annotated-transformer
annotated-transformer copied to clipboard
About Shared Embedding
if False:
model.src_embed[0].lut.weight = model.tgt_embeddings[0].lut.weight
model.generator.lut.weight = model.tgt_embed[0].lut.weight
Hi, I can't find tgt_embeddings
in your code. Maybe it is model.src_embed[0].lut.weight = model.tgt_embed[0].lut.weight
.
And if shared Embedding, Should the code be add to make_model()? Like
def make_model(src_vocab, tgt_vocab, N=6, d_model=512, d_ff=2048, h=8, dropout=0.1):
...
model = EncoderDecoder(...)
model.src_embed[0].lut.weight = model.tgt_embed[0].lut.weight
model.generator.lut.weight = model.tgt_embed[0].lut.weight
...
return model
i have the same question
It is a typo and it is a False
condition. Or else you can try BPE.