Join the conversation

Join the community of Machine Learners and AI enthusiasts.

Sign Up

All HF Hub posts

openfreeย 
posted an update about 19 hours ago
view post
Post
1899
๐Ÿค– AI-Generated 6-Nation Military Simulator in a Single HTML File

๐Ÿš€ Project Highlight
A full-scale military strategy simulator that runs in a single HTML file! This AI-generated wargame implements real military equipment and tactical doctrines from 6 nations (๐Ÿ‡ฐ๐Ÿ‡ท,๐Ÿ‡ฐ๐Ÿ‡ต,๐Ÿ‡บ๐Ÿ‡ธ,๐Ÿ‡ท๐Ÿ‡บ,๐Ÿ‡บ๐Ÿ‡ฆ,๐Ÿ‡จ๐Ÿ‡ณ) using pure JavaScript only, without any external libraries.

openfree/WAR-Game-Simul

๐Ÿ’ก Amazing Achievement of AI Auto-Generation

๐Ÿ“ Single File Magic
One-Click Launch: Just open the HTML file and play instantly
Zero Dependencies: No npm, webpack, or external libraries
Pure Vanilla JS: Implemented with Canvas API only, no frameworks
All-in-One: Rendering, physics engine, AI, and UI in a single file

๐ŸŽฎ Advanced Features AI Implemented
โœ… Perlin Noise terrain generation algorithm
โœ… Marching Squares contour rendering
โœ… Lanchester combat equations
โœ… A* pathfinding algorithm
โœ… Real-time Line of Sight (LOS) calculations
โœ… 40-second battlefield sound loop
๐ŸŽ–๏ธ Implemented Military Systems
๐Ÿ”ฅ Real Weapon Systems from 6 Nations (100+ Units)

Real tanks: K2 Black Panther, M1A2 Abrams, T-90M
Attack helicopters: Apache, Ka-52, WZ-10
MLRS systems: HIMARS, Chunmoo, BM-21 Grad
Actual performance specs for each nation's unique equipment

โš”๏ธ Realistic Combat Mechanics

Urban Warfare: 70% damage reduction for defenders, 2.5x road mobility
Mountain Combat: 20% range bonus from high ground, ridge defense bonus
Encirclement: Instant victory with 4+ units from 3+ directions
Real-time Combat: 10 engagement calculations per second

๐Ÿ“Š Technical Achievement of AI-Generated Code
javascript// Everything in a single file!

๐Ÿ† New Standard for AI Wargame Generation
This project proves that AI can auto-generate complex military simulations in a single HTML file. It implements all core mechanics of AAA strategy games using pure JavaScript without any external dependencies.

#AIGenerated #SingleFile #Wargame #NoFramework #VanillaJS #ZeroDependency #HTML5Game #MilitarySimulator
LPX55ย 
posted an update 1 day ago
view post
Post
2926
Qwen's latest Image Edit model has been implemented with lightx2v's LoRA for 8-step lightning fast inferencing. Still a WIP, so YMMV.

LPX55/Qwen-Image-Edit_Fast-Presets

  • 2 replies
ยท
RakshitAralimattiย 
posted an update 2 days ago
view post
Post
6285
When you ask ChatGPT, Claude, or Gemini a really tough question,
you might notice that little "thinking..." moment before it answers.

But what does it actually mean when an LLM is โ€œthinkingโ€?

Imagine a chess player pausing before their next move not because they donโ€™t know how to play, but because theyโ€™re running through possibilities, weighing options, and choosing the best one.
LLMs do something similarโ€ฆ except theyโ€™re not really thinking like us.

Hereโ€™s the surprising part :-
You might think these reasoning skills come from futuristic architectures or alien neural networks.
In reality, most reasoning LLMs still use the same transformer decoder-only architecture as other models
The real magic?
Itโ€™s in how theyโ€™re trained and what data they learn from.

Can AI actually think, or is it just insanely good at faking it?
I broke it down in a simple, 4-minute Medium read.
Bet youโ€™ll walk away with at least one โ€œaha!โ€ moment. ๐Ÿš€

Read here - https://lnkd.in/edZ8Ceyg
ยท
ginipickย 
posted an update 3 days ago
view post
Post
3036
โœจ HairPick | Preview Your Perfect Hair Transformation in 360ยฐ โœจ

๐ŸŽŠ Free Trial for Hugging Face Launch! Hurry! โฐ
Hello! Introducing an innovative AI service that helps you choose the perfect hairstyle without any regrets before visiting the salon!

๐ŸŽฏ Try It Now
ginigen/Hair-Pick

๐Ÿ”„ What Makes HairPick Special? 360ยฐ Complete Preview!
Other hair simulators only show the front view? ๐Ÿ˜‘

HairPick is different!
โœ… Front + 4 random angles = Total 5 multi-angle images generated
โœ… Perfect check from side profile ๐Ÿ‘ค diagonal ๐Ÿ“ back view ๐Ÿ‘ฅ!
โœ… 100+ trendy hairstyle library ๐Ÿ’‡โ€โ™€๏ธ

๐Ÿ’ก Highly Recommended For:
๐ŸŽฏ "I really don't want to fail this time!"
โ†’ Check side volume and back lines thoroughly
๐ŸŽฏ "It's hard to explain exactly to my stylist"
โ†’ Perfect communication with 360ยฐ result images!
๐ŸŽฏ "I have a profile photo/photoshoot coming up"
โ†’ Preview your best look from every angle
๐Ÿš€ Super Simple Usage (Just 1 Minute!)

1๏ธโƒฃ One Selfie ๐Ÿ“ธ
Take a front-facing photo in bright light (show your forehead and face outline clearly!)
2๏ธโƒฃ Choose Your Style ๐Ÿ’ซ
Select from 100+ options: short cuts, medium, long hair, layered, bangs, and more
3๏ธโƒฃ Check 360ยฐ Results ๐Ÿ”„
Compare front + side + back + diagonal angles all at once!
4๏ธโƒฃ Go to the Salon! โœ‚๏ธ
Save your favorite result โ†’ Show it to your stylist

๐Ÿ“ธ Pro Tips for Perfect Results!
๐Ÿ’ก Lighting: Natural light or bright, even indoor lighting
๐Ÿ’ก Angle: Camera at eye level, facing straight ahead
๐Ÿ’ก Preparation: No hatsโŒ No sunglassesโŒ Hair tucked behind earsโญ•

๐ŸŽ Now's Your Chance!
"The era of deciding based on front view only is over!"
HairPick isn't just simple hair synthesis, it's a next-level AI hair simulator that predicts your actual appearance in 360ยฐ.

๐Ÿ”ฅ Limited free access for Hugging Face launch!
๐Ÿ”ฅ 100+ latest trend styles!
๐Ÿ”ฅ ZERO failures with 360ยฐ perfect prediction!

โœ‚๏ธ Click before you cut! Take on the perfect hair transformation with HairPick! ๐ŸŒŸ

#HairPick #AIHairSimulator #360HairPreview
  • 1 reply
ยท
prithivMLmodsย 
posted an update 1 day ago
AdinaYย 
posted an update about 11 hours ago
etemizย 
posted an update 3 days ago
view post
Post
5559
benchmarked 9 models in 3 days. they were mostly below average in AHA score. p(doom) probably increased :(
prithivMLmodsย 
posted an update 3 days ago
view post
Post
4507
Excited to introduce the Tiny VLMs Lab App for experiencing 15+ multimodal VLMs, ranging from a 250M parameter model to a 4B parameter model, for tasks like OCR, reasoning, small models for single-shot answering, and captioning (abliterated), across a broad range of visual categories including images with complex, sensitive, or nuanced content, while handling varying aspect ratios and resolutions.๐Ÿงช

๐Ÿค— Space/App: prithivMLmods/Tiny-VLMs-Lab

โœฆ๏ธŽ Also introducing prithivMLmods/Qwen2.5-VL-3B-Abliterated-Caption-it, tailored for Abliterated Captioning / Uncensored Image Captioning. This release comes as a lighter alternative to the existing Qwen2.5-VL-7B-Abliterated-Caption-it prithivMLmods/Qwen2.5-VL-7B-Abliterated-Caption-it model, making it usable on mid-range GPUs and even experimental on T4 GPUs.

โœฆ๏ธŽ Collection: prithivMLmods/vl-abliterated-caption-68a0443b63182e97a15c47a3
โœฆ๏ธŽ GitHub: https://github.com/PRITHIVSAKTHIUR/Tiny-VLMs-Lab
.
.
.
To know more about it, visit the app page or the respective model page!!
Jawardย 
posted an update about 19 hours ago
view post
Post
866
fascinating read!
staying bullish on search with rl might just help us get rid of hallucination entirely. I really like their approach:
1) <think>on prompt/context && what u know </think>
2) self <search>when u donโ€™t know</search> (iteratively) with no external tool
3) <information>cite sources to support claim(s)</information>
4) <answer>final answer</answer>
their rl training was done cost efficiently too, see code: https://github.com/TsinghuaC3I/SSRL
seawolf2357ย 
posted an update 1 day ago
view post
Post
852
๐ŸŽจ Open Nano-Banana: Revolution in Ultra-Fast AI Image Editing!

๐Ÿš€ Introduction
**Open Nano-Banana** is an innovative image editing tool based on the Qwen-Image-Edit model. Experience amazing quality image editing in just 8 steps!

Heartsync/Nano-Banana

โœจ Core Features

โšก Lightning-Fast Editing
* **8-Step Generation**: Ultra-fast processing with Qwen-Image-Lightning LoRA
* **Real-time Editing**: 10x faster than conventional methods
* **GPU Optimization**: Maximized memory efficiency with xformers

๐Ÿค– AI Prompt Enhancement
* **Automatic Prompt Improvement**: Intelligent rewriting with Cerebras' Qwen3-235B model
* **Multilingual Support**: Auto-detection for Korean/Chinese/English
* **Context Understanding**: Sophisticated command generation aligned with image context

๐ŸŽฏ Versatile Editing Functions
โœ… Add/Delete/Replace objects
โœ… Text editing and style transformation
โœ… Person editing (expressions, hairstyles)
โœ… Vintage restoration and style conversion
โœ… Background replacement and enhancement

๐Ÿ› ๏ธ Tech Stack
* Base Model: Qwen-Image-Edit
* Acceleration: Qwen-Image-Lightning LoRA
* Prompt AI: Qwen3-235B (Cerebras)
* Framework: Gradio + Diffusers
* Optimization: bfloat16 precision

๐ŸŒŸ Why Open Nano-Banana?
* โšก Speed: Instant results with 8 steps
* ๐ŸŽจ Quality: Perfect editing with Prompt AI
* ๐Ÿ”’ Security: Token-based secure processing
* ๐Ÿ’œ Design: Beautiful gradient UI

๐Ÿท๏ธ Tags
#image-editing #ai-image-generation #qwen-image-edit #image-to-image #diffusers
#gradio #huggingface-spaces #lightning-lora #prompt-engineering #cerebras
#multilingual #real-time-editing #gpu-optimization #open-source #computer-vision
#deep-learning #machine-learning #artificial-intelligence #image-processing #creative-ai