onsite
LLM Training Operation (Language and Creative), Specialist
LLM Training Operation (Language and Creative), Specialist
The LLM Training Operation Specialist will be a core member of ByteDance's LLM Global Data Team, gaining hands-on experience in training Large Language Models with diverse data sets. This role involves direct data production, developing specialized skills in data and content domains, and delivering high-quality annotation. The specialist will also provide mentorship and guidance while managing annotation guidelines.
About the role
About The Team
As a core member of our LLM Global Data Team, you'll be at the heart of our operations. Gain first-hand experience in understanding the intricacies of training Large Language Models (LLMs) with diverse data sets. Through our carefully designed rotation program, you'll witness how high-quality data is meticulously crafted and used. After completing your rotation, you will have the chance to select your area of expertise.
Your Role Will Involve
- Gaining direct experience in data production, including drafting, revising, and quality-checking, to understand the synergy between humans and data in LLM training.
- Developing specialized skills in one of our data and content domains (including but not limited to language, reasoning and safety * SFT/RM).
- Delivering high-quality annotation independently with limited supervision.
- Checking annotation quality and providing mentorship and guidance to internal colleagues and external partners on specific topics following successful incubation.
- Drafting and managing annotation guidelines and casebooks for specific annotation projects and tasks.
Minimum Qualifications
- Excellent proficiency in English with outstanding writing and critique skills.
- Comfortable working in a project-based fast-paced environment, where quality standards iterate quickly with research and experiment progress.
Preferred Qualifications
- At least 1 working year of experience in journalism, teaching, writing, text editing is highly valued. Experience as RLHF annotators for leading AI/LLM companies is strongly preferred.
- Candidates should be self-driven, intelligent, diligent, and skilled in providing mentorship and guidance to new members of the team.
- A deep interest in LLMs, human behavior and user experience is essential. The ideal candidate is an enthusiastic learner who finds engagement with diverse case studies and annotators stimulating.