Testing

SORA from OpenAI

I will be uploading the learning of SORA-like models in this post

---

Bit of Intruction

Sora is a text-to-video model that was released by OpenAI on 15th Feb 2024. As other products from OpenAI, it earns the respect from industrial and acdemical aspects due to its outstanding performance of text-to-video generating.

Authors

| Authors | Background | Experience | | ————- |:————-:| —–:| | Tim Brookes | right-aligned | $1600 | | col 2 is | centered | $12 | | zebra stripes | are neat | $1 |

Authors	Role @ OpenAI	Background	Prior Experience	More Info
Tim Brookes	Leader of Sora:Developing large-generative-model for simulating the real world	PhD @ UC Berkeley	① Focused on Video generation & AI powered Pixel phone camera @ NVDIA/Google/Meta. ② Primary researcher of DALL·3. ③ Published structPix2Pix during PhD	① Won National Geographic photo competition. ② Performed at Beacon Theatre(New York Boardway).
Bill Peebles	Co-leader of Sora :Focused on video generation and real-world simulation technology development	PhD @ UC Berkeley	① Interned at FAIR/Adobe Research/Nvidia. ② Published Scalable Diffusion Models with Transformers with Prof Saining Xie, established the fundamental framework of Sora.	Hard-working guy.
Connor Holmes	Focused on system effiency of model training	BSc EEE and PhD High Performance COmputing @ Colorado School of Mines	① Contributor of Sora/Dall·E. ② Previous Microsoft employee focuseing on LLM/BERT/RNNs/UNets.	Swimming: Life Guard and Assistant Swim Coach.
Will DePue	Developer	BCs Computer Science	① Software engineer at Commsor and InternetActivism ② Technical at OpenAI: worked on Applied research projects. Jailbreak and prompt injection mitigations, custom models, model capability evaluations, API finetuning, etc.	Entrepreneurship: Founder&CEO of DeepResearch during High School.

Model Description

SORA is a model that deployed diffusion transformer

Contributor