imperative.cc 86.0 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
/* Copyright (c) 2018 PaddlePaddle Authors. All Rights Reserved.

Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License. */

#include "paddle/fluid/pybind/imperative.h"
16

17
#include <Python.h>
18 19 20 21
#include <pybind11/chrono.h>
#include <pybind11/complex.h>
#include <pybind11/functional.h>
#include <pybind11/stl.h>
22

23
#include <algorithm>
24
#include <memory>
25
#include <set>
J
Jiabin Yang 已提交
26
#include <string>
27
#include <unordered_map>
28
#include <unordered_set>
29
#include <utility>
J
Jiabin Yang 已提交
30
#include <vector>
31

32
#include "paddle/fluid/imperative/all_reduce.h"
33
#include "paddle/fluid/imperative/amp_auto_cast.h"
34
#include "paddle/fluid/imperative/basic_engine.h"
35
#include "paddle/fluid/imperative/bkcl_context.h"
36
#include "paddle/fluid/imperative/data_loader.h"
37
#include "paddle/fluid/imperative/hooks.h"
38
#include "paddle/fluid/imperative/layer.h"
J
Jiabin Yang 已提交
39
#include "paddle/fluid/imperative/nccl_context.h"
40
#include "paddle/fluid/imperative/partial_grad_engine.h"
41
#include "paddle/fluid/imperative/profiler.h"
42
#include "paddle/fluid/imperative/py_layer_fwd.h"
43
#include "paddle/fluid/imperative/reducer.h"
44
#include "paddle/fluid/imperative/tracer.h"
M
minqiyang 已提交
45
#include "paddle/fluid/imperative/type_defs.h"
46
#include "paddle/fluid/memory/allocation/mmap_allocator.h"
47
#include "paddle/fluid/pybind/op_function.h"
48
#include "paddle/fluid/pybind/pybind_boost_headers.h"
L
Leo Chen 已提交
49
#include "paddle/fluid/pybind/tensor_py.h"
50

51 52 53
namespace paddle {
namespace pybind {

54 55
PyTypeObject *g_varbase_pytype = nullptr;

56 57
namespace py = ::pybind11;

58 59 60 61
class Layer : public imperative::Layer {
 public:
  using imperative::Layer::Layer;  // Inherit constructors

62 63 64 65
  std::vector<std::shared_ptr<imperative::VarBase>> Forward(
      const std::vector<std::shared_ptr<imperative::VarBase>> &inputs)
      override {
    PYBIND11_OVERLOAD(std::vector<std::shared_ptr<imperative::VarBase>>, Layer,
J
Jiabin Yang 已提交
66
                      Forward, inputs);  // NOLINT
67 68 69
  }
};

70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128
template <typename T>
static T PyObjectCast(PyObject *obj) {
  try {
    return py::cast<T>(py::handle(obj));
  } catch (py::cast_error &) {
    PADDLE_THROW(platform::errors::InvalidArgument(
        "Python object is not type of %s", typeid(T).name()));
  }
}

class PyVariableWrapperHook : public imperative::VariableWrapperHook {
 public:
  explicit PyVariableWrapperHook(PyObject *func) : py_func_(func) {
    Py_INCREF(py_func_);
  }

  ~PyVariableWrapperHook() {
    py::gil_scoped_acquire gil;
    Py_DECREF(py_func_);
  }

  std::shared_ptr<imperative::VariableWrapper> operator()(
      const std::shared_ptr<imperative::VariableWrapper> &var) override {
    py::gil_scoped_acquire gil;
    VLOG(3) << "Call PyVariableWrapperHook for var " << var->Name();

    // 1. unpack temp VarBase from VariableWrapper
    std::shared_ptr<imperative::VarBase> tmp_varbase =
        std::make_shared<imperative::VarBase>(var);

    // 2. call hook and return
    PyObject *res = nullptr;
    try {
      res = PyObject_CallFunctionObjArgs(py_func_, py::cast(tmp_varbase).ptr(),
                                         nullptr);
    } catch (platform::EnforceNotMet &e) {
      throw std::move(e);
    } catch (std::exception &e) {
      PADDLE_THROW(platform::errors::Unavailable(
          "Hook function of Tensor raises an exception: %s.", e.what()));
    } catch (...) {
      PADDLE_THROW(platform::errors::Fatal(
          "Hook function of Tensor raises an unknown exception."));
    }

    PADDLE_ENFORCE_NOT_NULL(res,
                            platform::errors::Unavailable(
                                "Hook function of Tensor return a nullptr."));
    if (res == Py_None) {
      return var;
    }

    return PyObjectCast<std::shared_ptr<imperative::VarBase>>(res)->SharedVar();
  }

 private:
  PyObject *py_func_;
};

L
Leo Chen 已提交
129 130 131 132 133
static const platform::Place PyObjectToPlace(const py::object &place_obj) {
  if (py::isinstance<platform::CPUPlace>(place_obj)) {
    return place_obj.cast<platform::CPUPlace>();
  } else if (py::isinstance<platform::CUDAPlace>(place_obj)) {
    return place_obj.cast<platform::CUDAPlace>();
134 135
  } else if (py::isinstance<platform::XPUPlace>(place_obj)) {
    return place_obj.cast<platform::XPUPlace>();
L
Leo Chen 已提交
136 137
  } else if (py::isinstance<platform::CUDAPinnedPlace>(place_obj)) {
    return place_obj.cast<platform::CUDAPinnedPlace>();
138 139
  } else if (py::isinstance<platform::NPUPlace>(place_obj)) {
    return place_obj.cast<platform::NPUPlace>();
140 141
  } else if (py::isinstance<platform::Place>(place_obj)) {
    return place_obj.cast<platform::Place>();
L
Leo Chen 已提交
142 143
  } else {
    PADDLE_THROW(platform::errors::InvalidArgument(
144
        "Place should be one of "
145
        "Place/CPUPlace/XPUPlace/CUDAPlace/CUDAPinnedPlace/NPUPlace"));
L
Leo Chen 已提交
146 147 148
  }
}

L
Leo Chen 已提交
149 150 151 152 153 154 155 156 157 158
// only initialize varbase, but not its tensor.
static void InitVarBaseOnly(imperative::VarBase *self, const std::string &name,
                            bool persistable = false, int stop_gradient = -1) {
  auto name_ = name == ""
                   ? imperative::GetCurrentTracer()->GenerateUniqueName(
                         "generated_tensor")
                   : name;

  VLOG(5) << "Init Tensor as: / name: " << name_
          << " / persistable: " << persistable
159
          << " / stop_gradient: " << stop_gradient;
L
Leo Chen 已提交
160 161 162 163 164 165 166 167 168 169 170 171 172 173
  new (self) imperative::VarBase(name_);
  if (stop_gradient != -1) {
    self->SetOverridedStopGradient(stop_gradient);
  }
  self->SetPersistable(persistable);
  self->SetType(framework::proto::VarType::LOD_TENSOR);
}

// initialize varbase and its tensor.
static void InitVarBaseAndTensor(
    imperative::VarBase *self, const py::array &array,
    const platform::Place &place, const std::string &name,
    bool persistable = false, bool zero_copy = false, int stop_gradient = -1) {
  InitVarBaseOnly(self, name, persistable, stop_gradient);
174
  auto *tensor = self->MutableVar()->GetMutable<framework::LoDTensor>();
L
Leo Chen 已提交
175
  VLOG(4) << "zero_copy: " << zero_copy;
L
Leo Chen 已提交
176 177
  if (platform::is_cpu_place(place)) {
    SetTensorFromPyArray<platform::CPUPlace>(
178
        tensor, array, BOOST_GET_CONST(platform::CPUPlace, place), zero_copy);
179 180 181
  } else if (platform::is_xpu_place(place)) {
    SetTensorFromPyArray<platform::XPUPlace>(
        tensor, array, BOOST_GET_CONST(platform::XPUPlace, place), zero_copy);
L
Leo Chen 已提交
182 183
  } else if (platform::is_gpu_place(place)) {
    SetTensorFromPyArray<platform::CUDAPlace>(
184
        tensor, array, BOOST_GET_CONST(platform::CUDAPlace, place), zero_copy);
L
Leo Chen 已提交
185 186
  } else if (platform::is_cuda_pinned_place(place)) {
    SetTensorFromPyArray<platform::CUDAPinnedPlace>(
187 188
        tensor, array, BOOST_GET_CONST(platform::CUDAPinnedPlace, place),
        zero_copy);
189 190 191
  } else if (platform::is_npu_place(place)) {
    SetTensorFromPyArray<platform::NPUPlace>(
        tensor, array, BOOST_GET_CONST(platform::NPUPlace, place), zero_copy);
192
  } else {
L
Leo Chen 已提交
193
    PADDLE_THROW(platform::errors::InvalidArgument(
194 195
        "Place should be one of "
        "CPUPlace/XPUPlace/CUDAPlace/CUDAPinnedPlace/NPUPlace"));
J
Jiabin Yang 已提交
196
  }
197 198 199 200 201
  self->SetDataType(tensor->type());
}

static void InitVarBaseFromNumpyWithKwargs(imperative::VarBase *self,
                                           const py::kwargs &kwargs) {
202
  VLOG(4) << "Init VarBase from kwargs: ";
L
Leo Chen 已提交
203 204 205 206 207 208
  auto persistable = kwargs.contains("persistable")
                         ? kwargs["persistable"].cast<bool>()
                         : false;
  auto zero_copy =
      kwargs.contains("zero_copy") ? kwargs["zero_copy"].cast<bool>() : false;
  auto name = kwargs.contains("name") ? kwargs["name"].cast<std::string>() : "";
209 210 211
  auto stop_gradient = kwargs.contains("stop_gradient")
                           ? kwargs["stop_gradient"].cast<int>()
                           : -1;
L
Leo Chen 已提交
212
  auto default_place = imperative::GetCurrentTracer()->ExpectedPlace();
L
Leo Chen 已提交
213 214 215 216 217 218 219 220 221 222 223 224

  if (kwargs.contains("value")) {
    auto array = kwargs["value"].cast<py::array>();
    // place is only used when array is given, otherwise, it is meaningless and
    // ignored
    auto place = kwargs.contains("place") ? PyObjectToPlace(kwargs["place"])
                                          : default_place;
    InitVarBaseAndTensor(self, array, place, name, persistable, zero_copy,
                         stop_gradient);
  } else {
    InitVarBaseOnly(self, name, persistable, stop_gradient);
  }
225
}
226

227 228 229
template <typename P>
static void InitVarBaseFromNumpyWithArg(imperative::VarBase *self,
                                        const py::array &array, const P &place,
L
Leo Chen 已提交
230 231
                                        bool persistable = false,
                                        bool zero_copy = false,
232 233 234 235 236
                                        std::string name = "",
                                        int stop_gradient = -1) {
  VLOG(4) << "Init VarBase from Arg: ";
  // 0: self, 1: value, 2: place, 3: persistable, 4: zero_copy, 5: name , 6:
  // stop_gradient
L
Leo Chen 已提交
237
  if (name == "") {
238 239
    name =
        imperative::GetCurrentTracer()->GenerateUniqueName("generated_tensor");
L
Leo Chen 已提交
240
  }
241 242
  VLOG(5) << "Init Tensor as: / name: " << name
          << " / persistable: " << persistable << " / zero_copy: " << zero_copy
243
          << " / stop_gradient: " << stop_gradient << " / at " << place;
L
Leo Chen 已提交
244
  new (self) imperative::VarBase(name);
245 246
  self->SetPersistable(persistable);
  auto *tensor = self->MutableVar()->GetMutable<framework::LoDTensor>();
247 248 249
  if (stop_gradient != -1) {
    self->SetOverridedStopGradient(stop_gradient);
  }
250 251 252 253 254 255
  SetTensorFromPyArray<P>(tensor, array, place, zero_copy);
  self->SetType(framework::proto::VarType::LOD_TENSOR);
  self->SetDataType(tensor->type());
}

static void InitVarBaseFromNumpyWithArgDefault(imperative::VarBase *self,
L
Leo Chen 已提交
256 257
                                               const py::array &array) {
  auto place = imperative::GetCurrentTracer()->ExpectedPlace();
258
  VLOG(4) << "Init VarBase from numpy at " << place;
L
Leo Chen 已提交
259
  InitVarBaseAndTensor(self, array, place, "");
260
}
261

262
static void InitVarBaseFromTensorWithArgDefault(
263
    imperative::VarBase *self, const framework::Tensor &tensor) {
264 265 266
  VLOG(4) << "Init VarBase";
  auto place = imperative::GetCurrentTracer()->ExpectedPlace();
  new (self) imperative::VarBase(
267
      imperative::GetCurrentTracer()->GenerateUniqueName("generated_tensor"));
268 269 270 271 272 273 274 275 276 277 278 279 280 281
  self->SetPersistable(false);
  self->SetType(framework::proto::VarType::LOD_TENSOR);
  self->SetDataType(tensor.type());
  auto *new_tensor = self->MutableVar()->GetMutable<framework::LoDTensor>();
  // Same place,share data directly
  if (place == tensor.place()) {
    new_tensor->ShareDataWith(tensor);
    VLOG(4) << "Same place, do ShareDataWith";
  } else {
    framework::TensorCopy(tensor, place, new_tensor);
    VLOG(4) << "Different place, do TensorCopy";
  }
}

282 283 284 285 286
static std::string GetTypeName(const imperative::VarBase &var) {
  if (var.Type() == framework::proto::VarType::RAW) {
    return "RAW";
  } else if (!var.Var().IsInitialized()) {
    return "nullptr";
287
  } else {
288
    return framework::ToTypeName(var.Var().Type());
289 290
  }
}
L
Leo Chen 已提交
291

292
using PyNameVarBaseMap = std::unordered_map<std::string, py::handle>;
293 294 295 296 297 298 299 300 301 302 303 304 305

// NOTE(zjl): py::handle is a very light wrapper of PyObject *.
// Unlike py::object, py::handle does not change reference count of PyObject *.
static std::vector<std::shared_ptr<imperative::VarBase>>
GetVarBaseListFromPyHandle(const py::handle &handle) {
  PyObject *py_obj = handle.ptr();  // get underlying PyObject
  // Python None is not nullptr in C++!
  if (!py_obj || py_obj == Py_None) {
    return {};
  }

  std::vector<std::shared_ptr<imperative::VarBase>> result;

306
  if (PyList_Check(py_obj)) {  // List of VarBase
307 308 309
    size_t len = PyList_GET_SIZE(py_obj);
    result.reserve(len);
    for (size_t i = 0; i < len; ++i) {
310 311 312
      PyObject *py_ivar = PyList_GET_ITEM(py_obj, i);
      PADDLE_ENFORCE_NOT_NULL(
          py_ivar, platform::errors::InvalidArgument("Python Object is NULL"));
313 314 315
      result.emplace_back(
          PyObjectCast<std::shared_ptr<imperative::VarBase>>(py_ivar));
    }
316
  } else if (PyTuple_Check(py_obj)) {  // Tuple of VarBase
317 318 319
    size_t len = PyTuple_GET_SIZE(py_obj);
    result.reserve(len);
    for (size_t i = 0; i < len; ++i) {
320 321 322
      PyObject *py_ivar = PyTuple_GET_ITEM(py_obj, i);
      PADDLE_ENFORCE_NOT_NULL(
          py_ivar, platform::errors::InvalidArgument("Python Object is NULL"));
323 324 325
      result.emplace_back(
          PyObjectCast<std::shared_ptr<imperative::VarBase>>(py_ivar));
    }
326 327 328
  } else {  // VarBase
    result.emplace_back(
        PyObjectCast<std::shared_ptr<imperative::VarBase>>(py_obj));
329 330 331 332 333
  }

  return result;
}

J
Jiabin Yang 已提交
334 335 336
static imperative::NameVarBaseMap ConvertToNameVarBaseMap(
    const PyNameVarBaseMap &map) {
  imperative::NameVarBaseMap result;
337 338 339 340 341 342
  for (auto &pair : map) {
    auto var_vec = GetVarBaseListFromPyHandle(pair.second);
    if (!var_vec.empty()) {
      result.emplace(pair.first, std::move(var_vec));
    }
  }
J
Jiabin Yang 已提交
343

344 345 346
  PADDLE_ENFORCE_EQ(
      PyErr_Occurred(), nullptr,
      platform::errors::InvalidArgument(py::str(py::handle(PyErr_Occurred()))));
347 348 349
  return result;
}

350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394
static bool PyCheckInteger(PyObject *obj) {
#if PY_VERSION_HEX < 0x03000000
  return (PyLong_Check(obj) || PyInt_Check(obj)) && !PyBool_Check(obj);
#else
  return PyLong_Check(obj) && !PyBool_Check(obj);
#endif
}

// NOTE(zhiqiu): Revised version of PySlice_GetIndices. From:
// https://github.com/python/cpython/blob/8d21aa21f2cbc6d50aab3f420bb23be1d081dac4/Objects/sliceobject.c#L103
// Original PySlice_GetIndices return wrong result when
// slice_item contains long int, such as arr[:180L].
// NOT sure why this happens !!!
// Besides, PySlice_GetIndices cannot raise error when float in slice item.
// So, I make a revised version of PySlice_GetIndices, named to
// _PySlice_GetIndices. Try to use _PySlice_Unpack which is more robust than
// PySlice_GetIndices in the future.
static int _PySlice_GetIndices(PySliceObject *r, Py_ssize_t length,
                               Py_ssize_t *start, Py_ssize_t *stop,
                               Py_ssize_t *step) {
  /* XXX support long ints */
  if (r->step == Py_None) {
    *step = 1;
  } else {
    if (PyCheckInteger(r->step)) {
      *step = PyLong_AsLong(r->step);
    } else {
      PADDLE_THROW(platform::errors::InvalidArgument(
          "Currently, VarBase.__getitem__() only allows None or integers in "
          "slice item, but received %s.",
          std::string(Py_TYPE(r->step)->tp_name)));
    }
  }
  if (r->start == Py_None) {
    *start = *step < 0 ? length - 1 : 0;
  } else {
    if (PyCheckInteger(r->start)) {
      *start = PyLong_AsLong(r->start);
    } else {
      PADDLE_THROW(platform::errors::InvalidArgument(
          "Currently, VarBase.__getitem__() only allows None or integers in "
          "slice item, but received %s.",
          std::string(Py_TYPE(r->start)->tp_name)));
    }
    if (*start < 0) *start += length;
395
    *start = std::max(*start, static_cast<Py_ssize_t>(0));
396 397 398 399 400 401 402 403 404 405 406 407 408
  }
  if (r->stop == Py_None) {
    *stop = *step < 0 ? -1 : length;
  } else {
    if (PyCheckInteger(r->stop)) {
      *stop = PyLong_AsLong(r->stop);
    } else {
      PADDLE_THROW(platform::errors::InvalidArgument(
          "Currently, VarBase.__getitem__() only allows None or integers in "
          "slice item, but received %s.",
          std::string(Py_TYPE(r->stop)->tp_name)));
    }
    if (*stop < 0) *stop += length;
409
    *stop = std::min(*stop, length);
410 411 412 413 414 415 416
  }
  if (*stop > length) return -1;
  if (*start >= length) return -1;
  if (*step == 0) return -1;
  return 0;
}

Z
zyfncg 已提交
417 418 419 420 421 422 423 424 425
static void ParseIndexingSlice(
    framework::LoDTensor *tensor, PyObject *_index,
    std::vector<int> *slice_axes, std::vector<int> *slice_starts,
    std::vector<int> *slice_ends, std::vector<int> *slice_strides,
    std::vector<int> *decrease_axis, std::vector<int> *none_axes,
    std::vector<int> *infer_flags, std::vector<int> *list_select_idxs,
    bool *list_select_flag) {
  // We allow indexing by Integers, Slices, Ellipsis, None, tuples of those
  // types, and list of Bool and Integers.
S
songyouwei 已提交
426 427 428 429 430 431 432 433
  // wrap to tuple
  PyObject *index = !PyTuple_Check(_index) ? PyTuple_Pack(1, _index) : _index;
  PADDLE_ENFORCE_EQ(
      tensor->IsInitialized(), true,
      platform::errors::InvalidArgument("tensor has not been initialized"));
  const auto &shape = tensor->dims();
  const int rank = shape.size();
  const int size = PyTuple_GET_SIZE(index);
434 435 436 437 438 439 440 441 442 443 444 445 446 447

  // specified_dims is the number of dimensions which indexed by Interger,
  // Slices.
  int specified_dims = 0;
  for (int dim = 0; dim < size; ++dim) {
    PyObject *slice_item = PyTuple_GetItem(index, dim);
    if (PyCheckInteger(slice_item) || PySlice_Check(slice_item)) {
      specified_dims++;
    }
  }

  for (int i = 0, dim = 0; i < size; ++i) {
    PyObject *slice_item = PyTuple_GetItem(index, i);

S
songyouwei 已提交
448 449
    infer_flags->push_back(1);
    int dim_len = shape[dim];
450 451
    if (PyCheckInteger(slice_item)) {
      // integer, PyLong_AsLong supports both int and long
S
songyouwei 已提交
452
      int start = static_cast<int>(PyLong_AsLong(slice_item));
H
hong 已提交
453
      auto s_t = start;
S
songyouwei 已提交
454
      start = start < 0 ? start + dim_len : start;
455
      if (start >= dim_len || start < 0) {
H
hong 已提交
456 457 458 459 460 461 462 463 464
        std::string str_error_message =
            "The starting index " + std::to_string(s_t) +
            " of slice is out of bounds in tensor " + std::to_string(dim) +
            "-th axis, it shound be in the range of [" +
            std::to_string(-dim_len) + ", " + std::to_string(dim_len) + ")";
        // py::index_error is corresponding to IndexError in Python
        // Used to indicate out of bounds access in __getitem__, __setitem__
        throw py::index_error(str_error_message);
      }
S
songyouwei 已提交
465 466 467 468 469
      slice_axes->push_back(dim);
      slice_starts->push_back(start);
      slice_ends->push_back(start + 1);
      slice_strides->push_back(1);
      decrease_axis->push_back(dim);
470 471
      dim++;
    } else if (PySlice_Check(slice_item)) {
472
      // slice item
S
songyouwei 已提交
473
      Py_ssize_t start, end, step;
474 475 476
      PySliceObject *p = reinterpret_cast<PySliceObject *>(slice_item);
      _PySlice_GetIndices(p, dim_len, &start, &end, &step);

S
songyouwei 已提交
477
      // :: or : or 0:dim_len:1
478
      if (start == 0 && end == dim_len && step == 1) {
479
        dim++;
480 481
        continue;
      }
S
songyouwei 已提交
482 483 484 485
      slice_axes->push_back(dim);
      slice_starts->push_back(start);
      slice_ends->push_back(end);
      slice_strides->push_back(step);
486 487 488
      dim++;
    } else if (slice_item == Py_Ellipsis) {
      dim += rank - specified_dims;
489 490
    } else if (slice_item == Py_None) {
      none_axes->push_back(dim);
Z
zyfncg 已提交
491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536
    } else if (PyList_Check(slice_item)) {
      *list_select_flag = true;
      if (size != 1) {
        PADDLE_THROW(platform::errors::InvalidArgument(
            "When index contains a list, its length is excepted to 1, "
            "but received %d",
            size));
      }
      bool all_bool = true;
      int list_size = PyList_GET_SIZE(slice_item);
      for (int j = 0; j < list_size; ++j) {
        PyObject *list_item = PyList_GetItem(slice_item, j);
        if (PyCheckInteger(list_item)) {
          all_bool = false;
        } else if (!PyBool_Check(list_item)) {
          PADDLE_THROW(platform::errors::InvalidArgument(
              "Only support int or bool in index list."));
        }
      }
      if (all_bool) {
        if (list_size != shape[0]) {
          PADDLE_THROW(platform::errors::InvalidArgument(
              "The dimension of bool index doesn't match indexed array along "
              "dimension 0, the target dimension is %d, but received %d.",
              shape[0], list_size));
        }
        for (int j = 0; j < list_size; ++j) {
          PyObject *list_item = PyList_GetItem(slice_item, j);
          if (list_item == Py_True) {
            list_select_idxs->push_back(j);
          }
        }
      } else {
        for (int j = 0; j < list_size; ++j) {
          PyObject *list_item = PyList_GetItem(slice_item, j);
          if (PyCheckInteger(list_item)) {
            list_select_idxs->push_back(
                static_cast<int>(PyLong_AsLong(list_item)));
          } else if (list_item == Py_True) {
            list_select_idxs->push_back(1);
          } else {
            list_select_idxs->push_back(0);
          }
        }
      }

537 538
    } else {
      PADDLE_THROW(platform::errors::InvalidArgument(
Z
zyfncg 已提交
539 540 541 542
          "Currently, VarBase.__getitem__() only allows indexing "
          "by Integers, Slices, Ellipsis, None, tuples of these types "
          "and list of Bool and Integers, but received "
          "%s in %dth slice item",
543
          std::string(Py_TYPE(slice_item)->tp_name), i + 1));
S
songyouwei 已提交
544 545
    }
  }
546 547 548 549 550 551 552 553

  // valid_index is the number of dimensions exclude None index
  const int valid_indexs = size - none_axes->size();
  PADDLE_ENFORCE_EQ(valid_indexs <= rank, true,
                    platform::errors::InvalidArgument(
                        "Too many indices (%d) for tensor of dimension %d.",
                        valid_indexs, rank));

S
songyouwei 已提交
554 555 556
  if (!PyTuple_Check(_index)) Py_DecRef(index);
}

557
template <typename P>
558 559 560
static void VarBaseCopy(std::shared_ptr<imperative::VarBase> &src,  // NOLINT
                        imperative::VarBase &dst,                   // NOLINT
                        const P &dst_device, const bool blocking) {
561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612
  if (dst.SharedVar()->IsEmpty()) {
    VLOG(3) << "deep copy Variable from " << src->Name() << " to "
            << dst.Name();
    dst.SetPersistable(src->Persistable());
    dst.SetDataType(src->DataType());
    dst.SetType(src->Type());
    dst.SetOverridedStopGradient(src->OverridedStopGradient());
    if (!src->SharedVar()->IsEmpty()) {
      if (src->Var().IsType<framework::LoDTensor>()) {
        auto &src_tensor = src->Var().Get<framework::LoDTensor>();
        auto *dst_tensor = dst.MutableVar()->GetMutable<framework::LoDTensor>();
        dst_tensor->set_lod(src_tensor.lod());
        framework::TensorCopy(src_tensor, dst_device, dst_tensor);
        if (blocking) {
          platform::DeviceContextPool::Instance().Get(dst_device)->Wait();
          auto src_device = src_tensor.place();
          if (!(src_device == dst_device)) {
            platform::DeviceContextPool::Instance().Get(src_device)->Wait();
          }
        }
      } else if (src->Var().IsType<framework::SelectedRows>()) {
        auto &src_selected_rows = src->Var().Get<framework::SelectedRows>();
        auto *dst_selected_rows =
            dst.MutableVar()->GetMutable<framework::SelectedRows>();
        dst_selected_rows->set_height(src_selected_rows.height());
        dst_selected_rows->set_rows(src_selected_rows.rows());
        framework::TensorCopy(src_selected_rows.value(), dst_device,
                              dst_selected_rows->mutable_value());
        if (blocking) {
          platform::DeviceContextPool::Instance().Get(dst_device)->Wait();
          auto src_device = src_selected_rows.value().place();
          if (!(src_device == dst_device)) {
            platform::DeviceContextPool::Instance().Get(src_device)->Wait();
          }
        }
      }

      if (!blocking) {
        IncreaseVarbaseReferenceCountUntilCopyComplete(src, dst_device);
      }

    } else {
      PADDLE_THROW(platform::errors::InvalidArgument(
          "The source Tensor(%s) can not copy when it is empty.", src->Name()));
    }
  } else {
    PADDLE_THROW(platform::errors::InvalidArgument(
        "The destion Tensor(%s) can not copy when it is not empty.",
        dst.Name()));
  }
}

613
// Bind Methods
J
Jiabin Yang 已提交
614
void BindImperative(py::module *m_ptr) {
615 616
  auto &m = *m_ptr;

617 618
  BindOpFunctions(&m);

619 620
#ifndef _WIN32
  // Dygraph DataLoader signal handler
621 622 623 624 625 626 627 628 629 630 631 632 633
  m.def("_set_process_pids", [](int64_t key, py::object &obj) {
    PADDLE_ENFORCE_EQ(
        py::isinstance<py::tuple>(obj) || py::isinstance<py::list>(obj), true,
        platform::errors::InvalidArgument(
            "The subprocess ids set in DataLoader is illegal."
            "Expected data type is tuple or list, but received %s",
            obj.get_type()));
    py::list pids = py::cast<py::list>(obj);
    std::set<pid_t> pids_set = {};
    for (size_t i = 0; i < pids.size(); i++) {
      pids_set.insert(pids[i].cast<pid_t>());
    }
    imperative::SetLoadProcessPIDs(key, pids_set);
634
  });
635 636
  m.def("_erase_process_pids",
        [](int64_t key) { imperative::EraseLoadProcessPIDs(key); });
637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688
  m.def("_set_process_signal_handler",
        []() { imperative::SetLoadProcessSignalHandler(); });
  m.def("_throw_error_if_process_failed",
        []() { imperative::ThrowErrorIfLoadProcessFailed(); });

  // Dygraph DataLoader reader process & thread related functions
  m.def(
      "_convert_to_tensor_list",
      [](py::object &obj) -> py::list {
        // 0. input data check
        PADDLE_ENFORCE(
            py::isinstance<py::tuple>(obj) || py::isinstance<py::list>(obj),
            platform::errors::InvalidArgument(
                "The batch data read into DataLoader is illegal."
                "Expected data type is tuple or list, but received %s",
                obj.get_type()));
        py::list batch = py::cast<py::list>(obj);
        py::list tensors;
        for (size_t i = 0; i < batch.size(); ++i) {
          // 1. cast to python array
          auto array = batch[i].cast<py::array>();
          PADDLE_ENFORCE_NE(
              string::Sprintf("%s", array.dtype()).compare("object"), 0,
              platform::errors::InvalidArgument(
                  "Faild to convert input data to a regular ndarray.\n  * "
                  "Usually this means the input data contains nested "
                  "lists with different lengths.\n  * Check the reader "
                  "function passed to 'set_(sample/sample_list/batch)"
                  "_generator' to locate the data causes this issue."));
          // 2. construcct LoDTensor
          framework::LoDTensor t;
          SetTensorFromPyArray<platform::CPUPlace>(&t, array,
                                                   platform::CPUPlace(), true);
          // 3. allocate shared memory
          void *data_ptr = t.data<void>();
          size_t data_size = t.numel() * framework::SizeOfType(t.type());
          auto shared_writer_holder =
              memory::allocation::AllocateMemoryMapWriterAllocation(data_size);
          // 4. maintain mmap fd set & backup ipc_name
          const std::string &ipc_name = shared_writer_holder->ipc_name();
          memory::allocation::MemoryMapFdSet::Instance().Insert(ipc_name);
          // 5. copy data & reset holder
          memory::Copy(platform::CPUPlace(), shared_writer_holder->ptr(),
                       platform::CPUPlace(), data_ptr, data_size);
          t.ResetHolder(shared_writer_holder);
          // 6. append to result list
          tensors.append(t);
        }
        return tensors;
      },
      py::return_value_policy::take_ownership);

K
Kaipeng Deng 已提交
689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721
  m.def("_array_to_share_memory_tensor",
        [](py::object &obj) {
          // 1. cast to python array
          auto array = obj.cast<py::array>();
          PADDLE_ENFORCE_NE(
              string::Sprintf("%s", array.dtype()).compare("object"), 0,
              platform::errors::InvalidArgument(
                  "Faild to convert input data to a regular ndarray.\n  * "
                  "Usually this means the input data contains nested "
                  "lists with different lengths.\n  * Check the reader "
                  "function passed to 'set_(sample/sample_list/batch)"
                  "_generator' to locate the data causes this issue."));
          // 2. construcct LoDTensor
          framework::LoDTensor t;
          SetTensorFromPyArray<platform::CPUPlace>(&t, array,
                                                   platform::CPUPlace(), true);
          // 3. allocate shared memory
          void *data_ptr = t.data<void>();
          size_t data_size = t.numel() * framework::SizeOfType(t.type());
          auto shared_writer_holder =
              memory::allocation::AllocateMemoryMapWriterAllocation(data_size);
          // 4. maintain mmap fd set & backup ipc_name
          const std::string &ipc_name = shared_writer_holder->ipc_name();
          memory::allocation::MemoryMapFdSet::Instance().Insert(ipc_name);
          // 5. copy data & reset holder
          memory::Copy(platform::CPUPlace(), shared_writer_holder->ptr(),
                       platform::CPUPlace(), data_ptr, data_size);
          t.ResetHolder(shared_writer_holder);

          return t;
        },
        py::return_value_policy::take_ownership);

722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741
  m.def("_remove_tensor_list_mmap_fds", [](py::list &tensor_list) {
    for (size_t i = 0; i < tensor_list.size(); ++i) {
      auto t = tensor_list[i].cast<framework::LoDTensor>();
      auto *mmap_writer_allocation =
          dynamic_cast<memory::allocation::MemoryMapWriterAllocation *>(
              t.Holder().get());
      PADDLE_ENFORCE_NOT_NULL(
          mmap_writer_allocation,
          platform::errors::NotFound("The shared memory of LoDTensor in "
                                     "DataLoader's child process has been "
                                     "released."));
      memory::allocation::MemoryMapFdSet::Instance().Remove(
          mmap_writer_allocation->ipc_name());
    }
  });

  m.def("_cleanup_mmap_fds",
        []() { memory::allocation::MemoryMapFdSet::Instance().Clear(); });
#endif

742 743 744 745 746
  m.def("start_imperative_gperf_profiler",
        []() { imperative::StartProfile(); });

  m.def("stop_imperative_gperf_profiler", []() { imperative::StopProfile(); });

Z
Zeng Jinle 已提交
747 748 749
  m.def("_is_dygraph_debug_enabled",
        []() { return imperative::IsDebugEnabled(); });
  m.def("_dygraph_debug_level", []() { return imperative::GetDebugLevel(); });
750 751 752 753
  m.def("_switch_tracer",
        [](const std::shared_ptr<imperative::Tracer> &tracer) {
          imperative::SetCurrentTracer(tracer);
        });
Z
Zeng Jinle 已提交
754

755 756 757 758
  py::class_<imperative::VarBase, std::shared_ptr<imperative::VarBase>> varbase(
      m, "VarBase", R"DOC()DOC");
  g_varbase_pytype = (PyTypeObject *)varbase.ptr();  // NOLINT
  varbase.def_static("_alive_vars", &imperative::VarBase::AliveVarNames)
759 760 761 762 763 764 765
      .def("__init__",
           [](imperative::VarBase &self) {
             std::string name =
                 imperative::GetCurrentTracer()->GenerateUniqueName(
                     "generated_tensor");
             new (&self) imperative::VarBase(name);
           })
J
Jiabin Yang 已提交
766
      .def("__init__",
767 768 769
           [](imperative::VarBase &self, framework::proto::VarType::Type dtype,
              const std::vector<int> &dims, const py::handle &name,
              framework::proto::VarType::Type type, bool persistable) {
770
             VLOG(4) << "Init VarBase";
771 772 773
             std::string act_name = "";
             if (!name.ptr() || name.ptr() == Py_None) {
               act_name = imperative::GetCurrentTracer()->GenerateUniqueName(
774
                   "generated_tensor");
775 776 777 778
             } else {
               act_name = name.cast<std::string>();
             }
             new (&self) imperative::VarBase(act_name);
J
Jiabin Yang 已提交
779 780 781 782 783 784 785 786 787
             self.SetPersistable(persistable);
             self.SetType(type);
             self.SetDataType(dtype);
             if (type == framework::proto::VarType::LOD_TENSOR) {
               auto *tensor =
                   self.MutableVar()->GetMutable<framework::LoDTensor>();
               tensor->Resize(framework::make_ddim(dims));
             }
           })
788 789
      .def("__init__", &InitVarBaseFromNumpyWithArg<platform::CPUPlace>,
           py::arg("value"), py::arg("place"), py::arg("persistable") = false,
790 791
           py::arg("zero_copy") = false, py::arg("name") = "",
           py::arg("stop_gradient") = -1)
792 793 794 795
      .def("__init__", &InitVarBaseFromNumpyWithArg<platform::XPUPlace>,
           py::arg("value"), py::arg("place"), py::arg("persistable") = false,
           py::arg("zero_copy") = false, py::arg("name") = "",
           py::arg("stop_gradient") = -1)
796 797
      .def("__init__", &InitVarBaseFromNumpyWithArg<platform::CUDAPlace>,
           py::arg("value"), py::arg("place"), py::arg("persistable") = false,
798 799
           py::arg("zero_copy") = false, py::arg("name") = "",
           py::arg("stop_gradient") = -1)
800 801
      .def("__init__", &InitVarBaseFromNumpyWithArg<platform::CUDAPinnedPlace>,
           py::arg("value"), py::arg("place"), py::arg("persistable") = false,
802 803
           py::arg("zero_copy") = false, py::arg("name") = "",
           py::arg("stop_gradient") = -1)
804 805 806 807
      .def("__init__", &InitVarBaseFromNumpyWithArg<platform::NPUPlace>,
           py::arg("value"), py::arg("place"), py::arg("persistable") = false,
           py::arg("zero_copy") = false, py::arg("name") = "",
           py::arg("stop_gradient") = -1)
L
Leo Chen 已提交
808
      .def("__init__", &InitVarBaseFromNumpyWithArgDefault, py::arg("value"))
809
      .def("__init__", &InitVarBaseFromTensorWithArgDefault, py::arg("tensor"))
810
      .def("__init__", &InitVarBaseFromNumpyWithKwargs)
811 812 813 814 815
      .def("__setitem__",
           [](std::shared_ptr<imperative::VarBase> &self, py::handle _index,
              py::object &value_obj) {
             auto self_tensor =
                 self->MutableVar()->GetMutable<framework::LoDTensor>();
816 817 818 819
             PyObject *index_ptr = !PyTuple_Check(_index.ptr())
                                       ? PyTuple_Pack(1, _index.ptr())
                                       : _index.ptr();
             // 1. Check argumnets
820 821
             // 1.1 Check whether value obj is a tensor.
             bool value_is_tensor = true;
822
             bool parse_index = true;
823 824 825 826 827 828 829
             if (py::isinstance<py::array>(value_obj) ||
                 py::isinstance<py::int_>(value_obj) ||
                 py::isinstance<py::float_>(value_obj)) {
               value_is_tensor = false;
             }

             // 1.2 Check whether _index can be parsed.
830 831 832
             const int size = PyTuple_GET_SIZE(index_ptr);
             for (int dim = 0; dim < size; ++dim) {
               PyObject *slice_item = PyTuple_GetItem(index_ptr, dim);
833 834
               if (!(PyCheckInteger(slice_item) || PySlice_Check(slice_item) ||
                     slice_item == Py_Ellipsis || slice_item == Py_None)) {
835 836 837 838 839 840 841 842 843
                 parse_index = false;
                 break;
               }
             }

             // 2. Call op set_value to speed up if the condition is met,
             // otherwise call TensorToPyArray.
             // TODO(liym27): Try not to call TensorToPyArray because it always
             // copys data to cpu place, which reduces performance.
844
             if (parse_index && value_is_tensor) {
845
               std::vector<int> axes, starts, ends, steps, decrease_axes,
Z
zyfncg 已提交
846 847 848
                   none_axes, infer_flags, list_select_idxs;
               // if index is a list, list_select_flag will be true
               bool list_select_flag;
849
               ParseIndexingSlice(self_tensor, index_ptr, &axes, &starts, &ends,
850
                                  &steps, &decrease_axes, &none_axes,
Z
zyfncg 已提交
851 852
                                  &infer_flags, &list_select_idxs,
                                  &list_select_flag);
853 854 855 856 857 858

               framework::AttributeMap attrs = {
                   {"axes", axes},
                   {"starts", starts},
                   {"ends", ends},
                   {"steps", steps},
859 860
                   {"decrease_axes", decrease_axes},
                   {"none_axes", none_axes}};
861 862 863

               imperative::NameVarBaseMap ins = {{"Input", {self}}};
               imperative::NameVarBaseMap outs = {{"Out", {self}}};
864

865 866 867 868 869 870 871
               PADDLE_ENFORCE_EQ(
                   self->IsLeaf() && !self->OverridedStopGradient(), false,
                   platform::errors::InvalidArgument(
                       "Leaf Tensor (%s) that doesn't stop gradient can't use "
                       "inplace strategy.",
                       self->Name()));

872 873 874 875 876 877 878 879
               auto value_tensor =
                   value_obj.cast<std::shared_ptr<imperative::VarBase>>();
               ins.insert({"ValueTensor", {value_tensor}});

               const auto &tracer = imperative::GetCurrentTracer();
               {
                 // Release gil and do tracing
                 py::gil_scoped_release release;
880 881
                 tracer->TraceOp("set_value", ins, outs, std::move(attrs),
                                 {{"Input", "Out"}});
882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901
               }
             } else {
               auto self_numpy = TensorToPyArray(*self_tensor);

               if (value_is_tensor) {
                 auto value =
                     value_obj.cast<std::shared_ptr<imperative::VarBase>>();
                 auto value_tensor =
                     value->MutableVar()->GetMutable<framework::LoDTensor>();
                 auto value_numpy = TensorToPyArray(*value_tensor);

                 self_numpy[_index] = value_numpy;
                 SetTensorFromPyArray(self_tensor, self_numpy,
                                      self_tensor->place(), true);
               } else {
                 auto value_numpy = value_obj;
                 self_numpy[_index] = value_numpy;
                 SetTensorFromPyArray(self_tensor, self_numpy,
                                      self_tensor->place(), true);
               }
902
             }
903 904 905 906
             // NOTE(liym27):
             // Increase the version of VarBase self because __setitem__ is an
             // inplace operator for the VarBase self.
             self->BumpInplaceVersion();
907
           })
908
      .def("_getitem_index_not_tensor",
S
songyouwei 已提交
909
           [](std::shared_ptr<imperative::VarBase> &self, py::handle _index) {
910
             std::vector<int> slice_axes, slice_starts, slice_ends,
Z
zyfncg 已提交
911 912 913 914
                 slice_strides, decrease_axis, none_axes, infer_flags,
                 list_select_idxs;
             // if index is a list, list_select_flag will be true
             bool list_select_flag = false;
S
songyouwei 已提交
915 916 917 918
             auto tensor =
                 self->MutableVar()->GetMutable<framework::LoDTensor>();
             ParseIndexingSlice(tensor, _index.ptr(), &slice_axes,
                                &slice_starts, &slice_ends, &slice_strides,
Z
zyfncg 已提交
919 920
                                &decrease_axis, &none_axes, &infer_flags,
                                &list_select_idxs, &list_select_flag);
921 922 923
             // release gil and do tracing
             py::gil_scoped_release release;
             const auto &tracer = imperative::GetCurrentTracer();
924

Z
zyfncg 已提交
925
             auto out = slice_axes.empty() && !list_select_flag
926 927 928 929
                            ? self
                            : std::shared_ptr<imperative::VarBase>(
                                  new imperative::VarBase(
                                      tracer->GenerateUniqueName()));
Z
zyfncg 已提交
930

931
             if (!slice_axes.empty()) {
S
songyouwei 已提交
932
               imperative::NameVarBaseMap ins = {{"Input", {self}}};
933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950
               framework::AttributeMap attrs = {
                   {"axes", slice_axes},
                   {"starts", slice_starts},
                   {"ends", slice_ends},
                   {"infer_flags", infer_flags},
                   {"decrease_axis", decrease_axis}};
               imperative::NameVarBaseMap outs = {{"Out", {out}}};
               std::string op_type = "slice";
               for (auto stride : slice_strides) {
                 if (stride != 1) {
                   op_type = "strided_slice";
                   attrs.insert({"strides", slice_strides});
                   attrs.erase("decrease_axis");
                   break;
                 }
               }
               tracer->TraceOp(op_type, ins, outs, std::move(attrs));
             }
951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978
             if (!none_axes.empty()) {
               // Deal with cases when all axes are decreased.
               // After slice, the shape of out is [1], which should have been
               // [], but Paddle doesn't support scalar.
               // In order to ensure the correctness of the final shape of out,
               // one dimension of out needs to be decreased.
               // For example:
               // # x.shape: (2,3,4)
               // out = x[0, 1, 1, None] # out.shape : (1)
               if (static_cast<int>(decrease_axis.size()) ==
                   tensor->dims().size()) {
                 none_axes.pop_back();
               }
               if (!none_axes.empty()) {
                 // Deal with cases that decrease_axes is not empty
                 // For example:
                 // # x.shape: (2,3,4)
                 // out = x[0, 0:2, None] # out.shape : (2, 1, 4)
                 for (auto &axis : none_axes) {
                   int len = 0;
                   for (int da : decrease_axis) {
                     if (da < axis) {
                       len++;
                     }
                   }
                   axis -= len;
                 }

979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001
                 // Deal with cases that there are more than one
                 // prefix none index, For example:
                 // [None, None, :, :, None]
                 // the none_axes int the return of ParseIndexingSlice is:
                 // [0,    0,          2   ]
                 // according to the interface of "unsqueeze2",
                 // we should convert it to:
                 // [0,    0,          4   ]
                 int prefix_zero_cnt = 0;
                 for (const auto &axis : none_axes) {
                   if (axis == 0) {
                     prefix_zero_cnt++;
                   } else {
                     break;
                   }
                 }
                 if (prefix_zero_cnt > 0) {
                   int none_axes_num = static_cast<int>(none_axes.size());
                   for (int i = prefix_zero_cnt; i < none_axes_num; ++i) {
                     none_axes[i] += prefix_zero_cnt;
                   }
                 }

1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015
                 imperative::NameVarBaseMap ins = {{"X", {out}}};
                 framework::AttributeMap attrs = {{"axes", none_axes}};
                 auto new_out = std::shared_ptr<imperative::VarBase>(
                     new imperative::VarBase(tracer->GenerateUniqueName()));
                 auto out_xshape = std::shared_ptr<imperative::VarBase>(
                     new imperative::VarBase(tracer->GenerateUniqueName()));
                 imperative::NameVarBaseMap outs = {{"Out", {new_out}},
                                                    {"XShape", {out_xshape}}};
                 tracer->TraceOp("unsqueeze2", ins, outs, std::move(attrs));

                 return new_out;
               }
             }

Z
zyfncg 已提交
1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031
             // the index is a list
             if (list_select_flag) {
               auto select_index = std::shared_ptr<imperative::VarBase>(
                   new imperative::VarBase(tracer->GenerateUniqueName()));
               auto *idx_tensor = select_index->MutableVar()
                                      ->GetMutable<framework::LoDTensor>();
               auto *dev_ctx = platform::DeviceContextPool::Instance().Get(
                   tracer->ExpectedPlace());
               TensorFromVector(list_select_idxs, *dev_ctx, idx_tensor);

               imperative::NameVarBaseMap ins = {{"X", {self}},
                                                 {"Index", {select_index}}};
               imperative::NameVarBaseMap outs = {{"Out", {out}}};
               tracer->TraceOp("index_select", ins, outs, {{"dim", 0}});
             }

1032
             return out;
1033
           })
1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097
      .def(
          "_getitem_from_offset",
          [](std::shared_ptr<imperative::VarBase> &self, const py::args &args) {
            const auto &tensor = self->Var().Get<framework::LoDTensor>();
            PADDLE_ENFORCE_EQ(
                tensor.IsInitialized(), true,
                platform::errors::InvalidArgument(
                    "Tensor of %s is Empty, please check if it has no data.",
                    self->Name()));

            const auto &tensor_dims = tensor.dims();

            std::vector<size_t> dims(tensor_dims.size());
            std::vector<size_t> strides(tensor_dims.size());

            size_t numel = 1;
            for (int i = tensor_dims.size() - 1; i >= 0; --i) {
              strides[i] = numel;
              dims[i] = static_cast<size_t>(tensor_dims[i]);
              numel *= dims[i];
            }
            size_t offset = 0;
            if (args.empty()) {
              PADDLE_ENFORCE_EQ(
                  numel, 1,
                  platform::errors::InvalidArgument(
                      "only one element tensors can be converted to Python "
                      "scalars when no input coordinates"));
            } else if (args.size() == 1) {
              offset = args[0].cast<size_t>();
              PADDLE_ENFORCE_LT(
                  offset, numel,
                  platform::errors::InvalidArgument(
                      "index %d is out of bounds for size %d", offset, numel));
            } else {
              PADDLE_ENFORCE_EQ(args.size(), dims.size(),
                                platform::errors::InvalidArgument(
                                    "incorrect number of indices for Tensor"));

              for (size_t i = 0; i < args.size(); ++i) {
                size_t index = args[i].cast<size_t>();
                PADDLE_ENFORCE_LT(
                    index, dims[i],
                    platform::errors::InvalidArgument(
                        "index %d is out fo bounds for axis %d with size %d",
                        index, i, dims[i]));
                offset += index * strides[i];
              }
            }
#define TENSOR_TO_PY_SCALAR(T, proto_type)                                   \
  if (tensor.type() == proto_type) {                                         \
    std::string py_dtype_str = details::TensorDTypeToPyDTypeStr(proto_type); \
    T b = TensorGetElement<T>(tensor, offset);                               \
    return py::array(py::dtype(py_dtype_str.c_str()), {}, {},                \
                     static_cast<void *>(&b));                               \
  }

            _ForEachDataType_(TENSOR_TO_PY_SCALAR);
#undef TENSOR_TO_PY_SCALAR
            PADDLE_THROW(platform::errors::Unimplemented(
                "Unsupported tensor data type: %s",
                framework::DataTypeToString(tensor.type())));
          },
          py::return_value_policy::copy)
1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119
      .def("_inplace_version",
           [](imperative::VarBase &self) -> uint32_t {
             const auto &var = self.MutableVar();
             PADDLE_ENFORCE_EQ(
                 var->IsInitialized(), true,
                 platform::errors::InvalidArgument(
                     "Tensor of %s is Empty, please check if it has no data.",
                     self.Name()));
             return var->CurrentInplaceVersion();
           })
      .def("_bump_inplace_version",
           [](std::shared_ptr<imperative::VarBase> &self) {
             // NOTE(liym27): _bump_inplace_version is only used for inplace
             // operation
             self->BumpInplaceVersion();
           },
           R"DOC(
        **Notes**:
            **This API is ONLY available in Dygraph mode.**
            **This is a very low level API. Users should not use it directly. **
         Bump the version whenever the Tensor is modified through an inplace operation.
            )DOC")
1120
      .def("numpy",
1121

1122 1123 1124 1125 1126 1127
           [](imperative::VarBase &self) -> py::array {
             const auto &tensor =
                 self.MutableVar()->Get<framework::LoDTensor>();
             PADDLE_ENFORCE_EQ(
                 tensor.IsInitialized(), true,
                 platform::errors::InvalidArgument(
1128
                     "Tensor of %s is Empty, please check if it has no data.",
1129 1130 1131 1132
                     self.Name()));
             return TensorToPyArray(tensor, true);
           },
           R"DOC(
Z
Zhou Wei 已提交
1133 1134
        Returns a numpy array shows the value of current Tensor.
        
1135
        Returns:
Z
Zhou Wei 已提交
1136
            ndarray: The numpy value of current Tensor.
1137 1138

        Returns type:
Z
Zhou Wei 已提交
1139
            ndarray: dtype is same as current Tensor
1140 1141 1142 1143

        Examples:
            .. code-block:: python

Z
Zhou Wei 已提交
1144
                import paddle
1145 1146
                import numpy as np
                data = np.random.uniform(-1, 1, [30, 10, 32]).astype('float32')
Z
Zhou Wei 已提交
1147 1148 1149 1150
                linear = paddle.nn.Linear(32, 64)
                data = paddle.to_tensor(data)
                x = linear(data)
                print(x.numpy())
1151
       )DOC")
1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214
      .def("detach",
           [](const imperative::VarBase
                  &self) -> std::shared_ptr<imperative::VarBase> {
             PADDLE_ENFORCE_EQ(
                 self.Var().IsInitialized(), true,
                 platform::errors::InvalidArgument(
                     "Tensor %s has not been initialized!", self.Name()));

             PADDLE_ENFORCE_EQ(
                 self.Var().IsType<framework::LoDTensor>() ||
                     self.Var().IsType<framework::SelectedRows>(),
                 true,
                 platform::errors::InvalidArgument(
                     "Type of Tensor[%s] must be LoDTensor or SelectedRows!",
                     self.Name()));

             auto detach_var = std::make_shared<imperative::VarBase>(
                 true, "detach_" + self.Name());

             detach_var->SetPersistable(self.Persistable());
             detach_var->SetType(self.Type());
             detach_var->SetDataType(self.DataType());

             if (self.Var().IsType<framework::LoDTensor>()) {
               const auto &origin_tensor =
                   self.Var().Get<framework::LoDTensor>();
               PADDLE_ENFORCE_EQ(
                   origin_tensor.IsInitialized(), true,
                   platform::errors::InvalidArgument(
                       "Tensor %s has not been initialized!", self.Name()));

               auto *detach_tensor =
                   detach_var->MutableVar()->GetMutable<framework::LoDTensor>();
               detach_tensor->ShareDataWith(origin_tensor);
               // NOTE(liym27): Call ShareInplaceVersionCounterWith to share the
               // same TensorInplaceVersion, which is used to check whether
               // inplace
               // operations are correct.
               detach_tensor->ShareInplaceVersionCounterWith(origin_tensor);
             } else {
               const auto &origin_selected_rows =
                   self.Var().Get<framework::SelectedRows>();
               PADDLE_ENFORCE_EQ(
                   origin_selected_rows.value().IsInitialized(), true,
                   platform::errors::InvalidArgument(
                       "Tensor %s has not been initialized!", self.Name()));

               auto *detach_selected_rows =
                   detach_var->MutableVar()
                       ->GetMutable<framework::SelectedRows>();
               detach_selected_rows->set_height(origin_selected_rows.height());
               detach_selected_rows->set_rows(origin_selected_rows.rows());
               detach_selected_rows->mutable_value()->ShareDataWith(
                   origin_selected_rows.value());
               detach_selected_rows->mutable_value()
                   ->ShareInplaceVersionCounterWith(
                       origin_selected_rows.value());
             }
             VLOG(3) << "The detached Tensor(" << detach_var->Name()
                     << ") share data with " << self.Name();
             return detach_var;
           },
           py::return_value_policy::take_ownership, R"DOC(
1215

1216
        Returns a new Tensor, detached from the current graph.
Z
Zhou Wei 已提交
1217 1218
        It will share data with origin Tensor and always doesn't have a Tensor copy.
        In addition, the detached Tensor doesn't provide gradient propagation.
1219

1220
        Returns: The detached Tensor.
1221 1222 1223 1224

        Examples:
            .. code-block:: python

1225
                import paddle
Z
Zhou Wei 已提交
1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250

                x = paddle.to_tensor(1.0, stop_gradient=False)
                detach_x = x.detach()
                detach_x[:] = 10.0
                print(x)  # Tensor(shape=[1], dtype=float32, place=CPUPlace, stop_gradient=False,
                          #        [10.])
                y = x**2
                y.backward()
                print(x.grad)         # [20.0]
                print(detach_x.grad)  # None, 'stop_gradient=True' by default

                detach_x.stop_gradient = False # Set stop_gradient to be False, supported auto-grad
                z = detach_x**3
                z.backward()

                print(x.grad)         # [20.0], detach_x is detached from x's graph, not affect each other
                print(detach_x.grad)  # [300.0], detach_x has its own graph

                # Due to sharing of data with origin Tensor, There are some unsafe operations:
                y = 2 * x
                detach_x[:] = 5.0
                y.backward() 
                # It will raise Error:
                #   one of the variables needed for gradient computation has been modified by an inplace operation.
             
1251 1252 1253
       )DOC")
      .def("clear_gradient", &imperative::VarBase::ClearGradient, R"DOC(

1254
        Only for Tensor that has gradient, normally we use this for Parameters since other temporary Tensor doesen't has gradient.
1255

1256
        The Gradient of current Tensor will be set to ``0`` .
1257 1258 1259 1260 1261 1262

        Returns:  None

        Examples:
             .. code-block:: python

1263
                import paddle
Z
Zhou Wei 已提交
1264 1265 1266 1267 1268 1269 1270
                input = paddle.uniform([10, 2])
                linear = paddle.nn.Linear(2, 3)
                out = linear(input)
                out.backward()
                print("Before clear_gradient, linear.weight.grad: {}".format(linear.weight.grad))
                linear.weight.clear_gradient()
                print("After clear_gradient, linear.weight.grad: {}".format(linear.weight.grad))
1271
      )DOC")
Z
Zhou Wei 已提交
1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319
      .def("clone",
           [](std::shared_ptr<imperative::VarBase> &self) {
             const auto &tensor = self->Var().Get<framework::LoDTensor>();
             PADDLE_ENFORCE_EQ(
                 tensor.IsInitialized(), true,
                 platform::errors::InvalidArgument(
                     "%s has not been initialized", self->Name()));
             auto tracer = imperative::GetCurrentTracer();
             auto new_var = std::make_shared<imperative::VarBase>(
                 true, tracer->GenerateUniqueName(self->Name() + "_clone"));
             framework::AttributeMap attrs;
             imperative::NameVarBaseMap ins = {{"X", {self}}};
             imperative::NameVarBaseMap outs = {{"Out", {new_var}}};
             tracer->TraceOp("assign", ins, outs, attrs);
             return new_var;
           },
           py::return_value_policy::copy, R"DOC(

        Returns a new Tensor, which is clone of origin Tensor, and it remains in the current graph.
        It will always have a Tensor copy.
        Tn addition, the cloned Tensor provides gradient propagation.

        Returns: The cloned Tensor.

        Examples:
            .. code-block:: python

              import paddle

              x = paddle.to_tensor(1.0, stop_gradient=False)
              clone_x = x.clone()
              y = clone_x**2
              y.backward()
              print(clone_x.stop_gradient) # False
              print(clone_x.grad)          # [2.0], support gradient propagation
              print(x.stop_gradient)       # False
              print(x.grad)                # [2.0], clone_x support gradient propagation for x

              x = paddle.to_tensor(1.0)
              clone_x = x.clone()
              clone_x.stop_gradient = False
              z = clone_x**3
              z.backward()
              print(clone_x.stop_gradient) # False
              print(clone_x.grad)          # [3.0], support gradient propagation
              print(x.stop_gradient) # True
              print(x.grad)          # None
       )DOC")
L
Leo Chen 已提交
1320 1321 1322 1323 1324 1325
      .def("_grad_name", &imperative::VarBase::GradVarName)
      .def("_grad_value",
           [](imperative::VarBase &self) {
             return self.MutableGradVar()->Get<framework::LoDTensor>();
           },
           py::return_value_policy::reference)
1326 1327 1328 1329
      .def("_set_grad_type",
           [](imperative::VarBase &self, framework::proto::VarType::Type type) {
             self.MutableGradVarBase()->SetType(type);
           })
1330
      .def("_grad_ivar",
J
Jiabin Yang 已提交
1331 1332
           [](const imperative::VarBase &self) {
             auto &grad_var = self.GradVarBase();
1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343
             if (grad_var && grad_var->Var().IsInitialized()) {
               auto *tensor =
                   grad_var->MutableVar()->IsType<framework::LoDTensor>()
                       ? grad_var->MutableVar()
                             ->GetMutable<framework::LoDTensor>()
                       : grad_var->MutableVar()
                             ->GetMutable<framework::SelectedRows>()
                             ->mutable_value();
               if (tensor->IsInitialized()) {
                 return grad_var;
               }
J
Jiabin Yang 已提交
1344
             }
1345
             return std::shared_ptr<imperative::VarBase>(nullptr);
J
Jiabin Yang 已提交
1346 1347
           },
           py::return_value_policy::copy)
C
chentianyu03 已提交
1348 1349 1350 1351
      .def("_set_grad_ivar",
           [](imperative::VarBase &self, imperative::VarBase &grad) {
             self.SetGradVarBase(grad);
           })
1352 1353 1354 1355 1356 1357 1358 1359
      .def("_is_sparse",
           [](imperative::VarBase &self) {
             return self.Var().IsType<framework::SelectedRows>();
           })
      .def("_allreduce",
           [](imperative::VarBase &self,
              const imperative::ParallelStrategy &strategy) {
             if (strategy.nranks_ > 1) {
1360
#if defined(PADDLE_WITH_NCCL) || defined(PADDLE_WITH_RCCL)
1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377
#if NCCL_VERSION_CODE >= 2212
               imperative::AllReduce(self.Var(), self.MutableVar(), strategy);
#else
               if (!self.Var().IsType<framework::SelectedRows>()) {
                 imperative::AllReduce(self.Var(), self.MutableVar(), strategy);
               } else {
                 PADDLE_THROW(platform::errors::Unimplemented(
                     "Imperative SelectedRows allreduce is not supported when "
                     "paddle is compiled with NCCL verison lower than v2.2.12. "
                     "You can set is_sparse=False for the Layer containing "
                     "this argument, such as Embedding(is_sparse=False)."));
               }
#endif  // NCCL_VERSION_CODE
#else
               PADDLE_THROW(platform::errors::Unimplemented(
                   "Imperative allreduce is not supported when paddle is "
                   "not compiled with NCCL."));
1378
#endif  // PADDLE_WITH_NCCL or PADDLE_WITH_RCCL
1379 1380 1381
             }
           },
           py::call_guard<py::gil_scoped_release>())
1382 1383 1384
      .def("_register_grad_hook",
           [](imperative::VarBase &self, const py::handle &hook) {
             PADDLE_ENFORCE_EQ(
1385
                 !self.OverridedStopGradient() && self.HasGradVar(), true,
1386
                 platform::errors::InvalidArgument(
1387 1388 1389
                     "Cannot register gradient hook on a Tensor that stop "
                     "gradient or without gradient."));
             return self.GradVarBase()->AddVariableWrapperHook(
1390 1391 1392 1393 1394
                 std::make_shared<PyVariableWrapperHook>(hook.ptr()));
           })
      .def("_remove_grad_hook",
           [](imperative::VarBase &self, int64_t hook_id) {
             PADDLE_ENFORCE_EQ(
1395
                 !self.OverridedStopGradient() && self.HasGradVar(), true,
1396
                 platform::errors::InvalidArgument(
1397 1398 1399
                     "Cannot remove gradient hook on a Tensor that stop "
                     "gradient or without gradient."));
             return self.GradVarBase()->RemoveVariableWrapperHook(hook_id);
1400
           })
1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436
      .def("_register_backward_hook",
           [](imperative::VarBase &self, const py::handle &hook) {
             PADDLE_ENFORCE_EQ(
                 self.IsLeaf(), true,
                 platform::errors::InvalidArgument(
                     "Only can register backward hook for leaf Tensor."));
             PADDLE_ENFORCE_EQ(
                 !self.OverridedStopGradient() && self.HasGradVar(), true,
                 platform::errors::InvalidArgument(
                     "Cannot register backward hook on a Tensor that stop "
                     "gradient or without gradient."));
             auto py_func = PyObjectCast<std::function<void()>>(hook.ptr());
             self.GradVarBase()->AddVoidHook(
                 std::make_shared<std::function<void()>>(py_func));
           },
           R"DOC(
             Registers a backward hook for current Tensor.

             This hook will be called every time the gradient of current Tensor has been fully calculated.

             There are two differences with `_register_grad_hook`:
             1. This backward hook will be executed after the gradient accumulation completed across batchs,
                but the hook registered by `_register_grad_hook` will be executed the gradient accumulation
                completed in current batch.
             2. This backward hook function should have the following signature:

                  hook() -> None

                It requires no input and no return value.

             Args:
                 hook(function): A backward hook to be registered for Tensor.gradient

             Returns:
                 None
           )DOC")
1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464
      .def("cpu",
           [](const std::shared_ptr<imperative::VarBase> &self) {
             if (platform::is_cpu_place(self->Place())) {
               return self;
             } else {
               auto new_var = self->NewVarBase(platform::CPUPlace(), true);
               new_var->SetOverridedStopGradient(self->OverridedStopGradient());
               return new_var;
             }
           },
           R"DOC(
        Returns a copy of this Tensor in CPU memory.

        If this Tensor is already in CPU memory, then no copy is performed and the original Tensor is returned.

        Examples:
            .. code-block:: python

              import paddle
              x = paddle.to_tensor(1.0, place=paddle.CUDAPlace(0))
              print(x.place)    # CUDAPlace(0)
              
              y = x.cpu()
              print(y.place)    # CPUPlace

              )DOC")
      .def("pin_memory",
           [](const std::shared_ptr<imperative::VarBase> &self) {
1465
#if !defined(PADDLE_WITH_CUDA) && !defined(PADDLE_WITH_HIP)
1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496
             PADDLE_THROW(platform::errors::PermissionDenied(
                 "Cannot copy this Tensor to pinned memory in CPU version "
                 "Paddle, "
                 "Please recompile or reinstall Paddle with CUDA support."));
#endif
             if (platform::is_cuda_pinned_place(self->Place())) {
               return self;
             } else {
               auto new_var =
                   self->NewVarBase(platform::CUDAPinnedPlace(), true);
               new_var->SetOverridedStopGradient(self->OverridedStopGradient());
               return new_var;
             }
           },
           R"DOC(
        Returns a copy of this Tensor in pin memory.

        If this Tensor is already in pin memory, then no copy is performed and the original Tensor is returned.

        Examples:
            .. code-block:: python

              import paddle
              x = paddle.to_tensor(1.0, place=paddle.CUDAPlace(0))
              print(x.place)      # CUDAPlace(0)

              y = x.pin_memory()
              print(y.place)      # CUDAPinnedPlace

      )DOC")
      .def("cuda",
1497 1498
           [](const std::shared_ptr<imperative::VarBase> &self,
              py::handle &handle, bool blocking) {
1499
#if !defined(PADDLE_WITH_CUDA) && !defined(PADDLE_WITH_HIP)
1500 1501 1502 1503 1504
             PADDLE_THROW(platform::errors::PermissionDenied(
                 "Cannot copy this Tensor to GPU in CPU version Paddle, "
                 "Please recompile or reinstall Paddle with CUDA support."));
#else
             int device_count = platform::GetCUDADeviceCount();
1505 1506
             int device_id = 0;
             if (handle == py::none()) {
1507 1508 1509
               if (platform::is_gpu_place(self->Place())) {
                 return self;
               }
1510 1511 1512 1513 1514 1515 1516
             } else {
               PyObject *py_obj = handle.ptr();
               PADDLE_ENFORCE_EQ(
                   PyCheckInteger(py_obj), true,
                   platform::errors::InvalidArgument(
                       " 'device_id' must be a positive integer"));
               device_id = py::cast<int>(handle);
1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539
             }
             PADDLE_ENFORCE_GE(
                 device_id, 0,
                 platform::errors::InvalidArgument(
                     "Can not copy Tensor to Invalid CUDAPlace(%d), device id "
                     "must inside [0, %d)",
                     device_id, device_count));
             PADDLE_ENFORCE_LT(
                 device_id, device_count,
                 platform::errors::InvalidArgument(
                     "Can not copy Tensor to Invalid CUDAPlace(%d), device id "
                     "must inside [0, %d)",
                     device_id, device_count));
             platform::CUDAPlace place = platform::CUDAPlace(device_id);
             if (platform::is_same_place(self->Place(), place)) {
               return self;
             } else {
               auto new_var = self->NewVarBase(place, blocking);
               new_var->SetOverridedStopGradient(self->OverridedStopGradient());
               return new_var;
             }
#endif
           },
1540
           py::arg("device_id") = py::none(), py::arg("blocking") = true, R"DOC(
1541 1542 1543 1544 1545 1546
        Returns a copy of this Tensor in GPU memory.

        If this Tensor is already in GPU memory and device_id is default, 
        then no copy is performed and the original Tensor is returned.
        
        Args:
1547
            device_id(int, optional): The destination GPU device id. Default: None, means current device.
1548 1549 1550 1551 1552 1553
            blocking(bool, optional): If False and the source is in pinned memory, the copy will be 
              asynchronous with respect to the host. Otherwise, the argument has no effect. Default: False.

        Examples:
            .. code-block:: python

1554
              # required: gpu
1555 1556 1557 1558 1559 1560
              import paddle
              x = paddle.to_tensor(1.0, place=paddle.CPUPlace())
              print(x.place)        # CPUPlace

              y = x.cuda()
              print(y.place)        # CUDAPlace(0)
1561 1562 1563
            
              y = x.cuda(None)
              print(y.place)        # CUDAPlace(0)
1564 1565 1566 1567

              y = x.cuda(1)
              print(y.place)        # CUDAPlace(1)
       )DOC")
K
Kaipeng Deng 已提交
1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596
      .def("_share_memory",
           [](const std::shared_ptr<imperative::VarBase> &self) {
#ifndef _WIN32
             PADDLE_ENFORCE_EQ(
                 platform::is_cpu_place(self->Place()), true,
                 platform::errors::InvalidArgument(
                     "Sharing memory only support CPU Tensor currently"));
             // 1. get LoDTensor
             auto *t = self->MutableVar()->GetMutable<framework::LoDTensor>();
             // 2. allocate shared memory
             void *data_ptr = t->data<void>();
             size_t data_size = t->numel() * framework::SizeOfType(t->type());
             auto shared_writer_holder =
                 memory::allocation::AllocateMemoryMapWriterAllocation(
                     data_size);
             // 3. maintain mmap fd set & backup ipc_name
             const std::string &ipc_name = shared_writer_holder->ipc_name();
             memory::allocation::MemoryMapFdSet::Instance().Insert(ipc_name);
             // 4. copy data & reset holder
             memory::Copy(platform::CPUPlace(), shared_writer_holder->ptr(),
                          platform::CPUPlace(), data_ptr, data_size);
             t->ResetHolder(shared_writer_holder);
             return *t;
#else
             PADDLE_THROW(platform::errors::PermissionDenied(
                 "Sharing memory in Windows OS is not supported currently"));
#endif
           },
           py::return_value_policy::reference)
1597
      .def("copy_", &imperative::VarBase::CopyFrom)
1598
      .def("_copy_to",
1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614
           [](const std::shared_ptr<imperative::VarBase> &self,
              const platform::CPUPlace &place, bool blocking) {
             auto new_var = self->NewVarBase(place, blocking);
             // Note(zhiqiu): Since NewVarBase may use GpuCopyAsync to
             // copy data from the tensor of self to the tensor of new varbase,
             // we need to ensure that the varbase self is not destructed until
             // the GpuCopyAsync is completed. Otherwise, the memory may be
             // freed
             // when varbase self is destructed.
             // To do that, we increase the reference count of self by 1 and
             // add a cuda event to wait the GpuCopyAsync's completion.
             if (!blocking) {
               IncreaseVarbaseReferenceCountUntilCopyComplete(self, place);
             }
             return new_var;
           },
J
Jiabin Yang 已提交
1615
           py::return_value_policy::copy)
1616
      .def("_copy_to",
1617 1618 1619 1620 1621 1622 1623 1624
           [](const std::shared_ptr<imperative::VarBase> &self,
              const platform::CUDAPinnedPlace &place, bool blocking) {
             auto new_var = self->NewVarBase(place, blocking);
             if (!blocking) {
               IncreaseVarbaseReferenceCountUntilCopyComplete(self, place);
             }
             return new_var;
           },
1625
           py::return_value_policy::copy)
1626
      .def("_copy_to",
1627 1628 1629 1630 1631 1632 1633 1634
           [](const std::shared_ptr<imperative::VarBase> &self,
              const platform::XPUPlace &place, bool blocking) {
             auto new_var = self->NewVarBase(place, blocking);
             if (!blocking) {
               IncreaseVarbaseReferenceCountUntilCopyComplete(self, place);
             }
             return new_var;
           },
1635
           py::return_value_policy::copy)
1636
      .def("_copy_to",
1637 1638 1639 1640 1641 1642 1643 1644
           [](const std::shared_ptr<imperative::VarBase> &self,
              const platform::CUDAPlace &place, bool blocking) {
             auto new_var = self->NewVarBase(place, blocking);
             if (!blocking) {
               IncreaseVarbaseReferenceCountUntilCopyComplete(self, place);
             }
             return new_var;
           },
J
Jiabin Yang 已提交
1645
           py::return_value_policy::copy)
1646 1647 1648 1649 1650 1651 1652 1653 1654 1655
      .def("_copy_to",
           [](const std::shared_ptr<imperative::VarBase> &self,
              const platform::NPUPlace &place, bool blocking) {
             auto new_var = self->NewVarBase(place, blocking);
             if (!blocking) {
               IncreaseVarbaseReferenceCountUntilCopyComplete(self, place);
             }
             return new_var;
           },
           py::return_value_policy::copy)
C
chentianyu03 已提交
1656 1657 1658 1659 1660 1661 1662 1663 1664 1665
      .def("_copy_to",
           [](const std::shared_ptr<imperative::VarBase> &self,
              const platform::Place &place, bool blocking) {
             auto new_var = self->NewVarBase(place, blocking);
             if (!blocking) {
               IncreaseVarbaseReferenceCountUntilCopyComplete(self, place);
             }
             return new_var;
           },
           py::return_value_policy::copy)
J
Jiabin Yang 已提交
1666
      .def("value", [](imperative::VarBase &self) { return self.MutableVar(); },
1667 1668 1669
           py::return_value_policy::reference)
      .def_property("name", &imperative::VarBase::Name,
                    &imperative::VarBase::SetName)
L
Leo Chen 已提交
1670 1671 1672 1673 1674
      .def_property("stop_gradient",
                    &imperative::VarBase::OverridedStopGradient,
                    &imperative::VarBase::SetOverridedStopGradient)
      .def_property("persistable", &imperative::VarBase::Persistable,
                    &imperative::VarBase::SetPersistable)
1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690
      .def_property_readonly(
          "shape",
          [](imperative::VarBase &self) {
            if (self.Var().IsType<framework::LoDTensor>()) {
              return framework::vectorize<int>(
                  self.Var().Get<framework::LoDTensor>().dims());
            } else if (self.Var().IsType<framework::SelectedRows>()) {
              return framework::vectorize<int>(
                  self.Var().Get<framework::SelectedRows>().value().dims());
            } else {
              VLOG(2) << "It is meaningless to get shape of "
                         "variable type "
                      << GetTypeName(self);
              return std::vector<int>();
            }
          })
1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719
      .def_property_readonly("is_leaf", &imperative::VarBase::IsLeaf,
                             R"DOC(
      Whether a Tensor is leaf Tensor.

      For the Tensor whose stop_gradient is ``True`` , it will be leaf Tensor. 
      
      For the Tensor whose stop_gradient is ``False`` , it will be leaf Tensor too if it is created by user.

      Returns:
          bool: Whether a Tensor is leaf Tensor.

      Examples:
          .. code-block:: python

              import paddle

              x = paddle.to_tensor(1.)
              print(x.is_leaf) # True

              x = paddle.to_tensor(1., stop_gradient=True)
              y = x + 1
              print(x.is_leaf) # True
              print(y.is_leaf) # True

              x = paddle.to_tensor(1., stop_gradient=False)
              y = x + 1
              print(x.is_leaf) # True
              print(y.is_leaf) # False
       )DOC")
1720 1721 1722
      .def_property_readonly(
          "place", [](imperative::VarBase &self) { return self.Place(); },
          py::return_value_policy::copy)
1723 1724 1725 1726 1727 1728
      .def_property_readonly("_place_str",
                             [](imperative::VarBase &self) {
                               std::stringstream ostr;
                               ostr << self.Place();
                               return ostr.str();
                             })
J
Jiabin Yang 已提交
1729
      .def_property_readonly("type", &imperative::VarBase::Type)
L
Leo Chen 已提交
1730
      .def_property_readonly("dtype", &imperative::VarBase::DataType);
1731 1732 1733

  py::class_<imperative::Layer, Layer /* <--- trampoline*/> layer(m, "Layer");
  layer.def(py::init<>())
1734 1735 1736 1737 1738
      .def("forward",
           [](imperative::Layer &self,
              const std::vector<std::shared_ptr<imperative::VarBase>> &inputs) {
             return self.Forward(inputs);
           });
1739

1740 1741 1742 1743 1744
  py::class_<imperative::jit::ProgramDescTracer>(m, "ProgramDescTracer", "")
      .def("create_program_desc",
           &imperative::jit::ProgramDescTracer::CreateProgramDesc)
      .def("reset", &imperative::jit::ProgramDescTracer::Reset);

1745
  py::class_<imperative::Tracer, std::shared_ptr<imperative::Tracer>>(
1746
      m, "Tracer", R"DOC()DOC")
1747
      .def("__init__",
J
Jiabin Yang 已提交
1748
           [](imperative::Tracer &self) { new (&self) imperative::Tracer(); })
1749 1750 1751
      .def_property("_enable_program_desc_tracing",
                    &imperative::Tracer::IsProgramDescTracingEnabled,
                    &imperative::Tracer::SetEnableProgramDescTracing)
1752 1753
      .def_property("_enable_autocast", &imperative::Tracer::IsAutoCastEnabled,
                    &imperative::Tracer::SetEnableAutoCast)
1754
      .def_property("_has_grad", &imperative::Tracer::HasGrad,
1755
                    &imperative::Tracer::SetHasGrad)
1756 1757 1758 1759 1760 1761 1762 1763
      .def_property(
          "_expected_place",
          [](const imperative::Tracer &self) -> py::object {
            return py::cast(self.ExpectedPlace());
          },
          [](imperative::Tracer &self, const py::object &obj) {
            if (py::isinstance<platform::CUDAPlace>(obj)) {
              auto p = obj.cast<platform::CUDAPlace *>();
L
Leo Chen 已提交
1764
              self.SetExpectedPlace(*p);
1765 1766
              VLOG(4) << "Tracer(" << &self << ")"
                      << " set expected place " << *p;
1767 1768 1769
            } else if (py::isinstance<platform::XPUPlace>(obj)) {
              auto p = obj.cast<platform::XPUPlace *>();
              self.SetExpectedPlace(*p);
1770 1771
              VLOG(4) << "Tracer(" << &self << ")"
                      << " set expected place " << *p;
1772 1773
            } else if (py::isinstance<platform::CPUPlace>(obj)) {
              auto p = obj.cast<platform::CPUPlace *>();
L
Leo Chen 已提交
1774
              self.SetExpectedPlace(*p);
1775 1776
              VLOG(4) << "Tracer(" << &self << ")"
                      << " set expected place " << *p;
1777 1778
            } else if (py::isinstance<platform::CUDAPinnedPlace>(obj)) {
              auto p = obj.cast<platform::CUDAPinnedPlace *>();
L
Leo Chen 已提交
1779
              self.SetExpectedPlace(*p);
1780 1781
              VLOG(4) << "Tracer(" << &self << ")"
                      << " set expected place " << *p;
1782 1783 1784 1785 1786
            } else if (py::isinstance<platform::NPUPlace>(obj)) {
              auto p = obj.cast<platform::NPUPlace *>();
              self.SetExpectedPlace(*p);
              VLOG(4) << "Tracer(" << &self << ")"
                      << " set expected place " << *p;
1787 1788 1789 1790 1791
            } else if (py::isinstance<platform::Place>(obj)) {
              auto p = obj.cast<platform::Place *>();
              self.SetExpectedPlace(*p);
              VLOG(4) << "Tracer(" << &self << ")"
                      << " set expected place " << *p;
1792
            } else {
L
Leo Chen 已提交
1793
              PADDLE_THROW(platform::errors::InvalidArgument(
1794
                  "Incompatible Place Type: supports XPUPlace, CUDAPlace, "
1795
                  "CPUPlace, NPUPlace"
L
Leo Chen 已提交
1796 1797
                  "and CUDAPinnedPlace, "
                  "but got Unknown Type!"));
1798 1799
            }
          })
1800 1801 1802
      .def("_get_program_desc_tracer",
           &imperative::Tracer::GetProgramDescTracer,
           py::return_value_policy::reference)
1803
      .def("_generate_unique_name", &imperative::Tracer::GenerateUniqueName,
1804
           py::arg("key") = "dygraph_tmp")
1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820
      .def("_set_amp_op_list",
           [](imperative::Tracer &self,
              std::unordered_set<std::string> &allow_ops,
              std::unordered_set<std::string> &block_ops) {
             // NOTE(zhiqiu): The automatic conversion in pybind11 between
             // c++
             // STL and python set/list/dict involve a copy operation that
             // prevents pass-by-reference semantics, so it is ok to swap.
             // The reaseon why not directly pass
             // std::shared_ptr<std::unordered_set<std::string>>
             // is that pybind11 forbid shared_ptr<T> where T is not custom
             // type.
             imperative::AmpOperators::Instance().GetMutableAllowOps()->swap(
                 allow_ops);
             imperative::AmpOperators::Instance().GetMutableBlockOps()->swap(
                 block_ops);
1821
             VLOG(5) << "AMP operators changed, "
1822 1823
                     << imperative::AmpOperators::Instance();
           })
1824 1825 1826
      .def("_get_amp_op_list",
           [](imperative::Tracer &self) {
             return std::make_tuple(
1827 1828
                 *(imperative::AmpOperators::Instance().GetMutableAllowOps()),
                 *(imperative::AmpOperators::Instance().GetMutableBlockOps()));
1829
           })
1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842
      .def("trace",
           [](imperative::Tracer &self, const std::string &type,
              const PyNameVarBaseMap &ins, const PyNameVarBaseMap &outs,
              framework::AttributeMap attrs, const platform::XPUPlace &place,
              bool trace_backward) {
             auto ins_map = ConvertToNameVarBaseMap(ins);
             auto outs_map = ConvertToNameVarBaseMap(outs);
             {
               py::gil_scoped_release release;
               self.TraceOp(type, std::move(ins_map), std::move(outs_map),
                            std::move(attrs), place, trace_backward);
             }
           })
M
minqiyang 已提交
1843
      .def("trace",
J
Jiabin Yang 已提交
1844 1845 1846 1847 1848 1849
           [](imperative::Tracer &self, const std::string &type,
              const PyNameVarBaseMap &ins, const PyNameVarBaseMap &outs,
              framework::AttributeMap attrs, const platform::CUDAPlace &place,
              bool trace_backward) {
             auto ins_map = ConvertToNameVarBaseMap(ins);
             auto outs_map = ConvertToNameVarBaseMap(outs);
1850 1851
             {
               py::gil_scoped_release release;
J
Jiabin Yang 已提交
1852 1853
               self.TraceOp(type, std::move(ins_map), std::move(outs_map),
                            std::move(attrs), place, trace_backward);
1854
             }
M
minqiyang 已提交
1855
           })
1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868
      .def("trace",
           [](imperative::Tracer &self, const std::string &type,
              const PyNameVarBaseMap &ins, const PyNameVarBaseMap &outs,
              framework::AttributeMap attrs, const platform::NPUPlace &place,
              bool trace_backward) {
             auto ins_map = ConvertToNameVarBaseMap(ins);
             auto outs_map = ConvertToNameVarBaseMap(outs);
             {
               py::gil_scoped_release release;
               self.TraceOp(type, std::move(ins_map), std::move(outs_map),
                            std::move(attrs), place, trace_backward);
             }
           })
J
Jiabin Yang 已提交
1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881
      .def("trace",
           [](imperative::Tracer &self, const std::string &type,
              const PyNameVarBaseMap &ins, const PyNameVarBaseMap &outs,
              framework::AttributeMap attrs, const platform::CPUPlace &place,
              bool trace_backward) {
             auto ins_map = ConvertToNameVarBaseMap(ins);
             auto outs_map = ConvertToNameVarBaseMap(outs);
             {
               py::gil_scoped_release release;
               self.TraceOp(type, std::move(ins_map), std::move(outs_map),
                            std::move(attrs), place, trace_backward);
             }
           });
1882 1883

  // define parallel context
1884 1885 1886
  py::class_<imperative::ParallelStrategy> parallel_strategy(
      m, "ParallelStrategy", "");
  parallel_strategy.def(py::init())
1887 1888
      .def_property(
          "nranks",
1889 1890
          [](const imperative::ParallelStrategy &self) { return self.nranks_; },
          [](imperative::ParallelStrategy &self, int nranks) {
1891 1892 1893
            self.nranks_ = nranks;
          })
      .def_property("local_rank",
1894
                    [](const imperative::ParallelStrategy &self) {
1895 1896
                      return self.local_rank_;
                    },
1897
                    [](imperative::ParallelStrategy &self, int local_rank) {
1898 1899 1900 1901
                      self.local_rank_ = local_rank;
                    })
      .def_property(
          "trainer_endpoints",
1902
          [](const imperative::ParallelStrategy &self) {
1903 1904
            return self.trainer_endpoints_;
          },
1905
          [](imperative::ParallelStrategy &self, std::vector<std::string> eps) {
1906 1907 1908
            self.trainer_endpoints_ = eps;
          })
      .def_property("current_endpoint",
1909
                    [](const imperative::ParallelStrategy &self) {
1910 1911
                      return self.current_endpoint_;
                    },
1912
                    [](imperative::ParallelStrategy &self,
1913 1914 1915 1916 1917 1918 1919
                       const std::string &ep) { self.current_endpoint_ = ep; })
      .def_property(
          "nrings",
          [](const imperative::ParallelStrategy &self) { return self.nrings_; },
          [](imperative::ParallelStrategy &self, int nrings) {
            self.nrings_ = nrings;
          });
1920

1921 1922 1923 1924
  m.def("varbase_copy", &VarBaseCopy<platform::Place>);
  m.def("varbase_copy", &VarBaseCopy<platform::CPUPlace>);
  m.def("varbase_copy", &VarBaseCopy<platform::CUDAPlace>);
  m.def("varbase_copy", &VarBaseCopy<platform::XPUPlace>);
1925
  m.def("varbase_copy", &VarBaseCopy<platform::CUDAPinnedPlace>);
1926
  m.def("varbase_copy", &VarBaseCopy<platform::NPUPlace>);
1927

1928 1929 1930 1931 1932 1933 1934
  m.def(
      "dygraph_partial_grad",
      [](const std::vector<std::shared_ptr<imperative::VarBase>> &input_targets,
         const std::vector<std::shared_ptr<imperative::VarBase>>
             &output_targets,
         const std::vector<std::shared_ptr<imperative::VarBase>> &output_grads,
         const std::vector<std::shared_ptr<imperative::VarBase>> &no_grad_vars,
1935 1936
         const platform::Place &place, bool create_graph, bool retain_graph,
         bool allow_unused, bool only_inputs) {
Z
Zeng Jinle 已提交
1937 1938
        imperative::PartialGradEngine engine(
            input_targets, output_targets, output_grads, no_grad_vars, place,
1939
            create_graph, retain_graph, allow_unused, only_inputs);
1940 1941 1942 1943 1944
        engine.Execute();
        return engine.GetResult();
      },
      py::call_guard<py::gil_scoped_release>());

1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957
  m.def(
      "dygraph_run_backward",
      [](const std::vector<std::shared_ptr<imperative::VarBase>> &tensors,
         const std::vector<std::shared_ptr<imperative::VarBase>> &grad_tensors,
         bool retain_graph, const imperative::Tracer &tracer) {
        auto *engine = tracer.GetEngine();
        engine->Init(tensors, grad_tensors, retain_graph);
        VLOG(3) << "Start backward";
        engine->Execute();
        VLOG(3) << "Finish backward";
      },
      py::call_guard<py::gil_scoped_release>());

1958 1959
#if defined(PADDLE_WITH_NCCL) || defined(PADDLE_WITH_RCCL) || \
    defined(PADDLE_WITH_XPU_BKCL)
1960 1961 1962 1963 1964 1965
  py::class_<imperative::ParallelContext,
             std::shared_ptr<imperative::ParallelContext>>(m,
                                                           "ParallelContext");

  py::class_<imperative::Reducer, std::shared_ptr<imperative::Reducer>>(
      m, "Reducer", R"DOC()DOC")
S
ShenLiang 已提交
1966 1967 1968 1969 1970
      .def(py::init<const std::vector<std::shared_ptr<imperative::VarBase>> &,
                    const std::vector<std::vector<size_t>> &,
                    const std::vector<bool> &,
                    std::shared_ptr<imperative::ParallelContext>,
                    const std::vector<size_t> &, bool>())
1971
      .def("prepare_for_backward", &imperative::Reducer::PrepareForBackward,
1972
           py::arg("vars"), py::call_guard<py::gil_scoped_release>());
1973 1974 1975 1976

  m.def("assign_group_by_size", &imperative::AssignGroupBySize, py::arg("vars"),
        py::arg("is_sparse_gradient"),
        py::arg("group_size_limits") = std::vector<size_t>{25 * 1024 * 1024},
1977
        py::arg("tensor_indices") = std::vector<int64_t>{},
1978
        py::call_guard<py::gil_scoped_release>());
1979
#endif
1980

1981
#if defined(PADDLE_WITH_NCCL) || defined(PADDLE_WITH_RCCL)
1982 1983 1984 1985 1986
  py::class_<imperative::NCCLParallelContext, imperative::ParallelContext,
             std::shared_ptr<imperative::NCCLParallelContext>>(
      m, "NCCLParallelContext")
      .def(py::init<const imperative::ParallelStrategy &,
                    const platform::CUDAPlace &>())
K
kuizhiqing 已提交
1987 1988 1989 1990
      .def("init", [](imperative::NCCLParallelContext &self) { self.Init(); })
      .def("init_with_ring_id",
           &imperative::NCCLParallelContext::InitWithRingID,
           py::arg("ring_id"));
1991 1992 1993 1994 1995 1996 1997 1998
#endif

#if defined(PADDLE_WITH_XPU_BKCL)
  py::class_<imperative::BKCLParallelContext, imperative::ParallelContext,
             std::shared_ptr<imperative::BKCLParallelContext>>(
      m, "BKCLParallelContext")
      .def(py::init<const imperative::ParallelStrategy &,
                    const platform::XPUPlace &>())
K
kuizhiqing 已提交
1999 2000 2001 2002
      .def("init", [](imperative::BKCLParallelContext &self) { self.Init(); })
      .def("init_with_ring_id",
           &imperative::BKCLParallelContext::InitWithRingID,
           py::arg("ring_id"));
2003
#endif
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026
  m.def("pylayer_apply",
        [](const platform::CPUPlace &place, const py::object &cls,
           const py::args args, const py::kwargs kwargs) {
          return imperative::PyLayerApply(place, cls, args, kwargs);
        });

  m.def("pylayer_apply",
        [](const platform::CUDAPlace &place, const py::object &cls,
           const py::args args, const py::kwargs kwargs) {
          return imperative::PyLayerApply(place, cls, args, kwargs);
        });

  m.def("pylayer_apply",
        [](const platform::XPUPlace &place, const py::object &cls,
           const py::args args, const py::kwargs kwargs) {
          return imperative::PyLayerApply(place, cls, args, kwargs);
        });

  m.def("pylayer_apply",
        [](const platform::CUDAPinnedPlace &place, const py::object &cls,
           const py::args args, const py::kwargs kwargs) {
          return imperative::PyLayerApply(place, cls, args, kwargs);
        });
2027 2028 2029 2030 2031 2032

  m.def("pylayer_apply",
        [](const platform::NPUPlace &place, const py::object &cls,
           const py::args args, const py::kwargs kwargs) {
          return imperative::PyLayerApply(place, cls, args, kwargs);
        });
2033 2034 2035 2036
}

}  // namespace pybind
}  // namespace paddle