Skip to content

text-classification

Directory actions

More options

Directory actions

More options

Latest commit

 

History

History
 
 

text-classification

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
data
data
 
 
1.basic-rnn.ipynb
1.basic-rnn.ipynb
 
 
10.lstm-rnn-bidirectional.ipynb
10.lstm-rnn-bidirectional.ipynb
 
 
11.lstm-rnn-bidirectional-huber.ipynb
11.lstm-rnn-bidirectional-huber.ipynb
 
 
12.lstm-rnn-dropout-l2.ipynb
12.lstm-rnn-dropout-l2.ipynb
 
 
13.gru-rnn.ipynb
13.gru-rnn.ipynb
 
 
14.gru-rnn-hinge.ipynb
14.gru-rnn-hinge.ipynb
 
 
15.gru-rnn-huber.ipynb
15.gru-rnn-huber.ipynb
 
 
16.gru-rnn-bidirectional.ipynb
16.gru-rnn-bidirectional.ipynb
 
 
17.gru-rnn-bidirectional-hinge.ipynb
17.gru-rnn-bidirectional-hinge.ipynb
 
 
18.gru-rnn-bidirectional-huber.ipynb
18.gru-rnn-bidirectional-huber.ipynb
 
 
19.lstm-cnn-rnn.ipynb
19.lstm-cnn-rnn.ipynb
 
 
2.basic-rnn-hinge.ipynb
2.basic-rnn-hinge.ipynb
 
 
20.kmax-cnn.ipynb
20.kmax-cnn.ipynb
 
 
21.lstm-cnn-rnn-highway.ipynb
21.lstm-cnn-rnn-highway.ipynb
 
 
22.lstm-rnn-attention.ipynb
22.lstm-rnn-attention.ipynb
 
 
23.dilated-rnn-lstm.ipynb
23.dilated-rnn-lstm.ipynb
 
 
24.lnlstm-rnn.ipynb
24.lnlstm-rnn.ipynb
 
 
25.only-attention.ipynb
25.only-attention.ipynb
 
 
26.multihead-attention.ipynb
26.multihead-attention.ipynb
 
 
27.neural-turing-machine.ipynb
27.neural-turing-machine.ipynb
 
 
28.lstm-seq2seq.ipynb
28.lstm-seq2seq.ipynb
 
 
29.lstm-seq2seq-luong.ipynb
29.lstm-seq2seq-luong.ipynb
 
 
3.basic-rnn-huber.ipynb
3.basic-rnn-huber.ipynb
 
 
30.lstm-seq2seq-bahdanau.ipynb
30.lstm-seq2seq-bahdanau.ipynb
 
 
31.lstm-seq2seq-beam.ipynb
31.lstm-seq2seq-beam.ipynb
 
 
32.lstm-seq2seq-birnn.ipynb
32.lstm-seq2seq-birnn.ipynb
 
 
33.pointer-net.ipynb
33.pointer-net.ipynb
 
 
34.lstm-rnn-bahdanau.ipynb
34.lstm-rnn-bahdanau.ipynb
 
 
35.lstm-rnn-luong.ipynb
35.lstm-rnn-luong.ipynb
 
 
36.lstm-rnn-bahdanau-luong.ipynb
36.lstm-rnn-bahdanau-luong.ipynb
 
 
37.lstm-birnn-bahdanau-luong.ipynb
37.lstm-birnn-bahdanau-luong.ipynb
 
 
38.bytenet.ipynb
38.bytenet.ipynb
 
 
39.fast-slow-lstm.ipynb
39.fast-slow-lstm.ipynb
 
 
4.basic-rnn-bidirectional.ipynb
4.basic-rnn-bidirectional.ipynb
 
 
40.siamese-network.ipynb
40.siamese-network.ipynb
 
 
41.estimator.ipynb
41.estimator.ipynb
 
 
42.capsule-rnn-lstm.ipynb
42.capsule-rnn-lstm.ipynb
 
 
43.capsule-seq2seq-lstm.ipynb
43.capsule-seq2seq-lstm.ipynb
 
 
44.capsule-birrn-seq2seq-lstm.ipynb
44.capsule-birrn-seq2seq-lstm.ipynb
 
 
45.nested-lstm.ipynb
45.nested-lstm.ipynb
 
 
46.lstm-seq2seq-highway.ipynb
46.lstm-seq2seq-highway.ipynb
 
 
47.triplet-loss-lstm.ipynb
47.triplet-loss-lstm.ipynb
 
 
48.dnc.ipynb
48.dnc.ipynb
 
 
49.convlstm.ipynb
49.convlstm.ipynb
 
 
5.basic-rnn-bidirectional-hinge.ipynb
5.basic-rnn-bidirectional-hinge.ipynb
 
 
50.temporalconvd.ipynb
50.temporalconvd.ipynb
 
 
51.batch-all-triplet-loss-lstm.ipynb
51.batch-all-triplet-loss-lstm.ipynb
 
 
52.fast-text.ipynb
52.fast-text.ipynb
 
 
53.gated-convolution-network.ipynb
53.gated-convolution-network.ipynb
 
 
54.simple-recurrent-units.ipynb
54.simple-recurrent-units.ipynb
 
 
55.lstm-han.ipynb
55.lstm-han.ipynb
 
 
56.bert.ipynb
56.bert.ipynb
 
 
57.dynamic-memory-network.ipynb
57.dynamic-memory-network.ipynb
 
 
58.entity-network.ipynb
58.entity-network.ipynb
 
 
59.memory-network.ipynb
59.memory-network.ipynb
 
 
6.basic-rnn-bidirectional-huber.ipynb
6.basic-rnn-bidirectional-huber.ipynb
 
 
60.char-sparse.ipynb
60.char-sparse.ipynb
 
 
61.residual-network.ipynb
61.residual-network.ipynb
 
 
62.residual-network-bahdanau.ipynb
62.residual-network-bahdanau.ipynb
 
 
63.deep-pyramid-cnn.ipynb
63.deep-pyramid-cnn.ipynb
 
 
63.deep-pyramid-cnn.py
63.deep-pyramid-cnn.py
 
 
64.transformer-xl.ipynb
64.transformer-xl.ipynb
 
 
65.gpt-2.ipynb
65.gpt-2.ipynb
 
 
66.quasi-rnn.ipynb
66.quasi-rnn.ipynb
 
 
7.lstm-rnn.ipynb
7.lstm-rnn.ipynb
 
 
8.lstm-rnn-hinge.ipynb
8.lstm-rnn-hinge.ipynb
 
 
9.lstm-rnn-huber.ipynb
9.lstm-rnn-huber.ipynb
 
 
README.md
README.md
 
 
bert_model.py
bert_model.py
 
 
dynamic_memory_network.py
dynamic_memory_network.py
 
 
entity_network.py
entity_network.py
 
 
gpt_2.py
gpt_2.py
 
 
utils.py
utils.py
 
 
xl.py
xl.py
 
 

README.md

How-to

  1. Make sure data folder in the same directory of the notebooks.

  2. Run any notebooks.

Score and average time taken per epoch, not sorted

Based on 20% validation. The results will be different on different dataset. Trained on a GTX 960, 4GB VRAM.

name accuracy time taken (s)
1. basic-rnn 0.68 1.3219
2. basic-rnn-hinge 0.65 1.2455
3. basic-rnn-huber 0.68 1.2468
4. basic-rnn-bidirectional 0.71 3.8174
5. basic-rnn-bidirectional-hinge 0.68 2.5127
6. basic-rnn-bidirectional-huber 0.63 3.5095
7. lstm-rnn 0.73 2.69683
8. lstm-rnn-hinge 0.72 8.2088
9. lstm-rnn-huber 0.73 10.1754
10. lstm-rnn-bidirectional 0.71 11.0388
11. lstm-rnn-bidirectional-huber 0.71 5.5258
12. lstm-rnn-dropout-l2 0.74 3.2420
13. gru-rnn 0.72 3.16123
14. gru-rnn-hinge 0.72 6.71951
15. gru-rnn-huber 0.70 7.93373
16. gru-rnn-bidirectional 0.73 2.91590
17. gru-rnn-bidirectional-hinge 0.72 5.66385
18. gru-rnn-bidirectional-huber 0.70 18.01133
19. lstm-cnn-rnn 0.65 4.42849
20. kmax-cnn 0.73 18.89667
21. lstm-cnn-rnn-highway 0.68 3.23122
22. lstm-rnn-attention 0.75 13.97496
23. dilated-rnn-lstm 0.25 24.54002
24. lnlstm-rnn 0.68 24.86363
25. only-attention 0.74 2.63291
26. multihead-attention 0.69 9.033228
27. neural-turing-machine
28. lstm-seq2seq 0.72 9.63291
29. lstm-seq2seq-luong
30. lstm-seq2seq-bahdanau
31. lstm-seq2seq-beam
32. lstm-seq2seq-birnn
33. pointer-net
34. lstm-rnn-bahdanau 0.71 9.81993
35. lstm-rnn-luong 0.66 27.73932
36. lstm-rnn-bahdanau-luong 0.69 36.97628
37. lstm-birnn-bahdanau-luong 0.70 38.86009
38. bytenet
39. fast-slow-lstm
40. siamese-network 0.52 7.13535
41. estimator
42. capsule-rnn-lstm
43. capsule-seq2seq-lstm
44. capsule-birrn-seq2seq-lstm
45. nested-lstm
46. lstm-seq2seq-highway
47. triplet-loss-lstm 0.50
48. dnc 0.68 85.98529
49. convlstm 0.69 2.66726
50. temporalconvd 0.66 11.90590
51. batch-all-triplet-loss-lstm 0.70
52. fast-text 0.76 0.49499
53. gated-convolution-network 0.67 3.37712
54. simple-recurrent-units 0.65 3.12624
55. lstm-han 0.50 3.47965
56. bert 0.73 6.31015
57. dynamic-memory-network 0.71 3.25820
58. entity-network 0.74 1.10458
59. memory-network 0.58 1.157306
60. char-sparse 0.76 2.350096
61. residual-network 0.72 9.557085
62. residual-network-bahdanau 0.71 11.53799
63. deep-pyramid-cnn 0.68 6.980528
64. transformer-xl 0.51 38.66338
65. gpt-2 0.62 105.2107
66. quasi-rnn 0.66 166.1456