ONNX import/export: Sample_multinomial

python/mxnet/contrib/onnx/mx2onnx/_op_translations.py

@@ -1655,3 +1655,25 @@ def convert_size(node, **kwargs):
     and return the created node.
     """
     return create_basic_op_node('Size', node, kwargs)
+
+@mx_op.register("_sample_multinomial")
+def convert_multinomial(node, **kwargs):
+    """Map MXNet's multinomial operator attributes to onnx's
+    Multinomial operator and return the created node.
+    """
+    name, input_nodes, attrs = get_inputs(node, kwargs)
+    dtype = onnx.mapping.NP_TYPE_TO_TENSOR_TYPE[np.dtype(attrs.get("dtype", 'int32'))]
+    sample_size = convert_string_to_list(attrs.get("shape", '1'))
+    if len(sample_size) < 2:
+        sample_size = sample_size[-1]
+    else:
+        raise AttributeError("ONNX currently supports integer sample_size only")
+    node = onnx.helper.make_node(
+        "Multinomial",
+        input_nodes,
+        [name],
+        dtype=dtype,
+        sample_size=sample_size,
+        name=name,
+    )
+    return [node]

python/mxnet/contrib/onnx/onnx2mx/_import_helper.py

@@ -18,7 +18,7 @@
 # coding: utf-8_
 # pylint: disable=invalid-name
 """Operator attributes conversion"""
-from ._op_translations import identity, random_uniform, random_normal
+from ._op_translations import identity, random_uniform, random_normal, sample_multinomial
 from ._op_translations import add, subtract, multiply, divide, absolute, negative, add_n
 from ._op_translations import tanh, arccos, arcsin, arctan, _cos, _sin, _tan
 from ._op_translations import softplus, shape, gather, lp_pooling, size
@@ -48,6 +48,7 @@
     'RandomNormal'      : random_normal,
     'RandomUniformLike' : random_uniform,
     'RandomNormalLike'  : random_normal,
+    'Multinomial'       : sample_multinomial,
     # Arithmetic Operators
     'Add'               : add,
     'Sub'               : subtract,

python/mxnet/contrib/onnx/onnx2mx/_op_translations.py

@@ -38,6 +38,19 @@ def random_normal(attrs, inputs, proto_obj):
     new_attr = translation_utils._fix_attribute_names(new_attr, {'mean' : 'loc'})
     return 'random_uniform', new_attr, inputs
 
+def sample_multinomial(attrs, inputs, proto_obj):
+    """Draw random samples from a multinomial distribution."""
+    try:
+        from onnx.mapping import TENSOR_TYPE_TO_NP_TYPE
+    except ImportError:
+        raise ImportError("Onnx and protobuf need to be installed. "
+                          + "Instructions to install - https://github.com/onnx/onnx")
+    new_attrs = translation_utils._remove_attributes(attrs, ['seed'])
+    new_attrs = translation_utils._fix_attribute_names(new_attrs, {'sample_size': 'shape'})
+    new_attrs['dtype'] = TENSOR_TYPE_TO_NP_TYPE[int(new_attrs['dtype'])]
+    return 'sample_multinomial', new_attrs, inputs
+
+
 # Arithmetic Operations
 def add(attrs, inputs, proto_obj):
     """Adding two tensors"""

tests/python-pytest/onnx/export/mxnet_export_test.py

@@ -263,6 +263,22 @@ def test_softmax():
     # Comparing result of forward pass before using onnx export, import
     npt.assert_almost_equal(result, softmax_out)
 
+@with_seed()
+def test_multinomial():
+    input1 = np.array([0, 0.1, 0.2, 0.3, 0.4]).astype("float32")
+    shape = (10,)
+    inputs = [helper.make_tensor_value_info("input1", TensorProto.FLOAT, shape=np.shape(input1))]
+    outputs = [helper.make_tensor_value_info("output", TensorProto.FLOAT, shape=shape)]
+    nodes = [helper.make_node("Multinomial", ["input1"], ["output"], sample_size=shape, dtype=6)]
+    graph = helper.make_graph(nodes,
+                              "multinomial_test",
+                              inputs,
+                              outputs)
+    multinomial_model = helper.make_model(graph)
+    bkd_rep = backend.prepare(multinomial_model)
+    output = bkd_rep.run([input1])
+    assert output[0].shape == shape
+
 @with_seed()
 def test_comparison_ops():
     """Test greater, lesser, equal"""