cosmos3-setup

Guide users through Cosmos3 installation, environment setup, checkpoint downloading, and verification. Use when the user asks "how do I install cosmos3", "how do I set up the environment", "how do I download checkpoints", "how do I use Docker", or any question about getting the package running for the first time.

설치로 이동 파일 검토

저장소: NVIDIA/cosmos-framework
최근 소스 활동: 2026년 10월 24일 15:23
감지된 SKILL.md 언어: 영어
스타: 588
포크: 188

설치 방법

기본적으로 소스를 먼저 확인하는 Prompt가 선택됩니다. 직접 명령으로 전환하거나 로컬 사본을 다운로드할 수도 있습니다.

AI 어시스턴트로 설치

이 프롬프트를 복사해 사용 중인 AI 어시스턴트에 입력하세요.

Follow https://skillsmp.com/skill-install/prompt.md to install only the Agent Skill at https://github.com/NVIDIA/cosmos-framework/tree/main/.agents/skills/cosmos3-setup.

직접 명령은 검토 Prompt를 거치지 않습니다. 실행하기 전에 소스를 확인하세요.

npx skills add https://github.com/NVIDIA/cosmos-framework --skill cosmos3-setup

명령은 한 줄로 유지됩니다. 복사하기 전에 가로로 스크롤해 전체 내용을 확인하세요.

로컬 사본을 원하시나요? SkillsMP에서 현재 제공할 수 있는 파일을 다운로드하세요.

소스 파일 검토

설치 여부를 결정하기 전에 SKILL.md와 SkillsMP에 표시된 보조 파일을 읽어 보세요.

SKILL.md

소스 지침 · 읽기 전용 미리보기

name: cosmos3-setup
description: Guide users through Cosmos3 installation, environment setup, checkpoint downloading, and verification. Use when the user asks "how do I install cosmos3", "how do I set up the environment", "how do I download checkpoints", "how do I use Docker", or any question about getting the package running for the first time.

# Cosmos3 Setup ## When to use this skill - Use when a user wants to install Cosmos3 or set up a development environment - Use when a user asks about system requirements, CUDA versions, or GPU compatibility - Use when a user needs to download model checkpoints or configure HuggingFace auth - Use when a user wants to run Cosmos3 inside a Docker container or NGC container - For errors during setup, hand off to the **cosmos3-env-troubleshoot** skill ## Path convention All paths below are relative to the cosmos3 package root (`../../../` from this skill file). All `uv run` / `python` commands should also be run from there. ## Where to find answers The canonical setup reference is `docs/setup.md`. The README (`README.md` § Setup) has the shortest quickstart. | User question | Go to | | ------------------------------------------------------ | -------------------------------------------------------------------- | | What are the system requirements? | `docs/setup.md` § System Requirements | | How do I install with uv? (sync, pip venv, pip system) | `docs/setup.md` § Virtual Environment | | How do I install with Docker? | `docs/setup.md` § Docker Container | | Custom torch/CUDA versions or attention backends? | `docs/setup.md` § Advanced | | Which CUDA version? (cu130 vs cu128) | `docs/setup.md` § CUDA Variants, `docs/faq.md` § Which CUDA version? | | How do I download checkpoints? | `docs/setup.md` § Downloading Base Checkpoints | | NGC container issues? | `docs/setup.md` § PyTorch Import Issue | | Any installation error | `../cosmos3-env-troubleshoot/SKILL.md` | ## Setup steps at a glance 1. **Clone** the repository and `cd` into the project root (the directory containing `pyproject.toml`) 2. **System deps**: `sudo apt-get install -y --no-install-recommends curl ffmpeg git-lfs libgl1 libglib2.0-0 libx11-dev tree wget` 3. **Install uv**: `curl -LsSf https://astral.sh/uv/0.11.29/install.sh | sh && source $HOME/.local/bin/env`. Verify `uv --version` matches `tool.uv.required-version`; update an existing installation with `uv self update 0.11.29`. 4. **Install package**: `uv sync --locked --all-extras --group=cu130-train && source .venv/bin/activate` (use `cu128-train` on older drivers without a supported CUDA 13 compatibility setup; the inference-only `cu130` / `cu128` groups omit the training extras) 5. **Checkpoints**: auto-downloaded during inference; requires HuggingFace auth (see docs) 6. **Verify**: follow `docs/setup.md` § Verify the GPU and media runtime. Check CUDA initialization and actual video decoding, not just package imports. ## Things not obvious from the docs - **NGC container caveat**: inherited libraries can conflict with PyTorch, but clearing `LD_LIBRARY_PATH` can also remove required CUDA compatibility libraries. Follow the setup guide's conditional troubleshooting instead of clearing it automatically. - **CUDA compatibility**: compare the host driver with the selected CUDA build's requirements. `nvidia-smi` reports driver capability, not the installed toolkit; its CUDA version need not equal `torch.version.cuda`. - **`HF_HOME`**: controls where checkpoints are cached (default: `~/.cache/huggingface`). Set this if disk space is tight or you want a shared cache. - **Conflicting env vars**: stale `HF_TOKEN` or `HUGGING_FACE_HUB_TOKEN` env vars can silently override CLI auth. Check with `printenv | grep HF_`. - **AutoML is a separate source install**: `uv sync --all-extras` does not install the core engine. For AutoML setup, follow `docs/automl.md#installation-from-source` and the `cosmos-automl` skill: ask whether to use `NVIDIA-TAO/tao-automl`'s `main` branch or a user-selected development branch before a new installation. Keep the recorded source revision for existing experiments; do not upgrade on resume. ## Related skills | Skill | When to use | | -------------------------------------- | ---------------------------------------------- | | `../cosmos3-inference/SKILL.md` | Running inference after setup is complete | | `../cosmos3-codebase-nav/SKILL.md` | Finding files, parameters, and configs in code | | `../cosmos3-env-troubleshoot/SKILL.md` | Debugging environment and runtime errors |

GitHub에서 보기

검색으로 돌아가기 설치로 돌아가기

cosmos3-qwen3-vl-robovqa-sft

Fine-tune Qwen3-VL-8B-Instruct, a vision-language model from outside the Cosmos family, with the Cosmos Framework posttrain SFT path on robot-video question answering (RoboVQA), on one GPU, and score it before and after with the same evaluator. Use when the user asks to "fine-tune Qwen3-VL with Cosmos Framework", "train a non-Cosmos model or a VLM in the framework", "run SFT on robot videos / RoboVQA", "reproduce the Qwen3-VL RoboVQA result", or wants a worked post-training example with a measured before and after (environments, data preparation, preflight, training, Hugging Face export, evaluation). Not for Cosmos3 checkpoints (use cosmos3-post-training) and not for LoRA (the SFT path trains every parameter).

2026-10-31

cosmos3-generate-media

Generate images, videos, and audio-synchronized videos with the Cosmos3 audiovisual generator models (Edge / Nano / Super / Super-4Step) through any of the seven supported serving paths: diffusers in-process, cosmos_framework torchrun, vLLM-Omni, SGLang, TensorRT-LLM, distilled TensorRT-LLM, or NIM. Use when the user asks "generate a video from a prompt", "run text2image / text2video / image2video with Cosmos3", "serve a Cosmos3 generator", or asks which Cosmos3 generation backend fits their GPU. Self-contained: sample prompts, reference images and negative prompts ship in this skill's assets/.

2026-10-30

cosmos3-skill-authoring

Author new Cosmos3 agent skills and their eval configs, and get them through the skill eval. Use when adding a skill under .agents/skills/, migrating a cookbook notebook or README into a skill, writing or updating an eval.config, choosing eval cases and assertions, fixing a skill after a failed or self-corrected eval cell, reading an eval report on the dashboard, or preparing a skill PR. Covers the required structure, the writing rules that decide first-attempt pass/fail, the eval.config schema (multi-case evals), how the evaluator judges, and how to run the eval in GitHub CI or locally. Do not use it to run the Cosmos3 workloads themselves (the per-workload skills do that) or to change the evaluator.

2026-10-30

이 저장소의 모든 Skills 보기

cosmos3-setup

소스 정보

설치 방법

AI 어시스턴트로 설치

소스 파일 검토

이 저장소의 다른 Skills

AI 어시스턴트로 설치

이 저장소의 다른 Skills