| use std::str; | |
| fn main() { | |
| // -- FROM: vec of chars -- | |
| let src1: Vec<char> = vec!['j','{','"','i','m','m','y','"','}']; | |
| // to String | |
| let string1: String = src1.iter().collect::<String>(); | |
| // to str | |
| let str1: &str = &src1.iter().collect::<String>(); | |
| // to vec of byte |
| # /------------------------------------------\ | |
| # | don't forget to download the .tp file | | |
| # | and place it in the user's directory :› | | |
| # | | | |
| # | also install lolcat: | | |
| # | https://github.com/busyloop/lolcat | | |
| # \------------------------------------------/ | |
| alias test-passed='if [ "$?" -eq "0" ]; then lolcat ~/.tp -a -s 40 -d 2; fi;' |
This is a short post that explains how to write a high-performance matrix multiplication program on modern processors. In this tutorial I will use a single core of the Skylake-client CPU with AVX2, but the principles in this post also apply to other processors with different instruction sets (such as AVX512).
Matrix multiplication is a mathematical operation that defines the product of
| 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]) |
| \usepackage{listings} | |
| \usepackage{color} | |
| \definecolor{dkgreen}{rgb}{0,0.6,0} | |
| \definecolor{gray}{rgb}{0.5,0.5,0.5} | |
| \definecolor{mauve}{rgb}{0.58,0,0.82} | |
| \lstset{language=SQL, | |
| basicstyle={\small\ttfamily}, | |
| belowskip=3mm, | |
| breakatwhitespace=true, |
| """ | |
| Author: Awni Hannun | |
| This is an example CTC decoder written in Python. The code is | |
| intended to be a simple example and is not designed to be | |
| especially efficient. | |
| The algorithm is a prefix beam search for a model trained | |
| with the CTC loss function. |
| import gc | |
| import os | |
| import time | |
| import numpy as np | |
| import pandas as pd | |
| from pyarrow.compat import guid | |
| import pyarrow as pa | |
| import pyarrow.parquet as pq | |
| import snappy |
| #!/bin/sh | |
| # ngrok's web interface is HTML, but configuration is bootstrapped as a JSON | |
| # string. We can hack out the forwarded hostname by extracting the next | |
| # `*.ngrok.io` string from the JSON | |
| # | |
| # Brittle as all get out--YMMV. If you're still reading, usage is: | |
| # | |
| # $ ./ngrok_hostname.sh <proto> <addr> | |
| # |
| etcd-lb: | |
| image: rancher/load-balancer-service | |
| labels: | |
| io.rancher.scheduler.affinity:host_label: cluster=load-balancers | |
| ports: | |
| - 2379 | |
| links: | |
| - etcd0 | |
| - etcd1 | |
| - etcd2 |
