zhicongchen

def smart_procrustes_align_gensim(base_embed, other_embed, words=None):
    """
    Original script: https://gist.github.com/quadrismegistus/09a93e219a6ffc4f216fb85235535faf
    Procrustes align two gensim word2vec models (to allow for comparison between same word across models).
    Code ported from HistWords <https://github.com/williamleif/histwords> by William Hamilton <wleif@stanford.edu>.
        
    First, intersect the vocabularies (see `intersection_align_gensim` documentation).
    Then do the alignment on the other_embed model.
    Replace the other_embed model's syn0 and syn0norm numpy matrices with the aligned version.
    Return other_embed.

TengdaHan

Multi-node-training on slurm with PyTorch

What's this?

• A simple note for how to start multi-node-training on slurm scheduler with PyTorch.
• Useful especially when scheduler is too busy that you cannot get multiple GPUs allocated, or you need more than 4 GPUs for a single job.
• Requirement: Have to use PyTorch DistributedDataParallel(DDP) for this purpose.
• Warning: might need to re-factor your own code.
• Warning: might be secretly condemned by your colleagues because using too many GPUs.

jetnew

from keras.layers import LSTM, Dense, RepeatVector, TimeDistributed
from keras.models import Sequential

class LSTM_Autoencoder:
  def __init__(self, optimizer='adam', loss='mse'):
    self.optimizer = optimizer
    self.loss = loss
    self.n_features = 1
    
  def build_model(self):

koaning

import tensorflow as tf
import numpy as np
import uuid

x = tf.placeholder(shape=[None, 3], dtype=tf.float32)
nn = tf.layers.dense(x, 3, activation=tf.nn.sigmoid)
nn = tf.layers.dense(nn, 5, activation=tf.nn.sigmoid)
encoded = tf.layers.dense(nn, 2, activation=tf.nn.sigmoid)
nn = tf.layers.dense(encoded, 5, activation=tf.nn.sigmoid)
nn = tf.layers.dense(nn, 3, activation=tf.nn.sigmoid)

quadrismegistus

def smart_procrustes_align_gensim(base_embed, other_embed, words=None):
	"""Procrustes align two gensim word2vec models (to allow for comparison between same word across models).
	Code ported from HistWords <https://github.com/williamleif/histwords> by William Hamilton <wleif@stanford.edu>.
		(With help from William. Thank you!)

	First, intersect the vocabularies (see `intersection_align_gensim` documentation).
	Then do the alignment on the other_embed model.
	Replace the other_embed model's syn0 and syn0norm numpy matrices with the aligned version.
	Return other_embed.

nepsilon

How to change your commit messages in Git?

At some point you’ll find yourself in a situation where you need edit a commit message. That commit might already be pushed or not, be the most recent or burried below 10 other commits, but fear not, git has your back 🙂.

Not pushed + most recent commit:

git commit --amend

This will open your $EDITOR and let you change the message. Continue with your usual git push origin master.

ctokheim

Stacked and Grouped Bar Plot

Oddly enough ggplot2 has no support for a stacked and grouped (position="dodge") bar plot. The seaborn python package, although excellent, also does not provide an alternative. However, I knew it was surely possible to make such a plot in regular matplotlib. Matplotlib, although sometimes clunky, gives you enough flexibility to precisely place plotting elements which is needed for a stacked and grouped bar plot.

Below is a working example of making a stacked and grouped bar plot.

import seaborn as sns
import matplotlib.pyplot as plt
import numpy as np

aymanfarhat

from itertools import chain

intersection = []

sets = [[0,4,5,2,1],[1,3,6,2,4],[4,1,2,5,7,0]]

merged = list(chain.from_iterable(sets))

merged.sort()