83c2a211d87c89ab460dc08c105aba1e4673529d,preprocess.py,,main,#,52

Before Change

def main():
    print("Preparing training ...")
    with codecs.open(opt.train_src, "r", "utf-8") as src_file:
        src_line = src_file.readline().strip().split()
        _, _, nFeatures = onmt.IO.extract_features(src_line)

    fields = onmt.IO.ONMTDataset.get_fields(nFeatures)
    print("Building Training...")
    train = onmt.IO.ONMTDataset(opt.train_src, opt.train_tgt, fields, opt)
    print("Building Vocab...")

After Change

def main():
    if opt.data_type == "text":
        print("Preparing training ...")
        with codecs.open(opt.train_src, "r", "utf-8") as src_file:
            src_line = src_file.readline().strip().split()
            _, _, nFeatures = onmt.IO.extract_features(src_line)

        fields = onmt.IO.ONMTDataset.get_fields(nFeatures)
    elif opt.data_type == "img":
        fields = onmt.IO.ONMTDataset.get_fields(data_type="img")

    print("Building Training...")
    train = onmt.IO.ONMTDataset(opt.train_src, opt.train_tgt, fields, opt)
    print("Building Vocab...")
    onmt.IO.ONMTDataset.build_vocab(train, opt)

Instances