Further finetuneing

[rezzie-rich]

If i finetune an unsloth model and push it to huggingface, can i further finetune the trained model later using unsloth? If so, then how since the examples only load the models using 'fastlanguagemodel'. Can you please give an example?

Also, unsloth seems to be the best tool available for training using the least memory. It would be great if you guys made an Udemy course for unsloth. I'm confident it will be greatly appreciated by the community.

[danielhanchen]

Interesting idea on Udemy!! You're looking for https://github.com/unslothai/unsloth/wiki#loading-lora-adapters-for-continued-finetuning :)

[rezzie-rich]

Thank you, but it's still kinda confusing 😭.

Having a detailed tutorial, hence the course would be great for the newcomers.

[danielhanchen]

Oh so sorry - hmm we might make some Youtube videos :)

[rezzie-rich]

Will be much appreciated by many 🫶 lol