cap = u'[CLS] '+cap ???

[y78h11b09]

       embeddings = []
        for cap_idx in  encoded_captions:

            # padd caption to correct size
            while len(cap_idx) < max_dec_len:
                cap_idx.append(PAD)

            cap = ' '.join([vocab.idx2word[word_idx.item()] for word_idx in cap_idx])
            cap = u'[CLS] '+cap

Hello, I want to know why here is the code "cap = u'[CLS] '+cap ", is not "cap = u'[CLS]'+cap+u'[SEP]'

[mathshangw]

i know that your question is old, but excuse me do you know why? and what is encoded_captions, does that the captions or the sentences?

[y78h11b09]

Yes, it denotes the sentence about the content of image.

	y78h11b09

@. | On 1/27/2022 @.> wrote：

i know that your question is old, but excuse me do you know why? and what is encoded_captions, does that the captions or the sentences?

— Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android. You are receiving this because you authored the thread.Message ID: @.***>

[mathshangw]

thanks so cap_idx is string contain sentence or id ?

[y78h11b09]

I'am sorry to reply you lately, cap_idx is long number that denoted the index of sentence. Kind regards.

	y78h11b09

@. | On 1/31/2022 @.> wrote：

thanks so cap_idx is string contain sentence or id ?

— Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android. You are receiving this because you authored the thread.Message ID: @.***>

[mathshangw]

Thanks a lot.

  for cap_idx in  encoded_captions:
          # padd caption to correct size
          while len(cap_idx) < max_dec_len:
                  cap_idx.append(PAD)

len(cap_idx)does means the sentence after encoding ? or the index ? because there is padding for the length, not the name. sorry for my confusion