ML Staff Engineer - Scale Large Multimodal Models (Remote)

United States Digital Space LLC; United States; Remote; USD 140,000 - 200,000; Full time

Benefits offered by this job

Visa support (H1B/OPT transfers)

Remote-first culture

5 weeks paid leave

Healthcare including vision and dental

Job summary

United States Digital Space LLC is seeking a Member of Technical Staff on Machine Learning to contribute to the development cycle of our cutting-edge large deep learning models. You will prepare datasets, design architectures, implement solutions, and train and evaluate models to improve our products.

Collaborate closely with engineers and researchers to translate research into practical applications, joining us at an exciting moment as we build something from the ground up in a remote-first

Qualifications

• Experience with training and evaluating large DL models.
• Proficiency with PyTorch and/or JAX frameworks.
• Experience deploying ML models into software applications at scale.
• Ability to work in a fast-paced environment with some uncertainty.

Responsibilities

• Contribute to the development cycle of large deep learning models.
• Prepare datasets, design architectures, implement solutions, train and evaluate models.
• Collaborate with engineers and researchers to translate cutting-edge research into practical applications.
• Join us at an exciting moment and help build something from the ground up.

Skills

Training large DL models

Applying ML algorithms at scale

Collaborative teamwork

Tools

PyTorch

JAX

Job description

United States Digital Space LLC is seeking a Member of Technical Staff on Machine Learning to contribute to the development cycle of our cutting-edge large deep learning models. You will prepare datasets, design architectures, implement solutions, and train and evaluate models to improve our products.

Collaborate closely with engineers and researchers to translate research into practical applications, joining us at an exciting moment as we build something from the ground up in a remote-first