这个模型比较倔强

[微擎团队]

[David]

We are currently primarily focused on code generation capabilities. Although we used some dialogue data during the fine-tuning phase, there are still clear limitations in our Chinese and English dialogue capabilities. We will enhance this aspect in the near future. Stay tuned :)

[jungyin]

兄弟，你这个是怎么调的，为什么我这个在发疯？

[微擎团队]

兄弟，你这个是怎么调的，为什么我这个在发疯？

你是用的gpu吗？我就是按照文档来的，不过这个模型确实还在初期阶段，会有一些未知问题