About the role:
At MedARC we are passionate about utilizing open generative AI for medical applications. This also fits within the broader interest of Stability AI as a community and mission-driven, open AI company that cares deeply about real-world implications and applications.
We are looking for exceptional ML research engineers passionate about medical AI's potential and open importance. We especially value creative and unconventional approaches to research, and we care only about what you have built, not your credentials. You will have access to state-of-the-art high-performance computing resources, and you will work alongside top researchers in the field (both within MedARC and through collaborations with other institutions) to work on applying generative AI to the impactful field of medicine.
Responsibilities:
- Leading efforts to pretrain and finetune LLMs for medical and clinical data
- Working with the MedARC research team to develop the next generation of open medical LLMs.
- Writing papers and publishing at top conferences and scientific journals
Qualifications:
- 3+ years working on ML projects (training and fine tuning ML models, especially for research projects)
- 1+ years working on pretraining and finetuning of state-of-the-art LLMs (>7b parameter models)
- Experience with PyTorch and similar ML frameworks
- Ideally (but optionally) we are also looking for:
- Some familiarity with the medical NLP space
- Experience with distributed training of models
- Experience with the open ML ecosystem (HuggingFace, W&B, etc.)
- Experience with performance optimization / CUDA / JAX / TPUs a plus
- Communicate ML insights and results effectively: verbally, in writing and through visualization
- Self-motivated, well-organized (e.g. keep track of your own priorities and refine the plan according to new insights), self-reliant (able to debug complex issues with minimal supervision)
Equal Employment Opportunity:
We are an equal opportunity employer and do not discriminate on the basis of race, religion, national origin, gender, sexual orientation, age, veteran status, disability or other legally protected statuses.
