ML Research Hub
32.9K subscribers
4.45K photos
273 videos
23 files
4.81K links
Advancing research in Machine Learning – practical insights, tools, and techniques for researchers.

Admin: @HusseinSheikho || @Hussein_Sheikho
Download Telegram
πŸ§‘β€πŸŽ“ Machine learning approaches for personalized medicine

πŸ“•PhD Thesis from UniversitΓ  del Piemonte Orientale, Italy

πŸ—“Publish year: 2024

πŸ“Ž Study thesis

βœ…οΈ https://t.me/DataScienceT βœ…οΈ
Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
πŸ‘2
This media is not supported in your browser
VIEW IN TELEGRAM
πŸ˜€ ViViD: Diffusion Virtual Try-ON πŸ§‘β€πŸŽ“

πŸ‘‰ ViViD is a novel framework employing powerful diffusion models to tackle the task of video virtual try-on. Code announced, not released yet 😒

πŸ‘‰ Review: https://t.ly/h_SyP

πŸ‘‰ Paper: arxiv.org/pdf/2405.11794

πŸ‘‰ Repo: https://lnkd.in/dT4_bzPw

πŸ‘‰ Project: https://lnkd.in/dCK5ug4v

βœ…οΈ https://t.me/DataScienceT βœ…οΈ
Please open Telegram to view this post
VIEW IN TELEGRAM
❀‍πŸ”₯2πŸ‘2😍2πŸ†1
TON is Hosting Three-Day Offline Marathons for Developers in 13 Cities Worldwide

TON (The Open Network), a blockchain integrated into Telegram, is currently holding the largest hackathon in its history, β€œThe Open League Hackathon,” with a prize pool of $2,000,000!

To support the hackathon, TON Society is organizing offline events around the world, giving Web3 enthusiasts the opportunity to spend three days immersed in development and networking.

πŸ“ The marathons will take place in Prague, Berlin, Kyiv, Warsaw, Tbilisi, Belgrade, Seoul, Taipei (Taiwan), Gurugram (India), Hong Kong, Minsk, Moscow, and Saint Petersburg. The first events start on May 24th, so hurry up and register!

πŸ”₯ Why you should participate:

β€” 3 days of networking, lectures, competitions, and working on your own projects with continuous support from TON Foundation representatives and teams from the TON ecosystem. You can also join online.

β€” $5,000 prize for the top three projects at each offline event + plenty of merchandise and other bonuses.

Don’t miss the chance to present your mini-application to 900 million active Telegram users with TON.

Check the marathon schedule and details here - sign up and don't miss out on this great opportunity!

For guaranteed application review and other inquiries, contact the community manager @kate_shuffle
πŸ‘6❀‍πŸ”₯1πŸ†1
This media is not supported in your browser
VIEW IN TELEGRAM
🌹 Face Adapter for Pre-Trained Diffusion Models with Fine-Grained ID and Attribute Control

🀯 Face Adapter is something!

πŸš€ New powerful face transfer adapter that works with pre-trained diffusion models.

πŸ”„ Provides precise control over facial expressions and features.

Works with video and photos.

🟒 Github: https://github.com/FaceAdapter/Face-Adapter

🟒 Paper: https://arxiv.org/abs/2405.12970

🟒 HF: https://huggingface.co/FaceAdapter/FaceAdapter

🟒 Project: https://faceadapter.github.io/face-adapter.github.io/

βœ…οΈ https://t.me/DataScienceT βœ…οΈ
Please open Telegram to view this post
VIEW IN TELEGRAM
πŸ‘3❀‍πŸ”₯1❀1πŸ†1
This media is not supported in your browser
VIEW IN TELEGRAM
🧰 A comprehensive toolbox of data scientists
βœ… Access to 500+ TB of data science content

πŸ–₯ The Data Scientist Toolkit is the result of my five-year effort in the field of data science, which is an extremely comprehensive and extensive resource for those who want to become a professional data scientist. So don't miss this source

βš™οΈ This toolbox includes the following sections:

πŸ“š Access to a wide bank of scientific files, training courses and professional resources in the field of data science.

πŸ’― More than a decade of applied data science theses in finance, medicine, logistics and security.

πŸ› The latest data science courses from leading universities in the world such as Stanford, MIT and Berkeley.

πŸš€ And over 300+ terabytes of data science courses for experienced data scientists.

β”Œ
🐱 GitHub Repository
β””
⏩ The Data Scientist's Toolbox

🌐 http://t.me/codeprogrammer βœ…οΈ
Please open Telegram to view this post
VIEW IN TELEGRAM
πŸ‘11
⚑️ Adobe has released DMD2!

A new and improved distillation method that can turn diffusion models such as SDXL into powerful one-step image generators.

It's been a while since we've seen any interesting updates to the SD ecosystem, so this is cool πŸ”₯

β–ͺ Project page: https://tianweiy.github.io/dmd2/

β–ͺ Code: https://github.com/tianweiy/DMD2

β–ͺ Demo: https://4e4a5c6a8b08f76802.gradio.live

🌐 http://t.me/codeprogrammer βœ…οΈ
Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
πŸ‘7❀1
πŸŽ“ Empowering Graph Neural Networks for Real-World Tasks

πŸ“˜ Doctoral thesis
πŸ—“ Publish year: 2024

πŸ§‘β€πŸ’» Author: Zhichun Guo
🏒 University: University of Notre Dame, Indiana

πŸ“Ž Study Thesis 

🌐 http://t.me/codeprogrammer βœ…οΈ
Please open Telegram to view this post
VIEW IN TELEGRAM
πŸ‘4
Large Brain Model for Learning Generic Representations with Tremendous EEG Data in BCI

πŸ–₯ Github: https://github.com/935963004/labram

πŸ“•Paper: https://arxiv.org/abs/2405.18765v1

#machineLearning

βœ…https://t.me/DataScienceT
Please open Telegram to view this post
VIEW IN TELEGRAM
πŸ‘3
https://t.me/Hamster_kombat_bot/start?startapp=kentId418788114


πŸ’΅ 2 thousand coins as a gift for the first time
β˜„οΈ 25 thousand coins if you have Telegram Premium

Hamster will be listed on Ton Network soon

The experience is quite similar to that of the NOTCOIN bot πŸ’ 
Please open Telegram to view this post
VIEW IN TELEGRAM
πŸ‘5
🌟 InstaDrag: complex image editing (rotating in space, changing shape, etc.) using simple gestures

InstaDrag is an approach that allows you to edit an image using a simple drag gesture in ~1 second.
At the same time, InstaDrag also copes with complex deformations of part of the image that are not represented in the training data (such deformations are, for example, elongation of hair, curvature of the rainbow, etc.).
There are plans to integrate InstaDrag with SDXL in the future

πŸ–₯ GitHub
🟑 InstaDrag page
🟑 Arxiv

βœ…https://t.me/DataScienceT
Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
πŸ‘3
This media is not supported in your browser
VIEW IN TELEGRAM
⚑️ Flash version of Scribble SDXL

Utilizes SDXL Flash and Scribble SDXL, making it even faster to get quality images from simple sketches and doodles

πŸ€— Launch on Hugging Face

https://t.me/DataScienceT βœ…
Please open Telegram to view this post
VIEW IN TELEGRAM
❀‍πŸ”₯2πŸ‘1
Forwarded from πŸ³
This media is not supported in your browser
VIEW IN TELEGRAM
🐳Meet @whale – your new go-to gaming platform, now available on Telegram!

There are many games on our licensed platform where you can make a big jackpot!

🏎Over 1000 games with impressive winning odds.
πŸ€‘Accepts BTC, USDT, TON, and CELO.
🏎Up to 20% cashback.
πŸ€‘Ongoing promotions and contests with valuable rewards.
🏎Sportsbook with seamless betting, and the best odds that you could only imagine!

πŸ’ŽFrom May 16, you can play and withdraw notcoin!

Forget about registration – play and cash out your prizes right from Telegram. ⬆️

Join now and win big with @whale πŸ₯°
Please open Telegram to view this post
VIEW IN TELEGRAM
πŸ‘5❀‍πŸ”₯2πŸ†2
This media is not supported in your browser
VIEW IN TELEGRAM
🌟 V-Express is a method of animating a static photograph of a face, taking into account pose, sound and the image itself

The V-Express method was recently introduced by Cong Wang, Kuan Tian, ​​Jun Zhang and others as one of the most efficient methods for generating such animations, the code and model are published

πŸ–₯ GitHub

🟑 V-Express page

πŸ€— Model on Hugging Face

https://t.me/DataScienceT βœ…
Please open Telegram to view this post
VIEW IN TELEGRAM
πŸ‘3
πŸ”₯ unsloth - fine tuning Llama 3, Mistral, Phi and Gemma LLM, reducing memory consumption by 80%, speeding up 2-5 times

conda create --name unsloth_env python=3.10
conda activate unsloth_env

conda install pytorch-cuda=<12.1/11.8> pytorch cudatoolkit xformers -c pytorch -c nvidia -c xformers

pip install "unsloth[colab-new] @ git+https://github.com/unslothai/unsloth.git"

pip install --no-deps trl peft accelerate bitsandbytes


Some unsloth benchmarks are given here

πŸ–₯ GitHub
🟑 Wiki

https://t.me/DataScienceT βœ…
Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
πŸ‘7