vllm.models.glm5next.nvidia
¶
Modules:
-
mega_moe–GLM-5.3-Flash adapter for the DeepGEMM MegaMoE kernel.
-
ops– -
sparse_indexer–Custom Sparse Attention Indexer layers.
vllm.models.glm5next.nvidia
¶Modules:
mega_moe
–
GLM-5.3-Flash adapter for the DeepGEMM MegaMoE kernel.
ops
–
sparse_indexer
–
Custom Sparse Attention Indexer layers.