The fastai library simplifies training fast and accurate neural nets using modern best practices. See the fastai website to get started. The library is based on research into deep learning best practices undertaken at fast.ai
, and includes “out of the box” support for vision
, text
, tabular
, and collab
(collaborative filtering) models.
Download and prepare data:
URLs_MNIST_SAMPLE()
Transformations:
# transformations
= aug_transforms(do_flip = FALSE)
tfms = 'mnist_sample'
path = 20
bs
#load into memory
= ImageDataLoaders_from_folder(path, batch_tfms = tfms, size = 26, bs = bs)
data
= cnn_learner(data, resnet18(), metrics = accuracy) learn
Cbs
argument means callbacks:
%>% fit_one_cycle(1, cbs = TerminateOnNaNCallback()) learn
%>% fit_one_cycle(10, cbs = EarlyStoppingCallback(monitor='valid_loss', patience = 1)) learn
epoch train_loss valid_loss accuracy time
0 0.023524 0.009781 0.996565 00:16
1 0.033328 0.019839 0.993621 00:16
No improvement since epoch 0: early stopping
Save best model for each epoch:
= cnn_learner(data, resnet18(), metrics = accuracy, path = getwd())
learn
%>% fit_one_cycle(3, cbs = SaveModelCallback(every_epoch = TRUE, fname = 'model')) learn
See folder:
list.files('models')
# [1] "model_0.pth" "model_1.pth" "model_2.pth"
# [1] "model_0.pth" "model_1.pth" "model_2.pth"
Decrease learning rate if loss is not improved:
%>% fit_one_cycle(10, 1e-2, cbs = ReduceLROnPlateau(monitor='valid_loss', patience = 1)) learn
epoch train_loss valid_loss accuracy time
0 0.117138 0.038180 0.987242 00:17
1 0.140064 0.006160 0.996565 00:16
2 0.133680 0.061945 0.985770 00:16
Epoch 2: reducing lr to 0.0009891441414237997
3 0.049780 0.005699 0.998037 00:16
4 0.040660 0.019514 0.994112 00:16
Epoch 4: reducing lr to 0.0007502954607977343
5 0.027146 0.009783 0.997056 00:16
Epoch 5: reducing lr to 0.0005526052040192481
6 0.024709 0.008050 0.998528 00:16
Epoch 6: reducing lr to 0.0003458198506447947
7 0.016352 0.010778 0.998037 00:16
Epoch 7: reducing lr to 0.0001656946233635187
8 0.071180 0.009519 0.998528 00:16
Epoch 8: reducing lr to 4.337456332530222e-05
9 0.014804 0.005769 0.998528 00:16
Epoch 9: reducing lr to 1.0114427793916913e-08
Or add new parameter min_lr
:
%>% fit_one_cycle(10, 1e-2, cbs = ReduceLROnPlateau(monitor='valid_loss',
learn min_delta=0.1, patience = 1, min_lr = 1e-8))
Save train history. In addition, for multiple callbacks it is important to pass them within list:
= cnn_learner(data, resnet18(), metrics = accuracy, path = getwd())
learn
%>% fit_one_cycle(2, cbs = list(CSVLogger(),
learn ReduceLROnPlateau(monitor='valid_loss',
min_delta=0.1, patience = 1, min_lr = 1e-8)))
= read.csv('history.csv')
history history
epoch train_loss valid_loss accuracy time
1 0 0.15677054 0.09788394 0.9646713 00:17
2 1 0.08268011 0.05654754 0.9803729 00:17