Ze-Yi LIN
Zeyi-Lin
CPO@SwanLab; Ph.D. Student@XDU;
Languages
Top Repositories
⚡️HivisionIDPhotos: a lightweight and efficient AI ID photos tools. 一个轻量级的AI证件照制作算法。
大语言模型微调，Qwen2VL、Qwen2、GLM4指令微调
Qwen3 Fine-tuning: Medical R1 Style Chat
Stable Diffusion模型训练样例代码
BERT-IMDB电影评论情感分类实战：SwanLab可视化训练
Repositories
51⚡️HivisionIDPhotos: a lightweight and efficient AI ID photos tools. 一个轻量级的AI证件照制作算法。
大语言模型微调，Qwen2VL、Qwen2、GLM4指令微调
BERT-IMDB电影评论情感分类实战：SwanLab可视化训练
Train deepseek r1-like reasoning LLM with ease | 轻松训练1个deepseek r1类的推理LLM
Qwen3 Fine-tuning: Medical R1 Style Chat
No description provided.
PyTorch音频分类实战
Medical Image Segmentation Tutorial Case Studies - 医学影像分割教程案例
No description provided.
No description provided.
verl: Volcano Engine Reinforcement Learning for LLMs
Stable Diffusion模型训练样例代码
No description provided.
强化学习实战教程：从QLearning, DQN, PPO 到 LLM RL
🧐SwanLab: a tool for your machine learning log tracking and experiment management.
Scalable toolkit for efficient model reinforcement
Train speculative decoding models effortlessly and port them smoothly to SGLang serving.
Visualize the PaddleOCR2PyTorch project online to make the PaddleOCR experience and deployment easier.
🤗 Transformers: the model-definition framework for state-of-the-art machine learning models in text, vision, audio, and multimodal models, for both inference and training.
《开源大模型食用指南》基于Linux环境快速部署开源大模型，更适合中国宝宝的部署教程
📚 从零开始的大语言模型原理与实践教程
将SmolVLM2的视觉头与Qwen3-0.6B模型进行了拼接微调
Use PEFT or Full-parameter to finetune 450+ LLMs (Qwen2.5, InternLM3, GLM4, Llama3.3, Mistral, Yi1.5, Baichuan2, DeepSeek-R1, ...) and 150+ MLLMs (Qwen2.5-VL, Qwen2-Audio, Llama3.2-Vision, Llava, InternVL2.5, MiniCPM-V-2.6, GLM4v, Xcomposer2.5, Yi-VL, DeepSeek-VL2, Phi3.5-Vision, GOT-OCR2, ...).
EasyR1: An Efficient, Scalable, Multi-Modality RL Training Framework based on veRL
Unified Efficient Fine-Tuning of 100+ LLMs (ACL 2024)
No description provided.
No description provided.
No description provided.
🤗 LeRobot: Making AI for Robotics more accessible with end-to-end learning
An Easy-to-use, Scalable and High-performance RLHF Framework (70B+ PPO Full Tuning & Iterative DPO & LoRA & RingAttention & RFT)