Han-Cheol Cho priancho

AI Coding Assistant Rules for Research Engineering

Core Principles

You are an expert Python and PyTorch coding assistant specializing in LLM/VLM research and development.
Prioritize code quality, reproducibility, and research best practices.
Write production-ready code that balances clarity with performance.

Code Style & Structure

Python Conventions

Purpose

TensorBoard나 Jupyter notebook은 매우 편리한 도구이나, 적절한 세큐리티 (SSL, Authentication, Authorization 등) 방책이 없이는 사용하기 곤란한 경우가 많다.

여기서는 SSH dynamic port forwarding을 통해서 local computer (노트북, 데스크탑 등)으로부터 remote server (작업용 서버, 주로 회사 컴퓨터) 에서 동작중인 웹기반 어플리케이션에 액세스하는 방법을 설명한다. (물론 remote server의 port에 접근하는 것이므로 웹 기반 앱이 아니어도 됨)

	""" Convert jp2k image to jpeg.

	# Install JPEG2k decoder first
	sudo yum install -y libjpeg-devel
	"""
	import os
	import argparse
	from pathlib import Path

	import cv2

	#--------------------------#
	#----- Basic settings -----#
	#--------------------------#
	# Define colors
	C_DEFAULT="\[\033[m\]"
	C_WHITE="\[\033[1m\]"
	C_BLACK="\[\033[30m\]"
	C_RED="\[\033[31m\]"
	C_GREEN="\[\033[32m\]"
	C_YELLOW="\[\033[33m\]"

	#!/bin/bash

	# replace ${} with real names
	PROJECT_NAME=[project name]
	USER=[user name]
	SRC_SERVER=[hostname having the original data]
	SRC_DIR=[path to the original data]
	DST_DIR=[path to archive data]
	LOG=[path to the log file]

	class MultiGPUModelCheckpoint(ModelCheckpoint):
	"""MultiGPU model checkpointing.

	Save a template model, not multi_gpu_model instance.
	Save the model at the beginning (a random init model) and at the
	end of the training.

	> https://github.com/keras-team/keras/issues/8463
	"""
	def __init__(self, args, *kwargs):

	class CustomModelCheckpoint(ModelCheckpoint):
	"""Custom ModelCheckpoint.

	Save the model at the beginning (a random init model) and at the
	end of the training.
	"""
	def __init__(self, args, *kwargs):
	super(CustomModelCheckpoint, self).__init__(args, *kwargs)
	self.current_epoch = 0

	# Create this pyrightconfig.json at the workspace's root directory.
	#
	# Refer: https://github.com/microsoft/pyright/issues/168
	{
	"exclude": ["./data/"]
	}

	# Run the following command to create Jupyter-lab config file under
	# ~/.jupyter/jupyter_notebook_config.py
	# $ jupyter-lab --generate-config
	#

	############################
	### Some custom settings ###
	############################

	# Make it accessible from localhost only

	name: py3_main
	channels:
	- anaconda
	- conda-forge
	- defaults
	dependencies:
	- _nb_ext_conf=0.4.0=py36_1
	- anaconda-client=1.6.3=py36_0
	- beautifulsoup4=4.6.0=py36h49b8c8c_1
	- bleach=1.5.0=py36_0