Fix for CURAND issue

README.md

@@ -12,26 +12,55 @@ Please see <http://zju-capg.org/myo> for details.
 * [Docker](http://docker.io/)
 * [Nvidia Docker](https://github.com/NVIDIA/nvidia-docker)
 
-## Usage
+## Quick Usage
 
 Following commands will
 (1) pull docker image (see `docker/Dockerfile` for details);
 (2) train ConvNets on the training sets of NinaPro DB1, CapgMyo DB-a and CSL-HDEMG, respectively;
 and (3) test trained ConvNets on the test sets.
 
+Navigate to the downloaded version of this repo then run the following commands to use images from Docker Hub
 ```
 mkdir .cache
 # put NinaPro DB1 in .cache/ninapro-db1
 # put CapgMyo DB-a in .cache/dba
 # put CSL-HDEMG in .cache/csl
-docker pull answeror/sigr:2016-09-21
-scripts/trainsrep.sh
-scripts/testsrep.sh
+docker pull lif3line/sigr:latest
+sudo scripts/trainsrep.sh
+sudo scripts/testsrep.sh
 ```
 
+## Building from Source
+
+Alternatively you can rebuild the images from the dockerfiles in this repo:
+```
+mkdir .cache
+# put NinaPro DB1 in .cache/ninapro-db1
+# put CapgMyo DB-a in .cache/dba
+# put CSL-HDEMG in .cache/csl
+cd docker/mxnet
+sudo docker build .
+# Note #ID of image on completion
+cd ..
+```
+Edit the `Dockerfile` in this directory replacing `FROM lif3line/mxnet:latest` with `FROM #ID:latest`
+```
+sudo docker build .
+# Note #ID of this new image
+cd ..
+cd scripts
+```
+Edit `runsrep` replacing `lif3line/sigr:latest` with `#ID:latest`
+```
+cd ..
+sudo scripts/trainsrep.sh
+sudo scripts/testsrep.sh
+```
+
+## Notes
 Training on NinaPro and CapgMyo will take 1 to 2 hours depending on your GPU.
 Training on CSL-HDEMG will take several days.
-You can accelerate traning and testing by distribute different folds on different GPUs with the `gpu` parameter.
+You can accelerate training and testing by distribute different folds on different GPUs with the `gpu` parameter.
 
 The NinaPro DB1 should be segmented according to the gesture labels and stored in Matlab format as follows.
 `.cache/ninapro-db1/data/sss/ggg/sss_ggg_ttt.mat` contains a field `data` (frames x channels) represents the trial `ttt` of gesture `ggg` of subject `sss`.

docker/Dockerfile

@@ -1,4 +1,5 @@
-FROM answeror/mxnet:f2684a6
+# Self Made mxnet docker container
+FROM lif3line/mxnet:latest
 MAINTAINER answeror <[email protected]>
 
 RUN apt-get install -y python-pip python-scipy
@@ -27,6 +28,10 @@ RUN cd /mxnet && \
   make clean && \
   make -j8 ADD_LDFLAGS=-L/usr/local/cuda/lib64/stubs
 
+# Necessary for newer modules
+RUN pip install --upgrade setuptools pip
+# Hack to avoid needing python 2.7.9 security features
+RUN pip install tornado==4.0
 RUN pip install jupyter pandas matplotlib seaborn scikit-learn
 RUN mkdir -p -m 700 /root/.jupyter/ && \
   echo "c.NotebookApp.ip = '*'" >> /root/.jupyter/jupyter_notebook_config.py

docker/mxnet/Dockerfile

@@ -1,4 +1,5 @@
-FROM nvidia/cuda:7.5-cudnn5-devel
+# Use CUDA 8.0 to avoid issue on newer GPUs
+FROM nvidia/cuda:8.0-cudnn5-devel-ubuntu14.04 
 MAINTAINER answeror <[email protected]>
 
 RUN echo "deb http://mirrors.zju.edu.cn/ubuntu/ trusty main restricted universe multiverse" > /etc/apt/sources.list && \
@@ -23,7 +24,9 @@ RUN apt-get update && apt-get install -y \
   wget \
   unzip
 RUN git clone --recursive https://github.com/dmlc/mxnet/ && cd mxnet && \
-  git checkout f2684a6 && \
+  git checkout 25720d0 && \
+  git submodule update --init && \
+  git submodule update --recursive && \
   sed -i -e 's/CHECK(ksize_x <= dshape\[3\] && ksize_y <= dshape\[2\])/CHECK(ksize_x <= dshape[3] + 2 * param_.pad[1] \&\& ksize_y <= dshape[2] + 2 * param_.pad[0])/' src/operator/convolution-inl.h && \
   cp make/config.mk . && \
   echo "USE_CUDA=1" >>config.mk && \

scripts/runsrep

@@ -1,3 +1,3 @@
 #!/usr/bin/env bash
 
-nvidia-docker run -ti -v $(pwd):/code answeror/sigr:2016-09-21 $@
+nvidia-docker run -ti -v $(pwd):/code lif3line/sigr:latest $@