Pretrained Text-Generation Models Below 250M Parameters - a Felladrin Collection

Felladrin 's Collections

Trained Models 🏋️

Frequently Used Spaces

Pretrained Text-Generation Models Below 250M Parameters

Pretrained Text-Generation Models Below 250M Parameters

updated Aug 10

Great candidates for fine-tuning targeting Transformers.js, ordered by number of parameters.

Locutusque/TinyMistral-248M

Text Generation • Updated May 9 • 6.91k • 40

Note License: apache-2.0
princeton-nlp/Sheared-Pythia-160m

Text Generation • Updated Dec 4, 2023 • 606 • 4

Note License: apache-2.0
JackFram/llama-160m

Text Generation • Updated Jan 4 • 264k • 31

Note License: apache-2.0
BEE-spoke-data/smol_llama-101M-GQA

Text Generation • Updated Dec 25, 2023 • 3.97k • 26

Note License: apache-2.0
JackFram/llama-68m

Text Generation • Updated May 23 • 520k • 23

Note License: apache-2.0
Felladrin/Minueza-32M-Base

Text Generation • Updated Mar 10 • 851 • 14

Note License: apache-2.0
EleutherAI/pythia-31m

Text Generation • Updated Jul 26, 2023 • 14k • 5

Note License: apache-2.0
HuggingFaceTB/SmolLM-135M

Text Generation • Updated Aug 1 • 40.6k • 170