Support temporary implicit input

[disktnk]

fixes #201

def forward(self, x):
    x + chainer.Variable(...)

Then, second value's reference is lost when convert to graph, so converter cannot set the value into ONNX graph .

To keep such a value reference, exporter retains input values in forwarding using LinkHook and patched apply. By introducing LinkHook, the hook function only retains temporary values to reduce memory usage, but the target model is required to use forward function, not __call__ function.

• temporary implicit inputs are set as "value"(initializer), not as "constant".
• drop export_implicit_inputs_public option, because converter cannot distinguish the temporary value is covered by converter or not. For example, beta of batch-normalization is sometimes None but followed by converter.

[codecov-io]

Codecov Report

Merging #202 into master will increase coverage by 0.18%. The diff coverage is 97.05%.

@@            Coverage Diff             @@
##           master     #202      +/-   ##
==========================================
+ Coverage   89.69%   89.88%   +0.18%     
==========================================
  Files          24       24              
  Lines        1417     1473      +56     
==========================================
+ Hits         1271     1324      +53     
- Misses        146      149       +3

Impacted Files	Coverage Δ
onnx_chainer/graph.py	95.65% <100%> (ø)	:arrow_up:
onnx_chainer/export.py	91.47% <97.01%> (+1.47%)	:arrow_up:
onnx_chainer/functions/normalization.py	86.44% <0%> (-6.78%)	:arrow_down:
onnx_chainer/functions/math.py	85.71% <0%> (-0.54%)	:arrow_down:
onnx_chainer/functions/__init__.py	100% <0%> (ø)	:arrow_up:
onnx_chainer/mapping.py	90% <0%> (ø)	:arrow_up:
onnx_chainer/functions/pooling.py	88.73% <0%> (+5.59%)	:arrow_up:

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update 88994fb...8148f6e. Read the comment docs.

[disktnk]

/test

[pfn-ci-bot]

Successfully created a job for commit 0a2d8ab:

• Dashboard for commit 0a2d8ab

[shinh]

Sorry if I'm confused, but this does not work for the following case?

    def test_implicit_input_value(self):
        class Model(chainer.Chain):

            def __init__(self):
                super(Model, self).__init__()
                self.frac = np.array(4, dtype=np.float32)

            def forward(self, x):
                return x / self.frac

        x = chainer.Variable(np.array(1, dtype=np.float32))
        self.expect(Model(), x, name='implicit_input_value')

[disktnk]

Good catch. If inputs are ndarray type, function node converts them to chainer.Variable within itself, and will be lost out of function node (apply). I fixed it.

[disktnk]

/test

[pfn-ci-bot]

Successfully created a job for commit 8148f6e:

• Dashboard for commit 8148f6e