infinite-livestream README

The project's README from github.com/reactor-team/infinite-livestream. As one markdown file: README.md.

infinite-livestream

An end-to-end, chat-driven, never-ending AI video broadcast. Viewers type !prompt <idea> in Twitch or YouTube chat; an LLM expands each idea into a styled sequence of scenes; the fast-h3 model generates them as 768p video clips with synchronized audio; and the stream goes out over RTMP as one uninterrupted broadcast.

Two halves, one contract:

Folder What it is Runs on
fast-h3/ The model: a queue of prompt-driven clip generations with explicit/auto playback. A reactor CLI workspace. Reactor Runtime, 8x B200
streaming-client/ The client: chat → prompt upsampling → scene groups → the model's queue → paced RTMP output. reactor-sdk (Python), any box with ffmpeg

They meet on the wire: fast-h3/fasth3_types.py is the client-facing contract (commands, messages, tracks), and the streaming client speaks exactly that — through reactor-sdk, the Reactor Python SDK, which carries the session, the commands and messages, and the WebRTC media tracks.

The model

The video generator is FastH3 Preview v1: MiniMax-H3 (35B) distilled by the FastVideo project down to four transformer forwards with 90% sparse video attention, generating video and audio jointly from text. The model, the distillation, and the inference engine this repo serves are FastVideo's work — fast-h3/ wraps them in a queue-and-playout contract for the Reactor platform.

Quickstart

Serve the model (locally with reactor run from fast-h3/, or deploy it), then:

cd streaming-client
pip install -r requirements.txt     # ffmpeg must be on PATH for RTMP
cp .env.example .env                # keys, style, sink, chat channel
python main.py --local --sink noop  # dry run against a local runtime
python main.py                      # everything from .env

Documentation

License

The code is Apache License 2.0 — see LICENSE and NOTICE. The model weights are licensed separately under the MiniMax H3 Community License, inherited from the base model MiniMaxAI/MiniMax-H3.

infinite-livestream (reactor-team/infinite-livestream)

Searching Database…

241 ·GitHub
Infinite-livestream
这是什么？一个端到端、由聊天驱动、永不落幕的 AI 视频直播：观众在 Twitch 或 YouTube 聊天区输入 !prompt，LLM 把这个想法扩写成带风格的场景，FastH3 模型将其渲染为带同步音频的 768p 片段，并通过 RTMP 持续播出。
这是一个应用，还是两个独立的部分？它分为两半：fast-h3/ 运行视频生成模型，需要配备 8 块 B200 GPU 的 Reactor Runtime；streaming-client/ 负责聊天接入、提示词上采样和按节奏控制的 RTMP 输出，只需要带 reactor-sdk 的 Python 外加 ffmpeg。两者通过在 fasth3types.py 中定义的共享契约进行通信。
怎么让它跑起来？先启动模型服务，可以用 fast-h3/ 里的 `reactor run` 在本地运行，也可以部署出去；然后启动 streaming-client，它会拉取聊天命令、将场景生成请求排队发给模型，并按节奏通过 RTMP 推送输出。
运行模型需要什么硬件？模型需要 8 块 B200 GPU 上的 Reactor Runtime 来服务 FastH3 Preview v1；流媒体客户端本身只需要带 reactor-sdk 的 Python，以及任何装了 ffmpeg 的机器。
我能在普通的单 GPU 机器上运行这个模型吗？这个模型是围绕 8 块 B200 GPU 上的 Reactor Runtime 构建的，详细的权重布局、GPU/CUDA 前提条件和性能表现单独记录在 fast-h3/README.md 中，而不是这份顶层 README 里。
FastH3 Preview v1 是什么？它是 MiniMax-H3（350 亿参数）经 FastVideo 项目蒸馏后只需四次 transformer 前向传播、采用 90% 稀疏视频注意力的版本，可直接从文本联合生成视频和音频。
直播的提示词实际来自哪里？观众直接在 Twitch 或 YouTube 聊天区输入 !prompt，LLM 会把每个想法扩写成带风格的场景序列，然后才进入生成队列。
有没有可以直接测试模型队列的最小客户端？有，fast-h3/client/README.md 记录了一个最小的 reactor-sdk 冒烟测试客户端，它只走一遍原始队列契约，不涉及完整的流媒体流水线。
What are similar projects?FreeVideo, hyperflow, MiniMax-H3, ComfyUI-MiniMax-H3-LongMedia, CodeVideoModel, vllm-omni, sglang, ComfyUI-ALLinONE-MinimaxH3
Which topics is it in?knowledge-distillation, rtmp, minimax-h3, livestream, minimax, ai-video, inference-engine, audio-generation, webrtc, text-to-video, pretrained-models, youtube
Creator
reactor-team
Size
0.8 MB
Stars
241
Updated
CommitAug 30
Age
1mo
Commits · 90 Days