Video & Creative AI
AI Model Trained to Paint Watercolours
Hugging Face Bloghigh confidence
Why it matters
- →Demonstrates AI's potential in creative fields beyond traditional applications.
- →Opens new avenues for exploring human-AI collaboration in art.
- →Provides an open-source framework for further experimentation and development.
A language model has been trained to paint watercolours using JavaScript and the p5.brush library, capturing the attention of over 1.5 million viewers. The project uses reinforcement learning to teach the model aesthetic preferences, allowing it to create unique, handmade-style paintings. This approach contrasts with the typical perfection of AI-generated images, offering a fresh perspective on AI's creative potential. The open-source nature of the project invites further exploration and replication by others interested in the intersection of AI and art.
Read originalMore from Hugging Face Blog
© Hugging Face BlogModels & Labsmodels
NeoMME: Efficient Multimodal and Multilingual Encoder Released
Hugging Face has unveiled NeoMME, a new family of multilingual multimodal encoders that integrate text and image processing into a single bidirectional Transformer. Unlike traditional models that rely on separate pretrained vision towers, NeoMME processes both text tokens and image patches simultaneously, enhancing efficiency and reducing computational overhead. The model's innovative architecture allows for dynamic image resolution and long bidirectional context, making it highly adaptable for various applications. With its release, NeoMME offers a streamlined approach to visual document retrieval, outperforming existing models in efficiency and storage requirements.
Hugging Face Blog
© Hugging Face BlogResearchresearch
Fine-tuning 350M Model for Structured Outputs
Hugging Face has demonstrated how fine-tuning a 350M model can significantly enhance its ability to produce structured outputs, a crucial task for many real-world applications. By using a targeted fine-tuning approach with a LoRA adapter and specific reward functions, the model's performance on the IFStruct benchmark improved, achieving a 22.6% pass rate. This approach shows that smaller models can be optimized to match the performance of larger models in specific tasks, making them more viable for integration into downstream systems. The process is accessible, with the fine-tuning runnable on a free-tier GPU, making it a practical option for developers looking to enhance model performance without extensive resources.
Hugging Face Blog
© Hugging Face BlogCoding Toolscoding
Hugging Face Introduces Funes for Agent Memory
Hugging Face has launched Funes, a tool that provides a durable memory layer for coding agents like Claude Code and Codex. This innovation allows agents to index, retrieve, and rank session logs, enabling them to recall past decisions and rationales during their workflow. Funes operates locally, ensuring privacy and control, and can also sync with a private Hugging Face dataset for broader accessibility. This development enhances the functionality of coding agents by allowing them to seamlessly integrate past knowledge into current tasks, making them more efficient and context-aware.
Hugging Face Blog
More in Video & Creative AI
© The Verge AIVideo & Creative AImusic
Roland Enters AI Music with Melody Flip
Roland's new Melody Flip tool marks its entry into the generative AI music space, offering a plugin for digital audio workstations that generates musical ideas rather than complete tracks. With around 250 themed 'Palettes', users can create melodies, chord progressions, basslines, or drums, either from scratch or by building on a reference track. Unlike some competitors, Melody Flip focuses on providing creative sparks rather than polished songs, with outputs that require further development in a DAW. This move reflects Roland's attempt to innovate in the AI music domain, though it may not win over all music enthusiasts.
The Verge AI
© The Verge AIVideo & Creative AIimage
AI Struggles with Realistic Food Imagery
AI-generated food images often appear unsettling due to the technical limitations of diffusion models, which struggle with creating thin, continuous structures like noodles and tendrils. These models start with noise and refine images, but can misinterpret the basic structure, leading to bizarre and unappetizing results. The lack of understanding of real-world objects means AI can only mimic appearances without grasping their context or purpose. This results in food images that trigger human disgust, as they often resemble non-food textures or contain unsettling patterns.
The Verge AI
© TechCrunch AIVideo & Creative AIimage
AI Menus Face Criticism for Uncanny Aesthetics
AI-generated menus in restaurants are being criticized for their unnaturally perfect food illustrations, which often appear unsettling to customers. This issue arises from AI models trained on datasets that emphasize a pleasing aesthetic, resulting in homogenized and unrealistic images. The problem is exacerbated when AI-generated content is used in further training, leading to a degradation of quality known as 'convergence.' This situation highlights the broader implications of AI in content creation, where the balance between realism and aesthetic appeal remains a challenge. As AI continues to influence visual content, the discomfort with these images points to the need for more nuanced training data and model development.
TechCrunch AI