AI标注/评估兼职 (AI Annotation/Evaluation Part-time)
Responsible for multi-turn LLM dialogue annotation and quality review with a precise adherence to guidelines and consistency. Authored high-quality prompt-response pairs suitable for SFT fine-tuning applications. Maintained a perfect record of data compliance, accuracy, and no rework. • Completed over 600 annotation/review tasks with a 98.5% accuracy rate. • Authored 300+ prompt-response sets for instruction fine-tuning scenarios. • Utilized Labelbox and OpenTrain platform for annotation and evaluation tasks. • Supported quality assurance and compliance throughout the data lifecycle.