Upgrade to cu130

[RunningLeon]

Thanks for your contribution and we appreciate it a lot. The following instructions would make your pull request more healthy and more easily receiving feedbacks. If you do not understand some items, don't worry, just make the pull request and seek help from maintainers.

Motivation

Please describe the motivation of this PR and the goal you want to achieve through this PR.

Modification

Please briefly describe what modification is made in this PR.

BC-breaking (Optional)

Does the modification introduce changes that break the backward-compatibility of the downstream repositories?
If so, please describe how it breaks the compatibility and how the downstream projects should modify their code to keep compatibility with this PR.

Use cases (Optional)

If this PR introduces a new feature, it is better to list some use cases here, and update the documentation.

Checklist

1. Pre-commit or other linting tools are used to fix the potential lint issues.
2. The modification is covered by complete unit tests. If not, please add more unit tests to ensure the correctness.
3. If the modification has a dependency on downstream projects of a newer version, this PR should be tested with all supported versions of downstream projects.
4. The documentation has been modified accordingly, like docstring or example tutorials.

[Copilot]

Pull request overview

Updates the CUDA 13 (cu130) container stack and runtime dependency caps to support newer PyTorch / Triton versions, while improving Docker build determinism by installing torch/torchvision from the correct PyTorch CUDA index before resolving the rest of the requirements.

Changes:

• Bump CUDA runtime dependency upper bounds (torch/torchvision/triton) and update Triton environment checker max version.
• Refresh Docker images/stages to CUDA 13.0.3 / CUDA 12.8.1 on Ubuntu 24.04 and introduce PYTORCH_CUDA_VERSION to decouple “base CUDA” from “PyTorch wheel CUDA channel”.
• Harden Docker install/build scripts: safer apt mirror rewrites, conditional Python PPA use, explicit torch install from the matching PyTorch index, and wheel presence/import checks.

Reviewed changes

Copilot reviewed 8 out of 8 changed files in this pull request and generated 1 comment.

Show a summary per file

File	Description
requirements/runtime_cuda.txt	Updates torch/torchvision/triton version constraints for CUDA runtime installs.
lmdeploy/pytorch/check_env/triton.py	Raises the maximum Triton version considered “tested”.
docker/prepare_3rdparty_wheel.sh	Adjusts CUDA 13 detection and changes flash-attn-3 wheel acquisition strategy.
docker/install.sh	Makes Docker installs more deterministic (torch from correct index), adds validation and missing-wheel checks.
docker/Dockerfile_dev	Moves to Ubuntu 24.04 and installs torch/torchvision from PYTORCH_CUDA_VERSION index first.
docker/Dockerfile	Updates CUDA stages/images, adds PYTORCH_CUDA_VERSION, and improves wheel mounting/install flow.
docker/build.sh	Generalizes CUDA 13 detection for NCCL wheel selection.
builder/manywheel/README.md	Updates builder image examples to include CUDA 13.0.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Pull request overview

Copilot reviewed 8 out of 8 changed files in this pull request and generated no new comments.