Hansimov

List GPUs in machine

sudo lshw -C display

  *-display UNCLAIMED
       description: VGA compatible controller
       product: GP102 [TITAN Xp]

Hansimov

import datetime
import functools
import inspect
import logging
import os
import shutil
import subprocess
from termcolor import colored

Hansimov

Following this guide

• https://docs.docker.com/engine/install/ubuntu/#install-using-the-repository

Add Docker's official GPG key:

sudo apt-get update
sudo apt-get install ca-certificates curl gnupg
sudo install -m 0755 -d /etc/apt/keyrings
curl -fsSL https://download.docker.com/linux/ubuntu/gpg | sudo gpg --dearmor -o /etc/apt/keyrings/docker.gpg

Hansimov

Get current time:

• https://time.is/zh/Beijing

sudo date -s "29 Nov 2023 11:25:00+08:00"

Hansimov

Push Github Repo to Huggingface Space

Create a huggingface token

• https://huggingface.co/settings/tokens

Put it in the repo secrets with name like HF_TOKEN in related Github repo setitngs:

• https://github.com/<user_name>/<repo_name>/settings/secrets/actions

Run following commands in the git root folder:

Hansimov

1. Create a new folder where would place the new repo
2. git clone the original repo under this folder
3. Download git-filter-repo from here: (Do not chagne its extension)
   • https://raw.githubusercontent.com/newren/git-filter-repo/main/git-filter-repo
   • Place git-filter-repo to the cloned repo root folder
4. cd the cloned repo root folder, and run following: python git-filter-repo --path <directory_to_separate_to_new_repo> --force
5. Copy follwing files to root path of the new folder created in step 1:
   • Remaining files under <directory_to_separate_to_new_repo>
   • .git directory in original repo root folder

Hansimov

Do following steps:

1. Apply a student account of Azure, with @*.edu email:
   • https://azure.microsoft.com/zh-cn/free/students/
   • View balance here:
     • https://www.microsoftazuresponsorships.com/Balance#
2. Start to create a B1S (Linux) server via this page:
   • https://portal.azure.com/#view/Microsoft_Azure_Billing/FreeServicesBlade
   • Must use this entrance. Create with default method would cost money.
3. Config the server and create:

Hansimov

import torch
import whisper

from pathlib import Path
from termcolor import colored


class SpeechToTextConverter:
    def __init__(self, folders=["."], exts=[".mp3"], files=None):
        self.folders = folders

Hansimov

Check if CUDA is available by torch:

import torch

def check_cuda():
    print(torch.version.cuda)
    cuda_is_ok = torch.cuda.is_available()
    print(f"CUDA Enabled: {cuda_is_ok}")

Hansimov

tmux

sudo apt install tmux

• .tmux.conf
  • https://gist.github.com/Hansimov/55f92e1f75923fee178d8a5e49e40143

miniconda