torch.cuda.empty_cache()
cudnn.benchmark = True
else:
print("no cuda")
quit()
sys.path.append("/d/achanhon/github/EfficientNet-PyTorch")
sys.path.append("/d/achanhon/github/pytorch-image-models")
sys.path.append("/d/achanhon/github/pretrained-models.pytorch")
sys.path.append("/d/achanhon/github/segmentation_models.pytorch")
import segmentation_models_pytorch as smp
import miniworld
print("load data")
miniworlddataset = miniworld.MiniWorld("/train/")
print("define model")
net = smp.Unet(
encoder_name="efficientnet-b7",
encoder_weights="imagenet",
in_channels=3,
classes=2,
)
net = net.cuda()
net.train()
print("train")
weights = torch.Tensor([1, 1]).cuda()
criterion = torch.nn.CrossEntropyLoss(weight=weights, reduction="none")
optimizer = torch.optim.Adam(net.parameters(), lr=0.0001)
for col in range(0, image.shape[3] - tilesize + 1, stride):
tmp = net(image[:, :, row:row + tilesize, col:col + tilesize])
pred[0, :, row:row + tilesize, col:col + tilesize] += tmp[0]
return pred
sys.path.append("/d/achanhon/github/EfficientNet-PyTorch")
sys.path.append("/d/achanhon/github/pytorch-image-models")
sys.path.append("/d/achanhon/github/pretrained-models.pytorch")
sys.path.append("/d/achanhon/github/segmentation_models.pytorch")
import segmentation_models_pytorch as smp
import miniworld
print("load data")
dataset = miniworld.MiniWorld("/test/")
print("load model")
with torch.no_grad():
net = torch.load("build/model.pth")
net = net.cuda()
net.eval()
print("val", name)
cm = torch.zeros((len(dataset.cities), 2, 2)).cuda()
with torch.no_grad():
for k, city in enumerate(dataset.cities):
print(k, city)
for i in range(dataset.data[city].NB):
x, y = dataset.data[city].getImageAndLabel(i, torchformat=True)