[FRONTEND][ONNX]Gather operator added which is required for seq2seq models

nnvm/python/nnvm/frontend/onnx.py

@@ -446,7 +446,6 @@ def _impl_v1(cls, inputs, attr, params):
             inputs[0] = _sym.expand_dims(inputs[0], axis=axes, num_newaxis=1)
         return inputs[0]
 
-
 class Slice(OnnxOpConverter):
     """ Operator converter for Slice.
     """
@@ -487,6 +486,19 @@ def _impl_v1(cls, inputs, attr, params):
                                    'ends': 'end'},
                        ignores=['axes'])(inputs, attr)
 
+class Gather(OnnxOpConverter):
+    """ Operator converter for Gather.
+    """
+
+    @classmethod
+    def _impl_v1(cls, inputs, attr, params):
+        axis = attr['axis']
+        indices = np.array(attr['indices'], dtype='int32')
+        name = 'gather_indices'
+        gather_indices = _sym.Variable(name=name, init=indices)
+        params[name] = indices
+        return _sym.take(inputs[0], gather_indices, axis=axis)
+
 # compatible operators that do NOT require any conversion.
 _identity_list = []
 
@@ -593,7 +605,7 @@ def _get_convert_map(opset):
         'Split': AttrCvt('split', {'split': 'indices_or_sections'}),
         'Slice': Slice.get_converter(opset),
         'Transpose': AttrCvt('transpose', {'perm': 'axes'}),
-        # 'Gather'
+        'Gather': Gather.get_converter(opset),
         'Squeeze': Renamer('squeeze'),
         'Unsqueeze': Unsqueeze.get_converter(opset),
         'Pad': Pad.get_converter(opset),

nnvm/tests/python/frontend/onnx/test_forward.py

@@ -189,6 +189,34 @@ def test_unsqueeze():
 
     np.testing.assert_allclose(out_shape, tvm_out.shape)
 
+def verify_gather(in_shape, indices, axis=0):
+    indices_src = np.array(indices, dtype="int32")
+
+    x = np.random.uniform(size=in_shape)
+    out_np = np.take(x, indices_src, axis=axis)
+
+    y = helper.make_node("Gather", ['in'], ['out'], indices=indices, axis=axis)
+
+    graph = helper.make_graph([y],
+                              'gather_test',
+                              inputs = [helper.make_tensor_value_info("in",
+                                  TensorProto.FLOAT, list(in_shape))],
+                              outputs = [helper.make_tensor_value_info("out",
+                                  TensorProto.FLOAT, list(out_np.shape))])
+
+    model = helper.make_model(graph, producer_name='gather_test')
+
+    for target, ctx in ctx_list():
+        tvm_out = get_tvm_output(model, x, target, ctx, out_np.shape, 'float32')
+
+    np.testing.assert_allclose(out_np, tvm_out)
+
+def test_gather():
+    verify_gather((4,), [1])
+    verify_gather((4,), [0, 1, 2, 3])
+    verify_gather((4, 2), [1], 1)
+    verify_gather((4, 3, 5, 6), [2, 1, 0, 0], -2)
+
 def _test_slice_iteration(indata, outdata, starts, ends, axes=None):
     if axes:
         y = helper.make_node("Slice", ['in'], ['out'], axes=axes, starts=starts, ends=ends)
@@ -299,3 +327,4 @@ def test_matmul():
     test_ceil()
     test_clip()
     test_matmul()
+    test_gather()