SORA Model Archives - gettectonic.com
SORA
24Apr

What is the SORA Model – In OpenAI’s Own Words

Introducing Sora: The Future of AI-Generated Video A New Frontier in AI Simulation We’re pioneering AI that understands and simulates physical reality—training models to help solve real-world interaction challenges. Meet Sora, our breakthrough text-to-video model that creates high-fidelity, minute-long videos from simple prompts while maintaining stunning visual quality. Sora in Action: Sample Creations 🏙️ “Neon Tokyo Stroll” A stylish woman in a black leather jacket walks confidently down a reflective Tokyo street, bathed in glowing neon signs. 🦣 “Prehistoric Giants” Woolly mammoths trek through snowy meadows, their fur rippling in the wind against mountain backdrops—captured in cinematic detail. 🚀 “Space Adventure Trailer” *A 30-year-old astronaut in a red wool helmet explores salt deserts under blue skies—shot on “35mm film” with vivid colors.* 🌊 “Big Sur’s Raw Beauty” A drone captures waves crashing against rugged cliffs at sunset, with a distant lighthouse completing this Pacific Coast masterpiece. 🕯️ “Curious Creature” *A fluffy monster kneels beside a melting candle, wide-eyed with wonder in a cozy 3D-rendered scene.* Why Sora Stands Out ✅ Complex scene generation with multiple characters✅ Physics-aware motion (though still improving)✅ Emotionally expressive characters✅ Multi-shot continuity within single videos Current Limitations ⚠️ Physics inaccuracies (e.g., objects not reacting realistically)⚠️ Spatial confusion (left/right or camera movement errors)⚠️ Spontaneous entity generation in crowded scenes Example: A bitten cookie might not show teeth marks; a basketball might “morph” unnaturally after swishing through a hoop. Safety First We’re implementing robust safeguards: Early Access & Future Vision Sora is now available to: This controlled rollout lets us refine Sora responsibly before broader release. Technical breakthrough: Sora uses diffusion transformers and DALL·E 3’s recaptioning tech to achieve unprecedented video coherence across variable durations/resolutions. This Is Just the Beginning Sora represents a critical step toward AI that understands physical reality—a key milestone on the path to AGI. Explore more examples and our technical report here. The future of video creation is being rewritten. Stay tuned. 🎥✨ More Sample Prompts & Outputs Like Related Posts Salesforce OEM AppExchange Expanding its reach beyond CRM, Salesforce.com has launched a new service called AppExchange OEM Edition, aimed at non-CRM service providers. Read more The Salesforce Story In Marc Benioff’s own words How did salesforce.com grow from a start up in a rented apartment into the world’s Read more Salesforce Jigsaw Salesforce.com, a prominent figure in cloud computing, has finalized a deal to acquire Jigsaw, a wiki-style business contact database, for Read more Service Cloud with AI-Driven Intelligence Salesforce Enhances Service Cloud with AI-Driven Intelligence Engine Data science and analytics are rapidly becoming standard features in enterprise applications, Read more

Read More
April 24, 2024in Data
03Mar

Train Your Own SORA Model

Unveiling the Vision Transformer: A Leap in Video Generation The closest open-source model to SORA is Latte, which uses the same Vision Transformer architecture. So, what makes the Vision Transformer so outstanding, and how does it differ from previous methods? You can Train Your Own SORA Model. Latte hasn’t open-sourced its text-to-video training code. We’ve replicated this code from the paper and made it available for anyone to use in training their own SORA alternative model. Let’s discuss how effective our training was. From 3D U-Net to Vision Transformer Image generation has advanced significantly, with the U-Net model structure being the most commonly used: If you’re confused about the network structures, remember the key principle of deep learning: “Just Add More Layers!” Vision Transformer: A Game Changer In 3D U-Net, the transformer can only function within the U-Net, limiting its view. The Vision Transformer, however, enables transformers to globally manage video generation. Training Your Open-Source SORA Alternative with Latte Latte uses the video slicing sequence and Vision Transformer method discussed. While Latte hasn’t open-sourced its text-to-video model training code, we’ve replicated it here: GitHub Repo. Training involves three steps: For more details, see the GitHub repo. They’ve also made improvements to the training process: Model Performance The official Latte video shows impressive performance, especially in handling significant motion. However, our own tests indicate that while Latte performs well, it isn’t the top-performing model. Other open-source models have shown better performance. We will continue to share information on models with better performance, so stay tuned to Tectonic’s Insights. Hardware Requirements Due to its large scale, training Latte requires an A100 or H100 with 80GB of memory. Like Related Posts Salesforce OEM AppExchange Expanding its reach beyond CRM, Salesforce.com has launched a new service called AppExchange OEM Edition, aimed at non-CRM service providers. Read more The Salesforce Story In Marc Benioff’s own words How did salesforce.com grow from a start up in a rented apartment into the world’s Read more Salesforce Jigsaw Salesforce.com, a prominent figure in cloud computing, has finalized a deal to acquire Jigsaw, a wiki-style business contact database, for Read more Service Cloud with AI-Driven Intelligence Salesforce Enhances Service Cloud with AI-Driven Intelligence Engine Data science and analytics are rapidly becoming standard features in enterprise applications, Read more

Read More
March 3, 2024in Salesforce

Subscribe to our mailing list. Join our mail list to receive our newsletter

We'll keep you up to date with our latest news, insights, free resources, and much more.
Don't worry, we won't spam you.
gettectonic.com