Categorical('resnet18_v1b', 'mobilenetv3_small')

# you may choose more than 70+ available model in the model zoo provided by GluonCV:
model_list = ImagePredictor.list_models()

Specify the training hyper-parameters

Similarly, we can manually specify many crucial hyper-parameters, with specific value or search space(autogluon.core.space).

batch_size = 8
lr = ag.Categorical(1e-2, 1e-3)

Search Algorithms

In AutoGluon, autogluon.core.searcher supports different search search strategies for both hyperparameter optimization and architecture search. Beyond simply specifying the space of hyperparameter configurations to search over, you can also tell AutoGluon what strategy it should employ to actually search through this space. This process of finding good hyperparameters from a given search space is commonly referred to as hyperparameter optimization (HPO) or hyperparameter tuning. autogluon.core.scheduler orchestrates how individual training jobs are scheduled. We currently support FIFO (standard) and Hyperband scheduling, along with search by random sampling or Bayesian optimization. These basic techniques are rendered surprisingly powerful by AutoGluon’s support of asynchronous parallel execution.

Bayesian Optimization

Here is an example of using Bayesian Optimization using autogluon.core.searcher.GPFIFOSearcher.

Bayesian Optimization fits a probabilistic surrogate model to estimate the function that relates each hyperparameter configuration to the resulting performance of a model trained under this hyperparameter configuration. Our implementation makes use of a Gaussian process surrogate model along with expected improvement as acquisition function. It has been developed specifically to support asynchronous parallel evaluations.

hyperparameters={'model': model, 'batch_size': batch_size, 'lr': lr, 'epochs': 2}
predictor = ImagePredictor()
predictor.fit(train_data, time_limit=60*10, hyperparameters=hyperparameters,
              hyperparameter_tune_kwargs={'searcher': 'bayesopt', 'num_trials': 2})
print('Top-1 val acc: %.3f' % predictor.fit_summary()['valid_acc'])
Reset labels to [0, 1, 2, 3]
Randomly split train_data into train[720]/validation[80] splits.
The number of requested GPUs is greater than the number of available GPUs.Reduce the number to 1
Starting HPO experiments
  0%|          | 0/2 [00:00<?, ?it/s]
modified configs(<old> != <new>): {
root.train.data_dir  ~/.mxnet/datasets/imagenet != auto
root.train.rec_val_idx ~/.mxnet/datasets/imagenet/rec/val.idx != auto
root.train.epochs    10 != 2
root.train.num_training_samples 1281167 != -1
root.train.num_workers 4 != 8
root.train.batch_size 128 != 8
root.train.lr        0.1 != 0.01
root.train.early_stop_patience -1 != 10
root.train.early_stop_max_value 1.0 != inf
root.train.rec_val   ~/.mxnet/datasets/imagenet/rec/val.rec != auto
root.train.rec_train_idx ~/.mxnet/datasets/imagenet/rec/train.idx != auto
root.train.early_stop_baseline 0.0 != -inf
root.train.rec_train ~/.mxnet/datasets/imagenet/rec/train.rec != auto
root.img_cls.model   resnet50_v1 != resnet18_v1b
root.valid.num_workers 4 != 8
root.valid.batch_size 128 != 8
}
Saved config to /var/lib/jenkins/workspace/workspace/autogluon-tutorial-image-classification-v3/docs/_build/eval/tutorials/image_prediction/d74b7b45/.trial_0/config.yaml
Start training from [Epoch 0]
Epoch[0] Batch [49] Speed: 231.582461 samples/sec   accuracy=0.392500       lr=0.010000
[Epoch 0] training: accuracy=0.473611
[Epoch 0] speed: 243 samples/sec    time cost: 2.924711
[Epoch 0] validation: top1=0.737500 top5=1.000000
[Epoch 0] Current best top-1: 0.737500 vs previous -inf, saved to /var/lib/jenkins/workspace/workspace/autogluon-tutorial-image-classification-v3/docs/_build/eval/tutorials/image_prediction/d74b7b45/.trial_0/best_checkpoint.pkl
Epoch[1] Batch [49] Speed: 232.298602 samples/sec   accuracy=0.625000       lr=0.010000
[Epoch 1] training: accuracy=0.623611
[Epoch 1] speed: 243 samples/sec    time cost: 2.927748
[Epoch 1] validation: top1=0.812500 top5=1.000000
[Epoch 1] Current best top-1: 0.812500 vs previous 0.737500, saved to /var/lib/jenkins/workspace/workspace/autogluon-tutorial-image-classification-v3/docs/_build/eval/tutorials/image_prediction/d74b7b45/.trial_0/best_checkpoint.pkl
Applying the state from the best checkpoint...
modified configs(<old> != <new>): {
root.train.data_dir  ~/.mxnet/datasets/imagenet != auto
root.train.rec_val_idx ~/.mxnet/datasets/imagenet/rec/val.idx != auto
root.train.epochs    10 != 2
root.train.num_training_samples 1281167 != -1
root.train.num_workers 4 != 8
root.train.batch_size 128 != 8
root.train.lr        0.1 != 0.001
root.train.early_stop_patience -1 != 10
root.train.early_stop_max_value 1.0 != inf
root.train.rec_val   ~/.mxnet/datasets/imagenet/rec/val.rec != auto
root.train.rec_train_idx ~/.mxnet/datasets/imagenet/rec/train.idx != auto
root.train.early_stop_baseline 0.0 != -inf
root.train.rec_train ~/.mxnet/datasets/imagenet/rec/train.rec != auto
root.img_cls.model   resnet50_v1 != resnet18_v1b
root.valid.num_workers 4 != 8
root.valid.batch_size 128 != 8
}
Saved config to /var/lib/jenkins/workspace/workspace/autogluon-tutorial-image-classification-v3/docs/_build/eval/tutorials/image_prediction/d74b7b45/.trial_1/config.yaml
Start training from [Epoch 0]
Epoch[0] Batch [49] Speed: 225.250351 samples/sec   accuracy=0.305000       lr=0.001000
[Epoch 0] training: accuracy=0.293056
[Epoch 0] speed: 238 samples/sec    time cost: 2.986606
[Epoch 0] validation: top1=0.412500 top5=1.000000
[Epoch 0] Current best top-1: 0.412500 vs previous -inf, saved to /var/lib/jenkins/workspace/workspace/autogluon-tutorial-image-classification-v3/docs/_build/eval/tutorials/image_prediction/d74b7b45/.trial_1/best_checkpoint.pkl
Epoch[1] Batch [49] Speed: 249.459677 samples/sec   accuracy=0.360000       lr=0.001000
[Epoch 1] training: accuracy=0.393056
[Epoch 1] speed: 253 samples/sec    time cost: 2.810767
[Epoch 1] validation: top1=0.475000 top5=1.000000
[Epoch 1] Current best top-1: 0.475000 vs previous 0.412500, saved to /var/lib/jenkins/workspace/workspace/autogluon-tutorial-image-classification-v3/docs/_build/eval/tutorials/image_prediction/d74b7b45/.trial_1/best_checkpoint.pkl
Applying the state from the best checkpoint...
Finished, total runtime is 22.65 s
{ 'best_config': { 'estimator': <class 'gluoncv.auto.estimators.image_classification.image_classification.ImageClassificationEstimator'>,
                   'gpus': [0],
                   'img_cls': { 'batch_norm': False,
                                'last_gamma': False,
                                'model': 'resnet18_v1b',
                                'use_gn': False,
                                'use_pretrained': True,
                                'use_se': False},
                   'train': { 'batch_size': 8,
                              'crop_ratio': 0.875,
                              'data_dir': 'auto',
                              'dtype': 'float32',
                              'early_stop_baseline': -inf,
                              'early_stop_max_value': inf,
                              'early_stop_min_delta': 0.001,
                              'early_stop_patience': 10,
                              'epochs': 2,
                              'hard_weight': 0.5,
                              'input_size': 224,
                              'label_smoothing': False,
                              'log_interval': 50,
                              'lr': 0.01,
                              'lr_decay': 0.1,
                              'lr_decay_epoch': '40, 60',
                              'lr_decay_period': 0,
                              'lr_mode': 'step',
                              'mixup': False,
                              'mixup_alpha': 0.2,
                              'mixup_off_epoch': 0,
                              'mode': '',
                              'momentum': 0.9,
                              'no_wd': False,
                              'num_training_samples': -1,
                              'num_workers': 8,
                              'output_lr_mult': 0.1,
                              'pretrained_base': True,
                              'rec_train': 'auto',
                              'rec_train_idx': 'auto',
                              'rec_val': 'auto',
                              'rec_val_idx': 'auto',
                              'resume_epoch': 0,
                              'start_epoch': 0,
                              'teacher': None,
                              'temperature': 20,
                              'transfer_lr_mult': 0.01,
                              'use_rec': False,
                              'warmup_epochs': 0,
                              'warmup_lr': 0.0,
                              'wd': 0.0001},
                   'valid': {'batch_size': 8, 'num_workers': 8}},
  'total_time': 22.650203227996826,
  'train_acc': 0.39305555555555555,
  'valid_acc': 0.475}
Top-1 val acc: 0.475

The BO searcher can be configured by search_options, see autogluon.core.searcher.GPFIFOSearcher. Load the test dataset and evaluate:

results = predictor.evaluate(test_data)
print('Test acc on hold-out data:', results)
Test acc on hold-out data: {'top1': 0.7625, 'top5': 1.0}

Note that num_trials=2 above is only used to speed up the tutorial. In normal practice, it is common to only use time_limit and drop num_trials.