Skip to content

Instantly share code, notes, and snippets.

View igorcosta's full-sized avatar
:octocat:
I may be slow to respond.

Igor Costa igorcosta

:octocat:
I may be slow to respond.
98 followers · 72 following
View GitHub Profile
@igorcosta
igorcosta / GPT_logprobs.ipynb
Created January 12, 2025 23:16 — forked from anjsimmo/GPT_logprobs.ipynb
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
@igorcosta
igorcosta / windows-on-mac.md
Created April 13, 2023 04:50 — forked from mislav/windows-on-mac.md
Install Windows 11 in a virtual machine on macOS with an M1 CPU

Based on https://docs.getutm.app/guides/windows/

  1. Get UTM for Mac - the App Store version is offered merely as a way to support the developers, since it's paid, but has the same features as the free build;
  2. Obtain scripts that bootstrap an arm64 Windows 11 22H2 (I've chosen Windows Home edition and not Pro);
  3. brew install aria2 cabextract wimlib cdrtools minacle/chntpw/chntpw;
  4. cd into the directory extracted in step 1 and bash uup_download_macos.sh - this will produce an ISO;
  5. In the UTM app, create a new “Virtualize” machine and mount the ISO file from the previous step;
  6. Check “Install drivers and SPICE tools”;
  7. Boot the VM and go through the Windows installer;
  8. Boot the installed OS and finish the SPICE tools installer that should pop up by itself, then reboot the OS;
@igorcosta
igorcosta / README.md
Created August 17, 2022 04:29 — forked from osy/README.md
Local caching for GitHub Actions self hosted runner using Squid Proxy

One of the biggest issues with using a self hosted GitHub runner is that actions that require downloading large amounts of data will bottleneck at the network. [actions/cache][1] does not support locally caching objects and artifacts stored on GitHub's servers will require a lot of bandwidth to fetch on every job. We can, however, set up a content proxy using [Squid][2] with SSL bumping to locally cache requests from jobs.

Patching Squid

A major challenge is that [actions/cache][1] uses Azure storage APIs which makes HTTP range requests. While Squid supports range requests, it is not good at caching them. There is an option, range_offset_limit none which, according to the [documentation][3]:

A size of 'none' causes Squid to always fetch the object from the beginning so it may cache the result. (2.0 style)

However, after extensive debugging, I discovered that the feature does not work if the client closes the connection. When range_offset_limit is set, Squid will make a full request to the server,

@igorcosta
igorcosta / MainActivity.java
Created July 3, 2018 04:19 — forked from mjohnsullivan/MainActivity.java
Android Wear activity that reads and displays sensor data from the device
package com.example.wear;
import android.app.Activity;
import android.hardware.Sensor;
import android.hardware.SensorEvent;
import android.hardware.SensorEventListener;
import android.hardware.SensorManager;
import android.os.Bundle;
import android.support.wearable.view.WatchViewStub;
import android.util.Log;
@igorcosta
igorcosta / gist:835152315d756377a9d297dcdad0f5ff
Created March 27, 2018 02:34 — forked from kyledrake/gist:d7457a46a03d7408da31
Creating a self-signed SSL certificate, and then verifying it on another Linux machine
# Procedure is for Ubuntu 14.04 LTS.
# Using these guides:
# http://datacenteroverlords.com/2012/03/01/creating-your-own-ssl-certificate-authority/
# https://turboflash.wordpress.com/2009/06/23/curl-adding-installing-trusting-new-self-signed-certificate/
# https://jamielinux.com/articles/2013/08/act-as-your-own-certificate-authority/
# Generate the root (GIVE IT A PASSWORD IF YOU'RE NOT AUTOMATING SIGNING!):
openssl genrsa -aes256 -out ca.key 2048
openssl req -new -x509 -days 7300 -key ca.key -sha256 -extensions v3_ca -out ca.crt
@igorcosta
igorcosta / elastic_bulk_ingest.py
Created January 6, 2017 00:49 — forked from scotthaleen/elastic_bulk_ingest.py
Bulk Index json to elastic search
from pyspark import SparkContext, SparkConf
import json
import argparse
def fn_to_doc(line):
try:
doc = {}
data = json.loads(line)
doc['data'] = data
@igorcosta
igorcosta / ElasticSearch.md
Created January 6, 2017 00:44 — forked from tokhi/ElasticSearch.md
Elastic Search in simple words

Elastic Search

Elasticsearch is a real time search engine where a change to an index will be propegated to the whole cluster within a second.

An elasticsearch cluster indicated as one or more nodes, collection of nodes containing all the data, default cluster name is elasticserach.

A node is a single server and part of a cluster, node participate in searching and indexing.

Index is collection of documents equavalent to a database within a relational system, index name must be lowercase Type is represetn a class = table

@igorcosta
igorcosta / elastic_cheat_sheet.md
Created January 6, 2017 00:42 — forked from quantizor/elastic_cheat_sheet.md
Elastic Cheat Sheet
@igorcosta
igorcosta / es-attach-full.py
Created January 6, 2017 00:38 — forked from stevehanson/es-attach-full.py
Interactive Python script to recursively index files in directory tree to elasticSearch using the elasticsearch-mapper-attachments (https://github.com/elasticsearch/elasticsearch-mapper-attachments) plugin to index files (pdf, docx, html, etc).
import os
import sys
# constants, configure to match your environment
HOST = 'http://localhost:9200'
INDEX = 'test'
TYPE = 'attachment'
TMP_FILE_NAME = 'tmp.json'
# for supported formats, see apache tika - http://tika.apache.org/1.4/formats.html
INDEX_FILE_TYPES = ['html','pdf', 'doc', 'docx', 'xls', 'xlsx', 'xml']
@igorcosta
igorcosta / onpubcom_array_date_sort.js
Created August 20, 2014 14:59 — forked from onpubcom/onpubcom_array_date_sort.js
<script type="text/javascript">
// First let's create an array of JavaScript Date
// objects.
// More info about the Date class:
// http://w3schools.com/js/js_obj_date.asp
var dates = [
new Date(2010, 4, 10, 10, 07, 16),
new Date(2010, 4, 8, 9, 16, 09),
new Date(2010, 3, 30, 0, 15, 49),