bug fixed logs print copy

zhangmeng

2020-01-17 f7c4a3cfd07adede3308f8d9d3d7315427d90a7c

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
#pragma once
 
#include "rebatching_queue.h"
 
namespace caffe2 {
 
using RebatchingQueuePtr = std::unique_ptr<RebatchingQueue>;
 
class CreateRebatchingQueueOp : public Operator<CPUContext> {
 public:
  CreateRebatchingQueueOp(const OperatorDef& operator_def, Workspace* ws)
      : Operator(operator_def, ws) {}
 
  bool RunOnDevice() override {
    *OperatorBase::Output<RebatchingQueuePtr>(0) =
        RebatchingQueuePtr(new RebatchingQueue(
            OperatorBase::GetSingleArgument<int>("capacity", 1),
            OperatorBase::GetSingleArgument<int>("num_blobs", 1)));
    return true;
  }
};
 
class EnqueueRebatchingQueueOp : public Operator<CPUContext> {
 public:
  EnqueueRebatchingQueueOp(const OperatorDef& operator_def, Workspace* ws)
      : Operator(operator_def, ws),
        enqueueBatch_(
            OperatorBase::GetSingleArgument<bool>("enqueue_batch", false)) {}
  bool RunOnDevice() override {
    auto& queue = Inputs()[0]->template Get<RebatchingQueuePtr>();
    CHECK(queue);
    CAFFE_ENFORCE_EQ(InputSize(), queue->numBlobs() + 1);
    std::vector<const Tensor*> inputTensors;
    inputTensors.reserve(InputSize() - 1);
    for (int i = 1; i < InputSize(); ++i) {
      inputTensors.push_back(&Input(i));
    }
 
    return enqueueBatch_ ? queue->enqueueMany(context_, inputTensors)
                         : queue->enqueueOne(context_, inputTensors);
  }
 
 private:
  const bool enqueueBatch_;
};
 
class DequeueRebatchingQueueOp : public Operator<CPUContext> {
 public:
  DequeueRebatchingQueueOp(const OperatorDef& operator_def, Workspace* ws)
      : Operator(operator_def, ws),
        numElements_(OperatorBase::GetSingleArgument<int>("num_elements", 1)) {}
 
  bool RunOnDevice() override {
    auto& queue = Inputs()[0]->template Get<RebatchingQueuePtr>();
    CHECK(queue);
 
    std::vector<Tensor*> outputTensors;
    outputTensors.reserve(OutputSize());
    for (int i = 0; i < OutputSize(); ++i) {
      outputTensors.push_back(Output(i));
    }
 
    return queue->dequeue(context_, numElements_, outputTensors);
  }
 
 private:
  int numElements_;
};
 
class CloseRebatchingQueueOp : public Operator<CPUContext> {
 public:
  CloseRebatchingQueueOp(const OperatorDef& operator_def, Workspace* ws)
      : Operator(operator_def, ws) {}
 
  bool RunOnDevice() override {
    CAFFE_ENFORCE_EQ(InputSize(), 1);
    auto& queue = Inputs()[0]->template Get<RebatchingQueuePtr>();
    CAFFE_ENFORCE(queue);
    queue->close();
    return true;
  }
};
} // caffe2