11 часов назад
Researcher (AI)
Вакансия из Hirify Global, списка международных tech-компанийДля мэтча и отклика нужен Plus
Мэтч & Сопровод
Для мэтча с этой вакансией нужен Plus
Описание вакансии
Текст:
TL;DR
Researcher (AI) (image and video generation): Training and improving large-scale diffusion models on GPU clusters with an accent on distributed training, model quality, and reliable experimentation. Focus on optimizing parallel training strategies, debugging infrastructure and numerical issues, and designing data pipelines and evaluations for scalable model performance.
Location: On-site in San Francisco, United States
Company
Krea AI builds creative tools powered by controllable AI for text, images, video, sound, and 3D.
What you will do
- Train diffusion models for image and video generation on large GPU clusters.
- Optimize and profile distributed training across architectures, kernels, data loading, memory, and communication.
- Implement and improve distributed training strategies including FSDP, CP, SP, TP, and EP.
- Improve model quality and reliability through data, architecture, training pipelines, experiments, and evaluations.
- Debug distributed training failures, numerical errors, NCCL issues, and hardware components such as GPUs, NVLink, and InfiniBand.
- Run ablation studies across architectures, attention, optimizers, data, and algorithms to improve efficiency and performance.
Requirements
- Proven experience working with image or video models at scale.
- Strong proficiency in PyTorch and understanding of its internals.
- Strong knowledge of distributed training paradigms including FSDP, CP, SP, USP, TP, and EP.
- Experience profiling and debugging large distributed training runs and analyzing traces for bottlenecks.
- Knowledge of low-precision training and inference with FP8, NVFP4, and MXFP8.
- Understanding of diffusion model training across pretraining, midtraining, preference optimization, and reinforcement learning.
Nice to have
- Publications or open-source contributions in image or video modeling.
- Knowledge of LLM, VLM, representation learning, and robotics research.
Culture & Benefits
- Full-time, in-person work from the North Beach office in San Francisco.
- Goal-oriented research environment emphasizing independence, rapid iteration, and creative research directions.
- International visa sponsorship may be available, including STEM OPT, OPT, H-1B, O-1, and E-3.
- Competitive salary and equity, health, dental, and vision coverage, health FSA, long-term disability coverage, flexible PTO, and a 401(k) with a 4% company match.
- Office meals and covered Uber transit to and from the office.
Будьте осторожны: если работодатель просит войти в их систему, используя iCloud/Google, прислать код/пароль, запустить код/ПО, не делайте этого - это мошенники. Обязательно жмите "Пожаловаться" или пишите в поддержку. Подробнее в гайде →