ComfyUI V1 ComfyUI V1 is a packaged desktop application that offers a closed beta experience. Key features include: • One-click install and auto-update for Windows, macOS, and Linux• Code-signed to ensure security•…
Llama-3.1-Nemotron-70B-Instruct-HF
New model released by Nvidia available here: https://huggingface.co/chat/ it’s said that it has better performance than 4o or Claude 3.5 Sonnet (https://github.com/lmarena/arena-hard-auto). Also available in ollama: And it counts strawberry R letter…
Pinokio, Ichigo
Interesting software for using and testing AI software locally: https://pinokio.computer/ I’ve not yet installed it, but the amount of systems it gathered together is impressive. From ComfyUI to Voice Cloning, and probably…
Spring Framework 7.0 announcement, Gradio 5
I have completely missed the announcement of Spring Framework 7.0 (https://spring.io/blog/2024/10/01/from-spring-framework-6-2-to-7-0) which will happen at the beginning of next year. The decision was made to stick with Java 17, though most comments…
Meta Movie Gen, OpenAI Canvas
Meta Movie Gen – it seems to be the biggest news for this weekend – there are already tons of sample creations. But also other things in movie gen world which happened…
AI News
News that caught my attention: Meta published Llama Stack: From GitHub: The Llama Stack defines and standardizes the building blocks needed to bring generative AI applications to market. These blocks span the…
Small package of AI News
HuggingFace crossed now 1.000.000 public models. Mystic v2 is out – on X there are plenty examples right now of that upscaler (4k and 8k images). Molmo was released, this is direct…
WIP: LLMs
Ollama is my preferred choice, but here I want to gather the alternatives I’ve found. mlx-lm Repo: https://github.com/ml-explore/mlx-examples/blob/main/llms/mlx_lm/README.md it’s part of MLX:MLX is an array framework for machine learning research on Apple…
Llama 3.2 released
New models are in versions 1B, 3B, 11B or 90B. The smallest ones are described as: Use our 1B or 3B models for on device applications such as summarizing a discussion from…
Moshi Foundation Model for Speech-Text Processing
If you’re looking for an alternative to the Whisper stack, one option worth considering is Moshi. I’ve originally found it mentioned on X: Summary: Key Features: Mimi Codec: Training and Evaluation: