https://github.com/tranbavinhson/sequence-labeling
Sequence Labeling implemented by BiLSTM using Tensorflow
https://github.com/tranbavinhson/sequence-labeling
bilstm deep-learning lstm lstm-model sequence-labeling tensorflow tensorflow-tutorials
Last synced: about 1 month ago
JSON representation
Sequence Labeling implemented by BiLSTM using Tensorflow
- Host: GitHub
- URL: https://github.com/tranbavinhson/sequence-labeling
- Owner: TranBaVinhSon
- Created: 2018-04-13T09:41:46.000Z (over 7 years ago)
- Default Branch: master
- Last Pushed: 2018-06-06T17:10:01.000Z (over 7 years ago)
- Last Synced: 2025-08-04T20:19:55.102Z (2 months ago)
- Topics: bilstm, deep-learning, lstm, lstm-model, sequence-labeling, tensorflow, tensorflow-tutorials
- Language: Python
- Homepage:
- Size: 7.81 KB
- Stars: 6
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Sequence-labeling
Sequence Labeling implemented by Bi-LSTM using Tensorflow.
## Benchmark for building a (Bidirectional) LSTM model.
The hyperparameters used in the model:
- **learning_rate** - the initial value of the learning rate
- **max_lr_epoch** - after **max_lr_epoch**, the learning rate will be decreased
- **num_layers** - the number of (Bi)LSTM layers
- **num_steps** - the number of unrolled steps of (Bi)LSTM
- **hidden_size** - the number of (Bi)LSTM units
- **num_epochs** - the total number of epochs for training
- **keep_prob** - the probability of keeping weights in the dropout layer
- **lr_decay** - the decay of the learning rate
- **batch_size** - number of inputs
## Evaluation
- Accuracy
- F1 Score
## Todo
- [ ] BiLSTM + CRF
- [ ] BiLSTM + CRF + CNN