one by one

Toshihiro Kamiya tos-kamiya

one by one

tos-kamiya / timeg

Created September 15, 2024 18:26

A script to monitor CPU/GPU RAM usage of a command on NVIDIA systems, with process search by keyword and JSON support.

	#!/usr/bin/env python3

	import argparse
	import json
	import os
	import subprocess
	import sys
	import time
	from typing import List, Tuple, Dict

tos-kamiya / ollama-unload

Created September 14, 2024 16:04

A script to immediately free up VRAM used by ollama services.

	#!/bin/bash

	# The default ollama server address
	OLLAMA_URL="http://localhost:11434"

	# Help function
	show_help() {
	echo "Usage: $0 [OPTIONS] [URL]"
	echo "Unload all loaded models in ollama."
	echo ""

tos-kamiya / trans

Last active September 4, 2024 04:21

A simple text translator script

	#!/bin/bash

	# Function to display help message
	function show_help() {
	echo "Usage: $0 <language_code> [-p] <text_file_or_text>"
	echo
	echo "Arguments:"
	echo " <language_code> The target language code (e.g., 'en' for English, 'ja' for Japanese)."
	echo " <text_file_or_text> The path to the text file to be translated, or the text itself if -p is used. Use '-' to read from stdin."
	echo

tos-kamiya / git-status-summary.py

Created June 2, 2024 22:13

VSCode like Git status viewer

	#!/usr/bin/env python3

	from typing import List, Iterable, Iterator, Tuple
	import subprocess
	import sys


	def colorize(text: str, color: str) -> str:
	"""
	Colorize the text for terminal output.

tos-kamiya / flexcomm.py

Last active March 25, 2024 19:21

flexcomm: flexible comm utility, handle three or more files with use-specified predicates.

	import argparse
	import ast
	import sys
	from typing import List, TextIO, Optional


	def get_variables(expr: str) -> List[str]:
	"""
	Given an expression, return a sorted list of variable names used in the expression.

tos-kamiya / csv_to_latex.py

Created December 24, 2023 06:59

Convert CSV file to LaTeX table

	#!/usr/bin/env python3

	import argparse
	import sys
	import os
	import pandas as pd


	def csv_to_latex(input_file, output_file = None):
	# Read csv file

tos-kamiya / ai-howdoi

Last active September 2, 2024 15:22

howdoi command with leverage of Phind-CodeLlama

	#!/usr/bin/env python3

	# ref: https://stackoverflow.com/questions/2804543/read-subprocess-stdout-line-by-line#2813530

	import io
	import os
	import subprocess
	import sys

	from pygments import highlight

tos-kamiya / lang_ext_counter.py

Created November 19, 2023 16:14

A tool for aggregating file extensions and languages in a directory, compatible with the tokei project (https://github.com/XAMPPRocky/tokei).

	# The languages.json is obtained from the tokei project
	# Source: https://github.com/XAMPPRocky/tokei/blob/c8e4d0703252c87b1df45382b365c6bb00769dbe/languages.json

	from typing import Dict, Counter as CounterType

	from collections import Counter
	import json
	import os
	import sys

tos-kamiya / ipynb_to_md

Last active November 15, 2023 18:37

A tool to convert Jupyter Notebook (.ipynb) to Markdown (.md)

	#!/usr/bin/env python3

	# https://chat.openai.com/share/448d1592-749e-49f4-8e44-948b0207d075

	def to_markdown(jypyter_notebook_json):
	cells = jypyter_notebook_json["cells"]

	for cell in cells:
	ct = cell["cell_type"]
	if ct == "markdown":

tos-kamiya / trans_nllb.py

Created September 27, 2023 19:59

A command-line translator using Facebook's NLLB LLM (proof of concept)

	# ref https://zenn.dev/syoyo/articles/9a159ee747835a

	import sys

	from transformers import AutoModelForSeq2SeqLM, AutoTokenizer

	max_length = 512

	# ref https://huggingface.co/facebook/nllb-200-distilled-1.3B
	# The model was trained with input lengths not exceeding 512 tokens, therefore translating longer sequences might result in quality degradation.