Vadimbuildercxx/looped_transformer

Experimental implementation of "Looped Transformers are Better at Learning Learning Algorithms" showing superior performance with 12x fewer parameters. Includes complete environment setup, pre-trained weights, and extensive experiments comparing Looped TFs vs traditional Transformers for In-Context Learning.

20
/ 100
Experimental

This project helps machine learning researchers understand and apply 'Looped Transformers,' a new neural network architecture. It takes data for in-context learning tasks, specifically linear regression examples, and outputs insights into the performance and efficiency of Looped Transformers compared to traditional Transformers. The ideal user is a machine learning researcher or practitioner exploring advanced transformer architectures for improved learning algorithms.

No commits in the last 6 months.

Use this if you are a machine learning researcher interested in the theoretical and practical benefits of Looped Transformers for in-context learning tasks.

Not ideal if you are looking for a ready-to-use production model or a general-purpose transformer for tasks outside of algorithm learning research.

machine-learning-research neural-network-architecture in-context-learning deep-learning-efficiency algorithm-learning
No License Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 4 / 25
Maturity 8 / 25
Community 8 / 25

How are scores calculated?

Stars

8

Forks

1

Language

Jupyter Notebook

License

Last pushed

Aug 05, 2024

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/transformers/Vadimbuildercxx/looped_transformer"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.