OpenAI launched (*2*) a text-to-video-and-audio mannequin centered on bodily plausibility, multi-shot controllability, and synchronized dialogue/SFX. The OpenAI workforce has additionally launched a new invite-only Sora iOS app (U.S. and Canada first) that allows social creation, remixing, and consent-controlled “cameos” for inserting a verified likeness into generated scenes. Model capabilities Sora 2 claims materially higher world…
Local LLMs matured quick in 2025: open-weight households like Llama 3.1 (128K context size (ctx)), Qwen3 (Apache-2.0, dense + MoE), Gemma 2 (9B/27B, 8K ctx), Mixtral 8×7B (Apache-2.0 SMoE), and Phi-4-mini (3.8B, 128K ctx) now ship dependable specs and first-class native runners (GGUF/llama.cpp, LM Studio, Ollama), making on-prem and even laptop computer inference sensible should…
Alibaba’s Qwen group has simply launched FP8-quantized checkpoints for its new Qwen3-Next-80B-A3B fashions in two post-training variants—Instruct and Thinking—aimed toward high-throughput inference with ultra-long context and MoE effectivity. The FP8 repos mirror the BF16 releases however package deal “fine-grained FP8” weights (block measurement 128) and deployment notes for sglang and vLLM nightly builds. Benchmarks within…
In this tutorial, we introduce a Jailbreak Defense that we constructed step-by-step to detect and safely deal with policy-evasion prompts. We generate life like assault and benign examples, craft rule-based alerts, and mix these with TF-IDF options into a compact, interpretable classifier so we will catch evasive prompts with out blocking official requests. We exhibit…
IBM has launched Granite-Docling-258M, an open-source (Apache-2.0) vision-language mannequin designed particularly for end-to-end doc conversion. The mannequin targets layout-faithful extraction—tables, code, equations, lists, captions, and studying order—emitting a structured, machine-readable illustration relatively than lossy Markdown. It is obtainable on Hugging Face with a dwell demo and MLX construct for Apple Silicon. What’s new in comparison…
In this tutorial, we take a deep dive into the capabilities of Zarr, a library designed for environment friendly storage & manipulation of enormous, multidimensional arrays. We start by exploring the fundamentals, creating arrays, setting chunking methods, and modifying values immediately on disk. From there, we broaden into extra superior operations comparable to experimenting with…
How do you create 3D datasets to coach AI for Robotics with out costly conventional approaches? A group of researchers from NVIDIA launched “ViPE: Video Pose Engine for 3D Geometric Perception” bringing a key enchancment for Spatial AI. It addresses the central, agonizing bottleneck that has constrained the sector of 3D laptop imaginative and prescient…
Table of contents What architecture powers MobileLLM-R1? How efficient is the training? How does it perform against other open models? Where does MobileLLM-R1 fall short? How does MobileLLM-R1 compare to Qwen3, SmolLM2, and OLMo? Summary Meta has launched MobileLLM-R1, a household of light-weight edge reasoning fashions now obtainable on Hugging Face. The launch contains fashions…
IBM has quietly constructed a robust presence in the open-source AI ecosystem, and its newest launch exhibits why it shouldn’t be neglected. The firm has launched two new embedding fashions—granite-embedding-english-r2 and granite-embedding-small-english-r2—designed particularly for high-performance retrieval and RAG (retrieval-augmented technology) techniques. These fashions should not solely compact and environment friendly but additionally licensed underneath Apache…
BentoML has just lately launched llm-optimizer, an open-source framework designed to streamline the benchmarking and efficiency tuning of self-hosted giant language fashions (LLMs). The software addresses a standard problem in LLM deployment: discovering optimum configurations for latency, throughput, and price with out counting on guide trial-and-error. Why is tuning the LLM efficiency tough? Tuning LLM…