BinghengWu's picture

BinghengWu

wubingheng

·

https://github.com/wubingheng111

AI & ML interests

I like to fine-tune the small models of the Doge series.

Organizations

Articles 1

Article

7

Trainable Dynamic Mask Sparse Attention: Bridging Efficiency and Effectiveness in Long-Context Language Models

Papers 3

arxiv:2508.02124

arxiv:2505.19716

arxiv:2412.11834

models 3

wubingheng/Doge-20M-Medical-SFT

Text Generation • 13.1M • Updated Apr 16, 2025 • 5

wubingheng/Doge-20M-Chinese

Text Generation • 13.1M • Updated Apr 15, 2025 • 7 • 2

wubingheng/Doge-197M-Medical-SFT

Question Answering • 0.2B • Updated Jan 31, 2025 • 12 • 2

datasets 15

wubingheng/MixtureOfThoughts-Chinese-tryrun

Viewer • Updated Jul 18, 2025 • 10 • 12

wubingheng/Mixture-of-Thoughts-zh-try-run

Viewer • Updated Jul 17, 2025 • 10 • 15

wubingheng/Budget-aware-2048

Viewer • Updated Apr 29, 2025 • 25k • 30

wubingheng/Budget-aware-2048-in

Viewer • Updated Apr 29, 2025 • 25k • 15

wubingheng/Budget-aware-2048-in-try-run

Viewer • Updated Apr 29, 2025 • 2 • 18

wubingheng/Budget-aware-2048-try-run

Viewer • Updated Apr 29, 2025 • 2 • 15

wubingheng/L1-2048

Viewer • Updated Apr 28, 2025 • 25k • 9

wubingheng/L1-1024

Viewer • Updated Apr 28, 2025 • 25k • 15

wubingheng/compressed-openthoughts-50

Viewer • Updated Apr 28, 2025 • 25k • 23

wubingheng/compressed-openthoughts-90

Viewer • Updated Apr 28, 2025 • 25k • 9

View 15 datasets