Responsibilities
Team introduction: The ByteDance Beanbao Large Model Team was established in 2023. It is committed to developing the most advanced AI large model technology in the industry, becoming a world-class research team, and contributing to technological and social development. The Doubao Big Model team has long-term vision and determination in the field of AI. Its research directions cover NLP, CV, speech, etc., and it has laboratories and research positions in China, Singapore, the United States and other places. Relying on the platform's sufficient data, computing and other resources, the team continues to invest in related fields. It has launched a self-developed general large model to provide multi-modal capabilities, downstream supports 50+ businesses such as Doubao, Button, and Jimeng, and is open to corporate customers through the Volcano engine. At present, Doubao APP has become the AIGC application with the largest number of users in the Chinese market. 1. The team is responsible for the R&D and application of the company's large models, researching new applications and solutions of related technologies in the fields of search, recommendation, advertising, creation, dialogue and customer service, to meet users growing needs for intelligent interaction, and comprehensively improve the way users live and communicate in the future world the main work directions include: 1) Optimizing & innovating RLHF algorithm training efficiency and model generalization capabilities 2) Implementation and application of Long CoT technology 3) Posttraining algorithm for multi-modal large models (text, image, voice) 4) Construct high-quality, multi-domain data synthesis methods 5) Explore the application of LLM in scenarios such as emotional dialogue and creation.
Qualifications
1. Master degree or above, computer, communications, mathematics and other related majors (extra points for those who have mathematics and programming competitions) 2. Many years of NLP/deep learning R&D experience, at least 1 year of practical experience in large model applications 3. In-depth understanding of LLM technology stack (such as RLHF, Prompt Engineering, LoRA fine-tuning, etc.) 4. Familiar with Python/PyTorch, CUDA optimization, and core network architectures such as Transformer and MoE have solid coding capabilities (Python/C++) 5. Publish LLM-related papers at top conferences such as ACL/EMNLP/NeurIPS.
