Sequence Modeling Python Packages

fla-core

🚀 Efficient implementations for emerging model architectures

flash-linear-attention

🚀 Efficient implementations for emerging model architectures

gateloop-transformer

Implementation of GateLoop Transformer in Pytorch and Jax

rwkv-fla

🚀 Efficient implementations for emerging model architectures

s5-pytorch

Pytorch implementation of Simplified Structured State-Spaces for Sequence Modeling (S5)

gp3sequencespy

Transparent, reproducible sequence analysis in Python. Audit ordered categorical data, discover recurring structure, compare trajectories, model transitions and latent states, test group differences, and report every analytical choice.

arabicnlp

Python package for Arabic natural language processing

decepticons

Backend-neutral kernel of predictive primitives. O(n) attention is deception.

typeseg

Content-Type Segmentation for Mixed Text

pararnn-torch

Train nonlinear RNNs in parallel (Newton+scan); decode one step at a time.

megalodon-jax

jax rewrite for CEMA support + other training speedups.

pyxlstm

Efficient Python library for Extended LSTM with exponential gating, memory mixing, and matrix memory for superior sequence modeling.

discretax

Discretax is a light weight collection of state space models implemented in JAX ⚡️

mingru-scans

minGRU (Feng et al.) in PyTorch, plus a measured ladder of state-tracking recoveries — signed, rotation, Givens, and delta (DeltaNet-style Householder products) mixers with time decay, all under one parallel scan — with fused Triton GPU kernels as an optional backend

permea-core

Open toolkit and benchmarks for sequence-first delivery and mRNA expression engineering.

transact-tf

An unofficial implementation of "TransAct: Transformer-based Realtime User Action Model for Recommendation at Pinterest" in Tensorflow

linax

Discretax is a light weight collection of state space models implemented in JAX ⚡️

lrnnx

A unified PyTorch library for Linear RNNs

spectrans

Modular spectral transformer implementations in PyTorch with Fourier, wavelet, and other frequency-domain operations for efficient sequence modeling.