deepseek-vl系列的微调支持 (finetune)

[Jintao-Huang]

ms-swift大模型训练框架已经支持了deepseek-vl系列模型的推理和微调～

最佳实践可以查看这里:
https://github.com/modelscope/swift/blob/main/docs/source/Multi-Modal/deepseek-vl%E6%9C%80%E4%BD%B3%E5%AE%9E%E8%B7%B5.md

[RERV]

Thank you for supporting DeepSeek-VL!

[soloice]

@Jintao-Huang Can you kindly confirm if swift can be used to finetune visual encoder? If so, how? If not, what's the simplest way to support it?

[xs818818]

还有就是lora微调后怎么部署使用

[SinanAkkoyun]

@soloice Thank you very much for asking

[Jintao-Huang]

LoRA fine-tuning and merge-LoRA have been supported for both the visual encoder and aligner.
Full parameter fine-tuning is also supported.
😊

[SinanAkkoyun]

I am super grateful for your work, thank you a lot!!! ❤️