Unchun Yang's picture

Unchun Yang

ucyang

·

https://ucyang.com/

AI & ML interests

None yet

Recent Activity

liked a model about 2 hours ago

Qwen/Qwen3-30B-A3B-Instruct-2507-FP8

liked a model about 2 hours ago

Qwen/Qwen3-30B-A3B-Instruct-2507

liked a Space about 10 hours ago

Inferless/LLM-Inference-Benchmark

View all activity

Organizations

upvoted a paper 1 day ago

Group Sequence Policy Optimization

Paper • 2507.18071 • Published 6 days ago • 228

upvoted a paper 4 days ago

Llama-Nemotron: Efficient Reasoning Models

Paper • 2505.00949 • Published May 2 • 40

upvoted a collection 4 days ago

OpenReasoning-Nemotron

Collection of models for OpenReasoning-Nemotron which are trained on 5M reasoning traces for Math, Code and Science. • 6 items • Updated 8 days ago • 39

upvoted a collection 6 days ago

Kanana-1.5

Open Source Kanana-1.5 • 7 items • Updated 6 days ago • 27

upvoted a paper 6 days ago

EfficientQAT: Efficient Quantization-Aware Training for Large Language Models

Paper • 2407.11062 • Published Jul 10, 2024 • 10

upvoted an article 10 days ago

Article

A Gentle Introduction to 8-bit Matrix Multiplication for transformers at scale using transformers, accelerate and bitsandbytes

By

and 1 other •

Aug 17, 2022

• 102

upvoted a collection 10 days ago

A.X 4

5 items • Updated 1 day ago • 38

upvoted an article 11 days ago

Article

OpenReasoning-Nemotron: A Family of State-of-the-Art Distilled Reasoning Models

By

and 3 others •

11 days ago

• 47

upvoted a paper 12 days ago

Seq vs Seq: An Open Suite of Paired Encoders and Decoders

Paper • 2507.11412 • Published 14 days ago • 25

upvoted a collection 12 days ago

Encoders vs Decoders: the Ettin Suite

A collection of SOTA, open-data, paired encoder-only and decoder only models ranging from 17M params to 1B. See the paper at https://arxiv.org/abs/250 • 32 items • Updated 13 days ago • 14

upvoted an article 12 days ago

Article

Seq vs Seq: the Ettin Suite of Paired Encoders and Decoders

By

and 5 others •

14 days ago

• 50

upvoted 2 collections 14 days ago

Mi:dm 2.0

KT's Korea-centric AI models - Mi:dm 2.0 • 2 items • Updated 26 days ago • 14

H-Net

The family of hierarchical networks (H-Nets) from https://arxiv.org/abs/2507.07955 • 8 items • Updated 18 days ago • 18

upvoted a collection 15 days ago

EXAONE-4.0

EXAONE unified model series of 1.2B and 32B, integrating non-reasoning and reasoning modes. • 20 items • Updated about 12 hours ago • 40

upvoted 2 articles 15 days ago

Article

Creating custom kernels for the AMD MI300

By

and 1 other •

21 days ago

• 43

Article

Reachy Mini - The Open-Source Robot for Today's and Tomorrow's AI Builders

By

and 1 other •

21 days ago

• 615

upvoted a paper 16 days ago

MiniMax-M1: Scaling Test-Time Compute Efficiently with Lightning Attention

Paper • 2506.13585 • Published Jun 16 • 260

upvoted a collection 16 days ago

MiniMax-M1

MiniMax-M1, the world's first open-weight, large-scale hybrid-attention reasoning model. • 6 items • Updated 27 days ago • 110

upvoted a collection 18 days ago

Kimi-K2

Moonshot's MoE LLMs with 1 trillion parameters, exceptional on agentic intellegence • 2 items • Updated 17 days ago • 111

upvoted a collection 19 days ago

💧 LFM2

LFM2 is a new generation of hybrid models, designed for on-device deployment. • 15 items • Updated 1 day ago • 78