这是一个从头训练大语言模型的项目，包括预训练、微调和直接偏好优化，模型拥有1B参数，支持中英文。
☆854Feb 18, 2025Updated last year
Alternatives and similar repositories for mini_qwen
Users that are interested in mini_qwen are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.
Sorting:
- Train a 1B LLM with 1T tokens from scratch by personal☆810Apr 27, 2025Updated last year
- This is a repository used by individuals to experiment and reproduce the pre-training process of LLM.☆505May 1, 2025Updated last year
- 这是一个open-r1的复现项目，对0.5B、1.5B、3B、7B的qwen模型进行GRPO训练，观察到一些有趣的现象。☆63Apr 13, 2025Updated last year
- 本项目旨在分享大模型相关技术原理以及实战经验（大模型工程化、大模型应用落地）☆24,649Jul 1, 2026Updated last week
- 训练一个对中文支持更好的LLaVA模型，并开源训练代码和数据。☆82Sep 6, 2024Updated last year
- 1-Click AI Models by DigitalOcean Gradient • AdDeploy popular AI models on DigitalOcean Gradient GPU virtual machines with just a single click. Zero configuration with optimized deployments.
- 从零实现一个小参数量中文大语言模型。☆1,053Aug 22, 2024Updated last year
- 🧠「大模型」2小时完全从0训练64M的小参数LLM！Train a 64M-parameter LLM from scratch in just 2h!☆52,626Jun 28, 2026Updated last week
- Reproduce R1 Zero on Logic Puzzle☆2,453Mar 20, 2025Updated last year
- 轻量级 LLM Post-training 框架，支持 SFT、RLVR、On-Policy KD、Guide KD 及混合训练；实现单轮/多轮 Guide 蒸馏、多教师蒸馏、Reward 混合训练与自动化数据分流👩🎓👨🎓☆938Mar 8, 2026Updated 4 months ago
- minimal-cost for training 0.5B R1-Zero☆814May 14, 2025Updated last year
- 本项目对Deepseek-R1-Distill-Qwen-7B进行心理咨询CoT数据的LoRA微调，以进一步提升Deepseek-R1-Distill-Qwen-7B在心理咨询领域的慢思考能力。☆12Mar 11, 2025Updated last year
- 主要记录大语言大模型（LLMs） 算法（应用）工程师相关的知识及面试题☆14,664Jun 14, 2026Updated 3 weeks ago
- 从零构建大模型：从预训练到RLHF的完整实践☆2,674May 20, 2026Updated last month
- Unified Efficient Fine-Tuning of 100+ LLMs & VLMs (ACL 2024)☆72,921Updated this week
- Wordpress hosting with auto-scaling - Free Trial Offer • AdFully Managed hosting for WordPress and WooCommerce businesses that need reliable, auto-scalable performance. Cloudways SafeUpdates now available.
- 复现大模型相关算法及一些学习记录☆3,452Jul 2, 2026Updated last week
- 中文对话0.2B小模型（ChatLM-Chinese-0.2B），开源所有数据集来源、数据清洗、tokenizer训练、模型预训练、SFT指令微调、RLHF优化等流程的全部代码。支持下游任务sft微调，给出三元组信息抽取微调示例。☆1,717Apr 20, 2024Updated 2 years ago
- Train deepseek r1-like reasoning LLM with ease | 轻松训练1个deepseek r1类的推理LLM☆20Feb 15, 2025Updated last year
- 用于从头预训练+SFT一个小参数量的中文LLaMa2的仓库；24G单卡即可运行得到一个具备简单中文问答能力的chat-llama2.☆2,925May 21, 2024Updated 2 years ago
- 👀「大模型」2小时从0训练65M参数的视觉多模态VLM！Train a 65M-parameter VLM from scratch in just 2h!☆8,292Jun 28, 2026Updated last week
- 《开源大模型食用指南》针对中国宝宝量身打造的基于Linux环境快速微调（全参数/Lora）、部署国内外开源大模型（LLM）/多模态大模型（MLLM）教程☆31,200Jun 17, 2026Updated 3 weeks ago
- 从0到1构建一个MiniLLM (pretrain+sft+dpo实践中)☆552Mar 23, 2025Updated last year
- Fully open reproduction of DeepSeek-R1☆26,381Apr 2, 2026Updated 3 months ago
- verl/HybridFlow: A Flexible and Efficient RL Post-Training Framework☆22,312Updated this week
- GPU virtual machines on DigitalOcean Gradient AI • AdGet to production fast with high-performance AMD and NVIDIA GPUs you can spin up in seconds. The definition of operational simplicity.
- MedicalGPT: Training Your Own Medical GPT Model with ChatGPT Training Pipeline. 训练医疗大模型，实现了包括增量预训练(PT)、有监督微调(SFT)、RLHF、DPO、ORPO、GRPO。☆5,575Jun 3, 2026Updated last month
- 《大模型白盒子构建指南》：一个全手搓的Tiny-Universe☆4,945Feb 12, 2026Updated 4 months ago
- An Easy-to-use, Scalable and High-performance Agentic RL Framework based on Ray (PPO & DAPO & REINFORCE++ & VLM & TIS & vLLM & Ray & Asy…☆9,756Updated this week
- Use PEFT or Full-parameter to CPT/SFT/DPO/GRPO 600+ LLMs (Qwen3.6, DeepSeek-V4, GLM-5.1, InternLM3, Llama4, ...) and 300+ MLLMs (Qwen3-VL…☆14,701Updated this week
- 整理开源的中文大语言模型，以规模较小、可私有化部署、训练成本较低的模型为主，包括底座模型，垂直领域微调及应用，数据集与教程等。☆22,666May 10, 2026Updated last month
- Building a VLM model starts from the basic module.☆18Apr 7, 2024Updated 2 years ago
- Phi2-Chinese-0.2B 从0开始训练自己的Phi2中文小模型，支持接入langchain加载本地知识库做检索增强生成RAG。Training your own Phi2 small chat model from scratch.☆592Jul 11, 2024Updated last year
- Building DeepSeek R1 from Scratch☆780Mar 21, 2025Updated last year
- Search-R1: An Efficient, Scalable RL Training Framework for Reasoning & Search Engine Calling interleaved LLM based on veRL☆5,051Nov 13, 2025Updated 7 months ago
- Managed Kubernetes at scale on DigitalOcean • AdDigitalOcean Kubernetes includes the control plane, bandwidth allowance, container registry, automatic updates, and more for free.
- EasyR1: An Efficient, Scalable, Multi-Modality RL Training Framework based on veRL☆5,046Apr 6, 2026Updated 3 months ago
- 本项目利用医学领域的 CoT 数据对 Deepseek-R1-Distill-Qwen-7B 进行微调，通过 QLoRA 量化和 Unsloth 加速训练，显著提升模型在复杂医学推理任务中的慢思考能力。知识蒸馏技术使轻量级模型获得大模型的推理优势，实现高效、准确且具有解释性…☆47Mar 10, 2025Updated last year
- something for paper agent☆11Dec 18, 2024Updated last year
- ☆147Sep 29, 2024Updated last year
- ☆34Jul 8, 2025Updated last year
- A very simple GRPO implement for reproducing r1-like LLM thinking.☆1,700Nov 21, 2025Updated 7 months ago
- RLHF experiments on a single A100 40G GPU. Support PPO, GRPO, REINFORCE, RAFT, RLOO, ReMax, DeepSeek R1-Zero reproducing.☆80Feb 19, 2025Updated last year