AI研究論文デイリー

LLaVaOLMoBitnet1B: Ternary LLM goes Multimodal!

SwiftBrush v2: Make Your One-step Diffusion Model Better Than Its Teacher

K-Sort Arena: Efficient and Reliable Benchmarking for Generative Models via K-wise Human Preferences

Foundation Models for Music: A Survey

LlamaDuo: LLMOps Pipeline for Seamless Migration from Service LLMs to Small-Scale Local LLMs

NanoFlow: Towards Optimal Large Language Model Serving Throughput

Learning to Move Like Professional Counter-Strike Players

Power Scheduler: A Batch Size and Token Number Agnostic Learning Rate Scheduler

Efficient Detection of Toxic Prompts in Large Language Models

TVG: A Training-free Transition Video Generation Method with Diffusion Models

SWE-bench-java: A GitHub Issue Resolving Benchmark for Java

MagicMan: Generative Novel View Synthesis of Humans with 3D-Aware Diffusion and Iterative Refinement

Training-free Long Video Generation with Chain of Diffusion Model Experts

MobileQuant: Mobile-friendly Quantization for On-device Language Models