[ckpt] feat: add mooncake backend

[x1314aq]

What does this PR do?

Add mooncake-transfer-engine backend to checkpoint-engine.

TODO

• Add detailed performance testing results in checkpoint engine README.

benchmark result test with bucket_size=2GB

Model	Device Info	time cost (s)	bandwidth (GB/s)
Qwen3-8B	8x Ascend 910C	0.27	50.26
Qwen3-8B	8x Nvidia H100, ConnectX-7 400Gbps (InfiniBand)	1.90	7.28

Checklist Before Starting

• Search for similar PRs. Paste at least one query link here: [ckpt] feat: add kimi ckpt engine backend
• Format the PR title as [{modules}] {type}: {description} (This will be checked by the CI)
  • {modules} include fsdp, megatron, veomni, sglang, vllm, rollout, trainer, ci, training_utils, recipe, hardware, deployment, ray, worker, single_controller, misc, perf, model, algo, env, tool, ckpt, doc, data, cfg, reward
  • If this PR involves multiple modules, separate them with , like [megatron, fsdp, doc]
  • {type} is in feat, fix, refactor, chore, test
  • If this PR breaks any API (CLI arguments, config, function signature, etc.), add [BREAKING] to the beginning of the title.
  • Example: [BREAKING][fsdp, megatron] feat: dynamic batching

Test

For changes that can not be tested by CI (e.g., algorithm implementation, new model support), validate by experiment(s) and show results like training curve plots, evaluation results, etc.

API and Usage Example

Demonstrate how the API changes if any, and provide usage example(s) if possible.

# Add code snippet or script demonstrating how to use this

Design & Code Changes

Demonstrate the high-level design if this PR is complex, and list the specific changes.

Checklist Before Submitting

Important

Please check all the following items before requesting a review, otherwise the reviewer might deprioritize this PR for review.

• Read the Contribute Guide.
• Apply pre-commit checks: pre-commit install && pre-commit run --all-files --show-diff-on-failure --color=always
• Add / Update the documentation.
• Add unit or end-to-end test(s) to the CI workflow to cover all the code. If not feasible, explain why: ...
• Once your PR is ready for CI, send a message in the ci-request channel in the verl Slack workspace. (If not accessible, please try the Feishu group (飞书群).)
• If your PR is related to the recipe submodule, please also update the reference to the submodule commit via git submodule update --remote or cd recipe && git pull origin main.

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.

---

yexin seems not to be a GitHub user. You need a GitHub account to be able to sign the CLA. If you have already a GitHub account, please add the email address used for this commit to your account.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}

[wuxibin89]

There's a ongoing work on this: #4954

[wuxibin89]

What is the difference between kimi_ckpt_engine and mooncake?

[x1314aq]

What is the difference between kimi_ckpt_engine and mooncake?

The mooncake backend implements D2D communication, so tensors don’t need to be offloaded to host DRAM. The kimi_ckpt_engine backend mainly uses H2D transfers and broadcast.