Alibaba_Qwen 发布 Qwen3.8-Flash-Next，与 SGLang 合作，模型容量大，嵌入量大，支持长任务，值得期待。
Qwen3.8-Flash-Next 与 SGLang 部署，包含 125B 主模型和 51B N-gram 嵌入，GDN + QSA 混合注意力机制，Gated Residual 结构，Muon 训练。SGLang 为 Qwen4 新架构提供支持。
Big thanks to @sgl_project for the day-0 support! 🙌 Qwen3.8-Flash-Next is ready to deploy with SGLa...
Big thanks to @sgl_project for the day-0 support! 🙌 Qwen3.8-Flash-Next is ready to deploy with SGLang today. SGLang @sgl_project Congrats to @Alibaba_Qwen on launching Qwen3.8-Flash! SGLang is proud to be a day-0 partner supporting the new architecture preview for Qwen4. It's a 125B main model with 51B of N-gram embeddings and 6B activated per token. The 51B N-gram embeddings scale model capacity with almost no extra compute per token, and can sit in host memory with async prefetch instead of occupying GPU memory. The GDN + QSA hybrid attention gives you efficient memory and precise retrieval at the same time on long-horizon tasks, while Gated Residual gives the model 4 lanes instead of 1 to pass information between layers. And it's trained with Muon! We're excited for what's next with Qwen4, and we already have plenty of ideas for how to use the N-gram embeddings in new deployment setups. Stay tuned! Blog and cookbook in the comments👇 🔗 View Quoted Tweet 💬 7 🔄 2 ❤️ 120 👀 11800 📊 13 ⚡