3 ms·
Both are saying the same thing: in order for the base model that is phi to perform well as a chat agent, it would need to be tuned for that purpose before its b
by svnt 3y ago
Both are saying the same thing: in order for the base model that is phi to perform well as a chat agent, it would need to be tuned for that purpose before its benchmark results could have real-world value.
- imjonse 3y agoFrom this report. Phi-2 was not instruct tuned indeed. "Our models went through post-training with both supervised instruction fine-tuning, and preference tuning with DPO. We have worked on generating and curating various instruction and preference data. This has improved the model chat capabilities, robustness, as well as its safety."