[Feat.][310P]: weightNZ feature with quant or unquant.

[Tflowers-0129]

NZ Format Support for Linear Layers: Implemented support for the NZ (N-dimensional Z-order) format for linear layer weights on Ascend 310P, enhancing performance for both quantized and unquantized layers.
Unquantized Linear Method for Ascend 310P: Introduced AscendUnquantizedLinearMethod310 to specifically handle and apply NZ format casting to unquantized linear layer weights during the loading process.
MRotaryEmbedding Integration: Extended Rotary Embedding support by adding AscendMRotaryEmbedding310 to provide an Ascend-specific implementation for MRotaryEmbedding.
Quantization Method Updates: Updated the w8a8_static quantization method to directly transpose weights and apply NZ format casting, ensuring consistency with the new format.

• vLLM version: v0.15.0
• vLLM main: vllm-project/vllm@9562912

[gemini-code-assist]

Summary of Changes

Hello @Tflowers-0129, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

This pull request introduces significant optimizations for linear layer operations on Ascend 310P by enabling the use of the NZ (N-dimensional Z-order) tensor format for weights. This change aims to improve computational efficiency for both quantized and unquantized linear layers by optimizing memory access patterns. Additionally, it expands the compatibility of Rotary Embedding by integrating support for MRotaryEmbedding within the Ascend 310P framework, further enhancing model performance and feature parity.

Highlights

• NZ Format Support for Linear Layers: Implemented support for the NZ (N-dimensional Z-order) format for linear layer weights on Ascend 310P, enhancing performance for both quantized and unquantized layers.
• Unquantized Linear Method for Ascend 310P: Introduced AscendUnquantizedLinearMethod310 to specifically handle and apply NZ format casting to unquantized linear layer weights during the loading process.
• MRotaryEmbedding Integration: Extended Rotary Embedding support by adding AscendMRotaryEmbedding310 to provide an Ascend-specific implementation for MRotaryEmbedding.
• Quantization Method Updates: Updated the w8a8_static quantization method to directly transpose weights and apply NZ format casting, ensuring consistency with the new format.

🧠 New Feature in Public Preview: You can now enable Memory to help Gemini Code Assist learn from your team's feedback. This makes future code reviews more consistent and personalized to your project's style. Click here to enable Memory in your admin console.

Changelog

• tests/ut/_310p/quantization/test_w8a8_static_310.py
  • Updated npu_quant_matmul mock assertions to explicitly verify positional and keyword arguments.
  • Added a new test case to confirm that npu_format_cast is invoked during the weight processing phase for NZ format.
• vllm_ascend/_310p/ops/linear.py
  • Added AscendUnquantizedLinearMethod310 to apply NZ format casting to unquantized linear layer weights.
  • Introduced AscendLinearBase310 to integrate the new unquantized linear method into the Ascend 310P operations.
• vllm_ascend/_310p/ops/rotary_embedding.py
  • Imported MRotaryEmbedding from vllm.model_executor.layers.rotary_embedding.
  • Added AscendMRotaryEmbedding310 to provide an Ascend-specific implementation for MRotaryEmbedding.
• vllm_ascend/_310p/quantization/methods/w8a8_static.py
  • Modified the apply method to transpose layer.weight.data before passing it to torch_npu.npu_quant_matmul.
  • Updated process_weights_after_loading to apply torch_npu.npu_format_cast for NZ format to layer.weight.data.
• vllm_ascend/_310p/quantization/modelslim_config.py
  • Updated the get_quant_method function to utilize the newly introduced AscendUnquantizedLinearMethod310 for skipped linear layers.
• vllm_ascend/utils.py
  • Imported AscendMRotaryEmbedding310.
  • Registered AscendMRotaryEmbedding310 in the REGISTERED_ASCEND_OPS dictionary for Ascend 310P operations.

Activity

• No specific activity (comments, reviews, progress) has been recorded for this pull request yet.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

Feature	Command	Description
Code Review	/gemini review	Performs a code review for the current pull request in its current state.
Pull Request Summary	/gemini summary	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in pull request comments and review comments.
Help	/gemini help	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Footnotes

1. Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

[gemini-code-assist]

Code Review

This pull request introduces support for the weightNZ format on the 310P platform for both quantized and unquantized weights. The changes are generally in the right direction, but I've identified two significant issues. First, there's a critical performance regression in AscendMRotaryEmbedding310 where an optimized NPU kernel is bypassed. Second, there's an inefficiency in the w8a8_static quantization method due to transposing the weight tensor on every forward pass. I've provided detailed feedback and suggestions to address these points.

[github-actions]

👋 Hi! Thank you for contributing to the vLLM Ascend project. The following points will speed up your PR merge:‌‌

• A PR should do only one thing, smaller PRs enable faster reviews.
• Every PR should include unit tests and end-to-end tests ‌to ensure it works and is not broken by other future PRs.
• Write the commit message by fulfilling the PR description to help reviewer and future developers understand.

If CI fails, you can run linting and testing checks locally according Contributing and Testing.