Global

Fine-tune an 8B model on a 4 GB laptop GPU. Layer streaming keeps the frozen base out of VRAM and feeds it to the GPU one decoder layer at a time. 32 GB.

🤖 Agent 应用层 - ✅ Agent / Workflow - LangGraph、OpenAI Agents SDK - AutoGen、CrewAI、Pydantic AI - Dify、n8n、Flowise - ✅ 任务形态 - Research / Coding / Web Agent - Multi-Agent / Supervisor / Handoff - Computer Use / Browser Automation

> 第八章聚焦 GRPO、OPD、Search-R1 与 ReTool。更多 Agentic RL 算法、训练代码与实验实践，可以前往作者持续维护的另一个仓库 agentic-rl-lab；该仓库更新频率更高，会持续跟进新的算法与环境。

🚀 An open-source, hands-on curriculum bridging the gap from basic RL concepts to LLM alignment, RLVR, and advanced Agentic systems.

I implemented a transformer model from scratch using PyTorch, based on the paper Attention is All You Need. You can use my scripts to train your own billion or million parameter LLM using a single GPU.

🍔 In addition, ms-swift integrates the latest training technologies, including Megatron parallelism techniques such as TP, PP, CP, EP to accelerate training, as well as numerous GRPO algorithm family reinforcement learning algorithms including: GRPO, DAPO, GSPO, SAPO, CISPO…

A high-performance framework for training LLMs, VLMs, diffusion, and embodied models on NVIDIA GPUs and Kunlun XPUs. 3, Kimi-K3 and more.

5, our scalable data pipeline of environments, solvers, and verifiers. 5 Computer Use on Online-Mind2Web.

> New project: JevHarness — Let an LLM build task-specific decision harnesses with Jev, with optional refinement using rewards and execution traces. Includes an interactive Pokémon demo.

A fully open reproduction of DeepSeek-R1. This repo is a work in progress, let's build it together!

BISHENG is an open LLM devops platform for next generation Enterprise AI applications.

8-Flash-Next is the foundation model developed by Qwen Team, Alibaba Group.

\ No Kev-9B JevBench result is published. These are Kev-8B (Qwen3).

A local-first trajectory compiler built on DeepSeek Harness.