Skip to content

Instantly share code, notes, and snippets.

View bryant1410's full-sized avatar

Santiago Castro bryant1410

666 followers · 303 following
View GitHub Profile
@bryant1410
bryant1410 / layernorm_vs_fused.py
Created September 22, 2019 23:17 — forked from ptrblck/layernorm_vs_fused
import time
import torch
import torch.nn as nn
from apex.normalization import FusedLayerNorm
torch.backends.cudnn.benchmark = True
@bryant1410
bryant1410 / make-arXiv-package.sh
Last active January 24, 2017 01:31 — forked from robince/make-arXiv-package.sh
Script to prepare arXiv package of a document that depends on a recent texlive version of biblatex (using pdflatex)
#!/usr/bin/env bash
# This script is useful if:
# - you have a manuscript that you want to upload to the arXiv,
# - you are using biblatex, and
# - you are using a recent version of texlive while arXiv is still on texlive2011
#
# Put this file in a directory containing the manuscript you want to
# upload to arXiv.org, and adapt the paths below.
@bryant1410
bryant1410 / onehot_pandas_scikit.py
Last active August 29, 2015 14:07 — forked from kljensen/onehot_pandas_scikit.py
This function helps to do a one hot encoding of a pandas' dataframe instead of a features numpy matrix. This has some advantages, for instance the fact of knowing which new columns have been created (identifying them easily).
# -*- coding: utf-8 -*-
""" Small script that shows hot to do one hot encoding
of categorical columns in a pandas DataFrame.
See:
http://scikit-learn.org/dev/modules/generated/sklearn.preprocessing.OneHotEncoder.html#sklearn.preprocessing.OneHotEncoder
http://scikit-learn.org/dev/modules/generated/sklearn.feature_extraction.DictVectorizer.html
"""
import pandas
import random