MegEngine/imperative/python/test/unit/jit/test_tracing.py

# -*- coding: utf-8 -*-
# MegEngine is Licensed under the Apache License, Version 2.0 (the "License")
#
# Copyright (c) 2014-2021 Megvii Inc. All rights reserved.
#
# Unless required by applicable law or agreed to in writing,
# software distributed under the License is distributed on an
# "AS IS" BASIS, WITHOUT ARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
import inspect
import io
import itertools
from tempfile import mkstemp

import numpy as np
import pytest

import megengine.core.tensor.megbrain_graph as G
import megengine.functional as F
import megengine.optimizer as optim
import megengine.utils.comp_graph_tools as cgtools
from megengine import Parameter, tensor
from megengine.autodiff import GradManager
from megengine.core._trace_option import set_symbolic_shape
from megengine.core.ops import builtin as ops
from megengine.core.ops.builtin import Elemwise
from megengine.core.tensor.utils import isscalar
from megengine.functional import exp, log
from megengine.jit import exclude_from_trace, trace
from megengine.module import Module
from megengine.random import normal, uniform
from megengine.utils.naming import AutoNaming


@pytest.mark.parametrize("trace_mode", [False, True])
@pytest.mark.parametrize("return_mode", ["Value", "Tuple", "List", "Dict"])
def test_trace(trace_mode, return_mode):
    @trace(symbolic=trace_mode)
    def f(x):
        if return_mode == "Tuple":
            return (-x,)
        elif return_mode == "List":
            return [-x]
        elif return_mode == "Dict":
            return {"neg": -x}
        else:
            return -x

    def get_numpy(y):
        if return_mode == "Tuple" or return_mode == "List":
            return y[0].numpy()
        elif return_mode == "Dict":
            return y["neg"].numpy()
        return y.numpy()

    x = tensor([1])
    y = get_numpy(f(x))

    for i in range(3):
        np.testing.assert_equal(get_numpy(f(x)), y)


def test_output_copy_trace():
    class Simple(Module):
        def __init__(self):
            super().__init__()
            self.a = Parameter([1.0], dtype=np.float32)

        def forward(self, x):
            x = x * self.a
            # will result into a copy of output in grad
            x = F.exp(x)
            return x

    ys = {False: [], True: []}

    for symbolic in [False, True]:
        net = Simple()
        gm = GradManager().attach(net.parameters())
        opt = optim.SGD(net.parameters(), 1e-3, momentum=0.9)
        data = tensor(np.arange(4).reshape(2, 2), dtype="float32")

        @trace(symbolic=symbolic)
        def train_func(d):
            with gm:
                loss = net(d)
                gm.backward(loss)
                opt.step().clear_grad()
            return loss

        for i in range(3):
            y = train_func(data).numpy()
            ys[symbolic].append(y)

    for i in range(3):
        np.testing.assert_equal(ys[False][i], ys[True][i])


@pytest.mark.parametrize("trace_mode", [False, True])
def test_exclude_from_trace(trace_mode):
    @trace(symbolic=trace_mode)
    def f(x):
        x = -x
        with exclude_from_trace():
            if i % 2:
                x = -x
        x = -x
        return x

    x = tensor([1])

    for i in range(3):
        y = f(x).numpy()
        np.testing.assert_equal(f(x).numpy(), y)


@pytest.mark.parametrize("trace_mode", [False, True])
def test_elemwise_fuse(trace_mode):
    # explicitly declare opt_level as 2
    @trace(symbolic=trace_mode, opt_level=2)
    def f(a, b):
        base = 0
        c = b - a
        _, idx = F.topk(c, 3)
        # internally, biased_idx will be idx as gopt will ignore the addition
        biased_idx = base + idx
        return biased_idx

    a = tensor(np.ones((7, 2)), dtype=np.int32)
    b = tensor(2 * np.ones((7, 2)), dtype=np.float32)

    for i in range(3):
        y = f(a, b)
        y.numpy()


@pytest.mark.parametrize("trace_mode", [False, True])
def test_elemwise_fuse_in_grad(trace_mode):
    w = Parameter(np.ones([4, 6]), dtype="float32")

    gm = GradManager().attach(w)
    opt = optim.SGD([w], lr=0.01, momentum=0.9, weight_decay=5e-4)

    # explicitly declare opt_level as 2
    @trace(symbolic=trace_mode, opt_level=2)
    def f():
        with gm:
            wm = F.sum(w ** 2, axis=1) ** 0.5
            loss = wm.mean()
            gm.backward(loss)
            opt.step().clear_grad()
        return loss

    for i in range(3):
        y = f()
        y.numpy()


def test_print_in_trace():
    for symbolic in [False]:  # cannot read value in symbolic mode

        @trace(symbolic=symbolic)
        def f(x):
            nonlocal buf
            x = -x
            buf = x.numpy()
            x = -x
            return x

        buf = None
        x = tensor([1])

        for i in range(3):
            y = f(x).numpy()
            z = buf
            buf = None
            np.testing.assert_equal(f(x).numpy(), y)
            np.testing.assert_equal(z, buf)


def test_dump():
    @trace(symbolic=True, capture_as_const=True)
    def f(a, b):
        return a + b

    # prevent from remaining scope from exception test
    AutoNaming.clear()
    a = tensor([2])
    b = tensor([4])
    y = f(a, b).numpy()

    for i in range(3):
        np.testing.assert_equal(f(a, b).numpy(), y)

    file = io.BytesIO()
    dump_info = f.dump(file)
    assert dump_info.nr_opr == 3
    np.testing.assert_equal(dump_info.inputs, ["arg_0", "arg_1"])
    np.testing.assert_equal(dump_info.outputs, ["ADD"])
    file.seek(0)
    infer_cg = cgtools.GraphInference(file)
    result = list((infer_cg.run(a, b)).values())[0]
    np.testing.assert_equal(result[0], y)


def test_capture_dump():
    a = tensor([2])

    @trace(symbolic=True, capture_as_const=True)
    def f(x):
        return x * a

    x = tensor([3])
    y = f(x).numpy()

    for i in range(3):
        np.testing.assert_equal(f(x).numpy(), y)

    file = io.BytesIO()
    f.dump(file)
    file.seek(0)
    infer_cg = cgtools.GraphInference(file)
    result = list((infer_cg.run(x)).values())[0]
    np.testing.assert_equal(result[0], y)


def test_dump_volatile():
    p = tensor([2])

    @trace(symbolic=True, capture_as_const=True)
    def f(x):
        return x * p

    x = tensor([3])
    y = f(x).numpy()

    for i in range(3):
        np.testing.assert_equal(f(x).numpy(), y)

    file = io.BytesIO()
    f.dump(file, optimize_for_inference=False)
    file.seek(0)
    cg, _, outputs = G.load_graph(file)
    (out,) = outputs
    assert (
        cgtools.get_owner_opr_type(cgtools.get_owner_opr_inputs(out)[1])
        == "ImmutableTensor"
    )


def test_dump_backward_graph():
    x0 = tensor(np.random.randn(3, 4))
    x1 = tensor(np.random.randn(3, 4))

    gm = GradManager().attach(x0)

    @trace(symbolic=True, capture_as_const=True)
    def f(x0, x1):
        with gm:
            y = x0 * x1
            gm.backward(y, F.ones_like(y))
            dx0 = x0.grad
        return y, dx0

    y, dx0 = f(x0, x1)
    np.testing.assert_equal(dx0.numpy(), x1)

    file = io.BytesIO()
    f.dump(file, optimize_for_inference=False)
    file.seek(0)

    infer_cg = cgtools.GraphInference(file)
    results = list((infer_cg.run(x0, x1)).values())

    np.testing.assert_equal(results[0], y)
    np.testing.assert_equal(results[1], dx0)


@pytest.mark.parametrize("trace_mode", [False, True])
def test_trace_profiler(trace_mode):
    @trace(symbolic=trace_mode, profiling=True)
    def f(x):
        return -x

    x = tensor([1])
    y = f(x).numpy()

    f(x)
    f(x)  # XXX: has to run twice

    out = f.get_profile()
    assert out.get("profiler")


def test_goptions():
    @trace(symbolic=True, opt_level=0, capture_as_const=True)
    def f(x):
        # directly return x / x will not trigger gopt
        # since there's no way to tell the two x are the same
        y = 2.0 * x
        return y / y

    @trace(symbolic=True, opt_level=1, capture_as_const=True)
    def g(x):
        y = 2.0 * x
        return y / y

    d = tensor(0.0)
    assert not np.isfinite(f(d).numpy())
    np.testing.assert_equal(g(d).numpy().item(), 1.0)


def test_goptions_log_sum_exp():
    @trace(symbolic=True, opt_level=0, capture_as_const=True)
    def f(x, y):
        return log(exp(x) + exp(y))

    @trace(symbolic=True, opt_level=1, capture_as_const=True)
    def g(x, y):
        return log(exp(x) + exp(y))

    val = 1.0e4
    d = tensor(val)
    o = tensor(0.0)
    assert not np.isfinite(f(d, o).numpy())
    np.testing.assert_almost_equal(g(d, o), val)


def test_goptions_log_exp():
    @trace(symbolic=True, opt_level=0, capture_as_const=True)
    def f(x):
        return log(exp(x))

    @trace(symbolic=True, opt_level=1, capture_as_const=True)
    def g(x):
        return log(exp(x))

    f(tensor(1.0))
    _, out = mkstemp()
    f.dump(out, optimize_for_inference=False)
    *_, outputs = G.load_graph(out)
    oprs_1 = cgtools.get_oprs_seq(outputs)

    g(tensor(1.0))
    g.dump(out, optimize_for_inference=False)
    *_, outputs = G.load_graph(out)
    oprs_2 = cgtools.get_oprs_seq(outputs)

    assert len(oprs_1) - len(oprs_2) == 2


def test_optimize_for_inference():
    @trace(symbolic=True, capture_as_const=True)
    def f(x):
        return exp(x)

    _, out = mkstemp()
    f(tensor(5.0))
    f.dump(out, enable_io16xc32=True)

    res = G.load_graph(out)
    computing_input = res.output_vars_list[0].owner.inputs[0]
    assert computing_input.dtype == np.float16


def test_optimize_for_inference_broadcast():
    a = tensor(np.ones(1, dtype=np.float32))

    @trace(capture_as_const=True, symbolic_shape=True)
    def f():
        return a._broadcast(tensor([1, 10], dtype=np.int32))

    f()
    f.dump(io.BytesIO())


def test_trace_cvt_bool():
    x = tensor([0], dtype=np.int32)

    @trace(symbolic=True)
    def f(x):
        a = x.shape
        b = a[0]
        assert isscalar(b)
        return b == 0

    for i in range(3):
        np.testing.assert_equal(f(x).numpy(), False)


@pytest.mark.parametrize("trace_mode", [False, True])
def test_trace_reshape(trace_mode):
    x1 = tensor(np.random.randn(2, 10, 10))
    x2 = tensor(np.random.randn(4, 10, 10))
    x3 = tensor(np.random.randn(8, 10, 10))

    @trace(symbolic=trace_mode, capture_as_const=True)
    def f(x):
        y = x.reshape(x.shape[0], 100)
        return y

    f(x1)
    f(x2)
    f(x3)


def test_trace_topk():
    x = tensor([5, 2, 7, 1, 0, 3, 2])

    @trace(symbolic=True)
    def f(x):
        y = F.topk(x, 3)
        np.testing.assert_equal(y[0].shape.numpy(), np.array([3,]))
        return y

    for i in range(3):
        f(x)


def test_trace_warp_perspective():
    inp_shape = (1, 1, 4, 4)
    x = tensor(np.arange(16, dtype=np.float32).reshape(inp_shape))
    M_shape = (1, 3, 3)
    M = tensor(
        np.array(
            [[1.0, 0.0, 1.0], [0.0, 1.0, 1.0], [0.0, 0.0, 1.0]], dtype=np.float32
        ).reshape(M_shape)
    )

    @trace(symbolic=True)
    def f(x, M):
        out = F.vision.warp_perspective(x, M, (2, 2))
        np.testing.assert_equal(out.shape.numpy(), np.array([1, 1, 2, 2]))
        return out

    for i in range(3):
        f(x, M)


def test_raise_on_trace():
    step_count = 0
    catch_count = 0
    bad_step = 10

    class CatchMe(Exception):
        pass

    a = tensor([1, 2, 3, 4])
    b = tensor([5, 6, 7, 8])
    c = tensor([9, 0, 1, 2])

    @trace
    def add_abc(a, b, c):
        ps = a + b
        result = ps + c
        if step_count == bad_step:
            raise CatchMe("catch me")
        return result

    for i in range(100):
        try:
            d = add_abc(a, b, c)
        except CatchMe as e:
            catch_count += 1
        else:
            np.testing.assert_equal(d.numpy(), (a + b + c).numpy())
        step_count += 1

    assert catch_count == 1


@pytest.mark.parametrize("trace_mode", [False, True])
def test_trace_broadcast(trace_mode):
    x1 = tensor(np.random.randn(3, 1, 1))
    x2 = tensor(np.random.randn(1, 4, 1))
    x3 = tensor(np.random.randn(1, 1, 5))

    @trace(symbolic=trace_mode, capture_as_const=True)
    def f(x):
        y = F.broadcast_to(x, (3, 4, 5))
        return y

    f(x1)
    f(x2)
    f(x3)


def test_trace_nms():
    def make_inputs(n):
        boxes = np.zeros((n, 4))
        boxes[:, :2] = np.random.rand(n, 2) * 100
        boxes[:, 2:] = np.random.rand(n, 2) * 100 + 100

        scores = np.random.rand(n)

        return tensor(boxes), tensor(scores)

    @trace(symbolic=False)
    def f(boxes, scores):
        # with tracing, max_output must be specified
        results = F.vision.nms(boxes, scores=scores, iou_thresh=0.5, max_output=20)
        # without tracing, max output can be inferred inside nms
        with exclude_from_trace():
            _ = F.vision.nms(boxes, scores=scores, iou_thresh=0.5)
        return results

    f(*make_inputs(10))
    f(*make_inputs(20))
    f(*make_inputs(30))


def test_trace_valid_broadcast():
    x1 = tensor(np.random.randn(1, 1))
    x2 = tensor(np.random.randn(1, 2))
    shape = (tensor([2]), tensor([2]))

    @trace(symbolic=False)
    def f(x, shape):
        y = F.broadcast_to(x, shape)
        return y

    f(x1, shape)
    f(x2, shape)


def test_clip():
    x = tensor(np.random.randn(10, 10))

    @trace(symbolic=True)
    def f(x, lower, upper):
        y = F.clip(x, lower, upper)
        return y

    for i in range(3):
        f(x, tensor([0]), tensor([1]))


# test returning noncontiguous tensor from trace
def test_slice():
    @trace
    def f(x):
        return x[:, 1::2]

    x = F.arange(8).reshape(2, 4)
    f(x)
    y = f(x)
    np.testing.assert_array_equal(y.numpy(), x.numpy()[:, 1::2])
    y + y


@pytest.mark.parametrize("shape_mode", [False, True])
def test_random(shape_mode):
    def run_test(op):
        @trace(symbolic=True, symbolic_shape=shape_mode)
        def f():
            out = op(size=[10, 10])
            out_shape = out.shape
            assert out_shape is not None
            if not isinstance(out_shape, tuple):
                assert out.shape.numpy() is not None
            return out

        for _ in range(3):
            f()

    run_test(uniform)
    run_test(normal)


@pytest.mark.parametrize("shape_mode", [False, True])
def test_trace_advance_indexing(shape_mode):
    funcs = [
        lambda x, i: x[i],
        # lambda x, i, j: x[i, j],  # FIXME
        lambda x, i, j: x[i, :, j, ...],
        # lambda x, start, end: x[start:end],  # FIXME
        lambda x, start, end: x[:, 0, start:end, ..., 1],
        lambda x, vec: x[vec],
        lambda x, vec: x[vec, ..., 0, 1:3],
        lambda x, vec: x[vec, vec[0], vec[1]],
        # lambda x, i, start, end, vec: x[i, ..., :, vec, start:end],  # FIXME
        lambda x, mask: x[mask],
    ]

    inputs = {
        "x": np.random.randn(5, 5, 5, 5, 5).astype("float32"),
        "i": 0,
        "j": 2,
        "start": 1,
        "end": 3,
        "vec": [1, 2, 3],
        "mask": np.random.randn(5, 5, 5, 5, 5) >= 0,
    }
    for f in funcs:
        sig = inspect.signature(f)
        param_names = list(sig._parameters.keys())
        params = {}
        params_np = {}
        f_traced = trace(f, symbolic=False, symbolic_shape=shape_mode)
        for name in param_names:
            params[name] = tensor(inputs[name])
            params_np[name] = inputs[name]
        expected = f(**params_np)
        result_imperative = f(**params)
        np.testing.assert_equal(expected, result_imperative.numpy())
        for _ in range(3):
            result_trace = f_traced(**params)
            np.testing.assert_equal(expected, result_trace.numpy())