Fine tune model
#1
by
duchm
- opened
Hi,
How can I fine tune your model with my midi files?
Hello! This is a simple fine-tuned version of Llama-3.2-1B. I haven't modified the model architecture or tokenizer yet. The only thing that changed is the model's parameters.
So in theory, if you already have code that downloads Llama from Hugging Face and fine-tunes it with PyTorch on text data, you just need to replace the model name with mine. You will also need to convert your MIDI files into text.
I am currently training a better model with a new dataset. If it goes well, I will upload a new version. Thanks!