zhutian
36 lines
994 B
Python
36 lines
994 B
Python
import numpy as np
|
|
import tensorflow as tf
|
|
|
|
from onnx_tf.handlers.backend_handler import BackendHandler
|
|
from onnx_tf.handlers.handler import onnx_op
|
|
from onnx_tf.handlers.handler import tf_func
|
|
|
|
|
|
@onnx_op("Hardmax")
|
|
@tf_func(tf.contrib.seq2seq.hardmax)
|
|
class Hardmax(BackendHandler):
|
|
|
|
@classmethod
|
|
def _common(cls, node, **kwargs):
|
|
x = kwargs["tensor_dict"][node.inputs[0]]
|
|
axis = node.attrs.get("axis", 1)
|
|
axis = axis if axis >= 0 else len(np.shape(x)) + axis
|
|
|
|
if axis == len(np.shape(x)) - 1:
|
|
return [cls.make_tensor_from_onnx_node(node, **kwargs)]
|
|
|
|
shape = tf.shape(x)
|
|
cal_shape = (tf.reduce_prod(shape[0:axis]),
|
|
tf.reduce_prod(shape[axis:tf.size(shape)]))
|
|
x = tf.reshape(x, cal_shape)
|
|
|
|
return [tf.reshape(tf.contrib.seq2seq.hardmax(x), shape)]
|
|
|
|
@classmethod
|
|
def version_1(cls, node, **kwargs):
|
|
return cls._common(node, **kwargs)
|
|
|
|
@classmethod
|
|
def version_11(cls, node, **kwargs):
|
|
return cls._common(node, **kwargs)
|