vv111y

Naming axis in tensors (Thu Jan 10 2019)

• Tensor Considered Harmful Part I (http://nlp.seas.harvard.edu/NamedTensor), Part II http://nlp.seas.harvard.edu/NamedTensor2
  • First discussion on the topic : [Proposal] Named Axes/Dimensions or Tensor Shape Annotations · Issue #4164 · pytorch/pytorch (pytorch/pytorch#4164)
  • Repo : harvardnlp/namedtensor: Proof of concept for a dynamic named tensor for pytorch (https://github.com/harvardnlp/namedtensor)
  • Discussion:
    • [D] Tensor Considered Harmful (A polemic against numpy / pytorch and a proposal for a named tensor) : MachineLearning (https://www.reddit.com/r/MachineLearning/comments/accmek/d_tensor_considered_harmful_a_polemic_against/)
    • Tensor considered harmful | Hacker News (https://news.ycombinator.com/item?id=18823777)
    • harvardnlp on Twitter: “”Tensor Considered Harmful” (https://t.co/iueFvrYT6O). A polemic against numpy / pytorch and a proposal for a named tensor (https://t.co/MVBUm7OyBq). (New year’

vv111y

Naming axis in tensors (Thu Jan 10 2019)

• Tensor Considered Harmful Part I (http://nlp.seas.harvard.edu/NamedTensor), Part II http://nlp.seas.harvard.edu/NamedTensor2
  • First discussion on the topic : [Proposal] Named Axes/Dimensions or Tensor Shape Annotations · Issue #4164 · pytorch/pytorch (pytorch/pytorch#4164)
  • Repo : harvardnlp/namedtensor: Proof of concept for a dynamic named tensor for pytorch (https://github.com/harvardnlp/namedtensor)
  • Discussion:
    • [D] Tensor Considered Harmful (A polemic against numpy / pytorch and a proposal for a named tensor) : MachineLearning (https://www.reddit.com/r/MachineLearning/comments/accmek/d_tensor_considered_harmful_a_polemic_against/)
    • Tensor considered harmful | Hacker News (https://news.ycombinator.com/item?id=18823777)
    • harvardnlp on Twitter: “”Tensor Considered Harmful” (https://t.co/iueFvrYT6O). A polemic against numpy / pytorch and a proposal for a named tensor (https://t.co/MVBUm7OyBq). (New year’

vv111y

naming axis in tensors (Thu Jan 10 2019) NVIDIA/OpenSeq2Seq: Toolkit for efficient experimentation with various sequence-to-sequence models (https://github.com/NVIDIA/OpenSeq2Seq) ctongfei/nexus: Experimental typesafe tensors / deep learning / probabilistic programming in Scala (https://github.com/ctongfei/nexus) harvardnlp/namedtensor: Proof of concept for a dynamic named tensor for pytorch (https://github.com/harvardnlp/namedtensor) Tensor Considered Harmful (http://nlp.seas.harvard.edu/NamedTensor) [D] Tensor Considered Harmful (A polemic against numpy / pytorch and a proposal for a named tensor) : MachineLearning (https://www.reddit.com/r/MachineLearning/comments/accmek/d_tensor_considered_harmful_a_polemic_against/) harvardnlp on Twitter: “”Tensor Considered Harmful” (https://t.co/iueFvrYT6O). A polemic against numpy / pytorch and a proposal for a named tensor (https://t.co/MVBUm7OyBq). (New year’s goal, be more troublesome.)… https://t.co/fLmk8RR4Xy” (http

vv111y

#!/usr/bin/env python
import sys
if sys.version_info[0] >= 3:
    import PySimpleGUI as sg
else:
    import PySimpleGUI27 as sg
import cv2 as cv
from PIL import Image
import io
from sys import exit as exit

vv111y

;;; This file is loaded by Spacemacs at startup.
;; It must be stored in your home directory.
;; -*- mode: emacs-lisp -*-


;;; defun dotspacemacs/layers
(defun dotspacemacs/layers ()
  "Configuration Layers declaration.
You should not put any user code in this function besides modifying the variable
values."

vv111y

friedman.test.with.post.hoc <- function(data, alpha = 0.05)
{
  print("Check if you missing the packages 'graph' and 'Rgraphviz'. Try to install them using bioconductor")
  
  #source("http://bioconductor.org/biocLite.R")
  #biocLite(c("graph","Rgraphviz"))
  
  # Loading needed packages
  if(!require(ggplot2))
  {

vv111y

2018 - Machine learning paradigms

Abdolhosseini et al_2018_Cell Identity Codes

AlQuraishi_2018_End-to-end differentiable learning of protein structure

vv111y

local alert    = require("hs.alert")
local timer    = require("hs.timer")
local eventtap = require("hs.eventtap")

local events   = eventtap.event.types

local module   = {}

-- ** Changed original gist from "ctrl" to "cmd" **
-- Save this in your Hammerspoon configuration directiorn (~/.hammerspoon/) 

vv111y

def windowfn(X):
    return X

def MDCT(X, blocksize=64):
    originalN = len(X)
    N = blocksize

    # zero pad the signal right away. Calls within FFT functions will be
    # effectively ignored.
    nzeros, X = padData(X)

vv111y

EDIT3: server verbose log. I removed the contents of ```user-key-allowed``` in the log, not sure if they are sensitive or not
    
    Jul 27 14:01:49 gpgpu sshd[15719]: Connection from 192.168.0.200 port 56667 on xxx.xxx.xxx.xxx port 22
    Jul 27 14:01:49 gpgpu sshd[15719]: debug1: Client protocol version 2.0; client software version OpenSSH_6.9
    Jul 27 14:01:49 gpgpu sshd[15719]: debug1: match: OpenSSH_6.9 pat OpenSSH* compat 0x04000000
    Jul 27 14:01:49 gpgpu sshd[15719]: debug1: Enabling compatibility mode for protocol 2.0
    Jul 27 14:01:49 gpgpu sshd[15719]: debug1: Local version string SSH-2.0-OpenSSH_7.2p2 Ubuntu-4ubuntu2.2
    Jul 27 14:01:49 gpgpu sshd[15719]: debug2: fd 3 setting O_NONBLOCK
    Jul 27 14:01:49 gpgpu sshd[15719]: debug2: Network child is on pid 15720