feat(functional/ops): add _assert_equal

GitOrigin-RevId: b7ce4158b7087886e7a9aef5c89b682cae26c646

imperative/python/megengine/functional/__init__.py

@@ -7,7 +7,7 @@
 # software distributed under the License is distributed on an
 # "AS IS" BASIS, WITHOUT ARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # pylint: disable=redefined-builtin
-from . import metric, vision
+from . import metric, utils, vision
 from .elemwise import *
 from .math import *
 from .nn import *

imperative/python/megengine/functional/metric.py

@@ -11,6 +11,7 @@ from typing import Iterable, Union
 import numpy as np
 
 from ..tensor import Tensor
+from .elemwise import abs, maximum, minimum
 from .math import topk as _topk
 from .tensor import broadcast_to, transpose
 

imperative/python/megengine/functional/utils.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.
+from ..core._imperative_rt.core2 import apply
+from ..core._imperative_rt.core2 import sync as _sync
+from ..core.ops.builtin import AssertEqual
+from ..tensor import Tensor
+from .elemwise import abs, maximum, minimum
+
+
+def _assert_equal(
+    expect: Tensor, actual: Tensor, *, maxerr: float = 0.0001, verbose: bool = False
+):
+    r"""
+    Asserts two tensors equal and returns expected value (first input).
+    It is a variant of python assert which is symbolically traceable (similar to ``numpy.testing.assert_equal``).
+    If we want to verify the correctness of model, just ``assert`` its states and outputs.
+    While sometimes we need to verify the correctness at different backends for *dumped* model
+    (or in :class:`~jit.trace` context), and no python code could be executed in that case.
+    Thus we have to use :func:`~functional.utils._assert_equal` instead.
+
+    :param expect: expected tensor value
+    :param actual: tensor to check value
+    :param maxerr: max allowed error; error is defined as the minimal of absolute and relative error
+    :param verbose: whether to print maxerr to stdout during opr exec
+    :return: expected tensor
+
+    Examples:
+
+    .. testcode::
+
+        import numpy as np
+        from megengine import tensor
+        import megengine.functional as F
+
+        x = tensor([1, 2, 3], np.float32)
+        y = tensor([1, 2, 3], np.float32)
+        print(F.utils._assert_equal(x, y, maxerr=0).numpy())
+
+    Outputs:
+
+    .. testoutput::
+
+        [1. 2. 3.]
+    """
+    err = (
+        abs(expect - actual)
+        / maximum(minimum(abs(expect), abs(actual)), Tensor(1.0, dtype="float32"))
+    ).max()
+    result = apply(AssertEqual(maxerr=maxerr, verbose=verbose), expect, actual, err)[0]
+    _sync()  # sync interpreter to get exception
+    return result

imperative/python/megengine/jit/tracing.py

@@ -28,7 +28,12 @@ from ..core._imperative_rt.core2 import (
     unset_compiled,
     unset_tracing,
 )
-from ..core._imperative_rt.ops import CollectiveComm, RemoteRecv, RemoteSend
+from ..core._imperative_rt.ops import (
+    AssertEqual,
+    CollectiveComm,
+    RemoteRecv,
+    RemoteSend,
+)
 from ..core._trace_option import set_symbolic_shape
 from ..core._wrap import device as as_device
 from ..core.ops.builtin import BackwardGraph, OpDef
@@ -110,7 +115,7 @@ class TensorInfo:
         self.data_reader = None
 
 
-_io_op_types = {CollectiveComm, RemoteSend, RemoteRecv}
+_io_op_types = {AssertEqual, CollectiveComm, RemoteSend, RemoteRecv}
 
 
 class trace:

imperative/python/test/unit/functional/test_functional.py

@@ -21,6 +21,7 @@ from megengine.core._trace_option import use_symbolic_shape
 from megengine.core.autodiff.grad import Grad
 from megengine.core.tensor.utils import make_shape_tuple
 from megengine.distributed.helper import get_device_count_by_fork
+from megengine.jit import trace
 
 
 def test_where():
@@ -746,3 +747,18 @@ def test_ones(val):
     shp = tensor(val)
     np_shp = np.array(val)
     np.testing.assert_equal(F.ones(shp), np.ones(np_shp))
+
+
+def test_assert_equal():
+    shape = (2, 3, 4, 5)
+    x = F.ones(shape, dtype=np.float32)
+    y = F.zeros(shape, dtype=np.float32) + 1.00001
+    z = F.utils._assert_equal(x, y)
+
+
+def test_assert_not_equal():
+    shape = (2, 3, 4, 5)
+    x = F.ones(shape, dtype=np.float32)
+    y = F.zeros(shape, dtype=np.float32) + 1.1
+    with pytest.raises(RuntimeError):
+        z = F.utils._assert_equal(x, y)

imperative/src/impl/ops/specializations.cpp

@@ -453,18 +453,20 @@ namespace { namespace assert_equal {
 auto apply_on_var_node(
         const OpDef& def,
         const VarNodeArray& inputs) {
-        auto&& op = static_cast<const AssertEqual&>(def);
-        mgb_assert(inputs.size() == 2);
-        OperatorNodeConfig config{op.make_name()};
-        return opr::AssertEqual::make(inputs[0], inputs[1], op.param(), config);
-
+    auto&& op = def.cast_final<AssertEqual>();
+    if (inputs.size() == 2) {
+        return opr::AssertEqual::make(inputs[0], inputs[1], op.param());
+    } else {
+        // workaround for MiniGraph, which only allow one opr in the graph
+        mgb_assert(inputs.size() == 3);
+        return opr::AssertEqual::make(inputs[0], inputs[1], inputs[2], op.param(), {});
     }
+}
 
 OP_TRAIT_REG(AssertEqual, AssertEqual)
     .apply_on_var_node(apply_on_var_node)
     .fallback();
-
-}}
+}} // assert_equal
 
 namespace { namespace uniform_rng {
 auto apply_on_var_node(

imperative/src/impl/proxy_graph.cpp

@@ -445,6 +445,12 @@ public:
 
     size_t nr_oprs_in_graph() const override {return m_opr_refkeeper.size();}
 
+    void record_async_error(std::unique_ptr<MegBrainError> async_exc) override {
+        if (!ProxyGraph::tm_async_error) {
+            std::swap(async_exc, tm_async_error);
+        }
+    }
+
     std::unique_ptr<cg::AsyncExecutable> compile(const OutputSpec &out_spec) override {mgb_assert(0);}
     SmallVector<std::unique_ptr<cg::AsyncExecutable>> compile_multi_part(
             const SmallVector<OutputSpec>& out_specs) override {mgb_assert(0);}
@@ -457,7 +463,6 @@ public:
     size_t get_device_memory_size(CompNode cn) override {mgb_assert(0);}
     size_t clear_device_memory() override {mgb_assert(0);}
     void set_as_subgraph(ComputingGraph &par_graph) override {mgb_assert(0);}
-    void record_async_error(std::unique_ptr<MegBrainError> async_exc) override {mgb_assert(0);}
 };
 
 std::atomic<size_t> ProxyGraph::ProxyGraphImpl::m_node_id = 0;
@@ -861,6 +866,8 @@ TensorPtr ProxyGraph::as_tensor(cg::OperatorNodeBase* opr, bool share) {
     }
 }
 
+thread_local std::unique_ptr<MegBrainError> ProxyGraph::tm_async_error;
+
 } // namespace imperative
 } // namespace mgb
 

imperative/src/impl/proxy_graph.h

@@ -24,6 +24,9 @@ namespace imperative {
 class ProxyGraph : public NonCopyableObj {
 public:
     static ProxyGraph* get_default_graph();
+    static std::unique_ptr<MegBrainError> get_async_error() {
+        return std::move(tm_async_error);
+    }
 
     /********************** Physical Tensor API **********************/
 
@@ -98,6 +101,8 @@ private:
     std::unique_ptr<ExecEnv> m_env;
     std::unique_ptr<StaticInferManager> m_static_infer_manager;
     std::unique_ptr<SeqCompNodeOptimizer> m_seq_comp_node_optimizer;
+
+    static thread_local std::unique_ptr<MegBrainError> tm_async_error;
 };
 
 } // namespace imperative

imperative/src/impl/proxy_graph_detail.cpp

@@ -101,6 +101,10 @@ apply_on_physical_tensor(const OpDef& def,
         }
     }
     exec(def, inputs, outputs);
+    auto async_error = ProxyGraph::get_async_error();
+    if (async_error) {
+        throw *async_error;
+    }
     return outputs;
 }