linfeng-yuan 96309e2b79 [ops] support advanced apply_top_k_top_p without top_k constraint (#6098) ...

### What this PR does / why we need it?
Implement `apply_top_k_top_p` via ascendC to eliminate the constraint of
k [1,1024]. It enables high performance TopKTopP calculation and avoid
D2H synchronization introduced by k validation.

### Does this PR introduce _any_ user-facing change?
No.

### How was this patch tested?
E2E serving with `k=4096` and  `p=0.95`
- vLLM version: v0.13.0
- vLLM main:
d68209402d

---------

Signed-off-by: linfeng-yuan <1102311262@qq.com>
Signed-off-by: SlightwindSec <slightwindsec@gmail.com>
Co-authored-by: SlightwindSec <slightwindsec@gmail.com>

2026-01-26 09:08:42 +08:00

..

__init__.py

Spec decode support for V1 Engine (#874)

2025-05-23 14:25:46 +08:00

rejection_sampler.py

[Feature] add the magicmtp speculative decoding acceleration algorithm (#5542)

2026-01-08 09:15:55 +08:00

sampler.py

[ops] support advanced apply_top_k_top_p without top_k constraint (#6098)

2026-01-26 09:08:42 +08:00