By leveraging large language and speech models, generative AI is creating intelligent 3D avatars that can engage users in natural conversation, from video games to customer service. To make these characters truly lifelike, they need human-like expressions. NVIDIA Audio2Face accelerates the creation of realistic digital characters by providing real-time facial animation and lip-sync driven by generative AI. Today, NVIDIA is open sourcing our Audio2Face technology to accelerate adoption of AI-powered avatars in games and 3D applications. https://developer.nvidia.com/blog/nvidia-open-sources-audio2face-animation-model/

Hi everyone! I’m really excited to share some good news. I’ve been selected to join an MBA program in Data Science starting this October! I can’t wait to begin this journey and wanted to celebrate this win with you all.

I have noticed a certain decline in content posting for the last few weeks and I am not sure if it's me who finds less interest on what is posted or it is a real thing here. I have been in similar places before, so I know there are ups and downs, and I feel we're on one. What do you think?

Where to add Hugging Face access token to a serverless RunPod endpoint?

"Large language models (LLMs) have demonstrated impressive capabilities across diverse tasks, yet their ability to perform structured symbolic planning remains limited, particularly in domains requiring formal representations like the Planning Domain Definition Language (PDDL). In this paper, we present a novel instruction tuning framework, PDDL-Instruct, designed to enhance LLMs' symbolic planning capabilities through logical chain-of-thought reasoning. Our approach focuses on teaching models to rigorously reason about action applicability, state transitions, and plan validity using explicit logical inference steps. By developing instruction prompts that guide models through the precise logical reasoning required to determine when actions can be applied in a given state, we enable LLMs to self-correct their planning processes through structured reflection. The framework systematically builds verification skills by decomposing the planning process into explicit reasoning chains about precondition satisfaction, effect application, and invariant preservation. Experimental results on multiple planning domains show that our chain-of-thought reasoning based instruction-tuned models are significantly better at planning, achieving planning accuracy of up to 94% on standard benchmarks, representing a 66% absolute improvement over baseline models. This work bridges the gap between the general reasoning capabilities of LLMs and the logical precision required for automated planning, offering a promising direction for developing better AI planning systems. https://arxiv.org/abs/2509.13351