AI DATA ANNOTATOR
Text
Fine-tuningProject: Instruction Fine-Tuning Grading In this project, I worked on grading and evaluating AI-generated responses as part of an instruction fine-tuning pipeline. My role was to review prompts and the responses generated by AI models, then score and grade them based on how accurate, helpful, and well-structured the answers were. I followed strict quality guidelines to make sure my evaluations were consistent and unbiased across all tasks. The project involved carefully reading each instruction and comparing the AI response against the expected standard, then assigning grades based on criteria such as correctness, clarity, and relevance. I worked on a large number of tasks, maintaining a high level of attention to detail throughout. Quality measures I followed included sticking to the provided rubric, flagging responses that were unclear or incorrect, and ensuring my grades were fair and well-reasoned across every submission.