[sgl-kernel] Support FlashInfer top_k_top_p_sampling_from_logits (#9060)

Co-authored-by: luoyuan.luo <luoyuan.luo@antgroup.com>
2025-08-14 10:56:36 -07:00
parent 432f2053dd
commit 53dcc750b6
6 changed files with 349 additions and 5 deletions
@@ -85,7 +85,9 @@ from sgl_kernel.moe import (
 )
 from sgl_kernel.sampling import (
    min_p_sampling_from_probs,
+    top_k_mask_logits,
    top_k_renorm_prob,
+    top_k_top_p_sampling_from_logits,
    top_k_top_p_sampling_from_probs,
    top_p_renorm_prob,
    top_p_sampling_from_probs,