feat(humming): FP8 DeepEP dispatch for humming MoE backend (#31429)

2026-08-24 18:59:07 +08:00
parent 46b92b22e2
commit 21258b7a35
6 changed files with 543 additions and 38 deletions
@@ -35,7 +35,7 @@ dependencies = [
  "flash-attn-4>=4.0.0b18",
  "flashinfer_python[cu13]==0.6.17", # keep it aligned with jit-cache version in Dockerfile
  "gguf",
-  "humming-kernels[cu13]==0.1.10",
+  "humming-kernels[cu13]==0.1.12",
  "interegular",
  "IPython",
  "kernels>=0.14.1,<0.15",