The hardware and bandwidth for this mirror is donated by METANET, the Webhosting and Full Service-Cloud Provider.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]metanet.ch.
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.
Grab the pets dataset and Specify folders:
URLs_PETS()
path = 'oxford-iiit-pet'
path_hr = paste(path, 'images', sep = '/')
path_lr = paste(path, 'crappy', sep = '/')
Prepare the input data by crappifying images:
bs = 10
size = 64
arch = resnet34()
get_dls = function(bs, size) {
dblock = DataBlock(blocks = list(ImageBlock, ImageBlock),
get_items = get_image_files,
get_y = function(x) {paste(path_hr, as.character(x$name), sep = '/')},
splitter = RandomSplitter(),
item_tfms = Resize(size),
batch_tfms = list(
aug_transforms(max_zoom = 2.),
Normalize_from_stats( imagenet_stats() )
))
dls = dblock %>% dataloaders(path_lr, bs = bs, path = path)
dls$c = 3L
dls
}
dls_gen = get_dls(bs, size)
See batch:
Define loss function and create unet_learner
:
wd = 1e-3
y_range = c(-3.,3.)
loss_gen = MSELossFlat()
create_gen_learner = function() {
unet_learner(dls_gen, arch, loss_func = loss_gen,
config = unet_config(blur=TRUE, norm_type = "Weight",
self_attention = TRUE, y_range = y_range))
}
learn_gen = create_gen_learner()
learn_gen %>% fit_one_cycle(2, pct_start = 0.8, wd = wd)
epoch train_loss valid_loss time
0 0.025911 0.035153 00:42
1 0.019524 0.019408 00:39
These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.