def forward(self, inps):
x = F.matmul(inps[0], inps[1], self.param["transA"],
self.param["transB"])
if self.param["alpha"] != 1.0:
x = F.mul(x, self.param["alpha"])
if len(inps) == 3:
if self.param["beta"] != 1.0:
x = F.add(x, F.mul(inps[2], self.param["beta"]))
else:
x = F.add(x, inps[2])
return x
def test_as_raw_tensor_from_int64():
x = np.arange(6, dtype="int64").reshape(2, 3)
xx = Tensor(x, dtype="float32", device="xpux")
yy = F.add(xx, 1).numpy()
assert xx.dtype == np.float32
assert xx.device == "xpux"
np.testing.assert_almost_equal(yy, x.astype("float32") + 1)
def test_add_input():
a = Tensor([1, 2])
b = Tensor([3, 4])
@trace(symbolic=True, capture_as_const=True)
def fwd(a, b):
return (a + b) * 2
fwd(a, b)
orig_model = io.BytesIO()
fwd.dump(orig_model,
arg_names=["a", "b"],
output_names="o",
optimize_for_inference=False)
orig_model.seek(0)
graph = Net.load(orig_model)
inp_c = graph.make_input_node((2, ), np.int32, name="c")
varo = graph.var_filter.name("o").as_unique()
out = F.add(varo, inp_c)
out.name = "o1"
graph.remove_output(varo)
graph.add_output(out)
modified_model = io.BytesIO()
graph.dump(modified_model)
modified_model.seek(0)
load_graph = GraphInference(modified_model)
out = load_graph.run(a, b, a)
np.testing.assert_equal(out["o1"], ((a + b) * 2 + a).numpy())
def test_add_output():
a = Tensor([1.0, 2.0])
b = Tensor([3.0, 4.0])
@trace(symbolic=True, capture_as_const=True)
def fwd(a, b):
return (a + b) * 2
fwd(a, b)
orig_model = io.BytesIO()
fwd.dump(
orig_model, arg_names=["a", "b"], output_names="o", optimize_for_inference=False
)
orig_model.seek(0)
net = Net.load(orig_model)
var_a = net.var_filter.name("a").as_unique()
var_b = net.var_filter.name("b").as_unique()
y = F.add(var_a, var_b)
y = F.sigmoid(y)
y.name = "o1"
net.add_output(y)
modified_model = io.BytesIO()
net.dump(modified_model)
modified_model.seek(0)
g = GraphInference(modified_model)
out = g.run(a.numpy(), b.numpy())
np.testing.assert_equal(out["o"], ((a + b) * 2).numpy())
np.testing.assert_equal(out["o1"], (F.sigmoid((a + b))).numpy())
def test_replace_vars():
g = mgb_graph.Graph()
g.options.async_exec_level = 0b100
device = "xpux"
dtype = np.float32
a = mgb_graph.InputNode(device=device, dtype=dtype, graph=g)
const = g.make_const(1.234)
a_plus_a = F.add(a.outputs[0], a.outputs[0])
a_plus_a_mul_const = F.mul(a_plus_a, const)
rst = F.add(a_plus_a_mul_const, a.outputs[0])
(new, ) = cgtools.replace_vars([rst._node], {const._node: a_plus_a._node})
out = mgb_graph.OutputNode(mgb_graph.VarNode(new))
func = g.compile(out.outputs[0])
func.execute()
x = make_dev_tensor(5.0, device=device)
a.set_value(x)
res = out.get_value().numpy()
np.testing.assert_equal(res, np.array([105.0]))
def test_module_elemwise():
def test_func(method, *inps):
elemwise = Elemwise(method)
outputs = elemwise(*inps)
return outputs.numpy()
x = np.random.rand(100).astype("float32")
y = np.random.rand(100).astype("float32")
x, y = tensor(x), tensor(y)
np.testing.assert_almost_equal(test_func("H_SWISH", x),
F.hswish(x).numpy(),
decimal=6)
np.testing.assert_almost_equal(test_func("ADD", x, y),
F.add(x, y).numpy(),
decimal=6)
def forward(self, inps):
return F.add(inps[0], inps[1])
import megengine as mge
import megengine.functional as F
A = mge.tensor([[2., 4., 2.],
[2., 4., 2.]])
B = mge.tensor([[1., 2., 1.],
[1., 2., 1.]])
print(A + B)
print(A - B)
print(A * B)
print(A / B)
print(F.add(A, B))
print(F.sub(A, B))
print(F.mul(A, B))
print(F.div(A, B))
A = mge.tensor([[1., 2., 3.],
[4., 5., 6.]])
print(A[1, :2])
A = mge.tensor([[1., 2., 3.],
[4., 5., 6.]])
print(A.shape)
A = A.reshape(3, 2)
print(A.shape)
x = mge.tensor([[1., 3., 5.],