feat(mge/opr): add cumsum

GitOrigin-RevId: 740f00a8e5c66253934c676c2fcdf02fe8e2d313

imperative/python/megengine/functional/tensor.py

@@ -27,6 +27,7 @@ __all__ = [
     "broadcast_to",
     "concat",
     "cond_take",
+    "cumsum",
     "expand_dims",
     "eye",
     "flatten",
@@ -1328,3 +1329,35 @@ def roll(
     if shp_bak is not None:
         out = out.reshape(shp_bak)
     return out
+
+
+def cumsum(inp: Tensor, axis: int):
+    """
+    Computes the cumulative sum of elements along given axis.
+
+    :param inp: input tensor.
+    :param axis: axis along which cumsum is performed.
+
+    Examples:
+
+    .. testcode::
+
+        from megengine import tensor
+        import megengine.functional as F
+
+        x = tensor([[1, 2, 3], [4, 5, 6]], "int32")
+        y = F.cumsum(x, 1)
+        print(y.numpy())
+
+    Outputs:
+
+    .. testoutput::
+
+        [[ 1  3  6]
+         [ 4  9 15]]
+
+    """
+    assert isinstance(inp, Tensor), "input of cumsum must be type of Tensor"
+    assert axis >= 0 and axis < inp.ndim, "input axis {} out of bound".format(axis)
+    op = builtin.Cumsum(axis=axis, exclusive=False, reverse=False)
+    return apply(op, inp)[0]

imperative/src/impl/ops/specializations.cpp

@@ -673,4 +673,16 @@ OP_TRAIT_REG(SlidingWindowTranspose, SlidingWindowTranspose)
     .fallback();
 }} // sliding_window_transpose
 
+namespace {
+namespace cumsum {
+auto apply_on_var_node(const OpDef& def, const VarNodeArray& inputs) {
+    auto&& op = static_cast<const Cumsum&>(def);
+    OperatorNodeConfig config{op.make_name()};
+    return opr::Cumsum::make(inputs[0], op.param(), config);
+}
+
+OP_TRAIT_REG(Cumsum, Cumsum).apply_on_var_node(apply_on_var_node).fallback();
+}  // namespace cumsum
+}  // namespace
+
 } // namespace mgb::imperative

src/core/include/megbrain/ir/ops.td

@@ -377,4 +377,6 @@ def CheckHasInf: MgbHashableOp<"CheckHasInf", [EmptyParam]>;
 
 def FastpathCopy: MgbHashableOp<"FastpathCopy">;
 
+def Cumsum: MgbHashableOp<"Cumsum", [CumsumParam]>;
+
 #endif // MGB_OPS