Hieu C B cobaohieu

Prompt Mountain Image:

A panoramic illustration of a majestic, snow-capped mountain range with sharp, detailed peaks, illustrated in a clean anime landscape style combined with refined comic book line art. The scene should feel like a high-quality anime background asset for a cinematic nature environment. Use precise black and dark navy outlines, crisp hand-drawn contour lines, and clean angular details across the mountain surfaces.

The mountains are viewed from a wide, low-angle panoramic perspective, creating a grand and dramatic sense of scale. The snow-covered peaks feature distinct flat cel-shaded planes in cool blue-grey, ice-grey, pale cyan-grey, and clean off-white highlights. Avoid gradients, blending, and airbrushing. Each snow plane should be clearly separated by sharp angular shadow shapes.

Below the snow line, the rugged rocky terrain is rendered with clear blocks of flat colors, including muted tans, earthy browns, bluish-greys, slate blue, and charcoal tones. The rock faces should have g

LLM Wiki v2

A pattern for building personal knowledge bases using LLMs. Extended with lessons from building agentmemory 20K+ Stars ⭐️, a persistent memory engine for AI coding agents.

This builds on Andrej Karpathy's original LLM Wiki idea file. Everything in the original still applies. This document adds what we learned running the pattern in production: what breaks at scale, what's missing, and what separates a wiki that stays useful from one that rots.

What the original gets right

The core insight is correct: stop re-deriving, start compiling. RAG retrieves and forgets. A wiki accumulates and compounds. The three-layer architecture (raw sources, wiki, schema) works. The operations (ingest, query, lint) cover the basics. If you haven't read the original, start there.

LLM Wiki

A pattern for building personal knowledge bases using LLMs.

This is an idea file, it is designed to be copy pasted to your own LLM Agent (e.g. OpenAI Codex, Claude Code, OpenCode / Pi, or etc.). Its goal is to communicate the high level idea, but your agent will build out the specifics in collaboration with you.

The core idea

Most people's experience with LLMs and documents looks like RAG: you upload a collection of files, the LLM retrieves relevant chunks at query time, and generates an answer. This works, but the LLM is rediscovering knowledge from scratch on every question. There's no accumulation. Ask a subtle question that requires synthesizing five documents, and the LLM has to find and piece together the relevant fragments every time. Nothing is built up. NotebookLM, ChatGPT file uploads, and most RAG systems work this way.

Concurrency and Synchronization using Modern C++

Thread Management

Each running program has at least one thread which is occupied by its entry point. There can be many threads for a single program and the memory is shared among all threads. The objects possessed by the main thread (and shared by other threads as a reference to it) would be destroyed as soon as they go out of scope. This might lead to corrupt memory access by the rest of the threads.

Hence, sometimes, it's very important to make sure that all the threads are joined using std::thread::join() before the main thread terminates. Other times, you basically want a daemon thread running in background even after the termination of main thread and that can be achieved by calling std::thread::detach(), but it may or may not be possible to join a detachable thread.

Resource Acquisition Is Initialization (RAII) is a famous programming idiom in C++ (a misnomer actually) also (better) known and understood as **Scope-Bound Resource Managemen

sudo apt install curl gnupg
curl -fsSL https://bazel.build/bazel-release.pub.gpg | gpg --dearmor > bazel.gpg
sudo mv bazel.gpg /etc/apt/trusted.gpg.d/
echo "deb [arch=amd64] https://storage.googleapis.com/bazel-apt stable jdk1.8" | sudo tee /etc/apt/sources.list.d/bazel.list

sudo apt update && sudo apt install bazel
sudo apt update && sudo apt full-upgrade
bazel --version

Problem

The MPOW Bluetooth 5 dongle (Model: BH456A) does not work out of the box on Ubuntu 20.04 (kernel 5.4.0-42).

Solution

Patch the bluetooth kernel module and copy the firmware binaries to /lib/firmware.

	# A basic script for uninstalling app packages in Windows 10/11, including those pre-installed with Windows
	#
	# Note: If you get an error about the script not being allowed to run, the below command will change the execution polciy temporarily for one session only:
	# Set-ExecutionPolicy -ExecutionPolicy RemoteSigned -Scope Process
	#
	# To execute the script, open a Powershell window to the directory with the script and run the following command using your scripts file name (and don't forget the .\ )
	# .\WhateverScriptName.ps1
	# -------------------------------------------------------------------------------------------
	# Script by ThioJoe - https://github.com/ThioJoe

	{
	"final_space": true,
	"console_title": true,
	"console_title_style": "folder",
	"blocks": [
	{
	"type": "prompt",
	"alignment": "left",
	"horizontal_offset": 0,
	"vertical_offset": 0,

	FROM nvcr.io/nvidia/l4t-pytorch:r32.4.4-pth1.6-py3

	ENV DEBIAN_FRONTEND=noninteractive

	RUN apt update
	RUN apt install -y cmake libgtk2.0-dev wget

	RUN wget https://nvidia.box.com/shared/static/9eptse6jyly1ggt9axbja2yrmj6pbarc.whl -O torch-1.6.0-cp36-cp36m-linux_aarch64.whl
	RUN python3 -m pip install torch-1.6.0-cp36-cp36m-linux_aarch64.whl
	RUN python3 -m pip install scikit-build

	import numpy as np
	import cv2

	def correction(
	img,
	shadow_amount_percent, shadow_tone_percent, shadow_radius,
	highlight_amount_percent, highlight_tone_percent, highlight_radius,
	color_percent
	):
	"""