Skip to content

Instantly share code, notes, and snippets.

View Randl's full-sized avatar

Evgenii Zheltonozhskii Randl

166 followers · 45 following
View GitHub Profile
@lucidyan
lucidyan / gpu-control.md
Last active March 19, 2023 09:37
Prevent NVIDIA GPUs' throttling on headless server

Prevent NVIDIA GPUs' throttling on headless server

  • Unlock manual fan & overclock settings
    sudo nvidia-xconfig -a --cool-bits=28 --allow-empty-initial-configuration
  • Reboot system
  • Create script /usr/local/bin/gpu-fan-control.sh
#!/bin/bash
@MInner
MInner / gpu_profile.py
Created September 12, 2017 16:11
A script to generate per-line GPU memory usage trace. For more meaningful results set `CUDA_LAUNCH_BLOCKING=1`.
import datetime
import linecache
import os
import pynvml3
import torch
print_tensor_sizes = True
last_tensor_sizes = set()
gpu_profile_fn = f'{datetime.datetime.now():%d-%b-%y-%H:%M:%S}-gpu_mem_prof.txt'
@Kylmakalle
Kylmakalle / telethon_get_participants_request.py
Created May 10, 2017 12:12
Get info about Telagram chat participants using https://github.com/LonamiWebs/Telethon
from telethon import TelegramClient
from telethon.tl.functions.channels import GetParticipantsRequest
from telethon.tl.types import ChannelParticipantsRecent
from telethon.utils import get_input_peer
api_id = 123456 # Use your own values here. https://my.telegram.org
api_hash = '0123456789abcdef0123456789abcdef'
phone_number = '+34600000000'
client = TelegramClient('%sessionname%', api_id, api_hash) # feel free to edit %sessionname% as you want
@ncullen93
ncullen93 / affine_transforms.py
Created February 28, 2017 21:08
"""
Affine transforms implemented on torch tensors, and
only requiring one interpolation
Included:
- Affine()
- AffineCompose()
- Rotation()
- Translation()
- Shear()
@aras-p
aras-p / preprocessor_fun.h
Last active December 5, 2025 05:33
Things to commit just before leaving your job
// Just before switching jobs:
// Add one of these.
// Preferably into the same commit where you do a large merge.
//
// This started as a tweet with a joke of "C++ pro-tip: #define private public",
// and then it quickly escalated into more and more evil suggestions.
// I've tried to capture interesting suggestions here.
//
// Contributors: @r2d2rigo, @joeldevahl, @msinilo, @_Humus_,
// @YuriyODonnell, @rygorous, @cmuratori, @mike_acton, @grumpygiant,
@piscisaureus
piscisaureus / pr.md
Created August 13, 2012 16:12
Checkout github pull requests locally

Locate the section for your github remote in the .git/config file. It looks like this:

[remote "origin"]
	fetch = +refs/heads/*:refs/remotes/origin/*
	url = git@github.com:joyent/node.git

Now add the line fetch = +refs/pull/*/head:refs/remotes/origin/pr/* to this section. Obviously, change the github url to match your project's URL. It ends up looking like this: