Hakim Sidahmed

large language model (2) synthetic datum (1) lora fine-tuning (1) client heterogeneity (1) generator-critic framework (1) dataset generation (1) conversational dataset (1) llm alignment (1) llm prompting (1) synthetic dataset generation (1) neutral point of view (1) parameter-efficient reinforcement learning (1) impartial generation (1) lora finetuning (1) model-agnostic meta learning (1) persona-based dialogue (1) persona-based conversation (1) federated reconstruction (1) partially local federated learning (1) collaborative filtering (1)