Ducati

ducatiqwq

·

ducatiqwq

AI & ML interests

None yet

Recent Activity

authored a paper 23 days ago

WavePrune: One period is often enough for RoPE

upvoted a paper 24 days ago

WavePrune: One period is often enough for RoPE

upvoted a paper about 2 months ago

SMELT: Scaling Laws for Compute-Matched MoE Looped Transformers

View all activity

Organizations

None yet

liked 3 datasets over 1 year ago

nvidia/Llama-Nemotron-Post-Training-Dataset

Viewer • Updated May 8, 2025 • 3.91M • 6.87k • 721

HuggingFaceFW/fineweb

Viewer • Updated Jul 11, 2025 • 52.5B • 369k • 3.71k

cais/hle

Benchmark • Updated Jan 20 • 2.5k • 28.4k • 1.07k

liked a model almost 2 years ago

deepseek-ai/Janus-Pro-7B

Any-to-Any • Updated Feb 1, 2025 • 10.7k • 3.72k

liked a dataset almost 2 years ago

NovelQA/NovelQA

Updated Jun 4 • 144 • 17