ucyang (Unchun Yang)

upvoted 3 articles 4 days ago

Article

What Makes a Dialog Agent Useful?

Jan 24, 2023

• 1

Article

Can We Train Chat Models with Raw Data?

By

•

6 days ago

• 16

Article

Making LLMs even more accessible with bitsandbytes, 4-bit quantization and QLoRA

May 24, 2023

• 19

upvoted a collection 5 days ago

Korean Datasets I've released so far.

Collection

지금까지 업로드한 한국어 데이터셋 콜렉션입니다. • 6 items • Updated Dec 29, 2023 • 14

upvoted a paper 5 days ago

PoSE: Efficient Context Window Extension of LLMs via Positional Skip-wise Training

Paper • 2309.10400 • Published Sep 19, 2023 • 18

upvoted 3 collections 7 days ago

upvoted a collection 8 days ago

FewMany

Collection

Benchmark For Few Shot Classification with Many Classes • 8 items • Updated 13 days ago • 6

upvoted a paper 8 days ago

Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone

Paper • 2404.14219 • Published 9 days ago • 222

upvoted 3 collections 8 days ago

LayoutLM

Collection

The LayoutLM series are Transformer encoders useful for document AI tasks such as invoice parsing, document image classification and DocVQA. • 5 items • Updated Nov 27, 2023 • 7

Table Transformer

Collection

The Table Transformer (TATR) is a series of object detection models useful for table extraction from PDF images. • 5 items • Updated Dec 1, 2023 • 10

Phi-3

Collection

Phi-3 family of models • 6 items • Updated 6 days ago • 168

upvoted an article 8 days ago

Article

Faster fine-tuning using TRL & Unsloth

Jan 10

• 12

upvoted an article 9 days ago

Article

Jack of All Trades, Master of Some, a Multi-Purpose Transformer Agent

10 days ago

• 59

upvoted 2 articles 10 days ago

Article

Fine-tune Llama 3 with ORPO

By

•

9 days ago

• 147

Article

Fine-tuning 20B LLMs with RLHF on a 24GB consumer GPU

Mar 9, 2023

• 6

upvoted a collection 13 days ago

Meta Llama 3

Collection

This collection hosts the transformers and original repos of the Meta Llama 3 and Llama Guard 2 releases • 5 items • Updated 13 days ago • 444

upvoted an article 13 days ago

Article

Welcome Llama 3 - Meta's new open LLM

14 days ago

• 226

upvoted an article 14 days ago

Article

Synthetic data: save money, time and carbon with open source

Feb 16

• 7

upvoted a paper 18 days ago

Llama 2: Open Foundation and Fine-Tuned Chat Models

Paper • 2307.09288 • Published Jul 18, 2023 • 233

upvoted an article 19 days ago

Article

Making LLMs lighter with AutoGPTQ and transformers

Aug 23, 2023

• 9

upvoted a paper 20 days ago

ORPO: Monolithic Preference Optimization without Reference Model

Paper • 2403.07691 • Published Mar 12 • 52

upvoted a paper 26 days ago

Long-context LLMs Struggle with Long In-context Learning

Paper • 2404.02060 • Published 29 days ago • 31

upvoted a paper 28 days ago

ViTAR: Vision Transformer with Any Resolution

Paper • 2403.18361 • Published Mar 27 • 48

upvoted a collection 28 days ago

HF-curated models available on Workers AI

Collection

A collection of models curated with Hugging Face that can be run on Cloudflare's Workers AI serverless inference platform. • 15 items • Updated 29 days ago • 44

upvoted a paper 30 days ago

Bootstrapping LLM-based Task-Oriented Dialogue Agents via Self-Talk

Paper • 2401.05033 • Published Jan 10 • 14

upvoted a collection about 1 month ago

DBRX

Collection

DBRX is a mixture-of-experts (MoE) large language model trained from scratch by Databricks. • 3 items • Updated Mar 27 • 85

upvoted 2 papers about 1 month ago

Arcee's MergeKit: A Toolkit for Merging Large Language Models

Paper • 2403.13257 • Published Mar 20 • 16

Simple and Scalable Strategies to Continually Pre-train Large Language Models

Paper • 2403.08763 • Published Mar 13 • 48

upvoted a collection about 1 month ago

StructLM

Collection

The structure knowledge grounded language model • 6 items • Updated 25 days ago • 6

upvoted 7 papers about 2 months ago

StableDrag: Stable Dragging for Point-based Image Editing

Paper • 2403.04437 • Published Mar 7 • 23

Chatbot Arena: An Open Platform for Evaluating LLMs by Human Preference

Paper • 2403.04132 • Published Mar 7 • 38

DenseMamba: State Space Models with Dense Hidden Connection for Efficient Large Language Models

Paper • 2403.00818 • Published Feb 26 • 12

VisionLLaMA: A Unified LLaMA Interface for Vision Tasks

Paper • 2403.00522 • Published Mar 1 • 40

LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init Attention

Paper • 2303.16199 • Published Mar 28, 2023 • 3

The Era of 1-bit LLMs: All Large Language Models are in 1.58 Bits

Paper • 2402.17764 • Published Feb 27 • 560

GPTVQ: The Blessing of Dimensionality for LLM Quantization

Paper • 2402.15319 • Published Feb 23 • 19

upvoted a collection about 2 months ago

OpenMath

Collection

A collection of models and datasets introduced in "OpenMathInstruct-1: A 1.8 Million Math Instruction Tuning Dataset" • 15 items • Updated Feb 19 • 26

upvoted a paper about 2 months ago

OpenMathInstruct-1: A 1.8 Million Math Instruction Tuning Dataset

Paper • 2402.10176 • Published Feb 15 • 32

upvoted 3 collections about 2 months ago

LLM Leaderboard best models ❤️‍🔥

Collection

A daily uploaded list of models with best evaluations on the LLM leaderboard: • 65 items • Updated 2 days ago • 287

The Big Benchmarks Collection

Collection

Gathering benchmark spaces on the hub (beyond the Open LLM Leaderboard) • 12 items • Updated Feb 7 • 73

MPT

Collection

The MPT collections is a series of decoder-style transformer models trained from scratch by MosaicML. Details: https://www.mosaicml.com/mpt • 9 items • Updated Nov 27, 2023 • 7

upvoted a paper about 2 months ago

Judging LLM-as-a-judge with MT-Bench and Chatbot Arena

Paper • 2306.05685 • Published Jun 9, 2023 • 20

upvoted a paper 2 months ago

Textbooks Are All You Need II: phi-1.5 technical report

Paper • 2309.05463 • Published Sep 11, 2023 • 84

upvoted a collection 2 months ago

Zephyr 7B Gemma

Collection

Models, dataset, and Demo for Zephyr 7B Gemma. For code to train the models, see: https://github.com/huggingface/alignment-handbook • 5 items • Updated 19 days ago • 15

upvoted a paper 2 months ago

Self-Discover: Large Language Models Self-Compose Reasoning Structures

Paper • 2402.03620 • Published Feb 6 • 102

upvoted 13 papers 3 months ago

AutoMathText: Autonomous Data Selection with Language Models for Mathematical Texts

Paper • 2402.07625 • Published Feb 12 • 10

LiRank: Industrial Large Scale Ranking Models at LinkedIn

Paper • 2402.06859 • Published Feb 10 • 8

Code Representation Learning At Scale

Paper • 2402.01935 • Published Feb 2 • 12

InternLM-Math: Open Math Large Language Models Toward Verifiable Reasoning

Paper • 2402.06332 • Published Feb 9 • 17

DeepSpeed-FastGen: High-throughput Text Generation for LLMs via MII and DeepSpeed-Inference

Paper • 2401.08671 • Published Jan 9 • 12

Hydragen: High-Throughput LLM Inference with Shared Prefixes

Paper • 2402.05099 • Published Feb 7 • 16

Multilingual E5 Text Embeddings: A Technical Report

Paper • 2402.05672 • Published Feb 8 • 15

RLAIF: Scaling Reinforcement Learning from Human Feedback with AI Feedback

Paper • 2309.00267 • Published Sep 1, 2023 • 45

Nomic Embed: Training a Reproducible Long Context Text Embedder

Paper • 2402.01613 • Published Feb 2 • 13

EAGLE: Speculative Sampling Requires Rethinking Feature Uncertainty

Paper • 2401.15077 • Published Jan 26 • 17

SOLAR 10.7B: Scaling Large Language Models with Simple yet Effective Depth Up-Scaling

Paper • 2312.15166 • Published Dec 23, 2023 • 55

Dr. LLaMA: Improving Small Language Models in Domain-Specific QA via Generative Data Augmentation

Paper • 2305.07804 • Published May 12, 2023 • 2

HAE-RAE Bench: Evaluation of Korean Knowledge in Language Models

Paper • 2309.02706 • Published Sep 6, 2023 • 2

Unchun Yang

AI & ML interests

Organizations

ucyang's activity

What Makes a Dialog Agent Useful?

Can We Train Chat Models with Raw Data?

Making LLMs even more accessible with bitsandbytes, 4-bit quantization and QLoRA

Faster fine-tuning using TRL & Unsloth

Jack of All Trades, Master of Some, a Multi-Purpose Transformer Agent

Fine-tune Llama 3 with ORPO

Fine-tuning 20B LLMs with RLHF on a 24GB consumer GPU

Welcome Llama 3 - Meta's new open LLM

Synthetic data: save money, time and carbon with open source

Making LLMs lighter with AutoGPTQ and transformers