MCPcopy
hub / github.com/NVIDIA/TensorRT-LLM / optimize

Function optimize

tensorrt_llm/graph_rewriting.py:639–645  ·  view source on GitHub ↗
(net)

Source from the content-addressed store, hash-verified

637
638
639def optimize(net):
640 patterns = RewritePatternManager()
641 patterns.add(
642 label="fuse_attention_with_bias",
643 pattern=FuseAttentionWithBiasPass(),
644 )
645 patterns.rewrite(net)

Callers 1

buildFunction · 0.85

Calls 4

rewriteMethod · 0.95
RewritePatternManagerClass · 0.85
FuseAttentionWithBiasPassClass · 0.85
addMethod · 0.45

Tested by

no test coverage detected