Data Labelling
AppenTextPrompt Response Writing SFT
I worked on Project Fireweed as a Hindi prompt-response writer and evaluator. My role involved creating complex Hindi prompts, reviewing model-generated responses, selecting the best ones, and refining them for clarity and cultural relevance. I also rated localized Hindi content for quality and appropriateness.
This was part of an RLHF (Reinforcement Learning from Human Feedback) process—my judgments and edits were used to fine-tune AI models and improve their performance in Hindi.
I worked on Project Fireweed as a Hindi prompt-response writer and evaluator. My role involved creating complex Hindi prompts, reviewing model-generated responses, selecting the best ones, and refining them for clarity and cultural relevance. I also rated localized Hindi content for quality and appropriateness. This was part of an RLHF (Reinforcement Learning from Human Feedback) process—my judgments and edits were used to fine-tune AI models and improve their performance in Hindi.
2024 - 2024