View censored normal moments.ipynb
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
View patch_biblatex.sty
% biblatex authoryear with natbib does a million stray commas in versions < 3.5
% see: https://tex.stackexchange.com/q/312539/
% this works around it for versions 3.3 and 3.4
\ifthenelse{\equal{\abx@version}{3.4}}{%
\DeclareDelimFormat[cbx@textcite]{nameyeardelim}{\addspace}
}{\ifthenelse{\equal{\abx@version}{3.3}}{%
% Workaround following http://golatex.de/neue-biber-version-ploetzlich-komma-bei-zitierung-t17005.html
% (in German...), which changes \nameyeardelim / \nonameyeardelim to \addspace
% in the citation style's textcite macro.
% sigh....
View alphanumeric-hybrid.bbx
% This is an unholy mashup of the numeric.bbx and alphabetic.bbx files from
% the standard biblatex distribution, made so we can switch between numeric
% style for the main body and alphabetic style for the appendix.
\ProvidesFile{alphanumeric-hybrid.bbx}
[\abx@bbxid]
\RequireBibliographyStyle{standard}
\ExecuteBibliographyOptions{labelalpha,labelnumber,sorting=anyt}
% To use the alphabetic section in one refsection, \setbool{bbx:alpha}{true} at the start;
View two-sample tests via set classification.ipynb
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
View two-sample tests via set classification.ipynb
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
View mkl dgemm_batch.ipynb
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
View autoencoder_stuff.py
from __future__ import division
import json
import numpy as np
import keras
from keras.datasets import mnist
from keras import backend as K, layers as L
from keras.models import load_model, Model, model_from_config
View fix_paths.sh
#!/bin/bash
CUDA_LIB_DIR=/usr/local/cuda/lib
CUDA_VERSION=7.5
CUDA_LIBS="cublas cudart curand"
CUDNN_LIB_DIR=/usr/local/cuda/cudnn-3/lib
CUDNN_VERSION=7.0
CUDNN_LIBS="cudnn"
View monitor.sh
#!/bin/bash
pid=$1
echo "Watching for spot instance termination; will SIGHUP $pid if so."
while kill -0 $pid >/dev/null 2>/dev/null; do
if curl -s http://169.254.169.254/latest/meta-data/spot/termination-time | grep -q .*T.*Z; then
echo 'oh noes! telling caffe to snapshot.'
kill -SIGHUP $pid
fi
View fastfood.py
'''
Implementation of Fastfood (Le, Sarlos, and Smola, ICML 2013).
Primarily by @esc (Valentin Haenel) and felixmaximilian
from https://github.com/scikit-learn/scikit-learn/pull/3665.
Modified by @dougalsutherland.
FHT implementation was "inspired by" https://github.com/nbarbey/fht.
'''