13bbee615c87c812ddcfff73746cd504222e3fb0,onmt/Models.py,,make_base_model,#Any#Any#Any#Any#Any#,561

Before Change

    // HACK: collect source feature vocabs.
    feature_vocabs = []
    for j in range(100):
        key = "src_feat_" + str(j)
        if key not in fields:
            break
        feature_vocabs.append(fields[key].vocab)

    emb_opts = {"src_word_vec_size": model_opt.src_word_vec_size,
                "position_encoding": model_opt.position_encoding,

After Change

    // Make Encoder.
    src_vocab = fields["src"].vocab
    num_feat_embeddings = [len(feat_dict) for feat_dict in
                           ONMTDataset.collect_feature_dicts(fields)]
    embeddings = build_embeddings(
                model_opt, src_vocab.stoi[onmt.IO.PAD_WORD],

Instances