Description

In this role, you will focus on enhancing image and video generation by developing control mechanisms and training advanced models. Your work will involve analyzing design choices and collaborating with the team to ensure high-quality outputs.

Requirements

• •Experience training large scale Diffusion models for image and video data
• •Finetuning Diffusion models for image and video applications
• •Deep understanding of evaluating image and video generative models
• •Strong proficiency in PyTorch and transformer models
• •Deep understanding of training techniques such as FSDP and model parallelism
• •Experience with writing forward and backward Triton kernels
• •Profiling, debugging, and optimizing multi-GPU operations

Tasks

• •Develop precise control mechanisms for image and video generation models
• •Train large-scale diffusion models with advanced control features
• •Create conditioning mechanisms for production requirements
• •Analyze design choices for applied controls
• •Communicate results and decisions with the team
• •Evaluate speed and quality trade-offs of control architectures

Tools & Technologies