Skip to main content

Conditional random field in PyTorch

Project description

Conditional random field in PyTorch.


This package provides an implementation of conditional random field (CRF) in PyTorch. This implementation borrows mostly from AllenNLP CRF module with some modifications.


  • Python 3.6
  • PyTorch 0.4.1


You can install with pip

pip install pytorch-crf

Or, you can install from Github directly

pip install git+


In the examples below, we will assume that these lines have been executed

>>> import torch
>>> from torchcrf import CRF
>>> seq_length, batch_size, num_tags = 3, 2, 5
>>> emissions = torch.randn(seq_length, batch_size, num_tags)
>>> tags = torch.tensor([[0, 1], [2, 4], [3, 1]], dtype=torch.long)  # (seq_length, batch_size)
>>> model = CRF(num_tags)

Computing log likelihood

>>> model(emissions, tags)
tensor(-12.7431, grad_fn=<SumBackward0>)

Computing log likelihood with mask

>>> mask = torch.tensor([[1, 1], [1, 1], [1, 0]], dtype=torch.uint8)  # (seq_length, batch_size)
>>> model(emissions, tags, mask=mask)
tensor(-10.8390, grad_fn=<SumBackward0>)


>>> model.decode(emissions)
[[3, 1, 3], [0, 1, 0]]

Decoding with mask

>>> model.decode(emissions, mask=mask)
[[3, 1, 3], [0, 1]]

See tests/ for more examples.


MIT. See LICENSE for details.


Contributions are welcome! Please follow these instructions to setup dependencies and running the tests and linter. Make a pull request once your contribution is ready.

Installing dependencies

Make sure you setup a virtual environment with Python 3.6 and PyTorch installed. Then, install all the dependencies in requirements.txt file and install this package in development mode.

pip install -r requirements.txt
pip install -e .

Running tests

Run pytest in the project root directory.

Running linter

Run flake8 in the project root directory. This will also run mypy, thanks to flake8-mypy package.

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Filename, size & hash SHA256 hash help File type Python version Upload date
pytorch_crf-0.6.0-py3-none-any.whl (9.9 kB) Copy SHA256 hash SHA256 Wheel py3

Supported by

Elastic Elastic Search Pingdom Pingdom Monitoring Google Google BigQuery Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN SignalFx SignalFx Supporter DigiCert DigiCert EV certificate StatusPage StatusPage Status page