TensorFlow 1 version | View source on GitHub |
Representative dataset to evaluate optimizations.
tf.lite.RepresentativeDataset(
input_gen
)
A representative dataset that can be used to evaluate optimizations by the converter. E.g. converter can use these examples to estimate (min, max) ranges by calibrating the model on inputs. This can allow converter to quantize a converted floating point model.
Args | |
---|---|
input_gen
|
an input generator that can be used to generate input samples
for the model. This must be a callable object that returns an object
that supports the iter() protocol (e.g. a generator function). The
elements generated must have same type and shape as inputs to the model.
|