About the role#
The Hunyuan Multimodal Algorithm Researcher Intern will join the Technology Engineering Group to contribute to the development of Omni multimodal large models. This position focuses on the design, optimization, and research of foundational model architectures and training paradigms to advance multimodal understanding and generation capabilities.
What you'll do#
- Design and construct training data and foundational model algorithms.
- Optimize processes related to pre-training, Supervised Fine-Tuning (SFT), and Reinforcement Learning (RL).
- Analyze R&D challenges to identify performance bottlenecks and develop solutions using first-principles thinking.
- Explore next-generation model architectures and diverse paradigms for Omni-modal capabilities.
What you'll need#
- Bachelor’s degree in Computer Science, Artificial Intelligence, Mathematics, or related fields; graduate degrees are prioritized.
- Hands-on experience in large-scale multimodal data processing and high-quality data generation.
- A solid foundation in deep learning algorithms with practical experience in large model development.
- Proficiency in deep learning network implementation, model tuning for training and inference, CPU/GPU acceleration, and distributed training optimization.
- Participation in ACM or NOI competitions is highly valued. Familiarity with Diffusion Models, Autoregressive Models, or experience in cross-modal research is advantageous.
Location & details#
- Location: Palo Alto, California.
- Modality: On-site.
- Employment: Full-time internship.
- Compensation: $28.27 to $60.00 per hour.
- Benefits: Eligible for paid sick leave, 13 paid holidays, and enrollment in the company-sponsored medical plan.
