memory_optimization_transpiler.py 15.6 KB
Newer Older
1
#   Copyright (c) 2018 PaddlePaddle Authors. All Rights Reserved.
D
dzhwinter 已提交
2
#
D
dzhwinter 已提交
3 4 5
# 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
D
dzhwinter 已提交
6
#
D
dzhwinter 已提交
7
#     http://www.apache.org/licenses/LICENSE-2.0
D
dzhwinter 已提交
8
#
D
dzhwinter 已提交
9 10 11 12 13 14
# 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.

15
from collections import defaultdict
16 17 18
from .. import core
from ..framework import Program, default_main_program, Parameter, Variable
from ..backward import _rename_arg_
19 20

dtype_to_size = {
21 22 23 24 25 26
    core.VarDesc.VarType.FP16: 2,
    core.VarDesc.VarType.FP32: 4,
    core.VarDesc.VarType.FP64: 8,
    core.VarDesc.VarType.INT16: 2,
    core.VarDesc.VarType.INT32: 4,
    core.VarDesc.VarType.INT64: 8,
27 28
    core.VarDesc.VarType.BOOL: 1,
    core.VarDesc.VarType.UINT8: 1,
29
}
30

31
SUB_BLOCK_OPS = [
32 33 34
    "while", "while_grad", "parallel_do", "parallel_do_grad",
    "conditional_block", "conditional_block_grad"
]
35

36 37 38
SUB_BLOCK_PAIR = [("while", "while_grad"), ("parallel_do", "parallel_do_grad"),
                  ("conditional_block", "conditional_block_grad")]

Q
qiaolongfei 已提交
39 40
PRINT_LOG = False

41 42

class ControlFlowGraph(object):
43 44
    def __init__(self, program, ops, forward_num, skip_opt):
        self._program = program
45 46 47 48
        self._ops = ops
        self._forward_num = forward_num
        self._successors = defaultdict(set)
        self._presuccessors = defaultdict(set)
49 50 51 52
        self._uses = defaultdict(set)
        self._defs = defaultdict(set)
        self._live_in = defaultdict(set)
        self._live_out = defaultdict(set)
53
        self._skip_opt = skip_opt
54 55

    def _add_connections(self, connections):
56
        """Populates _successors and _presuccessors for two neighbor nodes."""
57 58 59 60
        for node1, node2 in connections:
            self._add(node1, node2)

    def _add(self, node1, node2):
61 62
        self._successors[node1].add(node2)
        self._presuccessors[node2].add(node1)
63

64 65
    # TODO(panyx0718): We need to have a unified way of building intermediate
    # representation.
66
    def _build_graph(self):
67 68
        """Build a graph based on op sequence.
        """
69
        self.op_size = len(self._ops)
70 71 72
        op_node_connections = [(i, i + 1) for i in range(self.op_size - 1)]
        self._add_connections(op_node_connections)
        for i in range(self.op_size):
73 74
            self._uses[i].update(self._ops[i].input_arg_names())
            self._defs[i].update(self._ops[i].output_arg_names())
75

76 77 78 79 80 81 82 83 84 85 86 87 88 89 90
    def _update_graph(self, old_name, new_name, begin_idx=0):
        for i in range(begin_idx, self.op_size):
            if old_name in self._uses[i]:
                self._uses[i].remove(old_name)
                self._uses[i].add(new_name)
            if old_name in self._defs[i]:
                self._defs[i].remove(old_name)
                self._defs[i].add(new_name)
            if old_name in self._live_in[i]:
                self._live_in[i].remove(old_name)
                self._live_out[i].add(new_name)
            if old_name in self._live_out[i]:
                self._live_out[i].remove(old_name)
                self._live_out[i].add(new_name)

91
    def _reach_fixed_point(self, live_in, live_out):
92
        """Check if the liveness set has stablized."""
93 94 95 96 97
        if len(live_in) != len(self._live_in):
            return False
        if len(live_out) != len(self._live_out):
            return False
        for i in range(self.op_size):
98 99
            if (live_in[i] != self._live_in[i] or
                    live_out[i] != self._live_out[i]):
100 101 102 103 104 105 106
                return False
        return True

    def _dataflow_analyze(self):
        self._build_graph()
        live_in = defaultdict(set)
        live_out = defaultdict(set)
107 108
        # Repeatedly apply liveness updates until the algorithm stablize
        # on a complete set live input vars and live output vars.
109
        while True:
110
            for i in reversed(range(self.op_size)):
111 112
                live_in[i] = set(self._live_in[i])
                live_out[i] = set(self._live_out[i])
113
                for s in self._successors[i]:
114
                    self._live_out[i] |= self._live_in[s]
Q
QI JUN 已提交
115 116
                self._live_in[i] = self._uses[i] | (
                    self._live_out[i] - self._defs[i])
117 118 119 120 121 122 123
            if self._reach_fixed_point(live_in, live_out):
                break

    def _get_diff(self, a, b):
        u = a & b
        return a - u, b - u

124 125 126 127 128 129 130 131 132 133 134 135
    def _has_var(self, block_desc, var_name, is_forward):
        if is_forward:
            return block_desc.has_var(str(var_name))
        else:
            return block_desc.has_var_recursive(str(var_name))

    def _find_var(self, block_desc, var_name, is_forward):
        if is_forward:
            return block_desc.find_var(str(var_name))
        else:
            return block_desc.find_var_recursive(str(var_name))

136 137 138 139 140 141 142 143 144 145 146 147 148 149 150
    def _check_var_validity(self, block_desc, x, is_forward):
        if str(x) == "@EMPTY@":
            return False
        if not self._has_var(block_desc, x, is_forward):
            return False
        if self._find_var(block_desc, x, is_forward).persistable():
            return False
        if self._find_var(block_desc, x,
                          is_forward).type() != core.VarDesc.VarType.LOD_TENSOR:
            return False
        if x in self._skip_opt:
            return False
        if not self._find_var(block_desc, x, is_forward).shape():
            return False
        return True
151

152 153
    # TODO(panyx0718): This needs to be less hacky. It seems memory optimization
    # doesn't consider vars copied between cpu and gpu.
154 155 156 157 158 159
    def _update_skip_opt_set(self):
        for i in range(self.op_size):
            op = self._ops[i]
            if op.type() == "fill_constant" and op.attr("force_cpu") == True:
                self._skip_opt.update(op.output_arg_names())

160
    def release_memory(self, skip_opt_set=None):
161
        self._dataflow_analyze()
162
        self._update_skip_opt_set()
163 164
        if skip_opt_set:
            self._skip_opt.update(skip_opt_set)
165 166 167 168
        fwd_id = 0
        bwd_id = 0
        for i in range(self.op_size):
            op = self._ops[i]
169
            if op.type() in SUB_BLOCK_OPS:
170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187
                continue
            block_desc = op.block()
            is_forward = i < self._forward_num
            in_diff, out_diff = self._get_diff(self._live_in[i],
                                               self._live_out[i])
            can_optimize = filter(
                lambda x: self._check_var_validity(block_desc, x, is_forward),
                in_diff)
            if can_optimize:
                index = i + fwd_id + 1 if is_forward else i - self._forward_num + bwd_id + 1
                delete_op = block_desc.insert_op(index)
                delete_op.set_type("delete_var")
                delete_op.set_input("X", can_optimize)
                if is_forward:
                    fwd_id += 1
                else:
                    bwd_id += 1

188
    def memory_optimize(self, skip_opt_set=None, level=0):
189 190 191
        def compare_shape(x_shape, cache_shape, opt_level):
            if opt_level == 0:
                return x_shape == cache_shape
192
            elif opt_level == 1:
193 194 195 196 197 198
                if (x_shape[0] == -1) ^ (cache_shape[0] == -1):
                    return False
                x_size = abs(reduce(lambda x, y: x * y, x_shape))
                cache_size = abs(reduce(lambda x, y: x * y, cache_shape))
                if x_size <= cache_size:
                    return True
199 200
            else:
                raise ValueError("only support opt_level 0 or 1.")
201 202 203 204
            return False

        self._dataflow_analyze()
        self._update_skip_opt_set()
205 206 207
        # update skip set to meet users' demand
        if skip_opt_set:
            self._skip_opt.update(skip_opt_set)
208 209
        self.pool = []
        for i in range(self.op_size):
210
            op = self._ops[i]
211
            if op.type() in SUB_BLOCK_OPS:
212 213 214
                continue
            block_desc = op.block()
            is_forward = i < self._forward_num
215
            if self.pool:
216
                defs_can_optimize = filter(
217
                    lambda x: self._check_var_validity(block_desc, x, is_forward),
218 219 220 221 222
                    self._defs[i])
                out_pair = [
                    (x, self._find_var(block_desc, x, is_forward).shape())
                    for x in defs_can_optimize
                ]
223
                for x, x_shape in out_pair:
224 225 226
                    # If x is both in uses and defs, it can not be optimized!
                    if x in self._uses[i]:
                        continue
227 228 229
                    for index, cache_pair in enumerate(self.pool):
                        cache_var = cache_pair[0]
                        cache_shape = cache_pair[1]
230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263
                        if not compare_shape(x_shape, cache_shape, level):
                            continue

                        if not self._has_var(block_desc, cache_var, is_forward):
                            continue

                        x_dtype = self._find_var(block_desc, x,
                                                 is_forward).dtype()
                        cache_dtype = self._find_var(block_desc, cache_var,
                                                     is_forward).dtype()
                        # TODO(qijun): actually, we should compare
                        # dtype_to_size[x_dtype] and dtype_to_size[cache_dtype]
                        if x_dtype != cache_dtype:
                            continue

                        if PRINT_LOG:
                            print(("Hit Cache !!!! cache pool index "
                                   "is %d, var name is %s, "
                                   "cached var name is %s, "
                                   "var shape is %s ") % (index, x, cache_var,
                                                          str(cache_shape)))
                        self.pool.pop(index)
                        if x == cache_var:
                            break
                        # Rename the var to the cache var already with
                        # memory allocated in order to reuse the memory.
                        _rename_arg_(self._ops, x, cache_var, begin_idx=i)
                        self._program.block(block_desc.id).var(str(
                            x)).desc = self._find_var(block_desc, cache_var,
                                                      is_forward)
                        self._update_graph(x, cache_var, begin_idx=i)
                        break

            in_diff, _ = self._get_diff(self._live_in[i], self._live_out[i])
264
            can_optimize = filter(
265
                lambda x: self._check_var_validity(block_desc, x, is_forward),
266 267 268
                in_diff)
            if can_optimize:
                for var_name in can_optimize:
269 270 271 272
                    self.pool.append((var_name, self._find_var(
                        block_desc, var_name, is_forward).shape()))


273
def _process_sub_block_pair(pdesc, sub_block_pair):
274 275 276 277 278 279 280 281 282 283 284 285 286
    """Creates a list of tuple each of which tracks info of a subblock.

      Note: this function doesn't handle nested subblocks yet.
      TODO(panyx0718): assert if case nested subblocks happen.

    :param pdesc: ProgramDesc.
    :param sub_block_pair: A list op pairs. Each op pair is the forward
        op and backward op. The ops in the list are special that they contain
        a subblock of ops.
    :return: A list of tuples, each tuple is (all ops in a subblock pair
        including forward and backward, number of forward ops,
        all output args names of the ops in the subblock pairs).
    """
287 288 289
    ops_list = []
    block_desc = pdesc.block(0)
    op_size = block_desc.op_size()
290 291 292 293 294 295 296 297 298 299 300 301 302
    for fwd_op, bwd_op in sub_block_pair:
        sub_block_ids = []
        grad_sub_block_ids = []
        sub_block_id_pair = []
        sub_op_dict = {}
        for i in range(op_size):
            op = block_desc.op(i)
            if op.type() == fwd_op:
                sub_block_ids.append(op.attr("sub_block").id)
                sub_op_dict[op.attr("sub_block").id] = op
            elif op.type() == bwd_op:
                grad_sub_block_ids.append(op.attr("sub_block").id)
                sub_op_dict[op.attr("sub_block").id] = op
303

304 305
        # Find fwd_op/bwd_op block pair
        for grad_id in grad_sub_block_ids:
Q
qijun 已提交
306 307 308 309
            fwd_id = pdesc.block(grad_id).get_forward_block_idx()
            if fwd_id in sub_block_ids:
                sub_block_id_pair.append((fwd_id, grad_id))
                sub_block_ids.remove(fwd_id)
310

311
        # Get fwd_op/bwd_op block ops
Q
qijun 已提交
312
        for fwd_id, grad_id in sub_block_id_pair:
313
            sub_block_ops = []
Q
qijun 已提交
314
            sub_block = pdesc.block(fwd_id)
315 316 317
            block_op_size = sub_block.op_size()
            for i in range(block_op_size):
                sub_block_ops.append(sub_block.op(i))
318

319 320 321 322
            grad_sub_block = pdesc.block(grad_id)
            grad_sub_block_op_size = grad_sub_block.op_size()
            for i in range(grad_sub_block_op_size):
                sub_block_ops.append(grad_sub_block.op(i))
323

324
            sub_op_output = set()
Q
qijun 已提交
325
            sub_op_output.update(sub_op_dict[fwd_id].output_arg_names())
326 327
            sub_op_output.update(sub_op_dict[grad_id].output_arg_names())
            ops_list.append((sub_block_ops, block_op_size, sub_op_output))
328

329
        # Process rest fwd_op block ops
Q
qijun 已提交
330
        for fwd_id in sub_block_ids:
331
            sub_block_ops = []
Q
qijun 已提交
332
            sub_block = pdesc.block(fwd_id)
333 334 335 336
            sub_block_op_size = sub_block.op_size()
            for i in range(sub_block_op_size):
                sub_block_ops.append(sub_block.op(i))
            sub_op_output = set()
Q
qijun 已提交
337
            sub_op_output.update(sub_op_dict[fwd_id].output_arg_names())
338 339
            ops_list.append((sub_block_ops, sub_block_op_size, sub_op_output))
    return ops_list
340

341

342
def _get_cfgs(input_program):
343 344 345 346 347
    """Process each block and create ControlFlowGraph for each of them.

    :param input_program: Program object.
    :return: A list of ControlFlowGraph, each corresponds to a block.
    """
348 349 350 351 352 353 354
    ops_list = []
    pdesc = input_program.get_desc()
    block_desc = pdesc.block(0)
    op_size = block_desc.op_size()
    # Get global block ops
    ops_list.append(
        ([block_desc.op(i) for i in range(op_size)], op_size, set()))
355

356 357
    # Only process one level of nested subblock.
    ops_list.extend(_process_sub_block_pair(pdesc, SUB_BLOCK_PAIR))
358

359 360 361 362
    cfgs = [
        ControlFlowGraph(input_program, ops, forward_num, skip_opt)
        for ops, forward_num, skip_opt in ops_list
    ]
363
    return cfgs
364 365


366
def memory_optimize(input_program, skip_opt_set=None, print_log=False, level=0):
367 368 369 370 371 372 373 374 375 376 377
    """Optimize memory by reusing var memory.

      Note: it doesn't not support subblock nested in subblock.

    :param input_program: Input Program
    :param print_log: whether to print debug log.
    :param level: If level=0, reuse if the shape is completely equal, o
    :return:
    """
    if level != 0 and level != 1:
        raise ValueError("only support opt_level 0 or 1.")
Q
qiaolongfei 已提交
378 379
    global PRINT_LOG
    PRINT_LOG = print_log
380
    cfgs = _get_cfgs(input_program)
381
    for cfg in cfgs:
382
        cfg.memory_optimize(skip_opt_set=skip_opt_set, level=level)
383 384


385
def release_memory(input_program, skip_opt_set=None):
Y
yuyang18 已提交
386 387 388 389 390 391 392 393 394 395
    """
    Modify the input program and insert :code:`delete_op` to early drop not used
    variables. The modification will be performed inplace.

    Notes: This is an experimental API and could be removed in next few
    releases. Users should not use this API.

    Args:
        input_program(Program): The program will be inserted :code:`delete_op`.
    """
396 397
    cfgs = _get_cfgs(input_program)
    for cfg in cfgs:
398
        cfg.release_memory(skip_opt_set=skip_opt_set)