Unchun Yang's picture

Unchun Yang

ucyang

·

https://ucyang.com/

AI & ML interests

None yet

Recent Activity

upvoted an article about 9 hours ago

You could have designed state of the art positional encoding

liked a model about 10 hours ago

google/gemma-3-27b-pt-qat-q4_0-gguf

View all activity

Organizations

ucyang's activity

upvoted an article about 9 hours ago

Article

You could have designed state of the art positional encoding

Nov 25, 2024

• 201

upvoted a collection about 10 hours ago

Gemma 3 QAT

Quantization Aware Trained (QAT) Gemma 3 checkpoints. The model preserves similar quality as half precision while using 3x less memory • 8 items • Updated about 16 hours ago • 58

upvoted an article 2 days ago

Article

Training and Finetuning Reranker Models with Sentence Transformers v4

9 days ago

• 93

upvoted a paper 3 days ago

MoCha: Towards Movie-Grade Talking Character Synthesis

Paper • 2503.23307 • Published 5 days ago • 65

upvoted an article 3 days ago

Article

Bamba: Inference-Efficient Hybrid Mamba2 Model

Dec 18, 2024

• 49

upvoted a paper 4 days ago

Qwen2.5-Omni Technical Report

Paper • 2503.20215 • Published 9 days ago • 113

upvoted a collection 7 days ago

CoRNStack

State-of-the-art code retrieval and re-ranking models and datasets • 9 items • Updated 8 days ago • 15

upvoted 3 collections 8 days ago

Qwen2.5-Omni

End-to-End Omni (text, audio, image, video, and natural speech interaction) model based Qwen2.5 • 3 items • Updated 8 days ago • 76

Ling

6 items • Updated 25 days ago • 8

TxGemma Release

Collection of open models to accelerate the development of therapeutics. • 5 items • Updated about 16 hours ago • 39

upvoted a paper 12 days ago

Why Do Multi-Agent LLM Systems Fail?

Paper • 2503.13657 • Published 17 days ago • 41

upvoted a paper 13 days ago

SynCity: Training-Free Generation of 3D Worlds

Paper • 2503.16420 • Published 14 days ago • 25

upvoted a collection 13 days ago

💫StarVector Models

StarVector is a multimodal LLM for Scalable Vector Graphics (SVG) generation, producing structured SVG code directly from images and text. • 2 items • Updated 14 days ago • 90

upvoted a paper 13 days ago

Survey on Evaluation of LLM-based Agents

Paper • 2503.16416 • Published 14 days ago • 81

upvoted a collection 13 days ago

MoshiVis v0.1

MoshiVis is a Vision Speech Model built as a perceptually-augmented version of Moshi v0.1 for conversing about image inputs • 8 items • Updated 14 days ago • 20

upvoted 2 collections 14 days ago

DeepHermes

Preview models of hybrid reasoner Hermes series • 6 items • Updated 21 days ago • 27

Llama Nemotron

Open, Production-ready Enterprise Models • 3 items • Updated about 3 hours ago • 25

upvoted a collection 15 days ago

EXAONE-Deep

EXAONE reasoning model series of 2.4B, 7.8B, and 32B, optimized for reasoning tasks including math and coding • 9 items • Updated 17 days ago • 85