hub / github.com/benoitsteiner/tensorflow-opencl / gradient

Method gradient

tensorflow/python/eager/backprop.py:754–779 · view source on GitHub ↗

Computes the gradient using information traced by the tape. Args: target: the tensor to be differentiated. sources: a list of Tensors or Variables, the target will be differentiated with respect to the sources. Returns: a list of Tensors (or IndexedSlices, or None)

(self, target, sources)

Source from the content-addressed store, hash-verified

source not stored for this graph (policy: none)

Callers 6

testGradientTapeMethod · 0.80

testGradientTapeVariableMethod · 0.80

FunctionLibraryDefinitionMethod · 0.80

AddLibraryMethod · 0.80

TEST_FFunction · 0.80

GetGradDefsFunction · 0.80

Calls

no outgoing calls

Tested by 4

testGradientTapeMethod · 0.64

testGradientTapeVariableMethod · 0.64

TEST_FFunction · 0.64

GetGradDefsFunction · 0.64