BTW, this cake is real, not AI-generated.
CREATIVE AI
-
AI-Generated 30-Minute Show Predicted Within Next Year
By
–
First compelling AI 30 minute show will be no later than next year. Movie length in 2026.
-
Google wins voice AI race, OpenAI redesigns coding benchmark
By
–
Top stories in AI today: -Google beats OpenAI in voice mode race
-OpenAI redesigns coding benchmark
-Bring images to life with Kling AI
-Become a tennis pro with AI
-6 new AI tools & 4 new AI jobs Read more: http://
therundown.ai/p/google-wins-
race-voice-ai
… -
Google announces Imagen 3 latent diffusion model for text-to-image generation
By
–
Google announces Imagen 3 discuss: https://
huggingface.co/papers/2408.07
009
… We introduce Imagen 3, a latent diffusion model that generates high quality images from text prompts. We describe our quality and responsibility evaluations. Imagen 3 is preferred over other state-of-the-art (SOTA) -
Fine-tuning AI model optimization exploring style and facial patterns
By
–
Otro ejemplo con diferentes steps (500, 1000, 1500, 4000) Parece que en entre 1000-1500 steps hay un buen equilibrio entre estilo y patrones de mi cara. Así que voy a seguir explorando con este modelo.
-
OpenAI announces GPT-4o system card and new AI initiatives
By
–
Someone pressed publish on too many new (and test?) articles too early in the last hours – expect GPT-4o system card, SWE-bench Verified, new customer stories, and more soon – Collaborating with The Met Museum to Awaken Sleeping Beauties with AI (Aug 9, 2024)
– Upwork puts AI to -
Flux AI Model Training with Multiple Artistic Style LoRAs
By
–
Vaya bestia es Flux. En breves termina de entrenarse el segundo Lora y probamos con más estilos artísticos.
-
Image Quality Improvement at 3500 Token Threshold
By
–
En mi caso fue así. Pero el salto a decente noté que fue sobre los 3500. De repente todas mis imágenes salía bien la cara, pero eso sí, los prompts con estilos de ilustración se convirtieron en fotos.
