Merge branch 'release' of https://github.com/yihui-he/lowrank into re…

…lease

.gitignore

@@ -0,0 +1,102 @@
+#lib/cfgs.py
+*.pickle
+*.caffemodel
+*.solverstate
+*.prototxt
+!temp/vgg.prototxt
+!temp/solver.prototxt
+cifar-10-batches-py
+*.swp
+.vscode/
+core
+*.tar*
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+env/
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+# lib/
+lib64/
+parts/
+sdist/
+var/
+*.egg-info/
+.installed.cfg
+*.egg
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*,cover
+.hypothesis/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+target/
+
+# IPython Notebook
+.ipynb_checkpoints
+
+# pyenv
+.python-version
+
+# celery beat schedule file
+celerybeat-schedule
+
+# dotenv
+.env
+
+# virtualenv
+.venv/
+venv/
+ENV/
+
+# Spyder project settings
+.spyderproject
+
+# Rope project settings
+.ropeproject

.gitmodules

@@ -0,0 +1,3 @@
+[submodule "caffe"]
+	path = caffe
+	url = https://github.com/yihui-he/caffe-pro

README.md

@@ -0,0 +1,72 @@
+# Channel Pruning for Accelerating Very Deep Neural Networks
+By [Yihui He](http://yihui-he.github.io/) (Xi'an Jiaotong University), Xiangyu Zhang and [Jian Sun](http://jiansun.org/) (Megvii)
+
+In this repository, we illustrate channel pruning VGG-16 **4X** with our 3C method, after finetuning the Top-5 accuracy is **89.9%** (suffers no performance degradation).
+
+![i2](http://yihui-he.github.io/assets_files/structure-1.png) | ![i1](http://yihui-he.github.io/assets_files/ill-1.png)
+:-------------------------:|:-------------------------:
+Structured simplification methods             |  Channel pruning (d)
+
+### Citation
+If you find the code useful in your research, please consider citing:
+
+    @article{he2017channel,
+      title={Channel Pruning for Accelerating Very Deep Neural Networks},
+      author={He, Yihui and Zhang, Xiangyu and Sun, Jian},
+      journal={arXiv preprint arXiv:1707.06168},
+      year={2017}
+    }
+
+### Contents
+1. [Requirements](#requirements)
+2. [Installation](#installation-sufficient-for-the-demo)
+3. [Channel Pruning and finetuning](#channel-pruning)  
+
+### requirements
+1. Python3 packages you might not have: `scipy`, `sklearn`, `easydict`
+2. For finetuning with 128 batch size, 4 GPUs (~11G of memory)
+
+### Installation (sufficient for the demo)
+1. Clone the repository
+    ```Shell
+    # Make sure to clone with --recursive
+    git clone --recursive https://github.com/yihui-he/channel-pruning.git
+    ```
+2. Build my Caffe fork
+    ```Shell
+    cd caffe
+
+    # If you're experienced with Caffe and have all of the requirements installed, then simply do:
+    make -j8 && make pycaffe
+    # Or follow the Caffe installation instructions here:
+    # http://caffe.berkeleyvision.org/installation.html
+    ```
+3. Download ImageNet classification dataset
+    http://www.image-net.org/download-images  
+   Specify imagenet `source` path in `temp/vgg.prototxt` (line 12 and 36)
+
+### Channel Pruning  
+1. Download the original VGG-16 model
+    http://www.robots.ox.ac.uk/~vgg/software/very_deep/caffe/VGG_ILSVRC_16_layers.caffemodel  
+   move it to `temp/vgg.caffemodel` (or create a softlink instead)
+
+2. Start Channel Pruning
+    ```Shell
+    python3 train.py -action c3 -gpu GPU0
+    # or log it with ./run.sh python3 train.py -action c3 -gpu GPU0
+    ```
+3. Combine some factorized layers for further compression, and calculate the acceleration ratio
+    ```Shell
+    ./combine.sh | xargs ./calflop.sh
+    ```
+
+4. Finetuning
+    ```Shell
+    ./finetune.sh GPU0,GPU1,GPU2,GPU3
+    ```
+
+5. Testing
+    Though testing is done while finetuning, you can test anytime with:
+    ```Shell
+    caffe test -model path/to/prototxt -weights path/to/caffemodel -iterations 5000 -gpu GPU0
+    ```

calflop.sh

@@ -0,0 +1,2 @@
+python3 tools.py flop $@
+#-model temp/cb_3c_3C4x_mem_bn_vgg.prototxt -weights ../caffe/models/vgg16/cb_3c_VGG16.v2.caffemodel

combine.sh

@@ -0,0 +1 @@
+python3 train.py -action combine -prototxt temp/3c_3C4x_mem_bn_vgg.prototxt -weights temp/3c_vgg.caffemodel

finetune.sh

@@ -0,0 +1 @@
+caffe train -solver temp/solver.prototxt -weights temp/3c_vgg.caffemodel -gpu $1