Research Engineer / Research Scientist (Post-training)

Ideogram•Toronto, ON

1d•Onsite

About The Position

In this role, you will develop the post-training pipeline for our text-to-image foundation models end to end, from data strategy to deployment, advancing techniques such as RLHF, RLAIF, and work on personalization/customization. You will contribute to post-training research that drives measurable gains, and implement and maintain high-throughput fine-tune/eval pipelines. You'll work with a creative and ambitious team of engineers and researchers who are building the future of the creative economy.

Requirements

5+ years of experience in developing machine learning models in JAX, PyTorch, or TensorFlow.
Experience in implementing Machine Learning foundations (e.g., Transformer, VAE, Denoising Diffusion models) from scratch.
Track record in machine learning innovation and familiarity with Deep Learning and advanced Machine Learning.
End-to-end understanding of generative media applications and excitement for pushing the state-of-the-art in generative AI.
Ability to debug machine learning models to iteratively improve model quality and performance.

Nice To Haves

Familiarity with kubernetes and docker.
Experience in low-level machine learning optimization, e.g., writing CUDA kernel code.

Responsibilities

Develop the post-training pipeline for text-to-image foundation models end to end, from data strategy to deployment.
Advance techniques such as RLHF and RLAIF.
Work on personalization/customization of models.
Contribute to post-training research that drives measurable gains.
Implement and maintain high-throughput fine-tune/eval pipelines.
Work with a creative and ambitious team of engineers and researchers.