wehub-resource-sync
38 lines
1.1 KiB
Python
38 lines
1.1 KiB
Python
#!/usr/bin/env python
|
|
# coding: utf-8
|
|
|
|
from fastai.vision.all import *
|
|
|
|
def get_data(url, presize, resize):
|
|
path = untar_data(url)
|
|
#print(Normalize.from_stats(*imagenet_stats))
|
|
return DataBlock(
|
|
blocks=(ImageBlock, CategoryBlock), get_items=get_image_files,
|
|
splitter=GrandparentSplitter(valid_name='val'),
|
|
get_y=parent_label, item_tfms=Resize(presize),
|
|
batch_tfms=aug_transforms(min_scale=0.5, size=resize),
|
|
).dataloaders(path, bs=128)
|
|
|
|
def block(ni, nf): return ConvLayer(ni, nf, stride=2)
|
|
def get_model():
|
|
return nn.Sequential(
|
|
block(3, 16),
|
|
block(16, 32),
|
|
block(32, 64),
|
|
block(64, 128),
|
|
block(128, 256),
|
|
nn.AdaptiveAvgPool2d(1),
|
|
Flatten(),
|
|
nn.Linear(256, dls.c))
|
|
|
|
def get_learner(dls, m):
|
|
return Learner(dls, m, loss_func=nn.CrossEntropyLoss(), metrics=accuracy
|
|
)
|
|
|
|
if __name__ == "__main__":
|
|
multiprocessing.set_start_method('spawn')
|
|
dls = get_data(URLs.IMAGENETTE_160, 160, 128)
|
|
resnet_model = get_model()
|
|
learn = get_learner(dls, resnet_model)
|
|
learn.lr_find()
|