Fine-Tuning with LoRA: LoRA image classification

shreydan / computer-vision-course

This repo is the homebase of a community driven course on Computer Vision with Neural Networks. Feel free to join us on the Hugging Face discord: hf.co/join/discord

MIT License

0 stars 1 forks source link

Fine-Tuning with LoRA: LoRA image classification #4

Closed asusevski closed 11 months ago

asusevski commented 11 months ago

Setup:

model: google/vit-base-patch16-224: Same reason as Shrey shared, most well known model
dataset: oxford-iiit pets dataset: Can compare easier with Shrey's results
Will rely on HF PEFT library as I think this will provide the best learning experience

I will include showing how long it took to train on the setup and compare with Shrey's results with Transfer learning

Let me know your guys' thoughts!

alanahmet commented 11 months ago

Hi @asusevski I think there might be a mistake. Right now, I'm working on fine-tuning the LoRA chapter. You can see the distribution of work from this link: https://github.com/users/sezan92/projects/30/views/1

asusevski commented 11 months ago

Got it, removing this issue. Will take on KD instead.