精选理由
Robbyant 把 LingBot-VLA 2.0 的训练后代码开源了，单张 RTX 4090D 跑出 130 毫秒推理，不用集群就能自己玩。
Robbyant 开源了 LingBot-VLA 2.0 的训练后代码。在单张 NVIDIA GeForce RTX 4090D 上，仅需 10 步去噪，推理耗时约 130 毫秒。这意味着用户无需集群即可进行适配和测试，大幅降低了部署门槛。
原文 · elvis
LingBot-VLA 2.0 shows practicality too. Robbyant open-sourced the post-training code. Inference tak...
LingBot-VLA 2.0 shows practicality too. Robbyant open-sourced the post-training code. Inference takes about 130 ms on a single NVIDIA GeForce RTX 4090D using 10 denoising steps. That means you can adapt and test it without a cluster. 💬 1 🔄 0 ❤️ 3 👀 656 📊 2 ⚡