titans-pytorch

Unofficial implementation of Titans, SOTA memory for transformers, in Pytorch

Project.

Catalog snapshot

License: MIT
Platforms: Linux
Docker: No
Commercial use: Allowed
Stars: 2k
Language: Python
Last commit: Jul 13, 2026
Tags: artificial-intelligencedeep-learninglong-term-memorytest-time-training

Repository health snapshot: 206 forks and 38 open issues at review time. Platform support is inferred conservatively from repository topics.

About.

Unofficial implementation of Titans, SOTA memory for transformers, in Pytorch Fossfits includes titans-pytorch as an actively maintained fine tuning option: its public repository was updated on 2026-07-13, uses the MIT license, and had 1,982 stars when reviewed. Check its documentation and release notes before production adoption.

artificial intelligence · deep learning · long term memory · test time training

Nearby.

All Fine-tuning

LLLaMA-FactoryWeb UI and YAML recipes for fine-tuning many open LLMs.Compare
TtorchtunePyTorch-native library for fine-tuning LLMs with recipes.Compare
DDeepSpeedMicrosoft’s Apache-2.0 ZeRO optimizer for multi-GPU, multi-node LLM training.Compare
Ppytorch-lightningPretrain, finetune ANY AI model of ANY size on 1 or 10,000+ GPUs with zero code changes.Compare
UUnslothFaster LoRA fine-tuning with lower VRAM on consumer GPUs.Compare