tiny_shakespeare

  • Mô tả :

40.000 dòng Shakespeare từ nhiều vở kịch của Shakespeare. Nổi bật trong bài đăng trên blog của Andrej Karpathy 'Hiệu quả phi lý của mạng thần kinh tái phát': http://karpathy.github.io/2015/05/21/rnn-efficiency/

Để sử dụng ví dụ như mô hình nhân vật:

d = tfds.load(name='tiny_shakespeare')['train']
d = d.map(lambda x: tf.strings.unicode_split(x['text'], 'UTF-8'))
# train split includes vocabulary for other splits
vocabulary = sorted(set(next(iter(d)).numpy()))
d = d.map(lambda x: {'cur_char': x[:-1], 'next_char': x[1:]})
d = d.unbatch()
seq_len = 100
batch_size = 2
d = d.batch(seq_len)
d = d.batch(batch_size)
Tách ra ví dụ
'test' 1
'train' 1
'validation' 1
  • Cấu trúc tính năng :
FeaturesDict({
    'text': Text(shape=(), dtype=string),
})
  • Tài liệu tính năng :
Tính năng Lớp học Hình dạng Dtype Sự miêu tả
Tính năngDict
chữ Chữ sợi dây
  • trích dẫn :
@misc{
  author={Karpathy, Andrej},
  title={char-rnn},
  year={2015},
  howpublished={\url{https://github.com/karpathy/char-rnn} }
}
,

  • Mô tả :

40.000 dòng Shakespeare từ nhiều vở kịch của Shakespeare. Nổi bật trong bài đăng trên blog của Andrej Karpathy 'Hiệu quả phi lý của mạng thần kinh tái phát': http://karpathy.github.io/2015/05/21/rnn-efficiency/

Để sử dụng ví dụ như mô hình nhân vật:

d = tfds.load(name='tiny_shakespeare')['train']
d = d.map(lambda x: tf.strings.unicode_split(x['text'], 'UTF-8'))
# train split includes vocabulary for other splits
vocabulary = sorted(set(next(iter(d)).numpy()))
d = d.map(lambda x: {'cur_char': x[:-1], 'next_char': x[1:]})
d = d.unbatch()
seq_len = 100
batch_size = 2
d = d.batch(seq_len)
d = d.batch(batch_size)
Tách ra ví dụ
'test' 1
'train' 1
'validation' 1
  • Cấu trúc tính năng :
FeaturesDict({
    'text': Text(shape=(), dtype=string),
})
  • Tài liệu tính năng :
Tính năng Lớp học Hình dạng Dtype Sự miêu tả
Tính năngDict
chữ Chữ sợi dây
  • trích dẫn :
@misc{
  author={Karpathy, Andrej},
  title={char-rnn},
  year={2015},
  howpublished={\url{https://github.com/karpathy/char-rnn} }
}