LLMs and Guardrails
The discussion highlights the varying sizes of LLMs, emphasizing that smaller models can be highly effective for specific tasks like classification and retrieval. It also reveals that using LLMs for guardrails isn't technically feasible due to delays, leading the team to pivot towards fine-tuning BERT models instead. A significant release in this area is noted, with plans to invite a contributor for further insights.In this clip
From this podcast
ThursdAI
🎄ThursdAI - Dec19 - o1 vs gemini reasoning, VEO vs SORA, and holiday season full of AI surprises
Related Questions