Mr.WXS dba34d4915 [v0.18.0][Triton][Qwen3.5] delete expr for kernels args (#7646) ...

### What this PR does / why we need it?
Some parameters of Triton operators are unnecessarily modified with the
"constexpr" modifier. When these parameters change, recompilation is
triggered, which significantly affects the model performance. Therefore,
these parameters need to be rectified.
backport: https://github.com/vllm-project/vllm-ascend/pull/7482


Signed-off-by: w30012745 <wangxiaoshuai2@h-partners.com>
Co-authored-by: w30012745 <wangxiaoshuai2@h-partners.com>

2026-03-25 23:31:27 +08:00

..

__init__.py

mkdir triton package and move triton files (#4420)

2025-11-26 11:06:12 +08:00

chunk_delta_h.py

[Feature] Optimize Qwen3.5/Qwen3Next GDN prefill by prebuilding chunk metadata (#7487)

2026-03-22 23:09:23 +08:00

chunk_delta_hupdate.py

[Feature] Optimize Qwen3.5/Qwen3Next GDN prefill by prebuilding chunk metadata (#7487)

2026-03-22 23:09:23 +08:00

chunk_o_update.py

[Feat] support basic pcp&dcp for qwen3next (#6091)

2026-02-28 21:44:08 +08:00

chunk_o.py

[v0.18.0][Triton][Qwen3.5] delete expr for kernels args (#7646)

2026-03-25 23:31:27 +08:00

chunk_scaled_dot_kkt.py

[Feature] Optimize Qwen3.5/Qwen3Next GDN prefill by prebuilding chunk metadata (#7487)

2026-03-22 23:09:23 +08:00

chunk.py

[Feature] Optimize Qwen3.5/Qwen3Next GDN prefill by prebuilding chunk metadata (#7487)

2026-03-22 23:09:23 +08:00

cumsum.py

[Feature] Optimize Qwen3.5/Qwen3Next GDN prefill by prebuilding chunk metadata (#7487)

2026-03-22 23:09:23 +08:00

fused_qkvzba_split_reshape.py

[Feat]fused_qkvzba_split_reshape supports token number greater than 65536 (#6740)

2026-03-05 14:41:38 +08:00

l2norm.py

[v0.18.0][Triton][Qwen3.5] delete expr for kernels args (#7646)

2026-03-25 23:31:27 +08:00

layernorm_guard.py

[Lint]Style: Convert vllm-ascend/ to ruff format(Batch #12) (#6177)

2026-01-23 14:59:19 +08:00

sigmoid_gating.py

[Lint]Style: Convert vllm-ascend/ to ruff format(Batch #12) (#6177)

2026-01-23 14:59:19 +08:00

solve_tril.py

[Feature] Optimize Qwen3.5/Qwen3Next GDN prefill by prebuilding chunk metadata (#7487)

2026-03-22 23:09:23 +08:00

utils.py

[Feat] support basic pcp&dcp for qwen3next (#6091)

2026-02-28 21:44:08 +08:00

wy_fast.py

[v0.18.0][Triton][Qwen3.5] delete expr for kernels args (#7646)

2026-03-25 23:31:27 +08:00