Stars
AI agents running research on single-GPU nanochat training automatically
A reference implementation of Hierarchical Federated Learning using Flower and PyTorch. HFL has been implemented simulating an automatic horizontal scaling of an intermediate level of edge servers.
The awesome collection of OpenClaw skills. 5,400+ skills filtered and categorized from the official OpenClaw Skills Registry.🦞
Repo for Visual Cue Enhancement and Dual Low-Rank Adaptation for Efficient Visual Instruction Fine-Tuning
a family of highly capabale yet efficient large multimodal models
论文X光机 — Claude Code Skill，解构学术论文，提炼餐巾纸公式
LoRA-FAIR: Federated LoRA Fine-Tuning with Aggregation and Initialization Refinement. The implementation is based on the paper accepted by ICCV 2025. Paper: https://arxiv.org/abs/2411.14961
🐫 CAMEL: The first and the best multi-agent framework. Finding the Scaling Law of Agents. https://www.camel-ai.org
"Paper2Slides: From Paper to Presentation in One Click"
TradingAgents: Multi-Agents LLM Financial Trading Framework
Every work on Federated Learning Pruning
[TMLR 2025] Stop Overthinking: A Survey on Efficient Reasoning for Large Language Models
CoCoFL: Communication- and Computation-Aware Federated Learning via Partial NN Freezing and Quantization
The official implementation of "PointLoRA: Low-Rank Adaptation with Token Selection for Point Cloud Learning" (CVPR 2025)
[ICML 2025] The Official implementation of our paper "Clients Collaborate: Flexible Differentially Private Federated Learning with Guaranteed Improvement of Utility-Privacy Trade-off"
Create Epic Math and Physics Animations & Study Notes From Text and Images.
Post-training with Tinker
A toolkit to optimize ML models for deployment for Keras and TensorFlow, including quantization and pruning.
Code for ICCV2025 paper——IDEATOR: Jailbreaking and Benchmarking Large Vision-Language Models Using Themselves
The Oyster series is a set of safety models developed in-house by Alibaba-AAIG, devoted to building a responsible AI ecosystem. | Oyster 系列是 Alibaba-AAIG 自研的安全模型，致力于构建负责任的 AI 生态。
A Framework of Small-scale Large Multimodal Models
Visual Instruction Tuning with 500x Fewer Parameters through Modality Linear Representation-Steering
EMER, OV-MER (ICML25), AffectGPT (ICML25, Oral), EmoPrefer (ICLR26)