Cypherpunk Samurai CypherpunkSamurai

Devuan Bootstrap Installation Guide

This guide is for installing Devuan Unstable (Ceres) edition using debootstrap for a more minimal setup. This process is for UEFI amd64 system.

Problem

I was training a deep learning model on Google Colab. I had to go somewhere for an emergency work. When I came back, I realized that Colab was disconnected; Therefore, I had to run the whole model again.

Solution

I searched on the Internet to solve this problem. I noticed that Colab will be disconnected if you don't click on the page. Here is an answer that I find:

function ClickConnect() {
  console.log('Working')
 document

Pre-requisites

Your phone must have the developer mode unlocked
You must unlock the USB debug

1 - Download the SDK

The last SDK is available in https://dl.google.com/android/repository/platform-tools-latest-windows.zip

2 - Extract to any folder

Please do a favor to yourself and dont extract to root of your C drive.

	# Install bitsandbytes:
	# `nvcc --version` to get CUDA version.
	# `pip install -i https://test.pypi.org/simple/ bitsandbytes-cudaXXX` to install for current CUDA.
	# Example Usage:
	# Single GPU: torchrun --nproc_per_node=1 trainer/diffusers_trainer.py --model="CompVis/stable-diffusion-v1-4" --run_name="liminal" --dataset="liminal-dataset" --hf_token="hf_blablabla" --bucket_side_min=64 --use_8bit_adam=True --gradient_checkpointing=True --batch_size=1 --fp16=True --image_log_steps=250 --epochs=20 --resolution=768 --use_ema=True
	# Multiple GPUs: torchrun --nproc_per_node=N trainer/diffusers_trainer.py --model="CompVis/stable-diffusion-v1-4" --run_name="liminal" --dataset="liminal-dataset" --hf_token="hf_blablabla" --bucket_side_min=64 --use_8bit_adam=True --gradient_checkpointing=True --batch_size=10 --fp16=True --image_log_steps=250 --epochs=20 --resolution=768 --use_ema=True

	import argparse
	import socket
	import sys

	# Install bitsandbytes:
	# `nvcc --version` to get CUDA version.
	# `pip install -i https://test.pypi.org/simple/ bitsandbytes-cudaXXX` to install for current CUDA.
	# Example Usage:
	# Single GPU: torchrun --nproc_per_node=1 trainer/diffusers_trainer.py --model="CompVis/stable-diffusion-v1-4" --run_name="liminal" --dataset="liminal-dataset" --hf_token="hf_blablabla" --bucket_side_min=64 --use_8bit_adam=True --gradient_checkpointing=True --batch_size=1 --fp16=True --image_log_steps=250 --epochs=20 --resolution=768 --use_ema=True
	# Multiple GPUs: torchrun --nproc_per_node=N trainer/diffusers_trainer.py --model="CompVis/stable-diffusion-v1-4" --run_name="liminal" --dataset="liminal-dataset" --hf_token="hf_blablabla" --bucket_side_min=64 --use_8bit_adam=True --gradient_checkpointing=True --batch_size=10 --fp16=True --image_log_steps=250 --epochs=20 --resolution=768 --use_ema=True

	import argparse
	import socket
	import torch

	# Cool Snippet Storage Utilities

	* Gisto - https://web.gistoapp.com/
	* Masscode - https://masscode.io/
	* 3Cols - https://3cols.com/
	* Cacher - https://www.cacher.io/

	# What do i prefer?
	I'm currently using Gisto, as it's pretty cool and uses Github Gists for storage. I also like the VSCode extension for Masscode and would like to try it.

	from tqdm import tqdm
	import requests

	# file url
	url = 'https://dumps.wikimedia.org/enwiki/latest/enwiki-latest-stub-meta-history.xml.gz' #file size is '70.5GB'
	# stream true is required
	response = requests.get(url, stream=True)
	# total file size
	t = int(response.headers.get('content-length', 0))
	block_size = 1024**2 #1 Mbit

	import os


	sublime_binary_path = "/tmp/sublime_text" # FIXME: this is the absolute path to writable sublime_text binary.
	version_magic_string = "4126"
	sz_magic_string = 4
	version_magic_string_offset = 0x0002d78a # (Real offset from xxd)

	is_file_read = os.access(sublime_binary_path, os.R_OK)
	if not is_file_read:

	To rename WSL Distros on Windows follow the steps:

	1. Stop all instances of WSL
	On PowerShell run the command: wsl --shutdown

	2. Open Registry Editor and go to HKEY_CURRENT_USER\SOFTWARE\Microsoft\Windows\CurrentVersion\Lxss
	Each subfolder on Lxss represents a distro

	3. Locate and rename the distro you want
	Inside each distro folder you have the DistributionName, change it and click ok

	import os


	sublime_binary_path = "/tmp/sublime_text"
	version_magic_string = "/updates/4/stable_update_check?version=4121&platform=linux&arch=x64"
	sz_magic_string = 67
	version_magic_string_offset = 0x000106bd # (Real offset from xxd)

	is_file_read = os.access(sublime_binary_path, os.R_OK)
	if not is_file_read: