Homoglyph Sproofing / Typo

 "rn" -> "m"

"cl" -> "d"

"ci" -> "a"

"vv" -> "w"

"Il" -> "ll"

"1" -> "l"

"I" -> "l"

"nn" -> "m"

"M0N" -> "MON"

[LLM] llmster

 Switch to a working folder:

# cd /usr/local/sbin

Show what kernel my VPC owns:

# uname -a

Linux demonalex-VirtualBox 7.0.0-14-generic #14-Ubuntu SMP PREEMPT_DYNAMIC Mon Apr 13 11:09:53 UTC 2026 x86_64 GNU/Linux

Install llmster:

# curl -fsSL https://lmstudio.ai/install.sh | bash

Add the following into /etc/profile:

export PATH="/root/.lmstudio/bin:$PATH"

Open a new terminal and execute the following to see the model list:

# lms get

Press Ctrl+C to end the showing.

Download the model:

# lms get google/gemma-4-e2b

Check if the model has already been downloaded:

# lms ls

Have an interactive chatbot:

# lms chat

Type "/exit" can exit the chatbot conversation.

[AI][ML][LLM]Run Llama 3B through Docker Model

 Install necessary plugin:

# curl -fsSL https://download.docker.com/linux/ubuntu/gpg | sudo gpg --dearmor -o /usr/share/keyrings/docker-archive-keyring.gpg

# echo "deb [arch=amd64 signed-by=/usr/share/keyrings/docker-archive-keyring.gpg] https://download.docker.com/linux/ubuntu $(lsb_release -cs) stable" | sudo tee /etc/apt/sources.list.d/docker.list > /dev/null

# apt-get update

# apt-get install docker-model-plugin

Pull:

# docker model pull ai/llama3.2

Run in Interactive mode:

# docker model run ai/llama3.2

Exit Interactive mode:

> /bye

Run in Single-prompt mode:

# docker model run ai/llama3.2 "Explain how Docker containers work in one sentence."

-----Failed to carry out from here:

Enable OpenAI mode:

# docker desktop enable model-runner --tcp 12434

Test OpenAI:

# curl http://localhost:12434/engines/v1/models

Remotely use through OpenAI:

# curl http://localhost:12434/engines/v1/chat/completions \

  -H "Content-Type: application/json" \

  -d '{

    "model": "ai/llama3.2",

    "messages": [{"role": "user", "content": "Hello!"}]

  }'

[Windows][Sniffer] Microsoft Pktmon command-based sniffer

#Add filter:

pktmon filter add MyFilter -i 192.168.215.11

#OR

pktmon filter add MyFilter -i 192.168.1.100 -p 443

pktmon filter add MyFilter -i 192.168.1.100 -p 443 -t tcp

#Start:

pktmon start --capture --log-mode real-time

#Stop:

pktmon stop

#Remove filter:

pktmon filter remove

[Linux][Shell] tmux substituting the screen command

 Installation:

#apt-get update

#apt-get install tmux

How to use:

#tmux

Shortcut:

Ctrl+b c: creation

Ctrl+b x: Exit

Ctrl+b %: Split screen

Ctrl+b ;: Switch screen

[Weka][Machine Learning][Kmeans]Weka labelling result of Kmeans

### The Step-by-Step Process

1. **Run your Clusterer:**

* Go to the **Cluster** tab.

* Select **SimpleKMeans** and configure your  (number of clusters).

* Run the algorithm. Ensure it completes successfully.

2. **Add the Labels via Filter:**

* Switch back to the **Preprocess** tab.

* Click **Choose** under the Filters section.

* Navigate to: `filters` -> `unsupervised` -> `attribute` -> **AddCluster**.

3. **Configure the Filter:**

* Click on the **AddCluster** text to open its configuration box.

* In the `clusterer` field, click "Choose" and select **SimpleKMeans** again (ensure the settings match what you used in the Cluster tab).

* Click **OK**.

4. **Apply and Save:**

* Click **Apply**. You will see a new attribute appear at the end of your attribute list, typically named "cluster."

* Click **Save...** to export your new labeled dataset as an `.arff` or `.csv` file.

---

### Pro-Tips for Labeling

* **The "Ignored" Attributes:** If you want to cluster based on certain features but keep an ID or Name column in the final file, make sure to use the `ignoredAttributeIndices` property within the SimpleKMeans settings inside the **AddCluster** filter.

* **Result Verification:** Before saving, you can click the **Edit...** button in the Preprocess tab to view the data table and confirm that the cluster assignments look correct (e.g., Cluster 0, Cluster 1, etc.).

---

[perl][ML][Machine Learning]K-means in Perl

#!/usr/bin/perl -w

use strict;

use warnings;

use PDL::LiteF;

use PDL::Stats;

# 1. Define 4 separate arrays (Features)

my $f1 = pdl(1, 2, 1, 10, 11, 12); # Feature 1

my $f2 = pdl(1, 1, 2, 10, 12, 11); # Feature 2

my $f3 = pdl(0, 1, 0, 15, 14, 15); # Feature 3

my $f4 = pdl(2, 2, 1, 20, 21, 22); # Feature 4

# 2. Combine them into a 4-column matrix

# cat() joins them, transpose() makes each row a "person/object"

my $data = cat($f1, $f2, $f3, $f4)->transpose;

# 3. Run K-means for 2 groups

my $k = 2;

my %result = $data->kmeans({k => $k, NCLUS => 2});

# 4. Display which group each of the 6 items belongs to

print "Data Matrix (Rows = Items, Cols = Features):\n$data\n";

print "Cluster Assignments:\n", $result{cluster}, "\n";

print "Final Centroids (The average of each group):\n", $result{centroid}, "\n";