search

meta-math / MetaMath

MetaMath: Bootstrap Your Own Mathematical Questions for Large Language Models
https://meta-math.github.io
Apache License 2.0
388 stars 35 forks source link

the few-shot in-context learning issues. #21

Open runzeer opened 11 months ago

runzeer commented 11 months ago

I just want to test the performance of the few-shot in-context learning capability. But I found an issue. I added the Instruction and response few-shot examples before the question and the generated result after the llm.generate function remains the same. And no matter how many examples I added, the inferenced results remain the same as the zero-shot result. So could you help me with the issues?

yulonghui commented 11 months ago

hi, runzeer I'm happy to help. You can share a repository with me so that I can investigate where the issue might be.