def test_adaptive_pool(self, hexagon_session: Session, dshape, out_size,
pool_type, layout):
dtype = "float32"
np_data = np.random.uniform(low=0, high=255, size=dshape).astype(dtype)
np_out = tvm.topi.testing.adaptive_pool(np_data, out_size, pool_type,
layout)
oshape = np_out.shape
data = te.placeholder(dshape, name="data", dtype=dtype)
if len(out_size) == 2:
out = topi.nn.adaptive_pool(data, out_size, pool_type, layout)
else:
assert len(out_size) == 3
out = topi.nn.adaptive_pool3d(data, out_size, pool_type, layout)
target_hexagon = tvm.target.hexagon("v68")
with tvm.target.Target(target_hexagon):
fschedule = topi.hexagon.schedule_adaptive_pool
s = fschedule(out)
func = tvm.build(
s,
[data, out],
tvm.target.Target(target_hexagon, host=target_hexagon),
name="adaptive-pool",
)
mod = hexagon_session.load_module(func)
dev = hexagon_session.device
a = tvm.nd.array(np_data, dev)
b = tvm.nd.array(np.zeros(get_const_tuple(oshape), dtype=out.dtype),
dev)
mod["adaptive-pool"](a, b)
tvm.testing.assert_allclose(b.numpy(), np_out, rtol=4e-5, atol=1e-6)
def test_add(hexagon_session: Session):
dtype = "int8"
A = tvm.te.placeholder((2, ), dtype=dtype)
B = tvm.te.placeholder((1, ), dtype=dtype)
C = tvm.te.compute(A.shape, lambda i: A[i] + B[0], name="C")
sched = tvm.te.create_schedule(C.op)
target_hexagon = tvm.target.hexagon("v68", link_params=True)
func = tvm.build(sched, [A, B, C],
tvm.target.Target(target_hexagon, host=target_hexagon),
name="add")
mod = hexagon_session.load_module(func)
A_data = tvm.nd.array(np.array([2, 3], dtype=dtype),
device=hexagon_session.device)
assert (A_data.numpy() == np.array([2, 3])).all()
B_data = tvm.nd.array(np.array([4], dtype=dtype),
device=hexagon_session.device)
assert (B_data.numpy() == np.array([4])).all()
C_data = tvm.nd.array(np.array([0, 0], dtype=dtype),
device=hexagon_session.device)
assert (C_data.numpy() == np.array([0, 0])).all()
mod["add"](A_data, B_data, C_data)
assert (C_data.numpy() == np.array([6, 7])).all()
def test_add_vtcm(hexagon_session: Session):
dtype = "int8"
A = tvm.te.placeholder((2, ), dtype=dtype)
B = tvm.te.placeholder((1, ), dtype=dtype)
C = tvm.te.compute(A.shape, lambda i: A[i] + B[0], name="C")
sched = tvm.te.create_schedule(C.op)
target_hexagon = tvm.target.hexagon("v68", link_params=True)
func = tvm.build(sched, [A, B, C],
tvm.target.Target(target_hexagon, host=target_hexagon),
name="add")
mod = hexagon_session.load_module(func)
A_data = tvm.nd.empty(A.shape, A.dtype, hexagon_session.device,
"global.vtcm")
A_data.copyfrom(np.array([2, 3]))
B_data = tvm.nd.empty(B.shape, B.dtype, hexagon_session.device,
"global.vtcm")
B_data.copyfrom(np.array([4]))
C_data = tvm.nd.empty(C.shape, C.dtype, hexagon_session.device,
"global.vtcm")
C_data.copyfrom(np.array([0, 0]))
mod["add"](A_data, B_data, C_data)
result = C_data.numpy()
assert (result == np.array([6, 7])).all()
def verify(hexagon_session: Session, schedule, x_tensor, y_tensor, z_tensor,
size):
"""Verify correctness with reference from numpy"""
print(tvm.lower(schedule, [x_tensor, y_tensor, z_tensor]))
target_hexagon = tvm.target.hexagon("v68", link_params=True)
func = tvm.build(
schedule,
[x_tensor, y_tensor, z_tensor],
tvm.target.Target(target_hexagon, host=target_hexagon),
name="dmacpy",
)
mod = hexagon_session.load_module(func)
x_array = tvm.nd.array(
np.random.randint(low=-128, high=127, size=size, dtype=x_tensor.dtype),
device=hexagon_session.device,
)
y_array = tvm.nd.array(
np.random.randint(low=-128, high=127, size=size, dtype=y_tensor.dtype),
device=hexagon_session.device,
)
z_array = tvm.nd.array(
np.random.randint(low=-128, high=127, size=size, dtype=z_tensor.dtype),
device=hexagon_session.device,
)
mod["dmacpy"](x_array, y_array, z_array)
ref = x_array.numpy() + y_array.numpy()
np.testing.assert_equal(z_array.numpy(), ref)
def test_add_vtcm(hexagon_session: Session):
"""Test add on VTCM"""
dtype = "int8"
placeholder_a = tvm.te.placeholder((2, ), dtype=dtype)
placeholder_b = tvm.te.placeholder((1, ), dtype=dtype)
compute_c = tvm.te.compute(placeholder_a.shape,
lambda i: placeholder_a[i] + placeholder_b[0],
name="C")
sched = tvm.te.create_schedule(compute_c.op)
target_hexagon = tvm.target.hexagon("v68", link_params=True)
func = tvm.build(
sched,
[placeholder_a, placeholder_b, compute_c],
tvm.target.Target(target_hexagon, host=target_hexagon),
name="add",
)
mod = hexagon_session.load_module(func)
a_data = tvm.nd.empty(placeholder_a.shape, placeholder_a.dtype,
hexagon_session.device, "global.vtcm")
a_data.copyfrom(np.array([2, 3]))
b_data = tvm.nd.empty(placeholder_b.shape, placeholder_b.dtype,
hexagon_session.device, "global.vtcm")
b_data.copyfrom(np.array([4]))
c_data = tvm.nd.empty(compute_c.shape, compute_c.dtype,
hexagon_session.device, "global.vtcm")
c_data.copyfrom(np.array([0, 0]))
mod["add"](a_data, b_data, c_data)
result = c_data.numpy()
assert (result == np.array([6, 7])).all()
def test_add(hexagon_session: Session):
"""Test simple add"""
dtype = "int8"
placeholder_a = tvm.te.placeholder((2, ), dtype=dtype)
placeholder_b = tvm.te.placeholder((1, ), dtype=dtype)
compute_c = tvm.te.compute(placeholder_a.shape,
lambda i: placeholder_a[i] + placeholder_b[0],
name="C")
sched = tvm.te.create_schedule(compute_c.op)
target_hexagon = tvm.target.hexagon("v68", link_params=True)
func = tvm.build(
sched,
[placeholder_a, placeholder_b, compute_c],
tvm.target.Target(target_hexagon, host=target_hexagon),
name="add",
)
mod = hexagon_session.load_module(func)
a_data = tvm.nd.array(np.array([2, 3], dtype=dtype),
device=hexagon_session.device)
assert (a_data.numpy() == np.array([2, 3])).all()
b_data = tvm.nd.array(np.array([4], dtype=dtype),
device=hexagon_session.device)
assert (b_data.numpy() == np.array([4])).all()
c_data = tvm.nd.array(np.array([0, 0], dtype=dtype),
device=hexagon_session.device)
assert (c_data.numpy() == np.array([0, 0])).all()
mod["add"](a_data, b_data, c_data)
assert (c_data.numpy() == np.array([6, 7])).all()
def verify(hexagon_session: Session, s, x, y, z, size):
print(tvm.lower(s, [x, y, z]))
target_hexagon = tvm.target.hexagon("v68", link_params=True)
func = tvm.build(s, [x, y, z],
tvm.target.Target(target_hexagon, host=target_hexagon),
name="dmacpy")
mod = hexagon_session.load_module(func)
xt = tvm.nd.array(
np.random.randint(low=-128, high=127, size=size, dtype=x.dtype),
device=hexagon_session.device,
)
yt = tvm.nd.array(
np.random.randint(low=-128, high=127, size=size, dtype=y.dtype),
device=hexagon_session.device,
)
zt = tvm.nd.array(
np.random.randint(low=-128, high=127, size=size, dtype=z.dtype),
device=hexagon_session.device,
)
mod["dmacpy"](xt, yt, zt)
ref = xt.numpy() + yt.numpy()
np.testing.assert_equal(zt.numpy(), ref)
def test_graph_executor(hexagon_session: Session):
"""Test graph executor"""
dtype = "float32"
data = relay.var("data", relay.TensorType((1, 64, 64, 3), dtype))
weight = relay.var("weight", relay.TensorType((5, 5, 3, 8), dtype))
conv2d_op = relay.nn.conv2d(
data,
weight,
padding=(2, 2),
kernel_size=(5, 5),
data_layout="NHWC",
kernel_layout="HWIO",
out_dtype="float32",
)
f = relay.Function([data, weight], conv2d_op)
relay_mod = tvm.IRModule.from_expr(f)
relay_mod = relay.transform.InferType()(relay_mod)
target_hexagon = tvm.target.hexagon("v68")
runtime = Runtime("cpp")
executor = Executor("graph")
weight_in = np.random.rand(5, 5, 3, 8).astype(dtype=dtype)
data_in = np.random.rand(1, 64, 64, 3).astype(dtype=dtype)
params = {"weight": weight_in}
inputs = {"data": data_in}
with tvm.transform.PassContext(opt_level=3):
lowered = tvm.relay.build(
relay_mod,
tvm.target.Target(target_hexagon, host=target_hexagon),
runtime=runtime,
executor=executor,
)
graph_mod = hexagon_session.get_executor_from_factory(lowered)
graph_mod.set_input(**params)
graph_mod.run(**inputs)
hexagon_output = graph_mod.get_output(0).numpy()
target_llvm = tvm.target.Target("llvm")
with tvm.transform.PassContext(opt_level=3):
llvm_lowered = tvm.relay.build(
relay_mod,
tvm.target.Target(target_llvm, host=target_llvm),
runtime=runtime,
executor=executor,
)
llvm_graph_mod = tvm.contrib.graph_executor.GraphModule(
llvm_lowered["default"](tvm.cpu(0)))
llvm_graph_mod.set_input(**params)
llvm_graph_mod.run(**inputs)
expected_output = llvm_graph_mod.get_output(0).numpy()
tvm.testing.assert_allclose(hexagon_output,
expected_output,
rtol=1e-4,
atol=1e-5)
def test_elemwise_sum_parallel(hexagon_session: Session):
target_hexagon = tvm.target.hexagon("v68", link_params=True)
func = tvm.build(
ElemwiseSumIRModule, target=tvm.target.Target(target_hexagon, host=target_hexagon)
)
mod = hexagon_session.load_module(func)
(a, b, c, n) = generate_add_test_data(hexagon_session)
mod["elemwise_sum_parallel"](a, b, c, n)
tvm.testing.assert_allclose(c.numpy(), a.numpy() + b.numpy())
def test_mobilenet_aot(hexagon_session: Session, aot_host_target, aot_target,
enable_usmp):
if hexagon_session._launcher._serial_number == "simulator":
pytest.skip(msg="Skip on simulator due to long runtime.")
dtype = "float32"
onnx_model = get_mobilenet()
data_in = np.random.rand(1, 3, 224, 224).astype(dtype=dtype)
input_name = "input"
shape_dict = {input_name: data_in.shape}
relay_mod, params = relay.frontend.from_onnx(onnx_model,
shape_dict,
freeze_params=True)
inputs = {input_name: data_in}
target_llvm = tvm.target.Target("llvm")
config = {"tir.usmp.enable": enable_usmp}
with tvm.transform.PassContext(opt_level=3, config=config):
hexagon_lowered = tvm.relay.build(
relay_mod,
tvm.target.Target(aot_target, host=aot_host_target),
runtime=Runtime("cpp"),
executor=Executor("aot", {
"unpacked-api": False,
"interface-api": "packed"
}),
params=params,
)
aot_mod = hexagon_session.get_executor_from_factory(hexagon_lowered)
aot_mod.set_input(**inputs)
aot_mod.run()
hexagon_output = aot_mod.get_output(0).numpy()
with tvm.transform.PassContext(opt_level=3):
llvm_lowered = tvm.relay.build(
relay_mod,
tvm.target.Target(target_llvm, host=target_llvm),
runtime=Runtime("cpp"),
executor=Executor("graph", {"link-params": True}),
params=params,
)
llvm_graph_mod = tvm.contrib.graph_executor.GraphModule(
llvm_lowered["default"](tvm.cpu(0)))
llvm_graph_mod.set_input(**inputs)
llvm_graph_mod.run()
expected_output = llvm_graph_mod.get_output(0).numpy()
tvm.testing.assert_allclose(hexagon_output,
expected_output,
rtol=1e-4,
atol=1e-5)
def test_speedup(hexagon_session: Session, capsys):
target_hexagon = tvm.target.hexagon("v68", link_params=True)
func = tvm.build(
ElemwiseSumIRModule, target=tvm.target.Target(target_hexagon, host=target_hexagon)
)
mod = hexagon_session.load_module(func)
args = generate_add_test_data(hexagon_session)
parallel_mean = benchmark_func(mod, "elemwise_sum_parallel", args, hexagon_session)
serial_mean = benchmark_func(mod, "elemwise_sum_serial", args, hexagon_session)
with capsys.disabled():
print("... speedup of {:.2f}".format(serial_mean / parallel_mean), end=" ")
def test_conv2d_nhwc(
self,
hexagon_session: Session,
ref_data,
batch,
in_channel,
in_size,
num_filter,
kernel,
dtype,
stride,
padding,
dilation,
):
target_hexagon = tvm.target.hexagon("v68")
a_np, w_np, b_np = ref_data
A = te.placeholder(a_np.shape, name="A", dtype=dtype)
W = te.placeholder(w_np.shape, name="W", dtype=dtype)
with tvm.target.Target(target_hexagon):
fcompute = topi.nn.conv2d_nhwc
fschedule = topi.hexagon.schedule_conv2d_nhwc
B = fcompute(A, W, stride, padding, dilation, dtype)
s = fschedule([B])
func_name = "conv2d_{}_{}_{}_{}_{}_{}_{}_{}_{}".format(
dtype,
batch,
in_channel,
in_size,
num_filter,
kernel,
stride,
padding,
dilation,
)
func = tvm.build(s, [A, W, B],
tvm.target.Target(target_hexagon,
host=target_hexagon),
name=func_name)
mod = hexagon_session.load_module(func)
dev = hexagon_session.device
a = tvm.nd.array(a_np, dev)
w = tvm.nd.array(w_np, dev)
b = tvm.nd.array(np.zeros(get_const_tuple(B.shape), dtype=B.dtype),
dev)
mod[func_name](a, w, b)
tvm.testing.assert_allclose(b.numpy(), b_np, rtol=1e-5)
def test_mobilenet(hexagon_session: Session):
dtype = "float32"
onnx_model = get_mobilenet()
target_hexagon = tvm.target.hexagon("v68")
target_llvm = tvm.target.Target("llvm")
runtime = Runtime("cpp")
executor = Executor("graph", {"link-params": True})
data_in = np.random.rand(1, 3, 224, 224).astype(dtype=dtype)
input_name = "input"
shape_dict = {input_name: data_in.shape}
relay_mod, params = relay.frontend.from_onnx(onnx_model,
shape_dict,
freeze_params=True)
inputs = {input_name: data_in}
with tvm.transform.PassContext(opt_level=3):
hexagon_lowered = tvm.relay.build(
relay_mod,
tvm.target.Target(target_hexagon, host=target_hexagon),
runtime=runtime,
executor=executor,
params=params,
)
llvm_lowered = tvm.relay.build(
relay_mod,
tvm.target.Target(target_llvm, host=target_llvm),
runtime=runtime,
executor=executor,
params=params,
)
graph_mod = hexagon_session.get_executor_from_factory(hexagon_lowered)
graph_mod.set_input(**inputs)
graph_mod.run()
hexagon_output = graph_mod.get_output(0).numpy()
llvm_graph_mod = tvm.contrib.graph_executor.GraphModule(
llvm_lowered["default"](tvm.cpu(0)))
llvm_graph_mod.set_input(**inputs)
llvm_graph_mod.run()
expected_output = llvm_graph_mod.get_output(0).numpy()
tvm.testing.assert_allclose(hexagon_output,
expected_output,
rtol=1e-4,
atol=1e-5)
def test_dense(
hexagon_session: Session,
batch_size,
in_dim,
out_dim,
use_bias,
in_dtype,
out_dtype,
dense_ref_data,
):
if in_dtype == "float16":
pytest.xfail("float16 is not supported.")
if "int" in in_dtype:
tol = {"atol": 0, "rtol": 0}
elif in_dtype == "float32":
tol = {"rtol": 1e-5, "atol": 1e-5}
A = te.placeholder((batch_size, in_dim), name="A", dtype=in_dtype)
B = te.placeholder((out_dim, in_dim), name="B", dtype=in_dtype)
C = te.placeholder((out_dim, ), name="C", dtype=out_dtype)
a_np, b_np, c_np, d_np = dense_ref_data
fcompute = topi.nn.dense
fschedule = topi.hexagon.schedule_dense
target_hexagon = tvm.target.hexagon("v68")
with tvm.target.Target(target_hexagon):
D = fcompute(A, B, C if use_bias else None, out_dtype)
D = topi.nn.relu(D)
s = fschedule([D])
func = tvm.build(s, [A, B, C, D],
tvm.target.Target(target_hexagon, host=target_hexagon),
name="dense")
mod = hexagon_session.load_module(func)
dev = hexagon_session.device
a = tvm.nd.array(a_np, dev)
b = tvm.nd.array(b_np, dev)
c = tvm.nd.array(c_np, dev)
d = tvm.nd.array(np.zeros(get_const_tuple(D.shape), dtype=out_dtype), dev)
mod["dense"](a, b, c, d)
tvm.testing.assert_allclose(d.numpy(), d_np, **tol)
def test_batch_matmul_int8(self, hexagon_session: Session, x_batch,
y_batch, M, N, K):
dtype = "int8"
out_dtype = "int8"
assert x_batch == y_batch or x_batch == 1 or y_batch == 1
x = te.placeholder((x_batch, M, K), name="x", dtype=dtype)
y = te.placeholder((y_batch, N, K), name="y", dtype=dtype)
def get_ref_data():
a_np = np.random.randint(low=-128, high=127,
size=(x_batch, M, K)).astype(dtype)
b_np = np.random.randint(low=-128, high=127,
size=(y_batch, N, K)).astype(dtype)
c_np = tvm.topi.testing.batch_matmul(a_np,
b_np,
out_dtype=out_dtype)
return (a_np, b_np, c_np)
# get the test data
a_np, b_np, c_np = get_ref_data()
target_hexagon = tvm.target.hexagon("v68")
with tvm.target.Target(target_hexagon):
fcompute = topi.nn.batch_matmul
fschedule = topi.hexagon.schedule_batch_matmul
out = fcompute(x, y)
s = fschedule([out])
func = tvm.build(
s,
[x, y, out],
tvm.target.Target(target_hexagon, host=target_hexagon),
name="batch_matmul_int8",
)
mod = hexagon_session.load_module(func)
dev = hexagon_session.device
a = tvm.nd.array(a_np, dev)
b = tvm.nd.array(b_np, dev)
c = tvm.nd.array(np.zeros(get_const_tuple(out.shape), dtype=out_dtype),
dev)
mod["batch_matmul_int8"](a, b, c)
tvm.testing.assert_allclose(c.numpy(), c_np, rtol=1e-5)
def test_batch_matmul(self, hexagon_session: Session, x_batch, y_batch, M,
N, K, dtype):
if dtype == "float16":
pytest.xfail("float16 is not supported.")
x = te.placeholder((x_batch, M, K), name="x")
y = te.placeholder((y_batch, N, K), name="y")
def get_ref_data():
a_np = np.random.uniform(size=(x_batch, M, K)).astype(dtype)
b_np = np.random.uniform(size=(y_batch, N, K)).astype(dtype)
c_np = tvm.topi.testing.batch_matmul(a_np, b_np)
return (a_np, b_np, c_np)
# get the test data
a_np, b_np, c_np = get_ref_data()
target_hexagon = tvm.target.hexagon("v68")
with tvm.target.Target(target_hexagon):
fcompute = topi.nn.batch_matmul
fschedule = topi.hexagon.schedule_batch_matmul
out = fcompute(x, y)
s = fschedule([out])
out_shape = out.shape
func = tvm.build(
s,
[x, y, out],
tvm.target.Target(target_hexagon, host=target_hexagon),
name="batch_matmul",
)
mod = hexagon_session.load_module(func)
dev = hexagon_session.device
a = tvm.nd.array(a_np, dev)
b = tvm.nd.array(b_np, dev)
c = tvm.nd.array(np.zeros(get_const_tuple(out_shape), dtype=dtype),
dev)
mod["batch_matmul"](a, b, c)
tvm.testing.assert_allclose(c.numpy(), c_np, rtol=1e-5)
def test_softmax(hexagon_session: Session, shape, dtype, softmax_operation):
if dtype == "float16":
pytest.xfail("float16 is not supported.")
A = te.placeholder(shape, dtype=dtype, name="A")
topi_op = configs[softmax_operation]["topi"]
B = topi_op(A, axis=1)
def get_ref_data(shape):
ref_func = tvm.topi.testing.softmax_python
a_np = np.random.uniform(size=shape).astype(dtype)
if len(shape) == 2:
b_np = ref_func(a_np)
elif len(shape) == 4:
_, c, h, w = a_np.shape
a_np_2d = a_np.transpose(0, 2, 3, 1).reshape(h * w, c)
b_np_2d = tvm.topi.testing.softmax_python(a_np_2d)
b_np = b_np_2d.reshape(1, h, w, c).transpose(0, 3, 1, 2)
return a_np, b_np
# get the test data
a_np, b_np = get_ref_data(shape)
target_hexagon = tvm.target.hexagon("v68")
with tvm.target.Target(target_hexagon):
fschedule = topi.hexagon.schedule_softmax
s = fschedule(B)
func = tvm.build(s, [A, B],
tvm.target.Target(target_hexagon, host=target_hexagon),
name="softmax")
mod = hexagon_session.load_module(func)
dev = hexagon_session.device
a = tvm.nd.array(a_np, dev)
b = tvm.nd.array(np.zeros(get_const_tuple(B.shape), dtype=B.dtype), dev)
mod["softmax"](a, b)
tvm.testing.assert_allclose(b.numpy(), b_np, rtol=1e-5)
def test_conv2d_nchw(
self,
hexagon_session: Session,
batch,
in_channel,
in_size,
num_filter,
kernel,
stride,
padding,
dtype,
ref_data,
dilation,
add_bias,
apply_relu,
):
target_hexagon = tvm.target.hexagon("v68")
pad_top, pad_left, pad_bottom, pad_right = get_pad_tuple(
padding, (kernel, kernel))
padding_sum = pad_top + pad_left + pad_bottom + pad_right
a_np, w_np, b_np, c_np = ref_data
A = te.placeholder(a_np.shape, name="A", dtype=dtype)
W = te.placeholder(w_np.shape, name="W", dtype=dtype)
bias = te.placeholder(b_np.shape, name="bias", dtype=dtype)
if "int" in dtype:
tol = {"atol": 0, "rtol": 0}
elif dtype == "float32":
tol = {"rtol": 1e-4, "atol": 2e-4}
elif dtype == "float16":
# A summation in float16 with a single accumulator very
# quickly runs into large rounding errors. At some point,
# this tolerance should be schedule-dependent for to avoid
# false negatives.
num_values_summed = in_channel * kernel * kernel
gap_size = np.nextafter(c_np.max(), np.inf,
dtype=c_np.dtype) - c_np.max()
tol = {"rtol": 1e-3, "atol": num_values_summed * gap_size / 2}
with tvm.target.Target(target_hexagon):
fcompute = topi.nn.conv2d_nchw
fschedule = topi.hexagon.schedule_conv2d_nchw
C = fcompute(A, W, (stride, stride), padding, (dilation, dilation),
dtype)
if add_bias:
C = topi.add(C, bias)
if apply_relu:
C = topi.nn.relu(C)
s = fschedule([C])
func_name = "conv2d_{}_{}_{}_{}_{}_{}_{}_{}_{}".format(
dtype,
batch,
in_channel,
in_size,
num_filter,
kernel,
stride,
padding_sum,
dilation,
)
func = tvm.build(
s,
[A, W, bias, C],
tvm.target.Target(target_hexagon, host=target_hexagon),
name=func_name,
)
mod = hexagon_session.load_module(func)
dev = hexagon_session.device
a = tvm.nd.array(a_np, dev)
w = tvm.nd.array(w_np, dev)
b = tvm.nd.array(b_np, dev)
c = tvm.nd.array(np.zeros(get_const_tuple(C.shape), dtype=C.dtype),
dev)
mod[func_name](a, w, b, c)
tvm.testing.assert_allclose(c.numpy(), c_np, **tol)
def test_conv2d(hexagon_session: Session, aot_host_target, aot_target,
usmp_enabled):
"""Try conv2d on AOT target with usmp_enabled and check for TVMBackendAllocWorkspace calls"""
dtype = "float32"
input_shape = (1, 8, 8, 3)
w1_shape = (5, 5, 3, 1)
w2_shape = (5, 5, 1, 3)
data = relay.var("data", relay.TensorType(input_shape, dtype))
weight1 = relay.var("weight1", relay.TensorType(w1_shape, dtype))
weight2 = relay.var("weight2", relay.TensorType(w2_shape, dtype))
outpu1 = relay.nn.conv2d(
data,
weight1,
padding=(2, 2),
kernel_size=(5, 5),
data_layout="NHWC",
kernel_layout="HWIO",
out_dtype="float32",
)
output2 = relay.nn.conv2d(
outpu1,
weight2,
padding=(2, 2),
kernel_size=(5, 5),
data_layout="NHWC",
kernel_layout="HWIO",
out_dtype="float32",
)
f = relay.Function([data, weight1, weight2], output2)
relay_mod = tvm.IRModule.from_expr(f)
relay_mod = relay.transform.InferType()(relay_mod)
weight1_data = np.random.rand(w1_shape[0], w1_shape[1], w1_shape[2],
w1_shape[3]).astype(dtype=dtype)
weight2_data = np.random.rand(w2_shape[0], w2_shape[1], w2_shape[2],
w2_shape[3]).astype(dtype=dtype)
input_data = np.random.rand(input_shape[0], input_shape[1], input_shape[2],
input_shape[3]).astype(dtype=dtype)
params = {"weight1": weight1_data, "weight2": weight2_data}
inputs = {"data": input_data}
with tvm.transform.PassContext(opt_level=3,
config={"tir.usmp.enable": usmp_enabled}):
lowered = tvm.relay.build(
relay_mod,
params=params,
target=tvm.target.Target(aot_target, host=aot_host_target),
runtime=Runtime("cpp"),
executor=Executor("aot", {
"unpacked-api": False,
"interface-api": "packed"
}),
)
assert is_tvm_backendallocworkspace_calls(lowered.lib) != usmp_enabled
aot_mod = hexagon_session.get_executor_from_factory(lowered)
aot_mod.set_input(**inputs)
aot_mod.run()
hexagon_output = aot_mod.get_output(0).numpy()
target_llvm = tvm.target.Target("llvm")
with tvm.transform.PassContext(opt_level=3):
llvm_lowered = tvm.relay.build(
relay_mod,
tvm.target.Target(target_llvm, host=target_llvm),
runtime=Runtime("cpp"),
executor=Executor("graph"),
)
llvm_graph_mod = tvm.contrib.graph_executor.GraphModule(
llvm_lowered["default"](tvm.cpu(0)))
llvm_graph_mod.set_input(**params)
llvm_graph_mod.run(**inputs)
expected_output = llvm_graph_mod.get_output(0).numpy()
tvm.testing.assert_allclose(hexagon_output,
expected_output,
rtol=1e-4,
atol=1e-5)
def test_conv2d(
self,
hexagon_session: Session,
batch,
in_channel,
in_size,
num_filter,
stride,
padding,
output_padding,
random_seed,
):
target_hexagon = tvm.target.hexagon("v68")
in_height, in_width = in_size
kernel_height, kernel_width = (1, 1)
stride_height, stride_width = stride
pad_top, pad_left, pad_bottom, pad_right = padding
A = te.placeholder((batch, in_channel, in_height, in_width), name="A")
W = te.placeholder(
(in_channel, num_filter, kernel_height, kernel_width), name="W")
a_shape = get_const_tuple(A.shape)
w_shape = get_const_tuple(W.shape)
dtype = A.dtype
def get_ref_data():
np.random.seed(random_seed)
a_np = np.random.uniform(size=a_shape).astype(dtype)
w_np = np.random.uniform(size=w_shape).astype(dtype)
b_np = tvm.topi.testing.conv2d_transpose_nchw_python(
a_np, w_np, stride, padding, output_padding)
c_np = np.maximum(b_np, 0)
return a_np, w_np, b_np, c_np
a_np, w_np, b_np, c_np = get_ref_data()
fcompute_args = (
A,
W,
[stride_height, stride_width],
[pad_top, pad_left, pad_bottom, pad_right],
A.dtype,
output_padding,
)
with tvm.target.Target(target_hexagon):
fcompute = topi.nn.conv2d_transpose_nchw
fschedule = topi.hexagon.schedule_conv2d_transpose_nchw
B = fcompute(*fcompute_args)
C = topi.nn.relu(B)
s1 = fschedule([B])
s2 = fschedule([C])
dev = hexagon_session.device
a = tvm.nd.array(a_np, dev)
w = tvm.nd.array(w_np, dev)
b = tvm.nd.array(np.zeros(get_const_tuple(B.shape), dtype=B.dtype),
dev)
c = tvm.nd.array(np.zeros(get_const_tuple(C.shape), dtype=C.dtype),
dev)
func1 = tvm.build(
s1, [A, W, B],
tvm.target.Target(target_hexagon, host=target_hexagon))
func2 = tvm.build(
s2, [A, W, C],
tvm.target.Target(target_hexagon, host=target_hexagon))
mod1 = hexagon_session.load_module(func1)
mod2 = hexagon_session.load_module(func2)
mod1(a, w, b)
mod2(a, w, c)
tvm.testing.assert_allclose(b.numpy(), b_np, rtol=1e-5)
tvm.testing.assert_allclose(c.numpy(), c_np, rtol=1e-5)
def test_graph_executor_multiple_conv2d(hexagon_session: Session):
dtype = "float32"
input_shape = (1, 8, 8, 3)
w1_shape = (5, 5, 3, 1)
w2_shape = (5, 5, 1, 3)
data = relay.var("data", relay.TensorType(input_shape, dtype))
weight1 = relay.var("weight1", relay.TensorType(w1_shape, dtype))
weight2 = relay.var("weight2", relay.TensorType(w2_shape, dtype))
y1 = relay.nn.conv2d(
data,
weight1,
padding=(2, 2),
kernel_size=(5, 5),
data_layout="NHWC",
kernel_layout="HWIO",
out_dtype="float32",
)
y2 = relay.nn.conv2d(
y1,
weight2,
padding=(2, 2),
kernel_size=(5, 5),
data_layout="NHWC",
kernel_layout="HWIO",
out_dtype="float32",
)
f = relay.Function([data, weight1, weight2], y2)
relay_mod = tvm.IRModule.from_expr(f)
relay_mod = relay.transform.InferType()(relay_mod)
target_hexagon = tvm.target.hexagon("v68")
runtime = Runtime("cpp")
executor = Executor("graph")
with tvm.transform.PassContext(opt_level=3):
lowered = tvm.relay.build(
relay_mod,
tvm.target.Target(target_hexagon, host=target_hexagon),
runtime=runtime,
executor=executor,
)
weight1_data = np.random.rand(w1_shape[0], w1_shape[1], w1_shape[2],
w1_shape[3]).astype(dtype=dtype)
weight2_data = np.random.rand(w2_shape[0], w2_shape[1], w2_shape[2],
w2_shape[3]).astype(dtype=dtype)
input_data = np.random.rand(input_shape[0], input_shape[1], input_shape[2],
input_shape[3]).astype(dtype=dtype)
params = {"weight1": weight1_data, "weight2": weight2_data}
inputs = {"data": input_data}
graph_mod = hexagon_session.get_executor_from_factory(lowered)
graph_mod.set_input(**params)
graph_mod.run(**inputs)
hexagon_output = graph_mod.get_output(0).numpy()
target_llvm = tvm.target.Target("llvm")
with tvm.transform.PassContext(opt_level=3):
llvm_lowered = tvm.relay.build(
relay_mod,
tvm.target.Target(target_llvm, host=target_llvm),
runtime=runtime,
executor=executor,
)
llvm_graph_mod = tvm.contrib.graph_executor.GraphModule(
llvm_lowered["default"](tvm.cpu(0)))
llvm_graph_mod.set_input(**params)
llvm_graph_mod.run(**inputs)
expected_output = llvm_graph_mod.get_output(0).numpy()
tvm.testing.assert_allclose(hexagon_output,
expected_output,
rtol=1e-4,
atol=1e-5)
def test_reduce_map(hexagon_session: Session, ref_data, in_shape, axis,
keepdims, reduce_type, dtype):
in_npy, in_npy_map, out_npy = ref_data
# Build the logic and compile the function
A = te.placeholder(shape=in_shape, name="A", dtype=dtype)
A1 = topi.sqrt(topi.exp(A))
out_dtype = dtype
if reduce_type == "sum":
B = topi.sum(A1, axis=axis, keepdims=keepdims)
elif reduce_type == "all":
B = topi.all(A, axis=axis, keepdims=keepdims)
elif reduce_type == "any":
B = topi.any(A, axis=axis, keepdims=keepdims)
elif reduce_type == "max":
B = topi.max(A1, axis=axis, keepdims=keepdims)
elif reduce_type == "min":
B = topi.min(A1, axis=axis, keepdims=keepdims)
elif reduce_type == "argmax":
B = topi.argmax(A1, axis=axis, keepdims=keepdims)
out_dtype = "int32"
elif reduce_type == "argmin":
B = topi.argmin(A1, axis=axis, keepdims=keepdims)
out_dtype = "int32"
else:
raise NotImplementedError
target_hexagon = tvm.target.hexagon("v68")
with tvm.target.Target(target_hexagon):
fschedule = topi.hexagon.schedule_reduce
s = fschedule(B)
func = tvm.build(s, [A, B],
tvm.target.Target(target_hexagon, host=target_hexagon),
name=reduce_type)
mod = hexagon_session.load_module(func)
dev = hexagon_session.device
data_tvm = tvm.nd.array(in_npy, device=dev)
out_tvm = tvm.nd.empty(shape=out_npy.shape, device=dev, dtype=out_dtype)
mod[reduce_type](data_tvm, out_tvm)
if reduce_type == "argmax" or reduce_type == "argmin":
out_tvm_indices = out_tvm.numpy()
if keepdims:
out_tvm_indices = np.take(out_tvm_indices, indices=0, axis=axis)
if axis is None:
out_tvm_val = in_npy_map.ravel()[out_tvm_indices]
else:
other_indices = tuple(
np.indices(in_shape[0:axis] + in_shape[(axis + 1):]))
sel_indices = other_indices[0:axis] + (
out_tvm_indices, ) + other_indices[axis:]
out_tvm_val = in_npy_map[sel_indices]
if reduce_type == "argmax":
tvm.testing.assert_allclose(out_tvm_val, in_npy_map.max(axis=axis),
1e-3, 1e-3)
elif reduce_type == "argmin":
tvm.testing.assert_allclose(out_tvm_val, in_npy_map.min(axis=axis),
1e-3, 1e-3)
else:
tvm.testing.assert_allclose(out_tvm.numpy(), out_npy, 1e-3, 1e-3)
def test_aot_executor(hexagon_session: Session, aot_host_target, aot_target):
dtype = "float32"
input_shape = (1, 128, 128, 3)
w_shape = (5, 5, 3, 8)
data = relay.var("data", relay.TensorType(input_shape, dtype))
weight = relay.var("weight", relay.TensorType(w_shape, dtype))
y = relay.nn.conv2d(
data,
weight,
padding=(2, 2),
kernel_size=(5, 5),
data_layout="NHWC",
kernel_layout="HWIO",
out_dtype="float32",
)
f = relay.Function([data, weight], y)
relay_mod = tvm.IRModule.from_expr(f)
relay_mod = relay.transform.InferType()(relay_mod)
weight_data = np.random.rand(w_shape[0], w_shape[1], w_shape[2],
w_shape[3]).astype(dtype=dtype)
input_data = np.random.rand(input_shape[0], input_shape[1], input_shape[2],
input_shape[3]).astype(dtype=dtype)
params = {"weight": weight_data}
inputs = {"data": input_data}
with tvm.transform.PassContext(opt_level=3):
lowered = tvm.relay.build(
relay_mod,
params=params,
target=tvm.target.Target(aot_target, host=aot_host_target),
runtime=Runtime("cpp"),
executor=Executor("aot", {
"unpacked-api": False,
"interface-api": "packed"
}),
)
aot_mod = hexagon_session.get_executor_from_factory(lowered)
aot_mod.set_input(**inputs)
aot_mod.run()
hexagon_output = aot_mod.get_output(0).numpy()
target_llvm = tvm.target.Target("llvm")
with tvm.transform.PassContext(opt_level=3):
llvm_lowered = tvm.relay.build(
relay_mod,
tvm.target.Target(target_llvm, host=target_llvm),
runtime=Runtime("cpp"),
executor=Executor("graph"),
)
llvm_graph_mod = tvm.contrib.graph_executor.GraphModule(
llvm_lowered["default"](tvm.cpu(0)))
llvm_graph_mod.set_input(**params)
llvm_graph_mod.run(**inputs)
expected_output = llvm_graph_mod.get_output(0).numpy()
tvm.testing.assert_allclose(hexagon_output,
expected_output,
rtol=1e-4,
atol=1e-5)
def test_maxpool2d_nhwc(
self,
N,
H,
W,
C,
DTYPE,
KERNEL,
STRIDE,
DILATION,
PADDING,
IO_TENSOR_MEM_SCOPE,
hexagon_session: Session,
):
keys_dict = {
"basic_kernel": "max_pool2d",
"sched_type": 1,
"input_shape_4d": [N, H, W, C],
"block_shape": [8, 8, 32],
"DTYPE": DTYPE,
"KERNEL": KERNEL,
"STRIDE": STRIDE,
"DILATION": DILATION,
"PADDING": PADDING,
"IO_TENSOR_MEM_SCOPE": IO_TENSOR_MEM_SCOPE,
}
desc = bu.get_benchmark_decription(keys_dict)
# Create the host-side directory for this benchmark run's files / logs...
host_files_dir_name = bu.get_benchmark_id(keys_dict)
host_files_dir_path = os.path.join(self.working_dir, host_files_dir_name)
os.mkdir(host_files_dir_path)
keys_dict["host_files_dir_path"] = host_files_dir_path
log_file_path = os.path.join(host_files_dir_path, "out.txt")
with open(log_file_path, "w") as log_file:
print(f"CONFIGURATION: {desc}")
log_file.write(f"CONFIGURATION: {desc}\n")
try:
input_tensor_shape_4d = [N, H, W, C]
input_tensor_shape_7d = _int8_nhwc_8h8w32c_shape(N, H, W, C)
data = te.placeholder(tuple(input_tensor_shape_4d), dtype=DTYPE)
output = topi.nn.pool2d(
data, KERNEL, STRIDE, DILATION, PADDING, "max", layout="NHWC"
)
primfunc = te.create_prim_func([data, output])
sch = tir.Schedule(primfunc, debug_mask="all")
sch.transform_layout(
block="tensor", buffer="placeholder", index_map=_int8_nhwc_8h8w32c_map
)
target_hexagon = tvm.target.hexagon("v69", link_params=True)
# func = tvm.build(sch.mod, target=tvm.target.Target(target_hexagon, host=target_hexagon))
built_module = tvm.build(
sch.mod, target=tvm.target.Target(target_hexagon, host=target_hexagon)
)
# Save a local copy of the Hexagon object code (in the form of a .so file)
# to allow post-mortem inspection.
host_dso_binary_path = os.path.join(host_files_dir_path, "test_binary.so")
built_module.save(host_dso_binary_path)
print(f"SAVED BINARY TO HOST PATH: {host_dso_binary_path}")
hexagon_mod = hexagon_session.load_module(built_module)
# Generate the input tensor's data.
# Note that we'll eventually need it in two different layouts:
# (1) NHWC as an argument to testing.poolnd_python.
# (2) NHWC_8h8w32c for as an argument to our Hexagon primfunc.
# a_numpy_4d = np.random.randint(low=-128, high=127, size=input_tensor_shape_4d, dtype=np.int8)
a_numpy_4d = _create_test_input(input_tensor_shape_4d, DTYPE)
ref_output_4d = testing.poolnd_python(
a_numpy_4d.astype("int32"),
KERNEL,
STRIDE,
DILATION,
PADDING[0:2],
PADDING[2:],
pool_type="max",
dtype="int32",
layout="NHWC",
).astype(DTYPE)
output_tensor_shape_4d = ref_output_4d.shape
a_numpy_7d = _int8_nhwc_8h8w32c_xform_immediate(a_numpy_4d)
a_hexagon_7d = allocate_hexagon_array(
hexagon_session.device,
tensor_shape=input_tensor_shape_7d,
axis_separators=[4],
dtype=DTYPE,
mem_scope=IO_TENSOR_MEM_SCOPE,
)
c_hexagon_4d = allocate_hexagon_array(
hexagon_session.device,
tensor_shape=output_tensor_shape_4d,
axis_separators=[],
dtype=DTYPE,
mem_scope=IO_TENSOR_MEM_SCOPE,
)
a_hexagon_7d.copyfrom(a_numpy_7d)
if DTYPE == "int8":
rel_tolerance = 0
abs_tolerance = 0
else:
assert False, f"TODO: decide acceptable tolerances for DTYPE {DTYPE}"
# hexagon_mod(a_hexagon_7d, c_hexagon_4d)
# tvm.testing.assert_allclose(ref_output_4d, c_hexagon_4d.numpy(), rtol=rel_tolerance, atol=abs_tolerance)
timer = hexagon_mod.time_evaluator(
"main", hexagon_session.device, number=10, repeat=1
)
timing_result = timer(a_hexagon_7d, c_hexagon_4d)
try:
tvm.testing.assert_allclose(
ref_output_4d, c_hexagon_4d.numpy(), rtol=rel_tolerance, atol=abs_tolerance
)
except AssertionError as e:
raise bu.NumericalAccuracyException(str(e))
except bu.NumericalAccuracyException as e:
print()
print(f"FAIL: Numerical accuracy error. See log file.")
log_file.write("\n")
log_file.write(f"FAIL: {e}\n")
self.benchmark_table.record_fail(
**keys_dict, comments=f"Numerical accuracy error. See log file."
)
except bu.UnsupportedException as e:
print()
print(f"SKIP: {e}")
log_file.write("\n")
log_file.write(f"SKIP: {e}\n")
self.benchmark_table.record_skip(
**keys_dict, comments=f"Unsupported configuration: {e}"
)
self.benchmark_table.record_success(timing_result, **keys_dict)
def test_avg_pool2d_slice(
self,
stride,
kernel,
dtype,
dilation,
padding,
count_include_pad,
input_layout,
output_layout,
output_shape,
input_shape,
input_shape_padded,
input_np,
input_np_padded,
transformed_input_np_padded,
transformed_expected_output_np,
expected_output_np,
hexagon_session: Session,
):
if hexagon_session._launcher._serial_number != "simulator":
pytest.skip(msg="Due to https://github.com/apache/tvm/issues/11928")
target_hexagon = tvm.target.hexagon("v69")
A = te.placeholder(input_shape_padded, name="A", dtype=dtype)
M = sl.avg_pool2d_compute(A, output_shape, kernel, stride, dilation)
# tir schedule
tir_schedule = sl.avg_pool2d_STIR_schedule(M, A, output_layout, input_layout)
sch = tir_schedule.mod
input_axis_separator = [4]
if output_layout == "nhwc-8h2w32c2w-2d":
output_axis_separator = [4]
elif output_layout == "n11c-1024c-2d":
output_axis_separator = [4]
else:
raise RuntimeError(f"Unexpected layout '{output_layout}'")
with tvm.transform.PassContext(opt_level=3):
func = tvm.build(
sch,
[A, M],
tvm.target.Target(target_hexagon, host=target_hexagon),
name="avg_pool2d",
)
input_arr = allocate_hexagon_array(
hexagon_session.device,
data=transformed_input_np_padded,
axis_separators=input_axis_separator,
mem_scope="global.vtcm",
)
output_arr = allocate_hexagon_array(
hexagon_session.device,
transformed_expected_output_np.shape,
dtype,
axis_separators=output_axis_separator,
mem_scope="global.vtcm",
)
mod = hexagon_session.load_module(func)
mod(input_arr, output_arr)
b, h, w, c = output_shape
if output_layout == "nhwc-8h2w32c2w-2d":
output_np = output_arr.numpy().reshape([b, h // 8, w // 4, c // 32, 8, 2, 32, 2])
elif output_layout == "n11c-1024c-2d":
output_np = output_arr.numpy().reshape([b, 1, 1, c // 1024, 1024])
else:
raise RuntimeError(f"Unexpected layout '{output_layout}'")
np.testing.assert_allclose(output_np, transformed_expected_output_np, rtol=1e-3, atol=1e-3)
def test_conv2d(
self,
hexagon_session: Session,
in_dtype,
out_dtype,
layout,
input_shape,
filter_shape,
scale_shape,
shift_shape,
use_scale_shift,
apply_relu,
batch,
in_channel,
channel_multiplier,
kernel,
stride,
padding,
dilation,
ref_data,
):
target_hexagon = tvm.target.hexagon("v68")
# Transform the padding argument from 'str' to 'tuple' to
# match the "workload" tuple in TopHub. Which padding_args to
# use for each layout chosen to reproduce previous behavior.
if dilation == 1:
padding_args = get_pad_tuple(padding, (kernel, kernel))
padding_args_i = [0, 1, 2, 3] if layout == "NCHW" else [0, 1]
padding_args = [padding_args[i] for i in padding_args_i]
else:
padding_args = padding
# placeholder
Input = te.placeholder(input_shape, name="Input", dtype=in_dtype)
Filter = te.placeholder(filter_shape, name="Filter", dtype=in_dtype)
Scale = te.placeholder(scale_shape, name="Scale", dtype=out_dtype)
Shift = te.placeholder(shift_shape, name="Shift", dtype=out_dtype)
if layout == "NCHW":
topi_scale_shift = topi.nn.scale_shift_nchw
fcompute_args = (Input, Filter, stride, padding_args, dilation,
out_dtype)
elif layout == "NHWC":
topi_scale_shift = topi.nn.scale_shift_nhwc
fcompute_args = (Input, Filter, stride, padding_args, dilation,
out_dtype)
elif layout == "NCHWc":
topi_scale_shift = topi.nn.scale_shift_nchwc
in_layout = "NCHW{}c".format(input_shape[-1])
out_layout = "NCHW{}c".format(filter_shape[-1])
fcompute_args = (
Input,
Filter,
stride,
padding,
dilation,
in_layout,
out_layout,
out_dtype,
)
with tvm.target.Target(target_hexagon):
# Declare, build schedule
if layout == "NCHW":
fcompute = topi.nn.depthwise_conv2d_nchw
fschedule = topi.hexagon.schedule_depthwise_conv2d_nchw
elif layout == "NHWC":
fcompute = topi.nn.depthwise_conv2d_nhwc
fschedule = topi.hexagon.schedule_depthwise_conv2d_nhwc
C = fcompute(*fcompute_args)
if use_scale_shift:
C = topi_scale_shift(C, Scale, Shift)
if apply_relu:
C = topi.nn.relu(C)
s = fschedule([C])
# Build and run
f = tvm.build(
s,
[Input, Filter, Scale, Shift, C],
tvm.target.Target(target_hexagon, host=target_hexagon),
)
mod = hexagon_session.load_module(f)
input_np, filter_np, scale_np, shift_np, output_np = ref_data
dev = hexagon_session.device
input_tvm = tvm.nd.array(input_np, dev)
filter_tvm = tvm.nd.array(filter_np, dev)
scale_tvm = tvm.nd.array(scale_np, dev)
shift_tvm = tvm.nd.array(shift_np, dev)
output_tvm = tvm.nd.array(
np.zeros(shape=get_const_tuple(C.shape), dtype=C.dtype),
dev,
)
mod(input_tvm, filter_tvm, scale_tvm, shift_tvm, output_tvm)
tol = {"rtol": 1e-4, "atol": 1e-5}
tvm.testing.assert_allclose(output_np, output_tvm.numpy(), **tol)