Back-merge changes in master

* master: bundle presentation in gh-pages for now... fix typo pointed out by @yinxusen note support for non-MKL installation in dev include pretrained snapshot and performance details Document AlexNet model, include download script define AlexNet architecture polished ignore
BVLC · May 20, 2014 · c6b00a3 · c6b00a3
2 parents f048bea + 86a3263
commit c6b00a3
Show file tree

Hide file tree

Showing 9 changed files with 899 additions and 3 deletions.
diff --git a/README.md b/README.md
@@ -22,7 +22,7 @@ line of code: `Caffe::set_mode(Caffe::CPU)`.
 Even in CPU mode, computing predictions on an image takes only 20 ms when images
 are processed in batch mode.
 
-* [Caffe introductory presentation](https://www.dropbox.com/s/10fx16yp5etb8dv/caffe-presentation.pdf)
+* [Caffe introductory presentation](http://berkeleyvision.org/caffe-presentation.pdf)
 * [Installation instructions](http://caffe.berkeleyvision.org/installation.html)
 
 \* When measured with the [SuperVision](http://www.image-net.org/challenges/LSVRC/2012/supervision.pdf) model that won the ImageNet Large Scale Visual Recognition Challenge 2012.

diff --git a/docs/getting_pretrained_models.md b/docs/getting_pretrained_models.md
@@ -12,6 +12,16 @@ This page will be updated as more models become available.
 
 ### ImageNet
 
-Our reference implementation of the AlexNet model trained on ILSVRC-2012 can be downloaded (232.57MB) by running `examples/imagenet/get_caffe_reference_imagenet_model.sh` from the Caffe root directory.
+**Caffe Reference ImageNet Model**: Our reference implementation of an ImageNet model trained on ILSVRC-2012 can be downloaded (232.6MB) by running `examples/imagenet/get_caffe_reference_imagenet_model.sh` from the Caffe root directory.
+
+- The bundled model is the iteration 310,000 snapshot.
+- The best validation performance during training was iteration 313,000 with
+  validation accuracy 57.412% and loss 1.82328.
+
+**AlexNet**: Our training of the Krizhevsky architecture, which differs from the paper's methodology by (1) not training with the relighting data-augmentation and (2) initializing non-zero biases to 0.1 instead of 1. (2) was found necessary for training, as initialization to 1 gave flat loss. Download the model (243.9MB) by running `examples/imagenet/get_caffe_alexnet_model.sh` from the Caffe root directory.
+
+- The bundled model is the iteration 360,000 snapshot.
+- The best validation performance during training was iteration 358,000 with
+  validation accuracy 57.258% and loss 1.83948.
 
 Additionally, you will probably eventually need some auxiliary data (mean image, synset list, etc.): run `data/ilsvrc12/get_ilsvrc_aux.sh` from the root directory to obtain it.
diff --git a/docs/index.md b/docs/index.md
@@ -23,7 +23,7 @@ Even in CPU mode, computing predictions on an image takes only 20 ms when images
 
 ## Documentation
 
-* [Introductory slides](https://www.dropbox.com/s/10fx16yp5etb8dv/caffe-presentation.pdf): slides about the Caffe architecture, *updated 03/14*.
+* [Introductory slides](http://caffe.berkeleyvision.org/caffe-presentation.pdf): slides about the Caffe architecture, *updated 03/14*.
 * [Installation](/installation.html): Instructions on installing Caffe (works on Ubuntu, Red Hat, OS X).
 * [Pre-trained models](/getting_pretrained_models.html): BVLC provides some pre-trained models for academic / non-commercial use.
 * [Development](/development.html): Guidelines for development and contributing to Caffe.

diff --git a/examples/imagenet/alexnet_deploy.prototxt b/examples/imagenet/alexnet_deploy.prototxt
@@ -0,0 +1,260 @@
+name: "AlexNet"
+input: "data"
+input_dim: 10
+input_dim: 3
+input_dim: 227
+input_dim: 227
+layers {
+  layer {
+    name: "conv1"
+    type: "conv"
+    num_output: 96
+    kernelsize: 11
+    stride: 4
+    blobs_lr: 1.
+    blobs_lr: 2.
+    weight_decay: 1.
+    weight_decay: 0.
+  }
+  bottom: "data"
+  top: "conv1"
+}
+layers {
+  layer {
+    name: "relu1"
+    type: "relu"
+  }
+  bottom: "conv1"
+  top: "conv1"
+}
+layers {
+  layer {
+    name: "norm1"
+    type: "lrn"
+    local_size: 5
+    alpha: 0.0001
+    beta: 0.75
+  }
+  bottom: "conv1"
+  top: "norm1"
+}
+layers {
+  layer {
+    name: "pool1"
+    type: "pool"
+    pool: MAX
+    kernelsize: 3
+    stride: 2
+  }
+  bottom: "norm1"
+  top: "pool1"
+}
+layers {
+  layer {
+    name: "conv2"
+    type: "conv"
+    num_output: 256
+    group: 2
+    kernelsize: 5
+    pad: 2
+    blobs_lr: 1.
+    blobs_lr: 2.
+    weight_decay: 1.
+    weight_decay: 0.
+  }
+  bottom: "pool1"
+  top: "conv2"
+}
+layers {
+  layer {
+    name: "relu2"
+    type: "relu"
+  }
+  bottom: "conv2"
+  top: "conv2"
+}
+layers {
+  layer {
+    name: "norm2"
+    type: "lrn"
+    local_size: 5
+    alpha: 0.0001
+    beta: 0.75
+  }
+  bottom: "conv2"
+  top: "norm2"
+}
+layers {
+  layer {
+    name: "pool2"
+    type: "pool"
+    pool: MAX
+    kernelsize: 3
+    stride: 2
+  }
+  bottom: "norm2"
+  top: "pool2"
+}
+layers {
+  layer {
+    name: "conv3"
+    type: "conv"
+    num_output: 384
+    kernelsize: 3
+    pad: 1
+    blobs_lr: 1.
+    blobs_lr: 2.
+    weight_decay: 1.
+    weight_decay: 0.
+  }
+  bottom: "pool2"
+  top: "conv3"
+}
+layers {
+  layer {
+    name: "relu3"
+    type: "relu"
+  }
+  bottom: "conv3"
+  top: "conv3"
+}
+layers {
+  layer {
+    name: "conv4"
+    type: "conv"
+    num_output: 384
+    group: 2
+    kernelsize: 3
+    pad: 1
+    blobs_lr: 1.
+    blobs_lr: 2.
+    weight_decay: 1.
+    weight_decay: 0.
+  }
+  bottom: "conv3"
+  top: "conv4"
+}
+layers {
+  layer {
+    name: "relu4"
+    type: "relu"
+  }
+  bottom: "conv4"
+  top: "conv4"
+}
+layers {
+  layer {
+    name: "conv5"
+    type: "conv"
+    num_output: 256
+    group: 2
+    kernelsize: 3
+    pad: 1
+    blobs_lr: 1.
+    blobs_lr: 2.
+    weight_decay: 1.
+    weight_decay: 0.
+  }
+  bottom: "conv4"
+  top: "conv5"
+}
+layers {
+  layer {
+    name: "relu5"
+    type: "relu"
+  }
+  bottom: "conv5"
+  top: "conv5"
+}
+layers {
+  layer {
+    name: "pool5"
+    type: "pool"
+    kernelsize: 3
+    pool: MAX
+    stride: 2
+  }
+  bottom: "conv5"
+  top: "pool5"
+}
+layers {
+  layer {
+    name: "fc6"
+    type: "innerproduct"
+    num_output: 4096
+    blobs_lr: 1.
+    blobs_lr: 2.
+    weight_decay: 1.
+    weight_decay: 0.
+  }
+  bottom: "pool5"
+  top: "fc6"
+}
+layers {
+  layer {
+    name: "relu6"
+    type: "relu"
+  }
+  bottom: "fc6"
+  top: "fc6"
+}
+layers {
+  layer {
+    name: "drop6"
+    type: "dropout"
+    dropout_ratio: 0.5
+  }
+  bottom: "fc6"
+  top: "fc6"
+}
+layers {
+  layer {
+    name: "fc7"
+    type: "innerproduct"
+    num_output: 4096
+    blobs_lr: 1.
+    blobs_lr: 2.
+    weight_decay: 1.
+    weight_decay: 0.
+  }
+  bottom: "fc6"
+  top: "fc7"
+}
+layers {
+  layer {
+    name: "relu7"
+    type: "relu"
+  }
+  bottom: "fc7"
+  top: "fc7"
+}
+layers {
+  layer {
+    name: "drop7"
+    type: "dropout"
+    dropout_ratio: 0.5
+  }
+  bottom: "fc7"
+  top: "fc7"
+}
+layers {
+  layer {
+    name: "fc8"
+    type: "innerproduct"
+    num_output: 1000
+    blobs_lr: 1.
+    blobs_lr: 2.
+    weight_decay: 1.
+    weight_decay: 0.
+  }
+  bottom: "fc7"
+  top: "fc8"
+}
+layers {
+  layer {
+    name: "prob"
+    type: "softmax"
+  }
+  bottom: "fc8"
+  top: "prob"
+}
diff --git a/examples/imagenet/alexnet_solver.prototxt b/examples/imagenet/alexnet_solver.prototxt
@@ -0,0 +1,14 @@
+train_net: "alexnet_train.prototxt"
+test_net: "alexnet_val.prototxt"
+test_iter: 1000
+test_interval: 1000
+base_lr: 0.01
+lr_policy: "step"
+gamma: 0.1
+stepsize: 100000
+display: 20
+max_iter: 450000
+momentum: 0.9
+weight_decay: 0.0005
+snapshot: 10000
+snapshot_prefix: "caffe_alexnet_train"