Used Tools & Technologies
Not specified
Required Skills & Competences ?
Python @ 6 GitHub @ 6 Communication @ 6 Microservices @ 3 LLM @ 3 PyTorch @ 6 GPU @ 3Details
The NVIDIA Retriever Team is seeking an Applied Research Intern who will work on the next generation of retrieval pipelines for Retrieval-Augmented Generation (RAG), focusing on modalities beyond text. You will join experienced Research Scientists, ML and Software Engineers developing NVIDIA’s components for enterprise RAG applications, including embedding, ranking, object/text detection, OCR, and LLM-as-a-judge models or highly optimized containers.
At NVIDIA, the team builds the framework upon which production RAG systems are based. They have contributed to top research models in the text embedding space, topping the MTEB leaderboard and developed commercially viable versions for production use.
Responsibilities
- Work with researchers to fine-tune information retrieval models and develop pipelines for text, image, video, audio, and other modalities.
- Explore and craft datasets, design metrics, run experiments, and evaluate models to develop standard methodologies.
- Assist ML Engineers in bringing new Retrieval models to production as NVIDIA Inference Microservices (NIMs) or blueprints.
- Write blog posts, documentation, training materials, and potentially research papers to help customers understand the research.
- Stay updated with the latest developments in Retrieval across academia and industry.
Requirements
- Pursuing a PhD in Computer Science or other relevant technical fields.
- Excellent Python programming skills and strong understanding of the Python deep learning ecosystem, particularly PyTorch.
- Excellent knowledge of Deep Learning, including experience fine-tuning state-of-the-art Large Language Models and Computer Vision models.
- Strong communication skills to share ideas clearly via blog posts, papers, kernels, GitHub, etc.
Ways to Stand Out
- Strong research and publication record at top-tier conferences.
- Knowledge of multi-GPU and multi-node training.
- Prior background or academic publication in Retrieval research.
- Prior experience or publications in multimodal Large Language Models.
Benefits
- Hourly intern pay ranging from $30 to $90 depending on role, location, year in school, degree, and experience.
- Eligibility for intern benefits.
- NVIDIA is an equal opportunity employer committed to diversity and inclusion.