batch size is creating confusion [ when we compare with Research Paper ]

[ahmadshabbir2468]

Research Paper Says :

Each batch contains a list of sentences which is determined by the parameter of batch size. In our experiments, we use batch size of 100 which means to include sentences whose total length is no greater than 100 (what i understand Total length is no greater than 100 means : Sentence length including white space should less then 100 i.e this sentence "LSTM CRF" length is 8 as it include 8 character in it ) am i Right?

[](Research Paper links https://arxiv.org/pdf/1508.01991.pdf)

what should we do for long sentence ??

lstm_crf main.py

   params = {
        'dim': 300,
        'dropout': 0.5,
        'num_oov_buckets': 1,
        'epochs': 25,
        'batch_size': 20,
        'buffer': 15000,
        'lstm_size': 100,
        'words': str(Path(DATADIR, 'vocab.words.txt')),
        'chars': str(Path(DATADIR, 'vocab.chars.txt')),
        'tags': str(Path(DATADIR, 'vocab.tags.txt')),
        'glove': str(Path(DATADIR, 'glove.npz'))
    }