🎵 Prompt construction techniques The order of information you give your prompt should look like: 1️⃣ Core style/genre 2️⃣ Key musical elements 3️⃣ Mood and emotion 4️⃣ Specific details 5️⃣ Additional instructions Example prompt: “A cinematic outlaw country instrumental perfect for a long drive featuring a blues pedal steel guitar, a rustic mandolin, a fiddle playing call and response, a tape-driven rattly drum-kit, an autoharp, and a soaring accordion solo. The mood has southern soul, is raw, emotional, expansive, and full of the blues.”

In this week’s Stability Seconds, we’re showing you how to generate custom, commercially safe audio for your next video project with Stable Audio 2.5 🎵 Here’s how you can do it: ▶️ Find a short video clip that you want to add sound to. ▶️ Prompt Stable Audio 2.5 for genre, overall mood, and instruments that match your video. In our prompt, we used terms like “cinematic,” “awe-inspiring,” and “dramatic horn section.” ▶️ Generate a 3-minute track, then scrub through to find the best cut. ▶️ Add that audio to your clip to create a polished soundtrack. You can learn more and start using Stable Audio 2.5 here 👉 https://t.co/qBxRMynFbk

How are artists using AI to make music? 🎶 That’s what our Audio Research team set out to understand when they analyzed 337 musical works. The research examines how artists are using AI tools in their creative and production process. We’re sharing the key findings, with insight on emerging AI music creation practices that artists and creative professionals can draw inspiration from. You can read the full paper and learn more here 👉 https://t.co/aeA4RnudsL

This week’s Stability Seconds shows how you can quickly iterate across multiple musical directions with Stable Audio 2.5 ⏩ Starting with one short video clip, we applied three different prompts to shift the genre and mood of the video’s track in seconds: 1️⃣ Shoegaze synth: Dreamy and hazy 2️⃣ Ethereal classical: Airy and uplifting 3️⃣ Experimental drone: Textured and cinematic You can learn more and start using Stable Audio 2.5 here 👉 https://t.co/uBiALIF44n

We joined our friends at @ComfyUI for a live chat about what's new with Stable Audio 2.5 🥁 Check out CJ Carr from our Audio Research team demoing the model’s inpainting capability. You can watch the full episode here for more insights and examples 👉 https://t.co/430lf6kve3 https://t.co/2A0KHZieQB

Today we announced that we’ve formed a strategic partnership with @EA to co-develop transformative generative AI models, tools, and workflows that empower EA’s artists, designers, and developers to reimagine how games are made. You can learn more about our partnership here 👉 https://t.co/bukowa6gp2

Today we announced a strategic alliance with @UMG to co-develop professional AI music creation tools, powered by responsibly trained generative AI and built to support the creative process of artists, producers and songwriters globally. You can learn more about our collaboration here 👉 https://t.co/CnpNDsehUZ

Today we announced a partnership with @warnermusic to develop professional-grade tools that enable artists, songwriters, and producers to experiment, compose, and produce using ethically trained models. Our work together will unlock new forms of creative expression while protecting creators’ rights and opening new pathways for revenue. You can learn more about our partnership here 👉 https://t.co/zGsM9ZJrSD

KernelFalcon achieves 100% correctness across all 250 KernelBench L1–L3 tasks through a deep agent architecture that structures the problem instead of prompting harder. The system combines hierarchical task decomposition, deterministic orchestration, grounded execution, and parallel verification to generate GPU kernels that compile to PTX, execute on real hardware, and preserve PyTorch semantics. 💡Read our latest blog from @LaurawlyLaura and collaborators at Team PyTorch: https://t.co/AwH0dOFxET #PyTorch #KernelFalcon #AIInfrastructure #OpenSourceAI

As the #opensourceAI ecosystem matures, defining what “open” truly means has become increasingly complex. In this clip from today's webinar, IBM Research’s Mo McElaney explains why the community may need two complementary approaches: one grounded in open science for transparency and research, and another in #opensource for practical implementation and governance. 🔗 Watch the full recording: https://t.co/IfhQ0snKsY #PyTorch #vLLM

More than 3,000 developers, researchers, and innovators came together in San Francisco for #PyTorchCon 2025 for two days of keynotes, technical sessions, and community connection across the global AI ecosystem. The recap video captures the collaboration and energy that defined this year’s event. Watch the highlights below and explore all session recordings from the conference and co-located events on the PyTorch YouTube channel. 🖇️ PyTorch Conference Sessions: https://t.co/Fq70PzZg4t 📷 Open Agent Summit: https://t.co/4BHDCiXu8x 💡 Measuring Intelligence Summit: https://t.co/mlkcHIB3Ee 🎬 AI Infra Summit: https://t.co/ZDZOjcDlTG 🎥 Watch the recap video below Thank you to our Planning Committee, Program Chairs, Program Committee, and Sponsors for making PyTorch Conference 2025 possible. 🔗 More on the incredible people and organizations who made PyTorch Conference 2025 possible: https://t.co/9DpgU90m8c #PyTorchCon #PyTorch #OpenSourceAI #AIInfrastructure

+4 more

Today at AI Summit & Seoul Expo 2025, Matt White presents “From Code to Agency: Open Source and Standards for the Next Generation of AI Agents” from 2:00–2:45 PM (GMT +9) The session explores how open source, open standards, and new licensing models like OpenMDW address gaps in traditional software licenses across the AI lifecycle. It also highlights emerging protocols such as Model Context Protocol (MCP), Agent-to-Agent (A2A) communication, and the AGNTCY framework that enable interoperability, governance, and accountability for autonomous systems built on shared values. 🔗 Learn more: https://t.co/5wrvqrfpFy #PyTorch #OpenSourceAI #AIStandards #AIInfrastructure

In addition to PyTorch itself, the PyTorch Foundation hosts projects like vLLM, DeepSpeed, and Ray, a clear signal that PyTorch is becoming the foundational language of AI. For Luca Antiga, @LightningAI CTO and the first PyTorch Foundation TAC Chair, the PyTorch Conference was about connecting with the community driving PyTorch forward, including the millions of developers building with PyTorch Lightning. Part of what makes the event so remarkable is that you bump into people that are building on top of PyTorch to create amazing things. 🎥 View the video on YouTube: https://t.co/Cn8PLYq3RZ #PyTorchFoundation #PyTorchCon #PyTorch #OpenSourceAI #LightningAI

Our latest community blog explores how sparsity can unlock new efficiency for LLM inference in PyTorch. As model sizes and compute needs grow, sparsity offers a clear path to faster, more energy-efficient deployment. In this post, Kira Selby and Varun Khare (@AKAvkkhare), @NimbleEdgeInc, share techniques like CETT thresholding, Relufication, weight caching, and statistical top-k that enable up to 6x faster inference and advance a unified framework for sparse inference. 🔗 Read the blog: https://t.co/fzt5KLTK5x #PyTorch #AIInfrastructure #EdgeAI #OpenSource

Enable fast and scalable simulation of atomic systems for chemistry and materials science research by integrating PyTorch-based MLIPs into the LAMMPS MD package through the ML-IAP-Kokkos interface. Learn more about how developers can simplify the process of interfacing community models and scale MD simulations in NVIDIA’s developer blog: 🔗 https://t.co/Bw7d0QrdFO #PyTorch #OpenSourceAI #AI #Inference #Innovation

Our latest PyTorch Foundation Spotlight features PyTorch Ambassador Ankita Guha, whose work with neural networks and survival analysis led to finding PyTorch and the community behind it. In the video, Ankita talks about learning through experimentation, realizing the scale of the PyTorch community, and plans to help foster local communities in Michigan and beyond so people can learn and contribute together. 🎥 Watch the PyTorch Foundation Spotlight with Ankita Guha: https://t.co/GFBP8QxMOx #PyTorch #OpenSourceAI #AIInfrastructure #PyTorchAmbassador

Hear directly from Dwarak Rajagopal (Vice President of AI Engineering and Research, @Snowflake on why @Snowflake chose premier membership with the PyTorch Foundation and how open source collaboration shapes their approach to enterprise AI. In this video, Dwarak explains how PyTorch powers the Snowflake AI Research Team’s applied research, connects innovation to production, and enables Snowflake to contribute training and inference optimizations back to the community. “PyTorch plays a critical role in our research, bridging the gap between open innovation and enterprise AI,” said Rajagopal. Watch to learn how Snowflake is advancing open source AI and driving the future of secure, scalable AI research and deployment, and check out Snowflake's blog post to learn more about Snowflake's sessions at PyTorch Conference. 📹 https://t.co/FHavTdlP67 🔗 Snowflake's blog: https://t.co/k0jZ2JxsCj #PyTorch #AIInfrastructure #OpenSourceAI #AIBuilders

PINA has joined the PyTorch Ecosystem, bringing a unified framework for Scientific Machine Learning to researchers and engineers. PINA is an open source Python library built on PyTorch, PyTorch Lightning, and compatible with PyTorch Geometric. It streamlines SciML workflows for modelling physical systems, solving PDEs, building machine learning force fields, simulating dynamics, and more. Its modular structure spans problem definition, model design, solver selection, and training, offering both flexibility and reproducibility. 🔗 Learn more and explore how to get involved: https://t.co/IHgoqDXAD4 #PyTorch #OpenSourceAI #SciML #MachineLearning

Need to optimize applications with large models and stretched memory resources? Learn how to accelerate large-scale LLM inference and Cache Offload with CPU-GPU Memory Sharing in NVIDIA’s recent developer tech blog: 📎: https://t.co/eyx4DLi6ta #PyTorch #OpenSourceAI #AI #Inference #Innovation

AI innovation at AMD starts with solving challenges across the full stack, from the behavior of large foundation models to ensuring core operators run quickly and reliably at the hardware level. In this PyTorch Conference conversation, Niles Burbank (Director, Solutions Architecture at AMD) explains how AMD provides hardware and software that accelerate basic operations to support the vision and promise of larger models, and shares why the event brings together users, practitioners, and developers across the AI spectrum in a way few others do. 📎 https://t.co/igLYa2Er3j #PyTorch #PyTorchCon #AMD #AIInfrastructure

OpenReg introduces a self-contained simulator that provides a stable reference for integrating new accelerators with PyTorch through the PrivateUse1 DispatchKey. It functions as a test backend, an integration example, and a source of matching documentation. The blog covers the CPU-based simulator, CI integration, operator registration, runtime support, and cross-platform availability, along with planned work in distributed computing, ONNX support, and compiler enhancements. 🔗 Read our latest blog from Jiahao Chen (Huawei), Jiawei Li (Huawei), and Zesheng Zong (Huawei): https://t.co/OUd8c78tUf #PyTorch #AIInfrastructure #OpenSourceAI #PyTorchEcosystem

At NeurIPS next week? Join us at our Open Source AI Reception, an evening focused on open source collaboration hosted by @CloudNativeFdn and @PyTorch with @anyscalecompute, @FeatherlessAI, @huggingface, and @UnslothAI Join AI enthusiasts, developers, and researchers for an evening of networking and conversation outside NeurIPS 2025. Drinks and light bites provided. 🔗 Register to secure your spot: https://t.co/avDb5a1j1O Wednesday, December 3, 6:00–9:00 PM PT Union Kitchen and Tap Gaslamp, San Diego, California, USA #PyTorch #NeurIPS2025 #NeurIPS #OpenSourceAI

Torch Ignite is coming to Patra, Greece on 28 November 2025 for a 3-hour, hands-on introduction to PyTorch. This workshop is built for beginners who want practical experience with tensors, autograd, optimization, and training a first neural network. Participants will work directly in Google Colab, follow guided tutorials, and leave with runnable notebooks and a complete beginner-level model. Event: Torch Ignite: From Tensors to Your First Neural Net 📆 November 28, Patra, Greece 🇬🇷 🔗 Register: https://t.co/Jg4kjVmWAf #PyTorch #OpenSourceAI #AIInfrastructure #AIBuilders

📸 Scenes from the Future of Inferencing! PyTorch ATX, the @vllm_project community, and @RedHat brought together 90+ AI builders at Capital Factory back in September, to dive into the latest in LLM inference -> from quantization and PagedAttention to multi-node deployment. Amazing energy, collaboration, and innovation from Austin’s growing AI community. 💡💪 Read more about it: https://t.co/nkkRgWjZTg Featuring sessions by @KamaraiCode @wattsteve @luka_govedic @root_fs Greg Pereira #PyTorchATX #vLLM #RedHat #AICommunity #LLM #AustinTech #InferenceOptimization

Open source is not just about code. It's also about the collaboration and the trust. Our latest PyTorch Foundation Spotlight features PyTorch Ambassador @Zesheng_Zong who is working to make the community better and who recently received a 2025 PyTorch Contributor Award for Outstanding PyTorch Ambassador. 🎥 Watch the PyTorch Foundation Spotlight with Zesheng Zong on the PyTorch YouTube channel: https://t.co/7fMHKT5pYz #OpenSource #PyTorch #OpenSourceAI #PyTorchAmbassador

【経営企画メンバー募集！】 Sakana AIは、研究開発と事業開発の両輪で成長をとげ、日本発でありながら既にグローバルAI企業や投資家からも注目される存在となっています。この度、当社の成長をさらに進化・スケールさせるため、「AI技術 x 産業変革 x 資本戦略」の交差点で戦略を一緒に描いてくれる 経営企画メンバーを募集します！ CEO/COOと密に連携し、資本政策、M&A、戦略的パートナーシップなどを通じて、当社の次の成長を一緒に創り上げてくれる方をお待ちしています。 The next frontier of finance is not in Wall Street — it’s AI. Business Role (Corporate Development): https://t.co/Iik4DnIYAf

We are expanding Sakana AI’s corporate development team. If you have worked in corp dev, IBD at a global investment bank in Japan, or PE in Japan, please have a look. Business Role (Corporate Development): https://t.co/2VbQQ7FCzw

Cold weather makes me grumpy 😅😅😅 https://t.co/soo4ulfIG5 https://t.co/SowT5xL0lb

Excited to announce our MIT Press book “Neuroevolution: Harnessing Creativity in AI Agent Design” by Sebastian Risi (@risi1979), Yujin Tang (@yujin_tang), Risto Miikkulainen, and myself. We explore decades of work on evolving intelligent agents and shows how neuroevolution can drive creativity in deep learning, RL, LLMs and AI Agents! 📖 Free open-access edition: https://t.co/1VraVue7Sk In addition to our own works, this video features work by Jürgen Schmidhuber (@SchmidhuberAI), Seth Bling (@SethBling), Igor Karpov, Jacob Schrum, Yulu Gan (@yule_gan), Ken Stanley (@kenneth0stanley), Joel Lehman (@joelbot3000), Jeff Clune (@jeffclune), Nick Cheney (@CheneyLab), Richard Song (@XingyouSong), Chelsea Finn (@chelseabfinn), Julian Togelius (@togelius), Sam Earle (@Smearle_RH), Hod Lipson (@hodlipson), and Jean-Baptiste Mouret (@jb_mouret).

📖 We also made a free, open-access edition of the book: https://t.co/D4tiv3D7Bj https://t.co/XweGKeBfMS

Just realized I knew all about hedging ‘AI risks’ > 15 years ago 😅 https://t.co/VDbDfhVGVg

“In my view, AI is ultimately going to be a normal technology. In 20 years, our kids will just be using a chat bot like it’s a fax machine. It won’t be magical anymore. This will just be integrated and adapted into our collective system.” Gave my 2c at #BloombergNewEconomy Forum https://t.co/Ov9j6xYech