MCPcopy Create free account
hub / github.com/alibaba/nann / hessians

Function hessians

tensorflow/tensorflow/python/ops/gradients_impl.py:333–396  ·  view source on GitHub ↗

Constructs the Hessian of sum of `ys` with respect to `x` in `xs`. `hessians()` adds ops to the graph to output the Hessian matrix of `ys` with respect to `xs`. It returns a list of `Tensor` of length `len(xs)` where each tensor is the Hessian of `sum(ys)`. The Hessian is a matrix of seco

(ys,
             xs,
             name="hessians",
             colocate_gradients_with_ops=False,
             gate_gradients=False,
             aggregation_method=None)

Source from the content-addressed store, hash-verified

source not stored for this graph (policy: none)

Callers 8

AddToScalarAccumulatorFunction · 0.85
AddToTensorAccumulatorFunction · 0.85
HessiansV2Function · 0.85
ComputeMethod · 0.85
ComputeMethod · 0.85
AddInstanceStatsToMapFunction · 0.85

Calls 11

reshapeMethod · 0.80
TensorArrayMethod · 0.80
gradientsFunction · 0.70
sizeMethod · 0.45
constantMethod · 0.45
while_loopMethod · 0.45
writeMethod · 0.45
shapeMethod · 0.45
stackMethod · 0.45
concatMethod · 0.45
appendMethod · 0.45

Tested by

no test coverage detected