def forward_can_be_applied(node: onnx_op.ONNXOp, state: SDFGState,
sdfg: SDFG) -> bool:
input0_dim = len(in_desc_with_name(node, state, sdfg, "A").shape)
input1_dim = len(in_desc_with_name(node, state, sdfg, "B").shape)
if input0_dim == 1 or input1_dim == 1:
return False
return True
def forward(node: onnx_op.ONNXOp, state: SDFGState,
sdfg: SDFG) -> typing.Union[Node, SDFG]:
input_dtype = in_desc_with_name(node, state, sdfg, "X").dtype
cast_lambda = "lambda x: max(x, dace.{}(0))".format(
input_dtype.to_string())
def prog(X, Y):
Y[:] = dace.elementwise(cast_lambda, X)
return program_for_node(prog, sdfg, state, node)
def forward_in_desc_with_name(forward_node: nd.Node, context: BackwardContext,
name) -> dt.Data:
""" Find the descriptor of the data that connects to input connector `name`.
:param forward_node: the node.
:param context: the backward context.
:param name: the input connector name.
:return: the descriptor of the data that connects to connector `name`.
"""
return utils.in_desc_with_name(forward_node, context.forward_state,
context.forward_sdfg, name)
def forward(node: onnx_op.ONNXOp, state: SDFGState,
sdfg: SDFG) -> typing.Union[Node, SDFG]:
node.validate(sdfg, state)
dtype = in_desc_with_name(node, state, sdfg, 'X').dtype
tanh_lambda = "lambda x: dace.{}(1) / x".format(dtype.to_string())
def prog(X, Y):
Y[:] = dace.elementwise(tanh_lambda, X)
return program_for_node(prog, sdfg, state, node)
def forward(node: onnx_op.ONNXOp, state: SDFGState,
sdfg: SDFG) -> typing.Union[Node, SDFG]:
input_desc = in_desc_with_name(node, state, sdfg, "input")
output_desc = out_desc_with_name(node, state, sdfg, "output")
if (input_desc.dtype == output_desc.dtype):
def prog(input, output):
output[:] = input
else:
def prog(input, output):
output[:] = dace.elementwise(lambda x: x, input)
return program_for_node(prog, sdfg, state, node)
def forward_can_be_applied(node: onnx_op.ONNXOp, state: SDFGState,
sdfg: SDFG) -> bool:
if (in_desc_with_name(node, state, sdfg,
"input").dtype == out_desc_with_name(
node, state, sdfg, "output").dtype):
return True
target_type = node.to
try:
converters.onnx_tensor_type_to_typeclass(target_type)
except ValueError:
return False
return True
def program_for_node(program, sdfg: SDFG, state: SDFGState,
node: onnx_op.ONNXOp) -> SDFG:
""" Expand a function to a dace program.
The dtypes for the arguments will be extracted by matching the parameter names to edges.
"""
input_names = node.schema.non_variadic_inputs()
variadic_input_names = node.schema.variadic_inputs()
output_names = node.schema.non_variadic_outputs()
variadic_output_names = node.schema.variadic_outputs()
if set(input_names).intersection(output_names):
# this is currently the case for only one onnx op
raise ValueError(
"program_for_node cannot be applied on nodes of this type;"
" '{}' is both an input and an output".format(
next(input_names.intersection(output_names))))
params = inspect.signature(program).parameters
annotations = {}
for name, param in params.items():
if name in input_names or ("__" in name
and parse_variadic_param(name)[0]
in variadic_input_names):
annotations[name] = in_desc_with_name(node, state, sdfg, name)
elif name in output_names or ("__" in name
and parse_variadic_param(name)[0]
in variadic_output_names):
annotations[name] = out_desc_with_name(node, state, sdfg, name)
else:
raise ValueError(
"'{}' was not found as an input or output for {}".format(
name, node.schema.name))
program.__annotations__ = annotations
result = DaceProgram(program, (), {}, False, dace.DeviceType.CPU)
result.name = node.label + "_expansion"
sdfg = result.to_sdfg()
if node.schedule in [dtypes.ScheduleType.GPU_Default
] + dtypes.GPU_SCHEDULES:
sdfg.apply_gpu_transformations()
return sdfg
def forward(node: onnx_op.ONNXOp, state: SDFGState,
sdfg: SDFG) -> typing.Union[Node, SDFG]:
axis = node.axis
reduced_shape = list(
copy.deepcopy(in_desc_with_name(node, state, sdfg, "input").shape))
reduced_shape[axis] = 1
def prog(input, output):
max = np.max(input, axis=axis)
max_keepdims = np.reshape(max, reduced_shape)
exp_arr = np.exp(input - max_keepdims)
sum = np.sum(exp_arr, axis=axis)
sum_keepdims = np.reshape(sum, reduced_shape)
output[:] = exp_arr / sum_keepdims
return program_for_node(prog, sdfg, state, node)
def forward(node: onnx_op.ONNXOp, state: SDFGState,
sdfg: SDFG) -> typing.Union[Node, SDFG]:
node.validate(sdfg, state)
nsdfg = dace.SDFG(node.label + "_expansion")
nstate = nsdfg.add_state()
for e in node.iter_inputs_in_onnx_order(state):
nsdfg.add_datadesc(
e.dst_conn, in_desc_with_name(node, state, sdfg, e.dst_conn))
for e in node.iter_outputs_in_onnx_order(state):
nsdfg.add_datadesc(
e.src_conn, out_desc_with_name(node, state, sdfg, e.src_conn))
create_einsum_sdfg(None,
nsdfg,
nstate,
node.equation.replace(" ", ""),
*(e.dst_conn
for e in node.iter_inputs_in_onnx_order(state)),
output="Output")
return nsdfg
def backward(
forward_node: Node, context: BackwardContext,
given_gradients: typing.List[typing.Optional[str]],
required_gradients: typing.List[typing.Optional[str]]
) -> typing.Tuple[Node, BackwardResult]:
reduction_type = detect_reduction_type(forward_node.wcr)
if len(given_gradients) != 1:
raise AutoDiffException(
"recieved invalid SDFG: reduce node {} should have exactly one output edge"
.format(forward_node))
if len(required_gradients) != 1:
raise AutoDiffException(
"recieved invalid SDFG: reduce node {} should have exactly one input edge"
.format(forward_node))
input_name = next(iter(required_gradients))
in_desc = in_desc_with_name(forward_node, context.forward_state,
context.forward_sdfg, input_name)
output_name = next(iter(given_gradients))
out_desc = out_desc_with_name(forward_node, context.forward_state,
context.forward_sdfg, output_name)
all_axes: typing.List[int] = list(range(len(in_desc.shape)))
reduce_axes: typing.List[
int] = all_axes if forward_node.axes is None else forward_node.axes
non_reduce_axes: typing.List[int] = [
i for i in all_axes if i not in reduce_axes
]
result = BackwardResult.empty()
if reduction_type is dtypes.ReductionType.Sum:
# in this case, we need to simply scatter the grad across the axes that were reduced
sdfg = SDFG("_reverse_" + str(reduction_type).replace(".", "_") +
"_")
state = sdfg.add_state()
rev_input_conn_name = "input_gradient"
rev_output_conn_name = "output_gradient"
result.required_grad_names[output_name] = rev_output_conn_name
result.given_grad_names[input_name] = rev_input_conn_name
_, rev_input_arr = sdfg.add_array(rev_input_conn_name,
shape=out_desc.shape,
dtype=out_desc.dtype)
_, rev_output_arr = sdfg.add_array(rev_output_conn_name,
shape=in_desc.shape,
dtype=in_desc.dtype)
state.add_mapped_tasklet(
"_distribute_grad_" + str(reduction_type).replace(".", "_") +
"_", {
"i" + str(i): "0:{}".format(shape)
for i, shape in enumerate(in_desc.shape)
}, {
"__in":
Memlet.simple(
rev_input_conn_name,
"0" if forward_node.axes is None else ",".join(
"i" + str(i) for i in non_reduce_axes))
},
"__out = __in", {
"__out":
Memlet.simple(rev_output_conn_name,
",".join("i" + str(i) for i in all_axes),
wcr_str="lambda x, y: x + y")
},
external_edges=True)
return context.backward_state.add_nested_sdfg(
sdfg, None, {rev_input_conn_name},
{rev_output_conn_name}), result
else:
raise AutoDiffException(
"Unsupported reduction type '{}'".format(reduction_type))
def forward_can_be_applied(node: onnx_op.ONNXOp, state: SDFGState,
sdfg: SDFG) -> bool:
return in_desc_with_name(node, state, sdfg, 'X').dtype in [
dace.float16, dace.float32, dace.float64
]
def forward(node: onnx_op.ONNXOp, state: SDFGState,
sdfg: SDFG) -> typing.Union[Node, SDFG]:
node.validate(sdfg, state)
A_desc = in_desc_with_name(node, state, sdfg, "A")
B_desc = in_desc_with_name(node, state, sdfg, "B")
Y_desc = out_desc_with_name(node, state, sdfg, "Y")
input0_dim = A_desc.shape
input1_dim = B_desc.shape
# list containing letters from z-a
letters = [chr(ord('z') - i) for i in range(26)]
# i j k are used for the last dimensions
letters = [l for l in letters if l not in ['i', 'j', 'k']]
if len(input0_dim) == 1:
if len(input1_dim) != 2:
raise ValueError("invalid dimensions")
arg1 = 'k'
arg2 = 'kj'
result = 'j'
elif len(input1_dim) == 1:
if len(input0_dim) != 2:
raise ValueError("invalid dimensions")
arg1 = 'ik'
arg2 = 'k'
result = 'i'
else:
# build the einsum. The last two dimensions are always just the matrix multiply einsum
# dace will later specialize to a batched matmul if possible
arg1 = 'ik'
arg2 = 'kj'
result = 'ij'
if input0_dim[-2] != input0_dim[-1]:
if dace.symbolic.issymbolic(input0_dim[-2]):
log.warning(
f"overriding symbol {input0_dim[-2]} with value {input1_dim[-1]} in descriptor of input A of node {node}"
)
new_shape = list(A_desc.shape)
new_shape[-1] = input1_dim[-2]
A_desc.shape = new_shape
elif dace.symbolic.issymbolic(input1_dim[-1]):
log.warning(
f"overriding symbol {input0_dim[-1]} with value {input0_dim[-2]} in descriptor of input B of node {node}"
)
new_shape = list(B_desc.shape)
new_shape[-2] = input0_dim[-1]
B_desc.shape = new_shape
input0_dim = input0_dim[:-2]
input1_dim = input1_dim[:-2]
for dim0, dim1 in itertools.zip_longest(reversed(input0_dim),
reversed(input1_dim)):
if dim0 is None:
# only dim0 exists
letter = letters.pop()
arg2 = letter + arg2
result = letter + result
elif dim1 is None:
# only dim1 exists
letter = letters.pop()
arg1 = letter + arg1
result = letter + result
else:
# both exist
letter = letters.pop()
arg1 = letter + arg1
arg2 = letter + arg2
result = letter + result
einsum_str = '{},{}->{}'.format(arg1, arg2, result)
# we lower to an ONNXEinsum node instead straight to the dace einsum to make the autodiff simpler
nsdfg = dace.SDFG(node.label + "_expansion")
nstate = nsdfg.add_state()
einsum_node: nodes.LibraryNode = onnx_op.ONNXEinsum(
node.label + "_einsum_expansion", equation=einsum_str)
nstate.add_node(einsum_node)
einsum_node.add_in_connector("Inputs__0")
einsum_node.add_in_connector("Inputs__1")
nsdfg.add_datadesc("A", copy.deepcopy(A_desc))
nsdfg.add_datadesc("B", copy.deepcopy(B_desc))
nsdfg.add_datadesc("Y", copy.deepcopy(Y_desc))
nsdfg.arrays["A"].transient = False
nsdfg.arrays["B"].transient = False
nsdfg.arrays["Y"].transient = False
nstate.add_edge(nstate.add_read("A"), None, einsum_node, "Inputs__0",
nsdfg.make_array_memlet("A"))
nstate.add_edge(nstate.add_read("B"), None, einsum_node, "Inputs__1",
nsdfg.make_array_memlet("B"))
nstate.add_edge(einsum_node, "Output", nstate.add_write("Y"), None,
nsdfg.make_array_memlet("Y"))
return nsdfg