ISubGVQA.models.question_encoder

Classes

QuestionEncoder

Module Contents

class ISubGVQA.models.question_encoder.QuestionEncoder(text_vocab_embedding, text_emb_dim, ninp, nhead, nhid, nlayers, dropout=0.5)

Bases: torch.nn.Module

text_vocab_embedding

model_type = 'Transformer'

emb_proj

pos_encoder

transformer_encoder

ninp

forward(src, mask)