Finetuning the Model
Hello,
Do yo have any resources on how you finetuned this? or anything on how can i finetune a model for example on a large set of a software documentations? knowing i have a QA format dataset...
Thank you
Hi. I used Llama-factory.
https://github.com/hiyouga/LLaMA-Factory
There's a description and examples of how you can do fine-tuning. In what format you need a dataset, etc.
But you still need to know about machine learning. Unless you already have skills in this area. As it is not an easy task to select hyperparameters for LLM. I hope this information will help you.
Yes, I have a good understanding theorically. I come from vision world, i just dived into text. Still discovering tools etc.
Thank you for your help. Also, i want to know if you have an idea about finetuning on a documentation, the data isnt QA, and i cant make them in a QA format as its very very big... and also dont find the proper example.