How to replace the word_embedded layer in the model with a trained word 2vec (512 dimensions per word)?

DeepRNN / image_captioning

Tensorflow implementation of "Show, Attend and Tell: Neural Image Caption Generation with Visual Attention"

MIT License

781 stars 351 forks source link

Open ThatNight opened 5 years ago

ThatNight commented 5 years ago

but

For help, I'd like to add a trained word 2vec to the model.