monjue

francisgroup

AI & ML interests

None yet

Recent Activity

reacted to openfree's post with ๐Ÿ”ฅ 3 days ago
๐ŸŒฑ Darwin-Qwen3-4B Released! We are excited to introduce the first achievement of the Darwin Project. We have evolved an embryonic model that simultaneously inherits the strengths of both non-reasoning models and reasoning models based on Qwen3-4B, creating a new offspring model. https://huggingface.co/openfree/Darwin-Qwen3-4B โœ… Benchmarking results confirm that the traits and advantages of both parent models are fully reflected. โœ… We are releasing Darwin-Qwen3-4B on Hugging Face. ๐Ÿ”ฌ Technical Specifications Model Configuration -Base Architecture: Qwen3-4B -Parent Model 1: Qwen/Qwen3-4B-Instruct-2507 -Parent Model 2: Qwen/Qwen3-4B-Thinking-2507 -Merge Algorithm: Darwin A2AP Enhanced v3.2 -Optimization Method: Evolutionary Algorithm with Adaptive Fitness Evolution Process -Population Size: 20 individuals -Evolution Cycles: 5,000+ generations -Fitness Objectives: Multi-objective optimization (accuracy, robustness, generalization) -Validation Metric: 88.56% on proxy task (digit classification) ๐Ÿ“Š Technical Innovations 1. Automated Model Merging Automatic exploration of optimal merge ratios without manual hyperparameter tuning Architecture-aware parameter alignment Late-stage performance optimization 2. Cost Efficiency Creation of new models at 1/10,000 the cost of foundation model training Repeatable experimental process ๐Ÿ”ฎ Significance and Outlook Darwin-Qwen3-4B demonstrates a new model creation paradigm: "Foundation A + Foundation B = Foundation Aร—B+C". This represents: -Instant Domain Fusion: Immediate integration of models with different characteristics -Sustainable AI Development: Reduced environmental burden through existing model recycling -Democratized AI Research: Small research teams can create new models ๐Ÿ“š Additional Resources GitHub Repository: [Coming Soon] Technical Paper: [In Preparation]
reacted to fantaxy's post with ๐Ÿ”ฅ 2 months ago
# ๐Ÿš€ AGI Turing Test Leaderboard: Evaluating AI's Novel Writing Abilities! ## ๐Ÿค” Can Machines Truly Create? In 2025, we've reached an era where we question whether AI can transcend being mere tools to become genuine creators. The **AGI Turing Test Leaderboard** is a revolutionary evaluation system that measures AI's true intelligence through long-form fiction writing! https://huggingface.co/spaces/fantaxy/AGI-LEADERBOARD ## ๐Ÿ“š Why Novel Writing? Novel writing is humanity's most complex cognitive task. Maintaining coherent worldbuilding across tens of thousands of words, creating multidimensional characters, crafting narratives that move the human heartโ€”all demanding a **symphony of cognitive abilities**! ๐ŸŽญ ## ๐Ÿ† 10-Tier Literary Evaluation System **10.0** โœจ Perfect Literary Achievement (Theoretical Apex) **9.1** ๐Ÿ… Nobel Prize Level (*Beloved*, *Never Let Me Go*) **8.1** ๐Ÿ“– Timeless Classic (*1984*, *Pride and Prejudice*) **7.1** ๐ŸŒ Global Bestseller (*Harry Potter*, *The Da Vinci Code*) **6.1** ๐ŸŽ–๏ธ International Literary Award (*The Handmaid's Tale*, *Underground Railroad*) **5.1** โœ๏ธ Professional Writer Level (Academy Award Screenplays) **0.0** โŒ Plagiarism/Human Work (Disqualified) ## ๐ŸŽฏ Current Record: 6.5 Points! The recently submitted work "**Commotion as a meteorite crashes through the roof**" achieved **6.5 points**! ### ๐Ÿ“Š Evaluation Breakdown - **Base Score**: 6.1 points (International Literary Award level) - **Volume Bonus**: +0.4 points (9,000 words) - **Final Score**: 6.5/10 points This work demonstrates quality comparable to major international literary award winners like the Booker Prize or Nebula Award! ## ๐Ÿค Join This Historic Experiment! The current record stands at 6.5 points. **Can your AI surpass this?** This platform is more than an evaluation tool. It's a grand experiment in understanding creativity itself! We welcome submissions at all levels. https://huggingface.co/blog/fantaxy/agi-leaderboard
reacted to ginipick's post with ๐Ÿš€ 2 months ago
๐ŸŽจ Flux Styler - AI Art Style Transfer ๐Ÿ“ Project Overview Flux Styler is a cutting-edge AI web application that transforms ordinary images into stunning artworks using FLUX.1-Kontext-dev model with 22 professional style LoRAs. https://huggingface.co/spaces/ginigen/Flux-Kontext-Style โœจ Key Features ๐Ÿ–ผ๏ธ One-Click Style Selection: Simply click thumbnails to apply styles instantly ๐ŸŽฏ 22 Premium Art Styles: From Ghibli to Van Gogh, Pixel Art to LEGO โšก High-Speed GPU Processing: Generate 1024x1024 images in 30-60 seconds ๐ŸŽฎ Intuitive Interface: No complex settings - just upload and transform! ๐ŸŽจ Style Categories ๐ŸŒธ Anime & Cartoon Ghibli | American Cartoon | JoJo | Snoopy | Rick & Morty ๐ŸŽช 3D & Geometric 3D Chibi | Low Poly | LEGO | Clay Toy ๐Ÿ–Œ๏ธ Traditional Art Chinese Ink | Oil Painting | Van Gogh | Picasso | Pop Art ๐Ÿงต Craft & Material Fabric | Origami | Paper Cutting | Macaron ๐Ÿ’ป Digital Art Pixel Art | Line Art | Vector ๐Ÿš€ How to Use Upload your image (or use default) ๐Ÿ“ค Click any style thumbnail ๐Ÿ–ฑ๏ธ (Optional) Add custom instructions โœ๏ธ Hit "Transform Image" ๐ŸŽจ Download your masterpiece! ๐Ÿ’พ ๐Ÿ› ๏ธ Tech Stack Model: FLUX.1-Kontext-dev by Black Forest Labs Style LoRAs: Owen777/Kontext-Style-Loras Framework: Gradio + Diffusers Acceleration: CUDA GPU (12GB+ VRAM) ๐Ÿ’ก Use Cases ๐Ÿ“ฑ Social Media Content Creation ๐Ÿ–ผ๏ธ NFT Art Generation ๐ŸŽ Personalized Gift Design ๐Ÿ“š Educational Visual Materials ๐ŸŽฏ Brand Marketing Assets
View all activity

Organizations

OpenFree_AI's profile picture