INNER CODE UNIT · Python

pad_length

graviraja/100-Days-of-NLP · applications/classification/ner_tagging/app/app.py:91

            pad_length = MAX_WORD_LEN - word_len
            word = word + [chars_vocab.PAD] * pad_length
        
        # convert the chars into numerical format
        char_ids = []
        for each_char in word: 
            char_ids.append(chars_vocab[each_char])
        char_seq.append(char_ids)

    # numericalize
    token_ids = [words_vocab[tok] for tok in tokens]
    
    # seq length
    sent_length = [len(token_ids)]

    # create tensors
    sent_tensor = torch.LongTensor(token_ids)
    sent_tensor = sent_tensor.unsqueeze(0)

View source record →