[examples] sequence and revise notebooks

- combine classification + filter visualization - order by classification, learning LeNet, brewing logreg, and fine-tuning to flickr style - improve flow of content in classification + filter visualization - include solver needed for learning LeNet - edit notebook descriptions for site catalogue
chengyangfu · Jun 30, 2015 · a6cb8ec · a6cb8ec
1 parent b8cc297
commit a6cb8ec
Show file tree

Hide file tree

Showing 19 changed files with 22,296 additions and 26,009 deletions.
diff --git a/examples/00-classification.ipynb b/examples/00-classification.ipynb
diff --git a/examples/python_solving.ipynb → examples/01-learning-lenet.ipynb b/examples/python_solving.ipynb → examples/01-learning-lenet.ipynb
diff --git a/examples/02-brewing-logreg.ipynb b/examples/02-brewing-logreg.ipynb
diff --git a/examples/03-fine-tuning.ipynb b/examples/03-fine-tuning.ipynb
diff --git a/examples/Finetune with Flickr Style Data.ipynb b/examples/Finetune with Flickr Style Data.ipynb
diff --git a/examples/classification.ipynb b/examples/classification.ipynb
diff --git a/examples/detection.ipynb b/examples/detection.ipynb
@@ -8385,7 +8385,7 @@
    "pygments_lexer": "ipython2",
    "version": "2.7.9"
   },
-  "priority": 3
+  "priority": 6
  },
  "nbformat": 4,
  "nbformat_minor": 0

diff --git a/examples/filter_visualization.ipynb b/examples/filter_visualization.ipynb
diff --git a/examples/hdf5_classification.ipynb b/examples/hdf5_classification.ipynb
diff --git a/examples/hdf5_classification/nonlinear_auto_test.prototxt b/examples/hdf5_classification/nonlinear_auto_test.prototxt
@@ -0,0 +1,54 @@
+layer {
+  name: "data"
+  type: "HDF5Data"
+  top: "data"
+  top: "label"
+  hdf5_data_param {
+    source: "examples/hdf5_classification/data/test.txt"
+    batch_size: 10
+  }
+}
+layer {
+  name: "ip1"
+  type: "InnerProduct"
+  bottom: "data"
+  top: "ip1"
+  inner_product_param {
+    num_output: 40
+    weight_filler {
+      type: "xavier"
+    }
+  }
+}
+layer {
+  name: "relu1"
+  type: "ReLU"
+  bottom: "ip1"
+  top: "ip1"
+}
+layer {
+  name: "ip2"
+  type: "InnerProduct"
+  bottom: "ip1"
+  top: "ip2"
+  inner_product_param {
+    num_output: 2
+    weight_filler {
+      type: "xavier"
+    }
+  }
+}
+layer {
+  name: "accuracy"
+  type: "Accuracy"
+  bottom: "ip2"
+  bottom: "label"
+  top: "accuracy"
+}
+layer {
+  name: "loss"
+  type: "SoftmaxWithLoss"
+  bottom: "ip2"
+  bottom: "label"
+  top: "loss"
+}
diff --git a/examples/hdf5_classification/nonlinear_auto_train.prototxt b/examples/hdf5_classification/nonlinear_auto_train.prototxt
@@ -0,0 +1,54 @@
+layer {
+  name: "data"
+  type: "HDF5Data"
+  top: "data"
+  top: "label"
+  hdf5_data_param {
+    source: "examples/hdf5_classification/data/train.txt"
+    batch_size: 10
+  }
+}
+layer {
+  name: "ip1"
+  type: "InnerProduct"
+  bottom: "data"
+  top: "ip1"
+  inner_product_param {
+    num_output: 40
+    weight_filler {
+      type: "xavier"
+    }
+  }
+}
+layer {
+  name: "relu1"
+  type: "ReLU"
+  bottom: "ip1"
+  top: "ip1"
+}
+layer {
+  name: "ip2"
+  type: "InnerProduct"
+  bottom: "ip1"
+  top: "ip2"
+  inner_product_param {
+    num_output: 2
+    weight_filler {
+      type: "xavier"
+    }
+  }
+}
+layer {
+  name: "accuracy"
+  type: "Accuracy"
+  bottom: "ip2"
+  bottom: "label"
+  top: "accuracy"
+}
+layer {
+  name: "loss"
+  type: "SoftmaxWithLoss"
+  bottom: "ip2"
+  bottom: "label"
+  top: "loss"
+}
diff --git a/examples/hdf5_classification/nonlinear_solver.prototxt b/examples/hdf5_classification/nonlinear_solver.prototxt
@@ -0,0 +1,15 @@
+train_net: "examples/hdf5_classification/nonlinear_auto_train.prototxt"
+test_net: "examples/hdf5_classification/nonlinear_auto_test.prototxt"
+test_iter: 250
+test_interval: 1000
+base_lr: 0.01
+lr_policy: "step"
+gamma: 0.1
+stepsize: 5000
+display: 1000
+max_iter: 10000
+momentum: 0.9
+weight_decay: 0.0005
+snapshot: 10000
+snapshot_prefix: "examples/hdf5_classification/data/train"
+solver_mode: CPU
diff --git a/...s/hdf5_classification/train_val2.prototxt → ...assification/nonlinear_train_val.prototxt b/...s/hdf5_classification/train_val2.prototxt → ...assification/nonlinear_train_val.prototxt
@@ -8,7 +8,7 @@ layer {
     phase: TRAIN
   }
   hdf5_data_param {
-    source: "hdf5_classification/data/train.txt"
+    source: "examples/hdf5_classification/data/train.txt"
     batch_size: 10
   }
 }
@@ -21,7 +21,7 @@ layer {
     phase: TEST
   }
   hdf5_data_param {
-    source: "hdf5_classification/data/test.txt"
+    source: "examples/hdf5_classification/data/test.txt"
     batch_size: 10
   }
 }
@@ -41,8 +41,7 @@ layer {
   inner_product_param {
     num_output: 40
     weight_filler {
-      type: "gaussian"
-      std: 0.01
+      type: "xavier"
     }
     bias_filler {
       type: "constant"
@@ -72,8 +71,7 @@ layer {
   inner_product_param {
     num_output: 2
     weight_filler {
-      type: "gaussian"
-      std: 0.01
+      type: "xavier"
     }
     bias_filler {
       type: "constant"

diff --git a/examples/hdf5_classification/solver.prototxt b/examples/hdf5_classification/solver.prototxt
@@ -1,4 +1,5 @@
-net: "hdf5_classification/train_val.prototxt"
+train_net: "examples/hdf5_classification/logreg_auto_train.prototxt"
+test_net: "examples/hdf5_classification/logreg_auto_test.prototxt"
 test_iter: 250
 test_interval: 1000
 base_lr: 0.01
@@ -10,5 +11,5 @@ max_iter: 10000
 momentum: 0.9
 weight_decay: 0.0005
 snapshot: 10000
-snapshot_prefix: "hdf5_classification/data/train"
+snapshot_prefix: "examples/hdf5_classification/data/train"
 solver_mode: CPU
diff --git a/examples/hdf5_classification/solver2.prototxt b/examples/hdf5_classification/solver2.prototxt
diff --git a/examples/hdf5_classification/train_val.prototxt b/examples/hdf5_classification/train_val.prototxt
@@ -8,7 +8,7 @@ layer {
     phase: TRAIN
   }
   hdf5_data_param {
-    source: "hdf5_classification/data/train.txt"
+    source: "examples/hdf5_classification/data/train.txt"
     batch_size: 10
   }
 }
@@ -21,7 +21,7 @@ layer {
     phase: TEST
   }
   hdf5_data_param {
-    source: "hdf5_classification/data/test.txt"
+    source: "examples/hdf5_classification/data/test.txt"
     batch_size: 10
   }
 }
@@ -41,8 +41,7 @@ layer {
   inner_product_param {
     num_output: 2
     weight_filler {
-      type: "gaussian"
-      std: 0.01
+      type: "xavier"
     }
     bias_filler {
       type: "constant"

diff --git a/examples/mnist/lenet_auto_solver.prototxt b/examples/mnist/lenet_auto_solver.prototxt
@@ -0,0 +1,24 @@
+# The train/test net protocol buffer definition
+train_net: "examples/mnist/lenet_auto_train.prototxt"
+test_net: "examples/mnist/lenet_auto_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: 100
+# Carry out testing every 500 training iterations.
+test_interval: 500
+# The base learning rate, momentum and the weight decay of the network.
+base_lr: 0.01
+momentum: 0.9
+weight_decay: 0.0005
+# The learning rate policy
+lr_policy: "inv"
+gamma: 0.0001
+power: 0.75
+# Display every 100 iterations
+display: 100
+# The maximum number of iterations
+max_iter: 10000
+# snapshot intermediate results
+snapshot: 5000
+snapshot_prefix: "examples/mnist/lenet"
diff --git a/examples/net_surgery.ipynb b/examples/net_surgery.ipynb
@@ -6884,7 +6884,7 @@
   }
  ],
  "metadata": {
-  "description": "How to do net surgery and manually change model parameters, making a fully-convolutional classifier for dense feature extraction.",
+  "description": "How to do net surgery and manually change model parameters for custom use.",
   "example_name": "Editing model parameters",
   "include_in_docs": true,
   "kernelspec": {

diff --git a/examples/siamese/mnist_siamese.ipynb b/examples/siamese/mnist_siamese.ipynb
@@ -1902,7 +1902,7 @@
    "pygments_lexer": "ipython2",
    "version": "2.7.9"
   },
-  "priority": 6
+  "priority": 7
  },
  "nbformat": 4,
  "nbformat_minor": 0