zero123plus
Guides Zero123Plus generation, ControlNet variants, demo launch, and Cog deployment workflows for single-image-to-multiview inference.
- License 许可
- Apache-2.0
- 最近更新
- 2026-09-03
- Stars
- ★ 351
- Forks
- 27
01 前言 Front Matter
- name
- zero123plus
- license
- Apache 2.0
- metadata
- { "disco-role": "operating" }
- description
- Guides Zero123Plus generation, ControlNet variants, demo launch, and Cog deployment workflows for single-image-to-multiview inference.
- disable-model-invocation
- true
02 资源文件 5
- 参考文档references/dependency-and-runtime.md3.8 KB
- 参考文档references/repo-provenance.md2.1 KB
- 参考文档references/repo-routing-metadata.json420 B
- 参考文档references/troubleshooting.md2.9 KB
- 脚本scripts/check_zero123plus_env.py6.9 KB
03 SKILL.md 正文
Zero123Plus
Use this skill when the task mentions Zero123Plus, single-image-to-multiview inference, depth ControlNet, normal generation, matting cleanup, Streamlit, Gradio, Docker/Gitpod, or Cog/Replicate deployment.
Start here
references/dependency-and-runtime.md: verified Python/runtime stack, CUDA expectations, and install recipe.references/troubleshooting.md: cross-cutting install/import/cache/backend issues and where to go next.references/repo-provenance.md: snapshot of the source checkout used to build this skill.scripts/check_zero123plus_env.py: safe environment checker for torch, CUDA, and the bundled optional modules.
Route map
- For base image-to-multiview generation, depth ControlNet, normal generation,
camera layout, and matting cleanup, read
sub-skills/generation/SKILL.md. - For Streamlit, Gradio, Docker/Gitpod, Cog/Replicate, weights, and demo
deployment behavior, read
sub-skills/deployment/SKILL.md.
Installation summary
This repository is GPU-first. The verified inspection stack used Python 3.11 and CUDA-capable PyTorch.
A representative setup is:
- Create a private Python 3.11 environment.
- Install a CUDA wheel for PyTorch and torchvision.
- Install the repo runtime stack summarized in
references/dependency-and-runtime.md. - Run
python scripts/check_zero123plus_env.py --check-onlyor--require-cudabefore a real generation run.
If you only need the generation path, start with the generation sub-skill. If you need local demos or deployment, start with the deployment sub-skill.
Minimal checks
python scripts/check_zero123plus_env.py --check-onlypython scripts/check_zero123plus_env.py --require-cuda
Notes
- The model weights are non-commercial even though the code is Apache 2.0.
- CUDA is required for actual generation; CPU is not a full substitute.
- Bundled scripts default to local-only loading and require
--allow-downloadbefore they fetch missing model weights.
04 同仓库更多 Skills
passnet-triton-opt
Make a single PassNet Triton kernel fast AND numerically faithful: performance model (when a replacement can win at all), block/grid/warp tuning, autotune policy, launch overhead, and per-op numeric recipes to pass the dtype baseline tolerances. Use when a pass matches and is correct but speedup below expected, or when correctness fails by small numeric margins.
passnet-solve
END-TO-END playbook for solving one PassNet sample: analyze the computation graphs, decide the optimization strategy (what to fuse, what to replace, what to leave alone), drive the iteration loop, and maximize the sample score. This is the ENTRY skill — invoke it first for any PassNet optimization task; it tells you when to use passnet-pattern-fusion, passnet-triton-opt and passnet-feedback.
passnet-skill
PassNet GPU kernel optimization via compiler passes. Design and implement Triton-based optimization passes, create pass files under ./pass_dir/, self-evaluate with pass_evaluator, and iterate to maximize GPU speedup.
passnet-pattern-fusion
Author PassNet pass files: write patterns that actually MATCH the FX graph, pick fusion regions (which ops to absorb into one Triton kernel), and structure multi-pass submissions with the mandatory shared-dispatch architecture. Use when creating or fixing pattern()/replacement_args()/replacement_func() files, when a pass "failed to match", or when deciding how to fuse multiple kernels/ops.
passnet-orchestrate
Round/batch planning and decision support for PassNet graph optimization. Use for multi-sample triage, worker allocation, eval-budget policy, and post-eval keep/revert decisions. For solving one concrete sample end to end, use passnet-solve as the entry skill; consult this skill only when you need broader planning or a specific decision gate.
passnet-feedback
Fast, deterministic feedback for PassNet work: pre-flight pattern/match verification WITHOUT burning a GPU evaluation, per-node bottleneck analysis of a sample's graphs, and eval-log parsing into per-variant status + estimated score + failure classification. Use BEFORE every GPU evaluation (check_pattern), at the START of a sample (analyze_graph), and AFTER every evaluation (parse_eval_log).