A simple visualization toolbox (script) for transformer attention visualization

This is a super simple visualization toolbox (script) for transformer attention visualization ✌ 1. How to prepare your attention matrix? Just convert it to numpy array like this 👇 # build an attetion matrixs as torch-output like token_num = 6 case_num = 3 layer_num = 2 head_num = 4 attention_map_mhml = [np.stack([make_attention_map_mh(head_num, token_num)]*case_num, 0) for _ in range(layer_num)] # 4cases’ 3 layers attention, with 3 head per layer( 每个case相同) _ = [print(i.shape) for i in attention_map_mhml] “”” >>>(3, 4, 6, […]

Read more

Simple script for looping a Denial Of Service (DoS) attack over one single mac address in range

Legal Note This project is made only for educational purposes and for helping in Proofs of Concept. The author does not encourage anyone to use it illegally or without permission of your noisy neighbours ;). Description This is a simple script for looping a Denial Of Service attack over one single mac address in range. If Mac address is down or not in range, it will keep trying to attack the target until user press CTRL+C. WARNING: It only works […]

Read more

A Python script that exports users from one Telegram group to another using one or more concurrent user bots

A Python script that exports users from one Telegram group to another using one or more concurrent user bots. Make sure to set all of the required information for each user bot in the config.yaml file, as well as the target group URL to which the bots will add users in the main.py file. For more details on how it works, see this article: https://medium.com/@fasiladam1290/how-i-export-one-telegram-group-users-to-another-using-a-python-script-6da69150182e GitHub View Github    

Read more

Personalized list of today’s articles from ArXiv

Print and/or send to your gmail a personalized list of today’s articles published in ArXiv based on your pre-defined multiple sets of keywords. The script returns: title, abstract and the ArXiv link for each article. Entries are grouped by your pre-defined key. Output example Info This is a very simple code that I wrote for my own personal usage. Some possible future implementations are: Add information to run the code everyday automatically at a certain fixed time using crontab Improve […]

Read more

Toy example of a python script that instantiates and trains a PyTorch neural network on the FashionMNIST dataset with several common and useful featur

This simple_pytorch_example project is a toy example of a python script that instantiates and trains a PyTorch neural network on the FashionMNIST dataset with several common and useful features: Choose between two different neural network architectures Make architectures parametrizable Read input arguments from config file or command line (command line arguments override config file ones) Download FashionMNIST dataset if not already downloaded Monitor training progress on the terminal and/or with TensorBoard logs Accuracy, loss, confusion matrix More details about FashionMNIST […]

Read more

Ghidra scripts to help with 3ds reverse engineering

These are ghidra scripts to help with 3ds reverse engineering. Features: Labels, comments (when inlined), and bookmarks svc use Labels service handles, given ctr::srv::GetServiceHandleDirect Labels IPC functions and uses handles to better identify functions Adds ThreadLocalStorage and types thread local storage Renames thread local storage to ‘tls’ These have been built over time for my personal use as needs came up, so results may vary and improvements can be made. If you run into a situation where these don’t work […]

Read more

A script utilises payload dumper and image extractor tools to extract the apps from the system.img of an android OTA file

This script utilises payload dumper and image extractor tools to extract the apps from the system.img of an android OTA file. It works on Windows OS. Requirements: Windows 10 Python3 installed and added to environment variable pip3 installed and added to environment variable Python Modules -> protobuf==3.6.0, six==1.11.0, bsdiff4>=1.1.5 If you face issues, install Windows 10 SDK & C++ build tools. Download the build tools here. Launch the Visual Studio Installer and select the Desktop development with C++ option. Install […]

Read more

This repo is about steps to create a effective custom wordlist in a few clicks

Custom Wordlist This repo is about steps to take in order to create a effective custom wordlist in a few clicks. this comes handing in pentesting engagement where you need to do brute force attacks.So the main probleme is that you have 0 chance to get that valide account when you use a random wordlist that contain passwords that don’t match your client password policy,Well to build this effective custom wordlist that match what ever password policy your client uses: […]

Read more

A python script that uses webtorrent to stream nyaa videos directly to mpv

A rather shitty script that uses webtorrent to stream nyaa videos directly to mpv. nyon_demo.mp4 python3 nyaascraperpy For now the script only works on Linux with dmenu and webtorrent (webtorrent-cli) installed. It’s really not hard to make it compatible with windows but i’ll just let someone do it for me. python dependencies BeautifulSouprequestsdmenu external dependencies dmenu (https://archlinux.org/packages/community/x86_64/dmenu/)webtorrent (https://aur.archlinux.org/packages/webtorrent-cli) You can edit the 16th line of the script to have it scrap default/danger (white/red) entries on nyaa GitHub View Github   […]

Read more

Run shell script in Jupyter with live output

ipylivebash is a library to run shell script in Jupyter with live output. Example %%livebash –save log.txt –save-timestamp find . Run find . and show the output in the Jupyter notebook and also save to log-${current_timestamp}.txt %%livebash –ask-confirm –notify set -e deploy_script Before running the deploy_script show a panel to ask for confirmation. Once it is finished, show a notification. Features Run shell script in Jupyter with live output The output in the notebook cell is not saved in the […]

Read more
1 6 7 8 9 10 18