Simple Caffe model replicating the config used in MLPMnistSingleLayerExample

simple_solver.prototxt

# The train/test net protocol buffer definition
net: "examples/mnist/simple_train_test.prototxt"
# test_iter specifies how many forward passes the test should carry out.
# In the case of MNIST, we have test batch size 100 and 100 test iterations,
# covering the full 10,000 testing images.
test_iter: 78
# Carry out testing every 500 training iterations.
test_interval: 7035
# The base learning rate, momentum and the weight decay of the network.
base_lr: 0.006
momentum: 0.9
weight_decay: 1e-4
# The learning rate policy
lr_policy: "fixed"
gamma: 0
power: 0
# Display every 100 iterations
display: 100
# The maximum number of iterations
max_iter: 7035
# snapshot intermediate results
snapshot: 7035
snapshot_prefix: "examples/mnist/simple"
# solver mode: CPU or GPU
solver_mode: CPU
type: "Nesterov"

simple_train_test.prototxt

name: "Simple"
layer {
  name: "mnist"
  type: "Data"
  top: "data"
  top: "label"
  include {
    phase: TRAIN
  }
  transform_param {
    scale: 0.00390625
  }
  data_param {
    source: "examples/mnist/mnist_train_lmdb"
    batch_size: 128
    backend: LMDB
  }
}
layer {
  name: "mnist"
  type: "Data"
  top: "data"
  top: "label"
  include {
    phase: TEST
  }
  transform_param {
    scale: 0.00390625
  }
  data_param {
    source: "examples/mnist/mnist_test_lmdb"
    batch_size: 128
    backend: LMDB
  }
}
layer {
  name: "ip1"
  type: "InnerProduct"
  bottom: "data"
  top: "ip1"
  inner_product_param {
    num_output: 1000
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
    }
  }
}
layer {
  name: "relu1"
  type: "ReLU"
  bottom: "ip1"
  top: "ip1"
}
layer {
  name: "ip2"
  type: "InnerProduct"
  bottom: "ip1"
  top: "ip2"
  inner_product_param {
    num_output: 10
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
    }
  }
}
layer {
  name: "accuracy"
  type: "Accuracy"
  bottom: "ip2"
  bottom: "label"
  top: "accuracy"
  include {
    phase: TEST
  }
}
layer {
  name: "loss"
  type: "SoftmaxWithLoss"
  bottom: "ip2"
  bottom: "label"
  top: "loss"
}

train_simple.sh

#!/usr/bin/env sh

./build/tools/caffe train --solver=examples/mnist/simple_solver.prototxt