Skip to content

Instantly share code, notes, and snippets.

Youssef Kashef kashefy

  • Berlin, Germany
Block or report user

Report or block kashefy

Hide content and notifications from this user.

Learn more about blocking users

Contact Support about this user’s behavior.

Learn more about reporting abuse

Report abuse
View GitHub Profile
kashefy /
Created Sep 7, 2017
Try out saving and restoring Tensorflow variables defined using tf.get_variable
Created on Aug 18, 2017
@author: kashefy
import os
import shutil
import tensorflow as tf
from tensorflow.examples.tutorials.mnist import input_data
import numpy as np
import tensorflow as tf
# adapted from
if __name__ == '__main__':
path_to_events_file = '/home/kashefy/models/ae/log_simple_stats/2017-07-18_12-22-45/events.out.tfevents.1500373527.hobbes'
tag = 'mlp_layer-1/loss_classification_1'
# with open(fpath, "r") as f:
kashefy / find_circV
Created Jan 6, 2017
grep -nHIrF --include=*.m -- circV
View find_circV
./auditory-front-end/src/Processors/pitchProc.m:236: pObj.maxConfBuf = circVBuf(bufferDurSec*pObj.FsHzIn,1);
./auditory-front-end/src/Processors/pitchProc.m:237: pObj.maxConf = circVBufArrayInterface(pObj.maxConfBuf);
./auditory-front-end/src/Signals/circVBuf.m:1:classdef circVBuf < handle
./auditory-front-end/src/Signals/circVBuf.m:2: %circVBuf class defines a circular double buffered vector buffer
./auditory-front-end/src/Signals/circVBuf.m:51: % example 1 - loop over new vectors of a circVBuf (slow because of copy):
./auditory-front-end/src/Signals/circVBuf.m:52: % for
./auditory-front-end/src/Signals/circVBuf.m:53: % vec(:) = circVBuf.VBuf.raw(:,bId);
./auditory-front-end/src/Signals/circVBuf.m:57: % new =;
./auditory-front-end/src/Signals/circVBuf.m:58: % lst = circVBuf.VBuf.lst;
./auditory-front-end/src/Signals/circVBuf.m:59: % mean = mean(circVBuf.VBuf.raw(3:7,new:lst));
kashefy /
Last active Jul 29, 2016
Moving average for a time series
import numpy as np
def moving_avg(x, window_size):
window = np.ones(int(window_size)) / float(window_size)
return np.convolve(x, window, 'valid')
kashefy /
Last active May 26, 2016
downsample sound files using ffmpeg
def is_wav(p):
return os.path.splitext(p)[-1] == '.wav'
def lower_samplerate_cmds(dir_src, samplerate, dir_dst, fpath_exec_script):
paths = fs.gen_paths(dir_src, is_wav)
commands = []
for fpath in paths:
path_dst = os.path.join(dir_dst, os.path.basename(fpath))
kashefy / 59_labels.txt
Last active Feb 20, 2016
The 59 labels subset for PASCAL Context dataset. Originally from here:
View 59_labels.txt
1: aeroplane
2: bicycle
3: bird
4: boat
5: bottle
6: bus
7: car
8: cat
9: chair
10: cow
kashefy / mnist_ae_tied_weights_trainval.prototxt
Last active Jan 29, 2016
network definition for an Auto-encoder with tied weights. The weight are tied by sharing parameter names in 'permissive mode' and setting the decoder's ip params to use tranposed weigts.
View mnist_ae_tied_weights_trainval.prototxt
name: "MNISTAutoencoderTiedWeights"
layer {
name: "mnist"
type: "Data"
top: "data"
top: "label"
include {
phase: TRAIN
transform_param {
View print_caffe_blob.cpp
template <typename Dtype>
void printMat(const Dtype* data, int rows, int cols) {
int j = 0;
for (int r=0; r<rows; r++) {
for (int c=0; c<cols; c++) {
if (c<cols-1) std::cout<<",";
kashefy /
Created Jan 13, 2016
2d convolution using numpy
Created on Jul 13, 2015
@author: kashefy
import numpy as np
from scipy import signal
if __name__ == '__main__':
# Installing Caffe dependencies without sudo privileges:
# The prefix serves as the installation directory for caffe dependencies
# that are not already installed on the system
# Set "inputs" according to host machine
export PREFIX_HOST=/mnt/scratch/$USER/caffe_deps
export PREFIX_CUDNN=$PREFIX_HOST/../build/cudnn-7.5
You can’t perform that action at this time.