Tag
ModelLens is a unified framework that recommends AI models for unseen datasets by learning from public leaderboard data, eliminating the need for costly direct evaluations. It constructs a performance-aware latent space to rank candidates across diverse tasks, outperforming existing baselines on large-scale benchmarks.
A novice asks for recommendations on small language models and prompting strategies to build an employee note summarization engine under 2000 tokens, after experiencing hallucinations with Qwen2.5-7B-Instruct.