caffe.proto是caffe数据结构定义的主要文件，本文主要是在caffe.proto代码的基础上加上了部分中文注释，其中的内容与caffe的prototxt文件中的结构相对应。
代码语言：javascript复制// syntax用来指定protobuf的版本
syntax = "proto2";
// package可以看作C 中的namespace，与Caffe C 代码中的namespace caffe对应
// package用来避免名称冲突
package caffe;
// 在消息定义中，每个字段都有唯一的一个数字标识符。这些标识符是用来在消息的二进制格式中识别各个字段的，一旦开始使用就不能够再改变。
// 注：[1,15]之内的标识号在编码的时候会占用一个字节。[16,2047]之内的标识号则占用2个字节。所以应该为那些频繁出现的消息元素保留 [1,15]之内的标识号。
// required：一个格式良好的消息一定要含有一个这种字段，表示该值是必须要设置的。
// optional：消息格式中该字段可以有0个或1个值（不超过1个）。
// repeated：在一个格式良好的消息中，这种字段可以重复任意多次（包括0次）。重复的值的顺序会被保留，表示该值可以重复，相当于Java中的List。
// Specifies the shape (dimensions) of a Blob.
// 指定Blob的shape，4-D shape
message BlobShape {
//数据块形状定义为Num * Channel * Height * Wight, 原因在于caffe基于容器的多维嵌套来实现高维数据的封装, 即vector。
repeated int64 dim = 1 [packed = true];
}
// Blob数据块，包括Blob shape，数据和微分
message BlobProto {
// Blob的shape, 即numpy中的shape
optional BlobShape shape = 7;
// Blob的数据部分
repeated float data = 5 [packed = true];
// Blob的微分部分
repeated float diff = 6 [packed = true];
// Blob中的数据部分(double类型)
repeated double double_data = 8 [packed = true];
// Blob的微分部分(double类型)
repeated double double_diff = 9 [packed = true];
// 4D dimensions -- deprecated. Use "shape" instead.
// Blob的4个维度，已被Blob shape代替
// Blob中数据的个数(例如卷积核的个数)
optional int32 num = 1 [default = 0];
// Blob中数据的通道数
optional int32 channels = 2 [default = 0];
// Blob中数据的高度
optional int32 height = 3 [default = 0];
// Blob中数据的宽度
optional int32 width = 4 [default = 0];
}
// The BlobProtoVector is simply a way to pass multiple blobproto instances
// around.
// BlobProtoVector, 用来保存多个BlobProb对象的Vector
message BlobProtoVector {
repeated BlobProto blobs = 1;
}
//图像数据, channel-图像通道数, height-高度, width-宽度, data-图像像素数据, label-图像标签, float_data-图像浮点型数据(0-1之间), encoded-图像编码方式
message Datum {
// 图像的通道数
optional int32 channels = 1;
// 图像的高度
optional int32 height = 2;
// 图像的宽度
optional int32 width = 3;
// the actual image data, in bytes
// 实际的图像数据，以字节形式(uint8)表示
optional bytes data = 4;
// 图像对应的标签，必须为整形
optional int32 label = 5;
// Optionally, the datum could also hold float data.
// 可选表示，图像数据表示为float数据，即0-255归一化到0-1之间
repeated float float_data = 6;
// If true data contains an encoded image that need to be decoded
// encoded为true表示图像采用压缩表示，需要解码
optional bool encoded = 7 [default = false];
}
// Filler参数, filler主要对网络权重进行初始化
// Filler类型分为常量初始化（constant）、高斯分布初始化（gaussian）、positive_unitball初始化、均匀分布初始化（uniform）、xavier初始化、msra初始化、双线性初始化（bilinear）
message FillerParameter {
// The filler type.
// Filler的类型
optional string type = 1 [default = 'constant'];
// 常量初始化的值
optional float value = 2 [default = 0]; // the value in constant filler
// 均匀分布初始化中的最小值
optional float min = 3 [default = 0]; // the min value in uniform filler
// 均匀分布初始化中的最大值
optional float max = 4 [default = 1]; // the max value in uniform filler
// 高斯分布初始化中的均值
optional float mean = 5 [default = 0]; // the mean value in Gaussian filler
// 高斯分布初始化中的标准差
optional float std = 6 [default = 1]; // the std value in Gaussian filler
// The expected number of non-zero output weights for a given input in
// Gaussian filler -- the default -1 means don't perform sparsification.
// 在高斯分布初始化中给定输入及权重，期望输出非0值，默认值-1表示不进行稀疏化
optional int32 sparse = 7 [default = -1];
// Normalize the filler variance by fan_in, fan_out, or their average.
// Applies to 'xavier' and 'msra' fillers.
// 通过fan_in, fan_out或average来归一化filler方差，主要应用到'xavier'和'msra' filler中
enum VarianceNorm {
FAN_IN = 0;
FAN_OUT = 1;
AVERAGE = 2;
}
// 定义filler方差归一化，默认为FAN_IN
optional VarianceNorm variance_norm = 8 [default = FAN_IN];
}
//神经网络参数
message NetParameter {
// 神经网络名字
optional string name = 1; // consider giving the network a name
// DEPRECATED. See InputParameter. The input blobs to the network.
// 已废弃。网络的输入部分，具体请看InputParameter。
repeated string input = 3;
// DEPRECATED. See InputParameter. The shape of the input blobs.
// 已废弃。输入blob的shape，具体请看InputParameter。
repeated BlobShape input_shape = 8;
// 4D input dimensions -- deprecated. Use "input_shape" instead.
// If specified, for each input blob there should be four
// values specifying the num, channels, height and width of the input blob.
// Thus, there should be a total of (4 * #input) numbers.
// 已废弃。用input_shape代替。
repeated int32 input_dim = 4;
// Whether the network will force every layer to carry out backward operation.
// If set False, then whether to carry out backward is determined
// automatically according to the net structure and learning rates.
// 网络中是否每一层都执行反向传播的标志，如果设为false，反向传播会根据网络结构和学习率自动进行。
optional bool force_backward = 5 [default = false];
// The current "state" of the network, including the phase, level, and stage.
// Some layers may be included/excluded depending on this state and the states
// specified in the layers' include and exclude fields.
// 网络的当前状态，包括phase, level和stage，(phase应该是对应prototxt文件中的TRAIN,TEST)
// 某些层是否included/excluded依赖于层中include，exclue字段指定的state。
optional NetState state = 6;
// Print debugging information about results while running Net::Forward,
// Net::Backward, and Net::Update.
// 在执行Net::Forward,Net::Backward, Net::Update时是否打印调试信息。
optional bool debug_info = 7 [default = false];
// The layers that make up the net. Each of their configurations, including
// connectivity and behavior, is specified as a LayerParameter.
// 构成网络的layer，每一个layer的配置，包括连接性和行为都在LayerParameter中指定。
repeated LayerParameter layer = 100; // ID 100 so layers are printed last.
// DEPRECATED: use 'layer' instead.
// 已废弃，用layer代替。
repeated V1LayerParameter layers = 2;
}
// NOTE
// Update the next available ID when you add a new SolverParameter field.
// 注意：当你添加一个新的SolverParameter字段时，要更新下一个可获取的ID
// SolverParameter next available ID: 41 (last added: type)
// Solver参数
message SolverParameter {
//////////////////////////////////////////////////////////////////////////////
// Specifying the train and test networks
//
// Exactly one train net must be specified using one of the following fields:
// train_net_param, train_net, net_param, net
// One or more test nets may be specified using any of the following fields:
// test_net_param, test_net, net_param, net
// If more than one test net field is specified (e.g., both net and
// test_net are specified), they will be evaluated in the field order given
// above: (1) test_net_param, (2) test_net, (3) net_param/net.
// A test_iter must be specified for each test_net.
// A test_level and/or a test_stage may also be specified for each test_net.
//////////////////////////////////////////////////////////////////////////////
// Proto filename for the train net, possibly combined with one or more test nets.
// 训练网络的prototxt文件名，可能结合一个或多个测试网络
optional string net = 24;
// Inline train net param, possibly combined with one or more test nets.
// 内联训练网络参数，可能结合一个或多个测试网络
optional NetParameter net_param = 25;
// 训练网络的proto文件名
optional string train_net = 1; // Proto filename for the train net.
// 测试网络的proto文件名
repeated string test_net = 2; // Proto filenames for the test nets.
// 内联训练网络参数
optional NetParameter train_net_param = 21; // Inline train net params.
// 内联测试网络参数
repeated NetParameter test_net_param = 22; // Inline test net params.
// The states for the train/test nets. Must be unspecified or specified once per net.
// By default, all states will have solver = true;
// train_state will have phase = TRAIN,
// and all test_state's will have phase = TEST.
// Other defaults are set according to the NetState defaults.
// train/test网络的状态，必须不指定或每个网络指定一次
// 默认情况下，所有的状态都有solver = true，train_state的phase = TRAIN，其它默认情况根据NetState默认值设定。
// train网络的状态，必须不指定或每个网络指定一次
optional NetState train_state = 26;
// test网络的状态，必须不指定或每个网络指定一次
repeated NetState test_state = 27;
// The number of iterations for each test net.
// 每个测试网络的迭代次数，即测试数据的迭代次数，测试数据总数=测试迭代次数*测试数据的batch_size。
repeated int32 test_iter = 3;
// The number of iterations between two testing phases.
// 两次测试间隔的迭代次数，即训练数据迭代多少次进行一次测试。
optional int32 test_interval = 4 [default = 0];
// 测试数据的loss，默认情况下不计算
optional bool test_compute_loss = 19 [default = false];
// If true, run an initial test pass before the first iteration,
// ensuring memory availability and printing the starting value of the loss.
// 如果为true，在第一次迭代之前进行一次初始测试，从而确保内存可用性并输出初始损失值。
optional bool test_initialization = 32 [default = true];
// 基本学习率
optional float base_lr = 5; // The base learning rate
// the number of iterations between displaying info. If display = 0, no info will be displayed.
// 执行多少次迭代显示一次信息，如果display = 0，不输出信息。
optional int32 display = 6;
// Display the loss averaged over the last average_loss iterations
// 输出的平均损失是之前多少次迭代的平均损失。
optional int32 average_loss = 33 [default = 1];
// 训练的最大迭代次数
optional int32 max_iter = 7; // the maximum number of iterations
// accumulate gradients over `iter_size` x `batch_size` instances
// 累积`iter_size` x `batch_size`个实例的梯度
optional int32 iter_size = 36 [default = 1];
// The learning rate decay policy. The currently implemented learning rate
// policies are as follows:
// - fixed: always return base_lr.
// - step: return base_lr * gamma ^ (floor(iter / step))
// - exp: return base_lr * gamma ^ iter
// - inv: return base_lr * (1 gamma * iter) ^ (- power)
// - multistep: similar to step but it allows non uniform steps defined by
// stepvalue
// - poly: the effective learning rate follows a polynomial decay, to be
// zero by the max_iter. return base_lr (1 - iter/max_iter) ^ (power)
// - sigmoid: the effective learning rate follows a sigmod decay
// return base_lr ( 1/(1 exp(-gamma * (iter - stepsize))))
//
// where base_lr, max_iter, gamma, step, stepvalue and power are defined
// in the solver parameter protocol buffer, and iter is the current iteration.
// 学习率的变化策略
optional string lr_policy = 8;
// 学习率的计算参数
optional float gamma = 9; // The parameter to compute the learning rate.
// 学习率的计算参数
optional float power = 10; // The parameter to compute the learning rate.
// 动量参数
optional float momentum = 11; // The momentum value.
// 权重衰减，权重衰减主要影响神经网络的正则项，具体可参考Caffe文档
optional float weight_decay = 12; // The weight decay.
// regularization types supported: L1 and L2, controlled by weight_decay
// 正则化类型支持L1和L2，受weight_decay控制。
optional string regularization_type = 29 [default = "L2"];
// the stepsize for learning rate policy "step"
// 学习率方案为step时的参数
optional int32 stepsize = 13;
// the stepsize for learning rate policy "multistep"
// 学习率方案为multistep时的参数
repeated int32 stepvalue = 34;
// Set clip_gradients to >= 0 to clip parameter gradients to that L2 norm,
// whenever their actual L2 norm is larger.
// 设置clip_gradients >= 0可以削减L2范数的梯度，当真实L2范数的梯度大于clip_gradients，将L2范数的梯度设为clip_gradients
optional float clip_gradients = 35 [default = -1];
// snapshot的间隔，即迭代多少次保存一次snapshot
optional int32 snapshot = 14 [default = 0]; // The snapshot interval
// snapshot的前缀
optional string snapshot_prefix = 15; // The prefix for the snapshot.
// whether to snapshot diff in the results or not. Snapshotting diff will help
// debugging but the final protocol buffer size will be much larger.
// 是否在结果中保存snapshot的差分，snapshot diff有助于调试，但snapshot的文件会更大。
optional bool snapshot_diff = 16 [default = false];
// snapshot的保存格式（hdf5,binaryproto）。
enum SnapshotFormat {
HDF5 = 0;
BINARYPROTO = 1;
}
// snapshot默认保存为BINARYPROTO。
optional SnapshotFormat snapshot_format = 37 [default = BINARYPROTO];
// the mode solver will use: 0 for CPU and 1 for GPU. Use GPU in default.
// 求解神经网络的方式，0 CPU, 1 GPU。默认使用GPU
enum SolverMode {
CPU = 0;
GPU = 1;
}
// 求解神经网络的模式，0 CPU, 1 GPU。默认使用GPU
optional SolverMode solver_mode = 17 [default = GPU];
// the device_id will that be used in GPU mode. Use device_id = 0 in default.
// device_id是GPU模式下GPU的ID。
optional int32 device_id = 18 [default = 0];
// If non-negative, the seed with which the Solver will initialize the Caffe
// random number generator -- useful for reproducible results. Otherwise,
// (and by default) initialize using a seed derived from the system clock.
// 如果是非负值，seed用来初始化Caffe的随机数生成器，对于再见结果是很有用的，默认情况下，seed的是从系统时钟获取。
optional int64 random_seed = 20 [default = -1];
// type of the solver
// 神经网络求解的类型, 默认为SGD
optional string type = 40 [default = "SGD"];
// numerical stability for RMSProp, AdaGrad and AdaDelta and Adam
// RMSProp, AdaGrad, AdaDelta, Adam求解类型的参数
optional float delta = 31 [default = 1e-8];
// parameters for the Adam solver
// Adam求解类型的参数
optional float momentum2 = 39 [default = 0.999];
// RMSProp decay value
// MeanSquare(t) = rms_decay*MeanSquare(t-1) (1-rms_decay)*SquareGradient(t)
// RMSProp类型的衰减值
optional float rms_decay = 38 [default = 0.99];
// If true, print information about the state of the net that may help with
// debugging learning problems.
// 如果设为true，会输出网络的状态信息，有助于调试
optional bool debug_info = 23 [default = false];
// If false, don't save a snapshot after training finishes.
// 如果设为false，不保存训练结束的snapshot。
optional bool snapshot_after_train = 28 [default = true];
// DEPRECATED: old solver enum types, use string instead
// 已废弃，使用string代替
enum SolverType {
SGD = 0;
NESTEROV = 1;
ADAGRAD = 2;
RMSPROP = 3;
ADADELTA = 4;
ADAM = 5;
}
// DEPRECATED: use type instead of solver_type
// 已废弃：使用type代替
optional SolverType solver_type = 30 [default = SGD];
}
// A message that stores the solver snapshots
// 保存solver snapshots
message SolverState {
// 当前的迭代次数
optional int32 iter = 1; // The current iteration
// 保存学习到的网络
optional string learned_net = 2; // The file that stores the learned net.
// sgd的求解历史
repeated BlobProto history = 3; // The history for sgd solvers
// 学习的当前step
optional int32 current_step = 4 [default = 0]; // The current step for learning rate
}
// 定义phase
enum Phase {
TRAIN = 0;
TEST = 1;
}
// 网络状态
message NetState {
// 属于哪个phase
optional Phase phase = 1 [default = TEST];
optional int32 level = 2 [default = 0];
repeated string stage = 3;
}
// 网络状态分类
message NetStateRule {
// Set phase to require the NetState have a particular phase (TRAIN or TEST)
// to meet this rule.
// 设置phase
optional Phase phase = 1;
// Set the minimum and/or maximum levels in which the layer should be used.
// Leave undefined to meet the rule regardless of level.
// 设置layer的level
optional int32 min_level = 2;
optional int32 max_level = 3;
// Customizable sets of stages to include or exclude.
// The net must have ALL of the specified stages and NONE of the specified
// "not_stage"s to meet the rule.
// (Use multiple NetStateRules to specify conjunctions of stages.)
// 可定制的stage集合
repeated string stage = 4;
repeated string not_stage = 5;
}
// Specifies training parameters (multipliers on global learning constants,
// and the name and other settings used for weight sharing).
// 指定训练参数及名称以及权重共享的其它设置
message ParamSpec {
// The names of the parameter blobs -- useful for sharing parameters among
// layers, but never required otherwise. To share a parameter between two
// layers, give it a (non-empty) name.
// 两个layer之间进行参数共享的blob名字
optional string name = 1;
// Whether to require shared weights to have the same shape, or just the same
// count -- defaults to STRICT if unspecified.
// 参数共享时是否需要具有相同的shape，默认情况下需要有相同的shape
optional DimCheckMode share_mode = 2;
// 参数共享时的维度检查
enum DimCheckMode {
// STRICT (default) requires that num, channels, height, width each match.
STRICT = 0;
// PERMISSIVE requires only the count (num*channels*height*width) to match.
PERMISSIVE = 1;
}
// The multiplier on the global learning rate for this parameter.
// 学习率参数, learning rate = base_lr * lr_mult
optional float lr_mult = 3 [default = 1.0];
// The multiplier on the global weight decay for this parameter.
// 权重衰减参数, weight = weight_decay * decay_mult
optional float decay_mult = 4 [default = 1.0];
}
// NOTE
// Update the next available ID when you add a new LayerParameter field.
// LayerParameter next available layer-specific ID: 147 (last added: recurrent_param)
// 注意：当你添加一个新的LayerParameter字段时，要更新下一个可获取的ID
message LayerParameter {
// layer名称
optional string name = 1; // the layer name
// layer类型
optional string type = 2; // the layer type
// layer的输入
repeated string bottom = 3; // the name of each bottom blob
// layer的输出
repeated string top = 4; // the name of each top blob
// The train / test phase for computation.
// layer用在train/test phase
optional Phase phase = 10;
// The amount of weight to assign each top blob in the objective.
// Each layer assigns a default value, usually of either 0 or 1,
// to each top blob.
// layer对最终的loss损失值的贡献率
repeated float loss_weight = 5;
// Specifies training parameters (multipliers on global learning constants,
// and the name and other settings used for weight sharing).
// 指定训练参数
repeated ParamSpec param = 6;
// The blobs containing the numeric parameters of the layer.
// layer的blobs
repeated BlobProto blobs = 7;
// Specifies whether to backpropagate to each bottom. If unspecified,
// Caffe will automatically infer whether each input needs backpropagation
// to compute parameter gradients. If set to true for some inputs,
// backpropagation to those inputs is forced; if set false for some inputs,
// backpropagation to those inputs is skipped.
//
// The size must be either 0 or equal to the number of bottoms.
// 指定反向传播是否传播到每一个bottom，如果不指定，caffe会自动检查推断是否每一个输入都需要反向传播来计算梯度。如果一些输入设为true,
// 则这些layer强制进行反向传播，如果设为false，这些layer将跳过反向传播。
repeated bool propagate_down = 11;
// Rules controlling whether and when a layer is included in the network,
// based on the current NetState. You may specify a non-zero number of rules
// to include OR exclude, but not both. If no include or exclude rules are
// specified, the layer is always included. If the current NetState meets
// ANY (i.e., one or more) of the specified rules, the layer is
// included/excluded.
// 控制layer included/excluded
repeated NetStateRule include = 8;
repeated NetStateRule exclude = 9;
// Parameters for data pre-processing.
// 数据预处理参数
optional TransformationParameter transform_param = 100;
// Parameters shared by loss layers.
// loss layer的参数共享
optional LossParameter loss_param = 101;
// Layer type-specific parameters.
//
// Note: certain layers may have more than one computational engine
// for their implementation. These layers include an Engine type and
// engine parameter for selecting the implementation.
// The default for the engine is set by the ENGINE switch at compile-time.
// 特定layer的参数
optional AccuracyParameter accuracy_param = 102;
optional ArgMaxParameter argmax_param = 103;
optional BatchNormParameter batch_norm_param = 139;
optional BiasParameter bias_param = 141;
optional ConcatParameter concat_param = 104;
optional ContrastiveLossParameter contrastive_loss_param = 105;
optional ConvolutionParameter convolution_param = 106;
optional CropParameter crop_param = 144;
optional DataParameter data_param = 107;
optional DropoutParameter dropout_param = 108;
optional DummyDataParameter dummy_data_param = 109;
optional EltwiseParameter eltwise_param = 110;
optional ELUParameter elu_param = 140;
optional EmbedParameter embed_param = 137;
optional ExpParameter exp_param = 111;
optional FlattenParameter flatten_param = 135;
optional HDF5DataParameter hdf5_data_param = 112;
optional HDF5OutputParameter hdf5_output_param = 113;
optional HingeLossParameter hinge_loss_param = 114;
optional ImageDataParameter image_data_param = 115;
optional InfogainLossParameter infogain_loss_param = 116;
optional InnerProductParameter inner_product_param = 117;
optional InputParameter input_param = 143;
optional LogParameter log_param = 134;
optional LRNParameter lrn_param = 118;
optional MemoryDataParameter memory_data_param = 119;
optional MVNParameter mvn_param = 120;
optional ParameterParameter parameter_param = 145;
optional PoolingParameter pooling_param = 121;
optional PowerParameter power_param = 122;
optional PReLUParameter prelu_param = 131;
optional PythonParameter python_param = 130;
optional RecurrentParameter recurrent_param = 146;
optional ReductionParameter reduction_param = 136;
optional ReLUParameter relu_param = 123;
optional ReshapeParameter reshape_param = 133;
optional ScaleParameter scale_param = 142;
optional SigmoidParameter sigmoid_param = 124;
optional SoftmaxParameter softmax_param = 125;
optional SPPParameter spp_param = 132;
optional SliceParameter slice_param = 126;
optional TanHParameter tanh_param = 127;
optional ThresholdParameter threshold_param = 128;
optional TileParameter tile_param = 138;
optional WindowDataParameter window_data_param = 129;
}
// Message that stores parameters used to apply transformation to the data layer's data
// 用来进行数据层（图像）变换的参数
message TransformationParameter {
// For data pre-processing, we can do simple scaling and subtracting the
// data mean, if provided. Note that the mean subtraction is always carried
// out before scaling.
// 像素归一化，归一化之前会减去均值
optional float scale = 1 [default = 1];
// Specify if we want to randomly mirror data.
// 图像进行随机mirror操作
optional bool mirror = 2 [default = false];
// Specify if we would like to randomly crop an image.
// 图像随机crop操作
optional uint32 crop_size = 3 [default = 0];
// mean_file and mean_value cannot be specified at the same time
// 图像的均值文件
optional string mean_file = 4;
// if specified can be repeated once (would subtract it from all the channels)
// or can be repeated the same number of times as channels
// (would subtract them from the corresponding channel)
// 图像的均值，手动指定，通常是三个
repeated float mean_value = 5;
// Force the decoded image to have 3 color channels.
// 强制图像必须有三个颜色通道
optional bool force_color = 6 [default = false];
// Force the decoded image to have 1 color channels.
// 强制图像为灰度图像
optional bool force_gray = 7 [default = false];
}
// Message that stores parameters shared by loss layers
// loss层参数
message LossParameter {
// If specified, ignore instances with the given label.
// 如果指定，则label等于ignore_label的样本将不参与Loss计算，并且反向传播时梯度直接置0。
optional int32 ignore_label = 1;
// How to normalize the loss for loss layers that aggregate across batches,
// spatial dimensions, or other dimensions. Currently only implemented in
// SoftmaxWithLoss and SigmoidCrossEntropyLoss layers.
// 指定loss归一化的方式
enum NormalizationMode {
// Divide by the number of examples in the batch times spatial dimensions.
// Outputs that receive the ignore label will NOT be ignored in computing
// the normalization factor.
// 所有样本都参与计算，包括ignore label
FULL = 0;
// Divide by the total number of output locations that do not take the
// ignore_label. If ignore_label is not set, this behaves like FULL.
// 所有样本都参与计算，不包括ignore label
VALID = 1;
// Divide by the batch size.
// 除以给定的batch size。
BATCH_SIZE = 2;
// Do not normalize the loss.
// 不归一化loss
NONE = 3;
}
// For historical reasons, the default normalization for
// SigmoidCrossEntropyLoss is BATCH_SIZE and *not* VALID.
// loss归一化方式
optional NormalizationMode normalization = 3 [default = VALID];
// Deprecated. Ignored if normalization is specified. If normalization
// is not specified, then setting this to false will be equivalent to
// normalization = BATCH_SIZE to be consistent with previous behavior.
// 已废弃。Loss会除以参与计算的样本总数；否则Loss等于直接求和
optional bool normalize = 2;
}
// Messages that store parameters used by individual layer types follow, in
// alphabetical order.
// accuracy层参数
message AccuracyParameter {
// When computing accuracy, count as correct by comparing the true label to
// the top k scoring classes. By default, only compare to the top scoring
// class (i.e. argmax).
// 计算前top-k的准确率，默认计算top-1准确率
optional uint32 top_k = 1 [default = 1];
// The "label" axis of the prediction blob, whose argmax corresponds to the
// predicted label -- may be negative to index from the end (e.g., -1 for the
// last axis). For example, if axis == 1 and the predictions are
// (N x C x H x W), the label blob is expected to contain N*H*W ground truth
// labels with integer values in {0, 1, ..., C-1}.
// 指定在哪个维度上计算label
optional int32 axis = 2 [default = 1];
// If specified, ignore instances with the given label.
// 如果指定，则忽略给定标签的实例
optional int32 ignore_label = 3;
}
// 标签最大化参数，标签最大化即确定概率最大的label
message ArgMaxParameter {
// If true produce pairs (argmax, maxval)
// 如果为真，则生成(argmax, maxval)
optional bool out_max_val = 1 [default = false];
// 类别的top-k
optional uint32 top_k = 2 [default = 1];
// The axis along which to maximise -- may be negative to index from the
// end (e.g., -1 for the last axis).
// By default ArgMaxLayer maximizes over the flattened trailing dimensions
// for each index of the first / num dimension.
// 根据axis进行标签最大化
optional int32 axis = 3;
}
// 参数拼接，在deconv的prototxt文件中见过
message ConcatParameter {
// The axis along which to concatenate -- may be negative to index from the
// end (e.g., -1 for the last axis). Other axes must have the
// same dimension for all the bottom blobs.
// By default, ConcatLayer concatenates blobs along the "channels" axis (1).
// 参数拼接时的维度，按axis进行拼接
optional int32 axis = 2 [default = 1];
// DEPRECATED: alias for "axis" -- does not support negative indexing.
// 已废弃。与axis一样。
optional uint32 concat_dim = 1 [default = 1];
}
// batch norm层的相关参数, batch norm layer通常配与scale layer一起使用，具体用法可参考Resnet结构
message BatchNormParameter {
// If false, accumulate global mean/variance values via a moving average. If
// true, use those accumulated values instead of computing mean/variance
// across the batch.
// 如果设为false，累计全部的mean/variance，如果为true，使用累计值代替batch上mean/variance的计算
// true是使用了caffe内部的均值和方差，false是使用了每个Batch里的数据的均值和方差
optional bool use_global_stats = 1;
// How much does the moving average decay each iteration?
// 每次迭代平均值衰减比例
optional float moving_average_fraction = 2 [default = .999];
// Small value to add to the variance estimate so that we don't divide by
// zero.
// variance估计时为了使除数不为0，需要加上eps
optional float eps = 3 [default = 1e-5];
}
// bias层参数，没找到实际的应用例子
message BiasParameter {
// The first axis of bottom[0] (the first input Blob) along which to apply
// bottom[1] (the second input Blob). May be negative to index from the end
// (e.g., -1 for the last axis).
//
// For example, if bottom[0] is 4D with shape 100x3x40x60, the output
// top[0] will have the same shape, and bottom[1] may have any of the
// following shapes (for the given value of axis):
// (axis == 0 == -4) 100; 100x3; 100x3x40; 100x3x40x60
// (axis == 1 == -3) 3; 3x40; 3x40x60
// (axis == 2 == -2) 40; 40x60
// (axis == 3 == -1) 60
// Furthermore, bottom[1] may have the empty shape (regardless of the value of
// "axis") -- a scalar bias.
optional int32 axis = 1 [default = 1];
// (num_axes is ignored unless just one bottom is given and the bias is
// a learned parameter of the layer. Otherwise, num_axes is determined by the
// number of axes by the second bottom.)
// The number of axes of the input (bottom[0]) covered by the bias
// parameter, or -1 to cover all axes of bottom[0] starting from `axis`.
// Set num_axes := 0, to add a zero-axis Blob: a scalar.
optional int32 num_axes = 2 [default = 1];
// (filler is ignored unless just one bottom is given and the bias is
// a learned parameter of the layer.)
// The initialization for the learned bias parameter.
// Default is the zero (0) initialization, resulting in the BiasLayer
// initially performing the identity operation.
optional FillerParameter filler = 3;
}
// 对比损失层，siamese network中使用了对比损失
message ContrastiveLossParameter {
// margin for dissimilar pair
// 不相似的样本对的距离保持在margin以上
optional float margin = 1 [default = 1.0];
// The first implementation of this cost did not exactly match the cost of
// Hadsell et al 2006 -- using (margin - d^2) instead of (margin - d)^2.
// legacy_version = false (the default) uses (margin - d)^2 as proposed in the
// Hadsell paper. New models should probably use this version.
// legacy_version = true uses (margin - d^2). This is kept to support /
// reproduce existing models and results
// 第一版对比损失没有完全按论文写，如果为false，则按照论文原来的公式计算
optional bool legacy_version = 2 [default = false];
}
// 卷积层参数
message ConvolutionParameter {
// 输出数据的个数
optional uint32 num_output = 1; // The number of outputs for the layer
// 是否有偏置项
optional bool bias_term = 2 [default = true]; // whether to have bias terms
// Pad, kernel size, and stride are all given as a single value for equal
// dimensions in all spatial dimensions, or once per spatial dimension.
// 卷积padding的大小
repeated uint32 pad = 3; // The padding size; defaults to 0
// 卷积核的大小
repeated uint32 kernel_size = 4; // The kernel size
// 卷积的步长
repeated uint32 stride = 6; // The stride; defaults to 1
// Factor used to dilate the kernel, (implicitly) zero-filling the resulting
// holes. (Kernel dilation is sometimes referred to by its use in the
// algorithme à trous from Holschneider et al. 1987.)
// 卷积膨胀，在卷积的时候可以skip一定长度的像素
repeated uint32 dilation = 18; // The dilation; defaults to 1
// For 2D convolution only, the *_h and *_w versions may also be used to
// specify both spatial dimensions.
// padding, kernel, stride的宽度和高度
optional uint32 pad_h = 9 [default = 0]; // The padding height (2D only)
optional uint32 pad_w = 10 [default = 0]; // The padding width (2D only)
optional uint32 kernel_h = 11; // The kernel height (2D only)
optional uint32 kernel_w = 12; // The kernel width (2D only)
optional uint32 stride_h = 13; // The stride height (2D only)
optional uint32 stride_w = 14; // The stride width (2D only)
// 来自于AlexNet论文
optional uint32 group = 5 [default = 1]; // The group size for group conv
// 权重初始化
optional FillerParameter weight_filler = 7; // The filler for the weight
// 偏置初始化
optional FillerParameter bias_filler = 8; // The filler for the bias
enum Engine {
DEFAULT = 0;
CAFFE = 1;
CUDNN = 2;
}
// 卷积的方式的选择，default是正常的卷积，caffe是矩阵乘法的卷积，cudnn是cuda库流并行式的卷积
optional Engine engine = 15 [default = DEFAULT];
// The axis to interpret as "channels" when performing convolution.
// Preceding dimensions are treated as independent inputs;
// succeeding dimensions are treated as "spatial".
// With (N, C, H, W) inputs, and axis == 1 (the default), we perform
// N independent 2D convolutions, sliding C-channel (or (C/g)-channels, for
// groups g>1) filters across the spatial axes (H, W) of the input.
// With (N, C, D, H, W) inputs, and axis == 1, we perform
// N independent 3D convolutions, sliding (C/g)-channels
// filters across the spatial axes (D, H, W) of the input.
// 通道channel所在的维度
optional int32 axis = 16 [default = 1];
// Whether to force use of the general ND convolution, even if a specific
// implementation for blobs of the appropriate number of spatial dimensions
// is available. (Currently, there is only a 2D-specific convolution
// implementation; for input blobs with num_axes != 2, this option is
// ignored and the ND implementation will be used.)
// 如果输入数据维度等于2，则执行通用的ND卷积，否则正常执行卷积
optional bool force_nd_im2col = 17 [default = false];
}
// 图像裁剪参数
message CropParameter {
// To crop, elements of the first bottom are selected to fit the dimensions
// of the second, reference bottom. The crop is configured by
// - the crop `axis` to pick the dimensions for cropping
// - the crop `offset` to set the shift for all/each dimension
// to align the cropped bottom with the reference bottom.
// All dimensions up to but excluding `axis` are preserved, while
// the dimensions including and trailing `axis` are cropped.
// If only one `offset` is set, then all dimensions are offset by this amount.
// Otherwise, the number of offsets must equal the number of cropped axes to
// shift the crop in each dimension accordingly.
// Note: standard dimensions are N,C,H,W so the default is a spatial crop,
// and `axis` may be negative to index from the end (e.g., -1 for the last
// axis).
// axis是在哪个维度上进行裁剪，会裁剪轴2及之后的所有轴
optional int32 axis = 1 [default = 2];
// offset设置是每个维度进行裁剪时的偏移量
repeated uint32 offset = 2;
}
// 数据层参数
message DataParameter {
enum DB {
LEVELDB = 0;
LMDB = 1;
}
// Specify the data source.
// 设定数据源路径
optional string source = 1;
// Specify the batch size.
// 指定一次处理的图片数量
optional uint32 batch_size = 4;
// The rand_skip variable is for the data layer to skip a few data points
// to avoid all asynchronous sgd clients to start at the same point. The skip
// point would be set as rand_skip * rand(0,1). Note that rand_skip should not
// be larger than the number of keys in the database.
// DEPRECATED. Each solver accesses a different subset of the database.
// rand_skip跳过指定的数据点，避免异步的sgd从同一个数据点开始
optional uint32 rand_skip = 7 [default = 0];
// 使用的数据库类型，LMDB or LEVELDB
optional DB backend = 8 [default = LEVELDB];
// DEPRECATED. See TransformationParameter. For data pre-processing, we can do
// simple scaling and subtracting the data mean, if provided. Note that the
// mean subtraction is always carried out before scaling.
// 已废弃。图像归一化，在TransformationParameter中。
optional float scale = 2 [default = 1];
// 已废弃。均值文件，在TransformationParameter中。
optional string mean_file = 3;
// DEPRECATED. See TransformationParameter. Specify if we would like to randomly
// crop an image.
// 已废弃。图像裁剪，在TransformationParameter中。
optional uint32 crop_size = 5 [default = 0];
// DEPRECATED. See TransformationParameter. Specify if we want to randomly mirror
// data.
// 已废弃。图像翻转，在TransformationParameter中。
optional bool mirror = 6 [default = false];
// Force the encoded image to have 3 color channels
// 强制图像数据有三个颜色通道
optional bool force_encoded_color = 9 [default = false];
// Prefetch queue (Number of batches to prefetch to host memory, increase if
// data access bandwidth varies).
// 预先拉取batch的数目
optional uint32 prefetch = 10 [default = 4];
}
// dropout层参数
message DropoutParameter {
// 为了避免过拟合，参数随机失活的比例
optional float dropout_ratio = 1 [default = 0.5]; // dropout ratio
}
// DummyDataLayer fills any number of arbitrarily shaped blobs with random
// (or constant) data generated by "Fillers" (see "message FillerParameter").
// DummyData层的参数
message DummyDataParameter {
// This layer produces N >= 1 top blobs. DummyDataParameter must specify 1 or N
// shape fields, and 0, 1 or N data_fillers.
// If 0 data_fillers are specified, ConstantFiller with a value of 0 is used.
// If 1 data_filler is specified, it is applied to all top blobs. If N are
// specified, the ith is applied to the ith top blob.
// blob数据的生成方式
repeated FillerParameter data_filler = 1;
// 数据的维度
repeated BlobShape shape = 6;
// 4D dimensions -- deprecated. Use "shape" instead.
// 已废弃。使用shape代替。
repeated uint32 num = 2;
repeated uint32 channels = 3;
repeated uint32 height = 4;
repeated uint32 width = 5;
}
//Eltwise层的参数
message EltwiseParameter {
// 操作的类型
enum EltwiseOp {
PROD = 0;
SUM = 1;
MAX = 2;
}
// 数据操作分三种：点乘，相加，取最大值
optional EltwiseOp operation = 1 [default = SUM]; // element-wise operation
// SUM操作时各个blob对应的系数
repeated float coeff = 2; // blob-wise coefficient for SUM operation
// Whether to use an asymptotically slower (for >2 inputs) but stabler method
// of computing the gradient for the PROD operation. (No effect for SUM op.)
// 在进行PROD操作，即乘法时是否使用异步操作来计算梯度，更慢但更稳定。
optional bool stable_prod_grad = 3 [default = true];
}
// Message that stores parameters used by ELULayer
// ELU层的参数，具体看论文
message ELUParameter {
// Described in:
// Clevert, D.-A., Unterthiner, T., & Hochreiter, S. (2015). Fast and Accurate
// Deep Network Learning by Exponential Linear Units (ELUs). arXiv
optional float alpha = 1 [default = 1];
}
// Message that stores parameters used by EmbedLayer
// Embed层的参数，主要用于LSTM等翻译网络
message EmbedParameter {
// Embed层的输出
optional uint32 num_output = 1; // The number of outputs for the layer
// The input is given as integers to be interpreted as one-hot
// vector indices with dimension num_input. Hence num_input should be
// 1 greater than the maximum possible input value.
// Embed层的输入
optional uint32 input_dim = 2;
// 是否使用偏置项
optional bool bias_term = 3 [default = true]; // Whether to use a bias term
// 权重生成
optional FillerParameter weight_filler = 4; // The filler for the weight
// 偏置生成
optional FillerParameter bias_filler = 5; // The filler for the bias
}
// Message that stores parameters used by ExpLayer
// Exp层的参数，即指数层参数
message ExpParameter {
// ExpLayer computes outputs y = base ^ (shift scale * x), for base > 0.
// Or if base is set to the default (-1), base is set to e,
// so y = exp(shift scale * x).
// 指数层的计算是y = base ^ (shift scale * x)，下面分别是公式中的三个参数
optional float base = 1 [default = -1.0];
optional float scale = 2 [default = 1.0];
optional float shift = 3 [default = 0.0];
}
// Message that stores parameters used by FlattenLayer
// Flatten层的参数，主要是按某个轴展开（平铺），mnist demo的mnist_autoencode就使用了Flatten层
message FlattenParameter {
// The first axis to flatten: all preceding axes are retained in the output.
// May be negative to index from the end (e.g., -1 for the last axis).
// 从哪一层开始展开
optional int32 axis = 1 [default = 1];
// The last axis to flatten: all following axes are retained in the output.
// May be negative to index from the end (e.g., the default -1 for the last
// axis).
// 展开到哪一层结束
optional int32 end_axis = 2 [default = -1];
}
// Message that stores parameters used by HDF5DataLayer
// HDF5数据层的参数
message HDF5DataParameter {
// Specify the data source.
// HDF5层输入数据的数据源
optional string source = 1;
// Specify the batch size.
// 训练的batch_size
optional uint32 batch_size = 2;
// Specify whether to shuffle the data.
// If shuffle == true, the ordering of the HDF5 files is shuffled,
// and the ordering of data within any given HDF5 file is shuffled,
// but data between different files are not interleaved; all of a file's
// data are output (in a random order) before moving onto another file.
// 是否对HDF5的输入数据进行shuffle
optional bool shuffle = 3 [default = false];
}
// HDF5输出层参数
message HDF5OutputParameter {
// 输出的HDF5文件的文件名
optional string file_name = 1;
}
// HingeLoss层参数
message HingeLossParameter {
enum Norm {
L1 = 1;
L2 = 2;
}
// Specify the Norm to use L1 or L2
// 指定HingeLoss的类型
optional Norm norm = 1 [default = L1];
}
// ImageData层参数，网络中直接输入原图
message ImageDataParameter {
// Specify the data source.
// 描述图像路径及标签的文件
optional string source = 1;
// Specify the batch size.
// 训练的batch size
optional uint32 batch_size = 4 [default = 1];
// The rand_skip variable is for the data layer to skip a few data points
// to avoid all asynchronous sgd clients to start at the same point. The skip
// point would be set as rand_skip * rand(0,1). Note that rand_skip should not
// be larger than the number of keys in the database.
// rand_skip跳过指定的数据点，避免异步的sgd从同一个数据点开始，与Data层中是一样的
optional uint32 rand_skip = 7 [default = 0];
// Whether or not ImageLayer should shuffle the list of files at every epoch.
// 是否对图像顺序进行shuffle
optional bool shuffle = 8 [default = false];
// It will also resize images if new_height or new_width are not zero.
// 图像resize的高度
optional uint32 new_height = 9 [default = 0];
// 图像resize的宽度
optional uint32 new_width = 10 [default = 0];
// Specify if the images are color or gray
// 指定图像彩色图像还是灰度图像，默认彩色
optional bool is_color = 11 [default = true];
// DEPRECATED. See TransformationParameter. For data pre-processing, we can do
// simple scaling and subtracting the data mean, if provided. Note that the
// mean subtraction is always carried out before scaling.
// 已废弃。参考TransformationParameter中的scale
optional float scale = 2 [default = 1];
// 指定均值文件
optional string mean_file = 3;
// DEPRECATED. See TransformationParameter. Specify if we would like to randomly
// crop an image.
// 已废弃。参考TransformationParameter中的crop_size
optional uint32 crop_size = 5 [default = 0];
// DEPRECATED. See TransformationParameter. Specify if we want to randomly mirror
// data.
// 已废弃，参考TransformationParameter的mirror。
optional bool mirror = 6 [default = false];
// 不太清楚root_folder具体是什么
optional string root_folder = 12 [default = ""];
}
// 信息增益损失层参数
message InfogainLossParameter {
// Specify the infogain matrix source.
// 指定存储信息增益矩阵的源文件
optional string source = 1;
}
// InnerProduct层的参数
message InnerProductParameter {
// InnerProduct层的输出
optional uint32 num_output = 1; // The number of outputs for the layer
// 是否有偏置项
optional bool bias_term = 2 [default = true]; // whether to have bias terms
// 权重初始化，随机生成
optional FillerParameter weight_filler = 3; // The filler for the weight
// 偏置初始化，随机生成
optional FillerParameter bias_filler = 4; // The filler for the bias
// The first axis to be lumped into a single inner product computation;
// all preceding axes are retained in the output.
// May be negative to index from the end (e.g., -1 for the last axis).
// 从某一维度开始进行内积计算，前面的维度保留
optional int32 axis = 5 [default = 1];
// Specify whether to transpose the weight matrix or not.
// If transpose == true, any operations will be performed on the transpose
// of the weight matrix. The weight matrix itself is not going to be transposed
// but rather the transfer flag of operations will be toggled accordingly.
// 是否对权重矩阵进行转置
optional bool transpose = 6 [default = false];
}
// Input参数，caffe网络部署时会用到
message InputParameter {
// This layer produces N >= 1 top blob(s) to be assigned manually.
// Define N shapes to set a shape for each top.
// Define 1 shape to set the same shape for every top.
// Define no shape to defer to reshaping manually.
// 输入数据的shape
repeated BlobShape shape = 1;
}
// Message that stores parameters used by LogLayer
// Log层参数，对数据进行Log运算
message LogParameter {
// LogLayer computes outputs y = log_base(shift scale * x), for base > 0.
// Or if base is set to the default (-1), base is set to e,
// so y = ln(shift scale * x) = log_e(shift scale * x)
// Log层计算公式为y = log_base(shift scale * x)，下面分别是公式中的三个参数
optional float base = 1 [default = -1.0];
optional float scale = 2 [default = 1.0];
optional float shift = 3 [default = 0.0];
}
// Message that stores parameters used by LRNLayer
// LRN层的参数，局部归一化，AlexNet中的LRN
message LRNParameter {
// 如果是跨通道LRN，则表示求和的通道数；如果是在通道内LRN，则表示求和的正方形区域长度。
optional uint32 local_size = 1 [default = 5];
// 归一化公式中的参数
optional float alpha = 2 [default = 1.];
optional float beta = 3 [default = 0.75];
enum NormRegion {
ACROSS_CHANNELS = 0;
WITHIN_CHANNEL = 1;
}
// 归一化的区域，分为通道内和跨通道两种
optional NormRegion norm_region = 4 [default = ACROSS_CHANNELS];
optional float k = 5 [default = 1.];
enum Engine {
DEFAULT = 0;
CAFFE = 1;
CUDNN = 2;
}
// 与前面的engine是一样的
optional Engine engine = 6 [default = DEFAULT];
}
// 内存数据层参数
message MemoryDataParameter {
// 训练的batch_size
optional uint32 batch_size = 1;
// 图像通道数
optional uint32 channels = 2;
// 图像高度
optional uint32 height = 3;
// 图像宽度
optional uint32 width = 4;
}
// mean-variance normalization层参数
message MVNParameter {
// This parameter can be set to false to normalize mean only
// 是否对方差进行归一化
optional bool normalize_variance = 1 [default = true];
// This parameter can be set to true to perform DNN-like MVN
// 是否进行跨通道的MVN
optional bool across_channels = 2 [default = false];
// Epsilon for not dividing by zero while normalizing variance
// 避免除数为0，与前面的一样
optional float eps = 3 [default = 1e-9];
}
// 参数层参数
message ParameterParameter {
// 用户自己定义的shape
optional BlobShape shape = 1;
}
// 池化层参数
message PoolingParameter {
enum PoolMethod {
MAX = 0;
AVE = 1;
STOCHASTIC = 2;
}
// 池化的方式
optional PoolMethod pool = 1 [default = MAX]; // The pooling method
// Pad, kernel size, and stride are all given as a single value for equal
// dimensions in height and width or as Y, X pairs.
// padding的大小
optional uint32 pad = 4 [default = 0]; // The padding size (equal in Y, X)
// padding的高度
optional uint32 pad_h = 9 [default = 0]; // The padding height
// padding的宽度
optional uint32 pad_w = 10 [default = 0]; // The padding width
// 池化的核大小
optional uint32 kernel_size = 2; // The kernel size (square)
// 核高度
optional uint32 kernel_h = 5; // The kernel height
// 核宽度
optional uint32 kernel_w = 6; // The kernel width
// 池化的步长
optional uint32 stride = 3 [default = 1]; // The stride (equal in Y, X)
// 步长的高度
optional uint32 stride_h = 7; // The stride height
// 步长的宽度
optional uint32 stride_w = 8; // The stride width
enum Engine {
DEFAULT = 0;
CAFFE = 1;
CUDNN = 2;
}
// 执行池化操作的类型，与前面的一样
optional Engine engine = 11 [default = DEFAULT];
// If global_pooling then it will pool over the size of the bottom by doing
// kernel_h = bottom->height and kernel_w = bottom->width
// global_pooling是对多个通道进行pooling，例如从三通道pooling为单通道
optional bool global_pooling = 12 [default = false];
}
// Power层参数
message PowerParameter {
// PowerLayer computes outputs y = (shift scale * x) ^ power.
// Power的计算公式为y = (shift scale * x) ^ power，下面是公式中的参数
optional float power = 1 [default = 1.0];
optional float scale = 2 [default = 1.0];
optional float shift = 3 [default = 0.0];
}
// python layer参数，在faster rcnn中有应用
message PythonParameter {
// python模块名称
optional string module = 1;
// python模块中层的名字，即类名
optional string layer = 2;
// This value is set to the attribute `param_str` of the `PythonLayer` object
// in Python before calling the `setup()` method. This could be a number,
// string, dictionary in Python dict format, JSON, etc. You may parse this
// string in `setup` method and use it in `forward` and `backward`.
// 可以用来设置参数，key-value形式，可以参考faster rcnn中模型的train.prototxt
optional string param_str = 3 [default = ''];
// Whether this PythonLayer is shared among worker solvers during data parallelism.
// If true, each worker solver sequentially run forward from this layer.
// This value should be set true if you are using it as a data layer.
// 是否需要在并行时共享layer
optional bool share_in_parallel = 4 [default = false];
}
// Message that stores parameters used by RecurrentLayer
// Recurrent层参数
message RecurrentParameter {
// The dimension of the output (and usually hidden state) representation --
// must be explicitly set to non-zero.
// Recurrent层的输出——必须非零
optional uint32 num_output = 1 [default = 0];
// 权重初始化，随机生成初始化
optional FillerParameter weight_filler = 2; // The filler for the weight
// 偏置初始化，随机生成
optional FillerParameter bias_filler = 3; // The filler for the bias
// Whether to enable displaying debug_info in the unrolled recurrent net.
// 是否输出调试信息
optional bool debug_info = 4 [default = false];
// Whether to add as additional inputs (bottoms) the initial hidden state
// blobs, and add as additional outputs (tops) the final timestep hidden state
// blobs. The number of additional bottom/top blobs required depends on the
// recurrent architecture -- e.g., 1 for RNNs, 2 for LSTMs.
// 是否添加额外的输入
optional bool expose_hidden = 5 [default = false];
}
// Message that stores parameters used by ReductionLayer
// Reduction层参数
message ReductionParameter {
enum ReductionOp {
SUM = 1;
ASUM = 2;
SUMSQ = 3;
MEAN = 4;
}
// 通过reduction操作来将数据减少到一维，可以通过上面的四种方式
optional ReductionOp operation = 1 [default = SUM]; // reduction operation
// The first axis to reduce to a scalar -- may be negative to index from the
// end (e.g., -1 for the last axis).
// (Currently, only reduction along ALL "tail" axes is supported; reduction
// of axis M through N, where N < num_axes - 1, is unsupported.)
// Suppose we have an n-axis bottom Blob with shape:
// (d0, d1, d2, ..., d(m-1), dm, d(m 1), ..., d(n-1)).
// If axis == m, the output Blob will have shape
// (d0, d1, d2, ..., d(m-1)),
// and the ReductionOp operation is performed (d0 * d1 * d2 * ... * d(m-1))
// times, each including (dm * d(m 1) * ... * d(n-1)) individual data.
// If axis == 0 (the default), the output Blob always has the empty shape
// (count 1), performing reduction across the entire input --
// often useful for creating new loss functions.
// 在哪个轴上执行reduction操作
optional int32 axis = 2 [default = 0];
// 输出系数
optional float coeff = 3 [default = 1.0]; // coefficient for output
}
// Message that stores parameters used by ReLULayer
// ReLU层参数
message ReLUParameter {
// Allow non-zero slope for negative inputs to speed up optimization
// Described in:
// Maas, A. L., Hannun, A. Y., & Ng, A. Y. (2013). Rectifier nonlinearities
// improve neural network acoustic models. In ICML Workshop on Deep Learning
// for Audio, Speech, and Language Processing.
// ReLUU操作的阈值
optional float negative_slope = 1 [default = 0];
enum Engine {
DEFAULT = 0;
CAFFE = 1;
CUDNN = 2;
}
// 执行ReLU操作的类型，与前面的一样
optional Engine engine = 2 [default = DEFAULT];
}
// Reshape层参数，与numpy中的Reshape作用是一样的
message ReshapeParameter {
// Specify the output dimensions. If some of the dimensions are set to 0,
// the corresponding dimension from the bottom layer is used (unchanged).
// Exactly one dimension may be set to -1, in which case its value is
// inferred from the count of the bottom blob and the remaining dimensions.
// For example, suppose we want to reshape a 2D blob "input" with shape 2 x 8:
//
// layer {
// type: "Reshape" bottom: "input" top: "output"
// reshape_param { ... }
// }
//
// If "input" is 2D with shape 2 x 8, then the following reshape_param
// specifications are all equivalent, producing a 3D blob "output" with shape
// 2 x 2 x 4:
//
// reshape_param { shape { dim: 2 dim: 2 dim: 4 } }
// reshape_param { shape { dim: 0 dim: 2 dim: 4 } }
// reshape_param { shape { dim: 0 dim: 2 dim: -1 } }
// reshape_param { shape { dim: 0 dim:-1 dim: 4 } }
// reshape之后输出的维度
optional BlobShape shape = 1;
// axis and num_axes control the portion of the bottom blob's shape that are
// replaced by (included in) the reshape. By default (axis == 0 and
// num_axes == -1), the entire bottom blob shape is included in the reshape,
// and hence the shape field must specify the entire output shape.
//
// axis may be non-zero to retain some portion of the beginning of the input
// shape (and may be negative to index from the end; e.g., -1 to begin the
// reshape after the last axis, including nothing in the reshape,
// -2 to include only the last axis, etc.).
//
// For example, suppose "input" is a 2D blob with shape 2 x 8.
// Then the following ReshapeLayer specifications are all equivalent,
// producing a blob "output" with shape 2 x 2 x 4:
//
// reshape_param { shape { dim: 2 dim: 2 dim: 4 } }
// reshape_param { shape { dim: 2 dim: 4 } axis: 1 }
// reshape_param { shape { dim: 2 dim: 4 } axis: -3 }
//
// num_axes specifies the extent of the reshape.
// If num_axes >= 0 (and axis >= 0), the reshape will be performed only on
// input axes in the range [axis, axis num_axes].
// num_axes may also be -1, the default, to include all remaining axes
// (starting from axis).
//
// For example, suppose "input" is a 2D blob with shape 2 x 8.
// Then the following ReshapeLayer specifications are equivalent,
// producing a blob "output" with shape 1 x 2 x 8.
//
// reshape_param { shape { dim: 1 dim: 2 dim: 8 } }
// reshape_param { shape { dim: 1 dim: 2 } num_axes: 1 }
// reshape_param { shape { dim: 1 } num_axes: 0 }
//
// On the other hand, these would produce output blob shape 2 x 1 x 8:
//
// reshape_param { shape { dim: 2 dim: 1 dim: 8 } }
// reshape_param { shape { dim: 1 } axis: 1 num_axes: 0 }
optional int32 axis = 2 [default = 0];
optional int32 num_axes = 3 [default = -1];
}
// Scale层参数，与batch norm layer配合使用，可参考Resnet结构
message ScaleParameter {
// The first axis of bottom[0] (the first input Blob) along which to apply
// bottom[1] (the second input Blob). May be negative to index from the end
// (e.g., -1 for the last axis).
//
// For example, if bottom[0] is 4D with shape 100x3x40x60, the output
// top[0] will have the same shape, and bottom[1] may have any of the
// following shapes (for the given value of axis):
// (axis == 0 == -4) 100; 100x3; 100x3x40; 100x3x40x60
// (axis == 1 == -3) 3; 3x40; 3x40x60
// (axis == 2 == -2) 40; 40x60
// (axis == 3 == -1) 60
// Furthermore, bottom[1] may have the empty shape (regardless of the value of
// "axis") -- a scalar multiplier.
optional int32 axis = 1 [default = 1];
// (num_axes is ignored unless just one bottom is given and the scale is
// a learned parameter of the layer. Otherwise, num_axes is determined by the
// number of axes by the second bottom.)
// The number of axes of the input (bottom[0]) covered by the scale
// parameter, or -1 to cover all axes of bottom[0] starting from `axis`.
// Set num_axes := 0, to multiply with a zero-axis Blob: a scalar.
optional int32 num_axes = 2 [default = 1];
// (filler is ignored unless just one bottom is given and the scale is
// a learned parameter of the layer.)
// The initialization for the learned scale parameter.
// Default is the unit (1) initialization, resulting in the ScaleLayer
// initially performing the identity operation.
optional FillerParameter filler = 3;
// Whether to also learn a bias (equivalent to a ScaleLayer BiasLayer, but
// may be more efficient). Initialized with bias_filler (defaults to 0).
// 是否使用偏置项
optional bool bias_term = 4 [default = false];
// 偏置项初始化
optional FillerParameter bias_filler = 5;
}
// Sigmoid层参数
message SigmoidParameter {
enum Engine {
DEFAULT = 0;
CAFFE = 1;
CUDNN = 2;
}
// 使用哪种sigmoid实现
optional Engine engine = 1 [default = DEFAULT];
}
// Slice层参数
message SliceParameter {
// The axis along which to slice -- may be negative to index from the end
// (e.g., -1 for the last axis).
// By default, SliceLayer concatenates blobs along the "channels" axis (1).
// 在哪个维度上进行拆分
optional int32 axis = 3 [default = 1];
// 指定拆分点
repeated uint32 slice_point = 2;
// DEPRECATED: alias for "axis" -- does not support negative indexing.
// 已废弃。
optional uint32 slice_dim = 1 [default = 1];
}
// Message that stores parameters used by SoftmaxLayer, SoftmaxWithLossLayer
// Softmax层参数
message SoftmaxParameter {
enum Engine {
DEFAULT = 0;
CAFFE = 1;
CUDNN = 2;
}
// 使用哪种softmax实现
optional Engine engine = 1 [default = DEFAULT];
// The axis along which to perform the softmax -- may be negative to index
// from the end (e.g., -1 for the last axis).
// Any other axes will be evaluated as independent softmaxes.
// 在哪个维度上进行softmax
optional int32 axis = 2 [default = 1];
}
// TanH层参数
message TanHParameter {
enum Engine {
DEFAULT = 0;
CAFFE = 1;
CUDNN = 2;
}
// 执行tanh激活函数的类型
optional Engine engine = 1 [default = DEFAULT];
}
// Message that stores parameters used by TileLayer
// Tile层参数，扩大某一维度
message TileParameter {
// The index of the axis to tile.
// 扩大哪个维度
optional int32 axis = 1 [default = 1];
// The number of copies (tiles) of the blob to output.
// 创建多少个副本
optional int32 tiles = 2;
}
// Message that stores parameters used by ThresholdLayer
// Threshold层参数，主要用来测试输入是否超过阈值
message ThresholdParameter {
// 设置阈值
optional float threshold = 1 [default = 0]; // Strictly positive values
}
// WindowData层参数
message WindowDataParameter {
// Specify the data source.
// 指定数据源
optional string source = 1;
// For data pre-processing, we can do simple scaling and subtracting the
// data mean, if provided. Note that the mean subtraction is always carried
// out before scaling.
// 是否归一化
optional float scale = 2 [default = 1];
// 图像均值文件
optional string mean_file = 3;
// Specify the batch size.
// 训练的batch_size
optional uint32 batch_size = 4;
// Specify if we would like to randomly crop an image.
// 是否随机crop
optional uint32 crop_size = 5 [default = 0];
// Specify if we want to randomly mirror data.
// 是否随机mirror
optional bool mirror = 6 [default = false];
// Foreground (object) overlap threshold
// 前景重叠阈值
optional float fg_threshold = 7 [default = 0.5];
// Background (non-object) overlap threshold
// 背景重叠阈值
optional float bg_threshold = 8 [default = 0.5];
// Fraction of batch that should be foreground objects
// 前景比例
optional float fg_fraction = 9 [default = 0.25];
// Amount of contextual padding to add around a window
// (used only by the window_data_layer)
// 是否padding
optional uint32 context_pad = 10 [default = 0];
// Mode for cropping out a detection window
// warp: cropped window is warped to a fixed size and aspect ratio
// square: the tightest square around the window is cropped
// crop的方式
optional string crop_mode = 11 [default = "warp"];
// cache_images: will load all images in memory for faster access
// 是否缓存图像，即将图像都转入内存
optional bool cache_images = 12 [default = false];
// append root_folder to locate images
// 图像文件的根目录
optional string root_folder = 13 [default = ""];
}
// SPP层参数，SPP是spatial pyramid pooling，空间金字塔池化，具体可参考何凯明论文Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition
message SPPParameter {
enum PoolMethod {
MAX = 0;
AVE = 1;
STOCHASTIC = 2;
}
// 空间金字塔高度
optional uint32 pyramid_height = 1;
// 池化方法
optional PoolMethod pool = 2 [default = MAX]; // The pooling method
enum Engine {
DEFAULT = 0;
CAFFE = 1;
CUDNN = 2;
}
// 执行SPP的方式
optional Engine engine = 6 [default = DEFAULT];
}
// DEPRECATED: use LayerParameter.
// 已废弃，使用LayerParameter。
message V1LayerParameter {
repeated string bottom = 2;
repeated string top = 3;
optional string name = 4;
repeated NetStateRule include = 32;
repeated NetStateRule exclude = 33;
enum LayerType {
NONE = 0;
ABSVAL = 35;
ACCURACY = 1;
ARGMAX = 30;
BNLL = 2;
CONCAT = 3;
CONTRASTIVE_LOSS = 37;
CONVOLUTION = 4;
DATA = 5;
DECONVOLUTION = 39;
DROPOUT = 6;
DUMMY_DATA = 32;
EUCLIDEAN_LOSS = 7;
ELTWISE = 25;
EXP = 38;
FLATTEN = 8;
HDF5_DATA = 9;
HDF5_OUTPUT = 10;
HINGE_LOSS = 28;
IM2COL = 11;
IMAGE_DATA = 12;
INFOGAIN_LOSS = 13;
INNER_PRODUCT = 14;
LRN = 15;
MEMORY_DATA = 29;
MULTINOMIAL_LOGISTIC_LOSS = 16;
MVN = 34;
POOLING = 17;
POWER = 26;
RELU = 18;
SIGMOID = 19;
SIGMOID_CROSS_ENTROPY_LOSS = 27;
SILENCE = 36;
SOFTMAX = 20;
SOFTMAX_LOSS = 21;
SPLIT = 22;
SLICE = 33;
TANH = 23;
WINDOW_DATA = 24;
THRESHOLD = 31;
}
optional LayerType type = 5;
repeated BlobProto blobs = 6;
repeated string param = 1001;
repeated DimCheckMode blob_share_mode = 1002;
enum DimCheckMode {
STRICT = 0;
PERMISSIVE = 1;
}
repeated float blobs_lr = 7;
repeated float weight_decay = 8;
repeated float loss_weight = 35;
optional AccuracyParameter accuracy_param = 27;
optional ArgMaxParameter argmax_param = 23;
optional ConcatParameter concat_param = 9;
optional ContrastiveLossParameter contrastive_loss_param = 40;
optional ConvolutionParameter convolution_param = 10;
optional DataParameter data_param = 11;
optional DropoutParameter dropout_param = 12;
optional DummyDataParameter dummy_data_param = 26;
optional EltwiseParameter eltwise_param = 24;
optional ExpParameter exp_param = 41;
optional HDF5DataParameter hdf5_data_param = 13;
optional HDF5OutputParameter hdf5_output_param = 14;
optional HingeLossParameter hinge_loss_param = 29;
optional ImageDataParameter image_data_param = 15;
optional InfogainLossParameter infogain_loss_param = 16;
optional InnerProductParameter inner_product_param = 17;
optional LRNParameter lrn_param = 18;
optional MemoryDataParameter memory_data_param = 22;
optional MVNParameter mvn_param = 34;
optional PoolingParameter pooling_param = 19;
optional PowerParameter power_param = 21;
optional ReLUParameter relu_param = 30;
optional SigmoidParameter sigmoid_param = 38;
optional SoftmaxParameter softmax_param = 39;
optional SliceParameter slice_param = 31;
optional TanHParameter tanh_param = 37;
optional ThresholdParameter threshold_param = 25;
optional WindowDataParameter window_data_param = 20;
optional TransformationParameter transform_param = 36;
optional LossParameter loss_param = 42;
optional V0LayerParameter layer = 1;
}
// DEPRECATED: V0LayerParameter is the old way of specifying layer parameters
// in Caffe. We keep this message type around for legacy support.
// 已废弃。
message V0LayerParameter {
optional string name = 1; // the layer name
optional string type = 2; // the string to specify the layer type
// Parameters to specify layers with inner products.
optional uint32 num_output = 3; // The number of outputs for the layer
optional bool biasterm = 4 [default = true]; // whether to have bias terms
optional FillerParameter weight_filler = 5; // The filler for the weight
optional FillerParameter bias_filler = 6; // The filler for the bias
optional uint32 pad = 7 [default = 0]; // The padding size
optional uint32 kernelsize = 8; // The kernel size
optional uint32 group = 9 [default = 1]; // The group size for group conv
optional uint32 stride = 10 [default = 1]; // The stride
enum PoolMethod {
MAX = 0;
AVE = 1;
STOCHASTIC = 2;
}
optional PoolMethod pool = 11 [default = MAX]; // The pooling method
optional float dropout_ratio = 12 [default = 0.5]; // dropout ratio
optional uint32 local_size = 13 [default = 5]; // for local response norm
optional float alpha = 14 [default = 1.]; // for local response norm
optional float beta = 15 [default = 0.75]; // for local response norm
optional float k = 22 [default = 1.];
// For data layers, specify the data source
optional string source = 16;
// For data pre-processing, we can do simple scaling and subtracting the
// data mean, if provided. Note that the mean subtraction is always carried
// out before scaling.
optional float scale = 17 [default = 1];
optional string meanfile = 18;
// For data layers, specify the batch size.
optional uint32 batchsize = 19;
// For data layers, specify if we would like to randomly crop an image.
optional uint32 cropsize = 20 [default = 0];
// For data layers, specify if we want to randomly mirror data.
optional bool mirror = 21 [default = false];
// The blobs containing the numeric parameters of the layer
repeated BlobProto blobs = 50;
// The ratio that is multiplied on the global learning rate. If you want to
// set the learning ratio for one blob, you need to set it for all blobs.
repeated float blobs_lr = 51;
// The weight decay that is multiplied on the global weight decay.
repeated float weight_decay = 52;
// The rand_skip variable is for the data layer to skip a few data points
// to avoid all asynchronous sgd clients to start at the same point. The skip
// point would be set as rand_skip * rand(0,1). Note that rand_skip should not
// be larger than the number of keys in the database.
optional uint32 rand_skip = 53 [default = 0];
// Fields related to detection (det_*)
// foreground (object) overlap threshold
optional float det_fg_threshold = 54 [default = 0.5];
// background (non-object) overlap threshold
optional float det_bg_threshold = 55 [default = 0.5];
// Fraction of batch that should be foreground objects
optional float det_fg_fraction = 56 [default = 0.25];
// optional bool OBSOLETE_can_clobber = 57 [default = true];
// Amount of contextual padding to add around a window
// (used only by the window_data_layer)
optional uint32 det_context_pad = 58 [default = 0];
// Mode for cropping out a detection window
// warp: cropped window is warped to a fixed size and aspect ratio
// square: the tightest square around the window is cropped
optional string det_crop_mode = 59 [default = "warp"];
// For ReshapeLayer, one needs to specify the new dimensions.
optional int32 new_num = 60 [default = 0];
optional int32 new_channels = 61 [default = 0];
optional int32 new_height = 62 [default = 0];
optional int32 new_width = 63 [default = 0];
// Whether or not ImageLayer should shuffle the list of files at every epoch.
// It will also resize images if new_height or new_width are not zero.
optional bool shuffle_images = 64 [default = false];
// For ConcatLayer, one needs to specify the dimension for concatenation, and
// the other dimensions must be the same for all the bottom blobs.
// By default it will concatenate blobs along the channels dimension.
optional uint32 concat_dim = 65 [default = 1];
optional HDF5OutputParameter hdf5_output_param = 1001;
}
// PReLU层参数，ReLU的进化版本
message PReLUParameter {
// Parametric ReLU described in K. He et al, Delving Deep into Rectifiers:
// Surpassing Human-Level Performance on ImageNet Classification, 2015.
// Initial value of a_i. Default is a_i=0.25 for all i.
// 参数初始化
optional FillerParameter filler = 1;
// Whether or not slope parameters are shared across channels.
// 是否在各通道共享参数
optional bool channel_shared = 2 [default = false];
} 
