Quazim0t0 / Escarda-86M-Base-JL
Text Generation97M330Updated Jul 11, 2026
Downloads
33
Likes
0
Last updated
Jul 11, 2026
License
apache-2.0
This is the Jet-Long edition of Escarda-86M-Base. It extends the usable context from the native 4,096-token training window to 10,240 tokens with no fine-tuning and no change to short-context behaviour, by adding dynamic bifocal RoPE from Jet-Long (arXiv:2607.07740, NVIDIA).
Tasks & library
- text generation
- transformers
Tags
- transformers
- spike_whale
- feature-extraction
- text-generation
- small-models
- base-model
Technical specifications
Parameters97M
Librarytransformers
Languagesen
Licenseapache-2.0