93fde518ce45a1373c4b635c15e25312b2cc3a0f,examples/semantic_segmentation/fcn/train_fcn.py,,main,#Any#Any#Any#Any#Any#Any#,17

Before Change


def main(gpu=-1, batch_size=1, iterations=100000,
         lr=1e-10, out="result", resume=""):
    // prepare datasets
    wrappers = [lambda d: SubtractWrapper(d),
                lambda d: PadWrapper(
                    d, max_size=(512, 512), preprocess_idx=[0, 1],
                    bg_values={0: 0, 1: -1})]
    train_data = VOCSemanticSegmentationDataset(mode="train")
    test_data = VOCSemanticSegmentationDataset(mode="val")
    for wrapper in wrappers:
        train_data = wrapper(train_data)
        test_data = wrapper(test_data)

    // set up FCN32s
    n_class = 21
    model = FCN32s(n_class=n_class)
    if gpu != -1:
        model.to_gpu(gpu)

After Change



    train_data = VOCSemanticSegmentationDataset(mode="train")
    test_data = VOCSemanticSegmentationDataset(mode="val")
    extend(train_data, transform)
    extend(test_data, transform)

    // set up FCN32s
    n_class = 21
Italian Trulli
In pattern: SUPERPATTERN

Frequency: 4

Non-data size: 4

Instances


Project Name: chainer/chainercv
Commit Name: 93fde518ce45a1373c4b635c15e25312b2cc3a0f
Time: 2017-03-10
Author: yuyuniitani@gmail.com
File Name: examples/semantic_segmentation/fcn/train_fcn.py
Class Name:
Method Name: main


Project Name: chainer/chainercv
Commit Name: a7706fbde22887909db42f96a696437c084c05db
Time: 2017-05-31
Author: Hakuyume@users.noreply.github.com
File Name: chainercv/evaluations/eval_detection_voc.py
Class Name:
Method Name: eval_detection_voc


Project Name: OpenNMT/OpenNMT-py
Commit Name: f1aadee66689491a6993070955abe7987b9818fc
Time: 2019-01-23
Author: guillaumekln@users.noreply.github.com
File Name: onmt/utils/loss.py
Class Name:
Method Name: filter_shard_state


Project Name: chainer/chainercv
Commit Name: 7e470179b0d874c43ec6d5dba11d7981ad83dff1
Time: 2017-03-10
Author: yuyuniitani@gmail.com
File Name: examples/detection/faster_rcnn/train_faster_rcnn.py
Class Name:
Method Name: main