AlphaUncommon
guolinke/TUPE
Transformer with Untied Positional Encoding (TUPE). Code of paper "Rethinking Positional Encoding in Language Pre-training". Improve existing models like BERT.
- Owner
- guolinke (personal account)
- Stars
- Stars: 250
- Forks
- Forks: 27
- Language
- Python
- License
- MIT
- Created
- 2020
- Topics
- bertpretraininglanguage-modeltransformer
- Tier
- Alpha (250 to 699 ★) What this tier means
More repos by guolinke
All 2 repos by guolinke- guolinke/SphereARStars: 114PythonPrototype
Nearby in Python
All Python repos- Ahren09/ScienceUtopiaStars: 250PythonAlpha
- ryan-bergio03h1/Instagram-View-Booster-Free-Reel-Story-Views-Tool-2026Stars: 250PythonAlpha
- aa1143/NamiAgentStars: 250PythonAlpha
- alex/django-ajax-validationStars: 250PythonAlpha
- bbangert/velruseStars: 250PythonAlpha
- Cue/fast-python-pbStars: 250PythonAlpha