Tensors and Dynamic neural networks in Python with strong GPU acceleration.
DyNet
DyNet provides tools for building dynamic neural networks and deep learning models.
About DyNet
DyNet is a neural network library written in C++ with bindings in Python. It is built for networks with dynamic structures that change for every training instance, a pattern used in natural language processing tasks such as syntactic parsing, machine translation, and morphological inflection. It runs on CPU or GPU and includes an auto-batching feature.
Installation requires external tools including CMake and a specific version of Eigen, with separate build steps for C++ and installation from GitHub for Python. Examples and tutorials are provided for both C++ and Python.
Key features
- Handles dynamic structures that change per training instance
- Written in C++ with bindings for Python
- Runs on CPU or GPU
- Includes auto-batching feature
- Provides C++ and Python examples and tutorials
Best for: Researchers and developers building dynamic neural networks for natural language processing tasks.
Good for
- Supports networks that change structure per training instance
- Works on both CPU and GPU
- Provides C++ and Python interfaces with examples
Watch out for
- Requires CMake and a specific version of Eigen to compile
- C++ use requires cloning the repository and building with CMake
DyNet at a glance
- Website
- github.com/clab/dynet
- Type
- Open-source project
- Pricing
- Not confirmed. Check the site
- Open source
- Yes · Source code
- Language
- English
- Last checked
- Oct 6, 2026 · Online when checked
- Listed
- Oct 5, 2026
DyNet alternatives
Tools and sites you could use instead, ranked by how closely they match.
Gymnasium provides a standard API for single agent reinforcement learning environments and related utilities.
Open source
24 alternatives
Tensorforce provides a TensorFlow library for applied reinforcement learning tasks.
Open source
24 alternatives
Extensible Toolkit for Finetuning and Inference of Large Foundation Models.
Open source
24 alternatives
GPT-NeoX implements model parallel autoregressive transformers on GPUs using Megatron and DeepSpeed libraries.
Open source
24 alternatives
oneDNN provides a library of optimized deep neural network operations for oneAPI.
Open source
24 alternativesPopular in AI Infrastructure & MLOps
Well-known listings from the same topic.
Guild AIguild.aiFlow AIflow-ai.comsystemprompt.iosystemprompt.ioKili Technologykili-technology.comAPI docsdocs.x.aiEyePop.aieyepop.ai
DyNet FAQ
What types of networks is DyNet for?
Networks with dynamic structures that change for every training instance, often used in natural language processing.
How do I install DyNet for Python?
Use pip install git+https://github.com/clab/dynet#egg=dynet, with further details in the documentation.
What is needed to install DyNet for C++?
Clone the repository, point CMake to the Eigen include directory, then compile with make.
Is something wrong or out of date?
Tell us and an editor will check it.
Own this website?
Claim this listing
Claim the listing and suggest updates yourself. An editor checks every claim first.