MCPcopy Index your code
hub / github.com/NVIDIA/TensorRT-LLM / has_moe_tp

Method has_moe_tp

tensorrt_llm/mapping.py:309–310  ·  view source on GitHub ↗
(self)

Source from the content-addressed store, hash-verified

307 return self.moe_cluster_size > 1
308
309 def has_moe_tp(self):
310 return self.moe_tp_size > 1
311
312 def has_moe_ep(self):
313 return self.moe_ep_size > 1

Callers 6

__init__Method · 0.80
convert_layerFunction · 0.80
convert_layerFunction · 0.80
load_weights_from_hf_safetensorsFunction · 0.80
convert_layerFunction · 0.80
convert_hf_qwenFunction · 0.80

Calls

no outgoing calls

Tested by

no test coverage detected