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

Method permute

tensorrt_llm/functional.py:420–424  ·  view source on GitHub ↗

See functional.permute.

(self, dims)

Source from the content-addressed store, hash-verified

418 return flatten(self, start_dim, end_dim)
419
420 def permute(self, dims):
421 '''
422 See functional.permute.
423 '''
424 return permute(self, dims)
425
426 def transpose(self, dim0, dim1):
427 '''

Callers 15

_p_matmul_ogsFunction · 0.45
swizzle_dataMethod · 0.45
unswizzle_mx_scale_cdna4Function · 0.45
swizzle_dataMethod · 0.45
unswizzle_dataMethod · 0.45
swizzle_dataMethod · 0.45
unswizzle_dataMethod · 0.45
swizzle_dataMethod · 0.45
run_testFunction · 0.45
reshape_softmaxFunction · 0.45
reshape_softmax_newFunction · 0.45
mha_refFunction · 0.45

Calls 1

permuteFunction · 0.70

Tested by 15

run_testFunction · 0.36
verify_kv_cacheMethod · 0.36
_split_headsMethod · 0.36
_merge_headsMethod · 0.36
interleave_scalesMethod · 0.36
test_gpt_float32Method · 0.36
test_loadersMethod · 0.36
test_mamba_conv1dMethod · 0.36
test_permuteMethod · 0.36
run_testFunction · 0.36