Fine tuning this model with Proprietary Code

#6
by vtraghu - opened

Hi, I am practically new to fine tuning of LLM models.
I would like to fine tune a code generator model with proprietary code mostly C++ for for embedded system . Can some one confirm

  1. Whether I can use Qwen/CodeQwen1.5-7B-Chat as a base a do PEFT ( LORA) with my custom data set
  2. Does this licencing term of the model allow me to use for commercial purpose.
  3. Any specific data format I have to follow while fine tuning?

Thanks

Qwen org
JustinLin610 changed discussion status to closed

@vtraghu , Did you make any progress in your tuning of Qwen/CodeQwen1.5-7B-Chat ?

Did you use https://qwen.readthedocs.io/en/latest/training/SFT/llama_factory.html ?

Thanks

Sign up or log in to comment