Hao AI Lab 发布 FastVideo FastH3,可在单张消费级显卡本地生成带音频视频
1/8) FastVideo FastH3 now runs on a single consumer machine: @NVIDIA RTX GPUs 💚, DGX Spark and @Apple Silicon 🍎
Hao AI Lab 发布 FastVideo FastH3,可在单台消费级设备上运行,支持 NVIDIA RTX GPU、DGX Spark 和 Apple Silicon。FastH3 V2 在 8 步内超过 MiniMax H3 基座模型,单张 RTX 5090 生成 5 秒 480p 带音频片段耗时 15 秒,768p 约 35–39 秒。同时发布 FastH3 Trim,体积比基座 H3 小 4.2 倍,最低 8 GB 显存即可运行,并已开放模型、Cookbook 与代码。
正文
当前显示原文,中文全文译文尚未就绪。
1/8) FastVideo FastH3 now runs on a single consumer machine: @NVIDIA RTX GPUs 💚, DGX Spark and @Apple Silicon 🍎
FastH3 V2 beats base @MiniMax_AI H3 in 8 steps. A 5 s 480p clip with audio takes 15 s on one RTX 5090.
We also release FastH3 Trim: 4.2× smaller than base H3, and it runs in as little as 8 GB of GPU memory.
Every clip here was generated locally. @NVIDIAAI @NVIDIARTXSpark