[Ascend] support qwen3.5 35BA3B

[wanfengcxz]

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

Adds Ascend/DLinfer support plumbing needed to run Qwen3.5 35B A3B (notably the linear-attention / gated-delta path) by propagating device_type, extending attention metadata, and initializing Ascend-specific Triton device properties.

Changes:

• Propagate device_type into HF→ModelConfig build path and use it to pick bf16/fp16 state dtypes for Qwen3.5 configs.
• Extend DLInfer attention metadata and populate Ascend attention metadata with GDN-related sequence info.
• Add Ascend backend initialization for Triton device properties with fallback warnings when triton-ascend is missing.

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 4 comments.

File	Description
lmdeploy/pytorch/configurations/qwen3_5.py	Use device_type when selecting bf16 vs fp16 for Qwen3.5 state caches.
lmdeploy/pytorch/config.py	Thread device_type through ModelConfig.from_hf_config into the model config builder.
lmdeploy/pytorch/backends/dlinfer/attention.py	Add has_initial_state field to DLInfer attention metadata.
lmdeploy/pytorch/backends/dlinfer/ascend/op_backend.py	Populate GDN sequence metadata and initialize Triton device properties for Ascend backend.

---

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

[grimoire]

LGTM