link89

2.2.8

CONDA_OVERRIDE_CUDA="11.8" conda create -p deepmd/2.2.8/gpu deepmd-kit=2.2.8=cuda118py310*  lammps horovod -c conda-forge

To work around a known issue of conda-forge's tensorflow package, one has to run pip install nvidia-cuda-nvcc-cu11 in the conda environment.

Before 2.2.8

Build DeepMD-kit + LAMMPS in a conda envrionment

Steps

Create conda environment

conda create -p 2.2.5/gpu python==3.10.13  tensorflow=2.10.*=gpu*

Note:

Install tensorflow with conda instead of pip will have less trouble (for example, conda will install the right cudnn for you)

POC: use typescript as container of structured data for LLM

Prompt

The following is Typescript code that describe the data structure of reactions. The reactions to be exported are not implement yet.

/**
 * This is a typescript prompt file to export reactions data.
 */

Improve Download Speed

Use BBR congestion control

Ubuntu

Update /etc/sysctl.conf

net.core.default_qdisc=fq
net.ipv4.tcp_congestion_control=bbr

Installation

https://medium.com/@anarmammadli/how-to-install-openpbs-20-0-1-on-ubuntu-b1cc62cfb0ec

Issue

pbs_server doesn't run
postgresql doesn't run
Similar issues
- https://community.openpbs.org/t/can-not-start-pbs-server-all-other-jobs-are-running/3048

Fix

Note

Use autoconfig when <= 2.6, and cmake > 2.7
Patch fortran Makefile
Use CP2K's toolchain as reference: https://github.com/cp2k/cp2k/blob/master/tools/toolchain/scripts/stage3/install_libint.sh

Build

# download
wget https://www.cp2k.org/static/downloads/libint-v2.6.0-cp2k-lmax-5.tgz --no-check-certificate
tar -xvf libint-v2.6.0-cp2k-lmax-5.tgz

Note

use mkl's fftw build

Build

wget https://www.cp2k.org/static/downloads/plumed-src-2.8.2.tgz --no-check-certificate
tar -xvf plumed-src-2.8.2.tgz
cd plumed-2.8.2

./configure FC=mpiifort CC=mpiicc CXX=mpiicpc CXXFLAGS="-O3 -mtune=icelake-server" --enable-mpi \

	"""
	Convert DIYP TV List to m3u format

	Usage:
	python dipy_to_m3u.py diyp.txt out.m3u

	DIPY format example:
	cctv\|channel,#genre#
	CCTV1,http://example.org
	"""

	// ==UserScript==
	// @name Reddit Crawler
	// @namespace http://tampermonkey.net/
	// @version 0.1
	// @description Crawl posts and comments on a specific reddit sub and export as jsonl
	// @author Bing
	// @match https://www.reddit.com/*
	// @grant GM_registerMenuCommand
	// ==/UserScript==

	import numpy as np

	def sample_data_frame(path: str, frame_lines: int, nsample: int, out_path: str,
	total_lines: int = -1, encoding='utf-8',
	write_buffer_size=10 * 1024 * 1024,
	):
	"""
	Sample data frame from file randomly

	A data file is composed of multiple frames, each frame has fixed number of lines.