Unsloth 推出 AMD 版本，可在 AMD GPU 上训练和运行 LLM
Introducing Unsloth for AMD 🚀
Unsloth 发布 AMD 版本，用户现在可以在 AMD 硬件上训练和运行 LLM。该版本与 AMD 合作实现，支持在 AMD GPU 上训练和运行 500+ 模型，覆盖 Windows、WSL、Linux，并可在 3GB VRAM 上训练 Qwen、Gemma。官方称借助自研 Triton kernel 和数学算法，在 Radeon、Instinct、Ryzen 及数据中心 GPU 上最高提速 2 倍、显存占用减少 70% 且无精度损失，同时提供针对 GGUF 与 Safetensors 推理优化的 ROCm 构建。
正文
当前显示原文，中文全文译文尚未就绪。
Introducing Unsloth for AMD 🚀 You can now train & run LLMs on your AMD hardware
• We collaborated with AMD to enable you to train & run 500+ models on AMD GPUs • Works on Windows, WSL, Linux • Train Qwen, Gemma on 3GB VRAM
Works on Radeon, Instinct, Ryzen and data center GPUs with up to 2× faster with 70% less VRAM and no accuracy loss via our custom Triton kernels and math algorithms. We also support optimized ROCm builds for GGUF & Safetensors inference.
Unsloth is an open-source local UI for faster LLM training and inference, with tool-call healing, code execution, secure web search, remote APIs, and HTTPS deployment. Connect local models to Claude Code, Codex agents and run the latest Kimi, GLM, DeepSeek, Qwen3.6, and Gemma 4 models.