banach-space

// The following matmul maps very nicely onto SME with SVL=128bits. For f32, there are 4 4x4 tiles,
// that can be assembled as a 8x8 matrix.

// %mat_A_tr - transpose of A
func.func @matmul(%mat_A_tr: memref<6x8xf32>, %mat_B: memref<6x8xf32>, %mat_C: memref<8x8xf32>) {
  linalg.matmul ins(%mat_A_tr, %mat_B: memref<6x8xf32>, memref<6x8xf32>)
            outs(%mat_C: memref<8x8xf32>)
  return
}

caspg

defmodule TravelerWeb.SearchbarLive do
  use TravelerWeb, :live_view

  alias Phoenix.LiveView.JS
  alias Traveler.Places

  def mount(_params, _session, socket) do
    socket = assign(socket, places: [])
    {:ok, socket, layout: false}
  end

shingohry

import UIKit
import Combine

/*
 # References
 - [ra1028/SwiftUI-Combine: This is an example project of SwiftUI and Combine using GitHub API.](https://github.com/ra1028/SwiftUI-Combine)
 - [marty-suzuki/GitHubSearchWithSwiftUI: SwiftUI and Combine based GitHubSearch example.](https://github.com/marty-suzuki/GitHubSearchWithSwiftUI)
 - [DataTaskPublisherを作ってみた](https://gist.github.com/yamoridon/16c1cc70ac46e50def4ca6695ceff772)
 - [【iOS】Combineフレームワークまとめ(2019/6/9時点) - Qiita](https://qiita.com/shiz/items/5efac86479db77a52ccc)
 */

alexito4

import SwiftUI
import Combine
import AlamofireImage

let imageDownloader = ImageDownloader(
    configuration: ImageDownloader.defaultURLSessionConfiguration(),
    downloadPrioritization: .fifo,
    maximumActiveDownloads: 4,
    imageCache: AutoPurgingImageCache()
)

mjambon

#! /usr/bin/env bash
#
# Run parallel commands and fail if any of them fails.
#

set -eu

pids=()

for x in 1 2 3; do

TomFaulkner

Ubuntu 18.04 VFIO GPU passthrough with a single GPU (onboard automatically disables itself)

Hardware

https://www.reddit.com/r/JDM_WAAAT/comments/8zgkfj/server_build_nas_killer_v_20_the_terminator_dual/ 2x X5650 from this thread Aside from that build guide, a generic USB3 card and an AMD R9-200

Grub

Modify /etc/default/grub and add vfio-pci.ids=dead:beef video=efifb:off to GRUB_CMDLINE_LINUX_DEFAULT (use the PCI vid/pid of your primary GPU, not dead:beef)

gangliao

import matplotlib.pyplot as plt
#for plotting
import numpy as np 

# create plot
fig, ax = plt.subplots()
bar_width = 0.15
opacity = 0.8

xlabel= np.array([8, 16, 32, 64])

peterroelants

from __future__ import division, print_function, absolute_import, \
    unicode_literals

import tensorflow as tf
from tensorflow.examples.tutorials.mnist import input_data as mnist_data
from tensorflow.contrib import slim
from tensorflow.contrib.learn import ModeKeys
from tensorflow.contrib.learn.python.learn import learn_runner

multiplemonomials

# - Find Intel MKL
# modified for AMBER
# Find the MKL libraries
#
# NOTE: MKL_MULTI_THREADED requires the patched FindOpenMPFixed module from the Amber-MD/cmake-buildscripts repository.
# 
# Options:
#
#   MKL_STATIC        :   use static linking.  Requires linker support for the -Wl,--start-group flag.
#   MKL_MULTI_THREADED:   use multi-threading. Requires the FindOpenMP module 

rafaspadilha

How to create a custom Caffe layer in Python?

This tutorial will guide through the steps to create a simple custom layer for Caffe using python. By the end of it, there are some examples of custom layers.

- Why would I want to do that?

Usually you would create a custom layer to implement a funcionality that isn't available in Caffe, tuning it for your requirements.

- What will I need?

Probably just Python and Caffe installed.

- Is there any downside?