2 Commits

Author	SHA1	Message	Date
pu-zhe	57fd6e4bd9	[Refact.]: refactoring 310p-kv cache allocator, align with main branch (#6270) ... ### What this PR does / why we need it? refactoring 310p-kv cache allocator, align with main branch vLLM version: v0.14.0 vLLM main: https://github.com/vllm-project/vllm-ascend/pull/6270 Qwen2.5-7B E2E Test --------- Signed-off-by: pu-zhe <puzhe1@h-partners.com> Signed-off-by: pu-zhe <zpuaa@outlook.com> Co-authored-by: pu-zhe <puzhe1@h-partners.com>	2026-01-27 16:26:48 +08:00
Shaoxu Cheng	fbae41697e	[310P]: refactoring for 310p kvcache and some ops class (#6117) ... ### What this PR does / why we need it? * Refactor the LayerNorm and activation operator classes to decouple the 310P device implementation from the main branch. * Refactor `mm_encoder_attention` on 310P to use the `torch_npu._npu_flash_attention_unpad` operator. * Refactor the QKV inputs in the prefill stage of `attention_v1` on 310P so they are no longer padded to 16× alignment. * Refactor `model_runner` on 310P to align the KV-cache initialization logic with the mainline implementation. ### Does this PR introduce _any_ user-facing change? NO ### How was this patch tested? use the e2e tests. - vLLM version: v0.13.0 - vLLM main: d68209402d --------- Signed-off-by: Tflowers-0129 <2906339855@qq.com>	2026-01-24 20:34:29 +08:00